1082 files changed, 63174 insertions, 30075 deletions

diff --git a/drivers/gpu/drm/Kconfig b/drivers/gpu/drm/Kconfig
index e67c194c2aca..1168351267fd 100644
--- a/drivers/gpu/drm/Kconfig
+++ b/drivers/gpu/drm/Kconfig
@@ -93,6 +93,20 @@ config DRM_KMS_FB_HELPER
 	help
 	  FBDEV helpers for KMS drivers.
 
+config DRM_DEBUG_DP_MST_TOPOLOGY_REFS
+        bool "Enable refcount backtrace history in the DP MST helpers"
+        select STACKDEPOT
+        depends on DRM_KMS_HELPER
+        depends on DEBUG_KERNEL
+        depends on EXPERT
+        help
+          Enables debug tracing for topology refs in DRM's DP MST helpers. A
+          history of each topology reference/dereference will be printed to the
+          kernel log once a port or branch device's topology refcount reaches 0.
+
+          This has the potential to use a lot of memory and print some very
+          large kernel messages. If in doubt, say "N".
+
 config DRM_FBDEV_EMULATION
 	bool "Enable legacy fbdev support for your modesetting driver"
 	depends on DRM
@@ -165,13 +179,26 @@ config DRM_TTM
 	  GPU memory types. Will be enabled automatically if a device driver
 	  uses it.
 
+config DRM_TTM_DMA_PAGE_POOL
+	bool
+	depends on DRM_TTM && (SWIOTLB || INTEL_IOMMU)
+	default y
+	help
+	  Choose this if you need the TTM dma page pool
+
 config DRM_VRAM_HELPER
 	tristate
 	depends on DRM
-	select DRM_TTM
 	help
 	  Helpers for VRAM memory management
 
+config DRM_TTM_HELPER
+	tristate
+	depends on DRM
+	select DRM_TTM
+	help
+	  Helpers for ttm-based gem objects
+
 config DRM_GEM_CMA_HELPER
 	bool
 	depends on DRM
@@ -226,9 +253,9 @@ config DRM_AMDGPU
 	tristate "AMD GPU"
 	depends on DRM && PCI && MMU
 	select FW_LOADER
-        select DRM_KMS_HELPER
+	select DRM_KMS_HELPER
 	select DRM_SCHED
-        select DRM_TTM
+	select DRM_TTM
 	select POWER_SUPPLY
 	select HWMON
 	select BACKLIGHT_CLASS_DEVICE
@@ -257,6 +284,7 @@ config DRM_VKMS
 	tristate "Virtual KMS (EXPERIMENTAL)"
 	depends on DRM
 	select DRM_KMS_HELPER
+	select CRC32
 	default n
 	help
 	  Virtual Kernel Mode-Setting (VKMS) is used for testing or for
@@ -397,7 +425,7 @@ config DRM_R128
 
 config DRM_I810
 	tristate "Intel I810"
-	# !PREEMPT because of missing ioctl locking
+	# !PREEMPTION because of missing ioctl locking
 	depends on DRM && AGP && AGP_INTEL && (!PREEMPTION || BROKEN)
 	help
 	  Choose this option if you have an Intel I810 graphics card.  If M is
diff --git a/drivers/gpu/drm/Makefile b/drivers/gpu/drm/Makefile
index 82ff826b33cc..9f1c7c486f88 100644
--- a/drivers/gpu/drm/Makefile
+++ b/drivers/gpu/drm/Makefile
@@ -33,10 +33,12 @@ drm-$(CONFIG_DEBUG_FS) += drm_debugfs.o drm_debugfs_crc.o
 drm-$(CONFIG_DRM_LOAD_EDID_FIRMWARE) += drm_edid_load.o
 
 drm_vram_helper-y := drm_gem_vram_helper.o \
-		     drm_vram_helper_common.o \
-		     drm_vram_mm_helper.o
+		     drm_vram_helper_common.o
 obj-$(CONFIG_DRM_VRAM_HELPER) += drm_vram_helper.o
 
+drm_ttm_helper-y := drm_gem_ttm_helper.o
+obj-$(CONFIG_DRM_TTM_HELPER) += drm_ttm_helper.o
+
 drm_kms_helper-y := drm_crtc_helper.o drm_dp_helper.o drm_dsc.o drm_probe_helper.o \
 		drm_plane_helper.o drm_dp_mst_topology.o drm_atomic_helper.o \
 		drm_kms_helper_common.o drm_dp_dual_mode_helper.o \
diff --git a/drivers/gpu/drm/amd/amdgpu/Makefile b/drivers/gpu/drm/amd/amdgpu/Makefile
index 42e2c1f57152..ca0e435559d5 100644
--- a/drivers/gpu/drm/amd/amdgpu/Makefile
+++ b/drivers/gpu/drm/amd/amdgpu/Makefile
@@ -53,8 +53,9 @@ amdgpu-y += amdgpu_device.o amdgpu_kms.o \
 	amdgpu_ucode.o amdgpu_bo_list.o amdgpu_ctx.o amdgpu_sync.o \
 	amdgpu_gtt_mgr.o amdgpu_vram_mgr.o amdgpu_virt.o amdgpu_atomfirmware.o \
 	amdgpu_vf_error.o amdgpu_sched.o amdgpu_debugfs.o amdgpu_ids.o \
-	amdgpu_gmc.o amdgpu_xgmi.o amdgpu_csa.o amdgpu_ras.o amdgpu_vm_cpu.o \
-	amdgpu_vm_sdma.o amdgpu_pmu.o amdgpu_discovery.o amdgpu_ras_eeprom.o smu_v11_0_i2c.o
+	amdgpu_gmc.o amdgpu_mmhub.o amdgpu_xgmi.o amdgpu_csa.o amdgpu_ras.o amdgpu_vm_cpu.o \
+	amdgpu_vm_sdma.o amdgpu_discovery.o amdgpu_ras_eeprom.o amdgpu_nbio.o \
+	amdgpu_umc.o smu_v11_0_i2c.o
 
 amdgpu-$(CONFIG_PERF_EVENTS) += amdgpu_pmu.o
 
@@ -67,7 +68,7 @@ amdgpu-$(CONFIG_DRM_AMDGPU_SI)+= si.o gmc_v6_0.o gfx_v6_0.o si_ih.o si_dma.o dce
 amdgpu-y += \
 	vi.o mxgpu_vi.o nbio_v6_1.o soc15.o emu_soc.o mxgpu_ai.o nbio_v7_0.o vega10_reg_init.o \
 	vega20_reg_init.o nbio_v7_4.o nbio_v2_3.o nv.o navi10_reg_init.o navi14_reg_init.o \
-	arct_reg_init.o navi12_reg_init.o
+	arct_reg_init.o navi12_reg_init.o mxgpu_nv.o
 
 # add DF block
 amdgpu-y += \
@@ -83,7 +84,7 @@ amdgpu-y += \
 
 # add UMC block
 amdgpu-y += \
-	umc_v6_1.o
+	umc_v6_1.o umc_v6_0.o
 
 # add IH block
 amdgpu-y += \
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
index bd37df5dd6d0..0c229a92a24b 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
@@ -73,6 +73,7 @@
 #include "amdgpu_gmc.h"
 #include "amdgpu_gfx.h"
 #include "amdgpu_sdma.h"
+#include "amdgpu_nbio.h"
 #include "amdgpu_dm.h"
 #include "amdgpu_virt.h"
 #include "amdgpu_csa.h"
@@ -106,6 +107,8 @@ struct amdgpu_mgpu_info
 	uint32_t			num_apu;
 };
 
+#define AMDGPU_MAX_TIMEOUT_PARAM_LENGTH	256
+
 /*
  * Modules parameters.
  */
@@ -122,6 +125,7 @@ extern int amdgpu_disp_priority;
 extern int amdgpu_hw_i2c;
 extern int amdgpu_pcie_gen2;
 extern int amdgpu_msi;
+extern char amdgpu_lockup_timeout[AMDGPU_MAX_TIMEOUT_PARAM_LENGTH];
 extern int amdgpu_dpm;
 extern int amdgpu_fw_load_type;
 extern int amdgpu_aspm;
@@ -135,6 +139,7 @@ extern int amdgpu_vm_fragment_size;
 extern int amdgpu_vm_fault_stop;
 extern int amdgpu_vm_debug;
 extern int amdgpu_vm_update_mode;
+extern int amdgpu_exp_hw_support;
 extern int amdgpu_dc;
 extern int amdgpu_sched_jobs;
 extern int amdgpu_sched_hw_submission;
@@ -146,11 +151,7 @@ extern uint amdgpu_sdma_phase_quantum;
 extern char *amdgpu_disable_cu;
 extern char *amdgpu_virtual_display;
 extern uint amdgpu_pp_feature_mask;
-extern int amdgpu_ngg;
-extern int amdgpu_prim_buf_per_se;
-extern int amdgpu_pos_buf_per_se;
-extern int amdgpu_cntl_sb_buf_per_se;
-extern int amdgpu_param_buf_per_se;
+extern uint amdgpu_force_long_training;
 extern int amdgpu_job_hang_limit;
 extern int amdgpu_lbpw;
 extern int amdgpu_compute_multipipe;
@@ -167,6 +168,12 @@ extern int amdgpu_mcbp;
 extern int amdgpu_discovery;
 extern int amdgpu_mes;
 extern int amdgpu_noretry;
+extern int amdgpu_force_asic_type;
+#ifdef CONFIG_HSA_AMD
+extern int sched_policy;
+#else
+static const int sched_policy = KFD_SCHED_POLICY_HWS;
+#endif
 
 #ifdef CONFIG_DRM_AMDGPU_SI
 extern int amdgpu_si_support;
@@ -283,6 +290,9 @@ struct amdgpu_ip_block_version {
 	const struct amd_ip_funcs *funcs;
 };
 
+#define HW_REV(_Major, _Minor, _Rev) \
+	((((uint32_t) (_Major)) << 16) | ((uint32_t) (_Minor) << 8) | ((uint32_t) (_Rev)))
+
 struct amdgpu_ip_block {
 	struct amdgpu_ip_block_status status;
 	const struct amdgpu_ip_block_version *version;
@@ -425,7 +435,6 @@ struct amdgpu_fpriv {
 };
 
 int amdgpu_file_to_fpriv(struct file *filp, struct amdgpu_fpriv **fpriv);
-int amdgpu_device_get_job_timeout_settings(struct amdgpu_device *adev);
 
 int amdgpu_ib_get(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 		  unsigned size, struct amdgpu_ib *ib);
@@ -477,7 +486,6 @@ struct amdgpu_cs_parser {
 	uint64_t			bytes_moved_vis_threshold;
 	uint64_t			bytes_moved;
 	uint64_t			bytes_moved_vis;
-	struct amdgpu_bo_list_entry	*evictable;
 
 	/* user fence */
 	struct amdgpu_bo_list_entry	uf_entry;
@@ -624,6 +632,11 @@ struct amdgpu_fw_vram_usage {
 	u64 size;
 	struct amdgpu_bo *reserved_bo;
 	void *va;
+
+	/* Offset on the top of VRAM, used as c2p write buffer.
+	*/
+	u64 mem_train_fb_loc;
+	bool mem_train_support;
 };
 
 /*
@@ -644,71 +657,14 @@ typedef void (*amdgpu_wreg64_t)(struct amdgpu_device*, uint32_t, uint64_t);
 typedef uint32_t (*amdgpu_block_rreg_t)(struct amdgpu_device*, uint32_t, uint32_t);
 typedef void (*amdgpu_block_wreg_t)(struct amdgpu_device*, uint32_t, uint32_t, uint32_t);
 
-
-/*
- * amdgpu nbio functions
- *
- */
-struct nbio_hdp_flush_reg {
-	u32 ref_and_mask_cp0;
-	u32 ref_and_mask_cp1;
-	u32 ref_and_mask_cp2;
-	u32 ref_and_mask_cp3;
-	u32 ref_and_mask_cp4;
-	u32 ref_and_mask_cp5;
-	u32 ref_and_mask_cp6;
-	u32 ref_and_mask_cp7;
-	u32 ref_and_mask_cp8;
-	u32 ref_and_mask_cp9;
-	u32 ref_and_mask_sdma0;
-	u32 ref_and_mask_sdma1;
-	u32 ref_and_mask_sdma2;
-	u32 ref_and_mask_sdma3;
-	u32 ref_and_mask_sdma4;
-	u32 ref_and_mask_sdma5;
-	u32 ref_and_mask_sdma6;
-	u32 ref_and_mask_sdma7;
-};
-
 struct amdgpu_mmio_remap {
 	u32 reg_offset;
 	resource_size_t bus_addr;
 };
 
-struct amdgpu_nbio_funcs {
-	const struct nbio_hdp_flush_reg *hdp_flush_reg;
-	u32 (*get_hdp_flush_req_offset)(struct amdgpu_device *adev);
-	u32 (*get_hdp_flush_done_offset)(struct amdgpu_device *adev);
-	u32 (*get_pcie_index_offset)(struct amdgpu_device *adev);
-	u32 (*get_pcie_data_offset)(struct amdgpu_device *adev);
-	u32 (*get_rev_id)(struct amdgpu_device *adev);
-	void (*mc_access_enable)(struct amdgpu_device *adev, bool enable);
-	void (*hdp_flush)(struct amdgpu_device *adev, struct amdgpu_ring *ring);
-	u32 (*get_memsize)(struct amdgpu_device *adev);
-	void (*sdma_doorbell_range)(struct amdgpu_device *adev, int instance,
-			bool use_doorbell, int doorbell_index, int doorbell_size);
-	void (*vcn_doorbell_range)(struct amdgpu_device *adev, bool use_doorbell,
-				   int doorbell_index, int instance);
-	void (*enable_doorbell_aperture)(struct amdgpu_device *adev,
-					 bool enable);
-	void (*enable_doorbell_selfring_aperture)(struct amdgpu_device *adev,
-						  bool enable);
-	void (*ih_doorbell_range)(struct amdgpu_device *adev,
-				  bool use_doorbell, int doorbell_index);
-	void (*update_medium_grain_clock_gating)(struct amdgpu_device *adev,
-						 bool enable);
-	void (*update_medium_grain_light_sleep)(struct amdgpu_device *adev,
-						bool enable);
-	void (*get_clockgating_state)(struct amdgpu_device *adev,
-				      u32 *flags);
-	void (*ih_control)(struct amdgpu_device *adev);
-	void (*init_registers)(struct amdgpu_device *adev);
-	void (*detect_hw_virt)(struct amdgpu_device *adev);
-	void (*remap_hdp_registers)(struct amdgpu_device *adev);
-};
-
 struct amdgpu_df_funcs {
 	void (*sw_init)(struct amdgpu_device *adev);
+	void (*sw_fini)(struct amdgpu_device *adev);
 	void (*enable_broadcast_mode)(struct amdgpu_device *adev,
 				      bool enable);
 	u32 (*get_fb_channel_number)(struct amdgpu_device *adev);
@@ -813,6 +769,7 @@ struct amdgpu_device {
 	uint8_t				*bios;
 	uint32_t			bios_size;
 	struct amdgpu_bo		*stolen_vga_memory;
+	struct amdgpu_bo		*discovery_memory;
 	uint32_t			bios_scratch_reg_offset;
 	uint32_t			bios_scratch[AMDGPU_BIOS_NUM_SCRATCH];
 
@@ -921,6 +878,12 @@ struct amdgpu_device {
 	u32				cg_flags;
 	u32				pg_flags;
 
+	/* nbio */
+	struct amdgpu_nbio		nbio;
+
+	/* mmhub */
+	struct amdgpu_mmhub		mmhub;
+
 	/* gfx */
 	struct amdgpu_gfx		gfx;
 
@@ -974,9 +937,7 @@ struct amdgpu_device {
 	/* soc15 register offset based on ip, instance and  segment */
 	uint32_t 		*reg_offset[MAX_HWIP][HWIP_MAX_INSTANCE];
 
-	const struct amdgpu_nbio_funcs	*nbio_funcs;
 	const struct amdgpu_df_funcs	*df_funcs;
-	const struct amdgpu_mmhub_funcs	*mmhub_funcs;
 
 	/* delayed work_func for deferring clockgating during resume */
 	struct delayed_work     delayed_init_work;
@@ -1006,11 +967,11 @@ struct amdgpu_device {
 	struct mutex  lock_reset;
 	struct amdgpu_doorbell_index doorbell_index;
 
+	struct mutex			notifier_lock;
+
 	int asic_reset_res;
 	struct work_struct		xgmi_reset_work;
 
-	bool                            in_baco_reset;
-
 	long				gfx_timeout;
 	long				sdma_timeout;
 	long				video_timeout;
@@ -1018,6 +979,9 @@ struct amdgpu_device {
 
 	uint64_t			unique_id;
 	uint64_t	df_perfmon_config_assign_mask[AMDGPU_MAX_DF_PERFMONS];
+
+	/* device pstate */
+	int				pstate;
 };
 
 static inline struct amdgpu_device *amdgpu_ttm_adev(struct ttm_bo_device *bdev)
@@ -1032,6 +996,8 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 void amdgpu_device_fini(struct amdgpu_device *adev);
 int amdgpu_gpu_wait_for_idle(struct amdgpu_device *adev);
 
+void amdgpu_device_vram_access(struct amdgpu_device *adev, loff_t pos,
+			       uint32_t *buf, size_t size, bool write);
 uint32_t amdgpu_mm_rreg(struct amdgpu_device *adev, uint32_t reg,
 			uint32_t acc_flags);
 void amdgpu_mm_wreg(struct amdgpu_device *adev, uint32_t reg, uint32_t v,
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_acp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_acp.c
index eba42c752bca..82155ac3288a 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_acp.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_acp.c
@@ -189,7 +189,7 @@ static int acp_hw_init(void *handle)
 	u32 val = 0;
 	u32 count = 0;
 	struct device *dev;
-	struct i2s_platform_data *i2s_pdata;
+	struct i2s_platform_data *i2s_pdata = NULL;
 
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
@@ -231,20 +231,21 @@ static int acp_hw_init(void *handle)
 	adev->acp.acp_cell = kcalloc(ACP_DEVS, sizeof(struct mfd_cell),
 							GFP_KERNEL);
 
-	if (adev->acp.acp_cell == NULL)
-		return -ENOMEM;
+	if (adev->acp.acp_cell == NULL) {
+		r = -ENOMEM;
+		goto failure;
+	}
 
 	adev->acp.acp_res = kcalloc(5, sizeof(struct resource), GFP_KERNEL);
 	if (adev->acp.acp_res == NULL) {
-		kfree(adev->acp.acp_cell);
-		return -ENOMEM;
+		r = -ENOMEM;
+		goto failure;
 	}
 
 	i2s_pdata = kcalloc(3, sizeof(struct i2s_platform_data), GFP_KERNEL);
 	if (i2s_pdata == NULL) {
-		kfree(adev->acp.acp_res);
-		kfree(adev->acp.acp_cell);
-		return -ENOMEM;
+		r = -ENOMEM;
+		goto failure;
 	}
 
 	switch (adev->asic_type) {
@@ -341,14 +342,14 @@ static int acp_hw_init(void *handle)
 	r = mfd_add_hotplug_devices(adev->acp.parent, adev->acp.acp_cell,
 								ACP_DEVS);
 	if (r)
-		return r;
+		goto failure;
 
 	for (i = 0; i < ACP_DEVS ; i++) {
 		dev = get_mfd_cell_dev(adev->acp.acp_cell[i].name, i);
 		r = pm_genpd_add_device(&adev->acp.acp_genpd->gpd, dev);
 		if (r) {
 			dev_err(dev, "Failed to add dev to genpd\n");
-			return r;
+			goto failure;
 		}
 	}
 
@@ -367,7 +368,8 @@ static int acp_hw_init(void *handle)
 			break;
 		if (--count == 0) {
 			dev_err(&adev->pdev->dev, "Failed to reset ACP\n");
-			return -ETIMEDOUT;
+			r = -ETIMEDOUT;
+			goto failure;
 		}
 		udelay(100);
 	}
@@ -384,7 +386,8 @@ static int acp_hw_init(void *handle)
 			break;
 		if (--count == 0) {
 			dev_err(&adev->pdev->dev, "Failed to reset ACP\n");
-			return -ETIMEDOUT;
+			r = -ETIMEDOUT;
+			goto failure;
 		}
 		udelay(100);
 	}
@@ -393,6 +396,13 @@ static int acp_hw_init(void *handle)
 	val &= ~ACP_SOFT_RESET__SoftResetAud_MASK;
 	cgs_write_register(adev->acp.cgs_device, mmACP_SOFT_RESET, val);
 	return 0;
+
+failure:
+	kfree(i2s_pdata);
+	kfree(adev->acp.acp_res);
+	kfree(adev->acp.acp_cell);
+	kfree(adev->acp.acp_genpd);
+	return r;
 }
 
 /**
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c
index 07eb29885372..d3da9dde4ee1 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c
@@ -63,45 +63,10 @@ void amdgpu_amdkfd_fini(void)
 
 void amdgpu_amdkfd_device_probe(struct amdgpu_device *adev)
 {
-	const struct kfd2kgd_calls *kfd2kgd;
-
-	switch (adev->asic_type) {
-#ifdef CONFIG_DRM_AMDGPU_CIK
-	case CHIP_KAVERI:
-	case CHIP_HAWAII:
-		kfd2kgd = amdgpu_amdkfd_gfx_7_get_functions();
-		break;
-#endif
-	case CHIP_CARRIZO:
-	case CHIP_TONGA:
-	case CHIP_FIJI:
-	case CHIP_POLARIS10:
-	case CHIP_POLARIS11:
-	case CHIP_POLARIS12:
-	case CHIP_VEGAM:
-		kfd2kgd = amdgpu_amdkfd_gfx_8_0_get_functions();
-		break;
-	case CHIP_VEGA10:
-	case CHIP_VEGA12:
-	case CHIP_VEGA20:
-	case CHIP_RAVEN:
-		kfd2kgd = amdgpu_amdkfd_gfx_9_0_get_functions();
-		break;
-	case CHIP_ARCTURUS:
-		kfd2kgd = amdgpu_amdkfd_arcturus_get_functions();
-		break;
-	case CHIP_NAVI10:
-	case CHIP_NAVI14:
-	case CHIP_NAVI12:
-		kfd2kgd = amdgpu_amdkfd_gfx_10_0_get_functions();
-		break;
-	default:
-		dev_info(adev->dev, "kfd not supported on this ASIC\n");
-		return;
-	}
+	bool vf = amdgpu_sriov_vf(adev);
 
 	adev->kfd.dev = kgd2kfd_probe((struct kgd_dev *)adev,
-				      adev->pdev, kfd2kgd);
+				      adev->pdev, adev->asic_type, vf);
 
 	if (adev->kfd.dev)
 		amdgpu_amdkfd_total_mem_size += adev->gmc.real_vram_size;
@@ -165,14 +130,6 @@ void amdgpu_amdkfd_device_init(struct amdgpu_device *adev)
 				  adev->gfx.mec.queue_bitmap,
 				  KGD_MAX_QUEUES);
 
-		/* remove the KIQ bit as well */
-		if (adev->gfx.kiq.ring.sched.ready)
-			clear_bit(amdgpu_gfx_mec_queue_to_bit(adev,
-							  adev->gfx.kiq.ring.me - 1,
-							  adev->gfx.kiq.ring.pipe,
-							  adev->gfx.kiq.ring.queue),
-				  gpu_resources.queue_bitmap);
-
 		/* According to linux/bitmap.h we shouldn't use bitmap_clear if
 		 * nbits is not compile time constant
 		 */
@@ -202,7 +159,7 @@ void amdgpu_amdkfd_device_init(struct amdgpu_device *adev)
 					adev->doorbell_index.last_non_cp;
 		}
 
-		kgd2kfd_device_init(adev->kfd.dev, &gpu_resources);
+		kgd2kfd_device_init(adev->kfd.dev, adev->ddev, &gpu_resources);
 	}
 }
 
@@ -709,38 +666,14 @@ int amdgpu_amdkfd_evict_userptr(struct kgd_mem *mem, struct mm_struct *mm)
 	return 0;
 }
 
-struct kfd2kgd_calls *amdgpu_amdkfd_gfx_7_get_functions(void)
-{
-	return NULL;
-}
-
-struct kfd2kgd_calls *amdgpu_amdkfd_gfx_8_0_get_functions(void)
-{
-	return NULL;
-}
-
-struct kfd2kgd_calls *amdgpu_amdkfd_gfx_9_0_get_functions(void)
-{
-	return NULL;
-}
-
-struct kfd2kgd_calls *amdgpu_amdkfd_arcturus_get_functions(void)
-{
-	return NULL;
-}
-
-struct kfd2kgd_calls *amdgpu_amdkfd_gfx_10_0_get_functions(void)
-{
-	return NULL;
-}
-
 struct kfd_dev *kgd2kfd_probe(struct kgd_dev *kgd, struct pci_dev *pdev,
-			      const struct kfd2kgd_calls *f2g)
+			      unsigned int asic_type, bool vf)
 {
 	return NULL;
 }
 
 bool kgd2kfd_device_init(struct kfd_dev *kfd,
+			 struct drm_device *ddev,
 			 const struct kgd2kfd_shared_resources *gpu_resources)
 {
 	return false;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h
index e519df3fd2b6..069d5d230810 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h
@@ -57,7 +57,7 @@ struct kgd_mem {
 	unsigned int mapped_to_gpu_memory;
 	uint64_t va;
 
-	uint32_t mapping_flags;
+	uint32_t alloc_flags;
 
 	atomic_t invalid;
 	struct amdkfd_process_info *process_info;
@@ -137,12 +137,6 @@ int amdgpu_amdkfd_submit_ib(struct kgd_dev *kgd, enum kgd_engine_type engine,
 void amdgpu_amdkfd_set_compute_idle(struct kgd_dev *kgd, bool idle);
 bool amdgpu_amdkfd_have_atomics_support(struct kgd_dev *kgd);
 
-struct kfd2kgd_calls *amdgpu_amdkfd_gfx_7_get_functions(void);
-struct kfd2kgd_calls *amdgpu_amdkfd_gfx_8_0_get_functions(void);
-struct kfd2kgd_calls *amdgpu_amdkfd_gfx_9_0_get_functions(void);
-struct kfd2kgd_calls *amdgpu_amdkfd_arcturus_get_functions(void);
-struct kfd2kgd_calls *amdgpu_amdkfd_gfx_10_0_get_functions(void);
-
 bool amdgpu_amdkfd_is_kfd_vmid(struct amdgpu_device *adev, u32 vmid);
 
 int amdgpu_amdkfd_pre_reset(struct amdgpu_device *adev);
@@ -179,10 +173,17 @@ uint64_t amdgpu_amdkfd_get_mmio_remap_phys_addr(struct kgd_dev *kgd);
 uint32_t amdgpu_amdkfd_get_num_gws(struct kgd_dev *kgd);
 uint8_t amdgpu_amdkfd_get_xgmi_hops_count(struct kgd_dev *dst, struct kgd_dev *src);
 
+/* Read user wptr from a specified user address space with page fault
+ * disabled. The memory must be pinned and mapped to the hardware when
+ * this is called in hqd_load functions, so it should never fault in
+ * the first place. This resolves a circular lock dependency involving
+ * four locks, including the DQM lock and mmap_sem.
+ */
 #define read_user_wptr(mmptr, wptr, dst)				\
 	({								\
 		bool valid = false;					\
 		if ((mmptr) && (wptr)) {				\
+			pagefault_disable();				\
 			if ((mmptr) == current->mm) {			\
 				valid = !get_user((dst), (wptr));	\
 			} else if (current->mm == NULL) {		\
@@ -190,6 +191,7 @@ uint8_t amdgpu_amdkfd_get_xgmi_hops_count(struct kgd_dev *dst, struct kgd_dev *s
 				valid = !get_user((dst), (wptr));	\
 				unuse_mm(mmptr);			\
 			}						\
+			pagefault_enable();				\
 		}							\
 		valid;							\
 	})
@@ -240,8 +242,9 @@ void amdgpu_amdkfd_unreserve_memory_limit(struct amdgpu_bo *bo);
 int kgd2kfd_init(void);
 void kgd2kfd_exit(void);
 struct kfd_dev *kgd2kfd_probe(struct kgd_dev *kgd, struct pci_dev *pdev,
-			      const struct kfd2kgd_calls *f2g);
+			      unsigned int asic_type, bool vf);
 bool kgd2kfd_device_init(struct kfd_dev *kfd,
+			 struct drm_device *ddev,
 			 const struct kgd2kfd_shared_resources *gpu_resources);
 void kgd2kfd_device_exit(struct kfd_dev *kfd);
 void kgd2kfd_suspend(struct kfd_dev *kfd);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_arcturus.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_arcturus.c
index c79aaebeeaf0..b6713e0ed1b2 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_arcturus.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_arcturus.c
@@ -19,10 +19,6 @@
  * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
  * OTHER DEALINGS IN THE SOFTWARE.
  */
-
-#undef pr_fmt
-#define pr_fmt(fmt) "kfd2kgd: " fmt
-
 #include <linux/module.h>
 #include <linux/fdtable.h>
 #include <linux/uaccess.h>
@@ -69,11 +65,11 @@ static inline struct v9_sdma_mqd *get_sdma_mqd(void *mqd)
 	return (struct v9_sdma_mqd *)mqd;
 }
 
-static uint32_t get_sdma_base_addr(struct amdgpu_device *adev,
+static uint32_t get_sdma_rlc_reg_offset(struct amdgpu_device *adev,
 				unsigned int engine_id,
 				unsigned int queue_id)
 {
-	uint32_t base[8] = {
+	uint32_t sdma_engine_reg_base[8] = {
 		SOC15_REG_OFFSET(SDMA0, 0,
 				 mmSDMA0_RLC0_RB_CNTL) - mmSDMA0_RLC0_RB_CNTL,
 		SOC15_REG_OFFSET(SDMA1, 0,
@@ -91,111 +87,82 @@ static uint32_t get_sdma_base_addr(struct amdgpu_device *adev,
 		SOC15_REG_OFFSET(SDMA7, 0,
 				 mmSDMA7_RLC0_RB_CNTL) - mmSDMA7_RLC0_RB_CNTL
 	};
-	uint32_t retval;
 
-	retval = base[engine_id] + queue_id * (mmSDMA0_RLC1_RB_CNTL -
-					       mmSDMA0_RLC0_RB_CNTL);
+	uint32_t retval = sdma_engine_reg_base[engine_id]
+		+ queue_id * (mmSDMA0_RLC1_RB_CNTL - mmSDMA0_RLC0_RB_CNTL);
 
-	pr_debug("sdma base address: 0x%x\n", retval);
+	pr_debug("RLC register offset for SDMA%d RLC%d: 0x%x\n", engine_id,
+			queue_id, retval);
 
 	return retval;
 }
 
-static u32 sdma_v4_0_get_reg_offset(struct amdgpu_device *adev,
-		u32 instance, u32 offset)
-{
-	switch (instance) {
-	case 0:
-		return (adev->reg_offset[SDMA0_HWIP][0][0] + offset);
-	case 1:
-		return (adev->reg_offset[SDMA1_HWIP][0][1] + offset);
-	case 2:
-		return (adev->reg_offset[SDMA2_HWIP][0][1] + offset);
-	case 3:
-		return (adev->reg_offset[SDMA3_HWIP][0][1] + offset);
-	case 4:
-		return (adev->reg_offset[SDMA4_HWIP][0][1] + offset);
-	case 5:
-		return (adev->reg_offset[SDMA5_HWIP][0][1] + offset);
-	case 6:
-		return (adev->reg_offset[SDMA6_HWIP][0][1] + offset);
-	case 7:
-		return (adev->reg_offset[SDMA7_HWIP][0][1] + offset);
-	default:
-		break;
-	}
-	return 0;
-}
-
 static int kgd_hqd_sdma_load(struct kgd_dev *kgd, void *mqd,
 			     uint32_t __user *wptr, struct mm_struct *mm)
 {
 	struct amdgpu_device *adev = get_amdgpu_device(kgd);
 	struct v9_sdma_mqd *m;
-	uint32_t sdma_base_addr, sdmax_gfx_context_cntl;
+	uint32_t sdma_rlc_reg_offset;
 	unsigned long end_jiffies;
 	uint32_t data;
 	uint64_t data64;
 	uint64_t __user *wptr64 = (uint64_t __user *)wptr;
 
 	m = get_sdma_mqd(mqd);
-	sdma_base_addr = get_sdma_base_addr(adev, m->sdma_engine_id,
+	sdma_rlc_reg_offset = get_sdma_rlc_reg_offset(adev, m->sdma_engine_id,
 					    m->sdma_queue_id);
-	sdmax_gfx_context_cntl = sdma_v4_0_get_reg_offset(adev,
-			m->sdma_engine_id, mmSDMA0_GFX_CONTEXT_CNTL);
 
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL,
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL,
 		m->sdmax_rlcx_rb_cntl & (~SDMA0_RLC0_RB_CNTL__RB_ENABLE_MASK));
 
 	end_jiffies = msecs_to_jiffies(2000) + jiffies;
 	while (true) {
-		data = RREG32(sdma_base_addr + mmSDMA0_RLC0_CONTEXT_STATUS);
+		data = RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_CONTEXT_STATUS);
 		if (data & SDMA0_RLC0_CONTEXT_STATUS__IDLE_MASK)
 			break;
-		if (time_after(jiffies, end_jiffies))
+		if (time_after(jiffies, end_jiffies)) {
+			pr_err("SDMA RLC not idle in %s\n", __func__);
 			return -ETIME;
+		}
 		usleep_range(500, 1000);
 	}
-	data = RREG32(sdmax_gfx_context_cntl);
-	data = REG_SET_FIELD(data, SDMA0_GFX_CONTEXT_CNTL,
-			     RESUME_CTX, 0);
-	WREG32(sdmax_gfx_context_cntl, data);
 
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_DOORBELL_OFFSET,
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_DOORBELL_OFFSET,
 	       m->sdmax_rlcx_doorbell_offset);
 
 	data = REG_SET_FIELD(m->sdmax_rlcx_doorbell, SDMA0_RLC0_DOORBELL,
 			     ENABLE, 1);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_DOORBELL, data);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR, m->sdmax_rlcx_rb_rptr);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR_HI,
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_DOORBELL, data);
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_RPTR,
+				m->sdmax_rlcx_rb_rptr);
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_RPTR_HI,
 				m->sdmax_rlcx_rb_rptr_hi);
 
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_MINOR_PTR_UPDATE, 1);
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_MINOR_PTR_UPDATE, 1);
 	if (read_user_wptr(mm, wptr64, data64)) {
-		WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_WPTR,
+		WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_WPTR,
 		       lower_32_bits(data64));
-		WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_WPTR_HI,
+		WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_WPTR_HI,
 		       upper_32_bits(data64));
 	} else {
-		WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_WPTR,
+		WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_WPTR,
 		       m->sdmax_rlcx_rb_rptr);
-		WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_WPTR_HI,
+		WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_WPTR_HI,
 		       m->sdmax_rlcx_rb_rptr_hi);
 	}
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_MINOR_PTR_UPDATE, 0);
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_MINOR_PTR_UPDATE, 0);
 
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_BASE, m->sdmax_rlcx_rb_base);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_BASE_HI,
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_BASE, m->sdmax_rlcx_rb_base);
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_BASE_HI,
 			m->sdmax_rlcx_rb_base_hi);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR_ADDR_LO,
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_RPTR_ADDR_LO,
 			m->sdmax_rlcx_rb_rptr_addr_lo);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR_ADDR_HI,
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_RPTR_ADDR_HI,
 			m->sdmax_rlcx_rb_rptr_addr_hi);
 
 	data = REG_SET_FIELD(m->sdmax_rlcx_rb_cntl, SDMA0_RLC0_RB_CNTL,
 			     RB_ENABLE, 1);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL, data);
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL, data);
 
 	return 0;
 }
@@ -205,7 +172,8 @@ static int kgd_hqd_sdma_dump(struct kgd_dev *kgd,
 			     uint32_t (**dump)[2], uint32_t *n_regs)
 {
 	struct amdgpu_device *adev = get_amdgpu_device(kgd);
-	uint32_t sdma_base_addr = get_sdma_base_addr(adev, engine_id, queue_id);
+	uint32_t sdma_rlc_reg_offset = get_sdma_rlc_reg_offset(adev,
+			engine_id, queue_id);
 	uint32_t i = 0, reg;
 #undef HQD_N_REGS
 #define HQD_N_REGS (19+6+7+10)
@@ -215,15 +183,15 @@ static int kgd_hqd_sdma_dump(struct kgd_dev *kgd,
 		return -ENOMEM;
 
 	for (reg = mmSDMA0_RLC0_RB_CNTL; reg <= mmSDMA0_RLC0_DOORBELL; reg++)
-		DUMP_REG(sdma_base_addr + reg);
+		DUMP_REG(sdma_rlc_reg_offset + reg);
 	for (reg = mmSDMA0_RLC0_STATUS; reg <= mmSDMA0_RLC0_CSA_ADDR_HI; reg++)
-		DUMP_REG(sdma_base_addr + reg);
+		DUMP_REG(sdma_rlc_reg_offset + reg);
 	for (reg = mmSDMA0_RLC0_IB_SUB_REMAIN;
 	     reg <= mmSDMA0_RLC0_MINOR_PTR_UPDATE; reg++)
-		DUMP_REG(sdma_base_addr + reg);
+		DUMP_REG(sdma_rlc_reg_offset + reg);
 	for (reg = mmSDMA0_RLC0_MIDCMD_DATA0;
 	     reg <= mmSDMA0_RLC0_MIDCMD_CNTL; reg++)
-		DUMP_REG(sdma_base_addr + reg);
+		DUMP_REG(sdma_rlc_reg_offset + reg);
 
 	WARN_ON_ONCE(i != HQD_N_REGS);
 	*n_regs = i;
@@ -235,14 +203,14 @@ static bool kgd_hqd_sdma_is_occupied(struct kgd_dev *kgd, void *mqd)
 {
 	struct amdgpu_device *adev = get_amdgpu_device(kgd);
 	struct v9_sdma_mqd *m;
-	uint32_t sdma_base_addr;
+	uint32_t sdma_rlc_reg_offset;
 	uint32_t sdma_rlc_rb_cntl;
 
 	m = get_sdma_mqd(mqd);
-	sdma_base_addr = get_sdma_base_addr(adev, m->sdma_engine_id,
+	sdma_rlc_reg_offset = get_sdma_rlc_reg_offset(adev, m->sdma_engine_id,
 					    m->sdma_queue_id);
 
-	sdma_rlc_rb_cntl = RREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL);
+	sdma_rlc_rb_cntl = RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL);
 
 	if (sdma_rlc_rb_cntl & SDMA0_RLC0_RB_CNTL__RB_ENABLE_MASK)
 		return true;
@@ -255,40 +223,42 @@ static int kgd_hqd_sdma_destroy(struct kgd_dev *kgd, void *mqd,
 {
 	struct amdgpu_device *adev = get_amdgpu_device(kgd);
 	struct v9_sdma_mqd *m;
-	uint32_t sdma_base_addr;
+	uint32_t sdma_rlc_reg_offset;
 	uint32_t temp;
 	unsigned long end_jiffies = (utimeout * HZ / 1000) + jiffies;
 
 	m = get_sdma_mqd(mqd);
-	sdma_base_addr = get_sdma_base_addr(adev, m->sdma_engine_id,
+	sdma_rlc_reg_offset = get_sdma_rlc_reg_offset(adev, m->sdma_engine_id,
 					    m->sdma_queue_id);
 
-	temp = RREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL);
+	temp = RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL);
 	temp = temp & ~SDMA0_RLC0_RB_CNTL__RB_ENABLE_MASK;
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL, temp);
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL, temp);
 
 	while (true) {
-		temp = RREG32(sdma_base_addr + mmSDMA0_RLC0_CONTEXT_STATUS);
+		temp = RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_CONTEXT_STATUS);
 		if (temp & SDMA0_RLC0_CONTEXT_STATUS__IDLE_MASK)
 			break;
-		if (time_after(jiffies, end_jiffies))
+		if (time_after(jiffies, end_jiffies)) {
+			pr_err("SDMA RLC not idle in %s\n", __func__);
 			return -ETIME;
+		}
 		usleep_range(500, 1000);
 	}
 
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_DOORBELL, 0);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL,
-		RREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL) |
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_DOORBELL, 0);
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL,
+		RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL) |
 		SDMA0_RLC0_RB_CNTL__RB_ENABLE_MASK);
 
-	m->sdmax_rlcx_rb_rptr = RREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR);
+	m->sdmax_rlcx_rb_rptr = RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_RPTR);
 	m->sdmax_rlcx_rb_rptr_hi =
-		RREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR_HI);
+		RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_RPTR_HI);
 
 	return 0;
 }
 
-static const struct kfd2kgd_calls kfd2kgd = {
+const struct kfd2kgd_calls arcturus_kfd2kgd = {
 	.program_sh_mem_settings = kgd_gfx_v9_program_sh_mem_settings,
 	.set_pasid_vmid_mapping = kgd_gfx_v9_set_pasid_vmid_mapping,
 	.init_interrupts = kgd_gfx_v9_init_interrupts,
@@ -304,20 +274,11 @@ static const struct kfd2kgd_calls kfd2kgd = {
 	.address_watch_execute = kgd_gfx_v9_address_watch_execute,
 	.wave_control_execute = kgd_gfx_v9_wave_control_execute,
 	.address_watch_get_offset = kgd_gfx_v9_address_watch_get_offset,
-	.get_atc_vmid_pasid_mapping_pasid =
-			kgd_gfx_v9_get_atc_vmid_pasid_mapping_pasid,
-	.get_atc_vmid_pasid_mapping_valid =
-			kgd_gfx_v9_get_atc_vmid_pasid_mapping_valid,
-	.set_scratch_backing_va = kgd_gfx_v9_set_scratch_backing_va,
+	.get_atc_vmid_pasid_mapping_info =
+			kgd_gfx_v9_get_atc_vmid_pasid_mapping_info,
 	.get_tile_config = kgd_gfx_v9_get_tile_config,
 	.set_vm_context_page_table_base = kgd_gfx_v9_set_vm_context_page_table_base,
 	.invalidate_tlbs = kgd_gfx_v9_invalidate_tlbs,
 	.invalidate_tlbs_vmid = kgd_gfx_v9_invalidate_tlbs_vmid,
 	.get_hive_id = amdgpu_amdkfd_get_hive_id,
 };
-
-struct kfd2kgd_calls *amdgpu_amdkfd_arcturus_get_functions(void)
-{
-	return (struct kfd2kgd_calls *)&kfd2kgd;
-}
-
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v10.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v10.c
index d10f483f5e27..61cd707158e4 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v10.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v10.c
@@ -19,18 +19,9 @@
  * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
  * OTHER DEALINGS IN THE SOFTWARE.
  */
-#undef pr_fmt
-#define pr_fmt(fmt) "kfd2kgd: " fmt
-
-#include <linux/module.h>
-#include <linux/fdtable.h>
-#include <linux/uaccess.h>
-#include <linux/firmware.h>
 #include <linux/mmu_context.h>
 #include "amdgpu.h"
 #include "amdgpu_amdkfd.h"
-#include "amdgpu_ucode.h"
-#include "soc15_hw_ip.h"
 #include "gc/gc_10_1_0_offset.h"
 #include "gc/gc_10_1_0_sh_mask.h"
 #include "navi10_enum.h"
@@ -42,6 +33,7 @@
 #include "v10_structs.h"
 #include "nv.h"
 #include "nvd.h"
+#include "gfxhub_v2_0.h"
 
 enum hqd_dequeue_request_type {
 	NO_ACTION = 0,
@@ -50,63 +42,6 @@ enum hqd_dequeue_request_type {
 	SAVE_WAVES
 };
 
-/*
- * Register access functions
- */
-
-static void kgd_program_sh_mem_settings(struct kgd_dev *kgd, uint32_t vmid,
-		uint32_t sh_mem_config,
-		uint32_t sh_mem_ape1_base, uint32_t sh_mem_ape1_limit,
-		uint32_t sh_mem_bases);
-static int kgd_set_pasid_vmid_mapping(struct kgd_dev *kgd, unsigned int pasid,
-		unsigned int vmid);
-static int kgd_init_interrupts(struct kgd_dev *kgd, uint32_t pipe_id);
-static int kgd_hqd_load(struct kgd_dev *kgd, void *mqd, uint32_t pipe_id,
-			uint32_t queue_id, uint32_t __user *wptr,
-			uint32_t wptr_shift, uint32_t wptr_mask,
-			struct mm_struct *mm);
-static int kgd_hqd_dump(struct kgd_dev *kgd,
-			uint32_t pipe_id, uint32_t queue_id,
-			uint32_t (**dump)[2], uint32_t *n_regs);
-static int kgd_hqd_sdma_load(struct kgd_dev *kgd, void *mqd,
-			     uint32_t __user *wptr, struct mm_struct *mm);
-static int kgd_hqd_sdma_dump(struct kgd_dev *kgd,
-			     uint32_t engine_id, uint32_t queue_id,
-			     uint32_t (**dump)[2], uint32_t *n_regs);
-static bool kgd_hqd_is_occupied(struct kgd_dev *kgd, uint64_t queue_address,
-		uint32_t pipe_id, uint32_t queue_id);
-static bool kgd_hqd_sdma_is_occupied(struct kgd_dev *kgd, void *mqd);
-static int kgd_hqd_destroy(struct kgd_dev *kgd, void *mqd,
-				enum kfd_preempt_type reset_type,
-				unsigned int utimeout, uint32_t pipe_id,
-				uint32_t queue_id);
-static int kgd_hqd_sdma_destroy(struct kgd_dev *kgd, void *mqd,
-				unsigned int utimeout);
-#if 0
-static uint32_t get_watch_base_addr(struct amdgpu_device *adev);
-#endif
-static int kgd_address_watch_disable(struct kgd_dev *kgd);
-static int kgd_address_watch_execute(struct kgd_dev *kgd,
-					unsigned int watch_point_id,
-					uint32_t cntl_val,
-					uint32_t addr_hi,
-					uint32_t addr_lo);
-static int kgd_wave_control_execute(struct kgd_dev *kgd,
-					uint32_t gfx_index_val,
-					uint32_t sq_cmd);
-static uint32_t kgd_address_watch_get_offset(struct kgd_dev *kgd,
-					unsigned int watch_point_id,
-					unsigned int reg_offset);
-
-static bool get_atc_vmid_pasid_mapping_valid(struct kgd_dev *kgd,
-		uint8_t vmid);
-static uint16_t get_atc_vmid_pasid_mapping_pasid(struct kgd_dev *kgd,
-		uint8_t vmid);
-static void set_vm_context_page_table_base(struct kgd_dev *kgd, uint32_t vmid,
-		uint64_t page_table_base);
-static int invalidate_tlbs(struct kgd_dev *kgd, uint16_t pasid);
-static int invalidate_tlbs_vmid(struct kgd_dev *kgd, uint16_t vmid);
-
 /* Because of REG_GET_FIELD() being used, we put this function in the
  * asic specific file.
  */
@@ -139,37 +74,6 @@ static int amdgpu_amdkfd_get_tile_config(struct kgd_dev *kgd,
 	return 0;
 }
 
-static const struct kfd2kgd_calls kfd2kgd = {
-	.program_sh_mem_settings = kgd_program_sh_mem_settings,
-	.set_pasid_vmid_mapping = kgd_set_pasid_vmid_mapping,
-	.init_interrupts = kgd_init_interrupts,
-	.hqd_load = kgd_hqd_load,
-	.hqd_sdma_load = kgd_hqd_sdma_load,
-	.hqd_dump = kgd_hqd_dump,
-	.hqd_sdma_dump = kgd_hqd_sdma_dump,
-	.hqd_is_occupied = kgd_hqd_is_occupied,
-	.hqd_sdma_is_occupied = kgd_hqd_sdma_is_occupied,
-	.hqd_destroy = kgd_hqd_destroy,
-	.hqd_sdma_destroy = kgd_hqd_sdma_destroy,
-	.address_watch_disable = kgd_address_watch_disable,
-	.address_watch_execute = kgd_address_watch_execute,
-	.wave_control_execute = kgd_wave_control_execute,
-	.address_watch_get_offset = kgd_address_watch_get_offset,
-	.get_atc_vmid_pasid_mapping_pasid =
-			get_atc_vmid_pasid_mapping_pasid,
-	.get_atc_vmid_pasid_mapping_valid =
-			get_atc_vmid_pasid_mapping_valid,
-	.invalidate_tlbs = invalidate_tlbs,
-	.invalidate_tlbs_vmid = invalidate_tlbs_vmid,
-	.set_vm_context_page_table_base = set_vm_context_page_table_base,
-	.get_tile_config = amdgpu_amdkfd_get_tile_config,
-};
-
-struct kfd2kgd_calls *amdgpu_amdkfd_gfx_10_0_get_functions()
-{
-	return (struct kfd2kgd_calls *)&kfd2kgd;
-}
-
 static inline struct amdgpu_device *get_amdgpu_device(struct kgd_dev *kgd)
 {
 	return (struct amdgpu_device *)kgd;
@@ -250,11 +154,6 @@ static int kgd_set_pasid_vmid_mapping(struct kgd_dev *kgd, unsigned int pasid,
 			ATC_VMID0_PASID_MAPPING__VALID_MASK;
 
 	pr_debug("pasid 0x%x vmid %d, reg value %x\n", pasid, vmid, pasid_mapping);
-	/*
-	 * need to do this twice, once for gfx and once for mmhub
-	 * for ATC add 16 to VMID for mmhub, for IH different registers.
-	 * ATC_VMID0..15 registers are separate from ATC_VMID16..31.
-	 */
 
 	pr_debug("ATHUB, reg %x\n", SOC15_REG_OFFSET(ATHUB, 0, mmATC_VMID0_PASID_MAPPING) + vmid);
 	WREG32(SOC15_REG_OFFSET(ATHUB, 0, mmATC_VMID0_PASID_MAPPING) + vmid,
@@ -306,11 +205,11 @@ static int kgd_init_interrupts(struct kgd_dev *kgd, uint32_t pipe_id)
 	return 0;
 }
 
-static uint32_t get_sdma_base_addr(struct amdgpu_device *adev,
+static uint32_t get_sdma_rlc_reg_offset(struct amdgpu_device *adev,
 				unsigned int engine_id,
 				unsigned int queue_id)
 {
-	uint32_t base[2] = {
+	uint32_t sdma_engine_reg_base[2] = {
 		SOC15_REG_OFFSET(SDMA0, 0,
 				 mmSDMA0_RLC0_RB_CNTL) - mmSDMA0_RLC0_RB_CNTL,
 		/* On gfx10, mmSDMA1_xxx registers are defined NOT based
@@ -322,12 +221,12 @@ static uint32_t get_sdma_base_addr(struct amdgpu_device *adev,
 		SOC15_REG_OFFSET(SDMA1, 0,
 				 mmSDMA1_RLC0_RB_CNTL) - mmSDMA0_RLC0_RB_CNTL
 	};
-	uint32_t retval;
 
-	retval = base[engine_id] + queue_id * (mmSDMA0_RLC1_RB_CNTL -
-					       mmSDMA0_RLC0_RB_CNTL);
+	uint32_t retval = sdma_engine_reg_base[engine_id]
+		+ queue_id * (mmSDMA0_RLC1_RB_CNTL - mmSDMA0_RLC0_RB_CNTL);
 
-	pr_debug("sdma base address: 0x%x\n", retval);
+	pr_debug("RLC register offset for SDMA%d RLC%d: 0x%x\n", engine_id,
+			queue_id, retval);
 
 	return retval;
 }
@@ -488,72 +387,67 @@ static int kgd_hqd_sdma_load(struct kgd_dev *kgd, void *mqd,
 {
 	struct amdgpu_device *adev = get_amdgpu_device(kgd);
 	struct v10_sdma_mqd *m;
-	uint32_t sdma_base_addr, sdmax_gfx_context_cntl;
+	uint32_t sdma_rlc_reg_offset;
 	unsigned long end_jiffies;
 	uint32_t data;
 	uint64_t data64;
 	uint64_t __user *wptr64 = (uint64_t __user *)wptr;
 
 	m = get_sdma_mqd(mqd);
-	sdma_base_addr = get_sdma_base_addr(adev, m->sdma_engine_id,
+	sdma_rlc_reg_offset = get_sdma_rlc_reg_offset(adev, m->sdma_engine_id,
 					    m->sdma_queue_id);
-	pr_debug("sdma load base addr %x for engine %d, queue %d\n", sdma_base_addr, m->sdma_engine_id, m->sdma_queue_id);
-	sdmax_gfx_context_cntl = m->sdma_engine_id ?
-		SOC15_REG_OFFSET(SDMA1, 0, mmSDMA1_GFX_CONTEXT_CNTL) :
-		SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_GFX_CONTEXT_CNTL);
 
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL,
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL,
 		m->sdmax_rlcx_rb_cntl & (~SDMA0_RLC0_RB_CNTL__RB_ENABLE_MASK));
 
 	end_jiffies = msecs_to_jiffies(2000) + jiffies;
 	while (true) {
-		data = RREG32(sdma_base_addr + mmSDMA0_RLC0_CONTEXT_STATUS);
+		data = RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_CONTEXT_STATUS);
 		if (data & SDMA0_RLC0_CONTEXT_STATUS__IDLE_MASK)
 			break;
-		if (time_after(jiffies, end_jiffies))
+		if (time_after(jiffies, end_jiffies)) {
+			pr_err("SDMA RLC not idle in %s\n", __func__);
 			return -ETIME;
+		}
 		usleep_range(500, 1000);
 	}
-	data = RREG32(sdmax_gfx_context_cntl);
-	data = REG_SET_FIELD(data, SDMA0_GFX_CONTEXT_CNTL,
-			     RESUME_CTX, 0);
-	WREG32(sdmax_gfx_context_cntl, data);
 
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_DOORBELL_OFFSET,
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_DOORBELL_OFFSET,
 	       m->sdmax_rlcx_doorbell_offset);
 
 	data = REG_SET_FIELD(m->sdmax_rlcx_doorbell, SDMA0_RLC0_DOORBELL,
 			     ENABLE, 1);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_DOORBELL, data);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR, m->sdmax_rlcx_rb_rptr);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR_HI,
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_DOORBELL, data);
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_RPTR,
+				m->sdmax_rlcx_rb_rptr);
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_RPTR_HI,
 				m->sdmax_rlcx_rb_rptr_hi);
 
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_MINOR_PTR_UPDATE, 1);
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_MINOR_PTR_UPDATE, 1);
 	if (read_user_wptr(mm, wptr64, data64)) {
-		WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_WPTR,
+		WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_WPTR,
 		       lower_32_bits(data64));
-		WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_WPTR_HI,
+		WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_WPTR_HI,
 		       upper_32_bits(data64));
 	} else {
-		WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_WPTR,
+		WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_WPTR,
 		       m->sdmax_rlcx_rb_rptr);
-		WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_WPTR_HI,
+		WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_WPTR_HI,
 		       m->sdmax_rlcx_rb_rptr_hi);
 	}
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_MINOR_PTR_UPDATE, 0);
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_MINOR_PTR_UPDATE, 0);
 
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_BASE, m->sdmax_rlcx_rb_base);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_BASE_HI,
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_BASE, m->sdmax_rlcx_rb_base);
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_BASE_HI,
 			m->sdmax_rlcx_rb_base_hi);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR_ADDR_LO,
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_RPTR_ADDR_LO,
 			m->sdmax_rlcx_rb_rptr_addr_lo);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR_ADDR_HI,
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_RPTR_ADDR_HI,
 			m->sdmax_rlcx_rb_rptr_addr_hi);
 
 	data = REG_SET_FIELD(m->sdmax_rlcx_rb_cntl, SDMA0_RLC0_RB_CNTL,
 			     RB_ENABLE, 1);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL, data);
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL, data);
 
 	return 0;
 }
@@ -563,28 +457,26 @@ static int kgd_hqd_sdma_dump(struct kgd_dev *kgd,
 			     uint32_t (**dump)[2], uint32_t *n_regs)
 {
 	struct amdgpu_device *adev = get_amdgpu_device(kgd);
-	uint32_t sdma_base_addr = get_sdma_base_addr(adev, engine_id, queue_id);
+	uint32_t sdma_rlc_reg_offset = get_sdma_rlc_reg_offset(adev,
+			engine_id, queue_id);
 	uint32_t i = 0, reg;
 #undef HQD_N_REGS
 #define HQD_N_REGS (19+6+7+10)
 
-	pr_debug("sdma dump engine id %d queue_id %d\n", engine_id, queue_id);
-	pr_debug("sdma base addr %x\n", sdma_base_addr);
-
 	*dump = kmalloc(HQD_N_REGS*2*sizeof(uint32_t), GFP_KERNEL);
 	if (*dump == NULL)
 		return -ENOMEM;
 
 	for (reg = mmSDMA0_RLC0_RB_CNTL; reg <= mmSDMA0_RLC0_DOORBELL; reg++)
-		DUMP_REG(sdma_base_addr + reg);
+		DUMP_REG(sdma_rlc_reg_offset + reg);
 	for (reg = mmSDMA0_RLC0_STATUS; reg <= mmSDMA0_RLC0_CSA_ADDR_HI; reg++)
-		DUMP_REG(sdma_base_addr + reg);
+		DUMP_REG(sdma_rlc_reg_offset + reg);
 	for (reg = mmSDMA0_RLC0_IB_SUB_REMAIN;
 	     reg <= mmSDMA0_RLC0_MINOR_PTR_UPDATE; reg++)
-		DUMP_REG(sdma_base_addr + reg);
+		DUMP_REG(sdma_rlc_reg_offset + reg);
 	for (reg = mmSDMA0_RLC0_MIDCMD_DATA0;
 	     reg <= mmSDMA0_RLC0_MIDCMD_CNTL; reg++)
-		DUMP_REG(sdma_base_addr + reg);
+		DUMP_REG(sdma_rlc_reg_offset + reg);
 
 	WARN_ON_ONCE(i != HQD_N_REGS);
 	*n_regs = i;
@@ -618,14 +510,14 @@ static bool kgd_hqd_sdma_is_occupied(struct kgd_dev *kgd, void *mqd)
 {
 	struct amdgpu_device *adev = get_amdgpu_device(kgd);
 	struct v10_sdma_mqd *m;
-	uint32_t sdma_base_addr;
+	uint32_t sdma_rlc_reg_offset;
 	uint32_t sdma_rlc_rb_cntl;
 
 	m = get_sdma_mqd(mqd);
-	sdma_base_addr = get_sdma_base_addr(adev, m->sdma_engine_id,
+	sdma_rlc_reg_offset = get_sdma_rlc_reg_offset(adev, m->sdma_engine_id,
 					    m->sdma_queue_id);
 
-	sdma_rlc_rb_cntl = RREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL);
+	sdma_rlc_rb_cntl = RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL);
 
 	if (sdma_rlc_rb_cntl & SDMA0_RLC0_RB_CNTL__RB_ENABLE_MASK)
 		return true;
@@ -746,59 +638,52 @@ static int kgd_hqd_sdma_destroy(struct kgd_dev *kgd, void *mqd,
 {
 	struct amdgpu_device *adev = get_amdgpu_device(kgd);
 	struct v10_sdma_mqd *m;
-	uint32_t sdma_base_addr;
+	uint32_t sdma_rlc_reg_offset;
 	uint32_t temp;
 	unsigned long end_jiffies = (utimeout * HZ / 1000) + jiffies;
 
 	m = get_sdma_mqd(mqd);
-	sdma_base_addr = get_sdma_base_addr(adev, m->sdma_engine_id,
+	sdma_rlc_reg_offset = get_sdma_rlc_reg_offset(adev, m->sdma_engine_id,
 					    m->sdma_queue_id);
 
-	temp = RREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL);
+	temp = RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL);
 	temp = temp & ~SDMA0_RLC0_RB_CNTL__RB_ENABLE_MASK;
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL, temp);
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL, temp);
 
 	while (true) {
-		temp = RREG32(sdma_base_addr + mmSDMA0_RLC0_CONTEXT_STATUS);
+		temp = RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_CONTEXT_STATUS);
 		if (temp & SDMA0_RLC0_CONTEXT_STATUS__IDLE_MASK)
 			break;
-		if (time_after(jiffies, end_jiffies))
+		if (time_after(jiffies, end_jiffies)) {
+			pr_err("SDMA RLC not idle in %s\n", __func__);
 			return -ETIME;
+		}
 		usleep_range(500, 1000);
 	}
 
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_DOORBELL, 0);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL,
-		RREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL) |
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_DOORBELL, 0);
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL,
+		RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL) |
 		SDMA0_RLC0_RB_CNTL__RB_ENABLE_MASK);
 
-	m->sdmax_rlcx_rb_rptr = RREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR);
+	m->sdmax_rlcx_rb_rptr = RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_RPTR);
 	m->sdmax_rlcx_rb_rptr_hi =
-		RREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR_HI);
+		RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_RPTR_HI);
 
 	return 0;
 }
 
-static bool get_atc_vmid_pasid_mapping_valid(struct kgd_dev *kgd,
-							uint8_t vmid)
+static bool get_atc_vmid_pasid_mapping_info(struct kgd_dev *kgd,
+					uint8_t vmid, uint16_t *p_pasid)
 {
-	uint32_t reg;
+	uint32_t value;
 	struct amdgpu_device *adev = (struct amdgpu_device *) kgd;
 
-	reg = RREG32(SOC15_REG_OFFSET(ATHUB, 0, mmATC_VMID0_PASID_MAPPING)
+	value = RREG32(SOC15_REG_OFFSET(ATHUB, 0, mmATC_VMID0_PASID_MAPPING)
 		     + vmid);
-	return reg & ATC_VMID0_PASID_MAPPING__VALID_MASK;
-}
-
-static uint16_t get_atc_vmid_pasid_mapping_pasid(struct kgd_dev *kgd,
-								uint8_t vmid)
-{
-	uint32_t reg;
-	struct amdgpu_device *adev = (struct amdgpu_device *) kgd;
+	*p_pasid = value & ATC_VMID0_PASID_MAPPING__PASID_MASK;
 
-	reg = RREG32(SOC15_REG_OFFSET(ATHUB, 0, mmATC_VMID0_PASID_MAPPING)
-		     + vmid);
-	return reg & ATC_VMID0_PASID_MAPPING__PASID_MASK;
+	return !!(value & ATC_VMID0_PASID_MAPPING__VALID_MASK);
 }
 
 static int invalidate_tlbs_with_kiq(struct amdgpu_device *adev, uint16_t pasid)
@@ -830,6 +715,8 @@ static int invalidate_tlbs(struct kgd_dev *kgd, uint16_t pasid)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *) kgd;
 	int vmid;
+	uint16_t queried_pasid;
+	bool ret;
 	struct amdgpu_ring *ring = &adev->gfx.kiq.ring;
 
 	if (amdgpu_emu_mode == 0 && ring->sched.ready)
@@ -838,13 +725,13 @@ static int invalidate_tlbs(struct kgd_dev *kgd, uint16_t pasid)
 	for (vmid = 0; vmid < 16; vmid++) {
 		if (!amdgpu_amdkfd_is_kfd_vmid(adev, vmid))
 			continue;
-		if (get_atc_vmid_pasid_mapping_valid(kgd, vmid)) {
-			if (get_atc_vmid_pasid_mapping_pasid(kgd, vmid)
-				== pasid) {
-				amdgpu_gmc_flush_gpu_tlb(adev, vmid,
-						AMDGPU_GFXHUB_0, 0);
-				break;
-			}
+
+		ret = get_atc_vmid_pasid_mapping_info(kgd, vmid,
+				&queried_pasid);
+		if (ret	&& queried_pasid == pasid) {
+			amdgpu_gmc_flush_gpu_tlb(adev, vmid,
+					AMDGPU_GFXHUB_0, 0);
+			break;
 		}
 	}
 
@@ -914,7 +801,6 @@ static void set_vm_context_page_table_base(struct kgd_dev *kgd, uint32_t vmid,
 		uint64_t page_table_base)
 {
 	struct amdgpu_device *adev = get_amdgpu_device(kgd);
-	uint64_t base = page_table_base | AMDGPU_PTE_VALID;
 
 	if (!amdgpu_amdkfd_is_kfd_vmid(adev, vmid)) {
 		pr_err("trying to set page table base for wrong VMID %u\n",
@@ -922,18 +808,31 @@ static void set_vm_context_page_table_base(struct kgd_dev *kgd, uint32_t vmid,
 		return;
 	}
 
-	/* TODO: take advantage of per-process address space size. For
-	 * now, all processes share the same address space size, like
-	 * on GFX8 and older.
-	 */
-	WREG32(SOC15_REG_OFFSET(GC, 0, mmGCVM_CONTEXT0_PAGE_TABLE_START_ADDR_LO32) + (vmid*2), 0);
-	WREG32(SOC15_REG_OFFSET(GC, 0, mmGCVM_CONTEXT0_PAGE_TABLE_START_ADDR_HI32) + (vmid*2), 0);
-
-	WREG32(SOC15_REG_OFFSET(GC, 0, mmGCVM_CONTEXT0_PAGE_TABLE_END_ADDR_LO32) + (vmid*2),
-			lower_32_bits(adev->vm_manager.max_pfn - 1));
-	WREG32(SOC15_REG_OFFSET(GC, 0, mmGCVM_CONTEXT0_PAGE_TABLE_END_ADDR_HI32) + (vmid*2),
-			upper_32_bits(adev->vm_manager.max_pfn - 1));
-
-	WREG32(SOC15_REG_OFFSET(GC, 0, mmGCVM_CONTEXT0_PAGE_TABLE_BASE_ADDR_LO32) + (vmid*2), lower_32_bits(base));
-	WREG32(SOC15_REG_OFFSET(GC, 0, mmGCVM_CONTEXT0_PAGE_TABLE_BASE_ADDR_HI32) + (vmid*2), upper_32_bits(base));
+	/* SDMA is on gfxhub as well for Navi1* series */
+	gfxhub_v2_0_setup_vm_pt_regs(adev, vmid, page_table_base);
 }
+
+const struct kfd2kgd_calls gfx_v10_kfd2kgd = {
+	.program_sh_mem_settings = kgd_program_sh_mem_settings,
+	.set_pasid_vmid_mapping = kgd_set_pasid_vmid_mapping,
+	.init_interrupts = kgd_init_interrupts,
+	.hqd_load = kgd_hqd_load,
+	.hqd_sdma_load = kgd_hqd_sdma_load,
+	.hqd_dump = kgd_hqd_dump,
+	.hqd_sdma_dump = kgd_hqd_sdma_dump,
+	.hqd_is_occupied = kgd_hqd_is_occupied,
+	.hqd_sdma_is_occupied = kgd_hqd_sdma_is_occupied,
+	.hqd_destroy = kgd_hqd_destroy,
+	.hqd_sdma_destroy = kgd_hqd_sdma_destroy,
+	.address_watch_disable = kgd_address_watch_disable,
+	.address_watch_execute = kgd_address_watch_execute,
+	.wave_control_execute = kgd_wave_control_execute,
+	.address_watch_get_offset = kgd_address_watch_get_offset,
+	.get_atc_vmid_pasid_mapping_info =
+			get_atc_vmid_pasid_mapping_info,
+	.get_tile_config = amdgpu_amdkfd_get_tile_config,
+	.set_vm_context_page_table_base = set_vm_context_page_table_base,
+	.invalidate_tlbs = invalidate_tlbs,
+	.invalidate_tlbs_vmid = invalidate_tlbs_vmid,
+	.get_hive_id = amdgpu_amdkfd_get_hive_id,
+};
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c
index 5f459bf5f622..6e6f0a99ec06 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c
@@ -20,8 +20,6 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  */
 
-#include <linux/fdtable.h>
-#include <linux/uaccess.h>
 #include <linux/mmu_context.h>
 
 #include "amdgpu.h"
@@ -86,65 +84,6 @@ union TCP_WATCH_CNTL_BITS {
 	float f32All;
 };
 
-/*
- * Register access functions
- */
-
-static void kgd_program_sh_mem_settings(struct kgd_dev *kgd, uint32_t vmid,
-		uint32_t sh_mem_config,	uint32_t sh_mem_ape1_base,
-		uint32_t sh_mem_ape1_limit, uint32_t sh_mem_bases);
-
-static int kgd_set_pasid_vmid_mapping(struct kgd_dev *kgd, unsigned int pasid,
-					unsigned int vmid);
-
-static int kgd_init_interrupts(struct kgd_dev *kgd, uint32_t pipe_id);
-static int kgd_hqd_load(struct kgd_dev *kgd, void *mqd, uint32_t pipe_id,
-			uint32_t queue_id, uint32_t __user *wptr,
-			uint32_t wptr_shift, uint32_t wptr_mask,
-			struct mm_struct *mm);
-static int kgd_hqd_dump(struct kgd_dev *kgd,
-			uint32_t pipe_id, uint32_t queue_id,
-			uint32_t (**dump)[2], uint32_t *n_regs);
-static int kgd_hqd_sdma_load(struct kgd_dev *kgd, void *mqd,
-			     uint32_t __user *wptr, struct mm_struct *mm);
-static int kgd_hqd_sdma_dump(struct kgd_dev *kgd,
-			     uint32_t engine_id, uint32_t queue_id,
-			     uint32_t (**dump)[2], uint32_t *n_regs);
-static bool kgd_hqd_is_occupied(struct kgd_dev *kgd, uint64_t queue_address,
-				uint32_t pipe_id, uint32_t queue_id);
-
-static int kgd_hqd_destroy(struct kgd_dev *kgd, void *mqd,
-				enum kfd_preempt_type reset_type,
-				unsigned int utimeout, uint32_t pipe_id,
-				uint32_t queue_id);
-static bool kgd_hqd_sdma_is_occupied(struct kgd_dev *kgd, void *mqd);
-static int kgd_hqd_sdma_destroy(struct kgd_dev *kgd, void *mqd,
-				unsigned int utimeout);
-static int kgd_address_watch_disable(struct kgd_dev *kgd);
-static int kgd_address_watch_execute(struct kgd_dev *kgd,
-					unsigned int watch_point_id,
-					uint32_t cntl_val,
-					uint32_t addr_hi,
-					uint32_t addr_lo);
-static int kgd_wave_control_execute(struct kgd_dev *kgd,
-					uint32_t gfx_index_val,
-					uint32_t sq_cmd);
-static uint32_t kgd_address_watch_get_offset(struct kgd_dev *kgd,
-					unsigned int watch_point_id,
-					unsigned int reg_offset);
-
-static bool get_atc_vmid_pasid_mapping_valid(struct kgd_dev *kgd, uint8_t vmid);
-static uint16_t get_atc_vmid_pasid_mapping_pasid(struct kgd_dev *kgd,
-							uint8_t vmid);
-
-static void set_scratch_backing_va(struct kgd_dev *kgd,
-					uint64_t va, uint32_t vmid);
-static void set_vm_context_page_table_base(struct kgd_dev *kgd, uint32_t vmid,
-		uint64_t page_table_base);
-static int invalidate_tlbs(struct kgd_dev *kgd, uint16_t pasid);
-static int invalidate_tlbs_vmid(struct kgd_dev *kgd, uint16_t vmid);
-static uint32_t read_vmid_from_vmfault_reg(struct kgd_dev *kgd);
-
 /* Because of REG_GET_FIELD() being used, we put this function in the
  * asic specific file.
  */
@@ -170,37 +109,6 @@ static int get_tile_config(struct kgd_dev *kgd,
 	return 0;
 }
 
-static const struct kfd2kgd_calls kfd2kgd = {
-	.program_sh_mem_settings = kgd_program_sh_mem_settings,
-	.set_pasid_vmid_mapping = kgd_set_pasid_vmid_mapping,
-	.init_interrupts = kgd_init_interrupts,
-	.hqd_load = kgd_hqd_load,
-	.hqd_sdma_load = kgd_hqd_sdma_load,
-	.hqd_dump = kgd_hqd_dump,
-	.hqd_sdma_dump = kgd_hqd_sdma_dump,
-	.hqd_is_occupied = kgd_hqd_is_occupied,
-	.hqd_sdma_is_occupied = kgd_hqd_sdma_is_occupied,
-	.hqd_destroy = kgd_hqd_destroy,
-	.hqd_sdma_destroy = kgd_hqd_sdma_destroy,
-	.address_watch_disable = kgd_address_watch_disable,
-	.address_watch_execute = kgd_address_watch_execute,
-	.wave_control_execute = kgd_wave_control_execute,
-	.address_watch_get_offset = kgd_address_watch_get_offset,
-	.get_atc_vmid_pasid_mapping_pasid = get_atc_vmid_pasid_mapping_pasid,
-	.get_atc_vmid_pasid_mapping_valid = get_atc_vmid_pasid_mapping_valid,
-	.set_scratch_backing_va = set_scratch_backing_va,
-	.get_tile_config = get_tile_config,
-	.set_vm_context_page_table_base = set_vm_context_page_table_base,
-	.invalidate_tlbs = invalidate_tlbs,
-	.invalidate_tlbs_vmid = invalidate_tlbs_vmid,
-	.read_vmid_from_vmfault_reg = read_vmid_from_vmfault_reg,
-};
-
-struct kfd2kgd_calls *amdgpu_amdkfd_gfx_7_get_functions(void)
-{
-	return (struct kfd2kgd_calls *)&kfd2kgd;
-}
-
 static inline struct amdgpu_device *get_amdgpu_device(struct kgd_dev *kgd)
 {
 	return (struct amdgpu_device *)kgd;
@@ -303,14 +211,15 @@ static int kgd_init_interrupts(struct kgd_dev *kgd, uint32_t pipe_id)
 	return 0;
 }
 
-static inline uint32_t get_sdma_base_addr(struct cik_sdma_rlc_registers *m)
+static inline uint32_t get_sdma_rlc_reg_offset(struct cik_sdma_rlc_registers *m)
 {
 	uint32_t retval;
 
 	retval = m->sdma_engine_id * SDMA1_REGISTER_OFFSET +
 			m->sdma_queue_id * KFD_CIK_SDMA_QUEUE_OFFSET;
 
-	pr_debug("sdma base address: 0x%x\n", retval);
+	pr_debug("RLC register offset for SDMA%d RLC%d: 0x%x\n",
+			m->sdma_engine_id, m->sdma_queue_id, retval);
 
 	return retval;
 }
@@ -413,60 +322,52 @@ static int kgd_hqd_sdma_load(struct kgd_dev *kgd, void *mqd,
 	struct amdgpu_device *adev = get_amdgpu_device(kgd);
 	struct cik_sdma_rlc_registers *m;
 	unsigned long end_jiffies;
-	uint32_t sdma_base_addr;
+	uint32_t sdma_rlc_reg_offset;
 	uint32_t data;
 
 	m = get_sdma_mqd(mqd);
-	sdma_base_addr = get_sdma_base_addr(m);
+	sdma_rlc_reg_offset = get_sdma_rlc_reg_offset(m);
 
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL,
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL,
 		m->sdma_rlc_rb_cntl & (~SDMA0_RLC0_RB_CNTL__RB_ENABLE_MASK));
 
 	end_jiffies = msecs_to_jiffies(2000) + jiffies;
 	while (true) {
-		data = RREG32(sdma_base_addr + mmSDMA0_RLC0_CONTEXT_STATUS);
+		data = RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_CONTEXT_STATUS);
 		if (data & SDMA0_RLC0_CONTEXT_STATUS__IDLE_MASK)
 			break;
-		if (time_after(jiffies, end_jiffies))
+		if (time_after(jiffies, end_jiffies)) {
+			pr_err("SDMA RLC not idle in %s\n", __func__);
 			return -ETIME;
+		}
 		usleep_range(500, 1000);
 	}
-	if (m->sdma_engine_id) {
-		data = RREG32(mmSDMA1_GFX_CONTEXT_CNTL);
-		data = REG_SET_FIELD(data, SDMA1_GFX_CONTEXT_CNTL,
-				RESUME_CTX, 0);
-		WREG32(mmSDMA1_GFX_CONTEXT_CNTL, data);
-	} else {
-		data = RREG32(mmSDMA0_GFX_CONTEXT_CNTL);
-		data = REG_SET_FIELD(data, SDMA0_GFX_CONTEXT_CNTL,
-				RESUME_CTX, 0);
-		WREG32(mmSDMA0_GFX_CONTEXT_CNTL, data);
-	}
 
 	data = REG_SET_FIELD(m->sdma_rlc_doorbell, SDMA0_RLC0_DOORBELL,
 			     ENABLE, 1);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_DOORBELL, data);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR, m->sdma_rlc_rb_rptr);
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_DOORBELL, data);
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_RPTR,
+				m->sdma_rlc_rb_rptr);
 
 	if (read_user_wptr(mm, wptr, data))
-		WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_WPTR, data);
+		WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_WPTR, data);
 	else
-		WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_WPTR,
+		WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_WPTR,
 		       m->sdma_rlc_rb_rptr);
 
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_VIRTUAL_ADDR,
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_VIRTUAL_ADDR,
 				m->sdma_rlc_virtual_addr);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_BASE, m->sdma_rlc_rb_base);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_BASE_HI,
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_BASE, m->sdma_rlc_rb_base);
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_BASE_HI,
 			m->sdma_rlc_rb_base_hi);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR_ADDR_LO,
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_RPTR_ADDR_LO,
 			m->sdma_rlc_rb_rptr_addr_lo);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR_ADDR_HI,
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_RPTR_ADDR_HI,
 			m->sdma_rlc_rb_rptr_addr_hi);
 
 	data = REG_SET_FIELD(m->sdma_rlc_rb_cntl, SDMA0_RLC0_RB_CNTL,
 			     RB_ENABLE, 1);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL, data);
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL, data);
 
 	return 0;
 }
@@ -524,13 +425,13 @@ static bool kgd_hqd_sdma_is_occupied(struct kgd_dev *kgd, void *mqd)
 {
 	struct amdgpu_device *adev = get_amdgpu_device(kgd);
 	struct cik_sdma_rlc_registers *m;
-	uint32_t sdma_base_addr;
+	uint32_t sdma_rlc_reg_offset;
 	uint32_t sdma_rlc_rb_cntl;
 
 	m = get_sdma_mqd(mqd);
-	sdma_base_addr = get_sdma_base_addr(m);
+	sdma_rlc_reg_offset = get_sdma_rlc_reg_offset(m);
 
-	sdma_rlc_rb_cntl = RREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL);
+	sdma_rlc_rb_cntl = RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL);
 
 	if (sdma_rlc_rb_cntl & SDMA0_RLC0_RB_CNTL__RB_ENABLE_MASK)
 		return true;
@@ -645,32 +546,34 @@ static int kgd_hqd_sdma_destroy(struct kgd_dev *kgd, void *mqd,
 {
 	struct amdgpu_device *adev = get_amdgpu_device(kgd);
 	struct cik_sdma_rlc_registers *m;
-	uint32_t sdma_base_addr;
+	uint32_t sdma_rlc_reg_offset;
 	uint32_t temp;
 	unsigned long end_jiffies = (utimeout * HZ / 1000) + jiffies;
 
 	m = get_sdma_mqd(mqd);
-	sdma_base_addr = get_sdma_base_addr(m);
+	sdma_rlc_reg_offset = get_sdma_rlc_reg_offset(m);
 
-	temp = RREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL);
+	temp = RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL);
 	temp = temp & ~SDMA0_RLC0_RB_CNTL__RB_ENABLE_MASK;
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL, temp);
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL, temp);
 
 	while (true) {
-		temp = RREG32(sdma_base_addr + mmSDMA0_RLC0_CONTEXT_STATUS);
+		temp = RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_CONTEXT_STATUS);
 		if (temp & SDMA0_RLC0_CONTEXT_STATUS__IDLE_MASK)
 			break;
-		if (time_after(jiffies, end_jiffies))
+		if (time_after(jiffies, end_jiffies)) {
+			pr_err("SDMA RLC not idle in %s\n", __func__);
 			return -ETIME;
+		}
 		usleep_range(500, 1000);
 	}
 
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_DOORBELL, 0);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL,
-		RREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL) |
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_DOORBELL, 0);
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL,
+		RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL) |
 		SDMA0_RLC0_RB_CNTL__RB_ENABLE_MASK);
 
-	m->sdma_rlc_rb_rptr = RREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR);
+	m->sdma_rlc_rb_rptr = RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_RPTR);
 
 	return 0;
 }
@@ -758,24 +661,16 @@ static uint32_t kgd_address_watch_get_offset(struct kgd_dev *kgd,
 	return watchRegs[watch_point_id * ADDRESS_WATCH_REG_MAX + reg_offset];
 }
 
-static bool get_atc_vmid_pasid_mapping_valid(struct kgd_dev *kgd,
-							uint8_t vmid)
+static bool get_atc_vmid_pasid_mapping_info(struct kgd_dev *kgd,
+					uint8_t vmid, uint16_t *p_pasid)
 {
-	uint32_t reg;
+	uint32_t value;
 	struct amdgpu_device *adev = (struct amdgpu_device *) kgd;
 
-	reg = RREG32(mmATC_VMID0_PASID_MAPPING + vmid);
-	return reg & ATC_VMID0_PASID_MAPPING__VALID_MASK;
-}
-
-static uint16_t get_atc_vmid_pasid_mapping_pasid(struct kgd_dev *kgd,
-								uint8_t vmid)
-{
-	uint32_t reg;
-	struct amdgpu_device *adev = (struct amdgpu_device *) kgd;
+	value = RREG32(mmATC_VMID0_PASID_MAPPING + vmid);
+	*p_pasid = value & ATC_VMID0_PASID_MAPPING__PASID_MASK;
 
-	reg = RREG32(mmATC_VMID0_PASID_MAPPING + vmid);
-	return reg & ATC_VMID0_PASID_MAPPING__PASID_MASK;
+	return !!(value & ATC_VMID0_PASID_MAPPING__VALID_MASK);
 }
 
 static void set_scratch_backing_va(struct kgd_dev *kgd,
@@ -855,3 +750,28 @@ static uint32_t read_vmid_from_vmfault_reg(struct kgd_dev *kgd)
 
 	return REG_GET_FIELD(status, VM_CONTEXT1_PROTECTION_FAULT_STATUS, VMID);
 }
+
+const struct kfd2kgd_calls gfx_v7_kfd2kgd = {
+	.program_sh_mem_settings = kgd_program_sh_mem_settings,
+	.set_pasid_vmid_mapping = kgd_set_pasid_vmid_mapping,
+	.init_interrupts = kgd_init_interrupts,
+	.hqd_load = kgd_hqd_load,
+	.hqd_sdma_load = kgd_hqd_sdma_load,
+	.hqd_dump = kgd_hqd_dump,
+	.hqd_sdma_dump = kgd_hqd_sdma_dump,
+	.hqd_is_occupied = kgd_hqd_is_occupied,
+	.hqd_sdma_is_occupied = kgd_hqd_sdma_is_occupied,
+	.hqd_destroy = kgd_hqd_destroy,
+	.hqd_sdma_destroy = kgd_hqd_sdma_destroy,
+	.address_watch_disable = kgd_address_watch_disable,
+	.address_watch_execute = kgd_address_watch_execute,
+	.wave_control_execute = kgd_wave_control_execute,
+	.address_watch_get_offset = kgd_address_watch_get_offset,
+	.get_atc_vmid_pasid_mapping_info = get_atc_vmid_pasid_mapping_info,
+	.set_scratch_backing_va = set_scratch_backing_va,
+	.get_tile_config = get_tile_config,
+	.set_vm_context_page_table_base = set_vm_context_page_table_base,
+	.invalidate_tlbs = invalidate_tlbs,
+	.invalidate_tlbs_vmid = invalidate_tlbs_vmid,
+	.read_vmid_from_vmfault_reg = read_vmid_from_vmfault_reg,
+};
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c
index 6d2f61449606..bfbddedb2380 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c
@@ -20,9 +20,6 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  */
 
-#include <linux/module.h>
-#include <linux/fdtable.h>
-#include <linux/uaccess.h>
 #include <linux/mmu_context.h>
 
 #include "amdgpu.h"
@@ -44,62 +41,6 @@ enum hqd_dequeue_request_type {
 	RESET_WAVES
 };
 
-/*
- * Register access functions
- */
-
-static void kgd_program_sh_mem_settings(struct kgd_dev *kgd, uint32_t vmid,
-		uint32_t sh_mem_config,
-		uint32_t sh_mem_ape1_base, uint32_t sh_mem_ape1_limit,
-		uint32_t sh_mem_bases);
-static int kgd_set_pasid_vmid_mapping(struct kgd_dev *kgd, unsigned int pasid,
-		unsigned int vmid);
-static int kgd_init_interrupts(struct kgd_dev *kgd, uint32_t pipe_id);
-static int kgd_hqd_load(struct kgd_dev *kgd, void *mqd, uint32_t pipe_id,
-			uint32_t queue_id, uint32_t __user *wptr,
-			uint32_t wptr_shift, uint32_t wptr_mask,
-			struct mm_struct *mm);
-static int kgd_hqd_dump(struct kgd_dev *kgd,
-			uint32_t pipe_id, uint32_t queue_id,
-			uint32_t (**dump)[2], uint32_t *n_regs);
-static int kgd_hqd_sdma_load(struct kgd_dev *kgd, void *mqd,
-			     uint32_t __user *wptr, struct mm_struct *mm);
-static int kgd_hqd_sdma_dump(struct kgd_dev *kgd,
-			     uint32_t engine_id, uint32_t queue_id,
-			     uint32_t (**dump)[2], uint32_t *n_regs);
-static bool kgd_hqd_is_occupied(struct kgd_dev *kgd, uint64_t queue_address,
-		uint32_t pipe_id, uint32_t queue_id);
-static bool kgd_hqd_sdma_is_occupied(struct kgd_dev *kgd, void *mqd);
-static int kgd_hqd_destroy(struct kgd_dev *kgd, void *mqd,
-				enum kfd_preempt_type reset_type,
-				unsigned int utimeout, uint32_t pipe_id,
-				uint32_t queue_id);
-static int kgd_hqd_sdma_destroy(struct kgd_dev *kgd, void *mqd,
-				unsigned int utimeout);
-static int kgd_address_watch_disable(struct kgd_dev *kgd);
-static int kgd_address_watch_execute(struct kgd_dev *kgd,
-					unsigned int watch_point_id,
-					uint32_t cntl_val,
-					uint32_t addr_hi,
-					uint32_t addr_lo);
-static int kgd_wave_control_execute(struct kgd_dev *kgd,
-					uint32_t gfx_index_val,
-					uint32_t sq_cmd);
-static uint32_t kgd_address_watch_get_offset(struct kgd_dev *kgd,
-					unsigned int watch_point_id,
-					unsigned int reg_offset);
-
-static bool get_atc_vmid_pasid_mapping_valid(struct kgd_dev *kgd,
-		uint8_t vmid);
-static uint16_t get_atc_vmid_pasid_mapping_pasid(struct kgd_dev *kgd,
-		uint8_t vmid);
-static void set_scratch_backing_va(struct kgd_dev *kgd,
-					uint64_t va, uint32_t vmid);
-static void set_vm_context_page_table_base(struct kgd_dev *kgd, uint32_t vmid,
-		uint64_t page_table_base);
-static int invalidate_tlbs(struct kgd_dev *kgd, uint16_t pasid);
-static int invalidate_tlbs_vmid(struct kgd_dev *kgd, uint16_t vmid);
-
 /* Because of REG_GET_FIELD() being used, we put this function in the
  * asic specific file.
  */
@@ -125,38 +66,6 @@ static int get_tile_config(struct kgd_dev *kgd,
 	return 0;
 }
 
-static const struct kfd2kgd_calls kfd2kgd = {
-	.program_sh_mem_settings = kgd_program_sh_mem_settings,
-	.set_pasid_vmid_mapping = kgd_set_pasid_vmid_mapping,
-	.init_interrupts = kgd_init_interrupts,
-	.hqd_load = kgd_hqd_load,
-	.hqd_sdma_load = kgd_hqd_sdma_load,
-	.hqd_dump = kgd_hqd_dump,
-	.hqd_sdma_dump = kgd_hqd_sdma_dump,
-	.hqd_is_occupied = kgd_hqd_is_occupied,
-	.hqd_sdma_is_occupied = kgd_hqd_sdma_is_occupied,
-	.hqd_destroy = kgd_hqd_destroy,
-	.hqd_sdma_destroy = kgd_hqd_sdma_destroy,
-	.address_watch_disable = kgd_address_watch_disable,
-	.address_watch_execute = kgd_address_watch_execute,
-	.wave_control_execute = kgd_wave_control_execute,
-	.address_watch_get_offset = kgd_address_watch_get_offset,
-	.get_atc_vmid_pasid_mapping_pasid =
-			get_atc_vmid_pasid_mapping_pasid,
-	.get_atc_vmid_pasid_mapping_valid =
-			get_atc_vmid_pasid_mapping_valid,
-	.set_scratch_backing_va = set_scratch_backing_va,
-	.get_tile_config = get_tile_config,
-	.set_vm_context_page_table_base = set_vm_context_page_table_base,
-	.invalidate_tlbs = invalidate_tlbs,
-	.invalidate_tlbs_vmid = invalidate_tlbs_vmid,
-};
-
-struct kfd2kgd_calls *amdgpu_amdkfd_gfx_8_0_get_functions(void)
-{
-	return (struct kfd2kgd_calls *)&kfd2kgd;
-}
-
 static inline struct amdgpu_device *get_amdgpu_device(struct kgd_dev *kgd)
 {
 	return (struct amdgpu_device *)kgd;
@@ -260,13 +169,15 @@ static int kgd_init_interrupts(struct kgd_dev *kgd, uint32_t pipe_id)
 	return 0;
 }
 
-static inline uint32_t get_sdma_base_addr(struct vi_sdma_mqd *m)
+static inline uint32_t get_sdma_rlc_reg_offset(struct vi_sdma_mqd *m)
 {
 	uint32_t retval;
 
 	retval = m->sdma_engine_id * SDMA1_REGISTER_OFFSET +
 		m->sdma_queue_id * KFD_VI_SDMA_QUEUE_OFFSET;
-	pr_debug("sdma base address: 0x%x\n", retval);
+
+	pr_debug("RLC register offset for SDMA%d RLC%d: 0x%x\n",
+			m->sdma_engine_id, m->sdma_queue_id, retval);
 
 	return retval;
 }
@@ -398,59 +309,51 @@ static int kgd_hqd_sdma_load(struct kgd_dev *kgd, void *mqd,
 	struct amdgpu_device *adev = get_amdgpu_device(kgd);
 	struct vi_sdma_mqd *m;
 	unsigned long end_jiffies;
-	uint32_t sdma_base_addr;
+	uint32_t sdma_rlc_reg_offset;
 	uint32_t data;
 
 	m = get_sdma_mqd(mqd);
-	sdma_base_addr = get_sdma_base_addr(m);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL,
+	sdma_rlc_reg_offset = get_sdma_rlc_reg_offset(m);
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL,
 		m->sdmax_rlcx_rb_cntl & (~SDMA0_RLC0_RB_CNTL__RB_ENABLE_MASK));
 
 	end_jiffies = msecs_to_jiffies(2000) + jiffies;
 	while (true) {
-		data = RREG32(sdma_base_addr + mmSDMA0_RLC0_CONTEXT_STATUS);
+		data = RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_CONTEXT_STATUS);
 		if (data & SDMA0_RLC0_CONTEXT_STATUS__IDLE_MASK)
 			break;
-		if (time_after(jiffies, end_jiffies))
+		if (time_after(jiffies, end_jiffies)) {
+			pr_err("SDMA RLC not idle in %s\n", __func__);
 			return -ETIME;
+		}
 		usleep_range(500, 1000);
 	}
-	if (m->sdma_engine_id) {
-		data = RREG32(mmSDMA1_GFX_CONTEXT_CNTL);
-		data = REG_SET_FIELD(data, SDMA1_GFX_CONTEXT_CNTL,
-				RESUME_CTX, 0);
-		WREG32(mmSDMA1_GFX_CONTEXT_CNTL, data);
-	} else {
-		data = RREG32(mmSDMA0_GFX_CONTEXT_CNTL);
-		data = REG_SET_FIELD(data, SDMA0_GFX_CONTEXT_CNTL,
-				RESUME_CTX, 0);
-		WREG32(mmSDMA0_GFX_CONTEXT_CNTL, data);
-	}
 
 	data = REG_SET_FIELD(m->sdmax_rlcx_doorbell, SDMA0_RLC0_DOORBELL,
 			     ENABLE, 1);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_DOORBELL, data);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR, m->sdmax_rlcx_rb_rptr);
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_DOORBELL, data);
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_RPTR,
+				m->sdmax_rlcx_rb_rptr);
 
 	if (read_user_wptr(mm, wptr, data))
-		WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_WPTR, data);
+		WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_WPTR, data);
 	else
-		WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_WPTR,
+		WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_WPTR,
 		       m->sdmax_rlcx_rb_rptr);
 
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_VIRTUAL_ADDR,
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_VIRTUAL_ADDR,
 				m->sdmax_rlcx_virtual_addr);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_BASE, m->sdmax_rlcx_rb_base);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_BASE_HI,
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_BASE, m->sdmax_rlcx_rb_base);
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_BASE_HI,
 			m->sdmax_rlcx_rb_base_hi);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR_ADDR_LO,
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_RPTR_ADDR_LO,
 			m->sdmax_rlcx_rb_rptr_addr_lo);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR_ADDR_HI,
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_RPTR_ADDR_HI,
 			m->sdmax_rlcx_rb_rptr_addr_hi);
 
 	data = REG_SET_FIELD(m->sdmax_rlcx_rb_cntl, SDMA0_RLC0_RB_CNTL,
 			     RB_ENABLE, 1);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL, data);
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL, data);
 
 	return 0;
 }
@@ -517,13 +420,13 @@ static bool kgd_hqd_sdma_is_occupied(struct kgd_dev *kgd, void *mqd)
 {
 	struct amdgpu_device *adev = get_amdgpu_device(kgd);
 	struct vi_sdma_mqd *m;
-	uint32_t sdma_base_addr;
+	uint32_t sdma_rlc_reg_offset;
 	uint32_t sdma_rlc_rb_cntl;
 
 	m = get_sdma_mqd(mqd);
-	sdma_base_addr = get_sdma_base_addr(m);
+	sdma_rlc_reg_offset = get_sdma_rlc_reg_offset(m);
 
-	sdma_rlc_rb_cntl = RREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL);
+	sdma_rlc_rb_cntl = RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL);
 
 	if (sdma_rlc_rb_cntl & SDMA0_RLC0_RB_CNTL__RB_ENABLE_MASK)
 		return true;
@@ -641,54 +544,48 @@ static int kgd_hqd_sdma_destroy(struct kgd_dev *kgd, void *mqd,
 {
 	struct amdgpu_device *adev = get_amdgpu_device(kgd);
 	struct vi_sdma_mqd *m;
-	uint32_t sdma_base_addr;
+	uint32_t sdma_rlc_reg_offset;
 	uint32_t temp;
 	unsigned long end_jiffies = (utimeout * HZ / 1000) + jiffies;
 
 	m = get_sdma_mqd(mqd);
-	sdma_base_addr = get_sdma_base_addr(m);
+	sdma_rlc_reg_offset = get_sdma_rlc_reg_offset(m);
 
-	temp = RREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL);
+	temp = RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL);
 	temp = temp & ~SDMA0_RLC0_RB_CNTL__RB_ENABLE_MASK;
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL, temp);
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL, temp);
 
 	while (true) {
-		temp = RREG32(sdma_base_addr + mmSDMA0_RLC0_CONTEXT_STATUS);
+		temp = RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_CONTEXT_STATUS);
 		if (temp & SDMA0_RLC0_CONTEXT_STATUS__IDLE_MASK)
 			break;
-		if (time_after(jiffies, end_jiffies))
+		if (time_after(jiffies, end_jiffies)) {
+			pr_err("SDMA RLC not idle in %s\n", __func__);
 			return -ETIME;
+		}
 		usleep_range(500, 1000);
 	}
 
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_DOORBELL, 0);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL,
-		RREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL) |
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_DOORBELL, 0);
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL,
+		RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL) |
 		SDMA0_RLC0_RB_CNTL__RB_ENABLE_MASK);
 
-	m->sdmax_rlcx_rb_rptr = RREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR);
+	m->sdmax_rlcx_rb_rptr = RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_RPTR);
 
 	return 0;
 }
 
-static bool get_atc_vmid_pasid_mapping_valid(struct kgd_dev *kgd,
-							uint8_t vmid)
+static bool get_atc_vmid_pasid_mapping_info(struct kgd_dev *kgd,
+					uint8_t vmid, uint16_t *p_pasid)
 {
-	uint32_t reg;
+	uint32_t value;
 	struct amdgpu_device *adev = (struct amdgpu_device *) kgd;
 
-	reg = RREG32(mmATC_VMID0_PASID_MAPPING + vmid);
-	return reg & ATC_VMID0_PASID_MAPPING__VALID_MASK;
-}
+	value = RREG32(mmATC_VMID0_PASID_MAPPING + vmid);
+	*p_pasid = value & ATC_VMID0_PASID_MAPPING__PASID_MASK;
 
-static uint16_t get_atc_vmid_pasid_mapping_pasid(struct kgd_dev *kgd,
-								uint8_t vmid)
-{
-	uint32_t reg;
-	struct amdgpu_device *adev = (struct amdgpu_device *) kgd;
-
-	reg = RREG32(mmATC_VMID0_PASID_MAPPING + vmid);
-	return reg & ATC_VMID0_PASID_MAPPING__PASID_MASK;
+	return !!(value & ATC_VMID0_PASID_MAPPING__VALID_MASK);
 }
 
 static int kgd_address_watch_disable(struct kgd_dev *kgd)
@@ -798,3 +695,28 @@ static int invalidate_tlbs_vmid(struct kgd_dev *kgd, uint16_t vmid)
 	RREG32(mmVM_INVALIDATE_RESPONSE);
 	return 0;
 }
+
+const struct kfd2kgd_calls gfx_v8_kfd2kgd = {
+	.program_sh_mem_settings = kgd_program_sh_mem_settings,
+	.set_pasid_vmid_mapping = kgd_set_pasid_vmid_mapping,
+	.init_interrupts = kgd_init_interrupts,
+	.hqd_load = kgd_hqd_load,
+	.hqd_sdma_load = kgd_hqd_sdma_load,
+	.hqd_dump = kgd_hqd_dump,
+	.hqd_sdma_dump = kgd_hqd_sdma_dump,
+	.hqd_is_occupied = kgd_hqd_is_occupied,
+	.hqd_sdma_is_occupied = kgd_hqd_sdma_is_occupied,
+	.hqd_destroy = kgd_hqd_destroy,
+	.hqd_sdma_destroy = kgd_hqd_sdma_destroy,
+	.address_watch_disable = kgd_address_watch_disable,
+	.address_watch_execute = kgd_address_watch_execute,
+	.wave_control_execute = kgd_wave_control_execute,
+	.address_watch_get_offset = kgd_address_watch_get_offset,
+	.get_atc_vmid_pasid_mapping_info =
+			get_atc_vmid_pasid_mapping_info,
+	.set_scratch_backing_va = set_scratch_backing_va,
+	.get_tile_config = get_tile_config,
+	.set_vm_context_page_table_base = set_vm_context_page_table_base,
+	.invalidate_tlbs = invalidate_tlbs,
+	.invalidate_tlbs_vmid = invalidate_tlbs_vmid,
+};
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c
index e262f2ac07a3..47c853ef1051 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c
@@ -19,17 +19,10 @@
  * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
  * OTHER DEALINGS IN THE SOFTWARE.
  */
-
-#define pr_fmt(fmt) "kfd2kgd: " fmt
-
-#include <linux/module.h>
-#include <linux/fdtable.h>
-#include <linux/uaccess.h>
 #include <linux/mmu_context.h>
 
 #include "amdgpu.h"
 #include "amdgpu_amdkfd.h"
-#include "soc15_hw_ip.h"
 #include "gc/gc_9_0_offset.h"
 #include "gc/gc_9_0_sh_mask.h"
 #include "vega10_enum.h"
@@ -50,9 +43,6 @@
 #include "gmc_v9_0.h"
 
 
-#define V9_PIPE_PER_MEC		(4)
-#define V9_QUEUES_PER_PIPE_MEC	(8)
-
 enum hqd_dequeue_request_type {
 	NO_ACTION = 0,
 	DRAIN_PIPE,
@@ -226,22 +216,21 @@ int kgd_gfx_v9_init_interrupts(struct kgd_dev *kgd, uint32_t pipe_id)
 	return 0;
 }
 
-static uint32_t get_sdma_base_addr(struct amdgpu_device *adev,
+static uint32_t get_sdma_rlc_reg_offset(struct amdgpu_device *adev,
 				unsigned int engine_id,
 				unsigned int queue_id)
 {
-	uint32_t base[2] = {
+	uint32_t sdma_engine_reg_base[2] = {
 		SOC15_REG_OFFSET(SDMA0, 0,
 				 mmSDMA0_RLC0_RB_CNTL) - mmSDMA0_RLC0_RB_CNTL,
 		SOC15_REG_OFFSET(SDMA1, 0,
 				 mmSDMA1_RLC0_RB_CNTL) - mmSDMA1_RLC0_RB_CNTL
 	};
-	uint32_t retval;
+	uint32_t retval = sdma_engine_reg_base[engine_id]
+		+ queue_id * (mmSDMA0_RLC1_RB_CNTL - mmSDMA0_RLC0_RB_CNTL);
 
-	retval = base[engine_id] + queue_id * (mmSDMA0_RLC1_RB_CNTL -
-					       mmSDMA0_RLC0_RB_CNTL);
-
-	pr_debug("sdma base address: 0x%x\n", retval);
+	pr_debug("RLC register offset for SDMA%d RLC%d: 0x%x\n", engine_id,
+			queue_id, retval);
 
 	return retval;
 }
@@ -388,71 +377,67 @@ static int kgd_hqd_sdma_load(struct kgd_dev *kgd, void *mqd,
 {
 	struct amdgpu_device *adev = get_amdgpu_device(kgd);
 	struct v9_sdma_mqd *m;
-	uint32_t sdma_base_addr, sdmax_gfx_context_cntl;
+	uint32_t sdma_rlc_reg_offset;
 	unsigned long end_jiffies;
 	uint32_t data;
 	uint64_t data64;
 	uint64_t __user *wptr64 = (uint64_t __user *)wptr;
 
 	m = get_sdma_mqd(mqd);
-	sdma_base_addr = get_sdma_base_addr(adev, m->sdma_engine_id,
+	sdma_rlc_reg_offset = get_sdma_rlc_reg_offset(adev, m->sdma_engine_id,
 					    m->sdma_queue_id);
-	sdmax_gfx_context_cntl = m->sdma_engine_id ?
-		SOC15_REG_OFFSET(SDMA1, 0, mmSDMA1_GFX_CONTEXT_CNTL) :
-		SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_GFX_CONTEXT_CNTL);
 
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL,
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL,
 		m->sdmax_rlcx_rb_cntl & (~SDMA0_RLC0_RB_CNTL__RB_ENABLE_MASK));
 
 	end_jiffies = msecs_to_jiffies(2000) + jiffies;
 	while (true) {
-		data = RREG32(sdma_base_addr + mmSDMA0_RLC0_CONTEXT_STATUS);
+		data = RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_CONTEXT_STATUS);
 		if (data & SDMA0_RLC0_CONTEXT_STATUS__IDLE_MASK)
 			break;
-		if (time_after(jiffies, end_jiffies))
+		if (time_after(jiffies, end_jiffies)) {
+			pr_err("SDMA RLC not idle in %s\n", __func__);
 			return -ETIME;
+		}
 		usleep_range(500, 1000);
 	}
-	data = RREG32(sdmax_gfx_context_cntl);
-	data = REG_SET_FIELD(data, SDMA0_GFX_CONTEXT_CNTL,
-			     RESUME_CTX, 0);
-	WREG32(sdmax_gfx_context_cntl, data);
 
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_DOORBELL_OFFSET,
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_DOORBELL_OFFSET,
 	       m->sdmax_rlcx_doorbell_offset);
 
 	data = REG_SET_FIELD(m->sdmax_rlcx_doorbell, SDMA0_RLC0_DOORBELL,
 			     ENABLE, 1);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_DOORBELL, data);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR, m->sdmax_rlcx_rb_rptr);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR_HI,
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_DOORBELL, data);
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_RPTR,
+				m->sdmax_rlcx_rb_rptr);
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_RPTR_HI,
 				m->sdmax_rlcx_rb_rptr_hi);
 
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_MINOR_PTR_UPDATE, 1);
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_MINOR_PTR_UPDATE, 1);
 	if (read_user_wptr(mm, wptr64, data64)) {
-		WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_WPTR,
+		WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_WPTR,
 		       lower_32_bits(data64));
-		WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_WPTR_HI,
+		WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_WPTR_HI,
 		       upper_32_bits(data64));
 	} else {
-		WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_WPTR,
+		WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_WPTR,
 		       m->sdmax_rlcx_rb_rptr);
-		WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_WPTR_HI,
+		WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_WPTR_HI,
 		       m->sdmax_rlcx_rb_rptr_hi);
 	}
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_MINOR_PTR_UPDATE, 0);
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_MINOR_PTR_UPDATE, 0);
 
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_BASE, m->sdmax_rlcx_rb_base);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_BASE_HI,
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_BASE, m->sdmax_rlcx_rb_base);
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_BASE_HI,
 			m->sdmax_rlcx_rb_base_hi);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR_ADDR_LO,
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_RPTR_ADDR_LO,
 			m->sdmax_rlcx_rb_rptr_addr_lo);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR_ADDR_HI,
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_RPTR_ADDR_HI,
 			m->sdmax_rlcx_rb_rptr_addr_hi);
 
 	data = REG_SET_FIELD(m->sdmax_rlcx_rb_cntl, SDMA0_RLC0_RB_CNTL,
 			     RB_ENABLE, 1);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL, data);
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL, data);
 
 	return 0;
 }
@@ -462,7 +447,8 @@ static int kgd_hqd_sdma_dump(struct kgd_dev *kgd,
 			     uint32_t (**dump)[2], uint32_t *n_regs)
 {
 	struct amdgpu_device *adev = get_amdgpu_device(kgd);
-	uint32_t sdma_base_addr = get_sdma_base_addr(adev, engine_id, queue_id);
+	uint32_t sdma_rlc_reg_offset = get_sdma_rlc_reg_offset(adev,
+			engine_id, queue_id);
 	uint32_t i = 0, reg;
 #undef HQD_N_REGS
 #define HQD_N_REGS (19+6+7+10)
@@ -472,15 +458,15 @@ static int kgd_hqd_sdma_dump(struct kgd_dev *kgd,
 		return -ENOMEM;
 
 	for (reg = mmSDMA0_RLC0_RB_CNTL; reg <= mmSDMA0_RLC0_DOORBELL; reg++)
-		DUMP_REG(sdma_base_addr + reg);
+		DUMP_REG(sdma_rlc_reg_offset + reg);
 	for (reg = mmSDMA0_RLC0_STATUS; reg <= mmSDMA0_RLC0_CSA_ADDR_HI; reg++)
-		DUMP_REG(sdma_base_addr + reg);
+		DUMP_REG(sdma_rlc_reg_offset + reg);
 	for (reg = mmSDMA0_RLC0_IB_SUB_REMAIN;
 	     reg <= mmSDMA0_RLC0_MINOR_PTR_UPDATE; reg++)
-		DUMP_REG(sdma_base_addr + reg);
+		DUMP_REG(sdma_rlc_reg_offset + reg);
 	for (reg = mmSDMA0_RLC0_MIDCMD_DATA0;
 	     reg <= mmSDMA0_RLC0_MIDCMD_CNTL; reg++)
-		DUMP_REG(sdma_base_addr + reg);
+		DUMP_REG(sdma_rlc_reg_offset + reg);
 
 	WARN_ON_ONCE(i != HQD_N_REGS);
 	*n_regs = i;
@@ -514,14 +500,14 @@ static bool kgd_hqd_sdma_is_occupied(struct kgd_dev *kgd, void *mqd)
 {
 	struct amdgpu_device *adev = get_amdgpu_device(kgd);
 	struct v9_sdma_mqd *m;
-	uint32_t sdma_base_addr;
+	uint32_t sdma_rlc_reg_offset;
 	uint32_t sdma_rlc_rb_cntl;
 
 	m = get_sdma_mqd(mqd);
-	sdma_base_addr = get_sdma_base_addr(adev, m->sdma_engine_id,
+	sdma_rlc_reg_offset = get_sdma_rlc_reg_offset(adev, m->sdma_engine_id,
 					    m->sdma_queue_id);
 
-	sdma_rlc_rb_cntl = RREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL);
+	sdma_rlc_rb_cntl = RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL);
 
 	if (sdma_rlc_rb_cntl & SDMA0_RLC0_RB_CNTL__RB_ENABLE_MASK)
 		return true;
@@ -584,59 +570,52 @@ static int kgd_hqd_sdma_destroy(struct kgd_dev *kgd, void *mqd,
 {
 	struct amdgpu_device *adev = get_amdgpu_device(kgd);
 	struct v9_sdma_mqd *m;
-	uint32_t sdma_base_addr;
+	uint32_t sdma_rlc_reg_offset;
 	uint32_t temp;
 	unsigned long end_jiffies = (utimeout * HZ / 1000) + jiffies;
 
 	m = get_sdma_mqd(mqd);
-	sdma_base_addr = get_sdma_base_addr(adev, m->sdma_engine_id,
+	sdma_rlc_reg_offset = get_sdma_rlc_reg_offset(adev, m->sdma_engine_id,
 					    m->sdma_queue_id);
 
-	temp = RREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL);
+	temp = RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL);
 	temp = temp & ~SDMA0_RLC0_RB_CNTL__RB_ENABLE_MASK;
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL, temp);
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL, temp);
 
 	while (true) {
-		temp = RREG32(sdma_base_addr + mmSDMA0_RLC0_CONTEXT_STATUS);
+		temp = RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_CONTEXT_STATUS);
 		if (temp & SDMA0_RLC0_CONTEXT_STATUS__IDLE_MASK)
 			break;
-		if (time_after(jiffies, end_jiffies))
+		if (time_after(jiffies, end_jiffies)) {
+			pr_err("SDMA RLC not idle in %s\n", __func__);
 			return -ETIME;
+		}
 		usleep_range(500, 1000);
 	}
 
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_DOORBELL, 0);
-	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL,
-		RREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL) |
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_DOORBELL, 0);
+	WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL,
+		RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL) |
 		SDMA0_RLC0_RB_CNTL__RB_ENABLE_MASK);
 
-	m->sdmax_rlcx_rb_rptr = RREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR);
+	m->sdmax_rlcx_rb_rptr = RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_RPTR);
 	m->sdmax_rlcx_rb_rptr_hi =
-		RREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR_HI);
+		RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_RPTR_HI);
 
 	return 0;
 }
 
-bool kgd_gfx_v9_get_atc_vmid_pasid_mapping_valid(struct kgd_dev *kgd,
-							uint8_t vmid)
+bool kgd_gfx_v9_get_atc_vmid_pasid_mapping_info(struct kgd_dev *kgd,
+					uint8_t vmid, uint16_t *p_pasid)
 {
-	uint32_t reg;
+	uint32_t value;
 	struct amdgpu_device *adev = (struct amdgpu_device *) kgd;
 
-	reg = RREG32(SOC15_REG_OFFSET(ATHUB, 0, mmATC_VMID0_PASID_MAPPING)
+	value = RREG32(SOC15_REG_OFFSET(ATHUB, 0, mmATC_VMID0_PASID_MAPPING)
 		     + vmid);
-	return reg & ATC_VMID0_PASID_MAPPING__VALID_MASK;
-}
-
-uint16_t kgd_gfx_v9_get_atc_vmid_pasid_mapping_pasid(struct kgd_dev *kgd,
-								uint8_t vmid)
-{
-	uint32_t reg;
-	struct amdgpu_device *adev = (struct amdgpu_device *) kgd;
+	*p_pasid = value & ATC_VMID0_PASID_MAPPING__PASID_MASK;
 
-	reg = RREG32(SOC15_REG_OFFSET(ATHUB, 0, mmATC_VMID0_PASID_MAPPING)
-		     + vmid);
-	return reg & ATC_VMID0_PASID_MAPPING__PASID_MASK;
+	return !!(value & ATC_VMID0_PASID_MAPPING__VALID_MASK);
 }
 
 static int invalidate_tlbs_with_kiq(struct amdgpu_device *adev, uint16_t pasid,
@@ -671,6 +650,8 @@ int kgd_gfx_v9_invalidate_tlbs(struct kgd_dev *kgd, uint16_t pasid)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *) kgd;
 	int vmid, i;
+	uint16_t queried_pasid;
+	bool ret;
 	struct amdgpu_ring *ring = &adev->gfx.kiq.ring;
 	uint32_t flush_type = 0;
 
@@ -686,14 +667,14 @@ int kgd_gfx_v9_invalidate_tlbs(struct kgd_dev *kgd, uint16_t pasid)
 	for (vmid = 0; vmid < 16; vmid++) {
 		if (!amdgpu_amdkfd_is_kfd_vmid(adev, vmid))
 			continue;
-		if (kgd_gfx_v9_get_atc_vmid_pasid_mapping_valid(kgd, vmid)) {
-			if (kgd_gfx_v9_get_atc_vmid_pasid_mapping_pasid(kgd, vmid)
-				== pasid) {
-				for (i = 0; i < adev->num_vmhubs; i++)
-					amdgpu_gmc_flush_gpu_tlb(adev, vmid,
-								i, flush_type);
-				break;
-			}
+
+		ret = kgd_gfx_v9_get_atc_vmid_pasid_mapping_info(kgd, vmid,
+				&queried_pasid);
+		if (ret && queried_pasid == pasid) {
+			for (i = 0; i < adev->num_vmhubs; i++)
+				amdgpu_gmc_flush_gpu_tlb(adev, vmid,
+							i, flush_type);
+			break;
 		}
 	}
 
@@ -777,15 +758,6 @@ uint32_t kgd_gfx_v9_address_watch_get_offset(struct kgd_dev *kgd,
 	return 0;
 }
 
-void kgd_gfx_v9_set_scratch_backing_va(struct kgd_dev *kgd,
-					uint64_t va, uint32_t vmid)
-{
-	/* No longer needed on GFXv9. The scratch base address is
-	 * passed to the shader by the CP. It's the user mode driver's
-	 * responsibility.
-	 */
-}
-
 void kgd_gfx_v9_set_vm_context_page_table_base(struct kgd_dev *kgd, uint32_t vmid,
 		uint64_t page_table_base)
 {
@@ -811,7 +783,7 @@ void kgd_gfx_v9_set_vm_context_page_table_base(struct kgd_dev *kgd, uint32_t vmi
 	gfxhub_v1_0_setup_vm_pt_regs(adev, vmid, page_table_base);
 }
 
-static const struct kfd2kgd_calls kfd2kgd = {
+const struct kfd2kgd_calls gfx_v9_kfd2kgd = {
 	.program_sh_mem_settings = kgd_gfx_v9_program_sh_mem_settings,
 	.set_pasid_vmid_mapping = kgd_gfx_v9_set_pasid_vmid_mapping,
 	.init_interrupts = kgd_gfx_v9_init_interrupts,
@@ -827,19 +799,11 @@ static const struct kfd2kgd_calls kfd2kgd = {
 	.address_watch_execute = kgd_gfx_v9_address_watch_execute,
 	.wave_control_execute = kgd_gfx_v9_wave_control_execute,
 	.address_watch_get_offset = kgd_gfx_v9_address_watch_get_offset,
-	.get_atc_vmid_pasid_mapping_pasid =
-			kgd_gfx_v9_get_atc_vmid_pasid_mapping_pasid,
-	.get_atc_vmid_pasid_mapping_valid =
-			kgd_gfx_v9_get_atc_vmid_pasid_mapping_valid,
-	.set_scratch_backing_va = kgd_gfx_v9_set_scratch_backing_va,
+	.get_atc_vmid_pasid_mapping_info =
+			kgd_gfx_v9_get_atc_vmid_pasid_mapping_info,
 	.get_tile_config = kgd_gfx_v9_get_tile_config,
 	.set_vm_context_page_table_base = kgd_gfx_v9_set_vm_context_page_table_base,
 	.invalidate_tlbs = kgd_gfx_v9_invalidate_tlbs,
 	.invalidate_tlbs_vmid = kgd_gfx_v9_invalidate_tlbs_vmid,
 	.get_hive_id = amdgpu_amdkfd_get_hive_id,
 };
-
-struct kfd2kgd_calls *amdgpu_amdkfd_gfx_9_0_get_functions(void)
-{
-	return (struct kfd2kgd_calls *)&kfd2kgd;
-}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.h
index 26d8879bff9d..d9e9ad22b2bd 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.h
@@ -55,14 +55,10 @@ uint32_t kgd_gfx_v9_address_watch_get_offset(struct kgd_dev *kgd,
 					unsigned int watch_point_id,
 					unsigned int reg_offset);
 
-bool kgd_gfx_v9_get_atc_vmid_pasid_mapping_valid(struct kgd_dev *kgd,
-		uint8_t vmid);
-uint16_t kgd_gfx_v9_get_atc_vmid_pasid_mapping_pasid(struct kgd_dev *kgd,
-		uint8_t vmid);
+bool kgd_gfx_v9_get_atc_vmid_pasid_mapping_info(struct kgd_dev *kgd,
+					uint8_t vmid, uint16_t *p_pasid);
 void kgd_gfx_v9_set_vm_context_page_table_base(struct kgd_dev *kgd, uint32_t vmid,
 		uint64_t page_table_base);
-void kgd_gfx_v9_set_scratch_backing_va(struct kgd_dev *kgd,
-					uint64_t va, uint32_t vmid);
 int kgd_gfx_v9_invalidate_tlbs(struct kgd_dev *kgd, uint16_t pasid);
 int kgd_gfx_v9_invalidate_tlbs_vmid(struct kgd_dev *kgd, uint16_t vmid);
 int kgd_gfx_v9_get_tile_config(struct kgd_dev *kgd,
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
index 6d021ecc8d59..7d35b5b66229 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
@@ -19,9 +19,6 @@
  * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
  * OTHER DEALINGS IN THE SOFTWARE.
  */
-
-#define pr_fmt(fmt) "kfd2kgd: " fmt
-
 #include <linux/dma-buf.h>
 #include <linux/list.h>
 #include <linux/pagemap.h>
@@ -33,11 +30,6 @@
 #include "amdgpu_amdkfd.h"
 #include "amdgpu_dma_buf.h"
 
-/* Special VM and GART address alignment needed for VI pre-Fiji due to
- * a HW bug.
- */
-#define VI_BO_SIZE_ALIGN (0x8000)
-
 /* BO flag to indicate a KFD userptr BO */
 #define AMDGPU_AMDKFD_USERPTR_BO (1ULL << 63)
 
@@ -349,13 +341,46 @@ static int vm_update_pds(struct amdgpu_vm *vm, struct amdgpu_sync *sync)
 	struct amdgpu_device *adev = amdgpu_ttm_adev(pd->tbo.bdev);
 	int ret;
 
-	ret = amdgpu_vm_update_directories(adev, vm);
+	ret = amdgpu_vm_update_pdes(adev, vm, false);
 	if (ret)
 		return ret;
 
 	return amdgpu_sync_fence(NULL, sync, vm->last_update, false);
 }
 
+static uint64_t get_pte_flags(struct amdgpu_device *adev, struct kgd_mem *mem)
+{
+	struct amdgpu_device *bo_adev = amdgpu_ttm_adev(mem->bo->tbo.bdev);
+	bool coherent = mem->alloc_flags & ALLOC_MEM_FLAGS_COHERENT;
+	uint32_t mapping_flags;
+
+	mapping_flags = AMDGPU_VM_PAGE_READABLE;
+	if (mem->alloc_flags & ALLOC_MEM_FLAGS_WRITABLE)
+		mapping_flags |= AMDGPU_VM_PAGE_WRITEABLE;
+	if (mem->alloc_flags & ALLOC_MEM_FLAGS_EXECUTABLE)
+		mapping_flags |= AMDGPU_VM_PAGE_EXECUTABLE;
+
+	switch (adev->asic_type) {
+	case CHIP_ARCTURUS:
+		if (mem->alloc_flags & ALLOC_MEM_FLAGS_VRAM) {
+			if (bo_adev == adev)
+				mapping_flags |= coherent ?
+					AMDGPU_VM_MTYPE_CC : AMDGPU_VM_MTYPE_RW;
+			else
+				mapping_flags |= AMDGPU_VM_MTYPE_UC;
+		} else {
+			mapping_flags |= coherent ?
+				AMDGPU_VM_MTYPE_UC : AMDGPU_VM_MTYPE_NC;
+		}
+		break;
+	default:
+		mapping_flags |= coherent ?
+			AMDGPU_VM_MTYPE_UC : AMDGPU_VM_MTYPE_NC;
+	}
+
+	return amdgpu_gem_va_map_flags(adev, mapping_flags);
+}
+
 /* add_bo_to_vm - Add a BO to a VM
  *
  * Everything that needs to bo done only once when a BO is first added
@@ -404,8 +429,7 @@ static int add_bo_to_vm(struct amdgpu_device *adev, struct kgd_mem *mem,
 	}
 
 	bo_va_entry->va = va;
-	bo_va_entry->pte_flags = amdgpu_gmc_get_pte_flags(adev,
-							 mem->mapping_flags);
+	bo_va_entry->pte_flags = get_pte_flags(adev, mem);
 	bo_va_entry->kgd_dev = (void *)adev;
 	list_add(&bo_va_entry->bo_list, list_bo_va);
 
@@ -481,8 +505,7 @@ static void remove_kgd_mem_from_kfd_bo_list(struct kgd_mem *mem,
  *
  * Returns 0 for success, negative errno for errors.
  */
-static int init_user_pages(struct kgd_mem *mem, struct mm_struct *mm,
-			   uint64_t user_addr)
+static int init_user_pages(struct kgd_mem *mem, uint64_t user_addr)
 {
 	struct amdkfd_process_info *process_info = mem->process_info;
 	struct amdgpu_bo *bo = mem->bo;
@@ -586,7 +609,7 @@ static int reserve_bo_and_vm(struct kgd_mem *mem,
 	amdgpu_vm_get_pd_bo(vm, &ctx->list, &ctx->vm_pd[0]);
 
 	ret = ttm_eu_reserve_buffers(&ctx->ticket, &ctx->list,
-				     false, &ctx->duplicates, true);
+				     false, &ctx->duplicates);
 	if (!ret)
 		ctx->reserved = true;
 	else {
@@ -659,7 +682,7 @@ static int reserve_bo_and_cond_vms(struct kgd_mem *mem,
 	}
 
 	ret = ttm_eu_reserve_buffers(&ctx->ticket, &ctx->list,
-				     false, &ctx->duplicates, true);
+				     false, &ctx->duplicates);
 	if (!ret)
 		ctx->reserved = true;
 	else
@@ -1079,10 +1102,8 @@ int amdgpu_amdkfd_gpuvm_alloc_memory_of_gpu(
 	uint64_t user_addr = 0;
 	struct amdgpu_bo *bo;
 	struct amdgpu_bo_param bp;
-	int byte_align;
 	u32 domain, alloc_domain;
 	u64 alloc_flags;
-	uint32_t mapping_flags;
 	int ret;
 
 	/*
@@ -1135,25 +1156,7 @@ int amdgpu_amdkfd_gpuvm_alloc_memory_of_gpu(
 	if ((*mem)->aql_queue)
 		size = size >> 1;
 
-	/* Workaround for TLB bug on older VI chips */
-	byte_align = (adev->family == AMDGPU_FAMILY_VI &&
-			adev->asic_type != CHIP_FIJI &&
-			adev->asic_type != CHIP_POLARIS10 &&
-			adev->asic_type != CHIP_POLARIS11 &&
-			adev->asic_type != CHIP_POLARIS12 &&
-			adev->asic_type != CHIP_VEGAM) ?
-			VI_BO_SIZE_ALIGN : 1;
-
-	mapping_flags = AMDGPU_VM_PAGE_READABLE;
-	if (flags & ALLOC_MEM_FLAGS_WRITABLE)
-		mapping_flags |= AMDGPU_VM_PAGE_WRITEABLE;
-	if (flags & ALLOC_MEM_FLAGS_EXECUTABLE)
-		mapping_flags |= AMDGPU_VM_PAGE_EXECUTABLE;
-	if (flags & ALLOC_MEM_FLAGS_COHERENT)
-		mapping_flags |= AMDGPU_VM_MTYPE_UC;
-	else
-		mapping_flags |= AMDGPU_VM_MTYPE_NC;
-	(*mem)->mapping_flags = mapping_flags;
+	(*mem)->alloc_flags = flags;
 
 	amdgpu_sync_create(&(*mem)->sync);
 
@@ -1168,7 +1171,7 @@ int amdgpu_amdkfd_gpuvm_alloc_memory_of_gpu(
 
 	memset(&bp, 0, sizeof(bp));
 	bp.size = size;
-	bp.byte_align = byte_align;
+	bp.byte_align = 1;
 	bp.domain = alloc_domain;
 	bp.flags = alloc_flags;
 	bp.type = bo_type;
@@ -1195,7 +1198,7 @@ int amdgpu_amdkfd_gpuvm_alloc_memory_of_gpu(
 	add_kgd_mem_to_kfd_bo_list(*mem, avm->process_info, user_addr);
 
 	if (user_addr) {
-		ret = init_user_pages(*mem, current->mm, user_addr);
+		ret = init_user_pages(*mem, user_addr);
 		if (ret)
 			goto allocate_init_user_pages_failed;
 	}
@@ -1626,9 +1629,10 @@ int amdgpu_amdkfd_gpuvm_import_dmabuf(struct kgd_dev *kgd,
 
 	INIT_LIST_HEAD(&(*mem)->bo_va_list);
 	mutex_init(&(*mem)->lock);
-	(*mem)->mapping_flags =
-		AMDGPU_VM_PAGE_READABLE | AMDGPU_VM_PAGE_WRITEABLE |
-		AMDGPU_VM_PAGE_EXECUTABLE | AMDGPU_VM_MTYPE_NC;
+	(*mem)->alloc_flags =
+		((bo->preferred_domains & AMDGPU_GEM_DOMAIN_VRAM) ?
+		 ALLOC_MEM_FLAGS_VRAM : ALLOC_MEM_FLAGS_GTT) |
+		ALLOC_MEM_FLAGS_WRITABLE | ALLOC_MEM_FLAGS_EXECUTABLE;
 
 	(*mem)->bo = amdgpu_bo_ref(bo);
 	(*mem)->va = va;
@@ -1739,6 +1743,10 @@ static int update_invalid_user_pages(struct amdkfd_process_info *process_info,
 			return ret;
 		}
 
+		/*
+		 * FIXME: Cannot ignore the return code, must hold
+		 * notifier_lock
+		 */
 		amdgpu_ttm_tt_get_user_pages_done(bo->tbo.ttm);
 
 		/* Mark the BO as valid unless it was invalidated
@@ -1797,8 +1805,7 @@ static int validate_invalid_user_pages(struct amdkfd_process_info *process_info)
 	}
 
 	/* Reserve all BOs and page tables for validation */
-	ret = ttm_eu_reserve_buffers(&ticket, &resv_list, false, &duplicates,
-				     true);
+	ret = ttm_eu_reserve_buffers(&ticket, &resv_list, false, &duplicates);
 	WARN(!list_empty(&duplicates), "Duplicates should be empty");
 	if (ret)
 		goto out_free;
@@ -1996,7 +2003,7 @@ int amdgpu_amdkfd_gpuvm_restore_process_bos(void *info, struct dma_fence **ef)
 	}
 
 	ret = ttm_eu_reserve_buffers(&ctx.ticket, &ctx.list,
-				     false, &duplicate_save, true);
+				     false, &duplicate_save);
 	if (ret) {
 		pr_debug("Memory eviction: TTM Reserve Failed. Try again\n");
 		goto ttm_reserve_fail;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.c
index 1c9d40f97a9b..72232fccf61a 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.c
@@ -2038,6 +2038,11 @@ int amdgpu_atombios_init(struct amdgpu_device *adev)
 	if (adev->is_atom_fw) {
 		amdgpu_atomfirmware_scratch_regs_init(adev);
 		amdgpu_atomfirmware_allocate_fb_scratch(adev);
+		ret = amdgpu_atomfirmware_get_mem_train_fb_loc(adev);
+		if (ret) {
+			DRM_ERROR("Failed to get mem train fb location.\n");
+			return ret;
+		}
 	} else {
 		amdgpu_atombios_scratch_regs_init(adev);
 		amdgpu_atombios_allocate_fb_scratch(adev);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c
index daf687428cdb..ff4eb96bdfb5 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c
@@ -27,6 +27,7 @@
 #include "amdgpu_atomfirmware.h"
 #include "atom.h"
 #include "atombios.h"
+#include "soc15_hw_ip.h"
 
 bool amdgpu_atomfirmware_gpu_supports_virtualization(struct amdgpu_device *adev)
 {
@@ -120,65 +121,14 @@ union vram_info {
 	struct atom_vram_info_header_v2_3 v23;
 	struct atom_vram_info_header_v2_4 v24;
 };
-/*
- * Return vram width from integrated system info table, if available,
- * or 0 if not.
- */
-int amdgpu_atomfirmware_get_vram_width(struct amdgpu_device *adev)
-{
-	struct amdgpu_mode_info *mode_info = &adev->mode_info;
-	int index;
-	u16 data_offset, size;
-	union igp_info *igp_info;
-	union vram_info *vram_info;
-	u32 mem_channel_number;
-	u32 mem_channel_width;
-	u8 frev, crev;
-
-	if (adev->flags & AMD_IS_APU)
-		index = get_index_into_master_table(atom_master_list_of_data_tables_v2_1,
-						    integratedsysteminfo);
-	else
-		index = get_index_into_master_table(atom_master_list_of_data_tables_v2_1,
-						    vram_info);
 
-	/* get any igp specific overrides */
-	if (amdgpu_atom_parse_data_header(mode_info->atom_context, index, &size,
-				   &frev, &crev, &data_offset)) {
-		if (adev->flags & AMD_IS_APU) {
-			igp_info = (union igp_info *)
-				(mode_info->atom_context->bios + data_offset);
-			switch (crev) {
-			case 11:
-				mem_channel_number = igp_info->v11.umachannelnumber;
-				/* channel width is 64 */
-				return mem_channel_number * 64;
-			default:
-				return 0;
-			}
-		} else {
-			vram_info = (union vram_info *)
-				(mode_info->atom_context->bios + data_offset);
-			switch (crev) {
-			case 3:
-				mem_channel_number = vram_info->v23.vram_module[0].channel_num;
-				mem_channel_width = vram_info->v23.vram_module[0].channel_width;
-				return mem_channel_number * (1 << mem_channel_width);
-			case 4:
-				mem_channel_number = vram_info->v24.vram_module[0].channel_num;
-				mem_channel_width = vram_info->v24.vram_module[0].channel_width;
-				return mem_channel_number * (1 << mem_channel_width);
-			default:
-				return 0;
-			}
-		}
-	}
-
-	return 0;
-}
+union vram_module {
+	struct atom_vram_module_v9 v9;
+	struct atom_vram_module_v10 v10;
+};
 
-static int convert_atom_mem_type_to_vram_type (struct amdgpu_device *adev,
-					       int atom_mem_type)
+static int convert_atom_mem_type_to_vram_type(struct amdgpu_device *adev,
+					      int atom_mem_type)
 {
 	int vram_type;
 
@@ -219,19 +169,25 @@ static int convert_atom_mem_type_to_vram_type (struct amdgpu_device *adev,
 
 	return vram_type;
 }
-/*
- * Return vram type from either integrated system info table
- * or umc info table, if available, or 0 (TYPE_UNKNOWN) if not
- */
-int amdgpu_atomfirmware_get_vram_type(struct amdgpu_device *adev)
+
+
+int
+amdgpu_atomfirmware_get_vram_info(struct amdgpu_device *adev,
+				  int *vram_width, int *vram_type,
+				  int *vram_vendor)
 {
 	struct amdgpu_mode_info *mode_info = &adev->mode_info;
-	int index;
+	int index, i = 0;
 	u16 data_offset, size;
 	union igp_info *igp_info;
 	union vram_info *vram_info;
+	union vram_module *vram_module;
 	u8 frev, crev;
 	u8 mem_type;
+	u8 mem_vendor;
+	u32 mem_channel_number;
+	u32 mem_channel_width;
+	u32 module_id;
 
 	if (adev->flags & AMD_IS_APU)
 		index = get_index_into_master_table(atom_master_list_of_data_tables_v2_1,
@@ -239,6 +195,7 @@ int amdgpu_atomfirmware_get_vram_type(struct amdgpu_device *adev)
 	else
 		index = get_index_into_master_table(atom_master_list_of_data_tables_v2_1,
 						    vram_info);
+
 	if (amdgpu_atom_parse_data_header(mode_info->atom_context,
 					  index, &size,
 					  &frev, &crev, &data_offset)) {
@@ -247,25 +204,67 @@ int amdgpu_atomfirmware_get_vram_type(struct amdgpu_device *adev)
 				(mode_info->atom_context->bios + data_offset);
 			switch (crev) {
 			case 11:
+				mem_channel_number = igp_info->v11.umachannelnumber;
+				/* channel width is 64 */
+				if (vram_width)
+					*vram_width = mem_channel_number * 64;
 				mem_type = igp_info->v11.memorytype;
-				return convert_atom_mem_type_to_vram_type(adev, mem_type);
+				if (vram_type)
+					*vram_type = convert_atom_mem_type_to_vram_type(adev, mem_type);
+				break;
 			default:
-				return 0;
+				return -EINVAL;
 			}
 		} else {
 			vram_info = (union vram_info *)
 				(mode_info->atom_context->bios + data_offset);
+			module_id = (RREG32(adev->bios_scratch_reg_offset + 4) & 0x00ff0000) >> 16;
 			switch (crev) {
 			case 3:
-				mem_type = vram_info->v23.vram_module[0].memory_type;
-				return convert_atom_mem_type_to_vram_type(adev, mem_type);
+				if (module_id > vram_info->v23.vram_module_num)
+					module_id = 0;
+				vram_module = (union vram_module *)vram_info->v23.vram_module;
+				while (i < module_id) {
+					vram_module = (union vram_module *)
+						((u8 *)vram_module + vram_module->v9.vram_module_size);
+					i++;
+				}
+				mem_type = vram_module->v9.memory_type;
+				if (vram_type)
+					*vram_type = convert_atom_mem_type_to_vram_type(adev, mem_type);
+				mem_channel_number = vram_module->v9.channel_num;
+				mem_channel_width = vram_module->v9.channel_width;
+				if (vram_width)
+					*vram_width = mem_channel_number * (1 << mem_channel_width);
+				mem_vendor = (vram_module->v9.vender_rev_id) & 0xF;
+				if (vram_vendor)
+					*vram_vendor = mem_vendor;
+				break;
 			case 4:
-				mem_type = vram_info->v24.vram_module[0].memory_type;
-				return convert_atom_mem_type_to_vram_type(adev, mem_type);
+				if (module_id > vram_info->v24.vram_module_num)
+					module_id = 0;
+				vram_module = (union vram_module *)vram_info->v24.vram_module;
+				while (i < module_id) {
+					vram_module = (union vram_module *)
+						((u8 *)vram_module + vram_module->v10.vram_module_size);
+					i++;
+				}
+				mem_type = vram_module->v10.memory_type;
+				if (vram_type)
+					*vram_type = convert_atom_mem_type_to_vram_type(adev, mem_type);
+				mem_channel_number = vram_module->v10.channel_num;
+				mem_channel_width = vram_module->v10.channel_width;
+				if (vram_width)
+					*vram_width = mem_channel_number * (1 << mem_channel_width);
+				mem_vendor = (vram_module->v10.vender_rev_id) & 0xF;
+				if (vram_vendor)
+					*vram_vendor = mem_vendor;
+				break;
 			default:
-				return 0;
+				return -EINVAL;
 			}
 		}
+
 	}
 
 	return 0;
@@ -464,3 +463,138 @@ int amdgpu_atomfirmware_get_gfx_info(struct amdgpu_device *adev)
 	}
 	return -EINVAL;
 }
+
+/*
+ * Check if VBIOS supports GDDR6 training data save/restore
+ */
+static bool gddr6_mem_train_vbios_support(struct amdgpu_device *adev)
+{
+	uint16_t data_offset;
+	int index;
+
+	index = get_index_into_master_table(atom_master_list_of_data_tables_v2_1,
+					    firmwareinfo);
+	if (amdgpu_atom_parse_data_header(adev->mode_info.atom_context, index, NULL,
+					  NULL, NULL, &data_offset)) {
+		struct atom_firmware_info_v3_1 *firmware_info =
+			(struct atom_firmware_info_v3_1 *)(adev->mode_info.atom_context->bios +
+							   data_offset);
+
+		DRM_DEBUG("atom firmware capability:0x%08x.\n",
+			  le32_to_cpu(firmware_info->firmware_capability));
+
+		if (le32_to_cpu(firmware_info->firmware_capability) &
+		    ATOM_FIRMWARE_CAP_ENABLE_2STAGE_BIST_TRAINING)
+			return true;
+	}
+
+	return false;
+}
+
+static int gddr6_mem_train_support(struct amdgpu_device *adev)
+{
+	int ret;
+	uint32_t major, minor, revision, hw_v;
+
+	if (gddr6_mem_train_vbios_support(adev)) {
+		amdgpu_discovery_get_ip_version(adev, MP0_HWID, &major, &minor, &revision);
+		hw_v = HW_REV(major, minor, revision);
+		/*
+		 * treat 0 revision as a special case since register for MP0 and MMHUB is missing
+		 * for some Navi10 A0, preventing driver from discovering the hwip information since
+		 * none of the functions will be initialized, it should not cause any problems
+		 */
+		switch (hw_v) {
+		case HW_REV(11, 0, 0):
+		case HW_REV(11, 0, 5):
+			ret = 1;
+			break;
+		default:
+			DRM_ERROR("memory training vbios supports but psp hw(%08x)"
+				  " doesn't support!\n", hw_v);
+			ret = -1;
+			break;
+		}
+	} else {
+		ret = 0;
+		hw_v = -1;
+	}
+
+
+	DRM_DEBUG("mp0 hw_v %08x, ret:%d.\n", hw_v, ret);
+	return ret;
+}
+
+int amdgpu_atomfirmware_get_mem_train_fb_loc(struct amdgpu_device *adev)
+{
+	struct atom_context *ctx = adev->mode_info.atom_context;
+	unsigned char *bios = ctx->bios;
+	struct vram_reserve_block *reserved_block;
+	int index, block_number;
+	uint8_t frev, crev;
+	uint16_t data_offset, size;
+	uint32_t start_address_in_kb;
+	uint64_t offset;
+	int ret;
+
+	adev->fw_vram_usage.mem_train_support = false;
+
+	if (adev->asic_type != CHIP_NAVI10 &&
+	    adev->asic_type != CHIP_NAVI14)
+		return 0;
+
+	if (amdgpu_sriov_vf(adev))
+		return 0;
+
+	ret = gddr6_mem_train_support(adev);
+	if (ret == -1)
+		return -EINVAL;
+	else if (ret == 0)
+		return 0;
+
+	index = get_index_into_master_table(atom_master_list_of_data_tables_v2_1,
+					    vram_usagebyfirmware);
+	ret = amdgpu_atom_parse_data_header(ctx, index, &size, &frev, &crev,
+					    &data_offset);
+	if (ret == 0) {
+		DRM_ERROR("parse data header failed.\n");
+		return -EINVAL;
+	}
+
+	DRM_DEBUG("atom firmware common table header size:0x%04x, frev:0x%02x,"
+		  " crev:0x%02x, data_offset:0x%04x.\n", size, frev, crev, data_offset);
+	/* only support 2.1+ */
+	if (((uint16_t)frev << 8 | crev) < 0x0201) {
+		DRM_ERROR("frev:0x%02x, crev:0x%02x < 2.1 !\n", frev, crev);
+		return -EINVAL;
+	}
+
+	reserved_block = (struct vram_reserve_block *)
+		(bios + data_offset + sizeof(struct atom_common_table_header));
+	block_number = ((unsigned int)size - sizeof(struct atom_common_table_header))
+		/ sizeof(struct vram_reserve_block);
+	reserved_block += (block_number > 0) ? block_number-1 : 0;
+	DRM_DEBUG("block_number:0x%04x, last block: 0x%08xkb sz, %dkb fw, %dkb drv.\n",
+		  block_number,
+		  le32_to_cpu(reserved_block->start_address_in_kb),
+		  le16_to_cpu(reserved_block->used_by_firmware_in_kb),
+		  le16_to_cpu(reserved_block->used_by_driver_in_kb));
+	if (reserved_block->used_by_firmware_in_kb > 0) {
+		start_address_in_kb = le32_to_cpu(reserved_block->start_address_in_kb);
+		offset = (uint64_t)start_address_in_kb * ONE_KiB;
+		if ((offset & (ONE_MiB - 1)) < (4 * ONE_KiB + 1) ) {
+			offset -= ONE_MiB;
+		}
+
+		offset &= ~(ONE_MiB - 1);
+		adev->fw_vram_usage.mem_train_fb_loc = offset;
+		adev->fw_vram_usage.mem_train_support = true;
+		DRM_DEBUG("mem_train_fb_loc:0x%09llx.\n", offset);
+		ret = 0;
+	} else {
+		DRM_ERROR("used_by_firmware_in_kb is 0!\n");
+		ret = -EINVAL;
+	}
+
+	return ret;
+}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.h
index 5ec6f92f353c..f871af5ea6f3 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.h
@@ -29,8 +29,9 @@
 bool amdgpu_atomfirmware_gpu_supports_virtualization(struct amdgpu_device *adev);
 void amdgpu_atomfirmware_scratch_regs_init(struct amdgpu_device *adev);
 int amdgpu_atomfirmware_allocate_fb_scratch(struct amdgpu_device *adev);
-int amdgpu_atomfirmware_get_vram_width(struct amdgpu_device *adev);
-int amdgpu_atomfirmware_get_vram_type(struct amdgpu_device *adev);
+int amdgpu_atomfirmware_get_vram_info(struct amdgpu_device *adev,
+	int *vram_width, int *vram_type, int *vram_vendor);
+int amdgpu_atomfirmware_get_mem_train_fb_loc(struct amdgpu_device *adev);
 int amdgpu_atomfirmware_get_clock_info(struct amdgpu_device *adev);
 int amdgpu_atomfirmware_get_gfx_info(struct amdgpu_device *adev);
 bool amdgpu_atomfirmware_mem_ecc_supported(struct amdgpu_device *adev);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c
index 3e35a8f2c5e5..a97fb759e2f4 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c
@@ -613,17 +613,7 @@ static bool amdgpu_atpx_detect(void)
 	bool d3_supported = false;
 	struct pci_dev *parent_pdev;
 
-	while ((pdev = pci_get_class(PCI_CLASS_DISPLAY_VGA << 8, pdev)) != NULL) {
-		vga_count++;
-
-		has_atpx |= (amdgpu_atpx_pci_probe_handle(pdev) == true);
-
-		parent_pdev = pci_upstream_bridge(pdev);
-		d3_supported |= parent_pdev && parent_pdev->bridge_d3;
-		amdgpu_atpx_get_quirks(pdev);
-	}
-
-	while ((pdev = pci_get_class(PCI_CLASS_DISPLAY_OTHER << 8, pdev)) != NULL) {
+	while ((pdev = pci_get_class(PCI_BASE_CLASS_DISPLAY << 16, pdev)) != NULL) {
 		vga_count++;
 
 		has_atpx |= (amdgpu_atpx_pci_probe_handle(pdev) == true);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_benchmark.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_benchmark.c
index 649e68c4479b..d1495e1c9289 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_benchmark.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_benchmark.c
@@ -33,7 +33,7 @@ static int amdgpu_benchmark_do_move(struct amdgpu_device *adev, unsigned size,
 {
 	unsigned long start_jiffies;
 	unsigned long end_jiffies;
-	struct dma_fence *fence = NULL;
+	struct dma_fence *fence;
 	int i, r;
 
 	start_jiffies = jiffies;
@@ -44,16 +44,14 @@ static int amdgpu_benchmark_do_move(struct amdgpu_device *adev, unsigned size,
 		if (r)
 			goto exit_do_move;
 		r = dma_fence_wait(fence, false);
+		dma_fence_put(fence);
 		if (r)
 			goto exit_do_move;
-		dma_fence_put(fence);
 	}
 	end_jiffies = jiffies;
 	r = jiffies_to_msecs(end_jiffies - start_jiffies);
 
 exit_do_move:
-	if (fence)
-		dma_fence_put(fence);
 	return r;
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c
index 7bcf86c61999..85b0515c0fdc 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c
@@ -140,7 +140,12 @@ int amdgpu_bo_list_create(struct amdgpu_device *adev, struct drm_file *filp,
 	return 0;
 
 error_free:
-	while (i--) {
+	for (i = 0; i < last_entry; ++i) {
+		struct amdgpu_bo *bo = ttm_to_amdgpu_bo(array[i].tv.bo);
+
+		amdgpu_bo_unref(&bo);
+	}
+	for (i = first_userptr; i < num_entries; ++i) {
 		struct amdgpu_bo *bo = ttm_to_amdgpu_bo(array[i].tv.bo);
 
 		amdgpu_bo_unref(&bo);
@@ -270,7 +275,7 @@ int amdgpu_bo_list_ioctl(struct drm_device *dev, void *data,
 
 	r = amdgpu_bo_create_list_entry_array(&args->in, &info);
 	if (r)
-		goto error_free;
+		return r;
 
 	switch (args->in.operation) {
 	case AMDGPU_BO_LIST_OP_CREATE:
@@ -283,8 +288,7 @@ int amdgpu_bo_list_ioctl(struct drm_device *dev, void *data,
 		r = idr_alloc(&fpriv->bo_list_handles, list, 1, 0, GFP_KERNEL);
 		mutex_unlock(&fpriv->bo_list_lock);
 		if (r < 0) {
-			amdgpu_bo_list_put(list);
-			return r;
+			goto error_put_list;
 		}
 
 		handle = r;
@@ -306,9 +310,8 @@ int amdgpu_bo_list_ioctl(struct drm_device *dev, void *data,
 		mutex_unlock(&fpriv->bo_list_lock);
 
 		if (IS_ERR(old)) {
-			amdgpu_bo_list_put(list);
 			r = PTR_ERR(old);
-			goto error_free;
+			goto error_put_list;
 		}
 
 		amdgpu_bo_list_put(old);
@@ -325,8 +328,10 @@ int amdgpu_bo_list_ioctl(struct drm_device *dev, void *data,
 
 	return 0;
 
+error_put_list:
+	amdgpu_bo_list_put(list);
+
 error_free:
-	if (info)
-		kvfree(info);
+	kvfree(info);
 	return r;
 }
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c
index ece55c8fa673..a62cbc8199de 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c
@@ -217,11 +217,10 @@ amdgpu_connector_update_scratch_regs(struct drm_connector *connector,
 	struct drm_encoder *encoder;
 	const struct drm_connector_helper_funcs *connector_funcs = connector->helper_private;
 	bool connected;
-	int i;
 
 	best_encoder = connector_funcs->best_encoder(connector);
 
-	drm_connector_for_each_possible_encoder(connector, encoder, i) {
+	drm_connector_for_each_possible_encoder(connector, encoder) {
 		if ((encoder == best_encoder) && (status == connector_status_connected))
 			connected = true;
 		else
@@ -236,9 +235,8 @@ amdgpu_connector_find_encoder(struct drm_connector *connector,
 			       int encoder_type)
 {
 	struct drm_encoder *encoder;
-	int i;
 
-	drm_connector_for_each_possible_encoder(connector, encoder, i) {
+	drm_connector_for_each_possible_encoder(connector, encoder) {
 		if (encoder->encoder_type == encoder_type)
 			return encoder;
 	}
@@ -347,10 +345,9 @@ static struct drm_encoder *
 amdgpu_connector_best_single_encoder(struct drm_connector *connector)
 {
 	struct drm_encoder *encoder;
-	int i;
 
 	/* pick the first one */
-	drm_connector_for_each_possible_encoder(connector, encoder, i)
+	drm_connector_for_each_possible_encoder(connector, encoder)
 		return encoder;
 
 	return NULL;
@@ -1022,8 +1019,12 @@ amdgpu_connector_dvi_detect(struct drm_connector *connector, bool force)
 			 */
 			if (amdgpu_connector->shared_ddc && (ret == connector_status_connected)) {
 				struct drm_connector *list_connector;
+				struct drm_connector_list_iter iter;
 				struct amdgpu_connector *list_amdgpu_connector;
-				list_for_each_entry(list_connector, &dev->mode_config.connector_list, head) {
+
+				drm_connector_list_iter_begin(dev, &iter);
+				drm_for_each_connector_iter(list_connector,
+							    &iter) {
 					if (connector == list_connector)
 						continue;
 					list_amdgpu_connector = to_amdgpu_connector(list_connector);
@@ -1040,6 +1041,7 @@ amdgpu_connector_dvi_detect(struct drm_connector *connector, bool force)
 						}
 					}
 				}
+				drm_connector_list_iter_end(&iter);
 			}
 		}
 	}
@@ -1065,9 +1067,8 @@ amdgpu_connector_dvi_detect(struct drm_connector *connector, bool force)
 	/* find analog encoder */
 	if (amdgpu_connector->dac_load_detect) {
 		struct drm_encoder *encoder;
-		int i;
 
-		drm_connector_for_each_possible_encoder(connector, encoder, i) {
+		drm_connector_for_each_possible_encoder(connector, encoder) {
 			if (encoder->encoder_type != DRM_MODE_ENCODER_DAC &&
 			    encoder->encoder_type != DRM_MODE_ENCODER_TVDAC)
 				continue;
@@ -1117,9 +1118,8 @@ amdgpu_connector_dvi_encoder(struct drm_connector *connector)
 {
 	struct amdgpu_connector *amdgpu_connector = to_amdgpu_connector(connector);
 	struct drm_encoder *encoder;
-	int i;
 
-	drm_connector_for_each_possible_encoder(connector, encoder, i) {
+	drm_connector_for_each_possible_encoder(connector, encoder) {
 		if (amdgpu_connector->use_digital == true) {
 			if (encoder->encoder_type == DRM_MODE_ENCODER_TMDS)
 				return encoder;
@@ -1134,7 +1134,7 @@ amdgpu_connector_dvi_encoder(struct drm_connector *connector)
 
 	/* then check use digitial */
 	/* pick the first one */
-	drm_connector_for_each_possible_encoder(connector, encoder, i)
+	drm_connector_for_each_possible_encoder(connector, encoder)
 		return encoder;
 
 	return NULL;
@@ -1271,9 +1271,8 @@ u16 amdgpu_connector_encoder_get_dp_bridge_encoder_id(struct drm_connector *conn
 {
 	struct drm_encoder *encoder;
 	struct amdgpu_encoder *amdgpu_encoder;
-	int i;
 
-	drm_connector_for_each_possible_encoder(connector, encoder, i) {
+	drm_connector_for_each_possible_encoder(connector, encoder) {
 		amdgpu_encoder = to_amdgpu_encoder(encoder);
 
 		switch (amdgpu_encoder->encoder_id) {
@@ -1292,10 +1291,9 @@ static bool amdgpu_connector_encoder_is_hbr2(struct drm_connector *connector)
 {
 	struct drm_encoder *encoder;
 	struct amdgpu_encoder *amdgpu_encoder;
-	int i;
 	bool found = false;
 
-	drm_connector_for_each_possible_encoder(connector, encoder, i) {
+	drm_connector_for_each_possible_encoder(connector, encoder) {
 		amdgpu_encoder = to_amdgpu_encoder(encoder);
 		if (amdgpu_encoder->caps & ATOM_ENCODER_CAP_RECORD_HBR2)
 			found = true;
@@ -1501,6 +1499,7 @@ amdgpu_connector_add(struct amdgpu_device *adev,
 {
 	struct drm_device *dev = adev->ddev;
 	struct drm_connector *connector;
+	struct drm_connector_list_iter iter;
 	struct amdgpu_connector *amdgpu_connector;
 	struct amdgpu_connector_atom_dig *amdgpu_dig_connector;
 	struct drm_encoder *encoder;
@@ -1515,10 +1514,12 @@ amdgpu_connector_add(struct amdgpu_device *adev,
 		return;
 
 	/* see if we already added it */
-	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
+	drm_connector_list_iter_begin(dev, &iter);
+	drm_for_each_connector_iter(connector, &iter) {
 		amdgpu_connector = to_amdgpu_connector(connector);
 		if (amdgpu_connector->connector_id == connector_id) {
 			amdgpu_connector->devices |= supported_device;
+			drm_connector_list_iter_end(&iter);
 			return;
 		}
 		if (amdgpu_connector->ddc_bus && i2c_bus->valid) {
@@ -1533,6 +1534,7 @@ amdgpu_connector_add(struct amdgpu_device *adev,
 			}
 		}
 	}
+	drm_connector_list_iter_end(&iter);
 
 	/* check if it's a dp bridge */
 	list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) {
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
index 2e53feed40e2..5ca905b4a0fb 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
@@ -35,6 +35,7 @@
 #include "amdgpu_trace.h"
 #include "amdgpu_gmc.h"
 #include "amdgpu_gem.h"
+#include "amdgpu_ras.h"
 
 static int amdgpu_cs_user_fence_chunk(struct amdgpu_cs_parser *p,
 				      struct drm_amdgpu_cs_chunk_fence *data,
@@ -449,75 +450,12 @@ retry:
 	return r;
 }
 
-/* Last resort, try to evict something from the current working set */
-static bool amdgpu_cs_try_evict(struct amdgpu_cs_parser *p,
-				struct amdgpu_bo *validated)
-{
-	uint32_t domain = validated->allowed_domains;
-	struct ttm_operation_ctx ctx = { true, false };
-	int r;
-
-	if (!p->evictable)
-		return false;
-
-	for (;&p->evictable->tv.head != &p->validated;
-	     p->evictable = list_prev_entry(p->evictable, tv.head)) {
-
-		struct amdgpu_bo_list_entry *candidate = p->evictable;
-		struct amdgpu_bo *bo = ttm_to_amdgpu_bo(candidate->tv.bo);
-		struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
-		bool update_bytes_moved_vis;
-		uint32_t other;
-
-		/* If we reached our current BO we can forget it */
-		if (bo == validated)
-			break;
-
-		/* We can't move pinned BOs here */
-		if (bo->pin_count)
-			continue;
-
-		other = amdgpu_mem_type_to_domain(bo->tbo.mem.mem_type);
-
-		/* Check if this BO is in one of the domains we need space for */
-		if (!(other & domain))
-			continue;
-
-		/* Check if we can move this BO somewhere else */
-		other = bo->allowed_domains & ~domain;
-		if (!other)
-			continue;
-
-		/* Good we can try to move this BO somewhere else */
-		update_bytes_moved_vis =
-				!amdgpu_gmc_vram_full_visible(&adev->gmc) &&
-				amdgpu_bo_in_cpu_visible_vram(bo);
-		amdgpu_bo_placement_from_domain(bo, other);
-		r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx);
-		p->bytes_moved += ctx.bytes_moved;
-		if (update_bytes_moved_vis)
-			p->bytes_moved_vis += ctx.bytes_moved;
-
-		if (unlikely(r))
-			break;
-
-		p->evictable = list_prev_entry(p->evictable, tv.head);
-		list_move(&candidate->tv.head, &p->validated);
-
-		return true;
-	}
-
-	return false;
-}
-
 static int amdgpu_cs_validate(void *param, struct amdgpu_bo *bo)
 {
 	struct amdgpu_cs_parser *p = param;
 	int r;
 
-	do {
-		r = amdgpu_cs_bo_validate(p, bo);
-	} while (r == -ENOMEM && amdgpu_cs_try_evict(p, bo));
+	r = amdgpu_cs_bo_validate(p, bo);
 	if (r)
 		return r;
 
@@ -536,7 +474,6 @@ static int amdgpu_cs_list_validate(struct amdgpu_cs_parser *p,
 
 	list_for_each_entry(lobj, validated, tv.head) {
 		struct amdgpu_bo *bo = ttm_to_amdgpu_bo(lobj->tv.bo);
-		bool binding_userptr = false;
 		struct mm_struct *usermm;
 
 		usermm = amdgpu_ttm_tt_get_usermm(bo->tbo.ttm);
@@ -553,20 +490,14 @@ static int amdgpu_cs_list_validate(struct amdgpu_cs_parser *p,
 
 			amdgpu_ttm_tt_set_user_pages(bo->tbo.ttm,
 						     lobj->user_pages);
-			binding_userptr = true;
 		}
 
-		if (p->evictable == lobj)
-			p->evictable = NULL;
-
 		r = amdgpu_cs_validate(p, bo);
 		if (r)
 			return r;
 
-		if (binding_userptr) {
-			kvfree(lobj->user_pages);
-			lobj->user_pages = NULL;
-		}
+		kvfree(lobj->user_pages);
+		lobj->user_pages = NULL;
 	}
 	return 0;
 }
@@ -607,8 +538,6 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p,
 		e->tv.num_shared = 2;
 
 	amdgpu_bo_list_get_list(p->bo_list, &p->validated);
-	if (p->bo_list->first_userptr != p->bo_list->num_entries)
-		p->mn = amdgpu_mn_get(p->adev, AMDGPU_MN_TYPE_GFX);
 
 	INIT_LIST_HEAD(&duplicates);
 	amdgpu_vm_get_pd_bo(&fpriv->vm, &p->validated, &p->vm_pd);
@@ -650,7 +579,7 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p,
 	}
 
 	r = ttm_eu_reserve_buffers(&p->ticket, &p->validated, true,
-				   &duplicates, false);
+				   &duplicates);
 	if (unlikely(r != 0)) {
 		if (r != -ERESTARTSYS)
 			DRM_ERROR("ttm_eu_reserve_buffers failed.\n");
@@ -661,9 +590,6 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p,
 					  &p->bytes_moved_vis_threshold);
 	p->bytes_moved = 0;
 	p->bytes_moved_vis = 0;
-	p->evictable = list_last_entry(&p->validated,
-				       struct amdgpu_bo_list_entry,
-				       tv.head);
 
 	r = amdgpu_vm_validate_pt_bos(p->adev, &fpriv->vm,
 				      amdgpu_cs_validate, p);
@@ -915,7 +841,7 @@ static int amdgpu_cs_vm_handling(struct amdgpu_cs_parser *p)
 	if (r)
 		return r;
 
-	r = amdgpu_vm_update_directories(adev, vm);
+	r = amdgpu_vm_update_pdes(adev, vm, false);
 	if (r)
 		return r;
 
@@ -1291,11 +1217,11 @@ static int amdgpu_cs_submit(struct amdgpu_cs_parser *p,
 	if (r)
 		goto error_unlock;
 
-	/* No memory allocation is allowed while holding the mn lock.
-	 * p->mn is hold until amdgpu_cs_submit is finished and fence is added
-	 * to BOs.
+	/* No memory allocation is allowed while holding the notifier lock.
+	 * The lock is held until amdgpu_cs_submit is finished and fence is
+	 * added to BOs.
 	 */
-	amdgpu_mn_lock(p->mn);
+	mutex_lock(&p->adev->notifier_lock);
 
 	/* If userptr are invalidated after amdgpu_cs_parser_bos(), return
 	 * -EAGAIN, drmIoctl in libdrm will restart the amdgpu_cs_ioctl.
@@ -1338,13 +1264,13 @@ static int amdgpu_cs_submit(struct amdgpu_cs_parser *p,
 	amdgpu_vm_move_to_lru_tail(p->adev, &fpriv->vm);
 
 	ttm_eu_fence_buffer_objects(&p->ticket, &p->validated, p->fence);
-	amdgpu_mn_unlock(p->mn);
+	mutex_unlock(&p->adev->notifier_lock);
 
 	return 0;
 
 error_abort:
 	drm_sched_job_cleanup(&job->base);
-	amdgpu_mn_unlock(p->mn);
+	mutex_unlock(&p->adev->notifier_lock);
 
 error_unlock:
 	amdgpu_job_free(job);
@@ -1359,6 +1285,9 @@ int amdgpu_cs_ioctl(struct drm_device *dev, void *data, struct drm_file *filp)
 	bool reserved_buffers = false;
 	int i, r;
 
+	if (amdgpu_ras_intr_triggered())
+		return -EHWPOISON;
+
 	if (!adev->accel_working)
 		return -EBUSY;
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_csa.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_csa.c
index 35a8d3c96fc9..08047bc4d588 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_csa.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_csa.c
@@ -80,7 +80,7 @@ int amdgpu_map_static_csa(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 	list_add(&csa_tv.head, &list);
 	amdgpu_vm_get_pd_bo(vm, &list, &pd);
 
-	r = ttm_eu_reserve_buffers(&ticket, &list, true, NULL, false);
+	r = ttm_eu_reserve_buffers(&ticket, &list, true, NULL);
 	if (r) {
 		DRM_ERROR("failed to reserve CSA,PD BOs: err=%d\n", r);
 		return r;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
index 6614d8a6f4c8..2cdaf3b2a721 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
@@ -604,8 +604,11 @@ void amdgpu_ctx_mgr_entity_fini(struct amdgpu_ctx_mgr *mgr)
 			continue;
 		}
 
-		for (i = 0; i < num_entities; i++)
+		for (i = 0; i < num_entities; i++) {
+			mutex_lock(&ctx->adev->lock_reset);
 			drm_sched_entity_fini(&ctx->entities[0][i].entity);
+			mutex_unlock(&ctx->adev->lock_reset);
+		}
 	}
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c
index 5652cc72ed3a..8e6726e0d035 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c
@@ -859,6 +859,9 @@ static int amdgpu_debugfs_test_ib(struct seq_file *m, void *data)
 	struct amdgpu_device *adev = dev->dev_private;
 	int r = 0, i;
 
+	/* Avoid accidently unparking the sched thread during GPU reset */
+	mutex_lock(&adev->lock_reset);
+
 	/* hold on the scheduler */
 	for (i = 0; i < AMDGPU_MAX_RINGS; i++) {
 		struct amdgpu_ring *ring = adev->rings[i];
@@ -884,6 +887,8 @@ static int amdgpu_debugfs_test_ib(struct seq_file *m, void *data)
 		kthread_unpark(ring->sched.thread);
 	}
 
+	mutex_unlock(&adev->lock_reset);
+
 	return 0;
 }
 
@@ -1036,6 +1041,9 @@ static int amdgpu_debugfs_ib_preempt(void *data, u64 val)
 	if (!fences)
 		return -ENOMEM;
 
+	/* Avoid accidently unparking the sched thread during GPU reset */
+	mutex_lock(&adev->lock_reset);
+
 	/* stop the scheduler */
 	kthread_park(ring->sched.thread);
 
@@ -1075,10 +1083,11 @@ failure:
 	/* restart the scheduler */
 	kthread_unpark(ring->sched.thread);
 
+	mutex_unlock(&adev->lock_reset);
+
 	ttm_bo_unlock_delayed_workqueue(&adev->mman.bdev, resched);
 
-	if (fences)
-		kfree(fences);
+	kfree(fences);
 
 	return 0;
 }
@@ -1090,8 +1099,8 @@ int amdgpu_debugfs_init(struct amdgpu_device *adev)
 {
 	adev->debugfs_preempt =
 		debugfs_create_file("amdgpu_preempt_ib", 0600,
-				    adev->ddev->primary->debugfs_root,
-				    (void *)adev, &fops_ib_preempt);
+				    adev->ddev->primary->debugfs_root, adev,
+				    &fops_ib_preempt);
 	if (!(adev->debugfs_preempt)) {
 		DRM_ERROR("unable to create amdgpu_preempt_ib debugsfs file\n");
 		return -EIO;
@@ -1103,8 +1112,7 @@ int amdgpu_debugfs_init(struct amdgpu_device *adev)
 
 void amdgpu_debugfs_preempt_cleanup(struct amdgpu_device *adev)
 {
-	if (adev->debugfs_preempt)
-		debugfs_remove(adev->debugfs_preempt);
+	debugfs_remove(adev->debugfs_preempt);
 }
 
 #else
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index 5a1939dbd4e3..c17505fba988 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -65,6 +65,8 @@
 #include "amdgpu_ras.h"
 #include "amdgpu_pmu.h"
 
+#include <linux/suspend.h>
+
 MODULE_FIRMWARE("amdgpu/vega10_gpu_info.bin");
 MODULE_FIRMWARE("amdgpu/vega12_gpu_info.bin");
 MODULE_FIRMWARE("amdgpu/raven_gpu_info.bin");
@@ -78,7 +80,7 @@ MODULE_FIRMWARE("amdgpu/navi12_gpu_info.bin");
 
 #define AMDGPU_RESUME_MS		2000
 
-static const char *amdgpu_asic_name[] = {
+const char *amdgpu_asic_name[] = {
 	"TAHITI",
 	"PITCAIRN",
 	"VERDE",
@@ -151,6 +153,36 @@ bool amdgpu_device_is_px(struct drm_device *dev)
 	return false;
 }
 
+/**
+ * VRAM access helper functions.
+ *
+ * amdgpu_device_vram_access - read/write a buffer in vram
+ *
+ * @adev: amdgpu_device pointer
+ * @pos: offset of the buffer in vram
+ * @buf: virtual address of the buffer in system memory
+ * @size: read/write size, sizeof(@buf) must > @size
+ * @write: true - write to vram, otherwise - read from vram
+ */
+void amdgpu_device_vram_access(struct amdgpu_device *adev, loff_t pos,
+			       uint32_t *buf, size_t size, bool write)
+{
+	uint64_t last;
+	unsigned long flags;
+
+	last = size - 4;
+	for (last += pos; pos <= last; pos += 4) {
+		spin_lock_irqsave(&adev->mmio_idx_lock, flags);
+		WREG32_NO_KIQ(mmMM_INDEX, ((uint32_t)pos) | 0x80000000);
+		WREG32_NO_KIQ(mmMM_INDEX_HI, pos >> 31);
+		if (write)
+			WREG32_NO_KIQ(mmMM_DATA, *buf++);
+		else
+			*buf++ = RREG32_NO_KIQ(mmMM_DATA);
+		spin_unlock_irqrestore(&adev->mmio_idx_lock, flags);
+	}
+}
+
 /*
  * MMIO register access helper functions.
  */
@@ -1023,12 +1055,6 @@ static int amdgpu_device_check_arguments(struct amdgpu_device *adev)
 
 	amdgpu_device_check_block_size(adev);
 
-	ret = amdgpu_device_get_job_timeout_settings(adev);
-	if (ret) {
-		dev_err(adev->dev, "invalid lockup_timeout parameter syntax\n");
-		return ret;
-	}
-
 	adev->firmware.load_type = amdgpu_ucode_get_load_type(adev, amdgpu_fw_load_type);
 
 	return ret;
@@ -1469,6 +1495,9 @@ static int amdgpu_device_parse_gpu_info_fw(struct amdgpu_device *adev)
 			(const struct gpu_info_firmware_v1_0 *)(adev->firmware.gpu_info_fw->data +
 								le32_to_cpu(hdr->header.ucode_array_offset_bytes));
 
+		if (amdgpu_discovery && adev->asic_type >= CHIP_NAVI10)
+			goto parse_soc_bounding_box;
+
 		adev->gfx.config.max_shader_engines = le32_to_cpu(gpu_info_fw->gc_num_se);
 		adev->gfx.config.max_cu_per_sh = le32_to_cpu(gpu_info_fw->gc_num_cu_per_sh);
 		adev->gfx.config.max_sh_per_se = le32_to_cpu(gpu_info_fw->gc_num_sh_per_se);
@@ -1496,7 +1525,13 @@ static int amdgpu_device_parse_gpu_info_fw(struct amdgpu_device *adev)
 			adev->gfx.config.num_packer_per_sc =
 				le32_to_cpu(gpu_info_fw->num_packer_per_sc);
 		}
+
+parse_soc_bounding_box:
 #ifdef CONFIG_DRM_AMD_DC_DCN2_0
+		/*
+		 * soc bounding box info is not integrated in disocovery table,
+		 * we always need to parse it from gpu info firmware.
+		 */
 		if (hdr->version_minor == 2) {
 			const struct gpu_info_firmware_v1_2 *gpu_info_fw =
 				(const struct gpu_info_firmware_v1_2 *)(adev->firmware.gpu_info_fw->data +
@@ -1613,6 +1648,9 @@ static int amdgpu_device_ip_early_init(struct amdgpu_device *adev)
 	if (r)
 		return r;
 
+	if (amdgpu_discovery && adev->asic_type >= CHIP_NAVI10)
+		amdgpu_discovery_get_gfx_info(adev);
+
 	amdgpu_amdkfd_device_probe(adev);
 
 	if (amdgpu_sriov_vf(adev)) {
@@ -1622,7 +1660,7 @@ static int amdgpu_device_ip_early_init(struct amdgpu_device *adev)
 	}
 
 	adev->pm.pp_feature = amdgpu_pp_feature_mask;
-	if (amdgpu_sriov_vf(adev))
+	if (amdgpu_sriov_vf(adev) || sched_policy == KFD_SCHED_POLICY_NO_HWS)
 		adev->pm.pp_feature &= ~PP_GFXOFF_MASK;
 
 	for (i = 0; i < adev->num_ip_blocks; i++) {
@@ -1839,6 +1877,19 @@ static int amdgpu_device_ip_init(struct amdgpu_device *adev)
 	if (r)
 		goto init_failed;
 
+	/*
+	 * retired pages will be loaded from eeprom and reserved here,
+	 * it should be called after amdgpu_device_ip_hw_init_phase2  since
+	 * for some ASICs the RAS EEPROM code relies on SMU fully functioning
+	 * for I2C communication which only true at this point.
+	 * recovery_init may fail, but it can free all resources allocated by
+	 * itself and its failure should not stop amdgpu init process.
+	 *
+	 * Note: theoretically, this should be called before all vram allocations
+	 * to protect retired page from abusing
+	 */
+	amdgpu_ras_recovery_init(adev);
+
 	if (adev->gmc.xgmi.num_physical_nodes > 1)
 		amdgpu_xgmi_add_device(adev);
 	amdgpu_amdkfd_device_init(adev);
@@ -2006,6 +2057,7 @@ out:
  */
 static int amdgpu_device_ip_late_init(struct amdgpu_device *adev)
 {
+	struct amdgpu_gpu_instance *gpu_instance;
 	int i = 0, r;
 
 	for (i = 0; i < adev->num_ip_blocks; i++) {
@@ -2031,8 +2083,39 @@ static int amdgpu_device_ip_late_init(struct amdgpu_device *adev)
 	if (r)
 		DRM_ERROR("enable mgpu fan boost failed (%d).\n", r);
 
-	/* set to low pstate by default */
-	amdgpu_xgmi_set_pstate(adev, 0);
+
+	if (adev->gmc.xgmi.num_physical_nodes > 1) {
+		mutex_lock(&mgpu_info.mutex);
+
+		/*
+		 * Reset device p-state to low as this was booted with high.
+		 *
+		 * This should be performed only after all devices from the same
+		 * hive get initialized.
+		 *
+		 * However, it's unknown how many device in the hive in advance.
+		 * As this is counted one by one during devices initializations.
+		 *
+		 * So, we wait for all XGMI interlinked devices initialized.
+		 * This may bring some delays as those devices may come from
+		 * different hives. But that should be OK.
+		 */
+		if (mgpu_info.num_dgpu == adev->gmc.xgmi.num_physical_nodes) {
+			for (i = 0; i < mgpu_info.num_gpu; i++) {
+				gpu_instance = &(mgpu_info.gpu_ins[i]);
+				if (gpu_instance->adev->flags & AMD_IS_APU)
+					continue;
+
+				r = amdgpu_xgmi_set_pstate(gpu_instance->adev, 0);
+				if (r) {
+					DRM_ERROR("pstate setting failed (%d).\n", r);
+					break;
+				}
+			}
+		}
+
+		mutex_unlock(&mgpu_info.mutex);
+	}
 
 	return 0;
 }
@@ -2220,6 +2303,12 @@ static int amdgpu_device_ip_suspend_phase2(struct amdgpu_device *adev)
 		/* displays are handled in phase1 */
 		if (adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_DCE)
 			continue;
+		/* PSP lost connection when err_event_athub occurs */
+		if (amdgpu_ras_intr_triggered() &&
+		    adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_PSP) {
+			adev->ip_blocks[i].status.hw = false;
+			continue;
+		}
 		/* XXX handle errors */
 		r = adev->ip_blocks[i].version->funcs->suspend(adev);
 		/* XXX handle errors */
@@ -2231,17 +2320,17 @@ static int amdgpu_device_ip_suspend_phase2(struct amdgpu_device *adev)
 		/* handle putting the SMC in the appropriate state */
 		if (adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_SMC) {
 			if (is_support_sw_smu(adev)) {
-				/* todo */
+				r = smu_set_mp1_state(&adev->smu, adev->mp1_state);
 			} else if (adev->powerplay.pp_funcs &&
 					   adev->powerplay.pp_funcs->set_mp1_state) {
 				r = adev->powerplay.pp_funcs->set_mp1_state(
 					adev->powerplay.pp_handle,
 					adev->mp1_state);
-				if (r) {
-					DRM_ERROR("SMC failed to set mp1 state %d, %d\n",
-						  adev->mp1_state, r);
-					return r;
-				}
+			}
+			if (r) {
+				DRM_ERROR("SMC failed to set mp1 state %d, %d\n",
+					  adev->mp1_state, r);
+				return r;
 			}
 		}
 
@@ -2556,6 +2645,73 @@ static void amdgpu_device_xgmi_reset_func(struct work_struct *__work)
 			 adev->asic_reset_res, adev->ddev->unique);
 }
 
+static int amdgpu_device_get_job_timeout_settings(struct amdgpu_device *adev)
+{
+	char *input = amdgpu_lockup_timeout;
+	char *timeout_setting = NULL;
+	int index = 0;
+	long timeout;
+	int ret = 0;
+
+	/*
+	 * By default timeout for non compute jobs is 10000.
+	 * And there is no timeout enforced on compute jobs.
+	 * In SR-IOV or passthrough mode, timeout for compute
+	 * jobs are 10000 by default.
+	 */
+	adev->gfx_timeout = msecs_to_jiffies(10000);
+	adev->sdma_timeout = adev->video_timeout = adev->gfx_timeout;
+	if (amdgpu_sriov_vf(adev) || amdgpu_passthrough(adev))
+		adev->compute_timeout = adev->gfx_timeout;
+	else
+		adev->compute_timeout = MAX_SCHEDULE_TIMEOUT;
+
+	if (strnlen(input, AMDGPU_MAX_TIMEOUT_PARAM_LENGTH)) {
+		while ((timeout_setting = strsep(&input, ",")) &&
+				strnlen(timeout_setting, AMDGPU_MAX_TIMEOUT_PARAM_LENGTH)) {
+			ret = kstrtol(timeout_setting, 0, &timeout);
+			if (ret)
+				return ret;
+
+			if (timeout == 0) {
+				index++;
+				continue;
+			} else if (timeout < 0) {
+				timeout = MAX_SCHEDULE_TIMEOUT;
+			} else {
+				timeout = msecs_to_jiffies(timeout);
+			}
+
+			switch (index++) {
+			case 0:
+				adev->gfx_timeout = timeout;
+				break;
+			case 1:
+				adev->compute_timeout = timeout;
+				break;
+			case 2:
+				adev->sdma_timeout = timeout;
+				break;
+			case 3:
+				adev->video_timeout = timeout;
+				break;
+			default:
+				break;
+			}
+		}
+		/*
+		 * There is only one value specified and
+		 * it should apply to all non-compute jobs.
+		 */
+		if (index == 1) {
+			adev->sdma_timeout = adev->video_timeout = adev->gfx_timeout;
+			if (amdgpu_sriov_vf(adev) || amdgpu_passthrough(adev))
+				adev->compute_timeout = adev->gfx_timeout;
+		}
+	}
+
+	return ret;
+}
 
 /**
  * amdgpu_device_init - initialize the driver
@@ -2583,7 +2739,12 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 	adev->ddev = ddev;
 	adev->pdev = pdev;
 	adev->flags = flags;
-	adev->asic_type = flags & AMD_ASIC_MASK;
+
+	if (amdgpu_force_asic_type >= 0 && amdgpu_force_asic_type < CHIP_LAST)
+		adev->asic_type = amdgpu_force_asic_type;
+	else
+		adev->asic_type = flags & AMD_ASIC_MASK;
+
 	adev->usec_timeout = AMDGPU_MAX_USEC_TIMEOUT;
 	if (amdgpu_emu_mode == 1)
 		adev->usec_timeout *= 2;
@@ -2633,6 +2794,7 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 	mutex_init(&adev->virt.vf_errors.lock);
 	hash_init(adev->mn_hash);
 	mutex_init(&adev->lock_reset);
+	mutex_init(&adev->notifier_lock);
 	mutex_init(&adev->virt.dpm_mutex);
 	mutex_init(&adev->psp.mutex);
 
@@ -2726,6 +2888,12 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 	if (r)
 		return r;
 
+	r = amdgpu_device_get_job_timeout_settings(adev);
+	if (r) {
+		dev_err(adev->dev, "invalid lockup_timeout parameter syntax\n");
+		return r;
+	}
+
 	/* doorbell bar mapping and doorbell index init*/
 	amdgpu_device_doorbell_init(adev);
 
@@ -2885,6 +3053,13 @@ fence_driver_init:
 			DRM_INFO("amdgpu: acceleration disabled, skipping benchmarks\n");
 	}
 
+	/*
+	 * Register gpu instance before amdgpu_device_enable_mgpu_fan_boost.
+	 * Otherwise the mgpu fan boost feature will be skipped due to the
+	 * gpu instance is counted less.
+	 */
+	amdgpu_register_gpu_instance(adev);
+
 	/* enable clockgating, etc. after ib tests, etc. since some blocks require
 	 * explicit gating rather than handling it automatically.
 	 */
@@ -2935,7 +3110,9 @@ void amdgpu_device_fini(struct amdgpu_device *adev)
 	int r;
 
 	DRM_INFO("amdgpu: finishing device.\n");
+	flush_delayed_work(&adev->delayed_init_work);
 	adev->shutdown = true;
+
 	/* disable all interrupts */
 	amdgpu_irq_disable_all(adev);
 	if (adev->mode_info.mode_config_initialized){
@@ -2953,7 +3130,6 @@ void amdgpu_device_fini(struct amdgpu_device *adev)
 		adev->firmware.gpu_info_fw = NULL;
 	}
 	adev->accel_working = false;
-	cancel_delayed_work_sync(&adev->delayed_init_work);
 	/* free i2c buses */
 	if (!amdgpu_device_has_dc_support(adev))
 		amdgpu_i2c_fini(adev);
@@ -3007,6 +3183,7 @@ int amdgpu_device_suspend(struct drm_device *dev, bool suspend, bool fbcon)
 	struct amdgpu_device *adev;
 	struct drm_crtc *crtc;
 	struct drm_connector *connector;
+	struct drm_connector_list_iter iter;
 	int r;
 
 	if (dev == NULL || dev->dev_private == NULL) {
@@ -3029,9 +3206,11 @@ int amdgpu_device_suspend(struct drm_device *dev, bool suspend, bool fbcon)
 	if (!amdgpu_device_has_dc_support(adev)) {
 		/* turn off display hw */
 		drm_modeset_lock_all(dev);
-		list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
-			drm_helper_connector_dpms(connector, DRM_MODE_DPMS_OFF);
-		}
+		drm_connector_list_iter_begin(dev, &iter);
+		drm_for_each_connector_iter(connector, &iter)
+			drm_helper_connector_dpms(connector,
+						  DRM_MODE_DPMS_OFF);
+		drm_connector_list_iter_end(&iter);
 		drm_modeset_unlock_all(dev);
 			/* unpin the front buffers and cursors */
 		list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {
@@ -3082,15 +3261,11 @@ int amdgpu_device_suspend(struct drm_device *dev, bool suspend, bool fbcon)
 	 */
 	amdgpu_bo_evict_vram(adev);
 
-	pci_save_state(dev->pdev);
 	if (suspend) {
+		pci_save_state(dev->pdev);
 		/* Shut down the device */
 		pci_disable_device(dev->pdev);
 		pci_set_power_state(dev->pdev, PCI_D3hot);
-	} else {
-		r = amdgpu_asic_reset(adev);
-		if (r)
-			DRM_ERROR("amdgpu asic reset failed\n");
 	}
 
 	return 0;
@@ -3110,6 +3285,7 @@ int amdgpu_device_suspend(struct drm_device *dev, bool suspend, bool fbcon)
 int amdgpu_device_resume(struct drm_device *dev, bool resume, bool fbcon)
 {
 	struct drm_connector *connector;
+	struct drm_connector_list_iter iter;
 	struct amdgpu_device *adev = dev->dev_private;
 	struct drm_crtc *crtc;
 	int r = 0;
@@ -3180,9 +3356,13 @@ int amdgpu_device_resume(struct drm_device *dev, bool resume, bool fbcon)
 
 			/* turn on display hw */
 			drm_modeset_lock_all(dev);
-			list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
-				drm_helper_connector_dpms(connector, DRM_MODE_DPMS_ON);
-			}
+
+			drm_connector_list_iter_begin(dev, &iter);
+			drm_for_each_connector_iter(connector, &iter)
+				drm_helper_connector_dpms(connector,
+							  DRM_MODE_DPMS_ON);
+			drm_connector_list_iter_end(&iter);
+
 			drm_modeset_unlock_all(dev);
 		}
 		amdgpu_fbdev_set_suspend(adev, 0);
@@ -3628,11 +3808,6 @@ static int amdgpu_do_asic_reset(struct amdgpu_hive_info *hive,
 						break;
 				}
 			}
-
-			list_for_each_entry(tmp_adev, device_list_handle,
-					gmc.xgmi.head) {
-				amdgpu_ras_reserve_bad_pages(tmp_adev);
-			}
 		}
 	}
 
@@ -3736,25 +3911,18 @@ static bool amdgpu_device_lock_adev(struct amdgpu_device *adev, bool trylock)
 		adev->mp1_state = PP_MP1_STATE_NONE;
 		break;
 	}
-	/* Block kfd: SRIOV would do it separately */
-	if (!amdgpu_sriov_vf(adev))
-                amdgpu_amdkfd_pre_reset(adev);
 
 	return true;
 }
 
 static void amdgpu_device_unlock_adev(struct amdgpu_device *adev)
 {
-	/*unlock kfd: SRIOV would do it separately */
-	if (!amdgpu_sriov_vf(adev))
-                amdgpu_amdkfd_post_reset(adev);
 	amdgpu_vf_error_trans_all(adev);
 	adev->mp1_state = PP_MP1_STATE_NONE;
 	adev->in_gpu_reset = 0;
 	mutex_unlock(&adev->lock_reset);
 }
 
-
 /**
  * amdgpu_device_gpu_recover - reset the asic and recover scheduler
  *
@@ -3774,11 +3942,24 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev,
 	struct amdgpu_hive_info *hive = NULL;
 	struct amdgpu_device *tmp_adev = NULL;
 	int i, r = 0;
+	bool in_ras_intr = amdgpu_ras_intr_triggered();
+
+	/*
+	 * Flush RAM to disk so that after reboot
+	 * the user can read log and see why the system rebooted.
+	 */
+	if (in_ras_intr && amdgpu_ras_get_context(adev)->reboot) {
+
+		DRM_WARN("Emergency reboot.");
+
+		ksys_sync_helper();
+		emergency_restart();
+	}
 
 	need_full_reset = job_signaled = false;
 	INIT_LIST_HEAD(&device_list);
 
-	dev_info(adev->dev, "GPU reset begin!\n");
+	dev_info(adev->dev, "GPU %s begin!\n", in_ras_intr ? "jobs stop":"reset");
 
 	cancel_delayed_work_sync(&adev->delayed_init_work);
 
@@ -3805,9 +3986,16 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev,
 		return 0;
 	}
 
+	/* Block kfd: SRIOV would do it separately */
+	if (!amdgpu_sriov_vf(adev))
+                amdgpu_amdkfd_pre_reset(adev);
+
 	/* Build list of devices to reset */
 	if  (adev->gmc.xgmi.num_physical_nodes > 1) {
 		if (!hive) {
+			/*unlock kfd: SRIOV would do it separately */
+			if (!amdgpu_sriov_vf(adev))
+		                amdgpu_amdkfd_post_reset(adev);
 			amdgpu_device_unlock_adev(adev);
 			return -ENODEV;
 		}
@@ -3823,17 +4011,22 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev,
 		device_list_handle = &device_list;
 	}
 
-	/*
-	 * Mark these ASICs to be reseted as untracked first
-	 * And add them back after reset completed
-	 */
-	list_for_each_entry(tmp_adev, device_list_handle, gmc.xgmi.head)
-		amdgpu_unregister_gpu_instance(tmp_adev);
-
 	/* block all schedulers and reset given job's ring */
 	list_for_each_entry(tmp_adev, device_list_handle, gmc.xgmi.head) {
+		if (tmp_adev != adev) {
+			amdgpu_device_lock_adev(tmp_adev, false);
+			if (!amdgpu_sriov_vf(tmp_adev))
+			                amdgpu_amdkfd_pre_reset(tmp_adev);
+		}
+
+		/*
+		 * Mark these ASICs to be reseted as untracked first
+		 * And add them back after reset completed
+		 */
+		amdgpu_unregister_gpu_instance(tmp_adev);
+
 		/* disable ras on ALL IPs */
-		if (amdgpu_device_ip_need_full_reset(tmp_adev))
+		if (!in_ras_intr && amdgpu_device_ip_need_full_reset(tmp_adev))
 			amdgpu_ras_suspend(tmp_adev);
 
 		for (i = 0; i < AMDGPU_MAX_RINGS; ++i) {
@@ -3843,10 +4036,16 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev,
 				continue;
 
 			drm_sched_stop(&ring->sched, job ? &job->base : NULL);
+
+			if (in_ras_intr)
+				amdgpu_job_stop_all_jobs_on_sched(&ring->sched);
 		}
 	}
 
 
+	if (in_ras_intr)
+		goto skip_sched_resume;
+
 	/*
 	 * Must check guilty signal here since after this point all old
 	 * HW fences are force signaled.
@@ -3857,9 +4056,6 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev,
 	    dma_fence_is_signaled(job->base.s_fence->parent))
 		job_signaled = true;
 
-	if (!amdgpu_device_ip_need_full_reset(adev))
-		device_list_handle = &device_list;
-
 	if (job_signaled) {
 		dev_info(adev->dev, "Guilty job already signaled, skipping HW reset");
 		goto skip_hw_reset;
@@ -3881,7 +4077,6 @@ retry:	/* Rest of adevs pre asic reset from XGMI hive. */
 		if (tmp_adev == adev)
 			continue;
 
-		amdgpu_device_lock_adev(tmp_adev, false);
 		r = amdgpu_device_pre_asic_reset(tmp_adev,
 						 NULL,
 						 &need_full_reset);
@@ -3909,6 +4104,7 @@ skip_hw_reset:
 
 	/* Post ASIC reset for all devs .*/
 	list_for_each_entry(tmp_adev, device_list_handle, gmc.xgmi.head) {
+
 		for (i = 0; i < AMDGPU_MAX_RINGS; ++i) {
 			struct amdgpu_ring *ring = tmp_adev->rings[i];
 
@@ -3930,12 +4126,18 @@ skip_hw_reset:
 
 		if (r) {
 			/* bad news, how to tell it to userspace ? */
-			dev_info(tmp_adev->dev, "GPU reset(%d) failed\n", atomic_read(&adev->gpu_reset_counter));
+			dev_info(tmp_adev->dev, "GPU reset(%d) failed\n", atomic_read(&tmp_adev->gpu_reset_counter));
 			amdgpu_vf_error_put(tmp_adev, AMDGIM_ERROR_VF_GPU_RESET_FAIL, 0, r);
 		} else {
-			dev_info(tmp_adev->dev, "GPU reset(%d) succeeded!\n", atomic_read(&adev->gpu_reset_counter));
+			dev_info(tmp_adev->dev, "GPU reset(%d) succeeded!\n", atomic_read(&tmp_adev->gpu_reset_counter));
 		}
+	}
 
+skip_sched_resume:
+	list_for_each_entry(tmp_adev, device_list_handle, gmc.xgmi.head) {
+		/*unlock kfd: SRIOV would do it separately */
+		if (!in_ras_intr && !amdgpu_sriov_vf(tmp_adev))
+	                amdgpu_amdkfd_post_reset(tmp_adev);
 		amdgpu_device_unlock_adev(tmp_adev);
 	}
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c
index 1481899f86c1..f95092741c38 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c
@@ -134,20 +134,10 @@ static int hw_id_map[MAX_HWIP] = {
 
 static int amdgpu_discovery_read_binary(struct amdgpu_device *adev, uint8_t *binary)
 {
-	uint32_t *p = (uint32_t *)binary;
 	uint64_t vram_size = (uint64_t)RREG32(mmRCC_CONFIG_MEMSIZE) << 20;
-	uint64_t pos = vram_size - BINARY_MAX_SIZE;
-	unsigned long flags;
-
-	while (pos < vram_size) {
-		spin_lock_irqsave(&adev->mmio_idx_lock, flags);
-		WREG32_NO_KIQ(mmMM_INDEX, ((uint32_t)pos) | 0x80000000);
-		WREG32_NO_KIQ(mmMM_INDEX_HI, pos >> 31);
-		*p++ = RREG32_NO_KIQ(mmMM_DATA);
-		spin_unlock_irqrestore(&adev->mmio_idx_lock, flags);
-		pos += 4;
-	}
+	uint64_t pos = vram_size - DISCOVERY_TMR_SIZE;
 
+	amdgpu_device_vram_access(adev, pos, (uint32_t *)binary, DISCOVERY_TMR_SIZE, false);
 	return 0;
 }
 
@@ -179,7 +169,7 @@ int amdgpu_discovery_init(struct amdgpu_device *adev)
 	uint16_t checksum;
 	int r;
 
-	adev->discovery = kzalloc(BINARY_MAX_SIZE, GFP_KERNEL);
+	adev->discovery = kzalloc(DISCOVERY_TMR_SIZE, GFP_KERNEL);
 	if (!adev->discovery)
 		return -ENOMEM;
 
@@ -333,7 +323,7 @@ int amdgpu_discovery_reg_base_init(struct amdgpu_device *adev)
 }
 
 int amdgpu_discovery_get_ip_version(struct amdgpu_device *adev, int hw_id,
-				    int *major, int *minor)
+				    int *major, int *minor, int *revision)
 {
 	struct binary_header *bhdr;
 	struct ip_discovery_header *ihdr;
@@ -369,6 +359,8 @@ int amdgpu_discovery_get_ip_version(struct amdgpu_device *adev, int hw_id,
 					*major = ip->major;
 				if (minor)
 					*minor = ip->minor;
+				if (revision)
+					*revision = ip->revision;
 				return 0;
 			}
 			ip_offset += sizeof(*ip) + 4 * (ip->num_base_address - 1);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.h
index 85b8c4d4d576..ba78e15d9b05 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.h
@@ -24,11 +24,13 @@
 #ifndef __AMDGPU_DISCOVERY__
 #define __AMDGPU_DISCOVERY__
 
+#define DISCOVERY_TMR_SIZE  (64 << 10)
+
 int amdgpu_discovery_init(struct amdgpu_device *adev);
 void amdgpu_discovery_fini(struct amdgpu_device *adev);
 int amdgpu_discovery_reg_base_init(struct amdgpu_device *adev);
 int amdgpu_discovery_get_ip_version(struct amdgpu_device *adev, int hw_id,
-                                    int *major, int *minor);
+                                    int *major, int *minor, int *revision);
 int amdgpu_discovery_get_gfx_info(struct amdgpu_device *adev);
 
 #endif /* __AMDGPU_DISCOVERY__ */
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c
index 1d4aaa9580f4..3cadb0b76f22 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c
@@ -370,11 +370,13 @@ void amdgpu_display_print_display_setup(struct drm_device *dev)
 	struct amdgpu_connector *amdgpu_connector;
 	struct drm_encoder *encoder;
 	struct amdgpu_encoder *amdgpu_encoder;
+	struct drm_connector_list_iter iter;
 	uint32_t devices;
 	int i = 0;
 
+	drm_connector_list_iter_begin(dev, &iter);
 	DRM_INFO("AMDGPU Display Connectors\n");
-	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
+	drm_for_each_connector_iter(connector, &iter) {
 		amdgpu_connector = to_amdgpu_connector(connector);
 		DRM_INFO("Connector %d:\n", i);
 		DRM_INFO("  %s\n", connector->name);
@@ -438,6 +440,7 @@ void amdgpu_display_print_display_setup(struct drm_device *dev)
 		}
 		i++;
 	}
+	drm_connector_list_iter_end(&iter);
 }
 
 /**
@@ -511,7 +514,7 @@ uint32_t amdgpu_display_supported_domains(struct amdgpu_device *adev,
 	 * Also, don't allow GTT domain if the BO doens't have USWC falg set.
 	 */
 	if (adev->asic_type >= CHIP_CARRIZO &&
-	    adev->asic_type <= CHIP_RAVEN &&
+	    adev->asic_type < CHIP_RAVEN &&
 	    (adev->flags & AMD_IS_APU) &&
 	    (bo_flags & AMDGPU_GEM_CREATE_CPU_GTT_USWC) &&
 	    amdgpu_bo_support_uswc(bo_flags) &&
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c
index 61f108ec2b5c..e2eec7b66334 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c
@@ -34,27 +34,12 @@
 #include "amdgpu.h"
 #include "amdgpu_display.h"
 #include "amdgpu_gem.h"
+#include "amdgpu_dma_buf.h"
 #include <drm/amdgpu_drm.h>
 #include <linux/dma-buf.h>
 #include <linux/dma-fence-array.h>
 
 /**
- * amdgpu_gem_prime_get_sg_table - &drm_driver.gem_prime_get_sg_table
- * implementation
- * @obj: GEM buffer object (BO)
- *
- * Returns:
- * A scatter/gather table for the pinned pages of the BO's memory.
- */
-struct sg_table *amdgpu_gem_prime_get_sg_table(struct drm_gem_object *obj)
-{
-	struct amdgpu_bo *bo = gem_to_amdgpu_bo(obj);
-	int npages = bo->tbo.num_pages;
-
-	return drm_prime_pages_to_sg(bo->tbo.ttm->pages, npages);
-}
-
-/**
  * amdgpu_gem_prime_vmap - &dma_buf_ops.vmap implementation
  * @obj: GEM BO
  *
@@ -179,92 +164,126 @@ err_fences_put:
 }
 
 /**
- * amdgpu_dma_buf_map_attach - &dma_buf_ops.attach implementation
- * @dma_buf: Shared DMA buffer
+ * amdgpu_dma_buf_attach - &dma_buf_ops.attach implementation
+ *
+ * @dmabuf: DMA-buf where we attach to
+ * @attach: attachment to add
+ *
+ * Add the attachment as user to the exported DMA-buf.
+ */
+static int amdgpu_dma_buf_attach(struct dma_buf *dmabuf,
+				 struct dma_buf_attachment *attach)
+{
+	struct drm_gem_object *obj = dmabuf->priv;
+	struct amdgpu_bo *bo = gem_to_amdgpu_bo(obj);
+	struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
+	int r;
+
+	if (attach->dev->driver == adev->dev->driver)
+		return 0;
+
+	r = amdgpu_bo_reserve(bo, false);
+	if (unlikely(r != 0))
+		return r;
+
+	/*
+	 * We only create shared fences for internal use, but importers
+	 * of the dmabuf rely on exclusive fences for implicitly
+	 * tracking write hazards. As any of the current fences may
+	 * correspond to a write, we need to convert all existing
+	 * fences on the reservation object into a single exclusive
+	 * fence.
+	 */
+	r = __dma_resv_make_exclusive(bo->tbo.base.resv);
+	if (r)
+		return r;
+
+	bo->prime_shared_count++;
+	amdgpu_bo_unreserve(bo);
+	return 0;
+}
+
+/**
+ * amdgpu_dma_buf_detach - &dma_buf_ops.detach implementation
+ *
+ * @dmabuf: DMA-buf where we remove the attachment from
+ * @attach: the attachment to remove
+ *
+ * Called when an attachment is removed from the DMA-buf.
+ */
+static void amdgpu_dma_buf_detach(struct dma_buf *dmabuf,
+				  struct dma_buf_attachment *attach)
+{
+	struct drm_gem_object *obj = dmabuf->priv;
+	struct amdgpu_bo *bo = gem_to_amdgpu_bo(obj);
+	struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
+
+	if (attach->dev->driver != adev->dev->driver && bo->prime_shared_count)
+		bo->prime_shared_count--;
+}
+
+/**
+ * amdgpu_dma_buf_map - &dma_buf_ops.map_dma_buf implementation
  * @attach: DMA-buf attachment
+ * @dir: DMA direction
  *
  * Makes sure that the shared DMA buffer can be accessed by the target device.
  * For now, simply pins it to the GTT domain, where it should be accessible by
  * all DMA devices.
  *
  * Returns:
- * 0 on success or a negative error code on failure.
+ * sg_table filled with the DMA addresses to use or ERR_PRT with negative error
+ * code.
  */
-static int amdgpu_dma_buf_map_attach(struct dma_buf *dma_buf,
-				     struct dma_buf_attachment *attach)
+static struct sg_table *amdgpu_dma_buf_map(struct dma_buf_attachment *attach,
+					   enum dma_data_direction dir)
 {
+	struct dma_buf *dma_buf = attach->dmabuf;
 	struct drm_gem_object *obj = dma_buf->priv;
 	struct amdgpu_bo *bo = gem_to_amdgpu_bo(obj);
-	struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
+	struct sg_table *sgt;
 	long r;
 
-	r = drm_gem_map_attach(dma_buf, attach);
-	if (r)
-		return r;
-
-	r = amdgpu_bo_reserve(bo, false);
-	if (unlikely(r != 0))
-		goto error_detach;
-
-
-	if (attach->dev->driver != adev->dev->driver) {
-		/*
-		 * We only create shared fences for internal use, but importers
-		 * of the dmabuf rely on exclusive fences for implicitly
-		 * tracking write hazards. As any of the current fences may
-		 * correspond to a write, we need to convert all existing
-		 * fences on the reservation object into a single exclusive
-		 * fence.
-		 */
-		r = __dma_resv_make_exclusive(bo->tbo.base.resv);
-		if (r)
-			goto error_unreserve;
-	}
-
-	/* pin buffer into GTT */
 	r = amdgpu_bo_pin(bo, AMDGPU_GEM_DOMAIN_GTT);
 	if (r)
-		goto error_unreserve;
+		return ERR_PTR(r);
 
-	if (attach->dev->driver != adev->dev->driver)
-		bo->prime_shared_count++;
+	sgt = drm_prime_pages_to_sg(bo->tbo.ttm->pages, bo->tbo.num_pages);
+	if (IS_ERR(sgt))
+		return sgt;
 
-error_unreserve:
-	amdgpu_bo_unreserve(bo);
+	if (!dma_map_sg_attrs(attach->dev, sgt->sgl, sgt->nents, dir,
+			      DMA_ATTR_SKIP_CPU_SYNC))
+		goto error_free;
 
-error_detach:
-	if (r)
-		drm_gem_map_detach(dma_buf, attach);
-	return r;
+	return sgt;
+
+error_free:
+	sg_free_table(sgt);
+	kfree(sgt);
+	return ERR_PTR(-ENOMEM);
 }
 
 /**
- * amdgpu_dma_buf_map_detach - &dma_buf_ops.detach implementation
- * @dma_buf: Shared DMA buffer
+ * amdgpu_dma_buf_unmap - &dma_buf_ops.unmap_dma_buf implementation
  * @attach: DMA-buf attachment
+ * @sgt: sg_table to unmap
+ * @dir: DMA direction
  *
  * This is called when a shared DMA buffer no longer needs to be accessible by
  * another device. For now, simply unpins the buffer from GTT.
  */
-static void amdgpu_dma_buf_map_detach(struct dma_buf *dma_buf,
-				      struct dma_buf_attachment *attach)
+static void amdgpu_dma_buf_unmap(struct dma_buf_attachment *attach,
+				 struct sg_table *sgt,
+				 enum dma_data_direction dir)
 {
-	struct drm_gem_object *obj = dma_buf->priv;
+	struct drm_gem_object *obj = attach->dmabuf->priv;
 	struct amdgpu_bo *bo = gem_to_amdgpu_bo(obj);
-	struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
-	int ret = 0;
-
-	ret = amdgpu_bo_reserve(bo, true);
-	if (unlikely(ret != 0))
-		goto error;
 
+	dma_unmap_sg(attach->dev, sgt->sgl, sgt->nents, dir);
+	sg_free_table(sgt);
+	kfree(sgt);
 	amdgpu_bo_unpin(bo);
-	if (attach->dev->driver != adev->dev->driver && bo->prime_shared_count)
-		bo->prime_shared_count--;
-	amdgpu_bo_unreserve(bo);
-
-error:
-	drm_gem_map_detach(dma_buf, attach);
 }
 
 /**
@@ -308,10 +327,11 @@ static int amdgpu_dma_buf_begin_cpu_access(struct dma_buf *dma_buf,
 }
 
 const struct dma_buf_ops amdgpu_dmabuf_ops = {
-	.attach = amdgpu_dma_buf_map_attach,
-	.detach = amdgpu_dma_buf_map_detach,
-	.map_dma_buf = drm_gem_map_dma_buf,
-	.unmap_dma_buf = drm_gem_unmap_dma_buf,
+	.dynamic_mapping = true,
+	.attach = amdgpu_dma_buf_attach,
+	.detach = amdgpu_dma_buf_detach,
+	.map_dma_buf = amdgpu_dma_buf_map,
+	.unmap_dma_buf = amdgpu_dma_buf_unmap,
 	.release = drm_gem_dmabuf_release,
 	.begin_cpu_access = amdgpu_dma_buf_begin_cpu_access,
 	.mmap = drm_gem_dmabuf_mmap,
@@ -321,7 +341,6 @@ const struct dma_buf_ops amdgpu_dmabuf_ops = {
 
 /**
  * amdgpu_gem_prime_export - &drm_driver.gem_prime_export implementation
- * @dev: DRM device
  * @gobj: GEM BO
  * @flags: Flags such as DRM_CLOEXEC and DRM_RDWR.
  *
@@ -350,31 +369,28 @@ struct dma_buf *amdgpu_gem_prime_export(struct drm_gem_object *gobj,
 }
 
 /**
- * amdgpu_gem_prime_import_sg_table - &drm_driver.gem_prime_import_sg_table
- * implementation
+ * amdgpu_dma_buf_create_obj - create BO for DMA-buf import
+ *
  * @dev: DRM device
- * @attach: DMA-buf attachment
- * @sg: Scatter/gather table
+ * @dma_buf: DMA-buf
  *
- * Imports shared DMA buffer memory exported by another device.
+ * Creates an empty SG BO for DMA-buf import.
  *
  * Returns:
  * A new GEM BO of the given DRM device, representing the memory
  * described by the given DMA-buf attachment and scatter/gather table.
  */
-struct drm_gem_object *
-amdgpu_gem_prime_import_sg_table(struct drm_device *dev,
-				 struct dma_buf_attachment *attach,
-				 struct sg_table *sg)
+static struct drm_gem_object *
+amdgpu_dma_buf_create_obj(struct drm_device *dev, struct dma_buf *dma_buf)
 {
-	struct dma_resv *resv = attach->dmabuf->resv;
+	struct dma_resv *resv = dma_buf->resv;
 	struct amdgpu_device *adev = dev->dev_private;
 	struct amdgpu_bo *bo;
 	struct amdgpu_bo_param bp;
 	int ret;
 
 	memset(&bp, 0, sizeof(bp));
-	bp.size = attach->dmabuf->size;
+	bp.size = dma_buf->size;
 	bp.byte_align = PAGE_SIZE;
 	bp.domain = AMDGPU_GEM_DOMAIN_CPU;
 	bp.flags = 0;
@@ -385,11 +401,9 @@ amdgpu_gem_prime_import_sg_table(struct drm_device *dev,
 	if (ret)
 		goto error;
 
-	bo->tbo.sg = sg;
-	bo->tbo.ttm->sg = sg;
 	bo->allowed_domains = AMDGPU_GEM_DOMAIN_GTT;
 	bo->preferred_domains = AMDGPU_GEM_DOMAIN_GTT;
-	if (attach->dmabuf->ops != &amdgpu_dmabuf_ops)
+	if (dma_buf->ops != &amdgpu_dmabuf_ops)
 		bo->prime_shared_count = 1;
 
 	dma_resv_unlock(resv);
@@ -405,15 +419,15 @@ error:
  * @dev: DRM device
  * @dma_buf: Shared DMA buffer
  *
- * The main work is done by the &drm_gem_prime_import helper, which in turn
- * uses &amdgpu_gem_prime_import_sg_table.
+ * Import a dma_buf into a the driver and potentially create a new GEM object.
  *
  * Returns:
  * GEM BO representing the shared DMA buffer for the given device.
  */
 struct drm_gem_object *amdgpu_gem_prime_import(struct drm_device *dev,
-					    struct dma_buf *dma_buf)
+					       struct dma_buf *dma_buf)
 {
+	struct dma_buf_attachment *attach;
 	struct drm_gem_object *obj;
 
 	if (dma_buf->ops == &amdgpu_dmabuf_ops) {
@@ -428,5 +442,17 @@ struct drm_gem_object *amdgpu_gem_prime_import(struct drm_device *dev,
 		}
 	}
 
-	return drm_gem_prime_import(dev, dma_buf);
+	obj = amdgpu_dma_buf_create_obj(dev, dma_buf);
+	if (IS_ERR(obj))
+		return obj;
+
+	attach = dma_buf_dynamic_attach(dma_buf, dev->dev, true);
+	if (IS_ERR(attach)) {
+		drm_gem_object_put(obj);
+		return ERR_CAST(attach);
+	}
+
+	get_dma_buf(dma_buf);
+	obj->import_attach = attach;
+	return obj;
 }
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.h
index 5012e6ab58f1..ec447a7b6b28 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.h
@@ -25,11 +25,6 @@
 
 #include <drm/drm_gem.h>
 
-struct sg_table *amdgpu_gem_prime_get_sg_table(struct drm_gem_object *obj);
-struct drm_gem_object *
-amdgpu_gem_prime_import_sg_table(struct drm_device *dev,
-				 struct dma_buf_attachment *attach,
-				 struct sg_table *sg);
 struct dma_buf *amdgpu_gem_prime_export(struct drm_gem_object *gobj,
 					int flags);
 struct drm_gem_object *amdgpu_gem_prime_import(struct drm_device *dev,
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.c
index 5803fcbae22f..9cc270efee7c 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.c
@@ -911,7 +911,8 @@ int amdgpu_dpm_get_sclk(struct amdgpu_device *adev, bool low)
 	if (is_support_sw_smu(adev)) {
 		ret = smu_get_dpm_freq_range(&adev->smu, SMU_GFXCLK,
 					     low ? &clk_freq : NULL,
-					     !low ? &clk_freq : NULL);
+					     !low ? &clk_freq : NULL,
+					     true);
 		if (ret)
 			return 0;
 		return clk_freq * 100;
@@ -928,7 +929,8 @@ int amdgpu_dpm_get_mclk(struct amdgpu_device *adev, bool low)
 	if (is_support_sw_smu(adev)) {
 		ret = smu_get_dpm_freq_range(&adev->smu, SMU_UCLK,
 					     low ? &clk_freq : NULL,
-					     !low ? &clk_freq : NULL);
+					     !low ? &clk_freq : NULL,
+					     true);
 		if (ret)
 			return 0;
 		return clk_freq * 100;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h
index 1c5c0fd76dbf..2cfb677272af 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h
@@ -298,12 +298,6 @@ enum amdgpu_pcie_gen {
 #define amdgpu_dpm_get_current_power_state(adev) \
 		((adev)->powerplay.pp_funcs->get_current_power_state((adev)->powerplay.pp_handle))
 
-#define amdgpu_smu_get_current_power_state(adev) \
-		((adev)->smu.ppt_funcs->get_current_power_state(&((adev)->smu)))
-
-#define amdgpu_smu_set_power_state(adev) \
-		((adev)->smu.ppt_funcs->set_power_state(&((adev)->smu)))
-
 #define amdgpu_dpm_get_pp_num_states(adev, data) \
 		((adev)->powerplay.pp_funcs->get_pp_num_states((adev)->powerplay.pp_handle, data))
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
index 264677ab248a..0ffc9447b573 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
@@ -43,6 +43,8 @@
 
 #include "amdgpu_amdkfd.h"
 
+#include "amdgpu_ras.h"
+
 /*
  * KMS wrapper.
  * - 3.0.0 - initial driver
@@ -81,13 +83,13 @@
  * - 3.32.0 - Add syncobj timeline support to AMDGPU_CS.
  * - 3.33.0 - Fixes for GDS ENOMEM failures in AMDGPU_CS.
  * - 3.34.0 - Non-DC can flip correctly between buffers with different pitches
+ * - 3.35.0 - Add drm_amdgpu_info_device::tcc_disabled_mask
+ * - 3.36.0 - Allow reading more status registers on si/cik
  */
 #define KMS_DRIVER_MAJOR	3
-#define KMS_DRIVER_MINOR	34
+#define KMS_DRIVER_MINOR	36
 #define KMS_DRIVER_PATCHLEVEL	0
 
-#define AMDGPU_MAX_TIMEOUT_PARAM_LENTH	256
-
 int amdgpu_vram_limit = 0;
 int amdgpu_vis_vram_limit = 0;
 int amdgpu_gart_size = -1; /* auto */
@@ -100,7 +102,7 @@ int amdgpu_disp_priority = 0;
 int amdgpu_hw_i2c = 0;
 int amdgpu_pcie_gen2 = -1;
 int amdgpu_msi = -1;
-char amdgpu_lockup_timeout[AMDGPU_MAX_TIMEOUT_PARAM_LENTH];
+char amdgpu_lockup_timeout[AMDGPU_MAX_TIMEOUT_PARAM_LENGTH];
 int amdgpu_dpm = -1;
 int amdgpu_fw_load_type = -1;
 int amdgpu_aspm = -1;
@@ -127,11 +129,7 @@ char *amdgpu_disable_cu = NULL;
 char *amdgpu_virtual_display = NULL;
 /* OverDrive(bit 14) disabled by default*/
 uint amdgpu_pp_feature_mask = 0xffffbfff;
-int amdgpu_ngg = 0;
-int amdgpu_prim_buf_per_se = 0;
-int amdgpu_pos_buf_per_se = 0;
-int amdgpu_cntl_sb_buf_per_se = 0;
-int amdgpu_param_buf_per_se = 0;
+uint amdgpu_force_long_training = 0;
 int amdgpu_job_hang_limit = 0;
 int amdgpu_lbpw = -1;
 int amdgpu_compute_multipipe = -1;
@@ -145,12 +143,13 @@ int amdgpu_mcbp = 0;
 int amdgpu_discovery = -1;
 int amdgpu_mes = 0;
 int amdgpu_noretry = 1;
+int amdgpu_force_asic_type = -1;
 
 struct amdgpu_mgpu_info mgpu_info = {
 	.mutex = __MUTEX_INITIALIZER(mgpu_info.mutex),
 };
 int amdgpu_ras_enable = -1;
-uint amdgpu_ras_mask = 0xfffffffb;
+uint amdgpu_ras_mask = 0xffffffff;
 
 /**
  * DOC: vramlimit (int)
@@ -243,16 +242,21 @@ module_param_named(msi, amdgpu_msi, int, 0444);
  *
  * The format can be [Non-Compute] or [GFX,Compute,SDMA,Video]. That is there can be one or
  * multiple values specified. 0 and negative values are invalidated. They will be adjusted
- * to default timeout.
- *  - With one value specified, the setting will apply to all non-compute jobs.
- *  - With multiple values specified, the first one will be for GFX. The second one is for Compute.
- *    And the third and fourth ones are for SDMA and Video.
+ * to the default timeout.
+ *
+ * - With one value specified, the setting will apply to all non-compute jobs.
+ * - With multiple values specified, the first one will be for GFX.
+ *   The second one is for Compute. The third and fourth ones are
+ *   for SDMA and Video.
+ *
  * By default(with no lockup_timeout settings), the timeout for all non-compute(GFX, SDMA and Video)
  * jobs is 10000. And there is no timeout enforced on compute jobs.
  */
-MODULE_PARM_DESC(lockup_timeout, "GPU lockup timeout in ms (default: 10000 for non-compute jobs and infinity timeout for compute jobs."
+MODULE_PARM_DESC(lockup_timeout, "GPU lockup timeout in ms (default: for bare metal 10000 for non-compute jobs and infinity timeout for compute jobs; "
+		"for passthrough or sriov, 10000 for all jobs."
 		" 0: keep default value. negative: infinity timeout), "
-		"format is [Non-Compute] or [GFX,Compute,SDMA,Video]");
+		"format: for bare metal [Non-Compute] or [GFX,Compute,SDMA,Video]; "
+		"for passthrough or sriov [all jobs] or [GFX,Compute,SDMA,Video].");
 module_param_string(lockup_timeout, amdgpu_lockup_timeout, sizeof(amdgpu_lockup_timeout), 0444);
 
 /**
@@ -391,6 +395,14 @@ MODULE_PARM_DESC(ppfeaturemask, "all power features enabled (default))");
 module_param_named(ppfeaturemask, amdgpu_pp_feature_mask, uint, 0444);
 
 /**
+ * DOC: forcelongtraining (uint)
+ * Force long memory training in resume.
+ * The default is zero, indicates short training in resume.
+ */
+MODULE_PARM_DESC(forcelongtraining, "force memory long training");
+module_param_named(forcelongtraining, amdgpu_force_long_training, uint, 0444);
+
+/**
  * DOC: pcie_gen_cap (uint)
  * Override PCIE gen speed capabilities. See the CAIL flags in drivers/gpu/drm/amd/include/amd_pcie.h.
  * The default is 0 (automatic for each asic).
@@ -448,42 +460,6 @@ MODULE_PARM_DESC(virtual_display,
 module_param_named(virtual_display, amdgpu_virtual_display, charp, 0444);
 
 /**
- * DOC: ngg (int)
- * Set to enable Next Generation Graphics (1 = enable). The default is 0 (disabled).
- */
-MODULE_PARM_DESC(ngg, "Next Generation Graphics (1 = enable, 0 = disable(default depending on gfx))");
-module_param_named(ngg, amdgpu_ngg, int, 0444);
-
-/**
- * DOC: prim_buf_per_se (int)
- * Override the size of Primitive Buffer per Shader Engine in Byte. The default is 0 (depending on gfx).
- */
-MODULE_PARM_DESC(prim_buf_per_se, "the size of Primitive Buffer per Shader Engine (default depending on gfx)");
-module_param_named(prim_buf_per_se, amdgpu_prim_buf_per_se, int, 0444);
-
-/**
- * DOC: pos_buf_per_se (int)
- * Override the size of Position Buffer per Shader Engine in Byte. The default is 0 (depending on gfx).
- */
-MODULE_PARM_DESC(pos_buf_per_se, "the size of Position Buffer per Shader Engine (default depending on gfx)");
-module_param_named(pos_buf_per_se, amdgpu_pos_buf_per_se, int, 0444);
-
-/**
- * DOC: cntl_sb_buf_per_se (int)
- * Override the size of Control Sideband per Shader Engine in Byte. The default is 0 (depending on gfx).
- */
-MODULE_PARM_DESC(cntl_sb_buf_per_se, "the size of Control Sideband per Shader Engine (default depending on gfx)");
-module_param_named(cntl_sb_buf_per_se, amdgpu_cntl_sb_buf_per_se, int, 0444);
-
-/**
- * DOC: param_buf_per_se (int)
- * Override the size of Off-Chip Parameter Cache per Shader Engine in Byte.
- * The default is 0 (depending on gfx).
- */
-MODULE_PARM_DESC(param_buf_per_se, "the size of Off-Chip Parameter Cache per Shader Engine (default depending on gfx)");
-module_param_named(param_buf_per_se, amdgpu_param_buf_per_se, int, 0444);
-
-/**
  * DOC: job_hang_limit (int)
  * Set how much time allow a job hang and not drop it. The default is 0.
  */
@@ -615,6 +591,16 @@ MODULE_PARM_DESC(noretry,
 	"Disable retry faults (0 = retry enabled, 1 = retry disabled (default))");
 module_param_named(noretry, amdgpu_noretry, int, 0644);
 
+/**
+ * DOC: force_asic_type (int)
+ * A non negative value used to specify the asic type for all supported GPUs.
+ */
+MODULE_PARM_DESC(force_asic_type,
+	"A non negative value used to specify the asic type for all supported GPUs");
+module_param_named(force_asic_type, amdgpu_force_asic_type, int, 0444);
+
+
+
 #ifdef CONFIG_HSA_AMD
 /**
  * DOC: sched_policy (int)
@@ -1012,15 +998,17 @@ static const struct pci_device_id pciidlist[] = {
 	{0x1002, 0x731B, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_NAVI10},
 	{0x1002, 0x731F, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_NAVI10},
 	/* Navi14 */
-	{0x1002, 0x7340, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_NAVI14|AMD_EXP_HW_SUPPORT},
-	{0x1002, 0x7341, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_NAVI14|AMD_EXP_HW_SUPPORT},
-	{0x1002, 0x7347, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_NAVI14|AMD_EXP_HW_SUPPORT},
+	{0x1002, 0x7340, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_NAVI14},
+	{0x1002, 0x7341, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_NAVI14},
+	{0x1002, 0x7347, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_NAVI14},
+	{0x1002, 0x734F, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_NAVI14},
 
 	/* Renoir */
 	{0x1002, 0x1636, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RENOIR|AMD_IS_APU|AMD_EXP_HW_SUPPORT},
 
 	/* Navi12 */
 	{0x1002, 0x7360, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_NAVI12|AMD_EXP_HW_SUPPORT},
+	{0x1002, 0x7362, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_NAVI12|AMD_EXP_HW_SUPPORT},
 
 	{0, 0, 0}
 };
@@ -1047,8 +1035,43 @@ static int amdgpu_pci_probe(struct pci_dev *pdev,
 		return -ENODEV;
 	}
 
+#ifdef CONFIG_DRM_AMDGPU_SI
+	if (!amdgpu_si_support) {
+		switch (flags & AMD_ASIC_MASK) {
+		case CHIP_TAHITI:
+		case CHIP_PITCAIRN:
+		case CHIP_VERDE:
+		case CHIP_OLAND:
+		case CHIP_HAINAN:
+			dev_info(&pdev->dev,
+				 "SI support provided by radeon.\n");
+			dev_info(&pdev->dev,
+				 "Use radeon.si_support=0 amdgpu.si_support=1 to override.\n"
+				);
+			return -ENODEV;
+		}
+	}
+#endif
+#ifdef CONFIG_DRM_AMDGPU_CIK
+	if (!amdgpu_cik_support) {
+		switch (flags & AMD_ASIC_MASK) {
+		case CHIP_KAVERI:
+		case CHIP_BONAIRE:
+		case CHIP_HAWAII:
+		case CHIP_KABINI:
+		case CHIP_MULLINS:
+			dev_info(&pdev->dev,
+				 "CIK support provided by radeon.\n");
+			dev_info(&pdev->dev,
+				 "Use radeon.cik_support=0 amdgpu.cik_support=1 to override.\n"
+				);
+			return -ENODEV;
+		}
+	}
+#endif
+
 	/* Get rid of things like offb */
-	ret = drm_fb_helper_remove_conflicting_pci_framebuffers(pdev, 0, "amdgpudrmfb");
+	ret = drm_fb_helper_remove_conflicting_pci_framebuffers(pdev, "amdgpudrmfb");
 	if (ret)
 		return ret;
 
@@ -1091,7 +1114,10 @@ amdgpu_pci_remove(struct pci_dev *pdev)
 {
 	struct drm_device *dev = pci_get_drvdata(pdev);
 
-	DRM_ERROR("Device removal is currently not supported outside of fbcon\n");
+#ifdef MODULE
+	if (THIS_MODULE->state != MODULE_STATE_GOING)
+#endif
+		DRM_ERROR("Hotplug removal is not supported\n");
 	drm_dev_unplug(dev);
 	drm_dev_put(dev);
 	pci_disable_device(pdev);
@@ -1104,6 +1130,9 @@ amdgpu_pci_shutdown(struct pci_dev *pdev)
 	struct drm_device *dev = pci_get_drvdata(pdev);
 	struct amdgpu_device *adev = dev->dev_private;
 
+	if (amdgpu_ras_intr_triggered())
+		return;
+
 	/* if we are running in a VM, make sure the device
 	 * torn down properly on reboot/shutdown.
 	 * unfortunately we can't detect certain
@@ -1138,8 +1167,13 @@ static int amdgpu_pmops_resume(struct device *dev)
 static int amdgpu_pmops_freeze(struct device *dev)
 {
 	struct drm_device *drm_dev = dev_get_drvdata(dev);
+	struct amdgpu_device *adev = drm_dev->dev_private;
+	int r;
 
-	return amdgpu_device_suspend(drm_dev, false, true);
+	r = amdgpu_device_suspend(drm_dev, false, true);
+	if (r)
+		return r;
+	return amdgpu_asic_reset(adev);
 }
 
 static int amdgpu_pmops_thaw(struct device *dev)
@@ -1311,66 +1345,6 @@ int amdgpu_file_to_fpriv(struct file *filp, struct amdgpu_fpriv **fpriv)
 	return 0;
 }
 
-int amdgpu_device_get_job_timeout_settings(struct amdgpu_device *adev)
-{
-	char *input = amdgpu_lockup_timeout;
-	char *timeout_setting = NULL;
-	int index = 0;
-	long timeout;
-	int ret = 0;
-
-	/*
-	 * By default timeout for non compute jobs is 10000.
-	 * And there is no timeout enforced on compute jobs.
-	 */
-	adev->gfx_timeout = msecs_to_jiffies(10000);
-	adev->sdma_timeout = adev->video_timeout = adev->gfx_timeout;
-	adev->compute_timeout = MAX_SCHEDULE_TIMEOUT;
-
-	if (strnlen(input, AMDGPU_MAX_TIMEOUT_PARAM_LENTH)) {
-		while ((timeout_setting = strsep(&input, ",")) &&
-				strnlen(timeout_setting, AMDGPU_MAX_TIMEOUT_PARAM_LENTH)) {
-			ret = kstrtol(timeout_setting, 0, &timeout);
-			if (ret)
-				return ret;
-
-			if (timeout == 0) {
-				index++;
-				continue;
-			} else if (timeout < 0) {
-				timeout = MAX_SCHEDULE_TIMEOUT;
-			} else {
-				timeout = msecs_to_jiffies(timeout);
-			}
-
-			switch (index++) {
-			case 0:
-				adev->gfx_timeout = timeout;
-				break;
-			case 1:
-				adev->compute_timeout = timeout;
-				break;
-			case 2:
-				adev->sdma_timeout = timeout;
-				break;
-			case 3:
-				adev->video_timeout = timeout;
-				break;
-			default:
-				break;
-			}
-		}
-		/*
-		 * There is only one value specified and
-		 * it should apply to all non-compute jobs.
-		 */
-		if (index == 1)
-			adev->sdma_timeout = adev->video_timeout = adev->gfx_timeout;
-	}
-
-	return ret;
-}
-
 static bool
 amdgpu_get_crtc_scanout_position(struct drm_device *dev, unsigned int pipe,
 				 bool in_vblank_irq, int *vpos, int *hpos,
@@ -1409,8 +1383,6 @@ static struct drm_driver kms_driver = {
 	.prime_fd_to_handle = drm_gem_prime_fd_to_handle,
 	.gem_prime_export = amdgpu_gem_prime_export,
 	.gem_prime_import = amdgpu_gem_prime_import,
-	.gem_prime_get_sg_table = amdgpu_gem_prime_get_sg_table,
-	.gem_prime_import_sg_table = amdgpu_gem_prime_import_sg_table,
 	.gem_prime_vmap = amdgpu_gem_prime_vmap,
 	.gem_prime_vunmap = amdgpu_gem_prime_vunmap,
 	.gem_prime_mmap = amdgpu_gem_prime_mmap,
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_encoders.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_encoders.c
index 571a6dfb473e..61fcf247a638 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_encoders.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_encoders.c
@@ -37,12 +37,14 @@ amdgpu_link_encoder_connector(struct drm_device *dev)
 {
 	struct amdgpu_device *adev = dev->dev_private;
 	struct drm_connector *connector;
+	struct drm_connector_list_iter iter;
 	struct amdgpu_connector *amdgpu_connector;
 	struct drm_encoder *encoder;
 	struct amdgpu_encoder *amdgpu_encoder;
 
+	drm_connector_list_iter_begin(dev, &iter);
 	/* walk the list and link encoders to connectors */
-	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
+	drm_for_each_connector_iter(connector, &iter) {
 		amdgpu_connector = to_amdgpu_connector(connector);
 		list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) {
 			amdgpu_encoder = to_amdgpu_encoder(encoder);
@@ -55,6 +57,7 @@ amdgpu_link_encoder_connector(struct drm_device *dev)
 			}
 		}
 	}
+	drm_connector_list_iter_end(&iter);
 }
 
 void amdgpu_encoder_set_active_device(struct drm_encoder *encoder)
@@ -62,8 +65,10 @@ void amdgpu_encoder_set_active_device(struct drm_encoder *encoder)
 	struct drm_device *dev = encoder->dev;
 	struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder);
 	struct drm_connector *connector;
+	struct drm_connector_list_iter iter;
 
-	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
+	drm_connector_list_iter_begin(dev, &iter);
+	drm_for_each_connector_iter(connector, &iter) {
 		if (connector->encoder == encoder) {
 			struct amdgpu_connector *amdgpu_connector = to_amdgpu_connector(connector);
 			amdgpu_encoder->active_device = amdgpu_encoder->devices & amdgpu_connector->devices;
@@ -72,6 +77,7 @@ void amdgpu_encoder_set_active_device(struct drm_encoder *encoder)
 				  amdgpu_connector->devices, encoder->encoder_type);
 		}
 	}
+	drm_connector_list_iter_end(&iter);
 }
 
 struct drm_connector *
@@ -79,15 +85,20 @@ amdgpu_get_connector_for_encoder(struct drm_encoder *encoder)
 {
 	struct drm_device *dev = encoder->dev;
 	struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder);
-	struct drm_connector *connector;
+	struct drm_connector *connector, *found = NULL;
+	struct drm_connector_list_iter iter;
 	struct amdgpu_connector *amdgpu_connector;
 
-	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
+	drm_connector_list_iter_begin(dev, &iter);
+	drm_for_each_connector_iter(connector, &iter) {
 		amdgpu_connector = to_amdgpu_connector(connector);
-		if (amdgpu_encoder->active_device & amdgpu_connector->devices)
-			return connector;
+		if (amdgpu_encoder->active_device & amdgpu_connector->devices) {
+			found = connector;
+			break;
+		}
 	}
-	return NULL;
+	drm_connector_list_iter_end(&iter);
+	return found;
 }
 
 struct drm_connector *
@@ -95,15 +106,20 @@ amdgpu_get_connector_for_encoder_init(struct drm_encoder *encoder)
 {
 	struct drm_device *dev = encoder->dev;
 	struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder);
-	struct drm_connector *connector;
+	struct drm_connector *connector, *found = NULL;
+	struct drm_connector_list_iter iter;
 	struct amdgpu_connector *amdgpu_connector;
 
-	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
+	drm_connector_list_iter_begin(dev, &iter);
+	drm_for_each_connector_iter(connector, &iter) {
 		amdgpu_connector = to_amdgpu_connector(connector);
-		if (amdgpu_encoder->devices & amdgpu_connector->devices)
-			return connector;
+		if (amdgpu_encoder->devices & amdgpu_connector->devices) {
+			found = connector;
+			break;
+		}
 	}
-	return NULL;
+	drm_connector_list_iter_end(&iter);
+	return found;
 }
 
 struct drm_encoder *amdgpu_get_external_encoder(struct drm_encoder *encoder)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c
index 23085b352cf2..377fe20bce23 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c
@@ -462,18 +462,7 @@ int amdgpu_fence_driver_init_ring(struct amdgpu_ring *ring,
 			timeout = adev->gfx_timeout;
 			break;
 		case AMDGPU_RING_TYPE_COMPUTE:
-			/*
-			 * For non-sriov case, no timeout enforce
-			 * on compute ring by default. Unless user
-			 * specifies a timeout for compute ring.
-			 *
-			 * For sriov case, always use the timeout
-			 * as gfx ring
-			 */
-			if (!amdgpu_sriov_vf(ring->adev))
-				timeout = adev->compute_timeout;
-			else
-				timeout = adev->gfx_timeout;
+			timeout = adev->compute_timeout;
 			break;
 		case AMDGPU_RING_TYPE_SDMA:
 			timeout = adev->sdma_timeout;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c
index 5e8bdded265f..19705e399905 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c
@@ -71,7 +71,7 @@
  */
 static int amdgpu_gart_dummy_page_init(struct amdgpu_device *adev)
 {
-	struct page *dummy_page = adev->mman.bdev.glob->dummy_read_page;
+	struct page *dummy_page = ttm_bo_glob.dummy_read_page;
 
 	if (adev->dummy_page_addr)
 		return 0;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
index 8ceb44925947..4277125a79ee 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
@@ -175,7 +175,7 @@ void amdgpu_gem_object_close(struct drm_gem_object *obj,
 
 	amdgpu_vm_get_pd_bo(vm, &list, &vm_pd);
 
-	r = ttm_eu_reserve_buffers(&ticket, &list, false, &duplicates, false);
+	r = ttm_eu_reserve_buffers(&ticket, &list, false, &duplicates);
 	if (r) {
 		dev_err(adev->dev, "leaking bo va because "
 			"we fail to reserve bo (%d)\n", r);
@@ -527,13 +527,41 @@ static void amdgpu_gem_va_update_vm(struct amdgpu_device *adev,
 			goto error;
 	}
 
-	r = amdgpu_vm_update_directories(adev, vm);
+	r = amdgpu_vm_update_pdes(adev, vm, false);
 
 error:
 	if (r && r != -ERESTARTSYS)
 		DRM_ERROR("Couldn't update BO_VA (%d)\n", r);
 }
 
+/**
+ * amdgpu_gem_va_map_flags - map GEM UAPI flags into hardware flags
+ *
+ * @adev: amdgpu_device pointer
+ * @flags: GEM UAPI flags
+ *
+ * Returns the GEM UAPI flags mapped into hardware for the ASIC.
+ */
+uint64_t amdgpu_gem_va_map_flags(struct amdgpu_device *adev, uint32_t flags)
+{
+	uint64_t pte_flag = 0;
+
+	if (flags & AMDGPU_VM_PAGE_EXECUTABLE)
+		pte_flag |= AMDGPU_PTE_EXECUTABLE;
+	if (flags & AMDGPU_VM_PAGE_READABLE)
+		pte_flag |= AMDGPU_PTE_READABLE;
+	if (flags & AMDGPU_VM_PAGE_WRITEABLE)
+		pte_flag |= AMDGPU_PTE_WRITEABLE;
+	if (flags & AMDGPU_VM_PAGE_PRT)
+		pte_flag |= AMDGPU_PTE_PRT;
+
+	if (adev->gmc.gmc_funcs->map_mtype)
+		pte_flag |= amdgpu_gmc_map_mtype(adev,
+						 flags & AMDGPU_VM_MTYPE_MASK);
+
+	return pte_flag;
+}
+
 int amdgpu_gem_va_ioctl(struct drm_device *dev, void *data,
 			  struct drm_file *filp)
 {
@@ -613,7 +641,7 @@ int amdgpu_gem_va_ioctl(struct drm_device *dev, void *data,
 
 	amdgpu_vm_get_pd_bo(&fpriv->vm, &list, &vm_pd);
 
-	r = ttm_eu_reserve_buffers(&ticket, &list, true, &duplicates, false);
+	r = ttm_eu_reserve_buffers(&ticket, &list, true, &duplicates);
 	if (r)
 		goto error_unref;
 
@@ -631,7 +659,7 @@ int amdgpu_gem_va_ioctl(struct drm_device *dev, void *data,
 
 	switch (args->operation) {
 	case AMDGPU_VA_OP_MAP:
-		va_flags = amdgpu_gmc_get_pte_flags(adev, args->flags);
+		va_flags = amdgpu_gem_va_map_flags(adev, args->flags);
 		r = amdgpu_vm_bo_map(adev, bo_va, args->va_address,
 				     args->offset_in_bo, args->map_size,
 				     va_flags);
@@ -646,7 +674,7 @@ int amdgpu_gem_va_ioctl(struct drm_device *dev, void *data,
 						args->map_size);
 		break;
 	case AMDGPU_VA_OP_REPLACE:
-		va_flags = amdgpu_gmc_get_pte_flags(adev, args->flags);
+		va_flags = amdgpu_gem_va_map_flags(adev, args->flags);
 		r = amdgpu_vm_bo_replace_map(adev, bo_va, args->va_address,
 					     args->offset_in_bo, args->map_size,
 					     va_flags);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.h
index 0b66d2e6b5d5..e0f025dd1b14 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.h
@@ -67,6 +67,7 @@ int amdgpu_gem_mmap_ioctl(struct drm_device *dev, void *data,
 			  struct drm_file *filp);
 int amdgpu_gem_wait_idle_ioctl(struct drm_device *dev, void *data,
 			      struct drm_file *filp);
+uint64_t amdgpu_gem_va_map_flags(struct amdgpu_device *adev, uint32_t flags);
 int amdgpu_gem_va_ioctl(struct drm_device *dev, void *data,
 			  struct drm_file *filp);
 int amdgpu_gem_op_ioctl(struct drm_device *dev, void *data,
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c
index f9bef3154b99..e00b46180d2e 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c
@@ -26,6 +26,7 @@
 #include "amdgpu.h"
 #include "amdgpu_gfx.h"
 #include "amdgpu_rlc.h"
+#include "amdgpu_ras.h"
 
 /* delay 0.1 second to enable gfx off feature */
 #define GFX_OFF_DELAY_ENABLE         msecs_to_jiffies(100)
@@ -231,12 +232,10 @@ void amdgpu_gfx_compute_queue_acquire(struct amdgpu_device *adev)
 
 void amdgpu_gfx_graphics_queue_acquire(struct amdgpu_device *adev)
 {
-	int i, queue, pipe, me;
+	int i, queue, me;
 
 	for (i = 0; i < AMDGPU_MAX_GFX_QUEUES; ++i) {
 		queue = i % adev->gfx.me.num_queue_per_pipe;
-		pipe = (i / adev->gfx.me.num_queue_per_pipe)
-			% adev->gfx.me.num_pipe_per_me;
 		me = (i / adev->gfx.me.num_queue_per_pipe)
 		      / adev->gfx.me.num_pipe_per_me;
 
@@ -320,8 +319,7 @@ int amdgpu_gfx_kiq_init_ring(struct amdgpu_device *adev,
 	return r;
 }
 
-void amdgpu_gfx_kiq_free_ring(struct amdgpu_ring *ring,
-			      struct amdgpu_irq_src *irq)
+void amdgpu_gfx_kiq_free_ring(struct amdgpu_ring *ring)
 {
 	amdgpu_device_wb_free(ring->adev, ring->adev->virt.reg_val_offs);
 	amdgpu_ring_fini(ring);
@@ -456,8 +454,6 @@ void amdgpu_gfx_mqd_sw_fini(struct amdgpu_device *adev)
 	}
 
 	ring = &adev->gfx.kiq.ring;
-	if (adev->asic_type >= CHIP_NAVI10 && amdgpu_async_gfx_ring)
-		kfree(adev->gfx.me.mqd_backup[AMDGPU_MAX_GFX_RINGS]);
 	kfree(adev->gfx.mec.mqd_backup[AMDGPU_MAX_COMPUTE_RINGS]);
 	amdgpu_bo_free_kernel(&ring->mqd_obj,
 			      &ring->mqd_gpu_addr,
@@ -569,3 +565,102 @@ void amdgpu_gfx_off_ctrl(struct amdgpu_device *adev, bool enable)
 
 	mutex_unlock(&adev->gfx.gfx_off_mutex);
 }
+
+int amdgpu_gfx_ras_late_init(struct amdgpu_device *adev)
+{
+	int r;
+	struct ras_fs_if fs_info = {
+		.sysfs_name = "gfx_err_count",
+		.debugfs_name = "gfx_err_inject",
+	};
+	struct ras_ih_if ih_info = {
+		.cb = amdgpu_gfx_process_ras_data_cb,
+	};
+
+	if (!adev->gfx.ras_if) {
+		adev->gfx.ras_if = kmalloc(sizeof(struct ras_common_if), GFP_KERNEL);
+		if (!adev->gfx.ras_if)
+			return -ENOMEM;
+		adev->gfx.ras_if->block = AMDGPU_RAS_BLOCK__GFX;
+		adev->gfx.ras_if->type = AMDGPU_RAS_ERROR__MULTI_UNCORRECTABLE;
+		adev->gfx.ras_if->sub_block_index = 0;
+		strcpy(adev->gfx.ras_if->name, "gfx");
+	}
+	fs_info.head = ih_info.head = *adev->gfx.ras_if;
+
+	r = amdgpu_ras_late_init(adev, adev->gfx.ras_if,
+				 &fs_info, &ih_info);
+	if (r)
+		goto free;
+
+	if (amdgpu_ras_is_supported(adev, adev->gfx.ras_if->block)) {
+		r = amdgpu_irq_get(adev, &adev->gfx.cp_ecc_error_irq, 0);
+		if (r)
+			goto late_fini;
+	} else {
+		/* free gfx ras_if if ras is not supported */
+		r = 0;
+		goto free;
+	}
+
+	return 0;
+late_fini:
+	amdgpu_ras_late_fini(adev, adev->gfx.ras_if, &ih_info);
+free:
+	kfree(adev->gfx.ras_if);
+	adev->gfx.ras_if = NULL;
+	return r;
+}
+
+void amdgpu_gfx_ras_fini(struct amdgpu_device *adev)
+{
+	if (amdgpu_ras_is_supported(adev, AMDGPU_RAS_BLOCK__GFX) &&
+			adev->gfx.ras_if) {
+		struct ras_common_if *ras_if = adev->gfx.ras_if;
+		struct ras_ih_if ih_info = {
+			.head = *ras_if,
+			.cb = amdgpu_gfx_process_ras_data_cb,
+		};
+
+		amdgpu_ras_late_fini(adev, ras_if, &ih_info);
+		kfree(ras_if);
+	}
+}
+
+int amdgpu_gfx_process_ras_data_cb(struct amdgpu_device *adev,
+		void *err_data,
+		struct amdgpu_iv_entry *entry)
+{
+	/* TODO ue will trigger an interrupt.
+	 *
+	 * When “Full RAS” is enabled, the per-IP interrupt sources should
+	 * be disabled and the driver should only look for the aggregated
+	 * interrupt via sync flood
+	 */
+	if (!amdgpu_ras_is_supported(adev, AMDGPU_RAS_BLOCK__GFX)) {
+		kgd2kfd_set_sram_ecc_flag(adev->kfd.dev);
+		if (adev->gfx.funcs->query_ras_error_count)
+			adev->gfx.funcs->query_ras_error_count(adev, err_data);
+		amdgpu_ras_reset_gpu(adev, 0);
+	}
+	return AMDGPU_RAS_SUCCESS;
+}
+
+int amdgpu_gfx_cp_ecc_error_irq(struct amdgpu_device *adev,
+				  struct amdgpu_irq_src *source,
+				  struct amdgpu_iv_entry *entry)
+{
+	struct ras_common_if *ras_if = adev->gfx.ras_if;
+	struct ras_dispatch_if ih_data = {
+		.entry = entry,
+	};
+
+	if (!ras_if)
+		return 0;
+
+	ih_data.head = *ras_if;
+
+	DRM_ERROR("CP ECC ERROR IRQ\n");
+	amdgpu_ras_interrupt_dispatch(adev, &ih_data);
+	return 0;
+}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h
index 554a59b3c4a6..0ae0a2715b0d 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h
@@ -165,6 +165,7 @@ struct amdgpu_gfx_config {
 	uint32_t num_sc_per_sh;
 	uint32_t num_packer_per_sc;
 	uint32_t pa_sc_tile_steering_override;
+	uint64_t tcc_disabled_mask;
 };
 
 struct amdgpu_cu_info {
@@ -200,28 +201,6 @@ struct amdgpu_gfx_funcs {
 	int (*query_ras_error_count) (struct amdgpu_device *adev, void *ras_error_status);
 };
 
-struct amdgpu_ngg_buf {
-	struct amdgpu_bo	*bo;
-	uint64_t		gpu_addr;
-	uint32_t		size;
-	uint32_t		bo_size;
-};
-
-enum {
-	NGG_PRIM = 0,
-	NGG_POS,
-	NGG_CNTL,
-	NGG_PARAM,
-	NGG_BUF_MAX
-};
-
-struct amdgpu_ngg {
-	struct amdgpu_ngg_buf	buf[NGG_BUF_MAX];
-	uint32_t		gds_reserve_addr;
-	uint32_t		gds_reserve_size;
-	bool			init;
-};
-
 struct sq_work {
 	struct work_struct	work;
 	unsigned ih_data;
@@ -246,7 +225,7 @@ struct amdgpu_me {
 	uint32_t			num_me;
 	uint32_t			num_pipe_per_me;
 	uint32_t			num_queue_per_pipe;
-	void				*mqd_backup[AMDGPU_MAX_GFX_RINGS + 1];
+	void				*mqd_backup[AMDGPU_MAX_GFX_RINGS];
 
 	/* These are the resources for which amdgpu takes ownership */
 	DECLARE_BITMAP(queue_bitmap, AMDGPU_MAX_GFX_QUEUES);
@@ -288,6 +267,7 @@ struct amdgpu_gfx {
 	uint32_t			mec2_feature_version;
 	bool				mec_fw_write_wait;
 	bool				me_fw_write_wait;
+	bool				cp_fw_write_wait;
 	struct amdgpu_ring		gfx_ring[AMDGPU_MAX_GFX_RINGS];
 	unsigned			num_gfx_rings;
 	struct amdgpu_ring		compute_ring[AMDGPU_MAX_COMPUTE_RINGS];
@@ -310,9 +290,6 @@ struct amdgpu_gfx {
 	uint32_t                        grbm_soft_reset;
 	uint32_t                        srbm_soft_reset;
 
-	/* NGG */
-	struct amdgpu_ngg		ngg;
-
 	/* gfx off */
 	bool                            gfx_off_state; /* true: enabled, false: disabled */
 	struct mutex                    gfx_off_mutex;
@@ -354,8 +331,7 @@ int amdgpu_gfx_kiq_init_ring(struct amdgpu_device *adev,
 			     struct amdgpu_ring *ring,
 			     struct amdgpu_irq_src *irq);
 
-void amdgpu_gfx_kiq_free_ring(struct amdgpu_ring *ring,
-			      struct amdgpu_irq_src *irq);
+void amdgpu_gfx_kiq_free_ring(struct amdgpu_ring *ring);
 
 void amdgpu_gfx_kiq_fini(struct amdgpu_device *adev);
 int amdgpu_gfx_kiq_init(struct amdgpu_device *adev,
@@ -383,5 +359,12 @@ void amdgpu_gfx_bit_to_me_queue(struct amdgpu_device *adev, int bit,
 bool amdgpu_gfx_is_me_queue_enabled(struct amdgpu_device *adev, int me,
 				    int pipe, int queue);
 void amdgpu_gfx_off_ctrl(struct amdgpu_device *adev, bool enable);
-
+int amdgpu_gfx_ras_late_init(struct amdgpu_device *adev);
+void amdgpu_gfx_ras_fini(struct amdgpu_device *adev);
+int amdgpu_gfx_process_ras_data_cb(struct amdgpu_device *adev,
+		void *err_data,
+		struct amdgpu_iv_entry *entry);
+int amdgpu_gfx_cp_ecc_error_irq(struct amdgpu_device *adev,
+				  struct amdgpu_irq_src *source,
+				  struct amdgpu_iv_entry *entry);
 #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c
index 5790db61fa2c..a12f33c0f5df 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c
@@ -27,6 +27,8 @@
 #include <linux/io-64-nonatomic-lo-hi.h>
 
 #include "amdgpu.h"
+#include "amdgpu_ras.h"
+#include "amdgpu_xgmi.h"
 
 /**
  * amdgpu_gmc_get_pde_for_bo - get the PDE for a BO
@@ -305,3 +307,29 @@ bool amdgpu_gmc_filter_faults(struct amdgpu_device *adev, uint64_t addr,
 	gmc->fault_hash[hash].idx = gmc->last_fault++;
 	return false;
 }
+
+int amdgpu_gmc_ras_late_init(struct amdgpu_device *adev)
+{
+	int r;
+
+	if (adev->umc.funcs && adev->umc.funcs->ras_late_init) {
+		r = adev->umc.funcs->ras_late_init(adev);
+		if (r)
+			return r;
+	}
+
+	if (adev->mmhub.funcs && adev->mmhub.funcs->ras_late_init) {
+		r = adev->mmhub.funcs->ras_late_init(adev);
+		if (r)
+			return r;
+	}
+
+	return amdgpu_xgmi_ras_late_init(adev);
+}
+
+void amdgpu_gmc_ras_fini(struct amdgpu_device *adev)
+{
+	amdgpu_umc_ras_fini(adev);
+	amdgpu_mmhub_ras_fini(adev);
+	amdgpu_xgmi_ras_fini(adev);
+}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h
index b6e1d98ef01e..b499a3de8bb6 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h
@@ -77,6 +77,7 @@ struct amdgpu_gmc_fault {
 struct amdgpu_vmhub {
 	uint32_t	ctx0_ptb_addr_lo32;
 	uint32_t	ctx0_ptb_addr_hi32;
+	uint32_t	vm_inv_eng0_sem;
 	uint32_t	vm_inv_eng0_req;
 	uint32_t	vm_inv_eng0_ack;
 	uint32_t	vm_context0_cntl;
@@ -99,12 +100,15 @@ struct amdgpu_gmc_funcs {
 				   unsigned pasid);
 	/* enable/disable PRT support */
 	void (*set_prt)(struct amdgpu_device *adev, bool enable);
-	/* set pte flags based per asic */
-	uint64_t (*get_vm_pte_flags)(struct amdgpu_device *adev,
-				     uint32_t flags);
+	/* map mtype to hardware flags */
+	uint64_t (*map_mtype)(struct amdgpu_device *adev, uint32_t flags);
 	/* get the pde for a given mc addr */
 	void (*get_vm_pde)(struct amdgpu_device *adev, int level,
 			   u64 *dst, u64 *flags);
+	/* get the pte flags to use for a BO VA mapping */
+	void (*get_vm_pte)(struct amdgpu_device *adev,
+			   struct amdgpu_bo_va_mapping *mapping,
+			   uint64_t *flags);
 };
 
 struct amdgpu_xgmi {
@@ -120,21 +124,52 @@ struct amdgpu_xgmi {
 	/* gpu list in the same hive */
 	struct list_head head;
 	bool supported;
+	struct ras_common_if *ras_if;
 };
 
 struct amdgpu_gmc {
+	/* FB's physical address in MMIO space (for CPU to
+	 * map FB). This is different compared to the agp/
+	 * gart/vram_start/end field as the later is from
+	 * GPU's view and aper_base is from CPU's view.
+	 */
 	resource_size_t		aper_size;
 	resource_size_t		aper_base;
 	/* for some chips with <= 32MB we need to lie
 	 * about vram size near mc fb location */
 	u64			mc_vram_size;
 	u64			visible_vram_size;
+	/* AGP aperture start and end in MC address space
+	 * Driver find a hole in the MC address space
+	 * to place AGP by setting MC_VM_AGP_BOT/TOP registers
+	 * Under VMID0, logical address == MC address. AGP
+	 * aperture maps to physical bus or IOVA addressed.
+	 * AGP aperture is used to simulate FB in ZFB case.
+	 * AGP aperture is also used for page table in system
+	 * memory (mainly for APU).
+	 *
+	 */
 	u64			agp_size;
 	u64			agp_start;
 	u64			agp_end;
+	/* GART aperture start and end in MC address space
+	 * Driver find a hole in the MC address space
+	 * to place GART by setting VM_CONTEXT0_PAGE_TABLE_START/END_ADDR
+	 * registers
+	 * Under VMID0, logical address inside GART aperture will
+	 * be translated through gpuvm gart page table to access
+	 * paged system memory
+	 */
 	u64			gart_size;
 	u64			gart_start;
 	u64			gart_end;
+	/* Frame buffer aperture of this GPU device. Different from
+	 * fb_start (see below), this only covers the local GPU device.
+	 * Driver get fb_start from MC_VM_FB_LOCATION_BASE (set by vbios)
+	 * and calculate vram_start of this local device by adding an
+	 * offset inside the XGMI hive.
+	 * Under VMID0, logical address == MC address
+	 */
 	u64			vram_start;
 	u64			vram_end;
 	/* FB region , it's same as local vram region in single GPU, in XGMI
@@ -153,6 +188,7 @@ struct amdgpu_gmc {
 	uint32_t                fw_version;
 	struct amdgpu_irq_src	vm_fault;
 	uint32_t		vram_type;
+	uint8_t			vram_vendor;
 	uint32_t                srbm_soft_reset;
 	bool			prt_warning;
 	uint64_t		stolen_size;
@@ -177,15 +213,14 @@ struct amdgpu_gmc {
 
 	struct amdgpu_xgmi xgmi;
 	struct amdgpu_irq_src	ecc_irq;
-	struct ras_common_if    *umc_ras_if;
-	struct ras_common_if    *mmhub_ras_if;
 };
 
 #define amdgpu_gmc_flush_gpu_tlb(adev, vmid, vmhub, type) ((adev)->gmc.gmc_funcs->flush_gpu_tlb((adev), (vmid), (vmhub), (type)))
 #define amdgpu_gmc_emit_flush_gpu_tlb(r, vmid, addr) (r)->adev->gmc.gmc_funcs->emit_flush_gpu_tlb((r), (vmid), (addr))
 #define amdgpu_gmc_emit_pasid_mapping(r, vmid, pasid) (r)->adev->gmc.gmc_funcs->emit_pasid_mapping((r), (vmid), (pasid))
+#define amdgpu_gmc_map_mtype(adev, flags) (adev)->gmc.gmc_funcs->map_mtype((adev),(flags))
 #define amdgpu_gmc_get_vm_pde(adev, level, dst, flags) (adev)->gmc.gmc_funcs->get_vm_pde((adev), (level), (dst), (flags))
-#define amdgpu_gmc_get_pte_flags(adev, flags) (adev)->gmc.gmc_funcs->get_vm_pte_flags((adev),(flags))
+#define amdgpu_gmc_get_vm_pte(adev, mapping, flags) (adev)->gmc.gmc_funcs->get_vm_pte((adev), (mapping), (flags))
 
 /**
  * amdgpu_gmc_vram_full_visible - Check if full VRAM is visible through the BAR
@@ -230,5 +265,7 @@ void amdgpu_gmc_agp_location(struct amdgpu_device *adev,
 			     struct amdgpu_gmc *mc);
 bool amdgpu_gmc_filter_faults(struct amdgpu_device *adev, uint64_t addr,
 			      uint16_t pasid, uint64_t timestamp);
+int amdgpu_gmc_ras_late_init(struct amdgpu_device *adev);
+void amdgpu_gmc_ras_fini(struct amdgpu_device *adev);
 
 #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.c
index 53734da1c2df..6f9289735e31 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.c
@@ -282,7 +282,7 @@ static int amdgpu_vmid_grab_reserved(struct amdgpu_vm *vm,
 	    !dma_fence_is_later(updates, (*id)->flushed_updates))
 	    updates = NULL;
 
-	if ((*id)->owner != vm->entity.fence_context ||
+	if ((*id)->owner != vm->direct.fence_context ||
 	    job->vm_pd_addr != (*id)->pd_gpu_addr ||
 	    updates || !(*id)->last_flush ||
 	    ((*id)->last_flush->context != fence_context &&
@@ -349,7 +349,7 @@ static int amdgpu_vmid_grab_used(struct amdgpu_vm *vm,
 		struct dma_fence *flushed;
 
 		/* Check all the prerequisites to using this VMID */
-		if ((*id)->owner != vm->entity.fence_context)
+		if ((*id)->owner != vm->direct.fence_context)
 			continue;
 
 		if ((*id)->pd_gpu_addr != job->vm_pd_addr)
@@ -449,7 +449,7 @@ int amdgpu_vmid_grab(struct amdgpu_vm *vm, struct amdgpu_ring *ring,
 	}
 
 	id->pd_gpu_addr = job->vm_pd_addr;
-	id->owner = vm->entity.fence_context;
+	id->owner = vm->direct.fence_context;
 
 	if (job->vm_needs_flush) {
 		dma_fence_put(id->last_flush);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c
index 2a3f5ec298db..30d540d23b77 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c
@@ -87,10 +87,13 @@ static void amdgpu_hotplug_work_func(struct work_struct *work)
 	struct drm_device *dev = adev->ddev;
 	struct drm_mode_config *mode_config = &dev->mode_config;
 	struct drm_connector *connector;
+	struct drm_connector_list_iter iter;
 
 	mutex_lock(&mode_config->mutex);
-	list_for_each_entry(connector, &mode_config->connector_list, head)
+	drm_connector_list_iter_begin(dev, &iter);
+	drm_for_each_connector_iter(connector, &iter)
 		amdgpu_connector_hotplug(connector);
+	drm_connector_list_iter_end(&iter);
 	mutex_unlock(&mode_config->mutex);
 	/* Just fire off a uevent and let userspace tell us what to do */
 	drm_helper_hpd_irq_event(dev);
@@ -153,6 +156,20 @@ irqreturn_t amdgpu_irq_handler(int irq, void *arg)
 	ret = amdgpu_ih_process(adev, &adev->irq.ih);
 	if (ret == IRQ_HANDLED)
 		pm_runtime_mark_last_busy(dev->dev);
+
+	/* For the hardware that cannot enable bif ring for both ras_controller_irq
+         * and ras_err_evnet_athub_irq ih cookies, the driver has to poll status
+	 * register to check whether the interrupt is triggered or not, and properly
+	 * ack the interrupt if it is there
+	 */
+	if (adev->nbio.funcs &&
+	    adev->nbio.funcs->handle_ras_controller_intr_no_bifring)
+		adev->nbio.funcs->handle_ras_controller_intr_no_bifring(adev);
+
+	if (adev->nbio.funcs &&
+	    adev->nbio.funcs->handle_ras_err_event_athub_intr_no_bifring)
+		adev->nbio.funcs->handle_ras_err_event_athub_intr_no_bifring(adev);
+
 	return ret;
 }
 
@@ -228,10 +245,19 @@ int amdgpu_irq_init(struct amdgpu_device *adev)
 	adev->irq.msi_enabled = false;
 
 	if (amdgpu_msi_ok(adev)) {
-		int ret = pci_enable_msi(adev->pdev);
-		if (!ret) {
+		int nvec = pci_msix_vec_count(adev->pdev);
+		unsigned int flags;
+
+		if (nvec <= 0) {
+			flags = PCI_IRQ_MSI;
+		} else {
+			flags = PCI_IRQ_MSI | PCI_IRQ_MSIX;
+		}
+		/* we only need one vector */
+		nvec = pci_alloc_irq_vectors(adev->pdev, 1, 1, flags);
+		if (nvec > 0) {
 			adev->irq.msi_enabled = true;
-			dev_dbg(adev->dev, "amdgpu: using MSI.\n");
+			dev_dbg(adev->dev, "amdgpu: using MSI/MSI-X.\n");
 		}
 	}
 
@@ -254,7 +280,8 @@ int amdgpu_irq_init(struct amdgpu_device *adev)
 	INIT_WORK(&adev->irq.ih2_work, amdgpu_irq_handle_ih2);
 
 	adev->irq.installed = true;
-	r = drm_irq_install(adev->ddev, adev->ddev->pdev->irq);
+	/* Use vector 0 for MSI-X */
+	r = drm_irq_install(adev->ddev, pci_irq_vector(adev->pdev, 0));
 	if (r) {
 		adev->irq.installed = false;
 		if (!amdgpu_device_has_dc_support(adev))
@@ -284,7 +311,7 @@ void amdgpu_irq_fini(struct amdgpu_device *adev)
 		drm_irq_uninstall(adev->ddev);
 		adev->irq.installed = false;
 		if (adev->irq.msi_enabled)
-			pci_disable_msi(adev->pdev);
+			pci_free_irq_vectors(adev->pdev);
 		if (!amdgpu_device_has_dc_support(adev))
 			flush_work(&adev->hotplug_work);
 	}
@@ -369,7 +396,7 @@ int amdgpu_irq_add_id(struct amdgpu_device *adev,
  * amdgpu_irq_dispatch - dispatch IRQ to IP blocks
  *
  * @adev: amdgpu device pointer
- * @entry: interrupt vector pointer
+ * @ih: interrupt ring instance
  *
  * Dispatches IRQ to IP blocks.
  */
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c
index 9d76e0923a5a..4fb20e870e63 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c
@@ -218,7 +218,7 @@ static struct dma_fence *amdgpu_job_run(struct drm_sched_job *sched_job)
 	struct amdgpu_ring *ring = to_amdgpu_ring(sched_job->sched);
 	struct dma_fence *fence = NULL, *finished;
 	struct amdgpu_job *job;
-	int r;
+	int r = 0;
 
 	job = to_amdgpu_job(sched_job);
 	finished = &job->base.s_fence->finished;
@@ -243,9 +243,49 @@ static struct dma_fence *amdgpu_job_run(struct drm_sched_job *sched_job)
 	job->fence = dma_fence_get(fence);
 
 	amdgpu_job_free_resources(job);
+
+	fence = r ? ERR_PTR(r) : fence;
 	return fence;
 }
 
+#define to_drm_sched_job(sched_job)		\
+		container_of((sched_job), struct drm_sched_job, queue_node)
+
+void amdgpu_job_stop_all_jobs_on_sched(struct drm_gpu_scheduler *sched)
+{
+	struct drm_sched_job *s_job;
+	struct drm_sched_entity *s_entity = NULL;
+	int i;
+
+	/* Signal all jobs not yet scheduled */
+	for (i = DRM_SCHED_PRIORITY_MAX - 1; i >= DRM_SCHED_PRIORITY_MIN; i--) {
+		struct drm_sched_rq *rq = &sched->sched_rq[i];
+
+		if (!rq)
+			continue;
+
+		spin_lock(&rq->lock);
+		list_for_each_entry(s_entity, &rq->entities, list) {
+			while ((s_job = to_drm_sched_job(spsc_queue_pop(&s_entity->job_queue)))) {
+				struct drm_sched_fence *s_fence = s_job->s_fence;
+
+				dma_fence_signal(&s_fence->scheduled);
+				dma_fence_set_error(&s_fence->finished, -EHWPOISON);
+				dma_fence_signal(&s_fence->finished);
+			}
+		}
+		spin_unlock(&rq->lock);
+	}
+
+	/* Signal all jobs already scheduled to HW */
+	list_for_each_entry(s_job, &sched->ring_mirror_list, node) {
+		struct drm_sched_fence *s_fence = s_job->s_fence;
+
+		dma_fence_set_error(&s_fence->finished, -EHWPOISON);
+		dma_fence_signal(&s_fence->finished);
+	}
+}
+
 const struct drm_sched_backend_ops amdgpu_sched_ops = {
 	.dependency = amdgpu_job_dependency,
 	.run_job = amdgpu_job_run,
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_job.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_job.h
index 51e62504c279..dc7ee9358dcd 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_job.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_job.h
@@ -76,4 +76,7 @@ int amdgpu_job_submit(struct amdgpu_job *job, struct drm_sched_entity *entity,
 		      void *owner, struct dma_fence **f);
 int amdgpu_job_submit_direct(struct amdgpu_job *job, struct amdgpu_ring *ring,
 			     struct dma_fence **fence);
+
+void amdgpu_job_stop_all_jobs_on_sched(struct drm_gpu_scheduler *sched);
+
 #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
index f6147528be64..b6db28a570c2 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
@@ -144,41 +144,6 @@ int amdgpu_driver_load_kms(struct drm_device *dev, unsigned long flags)
 	struct amdgpu_device *adev;
 	int r, acpi_status;
 
-#ifdef CONFIG_DRM_AMDGPU_SI
-	if (!amdgpu_si_support) {
-		switch (flags & AMD_ASIC_MASK) {
-		case CHIP_TAHITI:
-		case CHIP_PITCAIRN:
-		case CHIP_VERDE:
-		case CHIP_OLAND:
-		case CHIP_HAINAN:
-			dev_info(dev->dev,
-				 "SI support provided by radeon.\n");
-			dev_info(dev->dev,
-				 "Use radeon.si_support=0 amdgpu.si_support=1 to override.\n"
-				);
-			return -ENODEV;
-		}
-	}
-#endif
-#ifdef CONFIG_DRM_AMDGPU_CIK
-	if (!amdgpu_cik_support) {
-		switch (flags & AMD_ASIC_MASK) {
-		case CHIP_KAVERI:
-		case CHIP_BONAIRE:
-		case CHIP_HAWAII:
-		case CHIP_KABINI:
-		case CHIP_MULLINS:
-			dev_info(dev->dev,
-				 "CIK support provided by radeon.\n");
-			dev_info(dev->dev,
-				 "Use radeon.cik_support=0 amdgpu.cik_support=1 to override.\n"
-				);
-			return -ENODEV;
-		}
-	}
-#endif
-
 	adev = kzalloc(sizeof(struct amdgpu_device), GFP_KERNEL);
 	if (adev == NULL) {
 		return -ENOMEM;
@@ -225,7 +190,6 @@ int amdgpu_driver_load_kms(struct drm_device *dev, unsigned long flags)
 		pm_runtime_put_autosuspend(dev->dev);
 	}
 
-	amdgpu_register_gpu_instance(adev);
 out:
 	if (r) {
 		/* balance pm_runtime_get_sync in amdgpu_driver_unload_kms */
@@ -619,9 +583,12 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file
 		struct drm_amdgpu_info_vram_gtt vram_gtt;
 
 		vram_gtt.vram_size = adev->gmc.real_vram_size -
-			atomic64_read(&adev->vram_pin_size);
-		vram_gtt.vram_cpu_accessible_size = adev->gmc.visible_vram_size -
-			atomic64_read(&adev->visible_pin_size);
+			atomic64_read(&adev->vram_pin_size) -
+			AMDGPU_VM_RESERVED_VRAM;
+		vram_gtt.vram_cpu_accessible_size =
+			min(adev->gmc.visible_vram_size -
+			    atomic64_read(&adev->visible_pin_size),
+			    vram_gtt.vram_size);
 		vram_gtt.gtt_size = adev->mman.bdev.man[TTM_PL_TT].size;
 		vram_gtt.gtt_size *= PAGE_SIZE;
 		vram_gtt.gtt_size -= atomic64_read(&adev->gart_pin_size);
@@ -634,15 +601,18 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file
 		memset(&mem, 0, sizeof(mem));
 		mem.vram.total_heap_size = adev->gmc.real_vram_size;
 		mem.vram.usable_heap_size = adev->gmc.real_vram_size -
-			atomic64_read(&adev->vram_pin_size);
+			atomic64_read(&adev->vram_pin_size) -
+			AMDGPU_VM_RESERVED_VRAM;
 		mem.vram.heap_usage =
 			amdgpu_vram_mgr_usage(&adev->mman.bdev.man[TTM_PL_VRAM]);
 		mem.vram.max_allocation = mem.vram.usable_heap_size * 3 / 4;
 
 		mem.cpu_accessible_vram.total_heap_size =
 			adev->gmc.visible_vram_size;
-		mem.cpu_accessible_vram.usable_heap_size = adev->gmc.visible_vram_size -
-			atomic64_read(&adev->visible_pin_size);
+		mem.cpu_accessible_vram.usable_heap_size =
+			min(adev->gmc.visible_vram_size -
+			    atomic64_read(&adev->visible_pin_size),
+			    mem.vram.usable_heap_size);
 		mem.cpu_accessible_vram.heap_usage =
 			amdgpu_vram_mgr_vis_usage(&adev->mman.bdev.man[TTM_PL_VRAM]);
 		mem.cpu_accessible_vram.max_allocation =
@@ -685,15 +655,19 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file
 			return -ENOMEM;
 		alloc_size = info->read_mmr_reg.count * sizeof(*regs);
 
-		for (i = 0; i < info->read_mmr_reg.count; i++)
+		amdgpu_gfx_off_ctrl(adev, false);
+		for (i = 0; i < info->read_mmr_reg.count; i++) {
 			if (amdgpu_asic_read_register(adev, se_num, sh_num,
 						      info->read_mmr_reg.dword_offset + i,
 						      &regs[i])) {
 				DRM_DEBUG_KMS("unallowed offset %#x\n",
 					      info->read_mmr_reg.dword_offset + i);
 				kfree(regs);
+				amdgpu_gfx_off_ctrl(adev, true);
 				return -EFAULT;
 			}
+		}
+		amdgpu_gfx_off_ctrl(adev, true);
 		n = copy_to_user(out, regs, min(size, alloc_size));
 		kfree(regs);
 		return n ? -EFAULT : 0;
@@ -764,17 +738,6 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file
 		dev_info.vce_harvest_config = adev->vce.harvest_config;
 		dev_info.gc_double_offchip_lds_buf =
 			adev->gfx.config.double_offchip_lds_buf;
-
-		if (amdgpu_ngg) {
-			dev_info.prim_buf_gpu_addr = adev->gfx.ngg.buf[NGG_PRIM].gpu_addr;
-			dev_info.prim_buf_size = adev->gfx.ngg.buf[NGG_PRIM].size;
-			dev_info.pos_buf_gpu_addr = adev->gfx.ngg.buf[NGG_POS].gpu_addr;
-			dev_info.pos_buf_size = adev->gfx.ngg.buf[NGG_POS].size;
-			dev_info.cntl_sb_buf_gpu_addr = adev->gfx.ngg.buf[NGG_CNTL].gpu_addr;
-			dev_info.cntl_sb_buf_size = adev->gfx.ngg.buf[NGG_CNTL].size;
-			dev_info.param_buf_gpu_addr = adev->gfx.ngg.buf[NGG_PARAM].gpu_addr;
-			dev_info.param_buf_size = adev->gfx.ngg.buf[NGG_PARAM].size;
-		}
 		dev_info.wave_front_size = adev->gfx.cu_info.wave_front_size;
 		dev_info.num_shader_visible_vgprs = adev->gfx.config.max_gprs;
 		dev_info.num_cu_per_sh = adev->gfx.config.max_cu_per_sh;
@@ -787,6 +750,8 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file
 			dev_info.pa_sc_tile_steering_override =
 				adev->gfx.config.pa_sc_tile_steering_override;
 
+		dev_info.tcc_disabled_mask = adev->gfx.config.tcc_disabled_mask;
+
 		return copy_to_user(out, &dev_info,
 				    min((size_t)size, sizeof(dev_info))) ? -EFAULT : 0;
 	}
@@ -1001,6 +966,12 @@ int amdgpu_driver_open_kms(struct drm_device *dev, struct drm_file *file_priv)
 	/* Ensure IB tests are run on ring */
 	flush_delayed_work(&adev->delayed_init_work);
 
+
+	if (amdgpu_ras_intr_triggered()) {
+		DRM_ERROR("RAS Intr triggered, device disabled!!");
+		return -EHWPOISON;
+	}
+
 	file_priv->driver_priv = NULL;
 
 	r = pm_runtime_get_sync(dev->dev);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mmhub.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_mmhub.c
new file mode 100644
index 000000000000..676c48c02d77
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mmhub.c
@@ -0,0 +1,70 @@
+/*
+ * Copyright 2019 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#include "amdgpu.h"
+#include "amdgpu_ras.h"
+
+int amdgpu_mmhub_ras_late_init(struct amdgpu_device *adev)
+{
+	int r;
+	struct ras_ih_if ih_info = {
+		.cb = NULL,
+	};
+	struct ras_fs_if fs_info = {
+		.sysfs_name = "mmhub_err_count",
+		.debugfs_name = "mmhub_err_inject",
+	};
+
+	if (!adev->mmhub.ras_if) {
+		adev->mmhub.ras_if = kmalloc(sizeof(struct ras_common_if), GFP_KERNEL);
+		if (!adev->mmhub.ras_if)
+			return -ENOMEM;
+		adev->mmhub.ras_if->block = AMDGPU_RAS_BLOCK__MMHUB;
+		adev->mmhub.ras_if->type = AMDGPU_RAS_ERROR__MULTI_UNCORRECTABLE;
+		adev->mmhub.ras_if->sub_block_index = 0;
+		strcpy(adev->mmhub.ras_if->name, "mmhub");
+	}
+	ih_info.head = fs_info.head = *adev->mmhub.ras_if;
+	r = amdgpu_ras_late_init(adev, adev->mmhub.ras_if,
+				 &fs_info, &ih_info);
+	if (r || !amdgpu_ras_is_supported(adev, adev->mmhub.ras_if->block)) {
+		kfree(adev->mmhub.ras_if);
+		adev->mmhub.ras_if = NULL;
+	}
+
+	return r;
+}
+
+void amdgpu_mmhub_ras_fini(struct amdgpu_device *adev)
+{
+	if (amdgpu_ras_is_supported(adev, AMDGPU_RAS_BLOCK__MMHUB) &&
+			adev->mmhub.ras_if) {
+		struct ras_common_if *ras_if = adev->mmhub.ras_if;
+		struct ras_ih_if ih_info = {
+			.cb = NULL,
+		};
+
+		amdgpu_ras_late_fini(adev, ras_if, &ih_info);
+		kfree(ras_if);
+	}
+}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mmhub.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mmhub.h
index 2d75ecfa199b..1cd78940cf82 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mmhub.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mmhub.h
@@ -23,9 +23,17 @@
 
 struct amdgpu_mmhub_funcs {
 	void (*ras_init)(struct amdgpu_device *adev);
+	int (*ras_late_init)(struct amdgpu_device *adev);
 	void (*query_ras_error_count)(struct amdgpu_device *adev,
 					void *ras_error_status);
 };
 
+struct amdgpu_mmhub {
+	struct ras_common_if *ras_if;
+	const struct amdgpu_mmhub_funcs *funcs;
+};
+
+int amdgpu_mmhub_ras_late_init(struct amdgpu_device *adev);
+void amdgpu_mmhub_ras_fini(struct amdgpu_device *adev);
 #endif
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c
index 31d4deb5d294..828b5167ff12 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c
@@ -51,438 +51,107 @@
 #include "amdgpu_amdkfd.h"
 
 /**
- * struct amdgpu_mn_node
+ * amdgpu_mn_invalidate_gfx - callback to notify about mm change
  *
- * @it: interval node defining start-last of the affected address range
- * @bos: list of all BOs in the affected address range
- *
- * Manages all BOs which are affected of a certain range of address space.
- */
-struct amdgpu_mn_node {
-	struct interval_tree_node	it;
-	struct list_head		bos;
-};
-
-/**
- * amdgpu_mn_destroy - destroy the HMM mirror
- *
- * @work: previously sheduled work item
- *
- * Lazy destroys the notifier from a work item
- */
-static void amdgpu_mn_destroy(struct work_struct *work)
-{
-	struct amdgpu_mn *amn = container_of(work, struct amdgpu_mn, work);
-	struct amdgpu_device *adev = amn->adev;
-	struct amdgpu_mn_node *node, *next_node;
-	struct amdgpu_bo *bo, *next_bo;
-
-	mutex_lock(&adev->mn_lock);
-	down_write(&amn->lock);
-	hash_del(&amn->node);
-	rbtree_postorder_for_each_entry_safe(node, next_node,
-					     &amn->objects.rb_root, it.rb) {
-		list_for_each_entry_safe(bo, next_bo, &node->bos, mn_list) {
-			bo->mn = NULL;
-			list_del_init(&bo->mn_list);
-		}
-		kfree(node);
-	}
-	up_write(&amn->lock);
-	mutex_unlock(&adev->mn_lock);
-
-	hmm_mirror_unregister(&amn->mirror);
-	kfree(amn);
-}
-
-/**
- * amdgpu_hmm_mirror_release - callback to notify about mm destruction
- *
- * @mirror: the HMM mirror (mm) this callback is about
- *
- * Shedule a work item to lazy destroy HMM mirror.
- */
-static void amdgpu_hmm_mirror_release(struct hmm_mirror *mirror)
-{
-	struct amdgpu_mn *amn = container_of(mirror, struct amdgpu_mn, mirror);
-
-	INIT_WORK(&amn->work, amdgpu_mn_destroy);
-	schedule_work(&amn->work);
-}
-
-/**
- * amdgpu_mn_lock - take the write side lock for this notifier
- *
- * @mn: our notifier
- */
-void amdgpu_mn_lock(struct amdgpu_mn *mn)
-{
-	if (mn)
-		down_write(&mn->lock);
-}
-
-/**
- * amdgpu_mn_unlock - drop the write side lock for this notifier
- *
- * @mn: our notifier
- */
-void amdgpu_mn_unlock(struct amdgpu_mn *mn)
-{
-	if (mn)
-		up_write(&mn->lock);
-}
-
-/**
- * amdgpu_mn_read_lock - take the read side lock for this notifier
- *
- * @amn: our notifier
- */
-static int amdgpu_mn_read_lock(struct amdgpu_mn *amn, bool blockable)
-{
-	if (blockable)
-		down_read(&amn->lock);
-	else if (!down_read_trylock(&amn->lock))
-		return -EAGAIN;
-
-	return 0;
-}
-
-/**
- * amdgpu_mn_read_unlock - drop the read side lock for this notifier
- *
- * @amn: our notifier
- */
-static void amdgpu_mn_read_unlock(struct amdgpu_mn *amn)
-{
-	up_read(&amn->lock);
-}
-
-/**
- * amdgpu_mn_invalidate_node - unmap all BOs of a node
- *
- * @node: the node with the BOs to unmap
- * @start: start of address range affected
- * @end: end of address range affected
+ * @mni: the range (mm) is about to update
+ * @range: details on the invalidation
+ * @cur_seq: Value to pass to mmu_interval_set_seq()
  *
  * Block for operations on BOs to finish and mark pages as accessed and
  * potentially dirty.
  */
-static void amdgpu_mn_invalidate_node(struct amdgpu_mn_node *node,
-				      unsigned long start,
-				      unsigned long end)
+static bool amdgpu_mn_invalidate_gfx(struct mmu_interval_notifier *mni,
+				     const struct mmu_notifier_range *range,
+				     unsigned long cur_seq)
 {
-	struct amdgpu_bo *bo;
+	struct amdgpu_bo *bo = container_of(mni, struct amdgpu_bo, notifier);
+	struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
 	long r;
 
-	list_for_each_entry(bo, &node->bos, mn_list) {
-
-		if (!amdgpu_ttm_tt_affect_userptr(bo->tbo.ttm, start, end))
-			continue;
-
-		r = dma_resv_wait_timeout_rcu(bo->tbo.base.resv,
-			true, false, MAX_SCHEDULE_TIMEOUT);
-		if (r <= 0)
-			DRM_ERROR("(%ld) failed to wait for user bo\n", r);
-	}
-}
-
-/**
- * amdgpu_mn_sync_pagetables_gfx - callback to notify about mm change
- *
- * @mirror: the hmm_mirror (mm) is about to update
- * @update: the update start, end address
- *
- * Block for operations on BOs to finish and mark pages as accessed and
- * potentially dirty.
- */
-static int
-amdgpu_mn_sync_pagetables_gfx(struct hmm_mirror *mirror,
-			      const struct mmu_notifier_range *update)
-{
-	struct amdgpu_mn *amn = container_of(mirror, struct amdgpu_mn, mirror);
-	unsigned long start = update->start;
-	unsigned long end = update->end;
-	bool blockable = mmu_notifier_range_blockable(update);
-	struct interval_tree_node *it;
+	if (!mmu_notifier_range_blockable(range))
+		return false;
 
-	/* notification is exclusive, but interval is inclusive */
-	end -= 1;
+	mutex_lock(&adev->notifier_lock);
 
-	/* TODO we should be able to split locking for interval tree and
-	 * amdgpu_mn_invalidate_node
-	 */
-	if (amdgpu_mn_read_lock(amn, blockable))
-		return -EAGAIN;
+	mmu_interval_set_seq(mni, cur_seq);
 
-	it = interval_tree_iter_first(&amn->objects, start, end);
-	while (it) {
-		struct amdgpu_mn_node *node;
-
-		if (!blockable) {
-			amdgpu_mn_read_unlock(amn);
-			return -EAGAIN;
-		}
-
-		node = container_of(it, struct amdgpu_mn_node, it);
-		it = interval_tree_iter_next(it, start, end);
-
-		amdgpu_mn_invalidate_node(node, start, end);
-	}
-
-	amdgpu_mn_read_unlock(amn);
-
-	return 0;
-}
-
-/**
- * amdgpu_mn_sync_pagetables_hsa - callback to notify about mm change
- *
- * @mirror: the hmm_mirror (mm) is about to update
- * @update: the update start, end address
- *
- * We temporarily evict all BOs between start and end. This
- * necessitates evicting all user-mode queues of the process. The BOs
- * are restorted in amdgpu_mn_invalidate_range_end_hsa.
- */
-static int
-amdgpu_mn_sync_pagetables_hsa(struct hmm_mirror *mirror,
-			      const struct mmu_notifier_range *update)
-{
-	struct amdgpu_mn *amn = container_of(mirror, struct amdgpu_mn, mirror);
-	unsigned long start = update->start;
-	unsigned long end = update->end;
-	bool blockable = mmu_notifier_range_blockable(update);
-	struct interval_tree_node *it;
-
-	/* notification is exclusive, but interval is inclusive */
-	end -= 1;
-
-	if (amdgpu_mn_read_lock(amn, blockable))
-		return -EAGAIN;
-
-	it = interval_tree_iter_first(&amn->objects, start, end);
-	while (it) {
-		struct amdgpu_mn_node *node;
-		struct amdgpu_bo *bo;
-
-		if (!blockable) {
-			amdgpu_mn_read_unlock(amn);
-			return -EAGAIN;
-		}
-
-		node = container_of(it, struct amdgpu_mn_node, it);
-		it = interval_tree_iter_next(it, start, end);
-
-		list_for_each_entry(bo, &node->bos, mn_list) {
-			struct kgd_mem *mem = bo->kfd_bo;
-
-			if (amdgpu_ttm_tt_affect_userptr(bo->tbo.ttm,
-							 start, end))
-				amdgpu_amdkfd_evict_userptr(mem, amn->mm);
-		}
-	}
-
-	amdgpu_mn_read_unlock(amn);
-
-	return 0;
+	r = dma_resv_wait_timeout_rcu(bo->tbo.base.resv, true, false,
+				      MAX_SCHEDULE_TIMEOUT);
+	mutex_unlock(&adev->notifier_lock);
+	if (r <= 0)
+		DRM_ERROR("(%ld) failed to wait for user bo\n", r);
+	return true;
 }
 
-/* Low bits of any reasonable mm pointer will be unused due to struct
- * alignment. Use these bits to make a unique key from the mm pointer
- * and notifier type.
- */
-#define AMDGPU_MN_KEY(mm, type) ((unsigned long)(mm) + (type))
-
-static struct hmm_mirror_ops amdgpu_hmm_mirror_ops[] = {
-	[AMDGPU_MN_TYPE_GFX] = {
-		.sync_cpu_device_pagetables = amdgpu_mn_sync_pagetables_gfx,
-		.release = amdgpu_hmm_mirror_release
-	},
-	[AMDGPU_MN_TYPE_HSA] = {
-		.sync_cpu_device_pagetables = amdgpu_mn_sync_pagetables_hsa,
-		.release = amdgpu_hmm_mirror_release
-	},
+static const struct mmu_interval_notifier_ops amdgpu_mn_gfx_ops = {
+	.invalidate = amdgpu_mn_invalidate_gfx,
 };
 
 /**
- * amdgpu_mn_get - create HMM mirror context
+ * amdgpu_mn_invalidate_hsa - callback to notify about mm change
  *
- * @adev: amdgpu device pointer
- * @type: type of MMU notifier context
+ * @mni: the range (mm) is about to update
+ * @range: details on the invalidation
+ * @cur_seq: Value to pass to mmu_interval_set_seq()
  *
- * Creates a HMM mirror context for current->mm.
+ * We temporarily evict the BO attached to this range. This necessitates
+ * evicting all user-mode queues of the process.
  */
-struct amdgpu_mn *amdgpu_mn_get(struct amdgpu_device *adev,
-				enum amdgpu_mn_type type)
+static bool amdgpu_mn_invalidate_hsa(struct mmu_interval_notifier *mni,
+				     const struct mmu_notifier_range *range,
+				     unsigned long cur_seq)
 {
-	struct mm_struct *mm = current->mm;
-	struct amdgpu_mn *amn;
-	unsigned long key = AMDGPU_MN_KEY(mm, type);
-	int r;
-
-	mutex_lock(&adev->mn_lock);
-	if (down_write_killable(&mm->mmap_sem)) {
-		mutex_unlock(&adev->mn_lock);
-		return ERR_PTR(-EINTR);
-	}
-
-	hash_for_each_possible(adev->mn_hash, amn, node, key)
-		if (AMDGPU_MN_KEY(amn->mm, amn->type) == key)
-			goto release_locks;
-
-	amn = kzalloc(sizeof(*amn), GFP_KERNEL);
-	if (!amn) {
-		amn = ERR_PTR(-ENOMEM);
-		goto release_locks;
-	}
-
-	amn->adev = adev;
-	amn->mm = mm;
-	init_rwsem(&amn->lock);
-	amn->type = type;
-	amn->objects = RB_ROOT_CACHED;
-
-	amn->mirror.ops = &amdgpu_hmm_mirror_ops[type];
-	r = hmm_mirror_register(&amn->mirror, mm);
-	if (r)
-		goto free_amn;
+	struct amdgpu_bo *bo = container_of(mni, struct amdgpu_bo, notifier);
+	struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
 
-	hash_add(adev->mn_hash, &amn->node, AMDGPU_MN_KEY(mm, type));
+	if (!mmu_notifier_range_blockable(range))
+		return false;
 
-release_locks:
-	up_write(&mm->mmap_sem);
-	mutex_unlock(&adev->mn_lock);
+	mutex_lock(&adev->notifier_lock);
 
-	return amn;
+	mmu_interval_set_seq(mni, cur_seq);
 
-free_amn:
-	up_write(&mm->mmap_sem);
-	mutex_unlock(&adev->mn_lock);
-	kfree(amn);
+	amdgpu_amdkfd_evict_userptr(bo->kfd_bo, bo->notifier.mm);
+	mutex_unlock(&adev->notifier_lock);
 
-	return ERR_PTR(r);
+	return true;
 }
 
+static const struct mmu_interval_notifier_ops amdgpu_mn_hsa_ops = {
+	.invalidate = amdgpu_mn_invalidate_hsa,
+};
+
 /**
  * amdgpu_mn_register - register a BO for notifier updates
  *
  * @bo: amdgpu buffer object
  * @addr: userptr addr we should monitor
  *
- * Registers an HMM mirror for the given BO at the specified address.
+ * Registers a mmu_notifier for the given BO at the specified address.
  * Returns 0 on success, -ERRNO if anything goes wrong.
  */
 int amdgpu_mn_register(struct amdgpu_bo *bo, unsigned long addr)
 {
-	unsigned long end = addr + amdgpu_bo_size(bo) - 1;
-	struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
-	enum amdgpu_mn_type type =
-		bo->kfd_bo ? AMDGPU_MN_TYPE_HSA : AMDGPU_MN_TYPE_GFX;
-	struct amdgpu_mn *amn;
-	struct amdgpu_mn_node *node = NULL, *new_node;
-	struct list_head bos;
-	struct interval_tree_node *it;
-
-	amn = amdgpu_mn_get(adev, type);
-	if (IS_ERR(amn))
-		return PTR_ERR(amn);
-
-	new_node = kmalloc(sizeof(*new_node), GFP_KERNEL);
-	if (!new_node)
-		return -ENOMEM;
-
-	INIT_LIST_HEAD(&bos);
-
-	down_write(&amn->lock);
-
-	while ((it = interval_tree_iter_first(&amn->objects, addr, end))) {
-		kfree(node);
-		node = container_of(it, struct amdgpu_mn_node, it);
-		interval_tree_remove(&node->it, &amn->objects);
-		addr = min(it->start, addr);
-		end = max(it->last, end);
-		list_splice(&node->bos, &bos);
-	}
-
-	if (!node)
-		node = new_node;
-	else
-		kfree(new_node);
-
-	bo->mn = amn;
-
-	node->it.start = addr;
-	node->it.last = end;
-	INIT_LIST_HEAD(&node->bos);
-	list_splice(&bos, &node->bos);
-	list_add(&bo->mn_list, &node->bos);
-
-	interval_tree_insert(&node->it, &amn->objects);
-
-	up_write(&amn->lock);
-
-	return 0;
+	if (bo->kfd_bo)
+		return mmu_interval_notifier_insert(&bo->notifier, current->mm,
+						    addr, amdgpu_bo_size(bo),
+						    &amdgpu_mn_hsa_ops);
+	return mmu_interval_notifier_insert(&bo->notifier, current->mm, addr,
+					    amdgpu_bo_size(bo),
+					    &amdgpu_mn_gfx_ops);
 }
 
 /**
- * amdgpu_mn_unregister - unregister a BO for HMM mirror updates
+ * amdgpu_mn_unregister - unregister a BO for notifier updates
  *
  * @bo: amdgpu buffer object
  *
- * Remove any registration of HMM mirror updates from the buffer object.
+ * Remove any registration of mmu notifier updates from the buffer object.
  */
 void amdgpu_mn_unregister(struct amdgpu_bo *bo)
 {
-	struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
-	struct amdgpu_mn *amn;
-	struct list_head *head;
-
-	mutex_lock(&adev->mn_lock);
-
-	amn = bo->mn;
-	if (amn == NULL) {
-		mutex_unlock(&adev->mn_lock);
+	if (!bo->notifier.mm)
 		return;
-	}
-
-	down_write(&amn->lock);
-
-	/* save the next list entry for later */
-	head = bo->mn_list.next;
-
-	bo->mn = NULL;
-	list_del_init(&bo->mn_list);
-
-	if (list_empty(head)) {
-		struct amdgpu_mn_node *node;
-
-		node = container_of(head, struct amdgpu_mn_node, bos);
-		interval_tree_remove(&node->it, &amn->objects);
-		kfree(node);
-	}
-
-	up_write(&amn->lock);
-	mutex_unlock(&adev->mn_lock);
-}
-
-/* flags used by HMM internal, not related to CPU/GPU PTE flags */
-static const uint64_t hmm_range_flags[HMM_PFN_FLAG_MAX] = {
-		(1 << 0), /* HMM_PFN_VALID */
-		(1 << 1), /* HMM_PFN_WRITE */
-		0 /* HMM_PFN_DEVICE_PRIVATE */
-};
-
-static const uint64_t hmm_range_values[HMM_PFN_VALUE_MAX] = {
-		0xfffffffffffffffeUL, /* HMM_PFN_ERROR */
-		0, /* HMM_PFN_NONE */
-		0xfffffffffffffffcUL /* HMM_PFN_SPECIAL */
-};
-
-void amdgpu_hmm_init_range(struct hmm_range *range)
-{
-	if (range) {
-		range->flags = hmm_range_flags;
-		range->values = hmm_range_values;
-		range->pfn_shift = PAGE_SHIFT;
-	}
+	mmu_interval_notifier_remove(&bo->notifier);
+	bo->notifier.mm = NULL;
 }
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.h
index b8ed68943625..a292238f75eb 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.h
@@ -30,63 +30,10 @@
 #include <linux/workqueue.h>
 #include <linux/interval_tree.h>
 
-enum amdgpu_mn_type {
-	AMDGPU_MN_TYPE_GFX,
-	AMDGPU_MN_TYPE_HSA,
-};
-
-/**
- * struct amdgpu_mn
- *
- * @adev: amdgpu device pointer
- * @mm: process address space
- * @type: type of MMU notifier
- * @work: destruction work item
- * @node: hash table node to find structure by adev and mn
- * @lock: rw semaphore protecting the notifier nodes
- * @objects: interval tree containing amdgpu_mn_nodes
- * @mirror: HMM mirror function support
- *
- * Data for each amdgpu device and process address space.
- */
-struct amdgpu_mn {
-	/* constant after initialisation */
-	struct amdgpu_device	*adev;
-	struct mm_struct	*mm;
-	enum amdgpu_mn_type	type;
-
-	/* only used on destruction */
-	struct work_struct	work;
-
-	/* protected by adev->mn_lock */
-	struct hlist_node	node;
-
-	/* objects protected by lock */
-	struct rw_semaphore	lock;
-	struct rb_root_cached	objects;
-
-#ifdef CONFIG_HMM_MIRROR
-	/* HMM mirror */
-	struct hmm_mirror	mirror;
-#endif
-};
-
 #if defined(CONFIG_HMM_MIRROR)
-void amdgpu_mn_lock(struct amdgpu_mn *mn);
-void amdgpu_mn_unlock(struct amdgpu_mn *mn);
-struct amdgpu_mn *amdgpu_mn_get(struct amdgpu_device *adev,
-				enum amdgpu_mn_type type);
 int amdgpu_mn_register(struct amdgpu_bo *bo, unsigned long addr);
 void amdgpu_mn_unregister(struct amdgpu_bo *bo);
-void amdgpu_hmm_init_range(struct hmm_range *range);
 #else
-static inline void amdgpu_mn_lock(struct amdgpu_mn *mn) {}
-static inline void amdgpu_mn_unlock(struct amdgpu_mn *mn) {}
-static inline struct amdgpu_mn *amdgpu_mn_get(struct amdgpu_device *adev,
-					      enum amdgpu_mn_type type)
-{
-	return NULL;
-}
 static inline int amdgpu_mn_register(struct amdgpu_bo *bo, unsigned long addr)
 {
 	DRM_WARN_ONCE("HMM_MIRROR kernel config option is not enabled, "
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_nbio.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_nbio.c
new file mode 100644
index 000000000000..7d5c3a9de9ea
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_nbio.c
@@ -0,0 +1,84 @@
+/*
+ * Copyright (C) 2019  Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included
+ * in all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
+ * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN
+ * AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+ */
+
+#include "amdgpu.h"
+#include "amdgpu_ras.h"
+
+int amdgpu_nbio_ras_late_init(struct amdgpu_device *adev)
+{
+	int r;
+	struct ras_ih_if ih_info = {
+		.cb = NULL,
+	};
+	struct ras_fs_if fs_info = {
+		.sysfs_name = "pcie_bif_err_count",
+		.debugfs_name = "pcie_bif_err_inject",
+	};
+
+	if (!adev->nbio.ras_if) {
+		adev->nbio.ras_if = kmalloc(sizeof(struct ras_common_if), GFP_KERNEL);
+		if (!adev->nbio.ras_if)
+			return -ENOMEM;
+		adev->nbio.ras_if->block = AMDGPU_RAS_BLOCK__PCIE_BIF;
+		adev->nbio.ras_if->type = AMDGPU_RAS_ERROR__MULTI_UNCORRECTABLE;
+		adev->nbio.ras_if->sub_block_index = 0;
+		strcpy(adev->nbio.ras_if->name, "pcie_bif");
+	}
+	ih_info.head = fs_info.head = *adev->nbio.ras_if;
+	r = amdgpu_ras_late_init(adev, adev->nbio.ras_if,
+				 &fs_info, &ih_info);
+	if (r)
+		goto free;
+
+	if (amdgpu_ras_is_supported(adev, adev->nbio.ras_if->block)) {
+		r = amdgpu_irq_get(adev, &adev->nbio.ras_controller_irq, 0);
+		if (r)
+			goto late_fini;
+		r = amdgpu_irq_get(adev, &adev->nbio.ras_err_event_athub_irq, 0);
+		if (r)
+			goto late_fini;
+	} else {
+		r = 0;
+		goto free;
+	}
+
+	return 0;
+late_fini:
+	amdgpu_ras_late_fini(adev, adev->nbio.ras_if, &ih_info);
+free:
+	kfree(adev->nbio.ras_if);
+	adev->nbio.ras_if = NULL;
+	return r;
+}
+
+void amdgpu_nbio_ras_fini(struct amdgpu_device *adev)
+{
+	if (amdgpu_ras_is_supported(adev, AMDGPU_RAS_BLOCK__PCIE_BIF) &&
+			adev->nbio.ras_if) {
+		struct ras_common_if *ras_if = adev->nbio.ras_if;
+		struct ras_ih_if ih_info = {
+			.cb = NULL,
+		};
+
+		amdgpu_ras_late_fini(adev, ras_if, &ih_info);
+		kfree(ras_if);
+	}
+}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_nbio.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_nbio.h
new file mode 100644
index 000000000000..919bd566ba3c
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_nbio.h
@@ -0,0 +1,101 @@
+/*
+ * Copyright 2019 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#ifndef __AMDGPU_NBIO_H__
+#define __AMDGPU_NBIO_H__
+
+/*
+ * amdgpu nbio functions
+ */
+struct nbio_hdp_flush_reg {
+	u32 ref_and_mask_cp0;
+	u32 ref_and_mask_cp1;
+	u32 ref_and_mask_cp2;
+	u32 ref_and_mask_cp3;
+	u32 ref_and_mask_cp4;
+	u32 ref_and_mask_cp5;
+	u32 ref_and_mask_cp6;
+	u32 ref_and_mask_cp7;
+	u32 ref_and_mask_cp8;
+	u32 ref_and_mask_cp9;
+	u32 ref_and_mask_sdma0;
+	u32 ref_and_mask_sdma1;
+	u32 ref_and_mask_sdma2;
+	u32 ref_and_mask_sdma3;
+	u32 ref_and_mask_sdma4;
+	u32 ref_and_mask_sdma5;
+	u32 ref_and_mask_sdma6;
+	u32 ref_and_mask_sdma7;
+};
+
+struct amdgpu_nbio_funcs {
+	const struct nbio_hdp_flush_reg *hdp_flush_reg;
+	u32 (*get_hdp_flush_req_offset)(struct amdgpu_device *adev);
+	u32 (*get_hdp_flush_done_offset)(struct amdgpu_device *adev);
+	u32 (*get_pcie_index_offset)(struct amdgpu_device *adev);
+	u32 (*get_pcie_data_offset)(struct amdgpu_device *adev);
+	u32 (*get_rev_id)(struct amdgpu_device *adev);
+	void (*mc_access_enable)(struct amdgpu_device *adev, bool enable);
+	void (*hdp_flush)(struct amdgpu_device *adev, struct amdgpu_ring *ring);
+	u32 (*get_memsize)(struct amdgpu_device *adev);
+	void (*sdma_doorbell_range)(struct amdgpu_device *adev, int instance,
+			bool use_doorbell, int doorbell_index, int doorbell_size);
+	void (*vcn_doorbell_range)(struct amdgpu_device *adev, bool use_doorbell,
+				   int doorbell_index, int instance);
+	void (*enable_doorbell_aperture)(struct amdgpu_device *adev,
+					 bool enable);
+	void (*enable_doorbell_selfring_aperture)(struct amdgpu_device *adev,
+						  bool enable);
+	void (*ih_doorbell_range)(struct amdgpu_device *adev,
+				  bool use_doorbell, int doorbell_index);
+	void (*enable_doorbell_interrupt)(struct amdgpu_device *adev,
+					  bool enable);
+	void (*update_medium_grain_clock_gating)(struct amdgpu_device *adev,
+						 bool enable);
+	void (*update_medium_grain_light_sleep)(struct amdgpu_device *adev,
+						bool enable);
+	void (*get_clockgating_state)(struct amdgpu_device *adev,
+				      u32 *flags);
+	void (*ih_control)(struct amdgpu_device *adev);
+	void (*init_registers)(struct amdgpu_device *adev);
+	void (*detect_hw_virt)(struct amdgpu_device *adev);
+	void (*remap_hdp_registers)(struct amdgpu_device *adev);
+	void (*handle_ras_controller_intr_no_bifring)(struct amdgpu_device *adev);
+	void (*handle_ras_err_event_athub_intr_no_bifring)(struct amdgpu_device *adev);
+	int (*init_ras_controller_interrupt)(struct amdgpu_device *adev);
+	int (*init_ras_err_event_athub_interrupt)(struct amdgpu_device *adev);
+	void (*query_ras_error_count)(struct amdgpu_device *adev,
+					void *ras_error_status);
+	int (*ras_late_init)(struct amdgpu_device *adev);
+};
+
+struct amdgpu_nbio {
+	const struct nbio_hdp_flush_reg *hdp_flush_reg;
+	struct amdgpu_irq_src ras_controller_irq;
+	struct amdgpu_irq_src ras_err_event_athub_irq;
+	struct ras_common_if *ras_if;
+	const struct amdgpu_nbio_funcs *funcs;
+};
+
+int amdgpu_nbio_ras_late_init(struct amdgpu_device *adev);
+void amdgpu_nbio_ras_fini(struct amdgpu_device *adev);
+#endif
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
index 1fead0e8b890..e3f16b49e970 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
@@ -343,6 +343,70 @@ int amdgpu_bo_create_kernel(struct amdgpu_device *adev,
 }
 
 /**
+ * amdgpu_bo_create_kernel_at - create BO for kernel use at specific location
+ *
+ * @adev: amdgpu device object
+ * @offset: offset of the BO
+ * @size: size of the BO
+ * @domain: where to place it
+ * @bo_ptr:  used to initialize BOs in structures
+ * @cpu_addr: optional CPU address mapping
+ *
+ * Creates a kernel BO at a specific offset in the address space of the domain.
+ *
+ * Returns:
+ * 0 on success, negative error code otherwise.
+ */
+int amdgpu_bo_create_kernel_at(struct amdgpu_device *adev,
+			       uint64_t offset, uint64_t size, uint32_t domain,
+			       struct amdgpu_bo **bo_ptr, void **cpu_addr)
+{
+	struct ttm_operation_ctx ctx = { false, false };
+	unsigned int i;
+	int r;
+
+	offset &= PAGE_MASK;
+	size = ALIGN(size, PAGE_SIZE);
+
+	r = amdgpu_bo_create_reserved(adev, size, PAGE_SIZE, domain, bo_ptr,
+				      NULL, cpu_addr);
+	if (r)
+		return r;
+
+	/*
+	 * Remove the original mem node and create a new one at the request
+	 * position.
+	 */
+	if (cpu_addr)
+		amdgpu_bo_kunmap(*bo_ptr);
+
+	ttm_bo_mem_put(&(*bo_ptr)->tbo, &(*bo_ptr)->tbo.mem);
+
+	for (i = 0; i < (*bo_ptr)->placement.num_placement; ++i) {
+		(*bo_ptr)->placements[i].fpfn = offset >> PAGE_SHIFT;
+		(*bo_ptr)->placements[i].lpfn = (offset + size) >> PAGE_SHIFT;
+	}
+	r = ttm_bo_mem_space(&(*bo_ptr)->tbo, &(*bo_ptr)->placement,
+			     &(*bo_ptr)->tbo.mem, &ctx);
+	if (r)
+		goto error;
+
+	if (cpu_addr) {
+		r = amdgpu_bo_kmap(*bo_ptr, cpu_addr);
+		if (r)
+			goto error;
+	}
+
+	amdgpu_bo_unreserve(*bo_ptr);
+	return 0;
+
+error:
+	amdgpu_bo_unreserve(*bo_ptr);
+	amdgpu_bo_unref(bo_ptr);
+	return r;
+}
+
+/**
  * amdgpu_bo_free_kernel - free BO for kernel use
  *
  * @bo: amdgpu BO to free
@@ -451,9 +515,10 @@ static int amdgpu_bo_do_create(struct amdgpu_device *adev,
 {
 	struct ttm_operation_ctx ctx = {
 		.interruptible = (bp->type != ttm_bo_type_kernel),
-		.no_wait_gpu = false,
+		.no_wait_gpu = bp->no_wait_gpu,
 		.resv = bp->resv,
-		.flags = TTM_OPT_FLAG_ALLOW_RES_EVICT
+		.flags = bp->type != ttm_bo_type_kernel ?
+			TTM_OPT_FLAG_ALLOW_RES_EVICT : 0
 	};
 	struct amdgpu_bo *bo;
 	unsigned long page_align, size = bp->size;
@@ -1058,7 +1123,10 @@ void amdgpu_bo_fini(struct amdgpu_device *adev)
 int amdgpu_bo_fbdev_mmap(struct amdgpu_bo *bo,
 			     struct vm_area_struct *vma)
 {
-	return ttm_fbdev_mmap(vma, &bo->tbo);
+	if (vma->vm_pgoff != 0)
+		return -EACCES;
+
+	return ttm_bo_mmap_obj(vma, &bo->tbo);
 }
 
 /**
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
index 658f4c9779b7..36dec51d1ef1 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
@@ -30,6 +30,9 @@
 
 #include <drm/amdgpu_drm.h>
 #include "amdgpu.h"
+#ifdef CONFIG_MMU_NOTIFIER
+#include <linux/mmu_notifier.h>
+#endif
 
 #define AMDGPU_BO_INVALID_OFFSET	LONG_MAX
 #define AMDGPU_BO_MAX_PLACEMENTS	3
@@ -41,6 +44,7 @@ struct amdgpu_bo_param {
 	u32				preferred_domain;
 	u64				flags;
 	enum ttm_bo_type		type;
+	bool				no_wait_gpu;
 	struct dma_resv	*resv;
 };
 
@@ -100,10 +104,12 @@ struct amdgpu_bo {
 	struct ttm_bo_kmap_obj		dma_buf_vmap;
 	struct amdgpu_mn		*mn;
 
-	union {
-		struct list_head	mn_list;
-		struct list_head	shadow_list;
-	};
+
+#ifdef CONFIG_MMU_NOTIFIER
+	struct mmu_interval_notifier	notifier;
+#endif
+
+	struct list_head		shadow_list;
 
 	struct kgd_mem                  *kfd_bo;
 };
@@ -237,6 +243,9 @@ int amdgpu_bo_create_kernel(struct amdgpu_device *adev,
 			    unsigned long size, int align,
 			    u32 domain, struct amdgpu_bo **bo_ptr,
 			    u64 *gpu_addr, void **cpu_addr);
+int amdgpu_bo_create_kernel_at(struct amdgpu_device *adev,
+			       uint64_t offset, uint64_t size, uint32_t domain,
+			       struct amdgpu_bo **bo_ptr, void **cpu_addr);
 void amdgpu_bo_free_kernel(struct amdgpu_bo **bo, u64 *gpu_addr,
 			   void **cpu_addr);
 int amdgpu_bo_kmap(struct amdgpu_bo *bo, void **ptr);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
index 03930313c263..f205f56e3358 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
@@ -161,7 +161,7 @@ static ssize_t amdgpu_get_dpm_state(struct device *dev,
 
 	if (is_support_sw_smu(adev)) {
 		if (adev->smu.ppt_funcs->get_current_power_state)
-			pm = amdgpu_smu_get_current_power_state(adev);
+			pm = smu_get_current_power_state(&adev->smu);
 		else
 			pm = adev->pm.dpm.user_state;
 	} else if (adev->powerplay.pp_funcs->get_current_power_state) {
@@ -805,8 +805,7 @@ static ssize_t amdgpu_get_pp_feature_status(struct device *dev,
 }
 
 /**
- * DOC: pp_dpm_sclk pp_dpm_mclk pp_dpm_socclk pp_dpm_fclk pp_dpm_dcefclk
- * pp_dpm_pcie
+ * DOC: pp_dpm_sclk pp_dpm_mclk pp_dpm_socclk pp_dpm_fclk pp_dpm_dcefclk pp_dpm_pcie
  *
  * The amdgpu driver provides a sysfs API for adjusting what power levels
  * are enabled for a given power state.  The files pp_dpm_sclk, pp_dpm_mclk,
@@ -822,9 +821,15 @@ static ssize_t amdgpu_get_pp_feature_status(struct device *dev,
  *
  * To manually adjust these states, first select manual using
  * power_dpm_force_performance_level.
- * Secondly,Enter a new value for each level by inputing a string that
+ * Secondly, enter a new value for each level by inputing a string that
  * contains " echo xx xx xx > pp_dpm_sclk/mclk/pcie"
- * E.g., echo 4 5 6 to > pp_dpm_sclk will enable sclk levels 4, 5, and 6.
+ * E.g.,
+ *
+ * .. code-block:: bash
+ *
+ *	echo "4 5 6" > pp_dpm_sclk
+ *
+ * will enable sclk levels 4, 5, and 6.
  *
  * NOTE: change to the dcefclk max dpm level is not supported now
  */
@@ -902,7 +907,7 @@ static ssize_t amdgpu_set_pp_dpm_sclk(struct device *dev,
 		return ret;
 
 	if (is_support_sw_smu(adev))
-		ret = smu_force_clk_levels(&adev->smu, SMU_SCLK, mask);
+		ret = smu_force_clk_levels(&adev->smu, SMU_SCLK, mask, true);
 	else if (adev->powerplay.pp_funcs->force_clock_level)
 		ret = amdgpu_dpm_force_clock_level(adev, PP_SCLK, mask);
 
@@ -949,7 +954,7 @@ static ssize_t amdgpu_set_pp_dpm_mclk(struct device *dev,
 		return ret;
 
 	if (is_support_sw_smu(adev))
-		ret = smu_force_clk_levels(&adev->smu, SMU_MCLK, mask);
+		ret = smu_force_clk_levels(&adev->smu, SMU_MCLK, mask, true);
 	else if (adev->powerplay.pp_funcs->force_clock_level)
 		ret = amdgpu_dpm_force_clock_level(adev, PP_MCLK, mask);
 
@@ -989,7 +994,7 @@ static ssize_t amdgpu_set_pp_dpm_socclk(struct device *dev,
 		return ret;
 
 	if (is_support_sw_smu(adev))
-		ret = smu_force_clk_levels(&adev->smu, SMU_SOCCLK, mask);
+		ret = smu_force_clk_levels(&adev->smu, SMU_SOCCLK, mask, true);
 	else if (adev->powerplay.pp_funcs->force_clock_level)
 		ret = amdgpu_dpm_force_clock_level(adev, PP_SOCCLK, mask);
 
@@ -1029,7 +1034,7 @@ static ssize_t amdgpu_set_pp_dpm_fclk(struct device *dev,
 		return ret;
 
 	if (is_support_sw_smu(adev))
-		ret = smu_force_clk_levels(&adev->smu, SMU_FCLK, mask);
+		ret = smu_force_clk_levels(&adev->smu, SMU_FCLK, mask, true);
 	else if (adev->powerplay.pp_funcs->force_clock_level)
 		ret = amdgpu_dpm_force_clock_level(adev, PP_FCLK, mask);
 
@@ -1069,7 +1074,7 @@ static ssize_t amdgpu_set_pp_dpm_dcefclk(struct device *dev,
 		return ret;
 
 	if (is_support_sw_smu(adev))
-		ret = smu_force_clk_levels(&adev->smu, SMU_DCEFCLK, mask);
+		ret = smu_force_clk_levels(&adev->smu, SMU_DCEFCLK, mask, true);
 	else if (adev->powerplay.pp_funcs->force_clock_level)
 		ret = amdgpu_dpm_force_clock_level(adev, PP_DCEFCLK, mask);
 
@@ -1109,7 +1114,7 @@ static ssize_t amdgpu_set_pp_dpm_pcie(struct device *dev,
 		return ret;
 
 	if (is_support_sw_smu(adev))
-		ret = smu_force_clk_levels(&adev->smu, SMU_PCIE, mask);
+		ret = smu_force_clk_levels(&adev->smu, SMU_PCIE, mask, true);
 	else if (adev->powerplay.pp_funcs->force_clock_level)
 		ret = amdgpu_dpm_force_clock_level(adev, PP_PCIE, mask);
 
@@ -1301,7 +1306,7 @@ static ssize_t amdgpu_set_pp_power_profile_mode(struct device *dev,
 	}
 	parameter[parameter_size] = profile_mode;
 	if (is_support_sw_smu(adev))
-		ret = smu_set_power_profile_mode(&adev->smu, parameter, parameter_size);
+		ret = smu_set_power_profile_mode(&adev->smu, parameter, parameter_size, true);
 	else if (adev->powerplay.pp_funcs->set_power_profile_mode)
 		ret = amdgpu_dpm_set_power_profile_mode(adev, parameter, parameter_size);
 	if (!ret)
@@ -2010,7 +2015,7 @@ static ssize_t amdgpu_hwmon_show_power_cap_max(struct device *dev,
 	uint32_t limit = 0;
 
 	if (is_support_sw_smu(adev)) {
-		smu_get_power_limit(&adev->smu, &limit, true);
+		smu_get_power_limit(&adev->smu, &limit, true, true);
 		return snprintf(buf, PAGE_SIZE, "%u\n", limit * 1000000);
 	} else if (adev->powerplay.pp_funcs && adev->powerplay.pp_funcs->get_power_limit) {
 		adev->powerplay.pp_funcs->get_power_limit(adev->powerplay.pp_handle, &limit, true);
@@ -2028,7 +2033,7 @@ static ssize_t amdgpu_hwmon_show_power_cap(struct device *dev,
 	uint32_t limit = 0;
 
 	if (is_support_sw_smu(adev)) {
-		smu_get_power_limit(&adev->smu, &limit, false);
+		smu_get_power_limit(&adev->smu, &limit, false,  true);
 		return snprintf(buf, PAGE_SIZE, "%u\n", limit * 1000000);
 	} else if (adev->powerplay.pp_funcs && adev->powerplay.pp_funcs->get_power_limit) {
 		adev->powerplay.pp_funcs->get_power_limit(adev->powerplay.pp_handle, &limit, false);
@@ -2196,9 +2201,9 @@ static ssize_t amdgpu_hwmon_show_mclk_label(struct device *dev,
  *
  * - fan1_input: fan speed in RPM
  *
- * - fan[1-*]_target: Desired fan speed Unit: revolution/min (RPM)
+ * - fan[1-\*]_target: Desired fan speed Unit: revolution/min (RPM)
  *
- * - fan[1-*]_enable: Enable or disable the sensors.1: Enable 0: Disable
+ * - fan[1-\*]_enable: Enable or disable the sensors.1: Enable 0: Disable
  *
  * hwmon interfaces for GPU clocks:
  *
@@ -2825,6 +2830,19 @@ int amdgpu_pm_sysfs_init(struct amdgpu_device *adev)
 		DRM_ERROR("failed to create device file pp_dpm_sclk\n");
 		return ret;
 	}
+
+	/* Arcturus does not support standalone mclk/socclk/fclk level setting */
+	if (adev->asic_type == CHIP_ARCTURUS) {
+		dev_attr_pp_dpm_mclk.attr.mode &= ~S_IWUGO;
+		dev_attr_pp_dpm_mclk.store = NULL;
+
+		dev_attr_pp_dpm_socclk.attr.mode &= ~S_IWUGO;
+		dev_attr_pp_dpm_socclk.store = NULL;
+
+		dev_attr_pp_dpm_fclk.attr.mode &= ~S_IWUGO;
+		dev_attr_pp_dpm_fclk.store = NULL;
+	}
+
 	ret = device_create_file(adev->dev, &dev_attr_pp_dpm_mclk);
 	if (ret) {
 		DRM_ERROR("failed to create device file pp_dpm_mclk\n");
@@ -3008,7 +3026,8 @@ void amdgpu_pm_compute_clocks(struct amdgpu_device *adev)
 		struct smu_dpm_context *smu_dpm = &adev->smu.smu_dpm;
 		smu_handle_task(&adev->smu,
 				smu_dpm->dpm_level,
-				AMD_PP_TASK_DISPLAY_CONFIG_CHANGE);
+				AMD_PP_TASK_DISPLAY_CONFIG_CHANGE,
+				true);
 	} else {
 		if (adev->powerplay.pp_funcs->dispatch_tasks) {
 			if (!amdgpu_device_has_dc_support(adev)) {
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
index 4d71537a960d..2770cba56a6b 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
@@ -34,6 +34,8 @@
 #include "psp_v11_0.h"
 #include "psp_v12_0.h"
 
+#include "amdgpu_ras.h"
+
 static void psp_set_funcs(struct amdgpu_device *adev);
 
 static int psp_early_init(void *handle)
@@ -88,6 +90,17 @@ static int psp_sw_init(void *handle)
 		return ret;
 	}
 
+	ret = psp_mem_training_init(psp);
+	if (ret) {
+		DRM_ERROR("Failed to initialize memory training!\n");
+		return ret;
+	}
+	ret = psp_mem_training(psp, PSP_MEM_TRAIN_COLD_BOOT);
+	if (ret) {
+		DRM_ERROR("Failed to process memory training!\n");
+		return ret;
+	}
+
 	return 0;
 }
 
@@ -95,6 +108,7 @@ static int psp_sw_fini(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
+	psp_mem_training_fini(&adev->psp);
 	release_firmware(adev->psp.sos_fw);
 	adev->psp.sos_fw = NULL;
 	release_firmware(adev->psp.asd_fw);
@@ -151,10 +165,19 @@ psp_cmd_submit_buf(struct psp_context *psp,
 		return ret;
 	}
 
+	amdgpu_asic_invalidate_hdp(psp->adev, NULL);
 	while (*((unsigned int *)psp->fence_buf) != index) {
 		if (--timeout == 0)
 			break;
+		/*
+		 * Shouldn't wait for timeout when err_event_athub occurs,
+		 * because gpu reset thread triggered and lock resource should
+		 * be released for psp resume sequence.
+		 */
+		if (amdgpu_ras_intr_triggered())
+			break;
 		msleep(1);
+		amdgpu_asic_invalidate_hdp(psp->adev, NULL);
 	}
 
 	/* In some cases, psp response status is not 0 even there is no
@@ -168,8 +191,9 @@ psp_cmd_submit_buf(struct psp_context *psp,
 		if (ucode)
 			DRM_WARN("failed to load ucode id (%d) ",
 				  ucode->ucode_id);
-		DRM_WARN("psp command failed and response status is (0x%X)\n",
-			  psp->cmd_buf_mem->resp.status & GFX_CMD_STATUS_MASK);
+		DRM_DEBUG_DRIVER("psp command (0x%X) failed and response status is (0x%X)\n",
+			 psp->cmd_buf_mem->cmd_id,
+			 psp->cmd_buf_mem->resp.status & GFX_CMD_STATUS_MASK);
 		if (!timeout) {
 			mutex_unlock(&psp->mutex);
 			return -EINVAL;
@@ -253,7 +277,8 @@ static int psp_tmr_init(struct psp_context *psp)
 
 	/* For ASICs support RLC autoload, psp will parse the toc
 	 * and calculate the total size of TMR needed */
-	if (psp->toc_start_addr &&
+	if (!amdgpu_sriov_vf(psp->adev) &&
+	    psp->toc_start_addr &&
 	    psp->toc_bin_size &&
 	    psp->fw_pri_buf) {
 		ret = psp_load_toc(psp, &tmr_size);
@@ -287,15 +312,9 @@ static int psp_tmr_load(struct psp_context *psp)
 
 	ret = psp_cmd_submit_buf(psp, NULL, cmd,
 				 psp->fence_buf_mc_addr);
-	if (ret)
-		goto failed;
 
 	kfree(cmd);
 
-	return 0;
-
-failed:
-	kfree(cmd);
 	return ret;
 }
 
@@ -548,7 +567,9 @@ static int psp_xgmi_initialize(struct psp_context *psp)
 	struct ta_xgmi_shared_memory *xgmi_cmd;
 	int ret;
 
-	if (!psp->adev->psp.ta_fw)
+	if (!psp->adev->psp.ta_fw ||
+	    !psp->adev->psp.ta_xgmi_ucode_size ||
+	    !psp->adev->psp.ta_xgmi_start_addr)
 		return -ENOENT;
 
 	if (!psp->xgmi_context.initialized) {
@@ -737,6 +758,12 @@ static int psp_ras_terminate(struct psp_context *psp)
 {
 	int ret;
 
+	/*
+	 * TODO: bypass the terminate in sriov for now
+	 */
+	if (amdgpu_sriov_vf(psp->adev))
+		return 0;
+
 	if (!psp->ras.ras_initialized)
 		return 0;
 
@@ -758,6 +785,18 @@ static int psp_ras_initialize(struct psp_context *psp)
 {
 	int ret;
 
+	/*
+	 * TODO: bypass the initialize in sriov for now
+	 */
+	if (amdgpu_sriov_vf(psp->adev))
+		return 0;
+
+	if (!psp->adev->psp.ta_ras_ucode_size ||
+	    !psp->adev->psp.ta_ras_start_addr) {
+		dev_warn(psp->adev->dev, "RAS: ras ta ucode is not available\n");
+		return 0;
+	}
+
 	if (!psp->ras.ras_initialized) {
 		ret = psp_ras_init_shared_buf(psp);
 		if (ret)
@@ -772,6 +811,360 @@ static int psp_ras_initialize(struct psp_context *psp)
 }
 // ras end
 
+// HDCP start
+static void psp_prep_hdcp_ta_load_cmd_buf(struct psp_gfx_cmd_resp *cmd,
+					  uint64_t hdcp_ta_mc,
+					  uint64_t hdcp_mc_shared,
+					  uint32_t hdcp_ta_size,
+					  uint32_t shared_size)
+{
+	cmd->cmd_id = GFX_CMD_ID_LOAD_TA;
+	cmd->cmd.cmd_load_ta.app_phy_addr_lo = lower_32_bits(hdcp_ta_mc);
+	cmd->cmd.cmd_load_ta.app_phy_addr_hi = upper_32_bits(hdcp_ta_mc);
+	cmd->cmd.cmd_load_ta.app_len = hdcp_ta_size;
+
+	cmd->cmd.cmd_load_ta.cmd_buf_phy_addr_lo =
+		lower_32_bits(hdcp_mc_shared);
+	cmd->cmd.cmd_load_ta.cmd_buf_phy_addr_hi =
+		upper_32_bits(hdcp_mc_shared);
+	cmd->cmd.cmd_load_ta.cmd_buf_len = shared_size;
+}
+
+static int psp_hdcp_init_shared_buf(struct psp_context *psp)
+{
+	int ret;
+
+	/*
+	 * Allocate 16k memory aligned to 4k from Frame Buffer (local
+	 * physical) for hdcp ta <-> Driver
+	 */
+	ret = amdgpu_bo_create_kernel(psp->adev, PSP_HDCP_SHARED_MEM_SIZE,
+				      PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM,
+				      &psp->hdcp_context.hdcp_shared_bo,
+				      &psp->hdcp_context.hdcp_shared_mc_addr,
+				      &psp->hdcp_context.hdcp_shared_buf);
+
+	return ret;
+}
+
+static int psp_hdcp_load(struct psp_context *psp)
+{
+	int ret;
+	struct psp_gfx_cmd_resp *cmd;
+
+	/*
+	 * TODO: bypass the loading in sriov for now
+	 */
+	if (amdgpu_sriov_vf(psp->adev))
+		return 0;
+
+	cmd = kzalloc(sizeof(struct psp_gfx_cmd_resp), GFP_KERNEL);
+	if (!cmd)
+		return -ENOMEM;
+
+	memset(psp->fw_pri_buf, 0, PSP_1_MEG);
+	memcpy(psp->fw_pri_buf, psp->ta_hdcp_start_addr,
+	       psp->ta_hdcp_ucode_size);
+
+	psp_prep_hdcp_ta_load_cmd_buf(cmd, psp->fw_pri_mc_addr,
+				      psp->hdcp_context.hdcp_shared_mc_addr,
+				      psp->ta_hdcp_ucode_size,
+				      PSP_HDCP_SHARED_MEM_SIZE);
+
+	ret = psp_cmd_submit_buf(psp, NULL, cmd, psp->fence_buf_mc_addr);
+
+	if (!ret) {
+		psp->hdcp_context.hdcp_initialized = 1;
+		psp->hdcp_context.session_id = cmd->resp.session_id;
+	}
+
+	kfree(cmd);
+
+	return ret;
+}
+static int psp_hdcp_initialize(struct psp_context *psp)
+{
+	int ret;
+
+	/*
+	 * TODO: bypass the initialize in sriov for now
+	 */
+	if (amdgpu_sriov_vf(psp->adev))
+		return 0;
+
+	if (!psp->adev->psp.ta_hdcp_ucode_size ||
+	    !psp->adev->psp.ta_hdcp_start_addr) {
+		dev_warn(psp->adev->dev, "HDCP: hdcp ta ucode is not available\n");
+		return 0;
+	}
+
+	if (!psp->hdcp_context.hdcp_initialized) {
+		ret = psp_hdcp_init_shared_buf(psp);
+		if (ret)
+			return ret;
+	}
+
+	ret = psp_hdcp_load(psp);
+	if (ret)
+		return ret;
+
+	return 0;
+}
+static void psp_prep_hdcp_ta_unload_cmd_buf(struct psp_gfx_cmd_resp *cmd,
+					    uint32_t hdcp_session_id)
+{
+	cmd->cmd_id = GFX_CMD_ID_UNLOAD_TA;
+	cmd->cmd.cmd_unload_ta.session_id = hdcp_session_id;
+}
+
+static int psp_hdcp_unload(struct psp_context *psp)
+{
+	int ret;
+	struct psp_gfx_cmd_resp *cmd;
+
+	/*
+	 * TODO: bypass the unloading in sriov for now
+	 */
+	if (amdgpu_sriov_vf(psp->adev))
+		return 0;
+
+	cmd = kzalloc(sizeof(struct psp_gfx_cmd_resp), GFP_KERNEL);
+	if (!cmd)
+		return -ENOMEM;
+
+	psp_prep_hdcp_ta_unload_cmd_buf(cmd, psp->hdcp_context.session_id);
+
+	ret = psp_cmd_submit_buf(psp, NULL, cmd, psp->fence_buf_mc_addr);
+
+	kfree(cmd);
+
+	return ret;
+}
+
+static void psp_prep_hdcp_ta_invoke_cmd_buf(struct psp_gfx_cmd_resp *cmd,
+					    uint32_t ta_cmd_id,
+					    uint32_t hdcp_session_id)
+{
+	cmd->cmd_id = GFX_CMD_ID_INVOKE_CMD;
+	cmd->cmd.cmd_invoke_cmd.session_id = hdcp_session_id;
+	cmd->cmd.cmd_invoke_cmd.ta_cmd_id = ta_cmd_id;
+	/* Note: cmd_invoke_cmd.buf is not used for now */
+}
+
+int psp_hdcp_invoke(struct psp_context *psp, uint32_t ta_cmd_id)
+{
+	int ret;
+	struct psp_gfx_cmd_resp *cmd;
+
+	/*
+	 * TODO: bypass the loading in sriov for now
+	 */
+	if (amdgpu_sriov_vf(psp->adev))
+		return 0;
+
+	cmd = kzalloc(sizeof(struct psp_gfx_cmd_resp), GFP_KERNEL);
+	if (!cmd)
+		return -ENOMEM;
+
+	psp_prep_hdcp_ta_invoke_cmd_buf(cmd, ta_cmd_id,
+					psp->hdcp_context.session_id);
+
+	ret = psp_cmd_submit_buf(psp, NULL, cmd, psp->fence_buf_mc_addr);
+
+	kfree(cmd);
+
+	return ret;
+}
+
+static int psp_hdcp_terminate(struct psp_context *psp)
+{
+	int ret;
+
+	/*
+	 * TODO: bypass the terminate in sriov for now
+	 */
+	if (amdgpu_sriov_vf(psp->adev))
+		return 0;
+
+	if (!psp->hdcp_context.hdcp_initialized)
+		return 0;
+
+	ret = psp_hdcp_unload(psp);
+	if (ret)
+		return ret;
+
+	psp->hdcp_context.hdcp_initialized = 0;
+
+	/* free hdcp shared memory */
+	amdgpu_bo_free_kernel(&psp->hdcp_context.hdcp_shared_bo,
+			      &psp->hdcp_context.hdcp_shared_mc_addr,
+			      &psp->hdcp_context.hdcp_shared_buf);
+
+	return 0;
+}
+// HDCP end
+
+// DTM start
+static void psp_prep_dtm_ta_load_cmd_buf(struct psp_gfx_cmd_resp *cmd,
+					 uint64_t dtm_ta_mc,
+					 uint64_t dtm_mc_shared,
+					 uint32_t dtm_ta_size,
+					 uint32_t shared_size)
+{
+	cmd->cmd_id = GFX_CMD_ID_LOAD_TA;
+	cmd->cmd.cmd_load_ta.app_phy_addr_lo = lower_32_bits(dtm_ta_mc);
+	cmd->cmd.cmd_load_ta.app_phy_addr_hi = upper_32_bits(dtm_ta_mc);
+	cmd->cmd.cmd_load_ta.app_len = dtm_ta_size;
+
+	cmd->cmd.cmd_load_ta.cmd_buf_phy_addr_lo = lower_32_bits(dtm_mc_shared);
+	cmd->cmd.cmd_load_ta.cmd_buf_phy_addr_hi = upper_32_bits(dtm_mc_shared);
+	cmd->cmd.cmd_load_ta.cmd_buf_len = shared_size;
+}
+
+static int psp_dtm_init_shared_buf(struct psp_context *psp)
+{
+	int ret;
+
+	/*
+	 * Allocate 16k memory aligned to 4k from Frame Buffer (local
+	 * physical) for dtm ta <-> Driver
+	 */
+	ret = amdgpu_bo_create_kernel(psp->adev, PSP_DTM_SHARED_MEM_SIZE,
+				      PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM,
+				      &psp->dtm_context.dtm_shared_bo,
+				      &psp->dtm_context.dtm_shared_mc_addr,
+				      &psp->dtm_context.dtm_shared_buf);
+
+	return ret;
+}
+
+static int psp_dtm_load(struct psp_context *psp)
+{
+	int ret;
+	struct psp_gfx_cmd_resp *cmd;
+
+	/*
+	 * TODO: bypass the loading in sriov for now
+	 */
+	if (amdgpu_sriov_vf(psp->adev))
+		return 0;
+
+	cmd = kzalloc(sizeof(struct psp_gfx_cmd_resp), GFP_KERNEL);
+	if (!cmd)
+		return -ENOMEM;
+
+	memset(psp->fw_pri_buf, 0, PSP_1_MEG);
+	memcpy(psp->fw_pri_buf, psp->ta_dtm_start_addr, psp->ta_dtm_ucode_size);
+
+	psp_prep_dtm_ta_load_cmd_buf(cmd, psp->fw_pri_mc_addr,
+				     psp->dtm_context.dtm_shared_mc_addr,
+				     psp->ta_dtm_ucode_size,
+				     PSP_DTM_SHARED_MEM_SIZE);
+
+	ret = psp_cmd_submit_buf(psp, NULL, cmd, psp->fence_buf_mc_addr);
+
+	if (!ret) {
+		psp->dtm_context.dtm_initialized = 1;
+		psp->dtm_context.session_id = cmd->resp.session_id;
+	}
+
+	kfree(cmd);
+
+	return ret;
+}
+
+static int psp_dtm_initialize(struct psp_context *psp)
+{
+	int ret;
+
+	/*
+	 * TODO: bypass the initialize in sriov for now
+	 */
+	if (amdgpu_sriov_vf(psp->adev))
+		return 0;
+
+	if (!psp->adev->psp.ta_dtm_ucode_size ||
+	    !psp->adev->psp.ta_dtm_start_addr) {
+		dev_warn(psp->adev->dev, "DTM: dtm ta ucode is not available\n");
+		return 0;
+	}
+
+	if (!psp->dtm_context.dtm_initialized) {
+		ret = psp_dtm_init_shared_buf(psp);
+		if (ret)
+			return ret;
+	}
+
+	ret = psp_dtm_load(psp);
+	if (ret)
+		return ret;
+
+	return 0;
+}
+
+static void psp_prep_dtm_ta_invoke_cmd_buf(struct psp_gfx_cmd_resp *cmd,
+					   uint32_t ta_cmd_id,
+					   uint32_t dtm_session_id)
+{
+	cmd->cmd_id = GFX_CMD_ID_INVOKE_CMD;
+	cmd->cmd.cmd_invoke_cmd.session_id = dtm_session_id;
+	cmd->cmd.cmd_invoke_cmd.ta_cmd_id = ta_cmd_id;
+	/* Note: cmd_invoke_cmd.buf is not used for now */
+}
+
+int psp_dtm_invoke(struct psp_context *psp, uint32_t ta_cmd_id)
+{
+	int ret;
+	struct psp_gfx_cmd_resp *cmd;
+
+	/*
+	 * TODO: bypass the loading in sriov for now
+	 */
+	if (amdgpu_sriov_vf(psp->adev))
+		return 0;
+
+	cmd = kzalloc(sizeof(struct psp_gfx_cmd_resp), GFP_KERNEL);
+	if (!cmd)
+		return -ENOMEM;
+
+	psp_prep_dtm_ta_invoke_cmd_buf(cmd, ta_cmd_id,
+				       psp->dtm_context.session_id);
+
+	ret = psp_cmd_submit_buf(psp, NULL, cmd, psp->fence_buf_mc_addr);
+
+	kfree(cmd);
+
+	return ret;
+}
+
+static int psp_dtm_terminate(struct psp_context *psp)
+{
+	int ret;
+
+	/*
+	 * TODO: bypass the terminate in sriov for now
+	 */
+	if (amdgpu_sriov_vf(psp->adev))
+		return 0;
+
+	if (!psp->dtm_context.dtm_initialized)
+		return 0;
+
+	ret = psp_hdcp_unload(psp);
+	if (ret)
+		return ret;
+
+	psp->dtm_context.dtm_initialized = 0;
+
+	/* free hdcp shared memory */
+	amdgpu_bo_free_kernel(&psp->dtm_context.dtm_shared_bo,
+			      &psp->dtm_context.dtm_shared_mc_addr,
+			      &psp->dtm_context.dtm_shared_buf);
+
+	return 0;
+}
+// DTM end
+
 static int psp_hw_start(struct psp_context *psp)
 {
 	struct amdgpu_device *adev = psp->adev;
@@ -845,6 +1238,16 @@ static int psp_hw_start(struct psp_context *psp)
 		if (ret)
 			dev_err(psp->adev->dev,
 					"RAS: Failed to initialize RAS\n");
+
+		ret = psp_hdcp_initialize(psp);
+		if (ret)
+			dev_err(psp->adev->dev,
+				"HDCP: Failed to initialize HDCP\n");
+
+		ret = psp_dtm_initialize(psp);
+		if (ret)
+			dev_err(psp->adev->dev,
+				"DTM: Failed to initialize DTM\n");
 	}
 
 	return 0;
@@ -950,21 +1353,7 @@ static void psp_print_fw_hdr(struct psp_context *psp,
 			     struct amdgpu_firmware_info *ucode)
 {
 	struct amdgpu_device *adev = psp->adev;
-	const struct sdma_firmware_header_v1_0 *sdma_hdr =
-		(const struct sdma_firmware_header_v1_0 *)
-		adev->sdma.instance[ucode->ucode_id - AMDGPU_UCODE_ID_SDMA0].fw->data;
-	const struct gfx_firmware_header_v1_0 *ce_hdr =
-		(const struct gfx_firmware_header_v1_0 *)adev->gfx.ce_fw->data;
-	const struct gfx_firmware_header_v1_0 *pfp_hdr =
-		(const struct gfx_firmware_header_v1_0 *)adev->gfx.pfp_fw->data;
-	const struct gfx_firmware_header_v1_0 *me_hdr =
-		(const struct gfx_firmware_header_v1_0 *)adev->gfx.me_fw->data;
-	const struct gfx_firmware_header_v1_0 *mec_hdr =
-		(const struct gfx_firmware_header_v1_0 *)adev->gfx.mec_fw->data;
-	const struct rlc_firmware_header_v2_0 *rlc_hdr =
-		(const struct rlc_firmware_header_v2_0 *)adev->gfx.rlc_fw->data;
-	const struct smc_firmware_header_v1_0 *smc_hdr =
-		(const struct smc_firmware_header_v1_0 *)adev->pm.fw->data;
+	struct common_firmware_header *hdr;
 
 	switch (ucode->ucode_id) {
 	case AMDGPU_UCODE_ID_SDMA0:
@@ -975,25 +1364,33 @@ static void psp_print_fw_hdr(struct psp_context *psp,
 	case AMDGPU_UCODE_ID_SDMA5:
 	case AMDGPU_UCODE_ID_SDMA6:
 	case AMDGPU_UCODE_ID_SDMA7:
-		amdgpu_ucode_print_sdma_hdr(&sdma_hdr->header);
+		hdr = (struct common_firmware_header *)
+			adev->sdma.instance[ucode->ucode_id - AMDGPU_UCODE_ID_SDMA0].fw->data;
+		amdgpu_ucode_print_sdma_hdr(hdr);
 		break;
 	case AMDGPU_UCODE_ID_CP_CE:
-		amdgpu_ucode_print_gfx_hdr(&ce_hdr->header);
+		hdr = (struct common_firmware_header *)adev->gfx.ce_fw->data;
+		amdgpu_ucode_print_gfx_hdr(hdr);
 		break;
 	case AMDGPU_UCODE_ID_CP_PFP:
-		amdgpu_ucode_print_gfx_hdr(&pfp_hdr->header);
+		hdr = (struct common_firmware_header *)adev->gfx.pfp_fw->data;
+		amdgpu_ucode_print_gfx_hdr(hdr);
 		break;
 	case AMDGPU_UCODE_ID_CP_ME:
-		amdgpu_ucode_print_gfx_hdr(&me_hdr->header);
+		hdr = (struct common_firmware_header *)adev->gfx.me_fw->data;
+		amdgpu_ucode_print_gfx_hdr(hdr);
 		break;
 	case AMDGPU_UCODE_ID_CP_MEC1:
-		amdgpu_ucode_print_gfx_hdr(&mec_hdr->header);
+		hdr = (struct common_firmware_header *)adev->gfx.mec_fw->data;
+		amdgpu_ucode_print_gfx_hdr(hdr);
 		break;
 	case AMDGPU_UCODE_ID_RLC_G:
-		amdgpu_ucode_print_rlc_hdr(&rlc_hdr->header);
+		hdr = (struct common_firmware_header *)adev->gfx.rlc_fw->data;
+		amdgpu_ucode_print_rlc_hdr(hdr);
 		break;
 	case AMDGPU_UCODE_ID_SMC:
-		amdgpu_ucode_print_smc_hdr(&smc_hdr->header);
+		hdr = (struct common_firmware_header *)adev->pm.fw->data;
+		amdgpu_ucode_print_smc_hdr(hdr);
 		break;
 	default:
 		break;
@@ -1070,7 +1467,10 @@ out:
 		    || ucode->ucode_id == AMDGPU_UCODE_ID_SDMA5
 		    || ucode->ucode_id == AMDGPU_UCODE_ID_SDMA6
 		    || ucode->ucode_id == AMDGPU_UCODE_ID_SDMA7
-		    || ucode->ucode_id == AMDGPU_UCODE_ID_RLC_G))
+                    || ucode->ucode_id == AMDGPU_UCODE_ID_RLC_G
+	            || ucode->ucode_id == AMDGPU_UCODE_ID_RLC_RESTORE_LIST_CNTL
+	            || ucode->ucode_id == AMDGPU_UCODE_ID_RLC_RESTORE_LIST_GPM_MEM
+	            || ucode->ucode_id == AMDGPU_UCODE_ID_RLC_RESTORE_LIST_SRM_MEM))
 			/*skip ucode loading in SRIOV VF */
 			continue;
 
@@ -1079,10 +1479,6 @@ out:
 		     ucode->ucode_id == AMDGPU_UCODE_ID_CP_MEC2_JT))
 			/* skip mec JT when autoload is enabled */
 			continue;
-		/* Renoir only needs to load mec jump table one time */
-		if (adev->asic_type == CHIP_RENOIR &&
-		    ucode->ucode_id == AMDGPU_UCODE_ID_CP_MEC2_JT)
-			continue;
 
 		psp_print_fw_hdr(psp, ucode);
 
@@ -1091,7 +1487,8 @@ out:
 			return ret;
 
 		/* Start rlc autoload after psp recieved all the gfx firmware */
-		if (ucode->ucode_id == AMDGPU_UCODE_ID_RLC_RESTORE_LIST_SRM_MEM) {
+		if (psp->autoload_supported && ucode->ucode_id ==
+			AMDGPU_UCODE_ID_RLC_RESTORE_LIST_SRM_MEM) {
 			ret = psp_rlc_autoload(psp);
 			if (ret) {
 				DRM_ERROR("Failed to start rlc autoload\n");
@@ -1216,8 +1613,11 @@ static int psp_hw_fini(void *handle)
 	    psp->xgmi_context.initialized == 1)
                 psp_xgmi_terminate(psp);
 
-	if (psp->adev->psp.ta_fw)
+	if (psp->adev->psp.ta_fw) {
 		psp_ras_terminate(psp);
+		psp_dtm_terminate(psp);
+		psp_hdcp_terminate(psp);
+	}
 
 	psp_ring_destroy(psp, PSP_RING_TYPE__KM);
 
@@ -1259,6 +1659,16 @@ static int psp_suspend(void *handle)
 			DRM_ERROR("Failed to terminate ras ta\n");
 			return ret;
 		}
+		ret = psp_hdcp_terminate(psp);
+		if (ret) {
+			DRM_ERROR("Failed to terminate hdcp ta\n");
+			return ret;
+		}
+		ret = psp_dtm_terminate(psp);
+		if (ret) {
+			DRM_ERROR("Failed to terminate dtm ta\n");
+			return ret;
+		}
 	}
 
 	ret = psp_ring_stop(psp, PSP_RING_TYPE__KM);
@@ -1278,6 +1688,12 @@ static int psp_resume(void *handle)
 
 	DRM_INFO("PSP is resuming...\n");
 
+	ret = psp_mem_training(psp, PSP_MEM_TRAIN_RESUME);
+	if (ret) {
+		DRM_ERROR("Failed to process memory training!\n");
+		return ret;
+	}
+
 	mutex_lock(&adev->firmware.mutex);
 
 	ret = psp_hw_start(psp);
@@ -1317,9 +1733,6 @@ int psp_rlc_autoload_start(struct psp_context *psp)
 	int ret;
 	struct psp_gfx_cmd_resp *cmd;
 
-	if (amdgpu_sriov_vf(psp->adev))
-		return 0;
-
 	cmd = kzalloc(sizeof(struct psp_gfx_cmd_resp), GFP_KERNEL);
 	if (!cmd)
 		return -ENOMEM;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h
index bc0947f6bc8a..09c5474ebcc3 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h
@@ -37,6 +37,9 @@
 #define PSP_RAS_SHARED_MEM_SIZE 0x4000
 #define PSP_1_MEG		0x100000
 #define PSP_TMR_SIZE	0x400000
+#define PSP_HDCP_SHARED_MEM_SIZE	0x4000
+#define PSP_DTM_SHARED_MEM_SIZE	0x4000
+#define PSP_SHARED_MEM_SIZE		0x4000
 
 struct psp_context;
 struct psp_xgmi_node_info;
@@ -46,6 +49,8 @@ enum psp_bootloader_cmd {
 	PSP_BL__LOAD_SYSDRV		= 0x10000,
 	PSP_BL__LOAD_SOSDRV		= 0x20000,
 	PSP_BL__LOAD_KEY_DATABASE	= 0x80000,
+	PSP_BL__DRAM_LONG_TRAIN		= 0x100000,
+	PSP_BL__DRAM_SHORT_TRAIN	= 0x200000,
 };
 
 enum psp_ring_type
@@ -108,6 +113,9 @@ struct psp_funcs
 			struct ta_ras_trigger_error_input *info);
 	int (*ras_cure_posion)(struct psp_context *psp, uint64_t *mode_ptr);
 	int (*rlc_autoload_start)(struct psp_context *psp);
+	int (*mem_training_init)(struct psp_context *psp);
+	void (*mem_training_fini)(struct psp_context *psp);
+	int (*mem_training)(struct psp_context *psp, uint32_t ops);
 };
 
 #define AMDGPU_XGMI_MAX_CONNECTED_NODES		64
@@ -142,6 +150,65 @@ struct psp_ras_context {
 	struct amdgpu_ras	*ras;
 };
 
+struct psp_hdcp_context {
+	bool			hdcp_initialized;
+	uint32_t		session_id;
+	struct amdgpu_bo	*hdcp_shared_bo;
+	uint64_t		hdcp_shared_mc_addr;
+	void			*hdcp_shared_buf;
+};
+
+struct psp_dtm_context {
+	bool			dtm_initialized;
+	uint32_t		session_id;
+	struct amdgpu_bo	*dtm_shared_bo;
+	uint64_t		dtm_shared_mc_addr;
+	void			*dtm_shared_buf;
+};
+
+#define MEM_TRAIN_SYSTEM_SIGNATURE		0x54534942
+#define GDDR6_MEM_TRAINING_DATA_SIZE_IN_BYTES	0x1000
+#define GDDR6_MEM_TRAINING_OFFSET		0x8000
+
+enum psp_memory_training_init_flag {
+	PSP_MEM_TRAIN_NOT_SUPPORT	= 0x0,
+	PSP_MEM_TRAIN_SUPPORT		= 0x1,
+	PSP_MEM_TRAIN_INIT_FAILED	= 0x2,
+	PSP_MEM_TRAIN_RESERVE_SUCCESS	= 0x4,
+	PSP_MEM_TRAIN_INIT_SUCCESS	= 0x8,
+};
+
+enum psp_memory_training_ops {
+	PSP_MEM_TRAIN_SEND_LONG_MSG	= 0x1,
+	PSP_MEM_TRAIN_SAVE		= 0x2,
+	PSP_MEM_TRAIN_RESTORE		= 0x4,
+	PSP_MEM_TRAIN_SEND_SHORT_MSG	= 0x8,
+	PSP_MEM_TRAIN_COLD_BOOT		= PSP_MEM_TRAIN_SEND_LONG_MSG,
+	PSP_MEM_TRAIN_RESUME		= PSP_MEM_TRAIN_SEND_SHORT_MSG,
+};
+
+struct psp_memory_training_context {
+	/*training data size*/
+	u64 train_data_size;
+	/*
+	 * sys_cache
+	 * cpu virtual address
+	 * system memory buffer that used to store the training data.
+	 */
+	void *sys_cache;
+
+	/*vram offset of the p2c training data*/
+	u64 p2c_train_data_offset;
+	struct amdgpu_bo *p2c_bo;
+
+	/*vram offset of the c2p training data*/
+	u64 c2p_train_data_offset;
+	struct amdgpu_bo *c2p_bo;
+
+	enum psp_memory_training_init_flag init;
+	u32 training_cnt;
+};
+
 struct psp_context
 {
 	struct amdgpu_device            *adev;
@@ -206,9 +273,21 @@ struct psp_context
 	uint32_t			ta_ras_ucode_version;
 	uint32_t			ta_ras_ucode_size;
 	uint8_t				*ta_ras_start_addr;
+
+	uint32_t			ta_hdcp_ucode_version;
+	uint32_t			ta_hdcp_ucode_size;
+	uint8_t				*ta_hdcp_start_addr;
+
+	uint32_t			ta_dtm_ucode_version;
+	uint32_t			ta_dtm_ucode_size;
+	uint8_t				*ta_dtm_start_addr;
+
 	struct psp_xgmi_context		xgmi_context;
 	struct psp_ras_context		ras;
+	struct psp_hdcp_context 	hdcp_context;
+	struct psp_dtm_context		dtm_context;
 	struct mutex			mutex;
+	struct psp_memory_training_context mem_train_ctx;
 };
 
 struct amdgpu_psp_funcs {
@@ -251,6 +330,12 @@ struct amdgpu_psp_funcs {
 		(psp)->funcs->xgmi_set_topology_info((psp), (num_device), (topology)) : -EINVAL)
 #define psp_rlc_autoload(psp) \
 		((psp)->funcs->rlc_autoload_start ? (psp)->funcs->rlc_autoload_start((psp)) : 0)
+#define psp_mem_training_init(psp) \
+	((psp)->funcs->mem_training_init ? (psp)->funcs->mem_training_init((psp)) : 0)
+#define psp_mem_training_fini(psp) \
+	((psp)->funcs->mem_training_fini ? (psp)->funcs->mem_training_fini((psp)) : 0)
+#define psp_mem_training(psp, ops) \
+	((psp)->funcs->mem_training ? (psp)->funcs->mem_training((psp), (ops)) : 0)
 
 #define amdgpu_psp_check_fw_loading_status(adev, i) (adev)->firmware.funcs->check_fw_loading_status((adev), (i))
 
@@ -279,6 +364,8 @@ int psp_xgmi_invoke(struct psp_context *psp, uint32_t ta_cmd_id);
 int psp_ras_invoke(struct psp_context *psp, uint32_t ta_cmd_id);
 int psp_ras_enable_features(struct psp_context *psp,
 		union ta_ras_cmd_input *info, bool enable);
+int psp_hdcp_invoke(struct psp_context *psp, uint32_t ta_cmd_id);
+int psp_dtm_invoke(struct psp_context *psp, uint32_t ta_cmd_id);
 
 int psp_rlc_autoload_start(struct psp_context *psp);
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
index 016ea274b955..404483437bd3 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
@@ -25,10 +25,13 @@
 #include <linux/list.h>
 #include <linux/module.h>
 #include <linux/uaccess.h>
+#include <linux/reboot.h>
+#include <linux/syscalls.h>
 
 #include "amdgpu.h"
 #include "amdgpu_ras.h"
 #include "amdgpu_atomfirmware.h"
+#include "ivsrcid/nbio/irqsrcs_nbif_7_4.h"
 
 const char *ras_error_string[] = {
 	"none",
@@ -65,11 +68,16 @@ const char *ras_block_string[] = {
 /* inject address is 52 bits */
 #define	RAS_UMC_INJECT_ADDR_LIMIT	(0x1ULL << 52)
 
-static int amdgpu_ras_reserve_vram(struct amdgpu_device *adev,
-		uint64_t offset, uint64_t size,
-		struct amdgpu_bo **bo_ptr);
-static int amdgpu_ras_release_vram(struct amdgpu_device *adev,
-		struct amdgpu_bo **bo_ptr);
+enum amdgpu_ras_retire_page_reservation {
+	AMDGPU_RAS_RETIRE_PAGE_RESERVED,
+	AMDGPU_RAS_RETIRE_PAGE_PENDING,
+	AMDGPU_RAS_RETIRE_PAGE_FAULT,
+};
+
+atomic_t amdgpu_ras_in_intr = ATOMIC_INIT(0);
+
+static bool amdgpu_ras_check_bad_page(struct amdgpu_device *adev,
+				uint64_t addr);
 
 static ssize_t amdgpu_ras_debugfs_read(struct file *f, char __user *buf,
 					size_t size, loff_t *pos)
@@ -189,6 +197,10 @@ static int amdgpu_ras_debugfs_ctrl_parse_data(struct file *f,
 
 	return 0;
 }
+
+static struct ras_manager *amdgpu_ras_find_obj(struct amdgpu_device *adev,
+		struct ras_common_if *head);
+
 /**
  * DOC: AMDGPU RAS debugfs control interface
  *
@@ -208,31 +220,44 @@ static int amdgpu_ras_debugfs_ctrl_parse_data(struct file *f,
  * As their names indicate, inject operation will write the
  * value to the address.
  *
- * Second member: struct ras_debug_if::op.
+ * The second member: struct ras_debug_if::op.
  * It has three kinds of operations.
- *  0: disable RAS on the block. Take ::head as its data.
- *  1: enable RAS on the block. Take ::head as its data.
- *  2: inject errors on the block. Take ::inject as its data.
+ *
+ * - 0: disable RAS on the block. Take ::head as its data.
+ * - 1: enable RAS on the block. Take ::head as its data.
+ * - 2: inject errors on the block. Take ::inject as its data.
  *
  * How to use the interface?
- * programs:
- * copy the struct ras_debug_if in your codes and initialize it.
- * write the struct to the control node.
  *
- * bash:
- * echo op block [error [sub_blcok address value]] > .../ras/ras_ctrl
- *	op: disable, enable, inject
- *		disable: only block is needed
- *		enable: block and error are needed
- *		inject: error, address, value are needed
- *	block: umc, smda, gfx, .........
- *		see ras_block_string[] for details
- *	error: ue, ce
- *		ue: multi_uncorrectable
- *		ce: single_correctable
- *	sub_block: sub block index, pass 0 if there is no sub block
+ * Programs
+ *
+ * Copy the struct ras_debug_if in your codes and initialize it.
+ * Write the struct to the control node.
+ *
+ * Shells
+ *
+ * .. code-block:: bash
+ *
+ *	echo op block [error [sub_block address value]] > .../ras/ras_ctrl
+ *
+ * Parameters:
+ *
+ * op: disable, enable, inject
+ *	disable: only block is needed
+ *	enable: block and error are needed
+ *	inject: error, address, value are needed
+ * block: umc, sdma, gfx, .........
+ *	see ras_block_string[] for details
+ * error: ue, ce
+ *	ue: multi_uncorrectable
+ *	ce: single_correctable
+ * sub_block:
+ *	sub block index, pass 0 if there is no sub block
+ *
+ * here are some examples for bash commands:
+ *
+ * .. code-block:: bash
  *
- * here are some examples for bash commands,
  *	echo inject umc ue 0x0 0x0 0x0 > /sys/kernel/debug/dri/0/ras/ras_ctrl
  *	echo inject umc ce 0 0 0 > /sys/kernel/debug/dri/0/ras/ras_ctrl
  *	echo disable umc > /sys/kernel/debug/dri/0/ras/ras_ctrl
@@ -245,8 +270,11 @@ static int amdgpu_ras_debugfs_ctrl_parse_data(struct file *f,
  * For inject, please check corresponding err count at
  * /sys/class/drm/card[0/1/2...]/device/ras/[gfx/sdma/...]_err_count
  *
- * NOTE: operation is only allowed on blocks which are supported.
- * Please check ras mask at /sys/module/amdgpu/parameters/ras_mask
+ * .. note::
+ *	Operations are only allowed on blocks which are supported.
+ *	Please check ras mask at /sys/module/amdgpu/parameters/ras_mask
+ *	to see which blocks support RAS on a particular asic.
+ *
  */
 static ssize_t amdgpu_ras_debugfs_ctrl_write(struct file *f, const char __user *buf,
 		size_t size, loff_t *pos)
@@ -276,6 +304,14 @@ static ssize_t amdgpu_ras_debugfs_ctrl_write(struct file *f, const char __user *
 			break;
 		}
 
+		/* umc ce/ue error injection for a bad page is not allowed */
+		if ((data.head.block == AMDGPU_RAS_BLOCK__UMC) &&
+		    amdgpu_ras_check_bad_page(adev, data.inject.address)) {
+			DRM_WARN("RAS WARN: 0x%llx has been marked as bad before error injection!\n",
+					data.inject.address);
+			break;
+		}
+
 		/* data.inject.address is offset instead of absolute gpu address */
 		ret = amdgpu_ras_error_inject(adev, &data.inject);
 		break;
@@ -290,6 +326,33 @@ static ssize_t amdgpu_ras_debugfs_ctrl_write(struct file *f, const char __user *
 	return size;
 }
 
+/**
+ * DOC: AMDGPU RAS debugfs EEPROM table reset interface
+ *
+ * Some boards contain an EEPROM which is used to persistently store a list of
+ * bad pages which experiences ECC errors in vram.  This interface provides
+ * a way to reset the EEPROM, e.g., after testing error injection.
+ *
+ * Usage:
+ *
+ * .. code-block:: bash
+ *
+ *	echo 1 > ../ras/ras_eeprom_reset
+ *
+ * will reset EEPROM table to 0 entries.
+ *
+ */
+static ssize_t amdgpu_ras_debugfs_eeprom_write(struct file *f, const char __user *buf,
+		size_t size, loff_t *pos)
+{
+	struct amdgpu_device *adev = (struct amdgpu_device *)file_inode(f)->i_private;
+	int ret;
+
+	ret = amdgpu_ras_eeprom_reset_table(&adev->psp.ras.ras->eeprom_control);
+
+	return ret == 1 ? size : -EIO;
+}
+
 static const struct file_operations amdgpu_ras_debugfs_ctrl_ops = {
 	.owner = THIS_MODULE,
 	.read = NULL,
@@ -297,6 +360,34 @@ static const struct file_operations amdgpu_ras_debugfs_ctrl_ops = {
 	.llseek = default_llseek
 };
 
+static const struct file_operations amdgpu_ras_debugfs_eeprom_ops = {
+	.owner = THIS_MODULE,
+	.read = NULL,
+	.write = amdgpu_ras_debugfs_eeprom_write,
+	.llseek = default_llseek
+};
+
+/**
+ * DOC: AMDGPU RAS sysfs Error Count Interface
+ *
+ * It allows the user to read the error count for each IP block on the gpu through
+ * /sys/class/drm/card[0/1/2...]/device/ras/[gfx/sdma/...]_err_count
+ *
+ * It outputs the multiple lines which report the uncorrected (ue) and corrected
+ * (ce) error counts.
+ *
+ * The format of one line is below,
+ *
+ * [ce|ue]: count
+ *
+ * Example:
+ *
+ * .. code-block:: bash
+ *
+ *	ue: 0
+ *	ce: 1
+ *
+ */
 static ssize_t amdgpu_ras_sysfs_read(struct device *dev,
 		struct device_attribute *attr, char *buf)
 {
@@ -475,15 +566,17 @@ int amdgpu_ras_feature_enable(struct amdgpu_device *adev,
 	if (!(!!enable ^ !!amdgpu_ras_is_feature_enabled(adev, head)))
 		return 0;
 
-	ret = psp_ras_enable_features(&adev->psp, &info, enable);
-	if (ret) {
-		DRM_ERROR("RAS ERROR: %s %s feature failed ret %d\n",
-				enable ? "enable":"disable",
-				ras_block_str(head->block),
-				ret);
-		if (ret == TA_RAS_STATUS__RESET_NEEDED)
-			return -EAGAIN;
-		return -EINVAL;
+	if (!amdgpu_ras_intr_triggered()) {
+		ret = psp_ras_enable_features(&adev->psp, &info, enable);
+		if (ret) {
+			DRM_ERROR("RAS ERROR: %s %s feature failed ret %d\n",
+					enable ? "enable":"disable",
+					ras_block_str(head->block),
+					ret);
+			if (ret == TA_RAS_STATUS__RESET_NEEDED)
+				return -EAGAIN;
+			return -EINVAL;
+		}
 	}
 
 	/* setup the obj */
@@ -615,8 +708,12 @@ int amdgpu_ras_error_query(struct amdgpu_device *adev,
 			adev->gfx.funcs->query_ras_error_count(adev, &err_data);
 		break;
 	case AMDGPU_RAS_BLOCK__MMHUB:
-		if (adev->mmhub_funcs->query_ras_error_count)
-			adev->mmhub_funcs->query_ras_error_count(adev, &err_data);
+		if (adev->mmhub.funcs->query_ras_error_count)
+			adev->mmhub.funcs->query_ras_error_count(adev, &err_data);
+		break;
+	case AMDGPU_RAS_BLOCK__PCIE_BIF:
+		if (adev->nbio.funcs->query_ras_error_count)
+			adev->nbio.funcs->query_ras_error_count(adev, &err_data);
 		break;
 	default:
 		break;
@@ -628,12 +725,14 @@ int amdgpu_ras_error_query(struct amdgpu_device *adev,
 	info->ue_count = obj->err_data.ue_count;
 	info->ce_count = obj->err_data.ce_count;
 
-	if (err_data.ce_count)
+	if (err_data.ce_count) {
 		dev_info(adev->dev, "%ld correctable errors detected in %s block\n",
 			 obj->err_data.ce_count, ras_block_str(info->head.block));
-	if (err_data.ue_count)
+	}
+	if (err_data.ue_count) {
 		dev_info(adev->dev, "%ld uncorrectable errors detected in %s block\n",
 			 obj->err_data.ue_count, ras_block_str(info->head.block));
+	}
 
 	return 0;
 }
@@ -664,6 +763,8 @@ int amdgpu_ras_error_inject(struct amdgpu_device *adev,
 		break;
 	case AMDGPU_RAS_BLOCK__UMC:
 	case AMDGPU_RAS_BLOCK__MMHUB:
+	case AMDGPU_RAS_BLOCK__XGMI_WAFL:
+	case AMDGPU_RAS_BLOCK__PCIE_BIF:
 		ret = psp_ras_trigger_error(&adev->psp, &block_info);
 		break;
 	default:
@@ -723,18 +824,18 @@ static int amdgpu_ras_badpages_read(struct amdgpu_device *adev,
 static char *amdgpu_ras_badpage_flags_str(unsigned int flags)
 {
 	switch (flags) {
-	case 0:
+	case AMDGPU_RAS_RETIRE_PAGE_RESERVED:
 		return "R";
-	case 1:
+	case AMDGPU_RAS_RETIRE_PAGE_PENDING:
 		return "P";
-	case 2:
+	case AMDGPU_RAS_RETIRE_PAGE_FAULT:
 	default:
 		return "F";
 	};
 }
 
-/*
- * DOC: ras sysfs gpu_vram_bad_pages interface
+/**
+ * DOC: AMDGPU RAS sysfs gpu_vram_bad_pages Interface
  *
  * It allows user to read the bad pages of vram on the gpu through
  * /sys/class/drm/card[0/1/2...]/device/ras/gpu_vram_bad_pages
@@ -746,14 +847,21 @@ static char *amdgpu_ras_badpage_flags_str(unsigned int flags)
  *
  * gpu pfn and gpu page size are printed in hex format.
  * flags can be one of below character,
+ *
  * R: reserved, this gpu page is reserved and not able to use.
+ *
  * P: pending for reserve, this gpu page is marked as bad, will be reserved
- *    in next window of page_reserve.
+ * in next window of page_reserve.
+ *
  * F: unable to reserve. this gpu page can't be reserved due to some reasons.
  *
- * examples:
- * 0x00000001 : 0x00001000 : R
- * 0x00000002 : 0x00001000 : P
+ * Examples:
+ *
+ * .. code-block:: bash
+ *
+ *	0x00000001 : 0x00001000 : R
+ *	0x00000002 : 0x00001000 : P
+ *
  */
 
 static ssize_t amdgpu_ras_sysfs_badpages_read(struct file *f,
@@ -927,6 +1035,24 @@ static int amdgpu_ras_sysfs_remove_all(struct amdgpu_device *adev)
 }
 /* sysfs end */
 
+/**
+ * DOC: AMDGPU RAS Reboot Behavior for Unrecoverable Errors
+ *
+ * Normally when there is an uncorrectable error, the driver will reset
+ * the GPU to recover.  However, in the event of an unrecoverable error,
+ * the driver provides an interface to reboot the system automatically
+ * in that event.
+ *
+ * The following file in debugfs provides that interface:
+ * /sys/kernel/debug/dri/[0/1/2...]/ras/auto_reboot
+ *
+ * Usage:
+ *
+ * .. code-block:: bash
+ *
+ *	echo true > .../ras/auto_reboot
+ *
+ */
 /* debugfs begin */
 static void amdgpu_ras_debugfs_create_ctrl_node(struct amdgpu_device *adev)
 {
@@ -934,8 +1060,21 @@ static void amdgpu_ras_debugfs_create_ctrl_node(struct amdgpu_device *adev)
 	struct drm_minor *minor = adev->ddev->primary;
 
 	con->dir = debugfs_create_dir("ras", minor->debugfs_root);
-	con->ent = debugfs_create_file("ras_ctrl", S_IWUGO | S_IRUGO, con->dir,
-				       adev, &amdgpu_ras_debugfs_ctrl_ops);
+	debugfs_create_file("ras_ctrl", S_IWUGO | S_IRUGO, con->dir,
+				adev, &amdgpu_ras_debugfs_ctrl_ops);
+	debugfs_create_file("ras_eeprom_reset", S_IWUGO | S_IRUGO, con->dir,
+				adev, &amdgpu_ras_debugfs_eeprom_ops);
+
+	/*
+	 * After one uncorrectable error happens, usually GPU recovery will
+	 * be scheduled. But due to the known problem in GPU recovery failing
+	 * to bring GPU back, below interface provides one direct way to
+	 * user to reboot system automatically in such case within
+	 * ERREVENT_ATHUB_INTERRUPT generated. Normal GPU recovery routine
+	 * will never be called.
+	 */
+	debugfs_create_bool("auto_reboot", S_IWUGO | S_IRUGO, con->dir,
+				&con->reboot);
 }
 
 void amdgpu_ras_debugfs_create(struct amdgpu_device *adev,
@@ -980,10 +1119,8 @@ static void amdgpu_ras_debugfs_remove_all(struct amdgpu_device *adev)
 		amdgpu_ras_debugfs_remove(adev, &obj->head);
 	}
 
-	debugfs_remove(con->ent);
-	debugfs_remove(con->dir);
+	debugfs_remove_recursive(con->dir);
 	con->dir = NULL;
-	con->ent = NULL;
 }
 /* debugfs end */
 
@@ -1188,15 +1325,15 @@ static int amdgpu_ras_badpages_read(struct amdgpu_device *adev,
 
 	for (; i < data->count; i++) {
 		(*bps)[i] = (struct ras_badpage){
-			.bp = data->bps[i].bp,
+			.bp = data->bps[i].retired_page,
 			.size = AMDGPU_GPU_PAGE_SIZE,
-			.flags = 0,
+			.flags = AMDGPU_RAS_RETIRE_PAGE_RESERVED,
 		};
 
 		if (data->last_reserved <= i)
-			(*bps)[i].flags = 1;
-		else if (data->bps[i].bo == NULL)
-			(*bps)[i].flags = 2;
+			(*bps)[i].flags = AMDGPU_RAS_RETIRE_PAGE_PENDING;
+		else if (data->bps_bo[i] == NULL)
+			(*bps)[i].flags = AMDGPU_RAS_RETIRE_PAGE_FAULT;
 	}
 
 	*count = data->count;
@@ -1214,105 +1351,46 @@ static void amdgpu_ras_do_recovery(struct work_struct *work)
 	atomic_set(&ras->in_recovery, 0);
 }
 
-static int amdgpu_ras_release_vram(struct amdgpu_device *adev,
-		struct amdgpu_bo **bo_ptr)
-{
-	/* no need to free it actually. */
-	amdgpu_bo_free_kernel(bo_ptr, NULL, NULL);
-	return 0;
-}
-
-/* reserve vram with size@offset */
-static int amdgpu_ras_reserve_vram(struct amdgpu_device *adev,
-		uint64_t offset, uint64_t size,
-		struct amdgpu_bo **bo_ptr)
-{
-	struct ttm_operation_ctx ctx = { false, false };
-	struct amdgpu_bo_param bp;
-	int r = 0;
-	int i;
-	struct amdgpu_bo *bo;
-
-	if (bo_ptr)
-		*bo_ptr = NULL;
-	memset(&bp, 0, sizeof(bp));
-	bp.size = size;
-	bp.byte_align = PAGE_SIZE;
-	bp.domain = AMDGPU_GEM_DOMAIN_VRAM;
-	bp.flags = AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS |
-		AMDGPU_GEM_CREATE_NO_CPU_ACCESS;
-	bp.type = ttm_bo_type_kernel;
-	bp.resv = NULL;
-
-	r = amdgpu_bo_create(adev, &bp, &bo);
-	if (r)
-		return -EINVAL;
-
-	r = amdgpu_bo_reserve(bo, false);
-	if (r)
-		goto error_reserve;
-
-	offset = ALIGN(offset, PAGE_SIZE);
-	for (i = 0; i < bo->placement.num_placement; ++i) {
-		bo->placements[i].fpfn = offset >> PAGE_SHIFT;
-		bo->placements[i].lpfn = (offset + size) >> PAGE_SHIFT;
-	}
-
-	ttm_bo_mem_put(&bo->tbo, &bo->tbo.mem);
-	r = ttm_bo_mem_space(&bo->tbo, &bo->placement, &bo->tbo.mem, &ctx);
-	if (r)
-		goto error_pin;
-
-	r = amdgpu_bo_pin_restricted(bo,
-			AMDGPU_GEM_DOMAIN_VRAM,
-			offset,
-			offset + size);
-	if (r)
-		goto error_pin;
-
-	if (bo_ptr)
-		*bo_ptr = bo;
-
-	amdgpu_bo_unreserve(bo);
-	return r;
-
-error_pin:
-	amdgpu_bo_unreserve(bo);
-error_reserve:
-	amdgpu_bo_unref(&bo);
-	return r;
-}
-
 /* alloc/realloc bps array */
 static int amdgpu_ras_realloc_eh_data_space(struct amdgpu_device *adev,
 		struct ras_err_handler_data *data, int pages)
 {
 	unsigned int old_space = data->count + data->space_left;
 	unsigned int new_space = old_space + pages;
-	unsigned int align_space = ALIGN(new_space, 1024);
-	void *tmp = kmalloc(align_space * sizeof(*data->bps), GFP_KERNEL);
-
-	if (!tmp)
+	unsigned int align_space = ALIGN(new_space, 512);
+	void *bps = kmalloc(align_space * sizeof(*data->bps), GFP_KERNEL);
+	struct amdgpu_bo **bps_bo =
+			kmalloc(align_space * sizeof(*data->bps_bo), GFP_KERNEL);
+
+	if (!bps || !bps_bo) {
+		kfree(bps);
+		kfree(bps_bo);
 		return -ENOMEM;
+	}
 
 	if (data->bps) {
-		memcpy(tmp, data->bps,
+		memcpy(bps, data->bps,
 				data->count * sizeof(*data->bps));
 		kfree(data->bps);
 	}
+	if (data->bps_bo) {
+		memcpy(bps_bo, data->bps_bo,
+				data->count * sizeof(*data->bps_bo));
+		kfree(data->bps_bo);
+	}
 
-	data->bps = tmp;
+	data->bps = bps;
+	data->bps_bo = bps_bo;
 	data->space_left += align_space - old_space;
 	return 0;
 }
 
 /* it deal with vram only. */
 int amdgpu_ras_add_bad_pages(struct amdgpu_device *adev,
-		unsigned long *bps, int pages)
+		struct eeprom_table_record *bps, int pages)
 {
 	struct amdgpu_ras *con = amdgpu_ras_get_context(adev);
 	struct ras_err_handler_data *data;
-	int i = pages;
 	int ret = 0;
 
 	if (!con || !con->eh_data || !bps || pages <= 0)
@@ -1329,24 +1407,120 @@ int amdgpu_ras_add_bad_pages(struct amdgpu_device *adev,
 			goto out;
 		}
 
-	while (i--)
-		data->bps[data->count++].bp = bps[i];
-
+	memcpy(&data->bps[data->count], bps, pages * sizeof(*data->bps));
+	data->count += pages;
 	data->space_left -= pages;
+
 out:
 	mutex_unlock(&con->recovery_lock);
 
 	return ret;
 }
 
+/*
+ * write error record array to eeprom, the function should be
+ * protected by recovery_lock
+ */
+static int amdgpu_ras_save_bad_pages(struct amdgpu_device *adev)
+{
+	struct amdgpu_ras *con = amdgpu_ras_get_context(adev);
+	struct ras_err_handler_data *data;
+	struct amdgpu_ras_eeprom_control *control;
+	int save_count;
+
+	if (!con || !con->eh_data)
+		return 0;
+
+	control = &con->eeprom_control;
+	data = con->eh_data;
+	save_count = data->count - control->num_recs;
+	/* only new entries are saved */
+	if (save_count > 0)
+		if (amdgpu_ras_eeprom_process_recods(control,
+							&data->bps[control->num_recs],
+							true,
+							save_count)) {
+			DRM_ERROR("Failed to save EEPROM table data!");
+			return -EIO;
+		}
+
+	return 0;
+}
+
+/*
+ * read error record array in eeprom and reserve enough space for
+ * storing new bad pages
+ */
+static int amdgpu_ras_load_bad_pages(struct amdgpu_device *adev)
+{
+	struct amdgpu_ras_eeprom_control *control =
+					&adev->psp.ras.ras->eeprom_control;
+	struct eeprom_table_record *bps = NULL;
+	int ret = 0;
+
+	/* no bad page record, skip eeprom access */
+	if (!control->num_recs)
+		return ret;
+
+	bps = kcalloc(control->num_recs, sizeof(*bps), GFP_KERNEL);
+	if (!bps)
+		return -ENOMEM;
+
+	if (amdgpu_ras_eeprom_process_recods(control, bps, false,
+		control->num_recs)) {
+		DRM_ERROR("Failed to load EEPROM table records!");
+		ret = -EIO;
+		goto out;
+	}
+
+	ret = amdgpu_ras_add_bad_pages(adev, bps, control->num_recs);
+
+out:
+	kfree(bps);
+	return ret;
+}
+
+/*
+ * check if an address belongs to bad page
+ *
+ * Note: this check is only for umc block
+ */
+static bool amdgpu_ras_check_bad_page(struct amdgpu_device *adev,
+				uint64_t addr)
+{
+	struct amdgpu_ras *con = amdgpu_ras_get_context(adev);
+	struct ras_err_handler_data *data;
+	int i;
+	bool ret = false;
+
+	if (!con || !con->eh_data)
+		return ret;
+
+	mutex_lock(&con->recovery_lock);
+	data = con->eh_data;
+	if (!data)
+		goto out;
+
+	addr >>= AMDGPU_GPU_PAGE_SHIFT;
+	for (i = 0; i < data->count; i++)
+		if (addr == data->bps[i].retired_page) {
+			ret = true;
+			goto out;
+		}
+
+out:
+	mutex_unlock(&con->recovery_lock);
+	return ret;
+}
+
 /* called in gpu recovery/init */
 int amdgpu_ras_reserve_bad_pages(struct amdgpu_device *adev)
 {
 	struct amdgpu_ras *con = amdgpu_ras_get_context(adev);
 	struct ras_err_handler_data *data;
 	uint64_t bp;
-	struct amdgpu_bo *bo;
-	int i;
+	struct amdgpu_bo *bo = NULL;
+	int i, ret = 0;
 
 	if (!con || !con->eh_data)
 		return 0;
@@ -1357,18 +1531,29 @@ int amdgpu_ras_reserve_bad_pages(struct amdgpu_device *adev)
 		goto out;
 	/* reserve vram at driver post stage. */
 	for (i = data->last_reserved; i < data->count; i++) {
-		bp = data->bps[i].bp;
+		bp = data->bps[i].retired_page;
 
-		if (amdgpu_ras_reserve_vram(adev, bp << PAGE_SHIFT,
-					PAGE_SIZE, &bo))
-			DRM_ERROR("RAS ERROR: reserve vram %llx fail\n", bp);
+		/* There are two cases of reserve error should be ignored:
+		 * 1) a ras bad page has been allocated (used by someone);
+		 * 2) a ras bad page has been reserved (duplicate error injection
+		 *    for one page);
+		 */
+		if (amdgpu_bo_create_kernel_at(adev, bp << AMDGPU_GPU_PAGE_SHIFT,
+					       AMDGPU_GPU_PAGE_SIZE,
+					       AMDGPU_GEM_DOMAIN_VRAM,
+					       &bo, NULL))
+			DRM_WARN("RAS WARN: reserve vram for retired page %llx fail\n", bp);
 
-		data->bps[i].bo = bo;
+		data->bps_bo[i] = bo;
 		data->last_reserved = i + 1;
+		bo = NULL;
 	}
+
+	/* continue to save bad pages to eeprom even reesrve_vram fails */
+	ret = amdgpu_ras_save_bad_pages(adev);
 out:
 	mutex_unlock(&con->recovery_lock);
-	return 0;
+	return ret;
 }
 
 /* called when driver unload */
@@ -1388,11 +1573,11 @@ static int amdgpu_ras_release_bad_pages(struct amdgpu_device *adev)
 		goto out;
 
 	for (i = data->last_reserved - 1; i >= 0; i--) {
-		bo = data->bps[i].bo;
+		bo = data->bps_bo[i];
 
-		amdgpu_ras_release_vram(adev, &bo);
+		amdgpu_bo_free_kernel(&bo, NULL, NULL);
 
-		data->bps[i].bo = bo;
+		data->bps_bo[i] = bo;
 		data->last_reserved = i;
 	}
 out:
@@ -1400,41 +1585,54 @@ out:
 	return 0;
 }
 
-static int amdgpu_ras_save_bad_pages(struct amdgpu_device *adev)
-{
-	/* TODO
-	 * write the array to eeprom when SMU disabled.
-	 */
-	return 0;
-}
-
-static int amdgpu_ras_load_bad_pages(struct amdgpu_device *adev)
-{
-	/* TODO
-	 * read the array to eeprom when SMU disabled.
-	 */
-	return 0;
-}
-
-static int amdgpu_ras_recovery_init(struct amdgpu_device *adev)
+int amdgpu_ras_recovery_init(struct amdgpu_device *adev)
 {
 	struct amdgpu_ras *con = amdgpu_ras_get_context(adev);
-	struct ras_err_handler_data **data = &con->eh_data;
+	struct ras_err_handler_data **data;
+	int ret;
 
-	*data = kmalloc(sizeof(**data),
-			GFP_KERNEL|__GFP_ZERO);
-	if (!*data)
-		return -ENOMEM;
+	if (con)
+		data = &con->eh_data;
+	else
+		return 0;
+
+	*data = kmalloc(sizeof(**data), GFP_KERNEL | __GFP_ZERO);
+	if (!*data) {
+		ret = -ENOMEM;
+		goto out;
+	}
 
 	mutex_init(&con->recovery_lock);
 	INIT_WORK(&con->recovery_work, amdgpu_ras_do_recovery);
 	atomic_set(&con->in_recovery, 0);
 	con->adev = adev;
 
-	amdgpu_ras_load_bad_pages(adev);
-	amdgpu_ras_reserve_bad_pages(adev);
+	ret = amdgpu_ras_eeprom_init(&con->eeprom_control);
+	if (ret)
+		goto free;
+
+	if (con->eeprom_control.num_recs) {
+		ret = amdgpu_ras_load_bad_pages(adev);
+		if (ret)
+			goto free;
+		ret = amdgpu_ras_reserve_bad_pages(adev);
+		if (ret)
+			goto release;
+	}
 
 	return 0;
+
+release:
+	amdgpu_ras_release_bad_pages(adev);
+free:
+	kfree((*data)->bps);
+	kfree((*data)->bps_bo);
+	kfree(*data);
+	con->eh_data = NULL;
+out:
+	DRM_WARN("Failed to initialize ras recovery!\n");
+
+	return ret;
 }
 
 static int amdgpu_ras_recovery_fini(struct amdgpu_device *adev)
@@ -1442,13 +1640,17 @@ static int amdgpu_ras_recovery_fini(struct amdgpu_device *adev)
 	struct amdgpu_ras *con = amdgpu_ras_get_context(adev);
 	struct ras_err_handler_data *data = con->eh_data;
 
+	/* recovery_init failed to init it, fini is useless */
+	if (!data)
+		return 0;
+
 	cancel_work_sync(&con->recovery_work);
-	amdgpu_ras_save_bad_pages(adev);
 	amdgpu_ras_release_bad_pages(adev);
 
 	mutex_lock(&con->recovery_lock);
 	con->eh_data = NULL;
 	kfree(data->bps);
+	kfree(data->bps_bo);
 	kfree(data);
 	mutex_unlock(&con->recovery_lock);
 
@@ -1500,6 +1702,7 @@ static void amdgpu_ras_check_supported(struct amdgpu_device *adev,
 int amdgpu_ras_init(struct amdgpu_device *adev)
 {
 	struct amdgpu_ras *con = amdgpu_ras_get_context(adev);
+	int r;
 
 	if (con)
 		return 0;
@@ -1527,31 +1730,106 @@ int amdgpu_ras_init(struct amdgpu_device *adev)
 	/* Might need get this flag from vbios. */
 	con->flags = RAS_DEFAULT_FLAGS;
 
-	if (amdgpu_ras_recovery_init(adev))
-		goto recovery_out;
+	if (adev->nbio.funcs->init_ras_controller_interrupt) {
+		r = adev->nbio.funcs->init_ras_controller_interrupt(adev);
+		if (r)
+			return r;
+	}
+
+	if (adev->nbio.funcs->init_ras_err_event_athub_interrupt) {
+		r = adev->nbio.funcs->init_ras_err_event_athub_interrupt(adev);
+		if (r)
+			return r;
+	}
 
 	amdgpu_ras_mask &= AMDGPU_RAS_BLOCK_MASK;
 
 	if (amdgpu_ras_fs_init(adev))
 		goto fs_out;
 
-	/* ras init for each ras block */
-	if (adev->umc.funcs->ras_init)
-		adev->umc.funcs->ras_init(adev);
-
 	DRM_INFO("RAS INFO: ras initialized successfully, "
 			"hardware ability[%x] ras_mask[%x]\n",
 			con->hw_supported, con->supported);
 	return 0;
 fs_out:
-	amdgpu_ras_recovery_fini(adev);
-recovery_out:
 	amdgpu_ras_set_context(adev, NULL);
 	kfree(con);
 
 	return -EINVAL;
 }
 
+/* helper function to handle common stuff in ip late init phase */
+int amdgpu_ras_late_init(struct amdgpu_device *adev,
+			 struct ras_common_if *ras_block,
+			 struct ras_fs_if *fs_info,
+			 struct ras_ih_if *ih_info)
+{
+	int r;
+
+	/* disable RAS feature per IP block if it is not supported */
+	if (!amdgpu_ras_is_supported(adev, ras_block->block)) {
+		amdgpu_ras_feature_enable_on_boot(adev, ras_block, 0);
+		return 0;
+	}
+
+	r = amdgpu_ras_feature_enable_on_boot(adev, ras_block, 1);
+	if (r) {
+		if (r == -EAGAIN) {
+			/* request gpu reset. will run again */
+			amdgpu_ras_request_reset_on_boot(adev,
+					ras_block->block);
+			return 0;
+		} else if (adev->in_suspend || adev->in_gpu_reset) {
+			/* in resume phase, if fail to enable ras,
+			 * clean up all ras fs nodes, and disable ras */
+			goto cleanup;
+		} else
+			return r;
+	}
+
+	/* in resume phase, no need to create ras fs node */
+	if (adev->in_suspend || adev->in_gpu_reset)
+		return 0;
+
+	if (ih_info->cb) {
+		r = amdgpu_ras_interrupt_add_handler(adev, ih_info);
+		if (r)
+			goto interrupt;
+	}
+
+	amdgpu_ras_debugfs_create(adev, fs_info);
+
+	r = amdgpu_ras_sysfs_create(adev, fs_info);
+	if (r)
+		goto sysfs;
+
+	return 0;
+cleanup:
+	amdgpu_ras_sysfs_remove(adev, ras_block);
+sysfs:
+	amdgpu_ras_debugfs_remove(adev, ras_block);
+	if (ih_info->cb)
+		amdgpu_ras_interrupt_remove_handler(adev, ih_info);
+interrupt:
+	amdgpu_ras_feature_enable(adev, ras_block, 0);
+	return r;
+}
+
+/* helper function to remove ras fs node and interrupt handler */
+void amdgpu_ras_late_fini(struct amdgpu_device *adev,
+			  struct ras_common_if *ras_block,
+			  struct ras_ih_if *ih_info)
+{
+	if (!ras_block || !ih_info)
+		return;
+
+	amdgpu_ras_sysfs_remove(adev, ras_block);
+	amdgpu_ras_debugfs_remove(adev, ras_block);
+	if (ih_info->cb)
+                amdgpu_ras_interrupt_remove_handler(adev, ih_info);
+	amdgpu_ras_feature_enable(adev, ras_block, 0);
+}
+
 /* do some init work after IP late init as dependence.
  * and it runs in resume/gpu reset/booting up cases.
  */
@@ -1645,3 +1923,18 @@ int amdgpu_ras_fini(struct amdgpu_device *adev)
 
 	return 0;
 }
+
+void amdgpu_ras_global_ras_isr(struct amdgpu_device *adev)
+{
+	uint32_t hw_supported, supported;
+
+	amdgpu_ras_check_supported(adev, &hw_supported, &supported);
+	if (!hw_supported)
+		return;
+
+	if (atomic_cmpxchg(&amdgpu_ras_in_intr, 0, 1) == 0) {
+		DRM_WARN("RAS event of type ERREVENT_ATHUB_INTERRUPT detected!\n");
+
+		amdgpu_ras_reset_gpu(adev, false);
+	}
+}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h
index 6c76bb2a6843..f80fd3428c98 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h
@@ -317,8 +317,6 @@ struct amdgpu_ras {
 	struct list_head head;
 	/* debugfs */
 	struct dentry *dir;
-	/* debugfs ctrl */
-	struct dentry *ent;
 	/* sysfs */
 	struct device_attribute features_attr;
 	struct bin_attribute badpages_attr;
@@ -334,7 +332,7 @@ struct amdgpu_ras {
 	struct mutex recovery_lock;
 
 	uint32_t flags;
-
+	bool reboot;
 	struct amdgpu_ras_eeprom_control eeprom_control;
 };
 
@@ -347,15 +345,14 @@ struct ras_err_data {
 	unsigned long ue_count;
 	unsigned long ce_count;
 	unsigned long err_addr_cnt;
-	uint64_t *err_addr;
+	struct eeprom_table_record *err_addr;
 };
 
 struct ras_err_handler_data {
-	/* point to bad pages array */
-	struct {
-		unsigned long bp;
-		struct amdgpu_bo *bo;
-	} *bps;
+	/* point to bad page records array */
+	struct eeprom_table_record *bps;
+	/* point to reserved bo array */
+	struct amdgpu_bo **bps_bo;
 	/* the count of entries */
 	int count;
 	/* the space can place new entries */
@@ -365,7 +362,7 @@ struct ras_err_handler_data {
 };
 
 typedef int (*ras_ih_cb)(struct amdgpu_device *adev,
-		struct ras_err_data *err_data,
+		void *err_data,
 		struct amdgpu_iv_entry *entry);
 
 struct ras_ih_data {
@@ -481,6 +478,7 @@ static inline int amdgpu_ras_is_supported(struct amdgpu_device *adev,
 	return ras && (ras->supported & (1 << block));
 }
 
+int amdgpu_ras_recovery_init(struct amdgpu_device *adev);
 int amdgpu_ras_request_reset_on_boot(struct amdgpu_device *adev,
 		unsigned int block);
 
@@ -492,7 +490,7 @@ unsigned long amdgpu_ras_query_error_count(struct amdgpu_device *adev,
 
 /* error handling functions */
 int amdgpu_ras_add_bad_pages(struct amdgpu_device *adev,
-		unsigned long *bps, int pages);
+		struct eeprom_table_record *bps, int pages);
 
 int amdgpu_ras_reserve_bad_pages(struct amdgpu_device *adev);
 
@@ -501,6 +499,12 @@ static inline int amdgpu_ras_reset_gpu(struct amdgpu_device *adev,
 {
 	struct amdgpu_ras *ras = amdgpu_ras_get_context(adev);
 
+	/* save bad page to eeprom before gpu reset,
+	 * i2c may be unstable in gpu reset
+	 */
+	if (in_task())
+		amdgpu_ras_reserve_bad_pages(adev);
+
 	if (atomic_cmpxchg(&ras->in_recovery, 0, 1) == 0)
 		schedule_work(&ras->recovery_work);
 	return 0;
@@ -566,6 +570,13 @@ amdgpu_ras_error_to_ta(enum amdgpu_ras_error_type error) {
 int amdgpu_ras_init(struct amdgpu_device *adev);
 int amdgpu_ras_fini(struct amdgpu_device *adev);
 int amdgpu_ras_pre_fini(struct amdgpu_device *adev);
+int amdgpu_ras_late_init(struct amdgpu_device *adev,
+			 struct ras_common_if *ras_block,
+			 struct ras_fs_if *fs_info,
+			 struct ras_ih_if *ih_info);
+void amdgpu_ras_late_fini(struct amdgpu_device *adev,
+			  struct ras_common_if *ras_block,
+			  struct ras_ih_if *ih_info);
 
 int amdgpu_ras_feature_enable(struct amdgpu_device *adev,
 		struct ras_common_if *head, bool enable);
@@ -599,4 +610,14 @@ int amdgpu_ras_interrupt_remove_handler(struct amdgpu_device *adev,
 
 int amdgpu_ras_interrupt_dispatch(struct amdgpu_device *adev,
 		struct ras_dispatch_if *info);
+
+extern atomic_t amdgpu_ras_in_intr;
+
+static inline bool amdgpu_ras_intr_triggered(void)
+{
+	return !!atomic_read(&amdgpu_ras_in_intr);
+}
+
+void amdgpu_ras_global_ras_isr(struct amdgpu_device *adev);
+
 #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
index 8a32b5c93778..7de16c0c2f20 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
@@ -100,7 +100,101 @@ static int __update_table_header(struct amdgpu_ras_eeprom_control *control,
 	return ret;
 }
 
-static uint32_t  __calc_hdr_byte_sum(struct amdgpu_ras_eeprom_control *control);
+
+
+static uint32_t  __calc_hdr_byte_sum(struct amdgpu_ras_eeprom_control *control)
+{
+	int i;
+	uint32_t tbl_sum = 0;
+
+	/* Header checksum, skip checksum field in the calculation */
+	for (i = 0; i < sizeof(control->tbl_hdr) - sizeof(control->tbl_hdr.checksum); i++)
+		tbl_sum += *(((unsigned char *)&control->tbl_hdr) + i);
+
+	return tbl_sum;
+}
+
+static uint32_t  __calc_recs_byte_sum(struct eeprom_table_record *records,
+				      int num)
+{
+	int i, j;
+	uint32_t tbl_sum = 0;
+
+	/* Records checksum */
+	for (i = 0; i < num; i++) {
+		struct eeprom_table_record *record = &records[i];
+
+		for (j = 0; j < sizeof(*record); j++) {
+			tbl_sum += *(((unsigned char *)record) + j);
+		}
+	}
+
+	return tbl_sum;
+}
+
+static inline uint32_t  __calc_tbl_byte_sum(struct amdgpu_ras_eeprom_control *control,
+				  struct eeprom_table_record *records, int num)
+{
+	return __calc_hdr_byte_sum(control) + __calc_recs_byte_sum(records, num);
+}
+
+/* Checksum = 256 -((sum of all table entries) mod 256) */
+static void __update_tbl_checksum(struct amdgpu_ras_eeprom_control *control,
+				  struct eeprom_table_record *records, int num,
+				  uint32_t old_hdr_byte_sum)
+{
+	/*
+	 * This will update the table sum with new records.
+	 *
+	 * TODO: What happens when the EEPROM table is to be wrapped around
+	 * and old records from start will get overridden.
+	 */
+
+	/* need to recalculate updated header byte sum */
+	control->tbl_byte_sum -= old_hdr_byte_sum;
+	control->tbl_byte_sum += __calc_tbl_byte_sum(control, records, num);
+
+	control->tbl_hdr.checksum = 256 - (control->tbl_byte_sum % 256);
+}
+
+/* table sum mod 256 + checksum must equals 256 */
+static bool __validate_tbl_checksum(struct amdgpu_ras_eeprom_control *control,
+			    struct eeprom_table_record *records, int num)
+{
+	control->tbl_byte_sum = __calc_tbl_byte_sum(control, records, num);
+
+	if (control->tbl_hdr.checksum + (control->tbl_byte_sum % 256) != 256) {
+		DRM_WARN("Checksum mismatch, checksum: %u ", control->tbl_hdr.checksum);
+		return false;
+	}
+
+	return true;
+}
+
+int amdgpu_ras_eeprom_reset_table(struct amdgpu_ras_eeprom_control *control)
+{
+	unsigned char buff[EEPROM_ADDRESS_SIZE + EEPROM_TABLE_HEADER_SIZE] = { 0 };
+	struct amdgpu_ras_eeprom_table_header *hdr = &control->tbl_hdr;
+	int ret = 0;
+
+	mutex_lock(&control->tbl_mutex);
+
+	hdr->header = EEPROM_TABLE_HDR_VAL;
+	hdr->version = EEPROM_TABLE_VER;
+	hdr->first_rec_offset = EEPROM_RECORD_START;
+	hdr->tbl_size = EEPROM_TABLE_HEADER_SIZE;
+
+	control->tbl_byte_sum = 0;
+	__update_tbl_checksum(control, NULL, 0, 0);
+	control->next_addr = EEPROM_RECORD_START;
+
+	ret = __update_table_header(control, buff);
+
+	mutex_unlock(&control->tbl_mutex);
+
+	return ret;
+
+}
 
 int amdgpu_ras_eeprom_init(struct amdgpu_ras_eeprom_control *control)
 {
@@ -122,6 +216,10 @@ int amdgpu_ras_eeprom_init(struct amdgpu_ras_eeprom_control *control)
 		ret = smu_v11_0_i2c_eeprom_control_init(&control->eeprom_accessor);
 		break;
 
+	case CHIP_ARCTURUS:
+		ret = smu_i2c_eeprom_init(&adev->smu, &control->eeprom_accessor);
+		break;
+
 	default:
 		return 0;
 	}
@@ -143,25 +241,18 @@ int amdgpu_ras_eeprom_init(struct amdgpu_ras_eeprom_control *control)
 	if (hdr->header == EEPROM_TABLE_HDR_VAL) {
 		control->num_recs = (hdr->tbl_size - EEPROM_TABLE_HEADER_SIZE) /
 				    EEPROM_TABLE_RECORD_SIZE;
+		control->tbl_byte_sum = __calc_hdr_byte_sum(control);
+		control->next_addr = EEPROM_RECORD_START;
+
 		DRM_DEBUG_DRIVER("Found existing EEPROM table with %d records",
 				 control->num_recs);
 
 	} else {
 		DRM_INFO("Creating new EEPROM table");
 
-		hdr->header = EEPROM_TABLE_HDR_VAL;
-		hdr->version = EEPROM_TABLE_VER;
-		hdr->first_rec_offset = EEPROM_RECORD_START;
-		hdr->tbl_size = EEPROM_TABLE_HEADER_SIZE;
-
-		adev->psp.ras.ras->eeprom_control.tbl_byte_sum =
-				__calc_hdr_byte_sum(&adev->psp.ras.ras->eeprom_control);
-		ret = __update_table_header(control, buff);
+		ret = amdgpu_ras_eeprom_reset_table(control);
 	}
 
-	/* Start inserting records from here */
-	adev->psp.ras.ras->eeprom_control.next_addr = EEPROM_RECORD_START;
-
 	return ret == 1 ? 0 : -EIO;
 }
 
@@ -173,6 +264,9 @@ void amdgpu_ras_eeprom_fini(struct amdgpu_ras_eeprom_control *control)
 	case CHIP_VEGA20:
 		smu_v11_0_i2c_eeprom_control_fini(&control->eeprom_accessor);
 		break;
+	case CHIP_ARCTURUS:
+		smu_i2c_eeprom_fini(&adev->smu, &control->eeprom_accessor);
+		break;
 
 	default:
 		return;
@@ -226,8 +320,8 @@ static void __decode_table_record_from_buff(struct amdgpu_ras_eeprom_control *co
 	record->offset = (le64_to_cpu(tmp) & 0xffffffffffff);
 	i += 6;
 
-	buff[i++] = record->mem_channel;
-	buff[i++] = record->mcumc_id;
+	record->mem_channel = buff[i++];
+	record->mcumc_id = buff[i++];
 
 	memcpy(&tmp, buff + i,  6);
 	record->retired_page = (le64_to_cpu(tmp) & 0xffffffffffff);
@@ -266,87 +360,18 @@ static uint32_t __correct_eeprom_dest_address(uint32_t curr_address)
 	return curr_address;
 }
 
-
-static uint32_t  __calc_hdr_byte_sum(struct amdgpu_ras_eeprom_control *control)
-{
-	int i;
-	uint32_t tbl_sum = 0;
-
-	/* Header checksum, skip checksum field in the calculation */
-	for (i = 0; i < sizeof(control->tbl_hdr) - sizeof(control->tbl_hdr.checksum); i++)
-		tbl_sum += *(((unsigned char *)&control->tbl_hdr) + i);
-
-	return tbl_sum;
-}
-
-static uint32_t  __calc_recs_byte_sum(struct eeprom_table_record *records,
-				      int num)
-{
-	int i, j;
-	uint32_t tbl_sum = 0;
-
-	/* Records checksum */
-	for (i = 0; i < num; i++) {
-		struct eeprom_table_record *record = &records[i];
-
-		for (j = 0; j < sizeof(*record); j++) {
-			tbl_sum += *(((unsigned char *)record) + j);
-		}
-	}
-
-	return tbl_sum;
-}
-
-static inline uint32_t  __calc_tbl_byte_sum(struct amdgpu_ras_eeprom_control *control,
-				  struct eeprom_table_record *records, int num)
-{
-	return __calc_hdr_byte_sum(control) + __calc_recs_byte_sum(records, num);
-}
-
-/* Checksum = 256 -((sum of all table entries) mod 256) */
-static void __update_tbl_checksum(struct amdgpu_ras_eeprom_control *control,
-				  struct eeprom_table_record *records, int num,
-				  uint32_t old_hdr_byte_sum)
-{
-	/*
-	 * This will update the table sum with new records.
-	 *
-	 * TODO: What happens when the EEPROM table is to be wrapped around
-	 * and old records from start will get overridden.
-	 */
-
-	/* need to recalculate updated header byte sum */
-	control->tbl_byte_sum -= old_hdr_byte_sum;
-	control->tbl_byte_sum += __calc_tbl_byte_sum(control, records, num);
-
-	control->tbl_hdr.checksum = 256 - (control->tbl_byte_sum % 256);
-}
-
-/* table sum mod 256 + checksum must equals 256 */
-static bool __validate_tbl_checksum(struct amdgpu_ras_eeprom_control *control,
-			    struct eeprom_table_record *records, int num)
-{
-	control->tbl_byte_sum = __calc_tbl_byte_sum(control, records, num);
-
-	if (control->tbl_hdr.checksum + (control->tbl_byte_sum % 256) != 256) {
-		DRM_WARN("Checksum mismatch, checksum: %u ", control->tbl_hdr.checksum);
-		return false;
-	}
-
-	return true;
-}
-
 int amdgpu_ras_eeprom_process_recods(struct amdgpu_ras_eeprom_control *control,
 					    struct eeprom_table_record *records,
 					    bool write,
 					    int num)
 {
 	int i, ret = 0;
-	struct i2c_msg *msgs;
-	unsigned char *buffs;
+	struct i2c_msg *msgs, *msg;
+	unsigned char *buffs, *buff;
+	struct eeprom_table_record *record;
 	struct amdgpu_device *adev = to_amdgpu_device(control);
 
-	if (adev->asic_type != CHIP_VEGA20)
+	if (adev->asic_type != CHIP_VEGA20 && adev->asic_type != CHIP_ARCTURUS)
 		return 0;
 
 	buffs = kcalloc(num, EEPROM_ADDRESS_SIZE + EEPROM_TABLE_RECORD_SIZE,
@@ -373,9 +398,9 @@ int amdgpu_ras_eeprom_process_recods(struct amdgpu_ras_eeprom_control *control,
 	 * 256b
 	 */
 	for (i = 0; i < num; i++) {
-		unsigned char *buff = &buffs[i * (EEPROM_ADDRESS_SIZE + EEPROM_TABLE_RECORD_SIZE)];
-		struct eeprom_table_record *record = &records[i];
-		struct i2c_msg *msg = &msgs[i];
+		buff = &buffs[i * (EEPROM_ADDRESS_SIZE + EEPROM_TABLE_RECORD_SIZE)];
+		record = &records[i];
+		msg = &msgs[i];
 
 		control->next_addr = __correct_eeprom_dest_address(control->next_addr);
 
@@ -415,8 +440,8 @@ int amdgpu_ras_eeprom_process_recods(struct amdgpu_ras_eeprom_control *control,
 
 	if (!write) {
 		for (i = 0; i < num; i++) {
-			unsigned char *buff = &buffs[i*(EEPROM_ADDRESS_SIZE + EEPROM_TABLE_RECORD_SIZE)];
-			struct eeprom_table_record *record = &records[i];
+			buff = &buffs[i*(EEPROM_ADDRESS_SIZE + EEPROM_TABLE_RECORD_SIZE)];
+			record = &records[i];
 
 			__decode_table_record_from_buff(control, record, buff + EEPROM_ADDRESS_SIZE);
 		}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.h
index 41f3fcb9a29b..622269957c1b 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.h
@@ -79,6 +79,7 @@ struct eeprom_table_record {
 
 int amdgpu_ras_eeprom_init(struct amdgpu_ras_eeprom_control *control);
 void amdgpu_ras_eeprom_fini(struct amdgpu_ras_eeprom_control *control);
+int amdgpu_ras_eeprom_reset_table(struct amdgpu_ras_eeprom_control *control);
 
 int amdgpu_ras_eeprom_process_recods(struct amdgpu_ras_eeprom_control *control,
 					    struct eeprom_table_record *records,
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.c
index 5c13c503e61f..6010999d9020 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.c
@@ -23,6 +23,7 @@
 
 #include "amdgpu.h"
 #include "amdgpu_sdma.h"
+#include "amdgpu_ras.h"
 
 #define AMDGPU_CSA_SDMA_SIZE 64
 /* SDMA CSA reside in the 3rd page of CSA */
@@ -83,3 +84,101 @@ uint64_t amdgpu_sdma_get_csa_mc_addr(struct amdgpu_ring *ring,
 
 	return csa_mc_addr;
 }
+
+int amdgpu_sdma_ras_late_init(struct amdgpu_device *adev,
+			      void *ras_ih_info)
+{
+	int r, i;
+	struct ras_ih_if *ih_info = (struct ras_ih_if *)ras_ih_info;
+	struct ras_fs_if fs_info = {
+		.sysfs_name = "sdma_err_count",
+		.debugfs_name = "sdma_err_inject",
+	};
+
+	if (!ih_info)
+		return -EINVAL;
+
+	if (!adev->sdma.ras_if) {
+		adev->sdma.ras_if = kmalloc(sizeof(struct ras_common_if), GFP_KERNEL);
+		if (!adev->sdma.ras_if)
+			return -ENOMEM;
+		adev->sdma.ras_if->block = AMDGPU_RAS_BLOCK__SDMA;
+		adev->sdma.ras_if->type = AMDGPU_RAS_ERROR__MULTI_UNCORRECTABLE;
+		adev->sdma.ras_if->sub_block_index = 0;
+		strcpy(adev->sdma.ras_if->name, "sdma");
+	}
+	fs_info.head = ih_info->head = *adev->sdma.ras_if;
+
+	r = amdgpu_ras_late_init(adev, adev->sdma.ras_if,
+				 &fs_info, ih_info);
+	if (r)
+		goto free;
+
+	if (amdgpu_ras_is_supported(adev, adev->sdma.ras_if->block)) {
+		for (i = 0; i < adev->sdma.num_instances; i++) {
+			r = amdgpu_irq_get(adev, &adev->sdma.ecc_irq,
+				AMDGPU_SDMA_IRQ_INSTANCE0 + i);
+			if (r)
+				goto late_fini;
+		}
+	} else {
+		r = 0;
+		goto free;
+	}
+
+        return 0;
+
+late_fini:
+	amdgpu_ras_late_fini(adev, adev->sdma.ras_if, ih_info);
+free:
+	kfree(adev->sdma.ras_if);
+	adev->sdma.ras_if = NULL;
+	return r;
+}
+
+void amdgpu_sdma_ras_fini(struct amdgpu_device *adev)
+{
+	if (amdgpu_ras_is_supported(adev, AMDGPU_RAS_BLOCK__SDMA) &&
+			adev->sdma.ras_if) {
+		struct ras_common_if *ras_if = adev->sdma.ras_if;
+		struct ras_ih_if ih_info = {
+			.head = *ras_if,
+			/* the cb member will not be used by
+			 * amdgpu_ras_interrupt_remove_handler, init it only
+			 * to cheat the check in ras_late_fini
+			 */
+			.cb = amdgpu_sdma_process_ras_data_cb,
+		};
+
+		amdgpu_ras_late_fini(adev, ras_if, &ih_info);
+		kfree(ras_if);
+	}
+}
+
+int amdgpu_sdma_process_ras_data_cb(struct amdgpu_device *adev,
+		void *err_data,
+		struct amdgpu_iv_entry *entry)
+{
+	kgd2kfd_set_sram_ecc_flag(adev->kfd.dev);
+	amdgpu_ras_reset_gpu(adev, 0);
+
+	return AMDGPU_RAS_SUCCESS;
+}
+
+int amdgpu_sdma_process_ecc_irq(struct amdgpu_device *adev,
+				      struct amdgpu_irq_src *source,
+				      struct amdgpu_iv_entry *entry)
+{
+	struct ras_common_if *ras_if = adev->sdma.ras_if;
+	struct ras_dispatch_if ih_data = {
+		.entry = entry,
+	};
+
+	if (!ras_if)
+		return 0;
+
+	ih_data.head = *ras_if;
+
+	amdgpu_ras_interrupt_dispatch(adev, &ih_data);
+	return 0;
+}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.h
index a9ae0d8a0589..761ff8be6314 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.h
@@ -104,4 +104,13 @@ struct amdgpu_sdma_instance *
 amdgpu_sdma_get_instance_from_ring(struct amdgpu_ring *ring);
 int amdgpu_sdma_get_index_from_ring(struct amdgpu_ring *ring, uint32_t *index);
 uint64_t amdgpu_sdma_get_csa_mc_addr(struct amdgpu_ring *ring, unsigned vmid);
+int amdgpu_sdma_ras_late_init(struct amdgpu_device *adev,
+			      void *ras_ih_info);
+void amdgpu_sdma_ras_fini(struct amdgpu_device *adev);
+int amdgpu_sdma_process_ras_data_cb(struct amdgpu_device *adev,
+		void *err_data,
+		struct amdgpu_iv_entry *entry);
+int amdgpu_sdma_process_ecc_irq(struct amdgpu_device *adev,
+				      struct amdgpu_irq_src *source,
+				      struct amdgpu_iv_entry *entry);
 #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_test.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_test.c
index b66d29d5ffa2..b158230af8db 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_test.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_test.c
@@ -138,6 +138,7 @@ static void amdgpu_do_test_moves(struct amdgpu_device *adev)
 		}
 
 		dma_fence_put(fence);
+		fence = NULL;
 
 		r = amdgpu_bo_kmap(vram_obj, &vram_map);
 		if (r) {
@@ -183,6 +184,7 @@ static void amdgpu_do_test_moves(struct amdgpu_device *adev)
 		}
 
 		dma_fence_put(fence);
+		fence = NULL;
 
 		r = amdgpu_bo_kmap(gtt_obj[i], &gtt_map);
 		if (r) {
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h
index 77674a7b9616..63e734a125fb 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h
@@ -170,7 +170,7 @@ TRACE_EVENT(amdgpu_cs_ioctl,
 			     __field(unsigned int, context)
 			     __field(unsigned int, seqno)
 			     __field(struct dma_fence *, fence)
-			     __field(char *, ring_name)
+			     __string(ring, to_amdgpu_ring(job->base.sched)->name)
 			     __field(u32, num_ibs)
 			     ),
 
@@ -179,12 +179,12 @@ TRACE_EVENT(amdgpu_cs_ioctl,
 			   __assign_str(timeline, AMDGPU_JOB_GET_TIMELINE_NAME(job))
 			   __entry->context = job->base.s_fence->finished.context;
 			   __entry->seqno = job->base.s_fence->finished.seqno;
-			   __entry->ring_name = to_amdgpu_ring(job->base.sched)->name;
+			   __assign_str(ring, to_amdgpu_ring(job->base.sched)->name)
 			   __entry->num_ibs = job->num_ibs;
 			   ),
 	    TP_printk("sched_job=%llu, timeline=%s, context=%u, seqno=%u, ring_name=%s, num_ibs=%u",
 		      __entry->sched_job_id, __get_str(timeline), __entry->context,
-		      __entry->seqno, __entry->ring_name, __entry->num_ibs)
+		      __entry->seqno, __get_str(ring), __entry->num_ibs)
 );
 
 TRACE_EVENT(amdgpu_sched_run_job,
@@ -195,7 +195,7 @@ TRACE_EVENT(amdgpu_sched_run_job,
 			     __string(timeline, AMDGPU_JOB_GET_TIMELINE_NAME(job))
 			     __field(unsigned int, context)
 			     __field(unsigned int, seqno)
-			     __field(char *, ring_name)
+			     __string(ring, to_amdgpu_ring(job->base.sched)->name)
 			     __field(u32, num_ibs)
 			     ),
 
@@ -204,12 +204,12 @@ TRACE_EVENT(amdgpu_sched_run_job,
 			   __assign_str(timeline, AMDGPU_JOB_GET_TIMELINE_NAME(job))
 			   __entry->context = job->base.s_fence->finished.context;
 			   __entry->seqno = job->base.s_fence->finished.seqno;
-			   __entry->ring_name = to_amdgpu_ring(job->base.sched)->name;
+			   __assign_str(ring, to_amdgpu_ring(job->base.sched)->name)
 			   __entry->num_ibs = job->num_ibs;
 			   ),
 	    TP_printk("sched_job=%llu, timeline=%s, context=%u, seqno=%u, ring_name=%s, num_ibs=%u",
 		      __entry->sched_job_id, __get_str(timeline), __entry->context,
-		      __entry->seqno, __entry->ring_name, __entry->num_ibs)
+		      __entry->seqno, __get_str(ring), __entry->num_ibs)
 );
 
 
@@ -323,14 +323,15 @@ DEFINE_EVENT(amdgpu_vm_mapping, amdgpu_vm_bo_cs,
 
 TRACE_EVENT(amdgpu_vm_set_ptes,
 	    TP_PROTO(uint64_t pe, uint64_t addr, unsigned count,
-		     uint32_t incr, uint64_t flags),
-	    TP_ARGS(pe, addr, count, incr, flags),
+		     uint32_t incr, uint64_t flags, bool direct),
+	    TP_ARGS(pe, addr, count, incr, flags, direct),
 	    TP_STRUCT__entry(
 			     __field(u64, pe)
 			     __field(u64, addr)
 			     __field(u32, count)
 			     __field(u32, incr)
 			     __field(u64, flags)
+			     __field(bool, direct)
 			     ),
 
 	    TP_fast_assign(
@@ -339,28 +340,32 @@ TRACE_EVENT(amdgpu_vm_set_ptes,
 			   __entry->count = count;
 			   __entry->incr = incr;
 			   __entry->flags = flags;
+			   __entry->direct = direct;
 			   ),
-	    TP_printk("pe=%010Lx, addr=%010Lx, incr=%u, flags=%llx, count=%u",
-		      __entry->pe, __entry->addr, __entry->incr,
-		      __entry->flags, __entry->count)
+	    TP_printk("pe=%010Lx, addr=%010Lx, incr=%u, flags=%llx, count=%u, "
+		      "direct=%d", __entry->pe, __entry->addr, __entry->incr,
+		      __entry->flags, __entry->count, __entry->direct)
 );
 
 TRACE_EVENT(amdgpu_vm_copy_ptes,
-	    TP_PROTO(uint64_t pe, uint64_t src, unsigned count),
-	    TP_ARGS(pe, src, count),
+	    TP_PROTO(uint64_t pe, uint64_t src, unsigned count, bool direct),
+	    TP_ARGS(pe, src, count, direct),
 	    TP_STRUCT__entry(
 			     __field(u64, pe)
 			     __field(u64, src)
 			     __field(u32, count)
+			     __field(bool, direct)
 			     ),
 
 	    TP_fast_assign(
 			   __entry->pe = pe;
 			   __entry->src = src;
 			   __entry->count = count;
+			   __entry->direct = direct;
 			   ),
-	    TP_printk("pe=%010Lx, src=%010Lx, count=%u",
-		      __entry->pe, __entry->src, __entry->count)
+	    TP_printk("pe=%010Lx, src=%010Lx, count=%u, direct=%d",
+		      __entry->pe, __entry->src, __entry->count,
+		      __entry->direct)
 );
 
 TRACE_EVENT(amdgpu_vm_flush,
@@ -468,7 +473,7 @@ TRACE_EVENT(amdgpu_ib_pipe_sync,
 	    TP_PROTO(struct amdgpu_job *sched_job, struct dma_fence *fence),
 	    TP_ARGS(sched_job, fence),
 	    TP_STRUCT__entry(
-			     __field(const char *,name)
+			     __string(ring, sched_job->base.sched->name)
 			     __field(uint64_t, id)
 			     __field(struct dma_fence *, fence)
 			     __field(uint64_t, ctx)
@@ -476,14 +481,14 @@ TRACE_EVENT(amdgpu_ib_pipe_sync,
 			     ),
 
 	    TP_fast_assign(
-			   __entry->name = sched_job->base.sched->name;
+			   __assign_str(ring, sched_job->base.sched->name)
 			   __entry->id = sched_job->base.id;
 			   __entry->fence = fence;
 			   __entry->ctx = fence->context;
 			   __entry->seqno = fence->seqno;
 			   ),
 	    TP_printk("job ring=%s, id=%llu, need pipe sync to fence=%p, context=%llu, seq=%u",
-		      __entry->name, __entry->id,
+		      __get_str(ring), __entry->id,
 		      __entry->fence, __entry->ctx,
 		      __entry->seqno)
 );
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
index dff41d0a85fe..2616e2eafdeb 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
@@ -35,10 +35,12 @@
 #include <linux/hmm.h>
 #include <linux/pagemap.h>
 #include <linux/sched/task.h>
+#include <linux/sched/mm.h>
 #include <linux/seq_file.h>
 #include <linux/slab.h>
 #include <linux/swap.h>
 #include <linux/swiotlb.h>
+#include <linux/dma-buf.h>
 
 #include <drm/ttm/ttm_bo_api.h>
 #include <drm/ttm/ttm_bo_driver.h>
@@ -54,6 +56,7 @@
 #include "amdgpu_trace.h"
 #include "amdgpu_amdkfd.h"
 #include "amdgpu_sdma.h"
+#include "amdgpu_ras.h"
 #include "bif/bif_4_1_d.h"
 
 static int amdgpu_map_buffer(struct ttm_buffer_object *bo,
@@ -484,15 +487,12 @@ static int amdgpu_move_vram_ram(struct ttm_buffer_object *bo, bool evict,
 				struct ttm_operation_ctx *ctx,
 				struct ttm_mem_reg *new_mem)
 {
-	struct amdgpu_device *adev;
 	struct ttm_mem_reg *old_mem = &bo->mem;
 	struct ttm_mem_reg tmp_mem;
 	struct ttm_place placements;
 	struct ttm_placement placement;
 	int r;
 
-	adev = amdgpu_ttm_adev(bo->bdev);
-
 	/* create space/pages for new_mem in GTT space */
 	tmp_mem = *new_mem;
 	tmp_mem.mm_node = NULL;
@@ -543,15 +543,12 @@ static int amdgpu_move_ram_vram(struct ttm_buffer_object *bo, bool evict,
 				struct ttm_operation_ctx *ctx,
 				struct ttm_mem_reg *new_mem)
 {
-	struct amdgpu_device *adev;
 	struct ttm_mem_reg *old_mem = &bo->mem;
 	struct ttm_mem_reg tmp_mem;
 	struct ttm_placement placement;
 	struct ttm_place placements;
 	int r;
 
-	adev = amdgpu_ttm_adev(bo->bdev);
-
 	/* make space in GTT for old_mem buffer */
 	tmp_mem = *new_mem;
 	tmp_mem.mm_node = NULL;
@@ -763,6 +760,7 @@ static unsigned long amdgpu_ttm_io_mem_pfn(struct ttm_buffer_object *bo,
  */
 struct amdgpu_ttm_tt {
 	struct ttm_dma_tt	ttm;
+	struct drm_gem_object	*gobj;
 	u64			offset;
 	uint64_t		userptr;
 	struct task_struct	*usertask;
@@ -772,6 +770,20 @@ struct amdgpu_ttm_tt {
 #endif
 };
 
+#ifdef CONFIG_DRM_AMDGPU_USERPTR
+/* flags used by HMM internal, not related to CPU/GPU PTE flags */
+static const uint64_t hmm_range_flags[HMM_PFN_FLAG_MAX] = {
+	(1 << 0), /* HMM_PFN_VALID */
+	(1 << 1), /* HMM_PFN_WRITE */
+	0 /* HMM_PFN_DEVICE_PRIVATE */
+};
+
+static const uint64_t hmm_range_values[HMM_PFN_VALUE_MAX] = {
+	0xfffffffffffffffeUL, /* HMM_PFN_ERROR */
+	0, /* HMM_PFN_NONE */
+	0xfffffffffffffffcUL /* HMM_PFN_SPECIAL */
+};
+
 /**
  * amdgpu_ttm_tt_get_user_pages - get device accessible pages that back user
  * memory and start HMM tracking CPU page table update
@@ -779,85 +791,89 @@ struct amdgpu_ttm_tt {
  * Calling function must call amdgpu_ttm_tt_userptr_range_done() once and only
  * once afterwards to stop HMM tracking
  */
-#if IS_ENABLED(CONFIG_DRM_AMDGPU_USERPTR)
-
-#define MAX_RETRY_HMM_RANGE_FAULT	16
-
 int amdgpu_ttm_tt_get_user_pages(struct amdgpu_bo *bo, struct page **pages)
 {
-	struct hmm_mirror *mirror = bo->mn ? &bo->mn->mirror : NULL;
 	struct ttm_tt *ttm = bo->tbo.ttm;
 	struct amdgpu_ttm_tt *gtt = (void *)ttm;
-	struct mm_struct *mm = gtt->usertask->mm;
 	unsigned long start = gtt->userptr;
 	struct vm_area_struct *vma;
 	struct hmm_range *range;
+	unsigned long timeout;
+	struct mm_struct *mm;
 	unsigned long i;
-	uint64_t *pfns;
 	int r = 0;
 
-	if (!mm) /* Happens during process shutdown */
-		return -ESRCH;
-
-	if (unlikely(!mirror)) {
-		DRM_DEBUG_DRIVER("Failed to get hmm_mirror\n");
-		r = -EFAULT;
-		goto out;
+	mm = bo->notifier.mm;
+	if (unlikely(!mm)) {
+		DRM_DEBUG_DRIVER("BO is not registered?\n");
+		return -EFAULT;
 	}
 
-	vma = find_vma(mm, start);
-	if (unlikely(!vma || start < vma->vm_start)) {
-		r = -EFAULT;
-		goto out;
-	}
-	if (unlikely((gtt->userflags & AMDGPU_GEM_USERPTR_ANONONLY) &&
-		vma->vm_file)) {
-		r = -EPERM;
-		goto out;
-	}
+	/* Another get_user_pages is running at the same time?? */
+	if (WARN_ON(gtt->range))
+		return -EFAULT;
+
+	if (!mmget_not_zero(mm)) /* Happens during process shutdown */
+		return -ESRCH;
 
 	range = kzalloc(sizeof(*range), GFP_KERNEL);
 	if (unlikely(!range)) {
 		r = -ENOMEM;
 		goto out;
 	}
+	range->notifier = &bo->notifier;
+	range->flags = hmm_range_flags;
+	range->values = hmm_range_values;
+	range->pfn_shift = PAGE_SHIFT;
+	range->start = bo->notifier.interval_tree.start;
+	range->end = bo->notifier.interval_tree.last + 1;
+	range->default_flags = hmm_range_flags[HMM_PFN_VALID];
+	if (!amdgpu_ttm_tt_is_readonly(ttm))
+		range->default_flags |= range->flags[HMM_PFN_WRITE];
 
-	pfns = kvmalloc_array(ttm->num_pages, sizeof(*pfns), GFP_KERNEL);
-	if (unlikely(!pfns)) {
+	range->pfns = kvmalloc_array(ttm->num_pages, sizeof(*range->pfns),
+				     GFP_KERNEL);
+	if (unlikely(!range->pfns)) {
 		r = -ENOMEM;
 		goto out_free_ranges;
 	}
 
-	amdgpu_hmm_init_range(range);
-	range->default_flags = range->flags[HMM_PFN_VALID];
-	range->default_flags |= amdgpu_ttm_tt_is_readonly(ttm) ?
-				0 : range->flags[HMM_PFN_WRITE];
-	range->pfn_flags_mask = 0;
-	range->pfns = pfns;
-	range->start = start;
-	range->end = start + ttm->num_pages * PAGE_SIZE;
-
-	hmm_range_register(range, mirror);
+	down_read(&mm->mmap_sem);
+	vma = find_vma(mm, start);
+	if (unlikely(!vma || start < vma->vm_start)) {
+		r = -EFAULT;
+		goto out_unlock;
+	}
+	if (unlikely((gtt->userflags & AMDGPU_GEM_USERPTR_ANONONLY) &&
+		vma->vm_file)) {
+		r = -EPERM;
+		goto out_unlock;
+	}
+	up_read(&mm->mmap_sem);
+	timeout = jiffies + msecs_to_jiffies(HMM_RANGE_DEFAULT_TIMEOUT);
 
-	/*
-	 * Just wait for range to be valid, safe to ignore return value as we
-	 * will use the return value of hmm_range_fault() below under the
-	 * mmap_sem to ascertain the validity of the range.
-	 */
-	hmm_range_wait_until_valid(range, HMM_RANGE_DEFAULT_TIMEOUT);
+retry:
+	range->notifier_seq = mmu_interval_read_begin(&bo->notifier);
 
 	down_read(&mm->mmap_sem);
 	r = hmm_range_fault(range, 0);
 	up_read(&mm->mmap_sem);
-
-	if (unlikely(r < 0))
+	if (unlikely(r <= 0)) {
+		/*
+		 * FIXME: This timeout should encompass the retry from
+		 * mmu_interval_read_retry() as well.
+		 */
+		if ((r == 0 || r == -EBUSY) && !time_after(jiffies, timeout))
+			goto retry;
 		goto out_free_pfns;
+	}
 
 	for (i = 0; i < ttm->num_pages; i++) {
-		pages[i] = hmm_device_entry_to_page(range, pfns[i]);
+		/* FIXME: The pages cannot be touched outside the notifier_lock */
+		pages[i] = hmm_device_entry_to_page(range, range->pfns[i]);
 		if (unlikely(!pages[i])) {
 			pr_err("Page fault failed for pfn[%lu] = 0x%llx\n",
-			       i, pfns[i]);
+			       i, range->pfns[i]);
 			r = -ENOMEM;
 
 			goto out_free_pfns;
@@ -865,15 +881,18 @@ int amdgpu_ttm_tt_get_user_pages(struct amdgpu_bo *bo, struct page **pages)
 	}
 
 	gtt->range = range;
+	mmput(mm);
 
 	return 0;
 
+out_unlock:
+	up_read(&mm->mmap_sem);
 out_free_pfns:
-	hmm_range_unregister(range);
-	kvfree(pfns);
+	kvfree(range->pfns);
 out_free_ranges:
 	kfree(range);
 out:
+	mmput(mm);
 	return r;
 }
 
@@ -898,15 +917,18 @@ bool amdgpu_ttm_tt_get_user_pages_done(struct ttm_tt *ttm)
 		"No user pages to check\n");
 
 	if (gtt->range) {
-		r = hmm_range_valid(gtt->range);
-		hmm_range_unregister(gtt->range);
-
+		/*
+		 * FIXME: Must always hold notifier_lock for this, and must
+		 * not ignore the return code.
+		 */
+		r = mmu_interval_read_retry(gtt->range->notifier,
+					 gtt->range->notifier_seq);
 		kvfree(gtt->range->pfns);
 		kfree(gtt->range);
 		gtt->range = NULL;
 	}
 
-	return r;
+	return !r;
 }
 #endif
 
@@ -987,10 +1009,18 @@ static void amdgpu_ttm_tt_unpin_userptr(struct ttm_tt *ttm)
 	sg_free_table(ttm->sg);
 
 #if IS_ENABLED(CONFIG_DRM_AMDGPU_USERPTR)
-	if (gtt->range &&
-	    ttm->pages[0] == hmm_device_entry_to_page(gtt->range,
-						      gtt->range->pfns[0]))
-		WARN_ONCE(1, "Missing get_user_page_done\n");
+	if (gtt->range) {
+		unsigned long i;
+
+		for (i = 0; i < ttm->num_pages; i++) {
+			if (ttm->pages[i] !=
+				hmm_device_entry_to_page(gtt->range,
+					      gtt->range->pfns[i]))
+				break;
+		}
+
+		WARN((i == ttm->num_pages), "Missing get_user_page_done\n");
+	}
 #endif
 }
 
@@ -1217,16 +1247,14 @@ static struct ttm_backend_func amdgpu_backend_func = {
 static struct ttm_tt *amdgpu_ttm_tt_create(struct ttm_buffer_object *bo,
 					   uint32_t page_flags)
 {
-	struct amdgpu_device *adev;
 	struct amdgpu_ttm_tt *gtt;
 
-	adev = amdgpu_ttm_adev(bo->bdev);
-
 	gtt = kzalloc(sizeof(struct amdgpu_ttm_tt), GFP_KERNEL);
 	if (gtt == NULL) {
 		return NULL;
 	}
 	gtt->ttm.ttm.func = &amdgpu_backend_func;
+	gtt->gobj = &bo->base;
 
 	/* allocate space for the uninitialized page entries */
 	if (ttm_sg_tt_init(&gtt->ttm, bo, page_flags)) {
@@ -1247,7 +1275,6 @@ static int amdgpu_ttm_tt_populate(struct ttm_tt *ttm,
 {
 	struct amdgpu_device *adev = amdgpu_ttm_adev(ttm->bdev);
 	struct amdgpu_ttm_tt *gtt = (void *)ttm;
-	bool slave = !!(ttm->page_flags & TTM_PAGE_FLAG_SG);
 
 	/* user pages are bound by amdgpu_ttm_tt_pin_userptr() */
 	if (gtt && gtt->userptr) {
@@ -1260,7 +1287,19 @@ static int amdgpu_ttm_tt_populate(struct ttm_tt *ttm,
 		return 0;
 	}
 
-	if (slave && ttm->sg) {
+	if (ttm->page_flags & TTM_PAGE_FLAG_SG) {
+		if (!ttm->sg) {
+			struct dma_buf_attachment *attach;
+			struct sg_table *sgt;
+
+			attach = gtt->gobj->import_attach;
+			sgt = dma_buf_map_attachment(attach, DMA_BIDIRECTIONAL);
+			if (IS_ERR(sgt))
+				return PTR_ERR(sgt);
+
+			ttm->sg = sgt;
+		}
+
 		drm_prime_sg_to_page_addr_arrays(ttm->sg, ttm->pages,
 						 gtt->ttm.dma_address,
 						 ttm->num_pages);
@@ -1287,9 +1326,8 @@ static int amdgpu_ttm_tt_populate(struct ttm_tt *ttm,
  */
 static void amdgpu_ttm_tt_unpopulate(struct ttm_tt *ttm)
 {
-	struct amdgpu_device *adev;
 	struct amdgpu_ttm_tt *gtt = (void *)ttm;
-	bool slave = !!(ttm->page_flags & TTM_PAGE_FLAG_SG);
+	struct amdgpu_device *adev;
 
 	if (gtt && gtt->userptr) {
 		amdgpu_ttm_tt_set_user_pages(ttm, NULL);
@@ -1298,7 +1336,16 @@ static void amdgpu_ttm_tt_unpopulate(struct ttm_tt *ttm)
 		return;
 	}
 
-	if (slave)
+	if (ttm->sg && gtt->gobj->import_attach) {
+		struct dma_buf_attachment *attach;
+
+		attach = gtt->gobj->import_attach;
+		dma_buf_unmap_attachment(attach, ttm->sg, DMA_BIDIRECTIONAL);
+		ttm->sg = NULL;
+		return;
+	}
+
+	if (ttm->page_flags & TTM_PAGE_FLAG_SG)
 		return;
 
 	adev = amdgpu_ttm_adev(ttm->bdev);
@@ -1634,81 +1681,105 @@ static void amdgpu_ttm_fw_reserve_vram_fini(struct amdgpu_device *adev)
  */
 static int amdgpu_ttm_fw_reserve_vram_init(struct amdgpu_device *adev)
 {
-	struct ttm_operation_ctx ctx = { false, false };
-	struct amdgpu_bo_param bp;
-	int r = 0;
-	int i;
-	u64 vram_size = adev->gmc.visible_vram_size;
-	u64 offset = adev->fw_vram_usage.start_offset;
-	u64 size = adev->fw_vram_usage.size;
-	struct amdgpu_bo *bo;
-
-	memset(&bp, 0, sizeof(bp));
-	bp.size = adev->fw_vram_usage.size;
-	bp.byte_align = PAGE_SIZE;
-	bp.domain = AMDGPU_GEM_DOMAIN_VRAM;
-	bp.flags = AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED |
-		AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS;
-	bp.type = ttm_bo_type_kernel;
-	bp.resv = NULL;
+	uint64_t vram_size = adev->gmc.visible_vram_size;
+
 	adev->fw_vram_usage.va = NULL;
 	adev->fw_vram_usage.reserved_bo = NULL;
 
-	if (adev->fw_vram_usage.size > 0 &&
-		adev->fw_vram_usage.size <= vram_size) {
+	if (adev->fw_vram_usage.size == 0 ||
+	    adev->fw_vram_usage.size > vram_size)
+		return 0;
 
-		r = amdgpu_bo_create(adev, &bp,
-				     &adev->fw_vram_usage.reserved_bo);
-		if (r)
-			goto error_create;
+	return amdgpu_bo_create_kernel_at(adev,
+					  adev->fw_vram_usage.start_offset,
+					  adev->fw_vram_usage.size,
+					  AMDGPU_GEM_DOMAIN_VRAM,
+					  &adev->fw_vram_usage.reserved_bo,
+					  &adev->fw_vram_usage.va);
+}
 
-		r = amdgpu_bo_reserve(adev->fw_vram_usage.reserved_bo, false);
-		if (r)
-			goto error_reserve;
+/*
+ * Memoy training reservation functions
+ */
 
-		/* remove the original mem node and create a new one at the
-		 * request position
-		 */
-		bo = adev->fw_vram_usage.reserved_bo;
-		offset = ALIGN(offset, PAGE_SIZE);
-		for (i = 0; i < bo->placement.num_placement; ++i) {
-			bo->placements[i].fpfn = offset >> PAGE_SHIFT;
-			bo->placements[i].lpfn = (offset + size) >> PAGE_SHIFT;
-		}
+/**
+ * amdgpu_ttm_training_reserve_vram_fini - free memory training reserved vram
+ *
+ * @adev: amdgpu_device pointer
+ *
+ * free memory training reserved vram if it has been reserved.
+ */
+static int amdgpu_ttm_training_reserve_vram_fini(struct amdgpu_device *adev)
+{
+	struct psp_memory_training_context *ctx = &adev->psp.mem_train_ctx;
 
-		ttm_bo_mem_put(&bo->tbo, &bo->tbo.mem);
-		r = ttm_bo_mem_space(&bo->tbo, &bo->placement,
-				     &bo->tbo.mem, &ctx);
-		if (r)
-			goto error_pin;
+	ctx->init = PSP_MEM_TRAIN_NOT_SUPPORT;
+	amdgpu_bo_free_kernel(&ctx->c2p_bo, NULL, NULL);
+	ctx->c2p_bo = NULL;
 
-		r = amdgpu_bo_pin_restricted(adev->fw_vram_usage.reserved_bo,
-			AMDGPU_GEM_DOMAIN_VRAM,
-			adev->fw_vram_usage.start_offset,
-			(adev->fw_vram_usage.start_offset +
-			adev->fw_vram_usage.size));
-		if (r)
-			goto error_pin;
-		r = amdgpu_bo_kmap(adev->fw_vram_usage.reserved_bo,
-			&adev->fw_vram_usage.va);
-		if (r)
-			goto error_kmap;
+	amdgpu_bo_free_kernel(&ctx->p2c_bo, NULL, NULL);
+	ctx->p2c_bo = NULL;
+
+	return 0;
+}
 
-		amdgpu_bo_unreserve(adev->fw_vram_usage.reserved_bo);
+/**
+ * amdgpu_ttm_training_reserve_vram_init - create bo vram reservation from memory training
+ *
+ * @adev: amdgpu_device pointer
+ *
+ * create bo vram reservation from memory training.
+ */
+static int amdgpu_ttm_training_reserve_vram_init(struct amdgpu_device *adev)
+{
+	int ret;
+	struct psp_memory_training_context *ctx = &adev->psp.mem_train_ctx;
+
+	memset(ctx, 0, sizeof(*ctx));
+	if (!adev->fw_vram_usage.mem_train_support) {
+		DRM_DEBUG("memory training does not support!\n");
+		return 0;
 	}
-	return r;
 
-error_kmap:
-	amdgpu_bo_unpin(adev->fw_vram_usage.reserved_bo);
-error_pin:
-	amdgpu_bo_unreserve(adev->fw_vram_usage.reserved_bo);
-error_reserve:
-	amdgpu_bo_unref(&adev->fw_vram_usage.reserved_bo);
-error_create:
-	adev->fw_vram_usage.va = NULL;
-	adev->fw_vram_usage.reserved_bo = NULL;
-	return r;
+	ctx->c2p_train_data_offset = adev->fw_vram_usage.mem_train_fb_loc;
+	ctx->p2c_train_data_offset = (adev->gmc.mc_vram_size - GDDR6_MEM_TRAINING_OFFSET);
+	ctx->train_data_size = GDDR6_MEM_TRAINING_DATA_SIZE_IN_BYTES;
+
+	DRM_DEBUG("train_data_size:%llx,p2c_train_data_offset:%llx,c2p_train_data_offset:%llx.\n",
+		  ctx->train_data_size,
+		  ctx->p2c_train_data_offset,
+		  ctx->c2p_train_data_offset);
+
+	ret = amdgpu_bo_create_kernel_at(adev,
+					 ctx->p2c_train_data_offset,
+					 ctx->train_data_size,
+					 AMDGPU_GEM_DOMAIN_VRAM,
+					 &ctx->p2c_bo,
+					 NULL);
+	if (ret) {
+		DRM_ERROR("alloc p2c_bo failed(%d)!\n", ret);
+		goto Err_out;
+	}
+
+	ret = amdgpu_bo_create_kernel_at(adev,
+					 ctx->c2p_train_data_offset,
+					 ctx->train_data_size,
+					 AMDGPU_GEM_DOMAIN_VRAM,
+					 &ctx->c2p_bo,
+					 NULL);
+	if (ret) {
+		DRM_ERROR("alloc c2p_bo failed(%d)!\n", ret);
+		goto Err_out;
+	}
+
+	ctx->init = PSP_MEM_TRAIN_RESERVE_SUCCESS;
+	return 0;
+
+Err_out:
+	amdgpu_ttm_training_reserve_vram_fini(adev);
+	return ret;
 }
+
 /**
  * amdgpu_ttm_init - Init the memory management (ttm) as well as various
  * gtt/vram related fields.
@@ -1731,6 +1802,7 @@ int amdgpu_ttm_init(struct amdgpu_device *adev)
 	r = ttm_bo_device_init(&adev->mman.bdev,
 			       &amdgpu_bo_driver,
 			       adev->ddev->anon_inode->i_mapping,
+			       adev->ddev->vma_offset_manager,
 			       dma_addressing_limited(adev->dev));
 	if (r) {
 		DRM_ERROR("failed initializing buffer object driver(%d).\n", r);
@@ -1771,6 +1843,14 @@ int amdgpu_ttm_init(struct amdgpu_device *adev)
 		return r;
 	}
 
+	/*
+	 *The reserved vram for memory training must be pinned to the specified
+	 *place on the VRAM, so reserve it early.
+	 */
+	r = amdgpu_ttm_training_reserve_vram_init(adev);
+	if (r)
+		return r;
+
 	/* allocate memory as required for VGA
 	 * This is used for VGA emulation and pre-OS scanout buffers to
 	 * avoid display artifacts while transitioning between pre-OS
@@ -1781,6 +1861,20 @@ int amdgpu_ttm_init(struct amdgpu_device *adev)
 				    NULL, &stolen_vga_buf);
 	if (r)
 		return r;
+
+	/*
+	 * reserve one TMR (64K) memory at the top of VRAM which holds
+	 * IP Discovery data and is protected by PSP.
+	 */
+	r = amdgpu_bo_create_kernel_at(adev,
+				       adev->gmc.real_vram_size - DISCOVERY_TMR_SIZE,
+				       DISCOVERY_TMR_SIZE,
+				       AMDGPU_GEM_DOMAIN_VRAM,
+				       &adev->discovery_memory,
+				       NULL);
+	if (r)
+		return r;
+
 	DRM_INFO("amdgpu: %uM of VRAM memory ready\n",
 		 (unsigned) (adev->gmc.real_vram_size / (1024 * 1024)));
 
@@ -1856,7 +1950,11 @@ void amdgpu_ttm_fini(struct amdgpu_device *adev)
 		return;
 
 	amdgpu_ttm_debugfs_fini(adev);
+	amdgpu_ttm_training_reserve_vram_fini(adev);
+	/* return the IP Discovery TMR memory back to VRAM */
+	amdgpu_bo_free_kernel(&adev->discovery_memory, NULL, NULL);
 	amdgpu_ttm_fw_reserve_vram_fini(adev);
+
 	if (adev->mman.aper_base_kaddr)
 		iounmap(adev->mman.aper_base_kaddr);
 	adev->mman.aper_base_kaddr = NULL;
@@ -1952,10 +2050,7 @@ static int amdgpu_map_buffer(struct ttm_buffer_object *bo,
 	*addr += (u64)window * AMDGPU_GTT_MAX_TRANSFER_SIZE *
 		AMDGPU_GPU_PAGE_SIZE;
 
-	num_dw = adev->mman.buffer_funcs->copy_num_dw;
-	while (num_dw & 0x7)
-		num_dw++;
-
+	num_dw = ALIGN(adev->mman.buffer_funcs->copy_num_dw, 8);
 	num_bytes = num_pages * 8;
 
 	r = amdgpu_job_alloc_with_ib(adev, num_dw * 4 + num_bytes, &job);
@@ -2015,11 +2110,7 @@ int amdgpu_copy_buffer(struct amdgpu_ring *ring, uint64_t src_offset,
 
 	max_bytes = adev->mman.buffer_funcs->copy_max_bytes;
 	num_loops = DIV_ROUND_UP(byte_count, max_bytes);
-	num_dw = num_loops * adev->mman.buffer_funcs->copy_num_dw;
-
-	/* for IB padding */
-	while (num_dw & 0x7)
-		num_dw++;
+	num_dw = ALIGN(num_loops * adev->mman.buffer_funcs->copy_num_dw, 8);
 
 	r = amdgpu_job_alloc_with_ib(adev, num_dw * 4, &job);
 	if (r)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c
index 3a6115ad0196..833fc4b68940 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c
@@ -360,6 +360,7 @@ amdgpu_ucode_get_load_type(struct amdgpu_device *adev, int load_type)
 	case CHIP_RAVEN:
 	case CHIP_VEGA12:
 	case CHIP_VEGA20:
+	case CHIP_ARCTURUS:
 	case CHIP_RENOIR:
 	case CHIP_NAVI10:
 	case CHIP_NAVI14:
@@ -368,8 +369,6 @@ amdgpu_ucode_get_load_type(struct amdgpu_device *adev, int load_type)
 			return AMDGPU_FW_LOAD_DIRECT;
 		else
 			return AMDGPU_FW_LOAD_PSP;
-	case CHIP_ARCTURUS:
-		return AMDGPU_FW_LOAD_DIRECT;
 
 	default:
 		DRM_ERROR("Unknown firmware load type\n");
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h
index b34f00d42049..410587b950f3 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h
@@ -108,6 +108,12 @@ struct ta_firmware_header_v1_0 {
 	uint32_t ta_ras_ucode_version;
 	uint32_t ta_ras_offset_bytes;
 	uint32_t ta_ras_size_bytes;
+	uint32_t ta_hdcp_ucode_version;
+	uint32_t ta_hdcp_offset_bytes;
+	uint32_t ta_hdcp_size_bytes;
+	uint32_t ta_dtm_ucode_version;
+	uint32_t ta_dtm_offset_bytes;
+	uint32_t ta_dtm_size_bytes;
 };
 
 /* version_major=1, version_minor=0 */
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.c
new file mode 100644
index 000000000000..d4fb9cf27e21
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.c
@@ -0,0 +1,158 @@
+/*
+ * Copyright 2019 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#include "amdgpu_ras.h"
+
+int amdgpu_umc_ras_late_init(struct amdgpu_device *adev)
+{
+	int r;
+	struct ras_fs_if fs_info = {
+		.sysfs_name = "umc_err_count",
+		.debugfs_name = "umc_err_inject",
+	};
+	struct ras_ih_if ih_info = {
+		.cb = amdgpu_umc_process_ras_data_cb,
+	};
+
+	if (!adev->umc.ras_if) {
+		adev->umc.ras_if =
+			kmalloc(sizeof(struct ras_common_if), GFP_KERNEL);
+		if (!adev->umc.ras_if)
+			return -ENOMEM;
+		adev->umc.ras_if->block = AMDGPU_RAS_BLOCK__UMC;
+		adev->umc.ras_if->type = AMDGPU_RAS_ERROR__MULTI_UNCORRECTABLE;
+		adev->umc.ras_if->sub_block_index = 0;
+		strcpy(adev->umc.ras_if->name, "umc");
+	}
+	ih_info.head = fs_info.head = *adev->umc.ras_if;
+
+	r = amdgpu_ras_late_init(adev, adev->umc.ras_if,
+				 &fs_info, &ih_info);
+	if (r)
+		goto free;
+
+	if (amdgpu_ras_is_supported(adev, adev->umc.ras_if->block)) {
+		r = amdgpu_irq_get(adev, &adev->gmc.ecc_irq, 0);
+		if (r)
+			goto late_fini;
+	} else {
+		r = 0;
+		goto free;
+	}
+
+	/* ras init of specific umc version */
+	if (adev->umc.funcs && adev->umc.funcs->err_cnt_init)
+		adev->umc.funcs->err_cnt_init(adev);
+
+	return 0;
+
+late_fini:
+	amdgpu_ras_late_fini(adev, adev->umc.ras_if, &ih_info);
+free:
+	kfree(adev->umc.ras_if);
+	adev->umc.ras_if = NULL;
+	return r;
+}
+
+void amdgpu_umc_ras_fini(struct amdgpu_device *adev)
+{
+	if (amdgpu_ras_is_supported(adev, AMDGPU_RAS_BLOCK__UMC) &&
+			adev->umc.ras_if) {
+		struct ras_common_if *ras_if = adev->umc.ras_if;
+		struct ras_ih_if ih_info = {
+			.head = *ras_if,
+			.cb = amdgpu_umc_process_ras_data_cb,
+		};
+
+		amdgpu_ras_late_fini(adev, ras_if, &ih_info);
+		kfree(ras_if);
+	}
+}
+
+int amdgpu_umc_process_ras_data_cb(struct amdgpu_device *adev,
+		void *ras_error_status,
+		struct amdgpu_iv_entry *entry)
+{
+	struct ras_err_data *err_data = (struct ras_err_data *)ras_error_status;
+
+	/* When “Full RAS” is enabled, the per-IP interrupt sources should
+	 * be disabled and the driver should only look for the aggregated
+	 * interrupt via sync flood
+	 */
+	if (amdgpu_ras_is_supported(adev, AMDGPU_RAS_BLOCK__GFX))
+		return AMDGPU_RAS_SUCCESS;
+
+	kgd2kfd_set_sram_ecc_flag(adev->kfd.dev);
+	if (adev->umc.funcs &&
+	    adev->umc.funcs->query_ras_error_count)
+	    adev->umc.funcs->query_ras_error_count(adev, ras_error_status);
+
+	if (adev->umc.funcs &&
+	    adev->umc.funcs->query_ras_error_address &&
+	    adev->umc.max_ras_err_cnt_per_query) {
+		err_data->err_addr =
+			kcalloc(adev->umc.max_ras_err_cnt_per_query,
+				sizeof(struct eeprom_table_record), GFP_KERNEL);
+		/* still call query_ras_error_address to clear error status
+		 * even NOMEM error is encountered
+		 */
+		if(!err_data->err_addr)
+			DRM_WARN("Failed to alloc memory for umc error address record!\n");
+
+		/* umc query_ras_error_address is also responsible for clearing
+		 * error status
+		 */
+		adev->umc.funcs->query_ras_error_address(adev, ras_error_status);
+	}
+
+	/* only uncorrectable error needs gpu reset */
+	if (err_data->ue_count) {
+		if (err_data->err_addr_cnt &&
+		    amdgpu_ras_add_bad_pages(adev, err_data->err_addr,
+						err_data->err_addr_cnt))
+			DRM_WARN("Failed to add ras bad page!\n");
+
+		amdgpu_ras_reset_gpu(adev, 0);
+	}
+
+	kfree(err_data->err_addr);
+	return AMDGPU_RAS_SUCCESS;
+}
+
+int amdgpu_umc_process_ecc_irq(struct amdgpu_device *adev,
+		struct amdgpu_irq_src *source,
+		struct amdgpu_iv_entry *entry)
+{
+	struct ras_common_if *ras_if = adev->umc.ras_if;
+	struct ras_dispatch_if ih_data = {
+		.entry = entry,
+	};
+
+	if (!ras_if)
+		return 0;
+
+	ih_data.head = *ras_if;
+
+	amdgpu_ras_interrupt_dispatch(adev, &ih_data);
+	return 0;
+}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.h
index 975afa04df09..3283032a78e5 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.h
@@ -54,7 +54,8 @@
 	adev->umc.funcs->disable_umc_index_mode(adev);
 
 struct amdgpu_umc_funcs {
-	void (*ras_init)(struct amdgpu_device *adev);
+	void (*err_cnt_init)(struct amdgpu_device *adev);
+	int (*ras_late_init)(struct amdgpu_device *adev);
 	void (*query_ras_error_count)(struct amdgpu_device *adev,
 					void *ras_error_status);
 	void (*query_ras_error_address)(struct amdgpu_device *adev,
@@ -62,6 +63,7 @@ struct amdgpu_umc_funcs {
 	void (*enable_umc_index_mode)(struct amdgpu_device *adev,
 					uint32_t umc_instance);
 	void (*disable_umc_index_mode)(struct amdgpu_device *adev);
+	void (*init_registers)(struct amdgpu_device *adev);
 };
 
 struct amdgpu_umc {
@@ -75,8 +77,17 @@ struct amdgpu_umc {
 	uint32_t channel_offs;
 	/* channel index table of interleaved memory */
 	const uint32_t *channel_idx_tbl;
+	struct ras_common_if *ras_if;
 
 	const struct amdgpu_umc_funcs *funcs;
 };
 
+int amdgpu_umc_ras_late_init(struct amdgpu_device *adev);
+void amdgpu_umc_ras_fini(struct amdgpu_device *adev);
+int amdgpu_umc_process_ras_data_cb(struct amdgpu_device *adev,
+		void *ras_error_status,
+		struct amdgpu_iv_entry *entry);
+int amdgpu_umc_process_ecc_irq(struct amdgpu_device *adev,
+		struct amdgpu_irq_src *source,
+		struct amdgpu_iv_entry *entry);
 #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
index b2c364b8695f..e324bfe6c58f 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
@@ -39,6 +39,8 @@
 #include "cikd.h"
 #include "uvd/uvd_4_2_d.h"
 
+#include "amdgpu_ras.h"
+
 /* 1 second timeout */
 #define UVD_IDLE_TIMEOUT	msecs_to_jiffies(1000)
 
@@ -297,6 +299,7 @@ int amdgpu_uvd_sw_fini(struct amdgpu_device *adev)
 {
 	int i, j;
 
+	cancel_delayed_work_sync(&adev->uvd.idle_work);
 	drm_sched_entity_destroy(&adev->uvd.entity);
 
 	for (j = 0; j < adev->uvd.num_uvd_inst; ++j) {
@@ -372,7 +375,13 @@ int amdgpu_uvd_suspend(struct amdgpu_device *adev)
 		if (!adev->uvd.inst[j].saved_bo)
 			return -ENOMEM;
 
-		memcpy_fromio(adev->uvd.inst[j].saved_bo, ptr, size);
+		/* re-write 0 since err_event_athub will corrupt VCPU buffer */
+		if (amdgpu_ras_intr_triggered()) {
+			DRM_WARN("UVD VCPU state may lost due to RAS ERREVENT_ATHUB_INTERRUPT\n");
+			memset(adev->uvd.inst[j].saved_bo, 0, size);
+		} else {
+			memcpy_fromio(adev->uvd.inst[j].saved_bo, ptr, size);
+		}
 	}
 	return 0;
 }
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
index b70b3c45bb29..46b590af2fd2 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
@@ -80,6 +80,11 @@ MODULE_FIRMWARE(FIRMWARE_VEGA12);
 MODULE_FIRMWARE(FIRMWARE_VEGA20);
 
 static void amdgpu_vce_idle_work_handler(struct work_struct *work);
+static int amdgpu_vce_get_create_msg(struct amdgpu_ring *ring, uint32_t handle,
+				     struct amdgpu_bo *bo,
+				     struct dma_fence **fence);
+static int amdgpu_vce_get_destroy_msg(struct amdgpu_ring *ring, uint32_t handle,
+				      bool direct, struct dma_fence **fence);
 
 /**
  * amdgpu_vce_init - allocate memory, load vce firmware
@@ -211,6 +216,7 @@ int amdgpu_vce_sw_fini(struct amdgpu_device *adev)
 	if (adev->vce.vcpu_bo == NULL)
 		return 0;
 
+	cancel_delayed_work_sync(&adev->vce.idle_work);
 	drm_sched_entity_destroy(&adev->vce.entity);
 
 	amdgpu_bo_free_kernel(&adev->vce.vcpu_bo, &adev->vce.gpu_addr,
@@ -428,14 +434,15 @@ void amdgpu_vce_free_handles(struct amdgpu_device *adev, struct drm_file *filp)
  *
  * Open up a stream for HW test
  */
-int amdgpu_vce_get_create_msg(struct amdgpu_ring *ring, uint32_t handle,
-			      struct dma_fence **fence)
+static int amdgpu_vce_get_create_msg(struct amdgpu_ring *ring, uint32_t handle,
+				     struct amdgpu_bo *bo,
+				     struct dma_fence **fence)
 {
 	const unsigned ib_size_dw = 1024;
 	struct amdgpu_job *job;
 	struct amdgpu_ib *ib;
 	struct dma_fence *f = NULL;
-	uint64_t dummy;
+	uint64_t addr;
 	int i, r;
 
 	r = amdgpu_job_alloc_with_ib(ring->adev, ib_size_dw * 4, &job);
@@ -444,7 +451,7 @@ int amdgpu_vce_get_create_msg(struct amdgpu_ring *ring, uint32_t handle,
 
 	ib = &job->ibs[0];
 
-	dummy = ib->gpu_addr + 1024;
+	addr = amdgpu_bo_gpu_offset(bo);
 
 	/* stitch together an VCE create msg */
 	ib->length_dw = 0;
@@ -476,8 +483,8 @@ int amdgpu_vce_get_create_msg(struct amdgpu_ring *ring, uint32_t handle,
 
 	ib->ptr[ib->length_dw++] = 0x00000014; /* len */
 	ib->ptr[ib->length_dw++] = 0x05000005; /* feedback buffer */
-	ib->ptr[ib->length_dw++] = upper_32_bits(dummy);
-	ib->ptr[ib->length_dw++] = dummy;
+	ib->ptr[ib->length_dw++] = upper_32_bits(addr);
+	ib->ptr[ib->length_dw++] = addr;
 	ib->ptr[ib->length_dw++] = 0x00000001;
 
 	for (i = ib->length_dw; i < ib_size_dw; ++i)
@@ -507,8 +514,8 @@ err:
  *
  * Close up a stream for HW test or if userspace failed to do so
  */
-int amdgpu_vce_get_destroy_msg(struct amdgpu_ring *ring, uint32_t handle,
-			       bool direct, struct dma_fence **fence)
+static int amdgpu_vce_get_destroy_msg(struct amdgpu_ring *ring, uint32_t handle,
+				      bool direct, struct dma_fence **fence)
 {
 	const unsigned ib_size_dw = 1024;
 	struct amdgpu_job *job;
@@ -1110,13 +1117,20 @@ int amdgpu_vce_ring_test_ring(struct amdgpu_ring *ring)
 int amdgpu_vce_ring_test_ib(struct amdgpu_ring *ring, long timeout)
 {
 	struct dma_fence *fence = NULL;
+	struct amdgpu_bo *bo = NULL;
 	long r;
 
 	/* skip vce ring1/2 ib test for now, since it's not reliable */
 	if (ring != &ring->adev->vce.ring[0])
 		return 0;
 
-	r = amdgpu_vce_get_create_msg(ring, 1, NULL);
+	r = amdgpu_bo_create_reserved(ring->adev, 512, PAGE_SIZE,
+				      AMDGPU_GEM_DOMAIN_VRAM,
+				      &bo, NULL, NULL);
+	if (r)
+		return r;
+
+	r = amdgpu_vce_get_create_msg(ring, 1, bo, NULL);
 	if (r)
 		goto error;
 
@@ -1132,5 +1146,7 @@ int amdgpu_vce_ring_test_ib(struct amdgpu_ring *ring, long timeout)
 
 error:
 	dma_fence_put(fence);
+	amdgpu_bo_unreserve(bo);
+	amdgpu_bo_unref(&bo);
 	return r;
 }
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.h
index 30ea54dd9117..d6d83a3ec803 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.h
@@ -58,10 +58,6 @@ int amdgpu_vce_sw_fini(struct amdgpu_device *adev);
 int amdgpu_vce_entity_init(struct amdgpu_device *adev);
 int amdgpu_vce_suspend(struct amdgpu_device *adev);
 int amdgpu_vce_resume(struct amdgpu_device *adev);
-int amdgpu_vce_get_create_msg(struct amdgpu_ring *ring, uint32_t handle,
-			      struct dma_fence **fence);
-int amdgpu_vce_get_destroy_msg(struct amdgpu_ring *ring, uint32_t handle,
-			       bool direct, struct dma_fence **fence);
 void amdgpu_vce_free_handles(struct amdgpu_device *adev, struct drm_file *filp);
 int amdgpu_vce_ring_parse_cs(struct amdgpu_cs_parser *p, uint32_t ib_idx);
 int amdgpu_vce_ring_parse_cs_vm(struct amdgpu_cs_parser *p, uint32_t ib_idx);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
index 7a6beb2e7c4e..9d870444d7d6 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
@@ -193,6 +193,8 @@ int amdgpu_vcn_sw_fini(struct amdgpu_device *adev)
 {
 	int i, j;
 
+	cancel_delayed_work_sync(&adev->vcn.idle_work);
+
 	if (adev->vcn.indirect_sram) {
 		amdgpu_bo_free_kernel(&adev->vcn.dpg_sram_bo,
 				      &adev->vcn.dpg_sram_gpu_addr,
@@ -569,13 +571,14 @@ int amdgpu_vcn_enc_ring_test_ring(struct amdgpu_ring *ring)
 }
 
 static int amdgpu_vcn_enc_get_create_msg(struct amdgpu_ring *ring, uint32_t handle,
-			      struct dma_fence **fence)
+					 struct amdgpu_bo *bo,
+					 struct dma_fence **fence)
 {
 	const unsigned ib_size_dw = 16;
 	struct amdgpu_job *job;
 	struct amdgpu_ib *ib;
 	struct dma_fence *f = NULL;
-	uint64_t dummy;
+	uint64_t addr;
 	int i, r;
 
 	r = amdgpu_job_alloc_with_ib(ring->adev, ib_size_dw * 4, &job);
@@ -583,14 +586,14 @@ static int amdgpu_vcn_enc_get_create_msg(struct amdgpu_ring *ring, uint32_t hand
 		return r;
 
 	ib = &job->ibs[0];
-	dummy = ib->gpu_addr + 1024;
+	addr = amdgpu_bo_gpu_offset(bo);
 
 	ib->length_dw = 0;
 	ib->ptr[ib->length_dw++] = 0x00000018;
 	ib->ptr[ib->length_dw++] = 0x00000001; /* session info */
 	ib->ptr[ib->length_dw++] = handle;
-	ib->ptr[ib->length_dw++] = upper_32_bits(dummy);
-	ib->ptr[ib->length_dw++] = dummy;
+	ib->ptr[ib->length_dw++] = upper_32_bits(addr);
+	ib->ptr[ib->length_dw++] = addr;
 	ib->ptr[ib->length_dw++] = 0x0000000b;
 
 	ib->ptr[ib->length_dw++] = 0x00000014;
@@ -621,13 +624,14 @@ err:
 }
 
 static int amdgpu_vcn_enc_get_destroy_msg(struct amdgpu_ring *ring, uint32_t handle,
-				struct dma_fence **fence)
+					  struct amdgpu_bo *bo,
+					  struct dma_fence **fence)
 {
 	const unsigned ib_size_dw = 16;
 	struct amdgpu_job *job;
 	struct amdgpu_ib *ib;
 	struct dma_fence *f = NULL;
-	uint64_t dummy;
+	uint64_t addr;
 	int i, r;
 
 	r = amdgpu_job_alloc_with_ib(ring->adev, ib_size_dw * 4, &job);
@@ -635,14 +639,14 @@ static int amdgpu_vcn_enc_get_destroy_msg(struct amdgpu_ring *ring, uint32_t han
 		return r;
 
 	ib = &job->ibs[0];
-	dummy = ib->gpu_addr + 1024;
+	addr = amdgpu_bo_gpu_offset(bo);
 
 	ib->length_dw = 0;
 	ib->ptr[ib->length_dw++] = 0x00000018;
 	ib->ptr[ib->length_dw++] = 0x00000001;
 	ib->ptr[ib->length_dw++] = handle;
-	ib->ptr[ib->length_dw++] = upper_32_bits(dummy);
-	ib->ptr[ib->length_dw++] = dummy;
+	ib->ptr[ib->length_dw++] = upper_32_bits(addr);
+	ib->ptr[ib->length_dw++] = addr;
 	ib->ptr[ib->length_dw++] = 0x0000000b;
 
 	ib->ptr[ib->length_dw++] = 0x00000014;
@@ -675,13 +679,20 @@ err:
 int amdgpu_vcn_enc_ring_test_ib(struct amdgpu_ring *ring, long timeout)
 {
 	struct dma_fence *fence = NULL;
+	struct amdgpu_bo *bo = NULL;
 	long r;
 
-	r = amdgpu_vcn_enc_get_create_msg(ring, 1, NULL);
+	r = amdgpu_bo_create_reserved(ring->adev, 128 * 1024, PAGE_SIZE,
+				      AMDGPU_GEM_DOMAIN_VRAM,
+				      &bo, NULL, NULL);
+	if (r)
+		return r;
+
+	r = amdgpu_vcn_enc_get_create_msg(ring, 1, bo, NULL);
 	if (r)
 		goto error;
 
-	r = amdgpu_vcn_enc_get_destroy_msg(ring, 1, &fence);
+	r = amdgpu_vcn_enc_get_destroy_msg(ring, 1, bo, &fence);
 	if (r)
 		goto error;
 
@@ -693,6 +704,8 @@ int amdgpu_vcn_enc_ring_test_ib(struct amdgpu_ring *ring, long timeout)
 
 error:
 	dma_fence_put(fence);
+	amdgpu_bo_unreserve(bo);
+	amdgpu_bo_unref(&bo);
 	return r;
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
index e2fb141ff2e5..598c24505c73 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
@@ -130,7 +130,8 @@ static unsigned amdgpu_vm_num_entries(struct amdgpu_device *adev,
 
 	if (level == adev->vm_manager.root_level)
 		/* For the root directory */
-		return round_up(adev->vm_manager.max_pfn, 1ULL << shift) >> shift;
+		return round_up(adev->vm_manager.max_pfn, 1ULL << shift)
+			>> shift;
 	else if (level != AMDGPU_VM_PTB)
 		/* Everything in between */
 		return 512;
@@ -341,7 +342,7 @@ static struct amdgpu_vm_pt *amdgpu_vm_pt_parent(struct amdgpu_vm_pt *pt)
 	return container_of(parent->vm_bo, struct amdgpu_vm_pt, base);
 }
 
-/**
+/*
  * amdgpu_vm_pt_cursor - state for for_each_amdgpu_vm_pt
  */
 struct amdgpu_vm_pt_cursor {
@@ -482,6 +483,7 @@ static void amdgpu_vm_pt_next(struct amdgpu_device *adev,
  *
  * @adev: amdgpu_device structure
  * @vm: amdgpu_vm structure
+ * @start: optional cursor to start with
  * @cursor: state to initialize
  *
  * Starts a deep first traversal of the PD/PT tree.
@@ -535,7 +537,7 @@ static void amdgpu_vm_pt_next_dfs(struct amdgpu_device *adev,
 		amdgpu_vm_pt_ancestor(cursor);
 }
 
-/**
+/*
  * for_each_amdgpu_vm_pt_dfs_safe - safe deep first search of all PDs/PTs
  */
 #define for_each_amdgpu_vm_pt_dfs_safe(adev, vm, start, cursor, entry)		\
@@ -566,6 +568,14 @@ void amdgpu_vm_get_pd_bo(struct amdgpu_vm *vm,
 	list_add(&entry->tv.head, validated);
 }
 
+/**
+ * amdgpu_vm_del_from_lru_notify - update bulk_moveable flag
+ *
+ * @bo: BO which was removed from the LRU
+ *
+ * Make sure the bulk_moveable flag is updated when a BO is removed from the
+ * LRU.
+ */
 void amdgpu_vm_del_from_lru_notify(struct ttm_buffer_object *bo)
 {
 	struct amdgpu_bo *abo;
@@ -600,21 +610,18 @@ void amdgpu_vm_del_from_lru_notify(struct ttm_buffer_object *bo)
 void amdgpu_vm_move_to_lru_tail(struct amdgpu_device *adev,
 				struct amdgpu_vm *vm)
 {
-	struct ttm_bo_global *glob = adev->mman.bdev.glob;
 	struct amdgpu_vm_bo_base *bo_base;
 
-#if 0
 	if (vm->bulk_moveable) {
-		spin_lock(&glob->lru_lock);
+		spin_lock(&ttm_bo_glob.lru_lock);
 		ttm_bo_bulk_move_lru_tail(&vm->lru_bulk_move);
-		spin_unlock(&glob->lru_lock);
+		spin_unlock(&ttm_bo_glob.lru_lock);
 		return;
 	}
-#endif
 
 	memset(&vm->lru_bulk_move, 0, sizeof(vm->lru_bulk_move));
 
-	spin_lock(&glob->lru_lock);
+	spin_lock(&ttm_bo_glob.lru_lock);
 	list_for_each_entry(bo_base, &vm->idle, vm_status) {
 		struct amdgpu_bo *bo = bo_base->bo;
 
@@ -626,7 +633,7 @@ void amdgpu_vm_move_to_lru_tail(struct amdgpu_device *adev,
 			ttm_bo_move_to_lru_tail(&bo->shadow->tbo,
 						&vm->lru_bulk_move);
 	}
-	spin_unlock(&glob->lru_lock);
+	spin_unlock(&ttm_bo_glob.lru_lock);
 
 	vm->bulk_moveable = true;
 }
@@ -695,6 +702,7 @@ bool amdgpu_vm_ready(struct amdgpu_vm *vm)
  * @adev: amdgpu_device pointer
  * @vm: VM to clear BO from
  * @bo: BO to clear
+ * @direct: use a direct update
  *
  * Root PD needs to be reserved when calling this.
  *
@@ -703,7 +711,8 @@ bool amdgpu_vm_ready(struct amdgpu_vm *vm)
  */
 static int amdgpu_vm_clear_bo(struct amdgpu_device *adev,
 			      struct amdgpu_vm *vm,
-			      struct amdgpu_bo *bo)
+			      struct amdgpu_bo *bo,
+			      bool direct)
 {
 	struct ttm_operation_ctx ctx = { true, false };
 	unsigned level = adev->vm_manager.root_level;
@@ -762,6 +771,7 @@ static int amdgpu_vm_clear_bo(struct amdgpu_device *adev,
 	memset(&params, 0, sizeof(params));
 	params.adev = adev;
 	params.vm = vm;
+	params.direct = direct;
 
 	r = vm->update_funcs->prepare(&params, AMDGPU_FENCE_OWNER_KFD, NULL);
 	if (r)
@@ -815,10 +825,13 @@ static int amdgpu_vm_clear_bo(struct amdgpu_device *adev,
  *
  * @adev: amdgpu_device pointer
  * @vm: requesting vm
+ * @level: the page table level
+ * @direct: use a direct update
  * @bp: resulting BO allocation parameters
  */
 static void amdgpu_vm_bo_param(struct amdgpu_device *adev, struct amdgpu_vm *vm,
-			       int level, struct amdgpu_bo_param *bp)
+			       int level, bool direct,
+			       struct amdgpu_bo_param *bp)
 {
 	memset(bp, 0, sizeof(*bp));
 
@@ -833,6 +846,7 @@ static void amdgpu_vm_bo_param(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 	else if (!vm->root.base.bo || vm->root.base.bo->shadow)
 		bp->flags |= AMDGPU_GEM_CREATE_SHADOW;
 	bp->type = ttm_bo_type_kernel;
+	bp->no_wait_gpu = direct;
 	if (vm->root.base.bo)
 		bp->resv = vm->root.base.bo->tbo.base.resv;
 }
@@ -843,6 +857,7 @@ static void amdgpu_vm_bo_param(struct amdgpu_device *adev, struct amdgpu_vm *vm,
  * @adev: amdgpu_device pointer
  * @vm: VM to allocate page tables for
  * @cursor: Which page table to allocate
+ * @direct: use a direct update
  *
  * Make sure a specific page table or directory is allocated.
  *
@@ -852,7 +867,8 @@ static void amdgpu_vm_bo_param(struct amdgpu_device *adev, struct amdgpu_vm *vm,
  */
 static int amdgpu_vm_alloc_pts(struct amdgpu_device *adev,
 			       struct amdgpu_vm *vm,
-			       struct amdgpu_vm_pt_cursor *cursor)
+			       struct amdgpu_vm_pt_cursor *cursor,
+			       bool direct)
 {
 	struct amdgpu_vm_pt *entry = cursor->entry;
 	struct amdgpu_bo_param bp;
@@ -873,7 +889,7 @@ static int amdgpu_vm_alloc_pts(struct amdgpu_device *adev,
 	if (entry->base.bo)
 		return 0;
 
-	amdgpu_vm_bo_param(adev, vm, cursor->level, &bp);
+	amdgpu_vm_bo_param(adev, vm, cursor->level, direct, &bp);
 
 	r = amdgpu_bo_create(adev, &bp, &pt);
 	if (r)
@@ -885,7 +901,7 @@ static int amdgpu_vm_alloc_pts(struct amdgpu_device *adev,
 	pt->parent = amdgpu_bo_ref(cursor->parent->base.bo);
 	amdgpu_vm_bo_base_init(&entry->base, vm, pt);
 
-	r = amdgpu_vm_clear_bo(adev, vm, pt);
+	r = amdgpu_vm_clear_bo(adev, vm, pt, direct);
 	if (r)
 		goto error_free_pt;
 
@@ -1022,7 +1038,8 @@ bool amdgpu_vm_need_pipeline_sync(struct amdgpu_ring *ring,
  * Returns:
  * 0 on success, errno otherwise.
  */
-int amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job, bool need_pipe_sync)
+int amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job,
+		    bool need_pipe_sync)
 {
 	struct amdgpu_device *adev = ring->adev;
 	unsigned vmhub = ring->funcs->vmhub;
@@ -1036,10 +1053,8 @@ int amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job, bool need_
 		id->oa_base != job->oa_base ||
 		id->oa_size != job->oa_size);
 	bool vm_flush_needed = job->vm_needs_flush;
-	bool pasid_mapping_needed = id->pasid != job->pasid ||
-		!id->pasid_mapping ||
-		!dma_fence_is_signaled(id->pasid_mapping);
 	struct dma_fence *fence = NULL;
+	bool pasid_mapping_needed = false;
 	unsigned patch_offset = 0;
 	int r;
 
@@ -1049,6 +1064,12 @@ int amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job, bool need_
 		pasid_mapping_needed = true;
 	}
 
+	mutex_lock(&id_mgr->lock);
+	if (id->pasid != job->pasid || !id->pasid_mapping ||
+	    !dma_fence_is_signaled(id->pasid_mapping))
+		pasid_mapping_needed = true;
+	mutex_unlock(&id_mgr->lock);
+
 	gds_switch_needed &= !!ring->funcs->emit_gds_switch;
 	vm_flush_needed &= !!ring->funcs->emit_vm_flush  &&
 			job->vm_pd_addr != AMDGPU_BO_INVALID_OFFSET;
@@ -1088,9 +1109,11 @@ int amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job, bool need_
 	}
 
 	if (pasid_mapping_needed) {
+		mutex_lock(&id_mgr->lock);
 		id->pasid = job->pasid;
 		dma_fence_put(id->pasid_mapping);
 		id->pasid_mapping = dma_fence_get(fence);
+		mutex_unlock(&id_mgr->lock);
 	}
 	dma_fence_put(fence);
 
@@ -1174,10 +1197,10 @@ uint64_t amdgpu_vm_map_gart(const dma_addr_t *pages_addr, uint64_t addr)
 	return result;
 }
 
-/*
+/**
  * amdgpu_vm_update_pde - update a single level in the hierarchy
  *
- * @param: parameters for the update
+ * @params: parameters for the update
  * @vm: requested vm
  * @entry: entry to update
  *
@@ -1201,7 +1224,7 @@ static int amdgpu_vm_update_pde(struct amdgpu_vm_update_params *params,
 	return vm->update_funcs->update(params, bo, pde, pt, 1, 0, flags);
 }
 
-/*
+/**
  * amdgpu_vm_invalidate_pds - mark all PDs as invalid
  *
  * @adev: amdgpu_device pointer
@@ -1220,19 +1243,20 @@ static void amdgpu_vm_invalidate_pds(struct amdgpu_device *adev,
 			amdgpu_vm_bo_relocated(&entry->base);
 }
 
-/*
- * amdgpu_vm_update_directories - make sure that all directories are valid
+/**
+ * amdgpu_vm_update_pdes - make sure that all directories are valid
  *
  * @adev: amdgpu_device pointer
  * @vm: requested vm
+ * @direct: submit directly to the paging queue
  *
  * Makes sure all directories are up to date.
  *
  * Returns:
  * 0 for success, error for failure.
  */
-int amdgpu_vm_update_directories(struct amdgpu_device *adev,
-				 struct amdgpu_vm *vm)
+int amdgpu_vm_update_pdes(struct amdgpu_device *adev,
+			  struct amdgpu_vm *vm, bool direct)
 {
 	struct amdgpu_vm_update_params params;
 	int r;
@@ -1243,6 +1267,7 @@ int amdgpu_vm_update_directories(struct amdgpu_device *adev,
 	memset(&params, 0, sizeof(params));
 	params.adev = adev;
 	params.vm = vm;
+	params.direct = direct;
 
 	r = vm->update_funcs->prepare(&params, AMDGPU_FENCE_OWNER_VM, NULL);
 	if (r)
@@ -1270,7 +1295,7 @@ error:
 	return r;
 }
 
-/**
+/*
  * amdgpu_vm_update_flags - figure out flags for PTE updates
  *
  * Make sure to set the right flags for the PTEs at the desired level.
@@ -1393,7 +1418,11 @@ static int amdgpu_vm_update_ptes(struct amdgpu_vm_update_params *params,
 		uint64_t incr, entry_end, pe_start;
 		struct amdgpu_bo *pt;
 
-		r = amdgpu_vm_alloc_pts(params->adev, params->vm, &cursor);
+		/* make sure that the page tables covering the address range are
+		 * actually allocated
+		 */
+		r = amdgpu_vm_alloc_pts(params->adev, params->vm, &cursor,
+					params->direct);
 		if (r)
 			return r;
 
@@ -1465,7 +1494,12 @@ static int amdgpu_vm_update_ptes(struct amdgpu_vm_update_params *params,
 		} while (frag_start < entry_end);
 
 		if (amdgpu_vm_pt_descendant(adev, &cursor)) {
-			/* Free all child entries */
+			/* Free all child entries.
+			 * Update the tables with the flags and addresses and free up subsequent
+			 * tables in the case of huge pages or freed up areas.
+			 * This is the maximum you can free, because all other page tables are not
+			 * completely covered by the range and so potentially still in use.
+			 */
 			while (cursor.pfn < frag_start) {
 				amdgpu_vm_free_pts(adev, params->vm, &cursor);
 				amdgpu_vm_pt_next(adev, &cursor);
@@ -1484,13 +1518,14 @@ static int amdgpu_vm_update_ptes(struct amdgpu_vm_update_params *params,
  * amdgpu_vm_bo_update_mapping - update a mapping in the vm page table
  *
  * @adev: amdgpu_device pointer
- * @exclusive: fence we need to sync to
- * @pages_addr: DMA addresses to use for mapping
  * @vm: requested vm
+ * @direct: direct submission in a page fault
+ * @exclusive: fence we need to sync to
  * @start: start of mapped range
  * @last: last mapped entry
  * @flags: flags for the entries
  * @addr: addr to set the area to
+ * @pages_addr: DMA addresses to use for mapping
  * @fence: optional resulting fence
  *
  * Fill in the page table entries between @start and @last.
@@ -1499,11 +1534,11 @@ static int amdgpu_vm_update_ptes(struct amdgpu_vm_update_params *params,
  * 0 for success, -EINVAL for failure.
  */
 static int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev,
+				       struct amdgpu_vm *vm, bool direct,
 				       struct dma_fence *exclusive,
-				       dma_addr_t *pages_addr,
-				       struct amdgpu_vm *vm,
 				       uint64_t start, uint64_t last,
 				       uint64_t flags, uint64_t addr,
+				       dma_addr_t *pages_addr,
 				       struct dma_fence **fence)
 {
 	struct amdgpu_vm_update_params params;
@@ -1513,6 +1548,7 @@ static int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev,
 	memset(&params, 0, sizeof(params));
 	params.adev = adev;
 	params.vm = vm;
+	params.direct = direct;
 	params.pages_addr = pages_addr;
 
 	/* sync to everything except eviction fences on unmapping */
@@ -1571,27 +1607,8 @@ static int amdgpu_vm_bo_split_mapping(struct amdgpu_device *adev,
 	if (!(mapping->flags & AMDGPU_PTE_WRITEABLE))
 		flags &= ~AMDGPU_PTE_WRITEABLE;
 
-	flags &= ~AMDGPU_PTE_EXECUTABLE;
-	flags |= mapping->flags & AMDGPU_PTE_EXECUTABLE;
-
-	if (adev->asic_type >= CHIP_NAVI10) {
-		flags &= ~AMDGPU_PTE_MTYPE_NV10_MASK;
-		flags |= (mapping->flags & AMDGPU_PTE_MTYPE_NV10_MASK);
-	} else {
-		flags &= ~AMDGPU_PTE_MTYPE_VG10_MASK;
-		flags |= (mapping->flags & AMDGPU_PTE_MTYPE_VG10_MASK);
-	}
-
-	if ((mapping->flags & AMDGPU_PTE_PRT) &&
-	    (adev->asic_type >= CHIP_VEGA10)) {
-		flags |= AMDGPU_PTE_PRT;
-		if (adev->asic_type >= CHIP_NAVI10) {
-			flags |= AMDGPU_PTE_SNOOPED;
-			flags |= AMDGPU_PTE_LOG;
-			flags |= AMDGPU_PTE_SYSTEM;
-		}
-		flags &= ~AMDGPU_PTE_VALID;
-	}
+	/* Apply ASIC specific mapping flags */
+	amdgpu_gmc_get_vm_pte(adev, mapping, &flags);
 
 	trace_amdgpu_vm_bo_update(mapping);
 
@@ -1635,7 +1652,8 @@ static int amdgpu_vm_bo_split_mapping(struct amdgpu_device *adev,
 				dma_addr = pages_addr;
 			} else {
 				addr = pages_addr[pfn];
-				max_entries = count * AMDGPU_GPU_PAGES_IN_CPU_PAGE;
+				max_entries = count *
+					AMDGPU_GPU_PAGES_IN_CPU_PAGE;
 			}
 
 		} else if (flags & AMDGPU_PTE_VALID) {
@@ -1644,9 +1662,9 @@ static int amdgpu_vm_bo_split_mapping(struct amdgpu_device *adev,
 		}
 
 		last = min((uint64_t)mapping->last, start + max_entries - 1);
-		r = amdgpu_vm_bo_update_mapping(adev, exclusive, dma_addr, vm,
+		r = amdgpu_vm_bo_update_mapping(adev, vm, false, exclusive,
 						start, last, flags, addr,
-						fence);
+						dma_addr, fence);
 		if (r)
 			return r;
 
@@ -1674,8 +1692,7 @@ static int amdgpu_vm_bo_split_mapping(struct amdgpu_device *adev,
  * Returns:
  * 0 for success, -EINVAL for failure.
  */
-int amdgpu_vm_bo_update(struct amdgpu_device *adev,
-			struct amdgpu_bo_va *bo_va,
+int amdgpu_vm_bo_update(struct amdgpu_device *adev, struct amdgpu_bo_va *bo_va,
 			bool clear)
 {
 	struct amdgpu_bo *bo = bo_va->base.bo;
@@ -1702,7 +1719,7 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev,
 			ttm = container_of(bo->tbo.ttm, struct ttm_dma_tt, ttm);
 			pages_addr = ttm->dma_address;
 		}
-		exclusive = dma_resv_get_excl(bo->tbo.base.resv);
+		exclusive = bo->tbo.moving;
 	}
 
 	if (bo) {
@@ -1733,12 +1750,6 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev,
 			return r;
 	}
 
-	if (vm->use_cpu_for_update) {
-		/* Flush HDP */
-		mb();
-		amdgpu_asic_flush_hdp(adev, NULL);
-	}
-
 	/* If the BO is not in its preferred location add it back to
 	 * the evicted list so that it gets validated again on the
 	 * next command submission.
@@ -1746,7 +1757,8 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev,
 	if (bo && bo->tbo.base.resv == vm->root.base.bo->tbo.base.resv) {
 		uint32_t mem_type = bo->tbo.mem.mem_type;
 
-		if (!(bo->preferred_domains & amdgpu_mem_type_to_domain(mem_type)))
+		if (!(bo->preferred_domains &
+		      amdgpu_mem_type_to_domain(mem_type)))
 			amdgpu_vm_bo_evicted(&bo_va->base);
 		else
 			amdgpu_vm_bo_idle(&bo_va->base);
@@ -1940,9 +1952,9 @@ int amdgpu_vm_clear_freed(struct amdgpu_device *adev,
 		    mapping->start < AMDGPU_GMC_HOLE_START)
 			init_pte_value = AMDGPU_PTE_DEFAULT_ATC;
 
-		r = amdgpu_vm_bo_update_mapping(adev, NULL, NULL, vm,
+		r = amdgpu_vm_bo_update_mapping(adev, vm, false, NULL,
 						mapping->start, mapping->last,
-						init_pte_value, 0, &f);
+						init_pte_value, 0, NULL, &f);
 		amdgpu_vm_free_mapping(adev, vm, mapping, f);
 		if (r) {
 			dma_fence_put(f);
@@ -2684,12 +2696,17 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 	spin_lock_init(&vm->invalidated_lock);
 	INIT_LIST_HEAD(&vm->freed);
 
-	/* create scheduler entity for page table updates */
-	r = drm_sched_entity_init(&vm->entity, adev->vm_manager.vm_pte_rqs,
+	/* create scheduler entities for page table updates */
+	r = drm_sched_entity_init(&vm->direct, adev->vm_manager.vm_pte_rqs,
 				  adev->vm_manager.vm_pte_num_rqs, NULL);
 	if (r)
 		return r;
 
+	r = drm_sched_entity_init(&vm->delayed, adev->vm_manager.vm_pte_rqs,
+				  adev->vm_manager.vm_pte_num_rqs, NULL);
+	if (r)
+		goto error_free_direct;
+
 	vm->pte_support_ats = false;
 
 	if (vm_context == AMDGPU_VM_CONTEXT_COMPUTE) {
@@ -2704,7 +2721,8 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 	}
 	DRM_DEBUG_DRIVER("VM update mode is %s\n",
 			 vm->use_cpu_for_update ? "CPU" : "SDMA");
-	WARN_ONCE((vm->use_cpu_for_update && !amdgpu_gmc_vram_full_visible(&adev->gmc)),
+	WARN_ONCE((vm->use_cpu_for_update &&
+		   !amdgpu_gmc_vram_full_visible(&adev->gmc)),
 		  "CPU update of VM recommended only for large BAR system\n");
 
 	if (vm->use_cpu_for_update)
@@ -2713,12 +2731,12 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 		vm->update_funcs = &amdgpu_vm_sdma_funcs;
 	vm->last_update = NULL;
 
-	amdgpu_vm_bo_param(adev, vm, adev->vm_manager.root_level, &bp);
+	amdgpu_vm_bo_param(adev, vm, adev->vm_manager.root_level, false, &bp);
 	if (vm_context == AMDGPU_VM_CONTEXT_COMPUTE)
 		bp.flags &= ~AMDGPU_GEM_CREATE_SHADOW;
 	r = amdgpu_bo_create(adev, &bp, &root);
 	if (r)
-		goto error_free_sched_entity;
+		goto error_free_delayed;
 
 	r = amdgpu_bo_reserve(root, true);
 	if (r)
@@ -2730,7 +2748,7 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 
 	amdgpu_vm_bo_base_init(&vm->root.base, vm, root);
 
-	r = amdgpu_vm_clear_bo(adev, vm, root);
+	r = amdgpu_vm_clear_bo(adev, vm, root, false);
 	if (r)
 		goto error_unreserve;
 
@@ -2761,8 +2779,11 @@ error_free_root:
 	amdgpu_bo_unref(&vm->root.base.bo);
 	vm->root.base.bo = NULL;
 
-error_free_sched_entity:
-	drm_sched_entity_destroy(&vm->entity);
+error_free_delayed:
+	drm_sched_entity_destroy(&vm->delayed);
+
+error_free_direct:
+	drm_sched_entity_destroy(&vm->direct);
 
 	return r;
 }
@@ -2803,6 +2824,7 @@ static int amdgpu_vm_check_clean_reserved(struct amdgpu_device *adev,
  *
  * @adev: amdgpu_device pointer
  * @vm: requested vm
+ * @pasid: pasid to use
  *
  * This only works on GFX VMs that don't have any BOs added and no
  * page tables allocated yet.
@@ -2818,7 +2840,8 @@ static int amdgpu_vm_check_clean_reserved(struct amdgpu_device *adev,
  * Returns:
  * 0 for success, -errno for errors.
  */
-int amdgpu_vm_make_compute(struct amdgpu_device *adev, struct amdgpu_vm *vm, unsigned int pasid)
+int amdgpu_vm_make_compute(struct amdgpu_device *adev, struct amdgpu_vm *vm,
+			   unsigned int pasid)
 {
 	bool pte_support_ats = (adev->asic_type == CHIP_RAVEN);
 	int r;
@@ -2850,7 +2873,7 @@ int amdgpu_vm_make_compute(struct amdgpu_device *adev, struct amdgpu_vm *vm, uns
 	 */
 	if (pte_support_ats != vm->pte_support_ats) {
 		vm->pte_support_ats = pte_support_ats;
-		r = amdgpu_vm_clear_bo(adev, vm, vm->root.base.bo);
+		r = amdgpu_vm_clear_bo(adev, vm, vm->root.base.bo, false);
 		if (r)
 			goto free_idr;
 	}
@@ -2860,7 +2883,8 @@ int amdgpu_vm_make_compute(struct amdgpu_device *adev, struct amdgpu_vm *vm, uns
 				    AMDGPU_VM_USE_CPU_FOR_COMPUTE);
 	DRM_DEBUG_DRIVER("VM update mode is %s\n",
 			 vm->use_cpu_for_update ? "CPU" : "SDMA");
-	WARN_ONCE((vm->use_cpu_for_update && !amdgpu_gmc_vram_full_visible(&adev->gmc)),
+	WARN_ONCE((vm->use_cpu_for_update &&
+		   !amdgpu_gmc_vram_full_visible(&adev->gmc)),
 		  "CPU update of VM recommended only for large BAR system\n");
 
 	if (vm->use_cpu_for_update)
@@ -2939,19 +2963,38 @@ void amdgpu_vm_fini(struct amdgpu_device *adev, struct amdgpu_vm *vm)
 	struct amdgpu_bo_va_mapping *mapping, *tmp;
 	bool prt_fini_needed = !!adev->gmc.gmc_funcs->set_prt;
 	struct amdgpu_bo *root;
-	int i, r;
+	int i;
 
 	amdgpu_amdkfd_gpuvm_destroy_cb(adev, vm);
 
+	root = amdgpu_bo_ref(vm->root.base.bo);
+	amdgpu_bo_reserve(root, true);
 	if (vm->pasid) {
 		unsigned long flags;
 
 		spin_lock_irqsave(&adev->vm_manager.pasid_lock, flags);
 		idr_remove(&adev->vm_manager.pasid_idr, vm->pasid);
 		spin_unlock_irqrestore(&adev->vm_manager.pasid_lock, flags);
+		vm->pasid = 0;
 	}
 
-	drm_sched_entity_destroy(&vm->entity);
+	list_for_each_entry_safe(mapping, tmp, &vm->freed, list) {
+		if (mapping->flags & AMDGPU_PTE_PRT && prt_fini_needed) {
+			amdgpu_vm_prt_fini(adev, vm);
+			prt_fini_needed = false;
+		}
+
+		list_del(&mapping->list);
+		amdgpu_vm_free_mapping(adev, vm, mapping, NULL);
+	}
+
+	amdgpu_vm_free_pts(adev, vm, NULL);
+	amdgpu_bo_unreserve(root);
+	amdgpu_bo_unref(&root);
+	WARN_ON(vm->root.base.bo);
+
+	drm_sched_entity_destroy(&vm->direct);
+	drm_sched_entity_destroy(&vm->delayed);
 
 	if (!RB_EMPTY_ROOT(&vm->va.rb_root)) {
 		dev_err(adev->dev, "still active bo inside vm\n");
@@ -2964,26 +3007,7 @@ void amdgpu_vm_fini(struct amdgpu_device *adev, struct amdgpu_vm *vm)
 		list_del(&mapping->list);
 		kfree(mapping);
 	}
-	list_for_each_entry_safe(mapping, tmp, &vm->freed, list) {
-		if (mapping->flags & AMDGPU_PTE_PRT && prt_fini_needed) {
-			amdgpu_vm_prt_fini(adev, vm);
-			prt_fini_needed = false;
-		}
-
-		list_del(&mapping->list);
-		amdgpu_vm_free_mapping(adev, vm, mapping, NULL);
-	}
 
-	root = amdgpu_bo_ref(vm->root.base.bo);
-	r = amdgpu_bo_reserve(root, true);
-	if (r) {
-		dev_err(adev->dev, "Leaking page tables because BO reservation failed\n");
-	} else {
-		amdgpu_vm_free_pts(adev, vm, NULL);
-		amdgpu_bo_unreserve(root);
-	}
-	amdgpu_bo_unref(&root);
-	WARN_ON(vm->root.base.bo);
 	dma_fence_put(vm->last_update);
 	for (i = 0; i < AMDGPU_MAX_VMHUBS; i++)
 		amdgpu_vmid_free_reserved(adev, vm, i);
@@ -3067,8 +3091,9 @@ int amdgpu_vm_ioctl(struct drm_device *dev, void *data, struct drm_file *filp)
 
 	switch (args->in.op) {
 	case AMDGPU_VM_OP_RESERVE_VMID:
-		/* current, we only have requirement to reserve vmid from gfxhub */
-		r = amdgpu_vmid_alloc_reserved(adev, &fpriv->vm, AMDGPU_GFXHUB_0);
+		/* We only have requirement to reserve vmid from gfxhub */
+		r = amdgpu_vmid_alloc_reserved(adev, &fpriv->vm,
+					       AMDGPU_GFXHUB_0);
 		if (r)
 			return r;
 		break;
@@ -3111,13 +3136,88 @@ void amdgpu_vm_get_task_info(struct amdgpu_device *adev, unsigned int pasid,
  */
 void amdgpu_vm_set_task_info(struct amdgpu_vm *vm)
 {
-	if (!vm->task_info.pid) {
-		vm->task_info.pid = current->pid;
-		get_task_comm(vm->task_info.task_name, current);
+	if (vm->task_info.pid)
+		return;
 
-		if (current->group_leader->mm == current->mm) {
-			vm->task_info.tgid = current->group_leader->pid;
-			get_task_comm(vm->task_info.process_name, current->group_leader);
-		}
+	vm->task_info.pid = current->pid;
+	get_task_comm(vm->task_info.task_name, current);
+
+	if (current->group_leader->mm != current->mm)
+		return;
+
+	vm->task_info.tgid = current->group_leader->pid;
+	get_task_comm(vm->task_info.process_name, current->group_leader);
+}
+
+/**
+ * amdgpu_vm_handle_fault - graceful handling of VM faults.
+ * @adev: amdgpu device pointer
+ * @pasid: PASID of the VM
+ * @addr: Address of the fault
+ *
+ * Try to gracefully handle a VM fault. Return true if the fault was handled and
+ * shouldn't be reported any more.
+ */
+bool amdgpu_vm_handle_fault(struct amdgpu_device *adev, unsigned int pasid,
+			    uint64_t addr)
+{
+	struct amdgpu_bo *root;
+	uint64_t value, flags;
+	struct amdgpu_vm *vm;
+	long r;
+
+	spin_lock(&adev->vm_manager.pasid_lock);
+	vm = idr_find(&adev->vm_manager.pasid_idr, pasid);
+	if (vm)
+		root = amdgpu_bo_ref(vm->root.base.bo);
+	else
+		root = NULL;
+	spin_unlock(&adev->vm_manager.pasid_lock);
+
+	if (!root)
+		return false;
+
+	r = amdgpu_bo_reserve(root, true);
+	if (r)
+		goto error_unref;
+
+	/* Double check that the VM still exists */
+	spin_lock(&adev->vm_manager.pasid_lock);
+	vm = idr_find(&adev->vm_manager.pasid_idr, pasid);
+	if (vm && vm->root.base.bo != root)
+		vm = NULL;
+	spin_unlock(&adev->vm_manager.pasid_lock);
+	if (!vm)
+		goto error_unlock;
+
+	addr /= AMDGPU_GPU_PAGE_SIZE;
+	flags = AMDGPU_PTE_VALID | AMDGPU_PTE_SNOOPED |
+		AMDGPU_PTE_SYSTEM;
+
+	if (amdgpu_vm_fault_stop == AMDGPU_VM_FAULT_STOP_NEVER) {
+		/* Redirect the access to the dummy page */
+		value = adev->dummy_page_addr;
+		flags |= AMDGPU_PTE_EXECUTABLE | AMDGPU_PTE_READABLE |
+			AMDGPU_PTE_WRITEABLE;
+	} else {
+		/* Let the hw retry silently on the PTE */
+		value = 0;
 	}
+
+	r = amdgpu_vm_bo_update_mapping(adev, vm, true, NULL, addr, addr + 1,
+					flags, value, NULL, NULL);
+	if (r)
+		goto error_unlock;
+
+	r = amdgpu_vm_update_pdes(adev, vm, true);
+
+error_unlock:
+	amdgpu_bo_unreserve(root);
+	if (r < 0)
+		DRM_ERROR("Can't handle page fault (%ld)\n", r);
+
+error_unref:
+	amdgpu_bo_unref(&root);
+
+	return false;
 }
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
index 2eda3a8c330d..4dbbe1b6b413 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
@@ -99,6 +99,9 @@ struct amdgpu_bo_list_entry;
 #define AMDGPU_VM_FAULT_STOP_FIRST	1
 #define AMDGPU_VM_FAULT_STOP_ALWAYS	2
 
+/* Reserve 4MB VRAM for page tables */
+#define AMDGPU_VM_RESERVED_VRAM		(4ULL << 20)
+
 /* max number of VMHUB */
 #define AMDGPU_MAX_VMHUBS			3
 #define AMDGPU_GFXHUB_0				0
@@ -199,6 +202,11 @@ struct amdgpu_vm_update_params {
 	struct amdgpu_vm *vm;
 
 	/**
+	 * @direct: if changes should be made directly
+	 */
+	bool direct;
+
+	/**
 	 * @pages_addr:
 	 *
 	 * DMA addresses to use for mapping
@@ -254,8 +262,9 @@ struct amdgpu_vm {
 	struct amdgpu_vm_pt     root;
 	struct dma_fence	*last_update;
 
-	/* Scheduler entity for page table updates */
-	struct drm_sched_entity	entity;
+	/* Scheduler entities for page table updates */
+	struct drm_sched_entity	direct;
+	struct drm_sched_entity	delayed;
 
 	unsigned int		pasid;
 	/* dedicated to vm */
@@ -357,8 +366,8 @@ int amdgpu_vm_validate_pt_bos(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 			      int (*callback)(void *p, struct amdgpu_bo *bo),
 			      void *param);
 int amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job, bool need_pipe_sync);
-int amdgpu_vm_update_directories(struct amdgpu_device *adev,
-				 struct amdgpu_vm *vm);
+int amdgpu_vm_update_pdes(struct amdgpu_device *adev,
+			  struct amdgpu_vm *vm, bool direct);
 int amdgpu_vm_clear_freed(struct amdgpu_device *adev,
 			  struct amdgpu_vm *vm,
 			  struct dma_fence **fence);
@@ -404,6 +413,8 @@ void amdgpu_vm_check_compute_bug(struct amdgpu_device *adev);
 
 void amdgpu_vm_get_task_info(struct amdgpu_device *adev, unsigned int pasid,
 			     struct amdgpu_task_info *task_info);
+bool amdgpu_vm_handle_fault(struct amdgpu_device *adev, unsigned int pasid,
+			    uint64_t addr);
 
 void amdgpu_vm_set_task_info(struct amdgpu_vm *vm);
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_cpu.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_cpu.c
index 5222d165abfc..73fec7a0ced5 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_cpu.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_cpu.c
@@ -49,13 +49,6 @@ static int amdgpu_vm_cpu_prepare(struct amdgpu_vm_update_params *p, void *owner,
 {
 	int r;
 
-	/* Wait for PT BOs to be idle. PTs share the same resv. object
-	 * as the root PD BO
-	 */
-	r = amdgpu_bo_sync_wait(p->vm->root.base.bo, owner, true);
-	if (unlikely(r))
-		return r;
-
 	/* Wait for any BO move to be completed */
 	if (exclusive) {
 		r = dma_fence_wait(exclusive, true);
@@ -63,7 +56,14 @@ static int amdgpu_vm_cpu_prepare(struct amdgpu_vm_update_params *p, void *owner,
 			return r;
 	}
 
-	return 0;
+	/* Don't wait for submissions during page fault */
+	if (p->direct)
+		return 0;
+
+	/* Wait for PT BOs to be idle. PTs share the same resv. object
+	 * as the root PD BO
+	 */
+	return amdgpu_bo_sync_wait(p->vm->root.base.bo, owner, true);
 }
 
 /**
@@ -89,7 +89,7 @@ static int amdgpu_vm_cpu_update(struct amdgpu_vm_update_params *p,
 
 	pe += (unsigned long)amdgpu_bo_kptr(bo);
 
-	trace_amdgpu_vm_set_ptes(pe, addr, count, incr, flags);
+	trace_amdgpu_vm_set_ptes(pe, addr, count, incr, flags, p->direct);
 
 	for (i = 0; i < count; i++) {
 		value = p->pages_addr ?
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_sdma.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_sdma.c
index 61fc584cbb1a..832db59f441e 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_sdma.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_sdma.c
@@ -68,17 +68,19 @@ static int amdgpu_vm_sdma_prepare(struct amdgpu_vm_update_params *p,
 	if (r)
 		return r;
 
+	p->num_dw_left = ndw;
+
+	/* Wait for moves to be completed */
 	r = amdgpu_sync_fence(p->adev, &p->job->sync, exclusive, false);
 	if (r)
 		return r;
 
-	r = amdgpu_sync_resv(p->adev, &p->job->sync, root->tbo.base.resv,
-			     owner, false);
-	if (r)
-		return r;
+	/* Don't wait for any submissions during page fault handling */
+	if (p->direct)
+		return 0;
 
-	p->num_dw_left = ndw;
-	return 0;
+	return amdgpu_sync_resv(p->adev, &p->job->sync, root->tbo.base.resv,
+				owner, false);
 }
 
 /**
@@ -95,22 +97,23 @@ static int amdgpu_vm_sdma_commit(struct amdgpu_vm_update_params *p,
 {
 	struct amdgpu_bo *root = p->vm->root.base.bo;
 	struct amdgpu_ib *ib = p->job->ibs;
+	struct drm_sched_entity *entity;
 	struct amdgpu_ring *ring;
 	struct dma_fence *f;
 	int r;
 
-	ring = container_of(p->vm->entity.rq->sched, struct amdgpu_ring, sched);
+	entity = p->direct ? &p->vm->direct : &p->vm->delayed;
+	ring = container_of(entity->rq->sched, struct amdgpu_ring, sched);
 
 	WARN_ON(ib->length_dw == 0);
 	amdgpu_ring_pad_ib(ring, ib);
 	WARN_ON(ib->length_dw > p->num_dw_left);
-	r = amdgpu_job_submit(p->job, &p->vm->entity,
-			      AMDGPU_FENCE_OWNER_VM, &f);
+	r = amdgpu_job_submit(p->job, entity, AMDGPU_FENCE_OWNER_VM, &f);
 	if (r)
 		goto error;
 
 	amdgpu_bo_fence(root, f, true);
-	if (fence)
+	if (fence && !p->direct)
 		swap(*fence, f);
 	dma_fence_put(f);
 	return 0;
@@ -120,7 +123,6 @@ error:
 	return r;
 }
 
-
 /**
  * amdgpu_vm_sdma_copy_ptes - copy the PTEs from mapping
  *
@@ -141,7 +143,7 @@ static void amdgpu_vm_sdma_copy_ptes(struct amdgpu_vm_update_params *p,
 	src += p->num_dw_left * 4;
 
 	pe += amdgpu_bo_gpu_offset(bo);
-	trace_amdgpu_vm_copy_ptes(pe, src, count);
+	trace_amdgpu_vm_copy_ptes(pe, src, count, p->direct);
 
 	amdgpu_vm_copy_pte(p->adev, ib, pe, src, count);
 }
@@ -168,7 +170,7 @@ static void amdgpu_vm_sdma_set_ptes(struct amdgpu_vm_update_params *p,
 	struct amdgpu_ib *ib = p->job->ibs;
 
 	pe += amdgpu_bo_gpu_offset(bo);
-	trace_amdgpu_vm_set_ptes(pe, addr, count, incr, flags);
+	trace_amdgpu_vm_set_ptes(pe, addr, count, incr, flags, p->direct);
 	if (count < 3) {
 		amdgpu_vm_write_pte(p->adev, ib, pe, addr | flags,
 				    count, incr);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c
index 3a9d8c15fe9f..82a3299e53c0 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c
@@ -23,6 +23,9 @@
  */
 
 #include "amdgpu.h"
+#include "amdgpu_vm.h"
+#include "amdgpu_atomfirmware.h"
+#include "atom.h"
 
 struct amdgpu_vram_mgr {
 	struct drm_mm mm;
@@ -101,6 +104,39 @@ static ssize_t amdgpu_mem_info_vis_vram_used_show(struct device *dev,
 		amdgpu_vram_mgr_vis_usage(&adev->mman.bdev.man[TTM_PL_VRAM]));
 }
 
+static ssize_t amdgpu_mem_info_vram_vendor(struct device *dev,
+						 struct device_attribute *attr,
+						 char *buf)
+{
+	struct drm_device *ddev = dev_get_drvdata(dev);
+	struct amdgpu_device *adev = ddev->dev_private;
+
+	switch (adev->gmc.vram_vendor) {
+	case SAMSUNG:
+		return snprintf(buf, PAGE_SIZE, "samsung\n");
+	case INFINEON:
+		return snprintf(buf, PAGE_SIZE, "infineon\n");
+	case ELPIDA:
+		return snprintf(buf, PAGE_SIZE, "elpida\n");
+	case ETRON:
+		return snprintf(buf, PAGE_SIZE, "etron\n");
+	case NANYA:
+		return snprintf(buf, PAGE_SIZE, "nanya\n");
+	case HYNIX:
+		return snprintf(buf, PAGE_SIZE, "hynix\n");
+	case MOSEL:
+		return snprintf(buf, PAGE_SIZE, "mosel\n");
+	case WINBOND:
+		return snprintf(buf, PAGE_SIZE, "winbond\n");
+	case ESMT:
+		return snprintf(buf, PAGE_SIZE, "esmt\n");
+	case MICRON:
+		return snprintf(buf, PAGE_SIZE, "micron\n");
+	default:
+		return snprintf(buf, PAGE_SIZE, "unknown\n");
+	}
+}
+
 static DEVICE_ATTR(mem_info_vram_total, S_IRUGO,
 		   amdgpu_mem_info_vram_total_show, NULL);
 static DEVICE_ATTR(mem_info_vis_vram_total, S_IRUGO,
@@ -109,6 +145,8 @@ static DEVICE_ATTR(mem_info_vram_used, S_IRUGO,
 		   amdgpu_mem_info_vram_used_show, NULL);
 static DEVICE_ATTR(mem_info_vis_vram_used, S_IRUGO,
 		   amdgpu_mem_info_vis_vram_used_show, NULL);
+static DEVICE_ATTR(mem_info_vram_vendor, S_IRUGO,
+		   amdgpu_mem_info_vram_vendor, NULL);
 
 /**
  * amdgpu_vram_mgr_init - init VRAM manager and DRM MM
@@ -154,6 +192,11 @@ static int amdgpu_vram_mgr_init(struct ttm_mem_type_manager *man,
 		DRM_ERROR("Failed to create device file mem_info_vis_vram_used\n");
 		return ret;
 	}
+	ret = device_create_file(adev->dev, &dev_attr_mem_info_vram_vendor);
+	if (ret) {
+		DRM_ERROR("Failed to create device file mem_info_vram_vendor\n");
+		return ret;
+	}
 
 	return 0;
 }
@@ -180,6 +223,7 @@ static int amdgpu_vram_mgr_fini(struct ttm_mem_type_manager *man)
 	device_remove_file(adev->dev, &dev_attr_mem_info_vis_vram_total);
 	device_remove_file(adev->dev, &dev_attr_mem_info_vram_used);
 	device_remove_file(adev->dev, &dev_attr_mem_info_vis_vram_used);
+	device_remove_file(adev->dev, &dev_attr_mem_info_vram_vendor);
 	return 0;
 }
 
@@ -275,7 +319,7 @@ static int amdgpu_vram_mgr_new(struct ttm_mem_type_manager *man,
 	struct drm_mm_node *nodes;
 	enum drm_mm_insert_mode mode;
 	unsigned long lpfn, num_nodes, pages_per_node, pages_left;
-	uint64_t vis_usage = 0, mem_bytes;
+	uint64_t vis_usage = 0, mem_bytes, max_bytes;
 	unsigned i;
 	int r;
 
@@ -283,9 +327,13 @@ static int amdgpu_vram_mgr_new(struct ttm_mem_type_manager *man,
 	if (!lpfn)
 		lpfn = man->size;
 
+	max_bytes = adev->gmc.mc_vram_size;
+	if (tbo->type != ttm_bo_type_kernel)
+		max_bytes -= AMDGPU_VM_RESERVED_VRAM;
+
 	/* bail out quickly if there's likely not enough VRAM for this BO */
 	mem_bytes = (u64)mem->num_pages << PAGE_SHIFT;
-	if (atomic64_add_return(mem_bytes, &mgr->usage) > adev->gmc.mc_vram_size) {
+	if (atomic64_add_return(mem_bytes, &mgr->usage) > max_bytes) {
 		atomic64_sub(mem_bytes, &mgr->usage);
 		mem->mm_node = NULL;
 		return 0;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c
index 65aae75f80fd..61d13d8b7b20 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c
@@ -25,6 +25,7 @@
 #include "amdgpu.h"
 #include "amdgpu_xgmi.h"
 #include "amdgpu_smu.h"
+#include "amdgpu_ras.h"
 #include "df/df_3_6_offset.h"
 
 static DEFINE_MUTEX(xgmi_mutex);
@@ -273,22 +274,55 @@ int amdgpu_xgmi_set_pstate(struct amdgpu_device *adev, int pstate)
 {
 	int ret = 0;
 	struct amdgpu_hive_info *hive = amdgpu_get_xgmi_hive(adev, 0);
+	struct amdgpu_device *tmp_adev;
+	bool update_hive_pstate = true;
+	bool is_high_pstate = pstate && adev->asic_type == CHIP_VEGA20;
 
 	if (!hive)
 		return 0;
 
-	if (hive->pstate == pstate)
-		return 0;
+	mutex_lock(&hive->hive_lock);
+
+	if (hive->pstate == pstate) {
+		adev->pstate = is_high_pstate ? pstate : adev->pstate;
+		goto out;
+	}
 
 	dev_dbg(adev->dev, "Set xgmi pstate %d.\n", pstate);
 
 	if (is_support_sw_smu_xgmi(adev))
 		ret = smu_set_xgmi_pstate(&adev->smu, pstate);
-	if (ret)
+	else if (adev->powerplay.pp_funcs &&
+		 adev->powerplay.pp_funcs->set_xgmi_pstate)
+		ret = adev->powerplay.pp_funcs->set_xgmi_pstate(adev->powerplay.pp_handle,
+								pstate);
+
+	if (ret) {
 		dev_err(adev->dev,
 			"XGMI: Set pstate failure on device %llx, hive %llx, ret %d",
 			adev->gmc.xgmi.node_id,
 			adev->gmc.xgmi.hive_id, ret);
+		goto out;
+	}
+
+	/* Update device pstate */
+	adev->pstate = pstate;
+
+	/*
+	 * Update the hive pstate only all devices of the hive
+	 * are in the same pstate
+	 */
+	list_for_each_entry(tmp_adev, &hive->device_list, gmc.xgmi.head) {
+		if (tmp_adev->pstate != adev->pstate) {
+			update_hive_pstate = false;
+			break;
+		}
+	}
+	if (update_hive_pstate || is_high_pstate)
+		hive->pstate = pstate;
+
+out:
+	mutex_unlock(&hive->hive_lock);
 
 	return ret;
 }
@@ -363,6 +397,9 @@ int amdgpu_xgmi_add_device(struct amdgpu_device *adev)
 		goto exit;
 	}
 
+	/* Set default device pstate */
+	adev->pstate = -1;
+
 	top_info = &adev->psp.xgmi_context.top_info;
 
 	list_add_tail(&adev->gmc.xgmi.head, &hive->device_list);
@@ -437,3 +474,52 @@ void amdgpu_xgmi_remove_device(struct amdgpu_device *adev)
 		mutex_unlock(&hive->hive_lock);
 	}
 }
+
+int amdgpu_xgmi_ras_late_init(struct amdgpu_device *adev)
+{
+	int r;
+	struct ras_ih_if ih_info = {
+		.cb = NULL,
+	};
+	struct ras_fs_if fs_info = {
+		.sysfs_name = "xgmi_wafl_err_count",
+		.debugfs_name = "xgmi_wafl_err_inject",
+	};
+
+	if (!adev->gmc.xgmi.supported ||
+	    adev->gmc.xgmi.num_physical_nodes == 0)
+		return 0;
+
+	if (!adev->gmc.xgmi.ras_if) {
+		adev->gmc.xgmi.ras_if = kmalloc(sizeof(struct ras_common_if), GFP_KERNEL);
+		if (!adev->gmc.xgmi.ras_if)
+			return -ENOMEM;
+		adev->gmc.xgmi.ras_if->block = AMDGPU_RAS_BLOCK__XGMI_WAFL;
+		adev->gmc.xgmi.ras_if->type = AMDGPU_RAS_ERROR__MULTI_UNCORRECTABLE;
+		adev->gmc.xgmi.ras_if->sub_block_index = 0;
+		strcpy(adev->gmc.xgmi.ras_if->name, "xgmi_wafl");
+	}
+	ih_info.head = fs_info.head = *adev->gmc.xgmi.ras_if;
+	r = amdgpu_ras_late_init(adev, adev->gmc.xgmi.ras_if,
+				 &fs_info, &ih_info);
+	if (r || !amdgpu_ras_is_supported(adev, adev->gmc.xgmi.ras_if->block)) {
+		kfree(adev->gmc.xgmi.ras_if);
+		adev->gmc.xgmi.ras_if = NULL;
+	}
+
+	return r;
+}
+
+void amdgpu_xgmi_ras_fini(struct amdgpu_device *adev)
+{
+	if (amdgpu_ras_is_supported(adev, AMDGPU_RAS_BLOCK__XGMI_WAFL) &&
+			adev->gmc.xgmi.ras_if) {
+		struct ras_common_if *ras_if = adev->gmc.xgmi.ras_if;
+		struct ras_ih_if ih_info = {
+			.cb = NULL,
+		};
+
+		amdgpu_ras_late_fini(adev, ras_if, &ih_info);
+		kfree(ras_if);
+	}
+}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.h
index fbcee31788c4..bbf504ff7051 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.h
@@ -42,6 +42,8 @@ void amdgpu_xgmi_remove_device(struct amdgpu_device *adev);
 int amdgpu_xgmi_set_pstate(struct amdgpu_device *adev, int pstate);
 int amdgpu_xgmi_get_hops_count(struct amdgpu_device *adev,
 		struct amdgpu_device *peer_adev);
+int amdgpu_xgmi_ras_late_init(struct amdgpu_device *adev);
+void amdgpu_xgmi_ras_fini(struct amdgpu_device *adev);
 
 static inline bool amdgpu_xgmi_same_hive(struct amdgpu_device *adev,
 		struct amdgpu_device *bo_adev)
diff --git a/drivers/gpu/drm/amd/amdgpu/arct_reg_init.c b/drivers/gpu/drm/amd/amdgpu/arct_reg_init.c
index 4853899b1824..fda99c958c3b 100644
--- a/drivers/gpu/drm/amd/amdgpu/arct_reg_init.c
+++ b/drivers/gpu/drm/amd/amdgpu/arct_reg_init.c
@@ -24,7 +24,6 @@
 #include "soc15.h"
 
 #include "soc15_common.h"
-#include "soc15_hw_ip.h"
 #include "arct_ip_offset.h"
 
 int arct_reg_base_init(struct amdgpu_device *adev)
@@ -52,6 +51,8 @@ int arct_reg_base_init(struct amdgpu_device *adev)
 		adev->reg_offset[SDMA7_HWIP][i] = (uint32_t *)(&(SDMA7_BASE.instance[i]));
 		adev->reg_offset[SMUIO_HWIP][i] = (uint32_t *)(&(SMUIO_BASE.instance[i]));
 		adev->reg_offset[THM_HWIP][i] = (uint32_t *)(&(THM_BASE.instance[i]));
+		adev->reg_offset[UMC_HWIP][i] = (uint32_t *)(&(UMC_BASE.instance[i]));
+		adev->reg_offset[RSMU_HWIP][i] = (uint32_t *)(&(RSMU_BASE.instance[i]));
 	}
 	return 0;
 }
diff --git a/drivers/gpu/drm/amd/amdgpu/cik.c b/drivers/gpu/drm/amd/amdgpu/cik.c
index b81bb414fcb3..7a43993544c1 100644
--- a/drivers/gpu/drm/amd/amdgpu/cik.c
+++ b/drivers/gpu/drm/amd/amdgpu/cik.c
@@ -966,6 +966,25 @@ static bool cik_read_bios_from_rom(struct amdgpu_device *adev,
 
 static const struct amdgpu_allowed_register_entry cik_allowed_read_registers[] = {
 	{mmGRBM_STATUS},
+	{mmGRBM_STATUS2},
+	{mmGRBM_STATUS_SE0},
+	{mmGRBM_STATUS_SE1},
+	{mmGRBM_STATUS_SE2},
+	{mmGRBM_STATUS_SE3},
+	{mmSRBM_STATUS},
+	{mmSRBM_STATUS2},
+	{mmSDMA0_STATUS_REG + SDMA0_REGISTER_OFFSET},
+	{mmSDMA0_STATUS_REG + SDMA1_REGISTER_OFFSET},
+	{mmCP_STAT},
+	{mmCP_STALLED_STAT1},
+	{mmCP_STALLED_STAT2},
+	{mmCP_STALLED_STAT3},
+	{mmCP_CPF_BUSY_STAT},
+	{mmCP_CPF_STALLED_STAT1},
+	{mmCP_CPF_STATUS},
+	{mmCP_CPC_BUSY_STAT},
+	{mmCP_CPC_STALLED_STAT1},
+	{mmCP_CPC_STATUS},
 	{mmGB_ADDR_CONFIG},
 	{mmMC_ARB_RAMCFG},
 	{mmGB_TILE_MODE0},
@@ -1270,15 +1289,15 @@ static int cik_gpu_pci_config_reset(struct amdgpu_device *adev)
 }
 
 /**
- * cik_asic_reset - soft reset GPU
+ * cik_asic_pci_config_reset - soft reset GPU
  *
  * @adev: amdgpu_device pointer
  *
- * Look up which blocks are hung and attempt
- * to reset them.
+ * Use PCI Config method to reset the GPU.
+ *
  * Returns 0 for success.
  */
-static int cik_asic_reset(struct amdgpu_device *adev)
+static int cik_asic_pci_config_reset(struct amdgpu_device *adev)
 {
 	int r;
 
@@ -1294,7 +1313,45 @@ static int cik_asic_reset(struct amdgpu_device *adev)
 static enum amd_reset_method
 cik_asic_reset_method(struct amdgpu_device *adev)
 {
-	return AMD_RESET_METHOD_LEGACY;
+	bool baco_reset;
+
+	switch (adev->asic_type) {
+	case CHIP_BONAIRE:
+	case CHIP_HAWAII:
+		/* disable baco reset until it works */
+		/* smu7_asic_get_baco_capability(adev, &baco_reset); */
+		baco_reset = false;
+		break;
+	default:
+		baco_reset = false;
+		break;
+	}
+
+	if (baco_reset)
+		return AMD_RESET_METHOD_BACO;
+	else
+		return AMD_RESET_METHOD_LEGACY;
+}
+
+/**
+ * cik_asic_reset - soft reset GPU
+ *
+ * @adev: amdgpu_device pointer
+ *
+ * Look up which blocks are hung and attempt
+ * to reset them.
+ * Returns 0 for success.
+ */
+static int cik_asic_reset(struct amdgpu_device *adev)
+{
+	int r;
+
+	if (cik_asic_reset_method(adev) == AMD_RESET_METHOD_BACO)
+		r = smu7_asic_baco_reset(adev);
+	else
+		r = cik_asic_pci_config_reset(adev);
+
+	return r;
 }
 
 static u32 cik_get_config_memsize(struct amdgpu_device *adev)
@@ -1384,7 +1441,6 @@ static int cik_set_vce_clocks(struct amdgpu_device *adev, u32 evclk, u32 ecclk)
 static void cik_pcie_gen3_enable(struct amdgpu_device *adev)
 {
 	struct pci_dev *root = adev->pdev->bus->self;
-	int bridge_pos, gpu_pos;
 	u32 speed_cntl, current_data_rate;
 	int i;
 	u16 tmp16;
@@ -1419,12 +1475,7 @@ static void cik_pcie_gen3_enable(struct amdgpu_device *adev)
 		DRM_INFO("enabling PCIE gen 2 link speeds, disable with amdgpu.pcie_gen2=0\n");
 	}
 
-	bridge_pos = pci_pcie_cap(root);
-	if (!bridge_pos)
-		return;
-
-	gpu_pos = pci_pcie_cap(adev->pdev);
-	if (!gpu_pos)
+	if (!pci_is_pcie(root) || !pci_is_pcie(adev->pdev))
 		return;
 
 	if (adev->pm.pcie_gen_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_GEN3) {
@@ -1434,14 +1485,17 @@ static void cik_pcie_gen3_enable(struct amdgpu_device *adev)
 			u16 bridge_cfg2, gpu_cfg2;
 			u32 max_lw, current_lw, tmp;
 
-			pci_read_config_word(root, bridge_pos + PCI_EXP_LNKCTL, &bridge_cfg);
-			pci_read_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL, &gpu_cfg);
+			pcie_capability_read_word(root, PCI_EXP_LNKCTL,
+						  &bridge_cfg);
+			pcie_capability_read_word(adev->pdev, PCI_EXP_LNKCTL,
+						  &gpu_cfg);
 
 			tmp16 = bridge_cfg | PCI_EXP_LNKCTL_HAWD;
-			pci_write_config_word(root, bridge_pos + PCI_EXP_LNKCTL, tmp16);
+			pcie_capability_write_word(root, PCI_EXP_LNKCTL, tmp16);
 
 			tmp16 = gpu_cfg | PCI_EXP_LNKCTL_HAWD;
-			pci_write_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL, tmp16);
+			pcie_capability_write_word(adev->pdev, PCI_EXP_LNKCTL,
+						   tmp16);
 
 			tmp = RREG32_PCIE(ixPCIE_LC_STATUS1);
 			max_lw = (tmp & PCIE_LC_STATUS1__LC_DETECTED_LINK_WIDTH_MASK) >>
@@ -1465,15 +1519,23 @@ static void cik_pcie_gen3_enable(struct amdgpu_device *adev)
 
 			for (i = 0; i < 10; i++) {
 				/* check status */
-				pci_read_config_word(adev->pdev, gpu_pos + PCI_EXP_DEVSTA, &tmp16);
+				pcie_capability_read_word(adev->pdev,
+							  PCI_EXP_DEVSTA,
+							  &tmp16);
 				if (tmp16 & PCI_EXP_DEVSTA_TRPND)
 					break;
 
-				pci_read_config_word(root, bridge_pos + PCI_EXP_LNKCTL, &bridge_cfg);
-				pci_read_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL, &gpu_cfg);
+				pcie_capability_read_word(root, PCI_EXP_LNKCTL,
+							  &bridge_cfg);
+				pcie_capability_read_word(adev->pdev,
+							  PCI_EXP_LNKCTL,
+							  &gpu_cfg);
 
-				pci_read_config_word(root, bridge_pos + PCI_EXP_LNKCTL2, &bridge_cfg2);
-				pci_read_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL2, &gpu_cfg2);
+				pcie_capability_read_word(root, PCI_EXP_LNKCTL2,
+							  &bridge_cfg2);
+				pcie_capability_read_word(adev->pdev,
+							  PCI_EXP_LNKCTL2,
+							  &gpu_cfg2);
 
 				tmp = RREG32_PCIE(ixPCIE_LC_CNTL4);
 				tmp |= PCIE_LC_CNTL4__LC_SET_QUIESCE_MASK;
@@ -1486,26 +1548,45 @@ static void cik_pcie_gen3_enable(struct amdgpu_device *adev)
 				msleep(100);
 
 				/* linkctl */
-				pci_read_config_word(root, bridge_pos + PCI_EXP_LNKCTL, &tmp16);
+				pcie_capability_read_word(root, PCI_EXP_LNKCTL,
+							  &tmp16);
 				tmp16 &= ~PCI_EXP_LNKCTL_HAWD;
 				tmp16 |= (bridge_cfg & PCI_EXP_LNKCTL_HAWD);
-				pci_write_config_word(root, bridge_pos + PCI_EXP_LNKCTL, tmp16);
+				pcie_capability_write_word(root, PCI_EXP_LNKCTL,
+							   tmp16);
 
-				pci_read_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL, &tmp16);
+				pcie_capability_read_word(adev->pdev,
+							  PCI_EXP_LNKCTL,
+							  &tmp16);
 				tmp16 &= ~PCI_EXP_LNKCTL_HAWD;
 				tmp16 |= (gpu_cfg & PCI_EXP_LNKCTL_HAWD);
-				pci_write_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL, tmp16);
+				pcie_capability_write_word(adev->pdev,
+							   PCI_EXP_LNKCTL,
+							   tmp16);
 
 				/* linkctl2 */
-				pci_read_config_word(root, bridge_pos + PCI_EXP_LNKCTL2, &tmp16);
-				tmp16 &= ~((1 << 4) | (7 << 9));
-				tmp16 |= (bridge_cfg2 & ((1 << 4) | (7 << 9)));
-				pci_write_config_word(root, bridge_pos + PCI_EXP_LNKCTL2, tmp16);
-
-				pci_read_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL2, &tmp16);
-				tmp16 &= ~((1 << 4) | (7 << 9));
-				tmp16 |= (gpu_cfg2 & ((1 << 4) | (7 << 9)));
-				pci_write_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL2, tmp16);
+				pcie_capability_read_word(root, PCI_EXP_LNKCTL2,
+							  &tmp16);
+				tmp16 &= ~(PCI_EXP_LNKCTL2_ENTER_COMP |
+					   PCI_EXP_LNKCTL2_TX_MARGIN);
+				tmp16 |= (bridge_cfg2 &
+					  (PCI_EXP_LNKCTL2_ENTER_COMP |
+					   PCI_EXP_LNKCTL2_TX_MARGIN));
+				pcie_capability_write_word(root,
+							   PCI_EXP_LNKCTL2,
+							   tmp16);
+
+				pcie_capability_read_word(adev->pdev,
+							  PCI_EXP_LNKCTL2,
+							  &tmp16);
+				tmp16 &= ~(PCI_EXP_LNKCTL2_ENTER_COMP |
+					   PCI_EXP_LNKCTL2_TX_MARGIN);
+				tmp16 |= (gpu_cfg2 &
+					  (PCI_EXP_LNKCTL2_ENTER_COMP |
+					   PCI_EXP_LNKCTL2_TX_MARGIN));
+				pcie_capability_write_word(adev->pdev,
+							   PCI_EXP_LNKCTL2,
+							   tmp16);
 
 				tmp = RREG32_PCIE(ixPCIE_LC_CNTL4);
 				tmp &= ~PCIE_LC_CNTL4__LC_SET_QUIESCE_MASK;
@@ -1520,15 +1601,16 @@ static void cik_pcie_gen3_enable(struct amdgpu_device *adev)
 	speed_cntl &= ~PCIE_LC_SPEED_CNTL__LC_FORCE_DIS_SW_SPEED_CHANGE_MASK;
 	WREG32_PCIE(ixPCIE_LC_SPEED_CNTL, speed_cntl);
 
-	pci_read_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL2, &tmp16);
-	tmp16 &= ~0xf;
+	pcie_capability_read_word(adev->pdev, PCI_EXP_LNKCTL2, &tmp16);
+	tmp16 &= ~PCI_EXP_LNKCTL2_TLS;
+
 	if (adev->pm.pcie_gen_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_GEN3)
-		tmp16 |= 3; /* gen3 */
+		tmp16 |= PCI_EXP_LNKCTL2_TLS_8_0GT; /* gen3 */
 	else if (adev->pm.pcie_gen_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_GEN2)
-		tmp16 |= 2; /* gen2 */
+		tmp16 |= PCI_EXP_LNKCTL2_TLS_5_0GT; /* gen2 */
 	else
-		tmp16 |= 1; /* gen1 */
-	pci_write_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL2, tmp16);
+		tmp16 |= PCI_EXP_LNKCTL2_TLS_2_5GT; /* gen1 */
+	pcie_capability_write_word(adev->pdev, PCI_EXP_LNKCTL2, tmp16);
 
 	speed_cntl = RREG32_PCIE(ixPCIE_LC_SPEED_CNTL);
 	speed_cntl |= PCIE_LC_SPEED_CNTL__LC_INITIATE_LINK_SPEED_CHANGE_MASK;
diff --git a/drivers/gpu/drm/amd/amdgpu/cik.h b/drivers/gpu/drm/amd/amdgpu/cik.h
index 54c625a2e570..9870bf27870e 100644
--- a/drivers/gpu/drm/amd/amdgpu/cik.h
+++ b/drivers/gpu/drm/amd/amdgpu/cik.h
@@ -31,4 +31,7 @@ void cik_srbm_select(struct amdgpu_device *adev,
 int cik_set_ip_blocks(struct amdgpu_device *adev);
 
 void legacy_doorbell_index_init(struct amdgpu_device *adev);
+int smu7_asic_get_baco_capability(struct amdgpu_device *adev, bool *cap);
+int smu7_asic_baco_reset(struct amdgpu_device *adev);
+
 #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c
index 645550e7caf5..40d2ac723dd6 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c
@@ -330,9 +330,11 @@ static void dce_v10_0_hpd_init(struct amdgpu_device *adev)
 {
 	struct drm_device *dev = adev->ddev;
 	struct drm_connector *connector;
+	struct drm_connector_list_iter iter;
 	u32 tmp;
 
-	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
+	drm_connector_list_iter_begin(dev, &iter);
+	drm_for_each_connector_iter(connector, &iter) {
 		struct amdgpu_connector *amdgpu_connector = to_amdgpu_connector(connector);
 
 		if (amdgpu_connector->hpd.hpd >= adev->mode_info.num_hpd)
@@ -368,6 +370,7 @@ static void dce_v10_0_hpd_init(struct amdgpu_device *adev)
 		amdgpu_irq_get(adev, &adev->hpd_irq,
 			       amdgpu_connector->hpd.hpd);
 	}
+	drm_connector_list_iter_end(&iter);
 }
 
 /**
@@ -382,9 +385,11 @@ static void dce_v10_0_hpd_fini(struct amdgpu_device *adev)
 {
 	struct drm_device *dev = adev->ddev;
 	struct drm_connector *connector;
+	struct drm_connector_list_iter iter;
 	u32 tmp;
 
-	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
+	drm_connector_list_iter_begin(dev, &iter);
+	drm_for_each_connector_iter(connector, &iter) {
 		struct amdgpu_connector *amdgpu_connector = to_amdgpu_connector(connector);
 
 		if (amdgpu_connector->hpd.hpd >= adev->mode_info.num_hpd)
@@ -397,6 +402,7 @@ static void dce_v10_0_hpd_fini(struct amdgpu_device *adev)
 		amdgpu_irq_put(adev, &adev->hpd_irq,
 			       amdgpu_connector->hpd.hpd);
 	}
+	drm_connector_list_iter_end(&iter);
 }
 
 static u32 dce_v10_0_hpd_get_gpio_reg(struct amdgpu_device *adev)
@@ -1219,10 +1225,12 @@ static void dce_v10_0_afmt_audio_select_pin(struct drm_encoder *encoder)
 static void dce_v10_0_audio_write_latency_fields(struct drm_encoder *encoder,
 						struct drm_display_mode *mode)
 {
-	struct amdgpu_device *adev = encoder->dev->dev_private;
+	struct drm_device *dev = encoder->dev;
+	struct amdgpu_device *adev = dev->dev_private;
 	struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder);
 	struct amdgpu_encoder_atom_dig *dig = amdgpu_encoder->enc_priv;
 	struct drm_connector *connector;
+	struct drm_connector_list_iter iter;
 	struct amdgpu_connector *amdgpu_connector = NULL;
 	u32 tmp;
 	int interlace = 0;
@@ -1230,12 +1238,14 @@ static void dce_v10_0_audio_write_latency_fields(struct drm_encoder *encoder,
 	if (!dig || !dig->afmt || !dig->afmt->pin)
 		return;
 
-	list_for_each_entry(connector, &encoder->dev->mode_config.connector_list, head) {
+	drm_connector_list_iter_begin(dev, &iter);
+	drm_for_each_connector_iter(connector, &iter) {
 		if (connector->encoder == encoder) {
 			amdgpu_connector = to_amdgpu_connector(connector);
 			break;
 		}
 	}
+	drm_connector_list_iter_end(&iter);
 
 	if (!amdgpu_connector) {
 		DRM_ERROR("Couldn't find encoder's connector\n");
@@ -1261,10 +1271,12 @@ static void dce_v10_0_audio_write_latency_fields(struct drm_encoder *encoder,
 
 static void dce_v10_0_audio_write_speaker_allocation(struct drm_encoder *encoder)
 {
-	struct amdgpu_device *adev = encoder->dev->dev_private;
+	struct drm_device *dev = encoder->dev;
+	struct amdgpu_device *adev = dev->dev_private;
 	struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder);
 	struct amdgpu_encoder_atom_dig *dig = amdgpu_encoder->enc_priv;
 	struct drm_connector *connector;
+	struct drm_connector_list_iter iter;
 	struct amdgpu_connector *amdgpu_connector = NULL;
 	u32 tmp;
 	u8 *sadb = NULL;
@@ -1273,12 +1285,14 @@ static void dce_v10_0_audio_write_speaker_allocation(struct drm_encoder *encoder
 	if (!dig || !dig->afmt || !dig->afmt->pin)
 		return;
 
-	list_for_each_entry(connector, &encoder->dev->mode_config.connector_list, head) {
+	drm_connector_list_iter_begin(dev, &iter);
+	drm_for_each_connector_iter(connector, &iter) {
 		if (connector->encoder == encoder) {
 			amdgpu_connector = to_amdgpu_connector(connector);
 			break;
 		}
 	}
+	drm_connector_list_iter_end(&iter);
 
 	if (!amdgpu_connector) {
 		DRM_ERROR("Couldn't find encoder's connector\n");
@@ -1313,10 +1327,12 @@ static void dce_v10_0_audio_write_speaker_allocation(struct drm_encoder *encoder
 
 static void dce_v10_0_audio_write_sad_regs(struct drm_encoder *encoder)
 {
-	struct amdgpu_device *adev = encoder->dev->dev_private;
+	struct drm_device *dev = encoder->dev;
+	struct amdgpu_device *adev = dev->dev_private;
 	struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder);
 	struct amdgpu_encoder_atom_dig *dig = amdgpu_encoder->enc_priv;
 	struct drm_connector *connector;
+	struct drm_connector_list_iter iter;
 	struct amdgpu_connector *amdgpu_connector = NULL;
 	struct cea_sad *sads;
 	int i, sad_count;
@@ -1339,12 +1355,14 @@ static void dce_v10_0_audio_write_sad_regs(struct drm_encoder *encoder)
 	if (!dig || !dig->afmt || !dig->afmt->pin)
 		return;
 
-	list_for_each_entry(connector, &encoder->dev->mode_config.connector_list, head) {
+	drm_connector_list_iter_begin(dev, &iter);
+	drm_for_each_connector_iter(connector, &iter) {
 		if (connector->encoder == encoder) {
 			amdgpu_connector = to_amdgpu_connector(connector);
 			break;
 		}
 	}
+	drm_connector_list_iter_end(&iter);
 
 	if (!amdgpu_connector) {
 		DRM_ERROR("Couldn't find encoder's connector\n");
@@ -1352,10 +1370,10 @@ static void dce_v10_0_audio_write_sad_regs(struct drm_encoder *encoder)
 	}
 
 	sad_count = drm_edid_to_sad(amdgpu_connector_edid(connector), &sads);
-	if (sad_count <= 0) {
+	if (sad_count < 0)
 		DRM_ERROR("Couldn't read SADs: %d\n", sad_count);
+	if (sad_count <= 0)
 		return;
-	}
 	BUG_ON(!sads);
 
 	for (i = 0; i < ARRAY_SIZE(eld_reg_to_type); i++) {
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c
index d9f470632b2c..898ef72d423c 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c
@@ -348,9 +348,11 @@ static void dce_v11_0_hpd_init(struct amdgpu_device *adev)
 {
 	struct drm_device *dev = adev->ddev;
 	struct drm_connector *connector;
+	struct drm_connector_list_iter iter;
 	u32 tmp;
 
-	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
+	drm_connector_list_iter_begin(dev, &iter);
+	drm_for_each_connector_iter(connector, &iter) {
 		struct amdgpu_connector *amdgpu_connector = to_amdgpu_connector(connector);
 
 		if (amdgpu_connector->hpd.hpd >= adev->mode_info.num_hpd)
@@ -385,6 +387,7 @@ static void dce_v11_0_hpd_init(struct amdgpu_device *adev)
 		dce_v11_0_hpd_set_polarity(adev, amdgpu_connector->hpd.hpd);
 		amdgpu_irq_get(adev, &adev->hpd_irq, amdgpu_connector->hpd.hpd);
 	}
+	drm_connector_list_iter_end(&iter);
 }
 
 /**
@@ -399,9 +402,11 @@ static void dce_v11_0_hpd_fini(struct amdgpu_device *adev)
 {
 	struct drm_device *dev = adev->ddev;
 	struct drm_connector *connector;
+	struct drm_connector_list_iter iter;
 	u32 tmp;
 
-	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
+	drm_connector_list_iter_begin(dev, &iter);
+	drm_for_each_connector_iter(connector, &iter) {
 		struct amdgpu_connector *amdgpu_connector = to_amdgpu_connector(connector);
 
 		if (amdgpu_connector->hpd.hpd >= adev->mode_info.num_hpd)
@@ -413,6 +418,7 @@ static void dce_v11_0_hpd_fini(struct amdgpu_device *adev)
 
 		amdgpu_irq_put(adev, &adev->hpd_irq, amdgpu_connector->hpd.hpd);
 	}
+	drm_connector_list_iter_end(&iter);
 }
 
 static u32 dce_v11_0_hpd_get_gpio_reg(struct amdgpu_device *adev)
@@ -1245,10 +1251,12 @@ static void dce_v11_0_afmt_audio_select_pin(struct drm_encoder *encoder)
 static void dce_v11_0_audio_write_latency_fields(struct drm_encoder *encoder,
 						struct drm_display_mode *mode)
 {
-	struct amdgpu_device *adev = encoder->dev->dev_private;
+	struct drm_device *dev = encoder->dev;
+	struct amdgpu_device *adev = dev->dev_private;
 	struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder);
 	struct amdgpu_encoder_atom_dig *dig = amdgpu_encoder->enc_priv;
 	struct drm_connector *connector;
+	struct drm_connector_list_iter iter;
 	struct amdgpu_connector *amdgpu_connector = NULL;
 	u32 tmp;
 	int interlace = 0;
@@ -1256,12 +1264,14 @@ static void dce_v11_0_audio_write_latency_fields(struct drm_encoder *encoder,
 	if (!dig || !dig->afmt || !dig->afmt->pin)
 		return;
 
-	list_for_each_entry(connector, &encoder->dev->mode_config.connector_list, head) {
+	drm_connector_list_iter_begin(dev, &iter);
+	drm_for_each_connector_iter(connector, &iter) {
 		if (connector->encoder == encoder) {
 			amdgpu_connector = to_amdgpu_connector(connector);
 			break;
 		}
 	}
+	drm_connector_list_iter_end(&iter);
 
 	if (!amdgpu_connector) {
 		DRM_ERROR("Couldn't find encoder's connector\n");
@@ -1287,10 +1297,12 @@ static void dce_v11_0_audio_write_latency_fields(struct drm_encoder *encoder,
 
 static void dce_v11_0_audio_write_speaker_allocation(struct drm_encoder *encoder)
 {
-	struct amdgpu_device *adev = encoder->dev->dev_private;
+	struct drm_device *dev = encoder->dev;
+	struct amdgpu_device *adev = dev->dev_private;
 	struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder);
 	struct amdgpu_encoder_atom_dig *dig = amdgpu_encoder->enc_priv;
 	struct drm_connector *connector;
+	struct drm_connector_list_iter iter;
 	struct amdgpu_connector *amdgpu_connector = NULL;
 	u32 tmp;
 	u8 *sadb = NULL;
@@ -1299,12 +1311,14 @@ static void dce_v11_0_audio_write_speaker_allocation(struct drm_encoder *encoder
 	if (!dig || !dig->afmt || !dig->afmt->pin)
 		return;
 
-	list_for_each_entry(connector, &encoder->dev->mode_config.connector_list, head) {
+	drm_connector_list_iter_begin(dev, &iter);
+	drm_for_each_connector_iter(connector, &iter) {
 		if (connector->encoder == encoder) {
 			amdgpu_connector = to_amdgpu_connector(connector);
 			break;
 		}
 	}
+	drm_connector_list_iter_end(&iter);
 
 	if (!amdgpu_connector) {
 		DRM_ERROR("Couldn't find encoder's connector\n");
@@ -1339,10 +1353,12 @@ static void dce_v11_0_audio_write_speaker_allocation(struct drm_encoder *encoder
 
 static void dce_v11_0_audio_write_sad_regs(struct drm_encoder *encoder)
 {
-	struct amdgpu_device *adev = encoder->dev->dev_private;
+	struct drm_device *dev = encoder->dev;
+	struct amdgpu_device *adev = dev->dev_private;
 	struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder);
 	struct amdgpu_encoder_atom_dig *dig = amdgpu_encoder->enc_priv;
 	struct drm_connector *connector;
+	struct drm_connector_list_iter iter;
 	struct amdgpu_connector *amdgpu_connector = NULL;
 	struct cea_sad *sads;
 	int i, sad_count;
@@ -1365,12 +1381,14 @@ static void dce_v11_0_audio_write_sad_regs(struct drm_encoder *encoder)
 	if (!dig || !dig->afmt || !dig->afmt->pin)
 		return;
 
-	list_for_each_entry(connector, &encoder->dev->mode_config.connector_list, head) {
+	drm_connector_list_iter_begin(dev, &iter);
+	drm_for_each_connector_iter(connector, &iter) {
 		if (connector->encoder == encoder) {
 			amdgpu_connector = to_amdgpu_connector(connector);
 			break;
 		}
 	}
+	drm_connector_list_iter_end(&iter);
 
 	if (!amdgpu_connector) {
 		DRM_ERROR("Couldn't find encoder's connector\n");
@@ -1378,10 +1396,10 @@ static void dce_v11_0_audio_write_sad_regs(struct drm_encoder *encoder)
 	}
 
 	sad_count = drm_edid_to_sad(amdgpu_connector_edid(connector), &sads);
-	if (sad_count <= 0) {
+	if (sad_count < 0)
 		DRM_ERROR("Couldn't read SADs: %d\n", sad_count);
+	if (sad_count <= 0)
 		return;
-	}
 	BUG_ON(!sads);
 
 	for (i = 0; i < ARRAY_SIZE(eld_reg_to_type); i++) {
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c
index 3eb2e7429269..db15a112becc 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c
@@ -281,9 +281,11 @@ static void dce_v6_0_hpd_init(struct amdgpu_device *adev)
 {
 	struct drm_device *dev = adev->ddev;
 	struct drm_connector *connector;
+	struct drm_connector_list_iter iter;
 	u32 tmp;
 
-	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
+	drm_connector_list_iter_begin(dev, &iter);
+	drm_for_each_connector_iter(connector, &iter) {
 		struct amdgpu_connector *amdgpu_connector = to_amdgpu_connector(connector);
 
 		if (amdgpu_connector->hpd.hpd >= adev->mode_info.num_hpd)
@@ -309,7 +311,7 @@ static void dce_v6_0_hpd_init(struct amdgpu_device *adev)
 		dce_v6_0_hpd_set_polarity(adev, amdgpu_connector->hpd.hpd);
 		amdgpu_irq_get(adev, &adev->hpd_irq, amdgpu_connector->hpd.hpd);
 	}
-
+	drm_connector_list_iter_end(&iter);
 }
 
 /**
@@ -324,9 +326,11 @@ static void dce_v6_0_hpd_fini(struct amdgpu_device *adev)
 {
 	struct drm_device *dev = adev->ddev;
 	struct drm_connector *connector;
+	struct drm_connector_list_iter iter;
 	u32 tmp;
 
-	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
+	drm_connector_list_iter_begin(dev, &iter);
+	drm_for_each_connector_iter(connector, &iter) {
 		struct amdgpu_connector *amdgpu_connector = to_amdgpu_connector(connector);
 
 		if (amdgpu_connector->hpd.hpd >= adev->mode_info.num_hpd)
@@ -338,6 +342,7 @@ static void dce_v6_0_hpd_fini(struct amdgpu_device *adev)
 
 		amdgpu_irq_put(adev, &adev->hpd_irq, amdgpu_connector->hpd.hpd);
 	}
+	drm_connector_list_iter_end(&iter);
 }
 
 static u32 dce_v6_0_hpd_get_gpio_reg(struct amdgpu_device *adev)
@@ -1124,20 +1129,24 @@ static void dce_v6_0_audio_select_pin(struct drm_encoder *encoder)
 static void dce_v6_0_audio_write_latency_fields(struct drm_encoder *encoder,
 						struct drm_display_mode *mode)
 {
-	struct amdgpu_device *adev = encoder->dev->dev_private;
+	struct drm_device *dev = encoder->dev;
+	struct amdgpu_device *adev = dev->dev_private;
 	struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder);
 	struct amdgpu_encoder_atom_dig *dig = amdgpu_encoder->enc_priv;
 	struct drm_connector *connector;
+	struct drm_connector_list_iter iter;
 	struct amdgpu_connector *amdgpu_connector = NULL;
 	int interlace = 0;
 	u32 tmp;
 
-	list_for_each_entry(connector, &encoder->dev->mode_config.connector_list, head) {
+	drm_connector_list_iter_begin(dev, &iter);
+	drm_for_each_connector_iter(connector, &iter) {
 		if (connector->encoder == encoder) {
 			amdgpu_connector = to_amdgpu_connector(connector);
 			break;
 		}
 	}
+	drm_connector_list_iter_end(&iter);
 
 	if (!amdgpu_connector) {
 		DRM_ERROR("Couldn't find encoder's connector\n");
@@ -1164,21 +1173,25 @@ static void dce_v6_0_audio_write_latency_fields(struct drm_encoder *encoder,
 
 static void dce_v6_0_audio_write_speaker_allocation(struct drm_encoder *encoder)
 {
-	struct amdgpu_device *adev = encoder->dev->dev_private;
+	struct drm_device *dev = encoder->dev;
+	struct amdgpu_device *adev = dev->dev_private;
 	struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder);
 	struct amdgpu_encoder_atom_dig *dig = amdgpu_encoder->enc_priv;
 	struct drm_connector *connector;
+	struct drm_connector_list_iter iter;
 	struct amdgpu_connector *amdgpu_connector = NULL;
 	u8 *sadb = NULL;
 	int sad_count;
 	u32 tmp;
 
-	list_for_each_entry(connector, &encoder->dev->mode_config.connector_list, head) {
+	drm_connector_list_iter_begin(dev, &iter);
+	drm_for_each_connector_iter(connector, &iter) {
 		if (connector->encoder == encoder) {
 			amdgpu_connector = to_amdgpu_connector(connector);
 			break;
 		}
 	}
+	drm_connector_list_iter_end(&iter);
 
 	if (!amdgpu_connector) {
 		DRM_ERROR("Couldn't find encoder's connector\n");
@@ -1221,10 +1234,12 @@ static void dce_v6_0_audio_write_speaker_allocation(struct drm_encoder *encoder)
 
 static void dce_v6_0_audio_write_sad_regs(struct drm_encoder *encoder)
 {
-	struct amdgpu_device *adev = encoder->dev->dev_private;
+	struct drm_device *dev = encoder->dev;
+	struct amdgpu_device *adev = dev->dev_private;
 	struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder);
 	struct amdgpu_encoder_atom_dig *dig = amdgpu_encoder->enc_priv;
 	struct drm_connector *connector;
+	struct drm_connector_list_iter iter;
 	struct amdgpu_connector *amdgpu_connector = NULL;
 	struct cea_sad *sads;
 	int i, sad_count;
@@ -1244,12 +1259,14 @@ static void dce_v6_0_audio_write_sad_regs(struct drm_encoder *encoder)
 		{ ixAZALIA_F0_CODEC_PIN_CONTROL_AUDIO_DESCRIPTOR13, HDMI_AUDIO_CODING_TYPE_WMA_PRO },
 	};
 
-	list_for_each_entry(connector, &encoder->dev->mode_config.connector_list, head) {
+	drm_connector_list_iter_begin(dev, &iter);
+	drm_for_each_connector_iter(connector, &iter) {
 		if (connector->encoder == encoder) {
 			amdgpu_connector = to_amdgpu_connector(connector);
 			break;
 		}
 	}
+	drm_connector_list_iter_end(&iter);
 
 	if (!amdgpu_connector) {
 		DRM_ERROR("Couldn't find encoder's connector\n");
@@ -1257,10 +1274,10 @@ static void dce_v6_0_audio_write_sad_regs(struct drm_encoder *encoder)
 	}
 
 	sad_count = drm_edid_to_sad(amdgpu_connector_edid(connector), &sads);
-	if (sad_count <= 0) {
+	if (sad_count < 0)
 		DRM_ERROR("Couldn't read SADs: %d\n", sad_count);
+	if (sad_count <= 0)
 		return;
-	}
 
 	for (i = 0; i < ARRAY_SIZE(eld_reg_to_type); i++) {
 		u32 tmp = 0;
@@ -1632,6 +1649,7 @@ static void dce_v6_0_afmt_setmode(struct drm_encoder *encoder,
 	struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder);
 	struct amdgpu_encoder_atom_dig *dig = amdgpu_encoder->enc_priv;
 	struct drm_connector *connector;
+	struct drm_connector_list_iter iter;
 	struct amdgpu_connector *amdgpu_connector = NULL;
 	int em = amdgpu_atombios_encoder_get_encoder_mode(encoder);
 	int bpc = 8;
@@ -1639,12 +1657,14 @@ static void dce_v6_0_afmt_setmode(struct drm_encoder *encoder,
 	if (!dig || !dig->afmt)
 		return;
 
-	list_for_each_entry(connector, &encoder->dev->mode_config.connector_list, head) {
+	drm_connector_list_iter_begin(dev, &iter);
+	drm_for_each_connector_iter(connector, &iter) {
 		if (connector->encoder == encoder) {
 			amdgpu_connector = to_amdgpu_connector(connector);
 			break;
 		}
 	}
+	drm_connector_list_iter_end(&iter);
 
 	if (!amdgpu_connector) {
 		DRM_ERROR("Couldn't find encoder's connector\n");
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c
index a16c5e9e610e..f06c9022c1fd 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c
@@ -275,9 +275,11 @@ static void dce_v8_0_hpd_init(struct amdgpu_device *adev)
 {
 	struct drm_device *dev = adev->ddev;
 	struct drm_connector *connector;
+	struct drm_connector_list_iter iter;
 	u32 tmp;
 
-	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
+	drm_connector_list_iter_begin(dev, &iter);
+	drm_for_each_connector_iter(connector, &iter) {
 		struct amdgpu_connector *amdgpu_connector = to_amdgpu_connector(connector);
 
 		if (amdgpu_connector->hpd.hpd >= adev->mode_info.num_hpd)
@@ -303,6 +305,7 @@ static void dce_v8_0_hpd_init(struct amdgpu_device *adev)
 		dce_v8_0_hpd_set_polarity(adev, amdgpu_connector->hpd.hpd);
 		amdgpu_irq_get(adev, &adev->hpd_irq, amdgpu_connector->hpd.hpd);
 	}
+	drm_connector_list_iter_end(&iter);
 }
 
 /**
@@ -317,9 +320,11 @@ static void dce_v8_0_hpd_fini(struct amdgpu_device *adev)
 {
 	struct drm_device *dev = adev->ddev;
 	struct drm_connector *connector;
+	struct drm_connector_list_iter iter;
 	u32 tmp;
 
-	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
+	drm_connector_list_iter_begin(dev, &iter);
+	drm_for_each_connector_iter(connector, &iter) {
 		struct amdgpu_connector *amdgpu_connector = to_amdgpu_connector(connector);
 
 		if (amdgpu_connector->hpd.hpd >= adev->mode_info.num_hpd)
@@ -331,6 +336,7 @@ static void dce_v8_0_hpd_fini(struct amdgpu_device *adev)
 
 		amdgpu_irq_put(adev, &adev->hpd_irq, amdgpu_connector->hpd.hpd);
 	}
+	drm_connector_list_iter_end(&iter);
 }
 
 static u32 dce_v8_0_hpd_get_gpio_reg(struct amdgpu_device *adev)
@@ -1157,10 +1163,12 @@ static void dce_v8_0_afmt_audio_select_pin(struct drm_encoder *encoder)
 static void dce_v8_0_audio_write_latency_fields(struct drm_encoder *encoder,
 						struct drm_display_mode *mode)
 {
-	struct amdgpu_device *adev = encoder->dev->dev_private;
+	struct drm_device *dev = encoder->dev;
+	struct amdgpu_device *adev = dev->dev_private;
 	struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder);
 	struct amdgpu_encoder_atom_dig *dig = amdgpu_encoder->enc_priv;
 	struct drm_connector *connector;
+	struct drm_connector_list_iter iter;
 	struct amdgpu_connector *amdgpu_connector = NULL;
 	u32 tmp = 0, offset;
 
@@ -1169,12 +1177,14 @@ static void dce_v8_0_audio_write_latency_fields(struct drm_encoder *encoder,
 
 	offset = dig->afmt->pin->offset;
 
-	list_for_each_entry(connector, &encoder->dev->mode_config.connector_list, head) {
+	drm_connector_list_iter_begin(dev, &iter);
+	drm_for_each_connector_iter(connector, &iter) {
 		if (connector->encoder == encoder) {
 			amdgpu_connector = to_amdgpu_connector(connector);
 			break;
 		}
 	}
+	drm_connector_list_iter_end(&iter);
 
 	if (!amdgpu_connector) {
 		DRM_ERROR("Couldn't find encoder's connector\n");
@@ -1214,10 +1224,12 @@ static void dce_v8_0_audio_write_latency_fields(struct drm_encoder *encoder,
 
 static void dce_v8_0_audio_write_speaker_allocation(struct drm_encoder *encoder)
 {
-	struct amdgpu_device *adev = encoder->dev->dev_private;
+	struct drm_device *dev = encoder->dev;
+	struct amdgpu_device *adev = dev->dev_private;
 	struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder);
 	struct amdgpu_encoder_atom_dig *dig = amdgpu_encoder->enc_priv;
 	struct drm_connector *connector;
+	struct drm_connector_list_iter iter;
 	struct amdgpu_connector *amdgpu_connector = NULL;
 	u32 offset, tmp;
 	u8 *sadb = NULL;
@@ -1228,12 +1240,14 @@ static void dce_v8_0_audio_write_speaker_allocation(struct drm_encoder *encoder)
 
 	offset = dig->afmt->pin->offset;
 
-	list_for_each_entry(connector, &encoder->dev->mode_config.connector_list, head) {
+	drm_connector_list_iter_begin(dev, &iter);
+	drm_for_each_connector_iter(connector, &iter) {
 		if (connector->encoder == encoder) {
 			amdgpu_connector = to_amdgpu_connector(connector);
 			break;
 		}
 	}
+	drm_connector_list_iter_end(&iter);
 
 	if (!amdgpu_connector) {
 		DRM_ERROR("Couldn't find encoder's connector\n");
@@ -1263,11 +1277,13 @@ static void dce_v8_0_audio_write_speaker_allocation(struct drm_encoder *encoder)
 
 static void dce_v8_0_audio_write_sad_regs(struct drm_encoder *encoder)
 {
-	struct amdgpu_device *adev = encoder->dev->dev_private;
+	struct drm_device *dev = encoder->dev;
+	struct amdgpu_device *adev = dev->dev_private;
 	struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder);
 	struct amdgpu_encoder_atom_dig *dig = amdgpu_encoder->enc_priv;
 	u32 offset;
 	struct drm_connector *connector;
+	struct drm_connector_list_iter iter;
 	struct amdgpu_connector *amdgpu_connector = NULL;
 	struct cea_sad *sads;
 	int i, sad_count;
@@ -1292,12 +1308,14 @@ static void dce_v8_0_audio_write_sad_regs(struct drm_encoder *encoder)
 
 	offset = dig->afmt->pin->offset;
 
-	list_for_each_entry(connector, &encoder->dev->mode_config.connector_list, head) {
+	drm_connector_list_iter_begin(dev, &iter);
+	drm_for_each_connector_iter(connector, &iter) {
 		if (connector->encoder == encoder) {
 			amdgpu_connector = to_amdgpu_connector(connector);
 			break;
 		}
 	}
+	drm_connector_list_iter_end(&iter);
 
 	if (!amdgpu_connector) {
 		DRM_ERROR("Couldn't find encoder's connector\n");
@@ -1305,10 +1323,10 @@ static void dce_v8_0_audio_write_sad_regs(struct drm_encoder *encoder)
 	}
 
 	sad_count = drm_edid_to_sad(amdgpu_connector_edid(connector), &sads);
-	if (sad_count <= 0) {
+	if (sad_count < 0)
 		DRM_ERROR("Couldn't read SADs: %d\n", sad_count);
+	if (sad_count <= 0)
 		return;
-	}
 	BUG_ON(!sads);
 
 	for (i = 0; i < ARRAY_SIZE(eld_reg_to_type); i++) {
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_virtual.c b/drivers/gpu/drm/amd/amdgpu/dce_virtual.c
index c9608ae8643b..e4f94863332c 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_virtual.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_virtual.c
@@ -260,15 +260,14 @@ static struct drm_encoder *
 dce_virtual_encoder(struct drm_connector *connector)
 {
 	struct drm_encoder *encoder;
-	int i;
 
-	drm_connector_for_each_possible_encoder(connector, encoder, i) {
+	drm_connector_for_each_possible_encoder(connector, encoder) {
 		if (encoder->encoder_type == DRM_MODE_ENCODER_VIRTUAL)
 			return encoder;
 	}
 
 	/* pick the first one */
-	drm_connector_for_each_possible_encoder(connector, encoder, i)
+	drm_connector_for_each_possible_encoder(connector, encoder)
 		return encoder;
 
 	return NULL;
diff --git a/drivers/gpu/drm/amd/amdgpu/df_v1_7.c b/drivers/gpu/drm/amd/amdgpu/df_v1_7.c
index 844c03868248..d6221298b477 100644
--- a/drivers/gpu/drm/amd/amdgpu/df_v1_7.c
+++ b/drivers/gpu/drm/amd/amdgpu/df_v1_7.c
@@ -33,6 +33,10 @@ static void df_v1_7_sw_init(struct amdgpu_device *adev)
 {
 }
 
+static void df_v1_7_sw_fini(struct amdgpu_device *adev)
+{
+}
+
 static void df_v1_7_enable_broadcast_mode(struct amdgpu_device *adev,
                                           bool enable)
 {
@@ -111,6 +115,7 @@ static void df_v1_7_enable_ecc_force_par_wr_rmw(struct amdgpu_device *adev,
 
 const struct amdgpu_df_funcs df_v1_7_funcs = {
 	.sw_init = df_v1_7_sw_init,
+	.sw_fini = df_v1_7_sw_fini,
 	.enable_broadcast_mode = df_v1_7_enable_broadcast_mode,
 	.get_fb_channel_number = df_v1_7_get_fb_channel_number,
 	.get_hbm_channel_number = df_v1_7_get_hbm_channel_number,
diff --git a/drivers/gpu/drm/amd/amdgpu/df_v3_6.c b/drivers/gpu/drm/amd/amdgpu/df_v3_6.c
index 5850c8e34caa..16fbd2bc8ad1 100644
--- a/drivers/gpu/drm/amd/amdgpu/df_v3_6.c
+++ b/drivers/gpu/drm/amd/amdgpu/df_v3_6.c
@@ -99,8 +99,8 @@ static uint64_t df_v3_6_get_fica(struct amdgpu_device *adev,
 	unsigned long flags, address, data;
 	uint32_t ficadl_val, ficadh_val;
 
-	address = adev->nbio_funcs->get_pcie_index_offset(adev);
-	data = adev->nbio_funcs->get_pcie_data_offset(adev);
+	address = adev->nbio.funcs->get_pcie_index_offset(adev);
+	data = adev->nbio.funcs->get_pcie_data_offset(adev);
 
 	spin_lock_irqsave(&adev->pcie_idx_lock, flags);
 	WREG32(address, smnDF_PIE_AON_FabricIndirectConfigAccessAddress3);
@@ -122,8 +122,8 @@ static void df_v3_6_set_fica(struct amdgpu_device *adev, uint32_t ficaa_val,
 {
 	unsigned long flags, address, data;
 
-	address = adev->nbio_funcs->get_pcie_index_offset(adev);
-	data = adev->nbio_funcs->get_pcie_data_offset(adev);
+	address = adev->nbio.funcs->get_pcie_index_offset(adev);
+	data = adev->nbio.funcs->get_pcie_data_offset(adev);
 
 	spin_lock_irqsave(&adev->pcie_idx_lock, flags);
 	WREG32(address, smnDF_PIE_AON_FabricIndirectConfigAccessAddress3);
@@ -150,8 +150,8 @@ static void df_v3_6_perfmon_rreg(struct amdgpu_device *adev,
 {
 	unsigned long flags, address, data;
 
-	address = adev->nbio_funcs->get_pcie_index_offset(adev);
-	data = adev->nbio_funcs->get_pcie_data_offset(adev);
+	address = adev->nbio.funcs->get_pcie_index_offset(adev);
+	data = adev->nbio.funcs->get_pcie_data_offset(adev);
 
 	spin_lock_irqsave(&adev->pcie_idx_lock, flags);
 	WREG32(address, lo_addr);
@@ -172,8 +172,8 @@ static void df_v3_6_perfmon_wreg(struct amdgpu_device *adev, uint32_t lo_addr,
 {
 	unsigned long flags, address, data;
 
-	address = adev->nbio_funcs->get_pcie_index_offset(adev);
-	data = adev->nbio_funcs->get_pcie_data_offset(adev);
+	address = adev->nbio.funcs->get_pcie_index_offset(adev);
+	data = adev->nbio.funcs->get_pcie_data_offset(adev);
 
 	spin_lock_irqsave(&adev->pcie_idx_lock, flags);
 	WREG32(address, lo_addr);
@@ -220,6 +220,13 @@ static void df_v3_6_sw_init(struct amdgpu_device *adev)
 		adev->df_perfmon_config_assign_mask[i] = 0;
 }
 
+static void df_v3_6_sw_fini(struct amdgpu_device *adev)
+{
+
+	device_remove_file(adev->dev, &dev_attr_df_cntr_avail);
+
+}
+
 static void df_v3_6_enable_broadcast_mode(struct amdgpu_device *adev,
 					  bool enable)
 {
@@ -537,6 +544,7 @@ static void df_v3_6_pmc_get_count(struct amdgpu_device *adev,
 
 const struct amdgpu_df_funcs df_v3_6_funcs = {
 	.sw_init = df_v3_6_sw_init,
+	.sw_fini = df_v3_6_sw_fini,
 	.enable_broadcast_mode = df_v3_6_enable_broadcast_mode,
 	.get_fb_channel_number = df_v3_6_get_fb_channel_number,
 	.get_hbm_channel_number = df_v3_6_get_hbm_channel_number,
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c
index 638c821611ab..ca5f0e7ea1ac 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c
@@ -93,7 +93,7 @@ static const struct soc15_reg_golden golden_settings_gc_10_1[] =
 {
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCB_HW_CONTROL_4, 0xffffffff, 0x00400014),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCGTT_CPF_CLK_CTRL, 0xfcff8fff, 0xf8000100),
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCGTT_SPI_CLK_CTRL, 0xc0000000, 0xc0000100),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCGTT_SPI_CLK_CTRL, 0xcd000000, 0x0d000100),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCGTT_SQ_CLK_CTRL, 0x60000ff0, 0x60000100),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCGTT_SQG_CLK_CTRL, 0x40000000, 0x40000100),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCGTT_VGT_CLK_CTRL, 0xffff8fff, 0xffff8100),
@@ -127,7 +127,7 @@ static const struct soc15_reg_golden golden_settings_gc_10_1[] =
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmTA_CNTL_AUX, 0xfff7ffff, 0x01030000),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmTCP_CNTL, 0x60000010, 0x479c0010),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmUTCL1_CGTT_CLK_CTRL, 0xfeff0fff, 0x40000100),
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmUTCL1_CTRL, 0x00800000, 0x00800000)
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmUTCL1_CTRL, 0x00c00000, 0x00c00000)
 };
 
 static const struct soc15_reg_golden golden_settings_gc_10_0_nv10[] =
@@ -140,7 +140,7 @@ static const struct soc15_reg_golden golden_settings_gc_10_1_1[] =
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCB_HW_CONTROL_4, 0xffffffff, 0x003c0014),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCGTT_GS_NGG_CLK_CTRL, 0xffff8fff, 0xffff8100),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCGTT_IA_CLK_CTRL, 0xffff0fff, 0xffff0100),
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCGTT_SPI_CLK_CTRL, 0xc0000000, 0xc0000100),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCGTT_SPI_CLK_CTRL, 0xcd000000, 0x0d000100),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCGTT_SQ_CLK_CTRL, 0xf8ff0fff, 0x60000100),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCGTT_SQG_CLK_CTRL, 0x40000ff0, 0x40000100),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCGTT_VGT_CLK_CTRL, 0xffff8fff, 0xffff8100),
@@ -171,7 +171,7 @@ static const struct soc15_reg_golden golden_settings_gc_10_1_1[] =
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmSQ_LDS_CLK_CTRL, 0xffffffff, 0xffffffff),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmTA_CNTL_AUX, 0xfff7ffff, 0x01030000),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmTCP_CNTL, 0x60000010, 0x479c0010),
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmUTCL1_CTRL, 0x00800000, 0x00800000),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmUTCL1_CTRL, 0x00c00000, 0x00c00000),
 };
 
 static const struct soc15_reg_golden golden_settings_gc_10_1_2[] =
@@ -179,7 +179,7 @@ static const struct soc15_reg_golden golden_settings_gc_10_1_2[] =
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCB_HW_CONTROL_4, 0x003e001f, 0x003c0014),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCGTT_GS_NGG_CLK_CTRL, 0xffff8fff, 0xffff8100),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCGTT_IA_CLK_CTRL, 0xffff0fff, 0xffff0100),
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCGTT_SPI_CLK_CTRL, 0xff7f0fff, 0xc0000100),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCGTT_SPI_CLK_CTRL, 0xff7f0fff, 0x0d000100),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCGTT_SQ_CLK_CTRL, 0xffffcfff, 0x60000100),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCGTT_SQG_CLK_CTRL, 0xffff0fff, 0x40000100),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCGTT_VGT_CLK_CTRL, 0xffff8fff, 0xffff8100),
@@ -564,6 +564,32 @@ static void gfx_v10_0_free_microcode(struct amdgpu_device *adev)
 	kfree(adev->gfx.rlc.register_list_format);
 }
 
+static void gfx_v10_0_check_fw_write_wait(struct amdgpu_device *adev)
+{
+	adev->gfx.cp_fw_write_wait = false;
+
+	switch (adev->asic_type) {
+	case CHIP_NAVI10:
+	case CHIP_NAVI12:
+	case CHIP_NAVI14:
+		if ((adev->gfx.me_fw_version >= 0x00000046) &&
+		    (adev->gfx.me_feature_version >= 27) &&
+		    (adev->gfx.pfp_fw_version >= 0x00000068) &&
+		    (adev->gfx.pfp_feature_version >= 27) &&
+		    (adev->gfx.mec_fw_version >= 0x0000005b) &&
+		    (adev->gfx.mec_feature_version >= 27))
+			adev->gfx.cp_fw_write_wait = true;
+		break;
+	default:
+		break;
+	}
+
+	if (adev->gfx.cp_fw_write_wait == false)
+		DRM_WARN_ONCE("Warning: check cp_fw_version and update it to realize \
+			      GRBM requires 1-cycle delay in cp firmware\n");
+}
+
+
 static void gfx_v10_0_init_rlc_ext_microcode(struct amdgpu_device *adev)
 {
 	const struct rlc_firmware_header_v2_1 *rlc_hdr;
@@ -832,6 +858,7 @@ static int gfx_v10_0_init_microcode(struct amdgpu_device *adev)
 		}
 	}
 
+	gfx_v10_0_check_fw_write_wait(adev);
 out:
 	if (err) {
 		dev_err(adev->dev,
@@ -1442,7 +1469,7 @@ static int gfx_v10_0_sw_fini(void *handle)
 		amdgpu_ring_fini(&adev->gfx.compute_ring[i]);
 
 	amdgpu_gfx_mqd_sw_fini(adev);
-	amdgpu_gfx_kiq_free_ring(&adev->gfx.kiq.ring, &adev->gfx.kiq.irq);
+	amdgpu_gfx_kiq_free_ring(&adev->gfx.kiq.ring);
 	amdgpu_gfx_kiq_fini(adev);
 
 	gfx_v10_0_pfp_fini(adev);
@@ -1691,6 +1718,17 @@ static void gfx_v10_0_tcp_harvest(struct amdgpu_device *adev)
 	}
 }
 
+static void gfx_v10_0_get_tcc_info(struct amdgpu_device *adev)
+{
+	/* TCCs are global (not instanced). */
+	uint32_t tcc_disable = RREG32_SOC15(GC, 0, mmCGTS_TCC_DISABLE) |
+			       RREG32_SOC15(GC, 0, mmCGTS_USER_TCC_DISABLE);
+
+	adev->gfx.config.tcc_disabled_mask =
+		REG_GET_FIELD(tcc_disable, CGTS_TCC_DISABLE, TCC_DISABLE) |
+		(REG_GET_FIELD(tcc_disable, CGTS_TCC_DISABLE, HI_TCC_DISABLE) << 16);
+}
+
 static void gfx_v10_0_constants_init(struct amdgpu_device *adev)
 {
 	u32 tmp;
@@ -1702,6 +1740,7 @@ static void gfx_v10_0_constants_init(struct amdgpu_device *adev)
 
 	gfx_v10_0_setup_rb(adev);
 	gfx_v10_0_get_cu_info(adev, &adev->gfx.cu_info);
+	gfx_v10_0_get_tcc_info(adev);
 	adev->gfx.config.pa_sc_tile_steering_override =
 		gfx_v10_0_init_pa_sc_tile_steering_override(adev);
 
@@ -1746,27 +1785,52 @@ static void gfx_v10_0_enable_gui_idle_interrupt(struct amdgpu_device *adev,
 	WREG32_SOC15(GC, 0, mmCP_INT_CNTL_RING0, tmp);
 }
 
-static void gfx_v10_0_init_csb(struct amdgpu_device *adev)
+static int gfx_v10_0_init_csb(struct amdgpu_device *adev)
 {
+	int r;
+
+	if (adev->in_gpu_reset) {
+		r = amdgpu_bo_reserve(adev->gfx.rlc.clear_state_obj, false);
+		if (r)
+			return r;
+
+		r = amdgpu_bo_kmap(adev->gfx.rlc.clear_state_obj,
+				   (void **)&adev->gfx.rlc.cs_ptr);
+		if (!r) {
+			adev->gfx.rlc.funcs->get_csb_buffer(adev,
+					adev->gfx.rlc.cs_ptr);
+			amdgpu_bo_kunmap(adev->gfx.rlc.clear_state_obj);
+		}
+
+		amdgpu_bo_unreserve(adev->gfx.rlc.clear_state_obj);
+		if (r)
+			return r;
+	}
+
 	/* csib */
 	WREG32_SOC15(GC, 0, mmRLC_CSIB_ADDR_HI,
 		     adev->gfx.rlc.clear_state_gpu_addr >> 32);
 	WREG32_SOC15(GC, 0, mmRLC_CSIB_ADDR_LO,
 		     adev->gfx.rlc.clear_state_gpu_addr & 0xfffffffc);
 	WREG32_SOC15(GC, 0, mmRLC_CSIB_LENGTH, adev->gfx.rlc.clear_state_size);
+
+	return 0;
 }
 
-static void gfx_v10_0_init_pg(struct amdgpu_device *adev)
+static int gfx_v10_0_init_pg(struct amdgpu_device *adev)
 {
 	int i;
+	int r;
 
-	gfx_v10_0_init_csb(adev);
+	r = gfx_v10_0_init_csb(adev);
+	if (r)
+		return r;
 
 	for (i = 0; i < adev->num_vmhubs; i++)
 		amdgpu_gmc_flush_gpu_tlb(adev, 0, i, 0);
 
 	/* TODO: init power gating */
-	return;
+	return 0;
 }
 
 void gfx_v10_0_rlc_stop(struct amdgpu_device *adev)
@@ -1868,7 +1932,10 @@ static int gfx_v10_0_rlc_resume(struct amdgpu_device *adev)
 		r = gfx_v10_0_wait_for_rlc_autoload_complete(adev);
 		if (r)
 			return r;
-		gfx_v10_0_init_pg(adev);
+
+		r = gfx_v10_0_init_pg(adev);
+		if (r)
+			return r;
 
 		/* enable RLC SRM */
 		gfx_v10_0_rlc_enable_srm(adev);
@@ -1894,7 +1961,10 @@ static int gfx_v10_0_rlc_resume(struct amdgpu_device *adev)
 				return r;
 		}
 
-		gfx_v10_0_init_pg(adev);
+		r = gfx_v10_0_init_pg(adev);
+		if (r)
+			return r;
+
 		adev->gfx.rlc.funcs->start(adev);
 
 		if (adev->firmware.load_type == AMDGPU_FW_LOAD_RLC_BACKDOOR_AUTO) {
@@ -2361,7 +2431,7 @@ static int gfx_v10_0_wait_for_rlc_autoload_complete(struct amdgpu_device *adev)
 	return 0;
 }
 
-static void gfx_v10_0_cp_gfx_enable(struct amdgpu_device *adev, bool enable)
+static int gfx_v10_0_cp_gfx_enable(struct amdgpu_device *adev, bool enable)
 {
 	int i;
 	u32 tmp = RREG32_SOC15(GC, 0, mmCP_ME_CNTL);
@@ -2374,7 +2444,17 @@ static void gfx_v10_0_cp_gfx_enable(struct amdgpu_device *adev, bool enable)
 			adev->gfx.gfx_ring[i].sched.ready = false;
 	}
 	WREG32_SOC15(GC, 0, mmCP_ME_CNTL, tmp);
-	udelay(50);
+
+	for (i = 0; i < adev->usec_timeout; i++) {
+		if (RREG32_SOC15(GC, 0, mmCP_STAT) == 0)
+			break;
+		udelay(1);
+	}
+
+	if (i >= adev->usec_timeout)
+		DRM_ERROR("failed to %s cp gfx\n", enable ? "unhalt" : "halt");
+
+	return 0;
 }
 
 static int gfx_v10_0_cp_gfx_load_pfp_microcode(struct amdgpu_device *adev)
@@ -2431,7 +2511,7 @@ static int gfx_v10_0_cp_gfx_load_pfp_microcode(struct amdgpu_device *adev)
 	}
 
 	if (amdgpu_emu_mode == 1)
-		adev->nbio_funcs->hdp_flush(adev, NULL);
+		adev->nbio.funcs->hdp_flush(adev, NULL);
 
 	tmp = RREG32_SOC15(GC, 0, mmCP_PFP_IC_BASE_CNTL);
 	tmp = REG_SET_FIELD(tmp, CP_PFP_IC_BASE_CNTL, VMID, 0);
@@ -2501,7 +2581,7 @@ static int gfx_v10_0_cp_gfx_load_ce_microcode(struct amdgpu_device *adev)
 	}
 
 	if (amdgpu_emu_mode == 1)
-		adev->nbio_funcs->hdp_flush(adev, NULL);
+		adev->nbio.funcs->hdp_flush(adev, NULL);
 
 	tmp = RREG32_SOC15(GC, 0, mmCP_CE_IC_BASE_CNTL);
 	tmp = REG_SET_FIELD(tmp, CP_CE_IC_BASE_CNTL, VMID, 0);
@@ -2570,7 +2650,7 @@ static int gfx_v10_0_cp_gfx_load_me_microcode(struct amdgpu_device *adev)
 	}
 
 	if (amdgpu_emu_mode == 1)
-		adev->nbio_funcs->hdp_flush(adev, NULL);
+		adev->nbio.funcs->hdp_flush(adev, NULL);
 
 	tmp = RREG32_SOC15(GC, 0, mmCP_ME_IC_BASE_CNTL);
 	tmp = REG_SET_FIELD(tmp, CP_ME_IC_BASE_CNTL, VMID, 0);
@@ -2891,7 +2971,7 @@ static int gfx_v10_0_cp_compute_load_microcode(struct amdgpu_device *adev)
 	}
 
 	if (amdgpu_emu_mode == 1)
-		adev->nbio_funcs->hdp_flush(adev, NULL);
+		adev->nbio.funcs->hdp_flush(adev, NULL);
 
 	tmp = RREG32_SOC15(GC, 0, mmCP_CPC_IC_BASE_CNTL);
 	tmp = REG_SET_FIELD(tmp, CP_CPC_IC_BASE_CNTL, CACHE_POLICY, 0);
@@ -3075,6 +3155,7 @@ static int gfx_v10_0_gfx_init_queue(struct amdgpu_ring *ring)
 {
 	struct amdgpu_device *adev = ring->adev;
 	struct v10_gfx_mqd *mqd = ring->mqd_ptr;
+	int mqd_idx = ring - &adev->gfx.gfx_ring[0];
 
 	if (!adev->in_gpu_reset && !adev->in_suspend) {
 		memset((void *)mqd, 0, sizeof(*mqd));
@@ -3086,14 +3167,15 @@ static int gfx_v10_0_gfx_init_queue(struct amdgpu_ring *ring)
 #endif
 		nv_grbm_select(adev, 0, 0, 0, 0);
 		mutex_unlock(&adev->srbm_mutex);
-		if (adev->gfx.me.mqd_backup[AMDGPU_MAX_GFX_RINGS])
-			memcpy(adev->gfx.me.mqd_backup[AMDGPU_MAX_GFX_RINGS], mqd, sizeof(*mqd));
+		if (adev->gfx.me.mqd_backup[mqd_idx])
+			memcpy(adev->gfx.me.mqd_backup[mqd_idx], mqd, sizeof(*mqd));
 	} else if (adev->in_gpu_reset) {
 		/* reset mqd with the backup copy */
-		if (adev->gfx.me.mqd_backup[AMDGPU_MAX_GFX_RINGS])
-			memcpy(mqd, adev->gfx.me.mqd_backup[AMDGPU_MAX_GFX_RINGS], sizeof(*mqd));
+		if (adev->gfx.me.mqd_backup[mqd_idx])
+			memcpy(mqd, adev->gfx.me.mqd_backup[mqd_idx], sizeof(*mqd));
 		/* reset the ring */
 		ring->wptr = 0;
+		adev->wb.wb[ring->wptr_offs] = 0;
 		amdgpu_ring_clear_ring(ring);
 #ifdef BRING_UP_DEBUG
 		mutex_lock(&adev->srbm_mutex);
@@ -4345,7 +4427,7 @@ static void gfx_v10_0_ring_emit_hdp_flush(struct amdgpu_ring *ring)
 {
 	struct amdgpu_device *adev = ring->adev;
 	u32 ref_and_mask, reg_mem_engine;
-	const struct nbio_hdp_flush_reg *nbio_hf_reg = adev->nbio_funcs->hdp_flush_reg;
+	const struct nbio_hdp_flush_reg *nbio_hf_reg = adev->nbio.hdp_flush_reg;
 
 	if (ring->funcs->type == AMDGPU_RING_TYPE_COMPUTE) {
 		switch (ring->me) {
@@ -4365,8 +4447,8 @@ static void gfx_v10_0_ring_emit_hdp_flush(struct amdgpu_ring *ring)
 	}
 
 	gfx_v10_0_wait_reg_mem(ring, reg_mem_engine, 0, 1,
-			       adev->nbio_funcs->get_hdp_flush_req_offset(adev),
-			       adev->nbio_funcs->get_hdp_flush_done_offset(adev),
+			       adev->nbio.funcs->get_hdp_flush_req_offset(adev),
+			       adev->nbio.funcs->get_hdp_flush_done_offset(adev),
 			       ref_and_mask, ref_and_mask, 0x20);
 }
 
@@ -4753,6 +4835,24 @@ static void gfx_v10_0_ring_emit_reg_wait(struct amdgpu_ring *ring, uint32_t reg,
 	gfx_v10_0_wait_reg_mem(ring, 0, 0, 0, reg, 0, val, mask, 0x20);
 }
 
+static void gfx_v10_0_ring_emit_reg_write_reg_wait(struct amdgpu_ring *ring,
+						   uint32_t reg0, uint32_t reg1,
+						   uint32_t ref, uint32_t mask)
+{
+	int usepfp = (ring->funcs->type == AMDGPU_RING_TYPE_GFX);
+	struct amdgpu_device *adev = ring->adev;
+	bool fw_version_ok = false;
+
+	fw_version_ok = adev->gfx.cp_fw_write_wait;
+
+	if (fw_version_ok)
+		gfx_v10_0_wait_reg_mem(ring, usepfp, 0, 1, reg0, reg1,
+				       ref, mask, 0x20);
+	else
+		amdgpu_ring_emit_reg_write_reg_wait_helper(ring, reg0, reg1,
+							   ref, mask);
+}
+
 static void
 gfx_v10_0_set_gfx_eop_interrupt_state(struct amdgpu_device *adev,
 				      uint32_t me, uint32_t pipe,
@@ -5143,6 +5243,7 @@ static const struct amdgpu_ring_funcs gfx_v10_0_ring_funcs_gfx = {
 	.emit_tmz = gfx_v10_0_ring_emit_tmz,
 	.emit_wreg = gfx_v10_0_ring_emit_wreg,
 	.emit_reg_wait = gfx_v10_0_ring_emit_reg_wait,
+	.emit_reg_write_reg_wait = gfx_v10_0_ring_emit_reg_write_reg_wait,
 };
 
 static const struct amdgpu_ring_funcs gfx_v10_0_ring_funcs_compute = {
@@ -5176,6 +5277,7 @@ static const struct amdgpu_ring_funcs gfx_v10_0_ring_funcs_compute = {
 	.pad_ib = amdgpu_ring_generic_pad_ib,
 	.emit_wreg = gfx_v10_0_ring_emit_wreg,
 	.emit_reg_wait = gfx_v10_0_ring_emit_reg_wait,
+	.emit_reg_write_reg_wait = gfx_v10_0_ring_emit_reg_write_reg_wait,
 };
 
 static const struct amdgpu_ring_funcs gfx_v10_0_ring_funcs_kiq = {
@@ -5206,6 +5308,7 @@ static const struct amdgpu_ring_funcs gfx_v10_0_ring_funcs_kiq = {
 	.emit_rreg = gfx_v10_0_ring_emit_rreg,
 	.emit_wreg = gfx_v10_0_ring_emit_wreg,
 	.emit_reg_wait = gfx_v10_0_ring_emit_reg_wait,
+	.emit_reg_write_reg_wait = gfx_v10_0_ring_emit_reg_write_reg_wait,
 };
 
 static void gfx_v10_0_set_ring_funcs(struct amdgpu_device *adev)
@@ -5271,15 +5374,12 @@ static void gfx_v10_0_set_rlc_funcs(struct amdgpu_device *adev)
 
 static void gfx_v10_0_set_gds_init(struct amdgpu_device *adev)
 {
-	/* init asic gds info */
-	switch (adev->asic_type) {
-	case CHIP_NAVI10:
-	default:
-		adev->gds.gds_size = 0x10000;
-		adev->gds.gds_compute_max_wave_id = 0x4ff;
-		break;
-	}
+	unsigned total_cu = adev->gfx.config.max_cu_per_sh *
+			    adev->gfx.config.max_sh_per_se *
+			    adev->gfx.config.max_shader_engines;
 
+	adev->gds.gds_size = 0x10000;
+	adev->gds.gds_compute_max_wave_id = total_cu * 32 - 1;
 	adev->gds.gws_size = 64;
 	adev->gds.oa_size = 16;
 }
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
index 87dd55e9d72b..ffbde9136372 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
@@ -2103,7 +2103,7 @@ static int gfx_v8_0_sw_fini(void *handle)
 		amdgpu_ring_fini(&adev->gfx.compute_ring[i]);
 
 	amdgpu_gfx_mqd_sw_fini(adev);
-	amdgpu_gfx_kiq_free_ring(&adev->gfx.kiq.ring, &adev->gfx.kiq.irq);
+	amdgpu_gfx_kiq_free_ring(&adev->gfx.kiq.ring);
 	amdgpu_gfx_kiq_fini(adev);
 
 	gfx_v8_0_mec_fini(adev);
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
index dcadc73bffd2..faf2ffce5837 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
@@ -131,6 +131,18 @@ MODULE_FIRMWARE("amdgpu/renoir_rlc.bin");
 #define mmTCP_CHAN_STEER_5_ARCT								0x0b0c
 #define mmTCP_CHAN_STEER_5_ARCT_BASE_IDX							0
 
+struct ras_gfx_subblock_reg {
+	const char *name;
+	uint32_t hwip;
+	uint32_t inst;
+	uint32_t seg;
+	uint32_t reg_offset;
+	uint32_t sec_count_mask;
+	uint32_t sec_count_shift;
+	uint32_t ded_count_mask;
+	uint32_t ded_count_shift;
+};
+
 enum ta_ras_gfx_subblock {
 	/*CPC*/
 	TA_RAS_BLOCK__GFX_CPC_INDEX_START = 0,
@@ -517,9 +529,9 @@ static const struct soc15_reg_golden golden_settings_gc_9_0[] =
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmTCP_CHAN_STEER_LO, 0xffffffff, 0xb5d3f197),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmVGT_CACHE_INVALIDATION, 0x3fff3af3, 0x19200000),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmVGT_GS_MAX_WAVE_ID, 0x00000fff, 0x000003ff),
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCP_MEC1_F32_INT_DIS, 0x00000000, 0x00000800),
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCP_MEC2_F32_INT_DIS, 0x00000000, 0x00000800),
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCP_DEBUG, 0x00000000, 0x00008000)
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCP_MEC1_F32_INT_DIS, 0x00000800, 0x00000800),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCP_MEC2_F32_INT_DIS, 0x00000800, 0x00000800),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCP_DEBUG, 0x00008000, 0x00008000)
 };
 
 static const struct soc15_reg_golden golden_settings_gc_9_0_vg10[] =
@@ -582,9 +594,9 @@ static const struct soc15_reg_golden golden_settings_gc_9_1[] =
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmVGT_CACHE_INVALIDATION, 0x3fff3af3, 0x19200000),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmVGT_GS_MAX_WAVE_ID, 0x00000fff, 0x000000ff),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmWD_UTCL1_CNTL, 0x08000000, 0x08000080),
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCP_MEC1_F32_INT_DIS, 0x00000000, 0x00000800),
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCP_MEC2_F32_INT_DIS, 0x00000000, 0x00000800),
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCP_DEBUG, 0x00000000, 0x00008000)
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCP_MEC1_F32_INT_DIS, 0x00000800, 0x00000800),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCP_MEC2_F32_INT_DIS, 0x00000800, 0x00000800),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCP_DEBUG, 0x00008000, 0x00008000)
 };
 
 static const struct soc15_reg_golden golden_settings_gc_9_1_rv1[] =
@@ -676,9 +688,9 @@ static const struct soc15_reg_golden golden_settings_gc_9_2_1_vg12[] =
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmTCP_CHAN_STEER_HI, 0xffffffff, 0x00000000),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmTCP_CHAN_STEER_LO, 0xffffffff, 0x76325410),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmTD_CNTL, 0x01bd9f33, 0x01000000),
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCP_MEC1_F32_INT_DIS, 0x00000000, 0x00000800),
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCP_MEC2_F32_INT_DIS, 0x00000000, 0x00000800),
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCP_DEBUG, 0x00000000, 0x00008000)
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCP_MEC1_F32_INT_DIS, 0x00000800, 0x00000800),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCP_MEC2_F32_INT_DIS, 0x00000800, 0x00000800),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCP_DEBUG, 0x00008000, 0x00008000)
 };
 
 static const struct soc15_reg_golden golden_settings_gc_9_4_1_arct[] =
@@ -691,6 +703,8 @@ static const struct soc15_reg_golden golden_settings_gc_9_4_1_arct[] =
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmTCP_CHAN_STEER_3_ARCT, 0x3fffffff, 0x2ebd9fe3),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmTCP_CHAN_STEER_4_ARCT, 0x3fffffff, 0xb90f5b1),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmTCP_CHAN_STEER_5_ARCT, 0x3ff, 0x135),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmSQ_CONFIG, 0xffffffff, 0x011A0000),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmSQ_FIFO_SIZES, 0xffffffff, 0x00000f00),
 };
 
 static const u32 GFX_RLC_SRM_INDEX_CNTL_ADDR_OFFSETS[] =
@@ -973,6 +987,13 @@ static void gfx_v9_0_check_fw_write_wait(struct amdgpu_device *adev)
 	adev->gfx.me_fw_write_wait = false;
 	adev->gfx.mec_fw_write_wait = false;
 
+	if ((adev->gfx.mec_fw_version < 0x000001a5) ||
+	    (adev->gfx.mec_feature_version < 46) ||
+	    (adev->gfx.pfp_fw_version < 0x000000b7) ||
+	    (adev->gfx.pfp_feature_version < 46))
+		DRM_WARN_ONCE("Warning: check cp_fw_version and update it to realize \
+			      GRBM requires 1-cycle delay in cp firmware\n");
+
 	switch (adev->asic_type) {
 	case CHIP_VEGA10:
 		if ((adev->gfx.me_fw_version >= 0x0000009c) &&
@@ -1031,8 +1052,13 @@ static void gfx_v9_0_check_if_need_gfxoff(struct amdgpu_device *adev)
 	case CHIP_VEGA20:
 		break;
 	case CHIP_RAVEN:
-		if (!(adev->rev_id >= 0x8 || adev->pdev->device == 0x15d8)
-			&&((adev->gfx.rlc_fw_version != 106 &&
+		/* Disable GFXOFF on original raven.  There are combinations
+		 * of sbios and platforms that are not stable.
+		 */
+		if (!(adev->rev_id >= 0x8 || adev->pdev->device == 0x15d8))
+			adev->pm.pp_feature &= ~PP_GFXOFF_MASK;
+		else if (!(adev->rev_id >= 0x8 || adev->pdev->device == 0x15d8)
+			 &&((adev->gfx.rlc_fw_version != 106 &&
 			     adev->gfx.rlc_fw_version < 531) ||
 			    (adev->gfx.rlc_fw_version == 53815) ||
 			    (adev->gfx.rlc_feature_version < 1) ||
@@ -1044,6 +1070,12 @@ static void gfx_v9_0_check_if_need_gfxoff(struct amdgpu_device *adev)
 				AMD_PG_SUPPORT_CP |
 				AMD_PG_SUPPORT_RLC_SMU_HS;
 		break;
+	case CHIP_RENOIR:
+		if (adev->pm.pp_feature & PP_GFXOFF_MASK)
+			adev->pg_flags |= AMD_PG_SUPPORT_GFX_PG |
+				AMD_PG_SUPPORT_CP |
+				AMD_PG_SUPPORT_RLC_SMU_HS;
+		break;
 	default:
 		break;
 	}
@@ -1324,7 +1356,8 @@ static int gfx_v9_0_init_cp_compute_microcode(struct amdgpu_device *adev,
 
 			/* TODO: Determine if MEC2 JT FW loading can be removed
 				 for all GFX V9 asic and above */
-			if (adev->asic_type != CHIP_ARCTURUS) {
+			if (adev->asic_type != CHIP_ARCTURUS &&
+			    adev->asic_type != CHIP_RENOIR) {
 				info = &adev->firmware.ucode[AMDGPU_UCODE_ID_CP_MEC2_JT];
 				info->ucode_id = AMDGPU_UCODE_ID_CP_MEC2_JT;
 				info->fw = adev->gfx.mec2_fw;
@@ -1956,190 +1989,6 @@ static int gfx_v9_0_gpu_early_init(struct amdgpu_device *adev)
 	return 0;
 }
 
-static int gfx_v9_0_ngg_create_buf(struct amdgpu_device *adev,
-				   struct amdgpu_ngg_buf *ngg_buf,
-				   int size_se,
-				   int default_size_se)
-{
-	int r;
-
-	if (size_se < 0) {
-		dev_err(adev->dev, "Buffer size is invalid: %d\n", size_se);
-		return -EINVAL;
-	}
-	size_se = size_se ? size_se : default_size_se;
-
-	ngg_buf->size = size_se * adev->gfx.config.max_shader_engines;
-	r = amdgpu_bo_create_kernel(adev, ngg_buf->size,
-				    PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM,
-				    &ngg_buf->bo,
-				    &ngg_buf->gpu_addr,
-				    NULL);
-	if (r) {
-		dev_err(adev->dev, "(%d) failed to create NGG buffer\n", r);
-		return r;
-	}
-	ngg_buf->bo_size = amdgpu_bo_size(ngg_buf->bo);
-
-	return r;
-}
-
-static int gfx_v9_0_ngg_fini(struct amdgpu_device *adev)
-{
-	int i;
-
-	for (i = 0; i < NGG_BUF_MAX; i++)
-		amdgpu_bo_free_kernel(&adev->gfx.ngg.buf[i].bo,
-				      &adev->gfx.ngg.buf[i].gpu_addr,
-				      NULL);
-
-	memset(&adev->gfx.ngg.buf[0], 0,
-			sizeof(struct amdgpu_ngg_buf) * NGG_BUF_MAX);
-
-	adev->gfx.ngg.init = false;
-
-	return 0;
-}
-
-static int gfx_v9_0_ngg_init(struct amdgpu_device *adev)
-{
-	int r;
-
-	if (!amdgpu_ngg || adev->gfx.ngg.init == true)
-		return 0;
-
-	/* GDS reserve memory: 64 bytes alignment */
-	adev->gfx.ngg.gds_reserve_size = ALIGN(5 * 4, 0x40);
-	adev->gds.gds_size -= adev->gfx.ngg.gds_reserve_size;
-	adev->gfx.ngg.gds_reserve_addr = RREG32_SOC15(GC, 0, mmGDS_VMID0_BASE);
-	adev->gfx.ngg.gds_reserve_addr += RREG32_SOC15(GC, 0, mmGDS_VMID0_SIZE);
-
-	/* Primitive Buffer */
-	r = gfx_v9_0_ngg_create_buf(adev, &adev->gfx.ngg.buf[NGG_PRIM],
-				    amdgpu_prim_buf_per_se,
-				    64 * 1024);
-	if (r) {
-		dev_err(adev->dev, "Failed to create Primitive Buffer\n");
-		goto err;
-	}
-
-	/* Position Buffer */
-	r = gfx_v9_0_ngg_create_buf(adev, &adev->gfx.ngg.buf[NGG_POS],
-				    amdgpu_pos_buf_per_se,
-				    256 * 1024);
-	if (r) {
-		dev_err(adev->dev, "Failed to create Position Buffer\n");
-		goto err;
-	}
-
-	/* Control Sideband */
-	r = gfx_v9_0_ngg_create_buf(adev, &adev->gfx.ngg.buf[NGG_CNTL],
-				    amdgpu_cntl_sb_buf_per_se,
-				    256);
-	if (r) {
-		dev_err(adev->dev, "Failed to create Control Sideband Buffer\n");
-		goto err;
-	}
-
-	/* Parameter Cache, not created by default */
-	if (amdgpu_param_buf_per_se <= 0)
-		goto out;
-
-	r = gfx_v9_0_ngg_create_buf(adev, &adev->gfx.ngg.buf[NGG_PARAM],
-				    amdgpu_param_buf_per_se,
-				    512 * 1024);
-	if (r) {
-		dev_err(adev->dev, "Failed to create Parameter Cache\n");
-		goto err;
-	}
-
-out:
-	adev->gfx.ngg.init = true;
-	return 0;
-err:
-	gfx_v9_0_ngg_fini(adev);
-	return r;
-}
-
-static int gfx_v9_0_ngg_en(struct amdgpu_device *adev)
-{
-	struct amdgpu_ring *ring = &adev->gfx.gfx_ring[0];
-	int r;
-	u32 data, base;
-
-	if (!amdgpu_ngg)
-		return 0;
-
-	/* Program buffer size */
-	data = REG_SET_FIELD(0, WD_BUF_RESOURCE_1, INDEX_BUF_SIZE,
-			     adev->gfx.ngg.buf[NGG_PRIM].size >> 8);
-	data = REG_SET_FIELD(data, WD_BUF_RESOURCE_1, POS_BUF_SIZE,
-			     adev->gfx.ngg.buf[NGG_POS].size >> 8);
-	WREG32_SOC15(GC, 0, mmWD_BUF_RESOURCE_1, data);
-
-	data = REG_SET_FIELD(0, WD_BUF_RESOURCE_2, CNTL_SB_BUF_SIZE,
-			     adev->gfx.ngg.buf[NGG_CNTL].size >> 8);
-	data = REG_SET_FIELD(data, WD_BUF_RESOURCE_2, PARAM_BUF_SIZE,
-			     adev->gfx.ngg.buf[NGG_PARAM].size >> 10);
-	WREG32_SOC15(GC, 0, mmWD_BUF_RESOURCE_2, data);
-
-	/* Program buffer base address */
-	base = lower_32_bits(adev->gfx.ngg.buf[NGG_PRIM].gpu_addr);
-	data = REG_SET_FIELD(0, WD_INDEX_BUF_BASE, BASE, base);
-	WREG32_SOC15(GC, 0, mmWD_INDEX_BUF_BASE, data);
-
-	base = upper_32_bits(adev->gfx.ngg.buf[NGG_PRIM].gpu_addr);
-	data = REG_SET_FIELD(0, WD_INDEX_BUF_BASE_HI, BASE_HI, base);
-	WREG32_SOC15(GC, 0, mmWD_INDEX_BUF_BASE_HI, data);
-
-	base = lower_32_bits(adev->gfx.ngg.buf[NGG_POS].gpu_addr);
-	data = REG_SET_FIELD(0, WD_POS_BUF_BASE, BASE, base);
-	WREG32_SOC15(GC, 0, mmWD_POS_BUF_BASE, data);
-
-	base = upper_32_bits(adev->gfx.ngg.buf[NGG_POS].gpu_addr);
-	data = REG_SET_FIELD(0, WD_POS_BUF_BASE_HI, BASE_HI, base);
-	WREG32_SOC15(GC, 0, mmWD_POS_BUF_BASE_HI, data);
-
-	base = lower_32_bits(adev->gfx.ngg.buf[NGG_CNTL].gpu_addr);
-	data = REG_SET_FIELD(0, WD_CNTL_SB_BUF_BASE, BASE, base);
-	WREG32_SOC15(GC, 0, mmWD_CNTL_SB_BUF_BASE, data);
-
-	base = upper_32_bits(adev->gfx.ngg.buf[NGG_CNTL].gpu_addr);
-	data = REG_SET_FIELD(0, WD_CNTL_SB_BUF_BASE_HI, BASE_HI, base);
-	WREG32_SOC15(GC, 0, mmWD_CNTL_SB_BUF_BASE_HI, data);
-
-	/* Clear GDS reserved memory */
-	r = amdgpu_ring_alloc(ring, 17);
-	if (r) {
-		DRM_ERROR("amdgpu: NGG failed to lock ring %s (%d).\n",
-			  ring->name, r);
-		return r;
-	}
-
-	gfx_v9_0_write_data_to_reg(ring, 0, false,
-				   SOC15_REG_OFFSET(GC, 0, mmGDS_VMID0_SIZE),
-			           (adev->gds.gds_size +
-				    adev->gfx.ngg.gds_reserve_size));
-
-	amdgpu_ring_write(ring, PACKET3(PACKET3_DMA_DATA, 5));
-	amdgpu_ring_write(ring, (PACKET3_DMA_DATA_CP_SYNC |
-				PACKET3_DMA_DATA_DST_SEL(1) |
-				PACKET3_DMA_DATA_SRC_SEL(2)));
-	amdgpu_ring_write(ring, 0);
-	amdgpu_ring_write(ring, 0);
-	amdgpu_ring_write(ring, adev->gfx.ngg.gds_reserve_addr);
-	amdgpu_ring_write(ring, 0);
-	amdgpu_ring_write(ring, PACKET3_DMA_DATA_CMD_RAW_WAIT |
-				adev->gfx.ngg.gds_reserve_size);
-
-	gfx_v9_0_write_data_to_reg(ring, 0, false,
-				   SOC15_REG_OFFSET(GC, 0, mmGDS_VMID0_SIZE), 0);
-
-	amdgpu_ring_commit(ring);
-
-	return 0;
-}
-
 static int gfx_v9_0_compute_ring_init(struct amdgpu_device *adev, int ring_id,
 				      int mec, int pipe, int queue)
 {
@@ -2307,10 +2156,6 @@ static int gfx_v9_0_sw_init(void *handle)
 	if (r)
 		return r;
 
-	r = gfx_v9_0_ngg_init(adev);
-	if (r)
-		return r;
-
 	return 0;
 }
 
@@ -2320,19 +2165,7 @@ static int gfx_v9_0_sw_fini(void *handle)
 	int i;
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
-	if (amdgpu_ras_is_supported(adev, AMDGPU_RAS_BLOCK__GFX) &&
-			adev->gfx.ras_if) {
-		struct ras_common_if *ras_if = adev->gfx.ras_if;
-		struct ras_ih_if ih_info = {
-			.head = *ras_if,
-		};
-
-		amdgpu_ras_debugfs_remove(adev, ras_if);
-		amdgpu_ras_sysfs_remove(adev, ras_if);
-		amdgpu_ras_interrupt_remove_handler(adev,  &ih_info);
-		amdgpu_ras_feature_enable(adev, ras_if, 0);
-		kfree(ras_if);
-	}
+	amdgpu_gfx_ras_fini(adev);
 
 	for (i = 0; i < adev->gfx.num_gfx_rings; i++)
 		amdgpu_ring_fini(&adev->gfx.gfx_ring[i]);
@@ -2340,11 +2173,10 @@ static int gfx_v9_0_sw_fini(void *handle)
 		amdgpu_ring_fini(&adev->gfx.compute_ring[i]);
 
 	amdgpu_gfx_mqd_sw_fini(adev);
-	amdgpu_gfx_kiq_free_ring(&adev->gfx.kiq.ring, &adev->gfx.kiq.irq);
+	amdgpu_gfx_kiq_free_ring(&adev->gfx.kiq.ring);
 	amdgpu_gfx_kiq_fini(adev);
 
 	gfx_v9_0_mec_fini(adev);
-	gfx_v9_0_ngg_fini(adev);
 	amdgpu_bo_unref(&adev->gfx.rlc.clear_state_obj);
 	if (adev->asic_type == CHIP_RAVEN || adev->asic_type == CHIP_RENOIR) {
 		amdgpu_bo_free_kernel(&adev->gfx.rlc.cp_table_obj,
@@ -2912,7 +2744,10 @@ static void gfx_v9_0_init_pg(struct amdgpu_device *adev)
 	 * And it's needed by gfxoff feature.
 	 */
 	if (adev->gfx.rlc.is_rlc_v2_1) {
-		gfx_v9_1_init_rlc_save_restore_list(adev);
+		if (adev->asic_type == CHIP_VEGA12 ||
+		    (adev->asic_type == CHIP_RAVEN &&
+		     adev->rev_id >= 8))
+			gfx_v9_1_init_rlc_save_restore_list(adev);
 		gfx_v9_0_enable_save_restore_machine(adev);
 	}
 
@@ -3883,12 +3718,6 @@ static int gfx_v9_0_hw_init(void *handle)
 	if (r)
 		return r;
 
-	if (adev->asic_type != CHIP_ARCTURUS) {
-		r = gfx_v9_0_ngg_en(adev);
-		if (r)
-			return r;
-	}
-
 	return r;
 }
 
@@ -3930,8 +3759,10 @@ static int gfx_v9_0_hw_fini(void *handle)
 	amdgpu_irq_put(adev, &adev->gfx.priv_reg_irq, 0);
 	amdgpu_irq_put(adev, &adev->gfx.priv_inst_irq, 0);
 
-	/* disable KCQ to avoid CPC touch memory not valid anymore */
-	gfx_v9_0_kcq_disable(adev);
+	/* DF freeze and kcq disable will fail */
+	if (!amdgpu_ras_intr_triggered())
+		/* disable KCQ to avoid CPC touch memory not valid anymore */
+		gfx_v9_0_kcq_disable(adev);
 
 	if (amdgpu_sriov_vf(adev)) {
 		gfx_v9_0_cp_gfx_enable(adev, false);
@@ -4067,9 +3898,22 @@ static uint64_t gfx_v9_0_get_gpu_clock_counter(struct amdgpu_device *adev)
 	uint64_t clock;
 
 	mutex_lock(&adev->gfx.gpu_clock_mutex);
-	WREG32_SOC15(GC, 0, mmRLC_CAPTURE_GPU_CLOCK_COUNT, 1);
-	clock = (uint64_t)RREG32_SOC15(GC, 0, mmRLC_GPU_CLOCK_COUNT_LSB) |
-		((uint64_t)RREG32_SOC15(GC, 0, mmRLC_GPU_CLOCK_COUNT_MSB) << 32ULL);
+	if (adev->asic_type == CHIP_VEGA10 && amdgpu_sriov_runtime(adev)) {
+		uint32_t tmp, lsb, msb, i = 0;
+		do {
+			if (i != 0)
+				udelay(1);
+			tmp = RREG32_SOC15(GC, 0, mmRLC_REFCLOCK_TIMESTAMP_MSB);
+			lsb = RREG32_SOC15(GC, 0, mmRLC_REFCLOCK_TIMESTAMP_LSB);
+			msb = RREG32_SOC15(GC, 0, mmRLC_REFCLOCK_TIMESTAMP_MSB);
+			i++;
+		} while (unlikely(tmp != msb) && (i < adev->usec_timeout));
+		clock = (uint64_t)lsb | ((uint64_t)msb << 32ULL);
+	} else {
+		WREG32_SOC15(GC, 0, mmRLC_CAPTURE_GPU_CLOCK_COUNT, 1);
+		clock = (uint64_t)RREG32_SOC15(GC, 0, mmRLC_GPU_CLOCK_COUNT_LSB) |
+			((uint64_t)RREG32_SOC15(GC, 0, mmRLC_GPU_CLOCK_COUNT_MSB) << 32ULL);
+	}
 	mutex_unlock(&adev->gfx.gpu_clock_mutex);
 	return clock;
 }
@@ -4184,6 +4028,7 @@ static const struct soc15_reg_entry sec_ded_counter_registers[] = {
    { SOC15_REG_ENTRY(GC, 0, mmTCC_EDC_CNT), 0, 1, 16},
    { SOC15_REG_ENTRY(GC, 0, mmTCP_ATC_EDC_GATCL1_CNT), 0, 4, 16},
    { SOC15_REG_ENTRY(GC, 0, mmTCP_EDC_CNT), 0, 4, 16},
+   { SOC15_REG_ENTRY(GC, 0, mmTCP_EDC_CNT_NEW), 0, 4, 16},
    { SOC15_REG_ENTRY(GC, 0, mmTD_EDC_CNT), 0, 4, 16},
    { SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT2), 0, 4, 6},
    { SOC15_REG_ENTRY(GC, 0, mmSQ_EDC_CNT), 0, 4, 16},
@@ -4203,6 +4048,10 @@ static int gfx_v9_0_do_edc_gds_workarounds(struct amdgpu_device *adev)
 	struct amdgpu_ring *ring = &adev->gfx.compute_ring[0];
 	int i, r;
 
+	/* only support when RAS is enabled */
+	if (!amdgpu_ras_is_supported(adev, AMDGPU_RAS_BLOCK__GFX))
+		return 0;
+
 	r = amdgpu_ring_alloc(ring, 7);
 	if (r) {
 		DRM_ERROR("amdgpu: GDS workarounds failed to lock ring %s (%d).\n",
@@ -4393,33 +4242,14 @@ static int gfx_v9_0_early_init(void *handle)
 	return 0;
 }
 
-static int gfx_v9_0_process_ras_data_cb(struct amdgpu_device *adev,
-		struct ras_err_data *err_data,
-		struct amdgpu_iv_entry *entry);
-
 static int gfx_v9_0_ecc_late_init(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-	struct ras_common_if **ras_if = &adev->gfx.ras_if;
-	struct ras_ih_if ih_info = {
-		.cb = gfx_v9_0_process_ras_data_cb,
-	};
-	struct ras_fs_if fs_info = {
-		.sysfs_name = "gfx_err_count",
-		.debugfs_name = "gfx_err_inject",
-	};
-	struct ras_common_if ras_block = {
-		.block = AMDGPU_RAS_BLOCK__GFX,
-		.type = AMDGPU_RAS_ERROR__MULTI_UNCORRECTABLE,
-		.sub_block_index = 0,
-		.name = "gfx",
-	};
 	int r;
 
-	if (!amdgpu_ras_is_supported(adev, AMDGPU_RAS_BLOCK__GFX)) {
-		amdgpu_ras_feature_enable_on_boot(adev, &ras_block, 0);
-		return 0;
-	}
+	r = amdgpu_gfx_ras_late_init(adev);
+	if (r)
+		return r;
 
 	r = gfx_v9_0_do_edc_gds_workarounds(adev);
 	if (r)
@@ -4430,72 +4260,7 @@ static int gfx_v9_0_ecc_late_init(void *handle)
 	if (r)
 		return r;
 
-	/* handle resume path. */
-	if (*ras_if) {
-		/* resend ras TA enable cmd during resume.
-		 * prepare to handle failure.
-		 */
-		ih_info.head = **ras_if;
-		r = amdgpu_ras_feature_enable_on_boot(adev, *ras_if, 1);
-		if (r) {
-			if (r == -EAGAIN) {
-				/* request a gpu reset. will run again. */
-				amdgpu_ras_request_reset_on_boot(adev,
-						AMDGPU_RAS_BLOCK__GFX);
-				return 0;
-			}
-			/* fail to enable ras, cleanup all. */
-			goto irq;
-		}
-		/* enable successfully. continue. */
-		goto resume;
-	}
-
-	*ras_if = kmalloc(sizeof(**ras_if), GFP_KERNEL);
-	if (!*ras_if)
-		return -ENOMEM;
-
-	**ras_if = ras_block;
-
-	r = amdgpu_ras_feature_enable_on_boot(adev, *ras_if, 1);
-	if (r) {
-		if (r == -EAGAIN) {
-			amdgpu_ras_request_reset_on_boot(adev,
-					AMDGPU_RAS_BLOCK__GFX);
-			r = 0;
-		}
-		goto feature;
-	}
-
-	ih_info.head = **ras_if;
-	fs_info.head = **ras_if;
-
-	r = amdgpu_ras_interrupt_add_handler(adev, &ih_info);
-	if (r)
-		goto interrupt;
-
-	amdgpu_ras_debugfs_create(adev, &fs_info);
-
-	r = amdgpu_ras_sysfs_create(adev, &fs_info);
-	if (r)
-		goto sysfs;
-resume:
-	r = amdgpu_irq_get(adev, &adev->gfx.cp_ecc_error_irq, 0);
-	if (r)
-		goto irq;
-
 	return 0;
-irq:
-	amdgpu_ras_sysfs_remove(adev, *ras_if);
-sysfs:
-	amdgpu_ras_debugfs_remove(adev, *ras_if);
-	amdgpu_ras_interrupt_remove_handler(adev, &ih_info);
-interrupt:
-	amdgpu_ras_feature_enable(adev, *ras_if, 0);
-feature:
-	kfree(*ras_if);
-	*ras_if = NULL;
-	return r;
 }
 
 static int gfx_v9_0_late_init(void *handle)
@@ -4560,16 +4325,14 @@ static void gfx_v9_0_update_gfx_cg_power_gating(struct amdgpu_device *adev,
 {
 	amdgpu_gfx_rlc_enter_safe_mode(adev);
 
-	if (is_support_sw_smu(adev) && !enable)
-		smu_set_gfx_cgpg(&adev->smu, enable);
-
 	if ((adev->pg_flags & AMD_PG_SUPPORT_GFX_PG) && enable) {
 		gfx_v9_0_enable_gfx_cg_power_gating(adev, true);
 		if (adev->pg_flags & AMD_PG_SUPPORT_GFX_PIPELINE)
 			gfx_v9_0_enable_gfx_pipeline_powergating(adev, true);
 	} else {
 		gfx_v9_0_enable_gfx_cg_power_gating(adev, false);
-		gfx_v9_0_enable_gfx_pipeline_powergating(adev, false);
+		if (adev->pg_flags & AMD_PG_SUPPORT_GFX_PIPELINE)
+			gfx_v9_0_enable_gfx_pipeline_powergating(adev, false);
 	}
 
 	amdgpu_gfx_rlc_exit_safe_mode(adev);
@@ -4838,8 +4601,6 @@ static int gfx_v9_0_set_powergating_state(void *handle,
 			gfx_v9_0_enable_cp_power_gating(adev, false);
 
 		/* update gfx cgpg state */
-		if (is_support_sw_smu(adev) && enable)
-			smu_set_gfx_cgpg(&adev->smu, enable);
 		gfx_v9_0_update_gfx_cg_power_gating(adev, enable);
 
 		/* update mgcg state */
@@ -4970,7 +4731,7 @@ static void gfx_v9_0_ring_emit_hdp_flush(struct amdgpu_ring *ring)
 {
 	struct amdgpu_device *adev = ring->adev;
 	u32 ref_and_mask, reg_mem_engine;
-	const struct nbio_hdp_flush_reg *nbio_hf_reg = adev->nbio_funcs->hdp_flush_reg;
+	const struct nbio_hdp_flush_reg *nbio_hf_reg = adev->nbio.hdp_flush_reg;
 
 	if (ring->funcs->type == AMDGPU_RING_TYPE_COMPUTE) {
 		switch (ring->me) {
@@ -4990,8 +4751,8 @@ static void gfx_v9_0_ring_emit_hdp_flush(struct amdgpu_ring *ring)
 	}
 
 	gfx_v9_0_wait_reg_mem(ring, reg_mem_engine, 0, 1,
-			      adev->nbio_funcs->get_hdp_flush_req_offset(adev),
-			      adev->nbio_funcs->get_hdp_flush_done_offset(adev),
+			      adev->nbio.funcs->get_hdp_flush_req_offset(adev),
+			      adev->nbio.funcs->get_hdp_flush_done_offset(adev),
 			      ref_and_mask, ref_and_mask, 0x20);
 }
 
@@ -5723,313 +5484,446 @@ static int gfx_v9_0_priv_inst_irq(struct amdgpu_device *adev,
 	return 0;
 }
 
-static int gfx_v9_0_process_ras_data_cb(struct amdgpu_device *adev,
-		struct ras_err_data *err_data,
-		struct amdgpu_iv_entry *entry)
-{
-	/* TODO ue will trigger an interrupt. */
-	kgd2kfd_set_sram_ecc_flag(adev->kfd.dev);
-	if (adev->gfx.funcs->query_ras_error_count)
-		adev->gfx.funcs->query_ras_error_count(adev, err_data);
-	amdgpu_ras_reset_gpu(adev, 0);
-	return AMDGPU_RAS_SUCCESS;
-}
 
-static const struct {
-	const char *name;
-	uint32_t ip;
-	uint32_t inst;
-	uint32_t seg;
-	uint32_t reg_offset;
-	uint32_t per_se_instance;
-	int32_t num_instance;
-	uint32_t sec_count_mask;
-	uint32_t ded_count_mask;
-} gfx_ras_edc_regs[] = {
-	{ "CPC_SCRATCH", SOC15_REG_ENTRY(GC, 0, mmCPC_EDC_SCRATCH_CNT), 0, 1,
-	  REG_FIELD_MASK(CPC_EDC_SCRATCH_CNT, SEC_COUNT),
-	  REG_FIELD_MASK(CPC_EDC_SCRATCH_CNT, DED_COUNT) },
-	{ "CPC_UCODE", SOC15_REG_ENTRY(GC, 0, mmCPC_EDC_UCODE_CNT), 0, 1,
-	  REG_FIELD_MASK(CPC_EDC_UCODE_CNT, SEC_COUNT),
-	  REG_FIELD_MASK(CPC_EDC_UCODE_CNT, DED_COUNT) },
-	{ "CPF_ROQ_ME1", SOC15_REG_ENTRY(GC, 0, mmCPF_EDC_ROQ_CNT), 0, 1,
-	  REG_FIELD_MASK(CPF_EDC_ROQ_CNT, COUNT_ME1), 0 },
-	{ "CPF_ROQ_ME2", SOC15_REG_ENTRY(GC, 0, mmCPF_EDC_ROQ_CNT), 0, 1,
-	  REG_FIELD_MASK(CPF_EDC_ROQ_CNT, COUNT_ME2), 0 },
-	{ "CPF_TAG", SOC15_REG_ENTRY(GC, 0, mmCPF_EDC_TAG_CNT), 0, 1,
-	  REG_FIELD_MASK(CPF_EDC_TAG_CNT, SEC_COUNT),
-	  REG_FIELD_MASK(CPF_EDC_TAG_CNT, DED_COUNT) },
-	{ "CPG_DMA_ROQ", SOC15_REG_ENTRY(GC, 0, mmCPG_EDC_DMA_CNT), 0, 1,
-	  REG_FIELD_MASK(CPG_EDC_DMA_CNT, ROQ_COUNT), 0 },
-	{ "CPG_DMA_TAG", SOC15_REG_ENTRY(GC, 0, mmCPG_EDC_DMA_CNT), 0, 1,
-	  REG_FIELD_MASK(CPG_EDC_DMA_CNT, TAG_SEC_COUNT),
-	  REG_FIELD_MASK(CPG_EDC_DMA_CNT, TAG_DED_COUNT) },
-	{ "CPG_TAG", SOC15_REG_ENTRY(GC, 0, mmCPG_EDC_TAG_CNT), 0, 1,
-	  REG_FIELD_MASK(CPG_EDC_TAG_CNT, SEC_COUNT),
-	  REG_FIELD_MASK(CPG_EDC_TAG_CNT, DED_COUNT) },
-	{ "DC_CSINVOC", SOC15_REG_ENTRY(GC, 0, mmDC_EDC_CSINVOC_CNT), 0, 1,
-	  REG_FIELD_MASK(DC_EDC_CSINVOC_CNT, COUNT_ME1), 0 },
-	{ "DC_RESTORE", SOC15_REG_ENTRY(GC, 0, mmDC_EDC_RESTORE_CNT), 0, 1,
-	  REG_FIELD_MASK(DC_EDC_RESTORE_CNT, COUNT_ME1), 0 },
-	{ "DC_STATE", SOC15_REG_ENTRY(GC, 0, mmDC_EDC_STATE_CNT), 0, 1,
-	  REG_FIELD_MASK(DC_EDC_STATE_CNT, COUNT_ME1), 0 },
-	{ "GDS_MEM", SOC15_REG_ENTRY(GC, 0, mmGDS_EDC_CNT), 0, 1,
-	  REG_FIELD_MASK(GDS_EDC_CNT, GDS_MEM_SEC),
-	  REG_FIELD_MASK(GDS_EDC_CNT, GDS_MEM_DED) },
-	{ "GDS_INPUT_QUEUE", SOC15_REG_ENTRY(GC, 0, mmGDS_EDC_CNT), 0, 1,
-	  REG_FIELD_MASK(GDS_EDC_CNT, GDS_INPUT_QUEUE_SED), 0 },
+static const struct ras_gfx_subblock_reg ras_subblock_regs[] = {
+	{ "CPC_SCRATCH", SOC15_REG_ENTRY(GC, 0, mmCPC_EDC_SCRATCH_CNT),
+	  SOC15_REG_FIELD(CPC_EDC_SCRATCH_CNT, SEC_COUNT),
+	  SOC15_REG_FIELD(CPC_EDC_SCRATCH_CNT, DED_COUNT)
+	},
+	{ "CPC_UCODE", SOC15_REG_ENTRY(GC, 0, mmCPC_EDC_UCODE_CNT),
+	  SOC15_REG_FIELD(CPC_EDC_UCODE_CNT, SEC_COUNT),
+	  SOC15_REG_FIELD(CPC_EDC_UCODE_CNT, DED_COUNT)
+	},
+	{ "CPF_ROQ_ME1", SOC15_REG_ENTRY(GC, 0, mmCPF_EDC_ROQ_CNT),
+	  SOC15_REG_FIELD(CPF_EDC_ROQ_CNT, COUNT_ME1),
+	  0, 0
+	},
+	{ "CPF_ROQ_ME2", SOC15_REG_ENTRY(GC, 0, mmCPF_EDC_ROQ_CNT),
+	  SOC15_REG_FIELD(CPF_EDC_ROQ_CNT, COUNT_ME2),
+	  0, 0
+	},
+	{ "CPF_TAG", SOC15_REG_ENTRY(GC, 0, mmCPF_EDC_TAG_CNT),
+	  SOC15_REG_FIELD(CPF_EDC_TAG_CNT, SEC_COUNT),
+	  SOC15_REG_FIELD(CPF_EDC_TAG_CNT, DED_COUNT)
+	},
+	{ "CPG_DMA_ROQ", SOC15_REG_ENTRY(GC, 0, mmCPG_EDC_DMA_CNT),
+	  SOC15_REG_FIELD(CPG_EDC_DMA_CNT, ROQ_COUNT),
+	  0, 0
+	},
+	{ "CPG_DMA_TAG", SOC15_REG_ENTRY(GC, 0, mmCPG_EDC_DMA_CNT),
+	  SOC15_REG_FIELD(CPG_EDC_DMA_CNT, TAG_SEC_COUNT),
+	  SOC15_REG_FIELD(CPG_EDC_DMA_CNT, TAG_DED_COUNT)
+	},
+	{ "CPG_TAG", SOC15_REG_ENTRY(GC, 0, mmCPG_EDC_TAG_CNT),
+	  SOC15_REG_FIELD(CPG_EDC_TAG_CNT, SEC_COUNT),
+	  SOC15_REG_FIELD(CPG_EDC_TAG_CNT, DED_COUNT)
+	},
+	{ "DC_CSINVOC", SOC15_REG_ENTRY(GC, 0, mmDC_EDC_CSINVOC_CNT),
+	  SOC15_REG_FIELD(DC_EDC_CSINVOC_CNT, COUNT_ME1),
+	  0, 0
+	},
+	{ "DC_RESTORE", SOC15_REG_ENTRY(GC, 0, mmDC_EDC_RESTORE_CNT),
+	  SOC15_REG_FIELD(DC_EDC_RESTORE_CNT, COUNT_ME1),
+	  0, 0
+	},
+	{ "DC_STATE", SOC15_REG_ENTRY(GC, 0, mmDC_EDC_STATE_CNT),
+	  SOC15_REG_FIELD(DC_EDC_STATE_CNT, COUNT_ME1),
+	  0, 0
+	},
+	{ "GDS_MEM", SOC15_REG_ENTRY(GC, 0, mmGDS_EDC_CNT),
+	  SOC15_REG_FIELD(GDS_EDC_CNT, GDS_MEM_SEC),
+	  SOC15_REG_FIELD(GDS_EDC_CNT, GDS_MEM_DED)
+	},
+	{ "GDS_INPUT_QUEUE", SOC15_REG_ENTRY(GC, 0, mmGDS_EDC_CNT),
+	  SOC15_REG_FIELD(GDS_EDC_CNT, GDS_INPUT_QUEUE_SED),
+	  0, 0
+	},
 	{ "GDS_ME0_CS_PIPE_MEM", SOC15_REG_ENTRY(GC, 0, mmGDS_EDC_OA_PHY_CNT),
-	  0, 1, REG_FIELD_MASK(GDS_EDC_OA_PHY_CNT, ME0_CS_PIPE_MEM_SEC),
-	  REG_FIELD_MASK(GDS_EDC_OA_PHY_CNT, ME0_CS_PIPE_MEM_DED) },
+	  SOC15_REG_FIELD(GDS_EDC_OA_PHY_CNT, ME0_CS_PIPE_MEM_SEC),
+	  SOC15_REG_FIELD(GDS_EDC_OA_PHY_CNT, ME0_CS_PIPE_MEM_DED)
+	},
 	{ "GDS_OA_PHY_PHY_CMD_RAM_MEM",
-	  SOC15_REG_ENTRY(GC, 0, mmGDS_EDC_OA_PHY_CNT), 0, 1,
-	  REG_FIELD_MASK(GDS_EDC_OA_PHY_CNT, PHY_CMD_RAM_MEM_SEC),
-	  REG_FIELD_MASK(GDS_EDC_OA_PHY_CNT, PHY_CMD_RAM_MEM_DED) },
+	  SOC15_REG_ENTRY(GC, 0, mmGDS_EDC_OA_PHY_CNT),
+	  SOC15_REG_FIELD(GDS_EDC_OA_PHY_CNT, PHY_CMD_RAM_MEM_SEC),
+	  SOC15_REG_FIELD(GDS_EDC_OA_PHY_CNT, PHY_CMD_RAM_MEM_DED)
+	},
 	{ "GDS_OA_PHY_PHY_DATA_RAM_MEM",
-	  SOC15_REG_ENTRY(GC, 0, mmGDS_EDC_OA_PHY_CNT), 0, 1,
-	  REG_FIELD_MASK(GDS_EDC_OA_PHY_CNT, PHY_DATA_RAM_MEM_SED), 0 },
+	  SOC15_REG_ENTRY(GC, 0, mmGDS_EDC_OA_PHY_CNT),
+	  SOC15_REG_FIELD(GDS_EDC_OA_PHY_CNT, PHY_DATA_RAM_MEM_SED),
+	  0, 0
+	},
 	{ "GDS_OA_PIPE_ME1_PIPE0_PIPE_MEM",
-	  SOC15_REG_ENTRY(GC, 0, mmGDS_EDC_OA_PIPE_CNT), 0, 1,
-	  REG_FIELD_MASK(GDS_EDC_OA_PIPE_CNT, ME1_PIPE0_PIPE_MEM_SEC),
-	  REG_FIELD_MASK(GDS_EDC_OA_PIPE_CNT, ME1_PIPE0_PIPE_MEM_DED) },
+	  SOC15_REG_ENTRY(GC, 0, mmGDS_EDC_OA_PIPE_CNT),
+	  SOC15_REG_FIELD(GDS_EDC_OA_PIPE_CNT, ME1_PIPE0_PIPE_MEM_SEC),
+	  SOC15_REG_FIELD(GDS_EDC_OA_PIPE_CNT, ME1_PIPE0_PIPE_MEM_DED)
+	},
 	{ "GDS_OA_PIPE_ME1_PIPE1_PIPE_MEM",
-	  SOC15_REG_ENTRY(GC, 0, mmGDS_EDC_OA_PIPE_CNT), 0, 1,
-	  REG_FIELD_MASK(GDS_EDC_OA_PIPE_CNT, ME1_PIPE1_PIPE_MEM_SEC),
-	  REG_FIELD_MASK(GDS_EDC_OA_PIPE_CNT, ME1_PIPE1_PIPE_MEM_DED) },
+	  SOC15_REG_ENTRY(GC, 0, mmGDS_EDC_OA_PIPE_CNT),
+	  SOC15_REG_FIELD(GDS_EDC_OA_PIPE_CNT, ME1_PIPE1_PIPE_MEM_SEC),
+	  SOC15_REG_FIELD(GDS_EDC_OA_PIPE_CNT, ME1_PIPE1_PIPE_MEM_DED)
+	},
 	{ "GDS_OA_PIPE_ME1_PIPE2_PIPE_MEM",
-	  SOC15_REG_ENTRY(GC, 0, mmGDS_EDC_OA_PIPE_CNT), 0, 1,
-	  REG_FIELD_MASK(GDS_EDC_OA_PIPE_CNT, ME1_PIPE2_PIPE_MEM_SEC),
-	  REG_FIELD_MASK(GDS_EDC_OA_PIPE_CNT, ME1_PIPE2_PIPE_MEM_DED) },
+	  SOC15_REG_ENTRY(GC, 0, mmGDS_EDC_OA_PIPE_CNT),
+	  SOC15_REG_FIELD(GDS_EDC_OA_PIPE_CNT, ME1_PIPE2_PIPE_MEM_SEC),
+	  SOC15_REG_FIELD(GDS_EDC_OA_PIPE_CNT, ME1_PIPE2_PIPE_MEM_DED)
+	},
 	{ "GDS_OA_PIPE_ME1_PIPE3_PIPE_MEM",
-	  SOC15_REG_ENTRY(GC, 0, mmGDS_EDC_OA_PIPE_CNT), 0, 1,
-	  REG_FIELD_MASK(GDS_EDC_OA_PIPE_CNT, ME1_PIPE3_PIPE_MEM_SEC),
-	  REG_FIELD_MASK(GDS_EDC_OA_PIPE_CNT, ME1_PIPE3_PIPE_MEM_DED) },
-	{ "SPI_SR_MEM", SOC15_REG_ENTRY(GC, 0, mmSPI_EDC_CNT), 1, 1,
-	  REG_FIELD_MASK(SPI_EDC_CNT, SPI_SR_MEM_SED_COUNT), 0 },
-	{ "TA_FS_DFIFO", SOC15_REG_ENTRY(GC, 0, mmTA_EDC_CNT), 1, 16,
-	  REG_FIELD_MASK(TA_EDC_CNT, TA_FS_DFIFO_SEC_COUNT),
-	  REG_FIELD_MASK(TA_EDC_CNT, TA_FS_DFIFO_DED_COUNT) },
-	{ "TA_FS_AFIFO", SOC15_REG_ENTRY(GC, 0, mmTA_EDC_CNT), 1, 16,
-	  REG_FIELD_MASK(TA_EDC_CNT, TA_FS_AFIFO_SED_COUNT), 0 },
-	{ "TA_FL_LFIFO", SOC15_REG_ENTRY(GC, 0, mmTA_EDC_CNT), 1, 16,
-	  REG_FIELD_MASK(TA_EDC_CNT, TA_FL_LFIFO_SED_COUNT), 0 },
-	{ "TA_FX_LFIFO", SOC15_REG_ENTRY(GC, 0, mmTA_EDC_CNT), 1, 16,
-	  REG_FIELD_MASK(TA_EDC_CNT, TA_FX_LFIFO_SED_COUNT), 0 },
-	{ "TA_FS_CFIFO", SOC15_REG_ENTRY(GC, 0, mmTA_EDC_CNT), 1, 16,
-	  REG_FIELD_MASK(TA_EDC_CNT, TA_FS_CFIFO_SED_COUNT), 0 },
-	{ "TCA_HOLE_FIFO", SOC15_REG_ENTRY(GC, 0, mmTCA_EDC_CNT), 0, 2,
-	  REG_FIELD_MASK(TCA_EDC_CNT, HOLE_FIFO_SED_COUNT), 0 },
-	{ "TCA_REQ_FIFO", SOC15_REG_ENTRY(GC, 0, mmTCA_EDC_CNT), 0, 2,
-	  REG_FIELD_MASK(TCA_EDC_CNT, REQ_FIFO_SED_COUNT), 0 },
-	{ "TCC_CACHE_DATA", SOC15_REG_ENTRY(GC, 0, mmTCC_EDC_CNT), 0, 16,
-	  REG_FIELD_MASK(TCC_EDC_CNT, CACHE_DATA_SEC_COUNT),
-	  REG_FIELD_MASK(TCC_EDC_CNT, CACHE_DATA_DED_COUNT) },
-	{ "TCC_CACHE_DIRTY", SOC15_REG_ENTRY(GC, 0, mmTCC_EDC_CNT), 0, 16,
-	  REG_FIELD_MASK(TCC_EDC_CNT, CACHE_DIRTY_SEC_COUNT),
-	  REG_FIELD_MASK(TCC_EDC_CNT, CACHE_DIRTY_DED_COUNT) },
-	{ "TCC_HIGH_RATE_TAG", SOC15_REG_ENTRY(GC, 0, mmTCC_EDC_CNT), 0, 16,
-	  REG_FIELD_MASK(TCC_EDC_CNT, HIGH_RATE_TAG_SEC_COUNT),
-	  REG_FIELD_MASK(TCC_EDC_CNT, HIGH_RATE_TAG_DED_COUNT) },
-	{ "TCC_LOW_RATE_TAG", SOC15_REG_ENTRY(GC, 0, mmTCC_EDC_CNT), 0, 16,
-	  REG_FIELD_MASK(TCC_EDC_CNT, LOW_RATE_TAG_SEC_COUNT),
-	  REG_FIELD_MASK(TCC_EDC_CNT, LOW_RATE_TAG_DED_COUNT) },
-	{ "TCC_SRC_FIFO", SOC15_REG_ENTRY(GC, 0, mmTCC_EDC_CNT), 0, 16,
-	  REG_FIELD_MASK(TCC_EDC_CNT, SRC_FIFO_SEC_COUNT),
-	  REG_FIELD_MASK(TCC_EDC_CNT, SRC_FIFO_DED_COUNT) },
-	{ "TCC_IN_USE_DEC", SOC15_REG_ENTRY(GC, 0, mmTCC_EDC_CNT), 0, 16,
-	  REG_FIELD_MASK(TCC_EDC_CNT, IN_USE_DEC_SED_COUNT), 0 },
-	{ "TCC_IN_USE_TRANSFER", SOC15_REG_ENTRY(GC, 0, mmTCC_EDC_CNT), 0, 16,
-	  REG_FIELD_MASK(TCC_EDC_CNT, IN_USE_TRANSFER_SED_COUNT), 0 },
-	{ "TCC_LATENCY_FIFO", SOC15_REG_ENTRY(GC, 0, mmTCC_EDC_CNT), 0, 16,
-	  REG_FIELD_MASK(TCC_EDC_CNT, LATENCY_FIFO_SED_COUNT), 0 },
-	{ "TCC_RETURN_DATA", SOC15_REG_ENTRY(GC, 0, mmTCC_EDC_CNT), 0, 16,
-	  REG_FIELD_MASK(TCC_EDC_CNT, RETURN_DATA_SED_COUNT), 0 },
-	{ "TCC_RETURN_CONTROL", SOC15_REG_ENTRY(GC, 0, mmTCC_EDC_CNT), 0, 16,
-	  REG_FIELD_MASK(TCC_EDC_CNT, RETURN_CONTROL_SED_COUNT), 0 },
-	{ "TCC_UC_ATOMIC_FIFO", SOC15_REG_ENTRY(GC, 0, mmTCC_EDC_CNT), 0, 16,
-	  REG_FIELD_MASK(TCC_EDC_CNT, UC_ATOMIC_FIFO_SED_COUNT), 0 },
-	{ "TCC_WRITE_RETURN", SOC15_REG_ENTRY(GC, 0, mmTCC_EDC_CNT2), 0, 16,
-	  REG_FIELD_MASK(TCC_EDC_CNT2, WRITE_RETURN_SED_COUNT), 0 },
-	{ "TCC_WRITE_CACHE_READ", SOC15_REG_ENTRY(GC, 0, mmTCC_EDC_CNT2), 0, 16,
-	  REG_FIELD_MASK(TCC_EDC_CNT2, WRITE_CACHE_READ_SED_COUNT), 0 },
-	{ "TCC_SRC_FIFO_NEXT_RAM", SOC15_REG_ENTRY(GC, 0, mmTCC_EDC_CNT2), 0,
-	  16, REG_FIELD_MASK(TCC_EDC_CNT2, SRC_FIFO_NEXT_RAM_SED_COUNT), 0 },
+	  SOC15_REG_ENTRY(GC, 0, mmGDS_EDC_OA_PIPE_CNT),
+	  SOC15_REG_FIELD(GDS_EDC_OA_PIPE_CNT, ME1_PIPE3_PIPE_MEM_SEC),
+	  SOC15_REG_FIELD(GDS_EDC_OA_PIPE_CNT, ME1_PIPE3_PIPE_MEM_DED)
+	},
+	{ "SPI_SR_MEM", SOC15_REG_ENTRY(GC, 0, mmSPI_EDC_CNT),
+	  SOC15_REG_FIELD(SPI_EDC_CNT, SPI_SR_MEM_SED_COUNT),
+	  0, 0
+	},
+	{ "TA_FS_DFIFO", SOC15_REG_ENTRY(GC, 0, mmTA_EDC_CNT),
+	  SOC15_REG_FIELD(TA_EDC_CNT, TA_FS_DFIFO_SEC_COUNT),
+	  SOC15_REG_FIELD(TA_EDC_CNT, TA_FS_DFIFO_DED_COUNT)
+	},
+	{ "TA_FS_AFIFO", SOC15_REG_ENTRY(GC, 0, mmTA_EDC_CNT),
+	  SOC15_REG_FIELD(TA_EDC_CNT, TA_FS_AFIFO_SED_COUNT),
+	  0, 0
+	},
+	{ "TA_FL_LFIFO", SOC15_REG_ENTRY(GC, 0, mmTA_EDC_CNT),
+	  SOC15_REG_FIELD(TA_EDC_CNT, TA_FL_LFIFO_SED_COUNT),
+	  0, 0
+	},
+	{ "TA_FX_LFIFO", SOC15_REG_ENTRY(GC, 0, mmTA_EDC_CNT),
+	  SOC15_REG_FIELD(TA_EDC_CNT, TA_FX_LFIFO_SED_COUNT),
+	  0, 0
+	},
+	{ "TA_FS_CFIFO", SOC15_REG_ENTRY(GC, 0, mmTA_EDC_CNT),
+	  SOC15_REG_FIELD(TA_EDC_CNT, TA_FS_CFIFO_SED_COUNT),
+	  0, 0
+	},
+	{ "TCA_HOLE_FIFO", SOC15_REG_ENTRY(GC, 0, mmTCA_EDC_CNT),
+	  SOC15_REG_FIELD(TCA_EDC_CNT, HOLE_FIFO_SED_COUNT),
+	  0, 0
+	},
+	{ "TCA_REQ_FIFO", SOC15_REG_ENTRY(GC, 0, mmTCA_EDC_CNT),
+	  SOC15_REG_FIELD(TCA_EDC_CNT, REQ_FIFO_SED_COUNT),
+	  0, 0
+	},
+	{ "TCC_CACHE_DATA", SOC15_REG_ENTRY(GC, 0, mmTCC_EDC_CNT),
+	  SOC15_REG_FIELD(TCC_EDC_CNT, CACHE_DATA_SEC_COUNT),
+	  SOC15_REG_FIELD(TCC_EDC_CNT, CACHE_DATA_DED_COUNT)
+	},
+	{ "TCC_CACHE_DIRTY", SOC15_REG_ENTRY(GC, 0, mmTCC_EDC_CNT),
+	  SOC15_REG_FIELD(TCC_EDC_CNT, CACHE_DIRTY_SEC_COUNT),
+	  SOC15_REG_FIELD(TCC_EDC_CNT, CACHE_DIRTY_DED_COUNT)
+	},
+	{ "TCC_HIGH_RATE_TAG", SOC15_REG_ENTRY(GC, 0, mmTCC_EDC_CNT),
+	  SOC15_REG_FIELD(TCC_EDC_CNT, HIGH_RATE_TAG_SEC_COUNT),
+	  SOC15_REG_FIELD(TCC_EDC_CNT, HIGH_RATE_TAG_DED_COUNT)
+	},
+	{ "TCC_LOW_RATE_TAG", SOC15_REG_ENTRY(GC, 0, mmTCC_EDC_CNT),
+	  SOC15_REG_FIELD(TCC_EDC_CNT, LOW_RATE_TAG_SEC_COUNT),
+	  SOC15_REG_FIELD(TCC_EDC_CNT, LOW_RATE_TAG_DED_COUNT)
+	},
+	{ "TCC_SRC_FIFO", SOC15_REG_ENTRY(GC, 0, mmTCC_EDC_CNT),
+	  SOC15_REG_FIELD(TCC_EDC_CNT, SRC_FIFO_SEC_COUNT),
+	  SOC15_REG_FIELD(TCC_EDC_CNT, SRC_FIFO_DED_COUNT)
+	},
+	{ "TCC_IN_USE_DEC", SOC15_REG_ENTRY(GC, 0, mmTCC_EDC_CNT),
+	  SOC15_REG_FIELD(TCC_EDC_CNT, IN_USE_DEC_SED_COUNT),
+	  0, 0
+	},
+	{ "TCC_IN_USE_TRANSFER", SOC15_REG_ENTRY(GC, 0, mmTCC_EDC_CNT),
+	  SOC15_REG_FIELD(TCC_EDC_CNT, IN_USE_TRANSFER_SED_COUNT),
+	  0, 0
+	},
+	{ "TCC_LATENCY_FIFO", SOC15_REG_ENTRY(GC, 0, mmTCC_EDC_CNT),
+	  SOC15_REG_FIELD(TCC_EDC_CNT, LATENCY_FIFO_SED_COUNT),
+	  0, 0
+	},
+	{ "TCC_RETURN_DATA", SOC15_REG_ENTRY(GC, 0, mmTCC_EDC_CNT),
+	  SOC15_REG_FIELD(TCC_EDC_CNT, RETURN_DATA_SED_COUNT),
+	  0, 0
+	},
+	{ "TCC_RETURN_CONTROL", SOC15_REG_ENTRY(GC, 0, mmTCC_EDC_CNT),
+	  SOC15_REG_FIELD(TCC_EDC_CNT, RETURN_CONTROL_SED_COUNT),
+	  0, 0
+	},
+	{ "TCC_UC_ATOMIC_FIFO", SOC15_REG_ENTRY(GC, 0, mmTCC_EDC_CNT),
+	  SOC15_REG_FIELD(TCC_EDC_CNT, UC_ATOMIC_FIFO_SED_COUNT),
+	  0, 0
+	},
+	{ "TCC_WRITE_RETURN", SOC15_REG_ENTRY(GC, 0, mmTCC_EDC_CNT2),
+	  SOC15_REG_FIELD(TCC_EDC_CNT2, WRITE_RETURN_SED_COUNT),
+	  0, 0
+	},
+	{ "TCC_WRITE_CACHE_READ", SOC15_REG_ENTRY(GC, 0, mmTCC_EDC_CNT2),
+	  SOC15_REG_FIELD(TCC_EDC_CNT2, WRITE_CACHE_READ_SED_COUNT),
+	  0, 0
+	},
+	{ "TCC_SRC_FIFO_NEXT_RAM", SOC15_REG_ENTRY(GC, 0, mmTCC_EDC_CNT2),
+	  SOC15_REG_FIELD(TCC_EDC_CNT2, SRC_FIFO_NEXT_RAM_SED_COUNT),
+	  0, 0
+	},
 	{ "TCC_LATENCY_FIFO_NEXT_RAM", SOC15_REG_ENTRY(GC, 0, mmTCC_EDC_CNT2),
-	  0, 16, REG_FIELD_MASK(TCC_EDC_CNT2, LATENCY_FIFO_NEXT_RAM_SED_COUNT),
-	  0 },
-	{ "TCC_CACHE_TAG_PROBE_FIFO", SOC15_REG_ENTRY(GC, 0, mmTCC_EDC_CNT2), 0,
-	  16, REG_FIELD_MASK(TCC_EDC_CNT2, CACHE_TAG_PROBE_FIFO_SED_COUNT), 0 },
+	  SOC15_REG_FIELD(TCC_EDC_CNT2, LATENCY_FIFO_NEXT_RAM_SED_COUNT),
+	  0, 0
+	},
+	{ "TCC_CACHE_TAG_PROBE_FIFO", SOC15_REG_ENTRY(GC, 0, mmTCC_EDC_CNT2),
+	  SOC15_REG_FIELD(TCC_EDC_CNT2, CACHE_TAG_PROBE_FIFO_SED_COUNT),
+	  0, 0
+	},
 	{ "TCC_WRRET_TAG_WRITE_RETURN", SOC15_REG_ENTRY(GC, 0, mmTCC_EDC_CNT2),
-	  0, 16, REG_FIELD_MASK(TCC_EDC_CNT2, WRRET_TAG_WRITE_RETURN_SED_COUNT),
-	  0 },
-	{ "TCC_ATOMIC_RETURN_BUFFER", SOC15_REG_ENTRY(GC, 0, mmTCC_EDC_CNT2), 0,
-	  16, REG_FIELD_MASK(TCC_EDC_CNT2, ATOMIC_RETURN_BUFFER_SED_COUNT), 0 },
-	{ "TCI_WRITE_RAM", SOC15_REG_ENTRY(GC, 0, mmTCI_EDC_CNT), 0, 72,
-	  REG_FIELD_MASK(TCI_EDC_CNT, WRITE_RAM_SED_COUNT), 0 },
-	{ "TCP_CACHE_RAM", SOC15_REG_ENTRY(GC, 0, mmTCP_EDC_CNT_NEW), 1, 16,
-	  REG_FIELD_MASK(TCP_EDC_CNT_NEW, CACHE_RAM_SEC_COUNT),
-	  REG_FIELD_MASK(TCP_EDC_CNT_NEW, CACHE_RAM_DED_COUNT) },
-	{ "TCP_LFIFO_RAM", SOC15_REG_ENTRY(GC, 0, mmTCP_EDC_CNT_NEW), 1, 16,
-	  REG_FIELD_MASK(TCP_EDC_CNT_NEW, LFIFO_RAM_SEC_COUNT),
-	  REG_FIELD_MASK(TCP_EDC_CNT_NEW, LFIFO_RAM_DED_COUNT) },
-	{ "TCP_CMD_FIFO", SOC15_REG_ENTRY(GC, 0, mmTCP_EDC_CNT_NEW), 1, 16,
-	  REG_FIELD_MASK(TCP_EDC_CNT_NEW, CMD_FIFO_SED_COUNT), 0 },
-	{ "TCP_VM_FIFO", SOC15_REG_ENTRY(GC, 0, mmTCP_EDC_CNT_NEW), 1, 16,
-	  REG_FIELD_MASK(TCP_EDC_CNT_NEW, VM_FIFO_SEC_COUNT), 0 },
-	{ "TCP_DB_RAM", SOC15_REG_ENTRY(GC, 0, mmTCP_EDC_CNT_NEW), 1, 16,
-	  REG_FIELD_MASK(TCP_EDC_CNT_NEW, DB_RAM_SED_COUNT), 0 },
-	{ "TCP_UTCL1_LFIFO0", SOC15_REG_ENTRY(GC, 0, mmTCP_EDC_CNT_NEW), 1, 16,
-	  REG_FIELD_MASK(TCP_EDC_CNT_NEW, UTCL1_LFIFO0_SEC_COUNT),
-	  REG_FIELD_MASK(TCP_EDC_CNT_NEW, UTCL1_LFIFO0_DED_COUNT) },
-	{ "TCP_UTCL1_LFIFO1", SOC15_REG_ENTRY(GC, 0, mmTCP_EDC_CNT_NEW), 1, 16,
-	  REG_FIELD_MASK(TCP_EDC_CNT_NEW, UTCL1_LFIFO1_SEC_COUNT),
-	  REG_FIELD_MASK(TCP_EDC_CNT_NEW, UTCL1_LFIFO1_DED_COUNT) },
-	{ "TD_SS_FIFO_LO", SOC15_REG_ENTRY(GC, 0, mmTD_EDC_CNT), 1, 16,
-	  REG_FIELD_MASK(TD_EDC_CNT, SS_FIFO_LO_SEC_COUNT),
-	  REG_FIELD_MASK(TD_EDC_CNT, SS_FIFO_LO_DED_COUNT) },
-	{ "TD_SS_FIFO_HI", SOC15_REG_ENTRY(GC, 0, mmTD_EDC_CNT), 1, 16,
-	  REG_FIELD_MASK(TD_EDC_CNT, SS_FIFO_HI_SEC_COUNT),
-	  REG_FIELD_MASK(TD_EDC_CNT, SS_FIFO_HI_DED_COUNT) },
-	{ "TD_CS_FIFO", SOC15_REG_ENTRY(GC, 0, mmTD_EDC_CNT), 1, 16,
-	  REG_FIELD_MASK(TD_EDC_CNT, CS_FIFO_SED_COUNT), 0 },
-	{ "SQ_LDS_D", SOC15_REG_ENTRY(GC, 0, mmSQ_EDC_CNT), 1, 16,
-	  REG_FIELD_MASK(SQ_EDC_CNT, LDS_D_SEC_COUNT),
-	  REG_FIELD_MASK(SQ_EDC_CNT, LDS_D_DED_COUNT) },
-	{ "SQ_LDS_I", SOC15_REG_ENTRY(GC, 0, mmSQ_EDC_CNT), 1, 16,
-	  REG_FIELD_MASK(SQ_EDC_CNT, LDS_I_SEC_COUNT),
-	  REG_FIELD_MASK(SQ_EDC_CNT, LDS_I_DED_COUNT) },
-	{ "SQ_SGPR", SOC15_REG_ENTRY(GC, 0, mmSQ_EDC_CNT), 1, 16,
-	  REG_FIELD_MASK(SQ_EDC_CNT, SGPR_SEC_COUNT),
-	  REG_FIELD_MASK(SQ_EDC_CNT, SGPR_DED_COUNT) },
-	{ "SQ_VGPR0", SOC15_REG_ENTRY(GC, 0, mmSQ_EDC_CNT), 1, 16,
-	  REG_FIELD_MASK(SQ_EDC_CNT, VGPR0_SEC_COUNT),
-	  REG_FIELD_MASK(SQ_EDC_CNT, VGPR0_DED_COUNT) },
-	{ "SQ_VGPR1", SOC15_REG_ENTRY(GC, 0, mmSQ_EDC_CNT), 1, 16,
-	  REG_FIELD_MASK(SQ_EDC_CNT, VGPR1_SEC_COUNT),
-	  REG_FIELD_MASK(SQ_EDC_CNT, VGPR1_DED_COUNT) },
-	{ "SQ_VGPR2", SOC15_REG_ENTRY(GC, 0, mmSQ_EDC_CNT), 1, 16,
-	  REG_FIELD_MASK(SQ_EDC_CNT, VGPR2_SEC_COUNT),
-	  REG_FIELD_MASK(SQ_EDC_CNT, VGPR2_DED_COUNT) },
-	{ "SQ_VGPR3", SOC15_REG_ENTRY(GC, 0, mmSQ_EDC_CNT), 1, 16,
-	  REG_FIELD_MASK(SQ_EDC_CNT, VGPR3_SEC_COUNT),
-	  REG_FIELD_MASK(SQ_EDC_CNT, VGPR3_DED_COUNT) },
+	  SOC15_REG_FIELD(TCC_EDC_CNT2, WRRET_TAG_WRITE_RETURN_SED_COUNT),
+	  0, 0
+	},
+	{ "TCC_ATOMIC_RETURN_BUFFER", SOC15_REG_ENTRY(GC, 0, mmTCC_EDC_CNT2),
+	  SOC15_REG_FIELD(TCC_EDC_CNT2, ATOMIC_RETURN_BUFFER_SED_COUNT),
+	  0, 0
+	},
+	{ "TCI_WRITE_RAM", SOC15_REG_ENTRY(GC, 0, mmTCI_EDC_CNT),
+	  SOC15_REG_FIELD(TCI_EDC_CNT, WRITE_RAM_SED_COUNT),
+	  0, 0
+	},
+	{ "TCP_CACHE_RAM", SOC15_REG_ENTRY(GC, 0, mmTCP_EDC_CNT_NEW),
+	  SOC15_REG_FIELD(TCP_EDC_CNT_NEW, CACHE_RAM_SEC_COUNT),
+	  SOC15_REG_FIELD(TCP_EDC_CNT_NEW, CACHE_RAM_DED_COUNT)
+	},
+	{ "TCP_LFIFO_RAM", SOC15_REG_ENTRY(GC, 0, mmTCP_EDC_CNT_NEW),
+	  SOC15_REG_FIELD(TCP_EDC_CNT_NEW, LFIFO_RAM_SEC_COUNT),
+	  SOC15_REG_FIELD(TCP_EDC_CNT_NEW, LFIFO_RAM_DED_COUNT)
+	},
+	{ "TCP_CMD_FIFO", SOC15_REG_ENTRY(GC, 0, mmTCP_EDC_CNT_NEW),
+	  SOC15_REG_FIELD(TCP_EDC_CNT_NEW, CMD_FIFO_SED_COUNT),
+	  0, 0
+	},
+	{ "TCP_VM_FIFO", SOC15_REG_ENTRY(GC, 0, mmTCP_EDC_CNT_NEW),
+	  SOC15_REG_FIELD(TCP_EDC_CNT_NEW, VM_FIFO_SEC_COUNT),
+	  0, 0
+	},
+	{ "TCP_DB_RAM", SOC15_REG_ENTRY(GC, 0, mmTCP_EDC_CNT_NEW),
+	  SOC15_REG_FIELD(TCP_EDC_CNT_NEW, DB_RAM_SED_COUNT),
+	  0, 0
+	},
+	{ "TCP_UTCL1_LFIFO0", SOC15_REG_ENTRY(GC, 0, mmTCP_EDC_CNT_NEW),
+	  SOC15_REG_FIELD(TCP_EDC_CNT_NEW, UTCL1_LFIFO0_SEC_COUNT),
+	  SOC15_REG_FIELD(TCP_EDC_CNT_NEW, UTCL1_LFIFO0_DED_COUNT)
+	},
+	{ "TCP_UTCL1_LFIFO1", SOC15_REG_ENTRY(GC, 0, mmTCP_EDC_CNT_NEW),
+	  SOC15_REG_FIELD(TCP_EDC_CNT_NEW, UTCL1_LFIFO1_SEC_COUNT),
+	  SOC15_REG_FIELD(TCP_EDC_CNT_NEW, UTCL1_LFIFO1_DED_COUNT)
+	},
+	{ "TD_SS_FIFO_LO", SOC15_REG_ENTRY(GC, 0, mmTD_EDC_CNT),
+	  SOC15_REG_FIELD(TD_EDC_CNT, SS_FIFO_LO_SEC_COUNT),
+	  SOC15_REG_FIELD(TD_EDC_CNT, SS_FIFO_LO_DED_COUNT)
+	},
+	{ "TD_SS_FIFO_HI", SOC15_REG_ENTRY(GC, 0, mmTD_EDC_CNT),
+	  SOC15_REG_FIELD(TD_EDC_CNT, SS_FIFO_HI_SEC_COUNT),
+	  SOC15_REG_FIELD(TD_EDC_CNT, SS_FIFO_HI_DED_COUNT)
+	},
+	{ "TD_CS_FIFO", SOC15_REG_ENTRY(GC, 0, mmTD_EDC_CNT),
+	  SOC15_REG_FIELD(TD_EDC_CNT, CS_FIFO_SED_COUNT),
+	  0, 0
+	},
+	{ "SQ_LDS_D", SOC15_REG_ENTRY(GC, 0, mmSQ_EDC_CNT),
+	  SOC15_REG_FIELD(SQ_EDC_CNT, LDS_D_SEC_COUNT),
+	  SOC15_REG_FIELD(SQ_EDC_CNT, LDS_D_DED_COUNT)
+	},
+	{ "SQ_LDS_I", SOC15_REG_ENTRY(GC, 0, mmSQ_EDC_CNT),
+	  SOC15_REG_FIELD(SQ_EDC_CNT, LDS_I_SEC_COUNT),
+	  SOC15_REG_FIELD(SQ_EDC_CNT, LDS_I_DED_COUNT)
+	},
+	{ "SQ_SGPR", SOC15_REG_ENTRY(GC, 0, mmSQ_EDC_CNT),
+	  SOC15_REG_FIELD(SQ_EDC_CNT, SGPR_SEC_COUNT),
+	  SOC15_REG_FIELD(SQ_EDC_CNT, SGPR_DED_COUNT)
+	},
+	{ "SQ_VGPR0", SOC15_REG_ENTRY(GC, 0, mmSQ_EDC_CNT),
+	  SOC15_REG_FIELD(SQ_EDC_CNT, VGPR0_SEC_COUNT),
+	  SOC15_REG_FIELD(SQ_EDC_CNT, VGPR0_DED_COUNT)
+	},
+	{ "SQ_VGPR1", SOC15_REG_ENTRY(GC, 0, mmSQ_EDC_CNT),
+	  SOC15_REG_FIELD(SQ_EDC_CNT, VGPR1_SEC_COUNT),
+	  SOC15_REG_FIELD(SQ_EDC_CNT, VGPR1_DED_COUNT)
+	},
+	{ "SQ_VGPR2", SOC15_REG_ENTRY(GC, 0, mmSQ_EDC_CNT),
+	  SOC15_REG_FIELD(SQ_EDC_CNT, VGPR2_SEC_COUNT),
+	  SOC15_REG_FIELD(SQ_EDC_CNT, VGPR2_DED_COUNT)
+	},
+	{ "SQ_VGPR3", SOC15_REG_ENTRY(GC, 0, mmSQ_EDC_CNT),
+	  SOC15_REG_FIELD(SQ_EDC_CNT, VGPR3_SEC_COUNT),
+	  SOC15_REG_FIELD(SQ_EDC_CNT, VGPR3_DED_COUNT)
+	},
 	{ "SQC_DATA_CU0_WRITE_DATA_BUF", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT),
-	  1, 6, REG_FIELD_MASK(SQC_EDC_CNT, DATA_CU0_WRITE_DATA_BUF_SEC_COUNT),
-	  REG_FIELD_MASK(SQC_EDC_CNT, DATA_CU0_WRITE_DATA_BUF_DED_COUNT) },
-	{ "SQC_DATA_CU0_UTCL1_LFIFO", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT), 1,
-	  6, REG_FIELD_MASK(SQC_EDC_CNT, DATA_CU0_UTCL1_LFIFO_SEC_COUNT),
-	  REG_FIELD_MASK(SQC_EDC_CNT, DATA_CU0_UTCL1_LFIFO_DED_COUNT) },
+	  SOC15_REG_FIELD(SQC_EDC_CNT, DATA_CU0_WRITE_DATA_BUF_SEC_COUNT),
+	  SOC15_REG_FIELD(SQC_EDC_CNT, DATA_CU0_WRITE_DATA_BUF_DED_COUNT)
+	},
+	{ "SQC_DATA_CU0_UTCL1_LFIFO", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT),
+	  SOC15_REG_FIELD(SQC_EDC_CNT, DATA_CU0_UTCL1_LFIFO_SEC_COUNT),
+	  SOC15_REG_FIELD(SQC_EDC_CNT, DATA_CU0_UTCL1_LFIFO_DED_COUNT)
+	},
 	{ "SQC_DATA_CU1_WRITE_DATA_BUF", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT),
-	  1, 6, REG_FIELD_MASK(SQC_EDC_CNT, DATA_CU1_WRITE_DATA_BUF_SEC_COUNT),
-	  REG_FIELD_MASK(SQC_EDC_CNT, DATA_CU1_WRITE_DATA_BUF_DED_COUNT) },
-	{ "SQC_DATA_CU1_UTCL1_LFIFO", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT), 1,
-	  6, REG_FIELD_MASK(SQC_EDC_CNT, DATA_CU1_UTCL1_LFIFO_SEC_COUNT),
-	  REG_FIELD_MASK(SQC_EDC_CNT, DATA_CU1_UTCL1_LFIFO_DED_COUNT) },
+	  SOC15_REG_FIELD(SQC_EDC_CNT, DATA_CU1_WRITE_DATA_BUF_SEC_COUNT),
+	  SOC15_REG_FIELD(SQC_EDC_CNT, DATA_CU1_WRITE_DATA_BUF_DED_COUNT)
+	},
+	{ "SQC_DATA_CU1_UTCL1_LFIFO", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT),
+	  SOC15_REG_FIELD(SQC_EDC_CNT, DATA_CU1_UTCL1_LFIFO_SEC_COUNT),
+	  SOC15_REG_FIELD(SQC_EDC_CNT, DATA_CU1_UTCL1_LFIFO_DED_COUNT)
+	},
 	{ "SQC_DATA_CU2_WRITE_DATA_BUF", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT),
-	  1, 6, REG_FIELD_MASK(SQC_EDC_CNT, DATA_CU2_WRITE_DATA_BUF_SEC_COUNT),
-	  REG_FIELD_MASK(SQC_EDC_CNT, DATA_CU2_WRITE_DATA_BUF_DED_COUNT) },
-	{ "SQC_DATA_CU2_UTCL1_LFIFO", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT), 1,
-	  6, REG_FIELD_MASK(SQC_EDC_CNT, DATA_CU2_UTCL1_LFIFO_SEC_COUNT),
-	  REG_FIELD_MASK(SQC_EDC_CNT, DATA_CU2_UTCL1_LFIFO_DED_COUNT) },
-	{ "SQC_INST_BANKA_TAG_RAM", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT2), 1,
-	  6, REG_FIELD_MASK(SQC_EDC_CNT2, INST_BANKA_TAG_RAM_SEC_COUNT),
-	  REG_FIELD_MASK(SQC_EDC_CNT2, INST_BANKA_TAG_RAM_DED_COUNT) },
-	{ "SQC_INST_BANKA_BANK_RAM", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT2), 1,
-	  6, REG_FIELD_MASK(SQC_EDC_CNT2, INST_BANKA_BANK_RAM_SEC_COUNT),
-	  REG_FIELD_MASK(SQC_EDC_CNT2, INST_BANKA_BANK_RAM_DED_COUNT) },
-	{ "SQC_DATA_BANKA_TAG_RAM", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT2), 1,
-	  6, REG_FIELD_MASK(SQC_EDC_CNT2, DATA_BANKA_TAG_RAM_SEC_COUNT),
-	  REG_FIELD_MASK(SQC_EDC_CNT2, DATA_BANKA_TAG_RAM_DED_COUNT) },
-	{ "SQC_DATA_BANKA_BANK_RAM", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT2), 1,
-	  6, REG_FIELD_MASK(SQC_EDC_CNT2, DATA_BANKA_BANK_RAM_SEC_COUNT),
-	  REG_FIELD_MASK(SQC_EDC_CNT2, DATA_BANKA_BANK_RAM_DED_COUNT) },
-	{ "SQC_INST_BANKA_UTCL1_MISS_FIFO",
-	  SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT2), 1, 6,
-	  REG_FIELD_MASK(SQC_EDC_CNT2, INST_BANKA_UTCL1_MISS_FIFO_SED_COUNT),
-	  0 },
-	{ "SQC_INST_BANKA_MISS_FIFO", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT2), 1,
-	  6, REG_FIELD_MASK(SQC_EDC_CNT2, INST_BANKA_MISS_FIFO_SED_COUNT), 0 },
-	{ "SQC_DATA_BANKA_HIT_FIFO", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT2), 1,
-	  6, REG_FIELD_MASK(SQC_EDC_CNT2, DATA_BANKA_HIT_FIFO_SED_COUNT), 0 },
-	{ "SQC_DATA_BANKA_MISS_FIFO", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT2), 1,
-	  6, REG_FIELD_MASK(SQC_EDC_CNT2, DATA_BANKA_MISS_FIFO_SED_COUNT), 0 },
-	{ "SQC_DATA_BANKA_DIRTY_BIT_RAM",
-	  SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT2), 1, 6,
-	  REG_FIELD_MASK(SQC_EDC_CNT2, DATA_BANKA_DIRTY_BIT_RAM_SED_COUNT), 0 },
-	{ "SQC_INST_UTCL1_LFIFO", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT2), 1, 6,
-	  REG_FIELD_MASK(SQC_EDC_CNT2, INST_UTCL1_LFIFO_SEC_COUNT),
-	  REG_FIELD_MASK(SQC_EDC_CNT2, INST_UTCL1_LFIFO_DED_COUNT) },
-	{ "SQC_INST_BANKB_TAG_RAM", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT3), 1,
-	  6, REG_FIELD_MASK(SQC_EDC_CNT3, INST_BANKB_TAG_RAM_SEC_COUNT),
-	  REG_FIELD_MASK(SQC_EDC_CNT3, INST_BANKB_TAG_RAM_DED_COUNT) },
-	{ "SQC_INST_BANKB_BANK_RAM", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT3), 1,
-	  6, REG_FIELD_MASK(SQC_EDC_CNT3, INST_BANKB_BANK_RAM_SEC_COUNT),
-	  REG_FIELD_MASK(SQC_EDC_CNT3, INST_BANKB_BANK_RAM_DED_COUNT) },
-	{ "SQC_DATA_BANKB_TAG_RAM", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT3), 1,
-	  6, REG_FIELD_MASK(SQC_EDC_CNT3, DATA_BANKB_TAG_RAM_SEC_COUNT),
-	  REG_FIELD_MASK(SQC_EDC_CNT3, DATA_BANKB_TAG_RAM_DED_COUNT) },
-	{ "SQC_DATA_BANKB_BANK_RAM", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT3), 1,
-	  6, REG_FIELD_MASK(SQC_EDC_CNT3, DATA_BANKB_BANK_RAM_SEC_COUNT),
-	  REG_FIELD_MASK(SQC_EDC_CNT3, DATA_BANKB_BANK_RAM_DED_COUNT) },
-	{ "SQC_INST_BANKB_UTCL1_MISS_FIFO",
-	  SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT3), 1, 6,
-	  REG_FIELD_MASK(SQC_EDC_CNT3, INST_BANKB_UTCL1_MISS_FIFO_SED_COUNT),
-	  0 },
-	{ "SQC_INST_BANKB_MISS_FIFO", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT3), 1,
-	  6, REG_FIELD_MASK(SQC_EDC_CNT3, INST_BANKB_MISS_FIFO_SED_COUNT), 0 },
-	{ "SQC_DATA_BANKB_HIT_FIFO", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT3), 1,
-	  6, REG_FIELD_MASK(SQC_EDC_CNT3, DATA_BANKB_HIT_FIFO_SED_COUNT), 0 },
-	{ "SQC_DATA_BANKB_MISS_FIFO", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT3), 1,
-	  6, REG_FIELD_MASK(SQC_EDC_CNT3, DATA_BANKB_MISS_FIFO_SED_COUNT), 0 },
-	{ "SQC_DATA_BANKB_DIRTY_BIT_RAM",
-	  SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT3), 1, 6,
-	  REG_FIELD_MASK(SQC_EDC_CNT3, DATA_BANKB_DIRTY_BIT_RAM_SED_COUNT), 0 },
-	{ "EA_DRAMRD_CMDMEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT), 0, 32,
-	  REG_FIELD_MASK(GCEA_EDC_CNT, DRAMRD_CMDMEM_SEC_COUNT),
-	  REG_FIELD_MASK(GCEA_EDC_CNT, DRAMRD_CMDMEM_DED_COUNT) },
-	{ "EA_DRAMWR_CMDMEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT), 0, 32,
-	  REG_FIELD_MASK(GCEA_EDC_CNT, DRAMWR_CMDMEM_SEC_COUNT),
-	  REG_FIELD_MASK(GCEA_EDC_CNT, DRAMWR_CMDMEM_DED_COUNT) },
-	{ "EA_DRAMWR_DATAMEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT), 0, 32,
-	  REG_FIELD_MASK(GCEA_EDC_CNT, DRAMWR_DATAMEM_SEC_COUNT),
-	  REG_FIELD_MASK(GCEA_EDC_CNT, DRAMWR_DATAMEM_DED_COUNT) },
-	{ "EA_RRET_TAGMEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT), 0, 32,
-	  REG_FIELD_MASK(GCEA_EDC_CNT, RRET_TAGMEM_SEC_COUNT),
-	  REG_FIELD_MASK(GCEA_EDC_CNT, RRET_TAGMEM_DED_COUNT) },
-	{ "EA_WRET_TAGMEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT), 0, 32,
-	  REG_FIELD_MASK(GCEA_EDC_CNT, WRET_TAGMEM_SEC_COUNT),
-	  REG_FIELD_MASK(GCEA_EDC_CNT, WRET_TAGMEM_DED_COUNT) },
-	{ "EA_DRAMRD_PAGEMEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT), 0, 32,
-	  REG_FIELD_MASK(GCEA_EDC_CNT, DRAMRD_PAGEMEM_SED_COUNT), 0 },
-	{ "EA_DRAMWR_PAGEMEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT), 0, 32,
-	  REG_FIELD_MASK(GCEA_EDC_CNT, DRAMWR_PAGEMEM_SED_COUNT), 0 },
-	{ "EA_IORD_CMDMEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT), 0, 32,
-	  REG_FIELD_MASK(GCEA_EDC_CNT, IORD_CMDMEM_SED_COUNT), 0 },
-	{ "EA_IOWR_CMDMEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT), 0, 32,
-	  REG_FIELD_MASK(GCEA_EDC_CNT, IOWR_CMDMEM_SED_COUNT), 0 },
-	{ "EA_IOWR_DATAMEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT), 0, 32,
-	  REG_FIELD_MASK(GCEA_EDC_CNT, IOWR_DATAMEM_SED_COUNT), 0 },
-	{ "GMIRD_CMDMEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT2), 0, 32,
-	  REG_FIELD_MASK(GCEA_EDC_CNT2, GMIRD_CMDMEM_SEC_COUNT),
-	  REG_FIELD_MASK(GCEA_EDC_CNT2, GMIRD_CMDMEM_DED_COUNT) },
-	{ "GMIWR_CMDMEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT2), 0, 32,
-	  REG_FIELD_MASK(GCEA_EDC_CNT2, GMIWR_CMDMEM_SEC_COUNT),
-	  REG_FIELD_MASK(GCEA_EDC_CNT2, GMIWR_CMDMEM_DED_COUNT) },
-	{ "GMIWR_DATAMEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT2), 0, 32,
-	  REG_FIELD_MASK(GCEA_EDC_CNT2, GMIWR_DATAMEM_SEC_COUNT),
-	  REG_FIELD_MASK(GCEA_EDC_CNT2, GMIWR_DATAMEM_DED_COUNT) },
-	{ "GMIRD_PAGEMEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT2), 0, 32,
-	  REG_FIELD_MASK(GCEA_EDC_CNT2, GMIRD_PAGEMEM_SED_COUNT), 0 },
-	{ "GMIWR_PAGEMEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT2), 0, 32,
-	  REG_FIELD_MASK(GCEA_EDC_CNT2, GMIWR_PAGEMEM_SED_COUNT), 0 },
-	{ "MAM_D0MEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT2), 0, 32,
-	  REG_FIELD_MASK(GCEA_EDC_CNT2, MAM_D0MEM_SED_COUNT), 0 },
-	{ "MAM_D1MEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT2), 0, 32,
-	  REG_FIELD_MASK(GCEA_EDC_CNT2, MAM_D1MEM_SED_COUNT), 0 },
-	{ "MAM_D2MEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT2), 0, 32,
-	  REG_FIELD_MASK(GCEA_EDC_CNT2, MAM_D2MEM_SED_COUNT), 0 },
-	{ "MAM_D3MEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT2), 0, 32,
-	  REG_FIELD_MASK(GCEA_EDC_CNT2, MAM_D3MEM_SED_COUNT), 0 },
+	  SOC15_REG_FIELD(SQC_EDC_CNT, DATA_CU2_WRITE_DATA_BUF_SEC_COUNT),
+	  SOC15_REG_FIELD(SQC_EDC_CNT, DATA_CU2_WRITE_DATA_BUF_DED_COUNT)
+	},
+	{ "SQC_DATA_CU2_UTCL1_LFIFO", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT),
+	  SOC15_REG_FIELD(SQC_EDC_CNT, DATA_CU2_UTCL1_LFIFO_SEC_COUNT),
+	  SOC15_REG_FIELD(SQC_EDC_CNT, DATA_CU2_UTCL1_LFIFO_DED_COUNT)
+	},
+	{ "SQC_INST_BANKA_TAG_RAM", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT2),
+	  SOC15_REG_FIELD(SQC_EDC_CNT2, INST_BANKA_TAG_RAM_SEC_COUNT),
+	  SOC15_REG_FIELD(SQC_EDC_CNT2, INST_BANKA_TAG_RAM_DED_COUNT)
+	},
+	{ "SQC_INST_BANKA_BANK_RAM", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT2),
+	  SOC15_REG_FIELD(SQC_EDC_CNT2, INST_BANKA_BANK_RAM_SEC_COUNT),
+	  SOC15_REG_FIELD(SQC_EDC_CNT2, INST_BANKA_BANK_RAM_DED_COUNT)
+	},
+	{ "SQC_DATA_BANKA_TAG_RAM", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT2),
+	  SOC15_REG_FIELD(SQC_EDC_CNT2, DATA_BANKA_TAG_RAM_SEC_COUNT),
+	  SOC15_REG_FIELD(SQC_EDC_CNT2, DATA_BANKA_TAG_RAM_DED_COUNT)
+	},
+	{ "SQC_DATA_BANKA_BANK_RAM", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT2),
+	  SOC15_REG_FIELD(SQC_EDC_CNT2, DATA_BANKA_BANK_RAM_SEC_COUNT),
+	  SOC15_REG_FIELD(SQC_EDC_CNT2, DATA_BANKA_BANK_RAM_DED_COUNT)
+	},
+	{ "SQC_INST_BANKA_UTCL1_MISS_FIFO", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT2),
+	  SOC15_REG_FIELD(SQC_EDC_CNT2, INST_BANKA_UTCL1_MISS_FIFO_SED_COUNT),
+	  0, 0
+	},
+	{ "SQC_INST_BANKA_MISS_FIFO", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT2),
+	  SOC15_REG_FIELD(SQC_EDC_CNT2, INST_BANKA_MISS_FIFO_SED_COUNT),
+	  0, 0
+	},
+	{ "SQC_DATA_BANKA_HIT_FIFO", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT2),
+	  SOC15_REG_FIELD(SQC_EDC_CNT2, DATA_BANKA_HIT_FIFO_SED_COUNT),
+	  0, 0
+	},
+	{ "SQC_DATA_BANKA_MISS_FIFO", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT2),
+	  SOC15_REG_FIELD(SQC_EDC_CNT2, DATA_BANKA_MISS_FIFO_SED_COUNT),
+	  0, 0
+	},
+	{ "SQC_DATA_BANKA_DIRTY_BIT_RAM", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT2),
+	  SOC15_REG_FIELD(SQC_EDC_CNT2, DATA_BANKA_DIRTY_BIT_RAM_SED_COUNT),
+	  0, 0
+	},
+	{ "SQC_INST_UTCL1_LFIFO", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT2),
+	  SOC15_REG_FIELD(SQC_EDC_CNT2, INST_UTCL1_LFIFO_SEC_COUNT),
+	  SOC15_REG_FIELD(SQC_EDC_CNT2, INST_UTCL1_LFIFO_DED_COUNT)
+	},
+	{ "SQC_INST_BANKB_TAG_RAM", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT3),
+	  SOC15_REG_FIELD(SQC_EDC_CNT3, INST_BANKB_TAG_RAM_SEC_COUNT),
+	  SOC15_REG_FIELD(SQC_EDC_CNT3, INST_BANKB_TAG_RAM_DED_COUNT)
+	},
+	{ "SQC_INST_BANKB_BANK_RAM", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT3),
+	  SOC15_REG_FIELD(SQC_EDC_CNT3, INST_BANKB_BANK_RAM_SEC_COUNT),
+	  SOC15_REG_FIELD(SQC_EDC_CNT3, INST_BANKB_BANK_RAM_DED_COUNT)
+	},
+	{ "SQC_DATA_BANKB_TAG_RAM", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT3),
+	  SOC15_REG_FIELD(SQC_EDC_CNT3, DATA_BANKB_TAG_RAM_SEC_COUNT),
+	  SOC15_REG_FIELD(SQC_EDC_CNT3, DATA_BANKB_TAG_RAM_DED_COUNT)
+	},
+	{ "SQC_DATA_BANKB_BANK_RAM", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT3),
+	  SOC15_REG_FIELD(SQC_EDC_CNT3, DATA_BANKB_BANK_RAM_SEC_COUNT),
+	  SOC15_REG_FIELD(SQC_EDC_CNT3, DATA_BANKB_BANK_RAM_DED_COUNT)
+	},
+	{ "SQC_INST_BANKB_UTCL1_MISS_FIFO", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT3),
+	  SOC15_REG_FIELD(SQC_EDC_CNT3, INST_BANKB_UTCL1_MISS_FIFO_SED_COUNT),
+	  0, 0
+	},
+	{ "SQC_INST_BANKB_MISS_FIFO", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT3),
+	  SOC15_REG_FIELD(SQC_EDC_CNT3, INST_BANKB_MISS_FIFO_SED_COUNT),
+	  0, 0
+	},
+	{ "SQC_DATA_BANKB_HIT_FIFO", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT3),
+	  SOC15_REG_FIELD(SQC_EDC_CNT3, DATA_BANKB_HIT_FIFO_SED_COUNT),
+	  0, 0
+	},
+	{ "SQC_DATA_BANKB_MISS_FIFO", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT3),
+	  SOC15_REG_FIELD(SQC_EDC_CNT3, DATA_BANKB_MISS_FIFO_SED_COUNT),
+	  0, 0
+	},
+	{ "SQC_DATA_BANKB_DIRTY_BIT_RAM", SOC15_REG_ENTRY(GC, 0, mmSQC_EDC_CNT3),
+	  SOC15_REG_FIELD(SQC_EDC_CNT3, DATA_BANKB_DIRTY_BIT_RAM_SED_COUNT),
+	  0, 0
+	},
+	{ "EA_DRAMRD_CMDMEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT),
+	  SOC15_REG_FIELD(GCEA_EDC_CNT, DRAMRD_CMDMEM_SEC_COUNT),
+	  SOC15_REG_FIELD(GCEA_EDC_CNT, DRAMRD_CMDMEM_DED_COUNT)
+	},
+	{ "EA_DRAMWR_CMDMEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT),
+	  SOC15_REG_FIELD(GCEA_EDC_CNT, DRAMWR_CMDMEM_SEC_COUNT),
+	  SOC15_REG_FIELD(GCEA_EDC_CNT, DRAMWR_CMDMEM_DED_COUNT)
+	},
+	{ "EA_DRAMWR_DATAMEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT),
+	  SOC15_REG_FIELD(GCEA_EDC_CNT, DRAMWR_DATAMEM_SEC_COUNT),
+	  SOC15_REG_FIELD(GCEA_EDC_CNT, DRAMWR_DATAMEM_DED_COUNT)
+	},
+	{ "EA_RRET_TAGMEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT),
+	  SOC15_REG_FIELD(GCEA_EDC_CNT, RRET_TAGMEM_SEC_COUNT),
+	  SOC15_REG_FIELD(GCEA_EDC_CNT, RRET_TAGMEM_DED_COUNT)
+	},
+	{ "EA_WRET_TAGMEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT),
+	  SOC15_REG_FIELD(GCEA_EDC_CNT, WRET_TAGMEM_SEC_COUNT),
+	  SOC15_REG_FIELD(GCEA_EDC_CNT, WRET_TAGMEM_DED_COUNT)
+	},
+	{ "EA_DRAMRD_PAGEMEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT),
+	  SOC15_REG_FIELD(GCEA_EDC_CNT, DRAMRD_PAGEMEM_SED_COUNT),
+	  0, 0
+	},
+	{ "EA_DRAMWR_PAGEMEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT),
+	  SOC15_REG_FIELD(GCEA_EDC_CNT, DRAMWR_PAGEMEM_SED_COUNT),
+	  0, 0
+	},
+	{ "EA_IORD_CMDMEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT),
+	  SOC15_REG_FIELD(GCEA_EDC_CNT, IORD_CMDMEM_SED_COUNT),
+	  0, 0
+	},
+	{ "EA_IOWR_CMDMEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT),
+	  SOC15_REG_FIELD(GCEA_EDC_CNT, IOWR_CMDMEM_SED_COUNT),
+	  0, 0
+	},
+	{ "EA_IOWR_DATAMEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT),
+	  SOC15_REG_FIELD(GCEA_EDC_CNT, IOWR_DATAMEM_SED_COUNT),
+	  0, 0
+	},
+	{ "GMIRD_CMDMEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT2),
+	  SOC15_REG_FIELD(GCEA_EDC_CNT2, GMIRD_CMDMEM_SEC_COUNT),
+	  SOC15_REG_FIELD(GCEA_EDC_CNT2, GMIRD_CMDMEM_DED_COUNT)
+	},
+	{ "GMIWR_CMDMEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT2),
+	  SOC15_REG_FIELD(GCEA_EDC_CNT2, GMIWR_CMDMEM_SEC_COUNT),
+	  SOC15_REG_FIELD(GCEA_EDC_CNT2, GMIWR_CMDMEM_DED_COUNT)
+	},
+	{ "GMIWR_DATAMEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT2),
+	  SOC15_REG_FIELD(GCEA_EDC_CNT2, GMIWR_DATAMEM_SEC_COUNT),
+	  SOC15_REG_FIELD(GCEA_EDC_CNT2, GMIWR_DATAMEM_DED_COUNT)
+	},
+	{ "GMIRD_PAGEMEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT2),
+	  SOC15_REG_FIELD(GCEA_EDC_CNT2, GMIRD_PAGEMEM_SED_COUNT),
+	  0, 0
+	},
+	{ "GMIWR_PAGEMEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT2),
+	  SOC15_REG_FIELD(GCEA_EDC_CNT2, GMIWR_PAGEMEM_SED_COUNT),
+	  0, 0
+	},
+	{ "MAM_D0MEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT2),
+	  SOC15_REG_FIELD(GCEA_EDC_CNT2, MAM_D0MEM_SED_COUNT),
+	  0, 0
+	},
+	{ "MAM_D1MEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT2),
+	  SOC15_REG_FIELD(GCEA_EDC_CNT2, MAM_D1MEM_SED_COUNT),
+	  0, 0
+	},
+	{ "MAM_D2MEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT2),
+	  SOC15_REG_FIELD(GCEA_EDC_CNT2, MAM_D2MEM_SED_COUNT),
+	  0, 0
+	},
+	{ "MAM_D3MEM", SOC15_REG_ENTRY(GC, 0, mmGCEA_EDC_CNT2),
+	  SOC15_REG_FIELD(GCEA_EDC_CNT2, MAM_D3MEM_SED_COUNT),
+	  0, 0
+	}
 };
 
 static int gfx_v9_0_ras_error_inject(struct amdgpu_device *adev,
@@ -6078,14 +5972,217 @@ static int gfx_v9_0_ras_error_inject(struct amdgpu_device *adev,
 	return ret;
 }
 
+static const char *vml2_mems[] = {
+	"UTC_VML2_BANK_CACHE_0_BIGK_MEM0",
+	"UTC_VML2_BANK_CACHE_0_BIGK_MEM1",
+	"UTC_VML2_BANK_CACHE_0_4K_MEM0",
+	"UTC_VML2_BANK_CACHE_0_4K_MEM1",
+	"UTC_VML2_BANK_CACHE_1_BIGK_MEM0",
+	"UTC_VML2_BANK_CACHE_1_BIGK_MEM1",
+	"UTC_VML2_BANK_CACHE_1_4K_MEM0",
+	"UTC_VML2_BANK_CACHE_1_4K_MEM1",
+	"UTC_VML2_BANK_CACHE_2_BIGK_MEM0",
+	"UTC_VML2_BANK_CACHE_2_BIGK_MEM1",
+	"UTC_VML2_BANK_CACHE_2_4K_MEM0",
+	"UTC_VML2_BANK_CACHE_2_4K_MEM1",
+	"UTC_VML2_BANK_CACHE_3_BIGK_MEM0",
+	"UTC_VML2_BANK_CACHE_3_BIGK_MEM1",
+	"UTC_VML2_BANK_CACHE_3_4K_MEM0",
+	"UTC_VML2_BANK_CACHE_3_4K_MEM1",
+};
+
+static const char *vml2_walker_mems[] = {
+	"UTC_VML2_CACHE_PDE0_MEM0",
+	"UTC_VML2_CACHE_PDE0_MEM1",
+	"UTC_VML2_CACHE_PDE1_MEM0",
+	"UTC_VML2_CACHE_PDE1_MEM1",
+	"UTC_VML2_CACHE_PDE2_MEM0",
+	"UTC_VML2_CACHE_PDE2_MEM1",
+	"UTC_VML2_RDIF_LOG_FIFO",
+};
+
+static const char *atc_l2_cache_2m_mems[] = {
+	"UTC_ATCL2_CACHE_2M_BANK0_WAY0_MEM",
+	"UTC_ATCL2_CACHE_2M_BANK0_WAY1_MEM",
+	"UTC_ATCL2_CACHE_2M_BANK1_WAY0_MEM",
+	"UTC_ATCL2_CACHE_2M_BANK1_WAY1_MEM",
+};
+
+static const char *atc_l2_cache_4k_mems[] = {
+	"UTC_ATCL2_CACHE_4K_BANK0_WAY0_MEM0",
+	"UTC_ATCL2_CACHE_4K_BANK0_WAY0_MEM1",
+	"UTC_ATCL2_CACHE_4K_BANK0_WAY0_MEM2",
+	"UTC_ATCL2_CACHE_4K_BANK0_WAY0_MEM3",
+	"UTC_ATCL2_CACHE_4K_BANK0_WAY0_MEM4",
+	"UTC_ATCL2_CACHE_4K_BANK0_WAY0_MEM5",
+	"UTC_ATCL2_CACHE_4K_BANK0_WAY0_MEM6",
+	"UTC_ATCL2_CACHE_4K_BANK0_WAY0_MEM7",
+	"UTC_ATCL2_CACHE_4K_BANK0_WAY1_MEM0",
+	"UTC_ATCL2_CACHE_4K_BANK0_WAY1_MEM1",
+	"UTC_ATCL2_CACHE_4K_BANK0_WAY1_MEM2",
+	"UTC_ATCL2_CACHE_4K_BANK0_WAY1_MEM3",
+	"UTC_ATCL2_CACHE_4K_BANK0_WAY1_MEM4",
+	"UTC_ATCL2_CACHE_4K_BANK0_WAY1_MEM5",
+	"UTC_ATCL2_CACHE_4K_BANK0_WAY1_MEM6",
+	"UTC_ATCL2_CACHE_4K_BANK0_WAY1_MEM7",
+	"UTC_ATCL2_CACHE_4K_BANK1_WAY0_MEM0",
+	"UTC_ATCL2_CACHE_4K_BANK1_WAY0_MEM1",
+	"UTC_ATCL2_CACHE_4K_BANK1_WAY0_MEM2",
+	"UTC_ATCL2_CACHE_4K_BANK1_WAY0_MEM3",
+	"UTC_ATCL2_CACHE_4K_BANK1_WAY0_MEM4",
+	"UTC_ATCL2_CACHE_4K_BANK1_WAY0_MEM5",
+	"UTC_ATCL2_CACHE_4K_BANK1_WAY0_MEM6",
+	"UTC_ATCL2_CACHE_4K_BANK1_WAY0_MEM7",
+	"UTC_ATCL2_CACHE_4K_BANK1_WAY1_MEM0",
+	"UTC_ATCL2_CACHE_4K_BANK1_WAY1_MEM1",
+	"UTC_ATCL2_CACHE_4K_BANK1_WAY1_MEM2",
+	"UTC_ATCL2_CACHE_4K_BANK1_WAY1_MEM3",
+	"UTC_ATCL2_CACHE_4K_BANK1_WAY1_MEM4",
+	"UTC_ATCL2_CACHE_4K_BANK1_WAY1_MEM5",
+	"UTC_ATCL2_CACHE_4K_BANK1_WAY1_MEM6",
+	"UTC_ATCL2_CACHE_4K_BANK1_WAY1_MEM7",
+};
+
+static int gfx_v9_0_query_utc_edc_status(struct amdgpu_device *adev,
+					 struct ras_err_data *err_data)
+{
+	uint32_t i, data;
+	uint32_t sec_count, ded_count;
+
+	WREG32_SOC15(GC, 0, mmVM_L2_MEM_ECC_INDEX, 255);
+	WREG32_SOC15(GC, 0, mmVM_L2_MEM_ECC_CNT, 0);
+	WREG32_SOC15(GC, 0, mmVM_L2_WALKER_MEM_ECC_INDEX, 255);
+	WREG32_SOC15(GC, 0, mmVM_L2_WALKER_MEM_ECC_CNT, 0);
+	WREG32_SOC15(GC, 0, mmATC_L2_CACHE_2M_EDC_INDEX, 255);
+	WREG32_SOC15(GC, 0, mmATC_L2_CACHE_2M_EDC_CNT, 0);
+	WREG32_SOC15(GC, 0, mmATC_L2_CACHE_4K_EDC_INDEX, 255);
+	WREG32_SOC15(GC, 0, mmATC_L2_CACHE_4K_EDC_CNT, 0);
+
+	for (i = 0; i < 16; i++) {
+		WREG32_SOC15(GC, 0, mmVM_L2_MEM_ECC_INDEX, i);
+		data = RREG32_SOC15(GC, 0, mmVM_L2_MEM_ECC_CNT);
+
+		sec_count = REG_GET_FIELD(data, VM_L2_MEM_ECC_CNT, SEC_COUNT);
+		if (sec_count) {
+			DRM_INFO("Instance[%d]: SubBlock %s, SEC %d\n", i,
+				 vml2_mems[i], sec_count);
+			err_data->ce_count += sec_count;
+		}
+
+		ded_count = REG_GET_FIELD(data, VM_L2_MEM_ECC_CNT, DED_COUNT);
+		if (ded_count) {
+			DRM_INFO("Instance[%d]: SubBlock %s, DED %d\n", i,
+				 vml2_mems[i], ded_count);
+			err_data->ue_count += ded_count;
+		}
+	}
+
+	for (i = 0; i < 7; i++) {
+		WREG32_SOC15(GC, 0, mmVM_L2_WALKER_MEM_ECC_INDEX, i);
+		data = RREG32_SOC15(GC, 0, mmVM_L2_WALKER_MEM_ECC_CNT);
+
+		sec_count = REG_GET_FIELD(data, VM_L2_WALKER_MEM_ECC_CNT,
+						SEC_COUNT);
+		if (sec_count) {
+			DRM_INFO("Instance[%d]: SubBlock %s, SEC %d\n", i,
+				 vml2_walker_mems[i], sec_count);
+			err_data->ce_count += sec_count;
+		}
+
+		ded_count = REG_GET_FIELD(data, VM_L2_WALKER_MEM_ECC_CNT,
+						DED_COUNT);
+		if (ded_count) {
+			DRM_INFO("Instance[%d]: SubBlock %s, DED %d\n", i,
+				 vml2_walker_mems[i], ded_count);
+			err_data->ue_count += ded_count;
+		}
+	}
+
+	for (i = 0; i < 4; i++) {
+		WREG32_SOC15(GC, 0, mmATC_L2_CACHE_2M_EDC_INDEX, i);
+		data = RREG32_SOC15(GC, 0, mmATC_L2_CACHE_2M_EDC_CNT);
+
+		sec_count = (data & 0x00006000L) >> 0xd;
+		if (sec_count) {
+			DRM_INFO("Instance[%d]: SubBlock %s, SEC %d\n", i,
+				 atc_l2_cache_2m_mems[i], sec_count);
+			err_data->ce_count += sec_count;
+		}
+	}
+
+	for (i = 0; i < 32; i++) {
+		WREG32_SOC15(GC, 0, mmATC_L2_CACHE_4K_EDC_INDEX, i);
+		data = RREG32_SOC15(GC, 0, mmATC_L2_CACHE_4K_EDC_CNT);
+
+		sec_count = (data & 0x00006000L) >> 0xd;
+		if (sec_count) {
+			DRM_INFO("Instance[%d]: SubBlock %s, SEC %d\n", i,
+				 atc_l2_cache_4k_mems[i], sec_count);
+			err_data->ce_count += sec_count;
+		}
+
+		ded_count = (data & 0x00018000L) >> 0xf;
+		if (ded_count) {
+			DRM_INFO("Instance[%d]: SubBlock %s, DED %d\n", i,
+				 atc_l2_cache_4k_mems[i], ded_count);
+			err_data->ue_count += ded_count;
+		}
+	}
+
+	WREG32_SOC15(GC, 0, mmVM_L2_MEM_ECC_INDEX, 255);
+	WREG32_SOC15(GC, 0, mmVM_L2_WALKER_MEM_ECC_INDEX, 255);
+	WREG32_SOC15(GC, 0, mmATC_L2_CACHE_2M_EDC_INDEX, 255);
+	WREG32_SOC15(GC, 0, mmATC_L2_CACHE_4K_EDC_INDEX, 255);
+
+	return 0;
+}
+
+static int __get_ras_error_count(const struct soc15_reg_entry *reg,
+	uint32_t se_id, uint32_t inst_id, uint32_t value,
+	uint32_t *sec_count, uint32_t *ded_count)
+{
+	uint32_t i;
+	uint32_t sec_cnt, ded_cnt;
+
+	for (i = 0; i < ARRAY_SIZE(ras_subblock_regs); i++) {
+		if(ras_subblock_regs[i].reg_offset != reg->reg_offset ||
+			ras_subblock_regs[i].seg != reg->seg ||
+			ras_subblock_regs[i].inst != reg->inst)
+			continue;
+
+		sec_cnt = (value &
+				ras_subblock_regs[i].sec_count_mask) >>
+				ras_subblock_regs[i].sec_count_shift;
+		if (sec_cnt) {
+			DRM_INFO("GFX SubBlock %s, Instance[%d][%d], SEC %d\n",
+				ras_subblock_regs[i].name,
+				se_id, inst_id,
+				sec_cnt);
+			*sec_count += sec_cnt;
+		}
+
+		ded_cnt = (value &
+				ras_subblock_regs[i].ded_count_mask) >>
+				ras_subblock_regs[i].ded_count_shift;
+		if (ded_cnt) {
+			DRM_INFO("GFX SubBlock %s, Instance[%d][%d], DED %d\n",
+				ras_subblock_regs[i].name,
+				se_id, inst_id,
+				ded_cnt);
+			*ded_count += ded_cnt;
+		}
+	}
+
+	return 0;
+}
+
 static int gfx_v9_0_query_ras_error_count(struct amdgpu_device *adev,
 					  void *ras_error_status)
 {
 	struct ras_err_data *err_data = (struct ras_err_data *)ras_error_status;
-	uint32_t sec_count, ded_count;
-	uint32_t i;
+	uint32_t sec_count = 0, ded_count = 0;
+	uint32_t i, j, k;
 	uint32_t reg_value;
-	uint32_t se_id, instance_id;
 
 	if (adev->asic_type != CHIP_VEGA20)
 		return -EINVAL;
@@ -6094,71 +6191,29 @@ static int gfx_v9_0_query_ras_error_count(struct amdgpu_device *adev,
 	err_data->ce_count = 0;
 
 	mutex_lock(&adev->grbm_idx_mutex);
-	for (se_id = 0; se_id < adev->gfx.config.max_shader_engines; se_id++) {
-		for (instance_id = 0; instance_id < 256; instance_id++) {
-			for (i = 0;
-			     i < sizeof(gfx_ras_edc_regs) / sizeof(gfx_ras_edc_regs[0]);
-			     i++) {
-				if (se_id != 0 &&
-				    !gfx_ras_edc_regs[i].per_se_instance)
-					continue;
-				if (instance_id >= gfx_ras_edc_regs[i].num_instance)
-					continue;
 
-				gfx_v9_0_select_se_sh(adev, se_id, 0,
-						      instance_id);
-
-				reg_value = RREG32(
-					adev->reg_offset[gfx_ras_edc_regs[i].ip]
-							[gfx_ras_edc_regs[i].inst]
-							[gfx_ras_edc_regs[i].seg] +
-					gfx_ras_edc_regs[i].reg_offset);
-				sec_count = reg_value &
-					    gfx_ras_edc_regs[i].sec_count_mask;
-				ded_count = reg_value &
-					    gfx_ras_edc_regs[i].ded_count_mask;
-				if (sec_count) {
-					DRM_INFO(
-						"Instance[%d][%d]: SubBlock %s, SEC %d\n",
-						se_id, instance_id,
-						gfx_ras_edc_regs[i].name,
-						sec_count);
-					err_data->ce_count++;
-				}
-
-				if (ded_count) {
-					DRM_INFO(
-						"Instance[%d][%d]: SubBlock %s, DED %d\n",
-						se_id, instance_id,
-						gfx_ras_edc_regs[i].name,
-						ded_count);
-					err_data->ue_count++;
-				}
+	for (i = 0; i < ARRAY_SIZE(sec_ded_counter_registers); i++) {
+		for (j = 0; j < sec_ded_counter_registers[i].se_num; j++) {
+			for (k = 0; k < sec_ded_counter_registers[i].instance; k++) {
+				gfx_v9_0_select_se_sh(adev, j, 0, k);
+				reg_value =
+					RREG32(SOC15_REG_ENTRY_OFFSET(sec_ded_counter_registers[i]));
+				if (reg_value)
+					__get_ras_error_count(&sec_ded_counter_registers[i],
+							j, k, reg_value,
+							&sec_count, &ded_count);
 			}
 		}
 	}
-	gfx_v9_0_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff);
-	mutex_unlock(&adev->grbm_idx_mutex);
 
-	return 0;
-}
+	err_data->ce_count += sec_count;
+	err_data->ue_count += ded_count;
 
-static int gfx_v9_0_cp_ecc_error_irq(struct amdgpu_device *adev,
-				  struct amdgpu_irq_src *source,
-				  struct amdgpu_iv_entry *entry)
-{
-	struct ras_common_if *ras_if = adev->gfx.ras_if;
-	struct ras_dispatch_if ih_data = {
-		.entry = entry,
-	};
-
-	if (!ras_if)
-		return 0;
+	gfx_v9_0_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff);
+	mutex_unlock(&adev->grbm_idx_mutex);
 
-	ih_data.head = *ras_if;
+	gfx_v9_0_query_utc_edc_status(adev, err_data);
 
-	DRM_ERROR("CP ECC ERROR IRQ\n");
-	amdgpu_ras_interrupt_dispatch(adev, &ih_data);
 	return 0;
 }
 
@@ -6325,7 +6380,7 @@ static const struct amdgpu_irq_src_funcs gfx_v9_0_priv_inst_irq_funcs = {
 
 static const struct amdgpu_irq_src_funcs gfx_v9_0_cp_ecc_error_irq_funcs = {
 	.set = gfx_v9_0_set_cp_ecc_error_state,
-	.process = gfx_v9_0_cp_ecc_error_irq,
+	.process = amdgpu_gfx_cp_ecc_error_irq,
 };
 
 
diff --git a/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c b/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c
index 6ce37ce77d14..e91bd7945777 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c
@@ -178,6 +178,8 @@ static void gfxhub_v1_0_enable_system_domain(struct amdgpu_device *adev)
 	tmp = RREG32_SOC15(GC, 0, mmVM_CONTEXT0_CNTL);
 	tmp = REG_SET_FIELD(tmp, VM_CONTEXT0_CNTL, ENABLE_CONTEXT, 1);
 	tmp = REG_SET_FIELD(tmp, VM_CONTEXT0_CNTL, PAGE_TABLE_DEPTH, 0);
+	tmp = REG_SET_FIELD(tmp, VM_CONTEXT0_CNTL,
+			    RETRY_PERMISSION_OR_INVALID_PAGE_FAULT, 0);
 	WREG32_SOC15(GC, 0, mmVM_CONTEXT0_CNTL, tmp);
 }
 
@@ -365,6 +367,8 @@ void gfxhub_v1_0_init(struct amdgpu_device *adev)
 	hub->ctx0_ptb_addr_hi32 =
 		SOC15_REG_OFFSET(GC, 0,
 				 mmVM_CONTEXT0_PAGE_TABLE_BASE_ADDR_HI32);
+	hub->vm_inv_eng0_sem =
+		SOC15_REG_OFFSET(GC, 0, mmVM_INVALIDATE_ENG0_SEM);
 	hub->vm_inv_eng0_req =
 		SOC15_REG_OFFSET(GC, 0, mmVM_INVALIDATE_ENG0_REQ);
 	hub->vm_inv_eng0_ack =
diff --git a/drivers/gpu/drm/amd/amdgpu/gfxhub_v2_0.c b/drivers/gpu/drm/amd/amdgpu/gfxhub_v2_0.c
index 8b789f750b72..b70c7b483c24 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfxhub_v2_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfxhub_v2_0.c
@@ -46,21 +46,25 @@ u64 gfxhub_v2_0_get_mc_fb_offset(struct amdgpu_device *adev)
 	return (u64)RREG32_SOC15(GC, 0, mmGCMC_VM_FB_OFFSET) << 24;
 }
 
-static void gfxhub_v2_0_init_gart_pt_regs(struct amdgpu_device *adev)
+void gfxhub_v2_0_setup_vm_pt_regs(struct amdgpu_device *adev, uint32_t vmid,
+				uint64_t page_table_base)
 {
-	uint64_t value = amdgpu_gmc_pd_addr(adev->gart.bo);
+	/* two registers distance between mmGCVM_CONTEXT0_* to mmGCVM_CONTEXT1_* */
+	int offset = mmGCVM_CONTEXT1_PAGE_TABLE_BASE_ADDR_LO32
+			- mmGCVM_CONTEXT0_PAGE_TABLE_BASE_ADDR_LO32;
 
+	WREG32_SOC15_OFFSET(GC, 0, mmGCVM_CONTEXT0_PAGE_TABLE_BASE_ADDR_LO32,
+				offset * vmid, lower_32_bits(page_table_base));
 
-	WREG32_SOC15(GC, 0, mmGCVM_CONTEXT0_PAGE_TABLE_BASE_ADDR_LO32,
-		     lower_32_bits(value));
-
-	WREG32_SOC15(GC, 0, mmGCVM_CONTEXT0_PAGE_TABLE_BASE_ADDR_HI32,
-		     upper_32_bits(value));
+	WREG32_SOC15_OFFSET(GC, 0, mmGCVM_CONTEXT0_PAGE_TABLE_BASE_ADDR_HI32,
+				offset * vmid, upper_32_bits(page_table_base));
 }
 
 static void gfxhub_v2_0_init_gart_aperture_regs(struct amdgpu_device *adev)
 {
-	gfxhub_v2_0_init_gart_pt_regs(adev);
+	uint64_t pt_base = amdgpu_gmc_pd_addr(adev->gart.bo);
+
+	gfxhub_v2_0_setup_vm_pt_regs(adev, 0, pt_base);
 
 	WREG32_SOC15(GC, 0, mmGCVM_CONTEXT0_PAGE_TABLE_START_ADDR_LO32,
 		     (u32)(adev->gmc.gart_start >> 12));
@@ -151,6 +155,15 @@ static void gfxhub_v2_0_init_cache_regs(struct amdgpu_device *adev)
 	WREG32_SOC15(GC, 0, mmGCVM_L2_CNTL2, tmp);
 
 	tmp = mmGCVM_L2_CNTL3_DEFAULT;
+	if (adev->gmc.translate_further) {
+		tmp = REG_SET_FIELD(tmp, GCVM_L2_CNTL3, BANK_SELECT, 12);
+		tmp = REG_SET_FIELD(tmp, GCVM_L2_CNTL3,
+				    L2_CACHE_BIGK_FRAGMENT_SIZE, 9);
+	} else {
+		tmp = REG_SET_FIELD(tmp, GCVM_L2_CNTL3, BANK_SELECT, 9);
+		tmp = REG_SET_FIELD(tmp, GCVM_L2_CNTL3,
+				    L2_CACHE_BIGK_FRAGMENT_SIZE, 6);
+	}
 	WREG32_SOC15(GC, 0, mmGCVM_L2_CNTL3, tmp);
 
 	tmp = mmGCVM_L2_CNTL4_DEFAULT;
@@ -166,6 +179,8 @@ static void gfxhub_v2_0_enable_system_domain(struct amdgpu_device *adev)
 	tmp = RREG32_SOC15(GC, 0, mmGCVM_CONTEXT0_CNTL);
 	tmp = REG_SET_FIELD(tmp, GCVM_CONTEXT0_CNTL, ENABLE_CONTEXT, 1);
 	tmp = REG_SET_FIELD(tmp, GCVM_CONTEXT0_CNTL, PAGE_TABLE_DEPTH, 0);
+	tmp = REG_SET_FIELD(tmp, GCVM_CONTEXT0_CNTL,
+			    RETRY_PERMISSION_OR_INVALID_PAGE_FAULT, 0);
 	WREG32_SOC15(GC, 0, mmGCVM_CONTEXT0_CNTL, tmp);
 }
 
@@ -341,6 +356,8 @@ void gfxhub_v2_0_init(struct amdgpu_device *adev)
 	hub->ctx0_ptb_addr_hi32 =
 		SOC15_REG_OFFSET(GC, 0,
 				 mmGCVM_CONTEXT0_PAGE_TABLE_BASE_ADDR_HI32);
+	hub->vm_inv_eng0_sem =
+		SOC15_REG_OFFSET(GC, 0, mmGCVM_INVALIDATE_ENG0_SEM);
 	hub->vm_inv_eng0_req =
 		SOC15_REG_OFFSET(GC, 0, mmGCVM_INVALIDATE_ENG0_REQ);
 	hub->vm_inv_eng0_ack =
diff --git a/drivers/gpu/drm/amd/amdgpu/gfxhub_v2_0.h b/drivers/gpu/drm/amd/amdgpu/gfxhub_v2_0.h
index 06807940748b..392b8cd94fc0 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfxhub_v2_0.h
+++ b/drivers/gpu/drm/amd/amdgpu/gfxhub_v2_0.h
@@ -31,5 +31,7 @@ void gfxhub_v2_0_set_fault_enable_default(struct amdgpu_device *adev,
 					  bool value);
 void gfxhub_v2_0_init(struct amdgpu_device *adev);
 u64 gfxhub_v2_0_get_mc_fb_offset(struct amdgpu_device *adev);
+void gfxhub_v2_0_setup_vm_pt_regs(struct amdgpu_device *adev, uint32_t vmid,
+				uint64_t page_table_base);
 
 #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c
index 241a4e57cf4a..321f8a997be8 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c
@@ -235,6 +235,29 @@ static void gmc_v10_0_flush_vm_hub(struct amdgpu_device *adev, uint32_t vmid,
 	const unsigned eng = 17;
 	unsigned int i;
 
+	spin_lock(&adev->gmc.invalidate_lock);
+	/*
+	 * It may lose gpuvm invalidate acknowldege state across power-gating
+	 * off cycle, add semaphore acquire before invalidation and semaphore
+	 * release after invalidation to avoid entering power gated state
+	 * to WA the Issue
+	 */
+
+	/* TODO: It needs to continue working on debugging with semaphore for GFXHUB as well. */
+	if (vmhub == AMDGPU_MMHUB_0 ||
+	    vmhub == AMDGPU_MMHUB_1) {
+		for (i = 0; i < adev->usec_timeout; i++) {
+			/* a read return value of 1 means semaphore acuqire */
+			tmp = RREG32_NO_KIQ(hub->vm_inv_eng0_sem + eng);
+			if (tmp & 0x1)
+				break;
+			udelay(1);
+		}
+
+		if (i >= adev->usec_timeout)
+			DRM_ERROR("Timeout waiting for sem acquire in VM flush!\n");
+	}
+
 	WREG32_NO_KIQ(hub->vm_inv_eng0_req + eng, tmp);
 
 	/*
@@ -254,6 +277,17 @@ static void gmc_v10_0_flush_vm_hub(struct amdgpu_device *adev, uint32_t vmid,
 		udelay(1);
 	}
 
+	/* TODO: It needs to continue working on debugging with semaphore for GFXHUB as well. */
+	if (vmhub == AMDGPU_MMHUB_0 ||
+	    vmhub == AMDGPU_MMHUB_1)
+		/*
+		 * add semaphore release after invalidation,
+		 * write with 0 means semaphore release
+		 */
+		WREG32_NO_KIQ(hub->vm_inv_eng0_sem + eng, 0);
+
+	spin_unlock(&adev->gmc.invalidate_lock);
+
 	if (i < adev->usec_timeout)
 		return;
 
@@ -278,7 +312,7 @@ static void gmc_v10_0_flush_gpu_tlb(struct amdgpu_device *adev, uint32_t vmid,
 	int r;
 
 	/* flush hdp cache */
-	adev->nbio_funcs->hdp_flush(adev, NULL);
+	adev->nbio.funcs->hdp_flush(adev, NULL);
 
 	mutex_lock(&adev->mman.gtt_window_lock);
 
@@ -309,6 +343,7 @@ static void gmc_v10_0_flush_gpu_tlb(struct amdgpu_device *adev, uint32_t vmid,
 
 	job->vm_pd_addr = amdgpu_gmc_pd_addr(adev->gart.bo);
 	job->vm_needs_flush = true;
+	job->ibs->ptr[job->ibs->length_dw++] = ring->funcs->nop;
 	amdgpu_ring_pad_ib(ring, &job->ibs[0]);
 	r = amdgpu_job_submit(job, &adev->mman.entity,
 			      AMDGPU_FENCE_OWNER_UNDEFINED, &fence);
@@ -337,17 +372,38 @@ static uint64_t gmc_v10_0_emit_flush_gpu_tlb(struct amdgpu_ring *ring,
 	uint32_t req = gmc_v10_0_get_invalidate_req(vmid, 0);
 	unsigned eng = ring->vm_inv_eng;
 
+	/*
+	 * It may lose gpuvm invalidate acknowldege state across power-gating
+	 * off cycle, add semaphore acquire before invalidation and semaphore
+	 * release after invalidation to avoid entering power gated state
+	 * to WA the Issue
+	 */
+
+	/* TODO: It needs to continue working on debugging with semaphore for GFXHUB as well. */
+	if (ring->funcs->vmhub == AMDGPU_MMHUB_0 ||
+	    ring->funcs->vmhub == AMDGPU_MMHUB_1)
+		/* a read return value of 1 means semaphore acuqire */
+		amdgpu_ring_emit_reg_wait(ring,
+					  hub->vm_inv_eng0_sem + eng, 0x1, 0x1);
+
 	amdgpu_ring_emit_wreg(ring, hub->ctx0_ptb_addr_lo32 + (2 * vmid),
 			      lower_32_bits(pd_addr));
 
 	amdgpu_ring_emit_wreg(ring, hub->ctx0_ptb_addr_hi32 + (2 * vmid),
 			      upper_32_bits(pd_addr));
 
-	amdgpu_ring_emit_wreg(ring, hub->vm_inv_eng0_req + eng, req);
+	amdgpu_ring_emit_reg_write_reg_wait(ring, hub->vm_inv_eng0_req + eng,
+					    hub->vm_inv_eng0_ack + eng,
+					    req, 1 << vmid);
 
-	/* wait for the invalidate to complete */
-	amdgpu_ring_emit_reg_wait(ring, hub->vm_inv_eng0_ack + eng,
-				  1 << vmid, 1 << vmid);
+	/* TODO: It needs to continue working on debugging with semaphore for GFXHUB as well. */
+	if (ring->funcs->vmhub == AMDGPU_MMHUB_0 ||
+	    ring->funcs->vmhub == AMDGPU_MMHUB_1)
+		/*
+		 * add semaphore release after invalidation,
+		 * write with 0 means semaphore release
+		 */
+		amdgpu_ring_emit_wreg(ring, hub->vm_inv_eng0_sem + eng, 0);
 
 	return pd_addr;
 }
@@ -397,43 +453,23 @@ static void gmc_v10_0_emit_pasid_mapping(struct amdgpu_ring *ring, unsigned vmid
  * 1 system
  * 0 valid
  */
-static uint64_t gmc_v10_0_get_vm_pte_flags(struct amdgpu_device *adev,
-					   uint32_t flags)
-{
-	uint64_t pte_flag = 0;
-
-	if (flags & AMDGPU_VM_PAGE_EXECUTABLE)
-		pte_flag |= AMDGPU_PTE_EXECUTABLE;
-	if (flags & AMDGPU_VM_PAGE_READABLE)
-		pte_flag |= AMDGPU_PTE_READABLE;
-	if (flags & AMDGPU_VM_PAGE_WRITEABLE)
-		pte_flag |= AMDGPU_PTE_WRITEABLE;
 
-	switch (flags & AMDGPU_VM_MTYPE_MASK) {
+static uint64_t gmc_v10_0_map_mtype(struct amdgpu_device *adev, uint32_t flags)
+{
+	switch (flags) {
 	case AMDGPU_VM_MTYPE_DEFAULT:
-		pte_flag |= AMDGPU_PTE_MTYPE_NV10(MTYPE_NC);
-		break;
+		return AMDGPU_PTE_MTYPE_NV10(MTYPE_NC);
 	case AMDGPU_VM_MTYPE_NC:
-		pte_flag |= AMDGPU_PTE_MTYPE_NV10(MTYPE_NC);
-		break;
+		return AMDGPU_PTE_MTYPE_NV10(MTYPE_NC);
 	case AMDGPU_VM_MTYPE_WC:
-		pte_flag |= AMDGPU_PTE_MTYPE_NV10(MTYPE_WC);
-		break;
+		return AMDGPU_PTE_MTYPE_NV10(MTYPE_WC);
 	case AMDGPU_VM_MTYPE_CC:
-		pte_flag |= AMDGPU_PTE_MTYPE_NV10(MTYPE_CC);
-		break;
+		return AMDGPU_PTE_MTYPE_NV10(MTYPE_CC);
 	case AMDGPU_VM_MTYPE_UC:
-		pte_flag |= AMDGPU_PTE_MTYPE_NV10(MTYPE_UC);
-		break;
+		return AMDGPU_PTE_MTYPE_NV10(MTYPE_UC);
 	default:
-		pte_flag |= AMDGPU_PTE_MTYPE_NV10(MTYPE_NC);
-		break;
+		return AMDGPU_PTE_MTYPE_NV10(MTYPE_NC);
 	}
-
-	if (flags & AMDGPU_VM_PAGE_PRT)
-		pte_flag |= AMDGPU_PTE_PRT;
-
-	return pte_flag;
 }
 
 static void gmc_v10_0_get_vm_pde(struct amdgpu_device *adev, int level,
@@ -460,12 +496,32 @@ static void gmc_v10_0_get_vm_pde(struct amdgpu_device *adev, int level,
 	}
 }
 
+static void gmc_v10_0_get_vm_pte(struct amdgpu_device *adev,
+				 struct amdgpu_bo_va_mapping *mapping,
+				 uint64_t *flags)
+{
+	*flags &= ~AMDGPU_PTE_EXECUTABLE;
+	*flags |= mapping->flags & AMDGPU_PTE_EXECUTABLE;
+
+	*flags &= ~AMDGPU_PTE_MTYPE_NV10_MASK;
+	*flags |= (mapping->flags & AMDGPU_PTE_MTYPE_NV10_MASK);
+
+	if (mapping->flags & AMDGPU_PTE_PRT) {
+		*flags |= AMDGPU_PTE_PRT;
+		*flags |= AMDGPU_PTE_SNOOPED;
+		*flags |= AMDGPU_PTE_LOG;
+		*flags |= AMDGPU_PTE_SYSTEM;
+		*flags &= ~AMDGPU_PTE_VALID;
+	}
+}
+
 static const struct amdgpu_gmc_funcs gmc_v10_0_gmc_funcs = {
 	.flush_gpu_tlb = gmc_v10_0_flush_gpu_tlb,
 	.emit_flush_gpu_tlb = gmc_v10_0_emit_flush_gpu_tlb,
 	.emit_pasid_mapping = gmc_v10_0_emit_pasid_mapping,
-	.get_vm_pte_flags = gmc_v10_0_get_vm_pte_flags,
-	.get_vm_pde = gmc_v10_0_get_vm_pde
+	.map_mtype = gmc_v10_0_map_mtype,
+	.get_vm_pde = gmc_v10_0_get_vm_pde,
+	.get_vm_pte = gmc_v10_0_get_vm_pte
 };
 
 static void gmc_v10_0_set_gmc_funcs(struct amdgpu_device *adev)
@@ -519,8 +575,7 @@ static void gmc_v10_0_vram_gtt_location(struct amdgpu_device *adev,
 {
 	u64 base = 0;
 
-	if (!amdgpu_sriov_vf(adev))
-		base = gfxhub_v2_0_get_fb_location(adev);
+	base = gfxhub_v2_0_get_fb_location(adev);
 
 	amdgpu_gmc_vram_location(adev, &adev->gmc, base);
 	amdgpu_gmc_gart_location(adev, mc);
@@ -540,24 +595,13 @@ static void gmc_v10_0_vram_gtt_location(struct amdgpu_device *adev,
  */
 static int gmc_v10_0_mc_init(struct amdgpu_device *adev)
 {
-	int chansize, numchan;
-
-	if (!amdgpu_emu_mode)
-		adev->gmc.vram_width = amdgpu_atomfirmware_get_vram_width(adev);
-	else {
-		/* hard code vram_width for emulation */
-		chansize = 128;
-		numchan = 1;
-		adev->gmc.vram_width = numchan * chansize;
-	}
-
 	/* Could aper size report 0 ? */
 	adev->gmc.aper_base = pci_resource_start(adev->pdev, 0);
 	adev->gmc.aper_size = pci_resource_len(adev->pdev, 0);
 
 	/* size in MB on si */
 	adev->gmc.mc_vram_size =
-		adev->nbio_funcs->get_memsize(adev) * 1024ULL * 1024ULL;
+		adev->nbio.funcs->get_memsize(adev) * 1024ULL * 1024ULL;
 	adev->gmc.real_vram_size = adev->gmc.mc_vram_size;
 	adev->gmc.visible_vram_size = adev->gmc.aper_size;
 
@@ -636,7 +680,7 @@ static unsigned gmc_v10_0_get_vbios_fb_size(struct amdgpu_device *adev)
 
 static int gmc_v10_0_sw_init(void *handle)
 {
-	int r;
+	int r, vram_width = 0, vram_type = 0, vram_vendor = 0;
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
 	gfxhub_v2_0_init(adev);
@@ -644,7 +688,15 @@ static int gmc_v10_0_sw_init(void *handle)
 
 	spin_lock_init(&adev->gmc.invalidate_lock);
 
-	adev->gmc.vram_type = amdgpu_atomfirmware_get_vram_type(adev);
+	r = amdgpu_atomfirmware_get_vram_info(adev,
+		&vram_width, &vram_type, &vram_vendor);
+	if (!amdgpu_emu_mode)
+		adev->gmc.vram_width = vram_width;
+	else
+		adev->gmc.vram_width = 1 * 128; /* numchan * chansize */
+
+	adev->gmc.vram_type = vram_type;
+	adev->gmc.vram_vendor = vram_vendor;
 	switch (adev->asic_type) {
 	case CHIP_NAVI10:
 	case CHIP_NAVI14:
@@ -794,7 +846,7 @@ static int gmc_v10_0_gart_enable(struct amdgpu_device *adev)
 	WREG32_SOC15(HDP, 0, mmHDP_HOST_PATH_CNTL, tmp);
 
 	/* Flush HDP after it is initialized */
-	adev->nbio_funcs->hdp_flush(adev, NULL);
+	adev->nbio.funcs->hdp_flush(adev, NULL);
 
 	value = (amdgpu_vm_fault_stop == AMDGPU_VM_FAULT_STOP_ALWAYS) ?
 		false : true;
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c
index 9fb1765e92d1..b205039350b6 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c
@@ -386,27 +386,20 @@ static uint64_t gmc_v6_0_emit_flush_gpu_tlb(struct amdgpu_ring *ring,
 	return pd_addr;
 }
 
-static uint64_t gmc_v6_0_get_vm_pte_flags(struct amdgpu_device *adev,
-					  uint32_t flags)
-{
-	uint64_t pte_flag = 0;
-
-	if (flags & AMDGPU_VM_PAGE_READABLE)
-		pte_flag |= AMDGPU_PTE_READABLE;
-	if (flags & AMDGPU_VM_PAGE_WRITEABLE)
-		pte_flag |= AMDGPU_PTE_WRITEABLE;
-	if (flags & AMDGPU_VM_PAGE_PRT)
-		pte_flag |= AMDGPU_PTE_PRT;
-
-	return pte_flag;
-}
-
 static void gmc_v6_0_get_vm_pde(struct amdgpu_device *adev, int level,
 				uint64_t *addr, uint64_t *flags)
 {
 	BUG_ON(*addr & 0xFFFFFF0000000FFFULL);
 }
 
+static void gmc_v6_0_get_vm_pte(struct amdgpu_device *adev,
+				struct amdgpu_bo_va_mapping *mapping,
+				uint64_t *flags)
+{
+	*flags &= ~AMDGPU_PTE_EXECUTABLE;
+	*flags &= ~AMDGPU_PTE_PRT;
+}
+
 static void gmc_v6_0_set_fault_enable_default(struct amdgpu_device *adev,
 					      bool value)
 {
@@ -1153,7 +1146,7 @@ static const struct amdgpu_gmc_funcs gmc_v6_0_gmc_funcs = {
 	.emit_flush_gpu_tlb = gmc_v6_0_emit_flush_gpu_tlb,
 	.set_prt = gmc_v6_0_set_prt,
 	.get_vm_pde = gmc_v6_0_get_vm_pde,
-	.get_vm_pte_flags = gmc_v6_0_get_vm_pte_flags
+	.get_vm_pte = gmc_v6_0_get_vm_pte,
 };
 
 static const struct amdgpu_irq_src_funcs gmc_v6_0_irq_funcs = {
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c
index 0c3d9bc3a641..f08e5330642d 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c
@@ -463,27 +463,20 @@ static void gmc_v7_0_emit_pasid_mapping(struct amdgpu_ring *ring, unsigned vmid,
 	amdgpu_ring_emit_wreg(ring, mmIH_VMID_0_LUT + vmid, pasid);
 }
 
-static uint64_t gmc_v7_0_get_vm_pte_flags(struct amdgpu_device *adev,
-					  uint32_t flags)
-{
-	uint64_t pte_flag = 0;
-
-	if (flags & AMDGPU_VM_PAGE_READABLE)
-		pte_flag |= AMDGPU_PTE_READABLE;
-	if (flags & AMDGPU_VM_PAGE_WRITEABLE)
-		pte_flag |= AMDGPU_PTE_WRITEABLE;
-	if (flags & AMDGPU_VM_PAGE_PRT)
-		pte_flag |= AMDGPU_PTE_PRT;
-
-	return pte_flag;
-}
-
 static void gmc_v7_0_get_vm_pde(struct amdgpu_device *adev, int level,
 				uint64_t *addr, uint64_t *flags)
 {
 	BUG_ON(*addr & 0xFFFFFF0000000FFFULL);
 }
 
+static void gmc_v7_0_get_vm_pte(struct amdgpu_device *adev,
+				struct amdgpu_bo_va_mapping *mapping,
+				uint64_t *flags)
+{
+	*flags &= ~AMDGPU_PTE_EXECUTABLE;
+	*flags &= ~AMDGPU_PTE_PRT;
+}
+
 /**
  * gmc_v8_0_set_fault_enable_default - update VM fault handling
  *
@@ -1343,8 +1336,8 @@ static const struct amdgpu_gmc_funcs gmc_v7_0_gmc_funcs = {
 	.emit_flush_gpu_tlb = gmc_v7_0_emit_flush_gpu_tlb,
 	.emit_pasid_mapping = gmc_v7_0_emit_pasid_mapping,
 	.set_prt = gmc_v7_0_set_prt,
-	.get_vm_pte_flags = gmc_v7_0_get_vm_pte_flags,
-	.get_vm_pde = gmc_v7_0_get_vm_pde
+	.get_vm_pde = gmc_v7_0_get_vm_pde,
+	.get_vm_pte = gmc_v7_0_get_vm_pte
 };
 
 static const struct amdgpu_irq_src_funcs gmc_v7_0_irq_funcs = {
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
index ea764dd9245d..6d96d40fbcb8 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
@@ -686,29 +686,21 @@ static void gmc_v8_0_emit_pasid_mapping(struct amdgpu_ring *ring, unsigned vmid,
  * 0 valid
  */
 
-static uint64_t gmc_v8_0_get_vm_pte_flags(struct amdgpu_device *adev,
-					  uint32_t flags)
-{
-	uint64_t pte_flag = 0;
-
-	if (flags & AMDGPU_VM_PAGE_EXECUTABLE)
-		pte_flag |= AMDGPU_PTE_EXECUTABLE;
-	if (flags & AMDGPU_VM_PAGE_READABLE)
-		pte_flag |= AMDGPU_PTE_READABLE;
-	if (flags & AMDGPU_VM_PAGE_WRITEABLE)
-		pte_flag |= AMDGPU_PTE_WRITEABLE;
-	if (flags & AMDGPU_VM_PAGE_PRT)
-		pte_flag |= AMDGPU_PTE_PRT;
-
-	return pte_flag;
-}
-
 static void gmc_v8_0_get_vm_pde(struct amdgpu_device *adev, int level,
 				uint64_t *addr, uint64_t *flags)
 {
 	BUG_ON(*addr & 0xFFFFFF0000000FFFULL);
 }
 
+static void gmc_v8_0_get_vm_pte(struct amdgpu_device *adev,
+				struct amdgpu_bo_va_mapping *mapping,
+				uint64_t *flags)
+{
+	*flags &= ~AMDGPU_PTE_EXECUTABLE;
+	*flags |= mapping->flags & AMDGPU_PTE_EXECUTABLE;
+	*flags &= ~AMDGPU_PTE_PRT;
+}
+
 /**
  * gmc_v8_0_set_fault_enable_default - update VM fault handling
  *
@@ -1711,8 +1703,8 @@ static const struct amdgpu_gmc_funcs gmc_v8_0_gmc_funcs = {
 	.emit_flush_gpu_tlb = gmc_v8_0_emit_flush_gpu_tlb,
 	.emit_pasid_mapping = gmc_v8_0_emit_pasid_mapping,
 	.set_prt = gmc_v8_0_set_prt,
-	.get_vm_pte_flags = gmc_v8_0_get_vm_pte_flags,
-	.get_vm_pde = gmc_v8_0_get_vm_pde
+	.get_vm_pde = gmc_v8_0_get_vm_pde,
+	.get_vm_pte = gmc_v8_0_get_vm_pte
 };
 
 static const struct amdgpu_irq_src_funcs gmc_v8_0_irq_funcs = {
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
index f91337030dc0..3c355fb5d2b4 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
@@ -51,10 +51,12 @@
 #include "gfxhub_v1_1.h"
 #include "mmhub_v9_4.h"
 #include "umc_v6_1.h"
+#include "umc_v6_0.h"
 
 #include "ivsrcid/vmc/irqsrcs_vmc_1_0.h"
 
 #include "amdgpu_ras.h"
+#include "amdgpu_xgmi.h"
 
 /* add these here since we already include dce12 headers and these are for DCN */
 #define mmHUBP0_DCSURF_PRI_VIEWPORT_DIMENSION                                                          0x055d
@@ -243,44 +245,6 @@ static int gmc_v9_0_ecc_interrupt_state(struct amdgpu_device *adev,
 	return 0;
 }
 
-static int gmc_v9_0_process_ras_data_cb(struct amdgpu_device *adev,
-		struct ras_err_data *err_data,
-		struct amdgpu_iv_entry *entry)
-{
-	kgd2kfd_set_sram_ecc_flag(adev->kfd.dev);
-	if (adev->umc.funcs->query_ras_error_count)
-		adev->umc.funcs->query_ras_error_count(adev, err_data);
-	/* umc query_ras_error_address is also responsible for clearing
-	 * error status
-	 */
-	if (adev->umc.funcs->query_ras_error_address)
-		adev->umc.funcs->query_ras_error_address(adev, err_data);
-
-	/* only uncorrectable error needs gpu reset */
-	if (err_data->ue_count)
-		amdgpu_ras_reset_gpu(adev, 0);
-
-	return AMDGPU_RAS_SUCCESS;
-}
-
-static int gmc_v9_0_process_ecc_irq(struct amdgpu_device *adev,
-		struct amdgpu_irq_src *source,
-		struct amdgpu_iv_entry *entry)
-{
-	struct ras_common_if *ras_if = adev->gmc.umc_ras_if;
-	struct ras_dispatch_if ih_data = {
-		.entry = entry,
-	};
-
-	if (!ras_if)
-		return 0;
-
-	ih_data.head = *ras_if;
-
-	amdgpu_ras_interrupt_dispatch(adev, &ih_data);
-	return 0;
-}
-
 static int gmc_v9_0_vm_fault_interrupt_state(struct amdgpu_device *adev,
 					struct amdgpu_irq_src *src,
 					unsigned type,
@@ -355,6 +319,10 @@ static int gmc_v9_0_process_interrupt(struct amdgpu_device *adev,
 	}
 
 	/* If it's the first fault for this address, process it normally */
+	if (retry_fault && !in_interrupt() &&
+	    amdgpu_vm_handle_fault(adev, entry->pasid, addr))
+		return 1; /* This also prevents sending it to KFD */
+
 	if (!amdgpu_sriov_vf(adev)) {
 		/*
 		 * Issue a dummy read to wait for the status register to
@@ -417,7 +385,7 @@ static const struct amdgpu_irq_src_funcs gmc_v9_0_irq_funcs = {
 
 static const struct amdgpu_irq_src_funcs gmc_v9_0_ecc_funcs = {
 	.set = gmc_v9_0_ecc_interrupt_state,
-	.process = gmc_v9_0_process_ecc_irq,
+	.process = amdgpu_umc_process_ecc_irq,
 };
 
 static void gmc_v9_0_set_irq_funcs(struct amdgpu_device *adev)
@@ -491,6 +459,29 @@ static void gmc_v9_0_flush_gpu_tlb(struct amdgpu_device *adev, uint32_t vmid,
 	}
 
 	spin_lock(&adev->gmc.invalidate_lock);
+
+	/*
+	 * It may lose gpuvm invalidate acknowldege state across power-gating
+	 * off cycle, add semaphore acquire before invalidation and semaphore
+	 * release after invalidation to avoid entering power gated state
+	 * to WA the Issue
+	 */
+
+	/* TODO: It needs to continue working on debugging with semaphore for GFXHUB as well. */
+	if (vmhub == AMDGPU_MMHUB_0 ||
+	    vmhub == AMDGPU_MMHUB_1) {
+		for (j = 0; j < adev->usec_timeout; j++) {
+			/* a read return value of 1 means semaphore acuqire */
+			tmp = RREG32_NO_KIQ(hub->vm_inv_eng0_sem + eng);
+			if (tmp & 0x1)
+				break;
+			udelay(1);
+		}
+
+		if (j >= adev->usec_timeout)
+			DRM_ERROR("Timeout waiting for sem acquire in VM flush!\n");
+	}
+
 	WREG32_NO_KIQ(hub->vm_inv_eng0_req + eng, tmp);
 
 	/*
@@ -506,7 +497,18 @@ static void gmc_v9_0_flush_gpu_tlb(struct amdgpu_device *adev, uint32_t vmid,
 			break;
 		udelay(1);
 	}
+
+	/* TODO: It needs to continue working on debugging with semaphore for GFXHUB as well. */
+	if (vmhub == AMDGPU_MMHUB_0 ||
+	    vmhub == AMDGPU_MMHUB_1)
+		/*
+		 * add semaphore release after invalidation,
+		 * write with 0 means semaphore release
+		 */
+		WREG32_NO_KIQ(hub->vm_inv_eng0_sem + eng, 0);
+
 	spin_unlock(&adev->gmc.invalidate_lock);
+
 	if (j < adev->usec_timeout)
 		return;
 
@@ -521,6 +523,20 @@ static uint64_t gmc_v9_0_emit_flush_gpu_tlb(struct amdgpu_ring *ring,
 	uint32_t req = gmc_v9_0_get_invalidate_req(vmid, 0);
 	unsigned eng = ring->vm_inv_eng;
 
+	/*
+	 * It may lose gpuvm invalidate acknowldege state across power-gating
+	 * off cycle, add semaphore acquire before invalidation and semaphore
+	 * release after invalidation to avoid entering power gated state
+	 * to WA the Issue
+	 */
+
+	/* TODO: It needs to continue working on debugging with semaphore for GFXHUB as well. */
+	if (ring->funcs->vmhub == AMDGPU_MMHUB_0 ||
+	    ring->funcs->vmhub == AMDGPU_MMHUB_1)
+		/* a read return value of 1 means semaphore acuqire */
+		amdgpu_ring_emit_reg_wait(ring,
+					  hub->vm_inv_eng0_sem + eng, 0x1, 0x1);
+
 	amdgpu_ring_emit_wreg(ring, hub->ctx0_ptb_addr_lo32 + (2 * vmid),
 			      lower_32_bits(pd_addr));
 
@@ -531,6 +547,15 @@ static uint64_t gmc_v9_0_emit_flush_gpu_tlb(struct amdgpu_ring *ring,
 					    hub->vm_inv_eng0_ack + eng,
 					    req, 1 << vmid);
 
+	/* TODO: It needs to continue working on debugging with semaphore for GFXHUB as well. */
+	if (ring->funcs->vmhub == AMDGPU_MMHUB_0 ||
+	    ring->funcs->vmhub == AMDGPU_MMHUB_1)
+		/*
+		 * add semaphore release after invalidation,
+		 * write with 0 means semaphore release
+		 */
+		amdgpu_ring_emit_wreg(ring, hub->vm_inv_eng0_sem + eng, 0);
+
 	return pd_addr;
 }
 
@@ -584,44 +609,25 @@ static void gmc_v9_0_emit_pasid_mapping(struct amdgpu_ring *ring, unsigned vmid,
  * 0 valid
  */
 
-static uint64_t gmc_v9_0_get_vm_pte_flags(struct amdgpu_device *adev,
-						uint32_t flags)
+static uint64_t gmc_v9_0_map_mtype(struct amdgpu_device *adev, uint32_t flags)
 
 {
-	uint64_t pte_flag = 0;
-
-	if (flags & AMDGPU_VM_PAGE_EXECUTABLE)
-		pte_flag |= AMDGPU_PTE_EXECUTABLE;
-	if (flags & AMDGPU_VM_PAGE_READABLE)
-		pte_flag |= AMDGPU_PTE_READABLE;
-	if (flags & AMDGPU_VM_PAGE_WRITEABLE)
-		pte_flag |= AMDGPU_PTE_WRITEABLE;
-
-	switch (flags & AMDGPU_VM_MTYPE_MASK) {
+	switch (flags) {
 	case AMDGPU_VM_MTYPE_DEFAULT:
-		pte_flag |= AMDGPU_PTE_MTYPE_VG10(MTYPE_NC);
-		break;
+		return AMDGPU_PTE_MTYPE_VG10(MTYPE_NC);
 	case AMDGPU_VM_MTYPE_NC:
-		pte_flag |= AMDGPU_PTE_MTYPE_VG10(MTYPE_NC);
-		break;
+		return AMDGPU_PTE_MTYPE_VG10(MTYPE_NC);
 	case AMDGPU_VM_MTYPE_WC:
-		pte_flag |= AMDGPU_PTE_MTYPE_VG10(MTYPE_WC);
-		break;
+		return AMDGPU_PTE_MTYPE_VG10(MTYPE_WC);
+	case AMDGPU_VM_MTYPE_RW:
+		return AMDGPU_PTE_MTYPE_VG10(MTYPE_RW);
 	case AMDGPU_VM_MTYPE_CC:
-		pte_flag |= AMDGPU_PTE_MTYPE_VG10(MTYPE_CC);
-		break;
+		return AMDGPU_PTE_MTYPE_VG10(MTYPE_CC);
 	case AMDGPU_VM_MTYPE_UC:
-		pte_flag |= AMDGPU_PTE_MTYPE_VG10(MTYPE_UC);
-		break;
+		return AMDGPU_PTE_MTYPE_VG10(MTYPE_UC);
 	default:
-		pte_flag |= AMDGPU_PTE_MTYPE_VG10(MTYPE_NC);
-		break;
+		return AMDGPU_PTE_MTYPE_VG10(MTYPE_NC);
 	}
-
-	if (flags & AMDGPU_VM_PAGE_PRT)
-		pte_flag |= AMDGPU_PTE_PRT;
-
-	return pte_flag;
 }
 
 static void gmc_v9_0_get_vm_pde(struct amdgpu_device *adev, int level,
@@ -648,12 +654,34 @@ static void gmc_v9_0_get_vm_pde(struct amdgpu_device *adev, int level,
 	}
 }
 
+static void gmc_v9_0_get_vm_pte(struct amdgpu_device *adev,
+				struct amdgpu_bo_va_mapping *mapping,
+				uint64_t *flags)
+{
+	*flags &= ~AMDGPU_PTE_EXECUTABLE;
+	*flags |= mapping->flags & AMDGPU_PTE_EXECUTABLE;
+
+	*flags &= ~AMDGPU_PTE_MTYPE_VG10_MASK;
+	*flags |= mapping->flags & AMDGPU_PTE_MTYPE_VG10_MASK;
+
+	if (mapping->flags & AMDGPU_PTE_PRT) {
+		*flags |= AMDGPU_PTE_PRT;
+		*flags &= ~AMDGPU_PTE_VALID;
+	}
+
+	if (adev->asic_type == CHIP_ARCTURUS &&
+	    !(*flags & AMDGPU_PTE_SYSTEM) &&
+	    mapping->bo_va->is_xgmi)
+		*flags |= AMDGPU_PTE_SNOOPED;
+}
+
 static const struct amdgpu_gmc_funcs gmc_v9_0_gmc_funcs = {
 	.flush_gpu_tlb = gmc_v9_0_flush_gpu_tlb,
 	.emit_flush_gpu_tlb = gmc_v9_0_emit_flush_gpu_tlb,
 	.emit_pasid_mapping = gmc_v9_0_emit_pasid_mapping,
-	.get_vm_pte_flags = gmc_v9_0_get_vm_pte_flags,
-	.get_vm_pde = gmc_v9_0_get_vm_pde
+	.map_mtype = gmc_v9_0_map_mtype,
+	.get_vm_pde = gmc_v9_0_get_vm_pde,
+	.get_vm_pte = gmc_v9_0_get_vm_pte
 };
 
 static void gmc_v9_0_set_gmc_funcs(struct amdgpu_device *adev)
@@ -664,6 +692,9 @@ static void gmc_v9_0_set_gmc_funcs(struct amdgpu_device *adev)
 static void gmc_v9_0_set_umc_funcs(struct amdgpu_device *adev)
 {
 	switch (adev->asic_type) {
+	case CHIP_VEGA10:
+		adev->umc.funcs = &umc_v6_0_funcs;
+		break;
 	case CHIP_VEGA20:
 		adev->umc.max_ras_err_cnt_per_query = UMC_V6_1_TOTAL_CHANNEL_NUM;
 		adev->umc.channel_inst_num = UMC_V6_1_CHANNEL_INSTANCE_NUM;
@@ -681,7 +712,7 @@ static void gmc_v9_0_set_mmhub_funcs(struct amdgpu_device *adev)
 {
 	switch (adev->asic_type) {
 	case CHIP_VEGA20:
-		adev->mmhub_funcs = &mmhub_v1_0_funcs;
+		adev->mmhub.funcs = &mmhub_v1_0_funcs;
 		break;
 	default:
 		break;
@@ -762,140 +793,10 @@ static int gmc_v9_0_allocate_vm_inv_eng(struct amdgpu_device *adev)
 	return 0;
 }
 
-static int gmc_v9_0_ecc_ras_block_late_init(void *handle,
-			struct ras_fs_if *fs_info, struct ras_common_if *ras_block)
-{
-	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-	struct ras_common_if **ras_if = NULL;
-	struct ras_ih_if ih_info = {
-		.cb = gmc_v9_0_process_ras_data_cb,
-	};
-	int r;
-
-	if (ras_block->block == AMDGPU_RAS_BLOCK__UMC)
-		ras_if = &adev->gmc.umc_ras_if;
-	else if (ras_block->block == AMDGPU_RAS_BLOCK__MMHUB)
-		ras_if = &adev->gmc.mmhub_ras_if;
-	else
-		BUG();
-
-	if (!amdgpu_ras_is_supported(adev, ras_block->block)) {
-		amdgpu_ras_feature_enable_on_boot(adev, ras_block, 0);
-		return 0;
-	}
-
-	/* handle resume path. */
-	if (*ras_if) {
-		/* resend ras TA enable cmd during resume.
-		 * prepare to handle failure.
-		 */
-		ih_info.head = **ras_if;
-		r = amdgpu_ras_feature_enable_on_boot(adev, *ras_if, 1);
-		if (r) {
-			if (r == -EAGAIN) {
-				/* request a gpu reset. will run again. */
-				amdgpu_ras_request_reset_on_boot(adev,
-						ras_block->block);
-				return 0;
-			}
-			/* fail to enable ras, cleanup all. */
-			goto irq;
-		}
-		/* enable successfully. continue. */
-		goto resume;
-	}
-
-	*ras_if = kmalloc(sizeof(**ras_if), GFP_KERNEL);
-	if (!*ras_if)
-		return -ENOMEM;
-
-	**ras_if = *ras_block;
-
-	r = amdgpu_ras_feature_enable_on_boot(adev, *ras_if, 1);
-	if (r) {
-		if (r == -EAGAIN) {
-			amdgpu_ras_request_reset_on_boot(adev,
-					ras_block->block);
-			r = 0;
-		}
-		goto feature;
-	}
-
-	ih_info.head = **ras_if;
-	fs_info->head = **ras_if;
-
-	if (ras_block->block == AMDGPU_RAS_BLOCK__UMC) {
-		r = amdgpu_ras_interrupt_add_handler(adev, &ih_info);
-		if (r)
-			goto interrupt;
-	}
-
-	amdgpu_ras_debugfs_create(adev, fs_info);
-
-	r = amdgpu_ras_sysfs_create(adev, fs_info);
-	if (r)
-		goto sysfs;
-resume:
-	if (ras_block->block == AMDGPU_RAS_BLOCK__UMC) {
-		r = amdgpu_irq_get(adev, &adev->gmc.ecc_irq, 0);
-		if (r)
-			goto irq;
-	}
-
-	return 0;
-irq:
-	amdgpu_ras_sysfs_remove(adev, *ras_if);
-sysfs:
-	amdgpu_ras_debugfs_remove(adev, *ras_if);
-	if (ras_block->block == AMDGPU_RAS_BLOCK__UMC)
-		amdgpu_ras_interrupt_remove_handler(adev, &ih_info);
-interrupt:
-	amdgpu_ras_feature_enable(adev, *ras_if, 0);
-feature:
-	kfree(*ras_if);
-	*ras_if = NULL;
-	return r;
-}
-
-static int gmc_v9_0_ecc_late_init(void *handle)
-{
-	int r;
-
-	struct ras_fs_if umc_fs_info = {
-		.sysfs_name = "umc_err_count",
-		.debugfs_name = "umc_err_inject",
-	};
-	struct ras_common_if umc_ras_block = {
-		.block = AMDGPU_RAS_BLOCK__UMC,
-		.type = AMDGPU_RAS_ERROR__MULTI_UNCORRECTABLE,
-		.sub_block_index = 0,
-		.name = "umc",
-	};
-	struct ras_fs_if mmhub_fs_info = {
-		.sysfs_name = "mmhub_err_count",
-		.debugfs_name = "mmhub_err_inject",
-	};
-	struct ras_common_if mmhub_ras_block = {
-		.block = AMDGPU_RAS_BLOCK__MMHUB,
-		.type = AMDGPU_RAS_ERROR__MULTI_UNCORRECTABLE,
-		.sub_block_index = 0,
-		.name = "mmhub",
-	};
-
-	r = gmc_v9_0_ecc_ras_block_late_init(handle,
-			&umc_fs_info, &umc_ras_block);
-	if (r)
-		return r;
-
-	r = gmc_v9_0_ecc_ras_block_late_init(handle,
-			&mmhub_fs_info, &mmhub_ras_block);
-	return r;
-}
-
 static int gmc_v9_0_late_init(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-	bool r;
+	int r;
 
 	if (!gmc_v9_0_keep_stolen_memory(adev))
 		amdgpu_bo_late_init(adev);
@@ -929,7 +830,7 @@ static int gmc_v9_0_late_init(void *handle)
 		}
 	}
 
-	r = gmc_v9_0_ecc_late_init(handle);
+	r = amdgpu_gmc_ras_late_init(adev);
 	if (r)
 		return r;
 
@@ -970,33 +871,11 @@ static void gmc_v9_0_vram_gtt_location(struct amdgpu_device *adev,
  */
 static int gmc_v9_0_mc_init(struct amdgpu_device *adev)
 {
-	int chansize, numchan;
 	int r;
 
-	if (amdgpu_sriov_vf(adev)) {
-		/* For Vega10 SR-IOV, vram_width can't be read from ATOM as RAVEN,
-		 * and DF related registers is not readable, seems hardcord is the
-		 * only way to set the correct vram_width
-		 */
-		adev->gmc.vram_width = 2048;
-	} else if (amdgpu_emu_mode != 1) {
-		adev->gmc.vram_width = amdgpu_atomfirmware_get_vram_width(adev);
-	}
-
-	if (!adev->gmc.vram_width) {
-		/* hbm memory channel size */
-		if (adev->flags & AMD_IS_APU)
-			chansize = 64;
-		else
-			chansize = 128;
-
-		numchan = adev->df_funcs->get_hbm_channel_number(adev);
-		adev->gmc.vram_width = numchan * chansize;
-	}
-
 	/* size in MB on si */
 	adev->gmc.mc_vram_size =
-		adev->nbio_funcs->get_memsize(adev) * 1024ULL * 1024ULL;
+		adev->nbio.funcs->get_memsize(adev) * 1024ULL * 1024ULL;
 	adev->gmc.real_vram_size = adev->gmc.mc_vram_size;
 
 	if (!(adev->flags & AMD_IS_APU)) {
@@ -1108,7 +987,7 @@ static unsigned gmc_v9_0_get_vbios_fb_size(struct amdgpu_device *adev)
 
 static int gmc_v9_0_sw_init(void *handle)
 {
-	int r;
+	int r, vram_width = 0, vram_type = 0, vram_vendor = 0;
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
 	gfxhub_v1_0_init(adev);
@@ -1119,7 +998,32 @@ static int gmc_v9_0_sw_init(void *handle)
 
 	spin_lock_init(&adev->gmc.invalidate_lock);
 
-	adev->gmc.vram_type = amdgpu_atomfirmware_get_vram_type(adev);
+	r = amdgpu_atomfirmware_get_vram_info(adev,
+		&vram_width, &vram_type, &vram_vendor);
+	if (amdgpu_sriov_vf(adev))
+		/* For Vega10 SR-IOV, vram_width can't be read from ATOM as RAVEN,
+		 * and DF related registers is not readable, seems hardcord is the
+		 * only way to set the correct vram_width
+		 */
+		adev->gmc.vram_width = 2048;
+	else if (amdgpu_emu_mode != 1)
+		adev->gmc.vram_width = vram_width;
+
+	if (!adev->gmc.vram_width) {
+		int chansize, numchan;
+
+		/* hbm memory channel size */
+		if (adev->flags & AMD_IS_APU)
+			chansize = 64;
+		else
+			chansize = 128;
+
+		numchan = adev->df_funcs->get_hbm_channel_number(adev);
+		adev->gmc.vram_width = numchan * chansize;
+	}
+
+	adev->gmc.vram_type = vram_type;
+	adev->gmc.vram_vendor = vram_vendor;
 	switch (adev->asic_type) {
 	case CHIP_RAVEN:
 		adev->num_vmhubs = 2;
@@ -1240,33 +1144,7 @@ static int gmc_v9_0_sw_fini(void *handle)
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 	void *stolen_vga_buf;
 
-	if (amdgpu_ras_is_supported(adev, AMDGPU_RAS_BLOCK__UMC) &&
-			adev->gmc.umc_ras_if) {
-		struct ras_common_if *ras_if = adev->gmc.umc_ras_if;
-		struct ras_ih_if ih_info = {
-			.head = *ras_if,
-		};
-
-		/* remove fs first */
-		amdgpu_ras_debugfs_remove(adev, ras_if);
-		amdgpu_ras_sysfs_remove(adev, ras_if);
-		/* remove the IH */
-		amdgpu_ras_interrupt_remove_handler(adev, &ih_info);
-		amdgpu_ras_feature_enable(adev, ras_if, 0);
-		kfree(ras_if);
-	}
-
-	if (amdgpu_ras_is_supported(adev, AMDGPU_RAS_BLOCK__MMHUB) &&
-			adev->gmc.mmhub_ras_if) {
-		struct ras_common_if *ras_if = adev->gmc.mmhub_ras_if;
-
-		/* remove fs and disable ras feature */
-		amdgpu_ras_debugfs_remove(adev, ras_if);
-		amdgpu_ras_sysfs_remove(adev, ras_if);
-		amdgpu_ras_feature_enable(adev, ras_if, 0);
-		kfree(ras_if);
-	}
-
+	amdgpu_gmc_ras_fini(adev);
 	amdgpu_gem_force_release(adev);
 	amdgpu_vm_manager_fini(adev);
 
@@ -1316,13 +1194,7 @@ static void gmc_v9_0_init_golden_registers(struct amdgpu_device *adev)
  */
 static int gmc_v9_0_gart_enable(struct amdgpu_device *adev)
 {
-	int r, i;
-	bool value;
-	u32 tmp;
-
-	amdgpu_device_program_register_sequence(adev,
-						golden_settings_vega10_hdp,
-						ARRAY_SIZE(golden_settings_vega10_hdp));
+	int r;
 
 	if (adev->gart.bo == NULL) {
 		dev_err(adev->dev, "No VRAM object for PCIE GART.\n");
@@ -1332,15 +1204,6 @@ static int gmc_v9_0_gart_enable(struct amdgpu_device *adev)
 	if (r)
 		return r;
 
-	switch (adev->asic_type) {
-	case CHIP_RAVEN:
-		/* TODO for renoir */
-		mmhub_v1_0_update_power_gating(adev, true);
-		break;
-	default:
-		break;
-	}
-
 	r = gfxhub_v1_0_gart_enable(adev);
 	if (r)
 		return r;
@@ -1352,6 +1215,49 @@ static int gmc_v9_0_gart_enable(struct amdgpu_device *adev)
 	if (r)
 		return r;
 
+	DRM_INFO("PCIE GART of %uM enabled (table at 0x%016llX).\n",
+		 (unsigned)(adev->gmc.gart_size >> 20),
+		 (unsigned long long)amdgpu_bo_gpu_offset(adev->gart.bo));
+	adev->gart.ready = true;
+	return 0;
+}
+
+static int gmc_v9_0_hw_init(void *handle)
+{
+	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+	bool value;
+	int r, i;
+	u32 tmp;
+
+	/* The sequence of these two function calls matters.*/
+	gmc_v9_0_init_golden_registers(adev);
+
+	if (adev->mode_info.num_crtc) {
+		if (adev->asic_type != CHIP_ARCTURUS) {
+			/* Lockout access through VGA aperture*/
+			WREG32_FIELD15(DCE, 0, VGA_HDP_CONTROL, VGA_MEMORY_DISABLE, 1);
+
+			/* disable VGA render */
+			WREG32_FIELD15(DCE, 0, VGA_RENDER_CONTROL, VGA_VSTATUS_CNTL, 0);
+		}
+	}
+
+	amdgpu_device_program_register_sequence(adev,
+						golden_settings_vega10_hdp,
+						ARRAY_SIZE(golden_settings_vega10_hdp));
+
+	switch (adev->asic_type) {
+	case CHIP_RAVEN:
+		/* TODO for renoir */
+		mmhub_v1_0_update_power_gating(adev, true);
+		break;
+	case CHIP_ARCTURUS:
+		WREG32_FIELD15(HDP, 0, HDP_MMHUB_CNTL, HDP_MMHUB_GCC, 1);
+		break;
+	default:
+		break;
+	}
+
 	WREG32_FIELD15(HDP, 0, HDP_MISC_CNTL, FLUSH_INVALIDATE_CACHE, 1);
 
 	tmp = RREG32_SOC15(HDP, 0, mmHDP_HOST_PATH_CNTL);
@@ -1361,7 +1267,7 @@ static int gmc_v9_0_gart_enable(struct amdgpu_device *adev)
 	WREG32_SOC15(HDP, 0, mmHDP_NONSURFACE_BASE_HI, (adev->gmc.vram_start >> 40));
 
 	/* After HDP is initialized, flush HDP.*/
-	adev->nbio_funcs->hdp_flush(adev, NULL);
+	adev->nbio.funcs->hdp_flush(adev, NULL);
 
 	if (amdgpu_vm_fault_stop == AMDGPU_VM_FAULT_STOP_ALWAYS)
 		value = false;
@@ -1377,28 +1283,8 @@ static int gmc_v9_0_gart_enable(struct amdgpu_device *adev)
 	for (i = 0; i < adev->num_vmhubs; ++i)
 		gmc_v9_0_flush_gpu_tlb(adev, 0, i, 0);
 
-	DRM_INFO("PCIE GART of %uM enabled (table at 0x%016llX).\n",
-		 (unsigned)(adev->gmc.gart_size >> 20),
-		 (unsigned long long)amdgpu_bo_gpu_offset(adev->gart.bo));
-	adev->gart.ready = true;
-	return 0;
-}
-
-static int gmc_v9_0_hw_init(void *handle)
-{
-	int r;
-	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-
-	/* The sequence of these two function calls matters.*/
-	gmc_v9_0_init_golden_registers(adev);
-
-	if (adev->mode_info.num_crtc) {
-		/* Lockout access through VGA aperture*/
-		WREG32_FIELD15(DCE, 0, VGA_HDP_CONTROL, VGA_MEMORY_DISABLE, 1);
-
-		/* disable VGA render */
-		WREG32_FIELD15(DCE, 0, VGA_RENDER_CONTROL, VGA_VSTATUS_CNTL, 0);
-	}
+	if (adev->umc.funcs && adev->umc.funcs->init_registers)
+		adev->umc.funcs->init_registers(adev);
 
 	r = gmc_v9_0_gart_enable(adev);
 
diff --git a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c b/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c
index 04cd4b6f95d4..28105e4af507 100644
--- a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c
@@ -206,6 +206,8 @@ static void mmhub_v1_0_enable_system_domain(struct amdgpu_device *adev)
 	tmp = RREG32_SOC15(MMHUB, 0, mmVM_CONTEXT0_CNTL);
 	tmp = REG_SET_FIELD(tmp, VM_CONTEXT0_CNTL, ENABLE_CONTEXT, 1);
 	tmp = REG_SET_FIELD(tmp, VM_CONTEXT0_CNTL, PAGE_TABLE_DEPTH, 0);
+	tmp = REG_SET_FIELD(tmp, VM_CONTEXT0_CNTL,
+			    RETRY_PERMISSION_OR_INVALID_PAGE_FAULT, 0);
 	WREG32_SOC15(MMHUB, 0, mmVM_CONTEXT0_CNTL, tmp);
 }
 
@@ -418,6 +420,8 @@ void mmhub_v1_0_init(struct amdgpu_device *adev)
 	hub->ctx0_ptb_addr_hi32 =
 		SOC15_REG_OFFSET(MMHUB, 0,
 				 mmVM_CONTEXT0_PAGE_TABLE_BASE_ADDR_HI32);
+	hub->vm_inv_eng0_sem =
+		SOC15_REG_OFFSET(MMHUB, 0, mmVM_INVALIDATE_ENG0_SEM);
 	hub->vm_inv_eng0_req =
 		SOC15_REG_OFFSET(MMHUB, 0, mmVM_INVALIDATE_ENG0_REQ);
 	hub->vm_inv_eng0_ack =
@@ -616,5 +620,6 @@ static void mmhub_v1_0_query_ras_error_count(struct amdgpu_device *adev,
 }
 
 const struct amdgpu_mmhub_funcs mmhub_v1_0_funcs = {
+	.ras_late_init = amdgpu_mmhub_ras_late_init,
 	.query_ras_error_count = mmhub_v1_0_query_ras_error_count,
 };
diff --git a/drivers/gpu/drm/amd/amdgpu/mmhub_v2_0.c b/drivers/gpu/drm/amd/amdgpu/mmhub_v2_0.c
index 3542c203c3c8..a7cb185d639a 100644
--- a/drivers/gpu/drm/amd/amdgpu/mmhub_v2_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/mmhub_v2_0.c
@@ -31,20 +31,25 @@
 
 #include "soc15_common.h"
 
-static void mmhub_v2_0_init_gart_pt_regs(struct amdgpu_device *adev)
+void mmhub_v2_0_setup_vm_pt_regs(struct amdgpu_device *adev, uint32_t vmid,
+				uint64_t page_table_base)
 {
-	uint64_t value = amdgpu_gmc_pd_addr(adev->gart.bo);
+	/* two registers distance between mmMMVM_CONTEXT0_* to mmMMVM_CONTEXT1_* */
+	int offset = mmMMVM_CONTEXT1_PAGE_TABLE_BASE_ADDR_LO32
+			- mmMMVM_CONTEXT0_PAGE_TABLE_BASE_ADDR_LO32;
 
-	WREG32_SOC15(MMHUB, 0, mmMMVM_CONTEXT0_PAGE_TABLE_BASE_ADDR_LO32,
-		     lower_32_bits(value));
+	WREG32_SOC15_OFFSET(MMHUB, 0, mmMMVM_CONTEXT0_PAGE_TABLE_BASE_ADDR_LO32,
+			offset * vmid, lower_32_bits(page_table_base));
 
-	WREG32_SOC15(MMHUB, 0, mmMMVM_CONTEXT0_PAGE_TABLE_BASE_ADDR_HI32,
-		     upper_32_bits(value));
+	WREG32_SOC15_OFFSET(MMHUB, 0, mmMMVM_CONTEXT0_PAGE_TABLE_BASE_ADDR_HI32,
+			offset * vmid, upper_32_bits(page_table_base));
 }
 
 static void mmhub_v2_0_init_gart_aperture_regs(struct amdgpu_device *adev)
 {
-	mmhub_v2_0_init_gart_pt_regs(adev);
+	uint64_t pt_base = amdgpu_gmc_pd_addr(adev->gart.bo);
+
+	mmhub_v2_0_setup_vm_pt_regs(adev, 0, pt_base);
 
 	WREG32_SOC15(MMHUB, 0, mmMMVM_CONTEXT0_PAGE_TABLE_START_ADDR_LO32,
 		     (u32)(adev->gmc.gart_start >> 12));
@@ -137,6 +142,15 @@ static void mmhub_v2_0_init_cache_regs(struct amdgpu_device *adev)
 	WREG32_SOC15(MMHUB, 0, mmMMVM_L2_CNTL2, tmp);
 
 	tmp = mmMMVM_L2_CNTL3_DEFAULT;
+	if (adev->gmc.translate_further) {
+		tmp = REG_SET_FIELD(tmp, MMVM_L2_CNTL3, BANK_SELECT, 12);
+		tmp = REG_SET_FIELD(tmp, MMVM_L2_CNTL3,
+				    L2_CACHE_BIGK_FRAGMENT_SIZE, 9);
+	} else {
+		tmp = REG_SET_FIELD(tmp, MMVM_L2_CNTL3, BANK_SELECT, 9);
+		tmp = REG_SET_FIELD(tmp, MMVM_L2_CNTL3,
+				    L2_CACHE_BIGK_FRAGMENT_SIZE, 6);
+	}
 	WREG32_SOC15(MMHUB, 0, mmMMVM_L2_CNTL3, tmp);
 
 	tmp = mmMMVM_L2_CNTL4_DEFAULT;
@@ -152,6 +166,8 @@ static void mmhub_v2_0_enable_system_domain(struct amdgpu_device *adev)
 	tmp = RREG32_SOC15(MMHUB, 0, mmMMVM_CONTEXT0_CNTL);
 	tmp = REG_SET_FIELD(tmp, MMVM_CONTEXT0_CNTL, ENABLE_CONTEXT, 1);
 	tmp = REG_SET_FIELD(tmp, MMVM_CONTEXT0_CNTL, PAGE_TABLE_DEPTH, 0);
+	tmp = REG_SET_FIELD(tmp, MMVM_CONTEXT0_CNTL,
+			    RETRY_PERMISSION_OR_INVALID_PAGE_FAULT, 0);
 	WREG32_SOC15(MMHUB, 0, mmMMVM_CONTEXT0_CNTL, tmp);
 }
 
@@ -332,6 +348,8 @@ void mmhub_v2_0_init(struct amdgpu_device *adev)
 	hub->ctx0_ptb_addr_hi32 =
 		SOC15_REG_OFFSET(MMHUB, 0,
 				 mmMMVM_CONTEXT0_PAGE_TABLE_BASE_ADDR_HI32);
+	hub->vm_inv_eng0_sem =
+		SOC15_REG_OFFSET(MMHUB, 0, mmMMVM_INVALIDATE_ENG0_SEM);
 	hub->vm_inv_eng0_req =
 		SOC15_REG_OFFSET(MMHUB, 0, mmMMVM_INVALIDATE_ENG0_REQ);
 	hub->vm_inv_eng0_ack =
diff --git a/drivers/gpu/drm/amd/amdgpu/mmhub_v2_0.h b/drivers/gpu/drm/amd/amdgpu/mmhub_v2_0.h
index db16f3ece218..3ea4344f0315 100644
--- a/drivers/gpu/drm/amd/amdgpu/mmhub_v2_0.h
+++ b/drivers/gpu/drm/amd/amdgpu/mmhub_v2_0.h
@@ -31,5 +31,7 @@ void mmhub_v2_0_init(struct amdgpu_device *adev);
 int mmhub_v2_0_set_clockgating(struct amdgpu_device *adev,
 			       enum amd_clockgating_state state);
 void mmhub_v2_0_get_clockgating(struct amdgpu_device *adev, u32 *flags);
+void mmhub_v2_0_setup_vm_pt_regs(struct amdgpu_device *adev, uint32_t vmid,
+				uint64_t page_table_base);
 
 #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/mmhub_v9_4.c b/drivers/gpu/drm/amd/amdgpu/mmhub_v9_4.c
index 0cf7ef44b4b5..66efe2f7bd76 100644
--- a/drivers/gpu/drm/amd/amdgpu/mmhub_v9_4.c
+++ b/drivers/gpu/drm/amd/amdgpu/mmhub_v9_4.c
@@ -219,6 +219,15 @@ static void mmhub_v9_4_init_cache_regs(struct amdgpu_device *adev, int hubid)
 			    hubid * MMHUB_INSTANCE_REGISTER_OFFSET, tmp);
 
 	tmp = mmVML2PF0_VM_L2_CNTL3_DEFAULT;
+	if (adev->gmc.translate_further) {
+		tmp = REG_SET_FIELD(tmp, VML2PF0_VM_L2_CNTL3, BANK_SELECT, 12);
+		tmp = REG_SET_FIELD(tmp, VML2PF0_VM_L2_CNTL3,
+				    L2_CACHE_BIGK_FRAGMENT_SIZE, 9);
+	} else {
+		tmp = REG_SET_FIELD(tmp, VML2PF0_VM_L2_CNTL3, BANK_SELECT, 9);
+		tmp = REG_SET_FIELD(tmp, VML2PF0_VM_L2_CNTL3,
+				    L2_CACHE_BIGK_FRAGMENT_SIZE, 6);
+	}
 	WREG32_SOC15_OFFSET(MMHUB, 0, mmVML2PF0_VM_L2_CNTL3,
 			    hubid * MMHUB_INSTANCE_REGISTER_OFFSET, tmp);
 
@@ -240,6 +249,8 @@ static void mmhub_v9_4_enable_system_domain(struct amdgpu_device *adev,
 				  hubid * MMHUB_INSTANCE_REGISTER_OFFSET);
 	tmp = REG_SET_FIELD(tmp, VML2VC0_VM_CONTEXT0_CNTL, ENABLE_CONTEXT, 1);
 	tmp = REG_SET_FIELD(tmp, VML2VC0_VM_CONTEXT0_CNTL, PAGE_TABLE_DEPTH, 0);
+	tmp = REG_SET_FIELD(tmp, VML2VC0_VM_CONTEXT0_CNTL,
+			    RETRY_PERMISSION_OR_INVALID_PAGE_FAULT, 0);
 	WREG32_SOC15_OFFSET(MMHUB, 0, mmVML2VC0_VM_CONTEXT0_CNTL,
 			    hubid * MMHUB_INSTANCE_REGISTER_OFFSET, tmp);
 }
@@ -493,6 +504,10 @@ void mmhub_v9_4_init(struct amdgpu_device *adev)
 			SOC15_REG_OFFSET(MMHUB, 0,
 			    mmVML2VC0_VM_CONTEXT0_PAGE_TABLE_BASE_ADDR_HI32) +
 			    i * MMHUB_INSTANCE_REGISTER_OFFSET;
+		hub[i]->vm_inv_eng0_sem =
+			SOC15_REG_OFFSET(MMHUB, 0,
+					 mmVML2VC0_VM_INVALIDATE_ENG0_SEM) +
+					 i * MMHUB_INSTANCE_REGISTER_OFFSET;
 		hub[i]->vm_inv_eng0_req =
 			SOC15_REG_OFFSET(MMHUB, 0,
 					 mmVML2VC0_VM_INVALIDATE_ENG0_REQ) +
diff --git a/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c b/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c
new file mode 100644
index 000000000000..0d8767eb7a70
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c
@@ -0,0 +1,380 @@
+/*
+ * Copyright 2014 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#include "amdgpu.h"
+#include "nbio/nbio_2_3_offset.h"
+#include "nbio/nbio_2_3_sh_mask.h"
+#include "gc/gc_10_1_0_offset.h"
+#include "gc/gc_10_1_0_sh_mask.h"
+#include "soc15.h"
+#include "navi10_ih.h"
+#include "soc15_common.h"
+#include "mxgpu_nv.h"
+#include "mxgpu_ai.h"
+
+static void xgpu_nv_mailbox_send_ack(struct amdgpu_device *adev)
+{
+	WREG8(NV_MAIBOX_CONTROL_RCV_OFFSET_BYTE, 2);
+}
+
+static void xgpu_nv_mailbox_set_valid(struct amdgpu_device *adev, bool val)
+{
+	WREG8(NV_MAIBOX_CONTROL_TRN_OFFSET_BYTE, val ? 1 : 0);
+}
+
+/*
+ * this peek_msg could *only* be called in IRQ routine becuase in IRQ routine
+ * RCV_MSG_VALID filed of BIF_BX_PF_MAILBOX_CONTROL must already be set to 1
+ * by host.
+ *
+ * if called no in IRQ routine, this peek_msg cannot guaranteed to return the
+ * correct value since it doesn't return the RCV_DW0 under the case that
+ * RCV_MSG_VALID is set by host.
+ */
+static enum idh_event xgpu_nv_mailbox_peek_msg(struct amdgpu_device *adev)
+{
+	return RREG32_NO_KIQ(SOC15_REG_OFFSET(NBIO, 0,
+				mmBIF_BX_PF_MAILBOX_MSGBUF_RCV_DW0));
+}
+
+
+static int xgpu_nv_mailbox_rcv_msg(struct amdgpu_device *adev,
+				   enum idh_event event)
+{
+	u32 reg;
+
+	reg = RREG32_NO_KIQ(SOC15_REG_OFFSET(NBIO, 0,
+					     mmBIF_BX_PF_MAILBOX_MSGBUF_RCV_DW0));
+	if (reg != event)
+		return -ENOENT;
+
+	xgpu_nv_mailbox_send_ack(adev);
+
+	return 0;
+}
+
+static uint8_t xgpu_nv_peek_ack(struct amdgpu_device *adev)
+{
+	return RREG8(NV_MAIBOX_CONTROL_TRN_OFFSET_BYTE) & 2;
+}
+
+static int xgpu_nv_poll_ack(struct amdgpu_device *adev)
+{
+	int timeout  = NV_MAILBOX_POLL_ACK_TIMEDOUT;
+	u8 reg;
+
+	do {
+		reg = RREG8(NV_MAIBOX_CONTROL_TRN_OFFSET_BYTE);
+		if (reg & 2)
+			return 0;
+
+		mdelay(5);
+		timeout -= 5;
+	} while (timeout > 1);
+
+	pr_err("Doesn't get TRN_MSG_ACK from pf in %d msec\n", NV_MAILBOX_POLL_ACK_TIMEDOUT);
+
+	return -ETIME;
+}
+
+static int xgpu_nv_poll_msg(struct amdgpu_device *adev, enum idh_event event)
+{
+	int r, timeout = NV_MAILBOX_POLL_MSG_TIMEDOUT;
+
+	do {
+		r = xgpu_nv_mailbox_rcv_msg(adev, event);
+		if (!r)
+			return 0;
+
+		msleep(10);
+		timeout -= 10;
+	} while (timeout > 1);
+
+	pr_err("Doesn't get msg:%d from pf, error=%d\n", event, r);
+
+	return -ETIME;
+}
+
+static void xgpu_nv_mailbox_trans_msg (struct amdgpu_device *adev,
+	      enum idh_request req, u32 data1, u32 data2, u32 data3)
+{
+	u32 reg;
+	int r;
+	uint8_t trn;
+
+	/* IMPORTANT:
+	 * clear TRN_MSG_VALID valid to clear host's RCV_MSG_ACK
+	 * and with host's RCV_MSG_ACK cleared hw automatically clear host's RCV_MSG_ACK
+	 * which lead to VF's TRN_MSG_ACK cleared, otherwise below xgpu_nv_poll_ack()
+	 * will return immediatly
+	 */
+	do {
+		xgpu_nv_mailbox_set_valid(adev, false);
+		trn = xgpu_nv_peek_ack(adev);
+		if (trn) {
+			pr_err("trn=%x ACK should not assert! wait again !\n", trn);
+			msleep(1);
+		}
+	} while (trn);
+
+	reg = RREG32_NO_KIQ(SOC15_REG_OFFSET(NBIO, 0,
+					     mmBIF_BX_PF_MAILBOX_MSGBUF_TRN_DW0));
+	reg = REG_SET_FIELD(reg, BIF_BX_PF_MAILBOX_MSGBUF_TRN_DW0,
+			    MSGBUF_DATA, req);
+	WREG32_NO_KIQ(SOC15_REG_OFFSET(NBIO, 0, mmBIF_BX_PF_MAILBOX_MSGBUF_TRN_DW0),
+		      reg);
+	WREG32_NO_KIQ(SOC15_REG_OFFSET(NBIO, 0, mmBIF_BX_PF_MAILBOX_MSGBUF_TRN_DW1),
+				data1);
+	WREG32_NO_KIQ(SOC15_REG_OFFSET(NBIO, 0, mmBIF_BX_PF_MAILBOX_MSGBUF_TRN_DW2),
+				data2);
+	WREG32_NO_KIQ(SOC15_REG_OFFSET(NBIO, 0, mmBIF_BX_PF_MAILBOX_MSGBUF_TRN_DW3),
+				data3);
+
+	xgpu_nv_mailbox_set_valid(adev, true);
+
+	/* start to poll ack */
+	r = xgpu_nv_poll_ack(adev);
+	if (r)
+		pr_err("Doesn't get ack from pf, continue\n");
+
+	xgpu_nv_mailbox_set_valid(adev, false);
+}
+
+static int xgpu_nv_send_access_requests(struct amdgpu_device *adev,
+					enum idh_request req)
+{
+	int r;
+
+	xgpu_nv_mailbox_trans_msg(adev, req, 0, 0, 0);
+
+	/* start to check msg if request is idh_req_gpu_init_access */
+	if (req == IDH_REQ_GPU_INIT_ACCESS ||
+		req == IDH_REQ_GPU_FINI_ACCESS ||
+		req == IDH_REQ_GPU_RESET_ACCESS) {
+		r = xgpu_nv_poll_msg(adev, IDH_READY_TO_ACCESS_GPU);
+		if (r) {
+			pr_err("Doesn't get READY_TO_ACCESS_GPU from pf, give up\n");
+			return r;
+		}
+		/* Retrieve checksum from mailbox2 */
+		if (req == IDH_REQ_GPU_INIT_ACCESS || req == IDH_REQ_GPU_RESET_ACCESS) {
+			adev->virt.fw_reserve.checksum_key =
+				RREG32_NO_KIQ(SOC15_REG_OFFSET(NBIO, 0,
+					mmBIF_BX_PF_MAILBOX_MSGBUF_RCV_DW2));
+		}
+	}
+
+	return 0;
+}
+
+static int xgpu_nv_request_reset(struct amdgpu_device *adev)
+{
+	return xgpu_nv_send_access_requests(adev, IDH_REQ_GPU_RESET_ACCESS);
+}
+
+static int xgpu_nv_request_full_gpu_access(struct amdgpu_device *adev,
+					   bool init)
+{
+	enum idh_request req;
+
+	req = init ? IDH_REQ_GPU_INIT_ACCESS : IDH_REQ_GPU_FINI_ACCESS;
+	return xgpu_nv_send_access_requests(adev, req);
+}
+
+static int xgpu_nv_release_full_gpu_access(struct amdgpu_device *adev,
+					   bool init)
+{
+	enum idh_request req;
+	int r = 0;
+
+	req = init ? IDH_REL_GPU_INIT_ACCESS : IDH_REL_GPU_FINI_ACCESS;
+	r = xgpu_nv_send_access_requests(adev, req);
+
+	return r;
+}
+
+static int xgpu_nv_mailbox_ack_irq(struct amdgpu_device *adev,
+					struct amdgpu_irq_src *source,
+					struct amdgpu_iv_entry *entry)
+{
+	DRM_DEBUG("get ack intr and do nothing.\n");
+	return 0;
+}
+
+static int xgpu_nv_set_mailbox_ack_irq(struct amdgpu_device *adev,
+					struct amdgpu_irq_src *source,
+					unsigned type,
+					enum amdgpu_interrupt_state state)
+{
+	u32 tmp = RREG32_NO_KIQ(SOC15_REG_OFFSET(NBIO, 0, mmBIF_BX_PF_MAILBOX_INT_CNTL));
+
+	tmp = REG_SET_FIELD(tmp, BIF_BX_PF_MAILBOX_INT_CNTL, ACK_INT_EN,
+				(state == AMDGPU_IRQ_STATE_ENABLE) ? 1 : 0);
+	WREG32_NO_KIQ(SOC15_REG_OFFSET(NBIO, 0, mmBIF_BX_PF_MAILBOX_INT_CNTL), tmp);
+
+	return 0;
+}
+
+static void xgpu_nv_mailbox_flr_work(struct work_struct *work)
+{
+	struct amdgpu_virt *virt = container_of(work, struct amdgpu_virt, flr_work);
+	struct amdgpu_device *adev = container_of(virt, struct amdgpu_device, virt);
+	int timeout = NV_MAILBOX_POLL_FLR_TIMEDOUT;
+	int locked;
+
+	/* block amdgpu_gpu_recover till msg FLR COMPLETE received,
+	 * otherwise the mailbox msg will be ruined/reseted by
+	 * the VF FLR.
+	 *
+	 * we can unlock the lock_reset to allow "amdgpu_job_timedout"
+	 * to run gpu_recover() after FLR_NOTIFICATION_CMPL received
+	 * which means host side had finished this VF's FLR.
+	 */
+	locked = mutex_trylock(&adev->lock_reset);
+	if (locked)
+		adev->in_gpu_reset = 1;
+
+	do {
+		if (xgpu_nv_mailbox_peek_msg(adev) == IDH_FLR_NOTIFICATION_CMPL)
+			goto flr_done;
+
+		msleep(10);
+		timeout -= 10;
+	} while (timeout > 1);
+
+flr_done:
+	if (locked) {
+		adev->in_gpu_reset = 0;
+		mutex_unlock(&adev->lock_reset);
+	}
+
+	/* Trigger recovery for world switch failure if no TDR */
+	if (amdgpu_device_should_recover_gpu(adev))
+		amdgpu_device_gpu_recover(adev, NULL);
+}
+
+static int xgpu_nv_set_mailbox_rcv_irq(struct amdgpu_device *adev,
+				       struct amdgpu_irq_src *src,
+				       unsigned type,
+				       enum amdgpu_interrupt_state state)
+{
+	u32 tmp = RREG32_NO_KIQ(SOC15_REG_OFFSET(NBIO, 0, mmBIF_BX_PF_MAILBOX_INT_CNTL));
+
+	tmp = REG_SET_FIELD(tmp, BIF_BX_PF_MAILBOX_INT_CNTL, VALID_INT_EN,
+			    (state == AMDGPU_IRQ_STATE_ENABLE) ? 1 : 0);
+	WREG32_NO_KIQ(SOC15_REG_OFFSET(NBIO, 0, mmBIF_BX_PF_MAILBOX_INT_CNTL), tmp);
+
+	return 0;
+}
+
+static int xgpu_nv_mailbox_rcv_irq(struct amdgpu_device *adev,
+				   struct amdgpu_irq_src *source,
+				   struct amdgpu_iv_entry *entry)
+{
+	enum idh_event event = xgpu_nv_mailbox_peek_msg(adev);
+
+	switch (event) {
+	case IDH_FLR_NOTIFICATION:
+		if (amdgpu_sriov_runtime(adev))
+			schedule_work(&adev->virt.flr_work);
+		break;
+		/* READY_TO_ACCESS_GPU is fetched by kernel polling, IRQ can ignore
+		 * it byfar since that polling thread will handle it,
+		 * other msg like flr complete is not handled here.
+		 */
+	case IDH_CLR_MSG_BUF:
+	case IDH_FLR_NOTIFICATION_CMPL:
+	case IDH_READY_TO_ACCESS_GPU:
+	default:
+		break;
+	}
+
+	return 0;
+}
+
+static const struct amdgpu_irq_src_funcs xgpu_nv_mailbox_ack_irq_funcs = {
+	.set = xgpu_nv_set_mailbox_ack_irq,
+	.process = xgpu_nv_mailbox_ack_irq,
+};
+
+static const struct amdgpu_irq_src_funcs xgpu_nv_mailbox_rcv_irq_funcs = {
+	.set = xgpu_nv_set_mailbox_rcv_irq,
+	.process = xgpu_nv_mailbox_rcv_irq,
+};
+
+void xgpu_nv_mailbox_set_irq_funcs(struct amdgpu_device *adev)
+{
+	adev->virt.ack_irq.num_types = 1;
+	adev->virt.ack_irq.funcs = &xgpu_nv_mailbox_ack_irq_funcs;
+	adev->virt.rcv_irq.num_types = 1;
+	adev->virt.rcv_irq.funcs = &xgpu_nv_mailbox_rcv_irq_funcs;
+}
+
+int xgpu_nv_mailbox_add_irq_id(struct amdgpu_device *adev)
+{
+	int r;
+
+	r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_BIF, 135, &adev->virt.rcv_irq);
+	if (r)
+		return r;
+
+	r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_BIF, 138, &adev->virt.ack_irq);
+	if (r) {
+		amdgpu_irq_put(adev, &adev->virt.rcv_irq, 0);
+		return r;
+	}
+
+	return 0;
+}
+
+int xgpu_nv_mailbox_get_irq(struct amdgpu_device *adev)
+{
+	int r;
+
+	r = amdgpu_irq_get(adev, &adev->virt.rcv_irq, 0);
+	if (r)
+		return r;
+	r = amdgpu_irq_get(adev, &adev->virt.ack_irq, 0);
+	if (r) {
+		amdgpu_irq_put(adev, &adev->virt.rcv_irq, 0);
+		return r;
+	}
+
+	INIT_WORK(&adev->virt.flr_work, xgpu_nv_mailbox_flr_work);
+
+	return 0;
+}
+
+void xgpu_nv_mailbox_put_irq(struct amdgpu_device *adev)
+{
+	amdgpu_irq_put(adev, &adev->virt.ack_irq, 0);
+	amdgpu_irq_put(adev, &adev->virt.rcv_irq, 0);
+}
+
+const struct amdgpu_virt_ops xgpu_nv_virt_ops = {
+	.req_full_gpu	= xgpu_nv_request_full_gpu_access,
+	.rel_full_gpu	= xgpu_nv_release_full_gpu_access,
+	.reset_gpu = xgpu_nv_request_reset,
+	.wait_reset = NULL,
+	.trans_msg = xgpu_nv_mailbox_trans_msg,
+};
diff --git a/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.h b/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.h
new file mode 100644
index 000000000000..99b15f6865cb
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.h
@@ -0,0 +1,41 @@
+/*
+ * Copyright 2014 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#ifndef __MXGPU_NV_H__
+#define __MXGPU_NV_H__
+
+#define NV_MAILBOX_POLL_ACK_TIMEDOUT	500
+#define NV_MAILBOX_POLL_MSG_TIMEDOUT	12000
+#define NV_MAILBOX_POLL_FLR_TIMEDOUT	500
+
+extern const struct amdgpu_virt_ops xgpu_nv_virt_ops;
+
+void xgpu_nv_mailbox_set_irq_funcs(struct amdgpu_device *adev);
+int xgpu_nv_mailbox_add_irq_id(struct amdgpu_device *adev);
+int xgpu_nv_mailbox_get_irq(struct amdgpu_device *adev);
+void xgpu_nv_mailbox_put_irq(struct amdgpu_device *adev);
+
+#define NV_MAIBOX_CONTROL_TRN_OFFSET_BYTE (SOC15_REG_OFFSET(NBIO, 0, mmBIF_BX_PF_MAILBOX_CONTROL) * 4)
+#define NV_MAIBOX_CONTROL_RCV_OFFSET_BYTE (SOC15_REG_OFFSET(NBIO, 0, mmBIF_BX_PF_MAILBOX_CONTROL) * 4 + 1)
+
+#endif
diff --git a/drivers/gpu/drm/amd/amdgpu/navi10_ih.c b/drivers/gpu/drm/amd/amdgpu/navi10_ih.c
index 9fe08408db58..9af73567e716 100644
--- a/drivers/gpu/drm/amd/amdgpu/navi10_ih.c
+++ b/drivers/gpu/drm/amd/amdgpu/navi10_ih.c
@@ -117,7 +117,7 @@ static int navi10_ih_irq_init(struct amdgpu_device *adev)
 	/* disable irqs */
 	navi10_ih_disable_interrupts(adev);
 
-	adev->nbio_funcs->ih_control(adev);
+	adev->nbio.funcs->ih_control(adev);
 
 	/* Ring Buffer base. [39:8] of 40-bit address of the beginning of the ring buffer*/
 	WREG32_SOC15(OSSSYS, 0, mmIH_RB_BASE, ih->gpu_addr >> 8);
@@ -162,7 +162,7 @@ static int navi10_ih_irq_init(struct amdgpu_device *adev)
 	}
 	WREG32_SOC15(OSSSYS, 0, mmIH_DOORBELL_RPTR, ih_doorbell_rtpr);
 
-	adev->nbio_funcs->ih_doorbell_range(adev, ih->use_doorbell,
+	adev->nbio.funcs->ih_doorbell_range(adev, ih->use_doorbell,
 					    ih->doorbell_index);
 
 	tmp = RREG32_SOC15(OSSSYS, 0, mmIH_STORM_CLIENT_LIST_CNTL);
diff --git a/drivers/gpu/drm/amd/amdgpu/navi10_reg_init.c b/drivers/gpu/drm/amd/amdgpu/navi10_reg_init.c
index a56c93620e78..88efaecf9f70 100644
--- a/drivers/gpu/drm/amd/amdgpu/navi10_reg_init.c
+++ b/drivers/gpu/drm/amd/amdgpu/navi10_reg_init.c
@@ -24,7 +24,6 @@
 #include "nv.h"
 
 #include "soc15_common.h"
-#include "soc15_hw_ip.h"
 #include "navi10_ip_offset.h"
 
 int navi10_reg_base_init(struct amdgpu_device *adev)
diff --git a/drivers/gpu/drm/amd/amdgpu/navi12_reg_init.c b/drivers/gpu/drm/amd/amdgpu/navi12_reg_init.c
index cadc7603ca41..a786d159e5e9 100644
--- a/drivers/gpu/drm/amd/amdgpu/navi12_reg_init.c
+++ b/drivers/gpu/drm/amd/amdgpu/navi12_reg_init.c
@@ -24,7 +24,6 @@
 #include "nv.h"
 
 #include "soc15_common.h"
-#include "soc15_hw_ip.h"
 #include "navi12_ip_offset.h"
 
 int navi12_reg_base_init(struct amdgpu_device *adev)
diff --git a/drivers/gpu/drm/amd/amdgpu/navi14_reg_init.c b/drivers/gpu/drm/amd/amdgpu/navi14_reg_init.c
index 3b5f0f65e096..4ea1e8fbb601 100644
--- a/drivers/gpu/drm/amd/amdgpu/navi14_reg_init.c
+++ b/drivers/gpu/drm/amd/amdgpu/navi14_reg_init.c
@@ -24,7 +24,6 @@
 #include "nv.h"
 
 #include "soc15_common.h"
-#include "soc15_hw_ip.h"
 #include "navi14_ip_offset.h"
 
 int navi14_reg_base_init(struct amdgpu_device *adev)
diff --git a/drivers/gpu/drm/amd/amdgpu/nbio_v2_3.c b/drivers/gpu/drm/amd/amdgpu/nbio_v2_3.c
index c05d78d4efc6..f3a3fe746222 100644
--- a/drivers/gpu/drm/amd/amdgpu/nbio_v2_3.c
+++ b/drivers/gpu/drm/amd/amdgpu/nbio_v2_3.c
@@ -27,11 +27,21 @@
 #include "nbio/nbio_2_3_default.h"
 #include "nbio/nbio_2_3_offset.h"
 #include "nbio/nbio_2_3_sh_mask.h"
+#include <uapi/linux/kfd_ioctl.h>
 
 #define smnPCIE_CONFIG_CNTL	0x11180044
 #define smnCPM_CONTROL		0x11180460
 #define smnPCIE_CNTL2		0x11180070
 
+
+static void nbio_v2_3_remap_hdp_registers(struct amdgpu_device *adev)
+{
+	WREG32_SOC15(NBIO, 0, mmREMAP_HDP_MEM_FLUSH_CNTL,
+		adev->rmmio_remap.reg_offset + KFD_MMIO_REMAP_HDP_MEM_FLUSH_CNTL);
+	WREG32_SOC15(NBIO, 0, mmREMAP_HDP_REG_FLUSH_CNTL,
+		adev->rmmio_remap.reg_offset + KFD_MMIO_REMAP_HDP_REG_FLUSH_CNTL);
+}
+
 static u32 nbio_v2_3_get_rev_id(struct amdgpu_device *adev)
 {
 	u32 tmp = RREG32_SOC15(NBIO, 0, mmRCC_DEV0_EPF0_STRAP0);
@@ -56,10 +66,9 @@ static void nbio_v2_3_hdp_flush(struct amdgpu_device *adev,
 				struct amdgpu_ring *ring)
 {
 	if (!ring || !ring->funcs->emit_wreg)
-		WREG32_SOC15_NO_KIQ(NBIO, 0, mmBIF_BX_PF_HDP_MEM_COHERENCY_FLUSH_CNTL, 0);
+		WREG32_NO_KIQ((adev->rmmio_remap.reg_offset + KFD_MMIO_REMAP_HDP_MEM_FLUSH_CNTL) >> 2, 0);
 	else
-		amdgpu_ring_emit_wreg(ring, SOC15_REG_OFFSET(
-			NBIO, 0, mmBIF_BX_PF_HDP_MEM_COHERENCY_FLUSH_CNTL), 0);
+		amdgpu_ring_emit_wreg(ring, (adev->rmmio_remap.reg_offset + KFD_MMIO_REMAP_HDP_MEM_FLUSH_CNTL) >> 2, 0);
 }
 
 static u32 nbio_v2_3_get_memsize(struct amdgpu_device *adev)
@@ -311,7 +320,6 @@ static void nbio_v2_3_init_registers(struct amdgpu_device *adev)
 }
 
 const struct amdgpu_nbio_funcs nbio_v2_3_funcs = {
-	.hdp_flush_reg = &nbio_v2_3_hdp_flush_reg,
 	.get_hdp_flush_req_offset = nbio_v2_3_get_hdp_flush_req_offset,
 	.get_hdp_flush_done_offset = nbio_v2_3_get_hdp_flush_done_offset,
 	.get_pcie_index_offset = nbio_v2_3_get_pcie_index_offset,
@@ -331,4 +339,5 @@ const struct amdgpu_nbio_funcs nbio_v2_3_funcs = {
 	.ih_control = nbio_v2_3_ih_control,
 	.init_registers = nbio_v2_3_init_registers,
 	.detect_hw_virt = nbio_v2_3_detect_hw_virt,
+	.remap_hdp_registers = nbio_v2_3_remap_hdp_registers,
 };
diff --git a/drivers/gpu/drm/amd/amdgpu/nbio_v2_3.h b/drivers/gpu/drm/amd/amdgpu/nbio_v2_3.h
index 5ae52085f6b7..a43b60acf7f6 100644
--- a/drivers/gpu/drm/amd/amdgpu/nbio_v2_3.h
+++ b/drivers/gpu/drm/amd/amdgpu/nbio_v2_3.h
@@ -26,6 +26,7 @@
 
 #include "soc15_common.h"
 
+extern const struct nbio_hdp_flush_reg nbio_v2_3_hdp_flush_reg;
 extern const struct amdgpu_nbio_funcs nbio_v2_3_funcs;
 
 #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/nbio_v6_1.c b/drivers/gpu/drm/amd/amdgpu/nbio_v6_1.c
index 6590143c3f75..635d9e1fc0a3 100644
--- a/drivers/gpu/drm/amd/amdgpu/nbio_v6_1.c
+++ b/drivers/gpu/drm/amd/amdgpu/nbio_v6_1.c
@@ -226,7 +226,7 @@ static u32 nbio_v6_1_get_pcie_data_offset(struct amdgpu_device *adev)
 	return SOC15_REG_OFFSET(NBIO, 0, mmPCIE_DATA2);
 }
 
-static const struct nbio_hdp_flush_reg nbio_v6_1_hdp_flush_reg = {
+const struct nbio_hdp_flush_reg nbio_v6_1_hdp_flush_reg = {
 	.ref_and_mask_cp0 = BIF_BX_PF0_GPU_HDP_FLUSH_DONE__CP0_MASK,
 	.ref_and_mask_cp1 = BIF_BX_PF0_GPU_HDP_FLUSH_DONE__CP1_MASK,
 	.ref_and_mask_cp2 = BIF_BX_PF0_GPU_HDP_FLUSH_DONE__CP2_MASK,
@@ -277,7 +277,6 @@ static void nbio_v6_1_init_registers(struct amdgpu_device *adev)
 }
 
 const struct amdgpu_nbio_funcs nbio_v6_1_funcs = {
-	.hdp_flush_reg = &nbio_v6_1_hdp_flush_reg,
 	.get_hdp_flush_req_offset = nbio_v6_1_get_hdp_flush_req_offset,
 	.get_hdp_flush_done_offset = nbio_v6_1_get_hdp_flush_done_offset,
 	.get_pcie_index_offset = nbio_v6_1_get_pcie_index_offset,
diff --git a/drivers/gpu/drm/amd/amdgpu/nbio_v6_1.h b/drivers/gpu/drm/amd/amdgpu/nbio_v6_1.h
index 0743a6f016f3..6dc743b73218 100644
--- a/drivers/gpu/drm/amd/amdgpu/nbio_v6_1.h
+++ b/drivers/gpu/drm/amd/amdgpu/nbio_v6_1.h
@@ -26,6 +26,7 @@
 
 #include "soc15_common.h"
 
+extern const struct nbio_hdp_flush_reg nbio_v6_1_hdp_flush_reg;
 extern const struct amdgpu_nbio_funcs nbio_v6_1_funcs;
 
 #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/nbio_v7_0.c b/drivers/gpu/drm/amd/amdgpu/nbio_v7_0.c
index 74eecb768a82..d6cbf26074bc 100644
--- a/drivers/gpu/drm/amd/amdgpu/nbio_v7_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/nbio_v7_0.c
@@ -292,7 +292,6 @@ static void nbio_v7_0_init_registers(struct amdgpu_device *adev)
 }
 
 const struct amdgpu_nbio_funcs nbio_v7_0_funcs = {
-	.hdp_flush_reg = &nbio_v7_0_hdp_flush_reg,
 	.get_hdp_flush_req_offset = nbio_v7_0_get_hdp_flush_req_offset,
 	.get_hdp_flush_done_offset = nbio_v7_0_get_hdp_flush_done_offset,
 	.get_pcie_index_offset = nbio_v7_0_get_pcie_index_offset,
diff --git a/drivers/gpu/drm/amd/amdgpu/nbio_v7_0.h b/drivers/gpu/drm/amd/amdgpu/nbio_v7_0.h
index 508d549c5029..e7aefb252550 100644
--- a/drivers/gpu/drm/amd/amdgpu/nbio_v7_0.h
+++ b/drivers/gpu/drm/amd/amdgpu/nbio_v7_0.h
@@ -26,6 +26,7 @@
 
 #include "soc15_common.h"
 
+extern const struct nbio_hdp_flush_reg nbio_v7_0_hdp_flush_reg;
 extern const struct amdgpu_nbio_funcs nbio_v7_0_funcs;
 
 #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/nbio_v7_4.c b/drivers/gpu/drm/amd/amdgpu/nbio_v7_4.c
index 910fffced43b..0db458f9fafc 100644
--- a/drivers/gpu/drm/amd/amdgpu/nbio_v7_4.c
+++ b/drivers/gpu/drm/amd/amdgpu/nbio_v7_4.c
@@ -23,10 +23,12 @@
 #include "amdgpu.h"
 #include "amdgpu_atombios.h"
 #include "nbio_v7_4.h"
+#include "amdgpu_ras.h"
 
 #include "nbio/nbio_7_4_offset.h"
 #include "nbio/nbio_7_4_sh_mask.h"
 #include "nbio/nbio_7_4_0_smn.h"
+#include "ivsrcid/nbio/irqsrcs_nbif_7_4.h"
 #include <uapi/linux/kfd_ioctl.h>
 
 #define smnNBIF_MGCG_CTRL_LCLK	0x1013a21c
@@ -266,7 +268,7 @@ static u32 nbio_v7_4_get_pcie_data_offset(struct amdgpu_device *adev)
 	return SOC15_REG_OFFSET(NBIO, 0, mmPCIE_DATA2);
 }
 
-static const struct nbio_hdp_flush_reg nbio_v7_4_hdp_flush_reg = {
+const struct nbio_hdp_flush_reg nbio_v7_4_hdp_flush_reg = {
 	.ref_and_mask_cp0 = GPU_HDP_FLUSH_DONE__CP0_MASK,
 	.ref_and_mask_cp1 = GPU_HDP_FLUSH_DONE__CP1_MASK,
 	.ref_and_mask_cp2 = GPU_HDP_FLUSH_DONE__CP2_MASK,
@@ -306,17 +308,208 @@ static void nbio_v7_4_detect_hw_virt(struct amdgpu_device *adev)
 
 static void nbio_v7_4_init_registers(struct amdgpu_device *adev)
 {
-	uint32_t def, data;
 
-	def = data = RREG32_PCIE(smnPCIE_CI_CNTL);
-	data = REG_SET_FIELD(data, PCIE_CI_CNTL, CI_SLV_ORDERING_DIS, 1);
+}
 
-	if (def != data)
-		WREG32_PCIE(smnPCIE_CI_CNTL, data);
+static void nbio_v7_4_handle_ras_controller_intr_no_bifring(struct amdgpu_device *adev)
+{
+	uint32_t bif_doorbell_intr_cntl;
+
+	bif_doorbell_intr_cntl = RREG32_SOC15(NBIO, 0, mmBIF_DOORBELL_INT_CNTL);
+	if (REG_GET_FIELD(bif_doorbell_intr_cntl,
+		BIF_DOORBELL_INT_CNTL, RAS_CNTLR_INTERRUPT_STATUS)) {
+		/* driver has to clear the interrupt status when bif ring is disabled */
+		bif_doorbell_intr_cntl = REG_SET_FIELD(bif_doorbell_intr_cntl,
+						BIF_DOORBELL_INT_CNTL,
+						RAS_CNTLR_INTERRUPT_CLEAR, 1);
+		WREG32_SOC15(NBIO, 0, mmBIF_DOORBELL_INT_CNTL, bif_doorbell_intr_cntl);
+
+		amdgpu_ras_global_ras_isr(adev);
+	}
+}
+
+static void nbio_v7_4_handle_ras_err_event_athub_intr_no_bifring(struct amdgpu_device *adev)
+{
+	uint32_t bif_doorbell_intr_cntl;
+
+	bif_doorbell_intr_cntl = RREG32_SOC15(NBIO, 0, mmBIF_DOORBELL_INT_CNTL);
+	if (REG_GET_FIELD(bif_doorbell_intr_cntl,
+		BIF_DOORBELL_INT_CNTL, RAS_ATHUB_ERR_EVENT_INTERRUPT_STATUS)) {
+		/* driver has to clear the interrupt status when bif ring is disabled */
+		bif_doorbell_intr_cntl = REG_SET_FIELD(bif_doorbell_intr_cntl,
+						BIF_DOORBELL_INT_CNTL,
+						RAS_ATHUB_ERR_EVENT_INTERRUPT_CLEAR, 1);
+		WREG32_SOC15(NBIO, 0, mmBIF_DOORBELL_INT_CNTL, bif_doorbell_intr_cntl);
+
+		amdgpu_ras_global_ras_isr(adev);
+	}
+}
+
+
+static int nbio_v7_4_set_ras_controller_irq_state(struct amdgpu_device *adev,
+						  struct amdgpu_irq_src *src,
+						  unsigned type,
+						  enum amdgpu_interrupt_state state)
+{
+	/* The ras_controller_irq enablement should be done in psp bl when it
+	 * tries to enable ras feature. Driver only need to set the correct interrupt
+	 * vector for bare-metal and sriov use case respectively
+	 */
+	uint32_t bif_intr_cntl;
+
+	bif_intr_cntl = RREG32_SOC15(NBIO, 0, mmBIF_INTR_CNTL);
+	if (state == AMDGPU_IRQ_STATE_ENABLE) {
+		/* set interrupt vector select bit to 0 to select
+		 * vetcor 1 for bare metal case */
+		bif_intr_cntl = REG_SET_FIELD(bif_intr_cntl,
+					      BIF_INTR_CNTL,
+					      RAS_INTR_VEC_SEL, 0);
+		WREG32_SOC15(NBIO, 0, mmBIF_INTR_CNTL, bif_intr_cntl);
+	}
+
+	return 0;
+}
+
+static int nbio_v7_4_process_ras_controller_irq(struct amdgpu_device *adev,
+						struct amdgpu_irq_src *source,
+						struct amdgpu_iv_entry *entry)
+{
+	/* By design, the ih cookie for ras_controller_irq should be written
+	 * to BIFring instead of general iv ring. However, due to known bif ring
+	 * hw bug, it has to be disabled. There is no chance the process function
+	 * will be involked. Just left it as a dummy one.
+	 */
+	return 0;
+}
+
+static int nbio_v7_4_set_ras_err_event_athub_irq_state(struct amdgpu_device *adev,
+						       struct amdgpu_irq_src *src,
+						       unsigned type,
+						       enum amdgpu_interrupt_state state)
+{
+	/* The ras_controller_irq enablement should be done in psp bl when it
+	 * tries to enable ras feature. Driver only need to set the correct interrupt
+	 * vector for bare-metal and sriov use case respectively
+	 */
+	uint32_t bif_intr_cntl;
+
+	bif_intr_cntl = RREG32_SOC15(NBIO, 0, mmBIF_INTR_CNTL);
+	if (state == AMDGPU_IRQ_STATE_ENABLE) {
+		/* set interrupt vector select bit to 0 to select
+		 * vetcor 1 for bare metal case */
+		bif_intr_cntl = REG_SET_FIELD(bif_intr_cntl,
+					      BIF_INTR_CNTL,
+					      RAS_INTR_VEC_SEL, 0);
+		WREG32_SOC15(NBIO, 0, mmBIF_INTR_CNTL, bif_intr_cntl);
+	}
+
+	return 0;
+}
+
+static int nbio_v7_4_process_err_event_athub_irq(struct amdgpu_device *adev,
+						 struct amdgpu_irq_src *source,
+						 struct amdgpu_iv_entry *entry)
+{
+	/* By design, the ih cookie for err_event_athub_irq should be written
+	 * to BIFring instead of general iv ring. However, due to known bif ring
+	 * hw bug, it has to be disabled. There is no chance the process function
+	 * will be involked. Just left it as a dummy one.
+	 */
+	return 0;
+}
+
+static const struct amdgpu_irq_src_funcs nbio_v7_4_ras_controller_irq_funcs = {
+	.set = nbio_v7_4_set_ras_controller_irq_state,
+	.process = nbio_v7_4_process_ras_controller_irq,
+};
+
+static const struct amdgpu_irq_src_funcs nbio_v7_4_ras_err_event_athub_irq_funcs = {
+	.set = nbio_v7_4_set_ras_err_event_athub_irq_state,
+	.process = nbio_v7_4_process_err_event_athub_irq,
+};
+
+static int nbio_v7_4_init_ras_controller_interrupt (struct amdgpu_device *adev)
+{
+	int r;
+
+	/* init the irq funcs */
+	adev->nbio.ras_controller_irq.funcs =
+		&nbio_v7_4_ras_controller_irq_funcs;
+	adev->nbio.ras_controller_irq.num_types = 1;
+
+	/* register ras controller interrupt */
+	r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_BIF,
+			      NBIF_7_4__SRCID__RAS_CONTROLLER_INTERRUPT,
+			      &adev->nbio.ras_controller_irq);
+	if (r)
+		return r;
+
+	return 0;
+}
+
+static int nbio_v7_4_init_ras_err_event_athub_interrupt (struct amdgpu_device *adev)
+{
+
+	int r;
+
+	/* init the irq funcs */
+	adev->nbio.ras_err_event_athub_irq.funcs =
+		&nbio_v7_4_ras_err_event_athub_irq_funcs;
+	adev->nbio.ras_err_event_athub_irq.num_types = 1;
+
+	/* register ras err event athub interrupt */
+	r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_BIF,
+			      NBIF_7_4__SRCID__ERREVENT_ATHUB_INTERRUPT,
+			      &adev->nbio.ras_err_event_athub_irq);
+	if (r)
+		return r;
+
+	return 0;
+}
+
+static void nbio_v7_4_query_ras_error_count(struct amdgpu_device *adev,
+					void *ras_error_status)
+{
+	uint32_t global_sts, central_sts, int_eoi;
+	uint32_t corr, fatal, non_fatal;
+	struct ras_err_data *err_data = (struct ras_err_data *)ras_error_status;
+
+	global_sts = RREG32_PCIE(smnRAS_GLOBAL_STATUS_LO);
+	corr = REG_GET_FIELD(global_sts, RAS_GLOBAL_STATUS_LO, ParityErrCorr);
+	fatal = REG_GET_FIELD(global_sts, RAS_GLOBAL_STATUS_LO, ParityErrFatal);
+	non_fatal = REG_GET_FIELD(global_sts, RAS_GLOBAL_STATUS_LO,
+				ParityErrNonFatal);
+
+	if (corr)
+		err_data->ce_count++;
+	if (fatal)
+		err_data->ue_count++;
+
+	if (corr || fatal || non_fatal) {
+		central_sts = RREG32_PCIE(smnBIFL_RAS_CENTRAL_STATUS);
+		/* clear error status register */
+		WREG32_PCIE(smnRAS_GLOBAL_STATUS_LO, global_sts);
+
+		if (REG_GET_FIELD(central_sts, BIFL_RAS_CENTRAL_STATUS,
+				BIFL_RasContller_Intr_Recv)) {
+			/* clear interrupt status register */
+			WREG32_PCIE(smnBIFL_RAS_CENTRAL_STATUS, central_sts);
+			int_eoi = RREG32_PCIE(smnIOHC_INTERRUPT_EOI);
+			int_eoi = REG_SET_FIELD(int_eoi,
+					IOHC_INTERRUPT_EOI, SMI_EOI, 1);
+			WREG32_PCIE(smnIOHC_INTERRUPT_EOI, int_eoi);
+		}
+	}
+}
+
+static void nbio_v7_4_enable_doorbell_interrupt(struct amdgpu_device *adev,
+						bool enable)
+{
+	WREG32_FIELD15(NBIO, 0, BIF_DOORBELL_INT_CNTL,
+		       DOORBELL_INTERRUPT_DISABLE, enable ? 0 : 1);
 }
 
 const struct amdgpu_nbio_funcs nbio_v7_4_funcs = {
-	.hdp_flush_reg = &nbio_v7_4_hdp_flush_reg,
 	.get_hdp_flush_req_offset = nbio_v7_4_get_hdp_flush_req_offset,
 	.get_hdp_flush_done_offset = nbio_v7_4_get_hdp_flush_done_offset,
 	.get_pcie_index_offset = nbio_v7_4_get_pcie_index_offset,
@@ -330,6 +523,7 @@ const struct amdgpu_nbio_funcs nbio_v7_4_funcs = {
 	.enable_doorbell_aperture = nbio_v7_4_enable_doorbell_aperture,
 	.enable_doorbell_selfring_aperture = nbio_v7_4_enable_doorbell_selfring_aperture,
 	.ih_doorbell_range = nbio_v7_4_ih_doorbell_range,
+	.enable_doorbell_interrupt = nbio_v7_4_enable_doorbell_interrupt,
 	.update_medium_grain_clock_gating = nbio_v7_4_update_medium_grain_clock_gating,
 	.update_medium_grain_light_sleep = nbio_v7_4_update_medium_grain_light_sleep,
 	.get_clockgating_state = nbio_v7_4_get_clockgating_state,
@@ -337,4 +531,10 @@ const struct amdgpu_nbio_funcs nbio_v7_4_funcs = {
 	.init_registers = nbio_v7_4_init_registers,
 	.detect_hw_virt = nbio_v7_4_detect_hw_virt,
 	.remap_hdp_registers = nbio_v7_4_remap_hdp_registers,
+	.handle_ras_controller_intr_no_bifring = nbio_v7_4_handle_ras_controller_intr_no_bifring,
+	.handle_ras_err_event_athub_intr_no_bifring = nbio_v7_4_handle_ras_err_event_athub_intr_no_bifring,
+	.init_ras_controller_interrupt = nbio_v7_4_init_ras_controller_interrupt,
+	.init_ras_err_event_athub_interrupt = nbio_v7_4_init_ras_err_event_athub_interrupt,
+	.query_ras_error_count = nbio_v7_4_query_ras_error_count,
+	.ras_late_init = amdgpu_nbio_ras_late_init,
 };
diff --git a/drivers/gpu/drm/amd/amdgpu/nbio_v7_4.h b/drivers/gpu/drm/amd/amdgpu/nbio_v7_4.h
index c442865bac4f..b1ac82872752 100644
--- a/drivers/gpu/drm/amd/amdgpu/nbio_v7_4.h
+++ b/drivers/gpu/drm/amd/amdgpu/nbio_v7_4.h
@@ -26,6 +26,7 @@
 
 #include "soc15_common.h"
 
+extern const struct nbio_hdp_flush_reg nbio_v7_4_hdp_flush_reg;
 extern const struct amdgpu_nbio_funcs nbio_v7_4_funcs;
 
 #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/nv.c b/drivers/gpu/drm/amd/amdgpu/nv.c
index 85393a99a848..0ba66bef5746 100644
--- a/drivers/gpu/drm/amd/amdgpu/nv.c
+++ b/drivers/gpu/drm/amd/amdgpu/nv.c
@@ -40,12 +40,14 @@
 #include "gc/gc_10_1_0_sh_mask.h"
 #include "hdp/hdp_5_0_0_offset.h"
 #include "hdp/hdp_5_0_0_sh_mask.h"
+#include "smuio/smuio_11_0_0_offset.h"
 
 #include "soc15.h"
 #include "soc15_common.h"
 #include "gmc_v10_0.h"
 #include "gfxhub_v2_0.h"
 #include "mmhub_v2_0.h"
+#include "nbio_v2_3.h"
 #include "nv.h"
 #include "navi10_ih.h"
 #include "gfx_v10_0.h"
@@ -53,6 +55,7 @@
 #include "vcn_v2_0.h"
 #include "dce_virtual.h"
 #include "mes_v10_1.h"
+#include "mxgpu_nv.h"
 
 static const struct amd_ip_funcs nv_common_ip_funcs;
 
@@ -63,8 +66,8 @@ static u32 nv_pcie_rreg(struct amdgpu_device *adev, u32 reg)
 {
 	unsigned long flags, address, data;
 	u32 r;
-	address = adev->nbio_funcs->get_pcie_index_offset(adev);
-	data = adev->nbio_funcs->get_pcie_data_offset(adev);
+	address = adev->nbio.funcs->get_pcie_index_offset(adev);
+	data = adev->nbio.funcs->get_pcie_data_offset(adev);
 
 	spin_lock_irqsave(&adev->pcie_idx_lock, flags);
 	WREG32(address, reg);
@@ -78,8 +81,8 @@ static void nv_pcie_wreg(struct amdgpu_device *adev, u32 reg, u32 v)
 {
 	unsigned long flags, address, data;
 
-	address = adev->nbio_funcs->get_pcie_index_offset(adev);
-	data = adev->nbio_funcs->get_pcie_data_offset(adev);
+	address = adev->nbio.funcs->get_pcie_index_offset(adev);
+	data = adev->nbio.funcs->get_pcie_data_offset(adev);
 
 	spin_lock_irqsave(&adev->pcie_idx_lock, flags);
 	WREG32(address, reg);
@@ -119,7 +122,7 @@ static void nv_didt_wreg(struct amdgpu_device *adev, u32 reg, u32 v)
 
 static u32 nv_get_config_memsize(struct amdgpu_device *adev)
 {
-	return adev->nbio_funcs->get_memsize(adev);
+	return adev->nbio.funcs->get_memsize(adev);
 }
 
 static u32 nv_get_xclk(struct amdgpu_device *adev)
@@ -154,8 +157,27 @@ static bool nv_read_disabled_bios(struct amdgpu_device *adev)
 static bool nv_read_bios_from_rom(struct amdgpu_device *adev,
 				  u8 *bios, u32 length_bytes)
 {
-	/* TODO: will implement it when SMU header is available */
-	return false;
+	u32 *dw_ptr;
+	u32 i, length_dw;
+
+	if (bios == NULL)
+		return false;
+	if (length_bytes == 0)
+		return false;
+	/* APU vbios image is part of sbios image */
+	if (adev->flags & AMD_IS_APU)
+		return false;
+
+	dw_ptr = (u32 *)bios;
+	length_dw = ALIGN(length_bytes, 4) / 4;
+
+	/* set rom index to 0 */
+	WREG32(SOC15_REG_OFFSET(SMUIO, 0, mmROM_INDEX), 0);
+	/* read out the rom data */
+	for (i = 0; i < length_dw; i++)
+		dw_ptr[i] = RREG32(SOC15_REG_OFFSET(SMUIO, 0, mmROM_DATA));
+
+	return true;
 }
 
 static struct soc15_allowed_register_entry nv_allowed_read_registers[] = {
@@ -176,6 +198,7 @@ static struct soc15_allowed_register_entry nv_allowed_read_registers[] = {
 	{ SOC15_REG_ENTRY(GC, 0, mmCP_CPF_BUSY_STAT)},
 	{ SOC15_REG_ENTRY(GC, 0, mmCP_CPF_STALLED_STAT1)},
 	{ SOC15_REG_ENTRY(GC, 0, mmCP_CPF_STATUS)},
+	{ SOC15_REG_ENTRY(GC, 0, mmCP_CPC_BUSY_STAT)},
 	{ SOC15_REG_ENTRY(GC, 0, mmCP_CPC_STALLED_STAT1)},
 	{ SOC15_REG_ENTRY(GC, 0, mmCP_CPC_STATUS)},
 	{ SOC15_REG_ENTRY(GC, 0, mmGB_ADDR_CONFIG)},
@@ -279,7 +302,7 @@ static int nv_asic_mode1_reset(struct amdgpu_device *adev)
 
 	/* wait for asic to come out of reset */
 	for (i = 0; i < adev->usec_timeout; i++) {
-		u32 memsize = adev->nbio_funcs->get_memsize(adev);
+		u32 memsize = adev->nbio.funcs->get_memsize(adev);
 
 		if (memsize != 0xffffffff)
 			break;
@@ -296,7 +319,7 @@ nv_asic_reset_method(struct amdgpu_device *adev)
 {
 	struct smu_context *smu = &adev->smu;
 
-	if (smu_baco_is_support(smu))
+	if (!amdgpu_sriov_vf(adev) && smu_baco_is_support(smu))
 		return AMD_RESET_METHOD_BACO;
 	else
 		return AMD_RESET_METHOD_MODE1;
@@ -317,10 +340,12 @@ static int nv_asic_reset(struct amdgpu_device *adev)
 	struct smu_context *smu = &adev->smu;
 
 	if (nv_asic_reset_method(adev) == AMD_RESET_METHOD_BACO) {
-		amdgpu_inc_vram_lost(adev);
+		if (!adev->in_suspend)
+			amdgpu_inc_vram_lost(adev);
 		ret = smu_baco_reset(smu);
 	} else {
-		amdgpu_inc_vram_lost(adev);
+		if (!adev->in_suspend)
+			amdgpu_inc_vram_lost(adev);
 		ret = nv_asic_mode1_reset(adev);
 	}
 
@@ -366,8 +391,8 @@ static void nv_program_aspm(struct amdgpu_device *adev)
 static void nv_enable_doorbell_aperture(struct amdgpu_device *adev,
 					bool enable)
 {
-	adev->nbio_funcs->enable_doorbell_aperture(adev, enable);
-	adev->nbio_funcs->enable_doorbell_selfring_aperture(adev, enable);
+	adev->nbio.funcs->enable_doorbell_aperture(adev, enable);
+	adev->nbio.funcs->enable_doorbell_selfring_aperture(adev, enable);
 }
 
 static const struct amdgpu_ip_block_version nv_common_ip_block =
@@ -421,9 +446,13 @@ int nv_set_ip_blocks(struct amdgpu_device *adev)
 	if (r)
 		return r;
 
-	adev->nbio_funcs = &nbio_v2_3_funcs;
+	adev->nbio.funcs = &nbio_v2_3_funcs;
+	adev->nbio.hdp_flush_reg = &nbio_v2_3_hdp_flush_reg;
 
-	adev->nbio_funcs->detect_hw_virt(adev);
+	adev->nbio.funcs->detect_hw_virt(adev);
+
+	if (amdgpu_sriov_vf(adev))
+		adev->virt.ops = &xgpu_nv_virt_ops;
 
 	switch (adev->asic_type) {
 	case CHIP_NAVI10:
@@ -433,7 +462,7 @@ int nv_set_ip_blocks(struct amdgpu_device *adev)
 		amdgpu_device_ip_block_add(adev, &navi10_ih_ip_block);
 		amdgpu_device_ip_block_add(adev, &psp_v11_0_ip_block);
 		if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP &&
-		    is_support_sw_smu(adev))
+		    is_support_sw_smu(adev) && !amdgpu_sriov_vf(adev))
 			amdgpu_device_ip_block_add(adev, &smu_v11_0_ip_block);
 		if (adev->enable_virtual_display || amdgpu_sriov_vf(adev))
 			amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block);
@@ -444,7 +473,7 @@ int nv_set_ip_blocks(struct amdgpu_device *adev)
 		amdgpu_device_ip_block_add(adev, &gfx_v10_0_ip_block);
 		amdgpu_device_ip_block_add(adev, &sdma_v5_0_ip_block);
 		if (adev->firmware.load_type == AMDGPU_FW_LOAD_DIRECT &&
-		    is_support_sw_smu(adev))
+		    is_support_sw_smu(adev) && !amdgpu_sriov_vf(adev))
 			amdgpu_device_ip_block_add(adev, &smu_v11_0_ip_block);
 		amdgpu_device_ip_block_add(adev, &vcn_v2_0_ip_block);
 		if (adev->enable_mes)
@@ -456,7 +485,7 @@ int nv_set_ip_blocks(struct amdgpu_device *adev)
 		amdgpu_device_ip_block_add(adev, &navi10_ih_ip_block);
 		amdgpu_device_ip_block_add(adev, &psp_v11_0_ip_block);
 		if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP &&
-		    is_support_sw_smu(adev))
+		    is_support_sw_smu(adev) && !amdgpu_sriov_vf(adev))
 			amdgpu_device_ip_block_add(adev, &smu_v11_0_ip_block);
 		if (adev->enable_virtual_display || amdgpu_sriov_vf(adev))
 			amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block);
@@ -467,7 +496,7 @@ int nv_set_ip_blocks(struct amdgpu_device *adev)
 		amdgpu_device_ip_block_add(adev, &gfx_v10_0_ip_block);
 		amdgpu_device_ip_block_add(adev, &sdma_v5_0_ip_block);
 		if (adev->firmware.load_type == AMDGPU_FW_LOAD_DIRECT &&
-		    is_support_sw_smu(adev))
+		    is_support_sw_smu(adev) && !amdgpu_sriov_vf(adev))
 			amdgpu_device_ip_block_add(adev, &smu_v11_0_ip_block);
 		amdgpu_device_ip_block_add(adev, &vcn_v2_0_ip_block);
 		break;
@@ -480,12 +509,12 @@ int nv_set_ip_blocks(struct amdgpu_device *adev)
 
 static uint32_t nv_get_rev_id(struct amdgpu_device *adev)
 {
-	return adev->nbio_funcs->get_rev_id(adev);
+	return adev->nbio.funcs->get_rev_id(adev);
 }
 
 static void nv_flush_hdp(struct amdgpu_device *adev, struct amdgpu_ring *ring)
 {
-	adev->nbio_funcs->hdp_flush(adev, ring);
+	adev->nbio.funcs->hdp_flush(adev, ring);
 }
 
 static void nv_invalidate_hdp(struct amdgpu_device *adev,
@@ -530,6 +559,16 @@ static bool nv_need_reset_on_init(struct amdgpu_device *adev)
 	return false;
 }
 
+static uint64_t nv_get_pcie_replay_count(struct amdgpu_device *adev)
+{
+
+	/* TODO
+	 * dummy implement for pcie_replay_count sysfs interface
+	 * */
+
+	return 0;
+}
+
 static void nv_init_doorbell_index(struct amdgpu_device *adev)
 {
 	adev->doorbell_index.kiq = AMDGPU_NAVI10_DOORBELL_KIQ;
@@ -577,12 +616,16 @@ static const struct amdgpu_asic_funcs nv_asic_funcs =
 	.need_full_reset = &nv_need_full_reset,
 	.get_pcie_usage = &nv_get_pcie_usage,
 	.need_reset_on_init = &nv_need_reset_on_init,
+	.get_pcie_replay_count = &nv_get_pcie_replay_count,
 };
 
 static int nv_common_early_init(void *handle)
 {
+#define MMIO_REG_HOLE_OFFSET (0x80000 - PAGE_SIZE)
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
+	adev->rmmio_remap.reg_offset = MMIO_REG_HOLE_OFFSET;
+	adev->rmmio_remap.bus_addr = adev->rmmio_base + MMIO_REG_HOLE_OFFSET;
 	adev->smc_rreg = NULL;
 	adev->smc_wreg = NULL;
 	adev->pcie_rreg = &nv_pcie_rreg;
@@ -665,16 +708,31 @@ static int nv_common_early_init(void *handle)
 		return -EINVAL;
 	}
 
+	if (amdgpu_sriov_vf(adev)) {
+		amdgpu_virt_init_setting(adev);
+		xgpu_nv_mailbox_set_irq_funcs(adev);
+	}
+
 	return 0;
 }
 
 static int nv_common_late_init(void *handle)
 {
+	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+	if (amdgpu_sriov_vf(adev))
+		xgpu_nv_mailbox_get_irq(adev);
+
 	return 0;
 }
 
 static int nv_common_sw_init(void *handle)
 {
+	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+	if (amdgpu_sriov_vf(adev))
+		xgpu_nv_mailbox_add_irq_id(adev);
+
 	return 0;
 }
 
@@ -692,7 +750,13 @@ static int nv_common_hw_init(void *handle)
 	/* enable aspm */
 	nv_program_aspm(adev);
 	/* setup nbio registers */
-	adev->nbio_funcs->init_registers(adev);
+	adev->nbio.funcs->init_registers(adev);
+	/* remap HDP registers to a hole in mmio space,
+	 * for the purpose of expose those registers
+	 * to process space
+	 */
+	if (adev->nbio.funcs->remap_hdp_registers)
+		adev->nbio.funcs->remap_hdp_registers(adev);
 	/* enable the doorbell aperture */
 	nv_enable_doorbell_aperture(adev, true);
 
@@ -854,9 +918,9 @@ static int nv_common_set_clockgating_state(void *handle,
 	case CHIP_NAVI10:
 	case CHIP_NAVI14:
 	case CHIP_NAVI12:
-		adev->nbio_funcs->update_medium_grain_clock_gating(adev,
+		adev->nbio.funcs->update_medium_grain_clock_gating(adev,
 				state == AMD_CG_STATE_GATE ? true : false);
-		adev->nbio_funcs->update_medium_grain_light_sleep(adev,
+		adev->nbio.funcs->update_medium_grain_light_sleep(adev,
 				state == AMD_CG_STATE_GATE ? true : false);
 		nv_update_hdp_mem_power_gating(adev,
 				   state == AMD_CG_STATE_GATE ? true : false);
@@ -884,7 +948,7 @@ static void nv_common_get_clockgating_state(void *handle, u32 *flags)
 	if (amdgpu_sriov_vf(adev))
 		*flags = 0;
 
-	adev->nbio_funcs->get_clockgating_state(adev, flags);
+	adev->nbio.funcs->get_clockgating_state(adev, flags);
 
 	/* AMD_CG_SUPPORT_HDP_MGCG */
 	tmp = RREG32_SOC15(HDP, 0, mmHDP_CLK_CNTL);
diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c b/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c
index 5d95e614369a..b345e69ba246 100644
--- a/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c
@@ -40,6 +40,9 @@
 MODULE_FIRMWARE("amdgpu/raven_asd.bin");
 MODULE_FIRMWARE("amdgpu/picasso_asd.bin");
 MODULE_FIRMWARE("amdgpu/raven2_asd.bin");
+MODULE_FIRMWARE("amdgpu/picasso_ta.bin");
+MODULE_FIRMWARE("amdgpu/raven2_ta.bin");
+MODULE_FIRMWARE("amdgpu/raven_ta.bin");
 
 static int psp_v10_0_init_microcode(struct psp_context *psp)
 {
@@ -48,7 +51,7 @@ static int psp_v10_0_init_microcode(struct psp_context *psp)
 	char fw_name[30];
 	int err = 0;
 	const struct psp_firmware_header_v1_0 *hdr;
-
+	const struct ta_firmware_header_v1_0 *ta_hdr;
 	DRM_DEBUG("\n");
 
 	switch (adev->asic_type) {
@@ -79,7 +82,45 @@ static int psp_v10_0_init_microcode(struct psp_context *psp)
 	adev->psp.asd_start_addr = (uint8_t *)hdr +
 				le32_to_cpu(hdr->header.ucode_array_offset_bytes);
 
+	snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_ta.bin", chip_name);
+	err = request_firmware(&adev->psp.ta_fw, fw_name, adev->dev);
+	if (err) {
+		release_firmware(adev->psp.ta_fw);
+		adev->psp.ta_fw = NULL;
+		dev_info(adev->dev,
+			 "psp v10.0: Failed to load firmware \"%s\"\n",
+			 fw_name);
+	} else {
+		err = amdgpu_ucode_validate(adev->psp.ta_fw);
+		if (err)
+			goto out2;
+
+		ta_hdr = (const struct ta_firmware_header_v1_0 *)
+				 adev->psp.ta_fw->data;
+		adev->psp.ta_hdcp_ucode_version =
+			le32_to_cpu(ta_hdr->ta_hdcp_ucode_version);
+		adev->psp.ta_hdcp_ucode_size =
+			le32_to_cpu(ta_hdr->ta_hdcp_size_bytes);
+		adev->psp.ta_hdcp_start_addr =
+			(uint8_t *)ta_hdr +
+			le32_to_cpu(ta_hdr->header.ucode_array_offset_bytes);
+
+		adev->psp.ta_fw_version = le32_to_cpu(ta_hdr->header.ucode_version);
+
+		adev->psp.ta_dtm_ucode_version =
+			le32_to_cpu(ta_hdr->ta_dtm_ucode_version);
+		adev->psp.ta_dtm_ucode_size =
+			le32_to_cpu(ta_hdr->ta_dtm_size_bytes);
+		adev->psp.ta_dtm_start_addr =
+			(uint8_t *)adev->psp.ta_hdcp_start_addr +
+			le32_to_cpu(ta_hdr->ta_dtm_offset_bytes);
+	}
+
 	return 0;
+
+out2:
+	release_firmware(adev->psp.ta_fw);
+	adev->psp.ta_fw = NULL;
 out:
 	if (err) {
 		dev_err(adev->dev,
@@ -228,6 +269,7 @@ static int psp_v10_0_cmd_submit(struct psp_context *psp,
 	write_frame->fence_addr_hi = upper_32_bits(fence_mc_addr);
 	write_frame->fence_addr_lo = lower_32_bits(fence_mc_addr);
 	write_frame->fence_value = index;
+	amdgpu_asic_flush_hdp(adev, NULL);
 
 	/* Update the write Pointer in DWORDs */
 	psp_write_ptr_reg = (psp_write_ptr_reg + rb_frame_size_dw) % ring_size_dw;
diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v11_0.c b/drivers/gpu/drm/amd/amdgpu/psp_v11_0.c
index 10166104b8a3..ffeaa2f5588d 100644
--- a/drivers/gpu/drm/amd/amdgpu/psp_v11_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/psp_v11_0.c
@@ -49,6 +49,7 @@ MODULE_FIRMWARE("amdgpu/navi12_sos.bin");
 MODULE_FIRMWARE("amdgpu/navi12_asd.bin");
 MODULE_FIRMWARE("amdgpu/arcturus_sos.bin");
 MODULE_FIRMWARE("amdgpu/arcturus_asd.bin");
+MODULE_FIRMWARE("amdgpu/arcturus_ta.bin");
 
 /* address block */
 #define smnMP1_FIRMWARE_FLAGS		0x3010024
@@ -57,6 +58,8 @@ MODULE_FIRMWARE("amdgpu/arcturus_asd.bin");
 #define mmRLC_GPM_UCODE_DATA_NV10	0x5b62
 #define mmSDMA0_UCODE_ADDR_NV10		0x5880
 #define mmSDMA0_UCODE_DATA_NV10		0x5881
+/* memory training timeout define */
+#define MEM_TRAIN_SEND_MSG_TIMEOUT_US	3000000
 
 static int psp_v11_0_init_microcode(struct psp_context *psp)
 {
@@ -155,6 +158,7 @@ static int psp_v11_0_init_microcode(struct psp_context *psp)
 
 	switch (adev->asic_type) {
 	case CHIP_VEGA20:
+	case CHIP_ARCTURUS:
 		snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_ta.bin", chip_name);
 		err = request_firmware(&adev->psp.ta_fw, fw_name, adev->dev);
 		if (err) {
@@ -182,7 +186,6 @@ static int psp_v11_0_init_microcode(struct psp_context *psp)
 	case CHIP_NAVI10:
 	case CHIP_NAVI14:
 	case CHIP_NAVI12:
-	case CHIP_ARCTURUS:
 		break;
 	default:
 		BUG();
@@ -205,18 +208,26 @@ out:
 	return err;
 }
 
+static bool psp_v11_0_is_sos_alive(struct psp_context *psp)
+{
+	struct amdgpu_device *adev = psp->adev;
+	uint32_t sol_reg;
+
+	sol_reg = RREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_81);
+
+	return sol_reg != 0x0;
+}
+
 static int psp_v11_0_bootloader_load_kdb(struct psp_context *psp)
 {
 	int ret;
 	uint32_t psp_gfxdrv_command_reg = 0;
 	struct amdgpu_device *adev = psp->adev;
-	uint32_t sol_reg;
 
 	/* Check tOS sign of life register to confirm sys driver and sOS
 	 * are already been loaded.
 	 */
-	sol_reg = RREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_81);
-	if (sol_reg) {
+	if (psp_v11_0_is_sos_alive(psp)) {
 		psp->sos_fw_version = RREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_58);
 		dev_info(adev->dev, "sos fw version = 0x%x.\n", psp->sos_fw_version);
 		return 0;
@@ -233,7 +244,7 @@ static int psp_v11_0_bootloader_load_kdb(struct psp_context *psp)
 	/* Copy PSP KDB binary to memory */
 	memcpy(psp->fw_pri_buf, psp->kdb_start_addr, psp->kdb_bin_size);
 
-	/* Provide the sys driver to bootloader */
+	/* Provide the PSP KDB to bootloader */
 	WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_36,
 	       (uint32_t)(psp->fw_pri_mc_addr >> 20));
 	psp_gfxdrv_command_reg = PSP_BL__LOAD_KEY_DATABASE;
@@ -252,13 +263,11 @@ static int psp_v11_0_bootloader_load_sysdrv(struct psp_context *psp)
 	int ret;
 	uint32_t psp_gfxdrv_command_reg = 0;
 	struct amdgpu_device *adev = psp->adev;
-	uint32_t sol_reg;
 
 	/* Check sOS sign of life register to confirm sys driver and sOS
 	 * are already been loaded.
 	 */
-	sol_reg = RREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_81);
-	if (sol_reg) {
+	if (psp_v11_0_is_sos_alive(psp)) {
 		psp->sos_fw_version = RREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_58);
 		dev_info(adev->dev, "sos fw version = 0x%x.\n", psp->sos_fw_version);
 		return 0;
@@ -296,13 +305,11 @@ static int psp_v11_0_bootloader_load_sos(struct psp_context *psp)
 	int ret;
 	unsigned int psp_gfxdrv_command_reg = 0;
 	struct amdgpu_device *adev = psp->adev;
-	uint32_t sol_reg;
 
 	/* Check sOS sign of life register to confirm sys driver and sOS
 	 * are already been loaded.
 	 */
-	sol_reg = RREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_81);
-	if (sol_reg)
+	if (psp_v11_0_is_sos_alive(psp))
 		return 0;
 
 	/* Wait for bootloader to signify that is ready having bit 31 of C2PMSG_35 set to 1 */
@@ -398,6 +405,34 @@ static bool psp_v11_0_support_vmr_ring(struct psp_context *psp)
 	return false;
 }
 
+static int psp_v11_0_ring_stop(struct psp_context *psp,
+			      enum psp_ring_type ring_type)
+{
+	int ret = 0;
+	struct amdgpu_device *adev = psp->adev;
+
+	/* Write the ring destroy command*/
+	if (psp_v11_0_support_vmr_ring(psp))
+		WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_101,
+				     GFX_CTRL_CMD_ID_DESTROY_GPCOM_RING);
+	else
+		WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_64,
+				     GFX_CTRL_CMD_ID_DESTROY_RINGS);
+
+	/* there might be handshake issue with hardware which needs delay */
+	mdelay(20);
+
+	/* Wait for response flag (bit 31) */
+	if (psp_v11_0_support_vmr_ring(psp))
+		ret = psp_wait_for(psp, SOC15_REG_OFFSET(MP0, 0, mmMP0_SMN_C2PMSG_101),
+				   0x80000000, 0x80000000, false);
+	else
+		ret = psp_wait_for(psp, SOC15_REG_OFFSET(MP0, 0, mmMP0_SMN_C2PMSG_64),
+				   0x80000000, 0x80000000, false);
+
+	return ret;
+}
+
 static int psp_v11_0_ring_create(struct psp_context *psp,
 				enum psp_ring_type ring_type)
 {
@@ -407,6 +442,12 @@ static int psp_v11_0_ring_create(struct psp_context *psp,
 	struct amdgpu_device *adev = psp->adev;
 
 	if (psp_v11_0_support_vmr_ring(psp)) {
+		ret = psp_v11_0_ring_stop(psp, ring_type);
+		if (ret) {
+			DRM_ERROR("psp_v11_0_ring_stop_sriov failed!\n");
+			return ret;
+		}
+
 		/* Write low address of the ring to C2PMSG_102 */
 		psp_ring_reg = lower_32_bits(ring->ring_mem_mc_addr);
 		WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_102, psp_ring_reg);
@@ -426,6 +467,14 @@ static int psp_v11_0_ring_create(struct psp_context *psp,
 				   0x80000000, 0x8000FFFF, false);
 
 	} else {
+		/* Wait for sOS ready for ring creation */
+		ret = psp_wait_for(psp, SOC15_REG_OFFSET(MP0, 0, mmMP0_SMN_C2PMSG_64),
+				   0x80000000, 0x80000000, false);
+		if (ret) {
+			DRM_ERROR("Failed to wait for sOS ready for ring creation\n");
+			return ret;
+		}
+
 		/* Write low address of the ring to C2PMSG_69 */
 		psp_ring_reg = lower_32_bits(ring->ring_mem_mc_addr);
 		WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_69, psp_ring_reg);
@@ -451,33 +500,6 @@ static int psp_v11_0_ring_create(struct psp_context *psp,
 	return ret;
 }
 
-static int psp_v11_0_ring_stop(struct psp_context *psp,
-			      enum psp_ring_type ring_type)
-{
-	int ret = 0;
-	struct amdgpu_device *adev = psp->adev;
-
-	/* Write the ring destroy command*/
-	if (psp_v11_0_support_vmr_ring(psp))
-		WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_101,
-				     GFX_CTRL_CMD_ID_DESTROY_GPCOM_RING);
-	else
-		WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_64,
-				     GFX_CTRL_CMD_ID_DESTROY_RINGS);
-
-	/* there might be handshake issue with hardware which needs delay */
-	mdelay(20);
-
-	/* Wait for response flag (bit 31) */
-	if (psp_v11_0_support_vmr_ring(psp))
-		ret = psp_wait_for(psp, SOC15_REG_OFFSET(MP0, 0, mmMP0_SMN_C2PMSG_101),
-				   0x80000000, 0x80000000, false);
-	else
-		ret = psp_wait_for(psp, SOC15_REG_OFFSET(MP0, 0, mmMP0_SMN_C2PMSG_64),
-				   0x80000000, 0x80000000, false);
-
-	return ret;
-}
 
 static int psp_v11_0_ring_destroy(struct psp_context *psp,
 				 enum psp_ring_type ring_type)
@@ -541,6 +563,7 @@ static int psp_v11_0_cmd_submit(struct psp_context *psp,
 	write_frame->fence_addr_hi = upper_32_bits(fence_mc_addr);
 	write_frame->fence_addr_lo = lower_32_bits(fence_mc_addr);
 	write_frame->fence_value = index;
+	amdgpu_asic_flush_hdp(adev, NULL);
 
 	/* Update the write Pointer in DWORDs */
 	psp_write_ptr_reg = (psp_write_ptr_reg + rb_frame_size_dw) % ring_size_dw;
@@ -889,6 +912,162 @@ static int psp_v11_0_rlc_autoload_start(struct psp_context *psp)
 	return psp_rlc_autoload_start(psp);
 }
 
+static int psp_v11_0_memory_training_send_msg(struct psp_context *psp, int msg)
+{
+	int ret;
+	int i;
+	uint32_t data_32;
+	int max_wait;
+	struct amdgpu_device *adev = psp->adev;
+
+	data_32 = (psp->mem_train_ctx.c2p_train_data_offset >> 20);
+	WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_36, data_32);
+	WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_35, msg);
+
+	max_wait = MEM_TRAIN_SEND_MSG_TIMEOUT_US / adev->usec_timeout;
+	for (i = 0; i < max_wait; i++) {
+		ret = psp_wait_for(psp, SOC15_REG_OFFSET(MP0, 0, mmMP0_SMN_C2PMSG_35),
+				   0x80000000, 0x80000000, false);
+		if (ret == 0)
+			break;
+	}
+	if (i < max_wait)
+		ret = 0;
+	else
+		ret = -ETIME;
+
+	DRM_DEBUG("training %s %s, cost %d @ %d ms\n",
+		  (msg == PSP_BL__DRAM_SHORT_TRAIN) ? "short" : "long",
+		  (ret == 0) ? "succeed" : "failed",
+		  i, adev->usec_timeout/1000);
+	return ret;
+}
+
+static void psp_v11_0_memory_training_fini(struct psp_context *psp)
+{
+	struct psp_memory_training_context *ctx = &psp->mem_train_ctx;
+
+	ctx->init = PSP_MEM_TRAIN_NOT_SUPPORT;
+	kfree(ctx->sys_cache);
+	ctx->sys_cache = NULL;
+}
+
+static int psp_v11_0_memory_training_init(struct psp_context *psp)
+{
+	int ret;
+	struct psp_memory_training_context *ctx = &psp->mem_train_ctx;
+
+	if (ctx->init != PSP_MEM_TRAIN_RESERVE_SUCCESS) {
+		DRM_DEBUG("memory training is not supported!\n");
+		return 0;
+	}
+
+	ctx->sys_cache = kzalloc(ctx->train_data_size, GFP_KERNEL);
+	if (ctx->sys_cache == NULL) {
+		DRM_ERROR("alloc mem_train_ctx.sys_cache failed!\n");
+		ret = -ENOMEM;
+		goto Err_out;
+	}
+
+	DRM_DEBUG("train_data_size:%llx,p2c_train_data_offset:%llx,c2p_train_data_offset:%llx.\n",
+		  ctx->train_data_size,
+		  ctx->p2c_train_data_offset,
+		  ctx->c2p_train_data_offset);
+	ctx->init = PSP_MEM_TRAIN_INIT_SUCCESS;
+	return 0;
+
+Err_out:
+	psp_v11_0_memory_training_fini(psp);
+	return ret;
+}
+
+/*
+ * save and restore proces
+ */
+static int psp_v11_0_memory_training(struct psp_context *psp, uint32_t ops)
+{
+	int ret;
+	uint32_t p2c_header[4];
+	struct psp_memory_training_context *ctx = &psp->mem_train_ctx;
+	uint32_t *pcache = (uint32_t*)ctx->sys_cache;
+
+	if (ctx->init == PSP_MEM_TRAIN_NOT_SUPPORT) {
+		DRM_DEBUG("Memory training is not supported.\n");
+		return 0;
+	} else if (ctx->init != PSP_MEM_TRAIN_INIT_SUCCESS) {
+		DRM_ERROR("Memory training initialization failure.\n");
+		return -EINVAL;
+	}
+
+	if (psp_v11_0_is_sos_alive(psp)) {
+		DRM_DEBUG("SOS is alive, skip memory training.\n");
+		return 0;
+	}
+
+	amdgpu_device_vram_access(psp->adev, ctx->p2c_train_data_offset, p2c_header, sizeof(p2c_header), false);
+	DRM_DEBUG("sys_cache[%08x,%08x,%08x,%08x] p2c_header[%08x,%08x,%08x,%08x]\n",
+		  pcache[0], pcache[1], pcache[2], pcache[3],
+		  p2c_header[0], p2c_header[1], p2c_header[2], p2c_header[3]);
+
+	if (ops & PSP_MEM_TRAIN_SEND_SHORT_MSG) {
+		DRM_DEBUG("Short training depends on restore.\n");
+		ops |= PSP_MEM_TRAIN_RESTORE;
+	}
+
+	if ((ops & PSP_MEM_TRAIN_RESTORE) &&
+	    pcache[0] != MEM_TRAIN_SYSTEM_SIGNATURE) {
+		DRM_DEBUG("sys_cache[0] is invalid, restore depends on save.\n");
+		ops |= PSP_MEM_TRAIN_SAVE;
+	}
+
+	if (p2c_header[0] == MEM_TRAIN_SYSTEM_SIGNATURE &&
+	    !(pcache[0] == MEM_TRAIN_SYSTEM_SIGNATURE &&
+	      pcache[3] == p2c_header[3])) {
+		DRM_DEBUG("sys_cache is invalid or out-of-date, need save training data to sys_cache.\n");
+		ops |= PSP_MEM_TRAIN_SAVE;
+	}
+
+	if ((ops & PSP_MEM_TRAIN_SAVE) &&
+	    p2c_header[0] != MEM_TRAIN_SYSTEM_SIGNATURE) {
+		DRM_DEBUG("p2c_header[0] is invalid, save depends on long training.\n");
+		ops |= PSP_MEM_TRAIN_SEND_LONG_MSG;
+	}
+
+	if (ops & PSP_MEM_TRAIN_SEND_LONG_MSG) {
+		ops &= ~PSP_MEM_TRAIN_SEND_SHORT_MSG;
+		ops |= PSP_MEM_TRAIN_SAVE;
+	}
+
+	DRM_DEBUG("Memory training ops:%x.\n", ops);
+
+	if (ops & PSP_MEM_TRAIN_SEND_LONG_MSG) {
+		ret = psp_v11_0_memory_training_send_msg(psp, PSP_BL__DRAM_LONG_TRAIN);
+		if (ret) {
+			DRM_ERROR("Send long training msg failed.\n");
+			return ret;
+		}
+	}
+
+	if (ops & PSP_MEM_TRAIN_SAVE) {
+		amdgpu_device_vram_access(psp->adev, ctx->p2c_train_data_offset, ctx->sys_cache, ctx->train_data_size, false);
+	}
+
+	if (ops & PSP_MEM_TRAIN_RESTORE) {
+		amdgpu_device_vram_access(psp->adev, ctx->c2p_train_data_offset, ctx->sys_cache, ctx->train_data_size, true);
+	}
+
+	if (ops & PSP_MEM_TRAIN_SEND_SHORT_MSG) {
+		ret = psp_v11_0_memory_training_send_msg(psp, (amdgpu_force_long_training > 0) ?
+							 PSP_BL__DRAM_LONG_TRAIN : PSP_BL__DRAM_SHORT_TRAIN);
+		if (ret) {
+			DRM_ERROR("send training msg failed.\n");
+			return ret;
+		}
+	}
+	ctx->training_cnt++;
+	return 0;
+}
+
 static const struct psp_funcs psp_v11_0_funcs = {
 	.init_microcode = psp_v11_0_init_microcode,
 	.bootloader_load_kdb = psp_v11_0_bootloader_load_kdb,
@@ -909,6 +1088,9 @@ static const struct psp_funcs psp_v11_0_funcs = {
 	.ras_trigger_error = psp_v11_0_ras_trigger_error,
 	.ras_cure_posion = psp_v11_0_ras_cure_posion,
 	.rlc_autoload_start = psp_v11_0_rlc_autoload_start,
+	.mem_training_init = psp_v11_0_memory_training_init,
+	.mem_training_fini = psp_v11_0_memory_training_fini,
+	.mem_training = psp_v11_0_memory_training,
 };
 
 void psp_v11_0_set_psp_funcs(struct psp_context *psp)
diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v12_0.c b/drivers/gpu/drm/amd/amdgpu/psp_v12_0.c
index c72e43f8e0be..8f553f6f92d6 100644
--- a/drivers/gpu/drm/amd/amdgpu/psp_v12_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/psp_v12_0.c
@@ -378,6 +378,7 @@ static int psp_v12_0_cmd_submit(struct psp_context *psp,
 	write_frame->fence_addr_hi = upper_32_bits(fence_mc_addr);
 	write_frame->fence_addr_lo = lower_32_bits(fence_mc_addr);
 	write_frame->fence_value = index;
+	amdgpu_asic_flush_hdp(adev, NULL);
 
 	/* Update the write Pointer in DWORDs */
 	psp_write_ptr_reg = (psp_write_ptr_reg + rb_frame_size_dw) % ring_size_dw;
diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c b/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c
index d2c727f6a8bd..fdc00938327b 100644
--- a/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c
+++ b/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c
@@ -454,6 +454,7 @@ static int psp_v3_1_cmd_submit(struct psp_context *psp,
 	write_frame->fence_addr_hi = upper_32_bits(fence_mc_addr);
 	write_frame->fence_addr_lo = lower_32_bits(fence_mc_addr);
 	write_frame->fence_value = index;
+	amdgpu_asic_flush_hdp(adev, NULL);
 
 	/* Update the write Pointer in DWORDs */
 	psp_write_ptr_reg = (psp_write_ptr_reg + rb_frame_size_dw) % ring_size_dw;
diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c
index 78452cf0115d..4ef4d31f5231 100644
--- a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c
@@ -254,6 +254,7 @@ static const struct soc15_reg_golden golden_settings_sdma_4_3[] = {
 	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_RLC0_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000),
 	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_RLC1_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000),
 	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_UTCL1_PAGE, 0x000003ff, 0x000003c0),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_UTCL1_WATERMK, 0xfc000000, 0x00000000)
 };
 
 static u32 sdma_v4_0_get_reg_offset(struct amdgpu_device *adev,
@@ -746,13 +747,13 @@ static void sdma_v4_0_ring_emit_hdp_flush(struct amdgpu_ring *ring)
 {
 	struct amdgpu_device *adev = ring->adev;
 	u32 ref_and_mask = 0;
-	const struct nbio_hdp_flush_reg *nbio_hf_reg = adev->nbio_funcs->hdp_flush_reg;
+	const struct nbio_hdp_flush_reg *nbio_hf_reg = adev->nbio.hdp_flush_reg;
 
 	ref_and_mask = nbio_hf_reg->ref_and_mask_sdma0 << ring->me;
 
 	sdma_v4_0_wait_reg_mem(ring, 0, 1,
-			       adev->nbio_funcs->get_hdp_flush_done_offset(adev),
-			       adev->nbio_funcs->get_hdp_flush_req_offset(adev),
+			       adev->nbio.funcs->get_hdp_flush_done_offset(adev),
+			       adev->nbio.funcs->get_hdp_flush_req_offset(adev),
 			       ref_and_mask, ref_and_mask, 10);
 }
 
@@ -1690,102 +1691,17 @@ static int sdma_v4_0_early_init(void *handle)
 }
 
 static int sdma_v4_0_process_ras_data_cb(struct amdgpu_device *adev,
-		struct ras_err_data *err_data,
+		void *err_data,
 		struct amdgpu_iv_entry *entry);
 
 static int sdma_v4_0_late_init(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-	struct ras_common_if **ras_if = &adev->sdma.ras_if;
 	struct ras_ih_if ih_info = {
 		.cb = sdma_v4_0_process_ras_data_cb,
 	};
-	struct ras_fs_if fs_info = {
-		.sysfs_name = "sdma_err_count",
-		.debugfs_name = "sdma_err_inject",
-	};
-	struct ras_common_if ras_block = {
-		.block = AMDGPU_RAS_BLOCK__SDMA,
-		.type = AMDGPU_RAS_ERROR__MULTI_UNCORRECTABLE,
-		.sub_block_index = 0,
-		.name = "sdma",
-	};
-	int r, i;
-
-	if (!amdgpu_ras_is_supported(adev, AMDGPU_RAS_BLOCK__SDMA)) {
-		amdgpu_ras_feature_enable_on_boot(adev, &ras_block, 0);
-		return 0;
-	}
-
-	/* handle resume path. */
-	if (*ras_if) {
-		/* resend ras TA enable cmd during resume.
-		 * prepare to handle failure.
-		 */
-		ih_info.head = **ras_if;
-		r = amdgpu_ras_feature_enable_on_boot(adev, *ras_if, 1);
-		if (r) {
-			if (r == -EAGAIN) {
-				/* request a gpu reset. will run again. */
-				amdgpu_ras_request_reset_on_boot(adev,
-						AMDGPU_RAS_BLOCK__SDMA);
-				return 0;
-			}
-			/* fail to enable ras, cleanup all. */
-			goto irq;
-		}
-		/* enable successfully. continue. */
-		goto resume;
-	}
-
-	*ras_if = kmalloc(sizeof(**ras_if), GFP_KERNEL);
-	if (!*ras_if)
-		return -ENOMEM;
-
-	**ras_if = ras_block;
-
-	r = amdgpu_ras_feature_enable_on_boot(adev, *ras_if, 1);
-	if (r) {
-		if (r == -EAGAIN) {
-			amdgpu_ras_request_reset_on_boot(adev,
-					AMDGPU_RAS_BLOCK__SDMA);
-			r = 0;
-		}
-		goto feature;
-	}
-
-	ih_info.head = **ras_if;
-	fs_info.head = **ras_if;
-
-	r = amdgpu_ras_interrupt_add_handler(adev, &ih_info);
-	if (r)
-		goto interrupt;
-
-	amdgpu_ras_debugfs_create(adev, &fs_info);
 
-	r = amdgpu_ras_sysfs_create(adev, &fs_info);
-	if (r)
-		goto sysfs;
-resume:
-	for (i = 0; i < adev->sdma.num_instances; i++) {
-		r = amdgpu_irq_get(adev, &adev->sdma.ecc_irq,
-				   AMDGPU_SDMA_IRQ_INSTANCE0 + i);
-		if (r)
-			goto irq;
-	}
-
-	return 0;
-irq:
-	amdgpu_ras_sysfs_remove(adev, *ras_if);
-sysfs:
-	amdgpu_ras_debugfs_remove(adev, *ras_if);
-	amdgpu_ras_interrupt_remove_handler(adev, &ih_info);
-interrupt:
-	amdgpu_ras_feature_enable(adev, *ras_if, 0);
-feature:
-	kfree(*ras_if);
-	*ras_if = NULL;
-	return r;
+	return amdgpu_sdma_ras_late_init(adev, &ih_info);
 }
 
 static int sdma_v4_0_sw_init(void *handle)
@@ -1857,21 +1773,7 @@ static int sdma_v4_0_sw_fini(void *handle)
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 	int i;
 
-	if (amdgpu_ras_is_supported(adev, AMDGPU_RAS_BLOCK__SDMA) &&
-			adev->sdma.ras_if) {
-		struct ras_common_if *ras_if = adev->sdma.ras_if;
-		struct ras_ih_if ih_info = {
-			.head = *ras_if,
-		};
-
-		/*remove fs first*/
-		amdgpu_ras_debugfs_remove(adev, ras_if);
-		amdgpu_ras_sysfs_remove(adev, ras_if);
-		/*remove the IH*/
-		amdgpu_ras_interrupt_remove_handler(adev, &ih_info);
-		amdgpu_ras_feature_enable(adev, ras_if, 0);
-		kfree(ras_if);
-	}
+	amdgpu_sdma_ras_fini(adev);
 
 	for (i = 0; i < adev->sdma.num_instances; i++) {
 		amdgpu_ring_fini(&adev->sdma.instance[i].ring);
@@ -1891,7 +1793,7 @@ static int sdma_v4_0_hw_init(void *handle)
 
 	if ((adev->asic_type == CHIP_RAVEN && adev->powerplay.pp_funcs &&
 			adev->powerplay.pp_funcs->set_powergating_by_smu) ||
-			adev->asic_type == CHIP_RENOIR)
+			(adev->asic_type == CHIP_RENOIR && !adev->in_gpu_reset))
 		amdgpu_dpm_set_powergating_by_smu(adev, AMD_IP_BLOCK_TYPE_SDMA, false);
 
 	if (!amdgpu_sriov_vf(adev))
@@ -2024,52 +1926,28 @@ static int sdma_v4_0_process_trap_irq(struct amdgpu_device *adev,
 }
 
 static int sdma_v4_0_process_ras_data_cb(struct amdgpu_device *adev,
-		struct ras_err_data *err_data,
+		void *err_data,
 		struct amdgpu_iv_entry *entry)
 {
-	uint32_t err_source;
 	int instance;
 
+	/* When “Full RAS” is enabled, the per-IP interrupt sources should
+	 * be disabled and the driver should only look for the aggregated
+	 * interrupt via sync flood
+	 */
+	if (amdgpu_ras_is_supported(adev, AMDGPU_RAS_BLOCK__GFX))
+		goto out;
+
 	instance = sdma_v4_0_irq_id_to_seq(entry->client_id);
 	if (instance < 0)
-		return 0;
-
-	switch (entry->src_id) {
-	case SDMA0_4_0__SRCID__SDMA_SRAM_ECC:
-		err_source = 0;
-		break;
-	case SDMA0_4_0__SRCID__SDMA_ECC:
-		err_source = 1;
-		break;
-	default:
-		return 0;
-	}
-
-	kgd2kfd_set_sram_ecc_flag(adev->kfd.dev);
+		goto out;
 
-	amdgpu_ras_reset_gpu(adev, 0);
+	amdgpu_sdma_process_ras_data_cb(adev, err_data, entry);
 
+out:
 	return AMDGPU_RAS_SUCCESS;
 }
 
-static int sdma_v4_0_process_ecc_irq(struct amdgpu_device *adev,
-				      struct amdgpu_irq_src *source,
-				      struct amdgpu_iv_entry *entry)
-{
-	struct ras_common_if *ras_if = adev->sdma.ras_if;
-	struct ras_dispatch_if ih_data = {
-		.entry = entry,
-	};
-
-	if (!ras_if)
-		return 0;
-
-	ih_data.head = *ras_if;
-
-	amdgpu_ras_interrupt_dispatch(adev, &ih_data);
-	return 0;
-}
-
 static int sdma_v4_0_process_illegal_inst_irq(struct amdgpu_device *adev,
 					      struct amdgpu_irq_src *source,
 					      struct amdgpu_iv_entry *entry)
@@ -2417,7 +2295,7 @@ static const struct amdgpu_irq_src_funcs sdma_v4_0_illegal_inst_irq_funcs = {
 
 static const struct amdgpu_irq_src_funcs sdma_v4_0_ecc_irq_funcs = {
 	.set = sdma_v4_0_set_ecc_irq_state,
-	.process = sdma_v4_0_process_ecc_irq,
+	.process = amdgpu_sdma_process_ecc_irq,
 };
 
 
diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c
index fa2f70ce2e2b..f4ad2990f973 100644
--- a/drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c
@@ -406,7 +406,7 @@ static void sdma_v5_0_ring_emit_hdp_flush(struct amdgpu_ring *ring)
 {
 	struct amdgpu_device *adev = ring->adev;
 	u32 ref_and_mask = 0;
-	const struct nbio_hdp_flush_reg *nbio_hf_reg = adev->nbio_funcs->hdp_flush_reg;
+	const struct nbio_hdp_flush_reg *nbio_hf_reg = adev->nbio.hdp_flush_reg;
 
 	if (ring->me == 0)
 		ref_and_mask = nbio_hf_reg->ref_and_mask_sdma0;
@@ -416,8 +416,8 @@ static void sdma_v5_0_ring_emit_hdp_flush(struct amdgpu_ring *ring)
 	amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_POLL_REGMEM) |
 			  SDMA_PKT_POLL_REGMEM_HEADER_HDP_FLUSH(1) |
 			  SDMA_PKT_POLL_REGMEM_HEADER_FUNC(3)); /* == */
-	amdgpu_ring_write(ring, (adev->nbio_funcs->get_hdp_flush_done_offset(adev)) << 2);
-	amdgpu_ring_write(ring, (adev->nbio_funcs->get_hdp_flush_req_offset(adev)) << 2);
+	amdgpu_ring_write(ring, (adev->nbio.funcs->get_hdp_flush_done_offset(adev)) << 2);
+	amdgpu_ring_write(ring, (adev->nbio.funcs->get_hdp_flush_req_offset(adev)) << 2);
 	amdgpu_ring_write(ring, ref_and_mask); /* reference */
 	amdgpu_ring_write(ring, ref_and_mask); /* mask */
 	amdgpu_ring_write(ring, SDMA_PKT_POLL_REGMEM_DW5_RETRY_COUNT(0xfff) |
@@ -683,7 +683,7 @@ static int sdma_v5_0_gfx_resume(struct amdgpu_device *adev)
 		WREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_DOORBELL), doorbell);
 		WREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_DOORBELL_OFFSET), doorbell_offset);
 
-		adev->nbio_funcs->sdma_doorbell_range(adev, i, ring->use_doorbell,
+		adev->nbio.funcs->sdma_doorbell_range(adev, i, ring->use_doorbell,
 						      ring->doorbell_index, 20);
 
 		if (amdgpu_sriov_vf(adev))
@@ -1129,7 +1129,7 @@ static void sdma_v5_0_ring_emit_pipeline_sync(struct amdgpu_ring *ring)
 	amdgpu_ring_write(ring, addr & 0xfffffffc);
 	amdgpu_ring_write(ring, upper_32_bits(addr) & 0xffffffff);
 	amdgpu_ring_write(ring, seq); /* reference */
-	amdgpu_ring_write(ring, 0xfffffff); /* mask */
+	amdgpu_ring_write(ring, 0xffffffff); /* mask */
 	amdgpu_ring_write(ring, SDMA_PKT_POLL_REGMEM_DW5_RETRY_COUNT(0xfff) |
 			  SDMA_PKT_POLL_REGMEM_DW5_INTERVAL(4)); /* retry count, poll interval */
 }
@@ -1173,6 +1173,16 @@ static void sdma_v5_0_ring_emit_reg_wait(struct amdgpu_ring *ring, uint32_t reg,
 			  SDMA_PKT_POLL_REGMEM_DW5_INTERVAL(10));
 }
 
+static void sdma_v5_0_ring_emit_reg_write_reg_wait(struct amdgpu_ring *ring,
+						   uint32_t reg0, uint32_t reg1,
+						   uint32_t ref, uint32_t mask)
+{
+	amdgpu_ring_emit_wreg(ring, reg0, ref);
+	/* wait for a cycle to reset vm_inv_eng*_ack */
+	amdgpu_ring_emit_reg_wait(ring, reg0, 0, 0);
+	amdgpu_ring_emit_reg_wait(ring, reg1, mask, mask);
+}
+
 static int sdma_v5_0_early_init(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
@@ -1588,7 +1598,7 @@ static const struct amdgpu_ring_funcs sdma_v5_0_ring_funcs = {
 		6 + /* sdma_v5_0_ring_emit_pipeline_sync */
 		/* sdma_v5_0_ring_emit_vm_flush */
 		SOC15_FLUSH_GPU_TLB_NUM_WREG * 3 +
-		SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 6 +
+		SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 6 * 2 +
 		10 + 10 + 10, /* sdma_v5_0_ring_emit_fence x3 for user fence, vm fence */
 	.emit_ib_size = 7 + 6, /* sdma_v5_0_ring_emit_ib */
 	.emit_ib = sdma_v5_0_ring_emit_ib,
@@ -1602,6 +1612,7 @@ static const struct amdgpu_ring_funcs sdma_v5_0_ring_funcs = {
 	.pad_ib = sdma_v5_0_ring_pad_ib,
 	.emit_wreg = sdma_v5_0_ring_emit_wreg,
 	.emit_reg_wait = sdma_v5_0_ring_emit_reg_wait,
+	.emit_reg_write_reg_wait = sdma_v5_0_ring_emit_reg_write_reg_wait,
 	.init_cond_exec = sdma_v5_0_ring_init_cond_exec,
 	.patch_cond_exec = sdma_v5_0_ring_patch_cond_exec,
 	.preempt_ib = sdma_v5_0_ring_preempt_ib,
diff --git a/drivers/gpu/drm/amd/amdgpu/si.c b/drivers/gpu/drm/amd/amdgpu/si.c
index 493af42152f2..f2d70a47a3af 100644
--- a/drivers/gpu/drm/amd/amdgpu/si.c
+++ b/drivers/gpu/drm/amd/amdgpu/si.c
@@ -975,6 +975,17 @@ static void si_smc_wreg(struct amdgpu_device *adev, u32 reg, u32 v)
 
 static struct amdgpu_allowed_register_entry si_allowed_read_registers[] = {
 	{GRBM_STATUS},
+	{mmGRBM_STATUS2},
+	{mmGRBM_STATUS_SE0},
+	{mmGRBM_STATUS_SE1},
+	{mmSRBM_STATUS},
+	{mmSRBM_STATUS2},
+	{DMA_STATUS_REG + DMA0_REGISTER_OFFSET},
+	{DMA_STATUS_REG + DMA1_REGISTER_OFFSET},
+	{mmCP_STAT},
+	{mmCP_STALLED_STAT1},
+	{mmCP_STALLED_STAT2},
+	{mmCP_STALLED_STAT3},
 	{GB_ADDR_CONFIG},
 	{MC_ARB_RAMCFG},
 	{GB_TILE_MODE0},
@@ -1633,7 +1644,6 @@ static void si_init_golden_registers(struct amdgpu_device *adev)
 static void si_pcie_gen3_enable(struct amdgpu_device *adev)
 {
 	struct pci_dev *root = adev->pdev->bus->self;
-	int bridge_pos, gpu_pos;
 	u32 speed_cntl, current_data_rate;
 	int i;
 	u16 tmp16;
@@ -1668,12 +1678,7 @@ static void si_pcie_gen3_enable(struct amdgpu_device *adev)
 		DRM_INFO("enabling PCIE gen 2 link speeds, disable with amdgpu.pcie_gen2=0\n");
 	}
 
-	bridge_pos = pci_pcie_cap(root);
-	if (!bridge_pos)
-		return;
-
-	gpu_pos = pci_pcie_cap(adev->pdev);
-	if (!gpu_pos)
+	if (!pci_is_pcie(root) || !pci_is_pcie(adev->pdev))
 		return;
 
 	if (adev->pm.pcie_gen_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_GEN3) {
@@ -1682,14 +1687,17 @@ static void si_pcie_gen3_enable(struct amdgpu_device *adev)
 			u16 bridge_cfg2, gpu_cfg2;
 			u32 max_lw, current_lw, tmp;
 
-			pci_read_config_word(root, bridge_pos + PCI_EXP_LNKCTL, &bridge_cfg);
-			pci_read_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL, &gpu_cfg);
+			pcie_capability_read_word(root, PCI_EXP_LNKCTL,
+						  &bridge_cfg);
+			pcie_capability_read_word(adev->pdev, PCI_EXP_LNKCTL,
+						  &gpu_cfg);
 
 			tmp16 = bridge_cfg | PCI_EXP_LNKCTL_HAWD;
-			pci_write_config_word(root, bridge_pos + PCI_EXP_LNKCTL, tmp16);
+			pcie_capability_write_word(root, PCI_EXP_LNKCTL, tmp16);
 
 			tmp16 = gpu_cfg | PCI_EXP_LNKCTL_HAWD;
-			pci_write_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL, tmp16);
+			pcie_capability_write_word(adev->pdev, PCI_EXP_LNKCTL,
+						   tmp16);
 
 			tmp = RREG32_PCIE(PCIE_LC_STATUS1);
 			max_lw = (tmp & LC_DETECTED_LINK_WIDTH_MASK) >> LC_DETECTED_LINK_WIDTH_SHIFT;
@@ -1706,15 +1714,23 @@ static void si_pcie_gen3_enable(struct amdgpu_device *adev)
 			}
 
 			for (i = 0; i < 10; i++) {
-				pci_read_config_word(adev->pdev, gpu_pos + PCI_EXP_DEVSTA, &tmp16);
+				pcie_capability_read_word(adev->pdev,
+							  PCI_EXP_DEVSTA,
+							  &tmp16);
 				if (tmp16 & PCI_EXP_DEVSTA_TRPND)
 					break;
 
-				pci_read_config_word(root, bridge_pos + PCI_EXP_LNKCTL, &bridge_cfg);
-				pci_read_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL, &gpu_cfg);
+				pcie_capability_read_word(root, PCI_EXP_LNKCTL,
+							  &bridge_cfg);
+				pcie_capability_read_word(adev->pdev,
+							  PCI_EXP_LNKCTL,
+							  &gpu_cfg);
 
-				pci_read_config_word(root, bridge_pos + PCI_EXP_LNKCTL2, &bridge_cfg2);
-				pci_read_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL2, &gpu_cfg2);
+				pcie_capability_read_word(root, PCI_EXP_LNKCTL2,
+							  &bridge_cfg2);
+				pcie_capability_read_word(adev->pdev,
+							  PCI_EXP_LNKCTL2,
+							  &gpu_cfg2);
 
 				tmp = RREG32_PCIE_PORT(PCIE_LC_CNTL4);
 				tmp |= LC_SET_QUIESCE;
@@ -1726,25 +1742,44 @@ static void si_pcie_gen3_enable(struct amdgpu_device *adev)
 
 				mdelay(100);
 
-				pci_read_config_word(root, bridge_pos + PCI_EXP_LNKCTL, &tmp16);
+				pcie_capability_read_word(root, PCI_EXP_LNKCTL,
+							  &tmp16);
 				tmp16 &= ~PCI_EXP_LNKCTL_HAWD;
 				tmp16 |= (bridge_cfg & PCI_EXP_LNKCTL_HAWD);
-				pci_write_config_word(root, bridge_pos + PCI_EXP_LNKCTL, tmp16);
+				pcie_capability_write_word(root, PCI_EXP_LNKCTL,
+							   tmp16);
 
-				pci_read_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL, &tmp16);
+				pcie_capability_read_word(adev->pdev,
+							  PCI_EXP_LNKCTL,
+							  &tmp16);
 				tmp16 &= ~PCI_EXP_LNKCTL_HAWD;
 				tmp16 |= (gpu_cfg & PCI_EXP_LNKCTL_HAWD);
-				pci_write_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL, tmp16);
-
-				pci_read_config_word(root, bridge_pos + PCI_EXP_LNKCTL2, &tmp16);
-				tmp16 &= ~((1 << 4) | (7 << 9));
-				tmp16 |= (bridge_cfg2 & ((1 << 4) | (7 << 9)));
-				pci_write_config_word(root, bridge_pos + PCI_EXP_LNKCTL2, tmp16);
-
-				pci_read_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL2, &tmp16);
-				tmp16 &= ~((1 << 4) | (7 << 9));
-				tmp16 |= (gpu_cfg2 & ((1 << 4) | (7 << 9)));
-				pci_write_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL2, tmp16);
+				pcie_capability_write_word(adev->pdev,
+							   PCI_EXP_LNKCTL,
+							   tmp16);
+
+				pcie_capability_read_word(root, PCI_EXP_LNKCTL2,
+							  &tmp16);
+				tmp16 &= ~(PCI_EXP_LNKCTL2_ENTER_COMP |
+					   PCI_EXP_LNKCTL2_TX_MARGIN);
+				tmp16 |= (bridge_cfg2 &
+					  (PCI_EXP_LNKCTL2_ENTER_COMP |
+					   PCI_EXP_LNKCTL2_TX_MARGIN));
+				pcie_capability_write_word(root,
+							   PCI_EXP_LNKCTL2,
+							   tmp16);
+
+				pcie_capability_read_word(adev->pdev,
+							  PCI_EXP_LNKCTL2,
+							  &tmp16);
+				tmp16 &= ~(PCI_EXP_LNKCTL2_ENTER_COMP |
+					   PCI_EXP_LNKCTL2_TX_MARGIN);
+				tmp16 |= (gpu_cfg2 &
+					  (PCI_EXP_LNKCTL2_ENTER_COMP |
+					   PCI_EXP_LNKCTL2_TX_MARGIN));
+				pcie_capability_write_word(adev->pdev,
+							   PCI_EXP_LNKCTL2,
+							   tmp16);
 
 				tmp = RREG32_PCIE_PORT(PCIE_LC_CNTL4);
 				tmp &= ~LC_SET_QUIESCE;
@@ -1757,15 +1792,16 @@ static void si_pcie_gen3_enable(struct amdgpu_device *adev)
 	speed_cntl &= ~LC_FORCE_DIS_SW_SPEED_CHANGE;
 	WREG32_PCIE_PORT(PCIE_LC_SPEED_CNTL, speed_cntl);
 
-	pci_read_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL2, &tmp16);
-	tmp16 &= ~0xf;
+	pcie_capability_read_word(adev->pdev, PCI_EXP_LNKCTL2, &tmp16);
+	tmp16 &= ~PCI_EXP_LNKCTL2_TLS;
+
 	if (adev->pm.pcie_gen_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_GEN3)
-		tmp16 |= 3;
+		tmp16 |= PCI_EXP_LNKCTL2_TLS_8_0GT; /* gen3 */
 	else if (adev->pm.pcie_gen_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_GEN2)
-		tmp16 |= 2;
+		tmp16 |= PCI_EXP_LNKCTL2_TLS_5_0GT; /* gen2 */
 	else
-		tmp16 |= 1;
-	pci_write_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL2, tmp16);
+		tmp16 |= PCI_EXP_LNKCTL2_TLS_2_5GT; /* gen1 */
+	pcie_capability_write_word(adev->pdev, PCI_EXP_LNKCTL2, tmp16);
 
 	speed_cntl = RREG32_PCIE_PORT(PCIE_LC_SPEED_CNTL);
 	speed_cntl |= LC_INITIATE_LINK_SPEED_CHANGE;
diff --git a/drivers/gpu/drm/amd/amdgpu/si_ih.c b/drivers/gpu/drm/amd/amdgpu/si_ih.c
index 57bb5f9e08b2..88ae27a5a03d 100644
--- a/drivers/gpu/drm/amd/amdgpu/si_ih.c
+++ b/drivers/gpu/drm/amd/amdgpu/si_ih.c
@@ -64,7 +64,8 @@ static int si_ih_irq_init(struct amdgpu_device *adev)
 	u32 interrupt_cntl, ih_cntl, ih_rb_cntl;
 
 	si_ih_disable_interrupts(adev);
-	WREG32(INTERRUPT_CNTL2, adev->irq.ih.gpu_addr >> 8);
+	/* set dummy read address to dummy page address */
+	WREG32(INTERRUPT_CNTL2, adev->dummy_page_addr >> 8);
 	interrupt_cntl = RREG32(INTERRUPT_CNTL);
 	interrupt_cntl &= ~IH_DUMMY_RD_OVERRIDE;
 	interrupt_cntl &= ~IH_REQ_NONSNOOP_EN;
diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c
index f70658a536a9..8e1640bc07af 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc15.c
+++ b/drivers/gpu/drm/amd/amdgpu/soc15.c
@@ -58,6 +58,9 @@
 #include "mmhub_v1_0.h"
 #include "df_v1_7.h"
 #include "df_v3_6.h"
+#include "nbio_v6_1.h"
+#include "nbio_v7_0.h"
+#include "nbio_v7_4.h"
 #include "vega10_ih.h"
 #include "sdma_v4_0.h"
 #include "uvd_v7_0.h"
@@ -91,8 +94,8 @@ static u32 soc15_pcie_rreg(struct amdgpu_device *adev, u32 reg)
 {
 	unsigned long flags, address, data;
 	u32 r;
-	address = adev->nbio_funcs->get_pcie_index_offset(adev);
-	data = adev->nbio_funcs->get_pcie_data_offset(adev);
+	address = adev->nbio.funcs->get_pcie_index_offset(adev);
+	data = adev->nbio.funcs->get_pcie_data_offset(adev);
 
 	spin_lock_irqsave(&adev->pcie_idx_lock, flags);
 	WREG32(address, reg);
@@ -106,8 +109,8 @@ static void soc15_pcie_wreg(struct amdgpu_device *adev, u32 reg, u32 v)
 {
 	unsigned long flags, address, data;
 
-	address = adev->nbio_funcs->get_pcie_index_offset(adev);
-	data = adev->nbio_funcs->get_pcie_data_offset(adev);
+	address = adev->nbio.funcs->get_pcie_index_offset(adev);
+	data = adev->nbio.funcs->get_pcie_data_offset(adev);
 
 	spin_lock_irqsave(&adev->pcie_idx_lock, flags);
 	WREG32(address, reg);
@@ -121,8 +124,8 @@ static u64 soc15_pcie_rreg64(struct amdgpu_device *adev, u32 reg)
 {
 	unsigned long flags, address, data;
 	u64 r;
-	address = adev->nbio_funcs->get_pcie_index_offset(adev);
-	data = adev->nbio_funcs->get_pcie_data_offset(adev);
+	address = adev->nbio.funcs->get_pcie_index_offset(adev);
+	data = adev->nbio.funcs->get_pcie_data_offset(adev);
 
 	spin_lock_irqsave(&adev->pcie_idx_lock, flags);
 	/* read low 32 bit */
@@ -142,8 +145,8 @@ static void soc15_pcie_wreg64(struct amdgpu_device *adev, u32 reg, u64 v)
 {
 	unsigned long flags, address, data;
 
-	address = adev->nbio_funcs->get_pcie_index_offset(adev);
-	data = adev->nbio_funcs->get_pcie_data_offset(adev);
+	address = adev->nbio.funcs->get_pcie_index_offset(adev);
+	data = adev->nbio.funcs->get_pcie_data_offset(adev);
 
 	spin_lock_irqsave(&adev->pcie_idx_lock, flags);
 	/* write low 32 bit */
@@ -262,7 +265,7 @@ static void soc15_se_cac_wreg(struct amdgpu_device *adev, u32 reg, u32 v)
 
 static u32 soc15_get_config_memsize(struct amdgpu_device *adev)
 {
-	return adev->nbio_funcs->get_memsize(adev);
+	return adev->nbio.funcs->get_memsize(adev);
 }
 
 static u32 soc15_get_xclk(struct amdgpu_device *adev)
@@ -336,6 +339,7 @@ static struct soc15_allowed_register_entry soc15_allowed_read_registers[] = {
 	{ SOC15_REG_ENTRY(GC, 0, mmCP_CPF_BUSY_STAT)},
 	{ SOC15_REG_ENTRY(GC, 0, mmCP_CPF_STALLED_STAT1)},
 	{ SOC15_REG_ENTRY(GC, 0, mmCP_CPF_STATUS)},
+	{ SOC15_REG_ENTRY(GC, 0, mmCP_CPC_BUSY_STAT)},
 	{ SOC15_REG_ENTRY(GC, 0, mmCP_CPC_STALLED_STAT1)},
 	{ SOC15_REG_ENTRY(GC, 0, mmCP_CPC_STATUS)},
 	{ SOC15_REG_ENTRY(GC, 0, mmGB_ADDR_CONFIG)},
@@ -461,7 +465,7 @@ static int soc15_asic_mode1_reset(struct amdgpu_device *adev)
 
 	/* wait for asic to come out of reset */
 	for (i = 0; i < adev->usec_timeout; i++) {
-		u32 memsize = adev->nbio_funcs->get_memsize(adev);
+		u32 memsize = adev->nbio.funcs->get_memsize(adev);
 
 		if (memsize != 0xffffffff)
 			break;
@@ -475,42 +479,66 @@ static int soc15_asic_mode1_reset(struct amdgpu_device *adev)
 
 static int soc15_asic_get_baco_capability(struct amdgpu_device *adev, bool *cap)
 {
-	void *pp_handle = adev->powerplay.pp_handle;
-	const struct amd_pm_funcs *pp_funcs = adev->powerplay.pp_funcs;
+	if (is_support_sw_smu(adev)) {
+		struct smu_context *smu = &adev->smu;
 
-	if (!pp_funcs || !pp_funcs->get_asic_baco_capability) {
-		*cap = false;
-		return -ENOENT;
-	}
+		*cap = smu_baco_is_support(smu);
+		return 0;
+	} else {
+		void *pp_handle = adev->powerplay.pp_handle;
+		const struct amd_pm_funcs *pp_funcs = adev->powerplay.pp_funcs;
+
+		if (!pp_funcs || !pp_funcs->get_asic_baco_capability) {
+			*cap = false;
+			return -ENOENT;
+		}
 
-	return pp_funcs->get_asic_baco_capability(pp_handle, cap);
+		return pp_funcs->get_asic_baco_capability(pp_handle, cap);
+	}
 }
 
 static int soc15_asic_baco_reset(struct amdgpu_device *adev)
 {
-	void *pp_handle = adev->powerplay.pp_handle;
-	const struct amd_pm_funcs *pp_funcs = adev->powerplay.pp_funcs;
+	struct amdgpu_ras *ras = amdgpu_ras_get_context(adev);
 
-	if (!pp_funcs ||!pp_funcs->get_asic_baco_state ||!pp_funcs->set_asic_baco_state)
-		return -ENOENT;
+	/* avoid NBIF got stuck when do RAS recovery in BACO reset */
+	if (ras && ras->supported)
+		adev->nbio.funcs->enable_doorbell_interrupt(adev, false);
+
+	dev_info(adev->dev, "GPU BACO reset\n");
 
-	/* enter BACO state */
-	if (pp_funcs->set_asic_baco_state(pp_handle, 1))
-		return -EIO;
+	if (is_support_sw_smu(adev)) {
+		struct smu_context *smu = &adev->smu;
 
-	/* exit BACO state */
-	if (pp_funcs->set_asic_baco_state(pp_handle, 0))
-		return -EIO;
+		if (smu_baco_reset(smu))
+			return -EIO;
+	} else {
+		void *pp_handle = adev->powerplay.pp_handle;
+		const struct amd_pm_funcs *pp_funcs = adev->powerplay.pp_funcs;
 
-	dev_info(adev->dev, "GPU BACO reset\n");
+		if (!pp_funcs ||!pp_funcs->get_asic_baco_state ||!pp_funcs->set_asic_baco_state)
+			return -ENOENT;
 
-	adev->in_baco_reset = 1;
+		/* enter BACO state */
+		if (pp_funcs->set_asic_baco_state(pp_handle, 1))
+			return -EIO;
+
+		/* exit BACO state */
+		if (pp_funcs->set_asic_baco_state(pp_handle, 0))
+			return -EIO;
+	}
+
+	/* re-enable doorbell interrupt after BACO exit */
+	if (ras && ras->supported)
+		adev->nbio.funcs->enable_doorbell_interrupt(adev, true);
 
 	return 0;
 }
 
 static int soc15_mode2_reset(struct amdgpu_device *adev)
 {
+	if (is_support_sw_smu(adev))
+		return smu_mode2_reset(&adev->smu);
 	if (!adev->powerplay.pp_funcs ||
 	    !adev->powerplay.pp_funcs->asic_reset_mode_2)
 		return -ENOENT;
@@ -525,6 +553,7 @@ soc15_asic_reset_method(struct amdgpu_device *adev)
 
 	switch (adev->asic_type) {
 	case CHIP_RAVEN:
+	case CHIP_RENOIR:
 		return AMD_RESET_METHOD_MODE2;
 	case CHIP_VEGA10:
 	case CHIP_VEGA12:
@@ -558,12 +587,14 @@ static int soc15_asic_reset(struct amdgpu_device *adev)
 {
 	switch (soc15_asic_reset_method(adev)) {
 		case AMD_RESET_METHOD_BACO:
-			amdgpu_inc_vram_lost(adev);
+			if (!adev->in_suspend)
+				amdgpu_inc_vram_lost(adev);
 			return soc15_asic_baco_reset(adev);
 		case AMD_RESET_METHOD_MODE2:
 			return soc15_mode2_reset(adev);
 		default:
-			amdgpu_inc_vram_lost(adev);
+			if (!adev->in_suspend)
+				amdgpu_inc_vram_lost(adev);
 			return soc15_asic_mode1_reset(adev);
 	}
 }
@@ -624,8 +655,8 @@ static void soc15_program_aspm(struct amdgpu_device *adev)
 static void soc15_enable_doorbell_aperture(struct amdgpu_device *adev,
 					   bool enable)
 {
-	adev->nbio_funcs->enable_doorbell_aperture(adev, enable);
-	adev->nbio_funcs->enable_doorbell_selfring_aperture(adev, enable);
+	adev->nbio.funcs->enable_doorbell_aperture(adev, enable);
+	adev->nbio.funcs->enable_doorbell_selfring_aperture(adev, enable);
 }
 
 static const struct amdgpu_ip_block_version vega10_common_ip_block =
@@ -639,7 +670,7 @@ static const struct amdgpu_ip_block_version vega10_common_ip_block =
 
 static uint32_t soc15_get_rev_id(struct amdgpu_device *adev)
 {
-	return adev->nbio_funcs->get_rev_id(adev);
+	return adev->nbio.funcs->get_rev_id(adev);
 }
 
 int soc15_set_ip_blocks(struct amdgpu_device *adev)
@@ -665,13 +696,17 @@ int soc15_set_ip_blocks(struct amdgpu_device *adev)
 	if (adev->asic_type == CHIP_VEGA20 || adev->asic_type == CHIP_ARCTURUS)
 		adev->gmc.xgmi.supported = true;
 
-	if (adev->flags & AMD_IS_APU)
-		adev->nbio_funcs = &nbio_v7_0_funcs;
-	else if (adev->asic_type == CHIP_VEGA20 ||
-		adev->asic_type == CHIP_ARCTURUS)
-		adev->nbio_funcs = &nbio_v7_4_funcs;
-	else
-		adev->nbio_funcs = &nbio_v6_1_funcs;
+	if (adev->flags & AMD_IS_APU) {
+		adev->nbio.funcs = &nbio_v7_0_funcs;
+		adev->nbio.hdp_flush_reg = &nbio_v7_0_hdp_flush_reg;
+	} else if (adev->asic_type == CHIP_VEGA20 ||
+		   adev->asic_type == CHIP_ARCTURUS) {
+		adev->nbio.funcs = &nbio_v7_4_funcs;
+		adev->nbio.hdp_flush_reg = &nbio_v7_4_hdp_flush_reg;
+	} else {
+		adev->nbio.funcs = &nbio_v6_1_funcs;
+		adev->nbio.hdp_flush_reg = &nbio_v6_1_hdp_flush_reg;
+	}
 
 	if (adev->asic_type == CHIP_VEGA20 || adev->asic_type == CHIP_ARCTURUS)
 		adev->df_funcs = &df_v3_6_funcs;
@@ -679,7 +714,7 @@ int soc15_set_ip_blocks(struct amdgpu_device *adev)
 		adev->df_funcs = &df_v1_7_funcs;
 
 	adev->rev_id = soc15_get_rev_id(adev);
-	adev->nbio_funcs->detect_hw_virt(adev);
+	adev->nbio.funcs->detect_hw_virt(adev);
 
 	if (amdgpu_sriov_vf(adev))
 		adev->virt.ops = &xgpu_ai_virt_ops;
@@ -748,13 +783,26 @@ int soc15_set_ip_blocks(struct amdgpu_device *adev)
 	case CHIP_ARCTURUS:
 		amdgpu_device_ip_block_add(adev, &vega10_common_ip_block);
 		amdgpu_device_ip_block_add(adev, &gmc_v9_0_ip_block);
-		amdgpu_device_ip_block_add(adev, &vega10_ih_ip_block);
+
+		if (amdgpu_sriov_vf(adev)) {
+			if (likely(adev->firmware.load_type == AMDGPU_FW_LOAD_PSP))
+				amdgpu_device_ip_block_add(adev, &psp_v11_0_ip_block);
+			amdgpu_device_ip_block_add(adev, &vega10_ih_ip_block);
+		} else {
+			amdgpu_device_ip_block_add(adev, &vega10_ih_ip_block);
+			if (likely(adev->firmware.load_type == AMDGPU_FW_LOAD_PSP))
+				amdgpu_device_ip_block_add(adev, &psp_v11_0_ip_block);
+		}
+
 		if (adev->enable_virtual_display || amdgpu_sriov_vf(adev))
 			amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block);
 		amdgpu_device_ip_block_add(adev, &gfx_v9_0_ip_block);
 		amdgpu_device_ip_block_add(adev, &sdma_v4_0_ip_block);
-		amdgpu_device_ip_block_add(adev, &smu_v11_0_ip_block);
-		amdgpu_device_ip_block_add(adev, &vcn_v2_5_ip_block);
+		if (!amdgpu_sriov_vf(adev))
+			amdgpu_device_ip_block_add(adev, &smu_v11_0_ip_block);
+
+		if (unlikely(adev->firmware.load_type == AMDGPU_FW_LOAD_DIRECT))
+			amdgpu_device_ip_block_add(adev, &vcn_v2_5_ip_block);
 		break;
 	case CHIP_RENOIR:
 		amdgpu_device_ip_block_add(adev, &vega10_common_ip_block);
@@ -771,8 +819,6 @@ int soc15_set_ip_blocks(struct amdgpu_device *adev)
 #if defined(CONFIG_DRM_AMD_DC)
                 else if (amdgpu_device_has_dc_support(adev))
                         amdgpu_device_ip_block_add(adev, &dm_ip_block);
-#else
-#       warning "Enable CONFIG_DRM_AMD_DC for display support on SOC15."
 #endif
 		amdgpu_device_ip_block_add(adev, &vcn_v2_0_ip_block);
 		break;
@@ -785,7 +831,7 @@ int soc15_set_ip_blocks(struct amdgpu_device *adev)
 
 static void soc15_flush_hdp(struct amdgpu_device *adev, struct amdgpu_ring *ring)
 {
-	adev->nbio_funcs->hdp_flush(adev, ring);
+	adev->nbio.funcs->hdp_flush(adev, ring);
 }
 
 static void soc15_invalidate_hdp(struct amdgpu_device *adev,
@@ -1099,7 +1145,9 @@ static int soc15_common_early_init(void *handle)
 				AMD_CG_SUPPORT_SDMA_LS |
 				AMD_CG_SUPPORT_VCN_MGCG;
 
-			adev->pg_flags = AMD_PG_SUPPORT_SDMA | AMD_PG_SUPPORT_VCN;
+			adev->pg_flags = AMD_PG_SUPPORT_SDMA |
+				AMD_PG_SUPPORT_VCN |
+				AMD_PG_SUPPORT_VCN_DPG;
 		} else if (adev->pdev->device == 0x15d8) {
 			adev->cg_flags = AMD_CG_SUPPORT_GFX_MGCG |
 				AMD_CG_SUPPORT_GFX_MGLS |
@@ -1142,7 +1190,9 @@ static int soc15_common_early_init(void *handle)
 				AMD_CG_SUPPORT_SDMA_LS |
 				AMD_CG_SUPPORT_VCN_MGCG;
 
-			adev->pg_flags = AMD_PG_SUPPORT_SDMA | AMD_PG_SUPPORT_VCN;
+			adev->pg_flags = AMD_PG_SUPPORT_SDMA |
+				AMD_PG_SUPPORT_VCN |
+				AMD_PG_SUPPORT_VCN_DPG;
 		}
 		break;
 	case CHIP_ARCTURUS:
@@ -1157,7 +1207,8 @@ static int soc15_common_early_init(void *handle)
 			AMD_CG_SUPPORT_SDMA_MGCG |
 			AMD_CG_SUPPORT_SDMA_LS |
 			AMD_CG_SUPPORT_MC_MGCG |
-			AMD_CG_SUPPORT_MC_LS;
+			AMD_CG_SUPPORT_MC_LS |
+			AMD_CG_SUPPORT_IH_CG;
 		adev->pg_flags = 0;
 		adev->external_rev_id = adev->rev_id + 0x32;
 		break;
@@ -1186,11 +1237,6 @@ static int soc15_common_early_init(void *handle)
 				 AMD_PG_SUPPORT_VCN |
 				 AMD_PG_SUPPORT_VCN_DPG;
 		adev->external_rev_id = adev->rev_id + 0x91;
-
-		if (adev->pm.pp_feature & PP_GFXOFF_MASK)
-			adev->pg_flags |= AMD_PG_SUPPORT_GFX_PG |
-				AMD_PG_SUPPORT_CP |
-				AMD_PG_SUPPORT_RLC_SMU_HS;
 		break;
 	default:
 		/* FIXME: not supported yet */
@@ -1208,11 +1254,15 @@ static int soc15_common_early_init(void *handle)
 static int soc15_common_late_init(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+	int r = 0;
 
 	if (amdgpu_sriov_vf(adev))
 		xgpu_ai_mailbox_get_irq(adev);
 
-	return 0;
+	if (adev->nbio.funcs->ras_late_init)
+		r = adev->nbio.funcs->ras_late_init(adev);
+
+	return r;
 }
 
 static int soc15_common_sw_init(void *handle)
@@ -1229,6 +1279,10 @@ static int soc15_common_sw_init(void *handle)
 
 static int soc15_common_sw_fini(void *handle)
 {
+	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+	amdgpu_nbio_ras_fini(adev);
+	adev->df_funcs->sw_fini(adev);
 	return 0;
 }
 
@@ -1241,12 +1295,12 @@ static void soc15_doorbell_range_init(struct amdgpu_device *adev)
 	if (!amdgpu_sriov_vf(adev)) {
 		for (i = 0; i < adev->sdma.num_instances; i++) {
 			ring = &adev->sdma.instance[i].ring;
-			adev->nbio_funcs->sdma_doorbell_range(adev, i,
+			adev->nbio.funcs->sdma_doorbell_range(adev, i,
 				ring->use_doorbell, ring->doorbell_index,
 				adev->doorbell_index.sdma_doorbell_range);
 		}
 
-		adev->nbio_funcs->ih_doorbell_range(adev, adev->irq.ih.use_doorbell,
+		adev->nbio.funcs->ih_doorbell_range(adev, adev->irq.ih.use_doorbell,
 						adev->irq.ih.doorbell_index);
 	}
 }
@@ -1260,13 +1314,13 @@ static int soc15_common_hw_init(void *handle)
 	/* enable aspm */
 	soc15_program_aspm(adev);
 	/* setup nbio registers */
-	adev->nbio_funcs->init_registers(adev);
+	adev->nbio.funcs->init_registers(adev);
 	/* remap HDP registers to a hole in mmio space,
 	 * for the purpose of expose those registers
 	 * to process space
 	 */
-	if (adev->nbio_funcs->remap_hdp_registers)
-		adev->nbio_funcs->remap_hdp_registers(adev);
+	if (adev->nbio.funcs->remap_hdp_registers)
+		adev->nbio.funcs->remap_hdp_registers(adev);
 
 	/* enable the doorbell aperture */
 	soc15_enable_doorbell_aperture(adev, true);
@@ -1289,6 +1343,14 @@ static int soc15_common_hw_fini(void *handle)
 	if (amdgpu_sriov_vf(adev))
 		xgpu_ai_mailbox_put_irq(adev);
 
+	if (adev->nbio.ras_if &&
+	    amdgpu_ras_is_supported(adev, adev->nbio.ras_if->block)) {
+		if (adev->nbio.funcs->init_ras_controller_interrupt)
+			amdgpu_irq_put(adev, &adev->nbio.ras_controller_irq, 0);
+		if (adev->nbio.funcs->init_ras_err_event_athub_interrupt)
+			amdgpu_irq_put(adev, &adev->nbio.ras_err_event_athub_irq, 0);
+	}
+
 	return 0;
 }
 
@@ -1429,9 +1491,9 @@ static int soc15_common_set_clockgating_state(void *handle,
 	case CHIP_VEGA10:
 	case CHIP_VEGA12:
 	case CHIP_VEGA20:
-		adev->nbio_funcs->update_medium_grain_clock_gating(adev,
+		adev->nbio.funcs->update_medium_grain_clock_gating(adev,
 				state == AMD_CG_STATE_GATE ? true : false);
-		adev->nbio_funcs->update_medium_grain_light_sleep(adev,
+		adev->nbio.funcs->update_medium_grain_light_sleep(adev,
 				state == AMD_CG_STATE_GATE ? true : false);
 		soc15_update_hdp_light_sleep(adev,
 				state == AMD_CG_STATE_GATE ? true : false);
@@ -1446,9 +1508,9 @@ static int soc15_common_set_clockgating_state(void *handle,
 		break;
 	case CHIP_RAVEN:
 	case CHIP_RENOIR:
-		adev->nbio_funcs->update_medium_grain_clock_gating(adev,
+		adev->nbio.funcs->update_medium_grain_clock_gating(adev,
 				state == AMD_CG_STATE_GATE ? true : false);
-		adev->nbio_funcs->update_medium_grain_light_sleep(adev,
+		adev->nbio.funcs->update_medium_grain_light_sleep(adev,
 				state == AMD_CG_STATE_GATE ? true : false);
 		soc15_update_hdp_light_sleep(adev,
 				state == AMD_CG_STATE_GATE ? true : false);
@@ -1477,7 +1539,7 @@ static void soc15_common_get_clockgating_state(void *handle, u32 *flags)
 	if (amdgpu_sriov_vf(adev))
 		*flags = 0;
 
-	adev->nbio_funcs->get_clockgating_state(adev, flags);
+	adev->nbio.funcs->get_clockgating_state(adev, flags);
 
 	/* AMD_CG_SUPPORT_HDP_LS */
 	data = RREG32(SOC15_REG_OFFSET(HDP, 0, mmHDP_MEM_POWER_LS));
diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.h b/drivers/gpu/drm/amd/amdgpu/soc15.h
index a3dde0c31f57..57af489a5de3 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc15.h
+++ b/drivers/gpu/drm/amd/amdgpu/soc15.h
@@ -28,8 +28,8 @@
 #include "nbio_v7_0.h"
 #include "nbio_v7_4.h"
 
-#define SOC15_FLUSH_GPU_TLB_NUM_WREG		4
-#define SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT	1
+#define SOC15_FLUSH_GPU_TLB_NUM_WREG		6
+#define SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT	3
 
 extern const struct amd_ip_funcs soc15_common_ip_funcs;
 
@@ -67,6 +67,8 @@ struct soc15_allowed_register_entry {
 #define SOC15_REG_GOLDEN_VALUE(ip, inst, reg, and_mask, or_mask) \
 	{ ip##_HWIP, inst, reg##_BASE_IDX, reg, and_mask, or_mask }
 
+#define SOC15_REG_FIELD(reg, field) reg##__##field##_MASK, reg##__##field##__SHIFT
+
 void soc15_grbm_select(struct amdgpu_device *adev,
 		    u32 me, u32 pipe, u32 queue, u32 vmid);
 int soc15_set_ip_blocks(struct amdgpu_device *adev);
diff --git a/drivers/gpu/drm/amd/amdgpu/umc_v6_0.c b/drivers/gpu/drm/amd/amdgpu/umc_v6_0.c
new file mode 100644
index 000000000000..0d6b50528d76
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdgpu/umc_v6_0.c
@@ -0,0 +1,37 @@
+/*
+ * Copyright 2019 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#include "umc_v6_0.h"
+#include "amdgpu.h"
+
+static void umc_v6_0_init_registers(struct amdgpu_device *adev)
+{
+	unsigned i,j;
+
+	for (i = 0; i < 4; i++)
+		for (j = 0; j < 4; j++)
+			WREG32((i*0x100000 + 0x5010c + j*0x2000)/4, 0x1002);
+}
+
+const struct amdgpu_umc_funcs umc_v6_0_funcs = {
+	.init_registers = umc_v6_0_init_registers,
+};
diff --git a/drivers/gpu/drm/amd/amdgpu/umc_v6_0.h b/drivers/gpu/drm/amd/amdgpu/umc_v6_0.h
new file mode 100644
index 000000000000..109f1a57a46e
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdgpu/umc_v6_0.h
@@ -0,0 +1,31 @@
+/*
+ * Copyright 2019 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#ifndef __UMC_V6_0_H__
+#define __UMC_V6_0_H__
+
+#include "soc15_common.h"
+#include "amdgpu.h"
+
+extern const struct amdgpu_umc_funcs umc_v6_0_funcs;
+
+#endif
diff --git a/drivers/gpu/drm/amd/amdgpu/umc_v6_1.c b/drivers/gpu/drm/amd/amdgpu/umc_v6_1.c
index 8502e736f721..47c4b96b14d1 100644
--- a/drivers/gpu/drm/amd/amdgpu/umc_v6_1.c
+++ b/drivers/gpu/drm/amd/amdgpu/umc_v6_1.c
@@ -75,6 +75,17 @@ static void umc_v6_1_disable_umc_index_mode(struct amdgpu_device *adev)
 			RSMU_UMC_INDEX_MODE_EN, 0);
 }
 
+static uint32_t umc_v6_1_get_umc_inst(struct amdgpu_device *adev)
+{
+	uint32_t rsmu_umc_index;
+
+	rsmu_umc_index = RREG32_SOC15(RSMU, 0,
+				mmRSMU_UMC_INDEX_REGISTER_NBIF_VG20_GPU);
+	return REG_GET_FIELD(rsmu_umc_index,
+				RSMU_UMC_INDEX_REGISTER_NBIF_VG20_GPU,
+				RSMU_UMC_INDEX_INSTANCE);
+}
+
 static void umc_v6_1_query_correctable_error_count(struct amdgpu_device *adev,
 						   uint32_t umc_reg_offset,
 						   unsigned long *error_count)
@@ -165,7 +176,8 @@ static void umc_v6_1_query_error_address(struct amdgpu_device *adev,
 					 uint32_t umc_reg_offset, uint32_t channel_index)
 {
 	uint32_t lsb, mc_umc_status_addr;
-	uint64_t mc_umc_status, err_addr;
+	uint64_t mc_umc_status, err_addr, retired_page;
+	struct eeprom_table_record *err_rec;
 
 	mc_umc_status_addr =
 		SOC15_REG_OFFSET(UMC, 0, mmMCA_UMC_UMC0_MCUMC_STATUST0);
@@ -177,6 +189,7 @@ static void umc_v6_1_query_error_address(struct amdgpu_device *adev,
 		return;
 	}
 
+	err_rec = &err_data->err_addr[err_data->err_addr_cnt];
 	mc_umc_status = RREG64_UMC(mc_umc_status_addr + umc_reg_offset);
 
 	/* calculate error address if ue/ce error is detected */
@@ -191,12 +204,24 @@ static void umc_v6_1_query_error_address(struct amdgpu_device *adev,
 		err_addr &= ~((0x1ULL << lsb) - 1);
 
 		/* translate umc channel address to soc pa, 3 parts are included */
-		err_data->err_addr[err_data->err_addr_cnt] =
-						ADDR_OF_8KB_BLOCK(err_addr) |
-						ADDR_OF_256B_BLOCK(channel_index) |
-						OFFSET_IN_256B_BLOCK(err_addr);
-
-		err_data->err_addr_cnt++;
+		retired_page = ADDR_OF_8KB_BLOCK(err_addr) |
+				ADDR_OF_256B_BLOCK(channel_index) |
+				OFFSET_IN_256B_BLOCK(err_addr);
+
+		/* we only save ue error information currently, ce is skipped */
+		if (REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, UECC)
+				== 1) {
+			err_rec->address = err_addr;
+			/* page frame address is saved */
+			err_rec->retired_page = retired_page >> AMDGPU_GPU_PAGE_SHIFT;
+			err_rec->ts = (uint64_t)ktime_get_real_seconds();
+			err_rec->err_type = AMDGPU_RAS_EEPROM_ERR_NON_RECOVERABLE;
+			err_rec->cu = 0;
+			err_rec->mem_channel = channel_index;
+			err_rec->mcumc_id = umc_v6_1_get_umc_inst(adev);
+
+			err_data->err_addr_cnt++;
+		}
 	}
 
 	/* clear umc status */
@@ -209,7 +234,7 @@ static void umc_v6_1_query_ras_error_address(struct amdgpu_device *adev,
 	amdgpu_umc_for_each_channel(umc_v6_1_query_error_address);
 }
 
-static void umc_v6_1_ras_init_per_channel(struct amdgpu_device *adev,
+static void umc_v6_1_err_cnt_init_per_channel(struct amdgpu_device *adev,
 					 struct ras_err_data *err_data,
 					 uint32_t umc_reg_offset, uint32_t channel_index)
 {
@@ -239,15 +264,16 @@ static void umc_v6_1_ras_init_per_channel(struct amdgpu_device *adev,
 	WREG32(ecc_err_cnt_addr + umc_reg_offset, UMC_V6_1_CE_CNT_INIT);
 }
 
-static void umc_v6_1_ras_init(struct amdgpu_device *adev)
+static void umc_v6_1_err_cnt_init(struct amdgpu_device *adev)
 {
 	void *ras_error_status = NULL;
 
-	amdgpu_umc_for_each_channel(umc_v6_1_ras_init_per_channel);
+	amdgpu_umc_for_each_channel(umc_v6_1_err_cnt_init_per_channel);
 }
 
 const struct amdgpu_umc_funcs umc_v6_1_funcs = {
-	.ras_init = umc_v6_1_ras_init,
+	.err_cnt_init = umc_v6_1_err_cnt_init,
+	.ras_late_init = amdgpu_umc_ras_late_init,
 	.query_ras_error_count = umc_v6_1_query_ras_error_count,
 	.query_ras_error_address = umc_v6_1_query_ras_error_address,
 	.enable_umc_index_mode = umc_v6_1_enable_umc_index_mode,
diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c
index 670784a78512..217084d56ab8 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c
@@ -206,13 +206,14 @@ static int uvd_v6_0_enc_ring_test_ring(struct amdgpu_ring *ring)
  * Open up a stream for HW test
  */
 static int uvd_v6_0_enc_get_create_msg(struct amdgpu_ring *ring, uint32_t handle,
+				       struct amdgpu_bo *bo,
 				       struct dma_fence **fence)
 {
 	const unsigned ib_size_dw = 16;
 	struct amdgpu_job *job;
 	struct amdgpu_ib *ib;
 	struct dma_fence *f = NULL;
-	uint64_t dummy;
+	uint64_t addr;
 	int i, r;
 
 	r = amdgpu_job_alloc_with_ib(ring->adev, ib_size_dw * 4, &job);
@@ -220,15 +221,15 @@ static int uvd_v6_0_enc_get_create_msg(struct amdgpu_ring *ring, uint32_t handle
 		return r;
 
 	ib = &job->ibs[0];
-	dummy = ib->gpu_addr + 1024;
+	addr = amdgpu_bo_gpu_offset(bo);
 
 	ib->length_dw = 0;
 	ib->ptr[ib->length_dw++] = 0x00000018;
 	ib->ptr[ib->length_dw++] = 0x00000001; /* session info */
 	ib->ptr[ib->length_dw++] = handle;
 	ib->ptr[ib->length_dw++] = 0x00010000;
-	ib->ptr[ib->length_dw++] = upper_32_bits(dummy);
-	ib->ptr[ib->length_dw++] = dummy;
+	ib->ptr[ib->length_dw++] = upper_32_bits(addr);
+	ib->ptr[ib->length_dw++] = addr;
 
 	ib->ptr[ib->length_dw++] = 0x00000014;
 	ib->ptr[ib->length_dw++] = 0x00000002; /* task info */
@@ -268,13 +269,14 @@ err:
  */
 static int uvd_v6_0_enc_get_destroy_msg(struct amdgpu_ring *ring,
 					uint32_t handle,
+					struct amdgpu_bo *bo,
 					struct dma_fence **fence)
 {
 	const unsigned ib_size_dw = 16;
 	struct amdgpu_job *job;
 	struct amdgpu_ib *ib;
 	struct dma_fence *f = NULL;
-	uint64_t dummy;
+	uint64_t addr;
 	int i, r;
 
 	r = amdgpu_job_alloc_with_ib(ring->adev, ib_size_dw * 4, &job);
@@ -282,15 +284,15 @@ static int uvd_v6_0_enc_get_destroy_msg(struct amdgpu_ring *ring,
 		return r;
 
 	ib = &job->ibs[0];
-	dummy = ib->gpu_addr + 1024;
+	addr = amdgpu_bo_gpu_offset(bo);
 
 	ib->length_dw = 0;
 	ib->ptr[ib->length_dw++] = 0x00000018;
 	ib->ptr[ib->length_dw++] = 0x00000001; /* session info */
 	ib->ptr[ib->length_dw++] = handle;
 	ib->ptr[ib->length_dw++] = 0x00010000;
-	ib->ptr[ib->length_dw++] = upper_32_bits(dummy);
-	ib->ptr[ib->length_dw++] = dummy;
+	ib->ptr[ib->length_dw++] = upper_32_bits(addr);
+	ib->ptr[ib->length_dw++] = addr;
 
 	ib->ptr[ib->length_dw++] = 0x00000014;
 	ib->ptr[ib->length_dw++] = 0x00000002; /* task info */
@@ -327,13 +329,20 @@ err:
 static int uvd_v6_0_enc_ring_test_ib(struct amdgpu_ring *ring, long timeout)
 {
 	struct dma_fence *fence = NULL;
+	struct amdgpu_bo *bo = NULL;
 	long r;
 
-	r = uvd_v6_0_enc_get_create_msg(ring, 1, NULL);
+	r = amdgpu_bo_create_reserved(ring->adev, 128 * 1024, PAGE_SIZE,
+				      AMDGPU_GEM_DOMAIN_VRAM,
+				      &bo, NULL, NULL);
+	if (r)
+		return r;
+
+	r = uvd_v6_0_enc_get_create_msg(ring, 1, bo, NULL);
 	if (r)
 		goto error;
 
-	r = uvd_v6_0_enc_get_destroy_msg(ring, 1, &fence);
+	r = uvd_v6_0_enc_get_destroy_msg(ring, 1, bo, &fence);
 	if (r)
 		goto error;
 
@@ -345,6 +354,8 @@ static int uvd_v6_0_enc_ring_test_ib(struct amdgpu_ring *ring, long timeout)
 
 error:
 	dma_fence_put(fence);
+	amdgpu_bo_unreserve(bo);
+	amdgpu_bo_unref(&bo);
 	return r;
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
index 01f658fa72c6..0995378d8263 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
@@ -214,13 +214,14 @@ static int uvd_v7_0_enc_ring_test_ring(struct amdgpu_ring *ring)
  * Open up a stream for HW test
  */
 static int uvd_v7_0_enc_get_create_msg(struct amdgpu_ring *ring, uint32_t handle,
+				       struct amdgpu_bo *bo,
 				       struct dma_fence **fence)
 {
 	const unsigned ib_size_dw = 16;
 	struct amdgpu_job *job;
 	struct amdgpu_ib *ib;
 	struct dma_fence *f = NULL;
-	uint64_t dummy;
+	uint64_t addr;
 	int i, r;
 
 	r = amdgpu_job_alloc_with_ib(ring->adev, ib_size_dw * 4, &job);
@@ -228,15 +229,15 @@ static int uvd_v7_0_enc_get_create_msg(struct amdgpu_ring *ring, uint32_t handle
 		return r;
 
 	ib = &job->ibs[0];
-	dummy = ib->gpu_addr + 1024;
+	addr = amdgpu_bo_gpu_offset(bo);
 
 	ib->length_dw = 0;
 	ib->ptr[ib->length_dw++] = 0x00000018;
 	ib->ptr[ib->length_dw++] = 0x00000001; /* session info */
 	ib->ptr[ib->length_dw++] = handle;
 	ib->ptr[ib->length_dw++] = 0x00000000;
-	ib->ptr[ib->length_dw++] = upper_32_bits(dummy);
-	ib->ptr[ib->length_dw++] = dummy;
+	ib->ptr[ib->length_dw++] = upper_32_bits(addr);
+	ib->ptr[ib->length_dw++] = addr;
 
 	ib->ptr[ib->length_dw++] = 0x00000014;
 	ib->ptr[ib->length_dw++] = 0x00000002; /* task info */
@@ -275,13 +276,14 @@ err:
  * Close up a stream for HW test or if userspace failed to do so
  */
 static int uvd_v7_0_enc_get_destroy_msg(struct amdgpu_ring *ring, uint32_t handle,
-				struct dma_fence **fence)
+					struct amdgpu_bo *bo,
+					struct dma_fence **fence)
 {
 	const unsigned ib_size_dw = 16;
 	struct amdgpu_job *job;
 	struct amdgpu_ib *ib;
 	struct dma_fence *f = NULL;
-	uint64_t dummy;
+	uint64_t addr;
 	int i, r;
 
 	r = amdgpu_job_alloc_with_ib(ring->adev, ib_size_dw * 4, &job);
@@ -289,15 +291,15 @@ static int uvd_v7_0_enc_get_destroy_msg(struct amdgpu_ring *ring, uint32_t handl
 		return r;
 
 	ib = &job->ibs[0];
-	dummy = ib->gpu_addr + 1024;
+	addr = amdgpu_bo_gpu_offset(bo);
 
 	ib->length_dw = 0;
 	ib->ptr[ib->length_dw++] = 0x00000018;
 	ib->ptr[ib->length_dw++] = 0x00000001;
 	ib->ptr[ib->length_dw++] = handle;
 	ib->ptr[ib->length_dw++] = 0x00000000;
-	ib->ptr[ib->length_dw++] = upper_32_bits(dummy);
-	ib->ptr[ib->length_dw++] = dummy;
+	ib->ptr[ib->length_dw++] = upper_32_bits(addr);
+	ib->ptr[ib->length_dw++] = addr;
 
 	ib->ptr[ib->length_dw++] = 0x00000014;
 	ib->ptr[ib->length_dw++] = 0x00000002;
@@ -334,13 +336,20 @@ err:
 static int uvd_v7_0_enc_ring_test_ib(struct amdgpu_ring *ring, long timeout)
 {
 	struct dma_fence *fence = NULL;
+	struct amdgpu_bo *bo = NULL;
 	long r;
 
-	r = uvd_v7_0_enc_get_create_msg(ring, 1, NULL);
+	r = amdgpu_bo_create_reserved(ring->adev, 128 * 1024, PAGE_SIZE,
+				      AMDGPU_GEM_DOMAIN_VRAM,
+				      &bo, NULL, NULL);
+	if (r)
+		return r;
+
+	r = uvd_v7_0_enc_get_create_msg(ring, 1, bo, NULL);
 	if (r)
 		goto error;
 
-	r = uvd_v7_0_enc_get_destroy_msg(ring, 1, &fence);
+	r = uvd_v7_0_enc_get_destroy_msg(ring, 1, bo, &fence);
 	if (r)
 		goto error;
 
@@ -352,6 +361,8 @@ static int uvd_v7_0_enc_ring_test_ib(struct amdgpu_ring *ring, long timeout)
 
 error:
 	dma_fence_put(fence);
+	amdgpu_bo_unreserve(bo);
+	amdgpu_bo_unref(&bo);
 	return r;
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c
index 93b3500e522b..b4f84a820a44 100644
--- a/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c
@@ -202,7 +202,6 @@ static int vcn_v1_0_hw_init(void *handle)
 
 	for (i = 0; i < adev->vcn.num_enc_rings; ++i) {
 		ring = &adev->vcn.inst->ring_enc[i];
-		ring->sched.ready = true;
 		r = amdgpu_ring_test_helper(ring);
 		if (r)
 			goto done;
diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v2_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v2_0.c
index 36ad0c0e8efb..38f787a560cb 100644
--- a/drivers/gpu/drm/amd/amdgpu/vcn_v2_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/vcn_v2_0.c
@@ -244,33 +244,24 @@ static int vcn_v2_0_hw_init(void *handle)
 	struct amdgpu_ring *ring = &adev->vcn.inst->ring_dec;
 	int i, r;
 
-	adev->nbio_funcs->vcn_doorbell_range(adev, ring->use_doorbell,
+	adev->nbio.funcs->vcn_doorbell_range(adev, ring->use_doorbell,
 					     ring->doorbell_index, 0);
 
-	ring->sched.ready = true;
-	r = amdgpu_ring_test_ring(ring);
-	if (r) {
-		ring->sched.ready = false;
+	r = amdgpu_ring_test_helper(ring);
+	if (r)
 		goto done;
-	}
 
 	for (i = 0; i < adev->vcn.num_enc_rings; ++i) {
 		ring = &adev->vcn.inst->ring_enc[i];
-		ring->sched.ready = true;
-		r = amdgpu_ring_test_ring(ring);
-		if (r) {
-			ring->sched.ready = false;
+		r = amdgpu_ring_test_helper(ring);
+		if (r)
 			goto done;
-		}
 	}
 
 	ring = &adev->vcn.inst->ring_jpeg;
-	ring->sched.ready = true;
-	r = amdgpu_ring_test_ring(ring);
-	if (r) {
-		ring->sched.ready = false;
+	r = amdgpu_ring_test_helper(ring);
+	if (r)
 		goto done;
-	}
 
 done:
 	if (!r)
diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c b/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c
index 395c2259f979..93edf9193a7b 100644
--- a/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c
+++ b/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c
@@ -25,6 +25,7 @@
 
 #include "amdgpu.h"
 #include "amdgpu_vcn.h"
+#include "amdgpu_pm.h"
 #include "soc15.h"
 #include "soc15d.h"
 #include "vcn_v2_0.h"
@@ -255,32 +256,24 @@ static int vcn_v2_5_hw_init(void *handle)
 			continue;
 		ring = &adev->vcn.inst[j].ring_dec;
 
-		adev->nbio_funcs->vcn_doorbell_range(adev, ring->use_doorbell,
+		adev->nbio.funcs->vcn_doorbell_range(adev, ring->use_doorbell,
 						     ring->doorbell_index, j);
 
-		r = amdgpu_ring_test_ring(ring);
-		if (r) {
-			ring->sched.ready = false;
+		r = amdgpu_ring_test_helper(ring);
+		if (r)
 			goto done;
-		}
 
 		for (i = 0; i < adev->vcn.num_enc_rings; ++i) {
 			ring = &adev->vcn.inst[j].ring_enc[i];
-			ring->sched.ready = false;
-			continue;
-			r = amdgpu_ring_test_ring(ring);
-			if (r) {
-				ring->sched.ready = false;
+			r = amdgpu_ring_test_helper(ring);
+			if (r)
 				goto done;
-			}
 		}
 
 		ring = &adev->vcn.inst[j].ring_jpeg;
-		r = amdgpu_ring_test_ring(ring);
-		if (r) {
-			ring->sched.ready = false;
+		r = amdgpu_ring_test_helper(ring);
+		if (r)
 			goto done;
-		}
 	}
 done:
 	if (!r)
@@ -300,7 +293,7 @@ static int vcn_v2_5_hw_fini(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 	struct amdgpu_ring *ring;
-	int i;
+	int i, j;
 
 	for (i = 0; i < adev->vcn.num_vcn_inst; ++i) {
 		if (adev->vcn.harvest_config & (1 << i))
@@ -312,8 +305,8 @@ static int vcn_v2_5_hw_fini(void *handle)
 
 		ring->sched.ready = false;
 
-		for (i = 0; i < adev->vcn.num_enc_rings; ++i) {
-			ring = &adev->vcn.inst[i].ring_enc[i];
+		for (j = 0; j < adev->vcn.num_enc_rings; ++j) {
+			ring = &adev->vcn.inst[i].ring_enc[j];
 			ring->sched.ready = false;
 		}
 
@@ -423,7 +416,6 @@ static void vcn_v2_5_mc_resume(struct amdgpu_device *adev)
  * vcn_v2_5_disable_clock_gating - disable VCN clock gating
  *
  * @adev: amdgpu_device pointer
- * @sw: enable SW clock gating
  *
  * Disable clock gating for VCN block
  */
@@ -542,7 +534,6 @@ static void vcn_v2_5_disable_clock_gating(struct amdgpu_device *adev)
  * vcn_v2_5_enable_clock_gating - enable VCN clock gating
  *
  * @adev: amdgpu_device pointer
- * @sw: enable SW clock gating
  *
  * Enable clock gating for VCN block
  */
@@ -716,6 +707,9 @@ static int vcn_v2_5_start(struct amdgpu_device *adev)
 	uint32_t rb_bufsz, tmp;
 	int i, j, k, r;
 
+	if (adev->pm.dpm_enabled)
+		amdgpu_dpm_enable_uvd(adev, true);
+
 	for (i = 0; i < adev->vcn.num_vcn_inst; ++i) {
 		if (adev->vcn.harvest_config & (1 << i))
 			continue;
@@ -946,6 +940,9 @@ static int vcn_v2_5_stop(struct amdgpu_device *adev)
 			~UVD_POWER_STATUS__UVD_POWER_STATUS_MASK);
 	}
 
+	if (adev->pm.dpm_enabled)
+		amdgpu_dpm_enable_uvd(adev, false);
+
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/vega10_ih.c b/drivers/gpu/drm/amd/amdgpu/vega10_ih.c
index 9eae3536ddad..5cb7e231de5f 100644
--- a/drivers/gpu/drm/amd/amdgpu/vega10_ih.c
+++ b/drivers/gpu/drm/amd/amdgpu/vega10_ih.c
@@ -226,7 +226,7 @@ static int vega10_ih_irq_init(struct amdgpu_device *adev)
 	/* disable irqs */
 	vega10_ih_disable_interrupts(adev);
 
-	adev->nbio_funcs->ih_control(adev);
+	adev->nbio.funcs->ih_control(adev);
 
 	ih = &adev->irq.ih;
 	/* Ring Buffer base. [39:8] of 40-bit address of the beginning of the ring buffer*/
@@ -675,10 +675,49 @@ static int vega10_ih_soft_reset(void *handle)
 	return 0;
 }
 
+static void vega10_ih_update_clockgating_state(struct amdgpu_device *adev,
+					       bool enable)
+{
+	uint32_t data, def, field_val;
+
+	if (adev->cg_flags & AMD_CG_SUPPORT_IH_CG) {
+		def = data = RREG32_SOC15(OSSSYS, 0, mmIH_CLK_CTRL);
+		field_val = enable ? 0 : 1;
+		/**
+		 * Vega10 does not have IH_RETRY_INT_CAM_MEM_CLK_SOFT_OVERRIDE
+		 * and IH_BUFFER_MEM_CLK_SOFT_OVERRIDE field.
+		 */
+		if (adev->asic_type > CHIP_VEGA10) {
+			data = REG_SET_FIELD(data, IH_CLK_CTRL,
+				     IH_RETRY_INT_CAM_MEM_CLK_SOFT_OVERRIDE, field_val);
+			data = REG_SET_FIELD(data, IH_CLK_CTRL,
+				     IH_BUFFER_MEM_CLK_SOFT_OVERRIDE, field_val);
+		}
+
+		data = REG_SET_FIELD(data, IH_CLK_CTRL,
+				     DBUS_MUX_CLK_SOFT_OVERRIDE, field_val);
+		data = REG_SET_FIELD(data, IH_CLK_CTRL,
+				     OSSSYS_SHARE_CLK_SOFT_OVERRIDE, field_val);
+		data = REG_SET_FIELD(data, IH_CLK_CTRL,
+				     LIMIT_SMN_CLK_SOFT_OVERRIDE, field_val);
+		data = REG_SET_FIELD(data, IH_CLK_CTRL,
+				     DYN_CLK_SOFT_OVERRIDE, field_val);
+		data = REG_SET_FIELD(data, IH_CLK_CTRL,
+				     REG_CLK_SOFT_OVERRIDE, field_val);
+		if (def != data)
+			WREG32_SOC15(OSSSYS, 0, mmIH_CLK_CTRL, data);
+	}
+}
+
 static int vega10_ih_set_clockgating_state(void *handle,
 					  enum amd_clockgating_state state)
 {
+	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+	vega10_ih_update_clockgating_state(adev,
+				state == AMD_CG_STATE_GATE ? true : false);
 	return 0;
+
 }
 
 static int vega10_ih_set_powergating_state(void *handle,
diff --git a/drivers/gpu/drm/amd/amdgpu/vega10_reg_init.c b/drivers/gpu/drm/amd/amdgpu/vega10_reg_init.c
index bd0580334f83..6b52a539d51b 100644
--- a/drivers/gpu/drm/amd/amdgpu/vega10_reg_init.c
+++ b/drivers/gpu/drm/amd/amdgpu/vega10_reg_init.c
@@ -24,7 +24,6 @@
 #include "soc15.h"
 
 #include "soc15_common.h"
-#include "soc15_hw_ip.h"
 #include "vega10_ip_offset.h"
 
 int vega10_reg_base_init(struct amdgpu_device *adev)
diff --git a/drivers/gpu/drm/amd/amdgpu/vega20_reg_init.c b/drivers/gpu/drm/amd/amdgpu/vega20_reg_init.c
index 587e33f5dcce..556f854e3551 100644
--- a/drivers/gpu/drm/amd/amdgpu/vega20_reg_init.c
+++ b/drivers/gpu/drm/amd/amdgpu/vega20_reg_init.c
@@ -24,7 +24,6 @@
 #include "soc15.h"
 
 #include "soc15_common.h"
-#include "soc15_hw_ip.h"
 #include "vega20_ip_offset.h"
 
 int vega20_reg_base_init(struct amdgpu_device *adev)
diff --git a/drivers/gpu/drm/amd/amdgpu/vi.c b/drivers/gpu/drm/amd/amdgpu/vi.c
index 5f8c8786cac5..78e5cdc0c058 100644
--- a/drivers/gpu/drm/amd/amdgpu/vi.c
+++ b/drivers/gpu/drm/amd/amdgpu/vi.c
@@ -689,16 +689,50 @@ static int vi_gpu_pci_config_reset(struct amdgpu_device *adev)
 	return -EINVAL;
 }
 
+int smu7_asic_get_baco_capability(struct amdgpu_device *adev, bool *cap)
+{
+	void *pp_handle = adev->powerplay.pp_handle;
+	const struct amd_pm_funcs *pp_funcs = adev->powerplay.pp_funcs;
+
+	if (!pp_funcs || !pp_funcs->get_asic_baco_capability) {
+		*cap = false;
+		return -ENOENT;
+	}
+
+	return pp_funcs->get_asic_baco_capability(pp_handle, cap);
+}
+
+int smu7_asic_baco_reset(struct amdgpu_device *adev)
+{
+	void *pp_handle = adev->powerplay.pp_handle;
+	const struct amd_pm_funcs *pp_funcs = adev->powerplay.pp_funcs;
+
+	if (!pp_funcs ||!pp_funcs->get_asic_baco_state ||!pp_funcs->set_asic_baco_state)
+		return -ENOENT;
+
+	/* enter BACO state */
+	if (pp_funcs->set_asic_baco_state(pp_handle, 1))
+		return -EIO;
+
+	/* exit BACO state */
+	if (pp_funcs->set_asic_baco_state(pp_handle, 0))
+		return -EIO;
+
+	dev_info(adev->dev, "GPU BACO reset\n");
+
+	return 0;
+}
+
 /**
- * vi_asic_reset - soft reset GPU
+ * vi_asic_pci_config_reset - soft reset GPU
  *
  * @adev: amdgpu_device pointer
  *
- * Look up which blocks are hung and attempt
- * to reset them.
+ * Use PCI Config method to reset the GPU.
+ *
  * Returns 0 for success.
  */
-static int vi_asic_reset(struct amdgpu_device *adev)
+static int vi_asic_pci_config_reset(struct amdgpu_device *adev)
 {
 	int r;
 
@@ -714,7 +748,47 @@ static int vi_asic_reset(struct amdgpu_device *adev)
 static enum amd_reset_method
 vi_asic_reset_method(struct amdgpu_device *adev)
 {
-	return AMD_RESET_METHOD_LEGACY;
+	bool baco_reset;
+
+	switch (adev->asic_type) {
+	case CHIP_FIJI:
+	case CHIP_TONGA:
+	case CHIP_POLARIS10:
+	case CHIP_POLARIS11:
+	case CHIP_POLARIS12:
+	case CHIP_TOPAZ:
+		smu7_asic_get_baco_capability(adev, &baco_reset);
+		break;
+	default:
+		baco_reset = false;
+		break;
+	}
+
+	if (baco_reset)
+		return AMD_RESET_METHOD_BACO;
+	else
+		return AMD_RESET_METHOD_LEGACY;
+}
+
+/**
+ * vi_asic_reset - soft reset GPU
+ *
+ * @adev: amdgpu_device pointer
+ *
+ * Look up which blocks are hung and attempt
+ * to reset them.
+ * Returns 0 for success.
+ */
+static int vi_asic_reset(struct amdgpu_device *adev)
+{
+	int r;
+
+	if (vi_asic_reset_method(adev) == AMD_RESET_METHOD_BACO)
+		r = smu7_asic_baco_reset(adev);
+	else
+		r = vi_asic_pci_config_reset(adev);
+
+	return r;
 }
 
 static u32 vi_get_config_memsize(struct amdgpu_device *adev)
diff --git a/drivers/gpu/drm/amd/amdgpu/vi.h b/drivers/gpu/drm/amd/amdgpu/vi.h
index 8de0772f986c..40d4174913a4 100644
--- a/drivers/gpu/drm/amd/amdgpu/vi.h
+++ b/drivers/gpu/drm/amd/amdgpu/vi.h
@@ -31,4 +31,7 @@ void vi_srbm_select(struct amdgpu_device *adev,
 int vi_set_ip_blocks(struct amdgpu_device *adev);
 
 void legacy_doorbell_index_init(struct amdgpu_device *adev);
+int smu7_asic_get_baco_capability(struct amdgpu_device *adev, bool *cap);
+int smu7_asic_baco_reset(struct amdgpu_device *adev);
+
 #endif
diff --git a/drivers/gpu/drm/amd/amdkfd/cik_event_interrupt.c b/drivers/gpu/drm/amd/amdkfd/cik_event_interrupt.c
index 177d1e5329a5..9f59ba93cfe0 100644
--- a/drivers/gpu/drm/amd/amdkfd/cik_event_interrupt.c
+++ b/drivers/gpu/drm/amd/amdkfd/cik_event_interrupt.c
@@ -33,7 +33,9 @@ static bool cik_event_interrupt_isr(struct kfd_dev *dev,
 	const struct cik_ih_ring_entry *ihre =
 			(const struct cik_ih_ring_entry *)ih_ring_entry;
 	const struct kfd2kgd_calls *f2g = dev->kfd2kgd;
-	unsigned int vmid, pasid;
+	unsigned int vmid;
+	uint16_t pasid;
+	bool ret;
 
 	/* This workaround is due to HW/FW limitation on Hawaii that
 	 * VMID and PASID are not written into ih_ring_entry
@@ -48,13 +50,13 @@ static bool cik_event_interrupt_isr(struct kfd_dev *dev,
 		*tmp_ihre = *ihre;
 
 		vmid = f2g->read_vmid_from_vmfault_reg(dev->kgd);
-		pasid = f2g->get_atc_vmid_pasid_mapping_pasid(dev->kgd, vmid);
+		ret = f2g->get_atc_vmid_pasid_mapping_info(dev->kgd, vmid, &pasid);
 
 		tmp_ihre->ring_id &= 0x000000ff;
 		tmp_ihre->ring_id |= vmid << 8;
 		tmp_ihre->ring_id |= pasid << 16;
 
-		return (pasid != 0) &&
+		return ret && (pasid != 0) &&
 			vmid >= dev->vm_info.first_vmid_kfd &&
 			vmid <= dev->vm_info.last_vmid_kfd;
 	}
diff --git a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler.h b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler.h
index 901fe3590165..d3400da6ab64 100644
--- a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler.h
+++ b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler.h
@@ -905,7 +905,7 @@ static const uint32_t cwsr_trap_gfx10_hex[] = {
 	0x7a5d0000, 0x807c817c,
 	0x807aff7a, 0x00000080,
 	0xbf0a717c, 0xbf85fff8,
-	0xbf820141, 0xbef4037e,
+	0xbf820142, 0xbef4037e,
 	0x8775ff7f, 0x0000ffff,
 	0x8875ff75, 0x00040000,
 	0xbef60380, 0xbef703ff,
@@ -967,7 +967,7 @@ static const uint32_t cwsr_trap_gfx10_hex[] = {
 	0x725d0000, 0xe0304080,
 	0x725d0100, 0xe0304100,
 	0x725d0200, 0xe0304180,
-	0x725d0300, 0xbf820031,
+	0x725d0300, 0xbf820032,
 	0xbef603ff, 0x01000000,
 	0xbef20378, 0x8078ff78,
 	0x00000400, 0xbefc0384,
@@ -992,83 +992,84 @@ static const uint32_t cwsr_trap_gfx10_hex[] = {
 	0x725d0000, 0xe0304100,
 	0x725d0100, 0xe0304200,
 	0x725d0200, 0xe0304300,
-	0x725d0300, 0xb9782a05,
-	0x80788178, 0x907c9973,
-	0x877c817c, 0xbf06817c,
-	0xbf850002, 0x8f788978,
-	0xbf820001, 0x8f788a78,
-	0xb9721e06, 0x8f728a72,
-	0x80787278, 0x8078ff78,
-	0x00000200, 0x80f8ff78,
-	0x00000050, 0xbef603ff,
-	0x01000000, 0xbefc03ff,
-	0x0000006c, 0x80f89078,
-	0xf429003a, 0xf0000000,
-	0xbf8cc07f, 0x80fc847c,
-	0xbf800000, 0xbe803100,
-	0xbe823102, 0x80f8a078,
-	0xf42d003a, 0xf0000000,
-	0xbf8cc07f, 0x80fc887c,
-	0xbf800000, 0xbe803100,
-	0xbe823102, 0xbe843104,
-	0xbe863106, 0x80f8c078,
-	0xf431003a, 0xf0000000,
-	0xbf8cc07f, 0x80fc907c,
-	0xbf800000, 0xbe803100,
-	0xbe823102, 0xbe843104,
-	0xbe863106, 0xbe883108,
-	0xbe8a310a, 0xbe8c310c,
-	0xbe8e310e, 0xbf06807c,
-	0xbf84fff0, 0xb9782a05,
-	0x80788178, 0x907c9973,
-	0x877c817c, 0xbf06817c,
-	0xbf850002, 0x8f788978,
-	0xbf820001, 0x8f788a78,
-	0xb9721e06, 0x8f728a72,
-	0x80787278, 0x8078ff78,
-	0x00000200, 0xbef603ff,
-	0x01000000, 0xf4211bfa,
+	0x725d0300, 0xbf8c3f70,
+	0xb9782a05, 0x80788178,
+	0x907c9973, 0x877c817c,
+	0xbf06817c, 0xbf850002,
+	0x8f788978, 0xbf820001,
+	0x8f788a78, 0xb9721e06,
+	0x8f728a72, 0x80787278,
+	0x8078ff78, 0x00000200,
+	0x80f8ff78, 0x00000050,
+	0xbef603ff, 0x01000000,
+	0xbefc03ff, 0x0000006c,
+	0x80f89078, 0xf429003a,
+	0xf0000000, 0xbf8cc07f,
+	0x80fc847c, 0xbf800000,
+	0xbe803100, 0xbe823102,
+	0x80f8a078, 0xf42d003a,
+	0xf0000000, 0xbf8cc07f,
+	0x80fc887c, 0xbf800000,
+	0xbe803100, 0xbe823102,
+	0xbe843104, 0xbe863106,
+	0x80f8c078, 0xf431003a,
+	0xf0000000, 0xbf8cc07f,
+	0x80fc907c, 0xbf800000,
+	0xbe803100, 0xbe823102,
+	0xbe843104, 0xbe863106,
+	0xbe883108, 0xbe8a310a,
+	0xbe8c310c, 0xbe8e310e,
+	0xbf06807c, 0xbf84fff0,
+	0xb9782a05, 0x80788178,
+	0x907c9973, 0x877c817c,
+	0xbf06817c, 0xbf850002,
+	0x8f788978, 0xbf820001,
+	0x8f788a78, 0xb9721e06,
+	0x8f728a72, 0x80787278,
+	0x8078ff78, 0x00000200,
+	0xbef603ff, 0x01000000,
+	0xf4211bfa, 0xf0000000,
+	0x80788478, 0xf4211b3a,
 	0xf0000000, 0x80788478,
-	0xf4211b3a, 0xf0000000,
-	0x80788478, 0xf4211b7a,
+	0xf4211b7a, 0xf0000000,
+	0x80788478, 0xf4211eba,
 	0xf0000000, 0x80788478,
-	0xf4211eba, 0xf0000000,
-	0x80788478, 0xf4211efa,
+	0xf4211efa, 0xf0000000,
+	0x80788478, 0xf4211c3a,
 	0xf0000000, 0x80788478,
-	0xf4211c3a, 0xf0000000,
-	0x80788478, 0xf4211c7a,
+	0xf4211c7a, 0xf0000000,
+	0x80788478, 0xf4211e7a,
 	0xf0000000, 0x80788478,
-	0xf4211e7a, 0xf0000000,
-	0x80788478, 0xf4211cfa,
+	0xf4211cfa, 0xf0000000,
+	0x80788478, 0xf4211bba,
 	0xf0000000, 0x80788478,
+	0xbf8cc07f, 0xb9eef814,
 	0xf4211bba, 0xf0000000,
 	0x80788478, 0xbf8cc07f,
-	0xb9eef814, 0xf4211bba,
-	0xf0000000, 0x80788478,
-	0xbf8cc07f, 0xb9eef815,
-	0xbef2036d, 0x876dff72,
-	0x0000ffff, 0xbefc036f,
-	0xbefe037a, 0xbeff037b,
-	0x876f71ff, 0x000003ff,
-	0xb9ef4803, 0xb9f9f816,
-	0x876f71ff, 0xfffff800,
-	0x906f8b6f, 0xb9efa2c3,
-	0xb9f3f801, 0x876fff72,
-	0xfc000000, 0x906f9a6f,
-	0x8f6f906f, 0xbef30380,
+	0xb9eef815, 0xbef2036d,
+	0x876dff72, 0x0000ffff,
+	0xbefc036f, 0xbefe037a,
+	0xbeff037b, 0x876f71ff,
+	0x000003ff, 0xb9ef4803,
+	0xb9f9f816, 0x876f71ff,
+	0xfffff800, 0x906f8b6f,
+	0xb9efa2c3, 0xb9f3f801,
+	0x876fff72, 0xfc000000,
+	0x906f9a6f, 0x8f6f906f,
+	0xbef30380, 0x88736f73,
+	0x876fff72, 0x02000000,
+	0x906f996f, 0x8f6f8f6f,
 	0x88736f73, 0x876fff72,
-	0x02000000, 0x906f996f,
-	0x8f6f8f6f, 0x88736f73,
-	0x876fff72, 0x01000000,
-	0x906f986f, 0x8f6f996f,
-	0x88736f73, 0x876fff70,
-	0x00800000, 0x906f976f,
-	0xb9f3f807, 0x87fe7e7e,
-	0x87ea6a6a, 0xb9f0f802,
-	0xbf8a0000, 0xbe80226c,
-	0xbf810000, 0xbf9f0000,
+	0x01000000, 0x906f986f,
+	0x8f6f996f, 0x88736f73,
+	0x876fff70, 0x00800000,
+	0x906f976f, 0xb9f3f807,
+	0x87fe7e7e, 0x87ea6a6a,
+	0xb9f0f802, 0xbf8a0000,
+	0xbe80226c, 0xbf810000,
 	0xbf9f0000, 0xbf9f0000,
 	0xbf9f0000, 0xbf9f0000,
+	0xbf9f0000, 0x00000000,
 };
 static const uint32_t cwsr_trap_arcturus_hex[] = {
 	0xbf820001, 0xbf8202c4,
diff --git a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx10.asm b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx10.asm
index cdaa523ce6be..4433bda2ce25 100644
--- a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx10.asm
+++ b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx10.asm
@@ -758,6 +758,7 @@ L_RESTORE_V0:
 	buffer_load_dword	v1, v0, s_restore_buf_rsrc0, s_restore_mem_offset_save slc:1 glc:1 offset:256
 	buffer_load_dword	v2, v0, s_restore_buf_rsrc0, s_restore_mem_offset_save slc:1 glc:1 offset:256*2
 	buffer_load_dword	v3, v0, s_restore_buf_rsrc0, s_restore_mem_offset_save slc:1 glc:1 offset:256*3
+	s_waitcnt	vmcnt(0)
 
 	/* restore SGPRs */
 	//will be 2+8+16*6
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
index 1d3cd5c50d5f..1544007af34a 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
@@ -49,7 +49,7 @@ static const char kfd_dev_name[] = "kfd";
 static const struct file_operations kfd_fops = {
 	.owner = THIS_MODULE,
 	.unlocked_ioctl = kfd_ioctl,
-	.compat_ioctl = kfd_ioctl,
+	.compat_ioctl = compat_ptr_ioctl,
 	.open = kfd_open,
 	.mmap = kfd_mmap,
 };
@@ -282,7 +282,7 @@ static int kfd_ioctl_create_queue(struct file *filep, struct kfd_process *p,
 		goto err_bind_process;
 	}
 
-	pr_debug("Creating queue for PASID %d on gpu 0x%x\n",
+	pr_debug("Creating queue for PASID 0x%x on gpu 0x%x\n",
 			p->pasid,
 			dev->id);
 
@@ -332,7 +332,7 @@ static int kfd_ioctl_destroy_queue(struct file *filp, struct kfd_process *p,
 	int retval;
 	struct kfd_ioctl_destroy_queue_args *args = data;
 
-	pr_debug("Destroying queue id %d for pasid %d\n",
+	pr_debug("Destroying queue id %d for pasid 0x%x\n",
 				args->queue_id,
 				p->pasid);
 
@@ -378,7 +378,7 @@ static int kfd_ioctl_update_queue(struct file *filp, struct kfd_process *p,
 	properties.queue_percent = args->queue_percentage;
 	properties.priority = args->queue_priority;
 
-	pr_debug("Updating queue id %d for pasid %d\n",
+	pr_debug("Updating queue id %d for pasid 0x%x\n",
 			args->queue_id, p->pasid);
 
 	mutex_lock(&p->mutex);
@@ -855,7 +855,7 @@ static int kfd_ioctl_get_process_apertures(struct file *filp,
 	struct kfd_process_device_apertures *pAperture;
 	struct kfd_process_device *pdd;
 
-	dev_dbg(kfd_device, "get apertures for PASID %d", p->pasid);
+	dev_dbg(kfd_device, "get apertures for PASID 0x%x", p->pasid);
 
 	args->num_of_nodes = 0;
 
@@ -913,7 +913,7 @@ static int kfd_ioctl_get_process_apertures_new(struct file *filp,
 	uint32_t nodes = 0;
 	int ret;
 
-	dev_dbg(kfd_device, "get apertures for PASID %d", p->pasid);
+	dev_dbg(kfd_device, "get apertures for PASID 0x%x", p->pasid);
 
 	if (args->num_of_nodes == 0) {
 		/* Return number of nodes, so that user space can alloacate
@@ -1128,7 +1128,7 @@ static int kfd_ioctl_set_scratch_backing_va(struct file *filep,
 	mutex_unlock(&p->mutex);
 
 	if (dev->dqm->sched_policy == KFD_SCHED_POLICY_NO_HWS &&
-	    pdd->qpd.vmid != 0)
+	    pdd->qpd.vmid != 0 && dev->kfd2kgd->set_scratch_backing_va)
 		dev->kfd2kgd->set_scratch_backing_va(
 			dev->kgd, args->va_addr, pdd->qpd.vmid);
 
@@ -1801,7 +1801,7 @@ static long kfd_ioctl(struct file *filep, unsigned int cmd, unsigned long arg)
 	} else
 		goto err_i1;
 
-	dev_dbg(kfd_device, "ioctl cmd 0x%x (#%d), arg 0x%lx\n", cmd, nr, arg);
+	dev_dbg(kfd_device, "ioctl cmd 0x%x (#0x%x), arg 0x%lx\n", cmd, nr, arg);
 
 	process = kfd_get_process(current);
 	if (IS_ERR(process)) {
@@ -1856,7 +1856,8 @@ err_i1:
 		kfree(kdata);
 
 	if (retcode)
-		dev_dbg(kfd_device, "ret = %d\n", retcode);
+		dev_dbg(kfd_device, "ioctl cmd (#0x%x), arg 0x%lx, ret = %d\n",
+				nr, arg, retcode);
 
 	return retcode;
 }
@@ -1877,7 +1878,7 @@ static int kfd_mmio_mmap(struct kfd_dev *dev, struct kfd_process *process,
 
 	vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot);
 
-	pr_debug("Process %d mapping mmio page\n"
+	pr_debug("pasid 0x%x mapping mmio page\n"
 		 "     target user address == 0x%08llX\n"
 		 "     physical address    == 0x%08llX\n"
 		 "     vm_flags            == 0x%04lX\n"
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_crat.c b/drivers/gpu/drm/amd/amdkfd/kfd_crat.c
index 66387caf966e..de9f68d5c312 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_crat.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_crat.c
@@ -138,6 +138,7 @@ static struct kfd_gpu_cache_info carrizo_cache_info[] = {
 /* TODO - check & update Vega10 cache details */
 #define vega10_cache_info carrizo_cache_info
 #define raven_cache_info carrizo_cache_info
+#define renoir_cache_info carrizo_cache_info
 /* TODO - check & update Navi10 cache details */
 #define navi10_cache_info carrizo_cache_info
 
@@ -670,7 +671,13 @@ static int kfd_fill_gpu_cache_info(struct kfd_dev *kdev,
 		pcache_info = raven_cache_info;
 		num_of_cache_types = ARRAY_SIZE(raven_cache_info);
 		break;
+	case CHIP_RENOIR:
+		pcache_info = renoir_cache_info;
+		num_of_cache_types = ARRAY_SIZE(renoir_cache_info);
+		break;
 	case CHIP_NAVI10:
+	case CHIP_NAVI12:
+	case CHIP_NAVI14:
 		pcache_info = navi10_cache_info;
 		num_of_cache_types = ARRAY_SIZE(navi10_cache_info);
 		break;
@@ -703,7 +710,7 @@ static int kfd_fill_gpu_cache_info(struct kfd_dev *kdev,
 						pcache_info,
 						cu_info,
 						mem_available,
-						cu_info->cu_bitmap[i][j],
+						cu_info->cu_bitmap[i % 4][j + i / 4],
 						ct,
 						cu_processor_id,
 						k);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_dbgdev.c b/drivers/gpu/drm/amd/amdkfd/kfd_dbgdev.c
index a3441b0e385b..d59f2cd056c6 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_dbgdev.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_dbgdev.c
@@ -761,6 +761,7 @@ int dbgdev_wave_reset_wavefronts(struct kfd_dev *dev, struct kfd_process *p)
 {
 	int status = 0;
 	unsigned int vmid;
+	uint16_t queried_pasid;
 	union SQ_CMD_BITS reg_sq_cmd;
 	union GRBM_GFX_INDEX_BITS reg_gfx_index;
 	struct kfd_process_device *pdd;
@@ -782,19 +783,18 @@ int dbgdev_wave_reset_wavefronts(struct kfd_dev *dev, struct kfd_process *p)
 	 */
 
 	for (vmid = first_vmid_to_scan; vmid <= last_vmid_to_scan; vmid++) {
-		if (dev->kfd2kgd->get_atc_vmid_pasid_mapping_valid
-				(dev->kgd, vmid)) {
-			if (dev->kfd2kgd->get_atc_vmid_pasid_mapping_pasid
-					(dev->kgd, vmid) == p->pasid) {
-				pr_debug("Killing wave fronts of vmid %d and pasid %d\n",
-						vmid, p->pasid);
-				break;
-			}
+		status = dev->kfd2kgd->get_atc_vmid_pasid_mapping_info
+				(dev->kgd, vmid, &queried_pasid);
+
+		if (status && queried_pasid == p->pasid) {
+			pr_debug("Killing wave fronts of vmid %d and pasid 0x%x\n",
+					vmid, p->pasid);
+			break;
 		}
 	}
 
 	if (vmid > last_vmid_to_scan) {
-		pr_err("Didn't find vmid for pasid %d\n", p->pasid);
+		pr_err("Didn't find vmid for pasid 0x%x\n", p->pasid);
 		return -EFAULT;
 	}
 
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_dbgmgr.c b/drivers/gpu/drm/amd/amdkfd/kfd_dbgmgr.c
index 9d4af961c5d1..9bfa50633654 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_dbgmgr.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_dbgmgr.c
@@ -96,7 +96,7 @@ bool kfd_dbgmgr_create(struct kfd_dbgmgr **ppmgr, struct kfd_dev *pdev)
 long kfd_dbgmgr_register(struct kfd_dbgmgr *pmgr, struct kfd_process *p)
 {
 	if (pmgr->pasid != 0) {
-		pr_debug("H/W debugger is already active using pasid %d\n",
+		pr_debug("H/W debugger is already active using pasid 0x%x\n",
 				pmgr->pasid);
 		return -EBUSY;
 	}
@@ -117,7 +117,7 @@ long kfd_dbgmgr_unregister(struct kfd_dbgmgr *pmgr, struct kfd_process *p)
 {
 	/* Is the requests coming from the already registered process? */
 	if (pmgr->pasid != p->pasid) {
-		pr_debug("H/W debugger is not registered by calling pasid %d\n",
+		pr_debug("H/W debugger is not registered by calling pasid 0x%x\n",
 				p->pasid);
 		return -EINVAL;
 	}
@@ -134,7 +134,7 @@ long kfd_dbgmgr_wave_control(struct kfd_dbgmgr *pmgr,
 {
 	/* Is the requests coming from the already registered process? */
 	if (pmgr->pasid != wac_info->process->pasid) {
-		pr_debug("H/W debugger support was not registered for requester pasid %d\n",
+		pr_debug("H/W debugger support was not registered for requester pasid 0x%x\n",
 				wac_info->process->pasid);
 		return -EINVAL;
 	}
@@ -147,7 +147,7 @@ long kfd_dbgmgr_address_watch(struct kfd_dbgmgr *pmgr,
 {
 	/* Is the requests coming from the already registered process? */
 	if (pmgr->pasid != adw_info->process->pasid) {
-		pr_debug("H/W debugger support was not registered for requester pasid %d\n",
+		pr_debug("H/W debugger support was not registered for requester pasid 0x%x\n",
 				adw_info->process->pasid);
 		return -EINVAL;
 	}
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device.c b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
index 0dc1084b5e82..4fa8834ce7cb 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
@@ -39,6 +39,41 @@
  */
 static atomic_t kfd_locked = ATOMIC_INIT(0);
 
+#ifdef CONFIG_DRM_AMDGPU_CIK
+extern const struct kfd2kgd_calls gfx_v7_kfd2kgd;
+#endif
+extern const struct kfd2kgd_calls gfx_v8_kfd2kgd;
+extern const struct kfd2kgd_calls gfx_v9_kfd2kgd;
+extern const struct kfd2kgd_calls arcturus_kfd2kgd;
+extern const struct kfd2kgd_calls gfx_v10_kfd2kgd;
+
+static const struct kfd2kgd_calls *kfd2kgd_funcs[] = {
+#ifdef KFD_SUPPORT_IOMMU_V2
+#ifdef CONFIG_DRM_AMDGPU_CIK
+	[CHIP_KAVERI] = &gfx_v7_kfd2kgd,
+#endif
+	[CHIP_CARRIZO] = &gfx_v8_kfd2kgd,
+	[CHIP_RAVEN] = &gfx_v9_kfd2kgd,
+#endif
+#ifdef CONFIG_DRM_AMDGPU_CIK
+	[CHIP_HAWAII] = &gfx_v7_kfd2kgd,
+#endif
+	[CHIP_TONGA] = &gfx_v8_kfd2kgd,
+	[CHIP_FIJI] = &gfx_v8_kfd2kgd,
+	[CHIP_POLARIS10] = &gfx_v8_kfd2kgd,
+	[CHIP_POLARIS11] = &gfx_v8_kfd2kgd,
+	[CHIP_POLARIS12] = &gfx_v8_kfd2kgd,
+	[CHIP_VEGAM] = &gfx_v8_kfd2kgd,
+	[CHIP_VEGA10] = &gfx_v9_kfd2kgd,
+	[CHIP_VEGA12] = &gfx_v9_kfd2kgd,
+	[CHIP_VEGA20] = &gfx_v9_kfd2kgd,
+	[CHIP_RENOIR] = &gfx_v9_kfd2kgd,
+	[CHIP_ARCTURUS] = &arcturus_kfd2kgd,
+	[CHIP_NAVI10] = &gfx_v10_kfd2kgd,
+	[CHIP_NAVI12] = &gfx_v10_kfd2kgd,
+	[CHIP_NAVI14] = &gfx_v10_kfd2kgd,
+};
+
 #ifdef KFD_SUPPORT_IOMMU_V2
 static const struct kfd_device_info kaveri_device_info = {
 	.asic_family = CHIP_KAVERI,
@@ -351,6 +386,24 @@ static const struct kfd_device_info arcturus_device_info = {
 	.num_sdma_queues_per_engine = 8,
 };
 
+static const struct kfd_device_info renoir_device_info = {
+	.asic_family = CHIP_RENOIR,
+	.asic_name = "renoir",
+	.max_pasid_bits = 16,
+	.max_no_of_hqd  = 24,
+	.doorbell_size  = 8,
+	.ih_ring_entry_size = 8 * sizeof(uint32_t),
+	.event_interrupt_class = &event_interrupt_class_v9,
+	.num_of_watch_points = 4,
+	.mqd_size_aligned = MQD_SIZE_ALIGNED,
+	.supports_cwsr = true,
+	.needs_iommu_device = false,
+	.needs_pci_atomics = false,
+	.num_sdma_engines = 1,
+	.num_xgmi_sdma_engines = 0,
+	.num_sdma_queues_per_engine = 2,
+};
+
 static const struct kfd_device_info navi10_device_info = {
 	.asic_family = CHIP_NAVI10,
 	.asic_name = "navi10",
@@ -369,133 +422,64 @@ static const struct kfd_device_info navi10_device_info = {
 	.num_sdma_queues_per_engine = 8,
 };
 
-struct kfd_deviceid {
-	unsigned short did;
-	const struct kfd_device_info *device_info;
+static const struct kfd_device_info navi12_device_info = {
+	.asic_family = CHIP_NAVI12,
+	.asic_name = "navi12",
+	.max_pasid_bits = 16,
+	.max_no_of_hqd  = 24,
+	.doorbell_size  = 8,
+	.ih_ring_entry_size = 8 * sizeof(uint32_t),
+	.event_interrupt_class = &event_interrupt_class_v9,
+	.num_of_watch_points = 4,
+	.mqd_size_aligned = MQD_SIZE_ALIGNED,
+	.needs_iommu_device = false,
+	.supports_cwsr = true,
+	.needs_pci_atomics = false,
+	.num_sdma_engines = 2,
+	.num_xgmi_sdma_engines = 0,
+	.num_sdma_queues_per_engine = 8,
+};
+
+static const struct kfd_device_info navi14_device_info = {
+	.asic_family = CHIP_NAVI14,
+	.asic_name = "navi14",
+	.max_pasid_bits = 16,
+	.max_no_of_hqd  = 24,
+	.doorbell_size  = 8,
+	.ih_ring_entry_size = 8 * sizeof(uint32_t),
+	.event_interrupt_class = &event_interrupt_class_v9,
+	.num_of_watch_points = 4,
+	.mqd_size_aligned = MQD_SIZE_ALIGNED,
+	.needs_iommu_device = false,
+	.supports_cwsr = true,
+	.needs_pci_atomics = false,
+	.num_sdma_engines = 2,
+	.num_xgmi_sdma_engines = 0,
+	.num_sdma_queues_per_engine = 8,
 };
 
-static const struct kfd_deviceid supported_devices[] = {
+/* For each entry, [0] is regular and [1] is virtualisation device. */
+static const struct kfd_device_info *kfd_supported_devices[][2] = {
 #ifdef KFD_SUPPORT_IOMMU_V2
-	{ 0x1304, &kaveri_device_info },	/* Kaveri */
-	{ 0x1305, &kaveri_device_info },	/* Kaveri */
-	{ 0x1306, &kaveri_device_info },	/* Kaveri */
-	{ 0x1307, &kaveri_device_info },	/* Kaveri */
-	{ 0x1309, &kaveri_device_info },	/* Kaveri */
-	{ 0x130A, &kaveri_device_info },	/* Kaveri */
-	{ 0x130B, &kaveri_device_info },	/* Kaveri */
-	{ 0x130C, &kaveri_device_info },	/* Kaveri */
-	{ 0x130D, &kaveri_device_info },	/* Kaveri */
-	{ 0x130E, &kaveri_device_info },	/* Kaveri */
-	{ 0x130F, &kaveri_device_info },	/* Kaveri */
-	{ 0x1310, &kaveri_device_info },	/* Kaveri */
-	{ 0x1311, &kaveri_device_info },	/* Kaveri */
-	{ 0x1312, &kaveri_device_info },	/* Kaveri */
-	{ 0x1313, &kaveri_device_info },	/* Kaveri */
-	{ 0x1315, &kaveri_device_info },	/* Kaveri */
-	{ 0x1316, &kaveri_device_info },	/* Kaveri */
-	{ 0x1317, &kaveri_device_info },	/* Kaveri */
-	{ 0x1318, &kaveri_device_info },	/* Kaveri */
-	{ 0x131B, &kaveri_device_info },	/* Kaveri */
-	{ 0x131C, &kaveri_device_info },	/* Kaveri */
-	{ 0x131D, &kaveri_device_info },	/* Kaveri */
-	{ 0x9870, &carrizo_device_info },	/* Carrizo */
-	{ 0x9874, &carrizo_device_info },	/* Carrizo */
-	{ 0x9875, &carrizo_device_info },	/* Carrizo */
-	{ 0x9876, &carrizo_device_info },	/* Carrizo */
-	{ 0x9877, &carrizo_device_info },	/* Carrizo */
-	{ 0x15DD, &raven_device_info },		/* Raven */
-	{ 0x15D8, &raven_device_info },		/* Raven */
+	[CHIP_KAVERI] = {&kaveri_device_info, NULL},
+	[CHIP_CARRIZO] = {&carrizo_device_info, NULL},
+	[CHIP_RAVEN] = {&raven_device_info, NULL},
 #endif
-	{ 0x67A0, &hawaii_device_info },	/* Hawaii */
-	{ 0x67A1, &hawaii_device_info },	/* Hawaii */
-	{ 0x67A2, &hawaii_device_info },	/* Hawaii */
-	{ 0x67A8, &hawaii_device_info },	/* Hawaii */
-	{ 0x67A9, &hawaii_device_info },	/* Hawaii */
-	{ 0x67AA, &hawaii_device_info },	/* Hawaii */
-	{ 0x67B0, &hawaii_device_info },	/* Hawaii */
-	{ 0x67B1, &hawaii_device_info },	/* Hawaii */
-	{ 0x67B8, &hawaii_device_info },	/* Hawaii */
-	{ 0x67B9, &hawaii_device_info },	/* Hawaii */
-	{ 0x67BA, &hawaii_device_info },	/* Hawaii */
-	{ 0x67BE, &hawaii_device_info },	/* Hawaii */
-	{ 0x6920, &tonga_device_info },		/* Tonga */
-	{ 0x6921, &tonga_device_info },		/* Tonga */
-	{ 0x6928, &tonga_device_info },		/* Tonga */
-	{ 0x6929, &tonga_device_info },		/* Tonga */
-	{ 0x692B, &tonga_device_info },		/* Tonga */
-	{ 0x6938, &tonga_device_info },		/* Tonga */
-	{ 0x6939, &tonga_device_info },		/* Tonga */
-	{ 0x7300, &fiji_device_info },		/* Fiji */
-	{ 0x730F, &fiji_vf_device_info },	/* Fiji vf*/
-	{ 0x67C0, &polaris10_device_info },	/* Polaris10 */
-	{ 0x67C1, &polaris10_device_info },	/* Polaris10 */
-	{ 0x67C2, &polaris10_device_info },	/* Polaris10 */
-	{ 0x67C4, &polaris10_device_info },	/* Polaris10 */
-	{ 0x67C7, &polaris10_device_info },	/* Polaris10 */
-	{ 0x67C8, &polaris10_device_info },	/* Polaris10 */
-	{ 0x67C9, &polaris10_device_info },	/* Polaris10 */
-	{ 0x67CA, &polaris10_device_info },	/* Polaris10 */
-	{ 0x67CC, &polaris10_device_info },	/* Polaris10 */
-	{ 0x67CF, &polaris10_device_info },	/* Polaris10 */
-	{ 0x67D0, &polaris10_vf_device_info },	/* Polaris10 vf*/
-	{ 0x67DF, &polaris10_device_info },	/* Polaris10 */
-	{ 0x6FDF, &polaris10_device_info },	/* Polaris10 */
-	{ 0x67E0, &polaris11_device_info },	/* Polaris11 */
-	{ 0x67E1, &polaris11_device_info },	/* Polaris11 */
-	{ 0x67E3, &polaris11_device_info },	/* Polaris11 */
-	{ 0x67E7, &polaris11_device_info },	/* Polaris11 */
-	{ 0x67E8, &polaris11_device_info },	/* Polaris11 */
-	{ 0x67E9, &polaris11_device_info },	/* Polaris11 */
-	{ 0x67EB, &polaris11_device_info },	/* Polaris11 */
-	{ 0x67EF, &polaris11_device_info },	/* Polaris11 */
-	{ 0x67FF, &polaris11_device_info },	/* Polaris11 */
-	{ 0x6980, &polaris12_device_info },	/* Polaris12 */
-	{ 0x6981, &polaris12_device_info },	/* Polaris12 */
-	{ 0x6985, &polaris12_device_info },	/* Polaris12 */
-	{ 0x6986, &polaris12_device_info },	/* Polaris12 */
-	{ 0x6987, &polaris12_device_info },	/* Polaris12 */
-	{ 0x6995, &polaris12_device_info },	/* Polaris12 */
-	{ 0x6997, &polaris12_device_info },	/* Polaris12 */
-	{ 0x699F, &polaris12_device_info },	/* Polaris12 */
-	{ 0x694C, &vegam_device_info },		/* VegaM */
-	{ 0x694E, &vegam_device_info },		/* VegaM */
-	{ 0x694F, &vegam_device_info },		/* VegaM */
-	{ 0x6860, &vega10_device_info },	/* Vega10 */
-	{ 0x6861, &vega10_device_info },	/* Vega10 */
-	{ 0x6862, &vega10_device_info },	/* Vega10 */
-	{ 0x6863, &vega10_device_info },	/* Vega10 */
-	{ 0x6864, &vega10_device_info },	/* Vega10 */
-	{ 0x6867, &vega10_device_info },	/* Vega10 */
-	{ 0x6868, &vega10_device_info },	/* Vega10 */
-	{ 0x6869, &vega10_device_info },	/* Vega10 */
-	{ 0x686A, &vega10_device_info },	/* Vega10 */
-	{ 0x686B, &vega10_device_info },	/* Vega10 */
-	{ 0x686C, &vega10_vf_device_info },	/* Vega10  vf*/
-	{ 0x686D, &vega10_device_info },	/* Vega10 */
-	{ 0x686E, &vega10_device_info },	/* Vega10 */
-	{ 0x686F, &vega10_device_info },	/* Vega10 */
-	{ 0x687F, &vega10_device_info },	/* Vega10 */
-	{ 0x69A0, &vega12_device_info },	/* Vega12 */
-	{ 0x69A1, &vega12_device_info },	/* Vega12 */
-	{ 0x69A2, &vega12_device_info },	/* Vega12 */
-	{ 0x69A3, &vega12_device_info },	/* Vega12 */
-	{ 0x69AF, &vega12_device_info },	/* Vega12 */
-	{ 0x66a0, &vega20_device_info },	/* Vega20 */
-	{ 0x66a1, &vega20_device_info },	/* Vega20 */
-	{ 0x66a2, &vega20_device_info },	/* Vega20 */
-	{ 0x66a3, &vega20_device_info },	/* Vega20 */
-	{ 0x66a4, &vega20_device_info },	/* Vega20 */
-	{ 0x66a7, &vega20_device_info },	/* Vega20 */
-	{ 0x66af, &vega20_device_info },	/* Vega20 */
-	{ 0x738C, &arcturus_device_info },	/* Arcturus */
-	{ 0x7388, &arcturus_device_info },	/* Arcturus */
-	{ 0x738E, &arcturus_device_info },	/* Arcturus */
-	{ 0x7390, &arcturus_device_info },	/* Arcturus vf */
-	{ 0x7310, &navi10_device_info },	/* Navi10 */
-	{ 0x7312, &navi10_device_info },	/* Navi10 */
-	{ 0x7318, &navi10_device_info },	/* Navi10 */
-	{ 0x731a, &navi10_device_info },	/* Navi10 */
-	{ 0x731f, &navi10_device_info },	/* Navi10 */
+	[CHIP_HAWAII] = {&hawaii_device_info, NULL},
+	[CHIP_TONGA] = {&tonga_device_info, NULL},
+	[CHIP_FIJI] = {&fiji_device_info, &fiji_vf_device_info},
+	[CHIP_POLARIS10] = {&polaris10_device_info, &polaris10_vf_device_info},
+	[CHIP_POLARIS11] = {&polaris11_device_info, NULL},
+	[CHIP_POLARIS12] = {&polaris12_device_info, NULL},
+	[CHIP_VEGAM] = {&vegam_device_info, NULL},
+	[CHIP_VEGA10] = {&vega10_device_info, &vega10_vf_device_info},
+	[CHIP_VEGA12] = {&vega12_device_info, NULL},
+	[CHIP_VEGA20] = {&vega20_device_info, NULL},
+	[CHIP_RENOIR] = {&renoir_device_info, NULL},
+	[CHIP_ARCTURUS] = {&arcturus_device_info, &arcturus_device_info},
+	[CHIP_NAVI10] = {&navi10_device_info, NULL},
+	[CHIP_NAVI12] = {&navi12_device_info, &navi12_device_info},
+	[CHIP_NAVI14] = {&navi14_device_info, NULL},
 };
 
 static int kfd_gtt_sa_init(struct kfd_dev *kfd, unsigned int buf_size,
@@ -504,32 +488,25 @@ static void kfd_gtt_sa_fini(struct kfd_dev *kfd);
 
 static int kfd_resume(struct kfd_dev *kfd);
 
-static const struct kfd_device_info *lookup_device_info(unsigned short did)
+struct kfd_dev *kgd2kfd_probe(struct kgd_dev *kgd,
+	struct pci_dev *pdev, unsigned int asic_type, bool vf)
 {
-	size_t i;
+	struct kfd_dev *kfd;
+	const struct kfd_device_info *device_info;
+	const struct kfd2kgd_calls *f2g;
 
-	for (i = 0; i < ARRAY_SIZE(supported_devices); i++) {
-		if (supported_devices[i].did == did) {
-			WARN_ON(!supported_devices[i].device_info);
-			return supported_devices[i].device_info;
-		}
+	if (asic_type >= sizeof(kfd_supported_devices) / (sizeof(void *) * 2)
+		|| asic_type >= sizeof(kfd2kgd_funcs) / sizeof(void *)) {
+		dev_err(kfd_device, "asic_type %d out of range\n", asic_type);
+		return NULL; /* asic_type out of range */
 	}
 
-	dev_warn(kfd_device, "DID %04x is missing in supported_devices\n",
-		 did);
+	device_info = kfd_supported_devices[asic_type][vf];
+	f2g = kfd2kgd_funcs[asic_type];
 
-	return NULL;
-}
-
-struct kfd_dev *kgd2kfd_probe(struct kgd_dev *kgd,
-	struct pci_dev *pdev, const struct kfd2kgd_calls *f2g)
-{
-	struct kfd_dev *kfd;
-	const struct kfd_device_info *device_info =
-					lookup_device_info(pdev->device);
-
-	if (!device_info) {
-		dev_err(kfd_device, "kgd2kfd_probe failed\n");
+	if (!device_info || !f2g) {
+		dev_err(kfd_device, "%s %s not supported in kfd\n",
+			amdgpu_asic_name[asic_type], vf ? "VF" : "");
 		return NULL;
 	}
 
@@ -593,10 +570,12 @@ static void kfd_cwsr_init(struct kfd_dev *kfd)
 }
 
 bool kgd2kfd_device_init(struct kfd_dev *kfd,
+			 struct drm_device *ddev,
 			 const struct kgd2kfd_shared_resources *gpu_resources)
 {
 	unsigned int size;
 
+	kfd->ddev = ddev;
 	kfd->mec_fw_version = amdgpu_amdkfd_get_fw_version(kfd->kgd,
 			KGD_ENGINE_MEC1);
 	kfd->sdma_fw_version = amdgpu_amdkfd_get_fw_version(kfd->kgd,
@@ -751,9 +730,6 @@ int kgd2kfd_pre_reset(struct kfd_dev *kfd)
 		return 0;
 	kgd2kfd_suspend(kfd);
 
-	/* hold dqm->lock to prevent further execution*/
-	dqm_lock(kfd->dqm);
-
 	kfd_signal_reset_event(kfd);
 	return 0;
 }
@@ -771,8 +747,6 @@ int kgd2kfd_post_reset(struct kfd_dev *kfd)
 	if (!kfd->init_complete)
 		return 0;
 
-	dqm_unlock(kfd->dqm);
-
 	ret = kfd_resume(kfd);
 	if (ret)
 		return ret;
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
index d985e31fcc1e..984c2f2b24b6 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
@@ -195,20 +195,30 @@ static int allocate_vmid(struct device_queue_manager *dqm,
 			struct qcm_process_device *qpd,
 			struct queue *q)
 {
-	int bit, allocated_vmid;
+	int allocated_vmid = -1, i;
 
-	if (dqm->vmid_bitmap == 0)
-		return -ENOMEM;
+	for (i = dqm->dev->vm_info.first_vmid_kfd;
+			i <= dqm->dev->vm_info.last_vmid_kfd; i++) {
+		if (!dqm->vmid_pasid[i]) {
+			allocated_vmid = i;
+			break;
+		}
+	}
+
+	if (allocated_vmid < 0) {
+		pr_err("no more vmid to allocate\n");
+		return -ENOSPC;
+	}
+
+	pr_debug("vmid allocated: %d\n", allocated_vmid);
+
+	dqm->vmid_pasid[allocated_vmid] = q->process->pasid;
 
-	bit = ffs(dqm->vmid_bitmap) - 1;
-	dqm->vmid_bitmap &= ~(1 << bit);
+	set_pasid_vmid_mapping(dqm, q->process->pasid, allocated_vmid);
 
-	allocated_vmid = bit + dqm->dev->vm_info.first_vmid_kfd;
-	pr_debug("vmid allocation %d\n", allocated_vmid);
 	qpd->vmid = allocated_vmid;
 	q->properties.vmid = allocated_vmid;
 
-	set_pasid_vmid_mapping(dqm, q->process->pasid, q->properties.vmid);
 	program_sh_mem_settings(dqm, qpd);
 
 	/* qpd->page_table_base is set earlier when register_process()
@@ -220,8 +230,9 @@ static int allocate_vmid(struct device_queue_manager *dqm,
 	/* invalidate the VM context after pasid and vmid mapping is set up */
 	kfd_flush_tlb(qpd_to_pdd(qpd));
 
-	dqm->dev->kfd2kgd->set_scratch_backing_va(
-		dqm->dev->kgd, qpd->sh_hidden_private_base, qpd->vmid);
+	if (dqm->dev->kfd2kgd->set_scratch_backing_va)
+		dqm->dev->kfd2kgd->set_scratch_backing_va(dqm->dev->kgd,
+				qpd->sh_hidden_private_base, qpd->vmid);
 
 	return 0;
 }
@@ -248,8 +259,6 @@ static void deallocate_vmid(struct device_queue_manager *dqm,
 				struct qcm_process_device *qpd,
 				struct queue *q)
 {
-	int bit = qpd->vmid - dqm->dev->vm_info.first_vmid_kfd;
-
 	/* On GFX v7, CP doesn't flush TC at dequeue */
 	if (q->device->device_info->asic_family == CHIP_HAWAII)
 		if (flush_texture_cache_nocpsch(q->device, qpd))
@@ -259,8 +268,8 @@ static void deallocate_vmid(struct device_queue_manager *dqm,
 
 	/* Release the vmid mapping */
 	set_pasid_vmid_mapping(dqm, 0, qpd->vmid);
+	dqm->vmid_pasid[qpd->vmid] = 0;
 
-	dqm->vmid_bitmap |= (1 << bit);
 	qpd->vmid = 0;
 	q->properties.vmid = 0;
 }
@@ -331,6 +340,10 @@ static int create_queue_nocpsch(struct device_queue_manager *dqm,
 	mqd_mgr->init_mqd(mqd_mgr, &q->mqd, q->mqd_mem_obj,
 				&q->gart_mqd_addr, &q->properties);
 	if (q->properties.is_active) {
+		if (!dqm->sched_running) {
+			WARN_ONCE(1, "Load non-HWS mqd while stopped\n");
+			goto add_queue_to_list;
+		}
 
 		if (WARN(q->process->mm != current->mm,
 					"should only run in user thread"))
@@ -342,6 +355,7 @@ static int create_queue_nocpsch(struct device_queue_manager *dqm,
 			goto out_free_mqd;
 	}
 
+add_queue_to_list:
 	list_add(&q->list, &qpd->queues_list);
 	qpd->queue_count++;
 	if (q->properties.is_active)
@@ -449,6 +463,11 @@ static int destroy_queue_nocpsch_locked(struct device_queue_manager *dqm,
 
 	deallocate_doorbell(qpd, q);
 
+	if (!dqm->sched_running) {
+		WARN_ONCE(1, "Destroy non-HWS queue while stopped\n");
+		return 0;
+	}
+
 	retval = mqd_mgr->destroy_mqd(mqd_mgr, q->mqd,
 				KFD_PREEMPT_TYPE_WAVEFRONT_RESET,
 				KFD_UNMAP_LATENCY_MS,
@@ -524,6 +543,12 @@ static int update_queue(struct device_queue_manager *dqm, struct queue *q)
 		   (q->properties.type == KFD_QUEUE_TYPE_COMPUTE ||
 		    q->properties.type == KFD_QUEUE_TYPE_SDMA ||
 		    q->properties.type == KFD_QUEUE_TYPE_SDMA_XGMI)) {
+
+		if (!dqm->sched_running) {
+			WARN_ONCE(1, "Update non-HWS queue while stopped\n");
+			goto out_unlock;
+		}
+
 		retval = mqd_mgr->destroy_mqd(mqd_mgr, q->mqd,
 				KFD_PREEMPT_TYPE_WAVEFRONT_DRAIN,
 				KFD_UNMAP_LATENCY_MS, q->pipe, q->queue);
@@ -579,7 +604,7 @@ static int evict_process_queues_nocpsch(struct device_queue_manager *dqm,
 		goto out;
 
 	pdd = qpd_to_pdd(qpd);
-	pr_info_ratelimited("Evicting PASID %u queues\n",
+	pr_info_ratelimited("Evicting PASID 0x%x queues\n",
 			    pdd->process->pasid);
 
 	/* Mark all queues as evicted. Deactivate all active queues on
@@ -593,6 +618,11 @@ static int evict_process_queues_nocpsch(struct device_queue_manager *dqm,
 		mqd_mgr = dqm->mqd_mgrs[get_mqd_type_from_queue_type(
 				q->properties.type)];
 		q->properties.is_active = false;
+		dqm->queue_count--;
+
+		if (WARN_ONCE(!dqm->sched_running, "Evict when stopped\n"))
+			continue;
+
 		retval = mqd_mgr->destroy_mqd(mqd_mgr, q->mqd,
 				KFD_PREEMPT_TYPE_WAVEFRONT_DRAIN,
 				KFD_UNMAP_LATENCY_MS, q->pipe, q->queue);
@@ -601,7 +631,6 @@ static int evict_process_queues_nocpsch(struct device_queue_manager *dqm,
 			 * maintain a consistent eviction state
 			 */
 			ret = retval;
-		dqm->queue_count--;
 	}
 
 out:
@@ -621,7 +650,7 @@ static int evict_process_queues_cpsch(struct device_queue_manager *dqm,
 		goto out;
 
 	pdd = qpd_to_pdd(qpd);
-	pr_info_ratelimited("Evicting PASID %u queues\n",
+	pr_info_ratelimited("Evicting PASID 0x%x queues\n",
 			    pdd->process->pasid);
 
 	/* Mark all queues as evicted. Deactivate all active queues on
@@ -667,7 +696,7 @@ static int restore_process_queues_nocpsch(struct device_queue_manager *dqm,
 		goto out;
 	}
 
-	pr_info_ratelimited("Restoring PASID %u queues\n",
+	pr_info_ratelimited("Restoring PASID 0x%x queues\n",
 			    pdd->process->pasid);
 
 	/* Update PD Base in QPD */
@@ -702,6 +731,11 @@ static int restore_process_queues_nocpsch(struct device_queue_manager *dqm,
 		mqd_mgr = dqm->mqd_mgrs[get_mqd_type_from_queue_type(
 				q->properties.type)];
 		q->properties.is_active = true;
+		dqm->queue_count++;
+
+		if (WARN_ONCE(!dqm->sched_running, "Restore when stopped\n"))
+			continue;
+
 		retval = mqd_mgr->load_mqd(mqd_mgr, q->mqd, q->pipe,
 				       q->queue, &q->properties, mm);
 		if (retval && !ret)
@@ -709,7 +743,6 @@ static int restore_process_queues_nocpsch(struct device_queue_manager *dqm,
 			 * maintain a consistent eviction state
 			 */
 			ret = retval;
-		dqm->queue_count++;
 	}
 	qpd->evicted = 0;
 out:
@@ -739,7 +772,7 @@ static int restore_process_queues_cpsch(struct device_queue_manager *dqm,
 		goto out;
 	}
 
-	pr_info_ratelimited("Restoring PASID %u queues\n",
+	pr_info_ratelimited("Restoring PASID 0x%x queues\n",
 			    pdd->process->pasid);
 
 	/* Update PD Base in QPD */
@@ -879,7 +912,8 @@ static int initialize_nocpsch(struct device_queue_manager *dqm)
 				dqm->allocated_queues[pipe] |= 1 << queue;
 	}
 
-	dqm->vmid_bitmap = (1 << dqm->dev->vm_info.vmid_num_kfd) - 1;
+	memset(dqm->vmid_pasid, 0, sizeof(dqm->vmid_pasid));
+
 	dqm->sdma_bitmap = ~0ULL >> (64 - get_num_sdma_queues(dqm));
 	dqm->xgmi_sdma_bitmap = ~0ULL >> (64 - get_num_xgmi_sdma_queues(dqm));
 
@@ -902,12 +936,20 @@ static void uninitialize(struct device_queue_manager *dqm)
 static int start_nocpsch(struct device_queue_manager *dqm)
 {
 	init_interrupts(dqm);
-	return pm_init(&dqm->packets, dqm);
+	
+	if (dqm->dev->device_info->asic_family == CHIP_HAWAII)
+		return pm_init(&dqm->packets, dqm);
+	dqm->sched_running = true;
+
+	return 0;
 }
 
 static int stop_nocpsch(struct device_queue_manager *dqm)
 {
-	pm_uninit(&dqm->packets);
+	if (dqm->dev->device_info->asic_family == CHIP_HAWAII)
+		pm_uninit(&dqm->packets);
+	dqm->sched_running = false;
+
 	return 0;
 }
 
@@ -1058,6 +1100,7 @@ static int start_cpsch(struct device_queue_manager *dqm)
 	dqm_lock(dqm);
 	/* clear hang status when driver try to start the hw scheduler */
 	dqm->is_hws_hang = false;
+	dqm->sched_running = true;
 	execute_queues_cpsch(dqm, KFD_UNMAP_QUEUES_FILTER_DYNAMIC_QUEUES, 0);
 	dqm_unlock(dqm);
 
@@ -1073,6 +1116,7 @@ static int stop_cpsch(struct device_queue_manager *dqm)
 {
 	dqm_lock(dqm);
 	unmap_queues_cpsch(dqm, KFD_UNMAP_QUEUES_FILTER_ALL_QUEUES, 0);
+	dqm->sched_running = false;
 	dqm_unlock(dqm);
 
 	kfd_gtt_sa_free(dqm->dev, dqm->fence_mem);
@@ -1259,9 +1303,10 @@ static int map_queues_cpsch(struct device_queue_manager *dqm)
 {
 	int retval;
 
+	if (!dqm->sched_running)
+		return 0;
 	if (dqm->queue_count <= 0 || dqm->processes_count <= 0)
 		return 0;
-
 	if (dqm->active_runlist)
 		return 0;
 
@@ -1283,6 +1328,8 @@ static int unmap_queues_cpsch(struct device_queue_manager *dqm,
 {
 	int retval = 0;
 
+	if (!dqm->sched_running)
+		return 0;
 	if (dqm->is_hws_hang)
 		return -EIO;
 	if (!dqm->active_runlist)
@@ -1676,7 +1723,8 @@ static int allocate_hiq_sdma_mqd(struct device_queue_manager *dqm)
 	struct kfd_dev *dev = dqm->dev;
 	struct kfd_mem_obj *mem_obj = &dqm->hiq_sdma_mqd;
 	uint32_t size = dqm->mqd_mgrs[KFD_MQD_TYPE_SDMA]->mqd_size *
-		dev->device_info->num_sdma_engines *
+		(dev->device_info->num_sdma_engines +
+		dev->device_info->num_xgmi_sdma_engines) *
 		dev->device_info->num_sdma_queues_per_engine +
 		dqm->mqd_mgrs[KFD_MQD_TYPE_HIQ]->mqd_size;
 
@@ -1786,10 +1834,13 @@ struct device_queue_manager *device_queue_manager_init(struct kfd_dev *dev)
 	case CHIP_VEGA12:
 	case CHIP_VEGA20:
 	case CHIP_RAVEN:
+	case CHIP_RENOIR:
 	case CHIP_ARCTURUS:
 		device_queue_manager_init_v9(&dqm->asic_ops);
 		break;
 	case CHIP_NAVI10:
+	case CHIP_NAVI12:
+	case CHIP_NAVI14:
 		device_queue_manager_init_v10_navi10(&dqm->asic_ops);
 		break;
 	default:
@@ -1883,6 +1934,12 @@ int dqm_debugfs_hqds(struct seq_file *m, void *data)
 	int pipe, queue;
 	int r = 0;
 
+	if (!dqm->sched_running) {
+		seq_printf(m, " Device is stopped\n");
+
+		return 0;
+	}
+
 	r = dqm->dev->kfd2kgd->hqd_dump(dqm->dev->kgd,
 					KFD_CIK_HIQ_PIPE, KFD_CIK_HIQ_QUEUE,
 					&dump, &n_regs);
@@ -1917,7 +1974,8 @@ int dqm_debugfs_hqds(struct seq_file *m, void *data)
 		}
 	}
 
-	for (pipe = 0; pipe < get_num_sdma_engines(dqm); pipe++) {
+	for (pipe = 0; pipe < get_num_sdma_engines(dqm) +
+			get_num_xgmi_sdma_engines(dqm); pipe++) {
 		for (queue = 0;
 		     queue < dqm->dev->device_info->num_sdma_queues_per_engine;
 		     queue++) {
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h
index 90db2c9275f6..a8c37e6da027 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h
@@ -32,6 +32,8 @@
 #include "kfd_mqd_manager.h"
 
 
+#define VMID_NUM 16
+
 struct device_process_node {
 	struct qcm_process_device *qpd;
 	struct list_head list;
@@ -185,7 +187,8 @@ struct device_queue_manager {
 	unsigned int		*allocated_queues;
 	uint64_t		sdma_bitmap;
 	uint64_t		xgmi_sdma_bitmap;
-	unsigned int		vmid_bitmap;
+	/* the pasid mapping for each kfd vmid */
+	uint16_t		vmid_pasid[VMID_NUM];
 	uint64_t		pipelines_addr;
 	struct kfd_mem_obj	*pipeline_mem;
 	uint64_t		fence_gpu_addr;
@@ -198,6 +201,7 @@ struct device_queue_manager {
 	bool			is_hws_hang;
 	struct work_struct	hw_exception_work;
 	struct kfd_mem_obj	hiq_sdma_mqd;
+	bool			sched_running;
 };
 
 void device_queue_manager_init_cik(
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_events.c b/drivers/gpu/drm/amd/amdkfd/kfd_events.c
index d674d4b3340f..908081c85de1 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_events.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_events.c
@@ -852,8 +852,8 @@ static void lookup_events_by_type_and_signal(struct kfd_process *p,
 
 	if (type == KFD_EVENT_TYPE_MEMORY) {
 		dev_warn(kfd_device,
-			"Sending SIGSEGV to HSA Process with PID %d ",
-				p->lead_thread->pid);
+			"Sending SIGSEGV to process %d (pasid 0x%x)",
+				p->lead_thread->pid, p->pasid);
 		send_sig(SIGSEGV, p->lead_thread, 0);
 	}
 
@@ -861,13 +861,13 @@ static void lookup_events_by_type_and_signal(struct kfd_process *p,
 	if (send_signal) {
 		if (send_sigterm) {
 			dev_warn(kfd_device,
-				"Sending SIGTERM to HSA Process with PID %d ",
-					p->lead_thread->pid);
+				"Sending SIGTERM to process %d (pasid 0x%x)",
+					p->lead_thread->pid, p->pasid);
 			send_sig(SIGTERM, p->lead_thread, 0);
 		} else {
 			dev_err(kfd_device,
-				"HSA Process (PID %d) got unhandled exception",
-				p->lead_thread->pid);
+				"Process %d (pasid 0x%x) got unhandled exception",
+				p->lead_thread->pid, p->pasid);
 		}
 	}
 }
@@ -936,7 +936,8 @@ void kfd_signal_iommu_event(struct kfd_dev *dev, unsigned int pasid,
 	/* Workaround on Raven to not kill the process when memory is freed
 	 * before IOMMU is able to finish processing all the excessive PPRs
 	 */
-	if (dev->device_info->asic_family != CHIP_RAVEN) {
+	if (dev->device_info->asic_family != CHIP_RAVEN &&
+	    dev->device_info->asic_family != CHIP_RENOIR) {
 		mutex_lock(&p->event_mutex);
 
 		/* Lookup events by type and signal them */
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_flat_memory.c b/drivers/gpu/drm/amd/amdkfd/kfd_flat_memory.c
index 9dc4bff8085e..bb77b8890e77 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_flat_memory.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_flat_memory.c
@@ -369,8 +369,13 @@ int kfd_init_apertures(struct kfd_process *process)
 
 	/*Iterating over all devices*/
 	while (kfd_topology_enum_kfd_devices(id, &dev) == 0) {
-		if (!dev) {
-			id++; /* Skip non GPU devices */
+		if (!dev || kfd_devcgroup_check_permission(dev)) {
+			/* Skip non GPU devices and devices to which the
+			 * current process have no access to. Access can be
+			 * limited by placing the process in a specific
+			 * cgroup hierarchy
+			 */
+			id++;
 			continue;
 		}
 
@@ -405,8 +410,11 @@ int kfd_init_apertures(struct kfd_process *process)
 			case CHIP_VEGA12:
 			case CHIP_VEGA20:
 			case CHIP_RAVEN:
+			case CHIP_RENOIR:
 			case CHIP_ARCTURUS:
 			case CHIP_NAVI10:
+			case CHIP_NAVI12:
+			case CHIP_NAVI14:
 				kfd_init_apertures_v9(pdd, id);
 				break;
 			default:
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c b/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c
index 3ef67d2e0d9f..e05d75ecda21 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c
@@ -54,8 +54,7 @@ static bool event_interrupt_isr_v9(struct kfd_dev *dev,
 		memcpy(patched_ihre, ih_ring_entry,
 				dev->device_info->ih_ring_entry_size);
 
-		pasid = dev->kfd2kgd->get_atc_vmid_pasid_mapping_pasid(
-				dev->kgd, vmid);
+		pasid = dev->dqm->vmid_pasid[vmid];
 
 		/* Patch the pasid field */
 		patched_ihre[3] = cpu_to_le32((le32_to_cpu(patched_ihre[3])
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_interrupt.c b/drivers/gpu/drm/amd/amdkfd/kfd_interrupt.c
index c56ac47cd318..bc47f6a44456 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_interrupt.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_interrupt.c
@@ -62,6 +62,11 @@ int kfd_interrupt_init(struct kfd_dev *kfd)
 	}
 
 	kfd->ih_wq = alloc_workqueue("KFD IH", WQ_HIGHPRI, 1);
+	if (unlikely(!kfd->ih_wq)) {
+		kfifo_free(&kfd->ih_fifo);
+		dev_err(kfd_chardev(), "Failed to allocate KFD IH workqueue\n");
+		return -ENOMEM;
+	}
 	spin_lock_init(&kfd->interrupt_lock);
 
 	INIT_WORK(&kfd->interrupt_work, interrupt_wq);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_iommu.c b/drivers/gpu/drm/amd/amdkfd/kfd_iommu.c
index 5f35df23fb18..193e2835bd4d 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_iommu.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_iommu.c
@@ -160,7 +160,7 @@ static void iommu_pasid_shutdown_callback(struct pci_dev *pdev, int pasid)
 	if (!p)
 		return;
 
-	pr_debug("Unbinding process %d from IOMMU\n", pasid);
+	pr_debug("Unbinding process 0x%x from IOMMU\n", pasid);
 
 	mutex_lock(kfd_get_dbgmgr_mutex());
 
@@ -194,7 +194,7 @@ static int iommu_invalid_ppr_cb(struct pci_dev *pdev, int pasid,
 	struct kfd_dev *dev;
 
 	dev_warn_ratelimited(kfd_device,
-			"Invalid PPR device %x:%x.%x pasid %d address 0x%lX flags 0x%X",
+			"Invalid PPR device %x:%x.%x pasid 0x%x address 0x%lX flags 0x%X",
 			PCI_BUS_NUM(pdev->devfn),
 			PCI_SLOT(pdev->devfn),
 			PCI_FUNC(pdev->devfn),
@@ -235,7 +235,7 @@ static int kfd_bind_processes_to_device(struct kfd_dev *kfd)
 		err = amd_iommu_bind_pasid(kfd->pdev, p->pasid,
 				p->lead_thread);
 		if (err < 0) {
-			pr_err("Unexpected pasid %d binding failure\n",
+			pr_err("Unexpected pasid 0x%x binding failure\n",
 					p->pasid);
 			mutex_unlock(&p->mutex);
 			break;
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c
index 8b4564f71a7a..11d244891393 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c
@@ -330,10 +330,13 @@ struct kernel_queue *kernel_queue_init(struct kfd_dev *dev,
 	case CHIP_VEGA12:
 	case CHIP_VEGA20:
 	case CHIP_RAVEN:
+	case CHIP_RENOIR:
 	case CHIP_ARCTURUS:
 		kernel_queue_init_v9(&kq->ops_asic_specific);
 		break;
 	case CHIP_NAVI10:
+	case CHIP_NAVI12:
+	case CHIP_NAVI14:
 		kernel_queue_init_v10(&kq->ops_asic_specific);
 		break;
 	default:
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_module.c b/drivers/gpu/drm/amd/amdkfd/kfd_module.c
index 986ff52d5750..f4b7f7e6c40e 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_module.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_module.c
@@ -82,7 +82,7 @@ static void kfd_exit(void)
 	kfd_chardev_exit();
 }
 
-int kgd2kfd_init()
+int kgd2kfd_init(void)
 {
 	return kfd_init();
 }
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v10.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v10.c
index 9cd3eb2d90bd..4a236b2c2354 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v10.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v10.c
@@ -69,35 +69,13 @@ static void update_cu_mask(struct mqd_manager *mm, void *mqd,
 static struct kfd_mem_obj *allocate_mqd(struct kfd_dev *kfd,
 		struct queue_properties *q)
 {
-	int retval;
-	struct kfd_mem_obj *mqd_mem_obj = NULL;
+	struct kfd_mem_obj *mqd_mem_obj;
 
-	/* From V9,  for CWSR, the control stack is located on the next page
-	 * boundary after the mqd, we will use the gtt allocation function
-	 * instead of sub-allocation function.
-	 */
-	if (kfd->cwsr_enabled && (q->type == KFD_QUEUE_TYPE_COMPUTE)) {
-		mqd_mem_obj = kzalloc(sizeof(struct kfd_mem_obj), GFP_NOIO);
-		if (!mqd_mem_obj)
-			return NULL;
-		retval = amdgpu_amdkfd_alloc_gtt_mem(kfd->kgd,
-			ALIGN(q->ctl_stack_size, PAGE_SIZE) +
-				ALIGN(sizeof(struct v10_compute_mqd), PAGE_SIZE),
-			&(mqd_mem_obj->gtt_mem),
-			&(mqd_mem_obj->gpu_addr),
-			(void *)&(mqd_mem_obj->cpu_ptr), true);
-	} else {
-		retval = kfd_gtt_sa_allocate(kfd, sizeof(struct v10_compute_mqd),
-				&mqd_mem_obj);
-	}
-
-	if (retval) {
-		kfree(mqd_mem_obj);
+	if (kfd_gtt_sa_allocate(kfd, sizeof(struct v10_compute_mqd),
+			&mqd_mem_obj))
 		return NULL;
-	}
 
 	return mqd_mem_obj;
-
 }
 
 static void init_mqd(struct mqd_manager *mm, void **mqd,
@@ -250,14 +228,7 @@ static int destroy_mqd(struct mqd_manager *mm, void *mqd,
 static void free_mqd(struct mqd_manager *mm, void *mqd,
 			struct kfd_mem_obj *mqd_mem_obj)
 {
-	struct kfd_dev *kfd = mm->dev;
-
-	if (mqd_mem_obj->gtt_mem) {
-		amdgpu_amdkfd_free_gtt_mem(kfd->kgd, mqd_mem_obj->gtt_mem);
-		kfree(mqd_mem_obj);
-	} else {
-		kfd_gtt_sa_free(mm->dev, mqd_mem_obj);
-	}
+	kfd_gtt_sa_free(mm->dev, mqd_mem_obj);
 }
 
 static bool is_occupied(struct mqd_manager *mm, void *mqd,
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c
index 2c8624c5b42c..83ef4b3dd2fb 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c
@@ -239,10 +239,13 @@ int pm_init(struct packet_manager *pm, struct device_queue_manager *dqm)
 	case CHIP_VEGA12:
 	case CHIP_VEGA20:
 	case CHIP_RAVEN:
+	case CHIP_RENOIR:
 	case CHIP_ARCTURUS:
 		pm->pmf = &kfd_v9_pm_funcs;
 		break;
 	case CHIP_NAVI10:
+	case CHIP_NAVI12:
+	case CHIP_NAVI14:
 		pm->pmf = &kfd_v10_pm_funcs;
 		break;
 	default:
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
index c89326125d71..060a9e8b301e 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
@@ -36,6 +36,10 @@
 #include <linux/seq_file.h>
 #include <linux/kref.h>
 #include <linux/sysfs.h>
+#include <linux/device_cgroup.h>
+#include <drm/drm_file.h>
+#include <drm/drm_drv.h>
+#include <drm/drm_device.h>
 #include <kgd_kfd_interface.h>
 
 #include "amd_shared.h"
@@ -179,10 +183,6 @@ enum cache_policy {
 	cache_policy_noncoherent
 };
 
-#define KFD_IS_VI(chip) ((chip) >= CHIP_CARRIZO && (chip) <= CHIP_POLARIS11)
-#define KFD_IS_DGPU(chip) (((chip) >= CHIP_TONGA && \
-			   (chip) <= CHIP_NAVI10) || \
-			   (chip) == CHIP_HAWAII)
 #define KFD_IS_SOC15(chip) ((chip) >= CHIP_VEGA10)
 
 struct kfd_event_interrupt_class {
@@ -230,6 +230,7 @@ struct kfd_dev {
 
 	const struct kfd_device_info *device_info;
 	struct pci_dev *pdev;
+	struct drm_device *ddev;
 
 	unsigned int id;		/* topology stub index */
 
@@ -687,7 +688,7 @@ struct kfd_process {
 	/* We want to receive a notification when the mm_struct is destroyed */
 	struct mmu_notifier mmu_notifier;
 
-	unsigned int pasid;
+	uint16_t pasid;
 	unsigned int doorbell_index;
 
 	/*
@@ -1040,6 +1041,21 @@ bool kfd_is_locked(void);
 void kfd_inc_compute_active(struct kfd_dev *dev);
 void kfd_dec_compute_active(struct kfd_dev *dev);
 
+/* Cgroup Support */
+/* Check with device cgroup if @kfd device is accessible */
+static inline int kfd_devcgroup_check_permission(struct kfd_dev *kfd)
+{
+#if defined(CONFIG_CGROUP_DEVICE)
+	struct drm_device *ddev = kfd->ddev;
+
+	return devcgroup_check_permission(DEVCG_DEV_CHAR, ddev->driver->major,
+					  ddev->render->index,
+					  DEVCG_ACC_WRITE | DEVCG_ACC_READ);
+#else
+	return 0;
+#endif
+}
+
 /* Debugfs */
 #if defined(CONFIG_DEBUG_FS)
 
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process.c b/drivers/gpu/drm/amd/amdkfd/kfd_process.c
index 40e3fc0c6942..10f9af5784f2 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_process.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_process.c
@@ -416,7 +416,7 @@ static void kfd_process_destroy_pdds(struct kfd_process *p)
 
 	list_for_each_entry_safe(pdd, temp, &p->per_device_data,
 				 per_device_list) {
-		pr_debug("Releasing pdd (topology id %d) for process (pasid %d)\n",
+		pr_debug("Releasing pdd (topology id %d) for process (pasid 0x%x)\n",
 				pdd->dev->id, p->pasid);
 
 		if (pdd->drm_file) {
@@ -687,6 +687,8 @@ static int init_doorbell_bitmap(struct qcm_process_device *qpd,
 			struct kfd_dev *dev)
 {
 	unsigned int i;
+	int range_start = dev->shared_resources.non_cp_doorbells_start;
+	int range_end = dev->shared_resources.non_cp_doorbells_end;
 
 	if (!KFD_IS_SOC15(dev->device_info->asic_family))
 		return 0;
@@ -698,14 +700,16 @@ static int init_doorbell_bitmap(struct qcm_process_device *qpd,
 		return -ENOMEM;
 
 	/* Mask out doorbells reserved for SDMA, IH, and VCN on SOC15. */
+	pr_debug("reserved doorbell 0x%03x - 0x%03x\n", range_start, range_end);
+	pr_debug("reserved doorbell 0x%03x - 0x%03x\n",
+			range_start + KFD_QUEUE_DOORBELL_MIRROR_OFFSET,
+			range_end + KFD_QUEUE_DOORBELL_MIRROR_OFFSET);
+
 	for (i = 0; i < KFD_MAX_NUM_OF_QUEUES_PER_PROCESS / 2; i++) {
-		if (i >= dev->shared_resources.non_cp_doorbells_start
-			&& i <= dev->shared_resources.non_cp_doorbells_end) {
+		if (i >= range_start && i <= range_end) {
 			set_bit(i, qpd->doorbell_bitmap);
 			set_bit(i + KFD_QUEUE_DOORBELL_MIRROR_OFFSET,
 				qpd->doorbell_bitmap);
-			pr_debug("reserved doorbell 0x%03x and 0x%03x\n", i,
-				i + KFD_QUEUE_DOORBELL_MIRROR_OFFSET);
 		}
 	}
 
@@ -1020,7 +1024,7 @@ static void evict_process_worker(struct work_struct *work)
 	 */
 	flush_delayed_work(&p->restore_work);
 
-	pr_debug("Started evicting pasid %d\n", p->pasid);
+	pr_debug("Started evicting pasid 0x%x\n", p->pasid);
 	ret = kfd_process_evict_queues(p);
 	if (!ret) {
 		dma_fence_signal(p->ef);
@@ -1029,9 +1033,9 @@ static void evict_process_worker(struct work_struct *work)
 		queue_delayed_work(kfd_restore_wq, &p->restore_work,
 				msecs_to_jiffies(PROCESS_RESTORE_TIME_MS));
 
-		pr_debug("Finished evicting pasid %d\n", p->pasid);
+		pr_debug("Finished evicting pasid 0x%x\n", p->pasid);
 	} else
-		pr_err("Failed to evict queues of pasid %d\n", p->pasid);
+		pr_err("Failed to evict queues of pasid 0x%x\n", p->pasid);
 }
 
 static void restore_process_worker(struct work_struct *work)
@@ -1046,7 +1050,7 @@ static void restore_process_worker(struct work_struct *work)
 	 * lifetime of this thread, kfd_process p will be valid
 	 */
 	p = container_of(dwork, struct kfd_process, restore_work);
-	pr_debug("Started restoring pasid %d\n", p->pasid);
+	pr_debug("Started restoring pasid 0x%x\n", p->pasid);
 
 	/* Setting last_restore_timestamp before successful restoration.
 	 * Otherwise this would have to be set by KGD (restore_process_bos)
@@ -1062,7 +1066,7 @@ static void restore_process_worker(struct work_struct *work)
 	ret = amdgpu_amdkfd_gpuvm_restore_process_bos(p->kgd_process_info,
 						     &p->ef);
 	if (ret) {
-		pr_debug("Failed to restore BOs of pasid %d, retry after %d ms\n",
+		pr_debug("Failed to restore BOs of pasid 0x%x, retry after %d ms\n",
 			 p->pasid, PROCESS_BACK_OFF_TIME_MS);
 		ret = queue_delayed_work(kfd_restore_wq, &p->restore_work,
 				msecs_to_jiffies(PROCESS_BACK_OFF_TIME_MS));
@@ -1072,9 +1076,9 @@ static void restore_process_worker(struct work_struct *work)
 
 	ret = kfd_process_restore_queues(p);
 	if (!ret)
-		pr_debug("Finished restoring pasid %d\n", p->pasid);
+		pr_debug("Finished restoring pasid 0x%x\n", p->pasid);
 	else
-		pr_err("Failed to restore queues of pasid %d\n", p->pasid);
+		pr_err("Failed to restore queues of pasid 0x%x\n", p->pasid);
 }
 
 void kfd_suspend_all_processes(void)
@@ -1088,7 +1092,7 @@ void kfd_suspend_all_processes(void)
 		cancel_delayed_work_sync(&p->restore_work);
 
 		if (kfd_process_evict_queues(p))
-			pr_err("Failed to suspend process %d\n", p->pasid);
+			pr_err("Failed to suspend process 0x%x\n", p->pasid);
 		dma_fence_signal(p->ef);
 		dma_fence_put(p->ef);
 		p->ef = NULL;
@@ -1171,7 +1175,7 @@ int kfd_debugfs_mqds_by_process(struct seq_file *m, void *data)
 	int idx = srcu_read_lock(&kfd_processes_srcu);
 
 	hash_for_each_rcu(kfd_processes_table, temp, p, kfd_processes) {
-		seq_printf(m, "Process %d PASID %d:\n",
+		seq_printf(m, "Process %d PASID 0x%x:\n",
 			   p->lead_thread->tgid, p->pasid);
 
 		mutex_lock(&p->mutex);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c
index 7e6c3ee82f5b..2659d226c056 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c
@@ -53,7 +53,7 @@ static int find_available_queue_slot(struct process_queue_manager *pqm,
 	pr_debug("The new slot id %lu\n", found);
 
 	if (found >= KFD_MAX_NUM_OF_QUEUES_PER_PROCESS) {
-		pr_info("Cannot open more queues for process with pasid %d\n",
+		pr_info("Cannot open more queues for process with pasid 0x%x\n",
 				pqm->process->pasid);
 		return -ENOMEM;
 	}
@@ -298,7 +298,7 @@ int pqm_create_queue(struct process_queue_manager *pqm,
 	}
 
 	if (retval != 0) {
-		pr_err("Pasid %d DQM create queue %d failed. ret %d\n",
+		pr_err("Pasid 0x%x DQM create queue %d failed. ret %d\n",
 			pqm->process->pasid, type, retval);
 		goto err_create_queue;
 	}
@@ -377,7 +377,7 @@ int pqm_destroy_queue(struct process_queue_manager *pqm, unsigned int qid)
 		dqm = pqn->q->device->dqm;
 		retval = dqm->ops.destroy_queue(dqm, &pdd->qpd, pqn->q);
 		if (retval) {
-			pr_err("Pasid %d destroy queue %d failed, ret %d\n",
+			pr_err("Pasid 0x%x destroy queue %d failed, ret %d\n",
 				pqm->process->pasid,
 				pqn->q->properties.queue_id, retval);
 			if (retval != -ETIME)
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c
index 7551761f2aa9..69bd0628fdc6 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c
@@ -269,6 +269,8 @@ static ssize_t iolink_show(struct kobject *kobj, struct attribute *attr,
 	buffer[0] = 0;
 
 	iolink = container_of(attr, struct kfd_iolink_properties, attr);
+	if (iolink->gpu && kfd_devcgroup_check_permission(iolink->gpu))
+		return -EPERM;
 	sysfs_show_32bit_prop(buffer, "type", iolink->iolink_type);
 	sysfs_show_32bit_prop(buffer, "version_major", iolink->ver_maj);
 	sysfs_show_32bit_prop(buffer, "version_minor", iolink->ver_min);
@@ -305,6 +307,8 @@ static ssize_t mem_show(struct kobject *kobj, struct attribute *attr,
 	buffer[0] = 0;
 
 	mem = container_of(attr, struct kfd_mem_properties, attr);
+	if (mem->gpu && kfd_devcgroup_check_permission(mem->gpu))
+		return -EPERM;
 	sysfs_show_32bit_prop(buffer, "heap_type", mem->heap_type);
 	sysfs_show_64bit_prop(buffer, "size_in_bytes", mem->size_in_bytes);
 	sysfs_show_32bit_prop(buffer, "flags", mem->flags);
@@ -334,6 +338,8 @@ static ssize_t kfd_cache_show(struct kobject *kobj, struct attribute *attr,
 	buffer[0] = 0;
 
 	cache = container_of(attr, struct kfd_cache_properties, attr);
+	if (cache->gpu && kfd_devcgroup_check_permission(cache->gpu))
+		return -EPERM;
 	sysfs_show_32bit_prop(buffer, "processor_id_low",
 			cache->processor_id_low);
 	sysfs_show_32bit_prop(buffer, "level", cache->cache_level);
@@ -414,6 +420,8 @@ static ssize_t node_show(struct kobject *kobj, struct attribute *attr,
 	if (strcmp(attr->name, "gpu_id") == 0) {
 		dev = container_of(attr, struct kfd_topology_device,
 				attr_gpuid);
+		if (dev->gpu && kfd_devcgroup_check_permission(dev->gpu))
+			return -EPERM;
 		return sysfs_show_32bit_val(buffer, dev->gpu_id);
 	}
 
@@ -421,11 +429,15 @@ static ssize_t node_show(struct kobject *kobj, struct attribute *attr,
 		dev = container_of(attr, struct kfd_topology_device,
 				attr_name);
 
+		if (dev->gpu && kfd_devcgroup_check_permission(dev->gpu))
+			return -EPERM;
 		return sysfs_show_str_val(buffer, dev->node_props.name);
 	}
 
 	dev = container_of(attr, struct kfd_topology_device,
 			attr_props);
+	if (dev->gpu && kfd_devcgroup_check_permission(dev->gpu))
+		return -EPERM;
 	sysfs_show_32bit_prop(buffer, "cpu_cores_count",
 			dev->node_props.cpu_cores_count);
 	sysfs_show_32bit_prop(buffer, "simd_count",
@@ -1098,6 +1110,9 @@ static struct kfd_topology_device *kfd_assign_gpu(struct kfd_dev *gpu)
 {
 	struct kfd_topology_device *dev;
 	struct kfd_topology_device *out_dev = NULL;
+	struct kfd_mem_properties *mem;
+	struct kfd_cache_properties *cache;
+	struct kfd_iolink_properties *iolink;
 
 	down_write(&topology_lock);
 	list_for_each_entry(dev, &topology_device_list, list) {
@@ -1111,6 +1126,13 @@ static struct kfd_topology_device *kfd_assign_gpu(struct kfd_dev *gpu)
 		if (!dev->gpu && (dev->node_props.simd_count > 0)) {
 			dev->gpu = gpu;
 			out_dev = dev;
+
+			list_for_each_entry(mem, &dev->mem_props, list)
+				mem->gpu = dev->gpu;
+			list_for_each_entry(cache, &dev->cache_props, list)
+				cache->gpu = dev->gpu;
+			list_for_each_entry(iolink, &dev->io_link_props, list)
+				iolink->gpu = dev->gpu;
 			break;
 		}
 	}
@@ -1317,8 +1339,11 @@ int kfd_topology_add_device(struct kfd_dev *gpu)
 	case CHIP_VEGA12:
 	case CHIP_VEGA20:
 	case CHIP_RAVEN:
+	case CHIP_RENOIR:
 	case CHIP_ARCTURUS:
 	case CHIP_NAVI10:
+	case CHIP_NAVI12:
+	case CHIP_NAVI14:
 		dev->node_props.capability |= ((HSA_CAP_DOORBELL_TYPE_2_0 <<
 			HSA_CAP_DOORBELL_TYPE_TOTALBITS_SHIFT) &
 			HSA_CAP_DOORBELL_TYPE_TOTALBITS_MASK);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_topology.h b/drivers/gpu/drm/amd/amdkfd/kfd_topology.h
index d4718d58d0f2..15843e0fc756 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_topology.h
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_topology.h
@@ -102,6 +102,7 @@ struct kfd_mem_properties {
 	uint32_t		flags;
 	uint32_t		width;
 	uint32_t		mem_clk_max;
+	struct kfd_dev		*gpu;
 	struct kobject		*kobj;
 	struct attribute	attr;
 };
@@ -123,6 +124,7 @@ struct kfd_cache_properties {
 	uint32_t		cache_latency;
 	uint32_t		cache_type;
 	uint8_t			sibling_map[CRAT_SIBLINGMAP_SIZE];
+	struct kfd_dev		*gpu;
 	struct kobject		*kobj;
 	struct attribute	attr;
 };
@@ -141,6 +143,7 @@ struct kfd_iolink_properties {
 	uint32_t		max_bandwidth;
 	uint32_t		rec_transfer_size;
 	uint32_t		flags;
+	struct kfd_dev		*gpu;
 	struct kobject		*kobj;
 	struct attribute	attr;
 };
diff --git a/drivers/gpu/drm/amd/display/Kconfig b/drivers/gpu/drm/amd/display/Kconfig
index 71991a28a775..313183b80032 100644
--- a/drivers/gpu/drm/amd/display/Kconfig
+++ b/drivers/gpu/drm/amd/display/Kconfig
@@ -23,16 +23,16 @@ config DRM_AMD_DC_DCN2_0
 	depends on DRM_AMD_DC && X86
 	depends on DRM_AMD_DC_DCN1_0
 	help
-	    Choose this option if you want to have
-	    Navi support for display engine
+	  Choose this option if you want to have
+	  Navi support for display engine
 
 config DRM_AMD_DC_DCN2_1
-        bool "DCN 2.1 family"
-        depends on DRM_AMD_DC && X86
-        depends on DRM_AMD_DC_DCN2_0
-        help
-            Choose this option if you want to have
-            Renoir support for display engine
+	bool "DCN 2.1 family"
+	depends on DRM_AMD_DC && X86
+	depends on DRM_AMD_DC_DCN2_0
+	help
+	  Choose this option if you want to have
+	  Renoir support for display engine
 
 config DRM_AMD_DC_DSC_SUPPORT
 	bool "DSC support"
@@ -41,8 +41,16 @@ config DRM_AMD_DC_DSC_SUPPORT
 	depends on DRM_AMD_DC_DCN1_0
 	depends on DRM_AMD_DC_DCN2_0
 	help
-	    Choose this option if you want to have
-	    Dynamic Stream Compression support
+	  Choose this option if you want to have
+	  Dynamic Stream Compression support
+
+config DRM_AMD_DC_HDCP
+	bool "Enable HDCP support in DC"
+	depends on DRM_AMD_DC
+	help
+	 Choose this option
+	 if you want to support
+	 HDCP authentication
 
 config DEBUG_KERNEL_DC
 	bool "Enable kgdb break in DC"
diff --git a/drivers/gpu/drm/amd/display/Makefile b/drivers/gpu/drm/amd/display/Makefile
index 496cee000f10..36b3d6a5d04d 100644
--- a/drivers/gpu/drm/amd/display/Makefile
+++ b/drivers/gpu/drm/amd/display/Makefile
@@ -34,12 +34,19 @@ subdir-ccflags-y += -I$(FULL_AMD_DISPLAY_PATH)/modules/freesync
 subdir-ccflags-y += -I$(FULL_AMD_DISPLAY_PATH)/modules/color
 subdir-ccflags-y += -I$(FULL_AMD_DISPLAY_PATH)/modules/info_packet
 subdir-ccflags-y += -I$(FULL_AMD_DISPLAY_PATH)/modules/power
+ifdef CONFIG_DRM_AMD_DC_HDCP
+subdir-ccflags-y += -I$(FULL_AMD_DISPLAY_PATH)/modules/hdcp
+endif
 
 #TODO: remove when Timing Sync feature is complete
 subdir-ccflags-y += -DBUILD_FEATURE_TIMING_SYNC=0
 
 DAL_LIBS = amdgpu_dm dc	modules/freesync modules/color modules/info_packet modules/power
 
+ifdef CONFIG_DRM_AMD_DC_HDCP
+DAL_LIBS += modules/hdcp
+endif
+
 AMD_DAL = $(addsuffix /Makefile, $(addprefix $(FULL_AMD_DISPLAY_PATH)/,$(DAL_LIBS)))
 
 include $(AMD_DAL)
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile b/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile
index 94911871eb9b..9a3b7bf8ab0b 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile
@@ -31,6 +31,10 @@ ifneq ($(CONFIG_DRM_AMD_DC),)
 AMDGPUDM += amdgpu_dm_services.o amdgpu_dm_helpers.o amdgpu_dm_pp_smu.o
 endif
 
+ifdef CONFIG_DRM_AMD_DC_HDCP
+AMDGPUDM += amdgpu_dm_hdcp.o
+endif
+
 ifneq ($(CONFIG_DEBUG_FS),)
 AMDGPUDM += amdgpu_dm_crc.o amdgpu_dm_debugfs.o
 endif
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index 8cab6da512a0..7aac9568d3be 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -37,6 +37,9 @@
 #include "amdgpu_ucode.h"
 #include "atom.h"
 #include "amdgpu_dm.h"
+#ifdef CONFIG_DRM_AMD_DC_HDCP
+#include "amdgpu_dm_hdcp.h"
+#endif
 #include "amdgpu_pm.h"
 
 #include "amd_shared.h"
@@ -67,6 +70,7 @@
 #include <drm/drm_edid.h>
 #include <drm/drm_vblank.h>
 #include <drm/drm_audio_component.h>
+#include <drm/drm_hdcp.h>
 
 #if defined(CONFIG_DRM_AMD_DC_DCN1_0)
 #include "ivsrcid/dcn/irqsrcs_dcn_1_0.h"
@@ -143,6 +147,12 @@ static int amdgpu_dm_atomic_check(struct drm_device *dev,
 static void handle_cursor_update(struct drm_plane *plane,
 				 struct drm_plane_state *old_plane_state);
 
+static void amdgpu_dm_set_psr_caps(struct dc_link *link);
+static bool amdgpu_dm_psr_enable(struct dc_stream_state *stream);
+static bool amdgpu_dm_link_setup_psr(struct dc_stream_state *stream);
+static bool amdgpu_dm_psr_disable(struct dc_stream_state *stream);
+
+
 /*
  * dm_vblank_get_counter
  *
@@ -263,6 +273,13 @@ static inline bool amdgpu_dm_vrr_active(struct dm_crtc_state *dm_state)
 	       dm_state->freesync_config.state == VRR_STATE_ACTIVE_FIXED;
 }
 
+/**
+ * dm_pflip_high_irq() - Handle pageflip interrupt
+ * @interrupt_params: ignored
+ *
+ * Handles the pageflip interrupt by notifying all interested parties
+ * that the pageflip has been completed.
+ */
 static void dm_pflip_high_irq(void *interrupt_params)
 {
 	struct amdgpu_crtc *amdgpu_crtc;
@@ -407,6 +424,13 @@ static void dm_vupdate_high_irq(void *interrupt_params)
 	}
 }
 
+/**
+ * dm_crtc_high_irq() - Handles CRTC interrupt
+ * @interrupt_params: ignored
+ *
+ * Handles the CRTC/VSYNC interrupt by notfying DRM's VBLANK
+ * event handler.
+ */
 static void dm_crtc_high_irq(void *interrupt_params)
 {
 	struct common_irq_params *irq_params = interrupt_params;
@@ -646,11 +670,18 @@ void amdgpu_dm_audio_eld_notify(struct amdgpu_device *adev, int pin)
 static int amdgpu_dm_init(struct amdgpu_device *adev)
 {
 	struct dc_init_data init_data;
+#ifdef CONFIG_DRM_AMD_DC_HDCP
+	struct dc_callback_init init_params;
+#endif
+
 	adev->dm.ddev = adev->ddev;
 	adev->dm.adev = adev;
 
 	/* Zero all the fields */
 	memset(&init_data, 0, sizeof(init_data));
+#ifdef CONFIG_DRM_AMD_DC_HDCP
+	memset(&init_params, 0, sizeof(init_params));
+#endif
 
 	mutex_init(&adev->dm.dc_lock);
 	mutex_init(&adev->dm.audio_lock);
@@ -688,7 +719,7 @@ static int amdgpu_dm_init(struct amdgpu_device *adev)
 	 */
 	if (adev->flags & AMD_IS_APU &&
 	    adev->asic_type >= CHIP_CARRIZO &&
-	    adev->asic_type <= CHIP_RAVEN)
+	    adev->asic_type < CHIP_RAVEN)
 		init_data.flags.gpu_vm_support = true;
 
 	if (amdgpu_dc_feature_mask & DC_FBC_MASK)
@@ -697,6 +728,9 @@ static int amdgpu_dm_init(struct amdgpu_device *adev)
 	if (amdgpu_dc_feature_mask & DC_MULTI_MON_PP_MCLK_SWITCH_MASK)
 		init_data.flags.multi_mon_pp_mclk_switch = true;
 
+	if (amdgpu_dc_feature_mask & DC_DISABLE_FRACTIONAL_PWM_MASK)
+		init_data.flags.disable_fractional_pwm = true;
+
 	init_data.flags.power_down_display_on_boot = true;
 
 #ifdef CONFIG_DRM_AMD_DC_DCN2_0
@@ -713,6 +747,8 @@ static int amdgpu_dm_init(struct amdgpu_device *adev)
 		goto error;
 	}
 
+	dc_hardware_init(adev->dm.dc);
+
 	adev->dm.freesync_module = mod_freesync_create(adev->dm.dc);
 	if (!adev->dm.freesync_module) {
 		DRM_ERROR(
@@ -723,6 +759,18 @@ static int amdgpu_dm_init(struct amdgpu_device *adev)
 
 	amdgpu_dm_init_color_mod();
 
+#ifdef CONFIG_DRM_AMD_DC_HDCP
+	if (adev->asic_type >= CHIP_RAVEN) {
+		adev->dm.hdcp_workqueue = hdcp_create_workqueue(&adev->psp, &init_params.cp_psp, adev->dm.dc);
+
+		if (!adev->dm.hdcp_workqueue)
+			DRM_ERROR("amdgpu: failed to initialize hdcp_workqueue.\n");
+		else
+			DRM_DEBUG_DRIVER("amdgpu: hdcp_workqueue init done %p.\n", adev->dm.hdcp_workqueue);
+
+		dc_init_callbacks(adev->dm.dc, &init_params);
+	}
+#endif
 	if (amdgpu_dm_initialize_drm_device(adev)) {
 		DRM_ERROR(
 		"amdgpu: failed to initialize sw for display support.\n");
@@ -764,6 +812,16 @@ static void amdgpu_dm_fini(struct amdgpu_device *adev)
 
 	amdgpu_dm_destroy_drm_device(&adev->dm);
 
+#ifdef CONFIG_DRM_AMD_DC_HDCP
+	if (adev->dm.hdcp_workqueue) {
+		hdcp_destroy(adev->dm.hdcp_workqueue);
+		adev->dm.hdcp_workqueue = NULL;
+	}
+
+	if (adev->dm.dc)
+		dc_deinit_callbacks(adev->dm.dc);
+#endif
+
 	/* DC Destroy TODO: Replace destroy DAL */
 	if (adev->dm.dc)
 		dc_destroy(&adev->dm.dc);
@@ -897,27 +955,29 @@ static int detect_mst_link_for_all_connectors(struct drm_device *dev)
 {
 	struct amdgpu_dm_connector *aconnector;
 	struct drm_connector *connector;
+	struct drm_connector_list_iter iter;
 	int ret = 0;
 
-	drm_modeset_lock(&dev->mode_config.connection_mutex, NULL);
-
-	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
+	drm_connector_list_iter_begin(dev, &iter);
+	drm_for_each_connector_iter(connector, &iter) {
 		aconnector = to_amdgpu_dm_connector(connector);
 		if (aconnector->dc_link->type == dc_connection_mst_branch &&
 		    aconnector->mst_mgr.aux) {
 			DRM_DEBUG_DRIVER("DM_MST: starting TM on aconnector: %p [id: %d]\n",
-					aconnector, aconnector->base.base.id);
+					 aconnector,
+					 aconnector->base.base.id);
 
 			ret = drm_dp_mst_topology_mgr_set_mst(&aconnector->mst_mgr, true);
 			if (ret < 0) {
 				DRM_ERROR("DM_MST: Failed to start MST\n");
-				((struct dc_link *)aconnector->dc_link)->type = dc_connection_single;
-				return ret;
-				}
+				aconnector->dc_link->type =
+					dc_connection_single;
+				break;
 			}
+		}
 	}
+	drm_connector_list_iter_end(&iter);
 
-	drm_modeset_unlock(&dev->mode_config.connection_mutex);
 	return ret;
 }
 
@@ -940,6 +1000,11 @@ static int dm_late_init(void *handle)
 	params.backlight_lut_array_size = 16;
 	params.backlight_lut_array = linear_lut;
 
+	/* Min backlight level after ABM reduction,  Don't allow below 1%
+	 * 0xFFFF x 0.01 = 0x28F
+	 */
+	params.min_abm_backlight = 0x28F;
+
 	/* todo will enable for navi10 */
 	if (adev->asic_type <= CHIP_RAVEN) {
 		ret = dmcu_load_iram(dmcu, params);
@@ -955,14 +1020,13 @@ static void s3_handle_mst(struct drm_device *dev, bool suspend)
 {
 	struct amdgpu_dm_connector *aconnector;
 	struct drm_connector *connector;
+	struct drm_connector_list_iter iter;
 	struct drm_dp_mst_topology_mgr *mgr;
 	int ret;
 	bool need_hotplug = false;
 
-	drm_modeset_lock(&dev->mode_config.connection_mutex, NULL);
-
-	list_for_each_entry(connector, &dev->mode_config.connector_list,
-			    head) {
+	drm_connector_list_iter_begin(dev, &iter);
+	drm_for_each_connector_iter(connector, &iter) {
 		aconnector = to_amdgpu_dm_connector(connector);
 		if (aconnector->dc_link->type != dc_connection_mst_branch ||
 		    aconnector->mst_port)
@@ -973,15 +1037,14 @@ static void s3_handle_mst(struct drm_device *dev, bool suspend)
 		if (suspend) {
 			drm_dp_mst_topology_mgr_suspend(mgr);
 		} else {
-			ret = drm_dp_mst_topology_mgr_resume(mgr);
+			ret = drm_dp_mst_topology_mgr_resume(mgr, true);
 			if (ret < 0) {
 				drm_dp_mst_topology_mgr_set_mst(mgr, false);
 				need_hotplug = true;
 			}
 		}
 	}
-
-	drm_modeset_unlock(&dev->mode_config.connection_mutex);
+	drm_connector_list_iter_end(&iter);
 
 	if (need_hotplug)
 		drm_kms_helper_hotplug_event(dev);
@@ -989,7 +1052,7 @@ static void s3_handle_mst(struct drm_device *dev, bool suspend)
 
 /**
  * dm_hw_init() - Initialize DC device
- * @handle: The base driver device containing the amdpgu_dm device.
+ * @handle: The base driver device containing the amdgpu_dm device.
  *
  * Initialize the &struct amdgpu_display_manager device. This involves calling
  * the initializers of each DM component, then populating the struct with them.
@@ -1019,7 +1082,7 @@ static int dm_hw_init(void *handle)
 
 /**
  * dm_hw_fini() - Teardown DC device
- * @handle: The base driver device containing the amdpgu_dm device.
+ * @handle: The base driver device containing the amdgpu_dm device.
  *
  * Teardown components within &struct amdgpu_display_manager that require
  * cleanup. This involves cleaning up the DRM device, DC, and any modules that
@@ -1163,6 +1226,7 @@ static int dm_resume(void *handle)
 	struct amdgpu_display_manager *dm = &adev->dm;
 	struct amdgpu_dm_connector *aconnector;
 	struct drm_connector *connector;
+	struct drm_connector_list_iter iter;
 	struct drm_crtc *crtc;
 	struct drm_crtc_state *new_crtc_state;
 	struct dm_crtc_state *dm_new_crtc_state;
@@ -1185,17 +1249,18 @@ static int dm_resume(void *handle)
 	/* program HPD filter */
 	dc_resume(dm->dc);
 
-	/* On resume we need to  rewrite the MSTM control bits to enamble MST*/
-	s3_handle_mst(ddev, false);
-
 	/*
 	 * early enable HPD Rx IRQ, should be done before set mode as short
 	 * pulse interrupts are used for MST
 	 */
 	amdgpu_dm_irq_resume_early(adev);
 
+	/* On resume we need to rewrite the MSTM control bits to enable MST*/
+	s3_handle_mst(ddev, false);
+
 	/* Do detection*/
-	list_for_each_entry(connector, &ddev->mode_config.connector_list, head) {
+	drm_connector_list_iter_begin(ddev, &iter);
+	drm_for_each_connector_iter(connector, &iter) {
 		aconnector = to_amdgpu_dm_connector(connector);
 
 		/*
@@ -1223,6 +1288,7 @@ static int dm_resume(void *handle)
 		amdgpu_dm_update_connector_after_detect(aconnector);
 		mutex_unlock(&aconnector->hpd_lock);
 	}
+	drm_connector_list_iter_end(&iter);
 
 	/* Force mode set in atomic commit */
 	for_each_new_crtc_in_state(dm->cached_state, crtc, new_crtc_state, i)
@@ -1438,6 +1504,11 @@ amdgpu_dm_update_connector_after_detect(struct amdgpu_dm_connector *aconnector)
 		dc_sink_release(aconnector->dc_sink);
 		aconnector->dc_sink = NULL;
 		aconnector->edid = NULL;
+#ifdef CONFIG_DRM_AMD_DC_HDCP
+		/* Set CP to DESIRED if it was ENABLED, so we can re-enable it again on hotplug */
+		if (connector->state->content_protection == DRM_MODE_CONTENT_PROTECTION_ENABLED)
+			connector->state->content_protection = DRM_MODE_CONTENT_PROTECTION_DESIRED;
+#endif
 	}
 
 	mutex_unlock(&dev->mode_config.mutex);
@@ -1452,6 +1523,9 @@ static void handle_hpd_irq(void *param)
 	struct drm_connector *connector = &aconnector->base;
 	struct drm_device *dev = connector->dev;
 	enum dc_connection_type new_connection_type = dc_connection_none;
+#ifdef CONFIG_DRM_AMD_DC_HDCP
+	struct amdgpu_device *adev = dev->dev_private;
+#endif
 
 	/*
 	 * In case of failure or MST no need to update connector status or notify the OS
@@ -1459,6 +1533,10 @@ static void handle_hpd_irq(void *param)
 	 */
 	mutex_lock(&aconnector->hpd_lock);
 
+#ifdef CONFIG_DRM_AMD_DC_HDCP
+	if (adev->asic_type >= CHIP_RAVEN)
+		hdcp_reset_display(adev->dm.hdcp_workqueue, aconnector->dc_link->link_index);
+#endif
 	if (aconnector->fake_enable)
 		aconnector->fake_enable = false;
 
@@ -1577,6 +1655,12 @@ static void handle_hpd_rx_irq(void *param)
 	struct dc_link *dc_link = aconnector->dc_link;
 	bool is_mst_root_connector = aconnector->mst_mgr.mst_state;
 	enum dc_connection_type new_connection_type = dc_connection_none;
+#ifdef CONFIG_DRM_AMD_DC_HDCP
+	union hpd_irq_data hpd_irq_data;
+	struct amdgpu_device *adev = dev->dev_private;
+
+	memset(&hpd_irq_data, 0, sizeof(hpd_irq_data));
+#endif
 
 	/*
 	 * TODO:Temporary add mutex to protect hpd interrupt not have a gpio
@@ -1586,7 +1670,12 @@ static void handle_hpd_rx_irq(void *param)
 	if (dc_link->type != dc_connection_mst_branch)
 		mutex_lock(&aconnector->hpd_lock);
 
+
+#ifdef CONFIG_DRM_AMD_DC_HDCP
+	if (dc_link_handle_hpd_rx_irq(dc_link, &hpd_irq_data, NULL) &&
+#else
 	if (dc_link_handle_hpd_rx_irq(dc_link, NULL, NULL) &&
+#endif
 			!is_mst_root_connector) {
 		/* Downstream Port status changed. */
 		if (!dc_link_detect_sink(dc_link, &new_connection_type))
@@ -1621,6 +1710,10 @@ static void handle_hpd_rx_irq(void *param)
 			drm_kms_helper_hotplug_event(dev);
 		}
 	}
+#ifdef CONFIG_DRM_AMD_DC_HDCP
+	if (hpd_irq_data.bytes.device_service_irq.bits.CP_IRQ)
+		hdcp_handle_cpirq(adev->dm.hdcp_workqueue,  aconnector->base.index);
+#endif
 	if ((dc_link->cur_link_settings.lane_count != LANE_COUNT_UNKNOWN) ||
 	    (dc_link->type == dc_connection_mst_branch))
 		dm_handle_hpd_rx_irq(aconnector);
@@ -2334,6 +2427,8 @@ static int amdgpu_dm_initialize_drm_device(struct amdgpu_device *adev)
 		} else if (dc_link_detect(link, DETECT_REASON_BOOT)) {
 			amdgpu_dm_update_connector_after_detect(aconnector);
 			register_backlight_device(dm, link);
+			if (amdgpu_dc_feature_mask & DC_PSR_MASK)
+				amdgpu_dm_set_psr_caps(link);
 		}
 
 
@@ -2385,8 +2480,6 @@ static int amdgpu_dm_initialize_drm_device(struct amdgpu_device *adev)
 
 	if (adev->asic_type != CHIP_CARRIZO && adev->asic_type != CHIP_STONEY)
 		dm->dc->debug.disable_stutter = amdgpu_pp_feature_mask & PP_STUTTER_MODE ? false : true;
-	if (adev->asic_type == CHIP_RENOIR)
-		dm->dc->debug.disable_stutter = true;
 
 	return 0;
 fail:
@@ -3313,8 +3406,12 @@ static void fill_stream_properties_from_drm_display_mode(
 {
 	struct dc_crtc_timing *timing_out = &stream->timing;
 	const struct drm_display_info *info = &connector->display_info;
+	struct amdgpu_dm_connector *aconnector = to_amdgpu_dm_connector(connector);
+	struct hdmi_vendor_infoframe hv_frame;
+	struct hdmi_avi_infoframe avi_frame;
 
-	memset(timing_out, 0, sizeof(struct dc_crtc_timing));
+	memset(&hv_frame, 0, sizeof(hv_frame));
+	memset(&avi_frame, 0, sizeof(avi_frame));
 
 	timing_out->h_border_left = 0;
 	timing_out->h_border_right = 0;
@@ -3324,6 +3421,9 @@ static void fill_stream_properties_from_drm_display_mode(
 	if (drm_mode_is_420_only(info, mode_in)
 			&& stream->signal == SIGNAL_TYPE_HDMI_TYPE_A)
 		timing_out->pixel_encoding = PIXEL_ENCODING_YCBCR420;
+	else if (drm_mode_is_420_also(info, mode_in)
+			&& aconnector->force_yuv420_output)
+		timing_out->pixel_encoding = PIXEL_ENCODING_YCBCR420;
 	else if ((connector->display_info.color_formats & DRM_COLOR_FORMAT_YCRCB444)
 			&& stream->signal == SIGNAL_TYPE_HDMI_TYPE_A)
 		timing_out->pixel_encoding = PIXEL_ENCODING_YCBCR444;
@@ -3348,6 +3448,13 @@ static void fill_stream_properties_from_drm_display_mode(
 			timing_out->flags.VSYNC_POSITIVE_POLARITY = 1;
 	}
 
+	if (stream->signal == SIGNAL_TYPE_HDMI_TYPE_A) {
+		drm_hdmi_avi_infoframe_from_display_mode(&avi_frame, (struct drm_connector *)connector, mode_in);
+		timing_out->vic = avi_frame.video_code;
+		drm_hdmi_vendor_infoframe_from_display_mode(&hv_frame, (struct drm_connector *)connector, mode_in);
+		timing_out->hdmi_vic = hv_frame.vic;
+	}
+
 	timing_out->h_addressable = mode_in->crtc_hdisplay;
 	timing_out->h_total = mode_in->crtc_htotal;
 	timing_out->h_sync_width =
@@ -3568,6 +3675,9 @@ create_stream_for_sink(struct amdgpu_dm_connector *aconnector,
 
 	stream->dm_stream_context = aconnector;
 
+	stream->timing.flags.LTE_340MCSC_SCRAMBLE =
+		drm_connector->display_info.hdmi.scdc.scrambling.low_rates;
+
 	list_for_each_entry(preferred_mode, &aconnector->base.modes, head) {
 		/* Search for preferred mode */
 		if (preferred_mode->type & DRM_MODE_TYPE_PREFERRED) {
@@ -3623,8 +3733,9 @@ create_stream_for_sink(struct amdgpu_dm_connector *aconnector,
 							     dc_link_get_link_cap(aconnector->dc_link));
 
 		if (dsc_caps.is_dsc_supported)
-			if (dc_dsc_compute_config(aconnector->dc_link->ctx->dc,
+			if (dc_dsc_compute_config(aconnector->dc_link->ctx->dc->res_pool->dscs[0],
 						  &dsc_caps,
+						  aconnector->dc_link->ctx->dc->debug.dsc_min_slice_height_override,
 						  link_bandwidth_kbps,
 						  &stream->timing,
 						  &stream->timing.dsc_cfg))
@@ -3641,6 +3752,18 @@ create_stream_for_sink(struct amdgpu_dm_connector *aconnector,
 
 	update_stream_signal(stream, sink);
 
+	if (stream->signal == SIGNAL_TYPE_HDMI_TYPE_A)
+		mod_build_hf_vsif_infopacket(stream, &stream->vsp_infopacket, false, false);
+	if (stream->link->psr_feature_enabled)	{
+		struct dc  *core_dc = stream->link->ctx->dc;
+
+		if (dc_is_dmcu_initialized(core_dc)) {
+			struct dmcu *dmcu = core_dc->res_pool->dmcu;
+
+			stream->psr_version = dmcu->dmcu_version.psr_version;
+			mod_build_vsc_infopacket(stream, &stream->vsc_infopacket);
+		}
+	}
 finish:
 	dc_sink_release(sink);
 
@@ -4116,8 +4239,8 @@ enum drm_mode_status amdgpu_dm_connector_mode_valid(struct drm_connector *connec
 		result = MODE_OK;
 	else
 		DRM_DEBUG_KMS("Mode %dx%d (clk %d) failed DC validation with error %d\n",
-			      mode->vdisplay,
 			      mode->hdisplay,
+			      mode->vdisplay,
 			      mode->clock,
 			      dc_result);
 
@@ -4496,7 +4619,7 @@ static int dm_plane_helper_prepare_fb(struct drm_plane *plane,
 	tv.num_shared = 1;
 	list_add(&tv.head, &list);
 
-	r = ttm_eu_reserve_buffers(&ticket, &list, false, NULL, true);
+	r = ttm_eu_reserve_buffers(&ticket, &list, false, NULL);
 	if (r) {
 		dev_err(adev->dev, "fail to reserve bo (%d)\n", r);
 		return r;
@@ -4839,7 +4962,13 @@ static int to_drm_connector_type(enum signal_type st)
 
 static struct drm_encoder *amdgpu_dm_connector_to_encoder(struct drm_connector *connector)
 {
-	return drm_encoder_find(connector->dev, NULL, connector->encoder_ids[0]);
+	struct drm_encoder *encoder;
+
+	/* There is only one encoder per connector */
+	drm_connector_for_each_possible_encoder(connector, encoder)
+		return encoder;
+
+	return NULL;
 }
 
 static void amdgpu_dm_get_native_mode(struct drm_connector *connector)
@@ -5084,6 +5213,10 @@ void amdgpu_dm_connector_init_helper(struct amdgpu_display_manager *dm,
 
 		drm_connector_attach_vrr_capable_property(
 			&aconnector->base);
+#ifdef CONFIG_DRM_AMD_DC_HDCP
+		if (adev->asic_type >= CHIP_RAVEN)
+			drm_connector_attach_content_protection_property(&aconnector->base, false);
+#endif
 	}
 }
 
@@ -5326,6 +5459,53 @@ is_scaling_state_different(const struct dm_connector_state *dm_state,
 	return false;
 }
 
+#ifdef CONFIG_DRM_AMD_DC_HDCP
+static bool is_content_protection_different(struct drm_connector_state *state,
+					    const struct drm_connector_state *old_state,
+					    const struct drm_connector *connector, struct hdcp_workqueue *hdcp_w)
+{
+	struct amdgpu_dm_connector *aconnector = to_amdgpu_dm_connector(connector);
+
+	/* CP is being re enabled, ignore this */
+	if (old_state->content_protection == DRM_MODE_CONTENT_PROTECTION_ENABLED &&
+	    state->content_protection == DRM_MODE_CONTENT_PROTECTION_DESIRED) {
+		state->content_protection = DRM_MODE_CONTENT_PROTECTION_ENABLED;
+		return false;
+	}
+
+	/* S3 resume case, since old state will always be 0 (UNDESIRED) and the restored state will be ENABLED */
+	if (old_state->content_protection == DRM_MODE_CONTENT_PROTECTION_UNDESIRED &&
+	    state->content_protection == DRM_MODE_CONTENT_PROTECTION_ENABLED)
+		state->content_protection = DRM_MODE_CONTENT_PROTECTION_DESIRED;
+
+	/* Check if something is connected/enabled, otherwise we start hdcp but nothing is connected/enabled
+	 * hot-plug, headless s3, dpms
+	 */
+	if (state->content_protection == DRM_MODE_CONTENT_PROTECTION_DESIRED && connector->dpms == DRM_MODE_DPMS_ON &&
+	    aconnector->dc_sink != NULL)
+		return true;
+
+	if (old_state->content_protection == state->content_protection)
+		return false;
+
+	if (state->content_protection == DRM_MODE_CONTENT_PROTECTION_UNDESIRED)
+		return true;
+
+	return false;
+}
+
+static void update_content_protection(struct drm_connector_state *state, const struct drm_connector *connector,
+				      struct hdcp_workqueue *hdcp_w)
+{
+	struct amdgpu_dm_connector *aconnector = to_amdgpu_dm_connector(connector);
+
+	if (state->content_protection == DRM_MODE_CONTENT_PROTECTION_DESIRED)
+		hdcp_add_display(hdcp_w, aconnector->dc_link->link_index, aconnector);
+	else if (state->content_protection == DRM_MODE_CONTENT_PROTECTION_UNDESIRED)
+		hdcp_remove_display(hdcp_w, aconnector->dc_link->link_index, aconnector->base.index);
+
+}
+#endif
 static void remove_stream(struct amdgpu_device *adev,
 			  struct amdgpu_crtc *acrtc,
 			  struct dc_stream_state *stream)
@@ -5667,6 +5847,7 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state,
 	uint32_t target_vblank, last_flip_vblank;
 	bool vrr_active = amdgpu_dm_vrr_active(acrtc_state);
 	bool pflip_present = false;
+	bool swizzle = true;
 	struct {
 		struct dc_surface_update surface_updates[MAX_SURFACES];
 		struct dc_plane_info plane_infos[MAX_SURFACES];
@@ -5712,6 +5893,9 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state,
 
 		dc_plane = dm_new_plane_state->dc_state;
 
+		if (dc_plane && !dc_plane->tiling_info.gfx9.swizzle)
+			swizzle = false;
+
 		bundle->surface_updates[planes_count].surface = dc_plane;
 		if (new_pcrtc_state->color_mgmt_changed) {
 			bundle->surface_updates[planes_count].gamma = dc_plane->gamma_correction;
@@ -5866,6 +6050,7 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state,
 	/* Update the planes if changed or disable if we don't have any. */
 	if ((planes_count || acrtc_state->active_planes == 0) &&
 		acrtc_state->stream) {
+		bundle->stream_update.stream = acrtc_state->stream;
 		if (new_pcrtc_state->mode_changed) {
 			bundle->stream_update.src = acrtc_state->stream->src;
 			bundle->stream_update.dst = acrtc_state->stream->dst;
@@ -5901,14 +6086,29 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state,
 				&acrtc_state->vrr_params.adjust);
 			spin_unlock_irqrestore(&pcrtc->dev->event_lock, flags);
 		}
-
 		mutex_lock(&dm->dc_lock);
+		if ((acrtc_state->update_type > UPDATE_TYPE_FAST) &&
+				acrtc_state->stream->link->psr_allow_active)
+			amdgpu_dm_psr_disable(acrtc_state->stream);
+
 		dc_commit_updates_for_stream(dm->dc,
 						     bundle->surface_updates,
 						     planes_count,
 						     acrtc_state->stream,
 						     &bundle->stream_update,
 						     dc_state);
+
+		if ((acrtc_state->update_type > UPDATE_TYPE_FAST) &&
+						acrtc_state->stream->psr_version &&
+						!acrtc_state->stream->link->psr_feature_enabled)
+			amdgpu_dm_link_setup_psr(acrtc_state->stream);
+		else if ((acrtc_state->update_type == UPDATE_TYPE_FAST) &&
+						acrtc_state->stream->link->psr_feature_enabled &&
+						!acrtc_state->stream->link->psr_allow_active &&
+						swizzle) {
+			amdgpu_dm_psr_enable(acrtc_state->stream);
+		}
+
 		mutex_unlock(&dm->dc_lock);
 	}
 
@@ -6019,7 +6219,9 @@ static void amdgpu_dm_enable_crtc_interrupts(struct drm_device *dev,
 	struct drm_crtc *crtc;
 	struct drm_crtc_state *old_crtc_state, *new_crtc_state;
 	int i;
+#ifdef CONFIG_DEBUG_FS
 	enum amdgpu_dm_pipe_crc_source source;
+#endif
 
 	for_each_oldnew_crtc_in_state(state, crtc, old_crtc_state,
 				      new_crtc_state, i) {
@@ -6215,10 +6417,13 @@ static void amdgpu_dm_atomic_commit_tail(struct drm_atomic_state *state)
 			crtc->hwmode = new_crtc_state->mode;
 		} else if (modereset_required(new_crtc_state)) {
 			DRM_DEBUG_DRIVER("Atomic commit: RESET. crtc id %d:[%p]\n", acrtc->crtc_id, acrtc);
-
 			/* i.e. reset mode */
-			if (dm_old_crtc_state->stream)
+			if (dm_old_crtc_state->stream) {
+				if (dm_old_crtc_state->stream->link->psr_allow_active)
+					amdgpu_dm_psr_disable(dm_old_crtc_state->stream);
+
 				remove_stream(adev, acrtc, dm_old_crtc_state->stream);
+			}
 		}
 	} /* for_each_crtc_in_state() */
 
@@ -6248,6 +6453,30 @@ static void amdgpu_dm_atomic_commit_tail(struct drm_atomic_state *state)
 				acrtc->otg_inst = status->primary_otg_inst;
 		}
 	}
+#ifdef CONFIG_DRM_AMD_DC_HDCP
+	for_each_oldnew_connector_in_state(state, connector, old_con_state, new_con_state, i) {
+		struct dm_connector_state *dm_new_con_state = to_dm_connector_state(new_con_state);
+		struct amdgpu_crtc *acrtc = to_amdgpu_crtc(dm_new_con_state->base.crtc);
+		struct amdgpu_dm_connector *aconnector = to_amdgpu_dm_connector(connector);
+
+		new_crtc_state = NULL;
+
+		if (acrtc)
+			new_crtc_state = drm_atomic_get_new_crtc_state(state, &acrtc->base);
+
+		dm_new_crtc_state = to_dm_crtc_state(new_crtc_state);
+
+		if (dm_new_crtc_state && dm_new_crtc_state->stream == NULL &&
+		    connector->state->content_protection == DRM_MODE_CONTENT_PROTECTION_ENABLED) {
+			hdcp_reset_display(adev->dm.hdcp_workqueue, aconnector->dc_link->link_index);
+			new_con_state->content_protection = DRM_MODE_CONTENT_PROTECTION_DESIRED;
+			continue;
+		}
+
+		if (is_content_protection_different(new_con_state, old_con_state, connector, adev->dm.hdcp_workqueue))
+			update_content_protection(new_con_state, connector, adev->dm.hdcp_workqueue);
+	}
+#endif
 
 	/* Handle connector state changes */
 	for_each_oldnew_connector_in_state(state, connector, old_con_state, new_con_state, i) {
@@ -6287,9 +6516,10 @@ static void amdgpu_dm_atomic_commit_tail(struct drm_atomic_state *state)
 		if (!scaling_changed && !abm_changed && !hdr_changed)
 			continue;
 
+		stream_update.stream = dm_new_crtc_state->stream;
 		if (scaling_changed) {
 			update_stream_scaling_settings(&dm_new_con_state->base.crtc->mode,
-					dm_new_con_state, (struct dc_stream_state *)dm_new_crtc_state->stream);
+					dm_new_con_state, dm_new_crtc_state->stream);
 
 			stream_update.src = dm_new_crtc_state->stream->src;
 			stream_update.dst = dm_new_crtc_state->stream->dst;
@@ -7158,7 +7388,7 @@ dm_determine_update_type_for_commit(struct amdgpu_display_manager *dm,
 
 		status = dc_stream_get_status_from_state(old_dm_state->context,
 							 new_dm_crtc_state->stream);
-
+		stream_update.stream = new_dm_crtc_state->stream;
 		/*
 		 * TODO: DC modifies the surface during this call so we need
 		 * to lock here - find a way to do this without locking.
@@ -7569,3 +7799,92 @@ update:
 						       freesync_capable);
 }
 
+static void amdgpu_dm_set_psr_caps(struct dc_link *link)
+{
+	uint8_t dpcd_data[EDP_PSR_RECEIVER_CAP_SIZE];
+
+	if (!(link->connector_signal & SIGNAL_TYPE_EDP))
+		return;
+	if (link->type == dc_connection_none)
+		return;
+	if (dm_helpers_dp_read_dpcd(NULL, link, DP_PSR_SUPPORT,
+					dpcd_data, sizeof(dpcd_data))) {
+		link->psr_feature_enabled = dpcd_data[0] ? true:false;
+		DRM_INFO("PSR support:%d\n", link->psr_feature_enabled);
+	}
+}
+
+/*
+ * amdgpu_dm_link_setup_psr() - configure psr link
+ * @stream: stream state
+ *
+ * Return: true if success
+ */
+static bool amdgpu_dm_link_setup_psr(struct dc_stream_state *stream)
+{
+	struct dc_link *link = NULL;
+	struct psr_config psr_config = {0};
+	struct psr_context psr_context = {0};
+	struct dc *dc = NULL;
+	bool ret = false;
+
+	if (stream == NULL)
+		return false;
+
+	link = stream->link;
+	dc = link->ctx->dc;
+
+	psr_config.psr_version = dc->res_pool->dmcu->dmcu_version.psr_version;
+
+	if (psr_config.psr_version > 0) {
+		psr_config.psr_exit_link_training_required = 0x1;
+		psr_config.psr_frame_capture_indication_req = 0;
+		psr_config.psr_rfb_setup_time = 0x37;
+		psr_config.psr_sdp_transmit_line_num_deadline = 0x20;
+		psr_config.allow_smu_optimizations = 0x0;
+
+		ret = dc_link_setup_psr(link, stream, &psr_config, &psr_context);
+
+	}
+	DRM_DEBUG_DRIVER("PSR link: %d\n",	link->psr_feature_enabled);
+
+	return ret;
+}
+
+/*
+ * amdgpu_dm_psr_enable() - enable psr f/w
+ * @stream: stream state
+ *
+ * Return: true if success
+ */
+bool amdgpu_dm_psr_enable(struct dc_stream_state *stream)
+{
+	struct dc_link *link = stream->link;
+	struct dc_static_screen_events triggers = {0};
+
+	DRM_DEBUG_DRIVER("Enabling psr...\n");
+
+	triggers.cursor_update = true;
+	triggers.overlay_update = true;
+	triggers.surface_update = true;
+
+	dc_stream_set_static_screen_events(link->ctx->dc,
+					   &stream, 1,
+					   &triggers);
+
+	return dc_link_set_psr_allow_active(link, true, false);
+}
+
+/*
+ * amdgpu_dm_psr_disable() - disable psr f/w
+ * @stream:  stream state
+ *
+ * Return: true if success
+ */
+static bool amdgpu_dm_psr_disable(struct dc_stream_state *stream)
+{
+
+	DRM_DEBUG_DRIVER("Disabling psr...\n");
+
+	return dc_link_set_psr_allow_active(stream->link, false, true);
+}
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h
index c8c525a2b505..77c5166e6b08 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h
@@ -108,6 +108,12 @@ struct amdgpu_dm_backlight_caps {
  * @display_indexes_num: Max number of display streams supported
  * @irq_handler_list_table_lock: Synchronizes access to IRQ tables
  * @backlight_dev: Backlight control device
+ * @backlight_link: Link on which to control backlight
+ * @backlight_caps: Capabilities of the backlight device
+ * @freesync_module: Module handling freesync calculations
+ * @fw_dmcu: Reference to DMCU firmware
+ * @dmcu_fw_version: Version of the DMCU firmware
+ * @soc_bounding_box: SOC bounding box values provided by gpu_info FW
  * @cached_state: Caches device atomic state for suspend/resume
  * @compressor: Frame buffer compression buffer. See &struct dm_comressor_info
  */
@@ -128,7 +134,7 @@ struct amdgpu_display_manager {
 	u16 display_indexes_num;
 
 	/**
-	 * @atomic_obj
+	 * @atomic_obj:
 	 *
 	 * In combination with &dm_atomic_state it helps manage
 	 * global atomic state that doesn't map cleanly into existing
@@ -225,6 +231,9 @@ struct amdgpu_display_manager {
 	struct amdgpu_dm_backlight_caps backlight_caps;
 
 	struct mod_freesync *freesync_module;
+#ifdef CONFIG_DRM_AMD_DC_HDCP
+	struct hdcp_workqueue *hdcp_workqueue;
+#endif
 
 	struct drm_atomic_state *cached_state;
 
@@ -234,6 +243,8 @@ struct amdgpu_display_manager {
 	uint32_t dmcu_fw_version;
 #ifdef CONFIG_DRM_AMD_DC_DCN2_0
 	/**
+	 * @soc_bounding_box:
+	 *
 	 * gpu_info FW provided soc bounding box struct or 0 if not
 	 * available in FW
 	 */
@@ -287,6 +298,7 @@ struct amdgpu_dm_connector {
 	uint32_t debugfs_dpcd_address;
 	uint32_t debugfs_dpcd_size;
 #endif
+	bool force_yuv420_output;
 };
 
 #define to_amdgpu_dm_connector(x) container_of(x, struct amdgpu_dm_connector, base)
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
index b43bb7f90e4e..2233d293a707 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
@@ -210,6 +210,8 @@ static int __set_legacy_tf(struct dc_transfer_func *func,
 	res = mod_color_calculate_regamma_params(func, gamma, true, has_rom,
 						 NULL);
 
+	dc_gamma_release(&gamma);
+
 	return res ? 0 : -ENOMEM;
 }
 
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crc.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crc.c
index a549c7c717dd..eaad9099bc0b 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crc.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crc.c
@@ -122,11 +122,16 @@ int amdgpu_dm_crtc_configure_crc_source(struct drm_crtc *crtc,
 	}
 
 	/* Configure dithering */
-	if (!dm_need_crc_dither(source))
+	if (!dm_need_crc_dither(source)) {
 		dc_stream_set_dither_option(stream_state, DITHER_OPTION_TRUN8);
-	else
+		dc_stream_set_dyn_expansion(stream_state->ctx->dc, stream_state,
+					    DYN_EXPANSION_DISABLE);
+	} else {
 		dc_stream_set_dither_option(stream_state,
 					    DITHER_OPTION_DEFAULT);
+		dc_stream_set_dyn_expansion(stream_state->ctx->dc, stream_state,
+					    DYN_EXPANSION_AUTO);
+	}
 
 unlock:
 	mutex_unlock(&adev->dm.dc_lock);
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c
index f3dfb2887ae0..bdb37e611015 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c
@@ -942,6 +942,52 @@ static const struct {
 		{"aux_dpcd_data", &dp_dpcd_data_debugfs_fops}
 };
 
+/*
+ * Force YUV420 output if available from the given mode
+ */
+static int force_yuv420_output_set(void *data, u64 val)
+{
+	struct amdgpu_dm_connector *connector = data;
+
+	connector->force_yuv420_output = (bool)val;
+
+	return 0;
+}
+
+/*
+ * Check if YUV420 is forced when available from the given mode
+ */
+static int force_yuv420_output_get(void *data, u64 *val)
+{
+	struct amdgpu_dm_connector *connector = data;
+
+	*val = connector->force_yuv420_output;
+
+	return 0;
+}
+
+DEFINE_DEBUGFS_ATTRIBUTE(force_yuv420_output_fops, force_yuv420_output_get,
+			 force_yuv420_output_set, "%llu\n");
+
+/*
+ *  Read PSR state
+ */
+static int psr_get(void *data, u64 *val)
+{
+	struct amdgpu_dm_connector *connector = data;
+	struct dc_link *link = connector->dc_link;
+	uint32_t psr_state = 0;
+
+	dc_link_get_psr_state(link, &psr_state);
+
+	*val = psr_state;
+
+	return 0;
+}
+
+
+DEFINE_DEBUGFS_ATTRIBUTE(psr_fops, psr_get, NULL, "%llu\n");
+
 void connector_debugfs_init(struct amdgpu_dm_connector *connector)
 {
 	int i;
@@ -955,6 +1001,12 @@ void connector_debugfs_init(struct amdgpu_dm_connector *connector)
 					    dp_debugfs_entries[i].fops);
 		}
 	}
+	if (connector->base.connector_type == DRM_MODE_CONNECTOR_eDP)
+		debugfs_create_file_unsafe("psr_state", 0444, dir, connector, &psr_fops);
+
+	debugfs_create_file_unsafe("force_yuv420_output", 0644, dir, connector,
+				   &force_yuv420_output_fops);
+
 }
 
 /*
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_hdcp.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_hdcp.c
new file mode 100644
index 000000000000..77181ddf6c8e
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_hdcp.c
@@ -0,0 +1,346 @@
+/*
+ * Copyright 2019 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#include "amdgpu_dm_hdcp.h"
+#include "amdgpu.h"
+#include "amdgpu_dm.h"
+#include "dm_helpers.h"
+#include <drm/drm_hdcp.h>
+
+static bool
+lp_write_i2c(void *handle, uint32_t address, const uint8_t *data, uint32_t size)
+{
+
+	struct dc_link *link = handle;
+	struct i2c_payload i2c_payloads[] = {{true, address, size, (void *)data} };
+	struct i2c_command cmd = {i2c_payloads, 1, I2C_COMMAND_ENGINE_HW, link->dc->caps.i2c_speed_in_khz};
+
+	return dm_helpers_submit_i2c(link->ctx, link, &cmd);
+}
+
+static bool
+lp_read_i2c(void *handle, uint32_t address, uint8_t offset, uint8_t *data, uint32_t size)
+{
+	struct dc_link *link = handle;
+
+	struct i2c_payload i2c_payloads[] = {{true, address, 1, &offset}, {false, address, size, data} };
+	struct i2c_command cmd = {i2c_payloads, 2, I2C_COMMAND_ENGINE_HW, link->dc->caps.i2c_speed_in_khz};
+
+	return dm_helpers_submit_i2c(link->ctx, link, &cmd);
+}
+
+static bool
+lp_write_dpcd(void *handle, uint32_t address, const uint8_t *data, uint32_t size)
+{
+	struct dc_link *link = handle;
+
+	return dm_helpers_dp_write_dpcd(link->ctx, link, address, data, size);
+}
+
+static bool
+lp_read_dpcd(void *handle, uint32_t address, uint8_t *data, uint32_t size)
+{
+	struct dc_link *link = handle;
+
+	return dm_helpers_dp_read_dpcd(link->ctx, link, address, data, size);
+}
+
+static void process_output(struct hdcp_workqueue *hdcp_work)
+{
+	struct mod_hdcp_output output = hdcp_work->output;
+
+	if (output.callback_stop)
+		cancel_delayed_work(&hdcp_work->callback_dwork);
+
+	if (output.callback_needed)
+		schedule_delayed_work(&hdcp_work->callback_dwork,
+				      msecs_to_jiffies(output.callback_delay));
+
+	if (output.watchdog_timer_stop)
+		cancel_delayed_work(&hdcp_work->watchdog_timer_dwork);
+
+	if (output.watchdog_timer_needed)
+		schedule_delayed_work(&hdcp_work->watchdog_timer_dwork,
+				      msecs_to_jiffies(output.watchdog_timer_delay));
+
+}
+
+void hdcp_add_display(struct hdcp_workqueue *hdcp_work, unsigned int link_index, struct amdgpu_dm_connector *aconnector)
+{
+	struct hdcp_workqueue *hdcp_w = &hdcp_work[link_index];
+	struct mod_hdcp_display *display = &hdcp_work[link_index].display;
+	struct mod_hdcp_link *link = &hdcp_work[link_index].link;
+
+	mutex_lock(&hdcp_w->mutex);
+	hdcp_w->aconnector = aconnector;
+
+	mod_hdcp_add_display(&hdcp_w->hdcp, link, display, &hdcp_w->output);
+
+	schedule_delayed_work(&hdcp_w->property_validate_dwork, msecs_to_jiffies(DRM_HDCP_CHECK_PERIOD_MS));
+
+	process_output(hdcp_w);
+
+	mutex_unlock(&hdcp_w->mutex);
+
+}
+
+void hdcp_remove_display(struct hdcp_workqueue *hdcp_work, unsigned int link_index,  unsigned int display_index)
+{
+	struct hdcp_workqueue *hdcp_w = &hdcp_work[link_index];
+
+	mutex_lock(&hdcp_w->mutex);
+
+	mod_hdcp_remove_display(&hdcp_w->hdcp, display_index, &hdcp_w->output);
+
+	cancel_delayed_work(&hdcp_w->property_validate_dwork);
+	hdcp_w->encryption_status = MOD_HDCP_ENCRYPTION_STATUS_HDCP_OFF;
+
+	process_output(hdcp_w);
+
+	mutex_unlock(&hdcp_w->mutex);
+
+}
+
+void hdcp_reset_display(struct hdcp_workqueue *hdcp_work, unsigned int link_index)
+{
+	struct hdcp_workqueue *hdcp_w = &hdcp_work[link_index];
+
+	mutex_lock(&hdcp_w->mutex);
+
+	mod_hdcp_reset_connection(&hdcp_w->hdcp,  &hdcp_w->output);
+
+	cancel_delayed_work(&hdcp_w->property_validate_dwork);
+	hdcp_w->encryption_status = MOD_HDCP_ENCRYPTION_STATUS_HDCP_OFF;
+
+	process_output(hdcp_w);
+
+	mutex_unlock(&hdcp_w->mutex);
+}
+
+void hdcp_handle_cpirq(struct hdcp_workqueue *hdcp_work, unsigned int link_index)
+{
+	struct hdcp_workqueue *hdcp_w = &hdcp_work[link_index];
+
+	schedule_work(&hdcp_w->cpirq_work);
+}
+
+
+
+
+static void event_callback(struct work_struct *work)
+{
+	struct hdcp_workqueue *hdcp_work;
+
+	hdcp_work = container_of(to_delayed_work(work), struct hdcp_workqueue,
+				      callback_dwork);
+
+	mutex_lock(&hdcp_work->mutex);
+
+	cancel_delayed_work(&hdcp_work->watchdog_timer_dwork);
+
+	mod_hdcp_process_event(&hdcp_work->hdcp, MOD_HDCP_EVENT_CALLBACK,
+			       &hdcp_work->output);
+
+	process_output(hdcp_work);
+
+	mutex_unlock(&hdcp_work->mutex);
+
+
+}
+static void event_property_update(struct work_struct *work)
+{
+
+	struct hdcp_workqueue *hdcp_work = container_of(work, struct hdcp_workqueue, property_update_work);
+	struct amdgpu_dm_connector *aconnector = hdcp_work->aconnector;
+	struct drm_device *dev = hdcp_work->aconnector->base.dev;
+	long ret;
+
+	drm_modeset_lock(&dev->mode_config.connection_mutex, NULL);
+	mutex_lock(&hdcp_work->mutex);
+
+
+	if (aconnector->base.state->commit) {
+		ret = wait_for_completion_interruptible_timeout(&aconnector->base.state->commit->hw_done, 10 * HZ);
+
+		if (ret == 0) {
+			DRM_ERROR("HDCP state unknown! Setting it to DESIRED");
+			hdcp_work->encryption_status = MOD_HDCP_ENCRYPTION_STATUS_HDCP_OFF;
+		}
+	}
+
+	if (hdcp_work->encryption_status == MOD_HDCP_ENCRYPTION_STATUS_HDCP1_ON)
+		drm_hdcp_update_content_protection(&aconnector->base, DRM_MODE_CONTENT_PROTECTION_ENABLED);
+	else
+		drm_hdcp_update_content_protection(&aconnector->base, DRM_MODE_CONTENT_PROTECTION_DESIRED);
+
+
+	mutex_unlock(&hdcp_work->mutex);
+	drm_modeset_unlock(&dev->mode_config.connection_mutex);
+}
+
+static void event_property_validate(struct work_struct *work)
+{
+	struct hdcp_workqueue *hdcp_work =
+		container_of(to_delayed_work(work), struct hdcp_workqueue, property_validate_dwork);
+	struct mod_hdcp_display_query query;
+	struct amdgpu_dm_connector *aconnector = hdcp_work->aconnector;
+
+	mutex_lock(&hdcp_work->mutex);
+
+	query.encryption_status = MOD_HDCP_ENCRYPTION_STATUS_HDCP_OFF;
+	mod_hdcp_query_display(&hdcp_work->hdcp, aconnector->base.index, &query);
+
+	if (query.encryption_status != hdcp_work->encryption_status) {
+		hdcp_work->encryption_status = query.encryption_status;
+		schedule_work(&hdcp_work->property_update_work);
+	}
+
+	schedule_delayed_work(&hdcp_work->property_validate_dwork, msecs_to_jiffies(DRM_HDCP_CHECK_PERIOD_MS));
+
+	mutex_unlock(&hdcp_work->mutex);
+}
+
+static void event_watchdog_timer(struct work_struct *work)
+{
+	struct hdcp_workqueue *hdcp_work;
+
+	hdcp_work = container_of(to_delayed_work(work),
+				      struct hdcp_workqueue,
+				      watchdog_timer_dwork);
+
+	mutex_lock(&hdcp_work->mutex);
+
+	mod_hdcp_process_event(&hdcp_work->hdcp,
+			       MOD_HDCP_EVENT_WATCHDOG_TIMEOUT,
+			       &hdcp_work->output);
+
+	process_output(hdcp_work);
+
+	mutex_unlock(&hdcp_work->mutex);
+
+}
+
+static void event_cpirq(struct work_struct *work)
+{
+	struct hdcp_workqueue *hdcp_work;
+
+	hdcp_work = container_of(work, struct hdcp_workqueue, cpirq_work);
+
+	mutex_lock(&hdcp_work->mutex);
+
+	mod_hdcp_process_event(&hdcp_work->hdcp, MOD_HDCP_EVENT_CPIRQ, &hdcp_work->output);
+
+	process_output(hdcp_work);
+
+	mutex_unlock(&hdcp_work->mutex);
+
+}
+
+
+void hdcp_destroy(struct hdcp_workqueue *hdcp_work)
+{
+	int i = 0;
+
+	for (i = 0; i < hdcp_work->max_link; i++) {
+		cancel_delayed_work_sync(&hdcp_work[i].callback_dwork);
+		cancel_delayed_work_sync(&hdcp_work[i].watchdog_timer_dwork);
+	}
+
+	kfree(hdcp_work);
+
+}
+
+static void update_config(void *handle, struct cp_psp_stream_config *config)
+{
+	struct hdcp_workqueue *hdcp_work = handle;
+	struct amdgpu_dm_connector *aconnector = config->dm_stream_ctx;
+	int link_index = aconnector->dc_link->link_index;
+	struct mod_hdcp_display *display = &hdcp_work[link_index].display;
+	struct mod_hdcp_link *link = &hdcp_work[link_index].link;
+
+	memset(display, 0, sizeof(*display));
+	memset(link, 0, sizeof(*link));
+
+	display->index = aconnector->base.index;
+	display->state = MOD_HDCP_DISPLAY_ACTIVE;
+
+	if (aconnector->dc_sink != NULL)
+		link->mode = mod_hdcp_signal_type_to_operation_mode(aconnector->dc_sink->sink_signal);
+
+	display->controller = CONTROLLER_ID_D0 + config->otg_inst;
+	display->dig_fe = config->stream_enc_inst;
+	link->dig_be = config->link_enc_inst;
+	link->ddc_line = aconnector->dc_link->ddc_hw_inst + 1;
+	link->dp.rev = aconnector->dc_link->dpcd_caps.dpcd_rev.raw;
+	link->adjust.hdcp2.disable = 1;
+
+}
+
+struct hdcp_workqueue *hdcp_create_workqueue(void *psp_context, struct cp_psp *cp_psp, struct dc *dc)
+{
+
+	int max_caps = dc->caps.max_links;
+	struct hdcp_workqueue *hdcp_work = kzalloc(max_caps*sizeof(*hdcp_work), GFP_KERNEL);
+	int i = 0;
+
+	if (hdcp_work == NULL)
+		goto fail_alloc_context;
+
+	hdcp_work->max_link = max_caps;
+
+	for (i = 0; i < max_caps; i++) {
+
+		mutex_init(&hdcp_work[i].mutex);
+
+		INIT_WORK(&hdcp_work[i].cpirq_work, event_cpirq);
+		INIT_WORK(&hdcp_work[i].property_update_work, event_property_update);
+		INIT_DELAYED_WORK(&hdcp_work[i].callback_dwork, event_callback);
+		INIT_DELAYED_WORK(&hdcp_work[i].watchdog_timer_dwork, event_watchdog_timer);
+		INIT_DELAYED_WORK(&hdcp_work[i].property_validate_dwork, event_property_validate);
+
+		hdcp_work[i].hdcp.config.psp.handle =  psp_context;
+		hdcp_work[i].hdcp.config.ddc.handle = dc_get_link_at_index(dc, i);
+		hdcp_work[i].hdcp.config.ddc.funcs.write_i2c = lp_write_i2c;
+		hdcp_work[i].hdcp.config.ddc.funcs.read_i2c = lp_read_i2c;
+		hdcp_work[i].hdcp.config.ddc.funcs.write_dpcd = lp_write_dpcd;
+		hdcp_work[i].hdcp.config.ddc.funcs.read_dpcd = lp_read_dpcd;
+	}
+
+	cp_psp->funcs.update_stream_config = update_config;
+	cp_psp->handle = hdcp_work;
+
+	return hdcp_work;
+
+fail_alloc_context:
+	kfree(hdcp_work);
+
+	return NULL;
+
+
+
+}
+
+
+
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_hdcp.h b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_hdcp.h
new file mode 100644
index 000000000000..d3ba505d0696
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_hdcp.h
@@ -0,0 +1,66 @@
+/*
+ * Copyright 2019 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#ifndef AMDGPU_DM_AMDGPU_DM_HDCP_H_
+#define AMDGPU_DM_AMDGPU_DM_HDCP_H_
+
+#include "mod_hdcp.h"
+#include "hdcp.h"
+#include "dc.h"
+#include "dm_cp_psp.h"
+
+struct mod_hdcp;
+struct mod_hdcp_link;
+struct mod_hdcp_display;
+struct cp_psp;
+
+struct hdcp_workqueue {
+	struct work_struct cpirq_work;
+	struct work_struct property_update_work;
+	struct delayed_work callback_dwork;
+	struct delayed_work watchdog_timer_dwork;
+	struct delayed_work property_validate_dwork;
+	struct amdgpu_dm_connector *aconnector;
+	struct mutex mutex;
+
+	struct mod_hdcp hdcp;
+	struct mod_hdcp_output output;
+	struct mod_hdcp_display display;
+	struct mod_hdcp_link link;
+
+	enum mod_hdcp_encryption_status encryption_status;
+	uint8_t max_link;
+};
+
+void hdcp_add_display(struct hdcp_workqueue *hdcp_work, unsigned int link_index,
+		      struct amdgpu_dm_connector *aconnector);
+void hdcp_remove_display(struct hdcp_workqueue *work, unsigned int link_index, unsigned int display_index);
+void hdcp_reset_display(struct hdcp_workqueue *work, unsigned int link_index);
+void hdcp_handle_cpirq(struct hdcp_workqueue *work, unsigned int link_index);
+void hdcp_destroy(struct hdcp_workqueue *work);
+
+struct hdcp_workqueue *hdcp_create_workqueue(void *psp_context, struct cp_psp *cp_psp, struct dc *dc);
+
+#endif /* AMDGPU_DM_AMDGPU_DM_HDCP_H_ */
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c
index ee1dc75f5ddc..11e5784aa62a 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c
@@ -97,11 +97,10 @@ enum dc_edid_status dm_helpers_parse_edid_caps(
 			(struct edid *) edid->raw_edid);
 
 	sad_count = drm_edid_to_sad((struct edid *) edid->raw_edid, &sads);
-	if (sad_count <= 0) {
-		DRM_INFO("SADs count is: %d, don't need to read it\n",
-				sad_count);
+	if (sad_count < 0)
+		DRM_ERROR("Couldn't read SADs: %d\n", sad_count);
+	if (sad_count <= 0)
 		return result;
-	}
 
 	edid_caps->audio_mode_count = sad_count < DC_MAX_AUDIO_DESC_COUNT ? sad_count : DC_MAX_AUDIO_DESC_COUNT;
 	for (i = 0; i < edid_caps->audio_mode_count; ++i) {
@@ -282,7 +281,7 @@ void dm_helpers_dp_mst_clear_payload_allocation_table(
  * Polls for ACT (allocation change trigger) handled and sends
  * ALLOCATE_PAYLOAD message.
  */
-bool dm_helpers_dp_mst_poll_for_allocation_change_trigger(
+enum act_return_status dm_helpers_dp_mst_poll_for_allocation_change_trigger(
 		struct dc_context *ctx,
 		const struct dc_stream_state *stream)
 {
@@ -293,19 +292,19 @@ bool dm_helpers_dp_mst_poll_for_allocation_change_trigger(
 	aconnector = (struct amdgpu_dm_connector *)stream->dm_stream_context;
 
 	if (!aconnector || !aconnector->mst_port)
-		return false;
+		return ACT_FAILED;
 
 	mst_mgr = &aconnector->mst_port->mst_mgr;
 
 	if (!mst_mgr->mst_state)
-		return false;
+		return ACT_FAILED;
 
 	ret = drm_dp_check_act_status(mst_mgr);
 
 	if (ret)
-		return false;
+		return ACT_FAILED;
 
-	return true;
+	return ACT_SUCCESS;
 }
 
 bool dm_helpers_dp_mst_send_payload_allocation(
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.c
index fa5d503d379c..64445c4cc4c2 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.c
@@ -732,8 +732,10 @@ void amdgpu_dm_hpd_init(struct amdgpu_device *adev)
 {
 	struct drm_device *dev = adev->ddev;
 	struct drm_connector *connector;
+	struct drm_connector_list_iter iter;
 
-	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
+	drm_connector_list_iter_begin(dev, &iter);
+	drm_for_each_connector_iter(connector, &iter) {
 		struct amdgpu_dm_connector *amdgpu_dm_connector =
 				to_amdgpu_dm_connector(connector);
 
@@ -751,6 +753,7 @@ void amdgpu_dm_hpd_init(struct amdgpu_device *adev)
 					true);
 		}
 	}
+	drm_connector_list_iter_end(&iter);
 }
 
 /**
@@ -765,8 +768,10 @@ void amdgpu_dm_hpd_fini(struct amdgpu_device *adev)
 {
 	struct drm_device *dev = adev->ddev;
 	struct drm_connector *connector;
+	struct drm_connector_list_iter iter;
 
-	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
+	drm_connector_list_iter_begin(dev, &iter);
+	drm_for_each_connector_iter(connector, &iter) {
 		struct amdgpu_dm_connector *amdgpu_dm_connector =
 				to_amdgpu_dm_connector(connector);
 		const struct dc_link *dc_link = amdgpu_dm_connector->dc_link;
@@ -779,4 +784,5 @@ void amdgpu_dm_hpd_fini(struct amdgpu_device *adev)
 					false);
 		}
 	}
+	drm_connector_list_iter_end(&iter);
 }
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
index 16218a202b59..2bf8534c18fb 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
@@ -36,7 +36,9 @@
 #include "dc_link_ddc.h"
 
 #include "i2caux_interface.h"
-
+#if defined(CONFIG_DEBUG_FS)
+#include "amdgpu_dm_debugfs.h"
+#endif
 /* #define TRACE_DPCD */
 
 #ifdef TRACE_DPCD
@@ -113,6 +115,7 @@ static ssize_t dm_dp_aux_transfer(struct drm_dp_aux *aux,
 			result = -EIO;
 			break;
 		case AUX_CHANNEL_OPERATION_FAILED_INVALID_REPLY:
+		case AUX_CHANNEL_OPERATION_FAILED_ENGINE_ACQUIRE:
 			result = -EBUSY;
 			break;
 		case AUX_CHANNEL_OPERATION_FAILED_TIMEOUT:
@@ -123,31 +126,14 @@ static ssize_t dm_dp_aux_transfer(struct drm_dp_aux *aux,
 	return result;
 }
 
-static enum drm_connector_status
-dm_dp_mst_detect(struct drm_connector *connector, bool force)
-{
-	struct amdgpu_dm_connector *aconnector = to_amdgpu_dm_connector(connector);
-	struct amdgpu_dm_connector *master = aconnector->mst_port;
-
-	enum drm_connector_status status =
-		drm_dp_mst_detect_port(
-			connector,
-			&master->mst_mgr,
-			aconnector->port);
-
-	return status;
-}
-
 static void
 dm_dp_mst_connector_destroy(struct drm_connector *connector)
 {
 	struct amdgpu_dm_connector *amdgpu_dm_connector = to_amdgpu_dm_connector(connector);
 	struct amdgpu_encoder *amdgpu_encoder = amdgpu_dm_connector->mst_encoder;
 
-	if (amdgpu_dm_connector->edid) {
-		kfree(amdgpu_dm_connector->edid);
-		amdgpu_dm_connector->edid = NULL;
-	}
+	kfree(amdgpu_dm_connector->edid);
+	amdgpu_dm_connector->edid = NULL;
 
 	drm_encoder_cleanup(&amdgpu_encoder->base);
 	kfree(amdgpu_encoder);
@@ -163,6 +149,12 @@ amdgpu_dm_mst_connector_late_register(struct drm_connector *connector)
 		to_amdgpu_dm_connector(connector);
 	struct drm_dp_mst_port *port = amdgpu_dm_connector->port;
 
+#if defined(CONFIG_DEBUG_FS)
+	connector_debugfs_init(amdgpu_dm_connector);
+	amdgpu_dm_connector->debugfs_dpcd_address = 0;
+	amdgpu_dm_connector->debugfs_dpcd_size = 0;
+#endif
+
 	return drm_dp_mst_connector_late_register(connector, port);
 }
 
@@ -177,7 +169,6 @@ amdgpu_dm_mst_connector_early_unregister(struct drm_connector *connector)
 }
 
 static const struct drm_connector_funcs dm_dp_mst_connector_funcs = {
-	.detect = dm_dp_mst_detect,
 	.fill_modes = drm_helper_probe_single_connector_modes,
 	.destroy = dm_dp_mst_connector_destroy,
 	.reset = amdgpu_dm_connector_funcs_reset,
@@ -245,17 +236,29 @@ static int dm_dp_mst_get_modes(struct drm_connector *connector)
 	return ret;
 }
 
-static struct drm_encoder *dm_mst_best_encoder(struct drm_connector *connector)
+static struct drm_encoder *
+dm_mst_atomic_best_encoder(struct drm_connector *connector,
+			   struct drm_connector_state *connector_state)
 {
-	struct amdgpu_dm_connector *amdgpu_dm_connector = to_amdgpu_dm_connector(connector);
+	return &to_amdgpu_dm_connector(connector)->mst_encoder->base;
+}
+
+static int
+dm_dp_mst_detect(struct drm_connector *connector,
+		 struct drm_modeset_acquire_ctx *ctx, bool force)
+{
+	struct amdgpu_dm_connector *aconnector = to_amdgpu_dm_connector(connector);
+	struct amdgpu_dm_connector *master = aconnector->mst_port;
 
-	return &amdgpu_dm_connector->mst_encoder->base;
+	return drm_dp_mst_detect_port(connector, ctx, &master->mst_mgr,
+				      aconnector->port);
 }
 
 static const struct drm_connector_helper_funcs dm_dp_mst_connector_helper_funcs = {
 	.get_modes = dm_dp_mst_get_modes,
 	.mode_valid = amdgpu_dm_connector_mode_valid,
-	.best_encoder = dm_mst_best_encoder,
+	.atomic_best_encoder = dm_mst_atomic_best_encoder,
+	.detect_ctx = dm_dp_mst_detect,
 };
 
 static void amdgpu_dm_encoder_destroy(struct drm_encoder *encoder)
@@ -416,7 +419,11 @@ void amdgpu_dm_initialize_dp_connector(struct amdgpu_display_manager *dm,
 
 	drm_dp_aux_register(&aconnector->dm_dp_aux.aux);
 	drm_dp_cec_register_connector(&aconnector->dm_dp_aux.aux,
-				      aconnector->base.name, dm->adev->dev);
+				      &aconnector->base);
+
+	if (aconnector->base.connector_type == DRM_MODE_CONNECTOR_eDP)
+		return;
+
 	aconnector->mst_mgr.cbs = &dm_mst_cbs;
 	drm_dp_mst_topology_mgr_init(
 		&aconnector->mst_mgr,
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_pp_smu.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_pp_smu.c
index f4cfa0caeba8..55a520a63712 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_pp_smu.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_pp_smu.c
@@ -345,7 +345,7 @@ bool dm_pp_get_clock_levels_by_type(
 		/* Error in pplib. Provide default values. */
 			return true;
 		}
-	} else if (adev->smu.funcs && adev->smu.funcs->get_clock_by_type) {
+	} else if (adev->smu.ppt_funcs && adev->smu.ppt_funcs->get_clock_by_type) {
 		if (smu_get_clock_by_type(&adev->smu,
 					  dc_to_pp_clock_type(clk_type),
 					  &pp_clks)) {
@@ -365,7 +365,7 @@ bool dm_pp_get_clock_levels_by_type(
 			validation_clks.memory_max_clock = 80000;
 			validation_clks.level = 0;
 		}
-	} else if (adev->smu.funcs && adev->smu.funcs->get_max_high_clocks) {
+	} else if (adev->smu.ppt_funcs && adev->smu.ppt_funcs->get_max_high_clocks) {
 		if (smu_get_max_high_clocks(&adev->smu, &validation_clks)) {
 			DRM_INFO("DM_PPLIB: Warning: using default validation clocks!\n");
 			validation_clks.engine_max_clock = 72000;
@@ -506,8 +506,8 @@ bool dm_pp_apply_clock_for_voltage_request(
 		ret = adev->powerplay.pp_funcs->display_clock_voltage_request(
 			adev->powerplay.pp_handle,
 			&pp_clock_request);
-	else if (adev->smu.funcs &&
-		 adev->smu.funcs->display_clock_voltage_request)
+	else if (adev->smu.ppt_funcs &&
+		 adev->smu.ppt_funcs->display_clock_voltage_request)
 		ret = smu_display_clock_voltage_request(&adev->smu,
 							&pp_clock_request);
 	if (ret)
@@ -527,7 +527,7 @@ bool dm_pp_get_static_clocks(
 		ret = adev->powerplay.pp_funcs->get_current_clocks(
 			adev->powerplay.pp_handle,
 			&pp_clk_info);
-	else if (adev->smu.funcs)
+	else if (adev->smu.ppt_funcs)
 		ret = smu_get_current_clocks(&adev->smu, &pp_clk_info);
 	if (ret)
 		return false;
@@ -589,10 +589,9 @@ void pp_rv_set_wm_ranges(struct pp_smu *pp,
 	if (pp_funcs && pp_funcs->set_watermarks_for_clocks_ranges)
 		pp_funcs->set_watermarks_for_clocks_ranges(pp_handle,
 							   &wm_with_clock_ranges);
-	else if (adev->smu.funcs &&
-		 adev->smu.funcs->set_watermarks_for_clock_ranges)
+	else
 		smu_set_watermarks_for_clock_ranges(&adev->smu,
-						    &wm_with_clock_ranges);
+				&wm_with_clock_ranges);
 }
 
 void pp_rv_set_pme_wa_enable(struct pp_smu *pp)
@@ -604,7 +603,7 @@ void pp_rv_set_pme_wa_enable(struct pp_smu *pp)
 
 	if (pp_funcs && pp_funcs->notify_smu_enable_pwe)
 		pp_funcs->notify_smu_enable_pwe(pp_handle);
-	else if (adev->smu.funcs)
+	else if (adev->smu.ppt_funcs)
 		smu_notify_smu_enable_pwe(&adev->smu);
 }
 
@@ -665,7 +664,6 @@ enum pp_smu_status pp_nv_set_wm_ranges(struct pp_smu *pp,
 {
 	const struct dc_context *ctx = pp->dm;
 	struct amdgpu_device *adev = ctx->driver_context;
-	struct smu_context *smu = &adev->smu;
 	struct dm_pp_wm_sets_with_clock_ranges_soc15 wm_with_clock_ranges;
 	struct dm_pp_clock_range_for_dmif_wm_set_soc15 *wm_dce_clocks =
 			wm_with_clock_ranges.wm_dmif_clocks_ranges;
@@ -708,15 +706,7 @@ enum pp_smu_status pp_nv_set_wm_ranges(struct pp_smu *pp,
 			ranges->writer_wm_sets[i].min_drain_clk_mhz * 1000;
 	}
 
-	if (!smu->funcs)
-		return PP_SMU_RESULT_UNSUPPORTED;
-
-	/* 0: successful or smu.funcs->set_watermarks_for_clock_ranges = NULL;
-	 * 1: fail
-	 */
-	if (smu_set_watermarks_for_clock_ranges(&adev->smu,
-			&wm_with_clock_ranges))
-		return PP_SMU_RESULT_UNSUPPORTED;
+	smu_set_watermarks_for_clock_ranges(&adev->smu,	&wm_with_clock_ranges);
 
 	return PP_SMU_RESULT_OK;
 }
@@ -727,10 +717,10 @@ enum pp_smu_status pp_nv_set_pme_wa_enable(struct pp_smu *pp)
 	struct amdgpu_device *adev = ctx->driver_context;
 	struct smu_context *smu = &adev->smu;
 
-	if (!smu->funcs)
+	if (!smu->ppt_funcs)
 		return PP_SMU_RESULT_UNSUPPORTED;
 
-	/* 0: successful or smu.funcs->set_azalia_d3_pme = NULL;  1: fail */
+	/* 0: successful or smu.ppt_funcs->set_azalia_d3_pme = NULL;  1: fail */
 	if (smu_set_azalia_d3_pme(smu))
 		return PP_SMU_RESULT_FAIL;
 
@@ -743,10 +733,10 @@ enum pp_smu_status pp_nv_set_display_count(struct pp_smu *pp, int count)
 	struct amdgpu_device *adev = ctx->driver_context;
 	struct smu_context *smu = &adev->smu;
 
-	if (!smu->funcs)
+	if (!smu->ppt_funcs)
 		return PP_SMU_RESULT_UNSUPPORTED;
 
-	/* 0: successful or smu.funcs->set_display_count = NULL;  1: fail */
+	/* 0: successful or smu.ppt_funcs->set_display_count = NULL;  1: fail */
 	if (smu_set_display_count(smu, count))
 		return PP_SMU_RESULT_FAIL;
 
@@ -759,10 +749,10 @@ enum pp_smu_status pp_nv_set_min_deep_sleep_dcfclk(struct pp_smu *pp, int mhz)
 	struct amdgpu_device *adev = ctx->driver_context;
 	struct smu_context *smu = &adev->smu;
 
-	if (!smu->funcs)
+	if (!smu->ppt_funcs)
 		return PP_SMU_RESULT_UNSUPPORTED;
 
-	/* 0: successful or smu.funcs->set_deep_sleep_dcefclk = NULL;1: fail */
+	/* 0: successful or smu.ppt_funcs->set_deep_sleep_dcefclk = NULL;1: fail */
 	if (smu_set_deep_sleep_dcefclk(smu, mhz))
 		return PP_SMU_RESULT_FAIL;
 
@@ -777,13 +767,13 @@ enum pp_smu_status pp_nv_set_hard_min_dcefclk_by_freq(
 	struct smu_context *smu = &adev->smu;
 	struct pp_display_clock_request clock_req;
 
-	if (!smu->funcs)
+	if (!smu->ppt_funcs)
 		return PP_SMU_RESULT_UNSUPPORTED;
 
 	clock_req.clock_type = amd_pp_dcef_clock;
 	clock_req.clock_freq_in_khz = mhz * 1000;
 
-	/* 0: successful or smu.funcs->display_clock_voltage_request = NULL
+	/* 0: successful or smu.ppt_funcs->display_clock_voltage_request = NULL
 	 * 1: fail
 	 */
 	if (smu_display_clock_voltage_request(smu, &clock_req))
@@ -799,13 +789,13 @@ enum pp_smu_status pp_nv_set_hard_min_uclk_by_freq(struct pp_smu *pp, int mhz)
 	struct smu_context *smu = &adev->smu;
 	struct pp_display_clock_request clock_req;
 
-	if (!smu->funcs)
+	if (!smu->ppt_funcs)
 		return PP_SMU_RESULT_UNSUPPORTED;
 
 	clock_req.clock_type = amd_pp_mem_clock;
 	clock_req.clock_freq_in_khz = mhz * 1000;
 
-	/* 0: successful or smu.funcs->display_clock_voltage_request = NULL
+	/* 0: successful or smu.ppt_funcs->display_clock_voltage_request = NULL
 	 * 1: fail
 	 */
 	if (smu_display_clock_voltage_request(smu, &clock_req))
@@ -835,7 +825,7 @@ enum pp_smu_status pp_nv_set_voltage_by_freq(struct pp_smu *pp,
 	struct smu_context *smu = &adev->smu;
 	struct pp_display_clock_request clock_req;
 
-	if (!smu->funcs)
+	if (!smu->ppt_funcs)
 		return PP_SMU_RESULT_UNSUPPORTED;
 
 	switch (clock_id) {
@@ -853,7 +843,7 @@ enum pp_smu_status pp_nv_set_voltage_by_freq(struct pp_smu *pp,
 	}
 	clock_req.clock_freq_in_khz = mhz * 1000;
 
-	/* 0: successful or smu.funcs->display_clock_voltage_request = NULL
+	/* 0: successful or smu.ppt_funcs->display_clock_voltage_request = NULL
 	 * 1: fail
 	 */
 	if (smu_display_clock_voltage_request(smu, &clock_req))
@@ -869,13 +859,13 @@ enum pp_smu_status pp_nv_get_maximum_sustainable_clocks(
 	struct amdgpu_device *adev = ctx->driver_context;
 	struct smu_context *smu = &adev->smu;
 
-	if (!smu->funcs)
+	if (!smu->ppt_funcs)
 		return PP_SMU_RESULT_UNSUPPORTED;
 
-	if (!smu->funcs->get_max_sustainable_clocks_by_dc)
+	if (!smu->ppt_funcs->get_max_sustainable_clocks_by_dc)
 		return PP_SMU_RESULT_UNSUPPORTED;
 
-	if (!smu->funcs->get_max_sustainable_clocks_by_dc(smu, max_clocks))
+	if (!smu_get_max_sustainable_clocks_by_dc(smu, max_clocks))
 		return PP_SMU_RESULT_OK;
 
 	return PP_SMU_RESULT_FAIL;
@@ -894,13 +884,97 @@ enum pp_smu_status pp_nv_get_uclk_dpm_states(struct pp_smu *pp,
 	if (!smu->ppt_funcs->get_uclk_dpm_states)
 		return PP_SMU_RESULT_UNSUPPORTED;
 
-	if (!smu->ppt_funcs->get_uclk_dpm_states(smu,
+	if (!smu_get_uclk_dpm_states(smu,
 			clock_values_in_khz, num_states))
 		return PP_SMU_RESULT_OK;
 
 	return PP_SMU_RESULT_FAIL;
 }
 
+#ifdef CONFIG_DRM_AMD_DC_DCN2_1
+enum pp_smu_status pp_rn_get_dpm_clock_table(
+		struct pp_smu *pp, struct dpm_clocks *clock_table)
+{
+	const struct dc_context *ctx = pp->dm;
+	struct amdgpu_device *adev = ctx->driver_context;
+	struct smu_context *smu = &adev->smu;
+
+	if (!smu->ppt_funcs)
+		return PP_SMU_RESULT_UNSUPPORTED;
+
+	if (!smu->ppt_funcs->get_dpm_clock_table)
+		return PP_SMU_RESULT_UNSUPPORTED;
+
+	if (!smu_get_dpm_clock_table(smu, clock_table))
+		return PP_SMU_RESULT_OK;
+
+	return PP_SMU_RESULT_FAIL;
+}
+
+enum pp_smu_status pp_rn_set_wm_ranges(struct pp_smu *pp,
+		struct pp_smu_wm_range_sets *ranges)
+{
+	const struct dc_context *ctx = pp->dm;
+	struct amdgpu_device *adev = ctx->driver_context;
+	struct smu_context *smu = &adev->smu;
+	struct dm_pp_wm_sets_with_clock_ranges_soc15 wm_with_clock_ranges;
+	struct dm_pp_clock_range_for_dmif_wm_set_soc15 *wm_dce_clocks =
+			wm_with_clock_ranges.wm_dmif_clocks_ranges;
+	struct dm_pp_clock_range_for_mcif_wm_set_soc15 *wm_soc_clocks =
+			wm_with_clock_ranges.wm_mcif_clocks_ranges;
+	int32_t i;
+
+	if (!smu->ppt_funcs)
+		return PP_SMU_RESULT_UNSUPPORTED;
+
+	wm_with_clock_ranges.num_wm_dmif_sets = ranges->num_reader_wm_sets;
+	wm_with_clock_ranges.num_wm_mcif_sets = ranges->num_writer_wm_sets;
+
+	for (i = 0; i < wm_with_clock_ranges.num_wm_dmif_sets; i++) {
+		if (ranges->reader_wm_sets[i].wm_inst > 3)
+			wm_dce_clocks[i].wm_set_id = WM_SET_A;
+		else
+			wm_dce_clocks[i].wm_set_id =
+					ranges->reader_wm_sets[i].wm_inst;
+
+		wm_dce_clocks[i].wm_min_dcfclk_clk_in_khz =
+			ranges->reader_wm_sets[i].min_drain_clk_mhz;
+
+		wm_dce_clocks[i].wm_max_dcfclk_clk_in_khz =
+			ranges->reader_wm_sets[i].max_drain_clk_mhz;
+
+		wm_dce_clocks[i].wm_min_mem_clk_in_khz =
+			ranges->reader_wm_sets[i].min_fill_clk_mhz;
+
+		wm_dce_clocks[i].wm_max_mem_clk_in_khz =
+			ranges->reader_wm_sets[i].max_fill_clk_mhz;
+	}
+
+	for (i = 0; i < wm_with_clock_ranges.num_wm_mcif_sets; i++) {
+		if (ranges->writer_wm_sets[i].wm_inst > 3)
+			wm_soc_clocks[i].wm_set_id = WM_SET_A;
+		else
+			wm_soc_clocks[i].wm_set_id =
+					ranges->writer_wm_sets[i].wm_inst;
+		wm_soc_clocks[i].wm_min_socclk_clk_in_khz =
+				ranges->writer_wm_sets[i].min_fill_clk_mhz;
+
+		wm_soc_clocks[i].wm_max_socclk_clk_in_khz =
+			ranges->writer_wm_sets[i].max_fill_clk_mhz;
+
+		wm_soc_clocks[i].wm_min_mem_clk_in_khz =
+			ranges->writer_wm_sets[i].min_drain_clk_mhz;
+
+		wm_soc_clocks[i].wm_max_mem_clk_in_khz =
+			ranges->writer_wm_sets[i].max_drain_clk_mhz;
+	}
+
+	smu_set_watermarks_for_clock_ranges(&adev->smu, &wm_with_clock_ranges);
+
+	return PP_SMU_RESULT_OK;
+}
+#endif
+
 void dm_pp_get_funcs(
 		struct dc_context *ctx,
 		struct pp_smu_funcs *funcs)
@@ -945,6 +1019,15 @@ void dm_pp_get_funcs(
 		funcs->nv_funcs.set_pstate_handshake_support = pp_nv_set_pstate_handshake_support;
 		break;
 #endif
+
+#ifdef CONFIG_DRM_AMD_DC_DCN2_1
+	case DCN_VERSION_2_1:
+		funcs->ctx.ver = PP_SMU_VER_RN;
+		funcs->rn_funcs.pp_smu.dm = ctx;
+		funcs->rn_funcs.set_wm_ranges = pp_rn_set_wm_ranges;
+		funcs->rn_funcs.get_dpm_clock_table = pp_rn_get_dpm_clock_table;
+		break;
+#endif
 	default:
 		DRM_ERROR("smu version is not supported !\n");
 		break;
diff --git a/drivers/gpu/drm/amd/display/dc/Makefile b/drivers/gpu/drm/amd/display/dc/Makefile
index 627982cb15d2..a160512a2f04 100644
--- a/drivers/gpu/drm/amd/display/dc/Makefile
+++ b/drivers/gpu/drm/amd/display/dc/Makefile
@@ -48,6 +48,10 @@ DC_LIBS += dce110
 DC_LIBS += dce100
 DC_LIBS += dce80
 
+ifdef CONFIG_DRM_AMD_DC_HDCP
+DC_LIBS += hdcp
+endif
+
 AMD_DC = $(addsuffix /Makefile, $(addprefix $(FULL_AMD_DISPLAY_PATH)/dc/,$(DC_LIBS)))
 
 include $(AMD_DC)
diff --git a/drivers/gpu/drm/amd/display/dc/bios/bios_parser.c b/drivers/gpu/drm/amd/display/dc/bios/bios_parser.c
index 221e0f56389f..823843cd2613 100644
--- a/drivers/gpu/drm/amd/display/dc/bios/bios_parser.c
+++ b/drivers/gpu/drm/amd/display/dc/bios/bios_parser.c
@@ -2543,7 +2543,6 @@ static enum bp_result construct_integrated_info(
 
 	/* Sort voltage table from low to high*/
 	if (result == BP_RESULT_OK) {
-		struct clock_voltage_caps temp = {0, 0};
 		uint32_t i;
 		uint32_t j;
 
@@ -2553,10 +2552,8 @@ static enum bp_result construct_integrated_info(
 						info->disp_clk_voltage[j].max_supported_clk <
 						info->disp_clk_voltage[j-1].max_supported_clk) {
 					/* swap j and j - 1*/
-					temp = info->disp_clk_voltage[j-1];
-					info->disp_clk_voltage[j-1] =
-							info->disp_clk_voltage[j];
-					info->disp_clk_voltage[j] = temp;
+					swap(info->disp_clk_voltage[j - 1],
+					     info->disp_clk_voltage[j]);
 				}
 			}
 		}
diff --git a/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c b/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c
index dff65c0fe82f..7873abea4112 100644
--- a/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c
+++ b/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c
@@ -1613,8 +1613,6 @@ static enum bp_result construct_integrated_info(
 
 	struct atom_common_table_header *header;
 	struct atom_data_revision revision;
-
-	struct clock_voltage_caps temp = {0, 0};
 	uint32_t i;
 	uint32_t j;
 
@@ -1644,10 +1642,8 @@ static enum bp_result construct_integrated_info(
 				info->disp_clk_voltage[j-1].max_supported_clk
 				) {
 				/* swap j and j - 1*/
-				temp = info->disp_clk_voltage[j-1];
-				info->disp_clk_voltage[j-1] =
-					info->disp_clk_voltage[j];
-				info->disp_clk_voltage[j] = temp;
+				swap(info->disp_clk_voltage[j - 1],
+				     info->disp_clk_voltage[j]);
 			}
 		}
 	}
diff --git a/drivers/gpu/drm/amd/display/dc/calcs/Makefile b/drivers/gpu/drm/amd/display/dc/calcs/Makefile
index 985633c08a26..26c6d735cdc7 100644
--- a/drivers/gpu/drm/amd/display/dc/calcs/Makefile
+++ b/drivers/gpu/drm/amd/display/dc/calcs/Makefile
@@ -24,15 +24,20 @@
 # It calculates Bandwidth and Watermarks values for HW programming
 #
 
-ifneq ($(call cc-option, -mpreferred-stack-boundary=4),)
-	cc_stack_align := -mpreferred-stack-boundary=4
-else ifneq ($(call cc-option, -mstack-alignment=16),)
-	cc_stack_align := -mstack-alignment=16
-endif
+calcs_ccflags := -mhard-float -msse
 
-calcs_ccflags := -mhard-float -msse $(cc_stack_align)
+ifdef CONFIG_CC_IS_GCC
+ifeq ($(call cc-ifversion, -lt, 0701, y), y)
+IS_OLD_GCC = 1
+endif
+endif
 
-ifdef CONFIG_CC_IS_CLANG
+ifdef IS_OLD_GCC
+# Stack alignment mismatch, proceed with caution.
+# GCC < 7.1 cannot compile code using `double` and -mpreferred-stack-boundary=3
+# (8B stack alignment).
+calcs_ccflags += -mpreferred-stack-boundary=4
+else
 calcs_ccflags += -msse2
 endif
 
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/clk_mgr.c
index c43797bea413..8828dd9c3783 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/clk_mgr.c
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/clk_mgr.c
@@ -65,6 +65,31 @@ int clk_mgr_helper_get_active_display_cnt(
 	return display_count;
 }
 
+void clk_mgr_exit_optimized_pwr_state(const struct dc *dc, struct clk_mgr *clk_mgr)
+{
+	struct dc_link *edp_link = get_edp_link(dc);
+
+	if (dc->hwss.exit_optimized_pwr_state)
+		dc->hwss.exit_optimized_pwr_state(dc, dc->current_state);
+
+	if (edp_link) {
+		clk_mgr->psr_allow_active_cache = edp_link->psr_allow_active;
+		dc_link_set_psr_allow_active(edp_link, false, false);
+	}
+
+}
+
+void clk_mgr_optimize_pwr_state(const struct dc *dc, struct clk_mgr *clk_mgr)
+{
+	struct dc_link *edp_link = get_edp_link(dc);
+
+	if (edp_link)
+		dc_link_set_psr_allow_active(edp_link, clk_mgr->psr_allow_active_cache, false);
+
+	if (dc->hwss.optimize_pwr_state)
+		dc->hwss.optimize_pwr_state(dc, dc->current_state);
+
+}
 
 struct clk_mgr *dc_clk_mgr_create(struct dc_context *ctx, struct pp_smu_funcs *pp_smu, struct dccg *dccg)
 {
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dce100/dce_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dce100/dce_clk_mgr.c
index c5c8c4901eed..26db1c5d4e4d 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dce100/dce_clk_mgr.c
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dce100/dce_clk_mgr.c
@@ -147,7 +147,7 @@ int dce_get_dp_ref_freq_khz(struct clk_mgr *clk_mgr_base)
 
 	/* Calculate the current DFS clock, in kHz.*/
 	dp_ref_clk_khz = (DENTIST_DIVIDER_RANGE_SCALE_FACTOR
-		* clk_mgr->dentist_vco_freq_khz) / target_div;
+		* clk_mgr->base.dentist_vco_freq_khz) / target_div;
 
 	return dce_adjust_dp_ref_freq_for_ss(clk_mgr, dp_ref_clk_khz);
 }
@@ -239,7 +239,7 @@ int dce_set_clock(
 	/* Make sure requested clock isn't lower than minimum threshold*/
 	if (requested_clk_khz > 0)
 		requested_clk_khz = max(requested_clk_khz,
-				clk_mgr_dce->dentist_vco_freq_khz / 64);
+				clk_mgr_dce->base.dentist_vco_freq_khz / 64);
 
 	/* Prepare to program display clock*/
 	pxl_clk_params.target_pixel_clock_100hz = requested_clk_khz * 10;
@@ -276,11 +276,11 @@ static void dce_clock_read_integrated_info(struct clk_mgr_internal *clk_mgr_dce)
 	int i;
 
 	if (bp->integrated_info)
-		clk_mgr_dce->dentist_vco_freq_khz = bp->integrated_info->dentist_vco_freq;
-	if (clk_mgr_dce->dentist_vco_freq_khz == 0) {
-		clk_mgr_dce->dentist_vco_freq_khz = bp->fw_info.smu_gpu_pll_output_freq;
-		if (clk_mgr_dce->dentist_vco_freq_khz == 0)
-			clk_mgr_dce->dentist_vco_freq_khz = 3600000;
+		clk_mgr_dce->base.dentist_vco_freq_khz = bp->integrated_info->dentist_vco_freq;
+	if (clk_mgr_dce->base.dentist_vco_freq_khz == 0) {
+		clk_mgr_dce->base.dentist_vco_freq_khz = bp->fw_info.smu_gpu_pll_output_freq;
+		if (clk_mgr_dce->base.dentist_vco_freq_khz == 0)
+			clk_mgr_dce->base.dentist_vco_freq_khz = 3600000;
 	}
 
 	/*update the maximum display clock for each power state*/
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dce112/dce112_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dce112/dce112_clk_mgr.c
index 7c746ef1e32e..a6c46e903ff9 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dce112/dce112_clk_mgr.c
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dce112/dce112_clk_mgr.c
@@ -81,7 +81,7 @@ int dce112_set_clock(struct clk_mgr *clk_mgr_base, int requested_clk_khz)
 	/* Make sure requested clock isn't lower than minimum threshold*/
 	if (requested_clk_khz > 0)
 		requested_clk_khz = max(requested_clk_khz,
-				clk_mgr_dce->dentist_vco_freq_khz / 62);
+				clk_mgr_dce->base.dentist_vco_freq_khz / 62);
 
 	dce_clk_params.target_clock_frequency = requested_clk_khz;
 	dce_clk_params.pll_id = CLOCK_SOURCE_ID_DFS;
@@ -135,7 +135,7 @@ int dce112_set_dispclk(struct clk_mgr_internal *clk_mgr, int requested_clk_khz)
 	/* Make sure requested clock isn't lower than minimum threshold*/
 	if (requested_clk_khz > 0)
 		requested_clk_khz = max(requested_clk_khz,
-				clk_mgr->dentist_vco_freq_khz / 62);
+				clk_mgr->base.dentist_vco_freq_khz / 62);
 
 	dce_clk_params.target_clock_frequency = requested_clk_khz;
 	dce_clk_params.pll_id = CLOCK_SOURCE_ID_DFS;
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr.c
index 47f529ce280a..3fab9296918a 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr.c
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr.c
@@ -139,6 +139,9 @@ static void rv1_update_clocks(struct clk_mgr *clk_mgr_base,
 
 	ASSERT(clk_mgr->pp_smu);
 
+	if (dc->work_arounds.skip_clock_update)
+		return;
+
 	pp_smu = &clk_mgr->pp_smu->rv_funcs;
 
 	display_count = clk_mgr_helper_get_active_display_cnt(dc, context);
@@ -266,11 +269,11 @@ void rv1_clk_mgr_construct(struct dc_context *ctx, struct clk_mgr_internal *clk_
 	clk_mgr->base.dprefclk_khz = 600000;
 
 	if (bp->integrated_info)
-		clk_mgr->dentist_vco_freq_khz = bp->integrated_info->dentist_vco_freq;
-	if (bp->fw_info_valid && clk_mgr->dentist_vco_freq_khz == 0) {
-		clk_mgr->dentist_vco_freq_khz = bp->fw_info.smu_gpu_pll_output_freq;
-		if (clk_mgr->dentist_vco_freq_khz == 0)
-			clk_mgr->dentist_vco_freq_khz = 3600000;
+		clk_mgr->base.dentist_vco_freq_khz = bp->integrated_info->dentist_vco_freq;
+	if (bp->fw_info_valid && clk_mgr->base.dentist_vco_freq_khz == 0) {
+		clk_mgr->base.dentist_vco_freq_khz = bp->fw_info.smu_gpu_pll_output_freq;
+		if (clk_mgr->base.dentist_vco_freq_khz == 0)
+			clk_mgr->base.dentist_vco_freq_khz = 3600000;
 	}
 
 	if (!debug->disable_dfs_bypass && bp->integrated_info)
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn20/dcn20_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn20/dcn20_clk_mgr.c
index 3e8ac303bd52..25d7b7c6681c 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn20/dcn20_clk_mgr.c
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn20/dcn20_clk_mgr.c
@@ -104,84 +104,39 @@ void dcn20_update_clocks_update_dpp_dto(struct clk_mgr_internal *clk_mgr,
 {
 	int i;
 
+	clk_mgr->dccg->ref_dppclk = clk_mgr->base.clks.dppclk_khz;
 	for (i = 0; i < clk_mgr->base.ctx->dc->res_pool->pipe_count; i++) {
 		int dpp_inst, dppclk_khz;
 
-		if (!context->res_ctx.pipe_ctx[i].plane_state)
-			continue;
-
-		dpp_inst = context->res_ctx.pipe_ctx[i].plane_res.dpp->inst;
+		/* Loop index will match dpp->inst if resource exists,
+		 * and we want to avoid dependency on dpp object
+		 */
+		dpp_inst = i;
 		dppclk_khz = context->res_ctx.pipe_ctx[i].plane_res.bw.dppclk_khz;
+
 		clk_mgr->dccg->funcs->update_dpp_dto(
-				clk_mgr->dccg, dpp_inst, dppclk_khz, false);
+				clk_mgr->dccg, dpp_inst, dppclk_khz);
 	}
 }
 
-static void update_global_dpp_clk(struct clk_mgr_internal *clk_mgr, unsigned int khz)
+void dcn20_update_clocks_update_dentist(struct clk_mgr_internal *clk_mgr)
 {
 	int dpp_divider = DENTIST_DIVIDER_RANGE_SCALE_FACTOR
-			* clk_mgr->dentist_vco_freq_khz / khz;
-
-	uint32_t dppclk_wdivider = dentist_get_did_from_divider(dpp_divider);
-
-	REG_UPDATE(DENTIST_DISPCLK_CNTL,
-			DENTIST_DPPCLK_WDIVIDER, dppclk_wdivider);
-	REG_WAIT(DENTIST_DISPCLK_CNTL, DENTIST_DPPCLK_CHG_DONE, 1, 5, 100);
-}
-
-static void update_display_clk(struct clk_mgr_internal *clk_mgr, unsigned int khz)
-{
+			* clk_mgr->base.dentist_vco_freq_khz / clk_mgr->base.clks.dppclk_khz;
 	int disp_divider = DENTIST_DIVIDER_RANGE_SCALE_FACTOR
-			* clk_mgr->dentist_vco_freq_khz / khz;
+			* clk_mgr->base.dentist_vco_freq_khz / clk_mgr->base.clks.dispclk_khz;
 
+	uint32_t dppclk_wdivider = dentist_get_did_from_divider(dpp_divider);
 	uint32_t dispclk_wdivider = dentist_get_did_from_divider(disp_divider);
 
 	REG_UPDATE(DENTIST_DISPCLK_CNTL,
 			DENTIST_DISPCLK_WDIVIDER, dispclk_wdivider);
+//	REG_WAIT(DENTIST_DISPCLK_CNTL, DENTIST_DISPCLK_CHG_DONE, 1, 5, 100);
+	REG_UPDATE(DENTIST_DISPCLK_CNTL,
+			DENTIST_DPPCLK_WDIVIDER, dppclk_wdivider);
+	REG_WAIT(DENTIST_DISPCLK_CNTL, DENTIST_DPPCLK_CHG_DONE, 1, 5, 100);
 }
 
-static void request_voltage_and_program_disp_clk(struct clk_mgr *clk_mgr_base, unsigned int khz)
-{
-	struct clk_mgr_internal *clk_mgr = TO_CLK_MGR_INTERNAL(clk_mgr_base);
-	struct dc *dc = clk_mgr_base->ctx->dc;
-	struct pp_smu_funcs_nv *pp_smu = NULL;
-	bool going_up = clk_mgr->base.clks.dispclk_khz < khz;
-
-	if (dc->res_pool->pp_smu)
-		pp_smu = &dc->res_pool->pp_smu->nv_funcs;
-
-	clk_mgr->base.clks.dispclk_khz = khz;
-
-	if (going_up && pp_smu && pp_smu->set_voltage_by_freq)
-		pp_smu->set_voltage_by_freq(&pp_smu->pp_smu, PP_SMU_NV_DISPCLK, clk_mgr_base->clks.dispclk_khz / 1000);
-
-	update_display_clk(clk_mgr, khz);
-
-	if (!going_up && pp_smu && pp_smu->set_voltage_by_freq)
-		pp_smu->set_voltage_by_freq(&pp_smu->pp_smu, PP_SMU_NV_DISPCLK, clk_mgr_base->clks.dispclk_khz / 1000);
-}
-
-static void request_voltage_and_program_global_dpp_clk(struct clk_mgr *clk_mgr_base, unsigned int khz)
-{
-	struct clk_mgr_internal *clk_mgr = TO_CLK_MGR_INTERNAL(clk_mgr_base);
-	struct dc *dc = clk_mgr_base->ctx->dc;
-	struct pp_smu_funcs_nv *pp_smu = NULL;
-	bool going_up = clk_mgr->base.clks.dppclk_khz < khz;
-
-	if (dc->res_pool->pp_smu)
-		pp_smu = &dc->res_pool->pp_smu->nv_funcs;
-
-	clk_mgr->base.clks.dppclk_khz = khz;
-	clk_mgr->dccg->ref_dppclk = khz;
-
-	if (going_up && pp_smu && pp_smu->set_voltage_by_freq)
-		pp_smu->set_voltage_by_freq(&pp_smu->pp_smu, PP_SMU_NV_PIXELCLK, clk_mgr_base->clks.dppclk_khz / 1000);
-
-	update_global_dpp_clk(clk_mgr, khz);
-
-	if (!going_up && pp_smu && pp_smu->set_voltage_by_freq)
-		pp_smu->set_voltage_by_freq(&pp_smu->pp_smu, PP_SMU_NV_PIXELCLK, clk_mgr_base->clks.dppclk_khz / 1000);
-}
 
 void dcn2_update_clocks(struct clk_mgr *clk_mgr_base,
 			struct dc_state *context,
@@ -192,11 +147,12 @@ void dcn2_update_clocks(struct clk_mgr *clk_mgr_base,
 	struct dc *dc = clk_mgr_base->ctx->dc;
 	struct pp_smu_funcs_nv *pp_smu = NULL;
 	int display_count;
+	bool update_dppclk = false;
 	bool update_dispclk = false;
 	bool enter_display_off = false;
+	bool dpp_clock_lowered = false;
 	struct dmcu *dmcu = clk_mgr_base->ctx->dc->res_pool->dmcu;
 	bool force_reset = false;
-	int i;
 
 	if (dc->work_arounds.skip_clock_update)
 		return;
@@ -251,12 +207,10 @@ void dcn2_update_clocks(struct clk_mgr *clk_mgr_base,
 
 	if (should_update_pstate_support(safe_to_lower, new_clocks->p_state_change_support, clk_mgr_base->clks.p_state_change_support)) {
 		clk_mgr_base->clks.prev_p_state_change_support = clk_mgr_base->clks.p_state_change_support;
-
 		clk_mgr_base->clks.p_state_change_support = new_clocks->p_state_change_support;
 		if (pp_smu && pp_smu->set_pstate_handshake_support)
 			pp_smu->set_pstate_handshake_support(&pp_smu->pp_smu, clk_mgr_base->clks.p_state_change_support);
 	}
-	clk_mgr_base->clks.prev_p_state_change_support = clk_mgr_base->clks.p_state_change_support;
 
 	if (should_set_clock(safe_to_lower, new_clocks->dramclk_khz, clk_mgr_base->clks.dramclk_khz)) {
 		clk_mgr_base->clks.dramclk_khz = new_clocks->dramclk_khz;
@@ -264,50 +218,40 @@ void dcn2_update_clocks(struct clk_mgr *clk_mgr_base,
 			pp_smu->set_hard_min_uclk_by_freq(&pp_smu->pp_smu, clk_mgr_base->clks.dramclk_khz / 1000);
 	}
 
-	if (dc->config.forced_clocks == false) {
-		// First update display clock
-		if (should_set_clock(safe_to_lower, new_clocks->dispclk_khz, clk_mgr_base->clks.dispclk_khz))
-			request_voltage_and_program_disp_clk(clk_mgr_base, new_clocks->dispclk_khz);
-
-		// Updating DPP clock requires some more logic
-		if (!safe_to_lower) {
-			// For pre-programming, we need to make sure any DPP clock that will go up has to go up
+	if (should_set_clock(safe_to_lower, new_clocks->dppclk_khz, clk_mgr->base.clks.dppclk_khz)) {
+		if (clk_mgr->base.clks.dppclk_khz > new_clocks->dppclk_khz)
+			dpp_clock_lowered = true;
+		clk_mgr->base.clks.dppclk_khz = new_clocks->dppclk_khz;
 
-			// First raise the global reference if needed
-			if (new_clocks->dppclk_khz > clk_mgr_base->clks.dppclk_khz)
-				request_voltage_and_program_global_dpp_clk(clk_mgr_base, new_clocks->dppclk_khz);
+		if (pp_smu && pp_smu->set_voltage_by_freq)
+			pp_smu->set_voltage_by_freq(&pp_smu->pp_smu, PP_SMU_NV_PIXELCLK, clk_mgr_base->clks.dppclk_khz / 1000);
 
-			// Then raise any dividers that need raising
-			for (i = 0; i < clk_mgr->base.ctx->dc->res_pool->pipe_count; i++) {
-				int dpp_inst, dppclk_khz;
+		update_dppclk = true;
+	}
 
-				if (!context->res_ctx.pipe_ctx[i].plane_state)
-					continue;
+	if (should_set_clock(safe_to_lower, new_clocks->dispclk_khz, clk_mgr_base->clks.dispclk_khz)) {
+		clk_mgr_base->clks.dispclk_khz = new_clocks->dispclk_khz;
+		if (pp_smu && pp_smu->set_voltage_by_freq)
+			pp_smu->set_voltage_by_freq(&pp_smu->pp_smu, PP_SMU_NV_DISPCLK, clk_mgr_base->clks.dispclk_khz / 1000);
 
-				dpp_inst = context->res_ctx.pipe_ctx[i].plane_res.dpp->inst;
-				dppclk_khz = context->res_ctx.pipe_ctx[i].plane_res.bw.dppclk_khz;
+		update_dispclk = true;
+	}
 
-				clk_mgr->dccg->funcs->update_dpp_dto(clk_mgr->dccg, dpp_inst, dppclk_khz, true);
-			}
+	if (dc->config.forced_clocks == false || (force_reset && safe_to_lower)) {
+		if (dpp_clock_lowered) {
+			// if clock is being lowered, increase DTO before lowering refclk
+			dcn20_update_clocks_update_dpp_dto(clk_mgr, context);
+			dcn20_update_clocks_update_dentist(clk_mgr);
 		} else {
-			// For post-programming, we can lower ref clk if needed, and unconditionally set all the DTOs
-
-			if (new_clocks->dppclk_khz < clk_mgr_base->clks.dppclk_khz)
-				request_voltage_and_program_global_dpp_clk(clk_mgr_base, new_clocks->dppclk_khz);
-
-			for (i = 0; i < clk_mgr->base.ctx->dc->res_pool->pipe_count; i++) {
-				int dpp_inst, dppclk_khz;
-
-				if (!context->res_ctx.pipe_ctx[i].plane_state)
-					continue;
-
-				dpp_inst = context->res_ctx.pipe_ctx[i].plane_res.dpp->inst;
-				dppclk_khz = context->res_ctx.pipe_ctx[i].plane_res.bw.dppclk_khz;
-
-				clk_mgr->dccg->funcs->update_dpp_dto(clk_mgr->dccg, dpp_inst, dppclk_khz, false);
-			}
+			// if clock is being raised, increase refclk before lowering DTO
+			if (update_dppclk || update_dispclk)
+				dcn20_update_clocks_update_dentist(clk_mgr);
+			// always update dtos unless clock is lowered and not safe to lower
+			if (new_clocks->dppclk_khz >= dc->current_state->bw_ctx.bw.dcn.clk.dppclk_khz)
+				dcn20_update_clocks_update_dpp_dto(clk_mgr, context);
 		}
 	}
+
 	if (update_dispclk &&
 			dmcu && dmcu->funcs->is_dmcu_initialized(dmcu)) {
 		/*update dmcu for wait_loop count*/
@@ -320,6 +264,8 @@ void dcn2_update_clocks_fpga(struct clk_mgr *clk_mgr,
 		struct dc_state *context,
 		bool safe_to_lower)
 {
+	struct clk_mgr_internal *clk_mgr_int = TO_CLK_MGR_INTERNAL(clk_mgr);
+
 	struct dc_clocks *new_clocks = &context->bw_ctx.bw.dcn.clk;
 	/* Min fclk = 1.2GHz since all the extra scemi logic seems to run off of it */
 	int fclk_adj = new_clocks->fclk_khz > 1200000 ? new_clocks->fclk_khz : 1200000;
@@ -357,14 +303,18 @@ void dcn2_update_clocks_fpga(struct clk_mgr *clk_mgr,
 		clk_mgr->clks.dispclk_khz = new_clocks->dispclk_khz;
 	}
 
-	/* Both fclk and dppclk ref are run on the same scemi clock so we
-	 * need to keep the same value for both
+	/* Both fclk and ref_dppclk run on the same scemi clock.
+	 * So take the higher value since the DPP DTO is typically programmed
+	 * such that max dppclk is 1:1 with ref_dppclk.
 	 */
 	if (clk_mgr->clks.fclk_khz > clk_mgr->clks.dppclk_khz)
 		clk_mgr->clks.dppclk_khz = clk_mgr->clks.fclk_khz;
 	if (clk_mgr->clks.dppclk_khz > clk_mgr->clks.fclk_khz)
 		clk_mgr->clks.fclk_khz = clk_mgr->clks.dppclk_khz;
 
+	// Both fclk and ref_dppclk run on the same scemi clock.
+	clk_mgr_int->dccg->ref_dppclk = clk_mgr->clks.fclk_khz;
+
 	dm_set_dcn_clocks(clk_mgr->ctx, &clk_mgr->clks);
 }
 
@@ -409,12 +359,36 @@ void dcn2_get_clock(struct clk_mgr *clk_mgr,
 	}
 }
 
+static bool dcn2_are_clock_states_equal(struct dc_clocks *a,
+		struct dc_clocks *b)
+{
+	if (a->dispclk_khz != b->dispclk_khz)
+		return false;
+	else if (a->dppclk_khz != b->dppclk_khz)
+		return false;
+	else if (a->dcfclk_khz != b->dcfclk_khz)
+		return false;
+	else if (a->socclk_khz != b->socclk_khz)
+		return false;
+	else if (a->dcfclk_deep_sleep_khz != b->dcfclk_deep_sleep_khz)
+		return false;
+	else if (a->phyclk_khz != b->phyclk_khz)
+		return false;
+	else if (a->dramclk_khz != b->dramclk_khz)
+		return false;
+	else if (a->p_state_change_support != b->p_state_change_support)
+		return false;
+
+	return true;
+}
+
 static struct clk_mgr_funcs dcn2_funcs = {
 	.get_dp_ref_clk_frequency = dce12_get_dp_ref_freq_khz,
 	.update_clocks = dcn2_update_clocks,
 	.init_clocks = dcn2_init_clocks,
 	.enable_pme_wa = dcn2_enable_pme_wa,
 	.get_clock = dcn2_get_clock,
+	.are_clock_states_equal = dcn2_are_clock_states_equal,
 };
 
 
@@ -442,7 +416,7 @@ void dcn20_clk_mgr_construct(
 
 	if (IS_FPGA_MAXIMUS_DC(ctx->dce_environment)) {
 		dcn2_funcs.update_clocks = dcn2_update_clocks_fpga;
-		clk_mgr->dentist_vco_freq_khz = 3850000;
+		clk_mgr->base.dentist_vco_freq_khz = 3850000;
 
 	} else {
 		/* DFS Slice 2 should be used for DPREFCLK */
@@ -466,15 +440,15 @@ void dcn20_clk_mgr_construct(
 		pll_req = dc_fixpt_mul_int(pll_req, 100000);
 
 		/* integer part is now VCO frequency in kHz */
-		clk_mgr->dentist_vco_freq_khz = dc_fixpt_floor(pll_req);
+		clk_mgr->base.dentist_vco_freq_khz = dc_fixpt_floor(pll_req);
 
 		/* in case we don't get a value from the register, use default */
-		if (clk_mgr->dentist_vco_freq_khz == 0)
-			clk_mgr->dentist_vco_freq_khz = 3850000;
+		if (clk_mgr->base.dentist_vco_freq_khz == 0)
+			clk_mgr->base.dentist_vco_freq_khz = 3850000;
 
 		/* Calculate the DPREFCLK in kHz.*/
 		clk_mgr->base.dprefclk_khz = (DENTIST_DIVIDER_RANGE_SCALE_FACTOR
-			* clk_mgr->dentist_vco_freq_khz) / target_div;
+			* clk_mgr->base.dentist_vco_freq_khz) / target_div;
 	}
 	//Integrated_info table does not exist on dGPU projects so should not be referenced
 	//anywhere in code for dGPUs.
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn20/dcn20_clk_mgr.h b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn20/dcn20_clk_mgr.h
index ac31a9787305..c9fd824f3c23 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn20/dcn20_clk_mgr.h
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn20/dcn20_clk_mgr.h
@@ -50,4 +50,5 @@ void dcn2_get_clock(struct clk_mgr *clk_mgr,
 			enum dc_clock_type clock_type,
 			struct dc_clock_config *clock_cfg);
 
+void dcn20_update_clocks_update_dentist(struct clk_mgr_internal *clk_mgr);
 #endif //__DCN20_CLK_MGR_H__
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr.c
index 787f94d815f4..790a2d211bd6 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr.c
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr.c
@@ -52,6 +52,45 @@
 #define REG(reg_name) \
 	(CLK_BASE.instance[0].segment[mm ## reg_name ## _BASE_IDX] + mm ## reg_name)
 
+
+/* TODO: evaluate how to lower or disable all dcn clocks in screen off case */
+int rn_get_active_display_cnt_wa(
+		struct dc *dc,
+		struct dc_state *context)
+{
+	int i, display_count;
+	bool hdmi_present = false;
+
+	display_count = 0;
+	for (i = 0; i < context->stream_count; i++) {
+		const struct dc_stream_state *stream = context->streams[i];
+
+		if (stream->signal == SIGNAL_TYPE_HDMI_TYPE_A)
+			hdmi_present = true;
+	}
+
+	for (i = 0; i < dc->link_count; i++) {
+		const struct dc_link *link = dc->links[i];
+
+		/*
+		 * Only notify active stream or virtual stream.
+		 * Need to notify virtual stream to work around
+		 * headless case. HPD does not fire when system is in
+		 * S0i2.
+		 */
+		/* abusing the fact that the dig and phy are coupled to see if the phy is enabled */
+		if (link->connector_signal == SIGNAL_TYPE_VIRTUAL ||
+				link->link_enc->funcs->is_dig_enabled(link->link_enc))
+			display_count++;
+	}
+
+	/* WA for hang on HDMI after display off back back on*/
+	if (display_count == 0 && hdmi_present)
+		display_count = 1;
+
+	return display_count;
+}
+
 void rn_update_clocks(struct clk_mgr *clk_mgr_base,
 			struct dc_state *context,
 			bool safe_to_lower)
@@ -62,17 +101,36 @@ void rn_update_clocks(struct clk_mgr *clk_mgr_base,
 	int display_count;
 	bool update_dppclk = false;
 	bool update_dispclk = false;
-	bool enter_display_off = false;
 	bool dpp_clock_lowered = false;
-	struct dmcu *dmcu = clk_mgr_base->ctx->dc->res_pool->dmcu;
 
-	display_count = clk_mgr_helper_get_active_display_cnt(dc, context);
+	struct dmcu *dmcu = clk_mgr_base->ctx->dc->res_pool->dmcu;
 
-	if (display_count == 0)
-		enter_display_off = true;
+	if (dc->work_arounds.skip_clock_update)
+		return;
 
-	if (enter_display_off == safe_to_lower) {
-		rn_vbios_smu_set_display_count(clk_mgr, display_count);
+	/*
+	 * if it is safe to lower, but we are already in the lower state, we don't have to do anything
+	 * also if safe to lower is false, we just go in the higher state
+	 */
+	if (safe_to_lower) {
+		/* check that we're not already in lower */
+		if (clk_mgr_base->clks.pwr_state != DCN_PWR_STATE_LOW_POWER) {
+
+			display_count = rn_get_active_display_cnt_wa(dc, context);
+			/* if we can go lower, go lower */
+			if (display_count == 0) {
+				rn_vbios_smu_set_dcn_low_power_state(clk_mgr, DCN_PWR_STATE_LOW_POWER);
+				/* update power state */
+				clk_mgr_base->clks.pwr_state = DCN_PWR_STATE_LOW_POWER;
+			}
+		}
+	} else {
+		/* check that we're not already in D0 */
+		if (clk_mgr_base->clks.pwr_state != DCN_PWR_STATE_MISSION_MODE) {
+			rn_vbios_smu_set_dcn_low_power_state(clk_mgr, DCN_PWR_STATE_MISSION_MODE);
+			/* update power state */
+			clk_mgr_base->clks.pwr_state = DCN_PWR_STATE_MISSION_MODE;
+		}
 	}
 
 	if (should_set_clock(safe_to_lower, new_clocks->phyclk_khz, clk_mgr_base->clks.phyclk_khz)) {
@@ -113,7 +171,8 @@ void rn_update_clocks(struct clk_mgr *clk_mgr_base,
 		// if clock is being raised, increase refclk before lowering DTO
 		if (update_dppclk || update_dispclk)
 			rn_vbios_smu_set_dppclk(clk_mgr, clk_mgr_base->clks.dppclk_khz);
-		if (update_dppclk)
+		// always update dtos unless clock is lowered and not safe to lower
+		if (new_clocks->dppclk_khz >= dc->current_state->bw_ctx.bw.dcn.clk.dppclk_khz)
 			dcn20_update_clocks_update_dpp_dto(clk_mgr, context);
 	}
 
@@ -319,7 +378,7 @@ void rn_get_clk_states(struct clk_mgr *clk_mgr_base, struct clk_states *s)
 
 	rn_dump_clk_registers(&sb, clk_mgr_base, &log_info);
 
-	s->dprefclk_khz = sb.dprefclk;
+	s->dprefclk_khz = sb.dprefclk * 1000;
 }
 
 void rn_enable_pme_wa(struct clk_mgr *clk_mgr_base)
@@ -329,12 +388,96 @@ void rn_enable_pme_wa(struct clk_mgr *clk_mgr_base)
 	rn_vbios_smu_enable_pme_wa(clk_mgr);
 }
 
+void rn_init_clocks(struct clk_mgr *clk_mgr)
+{
+	memset(&(clk_mgr->clks), 0, sizeof(struct dc_clocks));
+	// Assumption is that boot state always supports pstate
+	clk_mgr->clks.p_state_change_support = true;
+	clk_mgr->clks.prev_p_state_change_support = true;
+	clk_mgr->clks.pwr_state = DCN_PWR_STATE_UNKNOWN;
+}
+
+void build_watermark_ranges(struct clk_bw_params *bw_params, struct pp_smu_wm_range_sets *ranges)
+{
+	int i, num_valid_sets;
+
+	num_valid_sets = 0;
+
+	for (i = 0; i < WM_SET_COUNT; i++) {
+		/* skip empty entries, the smu array has no holes*/
+		if (!bw_params->wm_table.entries[i].valid)
+			continue;
+
+		ranges->reader_wm_sets[num_valid_sets].wm_inst = bw_params->wm_table.entries[i].wm_inst;
+		ranges->reader_wm_sets[num_valid_sets].wm_type = bw_params->wm_table.entries[i].wm_type;;
+		/* We will not select WM based on dcfclk, so leave it as unconstrained */
+		ranges->reader_wm_sets[num_valid_sets].min_drain_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MIN;
+		ranges->reader_wm_sets[num_valid_sets].max_drain_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MAX;
+		/* fclk wil be used to select WM*/
+
+		if (ranges->reader_wm_sets[num_valid_sets].wm_type == WM_TYPE_PSTATE_CHG) {
+			if (i == 0)
+				ranges->reader_wm_sets[num_valid_sets].min_fill_clk_mhz = 0;
+			else {
+				/* add 1 to make it non-overlapping with next lvl */
+				ranges->reader_wm_sets[num_valid_sets].min_fill_clk_mhz = bw_params->clk_table.entries[i - 1].fclk_mhz + 1;
+			}
+			ranges->reader_wm_sets[num_valid_sets].max_fill_clk_mhz = bw_params->clk_table.entries[i].fclk_mhz;
+
+		} else {
+			/* unconstrained for memory retraining */
+			ranges->reader_wm_sets[num_valid_sets].min_fill_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MIN;
+			ranges->reader_wm_sets[num_valid_sets].max_fill_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MAX;
+
+			/* Modify previous watermark range to cover up to max */
+			ranges->reader_wm_sets[num_valid_sets - 1].max_fill_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MAX;
+		}
+		num_valid_sets++;
+	}
+
+	ASSERT(num_valid_sets != 0); /* Must have at least one set of valid watermarks */
+	ranges->num_reader_wm_sets = num_valid_sets;
+
+	/* modify the min and max to make sure we cover the whole range*/
+	ranges->reader_wm_sets[0].min_drain_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MIN;
+	ranges->reader_wm_sets[0].min_fill_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MIN;
+	ranges->reader_wm_sets[ranges->num_reader_wm_sets - 1].max_drain_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MAX;
+	ranges->reader_wm_sets[ranges->num_reader_wm_sets - 1].max_fill_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MAX;
+
+	/* This is for writeback only, does not matter currently as no writeback support*/
+	ranges->num_writer_wm_sets = 1;
+	ranges->writer_wm_sets[0].wm_inst = WM_A;
+	ranges->writer_wm_sets[0].min_fill_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MIN;
+	ranges->writer_wm_sets[0].max_fill_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MAX;
+	ranges->writer_wm_sets[0].min_drain_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MIN;
+	ranges->writer_wm_sets[0].max_drain_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MAX;
+
+}
+
+static void rn_notify_wm_ranges(struct clk_mgr *clk_mgr_base)
+{
+	struct dc_debug_options *debug = &clk_mgr_base->ctx->dc->debug;
+	struct pp_smu_wm_range_sets ranges = {0};
+	struct clk_mgr_internal *clk_mgr = TO_CLK_MGR_INTERNAL(clk_mgr_base);
+	struct pp_smu_funcs *pp_smu = clk_mgr->pp_smu;
+
+	if (!debug->disable_pplib_wm_range) {
+		build_watermark_ranges(clk_mgr_base->bw_params, &ranges);
+
+		/* Notify PP Lib/SMU which Watermarks to use for which clock ranges */
+		if (pp_smu && pp_smu->rn_funcs.set_wm_ranges)
+			pp_smu->rn_funcs.set_wm_ranges(&pp_smu->rn_funcs.pp_smu, &ranges);
+	}
+
+}
+
 static struct clk_mgr_funcs dcn21_funcs = {
 	.get_dp_ref_clk_frequency = dce12_get_dp_ref_freq_khz,
 	.update_clocks = rn_update_clocks,
-	.init_clocks = dcn2_init_clocks,
+	.init_clocks = rn_init_clocks,
 	.enable_pme_wa = rn_enable_pme_wa,
-	/* .dump_clk_registers = rn_dump_clk_registers */
+	/* .dump_clk_registers = rn_dump_clk_registers, */
+	.notify_wm_ranges = rn_notify_wm_ranges
 };
 
 struct clk_bw_params rn_bw_params = {
@@ -405,80 +548,50 @@ struct clk_bw_params rn_bw_params = {
 	}
 };
 
-void build_watermark_ranges(struct clk_bw_params *bw_params, struct pp_smu_wm_range_sets *ranges)
+static unsigned int find_dcfclk_for_voltage(struct dpm_clocks *clock_table, unsigned int voltage)
 {
-	int i, num_valid_sets;
-
-	num_valid_sets = 0;
-
-	for (i = 0; i < WM_SET_COUNT; i++) {
-		/* skip empty entries, the smu array has no holes*/
-		if (!bw_params->wm_table.entries[i].valid)
-			continue;
-
-		ranges->reader_wm_sets[num_valid_sets].wm_inst = bw_params->wm_table.entries[i].wm_inst;
-		ranges->reader_wm_sets[num_valid_sets].wm_type = bw_params->wm_table.entries[i].wm_type;;
-		/* We will not select WM based on dcfclk, so leave it as unconstrained */
-		ranges->reader_wm_sets[num_valid_sets].min_drain_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MIN;
-		ranges->reader_wm_sets[num_valid_sets].max_drain_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MAX;
-		/* fclk wil be used to select WM*/
-
-		if (ranges->reader_wm_sets[num_valid_sets].wm_type == WM_TYPE_PSTATE_CHG) {
-			if (i == 0)
-				ranges->reader_wm_sets[num_valid_sets].min_fill_clk_mhz = 0;
-			else {
-				/* add 1 to make it non-overlapping with next lvl */
-				ranges->reader_wm_sets[num_valid_sets].min_fill_clk_mhz = bw_params->clk_table.entries[i - 1].fclk_mhz + 1;
-			}
-			ranges->reader_wm_sets[num_valid_sets].max_fill_clk_mhz = bw_params->clk_table.entries[i].fclk_mhz;
-
-		} else {
-			/* unconstrained for memory retraining */
-			ranges->reader_wm_sets[num_valid_sets].min_fill_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MIN;
-			ranges->reader_wm_sets[num_valid_sets].max_fill_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MAX;
+	int i;
 
-			/* Modify previous watermark range to cover up to max */
-			ranges->reader_wm_sets[num_valid_sets - 1].max_fill_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MAX;
-		}
-		num_valid_sets++;
+	for (i = 0; i < PP_SMU_NUM_DCFCLK_DPM_LEVELS; i++) {
+		if (clock_table->DcfClocks[i].Vol == voltage)
+			return clock_table->DcfClocks[i].Freq;
 	}
 
-	ASSERT(num_valid_sets != 0); /* Must have at least one set of valid watermarks */
-	ranges->num_reader_wm_sets = num_valid_sets;
-
-	/* modify the min and max to make sure we cover the whole range*/
-	ranges->reader_wm_sets[0].min_drain_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MIN;
-	ranges->reader_wm_sets[0].min_fill_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MIN;
-	ranges->reader_wm_sets[ranges->num_reader_wm_sets - 1].max_drain_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MAX;
-	ranges->reader_wm_sets[ranges->num_reader_wm_sets - 1].max_fill_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MAX;
-
-	/* This is for writeback only, does not matter currently as no writeback support*/
-	ranges->num_writer_wm_sets = 1;
-	ranges->writer_wm_sets[0].wm_inst = WM_A;
-	ranges->writer_wm_sets[0].min_fill_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MIN;
-	ranges->writer_wm_sets[0].max_fill_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MAX;
-	ranges->writer_wm_sets[0].min_drain_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MIN;
-	ranges->writer_wm_sets[0].max_drain_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MAX;
-
+	ASSERT(0);
+	return 0;
 }
 
-void clk_mgr_helper_populate_bw_params(struct clk_bw_params *bw_params, struct dpm_clocks *clock_table, struct hw_asic_id *asic_id)
+static void rn_clk_mgr_helper_populate_bw_params(struct clk_bw_params *bw_params, struct dpm_clocks *clock_table, struct hw_asic_id *asic_id)
 {
-	int i;
+	int i, j = 0;
+
+	j = -1;
 
 	ASSERT(PP_SMU_NUM_FCLK_DPM_LEVELS <= MAX_NUM_DPM_LVL);
 
-	for (i = 0; i < PP_SMU_NUM_FCLK_DPM_LEVELS; i++) {
-		if (clock_table->FClocks[i].Freq == 0)
+	/* Find lowest DPM, FCLK is filled in reverse order*/
+
+	for (i = PP_SMU_NUM_FCLK_DPM_LEVELS - 1; i >= 0; i--) {
+		if (clock_table->FClocks[i].Freq != 0) {
+			j = i;
 			break;
+		}
+	}
+
+	if (j == -1) {
+		/* clock table is all 0s, just use our own hardcode */
+		ASSERT(0);
+		return;
+	}
+
+	bw_params->clk_table.num_entries = j + 1;
 
-		bw_params->clk_table.entries[i].dcfclk_mhz = clock_table->DcfClocks[i].Freq;
-		bw_params->clk_table.entries[i].fclk_mhz = clock_table->FClocks[i].Freq;
-		bw_params->clk_table.entries[i].memclk_mhz = clock_table->MemClocks[i].Freq;
-		bw_params->clk_table.entries[i].socclk_mhz = clock_table->SocClocks[i].Freq;
-		bw_params->clk_table.entries[i].voltage = clock_table->FClocks[i].Vol;
+	for (i = 0; i < bw_params->clk_table.num_entries; i++, j--) {
+		bw_params->clk_table.entries[i].fclk_mhz = clock_table->FClocks[j].Freq;
+		bw_params->clk_table.entries[i].memclk_mhz = clock_table->MemClocks[j].Freq;
+		bw_params->clk_table.entries[i].voltage = clock_table->FClocks[j].Vol;
+		bw_params->clk_table.entries[i].dcfclk_mhz = find_dcfclk_for_voltage(clock_table, clock_table->FClocks[j].Vol);
 	}
-	bw_params->clk_table.num_entries = i;
 
 	bw_params->vram_type = asic_id->vram_type;
 	bw_params->num_channels = asic_id->vram_width / DDR4_DRAM_WIDTH;
@@ -486,7 +599,7 @@ void clk_mgr_helper_populate_bw_params(struct clk_bw_params *bw_params, struct d
 	for (i = 0; i < WM_SET_COUNT; i++) {
 		bw_params->wm_table.entries[i].wm_inst = i;
 
-		if (clock_table->FClocks[i].Freq == 0) {
+		if (i >= bw_params->clk_table.num_entries) {
 			bw_params->wm_table.entries[i].valid = false;
 			continue;
 		}
@@ -534,57 +647,42 @@ void rn_clk_mgr_construct(
 
 	if (IS_FPGA_MAXIMUS_DC(ctx->dce_environment)) {
 		dcn21_funcs.update_clocks = dcn2_update_clocks_fpga;
-		clk_mgr->dentist_vco_freq_khz = 3600000;
+		clk_mgr->base.dentist_vco_freq_khz = 3600000;
 		clk_mgr->base.dprefclk_khz = 600000;
 	} else {
 		struct clk_log_info log_info = {0};
 
 		/* TODO: Check we get what we expect during bringup */
-		clk_mgr->dentist_vco_freq_khz = get_vco_frequency_from_reg(clk_mgr);
+		clk_mgr->base.dentist_vco_freq_khz = get_vco_frequency_from_reg(clk_mgr);
 
 		/* in case we don't get a value from the register, use default */
-		if (clk_mgr->dentist_vco_freq_khz == 0)
-			clk_mgr->dentist_vco_freq_khz = 3600000;
+		if (clk_mgr->base.dentist_vco_freq_khz == 0)
+			clk_mgr->base.dentist_vco_freq_khz = 3600000;
 
 		rn_dump_clk_registers(&s, &clk_mgr->base, &log_info);
-		clk_mgr->base.dprefclk_khz = s.dprefclk;
-
-		if (clk_mgr->base.dprefclk_khz != 600000) {
-			clk_mgr->base.dprefclk_khz = 600000;
-			ASSERT(1); //TODO: Renoir follow up.
-		}
+		/* Convert dprefclk units from MHz to KHz */
+		/* Value already divided by 10, some resolution lost */
+		clk_mgr->base.dprefclk_khz = s.dprefclk * 1000;
 
 		/* in case we don't get a value from the register, use default */
-		if (clk_mgr->base.dprefclk_khz == 0)
+		if (clk_mgr->base.dprefclk_khz == 0) {
+			ASSERT(clk_mgr->base.dprefclk_khz == 600000);
 			clk_mgr->base.dprefclk_khz = 600000;
+		}
 	}
 
 	dce_clock_read_ss_info(clk_mgr);
 
 	clk_mgr->base.bw_params = &rn_bw_params;
 
-	if (pp_smu) {
+	if (pp_smu && pp_smu->rn_funcs.get_dpm_clock_table) {
 		pp_smu->rn_funcs.get_dpm_clock_table(&pp_smu->rn_funcs.pp_smu, &clock_table);
-		clk_mgr_helper_populate_bw_params(clk_mgr->base.bw_params, &clock_table, &ctx->asic_id);
+		rn_clk_mgr_helper_populate_bw_params(clk_mgr->base.bw_params, &clock_table, &ctx->asic_id);
 	}
 
-	/*
-	 * Notify SMU which set of WM should be selected for different ranges of fclk
-	 * On Renoir there is a maximumum of 4 DF pstates supported, could be less
-	 * depending on DDR speed and fused maximum fclk.
-	 */
-	if (!debug->disable_pplib_wm_range) {
-		struct pp_smu_wm_range_sets ranges = {0};
-
-		build_watermark_ranges(clk_mgr->base.bw_params, &ranges);
-
-		/* Notify PP Lib/SMU which Watermarks to use for which clock ranges */
-		if (pp_smu && pp_smu->rn_funcs.set_wm_ranges)
-			pp_smu->rn_funcs.set_wm_ranges(&pp_smu->rn_funcs.pp_smu, &ranges);
+	if (!IS_FPGA_MAXIMUS_DC(ctx->dce_environment) && clk_mgr->smu_ver >= 0x00371500) {
+		/* enable powerfeatures when displaycount goes to 0 */
+		rn_vbios_smu_enable_48mhz_tmdp_refclk_pwrdwn(clk_mgr, !debug->disable_48mhz_pwrdwn);
 	}
-
-	/* enable powerfeatures when displaycount goes to 0 */
-	if (!debug->disable_48mhz_pwrdwn)
-		rn_vbios_smu_enable_48mhz_tmdp_refclk_pwrdwn(clk_mgr);
 }
 
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr.h b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr.h
index aadec06fde10..e4322fa5475b 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr.h
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr.h
@@ -26,11 +26,13 @@
 #ifndef __RN_CLK_MGR_H__
 #define __RN_CLK_MGR_H__
 
+#include "clk_mgr.h"
+#include "dm_pp_smu.h"
+
 struct rn_clk_registers {
 	uint32_t CLK1_CLK0_CURRENT_CNT; /* DPREFCLK */
 };
 
-
 void rn_clk_mgr_construct(struct dc_context *ctx,
 		struct clk_mgr_internal *clk_mgr,
 		struct pp_smu_funcs *pp_smu,
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr_vbios_smu.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr_vbios_smu.c
index 50984c1811bb..cb7c0e8b7e1b 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr_vbios_smu.c
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr_vbios_smu.c
@@ -33,7 +33,7 @@
 #include "mp/mp_12_0_0_sh_mask.h"
 
 #define REG(reg_name) \
-	(MP1_BASE.instance[0].segment[mm ## reg_name ## _BASE_IDX] + mm ## reg_name)
+	(MP0_BASE.instance[0].segment[mm ## reg_name ## _BASE_IDX] + mm ## reg_name)
 
 #define FN(reg_name, field) \
 	FD(reg_name##__##field)
@@ -84,16 +84,12 @@ int rn_vbios_smu_set_dispclk(struct clk_mgr_internal *clk_mgr, int requested_dis
 	int actual_dispclk_set_mhz = -1;
 	struct dc *core_dc = clk_mgr->base.ctx->dc;
 	struct dmcu *dmcu = core_dc->res_pool->dmcu;
-	uint32_t clk = requested_dispclk_khz / 1000;
-
-	if (clk <= 100)
-		clk = 101;
 
 	/*  Unit of SMU msg parameter is Mhz */
 	actual_dispclk_set_mhz = rn_vbios_smu_send_msg_with_param(
 			clk_mgr,
 			VBIOSSMC_MSG_SetDispclkFreq,
-			clk);
+			requested_dispclk_khz / 1000);
 
 	if (!IS_FPGA_MAXIMUS_DC(core_dc->ctx->dce_environment)) {
 		if (dmcu && dmcu->funcs->is_dmcu_initialized(dmcu)) {
@@ -124,7 +120,7 @@ int rn_vbios_smu_set_hard_min_dcfclk(struct clk_mgr_internal *clk_mgr, int reque
 {
 	int actual_dcfclk_set_mhz = -1;
 
-	if (clk_mgr->smu_ver < 0xFFFFFFFF)
+	if (clk_mgr->smu_ver < 0x370c00)
 		return actual_dcfclk_set_mhz;
 
 	actual_dcfclk_set_mhz = rn_vbios_smu_send_msg_with_param(
@@ -139,7 +135,7 @@ int rn_vbios_smu_set_min_deep_sleep_dcfclk(struct clk_mgr_internal *clk_mgr, int
 {
 	int actual_min_ds_dcfclk_mhz = -1;
 
-	if (clk_mgr->smu_ver < 0xFFFFFFFF)
+	if (clk_mgr->smu_ver < 0x370c00)
 		return actual_min_ds_dcfclk_mhz;
 
 	actual_min_ds_dcfclk_mhz = rn_vbios_smu_send_msg_with_param(
@@ -162,33 +158,35 @@ int rn_vbios_smu_set_dppclk(struct clk_mgr_internal *clk_mgr, int requested_dpp_
 {
 	int actual_dppclk_set_mhz = -1;
 
-	uint32_t clk = requested_dpp_khz / 1000;
-
-	if (clk <= 100)
-		clk = 101;
-
 	actual_dppclk_set_mhz = rn_vbios_smu_send_msg_with_param(
 			clk_mgr,
 			VBIOSSMC_MSG_SetDppclkFreq,
-			clk);
+			requested_dpp_khz / 1000);
 
 	return actual_dppclk_set_mhz * 1000;
 }
 
-void rn_vbios_smu_set_display_count(struct clk_mgr_internal *clk_mgr, int display_count)
+void rn_vbios_smu_set_dcn_low_power_state(struct clk_mgr_internal *clk_mgr, enum dcn_pwr_state state)
 {
+	int disp_count;
+
+	if (state == DCN_PWR_STATE_LOW_POWER)
+		disp_count = 0;
+	else
+		disp_count = 1;
+
 	rn_vbios_smu_send_msg_with_param(
-			clk_mgr,
-			VBIOSSMC_MSG_SetDisplayCount,
-			display_count);
+		clk_mgr,
+		VBIOSSMC_MSG_SetDisplayCount,
+		disp_count);
 }
 
-void rn_vbios_smu_enable_48mhz_tmdp_refclk_pwrdwn(struct clk_mgr_internal *clk_mgr)
+void rn_vbios_smu_enable_48mhz_tmdp_refclk_pwrdwn(struct clk_mgr_internal *clk_mgr, bool enable)
 {
 	rn_vbios_smu_send_msg_with_param(
 			clk_mgr,
 			VBIOSSMC_MSG_EnableTmdp48MHzRefclkPwrDown,
-			0);
+			enable);
 }
 
 void rn_vbios_smu_enable_pme_wa(struct clk_mgr_internal *clk_mgr)
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr_vbios_smu.h b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr_vbios_smu.h
index da3a49487c6d..ccc01879c9d4 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr_vbios_smu.h
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr_vbios_smu.h
@@ -33,8 +33,8 @@ int rn_vbios_smu_set_hard_min_dcfclk(struct clk_mgr_internal *clk_mgr, int reque
 int rn_vbios_smu_set_min_deep_sleep_dcfclk(struct clk_mgr_internal *clk_mgr, int requested_min_ds_dcfclk_khz);
 void rn_vbios_smu_set_phyclk(struct clk_mgr_internal *clk_mgr, int requested_phyclk_khz);
 int rn_vbios_smu_set_dppclk(struct clk_mgr_internal *clk_mgr, int requested_dpp_khz);
-void rn_vbios_smu_set_display_count(struct clk_mgr_internal *clk_mgr, int display_count);
-void rn_vbios_smu_enable_48mhz_tmdp_refclk_pwrdwn(struct clk_mgr_internal *clk_mgr);
+void rn_vbios_smu_set_dcn_low_power_state(struct clk_mgr_internal *clk_mgr, int display_count);
+void rn_vbios_smu_enable_48mhz_tmdp_refclk_pwrdwn(struct clk_mgr_internal *clk_mgr, bool enable);
 void rn_vbios_smu_enable_pme_wa(struct clk_mgr_internal *clk_mgr);
 
 #endif /* DAL_DC_DCN10_RV1_CLK_MGR_VBIOS_SMU_H_ */
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c
index 5d1adeda4d90..32f31bf91915 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc.c
@@ -194,7 +194,7 @@ static bool create_links(
 				}
 			}
 
-			if (!should_destory_link) {
+			if (dc->config.force_enum_edp || !should_destory_link) {
 				dc->links[dc->link_count] = link;
 				link->dc = dc;
 				++dc->link_count;
@@ -411,6 +411,27 @@ bool dc_stream_get_crc(struct dc *dc, struct dc_stream_state *stream,
 	return false;
 }
 
+void dc_stream_set_dyn_expansion(struct dc *dc, struct dc_stream_state *stream,
+		enum dc_dynamic_expansion option)
+{
+	/* OPP FMT dyn expansion updates*/
+	int i = 0;
+	struct pipe_ctx *pipe_ctx;
+
+	for (i = 0; i < MAX_PIPES; i++) {
+		if (dc->current_state->res_ctx.pipe_ctx[i].stream
+				== stream) {
+			pipe_ctx = &dc->current_state->res_ctx.pipe_ctx[i];
+			pipe_ctx->stream_res.opp->dyn_expansion = option;
+			pipe_ctx->stream_res.opp->funcs->opp_set_dyn_expansion(
+					pipe_ctx->stream_res.opp,
+					COLOR_SPACE_YCBCR601,
+					stream->timing.display_color_depth,
+					stream->signal);
+		}
+	}
+}
+
 void dc_stream_set_dither_option(struct dc_stream_state *stream,
 		enum dc_dither_option option)
 {
@@ -580,6 +601,10 @@ static bool construct(struct dc *dc,
 #ifdef CONFIG_DRM_AMD_DC_DCN2_0
 	// Allocate memory for the vm_helper
 	dc->vm_helper = kzalloc(sizeof(struct vm_helper), GFP_KERNEL);
+	if (!dc->vm_helper) {
+		dm_error("%s: failed to create dc->vm_helper\n", __func__);
+		goto fail;
+	}
 
 #endif
 	memcpy(&dc->bb_overrides, &init_params->bb_overrides, sizeof(dc->bb_overrides));
@@ -761,8 +786,13 @@ static void disable_dangling_plane(struct dc *dc, struct dc_state *context)
 #if defined(CONFIG_DRM_AMD_DC_DCN2_0)
 			disable_all_writeback_pipes_for_stream(dc, old_stream, dangling_context);
 #endif
-			dc->hwss.apply_ctx_for_surface(dc, old_stream, 0, dangling_context);
+			if (dc->hwss.apply_ctx_for_surface)
+				dc->hwss.apply_ctx_for_surface(dc, old_stream, 0, dangling_context);
 		}
+#if defined(CONFIG_DRM_AMD_DC_DCN2_0)
+		if (dc->hwss.program_front_end_for_ctx)
+			dc->hwss.program_front_end_for_ctx(dc, dangling_context);
+#endif
 	}
 
 	current_ctx = dc->current_state;
@@ -785,9 +815,6 @@ struct dc *dc_create(const struct dc_init_data *init_params)
 	if (false == construct(dc, init_params))
 		goto construct_fail;
 
-	/*TODO: separate HW and SW initialization*/
-	dc->hwss.init_hw(dc);
-
 	full_pipe_count = dc->res_pool->pipe_count;
 	if (dc->res_pool->underlay_pipe_index != NO_UNDERLAY_PIPE)
 		full_pipe_count--;
@@ -820,9 +847,24 @@ alloc_fail:
 	return NULL;
 }
 
+void dc_hardware_init(struct dc *dc)
+{
+	dc->hwss.init_hw(dc);
+}
+
 void dc_init_callbacks(struct dc *dc,
 		const struct dc_callback_init *init_params)
 {
+#ifdef CONFIG_DRM_AMD_DC_HDCP
+	dc->ctx->cp_psp = init_params->cp_psp;
+#endif
+}
+
+void dc_deinit_callbacks(struct dc *dc)
+{
+#ifdef CONFIG_DRM_AMD_DC_HDCP
+	memset(&dc->ctx->cp_psp, 0, sizeof(dc->ctx->cp_psp));
+#endif
 }
 
 void dc_destroy(struct dc **dc)
@@ -901,15 +943,11 @@ static void program_timing_sync(
 
 		/* set first pipe with plane as master */
 		for (j = 0; j < group_size; j++) {
-			struct pipe_ctx *temp;
-
 			if (pipe_set[j]->plane_state) {
 				if (j == 0)
 					break;
 
-				temp = pipe_set[0];
-				pipe_set[0] = pipe_set[j];
-				pipe_set[j] = temp;
+				swap(pipe_set[0], pipe_set[j]);
 				break;
 			}
 		}
@@ -966,40 +1004,87 @@ bool dc_validate_seamless_boot_timing(const struct dc *dc,
 				struct dc_crtc_timing *crtc_timing)
 {
 	struct timing_generator *tg;
+	struct stream_encoder *se = NULL;
+
+	struct dc_crtc_timing hw_crtc_timing = {0};
+
 	struct dc_link *link = sink->link;
-	unsigned int enc_inst, tg_inst;
+	unsigned int i, enc_inst, tg_inst = 0;
+
+	// Seamless port only support single DP and EDP so far
+	if (sink->sink_signal != SIGNAL_TYPE_DISPLAY_PORT &&
+		sink->sink_signal != SIGNAL_TYPE_EDP)
+		return false;
 
 	/* Check for enabled DIG to identify enabled display */
 	if (!link->link_enc->funcs->is_dig_enabled(link->link_enc))
 		return false;
 
-	/* Check for which front end is used by this encoder.
-	 * Note the inst is 1 indexed, where 0 is undefined.
-	 * Note that DIG_FE can source from different OTG but our
-	 * current implementation always map 1-to-1, so this code makes
-	 * the same assumption and doesn't check OTG source.
-	 */
 	enc_inst = link->link_enc->funcs->get_dig_frontend(link->link_enc);
 
-	/* Instance should be within the range of the pool */
-	if (enc_inst >= dc->res_pool->pipe_count)
+	if (enc_inst == ENGINE_ID_UNKNOWN)
 		return false;
 
-	if (enc_inst >= dc->res_pool->stream_enc_count)
-		return false;
+	for (i = 0; i < dc->res_pool->stream_enc_count; i++) {
+		if (dc->res_pool->stream_enc[i]->id == enc_inst) {
+
+			se = dc->res_pool->stream_enc[i];
+
+			tg_inst = dc->res_pool->stream_enc[i]->funcs->dig_source_otg(
+				dc->res_pool->stream_enc[i]);
+			break;
+		}
+	}
 
-	tg_inst = dc->res_pool->stream_enc[enc_inst]->funcs->dig_source_otg(
-		dc->res_pool->stream_enc[enc_inst]);
+	// tg_inst not found
+	if (i == dc->res_pool->stream_enc_count)
+		return false;
 
 	if (tg_inst >= dc->res_pool->timing_generator_count)
 		return false;
 
 	tg = dc->res_pool->timing_generators[tg_inst];
 
-	if (!tg->funcs->is_matching_timing)
+	if (!tg->funcs->get_hw_timing)
+		return false;
+
+	if (!tg->funcs->get_hw_timing(tg, &hw_crtc_timing))
+		return false;
+
+	if (crtc_timing->h_total != hw_crtc_timing.h_total)
+		return false;
+
+	if (crtc_timing->h_border_left != hw_crtc_timing.h_border_left)
+		return false;
+
+	if (crtc_timing->h_addressable != hw_crtc_timing.h_addressable)
+		return false;
+
+	if (crtc_timing->h_border_right != hw_crtc_timing.h_border_right)
+		return false;
+
+	if (crtc_timing->h_front_porch != hw_crtc_timing.h_front_porch)
+		return false;
+
+	if (crtc_timing->h_sync_width != hw_crtc_timing.h_sync_width)
+		return false;
+
+	if (crtc_timing->v_total != hw_crtc_timing.v_total)
+		return false;
+
+	if (crtc_timing->v_border_top != hw_crtc_timing.v_border_top)
 		return false;
 
-	if (!tg->funcs->is_matching_timing(tg, crtc_timing))
+	if (crtc_timing->v_addressable != hw_crtc_timing.v_addressable)
+		return false;
+
+	if (crtc_timing->v_border_bottom != hw_crtc_timing.v_border_bottom)
+		return false;
+
+	if (crtc_timing->v_front_porch != hw_crtc_timing.v_front_porch)
+		return false;
+
+	if (crtc_timing->v_sync_width != hw_crtc_timing.v_sync_width)
 		return false;
 
 	if (dc_is_dp_signal(link->connector_signal)) {
@@ -1012,6 +1097,20 @@ bool dc_validate_seamless_boot_timing(const struct dc *dc,
 		if (crtc_timing->pix_clk_100hz != pix_clk_100hz)
 			return false;
 
+		if (!se->funcs->dp_get_pixel_format)
+			return false;
+
+		if (!se->funcs->dp_get_pixel_format(
+			se,
+			&hw_crtc_timing.pixel_encoding,
+			&hw_crtc_timing.display_color_depth))
+			return false;
+
+		if (hw_crtc_timing.display_color_depth != crtc_timing->display_color_depth)
+			return false;
+
+		if (hw_crtc_timing.pixel_encoding != crtc_timing->pixel_encoding)
+			return false;
 	}
 
 	return true;
@@ -1073,15 +1172,20 @@ static enum dc_status dc_commit_state_no_check(struct dc *dc, struct dc_state *c
 	/* re-program planes for existing stream, in case we need to
 	 * free up plane resource for later use
 	 */
-	for (i = 0; i < context->stream_count; i++) {
-		if (context->streams[i]->mode_changed)
-			continue;
+	if (dc->hwss.apply_ctx_for_surface)
+		for (i = 0; i < context->stream_count; i++) {
+			if (context->streams[i]->mode_changed)
+				continue;
 
-		dc->hwss.apply_ctx_for_surface(
-			dc, context->streams[i],
-			context->stream_status[i].plane_count,
-			context); /* use new pipe config in new context */
-	}
+			dc->hwss.apply_ctx_for_surface(
+				dc, context->streams[i],
+				context->stream_status[i].plane_count,
+				context); /* use new pipe config in new context */
+		}
+#if defined(CONFIG_DRM_AMD_DC_DCN2_0)
+	if (dc->hwss.program_front_end_for_ctx)
+		dc->hwss.program_front_end_for_ctx(dc, context);
+#endif
 
 	/* Program hardware */
 	for (i = 0; i < dc->res_pool->pipe_count; i++) {
@@ -1100,16 +1204,21 @@ static enum dc_status dc_commit_state_no_check(struct dc *dc, struct dc_state *c
 	}
 
 	/* Program all planes within new context*/
+#if defined(CONFIG_DRM_AMD_DC_DCN2_0)
+	if (dc->hwss.program_front_end_for_ctx)
+		dc->hwss.program_front_end_for_ctx(dc, context);
+#endif
 	for (i = 0; i < context->stream_count; i++) {
 		const struct dc_link *link = context->streams[i]->link;
 
 		if (!context->streams[i]->mode_changed)
 			continue;
 
-		dc->hwss.apply_ctx_for_surface(
-				dc, context->streams[i],
-				context->stream_status[i].plane_count,
-				context);
+		if (dc->hwss.apply_ctx_for_surface)
+			dc->hwss.apply_ctx_for_surface(
+					dc, context->streams[i],
+					context->stream_status[i].plane_count,
+					context);
 
 		/*
 		 * enable stereo
@@ -1136,15 +1245,9 @@ static enum dc_status dc_commit_state_no_check(struct dc *dc, struct dc_state *c
 
 	dc_enable_stereo(dc, context, dc_streams, context->stream_count);
 
-	if (!dc->optimize_seamless_boot)
-		/* pplib is notified if disp_num changed */
-		dc->hwss.optimize_bandwidth(dc, context);
-
 	for (i = 0; i < context->stream_count; i++)
 		context->streams[i]->mode_changed = false;
 
-	memset(&context->commit_hints, 0, sizeof(context->commit_hints));
-
 	dc_release_state(dc->current_state);
 
 	dc->current_state = context;
@@ -1492,20 +1595,15 @@ static enum surface_update_type det_surface_update(const struct dc *dc,
 	enum surface_update_type overall_type = UPDATE_TYPE_FAST;
 	union surface_update_flags *update_flags = &u->surface->update_flags;
 
-	update_flags->raw = 0; // Reset all flags
-
 	if (u->flip_addr)
 		update_flags->bits.addr_update = 1;
 
-	if (!is_surface_in_context(context, u->surface)) {
-		update_flags->bits.new_plane = 1;
+	if (!is_surface_in_context(context, u->surface) || u->surface->force_full_update) {
+		update_flags->raw = 0xFFFFFFFF;
 		return UPDATE_TYPE_FULL;
 	}
 
-	if (u->surface->force_full_update) {
-		update_flags->bits.full_update = 1;
-		return UPDATE_TYPE_FULL;
-	}
+	update_flags->raw = 0; // Reset all flags
 
 	type = get_plane_info_update_type(u);
 	elevate_update_type(&overall_type, type);
@@ -1563,40 +1661,43 @@ static enum surface_update_type check_update_surfaces_for_stream(
 	enum surface_update_type overall_type = UPDATE_TYPE_FAST;
 
 	if (stream_status == NULL || stream_status->plane_count != surface_count)
-		return UPDATE_TYPE_FULL;
+		overall_type = UPDATE_TYPE_FULL;
 
 	/* some stream updates require passive update */
 	if (stream_update) {
-		if ((stream_update->src.height != 0) &&
-				(stream_update->src.width != 0))
-			return UPDATE_TYPE_FULL;
+		union stream_update_flags *su_flags = &stream_update->stream->update_flags;
 
-		if ((stream_update->dst.height != 0) &&
-				(stream_update->dst.width != 0))
-			return UPDATE_TYPE_FULL;
+		if ((stream_update->src.height != 0 && stream_update->src.width != 0) ||
+				(stream_update->dst.height != 0 && stream_update->dst.width != 0))
+			su_flags->bits.scaling = 1;
 
 		if (stream_update->out_transfer_func)
-			return UPDATE_TYPE_FULL;
+			su_flags->bits.out_tf = 1;
 
 		if (stream_update->abm_level)
-			return UPDATE_TYPE_FULL;
+			su_flags->bits.abm_level = 1;
 
 		if (stream_update->dpms_off)
-			return UPDATE_TYPE_FULL;
+			su_flags->bits.dpms_off = 1;
+
+		if (stream_update->gamut_remap)
+			su_flags->bits.gamut_remap = 1;
 
 #if defined(CONFIG_DRM_AMD_DC_DCN2_0)
 		if (stream_update->wb_update)
-			return UPDATE_TYPE_FULL;
+			su_flags->bits.wb_update = 1;
 #endif
+		if (su_flags->raw != 0)
+			overall_type = UPDATE_TYPE_FULL;
+
+		if (stream_update->output_csc_transform || stream_update->output_color_space)
+			su_flags->bits.out_csc = 1;
 	}
 
 	for (i = 0 ; i < surface_count; i++) {
 		enum surface_update_type type =
 				det_surface_update(dc, &updates[i]);
 
-		if (type == UPDATE_TYPE_FULL)
-			return type;
-
 		elevate_update_type(&overall_type, type);
 	}
 
@@ -1618,16 +1719,29 @@ enum surface_update_type dc_check_update_surfaces_for_stream(
 	int i;
 	enum surface_update_type type;
 
+	if (stream_update)
+		stream_update->stream->update_flags.raw = 0;
 	for (i = 0; i < surface_count; i++)
 		updates[i].surface->update_flags.raw = 0;
 
 	type = check_update_surfaces_for_stream(dc, updates, surface_count, stream_update, stream_status);
-	if (type == UPDATE_TYPE_FULL)
+	if (type == UPDATE_TYPE_FULL) {
+		if (stream_update)
+			stream_update->stream->update_flags.raw = 0xFFFFFFFF;
 		for (i = 0; i < surface_count; i++)
 			updates[i].surface->update_flags.raw = 0xFFFFFFFF;
+	}
 
-	if (type == UPDATE_TYPE_FAST && memcmp(&dc->current_state->bw_ctx.bw.dcn.clk, &dc->clk_mgr->clks, offsetof(struct dc_clocks, prev_p_state_change_support)) != 0)
-		dc->optimized_required = true;
+	if (type == UPDATE_TYPE_FAST) {
+		// If there's an available clock comparator, we use that.
+		if (dc->clk_mgr->funcs->are_clock_states_equal) {
+			if (!dc->clk_mgr->funcs->are_clock_states_equal(&dc->clk_mgr->clks, &dc->current_state->bw_ctx.bw.dcn.clk))
+				dc->optimized_required = true;
+		// Else we fallback to mem compare.
+		} else if (memcmp(&dc->current_state->bw_ctx.bw.dcn.clk, &dc->clk_mgr->clks, offsetof(struct dc_clocks, prev_p_state_change_support)) != 0) {
+			dc->optimized_required = true;
+		}
+	}
 
 	return type;
 }
@@ -1868,6 +1982,7 @@ static void commit_planes_do_stream_update(struct dc *dc,
 		struct dc_state *context)
 {
 	int j;
+	bool should_program_abm;
 
 	// Stream updates
 	for (j = 0; j < dc->res_pool->pipe_count; j++) {
@@ -1948,14 +2063,21 @@ static void commit_planes_do_stream_update(struct dc *dc,
 			}
 
 			if (stream_update->abm_level && pipe_ctx->stream_res.abm) {
-				if (pipe_ctx->stream_res.tg->funcs->is_blanked) {
-					// if otg funcs defined check if blanked before programming
-					if (!pipe_ctx->stream_res.tg->funcs->is_blanked(pipe_ctx->stream_res.tg))
+				should_program_abm = true;
+
+				// if otg funcs defined check if blanked before programming
+				if (pipe_ctx->stream_res.tg->funcs->is_blanked)
+					if (pipe_ctx->stream_res.tg->funcs->is_blanked(pipe_ctx->stream_res.tg))
+						should_program_abm = false;
+
+				if (should_program_abm) {
+					if (*stream_update->abm_level == ABM_LEVEL_IMMEDIATE_DISABLE) {
+						pipe_ctx->stream_res.abm->funcs->set_abm_immediate_disable(pipe_ctx->stream_res.abm);
+					} else {
 						pipe_ctx->stream_res.abm->funcs->set_abm_level(
 							pipe_ctx->stream_res.abm, stream->abm_level);
-				} else
-					pipe_ctx->stream_res.abm->funcs->set_abm_level(
-						pipe_ctx->stream_res.abm, stream->abm_level);
+					}
+				}
 			}
 		}
 	}
@@ -2000,7 +2122,13 @@ static void commit_planes_for_stream(struct dc *dc,
 		 * In case of turning off screen, no need to program front end a second time.
 		 * just return after program blank.
 		 */
-		dc->hwss.apply_ctx_for_surface(dc, stream, 0, context);
+		if (dc->hwss.apply_ctx_for_surface)
+			dc->hwss.apply_ctx_for_surface(dc, stream, 0, context);
+#if defined(CONFIG_DRM_AMD_DC_DCN2_0)
+		if (dc->hwss.program_front_end_for_ctx)
+			dc->hwss.program_front_end_for_ctx(dc, context);
+#endif
+
 		return;
 	}
 
@@ -2060,10 +2188,15 @@ static void commit_planes_for_stream(struct dc *dc,
 			stream_status =
 				stream_get_status(context, pipe_ctx->stream);
 
-			dc->hwss.apply_ctx_for_surface(
+			if (dc->hwss.apply_ctx_for_surface)
+				dc->hwss.apply_ctx_for_surface(
 					dc, pipe_ctx->stream, stream_status->plane_count, context);
 		}
 	}
+#if defined(CONFIG_DRM_AMD_DC_DCN2_0)
+	if (dc->hwss.program_front_end_for_ctx && update_type != UPDATE_TYPE_FAST)
+		dc->hwss.program_front_end_for_ctx(dc, context);
+#endif
 
 	// Update Type FAST, Surface updates
 	if (update_type == UPDATE_TYPE_FAST) {
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link.c b/drivers/gpu/drm/amd/display/dc/core/dc_link.c
index ca20b150afcc..12ba6fdf89b7 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_link.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_link.c
@@ -79,7 +79,6 @@ static void destruct(struct dc_link *link)
 	int i;
 
 	if (link->hpd_gpio != NULL) {
-		dal_gpio_close(link->hpd_gpio);
 		dal_gpio_destroy_irq(&link->hpd_gpio);
 		link->hpd_gpio = NULL;
 	}
@@ -520,7 +519,7 @@ static void link_disconnect_remap(struct dc_sink *prev_sink, struct dc_link *lin
 }
 
 
-static void read_edp_current_link_settings_on_detect(struct dc_link *link)
+static void read_current_link_settings_on_detect(struct dc_link *link)
 {
 	union lane_count_set lane_count_set = { {0} };
 	uint8_t link_bw_set;
@@ -555,17 +554,23 @@ static void read_edp_current_link_settings_on_detect(struct dc_link *link)
 			&link_bw_set, sizeof(link_bw_set));
 
 	if (link_bw_set == 0) {
-		/* If standard link rates are not being used,
-		 * Read DPCD 00115h to find the link rate set used
-		 */
-		core_link_read_dpcd(link, DP_LINK_RATE_SET,
-				&link_rate_set, sizeof(link_rate_set));
-
-		if (link_rate_set < link->dpcd_caps.edp_supported_link_rates_count) {
-			link->cur_link_settings.link_rate =
-				link->dpcd_caps.edp_supported_link_rates[link_rate_set];
-			link->cur_link_settings.link_rate_set = link_rate_set;
-			link->cur_link_settings.use_link_rate_set = true;
+		if (link->connector_signal == SIGNAL_TYPE_EDP) {
+			/* If standard link rates are not being used,
+			 * Read DPCD 00115h to find the edp link rate set used
+			 */
+			core_link_read_dpcd(link, DP_LINK_RATE_SET,
+					&link_rate_set, sizeof(link_rate_set));
+
+			// edp_supported_link_rates_count = 0 for DP
+			if (link_rate_set < link->dpcd_caps.edp_supported_link_rates_count) {
+				link->cur_link_settings.link_rate =
+						link->dpcd_caps.edp_supported_link_rates[link_rate_set];
+				link->cur_link_settings.link_rate_set = link_rate_set;
+				link->cur_link_settings.use_link_rate_set = true;
+			}
+		} else {
+			// Link Rate not found. Seamless boot may not work.
+			ASSERT(false);
 		}
 	} else {
 		link->cur_link_settings.link_rate = link_bw_set;
@@ -680,7 +685,7 @@ static bool is_same_edid(struct dc_edid *old_edid, struct dc_edid *new_edid)
 	return (memcmp(old_edid->raw_edid, new_edid->raw_edid, new_edid->length) == 0);
 }
 
-bool wait_for_alt_mode(struct dc_link *link)
+static bool wait_for_alt_mode(struct dc_link *link)
 {
 
 	/**
@@ -738,7 +743,8 @@ bool wait_for_alt_mode(struct dc_link *link)
  * This does not create remote sinks but will trigger DM
  * to start MST detection if a branch is detected.
  */
-bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason)
+static bool dc_link_detect_helper(struct dc_link *link,
+				  enum dc_detect_reason reason)
 {
 	struct dc_sink_init_data sink_init_data = { 0 };
 	struct display_sink_capability sink_caps = { 0 };
@@ -753,6 +759,8 @@ bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason)
 	struct dpcd_caps prev_dpcd_caps;
 	bool same_dpcd = true;
 	enum dc_connection_type new_connection_type = dc_connection_none;
+	bool perform_dp_seamless_boot = false;
+
 	DC_LOGGER_INIT(link->ctx->logger);
 
 	if (dc_is_virtual_signal(link->connector_signal))
@@ -809,15 +817,15 @@ bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason)
 		}
 
 		case SIGNAL_TYPE_EDP: {
-			read_edp_current_link_settings_on_detect(link);
+			read_current_link_settings_on_detect(link);
 			detect_edp_sink_caps(link);
-			sink_caps.transaction_type =
-				DDC_TRANSACTION_TYPE_I2C_OVER_AUX;
+			sink_caps.transaction_type = DDC_TRANSACTION_TYPE_I2C_OVER_AUX;
 			sink_caps.signal = SIGNAL_TYPE_EDP;
 			break;
 		}
 
 		case SIGNAL_TYPE_DISPLAY_PORT: {
+
 			/* wa HPD high coming too early*/
 			if (link->link_enc->features.flags.bits.DP_IS_USB_C == 1) {
 
@@ -865,12 +873,24 @@ bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason)
 				 * empty which leads to allocate_mst_payload() has "0"
 				 * pbn_per_slot value leading to exception on dc_fixpt_div()
 				 */
-				link->verified_link_cap = link->reported_link_cap;
+				dp_verify_mst_link_cap(link);
+
 				if (prev_sink != NULL)
 					dc_sink_release(prev_sink);
 				return false;
 			}
 
+			// For seamless boot, to skip verify link cap, we read UEFI settings and set them as verified.
+			if (reason == DETECT_REASON_BOOT &&
+					dc_ctx->dc->config.power_down_display_on_boot == false &&
+					link->link_status.link_active == true)
+				perform_dp_seamless_boot = true;
+
+			if (perform_dp_seamless_boot) {
+				read_current_link_settings_on_detect(link);
+				link->verified_link_cap = link->reported_link_cap;
+			}
+
 			break;
 		}
 
@@ -955,10 +975,11 @@ bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason)
 			 *  two link trainings
 			 */
 
-			/* deal with non-mst cases */
-			dp_verify_link_cap_with_retries(link,
-					&link->reported_link_cap,
-					LINK_TRAINING_MAX_VERIFY_RETRY);
+			// verify link cap for SST non-seamless boot
+			if (!perform_dp_seamless_boot)
+				dp_verify_link_cap_with_retries(link,
+						&link->reported_link_cap,
+						LINK_TRAINING_MAX_VERIFY_RETRY);
 		} else {
 			// If edid is the same, then discard new sink and revert back to original sink
 			if (same_edid) {
@@ -1047,6 +1068,23 @@ bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason)
 		dc_sink_release(prev_sink);
 
 	return true;
+
+}
+
+bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason)
+{
+	const struct dc *dc = link->dc;
+	bool ret;
+
+	/* get out of low power state */
+	clk_mgr_exit_optimized_pwr_state(dc, dc->clk_mgr);
+
+	ret = dc_link_detect_helper(link, reason);
+
+	/* Go back to power optimized state */
+	clk_mgr_optimize_pwr_state(dc, dc->clk_mgr);
+
+	return ret;
 }
 
 bool dc_link_get_hpd_state(struct dc_link *dc_link)
@@ -1492,7 +1530,7 @@ static enum dc_status enable_link_dp(
 
 	pipe_ctx->stream_res.pix_clk_params.requested_sym_clk =
 			link_settings.link_rate * LINK_RATE_REF_FREQ_IN_KHZ;
-	if (!apply_seamless_boot_optimization)
+	if (state->clk_mgr && !apply_seamless_boot_optimization)
 		state->clk_mgr->funcs->update_clocks(state->clk_mgr, state, false);
 
 	dp_enable_link_phy(
@@ -2169,8 +2207,10 @@ static void disable_link(struct dc_link *link, enum signal_type signal)
 			dp_set_fec_ready(link, false);
 		}
 #endif
-	} else
-		link->link_enc->funcs->disable_output(link->link_enc, signal);
+	} else {
+		if (signal != SIGNAL_TYPE_VIRTUAL)
+			link->link_enc->funcs->disable_output(link->link_enc, signal);
+	}
 
 	if (signal == SIGNAL_TYPE_DISPLAY_PORT_MST) {
 		/* MST disable link only when no stream use the link */
@@ -2217,7 +2257,7 @@ static bool dp_active_dongle_validate_timing(
 		break;
 	}
 
-	if (dongle_caps->dongle_type != DISPLAY_DONGLE_DP_HDMI_CONVERTER ||
+	if (dpcd_caps->dongle_type != DISPLAY_DONGLE_DP_HDMI_CONVERTER ||
 		dongle_caps->extendedCapValid == false)
 		return true;
 
@@ -2381,17 +2421,206 @@ bool dc_link_set_abm_disable(const struct dc_link *link)
 	return true;
 }
 
-bool dc_link_set_psr_enable(const struct dc_link *link, bool enable, bool wait)
+bool dc_link_set_psr_allow_active(struct dc_link *link, bool allow_active, bool wait)
 {
 	struct dc  *core_dc = link->ctx->dc;
 	struct dmcu *dmcu = core_dc->res_pool->dmcu;
 
-	if ((dmcu != NULL && dmcu->funcs->is_dmcu_initialized(dmcu)) && link->psr_enabled)
-		dmcu->funcs->set_psr_enable(dmcu, enable, wait);
+
+
+	if ((dmcu != NULL && dmcu->funcs->is_dmcu_initialized(dmcu)) && link->psr_feature_enabled)
+		dmcu->funcs->set_psr_enable(dmcu, allow_active, wait);
+
+	link->psr_allow_active = allow_active;
 
 	return true;
 }
 
+bool dc_link_get_psr_state(const struct dc_link *link, uint32_t *psr_state)
+{
+	struct dc  *core_dc = link->ctx->dc;
+	struct dmcu *dmcu = core_dc->res_pool->dmcu;
+
+	if (dmcu != NULL && link->psr_feature_enabled)
+		dmcu->funcs->get_psr_state(dmcu, psr_state);
+
+	return true;
+}
+
+static inline enum physical_phy_id
+transmitter_to_phy_id(enum transmitter transmitter_value)
+{
+	switch (transmitter_value) {
+	case TRANSMITTER_UNIPHY_A:
+		return PHYLD_0;
+	case TRANSMITTER_UNIPHY_B:
+		return PHYLD_1;
+	case TRANSMITTER_UNIPHY_C:
+		return PHYLD_2;
+	case TRANSMITTER_UNIPHY_D:
+		return PHYLD_3;
+	case TRANSMITTER_UNIPHY_E:
+		return PHYLD_4;
+	case TRANSMITTER_UNIPHY_F:
+		return PHYLD_5;
+	case TRANSMITTER_NUTMEG_CRT:
+		return PHYLD_6;
+	case TRANSMITTER_TRAVIS_CRT:
+		return PHYLD_7;
+	case TRANSMITTER_TRAVIS_LCD:
+		return PHYLD_8;
+	case TRANSMITTER_UNIPHY_G:
+		return PHYLD_9;
+	case TRANSMITTER_COUNT:
+		return PHYLD_COUNT;
+	case TRANSMITTER_UNKNOWN:
+		return PHYLD_UNKNOWN;
+	default:
+		WARN_ONCE(1, "Unknown transmitter value %d\n",
+			  transmitter_value);
+		return PHYLD_UNKNOWN;
+	}
+}
+
+bool dc_link_setup_psr(struct dc_link *link,
+		const struct dc_stream_state *stream, struct psr_config *psr_config,
+		struct psr_context *psr_context)
+{
+	struct dc *core_dc;
+	struct dmcu *dmcu;
+	int i;
+	/* updateSinkPsrDpcdConfig*/
+	union dpcd_psr_configuration psr_configuration;
+
+	psr_context->controllerId = CONTROLLER_ID_UNDEFINED;
+
+	if (!link)
+		return false;
+
+	core_dc = link->ctx->dc;
+	dmcu = core_dc->res_pool->dmcu;
+
+	if (!dmcu)
+		return false;
+
+
+	memset(&psr_configuration, 0, sizeof(psr_configuration));
+
+	psr_configuration.bits.ENABLE                    = 1;
+	psr_configuration.bits.CRC_VERIFICATION          = 1;
+	psr_configuration.bits.FRAME_CAPTURE_INDICATION  =
+			psr_config->psr_frame_capture_indication_req;
+
+	/* Check for PSR v2*/
+	if (psr_config->psr_version == 0x2) {
+		/* For PSR v2 selective update.
+		 * Indicates whether sink should start capturing
+		 * immediately following active scan line,
+		 * or starting with the 2nd active scan line.
+		 */
+		psr_configuration.bits.LINE_CAPTURE_INDICATION = 0;
+		/*For PSR v2, determines whether Sink should generate
+		 * IRQ_HPD when CRC mismatch is detected.
+		 */
+		psr_configuration.bits.IRQ_HPD_WITH_CRC_ERROR    = 1;
+	}
+
+	dm_helpers_dp_write_dpcd(
+		link->ctx,
+		link,
+		368,
+		&psr_configuration.raw,
+		sizeof(psr_configuration.raw));
+
+	psr_context->channel = link->ddc->ddc_pin->hw_info.ddc_channel;
+	psr_context->transmitterId = link->link_enc->transmitter;
+	psr_context->engineId = link->link_enc->preferred_engine;
+
+	for (i = 0; i < MAX_PIPES; i++) {
+		if (core_dc->current_state->res_ctx.pipe_ctx[i].stream
+				== stream) {
+			/* dmcu -1 for all controller id values,
+			 * therefore +1 here
+			 */
+			psr_context->controllerId =
+				core_dc->current_state->res_ctx.
+				pipe_ctx[i].stream_res.tg->inst + 1;
+			break;
+		}
+	}
+
+	/* Hardcoded for now.  Can be Pcie or Uniphy (or Unknown)*/
+	psr_context->phyType = PHY_TYPE_UNIPHY;
+	/*PhyId is associated with the transmitter id*/
+	psr_context->smuPhyId =
+		transmitter_to_phy_id(link->link_enc->transmitter);
+
+	psr_context->crtcTimingVerticalTotal = stream->timing.v_total;
+	psr_context->vsyncRateHz = div64_u64(div64_u64((stream->
+					timing.pix_clk_100hz * 100),
+					stream->timing.v_total),
+					stream->timing.h_total);
+
+	psr_context->psrSupportedDisplayConfig = true;
+	psr_context->psrExitLinkTrainingRequired =
+		psr_config->psr_exit_link_training_required;
+	psr_context->sdpTransmitLineNumDeadline =
+		psr_config->psr_sdp_transmit_line_num_deadline;
+	psr_context->psrFrameCaptureIndicationReq =
+		psr_config->psr_frame_capture_indication_req;
+
+	psr_context->skipPsrWaitForPllLock = 0; /* only = 1 in KV */
+
+	psr_context->numberOfControllers =
+			link->dc->res_pool->timing_generator_count;
+
+	psr_context->rfb_update_auto_en = true;
+
+	/* 2 frames before enter PSR. */
+	psr_context->timehyst_frames = 2;
+	/* half a frame
+	 * (units in 100 lines, i.e. a value of 1 represents 100 lines)
+	 */
+	psr_context->hyst_lines = stream->timing.v_total / 2 / 100;
+	psr_context->aux_repeats = 10;
+
+	psr_context->psr_level.u32all = 0;
+
+#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+	/*skip power down the single pipe since it blocks the cstate*/
+	if (ASICREV_IS_RAVEN(link->ctx->asic_id.hw_internal_rev))
+		psr_context->psr_level.bits.SKIP_CRTC_DISABLE = true;
+#endif
+
+	/* SMU will perform additional powerdown sequence.
+	 * For unsupported ASICs, set psr_level flag to skip PSR
+	 *  static screen notification to SMU.
+	 *  (Always set for DAL2, did not check ASIC)
+	 */
+	psr_context->allow_smu_optimizations = psr_config->allow_smu_optimizations;
+
+	/* Complete PSR entry before aborting to prevent intermittent
+	 * freezes on certain eDPs
+	 */
+	psr_context->psr_level.bits.DISABLE_PSR_ENTRY_ABORT = 1;
+
+	/* Controls additional delay after remote frame capture before
+	 * continuing power down, default = 0
+	 */
+	psr_context->frame_delay = 0;
+
+	link->psr_feature_enabled = dmcu->funcs->setup_psr(dmcu, link, psr_context);
+
+	/* psr_enabled == 0 indicates setup_psr did not succeed, but this
+	 * should not happen since firmware should be running at this point
+	 */
+	if (link->psr_feature_enabled == 0)
+		ASSERT(0);
+
+	return true;
+
+}
+
 const struct dc_link_status *dc_link_get_status(const struct dc_link *link)
 {
 	return &link->link_status;
@@ -2510,7 +2739,7 @@ static void update_mst_stream_alloc_table(
 /* convert link_mst_stream_alloc_table to dm dp_mst_stream_alloc_table
  * because stream_encoder is not exposed to dm
  */
-static enum dc_status allocate_mst_payload(struct pipe_ctx *pipe_ctx)
+enum dc_status dc_link_allocate_mst_payload(struct pipe_ctx *pipe_ctx)
 {
 	struct dc_stream_state *stream = pipe_ctx->stream;
 	struct dc_link *link = stream->link;
@@ -2521,6 +2750,7 @@ static enum dc_status allocate_mst_payload(struct pipe_ctx *pipe_ctx)
 	struct fixed31_32 pbn;
 	struct fixed31_32 pbn_per_slot;
 	uint8_t i;
+	enum act_return_status ret;
 	DC_LOGGER_INIT(link->ctx->logger);
 
 	/* enable_link_dp_mst already check link->enabled_stream_count
@@ -2568,14 +2798,16 @@ static enum dc_status allocate_mst_payload(struct pipe_ctx *pipe_ctx)
 		&link->mst_stream_alloc_table);
 
 	/* send down message */
-	dm_helpers_dp_mst_poll_for_allocation_change_trigger(
+	ret = dm_helpers_dp_mst_poll_for_allocation_change_trigger(
 			stream->ctx,
 			stream);
 
-	dm_helpers_dp_mst_send_payload_allocation(
-			stream->ctx,
-			stream,
-			true);
+	if (ret != ACT_LINK_LOST) {
+		dm_helpers_dp_mst_send_payload_allocation(
+				stream->ctx,
+				stream,
+				true);
+	}
 
 	/* slot X.Y for only current stream */
 	pbn_per_slot = get_pbn_per_slot(stream);
@@ -2667,6 +2899,24 @@ static enum dc_status deallocate_mst_payload(struct pipe_ctx *pipe_ctx)
 
 	return DC_OK;
 }
+#if defined(CONFIG_DRM_AMD_DC_HDCP)
+static void update_psp_stream_config(struct pipe_ctx *pipe_ctx, bool dpms_off)
+{
+	struct cp_psp *cp_psp = &pipe_ctx->stream->ctx->cp_psp;
+	if (cp_psp && cp_psp->funcs.update_stream_config) {
+		struct cp_psp_stream_config config;
+
+		memset(&config, 0, sizeof(config));
+
+		config.otg_inst = (uint8_t) pipe_ctx->stream_res.tg->inst;
+		config.stream_enc_inst = (uint8_t) pipe_ctx->stream_res.stream_enc->id;
+		config.link_enc_inst = pipe_ctx->stream->link->link_enc_hw_inst;
+		config.dpms_off = dpms_off;
+		config.dm_stream_ctx = pipe_ctx->stream->dm_stream_context;
+		cp_psp->funcs.update_stream_config(cp_psp->handle, &config);
+	}
+}
+#endif
 
 void core_link_enable_stream(
 		struct dc_state *state,
@@ -2677,6 +2927,10 @@ void core_link_enable_stream(
 	enum dc_status status;
 	DC_LOGGER_INIT(pipe_ctx->stream->ctx->logger);
 
+	if (!IS_FPGA_MAXIMUS_DC(core_dc->ctx->dce_environment) &&
+			dc_is_virtual_signal(pipe_ctx->stream->signal))
+		return;
+
 	if (!dc_is_virtual_signal(pipe_ctx->stream->signal)) {
 		stream->link->link_enc->funcs->setup(
 			stream->link->link_enc,
@@ -2727,6 +2981,9 @@ void core_link_enable_stream(
 		/* Do not touch link on seamless boot optimization. */
 		if (pipe_ctx->stream->apply_seamless_boot_optimization) {
 			pipe_ctx->stream->dpms_off = false;
+#if defined(CONFIG_DRM_AMD_DC_HDCP)
+			update_psp_stream_config(pipe_ctx, false);
+#endif
 			return;
 		}
 
@@ -2734,6 +2991,9 @@ void core_link_enable_stream(
 		if (pipe_ctx->stream->signal == SIGNAL_TYPE_EDP &&
 					apply_edp_fast_boot_optimization) {
 			pipe_ctx->stream->dpms_off = false;
+#if defined(CONFIG_DRM_AMD_DC_HDCP)
+			update_psp_stream_config(pipe_ctx, false);
+#endif
 			return;
 		}
 
@@ -2786,13 +3046,16 @@ void core_link_enable_stream(
 #endif
 
 		if (pipe_ctx->stream->signal == SIGNAL_TYPE_DISPLAY_PORT_MST)
-			allocate_mst_payload(pipe_ctx);
+			dc_link_allocate_mst_payload(pipe_ctx);
 
 		core_dc->hwss.unblank_stream(pipe_ctx,
 			&pipe_ctx->stream->link->cur_link_settings);
 
 		if (dc_is_dp_signal(pipe_ctx->stream->signal))
 			enable_stream_features(pipe_ctx);
+#if defined(CONFIG_DRM_AMD_DC_HDCP)
+		update_psp_stream_config(pipe_ctx, false);
+#endif
 	}
 #ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT
 	else { // if (IS_FPGA_MAXIMUS_DC(core_dc->ctx->dce_environment))
@@ -2810,6 +3073,14 @@ void core_link_disable_stream(struct pipe_ctx *pipe_ctx)
 	struct dc_stream_state *stream = pipe_ctx->stream;
 	struct dc_link *link = stream->sink->link;
 
+	if (!IS_FPGA_MAXIMUS_DC(core_dc->ctx->dce_environment) &&
+			dc_is_virtual_signal(pipe_ctx->stream->signal))
+		return;
+
+#if defined(CONFIG_DRM_AMD_DC_HDCP)
+	update_psp_stream_config(pipe_ctx, true);
+#endif
+
 	core_dc->hwss.blank_stream(pipe_ctx);
 
 	if (pipe_ctx->stream->signal == SIGNAL_TYPE_DISPLAY_PORT_MST)
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c
index 505967b48e14..7f904d55c1bc 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c
@@ -374,6 +374,7 @@ void dal_ddc_service_i2c_query_dp_dual_mode_adaptor(
 	enum display_dongle_type *dongle = &sink_cap->dongle_type;
 	uint8_t type2_dongle_buf[DP_ADAPTOR_TYPE2_SIZE];
 	bool is_type2_dongle = false;
+	int retry_count = 2;
 	struct dp_hdmi_dongle_signature_data *dongle_signature;
 
 	/* Assume we have no valid DP passive dongle connected */
@@ -386,13 +387,24 @@ void dal_ddc_service_i2c_query_dp_dual_mode_adaptor(
 		DP_HDMI_DONGLE_ADDRESS,
 		type2_dongle_buf,
 		sizeof(type2_dongle_buf))) {
-		*dongle = DISPLAY_DONGLE_DP_DVI_DONGLE;
-		sink_cap->max_hdmi_pixel_clock = DP_ADAPTOR_DVI_MAX_TMDS_CLK;
+		/* Passive HDMI dongles can sometimes fail here without retrying*/
+		while (retry_count > 0) {
+			if (i2c_read(ddc,
+				DP_HDMI_DONGLE_ADDRESS,
+				type2_dongle_buf,
+				sizeof(type2_dongle_buf)))
+				break;
+			retry_count--;
+		}
+		if (retry_count == 0) {
+			*dongle = DISPLAY_DONGLE_DP_DVI_DONGLE;
+			sink_cap->max_hdmi_pixel_clock = DP_ADAPTOR_DVI_MAX_TMDS_CLK;
 
-		CONN_DATA_DETECT(ddc->link, type2_dongle_buf, sizeof(type2_dongle_buf),
-				"DP-DVI passive dongle %dMhz: ",
-				DP_ADAPTOR_DVI_MAX_TMDS_CLK / 1000);
-		return;
+			CONN_DATA_DETECT(ddc->link, type2_dongle_buf, sizeof(type2_dongle_buf),
+					"DP-DVI passive dongle %dMhz: ",
+					DP_ADAPTOR_DVI_MAX_TMDS_CLK / 1000);
+			return;
+		}
 	}
 
 	/* Check if Type 2 dongle.*/
@@ -496,7 +508,7 @@ bool dal_ddc_service_query_ddc_data(
 	uint8_t *read_buf,
 	uint32_t read_size)
 {
-	bool ret;
+	bool ret = false;
 	uint32_t payload_size =
 		dal_ddc_service_is_in_aux_transaction_mode(ddc) ?
 			DEFAULT_AUX_MAX_DATA_SIZE : EDID_SEGMENT_SIZE;
@@ -515,34 +527,32 @@ bool dal_ddc_service_query_ddc_data(
 	/*TODO: len of payload data for i2c and aux is uint8!!!!,
 	 *  but we want to read 256 over i2c!!!!*/
 	if (dal_ddc_service_is_in_aux_transaction_mode(ddc)) {
-		struct aux_payload write_payload = {
-			.i2c_over_aux = true,
-			.write = true,
-			.mot = true,
-			.address = address,
-			.length = write_size,
-			.data = write_buf,
-			.reply = NULL,
-			.defer_delay = get_defer_delay(ddc),
-		};
-
-		struct aux_payload read_payload = {
-			.i2c_over_aux = true,
-			.write = false,
-			.mot = false,
-			.address = address,
-			.length = read_size,
-			.data = read_buf,
-			.reply = NULL,
-			.defer_delay = get_defer_delay(ddc),
-		};
-
-		ret = dc_link_aux_transfer_with_retries(ddc, &write_payload);
+		struct aux_payload payload;
+		bool read_available = true;
+
+		payload.i2c_over_aux = true;
+		payload.address = address;
+		payload.reply = NULL;
+		payload.defer_delay = get_defer_delay(ddc);
+
+		if (write_size != 0) {
+			payload.write = true;
+			payload.mot = false;
+			payload.length = write_size;
+			payload.data = write_buf;
+
+			ret = dal_ddc_submit_aux_command(ddc, &payload);
+			read_available = ret;
+		}
 
-		if (!ret)
-			return false;
+		if (read_size != 0 && read_available) {
+			payload.write = false;
+			payload.mot = false;
+			payload.length = read_size;
+			payload.data = read_buf;
 
-		ret = dc_link_aux_transfer_with_retries(ddc, &read_payload);
+			ret = dal_ddc_submit_aux_command(ddc, &payload);
+		}
 	} else {
 		struct i2c_payloads *payloads =
 			dal_ddc_i2c_payloads_create(ddc->ctx, payloads_num);
@@ -573,6 +583,41 @@ bool dal_ddc_service_query_ddc_data(
 	return ret;
 }
 
+bool dal_ddc_submit_aux_command(struct ddc_service *ddc,
+		struct aux_payload *payload)
+{
+	uint8_t retrieved = 0;
+	bool ret = 0;
+
+	if (!ddc)
+		return false;
+
+	if (!payload)
+		return false;
+
+	do {
+		struct aux_payload current_payload;
+		bool is_end_of_payload = (retrieved + DEFAULT_AUX_MAX_DATA_SIZE) >
+			payload->length ? true : false;
+
+		current_payload.address = payload->address;
+		current_payload.data = &payload->data[retrieved];
+		current_payload.defer_delay = payload->defer_delay;
+		current_payload.i2c_over_aux = payload->i2c_over_aux;
+		current_payload.length = is_end_of_payload ?
+			payload->length - retrieved : DEFAULT_AUX_MAX_DATA_SIZE;
+		current_payload.mot = !is_end_of_payload;
+		current_payload.reply = payload->reply;
+		current_payload.write = payload->write;
+
+		ret = dc_link_aux_transfer_with_retries(ddc, &current_payload);
+
+		retrieved += current_payload.length;
+	} while (retrieved < payload->length && ret == true);
+
+	return ret;
+}
+
 /* dc_link_aux_transfer_raw() - Attempt to transfer
  * the given aux payload.  This function does not perform
  * retries or handle error states.  The reply is returned
@@ -601,6 +646,20 @@ bool dc_link_aux_transfer_with_retries(struct ddc_service *ddc,
 	return dce_aux_transfer_with_retries(ddc, payload);
 }
 
+
+enum dc_status dc_link_aux_configure_timeout(struct ddc_service *ddc,
+		uint32_t timeout)
+{
+	enum dc_status status = DC_OK;
+	struct ddc *ddc_pin = ddc->ddc_pin;
+
+	if (ddc->ctx->dc->res_pool->engines[ddc_pin->pin_data->en]->funcs->configure_timeout == NULL)
+		return DC_ERROR_UNEXPECTED;
+	if (!ddc->ctx->dc->res_pool->engines[ddc_pin->pin_data->en]->funcs->configure_timeout(ddc, timeout))
+		status = DC_ERROR_UNEXPECTED;
+	return status;
+}
+
 /*test only function*/
 void dal_ddc_service_set_ddc_pin(
 	struct ddc_service *ddc_service,
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c
index f5742719b5d9..0f59b68aa4c2 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c
@@ -1409,6 +1409,9 @@ static struct dc_link_settings get_max_link_cap(struct dc_link *link)
 	if (link->link_enc->features.flags.bits.IS_HBR3_CAPABLE)
 		max_link_cap.link_rate = LINK_RATE_HIGH3;
 
+	if (link->link_enc->funcs->get_max_link_cap)
+		link->link_enc->funcs->get_max_link_cap(link->link_enc, &max_link_cap);
+
 	/* Lower link settings based on sink's link cap */
 	if (link->reported_link_cap.lane_count < max_link_cap.lane_count)
 		max_link_cap.lane_count =
@@ -1653,11 +1656,14 @@ bool dp_verify_link_cap_with_retries(
 
 	for (i = 0; i < attempts; i++) {
 		int fail_count = 0;
-		enum dc_connection_type type;
+		enum dc_connection_type type = dc_connection_none;
 
 		memset(&link->verified_link_cap, 0,
 				sizeof(struct dc_link_settings));
-		if (!dc_link_detect_sink(link, &type)) {
+		if (!dc_link_detect_sink(link, &type) || type == dc_connection_none) {
+			link->verified_link_cap.lane_count = LANE_COUNT_ONE;
+			link->verified_link_cap.link_rate = LINK_RATE_LOW;
+			link->verified_link_cap.link_spread = LINK_SPREAD_DISABLED;
 			break;
 		} else if (dp_verify_link_cap(link,
 				&link->reported_link_cap,
@@ -1670,6 +1676,19 @@ bool dp_verify_link_cap_with_retries(
 	return success;
 }
 
+bool dp_verify_mst_link_cap(
+	struct dc_link *link)
+{
+	struct dc_link_settings max_link_cap = {0};
+
+	max_link_cap = get_max_link_cap(link);
+	link->verified_link_cap = get_common_supported_link_settings(
+		link->reported_link_cap,
+		max_link_cap);
+
+	return true;
+}
+
 static struct dc_link_settings get_common_supported_link_settings(
 		struct dc_link_settings link_setting_a,
 		struct dc_link_settings link_setting_b)
@@ -2057,11 +2076,11 @@ static bool allow_hpd_rx_irq(const struct dc_link *link)
 	return false;
 }
 
-static bool handle_hpd_irq_psr_sink(const struct dc_link *link)
+static bool handle_hpd_irq_psr_sink(struct dc_link *link)
 {
 	union dpcd_psr_configuration psr_configuration;
 
-	if (!link->psr_enabled)
+	if (!link->psr_feature_enabled)
 		return false;
 
 	dm_helpers_dp_read_dpcd(
@@ -2100,8 +2119,8 @@ static bool handle_hpd_irq_psr_sink(const struct dc_link *link)
 				sizeof(psr_error_status.raw));
 
 			/* PSR error, disable and re-enable PSR */
-			dc_link_set_psr_enable(link, false, true);
-			dc_link_set_psr_enable(link, true, true);
+			dc_link_set_psr_allow_active(link, false, true);
+			dc_link_set_psr_allow_active(link, true, true);
 
 			return true;
 		} else if (psr_sink_psr_status.bits.SINK_SELF_REFRESH_STATUS ==
@@ -2364,6 +2383,8 @@ bool dc_link_handle_hpd_rx_irq(struct dc_link *link, union hpd_irq_data *out_hpd
 	enum dc_status result;
 
 	bool status = false;
+	struct pipe_ctx *pipe_ctx;
+	int i;
 
 	if (out_link_loss)
 		*out_link_loss = false;
@@ -2440,6 +2461,15 @@ bool dc_link_handle_hpd_rx_irq(struct dc_link *link, union hpd_irq_data *out_hpd
 			&link->cur_link_settings,
 			true, LINK_TRAINING_ATTEMPTS);
 
+		for (i = 0; i < MAX_PIPES; i++) {
+			pipe_ctx = &link->dc->current_state->res_ctx.pipe_ctx[i];
+			if (pipe_ctx && pipe_ctx->stream && pipe_ctx->stream->link == link &&
+					pipe_ctx->stream->dpms_off == false &&
+					pipe_ctx->stream->signal == SIGNAL_TYPE_DISPLAY_PORT_MST) {
+				dc_link_allocate_mst_payload(pipe_ctx);
+			}
+		}
+
 		status = false;
 		if (out_link_loss)
 			*out_link_loss = true;
@@ -2545,6 +2575,7 @@ static void get_active_converter_info(
 	uint8_t data, struct dc_link *link)
 {
 	union dp_downstream_port_present ds_port = { .byte = data };
+	memset(&link->dpcd_caps.dongle_caps, 0, sizeof(link->dpcd_caps.dongle_caps));
 
 	/* decode converter info*/
 	if (!ds_port.fields.PORT_PRESENT) {
@@ -2691,6 +2722,7 @@ static void dp_wa_power_up_0010FA(struct dc_link *link, uint8_t *dpcd_data,
 		 * keep receiver powered all the time.*/
 		case DP_BRANCH_DEVICE_ID_0010FA:
 		case DP_BRANCH_DEVICE_ID_0080E1:
+		case DP_BRANCH_DEVICE_ID_00E04C:
 			link->wa_flags.dp_keep_receiver_powered = true;
 			break;
 
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_hwss.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_hwss.c
index 79438c4f1e20..a519dbc5ecb6 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_link_hwss.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_hwss.c
@@ -277,7 +277,8 @@ void dp_retrain_link_dp_test(struct dc_link *link,
 		if (pipes[i].stream != NULL &&
 			!pipes[i].top_pipe && !pipes[i].prev_odm_pipe &&
 			pipes[i].stream->link != NULL &&
-			pipes[i].stream_res.stream_enc != NULL) {
+			pipes[i].stream_res.stream_enc != NULL &&
+			pipes[i].stream->link == link) {
 			udelay(100);
 
 			pipes[i].stream_res.stream_enc->funcs->dp_blank(
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
index 8f70295179ff..37698305a2dc 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
@@ -404,6 +404,9 @@ bool resource_are_streams_timing_synchronizable(
 	if (stream1->view_format != stream2->view_format)
 		return false;
 
+	if (stream1->ignore_msa_timing_param || stream2->ignore_msa_timing_param)
+		return false;
+
 	return true;
 }
 static bool is_dp_and_hdmi_sharable(
@@ -948,7 +951,7 @@ static void calculate_inits_and_adj_vp(struct pipe_ctx *pipe_ctx)
 	data->inits.v_c_bot = dc_fixpt_add(data->inits.v_c, data->ratios.vert_c);
 
 }
-static bool are_rect_integer_multiples(struct rect src, struct rect dest)
+static bool are_rects_integer_multiples(struct rect src, struct rect dest)
 {
 	if (dest.width  >= src.width  && dest.width  % src.width  == 0 &&
 		dest.height >= src.height && dest.height % src.height == 0)
@@ -956,6 +959,38 @@ static bool are_rect_integer_multiples(struct rect src, struct rect dest)
 
 	return false;
 }
+
+static void calculate_integer_scaling(struct pipe_ctx *pipe_ctx)
+{
+	if (!pipe_ctx->plane_state->scaling_quality.integer_scaling)
+		return;
+
+	//for Centered Mode
+	if (pipe_ctx->stream->dst.width  == pipe_ctx->stream->src.width &&
+		pipe_ctx->stream->dst.height == pipe_ctx->stream->src.height) {
+		// calculate maximum # of replication of src onto addressable
+		unsigned int integer_multiple = min(
+				pipe_ctx->stream->timing.h_addressable / pipe_ctx->stream->src.width,
+				pipe_ctx->stream->timing.v_addressable  / pipe_ctx->stream->src.height);
+
+		//scale dst
+		pipe_ctx->stream->dst.width  = integer_multiple * pipe_ctx->stream->src.width;
+		pipe_ctx->stream->dst.height = integer_multiple * pipe_ctx->stream->src.height;
+
+		//center dst onto addressable
+		pipe_ctx->stream->dst.x = (pipe_ctx->stream->timing.h_addressable - pipe_ctx->stream->dst.width)/2;
+		pipe_ctx->stream->dst.y = (pipe_ctx->stream->timing.v_addressable - pipe_ctx->stream->dst.height)/2;
+	}
+
+	//disable taps if src & dst are integer ratio
+	if (are_rects_integer_multiples(pipe_ctx->stream->src, pipe_ctx->stream->dst)) {
+		pipe_ctx->plane_state->scaling_quality.v_taps = 1;
+		pipe_ctx->plane_state->scaling_quality.h_taps = 1;
+		pipe_ctx->plane_state->scaling_quality.v_taps_c = 1;
+		pipe_ctx->plane_state->scaling_quality.h_taps_c = 1;
+	}
+}
+
 bool resource_build_scaling_params(struct pipe_ctx *pipe_ctx)
 {
 	const struct dc_plane_state *plane_state = pipe_ctx->plane_state;
@@ -969,6 +1004,8 @@ bool resource_build_scaling_params(struct pipe_ctx *pipe_ctx)
 	pipe_ctx->plane_res.scl_data.format = convert_pixel_format_to_dalsurface(
 			pipe_ctx->plane_state->format);
 
+	calculate_integer_scaling(pipe_ctx);
+
 	calculate_scaling_ratios(pipe_ctx);
 
 	calculate_viewport(pipe_ctx);
@@ -999,13 +1036,6 @@ bool resource_build_scaling_params(struct pipe_ctx *pipe_ctx)
 		res = pipe_ctx->plane_res.dpp->funcs->dpp_get_optimal_number_of_taps(
 				pipe_ctx->plane_res.dpp, &pipe_ctx->plane_res.scl_data, &plane_state->scaling_quality);
 
-	if (res &&
-	    plane_state->scaling_quality.integer_scaling &&
-	    are_rect_integer_multiples(pipe_ctx->plane_res.scl_data.viewport,
-				       pipe_ctx->plane_res.scl_data.recout)) {
-		pipe_ctx->plane_res.scl_data.taps.v_taps = 1;
-		pipe_ctx->plane_res.scl_data.taps.h_taps = 1;
-	}
 
 	if (!res) {
 		/* Try 24 bpp linebuffer */
@@ -1540,6 +1570,9 @@ bool dc_is_stream_unchanged(
 	if (!are_stream_backends_same(old_stream, stream))
 		return false;
 
+	if (old_stream->ignore_msa_timing_param != stream->ignore_msa_timing_param)
+		return false;
+
 	return true;
 }
 
@@ -1629,7 +1662,8 @@ static int acquire_first_free_pipe(
 static struct audio *find_first_free_audio(
 		struct resource_context *res_ctx,
 		const struct resource_pool *pool,
-		enum engine_id id)
+		enum engine_id id,
+		enum dce_version dc_version)
 {
 	int i, available_audio_count;
 
@@ -1848,28 +1882,28 @@ static int acquire_resource_from_hw_enabled_state(
 		struct dc_stream_state *stream)
 {
 	struct dc_link *link = stream->link;
-	unsigned int inst, tg_inst;
+	unsigned int i, inst, tg_inst = 0;
 
 	/* Check for enabled DIG to identify enabled display */
 	if (!link->link_enc->funcs->is_dig_enabled(link->link_enc))
 		return -1;
 
-	/* Check for which front end is used by this encoder.
-	 * Note the inst is 1 indexed, where 0 is undefined.
-	 * Note that DIG_FE can source from different OTG but our
-	 * current implementation always map 1-to-1, so this code makes
-	 * the same assumption and doesn't check OTG source.
-	 */
 	inst = link->link_enc->funcs->get_dig_frontend(link->link_enc);
 
-	/* Instance should be within the range of the pool */
-	if (inst >= pool->pipe_count)
-		return -1;
+	if (inst == ENGINE_ID_UNKNOWN)
+		return false;
 
-	if (inst >= pool->stream_enc_count)
-		return -1;
+	for (i = 0; i < pool->stream_enc_count; i++) {
+		if (pool->stream_enc[i]->id == inst) {
+			tg_inst = pool->stream_enc[i]->funcs->dig_source_otg(
+				pool->stream_enc[i]);
+			break;
+		}
+	}
 
-	tg_inst = pool->stream_enc[inst]->funcs->dig_source_otg(pool->stream_enc[inst]);
+	// tg_inst not found
+	if (i == pool->stream_enc_count)
+		return false;
 
 	if (tg_inst >= pool->timing_generator_count)
 		return false;
@@ -1965,7 +1999,7 @@ enum dc_status resource_map_pool_resources(
 	    dc_is_audio_capable_signal(pipe_ctx->stream->signal) &&
 	    stream->audio_info.mode_count && stream->audio_info.flags.all) {
 		pipe_ctx->stream_res.audio = find_first_free_audio(
-		&context->res_ctx, pool, pipe_ctx->stream_res.stream_enc->id);
+		&context->res_ctx, pool, pipe_ctx->stream_res.stream_enc->id, dc_ctx->dce_version);
 
 		/*
 		 * Audio assigned in order first come first get.
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_stream.c b/drivers/gpu/drm/amd/display/dc/core/dc_stream.c
index bf1d7bb90e0f..bb09243758fe 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_stream.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_stream.c
@@ -423,10 +423,10 @@ bool dc_stream_add_writeback(struct dc *dc,
 
 		if (dwb->funcs->is_enabled(dwb)) {
 			/* writeback pipe already enabled, only need to update */
-			dc->hwss.update_writeback(dc, stream_status, wb_info);
+			dc->hwss.update_writeback(dc, stream_status, wb_info, dc->current_state);
 		} else {
 			/* Enable writeback pipe from scratch*/
-			dc->hwss.enable_writeback(dc, stream_status, wb_info);
+			dc->hwss.enable_writeback(dc, stream_status, wb_info, dc->current_state);
 		}
 	}
 
diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h
index a82352a87808..0416a17b0897 100644
--- a/drivers/gpu/drm/amd/display/dc/dc.h
+++ b/drivers/gpu/drm/amd/display/dc/dc.h
@@ -39,7 +39,7 @@
 #include "inc/hw/dmcu.h"
 #include "dml/display_mode_lib.h"
 
-#define DC_VER "3.2.48"
+#define DC_VER "3.2.56"
 
 #define MAX_SURFACES 3
 #define MAX_PLANES 6
@@ -111,19 +111,20 @@ struct dc_caps {
 	bool force_dp_tps4_for_cp2520;
 	bool disable_dp_clk_share;
 	bool psp_setup_panel_mode;
+	bool extended_aux_timeout_support;
 #ifdef CONFIG_DRM_AMD_DC_DCN2_0
 	bool hw_3d_lut;
 #endif
 	struct dc_plane_cap planes[MAX_PLANES];
 };
 
-#if defined(CONFIG_DRM_AMD_DC_DCN2_0)
 struct dc_bug_wa {
+#if defined(CONFIG_DRM_AMD_DC_DCN2_0)
 	bool no_connect_phy_config;
 	bool dedcn20_305_wa;
+#endif
 	bool skip_clock_update;
 };
-#endif
 
 struct dc_dcc_surface_param {
 	struct dc_size surface_size;
@@ -219,7 +220,9 @@ struct dc_config {
 	bool allow_seamless_boot_optimization;
 	bool power_down_display_on_boot;
 	bool edp_not_connected;
+	bool force_enum_edp;
 	bool forced_clocks;
+	bool disable_extended_timeout_support; // Used to disable extended timeout and lttpr feature as well
 	bool multi_mon_pp_mclk_switch;
 };
 
@@ -227,6 +230,7 @@ enum visual_confirm {
 	VISUAL_CONFIRM_DISABLE = 0,
 	VISUAL_CONFIRM_SURFACE = 1,
 	VISUAL_CONFIRM_HDR = 2,
+	VISUAL_CONFIRM_MPCTREE = 4,
 };
 
 enum dcc_option {
@@ -245,6 +249,19 @@ enum wm_report_mode {
 	WM_REPORT_DEFAULT = 0,
 	WM_REPORT_OVERRIDE = 1,
 };
+enum dtm_pstate{
+	dtm_level_p0 = 0,/*highest voltage*/
+	dtm_level_p1,
+	dtm_level_p2,
+	dtm_level_p3,
+	dtm_level_p4,/*when active_display_count = 0*/
+};
+
+enum dcn_pwr_state {
+	DCN_PWR_STATE_UNKNOWN = -1,
+	DCN_PWR_STATE_MISSION_MODE = 0,
+	DCN_PWR_STATE_LOW_POWER = 3,
+};
 
 /*
  * For any clocks that may differ per pipe
@@ -252,11 +269,7 @@ enum wm_report_mode {
  */
 struct dc_clocks {
 	int dispclk_khz;
-	int max_supported_dppclk_khz;
-	int max_supported_dispclk_khz;
 	int dppclk_khz;
-	int bw_dppclk_khz; /*a copy of dppclk_khz*/
-	int bw_dispclk_khz;
 	int dcfclk_khz;
 	int socclk_khz;
 	int dcfclk_deep_sleep_khz;
@@ -264,12 +277,17 @@ struct dc_clocks {
 	int phyclk_khz;
 	int dramclk_khz;
 	bool p_state_change_support;
-
+	enum dcn_pwr_state pwr_state;
 	/*
 	 * Elements below are not compared for the purposes of
 	 * optimization required
 	 */
 	bool prev_p_state_change_support;
+	enum dtm_pstate dtm_level;
+	int max_supported_dppclk_khz;
+	int max_supported_dispclk_khz;
+	int bw_dppclk_khz; /*a copy of dppclk_khz*/
+	int bw_dispclk_khz;
 };
 
 struct dc_bw_validation_profile {
@@ -347,6 +365,7 @@ struct dc_debug_options {
 	bool disable_hubp_power_gate;
 #ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT
 	bool disable_dsc_power_gate;
+	int dsc_min_slice_height_override;
 #endif
 	bool disable_pplib_wm_range;
 	enum wm_report_mode pplib_wm_report_mode;
@@ -462,9 +481,7 @@ struct dc {
 	struct dc_config config;
 	struct dc_debug_options debug;
 	struct dc_bounding_box_overrides bb_overrides;
-#if defined(CONFIG_DRM_AMD_DC_DCN2_0)
 	struct dc_bug_wa work_arounds;
-#endif
 	struct dc_context *ctx;
 #ifdef CONFIG_DRM_AMD_DC_DCN2_0
 	struct dc_phy_addr_space_config vm_pa_config;
@@ -553,10 +570,16 @@ struct dc_init_data {
 };
 
 struct dc_callback_init {
+#ifdef CONFIG_DRM_AMD_DC_HDCP
+	struct cp_psp cp_psp;
+#else
 	uint8_t reserved;
+#endif
 };
 
 struct dc *dc_create(const struct dc_init_data *init_params);
+void dc_hardware_init(struct dc *dc);
+
 int dc_get_vmid_use_vector(struct dc *dc);
 #ifdef CONFIG_DRM_AMD_DC_DCN2_0
 void dc_setup_vm_context(struct dc *dc, struct dc_virtual_addr_space_config *va_config, int vmid);
@@ -565,6 +588,7 @@ int dc_setup_system_context(struct dc *dc, struct dc_phy_addr_space_config *pa_c
 #endif
 void dc_init_callbacks(struct dc *dc,
 		const struct dc_callback_init *init_params);
+void dc_deinit_callbacks(struct dc *dc);
 void dc_destroy(struct dc **dc);
 
 /*******************************************************************************
diff --git a/drivers/gpu/drm/amd/display/dc/dc_ddc_types.h b/drivers/gpu/drm/amd/display/dc/dc_ddc_types.h
index 4ef97f65e55d..4f8f576d5fcf 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_ddc_types.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_ddc_types.h
@@ -49,7 +49,8 @@ enum aux_channel_operation_result {
 	AUX_CHANNEL_OPERATION_FAILED_REASON_UNKNOWN,
 	AUX_CHANNEL_OPERATION_FAILED_INVALID_REPLY,
 	AUX_CHANNEL_OPERATION_FAILED_TIMEOUT,
-	AUX_CHANNEL_OPERATION_FAILED_HPD_DISCON
+	AUX_CHANNEL_OPERATION_FAILED_HPD_DISCON,
+	AUX_CHANNEL_OPERATION_FAILED_ENGINE_ACQUIRE
 };
 
 
diff --git a/drivers/gpu/drm/amd/display/dc/dc_dsc.h b/drivers/gpu/drm/amd/display/dc/dc_dsc.h
index 6e42209f0e20..0ed2962add5a 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_dsc.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_dsc.h
@@ -30,6 +30,7 @@
 #define DP_DSC_BRANCH_OVERALL_THROUGHPUT_0  0x0a0   /* DP 1.4a SCR */
 #define DP_DSC_BRANCH_OVERALL_THROUGHPUT_1  0x0a1
 #define DP_DSC_BRANCH_MAX_LINE_WIDTH        0x0a2
+#include "dc_types.h"
 
 struct dc_dsc_bw_range {
 	uint32_t min_kbps; /* Bandwidth if min_target_bpp_x16 is used */
@@ -39,13 +40,21 @@ struct dc_dsc_bw_range {
 	uint32_t stream_kbps; /* Uncompressed stream bandwidth */
 };
 
+struct display_stream_compressor {
+	const struct dsc_funcs *funcs;
+#ifndef AMD_EDID_UTILITY
+	struct dc_context *ctx;
+	int inst;
+#endif
+};
 
 bool dc_dsc_parse_dsc_dpcd(const uint8_t *dpcd_dsc_basic_data,
 		const uint8_t *dpcd_dsc_ext_data,
 		struct dsc_dec_dpcd_caps *dsc_sink_caps);
 
 bool dc_dsc_compute_bandwidth_range(
-		const struct dc *dc,
+		const struct display_stream_compressor *dsc,
+		const uint32_t dsc_min_slice_height_override,
 		const uint32_t min_kbps,
 		const uint32_t max_kbps,
 		const struct dsc_dec_dpcd_caps *dsc_sink_caps,
@@ -53,8 +62,9 @@ bool dc_dsc_compute_bandwidth_range(
 		struct dc_dsc_bw_range *range);
 
 bool dc_dsc_compute_config(
-		const struct dc *dc,
+		const struct display_stream_compressor *dsc,
 		const struct dsc_dec_dpcd_caps *dsc_sink_caps,
+		const uint32_t dsc_min_slice_height_override,
 		uint32_t target_bandwidth_kbps,
 		const struct dc_crtc_timing *timing,
 		struct dc_dsc_config *dsc_cfg);
diff --git a/drivers/gpu/drm/amd/display/dc/dc_hw_types.h b/drivers/gpu/drm/amd/display/dc/dc_hw_types.h
index 0b8700a8a94a..e0856bb8511f 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_hw_types.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_hw_types.h
@@ -26,6 +26,8 @@
 #ifndef DC_HW_TYPES_H
 #define DC_HW_TYPES_H
 
+#ifndef AMD_EDID_UTILITY
+
 #include "os_types.h"
 #include "fixed31_32.h"
 #include "signal_types.h"
@@ -124,20 +126,6 @@ struct plane_size {
 	int chroma_pitch;
 	struct rect surface_size;
 	struct rect chroma_size;
-
-	union {
-		struct {
-			struct rect surface_size;
-			int surface_pitch;
-		} grph;
-
-		struct {
-			struct rect luma_size;
-			int luma_pitch;
-			struct rect chroma_size;
-			int chroma_pitch;
-		} video;
-	};
 };
 
 struct dc_plane_dcc_param {
@@ -148,21 +136,6 @@ struct dc_plane_dcc_param {
 
 	int meta_pitch_c;
 	bool independent_64b_blks_c;
-
-	union {
-		struct {
-			int meta_pitch;
-			bool independent_64b_blks;
-		} grph;
-
-		struct {
-			int meta_pitch_l;
-			bool independent_64b_blks_l;
-
-			int meta_pitch_c;
-			bool independent_64b_blks_c;
-		} video;
-	};
 };
 
 /*Displayable pixel format in fb*/
@@ -605,6 +578,11 @@ enum dc_quantization_range {
 	QUANTIZATION_RANGE_LIMITED
 };
 
+enum dc_dynamic_expansion {
+	DYN_EXPANSION_AUTO,
+	DYN_EXPANSION_DISABLE
+};
+
 /* XFM */
 
 /* used in  struct dc_plane_state */
@@ -616,6 +594,8 @@ struct scaling_taps {
 	bool integer_scaling;
 };
 
+#endif /* AMD_EDID_UTILITY */
+
 enum dc_timing_standard {
 	DC_TIMING_STANDARD_UNDEFINED,
 	DC_TIMING_STANDARD_DMT,
@@ -737,30 +717,6 @@ enum dc_timing_3d_format {
 	TIMING_3D_FORMAT_MAX,
 };
 
-enum trigger_delay {
-	TRIGGER_DELAY_NEXT_PIXEL = 0,
-	TRIGGER_DELAY_NEXT_LINE,
-};
-
-enum crtc_event {
-	CRTC_EVENT_VSYNC_RISING = 0,
-	CRTC_EVENT_VSYNC_FALLING
-};
-
-struct crtc_trigger_info {
-	bool enabled;
-	struct dc_stream_state *event_source;
-	enum crtc_event event;
-	enum trigger_delay delay;
-};
-
-struct dc_crtc_timing_adjust {
-	uint32_t v_total_min;
-	uint32_t v_total_max;
-	uint32_t v_total_mid;
-	uint32_t v_total_mid_frame_num;
-};
-
 #ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT
 struct dc_dsc_config {
 	uint32_t num_slices_h; /* Number of DSC slices - horizontal */
@@ -804,6 +760,33 @@ struct dc_crtc_timing {
 #endif
 };
 
+#ifndef AMD_EDID_UTILITY
+
+enum trigger_delay {
+	TRIGGER_DELAY_NEXT_PIXEL = 0,
+	TRIGGER_DELAY_NEXT_LINE,
+};
+
+enum crtc_event {
+	CRTC_EVENT_VSYNC_RISING = 0,
+	CRTC_EVENT_VSYNC_FALLING
+};
+
+struct crtc_trigger_info {
+	bool enabled;
+	struct dc_stream_state *event_source;
+	enum crtc_event event;
+	enum trigger_delay delay;
+};
+
+struct dc_crtc_timing_adjust {
+	uint32_t v_total_min;
+	uint32_t v_total_max;
+	uint32_t v_total_mid;
+	uint32_t v_total_mid_frame_num;
+};
+
+
 /* Passed on init */
 enum vram_type {
 	VIDEO_MEMORY_TYPE_GDDR5  = 2,
@@ -874,5 +857,7 @@ struct tg_color {
 	uint16_t color_b_cb;
 };
 
+#endif /* AMD_EDID_UTILITY */
+
 #endif /* DC_HW_TYPES_H */
 
diff --git a/drivers/gpu/drm/amd/display/dc/dc_link.h b/drivers/gpu/drm/amd/display/dc/dc_link.h
index 9ea75db3484e..f24fd19ed93d 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_link.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_link.h
@@ -126,7 +126,8 @@ struct dc_link {
 	unsigned short chip_caps;
 	unsigned int dpcd_sink_count;
 	enum edp_revision edp_revision;
-	bool psr_enabled;
+	bool psr_feature_enabled;
+	bool psr_allow_active;
 
 	/* MST record stream using this link */
 	struct link_flags {
@@ -158,6 +159,18 @@ static inline struct dc_link *dc_get_link_at_index(struct dc *dc, uint32_t link_
 	return dc->links[link_index];
 }
 
+static inline struct dc_link *get_edp_link(const struct dc *dc)
+{
+	int i;
+
+	// report any eDP links, even unconnected DDI's
+	for (i = 0; i < dc->link_count; i++) {
+		if (dc->links[i]->connector_signal == SIGNAL_TYPE_EDP)
+			return dc->links[i];
+	}
+	return NULL;
+}
+
 /* Set backlight level of an embedded panel (eDP, LVDS).
  * backlight_pwm_u16_16 is unsigned 32 bit with 16 bit integer
  * and 16 bit fractional, where 1.0 is max backlight value.
@@ -170,7 +183,7 @@ int dc_link_get_backlight_level(const struct dc_link *dc_link);
 
 bool dc_link_set_abm_disable(const struct dc_link *dc_link);
 
-bool dc_link_set_psr_enable(const struct dc_link *dc_link, bool enable, bool wait);
+bool dc_link_set_psr_allow_active(struct dc_link *dc_link, bool enable, bool wait);
 
 bool dc_link_get_psr_state(const struct dc_link *dc_link, uint32_t *psr_state);
 
@@ -192,6 +205,7 @@ enum dc_detect_reason {
 
 bool dc_link_detect(struct dc_link *dc_link, enum dc_detect_reason reason);
 bool dc_link_get_hpd_state(struct dc_link *dc_link);
+enum dc_status dc_link_allocate_mst_payload(struct pipe_ctx *pipe_ctx);
 
 /* Notify DC about DP RX Interrupt (aka Short Pulse Interrupt).
  * Return:
diff --git a/drivers/gpu/drm/amd/display/dc/dc_stream.h b/drivers/gpu/drm/amd/display/dc/dc_stream.h
index 0fa1c26bc20d..fdb6adc37857 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_stream.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_stream.h
@@ -113,6 +113,21 @@ struct periodic_interrupt_config {
 	int lines_offset;
 };
 
+union stream_update_flags {
+	struct {
+		uint32_t scaling:1;
+		uint32_t out_tf:1;
+		uint32_t out_csc:1;
+		uint32_t abm_level:1;
+		uint32_t dpms_off:1;
+		uint32_t gamut_remap:1;
+#if defined(CONFIG_DRM_AMD_DC_DCN2_0)
+		uint32_t wb_update:1;
+#endif
+	} bits;
+
+	uint32_t raw;
+};
 
 struct dc_stream_state {
 	// sink is deprecated, new code should not reference
@@ -214,9 +229,14 @@ struct dc_stream_state {
 #ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT
 	bool is_dsc_enabled;
 #endif
+	union stream_update_flags update_flags;
 };
 
+#define ABM_LEVEL_IMMEDIATE_DISABLE 0xFFFFFFFF
+
 struct dc_stream_update {
+	struct dc_stream_state *stream;
+
 	struct rect src;
 	struct rect dst;
 	struct dc_transfer_func *out_transfer_func;
@@ -431,6 +451,9 @@ void dc_stream_set_static_screen_events(struct dc *dc,
 					int num_streams,
 					const struct dc_static_screen_events *events);
 
+void dc_stream_set_dyn_expansion(struct dc *dc, struct dc_stream_state *stream,
+		enum dc_dynamic_expansion option);
+
 void dc_stream_set_dither_option(struct dc_stream_state *stream,
 				 enum dc_dither_option option);
 
diff --git a/drivers/gpu/drm/amd/display/dc/dc_types.h b/drivers/gpu/drm/amd/display/dc/dc_types.h
index b273735b6a3e..d9be8fc3889f 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_types.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_types.h
@@ -25,6 +25,11 @@
 #ifndef DC_TYPES_H_
 #define DC_TYPES_H_
 
+#ifndef AMD_EDID_UTILITY
+/* AND EdidUtility only needs a portion
+ * of this file, including the rest only
+ * causes additional issues.
+ */
 #include "os_types.h"
 #include "fixed31_32.h"
 #include "irq_types.h"
@@ -33,6 +38,10 @@
 #include "dal_types.h"
 #include "grph_object_defs.h"
 
+#ifdef CONFIG_DRM_AMD_DC_HDCP
+#include "dm_cp_psp.h"
+#endif
+
 /* forward declarations */
 struct dc_plane_state;
 struct dc_stream_state;
@@ -100,6 +109,9 @@ struct dc_context {
 	uint32_t dc_sink_id_count;
 	uint32_t dc_stream_id_count;
 	uint64_t fbc_gpu_addr;
+#ifdef CONFIG_DRM_AMD_DC_HDCP
+	struct cp_psp cp_psp;
+#endif
 };
 
 
@@ -159,6 +171,12 @@ enum dc_edid_status {
 	EDID_THE_SAME,
 };
 
+enum act_return_status {
+	ACT_SUCCESS,
+	ACT_LINK_LOST,
+	ACT_FAILED
+};
+
 /* audio capability from EDID*/
 struct dc_cea_audio_mode {
 	uint8_t format_code; /* ucData[0] [6:3]*/
@@ -739,6 +757,9 @@ struct dc_clock_config {
 	uint32_t current_clock_khz;/*current clock in use*/
 };
 
+#endif /*AMD_EDID_UTILITY*/
+//AMD EDID UTILITY does not need any of the above structures
+
 #ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT
 /* DSC DPCD capabilities */
 union dsc_slice_caps1 {
@@ -810,4 +831,5 @@ struct dsc_dec_dpcd_caps {
 	uint32_t branch_max_line_width;
 };
 #endif
+
 #endif /* DC_TYPES_H_ */
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_abm.c b/drivers/gpu/drm/amd/display/dc/dce/dce_abm.c
index 58bd131d5b48..b8a3fc505c9b 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_abm.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_abm.c
@@ -77,6 +77,9 @@ static bool dce_abm_set_pipe(struct abm *abm, uint32_t controller_id)
 	/* notifyDMCUMsg */
 	REG_UPDATE(MASTER_COMM_CNTL_REG, MASTER_COMM_INTERRUPT, 1);
 
+	REG_WAIT(MASTER_COMM_CNTL_REG, MASTER_COMM_INTERRUPT, 0,
+			1, 80000);
+
 	return true;
 }
 
@@ -401,6 +404,10 @@ static bool dce_abm_init_backlight(struct abm *abm)
 	/* Enable the backlight output */
 	REG_UPDATE(BL_PWM_CNTL, BL_PWM_EN, 1);
 
+	/* Disable fractional pwm if configured */
+	REG_UPDATE(BL_PWM_CNTL, BL_PWM_FRACTIONAL_EN,
+		   abm->ctx->dc->config.disable_fractional_pwm ? 0 : 1);
+
 	/* Unlock group 2 backlight registers */
 	REG_UPDATE(BL_PWM_GRP1_REG_LOCK,
 			BL_PWM_GRP1_REG_LOCK, 0);
@@ -489,9 +496,6 @@ void dce_abm_destroy(struct abm **abm)
 {
 	struct dce_abm *abm_dce = TO_DCE_ABM(*abm);
 
-	if (abm_dce->base.dmcu_is_running == true)
-		abm_dce->base.funcs->set_abm_immediate_disable(*abm);
-
 	kfree(abm_dce);
 	*abm = NULL;
 }
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_aux.c b/drivers/gpu/drm/amd/display/dc/dce/dce_aux.c
index c3f9f4185ce8..e472608faf33 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_aux.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_aux.c
@@ -42,6 +42,10 @@
 
 #include "reg_helper.h"
 
+#undef FN
+#define FN(reg_name, field_name) \
+	aux110->shift->field_name, aux110->mask->field_name
+
 #define FROM_AUX_ENGINE(ptr) \
 	container_of((ptr), struct aux_engine_dce110, base)
 
@@ -55,6 +59,14 @@ enum {
 	AUX_TIMED_OUT_RETRY_COUNTER = 2,
 	AUX_DEFER_RETRY_COUNTER = 6
 };
+
+#define TIME_OUT_INCREMENT      1016
+#define TIME_OUT_MULTIPLIER_8 	8
+#define TIME_OUT_MULTIPLIER_16  16
+#define TIME_OUT_MULTIPLIER_32  32
+#define TIME_OUT_MULTIPLIER_64  64
+#define MAX_TIMEOUT_LENGTH      127
+
 static void release_engine(
 	struct dce_aux *engine)
 {
@@ -198,7 +210,7 @@ static void submit_channel_request(
 	REG_UPDATE(AUX_INTERRUPT_CONTROL, AUX_SW_DONE_ACK, 1);
 
 	REG_WAIT(AUX_SW_STATUS, AUX_SW_DONE, 0,
-				10, aux110->timeout_period/10);
+				10, aux110->polling_timeout_period/10);
 
 	/* set the delay and the number of bytes to write */
 
@@ -327,7 +339,7 @@ static enum aux_channel_operation_result get_channel_status(
 
 	/* poll to make sure that SW_DONE is asserted */
 	REG_WAIT(AUX_SW_STATUS, AUX_SW_DONE, 1,
-				10, aux110->timeout_period/10);
+				10, aux110->polling_timeout_period/10);
 
 	value = REG_READ(AUX_SW_STATUS);
 	/* in case HPD is LOW, exit AUX transaction */
@@ -414,20 +426,77 @@ void dce110_engine_destroy(struct dce_aux **engine)
 	*engine = NULL;
 
 }
+
+static bool dce_aux_configure_timeout(struct ddc_service *ddc,
+		uint32_t timeout_in_us)
+{
+	uint32_t multiplier = 0;
+	uint32_t length = 0;
+	struct ddc *ddc_pin = ddc->ddc_pin;
+	struct dce_aux *aux_engine = ddc->ctx->dc->res_pool->engines[ddc_pin->pin_data->en];
+	struct aux_engine_dce110 *aux110 = FROM_AUX_ENGINE(aux_engine);
+
+	/* 1-Update polling timeout period */
+	aux110->polling_timeout_period = timeout_in_us * SW_AUX_TIMEOUT_PERIOD_MULTIPLIER;
+
+	/* 2-Update aux timeout period length and multiplier */
+	if (timeout_in_us <= TIME_OUT_INCREMENT) {
+		multiplier = 0;
+		length = timeout_in_us/TIME_OUT_MULTIPLIER_8;
+		if (timeout_in_us % TIME_OUT_MULTIPLIER_8 != 0)
+			length++;
+	} else if (timeout_in_us <= 2 * TIME_OUT_INCREMENT) {
+		multiplier = 1;
+		length = timeout_in_us/TIME_OUT_MULTIPLIER_16;
+		if (timeout_in_us % TIME_OUT_MULTIPLIER_16 != 0)
+			length++;
+	} else if (timeout_in_us <= 4 * TIME_OUT_INCREMENT) {
+		multiplier = 2;
+		length = timeout_in_us/TIME_OUT_MULTIPLIER_32;
+		if (timeout_in_us % TIME_OUT_MULTIPLIER_32 != 0)
+			length++;
+	} else if (timeout_in_us > 4 * TIME_OUT_INCREMENT) {
+		multiplier = 3;
+		length = timeout_in_us/TIME_OUT_MULTIPLIER_64;
+		if (timeout_in_us % TIME_OUT_MULTIPLIER_64 != 0)
+			length++;
+	}
+
+	length = (length < MAX_TIMEOUT_LENGTH) ? length : MAX_TIMEOUT_LENGTH;
+
+	REG_UPDATE_SEQ_2(AUX_DPHY_RX_CONTROL1, AUX_RX_TIMEOUT_LEN, length, AUX_RX_TIMEOUT_LEN_MUL, multiplier);
+
+	return true;
+}
+
+static struct dce_aux_funcs aux_functions = {
+	.configure_timeout = NULL,
+	.destroy = NULL,
+};
+
 struct dce_aux *dce110_aux_engine_construct(struct aux_engine_dce110 *aux_engine110,
 		struct dc_context *ctx,
 		uint32_t inst,
 		uint32_t timeout_period,
-		const struct dce110_aux_registers *regs)
+		const struct dce110_aux_registers *regs,
+		const struct dce110_aux_registers_mask *mask,
+		const struct dce110_aux_registers_shift *shift,
+		bool is_ext_aux_timeout_configurable)
 {
 	aux_engine110->base.ddc = NULL;
 	aux_engine110->base.ctx = ctx;
 	aux_engine110->base.delay = 0;
 	aux_engine110->base.max_defer_write_retry = 0;
 	aux_engine110->base.inst = inst;
-	aux_engine110->timeout_period = timeout_period;
+	aux_engine110->polling_timeout_period = timeout_period;
 	aux_engine110->regs = regs;
 
+	aux_engine110->mask = mask;
+	aux_engine110->shift = shift;
+	aux_engine110->base.funcs = &aux_functions;
+	if (is_ext_aux_timeout_configurable)
+		aux_engine110->base.funcs->configure_timeout = &dce_aux_configure_timeout;
+
 	return &aux_engine110->base;
 }
 
@@ -464,8 +533,10 @@ int dce_aux_transfer_raw(struct ddc_service *ddc,
 	memset(&aux_rep, 0, sizeof(aux_rep));
 
 	aux_engine = ddc->ctx->dc->res_pool->engines[ddc_pin->pin_data->en];
-	if (!acquire(aux_engine, ddc_pin))
+	if (!acquire(aux_engine, ddc_pin)) {
+		*operation_result = AUX_CHANNEL_OPERATION_FAILED_ENGINE_ACQUIRE;
 		return -1;
+	}
 
 	if (payload->i2c_over_aux)
 		aux_req.type = AUX_TRANSACTION_TYPE_I2C;
@@ -475,7 +546,7 @@ int dce_aux_transfer_raw(struct ddc_service *ddc,
 	aux_req.action = i2caux_action_from_payload(payload);
 
 	aux_req.address = payload->address;
-	aux_req.delay = payload->defer_delay * 10;
+	aux_req.delay = 0;
 	aux_req.length = payload->length;
 	aux_req.data = payload->data;
 
@@ -544,8 +615,15 @@ bool dce_aux_transfer_with_retries(struct ddc_service *ddc,
 			case AUX_TRANSACTION_REPLY_AUX_DEFER:
 			case AUX_TRANSACTION_REPLY_I2C_OVER_AUX_NACK:
 			case AUX_TRANSACTION_REPLY_I2C_OVER_AUX_DEFER:
-				if (++aux_defer_retries >= AUX_MAX_DEFER_RETRIES)
+				if (++aux_defer_retries >= AUX_MAX_DEFER_RETRIES) {
 					goto fail;
+				} else {
+					if ((*payload->reply == AUX_TRANSACTION_REPLY_AUX_DEFER) ||
+						(*payload->reply == AUX_TRANSACTION_REPLY_I2C_OVER_AUX_DEFER)) {
+						if (payload->defer_delay > 0)
+							msleep(payload->defer_delay);
+					}
+				}
 				break;
 
 			case AUX_TRANSACTION_REPLY_I2C_DEFER:
@@ -582,6 +660,7 @@ bool dce_aux_transfer_with_retries(struct ddc_service *ddc,
 			break;
 
 		case AUX_CHANNEL_OPERATION_FAILED_HPD_DISCON:
+		case AUX_CHANNEL_OPERATION_FAILED_ENGINE_ACQUIRE:
 		case AUX_CHANNEL_OPERATION_FAILED_REASON_UNKNOWN:
 		default:
 			goto fail;
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_aux.h b/drivers/gpu/drm/amd/display/dc/dce/dce_aux.h
index ed7fec8fe253..b4b2c79a8073 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_aux.h
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_aux.h
@@ -29,6 +29,7 @@
 #include "i2caux_interface.h"
 #include "inc/hw/aux_engine.h"
 
+
 #ifdef CONFIG_DRM_AMD_DC_DCN2_0
 #define AUX_COMMON_REG_LIST0(id)\
 	SRI(AUX_CONTROL, DP_AUX, id), \
@@ -36,6 +37,7 @@
 	SRI(AUX_SW_DATA, DP_AUX, id), \
 	SRI(AUX_SW_CONTROL, DP_AUX, id), \
 	SRI(AUX_INTERRUPT_CONTROL, DP_AUX, id), \
+	SRI(AUX_DPHY_RX_CONTROL1, DP_AUX, id), \
 	SRI(AUX_SW_STATUS, DP_AUX, id)
 #endif
 
@@ -55,6 +57,7 @@ struct dce110_aux_registers {
 	uint32_t AUX_SW_DATA;
 	uint32_t AUX_SW_CONTROL;
 	uint32_t AUX_INTERRUPT_CONTROL;
+	uint32_t AUX_DPHY_RX_CONTROL1;
 	uint32_t AUX_SW_STATUS;
 	uint32_t AUXN_IMPCAL;
 	uint32_t AUXP_IMPCAL;
@@ -62,6 +65,156 @@ struct dce110_aux_registers {
 	uint32_t AUX_RESET_MASK;
 };
 
+#define DCE_AUX_REG_FIELD_LIST(type)\
+	type AUX_EN;\
+	type AUX_RESET;\
+	type AUX_RESET_DONE;\
+	type AUX_REG_RW_CNTL_STATUS;\
+	type AUX_SW_USE_AUX_REG_REQ;\
+	type AUX_SW_DONE_USING_AUX_REG;\
+	type AUX_SW_AUTOINCREMENT_DISABLE;\
+	type AUX_SW_DATA_RW;\
+	type AUX_SW_INDEX;\
+	type AUX_SW_GO;\
+	type AUX_SW_DATA;\
+	type AUX_SW_REPLY_BYTE_COUNT;\
+	type AUX_SW_DONE;\
+	type AUX_SW_DONE_ACK;\
+	type AUXN_IMPCAL_ENABLE;\
+	type AUXP_IMPCAL_ENABLE;\
+	type AUXN_IMPCAL_OVERRIDE_ENABLE;\
+	type AUXP_IMPCAL_OVERRIDE_ENABLE;\
+	type AUX_RX_TIMEOUT_LEN;\
+	type AUX_RX_TIMEOUT_LEN_MUL;\
+	type AUXN_CALOUT_ERROR_AK;\
+	type AUXP_CALOUT_ERROR_AK;\
+	type AUX_SW_START_DELAY;\
+	type AUX_SW_WR_BYTES
+
+#define DCE10_AUX_MASK_SH_LIST(mask_sh)\
+	AUX_SF(AUX_CONTROL, AUX_EN, mask_sh),\
+	AUX_SF(AUX_ARB_CONTROL, AUX_REG_RW_CNTL_STATUS, mask_sh),\
+	AUX_SF(AUX_ARB_CONTROL, AUX_SW_USE_AUX_REG_REQ, mask_sh),\
+	AUX_SF(AUX_ARB_CONTROL, AUX_SW_DONE_USING_AUX_REG, mask_sh),\
+	AUX_SF(AUX_SW_CONTROL, AUX_SW_START_DELAY, mask_sh),\
+	AUX_SF(AUX_SW_CONTROL, AUX_SW_WR_BYTES, mask_sh),\
+	AUX_SF(AUX_SW_CONTROL, AUX_SW_GO, mask_sh),\
+	AUX_SF(AUX_SW_DATA, AUX_SW_AUTOINCREMENT_DISABLE, mask_sh),\
+	AUX_SF(AUX_SW_DATA, AUX_SW_DATA_RW, mask_sh),\
+	AUX_SF(AUX_SW_DATA, AUX_SW_AUTOINCREMENT_DISABLE, mask_sh),\
+	AUX_SF(AUX_SW_DATA, AUX_SW_INDEX, mask_sh),\
+	AUX_SF(AUX_SW_DATA, AUX_SW_DATA, mask_sh),\
+	AUX_SF(AUX_SW_STATUS, AUX_SW_REPLY_BYTE_COUNT, mask_sh),\
+	AUX_SF(AUX_SW_STATUS, AUX_SW_DONE, mask_sh),\
+	AUX_SF(AUX_INTERRUPT_CONTROL, AUX_SW_DONE_ACK, mask_sh),\
+	AUX_SF(AUXN_IMPCAL, AUXN_CALOUT_ERROR_AK, mask_sh),\
+	AUX_SF(AUXP_IMPCAL, AUXP_CALOUT_ERROR_AK, mask_sh),\
+	AUX_SF(AUXN_IMPCAL, AUXN_IMPCAL_ENABLE, mask_sh),\
+	AUX_SF(AUXP_IMPCAL, AUXP_IMPCAL_ENABLE, mask_sh),\
+	AUX_SF(AUXP_IMPCAL, AUXP_IMPCAL_OVERRIDE_ENABLE, mask_sh),\
+	AUX_SF(AUXN_IMPCAL, AUXN_IMPCAL_OVERRIDE_ENABLE, mask_sh)
+
+#define DCE_AUX_MASK_SH_LIST(mask_sh)\
+	AUX_SF(AUX_CONTROL, AUX_EN, mask_sh),\
+	AUX_SF(AUX_CONTROL, AUX_RESET, mask_sh),\
+	AUX_SF(AUX_CONTROL, AUX_RESET_DONE, mask_sh),\
+	AUX_SF(AUX_ARB_CONTROL, AUX_REG_RW_CNTL_STATUS, mask_sh),\
+	AUX_SF(AUX_ARB_CONTROL, AUX_SW_USE_AUX_REG_REQ, mask_sh),\
+	AUX_SF(AUX_ARB_CONTROL, AUX_SW_DONE_USING_AUX_REG, mask_sh),\
+	AUX_SF(AUX_SW_CONTROL, AUX_SW_START_DELAY, mask_sh),\
+	AUX_SF(AUX_SW_CONTROL, AUX_SW_WR_BYTES, mask_sh),\
+	AUX_SF(AUX_SW_CONTROL, AUX_SW_GO, mask_sh),\
+	AUX_SF(AUX_SW_DATA, AUX_SW_AUTOINCREMENT_DISABLE, mask_sh),\
+	AUX_SF(AUX_SW_DATA, AUX_SW_DATA_RW, mask_sh),\
+	AUX_SF(AUX_SW_DATA, AUX_SW_AUTOINCREMENT_DISABLE, mask_sh),\
+	AUX_SF(AUX_SW_DATA, AUX_SW_INDEX, mask_sh),\
+	AUX_SF(AUX_SW_DATA, AUX_SW_DATA, mask_sh),\
+	AUX_SF(AUX_SW_STATUS, AUX_SW_REPLY_BYTE_COUNT, mask_sh),\
+	AUX_SF(AUX_SW_STATUS, AUX_SW_DONE, mask_sh),\
+	AUX_SF(AUX_INTERRUPT_CONTROL, AUX_SW_DONE_ACK, mask_sh),\
+	AUX_SF(AUXN_IMPCAL, AUXN_CALOUT_ERROR_AK, mask_sh),\
+	AUX_SF(AUXP_IMPCAL, AUXP_CALOUT_ERROR_AK, mask_sh),\
+	AUX_SF(AUXN_IMPCAL, AUXN_IMPCAL_ENABLE, mask_sh),\
+	AUX_SF(AUXP_IMPCAL, AUXP_IMPCAL_ENABLE, mask_sh),\
+	AUX_SF(AUXP_IMPCAL, AUXP_IMPCAL_OVERRIDE_ENABLE, mask_sh),\
+	AUX_SF(AUXN_IMPCAL, AUXN_IMPCAL_OVERRIDE_ENABLE, mask_sh)
+
+#define DCE12_AUX_MASK_SH_LIST(mask_sh)\
+	AUX_SF(DP_AUX0_AUX_CONTROL, AUX_EN, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_CONTROL, AUX_RESET, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_CONTROL, AUX_RESET_DONE, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_ARB_CONTROL, AUX_REG_RW_CNTL_STATUS, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_ARB_CONTROL, AUX_SW_USE_AUX_REG_REQ, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_ARB_CONTROL, AUX_SW_DONE_USING_AUX_REG, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_SW_CONTROL, AUX_SW_START_DELAY, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_SW_CONTROL, AUX_SW_WR_BYTES, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_SW_CONTROL, AUX_SW_GO, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_SW_DATA, AUX_SW_AUTOINCREMENT_DISABLE, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_SW_DATA, AUX_SW_DATA_RW, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_SW_DATA, AUX_SW_AUTOINCREMENT_DISABLE, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_SW_DATA, AUX_SW_INDEX, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_SW_DATA, AUX_SW_DATA, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_SW_STATUS, AUX_SW_REPLY_BYTE_COUNT, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_SW_STATUS, AUX_SW_DONE, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_INTERRUPT_CONTROL, AUX_SW_DONE_ACK, mask_sh),\
+	AUX_SF(AUXN_IMPCAL, AUXN_CALOUT_ERROR_AK, mask_sh),\
+	AUX_SF(AUXP_IMPCAL, AUXP_CALOUT_ERROR_AK, mask_sh),\
+	AUX_SF(AUXN_IMPCAL, AUXN_IMPCAL_ENABLE, mask_sh),\
+	AUX_SF(AUXP_IMPCAL, AUXP_IMPCAL_ENABLE, mask_sh),\
+	AUX_SF(AUXP_IMPCAL, AUXP_IMPCAL_OVERRIDE_ENABLE, mask_sh),\
+	AUX_SF(AUXN_IMPCAL, AUXN_IMPCAL_OVERRIDE_ENABLE, mask_sh)
+
+/* DCN10 MASK */
+#define DCN10_AUX_MASK_SH_LIST(mask_sh)\
+	AUX_SF(DP_AUX0_AUX_CONTROL, AUX_EN, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_CONTROL, AUX_RESET, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_CONTROL, AUX_RESET_DONE, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_ARB_CONTROL, AUX_REG_RW_CNTL_STATUS, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_ARB_CONTROL, AUX_SW_USE_AUX_REG_REQ, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_ARB_CONTROL, AUX_SW_DONE_USING_AUX_REG, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_SW_CONTROL, AUX_SW_START_DELAY, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_SW_CONTROL, AUX_SW_WR_BYTES, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_SW_CONTROL, AUX_SW_GO, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_SW_DATA, AUX_SW_AUTOINCREMENT_DISABLE, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_SW_DATA, AUX_SW_DATA_RW, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_SW_DATA, AUX_SW_AUTOINCREMENT_DISABLE, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_SW_DATA, AUX_SW_INDEX, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_SW_DATA, AUX_SW_DATA, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_SW_STATUS, AUX_SW_REPLY_BYTE_COUNT, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_SW_STATUS, AUX_SW_DONE, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_INTERRUPT_CONTROL, AUX_SW_DONE_ACK, mask_sh),\
+	AUX_SF(AUXN_IMPCAL, AUXN_CALOUT_ERROR_AK, mask_sh),\
+	AUX_SF(AUXP_IMPCAL, AUXP_CALOUT_ERROR_AK, mask_sh),\
+	AUX_SF(AUXN_IMPCAL, AUXN_IMPCAL_ENABLE, mask_sh),\
+	AUX_SF(AUXP_IMPCAL, AUXP_IMPCAL_ENABLE, mask_sh),\
+	AUX_SF(AUXP_IMPCAL, AUXP_IMPCAL_OVERRIDE_ENABLE, mask_sh),\
+	AUX_SF(AUXN_IMPCAL, AUXN_IMPCAL_OVERRIDE_ENABLE, mask_sh)
+
+/* for all other DCN */
+#define DCN_AUX_MASK_SH_LIST(mask_sh)\
+	AUX_SF(DP_AUX0_AUX_CONTROL, AUX_EN, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_CONTROL, AUX_RESET, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_CONTROL, AUX_RESET_DONE, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_ARB_CONTROL, AUX_REG_RW_CNTL_STATUS, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_ARB_CONTROL, AUX_SW_USE_AUX_REG_REQ, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_ARB_CONTROL, AUX_SW_DONE_USING_AUX_REG, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_SW_CONTROL, AUX_SW_START_DELAY, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_SW_CONTROL, AUX_SW_WR_BYTES, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_SW_CONTROL, AUX_SW_GO, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_SW_DATA, AUX_SW_AUTOINCREMENT_DISABLE, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_SW_DATA, AUX_SW_DATA_RW, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_SW_DATA, AUX_SW_AUTOINCREMENT_DISABLE, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_SW_DATA, AUX_SW_INDEX, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_SW_DATA, AUX_SW_DATA, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_SW_STATUS, AUX_SW_REPLY_BYTE_COUNT, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_SW_STATUS, AUX_SW_DONE, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_INTERRUPT_CONTROL, AUX_SW_DONE_ACK, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_DPHY_RX_CONTROL1, AUX_RX_TIMEOUT_LEN, mask_sh),\
+	AUX_SF(DP_AUX0_AUX_DPHY_RX_CONTROL1, AUX_RX_TIMEOUT_LEN_MUL, mask_sh)
+
+#define AUX_SF(reg_name, field_name, post_fix)\
+	.field_name = reg_name ## __ ## field_name ## post_fix
+
 enum {	/* This is the timeout as defined in DP 1.2a,
 	 * 2.3.4 "Detailed uPacket TX AUX CH State Description".
 	 */
@@ -97,20 +250,34 @@ struct dce_aux {
 	uint32_t max_defer_write_retry;
 
 	bool acquire_reset;
+	struct dce_aux_funcs *funcs;
+};
+
+struct dce110_aux_registers_mask {
+	DCE_AUX_REG_FIELD_LIST(uint32_t);
+};
+
+struct dce110_aux_registers_shift {
+	DCE_AUX_REG_FIELD_LIST(uint8_t);
 };
 
+
 struct aux_engine_dce110 {
 	struct dce_aux base;
 	const struct dce110_aux_registers *regs;
+	const struct dce110_aux_registers_mask *mask;
+	const struct dce110_aux_registers_shift *shift;
 	struct {
 		uint32_t aux_control;
 		uint32_t aux_arb_control;
 		uint32_t aux_sw_data;
 		uint32_t aux_sw_control;
 		uint32_t aux_interrupt_control;
+		uint32_t aux_dphy_rx_control1;
+		uint32_t aux_dphy_rx_control0;
 		uint32_t aux_sw_status;
 	} addr;
-	uint32_t timeout_period;
+	uint32_t polling_timeout_period;
 };
 
 struct aux_engine_dce110_init_data {
@@ -120,12 +287,15 @@ struct aux_engine_dce110_init_data {
 	const struct dce110_aux_registers *regs;
 };
 
-struct dce_aux *dce110_aux_engine_construct(
-		struct aux_engine_dce110 *aux_engine110,
+struct dce_aux *dce110_aux_engine_construct(struct aux_engine_dce110 *aux_engine110,
 		struct dc_context *ctx,
 		uint32_t inst,
 		uint32_t timeout_period,
-		const struct dce110_aux_registers *regs);
+		const struct dce110_aux_registers *regs,
+
+		const struct dce110_aux_registers_mask *mask,
+		const struct dce110_aux_registers_shift *shift,
+		bool is_ext_aux_timeout_configurable);
 
 void dce110_engine_destroy(struct dce_aux **engine);
 
@@ -139,4 +309,13 @@ int dce_aux_transfer_raw(struct ddc_service *ddc,
 
 bool dce_aux_transfer_with_retries(struct ddc_service *ddc,
 		struct aux_payload *cmd);
+
+struct dce_aux_funcs {
+	bool (*configure_timeout)
+		(struct ddc_service *ddc,
+		 uint32_t timeout);
+	void (*destroy)
+		(struct aux_engine **ptr);
+};
+
 #endif
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_dmcu.c b/drivers/gpu/drm/amd/display/dc/dce/dce_dmcu.c
index 0b86cee4876f..ba995d3f2318 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_dmcu.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_dmcu.c
@@ -907,9 +907,6 @@ void dce_dmcu_destroy(struct dmcu **dmcu)
 {
 	struct dce_dmcu *dmcu_dce = TO_DCE_DMCU(*dmcu);
 
-	if (dmcu_dce->base.dmcu_state == DMCU_RUNNING)
-		dmcu_dce->base.funcs->set_psr_enable(*dmcu, false, true);
-
 	kfree(dmcu_dce);
 	*dmcu = NULL;
 }
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_hwseq.h b/drivers/gpu/drm/amd/display/dc/dce/dce_hwseq.h
index ac04d77058f0..32d145a0d6fc 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_hwseq.h
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_hwseq.h
@@ -679,6 +679,7 @@ struct dce_hwseq_registers {
 	HWS_SF(, DOMAIN17_PG_STATUS, DOMAIN17_PGFSM_PWR_STATUS, mask_sh), \
 	HWS_SF(, DOMAIN18_PG_STATUS, DOMAIN18_PGFSM_PWR_STATUS, mask_sh), \
 	HWS_SF(, DC_IP_REQUEST_CNTL, IP_REQUEST_EN, mask_sh), \
+	HWSEQ_LVTMA_MASK_SH_LIST(mask_sh), \
 	HWS_SF(, LVTMA_PWRSEQ_CNTL, LVTMA_BLON, mask_sh), \
 	HWS_SF(, LVTMA_PWRSEQ_STATE, LVTMA_PWRSEQ_TARGET_STATE_R, mask_sh)
 #endif
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.c b/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.c
index 31b698bf9cfc..8aa937f496c4 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.c
@@ -606,11 +606,11 @@ static void dce_mi_allocate_dmif(
 	}
 
 	if (dce_mi->wa.single_head_rdreq_dmif_limit) {
-		uint32_t eanble =  (total_stream_num > 1) ? 0 :
+		uint32_t enable =  (total_stream_num > 1) ? 0 :
 				dce_mi->wa.single_head_rdreq_dmif_limit;
 
 		REG_UPDATE(MC_HUB_RDREQ_DMIF_LIMIT,
-				ENABLE, eanble);
+				ENABLE, enable);
 	}
 }
 
@@ -636,11 +636,11 @@ static void dce_mi_free_dmif(
 			10, 3500);
 
 	if (dce_mi->wa.single_head_rdreq_dmif_limit) {
-		uint32_t eanble =  (total_stream_num > 1) ? 0 :
+		uint32_t enable =  (total_stream_num > 1) ? 0 :
 				dce_mi->wa.single_head_rdreq_dmif_limit;
 
 		REG_UPDATE(MC_HUB_RDREQ_DMIF_LIMIT,
-				ENABLE, eanble);
+				ENABLE, enable);
 	}
 }
 
diff --git a/drivers/gpu/drm/amd/display/dc/dce100/dce100_resource.c b/drivers/gpu/drm/amd/display/dc/dce100/dce100_resource.c
index 1787b9bf800a..a5e122c721ec 100644
--- a/drivers/gpu/drm/amd/display/dc/dce100/dce100_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dce100/dce100_resource.c
@@ -399,6 +399,37 @@ static const struct dc_plane_cap plane_cap = {
 #define CC_DC_HDMI_STRAPS__AUDIO_STREAM_NUMBER__SHIFT 0x8
 #endif
 
+static int map_transmitter_id_to_phy_instance(
+	enum transmitter transmitter)
+{
+	switch (transmitter) {
+	case TRANSMITTER_UNIPHY_A:
+		return 0;
+	break;
+	case TRANSMITTER_UNIPHY_B:
+		return 1;
+	break;
+	case TRANSMITTER_UNIPHY_C:
+		return 2;
+	break;
+	case TRANSMITTER_UNIPHY_D:
+		return 3;
+	break;
+	case TRANSMITTER_UNIPHY_E:
+		return 4;
+	break;
+	case TRANSMITTER_UNIPHY_F:
+		return 5;
+	break;
+	case TRANSMITTER_UNIPHY_G:
+		return 6;
+	break;
+	default:
+		ASSERT(0);
+		return 0;
+	}
+}
+
 static void read_dce_straps(
 	struct dc_context *ctx,
 	struct resource_straps *straps)
@@ -506,6 +537,14 @@ static const struct dce_mem_input_mask mi_masks = {
 		.ENABLE = MC_HUB_RDREQ_DMIF_LIMIT__ENABLE_MASK
 };
 
+static const struct dce110_aux_registers_shift aux_shift = {
+	DCE10_AUX_MASK_SH_LIST(__SHIFT)
+};
+
+static const struct dce110_aux_registers_mask aux_mask = {
+	DCE10_AUX_MASK_SH_LIST(_MASK)
+};
+
 static struct mem_input *dce100_mem_input_create(
 	struct dc_context *ctx,
 	uint32_t inst)
@@ -571,14 +610,18 @@ struct link_encoder *dce100_link_encoder_create(
 {
 	struct dce110_link_encoder *enc110 =
 		kzalloc(sizeof(struct dce110_link_encoder), GFP_KERNEL);
+	int link_regs_id;
 
 	if (!enc110)
 		return NULL;
 
+	link_regs_id =
+		map_transmitter_id_to_phy_instance(enc_init_data->transmitter);
+
 	dce110_link_encoder_construct(enc110,
 				      enc_init_data,
 				      &link_enc_feature,
-				      &link_enc_regs[enc_init_data->transmitter],
+				      &link_enc_regs[link_regs_id],
 				      &link_enc_aux_regs[enc_init_data->channel - 1],
 				      &link_enc_hpd_regs[enc_init_data->hpd_source]);
 	return &enc110->base;
@@ -611,7 +654,10 @@ struct dce_aux *dce100_aux_engine_create(
 
 	dce110_aux_engine_construct(aux_engine, ctx, inst,
 				    SW_AUX_TIMEOUT_PERIOD_MULTIPLIER * AUX_TIMEOUT_PERIOD,
-				    &aux_engine_regs[inst]);
+				    &aux_engine_regs[inst],
+					&aux_mask,
+					&aux_shift,
+					ctx->dc->caps.extended_aux_timeout_support);
 
 	return &aux_engine->base;
 }
@@ -668,6 +714,7 @@ struct clock_source *dce100_clock_source_create(
 		return &clk_src->base;
 	}
 
+	kfree(clk_src);
 	BREAK_TO_DEBUGGER();
 	return NULL;
 }
@@ -996,6 +1043,8 @@ static bool construct(
 	dc->caps.max_cursor_size = 128;
 	dc->caps.dual_link_dvi = true;
 	dc->caps.disable_dp_clk_share = true;
+	dc->caps.extended_aux_timeout_support = false;
+
 	for (i = 0; i < pool->base.pipe_count; i++) {
 		pool->base.timing_generators[i] =
 			dce100_timing_generator_create(
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
index 01a924bf477a..f0e837d14000 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
@@ -944,7 +944,6 @@ void dce110_enable_audio_stream(struct pipe_ctx *pipe_ctx)
 {
 	/* notify audio driver for audio modes of monitor */
 	struct dc *core_dc;
-	struct pp_smu_funcs *pp_smu = NULL;
 	struct clk_mgr *clk_mgr;
 	unsigned int i, num_audio = 1;
 
@@ -957,9 +956,6 @@ void dce110_enable_audio_stream(struct pipe_ctx *pipe_ctx)
 	if (pipe_ctx->stream_res.audio && pipe_ctx->stream_res.audio->enabled == true)
 		return;
 
-	if (core_dc->res_pool->pp_smu)
-		pp_smu = core_dc->res_pool->pp_smu;
-
 	if (pipe_ctx->stream_res.audio) {
 		for (i = 0; i < MAX_PIPES; i++) {
 			/*current_state not updated yet*/
@@ -984,7 +980,6 @@ void dce110_enable_audio_stream(struct pipe_ctx *pipe_ctx)
 void dce110_disable_audio_stream(struct pipe_ctx *pipe_ctx)
 {
 	struct dc *dc;
-	struct pp_smu_funcs *pp_smu = NULL;
 	struct clk_mgr *clk_mgr;
 
 	if (!pipe_ctx || !pipe_ctx->stream)
@@ -1001,9 +996,6 @@ void dce110_disable_audio_stream(struct pipe_ctx *pipe_ctx)
 	if (pipe_ctx->stream_res.audio) {
 		pipe_ctx->stream_res.audio->enabled = false;
 
-		if (dc->res_pool->pp_smu)
-			pp_smu = dc->res_pool->pp_smu;
-
 		if (dc_is_dp_signal(pipe_ctx->stream->signal))
 			pipe_ctx->stream_res.stream_enc->funcs->dp_audio_disable(
 					pipe_ctx->stream_res.stream_enc);
@@ -1169,8 +1161,9 @@ static void build_audio_output(
 		}
 	}
 
-	if (pipe_ctx->stream->signal == SIGNAL_TYPE_DISPLAY_PORT ||
-			pipe_ctx->stream->signal == SIGNAL_TYPE_DISPLAY_PORT_MST) {
+	if (state->clk_mgr &&
+		(pipe_ctx->stream->signal == SIGNAL_TYPE_DISPLAY_PORT ||
+			pipe_ctx->stream->signal == SIGNAL_TYPE_DISPLAY_PORT_MST)) {
 		audio_output->pll_info.dp_dto_source_clock_in_khz =
 				state->clk_mgr->funcs->get_dp_ref_clk_frequency(
 						state->clk_mgr);
@@ -1418,7 +1411,7 @@ static enum dc_status apply_single_controller_ctx_to_hw(
 
 	pipe_ctx->plane_res.scl_data.lb_params.alpha_en = pipe_ctx->bottom_pipe != 0;
 
-	pipe_ctx->stream->link->psr_enabled = false;
+	pipe_ctx->stream->link->psr_feature_enabled = false;
 
 	return DC_OK;
 }
@@ -1428,8 +1421,6 @@ static enum dc_status apply_single_controller_ctx_to_hw(
 static void power_down_encoders(struct dc *dc)
 {
 	int i;
-	enum connector_id connector_id;
-	enum signal_type signal = SIGNAL_TYPE_NONE;
 
 	/* do not know BIOS back-front mapping, simply blank all. It will not
 	 * hurt for non-DP
@@ -1440,15 +1431,12 @@ static void power_down_encoders(struct dc *dc)
 	}
 
 	for (i = 0; i < dc->link_count; i++) {
-		connector_id = dal_graphics_object_id_get_connector_id(dc->links[i]->link_id);
-		if ((connector_id == CONNECTOR_ID_DISPLAY_PORT) ||
-			(connector_id == CONNECTOR_ID_EDP)) {
+		enum signal_type signal = dc->links[i]->connector_signal;
 
+		if ((signal == SIGNAL_TYPE_EDP) ||
+			(signal == SIGNAL_TYPE_DISPLAY_PORT))
 			if (!dc->links[i]->wa_flags.dp_keep_receiver_powered)
 				dp_receiver_power_ctrl(dc->links[i], false);
-			if (connector_id == CONNECTOR_ID_EDP)
-				signal = SIGNAL_TYPE_EDP;
-		}
 
 		dc->links[i]->link_enc->funcs->disable_output(
 				dc->links[i]->link_enc, signal);
@@ -1529,18 +1517,6 @@ static struct dc_stream_state *get_edp_stream(struct dc_state *context)
 	return NULL;
 }
 
-static struct dc_link *get_edp_link(struct dc *dc)
-{
-	int i;
-
-	// report any eDP links, even unconnected DDI's
-	for (i = 0; i < dc->link_count; i++) {
-		if (dc->links[i]->connector_signal == SIGNAL_TYPE_EDP)
-			return dc->links[i];
-	}
-	return NULL;
-}
-
 static struct dc_link *get_edp_link_with_sink(
 		struct dc *dc,
 		struct dc_state *context)
@@ -1834,7 +1810,7 @@ static bool should_enable_fbc(struct dc *dc,
 		return false;
 
 	/* PSR should not be enabled */
-	if (pipe_ctx->stream->link->psr_enabled)
+	if (pipe_ctx->stream->link->psr_feature_enabled)
 		return false;
 
 	/* Nothing to compress */
@@ -2464,7 +2440,6 @@ static void dce110_program_front_end_for_pipe(
 		struct dc *dc, struct pipe_ctx *pipe_ctx)
 {
 	struct mem_input *mi = pipe_ctx->plane_res.mi;
-	struct pipe_ctx *old_pipe = NULL;
 	struct dc_plane_state *plane_state = pipe_ctx->plane_state;
 	struct xfm_grph_csc_adjustment adjust;
 	struct out_csc_color_matrix tbl_entry;
@@ -2472,9 +2447,6 @@ static void dce110_program_front_end_for_pipe(
 	DC_LOGGER_INIT();
 	memset(&tbl_entry, 0, sizeof(tbl_entry));
 
-	if (dc->current_state)
-		old_pipe = &dc->current_state->res_ctx.pipe_ctx[pipe_ctx->pipe_idx];
-
 	memset(&adjust, 0, sizeof(adjust));
 	adjust.gamut_adjust_type = GRAPHICS_GAMUT_ADJUST_TYPE_BYPASS;
 
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_resource.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_resource.c
index 318e9c2e2ca8..83a4dbf6d76e 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_resource.c
@@ -275,6 +275,14 @@ static const struct dce_stream_encoder_mask se_mask = {
 		SE_COMMON_MASK_SH_LIST_DCE110(_MASK)
 };
 
+static const struct dce110_aux_registers_shift aux_shift = {
+	DCE_AUX_MASK_SH_LIST(__SHIFT)
+};
+
+static const struct dce110_aux_registers_mask aux_mask = {
+	DCE_AUX_MASK_SH_LIST(_MASK)
+};
+
 #define opp_regs(id)\
 [id] = {\
 	OPP_DCE_110_REG_LIST(id),\
@@ -440,6 +448,37 @@ static const struct dc_plane_cap underlay_plane_cap = {
 #define CC_DC_HDMI_STRAPS__AUDIO_STREAM_NUMBER__SHIFT 0x8
 #endif
 
+static int map_transmitter_id_to_phy_instance(
+	enum transmitter transmitter)
+{
+	switch (transmitter) {
+	case TRANSMITTER_UNIPHY_A:
+		return 0;
+	break;
+	case TRANSMITTER_UNIPHY_B:
+		return 1;
+	break;
+	case TRANSMITTER_UNIPHY_C:
+		return 2;
+	break;
+	case TRANSMITTER_UNIPHY_D:
+		return 3;
+	break;
+	case TRANSMITTER_UNIPHY_E:
+		return 4;
+	break;
+	case TRANSMITTER_UNIPHY_F:
+		return 5;
+	break;
+	case TRANSMITTER_UNIPHY_G:
+		return 6;
+	break;
+	default:
+		ASSERT(0);
+		return 0;
+	}
+}
+
 static void read_dce_straps(
 	struct dc_context *ctx,
 	struct resource_straps *straps)
@@ -617,14 +656,18 @@ static struct link_encoder *dce110_link_encoder_create(
 {
 	struct dce110_link_encoder *enc110 =
 		kzalloc(sizeof(struct dce110_link_encoder), GFP_KERNEL);
+	int link_regs_id;
 
 	if (!enc110)
 		return NULL;
 
+	link_regs_id =
+		map_transmitter_id_to_phy_instance(enc_init_data->transmitter);
+
 	dce110_link_encoder_construct(enc110,
 				      enc_init_data,
 				      &link_enc_feature,
-				      &link_enc_regs[enc_init_data->transmitter],
+				      &link_enc_regs[link_regs_id],
 				      &link_enc_aux_regs[enc_init_data->channel - 1],
 				      &link_enc_hpd_regs[enc_init_data->hpd_source]);
 	return &enc110->base;
@@ -657,7 +700,10 @@ struct dce_aux *dce110_aux_engine_create(
 
 	dce110_aux_engine_construct(aux_engine, ctx, inst,
 				    SW_AUX_TIMEOUT_PERIOD_MULTIPLIER * AUX_TIMEOUT_PERIOD,
-				    &aux_engine_regs[inst]);
+				    &aux_engine_regs[inst],
+					&aux_mask,
+					&aux_shift,
+					ctx->dc->caps.extended_aux_timeout_support);
 
 	return &aux_engine->base;
 }
@@ -714,6 +760,7 @@ struct clock_source *dce110_clock_source_create(
 		return &clk_src->base;
 	}
 
+	kfree(clk_src);
 	BREAK_TO_DEBUGGER();
 	return NULL;
 }
@@ -1292,6 +1339,7 @@ static bool construct(
 	dc->caps.i2c_speed_in_khz = 100;
 	dc->caps.max_cursor_size = 128;
 	dc->caps.is_apu = true;
+	dc->caps.extended_aux_timeout_support = false;
 
 	/*************************************************
 	 *  Create resources                             *
diff --git a/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.c b/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.c
index 83e1878161c9..97dcc5d0862b 100644
--- a/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.c
@@ -172,6 +172,14 @@ static const struct dce_abm_mask abm_mask = {
 		ABM_MASK_SH_LIST_DCE110(_MASK)
 };
 
+static const struct dce110_aux_registers_shift aux_shift = {
+	DCE_AUX_MASK_SH_LIST(__SHIFT)
+};
+
+static const struct dce110_aux_registers_mask aux_mask = {
+	DCE_AUX_MASK_SH_LIST(_MASK)
+};
+
 #define ipp_regs(id)\
 [id] = {\
 		IPP_DCE110_REG_LIST_DCE_BASE(id)\
@@ -417,6 +425,37 @@ static const struct dc_plane_cap plane_cap = {
 #define CC_DC_HDMI_STRAPS__AUDIO_STREAM_NUMBER__SHIFT 0x8
 #endif
 
+static int map_transmitter_id_to_phy_instance(
+	enum transmitter transmitter)
+{
+	switch (transmitter) {
+	case TRANSMITTER_UNIPHY_A:
+		return 0;
+	break;
+	case TRANSMITTER_UNIPHY_B:
+		return 1;
+	break;
+	case TRANSMITTER_UNIPHY_C:
+		return 2;
+	break;
+	case TRANSMITTER_UNIPHY_D:
+		return 3;
+	break;
+	case TRANSMITTER_UNIPHY_E:
+		return 4;
+	break;
+	case TRANSMITTER_UNIPHY_F:
+		return 5;
+	break;
+	case TRANSMITTER_UNIPHY_G:
+		return 6;
+	break;
+	default:
+		ASSERT(0);
+		return 0;
+	}
+}
+
 static void read_dce_straps(
 	struct dc_context *ctx,
 	struct resource_straps *straps)
@@ -575,14 +614,18 @@ struct link_encoder *dce112_link_encoder_create(
 {
 	struct dce110_link_encoder *enc110 =
 		kzalloc(sizeof(struct dce110_link_encoder), GFP_KERNEL);
+	int link_regs_id;
 
 	if (!enc110)
 		return NULL;
 
+	link_regs_id =
+		map_transmitter_id_to_phy_instance(enc_init_data->transmitter);
+
 	dce110_link_encoder_construct(enc110,
 				      enc_init_data,
 				      &link_enc_feature,
-				      &link_enc_regs[enc_init_data->transmitter],
+				      &link_enc_regs[link_regs_id],
 				      &link_enc_aux_regs[enc_init_data->channel - 1],
 				      &link_enc_hpd_regs[enc_init_data->hpd_source]);
 	return &enc110->base;
@@ -630,7 +673,10 @@ struct dce_aux *dce112_aux_engine_create(
 
 	dce110_aux_engine_construct(aux_engine, ctx, inst,
 				    SW_AUX_TIMEOUT_PERIOD_MULTIPLIER * AUX_TIMEOUT_PERIOD,
-				    &aux_engine_regs[inst]);
+				    &aux_engine_regs[inst],
+					&aux_mask,
+					&aux_shift,
+					ctx->dc->caps.extended_aux_timeout_support);
 
 	return &aux_engine->base;
 }
@@ -687,6 +733,7 @@ struct clock_source *dce112_clock_source_create(
 		return &clk_src->base;
 	}
 
+	kfree(clk_src);
 	BREAK_TO_DEBUGGER();
 	return NULL;
 }
@@ -1162,7 +1209,7 @@ static bool construct(
 	dc->caps.i2c_speed_in_khz = 100;
 	dc->caps.max_cursor_size = 128;
 	dc->caps.dual_link_dvi = true;
-
+	dc->caps.extended_aux_timeout_support = false;
 
 	/*************************************************
 	 *  Create resources                             *
diff --git a/drivers/gpu/drm/amd/display/dc/dce120/dce120_resource.c b/drivers/gpu/drm/amd/display/dc/dce120/dce120_resource.c
index 8b85e5274bba..63543f6918ff 100644
--- a/drivers/gpu/drm/amd/display/dc/dce120/dce120_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dce120/dce120_resource.c
@@ -293,6 +293,14 @@ static const struct dce_stream_encoder_mask se_mask = {
 		SE_COMMON_MASK_SH_LIST_DCE120(_MASK)
 };
 
+static const struct dce110_aux_registers_shift aux_shift = {
+	DCE12_AUX_MASK_SH_LIST(__SHIFT)
+};
+
+static const struct dce110_aux_registers_mask aux_mask = {
+	DCE12_AUX_MASK_SH_LIST(_MASK)
+};
+
 #define opp_regs(id)\
 [id] = {\
 	OPP_DCE_120_REG_LIST(id),\
@@ -356,6 +364,37 @@ static const struct dce_audio_mask audio_mask = {
 		DCE120_AUD_COMMON_MASK_SH_LIST(_MASK)
 };
 
+static int map_transmitter_id_to_phy_instance(
+	enum transmitter transmitter)
+{
+	switch (transmitter) {
+	case TRANSMITTER_UNIPHY_A:
+		return 0;
+	break;
+	case TRANSMITTER_UNIPHY_B:
+		return 1;
+	break;
+	case TRANSMITTER_UNIPHY_C:
+		return 2;
+	break;
+	case TRANSMITTER_UNIPHY_D:
+		return 3;
+	break;
+	case TRANSMITTER_UNIPHY_E:
+		return 4;
+	break;
+	case TRANSMITTER_UNIPHY_F:
+		return 5;
+	break;
+	case TRANSMITTER_UNIPHY_G:
+		return 6;
+	break;
+	default:
+		ASSERT(0);
+		return 0;
+	}
+}
+
 #define clk_src_regs(index, id)\
 [index] = {\
 	CS_COMMON_REG_LIST_DCE_112(id),\
@@ -404,7 +443,10 @@ struct dce_aux *dce120_aux_engine_create(
 
 	dce110_aux_engine_construct(aux_engine, ctx, inst,
 				    SW_AUX_TIMEOUT_PERIOD_MULTIPLIER * AUX_TIMEOUT_PERIOD,
-				    &aux_engine_regs[inst]);
+				    &aux_engine_regs[inst],
+					&aux_mask,
+					&aux_shift,
+					ctx->dc->caps.extended_aux_timeout_support);
 
 	return &aux_engine->base;
 }
@@ -500,6 +542,7 @@ static struct clock_source *dce120_clock_source_create(
 		return &clk_src->base;
 	}
 
+	kfree(clk_src);
 	BREAK_TO_DEBUGGER();
 	return NULL;
 }
@@ -654,14 +697,18 @@ static struct link_encoder *dce120_link_encoder_create(
 {
 	struct dce110_link_encoder *enc110 =
 		kzalloc(sizeof(struct dce110_link_encoder), GFP_KERNEL);
+	int link_regs_id;
 
 	if (!enc110)
 		return NULL;
 
+	link_regs_id =
+		map_transmitter_id_to_phy_instance(enc_init_data->transmitter);
+
 	dce110_link_encoder_construct(enc110,
 				      enc_init_data,
 				      &link_enc_feature,
-				      &link_enc_regs[enc_init_data->transmitter],
+				      &link_enc_regs[link_regs_id],
 				      &link_enc_aux_regs[enc_init_data->channel - 1],
 				      &link_enc_hpd_regs[enc_init_data->hpd_source]);
 
@@ -1005,7 +1052,7 @@ static bool construct(
 	dc->caps.max_cursor_size = 128;
 	dc->caps.dual_link_dvi = true;
 	dc->caps.psp_setup_panel_mode = true;
-
+	dc->caps.extended_aux_timeout_support = false;
 	dc->debug = debug_defaults;
 
 	/*************************************************
diff --git a/drivers/gpu/drm/amd/display/dc/dce80/dce80_resource.c b/drivers/gpu/drm/amd/display/dc/dce80/dce80_resource.c
index 4625df9f9fd2..3e8d4b49f279 100644
--- a/drivers/gpu/drm/amd/display/dc/dce80/dce80_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dce80/dce80_resource.c
@@ -288,6 +288,14 @@ static const struct dce_opp_mask opp_mask = {
 	OPP_COMMON_MASK_SH_LIST_DCE_80(_MASK)
 };
 
+static const struct dce110_aux_registers_shift aux_shift = {
+	DCE10_AUX_MASK_SH_LIST(__SHIFT)
+};
+
+static const struct dce110_aux_registers_mask aux_mask = {
+	DCE10_AUX_MASK_SH_LIST(_MASK)
+};
+
 #define aux_engine_regs(id)\
 [id] = {\
 	AUX_COMMON_REG_LIST(id), \
@@ -431,6 +439,37 @@ static const struct dce_abm_mask abm_mask = {
 #define CC_DC_HDMI_STRAPS__AUDIO_STREAM_NUMBER__SHIFT 0x8
 #endif
 
+static int map_transmitter_id_to_phy_instance(
+	enum transmitter transmitter)
+{
+	switch (transmitter) {
+	case TRANSMITTER_UNIPHY_A:
+		return 0;
+	break;
+	case TRANSMITTER_UNIPHY_B:
+		return 1;
+	break;
+	case TRANSMITTER_UNIPHY_C:
+		return 2;
+	break;
+	case TRANSMITTER_UNIPHY_D:
+		return 3;
+	break;
+	case TRANSMITTER_UNIPHY_E:
+		return 4;
+	break;
+	case TRANSMITTER_UNIPHY_F:
+		return 5;
+	break;
+	case TRANSMITTER_UNIPHY_G:
+		return 6;
+	break;
+	default:
+		ASSERT(0);
+		return 0;
+	}
+}
+
 static void read_dce_straps(
 	struct dc_context *ctx,
 	struct resource_straps *straps)
@@ -491,7 +530,10 @@ struct dce_aux *dce80_aux_engine_create(
 
 	dce110_aux_engine_construct(aux_engine, ctx, inst,
 				    SW_AUX_TIMEOUT_PERIOD_MULTIPLIER * AUX_TIMEOUT_PERIOD,
-				    &aux_engine_regs[inst]);
+				    &aux_engine_regs[inst],
+					&aux_mask,
+					&aux_shift,
+					ctx->dc->caps.extended_aux_timeout_support);
 
 	return &aux_engine->base;
 }
@@ -669,14 +711,18 @@ struct link_encoder *dce80_link_encoder_create(
 {
 	struct dce110_link_encoder *enc110 =
 		kzalloc(sizeof(struct dce110_link_encoder), GFP_KERNEL);
+	int link_regs_id;
 
 	if (!enc110)
 		return NULL;
 
+	link_regs_id =
+		map_transmitter_id_to_phy_instance(enc_init_data->transmitter);
+
 	dce110_link_encoder_construct(enc110,
 				      enc_init_data,
 				      &link_enc_feature,
-				      &link_enc_regs[enc_init_data->transmitter],
+				      &link_enc_regs[link_regs_id],
 				      &link_enc_aux_regs[enc_init_data->channel - 1],
 				      &link_enc_hpd_regs[enc_init_data->hpd_source]);
 	return &enc110->base;
@@ -701,6 +747,7 @@ struct clock_source *dce80_clock_source_create(
 		return &clk_src->base;
 	}
 
+	kfree(clk_src);
 	BREAK_TO_DEBUGGER();
 	return NULL;
 }
@@ -894,6 +941,7 @@ static bool dce80_construct(
 	dc->caps.i2c_speed_in_khz = 40;
 	dc->caps.max_cursor_size = 128;
 	dc->caps.dual_link_dvi = true;
+	dc->caps.extended_aux_timeout_support = false;
 
 	/*************************************************
 	 *  Create resources                             *
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_cm_common.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_cm_common.c
index 01c7e30b9ce1..bbd6e01b3eca 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_cm_common.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_cm_common.c
@@ -393,6 +393,10 @@ bool cm_helper_translate_curve_to_hw_format(
 	rgb_resulted[hw_points - 1].green = output_tf->tf_pts.green[start_index];
 	rgb_resulted[hw_points - 1].blue = output_tf->tf_pts.blue[start_index];
 
+	rgb_resulted[hw_points].red = rgb_resulted[hw_points - 1].red;
+	rgb_resulted[hw_points].green = rgb_resulted[hw_points - 1].green;
+	rgb_resulted[hw_points].blue = rgb_resulted[hw_points - 1].blue;
+
 	// All 3 color channels have same x
 	corner_points[0].red.x = dc_fixpt_pow(dc_fixpt_from_int(2),
 					     dc_fixpt_from_int(region_start));
@@ -464,13 +468,6 @@ bool cm_helper_translate_curve_to_hw_format(
 
 	i = 1;
 	while (i != hw_points + 1) {
-		if (dc_fixpt_lt(rgb_plus_1->red, rgb->red))
-			rgb_plus_1->red = rgb->red;
-		if (dc_fixpt_lt(rgb_plus_1->green, rgb->green))
-			rgb_plus_1->green = rgb->green;
-		if (dc_fixpt_lt(rgb_plus_1->blue, rgb->blue))
-			rgb_plus_1->blue = rgb->blue;
-
 		rgb->delta_red   = dc_fixpt_sub(rgb_plus_1->red,   rgb->red);
 		rgb->delta_green = dc_fixpt_sub(rgb_plus_1->green, rgb->green);
 		rgb->delta_blue  = dc_fixpt_sub(rgb_plus_1->blue,  rgb->blue);
@@ -562,6 +559,10 @@ bool cm_helper_translate_curve_to_degamma_hw_format(
 	rgb_resulted[hw_points - 1].green = output_tf->tf_pts.green[start_index];
 	rgb_resulted[hw_points - 1].blue = output_tf->tf_pts.blue[start_index];
 
+	rgb_resulted[hw_points].red = rgb_resulted[hw_points - 1].red;
+	rgb_resulted[hw_points].green = rgb_resulted[hw_points - 1].green;
+	rgb_resulted[hw_points].blue = rgb_resulted[hw_points - 1].blue;
+
 	corner_points[0].red.x = dc_fixpt_pow(dc_fixpt_from_int(2),
 					     dc_fixpt_from_int(region_start));
 	corner_points[0].green.x = corner_points[0].red.x;
@@ -624,13 +625,6 @@ bool cm_helper_translate_curve_to_degamma_hw_format(
 
 	i = 1;
 	while (i != hw_points + 1) {
-		if (dc_fixpt_lt(rgb_plus_1->red, rgb->red))
-			rgb_plus_1->red = rgb->red;
-		if (dc_fixpt_lt(rgb_plus_1->green, rgb->green))
-			rgb_plus_1->green = rgb->green;
-		if (dc_fixpt_lt(rgb_plus_1->blue, rgb->blue))
-			rgb_plus_1->blue = rgb->blue;
-
 		rgb->delta_red   = dc_fixpt_sub(rgb_plus_1->red,   rgb->red);
 		rgb->delta_green = dc_fixpt_sub(rgb_plus_1->green, rgb->green);
 		rgb->delta_blue  = dc_fixpt_sub(rgb_plus_1->blue,  rgb->blue);
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c
index d8b2da18db39..997e9582edc7 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c
@@ -129,7 +129,7 @@ void dpp_set_gamut_remap_bypass(struct dcn10_dpp *dpp)
 
 #define IDENTITY_RATIO(ratio) (dc_fixpt_u2d19(ratio) == (1 << 19))
 
-static bool dpp_get_optimal_number_of_taps(
+bool dpp1_get_optimal_number_of_taps(
 		struct dpp *dpp,
 		struct scaler_data *scl_data,
 		const struct scaling_taps *in_taps)
@@ -521,7 +521,7 @@ static const struct dpp_funcs dcn10_dpp_funcs = {
 		.dpp_read_state = dpp_read_state,
 		.dpp_reset = dpp_reset,
 		.dpp_set_scaler = dpp1_dscl_set_scaler_manual_scale,
-		.dpp_get_optimal_number_of_taps = dpp_get_optimal_number_of_taps,
+		.dpp_get_optimal_number_of_taps = dpp1_get_optimal_number_of_taps,
 		.dpp_set_gamut_remap = dpp1_cm_set_gamut_remap,
 		.dpp_set_csc_adjustment = dpp1_cm_set_output_csc_adjustment,
 		.dpp_set_csc_default = dpp1_cm_set_output_csc_default,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h
index e2c613611ac9..1d4a7d640334 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h
@@ -1504,6 +1504,11 @@ void dpp1_set_hdr_multiplier(
 		struct dpp *dpp_base,
 		uint32_t multiplier);
 
+bool dpp1_get_optimal_number_of_taps(
+		struct dpp *dpp,
+		struct scaler_data *scl_data,
+		const struct scaling_taps *in_taps);
+
 void dpp1_construct(struct dcn10_dpp *dpp1,
 	struct dc_context *ctx,
 	uint32_t inst,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c
index 001db49e4bb2..14d1be6c66e6 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c
@@ -841,6 +841,14 @@ void min_set_viewport(
 	REG_SET_2(DCSURF_PRI_VIEWPORT_START_C, 0,
 		  PRI_VIEWPORT_X_START_C, viewport_c->x,
 		  PRI_VIEWPORT_Y_START_C, viewport_c->y);
+
+	REG_SET_2(DCSURF_SEC_VIEWPORT_DIMENSION_C, 0,
+		  SEC_VIEWPORT_WIDTH_C, viewport_c->width,
+		  SEC_VIEWPORT_HEIGHT_C, viewport_c->height);
+
+	REG_SET_2(DCSURF_SEC_VIEWPORT_START_C, 0,
+		  SEC_VIEWPORT_X_START_C, viewport_c->x,
+		  SEC_VIEWPORT_Y_START_C, viewport_c->y);
 }
 
 void hubp1_read_state_common(struct hubp *hubp)
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h
index cb20d10288c0..ae70d9c0aa1d 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h
@@ -47,6 +47,8 @@
 	SRI(DCSURF_SEC_VIEWPORT_START, HUBP, id), \
 	SRI(DCSURF_PRI_VIEWPORT_DIMENSION_C, HUBP, id), \
 	SRI(DCSURF_PRI_VIEWPORT_START_C, HUBP, id), \
+	SRI(DCSURF_SEC_VIEWPORT_DIMENSION_C, HUBP, id), \
+	SRI(DCSURF_SEC_VIEWPORT_START_C, HUBP, id), \
 	SRI(DCSURF_PRIMARY_SURFACE_ADDRESS_HIGH, HUBPREQ, id),\
 	SRI(DCSURF_PRIMARY_SURFACE_ADDRESS, HUBPREQ, id),\
 	SRI(DCSURF_SECONDARY_SURFACE_ADDRESS_HIGH, HUBPREQ, id),\
@@ -57,8 +59,12 @@
 	SRI(DCSURF_SECONDARY_META_SURFACE_ADDRESS, HUBPREQ, id),\
 	SRI(DCSURF_PRIMARY_SURFACE_ADDRESS_HIGH_C, HUBPREQ, id),\
 	SRI(DCSURF_PRIMARY_SURFACE_ADDRESS_C, HUBPREQ, id),\
+	SRI(DCSURF_SECONDARY_SURFACE_ADDRESS_HIGH_C, HUBPREQ, id),\
+	SRI(DCSURF_SECONDARY_SURFACE_ADDRESS_C, HUBPREQ, id),\
 	SRI(DCSURF_PRIMARY_META_SURFACE_ADDRESS_HIGH_C, HUBPREQ, id),\
 	SRI(DCSURF_PRIMARY_META_SURFACE_ADDRESS_C, HUBPREQ, id),\
+	SRI(DCSURF_SECONDARY_META_SURFACE_ADDRESS_HIGH_C, HUBPREQ, id),\
+	SRI(DCSURF_SECONDARY_META_SURFACE_ADDRESS_C, HUBPREQ, id),\
 	SRI(DCSURF_SURFACE_INUSE, HUBPREQ, id),\
 	SRI(DCSURF_SURFACE_INUSE_HIGH, HUBPREQ, id),\
 	SRI(DCSURF_SURFACE_INUSE_C, HUBPREQ, id),\
@@ -150,6 +156,8 @@
 	uint32_t DCSURF_SEC_VIEWPORT_START; \
 	uint32_t DCSURF_PRI_VIEWPORT_DIMENSION_C; \
 	uint32_t DCSURF_PRI_VIEWPORT_START_C; \
+	uint32_t DCSURF_SEC_VIEWPORT_DIMENSION_C; \
+	uint32_t DCSURF_SEC_VIEWPORT_START_C; \
 	uint32_t DCSURF_PRIMARY_SURFACE_ADDRESS_HIGH; \
 	uint32_t DCSURF_PRIMARY_SURFACE_ADDRESS; \
 	uint32_t DCSURF_SECONDARY_SURFACE_ADDRESS_HIGH; \
@@ -160,8 +168,12 @@
 	uint32_t DCSURF_SECONDARY_META_SURFACE_ADDRESS; \
 	uint32_t DCSURF_PRIMARY_SURFACE_ADDRESS_HIGH_C; \
 	uint32_t DCSURF_PRIMARY_SURFACE_ADDRESS_C; \
+	uint32_t DCSURF_SECONDARY_SURFACE_ADDRESS_HIGH_C; \
+	uint32_t DCSURF_SECONDARY_SURFACE_ADDRESS_C; \
 	uint32_t DCSURF_PRIMARY_META_SURFACE_ADDRESS_HIGH_C; \
 	uint32_t DCSURF_PRIMARY_META_SURFACE_ADDRESS_C; \
+	uint32_t DCSURF_SECONDARY_META_SURFACE_ADDRESS_HIGH_C; \
+	uint32_t DCSURF_SECONDARY_META_SURFACE_ADDRESS_C; \
 	uint32_t DCSURF_SURFACE_INUSE; \
 	uint32_t DCSURF_SURFACE_INUSE_HIGH; \
 	uint32_t DCSURF_SURFACE_INUSE_C; \
@@ -279,6 +291,10 @@
 	HUBP_SF(HUBP0_DCSURF_PRI_VIEWPORT_DIMENSION_C, PRI_VIEWPORT_HEIGHT_C, mask_sh),\
 	HUBP_SF(HUBP0_DCSURF_PRI_VIEWPORT_START_C, PRI_VIEWPORT_X_START_C, mask_sh),\
 	HUBP_SF(HUBP0_DCSURF_PRI_VIEWPORT_START_C, PRI_VIEWPORT_Y_START_C, mask_sh),\
+	HUBP_SF(HUBP0_DCSURF_SEC_VIEWPORT_DIMENSION_C, SEC_VIEWPORT_WIDTH_C, mask_sh),\
+	HUBP_SF(HUBP0_DCSURF_SEC_VIEWPORT_DIMENSION_C, SEC_VIEWPORT_HEIGHT_C, mask_sh),\
+	HUBP_SF(HUBP0_DCSURF_SEC_VIEWPORT_START_C, SEC_VIEWPORT_X_START_C, mask_sh),\
+	HUBP_SF(HUBP0_DCSURF_SEC_VIEWPORT_START_C, SEC_VIEWPORT_Y_START_C, mask_sh),\
 	HUBP_SF(HUBPREQ0_DCSURF_PRIMARY_SURFACE_ADDRESS_HIGH, PRIMARY_SURFACE_ADDRESS_HIGH, mask_sh),\
 	HUBP_SF(HUBPREQ0_DCSURF_PRIMARY_SURFACE_ADDRESS, PRIMARY_SURFACE_ADDRESS, mask_sh),\
 	HUBP_SF(HUBPREQ0_DCSURF_SECONDARY_SURFACE_ADDRESS_HIGH, SECONDARY_SURFACE_ADDRESS_HIGH, mask_sh),\
@@ -289,8 +305,12 @@
 	HUBP_SF(HUBPREQ0_DCSURF_SECONDARY_META_SURFACE_ADDRESS, SECONDARY_META_SURFACE_ADDRESS, mask_sh),\
 	HUBP_SF(HUBPREQ0_DCSURF_PRIMARY_SURFACE_ADDRESS_HIGH_C, PRIMARY_SURFACE_ADDRESS_HIGH_C, mask_sh),\
 	HUBP_SF(HUBPREQ0_DCSURF_PRIMARY_SURFACE_ADDRESS_C, PRIMARY_SURFACE_ADDRESS_C, mask_sh),\
+	HUBP_SF(HUBPREQ0_DCSURF_SECONDARY_SURFACE_ADDRESS_HIGH_C, SECONDARY_SURFACE_ADDRESS_HIGH_C, mask_sh),\
+	HUBP_SF(HUBPREQ0_DCSURF_SECONDARY_SURFACE_ADDRESS_C, SECONDARY_SURFACE_ADDRESS_C, mask_sh),\
 	HUBP_SF(HUBPREQ0_DCSURF_PRIMARY_META_SURFACE_ADDRESS_HIGH_C, PRIMARY_META_SURFACE_ADDRESS_HIGH_C, mask_sh),\
 	HUBP_SF(HUBPREQ0_DCSURF_PRIMARY_META_SURFACE_ADDRESS_C, PRIMARY_META_SURFACE_ADDRESS_C, mask_sh),\
+	HUBP_SF(HUBPREQ0_DCSURF_SECONDARY_META_SURFACE_ADDRESS_HIGH_C, SECONDARY_META_SURFACE_ADDRESS_HIGH_C, mask_sh),\
+	HUBP_SF(HUBPREQ0_DCSURF_SECONDARY_META_SURFACE_ADDRESS_C, SECONDARY_META_SURFACE_ADDRESS_C, mask_sh),\
 	HUBP_SF(HUBPREQ0_DCSURF_SURFACE_INUSE, SURFACE_INUSE_ADDRESS, mask_sh),\
 	HUBP_SF(HUBPREQ0_DCSURF_SURFACE_INUSE_HIGH, SURFACE_INUSE_ADDRESS_HIGH, mask_sh),\
 	HUBP_SF(HUBPREQ0_DCSURF_SURFACE_INUSE_C, SURFACE_INUSE_ADDRESS_C, mask_sh),\
@@ -469,6 +489,10 @@
 	type PRI_VIEWPORT_HEIGHT_C; \
 	type PRI_VIEWPORT_X_START_C; \
 	type PRI_VIEWPORT_Y_START_C; \
+	type SEC_VIEWPORT_WIDTH_C; \
+	type SEC_VIEWPORT_HEIGHT_C; \
+	type SEC_VIEWPORT_X_START_C; \
+	type SEC_VIEWPORT_Y_START_C; \
 	type PRIMARY_SURFACE_ADDRESS_HIGH;\
 	type PRIMARY_SURFACE_ADDRESS;\
 	type SECONDARY_SURFACE_ADDRESS_HIGH;\
@@ -479,8 +503,12 @@
 	type SECONDARY_META_SURFACE_ADDRESS;\
 	type PRIMARY_SURFACE_ADDRESS_HIGH_C;\
 	type PRIMARY_SURFACE_ADDRESS_C;\
+	type SECONDARY_SURFACE_ADDRESS_HIGH_C;\
+	type SECONDARY_SURFACE_ADDRESS_C;\
 	type PRIMARY_META_SURFACE_ADDRESS_HIGH_C;\
 	type PRIMARY_META_SURFACE_ADDRESS_C;\
+	type SECONDARY_META_SURFACE_ADDRESS_HIGH_C;\
+	type SECONDARY_META_SURFACE_ADDRESS_C;\
 	type SURFACE_INUSE_ADDRESS;\
 	type SURFACE_INUSE_ADDRESS_HIGH;\
 	type SURFACE_INUSE_ADDRESS_C;\
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
index 60123db7ba02..eb91432621ab 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
@@ -670,6 +670,10 @@ static void dcn10_bios_golden_init(struct dc *dc)
 	int i;
 	bool allow_self_fresh_force_enable = true;
 
+#if defined(CONFIG_DRM_AMD_DC_DCN2_1)
+	if (dc->hwss.s0i3_golden_init_wa && dc->hwss.s0i3_golden_init_wa(dc))
+		return;
+#endif
 	if (dc->res_pool->hubbub->funcs->is_allow_self_refresh_enabled)
 		allow_self_fresh_force_enable =
 				dc->res_pool->hubbub->funcs->is_allow_self_refresh_enabled(dc->res_pool->hubbub);
@@ -1300,6 +1304,10 @@ static void dcn10_init_hw(struct dc *dc)
 	}
 
 	dc->hwss.enable_power_gating_plane(dc->hwseq, true);
+
+	if (dc->clk_mgr->funcs->notify_wm_ranges)
+		dc->clk_mgr->funcs->notify_wm_ranges(dc->clk_mgr);
+
 }
 
 static void dcn10_reset_hw_ctx_wrap(
@@ -1452,15 +1460,15 @@ static void log_tf(struct dc_context *ctx,
 	DC_LOG_ALL_TF_CHANNELS("Logging all channels...");
 
 	for (i = 0; i < hw_points_num; i++) {
-		DC_LOG_GAMMA("R\t%d\t%llu\n", i, tf->tf_pts.red[i].value);
-		DC_LOG_ALL_TF_CHANNELS("G\t%d\t%llu\n", i, tf->tf_pts.green[i].value);
-		DC_LOG_ALL_TF_CHANNELS("B\t%d\t%llu\n", i, tf->tf_pts.blue[i].value);
+		DC_LOG_GAMMA("R\t%d\t%llu", i, tf->tf_pts.red[i].value);
+		DC_LOG_ALL_TF_CHANNELS("G\t%d\t%llu", i, tf->tf_pts.green[i].value);
+		DC_LOG_ALL_TF_CHANNELS("B\t%d\t%llu", i, tf->tf_pts.blue[i].value);
 	}
 
 	for (i = hw_points_num; i < MAX_NUM_HW_POINTS; i++) {
-		DC_LOG_ALL_GAMMA("R\t%d\t%llu\n", i, tf->tf_pts.red[i].value);
-		DC_LOG_ALL_TF_CHANNELS("G\t%d\t%llu\n", i, tf->tf_pts.green[i].value);
-		DC_LOG_ALL_TF_CHANNELS("B\t%d\t%llu\n", i, tf->tf_pts.blue[i].value);
+		DC_LOG_ALL_GAMMA("R\t%d\t%llu", i, tf->tf_pts.red[i].value);
+		DC_LOG_ALL_TF_CHANNELS("G\t%d\t%llu", i, tf->tf_pts.green[i].value);
+		DC_LOG_ALL_TF_CHANNELS("B\t%d\t%llu", i, tf->tf_pts.blue[i].value);
 	}
 }
 
@@ -2304,8 +2312,7 @@ void update_dchubp_dpp(
 			dc->res_pool->dccg->funcs->update_dpp_dto(
 					dc->res_pool->dccg,
 					dpp->inst,
-					pipe_ctx->plane_res.bw.dppclk_khz,
-					false);
+					pipe_ctx->plane_res.bw.dppclk_khz);
 		else
 			dc->clk_mgr->clks.dppclk_khz = should_divided_by_2 ?
 						dc->clk_mgr->clks.dispclk_khz / 2 :
@@ -2512,8 +2519,10 @@ static void program_all_pipe_in_tree(
 		pipe_ctx->stream_res.tg->funcs->set_vtg_params(
 				pipe_ctx->stream_res.tg, &pipe_ctx->stream->timing);
 
-		dc->hwss.blank_pixel_data(dc, pipe_ctx, blank);
+		if (dc->hwss.setup_vupdate_interrupt)
+			dc->hwss.setup_vupdate_interrupt(pipe_ctx);
 
+		dc->hwss.blank_pixel_data(dc, pipe_ctx, blank);
 	}
 
 	if (pipe_ctx->plane_state != NULL)
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.h
index 8bf5f0f2301d..88fcc395adf5 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.h
@@ -113,6 +113,20 @@ struct dcn10_link_enc_registers {
 	uint32_t DIG_LANE_ENABLE;
 	/* UNIPHY */
 	uint32_t CHANNEL_XBAR_CNTL;
+	/* DPCS */
+	uint32_t RDPCSTX_PHY_CNTL3;
+	uint32_t RDPCSTX_PHY_CNTL4;
+	uint32_t RDPCSTX_PHY_CNTL5;
+	uint32_t RDPCSTX_PHY_CNTL6;
+	uint32_t RDPCSTX_PHY_CNTL7;
+	uint32_t RDPCSTX_PHY_CNTL8;
+	uint32_t RDPCSTX_PHY_CNTL9;
+	uint32_t RDPCSTX_PHY_CNTL10;
+	uint32_t RDPCSTX_PHY_CNTL11;
+	uint32_t RDPCSTX_PHY_CNTL12;
+	uint32_t RDPCSTX_PHY_CNTL13;
+	uint32_t RDPCSTX_PHY_CNTL14;
+	uint32_t RDPCSTX_PHY_CNTL15;
 	/* indirect registers */
 	uint32_t RAWLANE0_DIG_PCS_XF_RX_OVRD_IN_2;
 	uint32_t RAWLANE0_DIG_PCS_XF_RX_OVRD_IN_3;
@@ -250,6 +264,10 @@ struct dcn10_link_enc_registers {
 		type RDPCS_EXT_REFCLK_EN;\
 		type RDPCS_TX_FIFO_EN;\
 		type UNIPHY_LINK_ENABLE;\
+		type UNIPHY_CHANNEL0_XBAR_SOURCE;\
+		type UNIPHY_CHANNEL1_XBAR_SOURCE;\
+		type UNIPHY_CHANNEL2_XBAR_SOURCE;\
+		type UNIPHY_CHANNEL3_XBAR_SOURCE;\
 		type UNIPHY_CHANNEL0_INVERT;\
 		type UNIPHY_CHANNEL1_INVERT;\
 		type UNIPHY_CHANNEL2_INVERT;\
@@ -337,16 +355,46 @@ struct dcn10_link_enc_registers {
 		type RDPCS_TX_FIFO_ERROR_MASK;\
 		type RDPCS_DPALT_DISABLE_TOGGLE_MASK;\
 		type RDPCS_DPALT_4LANE_TOGGLE_MASK;\
+		type RDPCS_PHY_DPALT_DP4;\
 		type RDPCS_PHY_DPALT_DISABLE;\
 		type RDPCS_PHY_DPALT_DISABLE_ACK;\
 		type RDPCS_PHY_DP_MPLLB_V2I;\
 		type RDPCS_PHY_DP_MPLLB_FREQ_VCO;\
+		type RDPCS_PHY_DP_MPLLB_CP_INT_GS;\
+		type RDPCS_PHY_RX_VREF_CTRL;\
 		type RDPCS_PHY_DP_MPLLB_CP_INT;\
 		type RDPCS_PHY_DP_MPLLB_CP_PROP;\
 		type RDPCS_PHY_RX_REF_LD_VAL;\
 		type RDPCS_PHY_RX_VCO_LD_VAL;\
 		type DPCSTX_DEBUG_CONFIG; \
-		type RDPCSTX_DEBUG_CONFIG
+		type RDPCSTX_DEBUG_CONFIG; \
+		type RDPCS_PHY_DP_TX0_EQ_MAIN;\
+		type RDPCS_PHY_DP_TX0_EQ_PRE;\
+		type RDPCS_PHY_DP_TX0_EQ_POST;\
+		type RDPCS_PHY_DP_TX1_EQ_MAIN;\
+		type RDPCS_PHY_DP_TX1_EQ_PRE;\
+		type RDPCS_PHY_DP_TX1_EQ_POST;\
+		type RDPCS_PHY_DP_TX2_EQ_MAIN;\
+		type RDPCS_PHY_DP_MPLLB_CP_PROP_GS;\
+		type RDPCS_PHY_DP_TX2_EQ_PRE;\
+		type RDPCS_PHY_DP_TX2_EQ_POST;\
+		type RDPCS_PHY_DP_TX3_EQ_MAIN;\
+		type RDPCS_PHY_DCO_RANGE;\
+		type RDPCS_PHY_DCO_FINETUNE;\
+		type RDPCS_PHY_DP_TX3_EQ_PRE;\
+		type RDPCS_PHY_DP_TX3_EQ_POST;\
+		type RDPCS_PHY_SUP_PRE_HP;\
+		type RDPCS_PHY_DP_TX0_VREGDRV_BYP;\
+		type RDPCS_PHY_DP_TX1_VREGDRV_BYP;\
+		type RDPCS_PHY_DP_TX2_VREGDRV_BYP;\
+		type RDPCS_PHY_DP_TX3_VREGDRV_BYP;\
+		type RDPCS_DMCU_DPALT_DIS_BLOCK_REG;\
+		type UNIPHYA_SOFT_RESET;\
+		type UNIPHYB_SOFT_RESET;\
+		type UNIPHYC_SOFT_RESET;\
+		type UNIPHYD_SOFT_RESET;\
+		type UNIPHYE_SOFT_RESET;\
+		type UNIPHYF_SOFT_RESET
 
 #define DCN20_LINK_ENCODER_REG_FIELD_LIST(type) \
 	type DIG_LANE0EN;\
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_opp.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_opp.c
index e9ebbbe256b4..0a9ad692f541 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_opp.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_opp.c
@@ -168,7 +168,10 @@ static void opp1_set_pixel_encoding(
 		REG_UPDATE(FMT_CONTROL, FMT_PIXEL_ENCODING, 0);
 		break;
 	case PIXEL_ENCODING_YCBCR422:
-		REG_UPDATE(FMT_CONTROL, FMT_PIXEL_ENCODING, 1);
+		REG_UPDATE_3(FMT_CONTROL,
+				FMT_PIXEL_ENCODING, 1,
+				FMT_SUBSAMPLING_MODE, 2,
+				FMT_CBCR_BIT_REDUCTION_BYPASS, 0);
 		break;
 	case PIXEL_ENCODING_YCBCR420:
 		REG_UPDATE(FMT_CONTROL, FMT_PIXEL_ENCODING, 2);
@@ -237,6 +240,9 @@ void opp1_set_dyn_expansion(
 			FMT_DYNAMIC_EXP_EN, 0,
 			FMT_DYNAMIC_EXP_MODE, 0);
 
+	if (opp->dyn_expansion == DYN_EXPANSION_DISABLE)
+		return;
+
 	/*00 - 10-bit -> 12-bit dynamic expansion*/
 	/*01 - 8-bit  -> 12-bit dynamic expansion*/
 	if (signal == SIGNAL_TYPE_HDMI_TYPE_A ||
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_opp.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_opp.h
index 0f10adea000c..2c0ecfa5a643 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_opp.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_opp.h
@@ -116,6 +116,8 @@
 	type FMT_RAND_G_SEED; \
 	type FMT_RAND_B_SEED; \
 	type FMT_PIXEL_ENCODING; \
+	type FMT_SUBSAMPLING_MODE; \
+	type FMT_CBCR_BIT_REDUCTION_BYPASS; \
 	type FMT_CLAMP_DATA_EN; \
 	type FMT_CLAMP_COLOR_FORMAT; \
 	type FMT_DYNAMIC_EXP_EN; \
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c
index e74a07d03fde..dabccbd49ad4 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c
@@ -1230,59 +1230,25 @@ bool optc1_is_stereo_left_eye(struct timing_generator *optc)
 	return ret;
 }
 
-bool optc1_is_matching_timing(struct timing_generator *tg,
-		const struct dc_crtc_timing *otg_timing)
+bool optc1_get_hw_timing(struct timing_generator *tg,
+		struct dc_crtc_timing *hw_crtc_timing)
 {
-	struct dc_crtc_timing hw_crtc_timing = {0};
 	struct dcn_otg_state s = {0};
 
-	if (tg == NULL || otg_timing == NULL)
+	if (tg == NULL || hw_crtc_timing == NULL)
 		return false;
 
 	optc1_read_otg_state(DCN10TG_FROM_TG(tg), &s);
 
-	hw_crtc_timing.h_total = s.h_total + 1;
-	hw_crtc_timing.h_addressable = s.h_total - ((s.h_total - s.h_blank_start) + s.h_blank_end);
-	hw_crtc_timing.h_front_porch = s.h_total + 1 - s.h_blank_start;
-	hw_crtc_timing.h_sync_width = s.h_sync_a_end - s.h_sync_a_start;
+	hw_crtc_timing->h_total = s.h_total + 1;
+	hw_crtc_timing->h_addressable = s.h_total - ((s.h_total - s.h_blank_start) + s.h_blank_end);
+	hw_crtc_timing->h_front_porch = s.h_total + 1 - s.h_blank_start;
+	hw_crtc_timing->h_sync_width = s.h_sync_a_end - s.h_sync_a_start;
 
-	hw_crtc_timing.v_total = s.v_total + 1;
-	hw_crtc_timing.v_addressable = s.v_total - ((s.v_total - s.v_blank_start) + s.v_blank_end);
-	hw_crtc_timing.v_front_porch = s.v_total + 1 - s.v_blank_start;
-	hw_crtc_timing.v_sync_width = s.v_sync_a_end - s.v_sync_a_start;
-
-	if (otg_timing->h_total != hw_crtc_timing.h_total)
-		return false;
-
-	if (otg_timing->h_border_left != hw_crtc_timing.h_border_left)
-		return false;
-
-	if (otg_timing->h_addressable != hw_crtc_timing.h_addressable)
-		return false;
-
-	if (otg_timing->h_border_right != hw_crtc_timing.h_border_right)
-		return false;
-
-	if (otg_timing->h_front_porch != hw_crtc_timing.h_front_porch)
-		return false;
-
-	if (otg_timing->h_sync_width != hw_crtc_timing.h_sync_width)
-		return false;
-
-	if (otg_timing->v_total != hw_crtc_timing.v_total)
-		return false;
-
-	if (otg_timing->v_border_top != hw_crtc_timing.v_border_top)
-		return false;
-
-	if (otg_timing->v_addressable != hw_crtc_timing.v_addressable)
-		return false;
-
-	if (otg_timing->v_border_bottom != hw_crtc_timing.v_border_bottom)
-		return false;
-
-	if (otg_timing->v_sync_width != hw_crtc_timing.v_sync_width)
-		return false;
+	hw_crtc_timing->v_total = s.v_total + 1;
+	hw_crtc_timing->v_addressable = s.v_total - ((s.v_total - s.v_blank_start) + s.v_blank_end);
+	hw_crtc_timing->v_front_porch = s.v_total + 1 - s.v_blank_start;
+	hw_crtc_timing->v_sync_width = s.v_sync_a_end - s.v_sync_a_start;
 
 	return true;
 }
@@ -1486,7 +1452,6 @@ static const struct timing_generator_funcs dcn10_tg_funcs = {
 		.get_frame_count = optc1_get_vblank_counter,
 		.get_scanoutpos = optc1_get_crtc_scanoutpos,
 		.get_otg_active_size = optc1_get_otg_active_size,
-		.is_matching_timing = optc1_is_matching_timing,
 		.set_early_control = optc1_set_early_control,
 		/* used by enable_timing_synchronization. Not need for FPGA */
 		.wait_for_state = optc1_wait_for_state,
@@ -1514,7 +1479,8 @@ static const struct timing_generator_funcs dcn10_tg_funcs = {
 		.configure_crc = optc1_configure_crc,
 		.set_vtg_params = optc1_set_vtg_params,
 		.program_manual_trigger = optc1_program_manual_trigger,
-		.setup_manual_trigger = optc1_setup_manual_trigger
+		.setup_manual_trigger = optc1_setup_manual_trigger,
+		.get_hw_timing = optc1_get_hw_timing,
 };
 
 void dcn10_timing_generator_init(struct optc *optc1)
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h
index 83575599672e..c8d795b335ba 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h
@@ -547,9 +547,8 @@ struct dcn_otg_state {
 void optc1_read_otg_state(struct optc *optc1,
 		struct dcn_otg_state *s);
 
-bool optc1_is_matching_timing(
-	struct timing_generator *tg,
-	const struct dc_crtc_timing *otg_timing);
+bool optc1_get_hw_timing(struct timing_generator *tg,
+		struct dc_crtc_timing *hw_crtc_timing);
 
 bool optc1_validate_timing(
 	struct timing_generator *optc,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c
index 59305e411a66..15640aedd664 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c
@@ -319,6 +319,14 @@ static const struct dcn10_link_enc_mask le_mask = {
 		LINK_ENCODER_MASK_SH_LIST_DCN10(_MASK)
 };
 
+static const struct dce110_aux_registers_shift aux_shift = {
+	DCN10_AUX_MASK_SH_LIST(__SHIFT)
+};
+
+static const struct dce110_aux_registers_mask aux_mask = {
+	DCN10_AUX_MASK_SH_LIST(_MASK)
+};
+
 #define ipp_regs(id)\
 [id] = {\
 	IPP_REG_LIST_DCN10(id),\
@@ -471,6 +479,28 @@ static const struct dcn_hubbub_mask hubbub_mask = {
 		HUBBUB_MASK_SH_LIST_DCN10(_MASK)
 };
 
+static int map_transmitter_id_to_phy_instance(
+	enum transmitter transmitter)
+{
+	switch (transmitter) {
+	case TRANSMITTER_UNIPHY_A:
+		return 0;
+	break;
+	case TRANSMITTER_UNIPHY_B:
+		return 1;
+	break;
+	case TRANSMITTER_UNIPHY_C:
+		return 2;
+	break;
+	case TRANSMITTER_UNIPHY_D:
+		return 3;
+	break;
+	default:
+		ASSERT(0);
+		return 0;
+	}
+}
+
 #define clk_src_regs(index, pllid)\
 [index] = {\
 	CS_COMMON_REG_LIST_DCN1_0(index, pllid),\
@@ -642,7 +672,10 @@ struct dce_aux *dcn10_aux_engine_create(
 
 	dce110_aux_engine_construct(aux_engine, ctx, inst,
 				    SW_AUX_TIMEOUT_PERIOD_MULTIPLIER * AUX_TIMEOUT_PERIOD,
-				    &aux_engine_regs[inst]);
+				    &aux_engine_regs[inst],
+					&aux_mask,
+					&aux_shift,
+					ctx->dc->caps.extended_aux_timeout_support);
 
 	return &aux_engine->base;
 }
@@ -751,14 +784,18 @@ struct link_encoder *dcn10_link_encoder_create(
 {
 	struct dcn10_link_encoder *enc10 =
 		kzalloc(sizeof(struct dcn10_link_encoder), GFP_KERNEL);
+	int link_regs_id;
 
 	if (!enc10)
 		return NULL;
 
+	link_regs_id =
+		map_transmitter_id_to_phy_instance(enc_init_data->transmitter);
+
 	dcn10_link_encoder_construct(enc10,
 				      enc_init_data,
 				      &link_enc_feature,
-				      &link_enc_regs[enc_init_data->transmitter],
+				      &link_enc_regs[link_regs_id],
 				      &link_enc_aux_regs[enc_init_data->channel - 1],
 				      &link_enc_hpd_regs[enc_init_data->hpd_source],
 				      &le_shift,
@@ -786,6 +823,7 @@ struct clock_source *dcn10_clock_source_create(
 		return &clk_src->base;
 	}
 
+	kfree(clk_src);
 	BREAK_TO_DEBUGGER();
 	return NULL;
 }
@@ -1307,6 +1345,8 @@ static bool construct(
 	dc->caps.max_slave_planes = 1;
 	dc->caps.is_apu = true;
 	dc->caps.post_blend_color_processing = false;
+	dc->caps.extended_aux_timeout_support = false;
+
 	/* Raven DP PHY HBR2 eye diagram pattern is not stable. Use TP4 */
 	dc->caps.force_dp_tps4_for_cp2520 = true;
 
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.c
index 9aa258f3550b..06e5bbb4545c 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.c
@@ -1553,6 +1553,66 @@ unsigned int enc1_dig_source_otg(
 	return tg_inst;
 }
 
+bool enc1_stream_encoder_dp_get_pixel_format(
+	struct stream_encoder *enc,
+	enum dc_pixel_encoding *encoding,
+	enum dc_color_depth *depth)
+{
+	uint32_t hw_encoding = 0;
+	uint32_t hw_depth = 0;
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+
+	if (enc == NULL ||
+		encoding == NULL ||
+		depth == NULL)
+		return false;
+
+	REG_GET_2(DP_PIXEL_FORMAT,
+		DP_PIXEL_ENCODING, &hw_encoding,
+		DP_COMPONENT_DEPTH, &hw_depth);
+
+	switch (hw_depth) {
+	case DP_COMPONENT_PIXEL_DEPTH_6BPC:
+		*depth = COLOR_DEPTH_666;
+		break;
+	case DP_COMPONENT_PIXEL_DEPTH_8BPC:
+		*depth = COLOR_DEPTH_888;
+		break;
+	case DP_COMPONENT_PIXEL_DEPTH_10BPC:
+		*depth = COLOR_DEPTH_101010;
+		break;
+	case DP_COMPONENT_PIXEL_DEPTH_12BPC:
+		*depth = COLOR_DEPTH_121212;
+		break;
+	case DP_COMPONENT_PIXEL_DEPTH_16BPC:
+		*depth = COLOR_DEPTH_161616;
+		break;
+	default:
+		*depth = COLOR_DEPTH_UNDEFINED;
+		break;
+	}
+
+	switch (hw_encoding) {
+	case DP_PIXEL_ENCODING_TYPE_RGB444:
+		*encoding = PIXEL_ENCODING_RGB;
+		break;
+	case DP_PIXEL_ENCODING_TYPE_YCBCR422:
+		*encoding = PIXEL_ENCODING_YCBCR422;
+		break;
+	case DP_PIXEL_ENCODING_TYPE_YCBCR444:
+	case DP_PIXEL_ENCODING_TYPE_Y_ONLY:
+		*encoding = PIXEL_ENCODING_YCBCR444;
+		break;
+	case DP_PIXEL_ENCODING_TYPE_YCBCR420:
+		*encoding = PIXEL_ENCODING_YCBCR420;
+		break;
+	default:
+		*encoding = PIXEL_ENCODING_UNDEFINED;
+		break;
+	}
+	return true;
+}
+
 static const struct stream_encoder_funcs dcn10_str_enc_funcs = {
 	.dp_set_stream_attribute =
 		enc1_stream_encoder_dp_set_stream_attribute,
@@ -1589,6 +1649,8 @@ static const struct stream_encoder_funcs dcn10_str_enc_funcs = {
 	.dig_connect_to_otg  = enc1_dig_connect_to_otg,
 	.hdmi_reset_stream_attribute = enc1_reset_hdmi_stream_attribute,
 	.dig_source_otg = enc1_dig_source_otg,
+
+	.dp_get_pixel_format  = enc1_stream_encoder_dp_get_pixel_format,
 };
 
 void dcn10_stream_encoder_construct(
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.h
index a512cbea00d1..c9cbc21d121e 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.h
@@ -621,4 +621,9 @@ void get_audio_clock_info(
 void enc1_reset_hdmi_stream_attribute(
 	struct stream_encoder *enc);
 
+bool enc1_stream_encoder_dp_get_pixel_format(
+	struct stream_encoder *enc,
+	enum dc_pixel_encoding *encoding,
+	enum dc_color_depth *depth);
+
 #endif /* __DC_STREAM_ENCODER_DCN10_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/Makefile b/drivers/gpu/drm/amd/display/dc/dcn20/Makefile
index ddb8d5649e79..63f3bddba7da 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn20/Makefile
+++ b/drivers/gpu/drm/amd/display/dc/dcn20/Makefile
@@ -10,15 +10,20 @@ ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT
 DCN20 += dcn20_dsc.o
 endif
 
-ifneq ($(call cc-option, -mpreferred-stack-boundary=4),)
-	cc_stack_align := -mpreferred-stack-boundary=4
-else ifneq ($(call cc-option, -mstack-alignment=16),)
-	cc_stack_align := -mstack-alignment=16
-endif
+CFLAGS_$(AMDDALPATH)/dc/dcn20/dcn20_resource.o := -mhard-float -msse
 
-CFLAGS_$(AMDDALPATH)/dc/dcn20/dcn20_resource.o := -mhard-float -msse $(cc_stack_align)
+ifdef CONFIG_CC_IS_GCC
+ifeq ($(call cc-ifversion, -lt, 0701, y), y)
+IS_OLD_GCC = 1
+endif
+endif
 
-ifdef CONFIG_CC_IS_CLANG
+ifdef IS_OLD_GCC
+# Stack alignment mismatch, proceed with caution.
+# GCC < 7.1 cannot compile code using `double` and -mpreferred-stack-boundary=3
+# (8B stack alignment).
+CFLAGS_$(AMDDALPATH)/dc/dcn20/dcn20_resource.o += -mpreferred-stack-boundary=4
+else
 CFLAGS_$(AMDDALPATH)/dc/dcn20/dcn20_resource.o += -msse2
 endif
 
diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dccg.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dccg.c
index 16476ed25536..1e1151356e60 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dccg.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dccg.c
@@ -44,16 +44,12 @@
 #define DC_LOGGER \
 	dccg->ctx->logger
 
-void dccg2_update_dpp_dto(struct dccg *dccg,
-		int dpp_inst,
-		int req_dppclk,
-		bool reduce_divider_only)
+void dccg2_update_dpp_dto(struct dccg *dccg, int dpp_inst, int req_dppclk)
 {
 	struct dcn_dccg *dccg_dcn = TO_DCN_DCCG(dccg);
 
 	if (dccg->ref_dppclk && req_dppclk) {
 		int ref_dppclk = dccg->ref_dppclk;
-		int current_phase, current_modulo;
 
 		ASSERT(req_dppclk <= ref_dppclk);
 		/* need to clamp to 8 bits */
@@ -65,28 +61,9 @@ void dccg2_update_dpp_dto(struct dccg *dccg,
 			if (req_dppclk > ref_dppclk)
 				req_dppclk = ref_dppclk;
 		}
-
-		REG_GET_2(DPPCLK_DTO_PARAM[dpp_inst],
-				DPPCLK0_DTO_PHASE, &current_phase,
-				DPPCLK0_DTO_MODULO, &current_modulo);
-
-		if (reduce_divider_only) {
-			// requested phase/modulo greater than current
-			if (req_dppclk * current_modulo >= current_phase * ref_dppclk) {
-				REG_SET_2(DPPCLK_DTO_PARAM[dpp_inst], 0,
-						DPPCLK0_DTO_PHASE, req_dppclk,
-						DPPCLK0_DTO_MODULO, ref_dppclk);
-			} else {
-				REG_SET_2(DPPCLK_DTO_PARAM[dpp_inst], 0,
-						DPPCLK0_DTO_PHASE, current_phase,
-						DPPCLK0_DTO_MODULO, current_modulo);
-			}
-		} else {
-			REG_SET_2(DPPCLK_DTO_PARAM[dpp_inst], 0,
-					DPPCLK0_DTO_PHASE, req_dppclk,
-					DPPCLK0_DTO_MODULO, ref_dppclk);
-		}
-
+		REG_SET_2(DPPCLK_DTO_PARAM[dpp_inst], 0,
+				DPPCLK0_DTO_PHASE, req_dppclk,
+				DPPCLK0_DTO_MODULO, ref_dppclk);
 		REG_UPDATE(DPPCLK_DTO_CTRL,
 				DPPCLK_DTO_ENABLE[dpp_inst], 1);
 	} else {
@@ -119,32 +96,6 @@ void dccg2_get_dccg_ref_freq(struct dccg *dccg,
 
 void dccg2_init(struct dccg *dccg)
 {
-	struct dcn_dccg *dccg_dcn = TO_DCN_DCCG(dccg);
-
-	// Fallthrough intentional to program all available dpp_dto's
-	switch (dccg_dcn->base.ctx->dc->res_pool->pipe_count) {
-	case 6:
-		REG_UPDATE(DPPCLK_DTO_CTRL, DPPCLK_DTO_DB_EN[5], 1);
-		/* Fall through */
-	case 5:
-		REG_UPDATE(DPPCLK_DTO_CTRL, DPPCLK_DTO_DB_EN[4], 1);
-		/* Fall through */
-	case 4:
-		REG_UPDATE(DPPCLK_DTO_CTRL, DPPCLK_DTO_DB_EN[3], 1);
-		/* Fall through */
-	case 3:
-		REG_UPDATE(DPPCLK_DTO_CTRL, DPPCLK_DTO_DB_EN[2], 1);
-		/* Fall through */
-	case 2:
-		REG_UPDATE(DPPCLK_DTO_CTRL, DPPCLK_DTO_DB_EN[1], 1);
-		/* Fall through */
-	case 1:
-		REG_UPDATE(DPPCLK_DTO_CTRL, DPPCLK_DTO_DB_EN[0], 1);
-		break;
-	default:
-		ASSERT(false);
-		break;
-	}
 }
 
 static const struct dccg_funcs dccg2_funcs = {
diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dccg.h b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dccg.h
index 74a074a873cd..2205cb0204e7 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dccg.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dccg.h
@@ -97,7 +97,7 @@ struct dcn_dccg {
 	const struct dccg_mask *dccg_mask;
 };
 
-void dccg2_update_dpp_dto(struct dccg *dccg, int dpp_inst, int req_dppclk, bool raise_divider_only);
+void dccg2_update_dpp_dto(struct dccg *dccg, int dpp_inst, int req_dppclk);
 
 void dccg2_get_dccg_ref_freq(struct dccg *dccg,
 		unsigned int xtalin_freq_inKhz,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dpp.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dpp.c
index 2f5aade1e882..4d7e45892f08 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dpp.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dpp.c
@@ -376,13 +376,6 @@ bool dpp2_get_optimal_number_of_taps(
 		struct scaler_data *scl_data,
 		const struct scaling_taps *in_taps)
 {
-	uint32_t pixel_width;
-
-	if (scl_data->viewport.width > scl_data->recout.width)
-		pixel_width = scl_data->recout.width;
-	else
-		pixel_width = scl_data->viewport.width;
-
 	/* Some ASICs does not support  FP16 scaling, so we reject modes require this*/
 	if (scl_data->viewport.width  != scl_data->h_active &&
 		scl_data->viewport.height != scl_data->v_active &&
@@ -464,7 +457,7 @@ static struct dpp_funcs dcn20_dpp_funcs = {
 	.dpp_read_state = dpp20_read_state,
 	.dpp_reset = dpp_reset,
 	.dpp_set_scaler = dpp1_dscl_set_scaler_manual_scale,
-	.dpp_get_optimal_number_of_taps = dpp2_get_optimal_number_of_taps,
+	.dpp_get_optimal_number_of_taps = dpp1_get_optimal_number_of_taps,
 	.dpp_set_gamut_remap = dpp1_cm_set_gamut_remap,
 	.dpp_set_csc_adjustment = NULL,
 	.dpp_set_csc_default = NULL,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dpp.h b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dpp.h
index 290b2854bd2c..5b03b737b1d6 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dpp.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dpp.h
@@ -30,16 +30,20 @@
 #define TO_DCN20_DPP(dpp)\
 	container_of(dpp, struct dcn20_dpp, base)
 
-#define TF_REG_LIST_DCN20(id) \
-	TF_REG_LIST_DCN(id), \
+#define TF_REG_LIST_DCN20_COMMON_UPDATED(id) \
 	SRI(CM_BLNDGAM_LUT_WRITE_EN_MASK, CM, id), \
+	SRI(CM_BLNDGAM_RAMB_SLOPE_CNTL_B, CM, id), \
+	SRI(CM_BLNDGAM_RAMB_SLOPE_CNTL_G, CM, id), \
+	SRI(CM_BLNDGAM_RAMB_SLOPE_CNTL_R, CM, id), \
+	SRI(CM_BLNDGAM_RAMA_SLOPE_CNTL_B, CM, id), \
+	SRI(CM_BLNDGAM_RAMA_SLOPE_CNTL_G, CM, id), \
+	SRI(CM_BLNDGAM_RAMA_SLOPE_CNTL_R, CM, id)
+
+#define TF_REG_LIST_DCN20_COMMON(id) \
 	SRI(CM_BLNDGAM_CONTROL, CM, id), \
 	SRI(CM_BLNDGAM_RAMB_START_CNTL_B, CM, id), \
 	SRI(CM_BLNDGAM_RAMB_START_CNTL_G, CM, id), \
 	SRI(CM_BLNDGAM_RAMB_START_CNTL_R, CM, id), \
-	SRI(CM_BLNDGAM_RAMB_SLOPE_CNTL_B, CM, id), \
-	SRI(CM_BLNDGAM_RAMB_SLOPE_CNTL_G, CM, id), \
-	SRI(CM_BLNDGAM_RAMB_SLOPE_CNTL_R, CM, id), \
 	SRI(CM_BLNDGAM_RAMB_END_CNTL1_B, CM, id), \
 	SRI(CM_BLNDGAM_RAMB_END_CNTL2_B, CM, id), \
 	SRI(CM_BLNDGAM_RAMB_END_CNTL1_G, CM, id), \
@@ -66,9 +70,6 @@
 	SRI(CM_BLNDGAM_RAMA_START_CNTL_B, CM, id), \
 	SRI(CM_BLNDGAM_RAMA_START_CNTL_G, CM, id), \
 	SRI(CM_BLNDGAM_RAMA_START_CNTL_R, CM, id), \
-	SRI(CM_BLNDGAM_RAMA_SLOPE_CNTL_B, CM, id), \
-	SRI(CM_BLNDGAM_RAMA_SLOPE_CNTL_G, CM, id), \
-	SRI(CM_BLNDGAM_RAMA_SLOPE_CNTL_R, CM, id), \
 	SRI(CM_BLNDGAM_RAMA_END_CNTL1_B, CM, id), \
 	SRI(CM_BLNDGAM_RAMA_END_CNTL2_B, CM, id), \
 	SRI(CM_BLNDGAM_RAMA_END_CNTL1_G, CM, id), \
@@ -147,7 +148,12 @@
 	SRI(CM_SHAPER_RAMA_REGION_28_29, CM, id), \
 	SRI(CM_SHAPER_RAMA_REGION_30_31, CM, id), \
 	SRI(CM_SHAPER_RAMA_REGION_32_33, CM, id), \
-	SRI(CM_SHAPER_LUT_INDEX, CM, id), \
+	SRI(CM_SHAPER_LUT_INDEX, CM, id)
+
+#define TF_REG_LIST_DCN20(id) \
+	TF_REG_LIST_DCN(id), \
+	TF_REG_LIST_DCN20_COMMON(id), \
+	TF_REG_LIST_DCN20_COMMON_UPDATED(id), \
 	SRI(CURSOR_CONTROL, CURSOR0_, id), \
 	SRI(ALPHA_2BIT_LUT, CNVC_CFG, id), \
 	SRI(FCNV_FP_BIAS_R, CNVC_CFG, id), \
@@ -166,27 +172,41 @@
 	SRI(OBUF_MEM_PWR_CTRL, DSCL, id),\
 	SRI(DSCL_MEM_PWR_CTRL, DSCL, id)
 
-#define TF_REG_LIST_SH_MASK_DCN20(mask_sh)\
-	TF_REG_LIST_SH_MASK_DCN(mask_sh), \
+
+#define TF_REG_LIST_SH_MASK_DCN20_UPDATED(mask_sh)\
+	TF_SF(CM0_CM_BLNDGAM_RAMB_SLOPE_CNTL_B, CM_BLNDGAM_RAMB_EXP_REGION_LINEAR_SLOPE_B, mask_sh), \
+	TF_SF(CM0_CM_BLNDGAM_RAMB_SLOPE_CNTL_G, CM_BLNDGAM_RAMB_EXP_REGION_LINEAR_SLOPE_G, mask_sh), \
+	TF_SF(CM0_CM_BLNDGAM_RAMB_SLOPE_CNTL_R, CM_BLNDGAM_RAMB_EXP_REGION_LINEAR_SLOPE_R, mask_sh), \
+	TF_SF(CM0_CM_BLNDGAM_RAMB_END_CNTL1_B, CM_BLNDGAM_RAMB_EXP_REGION_END_B, mask_sh), \
+	TF_SF(CM0_CM_BLNDGAM_RAMB_END_CNTL1_G, CM_BLNDGAM_RAMB_EXP_REGION_END_G, mask_sh), \
+	TF_SF(CM0_CM_BLNDGAM_RAMB_END_CNTL1_R, CM_BLNDGAM_RAMB_EXP_REGION_END_R, mask_sh), \
+	TF_SF(CM0_CM_BLNDGAM_RAMA_SLOPE_CNTL_B, CM_BLNDGAM_RAMA_EXP_REGION_LINEAR_SLOPE_B, mask_sh), \
+	TF_SF(CM0_CM_BLNDGAM_RAMA_SLOPE_CNTL_G, CM_BLNDGAM_RAMA_EXP_REGION_LINEAR_SLOPE_G, mask_sh), \
+	TF_SF(CM0_CM_BLNDGAM_RAMA_SLOPE_CNTL_R, CM_BLNDGAM_RAMA_EXP_REGION_LINEAR_SLOPE_R, mask_sh), \
+	TF_SF(CM0_CM_BLNDGAM_RAMA_END_CNTL1_B, CM_BLNDGAM_RAMA_EXP_REGION_END_B, mask_sh), \
+	TF_SF(CM0_CM_BLNDGAM_RAMA_END_CNTL1_G, CM_BLNDGAM_RAMA_EXP_REGION_END_G, mask_sh), \
+	TF_SF(CM0_CM_BLNDGAM_RAMA_END_CNTL1_R, CM_BLNDGAM_RAMA_EXP_REGION_END_R, mask_sh), \
+	TF_SF(CM0_CM_BLNDGAM_RAMA_END_CNTL2_B, CM_BLNDGAM_RAMA_EXP_REGION_END_BASE_B, mask_sh), \
+	TF_SF(CM0_CM_BLNDGAM_RAMA_END_CNTL2_G, CM_BLNDGAM_RAMA_EXP_REGION_END_BASE_G, mask_sh), \
+	TF_SF(CM0_CM_BLNDGAM_RAMA_END_CNTL2_R, CM_BLNDGAM_RAMA_EXP_REGION_END_BASE_R, mask_sh), \
 	TF_SF(CM0_CM_BLNDGAM_CONTROL, CM_BLNDGAM_LUT_MODE, mask_sh), \
+	TF_SF(CM0_CM_BLNDGAM_LUT_WRITE_EN_MASK, CM_BLNDGAM_LUT_WRITE_EN_MASK, mask_sh), \
+	TF_SF(CM0_CM_BLNDGAM_LUT_WRITE_EN_MASK, CM_BLNDGAM_LUT_WRITE_SEL, mask_sh), \
+	TF_SF(CM0_CM_BLNDGAM_LUT_WRITE_EN_MASK, CM_BLNDGAM_CONFIG_STATUS, mask_sh), \
+	TF_SF(CM0_CM_SHAPER_CONTROL, CM_SHAPER_LUT_MODE, mask_sh)
+
+
+#define TF_REG_LIST_SH_MASK_DCN20_COMMON(mask_sh)\
+	TF_SF(CM0_CM_3DLUT_MODE, CM_3DLUT_MODE, mask_sh), \
 	TF_SF(CM0_CM_BLNDGAM_RAMB_START_CNTL_B, CM_BLNDGAM_RAMB_EXP_REGION_START_B, mask_sh), \
 	TF_SF(CM0_CM_BLNDGAM_RAMB_START_CNTL_B, CM_BLNDGAM_RAMB_EXP_REGION_START_SEGMENT_B, mask_sh), \
 	TF_SF(CM0_CM_BLNDGAM_RAMB_START_CNTL_G, CM_BLNDGAM_RAMB_EXP_REGION_START_G, mask_sh), \
 	TF_SF(CM0_CM_BLNDGAM_RAMB_START_CNTL_G, CM_BLNDGAM_RAMB_EXP_REGION_START_SEGMENT_G, mask_sh), \
 	TF_SF(CM0_CM_BLNDGAM_RAMB_START_CNTL_R, CM_BLNDGAM_RAMB_EXP_REGION_START_R, mask_sh), \
 	TF_SF(CM0_CM_BLNDGAM_RAMB_START_CNTL_R, CM_BLNDGAM_RAMB_EXP_REGION_START_SEGMENT_R, mask_sh), \
-	TF_SF(CM0_CM_BLNDGAM_RAMB_SLOPE_CNTL_B, CM_BLNDGAM_RAMB_EXP_REGION_LINEAR_SLOPE_B, mask_sh), \
-	TF_SF(CM0_CM_BLNDGAM_RAMB_SLOPE_CNTL_G, CM_BLNDGAM_RAMB_EXP_REGION_LINEAR_SLOPE_G, mask_sh), \
-	TF_SF(CM0_CM_BLNDGAM_RAMB_SLOPE_CNTL_R, CM_BLNDGAM_RAMB_EXP_REGION_LINEAR_SLOPE_R, mask_sh), \
-	TF_SF(CM0_CM_BLNDGAM_RAMB_END_CNTL1_B, CM_BLNDGAM_RAMB_EXP_REGION_END_B, mask_sh), \
 	TF_SF(CM0_CM_BLNDGAM_RAMB_END_CNTL2_B, CM_BLNDGAM_RAMB_EXP_REGION_END_SLOPE_B, mask_sh), \
-	TF_SF(CM0_CM_BLNDGAM_RAMB_END_CNTL2_B, CM_BLNDGAM_RAMB_EXP_REGION_END_BASE_B, mask_sh), \
-	TF_SF(CM0_CM_BLNDGAM_RAMB_END_CNTL1_G, CM_BLNDGAM_RAMB_EXP_REGION_END_G, mask_sh), \
 	TF_SF(CM0_CM_BLNDGAM_RAMB_END_CNTL2_G, CM_BLNDGAM_RAMB_EXP_REGION_END_SLOPE_G, mask_sh), \
-	TF_SF(CM0_CM_BLNDGAM_RAMB_END_CNTL2_G, CM_BLNDGAM_RAMB_EXP_REGION_END_BASE_G, mask_sh), \
-	TF_SF(CM0_CM_BLNDGAM_RAMB_END_CNTL1_R, CM_BLNDGAM_RAMB_EXP_REGION_END_R, mask_sh), \
 	TF_SF(CM0_CM_BLNDGAM_RAMB_END_CNTL2_R, CM_BLNDGAM_RAMB_EXP_REGION_END_SLOPE_R, mask_sh), \
-	TF_SF(CM0_CM_BLNDGAM_RAMB_END_CNTL2_R, CM_BLNDGAM_RAMB_EXP_REGION_END_BASE_R, mask_sh), \
 	TF_SF(CM0_CM_BLNDGAM_RAMB_REGION_0_1, CM_BLNDGAM_RAMB_EXP_REGION0_LUT_OFFSET, mask_sh), \
 	TF_SF(CM0_CM_BLNDGAM_RAMB_REGION_0_1, CM_BLNDGAM_RAMB_EXP_REGION0_NUM_SEGMENTS, mask_sh), \
 	TF_SF(CM0_CM_BLNDGAM_RAMB_REGION_0_1, CM_BLNDGAM_RAMB_EXP_REGION1_LUT_OFFSET, mask_sh), \
@@ -261,18 +281,9 @@
 	TF_SF(CM0_CM_BLNDGAM_RAMA_START_CNTL_G, CM_BLNDGAM_RAMA_EXP_REGION_START_SEGMENT_G, mask_sh), \
 	TF_SF(CM0_CM_BLNDGAM_RAMA_START_CNTL_R, CM_BLNDGAM_RAMA_EXP_REGION_START_R, mask_sh), \
 	TF_SF(CM0_CM_BLNDGAM_RAMA_START_CNTL_R, CM_BLNDGAM_RAMA_EXP_REGION_START_SEGMENT_R, mask_sh), \
-	TF_SF(CM0_CM_BLNDGAM_RAMA_SLOPE_CNTL_B, CM_BLNDGAM_RAMA_EXP_REGION_LINEAR_SLOPE_B, mask_sh), \
-	TF_SF(CM0_CM_BLNDGAM_RAMA_SLOPE_CNTL_G, CM_BLNDGAM_RAMA_EXP_REGION_LINEAR_SLOPE_G, mask_sh), \
-	TF_SF(CM0_CM_BLNDGAM_RAMA_SLOPE_CNTL_R, CM_BLNDGAM_RAMA_EXP_REGION_LINEAR_SLOPE_R, mask_sh), \
-	TF_SF(CM0_CM_BLNDGAM_RAMA_END_CNTL1_B, CM_BLNDGAM_RAMA_EXP_REGION_END_B, mask_sh), \
 	TF_SF(CM0_CM_BLNDGAM_RAMA_END_CNTL2_B, CM_BLNDGAM_RAMA_EXP_REGION_END_SLOPE_B, mask_sh), \
-	TF_SF(CM0_CM_BLNDGAM_RAMA_END_CNTL2_B, CM_BLNDGAM_RAMA_EXP_REGION_END_BASE_B, mask_sh), \
-	TF_SF(CM0_CM_BLNDGAM_RAMA_END_CNTL1_G, CM_BLNDGAM_RAMA_EXP_REGION_END_G, mask_sh), \
 	TF_SF(CM0_CM_BLNDGAM_RAMA_END_CNTL2_G, CM_BLNDGAM_RAMA_EXP_REGION_END_SLOPE_G, mask_sh), \
-	TF_SF(CM0_CM_BLNDGAM_RAMA_END_CNTL2_G, CM_BLNDGAM_RAMA_EXP_REGION_END_BASE_G, mask_sh), \
-	TF_SF(CM0_CM_BLNDGAM_RAMA_END_CNTL1_R, CM_BLNDGAM_RAMA_EXP_REGION_END_R, mask_sh), \
 	TF_SF(CM0_CM_BLNDGAM_RAMA_END_CNTL2_R, CM_BLNDGAM_RAMA_EXP_REGION_END_SLOPE_R, mask_sh), \
-	TF_SF(CM0_CM_BLNDGAM_RAMA_END_CNTL2_R, CM_BLNDGAM_RAMA_EXP_REGION_END_BASE_R, mask_sh), \
 	TF_SF(CM0_CM_BLNDGAM_RAMA_REGION_0_1, CM_BLNDGAM_RAMA_EXP_REGION0_LUT_OFFSET, mask_sh), \
 	TF_SF(CM0_CM_BLNDGAM_RAMA_REGION_0_1, CM_BLNDGAM_RAMA_EXP_REGION0_NUM_SEGMENTS, mask_sh), \
 	TF_SF(CM0_CM_BLNDGAM_RAMA_REGION_0_1, CM_BLNDGAM_RAMA_EXP_REGION1_LUT_OFFSET, mask_sh), \
@@ -341,9 +352,6 @@
 	TF_SF(CM0_CM_BLNDGAM_RAMA_REGION_32_33, CM_BLNDGAM_RAMA_EXP_REGION32_NUM_SEGMENTS, mask_sh), \
 	TF_SF(CM0_CM_BLNDGAM_RAMA_REGION_32_33, CM_BLNDGAM_RAMA_EXP_REGION33_LUT_OFFSET, mask_sh), \
 	TF_SF(CM0_CM_BLNDGAM_RAMA_REGION_32_33, CM_BLNDGAM_RAMA_EXP_REGION33_NUM_SEGMENTS, mask_sh), \
-	TF_SF(CM0_CM_BLNDGAM_LUT_WRITE_EN_MASK, CM_BLNDGAM_LUT_WRITE_EN_MASK, mask_sh), \
-	TF_SF(CM0_CM_BLNDGAM_LUT_WRITE_EN_MASK, CM_BLNDGAM_LUT_WRITE_SEL, mask_sh), \
-	TF_SF(CM0_CM_BLNDGAM_LUT_WRITE_EN_MASK, CM_BLNDGAM_CONFIG_STATUS, mask_sh), \
 	TF_SF(CM0_CM_BLNDGAM_LUT_INDEX, CM_BLNDGAM_LUT_INDEX, mask_sh), \
 	TF_SF(CM0_CM_BLNDGAM_LUT_DATA, CM_BLNDGAM_LUT_DATA, mask_sh), \
 	TF_SF(CM0_CM_MEM_PWR_CTRL, BLNDGAM_MEM_PWR_FORCE, mask_sh), \
@@ -356,7 +364,6 @@
 	TF_SF(CM0_CM_3DLUT_READ_WRITE_CONTROL, CM_3DLUT_WRITE_EN_MASK, mask_sh), \
 	TF_SF(CM0_CM_3DLUT_READ_WRITE_CONTROL, CM_3DLUT_RAM_SEL, mask_sh), \
 	TF_SF(CM0_CM_3DLUT_READ_WRITE_CONTROL, CM_3DLUT_30BIT_EN, mask_sh), \
-	TF_SF(CM0_CM_3DLUT_READ_WRITE_CONTROL, CM_3DLUT_CONFIG_STATUS, mask_sh), \
 	TF_SF(CM0_CM_3DLUT_READ_WRITE_CONTROL, CM_3DLUT_READ_SEL, mask_sh), \
 	TF_SF(CM0_CM_SHAPER_CONTROL, CM_SHAPER_LUT_MODE, mask_sh), \
 	TF_SF(CM0_CM_SHAPER_RAMB_START_CNTL_B, CM_SHAPER_RAMB_EXP_REGION_START_B, mask_sh), \
@@ -521,9 +528,14 @@
 	TF_SF(CM0_CM_SHAPER_RAMA_REGION_32_33, CM_SHAPER_RAMA_EXP_REGION33_NUM_SEGMENTS, mask_sh), \
 	TF_SF(CM0_CM_SHAPER_LUT_WRITE_EN_MASK, CM_SHAPER_LUT_WRITE_EN_MASK, mask_sh), \
 	TF_SF(CM0_CM_SHAPER_LUT_WRITE_EN_MASK, CM_SHAPER_LUT_WRITE_SEL, mask_sh), \
-	TF_SF(CM0_CM_SHAPER_LUT_WRITE_EN_MASK, CM_SHAPER_CONFIG_STATUS, mask_sh), \
 	TF_SF(CM0_CM_SHAPER_LUT_INDEX, CM_SHAPER_LUT_INDEX, mask_sh), \
-	TF_SF(CM0_CM_SHAPER_LUT_DATA, CM_SHAPER_LUT_DATA, mask_sh), \
+	TF_SF(CM0_CM_SHAPER_LUT_DATA, CM_SHAPER_LUT_DATA, mask_sh)
+
+
+#define TF_REG_LIST_SH_MASK_DCN20(mask_sh)\
+	TF_REG_LIST_SH_MASK_DCN(mask_sh), \
+	TF_REG_LIST_SH_MASK_DCN20_COMMON(mask_sh), \
+	TF_REG_LIST_SH_MASK_DCN20_UPDATED(mask_sh), \
 	TF_SF(CM0_CM_DGAM_LUT_WRITE_EN_MASK, CM_DGAM_CONFIG_STATUS, mask_sh), \
 	TF_SF(CM0_CM_CONTROL, CM_BYPASS, mask_sh), \
 	TF_SF(CURSOR0_0_CURSOR_CONTROL, CURSOR_MODE, mask_sh), \
@@ -560,6 +572,7 @@
 	TF_SF(DSCL0_OBUF_MEM_PWR_CTRL, OBUF_MEM_PWR_FORCE, mask_sh),\
 	TF_SF(DSCL0_DSCL_MEM_PWR_CTRL, LUT_MEM_PWR_FORCE, mask_sh)
 
+
 #define TF_REG_FIELD_LIST_DCN2_0(type) \
 	TF_REG_FIELD_LIST(type) \
 	type CM_BLNDGAM_LUT_DATA; \
@@ -593,6 +606,7 @@
 	type OBUF_MEM_PWR_FORCE;\
 	type LUT_MEM_PWR_FORCE
 
+
 struct dcn2_dpp_shift {
 	TF_REG_FIELD_LIST_DCN2_0(uint8_t);
 };
@@ -691,11 +705,6 @@ void dpp2_set_hdr_multiplier(
 		struct dpp *dpp_base,
 		uint32_t multiplier);
 
-bool dpp2_get_optimal_number_of_taps(
-		struct dpp *dpp,
-		struct scaler_data *scl_data,
-		const struct scaling_taps *in_taps);
-
 bool dpp2_construct(struct dcn20_dpp *dpp2,
 	struct dc_context *ctx,
 	uint32_t inst,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dsc.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dsc.c
index 1b419407af94..63eb377ed9c0 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dsc.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dsc.c
@@ -118,7 +118,7 @@ static void dsc2_get_enc_caps(struct dsc_enc_caps *dsc_enc_caps, int pixel_clock
 
 	dsc_enc_caps->color_formats.bits.RGB = 1;
 	dsc_enc_caps->color_formats.bits.YCBCR_444 = 1;
-	dsc_enc_caps->color_formats.bits.YCBCR_SIMPLE_422 = 0;
+	dsc_enc_caps->color_formats.bits.YCBCR_SIMPLE_422 = 1;
 	dsc_enc_caps->color_formats.bits.YCBCR_NATIVE_422 = 0;
 	dsc_enc_caps->color_formats.bits.YCBCR_NATIVE_420 = 1;
 
diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dwb_scl.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dwb_scl.c
index cd8bc92ce3ba..880954ac0b02 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dwb_scl.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dwb_scl.c
@@ -722,7 +722,6 @@ bool dwb_program_horz_scalar(struct dcn20_dwbc *dwbc20,
 		struct scaling_taps num_taps)
 {
 	uint32_t h_ratio_luma = 1;
-	uint32_t h_ratio_chroma = 1;
 	uint32_t h_taps_luma = num_taps.h_taps;
 	uint32_t h_taps_chroma = num_taps.h_taps_c;
 	int32_t h_init_phase_luma = 0;
@@ -747,7 +746,6 @@ bool dwb_program_horz_scalar(struct dcn20_dwbc *dwbc20,
 		h_ratio_luma = -1;
 	else
 		h_ratio_luma = dc_fixpt_u3d19(tmp_h_ratio_luma) << 5;
-	h_ratio_chroma = h_ratio_luma * 2;
 
 	/*Program ratio*/
 	REG_UPDATE(WBSCL_HORZ_FILTER_SCALE_RATIO, WBSCL_H_SCALE_RATIO, h_ratio_luma);
@@ -803,7 +801,6 @@ bool dwb_program_vert_scalar(struct dcn20_dwbc *dwbc20,
 		enum dwb_subsample_position subsample_position)
 {
 	uint32_t v_ratio_luma = 1;
-	uint32_t v_ratio_chroma = 1;
 	uint32_t v_taps_luma = num_taps.v_taps;
 	uint32_t v_taps_chroma = num_taps.v_taps_c;
 	int32_t v_init_phase_luma = 0;
@@ -827,7 +824,6 @@ bool dwb_program_vert_scalar(struct dcn20_dwbc *dwbc20,
 		v_ratio_luma = -1;
 	else
 		v_ratio_luma = dc_fixpt_u3d19(tmp_v_ratio_luma) << 5;
-	v_ratio_chroma = v_ratio_luma * 2;
 
 	/*Program ratio*/
 	REG_UPDATE(WBSCL_VERT_FILTER_SCALE_RATIO, WBSCL_V_SCALE_RATIO, v_ratio_luma);
diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubbub.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubbub.c
index b83c022e2c6f..8b8438566101 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubbub.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubbub.c
@@ -186,14 +186,13 @@ static void hubbub2_get_blk256_size(unsigned int *blk256_width, unsigned int *bl
 }
 
 static void hubbub2_det_request_size(
+		unsigned int detile_buf_size,
 		unsigned int height,
 		unsigned int width,
 		unsigned int bpe,
 		bool *req128_horz_wc,
 		bool *req128_vert_wc)
 {
-	unsigned int detile_buf_size = 164 * 1024;  /* 164KB for DCN1.0 */
-
 	unsigned int blk256_height = 0;
 	unsigned int blk256_width = 0;
 	unsigned int swath_bytes_horz_wc, swath_bytes_vert_wc;
@@ -236,7 +235,8 @@ bool hubbub2_get_dcc_compression_cap(struct hubbub *hubbub,
 			&segment_order_horz, &segment_order_vert))
 		return false;
 
-	hubbub2_det_request_size(input->surface_size.height,  input->surface_size.width,
+	hubbub2_det_request_size(TO_DCN20_HUBBUB(hubbub)->detile_buf_size,
+			input->surface_size.height,  input->surface_size.width,
 			bpe, &req128_horz_wc, &req128_vert_wc);
 
 	if (!req128_horz_wc && !req128_vert_wc) {
@@ -588,7 +588,7 @@ static void hubbub2_program_watermarks(
 			DCHUBBUB_ARB_SAT_LEVEL, 60 * refclk_mhz);
 	REG_UPDATE(DCHUBBUB_ARB_DF_REQ_OUTSTAND, DCHUBBUB_ARB_MIN_REQ_OUTSTAND, 180);
 
-	hubbub1_allow_self_refresh_control(hubbub, !hubbub->ctx->dc->debug.disable_stutter);
+	hubbub->funcs->allow_self_refresh_control(hubbub, !hubbub->ctx->dc->debug.disable_stutter);
 }
 
 static const struct hubbub_funcs hubbub2_funcs = {
@@ -600,7 +600,8 @@ static const struct hubbub_funcs hubbub2_funcs = {
 	.get_dcc_compression_cap = hubbub2_get_dcc_compression_cap,
 	.wm_read_state = hubbub2_wm_read_state,
 	.get_dchub_ref_freq = hubbub2_get_dchub_ref_freq,
-	.program_watermarks = hubbub2_program_watermarks
+	.program_watermarks = hubbub2_program_watermarks,
+	.allow_self_refresh_control = hubbub1_allow_self_refresh_control
 };
 
 void hubbub2_construct(struct dcn20_hubbub *hubbub,
@@ -618,4 +619,5 @@ void hubbub2_construct(struct dcn20_hubbub *hubbub,
 	hubbub->masks = hubbub_mask;
 
 	hubbub->debug_test_index_pstate = 0xB;
+	hubbub->detile_buf_size = 164 * 1024; /* 164KB for DCN2.0 */
 }
diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubbub.h b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubbub.h
index 626117d3b4e9..501532dd523a 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubbub.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubbub.h
@@ -81,6 +81,7 @@ struct dcn20_hubbub {
 	unsigned int debug_test_index_pstate;
 	struct dcn_watermark_set watermarks;
 	struct dcn20_vmid vmid[16];
+	unsigned int detile_buf_size;
 };
 
 void hubbub2_construct(struct dcn20_hubbub *hubbub,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c
index 1212da12c414..921a36668ced 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c
@@ -688,7 +688,7 @@ bool dcn20_set_output_transfer_func(struct pipe_ctx *pipe_ctx,
 	return true;
 }
 
-static bool dcn20_set_blend_lut(
+bool dcn20_set_blend_lut(
 	struct pipe_ctx *pipe_ctx, const struct dc_plane_state *plane_state)
 {
 	struct dpp *dpp_base = pipe_ctx->plane_res.dpp;
@@ -710,7 +710,7 @@ static bool dcn20_set_blend_lut(
 	return result;
 }
 
-static bool dcn20_set_shaper_3dlut(
+bool dcn20_set_shaper_3dlut(
 	struct pipe_ctx *pipe_ctx, const struct dc_plane_state *plane_state)
 {
 	struct dpp *dpp_base = pipe_ctx->plane_res.dpp;
@@ -999,72 +999,6 @@ void dcn20_enable_plane(
 }
 
 
-static void dcn20_program_pipe(
-		struct dc *dc,
-		struct pipe_ctx *pipe_ctx,
-		struct dc_state *context)
-{
-	pipe_ctx->plane_state->update_flags.bits.full_update =
-			context->commit_hints.full_update_needed ? 1 : pipe_ctx->plane_state->update_flags.bits.full_update;
-
-	if (pipe_ctx->plane_state->update_flags.bits.full_update)
-		dcn20_enable_plane(dc, pipe_ctx, context);
-
-	update_dchubp_dpp(dc, pipe_ctx, context);
-
-	set_hdr_multiplier(pipe_ctx);
-
-	if (pipe_ctx->plane_state->update_flags.bits.full_update ||
-			pipe_ctx->plane_state->update_flags.bits.in_transfer_func_change ||
-			pipe_ctx->plane_state->update_flags.bits.gamma_change)
-		dc->hwss.set_input_transfer_func(pipe_ctx, pipe_ctx->plane_state);
-
-	/* dcn10_translate_regamma_to_hw_format takes 750us to finish
-	 * only do gamma programming for full update.
-	 * TODO: This can be further optimized/cleaned up
-	 * Always call this for now since it does memcmp inside before
-	 * doing heavy calculation and programming
-	 */
-	if (pipe_ctx->plane_state->update_flags.bits.full_update)
-		dc->hwss.set_output_transfer_func(pipe_ctx, pipe_ctx->stream);
-}
-
-static void dcn20_program_all_pipe_in_tree(
-		struct dc *dc,
-		struct pipe_ctx *pipe_ctx,
-		struct dc_state *context)
-{
-	if (pipe_ctx->top_pipe == NULL && !pipe_ctx->prev_odm_pipe) {
-		bool blank = !is_pipe_tree_visible(pipe_ctx);
-
-		pipe_ctx->stream_res.tg->funcs->program_global_sync(
-				pipe_ctx->stream_res.tg,
-				pipe_ctx->pipe_dlg_param.vready_offset,
-				pipe_ctx->pipe_dlg_param.vstartup_start,
-				pipe_ctx->pipe_dlg_param.vupdate_offset,
-				pipe_ctx->pipe_dlg_param.vupdate_width);
-
-		pipe_ctx->stream_res.tg->funcs->set_vtg_params(
-				pipe_ctx->stream_res.tg, &pipe_ctx->stream->timing);
-
-		dc->hwss.blank_pixel_data(dc, pipe_ctx, blank);
-
-		if (dc->hwss.update_odm)
-			dc->hwss.update_odm(dc, context, pipe_ctx);
-	}
-
-	if (pipe_ctx->plane_state != NULL)
-		dcn20_program_pipe(dc, pipe_ctx, context);
-
-	if (pipe_ctx->bottom_pipe != NULL) {
-		ASSERT(pipe_ctx->bottom_pipe != pipe_ctx);
-		dcn20_program_all_pipe_in_tree(dc, pipe_ctx->bottom_pipe, context);
-	} else if (pipe_ctx->next_odm_pipe != NULL) {
-		ASSERT(pipe_ctx->next_odm_pipe != pipe_ctx);
-		dcn20_program_all_pipe_in_tree(dc, pipe_ctx->next_odm_pipe, context);
-	}
-}
-
 void dcn20_pipe_control_lock_global(
 		struct dc *dc,
 		struct pipe_ctx *pipe,
@@ -1124,114 +1058,456 @@ void dcn20_pipe_control_lock(
 	}
 }
 
-static void dcn20_apply_ctx_for_surface(
-		struct dc *dc,
-		const struct dc_stream_state *stream,
-		int num_planes,
-		struct dc_state *context)
+static void dcn20_detect_pipe_changes(struct pipe_ctx *old_pipe, struct pipe_ctx *new_pipe)
 {
-	const unsigned int TIMEOUT_FOR_PIPE_ENABLE_MS = 100;
-	int i;
-	struct timing_generator *tg;
-	bool removed_pipe[6] = { false };
-	bool interdependent_update = false;
-	struct pipe_ctx *top_pipe_to_program =
-			find_top_pipe_for_stream(dc, context, stream);
-	struct pipe_ctx *prev_top_pipe_to_program =
-			find_top_pipe_for_stream(dc, dc->current_state, stream);
-	DC_LOGGER_INIT(dc->ctx->logger);
+	new_pipe->update_flags.raw = 0;
 
-	if (!top_pipe_to_program)
+	/* Exit on unchanged, unused pipe */
+	if (!old_pipe->plane_state && !new_pipe->plane_state)
 		return;
+	/* Detect pipe enable/disable */
+	if (!old_pipe->plane_state && new_pipe->plane_state) {
+		new_pipe->update_flags.bits.enable = 1;
+		new_pipe->update_flags.bits.mpcc = 1;
+		new_pipe->update_flags.bits.dppclk = 1;
+		new_pipe->update_flags.bits.hubp_interdependent = 1;
+		new_pipe->update_flags.bits.hubp_rq_dlg_ttu = 1;
+		new_pipe->update_flags.bits.gamut_remap = 1;
+		new_pipe->update_flags.bits.scaler = 1;
+		new_pipe->update_flags.bits.viewport = 1;
+		if (!new_pipe->top_pipe && !new_pipe->prev_odm_pipe) {
+			new_pipe->update_flags.bits.odm = 1;
+			new_pipe->update_flags.bits.global_sync = 1;
+		}
+		return;
+	}
+	if (old_pipe->plane_state && !new_pipe->plane_state) {
+		new_pipe->update_flags.bits.disable = 1;
+		return;
+	}
 
-	/* Carry over GSL groups in case the context is changing. */
-	for (i = 0; i < dc->res_pool->pipe_count; i++) {
-		struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i];
-		struct pipe_ctx *old_pipe_ctx =
-			&dc->current_state->res_ctx.pipe_ctx[i];
-
-		if (pipe_ctx->stream == stream &&
-		    pipe_ctx->stream == old_pipe_ctx->stream)
-			pipe_ctx->stream_res.gsl_group =
-				old_pipe_ctx->stream_res.gsl_group;
+	/* Detect top pipe only changes */
+	if (!new_pipe->top_pipe && !new_pipe->prev_odm_pipe) {
+		/* Detect odm changes */
+		if ((old_pipe->next_odm_pipe && new_pipe->next_odm_pipe
+			&& old_pipe->next_odm_pipe->pipe_idx != new_pipe->next_odm_pipe->pipe_idx)
+				|| (!old_pipe->next_odm_pipe && new_pipe->next_odm_pipe)
+				|| (old_pipe->next_odm_pipe && !new_pipe->next_odm_pipe)
+				|| old_pipe->stream_res.opp != new_pipe->stream_res.opp)
+			new_pipe->update_flags.bits.odm = 1;
+
+		/* Detect global sync changes */
+		if (old_pipe->pipe_dlg_param.vready_offset != new_pipe->pipe_dlg_param.vready_offset
+				|| old_pipe->pipe_dlg_param.vstartup_start != new_pipe->pipe_dlg_param.vstartup_start
+				|| old_pipe->pipe_dlg_param.vupdate_offset != new_pipe->pipe_dlg_param.vupdate_offset
+				|| old_pipe->pipe_dlg_param.vupdate_width != new_pipe->pipe_dlg_param.vupdate_width)
+			new_pipe->update_flags.bits.global_sync = 1;
 	}
 
-	tg = top_pipe_to_program->stream_res.tg;
+	/*
+	 * Detect opp / tg change, only set on change, not on enable
+	 * Assume mpcc inst = pipe index, if not this code needs to be updated
+	 * since mpcc is what is affected by these. In fact all of our sequence
+	 * makes this assumption at the moment with how hubp reset is matched to
+	 * same index mpcc reset.
+	 */
+	if (old_pipe->stream_res.opp != new_pipe->stream_res.opp)
+		new_pipe->update_flags.bits.opp_changed = 1;
+	if (old_pipe->stream_res.tg != new_pipe->stream_res.tg)
+		new_pipe->update_flags.bits.tg_changed = 1;
+
+	/* Detect mpcc blending changes, only dpp inst and bot matter here */
+	if (old_pipe->plane_res.dpp != new_pipe->plane_res.dpp
+			|| old_pipe->stream_res.opp != new_pipe->stream_res.opp
+			|| (!old_pipe->bottom_pipe && new_pipe->bottom_pipe)
+			|| (old_pipe->bottom_pipe && !new_pipe->bottom_pipe)
+			|| (old_pipe->bottom_pipe && new_pipe->bottom_pipe
+				&& old_pipe->bottom_pipe->plane_res.mpcc_inst
+					!= new_pipe->bottom_pipe->plane_res.mpcc_inst))
+		new_pipe->update_flags.bits.mpcc = 1;
+
+	/* Detect dppclk change */
+	if (old_pipe->plane_res.bw.dppclk_khz != new_pipe->plane_res.bw.dppclk_khz)
+		new_pipe->update_flags.bits.dppclk = 1;
+
+	/* Check for scl update */
+	if (memcmp(&old_pipe->plane_res.scl_data, &new_pipe->plane_res.scl_data, sizeof(struct scaler_data)))
+			new_pipe->update_flags.bits.scaler = 1;
+	/* Check for vp update */
+	if (memcmp(&old_pipe->plane_res.scl_data.viewport, &new_pipe->plane_res.scl_data.viewport, sizeof(struct rect))
+			|| memcmp(&old_pipe->plane_res.scl_data.viewport_c,
+				&new_pipe->plane_res.scl_data.viewport_c, sizeof(struct rect)))
+		new_pipe->update_flags.bits.viewport = 1;
+
+	/* Detect dlg/ttu/rq updates */
+	{
+		struct _vcs_dpi_display_dlg_regs_st old_dlg_attr = old_pipe->dlg_regs;
+		struct _vcs_dpi_display_ttu_regs_st old_ttu_attr = old_pipe->ttu_regs;
+		struct _vcs_dpi_display_dlg_regs_st *new_dlg_attr = &new_pipe->dlg_regs;
+		struct _vcs_dpi_display_ttu_regs_st *new_ttu_attr = &new_pipe->ttu_regs;
+
+		/* Detect pipe interdependent updates */
+		if (old_dlg_attr.dst_y_prefetch != new_dlg_attr->dst_y_prefetch ||
+				old_dlg_attr.vratio_prefetch != new_dlg_attr->vratio_prefetch ||
+				old_dlg_attr.vratio_prefetch_c != new_dlg_attr->vratio_prefetch_c ||
+				old_dlg_attr.dst_y_per_vm_vblank != new_dlg_attr->dst_y_per_vm_vblank ||
+				old_dlg_attr.dst_y_per_row_vblank != new_dlg_attr->dst_y_per_row_vblank ||
+				old_dlg_attr.dst_y_per_vm_flip != new_dlg_attr->dst_y_per_vm_flip ||
+				old_dlg_attr.dst_y_per_row_flip != new_dlg_attr->dst_y_per_row_flip ||
+				old_dlg_attr.refcyc_per_meta_chunk_vblank_l != new_dlg_attr->refcyc_per_meta_chunk_vblank_l ||
+				old_dlg_attr.refcyc_per_meta_chunk_vblank_c != new_dlg_attr->refcyc_per_meta_chunk_vblank_c ||
+				old_dlg_attr.refcyc_per_meta_chunk_flip_l != new_dlg_attr->refcyc_per_meta_chunk_flip_l ||
+				old_dlg_attr.refcyc_per_line_delivery_pre_l != new_dlg_attr->refcyc_per_line_delivery_pre_l ||
+				old_dlg_attr.refcyc_per_line_delivery_pre_c != new_dlg_attr->refcyc_per_line_delivery_pre_c ||
+				old_ttu_attr.refcyc_per_req_delivery_pre_l != new_ttu_attr->refcyc_per_req_delivery_pre_l ||
+				old_ttu_attr.refcyc_per_req_delivery_pre_c != new_ttu_attr->refcyc_per_req_delivery_pre_c ||
+				old_ttu_attr.refcyc_per_req_delivery_pre_cur0 != new_ttu_attr->refcyc_per_req_delivery_pre_cur0 ||
+				old_ttu_attr.refcyc_per_req_delivery_pre_cur1 != new_ttu_attr->refcyc_per_req_delivery_pre_cur1 ||
+				old_ttu_attr.min_ttu_vblank != new_ttu_attr->min_ttu_vblank ||
+				old_ttu_attr.qos_level_flip != new_ttu_attr->qos_level_flip) {
+			old_dlg_attr.dst_y_prefetch = new_dlg_attr->dst_y_prefetch;
+			old_dlg_attr.vratio_prefetch = new_dlg_attr->vratio_prefetch;
+			old_dlg_attr.vratio_prefetch_c = new_dlg_attr->vratio_prefetch_c;
+			old_dlg_attr.dst_y_per_vm_vblank = new_dlg_attr->dst_y_per_vm_vblank;
+			old_dlg_attr.dst_y_per_row_vblank = new_dlg_attr->dst_y_per_row_vblank;
+			old_dlg_attr.dst_y_per_vm_flip = new_dlg_attr->dst_y_per_vm_flip;
+			old_dlg_attr.dst_y_per_row_flip = new_dlg_attr->dst_y_per_row_flip;
+			old_dlg_attr.refcyc_per_meta_chunk_vblank_l = new_dlg_attr->refcyc_per_meta_chunk_vblank_l;
+			old_dlg_attr.refcyc_per_meta_chunk_vblank_c = new_dlg_attr->refcyc_per_meta_chunk_vblank_c;
+			old_dlg_attr.refcyc_per_meta_chunk_flip_l = new_dlg_attr->refcyc_per_meta_chunk_flip_l;
+			old_dlg_attr.refcyc_per_line_delivery_pre_l = new_dlg_attr->refcyc_per_line_delivery_pre_l;
+			old_dlg_attr.refcyc_per_line_delivery_pre_c = new_dlg_attr->refcyc_per_line_delivery_pre_c;
+			old_ttu_attr.refcyc_per_req_delivery_pre_l = new_ttu_attr->refcyc_per_req_delivery_pre_l;
+			old_ttu_attr.refcyc_per_req_delivery_pre_c = new_ttu_attr->refcyc_per_req_delivery_pre_c;
+			old_ttu_attr.refcyc_per_req_delivery_pre_cur0 = new_ttu_attr->refcyc_per_req_delivery_pre_cur0;
+			old_ttu_attr.refcyc_per_req_delivery_pre_cur1 = new_ttu_attr->refcyc_per_req_delivery_pre_cur1;
+			old_ttu_attr.min_ttu_vblank = new_ttu_attr->min_ttu_vblank;
+			old_ttu_attr.qos_level_flip = new_ttu_attr->qos_level_flip;
+			new_pipe->update_flags.bits.hubp_interdependent = 1;
+		}
+		/* Detect any other updates to ttu/rq/dlg */
+		if (memcmp(&old_dlg_attr, &new_pipe->dlg_regs, sizeof(old_dlg_attr)) ||
+				memcmp(&old_ttu_attr, &new_pipe->ttu_regs, sizeof(old_ttu_attr)) ||
+				memcmp(&old_pipe->rq_regs, &new_pipe->rq_regs, sizeof(old_pipe->rq_regs)))
+			new_pipe->update_flags.bits.hubp_rq_dlg_ttu = 1;
+	}
+}
 
-	interdependent_update = top_pipe_to_program->plane_state &&
-		top_pipe_to_program->plane_state->update_flags.bits.full_update;
+static void dcn20_update_dchubp_dpp(
+	struct dc *dc,
+	struct pipe_ctx *pipe_ctx,
+	struct dc_state *context)
+{
+	struct hubp *hubp = pipe_ctx->plane_res.hubp;
+	struct dpp *dpp = pipe_ctx->plane_res.dpp;
+	struct dc_plane_state *plane_state = pipe_ctx->plane_state;
 
-	if (interdependent_update)
-		lock_all_pipes(dc, context, true);
-	else
-		dcn20_pipe_control_lock(dc, top_pipe_to_program, true);
+	if (pipe_ctx->update_flags.bits.dppclk)
+		dpp->funcs->dpp_dppclk_control(dpp, false, true);
 
-	if (num_planes == 0) {
-		/* OTG blank before remove all front end */
-		dc->hwss.blank_pixel_data(dc, top_pipe_to_program, true);
+	/* TODO: Need input parameter to tell current DCHUB pipe tie to which OTG
+	 * VTG is within DCHUBBUB which is commond block share by each pipe HUBP.
+	 * VTG is 1:1 mapping with OTG. Each pipe HUBP will select which VTG
+	 */
+	if (pipe_ctx->update_flags.bits.hubp_rq_dlg_ttu) {
+		hubp->funcs->hubp_vtg_sel(hubp, pipe_ctx->stream_res.tg->inst);
+
+		hubp->funcs->hubp_setup(
+			hubp,
+			&pipe_ctx->dlg_regs,
+			&pipe_ctx->ttu_regs,
+			&pipe_ctx->rq_regs,
+			&pipe_ctx->pipe_dlg_param);
+	}
+	if (pipe_ctx->update_flags.bits.hubp_interdependent)
+		hubp->funcs->hubp_setup_interdependent(
+			hubp,
+			&pipe_ctx->dlg_regs,
+			&pipe_ctx->ttu_regs);
+
+	if (pipe_ctx->update_flags.bits.enable ||
+			plane_state->update_flags.bits.bpp_change ||
+			plane_state->update_flags.bits.input_csc_change ||
+			plane_state->update_flags.bits.color_space_change ||
+			plane_state->update_flags.bits.coeff_reduction_change) {
+		struct dc_bias_and_scale bns_params = {0};
+
+		// program the input csc
+		dpp->funcs->dpp_setup(dpp,
+				plane_state->format,
+				EXPANSION_MODE_ZERO,
+				plane_state->input_csc_color_matrix,
+				plane_state->color_space,
+				NULL);
+
+		if (dpp->funcs->dpp_program_bias_and_scale) {
+			//TODO :for CNVC set scale and bias registers if necessary
+			dcn10_build_prescale_params(&bns_params, plane_state);
+			dpp->funcs->dpp_program_bias_and_scale(dpp, &bns_params);
+		}
 	}
 
-	/* Disconnect unused mpcc */
-	for (i = 0; i < dc->res_pool->pipe_count; i++) {
-		struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i];
-		struct pipe_ctx *old_pipe_ctx =
-				&dc->current_state->res_ctx.pipe_ctx[i];
-		/*
-		 * Powergate reused pipes that are not powergated
-		 * fairly hacky right now, using opp_id as indicator
-		 * TODO: After move dc_post to dc_update, this will
-		 * be removed.
-		 */
-		if (pipe_ctx->plane_state && !old_pipe_ctx->plane_state) {
-			if (old_pipe_ctx->stream_res.tg == tg &&
-			    old_pipe_ctx->plane_res.hubp &&
-			    old_pipe_ctx->plane_res.hubp->opp_id != OPP_ID_INVALID)
-				dc->hwss.disable_plane(dc, old_pipe_ctx);
+	if (pipe_ctx->update_flags.bits.mpcc
+			|| plane_state->update_flags.bits.global_alpha_change
+			|| plane_state->update_flags.bits.per_pixel_alpha_change) {
+		/* Need mpcc to be idle if changing opp */
+		if (pipe_ctx->update_flags.bits.opp_changed) {
+			struct pipe_ctx *old_pipe_ctx = &dc->current_state->res_ctx.pipe_ctx[pipe_ctx->pipe_idx];
+			int mpcc_inst;
+
+			for (mpcc_inst = 0; mpcc_inst < MAX_PIPES; mpcc_inst++) {
+				if (!old_pipe_ctx->stream_res.opp->mpcc_disconnect_pending[mpcc_inst])
+					continue;
+				dc->res_pool->mpc->funcs->wait_for_idle(dc->res_pool->mpc, mpcc_inst);
+				old_pipe_ctx->stream_res.opp->mpcc_disconnect_pending[mpcc_inst] = false;
+			}
 		}
+		dc->hwss.update_mpcc(dc, pipe_ctx);
+	}
 
-		if ((!pipe_ctx->plane_state ||
-		     pipe_ctx->stream_res.tg != old_pipe_ctx->stream_res.tg) &&
-		     old_pipe_ctx->plane_state &&
-		     old_pipe_ctx->stream_res.tg == tg) {
+	if (pipe_ctx->update_flags.bits.scaler ||
+			plane_state->update_flags.bits.scaling_change ||
+			plane_state->update_flags.bits.position_change ||
+			plane_state->update_flags.bits.per_pixel_alpha_change ||
+			pipe_ctx->stream->update_flags.bits.scaling) {
+		pipe_ctx->plane_res.scl_data.lb_params.alpha_en = pipe_ctx->plane_state->per_pixel_alpha;
+		ASSERT(pipe_ctx->plane_res.scl_data.lb_params.depth == LB_PIXEL_DEPTH_30BPP);
+		/* scaler configuration */
+		pipe_ctx->plane_res.dpp->funcs->dpp_set_scaler(
+				pipe_ctx->plane_res.dpp, &pipe_ctx->plane_res.scl_data);
+	}
 
-			dc->hwss.plane_atomic_disconnect(dc, old_pipe_ctx);
-			removed_pipe[i] = true;
+	if (pipe_ctx->update_flags.bits.viewport ||
+			(context == dc->current_state && plane_state->update_flags.bits.scaling_change) ||
+			(context == dc->current_state && pipe_ctx->stream->update_flags.bits.scaling))
+		hubp->funcs->mem_program_viewport(
+			hubp,
+			&pipe_ctx->plane_res.scl_data.viewport,
+			&pipe_ctx->plane_res.scl_data.viewport_c);
+
+	/* Any updates are handled in dc interface, just need to apply existing for plane enable */
+	if ((pipe_ctx->update_flags.bits.enable || pipe_ctx->update_flags.bits.opp_changed)
+			&& pipe_ctx->stream->cursor_attributes.address.quad_part != 0) {
+		dc->hwss.set_cursor_position(pipe_ctx);
+		dc->hwss.set_cursor_attribute(pipe_ctx);
+
+		if (dc->hwss.set_cursor_sdr_white_level)
+			dc->hwss.set_cursor_sdr_white_level(pipe_ctx);
+	}
 
-			DC_LOG_DC("Reset mpcc for pipe %d\n",
-					old_pipe_ctx->pipe_idx);
-		}
+	/* Any updates are handled in dc interface, just need
+	 * to apply existing for plane enable / opp change */
+	if (pipe_ctx->update_flags.bits.enable || pipe_ctx->update_flags.bits.opp_changed
+			|| pipe_ctx->stream->update_flags.bits.gamut_remap
+			|| pipe_ctx->stream->update_flags.bits.out_csc) {
+			/* dpp/cm gamut remap*/
+			dc->hwss.program_gamut_remap(pipe_ctx);
+
+		/*call the dcn2 method which uses mpc csc*/
+		dc->hwss.program_output_csc(dc,
+				pipe_ctx,
+				pipe_ctx->stream->output_color_space,
+				pipe_ctx->stream->csc_color_matrix.matrix,
+				hubp->opp_id);
 	}
 
-	if (num_planes > 0)
-		dcn20_program_all_pipe_in_tree(dc, top_pipe_to_program, context);
+	if (pipe_ctx->update_flags.bits.enable ||
+			pipe_ctx->update_flags.bits.opp_changed ||
+			plane_state->update_flags.bits.pixel_format_change ||
+			plane_state->update_flags.bits.horizontal_mirror_change ||
+			plane_state->update_flags.bits.rotation_change ||
+			plane_state->update_flags.bits.swizzle_change ||
+			plane_state->update_flags.bits.dcc_change ||
+			plane_state->update_flags.bits.bpp_change ||
+			plane_state->update_flags.bits.scaling_change ||
+			plane_state->update_flags.bits.plane_size_change) {
+		struct plane_size size = plane_state->plane_size;
+
+		size.surface_size = pipe_ctx->plane_res.scl_data.viewport;
+		hubp->funcs->hubp_program_surface_config(
+			hubp,
+			plane_state->format,
+			&plane_state->tiling_info,
+			&size,
+			plane_state->rotation,
+			&plane_state->dcc,
+			plane_state->horizontal_mirror,
+			0);
+		hubp->power_gated = false;
+	}
 
-	/* Program secondary blending tree and writeback pipes */
-	if ((stream->num_wb_info > 0) && (dc->hwss.program_all_writeback_pipes_in_tree))
-		dc->hwss.program_all_writeback_pipes_in_tree(dc, stream, context);
+	if (pipe_ctx->update_flags.bits.enable || plane_state->update_flags.bits.addr_update)
+		dc->hwss.update_plane_addr(dc, pipe_ctx);
 
-	if (interdependent_update)
-		for (i = 0; i < dc->res_pool->pipe_count; i++) {
+	if (pipe_ctx->update_flags.bits.enable)
+		hubp->funcs->set_blank(hubp, false);
+}
+
+
+static void dcn20_program_pipe(
+		struct dc *dc,
+		struct pipe_ctx *pipe_ctx,
+		struct dc_state *context)
+{
+	/* Only need to unblank on top pipe */
+	if ((pipe_ctx->update_flags.bits.enable || pipe_ctx->stream->update_flags.bits.abm_level)
+			&& !pipe_ctx->top_pipe && !pipe_ctx->prev_odm_pipe)
+		dc->hwss.blank_pixel_data(dc, pipe_ctx, !pipe_ctx->plane_state->visible);
+
+	if (pipe_ctx->update_flags.bits.global_sync) {
+		pipe_ctx->stream_res.tg->funcs->program_global_sync(
+				pipe_ctx->stream_res.tg,
+				pipe_ctx->pipe_dlg_param.vready_offset,
+				pipe_ctx->pipe_dlg_param.vstartup_start,
+				pipe_ctx->pipe_dlg_param.vupdate_offset,
+				pipe_ctx->pipe_dlg_param.vupdate_width);
+
+		pipe_ctx->stream_res.tg->funcs->set_vtg_params(
+				pipe_ctx->stream_res.tg, &pipe_ctx->stream->timing);
+
+		if (dc->hwss.setup_vupdate_interrupt)
+			dc->hwss.setup_vupdate_interrupt(pipe_ctx);
+	}
+
+	if (pipe_ctx->update_flags.bits.odm)
+		dc->hwss.update_odm(dc, context, pipe_ctx);
+
+	if (pipe_ctx->update_flags.bits.enable)
+		dcn20_enable_plane(dc, pipe_ctx, context);
+
+	if (pipe_ctx->update_flags.raw || pipe_ctx->plane_state->update_flags.raw || pipe_ctx->stream->update_flags.raw)
+		dcn20_update_dchubp_dpp(dc, pipe_ctx, context);
+
+	if (pipe_ctx->update_flags.bits.enable
+			|| pipe_ctx->plane_state->update_flags.bits.sdr_white_level)
+		set_hdr_multiplier(pipe_ctx);
+
+	if (pipe_ctx->update_flags.bits.enable ||
+			pipe_ctx->plane_state->update_flags.bits.in_transfer_func_change ||
+			pipe_ctx->plane_state->update_flags.bits.gamma_change)
+		dc->hwss.set_input_transfer_func(pipe_ctx, pipe_ctx->plane_state);
+
+	/* dcn10_translate_regamma_to_hw_format takes 750us to finish
+	 * only do gamma programming for powering on, internal memcmp to avoid
+	 * updating on slave planes
+	 */
+	if (pipe_ctx->update_flags.bits.enable || pipe_ctx->stream->update_flags.bits.out_tf)
+		dc->hwss.set_output_transfer_func(pipe_ctx, pipe_ctx->stream);
+
+	/* If the pipe has been enabled or has a different opp, we
+	 * should reprogram the fmt. This deals with cases where
+	 * interation between mpc and odm combine on different streams
+	 * causes a different pipe to be chosen to odm combine with.
+	 */
+	if (pipe_ctx->update_flags.bits.enable
+	    || pipe_ctx->update_flags.bits.opp_changed) {
+
+		pipe_ctx->stream_res.opp->funcs->opp_set_dyn_expansion(
+			pipe_ctx->stream_res.opp,
+			COLOR_SPACE_YCBCR601,
+			pipe_ctx->stream->timing.display_color_depth,
+			pipe_ctx->stream->signal);
+
+		pipe_ctx->stream_res.opp->funcs->opp_program_fmt(
+			pipe_ctx->stream_res.opp,
+			&pipe_ctx->stream->bit_depth_params,
+			&pipe_ctx->stream->clamping);
+	}
+}
+
+static bool does_pipe_need_lock(struct pipe_ctx *pipe)
+{
+	if ((pipe->plane_state && pipe->plane_state->update_flags.raw)
+			|| pipe->update_flags.raw)
+		return true;
+	if (pipe->bottom_pipe)
+		return does_pipe_need_lock(pipe->bottom_pipe);
+
+	return false;
+}
+
+static void dcn20_program_front_end_for_ctx(
+		struct dc *dc,
+		struct dc_state *context)
+{
+	const unsigned int TIMEOUT_FOR_PIPE_ENABLE_MS = 100;
+	int i;
+	bool pipe_locked[MAX_PIPES] = {false};
+	DC_LOGGER_INIT(dc->ctx->logger);
+
+	/* Carry over GSL groups in case the context is changing. */
+	for (i = 0; i < dc->res_pool->pipe_count; i++)
+		if (context->res_ctx.pipe_ctx[i].stream == dc->current_state->res_ctx.pipe_ctx[i].stream)
+			context->res_ctx.pipe_ctx[i].stream_res.gsl_group =
+				dc->current_state->res_ctx.pipe_ctx[i].stream_res.gsl_group;
+
+	/* Set pipe update flags and lock pipes */
+	for (i = 0; i < dc->res_pool->pipe_count; i++)
+		dcn20_detect_pipe_changes(&dc->current_state->res_ctx.pipe_ctx[i],
+				&context->res_ctx.pipe_ctx[i]);
+	for (i = 0; i < dc->res_pool->pipe_count; i++)
+		if (!context->res_ctx.pipe_ctx[i].top_pipe &&
+				does_pipe_need_lock(&context->res_ctx.pipe_ctx[i])) {
 			struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i];
 
-			/* Skip inactive pipes and ones already updated */
-			if (!pipe_ctx->stream || pipe_ctx->stream == stream ||
-			    !pipe_ctx->plane_state || !tg->funcs->is_tg_enabled(tg))
-				continue;
+			if (pipe_ctx->update_flags.bits.tg_changed || pipe_ctx->update_flags.bits.enable)
+				dc->hwss.pipe_control_lock(dc, pipe_ctx, true);
+			if (!pipe_ctx->update_flags.bits.enable)
+				dc->hwss.pipe_control_lock(dc, &dc->current_state->res_ctx.pipe_ctx[i], true);
+			pipe_locked[i] = true;
+		}
 
-			pipe_ctx->plane_res.hubp->funcs->hubp_setup_interdependent(
-				pipe_ctx->plane_res.hubp,
-				&pipe_ctx->dlg_regs,
-				&pipe_ctx->ttu_regs);
+	/* OTG blank before disabling all front ends */
+	for (i = 0; i < dc->res_pool->pipe_count; i++)
+		if (context->res_ctx.pipe_ctx[i].update_flags.bits.disable
+				&& !context->res_ctx.pipe_ctx[i].top_pipe
+				&& !context->res_ctx.pipe_ctx[i].prev_odm_pipe
+				&& context->res_ctx.pipe_ctx[i].stream)
+			dc->hwss.blank_pixel_data(dc, &context->res_ctx.pipe_ctx[i], true);
+
+	/* Disconnect mpcc */
+	for (i = 0; i < dc->res_pool->pipe_count; i++)
+		if (context->res_ctx.pipe_ctx[i].update_flags.bits.disable
+				|| context->res_ctx.pipe_ctx[i].update_flags.bits.opp_changed) {
+			dc->hwss.plane_atomic_disconnect(dc, &dc->current_state->res_ctx.pipe_ctx[i]);
+			DC_LOG_DC("Reset mpcc for pipe %d\n", dc->current_state->res_ctx.pipe_ctx[i].pipe_idx);
 		}
 
-	if (interdependent_update)
-		lock_all_pipes(dc, context, false);
-	else
-		dcn20_pipe_control_lock(dc, top_pipe_to_program, false);
+	/*
+	 * Program all updated pipes, order matters for mpcc setup. Start with
+	 * top pipe and program all pipes that follow in order
+	 */
+	for (i = 0; i < dc->res_pool->pipe_count; i++) {
+		struct pipe_ctx *pipe = &context->res_ctx.pipe_ctx[i];
+
+		if (pipe->plane_state && !pipe->top_pipe) {
+			while (pipe) {
+				dcn20_program_pipe(dc, pipe, context);
+				pipe = pipe->bottom_pipe;
+			}
+			/* Program secondary blending tree and writeback pipes */
+			pipe = &context->res_ctx.pipe_ctx[i];
+			if (!pipe->prev_odm_pipe && pipe->stream->num_wb_info > 0
+					&& (pipe->update_flags.raw || pipe->plane_state->update_flags.raw || pipe->stream->update_flags.raw)
+					&& dc->hwss.program_all_writeback_pipes_in_tree)
+				dc->hwss.program_all_writeback_pipes_in_tree(dc, pipe->stream, context);
+		}
+	}
 
+	/* Unlock all locked pipes */
 	for (i = 0; i < dc->res_pool->pipe_count; i++)
-		if (removed_pipe[i])
-			dcn20_disable_plane(dc, &dc->current_state->res_ctx.pipe_ctx[i]);
+		if (pipe_locked[i]) {
+			struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i];
+
+			if (pipe_ctx->update_flags.bits.tg_changed || pipe_ctx->update_flags.bits.enable)
+				dc->hwss.pipe_control_lock(dc, pipe_ctx, false);
+			if (!pipe_ctx->update_flags.bits.enable)
+				dc->hwss.pipe_control_lock(dc, &dc->current_state->res_ctx.pipe_ctx[i], false);
+		}
+
+	for (i = 0; i < dc->res_pool->pipe_count; i++)
+		if (context->res_ctx.pipe_ctx[i].update_flags.bits.disable)
+			dc->hwss.disable_plane(dc, &dc->current_state->res_ctx.pipe_ctx[i]);
 
 	/*
 	 * If we are enabling a pipe, we need to wait for pending clear as this is a critical
@@ -1239,15 +1515,22 @@ static void dcn20_apply_ctx_for_surface(
 	 * will cause HW to perform an "immediate enable" (as opposed to "vsync enable") which
 	 * is unsupported on DCN.
 	 */
-	i = 0;
-	if (num_planes > 0 && top_pipe_to_program &&
-			(prev_top_pipe_to_program == NULL || prev_top_pipe_to_program->plane_state == NULL)) {
-		while (i < TIMEOUT_FOR_PIPE_ENABLE_MS &&
-				top_pipe_to_program->plane_res.hubp->funcs->hubp_is_flip_pending(top_pipe_to_program->plane_res.hubp)) {
-			i += 1;
-			msleep(1);
+	for (i = 0; i < dc->res_pool->pipe_count; i++) {
+		struct pipe_ctx *pipe = &context->res_ctx.pipe_ctx[i];
+
+		if (pipe->plane_state && !pipe->top_pipe && pipe->update_flags.bits.enable) {
+			struct hubp *hubp = pipe->plane_res.hubp;
+			int j = 0;
+
+			for (j = 0; j < TIMEOUT_FOR_PIPE_ENABLE_MS
+					&& hubp->funcs->hubp_is_flip_pending(hubp); j++)
+				msleep(1);
 		}
 	}
+
+	/* WA to apply WM setting*/
+	if (dc->hwseq->wa.DEGVIDCN21)
+		dc->res_pool->hubbub->funcs->apply_DEDCN21_147_wa(dc->res_pool->hubbub);
 }
 
 
@@ -1319,8 +1602,12 @@ bool dcn20_update_bandwidth(
 
 			pipe_ctx->stream_res.tg->funcs->set_vtg_params(
 					pipe_ctx->stream_res.tg, &pipe_ctx->stream->timing);
+
 			if (pipe_ctx->prev_odm_pipe == NULL)
 				dc->hwss.blank_pixel_data(dc, pipe_ctx, blank);
+
+			if (dc->hwss.setup_vupdate_interrupt)
+				dc->hwss.setup_vupdate_interrupt(pipe_ctx);
 		}
 
 		pipe_ctx->plane_res.hubp->funcs->hubp_setup(
@@ -1337,7 +1624,8 @@ bool dcn20_update_bandwidth(
 static void dcn20_enable_writeback(
 		struct dc *dc,
 		const struct dc_stream_status *stream_status,
-		struct dc_writeback_info *wb_info)
+		struct dc_writeback_info *wb_info,
+		struct dc_state *context)
 {
 	struct dwbc *dwb;
 	struct mcif_wb *mcif_wb;
@@ -1354,7 +1642,7 @@ static void dcn20_enable_writeback(
 	optc->funcs->set_dwb_source(optc, wb_info->dwb_pipe_inst);
 	/* set MCIF_WB buffer and arbitration configuration */
 	mcif_wb->funcs->config_mcif_buf(mcif_wb, &wb_info->mcif_buf_params, wb_info->dwb_params.dest_height);
-	mcif_wb->funcs->config_mcif_arb(mcif_wb, &dc->current_state->bw_ctx.bw.dcn.bw_writeback.mcif_wb_arb[wb_info->dwb_pipe_inst]);
+	mcif_wb->funcs->config_mcif_arb(mcif_wb, &context->bw_ctx.bw.dcn.bw_writeback.mcif_wb_arb[wb_info->dwb_pipe_inst]);
 	/* Enable MCIF_WB */
 	mcif_wb->funcs->enable_mcif(mcif_wb);
 	/* Enable DWB */
@@ -1702,6 +1990,28 @@ static void dcn20_reset_hw_ctx_wrap(
 	}
 }
 
+void dcn20_get_mpctree_visual_confirm_color(
+		struct pipe_ctx *pipe_ctx,
+		struct tg_color *color)
+{
+	const struct tg_color pipe_colors[6] = {
+			{MAX_TG_COLOR_VALUE, 0, 0}, // red
+			{MAX_TG_COLOR_VALUE, 0, MAX_TG_COLOR_VALUE}, // yellow
+			{0, MAX_TG_COLOR_VALUE, 0}, // blue
+			{MAX_TG_COLOR_VALUE / 2, 0, MAX_TG_COLOR_VALUE / 2}, // purple
+			{0, 0, MAX_TG_COLOR_VALUE}, // green
+			{MAX_TG_COLOR_VALUE, MAX_TG_COLOR_VALUE * 2 / 3, 0}, // orange
+	};
+
+	struct pipe_ctx *top_pipe = pipe_ctx;
+
+	while (top_pipe->top_pipe) {
+		top_pipe = top_pipe->top_pipe;
+	}
+
+	*color = pipe_colors[top_pipe->pipe_idx];
+}
+
 static void dcn20_update_mpcc(struct dc *dc, struct pipe_ctx *pipe_ctx)
 {
 	struct hubp *hubp = pipe_ctx->plane_res.hubp;
@@ -1719,6 +2029,9 @@ static void dcn20_update_mpcc(struct dc *dc, struct pipe_ctx *pipe_ctx)
 	} else if (dc->debug.visual_confirm == VISUAL_CONFIRM_SURFACE) {
 		dcn10_get_surface_visual_confirm_color(
 				pipe_ctx, &blnd_cfg.black_color);
+	} else if (dc->debug.visual_confirm == VISUAL_CONFIRM_MPCTREE) {
+		dcn20_get_mpctree_visual_confirm_color(
+				pipe_ctx, &blnd_cfg.black_color);
 	}
 
 	if (per_pixel_alpha)
@@ -1919,8 +2232,10 @@ static void dcn20_enable_stream(struct pipe_ctx *pipe_ctx)
 	link->link_enc->funcs->connect_dig_be_to_fe(link->link_enc,
 						    pipe_ctx->stream_res.stream_enc->id, true);
 
-	if (link->dc->hwss.program_dmdata_engine)
-		link->dc->hwss.program_dmdata_engine(pipe_ctx);
+	if (pipe_ctx->plane_state && pipe_ctx->plane_state->flip_immediate != 1) {
+		if (link->dc->hwss.program_dmdata_engine)
+			link->dc->hwss.program_dmdata_engine(pipe_ctx);
+	}
 
 	link->dc->hwss.update_info_frame(pipe_ctx);
 
@@ -2095,7 +2410,8 @@ void dcn20_hw_sequencer_construct(struct dc *dc)
 	dc->hwss.program_triplebuffer = dcn20_program_tripleBuffer;
 	dc->hwss.set_input_transfer_func = dcn20_set_input_transfer_func;
 	dc->hwss.set_output_transfer_func = dcn20_set_output_transfer_func;
-	dc->hwss.apply_ctx_for_surface = dcn20_apply_ctx_for_surface;
+	dc->hwss.apply_ctx_for_surface = NULL;
+	dc->hwss.program_front_end_for_ctx = dcn20_program_front_end_for_ctx;
 	dc->hwss.pipe_control_lock = dcn20_pipe_control_lock;
 	dc->hwss.pipe_control_lock_global = dcn20_pipe_control_lock_global;
 	dc->hwss.optimize_bandwidth = dcn20_optimize_bandwidth;
diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.h b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.h
index 92ab3dd91814..3098f1049ed7 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.h
@@ -96,4 +96,20 @@ void dcn20_init_blank(
 	   struct dc *dc,
 	   struct timing_generator *tg);
 void dcn20_display_init(struct dc *dc);
+void dcn20_pipe_control_lock(
+	struct dc *dc,
+	struct pipe_ctx *pipe,
+	bool lock);
+void dcn20_disable_plane(struct dc *dc, struct pipe_ctx *pipe_ctx);
+void dcn20_enable_plane(
+	struct dc *dc,
+	struct pipe_ctx *pipe_ctx,
+	struct dc_state *context);
+bool dcn20_set_blend_lut(
+	struct pipe_ctx *pipe_ctx, const struct dc_plane_state *plane_state);
+bool dcn20_set_shaper_3dlut(
+	struct pipe_ctx *pipe_ctx, const struct dc_plane_state *plane_state);
+void dcn20_get_mpctree_visual_confirm_color(
+		struct pipe_ctx *pipe_ctx,
+		struct tg_color *color);
 #endif /* __DC_HWSS_DCN20_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_link_encoder.h b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_link_encoder.h
index 3736b5548a25..0c98a0bbbd14 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_link_encoder.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_link_encoder.h
@@ -91,6 +91,13 @@ struct mpll_cfg {
 	uint32_t ref_range;
 	uint32_t ref_clk;
 	bool hdmimode_enable;
+	bool sup_pre_hp;
+	bool dp_tx0_vergdrv_byp;
+	bool dp_tx1_vergdrv_byp;
+	bool dp_tx2_vergdrv_byp;
+	bool dp_tx3_vergdrv_byp;
+
+
 };
 
 struct dpcssys_phy_seq_cfg {
diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_optc.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_optc.c
index 2137e2be2140..3b613fb93ef8 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_optc.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_optc.c
@@ -287,6 +287,10 @@ void optc2_get_optc_source(struct timing_generator *optc,
 		*num_of_src_opp = 2;
 	else
 		*num_of_src_opp = 1;
+
+	/* Work around VBIOS not updating OPTC_NUM_OF_INPUT_SEGMENT */
+	if (*src_opp_id_1 == 0xf)
+		*num_of_src_opp = 1;
 }
 
 void optc2_set_dwb_source(struct timing_generator *optc,
@@ -456,7 +460,7 @@ static struct timing_generator_funcs dcn20_tg_funcs = {
 		.set_vtg_params = optc1_set_vtg_params,
 		.program_manual_trigger = optc2_program_manual_trigger,
 		.setup_manual_trigger = optc2_setup_manual_trigger,
-		.is_matching_timing = optc1_is_matching_timing
+		.get_hw_timing = optc1_get_hw_timing,
 };
 
 void dcn20_timing_generator_init(struct optc *optc1)
diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c
index b4e3ce22ed52..bbd1c98564be 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c
@@ -581,11 +581,13 @@ static const struct dcn2_dpp_registers tf_regs[] = {
 };
 
 static const struct dcn2_dpp_shift tf_shift = {
-		TF_REG_LIST_SH_MASK_DCN20(__SHIFT)
+		TF_REG_LIST_SH_MASK_DCN20(__SHIFT),
+		TF_DEBUG_REG_LIST_SH_DCN10
 };
 
 static const struct dcn2_dpp_mask tf_mask = {
-		TF_REG_LIST_SH_MASK_DCN20(_MASK)
+		TF_REG_LIST_SH_MASK_DCN20(_MASK),
+		TF_DEBUG_REG_LIST_MASK_DCN10
 };
 
 #define dwbc_regs_dcn2(id)\
@@ -732,6 +734,42 @@ static const struct dcn20_vmid_mask vmid_masks = {
 		DCN20_VMID_MASK_SH_LIST(_MASK)
 };
 
+static const struct dce110_aux_registers_shift aux_shift = {
+		DCN_AUX_MASK_SH_LIST(__SHIFT)
+};
+
+static const struct dce110_aux_registers_mask aux_mask = {
+		DCN_AUX_MASK_SH_LIST(_MASK)
+};
+
+static int map_transmitter_id_to_phy_instance(
+	enum transmitter transmitter)
+{
+	switch (transmitter) {
+	case TRANSMITTER_UNIPHY_A:
+		return 0;
+	break;
+	case TRANSMITTER_UNIPHY_B:
+		return 1;
+	break;
+	case TRANSMITTER_UNIPHY_C:
+		return 2;
+	break;
+	case TRANSMITTER_UNIPHY_D:
+		return 3;
+	break;
+	case TRANSMITTER_UNIPHY_E:
+		return 4;
+	break;
+	case TRANSMITTER_UNIPHY_F:
+		return 5;
+	break;
+	default:
+		ASSERT(0);
+		return 0;
+	}
+}
+
 #ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT
 #define dsc_regsDCN20(id)\
 [id] = {\
@@ -814,7 +852,7 @@ static const struct resource_caps res_cap_nv14 = {
 		.num_audio = 6,
 		.num_stream_encoder = 5,
 		.num_pll = 5,
-		.num_dwb = 0,
+		.num_dwb = 1,
 		.num_ddc = 5,
 };
 
@@ -825,7 +863,7 @@ static const struct dc_debug_options debug_defaults_drv = {
 		.clock_trace = true,
 		.disable_pplib_clock_request = true,
 		.pipe_split_policy = MPC_SPLIT_DYNAMIC,
-		.force_single_disp_pipe_split = true,
+		.force_single_disp_pipe_split = false,
 		.disable_dcc = DCC_ENABLE,
 		.vsr_support = true,
 		.performance_trace = false,
@@ -922,7 +960,10 @@ struct dce_aux *dcn20_aux_engine_create(
 
 	dce110_aux_engine_construct(aux_engine, ctx, inst,
 				    SW_AUX_TIMEOUT_PERIOD_MULTIPLIER * AUX_TIMEOUT_PERIOD,
-				    &aux_engine_regs[inst]);
+				    &aux_engine_regs[inst],
+					&aux_mask,
+					&aux_shift,
+					ctx->dc->caps.extended_aux_timeout_support);
 
 	return &aux_engine->base;
 }
@@ -1042,14 +1083,18 @@ struct link_encoder *dcn20_link_encoder_create(
 {
 	struct dcn20_link_encoder *enc20 =
 		kzalloc(sizeof(struct dcn20_link_encoder), GFP_KERNEL);
+	int link_regs_id;
 
 	if (!enc20)
 		return NULL;
 
+	link_regs_id =
+		map_transmitter_id_to_phy_instance(enc_init_data->transmitter);
+
 	dcn20_link_encoder_construct(enc20,
 				      enc_init_data,
 				      &link_enc_feature,
-				      &link_enc_regs[enc_init_data->transmitter],
+				      &link_enc_regs[link_regs_id],
 				      &link_enc_aux_regs[enc_init_data->channel - 1],
 				      &link_enc_hpd_regs[enc_init_data->hpd_source],
 				      &le_shift,
@@ -1077,6 +1122,7 @@ struct clock_source *dcn20_clock_source_create(
 		return &clk_src->base;
 	}
 
+	kfree(clk_src);
 	BREAK_TO_DEBUGGER();
 	return NULL;
 }
@@ -1106,6 +1152,11 @@ struct stream_encoder *dcn20_stream_encoder_create(
 	if (!enc1)
 		return NULL;
 
+	if (ASICREV_IS_NAVI14_M(ctx->asic_id.hw_internal_rev)) {
+		if (eng_id >= ENGINE_ID_DIGD)
+			eng_id++;
+	}
+
 	dcn20_stream_encoder_construct(enc1, ctx, ctx->dc_bios, eng_id,
 					&stream_enc_regs[eng_id],
 					&se_shift, &se_mask);
@@ -1153,6 +1204,8 @@ static const struct resource_create_funcs res_create_maximus_funcs = {
 	.create_hwseq = dcn20_hwseq_create,
 };
 
+static void dcn20_pp_smu_destroy(struct pp_smu_funcs **pp_smu);
+
 void dcn20_clock_source_destroy(struct clock_source **clk_src)
 {
 	kfree(TO_DCE110_CLK_SRC(*clk_src));
@@ -1595,7 +1648,7 @@ static void swizzle_to_dml_params(
 	}
 }
 
-static bool dcn20_split_stream_for_odm(
+bool dcn20_split_stream_for_odm(
 		struct resource_context *res_ctx,
 		const struct resource_pool *pool,
 		struct pipe_ctx *prev_odm_pipe,
@@ -1616,7 +1669,6 @@ static bool dcn20_split_stream_for_odm(
 	next_odm_pipe->stream_res.dsc = NULL;
 #endif
 	if (prev_odm_pipe->next_odm_pipe && prev_odm_pipe->next_odm_pipe != next_odm_pipe) {
-		ASSERT(!next_odm_pipe->next_odm_pipe);
 		next_odm_pipe->next_odm_pipe = prev_odm_pipe->next_odm_pipe;
 		next_odm_pipe->next_odm_pipe->prev_odm_pipe = next_odm_pipe;
 	}
@@ -1673,7 +1725,7 @@ static bool dcn20_split_stream_for_odm(
 	return true;
 }
 
-static void dcn20_split_stream_for_mpc(
+void dcn20_split_stream_for_mpc(
 		struct resource_context *res_ctx,
 		const struct resource_pool *pool,
 		struct pipe_ctx *primary_pipe,
@@ -1759,7 +1811,7 @@ int dcn20_populate_dml_pipes_from_context(
 			pipe_cnt = i;
 			continue;
 		}
-		if (!resource_are_streams_timing_synchronizable(
+		if (dc->debug.disable_timing_sync || !resource_are_streams_timing_synchronizable(
 				res_ctx->pipe_ctx[pipe_cnt].stream,
 				res_ctx->pipe_ctx[i].stream)) {
 			synchronized_vblank = false;
@@ -1891,7 +1943,7 @@ int dcn20_populate_dml_pipes_from_context(
 			break;
 		case PIXEL_ENCODING_YCBCR420:
 			pipes[pipe_cnt].dout.output_format = dm_420;
-			pipes[pipe_cnt].dout.output_bpp = (output_bpc * 3) / 2;
+			pipes[pipe_cnt].dout.output_bpp = (output_bpc * 3.0) / 2;
 			break;
 		case PIXEL_ENCODING_YCBCR422:
 			if (true) /* todo */
@@ -1905,6 +1957,11 @@ int dcn20_populate_dml_pipes_from_context(
 			pipes[pipe_cnt].dout.output_bpp = output_bpc * 3;
 		}
 
+#ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT
+		if (res_ctx->pipe_ctx[i].stream->timing.flags.DSC)
+			pipes[pipe_cnt].dout.output_bpp = res_ctx->pipe_ctx[i].stream->timing.dsc_cfg.bits_per_pixel / 16.0;
+#endif
+
 		/* todo: default max for now, until there is logic reflecting this in dc*/
 		pipes[pipe_cnt].dout.output_bpc = 12;
 		/*
@@ -2126,7 +2183,7 @@ void dcn20_set_mcif_arb_params(
 }
 
 #ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT
-static bool dcn20_validate_dsc(struct dc *dc, struct dc_state *new_ctx)
+bool dcn20_validate_dsc(struct dc *dc, struct dc_state *new_ctx)
 {
 	int i;
 
@@ -2161,7 +2218,7 @@ static bool dcn20_validate_dsc(struct dc *dc, struct dc_state *new_ctx)
 }
 #endif
 
-static struct pipe_ctx *dcn20_find_secondary_pipe(struct dc *dc,
+struct pipe_ctx *dcn20_find_secondary_pipe(struct dc *dc,
 		struct resource_context *res_ctx,
 		const struct resource_pool *pool,
 		const struct pipe_ctx *primary_pipe)
@@ -2201,7 +2258,8 @@ static struct pipe_ctx *dcn20_find_secondary_pipe(struct dc *dc,
 		 */
 		if (secondary_pipe == NULL) {
 			for (j = dc->res_pool->pipe_count - 1; j >= 0; j--) {
-				if (dc->current_state->res_ctx.pipe_ctx[j].top_pipe == NULL) {
+				if (dc->current_state->res_ctx.pipe_ctx[j].top_pipe == NULL
+						&& dc->current_state->res_ctx.pipe_ctx[j].prev_odm_pipe == NULL) {
 					preferred_pipe_idx = j;
 
 					if (res_ctx->pipe_ctx[preferred_pipe_idx].stream == NULL) {
@@ -2237,29 +2295,11 @@ static struct pipe_ctx *dcn20_find_secondary_pipe(struct dc *dc,
 	return secondary_pipe;
 }
 
-bool dcn20_fast_validate_bw(
+void dcn20_merge_pipes_for_validate(
 		struct dc *dc,
-		struct dc_state *context,
-		display_e2e_pipe_params_st *pipes,
-		int *pipe_cnt_out,
-		int *pipe_split_from,
-		int *vlevel_out)
+		struct dc_state *context)
 {
-	bool out = false;
-
-	int pipe_cnt, i, pipe_idx, vlevel, vlevel_unsplit;
-	bool odm_capable = context->bw_ctx.dml.ip.odm_capable;
-	bool force_split = false;
-#ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT
-	bool failed_non_odm_dsc = false;
-#endif
-	int split_threshold = dc->res_pool->pipe_count / 2;
-	bool avoid_split = dc->debug.pipe_split_policy != MPC_SPLIT_DYNAMIC;
-
-
-	ASSERT(pipes);
-	if (!pipes)
-		return false;
+	int i;
 
 	/* merge previously split odm pipes since mode support needs to make the decision */
 	for (i = 0; i < dc->res_pool->pipe_count; i++) {
@@ -2314,51 +2354,19 @@ bool dcn20_fast_validate_bw(
 		if (pipe->plane_state)
 			resource_build_scaling_params(pipe);
 	}
+}
 
-	if (dc->res_pool->funcs->populate_dml_pipes)
-		pipe_cnt = dc->res_pool->funcs->populate_dml_pipes(dc,
-			&context->res_ctx, pipes);
-	else
-		pipe_cnt = dcn20_populate_dml_pipes_from_context(dc,
-			&context->res_ctx, pipes);
-
-	*pipe_cnt_out = pipe_cnt;
-
-	if (!pipe_cnt) {
-		out = true;
-		goto validate_out;
-	}
-
-	context->bw_ctx.dml.ip.odm_capable = 0;
-
-	vlevel = dml_get_voltage_level(&context->bw_ctx.dml, pipes, pipe_cnt);
-
-	context->bw_ctx.dml.ip.odm_capable = odm_capable;
-
-#ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT
-	/* 1 dsc per stream dsc validation */
-	if (vlevel <= context->bw_ctx.dml.soc.num_states)
-		if (!dcn20_validate_dsc(dc, context)) {
-			failed_non_odm_dsc = true;
-			vlevel = context->bw_ctx.dml.soc.num_states + 1;
-		}
-#endif
-
-	if (vlevel > context->bw_ctx.dml.soc.num_states && odm_capable)
-		vlevel = dml_get_voltage_level(&context->bw_ctx.dml, pipes, pipe_cnt);
-
-	if (vlevel > context->bw_ctx.dml.soc.num_states)
-		goto validate_fail;
-
-	if ((context->stream_count > split_threshold && dc->current_state->stream_count <= split_threshold)
-		|| (context->stream_count <= split_threshold && dc->current_state->stream_count > split_threshold))
-		context->commit_hints.full_update_needed = true;
-
-	/*initialize pipe_just_split_from to invalid idx*/
-	for (i = 0; i < MAX_PIPES; i++)
-		pipe_split_from[i] = -1;
+int dcn20_validate_apply_pipe_split_flags(
+		struct dc *dc,
+		struct dc_state *context,
+		int vlevel,
+		bool *split)
+{
+	int i, pipe_idx, vlevel_split;
+	bool force_split = false;
+	bool avoid_split = dc->debug.pipe_split_policy != MPC_SPLIT_DYNAMIC;
 
-	/* Single display only conditionals get set here */
+	/* Single display loop, exits if there is more than one display */
 	for (i = 0; i < dc->res_pool->pipe_count; i++) {
 		struct pipe_ctx *pipe = &context->res_ctx.pipe_ctx[i];
 		bool exit_loop = false;
@@ -2385,38 +2393,107 @@ bool dcn20_fast_validate_bw(
 		if (exit_loop)
 			break;
 	}
-
-	if (context->stream_count > split_threshold)
+	/* TODO: fix dc bugs and remove this split threshold thing */
+	if (context->stream_count > dc->res_pool->pipe_count / 2)
 		avoid_split = true;
 
-	vlevel_unsplit = vlevel;
+	/* Avoid split loop looks for lowest voltage level that allows most unsplit pipes possible */
+	if (avoid_split) {
+		for (i = 0, pipe_idx = 0; i < dc->res_pool->pipe_count; i++) {
+			if (!context->res_ctx.pipe_ctx[i].stream)
+				continue;
+
+			for (vlevel_split = vlevel; vlevel <= context->bw_ctx.dml.soc.num_states; vlevel++)
+				if (context->bw_ctx.dml.vba.NoOfDPP[vlevel][0][pipe_idx] == 1)
+					break;
+			/* Impossible to not split this pipe */
+			if (vlevel > context->bw_ctx.dml.soc.num_states)
+				vlevel = vlevel_split;
+			pipe_idx++;
+		}
+		context->bw_ctx.dml.vba.maxMpcComb = 0;
+	}
+
+	/* Split loop sets which pipe should be split based on dml outputs and dc flags */
 	for (i = 0, pipe_idx = 0; i < dc->res_pool->pipe_count; i++) {
+		struct pipe_ctx *pipe = &context->res_ctx.pipe_ctx[i];
+
 		if (!context->res_ctx.pipe_ctx[i].stream)
 			continue;
-		for (; vlevel_unsplit <= context->bw_ctx.dml.soc.num_states; vlevel_unsplit++)
-			if (context->bw_ctx.dml.vba.NoOfDPP[vlevel_unsplit][0][pipe_idx] == 1)
-				break;
+
+		if (force_split || context->bw_ctx.dml.vba.NoOfDPP[vlevel][context->bw_ctx.dml.vba.maxMpcComb][pipe_idx] > 1)
+			split[i] = true;
+		if ((pipe->stream->view_format ==
+				VIEW_3D_FORMAT_SIDE_BY_SIDE ||
+				pipe->stream->view_format ==
+				VIEW_3D_FORMAT_TOP_AND_BOTTOM) &&
+				(pipe->stream->timing.timing_3d_format ==
+				TIMING_3D_FORMAT_TOP_AND_BOTTOM ||
+				 pipe->stream->timing.timing_3d_format ==
+				TIMING_3D_FORMAT_SIDE_BY_SIDE))
+			split[i] = true;
+		if (dc->debug.force_odm_combine & (1 << pipe->stream_res.tg->inst)) {
+			split[i] = true;
+			context->bw_ctx.dml.vba.ODMCombineEnablePerState[vlevel][pipe_idx] = true;
+		}
+		context->bw_ctx.dml.vba.ODMCombineEnabled[pipe_idx] =
+			context->bw_ctx.dml.vba.ODMCombineEnablePerState[vlevel][pipe_idx];
+		/* Adjust dppclk when split is forced, do not bother with dispclk */
+		if (split[i] && context->bw_ctx.dml.vba.NoOfDPP[vlevel][context->bw_ctx.dml.vba.maxMpcComb][pipe_idx] == 1)
+			context->bw_ctx.dml.vba.RequiredDPPCLK[vlevel][context->bw_ctx.dml.vba.maxMpcComb][pipe_idx] /= 2;
 		pipe_idx++;
 	}
 
+	return vlevel;
+}
+
+bool dcn20_fast_validate_bw(
+		struct dc *dc,
+		struct dc_state *context,
+		display_e2e_pipe_params_st *pipes,
+		int *pipe_cnt_out,
+		int *pipe_split_from,
+		int *vlevel_out)
+{
+	bool out = false;
+	bool split[MAX_PIPES] = { false };
+	int pipe_cnt, i, pipe_idx, vlevel;
+
+	ASSERT(pipes);
+	if (!pipes)
+		return false;
+
+	dcn20_merge_pipes_for_validate(dc, context);
+
+	pipe_cnt = dc->res_pool->funcs->populate_dml_pipes(dc, &context->res_ctx, pipes);
+
+	*pipe_cnt_out = pipe_cnt;
+
+	if (!pipe_cnt) {
+		out = true;
+		goto validate_out;
+	}
+
+	vlevel = dml_get_voltage_level(&context->bw_ctx.dml, pipes, pipe_cnt);
+
+	if (vlevel > context->bw_ctx.dml.soc.num_states)
+		goto validate_fail;
+
+	vlevel = dcn20_validate_apply_pipe_split_flags(dc, context, vlevel, split);
+
+	/*initialize pipe_just_split_from to invalid idx*/
+	for (i = 0; i < MAX_PIPES; i++)
+		pipe_split_from[i] = -1;
+
 	for (i = 0, pipe_idx = -1; i < dc->res_pool->pipe_count; i++) {
 		struct pipe_ctx *pipe = &context->res_ctx.pipe_ctx[i];
 		struct pipe_ctx *hsplit_pipe = pipe->bottom_pipe;
-		bool need_split = true;
-		bool need_split3d;
 
 		if (!pipe->stream || pipe_split_from[i] >= 0)
 			continue;
 
 		pipe_idx++;
 
-		if (dc->debug.force_odm_combine & (1 << pipe->stream_res.tg->inst)) {
-			force_split = true;
-			context->bw_ctx.dml.vba.ODMCombineEnabled[pipe_idx] = true;
-			context->bw_ctx.dml.vba.ODMCombineEnablePerState[vlevel][pipe_idx] = true;
-		}
-		if (force_split && context->bw_ctx.dml.vba.NoOfDPP[vlevel][context->bw_ctx.dml.vba.maxMpcComb][pipe_idx] == 1)
-			context->bw_ctx.dml.vba.RequiredDPPCLK[vlevel][context->bw_ctx.dml.vba.maxMpcComb][pipe_idx] /= 2;
 		if (!pipe->top_pipe && !pipe->plane_state && context->bw_ctx.dml.vba.ODMCombineEnabled[pipe_idx]) {
 			hsplit_pipe = dcn20_find_secondary_pipe(dc, &context->res_ctx, dc->res_pool, pipe);
 			ASSERT(hsplit_pipe);
@@ -2434,40 +2511,26 @@ bool dcn20_fast_validate_bw(
 		if (pipe->top_pipe && pipe->plane_state == pipe->top_pipe->plane_state)
 			continue;
 
-		need_split3d = ((pipe->stream->view_format ==
-				VIEW_3D_FORMAT_SIDE_BY_SIDE ||
-				pipe->stream->view_format ==
-				VIEW_3D_FORMAT_TOP_AND_BOTTOM) &&
-				(pipe->stream->timing.timing_3d_format ==
-				TIMING_3D_FORMAT_TOP_AND_BOTTOM ||
-				 pipe->stream->timing.timing_3d_format ==
-				TIMING_3D_FORMAT_SIDE_BY_SIDE));
-
-		if (avoid_split && vlevel_unsplit <= context->bw_ctx.dml.soc.num_states && !force_split && !need_split3d) {
-			need_split = false;
-			vlevel = vlevel_unsplit;
-			context->bw_ctx.dml.vba.maxMpcComb = 0;
-		} else
-			need_split = context->bw_ctx.dml.vba.NoOfDPP[vlevel][context->bw_ctx.dml.vba.maxMpcComb][pipe_idx] == 2;
-
 		/* We do not support mpo + odm at the moment */
 		if (hsplit_pipe && hsplit_pipe->plane_state != pipe->plane_state
 				&& context->bw_ctx.dml.vba.ODMCombineEnabled[pipe_idx])
 			goto validate_fail;
 
-		if (need_split3d || need_split || force_split) {
+		if (split[i]) {
 			if (!hsplit_pipe || hsplit_pipe->plane_state != pipe->plane_state) {
 				/* pipe not split previously needs split */
 				hsplit_pipe = dcn20_find_secondary_pipe(dc, &context->res_ctx, dc->res_pool, pipe);
-				ASSERT(hsplit_pipe || force_split);
-				if (!hsplit_pipe)
+				ASSERT(hsplit_pipe);
+				if (!hsplit_pipe) {
+					context->bw_ctx.dml.vba.RequiredDPPCLK[vlevel][context->bw_ctx.dml.vba.maxMpcComb][pipe_idx] *= 2;
 					continue;
-
+				}
 				if (context->bw_ctx.dml.vba.ODMCombineEnabled[pipe_idx]) {
 					if (!dcn20_split_stream_for_odm(
 							&context->res_ctx, dc->res_pool,
 							pipe, hsplit_pipe))
 						goto validate_fail;
+					dcn20_build_mapped_resource(dc, context, pipe->stream);
 				} else
 					dcn20_split_stream_for_mpc(
 						&context->res_ctx, dc->res_pool,
@@ -2481,7 +2544,7 @@ bool dcn20_fast_validate_bw(
 	}
 #ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT
 	/* Actual dsc count per stream dsc validation*/
-	if (failed_non_odm_dsc && !dcn20_validate_dsc(dc, context)) {
+	if (!dcn20_validate_dsc(dc, context)) {
 		context->bw_ctx.dml.vba.ValidationStatus[context->bw_ctx.dml.vba.soc.num_states] =
 				DML_FAIL_DSC_VALIDATION_FAILURE;
 		goto validate_fail;
@@ -2500,7 +2563,7 @@ validate_out:
 	return out;
 }
 
-void dcn20_calculate_wm(
+static void dcn20_calculate_wm(
 		struct dc *dc, struct dc_state *context,
 		display_e2e_pipe_params_st *pipes,
 		int *out_pipe_cnt,
@@ -2521,7 +2584,7 @@ void dcn20_calculate_wm(
 					context->bw_ctx.dml.vba.RequiredDPPCLK[vlevel][context->bw_ctx.dml.vba.maxMpcComb][pipe_idx];
 			if (context->bw_ctx.dml.vba.BlendingAndTiming[pipe_idx] == pipe_idx)
 				pipes[pipe_cnt].pipe.dest.odm_combine =
-						context->bw_ctx.dml.vba.ODMCombineEnablePerState[vlevel][pipe_idx];
+						context->bw_ctx.dml.vba.ODMCombineEnabled[pipe_idx];
 			else
 				pipes[pipe_cnt].pipe.dest.odm_combine = 0;
 			pipe_idx++;
@@ -2530,7 +2593,7 @@ void dcn20_calculate_wm(
 					context->bw_ctx.dml.vba.RequiredDPPCLK[vlevel][context->bw_ctx.dml.vba.maxMpcComb][pipe_split_from[i]];
 			if (context->bw_ctx.dml.vba.BlendingAndTiming[pipe_split_from[i]] == pipe_split_from[i])
 				pipes[pipe_cnt].pipe.dest.odm_combine =
-						context->bw_ctx.dml.vba.ODMCombineEnablePerState[vlevel][pipe_split_from[i]];
+						context->bw_ctx.dml.vba.ODMCombineEnabled[pipe_split_from[i]];
 			else
 				pipes[pipe_cnt].pipe.dest.odm_combine = 0;
 		}
@@ -2573,6 +2636,11 @@ void dcn20_calculate_wm(
 	context->bw_ctx.bw.dcn.watermarks.b.cstate_pstate.cstate_exit_ns = get_wm_stutter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000;
 	context->bw_ctx.bw.dcn.watermarks.b.cstate_pstate.pstate_change_ns = get_wm_dram_clock_change(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000;
 	context->bw_ctx.bw.dcn.watermarks.b.pte_meta_urgent_ns = get_wm_memory_trip(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000;
+#if defined(CONFIG_DRM_AMD_DC_DCN2_1)
+	context->bw_ctx.bw.dcn.watermarks.b.frac_urg_bw_nom = get_fraction_of_urgent_bandwidth(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000;
+	context->bw_ctx.bw.dcn.watermarks.b.frac_urg_bw_flip = get_fraction_of_urgent_bandwidth_imm_flip(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000;
+	context->bw_ctx.bw.dcn.watermarks.b.urgent_latency_ns = get_urgent_latency(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000;
+#endif
 
 	if (vlevel < 2) {
 		pipes[0].clks_cfg.voltage = 2;
@@ -2584,6 +2652,10 @@ void dcn20_calculate_wm(
 	context->bw_ctx.bw.dcn.watermarks.c.cstate_pstate.cstate_exit_ns = get_wm_stutter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000;
 	context->bw_ctx.bw.dcn.watermarks.c.cstate_pstate.pstate_change_ns = get_wm_dram_clock_change(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000;
 	context->bw_ctx.bw.dcn.watermarks.c.pte_meta_urgent_ns = get_wm_memory_trip(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000;
+#if defined(CONFIG_DRM_AMD_DC_DCN2_1)
+	context->bw_ctx.bw.dcn.watermarks.c.frac_urg_bw_nom = get_fraction_of_urgent_bandwidth(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000;
+	context->bw_ctx.bw.dcn.watermarks.c.frac_urg_bw_flip = get_fraction_of_urgent_bandwidth_imm_flip(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000;
+#endif
 
 	if (vlevel < 3) {
 		pipes[0].clks_cfg.voltage = 3;
@@ -2595,6 +2667,10 @@ void dcn20_calculate_wm(
 	context->bw_ctx.bw.dcn.watermarks.d.cstate_pstate.cstate_exit_ns = get_wm_stutter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000;
 	context->bw_ctx.bw.dcn.watermarks.d.cstate_pstate.pstate_change_ns = get_wm_dram_clock_change(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000;
 	context->bw_ctx.bw.dcn.watermarks.d.pte_meta_urgent_ns = get_wm_memory_trip(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000;
+#if defined(CONFIG_DRM_AMD_DC_DCN2_1)
+	context->bw_ctx.bw.dcn.watermarks.d.frac_urg_bw_nom = get_fraction_of_urgent_bandwidth(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000;
+	context->bw_ctx.bw.dcn.watermarks.d.frac_urg_bw_flip = get_fraction_of_urgent_bandwidth_imm_flip(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000;
+#endif
 
 	pipes[0].clks_cfg.voltage = vlevel;
 	pipes[0].clks_cfg.dcfclk_mhz = context->bw_ctx.dml.soc.clock_limits[vlevel].dcfclk_mhz;
@@ -2604,6 +2680,10 @@ void dcn20_calculate_wm(
 	context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.cstate_exit_ns = get_wm_stutter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000;
 	context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.pstate_change_ns = get_wm_dram_clock_change(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000;
 	context->bw_ctx.bw.dcn.watermarks.a.pte_meta_urgent_ns = get_wm_memory_trip(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000;
+#if defined(CONFIG_DRM_AMD_DC_DCN2_1)
+	context->bw_ctx.bw.dcn.watermarks.a.frac_urg_bw_nom = get_fraction_of_urgent_bandwidth(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000;
+	context->bw_ctx.bw.dcn.watermarks.a.frac_urg_bw_flip = get_fraction_of_urgent_bandwidth_imm_flip(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000;
+#endif
 }
 
 void dcn20_calculate_dlg_params(
@@ -2623,7 +2703,7 @@ void dcn20_calculate_dlg_params(
 	context->bw_ctx.bw.dcn.clk.socclk_khz = context->bw_ctx.dml.vba.SOCCLK * 1000;
 	context->bw_ctx.bw.dcn.clk.dramclk_khz = context->bw_ctx.dml.vba.DRAMSpeed * 1000 / 16;
 	context->bw_ctx.bw.dcn.clk.dcfclk_deep_sleep_khz = context->bw_ctx.dml.vba.DCFCLKDeepSleep * 1000;
-	context->bw_ctx.bw.dcn.clk.fclk_khz = 0;
+	context->bw_ctx.bw.dcn.clk.fclk_khz = context->bw_ctx.dml.vba.FabricClock * 1000;
 	context->bw_ctx.bw.dcn.clk.p_state_change_support =
 		context->bw_ctx.dml.vba.DRAMClockChangeSupport[vlevel][context->bw_ctx.dml.vba.maxMpcComb]
 							!= dm_dram_clock_change_unsupported;
@@ -2639,8 +2719,8 @@ void dcn20_calculate_dlg_params(
 			continue;
 
 		if (!visited[pipe_idx]) {
-			display_pipe_source_params_st *src = &pipes[pipe_idx_unsplit].pipe.src;
-			display_pipe_dest_params_st *dst = &pipes[pipe_idx_unsplit].pipe.dest;
+			display_pipe_source_params_st *src = &pipes[pipe_idx].pipe.src;
+			display_pipe_dest_params_st *dst = &pipes[pipe_idx].pipe.dest;
 
 			dst->vstartup_start = context->bw_ctx.dml.vba.VStartup[pipe_idx_unsplit];
 			dst->vupdate_offset = context->bw_ctx.dml.vba.VUpdateOffsetPix[pipe_idx_unsplit];
@@ -2800,7 +2880,6 @@ bool dcn20_validate_bandwidth(struct dc *dc, struct dc_state *context,
 	ASSERT(false);
 
 restore_dml_state:
-	memcpy(&context->bw_ctx.dml, &dc->dml, sizeof(struct display_mode_lib));
 	context->bw_ctx.dml.soc.dram_clock_change_latency_us = p_state_latency_us;
 
 	return voltage_supported;
@@ -2886,6 +2965,7 @@ static struct resource_funcs dcn20_res_pool_funcs = {
 	.populate_dml_writeback_from_context = dcn20_populate_dml_writeback_from_context,
 	.get_default_swizzle_mode = dcn20_get_default_swizzle_mode,
 	.set_mcif_arb_params = dcn20_set_mcif_arb_params,
+	.populate_dml_pipes = dcn20_populate_dml_pipes_from_context,
 	.find_first_free_match_stream_enc_for_link = dcn10_find_first_free_match_stream_enc_for_link
 };
 
@@ -2894,8 +2974,6 @@ bool dcn20_dwbc_create(struct dc_context *ctx, struct resource_pool *pool)
 	int i;
 	uint32_t pipe_count = pool->res_cap->num_dwb;
 
-	ASSERT(pipe_count > 0);
-
 	for (i = 0; i < pipe_count; i++) {
 		struct dcn20_dwbc *dwbc20 = kzalloc(sizeof(struct dcn20_dwbc),
 						    GFP_KERNEL);
@@ -2941,7 +3019,7 @@ bool dcn20_mmhubbub_create(struct dc_context *ctx, struct resource_pool *pool)
 	return true;
 }
 
-struct pp_smu_funcs *dcn20_pp_smu_create(struct dc_context *ctx)
+static struct pp_smu_funcs *dcn20_pp_smu_create(struct dc_context *ctx)
 {
 	struct pp_smu_funcs *pp_smu = kzalloc(sizeof(*pp_smu), GFP_KERNEL);
 
@@ -2956,7 +3034,7 @@ struct pp_smu_funcs *dcn20_pp_smu_create(struct dc_context *ctx)
 	return pp_smu;
 }
 
-void dcn20_pp_smu_destroy(struct pp_smu_funcs **pp_smu)
+static void dcn20_pp_smu_destroy(struct pp_smu_funcs **pp_smu)
 {
 	if (pp_smu && *pp_smu) {
 		kfree(*pp_smu);
@@ -2964,7 +3042,7 @@ void dcn20_pp_smu_destroy(struct pp_smu_funcs **pp_smu)
 	}
 }
 
-static void cap_soc_clocks(
+void dcn20_cap_soc_clocks(
 		struct _vcs_dpi_soc_bounding_box_st *bb,
 		struct pp_smu_nv_clock_table max_clocks)
 {
@@ -3031,10 +3109,10 @@ static void cap_soc_clocks(
 	}
 }
 
-static void update_bounding_box(struct dc *dc, struct _vcs_dpi_soc_bounding_box_st *bb,
+void dcn20_update_bounding_box(struct dc *dc, struct _vcs_dpi_soc_bounding_box_st *bb,
 		struct pp_smu_nv_clock_table *max_clocks, unsigned int *uclk_states, unsigned int num_states)
 {
-	struct _vcs_dpi_voltage_scaling_st calculated_states[MAX_CLOCK_LIMIT_STATES] = {0};
+	struct _vcs_dpi_voltage_scaling_st calculated_states[MAX_CLOCK_LIMIT_STATES];
 	int i;
 	int num_calculated_states = 0;
 	int min_dcfclk = 0;
@@ -3042,12 +3120,18 @@ static void update_bounding_box(struct dc *dc, struct _vcs_dpi_soc_bounding_box_
 	if (num_states == 0)
 		return;
 
+	memset(calculated_states, 0, sizeof(calculated_states));
+
 	if (dc->bb_overrides.min_dcfclk_mhz > 0)
 		min_dcfclk = dc->bb_overrides.min_dcfclk_mhz;
-	else
-		// Accounting for SOC/DCF relationship, we can go as high as
-		// 506Mhz in Vmin.  We need to code 507 since SMU will round down to 506.
-		min_dcfclk = 507;
+	else {
+		if (ASICREV_IS_NAVI12_P(dc->ctx->asic_id.hw_internal_rev))
+			min_dcfclk = 310;
+		else
+			// Accounting for SOC/DCF relationship, we can go as high as
+			// 506Mhz in Vmin.
+			min_dcfclk = 506;
+	}
 
 	for (i = 0; i < num_states; i++) {
 		int min_fclk_required_by_uclk;
@@ -3087,7 +3171,7 @@ static void update_bounding_box(struct dc *dc, struct _vcs_dpi_soc_bounding_box_
 	bb->clock_limits[num_calculated_states].state = bb->num_states;
 }
 
-static void patch_bounding_box(struct dc *dc, struct _vcs_dpi_soc_bounding_box_st *bb)
+void dcn20_patch_bounding_box(struct dc *dc, struct _vcs_dpi_soc_bounding_box_st *bb)
 {
 	kernel_fpu_begin();
 	if ((int)(bb->sr_exit_time_us * 1000) != dc->bb_overrides.sr_exit_time_ns
@@ -3286,14 +3370,14 @@ static bool init_soc_bounding_box(struct dc *dc,
 		}
 
 		if (clock_limits_available && uclk_states_available && num_states)
-			update_bounding_box(dc, loaded_bb, &max_clocks, uclk_states, num_states);
+			dcn20_update_bounding_box(dc, loaded_bb, &max_clocks, uclk_states, num_states);
 		else if (clock_limits_available)
-			cap_soc_clocks(loaded_bb, max_clocks);
+			dcn20_cap_soc_clocks(loaded_bb, max_clocks);
 	}
 
 	loaded_ip->max_num_otg = pool->base.res_cap->num_timing_generator;
 	loaded_ip->max_num_dpp = pool->base.pipe_count;
-	patch_bounding_box(dc, loaded_bb);
+	dcn20_patch_bounding_box(dc, loaded_bb);
 
 	return true;
 }
@@ -3339,6 +3423,7 @@ static bool construct(
 	dc->caps.post_blend_color_processing = true;
 	dc->caps.force_dp_tps4_for_cp2520 = true;
 	dc->caps.hw_3d_lut = true;
+	dc->caps.extended_aux_timeout_support = true;
 
 	if (dc->ctx->dce_environment == DCE_ENV_PRODUCTION_DRV) {
 		dc->debug = debug_defaults_drv;
diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.h b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.h
index 44f95aa0d61e..fef473d68a4a 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.h
@@ -95,9 +95,12 @@ struct display_stream_compressor *dcn20_dsc_create(
 	struct dc_context *ctx, uint32_t inst);
 void dcn20_dsc_destroy(struct display_stream_compressor **dsc);
 
-struct pp_smu_funcs *dcn20_pp_smu_create(struct dc_context *ctx);
-void dcn20_pp_smu_destroy(struct pp_smu_funcs **pp_smu);
-
+void dcn20_patch_bounding_box(struct dc *dc, struct _vcs_dpi_soc_bounding_box_st *bb);
+void dcn20_cap_soc_clocks(
+		struct _vcs_dpi_soc_bounding_box_st *bb,
+		struct pp_smu_nv_clock_table max_clocks);
+void dcn20_update_bounding_box(struct dc *dc, struct _vcs_dpi_soc_bounding_box_st *bb,
+		struct pp_smu_nv_clock_table *max_clocks, unsigned int *uclk_states, unsigned int num_states);
 struct hubp *dcn20_hubp_create(
 	struct dc_context *ctx,
 	uint32_t inst);
@@ -116,6 +119,31 @@ void dcn20_set_mcif_arb_params(
 		display_e2e_pipe_params_st *pipes,
 		int pipe_cnt);
 bool dcn20_validate_bandwidth(struct dc *dc, struct dc_state *context, bool fast_validate);
+void dcn20_merge_pipes_for_validate(
+		struct dc *dc,
+		struct dc_state *context);
+int dcn20_validate_apply_pipe_split_flags(
+		struct dc *dc,
+		struct dc_state *context,
+		int vlevel,
+		bool *split);
+#ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT
+bool dcn20_validate_dsc(struct dc *dc, struct dc_state *new_ctx);
+#endif
+void dcn20_split_stream_for_mpc(
+		struct resource_context *res_ctx,
+		const struct resource_pool *pool,
+		struct pipe_ctx *primary_pipe,
+		struct pipe_ctx *secondary_pipe);
+bool dcn20_split_stream_for_odm(
+		struct resource_context *res_ctx,
+		const struct resource_pool *pool,
+		struct pipe_ctx *prev_odm_pipe,
+		struct pipe_ctx *next_odm_pipe);
+struct pipe_ctx *dcn20_find_secondary_pipe(struct dc *dc,
+		struct resource_context *res_ctx,
+		const struct resource_pool *pool,
+		const struct pipe_ctx *primary_pipe);
 bool dcn20_fast_validate_bw(
 		struct dc *dc,
 		struct dc_state *context,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_stream_encoder.c
index 5ab9d6240498..4b3401616434 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_stream_encoder.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_stream_encoder.c
@@ -578,6 +578,10 @@ static const struct stream_encoder_funcs dcn20_str_enc_funcs = {
 	.set_avmute = enc1_stream_encoder_set_avmute,
 	.dig_connect_to_otg  = enc1_dig_connect_to_otg,
 	.dig_source_otg = enc1_dig_source_otg,
+
+	.dp_get_pixel_format =
+		enc1_stream_encoder_dp_get_pixel_format,
+
 #ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT
 	.enc_read_state = enc2_read_state,
 	.dp_set_dsc_config = enc2_dp_set_dsc_config,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn21/Makefile b/drivers/gpu/drm/amd/display/dc/dcn21/Makefile
index 8cd9de8b1a7a..14113ccf498d 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn21/Makefile
+++ b/drivers/gpu/drm/amd/display/dc/dcn21/Makefile
@@ -1,9 +1,24 @@
 #
 # Makefile for DCN21.
 
-DCN21 = dcn21_hubp.o dcn21_hubbub.o dcn21_resource.o
+DCN21 = dcn21_hubp.o dcn21_hubbub.o dcn21_resource.o dcn21_hwseq.o dcn21_link_encoder.o
 
-CFLAGS_$(AMDDALPATH)/dc/dcn21/dcn21_resource.o := -mhard-float -msse -mpreferred-stack-boundary=4
+CFLAGS_$(AMDDALPATH)/dc/dcn21/dcn21_resource.o := -mhard-float -msse
+
+ifdef CONFIG_CC_IS_GCC
+ifeq ($(call cc-ifversion, -lt, 0701, y), y)
+IS_OLD_GCC = 1
+endif
+endif
+
+ifdef IS_OLD_GCC
+# Stack alignment mismatch, proceed with caution.
+# GCC < 7.1 cannot compile code using `double` and -mpreferred-stack-boundary=3
+# (8B stack alignment).
+CFLAGS_$(AMDDALPATH)/dc/dcn21/dcn21_resource.o += -mpreferred-stack-boundary=4
+else
+CFLAGS_$(AMDDALPATH)/dc/dcn21/dcn21_resource.o += -msse2
+endif
 
 AMD_DAL_DCN21 = $(addprefix $(AMDDALPATH)/dc/dcn21/,$(DCN21))
 
diff --git a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hubbub.c b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hubbub.c
index d1266741763b..f546260c15b7 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hubbub.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hubbub.c
@@ -22,6 +22,7 @@
  * Authors: AMD
  *
  */
+#include <linux/delay.h>
 #include "dm_services.h"
 #include "dcn20/dcn20_hubbub.h"
 #include "dcn21_hubbub.h"
@@ -51,7 +52,7 @@
 #ifdef NUM_VMID
 #undef NUM_VMID
 #endif
-#define NUM_VMID 1
+#define NUM_VMID 16
 
 static uint32_t convert_and_clamp(
 	uint32_t wm_ns,
@@ -71,56 +72,76 @@ static uint32_t convert_and_clamp(
 void dcn21_dchvm_init(struct hubbub *hubbub)
 {
 	struct dcn20_hubbub *hubbub1 = TO_DCN20_HUBBUB(hubbub);
+	uint32_t riommu_active;
+	int i;
 
 	//Init DCHVM block
 	REG_UPDATE(DCHVM_CTRL0, HOSTVM_INIT_REQ, 1);
 
 	//Poll until RIOMMU_ACTIVE = 1
-	//TODO: Figure out interval us and retry count
-	REG_WAIT(DCHVM_RIOMMU_STAT0, RIOMMU_ACTIVE, 1, 5, 100);
+	for (i = 0; i < 100; i++) {
+		REG_GET(DCHVM_RIOMMU_STAT0, RIOMMU_ACTIVE, &riommu_active);
 
-	//Reflect the power status of DCHUBBUB
-	REG_UPDATE(DCHVM_RIOMMU_CTRL0, HOSTVM_POWERSTATUS, 1);
+		if (riommu_active)
+			break;
+		else
+			udelay(5);
+	}
+
+	if (riommu_active) {
+		//Reflect the power status of DCHUBBUB
+		REG_UPDATE(DCHVM_RIOMMU_CTRL0, HOSTVM_POWERSTATUS, 1);
 
-	//Start rIOMMU prefetching
-	REG_UPDATE(DCHVM_RIOMMU_CTRL0, HOSTVM_PREFETCH_REQ, 1);
+		//Start rIOMMU prefetching
+		REG_UPDATE(DCHVM_RIOMMU_CTRL0, HOSTVM_PREFETCH_REQ, 1);
 
-	// Enable dynamic clock gating
-	REG_UPDATE_4(DCHVM_CLK_CTRL,
-					HVM_DISPCLK_R_GATE_DIS, 0,
-					HVM_DISPCLK_G_GATE_DIS, 0,
-					HVM_DCFCLK_R_GATE_DIS, 0,
-					HVM_DCFCLK_G_GATE_DIS, 0);
+		// Enable dynamic clock gating
+		REG_UPDATE_4(DCHVM_CLK_CTRL,
+						HVM_DISPCLK_R_GATE_DIS, 0,
+						HVM_DISPCLK_G_GATE_DIS, 0,
+						HVM_DCFCLK_R_GATE_DIS, 0,
+						HVM_DCFCLK_G_GATE_DIS, 0);
 
-	//Poll until HOSTVM_PREFETCH_DONE = 1
-	//TODO: Figure out interval us and retry count
-	REG_WAIT(DCHVM_RIOMMU_STAT0, HOSTVM_PREFETCH_DONE, 1, 5, 100);
+		//Poll until HOSTVM_PREFETCH_DONE = 1
+		REG_WAIT(DCHVM_RIOMMU_STAT0, HOSTVM_PREFETCH_DONE, 1, 5, 100);
+	}
 }
 
-static int hubbub21_init_dchub(struct hubbub *hubbub,
+int hubbub21_init_dchub(struct hubbub *hubbub,
 		struct dcn_hubbub_phys_addr_config *pa_config)
 {
 	struct dcn20_hubbub *hubbub1 = TO_DCN20_HUBBUB(hubbub);
+	struct dcn_vmid_page_table_config phys_config;
 
 	REG_SET(DCN_VM_FB_LOCATION_BASE, 0,
-		FB_BASE, pa_config->system_aperture.fb_base);
+			FB_BASE, pa_config->system_aperture.fb_base >> 24);
 	REG_SET(DCN_VM_FB_LOCATION_TOP, 0,
-			FB_TOP, pa_config->system_aperture.fb_top);
+			FB_TOP, pa_config->system_aperture.fb_top >> 24);
 	REG_SET(DCN_VM_FB_OFFSET, 0,
-			FB_OFFSET, pa_config->system_aperture.fb_offset);
+			FB_OFFSET, pa_config->system_aperture.fb_offset >> 24);
 	REG_SET(DCN_VM_AGP_BOT, 0,
-			AGP_BOT, pa_config->system_aperture.agp_bot);
+			AGP_BOT, pa_config->system_aperture.agp_bot >> 24);
 	REG_SET(DCN_VM_AGP_TOP, 0,
-			AGP_TOP, pa_config->system_aperture.agp_top);
+			AGP_TOP, pa_config->system_aperture.agp_top >> 24);
 	REG_SET(DCN_VM_AGP_BASE, 0,
-			AGP_BASE, pa_config->system_aperture.agp_base);
+			AGP_BASE, pa_config->system_aperture.agp_base >> 24);
+
+	if (pa_config->gart_config.page_table_start_addr != pa_config->gart_config.page_table_end_addr) {
+		phys_config.page_table_start_addr = pa_config->gart_config.page_table_start_addr >> 12;
+		phys_config.page_table_end_addr = pa_config->gart_config.page_table_end_addr >> 12;
+		phys_config.page_table_base_addr = pa_config->gart_config.page_table_base_addr | 1; //Note: hack
+		phys_config.depth = 0;
+		phys_config.block_size = 0;
+		// Init VMID 0 based on PA config
+		dcn20_vmid_setup(&hubbub1->vmid[0], &phys_config);
+	}
 
 	dcn21_dchvm_init(hubbub);
 
 	return NUM_VMID;
 }
 
-static void hubbub21_program_urgent_watermarks(
+void hubbub21_program_urgent_watermarks(
 		struct hubbub *hubbub,
 		struct dcn_watermark_set *watermarks,
 		unsigned int refclk_mhz,
@@ -160,6 +181,13 @@ static void hubbub21_program_urgent_watermarks(
 		REG_SET(DCHUBBUB_ARB_FRAC_URG_BW_NOM_A, 0,
 				DCHUBBUB_ARB_FRAC_URG_BW_NOM_A, watermarks->a.frac_urg_bw_nom);
 	}
+	if (safe_to_lower || watermarks->a.urgent_latency_ns > hubbub1->watermarks.a.urgent_latency_ns) {
+		hubbub1->watermarks.a.urgent_latency_ns = watermarks->a.urgent_latency_ns;
+		prog_wm_value = convert_and_clamp(watermarks->a.urgent_latency_ns,
+				refclk_mhz, 0x1fffff);
+		REG_SET(DCHUBBUB_ARB_REFCYC_PER_TRIP_TO_MEMORY_A, 0,
+				DCHUBBUB_ARB_REFCYC_PER_TRIP_TO_MEMORY_A, prog_wm_value);
+	}
 
 	/* clock state B */
 	if (safe_to_lower || watermarks->b.urgent_ns > hubbub1->watermarks.b.urgent_ns) {
@@ -192,6 +220,14 @@ static void hubbub21_program_urgent_watermarks(
 				DCHUBBUB_ARB_FRAC_URG_BW_NOM_B, watermarks->a.frac_urg_bw_nom);
 	}
 
+	if (safe_to_lower || watermarks->b.urgent_latency_ns > hubbub1->watermarks.b.urgent_latency_ns) {
+		hubbub1->watermarks.b.urgent_latency_ns = watermarks->b.urgent_latency_ns;
+		prog_wm_value = convert_and_clamp(watermarks->b.urgent_latency_ns,
+				refclk_mhz, 0x1fffff);
+		REG_SET(DCHUBBUB_ARB_REFCYC_PER_TRIP_TO_MEMORY_B, 0,
+				DCHUBBUB_ARB_REFCYC_PER_TRIP_TO_MEMORY_B, prog_wm_value);
+	}
+
 	/* clock state C */
 	if (safe_to_lower || watermarks->c.urgent_ns > hubbub1->watermarks.c.urgent_ns) {
 		hubbub1->watermarks.c.urgent_ns = watermarks->c.urgent_ns;
@@ -223,6 +259,14 @@ static void hubbub21_program_urgent_watermarks(
 				DCHUBBUB_ARB_FRAC_URG_BW_NOM_C, watermarks->a.frac_urg_bw_nom);
 	}
 
+	if (safe_to_lower || watermarks->c.urgent_latency_ns > hubbub1->watermarks.c.urgent_latency_ns) {
+		hubbub1->watermarks.c.urgent_latency_ns = watermarks->c.urgent_latency_ns;
+		prog_wm_value = convert_and_clamp(watermarks->c.urgent_latency_ns,
+				refclk_mhz, 0x1fffff);
+		REG_SET(DCHUBBUB_ARB_REFCYC_PER_TRIP_TO_MEMORY_C, 0,
+				DCHUBBUB_ARB_REFCYC_PER_TRIP_TO_MEMORY_C, prog_wm_value);
+	}
+
 	/* clock state D */
 	if (safe_to_lower || watermarks->d.urgent_ns > hubbub1->watermarks.d.urgent_ns) {
 		hubbub1->watermarks.d.urgent_ns = watermarks->d.urgent_ns;
@@ -253,9 +297,17 @@ static void hubbub21_program_urgent_watermarks(
 		REG_SET(DCHUBBUB_ARB_FRAC_URG_BW_NOM_D, 0,
 				DCHUBBUB_ARB_FRAC_URG_BW_NOM_D, watermarks->a.frac_urg_bw_nom);
 	}
+
+	if (safe_to_lower || watermarks->d.urgent_latency_ns > hubbub1->watermarks.d.urgent_latency_ns) {
+		hubbub1->watermarks.d.urgent_latency_ns = watermarks->d.urgent_latency_ns;
+		prog_wm_value = convert_and_clamp(watermarks->d.urgent_latency_ns,
+				refclk_mhz, 0x1fffff);
+		REG_SET(DCHUBBUB_ARB_REFCYC_PER_TRIP_TO_MEMORY_D, 0,
+				DCHUBBUB_ARB_REFCYC_PER_TRIP_TO_MEMORY_D, prog_wm_value);
+	}
 }
 
-static void hubbub21_program_stutter_watermarks(
+void hubbub21_program_stutter_watermarks(
 		struct hubbub *hubbub,
 		struct dcn_watermark_set *watermarks,
 		unsigned int refclk_mhz,
@@ -389,7 +441,7 @@ static void hubbub21_program_stutter_watermarks(
 	}
 }
 
-static void hubbub21_program_pstate_watermarks(
+void hubbub21_program_pstate_watermarks(
 		struct hubbub *hubbub,
 		struct dcn_watermark_set *watermarks,
 		unsigned int refclk_mhz,
@@ -564,17 +616,26 @@ void hubbub21_wm_read_state(struct hubbub *hubbub,
 			DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_D, &s->dram_clk_chanage);
 }
 
+void hubbub21_apply_DEDCN21_147_wa(struct hubbub *hubbub)
+{
+	struct dcn20_hubbub *hubbub1 = TO_DCN20_HUBBUB(hubbub);
+	uint32_t prog_wm_value;
+
+	prog_wm_value = REG_READ(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_A);
+	REG_WRITE(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_A, prog_wm_value);
+}
 
 static const struct hubbub_funcs hubbub21_funcs = {
 	.update_dchub = hubbub2_update_dchub,
 	.init_dchub_sys_ctx = hubbub21_init_dchub,
-	.init_vm_ctx = NULL,
+	.init_vm_ctx = hubbub2_init_vm_ctx,
 	.dcc_support_swizzle = hubbub2_dcc_support_swizzle,
 	.dcc_support_pixel_format = hubbub2_dcc_support_pixel_format,
 	.get_dcc_compression_cap = hubbub2_get_dcc_compression_cap,
 	.wm_read_state = hubbub21_wm_read_state,
 	.get_dchub_ref_freq = hubbub2_get_dchub_ref_freq,
 	.program_watermarks = hubbub21_program_watermarks,
+	.apply_DEDCN21_147_wa = hubbub21_apply_DEDCN21_147_wa,
 };
 
 void hubbub21_construct(struct dcn20_hubbub *hubbub,
@@ -592,4 +653,5 @@ void hubbub21_construct(struct dcn20_hubbub *hubbub,
 	hubbub->masks = hubbub_mask;
 
 	hubbub->debug_test_index_pstate = 0xB;
+	hubbub->detile_buf_size = 164 * 1024; /* 164KB for DCN2.0 */
 }
diff --git a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hubbub.h b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hubbub.h
index 6ff3cdb89178..c4840dfb1fa5 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hubbub.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hubbub.h
@@ -36,6 +36,10 @@
 	SR(DCHUBBUB_ARB_FRAC_URG_BW_FLIP_B),\
 	SR(DCHUBBUB_ARB_FRAC_URG_BW_FLIP_C),\
 	SR(DCHUBBUB_ARB_FRAC_URG_BW_FLIP_D),\
+	SR(DCHUBBUB_ARB_REFCYC_PER_TRIP_TO_MEMORY_A),\
+	SR(DCHUBBUB_ARB_REFCYC_PER_TRIP_TO_MEMORY_B),\
+	SR(DCHUBBUB_ARB_REFCYC_PER_TRIP_TO_MEMORY_C),\
+	SR(DCHUBBUB_ARB_REFCYC_PER_TRIP_TO_MEMORY_D),\
 	SR(DCHUBBUB_ARB_HOSTVM_CNTL), \
 	SR(DCHVM_CTRL0), \
 	SR(DCHVM_MEM_CTRL), \
@@ -44,16 +48,9 @@
 	SR(DCHVM_RIOMMU_STAT0)
 
 #define HUBBUB_REG_LIST_DCN21()\
-	HUBBUB_REG_LIST_DCN_COMMON(), \
+	HUBBUB_REG_LIST_DCN20_COMMON(), \
 	HUBBUB_SR_WATERMARK_REG_LIST(), \
-	HUBBUB_HVM_REG_LIST(), \
-	SR(DCHUBBUB_CRC_CTRL), \
-	SR(DCN_VM_FB_LOCATION_BASE),\
-	SR(DCN_VM_FB_LOCATION_TOP),\
-	SR(DCN_VM_FB_OFFSET),\
-	SR(DCN_VM_AGP_BOT),\
-	SR(DCN_VM_AGP_TOP),\
-	SR(DCN_VM_AGP_BASE)
+	HUBBUB_HVM_REG_LIST()
 
 #define HUBBUB_MASK_SH_LIST_HVM(mask_sh) \
 	HUBBUB_SF(DCHUBBUB_ARB_DF_REQ_OUTSTAND, DCHUBBUB_ARB_MIN_REQ_OUTSTAND_COMMIT_THRESHOLD, mask_sh), \
@@ -102,7 +99,7 @@
 	HUBBUB_SF(DCHUBBUB_ARB_REFCYC_PER_TRIP_TO_MEMORY_D, DCHUBBUB_ARB_REFCYC_PER_TRIP_TO_MEMORY_D, mask_sh)
 
 #define HUBBUB_MASK_SH_LIST_DCN21(mask_sh)\
-	HUBBUB_MASK_SH_LIST_HVM(mask_sh),\
+	HUBBUB_MASK_SH_LIST_HVM(mask_sh), \
 	HUBBUB_MASK_SH_LIST_DCN_COMMON(mask_sh), \
 	HUBBUB_MASK_SH_LIST_STUTTER(mask_sh), \
 	HUBBUB_SF(DCHUBBUB_GLOBAL_TIMER_CNTL, DCHUBBUB_GLOBAL_TIMER_REFDIV, mask_sh), \
@@ -114,11 +111,28 @@
 	HUBBUB_SF(DCN_VM_AGP_BASE, AGP_BASE, mask_sh)
 
 void dcn21_dchvm_init(struct hubbub *hubbub);
+int hubbub21_init_dchub(struct hubbub *hubbub,
+		struct dcn_hubbub_phys_addr_config *pa_config);
 void hubbub21_program_watermarks(
 		struct hubbub *hubbub,
 		struct dcn_watermark_set *watermarks,
 		unsigned int refclk_mhz,
 		bool safe_to_lower);
+void hubbub21_program_urgent_watermarks(
+		struct hubbub *hubbub,
+		struct dcn_watermark_set *watermarks,
+		unsigned int refclk_mhz,
+		bool safe_to_lower);
+void hubbub21_program_stutter_watermarks(
+		struct hubbub *hubbub,
+		struct dcn_watermark_set *watermarks,
+		unsigned int refclk_mhz,
+		bool safe_to_lower);
+void hubbub21_program_pstate_watermarks(
+		struct hubbub *hubbub,
+		struct dcn_watermark_set *watermarks,
+		unsigned int refclk_mhz,
+		bool safe_to_lower);
 
 void hubbub21_wm_read_state(struct hubbub *hubbub,
 		struct dcn_hubbub_wm *wm);
diff --git a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hubp.c b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hubp.c
index a00af513aa2b..2f5a5867e674 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hubp.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hubp.c
@@ -22,6 +22,8 @@
  * Authors: AMD
  *
  */
+
+#include "dcn10/dcn10_hubp.h"
 #include "dcn21_hubp.h"
 
 #include "dm_services.h"
@@ -202,7 +204,7 @@ static struct hubp_funcs dcn21_hubp_funcs = {
 	.hubp_enable_tripleBuffer = hubp2_enable_triplebuffer,
 	.hubp_is_triplebuffer_enabled = hubp2_is_triplebuffer_enabled,
 	.hubp_program_surface_flip_and_addr = hubp2_program_surface_flip_and_addr,
-	.hubp_program_surface_config = hubp2_program_surface_config,
+	.hubp_program_surface_config = hubp1_program_surface_config,
 	.hubp_is_flip_pending = hubp1_is_flip_pending,
 	.hubp_setup = hubp21_setup,
 	.hubp_setup_interdependent = hubp2_setup_interdependent,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hwseq.c b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hwseq.c
new file mode 100644
index 000000000000..b25215cadf85
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hwseq.c
@@ -0,0 +1,122 @@
+/*
+ * Copyright 2016 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#include "dm_services.h"
+#include "dm_helpers.h"
+#include "core_types.h"
+#include "resource.h"
+#include "dce/dce_hwseq.h"
+#include "dcn20/dcn20_hwseq.h"
+#include "vmid.h"
+#include "reg_helper.h"
+#include "hw/clk_mgr.h"
+
+
+#define DC_LOGGER_INIT(logger)
+
+#define CTX \
+	hws->ctx
+#define REG(reg)\
+	hws->regs->reg
+
+#undef FN
+#define FN(reg_name, field_name) \
+	hws->shifts->field_name, hws->masks->field_name
+
+/* Temporary read settings, future will get values from kmd directly */
+static void mmhub_update_page_table_config(struct dcn_hubbub_phys_addr_config *config,
+		struct dce_hwseq *hws)
+{
+	uint32_t page_table_base_hi;
+	uint32_t page_table_base_lo;
+
+	REG_GET(VM_CONTEXT0_PAGE_TABLE_BASE_ADDR_HI32,
+			PAGE_DIRECTORY_ENTRY_HI32, &page_table_base_hi);
+	REG_GET(VM_CONTEXT0_PAGE_TABLE_BASE_ADDR_LO32,
+			PAGE_DIRECTORY_ENTRY_LO32, &page_table_base_lo);
+
+	config->gart_config.page_table_base_addr = ((uint64_t)page_table_base_hi << 32) | page_table_base_lo;
+
+}
+
+static int dcn21_init_sys_ctx(struct dce_hwseq *hws, struct dc *dc, struct dc_phy_addr_space_config *pa_config)
+{
+	struct dcn_hubbub_phys_addr_config config;
+
+	config.system_aperture.fb_top = pa_config->system_aperture.fb_top;
+	config.system_aperture.fb_offset = pa_config->system_aperture.fb_offset;
+	config.system_aperture.fb_base = pa_config->system_aperture.fb_base;
+	config.system_aperture.agp_top = pa_config->system_aperture.agp_top;
+	config.system_aperture.agp_bot = pa_config->system_aperture.agp_bot;
+	config.system_aperture.agp_base = pa_config->system_aperture.agp_base;
+	config.gart_config.page_table_start_addr = pa_config->gart_config.page_table_start_addr;
+	config.gart_config.page_table_end_addr = pa_config->gart_config.page_table_end_addr;
+	config.gart_config.page_table_base_addr = pa_config->gart_config.page_table_base_addr;
+
+	mmhub_update_page_table_config(&config, hws);
+
+	return dc->res_pool->hubbub->funcs->init_dchub_sys_ctx(dc->res_pool->hubbub, &config);
+}
+
+// work around for Renoir s0i3, if register is programmed, bypass golden init.
+
+static bool dcn21_s0i3_golden_init_wa(struct dc *dc)
+{
+	struct dce_hwseq *hws = dc->hwseq;
+	uint32_t value = 0;
+
+	value = REG_READ(MICROSECOND_TIME_BASE_DIV);
+
+	return value != 0x00120464;
+}
+
+void dcn21_exit_optimized_pwr_state(
+		const struct dc *dc,
+		struct dc_state *context)
+{
+	dc->clk_mgr->funcs->update_clocks(
+			dc->clk_mgr,
+			context,
+			false);
+}
+
+void dcn21_optimize_pwr_state(
+		const struct dc *dc,
+		struct dc_state *context)
+{
+	dc->clk_mgr->funcs->update_clocks(
+			dc->clk_mgr,
+			context,
+			true);
+}
+
+void dcn21_hw_sequencer_construct(struct dc *dc)
+{
+	dcn20_hw_sequencer_construct(dc);
+	dc->hwss.init_sys_ctx = dcn21_init_sys_ctx;
+	dc->hwss.s0i3_golden_init_wa = dcn21_s0i3_golden_init_wa;
+	dc->hwss.optimize_pwr_state = dcn21_optimize_pwr_state;
+	dc->hwss.exit_optimized_pwr_state = dcn21_exit_optimized_pwr_state;
+}
diff --git a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hwseq.h b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hwseq.h
new file mode 100644
index 000000000000..be67b62e6fb1
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hwseq.h
@@ -0,0 +1,33 @@
+/*
+* Copyright 2016 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#ifndef __DC_HWSS_DCN21_H__
+#define __DC_HWSS_DCN21_H__
+
+struct dc;
+
+void dcn21_hw_sequencer_construct(struct dc *dc);
+
+#endif /* __DC_HWSS_DCN21_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_link_encoder.c b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_link_encoder.c
new file mode 100644
index 000000000000..e8a504ca5890
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_link_encoder.c
@@ -0,0 +1,470 @@
+/*
+ * Copyright 2012-15 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#include "reg_helper.h"
+
+#include <linux/delay.h>
+#include "core_types.h"
+#include "link_encoder.h"
+#include "dcn21_link_encoder.h"
+#include "stream_encoder.h"
+
+#include "i2caux_interface.h"
+#include "dc_bios_types.h"
+
+#include "gpio_service_interface.h"
+
+#define CTX \
+	enc10->base.ctx
+#define DC_LOGGER \
+	enc10->base.ctx->logger
+
+#define REG(reg)\
+	(enc10->link_regs->reg)
+
+#undef FN
+#define FN(reg_name, field_name) \
+	enc10->link_shift->field_name, enc10->link_mask->field_name
+
+#define IND_REG(index) \
+	(enc10->link_regs->index)
+
+static struct mpll_cfg dcn21_mpll_cfg_ref[] = {
+	// RBR
+	{
+		.hdmimode_enable = 0,
+		.ref_range = 1,
+		.ref_clk_mpllb_div = 1,
+		.mpllb_ssc_en = 1,
+		.mpllb_div5_clk_en = 1,
+		.mpllb_multiplier = 238,
+		.mpllb_fracn_en = 0,
+		.mpllb_fracn_quot = 0,
+		.mpllb_fracn_rem = 0,
+		.mpllb_fracn_den = 1,
+		.mpllb_ssc_up_spread = 0,
+		.mpllb_ssc_peak = 44237,
+		.mpllb_ssc_stepsize = 59454,
+		.mpllb_div_clk_en = 0,
+		.mpllb_div_multiplier = 0,
+		.mpllb_hdmi_div = 0,
+		.mpllb_tx_clk_div = 2,
+		.tx_vboost_lvl = 5,
+		.mpllb_pmix_en = 1,
+		.mpllb_word_div2_en = 0,
+		.mpllb_ana_v2i = 2,
+		.mpllb_ana_freq_vco = 2,
+		.mpllb_ana_cp_int = 9,
+		.mpllb_ana_cp_prop = 15,
+		.hdmi_pixel_clk_div = 0,
+	},
+	// HBR
+	{
+		.hdmimode_enable = 0,
+		.ref_range = 1,
+		.ref_clk_mpllb_div = 1,
+		.mpllb_ssc_en = 1,
+		.mpllb_div5_clk_en = 1,
+		.mpllb_multiplier = 192,
+		.mpllb_fracn_en = 1,
+		.mpllb_fracn_quot = 32768,
+		.mpllb_fracn_rem = 0,
+		.mpllb_fracn_den = 1,
+		.mpllb_ssc_up_spread = 0,
+		.mpllb_ssc_peak = 36864,
+		.mpllb_ssc_stepsize = 49545,
+		.mpllb_div_clk_en = 0,
+		.mpllb_div_multiplier = 0,
+		.mpllb_hdmi_div = 0,
+		.mpllb_tx_clk_div = 1,
+		.tx_vboost_lvl = 5,
+		.mpllb_pmix_en = 1,
+		.mpllb_word_div2_en = 0,
+		.mpllb_ana_v2i = 2,
+		.mpllb_ana_freq_vco = 3,
+		.mpllb_ana_cp_int = 9,
+		.mpllb_ana_cp_prop = 15,
+		.hdmi_pixel_clk_div = 0,
+	},
+	//HBR2
+	{
+		.hdmimode_enable = 0,
+		.ref_range = 1,
+		.ref_clk_mpllb_div = 1,
+		.mpllb_ssc_en = 1,
+		.mpllb_div5_clk_en = 1,
+		.mpllb_multiplier = 192,
+		.mpllb_fracn_en = 1,
+		.mpllb_fracn_quot = 32768,
+		.mpllb_fracn_rem = 0,
+		.mpllb_fracn_den = 1,
+		.mpllb_ssc_up_spread = 0,
+		.mpllb_ssc_peak = 36864,
+		.mpllb_ssc_stepsize = 49545,
+		.mpllb_div_clk_en = 0,
+		.mpllb_div_multiplier = 0,
+		.mpllb_hdmi_div = 0,
+		.mpllb_tx_clk_div = 0,
+		.tx_vboost_lvl = 5,
+		.mpllb_pmix_en = 1,
+		.mpllb_word_div2_en = 0,
+		.mpllb_ana_v2i = 2,
+		.mpllb_ana_freq_vco = 3,
+		.mpllb_ana_cp_int = 9,
+		.mpllb_ana_cp_prop = 15,
+		.hdmi_pixel_clk_div = 0,
+	},
+	//HBR3
+	{
+		.hdmimode_enable = 0,
+		.ref_range = 1,
+		.ref_clk_mpllb_div = 1,
+		.mpllb_ssc_en = 1,
+		.mpllb_div5_clk_en = 1,
+		.mpllb_multiplier = 304,
+		.mpllb_fracn_en = 1,
+		.mpllb_fracn_quot = 49152,
+		.mpllb_fracn_rem = 0,
+		.mpllb_fracn_den = 1,
+		.mpllb_ssc_up_spread = 0,
+		.mpllb_ssc_peak = 55296,
+		.mpllb_ssc_stepsize = 74318,
+		.mpllb_div_clk_en = 0,
+		.mpllb_div_multiplier = 0,
+		.mpllb_hdmi_div = 0,
+		.mpllb_tx_clk_div = 0,
+		.tx_vboost_lvl = 5,
+		.mpllb_pmix_en = 1,
+		.mpllb_word_div2_en = 0,
+		.mpllb_ana_v2i = 2,
+		.mpllb_ana_freq_vco = 1,
+		.mpllb_ana_cp_int = 7,
+		.mpllb_ana_cp_prop = 16,
+		.hdmi_pixel_clk_div = 0,
+	},
+};
+
+
+static bool update_cfg_data(
+		struct dcn10_link_encoder *enc10,
+		const struct dc_link_settings *link_settings,
+		struct dpcssys_phy_seq_cfg *cfg)
+{
+	int i;
+
+	cfg->load_sram_fw = false;
+	cfg->use_calibration_setting = true;
+
+	//TODO: need to implement a proper lane mapping for Renoir.
+	for (i = 0; i < 4; i++)
+		cfg->lane_en[i] = true;
+
+	switch (link_settings->link_rate) {
+	case LINK_RATE_LOW:
+		cfg->mpll_cfg = dcn21_mpll_cfg_ref[0];
+		break;
+	case LINK_RATE_HIGH:
+		cfg->mpll_cfg = dcn21_mpll_cfg_ref[1];
+		break;
+	case LINK_RATE_HIGH2:
+		cfg->mpll_cfg = dcn21_mpll_cfg_ref[2];
+		break;
+	case LINK_RATE_HIGH3:
+		cfg->mpll_cfg = dcn21_mpll_cfg_ref[3];
+		break;
+	default:
+		DC_LOG_ERROR("%s: No supported link rate found %X!\n",
+				__func__, link_settings->link_rate);
+		return false;
+	}
+
+	return true;
+}
+
+void dcn21_link_encoder_get_max_link_cap(struct link_encoder *enc,
+	struct dc_link_settings *link_settings)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+	uint32_t value;
+
+	REG_GET(RDPCSTX_PHY_CNTL6, RDPCS_PHY_DPALT_DP4, &value);
+
+	if (!value && link_settings->lane_count > LANE_COUNT_TWO)
+		link_settings->lane_count = LANE_COUNT_TWO;
+}
+
+bool dcn21_link_encoder_is_in_alt_mode(struct link_encoder *enc)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+	uint32_t value;
+
+	REG_GET(RDPCSTX_PHY_CNTL6, RDPCS_PHY_DPALT_DISABLE, &value);
+
+	// if value == 1 alt mode is disabled, otherwise it is enabled
+	return !value;
+}
+
+bool dcn21_link_encoder_acquire_phy(struct link_encoder *enc)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+	int value;
+
+	if (enc->features.flags.bits.DP_IS_USB_C) {
+		REG_GET(RDPCSTX_PHY_CNTL6,
+				RDPCS_PHY_DPALT_DISABLE, &value);
+
+		if (value == 1) {
+			ASSERT(0);
+			return false;
+		}
+		REG_UPDATE(RDPCSTX_PHY_CNTL6,
+				RDPCS_PHY_DPALT_DISABLE_ACK, 0);
+
+		udelay(40);
+
+		REG_GET(RDPCSTX_PHY_CNTL6,
+						RDPCS_PHY_DPALT_DISABLE, &value);
+		if (value == 1) {
+			ASSERT(0);
+			REG_UPDATE(RDPCSTX_PHY_CNTL6,
+					RDPCS_PHY_DPALT_DISABLE_ACK, 1);
+			return false;
+		}
+	}
+
+	REG_UPDATE(RDPCSTX_PHY_CNTL6, RDPCS_PHY_DP_REF_CLK_EN, 1);
+
+	return true;
+}
+
+
+
+static void dcn21_link_encoder_release_phy(struct link_encoder *enc)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+
+	if (enc->features.flags.bits.DP_IS_USB_C) {
+		REG_UPDATE(RDPCSTX_PHY_CNTL6,
+				RDPCS_PHY_DPALT_DISABLE_ACK, 1);
+	}
+
+	REG_UPDATE(RDPCSTX_PHY_CNTL6, RDPCS_PHY_DP_REF_CLK_EN, 0);
+
+}
+
+void dcn21_link_encoder_enable_dp_output(
+	struct link_encoder *enc,
+	const struct dc_link_settings *link_settings,
+	enum clock_source_id clock_source)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+	struct dcn21_link_encoder *enc21 = (struct dcn21_link_encoder *) enc10;
+	struct dpcssys_phy_seq_cfg *cfg = &enc21->phy_seq_cfg;
+
+	if (!dcn21_link_encoder_acquire_phy(enc))
+		return;
+
+	if (!enc->ctx->dc->debug.avoid_vbios_exec_table) {
+		dcn10_link_encoder_enable_dp_output(enc, link_settings, clock_source);
+		return;
+	}
+
+	if (!update_cfg_data(enc10, link_settings, cfg))
+		return;
+
+	enc1_configure_encoder(enc10, link_settings);
+
+	dcn10_link_encoder_setup(enc, SIGNAL_TYPE_DISPLAY_PORT);
+
+}
+
+void dcn21_link_encoder_enable_dp_mst_output(
+	struct link_encoder *enc,
+	const struct dc_link_settings *link_settings,
+	enum clock_source_id clock_source)
+{
+	if (!dcn21_link_encoder_acquire_phy(enc))
+		return;
+
+	dcn10_link_encoder_enable_dp_mst_output(enc, link_settings, clock_source);
+}
+
+void dcn21_link_encoder_disable_output(
+	struct link_encoder *enc,
+	enum signal_type signal)
+{
+	dcn10_link_encoder_disable_output(enc, signal);
+
+	if (dc_is_dp_signal(signal))
+		dcn21_link_encoder_release_phy(enc);
+}
+
+
+static const struct link_encoder_funcs dcn21_link_enc_funcs = {
+#ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT
+	.read_state = link_enc2_read_state,
+#endif
+	.validate_output_with_stream =
+		dcn10_link_encoder_validate_output_with_stream,
+	.hw_init = enc2_hw_init,
+	.setup = dcn10_link_encoder_setup,
+	.enable_tmds_output = dcn10_link_encoder_enable_tmds_output,
+	.enable_dp_output = dcn21_link_encoder_enable_dp_output,
+	.enable_dp_mst_output = dcn21_link_encoder_enable_dp_mst_output,
+	.disable_output = dcn21_link_encoder_disable_output,
+	.dp_set_lane_settings = dcn10_link_encoder_dp_set_lane_settings,
+	.dp_set_phy_pattern = dcn10_link_encoder_dp_set_phy_pattern,
+	.update_mst_stream_allocation_table =
+		dcn10_link_encoder_update_mst_stream_allocation_table,
+	.psr_program_dp_dphy_fast_training =
+			dcn10_psr_program_dp_dphy_fast_training,
+	.psr_program_secondary_packet = dcn10_psr_program_secondary_packet,
+	.connect_dig_be_to_fe = dcn10_link_encoder_connect_dig_be_to_fe,
+	.enable_hpd = dcn10_link_encoder_enable_hpd,
+	.disable_hpd = dcn10_link_encoder_disable_hpd,
+	.is_dig_enabled = dcn10_is_dig_enabled,
+	.destroy = dcn10_link_encoder_destroy,
+	.fec_set_enable = enc2_fec_set_enable,
+	.fec_set_ready = enc2_fec_set_ready,
+	.fec_is_active = enc2_fec_is_active,
+	.get_dig_frontend = dcn10_get_dig_frontend,
+	.is_in_alt_mode = dcn21_link_encoder_is_in_alt_mode,
+	.get_max_link_cap = dcn21_link_encoder_get_max_link_cap,
+};
+
+void dcn21_link_encoder_construct(
+	struct dcn21_link_encoder *enc21,
+	const struct encoder_init_data *init_data,
+	const struct encoder_feature_support *enc_features,
+	const struct dcn10_link_enc_registers *link_regs,
+	const struct dcn10_link_enc_aux_registers *aux_regs,
+	const struct dcn10_link_enc_hpd_registers *hpd_regs,
+	const struct dcn10_link_enc_shift *link_shift,
+	const struct dcn10_link_enc_mask *link_mask)
+{
+	struct bp_encoder_cap_info bp_cap_info = {0};
+	const struct dc_vbios_funcs *bp_funcs = init_data->ctx->dc_bios->funcs;
+	enum bp_result result = BP_RESULT_OK;
+	struct dcn10_link_encoder *enc10 = &enc21->enc10;
+
+	enc10->base.funcs = &dcn21_link_enc_funcs;
+	enc10->base.ctx = init_data->ctx;
+	enc10->base.id = init_data->encoder;
+
+	enc10->base.hpd_source = init_data->hpd_source;
+	enc10->base.connector = init_data->connector;
+
+	enc10->base.preferred_engine = ENGINE_ID_UNKNOWN;
+
+	enc10->base.features = *enc_features;
+
+	enc10->base.transmitter = init_data->transmitter;
+
+	/* set the flag to indicate whether driver poll the I2C data pin
+	 * while doing the DP sink detect
+	 */
+
+/*	if (dal_adapter_service_is_feature_supported(as,
+		FEATURE_DP_SINK_DETECT_POLL_DATA_PIN))
+		enc10->base.features.flags.bits.
+			DP_SINK_DETECT_POLL_DATA_PIN = true;*/
+
+	enc10->base.output_signals =
+		SIGNAL_TYPE_DVI_SINGLE_LINK |
+		SIGNAL_TYPE_DVI_DUAL_LINK |
+		SIGNAL_TYPE_LVDS |
+		SIGNAL_TYPE_DISPLAY_PORT |
+		SIGNAL_TYPE_DISPLAY_PORT_MST |
+		SIGNAL_TYPE_EDP |
+		SIGNAL_TYPE_HDMI_TYPE_A;
+
+	/* For DCE 8.0 and 8.1, by design, UNIPHY is hardwired to DIG_BE.
+	 * SW always assign DIG_FE 1:1 mapped to DIG_FE for non-MST UNIPHY.
+	 * SW assign DIG_FE to non-MST UNIPHY first and MST last. So prefer
+	 * DIG is per UNIPHY and used by SST DP, eDP, HDMI, DVI and LVDS.
+	 * Prefer DIG assignment is decided by board design.
+	 * For DCE 8.0, there are only max 6 UNIPHYs, we assume board design
+	 * and VBIOS will filter out 7 UNIPHY for DCE 8.0.
+	 * By this, adding DIGG should not hurt DCE 8.0.
+	 * This will let DCE 8.1 share DCE 8.0 as much as possible
+	 */
+
+	enc10->link_regs = link_regs;
+	enc10->aux_regs = aux_regs;
+	enc10->hpd_regs = hpd_regs;
+	enc10->link_shift = link_shift;
+	enc10->link_mask = link_mask;
+
+	switch (enc10->base.transmitter) {
+	case TRANSMITTER_UNIPHY_A:
+		enc10->base.preferred_engine = ENGINE_ID_DIGA;
+	break;
+	case TRANSMITTER_UNIPHY_B:
+		enc10->base.preferred_engine = ENGINE_ID_DIGB;
+	break;
+	case TRANSMITTER_UNIPHY_C:
+		enc10->base.preferred_engine = ENGINE_ID_DIGC;
+	break;
+	case TRANSMITTER_UNIPHY_D:
+		enc10->base.preferred_engine = ENGINE_ID_DIGD;
+	break;
+	case TRANSMITTER_UNIPHY_E:
+		enc10->base.preferred_engine = ENGINE_ID_DIGE;
+	break;
+	case TRANSMITTER_UNIPHY_F:
+		enc10->base.preferred_engine = ENGINE_ID_DIGF;
+	break;
+	case TRANSMITTER_UNIPHY_G:
+		enc10->base.preferred_engine = ENGINE_ID_DIGG;
+	break;
+	default:
+		ASSERT_CRITICAL(false);
+		enc10->base.preferred_engine = ENGINE_ID_UNKNOWN;
+	}
+
+	/* default to one to mirror Windows behavior */
+	enc10->base.features.flags.bits.HDMI_6GB_EN = 1;
+
+	result = bp_funcs->get_encoder_cap_info(enc10->base.ctx->dc_bios,
+						enc10->base.id, &bp_cap_info);
+
+	/* Override features with DCE-specific values */
+	if (result == BP_RESULT_OK) {
+		enc10->base.features.flags.bits.IS_HBR2_CAPABLE =
+				bp_cap_info.DP_HBR2_EN;
+		enc10->base.features.flags.bits.IS_HBR3_CAPABLE =
+				bp_cap_info.DP_HBR3_EN;
+		enc10->base.features.flags.bits.HDMI_6GB_EN = bp_cap_info.HDMI_6GB_EN;
+		enc10->base.features.flags.bits.DP_IS_USB_C =
+				bp_cap_info.DP_IS_USB_C;
+	} else {
+		DC_LOG_WARNING("%s: Failed to get encoder_cap_info from VBIOS with error code %d!\n",
+				__func__,
+				result);
+	}
+	if (enc10->base.ctx->dc->debug.hdmi20_disable) {
+		enc10->base.features.flags.bits.HDMI_6GB_EN = 0;
+	}
+}
diff --git a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_link_encoder.h b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_link_encoder.h
new file mode 100644
index 000000000000..1d7a1a51f13d
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_link_encoder.h
@@ -0,0 +1,61 @@
+/*
+ * Copyright 2012-15 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ *  and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#ifndef __DC_LINK_ENCODER__DCN21_H__
+#define __DC_LINK_ENCODER__DCN21_H__
+
+#include "dcn20/dcn20_link_encoder.h"
+
+struct dcn21_link_encoder {
+	struct dcn10_link_encoder enc10;
+	struct dpcssys_phy_seq_cfg phy_seq_cfg;
+};
+
+#define LINK_ENCODER_MASK_SH_LIST_DCN21(mask_sh)\
+	LINK_ENCODER_MASK_SH_LIST_DCN20(mask_sh),\
+	LE_SF(UNIPHYA_CHANNEL_XBAR_CNTL, UNIPHY_CHANNEL0_XBAR_SOURCE, mask_sh),\
+	LE_SF(UNIPHYA_CHANNEL_XBAR_CNTL, UNIPHY_CHANNEL1_XBAR_SOURCE, mask_sh),\
+	LE_SF(UNIPHYA_CHANNEL_XBAR_CNTL, UNIPHY_CHANNEL2_XBAR_SOURCE, mask_sh),\
+	LE_SF(UNIPHYA_CHANNEL_XBAR_CNTL, UNIPHY_CHANNEL3_XBAR_SOURCE, mask_sh), \
+	SRI(RDPCSTX_PHY_FUSE2, RDPCSTX, id), \
+	SRI(RDPCSTX_PHY_FUSE3, RDPCSTX, id), \
+	SR(RDPCSTX0_RDPCSTX_SCRATCH)
+
+void dcn21_link_encoder_enable_dp_output(
+	struct link_encoder *enc,
+	const struct dc_link_settings *link_settings,
+	enum clock_source_id clock_source);
+
+void dcn21_link_encoder_construct(
+	struct dcn21_link_encoder *enc21,
+	const struct encoder_init_data *init_data,
+	const struct encoder_feature_support *enc_features,
+	const struct dcn10_link_enc_registers *link_regs,
+	const struct dcn10_link_enc_aux_registers *aux_regs,
+	const struct dcn10_link_enc_hpd_registers *hpd_regs,
+	const struct dcn10_link_enc_shift *link_shift,
+	const struct dcn10_link_enc_mask *link_mask);
+
+#endif
diff --git a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_resource.c b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_resource.c
index de182185fe1f..459bd9a5caed 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_resource.c
@@ -23,8 +23,6 @@
  *
  */
 
-#include <linux/slab.h>
-
 #include "dm_services.h"
 #include "dc.h"
 
@@ -42,11 +40,11 @@
 #include "irq/dcn21/irq_service_dcn21.h"
 #include "dcn20/dcn20_dpp.h"
 #include "dcn20/dcn20_optc.h"
-#include "dcn20/dcn20_hwseq.h"
+#include "dcn21/dcn21_hwseq.h"
 #include "dce110/dce110_hw_sequencer.h"
 #include "dcn20/dcn20_opp.h"
 #include "dcn20/dcn20_dsc.h"
-#include "dcn20/dcn20_link_encoder.h"
+#include "dcn21/dcn21_link_encoder.h"
 #include "dcn20/dcn20_stream_encoder.h"
 #include "dce/dce_clock_source.h"
 #include "dce/dce_audio.h"
@@ -84,8 +82,9 @@
 
 
 struct _vcs_dpi_ip_params_st dcn2_1_ip = {
-	.gpuvm_enable = 0,
-	.hostvm_enable = 0,
+	.odm_capable = 1,
+	.gpuvm_enable = 1,
+	.hostvm_enable = 1,
 	.gpuvm_max_page_table_levels = 1,
 	.hostvm_max_page_table_levels = 4,
 	.hostvm_cached_page_table_levels = 2,
@@ -205,11 +204,11 @@ struct _vcs_dpi_soc_bounding_box_st dcn2_1_soc = {
 				.state = 4,
 				.dcfclk_mhz = 810.0,
 				.fabricclk_mhz = 1600.0,
-				.dispclk_mhz = 1015.0,
-				.dppclk_mhz = 1015.0,
-				.phyclk_mhz = 810.0,
+				.dispclk_mhz = 1395.0,
+				.dppclk_mhz = 1285.0,
+				.phyclk_mhz = 1325.0,
 				.socclk_mhz = 953.0,
-				.dscclk_mhz = 318.334,
+				.dscclk_mhz = 489.0,
 				.dram_speed_mts = 4266.0,
 			},
 			/*Extra state, no dispclk ramping*/
@@ -217,18 +216,18 @@ struct _vcs_dpi_soc_bounding_box_st dcn2_1_soc = {
 				.state = 5,
 				.dcfclk_mhz = 810.0,
 				.fabricclk_mhz = 1600.0,
-				.dispclk_mhz = 1015.0,
-				.dppclk_mhz = 1015.0,
-				.phyclk_mhz = 810.0,
+				.dispclk_mhz = 1395.0,
+				.dppclk_mhz = 1285.0,
+				.phyclk_mhz = 1325.0,
 				.socclk_mhz = 953.0,
-				.dscclk_mhz = 318.334,
+				.dscclk_mhz = 489.0,
 				.dram_speed_mts = 4266.0,
 			},
 
 		},
 
-	.sr_exit_time_us = 9.0,
-	.sr_enter_plus_exit_time_us = 11.0,
+	.sr_exit_time_us = 12.5,
+	.sr_enter_plus_exit_time_us = 17.0,
 	.urgent_latency_us = 4.0,
 	.urgent_latency_pixel_data_only_us = 4.0,
 	.urgent_latency_pixel_mixed_with_vm_data_us = 4.0,
@@ -350,6 +349,30 @@ static const struct bios_registers bios_regs = {
 		NBIO_SR(BIOS_SCRATCH_6)
 };
 
+static const struct dce_dmcu_registers dmcu_regs = {
+		DMCU_DCN10_REG_LIST()
+};
+
+static const struct dce_dmcu_shift dmcu_shift = {
+		DMCU_MASK_SH_LIST_DCN10(__SHIFT)
+};
+
+static const struct dce_dmcu_mask dmcu_mask = {
+		DMCU_MASK_SH_LIST_DCN10(_MASK)
+};
+
+static const struct dce_abm_registers abm_regs = {
+		ABM_DCN20_REG_LIST()
+};
+
+static const struct dce_abm_shift abm_shift = {
+		ABM_MASK_SH_LIST_DCN20(__SHIFT)
+};
+
+static const struct dce_abm_mask abm_mask = {
+		ABM_MASK_SH_LIST_DCN20(_MASK)
+};
+
 #ifdef CONFIG_DRM_AMD_DC_DMUB
 static const struct dcn21_dmcub_registers dmcub_regs = {
 		DMCUB_REG_LIST_DCN()
@@ -628,6 +651,14 @@ static const struct dcn10_stream_enc_registers stream_enc_regs[] = {
 	stream_enc_regs(4),
 };
 
+static const struct dce110_aux_registers_shift aux_shift = {
+	DCN_AUX_MASK_SH_LIST(__SHIFT)
+};
+
+static const struct dce110_aux_registers_mask aux_mask = {
+	DCN_AUX_MASK_SH_LIST(_MASK)
+};
+
 static const struct dcn10_stream_encoder_shift se_shift = {
 		SE_COMMON_MASK_SH_LIST_DCN20(__SHIFT)
 };
@@ -636,6 +667,11 @@ static const struct dcn10_stream_encoder_mask se_mask = {
 		SE_COMMON_MASK_SH_LIST_DCN20(_MASK)
 };
 
+static void dcn21_pp_smu_destroy(struct pp_smu_funcs **pp_smu);
+
+static int dcn21_populate_dml_pipes_from_context(
+		struct dc *dc, struct resource_context *res_ctx, display_e2e_pipe_params_st *pipes);
+
 static struct input_pixel_processor *dcn21_ipp_create(
 	struct dc_context *ctx, uint32_t inst)
 {
@@ -683,7 +719,10 @@ static struct dce_aux *dcn21_aux_engine_create(
 
 	dce110_aux_engine_construct(aux_engine, ctx, inst,
 				    SW_AUX_TIMEOUT_PERIOD_MULTIPLIER * AUX_TIMEOUT_PERIOD,
-				    &aux_engine_regs[inst]);
+				    &aux_engine_regs[inst],
+					&aux_mask,
+					&aux_shift,
+					ctx->dc->caps.extended_aux_timeout_support);
 
 	return &aux_engine->base;
 }
@@ -726,11 +765,12 @@ static const struct resource_caps res_cap_rn = {
 		.num_timing_generator = 4,
 		.num_opp = 4,
 		.num_video_plane = 4,
-		.num_audio = 6, // 6 audio endpoints.  4 audio streams
+		.num_audio = 4, // 4 audio endpoints.  4 audio streams
 		.num_stream_encoder = 5,
 		.num_pll = 5,  // maybe 3 because the last two used for USB-c
 		.num_dwb = 1,
 		.num_ddc = 5,
+		.num_vmid = 1,
 #ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT
 		.num_dsc = 3,
 #endif
@@ -796,15 +836,15 @@ static const struct dc_debug_options debug_defaults_drv = {
 		.clock_trace = true,
 		.disable_pplib_clock_request = true,
 		.pipe_split_policy = MPC_SPLIT_AVOID_MULT_DISP,
-		.force_single_disp_pipe_split = true,
+		.force_single_disp_pipe_split = false,
 		.disable_dcc = DCC_ENABLE,
 		.vsr_support = true,
 		.performance_trace = false,
-		.max_downscale_src_width = 5120,/*upto 5K*/
+		.max_downscale_src_width = 3840,
 		.disable_pplib_wm_range = false,
 		.scl_reset_length10 = true,
 		.sanity_checks = true,
-		.disable_48mhz_pwrdwn = true,
+		.disable_48mhz_pwrdwn = false,
 };
 
 static const struct dc_debug_options debug_defaults_diags = {
@@ -939,7 +979,7 @@ static void destruct(struct dcn21_resource_pool *pool)
 		dcn_dccg_destroy(&pool->base.dccg);
 
 	if (pool->base.pp_smu != NULL)
-		dcn20_pp_smu_destroy(&pool->base.pp_smu);
+		dcn21_pp_smu_destroy(&pool->base.pp_smu);
 }
 
 
@@ -969,11 +1009,35 @@ static void calculate_wm_set_for_vlevel(
 #if defined(CONFIG_DRM_AMD_DC_DCN2_1)
 	wm_set->frac_urg_bw_nom = get_fraction_of_urgent_bandwidth(dml, pipes, pipe_cnt) * 1000;
 	wm_set->frac_urg_bw_flip = get_fraction_of_urgent_bandwidth_imm_flip(dml, pipes, pipe_cnt) * 1000;
+	wm_set->urgent_latency_ns = get_urgent_latency(dml, pipes, pipe_cnt) * 1000;
 #endif
 	dml->soc.dram_clock_change_latency_us = dram_clock_change_latency_cached;
 
 }
 
+static void patch_bounding_box(struct dc *dc, struct _vcs_dpi_soc_bounding_box_st *bb)
+{
+	kernel_fpu_begin();
+	if (dc->bb_overrides.sr_exit_time_ns) {
+		bb->sr_exit_time_us = dc->bb_overrides.sr_exit_time_ns / 1000.0;
+	}
+
+	if (dc->bb_overrides.sr_enter_plus_exit_time_ns) {
+		bb->sr_enter_plus_exit_time_us =
+				dc->bb_overrides.sr_enter_plus_exit_time_ns / 1000.0;
+	}
+
+	if (dc->bb_overrides.urgent_latency_ns) {
+		bb->urgent_latency_us = dc->bb_overrides.urgent_latency_ns / 1000.0;
+	}
+
+	if (dc->bb_overrides.dram_clock_change_latency_ns) {
+		bb->dram_clock_change_latency_us =
+				dc->bb_overrides.dram_clock_change_latency_ns / 1000.0;
+	}
+	kernel_fpu_end();
+}
+
 void dcn21_calculate_wm(
 		struct dc *dc, struct dc_state *context,
 		display_e2e_pipe_params_st *pipes,
@@ -988,6 +1052,8 @@ void dcn21_calculate_wm(
 
 	ASSERT(bw_params);
 
+	patch_bounding_box(dc, &context->bw_ctx.dml.soc);
+
 	for (i = 0, pipe_idx = 0, pipe_cnt = 0; i < dc->res_pool->pipe_count; i++) {
 			if (!context->res_ctx.pipe_ctx[i].stream)
 				continue;
@@ -1021,7 +1087,7 @@ void dcn21_calculate_wm(
 			pipe_cnt = dc->res_pool->funcs->populate_dml_pipes(dc,
 				&context->res_ctx, pipes);
 		else
-			pipe_cnt = dcn20_populate_dml_pipes_from_context(dc,
+			pipe_cnt = dcn21_populate_dml_pipes_from_context(dc,
 				&context->res_ctx, pipes);
 	}
 
@@ -1271,6 +1337,12 @@ struct display_stream_compressor *dcn21_dsc_create(
 
 static void update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_params)
 {
+	/*
+	TODO: Fix this function to calcualte correct values.
+	There are known issues with this function currently
+	that will need to be investigated. Use hardcoded known good values for now.
+
+
 	struct dcn21_resource_pool *pool = TO_DCN21_RES_POOL(dc->res_pool);
 	struct clk_limit_table *clk_table = &bw_params->clk_table;
 	int i;
@@ -1278,7 +1350,6 @@ static void update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_param
 	dcn2_1_ip.max_num_otg = pool->base.res_cap->num_timing_generator;
 	dcn2_1_ip.max_num_dpp = pool->base.pipe_count;
 	dcn2_1_soc.num_chans = bw_params->num_channels;
-	dcn2_1_soc.num_states = 0;
 
 	for (i = 0; i < clk_table->num_entries; i++) {
 
@@ -1286,10 +1357,11 @@ static void update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_param
 		dcn2_1_soc.clock_limits[i].dcfclk_mhz = clk_table->entries[i].dcfclk_mhz;
 		dcn2_1_soc.clock_limits[i].fabricclk_mhz = clk_table->entries[i].fclk_mhz;
 		dcn2_1_soc.clock_limits[i].socclk_mhz = clk_table->entries[i].socclk_mhz;
-		/* This is probably wrong, TODO: find correct calculation */
 		dcn2_1_soc.clock_limits[i].dram_speed_mts = clk_table->entries[i].memclk_mhz * 16 / 1000;
-		dcn2_1_soc.num_states++;
 	}
+	dcn2_1_soc.clock_limits[i] = dcn2_1_soc.clock_limits[i - i];
+	dcn2_1_soc.num_states = i;
+	*/
 }
 
 /* Temporary Place holder until we can get them from fuse */
@@ -1317,32 +1389,42 @@ static struct dpm_clocks dummy_clocks = {
 
 };
 
-enum pp_smu_status dummy_set_wm_ranges(struct pp_smu *pp,
+static enum pp_smu_status dummy_set_wm_ranges(struct pp_smu *pp,
 		struct pp_smu_wm_range_sets *ranges)
 {
 	return PP_SMU_RESULT_OK;
 }
 
-enum pp_smu_status dummy_get_dpm_clock_table(struct pp_smu *pp,
+static enum pp_smu_status dummy_get_dpm_clock_table(struct pp_smu *pp,
 		struct dpm_clocks *clock_table)
 {
 	*clock_table = dummy_clocks;
 	return PP_SMU_RESULT_OK;
 }
 
-struct pp_smu_funcs *dcn21_pp_smu_create(struct dc_context *ctx)
+static struct pp_smu_funcs *dcn21_pp_smu_create(struct dc_context *ctx)
 {
 	struct pp_smu_funcs *pp_smu = kzalloc(sizeof(*pp_smu), GFP_KERNEL);
 
-	pp_smu->ctx.ver = PP_SMU_VER_RN;
+	if (!pp_smu)
+		return pp_smu;
+
+	if (IS_FPGA_MAXIMUS_DC(ctx->dce_environment) || IS_DIAG_DC(ctx->dce_environment)) {
+		pp_smu->ctx.ver = PP_SMU_VER_RN;
+		pp_smu->rn_funcs.get_dpm_clock_table = dummy_get_dpm_clock_table;
+		pp_smu->rn_funcs.set_wm_ranges = dummy_set_wm_ranges;
+	} else {
 
-	pp_smu->rn_funcs.get_dpm_clock_table = dummy_get_dpm_clock_table;
-	pp_smu->rn_funcs.set_wm_ranges = dummy_set_wm_ranges;
+		dm_pp_get_funcs(ctx, pp_smu);
+
+		if (pp_smu->ctx.ver != PP_SMU_VER_RN)
+			pp_smu = memset(pp_smu, 0, sizeof(struct pp_smu_funcs));
+	}
 
 	return pp_smu;
 }
 
-void dcn21_pp_smu_destroy(struct pp_smu_funcs **pp_smu)
+static void dcn21_pp_smu_destroy(struct pp_smu_funcs **pp_smu)
 {
 	if (pp_smu && *pp_smu) {
 		kfree(*pp_smu);
@@ -1400,6 +1482,7 @@ static struct dce_hwseq *dcn21_hwseq_create(
 		hws->regs = &hwseq_reg;
 		hws->shifts = &hwseq_shift;
 		hws->masks = &hwseq_mask;
+		hws->wa.DEGVIDCN21 = true;
 	}
 	return hws;
 }
@@ -1418,10 +1501,152 @@ static const struct resource_create_funcs res_create_maximus_funcs = {
 	.create_hwseq = dcn21_hwseq_create,
 };
 
+static const struct encoder_feature_support link_enc_feature = {
+		.max_hdmi_deep_color = COLOR_DEPTH_121212,
+		.max_hdmi_pixel_clock = 600000,
+		.hdmi_ycbcr420_supported = true,
+		.dp_ycbcr420_supported = true,
+		.flags.bits.IS_HBR2_CAPABLE = true,
+		.flags.bits.IS_HBR3_CAPABLE = true,
+		.flags.bits.IS_TPS3_CAPABLE = true,
+		.flags.bits.IS_TPS4_CAPABLE = true
+};
+
+
+#define link_regs(id, phyid)\
+[id] = {\
+	LE_DCN10_REG_LIST(id), \
+	UNIPHY_DCN2_REG_LIST(phyid), \
+	SRI(DP_DPHY_INTERNAL_CTRL, DP, id) \
+}
+
+static const struct dcn10_link_enc_registers link_enc_regs[] = {
+	link_regs(0, A),
+	link_regs(1, B),
+	link_regs(2, C),
+	link_regs(3, D),
+	link_regs(4, E),
+};
+
+#define aux_regs(id)\
+[id] = {\
+	DCN2_AUX_REG_LIST(id)\
+}
+
+static const struct dcn10_link_enc_aux_registers link_enc_aux_regs[] = {
+		aux_regs(0),
+		aux_regs(1),
+		aux_regs(2),
+		aux_regs(3),
+		aux_regs(4)
+};
+
+#define hpd_regs(id)\
+[id] = {\
+	HPD_REG_LIST(id)\
+}
+
+static const struct dcn10_link_enc_hpd_registers link_enc_hpd_regs[] = {
+		hpd_regs(0),
+		hpd_regs(1),
+		hpd_regs(2),
+		hpd_regs(3),
+		hpd_regs(4)
+};
+
+static const struct dcn10_link_enc_shift le_shift = {
+	LINK_ENCODER_MASK_SH_LIST_DCN20(__SHIFT)
+};
+
+static const struct dcn10_link_enc_mask le_mask = {
+	LINK_ENCODER_MASK_SH_LIST_DCN20(_MASK)
+};
+
+static int map_transmitter_id_to_phy_instance(
+	enum transmitter transmitter)
+{
+	switch (transmitter) {
+	case TRANSMITTER_UNIPHY_A:
+		return 0;
+	break;
+	case TRANSMITTER_UNIPHY_B:
+		return 1;
+	break;
+	case TRANSMITTER_UNIPHY_C:
+		return 2;
+	break;
+	case TRANSMITTER_UNIPHY_D:
+		return 3;
+	break;
+	case TRANSMITTER_UNIPHY_E:
+		return 4;
+	break;
+	default:
+		ASSERT(0);
+		return 0;
+	}
+}
+
+static struct link_encoder *dcn21_link_encoder_create(
+	const struct encoder_init_data *enc_init_data)
+{
+	struct dcn21_link_encoder *enc21 =
+		kzalloc(sizeof(struct dcn21_link_encoder), GFP_KERNEL);
+	int link_regs_id;
+
+	if (!enc21)
+		return NULL;
+
+	link_regs_id =
+		map_transmitter_id_to_phy_instance(enc_init_data->transmitter);
+
+	dcn21_link_encoder_construct(enc21,
+				      enc_init_data,
+				      &link_enc_feature,
+				      &link_enc_regs[link_regs_id],
+				      &link_enc_aux_regs[enc_init_data->channel - 1],
+				      &link_enc_hpd_regs[enc_init_data->hpd_source],
+				      &le_shift,
+				      &le_mask);
+
+	return &enc21->enc10.base;
+}
+#define CTX ctx
+
+#define REG(reg_name) \
+	(DCN_BASE.instance[0].segment[mm ## reg_name ## _BASE_IDX] + mm ## reg_name)
+
+static uint32_t read_pipe_fuses(struct dc_context *ctx)
+{
+	uint32_t value = REG_READ(CC_DC_PIPE_DIS);
+	/* RV1 support max 4 pipes */
+	value = value & 0xf;
+	return value;
+}
+
+static int dcn21_populate_dml_pipes_from_context(
+		struct dc *dc, struct resource_context *res_ctx, display_e2e_pipe_params_st *pipes)
+{
+	uint32_t pipe_cnt = dcn20_populate_dml_pipes_from_context(dc, res_ctx, pipes);
+	int i;
+
+	for (i = 0; i < dc->res_pool->pipe_count; i++) {
+
+		if (!res_ctx->pipe_ctx[i].stream)
+			continue;
+
+		pipes[i].pipe.src.hostvm = 1;
+		pipes[i].pipe.src.gpuvm = 1;
+	}
+
+	return pipe_cnt;
+}
+
 static struct resource_funcs dcn21_res_pool_funcs = {
 	.destroy = dcn21_destroy_resource_pool,
-	.link_enc_create = dcn20_link_encoder_create,
+	.link_enc_create = dcn21_link_encoder_create,
 	.validate_bandwidth = dcn21_validate_bandwidth,
+	.populate_dml_pipes = dcn21_populate_dml_pipes_from_context,
 	.add_stream_to_ctx = dcn20_add_stream_to_ctx,
 	.remove_stream_from_ctx = dcn20_remove_stream_from_ctx,
 	.acquire_idle_pipe_for_layer = dcn20_acquire_idle_pipe_for_layer,
@@ -1437,9 +1662,11 @@ static bool construct(
 	struct dc *dc,
 	struct dcn21_resource_pool *pool)
 {
-	int i;
+	int i, j;
 	struct dc_context *ctx = dc->ctx;
 	struct irq_service_init_data init_data;
+	uint32_t pipe_fuses = read_pipe_fuses(ctx);
+	uint32_t num_pipes;
 
 	ctx->dc_bios->regs = &bios_regs;
 
@@ -1457,7 +1684,9 @@ static bool construct(
 	 *************************************************/
 	pool->base.underlay_pipe_index = NO_UNDERLAY_PIPE;
 
-	pool->base.pipe_count = 4;
+	/* max pipe num for ASIC before check pipe fuses */
+	pool->base.pipe_count = pool->base.res_cap->num_timing_generator;
+
 	dc->caps.max_downscale_ratio = 200;
 	dc->caps.i2c_speed_in_khz = 100;
 	dc->caps.max_cursor_size = 256;
@@ -1467,6 +1696,7 @@ static bool construct(
 	dc->caps.max_slave_planes = 1;
 	dc->caps.post_blend_color_processing = true;
 	dc->caps.force_dp_tps4_for_cp2520 = true;
+	dc->caps.extended_aux_timeout_support = true;
 
 	if (dc->ctx->dce_environment == DCE_ENV_PRODUCTION_DRV)
 		dc->debug = debug_defaults_drv;
@@ -1516,6 +1746,26 @@ static bool construct(
 		goto create_fail;
 	}
 
+	pool->base.dmcu = dcn20_dmcu_create(ctx,
+			&dmcu_regs,
+			&dmcu_shift,
+			&dmcu_mask);
+	if (pool->base.dmcu == NULL) {
+		dm_error("DC: failed to create dmcu!\n");
+		BREAK_TO_DEBUGGER();
+		goto create_fail;
+	}
+
+	pool->base.abm = dce_abm_create(ctx,
+			&abm_regs,
+			&abm_shift,
+			&abm_mask);
+	if (pool->base.abm == NULL) {
+		dm_error("DC: failed to create abm!\n");
+		BREAK_TO_DEBUGGER();
+		goto create_fail;
+	}
+
 #ifdef CONFIG_DRM_AMD_DC_DMUB
 	pool->base.dmcub = dcn21_dmcub_create(ctx,
 			&dmcub_regs,
@@ -1530,6 +1780,14 @@ static bool construct(
 
 	pool->base.pp_smu = dcn21_pp_smu_create(ctx);
 
+	num_pipes = dcn2_1_ip.max_num_dpp;
+
+	for (i = 0; i < dcn2_1_ip.max_num_dpp; i++)
+		if (pipe_fuses & 1 << i)
+			num_pipes--;
+	dcn2_1_ip.max_num_dpp = num_pipes;
+	dcn2_1_ip.max_num_otg = num_pipes;
+
 	dml_init_instance(&dc->dml, &dcn2_1_soc, &dcn2_1_ip, DML_PROJECT_DCN21);
 
 	init_data.ctx = dc->ctx;
@@ -1537,8 +1795,15 @@ static bool construct(
 	if (!pool->base.irqs)
 		goto create_fail;
 
+	j = 0;
 	/* mem input -> ipp -> dpp -> opp -> TG */
 	for (i = 0; i < pool->base.pipe_count; i++) {
+		/* if pipe is disabled, skip instance of HW pipe,
+		 * i.e, skip ASIC register instance
+		 */
+		if ((pipe_fuses & (1 << i)) != 0)
+			continue;
+
 		pool->base.hubps[i] = dcn21_hubp_create(ctx, i);
 		if (pool->base.hubps[i] == NULL) {
 			BREAK_TO_DEBUGGER();
@@ -1562,6 +1827,23 @@ static bool construct(
 				"DC: failed to create dpps!\n");
 			goto create_fail;
 		}
+
+		pool->base.opps[i] = dcn21_opp_create(ctx, i);
+		if (pool->base.opps[i] == NULL) {
+			BREAK_TO_DEBUGGER();
+			dm_error(
+				"DC: failed to create output pixel processor!\n");
+			goto create_fail;
+		}
+
+		pool->base.timing_generators[i] = dcn21_timing_generator_create(
+				ctx, i);
+		if (pool->base.timing_generators[i] == NULL) {
+			BREAK_TO_DEBUGGER();
+			dm_error("DC: failed to create tg!\n");
+			goto create_fail;
+		}
+		j++;
 	}
 
 	for (i = 0; i < pool->base.res_cap->num_ddc; i++) {
@@ -1582,27 +1864,9 @@ static bool construct(
 		pool->base.sw_i2cs[i] = NULL;
 	}
 
-	for (i = 0; i < pool->base.res_cap->num_opp; i++) {
-		pool->base.opps[i] = dcn21_opp_create(ctx, i);
-		if (pool->base.opps[i] == NULL) {
-			BREAK_TO_DEBUGGER();
-			dm_error(
-				"DC: failed to create output pixel processor!\n");
-			goto create_fail;
-		}
-	}
-
-	for (i = 0; i < pool->base.res_cap->num_timing_generator; i++) {
-		pool->base.timing_generators[i] = dcn21_timing_generator_create(
-				ctx, i);
-		if (pool->base.timing_generators[i] == NULL) {
-			BREAK_TO_DEBUGGER();
-			dm_error("DC: failed to create tg!\n");
-			goto create_fail;
-		}
-	}
-
-	pool->base.timing_generator_count = i;
+	pool->base.timing_generator_count = j;
+	pool->base.pipe_count = j;
+	pool->base.mpcc_count = j;
 
 	pool->base.mpc = dcn21_mpc_create(ctx);
 	if (pool->base.mpc == NULL) {
@@ -1645,7 +1909,7 @@ static bool construct(
 			&res_create_funcs : &res_create_maximus_funcs)))
 			goto create_fail;
 
-	dcn20_hw_sequencer_construct(dc);
+	dcn21_hw_sequencer_construct(dc);
 
 	dc->caps.max_planes =  pool->base.pipe_count;
 
diff --git a/drivers/gpu/drm/amd/display/dc/dm_cp_psp.h b/drivers/gpu/drm/amd/display/dc/dm_cp_psp.h
new file mode 100644
index 000000000000..626d22d437f4
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/dm_cp_psp.h
@@ -0,0 +1,49 @@
+/*
+ * Copyright 2018 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#ifndef DM_CP_PSP_IF__H
+#define DM_CP_PSP_IF__H
+
+struct dc_link;
+
+struct cp_psp_stream_config {
+	uint8_t otg_inst;
+	uint8_t link_enc_inst;
+	uint8_t stream_enc_inst;
+	void *dm_stream_ctx;
+	bool dpms_off;
+};
+
+struct cp_psp_funcs {
+	void (*update_stream_config)(void *handle, struct cp_psp_stream_config *config);
+};
+
+struct cp_psp {
+	void *handle;
+	struct cp_psp_funcs funcs;
+};
+
+
+#endif /* DM_CP_PSP_IF__H */
diff --git a/drivers/gpu/drm/amd/display/dc/dm_helpers.h b/drivers/gpu/drm/amd/display/dc/dm_helpers.h
index b6b4333737f2..94b75e942607 100644
--- a/drivers/gpu/drm/amd/display/dc/dm_helpers.h
+++ b/drivers/gpu/drm/amd/display/dc/dm_helpers.h
@@ -74,7 +74,7 @@ void dm_helpers_dp_mst_clear_payload_allocation_table(
 /*
  * Polls for ACT (allocation change trigger) handled and
  */
-bool dm_helpers_dp_mst_poll_for_allocation_change_trigger(
+enum act_return_status dm_helpers_dp_mst_poll_for_allocation_change_trigger(
 		struct dc_context *ctx,
 		const struct dc_stream_state *stream);
 /*
diff --git a/drivers/gpu/drm/amd/display/dc/dm_pp_smu.h b/drivers/gpu/drm/amd/display/dc/dm_pp_smu.h
index c03a441ee638..ef7df9ef6d7e 100644
--- a/drivers/gpu/drm/amd/display/dc/dm_pp_smu.h
+++ b/drivers/gpu/drm/amd/display/dc/dm_pp_smu.h
@@ -249,10 +249,8 @@ struct pp_smu_funcs_nv {
 };
 #endif
 
-#if defined(CONFIG_DRM_AMD_DC_DCN2_1)
-
 #define PP_SMU_NUM_SOCCLK_DPM_LEVELS  8
-#define PP_SMU_NUM_DCFCLK_DPM_LEVELS  4
+#define PP_SMU_NUM_DCFCLK_DPM_LEVELS  8
 #define PP_SMU_NUM_FCLK_DPM_LEVELS    4
 #define PP_SMU_NUM_MEMCLK_DPM_LEVELS  4
 
@@ -288,7 +286,6 @@ struct pp_smu_funcs_rn {
 	enum pp_smu_status (*get_dpm_clock_table) (struct pp_smu *pp,
 			struct dpm_clocks *clock_table);
 };
-#endif
 
 struct pp_smu_funcs {
 	struct pp_smu ctx;
diff --git a/drivers/gpu/drm/amd/display/dc/dml/Makefile b/drivers/gpu/drm/amd/display/dc/dml/Makefile
index 5b2a65b42403..8df251626e22 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/Makefile
+++ b/drivers/gpu/drm/amd/display/dc/dml/Makefile
@@ -24,15 +24,20 @@
 # It provides the general basic services required by other DAL
 # subcomponents.
 
-ifneq ($(call cc-option, -mpreferred-stack-boundary=4),)
-	cc_stack_align := -mpreferred-stack-boundary=4
-else ifneq ($(call cc-option, -mstack-alignment=16),)
-	cc_stack_align := -mstack-alignment=16
-endif
+dml_ccflags := -mhard-float -msse
 
-dml_ccflags := -mhard-float -msse $(cc_stack_align)
+ifdef CONFIG_CC_IS_GCC
+ifeq ($(call cc-ifversion, -lt, 0701, y), y)
+IS_OLD_GCC = 1
+endif
+endif
 
-ifdef CONFIG_CC_IS_CLANG
+ifdef IS_OLD_GCC
+# Stack alignment mismatch, proceed with caution.
+# GCC < 7.1 cannot compile code using `double` and -mpreferred-stack-boundary=3
+# (8B stack alignment).
+dml_ccflags += -mpreferred-stack-boundary=4
+else
 dml_ccflags += -msse2
 endif
 
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_mode_vba_20.c b/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_mode_vba_20.c
index 649883777f62..6c6c486b774a 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_mode_vba_20.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_mode_vba_20.c
@@ -2577,7 +2577,8 @@ static void dml20_DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPer
 			mode_lib->vba.MinActiveDRAMClockChangeMargin
 					+ mode_lib->vba.DRAMClockChangeLatency;
 
-	if (mode_lib->vba.MinActiveDRAMClockChangeMargin > 0) {
+	if (mode_lib->vba.MinActiveDRAMClockChangeMargin > 50) {
+		mode_lib->vba.DRAMClockChangeWatermark += 25;
 		mode_lib->vba.DRAMClockChangeSupport[0][0] = dm_dram_clock_change_vactive;
 	} else {
 		if (mode_lib->vba.SynchronizedVBlank || mode_lib->vba.NumberOfActivePlanes == 1) {
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_mode_vba_20v2.c b/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_mode_vba_20v2.c
index 0fafd693ffb4..3c70dd577292 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_mode_vba_20v2.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_mode_vba_20v2.c
@@ -38,6 +38,7 @@
 
 #define BPP_INVALID 0
 #define BPP_BLENDED_PIPE 0xffffffff
+#define DCN20_MAX_DSC_IMAGE_WIDTH 5184
 
 static double adjust_ReturnBW(
 		struct display_mode_lib *mode_lib,
@@ -2610,7 +2611,8 @@ static void dml20v2_DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndP
 			mode_lib->vba.MinActiveDRAMClockChangeMargin
 					+ mode_lib->vba.DRAMClockChangeLatency;
 
-	if (mode_lib->vba.MinActiveDRAMClockChangeMargin > 0) {
+	if (mode_lib->vba.MinActiveDRAMClockChangeMargin > 50) {
+		mode_lib->vba.DRAMClockChangeWatermark += 25;
 		mode_lib->vba.DRAMClockChangeSupport[0][0] = dm_dram_clock_change_vactive;
 	} else {
 		if (mode_lib->vba.SynchronizedVBlank || mode_lib->vba.NumberOfActivePlanes == 1) {
@@ -3901,6 +3903,10 @@ void dml20v2_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode
 				mode_lib->vba.MaximumSwathWidthInLineBuffer);
 	}
 	for (i = 0; i <= mode_lib->vba.soc.num_states; i++) {
+		double MaxMaxDispclkRoundedDown = RoundToDFSGranularityDown(
+			mode_lib->vba.MaxDispclk[mode_lib->vba.soc.num_states],
+			mode_lib->vba.DISPCLKDPPCLKVCOSpeed);
+
 		for (j = 0; j < 2; j++) {
 			mode_lib->vba.MaxDispclkRoundedDownToDFSGranularity = RoundToDFSGranularityDown(
 				mode_lib->vba.MaxDispclk[i],
@@ -3925,7 +3931,9 @@ void dml20v2_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode
 						&& i == mode_lib->vba.soc.num_states)
 					mode_lib->vba.PlaneRequiredDISPCLKWithODMCombine = mode_lib->vba.PixelClock[k] / 2
 							* (1 + mode_lib->vba.DISPCLKDPPCLKDSCCLKDownSpreading / 100.0);
-				if (mode_lib->vba.ODMCapability == false || mode_lib->vba.PlaneRequiredDISPCLKWithoutODMCombine <= mode_lib->vba.MaxDispclkRoundedDownToDFSGranularity) {
+				if (mode_lib->vba.ODMCapability == false ||
+						(locals->PlaneRequiredDISPCLKWithoutODMCombine <= MaxMaxDispclkRoundedDown
+							&& (!locals->DSCEnabled[k] || locals->HActive[k] <= DCN20_MAX_DSC_IMAGE_WIDTH))) {
 					locals->ODMCombineEnablePerState[i][k] = false;
 					mode_lib->vba.PlaneRequiredDISPCLK = mode_lib->vba.PlaneRequiredDISPCLKWithoutODMCombine;
 				} else {
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_rq_dlg_calc_20.c b/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_rq_dlg_calc_20.c
index 878bf4782ce6..2c7455e22a65 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_rq_dlg_calc_20.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_rq_dlg_calc_20.c
@@ -207,7 +207,7 @@ static void extract_rq_regs(struct display_mode_lib *mode_lib,
 	rq_regs->rq_regs_l.swath_height = dml_log2(rq_param.dlg.rq_l.swath_height);
 	rq_regs->rq_regs_c.swath_height = dml_log2(rq_param.dlg.rq_c.swath_height);
 
-	// FIXME: take the max between luma, chroma chunk size?
+	// TODO: take the max between luma, chroma chunk size?
 	// okay for now, as we are setting chunk_bytes to 8kb anyways
 	if (rq_param.sizing.rq_l.chunk_bytes >= 32 * 1024) { //32kb
 		rq_regs->drq_expansion_mode = 0;
@@ -677,7 +677,7 @@ static void get_surf_rq_param(struct display_mode_lib *mode_lib,
 	unsigned int meta_pitch = 0;
 	unsigned int ppe = mode_422 ? 2 : 1;
 
-	// FIXME check if ppe apply for both luma and chroma in 422 case
+	// TODO check if ppe apply for both luma and chroma in 422 case
 	if (is_chroma) {
 		vp_width = pipe_src_param.viewport_width_c / ppe;
 		vp_height = pipe_src_param.viewport_height_c;
@@ -959,7 +959,7 @@ static void dml20_rq_dlg_get_dlg_params(struct display_mode_lib *mode_lib,
 	// Source
 //             dcc_en              = src.dcc;
 	dual_plane = is_dual_plane((enum source_format_class)(src->source_format));
-	mode_422 = 0; // FIXME
+	mode_422 = 0; // TODO
 	access_dir = (src->source_scan == dm_vert); // vp access direction: horizontal or vertical accessed
 //      bytes_per_element_l = get_bytes_per_element(source_format_class(src.source_format), 0);
 //      bytes_per_element_c = get_bytes_per_element(source_format_class(src.source_format), 1);
@@ -1655,7 +1655,7 @@ static void calculate_ttu_cursor(struct display_mode_lib *mode_lib,
 		cur_width_ub = dml_ceil((double) cur_src_width / (double) cur_req_width, 1)
 				* (double) cur_req_width;
 		cur_req_per_width = cur_width_ub / (double) cur_req_width;
-		hactive_cur = (double) cur_src_width / hscl_ratio; // FIXME: oswin to think about what to do for cursor
+		hactive_cur = (double) cur_src_width / hscl_ratio; // TODO: oswin to think about what to do for cursor
 
 		if (vratio_pre_l <= 1.0) {
 			*refcyc_per_req_delivery_pre_cur = hactive_cur * ref_freq_to_pix_freq
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_rq_dlg_calc_20v2.c b/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_rq_dlg_calc_20v2.c
index ed8bf5f723c9..1e6aeb1bd2bf 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_rq_dlg_calc_20v2.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_rq_dlg_calc_20v2.c
@@ -207,7 +207,7 @@ static void extract_rq_regs(struct display_mode_lib *mode_lib,
 	rq_regs->rq_regs_l.swath_height = dml_log2(rq_param.dlg.rq_l.swath_height);
 	rq_regs->rq_regs_c.swath_height = dml_log2(rq_param.dlg.rq_c.swath_height);
 
-	// FIXME: take the max between luma, chroma chunk size?
+	// TODO: take the max between luma, chroma chunk size?
 	// okay for now, as we are setting chunk_bytes to 8kb anyways
 	if (rq_param.sizing.rq_l.chunk_bytes >= 32 * 1024) { //32kb
 		rq_regs->drq_expansion_mode = 0;
@@ -677,7 +677,7 @@ static void get_surf_rq_param(struct display_mode_lib *mode_lib,
 	unsigned int meta_pitch = 0;
 	unsigned int ppe = mode_422 ? 2 : 1;
 
-	// FIXME check if ppe apply for both luma and chroma in 422 case
+	// TODO check if ppe apply for both luma and chroma in 422 case
 	if (is_chroma) {
 		vp_width = pipe_src_param.viewport_width_c / ppe;
 		vp_height = pipe_src_param.viewport_height_c;
@@ -959,7 +959,7 @@ static void dml20v2_rq_dlg_get_dlg_params(struct display_mode_lib *mode_lib,
 	// Source
 //             dcc_en              = src.dcc;
 	dual_plane = is_dual_plane((enum source_format_class)(src->source_format));
-	mode_422 = 0; // FIXME
+	mode_422 = 0; // TODO
 	access_dir = (src->source_scan == dm_vert); // vp access direction: horizontal or vertical accessed
 //      bytes_per_element_l = get_bytes_per_element(source_format_class(src.source_format), 0);
 //      bytes_per_element_c = get_bytes_per_element(source_format_class(src.source_format), 1);
@@ -1655,7 +1655,7 @@ static void calculate_ttu_cursor(struct display_mode_lib *mode_lib,
 		cur_width_ub = dml_ceil((double) cur_src_width / (double) cur_req_width, 1)
 				* (double) cur_req_width;
 		cur_req_per_width = cur_width_ub / (double) cur_req_width;
-		hactive_cur = (double) cur_src_width / hscl_ratio; // FIXME: oswin to think about what to do for cursor
+		hactive_cur = (double) cur_src_width / hscl_ratio; // TODO: oswin to think about what to do for cursor
 
 		if (vratio_pre_l <= 1.0) {
 			*refcyc_per_req_delivery_pre_cur = hactive_cur * ref_freq_to_pix_freq
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn21/display_mode_vba_21.c b/drivers/gpu/drm/amd/display/dc/dml/dcn21/display_mode_vba_21.c
index 456cd0e3289c..ba77957aefe3 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn21/display_mode_vba_21.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn21/display_mode_vba_21.c
@@ -39,9 +39,6 @@
  * ways. Unless there is something clearly wrong with it the code should
  * remain as-is as it provides us with a guarantee from HW that it is correct.
  */
-
-typedef unsigned int uint;
-
 typedef struct {
 	double DPPCLK;
 	double DISPCLK;
@@ -68,6 +65,7 @@ typedef struct {
 
 #define BPP_INVALID 0
 #define BPP_BLENDED_PIPE 0xffffffff
+#define DCN21_MAX_DSC_IMAGE_WIDTH 5184
 
 static void DisplayPipeConfiguration(struct display_mode_lib *mode_lib);
 static void DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerformanceCalculation(
@@ -3382,6 +3380,8 @@ static unsigned int TruncToValidBPP(
 				return 30;
 			else if (DecimalBPP >= 24 && (DesiredBPP == 0 || DesiredBPP == 24))
 				return 24;
+			else if (DecimalBPP >= 18 && (DesiredBPP == 0 || DesiredBPP == 18))
+				return 18;
 			else
 				return BPP_INVALID;
 		}
@@ -3939,6 +3939,10 @@ void dml21_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l
 				mode_lib->vba.MaximumSwathWidthInLineBuffer);
 	}
 	for (i = 0; i <= mode_lib->vba.soc.num_states; i++) {
+		double MaxMaxDispclkRoundedDown = RoundToDFSGranularityDown(
+			mode_lib->vba.MaxDispclk[mode_lib->vba.soc.num_states],
+			mode_lib->vba.DISPCLKDPPCLKVCOSpeed);
+
 		for (j = 0; j < 2; j++) {
 			mode_lib->vba.MaxDispclkRoundedDownToDFSGranularity = RoundToDFSGranularityDown(
 				mode_lib->vba.MaxDispclk[i],
@@ -3968,7 +3972,9 @@ void dml21_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l
 						&& i == mode_lib->vba.soc.num_states)
 					mode_lib->vba.PlaneRequiredDISPCLKWithODMCombine = mode_lib->vba.PixelClock[k] / 2
 							* (1 + mode_lib->vba.DISPCLKDPPCLKDSCCLKDownSpreading / 100.0);
-				if (mode_lib->vba.ODMCapability == false || mode_lib->vba.PlaneRequiredDISPCLKWithoutODMCombine <= mode_lib->vba.MaxDispclkRoundedDownToDFSGranularity) {
+				if (mode_lib->vba.ODMCapability == false ||
+						(locals->PlaneRequiredDISPCLKWithoutODMCombine <= MaxMaxDispclkRoundedDown
+							&& (!locals->DSCEnabled[k] || locals->HActive[k] <= DCN21_MAX_DSC_IMAGE_WIDTH))) {
 					locals->ODMCombineEnablePerState[i][k] = false;
 					mode_lib->vba.PlaneRequiredDISPCLK = mode_lib->vba.PlaneRequiredDISPCLKWithoutODMCombine;
 				} else {
@@ -4774,7 +4780,7 @@ void dml21_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l
 				mode_lib->vba.MaximumReadBandwidthWithoutPrefetch = 0.0;
 				mode_lib->vba.MaximumReadBandwidthWithPrefetch = 0.0;
 				for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) {
-					uint m;
+					unsigned int m;
 
 					locals->cursor_bw[k] = 0;
 					locals->cursor_bw_pre[k] = 0;
@@ -5285,7 +5291,7 @@ static void CalculateWatermarksAndDRAMSpeedChangeSupport(
 	double SecondMinActiveDRAMClockChangeMarginOneDisplayInVBLank;
 	double FullDETBufferingTimeYStutterCriticalPlane = 0;
 	double TimeToFinishSwathTransferStutterCriticalPlane = 0;
-	uint k, j;
+	unsigned int k, j;
 
 	mode_lib->vba.TotalActiveDPP = 0;
 	mode_lib->vba.TotalDCCActiveDPP = 0;
@@ -5507,7 +5513,7 @@ static void CalculateDCFCLKDeepSleep(
 		double DPPCLK[],
 		double *DCFCLKDeepSleep)
 {
-	uint k;
+	unsigned int k;
 	double DisplayPipeLineDeliveryTimeLuma;
 	double DisplayPipeLineDeliveryTimeChroma;
 	//double   DCFCLKDeepSleepPerPlane[DC__NUM_DPP__MAX];
@@ -5727,7 +5733,7 @@ static void CalculatePixelDeliveryTimes(
 		double DisplayPipeRequestDeliveryTimeChromaPrefetch[])
 {
 	double req_per_swath_ub;
-	uint k;
+	unsigned int k;
 
 	for (k = 0; k < NumberOfActivePlanes; ++k) {
 		if (VRatio[k] <= 1) {
@@ -5869,7 +5875,7 @@ static void CalculateMetaAndPTETimes(
 	unsigned int dpte_groups_per_row_chroma_ub;
 	unsigned int num_group_per_lower_vm_stage;
 	unsigned int num_req_per_lower_vm_stage;
-	uint k;
+	unsigned int k;
 
 	for (k = 0; k < NumberOfActivePlanes; ++k) {
 		if (GPUVMEnable == true) {
diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h b/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h
index f4c1ef9046bf..cfacd6027467 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h
+++ b/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h
@@ -269,7 +269,7 @@ struct writeback_st {
 
 struct _vcs_dpi_display_output_params_st {
 	int dp_lanes;
-	int output_bpp;
+	double output_bpp;
 	int dsc_enable;
 	int wb_enable;
 	int num_active_wb;
@@ -318,6 +318,7 @@ struct _vcs_dpi_display_pipe_dest_params_st {
 	unsigned int vupdate_width;
 	unsigned int vready_offset;
 	unsigned char interlaced;
+	unsigned char embedded;
 	double pixel_rate_mhz;
 	unsigned char synchronized_vblank_all_planes;
 	unsigned char otg_inst;
diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.c b/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.c
index 65cf4edddaff..7f9a5621922f 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.c
@@ -375,6 +375,7 @@ static void fetch_pipe_params(struct display_mode_lib *mode_lib)
 
 		mode_lib->vba.pipe_plane[j] = mode_lib->vba.NumberOfActivePlanes;
 
+		mode_lib->vba.EmbeddedPanel[mode_lib->vba.NumberOfActivePlanes] = dst->embedded;
 		mode_lib->vba.DPPPerPlane[mode_lib->vba.NumberOfActivePlanes] = 1;
 		mode_lib->vba.SourceScan[mode_lib->vba.NumberOfActivePlanes] =
 				(enum scan_direction_class) (src->source_scan);
@@ -432,8 +433,12 @@ static void fetch_pipe_params(struct display_mode_lib *mode_lib)
 				dst->recout_width; // TODO: or should this be full_recout_width???...maybe only when in hsplit mode?
 		mode_lib->vba.ODMCombineEnabled[mode_lib->vba.NumberOfActivePlanes] =
 				dst->odm_combine;
+		mode_lib->vba.ODMCombineTypeEnabled[mode_lib->vba.NumberOfActivePlanes] =
+				dst->odm_combine;
 		mode_lib->vba.OutputFormat[mode_lib->vba.NumberOfActivePlanes] =
 				(enum output_format_class) (dout->output_format);
+		mode_lib->vba.OutputBpp[mode_lib->vba.NumberOfActivePlanes] =
+				dout->output_bpp;
 		mode_lib->vba.Output[mode_lib->vba.NumberOfActivePlanes] =
 				(enum output_encoder_class) (dout->output_type);
 
diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.h b/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.h
index 91decac50557..1540ffbe3979 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.h
+++ b/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.h
@@ -387,6 +387,7 @@ struct vba_vars_st {
 
 	/* vba mode support */
 	/*inputs*/
+	bool EmbeddedPanel[DC__NUM_DPP__MAX];
 	bool SupportGFX7CompatibleTilingIn32bppAnd64bpp;
 	double MaxHSCLRatio;
 	double MaxVSCLRatio;
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dml1_display_rq_dlg_calc.c b/drivers/gpu/drm/amd/display/dc/dml/dml1_display_rq_dlg_calc.c
index ad8571f5a142..4c3e9cc30167 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dml1_display_rq_dlg_calc.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/dml1_display_rq_dlg_calc.c
@@ -243,7 +243,7 @@ void dml1_extract_rq_regs(
 	rq_regs->rq_regs_l.swath_height = dml_log2(rq_param.dlg.rq_l.swath_height);
 	rq_regs->rq_regs_c.swath_height = dml_log2(rq_param.dlg.rq_c.swath_height);
 
-	/* FIXME: take the max between luma, chroma chunk size?
+	/* TODO: take the max between luma, chroma chunk size?
 	 * okay for now, as we are setting chunk_bytes to 8kb anyways
 	 */
 	if (rq_param.sizing.rq_l.chunk_bytes >= 32 * 1024) { /*32kb */
@@ -602,7 +602,7 @@ static void get_surf_rq_param(
 	unsigned int log2_dpte_group_length;
 	unsigned int func_meta_row_height, func_dpte_row_height;
 
-	/* FIXME check if ppe apply for both luma and chroma in 422 case */
+	/* TODO check if ppe apply for both luma and chroma in 422 case */
 	if (is_chroma) {
 		vp_width = pipe_src_param.viewport_width_c / ppe;
 		vp_height = pipe_src_param.viewport_height_c;
@@ -1141,7 +1141,7 @@ void dml1_rq_dlg_get_dlg_params(
 	ASSERT(disp_dlg_regs->refcyc_h_blank_end < (unsigned int) dml_pow(2, 13));
 	disp_dlg_regs->dlg_vblank_end = interlaced ? (vblank_end / 2) : vblank_end; /* 15 bits */
 
-	prefetch_xy_calc_in_dcfclk = 24.0; /* FIXME: ip_param */
+	prefetch_xy_calc_in_dcfclk = 24.0; /* TODO: ip_param */
 	min_dcfclk_mhz = dlg_sys_param.deepsleep_dcfclk_mhz;
 	t_calc_us = prefetch_xy_calc_in_dcfclk / min_dcfclk_mhz;
 	min_ttu_vblank = dlg_sys_param.t_urg_wm_us;
@@ -1182,7 +1182,7 @@ void dml1_rq_dlg_get_dlg_params(
 	dcc_en = e2e_pipe_param.pipe.src.dcc;
 	dual_plane = is_dual_plane(
 			(enum source_format_class) e2e_pipe_param.pipe.src.source_format);
-	mode_422 = 0; /* FIXME */
+	mode_422 = 0; /* TODO */
 	access_dir = (e2e_pipe_param.pipe.src.source_scan == dm_vert); /* vp access direction: horizontal or vertical accessed */
 	bytes_per_element_l = get_bytes_per_element(
 			(enum source_format_class) e2e_pipe_param.pipe.src.source_format,
@@ -1837,7 +1837,7 @@ void dml1_rq_dlg_get_dlg_params(
 		cur0_width_ub = dml_ceil((double) cur0_src_width / (double) cur0_req_width, 1)
 				* (double) cur0_req_width;
 		cur0_req_per_width = cur0_width_ub / (double) cur0_req_width;
-		hactive_cur0 = (double) cur0_src_width / hratios_cur0; /* FIXME: oswin to think about what to do for cursor */
+		hactive_cur0 = (double) cur0_src_width / hratios_cur0; /* TODO: oswin to think about what to do for cursor */
 
 		if (vratio_pre_l <= 1.0) {
 			refcyc_per_req_delivery_pre_cur0 = hactive_cur0 * ref_freq_to_pix_freq
diff --git a/drivers/gpu/drm/amd/display/dc/dsc/Makefile b/drivers/gpu/drm/amd/display/dc/dsc/Makefile
index b456cd23c6fa..970737217e53 100644
--- a/drivers/gpu/drm/amd/display/dc/dsc/Makefile
+++ b/drivers/gpu/drm/amd/display/dc/dsc/Makefile
@@ -1,15 +1,20 @@
 #
 # Makefile for the 'dsc' sub-component of DAL.
 
-ifneq ($(call cc-option, -mpreferred-stack-boundary=4),)
-	cc_stack_align := -mpreferred-stack-boundary=4
-else ifneq ($(call cc-option, -mstack-alignment=16),)
-	cc_stack_align := -mstack-alignment=16
-endif
+dsc_ccflags := -mhard-float -msse
 
-dsc_ccflags := -mhard-float -msse $(cc_stack_align)
+ifdef CONFIG_CC_IS_GCC
+ifeq ($(call cc-ifversion, -lt, 0701, y), y)
+IS_OLD_GCC = 1
+endif
+endif
 
-ifdef CONFIG_CC_IS_CLANG
+ifdef IS_OLD_GCC
+# Stack alignment mismatch, proceed with caution.
+# GCC < 7.1 cannot compile code using `double` and -mpreferred-stack-boundary=3
+# (8B stack alignment).
+dsc_ccflags += -mpreferred-stack-boundary=4
+else
 dsc_ccflags += -msse2
 endif
 
diff --git a/drivers/gpu/drm/amd/display/dc/dsc/dc_dsc.c b/drivers/gpu/drm/amd/display/dc/dsc/dc_dsc.c
index 5995bcdfed54..e60f760585e4 100644
--- a/drivers/gpu/drm/amd/display/dc/dsc/dc_dsc.c
+++ b/drivers/gpu/drm/amd/display/dc/dsc/dc_dsc.c
@@ -23,8 +23,7 @@
  */
 
 #ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT
-#include "dc.h"
-#include "core_types.h"
+#include "dc_hw_types.h"
 #include "dsc.h"
 #include <drm/drm_dp_helper.h>
 
@@ -47,6 +46,59 @@ const struct dc_dsc_policy dsc_policy = {
 
 /* This module's internal functions */
 
+static uint32_t dc_dsc_bandwidth_in_kbps_from_timing(
+	const struct dc_crtc_timing *timing)
+{
+	uint32_t bits_per_channel = 0;
+	uint32_t kbps;
+
+	if (timing->flags.DSC) {
+		kbps = (timing->pix_clk_100hz * timing->dsc_cfg.bits_per_pixel);
+		kbps = kbps / 160 + ((kbps % 160) ? 1 : 0);
+		return kbps;
+	}
+
+	switch (timing->display_color_depth) {
+	case COLOR_DEPTH_666:
+		bits_per_channel = 6;
+		break;
+	case COLOR_DEPTH_888:
+		bits_per_channel = 8;
+		break;
+	case COLOR_DEPTH_101010:
+		bits_per_channel = 10;
+		break;
+	case COLOR_DEPTH_121212:
+		bits_per_channel = 12;
+		break;
+	case COLOR_DEPTH_141414:
+		bits_per_channel = 14;
+		break;
+	case COLOR_DEPTH_161616:
+		bits_per_channel = 16;
+		break;
+	default:
+		break;
+	}
+
+	ASSERT(bits_per_channel != 0);
+
+	kbps = timing->pix_clk_100hz / 10;
+	kbps *= bits_per_channel;
+
+	if (timing->flags.Y_ONLY != 1) {
+		/*Only YOnly make reduce bandwidth by 1/3 compares to RGB*/
+		kbps *= 3;
+		if (timing->pixel_encoding == PIXEL_ENCODING_YCBCR420)
+			kbps /= 2;
+		else if (timing->pixel_encoding == PIXEL_ENCODING_YCBCR422)
+			kbps = kbps * 2 / 3;
+	}
+
+	return kbps;
+
+}
+
 static bool dsc_buff_block_size_from_dpcd(int dpcd_buff_block_size, int *buff_block_size)
 {
 
@@ -178,12 +230,11 @@ static bool dsc_bpp_increment_div_from_dpcd(int bpp_increment_dpcd, uint32_t *bp
 }
 
 static void get_dsc_enc_caps(
-	const struct dc *dc,
+	const struct display_stream_compressor *dsc,
 	struct dsc_enc_caps *dsc_enc_caps,
 	int pixel_clock_100Hz)
 {
 	// This is a static HW query, so we can use any DSC
-	struct display_stream_compressor *dsc = dc->res_pool->dscs[0];
 
 	memset(dsc_enc_caps, 0, sizeof(struct dsc_enc_caps));
 	if (dsc)
@@ -290,7 +341,7 @@ static void get_dsc_bandwidth_range(
 		struct dc_dsc_bw_range *range)
 {
 	/* native stream bandwidth */
-	range->stream_kbps = dc_bandwidth_in_kbps_from_timing(timing);
+	range->stream_kbps = dc_dsc_bandwidth_in_kbps_from_timing(timing);
 
 	/* max dsc target bpp */
 	range->max_kbps = dsc_div_by_10_round_up(max_bpp * timing->pix_clk_100hz);
@@ -512,6 +563,7 @@ static bool setup_dsc_config(
 		const struct dsc_enc_caps *dsc_enc_caps,
 		int target_bandwidth_kbps,
 		const struct dc_crtc_timing *timing,
+		int min_slice_height_override,
 		struct dc_dsc_config *dsc_cfg)
 {
 	struct dsc_enc_caps dsc_common_caps;
@@ -680,7 +732,10 @@ static bool setup_dsc_config(
 
 	// Slice height (i.e. number of slices per column): start with policy and pick the first one that height is divisible by.
 	// For 4:2:0 make sure the slice height is divisible by 2 as well.
-	slice_height = min(dsc_policy.min_sice_height, pic_height);
+	if (min_slice_height_override == 0)
+		slice_height = min(dsc_policy.min_sice_height, pic_height);
+	else
+		slice_height = min(min_slice_height_override, pic_height);
 
 	while (slice_height < pic_height && (pic_height % slice_height != 0 ||
 		(timing->pixel_encoding == PIXEL_ENCODING_YCBCR420 && slice_height % 2 != 0)))
@@ -802,7 +857,8 @@ bool dc_dsc_parse_dsc_dpcd(const uint8_t *dpcd_dsc_basic_data, const uint8_t *dp
  * If DSC is not possible, leave '*range' untouched.
  */
 bool dc_dsc_compute_bandwidth_range(
-		const struct dc *dc,
+		const struct display_stream_compressor *dsc,
+		const uint32_t dsc_min_slice_height_override,
 		const uint32_t min_bpp,
 		const uint32_t max_bpp,
 		const struct dsc_dec_dpcd_caps *dsc_sink_caps,
@@ -814,16 +870,14 @@ bool dc_dsc_compute_bandwidth_range(
 	struct dsc_enc_caps dsc_common_caps;
 	struct dc_dsc_config config;
 
-	get_dsc_enc_caps(dc, &dsc_enc_caps, timing->pix_clk_100hz);
+	get_dsc_enc_caps(dsc, &dsc_enc_caps, timing->pix_clk_100hz);
 
 	is_dsc_possible = intersect_dsc_caps(dsc_sink_caps, &dsc_enc_caps,
 			timing->pixel_encoding, &dsc_common_caps);
 
 	if (is_dsc_possible)
-		is_dsc_possible = setup_dsc_config(dsc_sink_caps,
-				&dsc_enc_caps,
-				0,
-				timing, &config);
+		is_dsc_possible = setup_dsc_config(dsc_sink_caps, &dsc_enc_caps, 0, timing,
+				dsc_min_slice_height_override, &config);
 
 	if (is_dsc_possible)
 		get_dsc_bandwidth_range(min_bpp, max_bpp, &dsc_common_caps, timing, range);
@@ -832,8 +886,9 @@ bool dc_dsc_compute_bandwidth_range(
 }
 
 bool dc_dsc_compute_config(
-		const struct dc *dc,
+		const struct display_stream_compressor *dsc,
 		const struct dsc_dec_dpcd_caps *dsc_sink_caps,
+		const uint32_t dsc_min_slice_height_override,
 		uint32_t target_bandwidth_kbps,
 		const struct dc_crtc_timing *timing,
 		struct dc_dsc_config *dsc_cfg)
@@ -841,11 +896,11 @@ bool dc_dsc_compute_config(
 	bool is_dsc_possible = false;
 	struct dsc_enc_caps dsc_enc_caps;
 
-	get_dsc_enc_caps(dc, &dsc_enc_caps, timing->pix_clk_100hz);
+	get_dsc_enc_caps(dsc, &dsc_enc_caps, timing->pix_clk_100hz);
 	is_dsc_possible = setup_dsc_config(dsc_sink_caps,
 			&dsc_enc_caps,
 			target_bandwidth_kbps,
-			timing, dsc_cfg);
+			timing, dsc_min_slice_height_override, dsc_cfg);
 	return is_dsc_possible;
 }
 #endif /* CONFIG_DRM_AMD_DC_DSC_SUPPORT */
diff --git a/drivers/gpu/drm/amd/display/dc/dsc/rc_calc.c b/drivers/gpu/drm/amd/display/dc/dsc/rc_calc.c
index ca51e83f8764..76c4b12d6824 100644
--- a/drivers/gpu/drm/amd/display/dc/dsc/rc_calc.c
+++ b/drivers/gpu/drm/amd/display/dc/dsc/rc_calc.c
@@ -177,7 +177,6 @@ void calc_rc_params(struct rc_params *rc, enum colour_mode cm, enum bits_per_com
 {
 	float bpp_group;
 	float initial_xmit_delay_factor;
-	int source_bpp;
 	int padding_pixels;
 	int i;
 
@@ -217,8 +216,6 @@ void calc_rc_params(struct rc_params *rc, enum colour_mode cm, enum bits_per_com
 			rc->initial_xmit_delay++;
 	}
 
-	source_bpp = MODE_SELECT(bpc * 3, bpc * 2, bpc * 1.5);
-
 	rc->flatness_min_qp     = ((bpc == BPC_8) ?  (3) : ((bpc == BPC_10) ? (7)  : (11))) - ((minor_version == 1 && cm == CM_444) ? 1 : 0);
 	rc->flatness_max_qp     = ((bpc == BPC_8) ? (12) : ((bpc == BPC_10) ? (16) : (20))) - ((minor_version == 1 && cm == CM_444) ? 1 : 0);
 	rc->flatness_det_thresh = 2 << (bpc - 8);
diff --git a/drivers/gpu/drm/amd/display/dc/gpio/gpio_base.c b/drivers/gpu/drm/amd/display/dc/gpio/gpio_base.c
index f8f85490e77e..f67c18375bfd 100644
--- a/drivers/gpu/drm/amd/display/dc/gpio/gpio_base.c
+++ b/drivers/gpu/drm/amd/display/dc/gpio/gpio_base.c
@@ -321,8 +321,6 @@ void dal_gpio_destroy(
 		return;
 	}
 
-	dal_gpio_close(*gpio);
-
 	switch ((*gpio)->id) {
 	case GPIO_ID_DDC_DATA:
 		kfree((*gpio)->hw_container.ddc);
diff --git a/drivers/gpu/drm/amd/display/dc/gpio/gpio_service.c b/drivers/gpu/drm/amd/display/dc/gpio/gpio_service.c
index d03165e71dc6..92280cc05e2d 100644
--- a/drivers/gpu/drm/amd/display/dc/gpio/gpio_service.c
+++ b/drivers/gpu/drm/amd/display/dc/gpio/gpio_service.c
@@ -169,7 +169,6 @@ void dal_gpio_destroy_generic_mux(
 		return;
 	}
 
-	dal_gpio_close(*mux);
 	dal_gpio_destroy(mux);
 	kfree(*mux);
 
@@ -460,7 +459,6 @@ void dal_gpio_destroy_irq(
 		return;
 	}
 
-	dal_gpio_close(*irq);
 	dal_gpio_destroy(irq);
 	kfree(*irq);
 
diff --git a/drivers/gpu/drm/amd/display/dc/hdcp/Makefile b/drivers/gpu/drm/amd/display/dc/hdcp/Makefile
new file mode 100644
index 000000000000..4170b6eb9ec0
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/hdcp/Makefile
@@ -0,0 +1,28 @@
+# Copyright 2019 Advanced Micro Devices, Inc.
+#
+# Permission is hereby granted, free of charge, to any person obtaining a
+# copy of this software and associated documentation files (the "Software"),
+# to deal in the Software without restriction, including without limitation
+# the rights to use, copy, modify, merge, publish, distribute, sublicense,
+# and/or sell copies of the Software, and to permit persons to whom the
+# Software is furnished to do so, subject to the following conditions:
+#
+# The above copyright notice and this permission notice shall be included in
+# all copies or substantial portions of the Software.
+#
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+# THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+# OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+# ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+# OTHER DEALINGS IN THE SOFTWARE.
+#
+# Makefile for the 'hdcp' sub-component of DAL.
+#
+
+HDCP_MSG = hdcp_msg.o
+
+AMD_DAL_HDCP_MSG = $(addprefix $(AMDDALPATH)/dc/hdcp/,$(HDCP_MSG))
+
+AMD_DISPLAY_FILES += $(AMD_DAL_HDCP_MSG)
diff --git a/drivers/gpu/drm/amd/display/dc/hdcp/hdcp_msg.c b/drivers/gpu/drm/amd/display/dc/hdcp/hdcp_msg.c
new file mode 100644
index 000000000000..6f730b5bfe42
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/hdcp/hdcp_msg.c
@@ -0,0 +1,324 @@
+/*
+ * Copyright 2019 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#include <linux/slab.h>
+
+#include "dm_services.h"
+#include "dm_helpers.h"
+#include "include/hdcp_types.h"
+#include "include/i2caux_interface.h"
+#include "include/signal_types.h"
+#include "core_types.h"
+#include "dc_link_ddc.h"
+#include "link_hwss.h"
+
+#define DC_LOGGER \
+	link->ctx->logger
+#define HDCP14_KSV_SIZE 5
+#define HDCP14_MAX_KSV_FIFO_SIZE 127*HDCP14_KSV_SIZE
+
+static const bool hdcp_cmd_is_read[] = {
+	[HDCP_MESSAGE_ID_READ_BKSV] = true,
+	[HDCP_MESSAGE_ID_READ_RI_R0] = true,
+	[HDCP_MESSAGE_ID_READ_PJ] = true,
+	[HDCP_MESSAGE_ID_WRITE_AKSV] = false,
+	[HDCP_MESSAGE_ID_WRITE_AINFO] = false,
+	[HDCP_MESSAGE_ID_WRITE_AN] = false,
+	[HDCP_MESSAGE_ID_READ_VH_X] = true,
+	[HDCP_MESSAGE_ID_READ_VH_0] = true,
+	[HDCP_MESSAGE_ID_READ_VH_1] = true,
+	[HDCP_MESSAGE_ID_READ_VH_2] = true,
+	[HDCP_MESSAGE_ID_READ_VH_3] = true,
+	[HDCP_MESSAGE_ID_READ_VH_4] = true,
+	[HDCP_MESSAGE_ID_READ_BCAPS] = true,
+	[HDCP_MESSAGE_ID_READ_BSTATUS] = true,
+	[HDCP_MESSAGE_ID_READ_KSV_FIFO] = true,
+	[HDCP_MESSAGE_ID_READ_BINFO] = true,
+	[HDCP_MESSAGE_ID_HDCP2VERSION] = true,
+	[HDCP_MESSAGE_ID_RX_CAPS] = true,
+	[HDCP_MESSAGE_ID_WRITE_AKE_INIT] = false,
+	[HDCP_MESSAGE_ID_READ_AKE_SEND_CERT] = true,
+	[HDCP_MESSAGE_ID_WRITE_AKE_NO_STORED_KM] = false,
+	[HDCP_MESSAGE_ID_WRITE_AKE_STORED_KM] = false,
+	[HDCP_MESSAGE_ID_READ_AKE_SEND_H_PRIME] = true,
+	[HDCP_MESSAGE_ID_READ_AKE_SEND_PAIRING_INFO] = true,
+	[HDCP_MESSAGE_ID_WRITE_LC_INIT] = false,
+	[HDCP_MESSAGE_ID_READ_LC_SEND_L_PRIME] = true,
+	[HDCP_MESSAGE_ID_WRITE_SKE_SEND_EKS] = false,
+	[HDCP_MESSAGE_ID_READ_REPEATER_AUTH_SEND_RECEIVERID_LIST] = true,
+	[HDCP_MESSAGE_ID_WRITE_REPEATER_AUTH_SEND_ACK] = false,
+	[HDCP_MESSAGE_ID_WRITE_REPEATER_AUTH_STREAM_MANAGE] = false,
+	[HDCP_MESSAGE_ID_READ_REPEATER_AUTH_STREAM_READY] = true,
+	[HDCP_MESSAGE_ID_READ_RXSTATUS] = true,
+	[HDCP_MESSAGE_ID_WRITE_CONTENT_STREAM_TYPE] = false
+};
+
+static const uint8_t hdcp_i2c_offsets[] = {
+	[HDCP_MESSAGE_ID_READ_BKSV] = 0x0,
+	[HDCP_MESSAGE_ID_READ_RI_R0] = 0x8,
+	[HDCP_MESSAGE_ID_READ_PJ] = 0xA,
+	[HDCP_MESSAGE_ID_WRITE_AKSV] = 0x10,
+	[HDCP_MESSAGE_ID_WRITE_AINFO] = 0x15,
+	[HDCP_MESSAGE_ID_WRITE_AN] = 0x18,
+	[HDCP_MESSAGE_ID_READ_VH_X] = 0x20,
+	[HDCP_MESSAGE_ID_READ_VH_0] = 0x20,
+	[HDCP_MESSAGE_ID_READ_VH_1] = 0x24,
+	[HDCP_MESSAGE_ID_READ_VH_2] = 0x28,
+	[HDCP_MESSAGE_ID_READ_VH_3] = 0x2C,
+	[HDCP_MESSAGE_ID_READ_VH_4] = 0x30,
+	[HDCP_MESSAGE_ID_READ_BCAPS] = 0x40,
+	[HDCP_MESSAGE_ID_READ_BSTATUS] = 0x41,
+	[HDCP_MESSAGE_ID_READ_KSV_FIFO] = 0x43,
+	[HDCP_MESSAGE_ID_READ_BINFO] = 0xFF,
+	[HDCP_MESSAGE_ID_HDCP2VERSION] = 0x50,
+	[HDCP_MESSAGE_ID_WRITE_AKE_INIT] = 0x60,
+	[HDCP_MESSAGE_ID_READ_AKE_SEND_CERT] = 0x80,
+	[HDCP_MESSAGE_ID_WRITE_AKE_NO_STORED_KM] = 0x60,
+	[HDCP_MESSAGE_ID_WRITE_AKE_STORED_KM] = 0x60,
+	[HDCP_MESSAGE_ID_READ_AKE_SEND_H_PRIME] = 0x80,
+	[HDCP_MESSAGE_ID_READ_AKE_SEND_PAIRING_INFO] = 0x80,
+	[HDCP_MESSAGE_ID_WRITE_LC_INIT] = 0x60,
+	[HDCP_MESSAGE_ID_READ_LC_SEND_L_PRIME] = 0x80,
+	[HDCP_MESSAGE_ID_WRITE_SKE_SEND_EKS] = 0x60,
+	[HDCP_MESSAGE_ID_READ_REPEATER_AUTH_SEND_RECEIVERID_LIST] = 0x80,
+	[HDCP_MESSAGE_ID_WRITE_REPEATER_AUTH_SEND_ACK] = 0x60,
+	[HDCP_MESSAGE_ID_WRITE_REPEATER_AUTH_STREAM_MANAGE] = 0x60,
+	[HDCP_MESSAGE_ID_READ_REPEATER_AUTH_STREAM_READY] = 0x80,
+	[HDCP_MESSAGE_ID_READ_RXSTATUS] = 0x70
+};
+
+struct protection_properties {
+	bool supported;
+	bool (*process_transaction)(
+		struct dc_link *link,
+		struct hdcp_protection_message *message_info);
+};
+
+static const struct protection_properties non_supported_protection = {
+	.supported = false
+};
+
+static bool hdmi_14_process_transaction(
+	struct dc_link *link,
+	struct hdcp_protection_message *message_info)
+{
+	uint8_t *buff = NULL;
+	bool result;
+	const uint8_t hdcp_i2c_addr_link_primary = 0x3a; /* 0x74 >> 1*/
+	const uint8_t hdcp_i2c_addr_link_secondary = 0x3b; /* 0x76 >> 1*/
+	struct i2c_command i2c_command;
+	uint8_t offset = hdcp_i2c_offsets[message_info->msg_id];
+	struct i2c_payload i2c_payloads[] = {
+		{ true, 0, 1, &offset },
+		/* actual hdcp payload, will be filled later, zeroed for now*/
+		{ 0 }
+	};
+
+	switch (message_info->link) {
+	case HDCP_LINK_SECONDARY:
+		i2c_payloads[0].address = hdcp_i2c_addr_link_secondary;
+		i2c_payloads[1].address = hdcp_i2c_addr_link_secondary;
+		break;
+	case HDCP_LINK_PRIMARY:
+	default:
+		i2c_payloads[0].address = hdcp_i2c_addr_link_primary;
+		i2c_payloads[1].address = hdcp_i2c_addr_link_primary;
+		break;
+	}
+
+	if (hdcp_cmd_is_read[message_info->msg_id]) {
+		i2c_payloads[1].write = false;
+		i2c_command.number_of_payloads = ARRAY_SIZE(i2c_payloads);
+		i2c_payloads[1].length = message_info->length;
+		i2c_payloads[1].data = message_info->data;
+	} else {
+		i2c_command.number_of_payloads = 1;
+		buff = kzalloc(message_info->length + 1, GFP_KERNEL);
+
+		if (!buff)
+			return false;
+
+		buff[0] = offset;
+		memmove(&buff[1], message_info->data, message_info->length);
+		i2c_payloads[0].length = message_info->length + 1;
+		i2c_payloads[0].data = buff;
+	}
+
+	i2c_command.payloads = i2c_payloads;
+	i2c_command.engine = I2C_COMMAND_ENGINE_HW;//only HW
+	i2c_command.speed = link->ddc->ctx->dc->caps.i2c_speed_in_khz;
+
+	result = dm_helpers_submit_i2c(
+			link->ctx,
+			link,
+			&i2c_command);
+	kfree(buff);
+
+	return result;
+}
+
+static const struct protection_properties hdmi_14_protection = {
+	.supported = true,
+	.process_transaction = hdmi_14_process_transaction
+};
+
+static const uint32_t hdcp_dpcd_addrs[] = {
+	[HDCP_MESSAGE_ID_READ_BKSV] = 0x68000,
+	[HDCP_MESSAGE_ID_READ_RI_R0] = 0x68005,
+	[HDCP_MESSAGE_ID_READ_PJ] = 0xFFFFFFFF,
+	[HDCP_MESSAGE_ID_WRITE_AKSV] = 0x68007,
+	[HDCP_MESSAGE_ID_WRITE_AINFO] = 0x6803B,
+	[HDCP_MESSAGE_ID_WRITE_AN] = 0x6800c,
+	[HDCP_MESSAGE_ID_READ_VH_X] = 0x68014,
+	[HDCP_MESSAGE_ID_READ_VH_0] = 0x68014,
+	[HDCP_MESSAGE_ID_READ_VH_1] = 0x68018,
+	[HDCP_MESSAGE_ID_READ_VH_2] = 0x6801c,
+	[HDCP_MESSAGE_ID_READ_VH_3] = 0x68020,
+	[HDCP_MESSAGE_ID_READ_VH_4] = 0x68024,
+	[HDCP_MESSAGE_ID_READ_BCAPS] = 0x68028,
+	[HDCP_MESSAGE_ID_READ_BSTATUS] = 0x68029,
+	[HDCP_MESSAGE_ID_READ_KSV_FIFO] = 0x6802c,
+	[HDCP_MESSAGE_ID_READ_BINFO] = 0x6802a,
+	[HDCP_MESSAGE_ID_RX_CAPS] = 0x6921d,
+	[HDCP_MESSAGE_ID_WRITE_AKE_INIT] = 0x69000,
+	[HDCP_MESSAGE_ID_READ_AKE_SEND_CERT] = 0x6900b,
+	[HDCP_MESSAGE_ID_WRITE_AKE_NO_STORED_KM] = 0x69220,
+	[HDCP_MESSAGE_ID_WRITE_AKE_STORED_KM] = 0x692a0,
+	[HDCP_MESSAGE_ID_READ_AKE_SEND_H_PRIME] = 0x692c0,
+	[HDCP_MESSAGE_ID_READ_AKE_SEND_PAIRING_INFO] = 0x692e0,
+	[HDCP_MESSAGE_ID_WRITE_LC_INIT] = 0x692f0,
+	[HDCP_MESSAGE_ID_READ_LC_SEND_L_PRIME] = 0x692f8,
+	[HDCP_MESSAGE_ID_WRITE_SKE_SEND_EKS] = 0x69318,
+	[HDCP_MESSAGE_ID_READ_REPEATER_AUTH_SEND_RECEIVERID_LIST] = 0x69330,
+	[HDCP_MESSAGE_ID_WRITE_REPEATER_AUTH_SEND_ACK] = 0x693e0,
+	[HDCP_MESSAGE_ID_WRITE_REPEATER_AUTH_STREAM_MANAGE] = 0x693f0,
+	[HDCP_MESSAGE_ID_READ_REPEATER_AUTH_STREAM_READY] = 0x69473,
+	[HDCP_MESSAGE_ID_READ_RXSTATUS] = 0x69493,
+	[HDCP_MESSAGE_ID_WRITE_CONTENT_STREAM_TYPE] = 0x69494
+};
+
+static bool dpcd_access_helper(
+	struct dc_link *link,
+	uint32_t length,
+	uint8_t *data,
+	uint32_t dpcd_addr,
+	bool is_read)
+{
+	enum dc_status status;
+	uint32_t cur_length = 0;
+	uint32_t offset = 0;
+	uint32_t ksv_read_size = 0x6803b - 0x6802c;
+
+	/* Read KSV, need repeatedly handle */
+	if (dpcd_addr == 0x6802c) {
+		if (length % HDCP14_KSV_SIZE) {
+			DC_LOG_ERROR("%s: KsvFifo Size(%d) is not a multiple of HDCP14_KSV_SIZE(%d)\n",
+				__func__,
+				length,
+				HDCP14_KSV_SIZE);
+		}
+		if (length > HDCP14_MAX_KSV_FIFO_SIZE) {
+			DC_LOG_ERROR("%s: KsvFifo Size(%d) is greater than HDCP14_MAX_KSV_FIFO_SIZE(%d)\n",
+				__func__,
+				length,
+				HDCP14_MAX_KSV_FIFO_SIZE);
+		}
+
+		DC_LOG_ERROR("%s: Reading %d Ksv(s) from KsvFifo\n",
+			__func__,
+			length / HDCP14_KSV_SIZE);
+
+		while (length > 0) {
+			if (length > ksv_read_size) {
+				status = core_link_read_dpcd(
+					link,
+					dpcd_addr + offset,
+					data + offset,
+					ksv_read_size);
+
+				data += ksv_read_size;
+				length -= ksv_read_size;
+			} else {
+				status = core_link_read_dpcd(
+					link,
+					dpcd_addr + offset,
+					data + offset,
+					length);
+
+				data += length;
+				length = 0;
+			}
+
+			if (status != DC_OK)
+				return false;
+		}
+	} else {
+		while (length > 0) {
+			if (length > DEFAULT_AUX_MAX_DATA_SIZE)
+				cur_length = DEFAULT_AUX_MAX_DATA_SIZE;
+			else
+				cur_length = length;
+
+			if (is_read) {
+				status = core_link_read_dpcd(
+					link,
+					dpcd_addr + offset,
+					data + offset,
+					cur_length);
+			} else {
+				status = core_link_write_dpcd(
+					link,
+					dpcd_addr + offset,
+					data + offset,
+					cur_length);
+			}
+
+			if (status != DC_OK)
+				return false;
+
+			length -= cur_length;
+			offset += cur_length;
+		}
+	}
+	return true;
+}
+
+static bool dp_11_process_transaction(
+	struct dc_link *link,
+	struct hdcp_protection_message *message_info)
+{
+	return dpcd_access_helper(
+		link,
+		message_info->length,
+		message_info->data,
+		hdcp_dpcd_addrs[message_info->msg_id],
+		hdcp_cmd_is_read[message_info->msg_id]);
+}
+
+static const struct protection_properties dp_11_protection = {
+	.supported = true,
+	.process_transaction = dp_11_process_transaction
+};
+
diff --git a/drivers/gpu/drm/amd/display/dc/inc/core_types.h b/drivers/gpu/drm/amd/display/dc/inc/core_types.h
index f189307750ab..a831079607cd 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/core_types.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/core_types.h
@@ -52,7 +52,9 @@ void enable_surface_flip_reporting(struct dc_plane_state *plane_state,
 #include "clock_source.h"
 #include "audio.h"
 #include "dm_pp_smu.h"
-
+#ifdef CONFIG_DRM_AMD_DC_HDCP
+#include "dm_cp_psp.h"
+#endif
 
 /************ link *****************/
 struct link_init_data {
@@ -231,6 +233,7 @@ struct resource_pool {
 
 struct dcn_fe_bandwidth {
 	int dppclk_khz;
+
 };
 
 struct stream_resource {
@@ -395,10 +398,6 @@ struct dc_state {
 
 	struct clk_mgr *clk_mgr;
 
-	struct {
-		bool full_update_needed : 1;
-	} commit_hints;
-
 	struct kref refcount;
 };
 
diff --git a/drivers/gpu/drm/amd/display/dc/inc/dc_link_ddc.h b/drivers/gpu/drm/amd/display/dc/inc/dc_link_ddc.h
index b1fab251c09b..14716ba35662 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/dc_link_ddc.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/dc_link_ddc.h
@@ -95,6 +95,9 @@ bool dal_ddc_service_query_ddc_data(
 		uint8_t *read_buf,
 		uint32_t read_size);
 
+bool dal_ddc_submit_aux_command(struct ddc_service *ddc,
+		struct aux_payload *payload);
+
 int dc_link_aux_transfer_raw(struct ddc_service *ddc,
 		struct aux_payload *payload,
 		enum aux_channel_operation_result *operation_result);
@@ -102,6 +105,9 @@ int dc_link_aux_transfer_raw(struct ddc_service *ddc,
 bool dc_link_aux_transfer_with_retries(struct ddc_service *ddc,
 		struct aux_payload *payload);
 
+enum dc_status dc_link_aux_configure_timeout(struct ddc_service *ddc,
+		uint32_t timeout);
+
 void dal_ddc_service_write_scdc_data(
 		struct ddc_service *ddc_service,
 		uint32_t pix_clk,
diff --git a/drivers/gpu/drm/amd/display/dc/inc/dc_link_dp.h b/drivers/gpu/drm/amd/display/dc/inc/dc_link_dp.h
index 08a4df2c61a8..045138dbdccb 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/dc_link_dp.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/dc_link_dp.h
@@ -28,6 +28,8 @@
 
 #define LINK_TRAINING_ATTEMPTS 4
 #define LINK_TRAINING_RETRY_DELAY 50 /* ms */
+#define LINK_AUX_DEFAULT_EXTENDED_TIMEOUT_PERIOD 32000 /*us*/
+#define LINK_AUX_DEFAULT_TIMEOUT_PERIOD 400 /*us*/
 
 struct dc_link;
 struct dc_stream_state;
@@ -43,6 +45,9 @@ bool dp_verify_link_cap_with_retries(
 	struct dc_link_settings *known_limit_link_setting,
 	int attempts);
 
+bool dp_verify_mst_link_cap(
+	struct dc_link *link);
+
 bool dp_validate_mode_timing(
 	struct dc_link *link,
 	const struct dc_crtc_timing *timing);
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/aux_engine.h b/drivers/gpu/drm/amd/display/dc/inc/hw/aux_engine.h
index e79cd4e92919..e77b3a76766d 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/aux_engine.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/aux_engine.h
@@ -140,6 +140,9 @@ struct write_command_context {
 
 
 struct aux_engine_funcs {
+	bool (*configure_timeout)(
+		struct ddc_service *ddc,
+		uint32_t timeout);
 	void (*destroy)(
 		struct aux_engine **ptr);
 	bool (*acquire_engine)(
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr.h b/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr.h
index 76f9ad1b23df..4e18e77dcf42 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr.h
@@ -47,7 +47,7 @@
 #ifdef CONFIG_DRM_AMD_DC_DCN2_1
 /* Will these bw structures be ASIC specific? */
 
-#define MAX_NUM_DPM_LVL		4
+#define MAX_NUM_DPM_LVL		8
 #define WM_SET_COUNT 		4
 
 
@@ -180,13 +180,19 @@ struct clk_mgr_funcs {
 			struct dc_state *context,
 			enum dc_clock_type clock_type,
 			struct dc_clock_config *clock_cfg);
+
+	bool (*are_clock_states_equal) (struct dc_clocks *a,
+			struct dc_clocks *b);
+	void (*notify_wm_ranges)(struct clk_mgr *clk_mgr);
 };
 
 struct clk_mgr {
 	struct dc_context *ctx;
 	struct clk_mgr_funcs *funcs;
 	struct dc_clocks clks;
+	bool psr_allow_active_cache;
 	int dprefclk_khz; // Used by program pixel clock in clock source funcs, need to figureout where this goes
+	int dentist_vco_freq_khz;
 #ifdef CONFIG_DRM_AMD_DC_DCN2_1
 	struct clk_bw_params *bw_params;
 #endif
@@ -199,4 +205,8 @@ struct clk_mgr *dc_clk_mgr_create(struct dc_context *ctx, struct pp_smu_funcs *p
 
 void dc_destroy_clk_mgr(struct clk_mgr *clk_mgr);
 
+void clk_mgr_exit_optimized_pwr_state(const struct dc *dc, struct clk_mgr *clk_mgr);
+
+void clk_mgr_optimize_pwr_state(const struct dc *dc, struct clk_mgr *clk_mgr);
+
 #endif /* __DAL_CLK_MGR_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr_internal.h b/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr_internal.h
index 7dd46eb96d67..a17a77192690 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr_internal.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr_internal.h
@@ -184,6 +184,21 @@ struct clk_mgr_registers {
 	uint32_t MP1_SMN_C2PMSG_91;
 };
 
+enum clock_type {
+	clock_type_dispclk = 1,
+	clock_type_dcfclk,
+	clock_type_socclk,
+	clock_type_pixelclk,
+	clock_type_phyclk,
+	clock_type_dppclk,
+	clock_type_fclk,
+	clock_type_dcfdsclk,
+	clock_type_dscclk,
+	clock_type_uclk,
+	clock_type_dramclk,
+};
+
+
 struct state_dependent_clocks {
 	int display_clk_khz;
 	int pixel_clk_khz;
@@ -210,8 +225,6 @@ struct clk_mgr_internal {
 	struct state_dependent_clocks max_clks_by_state[DM_PP_CLOCKS_MAX_STATES];
 
 	/*TODO: figure out which of the below fields should be here vs in asic specific portion */
-	int dentist_vco_freq_khz;
-
 	/* Cache the status of DFS-bypass feature*/
 	bool dfs_bypass_enabled;
 	/* True if the DFS-bypass feature is enabled and active. */
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/dccg.h b/drivers/gpu/drm/amd/display/dc/inc/hw/dccg.h
index d8e744f366e5..05ee5295d2c1 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/dccg.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/dccg.h
@@ -38,8 +38,7 @@ struct dccg {
 struct dccg_funcs {
 	void (*update_dpp_dto)(struct dccg *dccg,
 			int dpp_inst,
-			int req_dppclk,
-			bool reduce_divider_only);
+			int req_dppclk);
 	void (*get_dccg_ref_freq)(struct dccg *dccg,
 			unsigned int xtalin_freq_inKhz,
 			unsigned int *dccg_ref_freq_inKhz);
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/dchubbub.h b/drivers/gpu/drm/amd/display/dc/inc/hw/dchubbub.h
index a6297219d7fc..c81a17aeaa25 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/dchubbub.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/dchubbub.h
@@ -147,6 +147,7 @@ struct hubbub_funcs {
 	bool (*is_allow_self_refresh_enabled)(struct hubbub *hubbub);
 	void (*allow_self_refresh_control)(struct hubbub *hubbub, bool allow);
 
+	void (*apply_DEDCN21_147_wa)(struct hubbub *hubbub);
 };
 
 struct hubbub {
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/dsc.h b/drivers/gpu/drm/amd/display/dc/inc/hw/dsc.h
index 1ddb1c6fa149..c6ff3d78b435 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/dsc.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/dsc.h
@@ -28,7 +28,11 @@
 
 #include "dc_dsc.h"
 #include "dc_hw_types.h"
-#include "dc_dp_types.h"
+#include "dc_types.h"
+/* do not include any other headers
+ * or else it might break Edid Utility functionality.
+ */
+
 
 /* Input parameters for configuring DSC from the outside of DSC */
 struct dsc_config {
@@ -81,12 +85,6 @@ struct dsc_enc_caps {
 	uint32_t bpp_increment_div; /* bpp increment divisor, e.g. if 16, it's 1/16th of a bit */
 };
 
-struct display_stream_compressor {
-	const struct dsc_funcs *funcs;
-	struct dc_context *ctx;
-	int inst;
-};
-
 struct dsc_funcs {
 	void (*dsc_get_enc_caps)(struct dsc_enc_caps *dsc_enc_caps, int pixel_clock_100Hz);
 	void (*dsc_read_state)(struct display_stream_compressor *dsc, struct dcn_dsc_state *s);
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/link_encoder.h b/drivers/gpu/drm/amd/display/dc/inc/hw/link_encoder.h
index abb4e4237fb6..b21909216fb6 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/link_encoder.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/link_encoder.h
@@ -184,6 +184,10 @@ struct link_encoder_funcs {
 	bool (*fec_is_active)(struct link_encoder *enc);
 #endif
 	bool (*is_in_alt_mode) (struct link_encoder *enc);
+
+	void (*get_max_link_cap)(struct link_encoder *enc,
+		struct dc_link_settings *link_settings);
+
 	enum signal_type (*get_dig_mode)(
 		struct link_encoder *enc);
 };
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/mem_input.h b/drivers/gpu/drm/amd/display/dc/inc/hw/mem_input.h
index e8668388581b..67b610d6d91f 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/mem_input.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/mem_input.h
@@ -43,6 +43,7 @@ struct dcn_watermarks {
 #if defined(CONFIG_DRM_AMD_DC_DCN2_1)
 	uint32_t frac_urg_bw_nom;
 	uint32_t frac_urg_bw_flip;
+	int32_t urgent_latency_ns;
 #endif
 	struct cstate_pstate_watermarks_st cstate_pstate;
 };
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/opp.h b/drivers/gpu/drm/amd/display/dc/inc/hw/opp.h
index 957e9047381a..18def2b6fafe 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/opp.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/opp.h
@@ -208,6 +208,7 @@ struct output_pixel_processor {
 	struct mpc_tree mpc_tree_params;
 	bool mpcc_disconnect_pending[MAX_PIPES];
 	const struct opp_funcs *funcs;
+	uint32_t dyn_expansion;
 };
 
 enum fmt_stereo_action {
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/stream_encoder.h b/drivers/gpu/drm/amd/display/dc/inc/hw/stream_encoder.h
index fe9b7a10a1c3..6305e388612a 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/stream_encoder.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/stream_encoder.h
@@ -214,6 +214,11 @@ struct stream_encoder_funcs {
 	unsigned int (*dig_source_otg)(
 		struct stream_encoder *enc);
 
+	bool (*dp_get_pixel_format)(
+		struct stream_encoder *enc,
+		enum dc_pixel_encoding *encoding,
+		enum dc_color_depth *depth);
+
 #if defined(CONFIG_DRM_AMD_DC_DCN2_0)
 #ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT
 	void (*enc_read_state)(struct stream_encoder *enc, struct enc_state *s);
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h b/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h
index 6196cc32356e..27c73caf74ee 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h
@@ -261,6 +261,8 @@ struct timing_generator_funcs {
 
 	void (*program_manual_trigger)(struct timing_generator *optc);
 	void (*setup_manual_trigger)(struct timing_generator *optc);
+	bool (*get_hw_timing)(struct timing_generator *optc,
+			struct dc_crtc_timing *hw_crtc_timing);
 
 	void (*set_vtg_params)(struct timing_generator *optc,
 			const struct dc_crtc_timing *dc_crtc_timing);
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h b/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h
index 3a938cd414ea..d39c1e11def5 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h
@@ -114,6 +114,9 @@ struct hw_sequencer_funcs {
 			int opp_id);
 
 #if defined(CONFIG_DRM_AMD_DC_DCN2_0)
+	void (*program_front_end_for_ctx)(
+			struct dc *dc,
+			struct dc_state *context);
 	void (*program_triplebuffer)(
 		const struct dc *dc,
 		struct pipe_ctx *pipe_ctx,
@@ -229,6 +232,13 @@ struct hw_sequencer_funcs {
 			struct dc *dc,
 			struct dc_state *context);
 
+	void (*exit_optimized_pwr_state)(
+			const struct dc *dc,
+			struct dc_state *context);
+	void (*optimize_pwr_state)(
+			const struct dc *dc,
+			struct dc_state *context);
+
 #if defined(CONFIG_DRM_AMD_DC_DCN2_0)
 	bool (*update_bandwidth)(
 			struct dc *dc,
@@ -321,10 +331,12 @@ struct hw_sequencer_funcs {
 			struct dc_state *context);
 	void (*update_writeback)(struct dc *dc,
 			const struct dc_stream_status *stream_status,
-			struct dc_writeback_info *wb_info);
+			struct dc_writeback_info *wb_info,
+			struct dc_state *context);
 	void (*enable_writeback)(struct dc *dc,
 			const struct dc_stream_status *stream_status,
-			struct dc_writeback_info *wb_info);
+			struct dc_writeback_info *wb_info,
+			struct dc_state *context);
 	void (*disable_writeback)(struct dc *dc,
 			unsigned int dwb_pipe_inst);
 #endif
@@ -337,6 +349,9 @@ struct hw_sequencer_funcs {
 			enum dc_clock_type clock_type,
 			struct dc_clock_config *clock_cfg);
 
+#if defined(CONFIG_DRM_AMD_DC_DCN2_1)
+	bool (*s0i3_golden_init_wa)(struct dc *dc);
+#endif
 };
 
 void color_space_to_black_color(
diff --git a/drivers/gpu/drm/amd/display/include/ddc_service_types.h b/drivers/gpu/drm/amd/display/include/ddc_service_types.h
index 18961707db23..9ad49da50a17 100644
--- a/drivers/gpu/drm/amd/display/include/ddc_service_types.h
+++ b/drivers/gpu/drm/amd/display/include/ddc_service_types.h
@@ -31,6 +31,8 @@
 #define DP_BRANCH_DEVICE_ID_0022B9 0x0022B9
 #define DP_BRANCH_DEVICE_ID_00001A 0x00001A
 #define DP_BRANCH_DEVICE_ID_0080E1 0x0080e1
+#define DP_BRANCH_DEVICE_ID_90CC24 0x90CC24
+#define DP_BRANCH_DEVICE_ID_00E04C 0x00E04C
 
 enum ddc_result {
 	DDC_RESULT_UNKNOWN = 0,
diff --git a/drivers/gpu/drm/amd/display/include/hdcp_types.h b/drivers/gpu/drm/amd/display/include/hdcp_types.h
new file mode 100644
index 000000000000..f31e6befc8d6
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/include/hdcp_types.h
@@ -0,0 +1,96 @@
+/*
+* Copyright 2019 Advanced Micro Devices, Inc.
+*
+* Permission is hereby granted, free of charge, to any person obtaining a
+* copy of this software and associated documentation files (the "Software"),
+* to deal in the Software without restriction, including without limitation
+* the rights to use, copy, modify, merge, publish, distribute, sublicense,
+* and/or sell copies of the Software, and to permit persons to whom the
+* Software is furnished to do so, subject to the following conditions:
+*
+* The above copyright notice and this permission notice shall be included in
+* all copies or substantial portions of the Software.
+*
+* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+* THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+* OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+* ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+* OTHER DEALINGS IN THE SOFTWARE.
+*
+* Authors: AMD
+*
+*/
+
+#ifndef __DC_HDCP_TYPES_H__
+#define __DC_HDCP_TYPES_H__
+
+enum hdcp_message_id {
+	HDCP_MESSAGE_ID_INVALID = -1,
+
+	/* HDCP 1.4 */
+
+	HDCP_MESSAGE_ID_READ_BKSV = 0,
+	/* HDMI is called Ri', DP is called R0' */
+	HDCP_MESSAGE_ID_READ_RI_R0,
+	HDCP_MESSAGE_ID_READ_PJ,
+	HDCP_MESSAGE_ID_WRITE_AKSV,
+	HDCP_MESSAGE_ID_WRITE_AINFO,
+	HDCP_MESSAGE_ID_WRITE_AN,
+	HDCP_MESSAGE_ID_READ_VH_X,
+	HDCP_MESSAGE_ID_READ_VH_0,
+	HDCP_MESSAGE_ID_READ_VH_1,
+	HDCP_MESSAGE_ID_READ_VH_2,
+	HDCP_MESSAGE_ID_READ_VH_3,
+	HDCP_MESSAGE_ID_READ_VH_4,
+	HDCP_MESSAGE_ID_READ_BCAPS,
+	HDCP_MESSAGE_ID_READ_BSTATUS,
+	HDCP_MESSAGE_ID_READ_KSV_FIFO,
+	HDCP_MESSAGE_ID_READ_BINFO,
+
+	/* HDCP 2.2 */
+
+	HDCP_MESSAGE_ID_HDCP2VERSION,
+	HDCP_MESSAGE_ID_RX_CAPS,
+	HDCP_MESSAGE_ID_WRITE_AKE_INIT,
+	HDCP_MESSAGE_ID_READ_AKE_SEND_CERT,
+	HDCP_MESSAGE_ID_WRITE_AKE_NO_STORED_KM,
+	HDCP_MESSAGE_ID_WRITE_AKE_STORED_KM,
+	HDCP_MESSAGE_ID_READ_AKE_SEND_H_PRIME,
+	HDCP_MESSAGE_ID_READ_AKE_SEND_PAIRING_INFO,
+	HDCP_MESSAGE_ID_WRITE_LC_INIT,
+	HDCP_MESSAGE_ID_READ_LC_SEND_L_PRIME,
+	HDCP_MESSAGE_ID_WRITE_SKE_SEND_EKS,
+	HDCP_MESSAGE_ID_READ_REPEATER_AUTH_SEND_RECEIVERID_LIST,
+	HDCP_MESSAGE_ID_WRITE_REPEATER_AUTH_SEND_ACK,
+	HDCP_MESSAGE_ID_WRITE_REPEATER_AUTH_STREAM_MANAGE,
+	HDCP_MESSAGE_ID_READ_REPEATER_AUTH_STREAM_READY,
+	HDCP_MESSAGE_ID_READ_RXSTATUS,
+	HDCP_MESSAGE_ID_WRITE_CONTENT_STREAM_TYPE,
+
+	HDCP_MESSAGE_ID_MAX
+};
+
+enum hdcp_version {
+	HDCP_Unknown = 0,
+	HDCP_VERSION_14,
+	HDCP_VERSION_22,
+};
+
+enum hdcp_link {
+	HDCP_LINK_PRIMARY,
+	HDCP_LINK_SECONDARY
+};
+
+struct hdcp_protection_message {
+	enum hdcp_version version;
+	/* relevant only for DVI */
+	enum hdcp_link link;
+	enum hdcp_message_id msg_id;
+	uint32_t length;
+	uint8_t max_retries;
+	uint8_t *data;
+};
+
+#endif
diff --git a/drivers/gpu/drm/amd/display/modules/color/color_gamma.c b/drivers/gpu/drm/amd/display/modules/color/color_gamma.c
index 2d8f14b69117..1de4805cb8c7 100644
--- a/drivers/gpu/drm/amd/display/modules/color/color_gamma.c
+++ b/drivers/gpu/drm/amd/display/modules/color/color_gamma.c
@@ -373,7 +373,42 @@ static struct fixed31_32 translate_from_linear_space(
 		return dc_fixpt_mul(args->arg, args->a1);
 }
 
-static struct fixed31_32 calculate_gamma22(struct fixed31_32 arg)
+
+static struct fixed31_32 translate_from_linear_space_long(
+		struct translate_from_linear_space_args *args)
+{
+	const struct fixed31_32 one = dc_fixpt_from_int(1);
+
+	if (dc_fixpt_lt(one, args->arg))
+		return one;
+
+	if (dc_fixpt_le(args->arg, dc_fixpt_neg(args->a0)))
+		return dc_fixpt_sub(
+			args->a2,
+			dc_fixpt_mul(
+				dc_fixpt_add(
+					one,
+					args->a3),
+				dc_fixpt_pow(
+					dc_fixpt_neg(args->arg),
+					dc_fixpt_recip(args->gamma))));
+	else if (dc_fixpt_le(args->a0, args->arg))
+		return dc_fixpt_sub(
+			dc_fixpt_mul(
+				dc_fixpt_add(
+					one,
+					args->a3),
+				dc_fixpt_pow(
+						args->arg,
+					dc_fixpt_recip(args->gamma))),
+					args->a2);
+	else
+		return dc_fixpt_mul(
+			args->arg,
+			args->a1);
+}
+
+static struct fixed31_32 calculate_gamma22(struct fixed31_32 arg, bool use_eetf)
 {
 	struct fixed31_32 gamma = dc_fixpt_from_fraction(22, 10);
 
@@ -384,9 +419,13 @@ static struct fixed31_32 calculate_gamma22(struct fixed31_32 arg)
 	scratch_gamma_args.a3 = dc_fixpt_zero;
 	scratch_gamma_args.gamma = gamma;
 
+	if (use_eetf)
+		return translate_from_linear_space_long(&scratch_gamma_args);
+
 	return translate_from_linear_space(&scratch_gamma_args);
 }
 
+
 static struct fixed31_32 translate_to_linear_space(
 	struct fixed31_32 arg,
 	struct fixed31_32 a0,
@@ -920,11 +959,7 @@ static bool build_freesync_hdr(struct pwl_float_data_ex *rgb_regamma,
 	if (fs_params->max_display < 100) // cap at 100 at the top
 		max_display = dc_fixpt_from_int(100);
 
-	if (fs_params->min_content < fs_params->min_display)
-		use_eetf = true;
-	else
-		min_content = min_display;
-
+	// only max used, we don't adjust min luminance
 	if (fs_params->max_content > fs_params->max_display)
 		use_eetf = true;
 	else
@@ -950,7 +985,7 @@ static bool build_freesync_hdr(struct pwl_float_data_ex *rgb_regamma,
 				if (dc_fixpt_lt(scaledX, dc_fixpt_zero))
 					output = dc_fixpt_zero;
 				else
-					output = calculate_gamma22(scaledX);
+					output = calculate_gamma22(scaledX, use_eetf);
 
 				rgb->r = output;
 				rgb->g = output;
@@ -2173,5 +2208,3 @@ bool  mod_color_calculate_degamma_curve(enum dc_transfer_func_predefined trans,
 rgb_degamma_alloc_fail:
 	return ret;
 }
-
-
diff --git a/drivers/gpu/drm/amd/display/modules/freesync/freesync.c b/drivers/gpu/drm/amd/display/modules/freesync/freesync.c
index ec70c9b12e1a..16e69bbc69aa 100644
--- a/drivers/gpu/drm/amd/display/modules/freesync/freesync.c
+++ b/drivers/gpu/drm/amd/display/modules/freesync/freesync.c
@@ -37,8 +37,8 @@
 #define STATIC_SCREEN_RAMP_DELTA_REFRESH_RATE_PER_FRAME ((1000 / 60) * 65)
 /* Number of elements in the render times cache array */
 #define RENDER_TIMES_MAX_COUNT 10
-/* Threshold to exit BTR (to avoid frequent enter-exits at the lower limit) */
-#define BTR_EXIT_MARGIN 2000
+/* Threshold to exit/exit BTR (to avoid frequent enter-exits at the lower limit) */
+#define BTR_MAX_MARGIN 2500
 /* Threshold to change BTR multiplier (to avoid frequent changes) */
 #define BTR_DRIFT_MARGIN 2000
 /*Threshold to exit fixed refresh rate*/
@@ -234,6 +234,10 @@ static void update_v_total_for_static_ramp(
 			current_duration_in_us) * (stream->timing.pix_clk_100hz / 10)),
 				stream->timing.h_total), 1000);
 
+	/* v_total cannot be less than nominal */
+	if (v_total < stream->timing.v_total)
+		v_total = stream->timing.v_total;
+
 	in_out_vrr->adjust.v_total_min = v_total;
 	in_out_vrr->adjust.v_total_max = v_total;
 }
@@ -250,24 +254,22 @@ static void apply_below_the_range(struct core_freesync *core_freesync,
 	unsigned int delta_from_mid_point_in_us_1 = 0xFFFFFFFF;
 	unsigned int delta_from_mid_point_in_us_2 = 0xFFFFFFFF;
 	unsigned int frames_to_insert = 0;
-	unsigned int min_frame_duration_in_ns = 0;
-	unsigned int max_render_time_in_us = in_out_vrr->max_duration_in_us;
 	unsigned int delta_from_mid_point_delta_in_us;
-
-	min_frame_duration_in_ns = ((unsigned int) (div64_u64(
-		(1000000000ULL * 1000000),
-		in_out_vrr->max_refresh_in_uhz)));
+	unsigned int max_render_time_in_us =
+			in_out_vrr->max_duration_in_us - in_out_vrr->btr.margin_in_us;
 
 	/* Program BTR */
-	if (last_render_time_in_us + BTR_EXIT_MARGIN < max_render_time_in_us) {
+	if ((last_render_time_in_us + in_out_vrr->btr.margin_in_us / 2) < max_render_time_in_us) {
 		/* Exit Below the Range */
 		if (in_out_vrr->btr.btr_active) {
 			in_out_vrr->btr.frame_counter = 0;
 			in_out_vrr->btr.btr_active = false;
 		}
-	} else if (last_render_time_in_us > max_render_time_in_us) {
+	} else if (last_render_time_in_us > (max_render_time_in_us + in_out_vrr->btr.margin_in_us / 2)) {
 		/* Enter Below the Range */
-		in_out_vrr->btr.btr_active = true;
+		if (!in_out_vrr->btr.btr_active) {
+			in_out_vrr->btr.btr_active = true;
+		}
 	}
 
 	/* BTR set to "not active" so disengage */
@@ -323,7 +325,9 @@ static void apply_below_the_range(struct core_freesync *core_freesync,
 		/* Choose number of frames to insert based on how close it
 		 * can get to the mid point of the variable range.
 		 */
-		if (delta_from_mid_point_in_us_1 < delta_from_mid_point_in_us_2) {
+		if ((frame_time_in_us / mid_point_frames_ceil) > in_out_vrr->min_duration_in_us &&
+				(delta_from_mid_point_in_us_1 < delta_from_mid_point_in_us_2 ||
+						mid_point_frames_floor < 2)) {
 			frames_to_insert = mid_point_frames_ceil;
 			delta_from_mid_point_delta_in_us = delta_from_mid_point_in_us_2 -
 					delta_from_mid_point_in_us_1;
@@ -339,7 +343,7 @@ static void apply_below_the_range(struct core_freesync *core_freesync,
 		if (in_out_vrr->btr.frames_to_insert != 0 &&
 				delta_from_mid_point_delta_in_us < BTR_DRIFT_MARGIN) {
 			if (((last_render_time_in_us / in_out_vrr->btr.frames_to_insert) <
-					in_out_vrr->max_duration_in_us) &&
+					max_render_time_in_us) &&
 				((last_render_time_in_us / in_out_vrr->btr.frames_to_insert) >
 					in_out_vrr->min_duration_in_us))
 				frames_to_insert = in_out_vrr->btr.frames_to_insert;
@@ -743,6 +747,10 @@ void mod_freesync_build_vrr_params(struct mod_freesync *mod_freesync,
 	nominal_field_rate_in_uhz =
 			mod_freesync_calc_nominal_field_rate(stream);
 
+	/* Rounded to the nearest Hz */
+	nominal_field_rate_in_uhz = 1000000ULL *
+			div_u64(nominal_field_rate_in_uhz + 500000, 1000000);
+
 	min_refresh_in_uhz = in_config->min_refresh_in_uhz;
 	max_refresh_in_uhz = in_config->max_refresh_in_uhz;
 
@@ -788,6 +796,11 @@ void mod_freesync_build_vrr_params(struct mod_freesync *mod_freesync,
 		refresh_range = in_out_vrr->max_refresh_in_uhz -
 				in_out_vrr->min_refresh_in_uhz;
 
+		in_out_vrr->btr.margin_in_us = in_out_vrr->max_duration_in_us -
+				2 * in_out_vrr->min_duration_in_us;
+		if (in_out_vrr->btr.margin_in_us > BTR_MAX_MARGIN)
+			in_out_vrr->btr.margin_in_us = BTR_MAX_MARGIN;
+
 		in_out_vrr->supported = true;
 	}
 
@@ -803,6 +816,7 @@ void mod_freesync_build_vrr_params(struct mod_freesync *mod_freesync,
 	in_out_vrr->btr.inserted_duration_in_us = 0;
 	in_out_vrr->btr.frames_to_insert = 0;
 	in_out_vrr->btr.frame_counter = 0;
+
 	in_out_vrr->btr.mid_point_in_us =
 				(in_out_vrr->min_duration_in_us +
 				 in_out_vrr->max_duration_in_us) / 2;
@@ -975,13 +989,9 @@ void mod_freesync_get_settings(struct mod_freesync *mod_freesync,
 		unsigned int *inserted_frames,
 		unsigned int *inserted_duration_in_us)
 {
-	struct core_freesync *core_freesync = NULL;
-
 	if (mod_freesync == NULL)
 		return;
 
-	core_freesync = MOD_FREESYNC_TO_CORE(mod_freesync);
-
 	if (vrr->supported) {
 		*v_total_min = vrr->adjust.v_total_min;
 		*v_total_max = vrr->adjust.v_total_max;
@@ -996,14 +1006,13 @@ unsigned long long mod_freesync_calc_nominal_field_rate(
 			const struct dc_stream_state *stream)
 {
 	unsigned long long nominal_field_rate_in_uhz = 0;
+	unsigned int total = stream->timing.h_total * stream->timing.v_total;
 
-	/* Calculate nominal field rate for stream */
+	/* Calculate nominal field rate for stream, rounded up to nearest integer */
 	nominal_field_rate_in_uhz = stream->timing.pix_clk_100hz / 10;
 	nominal_field_rate_in_uhz *= 1000ULL * 1000ULL * 1000ULL;
-	nominal_field_rate_in_uhz = div_u64(nominal_field_rate_in_uhz,
-						stream->timing.h_total);
-	nominal_field_rate_in_uhz = div_u64(nominal_field_rate_in_uhz,
-						stream->timing.v_total);
+
+	nominal_field_rate_in_uhz =	div_u64(nominal_field_rate_in_uhz, total);
 
 	return nominal_field_rate_in_uhz;
 }
diff --git a/drivers/gpu/drm/amd/display/modules/hdcp/Makefile b/drivers/gpu/drm/amd/display/modules/hdcp/Makefile
new file mode 100644
index 000000000000..1c3c6d47973a
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/modules/hdcp/Makefile
@@ -0,0 +1,32 @@
+#
+# Copyright 2019 Advanced Micro Devices, Inc.
+#
+# Permission is hereby granted, free of charge, to any person obtaining a
+# copy of this software and associated documentation files (the "Software"),
+# to deal in the Software without restriction, including without limitation
+# the rights to use, copy, modify, merge, publish, distribute, sublicense,
+# and/or sell copies of the Software, and to permit persons to whom the
+# Software is furnished to do so, subject to the following conditions:
+#
+# The above copyright notice and this permission notice shall be included in
+# all copies or substantial portions of the Software.
+#
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+# THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+# OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+# ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+# OTHER DEALINGS IN THE SOFTWARE.
+#
+#
+# Makefile for the 'hdcp' sub-module of DAL.
+#
+
+HDCP = hdcp_ddc.o hdcp_log.o hdcp_psp.o hdcp.o \
+		hdcp1_execution.o hdcp1_transition.o
+
+AMD_DAL_HDCP = $(addprefix $(AMDDALPATH)/modules/hdcp/,$(HDCP))
+#$(info ************  DAL-HDCP_MAKEFILE ************)
+
+AMD_DISPLAY_FILES += $(AMD_DAL_HDCP)
diff --git a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp.c b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp.c
new file mode 100644
index 000000000000..d7ac445dec6f
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp.c
@@ -0,0 +1,426 @@
+/*
+ * Copyright 2019 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#include "hdcp.h"
+
+static void push_error_status(struct mod_hdcp *hdcp,
+		enum mod_hdcp_status status)
+{
+	struct mod_hdcp_trace *trace = &hdcp->connection.trace;
+
+	if (trace->error_count < MAX_NUM_OF_ERROR_TRACE) {
+		trace->errors[trace->error_count].status = status;
+		trace->errors[trace->error_count].state_id = hdcp->state.id;
+		trace->error_count++;
+		HDCP_ERROR_TRACE(hdcp, status);
+	}
+
+	hdcp->connection.hdcp1_retry_count++;
+}
+
+static uint8_t is_cp_desired_hdcp1(struct mod_hdcp *hdcp)
+{
+	int i, display_enabled = 0;
+
+	/* if all displays on the link are disabled, hdcp is not desired */
+	for (i = 0; i < MAX_NUM_OF_DISPLAYS; i++) {
+		if (hdcp->connection.displays[i].state != MOD_HDCP_DISPLAY_INACTIVE &&
+				!hdcp->connection.displays[i].adjust.disable) {
+			display_enabled = 1;
+			break;
+		}
+	}
+
+	return (hdcp->connection.hdcp1_retry_count < MAX_NUM_OF_ATTEMPTS) &&
+			display_enabled && !hdcp->connection.link.adjust.hdcp1.disable;
+}
+
+static enum mod_hdcp_status execution(struct mod_hdcp *hdcp,
+		struct mod_hdcp_event_context *event_ctx,
+		union mod_hdcp_transition_input *input)
+{
+	enum mod_hdcp_status status = MOD_HDCP_STATUS_SUCCESS;
+
+	if (is_in_initialized_state(hdcp)) {
+		if (event_ctx->event != MOD_HDCP_EVENT_CALLBACK) {
+			event_ctx->unexpected_event = 1;
+			goto out;
+		}
+		/* initialize transition input */
+		memset(input, 0, sizeof(union mod_hdcp_transition_input));
+	} else if (is_in_cp_not_desired_state(hdcp)) {
+		if (event_ctx->event != MOD_HDCP_EVENT_CALLBACK) {
+			event_ctx->unexpected_event = 1;
+			goto out;
+		}
+		/* update topology event if hdcp is not desired */
+		status = mod_hdcp_add_display_topology(hdcp);
+	} else if (is_in_hdcp1_states(hdcp)) {
+		status = mod_hdcp_hdcp1_execution(hdcp, event_ctx, &input->hdcp1);
+	} else if (is_in_hdcp1_dp_states(hdcp)) {
+		status = mod_hdcp_hdcp1_dp_execution(hdcp,
+				event_ctx, &input->hdcp1);
+	}
+out:
+	return status;
+}
+
+static enum mod_hdcp_status transition(struct mod_hdcp *hdcp,
+		struct mod_hdcp_event_context *event_ctx,
+		union mod_hdcp_transition_input *input,
+		struct mod_hdcp_output *output)
+{
+	enum mod_hdcp_status status = MOD_HDCP_STATUS_SUCCESS;
+
+	if (event_ctx->unexpected_event)
+		goto out;
+
+	if (is_in_initialized_state(hdcp)) {
+		if (is_dp_hdcp(hdcp))
+			if (is_cp_desired_hdcp1(hdcp)) {
+				callback_in_ms(0, output);
+				set_state_id(hdcp, output, D1_A0_DETERMINE_RX_HDCP_CAPABLE);
+			} else {
+				callback_in_ms(0, output);
+				set_state_id(hdcp, output, HDCP_CP_NOT_DESIRED);
+			}
+		else if (is_hdmi_dvi_sl_hdcp(hdcp))
+			if (is_cp_desired_hdcp1(hdcp)) {
+				callback_in_ms(0, output);
+				set_state_id(hdcp, output, H1_A0_WAIT_FOR_ACTIVE_RX);
+			} else {
+				callback_in_ms(0, output);
+				set_state_id(hdcp, output, HDCP_CP_NOT_DESIRED);
+			}
+		else {
+			callback_in_ms(0, output);
+			set_state_id(hdcp, output, HDCP_CP_NOT_DESIRED);
+		}
+	} else if (is_in_cp_not_desired_state(hdcp)) {
+		increment_stay_counter(hdcp);
+	} else if (is_in_hdcp1_states(hdcp)) {
+		status = mod_hdcp_hdcp1_transition(hdcp,
+				event_ctx, &input->hdcp1, output);
+	} else if (is_in_hdcp1_dp_states(hdcp)) {
+		status = mod_hdcp_hdcp1_dp_transition(hdcp,
+				event_ctx, &input->hdcp1, output);
+	} else {
+		status = MOD_HDCP_STATUS_INVALID_STATE;
+	}
+out:
+	return status;
+}
+
+static enum mod_hdcp_status reset_authentication(struct mod_hdcp *hdcp,
+		struct mod_hdcp_output *output)
+{
+	enum mod_hdcp_status status = MOD_HDCP_STATUS_SUCCESS;
+
+	if (is_hdcp1(hdcp)) {
+		if (hdcp->auth.trans_input.hdcp1.create_session != UNKNOWN)
+			mod_hdcp_hdcp1_destroy_session(hdcp);
+
+		if (hdcp->auth.trans_input.hdcp1.add_topology == PASS) {
+			status = mod_hdcp_remove_display_topology(hdcp);
+			if (status != MOD_HDCP_STATUS_SUCCESS) {
+				output->callback_needed = 0;
+				output->watchdog_timer_needed = 0;
+				goto out;
+			}
+		}
+		HDCP_TOP_RESET_AUTH_TRACE(hdcp);
+		memset(&hdcp->auth, 0, sizeof(struct mod_hdcp_authentication));
+		memset(&hdcp->state, 0, sizeof(struct mod_hdcp_state));
+		set_state_id(hdcp, output, HDCP_INITIALIZED);
+	} else if (is_in_cp_not_desired_state(hdcp)) {
+		status = mod_hdcp_remove_display_topology(hdcp);
+		if (status != MOD_HDCP_STATUS_SUCCESS) {
+			output->callback_needed = 0;
+			output->watchdog_timer_needed = 0;
+			goto out;
+		}
+		HDCP_TOP_RESET_AUTH_TRACE(hdcp);
+		memset(&hdcp->auth, 0, sizeof(struct mod_hdcp_authentication));
+		memset(&hdcp->state, 0, sizeof(struct mod_hdcp_state));
+		set_state_id(hdcp, output, HDCP_INITIALIZED);
+	}
+
+out:
+	/* stop callback and watchdog requests from previous authentication*/
+	output->watchdog_timer_stop = 1;
+	output->callback_stop = 1;
+	return status;
+}
+
+static enum mod_hdcp_status reset_connection(struct mod_hdcp *hdcp,
+		struct mod_hdcp_output *output)
+{
+	enum mod_hdcp_status status = MOD_HDCP_STATUS_SUCCESS;
+
+	memset(output, 0, sizeof(struct mod_hdcp_output));
+
+	status = reset_authentication(hdcp, output);
+	if (status != MOD_HDCP_STATUS_SUCCESS)
+		goto out;
+
+	if (current_state(hdcp) != HDCP_UNINITIALIZED) {
+		HDCP_TOP_RESET_CONN_TRACE(hdcp);
+		set_state_id(hdcp, output, HDCP_UNINITIALIZED);
+	}
+	memset(&hdcp->connection, 0, sizeof(hdcp->connection));
+out:
+	return status;
+}
+
+/*
+ * Implementation of functions in mod_hdcp.h
+ */
+size_t mod_hdcp_get_memory_size(void)
+{
+	return sizeof(struct mod_hdcp);
+}
+
+enum mod_hdcp_status mod_hdcp_setup(struct mod_hdcp *hdcp,
+		struct mod_hdcp_config *config)
+{
+	struct mod_hdcp_output output;
+	enum mod_hdcp_status status = MOD_HDCP_STATUS_SUCCESS;
+
+	memset(hdcp, 0, sizeof(struct mod_hdcp));
+	memset(&output, 0, sizeof(output));
+	hdcp->config = *config;
+	HDCP_TOP_INTERFACE_TRACE(hdcp);
+	status = reset_connection(hdcp, &output);
+	if (status != MOD_HDCP_STATUS_SUCCESS)
+		push_error_status(hdcp, status);
+	return status;
+}
+
+enum mod_hdcp_status mod_hdcp_teardown(struct mod_hdcp *hdcp)
+{
+	enum mod_hdcp_status status = MOD_HDCP_STATUS_SUCCESS;
+	struct mod_hdcp_output output;
+
+	HDCP_TOP_INTERFACE_TRACE(hdcp);
+	memset(&output, 0,  sizeof(output));
+	status = reset_connection(hdcp, &output);
+	if (status == MOD_HDCP_STATUS_SUCCESS)
+		memset(hdcp, 0, sizeof(struct mod_hdcp));
+	else
+		push_error_status(hdcp, status);
+	return status;
+}
+
+enum mod_hdcp_status mod_hdcp_add_display(struct mod_hdcp *hdcp,
+		struct mod_hdcp_link *link, struct mod_hdcp_display *display,
+		struct mod_hdcp_output *output)
+{
+	enum mod_hdcp_status status = MOD_HDCP_STATUS_SUCCESS;
+	struct mod_hdcp_display *display_container = NULL;
+
+	HDCP_TOP_INTERFACE_TRACE_WITH_INDEX(hdcp, display->index);
+	memset(output, 0, sizeof(struct mod_hdcp_output));
+
+	/* skip inactive display */
+	if (display->state != MOD_HDCP_DISPLAY_ACTIVE) {
+		status = MOD_HDCP_STATUS_SUCCESS;
+		goto out;
+	}
+
+	/* check existing display container */
+	if (get_active_display_at_index(hdcp, display->index)) {
+		status = MOD_HDCP_STATUS_SUCCESS;
+		goto out;
+	}
+
+	/* find an empty display container */
+	display_container = get_empty_display_container(hdcp);
+	if (!display_container) {
+		status = MOD_HDCP_STATUS_DISPLAY_OUT_OF_BOUND;
+		goto out;
+	}
+
+	/* reset existing authentication status */
+	status = reset_authentication(hdcp, output);
+	if (status != MOD_HDCP_STATUS_SUCCESS)
+		goto out;
+
+	/* add display to connection */
+	hdcp->connection.link = *link;
+	*display_container = *display;
+
+	/* reset retry counters */
+	reset_retry_counts(hdcp);
+
+	/* reset error trace */
+	memset(&hdcp->connection.trace, 0, sizeof(hdcp->connection.trace));
+
+	/* request authentication */
+	if (current_state(hdcp) != HDCP_INITIALIZED)
+		set_state_id(hdcp, output, HDCP_INITIALIZED);
+	callback_in_ms(hdcp->connection.link.adjust.auth_delay * 1000, output);
+out:
+	if (status != MOD_HDCP_STATUS_SUCCESS)
+		push_error_status(hdcp, status);
+
+	return status;
+}
+
+enum mod_hdcp_status mod_hdcp_remove_display(struct mod_hdcp *hdcp,
+		uint8_t index, struct mod_hdcp_output *output)
+{
+	enum mod_hdcp_status status = MOD_HDCP_STATUS_SUCCESS;
+	struct mod_hdcp_display *display = NULL;
+
+	HDCP_TOP_INTERFACE_TRACE_WITH_INDEX(hdcp, index);
+	memset(output, 0, sizeof(struct mod_hdcp_output));
+
+	/* find display in connection */
+	display = get_active_display_at_index(hdcp, index);
+	if (!display) {
+		status = MOD_HDCP_STATUS_SUCCESS;
+		goto out;
+	}
+
+	/* stop current authentication */
+	status = reset_authentication(hdcp, output);
+	if (status != MOD_HDCP_STATUS_SUCCESS)
+		goto out;
+
+	/* remove display */
+	display->state = MOD_HDCP_DISPLAY_INACTIVE;
+
+	/* clear retry counters */
+	reset_retry_counts(hdcp);
+
+	/* reset error trace */
+	memset(&hdcp->connection.trace, 0, sizeof(hdcp->connection.trace));
+
+	/* request authentication for remaining displays*/
+	if (get_active_display_count(hdcp) > 0)
+		callback_in_ms(hdcp->connection.link.adjust.auth_delay * 1000,
+				output);
+out:
+	if (status != MOD_HDCP_STATUS_SUCCESS)
+		push_error_status(hdcp, status);
+	return status;
+}
+
+enum mod_hdcp_status mod_hdcp_query_display(struct mod_hdcp *hdcp,
+		uint8_t index, struct mod_hdcp_display_query *query)
+{
+	enum mod_hdcp_status status = MOD_HDCP_STATUS_SUCCESS;
+	struct mod_hdcp_display *display = NULL;
+
+	/* find display in connection */
+	display = get_active_display_at_index(hdcp, index);
+	if (!display) {
+		status = MOD_HDCP_STATUS_DISPLAY_NOT_FOUND;
+		goto out;
+	}
+
+	/* populate query */
+	query->link = &hdcp->connection.link;
+	query->display = display;
+	query->trace = &hdcp->connection.trace;
+	query->encryption_status = MOD_HDCP_ENCRYPTION_STATUS_HDCP_OFF;
+
+	mod_hdcp_hdcp1_get_link_encryption_status(hdcp, &query->encryption_status);
+
+out:
+	return status;
+}
+
+enum mod_hdcp_status mod_hdcp_reset_connection(struct mod_hdcp *hdcp,
+		struct mod_hdcp_output *output)
+{
+	enum mod_hdcp_status status = MOD_HDCP_STATUS_SUCCESS;
+
+	HDCP_TOP_INTERFACE_TRACE(hdcp);
+	status = reset_connection(hdcp, output);
+	if (status != MOD_HDCP_STATUS_SUCCESS)
+		push_error_status(hdcp, status);
+
+	return status;
+}
+
+enum mod_hdcp_status mod_hdcp_process_event(struct mod_hdcp *hdcp,
+		enum mod_hdcp_event event, struct mod_hdcp_output *output)
+{
+	enum mod_hdcp_status exec_status, trans_status, reset_status, status;
+	struct mod_hdcp_event_context event_ctx;
+
+	HDCP_EVENT_TRACE(hdcp, event);
+	memset(output, 0, sizeof(struct mod_hdcp_output));
+	memset(&event_ctx, 0, sizeof(struct mod_hdcp_event_context));
+	event_ctx.event = event;
+
+	/* execute and transition */
+	exec_status = execution(hdcp, &event_ctx, &hdcp->auth.trans_input);
+	trans_status = transition(
+			hdcp, &event_ctx, &hdcp->auth.trans_input, output);
+	if (trans_status == MOD_HDCP_STATUS_SUCCESS) {
+		status = MOD_HDCP_STATUS_SUCCESS;
+	} else if (exec_status == MOD_HDCP_STATUS_SUCCESS) {
+		status = MOD_HDCP_STATUS_INTERNAL_POLICY_FAILURE;
+		push_error_status(hdcp, status);
+	} else {
+		status = exec_status;
+		push_error_status(hdcp, status);
+	}
+
+	/* reset authentication if needed */
+	if (trans_status == MOD_HDCP_STATUS_RESET_NEEDED) {
+		HDCP_FULL_DDC_TRACE(hdcp);
+		reset_status = reset_authentication(hdcp, output);
+		if (reset_status != MOD_HDCP_STATUS_SUCCESS)
+			push_error_status(hdcp, reset_status);
+	}
+	return status;
+}
+
+enum mod_hdcp_operation_mode mod_hdcp_signal_type_to_operation_mode(
+		enum signal_type signal)
+{
+	enum mod_hdcp_operation_mode mode = MOD_HDCP_MODE_OFF;
+
+	switch (signal) {
+	case SIGNAL_TYPE_DVI_SINGLE_LINK:
+	case SIGNAL_TYPE_HDMI_TYPE_A:
+		mode = MOD_HDCP_MODE_DEFAULT;
+		break;
+	case SIGNAL_TYPE_EDP:
+	case SIGNAL_TYPE_DISPLAY_PORT:
+		mode = MOD_HDCP_MODE_DP;
+		break;
+	case SIGNAL_TYPE_DISPLAY_PORT_MST:
+		mode = MOD_HDCP_MODE_DP_MST;
+		break;
+	default:
+		break;
+	};
+
+	return mode;
+}
diff --git a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp.h b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp.h
new file mode 100644
index 000000000000..5664bc0b5bd0
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp.h
@@ -0,0 +1,442 @@
+/*
+ * Copyright 2019 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#ifndef HDCP_H_
+#define HDCP_H_
+
+#include "mod_hdcp.h"
+#include "hdcp_log.h"
+
+#define BCAPS_READY_MASK				0x20
+#define BCAPS_REPEATER_MASK				0x40
+#define BSTATUS_DEVICE_COUNT_MASK			0X007F
+#define BSTATUS_MAX_DEVS_EXCEEDED_MASK			0x0080
+#define BSTATUS_MAX_CASCADE_EXCEEDED_MASK		0x0800
+#define BCAPS_HDCP_CAPABLE_MASK_DP			0x01
+#define BCAPS_REPEATER_MASK_DP				0x02
+#define BSTATUS_READY_MASK_DP				0x01
+#define BSTATUS_R0_P_AVAILABLE_MASK_DP			0x02
+#define BSTATUS_LINK_INTEGRITY_FAILURE_MASK_DP		0x04
+#define BSTATUS_REAUTH_REQUEST_MASK_DP			0x08
+#define BINFO_DEVICE_COUNT_MASK_DP			0X007F
+#define BINFO_MAX_DEVS_EXCEEDED_MASK_DP			0x0080
+#define BINFO_MAX_CASCADE_EXCEEDED_MASK_DP		0x0800
+
+#define RXSTATUS_MSG_SIZE_MASK				0x03FF
+#define RXSTATUS_READY_MASK				0x0400
+#define RXSTATUS_REAUTH_REQUEST_MASK			0x0800
+#define RXIDLIST_DEVICE_COUNT_LOWER_MASK		0xf0
+#define RXIDLIST_DEVICE_COUNT_UPPER_MASK		0x01
+#define RXCAPS_BYTE0_HDCP_CAPABLE_MASK_DP		0x02
+#define RXSTATUS_READY_MASK_DP				0x0001
+#define RXSTATUS_H_P_AVAILABLE_MASK_DP			0x0002
+#define RXSTATUS_PAIRING_AVAILABLE_MASK_DP		0x0004
+#define RXSTATUS_REAUTH_REQUEST_MASK_DP			0x0008
+#define RXSTATUS_LINK_INTEGRITY_FAILURE_MASK_DP		0x0010
+
+enum mod_hdcp_trans_input_result {
+	UNKNOWN = 0,
+	PASS,
+	FAIL
+};
+
+struct mod_hdcp_transition_input_hdcp1 {
+	uint8_t bksv_read;
+	uint8_t bksv_validation;
+	uint8_t add_topology;
+	uint8_t create_session;
+	uint8_t an_write;
+	uint8_t aksv_write;
+	uint8_t ainfo_write;
+	uint8_t bcaps_read;
+	uint8_t r0p_read;
+	uint8_t rx_validation;
+	uint8_t encryption;
+	uint8_t link_maintenance;
+	uint8_t ready_check;
+	uint8_t bstatus_read;
+	uint8_t max_cascade_check;
+	uint8_t max_devs_check;
+	uint8_t device_count_check;
+	uint8_t ksvlist_read;
+	uint8_t vp_read;
+	uint8_t ksvlist_vp_validation;
+
+	uint8_t hdcp_capable_dp;
+	uint8_t binfo_read_dp;
+	uint8_t r0p_available_dp;
+	uint8_t link_integiry_check;
+	uint8_t reauth_request_check;
+	uint8_t stream_encryption_dp;
+};
+
+union mod_hdcp_transition_input {
+	struct mod_hdcp_transition_input_hdcp1 hdcp1;
+};
+
+struct mod_hdcp_message_hdcp1 {
+	uint8_t		an[8];
+	uint8_t		aksv[5];
+	uint8_t		ainfo;
+	uint8_t		bksv[5];
+	uint16_t	r0p;
+	uint8_t		bcaps;
+	uint16_t	bstatus;
+	uint8_t		ksvlist[635];
+	uint16_t	ksvlist_size;
+	uint8_t		vp[20];
+
+	uint16_t	binfo_dp;
+};
+
+union mod_hdcp_message {
+	struct mod_hdcp_message_hdcp1 hdcp1;
+};
+
+struct mod_hdcp_auth_counters {
+	uint8_t stream_management_retry_count;
+};
+
+/* contains values per connection */
+struct mod_hdcp_connection {
+	struct mod_hdcp_link link;
+	struct mod_hdcp_display displays[MAX_NUM_OF_DISPLAYS];
+	uint8_t is_repeater;
+	uint8_t is_km_stored;
+	struct mod_hdcp_trace trace;
+	uint8_t hdcp1_retry_count;
+};
+
+/* contains values per authentication cycle */
+struct mod_hdcp_authentication {
+	uint32_t id;
+	union mod_hdcp_message msg;
+	union mod_hdcp_transition_input trans_input;
+	struct mod_hdcp_auth_counters count;
+};
+
+/* contains values per state change */
+struct mod_hdcp_state {
+	uint8_t id;
+	uint32_t stay_count;
+};
+
+/* per event in a state */
+struct mod_hdcp_event_context {
+	enum mod_hdcp_event event;
+	uint8_t rx_id_list_ready;
+	uint8_t unexpected_event;
+};
+
+struct mod_hdcp {
+	/* per link */
+	struct mod_hdcp_config config;
+	/* per connection */
+	struct mod_hdcp_connection connection;
+	/* per authentication attempt */
+	struct mod_hdcp_authentication auth;
+	/* per state in an authentication */
+	struct mod_hdcp_state state;
+	/* reserved memory buffer */
+	uint8_t buf[2025];
+};
+
+enum mod_hdcp_initial_state_id {
+	HDCP_UNINITIALIZED = 0x0,
+	HDCP_INITIAL_STATE_START = HDCP_UNINITIALIZED,
+	HDCP_INITIALIZED,
+	HDCP_CP_NOT_DESIRED,
+	HDCP_INITIAL_STATE_END = HDCP_CP_NOT_DESIRED
+};
+
+enum mod_hdcp_hdcp1_state_id {
+	HDCP1_STATE_START = HDCP_INITIAL_STATE_END,
+	H1_A0_WAIT_FOR_ACTIVE_RX,
+	H1_A1_EXCHANGE_KSVS,
+	H1_A2_COMPUTATIONS_A3_VALIDATE_RX_A6_TEST_FOR_REPEATER,
+	H1_A45_AUTHENTICATED,
+	H1_A8_WAIT_FOR_READY,
+	H1_A9_READ_KSV_LIST,
+	HDCP1_STATE_END = H1_A9_READ_KSV_LIST
+};
+
+enum mod_hdcp_hdcp1_dp_state_id {
+	HDCP1_DP_STATE_START = HDCP1_STATE_END,
+	D1_A0_DETERMINE_RX_HDCP_CAPABLE,
+	D1_A1_EXCHANGE_KSVS,
+	D1_A23_WAIT_FOR_R0_PRIME,
+	D1_A2_COMPUTATIONS_A3_VALIDATE_RX_A5_TEST_FOR_REPEATER,
+	D1_A4_AUTHENTICATED,
+	D1_A6_WAIT_FOR_READY,
+	D1_A7_READ_KSV_LIST,
+	HDCP1_DP_STATE_END = D1_A7_READ_KSV_LIST,
+};
+
+/* hdcp1 executions and transitions */
+typedef enum mod_hdcp_status (*mod_hdcp_action)(struct mod_hdcp *hdcp);
+uint8_t mod_hdcp_execute_and_set(
+		mod_hdcp_action func, uint8_t *flag,
+		enum mod_hdcp_status *status, struct mod_hdcp *hdcp, char *str);
+enum mod_hdcp_status mod_hdcp_hdcp1_execution(struct mod_hdcp *hdcp,
+	struct mod_hdcp_event_context *event_ctx,
+	struct mod_hdcp_transition_input_hdcp1 *input);
+enum mod_hdcp_status mod_hdcp_hdcp1_dp_execution(struct mod_hdcp *hdcp,
+	struct mod_hdcp_event_context *event_ctx,
+	struct mod_hdcp_transition_input_hdcp1 *input);
+enum mod_hdcp_status mod_hdcp_hdcp1_transition(struct mod_hdcp *hdcp,
+	struct mod_hdcp_event_context *event_ctx,
+	struct mod_hdcp_transition_input_hdcp1 *input,
+	struct mod_hdcp_output *output);
+enum mod_hdcp_status mod_hdcp_hdcp1_dp_transition(struct mod_hdcp *hdcp,
+	struct mod_hdcp_event_context *event_ctx,
+	struct mod_hdcp_transition_input_hdcp1 *input,
+	struct mod_hdcp_output *output);
+
+/* log functions */
+void mod_hdcp_dump_binary_message(uint8_t *msg, uint32_t msg_size,
+		uint8_t *buf, uint32_t buf_size);
+/* TODO: add adjustment log */
+
+/* psp functions */
+enum mod_hdcp_status mod_hdcp_add_display_topology(
+		struct mod_hdcp *hdcp);
+enum mod_hdcp_status mod_hdcp_remove_display_topology(
+		struct mod_hdcp *hdcp);
+enum mod_hdcp_status mod_hdcp_hdcp1_create_session(struct mod_hdcp *hdcp);
+enum mod_hdcp_status mod_hdcp_hdcp1_destroy_session(struct mod_hdcp *hdcp);
+enum mod_hdcp_status mod_hdcp_hdcp1_validate_rx(struct mod_hdcp *hdcp);
+enum mod_hdcp_status mod_hdcp_hdcp1_enable_encryption(struct mod_hdcp *hdcp);
+enum mod_hdcp_status mod_hdcp_hdcp1_validate_ksvlist_vp(struct mod_hdcp *hdcp);
+enum mod_hdcp_status mod_hdcp_hdcp1_enable_dp_stream_encryption(
+	struct mod_hdcp *hdcp);
+enum mod_hdcp_status mod_hdcp_hdcp1_link_maintenance(struct mod_hdcp *hdcp);
+enum mod_hdcp_status mod_hdcp_hdcp1_get_link_encryption_status(struct mod_hdcp *hdcp,
+							       enum mod_hdcp_encryption_status *encryption_status);
+/* ddc functions */
+enum mod_hdcp_status mod_hdcp_read_bksv(struct mod_hdcp *hdcp);
+enum mod_hdcp_status mod_hdcp_read_bcaps(struct mod_hdcp *hdcp);
+enum mod_hdcp_status mod_hdcp_read_bstatus(struct mod_hdcp *hdcp);
+enum mod_hdcp_status mod_hdcp_read_r0p(struct mod_hdcp *hdcp);
+enum mod_hdcp_status mod_hdcp_read_ksvlist(struct mod_hdcp *hdcp);
+enum mod_hdcp_status mod_hdcp_read_vp(struct mod_hdcp *hdcp);
+enum mod_hdcp_status mod_hdcp_read_binfo(struct mod_hdcp *hdcp);
+enum mod_hdcp_status mod_hdcp_write_aksv(struct mod_hdcp *hdcp);
+enum mod_hdcp_status mod_hdcp_write_ainfo(struct mod_hdcp *hdcp);
+enum mod_hdcp_status mod_hdcp_write_an(struct mod_hdcp *hdcp);
+enum mod_hdcp_status mod_hdcp_read_rxcaps(struct mod_hdcp *hdcp);
+enum mod_hdcp_status mod_hdcp_read_rxstatus(struct mod_hdcp *hdcp);
+enum mod_hdcp_status mod_hdcp_read_ake_cert(struct mod_hdcp *hdcp);
+enum mod_hdcp_status mod_hdcp_read_h_prime(struct mod_hdcp *hdcp);
+enum mod_hdcp_status mod_hdcp_read_pairing_info(struct mod_hdcp *hdcp);
+enum mod_hdcp_status mod_hdcp_read_l_prime(struct mod_hdcp *hdcp);
+enum mod_hdcp_status mod_hdcp_read_rx_id_list(struct mod_hdcp *hdcp);
+enum mod_hdcp_status mod_hdcp_read_stream_ready(struct mod_hdcp *hdcp);
+enum mod_hdcp_status mod_hdcp_write_ake_init(struct mod_hdcp *hdcp);
+enum mod_hdcp_status mod_hdcp_write_no_stored_km(struct mod_hdcp *hdcp);
+enum mod_hdcp_status mod_hdcp_write_stored_km(struct mod_hdcp *hdcp);
+enum mod_hdcp_status mod_hdcp_write_lc_init(struct mod_hdcp *hdcp);
+enum mod_hdcp_status mod_hdcp_write_eks(struct mod_hdcp *hdcp);
+enum mod_hdcp_status mod_hdcp_write_repeater_auth_ack(struct mod_hdcp *hdcp);
+enum mod_hdcp_status mod_hdcp_write_stream_manage(struct mod_hdcp *hdcp);
+enum mod_hdcp_status mod_hdcp_write_content_type(struct mod_hdcp *hdcp);
+
+/* hdcp version helpers */
+static inline uint8_t is_dp_hdcp(struct mod_hdcp *hdcp)
+{
+	return (hdcp->connection.link.mode == MOD_HDCP_MODE_DP ||
+			hdcp->connection.link.mode == MOD_HDCP_MODE_DP_MST);
+}
+
+static inline uint8_t is_dp_mst_hdcp(struct mod_hdcp *hdcp)
+{
+	return (hdcp->connection.link.mode == MOD_HDCP_MODE_DP_MST);
+}
+
+static inline uint8_t is_hdmi_dvi_sl_hdcp(struct mod_hdcp *hdcp)
+{
+	return (hdcp->connection.link.mode == MOD_HDCP_MODE_DEFAULT);
+}
+
+/* hdcp state helpers */
+static inline uint8_t current_state(struct mod_hdcp *hdcp)
+{
+	return hdcp->state.id;
+}
+
+static inline void set_state_id(struct mod_hdcp *hdcp,
+		struct mod_hdcp_output *output, uint8_t id)
+{
+	memset(&hdcp->state, 0, sizeof(hdcp->state));
+	hdcp->state.id = id;
+	/* callback timer should be reset per state */
+	output->callback_stop = 1;
+	output->watchdog_timer_stop = 1;
+	HDCP_NEXT_STATE_TRACE(hdcp, id, output);
+}
+
+static inline uint8_t is_in_hdcp1_states(struct mod_hdcp *hdcp)
+{
+	return (current_state(hdcp) > HDCP1_STATE_START &&
+			current_state(hdcp) <= HDCP1_STATE_END);
+}
+
+static inline uint8_t is_in_hdcp1_dp_states(struct mod_hdcp *hdcp)
+{
+	return (current_state(hdcp) > HDCP1_DP_STATE_START &&
+			current_state(hdcp) <= HDCP1_DP_STATE_END);
+}
+
+static inline uint8_t is_hdcp1(struct mod_hdcp *hdcp)
+{
+	return (is_in_hdcp1_states(hdcp) || is_in_hdcp1_dp_states(hdcp));
+}
+
+static inline uint8_t is_in_cp_not_desired_state(struct mod_hdcp *hdcp)
+{
+	return current_state(hdcp) == HDCP_CP_NOT_DESIRED;
+}
+
+static inline uint8_t is_in_initialized_state(struct mod_hdcp *hdcp)
+{
+	return current_state(hdcp) == HDCP_INITIALIZED;
+}
+
+/* transition operation helpers */
+static inline void increment_stay_counter(struct mod_hdcp *hdcp)
+{
+	hdcp->state.stay_count++;
+}
+
+static inline void fail_and_restart_in_ms(uint16_t time,
+		enum mod_hdcp_status *status,
+		struct mod_hdcp_output *output)
+{
+	output->callback_needed = 1;
+	output->callback_delay = time;
+	output->watchdog_timer_needed = 0;
+	output->watchdog_timer_delay = 0;
+	*status = MOD_HDCP_STATUS_RESET_NEEDED;
+}
+
+static inline void callback_in_ms(uint16_t time, struct mod_hdcp_output *output)
+{
+	output->callback_needed = 1;
+	output->callback_delay = time;
+}
+
+static inline void set_watchdog_in_ms(struct mod_hdcp *hdcp, uint16_t time,
+		struct mod_hdcp_output *output)
+{
+	output->watchdog_timer_needed = 1;
+	output->watchdog_timer_delay = time;
+}
+
+/* connection topology helpers */
+static inline uint8_t is_display_active(struct mod_hdcp_display *display)
+{
+	return display->state >= MOD_HDCP_DISPLAY_ACTIVE;
+}
+
+static inline uint8_t is_display_added(struct mod_hdcp_display *display)
+{
+	return display->state >= MOD_HDCP_DISPLAY_ACTIVE_AND_ADDED;
+}
+
+static inline uint8_t is_display_encryption_enabled(struct mod_hdcp_display *display)
+{
+	return display->state >= MOD_HDCP_DISPLAY_ENCRYPTION_ENABLED;
+}
+
+static inline uint8_t get_active_display_count(struct mod_hdcp *hdcp)
+{
+	uint8_t added_count = 0;
+	uint8_t i;
+
+	for (i = 0; i < MAX_NUM_OF_DISPLAYS; i++)
+		if (is_display_active(&hdcp->connection.displays[i]))
+			added_count++;
+	return added_count;
+}
+
+static inline uint8_t get_added_display_count(struct mod_hdcp *hdcp)
+{
+	uint8_t added_count = 0;
+	uint8_t i;
+
+	for (i = 0; i < MAX_NUM_OF_DISPLAYS; i++)
+		if (is_display_added(&hdcp->connection.displays[i]))
+			added_count++;
+	return added_count;
+}
+
+static inline struct mod_hdcp_display *get_first_added_display(
+		struct mod_hdcp *hdcp)
+{
+	uint8_t i;
+	struct mod_hdcp_display *display = NULL;
+
+	for (i = 0; i < MAX_NUM_OF_DISPLAYS; i++)
+		if (is_display_added(&hdcp->connection.displays[i])) {
+			display = &hdcp->connection.displays[i];
+			break;
+		}
+	return display;
+}
+
+static inline struct mod_hdcp_display *get_active_display_at_index(
+		struct mod_hdcp *hdcp, uint8_t index)
+{
+	uint8_t i;
+	struct mod_hdcp_display *display = NULL;
+
+	for (i = 0; i < MAX_NUM_OF_DISPLAYS; i++)
+		if (hdcp->connection.displays[i].index == index &&
+				is_display_active(&hdcp->connection.displays[i])) {
+			display = &hdcp->connection.displays[i];
+			break;
+		}
+	return display;
+}
+
+static inline struct mod_hdcp_display *get_empty_display_container(
+		struct mod_hdcp *hdcp)
+{
+	uint8_t i;
+	struct mod_hdcp_display *display = NULL;
+
+	for (i = 0; i < MAX_NUM_OF_DISPLAYS; i++)
+		if (!is_display_active(&hdcp->connection.displays[i])) {
+			display = &hdcp->connection.displays[i];
+			break;
+		}
+	return display;
+}
+
+static inline void reset_retry_counts(struct mod_hdcp *hdcp)
+{
+	hdcp->connection.hdcp1_retry_count = 0;
+}
+
+#endif /* HDCP_H_ */
diff --git a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp1_execution.c b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp1_execution.c
new file mode 100644
index 000000000000..3db4a7da414f
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp1_execution.c
@@ -0,0 +1,531 @@
+/*
+ * Copyright 2019 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#include "hdcp.h"
+
+static inline enum mod_hdcp_status validate_bksv(struct mod_hdcp *hdcp)
+{
+	uint64_t n = *(uint64_t *)hdcp->auth.msg.hdcp1.bksv;
+	uint8_t count = 0;
+
+	while (n) {
+		count++;
+		n &= (n - 1);
+	}
+	return (count == 20) ? MOD_HDCP_STATUS_SUCCESS :
+			MOD_HDCP_STATUS_HDCP1_INVALID_BKSV;
+}
+
+static inline enum mod_hdcp_status check_ksv_ready(struct mod_hdcp *hdcp)
+{
+	if (is_dp_hdcp(hdcp))
+		return (hdcp->auth.msg.hdcp1.bstatus & BSTATUS_READY_MASK_DP) ?
+				MOD_HDCP_STATUS_SUCCESS :
+				MOD_HDCP_STATUS_HDCP1_KSV_LIST_NOT_READY;
+	return (hdcp->auth.msg.hdcp1.bcaps & BCAPS_READY_MASK) ?
+			MOD_HDCP_STATUS_SUCCESS :
+			MOD_HDCP_STATUS_HDCP1_KSV_LIST_NOT_READY;
+}
+
+static inline enum mod_hdcp_status check_hdcp_capable_dp(struct mod_hdcp *hdcp)
+{
+	return (hdcp->auth.msg.hdcp1.bcaps & BCAPS_HDCP_CAPABLE_MASK_DP) ?
+			MOD_HDCP_STATUS_SUCCESS :
+			MOD_HDCP_STATUS_HDCP1_NOT_CAPABLE;
+}
+
+static inline enum mod_hdcp_status check_r0p_available_dp(struct mod_hdcp *hdcp)
+{
+	enum mod_hdcp_status status;
+	if (is_dp_hdcp(hdcp)) {
+		status = (hdcp->auth.msg.hdcp1.bstatus &
+				BSTATUS_R0_P_AVAILABLE_MASK_DP) ?
+			MOD_HDCP_STATUS_SUCCESS :
+			MOD_HDCP_STATUS_HDCP1_R0_PRIME_PENDING;
+	} else {
+		status = MOD_HDCP_STATUS_INVALID_OPERATION;
+	}
+	return status;
+}
+
+static inline enum mod_hdcp_status check_link_integrity_dp(
+		struct mod_hdcp *hdcp)
+{
+	return (hdcp->auth.msg.hdcp1.bstatus &
+			BSTATUS_LINK_INTEGRITY_FAILURE_MASK_DP) ?
+			MOD_HDCP_STATUS_HDCP1_LINK_INTEGRITY_FAILURE :
+			MOD_HDCP_STATUS_SUCCESS;
+}
+
+static inline enum mod_hdcp_status check_no_reauthentication_request_dp(
+		struct mod_hdcp *hdcp)
+{
+	return (hdcp->auth.msg.hdcp1.bstatus & BSTATUS_REAUTH_REQUEST_MASK_DP) ?
+			MOD_HDCP_STATUS_HDCP1_REAUTH_REQUEST_ISSUED :
+			MOD_HDCP_STATUS_SUCCESS;
+}
+
+static inline enum mod_hdcp_status check_no_max_cascade(struct mod_hdcp *hdcp)
+{
+	enum mod_hdcp_status status;
+
+	if (is_dp_hdcp(hdcp))
+		status = (hdcp->auth.msg.hdcp1.binfo_dp &
+				BINFO_MAX_CASCADE_EXCEEDED_MASK_DP) ?
+			MOD_HDCP_STATUS_HDCP1_MAX_CASCADE_EXCEEDED_FAILURE :
+			MOD_HDCP_STATUS_SUCCESS;
+	else
+		status = (hdcp->auth.msg.hdcp1.bstatus &
+				BSTATUS_MAX_CASCADE_EXCEEDED_MASK) ?
+				MOD_HDCP_STATUS_HDCP1_MAX_CASCADE_EXCEEDED_FAILURE :
+				MOD_HDCP_STATUS_SUCCESS;
+	return status;
+}
+
+static inline enum mod_hdcp_status check_no_max_devs(struct mod_hdcp *hdcp)
+{
+	enum mod_hdcp_status status;
+
+	if (is_dp_hdcp(hdcp))
+		status = (hdcp->auth.msg.hdcp1.binfo_dp &
+				BINFO_MAX_DEVS_EXCEEDED_MASK_DP) ?
+				MOD_HDCP_STATUS_HDCP1_MAX_DEVS_EXCEEDED_FAILURE :
+				MOD_HDCP_STATUS_SUCCESS;
+	else
+		status = (hdcp->auth.msg.hdcp1.bstatus &
+				BSTATUS_MAX_DEVS_EXCEEDED_MASK) ?
+				MOD_HDCP_STATUS_HDCP1_MAX_DEVS_EXCEEDED_FAILURE :
+				MOD_HDCP_STATUS_SUCCESS;
+	return status;
+}
+
+static inline uint8_t get_device_count(struct mod_hdcp *hdcp)
+{
+	return is_dp_hdcp(hdcp) ?
+			(hdcp->auth.msg.hdcp1.binfo_dp & BINFO_DEVICE_COUNT_MASK_DP) :
+			(hdcp->auth.msg.hdcp1.bstatus & BSTATUS_DEVICE_COUNT_MASK);
+}
+
+static inline enum mod_hdcp_status check_device_count(struct mod_hdcp *hdcp)
+{
+	/* device count must be greater than or equal to tracked hdcp displays */
+	return (get_device_count(hdcp) < get_added_display_count(hdcp)) ?
+			MOD_HDCP_STATUS_HDCP1_DEVICE_COUNT_MISMATCH_FAILURE :
+			MOD_HDCP_STATUS_SUCCESS;
+}
+
+static enum mod_hdcp_status wait_for_active_rx(struct mod_hdcp *hdcp,
+		struct mod_hdcp_event_context *event_ctx,
+		struct mod_hdcp_transition_input_hdcp1 *input)
+{
+	enum mod_hdcp_status status = MOD_HDCP_STATUS_SUCCESS;
+
+	if (event_ctx->event != MOD_HDCP_EVENT_CALLBACK) {
+		event_ctx->unexpected_event = 1;
+		goto out;
+	}
+
+	if (!mod_hdcp_execute_and_set(mod_hdcp_read_bksv,
+			&input->bksv_read, &status,
+			hdcp, "bksv_read"))
+		goto out;
+	if (!mod_hdcp_execute_and_set(mod_hdcp_read_bcaps,
+			&input->bcaps_read, &status,
+			hdcp, "bcaps_read"))
+		goto out;
+out:
+	return status;
+}
+
+static enum mod_hdcp_status exchange_ksvs(struct mod_hdcp *hdcp,
+		struct mod_hdcp_event_context *event_ctx,
+		struct mod_hdcp_transition_input_hdcp1 *input)
+{
+	enum mod_hdcp_status status = MOD_HDCP_STATUS_SUCCESS;
+
+	if (event_ctx->event != MOD_HDCP_EVENT_CALLBACK) {
+		event_ctx->unexpected_event = 1;
+		goto out;
+	}
+
+	if (!mod_hdcp_execute_and_set(mod_hdcp_add_display_topology,
+			&input->add_topology, &status,
+			hdcp, "add_topology"))
+		goto out;
+	if (!mod_hdcp_execute_and_set(mod_hdcp_hdcp1_create_session,
+			&input->create_session, &status,
+			hdcp, "create_session"))
+		goto out;
+	if (!mod_hdcp_execute_and_set(mod_hdcp_write_an,
+			&input->an_write, &status,
+			hdcp, "an_write"))
+		goto out;
+	if (!mod_hdcp_execute_and_set(mod_hdcp_write_aksv,
+			&input->aksv_write, &status,
+			hdcp, "aksv_write"))
+		goto out;
+	if (!mod_hdcp_execute_and_set(mod_hdcp_read_bksv,
+			&input->bksv_read, &status,
+			hdcp, "bksv_read"))
+		goto out;
+	if (!mod_hdcp_execute_and_set(validate_bksv,
+			&input->bksv_validation, &status,
+			hdcp, "bksv_validation"))
+		goto out;
+	if (hdcp->auth.msg.hdcp1.ainfo) {
+		if (!mod_hdcp_execute_and_set(mod_hdcp_write_ainfo,
+				&input->ainfo_write, &status,
+				hdcp, "ainfo_write"))
+			goto out;
+	}
+out:
+	return status;
+}
+
+static enum mod_hdcp_status computations_validate_rx_test_for_repeater(
+		struct mod_hdcp *hdcp,
+		struct mod_hdcp_event_context *event_ctx,
+		struct mod_hdcp_transition_input_hdcp1 *input)
+{
+	enum mod_hdcp_status status = MOD_HDCP_STATUS_SUCCESS;
+
+	if (event_ctx->event != MOD_HDCP_EVENT_CALLBACK) {
+		event_ctx->unexpected_event = 1;
+		goto out;
+	}
+
+	if (!mod_hdcp_execute_and_set(mod_hdcp_read_r0p,
+			&input->r0p_read, &status,
+			hdcp, "r0p_read"))
+		goto out;
+	if (!mod_hdcp_execute_and_set(mod_hdcp_hdcp1_validate_rx,
+			&input->rx_validation, &status,
+			hdcp, "rx_validation"))
+		goto out;
+	if (hdcp->connection.is_repeater) {
+		if (!hdcp->connection.link.adjust.hdcp1.postpone_encryption)
+			if (!mod_hdcp_execute_and_set(
+					mod_hdcp_hdcp1_enable_encryption,
+					&input->encryption, &status,
+					hdcp, "encryption"))
+				goto out;
+	} else {
+		if (!mod_hdcp_execute_and_set(mod_hdcp_hdcp1_enable_encryption,
+				&input->encryption, &status,
+				hdcp, "encryption"))
+			goto out;
+		if (is_dp_mst_hdcp(hdcp))
+			if (!mod_hdcp_execute_and_set(
+					mod_hdcp_hdcp1_enable_dp_stream_encryption,
+					&input->stream_encryption_dp, &status,
+					hdcp, "stream_encryption_dp"))
+				goto out;
+	}
+out:
+	return status;
+}
+
+static enum mod_hdcp_status authenticated(struct mod_hdcp *hdcp,
+		struct mod_hdcp_event_context *event_ctx,
+		struct mod_hdcp_transition_input_hdcp1 *input)
+{
+	enum mod_hdcp_status status = MOD_HDCP_STATUS_SUCCESS;
+
+	if (event_ctx->event != MOD_HDCP_EVENT_CALLBACK) {
+		event_ctx->unexpected_event = 1;
+		goto out;
+	}
+
+	if (!mod_hdcp_execute_and_set(mod_hdcp_hdcp1_link_maintenance,
+			&input->link_maintenance, &status,
+			hdcp, "link_maintenance"))
+		goto out;
+out:
+	return status;
+}
+
+static enum mod_hdcp_status wait_for_ready(struct mod_hdcp *hdcp,
+		struct mod_hdcp_event_context *event_ctx,
+		struct mod_hdcp_transition_input_hdcp1 *input)
+{
+	enum mod_hdcp_status status = MOD_HDCP_STATUS_SUCCESS;
+
+	if (event_ctx->event != MOD_HDCP_EVENT_CALLBACK &&
+			event_ctx->event != MOD_HDCP_EVENT_CPIRQ &&
+			event_ctx->event != MOD_HDCP_EVENT_WATCHDOG_TIMEOUT) {
+		event_ctx->unexpected_event = 1;
+		goto out;
+	}
+
+	if (is_dp_hdcp(hdcp)) {
+		if (!mod_hdcp_execute_and_set(mod_hdcp_read_bstatus,
+				&input->bstatus_read, &status,
+				hdcp, "bstatus_read"))
+			goto out;
+		if (!mod_hdcp_execute_and_set(check_link_integrity_dp,
+				&input->link_integiry_check, &status,
+				hdcp, "link_integiry_check"))
+			goto out;
+		if (!mod_hdcp_execute_and_set(check_no_reauthentication_request_dp,
+				&input->reauth_request_check, &status,
+				hdcp, "reauth_request_check"))
+			goto out;
+	} else {
+		if (!mod_hdcp_execute_and_set(mod_hdcp_read_bcaps,
+				&input->bcaps_read, &status,
+				hdcp, "bcaps_read"))
+			goto out;
+	}
+	if (!mod_hdcp_execute_and_set(check_ksv_ready,
+			&input->ready_check, &status,
+			hdcp, "ready_check"))
+		goto out;
+out:
+	return status;
+}
+
+static enum mod_hdcp_status read_ksv_list(struct mod_hdcp *hdcp,
+		struct mod_hdcp_event_context *event_ctx,
+		struct mod_hdcp_transition_input_hdcp1 *input)
+{
+	enum mod_hdcp_status status = MOD_HDCP_STATUS_SUCCESS;
+	uint8_t device_count;
+
+	if (event_ctx->event != MOD_HDCP_EVENT_CALLBACK) {
+		event_ctx->unexpected_event = 1;
+		goto out;
+	}
+
+	if (is_dp_hdcp(hdcp)) {
+		if (!mod_hdcp_execute_and_set(mod_hdcp_read_binfo,
+				&input->binfo_read_dp, &status,
+				hdcp, "binfo_read_dp"))
+			goto out;
+	} else {
+		if (!mod_hdcp_execute_and_set(mod_hdcp_read_bstatus,
+				&input->bstatus_read, &status,
+				hdcp, "bstatus_read"))
+			goto out;
+	}
+	if (!mod_hdcp_execute_and_set(check_no_max_cascade,
+			&input->max_cascade_check, &status,
+			hdcp, "max_cascade_check"))
+		goto out;
+	if (!mod_hdcp_execute_and_set(check_no_max_devs,
+			&input->max_devs_check, &status,
+			hdcp, "max_devs_check"))
+		goto out;
+	if (!mod_hdcp_execute_and_set(check_device_count,
+			&input->device_count_check, &status,
+			hdcp, "device_count_check"))
+		goto out;
+	device_count = get_device_count(hdcp);
+	hdcp->auth.msg.hdcp1.ksvlist_size = device_count*5;
+	if (!mod_hdcp_execute_and_set(mod_hdcp_read_ksvlist,
+			&input->ksvlist_read, &status,
+			hdcp, "ksvlist_read"))
+		goto out;
+	if (!mod_hdcp_execute_and_set(mod_hdcp_read_vp,
+			&input->vp_read, &status,
+			hdcp, "vp_read"))
+		goto out;
+	if (!mod_hdcp_execute_and_set(mod_hdcp_hdcp1_validate_ksvlist_vp,
+			&input->ksvlist_vp_validation, &status,
+			hdcp, "ksvlist_vp_validation"))
+		goto out;
+	if (input->encryption != PASS)
+		if (!mod_hdcp_execute_and_set(mod_hdcp_hdcp1_enable_encryption,
+				&input->encryption, &status,
+				hdcp, "encryption"))
+			goto out;
+	if (is_dp_mst_hdcp(hdcp))
+		if (!mod_hdcp_execute_and_set(
+				mod_hdcp_hdcp1_enable_dp_stream_encryption,
+				&input->stream_encryption_dp, &status,
+				hdcp, "stream_encryption_dp"))
+			goto out;
+out:
+	return status;
+}
+
+static enum mod_hdcp_status determine_rx_hdcp_capable_dp(struct mod_hdcp *hdcp,
+		struct mod_hdcp_event_context *event_ctx,
+		struct mod_hdcp_transition_input_hdcp1 *input)
+{
+	enum mod_hdcp_status status = MOD_HDCP_STATUS_SUCCESS;
+
+	if (event_ctx->event != MOD_HDCP_EVENT_CALLBACK) {
+		event_ctx->unexpected_event = 1;
+		goto out;
+	}
+
+	if (!mod_hdcp_execute_and_set(mod_hdcp_read_bcaps,
+			&input->bcaps_read, &status,
+			hdcp, "bcaps_read"))
+		goto out;
+	if (!mod_hdcp_execute_and_set(check_hdcp_capable_dp,
+			&input->hdcp_capable_dp, &status,
+			hdcp, "hdcp_capable_dp"))
+		goto out;
+out:
+	return status;
+}
+
+static enum mod_hdcp_status wait_for_r0_prime_dp(struct mod_hdcp *hdcp,
+		struct mod_hdcp_event_context *event_ctx,
+		struct mod_hdcp_transition_input_hdcp1 *input)
+{
+	enum mod_hdcp_status status = MOD_HDCP_STATUS_SUCCESS;
+
+	if (event_ctx->event != MOD_HDCP_EVENT_CPIRQ &&
+			event_ctx->event != MOD_HDCP_EVENT_WATCHDOG_TIMEOUT) {
+		event_ctx->unexpected_event = 1;
+		goto out;
+	}
+
+	if (!mod_hdcp_execute_and_set(mod_hdcp_read_bstatus,
+			&input->bstatus_read, &status,
+			hdcp, "bstatus_read"))
+		goto out;
+	if (!mod_hdcp_execute_and_set(check_r0p_available_dp,
+			&input->r0p_available_dp, &status,
+			hdcp, "r0p_available_dp"))
+		goto out;
+out:
+	return status;
+}
+
+static enum mod_hdcp_status authenticated_dp(struct mod_hdcp *hdcp,
+		struct mod_hdcp_event_context *event_ctx,
+		struct mod_hdcp_transition_input_hdcp1 *input)
+{
+	enum mod_hdcp_status status = MOD_HDCP_STATUS_SUCCESS;
+
+	if (event_ctx->event != MOD_HDCP_EVENT_CPIRQ) {
+		event_ctx->unexpected_event = 1;
+		goto out;
+	}
+
+	if (!mod_hdcp_execute_and_set(mod_hdcp_read_bstatus,
+			&input->bstatus_read, &status,
+			hdcp, "bstatus_read"))
+		goto out;
+	if (!mod_hdcp_execute_and_set(check_link_integrity_dp,
+			&input->link_integiry_check, &status,
+			hdcp, "link_integiry_check"))
+		goto out;
+	if (!mod_hdcp_execute_and_set(check_no_reauthentication_request_dp,
+			&input->reauth_request_check, &status,
+			hdcp, "reauth_request_check"))
+		goto out;
+out:
+	return status;
+}
+
+uint8_t mod_hdcp_execute_and_set(
+		mod_hdcp_action func, uint8_t *flag,
+		enum mod_hdcp_status *status, struct mod_hdcp *hdcp, char *str)
+{
+	*status = func(hdcp);
+	if (*status == MOD_HDCP_STATUS_SUCCESS && *flag != PASS) {
+		HDCP_INPUT_PASS_TRACE(hdcp, str);
+		*flag = PASS;
+	} else if (*status != MOD_HDCP_STATUS_SUCCESS && *flag != FAIL) {
+		HDCP_INPUT_FAIL_TRACE(hdcp, str);
+		*flag = FAIL;
+	}
+	return (*status == MOD_HDCP_STATUS_SUCCESS);
+}
+
+enum mod_hdcp_status mod_hdcp_hdcp1_execution(struct mod_hdcp *hdcp,
+		struct mod_hdcp_event_context *event_ctx,
+		struct mod_hdcp_transition_input_hdcp1 *input)
+{
+	enum mod_hdcp_status status = MOD_HDCP_STATUS_SUCCESS;
+
+	switch (current_state(hdcp)) {
+	case H1_A0_WAIT_FOR_ACTIVE_RX:
+		status = wait_for_active_rx(hdcp, event_ctx, input);
+		break;
+	case H1_A1_EXCHANGE_KSVS:
+		status = exchange_ksvs(hdcp, event_ctx, input);
+		break;
+	case H1_A2_COMPUTATIONS_A3_VALIDATE_RX_A6_TEST_FOR_REPEATER:
+		status = computations_validate_rx_test_for_repeater(hdcp,
+				event_ctx, input);
+		break;
+	case H1_A45_AUTHENTICATED:
+		status = authenticated(hdcp, event_ctx, input);
+		break;
+	case H1_A8_WAIT_FOR_READY:
+		status = wait_for_ready(hdcp, event_ctx, input);
+		break;
+	case H1_A9_READ_KSV_LIST:
+		status = read_ksv_list(hdcp, event_ctx, input);
+		break;
+	default:
+		status = MOD_HDCP_STATUS_INVALID_STATE;
+		break;
+	}
+
+	return status;
+}
+
+extern enum mod_hdcp_status mod_hdcp_hdcp1_dp_execution(struct mod_hdcp *hdcp,
+		struct mod_hdcp_event_context *event_ctx,
+		struct mod_hdcp_transition_input_hdcp1 *input)
+{
+	enum mod_hdcp_status status = MOD_HDCP_STATUS_SUCCESS;
+
+	switch (current_state(hdcp)) {
+	case D1_A0_DETERMINE_RX_HDCP_CAPABLE:
+		status = determine_rx_hdcp_capable_dp(hdcp, event_ctx, input);
+		break;
+	case D1_A1_EXCHANGE_KSVS:
+		status = exchange_ksvs(hdcp, event_ctx, input);
+		break;
+	case D1_A23_WAIT_FOR_R0_PRIME:
+		status = wait_for_r0_prime_dp(hdcp, event_ctx, input);
+		break;
+	case D1_A2_COMPUTATIONS_A3_VALIDATE_RX_A5_TEST_FOR_REPEATER:
+		status = computations_validate_rx_test_for_repeater(
+				hdcp, event_ctx, input);
+		break;
+	case D1_A4_AUTHENTICATED:
+		status = authenticated_dp(hdcp, event_ctx, input);
+		break;
+	case D1_A6_WAIT_FOR_READY:
+		status = wait_for_ready(hdcp, event_ctx, input);
+		break;
+	case D1_A7_READ_KSV_LIST:
+		status = read_ksv_list(hdcp, event_ctx, input);
+		break;
+	default:
+		status = MOD_HDCP_STATUS_INVALID_STATE;
+		break;
+	}
+
+	return status;
+}
diff --git a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp1_transition.c b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp1_transition.c
new file mode 100644
index 000000000000..136b8011ff3f
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp1_transition.c
@@ -0,0 +1,307 @@
+/*
+ * Copyright 2019 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#include "hdcp.h"
+
+enum mod_hdcp_status mod_hdcp_hdcp1_transition(struct mod_hdcp *hdcp,
+		struct mod_hdcp_event_context *event_ctx,
+		struct mod_hdcp_transition_input_hdcp1 *input,
+		struct mod_hdcp_output *output)
+{
+	enum mod_hdcp_status status = MOD_HDCP_STATUS_SUCCESS;
+	struct mod_hdcp_connection *conn = &hdcp->connection;
+	struct mod_hdcp_link_adjustment *adjust = &hdcp->connection.link.adjust;
+
+	switch (current_state(hdcp)) {
+	case H1_A0_WAIT_FOR_ACTIVE_RX:
+		if (input->bksv_read != PASS || input->bcaps_read != PASS) {
+			/* 1A-04: repeatedly attempts on port access failure */
+			callback_in_ms(500, output);
+			increment_stay_counter(hdcp);
+			break;
+		}
+		callback_in_ms(0, output);
+		set_state_id(hdcp, output, H1_A1_EXCHANGE_KSVS);
+		break;
+	case H1_A1_EXCHANGE_KSVS:
+		if (input->add_topology != PASS ||
+				input->create_session != PASS) {
+			/* out of sync with psp state */
+			adjust->hdcp1.disable = 1;
+			fail_and_restart_in_ms(0, &status, output);
+			break;
+		} else if (input->an_write != PASS ||
+				input->aksv_write != PASS ||
+				input->bksv_read != PASS ||
+				input->bksv_validation != PASS ||
+				input->ainfo_write == FAIL) {
+			/* 1A-05: consider invalid bksv a failure */
+			fail_and_restart_in_ms(0, &status, output);
+			break;
+		}
+		callback_in_ms(300, output);
+		set_state_id(hdcp, output,
+			H1_A2_COMPUTATIONS_A3_VALIDATE_RX_A6_TEST_FOR_REPEATER);
+		break;
+	case H1_A2_COMPUTATIONS_A3_VALIDATE_RX_A6_TEST_FOR_REPEATER:
+		if (input->bcaps_read != PASS ||
+				input->r0p_read != PASS ||
+				input->rx_validation != PASS ||
+				(!conn->is_repeater && input->encryption != PASS)) {
+			/* 1A-06: consider invalid r0' a failure */
+			/* 1A-08: consider bksv listed in SRM a failure */
+			fail_and_restart_in_ms(0, &status, output);
+			break;
+		}
+		if (conn->is_repeater) {
+			callback_in_ms(0, output);
+			set_watchdog_in_ms(hdcp, 5000, output);
+			set_state_id(hdcp, output, H1_A8_WAIT_FOR_READY);
+		} else {
+			callback_in_ms(0, output);
+			set_state_id(hdcp, output, H1_A45_AUTHENTICATED);
+			HDCP_FULL_DDC_TRACE(hdcp);
+		}
+		break;
+	case H1_A45_AUTHENTICATED:
+		if (input->link_maintenance != PASS) {
+			/* 1A-07: consider invalid ri' a failure */
+			/* 1A-07a: consider read ri' not returned a failure */
+			fail_and_restart_in_ms(0, &status, output);
+			break;
+		}
+		callback_in_ms(500, output);
+		increment_stay_counter(hdcp);
+		break;
+	case H1_A8_WAIT_FOR_READY:
+		if (input->ready_check != PASS) {
+			if (event_ctx->event ==
+					MOD_HDCP_EVENT_WATCHDOG_TIMEOUT) {
+				/* 1B-03: fail hdcp on ksv list READY timeout */
+				/* prevent black screen in next attempt */
+				adjust->hdcp1.postpone_encryption = 1;
+				fail_and_restart_in_ms(0, &status, output);
+			} else {
+				/* continue ksv list READY polling*/
+				callback_in_ms(500, output);
+				increment_stay_counter(hdcp);
+			}
+			break;
+		}
+		callback_in_ms(0, output);
+		set_state_id(hdcp, output, H1_A9_READ_KSV_LIST);
+		break;
+	case H1_A9_READ_KSV_LIST:
+		if (input->bstatus_read != PASS ||
+				input->max_cascade_check != PASS ||
+				input->max_devs_check != PASS ||
+				input->device_count_check != PASS ||
+				input->ksvlist_read != PASS ||
+				input->vp_read != PASS ||
+				input->ksvlist_vp_validation != PASS ||
+				input->encryption != PASS) {
+			/* 1B-06: consider MAX_CASCADE_EXCEEDED a failure */
+			/* 1B-05: consider MAX_DEVS_EXCEEDED a failure */
+			/* 1B-04: consider invalid v' a failure */
+			fail_and_restart_in_ms(0, &status, output);
+			break;
+		}
+		callback_in_ms(0, output);
+		set_state_id(hdcp, output, H1_A45_AUTHENTICATED);
+		HDCP_FULL_DDC_TRACE(hdcp);
+		break;
+	default:
+		status = MOD_HDCP_STATUS_INVALID_STATE;
+		fail_and_restart_in_ms(0, &status, output);
+		break;
+	}
+
+	return status;
+}
+
+enum mod_hdcp_status mod_hdcp_hdcp1_dp_transition(struct mod_hdcp *hdcp,
+		struct mod_hdcp_event_context *event_ctx,
+		struct mod_hdcp_transition_input_hdcp1 *input,
+		struct mod_hdcp_output *output)
+{
+	enum mod_hdcp_status status = MOD_HDCP_STATUS_SUCCESS;
+	struct mod_hdcp_connection *conn = &hdcp->connection;
+	struct mod_hdcp_link_adjustment *adjust = &hdcp->connection.link.adjust;
+
+	switch (current_state(hdcp)) {
+	case D1_A0_DETERMINE_RX_HDCP_CAPABLE:
+		if (input->bcaps_read != PASS) {
+			/* 1A-04: no authentication on bcaps read failure */
+			fail_and_restart_in_ms(0, &status, output);
+			break;
+		} else if (input->hdcp_capable_dp != PASS) {
+			adjust->hdcp1.disable = 1;
+			fail_and_restart_in_ms(0, &status, output);
+			break;
+		}
+		callback_in_ms(0, output);
+		set_state_id(hdcp, output, D1_A1_EXCHANGE_KSVS);
+		break;
+	case D1_A1_EXCHANGE_KSVS:
+		if (input->add_topology != PASS ||
+				input->create_session != PASS) {
+			/* out of sync with psp state */
+			adjust->hdcp1.disable = 1;
+			fail_and_restart_in_ms(0, &status, output);
+			break;
+		} else if (input->an_write != PASS ||
+				input->aksv_write != PASS ||
+				input->bksv_read != PASS ||
+				input->bksv_validation != PASS ||
+				input->ainfo_write == FAIL) {
+			/* 1A-05: consider invalid bksv a failure */
+			fail_and_restart_in_ms(0, &status, output);
+			break;
+		}
+		set_watchdog_in_ms(hdcp, 100, output);
+		set_state_id(hdcp, output, D1_A23_WAIT_FOR_R0_PRIME);
+		break;
+	case D1_A23_WAIT_FOR_R0_PRIME:
+		if (input->bstatus_read != PASS) {
+			fail_and_restart_in_ms(0, &status, output);
+			break;
+		} else if (input->r0p_available_dp != PASS) {
+			if (event_ctx->event == MOD_HDCP_EVENT_WATCHDOG_TIMEOUT)
+				fail_and_restart_in_ms(0, &status, output);
+			else
+				increment_stay_counter(hdcp);
+			break;
+		}
+		callback_in_ms(0, output);
+		set_state_id(hdcp, output, D1_A2_COMPUTATIONS_A3_VALIDATE_RX_A5_TEST_FOR_REPEATER);
+		break;
+	case D1_A2_COMPUTATIONS_A3_VALIDATE_RX_A5_TEST_FOR_REPEATER:
+		if (input->r0p_read != PASS) {
+			fail_and_restart_in_ms(0, &status, output);
+			break;
+		} else if (input->rx_validation != PASS) {
+			if (hdcp->state.stay_count < 2) {
+				/* allow 2 additional retries */
+				callback_in_ms(0, output);
+				increment_stay_counter(hdcp);
+			} else {
+				/*
+				 * 1A-06: consider invalid r0' a failure
+				 * after 3 attempts.
+				 * 1A-08: consider bksv listed in SRM a failure
+				 */
+				fail_and_restart_in_ms(0, &status, output);
+			}
+			break;
+		} else if ((!conn->is_repeater && input->encryption != PASS) ||
+				(!conn->is_repeater && is_dp_mst_hdcp(hdcp) && input->stream_encryption_dp != PASS)) {
+			fail_and_restart_in_ms(0, &status, output);
+			break;
+		}
+		if (conn->is_repeater) {
+			set_watchdog_in_ms(hdcp, 5000, output);
+			set_state_id(hdcp, output, D1_A6_WAIT_FOR_READY);
+		} else {
+			set_state_id(hdcp, output, D1_A4_AUTHENTICATED);
+			HDCP_FULL_DDC_TRACE(hdcp);
+		}
+		break;
+	case D1_A4_AUTHENTICATED:
+		if (input->link_integiry_check != PASS ||
+				input->reauth_request_check != PASS) {
+			/* 1A-07: restart hdcp on a link integrity failure */
+			fail_and_restart_in_ms(0, &status, output);
+			break;
+		}
+		break;
+	case D1_A6_WAIT_FOR_READY:
+		if (input->link_integiry_check == FAIL ||
+				input->reauth_request_check == FAIL) {
+			fail_and_restart_in_ms(0, &status, output);
+			break;
+		} else if (input->ready_check != PASS) {
+			if (event_ctx->event ==
+					MOD_HDCP_EVENT_WATCHDOG_TIMEOUT) {
+				/* 1B-04: fail hdcp on ksv list READY timeout */
+				/* prevent black screen in next attempt */
+				adjust->hdcp1.postpone_encryption = 1;
+				fail_and_restart_in_ms(0, &status, output);
+			} else {
+				increment_stay_counter(hdcp);
+			}
+			break;
+		}
+		callback_in_ms(0, output);
+		set_state_id(hdcp, output, D1_A7_READ_KSV_LIST);
+		break;
+	case D1_A7_READ_KSV_LIST:
+		if (input->binfo_read_dp != PASS ||
+				input->max_cascade_check != PASS ||
+				input->max_devs_check != PASS) {
+			/* 1B-06: consider MAX_DEVS_EXCEEDED a failure */
+			/* 1B-07: consider MAX_CASCADE_EXCEEDED a failure */
+			fail_and_restart_in_ms(0, &status, output);
+			break;
+		} else if (input->device_count_check != PASS) {
+			/*
+			 * some slow dongle doesn't update
+			 * device count as soon as downstream is connected.
+			 * give it more time to react.
+			 */
+			adjust->hdcp1.postpone_encryption = 1;
+			fail_and_restart_in_ms(1000, &status, output);
+			break;
+		} else if (input->ksvlist_read != PASS ||
+				input->vp_read != PASS) {
+			fail_and_restart_in_ms(0, &status, output);
+			break;
+		} else if (input->ksvlist_vp_validation != PASS) {
+			if (hdcp->state.stay_count < 2) {
+				/* allow 2 additional retries */
+				callback_in_ms(0, output);
+				increment_stay_counter(hdcp);
+			} else {
+				/*
+				 * 1B-05: consider invalid v' a failure
+				 * after 3 attempts.
+				 */
+				fail_and_restart_in_ms(0, &status, output);
+			}
+			break;
+		} else if (input->encryption != PASS ||
+				(is_dp_mst_hdcp(hdcp) && input->stream_encryption_dp != PASS)) {
+			fail_and_restart_in_ms(0, &status, output);
+			break;
+		}
+		set_state_id(hdcp, output, D1_A4_AUTHENTICATED);
+		HDCP_FULL_DDC_TRACE(hdcp);
+		break;
+	default:
+		fail_and_restart_in_ms(0, &status, output);
+		break;
+	}
+
+	return status;
+}
diff --git a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_ddc.c b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_ddc.c
new file mode 100644
index 000000000000..e7baae059b85
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_ddc.c
@@ -0,0 +1,305 @@
+/*
+ * Copyright 2019 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#include "hdcp.h"
+
+#define MIN(a, b) ((a) < (b) ? (a) : (b))
+#define HDCP_I2C_ADDR 0x3a	/* 0x74 >> 1*/
+#define KSV_READ_SIZE 0xf	/* 0x6803b - 0x6802c */
+#define HDCP_MAX_AUX_TRANSACTION_SIZE 16
+
+enum mod_hdcp_ddc_message_id {
+	MOD_HDCP_MESSAGE_ID_INVALID = -1,
+
+	/* HDCP 1.4 */
+
+	MOD_HDCP_MESSAGE_ID_READ_BKSV = 0,
+	MOD_HDCP_MESSAGE_ID_READ_RI_R0,
+	MOD_HDCP_MESSAGE_ID_WRITE_AKSV,
+	MOD_HDCP_MESSAGE_ID_WRITE_AINFO,
+	MOD_HDCP_MESSAGE_ID_WRITE_AN,
+	MOD_HDCP_MESSAGE_ID_READ_VH_X,
+	MOD_HDCP_MESSAGE_ID_READ_VH_0,
+	MOD_HDCP_MESSAGE_ID_READ_VH_1,
+	MOD_HDCP_MESSAGE_ID_READ_VH_2,
+	MOD_HDCP_MESSAGE_ID_READ_VH_3,
+	MOD_HDCP_MESSAGE_ID_READ_VH_4,
+	MOD_HDCP_MESSAGE_ID_READ_BCAPS,
+	MOD_HDCP_MESSAGE_ID_READ_BSTATUS,
+	MOD_HDCP_MESSAGE_ID_READ_KSV_FIFO,
+	MOD_HDCP_MESSAGE_ID_READ_BINFO,
+
+	MOD_HDCP_MESSAGE_ID_MAX
+};
+
+static const uint8_t hdcp_i2c_offsets[] = {
+	[MOD_HDCP_MESSAGE_ID_READ_BKSV] = 0x0,
+	[MOD_HDCP_MESSAGE_ID_READ_RI_R0] = 0x8,
+	[MOD_HDCP_MESSAGE_ID_WRITE_AKSV] = 0x10,
+	[MOD_HDCP_MESSAGE_ID_WRITE_AINFO] = 0x15,
+	[MOD_HDCP_MESSAGE_ID_WRITE_AN] = 0x18,
+	[MOD_HDCP_MESSAGE_ID_READ_VH_X] = 0x20,
+	[MOD_HDCP_MESSAGE_ID_READ_VH_0] = 0x20,
+	[MOD_HDCP_MESSAGE_ID_READ_VH_1] = 0x24,
+	[MOD_HDCP_MESSAGE_ID_READ_VH_2] = 0x28,
+	[MOD_HDCP_MESSAGE_ID_READ_VH_3] = 0x2C,
+	[MOD_HDCP_MESSAGE_ID_READ_VH_4] = 0x30,
+	[MOD_HDCP_MESSAGE_ID_READ_BCAPS] = 0x40,
+	[MOD_HDCP_MESSAGE_ID_READ_BSTATUS] = 0x41,
+	[MOD_HDCP_MESSAGE_ID_READ_KSV_FIFO] = 0x43,
+	[MOD_HDCP_MESSAGE_ID_READ_BINFO] = 0xFF,
+};
+
+static const uint32_t hdcp_dpcd_addrs[] = {
+	[MOD_HDCP_MESSAGE_ID_READ_BKSV] = 0x68000,
+	[MOD_HDCP_MESSAGE_ID_READ_RI_R0] = 0x68005,
+	[MOD_HDCP_MESSAGE_ID_WRITE_AKSV] = 0x68007,
+	[MOD_HDCP_MESSAGE_ID_WRITE_AINFO] = 0x6803B,
+	[MOD_HDCP_MESSAGE_ID_WRITE_AN] = 0x6800c,
+	[MOD_HDCP_MESSAGE_ID_READ_VH_X] = 0x68014,
+	[MOD_HDCP_MESSAGE_ID_READ_VH_0] = 0x68014,
+	[MOD_HDCP_MESSAGE_ID_READ_VH_1] = 0x68018,
+	[MOD_HDCP_MESSAGE_ID_READ_VH_2] = 0x6801c,
+	[MOD_HDCP_MESSAGE_ID_READ_VH_3] = 0x68020,
+	[MOD_HDCP_MESSAGE_ID_READ_VH_4] = 0x68024,
+	[MOD_HDCP_MESSAGE_ID_READ_BCAPS] = 0x68028,
+	[MOD_HDCP_MESSAGE_ID_READ_BSTATUS] = 0x68029,
+	[MOD_HDCP_MESSAGE_ID_READ_KSV_FIFO] = 0x6802c,
+	[MOD_HDCP_MESSAGE_ID_READ_BINFO] = 0x6802a,
+};
+
+static enum mod_hdcp_status read(struct mod_hdcp *hdcp,
+		enum mod_hdcp_ddc_message_id msg_id,
+		uint8_t *buf,
+		uint32_t buf_len)
+{
+	bool success = true;
+	uint32_t cur_size = 0;
+	uint32_t data_offset = 0;
+
+	if (is_dp_hdcp(hdcp)) {
+		while (buf_len > 0) {
+			cur_size = MIN(buf_len, HDCP_MAX_AUX_TRANSACTION_SIZE);
+			success = hdcp->config.ddc.funcs.read_dpcd(hdcp->config.ddc.handle,
+					hdcp_dpcd_addrs[msg_id] + data_offset,
+					buf + data_offset,
+					cur_size);
+
+			if (!success)
+				break;
+
+			buf_len -= cur_size;
+			data_offset += cur_size;
+		}
+	} else {
+		success = hdcp->config.ddc.funcs.read_i2c(
+				hdcp->config.ddc.handle,
+				HDCP_I2C_ADDR,
+				hdcp_i2c_offsets[msg_id],
+				buf,
+				(uint32_t)buf_len);
+	}
+
+	return success ? MOD_HDCP_STATUS_SUCCESS : MOD_HDCP_STATUS_DDC_FAILURE;
+}
+
+static enum mod_hdcp_status read_repeatedly(struct mod_hdcp *hdcp,
+		enum mod_hdcp_ddc_message_id msg_id,
+		uint8_t *buf,
+		uint32_t buf_len,
+		uint8_t read_size)
+{
+	enum mod_hdcp_status status = MOD_HDCP_STATUS_DDC_FAILURE;
+	uint32_t cur_size = 0;
+	uint32_t data_offset = 0;
+
+	while (buf_len > 0) {
+		cur_size = MIN(buf_len, read_size);
+		status = read(hdcp, msg_id, buf + data_offset, cur_size);
+
+		if (status != MOD_HDCP_STATUS_SUCCESS)
+			break;
+
+		buf_len -= cur_size;
+		data_offset += cur_size;
+	}
+
+	return status;
+}
+
+static enum mod_hdcp_status write(struct mod_hdcp *hdcp,
+		enum mod_hdcp_ddc_message_id msg_id,
+		uint8_t *buf,
+		uint32_t buf_len)
+{
+	bool success = true;
+	uint32_t cur_size = 0;
+	uint32_t data_offset = 0;
+
+	if (is_dp_hdcp(hdcp)) {
+		while (buf_len > 0) {
+			cur_size = MIN(buf_len, HDCP_MAX_AUX_TRANSACTION_SIZE);
+			success = hdcp->config.ddc.funcs.write_dpcd(
+					hdcp->config.ddc.handle,
+					hdcp_dpcd_addrs[msg_id] + data_offset,
+					buf + data_offset,
+					cur_size);
+
+			if (!success)
+				break;
+
+			buf_len -= cur_size;
+			data_offset += cur_size;
+		}
+	} else {
+		hdcp->buf[0] = hdcp_i2c_offsets[msg_id];
+		memmove(&hdcp->buf[1], buf, buf_len);
+		success = hdcp->config.ddc.funcs.write_i2c(
+				hdcp->config.ddc.handle,
+				HDCP_I2C_ADDR,
+				hdcp->buf,
+				(uint32_t)(buf_len+1));
+	}
+
+	return success ? MOD_HDCP_STATUS_SUCCESS : MOD_HDCP_STATUS_DDC_FAILURE;
+}
+
+enum mod_hdcp_status mod_hdcp_read_bksv(struct mod_hdcp *hdcp)
+{
+	return read(hdcp, MOD_HDCP_MESSAGE_ID_READ_BKSV,
+			hdcp->auth.msg.hdcp1.bksv,
+			sizeof(hdcp->auth.msg.hdcp1.bksv));
+}
+
+enum mod_hdcp_status mod_hdcp_read_bcaps(struct mod_hdcp *hdcp)
+{
+	return read(hdcp, MOD_HDCP_MESSAGE_ID_READ_BCAPS,
+			&hdcp->auth.msg.hdcp1.bcaps,
+			sizeof(hdcp->auth.msg.hdcp1.bcaps));
+}
+
+enum mod_hdcp_status mod_hdcp_read_bstatus(struct mod_hdcp *hdcp)
+{
+	enum mod_hdcp_status status;
+
+	if (is_dp_hdcp(hdcp))
+		status = read(hdcp, MOD_HDCP_MESSAGE_ID_READ_BSTATUS,
+					(uint8_t *)&hdcp->auth.msg.hdcp1.bstatus,
+					1);
+	else
+		status = read(hdcp, MOD_HDCP_MESSAGE_ID_READ_BSTATUS,
+				(uint8_t *)&hdcp->auth.msg.hdcp1.bstatus,
+				sizeof(hdcp->auth.msg.hdcp1.bstatus));
+	return status;
+}
+
+enum mod_hdcp_status mod_hdcp_read_r0p(struct mod_hdcp *hdcp)
+{
+	return read(hdcp, MOD_HDCP_MESSAGE_ID_READ_RI_R0,
+			(uint8_t *)&hdcp->auth.msg.hdcp1.r0p,
+			sizeof(hdcp->auth.msg.hdcp1.r0p));
+}
+
+/* special case, reading repeatedly at the same address, don't use read() */
+enum mod_hdcp_status mod_hdcp_read_ksvlist(struct mod_hdcp *hdcp)
+{
+	enum mod_hdcp_status status;
+
+	if (is_dp_hdcp(hdcp))
+		status = read_repeatedly(hdcp, MOD_HDCP_MESSAGE_ID_READ_KSV_FIFO,
+				hdcp->auth.msg.hdcp1.ksvlist,
+				hdcp->auth.msg.hdcp1.ksvlist_size,
+				KSV_READ_SIZE);
+	else
+		status = read(hdcp, MOD_HDCP_MESSAGE_ID_READ_KSV_FIFO,
+				(uint8_t *)&hdcp->auth.msg.hdcp1.ksvlist,
+				hdcp->auth.msg.hdcp1.ksvlist_size);
+	return status;
+}
+
+enum mod_hdcp_status mod_hdcp_read_vp(struct mod_hdcp *hdcp)
+{
+	enum mod_hdcp_status status;
+
+	status = read(hdcp, MOD_HDCP_MESSAGE_ID_READ_VH_0,
+			&hdcp->auth.msg.hdcp1.vp[0], 4);
+	if (status != MOD_HDCP_STATUS_SUCCESS)
+		goto out;
+
+	status = read(hdcp, MOD_HDCP_MESSAGE_ID_READ_VH_1,
+			&hdcp->auth.msg.hdcp1.vp[4], 4);
+	if (status != MOD_HDCP_STATUS_SUCCESS)
+		goto out;
+
+	status = read(hdcp, MOD_HDCP_MESSAGE_ID_READ_VH_2,
+			&hdcp->auth.msg.hdcp1.vp[8], 4);
+	if (status != MOD_HDCP_STATUS_SUCCESS)
+		goto out;
+
+	status = read(hdcp, MOD_HDCP_MESSAGE_ID_READ_VH_3,
+			&hdcp->auth.msg.hdcp1.vp[12], 4);
+	if (status != MOD_HDCP_STATUS_SUCCESS)
+		goto out;
+
+	status = read(hdcp, MOD_HDCP_MESSAGE_ID_READ_VH_4,
+			&hdcp->auth.msg.hdcp1.vp[16], 4);
+out:
+	return status;
+}
+
+enum mod_hdcp_status mod_hdcp_read_binfo(struct mod_hdcp *hdcp)
+{
+	enum mod_hdcp_status status;
+
+	if (is_dp_hdcp(hdcp))
+		status = read(hdcp, MOD_HDCP_MESSAGE_ID_READ_BINFO,
+				(uint8_t *)&hdcp->auth.msg.hdcp1.binfo_dp,
+				sizeof(hdcp->auth.msg.hdcp1.binfo_dp));
+	else
+		status = MOD_HDCP_STATUS_INVALID_OPERATION;
+
+	return status;
+}
+
+enum mod_hdcp_status mod_hdcp_write_aksv(struct mod_hdcp *hdcp)
+{
+	return write(hdcp, MOD_HDCP_MESSAGE_ID_WRITE_AKSV,
+			hdcp->auth.msg.hdcp1.aksv,
+			sizeof(hdcp->auth.msg.hdcp1.aksv));
+}
+
+enum mod_hdcp_status mod_hdcp_write_ainfo(struct mod_hdcp *hdcp)
+{
+	return write(hdcp, MOD_HDCP_MESSAGE_ID_WRITE_AINFO,
+			&hdcp->auth.msg.hdcp1.ainfo,
+			sizeof(hdcp->auth.msg.hdcp1.ainfo));
+}
+
+enum mod_hdcp_status mod_hdcp_write_an(struct mod_hdcp *hdcp)
+{
+	return write(hdcp, MOD_HDCP_MESSAGE_ID_WRITE_AN,
+			hdcp->auth.msg.hdcp1.an,
+			sizeof(hdcp->auth.msg.hdcp1.an));
+}
diff --git a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_log.c b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_log.c
new file mode 100644
index 000000000000..3982ced5f969
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_log.c
@@ -0,0 +1,163 @@
+/*
+ * Copyright 2019 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+
+#include "hdcp.h"
+
+void mod_hdcp_dump_binary_message(uint8_t *msg, uint32_t msg_size,
+		uint8_t *buf, uint32_t buf_size)
+{
+	const uint8_t bytes_per_line = 16,
+			byte_size = 3,
+			newline_size = 1,
+			terminator_size = 1;
+	uint32_t line_count = msg_size / bytes_per_line,
+			trailing_bytes = msg_size % bytes_per_line;
+	uint32_t target_size = (byte_size * bytes_per_line + newline_size) * line_count +
+			byte_size * trailing_bytes + newline_size + terminator_size;
+	uint32_t buf_pos = 0;
+	uint32_t i = 0;
+
+	if (buf_size >= target_size) {
+		for (i = 0; i < msg_size; i++) {
+			if (i % bytes_per_line == 0)
+				buf[buf_pos++] = '\n';
+			sprintf(&buf[buf_pos], "%02X ", msg[i]);
+			buf_pos += byte_size;
+		}
+		buf[buf_pos++] = '\0';
+	}
+}
+
+char *mod_hdcp_status_to_str(int32_t status)
+{
+	switch (status) {
+	case MOD_HDCP_STATUS_SUCCESS:
+		return "MOD_HDCP_STATUS_SUCCESS";
+	case MOD_HDCP_STATUS_FAILURE:
+		return "MOD_HDCP_STATUS_FAILURE";
+	case MOD_HDCP_STATUS_RESET_NEEDED:
+		return "MOD_HDCP_STATUS_RESET_NEEDED";
+	case MOD_HDCP_STATUS_DISPLAY_OUT_OF_BOUND:
+		return "MOD_HDCP_STATUS_DISPLAY_OUT_OF_BOUND";
+	case MOD_HDCP_STATUS_DISPLAY_NOT_FOUND:
+		return "MOD_HDCP_STATUS_DISPLAY_NOT_FOUND";
+	case MOD_HDCP_STATUS_INVALID_STATE:
+		return "MOD_HDCP_STATUS_INVALID_STATE";
+	case MOD_HDCP_STATUS_NOT_IMPLEMENTED:
+		return "MOD_HDCP_STATUS_NOT_IMPLEMENTED";
+	case MOD_HDCP_STATUS_INTERNAL_POLICY_FAILURE:
+		return "MOD_HDCP_STATUS_INTERNAL_POLICY_FAILURE";
+	case MOD_HDCP_STATUS_UPDATE_TOPOLOGY_FAILURE:
+		return "MOD_HDCP_STATUS_UPDATE_TOPOLOGY_FAILURE";
+	case MOD_HDCP_STATUS_CREATE_PSP_SERVICE_FAILURE:
+		return "MOD_HDCP_STATUS_CREATE_PSP_SERVICE_FAILURE";
+	case MOD_HDCP_STATUS_DESTROY_PSP_SERVICE_FAILURE:
+		return "MOD_HDCP_STATUS_DESTROY_PSP_SERVICE_FAILURE";
+	case MOD_HDCP_STATUS_HDCP1_CREATE_SESSION_FAILURE:
+		return "MOD_HDCP_STATUS_HDCP1_CREATE_SESSION_FAILURE";
+	case MOD_HDCP_STATUS_HDCP1_DESTROY_SESSION_FAILURE:
+		return "MOD_HDCP_STATUS_HDCP1_DESTROY_SESSION_FAILURE";
+	case MOD_HDCP_STATUS_HDCP1_VALIDATE_ENCRYPTION_FAILURE:
+		return "MOD_HDCP_STATUS_HDCP1_VALIDATE_ENCRYPTION_FAILURE";
+	case MOD_HDCP_STATUS_HDCP1_NOT_HDCP_REPEATER:
+		return "MOD_HDCP_STATUS_HDCP1_NOT_HDCP_REPEATER";
+	case MOD_HDCP_STATUS_HDCP1_NOT_CAPABLE:
+		return "MOD_HDCP_STATUS_HDCP1_NOT_CAPABLE";
+	case MOD_HDCP_STATUS_HDCP1_R0_PRIME_PENDING:
+		return "MOD_HDCP_STATUS_HDCP1_R0_PRIME_PENDING";
+	case MOD_HDCP_STATUS_HDCP1_VALIDATE_RX_FAILURE:
+		return "MOD_HDCP_STATUS_HDCP1_VALIDATE_RX_FAILURE";
+	case MOD_HDCP_STATUS_HDCP1_KSV_LIST_NOT_READY:
+		return "MOD_HDCP_STATUS_HDCP1_KSV_LIST_NOT_READY";
+	case MOD_HDCP_STATUS_HDCP1_VALIDATE_KSV_LIST_FAILURE:
+		return "MOD_HDCP_STATUS_HDCP1_VALIDATE_KSV_LIST_FAILURE";
+	case MOD_HDCP_STATUS_HDCP1_ENABLE_ENCRYPTION:
+		return "MOD_HDCP_STATUS_HDCP1_ENABLE_ENCRYPTION";
+	case MOD_HDCP_STATUS_HDCP1_ENABLE_STREAM_ENCRYPTION_FAILURE:
+		return "MOD_HDCP_STATUS_HDCP1_ENABLE_STREAM_ENCRYPTION_FAILURE";
+	case MOD_HDCP_STATUS_HDCP1_MAX_CASCADE_EXCEEDED_FAILURE:
+		return "MOD_HDCP_STATUS_HDCP1_MAX_CASCADE_EXCEEDED_FAILURE";
+	case MOD_HDCP_STATUS_HDCP1_MAX_DEVS_EXCEEDED_FAILURE:
+		return "MOD_HDCP_STATUS_HDCP1_MAX_DEVS_EXCEEDED_FAILURE";
+	case MOD_HDCP_STATUS_HDCP1_DEVICE_COUNT_MISMATCH_FAILURE:
+		return "MOD_HDCP_STATUS_HDCP1_DEVICE_COUNT_MISMATCH_FAILURE";
+	case MOD_HDCP_STATUS_HDCP1_LINK_INTEGRITY_FAILURE:
+		return "MOD_HDCP_STATUS_HDCP1_LINK_INTEGRITY_FAILURE";
+	case MOD_HDCP_STATUS_HDCP1_REAUTH_REQUEST_ISSUED:
+		return "MOD_HDCP_STATUS_HDCP1_REAUTH_REQUEST_ISSUED";
+	case MOD_HDCP_STATUS_HDCP1_LINK_MAINTENANCE_FAILURE:
+		return "MOD_HDCP_STATUS_HDCP1_LINK_MAINTENANCE_FAILURE";
+	case MOD_HDCP_STATUS_HDCP1_INVALID_BKSV:
+		return "MOD_HDCP_STATUS_HDCP1_INVALID_BKSV";
+	case MOD_HDCP_STATUS_DDC_FAILURE:
+		return "MOD_HDCP_STATUS_DDC_FAILURE";
+	case MOD_HDCP_STATUS_INVALID_OPERATION:
+		return "MOD_HDCP_STATUS_INVALID_OPERATION";
+	default:
+		return "MOD_HDCP_STATUS_UNKNOWN";
+	}
+}
+
+char *mod_hdcp_state_id_to_str(int32_t id)
+{
+	switch (id) {
+	case HDCP_UNINITIALIZED:
+		return "HDCP_UNINITIALIZED";
+	case HDCP_INITIALIZED:
+		return "HDCP_INITIALIZED";
+	case HDCP_CP_NOT_DESIRED:
+		return "HDCP_CP_NOT_DESIRED";
+	case H1_A0_WAIT_FOR_ACTIVE_RX:
+		return "H1_A0_WAIT_FOR_ACTIVE_RX";
+	case H1_A1_EXCHANGE_KSVS:
+		return "H1_A1_EXCHANGE_KSVS";
+	case H1_A2_COMPUTATIONS_A3_VALIDATE_RX_A6_TEST_FOR_REPEATER:
+		return "H1_A2_COMPUTATIONS_A3_VALIDATE_RX_A6_TEST_FOR_REPEATER";
+	case H1_A45_AUTHENTICATED:
+		return "H1_A45_AUTHENTICATED";
+	case H1_A8_WAIT_FOR_READY:
+		return "H1_A8_WAIT_FOR_READY";
+	case H1_A9_READ_KSV_LIST:
+		return "H1_A9_READ_KSV_LIST";
+	case D1_A0_DETERMINE_RX_HDCP_CAPABLE:
+		return "D1_A0_DETERMINE_RX_HDCP_CAPABLE";
+	case D1_A1_EXCHANGE_KSVS:
+		return "D1_A1_EXCHANGE_KSVS";
+	case D1_A23_WAIT_FOR_R0_PRIME:
+		return "D1_A23_WAIT_FOR_R0_PRIME";
+	case D1_A2_COMPUTATIONS_A3_VALIDATE_RX_A5_TEST_FOR_REPEATER:
+		return "D1_A2_COMPUTATIONS_A3_VALIDATE_RX_A5_TEST_FOR_REPEATER";
+	case D1_A4_AUTHENTICATED:
+		return "D1_A4_AUTHENTICATED";
+	case D1_A6_WAIT_FOR_READY:
+		return "D1_A6_WAIT_FOR_READY";
+	case D1_A7_READ_KSV_LIST:
+		return "D1_A7_READ_KSV_LIST";
+	default:
+		return "UNKNOWN_STATE_ID";
+	};
+}
+
diff --git a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_log.h b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_log.h
new file mode 100644
index 000000000000..2fd0e0a893ef
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_log.h
@@ -0,0 +1,139 @@
+/*
+ * Copyright 2019 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#ifndef MOD_HDCP_LOG_H_
+#define MOD_HDCP_LOG_H_
+
+#ifdef CONFIG_DRM_AMD_DC_HDCP
+#define HDCP_LOG_ERR(hdcp, ...) DRM_ERROR(__VA_ARGS__)
+#define HDCP_LOG_VER(hdcp, ...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define HDCP_LOG_FSM(hdcp, ...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define HDCP_LOG_TOP(hdcp, ...) pr_debug("[HDCP_TOP]:"__VA_ARGS__)
+#define HDCP_LOG_DDC(hdcp, ...) pr_debug("[HDCP_DDC]:"__VA_ARGS__)
+#endif
+
+/* default logs */
+#define HDCP_ERROR_TRACE(hdcp, status) \
+		HDCP_LOG_ERR(hdcp, \
+			"[Link %d] ERROR %s IN STATE %s", \
+			hdcp->config.index, \
+			mod_hdcp_status_to_str(status), \
+			mod_hdcp_state_id_to_str(hdcp->state.id))
+#define HDCP_HDCP1_ENABLED_TRACE(hdcp, displayIndex) \
+		HDCP_LOG_VER(hdcp, \
+			"[Link %d] HDCP 1.4 enabled on display %d", \
+			hdcp->config.index, displayIndex)
+/* state machine logs */
+#define HDCP_REMOVE_DISPLAY_TRACE(hdcp, displayIndex) \
+		HDCP_LOG_FSM(hdcp, \
+			"[Link %d] HDCP_REMOVE_DISPLAY index %d", \
+			hdcp->config.index, displayIndex)
+#define HDCP_INPUT_PASS_TRACE(hdcp, str) \
+		HDCP_LOG_FSM(hdcp, \
+			"[Link %d]\tPASS %s", \
+			hdcp->config.index, str)
+#define HDCP_INPUT_FAIL_TRACE(hdcp, str) \
+		HDCP_LOG_FSM(hdcp, \
+			"[Link %d]\tFAIL %s", \
+			hdcp->config.index, str)
+#define HDCP_NEXT_STATE_TRACE(hdcp, id, output) do { \
+		if (output->watchdog_timer_needed) \
+			HDCP_LOG_FSM(hdcp, \
+				"[Link %d] > %s with %d ms watchdog", \
+				hdcp->config.index, \
+				mod_hdcp_state_id_to_str(id), output->watchdog_timer_delay); \
+		else \
+			HDCP_LOG_FSM(hdcp, \
+				"[Link %d] > %s", hdcp->config.index, \
+				mod_hdcp_state_id_to_str(id)); \
+} while (0)
+#define HDCP_TIMEOUT_TRACE(hdcp) \
+		HDCP_LOG_FSM(hdcp, "[Link %d] --> TIMEOUT", hdcp->config.index)
+#define HDCP_CPIRQ_TRACE(hdcp) \
+		HDCP_LOG_FSM(hdcp, "[Link %d] --> CPIRQ", hdcp->config.index)
+#define HDCP_EVENT_TRACE(hdcp, event) \
+		if (event == MOD_HDCP_EVENT_WATCHDOG_TIMEOUT) \
+			HDCP_TIMEOUT_TRACE(hdcp); \
+		else if (event == MOD_HDCP_EVENT_CPIRQ) \
+			HDCP_CPIRQ_TRACE(hdcp)
+/* TODO: find some way to tell if logging is off to save time */
+#define HDCP_DDC_READ_TRACE(hdcp, msg_name, msg, msg_size) do { \
+		mod_hdcp_dump_binary_message(msg, msg_size, hdcp->buf, \
+				sizeof(hdcp->buf)); \
+		HDCP_LOG_DDC(hdcp, "[Link %d] Read %s%s", hdcp->config.index, \
+				msg_name, hdcp->buf); \
+} while (0)
+#define HDCP_DDC_WRITE_TRACE(hdcp, msg_name, msg, msg_size) do { \
+		mod_hdcp_dump_binary_message(msg, msg_size, hdcp->buf, \
+				sizeof(hdcp->buf)); \
+		HDCP_LOG_DDC(hdcp, "[Link %d] Write %s%s", \
+				hdcp->config.index, msg_name,\
+				hdcp->buf); \
+} while (0)
+#define HDCP_FULL_DDC_TRACE(hdcp) do { \
+	HDCP_DDC_READ_TRACE(hdcp, "BKSV", hdcp->auth.msg.hdcp1.bksv, \
+			sizeof(hdcp->auth.msg.hdcp1.bksv)); \
+	HDCP_DDC_READ_TRACE(hdcp, "BCAPS", &hdcp->auth.msg.hdcp1.bcaps, \
+			sizeof(hdcp->auth.msg.hdcp1.bcaps)); \
+	HDCP_DDC_WRITE_TRACE(hdcp, "AN", hdcp->auth.msg.hdcp1.an, \
+			sizeof(hdcp->auth.msg.hdcp1.an)); \
+	HDCP_DDC_WRITE_TRACE(hdcp, "AKSV", hdcp->auth.msg.hdcp1.aksv, \
+			sizeof(hdcp->auth.msg.hdcp1.aksv)); \
+	HDCP_DDC_WRITE_TRACE(hdcp, "AINFO", &hdcp->auth.msg.hdcp1.ainfo, \
+			sizeof(hdcp->auth.msg.hdcp1.ainfo)); \
+	HDCP_DDC_READ_TRACE(hdcp, "RI' / R0'", \
+			(uint8_t *)&hdcp->auth.msg.hdcp1.r0p, \
+			sizeof(hdcp->auth.msg.hdcp1.r0p)); \
+	HDCP_DDC_READ_TRACE(hdcp, "BINFO", \
+			(uint8_t *)&hdcp->auth.msg.hdcp1.binfo_dp, \
+			sizeof(hdcp->auth.msg.hdcp1.binfo_dp)); \
+	HDCP_DDC_READ_TRACE(hdcp, "KSVLIST", hdcp->auth.msg.hdcp1.ksvlist, \
+			hdcp->auth.msg.hdcp1.ksvlist_size); \
+	HDCP_DDC_READ_TRACE(hdcp, "V'", hdcp->auth.msg.hdcp1.vp, \
+			sizeof(hdcp->auth.msg.hdcp1.vp)); \
+} while (0)
+#define HDCP_TOP_ADD_DISPLAY_TRACE(hdcp, i) \
+		HDCP_LOG_TOP(hdcp, "[Link %d]\tadd display %d", \
+				hdcp->config.index, i)
+#define HDCP_TOP_REMOVE_DISPLAY_TRACE(hdcp, i) \
+		HDCP_LOG_TOP(hdcp, "[Link %d]\tremove display %d", \
+				hdcp->config.index, i)
+#define HDCP_TOP_HDCP1_DESTROY_SESSION_TRACE(hdcp) \
+		HDCP_LOG_TOP(hdcp, "[Link %d]\tdestroy hdcp1 session", \
+				hdcp->config.index)
+#define HDCP_TOP_RESET_AUTH_TRACE(hdcp) \
+		HDCP_LOG_TOP(hdcp, "[Link %d]\treset authentication", hdcp->config.index)
+#define HDCP_TOP_RESET_CONN_TRACE(hdcp) \
+		HDCP_LOG_TOP(hdcp, "[Link %d]\treset connection", hdcp->config.index)
+#define HDCP_TOP_INTERFACE_TRACE(hdcp) do { \
+		HDCP_LOG_TOP(hdcp, "\n"); \
+		HDCP_LOG_TOP(hdcp, "[Link %d] %s", hdcp->config.index, __func__); \
+} while (0)
+#define HDCP_TOP_INTERFACE_TRACE_WITH_INDEX(hdcp, i) do { \
+		HDCP_LOG_TOP(hdcp, "\n"); \
+		HDCP_LOG_TOP(hdcp, "[Link %d] %s display %d", hdcp->config.index, __func__, i); \
+} while (0)
+
+#endif // MOD_HDCP_LOG_H_
diff --git a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_psp.c b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_psp.c
new file mode 100644
index 000000000000..646d909bbc37
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_psp.c
@@ -0,0 +1,328 @@
+/*
+ * Copyright 2018 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#define MAX_NUM_DISPLAYS 24
+
+
+#include "hdcp.h"
+
+#include "amdgpu.h"
+#include "hdcp_psp.h"
+
+enum mod_hdcp_status mod_hdcp_remove_display_topology(struct mod_hdcp *hdcp)
+{
+
+	struct psp_context *psp = hdcp->config.psp.handle;
+	struct ta_dtm_shared_memory *dtm_cmd;
+	struct mod_hdcp_display *display = NULL;
+	uint8_t i;
+
+	dtm_cmd = (struct ta_dtm_shared_memory *)psp->dtm_context.dtm_shared_buf;
+
+	for (i = 0; i < MAX_NUM_OF_DISPLAYS; i++) {
+		if (hdcp->connection.displays[i].state == MOD_HDCP_DISPLAY_ACTIVE_AND_ADDED) {
+
+			memset(dtm_cmd, 0, sizeof(struct ta_dtm_shared_memory));
+
+			display = &hdcp->connection.displays[i];
+
+			dtm_cmd->cmd_id = TA_DTM_COMMAND__TOPOLOGY_UPDATE_V2;
+			dtm_cmd->dtm_in_message.topology_update_v2.display_handle = display->index;
+			dtm_cmd->dtm_in_message.topology_update_v2.is_active = 0;
+			dtm_cmd->dtm_status = TA_DTM_STATUS__GENERIC_FAILURE;
+
+			psp_dtm_invoke(psp, dtm_cmd->cmd_id);
+
+			if (dtm_cmd->dtm_status != TA_DTM_STATUS__SUCCESS)
+				return MOD_HDCP_STATUS_UPDATE_TOPOLOGY_FAILURE;
+
+			display->state = MOD_HDCP_DISPLAY_ACTIVE;
+			HDCP_TOP_REMOVE_DISPLAY_TRACE(hdcp, display->index);
+		}
+	}
+
+	return MOD_HDCP_STATUS_SUCCESS;
+}
+
+enum mod_hdcp_status mod_hdcp_add_display_topology(struct mod_hdcp *hdcp)
+{
+	struct psp_context *psp = hdcp->config.psp.handle;
+	struct ta_dtm_shared_memory *dtm_cmd;
+	struct mod_hdcp_display *display = NULL;
+	struct mod_hdcp_link *link = &hdcp->connection.link;
+	uint8_t i;
+
+	if (!psp->dtm_context.dtm_initialized) {
+		DRM_ERROR("Failed to add display topology, DTM TA is not initialized.");
+		return MOD_HDCP_STATUS_FAILURE;
+	}
+
+	dtm_cmd = (struct ta_dtm_shared_memory *)psp->dtm_context.dtm_shared_buf;
+
+	for (i = 0; i < MAX_NUM_OF_DISPLAYS; i++) {
+		if (hdcp->connection.displays[i].state == MOD_HDCP_DISPLAY_ACTIVE) {
+			display = &hdcp->connection.displays[i];
+
+			memset(dtm_cmd, 0, sizeof(struct ta_dtm_shared_memory));
+
+			dtm_cmd->cmd_id = TA_DTM_COMMAND__TOPOLOGY_UPDATE_V2;
+			dtm_cmd->dtm_in_message.topology_update_v2.display_handle = display->index;
+			dtm_cmd->dtm_in_message.topology_update_v2.is_active = 1;
+			dtm_cmd->dtm_in_message.topology_update_v2.controller = display->controller;
+			dtm_cmd->dtm_in_message.topology_update_v2.ddc_line = link->ddc_line;
+			dtm_cmd->dtm_in_message.topology_update_v2.dig_be = link->dig_be;
+			dtm_cmd->dtm_in_message.topology_update_v2.dig_fe = display->dig_fe;
+			dtm_cmd->dtm_in_message.topology_update_v2.dp_mst_vcid = display->vc_id;
+			dtm_cmd->dtm_in_message.topology_update_v2.max_hdcp_supported_version =
+				TA_DTM_HDCP_VERSION_MAX_SUPPORTED__1_x;
+			dtm_cmd->dtm_status = TA_DTM_STATUS__GENERIC_FAILURE;
+
+			psp_dtm_invoke(psp, dtm_cmd->cmd_id);
+
+			if (dtm_cmd->dtm_status != TA_DTM_STATUS__SUCCESS)
+				return MOD_HDCP_STATUS_UPDATE_TOPOLOGY_FAILURE;
+
+			display->state = MOD_HDCP_DISPLAY_ACTIVE_AND_ADDED;
+			HDCP_TOP_ADD_DISPLAY_TRACE(hdcp, display->index);
+		}
+	}
+
+	return MOD_HDCP_STATUS_SUCCESS;
+}
+
+enum mod_hdcp_status mod_hdcp_hdcp1_create_session(struct mod_hdcp *hdcp)
+{
+
+	struct psp_context *psp = hdcp->config.psp.handle;
+	struct mod_hdcp_display *display = get_first_added_display(hdcp);
+	struct ta_hdcp_shared_memory *hdcp_cmd;
+
+	if (!psp->hdcp_context.hdcp_initialized) {
+		DRM_ERROR("Failed to create hdcp session. HDCP TA is not initialized.");
+		return MOD_HDCP_STATUS_FAILURE;
+	}
+
+	hdcp_cmd = (struct ta_hdcp_shared_memory *)psp->hdcp_context.hdcp_shared_buf;
+	memset(hdcp_cmd, 0, sizeof(struct ta_hdcp_shared_memory));
+
+	hdcp_cmd->in_msg.hdcp1_create_session.display_handle = display->index;
+	hdcp_cmd->cmd_id = TA_HDCP_COMMAND__HDCP1_CREATE_SESSION;
+
+	psp_hdcp_invoke(psp, hdcp_cmd->cmd_id);
+
+	if (hdcp_cmd->hdcp_status != TA_HDCP_STATUS__SUCCESS)
+		return MOD_HDCP_STATUS_HDCP1_CREATE_SESSION_FAILURE;
+
+	hdcp->auth.id = hdcp_cmd->out_msg.hdcp1_create_session.session_handle;
+	hdcp->auth.msg.hdcp1.ainfo = hdcp_cmd->out_msg.hdcp1_create_session.ainfo_primary;
+	memcpy(hdcp->auth.msg.hdcp1.aksv, hdcp_cmd->out_msg.hdcp1_create_session.aksv_primary,
+		sizeof(hdcp->auth.msg.hdcp1.aksv));
+	memcpy(hdcp->auth.msg.hdcp1.an, hdcp_cmd->out_msg.hdcp1_create_session.an_primary,
+		sizeof(hdcp->auth.msg.hdcp1.an));
+
+	return MOD_HDCP_STATUS_SUCCESS;
+}
+
+enum mod_hdcp_status mod_hdcp_hdcp1_destroy_session(struct mod_hdcp *hdcp)
+{
+
+	struct psp_context *psp = hdcp->config.psp.handle;
+	struct ta_hdcp_shared_memory *hdcp_cmd;
+
+	hdcp_cmd = (struct ta_hdcp_shared_memory *)psp->hdcp_context.hdcp_shared_buf;
+	memset(hdcp_cmd, 0, sizeof(struct ta_hdcp_shared_memory));
+
+	hdcp_cmd->in_msg.hdcp1_destroy_session.session_handle = hdcp->auth.id;
+	hdcp_cmd->cmd_id = TA_HDCP_COMMAND__HDCP1_DESTROY_SESSION;
+
+	psp_hdcp_invoke(psp, hdcp_cmd->cmd_id);
+
+	if (hdcp_cmd->hdcp_status != TA_HDCP_STATUS__SUCCESS)
+		return MOD_HDCP_STATUS_HDCP1_DESTROY_SESSION_FAILURE;
+
+	HDCP_TOP_HDCP1_DESTROY_SESSION_TRACE(hdcp);
+
+	return MOD_HDCP_STATUS_SUCCESS;
+}
+
+enum mod_hdcp_status mod_hdcp_hdcp1_validate_rx(struct mod_hdcp *hdcp)
+{
+	struct psp_context *psp = hdcp->config.psp.handle;
+	struct ta_hdcp_shared_memory *hdcp_cmd;
+
+	hdcp_cmd = (struct ta_hdcp_shared_memory *)psp->hdcp_context.hdcp_shared_buf;
+	memset(hdcp_cmd, 0, sizeof(struct ta_hdcp_shared_memory));
+
+	hdcp_cmd->in_msg.hdcp1_first_part_authentication.session_handle = hdcp->auth.id;
+
+	memcpy(hdcp_cmd->in_msg.hdcp1_first_part_authentication.bksv_primary, hdcp->auth.msg.hdcp1.bksv,
+		TA_HDCP__HDCP1_KSV_SIZE);
+
+	hdcp_cmd->in_msg.hdcp1_first_part_authentication.r0_prime_primary = hdcp->auth.msg.hdcp1.r0p;
+	hdcp_cmd->in_msg.hdcp1_first_part_authentication.bcaps = hdcp->auth.msg.hdcp1.bcaps;
+	hdcp_cmd->cmd_id = TA_HDCP_COMMAND__HDCP1_FIRST_PART_AUTHENTICATION;
+
+	psp_hdcp_invoke(psp, hdcp_cmd->cmd_id);
+
+	if (hdcp_cmd->hdcp_status != TA_HDCP_STATUS__SUCCESS)
+		return MOD_HDCP_STATUS_HDCP1_VALIDATE_RX_FAILURE;
+
+	if (hdcp_cmd->out_msg.hdcp1_first_part_authentication.authentication_status ==
+	    TA_HDCP_AUTHENTICATION_STATUS__HDCP1_FIRST_PART_COMPLETE) {
+		/* needs second part of authentication */
+		hdcp->connection.is_repeater = 1;
+	} else if (hdcp_cmd->out_msg.hdcp1_first_part_authentication.authentication_status ==
+		   TA_HDCP_AUTHENTICATION_STATUS__HDCP1_AUTHENTICATED) {
+		hdcp->connection.is_repeater = 0;
+	} else
+		return MOD_HDCP_STATUS_HDCP1_VALIDATE_RX_FAILURE;
+
+
+	return MOD_HDCP_STATUS_SUCCESS;
+}
+
+enum mod_hdcp_status mod_hdcp_hdcp1_enable_encryption(struct mod_hdcp *hdcp)
+{
+	struct psp_context *psp = hdcp->config.psp.handle;
+	struct ta_hdcp_shared_memory *hdcp_cmd;
+	struct mod_hdcp_display *display = get_first_added_display(hdcp);
+
+	hdcp_cmd = (struct ta_hdcp_shared_memory *)psp->hdcp_context.hdcp_shared_buf;
+	memset(hdcp_cmd, 0, sizeof(struct ta_hdcp_shared_memory));
+
+	hdcp_cmd->in_msg.hdcp1_enable_encryption.session_handle = hdcp->auth.id;
+	hdcp_cmd->cmd_id = TA_HDCP_COMMAND__HDCP1_ENABLE_ENCRYPTION;
+
+	psp_hdcp_invoke(psp, hdcp_cmd->cmd_id);
+
+	if (hdcp_cmd->hdcp_status != TA_HDCP_STATUS__SUCCESS)
+		return MOD_HDCP_STATUS_HDCP1_ENABLE_ENCRYPTION;
+
+	if (!is_dp_mst_hdcp(hdcp)) {
+		display->state = MOD_HDCP_DISPLAY_ENCRYPTION_ENABLED;
+		HDCP_HDCP1_ENABLED_TRACE(hdcp, display->index);
+	}
+	return MOD_HDCP_STATUS_SUCCESS;
+}
+
+enum mod_hdcp_status mod_hdcp_hdcp1_validate_ksvlist_vp(struct mod_hdcp *hdcp)
+{
+	struct psp_context *psp = hdcp->config.psp.handle;
+	struct ta_hdcp_shared_memory *hdcp_cmd;
+
+	hdcp_cmd = (struct ta_hdcp_shared_memory *)psp->hdcp_context.hdcp_shared_buf;
+	memset(hdcp_cmd, 0, sizeof(struct ta_hdcp_shared_memory));
+
+	hdcp_cmd->in_msg.hdcp1_second_part_authentication.session_handle = hdcp->auth.id;
+
+	hdcp_cmd->in_msg.hdcp1_second_part_authentication.ksv_list_size = hdcp->auth.msg.hdcp1.ksvlist_size;
+	memcpy(hdcp_cmd->in_msg.hdcp1_second_part_authentication.ksv_list, hdcp->auth.msg.hdcp1.ksvlist,
+	       hdcp->auth.msg.hdcp1.ksvlist_size);
+
+	memcpy(hdcp_cmd->in_msg.hdcp1_second_part_authentication.v_prime, hdcp->auth.msg.hdcp1.vp,
+	       sizeof(hdcp->auth.msg.hdcp1.vp));
+
+	hdcp_cmd->in_msg.hdcp1_second_part_authentication.bstatus_binfo =
+		is_dp_hdcp(hdcp) ? hdcp->auth.msg.hdcp1.binfo_dp : hdcp->auth.msg.hdcp1.bstatus;
+	hdcp_cmd->cmd_id = TA_HDCP_COMMAND__HDCP1_SECOND_PART_AUTHENTICATION;
+
+	psp_hdcp_invoke(psp, hdcp_cmd->cmd_id);
+
+	if (hdcp_cmd->hdcp_status != TA_HDCP_STATUS__SUCCESS)
+		return MOD_HDCP_STATUS_HDCP1_VALIDATE_KSV_LIST_FAILURE;
+
+	return MOD_HDCP_STATUS_SUCCESS;
+}
+
+enum mod_hdcp_status mod_hdcp_hdcp1_enable_dp_stream_encryption(struct mod_hdcp *hdcp)
+{
+
+	struct psp_context *psp = hdcp->config.psp.handle;
+	struct ta_hdcp_shared_memory *hdcp_cmd;
+	int i = 0;
+
+	hdcp_cmd = (struct ta_hdcp_shared_memory *)psp->hdcp_context.hdcp_shared_buf;
+
+	for (i = 0; i < MAX_NUM_OF_DISPLAYS; i++) {
+
+		if (hdcp->connection.displays[i].state != MOD_HDCP_DISPLAY_ACTIVE_AND_ADDED ||
+		    hdcp->connection.displays[i].adjust.disable)
+			continue;
+
+		memset(hdcp_cmd, 0, sizeof(struct ta_hdcp_shared_memory));
+
+		hdcp_cmd->in_msg.hdcp1_enable_dp_stream_encryption.session_handle = hdcp->auth.id;
+		hdcp_cmd->in_msg.hdcp1_enable_dp_stream_encryption.display_handle = hdcp->connection.displays[i].index;
+		hdcp_cmd->cmd_id = TA_HDCP_COMMAND__HDCP1_ENABLE_DP_STREAM_ENCRYPTION;
+
+		psp_hdcp_invoke(psp, hdcp_cmd->cmd_id);
+
+		if (hdcp_cmd->hdcp_status != TA_HDCP_STATUS__SUCCESS)
+			return MOD_HDCP_STATUS_HDCP1_ENABLE_STREAM_ENCRYPTION_FAILURE;
+
+		hdcp->connection.displays[i].state = MOD_HDCP_DISPLAY_ENCRYPTION_ENABLED;
+		HDCP_HDCP1_ENABLED_TRACE(hdcp, hdcp->connection.displays[i].index);
+	}
+
+	return MOD_HDCP_STATUS_SUCCESS;
+}
+
+enum mod_hdcp_status mod_hdcp_hdcp1_link_maintenance(struct mod_hdcp *hdcp)
+{
+	struct psp_context *psp = hdcp->config.psp.handle;
+	struct ta_hdcp_shared_memory *hdcp_cmd;
+
+	hdcp_cmd = (struct ta_hdcp_shared_memory *)psp->hdcp_context.hdcp_shared_buf;
+
+	memset(hdcp_cmd, 0, sizeof(struct ta_hdcp_shared_memory));
+
+	hdcp_cmd->in_msg.hdcp1_get_encryption_status.session_handle = hdcp->auth.id;
+
+	hdcp_cmd->out_msg.hdcp1_get_encryption_status.protection_level = 0;
+	hdcp_cmd->cmd_id = TA_HDCP_COMMAND__HDCP1_GET_ENCRYPTION_STATUS;
+
+	psp_hdcp_invoke(psp, hdcp_cmd->cmd_id);
+
+	if (hdcp_cmd->hdcp_status != TA_HDCP_STATUS__SUCCESS)
+		return MOD_HDCP_STATUS_HDCP1_LINK_MAINTENANCE_FAILURE;
+
+	return (hdcp_cmd->out_msg.hdcp1_get_encryption_status.protection_level == 1)
+		       ? MOD_HDCP_STATUS_SUCCESS
+		       : MOD_HDCP_STATUS_HDCP1_LINK_MAINTENANCE_FAILURE;
+}
+
+enum mod_hdcp_status mod_hdcp_hdcp1_get_link_encryption_status(struct mod_hdcp *hdcp,
+							       enum mod_hdcp_encryption_status *encryption_status)
+{
+	*encryption_status = MOD_HDCP_ENCRYPTION_STATUS_HDCP_OFF;
+
+	if (mod_hdcp_hdcp1_link_maintenance(hdcp) != MOD_HDCP_STATUS_SUCCESS)
+		return MOD_HDCP_STATUS_FAILURE;
+
+	*encryption_status = MOD_HDCP_ENCRYPTION_STATUS_HDCP1_ON;
+
+	return MOD_HDCP_STATUS_SUCCESS;
+}
+
diff --git a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_psp.h b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_psp.h
new file mode 100644
index 000000000000..986fc07ea9ea
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_psp.h
@@ -0,0 +1,272 @@
+/*
+ * Copyright 2019 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#ifndef MODULES_HDCP_HDCP_PSP_H_
+#define MODULES_HDCP_HDCP_PSP_H_
+
+/*
+ * NOTE: These parameters are a one-to-one copy of the
+ * parameters required by PSP
+ */
+enum bgd_security_hdcp_encryption_level {
+	HDCP_ENCRYPTION_LEVEL__INVALID = 0,
+	HDCP_ENCRYPTION_LEVEL__OFF,
+	HDCP_ENCRYPTION_LEVEL__ON
+};
+
+enum ta_dtm_command {
+	TA_DTM_COMMAND__UNUSED_1 = 1,
+	TA_DTM_COMMAND__TOPOLOGY_UPDATE_V2,
+	TA_DTM_COMMAND__TOPOLOGY_ASSR_ENABLE
+};
+
+/* DTM related enumerations */
+/**********************************************************/
+
+enum ta_dtm_status {
+	TA_DTM_STATUS__SUCCESS = 0x00,
+	TA_DTM_STATUS__GENERIC_FAILURE = 0x01,
+	TA_DTM_STATUS__INVALID_PARAMETER = 0x02,
+	TA_DTM_STATUS__NULL_POINTER = 0x3
+};
+
+/* input/output structures for DTM commands */
+/**********************************************************/
+/**
+ * Input structures
+ */
+enum ta_dtm_hdcp_version_max_supported {
+	TA_DTM_HDCP_VERSION_MAX_SUPPORTED__NONE = 0,
+	TA_DTM_HDCP_VERSION_MAX_SUPPORTED__1_x = 10,
+	TA_DTM_HDCP_VERSION_MAX_SUPPORTED__2_0 = 20,
+	TA_DTM_HDCP_VERSION_MAX_SUPPORTED__2_1 = 21,
+	TA_DTM_HDCP_VERSION_MAX_SUPPORTED__2_2 = 22,
+	TA_DTM_HDCP_VERSION_MAX_SUPPORTED__2_3 = 23
+};
+
+struct ta_dtm_topology_update_input_v2 {
+	/* display handle is unique across the driver and is used to identify a display */
+	/* for all security interfaces which reference displays such as HDCP */
+	uint32_t display_handle;
+	uint32_t is_active;
+	uint32_t is_miracast;
+	uint32_t controller;
+	uint32_t ddc_line;
+	uint32_t dig_be;
+	uint32_t dig_fe;
+	uint32_t dp_mst_vcid;
+	uint32_t is_assr;
+	uint32_t max_hdcp_supported_version;
+};
+
+struct ta_dtm_topology_assr_enable {
+	uint32_t display_topology_dig_be_index;
+};
+
+/**
+ * Output structures
+ */
+
+/* No output structures yet */
+
+union ta_dtm_cmd_input {
+	struct ta_dtm_topology_update_input_v2 topology_update_v2;
+	struct ta_dtm_topology_assr_enable topology_assr_enable;
+};
+
+union ta_dtm_cmd_output {
+	uint32_t reserved;
+};
+
+struct ta_dtm_shared_memory {
+	uint32_t cmd_id;
+	uint32_t resp_id;
+	enum ta_dtm_status dtm_status;
+	uint32_t reserved;
+	union ta_dtm_cmd_input dtm_in_message;
+	union ta_dtm_cmd_output dtm_out_message;
+};
+
+int psp_cmd_submit_buf(struct psp_context *psp, struct amdgpu_firmware_info *ucode, struct psp_gfx_cmd_resp *cmd,
+		uint64_t fence_mc_addr);
+
+enum ta_hdcp_command {
+	TA_HDCP_COMMAND__INITIALIZE,
+	TA_HDCP_COMMAND__HDCP1_CREATE_SESSION,
+	TA_HDCP_COMMAND__HDCP1_DESTROY_SESSION,
+	TA_HDCP_COMMAND__HDCP1_FIRST_PART_AUTHENTICATION,
+	TA_HDCP_COMMAND__HDCP1_SECOND_PART_AUTHENTICATION,
+	TA_HDCP_COMMAND__HDCP1_ENABLE_ENCRYPTION,
+	TA_HDCP_COMMAND__HDCP1_ENABLE_DP_STREAM_ENCRYPTION,
+	TA_HDCP_COMMAND__HDCP1_GET_ENCRYPTION_STATUS,
+};
+
+
+/* HDCP related enumerations */
+/**********************************************************/
+#define TA_HDCP__INVALID_SESSION 0xFFFF
+#define TA_HDCP__HDCP1_AN_SIZE 8
+#define TA_HDCP__HDCP1_KSV_SIZE 5
+#define TA_HDCP__HDCP1_KSV_LIST_MAX_ENTRIES 127
+#define TA_HDCP__HDCP1_V_PRIME_SIZE 20
+
+enum ta_hdcp_status {
+	TA_HDCP_STATUS__SUCCESS = 0x00,
+	TA_HDCP_STATUS__GENERIC_FAILURE = 0x01,
+	TA_HDCP_STATUS__NULL_POINTER = 0x02,
+	TA_HDCP_STATUS__FAILED_ALLOCATING_SESSION = 0x03,
+	TA_HDCP_STATUS__FAILED_SETUP_TX = 0x04,
+	TA_HDCP_STATUS__INVALID_PARAMETER = 0x05,
+	TA_HDCP_STATUS__VHX_ERROR = 0x06,
+	TA_HDCP_STATUS__SESSION_NOT_CLOSED_PROPERLY = 0x07,
+	TA_HDCP_STATUS__SRM_FAILURE = 0x08,
+	TA_HDCP_STATUS__MST_AUTHENTICATED_ALREADY_STARTED = 0x09,
+	TA_HDCP_STATUS__AKE_SEND_CERT_FAILURE = 0x0A,
+	TA_HDCP_STATUS__AKE_NO_STORED_KM_FAILURE = 0x0B,
+	TA_HDCP_STATUS__AKE_SEND_HPRIME_FAILURE = 0x0C,
+	TA_HDCP_STATUS__LC_SEND_LPRIME_FAILURE = 0x0D,
+	TA_HDCP_STATUS__SKE_SEND_EKS_FAILURE = 0x0E,
+	TA_HDCP_STATUS__REPAUTH_SEND_RXIDLIST_FAILURE = 0x0F,
+	TA_HDCP_STATUS__REPAUTH_STREAM_READY_FAILURE = 0x10,
+	TA_HDCP_STATUS__ASD_GENERIC_FAILURE = 0x11,
+	TA_HDCP_STATUS__UNWRAP_SECRET_FAILURE = 0x12,
+	TA_HDCP_STATUS__ENABLE_ENCR_FAILURE = 0x13,
+	TA_HDCP_STATUS__DISABLE_ENCR_FAILURE = 0x14,
+	TA_HDCP_STATUS__NOT_ENOUGH_MEMORY_FAILURE = 0x15,
+	TA_HDCP_STATUS__UNKNOWN_MESSAGE = 0x16,
+	TA_HDCP_STATUS__TOO_MANY_STREAM = 0x17
+};
+
+enum ta_hdcp_authentication_status {
+	TA_HDCP_AUTHENTICATION_STATUS__NOT_STARTED = 0x00,
+	TA_HDCP_AUTHENTICATION_STATUS__HDCP1_FIRST_PART_FAILED = 0x01,
+	TA_HDCP_AUTHENTICATION_STATUS__HDCP1_FIRST_PART_COMPLETE = 0x02,
+	TA_HDCP_AUTHENTICATION_STATUS__HDCP1_SECOND_PART_FAILED = 0x03,
+	TA_HDCP_AUTHENTICATION_STATUS__HDCP1_AUTHENTICATED = 0x04,
+	TA_HDCP_AUTHENTICATION_STATUS__HDCP1_KSV_VALIDATION_FAILED = 0x09
+};
+
+
+/* input/output structures for HDCP commands */
+/**********************************************************/
+struct ta_hdcp_cmd_hdcp1_create_session_input {
+	uint8_t display_handle;
+};
+
+struct ta_hdcp_cmd_hdcp1_create_session_output {
+	uint32_t session_handle;
+	uint8_t an_primary[TA_HDCP__HDCP1_AN_SIZE];
+	uint8_t aksv_primary[TA_HDCP__HDCP1_KSV_SIZE];
+	uint8_t ainfo_primary;
+	uint8_t an_secondary[TA_HDCP__HDCP1_AN_SIZE];
+	uint8_t aksv_secondary[TA_HDCP__HDCP1_KSV_SIZE];
+	uint8_t ainfo_secondary;
+};
+
+struct ta_hdcp_cmd_hdcp1_destroy_session_input {
+	uint32_t session_handle;
+};
+
+struct ta_hdcp_cmd_hdcp1_first_part_authentication_input {
+	uint32_t session_handle;
+	uint8_t bksv_primary[TA_HDCP__HDCP1_KSV_SIZE];
+	uint8_t bksv_secondary[TA_HDCP__HDCP1_KSV_SIZE];
+	uint8_t bcaps;
+	uint16_t r0_prime_primary;
+	uint16_t r0_prime_secondary;
+};
+
+struct ta_hdcp_cmd_hdcp1_first_part_authentication_output {
+	enum ta_hdcp_authentication_status authentication_status;
+};
+
+struct ta_hdcp_cmd_hdcp1_second_part_authentication_input {
+	uint32_t session_handle;
+	uint16_t bstatus_binfo;
+	uint8_t ksv_list[TA_HDCP__HDCP1_KSV_LIST_MAX_ENTRIES][TA_HDCP__HDCP1_KSV_SIZE];
+	uint32_t ksv_list_size;
+	uint8_t pj_prime;
+	uint8_t v_prime[TA_HDCP__HDCP1_V_PRIME_SIZE];
+};
+
+struct ta_hdcp_cmd_hdcp1_second_part_authentication_output {
+	enum ta_hdcp_authentication_status authentication_status;
+};
+
+struct ta_hdcp_cmd_hdcp1_enable_encryption_input {
+	uint32_t session_handle;
+};
+
+struct ta_hdcp_cmd_hdcp1_enable_dp_stream_encryption_input {
+	uint32_t session_handle;
+	uint32_t display_handle;
+};
+
+struct ta_hdcp_cmd_hdcp1_get_encryption_status_input {
+	uint32_t session_handle;
+};
+
+struct ta_hdcp_cmd_hdcp1_get_encryption_status_output {
+	uint32_t protection_level;
+};
+
+/**********************************************************/
+/* Common input structure for HDCP callbacks */
+union ta_hdcp_cmd_input {
+	struct ta_hdcp_cmd_hdcp1_create_session_input hdcp1_create_session;
+	struct ta_hdcp_cmd_hdcp1_destroy_session_input hdcp1_destroy_session;
+	struct ta_hdcp_cmd_hdcp1_first_part_authentication_input hdcp1_first_part_authentication;
+	struct ta_hdcp_cmd_hdcp1_second_part_authentication_input hdcp1_second_part_authentication;
+	struct ta_hdcp_cmd_hdcp1_enable_encryption_input hdcp1_enable_encryption;
+	struct ta_hdcp_cmd_hdcp1_enable_dp_stream_encryption_input hdcp1_enable_dp_stream_encryption;
+	struct ta_hdcp_cmd_hdcp1_get_encryption_status_input hdcp1_get_encryption_status;
+};
+
+/* Common output structure for HDCP callbacks */
+union ta_hdcp_cmd_output {
+	struct ta_hdcp_cmd_hdcp1_create_session_output hdcp1_create_session;
+	struct ta_hdcp_cmd_hdcp1_first_part_authentication_output hdcp1_first_part_authentication;
+	struct ta_hdcp_cmd_hdcp1_second_part_authentication_output hdcp1_second_part_authentication;
+	struct ta_hdcp_cmd_hdcp1_get_encryption_status_output hdcp1_get_encryption_status;
+};
+/**********************************************************/
+
+struct ta_hdcp_shared_memory {
+	uint32_t cmd_id;
+	enum ta_hdcp_status hdcp_status;
+	uint32_t reserved;
+	union ta_hdcp_cmd_input in_msg;
+	union ta_hdcp_cmd_output out_msg;
+};
+
+enum psp_status {
+	PSP_STATUS__SUCCESS = 0,
+	PSP_STATUS__ERROR_INVALID_PARAMS,
+	PSP_STATUS__ERROR_GENERIC,
+	PSP_STATUS__ERROR_OUT_OF_MEMORY,
+	PSP_STATUS__ERROR_UNSUPPORTED_FEATURE
+};
+
+#endif /* MODULES_HDCP_HDCP_PSP_H_ */
diff --git a/drivers/gpu/drm/amd/display/modules/inc/mod_freesync.h b/drivers/gpu/drm/amd/display/modules/inc/mod_freesync.h
index dc187844d10b..dbe7835aabcf 100644
--- a/drivers/gpu/drm/amd/display/modules/inc/mod_freesync.h
+++ b/drivers/gpu/drm/amd/display/modules/inc/mod_freesync.h
@@ -92,6 +92,7 @@ struct mod_vrr_params_btr {
 	uint32_t inserted_duration_in_us;
 	uint32_t frames_to_insert;
 	uint32_t frame_counter;
+	uint32_t margin_in_us;
 };
 
 struct mod_vrr_params_fixed_refresh {
diff --git a/drivers/gpu/drm/amd/display/modules/inc/mod_hdcp.h b/drivers/gpu/drm/amd/display/modules/inc/mod_hdcp.h
new file mode 100644
index 000000000000..dea21702edff
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/modules/inc/mod_hdcp.h
@@ -0,0 +1,289 @@
+/*
+ * Copyright 2019 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#ifndef MOD_HDCP_H_
+#define MOD_HDCP_H_
+
+#include "os_types.h"
+#include "signal_types.h"
+
+/* Forward Declarations */
+struct mod_hdcp;
+
+#define MAX_NUM_OF_DISPLAYS 6
+#define MAX_NUM_OF_ATTEMPTS 4
+#define MAX_NUM_OF_ERROR_TRACE 10
+
+/* detailed return status */
+enum mod_hdcp_status {
+	MOD_HDCP_STATUS_SUCCESS = 0,
+	MOD_HDCP_STATUS_FAILURE,
+	MOD_HDCP_STATUS_RESET_NEEDED,
+	MOD_HDCP_STATUS_DISPLAY_OUT_OF_BOUND,
+	MOD_HDCP_STATUS_DISPLAY_NOT_FOUND,
+	MOD_HDCP_STATUS_INVALID_STATE,
+	MOD_HDCP_STATUS_NOT_IMPLEMENTED,
+	MOD_HDCP_STATUS_INTERNAL_POLICY_FAILURE,
+	MOD_HDCP_STATUS_UPDATE_TOPOLOGY_FAILURE,
+	MOD_HDCP_STATUS_CREATE_PSP_SERVICE_FAILURE,
+	MOD_HDCP_STATUS_DESTROY_PSP_SERVICE_FAILURE,
+	MOD_HDCP_STATUS_HDCP1_CREATE_SESSION_FAILURE,
+	MOD_HDCP_STATUS_HDCP1_DESTROY_SESSION_FAILURE,
+	MOD_HDCP_STATUS_HDCP1_VALIDATE_ENCRYPTION_FAILURE,
+	MOD_HDCP_STATUS_HDCP1_NOT_HDCP_REPEATER,
+	MOD_HDCP_STATUS_HDCP1_NOT_CAPABLE,
+	MOD_HDCP_STATUS_HDCP1_R0_PRIME_PENDING,
+	MOD_HDCP_STATUS_HDCP1_VALIDATE_RX_FAILURE,
+	MOD_HDCP_STATUS_HDCP1_KSV_LIST_NOT_READY,
+	MOD_HDCP_STATUS_HDCP1_VALIDATE_KSV_LIST_FAILURE,
+	MOD_HDCP_STATUS_HDCP1_ENABLE_ENCRYPTION,
+	MOD_HDCP_STATUS_HDCP1_ENABLE_STREAM_ENCRYPTION_FAILURE,
+	MOD_HDCP_STATUS_HDCP1_MAX_CASCADE_EXCEEDED_FAILURE,
+	MOD_HDCP_STATUS_HDCP1_MAX_DEVS_EXCEEDED_FAILURE,
+	MOD_HDCP_STATUS_HDCP1_DEVICE_COUNT_MISMATCH_FAILURE,
+	MOD_HDCP_STATUS_HDCP1_LINK_INTEGRITY_FAILURE,
+	MOD_HDCP_STATUS_HDCP1_REAUTH_REQUEST_ISSUED,
+	MOD_HDCP_STATUS_HDCP1_LINK_MAINTENANCE_FAILURE,
+	MOD_HDCP_STATUS_HDCP1_INVALID_BKSV,
+	MOD_HDCP_STATUS_DDC_FAILURE, /* TODO: specific errors */
+	MOD_HDCP_STATUS_INVALID_OPERATION,
+	MOD_HDCP_STATUS_HDCP2_NOT_CAPABLE,
+	MOD_HDCP_STATUS_HDCP2_CREATE_SESSION_FAILURE,
+	MOD_HDCP_STATUS_HDCP2_DESTROY_SESSION_FAILURE,
+	MOD_HDCP_STATUS_HDCP2_PREP_AKE_INIT_FAILURE,
+	MOD_HDCP_STATUS_HDCP2_AKE_CERT_PENDING,
+	MOD_HDCP_STATUS_HDCP2_H_PRIME_PENDING,
+	MOD_HDCP_STATUS_HDCP2_PAIRING_INFO_PENDING,
+	MOD_HDCP_STATUS_HDCP2_VALIDATE_AKE_CERT_FAILURE,
+	MOD_HDCP_STATUS_HDCP2_VALIDATE_H_PRIME_FAILURE,
+	MOD_HDCP_STATUS_HDCP2_VALIDATE_PAIRING_INFO_FAILURE,
+	MOD_HDCP_STATUS_HDCP2_PREP_LC_INIT_FAILURE,
+	MOD_HDCP_STATUS_HDCP2_L_PRIME_PENDING,
+	MOD_HDCP_STATUS_HDCP2_VALIDATE_L_PRIME_FAILURE,
+	MOD_HDCP_STATUS_HDCP2_PREP_EKS_FAILURE,
+	MOD_HDCP_STATUS_HDCP2_ENABLE_ENCRYPTION_FAILURE,
+	MOD_HDCP_STATUS_HDCP2_RX_ID_LIST_NOT_READY,
+	MOD_HDCP_STATUS_HDCP2_VALIDATE_RX_ID_LIST_FAILURE,
+	MOD_HDCP_STATUS_HDCP2_ENABLE_STREAM_ENCRYPTION,
+	MOD_HDCP_STATUS_HDCP2_STREAM_READY_PENDING,
+	MOD_HDCP_STATUS_HDCP2_VALIDATE_STREAM_READY_FAILURE,
+	MOD_HDCP_STATUS_HDCP2_PREPARE_STREAM_MANAGEMENT_FAILURE,
+	MOD_HDCP_STATUS_HDCP2_REAUTH_REQUEST,
+	MOD_HDCP_STATUS_HDCP2_REAUTH_LINK_INTEGRITY_FAILURE,
+	MOD_HDCP_STATUS_HDCP2_DEVICE_COUNT_MISMATCH_FAILURE,
+};
+
+struct mod_hdcp_displayport {
+	uint8_t rev;
+	uint8_t assr_supported;
+};
+
+struct mod_hdcp_hdmi {
+	uint8_t reserved;
+};
+enum mod_hdcp_operation_mode {
+	MOD_HDCP_MODE_OFF,
+	MOD_HDCP_MODE_DEFAULT,
+	MOD_HDCP_MODE_DP,
+	MOD_HDCP_MODE_DP_MST
+};
+
+enum mod_hdcp_display_state {
+	MOD_HDCP_DISPLAY_INACTIVE = 0,
+	MOD_HDCP_DISPLAY_ACTIVE,
+	MOD_HDCP_DISPLAY_ACTIVE_AND_ADDED,
+	MOD_HDCP_DISPLAY_ENCRYPTION_ENABLED
+};
+
+struct mod_hdcp_ddc {
+	void *handle;
+	struct {
+		bool (*read_i2c)(void *handle,
+				uint32_t address,
+				uint8_t offset,
+				uint8_t *data,
+				uint32_t size);
+		bool (*write_i2c)(void *handle,
+				uint32_t address,
+				const uint8_t *data,
+				uint32_t size);
+		bool (*read_dpcd)(void *handle,
+				uint32_t address,
+				uint8_t *data,
+				uint32_t size);
+		bool (*write_dpcd)(void *handle,
+				uint32_t address,
+				const uint8_t *data,
+				uint32_t size);
+	} funcs;
+};
+
+struct mod_hdcp_psp {
+	void *handle;
+	void *funcs;
+};
+
+struct mod_hdcp_display_adjustment {
+	uint8_t disable			: 1;
+	uint8_t reserved		: 7;
+};
+
+struct mod_hdcp_link_adjustment_hdcp1 {
+	uint8_t disable			: 1;
+	uint8_t postpone_encryption	: 1;
+	uint8_t reserved		: 6;
+};
+
+struct mod_hdcp_link_adjustment_hdcp2 {
+	uint8_t disable			: 1;
+	uint8_t disable_type1		: 1;
+	uint8_t force_no_stored_km	: 1;
+	uint8_t increase_h_prime_timeout: 1;
+	uint8_t reserved		: 4;
+};
+
+struct mod_hdcp_link_adjustment {
+	uint8_t auth_delay;
+	struct mod_hdcp_link_adjustment_hdcp1 hdcp1;
+	struct mod_hdcp_link_adjustment_hdcp2 hdcp2;
+};
+
+struct mod_hdcp_error {
+	enum mod_hdcp_status status;
+	uint8_t state_id;
+};
+
+struct mod_hdcp_trace {
+	struct mod_hdcp_error errors[MAX_NUM_OF_ERROR_TRACE];
+	uint8_t error_count;
+};
+
+enum mod_hdcp_encryption_status {
+	MOD_HDCP_ENCRYPTION_STATUS_HDCP_OFF = 0,
+	MOD_HDCP_ENCRYPTION_STATUS_HDCP1_ON,
+	MOD_HDCP_ENCRYPTION_STATUS_HDCP2_TYPE0_ON,
+	MOD_HDCP_ENCRYPTION_STATUS_HDCP2_TYPE1_ON
+};
+
+/* per link events dm has to notify to hdcp module */
+enum mod_hdcp_event {
+	MOD_HDCP_EVENT_CALLBACK = 0,
+	MOD_HDCP_EVENT_WATCHDOG_TIMEOUT,
+	MOD_HDCP_EVENT_CPIRQ
+};
+
+/* output flags from module requesting timer operations */
+struct mod_hdcp_output {
+	uint8_t callback_needed;
+	uint8_t callback_stop;
+	uint8_t watchdog_timer_needed;
+	uint8_t watchdog_timer_stop;
+	uint16_t callback_delay;
+	uint16_t watchdog_timer_delay;
+};
+
+/* used to represent per display info */
+struct mod_hdcp_display {
+	enum mod_hdcp_display_state state;
+	uint8_t index;
+	uint8_t controller;
+	uint8_t dig_fe;
+	union {
+		uint8_t vc_id;
+	};
+	struct mod_hdcp_display_adjustment adjust;
+};
+
+/* used to represent per link info */
+/* in case a link has multiple displays, they share the same link info */
+struct mod_hdcp_link {
+	enum mod_hdcp_operation_mode mode;
+	uint8_t dig_be;
+	uint8_t ddc_line;
+	union {
+		struct mod_hdcp_displayport dp;
+		struct mod_hdcp_hdmi hdmi;
+	};
+	struct mod_hdcp_link_adjustment adjust;
+};
+
+/* a query structure for a display's hdcp information */
+struct mod_hdcp_display_query {
+	const struct mod_hdcp_display *display;
+	const struct mod_hdcp_link *link;
+	const struct mod_hdcp_trace *trace;
+	enum mod_hdcp_encryption_status encryption_status;
+};
+
+/* contains values per on external display configuration change */
+struct mod_hdcp_config {
+	struct mod_hdcp_psp psp;
+	struct mod_hdcp_ddc ddc;
+	uint8_t index;
+};
+
+struct mod_hdcp;
+
+/* dm allocates memory of mod_hdcp per dc_link on dm init based on memory size*/
+size_t mod_hdcp_get_memory_size(void);
+
+/* called per link on link creation */
+enum mod_hdcp_status mod_hdcp_setup(struct mod_hdcp *hdcp,
+		struct mod_hdcp_config *config);
+
+/* called per link on link destroy */
+enum mod_hdcp_status mod_hdcp_teardown(struct mod_hdcp *hdcp);
+
+/* called per display on cp_desired set to true */
+enum mod_hdcp_status mod_hdcp_add_display(struct mod_hdcp *hdcp,
+		struct mod_hdcp_link *link, struct mod_hdcp_display *display,
+		struct mod_hdcp_output *output);
+
+/* called per display on cp_desired set to false */
+enum mod_hdcp_status mod_hdcp_remove_display(struct mod_hdcp *hdcp,
+		uint8_t index, struct mod_hdcp_output *output);
+
+/* called to query hdcp information on a specific index */
+enum mod_hdcp_status mod_hdcp_query_display(struct mod_hdcp *hdcp,
+		uint8_t index, struct mod_hdcp_display_query *query);
+
+/* called per link on connectivity change */
+enum mod_hdcp_status mod_hdcp_reset_connection(struct mod_hdcp *hdcp,
+		struct mod_hdcp_output *output);
+
+/* called per link on events (i.e. callback, watchdog, CP_IRQ) */
+enum mod_hdcp_status mod_hdcp_process_event(struct mod_hdcp *hdcp,
+		enum mod_hdcp_event event, struct mod_hdcp_output *output);
+
+/* called to convert enum mod_hdcp_status to c string */
+char *mod_hdcp_status_to_str(int32_t status);
+
+/* called to convert state id to c string */
+char *mod_hdcp_state_id_to_str(int32_t id);
+
+/* called to convert signal type to operation mode */
+enum mod_hdcp_operation_mode mod_hdcp_signal_type_to_operation_mode(
+		enum signal_type signal);
+#endif /* MOD_HDCP_H_ */
diff --git a/drivers/gpu/drm/amd/display/modules/inc/mod_info_packet.h b/drivers/gpu/drm/amd/display/modules/inc/mod_info_packet.h
index d930bdecb117..ca8ce3c55337 100644
--- a/drivers/gpu/drm/amd/display/modules/inc/mod_info_packet.h
+++ b/drivers/gpu/drm/amd/display/modules/inc/mod_info_packet.h
@@ -35,4 +35,7 @@ struct mod_vrr_params;
 void mod_build_vsc_infopacket(const struct dc_stream_state *stream,
 		struct dc_info_packet *info_packet);
 
+void mod_build_hf_vsif_infopacket(const struct dc_stream_state *stream,
+		struct dc_info_packet *info_packet, int ALLMEnabled, int ALLMValue);
+
 #endif
diff --git a/drivers/gpu/drm/amd/display/modules/info_packet/info_packet.c b/drivers/gpu/drm/amd/display/modules/info_packet/info_packet.c
index d885d642ed7f..db6b08f6d093 100644
--- a/drivers/gpu/drm/amd/display/modules/info_packet/info_packet.c
+++ b/drivers/gpu/drm/amd/display/modules/info_packet/info_packet.c
@@ -31,6 +31,7 @@
 #include "dc.h"
 
 #define HDMI_INFOFRAME_TYPE_VENDOR 0x81
+#define HF_VSIF_VERSION 1
 
 // VTEM Byte Offset
 #define VTEM_PB0		0
@@ -395,3 +396,100 @@ void mod_build_vsc_infopacket(const struct dc_stream_state *stream,
 
 }
 
+/**
+ *****************************************************************************
+ *  Function: mod_build_hf_vsif_infopacket
+ *
+ *  @brief
+ *     Prepare HDMI Vendor Specific info frame.
+ *     Follows HDMI Spec to build up Vendor Specific info frame
+ *
+ *  @param [in] stream: contains data we may need to construct VSIF (i.e. timing_3d_format, etc.)
+ *  @param [out] info_packet:   output structure where to store VSIF
+ *****************************************************************************
+ */
+void mod_build_hf_vsif_infopacket(const struct dc_stream_state *stream,
+		struct dc_info_packet *info_packet, int ALLMEnabled, int ALLMValue)
+{
+		unsigned int length = 5;
+		bool hdmi_vic_mode = false;
+		uint8_t checksum = 0;
+		uint32_t i = 0;
+		enum dc_timing_3d_format format;
+		bool bALLM = (bool)ALLMEnabled;
+		bool bALLMVal = (bool)ALLMValue;
+
+		info_packet->valid = false;
+		format = stream->timing.timing_3d_format;
+		if (stream->view_format == VIEW_3D_FORMAT_NONE)
+			format = TIMING_3D_FORMAT_NONE;
+
+		if (stream->timing.hdmi_vic != 0
+				&& stream->timing.h_total >= 3840
+				&& stream->timing.v_total >= 2160
+				&& format == TIMING_3D_FORMAT_NONE)
+			hdmi_vic_mode = true;
+
+		if ((format == TIMING_3D_FORMAT_NONE) && !hdmi_vic_mode && !bALLM)
+			return;
+
+		info_packet->sb[1] = 0x03;
+		info_packet->sb[2] = 0x0C;
+		info_packet->sb[3] = 0x00;
+
+		if (bALLM) {
+			info_packet->sb[1] = 0xD8;
+			info_packet->sb[2] = 0x5D;
+			info_packet->sb[3] = 0xC4;
+			info_packet->sb[4] = HF_VSIF_VERSION;
+		}
+
+		if (format != TIMING_3D_FORMAT_NONE)
+			info_packet->sb[4] = (2 << 5);
+
+		else if (hdmi_vic_mode)
+			info_packet->sb[4] = (1 << 5);
+
+		switch (format) {
+		case TIMING_3D_FORMAT_HW_FRAME_PACKING:
+		case TIMING_3D_FORMAT_SW_FRAME_PACKING:
+			info_packet->sb[5] = (0x0 << 4);
+			break;
+
+		case TIMING_3D_FORMAT_SIDE_BY_SIDE:
+		case TIMING_3D_FORMAT_SBS_SW_PACKED:
+			info_packet->sb[5] = (0x8 << 4);
+			length = 6;
+			break;
+
+		case TIMING_3D_FORMAT_TOP_AND_BOTTOM:
+		case TIMING_3D_FORMAT_TB_SW_PACKED:
+			info_packet->sb[5] = (0x6 << 4);
+			break;
+
+		default:
+			break;
+		}
+
+		if (hdmi_vic_mode)
+			info_packet->sb[5] = stream->timing.hdmi_vic;
+
+		info_packet->hb0 = HDMI_INFOFRAME_TYPE_VENDOR;
+		info_packet->hb1 = 0x01;
+		info_packet->hb2 = (uint8_t) (length);
+
+		if (bALLM)
+			info_packet->sb[5] = (info_packet->sb[5] & ~0x02) | (bALLMVal << 1);
+
+		checksum += info_packet->hb0;
+		checksum += info_packet->hb1;
+		checksum += info_packet->hb2;
+
+		for (i = 1; i <= length; i++)
+			checksum += info_packet->sb[i];
+
+		info_packet->sb[0] = (uint8_t) (0x100 - checksum);
+
+		info_packet->valid = true;
+}
+
diff --git a/drivers/gpu/drm/amd/display/modules/power/power_helpers.c b/drivers/gpu/drm/amd/display/modules/power/power_helpers.c
index 05e2be856037..4e2f615c3566 100644
--- a/drivers/gpu/drm/amd/display/modules/power/power_helpers.c
+++ b/drivers/gpu/drm/amd/display/modules/power/power_helpers.c
@@ -80,18 +80,18 @@ struct abm_parameters {
 
 static const struct abm_parameters abm_settings_config0[abm_defines_max_level] = {
 //  min_red  max_red  bright_pos  dark_pos  brightness_gain  contrast  deviation  min_knee  max_knee
-	{0xff,   0xbf,    0x20,       0x00,     0xff,            0x99,     0xb3,      0x40,     0xE0},
-	{0xff,   0x85,    0x20,       0x00,     0xff,            0x90,     0xa8,      0x40,     0xE0},
-	{0xff,   0x40,    0x20,       0x00,     0xff,            0x90,     0x68,      0x40,     0xE0},
-	{0x82,   0x4d,    0x20,       0x00,     0x00,            0x90,     0xb3,      0x70,     0x70},
+	{0xff,   0xbf,    0x20,       0x00,     0xff,            0x99,     0xb3,      0x40,     0xe0},
+	{0xde,   0x85,    0x20,       0x00,     0xff,            0x90,     0xa8,      0x40,     0xdf},
+	{0xb0,   0x50,    0x20,       0x00,     0xc0,            0x88,     0x78,      0x70,     0xa0},
+	{0x82,   0x40,    0x20,       0x00,     0x00,            0xff,     0xb3,      0x70,     0x70},
 };
 
 static const struct abm_parameters abm_settings_config1[abm_defines_max_level] = {
 //  min_red  max_red  bright_pos  dark_pos  brightness_gain  contrast  deviation  min_knee  max_knee
-	{0xf0,   0xd9,    0x20,       0x00,     0x00,            0xa8,     0xb3,      0x70,     0x70},
-	{0xcd,   0xa5,    0x20,       0x00,     0x00,            0xa8,     0xb3,      0x70,     0x70},
-	{0x99,   0x65,    0x20,       0x00,     0x00,            0xa8,     0xb3,      0x70,     0x70},
-	{0x82,   0x4d,    0x20,       0x00,     0x00,            0xa8,     0xb3,      0x70,     0x70},
+	{0xf0,   0xd9,    0x20,       0x00,     0x00,            0xff,     0xb3,      0x70,     0x70},
+	{0xcd,   0xa5,    0x20,       0x00,     0x00,            0xff,     0xb3,      0x70,     0x70},
+	{0x99,   0x65,    0x20,       0x00,     0x00,            0xff,     0xb3,      0x70,     0x70},
+	{0x82,   0x4d,    0x20,       0x00,     0x00,            0xff,     0xb3,      0x70,     0x70},
 };
 
 static const struct abm_parameters * const abm_settings[] = {
@@ -115,7 +115,7 @@ static const struct abm_parameters * const abm_settings[] = {
 /* NOTE: iRAM is 256B in size */
 struct iram_table_v_2 {
 	/* flags                      */
-	uint16_t flags;							/* 0x00 U16  */
+	uint16_t min_abm_backlight;					/* 0x00 U16  */
 
 	/* parameters for ABM2.0 algorithm */
 	uint8_t min_reduction[NUM_AMBI_LEVEL][NUM_AGGR_LEVEL];		/* 0x02 U0.8 */
@@ -140,10 +140,10 @@ struct iram_table_v_2 {
 
 	/* For reading PSR State directly from IRAM */
 	uint8_t psr_state;						/* 0xf0       */
-	uint8_t dmcu_mcp_interface_version;							/* 0xf1       */
-	uint8_t dmcu_abm_feature_version;							/* 0xf2       */
-	uint8_t dmcu_psr_feature_version;							/* 0xf3       */
-	uint16_t dmcu_version;										/* 0xf4       */
+	uint8_t dmcu_mcp_interface_version;				/* 0xf1       */
+	uint8_t dmcu_abm_feature_version;				/* 0xf2       */
+	uint8_t dmcu_psr_feature_version;				/* 0xf3       */
+	uint16_t dmcu_version;						/* 0xf4       */
 	uint8_t dmcu_state;						/* 0xf6       */
 
 	uint16_t blRampReduction;					/* 0xf7       */
@@ -164,42 +164,43 @@ struct iram_table_v_2_2 {
 	uint8_t max_reduction[NUM_AMBI_LEVEL][NUM_AGGR_LEVEL];		/* 0x16 U0.8 */
 	uint8_t bright_pos_gain[NUM_AMBI_LEVEL][NUM_AGGR_LEVEL];	/* 0x2a U2.6 */
 	uint8_t dark_pos_gain[NUM_AMBI_LEVEL][NUM_AGGR_LEVEL];		/* 0x3e U2.6 */
-	uint8_t hybrid_factor[NUM_AGGR_LEVEL];						/* 0x52 U0.8 */
-	uint8_t contrast_factor[NUM_AGGR_LEVEL];					/* 0x56 U0.8 */
-	uint8_t deviation_gain[NUM_AGGR_LEVEL];						/* 0x5a U0.8 */
-	uint8_t iir_curve[NUM_AMBI_LEVEL];							/* 0x5e U0.8 */
-	uint8_t min_knee[NUM_AGGR_LEVEL];							/* 0x63 U0.8 */
-	uint8_t max_knee[NUM_AGGR_LEVEL];							/* 0x67 U0.8 */
-	uint8_t pad[21];											/* 0x6b U0.8 */
+	uint8_t hybrid_factor[NUM_AGGR_LEVEL];				/* 0x52 U0.8 */
+	uint8_t contrast_factor[NUM_AGGR_LEVEL];			/* 0x56 U0.8 */
+	uint8_t deviation_gain[NUM_AGGR_LEVEL];				/* 0x5a U0.8 */
+	uint8_t iir_curve[NUM_AMBI_LEVEL];				/* 0x5e U0.8 */
+	uint8_t min_knee[NUM_AGGR_LEVEL];				/* 0x63 U0.8 */
+	uint8_t max_knee[NUM_AGGR_LEVEL];				/* 0x67 U0.8 */
+	uint16_t min_abm_backlight;					/* 0x6b U16  */
+	uint8_t pad[19];						/* 0x6d U0.8 */
 
 	/* parameters for crgb conversion */
-	uint16_t crgb_thresh[NUM_POWER_FN_SEGS];					/* 0x80 U3.13 */
-	uint16_t crgb_offset[NUM_POWER_FN_SEGS];					/* 0x90 U1.15 */
-	uint16_t crgb_slope[NUM_POWER_FN_SEGS];						/* 0xa0 U4.12 */
+	uint16_t crgb_thresh[NUM_POWER_FN_SEGS];			/* 0x80 U3.13 */
+	uint16_t crgb_offset[NUM_POWER_FN_SEGS];			/* 0x90 U1.15 */
+	uint16_t crgb_slope[NUM_POWER_FN_SEGS];				/* 0xa0 U4.12 */
 
 	/* parameters for custom curve */
 	/* thresholds for brightness --> backlight */
-	uint16_t backlight_thresholds[NUM_BL_CURVE_SEGS];			/* 0xb0 U16.0 */
+	uint16_t backlight_thresholds[NUM_BL_CURVE_SEGS];		/* 0xb0 U16.0 */
 	/* offsets for brightness --> backlight */
-	uint16_t backlight_offsets[NUM_BL_CURVE_SEGS];				/* 0xd0 U16.0 */
+	uint16_t backlight_offsets[NUM_BL_CURVE_SEGS];			/* 0xd0 U16.0 */
 
 	/* For reading PSR State directly from IRAM */
-	uint8_t psr_state;											/* 0xf0       */
-	uint8_t dmcu_mcp_interface_version;							/* 0xf1       */
-	uint8_t dmcu_abm_feature_version;							/* 0xf2       */
-	uint8_t dmcu_psr_feature_version;							/* 0xf3       */
-	uint16_t dmcu_version;										/* 0xf4       */
-	uint8_t dmcu_state;											/* 0xf6       */
-
-	uint8_t dummy1;												/* 0xf7       */
-	uint8_t dummy2;												/* 0xf8       */
-	uint8_t dummy3;												/* 0xf9       */
-	uint8_t dummy4;												/* 0xfa       */
-	uint8_t dummy5;												/* 0xfb       */
-	uint8_t dummy6;												/* 0xfc       */
-	uint8_t dummy7;												/* 0xfd       */
-	uint8_t dummy8;												/* 0xfe       */
-	uint8_t dummy9;												/* 0xff       */
+	uint8_t psr_state;						/* 0xf0       */
+	uint8_t dmcu_mcp_interface_version;				/* 0xf1       */
+	uint8_t dmcu_abm_feature_version;				/* 0xf2       */
+	uint8_t dmcu_psr_feature_version;				/* 0xf3       */
+	uint16_t dmcu_version;						/* 0xf4       */
+	uint8_t dmcu_state;						/* 0xf6       */
+
+	uint8_t dummy1;							/* 0xf7       */
+	uint8_t dummy2;							/* 0xf8       */
+	uint8_t dummy3;							/* 0xf9       */
+	uint8_t dummy4;							/* 0xfa       */
+	uint8_t dummy5;							/* 0xfb       */
+	uint8_t dummy6;							/* 0xfc       */
+	uint8_t dummy7;							/* 0xfd       */
+	uint8_t dummy8;							/* 0xfe       */
+	uint8_t dummy9;							/* 0xff       */
 };
 #pragma pack(pop)
 
@@ -271,7 +272,8 @@ void fill_iram_v_2(struct iram_table_v_2 *ram_table, struct dmcu_iram_parameters
 {
 	unsigned int set = params.set;
 
-	ram_table->flags = 0x0;
+	ram_table->min_abm_backlight =
+			cpu_to_be16(params.min_abm_backlight);
 	ram_table->deviation_gain = 0xb3;
 
 	ram_table->blRampReduction =
@@ -445,6 +447,9 @@ void fill_iram_v_2_2(struct iram_table_v_2_2 *ram_table, struct dmcu_iram_parame
 
 	ram_table->flags = 0x0;
 
+	ram_table->min_abm_backlight =
+			cpu_to_be16(params.min_abm_backlight);
+
 	ram_table->deviation_gain[0] = 0xb3;
 	ram_table->deviation_gain[1] = 0xa8;
 	ram_table->deviation_gain[2] = 0x98;
@@ -588,6 +593,10 @@ void fill_iram_v_2_3(struct iram_table_v_2_2 *ram_table, struct dmcu_iram_parame
 	unsigned int set = params.set;
 
 	ram_table->flags = 0x0;
+
+	ram_table->min_abm_backlight =
+			cpu_to_be16(params.min_abm_backlight);
+
 	for (i = 0; i < NUM_AGGR_LEVEL; i++) {
 		ram_table->hybrid_factor[i] = abm_settings[set][i].brightness_gain;
 		ram_table->contrast_factor[i] = abm_settings[set][i].contrast_factor;
diff --git a/drivers/gpu/drm/amd/display/modules/power/power_helpers.h b/drivers/gpu/drm/amd/display/modules/power/power_helpers.h
index da5df00fedce..e54157026330 100644
--- a/drivers/gpu/drm/amd/display/modules/power/power_helpers.h
+++ b/drivers/gpu/drm/amd/display/modules/power/power_helpers.h
@@ -38,6 +38,7 @@ struct dmcu_iram_parameters {
 	unsigned int backlight_lut_array_size;
 	unsigned int backlight_ramping_reduction;
 	unsigned int backlight_ramping_start;
+	unsigned int min_abm_backlight;
 	unsigned int set;
 };
 
diff --git a/drivers/gpu/drm/amd/include/amd_shared.h b/drivers/gpu/drm/amd/include/amd_shared.h
index 8889aaceec60..dc7eb28f0296 100644
--- a/drivers/gpu/drm/amd/include/amd_shared.h
+++ b/drivers/gpu/drm/amd/include/amd_shared.h
@@ -143,6 +143,8 @@ enum PP_FEATURE_MASK {
 enum DC_FEATURE_MASK {
 	DC_FBC_MASK = 0x1,
 	DC_MULTI_MON_PP_MCLK_SWITCH_MASK = 0x2,
+	DC_DISABLE_FRACTIONAL_PWM_MASK = 0x4,
+	DC_PSR_MASK = 0x8,
 };
 
 enum amd_dpm_forced_level;
diff --git a/drivers/gpu/drm/amd/include/asic_reg/bif/bif_4_1_d.h b/drivers/gpu/drm/amd/include/asic_reg/bif/bif_4_1_d.h
index a761ba07f937..fce965984e76 100644
--- a/drivers/gpu/drm/amd/include/asic_reg/bif/bif_4_1_d.h
+++ b/drivers/gpu/drm/amd/include/asic_reg/bif/bif_4_1_d.h
@@ -27,6 +27,7 @@
 #define mmMM_INDEX                                                              0x0
 #define mmMM_INDEX_HI                                                           0x6
 #define mmMM_DATA                                                               0x1
+#define mmCC_BIF_BX_FUSESTRAP0							0x14D7
 #define mmBUS_CNTL                                                              0x1508
 #define mmCONFIG_CNTL                                                           0x1509
 #define mmCONFIG_MEMSIZE                                                        0x150a
diff --git a/drivers/gpu/drm/amd/include/asic_reg/bif/bif_4_1_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/bif/bif_4_1_sh_mask.h
index 8fbfd0261d27..39cc4880beb4 100644
--- a/drivers/gpu/drm/amd/include/asic_reg/bif/bif_4_1_sh_mask.h
+++ b/drivers/gpu/drm/amd/include/asic_reg/bif/bif_4_1_sh_mask.h
@@ -32,6 +32,8 @@
 #define MM_INDEX_HI__MM_OFFSET_HI__SHIFT 0x0
 #define MM_DATA__MM_DATA_MASK 0xffffffff
 #define MM_DATA__MM_DATA__SHIFT 0x0
+#define CC_BIF_BX_FUSESTRAP0__STRAP_BIF_PX_CAPABLE_MASK    0x2
+#define CC_BIF_BX_FUSESTRAP0__STRAP_BIF_PX_CAPABLE__SHIFT  0x1
 #define BUS_CNTL__BIOS_ROM_WRT_EN_MASK 0x1
 #define BUS_CNTL__BIOS_ROM_WRT_EN__SHIFT 0x0
 #define BUS_CNTL__BIOS_ROM_DIS_MASK 0x2
diff --git a/drivers/gpu/drm/amd/include/asic_reg/bif/bif_5_0_d.h b/drivers/gpu/drm/amd/include/asic_reg/bif/bif_5_0_d.h
index 809759f7bb81..8d05d6ca1c8d 100644
--- a/drivers/gpu/drm/amd/include/asic_reg/bif/bif_5_0_d.h
+++ b/drivers/gpu/drm/amd/include/asic_reg/bif/bif_5_0_d.h
@@ -27,6 +27,7 @@
 #define mmMM_INDEX                                                              0x0
 #define mmMM_INDEX_HI                                                           0x6
 #define mmMM_DATA                                                               0x1
+#define mmCC_BIF_BX_FUSESTRAP0							0x14D7
 #define mmCC_BIF_BX_STRAP2							0x152A
 #define mmBIF_MM_INDACCESS_CNTL                                                 0x1500
 #define mmBIF_DOORBELL_APER_EN                                                  0x1501
diff --git a/drivers/gpu/drm/amd/include/asic_reg/bif/bif_5_0_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/bif/bif_5_0_sh_mask.h
index adc71b01f793..73435687d049 100644
--- a/drivers/gpu/drm/amd/include/asic_reg/bif/bif_5_0_sh_mask.h
+++ b/drivers/gpu/drm/amd/include/asic_reg/bif/bif_5_0_sh_mask.h
@@ -32,6 +32,8 @@
 #define MM_INDEX_HI__MM_OFFSET_HI__SHIFT 0x0
 #define MM_DATA__MM_DATA_MASK 0xffffffff
 #define MM_DATA__MM_DATA__SHIFT 0x0
+#define CC_BIF_BX_FUSESTRAP0__STRAP_BIF_PX_CAPABLE_MASK    0x2
+#define CC_BIF_BX_FUSESTRAP0__STRAP_BIF_PX_CAPABLE__SHIFT  0x1
 #define BIF_MM_INDACCESS_CNTL__MM_INDACCESS_DIS_MASK 0x2
 #define BIF_MM_INDACCESS_CNTL__MM_INDACCESS_DIS__SHIFT 0x1
 #define BIF_DOORBELL_APER_EN__BIF_DOORBELL_APER_EN_MASK 0x1
diff --git a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_2_1_0_offset.h b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_2_1_0_offset.h
index be4249adb356..eddf83ec1c39 100644
--- a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_2_1_0_offset.h
+++ b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_2_1_0_offset.h
@@ -9859,6 +9859,8 @@
 #define mmDP0_DP_STEER_FIFO_BASE_IDX                                                                   2
 #define mmDP0_DP_MSA_MISC                                                                              0x210e
 #define mmDP0_DP_MSA_MISC_BASE_IDX                                                                     2
+#define mmDP0_DP_DPHY_INTERNAL_CTRL                                                                    0x210f
+#define mmDP0_DP_DPHY_INTERNAL_CTRL_BASE_IDX                                                           2
 #define mmDP0_DP_VID_TIMING                                                                            0x2110
 #define mmDP0_DP_VID_TIMING_BASE_IDX                                                                   2
 #define mmDP0_DP_VID_N                                                                                 0x2111
@@ -10187,6 +10189,8 @@
 #define mmDP1_DP_STEER_FIFO_BASE_IDX                                                                   2
 #define mmDP1_DP_MSA_MISC                                                                              0x220e
 #define mmDP1_DP_MSA_MISC_BASE_IDX                                                                     2
+#define mmDP1_DP_DPHY_INTERNAL_CTRL                                                                    0x220f
+#define mmDP1_DP_DPHY_INTERNAL_CTRL_BASE_IDX                                                           2
 #define mmDP1_DP_VID_TIMING                                                                            0x2210
 #define mmDP1_DP_VID_TIMING_BASE_IDX                                                                   2
 #define mmDP1_DP_VID_N                                                                                 0x2211
@@ -10515,6 +10519,8 @@
 #define mmDP2_DP_STEER_FIFO_BASE_IDX                                                                   2
 #define mmDP2_DP_MSA_MISC                                                                              0x230e
 #define mmDP2_DP_MSA_MISC_BASE_IDX                                                                     2
+#define mmDP2_DP_DPHY_INTERNAL_CTRL                                                                    0x230f
+#define mmDP2_DP_DPHY_INTERNAL_CTRL_BASE_IDX                                                           2
 #define mmDP2_DP_VID_TIMING                                                                            0x2310
 #define mmDP2_DP_VID_TIMING_BASE_IDX                                                                   2
 #define mmDP2_DP_VID_N                                                                                 0x2311
@@ -10843,6 +10849,8 @@
 #define mmDP3_DP_STEER_FIFO_BASE_IDX                                                                   2
 #define mmDP3_DP_MSA_MISC                                                                              0x240e
 #define mmDP3_DP_MSA_MISC_BASE_IDX                                                                     2
+#define mmDP3_DP_DPHY_INTERNAL_CTRL                                                                    0x240f
+#define mmDP3_DP_DPHY_INTERNAL_CTRL_BASE_IDX                                                           2
 #define mmDP3_DP_VID_TIMING                                                                            0x2410
 #define mmDP3_DP_VID_TIMING_BASE_IDX                                                                   2
 #define mmDP3_DP_VID_N                                                                                 0x2411
@@ -11171,6 +11179,8 @@
 #define mmDP4_DP_STEER_FIFO_BASE_IDX                                                                   2
 #define mmDP4_DP_MSA_MISC                                                                              0x250e
 #define mmDP4_DP_MSA_MISC_BASE_IDX                                                                     2
+#define mmDP4_DP_DPHY_INTERNAL_CTRL                                                                    0x250f
+#define mmDP4_DP_DPHY_INTERNAL_CTRL_BASE_IDX                                                           2
 #define mmDP4_DP_VID_TIMING                                                                            0x2510
 #define mmDP4_DP_VID_TIMING_BASE_IDX                                                                   2
 #define mmDP4_DP_VID_N                                                                                 0x2511
diff --git a/drivers/gpu/drm/amd/include/asic_reg/gc/gc_9_0_offset.h b/drivers/gpu/drm/amd/include/asic_reg/gc/gc_9_0_offset.h
index ca16d9125fbc..2bfaaa8157d0 100644
--- a/drivers/gpu/drm/amd/include/asic_reg/gc/gc_9_0_offset.h
+++ b/drivers/gpu/drm/amd/include/asic_reg/gc/gc_9_0_offset.h
@@ -1146,7 +1146,14 @@
 #define mmATC_L2_MEM_POWER_LS_BASE_IDX                                                                 0
 #define mmATC_L2_CGTT_CLK_CTRL                                                                         0x080c
 #define mmATC_L2_CGTT_CLK_CTRL_BASE_IDX                                                                0
-
+#define mmATC_L2_CACHE_4K_EDC_INDEX                                                                    0x080e
+#define mmATC_L2_CACHE_4K_EDC_INDEX_BASE_IDX                                                           0
+#define mmATC_L2_CACHE_2M_EDC_INDEX                                                                    0x080f
+#define mmATC_L2_CACHE_2M_EDC_INDEX_BASE_IDX                                                           0
+#define mmATC_L2_CACHE_4K_EDC_CNT                                                                      0x0810
+#define mmATC_L2_CACHE_4K_EDC_CNT_BASE_IDX                                                             0
+#define mmATC_L2_CACHE_2M_EDC_CNT                                                                      0x0811
+#define mmATC_L2_CACHE_2M_EDC_CNT_BASE_IDX                                                             0
 
 // addressBlock: gc_utcl2_vml2pfdec
 // base address: 0xa100
@@ -1206,7 +1213,14 @@
 #define mmVM_L2_CACHE_PARITY_CNTL_BASE_IDX                                                             0
 #define mmVM_L2_CGTT_CLK_CTRL                                                                          0x085e
 #define mmVM_L2_CGTT_CLK_CTRL_BASE_IDX                                                                 0
-
+#define mmVM_L2_MEM_ECC_INDEX                                                                          0x0860
+#define mmVM_L2_MEM_ECC_INDEX_BASE_IDX                                                                 0
+#define mmVM_L2_WALKER_MEM_ECC_INDEX                                                                   0x0861
+#define mmVM_L2_WALKER_MEM_ECC_INDEX_BASE_IDX                                                          0
+#define mmVM_L2_MEM_ECC_CNT                                                                            0x0862
+#define mmVM_L2_MEM_ECC_CNT_BASE_IDX                                                                   0
+#define mmVM_L2_WALKER_MEM_ECC_CNT                                                                     0x0863
+#define mmVM_L2_WALKER_MEM_ECC_CNT_BASE_IDX                                                            0
 
 // addressBlock: gc_utcl2_vml2vcdec
 // base address: 0xa200
diff --git a/drivers/gpu/drm/amd/include/asic_reg/gc/gc_9_0_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/gc/gc_9_0_sh_mask.h
index 064c4bb1dc62..d4c613a85352 100644
--- a/drivers/gpu/drm/amd/include/asic_reg/gc/gc_9_0_sh_mask.h
+++ b/drivers/gpu/drm/amd/include/asic_reg/gc/gc_9_0_sh_mask.h
@@ -6661,7 +6661,6 @@
 #define ATC_L2_CGTT_CLK_CTRL__SOFT_STALL_OVERRIDE_MASK                                                        0x00FF0000L
 #define ATC_L2_CGTT_CLK_CTRL__SOFT_OVERRIDE_MASK                                                              0xFF000000L
 
-
 // addressBlock: gc_utcl2_vml2pfdec
 //VM_L2_CNTL
 #define VM_L2_CNTL__ENABLE_L2_CACHE__SHIFT                                                                    0x0
@@ -6991,7 +6990,22 @@
 #define VM_L2_CGTT_CLK_CTRL__MGLS_OVERRIDE_MASK                                                               0x00008000L
 #define VM_L2_CGTT_CLK_CTRL__SOFT_STALL_OVERRIDE_MASK                                                         0x00FF0000L
 #define VM_L2_CGTT_CLK_CTRL__SOFT_OVERRIDE_MASK                                                               0xFF000000L
-
+//VM_L2_MEM_ECC_INDEX
+#define VM_L2_MEM_ECC_INDEX__INDEX__SHIFT                                                                     0x0
+#define VM_L2_MEM_ECC_INDEX__INDEX_MASK                                                                       0x000000FFL
+//VM_L2_WALKER_MEM_ECC_INDEX
+#define VM_L2_WALKER_MEM_ECC_INDEX__INDEX__SHIFT                                                              0x0
+#define VM_L2_WALKER_MEM_ECC_INDEX__INDEX_MASK                                                                0x000000FFL
+//VM_L2_MEM_ECC_CNT
+#define VM_L2_MEM_ECC_CNT__SEC_COUNT__SHIFT                                                                   0xc
+#define VM_L2_MEM_ECC_CNT__DED_COUNT__SHIFT                                                                   0xe
+#define VM_L2_MEM_ECC_CNT__SEC_COUNT_MASK                                                                     0x00003000L
+#define VM_L2_MEM_ECC_CNT__DED_COUNT_MASK                                                                     0x0000C000L
+//VM_L2_WALKER_MEM_ECC_CNT
+#define VM_L2_WALKER_MEM_ECC_CNT__SEC_COUNT__SHIFT                                                            0xc
+#define VM_L2_WALKER_MEM_ECC_CNT__DED_COUNT__SHIFT                                                            0xe
+#define VM_L2_WALKER_MEM_ECC_CNT__SEC_COUNT_MASK                                                              0x00003000L
+#define VM_L2_WALKER_MEM_ECC_CNT__DED_COUNT_MASK                                                              0x0000C000L
 
 // addressBlock: gc_utcl2_vml2vcdec
 //VM_CONTEXT0_CNTL
diff --git a/drivers/gpu/drm/amd/include/asic_reg/nbio/nbio_7_4_0_smn.h b/drivers/gpu/drm/amd/include/asic_reg/nbio/nbio_7_4_0_smn.h
index 4bcacf529852..991128bb9476 100644
--- a/drivers/gpu/drm/amd/include/asic_reg/nbio/nbio_7_4_0_smn.h
+++ b/drivers/gpu/drm/amd/include/asic_reg/nbio/nbio_7_4_0_smn.h
@@ -22,6 +22,9 @@
 #ifndef _nbio_7_4_0_SMN_HEADER
 #define _nbio_7_4_0_SMN_HEADER
 
+// addressBlock: nbio_nbif0_bif_ras_bif_ras_regblk
+// base address: 0x10100000
+#define smnBIFL_RAS_CENTRAL_STATUS			0x10139040
 
 #define smnNBIF_MGCG_CTRL_LCLK				0x1013a21c
 #define smnCPM_CONTROL					0x11180460
@@ -53,4 +56,13 @@
 #define smnPCIE_RX_NUM_NAK				0x11180038
 #define smnPCIE_RX_NUM_NAK_GENERATED			0x1118003c
 
+// addressBlock: nbio_iohub_nb_misc_misc_cfgdec
+// base address: 0x13a10000
+#define smnIOHC_INTERRUPT_EOI				0x13a10120
+
+// addressBlock: nbio_iohub_nb_rascfg_ras_cfgdec
+// base address: 0x13a20000
+#define smnRAS_GLOBAL_STATUS_LO				0x13a20020
+#define smnRAS_GLOBAL_STATUS_HI				0x13a20024
+
 #endif	// _nbio_7_4_0_SMN_HEADER
diff --git a/drivers/gpu/drm/amd/include/asic_reg/nbio/nbio_7_4_offset.h b/drivers/gpu/drm/amd/include/asic_reg/nbio/nbio_7_4_offset.h
index 994e796a28d7..ce5830ebe095 100644
--- a/drivers/gpu/drm/amd/include/asic_reg/nbio/nbio_7_4_offset.h
+++ b/drivers/gpu/drm/amd/include/asic_reg/nbio/nbio_7_4_offset.h
@@ -2793,8 +2793,8 @@
 #define mmBIF_DOORBELL_INT_CNTL_BASE_IDX                                                               2
 #define mmBIF_FB_EN                                                                                    0x00ff
 #define mmBIF_FB_EN_BASE_IDX                                                                           2
-#define mmBIF_BUSY_DELAY_CNTR                                                                          0x0100
-#define mmBIF_BUSY_DELAY_CNTR_BASE_IDX                                                                 2
+#define mmBIF_INTR_CNTL                                                                                0x0100
+#define mmBIF_INTR_CNTL_BASE_IDX                                                                       2
 #define mmBIF_MST_TRANS_PENDING_VF                                                                     0x0109
 #define mmBIF_MST_TRANS_PENDING_VF_BASE_IDX                                                            2
 #define mmBIF_SLV_TRANS_PENDING_VF                                                                     0x010a
diff --git a/drivers/gpu/drm/amd/include/asic_reg/nbio/nbio_7_4_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/nbio/nbio_7_4_sh_mask.h
index d467b939c971..07f04b2b5bdd 100644
--- a/drivers/gpu/drm/amd/include/asic_reg/nbio/nbio_7_4_sh_mask.h
+++ b/drivers/gpu/drm/amd/include/asic_reg/nbio/nbio_7_4_sh_mask.h
@@ -20420,9 +20420,9 @@
 #define BIF_FB_EN__FB_WRITE_EN__SHIFT                                                                         0x1
 #define BIF_FB_EN__FB_READ_EN_MASK                                                                            0x00000001L
 #define BIF_FB_EN__FB_WRITE_EN_MASK                                                                           0x00000002L
-//BIF_BUSY_DELAY_CNTR
-#define BIF_BUSY_DELAY_CNTR__DELAY_CNT__SHIFT                                                                 0x0
-#define BIF_BUSY_DELAY_CNTR__DELAY_CNT_MASK                                                                   0x0000003FL
+//BIF_INTR_CNTL
+#define BIF_INTR_CNTL__RAS_INTR_VEC_SEL__SHIFT                                                                0x0
+#define BIF_INTR_CNTL__RAS_INTR_VEC_SEL_MASK                                                                  0x00000001L
 //BIF_MST_TRANS_PENDING_VF
 #define BIF_MST_TRANS_PENDING_VF__BIF_MST_TRANS_PENDING__SHIFT                                                0x0
 #define BIF_MST_TRANS_PENDING_VF__BIF_MST_TRANS_PENDING_MASK                                                  0x7FFFFFFFL
@@ -48436,4 +48436,47 @@
 #define RCC_DEV0_EPF0_VF15_GFXMSIX_PBA__MSIX_PENDING_BITS_1_MASK                                              0x00000002L
 #define RCC_DEV0_EPF0_VF15_GFXMSIX_PBA__MSIX_PENDING_BITS_2_MASK                                              0x00000004L
 
+//IOHC_INTERRUPT_EOI
+#define IOHC_INTERRUPT_EOI__SMI_EOI__SHIFT                                                                    0x0
+#define IOHC_INTERRUPT_EOI__SCI_EOI__SHIFT                                                                    0x1
+#define IOHC_INTERRUPT_EOI__NMI_EOI__SHIFT                                                                    0x2
+#define IOHC_INTERRUPT_EOI__SMI_EOI_MASK                                                                      0x00000001L
+#define IOHC_INTERRUPT_EOI__SCI_EOI_MASK                                                                      0x00000002L
+#define IOHC_INTERRUPT_EOI__NMI_EOI_MASK                                                                      0x00000004L
+
+//RAS_GLOBAL_STATUS_LO
+#define RAS_GLOBAL_STATUS_LO__ParityErrCorr__SHIFT                                                            0x0
+#define RAS_GLOBAL_STATUS_LO__ParityErrNonFatal__SHIFT                                                        0x1
+#define RAS_GLOBAL_STATUS_LO__ParityErrFatal__SHIFT                                                           0x2
+#define RAS_GLOBAL_STATUS_LO__ParityErrSerr__SHIFT                                                            0x3
+#define RAS_GLOBAL_STATUS_LO__HPLGWA_NMI__SHIFT                                                               0x6
+#define RAS_GLOBAL_STATUS_LO__HPLGWA_SCI__SHIFT                                                               0x7
+#define RAS_GLOBAL_STATUS_LO__HPLGWA_SMI__SHIFT                                                               0x8
+#define RAS_GLOBAL_STATUS_LO__SW_SMI__SHIFT                                                                   0x9
+#define RAS_GLOBAL_STATUS_LO__SW_SCI__SHIFT                                                                   0xa
+#define RAS_GLOBAL_STATUS_LO__SW_NMI__SHIFT                                                                   0xb
+#define RAS_GLOBAL_STATUS_LO__APML_NMI__SHIFT                                                                 0xc
+#define RAS_GLOBAL_STATUS_LO__APML_SyncFld__SHIFT                                                             0xd
+#define RAS_GLOBAL_STATUS_LO__PIN_SyncFld_NMI__SHIFT                                                          0xe
+#define RAS_GLOBAL_STATUS_LO__APML_SyncFld_Private__SHIFT                                                     0xf
+#define RAS_GLOBAL_STATUS_LO__ParityErrCorr_MASK                                                              0x00000001L
+#define RAS_GLOBAL_STATUS_LO__ParityErrNonFatal_MASK                                                          0x00000002L
+#define RAS_GLOBAL_STATUS_LO__ParityErrFatal_MASK                                                             0x00000004L
+#define RAS_GLOBAL_STATUS_LO__ParityErrSerr_MASK                                                              0x00000008L
+#define RAS_GLOBAL_STATUS_LO__HPLGWA_NMI_MASK                                                                 0x00000040L
+#define RAS_GLOBAL_STATUS_LO__HPLGWA_SCI_MASK                                                                 0x00000080L
+#define RAS_GLOBAL_STATUS_LO__HPLGWA_SMI_MASK                                                                 0x00000100L
+#define RAS_GLOBAL_STATUS_LO__SW_SMI_MASK                                                                     0x00000200L
+#define RAS_GLOBAL_STATUS_LO__SW_SCI_MASK                                                                     0x00000400L
+#define RAS_GLOBAL_STATUS_LO__SW_NMI_MASK                                                                     0x00000800L
+#define RAS_GLOBAL_STATUS_LO__APML_NMI_MASK                                                                   0x00001000L
+#define RAS_GLOBAL_STATUS_LO__APML_SyncFld_MASK                                                               0x00002000L
+#define RAS_GLOBAL_STATUS_LO__PIN_SyncFld_NMI_MASK                                                            0x00004000L
+#define RAS_GLOBAL_STATUS_LO__APML_SyncFld_Private_MASK                                                       0x00008000L
+//RAS_GLOBAL_STATUS_HI
+#define RAS_GLOBAL_STATUS_HI__PCIE0PortAErr__SHIFT                                                            0x0
+#define RAS_GLOBAL_STATUS_HI__NBIF0PortAErr__SHIFT                                                            0x1
+#define RAS_GLOBAL_STATUS_HI__PCIE0PortAErr_MASK                                                              0x00000001L
+#define RAS_GLOBAL_STATUS_HI__NBIF0PortAErr_MASK                                                              0x00000002L
+
 #endif
diff --git a/drivers/gpu/drm/amd/include/asic_reg/oss/osssys_4_0_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/oss/osssys_4_0_sh_mask.h
index dc9895a684fe..096d878eb1de 100644
--- a/drivers/gpu/drm/amd/include/asic_reg/oss/osssys_4_0_sh_mask.h
+++ b/drivers/gpu/drm/amd/include/asic_reg/oss/osssys_4_0_sh_mask.h
@@ -588,11 +588,15 @@
 #define IH_STORM_CLIENT_LIST_CNTL__CLIENT30_IS_STORM_CLIENT_MASK                                              0x40000000L
 #define IH_STORM_CLIENT_LIST_CNTL__CLIENT31_IS_STORM_CLIENT_MASK                                              0x80000000L
 //IH_CLK_CTRL
+#define IH_CLK_CTRL__IH_RETRY_INT_CAM_MEM_CLK_SOFT_OVERRIDE__SHIFT                                             0x19
+#define IH_CLK_CTRL__IH_BUFFER_MEM_CLK_SOFT_OVERRIDE__SHIFT                                             		  0x1a
 #define IH_CLK_CTRL__DBUS_MUX_CLK_SOFT_OVERRIDE__SHIFT                                                        0x1b
 #define IH_CLK_CTRL__OSSSYS_SHARE_CLK_SOFT_OVERRIDE__SHIFT                                                    0x1c
 #define IH_CLK_CTRL__LIMIT_SMN_CLK_SOFT_OVERRIDE__SHIFT                                                       0x1d
 #define IH_CLK_CTRL__DYN_CLK_SOFT_OVERRIDE__SHIFT                                                             0x1e
 #define IH_CLK_CTRL__REG_CLK_SOFT_OVERRIDE__SHIFT                                                             0x1f
+#define IH_CLK_CTRL__IH_RETRY_INT_CAM_MEM_CLK_SOFT_OVERRIDE_MASK                                              0x02000000L
+#define IH_CLK_CTRL__IH_BUFFER_MEM_CLK_SOFT_OVERRIDE_MASK                                             		  0x04000000L
 #define IH_CLK_CTRL__DBUS_MUX_CLK_SOFT_OVERRIDE_MASK                                                          0x08000000L
 #define IH_CLK_CTRL__OSSSYS_SHARE_CLK_SOFT_OVERRIDE_MASK                                                      0x10000000L
 #define IH_CLK_CTRL__LIMIT_SMN_CLK_SOFT_OVERRIDE_MASK                                                         0x20000000L
diff --git a/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_0_1_d.h b/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_0_1_d.h
index dbc2e723f659..71169daa701a 100644
--- a/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_0_1_d.h
+++ b/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_0_1_d.h
@@ -49,6 +49,7 @@
 #define ixCG_SPLL_FUNC_CNTL_5                                                   0xc0500150
 #define ixCG_SPLL_FUNC_CNTL_6                                                   0xc0500154
 #define ixCG_SPLL_FUNC_CNTL_7                                                   0xc0500158
+#define ixCG_SPLL_STATUS                                                        0xC050015C
 #define ixSPLL_CNTL_MODE                                                        0xc0500160
 #define ixCG_SPLL_SPREAD_SPECTRUM                                               0xc0500164
 #define ixCG_SPLL_SPREAD_SPECTRUM_2                                             0xc0500168
diff --git a/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_0_1_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_0_1_sh_mask.h
index 6af9f0217b34..61a9a84e0c3a 100644
--- a/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_0_1_sh_mask.h
+++ b/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_0_1_sh_mask.h
@@ -194,6 +194,8 @@
 #define CG_SPLL_FUNC_CNTL_6__SPLL_LF_CNTR__SHIFT 0x19
 #define CG_SPLL_FUNC_CNTL_7__SPLL_BW_CNTRL_MASK 0xfff
 #define CG_SPLL_FUNC_CNTL_7__SPLL_BW_CNTRL__SHIFT 0x0
+#define CG_SPLL_STATUS__SPLL_CHG_STATUS_MASK 0x2
+#define CG_SPLL_STATUS__SPLL_CHG_STATUS__SHIFT 0x1
 #define SPLL_CNTL_MODE__SPLL_SW_DIR_CONTROL_MASK 0x1
 #define SPLL_CNTL_MODE__SPLL_SW_DIR_CONTROL__SHIFT 0x0
 #define SPLL_CNTL_MODE__SPLL_LEGACY_PDIV_MASK 0x2
diff --git a/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_2_d.h b/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_2_d.h
index bd3685166779..351446754c72 100644
--- a/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_2_d.h
+++ b/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_2_d.h
@@ -49,6 +49,7 @@
 #define ixCG_SPLL_FUNC_CNTL_5                                                   0xc0500150
 #define ixCG_SPLL_FUNC_CNTL_6                                                   0xc0500154
 #define ixCG_SPLL_FUNC_CNTL_7                                                   0xc0500158
+#define ixCG_SPLL_STATUS                                                        0xC050015C
 #define ixSPLL_CNTL_MODE                                                        0xc0500160
 #define ixCG_SPLL_SPREAD_SPECTRUM                                               0xc0500164
 #define ixCG_SPLL_SPREAD_SPECTRUM_2                                             0xc0500168
diff --git a/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_2_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_2_sh_mask.h
index 627906674fe8..4bfd5f8ba66c 100644
--- a/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_2_sh_mask.h
+++ b/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_2_sh_mask.h
@@ -194,6 +194,8 @@
 #define CG_SPLL_FUNC_CNTL_6__SPLL_LF_CNTR__SHIFT 0x19
 #define CG_SPLL_FUNC_CNTL_7__SPLL_BW_CNTRL_MASK 0xfff
 #define CG_SPLL_FUNC_CNTL_7__SPLL_BW_CNTRL__SHIFT 0x0
+#define CG_SPLL_STATUS__SPLL_CHG_STATUS_MASK 0x2
+#define CG_SPLL_STATUS__SPLL_CHG_STATUS__SHIFT 0x1
 #define SPLL_CNTL_MODE__SPLL_SW_DIR_CONTROL_MASK 0x1
 #define SPLL_CNTL_MODE__SPLL_SW_DIR_CONTROL__SHIFT 0x0
 #define SPLL_CNTL_MODE__SPLL_LEGACY_PDIV_MASK 0x2
diff --git a/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_3_d.h b/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_3_d.h
index f35aba72e640..21da61c398f5 100644
--- a/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_3_d.h
+++ b/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_3_d.h
@@ -52,6 +52,7 @@
 #define ixCG_SPLL_FUNC_CNTL_5                                                   0xc0500150
 #define ixCG_SPLL_FUNC_CNTL_6                                                   0xc0500154
 #define ixCG_SPLL_FUNC_CNTL_7                                                   0xc0500158
+#define ixCG_SPLL_STATUS                                                        0xC050015C
 #define ixSPLL_CNTL_MODE                                                        0xc0500160
 #define ixCG_SPLL_SPREAD_SPECTRUM                                               0xc0500164
 #define ixCG_SPLL_SPREAD_SPECTRUM_2                                             0xc0500168
diff --git a/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_3_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_3_sh_mask.h
index 481ee6560aa9..f64fe0fbcb32 100644
--- a/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_3_sh_mask.h
+++ b/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_3_sh_mask.h
@@ -220,6 +220,8 @@
 #define CG_SPLL_FUNC_CNTL_6__SPLL_LF_CNTR__SHIFT 0x19
 #define CG_SPLL_FUNC_CNTL_7__SPLL_BW_CNTRL_MASK 0xfff
 #define CG_SPLL_FUNC_CNTL_7__SPLL_BW_CNTRL__SHIFT 0x0
+#define CG_SPLL_STATUS__SPLL_CHG_STATUS_MASK 0x2
+#define CG_SPLL_STATUS__SPLL_CHG_STATUS__SHIFT 0x1
 #define SPLL_CNTL_MODE__SPLL_SW_DIR_CONTROL_MASK 0x1
 #define SPLL_CNTL_MODE__SPLL_SW_DIR_CONTROL__SHIFT 0x0
 #define SPLL_CNTL_MODE__SPLL_LEGACY_PDIV_MASK 0x2
diff --git a/drivers/gpu/drm/amd/include/asic_reg/smuio/smuio_11_0_0_offset.h b/drivers/gpu/drm/amd/include/asic_reg/smuio/smuio_11_0_0_offset.h
index d3876052562b..687d6843c258 100644
--- a/drivers/gpu/drm/amd/include/asic_reg/smuio/smuio_11_0_0_offset.h
+++ b/drivers/gpu/drm/amd/include/asic_reg/smuio/smuio_11_0_0_offset.h
@@ -121,6 +121,98 @@
 #define mmCKSVII2C_IC_COMP_VERSION_BASE_IDX                                                            0
 #define mmCKSVII2C_IC_COMP_TYPE                                                                        0x006d
 #define mmCKSVII2C_IC_COMP_TYPE_BASE_IDX                                                               0
+#define mmCKSVII2C1_IC_CON                                                                             0x0080
+#define mmCKSVII2C1_IC_CON_BASE_IDX                                                                    0
+#define mmCKSVII2C1_IC_TAR                                                                             0x0081
+#define mmCKSVII2C1_IC_TAR_BASE_IDX                                                                    0
+#define mmCKSVII2C1_IC_SAR                                                                             0x0082
+#define mmCKSVII2C1_IC_SAR_BASE_IDX                                                                    0
+#define mmCKSVII2C1_IC_HS_MADDR                                                                        0x0083
+#define mmCKSVII2C1_IC_HS_MADDR_BASE_IDX                                                               0
+#define mmCKSVII2C1_IC_DATA_CMD                                                                        0x0084
+#define mmCKSVII2C1_IC_DATA_CMD_BASE_IDX                                                               0
+#define mmCKSVII2C1_IC_SS_SCL_HCNT                                                                     0x0085
+#define mmCKSVII2C1_IC_SS_SCL_HCNT_BASE_IDX                                                            0
+#define mmCKSVII2C1_IC_SS_SCL_LCNT                                                                     0x0086
+#define mmCKSVII2C1_IC_SS_SCL_LCNT_BASE_IDX                                                            0
+#define mmCKSVII2C1_IC_FS_SCL_HCNT                                                                     0x0087
+#define mmCKSVII2C1_IC_FS_SCL_HCNT_BASE_IDX                                                            0
+#define mmCKSVII2C1_IC_FS_SCL_LCNT                                                                     0x0088
+#define mmCKSVII2C1_IC_FS_SCL_LCNT_BASE_IDX                                                            0
+#define mmCKSVII2C1_IC_HS_SCL_HCNT                                                                     0x0089
+#define mmCKSVII2C1_IC_HS_SCL_HCNT_BASE_IDX                                                            0
+#define mmCKSVII2C1_IC_HS_SCL_LCNT                                                                     0x008a
+#define mmCKSVII2C1_IC_HS_SCL_LCNT_BASE_IDX                                                            0
+#define mmCKSVII2C1_IC_INTR_STAT                                                                       0x008b
+#define mmCKSVII2C1_IC_INTR_STAT_BASE_IDX                                                              0
+#define mmCKSVII2C1_IC_INTR_MASK                                                                       0x008c
+#define mmCKSVII2C1_IC_INTR_MASK_BASE_IDX                                                              0
+#define mmCKSVII2C1_IC_RAW_INTR_STAT                                                                   0x008d
+#define mmCKSVII2C1_IC_RAW_INTR_STAT_BASE_IDX                                                          0
+#define mmCKSVII2C1_IC_RX_TL                                                                           0x008e
+#define mmCKSVII2C1_IC_RX_TL_BASE_IDX                                                                  0
+#define mmCKSVII2C1_IC_TX_TL                                                                           0x008f
+#define mmCKSVII2C1_IC_TX_TL_BASE_IDX                                                                  0
+#define mmCKSVII2C1_IC_CLR_INTR                                                                        0x0090
+#define mmCKSVII2C1_IC_CLR_INTR_BASE_IDX                                                               0
+#define mmCKSVII2C1_IC_CLR_RX_UNDER                                                                    0x0091
+#define mmCKSVII2C1_IC_CLR_RX_UNDER_BASE_IDX                                                           0
+#define mmCKSVII2C1_IC_CLR_RX_OVER                                                                     0x0092
+#define mmCKSVII2C1_IC_CLR_RX_OVER_BASE_IDX                                                            0
+#define mmCKSVII2C1_IC_CLR_TX_OVER                                                                     0x0093
+#define mmCKSVII2C1_IC_CLR_TX_OVER_BASE_IDX                                                            0
+#define mmCKSVII2C1_IC_CLR_RD_REQ                                                                      0x0094
+#define mmCKSVII2C1_IC_CLR_RD_REQ_BASE_IDX                                                             0
+#define mmCKSVII2C1_IC_CLR_TX_ABRT                                                                     0x0095
+#define mmCKSVII2C1_IC_CLR_TX_ABRT_BASE_IDX                                                            0
+#define mmCKSVII2C1_IC_CLR_RX_DONE                                                                     0x0096
+#define mmCKSVII2C1_IC_CLR_RX_DONE_BASE_IDX                                                            0
+#define mmCKSVII2C1_IC_CLR_ACTIVITY                                                                    0x0097
+#define mmCKSVII2C1_IC_CLR_ACTIVITY_BASE_IDX                                                           0
+#define mmCKSVII2C1_IC_CLR_STOP_DET                                                                    0x0098
+#define mmCKSVII2C1_IC_CLR_STOP_DET_BASE_IDX                                                           0
+#define mmCKSVII2C1_IC_CLR_START_DET                                                                   0x0099
+#define mmCKSVII2C1_IC_CLR_START_DET_BASE_IDX                                                          0
+#define mmCKSVII2C1_IC_CLR_GEN_CALL                                                                    0x009a
+#define mmCKSVII2C1_IC_CLR_GEN_CALL_BASE_IDX                                                           0
+#define mmCKSVII2C1_IC_ENABLE                                                                          0x009b
+#define mmCKSVII2C1_IC_ENABLE_BASE_IDX                                                                 0
+#define mmCKSVII2C1_IC_STATUS                                                                          0x009c
+#define mmCKSVII2C1_IC_STATUS_BASE_IDX                                                                 0
+#define mmCKSVII2C1_IC_TXFLR                                                                           0x009d
+#define mmCKSVII2C1_IC_TXFLR_BASE_IDX                                                                  0
+#define mmCKSVII2C1_IC_RXFLR                                                                           0x009e
+#define mmCKSVII2C1_IC_RXFLR_BASE_IDX                                                                  0
+#define mmCKSVII2C1_IC_SDA_HOLD                                                                        0x009f
+#define mmCKSVII2C1_IC_SDA_HOLD_BASE_IDX                                                               0
+#define mmCKSVII2C1_IC_TX_ABRT_SOURCE                                                                  0x00a0
+#define mmCKSVII2C1_IC_TX_ABRT_SOURCE_BASE_IDX                                                         0
+#define mmCKSVII2C1_IC_SLV_DATA_NACK_ONLY                                                              0x00a1
+#define mmCKSVII2C1_IC_SLV_DATA_NACK_ONLY_BASE_IDX                                                     0
+#define mmCKSVII2C1_IC_DMA_CR                                                                          0x00a2
+#define mmCKSVII2C1_IC_DMA_CR_BASE_IDX                                                                 0
+#define mmCKSVII2C1_IC_DMA_TDLR                                                                        0x00a3
+#define mmCKSVII2C1_IC_DMA_TDLR_BASE_IDX                                                               0
+#define mmCKSVII2C1_IC_DMA_RDLR                                                                        0x00a4
+#define mmCKSVII2C1_IC_DMA_RDLR_BASE_IDX                                                               0
+#define mmCKSVII2C1_IC_SDA_SETUP                                                                       0x00a5
+#define mmCKSVII2C1_IC_SDA_SETUP_BASE_IDX                                                              0
+#define mmCKSVII2C1_IC_ACK_GENERAL_CALL                                                                0x00a6
+#define mmCKSVII2C1_IC_ACK_GENERAL_CALL_BASE_IDX                                                       0
+#define mmCKSVII2C1_IC_ENABLE_STATUS                                                                   0x00a7
+#define mmCKSVII2C1_IC_ENABLE_STATUS_BASE_IDX                                                          0
+#define mmCKSVII2C1_IC_FS_SPKLEN                                                                       0x00a8
+#define mmCKSVII2C1_IC_FS_SPKLEN_BASE_IDX                                                              0
+#define mmCKSVII2C1_IC_HS_SPKLEN                                                                       0x00a9
+#define mmCKSVII2C1_IC_HS_SPKLEN_BASE_IDX                                                              0
+#define mmCKSVII2C1_IC_CLR_RESTART_DET                                                                 0x00aa
+#define mmCKSVII2C1_IC_CLR_RESTART_DET_BASE_IDX                                                        0
+#define mmCKSVII2C1_IC_COMP_PARAM_1                                                                    0x00ab
+#define mmCKSVII2C1_IC_COMP_PARAM_1_BASE_IDX                                                           0
+#define mmCKSVII2C1_IC_COMP_VERSION                                                                    0x00ac
+#define mmCKSVII2C1_IC_COMP_VERSION_BASE_IDX                                                           0
+#define mmCKSVII2C1_IC_COMP_TYPE                                                                       0x00ad
+#define mmCKSVII2C1_IC_COMP_TYPE_BASE_IDX                                                              0
 #define mmSMUIO_MP_RESET_INTR                                                                          0x00c1
 #define mmSMUIO_MP_RESET_INTR_BASE_IDX                                                                 0
 #define mmSMUIO_SOC_HALT                                                                               0x00c2
diff --git a/drivers/gpu/drm/amd/include/asic_reg/smuio/smuio_11_0_0_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/smuio/smuio_11_0_0_sh_mask.h
index f8afa3518bf2..6905a9618127 100644
--- a/drivers/gpu/drm/amd/include/asic_reg/smuio/smuio_11_0_0_sh_mask.h
+++ b/drivers/gpu/drm/amd/include/asic_reg/smuio/smuio_11_0_0_sh_mask.h
@@ -268,6 +268,182 @@
 //CKSVII2C_IC_COMP_TYPE
 #define CKSVII2C_IC_COMP_TYPE__COMP_TYPE__SHIFT                                                               0x0
 #define CKSVII2C_IC_COMP_TYPE__COMP_TYPE_MASK                                                                 0xFFFFFFFFL
+//CKSVII2C1_IC_CON
+#define CKSVII2C1_IC_CON__IC1_MASTER_MODE__SHIFT                                                              0x0
+#define CKSVII2C1_IC_CON__IC1_MAX_SPEED_MODE__SHIFT                                                           0x1
+#define CKSVII2C1_IC_CON__IC1_10BITADDR_SLAVE__SHIFT                                                          0x3
+#define CKSVII2C1_IC_CON__IC1_10BITADDR_MASTER__SHIFT                                                         0x4
+#define CKSVII2C1_IC_CON__IC1_RESTART_EN__SHIFT                                                               0x5
+#define CKSVII2C1_IC_CON__IC1_SLAVE_DISABLE__SHIFT                                                            0x6
+#define CKSVII2C1_IC_CON__STOP1_DET_IFADDRESSED__SHIFT                                                        0x7
+#define CKSVII2C1_IC_CON__TX1_EMPTY_CTRL__SHIFT                                                               0x8
+#define CKSVII2C1_IC_CON__RX1_FIFO_FULL_HLD_CTRL__SHIFT                                                       0x9
+#define CKSVII2C1_IC_CON__IC1_MASTER_MODE_MASK                                                                0x00000001L
+#define CKSVII2C1_IC_CON__IC1_MAX_SPEED_MODE_MASK                                                             0x00000006L
+#define CKSVII2C1_IC_CON__IC1_10BITADDR_SLAVE_MASK                                                            0x00000008L
+#define CKSVII2C1_IC_CON__IC1_10BITADDR_MASTER_MASK                                                           0x00000010L
+#define CKSVII2C1_IC_CON__IC1_RESTART_EN_MASK                                                                 0x00000020L
+#define CKSVII2C1_IC_CON__IC1_SLAVE_DISABLE_MASK                                                              0x00000040L
+#define CKSVII2C1_IC_CON__STOP1_DET_IFADDRESSED_MASK                                                          0x00000080L
+#define CKSVII2C1_IC_CON__TX1_EMPTY_CTRL_MASK                                                                 0x00000100L
+#define CKSVII2C1_IC_CON__RX1_FIFO_FULL_HLD_CTRL_MASK                                                         0x00000200L
+//CKSVII2C1_IC_TAR
+#define CKSVII2C1_IC_TAR__IC1_TAR__SHIFT                                                                      0x0
+#define CKSVII2C1_IC_TAR__GC1_OR_START__SHIFT                                                                 0xa
+#define CKSVII2C1_IC_TAR__SPECIAL1__SHIFT                                                                     0xb
+#define CKSVII2C1_IC_TAR__IC1_10BITADDR_MASTER__SHIFT                                                         0xc
+#define CKSVII2C1_IC_TAR__IC1_TAR_MASK                                                                        0x000003FFL
+#define CKSVII2C1_IC_TAR__GC1_OR_START_MASK                                                                   0x00000400L
+#define CKSVII2C1_IC_TAR__SPECIAL1_MASK                                                                       0x00000800L
+#define CKSVII2C1_IC_TAR__IC1_10BITADDR_MASTER_MASK                                                           0x00001000L
+//CKSVII2C1_IC_SAR
+#define CKSVII2C1_IC_SAR__IC1_SAR__SHIFT                                                                      0x0
+#define CKSVII2C1_IC_SAR__IC1_SAR_MASK                                                                        0x000003FFL
+//CKSVII2C1_IC_HS_MADDR
+#define CKSVII2C1_IC_HS_MADDR__IC1_HS_MADDR__SHIFT                                                            0x0
+#define CKSVII2C1_IC_HS_MADDR__IC1_HS_MADDR_MASK                                                              0x00000007L
+//CKSVII2C1_IC_DATA_CMD
+#define CKSVII2C1_IC_DATA_CMD__DAT1__SHIFT                                                                    0x0
+#define CKSVII2C1_IC_DATA_CMD__CMD1__SHIFT                                                                    0x8
+#define CKSVII2C1_IC_DATA_CMD__STOP1__SHIFT                                                                   0x9
+#define CKSVII2C1_IC_DATA_CMD__RESTART1__SHIFT                                                                0xa
+#define CKSVII2C1_IC_DATA_CMD__DAT1_MASK                                                                      0x000000FFL
+#define CKSVII2C1_IC_DATA_CMD__CMD1_MASK                                                                      0x00000100L
+#define CKSVII2C1_IC_DATA_CMD__STOP1_MASK                                                                     0x00000200L
+#define CKSVII2C1_IC_DATA_CMD__RESTART1_MASK                                                                  0x00000400L
+//CKSVII2C1_IC_SS_SCL_HCNT
+#define CKSVII2C1_IC_SS_SCL_HCNT__IC1_SS_SCL_HCNT__SHIFT                                                      0x0
+#define CKSVII2C1_IC_SS_SCL_HCNT__IC1_SS_SCL_HCNT_MASK                                                        0x0000FFFFL
+//CKSVII2C1_IC_SS_SCL_LCNT
+#define CKSVII2C1_IC_SS_SCL_LCNT__IC1_SS_SCL_LCNT__SHIFT                                                      0x0
+#define CKSVII2C1_IC_SS_SCL_LCNT__IC1_SS_SCL_LCNT_MASK                                                        0x0000FFFFL
+//CKSVII2C1_IC_FS_SCL_HCNT
+#define CKSVII2C1_IC_FS_SCL_HCNT__IC1_FS_SCL_HCNT__SHIFT                                                      0x0
+#define CKSVII2C1_IC_FS_SCL_HCNT__IC1_FS_SCL_HCNT_MASK                                                        0x0000FFFFL
+//CKSVII2C1_IC_FS_SCL_LCNT
+#define CKSVII2C1_IC_FS_SCL_LCNT__IC1_FS_SCL_LCNT__SHIFT                                                      0x0
+#define CKSVII2C1_IC_FS_SCL_LCNT__IC1_FS_SCL_LCNT_MASK                                                        0x0000FFFFL
+//CKSVII2C1_IC_HS_SCL_HCNT
+#define CKSVII2C1_IC_HS_SCL_HCNT__IC1_HS_SCL_HCNT__SHIFT                                                      0x0
+#define CKSVII2C1_IC_HS_SCL_HCNT__IC1_HS_SCL_HCNT_MASK                                                        0x0000FFFFL
+//CKSVII2C1_IC_HS_SCL_LCNT
+#define CKSVII2C1_IC_HS_SCL_LCNT__IC1_HS_SCL_LCNT__SHIFT                                                      0x0
+#define CKSVII2C1_IC_HS_SCL_LCNT__IC1_HS_SCL_LCNT_MASK                                                        0x0000FFFFL
+//CKSVII2C1_IC_INTR_STAT
+#define CKSVII2C1_IC_INTR_STAT__R1_RX_UNDER__SHIFT                                                            0x0
+#define CKSVII2C1_IC_INTR_STAT__R1_RX_OVER__SHIFT                                                             0x1
+#define CKSVII2C1_IC_INTR_STAT__R1_RX_FULL__SHIFT                                                             0x2
+#define CKSVII2C1_IC_INTR_STAT__R1_TX_OVER__SHIFT                                                             0x3
+#define CKSVII2C1_IC_INTR_STAT__R1_TX_EMPTY__SHIFT                                                            0x4
+#define CKSVII2C1_IC_INTR_STAT__R1_RD_REQ__SHIFT                                                              0x5
+#define CKSVII2C1_IC_INTR_STAT__R1_TX_ABRT__SHIFT                                                             0x6
+#define CKSVII2C1_IC_INTR_STAT__R1_RX_DONE__SHIFT                                                             0x7
+#define CKSVII2C1_IC_INTR_STAT__R1_ACTIVITY__SHIFT                                                            0x8
+#define CKSVII2C1_IC_INTR_STAT__R1_STOP_DET__SHIFT                                                            0x9
+#define CKSVII2C1_IC_INTR_STAT__R1_START_DET__SHIFT                                                           0xa
+#define CKSVII2C1_IC_INTR_STAT__R1_GEN_CALL__SHIFT                                                            0xb
+#define CKSVII2C1_IC_INTR_STAT__R1_RESTART_DET__SHIFT                                                         0xc
+#define CKSVII2C1_IC_INTR_STAT__R1_MST_ON_HOLD__SHIFT                                                         0xd
+#define CKSVII2C1_IC_INTR_STAT__R1_RX_UNDER_MASK                                                              0x00000001L
+#define CKSVII2C1_IC_INTR_STAT__R1_RX_OVER_MASK                                                               0x00000002L
+#define CKSVII2C1_IC_INTR_STAT__R1_RX_FULL_MASK                                                               0x00000004L
+#define CKSVII2C1_IC_INTR_STAT__R1_TX_OVER_MASK                                                               0x00000008L
+#define CKSVII2C1_IC_INTR_STAT__R1_TX_EMPTY_MASK                                                              0x00000010L
+#define CKSVII2C1_IC_INTR_STAT__R1_RD_REQ_MASK                                                                0x00000020L
+#define CKSVII2C1_IC_INTR_STAT__R1_TX_ABRT_MASK                                                               0x00000040L
+#define CKSVII2C1_IC_INTR_STAT__R1_RX_DONE_MASK                                                               0x00000080L
+#define CKSVII2C1_IC_INTR_STAT__R1_ACTIVITY_MASK                                                              0x00000100L
+#define CKSVII2C1_IC_INTR_STAT__R1_STOP_DET_MASK                                                              0x00000200L
+#define CKSVII2C1_IC_INTR_STAT__R1_START_DET_MASK                                                             0x00000400L
+#define CKSVII2C1_IC_INTR_STAT__R1_GEN_CALL_MASK                                                              0x00000800L
+#define CKSVII2C1_IC_INTR_STAT__R1_RESTART_DET_MASK                                                           0x00001000L
+#define CKSVII2C1_IC_INTR_STAT__R1_MST_ON_HOLD_MASK                                                           0x00002000L
+//CKSVII2C1_IC_INTR_MASK
+#define CKSVII2C1_IC_INTR_MASK__M1_RX_UNDER__SHIFT                                                            0x0
+#define CKSVII2C1_IC_INTR_MASK__M1_RX_OVER__SHIFT                                                             0x1
+#define CKSVII2C1_IC_INTR_MASK__M1_RX_FULL__SHIFT                                                             0x2
+#define CKSVII2C1_IC_INTR_MASK__M1_TX_OVER__SHIFT                                                             0x3
+#define CKSVII2C1_IC_INTR_MASK__M1_TX_EMPTY__SHIFT                                                            0x4
+#define CKSVII2C1_IC_INTR_MASK__M1_RD_REQ__SHIFT                                                              0x5
+#define CKSVII2C1_IC_INTR_MASK__M1_TX_ABRT__SHIFT                                                             0x6
+#define CKSVII2C1_IC_INTR_MASK__M1_RX_DONE__SHIFT                                                             0x7
+#define CKSVII2C1_IC_INTR_MASK__M1_ACTIVITY__SHIFT                                                            0x8
+#define CKSVII2C1_IC_INTR_MASK__M1_STOP_DET__SHIFT                                                            0x9
+#define CKSVII2C1_IC_INTR_MASK__M1_START_DET__SHIFT                                                           0xa
+#define CKSVII2C1_IC_INTR_MASK__M1_GEN_CALL__SHIFT                                                            0xb
+#define CKSVII2C1_IC_INTR_MASK__M1_RESTART_DET__SHIFT                                                         0xc
+#define CKSVII2C1_IC_INTR_MASK__M1_MST_ON_HOLD__SHIFT                                                         0xd
+#define CKSVII2C1_IC_INTR_MASK__M1_RX_UNDER_MASK                                                              0x00000001L
+#define CKSVII2C1_IC_INTR_MASK__M1_RX_OVER_MASK                                                               0x00000002L
+#define CKSVII2C1_IC_INTR_MASK__M1_RX_FULL_MASK                                                               0x00000004L
+#define CKSVII2C1_IC_INTR_MASK__M1_TX_OVER_MASK                                                               0x00000008L
+#define CKSVII2C1_IC_INTR_MASK__M1_TX_EMPTY_MASK                                                              0x00000010L
+#define CKSVII2C1_IC_INTR_MASK__M1_RD_REQ_MASK                                                                0x00000020L
+#define CKSVII2C1_IC_INTR_MASK__M1_TX_ABRT_MASK                                                               0x00000040L
+#define CKSVII2C1_IC_INTR_MASK__M1_RX_DONE_MASK                                                               0x00000080L
+#define CKSVII2C1_IC_INTR_MASK__M1_ACTIVITY_MASK                                                              0x00000100L
+#define CKSVII2C1_IC_INTR_MASK__M1_STOP_DET_MASK                                                              0x00000200L
+#define CKSVII2C1_IC_INTR_MASK__M1_START_DET_MASK                                                             0x00000400L
+#define CKSVII2C1_IC_INTR_MASK__M1_GEN_CALL_MASK                                                              0x00000800L
+#define CKSVII2C1_IC_INTR_MASK__M1_RESTART_DET_MASK                                                           0x00001000L
+#define CKSVII2C1_IC_INTR_MASK__M1_MST_ON_HOLD_MASK                                                           0x00002000L
+//CKSVII2C1_IC_RAW_INTR_STAT
+//CKSVII2C1_IC_RX_TL
+//CKSVII2C1_IC_TX_TL
+//CKSVII2C1_IC_CLR_INTR
+//CKSVII2C1_IC_CLR_RX_UNDER
+//CKSVII2C1_IC_CLR_RX_OVER
+//CKSVII2C1_IC_CLR_TX_OVER
+//CKSVII2C1_IC_CLR_RD_REQ
+//CKSVII2C1_IC_CLR_TX_ABRT
+//CKSVII2C1_IC_CLR_RX_DONE
+//CKSVII2C1_IC_CLR_ACTIVITY
+//CKSVII2C1_IC_CLR_STOP_DET
+//CKSVII2C1_IC_CLR_START_DET
+//CKSVII2C1_IC_CLR_GEN_CALL
+//CKSVII2C1_IC_ENABLE
+#define CKSVII2C1_IC_ENABLE__ENABLE1__SHIFT                                                                   0x0
+#define CKSVII2C1_IC_ENABLE__ABORT1__SHIFT                                                                    0x1
+#define CKSVII2C1_IC_ENABLE__ENABLE1_MASK                                                                     0x00000001L
+#define CKSVII2C1_IC_ENABLE__ABORT1_MASK                                                                      0x00000002L
+//CKSVII2C1_IC_STATUS
+#define CKSVII2C1_IC_STATUS__ACTIVITY1__SHIFT                                                                 0x0
+#define CKSVII2C1_IC_STATUS__TFNF1__SHIFT                                                                     0x1
+#define CKSVII2C1_IC_STATUS__TFE1__SHIFT                                                                      0x2
+#define CKSVII2C1_IC_STATUS__RFNE1__SHIFT                                                                     0x3
+#define CKSVII2C1_IC_STATUS__RFF1__SHIFT                                                                      0x4
+#define CKSVII2C1_IC_STATUS__MST1_ACTIVITY__SHIFT                                                             0x5
+#define CKSVII2C1_IC_STATUS__SLV1_ACTIVITY__SHIFT                                                             0x6
+#define CKSVII2C1_IC_STATUS__ACTIVITY1_MASK                                                                   0x00000001L
+#define CKSVII2C1_IC_STATUS__TFNF1_MASK                                                                       0x00000002L
+#define CKSVII2C1_IC_STATUS__TFE1_MASK                                                                        0x00000004L
+#define CKSVII2C1_IC_STATUS__RFNE1_MASK                                                                       0x00000008L
+#define CKSVII2C1_IC_STATUS__RFF1_MASK                                                                        0x00000010L
+#define CKSVII2C1_IC_STATUS__MST1_ACTIVITY_MASK                                                               0x00000020L
+#define CKSVII2C1_IC_STATUS__SLV1_ACTIVITY_MASK                                                               0x00000040L
+//CKSVII2C1_IC_TXFLR
+//CKSVII2C1_IC_RXFLR
+//CKSVII2C1_IC_SDA_HOLD
+#define CKSVII2C1_IC_SDA_HOLD__IC1_SDA_HOLD__SHIFT                                                            0x0
+#define CKSVII2C1_IC_SDA_HOLD__IC1_SDA_HOLD_MASK                                                              0x00FFFFFFL
+//CKSVII2C1_IC_TX_ABRT_SOURCE
+//CKSVII2C1_IC_SLV_DATA_NACK_ONLY
+//CKSVII2C1_IC_DMA_CR
+//CKSVII2C1_IC_DMA_TDLR
+//CKSVII2C1_IC_DMA_RDLR
+//CKSVII2C1_IC_SDA_SETUP
+#define CKSVII2C1_IC_SDA_SETUP__SDA1_SETUP__SHIFT                                                             0x0
+#define CKSVII2C1_IC_SDA_SETUP__SDA1_SETUP_MASK                                                               0x000000FFL
+//CKSVII2C1_IC_ACK_GENERAL_CALL
+#define CKSVII2C1_IC_ACK_GENERAL_CALL__ACK1_GENERAL_CALL__SHIFT                                               0x0
+#define CKSVII2C1_IC_ACK_GENERAL_CALL__ACK1_GENERAL_CALL_MASK                                                 0x00000001L
+//CKSVII2C1_IC_ENABLE_STATUS
+#define CKSVII2C1_IC_ENABLE_STATUS__IC1_EN__SHIFT                                                             0x0
+#define CKSVII2C1_IC_ENABLE_STATUS__SLV1_RX_ABORTED__SHIFT                                                    0x1
+#define CKSVII2C1_IC_ENABLE_STATUS__SLV1_FIFO_FILLED_AND_FLUSHED__SHIFT                                       0x2
+#define CKSVII2C1_IC_ENABLE_STATUS__IC1_EN_MASK                                                               0x00000001L
+#define CKSVII2C1_IC_ENABLE_STATUS__SLV1_RX_ABORTED_MASK                                                      0x00000002L
+#define CKSVII2C1_IC_ENABLE_STATUS__SLV1_FIFO_FILLED_AND_FLUSHED_MASK                                         0x00000004L
 //SMUIO_MP_RESET_INTR
 #define SMUIO_MP_RESET_INTR__SMUIO_MP_RESET_INTR__SHIFT                                                       0x0
 #define SMUIO_MP_RESET_INTR__SMUIO_MP_RESET_INTR_MASK                                                         0x00000001L
diff --git a/drivers/gpu/drm/amd/include/asic_reg/vcn/vcn_2_5_offset.h b/drivers/gpu/drm/amd/include/asic_reg/vcn/vcn_2_5_offset.h
index cf2149cc12ee..90350f46a0c4 100644
--- a/drivers/gpu/drm/amd/include/asic_reg/vcn/vcn_2_5_offset.h
+++ b/drivers/gpu/drm/amd/include/asic_reg/vcn/vcn_2_5_offset.h
@@ -24,6 +24,18 @@
 
 // addressBlock: uvd0_mmsch_dec
 // base address: 0x1e000
+#define mmMMSCH_VF_VMID                                                                                0x000b
+#define mmMMSCH_VF_VMID_BASE_IDX                                                                       0
+#define mmMMSCH_VF_CTX_ADDR_LO                                                                         0x000c
+#define mmMMSCH_VF_CTX_ADDR_LO_BASE_IDX                                                                0
+#define mmMMSCH_VF_CTX_ADDR_HI                                                                         0x000d
+#define mmMMSCH_VF_CTX_ADDR_HI_BASE_IDX                                                                0
+#define mmMMSCH_VF_CTX_SIZE                                                                            0x000e
+#define mmMMSCH_VF_CTX_SIZE_BASE_IDX                                                                   0
+#define mmMMSCH_VF_MAILBOX_HOST                                                                        0x0012
+#define mmMMSCH_VF_MAILBOX_HOST_BASE_IDX                                                               0
+#define mmMMSCH_VF_MAILBOX_RESP                                                                        0x0013
+#define mmMMSCH_VF_MAILBOX_RESP_BASE_IDX                                                               0
 
 
 // addressBlock: uvd0_jpegnpdec
diff --git a/drivers/gpu/drm/amd/include/atomfirmware.h b/drivers/gpu/drm/amd/include/atomfirmware.h
index e88541d67aa0..dd7cbc00a0aa 100644
--- a/drivers/gpu/drm/amd/include/atomfirmware.h
+++ b/drivers/gpu/drm/amd/include/atomfirmware.h
@@ -492,12 +492,13 @@ struct atom_firmware_info_v3_1
 /* Total 32bit cap indication */
 enum atombios_firmware_capability
 {
-  ATOM_FIRMWARE_CAP_FIRMWARE_POSTED = 0x00000001,
-  ATOM_FIRMWARE_CAP_GPU_VIRTUALIZATION  = 0x00000002,
-  ATOM_FIRMWARE_CAP_WMI_SUPPORT  = 0x00000040,
-  ATOM_FIRMWARE_CAP_HWEMU_ENABLE  = 0x00000080,
-  ATOM_FIRMWARE_CAP_HWEMU_UMC_CFG = 0x00000100,
-  ATOM_FIRMWARE_CAP_SRAM_ECC      = 0x00000200,
+	ATOM_FIRMWARE_CAP_FIRMWARE_POSTED = 0x00000001,
+	ATOM_FIRMWARE_CAP_GPU_VIRTUALIZATION  = 0x00000002,
+	ATOM_FIRMWARE_CAP_WMI_SUPPORT  = 0x00000040,
+	ATOM_FIRMWARE_CAP_HWEMU_ENABLE  = 0x00000080,
+	ATOM_FIRMWARE_CAP_HWEMU_UMC_CFG = 0x00000100,
+	ATOM_FIRMWARE_CAP_SRAM_ECC      = 0x00000200,
+	ATOM_FIRMWARE_CAP_ENABLE_2STAGE_BIST_TRAINING  = 0x00000400,
 };
 
 enum atom_cooling_solution_id{
@@ -671,6 +672,20 @@ struct vram_usagebyfirmware_v2_1
   uint16_t  used_by_driver_in_kb; 
 };
 
+/* This is part of vram_usagebyfirmware_v2_1 */
+struct vram_reserve_block
+{
+	uint32_t start_address_in_kb;
+	uint16_t used_by_firmware_in_kb;
+	uint16_t used_by_driver_in_kb;
+};
+
+/* Definitions for constance */
+enum atomfirmware_internal_constants
+{
+	ONE_KiB	= 0x400,
+	ONE_MiB	= 0x100000,
+};
 
 /* 
   ***************************************************************************
diff --git a/drivers/gpu/drm/amd/include/discovery.h b/drivers/gpu/drm/amd/include/discovery.h
index 5dcb776548d8..7ec4331e67f2 100644
--- a/drivers/gpu/drm/amd/include/discovery.h
+++ b/drivers/gpu/drm/amd/include/discovery.h
@@ -25,7 +25,6 @@
 #define _DISCOVERY_H_
 
 #define PSP_HEADER_SIZE                 256
-#define BINARY_MAX_SIZE                 (64 << 10)
 #define BINARY_SIGNATURE                0x28211407
 #define DISCOVERY_TABLE_SIGNATURE       0x53445049
 
diff --git a/drivers/gpu/drm/amd/include/ivsrcid/nbio/irqsrcs_nbif_7_4.h b/drivers/gpu/drm/amd/include/ivsrcid/nbio/irqsrcs_nbif_7_4.h
new file mode 100644
index 000000000000..79af4258f259
--- /dev/null
+++ b/drivers/gpu/drm/amd/include/ivsrcid/nbio/irqsrcs_nbif_7_4.h
@@ -0,0 +1,42 @@
+/*
+ * Copyright 2019 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#ifndef __IRQSRCS_NBIF_7_4_H__
+#define __IRQSRCS_NBIF_7_4_H__
+
+#define NBIF_7_4__SRCID__CHIP_ERR_INT_EVENT            0x5E        // Error generated
+#define NBIF_7_4__SRCID__DOORBELL_INTERRUPT            0x5F        // Interrupt for doorbell event during VDDGFX off
+#define NBIF_7_4__SRCID__RAS_CONTROLLER_INTERRUPT      0x60        // Interrupt for ras_intr_valid from RAS controller
+#define NBIF_7_4__SRCID__ERREVENT_ATHUB_INTERRUPT      0x61        // Interrupt for SDP ErrEvent received from ATHUB
+#define NBIF_7_4__SRCID__PF_VF_MSGBUF_VALID            0x87        // Valid message in PF->VF mailbox message buffer (The interrupt is sent on behalf of PF)
+#define NBIF_7_4__SRCID__PF_VF_MSGBUF_ACK              0x88        // Acknowledge message in PF->VF mailbox message buffer (The interrupt is sent on behalf of VF)
+#define NBIF_7_4__SRCID__VF_PF_MSGBUF_VALID            0x89        // Valid message in VF->PF mailbox message buffer (The interrupt is sent on behalf of VF)
+#define NBIF_7_4__SRCID__VF_PF_MSGBUF_ACK              0x8A        // Acknowledge message in VF->PF mailbox message buffer (The interrupt is sent on behalf of PF)
+#define NBIF_7_4__SRCID__CHIP_DPA_INT_EVENT            0xA0        // BIF_CHIP_DPA_INT_EVENT
+#define NBIF_7_4__SRCID__CHIP_SLOT_POWER_CHG_INT_EVENT 0xA1        // BIF_CHIP_SLOT_POWER_CHG_INT_EVENT
+#define NBIF_7_4__SRCID__ATOMIC_UR_OPCODE              0xCE        // BIF receives unsupported atomic opcode from MC
+#define NBIF_7_4__SRCID__ATOMIC_REQESTEREN_LOW         0xCF        // BIF receive atomic request from MC while AtomicOp Requester is not enabled in PCIE config space
+
+#endif // __IRQSRCS_NBIF_7_4_H__
diff --git a/drivers/gpu/drm/amd/include/kgd_kfd_interface.h b/drivers/gpu/drm/amd/include/kgd_kfd_interface.h
index 98b9533e672b..2cd217e60125 100644
--- a/drivers/gpu/drm/amd/include/kgd_kfd_interface.h
+++ b/drivers/gpu/drm/amd/include/kgd_kfd_interface.h
@@ -291,15 +291,18 @@ struct kfd2kgd_calls {
 	uint32_t (*address_watch_get_offset)(struct kgd_dev *kgd,
 					unsigned int watch_point_id,
 					unsigned int reg_offset);
-	bool (*get_atc_vmid_pasid_mapping_valid)(
+	bool (*get_atc_vmid_pasid_mapping_info)(
 					struct kgd_dev *kgd,
-					uint8_t vmid);
-	uint16_t (*get_atc_vmid_pasid_mapping_pasid)(
-					struct kgd_dev *kgd,
-					uint8_t vmid);
+					uint8_t vmid,
+					uint16_t *p_pasid);
 
+	/* No longer needed from GFXv9 onward. The scratch base address is
+	 * passed to the shader by the CP. It's the user mode driver's
+	 * responsibility.
+	 */
 	void (*set_scratch_backing_va)(struct kgd_dev *kgd,
 				uint64_t va, uint32_t vmid);
+
 	int (*get_tile_config)(struct kgd_dev *kgd, struct tile_config *config);
 
 	void (*set_vm_context_page_table_base)(struct kgd_dev *kgd,
diff --git a/drivers/gpu/drm/amd/include/kgd_pp_interface.h b/drivers/gpu/drm/amd/include/kgd_pp_interface.h
index 27cf0afaa0b4..a7f92d0b3a90 100644
--- a/drivers/gpu/drm/amd/include/kgd_pp_interface.h
+++ b/drivers/gpu/drm/amd/include/kgd_pp_interface.h
@@ -179,6 +179,11 @@ enum pp_mp1_state {
 	PP_MP1_STATE_RESET,
 };
 
+enum pp_df_cstate {
+	DF_CSTATE_DISALLOW = 0,
+	DF_CSTATE_ALLOW,
+};
+
 #define PP_GROUP_MASK        0xF0000000
 #define PP_GROUP_SHIFT       28
 
@@ -215,6 +220,9 @@ enum pp_mp1_state {
 		((group) << PP_GROUP_SHIFT | (block) << PP_BLOCK_SHIFT | \
 		(support) << PP_STATE_SUPPORT_SHIFT | (state) << PP_STATE_SHIFT)
 
+#define XGMI_MODE_PSTATE_D3 0
+#define XGMI_MODE_PSTATE_D0 1
+
 struct seq_file;
 enum amd_pp_clock_type;
 struct amd_pp_simple_clock_info;
@@ -312,6 +320,8 @@ struct amd_pm_funcs {
 	int (*get_ppfeature_status)(void *handle, char *buf);
 	int (*set_ppfeature_status)(void *handle, uint64_t ppfeature_masks);
 	int (*asic_reset_mode_2)(void *handle);
+	int (*set_df_cstate)(void *handle, enum pp_df_cstate state);
+	int (*set_xgmi_pstate)(void *handle, uint32_t pstate);
 };
 
 #endif
diff --git a/drivers/gpu/drm/amd/include/renoir_ip_offset.h b/drivers/gpu/drm/amd/include/renoir_ip_offset.h
index 094648cac392..07633e22e99a 100644
--- a/drivers/gpu/drm/amd/include/renoir_ip_offset.h
+++ b/drivers/gpu/drm/amd/include/renoir_ip_offset.h
@@ -169,6 +169,11 @@ static const struct IP_BASE NBIF0_BASE ={ { { { 0x00000000, 0x00000014, 0x00000D
                                         { { 0, 0, 0, 0, 0 } },
                                         { { 0, 0, 0, 0, 0 } },
                                         { { 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE DCN_BASE   ={ { { { 0x00000012, 0x000000C0, 0x000034C0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0 } } } };
 static const struct IP_BASE OSSSYS_BASE ={ { { { 0x000010A0, 0x0240A000, 0, 0, 0 } },
                                         { { 0, 0, 0, 0, 0 } },
                                         { { 0, 0, 0, 0, 0 } },
@@ -1361,4 +1366,33 @@ static const struct IP_BASE UVD0_BASE ={ { { { 0x00007800, 0x00007E00, 0x0240300
 #define UVD0_BASE__INST6_SEG3                      0
 #define UVD0_BASE__INST6_SEG4                      0
 
+#define DCN_BASE__INST0_SEG0                      0x00000012
+#define DCN_BASE__INST0_SEG1                      0x000000C0
+#define DCN_BASE__INST0_SEG2                      0x000034C0
+#define DCN_BASE__INST0_SEG3                      0
+#define DCN_BASE__INST0_SEG4                      0
+
+#define DCN_BASE__INST1_SEG0                      0
+#define DCN_BASE__INST1_SEG1                      0
+#define DCN_BASE__INST1_SEG2                      0
+#define DCN_BASE__INST1_SEG3                      0
+#define DCN_BASE__INST1_SEG4                      0
+
+#define DCN_BASE__INST2_SEG0                      0
+#define DCN_BASE__INST2_SEG1                      0
+#define DCN_BASE__INST2_SEG2                      0
+#define DCN_BASE__INST2_SEG3                      0
+#define DCN_BASE__INST2_SEG4                      0
+
+#define DCN_BASE__INST3_SEG0                      0
+#define DCN_BASE__INST3_SEG1                      0
+#define DCN_BASE__INST3_SEG2                      0
+#define DCN_BASE__INST3_SEG3                      0
+#define DCN_BASE__INST3_SEG4                      0
+
+#define DCN_BASE__INST4_SEG0                      0
+#define DCN_BASE__INST4_SEG1                      0
+#define DCN_BASE__INST4_SEG2                      0
+#define DCN_BASE__INST4_SEG3                      0
+#define DCN_BASE__INST4_SEG4                      0
 #endif
diff --git a/drivers/gpu/drm/amd/include/vega10_enum.h b/drivers/gpu/drm/amd/include/vega10_enum.h
index c14ba65a2415..adf1b754666e 100644
--- a/drivers/gpu/drm/amd/include/vega10_enum.h
+++ b/drivers/gpu/drm/amd/include/vega10_enum.h
@@ -1037,6 +1037,7 @@ TCC_CACHE_POLICY_STREAM                  = 0x00000001,
 typedef enum MTYPE {
 MTYPE_NC                                 = 0x00000000,
 MTYPE_WC                                 = 0x00000001,
+MTYPE_RW                                 = 0x00000001,
 MTYPE_CC                                 = 0x00000002,
 MTYPE_UC                                 = 0x00000003,
 } MTYPE;
diff --git a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
index fa8ad7db2b3a..7932eb163a00 100644
--- a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
+++ b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
@@ -969,6 +969,14 @@ static int pp_dpm_switch_power_profile(void *handle,
 		workload = hwmgr->workload_setting[index];
 	}
 
+	if (type == PP_SMC_POWER_PROFILE_COMPUTE &&
+		hwmgr->hwmgr_func->disable_power_features_for_compute_performance) {
+			if (hwmgr->hwmgr_func->disable_power_features_for_compute_performance(hwmgr, en)) {
+				mutex_unlock(&hwmgr->smu_lock);
+				return -EINVAL;
+			}
+	}
+
 	if (hwmgr->dpm_level != AMD_DPM_FORCED_LEVEL_MANUAL)
 		hwmgr->hwmgr_func->set_power_profile_mode(hwmgr, &workload, 0);
 	mutex_unlock(&hwmgr->smu_lock);
@@ -1421,6 +1429,7 @@ static int pp_get_asic_baco_capability(void *handle, bool *cap)
 {
 	struct pp_hwmgr *hwmgr = handle;
 
+	*cap = false;
 	if (!hwmgr)
 		return -EINVAL;
 
@@ -1548,6 +1557,40 @@ static int pp_smu_i2c_bus_access(void *handle, bool acquire)
 	return ret;
 }
 
+static int pp_set_df_cstate(void *handle, enum pp_df_cstate state)
+{
+	struct pp_hwmgr *hwmgr = handle;
+
+	if (!hwmgr)
+		return -EINVAL;
+
+	if (!hwmgr->pm_en || !hwmgr->hwmgr_func->set_df_cstate)
+		return 0;
+
+	mutex_lock(&hwmgr->smu_lock);
+	hwmgr->hwmgr_func->set_df_cstate(hwmgr, state);
+	mutex_unlock(&hwmgr->smu_lock);
+
+	return 0;
+}
+
+static int pp_set_xgmi_pstate(void *handle, uint32_t pstate)
+{
+	struct pp_hwmgr *hwmgr = handle;
+
+	if (!hwmgr)
+		return -EINVAL;
+
+	if (!hwmgr->pm_en || !hwmgr->hwmgr_func->set_xgmi_pstate)
+		return 0;
+
+	mutex_lock(&hwmgr->smu_lock);
+	hwmgr->hwmgr_func->set_xgmi_pstate(hwmgr, pstate);
+	mutex_unlock(&hwmgr->smu_lock);
+
+	return 0;
+}
+
 static const struct amd_pm_funcs pp_dpm_funcs = {
 	.load_firmware = pp_dpm_load_fw,
 	.wait_for_fw_loading_complete = pp_dpm_fw_loading_complete,
@@ -1606,4 +1649,6 @@ static const struct amd_pm_funcs pp_dpm_funcs = {
 	.set_ppfeature_status = pp_set_ppfeature_status,
 	.asic_reset_mode_2 = pp_asic_reset_mode_2,
 	.smu_i2c_bus_access = pp_smu_i2c_bus_access,
+	.set_df_cstate = pp_set_df_cstate,
+	.set_xgmi_pstate = pp_set_xgmi_pstate,
 };
diff --git a/drivers/gpu/drm/amd/powerplay/amdgpu_smu.c b/drivers/gpu/drm/amd/powerplay/amdgpu_smu.c
index 33960fb38a5d..40b546c75fc2 100644
--- a/drivers/gpu/drm/amd/powerplay/amdgpu_smu.c
+++ b/drivers/gpu/drm/amd/powerplay/amdgpu_smu.c
@@ -25,11 +25,16 @@
 #include "pp_debug.h"
 #include "amdgpu.h"
 #include "amdgpu_smu.h"
+#include "smu_internal.h"
 #include "soc15_common.h"
 #include "smu_v11_0.h"
 #include "smu_v12_0.h"
 #include "atom.h"
 #include "amd_pcie.h"
+#include "vega20_ppt.h"
+#include "arcturus_ppt.h"
+#include "navi10_ppt.h"
+#include "renoir_ppt.h"
 
 #undef __SMU_DUMMY_MAP
 #define __SMU_DUMMY_MAP(type)	#type
@@ -67,6 +72,8 @@ size_t smu_sys_get_pp_feature_mask(struct smu_context *smu, char *buf)
 	uint32_t sort_feature[SMU_FEATURE_COUNT];
 	uint64_t hw_feature_count = 0;
 
+	mutex_lock(&smu->mutex);
+
 	ret = smu_feature_get_enabled_mask(smu, feature_mask, 2);
 	if (ret)
 		goto failed;
@@ -92,9 +99,57 @@ size_t smu_sys_get_pp_feature_mask(struct smu_context *smu, char *buf)
 	}
 
 failed:
+	mutex_unlock(&smu->mutex);
+
 	return size;
 }
 
+static int smu_feature_update_enable_state(struct smu_context *smu,
+					   uint64_t feature_mask,
+					   bool enabled)
+{
+	struct smu_feature *feature = &smu->smu_feature;
+	uint32_t feature_low = 0, feature_high = 0;
+	int ret = 0;
+
+	if (!smu->pm_enabled)
+		return ret;
+
+	feature_low = (feature_mask >> 0 ) & 0xffffffff;
+	feature_high = (feature_mask >> 32) & 0xffffffff;
+
+	if (enabled) {
+		ret = smu_send_smc_msg_with_param(smu, SMU_MSG_EnableSmuFeaturesLow,
+						  feature_low);
+		if (ret)
+			return ret;
+		ret = smu_send_smc_msg_with_param(smu, SMU_MSG_EnableSmuFeaturesHigh,
+						  feature_high);
+		if (ret)
+			return ret;
+	} else {
+		ret = smu_send_smc_msg_with_param(smu, SMU_MSG_DisableSmuFeaturesLow,
+						  feature_low);
+		if (ret)
+			return ret;
+		ret = smu_send_smc_msg_with_param(smu, SMU_MSG_DisableSmuFeaturesHigh,
+						  feature_high);
+		if (ret)
+			return ret;
+	}
+
+	mutex_lock(&feature->mutex);
+	if (enabled)
+		bitmap_or(feature->enabled, feature->enabled,
+				(unsigned long *)(&feature_mask), SMU_FEATURE_MAX);
+	else
+		bitmap_andnot(feature->enabled, feature->enabled,
+				(unsigned long *)(&feature_mask), SMU_FEATURE_MAX);
+	mutex_unlock(&feature->mutex);
+
+	return ret;
+}
+
 int smu_sys_set_pp_feature_mask(struct smu_context *smu, uint64_t new_mask)
 {
 	int ret = 0;
@@ -103,9 +158,11 @@ int smu_sys_set_pp_feature_mask(struct smu_context *smu, uint64_t new_mask)
 	uint64_t feature_2_disabled = 0;
 	uint64_t feature_enables = 0;
 
+	mutex_lock(&smu->mutex);
+
 	ret = smu_feature_get_enabled_mask(smu, feature_mask, 2);
 	if (ret)
-		return ret;
+		goto out;
 
 	feature_enables = ((uint64_t)feature_mask[1] << 32 | (uint64_t)feature_mask[0]);
 
@@ -115,14 +172,17 @@ int smu_sys_set_pp_feature_mask(struct smu_context *smu, uint64_t new_mask)
 	if (feature_2_enabled) {
 		ret = smu_feature_update_enable_state(smu, feature_2_enabled, true);
 		if (ret)
-			return ret;
+			goto out;
 	}
 	if (feature_2_disabled) {
 		ret = smu_feature_update_enable_state(smu, feature_2_disabled, false);
 		if (ret)
-			return ret;
+			goto out;
 	}
 
+out:
+	mutex_unlock(&smu->mutex);
+
 	return ret;
 }
 
@@ -159,8 +219,7 @@ int smu_get_smc_version(struct smu_context *smu, uint32_t *if_version, uint32_t
 int smu_set_soft_freq_range(struct smu_context *smu, enum smu_clk_type clk_type,
 			    uint32_t min, uint32_t max)
 {
-	int ret = 0, clk_id = 0;
-	uint32_t param;
+	int ret = 0;
 
 	if (min <= 0 && max <= 0)
 		return -EINVAL;
@@ -168,27 +227,7 @@ int smu_set_soft_freq_range(struct smu_context *smu, enum smu_clk_type clk_type,
 	if (!smu_clk_dpm_is_enabled(smu, clk_type))
 		return 0;
 
-	clk_id = smu_clk_get_index(smu, clk_type);
-	if (clk_id < 0)
-		return clk_id;
-
-	if (max > 0) {
-		param = (uint32_t)((clk_id << 16) | (max & 0xffff));
-		ret = smu_send_smc_msg_with_param(smu, SMU_MSG_SetSoftMaxByFreq,
-						  param);
-		if (ret)
-			return ret;
-	}
-
-	if (min > 0) {
-		param = (uint32_t)((clk_id << 16) | (min & 0xffff));
-		ret = smu_send_smc_msg_with_param(smu, SMU_MSG_SetSoftMinByFreq,
-						  param);
-		if (ret)
-			return ret;
-	}
-
-
+	ret = smu_set_soft_freq_limited_range(smu, clk_type, min, max);
 	return ret;
 }
 
@@ -229,7 +268,7 @@ int smu_set_hard_freq_range(struct smu_context *smu, enum smu_clk_type clk_type,
 }
 
 int smu_get_dpm_freq_range(struct smu_context *smu, enum smu_clk_type clk_type,
-			   uint32_t *min, uint32_t *max)
+			   uint32_t *min, uint32_t *max, bool lock_needed)
 {
 	uint32_t clock_limit;
 	int ret = 0;
@@ -237,6 +276,9 @@ int smu_get_dpm_freq_range(struct smu_context *smu, enum smu_clk_type clk_type,
 	if (!min && !max)
 		return -EINVAL;
 
+	if (lock_needed)
+		mutex_lock(&smu->mutex);
+
 	if (!smu_clk_dpm_is_enabled(smu, clk_type)) {
 		switch (clk_type) {
 		case SMU_MCLK:
@@ -260,14 +302,17 @@ int smu_get_dpm_freq_range(struct smu_context *smu, enum smu_clk_type clk_type,
 			*min = clock_limit / 100;
 		if (max)
 			*max = clock_limit / 100;
-
-		return 0;
+	} else {
+		/*
+		 * Todo: Use each asic(ASIC_ppt funcs) control the callbacks exposed to the
+		 * core driver and then have helpers for stuff that is common(SMU_v11_x | SMU_v12_x funcs).
+		 */
+		ret = smu_get_dpm_ultimate_freq(smu, clk_type, min, max);
 	}
-	/*
-	 * Todo: Use each asic(ASIC_ppt funcs) control the callbacks exposed to the
-	 * core driver and then have helpers for stuff that is common(SMU_v11_x | SMU_v12_x funcs).
-	 */
-	ret = smu_get_dpm_ultimate_freq(smu, clk_type, min, max);
+
+	if (lock_needed)
+		mutex_unlock(&smu->mutex);
+
 	return ret;
 }
 
@@ -338,7 +383,20 @@ bool smu_clk_dpm_is_enabled(struct smu_context *smu, enum smu_clk_type clk_type)
 	return true;
 }
 
-
+/**
+ * smu_dpm_set_power_gate - power gate/ungate the specific IP block
+ *
+ * @smu:        smu_context pointer
+ * @block_type: the IP block to power gate/ungate
+ * @gate:       to power gate if true, ungate otherwise
+ *
+ * This API uses no smu->mutex lock protection due to:
+ * 1. It is either called by other IP block(gfx/sdma/vcn/uvd/vce).
+ *    This is guarded to be race condition free by the caller.
+ * 2. Or get called on user setting request of power_dpm_force_performance_level.
+ *    Under this case, the smu->mutex lock protection is already enforced on
+ *    the parent API smu_force_performance_level of the call path.
+ */
 int smu_dpm_set_power_gate(struct smu_context *smu, uint32_t block_type,
 			   bool gate)
 {
@@ -364,12 +422,6 @@ int smu_dpm_set_power_gate(struct smu_context *smu, uint32_t block_type,
 	return ret;
 }
 
-enum amd_pm_state_type smu_get_current_power_state(struct smu_context *smu)
-{
-	/* not support power state */
-	return POWER_STATE_TYPE_DEFAULT;
-}
-
 int smu_get_power_num_states(struct smu_context *smu,
 			     struct pp_states_info *state_info)
 {
@@ -439,7 +491,7 @@ int smu_update_table(struct smu_context *smu, enum smu_table_id table_index, int
 	int ret = 0;
 	int table_id = smu_table_get_index(smu, table_index);
 
-	if (!table_data || table_id >= smu_table->table_count || table_id < 0)
+	if (!table_data || table_id >= SMU_TABLE_COUNT || table_id < 0)
 		return -EINVAL;
 
 	table = &smu_table->tables[table_index];
@@ -463,7 +515,7 @@ int smu_update_table(struct smu_context *smu, enum smu_table_id table_index, int
 		return ret;
 
 	/* flush hdp cache */
-	adev->nbio_funcs->hdp_flush(adev, NULL);
+	adev->nbio.funcs->hdp_flush(adev, NULL);
 
 	if (!drv2smu)
 		memcpy(table_data, table->cpu_addr, table->size);
@@ -483,7 +535,7 @@ bool is_support_sw_smu(struct amdgpu_device *adev)
 
 bool is_support_sw_smu_xgmi(struct amdgpu_device *adev)
 {
-	if (amdgpu_dpm != 1)
+	if (!is_support_sw_smu(adev))
 		return false;
 
 	if (adev->asic_type == CHIP_VEGA20)
@@ -495,16 +547,23 @@ bool is_support_sw_smu_xgmi(struct amdgpu_device *adev)
 int smu_sys_get_pp_table(struct smu_context *smu, void **table)
 {
 	struct smu_table_context *smu_table = &smu->smu_table;
+	uint32_t powerplay_table_size;
 
 	if (!smu_table->power_play_table && !smu_table->hardcode_pptable)
 		return -EINVAL;
 
+	mutex_lock(&smu->mutex);
+
 	if (smu_table->hardcode_pptable)
 		*table = smu_table->hardcode_pptable;
 	else
 		*table = smu_table->power_play_table;
 
-	return smu_table->power_play_table_size;
+	powerplay_table_size = smu_table->power_play_table_size;
+
+	mutex_unlock(&smu->mutex);
+
+	return powerplay_table_size;
 }
 
 int smu_sys_set_pp_table(struct smu_context *smu,  void *buf, size_t size)
@@ -531,13 +590,18 @@ int smu_sys_set_pp_table(struct smu_context *smu,  void *buf, size_t size)
 	memcpy(smu_table->hardcode_pptable, buf, size);
 	smu_table->power_play_table = smu_table->hardcode_pptable;
 	smu_table->power_play_table_size = size;
-	mutex_unlock(&smu->mutex);
+
+	/*
+	 * Special hw_fini action(for Navi1x, the DPMs disablement will be
+	 * skipped) may be needed for custom pptable uploading.
+	 */
+	smu->uploading_custom_pp_table = true;
 
 	ret = smu_reset(smu);
 	if (ret)
 		pr_info("smu reset failed, ret = %d\n", ret);
 
-	return ret;
+	smu->uploading_custom_pp_table = false;
 
 failed:
 	mutex_unlock(&smu->mutex);
@@ -569,41 +633,7 @@ int smu_feature_init_dpm(struct smu_context *smu)
 
 	return ret;
 }
-int smu_feature_update_enable_state(struct smu_context *smu, uint64_t feature_mask, bool enabled)
-{
-	uint32_t feature_low = 0, feature_high = 0;
-	int ret = 0;
-
-	if (!smu->pm_enabled)
-		return ret;
 
-	feature_low = (feature_mask >> 0 ) & 0xffffffff;
-	feature_high = (feature_mask >> 32) & 0xffffffff;
-
-	if (enabled) {
-		ret = smu_send_smc_msg_with_param(smu, SMU_MSG_EnableSmuFeaturesLow,
-						  feature_low);
-		if (ret)
-			return ret;
-		ret = smu_send_smc_msg_with_param(smu, SMU_MSG_EnableSmuFeaturesHigh,
-						  feature_high);
-		if (ret)
-			return ret;
-
-	} else {
-		ret = smu_send_smc_msg_with_param(smu, SMU_MSG_DisableSmuFeaturesLow,
-						  feature_low);
-		if (ret)
-			return ret;
-		ret = smu_send_smc_msg_with_param(smu, SMU_MSG_DisableSmuFeaturesHigh,
-						  feature_high);
-		if (ret)
-			return ret;
-
-	}
-
-	return ret;
-}
 
 int smu_feature_is_enabled(struct smu_context *smu, enum smu_feature_mask mask)
 {
@@ -633,8 +663,6 @@ int smu_feature_set_enabled(struct smu_context *smu, enum smu_feature_mask mask,
 {
 	struct smu_feature *feature = &smu->smu_feature;
 	int feature_id;
-	uint64_t feature_mask = 0;
-	int ret = 0;
 
 	feature_id = smu_feature_get_index(smu, mask);
 	if (feature_id < 0)
@@ -642,22 +670,9 @@ int smu_feature_set_enabled(struct smu_context *smu, enum smu_feature_mask mask,
 
 	WARN_ON(feature_id > feature->feature_num);
 
-	feature_mask = 1ULL << feature_id;
-
-	mutex_lock(&feature->mutex);
-	ret = smu_feature_update_enable_state(smu, feature_mask, enable);
-	if (ret)
-		goto failed;
-
-	if (enable)
-		test_and_set_bit(feature_id, feature->enabled);
-	else
-		test_and_clear_bit(feature_id, feature->enabled);
-
-failed:
-	mutex_unlock(&feature->mutex);
-
-	return ret;
+	return smu_feature_update_enable_state(smu,
+					       1ULL << feature_id,
+					       enable);
 }
 
 int smu_feature_is_supported(struct smu_context *smu, enum smu_feature_mask mask)
@@ -707,20 +722,27 @@ static int smu_set_funcs(struct amdgpu_device *adev)
 {
 	struct smu_context *smu = &adev->smu;
 
+	if (adev->pm.pp_feature & PP_OVERDRIVE_MASK)
+		smu->od_enabled = true;
+
 	switch (adev->asic_type) {
 	case CHIP_VEGA20:
+		adev->pm.pp_feature &= ~PP_GFXOFF_MASK;
+		vega20_set_ppt_funcs(smu);
+		break;
 	case CHIP_NAVI10:
 	case CHIP_NAVI14:
 	case CHIP_NAVI12:
+		navi10_set_ppt_funcs(smu);
+		break;
 	case CHIP_ARCTURUS:
-		if (adev->pm.pp_feature & PP_OVERDRIVE_MASK)
-			smu->od_enabled = true;
-		smu_v11_0_set_smu_funcs(smu);
+		adev->pm.pp_feature &= ~PP_GFXOFF_MASK;
+		arcturus_set_ppt_funcs(smu);
+		/* OD is not supported on Arcturus */
+		smu->od_enabled =false;
 		break;
 	case CHIP_RENOIR:
-		if (adev->pm.pp_feature & PP_OVERDRIVE_MASK)
-			smu->od_enabled = true;
-		smu_v12_0_set_smu_funcs(smu);
+		renoir_set_ppt_funcs(smu);
 		break;
 	default:
 		return -EINVAL;
@@ -736,6 +758,7 @@ static int smu_early_init(void *handle)
 
 	smu->adev = adev;
 	smu->pm_enabled = !!amdgpu_dpm;
+	smu->is_apu = false;
 	mutex_init(&smu->mutex);
 
 	return smu_set_funcs(adev);
@@ -749,11 +772,10 @@ static int smu_late_init(void *handle)
 	if (!smu->pm_enabled)
 		return 0;
 
-	mutex_lock(&smu->mutex);
 	smu_handle_task(&adev->smu,
 			smu->smu_dpm.dpm_level,
-			AMD_PP_TASK_COMPLETE_INIT);
-	mutex_unlock(&smu->mutex);
+			AMD_PP_TASK_COMPLETE_INIT,
+			false);
 
 	return 0;
 }
@@ -843,6 +865,8 @@ static int smu_sw_init(void *handle)
 	smu->smu_baco.state = SMU_BACO_STATE_EXIT;
 	smu->smu_baco.platform_support = false;
 
+	mutex_init(&smu->sensor_lock);
+
 	smu->watermarks_bitmap = 0;
 	smu->power_profile_mode = PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT;
 	smu->default_power_profile_mode = PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT;
@@ -917,14 +941,9 @@ static int smu_init_fb_allocations(struct smu_context *smu)
 	struct amdgpu_device *adev = smu->adev;
 	struct smu_table_context *smu_table = &smu->smu_table;
 	struct smu_table *tables = smu_table->tables;
-	uint32_t table_count = smu_table->table_count;
-	uint32_t i = 0;
-	int32_t ret = 0;
+	int ret, i;
 
-	if (table_count <= 0)
-		return -EINVAL;
-
-	for (i = 0 ; i < table_count; i++) {
+	for (i = 0; i < SMU_TABLE_COUNT; i++) {
 		if (tables[i].size == 0)
 			continue;
 		ret = amdgpu_bo_create_kernel(adev,
@@ -940,7 +959,7 @@ static int smu_init_fb_allocations(struct smu_context *smu)
 
 	return 0;
 failed:
-	for (; i > 0; i--) {
+	while (--i >= 0) {
 		if (tables[i].size == 0)
 			continue;
 		amdgpu_bo_free_kernel(&tables[i].bo,
@@ -955,13 +974,12 @@ static int smu_fini_fb_allocations(struct smu_context *smu)
 {
 	struct smu_table_context *smu_table = &smu->smu_table;
 	struct smu_table *tables = smu_table->tables;
-	uint32_t table_count = smu_table->table_count;
 	uint32_t i = 0;
 
-	if (table_count == 0 || tables == NULL)
+	if (!tables)
 		return 0;
 
-	for (i = 0 ; i < table_count; i++) {
+	for (i = 0; i < SMU_TABLE_COUNT; i++) {
 		if (tables[i].size == 0)
 			continue;
 		amdgpu_bo_free_kernel(&tables[i].bo,
@@ -972,50 +990,6 @@ static int smu_fini_fb_allocations(struct smu_context *smu)
 	return 0;
 }
 
-static int smu_override_pcie_parameters(struct smu_context *smu)
-{
-	struct amdgpu_device *adev = smu->adev;
-	uint32_t pcie_gen = 0, pcie_width = 0, smu_pcie_arg;
-	int ret;
-
-	if (adev->flags & AMD_IS_APU)
-		return 0;
-
-	if (adev->pm.pcie_gen_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_GEN4)
-		pcie_gen = 3;
-	else if (adev->pm.pcie_gen_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_GEN3)
-		pcie_gen = 2;
-	else if (adev->pm.pcie_gen_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_GEN2)
-		pcie_gen = 1;
-	else if (adev->pm.pcie_gen_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_GEN1)
-		pcie_gen = 0;
-
-	/* Bit 31:16: LCLK DPM level. 0 is DPM0, and 1 is DPM1
-	 * Bit 15:8:  PCIE GEN, 0 to 3 corresponds to GEN1 to GEN4
-	 * Bit 7:0:   PCIE lane width, 1 to 7 corresponds is x1 to x32
-	 */
-	if (adev->pm.pcie_mlw_mask & CAIL_PCIE_LINK_WIDTH_SUPPORT_X16)
-		pcie_width = 6;
-	else if (adev->pm.pcie_mlw_mask & CAIL_PCIE_LINK_WIDTH_SUPPORT_X12)
-		pcie_width = 5;
-	else if (adev->pm.pcie_mlw_mask & CAIL_PCIE_LINK_WIDTH_SUPPORT_X8)
-		pcie_width = 4;
-	else if (adev->pm.pcie_mlw_mask & CAIL_PCIE_LINK_WIDTH_SUPPORT_X4)
-		pcie_width = 3;
-	else if (adev->pm.pcie_mlw_mask & CAIL_PCIE_LINK_WIDTH_SUPPORT_X2)
-		pcie_width = 2;
-	else if (adev->pm.pcie_mlw_mask & CAIL_PCIE_LINK_WIDTH_SUPPORT_X1)
-		pcie_width = 1;
-
-	smu_pcie_arg = (1 << 16) | (pcie_gen << 8) | pcie_width;
-	ret = smu_send_smc_msg_with_param(smu,
-					  SMU_MSG_OverridePcieParameters,
-					  smu_pcie_arg);
-	if (ret)
-		pr_err("[%s] Attempt to override pcie params failed!\n", __func__);
-	return ret;
-}
-
 static int smu_smc_table_hw_init(struct smu_context *smu,
 				 bool initialize)
 {
@@ -1090,8 +1064,8 @@ static int smu_smc_table_hw_init(struct smu_context *smu,
 	if (ret)
 		return ret;
 
-	/* issue RunAfllBtc msg */
-	ret = smu_run_afll_btc(smu);
+	/* issue Run*Btc msg */
+	ret = smu_run_btc(smu);
 	if (ret)
 		return ret;
 
@@ -1104,10 +1078,6 @@ static int smu_smc_table_hw_init(struct smu_context *smu,
 		return ret;
 
 	if (adev->asic_type != CHIP_ARCTURUS) {
-		ret = smu_override_pcie_parameters(smu);
-		if (ret)
-			return ret;
-
 		ret = smu_notify_display_change(smu);
 		if (ret)
 			return ret;
@@ -1136,6 +1106,12 @@ static int smu_smc_table_hw_init(struct smu_context *smu,
 			return ret;
 	}
 
+	if (adev->asic_type != CHIP_ARCTURUS) {
+		ret = smu_override_pcie_parameters(smu);
+		if (ret)
+			return ret;
+	}
+
 	ret = smu_set_default_od_settings(smu, initialize);
 	if (ret)
 		return ret;
@@ -1145,7 +1121,7 @@ static int smu_smc_table_hw_init(struct smu_context *smu,
 		if (ret)
 			return ret;
 
-		ret = smu_get_power_limit(smu, &smu->default_power_limit, true);
+		ret = smu_get_power_limit(smu, &smu->default_power_limit, false, false);
 		if (ret)
 			return ret;
 	}
@@ -1224,29 +1200,46 @@ static int smu_free_memory_pool(struct smu_context *smu)
 	return ret;
 }
 
-static int smu_hw_init(void *handle)
+static int smu_start_smc_engine(struct smu_context *smu)
 {
-	int ret;
-	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-	struct smu_context *smu = &adev->smu;
+	struct amdgpu_device *adev = smu->adev;
+	int ret = 0;
 
 	if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP) {
 		if (adev->asic_type < CHIP_NAVI10) {
-			ret = smu_load_microcode(smu);
-			if (ret)
-				return ret;
+			if (smu->ppt_funcs->load_microcode) {
+				ret = smu->ppt_funcs->load_microcode(smu);
+				if (ret)
+					return ret;
+			}
 		}
 	}
 
-	ret = smu_check_fw_status(smu);
+	if (smu->ppt_funcs->check_fw_status) {
+		ret = smu->ppt_funcs->check_fw_status(smu);
+		if (ret)
+			pr_err("SMC is not ready\n");
+	}
+
+	return ret;
+}
+
+static int smu_hw_init(void *handle)
+{
+	int ret;
+	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+	struct smu_context *smu = &adev->smu;
+
+	ret = smu_start_smc_engine(smu);
 	if (ret) {
-		pr_err("SMC firmware status is not correct\n");
+		pr_err("SMU is not ready yet!\n");
 		return ret;
 	}
 
 	if (adev->flags & AMD_IS_APU) {
 		smu_powergate_sdma(&adev->smu, false);
 		smu_powergate_vcn(&adev->smu, false);
+		smu_set_gfx_cgpg(&adev->smu, true);
 	}
 
 	if (!smu->pm_enabled)
@@ -1289,6 +1282,11 @@ failed:
 	return ret;
 }
 
+static int smu_stop_dpms(struct smu_context *smu)
+{
+	return smu_send_smc_msg(smu, SMU_MSG_DisableAllSmuFeatures);
+}
+
 static int smu_hw_fini(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
@@ -1301,6 +1299,33 @@ static int smu_hw_fini(void *handle)
 		smu_powergate_vcn(&adev->smu, true);
 	}
 
+	ret = smu_stop_thermal_control(smu);
+	if (ret) {
+		pr_warn("Fail to stop thermal control!\n");
+		return ret;
+	}
+
+	/*
+	 * For custom pptable uploading, skip the DPM features
+	 * disable process on Navi1x ASICs.
+	 *   - As the gfx related features are under control of
+	 *     RLC on those ASICs. RLC reinitialization will be
+	 *     needed to reenable them. That will cost much more
+	 *     efforts.
+	 *
+	 *   - SMU firmware can handle the DPM reenablement
+	 *     properly.
+	 */
+	if (!smu->uploading_custom_pp_table ||
+	    !((adev->asic_type >= CHIP_NAVI10) &&
+	      (adev->asic_type <= CHIP_NAVI12))) {
+		ret = smu_stop_dpms(smu);
+		if (ret) {
+			pr_warn("Fail to stop Dpms!\n");
+			return ret;
+		}
+	}
+
 	kfree(table_context->driver_pptable);
 	table_context->driver_pptable = NULL;
 
@@ -1342,7 +1367,10 @@ static int smu_suspend(void *handle)
 	int ret;
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 	struct smu_context *smu = &adev->smu;
-	bool baco_feature_is_enabled = smu_feature_is_enabled(smu, SMU_FEATURE_BACO_BIT);
+	bool baco_feature_is_enabled = false;
+
+	if(!(adev->flags & AMD_IS_APU))
+		baco_feature_is_enabled = smu_feature_is_enabled(smu, SMU_FEATURE_BACO_BIT);
 
 	ret = smu_system_features_control(smu, false);
 	if (ret)
@@ -1361,6 +1389,8 @@ static int smu_suspend(void *handle)
 	if (adev->asic_type >= CHIP_NAVI10 &&
 	    adev->gfx.rlc.funcs->stop)
 		adev->gfx.rlc.funcs->stop(adev);
+	if (smu->is_apu)
+		smu_set_gfx_cgpg(&adev->smu, false);
 
 	return 0;
 }
@@ -1373,7 +1403,11 @@ static int smu_resume(void *handle)
 
 	pr_info("SMU is resuming...\n");
 
-	mutex_lock(&smu->mutex);
+	ret = smu_start_smc_engine(smu);
+	if (ret) {
+		pr_err("SMU is not ready yet!\n");
+		goto failed;
+	}
 
 	ret = smu_smc_table_hw_init(smu, false);
 	if (ret)
@@ -1383,13 +1417,16 @@ static int smu_resume(void *handle)
 	if (ret)
 		goto failed;
 
-	mutex_unlock(&smu->mutex);
+	if (smu->is_apu)
+		smu_set_gfx_cgpg(&adev->smu, true);
+
+	smu->disable_uclk_switch = 0;
 
 	pr_info("SMU is resumed successfully!\n");
 
 	return 0;
+
 failed:
-	mutex_unlock(&smu->mutex);
 	return ret;
 }
 
@@ -1407,8 +1444,9 @@ int smu_display_configuration_change(struct smu_context *smu,
 
 	mutex_lock(&smu->mutex);
 
-	smu_set_deep_sleep_dcefclk(smu,
-				   display_config->min_dcef_deep_sleep_set_clk / 100);
+	if (smu->ppt_funcs->set_deep_sleep_dcefclk)
+		smu->ppt_funcs->set_deep_sleep_dcefclk(smu,
+				display_config->min_dcef_deep_sleep_set_clk / 100);
 
 	for (index = 0; index < display_config->num_path_including_non_display; index++) {
 		if (display_config->displays[index].controller_id != 0)
@@ -1527,7 +1565,8 @@ static int smu_enable_umd_pstate(void *handle,
 
 	struct smu_context *smu = (struct smu_context*)(handle);
 	struct smu_dpm_context *smu_dpm_ctx = &(smu->smu_dpm);
-	if (!smu->pm_enabled || !smu_dpm_ctx->dpm_context)
+
+	if (!smu->is_apu && (!smu->pm_enabled || !smu_dpm_ctx->dpm_context))
 		return -EINVAL;
 
 	if (!(smu_dpm_ctx->dpm_level & profile_mode_mask)) {
@@ -1585,9 +1624,9 @@ static int smu_default_set_performance_level(struct smu_context *smu, enum amd_d
 						 &soc_mask);
 		if (ret)
 			return ret;
-		smu_force_clk_levels(smu, SMU_SCLK, 1 << sclk_mask);
-		smu_force_clk_levels(smu, SMU_MCLK, 1 << mclk_mask);
-		smu_force_clk_levels(smu, SMU_SOCCLK, 1 << soc_mask);
+		smu_force_clk_levels(smu, SMU_SCLK, 1 << sclk_mask, false);
+		smu_force_clk_levels(smu, SMU_MCLK, 1 << mclk_mask, false);
+		smu_force_clk_levels(smu, SMU_SOCCLK, 1 << soc_mask, false);
 		break;
 	case AMD_DPM_FORCED_LEVEL_MANUAL:
 	case AMD_DPM_FORCED_LEVEL_PROFILE_EXIT:
@@ -1651,7 +1690,7 @@ int smu_adjust_power_state_dynamic(struct smu_context *smu,
 		workload = smu->workload_setting[index];
 
 		if (smu->power_profile_mode != workload)
-			smu_set_power_profile_mode(smu, &workload, 0);
+			smu_set_power_profile_mode(smu, &workload, 0, false);
 	}
 
 	return ret;
@@ -1659,18 +1698,22 @@ int smu_adjust_power_state_dynamic(struct smu_context *smu,
 
 int smu_handle_task(struct smu_context *smu,
 		    enum amd_dpm_forced_level level,
-		    enum amd_pp_task task_id)
+		    enum amd_pp_task task_id,
+		    bool lock_needed)
 {
 	int ret = 0;
 
+	if (lock_needed)
+		mutex_lock(&smu->mutex);
+
 	switch (task_id) {
 	case AMD_PP_TASK_DISPLAY_CONFIG_CHANGE:
 		ret = smu_pre_display_config_changed(smu);
 		if (ret)
-			return ret;
+			goto out;
 		ret = smu_set_cpu_power_state(smu);
 		if (ret)
-			return ret;
+			goto out;
 		ret = smu_adjust_power_state_dynamic(smu, level, false);
 		break;
 	case AMD_PP_TASK_COMPLETE_INIT:
@@ -1681,6 +1724,10 @@ int smu_handle_task(struct smu_context *smu,
 		break;
 	}
 
+out:
+	if (lock_needed)
+		mutex_unlock(&smu->mutex);
+
 	return ret;
 }
 
@@ -1713,7 +1760,7 @@ int smu_switch_power_profile(struct smu_context *smu,
 	}
 
 	if (smu_dpm_ctx->dpm_level != AMD_DPM_FORCED_LEVEL_MANUAL)
-		smu_set_power_profile_mode(smu, &workload, 0);
+		smu_set_power_profile_mode(smu, &workload, 0, false);
 
 	mutex_unlock(&smu->mutex);
 
@@ -1725,7 +1772,7 @@ enum amd_dpm_forced_level smu_get_performance_level(struct smu_context *smu)
 	struct smu_dpm_context *smu_dpm_ctx = &(smu->smu_dpm);
 	enum amd_dpm_forced_level level;
 
-	if (!smu_dpm_ctx->dpm_context)
+	if (!smu->is_apu && !smu_dpm_ctx->dpm_context)
 		return -EINVAL;
 
 	mutex_lock(&(smu->mutex));
@@ -1740,15 +1787,22 @@ int smu_force_performance_level(struct smu_context *smu, enum amd_dpm_forced_lev
 	struct smu_dpm_context *smu_dpm_ctx = &(smu->smu_dpm);
 	int ret = 0;
 
-	if (!smu_dpm_ctx->dpm_context)
+	if (!smu->is_apu && !smu_dpm_ctx->dpm_context)
 		return -EINVAL;
 
+	mutex_lock(&smu->mutex);
+
 	ret = smu_enable_umd_pstate(smu, &level);
-	if (ret)
+	if (ret) {
+		mutex_unlock(&smu->mutex);
 		return ret;
+	}
 
 	ret = smu_handle_task(smu, level,
-			      AMD_PP_TASK_READJUST_POWER_STATE);
+			      AMD_PP_TASK_READJUST_POWER_STATE,
+			      false);
+
+	mutex_unlock(&smu->mutex);
 
 	return ret;
 }
@@ -1764,6 +1818,144 @@ int smu_set_display_count(struct smu_context *smu, uint32_t count)
 	return ret;
 }
 
+int smu_force_clk_levels(struct smu_context *smu,
+			 enum smu_clk_type clk_type,
+			 uint32_t mask,
+			 bool lock_needed)
+{
+	struct smu_dpm_context *smu_dpm_ctx = &(smu->smu_dpm);
+	int ret = 0;
+
+	if (smu_dpm_ctx->dpm_level != AMD_DPM_FORCED_LEVEL_MANUAL) {
+		pr_debug("force clock level is for dpm manual mode only.\n");
+		return -EINVAL;
+	}
+
+	if (lock_needed)
+		mutex_lock(&smu->mutex);
+
+	if (smu->ppt_funcs && smu->ppt_funcs->force_clk_levels)
+		ret = smu->ppt_funcs->force_clk_levels(smu, clk_type, mask);
+
+	if (lock_needed)
+		mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+int smu_set_mp1_state(struct smu_context *smu,
+		      enum pp_mp1_state mp1_state)
+{
+	uint16_t msg;
+	int ret;
+
+	/*
+	 * The SMC is not fully ready. That may be
+	 * expected as the IP may be masked.
+	 * So, just return without error.
+	 */
+	if (!smu->pm_enabled)
+		return 0;
+
+	mutex_lock(&smu->mutex);
+
+	switch (mp1_state) {
+	case PP_MP1_STATE_SHUTDOWN:
+		msg = SMU_MSG_PrepareMp1ForShutdown;
+		break;
+	case PP_MP1_STATE_UNLOAD:
+		msg = SMU_MSG_PrepareMp1ForUnload;
+		break;
+	case PP_MP1_STATE_RESET:
+		msg = SMU_MSG_PrepareMp1ForReset;
+		break;
+	case PP_MP1_STATE_NONE:
+	default:
+		mutex_unlock(&smu->mutex);
+		return 0;
+	}
+
+	/* some asics may not support those messages */
+	if (smu_msg_get_index(smu, msg) < 0) {
+		mutex_unlock(&smu->mutex);
+		return 0;
+	}
+
+	ret = smu_send_smc_msg(smu, msg);
+	if (ret)
+		pr_err("[PrepareMp1] Failed!\n");
+
+	mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+int smu_set_df_cstate(struct smu_context *smu,
+		      enum pp_df_cstate state)
+{
+	int ret = 0;
+
+	/*
+	 * The SMC is not fully ready. That may be
+	 * expected as the IP may be masked.
+	 * So, just return without error.
+	 */
+	if (!smu->pm_enabled)
+		return 0;
+
+	if (!smu->ppt_funcs || !smu->ppt_funcs->set_df_cstate)
+		return 0;
+
+	mutex_lock(&smu->mutex);
+
+	ret = smu->ppt_funcs->set_df_cstate(smu, state);
+	if (ret)
+		pr_err("[SetDfCstate] failed!\n");
+
+	mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+int smu_write_watermarks_table(struct smu_context *smu)
+{
+	int ret = 0;
+	struct smu_table_context *smu_table = &smu->smu_table;
+	struct smu_table *table = NULL;
+
+	table = &smu_table->tables[SMU_TABLE_WATERMARKS];
+
+	if (!table->cpu_addr)
+		return -EINVAL;
+
+	ret = smu_update_table(smu, SMU_TABLE_WATERMARKS, 0, table->cpu_addr,
+				true);
+
+	return ret;
+}
+
+int smu_set_watermarks_for_clock_ranges(struct smu_context *smu,
+		struct dm_pp_wm_sets_with_clock_ranges_soc15 *clock_ranges)
+{
+	int ret = 0;
+	struct smu_table *watermarks = &smu->smu_table.tables[SMU_TABLE_WATERMARKS];
+	void *table = watermarks->cpu_addr;
+
+	mutex_lock(&smu->mutex);
+
+	if (!smu->disable_watermark &&
+			smu_feature_is_enabled(smu, SMU_FEATURE_DPM_DCEFCLK_BIT) &&
+			smu_feature_is_enabled(smu, SMU_FEATURE_DPM_SOCCLK_BIT)) {
+		smu_set_watermarks_table(smu, table, clock_ranges);
+		smu->watermarks_bitmap |= WATERMARKS_EXIST;
+		smu->watermarks_bitmap &= ~WATERMARKS_LOADED;
+	}
+
+	mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
 const struct amd_ip_funcs smu_ip_funcs = {
 	.name = "smu",
 	.early_init = smu_early_init,
@@ -1800,3 +1992,559 @@ const struct amdgpu_ip_block_version smu_v12_0_ip_block =
 	.rev = 0,
 	.funcs = &smu_ip_funcs,
 };
+
+int smu_load_microcode(struct smu_context *smu)
+{
+	int ret = 0;
+
+	mutex_lock(&smu->mutex);
+
+	if (smu->ppt_funcs->load_microcode)
+		ret = smu->ppt_funcs->load_microcode(smu);
+
+	mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+int smu_check_fw_status(struct smu_context *smu)
+{
+	int ret = 0;
+
+	mutex_lock(&smu->mutex);
+
+	if (smu->ppt_funcs->check_fw_status)
+		ret = smu->ppt_funcs->check_fw_status(smu);
+
+	mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+int smu_set_gfx_cgpg(struct smu_context *smu, bool enabled)
+{
+	int ret = 0;
+
+	mutex_lock(&smu->mutex);
+
+	if (smu->ppt_funcs->set_gfx_cgpg)
+		ret = smu->ppt_funcs->set_gfx_cgpg(smu, enabled);
+
+	mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+int smu_set_fan_speed_rpm(struct smu_context *smu, uint32_t speed)
+{
+	int ret = 0;
+
+	mutex_lock(&smu->mutex);
+
+	if (smu->ppt_funcs->set_fan_speed_rpm)
+		ret = smu->ppt_funcs->set_fan_speed_rpm(smu, speed);
+
+	mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+int smu_get_power_limit(struct smu_context *smu,
+			uint32_t *limit,
+			bool def,
+			bool lock_needed)
+{
+	int ret = 0;
+
+	if (lock_needed)
+		mutex_lock(&smu->mutex);
+
+	if (smu->ppt_funcs->get_power_limit)
+		ret = smu->ppt_funcs->get_power_limit(smu, limit, def);
+
+	if (lock_needed)
+		mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+int smu_set_power_limit(struct smu_context *smu, uint32_t limit)
+{
+	int ret = 0;
+
+	mutex_lock(&smu->mutex);
+
+	if (smu->ppt_funcs->set_power_limit)
+		ret = smu->ppt_funcs->set_power_limit(smu, limit);
+
+	mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+int smu_print_clk_levels(struct smu_context *smu, enum smu_clk_type clk_type, char *buf)
+{
+	int ret = 0;
+
+	mutex_lock(&smu->mutex);
+
+	if (smu->ppt_funcs->print_clk_levels)
+		ret = smu->ppt_funcs->print_clk_levels(smu, clk_type, buf);
+
+	mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+int smu_get_od_percentage(struct smu_context *smu, enum smu_clk_type type)
+{
+	int ret = 0;
+
+	mutex_lock(&smu->mutex);
+
+	if (smu->ppt_funcs->get_od_percentage)
+		ret = smu->ppt_funcs->get_od_percentage(smu, type);
+
+	mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+int smu_set_od_percentage(struct smu_context *smu, enum smu_clk_type type, uint32_t value)
+{
+	int ret = 0;
+
+	mutex_lock(&smu->mutex);
+
+	if (smu->ppt_funcs->set_od_percentage)
+		ret = smu->ppt_funcs->set_od_percentage(smu, type, value);
+
+	mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+int smu_od_edit_dpm_table(struct smu_context *smu,
+			  enum PP_OD_DPM_TABLE_COMMAND type,
+			  long *input, uint32_t size)
+{
+	int ret = 0;
+
+	mutex_lock(&smu->mutex);
+
+	if (smu->ppt_funcs->od_edit_dpm_table)
+		ret = smu->ppt_funcs->od_edit_dpm_table(smu, type, input, size);
+
+	mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+int smu_read_sensor(struct smu_context *smu,
+		    enum amd_pp_sensors sensor,
+		    void *data, uint32_t *size)
+{
+	int ret = 0;
+
+	mutex_lock(&smu->mutex);
+
+	if (smu->ppt_funcs->read_sensor)
+		ret = smu->ppt_funcs->read_sensor(smu, sensor, data, size);
+
+	mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+int smu_get_power_profile_mode(struct smu_context *smu, char *buf)
+{
+	int ret = 0;
+
+	mutex_lock(&smu->mutex);
+
+	if (smu->ppt_funcs->get_power_profile_mode)
+		ret = smu->ppt_funcs->get_power_profile_mode(smu, buf);
+
+	mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+int smu_set_power_profile_mode(struct smu_context *smu,
+			       long *param,
+			       uint32_t param_size,
+			       bool lock_needed)
+{
+	int ret = 0;
+
+	if (lock_needed)
+		mutex_lock(&smu->mutex);
+
+	if (smu->ppt_funcs->set_power_profile_mode)
+		ret = smu->ppt_funcs->set_power_profile_mode(smu, param, param_size);
+
+	if (lock_needed)
+		mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+
+int smu_get_fan_control_mode(struct smu_context *smu)
+{
+	int ret = 0;
+
+	mutex_lock(&smu->mutex);
+
+	if (smu->ppt_funcs->get_fan_control_mode)
+		ret = smu->ppt_funcs->get_fan_control_mode(smu);
+
+	mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+int smu_set_fan_control_mode(struct smu_context *smu, int value)
+{
+	int ret = 0;
+
+	mutex_lock(&smu->mutex);
+
+	if (smu->ppt_funcs->set_fan_control_mode)
+		ret = smu->ppt_funcs->set_fan_control_mode(smu, value);
+
+	mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+int smu_get_fan_speed_percent(struct smu_context *smu, uint32_t *speed)
+{
+	int ret = 0;
+
+	mutex_lock(&smu->mutex);
+
+	if (smu->ppt_funcs->get_fan_speed_percent)
+		ret = smu->ppt_funcs->get_fan_speed_percent(smu, speed);
+
+	mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+int smu_set_fan_speed_percent(struct smu_context *smu, uint32_t speed)
+{
+	int ret = 0;
+
+	mutex_lock(&smu->mutex);
+
+	if (smu->ppt_funcs->set_fan_speed_percent)
+		ret = smu->ppt_funcs->set_fan_speed_percent(smu, speed);
+
+	mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+int smu_get_fan_speed_rpm(struct smu_context *smu, uint32_t *speed)
+{
+	int ret = 0;
+
+	mutex_lock(&smu->mutex);
+
+	if (smu->ppt_funcs->get_fan_speed_rpm)
+		ret = smu->ppt_funcs->get_fan_speed_rpm(smu, speed);
+
+	mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+int smu_set_deep_sleep_dcefclk(struct smu_context *smu, int clk)
+{
+	int ret = 0;
+
+	mutex_lock(&smu->mutex);
+
+	if (smu->ppt_funcs->set_deep_sleep_dcefclk)
+		ret = smu->ppt_funcs->set_deep_sleep_dcefclk(smu, clk);
+
+	mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+int smu_set_active_display_count(struct smu_context *smu, uint32_t count)
+{
+	int ret = 0;
+
+	mutex_lock(&smu->mutex);
+
+	if (smu->ppt_funcs->set_active_display_count)
+		ret = smu->ppt_funcs->set_active_display_count(smu, count);
+
+	mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+int smu_get_clock_by_type(struct smu_context *smu,
+			  enum amd_pp_clock_type type,
+			  struct amd_pp_clocks *clocks)
+{
+	int ret = 0;
+
+	mutex_lock(&smu->mutex);
+
+	if (smu->ppt_funcs->get_clock_by_type)
+		ret = smu->ppt_funcs->get_clock_by_type(smu, type, clocks);
+
+	mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+int smu_get_max_high_clocks(struct smu_context *smu,
+			    struct amd_pp_simple_clock_info *clocks)
+{
+	int ret = 0;
+
+	mutex_lock(&smu->mutex);
+
+	if (smu->ppt_funcs->get_max_high_clocks)
+		ret = smu->ppt_funcs->get_max_high_clocks(smu, clocks);
+
+	mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+int smu_get_clock_by_type_with_latency(struct smu_context *smu,
+				       enum smu_clk_type clk_type,
+				       struct pp_clock_levels_with_latency *clocks)
+{
+	int ret = 0;
+
+	mutex_lock(&smu->mutex);
+
+	if (smu->ppt_funcs->get_clock_by_type_with_latency)
+		ret = smu->ppt_funcs->get_clock_by_type_with_latency(smu, clk_type, clocks);
+
+	mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+int smu_get_clock_by_type_with_voltage(struct smu_context *smu,
+				       enum amd_pp_clock_type type,
+				       struct pp_clock_levels_with_voltage *clocks)
+{
+	int ret = 0;
+
+	mutex_lock(&smu->mutex);
+
+	if (smu->ppt_funcs->get_clock_by_type_with_voltage)
+		ret = smu->ppt_funcs->get_clock_by_type_with_voltage(smu, type, clocks);
+
+	mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+
+int smu_display_clock_voltage_request(struct smu_context *smu,
+				      struct pp_display_clock_request *clock_req)
+{
+	int ret = 0;
+
+	mutex_lock(&smu->mutex);
+
+	if (smu->ppt_funcs->display_clock_voltage_request)
+		ret = smu->ppt_funcs->display_clock_voltage_request(smu, clock_req);
+
+	mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+
+int smu_display_disable_memory_clock_switch(struct smu_context *smu, bool disable_memory_clock_switch)
+{
+	int ret = -EINVAL;
+
+	mutex_lock(&smu->mutex);
+
+	if (smu->ppt_funcs->display_disable_memory_clock_switch)
+		ret = smu->ppt_funcs->display_disable_memory_clock_switch(smu, disable_memory_clock_switch);
+
+	mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+int smu_notify_smu_enable_pwe(struct smu_context *smu)
+{
+	int ret = 0;
+
+	mutex_lock(&smu->mutex);
+
+	if (smu->ppt_funcs->notify_smu_enable_pwe)
+		ret = smu->ppt_funcs->notify_smu_enable_pwe(smu);
+
+	mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+int smu_set_xgmi_pstate(struct smu_context *smu,
+			uint32_t pstate)
+{
+	int ret = 0;
+
+	mutex_lock(&smu->mutex);
+
+	if (smu->ppt_funcs->set_xgmi_pstate)
+		ret = smu->ppt_funcs->set_xgmi_pstate(smu, pstate);
+
+	mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+int smu_set_azalia_d3_pme(struct smu_context *smu)
+{
+	int ret = 0;
+
+	mutex_lock(&smu->mutex);
+
+	if (smu->ppt_funcs->set_azalia_d3_pme)
+		ret = smu->ppt_funcs->set_azalia_d3_pme(smu);
+
+	mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+bool smu_baco_is_support(struct smu_context *smu)
+{
+	bool ret = false;
+
+	mutex_lock(&smu->mutex);
+
+	if (smu->ppt_funcs->baco_is_support)
+		ret = smu->ppt_funcs->baco_is_support(smu);
+
+	mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+int smu_baco_get_state(struct smu_context *smu, enum smu_baco_state *state)
+{
+	if (smu->ppt_funcs->baco_get_state)
+		return -EINVAL;
+
+	mutex_lock(&smu->mutex);
+	*state = smu->ppt_funcs->baco_get_state(smu);
+	mutex_unlock(&smu->mutex);
+
+	return 0;
+}
+
+int smu_baco_reset(struct smu_context *smu)
+{
+	int ret = 0;
+
+	mutex_lock(&smu->mutex);
+
+	if (smu->ppt_funcs->baco_reset)
+		ret = smu->ppt_funcs->baco_reset(smu);
+
+	mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+int smu_mode2_reset(struct smu_context *smu)
+{
+	int ret = 0;
+
+	mutex_lock(&smu->mutex);
+
+	if (smu->ppt_funcs->mode2_reset)
+		ret = smu->ppt_funcs->mode2_reset(smu);
+
+	mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+int smu_get_max_sustainable_clocks_by_dc(struct smu_context *smu,
+					 struct pp_smu_nv_clock_table *max_clocks)
+{
+	int ret = 0;
+
+	mutex_lock(&smu->mutex);
+
+	if (smu->ppt_funcs->get_max_sustainable_clocks_by_dc)
+		ret = smu->ppt_funcs->get_max_sustainable_clocks_by_dc(smu, max_clocks);
+
+	mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+int smu_get_uclk_dpm_states(struct smu_context *smu,
+			    unsigned int *clock_values_in_khz,
+			    unsigned int *num_states)
+{
+	int ret = 0;
+
+	mutex_lock(&smu->mutex);
+
+	if (smu->ppt_funcs->get_uclk_dpm_states)
+		ret = smu->ppt_funcs->get_uclk_dpm_states(smu, clock_values_in_khz, num_states);
+
+	mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+enum amd_pm_state_type smu_get_current_power_state(struct smu_context *smu)
+{
+	enum amd_pm_state_type pm_state = POWER_STATE_TYPE_DEFAULT;
+
+	mutex_lock(&smu->mutex);
+
+	if (smu->ppt_funcs->get_current_power_state)
+		pm_state = smu->ppt_funcs->get_current_power_state(smu);
+
+	mutex_unlock(&smu->mutex);
+
+	return pm_state;
+}
+
+int smu_get_dpm_clock_table(struct smu_context *smu,
+			    struct dpm_clocks *clock_table)
+{
+	int ret = 0;
+
+	mutex_lock(&smu->mutex);
+
+	if (smu->ppt_funcs->get_dpm_clock_table)
+		ret = smu->ppt_funcs->get_dpm_clock_table(smu, clock_table);
+
+	mutex_unlock(&smu->mutex);
+
+	return ret;
+}
+
+uint32_t smu_get_pptable_power_limit(struct smu_context *smu)
+{
+	uint32_t ret = 0;
+
+	if (smu->ppt_funcs->get_pptable_power_limit)
+		ret = smu->ppt_funcs->get_pptable_power_limit(smu);
+
+	return ret;
+}
diff --git a/drivers/gpu/drm/amd/powerplay/arcturus_ppt.c b/drivers/gpu/drm/amd/powerplay/arcturus_ppt.c
index f1f072012fac..58c7c4a3053e 100644
--- a/drivers/gpu/drm/amd/powerplay/arcturus_ppt.c
+++ b/drivers/gpu/drm/amd/powerplay/arcturus_ppt.c
@@ -25,6 +25,7 @@
 #include <linux/firmware.h>
 #include "amdgpu.h"
 #include "amdgpu_smu.h"
+#include "smu_internal.h"
 #include "atomfirmware.h"
 #include "amdgpu_atomfirmware.h"
 #include "smu_v11_0.h"
@@ -36,6 +37,12 @@
 #include "smu_v11_0_pptable.h"
 #include "arcturus_ppsmc.h"
 #include "nbio/nbio_7_4_sh_mask.h"
+#include "amdgpu_xgmi.h"
+#include <linux/i2c.h>
+#include <linux/pci.h>
+#include "amdgpu_ras.h"
+
+#define to_amdgpu_device(x) (container_of(x, struct amdgpu_ras, eeprom_control.eeprom_accessor))->adev
 
 #define CTF_OFFSET_EDGE			5
 #define CTF_OFFSET_HOTSPOT		5
@@ -112,8 +119,7 @@ static struct smu_11_0_cmn2aisc_mapping arcturus_message_map[SMU_MSG_MAX_COUNT]
 	MSG_MAP(PrepareMp1ForShutdown,		     PPSMC_MSG_PrepareMp1ForShutdown),
 	MSG_MAP(SoftReset,			     PPSMC_MSG_SoftReset),
 	MSG_MAP(RunAfllBtc,			     PPSMC_MSG_RunAfllBtc),
-	MSG_MAP(RunGfxDcBtc,			     PPSMC_MSG_RunGfxDcBtc),
-	MSG_MAP(RunSocDcBtc,			     PPSMC_MSG_RunSocDcBtc),
+	MSG_MAP(RunDcBtc,			     PPSMC_MSG_RunDcBtc),
 	MSG_MAP(DramLogSetDramAddrHigh,		     PPSMC_MSG_DramLogSetDramAddrHigh),
 	MSG_MAP(DramLogSetDramAddrLow,		     PPSMC_MSG_DramLogSetDramAddrLow),
 	MSG_MAP(DramLogSetDramSize,		     PPSMC_MSG_DramLogSetDramSize),
@@ -172,6 +178,7 @@ static struct smu_11_0_cmn2aisc_mapping arcturus_table_map[SMU_TABLE_COUNT] = {
 	TAB_MAP(SMU_METRICS),
 	TAB_MAP(DRIVER_SMU_CONFIG),
 	TAB_MAP(OVERDRIVE),
+	TAB_MAP(I2C_COMMANDS),
 };
 
 static struct smu_11_0_cmn2aisc_mapping arcturus_pwr_src_map[SMU_POWER_SOURCE_COUNT] = {
@@ -294,6 +301,9 @@ static int arcturus_tables_init(struct smu_context *smu, struct smu_table *table
 	SMU_TABLE_INIT(tables, SMU_TABLE_SMU_METRICS, sizeof(SmuMetrics_t),
 		       PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM);
 
+	SMU_TABLE_INIT(tables, SMU_TABLE_I2C_COMMANDS, sizeof(SwI2cRequest_t),
+			       PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM);
+
 	smu_table->metrics_table = kzalloc(sizeof(SmuMetrics_t), GFP_KERNEL);
 	if (!smu_table->metrics_table)
 		return -ENOMEM;
@@ -528,9 +538,17 @@ static int arcturus_append_powerplay_table(struct smu_context *smu)
 	return 0;
 }
 
-static int arcturus_run_btc_afll(struct smu_context *smu)
+static int arcturus_run_btc(struct smu_context *smu)
 {
-	return smu_send_smc_msg(smu, SMU_MSG_RunAfllBtc);
+	int ret = 0;
+
+	ret = smu_send_smc_msg(smu, SMU_MSG_RunAfllBtc);
+	if (ret) {
+		pr_err("RunAfllBtc failed!\n");
+		return ret;
+	}
+
+	return smu_send_smc_msg(smu, SMU_MSG_RunDcBtc);
 }
 
 static int arcturus_populate_umd_state_clk(struct smu_context *smu)
@@ -610,12 +628,17 @@ static int arcturus_print_clk_levels(struct smu_context *smu,
 			return ret;
 		}
 
+		/*
+		 * For DPM disabled case, there will be only one clock level.
+		 * And it's safe to assume that is always the current clock.
+		 */
 		for (i = 0; i < clocks.num_levels; i++)
 			size += sprintf(buf + size, "%d: %uMhz %s\n", i,
 					clocks.data[i].clocks_in_khz / 1000,
-					arcturus_freqs_in_same_level(
+					(clocks.num_levels == 1) ? "*" :
+					(arcturus_freqs_in_same_level(
 					clocks.data[i].clocks_in_khz / 1000,
-					now / 100) ? "*" : "");
+					now / 100) ? "*" : ""));
 		break;
 
 	case SMU_MCLK:
@@ -635,9 +658,10 @@ static int arcturus_print_clk_levels(struct smu_context *smu,
 		for (i = 0; i < clocks.num_levels; i++)
 			size += sprintf(buf + size, "%d: %uMhz %s\n",
 				i, clocks.data[i].clocks_in_khz / 1000,
-				arcturus_freqs_in_same_level(
+				(clocks.num_levels == 1) ? "*" :
+				(arcturus_freqs_in_same_level(
 				clocks.data[i].clocks_in_khz / 1000,
-				now / 100) ? "*" : "");
+				now / 100) ? "*" : ""));
 		break;
 
 	case SMU_SOCCLK:
@@ -657,9 +681,10 @@ static int arcturus_print_clk_levels(struct smu_context *smu,
 		for (i = 0; i < clocks.num_levels; i++)
 			size += sprintf(buf + size, "%d: %uMhz %s\n",
 				i, clocks.data[i].clocks_in_khz / 1000,
-				arcturus_freqs_in_same_level(
+				(clocks.num_levels == 1) ? "*" :
+				(arcturus_freqs_in_same_level(
 				clocks.data[i].clocks_in_khz / 1000,
-				now / 100) ? "*" : "");
+				now / 100) ? "*" : ""));
 		break;
 
 	case SMU_FCLK:
@@ -679,9 +704,10 @@ static int arcturus_print_clk_levels(struct smu_context *smu,
 		for (i = 0; i < single_dpm_table->count; i++)
 			size += sprintf(buf + size, "%d: %uMhz %s\n",
 				i, single_dpm_table->dpm_levels[i].value,
-				arcturus_freqs_in_same_level(
+				(clocks.num_levels == 1) ? "*" :
+				(arcturus_freqs_in_same_level(
 				clocks.data[i].clocks_in_khz / 1000,
-				now / 100) ? "*" : "");
+				now / 100) ? "*" : ""));
 		break;
 
 	default:
@@ -756,8 +782,6 @@ static int arcturus_force_clk_levels(struct smu_context *smu,
 	uint32_t soft_min_level, soft_max_level;
 	int ret = 0;
 
-	mutex_lock(&(smu->mutex));
-
 	soft_min_level = mask ? (ffs(mask) - 1) : 0;
 	soft_max_level = mask ? (fls(mask) - 1) : 0;
 
@@ -792,91 +816,19 @@ static int arcturus_force_clk_levels(struct smu_context *smu,
 		break;
 
 	case SMU_MCLK:
-		single_dpm_table = &(dpm_table->mem_table);
-
-		if (soft_max_level >= single_dpm_table->count) {
-			pr_err("Clock level specified %d is over max allowed %d\n",
-					soft_max_level, single_dpm_table->count - 1);
-			ret = -EINVAL;
-			break;
-		}
-
-		single_dpm_table->dpm_state.soft_min_level =
-			single_dpm_table->dpm_levels[soft_min_level].value;
-		single_dpm_table->dpm_state.soft_max_level =
-			single_dpm_table->dpm_levels[soft_max_level].value;
-
-		ret = arcturus_upload_dpm_level(smu, false, FEATURE_DPM_UCLK_MASK);
-		if (ret) {
-			pr_err("Failed to upload boot level to lowest!\n");
-			break;
-		}
-
-		ret = arcturus_upload_dpm_level(smu, true, FEATURE_DPM_UCLK_MASK);
-		if (ret)
-			pr_err("Failed to upload dpm max level to highest!\n");
-
-		break;
-
 	case SMU_SOCCLK:
-		single_dpm_table = &(dpm_table->soc_table);
-
-		if (soft_max_level >= single_dpm_table->count) {
-			pr_err("Clock level specified %d is over max allowed %d\n",
-					soft_max_level, single_dpm_table->count - 1);
-			ret = -EINVAL;
-			break;
-		}
-
-		single_dpm_table->dpm_state.soft_min_level =
-			single_dpm_table->dpm_levels[soft_min_level].value;
-		single_dpm_table->dpm_state.soft_max_level =
-			single_dpm_table->dpm_levels[soft_max_level].value;
-
-		ret = arcturus_upload_dpm_level(smu, false, FEATURE_DPM_SOCCLK_MASK);
-		if (ret) {
-			pr_err("Failed to upload boot level to lowest!\n");
-			break;
-		}
-
-		ret = arcturus_upload_dpm_level(smu, true, FEATURE_DPM_SOCCLK_MASK);
-		if (ret)
-			pr_err("Failed to upload dpm max level to highest!\n");
-
-		break;
-
 	case SMU_FCLK:
-		single_dpm_table = &(dpm_table->fclk_table);
-
-		if (soft_max_level >= single_dpm_table->count) {
-			pr_err("Clock level specified %d is over max allowed %d\n",
-					soft_max_level, single_dpm_table->count - 1);
-			ret = -EINVAL;
-			break;
-		}
-
-		single_dpm_table->dpm_state.soft_min_level =
-			single_dpm_table->dpm_levels[soft_min_level].value;
-		single_dpm_table->dpm_state.soft_max_level =
-			single_dpm_table->dpm_levels[soft_max_level].value;
-
-		ret = arcturus_upload_dpm_level(smu, false, FEATURE_DPM_FCLK_MASK);
-		if (ret) {
-			pr_err("Failed to upload boot level to lowest!\n");
-			break;
-		}
-
-		ret = arcturus_upload_dpm_level(smu, true, FEATURE_DPM_FCLK_MASK);
-		if (ret)
-			pr_err("Failed to upload dpm max level to highest!\n");
-
+		/*
+		 * Should not arrive here since Arcturus does not
+		 * support mclk/socclk/fclk softmin/softmax settings
+		 */
+		ret = -EINVAL;
 		break;
 
 	default:
 		break;
 	}
 
-	mutex_unlock(&(smu->mutex));
 	return ret;
 }
 
@@ -1018,6 +970,7 @@ static int arcturus_read_sensor(struct smu_context *smu,
 	if (!data || !size)
 		return -EINVAL;
 
+	mutex_lock(&smu->sensor_lock);
 	switch (sensor) {
 	case AMDGPU_PP_SENSOR_MAX_FAN_RPM:
 		*(uint32_t *)data = pptable->FanMaximumRpm;
@@ -1042,8 +995,9 @@ static int arcturus_read_sensor(struct smu_context *smu,
 		*size = 4;
 		break;
 	default:
-		ret = smu_smc_read_sensor(smu, sensor, data, size);
+		ret = smu_v11_0_read_sensor(smu, sensor, data, size);
 	}
+	mutex_unlock(&smu->sensor_lock);
 
 	return ret;
 }
@@ -1184,6 +1138,7 @@ static int arcturus_force_dpm_limit_value(struct smu_context *smu, bool highest)
 {
 	struct arcturus_dpm_table *dpm_table =
 		(struct arcturus_dpm_table *)smu->smu_dpm.dpm_context;
+	struct amdgpu_hive_info *hive = amdgpu_get_xgmi_hive(smu->adev, 0);
 	uint32_t soft_level;
 	int ret = 0;
 
@@ -1197,40 +1152,27 @@ static int arcturus_force_dpm_limit_value(struct smu_context *smu, bool highest)
 		dpm_table->gfx_table.dpm_state.soft_max_level =
 		dpm_table->gfx_table.dpm_levels[soft_level].value;
 
-	/* uclk */
-	if (highest)
-		soft_level = arcturus_find_highest_dpm_level(&(dpm_table->mem_table));
-	else
-		soft_level = arcturus_find_lowest_dpm_level(&(dpm_table->mem_table));
-
-	dpm_table->mem_table.dpm_state.soft_min_level =
-		dpm_table->mem_table.dpm_state.soft_max_level =
-		dpm_table->mem_table.dpm_levels[soft_level].value;
-
-	/* socclk */
-	if (highest)
-		soft_level = arcturus_find_highest_dpm_level(&(dpm_table->soc_table));
-	else
-		soft_level = arcturus_find_lowest_dpm_level(&(dpm_table->soc_table));
-
-	dpm_table->soc_table.dpm_state.soft_min_level =
-		dpm_table->soc_table.dpm_state.soft_max_level =
-		dpm_table->soc_table.dpm_levels[soft_level].value;
-
-	ret = arcturus_upload_dpm_level(smu, false, 0xFFFFFFFF);
+	ret = arcturus_upload_dpm_level(smu, false, FEATURE_DPM_GFXCLK_MASK);
 	if (ret) {
 		pr_err("Failed to upload boot level to %s!\n",
 				highest ? "highest" : "lowest");
 		return ret;
 	}
 
-	ret = arcturus_upload_dpm_level(smu, true, 0xFFFFFFFF);
+	ret = arcturus_upload_dpm_level(smu, true, FEATURE_DPM_GFXCLK_MASK);
 	if (ret) {
 		pr_err("Failed to upload dpm max level to %s!\n!",
 				highest ? "highest" : "lowest");
 		return ret;
 	}
 
+	if (hive)
+		/*
+		 * Force XGMI Pstate to highest or lowest
+		 * TODO: revise this when xgmi dpm is functional
+		 */
+		ret = smu_v11_0_set_xgmi_pstate(smu, highest ? 1 : 0);
+
 	return ret;
 }
 
@@ -1238,6 +1180,7 @@ static int arcturus_unforce_dpm_levels(struct smu_context *smu)
 {
 	struct arcturus_dpm_table *dpm_table =
 		(struct arcturus_dpm_table *)smu->smu_dpm.dpm_context;
+	struct amdgpu_hive_info *hive = amdgpu_get_xgmi_hive(smu->adev, 0);
 	uint32_t soft_min_level, soft_max_level;
 	int ret = 0;
 
@@ -1249,34 +1192,25 @@ static int arcturus_unforce_dpm_levels(struct smu_context *smu)
 	dpm_table->gfx_table.dpm_state.soft_max_level =
 		dpm_table->gfx_table.dpm_levels[soft_max_level].value;
 
-	/* uclk */
-	soft_min_level = arcturus_find_lowest_dpm_level(&(dpm_table->mem_table));
-	soft_max_level = arcturus_find_highest_dpm_level(&(dpm_table->mem_table));
-	dpm_table->mem_table.dpm_state.soft_min_level =
-		dpm_table->gfx_table.dpm_levels[soft_min_level].value;
-	dpm_table->mem_table.dpm_state.soft_max_level =
-		dpm_table->gfx_table.dpm_levels[soft_max_level].value;
-
-	/* socclk */
-	soft_min_level = arcturus_find_lowest_dpm_level(&(dpm_table->soc_table));
-	soft_max_level = arcturus_find_highest_dpm_level(&(dpm_table->soc_table));
-	dpm_table->soc_table.dpm_state.soft_min_level =
-		dpm_table->soc_table.dpm_levels[soft_min_level].value;
-	dpm_table->soc_table.dpm_state.soft_max_level =
-		dpm_table->soc_table.dpm_levels[soft_max_level].value;
-
-	ret = arcturus_upload_dpm_level(smu, false, 0xFFFFFFFF);
+	ret = arcturus_upload_dpm_level(smu, false, FEATURE_DPM_GFXCLK_MASK);
 	if (ret) {
 		pr_err("Failed to upload DPM Bootup Levels!");
 		return ret;
 	}
 
-	ret = arcturus_upload_dpm_level(smu, true, 0xFFFFFFFF);
+	ret = arcturus_upload_dpm_level(smu, true, FEATURE_DPM_GFXCLK_MASK);
 	if (ret) {
 		pr_err("Failed to upload DPM Max Levels!");
 		return ret;
 	}
 
+	if (hive)
+		/*
+		 * Reset XGMI Pstate back to default
+		 * TODO: revise this when xgmi dpm is functional
+		 */
+		ret = smu_v11_0_set_xgmi_pstate(smu, 0);
+
 	return ret;
 }
 
@@ -1327,15 +1261,14 @@ arcturus_get_profiling_clk_mask(struct smu_context *smu,
 
 static int arcturus_get_power_limit(struct smu_context *smu,
 				     uint32_t *limit,
-				     bool asic_default)
+				     bool cap)
 {
 	PPTable_t *pptable = smu->smu_table.driver_pptable;
 	uint32_t asic_default_power_limit = 0;
 	int ret = 0;
 	int power_src;
 
-	if (!smu->default_power_limit ||
-	    !smu->power_limit) {
+	if (!smu->power_limit) {
 		if (smu_feature_is_enabled(smu, SMU_FEATURE_PPT_BIT)) {
 			power_src = smu_power_get_index(smu, SMU_POWER_SOURCE_AC);
 			if (power_src < 0)
@@ -1358,17 +1291,11 @@ static int arcturus_get_power_limit(struct smu_context *smu,
 				pptable->SocketPowerLimitAc[PPT_THROTTLER_PPT0];
 		}
 
-		if (smu->od_enabled) {
-			asic_default_power_limit *= (100 + smu->smu_table.TDPODLimit);
-			asic_default_power_limit /= 100;
-		}
-
-		smu->default_power_limit = asic_default_power_limit;
 		smu->power_limit = asic_default_power_limit;
 	}
 
-	if (asic_default)
-		*limit = smu->default_power_limit;
+	if (cap)
+		*limit = smu_v11_0_get_max_power_limit(smu);
 	else
 		*limit = smu->power_limit;
 
@@ -1889,6 +1816,260 @@ static bool arcturus_is_dpm_running(struct smu_context *smu)
 	return !!(feature_enabled & SMC_DPM_FEATURE);
 }
 
+static int arcturus_dpm_set_uvd_enable(struct smu_context *smu, bool enable)
+{
+	struct smu_power_context *smu_power = &smu->smu_power;
+	struct smu_power_gate *power_gate = &smu_power->power_gate;
+	int ret = 0;
+
+	if (enable) {
+		if (!smu_feature_is_enabled(smu, SMU_FEATURE_VCN_PG_BIT)) {
+			ret = smu_feature_set_enabled(smu, SMU_FEATURE_VCN_PG_BIT, 1);
+			if (ret) {
+				pr_err("[EnableVCNDPM] failed!\n");
+				return ret;
+			}
+		}
+		power_gate->vcn_gated = false;
+	} else {
+		if (smu_feature_is_enabled(smu, SMU_FEATURE_VCN_PG_BIT)) {
+			ret = smu_feature_set_enabled(smu, SMU_FEATURE_VCN_PG_BIT, 0);
+			if (ret) {
+				pr_err("[DisableVCNDPM] failed!\n");
+				return ret;
+			}
+		}
+		power_gate->vcn_gated = true;
+	}
+
+	return ret;
+}
+
+
+static void arcturus_fill_eeprom_i2c_req(SwI2cRequest_t  *req, bool write,
+				  uint8_t address, uint32_t numbytes,
+				  uint8_t *data)
+{
+	int i;
+
+	BUG_ON(numbytes > MAX_SW_I2C_COMMANDS);
+
+	req->I2CcontrollerPort = 0;
+	req->I2CSpeed = 2;
+	req->SlaveAddress = address;
+	req->NumCmds = numbytes;
+
+	for (i = 0; i < numbytes; i++) {
+		SwI2cCmd_t *cmd =  &req->SwI2cCmds[i];
+
+		/* First 2 bytes are always write for lower 2b EEPROM address */
+		if (i < 2)
+			cmd->Cmd = 1;
+		else
+			cmd->Cmd = write;
+
+
+		/* Add RESTART for read  after address filled */
+		cmd->CmdConfig |= (i == 2 && !write) ? CMDCONFIG_RESTART_MASK : 0;
+
+		/* Add STOP in the end */
+		cmd->CmdConfig |= (i == (numbytes - 1)) ? CMDCONFIG_STOP_MASK : 0;
+
+		/* Fill with data regardless if read or write to simplify code */
+		cmd->RegisterAddr = data[i];
+	}
+}
+
+static int arcturus_i2c_eeprom_read_data(struct i2c_adapter *control,
+					       uint8_t address,
+					       uint8_t *data,
+					       uint32_t numbytes)
+{
+	uint32_t  i, ret = 0;
+	SwI2cRequest_t req;
+	struct amdgpu_device *adev = to_amdgpu_device(control);
+	struct smu_table_context *smu_table = &adev->smu.smu_table;
+	struct smu_table *table = &smu_table->tables[SMU_TABLE_I2C_COMMANDS];
+
+	memset(&req, 0, sizeof(req));
+	arcturus_fill_eeprom_i2c_req(&req, false, address, numbytes, data);
+
+	mutex_lock(&adev->smu.mutex);
+	/* Now read data starting with that address */
+	ret = smu_update_table(&adev->smu, SMU_TABLE_I2C_COMMANDS, 0, &req,
+					true);
+	mutex_unlock(&adev->smu.mutex);
+
+	if (!ret) {
+		SwI2cRequest_t *res = (SwI2cRequest_t *)table->cpu_addr;
+
+		/* Assume SMU  fills res.SwI2cCmds[i].Data with read bytes */
+		for (i = 0; i < numbytes; i++)
+			data[i] = res->SwI2cCmds[i].Data;
+
+		pr_debug("arcturus_i2c_eeprom_read_data, address = %x, bytes = %d, data :",
+				  (uint16_t)address, numbytes);
+
+		print_hex_dump(KERN_DEBUG, "data: ", DUMP_PREFIX_NONE,
+			       8, 1, data, numbytes, false);
+	} else
+		pr_err("arcturus_i2c_eeprom_read_data - error occurred :%x", ret);
+
+	return ret;
+}
+
+static int arcturus_i2c_eeprom_write_data(struct i2c_adapter *control,
+						uint8_t address,
+						uint8_t *data,
+						uint32_t numbytes)
+{
+	uint32_t ret;
+	SwI2cRequest_t req;
+	struct amdgpu_device *adev = to_amdgpu_device(control);
+
+	memset(&req, 0, sizeof(req));
+	arcturus_fill_eeprom_i2c_req(&req, true, address, numbytes, data);
+
+	mutex_lock(&adev->smu.mutex);
+	ret = smu_update_table(&adev->smu, SMU_TABLE_I2C_COMMANDS, 0, &req, true);
+	mutex_unlock(&adev->smu.mutex);
+
+	if (!ret) {
+		pr_debug("arcturus_i2c_write(), address = %x, bytes = %d , data: ",
+					 (uint16_t)address, numbytes);
+
+		print_hex_dump(KERN_DEBUG, "data: ", DUMP_PREFIX_NONE,
+			       8, 1, data, numbytes, false);
+		/*
+		 * According to EEPROM spec there is a MAX of 10 ms required for
+		 * EEPROM to flush internal RX buffer after STOP was issued at the
+		 * end of write transaction. During this time the EEPROM will not be
+		 * responsive to any more commands - so wait a bit more.
+		 */
+		msleep(10);
+
+	} else
+		pr_err("arcturus_i2c_write- error occurred :%x", ret);
+
+	return ret;
+}
+
+static int arcturus_i2c_eeprom_i2c_xfer(struct i2c_adapter *i2c_adap,
+			      struct i2c_msg *msgs, int num)
+{
+	uint32_t  i, j, ret, data_size, data_chunk_size, next_eeprom_addr = 0;
+	uint8_t *data_ptr, data_chunk[MAX_SW_I2C_COMMANDS] = { 0 };
+
+	for (i = 0; i < num; i++) {
+		/*
+		 * SMU interface allows at most MAX_SW_I2C_COMMANDS bytes of data at
+		 * once and hence the data needs to be spliced into chunks and sent each
+		 * chunk separately
+		 */
+		data_size = msgs[i].len - 2;
+		data_chunk_size = MAX_SW_I2C_COMMANDS - 2;
+		next_eeprom_addr = (msgs[i].buf[0] << 8 & 0xff00) | (msgs[i].buf[1] & 0xff);
+		data_ptr = msgs[i].buf + 2;
+
+		for (j = 0; j < data_size / data_chunk_size; j++) {
+			/* Insert the EEPROM dest addess, bits 0-15 */
+			data_chunk[0] = ((next_eeprom_addr >> 8) & 0xff);
+			data_chunk[1] = (next_eeprom_addr & 0xff);
+
+			if (msgs[i].flags & I2C_M_RD) {
+				ret = arcturus_i2c_eeprom_read_data(i2c_adap,
+								(uint8_t)msgs[i].addr,
+								data_chunk, MAX_SW_I2C_COMMANDS);
+
+				memcpy(data_ptr, data_chunk + 2, data_chunk_size);
+			} else {
+
+				memcpy(data_chunk + 2, data_ptr, data_chunk_size);
+
+				ret = arcturus_i2c_eeprom_write_data(i2c_adap,
+								 (uint8_t)msgs[i].addr,
+								 data_chunk, MAX_SW_I2C_COMMANDS);
+			}
+
+			if (ret) {
+				num = -EIO;
+				goto fail;
+			}
+
+			next_eeprom_addr += data_chunk_size;
+			data_ptr += data_chunk_size;
+		}
+
+		if (data_size % data_chunk_size) {
+			data_chunk[0] = ((next_eeprom_addr >> 8) & 0xff);
+			data_chunk[1] = (next_eeprom_addr & 0xff);
+
+			if (msgs[i].flags & I2C_M_RD) {
+				ret = arcturus_i2c_eeprom_read_data(i2c_adap,
+								(uint8_t)msgs[i].addr,
+								data_chunk, (data_size % data_chunk_size) + 2);
+
+				memcpy(data_ptr, data_chunk + 2, data_size % data_chunk_size);
+			} else {
+				memcpy(data_chunk + 2, data_ptr, data_size % data_chunk_size);
+
+				ret = arcturus_i2c_eeprom_write_data(i2c_adap,
+								 (uint8_t)msgs[i].addr,
+								 data_chunk, (data_size % data_chunk_size) + 2);
+			}
+
+			if (ret) {
+				num = -EIO;
+				goto fail;
+			}
+		}
+	}
+
+fail:
+	return num;
+}
+
+static u32 arcturus_i2c_eeprom_i2c_func(struct i2c_adapter *adap)
+{
+	return I2C_FUNC_I2C | I2C_FUNC_SMBUS_EMUL;
+}
+
+
+static const struct i2c_algorithm arcturus_i2c_eeprom_i2c_algo = {
+	.master_xfer = arcturus_i2c_eeprom_i2c_xfer,
+	.functionality = arcturus_i2c_eeprom_i2c_func,
+};
+
+static int arcturus_i2c_eeprom_control_init(struct i2c_adapter *control)
+{
+	struct amdgpu_device *adev = to_amdgpu_device(control);
+	int res;
+
+	control->owner = THIS_MODULE;
+	control->class = I2C_CLASS_SPD;
+	control->dev.parent = &adev->pdev->dev;
+	control->algo = &arcturus_i2c_eeprom_i2c_algo;
+	snprintf(control->name, sizeof(control->name), "RAS EEPROM");
+
+	res = i2c_add_adapter(control);
+	if (res)
+		DRM_ERROR("Failed to register hw i2c, err: %d\n", res);
+
+	return res;
+}
+
+static void arcturus_i2c_eeprom_control_fini(struct i2c_adapter *control)
+{
+	i2c_del_adapter(control);
+}
+
+static uint32_t arcturus_get_pptable_power_limit(struct smu_context *smu)
+{
+	PPTable_t *pptable = smu->smu_table.driver_pptable;
+
+	return pptable->SocketPowerLimitAc[PPT_THROTTLER_PPT0];
+}
+
 static const struct pptable_funcs arcturus_ppt_funcs = {
 	/* translate smu index into arcturus specific index */
 	.get_smu_msg_index = arcturus_get_smu_msg_index,
@@ -1907,7 +2088,7 @@ static const struct pptable_funcs arcturus_ppt_funcs = {
 	/* init dpm */
 	.get_allowed_feature_mask = arcturus_get_allowed_feature_mask,
 	/* btc */
-	.run_afll_btc = arcturus_run_btc_afll,
+	.run_btc = arcturus_run_btc,
 	/* dpm/clk tables */
 	.set_default_dpm_table = arcturus_set_default_dpm_table,
 	.populate_umd_state_clk = arcturus_populate_umd_state_clk,
@@ -1927,12 +2108,62 @@ static const struct pptable_funcs arcturus_ppt_funcs = {
 	.dump_pptable = arcturus_dump_pptable,
 	.get_power_limit = arcturus_get_power_limit,
 	.is_dpm_running = arcturus_is_dpm_running,
+	.dpm_set_uvd_enable = arcturus_dpm_set_uvd_enable,
+	.i2c_eeprom_init = arcturus_i2c_eeprom_control_init,
+	.i2c_eeprom_fini = arcturus_i2c_eeprom_control_fini,
+	.init_microcode = smu_v11_0_init_microcode,
+	.load_microcode = smu_v11_0_load_microcode,
+	.init_smc_tables = smu_v11_0_init_smc_tables,
+	.fini_smc_tables = smu_v11_0_fini_smc_tables,
+	.init_power = smu_v11_0_init_power,
+	.fini_power = smu_v11_0_fini_power,
+	.check_fw_status = smu_v11_0_check_fw_status,
+	.setup_pptable = smu_v11_0_setup_pptable,
+	.get_vbios_bootup_values = smu_v11_0_get_vbios_bootup_values,
+	.get_clk_info_from_vbios = smu_v11_0_get_clk_info_from_vbios,
+	.check_pptable = smu_v11_0_check_pptable,
+	.parse_pptable = smu_v11_0_parse_pptable,
+	.populate_smc_tables = smu_v11_0_populate_smc_pptable,
+	.check_fw_version = smu_v11_0_check_fw_version,
+	.write_pptable = smu_v11_0_write_pptable,
+	.set_min_dcef_deep_sleep = smu_v11_0_set_min_dcef_deep_sleep,
+	.set_tool_table_location = smu_v11_0_set_tool_table_location,
+	.notify_memory_pool_location = smu_v11_0_notify_memory_pool_location,
+	.system_features_control = smu_v11_0_system_features_control,
+	.send_smc_msg = smu_v11_0_send_msg,
+	.send_smc_msg_with_param = smu_v11_0_send_msg_with_param,
+	.read_smc_arg = smu_v11_0_read_arg,
+	.init_display_count = smu_v11_0_init_display_count,
+	.set_allowed_mask = smu_v11_0_set_allowed_mask,
+	.get_enabled_mask = smu_v11_0_get_enabled_mask,
+	.notify_display_change = smu_v11_0_notify_display_change,
+	.set_power_limit = smu_v11_0_set_power_limit,
+	.get_current_clk_freq = smu_v11_0_get_current_clk_freq,
+	.init_max_sustainable_clocks = smu_v11_0_init_max_sustainable_clocks,
+	.start_thermal_control = smu_v11_0_start_thermal_control,
+	.stop_thermal_control = smu_v11_0_stop_thermal_control,
+	.set_deep_sleep_dcefclk = smu_v11_0_set_deep_sleep_dcefclk,
+	.display_clock_voltage_request = smu_v11_0_display_clock_voltage_request,
+	.get_fan_control_mode = smu_v11_0_get_fan_control_mode,
+	.set_fan_control_mode = smu_v11_0_set_fan_control_mode,
+	.set_fan_speed_percent = smu_v11_0_set_fan_speed_percent,
+	.set_fan_speed_rpm = smu_v11_0_set_fan_speed_rpm,
+	.set_xgmi_pstate = smu_v11_0_set_xgmi_pstate,
+	.gfx_off_control = smu_v11_0_gfx_off_control,
+	.register_irq_handler = smu_v11_0_register_irq_handler,
+	.set_azalia_d3_pme = smu_v11_0_set_azalia_d3_pme,
+	.get_max_sustainable_clocks_by_dc = smu_v11_0_get_max_sustainable_clocks_by_dc,
+	.baco_is_support= smu_v11_0_baco_is_support,
+	.baco_get_state = smu_v11_0_baco_get_state,
+	.baco_set_state = smu_v11_0_baco_set_state,
+	.baco_reset = smu_v11_0_baco_reset,
+	.get_dpm_ultimate_freq = smu_v11_0_get_dpm_ultimate_freq,
+	.set_soft_freq_limited_range = smu_v11_0_set_soft_freq_limited_range,
+	.override_pcie_parameters = smu_v11_0_override_pcie_parameters,
+	.get_pptable_power_limit = arcturus_get_pptable_power_limit,
 };
 
 void arcturus_set_ppt_funcs(struct smu_context *smu)
 {
-	struct smu_table_context *smu_table = &smu->smu_table;
-
 	smu->ppt_funcs = &arcturus_ppt_funcs;
-	smu_table->table_count = TABLE_COUNT;
 }
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/Makefile b/drivers/gpu/drm/amd/powerplay/hwmgr/Makefile
index cc63705920dc..2773966ae434 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/Makefile
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/Makefile
@@ -36,7 +36,8 @@ HARDWARE_MGR = hwmgr.o processpptables.o \
 		pp_overdriver.o smu_helper.o \
 		vega20_processpptables.o vega20_hwmgr.o vega20_powertune.o \
 		vega20_thermal.o common_baco.o vega10_baco.o  vega20_baco.o \
-		vega12_baco.o smu9_baco.o
+		vega12_baco.o smu9_baco.o tonga_baco.o polaris_baco.o fiji_baco.o \
+		ci_baco.o smu7_baco.o
 
 AMD_PP_HWMGR = $(addprefix $(AMD_PP_PATH)/hwmgr/,$(HARDWARE_MGR))
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/ci_baco.c b/drivers/gpu/drm/amd/powerplay/hwmgr/ci_baco.c
new file mode 100644
index 000000000000..3be40114e63d
--- /dev/null
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/ci_baco.c
@@ -0,0 +1,195 @@
+/*
+ * Copyright 2019 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#include "amdgpu.h"
+#include "ci_baco.h"
+
+#include "gmc/gmc_7_1_d.h"
+#include "gmc/gmc_7_1_sh_mask.h"
+
+#include "bif/bif_4_1_d.h"
+#include "bif/bif_4_1_sh_mask.h"
+
+#include "dce/dce_8_0_d.h"
+#include "dce/dce_8_0_sh_mask.h"
+
+#include "smu/smu_7_0_1_d.h"
+#include "smu/smu_7_0_1_sh_mask.h"
+
+#include "gca/gfx_7_2_d.h"
+#include "gca/gfx_7_2_sh_mask.h"
+
+static const struct baco_cmd_entry gpio_tbl[] =
+{
+	{ CMD_WRITE, mmGPIOPAD_EN, 0, 0, 0, 0x0 },
+	{ CMD_WRITE, mmGPIOPAD_PD_EN, 0, 0, 0, 0x0 },
+	{ CMD_WRITE, mmGPIOPAD_PU_EN, 0, 0, 0, 0x0 },
+	{ CMD_WRITE, mmGPIOPAD_MASK, 0, 0, 0, 0xff77ffff },
+	{ CMD_WRITE, mmDC_GPIO_DVODATA_EN, 0, 0, 0, 0x0 },
+	{ CMD_WRITE, mmDC_GPIO_DVODATA_MASK, 0, 0, 0, 0xffffffff },
+	{ CMD_WRITE, mmDC_GPIO_GENERIC_EN, 0, 0, 0, 0x0 },
+	{ CMD_READMODIFYWRITE, mmDC_GPIO_GENERIC_MASK, 0, 0, 0, 0x03333333 },
+	{ CMD_WRITE, mmDC_GPIO_SYNCA_EN, 0, 0, 0, 0x0 },
+	{ CMD_READMODIFYWRITE, mmDC_GPIO_SYNCA_MASK, 0, 0, 0, 0x00001111 }
+};
+
+static const struct baco_cmd_entry enable_fb_req_rej_tbl[] =
+{
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, 0xC0300024 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, 0x1, 0x0, 0, 0x1 },
+	{ CMD_WRITE, mmBIF_FB_EN, 0, 0, 0, 0x0 }
+};
+
+static const struct baco_cmd_entry use_bclk_tbl[] =
+{
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixCG_SPLL_FUNC_CNTL },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, CG_SPLL_FUNC_CNTL__SPLL_BYPASS_EN_MASK, CG_SPLL_FUNC_CNTL__SPLL_BYPASS_EN__SHIFT, 0, 0x1 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixCG_SPLL_FUNC_CNTL_2 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, CG_SPLL_FUNC_CNTL_2__SPLL_BYPASS_CHG_MASK, CG_SPLL_FUNC_CNTL_2__SPLL_BYPASS_CHG__SHIFT, 0, 0x1 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixCG_SPLL_STATUS },
+	{ CMD_WAITFOR, mmGCK_SMC_IND_DATA, CG_SPLL_STATUS__SPLL_CHG_STATUS_MASK, 0, 0xffffffff, 0x2 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixCG_SPLL_FUNC_CNTL_2 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, CG_SPLL_FUNC_CNTL_2__SPLL_BYPASS_CHG_MASK, CG_SPLL_FUNC_CNTL_2__SPLL_BYPASS_CHG__SHIFT, 0, 0x0 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, CG_SPLL_FUNC_CNTL_2__SPLL_CTLREQ_CHG_MASK, CG_SPLL_FUNC_CNTL_2__SPLL_CTLREQ_CHG__SHIFT, 0, 0x1 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixCG_SPLL_STATUS },
+	{ CMD_WAITFOR, mmGCK_SMC_IND_DATA, CG_SPLL_STATUS__SPLL_CHG_STATUS_MASK, 0, 0xffffffff, 0x2 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixCG_SPLL_FUNC_CNTL_2 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, CG_SPLL_FUNC_CNTL_2__SPLL_CTLREQ_CHG_MASK, CG_SPLL_FUNC_CNTL_2__SPLL_CTLREQ_CHG__SHIFT, 0, 0x0 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, 0xC0500170 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, 0x4000000, 0x1a, 0, 0x1 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixMPLL_BYPASSCLK_SEL },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, MPLL_BYPASSCLK_SEL__MPLL_CLKOUT_SEL_MASK, MPLL_BYPASSCLK_SEL__MPLL_CLKOUT_SEL__SHIFT, 0, 0x2 },
+	{ CMD_READMODIFYWRITE, mmMPLL_CNTL_MODE, MPLL_CNTL_MODE__MPLL_SW_DIR_CONTROL_MASK, MPLL_CNTL_MODE__MPLL_SW_DIR_CONTROL__SHIFT, 0, 0x1 },
+	{ CMD_READMODIFYWRITE, mmMPLL_CNTL_MODE, MPLL_CNTL_MODE__MPLL_MCLK_SEL_MASK, MPLL_CNTL_MODE__MPLL_MCLK_SEL__SHIFT, 0, 0x0 }
+};
+
+static const struct baco_cmd_entry turn_off_plls_tbl[] =
+{
+	{ CMD_READMODIFYWRITE, mmDISPPLL_BG_CNTL, DISPPLL_BG_CNTL__DISPPLL_BG_PDN_MASK, DISPPLL_BG_CNTL__DISPPLL_BG_PDN__SHIFT, 0, 0x1 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixCG_CLKPIN_CNTL_DC },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, CG_CLKPIN_CNTL_DC__OSC_EN_MASK, CG_CLKPIN_CNTL_DC__OSC_EN__SHIFT, 0, 0x0 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, CG_CLKPIN_CNTL_DC__XTALIN_SEL_MASK, CG_CLKPIN_CNTL_DC__XTALIN_SEL__SHIFT, 0, 0x0 },
+	{ CMD_READMODIFYWRITE, mmPLL_CNTL, PLL_CNTL__PLL_RESET_MASK, PLL_CNTL__PLL_RESET__SHIFT, 0, 0x1 },
+	{ CMD_READMODIFYWRITE, mmPLL_CNTL, PLL_CNTL__PLL_POWER_DOWN_MASK, PLL_CNTL__PLL_POWER_DOWN__SHIFT, 0, 0x1 },
+	{ CMD_READMODIFYWRITE, mmPLL_CNTL, PLL_CNTL__PLL_BYPASS_CAL_MASK, PLL_CNTL__PLL_BYPASS_CAL__SHIFT, 0, 0x1 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixCG_SPLL_FUNC_CNTL },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, CG_SPLL_FUNC_CNTL__SPLL_RESET_MASK, CG_SPLL_FUNC_CNTL__SPLL_RESET__SHIFT, 0, 0x1 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, CG_SPLL_FUNC_CNTL__SPLL_PWRON_MASK, CG_SPLL_FUNC_CNTL__SPLL_PWRON__SHIFT, 0, 0x0 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, 0xC0500170 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, 0x2000000, 0x19, 0, 0x1 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, 0x8000000, 0x1b, 0, 0x0 },
+	{ CMD_READMODIFYWRITE, mmMPLL_CNTL_MODE, MPLL_CNTL_MODE__GLOBAL_MPLL_RESET_MASK, MPLL_CNTL_MODE__GLOBAL_MPLL_RESET__SHIFT, 0, 0x1 },
+	{ CMD_WRITE, mmMPLL_CONTROL, 0, 0, 0, 0x00000006 },
+	{ CMD_WRITE, mmMC_IO_RXCNTL_DPHY0_D0, 0, 0, 0, 0x00007740 },
+	{ CMD_WRITE, mmMC_IO_RXCNTL_DPHY0_D1, 0, 0, 0, 0x00007740 },
+	{ CMD_WRITE, mmMC_IO_RXCNTL_DPHY1_D0, 0, 0, 0, 0x00007740 },
+	{ CMD_WRITE, mmMC_IO_RXCNTL_DPHY1_D1, 0, 0, 0, 0x00007740 },
+	{ CMD_READMODIFYWRITE, mmMCLK_PWRMGT_CNTL, MCLK_PWRMGT_CNTL__MRDCK0_PDNB_MASK, MCLK_PWRMGT_CNTL__MRDCK0_PDNB__SHIFT, 0, 0x0 },
+	{ CMD_READMODIFYWRITE, mmMCLK_PWRMGT_CNTL, MCLK_PWRMGT_CNTL__MRDCK1_PDNB_MASK, MCLK_PWRMGT_CNTL__MRDCK1_PDNB__SHIFT, 0, 0x0 },
+	{ CMD_READMODIFYWRITE, mmMC_SEQ_CNTL_2, MC_SEQ_CNTL_2__DRST_PU_MASK, MC_SEQ_CNTL_2__DRST_PU__SHIFT, 0, 0x0 },
+	{ CMD_READMODIFYWRITE, mmMC_SEQ_CNTL_2, MC_SEQ_CNTL_2__DRST_PD_MASK, MC_SEQ_CNTL_2__DRST_PD__SHIFT, 0, 0x0 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixCG_CLKPIN_CNTL_2 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, CG_CLKPIN_CNTL_2__FORCE_BIF_REFCLK_EN_MASK, CG_CLKPIN_CNTL_2__FORCE_BIF_REFCLK_EN__SHIFT, 0, 0x0 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixMPLL_BYPASSCLK_SEL },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, MPLL_BYPASSCLK_SEL__MPLL_CLKOUT_SEL_MASK, MPLL_BYPASSCLK_SEL__MPLL_CLKOUT_SEL__SHIFT, 0, 0x4 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixMISC_CLK_CTRL },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, MISC_CLK_CTRL__DEEP_SLEEP_CLK_SEL_MASK, MISC_CLK_CTRL__DEEP_SLEEP_CLK_SEL__SHIFT, 0, 0x2 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, MISC_CLK_CTRL__ZCLK_SEL_MASK, MISC_CLK_CTRL__ZCLK_SEL__SHIFT, 0, 0x2 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, MISC_CLK_CTRL__DFT_SMS_PG_CLK_SEL_MASK, MISC_CLK_CTRL__DFT_SMS_PG_CLK_SEL__SHIFT, 0, 0x2 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixTHM_CLK_CNTL },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, THM_CLK_CNTL__CMON_CLK_SEL_MASK, THM_CLK_CNTL__CMON_CLK_SEL__SHIFT, 0, 0x2 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, THM_CLK_CNTL__TMON_CLK_SEL_MASK, THM_CLK_CNTL__TMON_CLK_SEL__SHIFT, 0, 0x2 }
+};
+
+static const struct baco_cmd_entry enter_baco_tbl[] =
+{
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_EN_MASK, BACO_CNTL__BACO_EN__SHIFT, 0, 0x01 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_BCLK_OFF_MASK, BACO_CNTL__BACO_BCLK_OFF__SHIFT, 0, 0x01 },
+	{ CMD_WAITFOR,         mmBACO_CNTL, BACO_CNTL__BACO_BCLK_OFF_MASK, 0, 5, 0x02 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_ISO_DIS_MASK, BACO_CNTL__BACO_ISO_DIS__SHIFT, 0, 0x00 },
+	{ CMD_WAITFOR,         mmBACO_CNTL, BACO_CNTL__BACO_ISO_DIS_MASK, 0, 5,  0x00 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_ANA_ISO_DIS_MASK, BACO_CNTL__BACO_ANA_ISO_DIS__SHIFT, 0, 0x00 },
+	{ CMD_WAITFOR,         mmBACO_CNTL, BACO_CNTL__BACO_ANA_ISO_DIS_MASK, 0, 5, 0x00 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_POWER_OFF_MASK, BACO_CNTL__BACO_POWER_OFF__SHIFT, 0, 0x01 },
+	{ CMD_WAITFOR,         mmBACO_CNTL, BACO_CNTL__BACO_POWER_OFF_MASK, 0, 5, 0x08 },
+	{ CMD_WAITFOR,         mmBACO_CNTL, BACO_CNTL__BACO_MODE_MASK, 0, 0xffffffff, 0x40 }
+};
+
+#define BACO_CNTL__PWRGOOD_MASK  BACO_CNTL__PWRGOOD_GPIO_MASK+BACO_CNTL__PWRGOOD_MEM_MASK+BACO_CNTL__PWRGOOD_DVO_MASK
+
+static const struct baco_cmd_entry exit_baco_tbl[] =
+{
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_RESET_EN_MASK, BACO_CNTL__BACO_RESET_EN__SHIFT, 0, 0x01 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_BCLK_OFF_MASK,           BACO_CNTL__BACO_BCLK_OFF__SHIFT, 0, 0x00 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_POWER_OFF_MASK,          BACO_CNTL__BACO_POWER_OFF__SHIFT, 0, 0x00 },
+	{ CMD_DELAY_MS, 0, 0, 0, 20, 0 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__PWRGOOD_BF_MASK, 0, 0xffffffff, 0x20 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_ISO_DIS_MASK, BACO_CNTL__BACO_ISO_DIS__SHIFT, 0, 0x01 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__PWRGOOD_MASK, 0, 5, 0x1c },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_ANA_ISO_DIS_MASK, BACO_CNTL__BACO_ANA_ISO_DIS__SHIFT, 0, 0x01 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_RESET_EN_MASK, BACO_CNTL__BACO_RESET_EN__SHIFT, 0, 0x00 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__RCU_BIF_CONFIG_DONE_MASK, 0, 5, 0x10 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_EN_MASK, BACO_CNTL__BACO_EN__SHIFT, 0, 0x00 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__BACO_MODE_MASK, 0, 0xffffffff, 0x00 }
+};
+
+static const struct baco_cmd_entry clean_baco_tbl[] =
+{
+	{ CMD_WRITE, mmBIOS_SCRATCH_6, 0, 0, 0, 0 },
+	{ CMD_WRITE, mmCP_PFP_UCODE_ADDR, 0, 0, 0, 0 }
+};
+
+int ci_baco_set_state(struct pp_hwmgr *hwmgr, enum BACO_STATE state)
+{
+	enum BACO_STATE cur_state;
+
+	smu7_baco_get_state(hwmgr, &cur_state);
+
+	if (cur_state == state)
+		/* aisc already in the target state */
+		return 0;
+
+	if (state == BACO_STATE_IN) {
+		baco_program_registers(hwmgr, gpio_tbl, ARRAY_SIZE(gpio_tbl));
+		baco_program_registers(hwmgr, enable_fb_req_rej_tbl,
+				       ARRAY_SIZE(enable_fb_req_rej_tbl));
+		baco_program_registers(hwmgr, use_bclk_tbl, ARRAY_SIZE(use_bclk_tbl));
+		baco_program_registers(hwmgr, turn_off_plls_tbl,
+				       ARRAY_SIZE(turn_off_plls_tbl));
+		if (baco_program_registers(hwmgr, enter_baco_tbl,
+					   ARRAY_SIZE(enter_baco_tbl)))
+			return 0;
+
+	} else if (state == BACO_STATE_OUT) {
+		/* HW requires at least 20ms between regulator off and on */
+		msleep(20);
+		/* Execute Hardware BACO exit sequence */
+		if (baco_program_registers(hwmgr, exit_baco_tbl,
+					   ARRAY_SIZE(exit_baco_tbl))) {
+			if (baco_program_registers(hwmgr, clean_baco_tbl,
+						   ARRAY_SIZE(clean_baco_tbl)))
+				return 0;
+		}
+	}
+
+	return -EINVAL;
+}
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/ci_baco.h b/drivers/gpu/drm/amd/powerplay/hwmgr/ci_baco.h
new file mode 100644
index 000000000000..17041f187020
--- /dev/null
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/ci_baco.h
@@ -0,0 +1,29 @@
+/*
+ * Copyright 2019 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#ifndef __CI_BACO_H__
+#define __CI_BACO_H__
+#include "smu7_baco.h"
+
+extern int ci_baco_set_state(struct pp_hwmgr *hwmgr, enum BACO_STATE state);
+
+#endif
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/common_baco.c b/drivers/gpu/drm/amd/powerplay/hwmgr/common_baco.c
index 9c57c1f67749..1c73776bd606 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/common_baco.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/common_baco.c
@@ -79,6 +79,25 @@ static bool baco_cmd_handler(struct pp_hwmgr *hwmgr, u32 command, u32 reg, u32 m
 	return ret;
 }
 
+bool baco_program_registers(struct pp_hwmgr *hwmgr,
+			    const struct baco_cmd_entry *entry,
+			    const u32 array_size)
+{
+	u32 i, reg = 0;
+
+	for (i = 0; i < array_size; i++) {
+		if ((entry[i].cmd == CMD_WRITE) ||
+		    (entry[i].cmd == CMD_READMODIFYWRITE) ||
+		    (entry[i].cmd == CMD_WAITFOR))
+			reg = entry[i].reg_offset;
+		if (!baco_cmd_handler(hwmgr, entry[i].cmd, reg, entry[i].mask,
+				     entry[i].shift, entry[i].val, entry[i].timeout))
+			return false;
+	}
+
+	return true;
+}
+
 bool soc15_baco_program_registers(struct pp_hwmgr *hwmgr,
 				 const struct soc15_baco_cmd_entry *entry,
 				 const u32 array_size)
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/common_baco.h b/drivers/gpu/drm/amd/powerplay/hwmgr/common_baco.h
index 95296c916f4e..8393eb62706d 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/common_baco.h
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/common_baco.h
@@ -33,6 +33,15 @@ enum baco_cmd_type {
 	CMD_DELAY_US,
 };
 
+struct baco_cmd_entry {
+	enum baco_cmd_type cmd;
+	uint32_t 	reg_offset;
+	uint32_t     	mask;
+	uint32_t     	shift;
+	uint32_t     	timeout;
+	uint32_t     	val;
+};
+
 struct soc15_baco_cmd_entry {
 	enum baco_cmd_type cmd;
 	uint32_t 	hwip;
@@ -44,6 +53,10 @@ struct soc15_baco_cmd_entry {
 	uint32_t     	timeout;
 	uint32_t     	val;
 };
+
+extern bool baco_program_registers(struct pp_hwmgr *hwmgr,
+				   const struct baco_cmd_entry *entry,
+				   const u32 array_size);
 extern bool soc15_baco_program_registers(struct pp_hwmgr *hwmgr,
 					const struct soc15_baco_cmd_entry *entry,
 					const u32 array_size);
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/fiji_baco.c b/drivers/gpu/drm/amd/powerplay/hwmgr/fiji_baco.c
new file mode 100644
index 000000000000..c0368f2dfb21
--- /dev/null
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/fiji_baco.c
@@ -0,0 +1,196 @@
+/*
+ * Copyright 2019 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#include "amdgpu.h"
+#include "fiji_baco.h"
+
+#include "gmc/gmc_8_1_d.h"
+#include "gmc/gmc_8_1_sh_mask.h"
+
+#include "bif/bif_5_0_d.h"
+#include "bif/bif_5_0_sh_mask.h"
+
+#include "dce/dce_10_0_d.h"
+#include "dce/dce_10_0_sh_mask.h"
+
+#include "smu/smu_7_1_3_d.h"
+#include "smu/smu_7_1_3_sh_mask.h"
+
+
+static const struct baco_cmd_entry gpio_tbl[] =
+{
+	{ CMD_WRITE, mmGPIOPAD_EN, 0, 0, 0, 0x0 },
+	{ CMD_WRITE, mmGPIOPAD_PD_EN, 0, 0, 0, 0x0 },
+	{ CMD_WRITE, mmGPIOPAD_PU_EN, 0, 0, 0, 0x0 },
+	{ CMD_WRITE, mmGPIOPAD_MASK, 0, 0, 0, 0xff77ffff },
+	{ CMD_WRITE, mmDC_GPIO_DVODATA_EN, 0, 0, 0, 0x0 },
+	{ CMD_WRITE, mmDC_GPIO_DVODATA_MASK, 0, 0, 0, 0xffffffff },
+	{ CMD_WRITE, mmDC_GPIO_GENERIC_EN, 0, 0, 0, 0x0 },
+	{ CMD_READMODIFYWRITE, mmDC_GPIO_GENERIC_MASK, 0, 0, 0, 0x03333333 },
+	{ CMD_WRITE, mmDC_GPIO_SYNCA_EN, 0, 0, 0, 0x0 },
+	{ CMD_READMODIFYWRITE, mmDC_GPIO_SYNCA_MASK, 0, 0, 0, 0x00001111 }
+};
+
+static const struct baco_cmd_entry enable_fb_req_rej_tbl[] =
+{
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, 0xC0300024 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, 0x1, 0x0, 0, 0x1 },
+	{ CMD_WRITE, mmBIF_FB_EN, 0, 0, 0, 0x0 }
+};
+
+static const struct baco_cmd_entry use_bclk_tbl[] =
+{
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixCG_SPLL_FUNC_CNTL },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, CG_SPLL_FUNC_CNTL__SPLL_BYPASS_EN_MASK, CG_SPLL_FUNC_CNTL__SPLL_BYPASS_EN__SHIFT, 0, 0x1 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixCG_SPLL_FUNC_CNTL_2 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, CG_SPLL_FUNC_CNTL_2__SPLL_BYPASS_CHG_MASK, CG_SPLL_FUNC_CNTL_2__SPLL_BYPASS_CHG__SHIFT, 0, 0x1 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixCG_SPLL_STATUS },
+	{ CMD_WAITFOR, mmGCK_SMC_IND_DATA, CG_SPLL_STATUS__SPLL_CHG_STATUS_MASK, 0, 0xffffffff, 0x2 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixCG_SPLL_FUNC_CNTL_2 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, CG_SPLL_FUNC_CNTL_2__SPLL_BYPASS_CHG_MASK, CG_SPLL_FUNC_CNTL_2__SPLL_BYPASS_CHG__SHIFT, 0,  0x0 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, CG_SPLL_FUNC_CNTL_2__SPLL_CTLREQ_CHG_MASK, CG_SPLL_FUNC_CNTL_2__SPLL_CTLREQ_CHG__SHIFT, 0,  0x1 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixCG_SPLL_STATUS },
+	{ CMD_WAITFOR, mmGCK_SMC_IND_DATA, CG_SPLL_STATUS__SPLL_CHG_STATUS_MASK, 0, 0xffffffff, 0x2 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixCG_SPLL_FUNC_CNTL_2 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, CG_SPLL_FUNC_CNTL_2__SPLL_CTLREQ_CHG_MASK, CG_SPLL_FUNC_CNTL_2__SPLL_CTLREQ_CHG__SHIFT, 0, 0x0 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, 0xC0500170 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, 0x4000000, 0x1a, 0, 0x1 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixMPLL_BYPASSCLK_SEL },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, MPLL_BYPASSCLK_SEL__MPLL_CLKOUT_SEL_MASK, MPLL_BYPASSCLK_SEL__MPLL_CLKOUT_SEL__SHIFT, 0, 0x2 }
+};
+
+static const struct baco_cmd_entry turn_off_plls_tbl[] =
+{
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixCG_SPLL_FUNC_CNTL },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, CG_SPLL_FUNC_CNTL__SPLL_RESET_MASK, CG_SPLL_FUNC_CNTL__SPLL_RESET__SHIFT, 0,     0x1 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, CG_SPLL_FUNC_CNTL__SPLL_PWRON_MASK, CG_SPLL_FUNC_CNTL__SPLL_PWRON__SHIFT, 0,     0x0 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, 0xC0500170 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, 0x2000000, 0x19, 0, 0x1 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, 0x8000000, 0x1b, 0, 0x0 }
+};
+
+static const struct baco_cmd_entry clk_req_b_tbl[] =
+{
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixCG_CLKPIN_CNTL_2 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, CG_CLKPIN_CNTL_2__FORCE_BIF_REFCLK_EN_MASK, CG_CLKPIN_CNTL_2__FORCE_BIF_REFCLK_EN__SHIFT, 0, 0x0 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixMPLL_BYPASSCLK_SEL },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, MPLL_BYPASSCLK_SEL__MPLL_CLKOUT_SEL_MASK, MPLL_BYPASSCLK_SEL__MPLL_CLKOUT_SEL__SHIFT, 0, 0x4 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixMISC_CLK_CTRL },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, MISC_CLK_CTRL__DEEP_SLEEP_CLK_SEL_MASK, MISC_CLK_CTRL__DEEP_SLEEP_CLK_SEL__SHIFT, 0, 0x1 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, MISC_CLK_CTRL__ZCLK_SEL_MASK, MISC_CLK_CTRL__ZCLK_SEL__SHIFT, 0, 0x1 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixCG_CLKPIN_CNTL },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, CG_CLKPIN_CNTL__BCLK_AS_XCLK_MASK, CG_CLKPIN_CNTL__BCLK_AS_XCLK__SHIFT, 0, 0x0 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixTHM_CLK_CNTL },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, THM_CLK_CNTL__CMON_CLK_SEL_MASK, THM_CLK_CNTL__CMON_CLK_SEL__SHIFT, 0, 0x1 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, THM_CLK_CNTL__TMON_CLK_SEL_MASK, THM_CLK_CNTL__TMON_CLK_SEL__SHIFT, 0, 0x1 }
+};
+
+static const struct baco_cmd_entry enter_baco_tbl[] =
+{
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_EN_MASK, BACO_CNTL__BACO_EN__SHIFT, 0, 0x01 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_BIF_SCLK_SWITCH_MASK, BACO_CNTL__BACO_BIF_SCLK_SWITCH__SHIFT, 0, 0x01 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__BACO_BIF_SCLK_SWITCH_MASK, 0, 5, 0x40000 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_BCLK_OFF_MASK, BACO_CNTL__BACO_BCLK_OFF__SHIFT, 0, 0x01 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__BACO_BCLK_OFF_MASK, 0, 5, 0x02 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_ISO_DIS_MASK, BACO_CNTL__BACO_ISO_DIS__SHIFT, 0, 0x00 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__BACO_ISO_DIS_MASK, 0, 5, 0x00 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_ANA_ISO_DIS_MASK, BACO_CNTL__BACO_ANA_ISO_DIS__SHIFT, 0, 0x00 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__BACO_ANA_ISO_DIS_MASK, 0, 5, 0x00 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_POWER_OFF_MASK, BACO_CNTL__BACO_POWER_OFF__SHIFT, 0, 0x01 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__BACO_POWER_OFF_MASK, 0, 5, 0x08 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__BACO_MODE_MASK, 0, 0xffffffff, 0x40 }
+};
+
+#define BACO_CNTL__PWRGOOD_MASK  BACO_CNTL__PWRGOOD_GPIO_MASK+BACO_CNTL__PWRGOOD_MEM_MASK+BACO_CNTL__PWRGOOD_DVO_MASK
+
+static const struct baco_cmd_entry exit_baco_tbl[] =
+{
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_RESET_EN_MASK, BACO_CNTL__BACO_RESET_EN__SHIFT, 0, 0x01 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_BCLK_OFF_MASK, BACO_CNTL__BACO_BCLK_OFF__SHIFT, 0, 0x00 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_POWER_OFF_MASK, BACO_CNTL__BACO_POWER_OFF__SHIFT, 0, 0x00 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__PWRGOOD_BF_MASK, 0, 0xffffffff, 0x200 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_ISO_DIS_MASK, BACO_CNTL__BACO_ISO_DIS__SHIFT, 0, 0x01 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__PWRGOOD_MASK, 0, 5, 0x1c00 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_ANA_ISO_DIS_MASK, BACO_CNTL__BACO_ANA_ISO_DIS__SHIFT, 0, 0x01 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_BIF_SCLK_SWITCH_MASK, BACO_CNTL__BACO_BIF_SCLK_SWITCH__SHIFT, 0, 0x00 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_RESET_EN_MASK, BACO_CNTL__BACO_RESET_EN__SHIFT, 0, 0x00 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__RCU_BIF_CONFIG_DONE_MASK, 0, 5, 0x100 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_EN_MASK, BACO_CNTL__BACO_EN__SHIFT, 0, 0x00 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__BACO_MODE_MASK, 0, 0xffffffff, 0x00 }
+};
+
+static const struct baco_cmd_entry clean_baco_tbl[] =
+{
+	{ CMD_WRITE, mmBIOS_SCRATCH_0, 0, 0, 0, 0 },
+	{ CMD_WRITE, mmBIOS_SCRATCH_1, 0, 0, 0, 0 },
+	{ CMD_WRITE, mmBIOS_SCRATCH_2, 0, 0, 0, 0 },
+	{ CMD_WRITE, mmBIOS_SCRATCH_3, 0, 0, 0, 0 },
+	{ CMD_WRITE, mmBIOS_SCRATCH_4, 0, 0, 0, 0 },
+	{ CMD_WRITE, mmBIOS_SCRATCH_5, 0, 0, 0, 0 },
+	{ CMD_WRITE, mmBIOS_SCRATCH_6, 0, 0, 0, 0 },
+	{ CMD_WRITE, mmBIOS_SCRATCH_7, 0, 0, 0, 0 },
+	{ CMD_WRITE, mmBIOS_SCRATCH_8, 0, 0, 0, 0 },
+	{ CMD_WRITE, mmBIOS_SCRATCH_9, 0, 0, 0, 0 },
+	{ CMD_WRITE, mmBIOS_SCRATCH_10, 0, 0, 0, 0 },
+	{ CMD_WRITE, mmBIOS_SCRATCH_11, 0, 0, 0, 0 },
+	{ CMD_WRITE, mmBIOS_SCRATCH_12, 0, 0, 0, 0 },
+	{ CMD_WRITE, mmBIOS_SCRATCH_13, 0, 0, 0, 0 },
+	{ CMD_WRITE, mmBIOS_SCRATCH_14, 0, 0, 0, 0 },
+	{ CMD_WRITE, mmBIOS_SCRATCH_15, 0, 0, 0, 0 }
+};
+
+int fiji_baco_set_state(struct pp_hwmgr *hwmgr, enum BACO_STATE state)
+{
+	enum BACO_STATE cur_state;
+
+	smu7_baco_get_state(hwmgr, &cur_state);
+
+	if (cur_state == state)
+		/* aisc already in the target state */
+		return 0;
+
+	if (state == BACO_STATE_IN) {
+		baco_program_registers(hwmgr, gpio_tbl, ARRAY_SIZE(gpio_tbl));
+		baco_program_registers(hwmgr, enable_fb_req_rej_tbl,
+				       ARRAY_SIZE(enable_fb_req_rej_tbl));
+		baco_program_registers(hwmgr, use_bclk_tbl, ARRAY_SIZE(use_bclk_tbl));
+		baco_program_registers(hwmgr, turn_off_plls_tbl,
+				       ARRAY_SIZE(turn_off_plls_tbl));
+		baco_program_registers(hwmgr, clk_req_b_tbl, ARRAY_SIZE(clk_req_b_tbl));
+		if (baco_program_registers(hwmgr, enter_baco_tbl,
+					   ARRAY_SIZE(enter_baco_tbl)))
+			return 0;
+
+	} else if (state == BACO_STATE_OUT) {
+		/* HW requires at least 20ms between regulator off and on */
+		msleep(20);
+		/* Execute Hardware BACO exit sequence */
+		if (baco_program_registers(hwmgr, exit_baco_tbl,
+					   ARRAY_SIZE(exit_baco_tbl))) {
+			if (baco_program_registers(hwmgr, clean_baco_tbl,
+						   ARRAY_SIZE(clean_baco_tbl)))
+				return 0;
+		}
+	}
+
+	return -EINVAL;
+}
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/fiji_baco.h b/drivers/gpu/drm/amd/powerplay/hwmgr/fiji_baco.h
new file mode 100644
index 000000000000..47f402900bdb
--- /dev/null
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/fiji_baco.h
@@ -0,0 +1,29 @@
+/*
+ * Copyright 2019 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#ifndef __FIJI_BACO_H__
+#define __FIJI_BACO_H__
+#include "smu7_baco.h"
+
+extern int fiji_baco_set_state(struct pp_hwmgr *hwmgr, enum BACO_STATE state);
+
+#endif
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c
index a24beaa4fb01..d2909c91d65b 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c
@@ -81,6 +81,8 @@ static void hwmgr_init_workload_prority(struct pp_hwmgr *hwmgr)
 
 int hwmgr_early_init(struct pp_hwmgr *hwmgr)
 {
+	struct amdgpu_device *adev;
+
 	if (!hwmgr)
 		return -EINVAL;
 
@@ -94,8 +96,11 @@ int hwmgr_early_init(struct pp_hwmgr *hwmgr)
 	hwmgr_init_workload_prority(hwmgr);
 	hwmgr->gfxoff_state_changed_by_workload = false;
 
+	adev = hwmgr->adev;
+
 	switch (hwmgr->chip_family) {
 	case AMDGPU_FAMILY_CI:
+		adev->pm.pp_feature &= ~PP_GFXOFF_MASK;
 		hwmgr->smumgr_funcs = &ci_smu_funcs;
 		ci_set_asic_special_caps(hwmgr);
 		hwmgr->feature_mask &= ~(PP_VBI_TIME_SUPPORT_MASK |
@@ -106,12 +111,14 @@ int hwmgr_early_init(struct pp_hwmgr *hwmgr)
 		smu7_init_function_pointers(hwmgr);
 		break;
 	case AMDGPU_FAMILY_CZ:
+		adev->pm.pp_feature &= ~PP_GFXOFF_MASK;
 		hwmgr->od_enabled = false;
 		hwmgr->smumgr_funcs = &smu8_smu_funcs;
 		hwmgr->feature_mask &= ~PP_GFXOFF_MASK;
 		smu8_init_function_pointers(hwmgr);
 		break;
 	case AMDGPU_FAMILY_VI:
+		adev->pm.pp_feature &= ~PP_GFXOFF_MASK;
 		hwmgr->feature_mask &= ~PP_GFXOFF_MASK;
 		switch (hwmgr->chip_id) {
 		case CHIP_TOPAZ:
@@ -153,6 +160,7 @@ int hwmgr_early_init(struct pp_hwmgr *hwmgr)
 	case AMDGPU_FAMILY_AI:
 		switch (hwmgr->chip_id) {
 		case CHIP_VEGA10:
+			adev->pm.pp_feature &= ~PP_GFXOFF_MASK;
 			hwmgr->feature_mask &= ~PP_GFXOFF_MASK;
 			hwmgr->smumgr_funcs = &vega10_smu_funcs;
 			vega10_hwmgr_init(hwmgr);
@@ -162,6 +170,7 @@ int hwmgr_early_init(struct pp_hwmgr *hwmgr)
 			vega12_hwmgr_init(hwmgr);
 			break;
 		case CHIP_VEGA20:
+			adev->pm.pp_feature &= ~PP_GFXOFF_MASK;
 			hwmgr->feature_mask &= ~PP_GFXOFF_MASK;
 			hwmgr->smumgr_funcs = &vega20_smu_funcs;
 			vega20_hwmgr_init(hwmgr);
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/polaris_baco.c b/drivers/gpu/drm/amd/powerplay/hwmgr/polaris_baco.c
new file mode 100644
index 000000000000..8f8e296f2fe9
--- /dev/null
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/polaris_baco.c
@@ -0,0 +1,222 @@
+/*
+ * Copyright 2019 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#include "amdgpu.h"
+#include "polaris_baco.h"
+
+#include "gmc/gmc_8_1_d.h"
+#include "gmc/gmc_8_1_sh_mask.h"
+
+#include "bif/bif_5_0_d.h"
+#include "bif/bif_5_0_sh_mask.h"
+
+#include "dce/dce_11_0_d.h"
+#include "dce/dce_11_0_sh_mask.h"
+
+#include "smu/smu_7_1_3_d.h"
+#include "smu/smu_7_1_3_sh_mask.h"
+
+static const struct baco_cmd_entry gpio_tbl[] =
+{
+	{ CMD_WRITE, mmGPIOPAD_EN, 0, 0, 0, 0x0 },
+	{ CMD_WRITE, mmGPIOPAD_PD_EN, 0, 0, 0, 0x0 },
+	{ CMD_WRITE, mmGPIOPAD_PU_EN, 0, 0, 0, 0x0 },
+	{ CMD_WRITE, mmGPIOPAD_MASK, 0, 0, 0, 0xff77ffff },
+	{ CMD_WRITE, mmDC_GPIO_DVODATA_EN, 0, 0, 0, 0x0 },
+	{ CMD_WRITE, mmDC_GPIO_DVODATA_MASK, 0, 0, 0, 0xffffffff },
+	{ CMD_WRITE, mmDC_GPIO_GENERIC_EN, 0, 0, 0, 0x0 },
+	{ CMD_READMODIFYWRITE, mmDC_GPIO_GENERIC_MASK, 0, 0, 0, 0x03333333 },
+	{ CMD_WRITE, mmDC_GPIO_SYNCA_EN, 0, 0, 0, 0x0 },
+	{ CMD_READMODIFYWRITE, mmDC_GPIO_SYNCA_MASK, 0, 0, 0, 0x00001111 }
+};
+
+static const struct baco_cmd_entry enable_fb_req_rej_tbl[] =
+{
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, 0xC0300024 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, 0x1, 0x0, 0, 0x1 },
+	{ CMD_WRITE, mmBIF_FB_EN, 0, 0, 0, 0x0 }
+};
+
+static const struct baco_cmd_entry use_bclk_tbl[] =
+{
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixCG_SPLL_FUNC_CNTL },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, CG_SPLL_FUNC_CNTL__SPLL_BYPASS_EN_MASK, CG_SPLL_FUNC_CNTL__SPLL_BYPASS_EN__SHIFT, 0, 0x1 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, 0xC0500170 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, 0x4000000, 0x1a, 0, 0x1 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixGCK_DFS_BYPASS_CNTL },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, GCK_DFS_BYPASS_CNTL__BYPASSACLK_MASK, GCK_DFS_BYPASS_CNTL__BYPASSACLK__SHIFT, 0, 0x1 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixMPLL_BYPASSCLK_SEL },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, MPLL_BYPASSCLK_SEL__MPLL_CLKOUT_SEL_MASK, MPLL_BYPASSCLK_SEL__MPLL_CLKOUT_SEL__SHIFT, 0, 0x2 },
+	{ CMD_READMODIFYWRITE, mmMPLL_CNTL_MODE, MPLL_CNTL_MODE__MPLL_SW_DIR_CONTROL_MASK, MPLL_CNTL_MODE__MPLL_SW_DIR_CONTROL__SHIFT, 0, 0x1 },
+	{ CMD_READMODIFYWRITE, mmMPLL_CNTL_MODE, MPLL_CNTL_MODE__MPLL_MCLK_SEL_MASK, MPLL_CNTL_MODE__MPLL_MCLK_SEL__SHIFT, 0, 0x0 }
+};
+
+static const struct baco_cmd_entry turn_off_plls_tbl[] =
+{
+	{ CMD_READMODIFYWRITE, mmDC_GPIO_PAD_STRENGTH_1, DC_GPIO_PAD_STRENGTH_1__GENLK_STRENGTH_SP_MASK, DC_GPIO_PAD_STRENGTH_1__GENLK_STRENGTH_SP__SHIFT, 0, 0x1 },
+	{ CMD_DELAY_US, 0, 0, 0, 1, 0x0 },
+	{ CMD_READMODIFYWRITE, mmMC_SEQ_DRAM, MC_SEQ_DRAM__RST_CTL_MASK, MC_SEQ_DRAM__RST_CTL__SHIFT, 0, 0x1 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, 0xC05002B0 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, 0x10, 0x4, 0, 0x1 },
+	{ CMD_WAITFOR, mmGCK_SMC_IND_DATA, 0x10, 0, 1, 0 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, 0xC050032C },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, 0x10, 0x4, 0, 0x1 },
+	{ CMD_WAITFOR, mmGCK_SMC_IND_DATA, 0x10, 0, 1, 0 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, 0xC0500080 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, 0x1, 0x0, 0, 0x1 },
+	{ CMD_READMODIFYWRITE, 0xda2, 0x40, 0x6, 0, 0x0 },
+	{ CMD_DELAY_US, 0, 0, 0, 3, 0x0 },
+	{ CMD_READMODIFYWRITE, 0xda2, 0x8, 0x3, 0, 0x0 },
+	{ CMD_READMODIFYWRITE, 0xda2, 0x3fff00, 0x8, 0, 0x32 },
+	{ CMD_DELAY_US, 0, 0, 0, 3, 0x0 },
+	{ CMD_READMODIFYWRITE, mmMPLL_FUNC_CNTL_2, MPLL_FUNC_CNTL_2__ISO_DIS_P_MASK, MPLL_FUNC_CNTL_2__ISO_DIS_P__SHIFT, 0, 0x0 },
+	{ CMD_DELAY_US, 0, 0, 0, 5, 0x0 }
+};
+
+static const struct baco_cmd_entry clk_req_b_tbl[] =
+{
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixTHM_CLK_CNTL },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, THM_CLK_CNTL__CMON_CLK_SEL_MASK, THM_CLK_CNTL__CMON_CLK_SEL__SHIFT, 0, 0x1 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, THM_CLK_CNTL__TMON_CLK_SEL_MASK, THM_CLK_CNTL__TMON_CLK_SEL__SHIFT, 0, 0x1 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixMISC_CLK_CTRL },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, MISC_CLK_CTRL__DEEP_SLEEP_CLK_SEL_MASK, MISC_CLK_CTRL__DEEP_SLEEP_CLK_SEL__SHIFT, 0, 0x1 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, MISC_CLK_CTRL__ZCLK_SEL_MASK, MISC_CLK_CTRL__ZCLK_SEL__SHIFT, 0, 0x1 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixCG_CLKPIN_CNTL },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, CG_CLKPIN_CNTL__BCLK_AS_XCLK_MASK, CG_CLKPIN_CNTL__BCLK_AS_XCLK__SHIFT, 0, 0x0 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixCG_CLKPIN_CNTL_2 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, CG_CLKPIN_CNTL_2__FORCE_BIF_REFCLK_EN_MASK, CG_CLKPIN_CNTL_2__FORCE_BIF_REFCLK_EN__SHIFT, 0, 0x0 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixMPLL_BYPASSCLK_SEL },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, MPLL_BYPASSCLK_SEL__MPLL_CLKOUT_SEL_MASK, MPLL_BYPASSCLK_SEL__MPLL_CLKOUT_SEL__SHIFT, 0,     0x4 }
+};
+
+static const struct baco_cmd_entry enter_baco_tbl[] =
+{
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_EN_MASK, BACO_CNTL__BACO_EN__SHIFT, 0, 0x01 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_BIF_SCLK_SWITCH_MASK, BACO_CNTL__BACO_BIF_SCLK_SWITCH__SHIFT, 0, 0x01 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__BACO_BIF_SCLK_SWITCH_MASK, 0, 5, 0x40000 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_BCLK_OFF_MASK, BACO_CNTL__BACO_BCLK_OFF__SHIFT, 0, 0x01 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__BACO_BCLK_OFF_MASK, 0, 5, 0x02 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_ISO_DIS_MASK, BACO_CNTL__BACO_ISO_DIS__SHIFT, 0, 0x00 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__BACO_ISO_DIS_MASK, 0, 5, 0x00 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_ANA_ISO_DIS_MASK, BACO_CNTL__BACO_ANA_ISO_DIS__SHIFT, 0, 0x00 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__BACO_ANA_ISO_DIS_MASK, 0, 5, 0x00 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_POWER_OFF_MASK, BACO_CNTL__BACO_POWER_OFF__SHIFT, 0, 0x01 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__BACO_POWER_OFF_MASK, 0, 5, 0x08 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__BACO_MODE_MASK, 0, 0xffffffff, 0x40 }
+};
+
+#define BACO_CNTL__PWRGOOD_MASK  BACO_CNTL__PWRGOOD_GPIO_MASK+BACO_CNTL__PWRGOOD_MEM_MASK+BACO_CNTL__PWRGOOD_DVO_MASK
+
+static const struct baco_cmd_entry exit_baco_tbl[] =
+{
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_RESET_EN_MASK, BACO_CNTL__BACO_RESET_EN__SHIFT, 0, 0x01 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_BCLK_OFF_MASK, BACO_CNTL__BACO_BCLK_OFF__SHIFT, 0, 0x00 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_POWER_OFF_MASK, BACO_CNTL__BACO_POWER_OFF__SHIFT, 0, 0x00 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__PWRGOOD_BF_MASK, 0, 0xffffffff, 0x200 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_ISO_DIS_MASK, BACO_CNTL__BACO_ISO_DIS__SHIFT, 0, 0x01 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__PWRGOOD_MASK, 0, 5, 0x1c00 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_ANA_ISO_DIS_MASK, BACO_CNTL__BACO_ANA_ISO_DIS__SHIFT, 0, 0x01 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_BIF_SCLK_SWITCH_MASK, BACO_CNTL__BACO_BIF_SCLK_SWITCH__SHIFT, 0, 0x00 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_RESET_EN_MASK, BACO_CNTL__BACO_RESET_EN__SHIFT, 0, 0x00 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__RCU_BIF_CONFIG_DONE_MASK, 0, 5, 0x100 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_EN_MASK, BACO_CNTL__BACO_EN__SHIFT, 0, 0x00 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__BACO_MODE_MASK, 0, 0xffffffff, 0x00 }
+};
+
+static const struct baco_cmd_entry clean_baco_tbl[] =
+{
+	{ CMD_WRITE, mmBIOS_SCRATCH_6, 0, 0, 0, 0 },
+	{ CMD_WRITE, mmBIOS_SCRATCH_7, 0, 0, 0, 0 }
+};
+
+static const struct baco_cmd_entry use_bclk_tbl_vg[] =
+{
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixCG_SPLL_FUNC_CNTL },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, CG_SPLL_FUNC_CNTL__SPLL_BYPASS_EN_MASK, CG_SPLL_FUNC_CNTL__SPLL_BYPASS_EN__SHIFT, 0, 0x1 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, 0xC0500170 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, 0x4000000, 0x1a, 0, 0x1 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixGCK_DFS_BYPASS_CNTL },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, GCK_DFS_BYPASS_CNTL__BYPASSACLK_MASK, GCK_DFS_BYPASS_CNTL__BYPASSACLK__SHIFT, 0, 0x1 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixMPLL_BYPASSCLK_SEL },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, MPLL_BYPASSCLK_SEL__MPLL_CLKOUT_SEL_MASK, MPLL_BYPASSCLK_SEL__MPLL_CLKOUT_SEL__SHIFT, 0, 0x2 }
+};
+
+static const struct baco_cmd_entry turn_off_plls_tbl_vg[] =
+{
+	{ CMD_READMODIFYWRITE, mmDC_GPIO_PAD_STRENGTH_1, DC_GPIO_PAD_STRENGTH_1__GENLK_STRENGTH_SP_MASK, DC_GPIO_PAD_STRENGTH_1__GENLK_STRENGTH_SP__SHIFT, 0, 0x1 },
+	{ CMD_DELAY_US, 0, 0, 0, 1, 0x0 },
+	{ CMD_READMODIFYWRITE, mmMC_SEQ_DRAM, MC_SEQ_DRAM__RST_CTL_MASK, MC_SEQ_DRAM__RST_CTL__SHIFT, 0, 0x1 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, 0xC05002B0 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, 0x10, 0x4, 0, 0x1 },
+	{ CMD_WAITFOR, mmGCK_SMC_IND_DATA, 0x10, 0, 1, 0 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, 0xC050032C },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, 0x10, 0x4, 0, 0x1 },
+	{ CMD_WAITFOR, mmGCK_SMC_IND_DATA, 0x10, 0, 1, 0 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, 0xC0500080 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, 0x1, 0x0, 0, 0x1 },
+	{ CMD_DELAY_US, 0, 0, 0, 3, 0x0 },
+	{ CMD_DELAY_US, 0, 0, 0, 3, 0x0 },
+	{ CMD_DELAY_US, 0, 0, 0, 5, 0x0 }
+};
+
+int polaris_baco_set_state(struct pp_hwmgr *hwmgr, enum BACO_STATE state)
+{
+	enum BACO_STATE cur_state;
+
+	smu7_baco_get_state(hwmgr, &cur_state);
+
+	if (cur_state == state)
+		/* aisc already in the target state */
+		return 0;
+
+	if (state == BACO_STATE_IN) {
+		baco_program_registers(hwmgr, gpio_tbl, ARRAY_SIZE(gpio_tbl));
+		baco_program_registers(hwmgr, enable_fb_req_rej_tbl,
+				       ARRAY_SIZE(enable_fb_req_rej_tbl));
+		if (hwmgr->chip_id == CHIP_VEGAM) {
+			baco_program_registers(hwmgr, use_bclk_tbl_vg, ARRAY_SIZE(use_bclk_tbl_vg));
+			baco_program_registers(hwmgr, turn_off_plls_tbl_vg,
+					       ARRAY_SIZE(turn_off_plls_tbl_vg));
+		} else {
+			baco_program_registers(hwmgr, use_bclk_tbl, ARRAY_SIZE(use_bclk_tbl));
+			baco_program_registers(hwmgr, turn_off_plls_tbl,
+					       ARRAY_SIZE(turn_off_plls_tbl));
+		}
+		baco_program_registers(hwmgr, clk_req_b_tbl, ARRAY_SIZE(clk_req_b_tbl));
+		if (baco_program_registers(hwmgr, enter_baco_tbl,
+					   ARRAY_SIZE(enter_baco_tbl)))
+			return 0;
+
+	} else if (state == BACO_STATE_OUT) {
+		/* HW requires at least 20ms between regulator off and on */
+		msleep(20);
+		/* Execute Hardware BACO exit sequence */
+		if (baco_program_registers(hwmgr, exit_baco_tbl,
+					   ARRAY_SIZE(exit_baco_tbl))) {
+			if (baco_program_registers(hwmgr, clean_baco_tbl,
+						   ARRAY_SIZE(clean_baco_tbl)))
+				return 0;
+		}
+	}
+
+	return -EINVAL;
+}
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/polaris_baco.h b/drivers/gpu/drm/amd/powerplay/hwmgr/polaris_baco.h
new file mode 100644
index 000000000000..87a5fa0a157a
--- /dev/null
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/polaris_baco.h
@@ -0,0 +1,29 @@
+/*
+ * Copyright 2019 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#ifndef __POLARIS_BACO_H__
+#define __POLARIS_BACO_H__
+#include "smu7_baco.h"
+
+extern int polaris_baco_set_state(struct pp_hwmgr *hwmgr, enum BACO_STATE state);
+
+#endif
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_baco.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_baco.c
new file mode 100644
index 000000000000..044cda005aed
--- /dev/null
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_baco.c
@@ -0,0 +1,91 @@
+/*
+ * Copyright 2019 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#include "amdgpu.h"
+#include "smu7_baco.h"
+#include "tonga_baco.h"
+#include "fiji_baco.h"
+#include "polaris_baco.h"
+#include "ci_baco.h"
+
+#include "bif/bif_5_0_d.h"
+#include "bif/bif_5_0_sh_mask.h"
+
+#include "smu/smu_7_1_2_d.h"
+#include "smu/smu_7_1_2_sh_mask.h"
+
+int smu7_baco_get_capability(struct pp_hwmgr *hwmgr, bool *cap)
+{
+	struct amdgpu_device *adev = (struct amdgpu_device *)(hwmgr->adev);
+	uint32_t reg;
+
+	*cap = false;
+	if (!phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_BACO))
+		return 0;
+
+	reg = RREG32(mmCC_BIF_BX_FUSESTRAP0);
+
+	if (reg & CC_BIF_BX_FUSESTRAP0__STRAP_BIF_PX_CAPABLE_MASK)
+		*cap = true;
+
+	return 0;
+}
+
+int smu7_baco_get_state(struct pp_hwmgr *hwmgr, enum BACO_STATE *state)
+{
+	struct amdgpu_device *adev = (struct amdgpu_device *)(hwmgr->adev);
+	uint32_t reg;
+
+	reg = RREG32(mmBACO_CNTL);
+
+	if (reg & BACO_CNTL__BACO_MODE_MASK)
+		/* gfx has already entered BACO state */
+		*state = BACO_STATE_IN;
+	else
+		*state = BACO_STATE_OUT;
+	return 0;
+}
+
+int smu7_baco_set_state(struct pp_hwmgr *hwmgr, enum BACO_STATE state)
+{
+	struct amdgpu_device *adev = (struct amdgpu_device *)(hwmgr->adev);
+
+	switch (adev->asic_type) {
+	case CHIP_TOPAZ:
+	case CHIP_TONGA:
+		return tonga_baco_set_state(hwmgr, state);
+	case CHIP_FIJI:
+		return fiji_baco_set_state(hwmgr, state);
+	case CHIP_POLARIS10:
+	case CHIP_POLARIS11:
+	case CHIP_POLARIS12:
+	case CHIP_VEGAM:
+		return polaris_baco_set_state(hwmgr, state);
+#ifdef CONFIG_DRM_AMDGPU_CIK
+	case CHIP_BONAIRE:
+	case CHIP_HAWAII:
+		return ci_baco_set_state(hwmgr, state);
+#endif
+	default:
+		return -EINVAL;
+	}
+}
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_baco.h b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_baco.h
new file mode 100644
index 000000000000..be0d98abb536
--- /dev/null
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_baco.h
@@ -0,0 +1,32 @@
+/*
+ * Copyright 2019 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#ifndef __SMU7_BACO_H__
+#define __SMU7_BACO_H__
+#include "hwmgr.h"
+#include "common_baco.h"
+
+extern int smu7_baco_get_capability(struct pp_hwmgr *hwmgr, bool *cap);
+extern int smu7_baco_get_state(struct pp_hwmgr *hwmgr, enum BACO_STATE *state);
+extern int smu7_baco_set_state(struct pp_hwmgr *hwmgr, enum BACO_STATE state);
+
+#endif
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
index 34f95e0e3ea4..f73dff68e799 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
@@ -48,6 +48,7 @@
 #include "smu7_clockpowergating.h"
 #include "processpptables.h"
 #include "pp_thermal.h"
+#include "smu7_baco.h"
 
 #include "ivsrcid/ivsrcid_vislands30.h"
 
@@ -1994,7 +1995,6 @@ static int smu7_sort_lookup_table(struct pp_hwmgr *hwmgr,
 		struct phm_ppt_v1_voltage_lookup_table *lookup_table)
 {
 	uint32_t table_size, i, j;
-	struct phm_ppt_v1_voltage_lookup_record tmp_voltage_lookup_record;
 	table_size = lookup_table->count;
 
 	PP_ASSERT_WITH_CODE(0 != lookup_table->count,
@@ -2005,9 +2005,8 @@ static int smu7_sort_lookup_table(struct pp_hwmgr *hwmgr,
 		for (j = i + 1; j > 0; j--) {
 			if (lookup_table->entries[j].us_vdd <
 					lookup_table->entries[j - 1].us_vdd) {
-				tmp_voltage_lookup_record = lookup_table->entries[j - 1];
-				lookup_table->entries[j - 1] = lookup_table->entries[j];
-				lookup_table->entries[j] = tmp_voltage_lookup_record;
+				swap(lookup_table->entries[j - 1],
+				     lookup_table->entries[j]);
 			}
 		}
 	}
@@ -3478,18 +3477,31 @@ static int smu7_get_pp_table_entry(struct pp_hwmgr *hwmgr,
 
 static int smu7_get_gpu_power(struct pp_hwmgr *hwmgr, u32 *query)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	int i;
 	u32 tmp = 0;
 
 	if (!query)
 		return -EINVAL;
 
-	smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_GetCurrPkgPwr, 0);
-	tmp = cgs_read_register(hwmgr->device, mmSMC_MSG_ARG_0);
-	*query = tmp;
+	/*
+	 * PPSMC_MSG_GetCurrPkgPwr is not supported on:
+	 *  - Hawaii
+	 *  - Bonaire
+	 *  - Fiji
+	 *  - Tonga
+	 */
+	if ((adev->asic_type != CHIP_HAWAII) &&
+	    (adev->asic_type != CHIP_BONAIRE) &&
+	    (adev->asic_type != CHIP_FIJI) &&
+	    (adev->asic_type != CHIP_TONGA)) {
+		smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_GetCurrPkgPwr, 0);
+		tmp = cgs_read_register(hwmgr->device, mmSMC_MSG_ARG_0);
+		*query = tmp;
 
-	if (tmp != 0)
-		return 0;
+		if (tmp != 0)
+			return 0;
+	}
 
 	smum_send_msg_to_smc(hwmgr, PPSMC_MSG_PmStatusLogStart);
 	cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC,
@@ -3970,6 +3982,13 @@ static int smu7_set_power_state_tasks(struct pp_hwmgr *hwmgr, const void *input)
 			"Failed to populate and upload SCLK MCLK DPM levels!",
 			result = tmp_result);
 
+	/*
+	 * If a custom pp table is loaded, set DPMTABLE_OD_UPDATE_VDDC flag.
+	 * That effectively disables AVFS feature.
+	 */
+	if (hwmgr->hardcode_pp_table != NULL)
+		data->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_VDDC;
+
 	tmp_result = smu7_update_avfs(hwmgr);
 	PP_ASSERT_WITH_CODE((0 == tmp_result),
 			"Failed to update avfs voltages!",
@@ -5145,6 +5164,9 @@ static const struct pp_hwmgr_func smu7_hwmgr_funcs = {
 	.get_power_profile_mode = smu7_get_power_profile_mode,
 	.set_power_profile_mode = smu7_set_power_profile_mode,
 	.get_performance_level = smu7_get_performance_level,
+	.get_asic_baco_capability = smu7_baco_get_capability,
+	.get_asic_baco_state = smu7_baco_get_state,
+	.set_asic_baco_state = smu7_baco_set_state,
 	.power_off_asic = smu7_power_off_asic,
 };
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_baco.c b/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_baco.c
new file mode 100644
index 000000000000..ea743bea8e29
--- /dev/null
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_baco.c
@@ -0,0 +1,231 @@
+/*
+ * Copyright 2019 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#include "amdgpu.h"
+#include "tonga_baco.h"
+
+#include "gmc/gmc_8_1_d.h"
+#include "gmc/gmc_8_1_sh_mask.h"
+
+#include "bif/bif_5_0_d.h"
+#include "bif/bif_5_0_sh_mask.h"
+
+#include "dce/dce_10_0_d.h"
+#include "dce/dce_10_0_sh_mask.h"
+
+#include "smu/smu_7_1_2_d.h"
+#include "smu/smu_7_1_2_sh_mask.h"
+
+
+static const struct baco_cmd_entry gpio_tbl[] =
+{
+	{ CMD_WRITE, mmGPIOPAD_EN, 0, 0, 0, 0x0 },
+	{ CMD_WRITE, mmGPIOPAD_PD_EN, 0, 0, 0, 0x0 },
+	{ CMD_WRITE, mmGPIOPAD_PU_EN, 0, 0, 0, 0x0 },
+	{ CMD_WRITE, mmGPIOPAD_MASK, 0, 0, 0, 0xff77ffff },
+	{ CMD_WRITE, mmDC_GPIO_DVODATA_EN, 0, 0, 0, 0x0 },
+	{ CMD_WRITE, mmDC_GPIO_DVODATA_MASK, 0, 0, 0, 0xffffffff },
+	{ CMD_WRITE, mmDC_GPIO_GENERIC_EN, 0, 0, 0, 0x0 },
+	{ CMD_READMODIFYWRITE, mmDC_GPIO_GENERIC_MASK, 0, 0, 0, 0x03333333 },
+	{ CMD_WRITE, mmDC_GPIO_SYNCA_EN, 0, 0, 0, 0x0 },
+	{ CMD_READMODIFYWRITE, mmDC_GPIO_SYNCA_MASK, 0, 0, 0, 0x00001111 }
+};
+
+static const struct baco_cmd_entry enable_fb_req_rej_tbl[] =
+{
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, 0xC0300024 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, 0x1, 0x0, 0, 0x1 },
+	{ CMD_WRITE, mmBIF_FB_EN, 0, 0, 0, 0x0 }
+};
+
+static const struct baco_cmd_entry use_bclk_tbl[] =
+{
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixCG_SPLL_FUNC_CNTL },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, CG_SPLL_FUNC_CNTL__SPLL_BYPASS_EN_MASK, CG_SPLL_FUNC_CNTL__SPLL_BYPASS_EN__SHIFT, 0, 0x1 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixCG_SPLL_FUNC_CNTL_2 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, CG_SPLL_FUNC_CNTL_2__SPLL_BYPASS_CHG_MASK, CG_SPLL_FUNC_CNTL_2__SPLL_BYPASS_CHG__SHIFT, 0, 0x1 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixCG_SPLL_STATUS },
+	{ CMD_WAITFOR, mmGCK_SMC_IND_DATA, CG_SPLL_STATUS__SPLL_CHG_STATUS_MASK, 0, 0xffffffff, 0x2 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixCG_SPLL_FUNC_CNTL_2 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, CG_SPLL_FUNC_CNTL_2__SPLL_BYPASS_CHG_MASK, CG_SPLL_FUNC_CNTL_2__SPLL_BYPASS_CHG__SHIFT, 0, 0x0 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, CG_SPLL_FUNC_CNTL_2__SPLL_CTLREQ_CHG_MASK, CG_SPLL_FUNC_CNTL_2__SPLL_CTLREQ_CHG__SHIFT, 0, 0x1 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixCG_SPLL_STATUS },
+	{ CMD_WAITFOR, mmGCK_SMC_IND_DATA, CG_SPLL_STATUS__SPLL_CHG_STATUS_MASK, 0, 0xffffffff, 0x2 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixCG_SPLL_FUNC_CNTL_2 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, CG_SPLL_FUNC_CNTL_2__SPLL_CTLREQ_CHG_MASK, CG_SPLL_FUNC_CNTL_2__SPLL_CTLREQ_CHG__SHIFT, 0, 0x0 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, 0xC0500170 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, 0x4000000, 0x1a, 0, 0x1 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixMPLL_BYPASSCLK_SEL },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, MPLL_BYPASSCLK_SEL__MPLL_CLKOUT_SEL_MASK, MPLL_BYPASSCLK_SEL__MPLL_CLKOUT_SEL__SHIFT, 0, 0x2 },
+	{ CMD_READMODIFYWRITE, mmMPLL_CNTL_MODE, MPLL_CNTL_MODE__MPLL_SW_DIR_CONTROL_MASK, MPLL_CNTL_MODE__MPLL_SW_DIR_CONTROL__SHIFT, 0, 0x1 },
+	{ CMD_READMODIFYWRITE, mmMPLL_CNTL_MODE, MPLL_CNTL_MODE__MPLL_MCLK_SEL_MASK, MPLL_CNTL_MODE__MPLL_MCLK_SEL__SHIFT, 0, 0x0 }
+};
+
+static const struct baco_cmd_entry turn_off_plls_tbl[] =
+{
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixCG_SPLL_FUNC_CNTL },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, CG_SPLL_FUNC_CNTL__SPLL_RESET_MASK, CG_SPLL_FUNC_CNTL__SPLL_RESET__SHIFT, 0, 0x1 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, CG_SPLL_FUNC_CNTL__SPLL_PWRON_MASK, CG_SPLL_FUNC_CNTL__SPLL_PWRON__SHIFT, 0, 0x0 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, 0xC0500170 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, 0x2000000, 0x19, 0, 0x1 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, 0x8000000, 0x1b, 0, 0x0 },
+	{ CMD_READMODIFYWRITE, mmMPLL_CNTL_MODE, MPLL_CNTL_MODE__GLOBAL_MPLL_RESET_MASK, MPLL_CNTL_MODE__GLOBAL_MPLL_RESET__SHIFT, 0, 0x1 },
+	{ CMD_WRITE, mmMPLL_CONTROL, 0, 0, 0, 0x00000006 },
+	{ CMD_WRITE, mmMC_IO_RXCNTL_DPHY0_D0, 0, 0, 0, 0x00007740 },
+	{ CMD_WRITE, mmMC_IO_RXCNTL_DPHY0_D1, 0, 0, 0, 0x00007740 },
+	{ CMD_WRITE, mmMC_IO_RXCNTL_DPHY1_D0, 0, 0, 0, 0x00007740 },
+	{ CMD_WRITE, mmMC_IO_RXCNTL_DPHY1_D1, 0, 0, 0, 0x00007740 },
+	{ CMD_READMODIFYWRITE, mmMCLK_PWRMGT_CNTL, MCLK_PWRMGT_CNTL__MRDCK0_PDNB_MASK, MCLK_PWRMGT_CNTL__MRDCK0_PDNB__SHIFT, 0, 0x0 },
+	{ CMD_READMODIFYWRITE, mmMCLK_PWRMGT_CNTL, MCLK_PWRMGT_CNTL__MRDCK1_PDNB_MASK, MCLK_PWRMGT_CNTL__MRDCK1_PDNB__SHIFT, 0, 0x0 },
+	{ CMD_READMODIFYWRITE, mmMC_SEQ_CNTL_2, MC_SEQ_CNTL_2__DRST_PU_MASK, MC_SEQ_CNTL_2__DRST_PU__SHIFT, 0, 0x0 },
+	{ CMD_READMODIFYWRITE, mmMC_SEQ_CNTL_2, MC_SEQ_CNTL_2__DRST_PD_MASK, MC_SEQ_CNTL_2__DRST_PD__SHIFT, 0, 0x0 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixCG_CLKPIN_CNTL_2 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, CG_CLKPIN_CNTL_2__FORCE_BIF_REFCLK_EN_MASK, CG_CLKPIN_CNTL_2__FORCE_BIF_REFCLK_EN__SHIFT, 0, 0x0 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixMPLL_BYPASSCLK_SEL },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, MPLL_BYPASSCLK_SEL__MPLL_CLKOUT_SEL_MASK, MPLL_BYPASSCLK_SEL__MPLL_CLKOUT_SEL__SHIFT, 0, 0x4 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixMISC_CLK_CTRL },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, MISC_CLK_CTRL__DEEP_SLEEP_CLK_SEL_MASK, MISC_CLK_CTRL__DEEP_SLEEP_CLK_SEL__SHIFT, 0, 0x1 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, MISC_CLK_CTRL__ZCLK_SEL_MASK, MISC_CLK_CTRL__ZCLK_SEL__SHIFT, 0, 0x1 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixCG_CLKPIN_CNTL },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, CG_CLKPIN_CNTL__BCLK_AS_XCLK_MASK, CG_CLKPIN_CNTL__BCLK_AS_XCLK__SHIFT, 0, 0x0 },
+	{ CMD_WRITE, mmGCK_SMC_IND_INDEX, 0, 0, 0, ixTHM_CLK_CNTL },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, THM_CLK_CNTL__CMON_CLK_SEL_MASK,  THM_CLK_CNTL__CMON_CLK_SEL__SHIFT, 0, 0x1 },
+	{ CMD_READMODIFYWRITE, mmGCK_SMC_IND_DATA, THM_CLK_CNTL__TMON_CLK_SEL_MASK,  THM_CLK_CNTL__TMON_CLK_SEL__SHIFT, 0, 0x1 }
+};
+
+static const struct baco_cmd_entry enter_baco_tbl[] =
+{
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_EN_MASK, BACO_CNTL__BACO_EN__SHIFT, 0, 0x01 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_BIF_SCLK_SWITCH_MASK, BACO_CNTL__BACO_BIF_SCLK_SWITCH__SHIFT, 0, 0x01 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__BACO_BIF_SCLK_SWITCH_MASK, 0, 5, 0x40000 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_BCLK_OFF_MASK, BACO_CNTL__BACO_BCLK_OFF__SHIFT, 0, 0x01 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__BACO_BCLK_OFF_MASK, 0, 5, 0x02 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_ISO_DIS_MASK, BACO_CNTL__BACO_ISO_DIS__SHIFT, 0, 0x00 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__BACO_ISO_DIS_MASK, 0, 5, 0x00 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_ANA_ISO_DIS_MASK, BACO_CNTL__BACO_ANA_ISO_DIS__SHIFT, 0, 0x00 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__BACO_ANA_ISO_DIS_MASK, 0, 5, 0x00 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_POWER_OFF_MASK, BACO_CNTL__BACO_POWER_OFF__SHIFT, 0, 0x01 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__BACO_POWER_OFF_MASK, 0, 5, 0x08 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__BACO_MODE_MASK, 0, 0xffffffff, 0x40 }
+};
+
+#define BACO_CNTL__PWRGOOD_MASK  BACO_CNTL__PWRGOOD_GPIO_MASK+BACO_CNTL__PWRGOOD_MEM_MASK+BACO_CNTL__PWRGOOD_DVO_MASK
+
+static const struct baco_cmd_entry exit_baco_tbl[] =
+{
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_RESET_EN_MASK, BACO_CNTL__BACO_RESET_EN__SHIFT, 0, 0x01 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_BCLK_OFF_MASK, BACO_CNTL__BACO_BCLK_OFF__SHIFT, 0, 0x00 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_POWER_OFF_MASK, BACO_CNTL__BACO_POWER_OFF__SHIFT, 0, 0x00 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__PWRGOOD_BF_MASK, 0, 0xffffffff, 0x200 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_ISO_DIS_MASK, BACO_CNTL__BACO_ISO_DIS__SHIFT, 0, 0x01 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__PWRGOOD_MASK, 0, 5, 0x1c00 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_ANA_ISO_DIS_MASK, BACO_CNTL__BACO_ANA_ISO_DIS__SHIFT, 0, 0x01 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_BIF_SCLK_SWITCH_MASK, BACO_CNTL__BACO_BIF_SCLK_SWITCH__SHIFT, 0, 0x00 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_RESET_EN_MASK, BACO_CNTL__BACO_RESET_EN__SHIFT, 0, 0x00 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__RCU_BIF_CONFIG_DONE_MASK, 0, 5, 0x100 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_EN_MASK, BACO_CNTL__BACO_EN__SHIFT, 0, 0x00 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__BACO_MODE_MASK, 0, 0xffffffff, 0x00 }
+};
+
+static const struct baco_cmd_entry clean_baco_tbl[] =
+{
+	{ CMD_WRITE, mmBIOS_SCRATCH_6, 0, 0, 0, 0 },
+	{ CMD_WRITE, mmBIOS_SCRATCH_7, 0, 0, 0, 0 }
+};
+
+static const struct baco_cmd_entry gpio_tbl_iceland[] =
+{
+	{ CMD_WRITE, mmGPIOPAD_EN, 0, 0, 0, 0x0 },
+	{ CMD_WRITE, mmGPIOPAD_PD_EN, 0, 0, 0, 0x0 },
+	{ CMD_WRITE, mmGPIOPAD_PU_EN, 0, 0, 0, 0x0 },
+	{ CMD_WRITE, mmGPIOPAD_MASK, 0, 0, 0, 0xff77ffff }
+};
+
+static const struct baco_cmd_entry exit_baco_tbl_iceland[] =
+{
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_RESET_EN_MASK, BACO_CNTL__BACO_RESET_EN__SHIFT, 0, 0x01 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_BCLK_OFF_MASK, BACO_CNTL__BACO_BCLK_OFF__SHIFT, 0, 0x00 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_POWER_OFF_MASK, BACO_CNTL__BACO_POWER_OFF__SHIFT, 0, 0x00 },
+	{ CMD_DELAY_MS, 0, 0, 0, 20, 0 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__PWRGOOD_BF_MASK, 0, 0xffffffff, 0x200 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_ISO_DIS_MASK, BACO_CNTL__BACO_ISO_DIS__SHIFT, 0, 0x01 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__PWRGOOD_MASK, 0, 5, 0x1c00 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_ANA_ISO_DIS_MASK, BACO_CNTL__BACO_ANA_ISO_DIS__SHIFT, 0, 0x01 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_BIF_SCLK_SWITCH_MASK, BACO_CNTL__BACO_BIF_SCLK_SWITCH__SHIFT, 0, 0x00 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_RESET_EN_MASK, BACO_CNTL__BACO_RESET_EN__SHIFT, 0, 0x00 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__RCU_BIF_CONFIG_DONE_MASK, 0, 5, 0x100 },
+	{ CMD_READMODIFYWRITE, mmBACO_CNTL, BACO_CNTL__BACO_EN_MASK, BACO_CNTL__BACO_EN__SHIFT, 0, 0x00 },
+	{ CMD_WAITFOR, mmBACO_CNTL, BACO_CNTL__BACO_MODE_MASK, 0, 0xffffffff, 0x00 }
+};
+
+static const struct baco_cmd_entry clean_baco_tbl_iceland[] =
+{
+	{ CMD_WRITE, mmBIOS_SCRATCH_7, 0, 0, 0, 0 }
+};
+
+int tonga_baco_set_state(struct pp_hwmgr *hwmgr, enum BACO_STATE state)
+{
+	enum BACO_STATE cur_state;
+
+	smu7_baco_get_state(hwmgr, &cur_state);
+
+	if (cur_state == state)
+		/* aisc already in the target state */
+		return 0;
+
+	if (state == BACO_STATE_IN) {
+		if (hwmgr->chip_id == CHIP_TOPAZ)
+			baco_program_registers(hwmgr, gpio_tbl_iceland, ARRAY_SIZE(gpio_tbl_iceland));
+		else
+			baco_program_registers(hwmgr, gpio_tbl, ARRAY_SIZE(gpio_tbl));
+		baco_program_registers(hwmgr, enable_fb_req_rej_tbl,
+				       ARRAY_SIZE(enable_fb_req_rej_tbl));
+		baco_program_registers(hwmgr, use_bclk_tbl, ARRAY_SIZE(use_bclk_tbl));
+		baco_program_registers(hwmgr, turn_off_plls_tbl,
+				       ARRAY_SIZE(turn_off_plls_tbl));
+		if (baco_program_registers(hwmgr, enter_baco_tbl,
+					   ARRAY_SIZE(enter_baco_tbl)))
+			return 0;
+
+	} else if (state == BACO_STATE_OUT) {
+		/* HW requires at least 20ms between regulator off and on */
+		msleep(20);
+		/* Execute Hardware BACO exit sequence */
+		if (hwmgr->chip_id == CHIP_TOPAZ) {
+			if (baco_program_registers(hwmgr, exit_baco_tbl_iceland,
+						   ARRAY_SIZE(exit_baco_tbl_iceland))) {
+				if (baco_program_registers(hwmgr, clean_baco_tbl_iceland,
+							   ARRAY_SIZE(clean_baco_tbl_iceland)))
+					return 0;
+			}
+		} else {
+			if (baco_program_registers(hwmgr, exit_baco_tbl,
+						   ARRAY_SIZE(exit_baco_tbl))) {
+				if (baco_program_registers(hwmgr, clean_baco_tbl,
+							   ARRAY_SIZE(clean_baco_tbl)))
+					return 0;
+			}
+		}
+	}
+
+	return -EINVAL;
+}
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_baco.h b/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_baco.h
new file mode 100644
index 000000000000..5dc16cc8a295
--- /dev/null
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_baco.h
@@ -0,0 +1,29 @@
+/*
+ * Copyright 2019 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#ifndef __TONGA_BACO_H__
+#define __TONGA_BACO_H__
+#include "smu7_baco.h"
+
+extern int tonga_baco_set_state(struct pp_hwmgr *hwmgr, enum BACO_STATE state);
+
+#endif
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
index d08493b67b67..d71a492c87a3 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
@@ -712,7 +712,6 @@ static int vega10_sort_lookup_table(struct pp_hwmgr *hwmgr,
 		struct phm_ppt_v1_voltage_lookup_table *lookup_table)
 {
 	uint32_t table_size, i, j;
-	struct phm_ppt_v1_voltage_lookup_record tmp_voltage_lookup_record;
 
 	PP_ASSERT_WITH_CODE(lookup_table && lookup_table->count,
 		"Lookup table is empty", return -EINVAL);
@@ -724,9 +723,8 @@ static int vega10_sort_lookup_table(struct pp_hwmgr *hwmgr,
 		for (j = i + 1; j > 0; j--) {
 			if (lookup_table->entries[j].us_vdd <
 					lookup_table->entries[j - 1].us_vdd) {
-				tmp_voltage_lookup_record = lookup_table->entries[j - 1];
-				lookup_table->entries[j - 1] = lookup_table->entries[j];
-				lookup_table->entries[j] = tmp_voltage_lookup_record;
+				swap(lookup_table->entries[j - 1],
+				     lookup_table->entries[j]);
 			}
 		}
 	}
@@ -3691,6 +3689,13 @@ static int vega10_set_power_state_tasks(struct pp_hwmgr *hwmgr,
 	PP_ASSERT_WITH_CODE(!result,
 			"Failed to upload PPtable!", return result);
 
+	/*
+	 * If a custom pp table is loaded, set DPMTABLE_OD_UPDATE_VDDC flag.
+	 * That effectively disables AVFS feature.
+	 */
+	if(hwmgr->hardcode_pp_table != NULL)
+		data->need_update_dpm_table |= DPMTABLE_OD_UPDATE_VDDC;
+
 	vega10_update_avfs(hwmgr);
 
 	/*
@@ -5098,9 +5103,7 @@ static void vega10_odn_update_soc_table(struct pp_hwmgr *hwmgr,
 
 	if (type == PP_OD_EDIT_SCLK_VDDC_TABLE) {
 		podn_vdd_dep = &data->odn_dpm_table.vdd_dep_on_sclk;
-		for (i = 0; i < podn_vdd_dep->count - 1; i++)
-			od_vddc_lookup_table->entries[i].us_vdd = podn_vdd_dep->entries[i].vddc;
-		if (od_vddc_lookup_table->entries[i].us_vdd < podn_vdd_dep->entries[i].vddc)
+		for (i = 0; i < podn_vdd_dep->count; i++)
 			od_vddc_lookup_table->entries[i].us_vdd = podn_vdd_dep->entries[i].vddc;
 	} else if (type == PP_OD_EDIT_MCLK_VDDC_TABLE) {
 		podn_vdd_dep = &data->odn_dpm_table.vdd_dep_on_mclk;
@@ -5267,6 +5270,59 @@ static int vega10_get_performance_level(struct pp_hwmgr *hwmgr, const struct pp_
 	return 0;
 }
 
+static int vega10_disable_power_features_for_compute_performance(struct pp_hwmgr *hwmgr, bool disable)
+{
+	struct vega10_hwmgr *data = hwmgr->backend;
+	uint32_t feature_mask = 0;
+
+	if (disable) {
+		feature_mask |= data->smu_features[GNLD_ULV].enabled ?
+			data->smu_features[GNLD_ULV].smu_feature_bitmap : 0;
+		feature_mask |= data->smu_features[GNLD_DS_GFXCLK].enabled ?
+			data->smu_features[GNLD_DS_GFXCLK].smu_feature_bitmap : 0;
+		feature_mask |= data->smu_features[GNLD_DS_SOCCLK].enabled ?
+			data->smu_features[GNLD_DS_SOCCLK].smu_feature_bitmap : 0;
+		feature_mask |= data->smu_features[GNLD_DS_LCLK].enabled ?
+			data->smu_features[GNLD_DS_LCLK].smu_feature_bitmap : 0;
+		feature_mask |= data->smu_features[GNLD_DS_DCEFCLK].enabled ?
+			data->smu_features[GNLD_DS_DCEFCLK].smu_feature_bitmap : 0;
+	} else {
+		feature_mask |= (!data->smu_features[GNLD_ULV].enabled) ?
+			data->smu_features[GNLD_ULV].smu_feature_bitmap : 0;
+		feature_mask |= (!data->smu_features[GNLD_DS_GFXCLK].enabled) ?
+			data->smu_features[GNLD_DS_GFXCLK].smu_feature_bitmap : 0;
+		feature_mask |= (!data->smu_features[GNLD_DS_SOCCLK].enabled) ?
+			data->smu_features[GNLD_DS_SOCCLK].smu_feature_bitmap : 0;
+		feature_mask |= (!data->smu_features[GNLD_DS_LCLK].enabled) ?
+			data->smu_features[GNLD_DS_LCLK].smu_feature_bitmap : 0;
+		feature_mask |= (!data->smu_features[GNLD_DS_DCEFCLK].enabled) ?
+			data->smu_features[GNLD_DS_DCEFCLK].smu_feature_bitmap : 0;
+	}
+
+	if (feature_mask)
+		PP_ASSERT_WITH_CODE(!vega10_enable_smc_features(hwmgr,
+				!disable, feature_mask),
+				"enable/disable power features for compute performance Failed!",
+				return -EINVAL);
+
+	if (disable) {
+		data->smu_features[GNLD_ULV].enabled = false;
+		data->smu_features[GNLD_DS_GFXCLK].enabled = false;
+		data->smu_features[GNLD_DS_SOCCLK].enabled = false;
+		data->smu_features[GNLD_DS_LCLK].enabled = false;
+		data->smu_features[GNLD_DS_DCEFCLK].enabled = false;
+	} else {
+		data->smu_features[GNLD_ULV].enabled = true;
+		data->smu_features[GNLD_DS_GFXCLK].enabled = true;
+		data->smu_features[GNLD_DS_SOCCLK].enabled = true;
+		data->smu_features[GNLD_DS_LCLK].enabled = true;
+		data->smu_features[GNLD_DS_DCEFCLK].enabled = true;
+	}
+
+	return 0;
+
+}
+
 static const struct pp_hwmgr_func vega10_hwmgr_funcs = {
 	.backend_init = vega10_hwmgr_backend_init,
 	.backend_fini = vega10_hwmgr_backend_fini,
@@ -5334,6 +5390,8 @@ static const struct pp_hwmgr_func vega10_hwmgr_funcs = {
 	.get_ppfeature_status = vega10_get_ppfeature_status,
 	.set_ppfeature_status = vega10_set_ppfeature_status,
 	.set_mp1_state = vega10_set_mp1_state,
+	.disable_power_features_for_compute_performance =
+			vega10_disable_power_features_for_compute_performance,
 };
 
 int vega10_hwmgr_init(struct pp_hwmgr *hwmgr)
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_baco.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_baco.c
index df6ff9252401..9b5e72bdceca 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_baco.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_baco.c
@@ -29,7 +29,7 @@
 #include "vega20_baco.h"
 #include "vega20_smumgr.h"
 
-
+#include "amdgpu_ras.h"
 
 static const struct soc15_baco_cmd_entry clean_baco_tbl[] =
 {
@@ -74,6 +74,7 @@ int vega20_baco_get_state(struct pp_hwmgr *hwmgr, enum BACO_STATE *state)
 int vega20_baco_set_state(struct pp_hwmgr *hwmgr, enum BACO_STATE state)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)(hwmgr->adev);
+	struct amdgpu_ras *ras = amdgpu_ras_get_context(adev);
 	enum BACO_STATE cur_state;
 	uint32_t data;
 
@@ -84,13 +85,19 @@ int vega20_baco_set_state(struct pp_hwmgr *hwmgr, enum BACO_STATE state)
 		return 0;
 
 	if (state == BACO_STATE_IN) {
-		data = RREG32_SOC15(THM, 0, mmTHM_BACO_CNTL);
-		data |= 0x80000000;
-		WREG32_SOC15(THM, 0, mmTHM_BACO_CNTL, data);
-
-
-		if(smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_EnterBaco, 0))
-			return -EINVAL;
+		if (!ras || !ras->supported) {
+			data = RREG32_SOC15(THM, 0, mmTHM_BACO_CNTL);
+			data |= 0x80000000;
+			WREG32_SOC15(THM, 0, mmTHM_BACO_CNTL, data);
+
+			if(smum_send_msg_to_smc_with_parameter(hwmgr,
+					PPSMC_MSG_EnterBaco, 0))
+				return -EINVAL;
+		} else {
+			if(smum_send_msg_to_smc_with_parameter(hwmgr,
+					PPSMC_MSG_EnterBaco, 1))
+				return -EINVAL;
+		}
 
 	} else if (state == BACO_STATE_OUT) {
 		if (smum_send_msg_to_smc(hwmgr, PPSMC_MSG_ExitBaco))
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c
index f5915308e643..5bcf0d684151 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c
@@ -183,6 +183,9 @@ static int vega20_set_features_platform_caps(struct pp_hwmgr *hwmgr)
 			PHM_PlatformCaps_TablelessHardwareInterface);
 
 	phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+			PHM_PlatformCaps_BACO);
+
+	phm_cap_set(hwmgr->platform_descriptor.platformCaps,
 			PHM_PlatformCaps_EnableSMU7ThermalManagement);
 
 	if (adev->pg_flags & AMD_PG_SUPPORT_UVD)
@@ -490,8 +493,8 @@ static int vega20_setup_asic_task(struct pp_hwmgr *hwmgr)
 			"Failed to init sclk threshold!",
 			return ret);
 
-	if (adev->in_baco_reset) {
-		adev->in_baco_reset = 0;
+	if (adev->in_gpu_reset &&
+	    (amdgpu_asic_reset_method(adev) == AMD_RESET_METHOD_BACO)) {
 
 		ret = vega20_baco_apply_vdci_flush_workaround(hwmgr);
 		if (ret)
@@ -4155,6 +4158,38 @@ static int vega20_smu_i2c_bus_access(struct pp_hwmgr *hwmgr, bool acquire)
 	return res;
 }
 
+static int vega20_set_df_cstate(struct pp_hwmgr *hwmgr,
+				enum pp_df_cstate state)
+{
+	int ret;
+
+	/* PPSMC_MSG_DFCstateControl is supported with 40.50 and later fws */
+	if (hwmgr->smu_version < 0x283200) {
+		pr_err("Df cstate control is supported with 40.50 and later SMC fw!\n");
+		return -EINVAL;
+	}
+
+	ret = smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_DFCstateControl, state);
+	if (ret)
+		pr_err("SetDfCstate failed!\n");
+
+	return ret;
+}
+
+static int vega20_set_xgmi_pstate(struct pp_hwmgr *hwmgr,
+				  uint32_t pstate)
+{
+	int ret;
+
+	ret = smum_send_msg_to_smc_with_parameter(hwmgr,
+						  PPSMC_MSG_SetXgmiMode,
+						  pstate ? XGMI_MODE_PSTATE_D0 : XGMI_MODE_PSTATE_D3);
+	if (ret)
+		pr_err("SetXgmiPstate failed!\n");
+
+	return ret;
+}
+
 static const struct pp_hwmgr_func vega20_hwmgr_funcs = {
 	/* init/fini related */
 	.backend_init = vega20_hwmgr_backend_init,
@@ -4223,6 +4258,8 @@ static const struct pp_hwmgr_func vega20_hwmgr_funcs = {
 	.set_asic_baco_state = vega20_baco_set_state,
 	.set_mp1_state = vega20_set_mp1_state,
 	.smu_i2c_bus_access = vega20_smu_i2c_bus_access,
+	.set_df_cstate = vega20_set_df_cstate,
+	.set_xgmi_pstate = vega20_set_xgmi_pstate,
 };
 
 int vega20_hwmgr_init(struct pp_hwmgr *hwmgr)
diff --git a/drivers/gpu/drm/amd/powerplay/inc/amdgpu_smu.h b/drivers/gpu/drm/amd/powerplay/inc/amdgpu_smu.h
index 6109815a0401..031e0c22fcc7 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/amdgpu_smu.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/amdgpu_smu.h
@@ -259,10 +259,8 @@ struct smu_table_context
 	struct smu_bios_boot_up_values	boot_values;
 	void                            *driver_pptable;
 	struct smu_table		*tables;
-	uint32_t			table_count;
 	struct smu_table		memory_pool;
 	uint8_t                         thermal_controller_type;
-	uint16_t			TDPODLimit;
 
 	void				*overdrive_table;
 };
@@ -322,6 +320,13 @@ struct mclock_latency_table {
 	struct mclk_latency_entries  entries[MAX_REGULAR_DPM_NUM];
 };
 
+enum smu_reset_mode
+{
+    SMU_RESET_MODE_0,
+    SMU_RESET_MODE_1,
+    SMU_RESET_MODE_2,
+};
+
 enum smu_baco_state
 {
 	SMU_BACO_STATE_ENTER = 0,
@@ -341,9 +346,9 @@ struct smu_context
 	struct amdgpu_device            *adev;
 	struct amdgpu_irq_src		*irq_source;
 
-	const struct smu_funcs		*funcs;
 	const struct pptable_funcs	*ppt_funcs;
 	struct mutex			mutex;
+	struct mutex			sensor_lock;
 	uint64_t pool_size;
 
 	struct smu_table_context	smu_table;
@@ -381,11 +386,15 @@ struct smu_context
 	uint32_t power_profile_mode;
 	uint32_t default_power_profile_mode;
 	bool pm_enabled;
+	bool is_apu;
 
 	uint32_t smc_if_version;
 
+	bool uploading_custom_pp_table;
 };
 
+struct i2c_adapter;
+
 struct pptable_funcs {
 	int (*alloc_dpm_context)(struct smu_context *smu);
 	int (*store_powerplay_table)(struct smu_context *smu);
@@ -397,7 +406,7 @@ struct pptable_funcs {
 	int (*get_smu_table_index)(struct smu_context *smu, uint32_t index);
 	int (*get_smu_power_index)(struct smu_context *smu, uint32_t index);
 	int (*get_workload_type)(struct smu_context *smu, enum PP_SMC_POWER_PROFILE profile);
-	int (*run_afll_btc)(struct smu_context *smu);
+	int (*run_btc)(struct smu_context *smu);
 	int (*get_allowed_feature_mask)(struct smu_context *smu, uint32_t *feature_mask, uint32_t num);
 	enum amd_pm_state_type (*get_current_power_state)(struct smu_context *smu);
 	int (*set_default_dpm_table)(struct smu_context *smu);
@@ -458,17 +467,19 @@ struct pptable_funcs {
 	int (*display_disable_memory_clock_switch)(struct smu_context *smu, bool disable_memory_clock_switch);
 	void (*dump_pptable)(struct smu_context *smu);
 	int (*get_power_limit)(struct smu_context *smu, uint32_t *limit, bool asic_default);
-	int (*get_dpm_uclk_limited)(struct smu_context *smu, uint32_t *clock, bool max);
-};
-
-struct smu_funcs
-{
+	int (*get_dpm_clk_limited)(struct smu_context *smu, enum smu_clk_type clk_type,
+				   uint32_t dpm_level, uint32_t *freq);
+	int (*set_df_cstate)(struct smu_context *smu, enum pp_df_cstate state);
+	int (*update_pcie_parameters)(struct smu_context *smu, uint32_t pcie_gen_cap, uint32_t pcie_width_cap);
+	int (*i2c_eeprom_init)(struct i2c_adapter *control);
+	void (*i2c_eeprom_fini)(struct i2c_adapter *control);
+	int (*get_dpm_clock_table)(struct smu_context *smu, struct dpm_clocks *clock_table);
 	int (*init_microcode)(struct smu_context *smu);
+	int (*load_microcode)(struct smu_context *smu);
 	int (*init_smc_tables)(struct smu_context *smu);
 	int (*fini_smc_tables)(struct smu_context *smu);
 	int (*init_power)(struct smu_context *smu);
 	int (*fini_power)(struct smu_context *smu);
-	int (*load_microcode)(struct smu_context *smu);
 	int (*check_fw_status)(struct smu_context *smu);
 	int (*setup_pptable)(struct smu_context *smu);
 	int (*get_vbios_bootup_values)(struct smu_context *smu);
@@ -484,7 +495,6 @@ struct smu_funcs
 	int (*set_min_dcef_deep_sleep)(struct smu_context *smu);
 	int (*set_tool_table_location)(struct smu_context *smu);
 	int (*notify_memory_pool_location)(struct smu_context *smu);
-	int (*write_watermarks_table)(struct smu_context *smu);
 	int (*set_last_dcef_min_deep_sleep_clk)(struct smu_context *smu);
 	int (*system_features_control)(struct smu_context *smu, bool en);
 	int (*send_smc_msg)(struct smu_context *smu, uint16_t msg);
@@ -498,8 +508,7 @@ struct smu_funcs
 	int (*get_current_clk_freq)(struct smu_context *smu, enum smu_clk_type clk_id, uint32_t *value);
 	int (*init_max_sustainable_clocks)(struct smu_context *smu);
 	int (*start_thermal_control)(struct smu_context *smu);
-	int (*read_sensor)(struct smu_context *smu, enum amd_pp_sensors sensor,
-			   void *data, uint32_t *size);
+	int (*stop_thermal_control)(struct smu_context *smu);
 	int (*set_deep_sleep_dcefclk)(struct smu_context *smu, uint32_t clk);
 	int (*set_active_display_count)(struct smu_context *smu, uint32_t count);
 	int (*store_cc6_data)(struct smu_context *smu, uint32_t separation_time,
@@ -521,8 +530,6 @@ struct smu_funcs
 	int (*get_current_shallow_sleep_clocks)(struct smu_context *smu,
 						struct smu_clock_info *clocks);
 	int (*notify_smu_enable_pwe)(struct smu_context *smu);
-	int (*set_watermarks_for_clock_ranges)(struct smu_context *smu,
-					       struct dm_pp_wm_sets_with_clock_ranges_soc15 *clock_ranges);
 	int (*conv_power_profile_to_pplib_workload)(int power_profile);
 	uint32_t (*get_fan_control_mode)(struct smu_context *smu);
 	int (*set_fan_control_mode)(struct smu_context *smu, uint32_t mode);
@@ -537,234 +544,90 @@ struct smu_funcs
 	enum smu_baco_state (*baco_get_state)(struct smu_context *smu);
 	int (*baco_set_state)(struct smu_context *smu, enum smu_baco_state state);
 	int (*baco_reset)(struct smu_context *smu);
+	int (*mode2_reset)(struct smu_context *smu);
 	int (*get_dpm_ultimate_freq)(struct smu_context *smu, enum smu_clk_type clk_type, uint32_t *min, uint32_t *max);
+	int (*set_soft_freq_limited_range)(struct smu_context *smu, enum smu_clk_type clk_type, uint32_t min, uint32_t max);
+	int (*override_pcie_parameters)(struct smu_context *smu);
+	uint32_t (*get_pptable_power_limit)(struct smu_context *smu);
 };
 
-#define smu_init_microcode(smu) \
-	((smu)->funcs->init_microcode ? (smu)->funcs->init_microcode((smu)) : 0)
-#define smu_init_smc_tables(smu) \
-	((smu)->funcs->init_smc_tables ? (smu)->funcs->init_smc_tables((smu)) : 0)
-#define smu_fini_smc_tables(smu) \
-	((smu)->funcs->fini_smc_tables ? (smu)->funcs->fini_smc_tables((smu)) : 0)
-#define smu_init_power(smu) \
-	((smu)->funcs->init_power ? (smu)->funcs->init_power((smu)) : 0)
-#define smu_fini_power(smu) \
-	((smu)->funcs->fini_power ? (smu)->funcs->fini_power((smu)) : 0)
-#define smu_load_microcode(smu) \
-	((smu)->funcs->load_microcode ? (smu)->funcs->load_microcode((smu)) : 0)
-#define smu_check_fw_status(smu) \
-	((smu)->funcs->check_fw_status ? (smu)->funcs->check_fw_status((smu)) : 0)
-#define smu_setup_pptable(smu) \
-	((smu)->funcs->setup_pptable ? (smu)->funcs->setup_pptable((smu)) : 0)
-#define smu_powergate_sdma(smu, gate) \
-	((smu)->funcs->powergate_sdma ? (smu)->funcs->powergate_sdma((smu), (gate)) : 0)
-#define smu_powergate_vcn(smu, gate) \
-	((smu)->funcs->powergate_vcn ? (smu)->funcs->powergate_vcn((smu), (gate)) : 0)
-#define smu_set_gfx_cgpg(smu, enabled) \
-	((smu)->funcs->set_gfx_cgpg ? (smu)->funcs->set_gfx_cgpg((smu), (enabled)) : 0)
-#define smu_get_vbios_bootup_values(smu) \
-	((smu)->funcs->get_vbios_bootup_values ? (smu)->funcs->get_vbios_bootup_values((smu)) : 0)
-#define smu_get_clk_info_from_vbios(smu) \
-	((smu)->funcs->get_clk_info_from_vbios ? (smu)->funcs->get_clk_info_from_vbios((smu)) : 0)
-#define smu_check_pptable(smu) \
-	((smu)->funcs->check_pptable ? (smu)->funcs->check_pptable((smu)) : 0)
-#define smu_parse_pptable(smu) \
-	((smu)->funcs->parse_pptable ? (smu)->funcs->parse_pptable((smu)) : 0)
-#define smu_populate_smc_tables(smu) \
-	((smu)->funcs->populate_smc_tables ? (smu)->funcs->populate_smc_tables((smu)) : 0)
-#define smu_check_fw_version(smu) \
-	((smu)->funcs->check_fw_version ? (smu)->funcs->check_fw_version((smu)) : 0)
-#define smu_write_pptable(smu) \
-	((smu)->funcs->write_pptable ? (smu)->funcs->write_pptable((smu)) : 0)
-#define smu_set_min_dcef_deep_sleep(smu) \
-	((smu)->funcs->set_min_dcef_deep_sleep ? (smu)->funcs->set_min_dcef_deep_sleep((smu)) : 0)
-#define smu_set_tool_table_location(smu) \
-	((smu)->funcs->set_tool_table_location ? (smu)->funcs->set_tool_table_location((smu)) : 0)
-#define smu_notify_memory_pool_location(smu) \
-	((smu)->funcs->notify_memory_pool_location ? (smu)->funcs->notify_memory_pool_location((smu)) : 0)
-#define smu_gfx_off_control(smu, enable) \
-	((smu)->funcs->gfx_off_control ? (smu)->funcs->gfx_off_control((smu), (enable)) : 0)
-
-#define smu_write_watermarks_table(smu) \
-	((smu)->funcs->write_watermarks_table ? (smu)->funcs->write_watermarks_table((smu)) : 0)
-#define smu_set_last_dcef_min_deep_sleep_clk(smu) \
-	((smu)->funcs->set_last_dcef_min_deep_sleep_clk ? (smu)->funcs->set_last_dcef_min_deep_sleep_clk((smu)) : 0)
-#define smu_system_features_control(smu, en) \
-	((smu)->funcs->system_features_control ? (smu)->funcs->system_features_control((smu), (en)) : 0)
-#define smu_init_max_sustainable_clocks(smu) \
-	((smu)->funcs->init_max_sustainable_clocks ? (smu)->funcs->init_max_sustainable_clocks((smu)) : 0)
-#define smu_set_default_od_settings(smu, initialize) \
-	((smu)->ppt_funcs->set_default_od_settings ? (smu)->ppt_funcs->set_default_od_settings((smu), (initialize)) : 0)
-#define smu_set_fan_speed_rpm(smu, speed) \
-	((smu)->funcs->set_fan_speed_rpm ? (smu)->funcs->set_fan_speed_rpm((smu), (speed)) : 0)
-#define smu_send_smc_msg(smu, msg) \
-	((smu)->funcs->send_smc_msg? (smu)->funcs->send_smc_msg((smu), (msg)) : 0)
-#define smu_send_smc_msg_with_param(smu, msg, param) \
-	((smu)->funcs->send_smc_msg_with_param? (smu)->funcs->send_smc_msg_with_param((smu), (msg), (param)) : 0)
-#define smu_read_smc_arg(smu, arg) \
-	((smu)->funcs->read_smc_arg? (smu)->funcs->read_smc_arg((smu), (arg)) : 0)
-#define smu_alloc_dpm_context(smu) \
-	((smu)->ppt_funcs->alloc_dpm_context ? (smu)->ppt_funcs->alloc_dpm_context((smu)) : 0)
-#define smu_init_display_count(smu, count) \
-	((smu)->funcs->init_display_count ? (smu)->funcs->init_display_count((smu), (count)) : 0)
-#define smu_feature_set_allowed_mask(smu) \
-	((smu)->funcs->set_allowed_mask? (smu)->funcs->set_allowed_mask((smu)) : 0)
-#define smu_feature_get_enabled_mask(smu, mask, num) \
-	((smu)->funcs->get_enabled_mask? (smu)->funcs->get_enabled_mask((smu), (mask), (num)) : 0)
-#define smu_is_dpm_running(smu) \
-	((smu)->ppt_funcs->is_dpm_running ? (smu)->ppt_funcs->is_dpm_running((smu)) : 0)
-#define smu_notify_display_change(smu) \
-	((smu)->funcs->notify_display_change? (smu)->funcs->notify_display_change((smu)) : 0)
-#define smu_store_powerplay_table(smu) \
-	((smu)->ppt_funcs->store_powerplay_table ? (smu)->ppt_funcs->store_powerplay_table((smu)) : 0)
-#define smu_check_powerplay_table(smu) \
-	((smu)->ppt_funcs->check_powerplay_table ? (smu)->ppt_funcs->check_powerplay_table((smu)) : 0)
-#define smu_append_powerplay_table(smu) \
-	((smu)->ppt_funcs->append_powerplay_table ? (smu)->ppt_funcs->append_powerplay_table((smu)) : 0)
-#define smu_set_default_dpm_table(smu) \
-	((smu)->ppt_funcs->set_default_dpm_table ? (smu)->ppt_funcs->set_default_dpm_table((smu)) : 0)
-#define smu_populate_umd_state_clk(smu) \
-	((smu)->ppt_funcs->populate_umd_state_clk ? (smu)->ppt_funcs->populate_umd_state_clk((smu)) : 0)
-#define smu_set_default_od8_settings(smu) \
-	((smu)->ppt_funcs->set_default_od8_settings ? (smu)->ppt_funcs->set_default_od8_settings((smu)) : 0)
-#define smu_get_power_limit(smu, limit, def) \
-	((smu)->ppt_funcs->get_power_limit ? (smu)->ppt_funcs->get_power_limit((smu), (limit), (def)) : 0)
-#define smu_set_power_limit(smu, limit) \
-	((smu)->funcs->set_power_limit ? (smu)->funcs->set_power_limit((smu), (limit)) : 0)
-#define smu_get_current_clk_freq(smu, clk_id, value) \
-	((smu)->funcs->get_current_clk_freq? (smu)->funcs->get_current_clk_freq((smu), (clk_id), (value)) : 0)
-#define smu_print_clk_levels(smu, clk_type, buf) \
-	((smu)->ppt_funcs->print_clk_levels ? (smu)->ppt_funcs->print_clk_levels((smu), (clk_type), (buf)) : 0)
-#define smu_force_clk_levels(smu, clk_type, level) \
-	((smu)->ppt_funcs->force_clk_levels ? (smu)->ppt_funcs->force_clk_levels((smu), (clk_type), (level)) : 0)
-#define smu_get_od_percentage(smu, type) \
-	((smu)->ppt_funcs->get_od_percentage ? (smu)->ppt_funcs->get_od_percentage((smu), (type)) : 0)
-#define smu_set_od_percentage(smu, type, value) \
-	((smu)->ppt_funcs->set_od_percentage ? (smu)->ppt_funcs->set_od_percentage((smu), (type), (value)) : 0)
-#define smu_od_edit_dpm_table(smu, type, input, size) \
-	((smu)->ppt_funcs->od_edit_dpm_table ? (smu)->ppt_funcs->od_edit_dpm_table((smu), (type), (input), (size)) : 0)
-#define smu_tables_init(smu, tab) \
-	((smu)->ppt_funcs->tables_init ? (smu)->ppt_funcs->tables_init((smu), (tab)) : 0)
-#define smu_set_thermal_fan_table(smu) \
-	((smu)->ppt_funcs->set_thermal_fan_table ? (smu)->ppt_funcs->set_thermal_fan_table((smu)) : 0)
-#define smu_start_thermal_control(smu) \
-	((smu)->funcs->start_thermal_control? (smu)->funcs->start_thermal_control((smu)) : 0)
-#define smu_read_sensor(smu, sensor, data, size) \
-	((smu)->ppt_funcs->read_sensor? (smu)->ppt_funcs->read_sensor((smu), (sensor), (data), (size)) : 0)
-#define smu_smc_read_sensor(smu, sensor, data, size) \
-	((smu)->funcs->read_sensor? (smu)->funcs->read_sensor((smu), (sensor), (data), (size)) : -EINVAL)
-#define smu_get_power_profile_mode(smu, buf) \
-	((smu)->ppt_funcs->get_power_profile_mode ? (smu)->ppt_funcs->get_power_profile_mode((smu), buf) : 0)
-#define smu_set_power_profile_mode(smu, param, param_size) \
-	((smu)->ppt_funcs->set_power_profile_mode ? (smu)->ppt_funcs->set_power_profile_mode((smu), (param), (param_size)) : 0)
-#define smu_pre_display_config_changed(smu) \
-	((smu)->ppt_funcs->pre_display_config_changed ? (smu)->ppt_funcs->pre_display_config_changed((smu)) : 0)
-#define smu_display_config_changed(smu) \
-	((smu)->ppt_funcs->display_config_changed ? (smu)->ppt_funcs->display_config_changed((smu)) : 0)
-#define smu_apply_clocks_adjust_rules(smu) \
-	((smu)->ppt_funcs->apply_clocks_adjust_rules ? (smu)->ppt_funcs->apply_clocks_adjust_rules((smu)) : 0)
-#define smu_notify_smc_dispaly_config(smu) \
-	((smu)->ppt_funcs->notify_smc_dispaly_config ? (smu)->ppt_funcs->notify_smc_dispaly_config((smu)) : 0)
-#define smu_force_dpm_limit_value(smu, highest) \
-	((smu)->ppt_funcs->force_dpm_limit_value ? (smu)->ppt_funcs->force_dpm_limit_value((smu), (highest)) : 0)
-#define smu_unforce_dpm_levels(smu) \
-	((smu)->ppt_funcs->unforce_dpm_levels ? (smu)->ppt_funcs->unforce_dpm_levels((smu)) : 0)
-#define smu_get_profiling_clk_mask(smu, level, sclk_mask, mclk_mask, soc_mask) \
-	((smu)->ppt_funcs->get_profiling_clk_mask ? (smu)->ppt_funcs->get_profiling_clk_mask((smu), (level), (sclk_mask), (mclk_mask), (soc_mask)) : 0)
-#define smu_set_cpu_power_state(smu) \
-	((smu)->ppt_funcs->set_cpu_power_state ? (smu)->ppt_funcs->set_cpu_power_state((smu)) : 0)
-#define smu_get_fan_control_mode(smu) \
-	((smu)->funcs->get_fan_control_mode ? (smu)->funcs->get_fan_control_mode((smu)) : 0)
-#define smu_set_fan_control_mode(smu, value) \
-	((smu)->funcs->set_fan_control_mode ? (smu)->funcs->set_fan_control_mode((smu), (value)) : 0)
-#define smu_get_fan_speed_percent(smu, speed) \
-	((smu)->ppt_funcs->get_fan_speed_percent ? (smu)->ppt_funcs->get_fan_speed_percent((smu), (speed)) : 0)
-#define smu_set_fan_speed_percent(smu, speed) \
-	((smu)->funcs->set_fan_speed_percent ? (smu)->funcs->set_fan_speed_percent((smu), (speed)) : 0)
-#define smu_get_fan_speed_rpm(smu, speed) \
-	((smu)->ppt_funcs->get_fan_speed_rpm ? (smu)->ppt_funcs->get_fan_speed_rpm((smu), (speed)) : 0)
-
-#define smu_msg_get_index(smu, msg) \
-	((smu)->ppt_funcs? ((smu)->ppt_funcs->get_smu_msg_index? (smu)->ppt_funcs->get_smu_msg_index((smu), (msg)) : -EINVAL) : -EINVAL)
-#define smu_clk_get_index(smu, msg) \
-	((smu)->ppt_funcs? ((smu)->ppt_funcs->get_smu_clk_index? (smu)->ppt_funcs->get_smu_clk_index((smu), (msg)) : -EINVAL) : -EINVAL)
-#define smu_feature_get_index(smu, msg) \
-	((smu)->ppt_funcs? ((smu)->ppt_funcs->get_smu_feature_index? (smu)->ppt_funcs->get_smu_feature_index((smu), (msg)) : -EINVAL) : -EINVAL)
-#define smu_table_get_index(smu, tab) \
-	((smu)->ppt_funcs? ((smu)->ppt_funcs->get_smu_table_index? (smu)->ppt_funcs->get_smu_table_index((smu), (tab)) : -EINVAL) : -EINVAL)
-#define smu_power_get_index(smu, src) \
-	((smu)->ppt_funcs? ((smu)->ppt_funcs->get_smu_power_index? (smu)->ppt_funcs->get_smu_power_index((smu), (src)) : -EINVAL) : -EINVAL)
-#define smu_workload_get_type(smu, profile) \
-	((smu)->ppt_funcs? ((smu)->ppt_funcs->get_workload_type? (smu)->ppt_funcs->get_workload_type((smu), (profile)) : -EINVAL) : -EINVAL)
-#define smu_run_afll_btc(smu) \
-	((smu)->ppt_funcs? ((smu)->ppt_funcs->run_afll_btc? (smu)->ppt_funcs->run_afll_btc((smu)) : 0) : 0)
-#define smu_get_allowed_feature_mask(smu, feature_mask, num) \
-	((smu)->ppt_funcs? ((smu)->ppt_funcs->get_allowed_feature_mask? (smu)->ppt_funcs->get_allowed_feature_mask((smu), (feature_mask), (num)) : 0) : 0)
-#define smu_set_deep_sleep_dcefclk(smu, clk) \
-	((smu)->funcs->set_deep_sleep_dcefclk ? (smu)->funcs->set_deep_sleep_dcefclk((smu), (clk)) : 0)
-#define smu_set_active_display_count(smu, count) \
-	((smu)->funcs->set_active_display_count ? (smu)->funcs->set_active_display_count((smu), (count)) : 0)
-#define smu_store_cc6_data(smu, st, cc6_dis, pst_dis, pst_sw_dis) \
-	((smu)->funcs->store_cc6_data ? (smu)->funcs->store_cc6_data((smu), (st), (cc6_dis), (pst_dis), (pst_sw_dis)) : 0)
-#define smu_get_clock_by_type(smu, type, clocks) \
-	((smu)->funcs->get_clock_by_type ? (smu)->funcs->get_clock_by_type((smu), (type), (clocks)) : 0)
-#define smu_get_max_high_clocks(smu, clocks) \
-	((smu)->funcs->get_max_high_clocks ? (smu)->funcs->get_max_high_clocks((smu), (clocks)) : 0)
-#define smu_get_clock_by_type_with_latency(smu, clk_type, clocks) \
-	((smu)->ppt_funcs->get_clock_by_type_with_latency ? (smu)->ppt_funcs->get_clock_by_type_with_latency((smu), (clk_type), (clocks)) : 0)
-#define smu_get_clock_by_type_with_voltage(smu, type, clocks) \
-	((smu)->ppt_funcs->get_clock_by_type_with_voltage ? (smu)->ppt_funcs->get_clock_by_type_with_voltage((smu), (type), (clocks)) : 0)
-#define smu_display_clock_voltage_request(smu, clock_req) \
-	((smu)->funcs->display_clock_voltage_request ? (smu)->funcs->display_clock_voltage_request((smu), (clock_req)) : 0)
-#define smu_display_disable_memory_clock_switch(smu, disable_memory_clock_switch) \
-	((smu)->ppt_funcs->display_disable_memory_clock_switch ? (smu)->ppt_funcs->display_disable_memory_clock_switch((smu), (disable_memory_clock_switch)) : -EINVAL)
-#define smu_get_dal_power_level(smu, clocks) \
-	((smu)->funcs->get_dal_power_level ? (smu)->funcs->get_dal_power_level((smu), (clocks)) : 0)
-#define smu_get_perf_level(smu, designation, level) \
-	((smu)->funcs->get_perf_level ? (smu)->funcs->get_perf_level((smu), (designation), (level)) : 0)
-#define smu_get_current_shallow_sleep_clocks(smu, clocks) \
-	((smu)->funcs->get_current_shallow_sleep_clocks ? (smu)->funcs->get_current_shallow_sleep_clocks((smu), (clocks)) : 0)
-#define smu_notify_smu_enable_pwe(smu) \
-	((smu)->funcs->notify_smu_enable_pwe ? (smu)->funcs->notify_smu_enable_pwe((smu)) : 0)
-#define smu_set_watermarks_for_clock_ranges(smu, clock_ranges) \
-	((smu)->funcs->set_watermarks_for_clock_ranges ? (smu)->funcs->set_watermarks_for_clock_ranges((smu), (clock_ranges)) : 0)
-#define smu_dpm_set_uvd_enable(smu, enable) \
-	((smu)->ppt_funcs->dpm_set_uvd_enable ? (smu)->ppt_funcs->dpm_set_uvd_enable((smu), (enable)) : 0)
-#define smu_dpm_set_vce_enable(smu, enable) \
-	((smu)->ppt_funcs->dpm_set_vce_enable ? (smu)->ppt_funcs->dpm_set_vce_enable((smu), (enable)) : 0)
-#define smu_set_xgmi_pstate(smu, pstate) \
-		((smu)->funcs->set_xgmi_pstate ? (smu)->funcs->set_xgmi_pstate((smu), (pstate)) : 0)
-#define smu_set_watermarks_table(smu, tab, clock_ranges) \
-	((smu)->ppt_funcs->set_watermarks_table ? (smu)->ppt_funcs->set_watermarks_table((smu), (tab), (clock_ranges)) : 0)
-#define smu_get_current_clk_freq_by_table(smu, clk_type, value) \
-	((smu)->ppt_funcs->get_current_clk_freq_by_table ? (smu)->ppt_funcs->get_current_clk_freq_by_table((smu), (clk_type), (value)) : 0)
-#define smu_thermal_temperature_range_update(smu, range, rw) \
-	((smu)->ppt_funcs->thermal_temperature_range_update? (smu)->ppt_funcs->thermal_temperature_range_update((smu), (range), (rw)) : 0)
-#define smu_get_thermal_temperature_range(smu, range) \
-	((smu)->ppt_funcs->get_thermal_temperature_range? (smu)->ppt_funcs->get_thermal_temperature_range((smu), (range)) : 0)
-#define smu_register_irq_handler(smu) \
-	((smu)->funcs->register_irq_handler ? (smu)->funcs->register_irq_handler(smu) : 0)
-#define smu_set_azalia_d3_pme(smu) \
-	((smu)->funcs->set_azalia_d3_pme ? (smu)->funcs->set_azalia_d3_pme((smu)) : 0)
-#define smu_get_dpm_ultimate_freq(smu, param, min, max) \
-		((smu)->funcs->get_dpm_ultimate_freq ? (smu)->funcs->get_dpm_ultimate_freq((smu), (param), (min), (max)) : 0)
-#define smu_get_max_sustainable_clocks_by_dc(smu, max_clocks) \
-	((smu)->funcs->get_max_sustainable_clocks_by_dc ? (smu)->funcs->get_max_sustainable_clocks_by_dc((smu), (max_clocks)) : 0)
-#define smu_get_uclk_dpm_states(smu, clocks_in_khz, num_states) \
-	((smu)->ppt_funcs->get_uclk_dpm_states ? (smu)->ppt_funcs->get_uclk_dpm_states((smu), (clocks_in_khz), (num_states)) : 0)
-#define smu_baco_is_support(smu) \
-	((smu)->funcs->baco_is_support? (smu)->funcs->baco_is_support((smu)) : false)
-#define smu_baco_get_state(smu, state) \
-	((smu)->funcs->baco_get_state? (smu)->funcs->baco_get_state((smu), (state)) : 0)
-#define smu_baco_reset(smu) \
-	((smu)->funcs->baco_reset? (smu)->funcs->baco_reset((smu)) : 0)
-#define smu_asic_set_performance_level(smu, level) \
-	((smu)->ppt_funcs->set_performance_level? (smu)->ppt_funcs->set_performance_level((smu), (level)) : -EINVAL);
-#define smu_dump_pptable(smu) \
-	((smu)->ppt_funcs->dump_pptable ? (smu)->ppt_funcs->dump_pptable((smu)) : 0)
-#define smu_get_dpm_uclk_limited(smu, clock, max) \
-		((smu)->ppt_funcs->get_dpm_uclk_limited ? (smu)->ppt_funcs->get_dpm_uclk_limited((smu), (clock), (max)) : -EINVAL)
+int smu_load_microcode(struct smu_context *smu);
+
+int smu_check_fw_status(struct smu_context *smu);
+
+int smu_set_gfx_cgpg(struct smu_context *smu, bool enabled);
+
+#define smu_i2c_eeprom_init(smu, control) \
+		((smu)->ppt_funcs->i2c_eeprom_init ? (smu)->ppt_funcs->i2c_eeprom_init((control)) : -EINVAL)
+#define smu_i2c_eeprom_fini(smu, control) \
+		((smu)->ppt_funcs->i2c_eeprom_fini ? (smu)->ppt_funcs->i2c_eeprom_fini((control)) : -EINVAL)
+
+int smu_set_fan_speed_rpm(struct smu_context *smu, uint32_t speed);
+
+int smu_get_power_limit(struct smu_context *smu,
+			uint32_t *limit,
+			bool def,
+			bool lock_needed);
+
+int smu_set_power_limit(struct smu_context *smu, uint32_t limit);
+int smu_print_clk_levels(struct smu_context *smu, enum smu_clk_type clk_type, char *buf);
+int smu_get_od_percentage(struct smu_context *smu, enum smu_clk_type type);
+int smu_set_od_percentage(struct smu_context *smu, enum smu_clk_type type, uint32_t value);
+
+int smu_od_edit_dpm_table(struct smu_context *smu,
+			  enum PP_OD_DPM_TABLE_COMMAND type,
+			  long *input, uint32_t size);
+
+int smu_read_sensor(struct smu_context *smu,
+		    enum amd_pp_sensors sensor,
+		    void *data, uint32_t *size);
+int smu_get_power_profile_mode(struct smu_context *smu, char *buf);
+
+int smu_set_power_profile_mode(struct smu_context *smu,
+			       long *param,
+			       uint32_t param_size,
+			       bool lock_needed);
+int smu_get_fan_control_mode(struct smu_context *smu);
+int smu_set_fan_control_mode(struct smu_context *smu, int value);
+int smu_get_fan_speed_percent(struct smu_context *smu, uint32_t *speed);
+int smu_set_fan_speed_percent(struct smu_context *smu, uint32_t speed);
+int smu_get_fan_speed_rpm(struct smu_context *smu, uint32_t *speed);
+
+int smu_set_deep_sleep_dcefclk(struct smu_context *smu, int clk);
+int smu_set_active_display_count(struct smu_context *smu, uint32_t count);
+
+int smu_get_clock_by_type(struct smu_context *smu,
+			  enum amd_pp_clock_type type,
+			  struct amd_pp_clocks *clocks);
+
+int smu_get_max_high_clocks(struct smu_context *smu,
+			    struct amd_pp_simple_clock_info *clocks);
+
+int smu_get_clock_by_type_with_latency(struct smu_context *smu,
+				       enum smu_clk_type clk_type,
+				       struct pp_clock_levels_with_latency *clocks);
+
+int smu_get_clock_by_type_with_voltage(struct smu_context *smu,
+				       enum amd_pp_clock_type type,
+				       struct pp_clock_levels_with_voltage *clocks);
+
+int smu_display_clock_voltage_request(struct smu_context *smu,
+				      struct pp_display_clock_request *clock_req);
+int smu_display_disable_memory_clock_switch(struct smu_context *smu, bool disable_memory_clock_switch);
+int smu_notify_smu_enable_pwe(struct smu_context *smu);
+
+int smu_set_xgmi_pstate(struct smu_context *smu,
+			uint32_t pstate);
+
+int smu_set_azalia_d3_pme(struct smu_context *smu);
+
+bool smu_baco_is_support(struct smu_context *smu);
+
+int smu_baco_get_state(struct smu_context *smu, enum smu_baco_state *state);
+
+int smu_baco_reset(struct smu_context *smu);
 
+int smu_mode2_reset(struct smu_context *smu);
 
 extern int smu_get_atom_data_table(struct smu_context *smu, uint32_t table,
 				   uint16_t *size, uint8_t *frev, uint8_t *crev,
@@ -798,6 +661,10 @@ int smu_sys_get_pp_table(struct smu_context *smu, void **table);
 int smu_sys_set_pp_table(struct smu_context *smu,  void *buf, size_t size);
 int smu_get_power_num_states(struct smu_context *smu, struct pp_states_info *state_info);
 enum amd_pm_state_type smu_get_current_power_state(struct smu_context *smu);
+int smu_write_watermarks_table(struct smu_context *smu);
+int smu_set_watermarks_for_clock_ranges(
+		struct smu_context *smu,
+		struct dm_pp_wm_sets_with_clock_ranges_soc15 *clock_ranges);
 
 /* smu to display interface */
 extern int smu_display_configuration_change(struct smu_context *smu, const
@@ -808,7 +675,8 @@ extern int smu_get_current_clocks(struct smu_context *smu,
 extern int smu_dpm_set_power_gate(struct smu_context *smu,uint32_t block_type, bool gate);
 extern int smu_handle_task(struct smu_context *smu,
 			   enum amd_dpm_forced_level level,
-			   enum amd_pp_task task_id);
+			   enum amd_pp_task task_id,
+			   bool lock_needed);
 int smu_switch_power_profile(struct smu_context *smu,
 			     enum PP_SMC_POWER_PROFILE type,
 			     bool en);
@@ -818,7 +686,7 @@ int smu_get_dpm_freq_by_index(struct smu_context *smu, enum smu_clk_type clk_typ
 int smu_get_dpm_level_count(struct smu_context *smu, enum smu_clk_type clk_type,
 			    uint32_t *value);
 int smu_get_dpm_freq_range(struct smu_context *smu, enum smu_clk_type clk_type,
-			   uint32_t *min, uint32_t *max);
+			   uint32_t *min, uint32_t *max, bool lock_needed);
 int smu_set_soft_freq_range(struct smu_context *smu, enum smu_clk_type clk_type,
 			    uint32_t min, uint32_t max);
 int smu_set_hard_freq_range(struct smu_context *smu, enum smu_clk_type clk_type,
@@ -827,10 +695,29 @@ enum amd_dpm_forced_level smu_get_performance_level(struct smu_context *smu);
 int smu_force_performance_level(struct smu_context *smu, enum amd_dpm_forced_level level);
 int smu_set_display_count(struct smu_context *smu, uint32_t count);
 bool smu_clk_dpm_is_enabled(struct smu_context *smu, enum smu_clk_type clk_type);
-int smu_feature_update_enable_state(struct smu_context *smu, uint64_t feature_mask, bool enabled);
 const char *smu_get_message_name(struct smu_context *smu, enum smu_message_type type);
 const char *smu_get_feature_name(struct smu_context *smu, enum smu_feature_mask feature);
 size_t smu_sys_get_pp_feature_mask(struct smu_context *smu, char *buf);
 int smu_sys_set_pp_feature_mask(struct smu_context *smu, uint64_t new_mask);
+int smu_force_clk_levels(struct smu_context *smu,
+			 enum smu_clk_type clk_type,
+			 uint32_t mask,
+			 bool lock_needed);
+int smu_set_mp1_state(struct smu_context *smu,
+		      enum pp_mp1_state mp1_state);
+int smu_set_df_cstate(struct smu_context *smu,
+		      enum pp_df_cstate state);
+
+int smu_get_max_sustainable_clocks_by_dc(struct smu_context *smu,
+					 struct pp_smu_nv_clock_table *max_clocks);
+
+int smu_get_uclk_dpm_states(struct smu_context *smu,
+			    unsigned int *clock_values_in_khz,
+			    unsigned int *num_states);
+
+int smu_get_dpm_clock_table(struct smu_context *smu,
+			    struct dpm_clocks *clock_table);
+
+uint32_t smu_get_pptable_power_limit(struct smu_context *smu);
 
 #endif
diff --git a/drivers/gpu/drm/amd/powerplay/inc/arcturus_ppsmc.h b/drivers/gpu/drm/amd/powerplay/inc/arcturus_ppsmc.h
index 78e5927b7711..e3291259b249 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/arcturus_ppsmc.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/arcturus_ppsmc.h
@@ -95,8 +95,7 @@
 
 //BTC
 #define PPSMC_MSG_RunAfllBtc                     0x30
-#define PPSMC_MSG_RunGfxDcBtc                    0x31
-#define PPSMC_MSG_RunSocDcBtc                    0x32
+#define PPSMC_MSG_RunDcBtc                       0x31
 
 //Debug
 #define PPSMC_MSG_DramLogSetDramAddrHigh         0x33
diff --git a/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h b/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
index 7bf9a14bfa0b..af977675fd33 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
@@ -355,6 +355,10 @@ struct pp_hwmgr_func {
 	int (*set_mp1_state)(struct pp_hwmgr *hwmgr, enum pp_mp1_state mp1_state);
 	int (*asic_reset)(struct pp_hwmgr *hwmgr, enum SMU_ASIC_RESET_MODE mode);
 	int (*smu_i2c_bus_access)(struct pp_hwmgr *hwmgr, bool aquire);
+	int (*set_df_cstate)(struct pp_hwmgr *hwmgr, enum pp_df_cstate state);
+	int (*set_xgmi_pstate)(struct pp_hwmgr *hwmgr, uint32_t pstate);
+	int (*disable_power_features_for_compute_performance)(struct pp_hwmgr *hwmgr,
+					bool disable);
 };
 
 struct pp_table_func {
diff --git a/drivers/gpu/drm/amd/powerplay/inc/smu11_driver_if_arcturus.h b/drivers/gpu/drm/amd/powerplay/inc/smu11_driver_if_arcturus.h
index e02950b505fa..a886f0644d24 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/smu11_driver_if_arcturus.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/smu11_driver_if_arcturus.h
@@ -137,29 +137,29 @@
 #define FEATURE_DS_SOCCLK_MASK            (1 << FEATURE_DS_SOCCLK_BIT            )
 #define FEATURE_DS_LCLK_MASK              (1 << FEATURE_DS_LCLK_BIT              )
 #define FEATURE_DS_FCLK_MASK              (1 << FEATURE_DS_FCLK_BIT              )
-#define FEATURE_DS_LCLK_MASK              (1 << FEATURE_DS_LCLK_BIT              )
+#define FEATURE_DS_UCLK_MASK              (1 << FEATURE_DS_UCLK_BIT              )
 #define FEATURE_GFX_ULV_MASK              (1 << FEATURE_GFX_ULV_BIT              )
-#define FEATURE_VCN_PG_MASK               (1 << FEATURE_VCN_PG_BIT               )
+#define FEATURE_DPM_VCN_MASK              (1 << FEATURE_DPM_VCN_BIT              )
 #define FEATURE_RSMU_SMN_CG_MASK          (1 << FEATURE_RSMU_SMN_CG_BIT          )
 #define FEATURE_WAFL_CG_MASK              (1 << FEATURE_WAFL_CG_BIT              )
 
 #define FEATURE_PPT_MASK                  (1 << FEATURE_PPT_BIT                  )
 #define FEATURE_TDC_MASK                  (1 << FEATURE_TDC_BIT                  )
-#define FEATURE_APCC_MASK                 (1 << FEATURE_APCC_BIT                 )
+#define FEATURE_APCC_PLUS_MASK            (1 << FEATURE_APCC_PLUS_BIT            )
 #define FEATURE_VR0HOT_MASK               (1 << FEATURE_VR0HOT_BIT               )
 #define FEATURE_VR1HOT_MASK               (1 << FEATURE_VR1HOT_BIT               )
 #define FEATURE_FW_CTF_MASK               (1 << FEATURE_FW_CTF_BIT               )
 #define FEATURE_FAN_CONTROL_MASK          (1 << FEATURE_FAN_CONTROL_BIT          )
 #define FEATURE_THERMAL_MASK              (1 << FEATURE_THERMAL_BIT              )
 
-#define FEATURE_OUT_OF_BAND_MONITOR_MASK  (1 << EATURE_OUT_OF_BAND_MONITOR_BIT   )
-#define FEATURE_TEMP_DEPENDENT_VMIN_MASK  (1 << FEATURE_TEMP_DEPENDENT_VMIN_MASK )
+#define FEATURE_OUT_OF_BAND_MONITOR_MASK  (1 << FEATURE_OUT_OF_BAND_MONITOR_BIT   )
+#define FEATURE_TEMP_DEPENDENT_VMIN_MASK  (1 << FEATURE_TEMP_DEPENDENT_VMIN_BIT )
 
 
 //FIXME need updating
 // Debug Overrides Bitmask
 #define DPM_OVERRIDE_DISABLE_UCLK_PID               0x00000001
-#define DPM_OVERRIDE_ENABLE_VOLT_LINK_VCN_FCLK      0x00000002
+#define DPM_OVERRIDE_DISABLE_VOLT_LINK_VCN_FCLK     0x00000002
 
 // I2C Config Bit Defines
 #define I2C_CONTROLLER_ENABLED           1
@@ -423,18 +423,30 @@ typedef enum {
 } PwrConfig_e;
 
 typedef enum {
-  XGMI_LINK_RATE_12 = 0,  // 12Gbps
-  XGMI_LINK_RATE_16,      // 16Gbps
-  XGMI_LINK_RATE_22,      // 22Gbps
-  XGMI_LINK_RATE_25,      // 25Gbps
+  XGMI_LINK_RATE_2 = 2,    // 2Gbps
+  XGMI_LINK_RATE_4 = 4,    // 4Gbps
+  XGMI_LINK_RATE_8 = 8,    // 8Gbps
+  XGMI_LINK_RATE_12 = 12,  // 12Gbps
+  XGMI_LINK_RATE_16 = 16,  // 16Gbps
+  XGMI_LINK_RATE_17 = 17,  // 17Gbps
+  XGMI_LINK_RATE_18 = 18,  // 18Gbps
+  XGMI_LINK_RATE_19 = 19,  // 19Gbps
+  XGMI_LINK_RATE_20 = 20,  // 20Gbps
+  XGMI_LINK_RATE_21 = 21,  // 21Gbps
+  XGMI_LINK_RATE_22 = 22,  // 22Gbps
+  XGMI_LINK_RATE_23 = 23,  // 23Gbps
+  XGMI_LINK_RATE_24 = 24,  // 24Gbps
+  XGMI_LINK_RATE_25 = 25,  // 25Gbps
   XGMI_LINK_RATE_COUNT
 } XGMI_LINK_RATE_e;
 
 typedef enum {
-  XGMI_LINK_WIDTH_2 = 0, // x2
-  XGMI_LINK_WIDTH_4,     // x4
-  XGMI_LINK_WIDTH_8,     // x8
-  XGMI_LINK_WIDTH_16,    // x16
+  XGMI_LINK_WIDTH_1 = 1,   // x1
+  XGMI_LINK_WIDTH_2 = 2,   // x2
+  XGMI_LINK_WIDTH_4 = 4,   // x4
+  XGMI_LINK_WIDTH_8 = 8,   // x8
+  XGMI_LINK_WIDTH_9 = 9,   // x9
+  XGMI_LINK_WIDTH_16 = 16, // x16
   XGMI_LINK_WIDTH_COUNT
 } XGMI_LINK_WIDTH_e;
 
@@ -696,7 +708,11 @@ typedef struct {
   uint8_t      GpioI2cSda;          // Serial Data
   uint16_t     GpioPadding;
 
-  uint32_t     BoardReserved[9];
+  // Platform input telemetry voltage coefficient
+  uint32_t     BoardVoltageCoeffA;    // decode by /1000
+  uint32_t     BoardVoltageCoeffB;    // decode by /1000
+
+  uint32_t     BoardReserved[7];
 
   // Padding for MMHUB - do not modify this
   uint32_t     MmHubPadding[8]; // SMU internal use
@@ -802,7 +818,7 @@ typedef struct {
 
   uint32_t P2VCharzFreq[AVFS_VOLTAGE_COUNT]; // in 10KHz units
 
-  uint32_t EnabledAvfsModules[2];
+  uint32_t EnabledAvfsModules[3];
 
   uint32_t MmHubPadding[8]; // SMU internal use
 } AvfsFuseOverride_t;
@@ -865,7 +881,8 @@ typedef struct {
 //#define TABLE_ACTIVITY_MONITOR_COEFF  7
 #define TABLE_OVERDRIVE               7
 #define TABLE_WAFL_XGMI_TOPOLOGY      8
-#define TABLE_COUNT                   9
+#define TABLE_I2C_COMMANDS            9
+#define TABLE_COUNT                   10
 
 // These defines are used with the SMC_MSG_SetUclkFastSwitch message.
 typedef enum {
diff --git a/drivers/gpu/drm/amd/powerplay/inc/smu_types.h b/drivers/gpu/drm/amd/powerplay/inc/smu_types.h
index b0dd05d431dd..d8c9b7f91fcc 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/smu_types.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/smu_types.h
@@ -114,6 +114,7 @@
        __SMU_DUMMY_MAP(PowerDownJpeg),                \
        __SMU_DUMMY_MAP(BacoAudioD3PME),               \
        __SMU_DUMMY_MAP(ArmD3),                        \
+       __SMU_DUMMY_MAP(RunDcBtc),                     \
        __SMU_DUMMY_MAP(RunGfxDcBtc),                  \
        __SMU_DUMMY_MAP(RunSocDcBtc),                  \
        __SMU_DUMMY_MAP(SetMemoryChannelEnable),       \
@@ -168,6 +169,7 @@
 	__SMU_DUMMY_MAP(PowerGateAtHub),              \
 	__SMU_DUMMY_MAP(SetSoftMinJpeg),              \
 	__SMU_DUMMY_MAP(SetHardMinFclkByFreq),        \
+	__SMU_DUMMY_MAP(DFCstateControl), \
 
 #undef __SMU_DUMMY_MAP
 #define __SMU_DUMMY_MAP(type)	SMU_MSG_##type
@@ -251,6 +253,7 @@ enum smu_clk_type {
        __SMU_DUMMY_MAP(TEMP_DEPENDENT_VMIN),           	\
        __SMU_DUMMY_MAP(MMHUB_PG),                      	\
        __SMU_DUMMY_MAP(ATHUB_PG),                      	\
+       __SMU_DUMMY_MAP(APCC_DFLL),                     	\
        __SMU_DUMMY_MAP(WAFL_CG),
 
 #undef __SMU_DUMMY_MAP
diff --git a/drivers/gpu/drm/amd/powerplay/inc/smu_v11_0.h b/drivers/gpu/drm/amd/powerplay/inc/smu_v11_0.h
index 5bda8539447a..606149085683 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/smu_v11_0.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/smu_v11_0.h
@@ -27,7 +27,7 @@
 
 #define SMU11_DRIVER_IF_VERSION_INV 0xFFFFFFFF
 #define SMU11_DRIVER_IF_VERSION_VG20 0x13
-#define SMU11_DRIVER_IF_VERSION_ARCT 0x09
+#define SMU11_DRIVER_IF_VERSION_ARCT 0x10
 #define SMU11_DRIVER_IF_VERSION_NV10 0x33
 #define SMU11_DRIVER_IF_VERSION_NV14 0x34
 
@@ -48,6 +48,8 @@
 
 #define SMU11_TOOL_SIZE			0x19000
 
+#define MAX_PCIE_CONF 2
+
 #define CLK_MAP(clk, index) \
 	[SMU_##clk] = {1, (index)}
 
@@ -88,6 +90,11 @@ struct smu_11_0_dpm_table {
 	uint32_t    max;        /* MHz */
 };
 
+struct smu_11_0_pcie_table {
+        uint8_t  pcie_gen[MAX_PCIE_CONF];
+        uint8_t  pcie_lane[MAX_PCIE_CONF];
+};
+
 struct smu_11_0_dpm_tables {
 	struct smu_11_0_dpm_table        soc_table;
 	struct smu_11_0_dpm_table        gfx_table;
@@ -100,6 +107,7 @@ struct smu_11_0_dpm_tables {
 	struct smu_11_0_dpm_table        display_table;
 	struct smu_11_0_dpm_table        phy_table;
 	struct smu_11_0_dpm_table        fclk_table;
+	struct smu_11_0_pcie_table       pcie_table;
 };
 
 struct smu_11_0_dpm_context {
@@ -130,6 +138,128 @@ enum smu_v11_0_baco_seq {
 	BACO_SEQ_COUNT,
 };
 
-void smu_v11_0_set_smu_funcs(struct smu_context *smu);
+int smu_v11_0_init_microcode(struct smu_context *smu);
+
+int smu_v11_0_load_microcode(struct smu_context *smu);
+
+int smu_v11_0_init_smc_tables(struct smu_context *smu);
+
+int smu_v11_0_fini_smc_tables(struct smu_context *smu);
+
+int smu_v11_0_init_power(struct smu_context *smu);
+
+int smu_v11_0_fini_power(struct smu_context *smu);
+
+int smu_v11_0_check_fw_status(struct smu_context *smu);
+
+int smu_v11_0_setup_pptable(struct smu_context *smu);
+
+int smu_v11_0_get_vbios_bootup_values(struct smu_context *smu);
+
+int smu_v11_0_get_clk_info_from_vbios(struct smu_context *smu);
+
+int smu_v11_0_check_pptable(struct smu_context *smu);
+
+int smu_v11_0_parse_pptable(struct smu_context *smu);
+
+int smu_v11_0_populate_smc_pptable(struct smu_context *smu);
+
+int smu_v11_0_check_fw_version(struct smu_context *smu);
+
+int smu_v11_0_write_pptable(struct smu_context *smu);
+
+int smu_v11_0_set_min_dcef_deep_sleep(struct smu_context *smu);
+
+int smu_v11_0_set_tool_table_location(struct smu_context *smu);
+
+int smu_v11_0_notify_memory_pool_location(struct smu_context *smu);
+
+int smu_v11_0_system_features_control(struct smu_context *smu,
+					     bool en);
+
+int smu_v11_0_send_msg(struct smu_context *smu, uint16_t msg);
+
+int
+smu_v11_0_send_msg_with_param(struct smu_context *smu, uint16_t msg,
+			      uint32_t param);
+
+int smu_v11_0_read_arg(struct smu_context *smu, uint32_t *arg);
+
+int smu_v11_0_init_display_count(struct smu_context *smu, uint32_t count);
+
+int smu_v11_0_set_allowed_mask(struct smu_context *smu);
+
+int smu_v11_0_get_enabled_mask(struct smu_context *smu,
+				      uint32_t *feature_mask, uint32_t num);
+
+int smu_v11_0_notify_display_change(struct smu_context *smu);
+
+int smu_v11_0_set_power_limit(struct smu_context *smu, uint32_t n);
+
+int smu_v11_0_get_current_clk_freq(struct smu_context *smu,
+					  enum smu_clk_type clk_id,
+					  uint32_t *value);
+
+int smu_v11_0_init_max_sustainable_clocks(struct smu_context *smu);
+
+int smu_v11_0_start_thermal_control(struct smu_context *smu);
+
+int smu_v11_0_stop_thermal_control(struct smu_context *smu);
+
+int smu_v11_0_read_sensor(struct smu_context *smu,
+				 enum amd_pp_sensors sensor,
+				 void *data, uint32_t *size);
+
+int smu_v11_0_set_deep_sleep_dcefclk(struct smu_context *smu, uint32_t clk);
+
+int
+smu_v11_0_display_clock_voltage_request(struct smu_context *smu,
+					struct pp_display_clock_request
+					*clock_req);
+
+uint32_t
+smu_v11_0_get_fan_control_mode(struct smu_context *smu);
+
+int
+smu_v11_0_set_fan_control_mode(struct smu_context *smu,
+			       uint32_t mode);
+
+int
+smu_v11_0_set_fan_speed_percent(struct smu_context *smu, uint32_t speed);
+
+int smu_v11_0_set_fan_speed_rpm(struct smu_context *smu,
+				       uint32_t speed);
+
+int smu_v11_0_set_xgmi_pstate(struct smu_context *smu,
+				     uint32_t pstate);
+
+int smu_v11_0_gfx_off_control(struct smu_context *smu, bool enable);
+
+int smu_v11_0_register_irq_handler(struct smu_context *smu);
+
+int smu_v11_0_set_azalia_d3_pme(struct smu_context *smu);
+
+int smu_v11_0_get_max_sustainable_clocks_by_dc(struct smu_context *smu,
+		struct pp_smu_nv_clock_table *max_clocks);
+
+bool smu_v11_0_baco_is_support(struct smu_context *smu);
+
+enum smu_baco_state smu_v11_0_baco_get_state(struct smu_context *smu);
+
+int smu_v11_0_baco_set_state(struct smu_context *smu, enum smu_baco_state state);
+
+int smu_v11_0_baco_reset(struct smu_context *smu);
+
+int smu_v11_0_get_dpm_ultimate_freq(struct smu_context *smu, enum smu_clk_type clk_type,
+						 uint32_t *min, uint32_t *max);
+
+int smu_v11_0_set_soft_freq_limited_range(struct smu_context *smu, enum smu_clk_type clk_type,
+			    uint32_t min, uint32_t max);
+
+int smu_v11_0_override_pcie_parameters(struct smu_context *smu);
+
+int smu_v11_0_set_default_od_settings(struct smu_context *smu, bool initialize, size_t overdrive_table_size);
+
+uint32_t smu_v11_0_get_max_power_limit(struct smu_context *smu);
 
 #endif
diff --git a/drivers/gpu/drm/amd/powerplay/inc/smu_v11_0_pptable.h b/drivers/gpu/drm/amd/powerplay/inc/smu_v11_0_pptable.h
index 86cdc3393eac..b2f96a101124 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/smu_v11_0_pptable.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/smu_v11_0_pptable.h
@@ -141,7 +141,9 @@ struct smu_11_0_powerplay_table
       struct smu_11_0_power_saving_clock_table      power_saving_clock;
       struct smu_11_0_overdrive_table               overdrive_table;
 
+#ifndef SMU_11_0_PARTIAL_PPTABLE
       PPTable_t smc_pptable;                        //PPTable_t in smu11_driver_if.h
+#endif
 } __attribute__((packed));
 
 #endif
diff --git a/drivers/gpu/drm/amd/powerplay/inc/smu_v12_0.h b/drivers/gpu/drm/amd/powerplay/inc/smu_v12_0.h
index acf3db12f59f..9b9f5df0911c 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/smu_v12_0.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/smu_v12_0.h
@@ -37,6 +37,45 @@ struct smu_12_0_cmn2aisc_mapping {
 	int	map_to;
 };
 
-void smu_v12_0_set_smu_funcs(struct smu_context *smu);
+int smu_v12_0_send_msg_without_waiting(struct smu_context *smu,
+					      uint16_t msg);
+
+int smu_v12_0_read_arg(struct smu_context *smu, uint32_t *arg);
+
+int smu_v12_0_wait_for_response(struct smu_context *smu);
+
+int smu_v12_0_send_msg(struct smu_context *smu, uint16_t msg);
+
+int
+smu_v12_0_send_msg_with_param(struct smu_context *smu, uint16_t msg,
+			      uint32_t param);
+
+int smu_v12_0_check_fw_status(struct smu_context *smu);
+
+int smu_v12_0_check_fw_version(struct smu_context *smu);
+
+int smu_v12_0_powergate_sdma(struct smu_context *smu, bool gate);
+
+int smu_v12_0_powergate_vcn(struct smu_context *smu, bool gate);
+
+int smu_v12_0_set_gfx_cgpg(struct smu_context *smu, bool enable);
+
+uint32_t smu_v12_0_get_gfxoff_status(struct smu_context *smu);
+
+int smu_v12_0_gfx_off_control(struct smu_context *smu, bool enable);
+
+int smu_v12_0_init_smc_tables(struct smu_context *smu);
+
+int smu_v12_0_fini_smc_tables(struct smu_context *smu);
+
+int smu_v12_0_populate_smc_tables(struct smu_context *smu);
+
+int smu_v12_0_get_dpm_ultimate_freq(struct smu_context *smu, enum smu_clk_type clk_type,
+						 uint32_t *min, uint32_t *max);
+
+int smu_v12_0_mode2_reset(struct smu_context *smu);
+
+int smu_v12_0_set_soft_freq_limited_range(struct smu_context *smu, enum smu_clk_type clk_type,
+			    uint32_t min, uint32_t max);
 
 #endif
diff --git a/drivers/gpu/drm/amd/powerplay/inc/vega20_ppsmc.h b/drivers/gpu/drm/amd/powerplay/inc/vega20_ppsmc.h
index a0883038f3c3..0c66f0fe1aaf 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/vega20_ppsmc.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/vega20_ppsmc.h
@@ -120,7 +120,8 @@
 #define PPSMC_MSG_SetMGpuFanBoostLimitRpm        0x5D
 #define PPSMC_MSG_GetAVFSVoltageByDpm            0x5F
 #define PPSMC_MSG_BacoWorkAroundFlushVDCI        0x60
-#define PPSMC_Message_Count                      0x61
+#define PPSMC_MSG_DFCstateControl                0x63
+#define PPSMC_Message_Count                      0x64
 
 typedef uint32_t PPSMC_Result;
 typedef uint32_t PPSMC_Msg;
diff --git a/drivers/gpu/drm/amd/powerplay/navi10_ppt.c b/drivers/gpu/drm/amd/powerplay/navi10_ppt.c
index 12c0e469bf35..aaec884d63ed 100644
--- a/drivers/gpu/drm/amd/powerplay/navi10_ppt.c
+++ b/drivers/gpu/drm/amd/powerplay/navi10_ppt.c
@@ -26,6 +26,7 @@
 #include <linux/pci.h>
 #include "amdgpu.h"
 #include "amdgpu_smu.h"
+#include "smu_internal.h"
 #include "atomfirmware.h"
 #include "amdgpu_atomfirmware.h"
 #include "smu_v11_0.h"
@@ -35,6 +36,7 @@
 #include "navi10_ppt.h"
 #include "smu_v11_0_pptable.h"
 #include "smu_v11_0_ppsmc.h"
+#include "nbio/nbio_7_4_sh_mask.h"
 
 #include "asic_reg/mp/mp_11_0_sh_mask.h"
 
@@ -177,6 +179,7 @@ static struct smu_11_0_cmn2aisc_mapping navi10_feature_mask_map[SMU_FEATURE_COUN
 	FEA_MAP(TEMP_DEPENDENT_VMIN),
 	FEA_MAP(MMHUB_PG),
 	FEA_MAP(ATHUB_PG),
+	FEA_MAP(APCC_DFLL),
 };
 
 static struct smu_11_0_cmn2aisc_mapping navi10_table_map[SMU_TABLE_COUNT] = {
@@ -205,7 +208,7 @@ static struct smu_11_0_cmn2aisc_mapping navi10_workload_map[PP_SMC_POWER_PROFILE
 	WORKLOAD_MAP(PP_SMC_POWER_PROFILE_POWERSAVING,		WORKLOAD_PPLIB_POWER_SAVING_BIT),
 	WORKLOAD_MAP(PP_SMC_POWER_PROFILE_VIDEO,		WORKLOAD_PPLIB_VIDEO_BIT),
 	WORKLOAD_MAP(PP_SMC_POWER_PROFILE_VR,			WORKLOAD_PPLIB_VR_BIT),
-	WORKLOAD_MAP(PP_SMC_POWER_PROFILE_COMPUTE,		WORKLOAD_PPLIB_CUSTOM_BIT),
+	WORKLOAD_MAP(PP_SMC_POWER_PROFILE_COMPUTE,		WORKLOAD_PPLIB_COMPUTE_BIT),
 	WORKLOAD_MAP(PP_SMC_POWER_PROFILE_CUSTOM,		WORKLOAD_PPLIB_CUSTOM_BIT),
 };
 
@@ -327,40 +330,52 @@ navi10_get_allowed_feature_mask(struct smu_context *smu,
 	memset(feature_mask, 0, sizeof(uint32_t) * num);
 
 	*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DPM_PREFETCHER_BIT)
-				| FEATURE_MASK(FEATURE_DPM_GFXCLK_BIT)
-				| FEATURE_MASK(FEATURE_DPM_SOCCLK_BIT)
 				| FEATURE_MASK(FEATURE_DPM_MP0CLK_BIT)
-				| FEATURE_MASK(FEATURE_DPM_LINK_BIT)
-				| FEATURE_MASK(FEATURE_GFX_ULV_BIT)
 				| FEATURE_MASK(FEATURE_RSMU_SMN_CG_BIT)
 				| FEATURE_MASK(FEATURE_DS_SOCCLK_BIT)
 				| FEATURE_MASK(FEATURE_PPT_BIT)
 				| FEATURE_MASK(FEATURE_TDC_BIT)
 				| FEATURE_MASK(FEATURE_GFX_EDC_BIT)
+				| FEATURE_MASK(FEATURE_APCC_PLUS_BIT)
 				| FEATURE_MASK(FEATURE_VR0HOT_BIT)
 				| FEATURE_MASK(FEATURE_FAN_CONTROL_BIT)
 				| FEATURE_MASK(FEATURE_THERMAL_BIT)
 				| FEATURE_MASK(FEATURE_LED_DISPLAY_BIT)
-				| FEATURE_MASK(FEATURE_DPM_DCEFCLK_BIT)
-				| FEATURE_MASK(FEATURE_DS_GFXCLK_BIT)
+				| FEATURE_MASK(FEATURE_DS_LCLK_BIT)
 				| FEATURE_MASK(FEATURE_DS_DCEFCLK_BIT)
 				| FEATURE_MASK(FEATURE_FW_DSTATE_BIT)
 				| FEATURE_MASK(FEATURE_BACO_BIT)
 				| FEATURE_MASK(FEATURE_ACDC_BIT)
 				| FEATURE_MASK(FEATURE_GFX_SS_BIT)
 				| FEATURE_MASK(FEATURE_APCC_DFLL_BIT)
-				| FEATURE_MASK(FEATURE_FW_CTF_BIT);
+				| FEATURE_MASK(FEATURE_FW_CTF_BIT)
+				| FEATURE_MASK(FEATURE_OUT_OF_BAND_MONITOR_BIT);
+
+	if (adev->pm.pp_feature & PP_SOCCLK_DPM_MASK)
+		*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DPM_SOCCLK_BIT);
+
+	if (adev->pm.pp_feature & PP_SCLK_DPM_MASK)
+		*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DPM_GFXCLK_BIT);
+
+	if (adev->pm.pp_feature & PP_PCIE_DPM_MASK)
+		*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DPM_LINK_BIT);
+
+	if (adev->pm.pp_feature & PP_DCEFCLK_DPM_MASK)
+		*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DPM_DCEFCLK_BIT);
 
 	if (adev->pm.pp_feature & PP_MCLK_DPM_MASK)
 		*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DPM_UCLK_BIT)
 				| FEATURE_MASK(FEATURE_MEM_VDDCI_SCALING_BIT)
 				| FEATURE_MASK(FEATURE_MEM_MVDD_SCALING_BIT);
 
-	if (adev->pm.pp_feature & PP_GFXOFF_MASK) {
+	if (adev->pm.pp_feature & PP_ULV_MASK)
+		*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_GFX_ULV_BIT);
+
+	if (adev->pm.pp_feature & PP_SCLK_DEEP_SLEEP_MASK)
+		*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DS_GFXCLK_BIT);
+
+	if (adev->pm.pp_feature & PP_GFXOFF_MASK)
 		*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_GFXOFF_BIT);
-		/* TODO: remove it once fw fix the bug */
-		*(uint64_t *)feature_mask &= ~FEATURE_MASK(FEATURE_FW_DSTATE_BIT);
-	}
 
 	if (smu->adev->pg_flags & AMD_PG_SUPPORT_MMHUB)
 		*(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_MMHUB_PG_BIT);
@@ -547,7 +562,7 @@ static int navi10_get_metrics_table(struct smu_context *smu,
 	struct smu_table_context *smu_table= &smu->smu_table;
 	int ret = 0;
 
-	if (!smu_table->metrics_time || time_after(jiffies, smu_table->metrics_time + HZ / 1000)) {
+	if (!smu_table->metrics_time || time_after(jiffies, smu_table->metrics_time + msecs_to_jiffies(100))) {
 		ret = smu_update_table(smu, SMU_TABLE_SMU_METRICS, 0,
 				(void *)smu_table->metrics_table, false);
 		if (ret) {
@@ -585,6 +600,7 @@ static int navi10_set_default_dpm_table(struct smu_context *smu)
 	struct smu_table_context *table_context = &smu->smu_table;
 	struct smu_11_0_dpm_context *dpm_context = smu_dpm->dpm_context;
 	PPTable_t *driver_ppt = NULL;
+	int i;
 
 	driver_ppt = table_context->driver_pptable;
 
@@ -615,6 +631,11 @@ static int navi10_set_default_dpm_table(struct smu_context *smu)
 	dpm_context->dpm_tables.phy_table.min = driver_ppt->FreqTablePhyclk[0];
 	dpm_context->dpm_tables.phy_table.max = driver_ppt->FreqTablePhyclk[NUM_PHYCLK_DPM_LEVELS - 1];
 
+	for (i = 0; i < MAX_PCIE_CONF; i++) {
+		dpm_context->dpm_tables.pcie_table.pcie_gen[i] = driver_ppt->PcieGenSpeed[i];
+		dpm_context->dpm_tables.pcie_table.pcie_lane[i] = driver_ppt->PcieLaneCount[i];
+	}
+
 	return 0;
 }
 
@@ -677,13 +698,29 @@ static bool navi10_is_support_fine_grained_dpm(struct smu_context *smu, enum smu
 	return dpm_desc->SnapToDiscrete == 0 ? true : false;
 }
 
+static inline bool navi10_od_feature_is_supported(struct smu_11_0_overdrive_table *od_table, enum SMU_11_0_ODFEATURE_ID feature)
+{
+	return od_table->cap[feature];
+}
+
+
 static int navi10_print_clk_levels(struct smu_context *smu,
 			enum smu_clk_type clk_type, char *buf)
 {
+	uint16_t *curve_settings;
 	int i, size = 0, ret = 0;
 	uint32_t cur_value = 0, value = 0, count = 0;
 	uint32_t freq_values[3] = {0};
 	uint32_t mark_index = 0;
+	struct smu_table_context *table_context = &smu->smu_table;
+	uint32_t gen_speed, lane_width;
+	struct smu_dpm_context *smu_dpm = &smu->smu_dpm;
+	struct smu_11_0_dpm_context *dpm_context = smu_dpm->dpm_context;
+	struct amdgpu_device *adev = smu->adev;
+	PPTable_t *pptable = (PPTable_t *)table_context->driver_pptable;
+	OverDriveTable_t *od_table =
+		(OverDriveTable_t *)table_context->overdrive_table;
+	struct smu_11_0_overdrive_table *od_settings = smu->od_settings;
 
 	switch (clk_type) {
 	case SMU_GFXCLK:
@@ -734,6 +771,69 @@ static int navi10_print_clk_levels(struct smu_context *smu,
 
 		}
 		break;
+	case SMU_PCIE:
+		gen_speed = (RREG32_PCIE(smnPCIE_LC_SPEED_CNTL) &
+			     PSWUSP0_PCIE_LC_SPEED_CNTL__LC_CURRENT_DATA_RATE_MASK)
+			>> PSWUSP0_PCIE_LC_SPEED_CNTL__LC_CURRENT_DATA_RATE__SHIFT;
+		lane_width = (RREG32_PCIE(smnPCIE_LC_LINK_WIDTH_CNTL) &
+			      PCIE_LC_LINK_WIDTH_CNTL__LC_LINK_WIDTH_RD_MASK)
+			>> PCIE_LC_LINK_WIDTH_CNTL__LC_LINK_WIDTH_RD__SHIFT;
+		for (i = 0; i < NUM_LINK_LEVELS; i++)
+			size += sprintf(buf + size, "%d: %s %s %dMhz %s\n", i,
+					(dpm_context->dpm_tables.pcie_table.pcie_gen[i] == 0) ? "2.5GT/s," :
+					(dpm_context->dpm_tables.pcie_table.pcie_gen[i] == 1) ? "5.0GT/s," :
+					(dpm_context->dpm_tables.pcie_table.pcie_gen[i] == 2) ? "8.0GT/s," :
+					(dpm_context->dpm_tables.pcie_table.pcie_gen[i] == 3) ? "16.0GT/s," : "",
+					(dpm_context->dpm_tables.pcie_table.pcie_lane[i] == 1) ? "x1" :
+					(dpm_context->dpm_tables.pcie_table.pcie_lane[i] == 2) ? "x2" :
+					(dpm_context->dpm_tables.pcie_table.pcie_lane[i] == 3) ? "x4" :
+					(dpm_context->dpm_tables.pcie_table.pcie_lane[i] == 4) ? "x8" :
+					(dpm_context->dpm_tables.pcie_table.pcie_lane[i] == 5) ? "x12" :
+					(dpm_context->dpm_tables.pcie_table.pcie_lane[i] == 6) ? "x16" : "",
+					pptable->LclkFreq[i],
+					(gen_speed == dpm_context->dpm_tables.pcie_table.pcie_gen[i]) &&
+					(lane_width == dpm_context->dpm_tables.pcie_table.pcie_lane[i]) ?
+					"*" : "");
+		break;
+	case SMU_OD_SCLK:
+		if (!smu->od_enabled || !od_table || !od_settings)
+			break;
+		if (!navi10_od_feature_is_supported(od_settings, SMU_11_0_ODFEATURE_GFXCLK_LIMITS))
+			break;
+		size += sprintf(buf + size, "OD_SCLK:\n");
+		size += sprintf(buf + size, "0: %uMhz\n1: %uMhz\n", od_table->GfxclkFmin, od_table->GfxclkFmax);
+		break;
+	case SMU_OD_MCLK:
+		if (!smu->od_enabled || !od_table || !od_settings)
+			break;
+		if (!navi10_od_feature_is_supported(od_settings, SMU_11_0_ODFEATURE_UCLK_MAX))
+			break;
+		size += sprintf(buf + size, "OD_MCLK:\n");
+		size += sprintf(buf + size, "0: %uMHz\n", od_table->UclkFmax);
+		break;
+	case SMU_OD_VDDC_CURVE:
+		if (!smu->od_enabled || !od_table || !od_settings)
+			break;
+		if (!navi10_od_feature_is_supported(od_settings, SMU_11_0_ODFEATURE_GFXCLK_CURVE))
+			break;
+		size += sprintf(buf + size, "OD_VDDC_CURVE:\n");
+		for (i = 0; i < 3; i++) {
+			switch (i) {
+			case 0:
+				curve_settings = &od_table->GfxclkFreq1;
+				break;
+			case 1:
+				curve_settings = &od_table->GfxclkFreq2;
+				break;
+			case 2:
+				curve_settings = &od_table->GfxclkFreq3;
+				break;
+			default:
+				break;
+			}
+			size += sprintf(buf + size, "%d: %uMHz @ %umV\n", i, curve_settings[0], curve_settings[1] / NAVI10_VOLTAGE_SCALE);
+		}
+		break;
 	default:
 		break;
 	}
@@ -759,6 +859,12 @@ static int navi10_force_clk_levels(struct smu_context *smu,
 	case SMU_UCLK:
 	case SMU_DCEFCLK:
 	case SMU_FCLK:
+		/* There is only 2 levels for fine grained DPM */
+		if (navi10_is_support_fine_grained_dpm(smu, clk_type)) {
+			soft_max_level = (soft_max_level >= 1 ? 1 : 0);
+			soft_min_level = (soft_min_level >= 1 ? 1 : 0);
+		}
+
 		ret = smu_get_dpm_freq_by_index(smu, clk_type, soft_min_level, &min_freq);
 		if (ret)
 			return size;
@@ -783,13 +889,13 @@ static int navi10_populate_umd_state_clk(struct smu_context *smu)
 	int ret = 0;
 	uint32_t min_sclk_freq = 0, min_mclk_freq = 0;
 
-	ret = smu_get_dpm_freq_range(smu, SMU_SCLK, &min_sclk_freq, NULL);
+	ret = smu_get_dpm_freq_range(smu, SMU_SCLK, &min_sclk_freq, NULL, false);
 	if (ret)
 		return ret;
 
 	smu->pstate_sclk = min_sclk_freq * 100;
 
-	ret = smu_get_dpm_freq_range(smu, SMU_MCLK, &min_mclk_freq, NULL);
+	ret = smu_get_dpm_freq_range(smu, SMU_MCLK, &min_mclk_freq, NULL, false);
 	if (ret)
 		return ret;
 
@@ -842,7 +948,7 @@ static int navi10_pre_display_config_changed(struct smu_context *smu)
 		return ret;
 
 	if (smu_feature_is_enabled(smu, SMU_FEATURE_DPM_UCLK_BIT)) {
-		ret = smu_get_dpm_freq_range(smu, SMU_UCLK, NULL, &max_freq);
+		ret = smu_get_dpm_freq_range(smu, SMU_UCLK, NULL, &max_freq, false);
 		if (ret)
 			return ret;
 		ret = smu_set_hard_freq_range(smu, SMU_UCLK, 0, max_freq);
@@ -892,7 +998,7 @@ static int navi10_force_dpm_limit_value(struct smu_context *smu, bool highest)
 
 	for (i = 0; i < ARRAY_SIZE(clks); i++) {
 		clk_type = clks[i];
-		ret = smu_get_dpm_freq_range(smu, clk_type, &min_freq, &max_freq);
+		ret = smu_get_dpm_freq_range(smu, clk_type, &min_freq, &max_freq, false);
 		if (ret)
 			return ret;
 
@@ -919,7 +1025,7 @@ static int navi10_unforce_dpm_levels(struct smu_context *smu)
 
 	for (i = 0; i < ARRAY_SIZE(clks); i++) {
 		clk_type = clks[i];
-		ret = smu_get_dpm_freq_range(smu, clk_type, &min_freq, &max_freq);
+		ret = smu_get_dpm_freq_range(smu, clk_type, &min_freq, &max_freq, false);
 		if (ret)
 			return ret;
 
@@ -1254,7 +1360,9 @@ static int navi10_notify_smc_dispaly_config(struct smu_context *smu)
 	if (smu_feature_is_supported(smu, SMU_FEATURE_DPM_DCEFCLK_BIT)) {
 		clock_req.clock_type = amd_pp_dcef_clock;
 		clock_req.clock_freq_in_khz = min_clocks.dcef_clock * 10;
-		if (!smu_display_clock_voltage_request(smu, &clock_req)) {
+
+		ret = smu_v11_0_display_clock_voltage_request(smu, &clock_req);
+		if (!ret) {
 			if (smu_feature_is_supported(smu, SMU_FEATURE_DS_DCEFCLK_BIT)) {
 				ret = smu_send_smc_msg_with_param(smu,
 								  SMU_MSG_SetMinDeepSleepDcefclk,
@@ -1386,6 +1494,7 @@ static int navi10_read_sensor(struct smu_context *smu,
 	if(!data || !size)
 		return -EINVAL;
 
+	mutex_lock(&smu->sensor_lock);
 	switch (sensor) {
 	case AMDGPU_PP_SENSOR_MAX_FAN_RPM:
 		*(uint32_t *)data = pptable->FanMaximumRpm;
@@ -1407,8 +1516,9 @@ static int navi10_read_sensor(struct smu_context *smu,
 		*size = 4;
 		break;
 	default:
-		ret = smu_smc_read_sensor(smu, sensor, data, size);
+		ret = smu_v11_0_read_sensor(smu, sensor, data, size);
 	}
+	mutex_unlock(&smu->sensor_lock);
 
 	return ret;
 }
@@ -1449,18 +1559,47 @@ static int navi10_set_peak_clock_by_device(struct smu_context *smu)
 	uint32_t sclk_freq = 0, uclk_freq = 0;
 	uint32_t uclk_level = 0;
 
-	switch (adev->pdev->revision) {
-	case 0xf0: /* XTX */
-	case 0xc0:
-		sclk_freq = NAVI10_PEAK_SCLK_XTX;
-		break;
-	case 0xf1: /* XT */
-	case 0xc1:
-		sclk_freq = NAVI10_PEAK_SCLK_XT;
+	switch (adev->asic_type) {
+	case CHIP_NAVI10:
+		switch (adev->pdev->revision) {
+		case 0xf0: /* XTX */
+		case 0xc0:
+			sclk_freq = NAVI10_PEAK_SCLK_XTX;
+			break;
+		case 0xf1: /* XT */
+		case 0xc1:
+			sclk_freq = NAVI10_PEAK_SCLK_XT;
+			break;
+		default: /* XL */
+			sclk_freq = NAVI10_PEAK_SCLK_XL;
+			break;
+		}
 		break;
-	default: /* XL */
-		sclk_freq = NAVI10_PEAK_SCLK_XL;
+	case CHIP_NAVI14:
+		switch (adev->pdev->revision) {
+		case 0xc7: /* XT */
+		case 0xf4:
+			sclk_freq = NAVI14_UMD_PSTATE_PEAK_XT_GFXCLK;
+			break;
+		case 0xc1: /* XTM */
+		case 0xf2:
+			sclk_freq = NAVI14_UMD_PSTATE_PEAK_XTM_GFXCLK;
+			break;
+		case 0xc3: /* XLM */
+		case 0xf3:
+			sclk_freq = NAVI14_UMD_PSTATE_PEAK_XLM_GFXCLK;
+			break;
+		case 0xc5: /* XTX */
+		case 0xf6:
+			sclk_freq = NAVI14_UMD_PSTATE_PEAK_XLM_GFXCLK;
+			break;
+		default: /* XL */
+			sclk_freq = NAVI14_UMD_PSTATE_PEAK_XL_GFXCLK;
+			break;
+		}
 		break;
+	default:
+		return -EINVAL;
 	}
 
 	ret = smu_get_dpm_level_count(smu, SMU_UCLK, &uclk_level);
@@ -1483,10 +1622,6 @@ static int navi10_set_peak_clock_by_device(struct smu_context *smu)
 static int navi10_set_performance_level(struct smu_context *smu, enum amd_dpm_forced_level level)
 {
 	int ret = 0;
-	struct amdgpu_device *adev = smu->adev;
-
-	if (adev->asic_type != CHIP_NAVI10)
-		return -EINVAL;
 
 	switch (level) {
 	case AMD_DPM_FORCED_LEVEL_PROFILE_PEAK:
@@ -1539,17 +1674,22 @@ static int navi10_display_disable_memory_clock_switch(struct smu_context *smu,
 	return ret;
 }
 
+static uint32_t navi10_get_pptable_power_limit(struct smu_context *smu)
+{
+	PPTable_t *pptable = smu->smu_table.driver_pptable;
+	return pptable->SocketPowerLimitAc[PPT_THROTTLER_PPT0];
+}
+
 static int navi10_get_power_limit(struct smu_context *smu,
 				     uint32_t *limit,
-				     bool asic_default)
+				     bool cap)
 {
 	PPTable_t *pptable = smu->smu_table.driver_pptable;
 	uint32_t asic_default_power_limit = 0;
 	int ret = 0;
 	int power_src;
 
-	if (!smu->default_power_limit ||
-	    !smu->power_limit) {
+	if (!smu->power_limit) {
 		if (smu_feature_is_enabled(smu, SMU_FEATURE_PPT_BIT)) {
 			power_src = smu_power_get_index(smu, SMU_POWER_SOURCE_AC);
 			if (power_src < 0)
@@ -1572,23 +1712,291 @@ static int navi10_get_power_limit(struct smu_context *smu,
 				pptable->SocketPowerLimitAc[PPT_THROTTLER_PPT0];
 		}
 
-		if (smu->od_enabled) {
-			asic_default_power_limit *= (100 + smu->smu_table.TDPODLimit);
-			asic_default_power_limit /= 100;
-		}
-
-		smu->default_power_limit = asic_default_power_limit;
 		smu->power_limit = asic_default_power_limit;
 	}
 
-	if (asic_default)
-		*limit = smu->default_power_limit;
+	if (cap)
+		*limit = smu_v11_0_get_max_power_limit(smu);
 	else
 		*limit = smu->power_limit;
 
 	return 0;
 }
 
+static int navi10_update_pcie_parameters(struct smu_context *smu,
+				     uint32_t pcie_gen_cap,
+				     uint32_t pcie_width_cap)
+{
+	PPTable_t *pptable = smu->smu_table.driver_pptable;
+	int ret, i;
+	uint32_t smu_pcie_arg;
+
+	struct smu_dpm_context *smu_dpm = &smu->smu_dpm;
+	struct smu_11_0_dpm_context *dpm_context = smu_dpm->dpm_context;
+
+	for (i = 0; i < NUM_LINK_LEVELS; i++) {
+		smu_pcie_arg = (i << 16) |
+			((pptable->PcieGenSpeed[i] <= pcie_gen_cap) ? (pptable->PcieGenSpeed[i] << 8) :
+				(pcie_gen_cap << 8)) | ((pptable->PcieLaneCount[i] <= pcie_width_cap) ?
+					pptable->PcieLaneCount[i] : pcie_width_cap);
+		ret = smu_send_smc_msg_with_param(smu,
+					  SMU_MSG_OverridePcieParameters,
+					  smu_pcie_arg);
+
+		if (ret)
+			return ret;
+
+		if (pptable->PcieGenSpeed[i] > pcie_gen_cap)
+			dpm_context->dpm_tables.pcie_table.pcie_gen[i] = pcie_gen_cap;
+		if (pptable->PcieLaneCount[i] > pcie_width_cap)
+			dpm_context->dpm_tables.pcie_table.pcie_lane[i] = pcie_width_cap;
+	}
+
+	return 0;
+}
+
+static inline void navi10_dump_od_table(OverDriveTable_t *od_table) {
+	pr_debug("OD: Gfxclk: (%d, %d)\n", od_table->GfxclkFmin, od_table->GfxclkFmax);
+	pr_debug("OD: Gfx1: (%d, %d)\n", od_table->GfxclkFreq1, od_table->GfxclkVolt1);
+	pr_debug("OD: Gfx2: (%d, %d)\n", od_table->GfxclkFreq2, od_table->GfxclkVolt2);
+	pr_debug("OD: Gfx3: (%d, %d)\n", od_table->GfxclkFreq3, od_table->GfxclkVolt3);
+	pr_debug("OD: UclkFmax: %d\n", od_table->UclkFmax);
+	pr_debug("OD: OverDrivePct: %d\n", od_table->OverDrivePct);
+}
+
+static int navi10_od_setting_check_range(struct smu_11_0_overdrive_table *od_table, enum SMU_11_0_ODSETTING_ID setting, uint32_t value)
+{
+	if (value < od_table->min[setting]) {
+		pr_warn("OD setting (%d, %d) is less than the minimum allowed (%d)\n", setting, value, od_table->min[setting]);
+		return -EINVAL;
+	}
+	if (value > od_table->max[setting]) {
+		pr_warn("OD setting (%d, %d) is greater than the maximum allowed (%d)\n", setting, value, od_table->max[setting]);
+		return -EINVAL;
+	}
+	return 0;
+}
+
+static int navi10_setup_od_limits(struct smu_context *smu) {
+	struct smu_11_0_overdrive_table *overdrive_table = NULL;
+	struct smu_11_0_powerplay_table *powerplay_table = NULL;
+
+	if (!smu->smu_table.power_play_table) {
+		pr_err("powerplay table uninitialized!\n");
+		return -ENOENT;
+	}
+	powerplay_table = (struct smu_11_0_powerplay_table *)smu->smu_table.power_play_table;
+	overdrive_table = &powerplay_table->overdrive_table;
+	if (!smu->od_settings) {
+		smu->od_settings = kmemdup(overdrive_table, sizeof(struct smu_11_0_overdrive_table), GFP_KERNEL);
+	} else {
+		memcpy(smu->od_settings, overdrive_table, sizeof(struct smu_11_0_overdrive_table));
+	}
+	return 0;
+}
+
+static int navi10_set_default_od_settings(struct smu_context *smu, bool initialize) {
+	OverDriveTable_t *od_table;
+	int ret = 0;
+
+	ret = smu_v11_0_set_default_od_settings(smu, initialize, sizeof(OverDriveTable_t));
+	if (ret)
+		return ret;
+
+	if (initialize) {
+		ret = navi10_setup_od_limits(smu);
+		if (ret) {
+			pr_err("Failed to retrieve board OD limits\n");
+			return ret;
+		}
+
+	}
+
+	od_table = (OverDriveTable_t *)smu->smu_table.overdrive_table;
+	if (od_table) {
+		navi10_dump_od_table(od_table);
+	}
+
+	return ret;
+}
+
+static int navi10_od_edit_dpm_table(struct smu_context *smu, enum PP_OD_DPM_TABLE_COMMAND type, long input[], uint32_t size) {
+	int i;
+	int ret = 0;
+	struct smu_table_context *table_context = &smu->smu_table;
+	OverDriveTable_t *od_table;
+	struct smu_11_0_overdrive_table *od_settings;
+	enum SMU_11_0_ODSETTING_ID freq_setting, voltage_setting;
+	uint16_t *freq_ptr, *voltage_ptr;
+	od_table = (OverDriveTable_t *)table_context->overdrive_table;
+
+	if (!smu->od_enabled) {
+		pr_warn("OverDrive is not enabled!\n");
+		return -EINVAL;
+	}
+
+	if (!smu->od_settings) {
+		pr_err("OD board limits are not set!\n");
+		return -ENOENT;
+	}
+
+	od_settings = smu->od_settings;
+
+	switch (type) {
+	case PP_OD_EDIT_SCLK_VDDC_TABLE:
+		if (!navi10_od_feature_is_supported(od_settings, SMU_11_0_ODFEATURE_GFXCLK_LIMITS)) {
+			pr_warn("GFXCLK_LIMITS not supported!\n");
+			return -ENOTSUPP;
+		}
+		if (!table_context->overdrive_table) {
+			pr_err("Overdrive is not initialized\n");
+			return -EINVAL;
+		}
+		for (i = 0; i < size; i += 2) {
+			if (i + 2 > size) {
+				pr_info("invalid number of input parameters %d\n", size);
+				return -EINVAL;
+			}
+			switch (input[i]) {
+			case 0:
+				freq_setting = SMU_11_0_ODSETTING_GFXCLKFMIN;
+				freq_ptr = &od_table->GfxclkFmin;
+				if (input[i + 1] > od_table->GfxclkFmax) {
+					pr_info("GfxclkFmin (%ld) must be <= GfxclkFmax (%u)!\n",
+						input[i + 1],
+						od_table->GfxclkFmin);
+					return -EINVAL;
+				}
+				break;
+			case 1:
+				freq_setting = SMU_11_0_ODSETTING_GFXCLKFMAX;
+				freq_ptr = &od_table->GfxclkFmax;
+				if (input[i + 1] < od_table->GfxclkFmin) {
+					pr_info("GfxclkFmax (%ld) must be >= GfxclkFmin (%u)!\n",
+						input[i + 1],
+						od_table->GfxclkFmax);
+					return -EINVAL;
+				}
+				break;
+			default:
+				pr_info("Invalid SCLK_VDDC_TABLE index: %ld\n", input[i]);
+				pr_info("Supported indices: [0:min,1:max]\n");
+				return -EINVAL;
+			}
+			ret = navi10_od_setting_check_range(od_settings, freq_setting, input[i + 1]);
+			if (ret)
+				return ret;
+			*freq_ptr = input[i + 1];
+		}
+		break;
+	case PP_OD_EDIT_MCLK_VDDC_TABLE:
+		if (!navi10_od_feature_is_supported(od_settings, SMU_11_0_ODFEATURE_UCLK_MAX)) {
+			pr_warn("UCLK_MAX not supported!\n");
+			return -ENOTSUPP;
+		}
+		if (size < 2) {
+			pr_info("invalid number of parameters: %d\n", size);
+			return -EINVAL;
+		}
+		if (input[0] != 1) {
+			pr_info("Invalid MCLK_VDDC_TABLE index: %ld\n", input[0]);
+			pr_info("Supported indices: [1:max]\n");
+			return -EINVAL;
+		}
+		ret = navi10_od_setting_check_range(od_settings, SMU_11_0_ODSETTING_UCLKFMAX, input[1]);
+		if (ret)
+			return ret;
+		od_table->UclkFmax = input[1];
+		break;
+	case PP_OD_COMMIT_DPM_TABLE:
+		navi10_dump_od_table(od_table);
+		ret = smu_update_table(smu, SMU_TABLE_OVERDRIVE, 0, (void *)od_table, true);
+		if (ret) {
+			pr_err("Failed to import overdrive table!\n");
+			return ret;
+		}
+		// no lock needed because smu_od_edit_dpm_table has it
+		ret = smu_handle_task(smu, smu->smu_dpm.dpm_level,
+			AMD_PP_TASK_READJUST_POWER_STATE,
+			false);
+		if (ret) {
+			return ret;
+		}
+		break;
+	case PP_OD_EDIT_VDDC_CURVE:
+		if (!navi10_od_feature_is_supported(od_settings, SMU_11_0_ODFEATURE_GFXCLK_CURVE)) {
+			pr_warn("GFXCLK_CURVE not supported!\n");
+			return -ENOTSUPP;
+		}
+		if (size < 3) {
+			pr_info("invalid number of parameters: %d\n", size);
+			return -EINVAL;
+		}
+		if (!od_table) {
+			pr_info("Overdrive is not initialized\n");
+			return -EINVAL;
+		}
+
+		switch (input[0]) {
+		case 0:
+			freq_setting = SMU_11_0_ODSETTING_VDDGFXCURVEFREQ_P1;
+			voltage_setting = SMU_11_0_ODSETTING_VDDGFXCURVEVOLTAGE_P1;
+			freq_ptr = &od_table->GfxclkFreq1;
+			voltage_ptr = &od_table->GfxclkVolt1;
+			break;
+		case 1:
+			freq_setting = SMU_11_0_ODSETTING_VDDGFXCURVEFREQ_P2;
+			voltage_setting = SMU_11_0_ODSETTING_VDDGFXCURVEVOLTAGE_P2;
+			freq_ptr = &od_table->GfxclkFreq2;
+			voltage_ptr = &od_table->GfxclkVolt2;
+			break;
+		case 2:
+			freq_setting = SMU_11_0_ODSETTING_VDDGFXCURVEFREQ_P3;
+			voltage_setting = SMU_11_0_ODSETTING_VDDGFXCURVEVOLTAGE_P3;
+			freq_ptr = &od_table->GfxclkFreq3;
+			voltage_ptr = &od_table->GfxclkVolt3;
+			break;
+		default:
+			pr_info("Invalid VDDC_CURVE index: %ld\n", input[0]);
+			pr_info("Supported indices: [0, 1, 2]\n");
+			return -EINVAL;
+		}
+		ret = navi10_od_setting_check_range(od_settings, freq_setting, input[1]);
+		if (ret)
+			return ret;
+		// Allow setting zero to disable the OverDrive VDDC curve
+		if (input[2] != 0) {
+			ret = navi10_od_setting_check_range(od_settings, voltage_setting, input[2]);
+			if (ret)
+				return ret;
+			*freq_ptr = input[1];
+			*voltage_ptr = ((uint16_t)input[2]) * NAVI10_VOLTAGE_SCALE;
+			pr_debug("OD: set curve %ld: (%d, %d)\n", input[0], *freq_ptr, *voltage_ptr);
+		} else {
+			// If setting 0, disable all voltage curve settings
+			od_table->GfxclkVolt1 = 0;
+			od_table->GfxclkVolt2 = 0;
+			od_table->GfxclkVolt3 = 0;
+		}
+		navi10_dump_od_table(od_table);
+		break;
+	default:
+		return -ENOSYS;
+	}
+	return ret;
+}
+
+static int navi10_run_btc(struct smu_context *smu)
+{
+	int ret = 0;
+
+	ret = smu_send_smc_msg(smu, SMU_MSG_RunBtc);
+	if (ret)
+		pr_err("RunBtc failed!\n");
+
+	return ret;
+}
+
 static const struct pptable_funcs navi10_ppt_funcs = {
 	.tables_init = navi10_tables_init,
 	.alloc_dpm_context = navi10_allocate_dpm_context,
@@ -1627,12 +2035,63 @@ static const struct pptable_funcs navi10_ppt_funcs = {
 	.get_thermal_temperature_range = navi10_get_thermal_temperature_range,
 	.display_disable_memory_clock_switch = navi10_display_disable_memory_clock_switch,
 	.get_power_limit = navi10_get_power_limit,
+	.update_pcie_parameters = navi10_update_pcie_parameters,
+	.init_microcode = smu_v11_0_init_microcode,
+	.load_microcode = smu_v11_0_load_microcode,
+	.init_smc_tables = smu_v11_0_init_smc_tables,
+	.fini_smc_tables = smu_v11_0_fini_smc_tables,
+	.init_power = smu_v11_0_init_power,
+	.fini_power = smu_v11_0_fini_power,
+	.check_fw_status = smu_v11_0_check_fw_status,
+	.setup_pptable = smu_v11_0_setup_pptable,
+	.get_vbios_bootup_values = smu_v11_0_get_vbios_bootup_values,
+	.get_clk_info_from_vbios = smu_v11_0_get_clk_info_from_vbios,
+	.check_pptable = smu_v11_0_check_pptable,
+	.parse_pptable = smu_v11_0_parse_pptable,
+	.populate_smc_tables = smu_v11_0_populate_smc_pptable,
+	.check_fw_version = smu_v11_0_check_fw_version,
+	.write_pptable = smu_v11_0_write_pptable,
+	.set_min_dcef_deep_sleep = smu_v11_0_set_min_dcef_deep_sleep,
+	.set_tool_table_location = smu_v11_0_set_tool_table_location,
+	.notify_memory_pool_location = smu_v11_0_notify_memory_pool_location,
+	.system_features_control = smu_v11_0_system_features_control,
+	.send_smc_msg = smu_v11_0_send_msg,
+	.send_smc_msg_with_param = smu_v11_0_send_msg_with_param,
+	.read_smc_arg = smu_v11_0_read_arg,
+	.init_display_count = smu_v11_0_init_display_count,
+	.set_allowed_mask = smu_v11_0_set_allowed_mask,
+	.get_enabled_mask = smu_v11_0_get_enabled_mask,
+	.notify_display_change = smu_v11_0_notify_display_change,
+	.set_power_limit = smu_v11_0_set_power_limit,
+	.get_current_clk_freq = smu_v11_0_get_current_clk_freq,
+	.init_max_sustainable_clocks = smu_v11_0_init_max_sustainable_clocks,
+	.start_thermal_control = smu_v11_0_start_thermal_control,
+	.stop_thermal_control = smu_v11_0_stop_thermal_control,
+	.set_deep_sleep_dcefclk = smu_v11_0_set_deep_sleep_dcefclk,
+	.display_clock_voltage_request = smu_v11_0_display_clock_voltage_request,
+	.get_fan_control_mode = smu_v11_0_get_fan_control_mode,
+	.set_fan_control_mode = smu_v11_0_set_fan_control_mode,
+	.set_fan_speed_percent = smu_v11_0_set_fan_speed_percent,
+	.set_fan_speed_rpm = smu_v11_0_set_fan_speed_rpm,
+	.set_xgmi_pstate = smu_v11_0_set_xgmi_pstate,
+	.gfx_off_control = smu_v11_0_gfx_off_control,
+	.register_irq_handler = smu_v11_0_register_irq_handler,
+	.set_azalia_d3_pme = smu_v11_0_set_azalia_d3_pme,
+	.get_max_sustainable_clocks_by_dc = smu_v11_0_get_max_sustainable_clocks_by_dc,
+	.baco_is_support= smu_v11_0_baco_is_support,
+	.baco_get_state = smu_v11_0_baco_get_state,
+	.baco_set_state = smu_v11_0_baco_set_state,
+	.baco_reset = smu_v11_0_baco_reset,
+	.get_dpm_ultimate_freq = smu_v11_0_get_dpm_ultimate_freq,
+	.set_soft_freq_limited_range = smu_v11_0_set_soft_freq_limited_range,
+	.override_pcie_parameters = smu_v11_0_override_pcie_parameters,
+	.set_default_od_settings = navi10_set_default_od_settings,
+	.od_edit_dpm_table = navi10_od_edit_dpm_table,
+	.get_pptable_power_limit = navi10_get_pptable_power_limit,
+	.run_btc = navi10_run_btc,
 };
 
 void navi10_set_ppt_funcs(struct smu_context *smu)
 {
-	struct smu_table_context *smu_table = &smu->smu_table;
-
 	smu->ppt_funcs = &navi10_ppt_funcs;
-	smu_table->table_count = TABLE_COUNT;
 }
diff --git a/drivers/gpu/drm/amd/powerplay/navi10_ppt.h b/drivers/gpu/drm/amd/powerplay/navi10_ppt.h
index 620ff17c2fef..ec03c7992f6d 100644
--- a/drivers/gpu/drm/amd/powerplay/navi10_ppt.h
+++ b/drivers/gpu/drm/amd/powerplay/navi10_ppt.h
@@ -27,6 +27,17 @@
 #define NAVI10_PEAK_SCLK_XT  		(1755)
 #define NAVI10_PEAK_SCLK_XL  		(1625)
 
+#define NAVI14_UMD_PSTATE_PEAK_XT_GFXCLK      (1670)
+#define NAVI14_UMD_PSTATE_PEAK_XTM_GFXCLK     (1448)
+#define NAVI14_UMD_PSTATE_PEAK_XLM_GFXCLK     (1181)
+#define NAVI14_UMD_PSTATE_PEAK_XTX_GFXCLK     (1717)
+#define NAVI14_UMD_PSTATE_PEAK_XL_GFXCLK      (1448)
+
+#define NAVI10_VOLTAGE_SCALE (4)
+
+#define smnPCIE_LC_SPEED_CNTL			0x11140290
+#define smnPCIE_LC_LINK_WIDTH_CNTL		0x11140288
+
 extern void navi10_set_ppt_funcs(struct smu_context *smu);
 
 #endif
diff --git a/drivers/gpu/drm/amd/powerplay/renoir_ppt.c b/drivers/gpu/drm/amd/powerplay/renoir_ppt.c
index e62bfba51562..04daf7e9fe05 100644
--- a/drivers/gpu/drm/amd/powerplay/renoir_ppt.c
+++ b/drivers/gpu/drm/amd/powerplay/renoir_ppt.c
@@ -23,6 +23,7 @@
 
 #include "amdgpu.h"
 #include "amdgpu_smu.h"
+#include "smu_internal.h"
 #include "soc15_common.h"
 #include "smu_v12_0_ppsmc.h"
 #include "smu12_driver_if.h"
@@ -160,21 +161,17 @@ static int renoir_tables_init(struct smu_context *smu, struct smu_table *tables)
  * This interface just for getting uclk ultimate freq and should't introduce
  * other likewise function result in overmuch callback.
  */
-static int renoir_get_dpm_uclk_limited(struct smu_context *smu, uint32_t *clock, bool max)
+static int renoir_get_dpm_clk_limited(struct smu_context *smu, enum smu_clk_type clk_type,
+						uint32_t dpm_level, uint32_t *freq)
 {
+	DpmClocks_t *clk_table = smu->smu_table.clocks_table;
 
-	DpmClocks_t *table = smu->smu_table.clocks_table;
-
-	if (!clock || !table)
+	if (!clk_table || clk_type >= SMU_CLK_COUNT)
 		return -EINVAL;
 
-	if (max)
-		*clock = table->FClocks[NUM_FCLK_DPM_LEVELS-1].Freq;
-	else
-		*clock = table->FClocks[0].Freq;
+	GET_DPM_CUR_FREQ(clk_table, clk_type, dpm_level, *freq);
 
 	return 0;
-
 }
 
 static int renoir_print_clk_levels(struct smu_context *smu,
@@ -183,11 +180,13 @@ static int renoir_print_clk_levels(struct smu_context *smu,
 	int i, size = 0, ret = 0;
 	uint32_t cur_value = 0, value = 0, count = 0, min = 0, max = 0;
 	DpmClocks_t *clk_table = smu->smu_table.clocks_table;
-	SmuMetrics_t metrics = {0};
+	SmuMetrics_t metrics;
 
 	if (!clk_table || clk_type >= SMU_CLK_COUNT)
 		return -EINVAL;
 
+	memset(&metrics, 0, sizeof(metrics));
+
 	ret = smu_update_table(smu, SMU_TABLE_SMU_METRICS, 0,
 			       (void *)&metrics, false);
 	if (ret)
@@ -198,7 +197,7 @@ static int renoir_print_clk_levels(struct smu_context *smu,
 	case SMU_SCLK:
 		/* retirve table returned paramters unit is MHz */
 		cur_value = metrics.ClockFrequency[CLOCK_GFXCLK];
-		ret = smu_get_dpm_freq_range(smu, SMU_GFXCLK, &min, &max);
+		ret = smu_get_dpm_freq_range(smu, SMU_GFXCLK, &min, &max, false);
 		if (!ret) {
 			/* driver only know min/max gfx_clk, Add level 1 for all other gfx clks */
 			if (cur_value  == max)
@@ -246,20 +245,474 @@ static int renoir_print_clk_levels(struct smu_context *smu,
 	return size;
 }
 
+static enum amd_pm_state_type renoir_get_current_power_state(struct smu_context *smu)
+{
+	enum amd_pm_state_type pm_type;
+	struct smu_dpm_context *smu_dpm_ctx = &(smu->smu_dpm);
+
+	if (!smu_dpm_ctx->dpm_context ||
+	    !smu_dpm_ctx->dpm_current_power_state)
+		return -EINVAL;
+
+	switch (smu_dpm_ctx->dpm_current_power_state->classification.ui_label) {
+	case SMU_STATE_UI_LABEL_BATTERY:
+		pm_type = POWER_STATE_TYPE_BATTERY;
+		break;
+	case SMU_STATE_UI_LABEL_BALLANCED:
+		pm_type = POWER_STATE_TYPE_BALANCED;
+		break;
+	case SMU_STATE_UI_LABEL_PERFORMANCE:
+		pm_type = POWER_STATE_TYPE_PERFORMANCE;
+		break;
+	default:
+		if (smu_dpm_ctx->dpm_current_power_state->classification.flags & SMU_STATE_CLASSIFICATION_FLAG_BOOT)
+			pm_type = POWER_STATE_TYPE_INTERNAL_BOOT;
+		else
+			pm_type = POWER_STATE_TYPE_DEFAULT;
+		break;
+	}
+
+	return pm_type;
+}
+
+static int renoir_dpm_set_uvd_enable(struct smu_context *smu, bool enable)
+{
+	struct smu_power_context *smu_power = &smu->smu_power;
+	struct smu_power_gate *power_gate = &smu_power->power_gate;
+	int ret = 0;
+
+	if (enable) {
+		/* vcn dpm on is a prerequisite for vcn power gate messages */
+		if (smu_feature_is_enabled(smu, SMU_FEATURE_VCN_PG_BIT)) {
+			ret = smu_send_smc_msg_with_param(smu, SMU_MSG_PowerUpVcn, 0);
+			if (ret)
+				return ret;
+		}
+		power_gate->vcn_gated = false;
+	} else {
+		if (smu_feature_is_enabled(smu, SMU_FEATURE_VCN_PG_BIT)) {
+			ret = smu_send_smc_msg(smu, SMU_MSG_PowerDownVcn);
+			if (ret)
+				return ret;
+		}
+		power_gate->vcn_gated = true;
+	}
+
+	return ret;
+}
+
+static int renoir_force_dpm_limit_value(struct smu_context *smu, bool highest)
+{
+	int ret = 0, i = 0;
+	uint32_t min_freq, max_freq, force_freq;
+	enum smu_clk_type clk_type;
+
+	enum smu_clk_type clks[] = {
+		SMU_GFXCLK,
+		SMU_MCLK,
+		SMU_SOCCLK,
+	};
+
+	for (i = 0; i < ARRAY_SIZE(clks); i++) {
+		clk_type = clks[i];
+		ret = smu_get_dpm_freq_range(smu, clk_type, &min_freq, &max_freq, false);
+		if (ret)
+			return ret;
+
+		force_freq = highest ? max_freq : min_freq;
+		ret = smu_set_soft_freq_range(smu, clk_type, force_freq, force_freq);
+		if (ret)
+			return ret;
+	}
+
+	return ret;
+}
+
+static int renoir_unforce_dpm_levels(struct smu_context *smu) {
+
+	int ret = 0, i = 0;
+	uint32_t min_freq, max_freq;
+	enum smu_clk_type clk_type;
+
+	struct clk_feature_map {
+		enum smu_clk_type clk_type;
+		uint32_t	feature;
+	} clk_feature_map[] = {
+		{SMU_GFXCLK, SMU_FEATURE_DPM_GFXCLK_BIT},
+		{SMU_MCLK,   SMU_FEATURE_DPM_UCLK_BIT},
+		{SMU_SOCCLK, SMU_FEATURE_DPM_SOCCLK_BIT},
+	};
+
+	for (i = 0; i < ARRAY_SIZE(clk_feature_map); i++) {
+		if (!smu_feature_is_enabled(smu, clk_feature_map[i].feature))
+		    continue;
+
+		clk_type = clk_feature_map[i].clk_type;
+
+		ret = smu_get_dpm_freq_range(smu, clk_type, &min_freq, &max_freq, false);
+		if (ret)
+			return ret;
+
+		ret = smu_set_soft_freq_range(smu, clk_type, min_freq, max_freq);
+		if (ret)
+			return ret;
+	}
+
+	return ret;
+}
+
+static int renoir_get_workload_type(struct smu_context *smu, uint32_t profile)
+{
+
+	uint32_t  pplib_workload = 0;
+
+	switch (profile) {
+	case PP_SMC_POWER_PROFILE_FULLSCREEN3D:
+		pplib_workload = WORKLOAD_PPLIB_FULL_SCREEN_3D_BIT;
+		break;
+	case PP_SMC_POWER_PROFILE_CUSTOM:
+		pplib_workload = WORKLOAD_PPLIB_COUNT;
+		break;
+	case PP_SMC_POWER_PROFILE_VIDEO:
+		pplib_workload = WORKLOAD_PPLIB_VIDEO_BIT;
+		break;
+	case PP_SMC_POWER_PROFILE_VR:
+		pplib_workload = WORKLOAD_PPLIB_VR_BIT;
+		break;
+	case PP_SMC_POWER_PROFILE_COMPUTE:
+		pplib_workload = WORKLOAD_PPLIB_COMPUTE_BIT;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	return pplib_workload;
+}
+
+static int renoir_get_profiling_clk_mask(struct smu_context *smu,
+					 enum amd_dpm_forced_level level,
+					 uint32_t *sclk_mask,
+					 uint32_t *mclk_mask,
+					 uint32_t *soc_mask)
+{
+
+	if (level == AMD_DPM_FORCED_LEVEL_PROFILE_MIN_SCLK) {
+		if (sclk_mask)
+			*sclk_mask = 0;
+	} else if (level == AMD_DPM_FORCED_LEVEL_PROFILE_MIN_MCLK) {
+		if (mclk_mask)
+			*mclk_mask = 0;
+	} else if (level == AMD_DPM_FORCED_LEVEL_PROFILE_PEAK) {
+		if(sclk_mask)
+			/* The sclk as gfxclk and has three level about max/min/current */
+			*sclk_mask = 3 - 1;
+
+		if(mclk_mask)
+			*mclk_mask = NUM_MEMCLK_DPM_LEVELS - 1;
+
+		if(soc_mask)
+			*soc_mask = NUM_SOCCLK_DPM_LEVELS - 1;
+	}
+
+	return 0;
+}
+
+/**
+ * This interface get dpm clock table for dc
+ */
+static int renoir_get_dpm_clock_table(struct smu_context *smu, struct dpm_clocks *clock_table)
+{
+	DpmClocks_t *table = smu->smu_table.clocks_table;
+	int i;
+
+	if (!clock_table || !table)
+		return -EINVAL;
+
+	for (i = 0; i < NUM_DCFCLK_DPM_LEVELS; i++) {
+		clock_table->DcfClocks[i].Freq = table->DcfClocks[i].Freq;
+		clock_table->DcfClocks[i].Vol = table->DcfClocks[i].Vol;
+	}
+
+	for (i = 0; i < NUM_SOCCLK_DPM_LEVELS; i++) {
+		clock_table->SocClocks[i].Freq = table->SocClocks[i].Freq;
+		clock_table->SocClocks[i].Vol = table->SocClocks[i].Vol;
+	}
+
+	for (i = 0; i < NUM_FCLK_DPM_LEVELS; i++) {
+		clock_table->FClocks[i].Freq = table->FClocks[i].Freq;
+		clock_table->FClocks[i].Vol = table->FClocks[i].Vol;
+	}
+
+	for (i = 0; i<  NUM_MEMCLK_DPM_LEVELS; i++) {
+		clock_table->MemClocks[i].Freq = table->MemClocks[i].Freq;
+		clock_table->MemClocks[i].Vol = table->MemClocks[i].Vol;
+	}
+
+	return 0;
+}
+
+static int renoir_force_clk_levels(struct smu_context *smu,
+				   enum smu_clk_type clk_type, uint32_t mask)
+{
+
+	int ret = 0 ;
+	uint32_t soft_min_level = 0, soft_max_level = 0, min_freq = 0, max_freq = 0;
+	DpmClocks_t *clk_table = smu->smu_table.clocks_table;
+
+	soft_min_level = mask ? (ffs(mask) - 1) : 0;
+	soft_max_level = mask ? (fls(mask) - 1) : 0;
+
+	switch (clk_type) {
+	case SMU_GFXCLK:
+	case SMU_SCLK:
+		if (soft_min_level > 2 || soft_max_level > 2) {
+			pr_info("Currently sclk only support 3 levels on APU\n");
+			return -EINVAL;
+		}
+
+		ret = smu_get_dpm_freq_range(smu, SMU_GFXCLK, &min_freq, &max_freq, false);
+		if (ret)
+			return ret;
+		ret = smu_send_smc_msg_with_param(smu, SMU_MSG_SetSoftMaxGfxClk,
+					soft_max_level == 0 ? min_freq :
+					soft_max_level == 1 ? RENOIR_UMD_PSTATE_GFXCLK : max_freq);
+		if (ret)
+			return ret;
+		ret = smu_send_smc_msg_with_param(smu, SMU_MSG_SetHardMinGfxClk,
+					soft_min_level == 2 ? max_freq :
+					soft_min_level == 1 ? RENOIR_UMD_PSTATE_GFXCLK : min_freq);
+		if (ret)
+			return ret;
+		break;
+	case SMU_SOCCLK:
+		GET_DPM_CUR_FREQ(clk_table, clk_type, soft_min_level, min_freq);
+		GET_DPM_CUR_FREQ(clk_table, clk_type, soft_max_level, max_freq);
+		ret = smu_send_smc_msg_with_param(smu, SMU_MSG_SetSoftMaxSocclkByFreq, max_freq);
+		if (ret)
+			return ret;
+		ret = smu_send_smc_msg_with_param(smu, SMU_MSG_SetHardMinSocclkByFreq, min_freq);
+		if (ret)
+			return ret;
+		break;
+	case SMU_MCLK:
+	case SMU_FCLK:
+		GET_DPM_CUR_FREQ(clk_table, clk_type, soft_min_level, min_freq);
+		GET_DPM_CUR_FREQ(clk_table, clk_type, soft_max_level, max_freq);
+		ret = smu_send_smc_msg_with_param(smu, SMU_MSG_SetSoftMaxFclkByFreq, max_freq);
+		if (ret)
+			return ret;
+		ret = smu_send_smc_msg_with_param(smu, SMU_MSG_SetHardMinFclkByFreq, min_freq);
+		if (ret)
+			return ret;
+		break;
+	default:
+		break;
+	}
+
+	return ret;
+}
+
+static int renoir_set_power_profile_mode(struct smu_context *smu, long *input, uint32_t size)
+{
+	int workload_type, ret;
+	uint32_t profile_mode = input[size];
+
+	if (profile_mode > PP_SMC_POWER_PROFILE_CUSTOM) {
+		pr_err("Invalid power profile mode %d\n", smu->power_profile_mode);
+		return -EINVAL;
+	}
+
+	/* conv PP_SMC_POWER_PROFILE* to WORKLOAD_PPLIB_*_BIT */
+	workload_type = smu_workload_get_type(smu, smu->power_profile_mode);
+	if (workload_type < 0) {
+		pr_err("Unsupported power profile mode %d on RENOIR\n",smu->power_profile_mode);
+		return -EINVAL;
+	}
+
+	ret = smu_send_smc_msg_with_param(smu, SMU_MSG_SetWorkloadMask,
+				    1 << workload_type);
+	if (ret) {
+		pr_err("Fail to set workload type %d\n", workload_type);
+		return ret;
+	}
+
+	smu->power_profile_mode = profile_mode;
+
+	return 0;
+}
+
+static int renoir_set_peak_clock_by_device(struct smu_context *smu)
+{
+	int ret = 0;
+	uint32_t sclk_freq = 0, uclk_freq = 0;
+
+	ret = smu_get_dpm_freq_range(smu, SMU_SCLK, NULL, &sclk_freq, false);
+	if (ret)
+		return ret;
+
+	ret = smu_set_soft_freq_range(smu, SMU_SCLK, sclk_freq, sclk_freq);
+	if (ret)
+		return ret;
+
+	ret = smu_get_dpm_freq_range(smu, SMU_UCLK, NULL, &uclk_freq, false);
+	if (ret)
+		return ret;
+
+	ret = smu_set_soft_freq_range(smu, SMU_UCLK, uclk_freq, uclk_freq);
+	if (ret)
+		return ret;
+
+	return ret;
+}
+
+static int renoir_set_performance_level(struct smu_context *smu, enum amd_dpm_forced_level level)
+{
+	int ret = 0;
+
+	switch (level) {
+	case AMD_DPM_FORCED_LEVEL_PROFILE_PEAK:
+		ret = renoir_set_peak_clock_by_device(smu);
+		break;
+	default:
+		ret = -EINVAL;
+		break;
+	}
+
+	return ret;
+}
+
+/* save watermark settings into pplib smu structure,
+ * also pass data to smu controller
+ */
+static int renoir_set_watermarks_table(
+		struct smu_context *smu,
+		void *watermarks,
+		struct dm_pp_wm_sets_with_clock_ranges_soc15 *clock_ranges)
+{
+	int i;
+	int ret = 0;
+	Watermarks_t *table = watermarks;
+
+	if (!table || !clock_ranges)
+		return -EINVAL;
+
+	if (clock_ranges->num_wm_dmif_sets > 4 ||
+			clock_ranges->num_wm_mcif_sets > 4)
+		return -EINVAL;
+
+	/* save into smu->smu_table.tables[SMU_TABLE_WATERMARKS]->cpu_addr*/
+	for (i = 0; i < clock_ranges->num_wm_dmif_sets; i++) {
+		table->WatermarkRow[WM_DCFCLK][i].MinClock =
+			cpu_to_le16((uint16_t)
+			(clock_ranges->wm_dmif_clocks_ranges[i].wm_min_dcfclk_clk_in_khz));
+		table->WatermarkRow[WM_DCFCLK][i].MaxClock =
+			cpu_to_le16((uint16_t)
+			(clock_ranges->wm_dmif_clocks_ranges[i].wm_max_dcfclk_clk_in_khz));
+		table->WatermarkRow[WM_DCFCLK][i].MinMclk =
+			cpu_to_le16((uint16_t)
+			(clock_ranges->wm_dmif_clocks_ranges[i].wm_min_mem_clk_in_khz));
+		table->WatermarkRow[WM_DCFCLK][i].MaxMclk =
+			cpu_to_le16((uint16_t)
+			(clock_ranges->wm_dmif_clocks_ranges[i].wm_max_mem_clk_in_khz));
+		table->WatermarkRow[WM_DCFCLK][i].WmSetting = (uint8_t)
+				clock_ranges->wm_dmif_clocks_ranges[i].wm_set_id;
+	}
+
+	for (i = 0; i < clock_ranges->num_wm_mcif_sets; i++) {
+		table->WatermarkRow[WM_SOCCLK][i].MinClock =
+			cpu_to_le16((uint16_t)
+			(clock_ranges->wm_mcif_clocks_ranges[i].wm_min_socclk_clk_in_khz));
+		table->WatermarkRow[WM_SOCCLK][i].MaxClock =
+			cpu_to_le16((uint16_t)
+			(clock_ranges->wm_mcif_clocks_ranges[i].wm_max_socclk_clk_in_khz));
+		table->WatermarkRow[WM_SOCCLK][i].MinMclk =
+			cpu_to_le16((uint16_t)
+			(clock_ranges->wm_mcif_clocks_ranges[i].wm_min_mem_clk_in_khz));
+		table->WatermarkRow[WM_SOCCLK][i].MaxMclk =
+			cpu_to_le16((uint16_t)
+			(clock_ranges->wm_mcif_clocks_ranges[i].wm_max_mem_clk_in_khz));
+		table->WatermarkRow[WM_SOCCLK][i].WmSetting = (uint8_t)
+				clock_ranges->wm_mcif_clocks_ranges[i].wm_set_id;
+	}
+
+	/* pass data to smu controller */
+	ret = smu_write_watermarks_table(smu);
+
+	return ret;
+}
+
+static int renoir_get_power_profile_mode(struct smu_context *smu,
+					   char *buf)
+{
+	static const char *profile_name[] = {
+					"BOOTUP_DEFAULT",
+					"3D_FULL_SCREEN",
+					"POWER_SAVING",
+					"VIDEO",
+					"VR",
+					"COMPUTE",
+					"CUSTOM"};
+	uint32_t i, size = 0;
+	int16_t workload_type = 0;
+
+	if (!smu->pm_enabled || !buf)
+		return -EINVAL;
+
+	for (i = 0; i <= PP_SMC_POWER_PROFILE_CUSTOM; i++) {
+		/*
+		 * Conv PP_SMC_POWER_PROFILE* to WORKLOAD_PPLIB_*_BIT
+		 * Not all profile modes are supported on arcturus.
+		 */
+		workload_type = smu_workload_get_type(smu, i);
+		if (workload_type < 0)
+			continue;
+
+		size += sprintf(buf + size, "%2d %14s%s\n",
+			i, profile_name[i], (i == smu->power_profile_mode) ? "*" : " ");
+	}
+
+	return size;
+}
+
 static const struct pptable_funcs renoir_ppt_funcs = {
 	.get_smu_msg_index = renoir_get_smu_msg_index,
 	.get_smu_table_index = renoir_get_smu_table_index,
 	.tables_init = renoir_tables_init,
 	.set_power_state = NULL,
-	.get_dpm_uclk_limited = renoir_get_dpm_uclk_limited,
+	.get_dpm_clk_limited = renoir_get_dpm_clk_limited,
 	.print_clk_levels = renoir_print_clk_levels,
+	.get_current_power_state = renoir_get_current_power_state,
+	.dpm_set_uvd_enable = renoir_dpm_set_uvd_enable,
+	.force_dpm_limit_value = renoir_force_dpm_limit_value,
+	.unforce_dpm_levels = renoir_unforce_dpm_levels,
+	.get_workload_type = renoir_get_workload_type,
+	.get_profiling_clk_mask = renoir_get_profiling_clk_mask,
+	.force_clk_levels = renoir_force_clk_levels,
+	.set_power_profile_mode = renoir_set_power_profile_mode,
+	.set_performance_level = renoir_set_performance_level,
+	.get_dpm_clock_table = renoir_get_dpm_clock_table,
+	.set_watermarks_table = renoir_set_watermarks_table,
+	.get_power_profile_mode = renoir_get_power_profile_mode,
+	.check_fw_status = smu_v12_0_check_fw_status,
+	.check_fw_version = smu_v12_0_check_fw_version,
+	.powergate_sdma = smu_v12_0_powergate_sdma,
+	.powergate_vcn = smu_v12_0_powergate_vcn,
+	.send_smc_msg = smu_v12_0_send_msg,
+	.send_smc_msg_with_param = smu_v12_0_send_msg_with_param,
+	.read_smc_arg = smu_v12_0_read_arg,
+	.set_gfx_cgpg = smu_v12_0_set_gfx_cgpg,
+	.gfx_off_control = smu_v12_0_gfx_off_control,
+	.init_smc_tables = smu_v12_0_init_smc_tables,
+	.fini_smc_tables = smu_v12_0_fini_smc_tables,
+	.populate_smc_tables = smu_v12_0_populate_smc_tables,
+	.get_dpm_ultimate_freq = smu_v12_0_get_dpm_ultimate_freq,
+	.mode2_reset = smu_v12_0_mode2_reset,
+	.set_soft_freq_limited_range = smu_v12_0_set_soft_freq_limited_range,
 };
 
 void renoir_set_ppt_funcs(struct smu_context *smu)
 {
-	struct smu_table_context *smu_table = &smu->smu_table;
-
 	smu->ppt_funcs = &renoir_ppt_funcs;
 	smu->smc_if_version = SMU12_DRIVER_IF_VERSION;
-	smu_table->table_count = TABLE_COUNT;
+	smu->is_apu = true;
 }
diff --git a/drivers/gpu/drm/amd/powerplay/smu_internal.h b/drivers/gpu/drm/amd/powerplay/smu_internal.h
new file mode 100644
index 000000000000..8bcda7871309
--- /dev/null
+++ b/drivers/gpu/drm/amd/powerplay/smu_internal.h
@@ -0,0 +1,204 @@
+/*
+ * Copyright 2019 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+
+#ifndef __SMU_INTERNAL_H__
+#define __SMU_INTERNAL_H__
+
+#include "amdgpu_smu.h"
+
+#define smu_init_microcode(smu) \
+	((smu)->ppt_funcs->init_microcode ? (smu)->ppt_funcs->init_microcode((smu)) : 0)
+#define smu_init_smc_tables(smu) \
+	((smu)->ppt_funcs->init_smc_tables ? (smu)->ppt_funcs->init_smc_tables((smu)) : 0)
+#define smu_fini_smc_tables(smu) \
+	((smu)->ppt_funcs->fini_smc_tables ? (smu)->ppt_funcs->fini_smc_tables((smu)) : 0)
+#define smu_init_power(smu) \
+	((smu)->ppt_funcs->init_power ? (smu)->ppt_funcs->init_power((smu)) : 0)
+#define smu_fini_power(smu) \
+	((smu)->ppt_funcs->fini_power ? (smu)->ppt_funcs->fini_power((smu)) : 0)
+
+#define smu_setup_pptable(smu) \
+	((smu)->ppt_funcs->setup_pptable ? (smu)->ppt_funcs->setup_pptable((smu)) : 0)
+#define smu_powergate_sdma(smu, gate) \
+	((smu)->ppt_funcs->powergate_sdma ? (smu)->ppt_funcs->powergate_sdma((smu), (gate)) : 0)
+#define smu_powergate_vcn(smu, gate) \
+	((smu)->ppt_funcs->powergate_vcn ? (smu)->ppt_funcs->powergate_vcn((smu), (gate)) : 0)
+
+#define smu_get_vbios_bootup_values(smu) \
+	((smu)->ppt_funcs->get_vbios_bootup_values ? (smu)->ppt_funcs->get_vbios_bootup_values((smu)) : 0)
+#define smu_get_clk_info_from_vbios(smu) \
+	((smu)->ppt_funcs->get_clk_info_from_vbios ? (smu)->ppt_funcs->get_clk_info_from_vbios((smu)) : 0)
+#define smu_check_pptable(smu) \
+	((smu)->ppt_funcs->check_pptable ? (smu)->ppt_funcs->check_pptable((smu)) : 0)
+#define smu_parse_pptable(smu) \
+	((smu)->ppt_funcs->parse_pptable ? (smu)->ppt_funcs->parse_pptable((smu)) : 0)
+#define smu_populate_smc_tables(smu) \
+	((smu)->ppt_funcs->populate_smc_tables ? (smu)->ppt_funcs->populate_smc_tables((smu)) : 0)
+#define smu_check_fw_version(smu) \
+	((smu)->ppt_funcs->check_fw_version ? (smu)->ppt_funcs->check_fw_version((smu)) : 0)
+#define smu_write_pptable(smu) \
+	((smu)->ppt_funcs->write_pptable ? (smu)->ppt_funcs->write_pptable((smu)) : 0)
+#define smu_set_min_dcef_deep_sleep(smu) \
+	((smu)->ppt_funcs->set_min_dcef_deep_sleep ? (smu)->ppt_funcs->set_min_dcef_deep_sleep((smu)) : 0)
+#define smu_set_tool_table_location(smu) \
+	((smu)->ppt_funcs->set_tool_table_location ? (smu)->ppt_funcs->set_tool_table_location((smu)) : 0)
+#define smu_notify_memory_pool_location(smu) \
+	((smu)->ppt_funcs->notify_memory_pool_location ? (smu)->ppt_funcs->notify_memory_pool_location((smu)) : 0)
+#define smu_gfx_off_control(smu, enable) \
+	((smu)->ppt_funcs->gfx_off_control ? (smu)->ppt_funcs->gfx_off_control((smu), (enable)) : 0)
+
+#define smu_set_last_dcef_min_deep_sleep_clk(smu) \
+	((smu)->ppt_funcs->set_last_dcef_min_deep_sleep_clk ? (smu)->ppt_funcs->set_last_dcef_min_deep_sleep_clk((smu)) : 0)
+#define smu_system_features_control(smu, en) \
+	((smu)->ppt_funcs->system_features_control ? (smu)->ppt_funcs->system_features_control((smu), (en)) : 0)
+#define smu_init_max_sustainable_clocks(smu) \
+	((smu)->ppt_funcs->init_max_sustainable_clocks ? (smu)->ppt_funcs->init_max_sustainable_clocks((smu)) : 0)
+#define smu_set_default_od_settings(smu, initialize) \
+	((smu)->ppt_funcs->set_default_od_settings ? (smu)->ppt_funcs->set_default_od_settings((smu), (initialize)) : 0)
+
+#define smu_send_smc_msg(smu, msg) \
+	((smu)->ppt_funcs->send_smc_msg? (smu)->ppt_funcs->send_smc_msg((smu), (msg)) : 0)
+#define smu_send_smc_msg_with_param(smu, msg, param) \
+	((smu)->ppt_funcs->send_smc_msg_with_param? (smu)->ppt_funcs->send_smc_msg_with_param((smu), (msg), (param)) : 0)
+#define smu_read_smc_arg(smu, arg) \
+	((smu)->ppt_funcs->read_smc_arg? (smu)->ppt_funcs->read_smc_arg((smu), (arg)) : 0)
+#define smu_alloc_dpm_context(smu) \
+	((smu)->ppt_funcs->alloc_dpm_context ? (smu)->ppt_funcs->alloc_dpm_context((smu)) : 0)
+#define smu_init_display_count(smu, count) \
+	((smu)->ppt_funcs->init_display_count ? (smu)->ppt_funcs->init_display_count((smu), (count)) : 0)
+#define smu_feature_set_allowed_mask(smu) \
+	((smu)->ppt_funcs->set_allowed_mask? (smu)->ppt_funcs->set_allowed_mask((smu)) : 0)
+#define smu_feature_get_enabled_mask(smu, mask, num) \
+	((smu)->ppt_funcs->get_enabled_mask? (smu)->ppt_funcs->get_enabled_mask((smu), (mask), (num)) : 0)
+#define smu_is_dpm_running(smu) \
+	((smu)->ppt_funcs->is_dpm_running ? (smu)->ppt_funcs->is_dpm_running((smu)) : 0)
+#define smu_notify_display_change(smu) \
+	((smu)->ppt_funcs->notify_display_change? (smu)->ppt_funcs->notify_display_change((smu)) : 0)
+#define smu_store_powerplay_table(smu) \
+	((smu)->ppt_funcs->store_powerplay_table ? (smu)->ppt_funcs->store_powerplay_table((smu)) : 0)
+#define smu_check_powerplay_table(smu) \
+	((smu)->ppt_funcs->check_powerplay_table ? (smu)->ppt_funcs->check_powerplay_table((smu)) : 0)
+#define smu_append_powerplay_table(smu) \
+	((smu)->ppt_funcs->append_powerplay_table ? (smu)->ppt_funcs->append_powerplay_table((smu)) : 0)
+#define smu_set_default_dpm_table(smu) \
+	((smu)->ppt_funcs->set_default_dpm_table ? (smu)->ppt_funcs->set_default_dpm_table((smu)) : 0)
+#define smu_populate_umd_state_clk(smu) \
+	((smu)->ppt_funcs->populate_umd_state_clk ? (smu)->ppt_funcs->populate_umd_state_clk((smu)) : 0)
+#define smu_set_default_od8_settings(smu) \
+	((smu)->ppt_funcs->set_default_od8_settings ? (smu)->ppt_funcs->set_default_od8_settings((smu)) : 0)
+
+#define smu_get_current_clk_freq(smu, clk_id, value) \
+	((smu)->ppt_funcs->get_current_clk_freq? (smu)->ppt_funcs->get_current_clk_freq((smu), (clk_id), (value)) : 0)
+
+#define smu_tables_init(smu, tab) \
+	((smu)->ppt_funcs->tables_init ? (smu)->ppt_funcs->tables_init((smu), (tab)) : 0)
+#define smu_set_thermal_fan_table(smu) \
+	((smu)->ppt_funcs->set_thermal_fan_table ? (smu)->ppt_funcs->set_thermal_fan_table((smu)) : 0)
+#define smu_start_thermal_control(smu) \
+	((smu)->ppt_funcs->start_thermal_control? (smu)->ppt_funcs->start_thermal_control((smu)) : 0)
+#define smu_stop_thermal_control(smu) \
+	((smu)->ppt_funcs->stop_thermal_control? (smu)->ppt_funcs->stop_thermal_control((smu)) : 0)
+
+#define smu_smc_read_sensor(smu, sensor, data, size) \
+	((smu)->ppt_funcs->read_sensor? (smu)->ppt_funcs->read_sensor((smu), (sensor), (data), (size)) : -EINVAL)
+
+#define smu_pre_display_config_changed(smu) \
+	((smu)->ppt_funcs->pre_display_config_changed ? (smu)->ppt_funcs->pre_display_config_changed((smu)) : 0)
+#define smu_display_config_changed(smu) \
+	((smu)->ppt_funcs->display_config_changed ? (smu)->ppt_funcs->display_config_changed((smu)) : 0)
+#define smu_apply_clocks_adjust_rules(smu) \
+	((smu)->ppt_funcs->apply_clocks_adjust_rules ? (smu)->ppt_funcs->apply_clocks_adjust_rules((smu)) : 0)
+#define smu_notify_smc_dispaly_config(smu) \
+	((smu)->ppt_funcs->notify_smc_dispaly_config ? (smu)->ppt_funcs->notify_smc_dispaly_config((smu)) : 0)
+#define smu_force_dpm_limit_value(smu, highest) \
+	((smu)->ppt_funcs->force_dpm_limit_value ? (smu)->ppt_funcs->force_dpm_limit_value((smu), (highest)) : 0)
+#define smu_unforce_dpm_levels(smu) \
+	((smu)->ppt_funcs->unforce_dpm_levels ? (smu)->ppt_funcs->unforce_dpm_levels((smu)) : 0)
+#define smu_get_profiling_clk_mask(smu, level, sclk_mask, mclk_mask, soc_mask) \
+	((smu)->ppt_funcs->get_profiling_clk_mask ? (smu)->ppt_funcs->get_profiling_clk_mask((smu), (level), (sclk_mask), (mclk_mask), (soc_mask)) : 0)
+#define smu_set_cpu_power_state(smu) \
+	((smu)->ppt_funcs->set_cpu_power_state ? (smu)->ppt_funcs->set_cpu_power_state((smu)) : 0)
+
+#define smu_msg_get_index(smu, msg) \
+	((smu)->ppt_funcs? ((smu)->ppt_funcs->get_smu_msg_index? (smu)->ppt_funcs->get_smu_msg_index((smu), (msg)) : -EINVAL) : -EINVAL)
+#define smu_clk_get_index(smu, msg) \
+	((smu)->ppt_funcs? ((smu)->ppt_funcs->get_smu_clk_index? (smu)->ppt_funcs->get_smu_clk_index((smu), (msg)) : -EINVAL) : -EINVAL)
+#define smu_feature_get_index(smu, msg) \
+	((smu)->ppt_funcs? ((smu)->ppt_funcs->get_smu_feature_index? (smu)->ppt_funcs->get_smu_feature_index((smu), (msg)) : -EINVAL) : -EINVAL)
+#define smu_table_get_index(smu, tab) \
+	((smu)->ppt_funcs? ((smu)->ppt_funcs->get_smu_table_index? (smu)->ppt_funcs->get_smu_table_index((smu), (tab)) : -EINVAL) : -EINVAL)
+#define smu_power_get_index(smu, src) \
+	((smu)->ppt_funcs? ((smu)->ppt_funcs->get_smu_power_index? (smu)->ppt_funcs->get_smu_power_index((smu), (src)) : -EINVAL) : -EINVAL)
+#define smu_workload_get_type(smu, profile) \
+	((smu)->ppt_funcs? ((smu)->ppt_funcs->get_workload_type? (smu)->ppt_funcs->get_workload_type((smu), (profile)) : -EINVAL) : -EINVAL)
+#define smu_run_btc(smu) \
+	((smu)->ppt_funcs? ((smu)->ppt_funcs->run_btc? (smu)->ppt_funcs->run_btc((smu)) : 0) : 0)
+#define smu_get_allowed_feature_mask(smu, feature_mask, num) \
+	((smu)->ppt_funcs? ((smu)->ppt_funcs->get_allowed_feature_mask? (smu)->ppt_funcs->get_allowed_feature_mask((smu), (feature_mask), (num)) : 0) : 0)
+
+
+#define smu_store_cc6_data(smu, st, cc6_dis, pst_dis, pst_sw_dis) \
+	((smu)->ppt_funcs->store_cc6_data ? (smu)->ppt_funcs->store_cc6_data((smu), (st), (cc6_dis), (pst_dis), (pst_sw_dis)) : 0)
+
+#define smu_get_dal_power_level(smu, clocks) \
+	((smu)->ppt_funcs->get_dal_power_level ? (smu)->ppt_funcs->get_dal_power_level((smu), (clocks)) : 0)
+#define smu_get_perf_level(smu, designation, level) \
+	((smu)->ppt_funcs->get_perf_level ? (smu)->ppt_funcs->get_perf_level((smu), (designation), (level)) : 0)
+#define smu_get_current_shallow_sleep_clocks(smu, clocks) \
+	((smu)->ppt_funcs->get_current_shallow_sleep_clocks ? (smu)->ppt_funcs->get_current_shallow_sleep_clocks((smu), (clocks)) : 0)
+
+#define smu_dpm_set_uvd_enable(smu, enable) \
+	((smu)->ppt_funcs->dpm_set_uvd_enable ? (smu)->ppt_funcs->dpm_set_uvd_enable((smu), (enable)) : 0)
+#define smu_dpm_set_vce_enable(smu, enable) \
+	((smu)->ppt_funcs->dpm_set_vce_enable ? (smu)->ppt_funcs->dpm_set_vce_enable((smu), (enable)) : 0)
+
+#define smu_set_watermarks_table(smu, tab, clock_ranges) \
+	((smu)->ppt_funcs->set_watermarks_table ? (smu)->ppt_funcs->set_watermarks_table((smu), (tab), (clock_ranges)) : 0)
+#define smu_get_current_clk_freq_by_table(smu, clk_type, value) \
+	((smu)->ppt_funcs->get_current_clk_freq_by_table ? (smu)->ppt_funcs->get_current_clk_freq_by_table((smu), (clk_type), (value)) : 0)
+#define smu_thermal_temperature_range_update(smu, range, rw) \
+	((smu)->ppt_funcs->thermal_temperature_range_update? (smu)->ppt_funcs->thermal_temperature_range_update((smu), (range), (rw)) : 0)
+#define smu_get_thermal_temperature_range(smu, range) \
+	((smu)->ppt_funcs->get_thermal_temperature_range? (smu)->ppt_funcs->get_thermal_temperature_range((smu), (range)) : 0)
+#define smu_register_irq_handler(smu) \
+	((smu)->ppt_funcs->register_irq_handler ? (smu)->ppt_funcs->register_irq_handler(smu) : 0)
+
+#define smu_get_dpm_ultimate_freq(smu, param, min, max) \
+		((smu)->ppt_funcs->get_dpm_ultimate_freq ? (smu)->ppt_funcs->get_dpm_ultimate_freq((smu), (param), (min), (max)) : 0)
+
+#define smu_asic_set_performance_level(smu, level) \
+	((smu)->ppt_funcs->set_performance_level? (smu)->ppt_funcs->set_performance_level((smu), (level)) : -EINVAL);
+#define smu_dump_pptable(smu) \
+	((smu)->ppt_funcs->dump_pptable ? (smu)->ppt_funcs->dump_pptable((smu)) : 0)
+#define smu_get_dpm_clk_limited(smu, clk_type, dpm_level, freq) \
+		((smu)->ppt_funcs->get_dpm_clk_limited ? (smu)->ppt_funcs->get_dpm_clk_limited((smu), (clk_type), (dpm_level), (freq)) : -EINVAL)
+
+#define smu_set_soft_freq_limited_range(smu, clk_type, min, max) \
+		((smu)->ppt_funcs->set_soft_freq_limited_range ? (smu)->ppt_funcs->set_soft_freq_limited_range((smu), (clk_type), (min), (max)) : -EINVAL)
+
+#define smu_override_pcie_parameters(smu) \
+		((smu)->ppt_funcs->override_pcie_parameters ? (smu)->ppt_funcs->override_pcie_parameters((smu)) : 0)
+
+#define smu_update_pcie_parameters(smu, pcie_gen_cap, pcie_width_cap) \
+		((smu)->ppt_funcs->update_pcie_parameters ? (smu)->ppt_funcs->update_pcie_parameters((smu), (pcie_gen_cap), (pcie_width_cap)) : 0)
+
+#endif
diff --git a/drivers/gpu/drm/amd/powerplay/smu_v11_0.c b/drivers/gpu/drm/amd/powerplay/smu_v11_0.c
index c5257ae3188a..fc9679ea2368 100644
--- a/drivers/gpu/drm/amd/powerplay/smu_v11_0.c
+++ b/drivers/gpu/drm/amd/powerplay/smu_v11_0.c
@@ -24,17 +24,19 @@
 #include <linux/module.h>
 #include <linux/pci.h>
 
+#define SMU_11_0_PARTIAL_PPTABLE
+
 #include "pp_debug.h"
 #include "amdgpu.h"
 #include "amdgpu_smu.h"
+#include "smu_internal.h"
 #include "atomfirmware.h"
 #include "amdgpu_atomfirmware.h"
 #include "smu_v11_0.h"
+#include "smu_v11_0_pptable.h"
 #include "soc15_common.h"
 #include "atom.h"
-#include "vega20_ppt.h"
-#include "arcturus_ppt.h"
-#include "navi10_ppt.h"
+#include "amd_pcie.h"
 
 #include "asic_reg/thm/thm_11_0_2_offset.h"
 #include "asic_reg/thm/thm_11_0_2_sh_mask.h"
@@ -61,7 +63,7 @@ static int smu_v11_0_send_msg_without_waiting(struct smu_context *smu,
 	return 0;
 }
 
-static int smu_v11_0_read_arg(struct smu_context *smu, uint32_t *arg)
+int smu_v11_0_read_arg(struct smu_context *smu, uint32_t *arg)
 {
 	struct amdgpu_device *adev = smu->adev;
 
@@ -88,7 +90,7 @@ static int smu_v11_0_wait_for_response(struct smu_context *smu)
 	return RREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_90) == 0x1 ? 0 : -EIO;
 }
 
-static int smu_v11_0_send_msg(struct smu_context *smu, uint16_t msg)
+int smu_v11_0_send_msg(struct smu_context *smu, uint16_t msg)
 {
 	struct amdgpu_device *adev = smu->adev;
 	int ret = 0, index = 0;
@@ -113,7 +115,7 @@ static int smu_v11_0_send_msg(struct smu_context *smu, uint16_t msg)
 
 }
 
-static int
+int
 smu_v11_0_send_msg_with_param(struct smu_context *smu, uint16_t msg,
 			      uint32_t param)
 {
@@ -144,7 +146,7 @@ smu_v11_0_send_msg_with_param(struct smu_context *smu, uint16_t msg,
 	return ret;
 }
 
-static int smu_v11_0_init_microcode(struct smu_context *smu)
+int smu_v11_0_init_microcode(struct smu_context *smu)
 {
 	struct amdgpu_device *adev = smu->adev;
 	const char *chip_name;
@@ -206,7 +208,7 @@ out:
 	return err;
 }
 
-static int smu_v11_0_load_microcode(struct smu_context *smu)
+int smu_v11_0_load_microcode(struct smu_context *smu)
 {
 	struct amdgpu_device *adev = smu->adev;
 	const uint32_t *src;
@@ -244,7 +246,7 @@ static int smu_v11_0_load_microcode(struct smu_context *smu)
 	return 0;
 }
 
-static int smu_v11_0_check_fw_status(struct smu_context *smu)
+int smu_v11_0_check_fw_status(struct smu_context *smu)
 {
 	struct amdgpu_device *adev = smu->adev;
 	uint32_t mp1_fw_flags;
@@ -259,7 +261,7 @@ static int smu_v11_0_check_fw_status(struct smu_context *smu)
 	return -EIO;
 }
 
-static int smu_v11_0_check_fw_version(struct smu_context *smu)
+int smu_v11_0_check_fw_version(struct smu_context *smu)
 {
 	uint32_t if_version = 0xff, smu_version = 0xff;
 	uint16_t smu_major;
@@ -354,7 +356,7 @@ static int smu_v11_0_set_pptable_v2_1(struct smu_context *smu, void **table,
 	return 0;
 }
 
-static int smu_v11_0_setup_pptable(struct smu_context *smu)
+int smu_v11_0_setup_pptable(struct smu_context *smu)
 {
 	struct amdgpu_device *adev = smu->adev;
 	const struct smc_firmware_header_v1_0 *hdr;
@@ -369,6 +371,7 @@ static int smu_v11_0_setup_pptable(struct smu_context *smu)
 	version_major = le16_to_cpu(hdr->header.header_version_major);
 	version_minor = le16_to_cpu(hdr->header.header_version_minor);
 	if (version_major == 2 && smu->smu_table.boot_values.pp_table_id > 0) {
+		pr_info("use driver provided pptable %d\n", smu->smu_table.boot_values.pp_table_id);
 		switch (version_minor) {
 		case 0:
 			ret = smu_v11_0_set_pptable_v2_0(smu, &table, &size);
@@ -385,6 +388,7 @@ static int smu_v11_0_setup_pptable(struct smu_context *smu)
 			return ret;
 
 	} else {
+		pr_info("use vbios provided pptable\n");
 		index = get_index_into_master_table(atom_master_list_of_data_tables_v2_1,
 						    powerplayinfo);
 
@@ -433,13 +437,13 @@ static int smu_v11_0_fini_dpm_context(struct smu_context *smu)
 	return 0;
 }
 
-static int smu_v11_0_init_smc_tables(struct smu_context *smu)
+int smu_v11_0_init_smc_tables(struct smu_context *smu)
 {
 	struct smu_table_context *smu_table = &smu->smu_table;
 	struct smu_table *tables = NULL;
 	int ret = 0;
 
-	if (smu_table->tables || smu_table->table_count == 0)
+	if (smu_table->tables)
 		return -EINVAL;
 
 	tables = kcalloc(SMU_TABLE_COUNT, sizeof(struct smu_table),
@@ -460,18 +464,17 @@ static int smu_v11_0_init_smc_tables(struct smu_context *smu)
 	return 0;
 }
 
-static int smu_v11_0_fini_smc_tables(struct smu_context *smu)
+int smu_v11_0_fini_smc_tables(struct smu_context *smu)
 {
 	struct smu_table_context *smu_table = &smu->smu_table;
 	int ret = 0;
 
-	if (!smu_table->tables || smu_table->table_count == 0)
+	if (!smu_table->tables)
 		return -EINVAL;
 
 	kfree(smu_table->tables);
 	kfree(smu_table->metrics_table);
 	smu_table->tables = NULL;
-	smu_table->table_count = 0;
 	smu_table->metrics_table = NULL;
 	smu_table->metrics_time = 0;
 
@@ -481,7 +484,7 @@ static int smu_v11_0_fini_smc_tables(struct smu_context *smu)
 	return 0;
 }
 
-static int smu_v11_0_init_power(struct smu_context *smu)
+int smu_v11_0_init_power(struct smu_context *smu)
 {
 	struct smu_power_context *smu_power = &smu->smu_power;
 
@@ -499,7 +502,7 @@ static int smu_v11_0_init_power(struct smu_context *smu)
 	return 0;
 }
 
-static int smu_v11_0_fini_power(struct smu_context *smu)
+int smu_v11_0_fini_power(struct smu_context *smu)
 {
 	struct smu_power_context *smu_power = &smu->smu_power;
 
@@ -576,7 +579,7 @@ int smu_v11_0_get_vbios_bootup_values(struct smu_context *smu)
 	return 0;
 }
 
-static int smu_v11_0_get_clk_info_from_vbios(struct smu_context *smu)
+int smu_v11_0_get_clk_info_from_vbios(struct smu_context *smu)
 {
 	int ret, index;
 	struct amdgpu_device *adev = smu->adev;
@@ -673,7 +676,7 @@ static int smu_v11_0_get_clk_info_from_vbios(struct smu_context *smu)
 	return 0;
 }
 
-static int smu_v11_0_notify_memory_pool_location(struct smu_context *smu)
+int smu_v11_0_notify_memory_pool_location(struct smu_context *smu)
 {
 	struct smu_table_context *smu_table = &smu->smu_table;
 	struct smu_table *memory_pool = &smu_table->memory_pool;
@@ -719,7 +722,7 @@ static int smu_v11_0_notify_memory_pool_location(struct smu_context *smu)
 	return ret;
 }
 
-static int smu_v11_0_check_pptable(struct smu_context *smu)
+int smu_v11_0_check_pptable(struct smu_context *smu)
 {
 	int ret;
 
@@ -727,7 +730,7 @@ static int smu_v11_0_check_pptable(struct smu_context *smu)
 	return ret;
 }
 
-static int smu_v11_0_parse_pptable(struct smu_context *smu)
+int smu_v11_0_parse_pptable(struct smu_context *smu)
 {
 	int ret;
 
@@ -751,7 +754,7 @@ static int smu_v11_0_parse_pptable(struct smu_context *smu)
 	return ret;
 }
 
-static int smu_v11_0_populate_smc_pptable(struct smu_context *smu)
+int smu_v11_0_populate_smc_pptable(struct smu_context *smu)
 {
 	int ret;
 
@@ -760,7 +763,7 @@ static int smu_v11_0_populate_smc_pptable(struct smu_context *smu)
 	return ret;
 }
 
-static int smu_v11_0_write_pptable(struct smu_context *smu)
+int smu_v11_0_write_pptable(struct smu_context *smu)
 {
 	struct smu_table_context *table_context = &smu->smu_table;
 	int ret = 0;
@@ -771,24 +774,7 @@ static int smu_v11_0_write_pptable(struct smu_context *smu)
 	return ret;
 }
 
-static int smu_v11_0_write_watermarks_table(struct smu_context *smu)
-{
-	int ret = 0;
-	struct smu_table_context *smu_table = &smu->smu_table;
-	struct smu_table *table = NULL;
-
-	table = &smu_table->tables[SMU_TABLE_WATERMARKS];
-
-	if (!table->cpu_addr)
-		return -EINVAL;
-
-	ret = smu_update_table(smu, SMU_TABLE_WATERMARKS, 0, table->cpu_addr,
-				true);
-
-	return ret;
-}
-
-static int smu_v11_0_set_deep_sleep_dcefclk(struct smu_context *smu, uint32_t clk)
+int smu_v11_0_set_deep_sleep_dcefclk(struct smu_context *smu, uint32_t clk)
 {
 	int ret;
 
@@ -800,7 +786,7 @@ static int smu_v11_0_set_deep_sleep_dcefclk(struct smu_context *smu, uint32_t cl
 	return ret;
 }
 
-static int smu_v11_0_set_min_dcef_deep_sleep(struct smu_context *smu)
+int smu_v11_0_set_min_dcef_deep_sleep(struct smu_context *smu)
 {
 	struct smu_table_context *table_context = &smu->smu_table;
 
@@ -809,11 +795,10 @@ static int smu_v11_0_set_min_dcef_deep_sleep(struct smu_context *smu)
 	if (!table_context)
 		return -EINVAL;
 
-	return smu_set_deep_sleep_dcefclk(smu,
-					  table_context->boot_values.dcefclk / 100);
+	return smu_v11_0_set_deep_sleep_dcefclk(smu, table_context->boot_values.dcefclk / 100);
 }
 
-static int smu_v11_0_set_tool_table_location(struct smu_context *smu)
+int smu_v11_0_set_tool_table_location(struct smu_context *smu)
 {
 	int ret = 0;
 	struct smu_table *tool_table = &smu->smu_table.tables[SMU_TABLE_PMSTATUSLOG];
@@ -831,7 +816,7 @@ static int smu_v11_0_set_tool_table_location(struct smu_context *smu)
 	return ret;
 }
 
-static int smu_v11_0_init_display_count(struct smu_context *smu, uint32_t count)
+int smu_v11_0_init_display_count(struct smu_context *smu, uint32_t count)
 {
 	int ret = 0;
 
@@ -843,7 +828,7 @@ static int smu_v11_0_init_display_count(struct smu_context *smu, uint32_t count)
 }
 
 
-static int smu_v11_0_set_allowed_mask(struct smu_context *smu)
+int smu_v11_0_set_allowed_mask(struct smu_context *smu)
 {
 	struct smu_feature *feature = &smu->smu_feature;
 	int ret = 0;
@@ -870,7 +855,7 @@ failed:
 	return ret;
 }
 
-static int smu_v11_0_get_enabled_mask(struct smu_context *smu,
+int smu_v11_0_get_enabled_mask(struct smu_context *smu,
 				      uint32_t *feature_mask, uint32_t num)
 {
 	uint32_t feature_mask_high = 0, feature_mask_low = 0;
@@ -899,7 +884,7 @@ static int smu_v11_0_get_enabled_mask(struct smu_context *smu,
 	return ret;
 }
 
-static int smu_v11_0_system_features_control(struct smu_context *smu,
+int smu_v11_0_system_features_control(struct smu_context *smu,
 					     bool en)
 {
 	struct smu_feature *feature = &smu->smu_feature;
@@ -925,7 +910,7 @@ static int smu_v11_0_system_features_control(struct smu_context *smu,
 	return ret;
 }
 
-static int smu_v11_0_notify_display_change(struct smu_context *smu)
+int smu_v11_0_notify_display_change(struct smu_context *smu)
 {
 	int ret = 0;
 
@@ -983,7 +968,7 @@ smu_v11_0_get_max_sustainable_clock(struct smu_context *smu, uint32_t *clock,
 	return ret;
 }
 
-static int smu_v11_0_init_max_sustainable_clocks(struct smu_context *smu)
+int smu_v11_0_init_max_sustainable_clocks(struct smu_context *smu)
 {
 	struct smu_11_0_max_sustainable_clocks *max_sustainable_clocks;
 	int ret = 0;
@@ -1063,13 +1048,44 @@ static int smu_v11_0_init_max_sustainable_clocks(struct smu_context *smu)
 	return 0;
 }
 
-static int smu_v11_0_set_power_limit(struct smu_context *smu, uint32_t n)
+uint32_t smu_v11_0_get_max_power_limit(struct smu_context *smu) {
+	uint32_t od_limit, max_power_limit;
+	struct smu_11_0_powerplay_table *powerplay_table = NULL;
+	struct smu_table_context *table_context = &smu->smu_table;
+	powerplay_table = table_context->power_play_table;
+
+	max_power_limit = smu_get_pptable_power_limit(smu);
+
+	if (!max_power_limit) {
+		// If we couldn't get the table limit, fall back on first-read value
+		if (!smu->default_power_limit)
+			smu->default_power_limit = smu->power_limit;
+		max_power_limit = smu->default_power_limit;
+	}
+
+	if (smu->od_enabled) {
+		od_limit = le32_to_cpu(powerplay_table->overdrive_table.max[SMU_11_0_ODSETTING_POWERPERCENTAGE]);
+
+		pr_debug("ODSETTING_POWERPERCENTAGE: %d (default: %d)\n", od_limit, smu->default_power_limit);
+
+		max_power_limit *= (100 + od_limit);
+		max_power_limit /= 100;
+	}
+
+	return max_power_limit;
+}
+
+int smu_v11_0_set_power_limit(struct smu_context *smu, uint32_t n)
 {
 	int ret = 0;
+	uint32_t max_power_limit;
 
-	if (n > smu->default_power_limit) {
-		pr_err("New power limit is over the max allowed %d\n",
-				smu->default_power_limit);
+	max_power_limit = smu_v11_0_get_max_power_limit(smu);
+
+	if (n > max_power_limit) {
+		pr_err("New power limit (%d) is over the max allowed %d\n",
+				n,
+				max_power_limit);
 		return -EINVAL;
 	}
 
@@ -1091,7 +1107,7 @@ static int smu_v11_0_set_power_limit(struct smu_context *smu, uint32_t n)
 	return 0;
 }
 
-static int smu_v11_0_get_current_clk_freq(struct smu_context *smu,
+int smu_v11_0_get_current_clk_freq(struct smu_context *smu,
 					  enum smu_clk_type clk_id,
 					  uint32_t *value)
 {
@@ -1170,7 +1186,7 @@ static int smu_v11_0_enable_thermal_alert(struct smu_context *smu)
 	return 0;
 }
 
-static int smu_v11_0_start_thermal_control(struct smu_context *smu)
+int smu_v11_0_start_thermal_control(struct smu_context *smu)
 {
 	int ret = 0;
 	struct smu_temperature_range range;
@@ -1212,6 +1228,15 @@ static int smu_v11_0_start_thermal_control(struct smu_context *smu)
 	return ret;
 }
 
+int smu_v11_0_stop_thermal_control(struct smu_context *smu)
+{
+	struct amdgpu_device *adev = smu->adev;
+
+	WREG32_SOC15(THM, 0, mmTHM_THERMAL_INT_ENA, 0);
+
+	return 0;
+}
+
 static uint16_t convert_to_vddc(uint8_t vid)
 {
 	return (uint16_t) ((6200 - (vid * 25)) / SMU11_VOLTAGE_SCALE);
@@ -1236,7 +1261,7 @@ static int smu_v11_0_get_gfx_vdd(struct smu_context *smu, uint32_t *value)
 
 }
 
-static int smu_v11_0_read_sensor(struct smu_context *smu,
+int smu_v11_0_read_sensor(struct smu_context *smu,
 				 enum amd_pp_sensors sensor,
 				 void *data, uint32_t *size)
 {
@@ -1273,7 +1298,7 @@ static int smu_v11_0_read_sensor(struct smu_context *smu,
 	return ret;
 }
 
-static int
+int
 smu_v11_0_display_clock_voltage_request(struct smu_context *smu,
 					struct pp_display_clock_request
 					*clock_req)
@@ -1316,9 +1341,7 @@ smu_v11_0_display_clock_voltage_request(struct smu_context *smu,
 		if (clk_select == SMU_UCLK && smu->disable_uclk_switch)
 			return 0;
 
-		mutex_lock(&smu->mutex);
 		ret = smu_set_hard_freq_range(smu, clk_select, clk_freq, 0);
-		mutex_unlock(&smu->mutex);
 
 		if(clk_select == SMU_UCLK)
 			smu->hard_min_uclk_req_from_dal = clk_freq;
@@ -1328,27 +1351,7 @@ failed:
 	return ret;
 }
 
-static int
-smu_v11_0_set_watermarks_for_clock_ranges(struct smu_context *smu, struct
-					  dm_pp_wm_sets_with_clock_ranges_soc15
-					  *clock_ranges)
-{
-	int ret = 0;
-	struct smu_table *watermarks = &smu->smu_table.tables[SMU_TABLE_WATERMARKS];
-	void *table = watermarks->cpu_addr;
-
-	if (!smu->disable_watermark &&
-	    smu_feature_is_enabled(smu, SMU_FEATURE_DPM_DCEFCLK_BIT) &&
-	    smu_feature_is_enabled(smu, SMU_FEATURE_DPM_SOCCLK_BIT)) {
-		smu_set_watermarks_table(smu, table, clock_ranges);
-		smu->watermarks_bitmap |= WATERMARKS_EXIST;
-		smu->watermarks_bitmap &= ~WATERMARKS_LOADED;
-	}
-
-	return ret;
-}
-
-static int smu_v11_0_gfx_off_control(struct smu_context *smu, bool enable)
+int smu_v11_0_gfx_off_control(struct smu_context *smu, bool enable)
 {
 	int ret = 0;
 	struct amdgpu_device *adev = smu->adev;
@@ -1361,12 +1364,10 @@ static int smu_v11_0_gfx_off_control(struct smu_context *smu, bool enable)
 	case CHIP_NAVI12:
 		if (!(adev->pm.pp_feature & PP_GFXOFF_MASK))
 			return 0;
-		mutex_lock(&smu->mutex);
 		if (enable)
 			ret = smu_send_smc_msg(smu, SMU_MSG_AllowGfxOff);
 		else
 			ret = smu_send_smc_msg(smu, SMU_MSG_DisallowGfxOff);
-		mutex_unlock(&smu->mutex);
 		break;
 	default:
 		break;
@@ -1375,7 +1376,7 @@ static int smu_v11_0_gfx_off_control(struct smu_context *smu, bool enable)
 	return ret;
 }
 
-static uint32_t
+uint32_t
 smu_v11_0_get_fan_control_mode(struct smu_context *smu)
 {
 	if (!smu_feature_is_enabled(smu, SMU_FEATURE_FAN_CONTROL_BIT))
@@ -1415,7 +1416,7 @@ smu_v11_0_set_fan_static_mode(struct smu_context *smu, uint32_t mode)
 	return 0;
 }
 
-static int
+int
 smu_v11_0_set_fan_speed_percent(struct smu_context *smu, uint32_t speed)
 {
 	struct amdgpu_device *adev = smu->adev;
@@ -1444,7 +1445,7 @@ smu_v11_0_set_fan_speed_percent(struct smu_context *smu, uint32_t speed)
 	return smu_v11_0_set_fan_static_mode(smu, FDO_PWM_MODE_STATIC);
 }
 
-static int
+int
 smu_v11_0_set_fan_control_mode(struct smu_context *smu,
 			       uint32_t mode)
 {
@@ -1472,7 +1473,7 @@ smu_v11_0_set_fan_control_mode(struct smu_context *smu,
 	return ret;
 }
 
-static int smu_v11_0_set_fan_speed_rpm(struct smu_context *smu,
+int smu_v11_0_set_fan_speed_rpm(struct smu_context *smu,
 				       uint32_t speed)
 {
 	struct amdgpu_device *adev = smu->adev;
@@ -1482,10 +1483,9 @@ static int smu_v11_0_set_fan_speed_rpm(struct smu_context *smu,
 	if (!speed)
 		return -EINVAL;
 
-	mutex_lock(&(smu->mutex));
 	ret = smu_v11_0_auto_fan_control(smu, 0);
 	if (ret)
-		goto set_fan_speed_rpm_failed;
+		return ret;
 
 	crystal_clock_freq = amdgpu_asic_get_xclk(adev);
 	tach_period = 60 * crystal_clock_freq * 10000 / (8 * speed);
@@ -1496,23 +1496,16 @@ static int smu_v11_0_set_fan_speed_rpm(struct smu_context *smu,
 
 	ret = smu_v11_0_set_fan_static_mode(smu, FDO_PWM_MODE_STATIC_RPM);
 
-set_fan_speed_rpm_failed:
-	mutex_unlock(&(smu->mutex));
 	return ret;
 }
 
-#define XGMI_STATE_D0 1
-#define XGMI_STATE_D3 0
-
-static int smu_v11_0_set_xgmi_pstate(struct smu_context *smu,
+int smu_v11_0_set_xgmi_pstate(struct smu_context *smu,
 				     uint32_t pstate)
 {
 	int ret = 0;
-	mutex_lock(&(smu->mutex));
 	ret = smu_send_smc_msg_with_param(smu,
 					  SMU_MSG_SetXgmiMode,
-					  pstate ? XGMI_STATE_D0 : XGMI_STATE_D3);
-	mutex_unlock(&(smu->mutex));
+					  pstate ? XGMI_MODE_PSTATE_D0 : XGMI_MODE_PSTATE_D3);
 	return ret;
 }
 
@@ -1559,7 +1552,7 @@ static const struct amdgpu_irq_src_funcs smu_v11_0_irq_funcs =
 	.process = smu_v11_0_irq_process,
 };
 
-static int smu_v11_0_register_irq_handler(struct smu_context *smu)
+int smu_v11_0_register_irq_handler(struct smu_context *smu)
 {
 	struct amdgpu_device *adev = smu->adev;
 	struct amdgpu_irq_src *irq_src = smu->irq_source;
@@ -1591,7 +1584,7 @@ static int smu_v11_0_register_irq_handler(struct smu_context *smu)
 	return ret;
 }
 
-static int smu_v11_0_get_max_sustainable_clocks_by_dc(struct smu_context *smu,
+int smu_v11_0_get_max_sustainable_clocks_by_dc(struct smu_context *smu,
 		struct pp_smu_nv_clock_table *max_clocks)
 {
 	struct smu_table_context *table_context = &smu->smu_table;
@@ -1621,13 +1614,11 @@ static int smu_v11_0_get_max_sustainable_clocks_by_dc(struct smu_context *smu,
 	return 0;
 }
 
-static int smu_v11_0_set_azalia_d3_pme(struct smu_context *smu)
+int smu_v11_0_set_azalia_d3_pme(struct smu_context *smu)
 {
 	int ret = 0;
 
-	mutex_lock(&smu->mutex);
 	ret = smu_send_smc_msg(smu, SMU_MSG_BacoAudioD3PME);
-	mutex_unlock(&smu->mutex);
 
 	return ret;
 }
@@ -1637,7 +1628,7 @@ static int smu_v11_0_baco_set_armd3_sequence(struct smu_context *smu, enum smu_v
 	return smu_send_smc_msg_with_param(smu, SMU_MSG_ArmD3, baco_seq);
 }
 
-static bool smu_v11_0_baco_is_support(struct smu_context *smu)
+bool smu_v11_0_baco_is_support(struct smu_context *smu)
 {
 	struct amdgpu_device *adev = smu->adev;
 	struct smu_baco_context *smu_baco = &smu->smu_baco;
@@ -1661,7 +1652,7 @@ static bool smu_v11_0_baco_is_support(struct smu_context *smu)
 	return false;
 }
 
-static enum smu_baco_state smu_v11_0_baco_get_state(struct smu_context *smu)
+enum smu_baco_state smu_v11_0_baco_get_state(struct smu_context *smu)
 {
 	struct smu_baco_context *smu_baco = &smu->smu_baco;
 	enum smu_baco_state baco_state;
@@ -1673,7 +1664,7 @@ static enum smu_baco_state smu_v11_0_baco_get_state(struct smu_context *smu)
 	return baco_state;
 }
 
-static int smu_v11_0_baco_set_state(struct smu_context *smu, enum smu_baco_state state)
+int smu_v11_0_baco_set_state(struct smu_context *smu, enum smu_baco_state state)
 {
 
 	struct smu_baco_context *smu_baco = &smu->smu_baco;
@@ -1697,7 +1688,7 @@ out:
 	return ret;
 }
 
-static int smu_v11_0_baco_reset(struct smu_context *smu)
+int smu_v11_0_baco_reset(struct smu_context *smu)
 {
 	int ret = 0;
 
@@ -1718,13 +1709,12 @@ static int smu_v11_0_baco_reset(struct smu_context *smu)
 	return ret;
 }
 
-static int smu_v11_0_get_dpm_ultimate_freq(struct smu_context *smu, enum smu_clk_type clk_type,
+int smu_v11_0_get_dpm_ultimate_freq(struct smu_context *smu, enum smu_clk_type clk_type,
 						 uint32_t *min, uint32_t *max)
 {
 	int ret = 0, clk_id = 0;
 	uint32_t param = 0;
 
-	mutex_lock(&smu->mutex);
 	clk_id = smu_clk_get_index(smu, clk_type);
 	if (clk_id < 0) {
 		ret = -EINVAL;
@@ -1751,80 +1741,102 @@ static int smu_v11_0_get_dpm_ultimate_freq(struct smu_context *smu, enum smu_clk
 	}
 
 failed:
-	mutex_unlock(&smu->mutex);
 	return ret;
 }
 
-static const struct smu_funcs smu_v11_0_funcs = {
-	.init_microcode = smu_v11_0_init_microcode,
-	.load_microcode = smu_v11_0_load_microcode,
-	.check_fw_status = smu_v11_0_check_fw_status,
-	.check_fw_version = smu_v11_0_check_fw_version,
-	.send_smc_msg = smu_v11_0_send_msg,
-	.send_smc_msg_with_param = smu_v11_0_send_msg_with_param,
-	.read_smc_arg = smu_v11_0_read_arg,
-	.setup_pptable = smu_v11_0_setup_pptable,
-	.init_smc_tables = smu_v11_0_init_smc_tables,
-	.fini_smc_tables = smu_v11_0_fini_smc_tables,
-	.init_power = smu_v11_0_init_power,
-	.fini_power = smu_v11_0_fini_power,
-	.get_vbios_bootup_values = smu_v11_0_get_vbios_bootup_values,
-	.get_clk_info_from_vbios = smu_v11_0_get_clk_info_from_vbios,
-	.notify_memory_pool_location = smu_v11_0_notify_memory_pool_location,
-	.check_pptable = smu_v11_0_check_pptable,
-	.parse_pptable = smu_v11_0_parse_pptable,
-	.populate_smc_tables = smu_v11_0_populate_smc_pptable,
-	.write_pptable = smu_v11_0_write_pptable,
-	.write_watermarks_table = smu_v11_0_write_watermarks_table,
-	.set_min_dcef_deep_sleep = smu_v11_0_set_min_dcef_deep_sleep,
-	.set_tool_table_location = smu_v11_0_set_tool_table_location,
-	.init_display_count = smu_v11_0_init_display_count,
-	.set_allowed_mask = smu_v11_0_set_allowed_mask,
-	.get_enabled_mask = smu_v11_0_get_enabled_mask,
-	.system_features_control = smu_v11_0_system_features_control,
-	.notify_display_change = smu_v11_0_notify_display_change,
-	.set_power_limit = smu_v11_0_set_power_limit,
-	.get_current_clk_freq = smu_v11_0_get_current_clk_freq,
-	.init_max_sustainable_clocks = smu_v11_0_init_max_sustainable_clocks,
-	.start_thermal_control = smu_v11_0_start_thermal_control,
-	.read_sensor = smu_v11_0_read_sensor,
-	.set_deep_sleep_dcefclk = smu_v11_0_set_deep_sleep_dcefclk,
-	.display_clock_voltage_request = smu_v11_0_display_clock_voltage_request,
-	.set_watermarks_for_clock_ranges = smu_v11_0_set_watermarks_for_clock_ranges,
-	.get_fan_control_mode = smu_v11_0_get_fan_control_mode,
-	.set_fan_control_mode = smu_v11_0_set_fan_control_mode,
-	.set_fan_speed_percent = smu_v11_0_set_fan_speed_percent,
-	.set_fan_speed_rpm = smu_v11_0_set_fan_speed_rpm,
-	.set_xgmi_pstate = smu_v11_0_set_xgmi_pstate,
-	.gfx_off_control = smu_v11_0_gfx_off_control,
-	.register_irq_handler = smu_v11_0_register_irq_handler,
-	.set_azalia_d3_pme = smu_v11_0_set_azalia_d3_pme,
-	.get_max_sustainable_clocks_by_dc = smu_v11_0_get_max_sustainable_clocks_by_dc,
-	.baco_is_support = smu_v11_0_baco_is_support,
-	.baco_get_state = smu_v11_0_baco_get_state,
-	.baco_set_state = smu_v11_0_baco_set_state,
-	.baco_reset = smu_v11_0_baco_reset,
-	.get_dpm_ultimate_freq = smu_v11_0_get_dpm_ultimate_freq,
-};
+int smu_v11_0_set_soft_freq_limited_range(struct smu_context *smu, enum smu_clk_type clk_type,
+			    uint32_t min, uint32_t max)
+{
+	int ret = 0, clk_id = 0;
+	uint32_t param;
 
-void smu_v11_0_set_smu_funcs(struct smu_context *smu)
+	clk_id = smu_clk_get_index(smu, clk_type);
+	if (clk_id < 0)
+		return clk_id;
+
+	if (max > 0) {
+		param = (uint32_t)((clk_id << 16) | (max & 0xffff));
+		ret = smu_send_smc_msg_with_param(smu, SMU_MSG_SetSoftMaxByFreq,
+						  param);
+		if (ret)
+			return ret;
+	}
+
+	if (min > 0) {
+		param = (uint32_t)((clk_id << 16) | (min & 0xffff));
+		ret = smu_send_smc_msg_with_param(smu, SMU_MSG_SetSoftMinByFreq,
+						  param);
+		if (ret)
+			return ret;
+	}
+
+	return ret;
+}
+
+int smu_v11_0_override_pcie_parameters(struct smu_context *smu)
 {
 	struct amdgpu_device *adev = smu->adev;
+	uint32_t pcie_gen = 0, pcie_width = 0;
+	int ret;
 
-	smu->funcs = &smu_v11_0_funcs;
-	switch (adev->asic_type) {
-	case CHIP_VEGA20:
-		vega20_set_ppt_funcs(smu);
-		break;
-	case CHIP_ARCTURUS:
-		arcturus_set_ppt_funcs(smu);
-		break;
-	case CHIP_NAVI10:
-	case CHIP_NAVI14:
-	case CHIP_NAVI12:
-		navi10_set_ppt_funcs(smu);
-		break;
-	default:
-		pr_warn("Unknown asic for smu11\n");
+	if (adev->pm.pcie_gen_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_GEN4)
+		pcie_gen = 3;
+	else if (adev->pm.pcie_gen_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_GEN3)
+		pcie_gen = 2;
+	else if (adev->pm.pcie_gen_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_GEN2)
+		pcie_gen = 1;
+	else if (adev->pm.pcie_gen_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_GEN1)
+		pcie_gen = 0;
+
+	/* Bit 31:16: LCLK DPM level. 0 is DPM0, and 1 is DPM1
+	 * Bit 15:8:  PCIE GEN, 0 to 3 corresponds to GEN1 to GEN4
+	 * Bit 7:0:   PCIE lane width, 1 to 7 corresponds is x1 to x32
+	 */
+	if (adev->pm.pcie_mlw_mask & CAIL_PCIE_LINK_WIDTH_SUPPORT_X16)
+		pcie_width = 6;
+	else if (adev->pm.pcie_mlw_mask & CAIL_PCIE_LINK_WIDTH_SUPPORT_X12)
+		pcie_width = 5;
+	else if (adev->pm.pcie_mlw_mask & CAIL_PCIE_LINK_WIDTH_SUPPORT_X8)
+		pcie_width = 4;
+	else if (adev->pm.pcie_mlw_mask & CAIL_PCIE_LINK_WIDTH_SUPPORT_X4)
+		pcie_width = 3;
+	else if (adev->pm.pcie_mlw_mask & CAIL_PCIE_LINK_WIDTH_SUPPORT_X2)
+		pcie_width = 2;
+	else if (adev->pm.pcie_mlw_mask & CAIL_PCIE_LINK_WIDTH_SUPPORT_X1)
+		pcie_width = 1;
+
+	ret = smu_update_pcie_parameters(smu, pcie_gen, pcie_width);
+
+	if (ret)
+		pr_err("[%s] Attempt to override pcie params failed!\n", __func__);
+
+	return ret;
+
+}
+
+int smu_v11_0_set_default_od_settings(struct smu_context *smu, bool initialize, size_t overdrive_table_size)
+{
+	struct smu_table_context *table_context = &smu->smu_table;
+	int ret = 0;
+
+	if (initialize) {
+		if (table_context->overdrive_table) {
+			return -EINVAL;
+		}
+		table_context->overdrive_table = kzalloc(overdrive_table_size, GFP_KERNEL);
+		if (!table_context->overdrive_table) {
+			return -ENOMEM;
+		}
+		ret = smu_update_table(smu, SMU_TABLE_OVERDRIVE, 0, table_context->overdrive_table, false);
+		if (ret) {
+			pr_err("Failed to export overdrive table!\n");
+			return ret;
+		}
 	}
+	ret = smu_update_table(smu, SMU_TABLE_OVERDRIVE, 0, table_context->overdrive_table, true);
+	if (ret) {
+		pr_err("Failed to import overdrive table!\n");
+		return ret;
+	}
+	return ret;
 }
diff --git a/drivers/gpu/drm/amd/powerplay/smu_v12_0.c b/drivers/gpu/drm/amd/powerplay/smu_v12_0.c
index 9d2280ca1f4b..139dd737eaa5 100644
--- a/drivers/gpu/drm/amd/powerplay/smu_v12_0.c
+++ b/drivers/gpu/drm/amd/powerplay/smu_v12_0.c
@@ -24,12 +24,12 @@
 #include <linux/firmware.h>
 #include "amdgpu.h"
 #include "amdgpu_smu.h"
+#include "smu_internal.h"
 #include "atomfirmware.h"
 #include "amdgpu_atomfirmware.h"
 #include "smu_v12_0.h"
 #include "soc15_common.h"
 #include "atom.h"
-#include "renoir_ppt.h"
 
 #include "asic_reg/mp/mp_12_0_0_offset.h"
 #include "asic_reg/mp/mp_12_0_0_sh_mask.h"
@@ -41,7 +41,7 @@
 #define SMUIO_GFX_MISC_CNTL__PWR_GFXOFF_STATUS_MASK          0x00000006L
 #define SMUIO_GFX_MISC_CNTL__PWR_GFXOFF_STATUS__SHIFT        0x1
 
-static int smu_v12_0_send_msg_without_waiting(struct smu_context *smu,
+int smu_v12_0_send_msg_without_waiting(struct smu_context *smu,
 					      uint16_t msg)
 {
 	struct amdgpu_device *adev = smu->adev;
@@ -50,7 +50,7 @@ static int smu_v12_0_send_msg_without_waiting(struct smu_context *smu,
 	return 0;
 }
 
-static int smu_v12_0_read_arg(struct smu_context *smu, uint32_t *arg)
+int smu_v12_0_read_arg(struct smu_context *smu, uint32_t *arg)
 {
 	struct amdgpu_device *adev = smu->adev;
 
@@ -58,7 +58,7 @@ static int smu_v12_0_read_arg(struct smu_context *smu, uint32_t *arg)
 	return 0;
 }
 
-static int smu_v12_0_wait_for_response(struct smu_context *smu)
+int smu_v12_0_wait_for_response(struct smu_context *smu)
 {
 	struct amdgpu_device *adev = smu->adev;
 	uint32_t cur_value, i;
@@ -77,7 +77,7 @@ static int smu_v12_0_wait_for_response(struct smu_context *smu)
 	return RREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_90) == 0x1 ? 0 : -EIO;
 }
 
-static int smu_v12_0_send_msg(struct smu_context *smu, uint16_t msg)
+int smu_v12_0_send_msg(struct smu_context *smu, uint16_t msg)
 {
 	struct amdgpu_device *adev = smu->adev;
 	int ret = 0, index = 0;
@@ -102,7 +102,7 @@ static int smu_v12_0_send_msg(struct smu_context *smu, uint16_t msg)
 
 }
 
-static int
+int
 smu_v12_0_send_msg_with_param(struct smu_context *smu, uint16_t msg,
 			      uint32_t param)
 {
@@ -132,7 +132,7 @@ smu_v12_0_send_msg_with_param(struct smu_context *smu, uint16_t msg,
 	return ret;
 }
 
-static int smu_v12_0_check_fw_status(struct smu_context *smu)
+int smu_v12_0_check_fw_status(struct smu_context *smu)
 {
 	struct amdgpu_device *adev = smu->adev;
 	uint32_t mp1_fw_flags;
@@ -147,7 +147,7 @@ static int smu_v12_0_check_fw_status(struct smu_context *smu)
 	return -EIO;
 }
 
-static int smu_v12_0_check_fw_version(struct smu_context *smu)
+int smu_v12_0_check_fw_version(struct smu_context *smu)
 {
 	uint32_t if_version = 0xff, smu_version = 0xff;
 	uint16_t smu_major;
@@ -181,7 +181,7 @@ static int smu_v12_0_check_fw_version(struct smu_context *smu)
 	return ret;
 }
 
-static int smu_v12_0_powergate_sdma(struct smu_context *smu, bool gate)
+int smu_v12_0_powergate_sdma(struct smu_context *smu, bool gate)
 {
 	if (!(smu->adev->flags & AMD_IS_APU))
 		return 0;
@@ -192,7 +192,7 @@ static int smu_v12_0_powergate_sdma(struct smu_context *smu, bool gate)
 		return smu_send_smc_msg(smu, SMU_MSG_PowerUpSdma);
 }
 
-static int smu_v12_0_powergate_vcn(struct smu_context *smu, bool gate)
+int smu_v12_0_powergate_vcn(struct smu_context *smu, bool gate)
 {
 	if (!(smu->adev->flags & AMD_IS_APU))
 		return 0;
@@ -203,7 +203,7 @@ static int smu_v12_0_powergate_vcn(struct smu_context *smu, bool gate)
 		return smu_send_smc_msg(smu, SMU_MSG_PowerUpVcn);
 }
 
-static int smu_v12_0_set_gfx_cgpg(struct smu_context *smu, bool enable)
+int smu_v12_0_set_gfx_cgpg(struct smu_context *smu, bool enable)
 {
 	if (!(smu->adev->pg_flags & AMD_PG_SUPPORT_GFX_PG))
 		return 0;
@@ -224,7 +224,7 @@ static int smu_v12_0_set_gfx_cgpg(struct smu_context *smu, bool enable)
  * Returns 2=Not in GFXOFF.
  * Returns 3=Transition into GFXOFF.
  */
-static uint32_t smu_v12_0_get_gfxoff_status(struct smu_context *smu)
+uint32_t smu_v12_0_get_gfxoff_status(struct smu_context *smu)
 {
 	uint32_t reg;
 	uint32_t gfxOff_Status = 0;
@@ -237,22 +237,13 @@ static uint32_t smu_v12_0_get_gfxoff_status(struct smu_context *smu)
 	return gfxOff_Status;
 }
 
-static int smu_v12_0_gfx_off_control(struct smu_context *smu, bool enable)
+int smu_v12_0_gfx_off_control(struct smu_context *smu, bool enable)
 {
 	int ret = 0, timeout = 500;
 
 	if (enable) {
 		ret = smu_send_smc_msg(smu, SMU_MSG_AllowGfxOff);
 
-		/* confirm gfx is back to "off" state, timeout is 5 seconds */
-		while (!(smu_v12_0_get_gfxoff_status(smu) == 0)) {
-			msleep(10);
-			timeout--;
-			if (timeout == 0) {
-				DRM_ERROR("enable gfxoff timeout and failed!\n");
-				break;
-			}
-		}
 	} else {
 		ret = smu_send_smc_msg(smu, SMU_MSG_DisallowGfxOff);
 
@@ -270,12 +261,12 @@ static int smu_v12_0_gfx_off_control(struct smu_context *smu, bool enable)
 	return ret;
 }
 
-static int smu_v12_0_init_smc_tables(struct smu_context *smu)
+int smu_v12_0_init_smc_tables(struct smu_context *smu)
 {
 	struct smu_table_context *smu_table = &smu->smu_table;
 	struct smu_table *tables = NULL;
 
-	if (smu_table->tables || smu_table->table_count == 0)
+	if (smu_table->tables)
 		return -EINVAL;
 
 	tables = kcalloc(SMU_TABLE_COUNT, sizeof(struct smu_table),
@@ -288,11 +279,11 @@ static int smu_v12_0_init_smc_tables(struct smu_context *smu)
 	return smu_tables_init(smu, tables);
 }
 
-static int smu_v12_0_fini_smc_tables(struct smu_context *smu)
+int smu_v12_0_fini_smc_tables(struct smu_context *smu)
 {
 	struct smu_table_context *smu_table = &smu->smu_table;
 
-	if (!smu_table->tables || smu_table->table_count == 0)
+	if (!smu_table->tables)
 		return -EINVAL;
 
 	kfree(smu_table->clocks_table);
@@ -304,7 +295,7 @@ static int smu_v12_0_fini_smc_tables(struct smu_context *smu)
 	return 0;
 }
 
-static int smu_v12_0_populate_smc_tables(struct smu_context *smu)
+int smu_v12_0_populate_smc_tables(struct smu_context *smu)
 {
 	struct smu_table_context *smu_table = &smu->smu_table;
 	struct smu_table *table = NULL;
@@ -319,14 +310,20 @@ static int smu_v12_0_populate_smc_tables(struct smu_context *smu)
 	return smu_update_table(smu, SMU_TABLE_DPMCLOCKS, 0, smu_table->clocks_table, false);
 }
 
-static int smu_v12_0_get_dpm_ultimate_freq(struct smu_context *smu, enum smu_clk_type clk_type,
+int smu_v12_0_get_dpm_ultimate_freq(struct smu_context *smu, enum smu_clk_type clk_type,
 						 uint32_t *min, uint32_t *max)
 {
 	int ret = 0;
-
-	mutex_lock(&smu->mutex);
+	uint32_t mclk_mask, soc_mask;
 
 	if (max) {
+		ret = smu_get_profiling_clk_mask(smu, AMD_DPM_FORCED_LEVEL_PROFILE_PEAK,
+						 NULL,
+						 &mclk_mask,
+						 &soc_mask);
+		if (ret)
+			goto failed;
+
 		switch (clk_type) {
 		case SMU_GFXCLK:
 		case SMU_SCLK:
@@ -340,14 +337,20 @@ static int smu_v12_0_get_dpm_ultimate_freq(struct smu_context *smu, enum smu_clk
 				goto failed;
 			break;
 		case SMU_UCLK:
-			ret = smu_get_dpm_uclk_limited(smu, max, true);
+		case SMU_FCLK:
+		case SMU_MCLK:
+			ret = smu_get_dpm_clk_limited(smu, clk_type, mclk_mask, max);
+			if (ret)
+				goto failed;
+			break;
+		case SMU_SOCCLK:
+			ret = smu_get_dpm_clk_limited(smu, clk_type, soc_mask, max);
 			if (ret)
 				goto failed;
 			break;
 		default:
 			ret = -EINVAL;
 			goto failed;
-
 		}
 	}
 
@@ -365,7 +368,14 @@ static int smu_v12_0_get_dpm_ultimate_freq(struct smu_context *smu, enum smu_clk
 				goto failed;
 			break;
 		case SMU_UCLK:
-			ret = smu_get_dpm_uclk_limited(smu, min, false);
+		case SMU_FCLK:
+		case SMU_MCLK:
+			ret = smu_get_dpm_clk_limited(smu, clk_type, 0, min);
+			if (ret)
+				goto failed;
+			break;
+		case SMU_SOCCLK:
+			ret = smu_get_dpm_clk_limited(smu, clk_type, 0, min);
 			if (ret)
 				goto failed;
 			break;
@@ -373,40 +383,65 @@ static int smu_v12_0_get_dpm_ultimate_freq(struct smu_context *smu, enum smu_clk
 			ret = -EINVAL;
 			goto failed;
 		}
-
 	}
 failed:
-	mutex_unlock(&smu->mutex);
 	return ret;
 }
 
-static const struct smu_funcs smu_v12_0_funcs = {
-	.check_fw_status = smu_v12_0_check_fw_status,
-	.check_fw_version = smu_v12_0_check_fw_version,
-	.powergate_sdma = smu_v12_0_powergate_sdma,
-	.powergate_vcn = smu_v12_0_powergate_vcn,
-	.send_smc_msg = smu_v12_0_send_msg,
-	.send_smc_msg_with_param = smu_v12_0_send_msg_with_param,
-	.read_smc_arg = smu_v12_0_read_arg,
-	.set_gfx_cgpg = smu_v12_0_set_gfx_cgpg,
-	.gfx_off_control = smu_v12_0_gfx_off_control,
-	.init_smc_tables = smu_v12_0_init_smc_tables,
-	.fini_smc_tables = smu_v12_0_fini_smc_tables,
-	.populate_smc_tables = smu_v12_0_populate_smc_tables,
-	.get_dpm_ultimate_freq = smu_v12_0_get_dpm_ultimate_freq,
-};
-
-void smu_v12_0_set_smu_funcs(struct smu_context *smu)
+int smu_v12_0_mode2_reset(struct smu_context *smu){
+	return smu_v12_0_send_msg_with_param(smu, SMU_MSG_GfxDeviceDriverReset, SMU_RESET_MODE_2);
+}
+
+int smu_v12_0_set_soft_freq_limited_range(struct smu_context *smu, enum smu_clk_type clk_type,
+			    uint32_t min, uint32_t max)
 {
-	struct amdgpu_device *adev = smu->adev;
+	int ret = 0;
 
-	smu->funcs = &smu_v12_0_funcs;
+	if (max < min)
+		return -EINVAL;
 
-	switch (adev->asic_type) {
-	case CHIP_RENOIR:
-		renoir_set_ppt_funcs(smu);
-		break;
+	switch (clk_type) {
+	case SMU_GFXCLK:
+	case SMU_SCLK:
+		ret = smu_send_smc_msg_with_param(smu, SMU_MSG_SetHardMinGfxClk, min);
+		if (ret)
+			return ret;
+
+		ret = smu_send_smc_msg_with_param(smu, SMU_MSG_SetSoftMaxGfxClk, max);
+		if (ret)
+			return ret;
+	break;
+	case SMU_FCLK:
+	case SMU_MCLK:
+		ret = smu_send_smc_msg_with_param(smu, SMU_MSG_SetHardMinFclkByFreq, min);
+		if (ret)
+			return ret;
+
+		ret = smu_send_smc_msg_with_param(smu, SMU_MSG_SetSoftMaxFclkByFreq, max);
+		if (ret)
+			return ret;
+	break;
+	case SMU_SOCCLK:
+		ret = smu_send_smc_msg_with_param(smu, SMU_MSG_SetHardMinSocclkByFreq, min);
+		if (ret)
+			return ret;
+
+		ret = smu_send_smc_msg_with_param(smu, SMU_MSG_SetSoftMaxSocclkByFreq, max);
+		if (ret)
+			return ret;
+	break;
+	case SMU_VCLK:
+		ret = smu_send_smc_msg_with_param(smu, SMU_MSG_SetHardMinVcn, min);
+		if (ret)
+			return ret;
+
+		ret = smu_send_smc_msg_with_param(smu, SMU_MSG_SetSoftMaxVcn, max);
+		if (ret)
+			return ret;
+	break;
 	default:
-		pr_warn("Unknown asic for smu12\n");
+		return -EINVAL;
 	}
+
+	return ret;
 }
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c
index dc754447f0dd..23c12018dbc1 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c
@@ -655,7 +655,7 @@ static int polaris10_populate_smc_mvdd_table(struct pp_hwmgr *hwmgr,
 			count = SMU_MAX_SMIO_LEVELS;
 		for (level = 0; level < count; level++) {
 			table->SmioTable2.Pattern[level].Voltage =
-				PP_HOST_TO_SMC_US(data->mvdd_voltage_table.entries[count].value * VOLTAGE_SCALE);
+				PP_HOST_TO_SMC_US(data->mvdd_voltage_table.entries[level].value * VOLTAGE_SCALE);
 			/* Index into DpmTable.Smio. Drive bits from Smio entry to get this voltage level.*/
 			table->SmioTable2.Pattern[level].Smio =
 				(uint8_t) level;
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smu10_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/smu10_smumgr.c
index 3f12cf341511..aa0ee2b46135 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/smu10_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/smu10_smumgr.c
@@ -137,7 +137,7 @@ static int smu10_copy_table_from_smc(struct pp_hwmgr *hwmgr,
 			priv->smu_tables.entry[table_id].table_id);
 
 	/* flush hdp cache */
-	adev->nbio_funcs->hdp_flush(adev, NULL);
+	adev->nbio.funcs->hdp_flush(adev, NULL);
 
 	memcpy(table, (uint8_t *)priv->smu_tables.entry[table_id].table,
 			priv->smu_tables.entry[table_id].size);
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smu8_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/smu8_smumgr.c
index 4728aa23a818..7dca04a89217 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/smu8_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/smu8_smumgr.c
@@ -177,12 +177,10 @@ static int smu8_load_mec_firmware(struct pp_hwmgr *hwmgr)
 	uint32_t tmp;
 	int ret = 0;
 	struct cgs_firmware_info info = {0};
-	struct smu8_smumgr *smu8_smu;
 
 	if (hwmgr == NULL || hwmgr->device == NULL)
 		return -EINVAL;
 
-	smu8_smu = hwmgr->smu_backend;
 	ret = cgs_get_firmware_info(hwmgr->device,
 						CGS_UCODE_ID_CP_MEC, &info);
 
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/vega10_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/vega10_smumgr.c
index 0dbdde69f2d9..0f3836fd9666 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/vega10_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/vega10_smumgr.c
@@ -58,7 +58,7 @@ static int vega10_copy_table_from_smc(struct pp_hwmgr *hwmgr,
 			priv->smu_tables.entry[table_id].table_id);
 
 	/* flush hdp cache */
-	adev->nbio_funcs->hdp_flush(adev, NULL);
+	adev->nbio.funcs->hdp_flush(adev, NULL);
 
 	memcpy(table, priv->smu_tables.entry[table_id].table,
 			priv->smu_tables.entry[table_id].size);
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/vega12_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/vega12_smumgr.c
index f9589806bf83..90c782c132d2 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/vega12_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/vega12_smumgr.c
@@ -66,7 +66,7 @@ static int vega12_copy_table_from_smc(struct pp_hwmgr *hwmgr,
 			return -EINVAL);
 
 	/* flush hdp cache */
-	adev->nbio_funcs->hdp_flush(adev, NULL);
+	adev->nbio.funcs->hdp_flush(adev, NULL);
 
 	memcpy(table, priv->smu_tables.entry[table_id].table,
 			priv->smu_tables.entry[table_id].size);
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/vega20_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/vega20_smumgr.c
index b9089c6bea85..f604612f411f 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/vega20_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/vega20_smumgr.c
@@ -189,7 +189,7 @@ static int vega20_copy_table_from_smc(struct pp_hwmgr *hwmgr,
 			return ret);
 
 	/* flush hdp cache */
-	adev->nbio_funcs->hdp_flush(adev, NULL);
+	adev->nbio.funcs->hdp_flush(adev, NULL);
 
 	memcpy(table, priv->smu_tables.entry[table_id].table,
 			priv->smu_tables.entry[table_id].size);
@@ -290,7 +290,7 @@ int vega20_get_activity_monitor_coeff(struct pp_hwmgr *hwmgr,
 			return ret);
 
 	/* flush hdp cache */
-	adev->nbio_funcs->hdp_flush(adev, NULL);
+	adev->nbio.funcs->hdp_flush(adev, NULL);
 
 	memcpy(table, priv->smu_tables.entry[TABLE_ACTIVITY_MONITOR_COEFF].table,
 			priv->smu_tables.entry[TABLE_ACTIVITY_MONITOR_COEFF].size);
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/vegam_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/vegam_smumgr.c
index 7c960b07746f..ae18fbcb26fb 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/vegam_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/vegam_smumgr.c
@@ -456,7 +456,7 @@ static int vegam_populate_smc_mvdd_table(struct pp_hwmgr *hwmgr,
 			count = SMU_MAX_SMIO_LEVELS;
 		for (level = 0; level < count; level++) {
 			table->SmioTable2.Pattern[level].Voltage = PP_HOST_TO_SMC_US(
-					data->mvdd_voltage_table.entries[count].value * VOLTAGE_SCALE);
+					data->mvdd_voltage_table.entries[level].value * VOLTAGE_SCALE);
 			/* Index into DpmTable.Smio. Drive bits from Smio entry to get this voltage level.*/
 			table->SmioTable2.Pattern[level].Smio =
 				(uint8_t) level;
diff --git a/drivers/gpu/drm/amd/powerplay/vega20_ppt.c b/drivers/gpu/drm/amd/powerplay/vega20_ppt.c
index 64386ee3f878..0b4892833808 100644
--- a/drivers/gpu/drm/amd/powerplay/vega20_ppt.c
+++ b/drivers/gpu/drm/amd/powerplay/vega20_ppt.c
@@ -25,6 +25,7 @@
 #include <linux/firmware.h>
 #include "amdgpu.h"
 #include "amdgpu_smu.h"
+#include "smu_internal.h"
 #include "atomfirmware.h"
 #include "amdgpu_atomfirmware.h"
 #include "smu_v11_0.h"
@@ -143,6 +144,7 @@ static struct smu_11_0_cmn2aisc_mapping vega20_message_map[SMU_MSG_MAX_COUNT] =
 	MSG_MAP(PrepareMp1ForShutdown),
 	MSG_MAP(SetMGpuFanBoostLimitRpm),
 	MSG_MAP(GetAVFSVoltageByDpm),
+	MSG_MAP(DFCstateControl),
 };
 
 static struct smu_11_0_cmn2aisc_mapping vega20_clk_map[SMU_CLK_COUNT] = {
@@ -219,7 +221,7 @@ static struct smu_11_0_cmn2aisc_mapping vega20_workload_map[PP_SMC_POWER_PROFILE
 	WORKLOAD_MAP(PP_SMC_POWER_PROFILE_POWERSAVING,		WORKLOAD_PPLIB_POWER_SAVING_BIT),
 	WORKLOAD_MAP(PP_SMC_POWER_PROFILE_VIDEO,		WORKLOAD_PPLIB_VIDEO_BIT),
 	WORKLOAD_MAP(PP_SMC_POWER_PROFILE_VR,			WORKLOAD_PPLIB_VR_BIT),
-	WORKLOAD_MAP(PP_SMC_POWER_PROFILE_COMPUTE,		WORKLOAD_PPLIB_CUSTOM_BIT),
+	WORKLOAD_MAP(PP_SMC_POWER_PROFILE_COMPUTE,		WORKLOAD_PPLIB_COMPUTE_BIT),
 	WORKLOAD_MAP(PP_SMC_POWER_PROFILE_CUSTOM,		WORKLOAD_PPLIB_CUSTOM_BIT),
 };
 
@@ -464,7 +466,6 @@ static int vega20_store_powerplay_table(struct smu_context *smu)
 	       sizeof(PPTable_t));
 
 	table_context->thermal_controller_type = powerplay_table->ucThermalControllerType;
-	table_context->TDPODLimit = le32_to_cpu(powerplay_table->OverDrive8Table.ODSettingsMax[ATOM_VEGA20_ODSETTING_POWERPERCENTAGE]);
 
 	return 0;
 }
@@ -634,7 +635,6 @@ amd_pm_state_type vega20_get_current_power_state(struct smu_context *smu)
 	    !smu_dpm_ctx->dpm_current_power_state)
 		return -EINVAL;
 
-	mutex_lock(&(smu->mutex));
 	switch (smu_dpm_ctx->dpm_current_power_state->classification.ui_label) {
 	case SMU_STATE_UI_LABEL_BATTERY:
 		pm_type = POWER_STATE_TYPE_BATTERY;
@@ -652,7 +652,6 @@ amd_pm_state_type vega20_get_current_power_state(struct smu_context *smu)
 			pm_type = POWER_STATE_TYPE_DEFAULT;
 		break;
 	}
-	mutex_unlock(&(smu->mutex));
 
 	return pm_type;
 }
@@ -1274,16 +1273,8 @@ static int vega20_force_clk_levels(struct smu_context *smu,
 	struct vega20_dpm_table *dpm_table;
 	struct vega20_single_dpm_table *single_dpm_table;
 	uint32_t soft_min_level, soft_max_level, hard_min_level;
-	struct smu_dpm_context *smu_dpm = &smu->smu_dpm;
 	int ret = 0;
 
-	if (smu_dpm->dpm_level != AMD_DPM_FORCED_LEVEL_MANUAL) {
-		pr_info("force clock level is for dpm manual mode only.\n");
-		return -EINVAL;
-	}
-
-	mutex_lock(&(smu->mutex));
-
 	soft_min_level = mask ? (ffs(mask) - 1) : 0;
 	soft_max_level = mask ? (fls(mask) - 1) : 0;
 
@@ -1436,7 +1427,6 @@ static int vega20_force_clk_levels(struct smu_context *smu,
 		break;
 	}
 
-	mutex_unlock(&(smu->mutex));
 	return ret;
 }
 
@@ -1451,8 +1441,6 @@ static int vega20_get_clock_by_type_with_latency(struct smu_context *smu,
 
 	dpm_table = smu_dpm->dpm_context;
 
-	mutex_lock(&smu->mutex);
-
 	switch (clk_type) {
 	case SMU_GFXCLK:
 		single_dpm_table = &(dpm_table->gfx_table);
@@ -1474,7 +1462,6 @@ static int vega20_get_clock_by_type_with_latency(struct smu_context *smu,
 		ret = -EINVAL;
 	}
 
-	mutex_unlock(&smu->mutex);
 	return ret;
 }
 
@@ -2260,7 +2247,7 @@ vega20_notify_smc_dispaly_config(struct smu_context *smu)
 	if (smu_feature_is_supported(smu, SMU_FEATURE_DPM_DCEFCLK_BIT)) {
 		clock_req.clock_type = amd_pp_dcef_clock;
 		clock_req.clock_freq_in_khz = min_clocks.dcef_clock * 10;
-		if (!smu->funcs->display_clock_voltage_request(smu, &clock_req)) {
+		if (!smu_v11_0_display_clock_voltage_request(smu, &clock_req)) {
 			if (smu_feature_is_supported(smu, SMU_FEATURE_DS_DCEFCLK_BIT)) {
 				ret = smu_send_smc_msg_with_param(smu,
 								  SMU_MSG_SetMinDeepSleepDcefclk,
@@ -2547,8 +2534,6 @@ static int vega20_set_od_percentage(struct smu_context *smu,
 	int feature_enabled;
 	PPCLK_e clk_id;
 
-	mutex_lock(&(smu->mutex));
-
 	dpm_table = smu_dpm->dpm_context;
 	golden_table = smu_dpm->golden_dpm_context;
 
@@ -2598,11 +2583,10 @@ static int vega20_set_od_percentage(struct smu_context *smu,
 	}
 
 	ret = smu_handle_task(smu, smu_dpm->dpm_level,
-			      AMD_PP_TASK_READJUST_POWER_STATE);
+			      AMD_PP_TASK_READJUST_POWER_STATE,
+			      false);
 
 set_od_failed:
-	mutex_unlock(&(smu->mutex));
-
 	return ret;
 }
 
@@ -2827,10 +2811,9 @@ static int vega20_odn_edit_dpm_table(struct smu_context *smu,
 	}
 
 	if (type == PP_OD_COMMIT_DPM_TABLE) {
-		mutex_lock(&(smu->mutex));
 		ret = smu_handle_task(smu, smu_dpm->dpm_level,
-				      AMD_PP_TASK_READJUST_POWER_STATE);
-		mutex_unlock(&(smu->mutex));
+				      AMD_PP_TASK_READJUST_POWER_STATE,
+				      false);
 	}
 
 	return ret;
@@ -3023,6 +3006,7 @@ static int vega20_read_sensor(struct smu_context *smu,
 	if(!data || !size)
 		return -EINVAL;
 
+	mutex_lock(&smu->sensor_lock);
 	switch (sensor) {
 	case AMDGPU_PP_SENSOR_MAX_FAN_RPM:
 		*(uint32_t *)data = pptable->FanMaximumRpm;
@@ -3046,8 +3030,9 @@ static int vega20_read_sensor(struct smu_context *smu,
 		*size = 4;
 		break;
 	default:
-		ret = smu_smc_read_sensor(smu, sensor, data, size);
+		ret = smu_v11_0_read_sensor(smu, sensor, data, size);
 	}
+	mutex_unlock(&smu->sensor_lock);
 
 	return ret;
 }
@@ -3139,6 +3124,49 @@ static int vega20_get_thermal_temperature_range(struct smu_context *smu,
 	return 0;
 }
 
+static int vega20_set_df_cstate(struct smu_context *smu,
+				enum pp_df_cstate state)
+{
+	uint32_t smu_version;
+	int ret;
+
+	ret = smu_get_smc_version(smu, NULL, &smu_version);
+	if (ret) {
+		pr_err("Failed to get smu version!\n");
+		return ret;
+	}
+
+	/* PPSMC_MSG_DFCstateControl is supported with 40.50 and later fws */
+	if (smu_version < 0x283200) {
+		pr_err("Df cstate control is supported with 40.50 and later SMC fw!\n");
+		return -EINVAL;
+	}
+
+	return smu_send_smc_msg_with_param(smu, SMU_MSG_DFCstateControl, state);
+}
+
+static int vega20_update_pcie_parameters(struct smu_context *smu,
+				     uint32_t pcie_gen_cap,
+				     uint32_t pcie_width_cap)
+{
+	PPTable_t *pptable = smu->smu_table.driver_pptable;
+	int ret, i;
+	uint32_t smu_pcie_arg;
+
+	for (i = 0; i < NUM_LINK_LEVELS; i++) {
+		smu_pcie_arg = (i << 16) |
+			((pptable->PcieGenSpeed[i] <= pcie_gen_cap) ? (pptable->PcieGenSpeed[i] << 8) :
+				(pcie_gen_cap << 8)) | ((pptable->PcieLaneCount[i] <= pcie_width_cap) ?
+					pptable->PcieLaneCount[i] : pcie_width_cap);
+		ret = smu_send_smc_msg_with_param(smu,
+					  SMU_MSG_OverridePcieParameters,
+					  smu_pcie_arg);
+	}
+
+	return ret;
+}
+
+
 static const struct pptable_funcs vega20_ppt_funcs = {
 	.tables_init = vega20_tables_init,
 	.alloc_dpm_context = vega20_allocate_dpm_context,
@@ -3151,7 +3179,7 @@ static const struct pptable_funcs vega20_ppt_funcs = {
 	.get_smu_table_index = vega20_get_smu_table_index,
 	.get_smu_power_index = vega20_get_pwr_src_index,
 	.get_workload_type = vega20_get_workload_type,
-	.run_afll_btc = vega20_run_btc_afll,
+	.run_btc = vega20_run_btc_afll,
 	.get_allowed_feature_mask = vega20_get_allowed_feature_mask,
 	.get_current_power_state = vega20_get_current_power_state,
 	.set_default_dpm_table = vega20_set_default_dpm_table,
@@ -3181,13 +3209,61 @@ static const struct pptable_funcs vega20_ppt_funcs = {
 	.get_fan_speed_percent = vega20_get_fan_speed_percent,
 	.get_fan_speed_rpm = vega20_get_fan_speed_rpm,
 	.set_watermarks_table = vega20_set_watermarks_table,
-	.get_thermal_temperature_range = vega20_get_thermal_temperature_range
+	.get_thermal_temperature_range = vega20_get_thermal_temperature_range,
+	.set_df_cstate = vega20_set_df_cstate,
+	.update_pcie_parameters = vega20_update_pcie_parameters,
+	.init_microcode = smu_v11_0_init_microcode,
+	.load_microcode = smu_v11_0_load_microcode,
+	.init_smc_tables = smu_v11_0_init_smc_tables,
+	.fini_smc_tables = smu_v11_0_fini_smc_tables,
+	.init_power = smu_v11_0_init_power,
+	.fini_power = smu_v11_0_fini_power,
+	.check_fw_status = smu_v11_0_check_fw_status,
+	.setup_pptable = smu_v11_0_setup_pptable,
+	.get_vbios_bootup_values = smu_v11_0_get_vbios_bootup_values,
+	.get_clk_info_from_vbios = smu_v11_0_get_clk_info_from_vbios,
+	.check_pptable = smu_v11_0_check_pptable,
+	.parse_pptable = smu_v11_0_parse_pptable,
+	.populate_smc_tables = smu_v11_0_populate_smc_pptable,
+	.check_fw_version = smu_v11_0_check_fw_version,
+	.write_pptable = smu_v11_0_write_pptable,
+	.set_min_dcef_deep_sleep = smu_v11_0_set_min_dcef_deep_sleep,
+	.set_tool_table_location = smu_v11_0_set_tool_table_location,
+	.notify_memory_pool_location = smu_v11_0_notify_memory_pool_location,
+	.system_features_control = smu_v11_0_system_features_control,
+	.send_smc_msg = smu_v11_0_send_msg,
+	.send_smc_msg_with_param = smu_v11_0_send_msg_with_param,
+	.read_smc_arg = smu_v11_0_read_arg,
+	.init_display_count = smu_v11_0_init_display_count,
+	.set_allowed_mask = smu_v11_0_set_allowed_mask,
+	.get_enabled_mask = smu_v11_0_get_enabled_mask,
+	.notify_display_change = smu_v11_0_notify_display_change,
+	.set_power_limit = smu_v11_0_set_power_limit,
+	.get_current_clk_freq = smu_v11_0_get_current_clk_freq,
+	.init_max_sustainable_clocks = smu_v11_0_init_max_sustainable_clocks,
+	.start_thermal_control = smu_v11_0_start_thermal_control,
+	.stop_thermal_control = smu_v11_0_stop_thermal_control,
+	.set_deep_sleep_dcefclk = smu_v11_0_set_deep_sleep_dcefclk,
+	.display_clock_voltage_request = smu_v11_0_display_clock_voltage_request,
+	.get_fan_control_mode = smu_v11_0_get_fan_control_mode,
+	.set_fan_control_mode = smu_v11_0_set_fan_control_mode,
+	.set_fan_speed_percent = smu_v11_0_set_fan_speed_percent,
+	.set_fan_speed_rpm = smu_v11_0_set_fan_speed_rpm,
+	.set_xgmi_pstate = smu_v11_0_set_xgmi_pstate,
+	.gfx_off_control = smu_v11_0_gfx_off_control,
+	.register_irq_handler = smu_v11_0_register_irq_handler,
+	.set_azalia_d3_pme = smu_v11_0_set_azalia_d3_pme,
+	.get_max_sustainable_clocks_by_dc = smu_v11_0_get_max_sustainable_clocks_by_dc,
+	.baco_is_support= smu_v11_0_baco_is_support,
+	.baco_get_state = smu_v11_0_baco_get_state,
+	.baco_set_state = smu_v11_0_baco_set_state,
+	.baco_reset = smu_v11_0_baco_reset,
+	.get_dpm_ultimate_freq = smu_v11_0_get_dpm_ultimate_freq,
+	.set_soft_freq_limited_range = smu_v11_0_set_soft_freq_limited_range,
+	.override_pcie_parameters = smu_v11_0_override_pcie_parameters,
 };
 
 void vega20_set_ppt_funcs(struct smu_context *smu)
 {
-	struct smu_table_context *smu_table = &smu->smu_table;
-
 	smu->ppt_funcs = &vega20_ppt_funcs;
-	smu_table->table_count = TABLE_COUNT;
 }
diff --git a/drivers/gpu/drm/arc/arcpgu_drv.c b/drivers/gpu/drm/arc/arcpgu_drv.c
index 6b7f791685ec..d6a6692db0ac 100644
--- a/drivers/gpu/drm/arc/arcpgu_drv.c
+++ b/drivers/gpu/drm/arc/arcpgu_drv.c
@@ -14,6 +14,7 @@
 #include <drm/drm_fb_helper.h>
 #include <drm/drm_gem_cma_helper.h>
 #include <drm/drm_gem_framebuffer_helper.h>
+#include <drm/drm_of.h>
 #include <drm/drm_probe_helper.h>
 #include <linux/dma-mapping.h>
 #include <linux/module.h>
@@ -45,7 +46,7 @@ static int arcpgu_load(struct drm_device *drm)
 {
 	struct platform_device *pdev = to_platform_device(drm->dev);
 	struct arcpgu_drm_private *arcpgu;
-	struct device_node *encoder_node;
+	struct device_node *encoder_node = NULL, *endpoint_node = NULL;
 	struct resource *res;
 	int ret;
 
@@ -80,14 +81,23 @@ static int arcpgu_load(struct drm_device *drm)
 	if (arc_pgu_setup_crtc(drm) < 0)
 		return -ENODEV;
 
-	/* find the encoder node and initialize it */
-	encoder_node = of_parse_phandle(drm->dev->of_node, "encoder-slave", 0);
+	/*
+	 * There is only one output port inside each device. It is linked with
+	 * encoder endpoint.
+	 */
+	endpoint_node = of_graph_get_next_endpoint(pdev->dev.of_node, NULL);
+	if (endpoint_node) {
+		encoder_node = of_graph_get_remote_port_parent(endpoint_node);
+		of_node_put(endpoint_node);
+	}
+
 	if (encoder_node) {
 		ret = arcpgu_drm_hdmi_init(drm, encoder_node);
 		of_node_put(encoder_node);
 		if (ret < 0)
 			return ret;
 	} else {
+		dev_info(drm->dev, "no encoder found. Assumed virtual LCD on simulation platform\n");
 		ret = arcpgu_drm_sim_init(drm, NULL);
 		if (ret < 0)
 			return ret;
diff --git a/drivers/gpu/drm/arc/arcpgu_hdmi.c b/drivers/gpu/drm/arc/arcpgu_hdmi.c
index 98aac743cc26..8fd7094beece 100644
--- a/drivers/gpu/drm/arc/arcpgu_hdmi.c
+++ b/drivers/gpu/drm/arc/arcpgu_hdmi.c
@@ -5,6 +5,7 @@
  * Copyright (C) 2016 Synopsys, Inc. (www.synopsys.com)
  */
 
+#include <drm/drm_bridge.h>
 #include <drm/drm_crtc.h>
 #include <drm/drm_encoder.h>
 #include <drm/drm_device.h>
diff --git a/drivers/gpu/drm/arm/display/Kconfig b/drivers/gpu/drm/arm/display/Kconfig
index cec0639e3aa1..e87ff8623076 100644
--- a/drivers/gpu/drm/arm/display/Kconfig
+++ b/drivers/gpu/drm/arm/display/Kconfig
@@ -12,3 +12,9 @@ config DRM_KOMEDA
 	  Processor driver. It supports the D71 variants of the hardware.
 
 	  If compiled as a module it will be called komeda.
+
+config DRM_KOMEDA_ERROR_PRINT
+	bool "Enable komeda error print"
+	depends on DRM_KOMEDA
+	help
+	  Choose this option to enable error printing.
diff --git a/drivers/gpu/drm/arm/display/komeda/Makefile b/drivers/gpu/drm/arm/display/komeda/Makefile
index 5c3900c2e764..f095a1c68ac7 100644
--- a/drivers/gpu/drm/arm/display/komeda/Makefile
+++ b/drivers/gpu/drm/arm/display/komeda/Makefile
@@ -22,4 +22,6 @@ komeda-y += \
 	d71/d71_dev.o \
 	d71/d71_component.o
 
+komeda-$(CONFIG_DRM_KOMEDA_ERROR_PRINT) += komeda_event.o
+
 obj-$(CONFIG_DRM_KOMEDA) += komeda.o
diff --git a/drivers/gpu/drm/arm/display/komeda/d71/d71_component.c b/drivers/gpu/drm/arm/display/komeda/d71/d71_component.c
index 55a8cc94808a..f0ba26e282c3 100644
--- a/drivers/gpu/drm/arm/display/komeda/d71/d71_component.c
+++ b/drivers/gpu/drm/arm/display/komeda/d71/d71_component.c
@@ -106,6 +106,23 @@ static void dump_block_header(struct seq_file *sf, void __iomem *reg)
 			   i, hdr.output_ids[i]);
 }
 
+/* On D71, we are using the global line size. From D32, every component have
+ * a line size register to indicate the fifo size.
+ */
+static u32 __get_blk_line_size(struct d71_dev *d71, u32 __iomem *reg,
+			       u32 max_default)
+{
+	if (!d71->periph_addr)
+		max_default = malidp_read32(reg, BLK_MAX_LINE_SIZE);
+
+	return max_default;
+}
+
+static u32 get_blk_line_size(struct d71_dev *d71, u32 __iomem *reg)
+{
+	return __get_blk_line_size(d71, reg, d71->max_line_size);
+}
+
 static u32 to_rot_ctrl(u32 rot)
 {
 	u32 lr_ctrl = 0;
@@ -332,7 +349,56 @@ static void d71_layer_dump(struct komeda_component *c, struct seq_file *sf)
 	seq_printf(sf, "%sAD_V_CROP:\t\t0x%X\n", prefix, v[2]);
 }
 
+static int d71_layer_validate(struct komeda_component *c,
+			      struct komeda_component_state *state)
+{
+	struct komeda_layer_state *st = to_layer_st(state);
+	struct komeda_layer *layer = to_layer(c);
+	struct drm_plane_state *plane_st;
+	struct drm_framebuffer *fb;
+	u32 fourcc, line_sz, max_line_sz;
+
+	plane_st = drm_atomic_get_new_plane_state(state->obj.state,
+						  state->plane);
+	fb = plane_st->fb;
+	fourcc = fb->format->format;
+
+	if (drm_rotation_90_or_270(st->rot))
+		line_sz = st->vsize - st->afbc_crop_t - st->afbc_crop_b;
+	else
+		line_sz = st->hsize - st->afbc_crop_l - st->afbc_crop_r;
+
+	if (fb->modifier) {
+		if ((fb->modifier & AFBC_FORMAT_MOD_BLOCK_SIZE_MASK) ==
+			AFBC_FORMAT_MOD_BLOCK_SIZE_32x8)
+			max_line_sz = layer->line_sz;
+		else
+			max_line_sz = layer->line_sz / 2;
+
+		if (line_sz > max_line_sz) {
+			DRM_DEBUG_ATOMIC("afbc request line_sz: %d exceed the max afbc line_sz: %d.\n",
+					 line_sz, max_line_sz);
+			return -EINVAL;
+		}
+	}
+
+	if (fourcc == DRM_FORMAT_YUV420_10BIT && line_sz > 2046 && (st->afbc_crop_l % 4)) {
+		DRM_DEBUG_ATOMIC("YUV420_10BIT input_hsize: %d exceed the max size 2046.\n",
+				 line_sz);
+		return -EINVAL;
+	}
+
+	if (fourcc == DRM_FORMAT_X0L2 && line_sz > 2046 && (st->addr[0] % 16)) {
+		DRM_DEBUG_ATOMIC("X0L2 input_hsize: %d exceed the max size 2046.\n",
+				 line_sz);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
 static const struct komeda_component_funcs d71_layer_funcs = {
+	.validate	= d71_layer_validate,
 	.update		= d71_layer_update,
 	.disable	= d71_layer_disable,
 	.dump_register	= d71_layer_dump,
@@ -365,7 +431,28 @@ static int d71_layer_init(struct d71_dev *d71,
 	else
 		layer->layer_type = KOMEDA_FMT_SIMPLE_LAYER;
 
-	set_range(&layer->hsize_in, 4, d71->max_line_size);
+	if (!d71->periph_addr) {
+		/* D32 or newer product */
+		layer->line_sz = malidp_read32(reg, BLK_MAX_LINE_SIZE);
+		layer->yuv_line_sz = L_INFO_YUV_MAX_LINESZ(layer_info);
+	} else if (d71->max_line_size > 2048) {
+		/* D71 4K */
+		layer->line_sz = d71->max_line_size;
+		layer->yuv_line_sz = layer->line_sz / 2;
+	} else	{
+		/* D71 2K */
+		if (layer->layer_type == KOMEDA_FMT_RICH_LAYER) {
+			/* rich layer is 4K configuration */
+			layer->line_sz = d71->max_line_size * 2;
+			layer->yuv_line_sz = layer->line_sz / 2;
+		} else {
+			layer->line_sz = d71->max_line_size;
+			layer->yuv_line_sz = 0;
+		}
+	}
+
+	set_range(&layer->hsize_in, 4, layer->line_sz);
+
 	set_range(&layer->vsize_in, 4, d71->max_vsize);
 
 	malidp_write32(reg, LAYER_PALPHA, D71_PALPHA_DEF_MAP);
@@ -456,9 +543,11 @@ static int d71_wb_layer_init(struct d71_dev *d71,
 
 	wb_layer = to_layer(c);
 	wb_layer->layer_type = KOMEDA_FMT_WB_LAYER;
+	wb_layer->line_sz = get_blk_line_size(d71, reg);
+	wb_layer->yuv_line_sz = wb_layer->line_sz;
 
-	set_range(&wb_layer->hsize_in, D71_MIN_LINE_SIZE, d71->max_line_size);
-	set_range(&wb_layer->vsize_in, D71_MIN_VERTICAL_SIZE, d71->max_vsize);
+	set_range(&wb_layer->hsize_in, 64, wb_layer->line_sz);
+	set_range(&wb_layer->vsize_in, 64, d71->max_vsize);
 
 	return 0;
 }
@@ -595,8 +684,8 @@ static int d71_compiz_init(struct d71_dev *d71,
 
 	compiz = to_compiz(c);
 
-	set_range(&compiz->hsize, D71_MIN_LINE_SIZE, d71->max_line_size);
-	set_range(&compiz->vsize, D71_MIN_VERTICAL_SIZE, d71->max_vsize);
+	set_range(&compiz->hsize, 64, get_blk_line_size(d71, reg));
+	set_range(&compiz->vsize, 64, d71->max_vsize);
 
 	return 0;
 }
@@ -703,7 +792,7 @@ static void d71_scaler_update(struct komeda_component *c,
 
 static void d71_scaler_dump(struct komeda_component *c, struct seq_file *sf)
 {
-	u32 v[9];
+	u32 v[10];
 
 	dump_block_header(sf, c->reg);
 
@@ -723,6 +812,18 @@ static void d71_scaler_dump(struct komeda_component *c, struct seq_file *sf)
 	seq_printf(sf, "SC_H_DELTA_PH:\t\t0x%X\n", v[6]);
 	seq_printf(sf, "SC_V_INIT_PH:\t\t0x%X\n", v[7]);
 	seq_printf(sf, "SC_V_DELTA_PH:\t\t0x%X\n", v[8]);
+
+	get_values_from_reg(c->reg, 0x130, 10, v);
+	seq_printf(sf, "SC_ENH_LIMITS:\t\t0x%X\n", v[0]);
+	seq_printf(sf, "SC_ENH_COEFF0:\t\t0x%X\n", v[1]);
+	seq_printf(sf, "SC_ENH_COEFF1:\t\t0x%X\n", v[2]);
+	seq_printf(sf, "SC_ENH_COEFF2:\t\t0x%X\n", v[3]);
+	seq_printf(sf, "SC_ENH_COEFF3:\t\t0x%X\n", v[4]);
+	seq_printf(sf, "SC_ENH_COEFF4:\t\t0x%X\n", v[5]);
+	seq_printf(sf, "SC_ENH_COEFF5:\t\t0x%X\n", v[6]);
+	seq_printf(sf, "SC_ENH_COEFF6:\t\t0x%X\n", v[7]);
+	seq_printf(sf, "SC_ENH_COEFF7:\t\t0x%X\n", v[8]);
+	seq_printf(sf, "SC_ENH_COEFF8:\t\t0x%X\n", v[9]);
 }
 
 static const struct komeda_component_funcs d71_scaler_funcs = {
@@ -753,7 +854,7 @@ static int d71_scaler_init(struct d71_dev *d71,
 	}
 
 	scaler = to_scaler(c);
-	set_range(&scaler->hsize, 4, 2048);
+	set_range(&scaler->hsize, 4, __get_blk_line_size(d71, reg, 2048));
 	set_range(&scaler->vsize, 4, 4096);
 	scaler->max_downscaling = 6;
 	scaler->max_upscaling = 64;
@@ -862,7 +963,7 @@ static int d71_splitter_init(struct d71_dev *d71,
 
 	splitter = to_splitter(c);
 
-	set_range(&splitter->hsize, 4, d71->max_line_size);
+	set_range(&splitter->hsize, 4, get_blk_line_size(d71, reg));
 	set_range(&splitter->vsize, 4, d71->max_vsize);
 
 	return 0;
@@ -933,7 +1034,8 @@ static int d71_merger_init(struct d71_dev *d71,
 
 	merger = to_merger(c);
 
-	set_range(&merger->hsize_merged, 4, 4032);
+	set_range(&merger->hsize_merged, 4,
+		  __get_blk_line_size(d71, reg, 4032));
 	set_range(&merger->vsize_merged, 4, 4096);
 
 	return 0;
@@ -944,13 +1046,26 @@ static void d71_improc_update(struct komeda_component *c,
 {
 	struct komeda_improc_state *st = to_improc_st(state);
 	u32 __iomem *reg = c->reg;
-	u32 index;
+	u32 index, mask = 0, ctrl = 0;
 
 	for_each_changed_input(state, index)
 		malidp_write32(reg, BLK_INPUT_ID0 + index * 4,
 			       to_d71_input_id(state, index));
 
 	malidp_write32(reg, BLK_SIZE, HV_SIZE(st->hsize, st->vsize));
+	malidp_write32(reg, IPS_DEPTH, st->color_depth);
+
+	mask |= IPS_CTRL_YUV | IPS_CTRL_CHD422 | IPS_CTRL_CHD420;
+
+	/* config color format */
+	if (st->color_format == DRM_COLOR_FORMAT_YCRCB420)
+		ctrl |= IPS_CTRL_YUV | IPS_CTRL_CHD422 | IPS_CTRL_CHD420;
+	else if (st->color_format == DRM_COLOR_FORMAT_YCRCB422)
+		ctrl |= IPS_CTRL_YUV | IPS_CTRL_CHD422;
+	else if (st->color_format == DRM_COLOR_FORMAT_YCRCB444)
+		ctrl |= IPS_CTRL_YUV;
+
+	malidp_write32_mask(reg, BLK_CONTROL, mask, ctrl);
 }
 
 static void d71_improc_dump(struct komeda_component *c, struct seq_file *sf)
@@ -1218,6 +1333,90 @@ int d71_probe_block(struct d71_dev *d71,
 	return err;
 }
 
+static void d71_gcu_dump(struct d71_dev *d71, struct seq_file *sf)
+{
+	u32 v[5];
+
+	seq_puts(sf, "\n------ GCU ------\n");
+
+	get_values_from_reg(d71->gcu_addr, 0, 3, v);
+	seq_printf(sf, "GLB_ARCH_ID:\t\t0x%X\n", v[0]);
+	seq_printf(sf, "GLB_CORE_ID:\t\t0x%X\n", v[1]);
+	seq_printf(sf, "GLB_CORE_INFO:\t\t0x%X\n", v[2]);
+
+	get_values_from_reg(d71->gcu_addr, 0x10, 1, v);
+	seq_printf(sf, "GLB_IRQ_STATUS:\t\t0x%X\n", v[0]);
+
+	get_values_from_reg(d71->gcu_addr, 0xA0, 5, v);
+	seq_printf(sf, "GCU_IRQ_RAW_STATUS:\t0x%X\n", v[0]);
+	seq_printf(sf, "GCU_IRQ_CLEAR:\t\t0x%X\n", v[1]);
+	seq_printf(sf, "GCU_IRQ_MASK:\t\t0x%X\n", v[2]);
+	seq_printf(sf, "GCU_IRQ_STATUS:\t\t0x%X\n", v[3]);
+	seq_printf(sf, "GCU_STATUS:\t\t0x%X\n", v[4]);
+
+	get_values_from_reg(d71->gcu_addr, 0xD0, 3, v);
+	seq_printf(sf, "GCU_CONTROL:\t\t0x%X\n", v[0]);
+	seq_printf(sf, "GCU_CONFIG_VALID0:\t0x%X\n", v[1]);
+	seq_printf(sf, "GCU_CONFIG_VALID1:\t0x%X\n", v[2]);
+}
+
+static void d71_lpu_dump(struct d71_pipeline *pipe, struct seq_file *sf)
+{
+	u32 v[6];
+
+	seq_printf(sf, "\n------ LPU%d ------\n", pipe->base.id);
+
+	dump_block_header(sf, pipe->lpu_addr);
+
+	get_values_from_reg(pipe->lpu_addr, 0xA0, 6, v);
+	seq_printf(sf, "LPU_IRQ_RAW_STATUS:\t0x%X\n", v[0]);
+	seq_printf(sf, "LPU_IRQ_CLEAR:\t\t0x%X\n", v[1]);
+	seq_printf(sf, "LPU_IRQ_MASK:\t\t0x%X\n", v[2]);
+	seq_printf(sf, "LPU_IRQ_STATUS:\t\t0x%X\n", v[3]);
+	seq_printf(sf, "LPU_STATUS:\t\t0x%X\n", v[4]);
+	seq_printf(sf, "LPU_TBU_STATUS:\t\t0x%X\n", v[5]);
+
+	get_values_from_reg(pipe->lpu_addr, 0xC0, 1, v);
+	seq_printf(sf, "LPU_INFO:\t\t0x%X\n", v[0]);
+
+	get_values_from_reg(pipe->lpu_addr, 0xD0, 3, v);
+	seq_printf(sf, "LPU_RAXI_CONTROL:\t0x%X\n", v[0]);
+	seq_printf(sf, "LPU_WAXI_CONTROL:\t0x%X\n", v[1]);
+	seq_printf(sf, "LPU_TBU_CONTROL:\t0x%X\n", v[2]);
+}
+
+static void d71_dou_dump(struct d71_pipeline *pipe, struct seq_file *sf)
+{
+	u32 v[5];
+
+	seq_printf(sf, "\n------ DOU%d ------\n", pipe->base.id);
+
+	dump_block_header(sf, pipe->dou_addr);
+
+	get_values_from_reg(pipe->dou_addr, 0xA0, 5, v);
+	seq_printf(sf, "DOU_IRQ_RAW_STATUS:\t0x%X\n", v[0]);
+	seq_printf(sf, "DOU_IRQ_CLEAR:\t\t0x%X\n", v[1]);
+	seq_printf(sf, "DOU_IRQ_MASK:\t\t0x%X\n", v[2]);
+	seq_printf(sf, "DOU_IRQ_STATUS:\t\t0x%X\n", v[3]);
+	seq_printf(sf, "DOU_STATUS:\t\t0x%X\n", v[4]);
+}
+
+static void d71_pipeline_dump(struct komeda_pipeline *pipe, struct seq_file *sf)
+{
+	struct d71_pipeline *d71_pipe = to_d71_pipeline(pipe);
+
+	d71_lpu_dump(d71_pipe, sf);
+	d71_dou_dump(d71_pipe, sf);
+}
+
 const struct komeda_pipeline_funcs d71_pipeline_funcs = {
-	.downscaling_clk_check = d71_downscaling_clk_check,
+	.downscaling_clk_check	= d71_downscaling_clk_check,
+	.dump_register		= d71_pipeline_dump,
 };
+
+void d71_dump(struct komeda_dev *mdev, struct seq_file *sf)
+{
+	struct d71_dev *d71 = mdev->chip_data;
+
+	d71_gcu_dump(d71, sf);
+}
diff --git a/drivers/gpu/drm/arm/display/komeda/d71/d71_dev.c b/drivers/gpu/drm/arm/display/komeda/d71/d71_dev.c
index d567ab7ed314..822b23a1ce75 100644
--- a/drivers/gpu/drm/arm/display/komeda/d71/d71_dev.c
+++ b/drivers/gpu/drm/arm/display/komeda/d71/d71_dev.c
@@ -195,7 +195,7 @@ d71_irq_handler(struct komeda_dev *mdev, struct komeda_events *evts)
 	if (gcu_status & GLB_IRQ_STATUS_PIPE1)
 		evts->pipes[1] |= get_pipeline_event(d71->pipes[1], gcu_status);
 
-	return gcu_status ? IRQ_HANDLED : IRQ_NONE;
+	return IRQ_RETVAL(gcu_status);
 }
 
 #define ENABLED_GCU_IRQS	(GCU_IRQ_CVAL0 | GCU_IRQ_CVAL1 | \
@@ -395,6 +395,22 @@ static int d71_enum_resources(struct komeda_dev *mdev)
 			err = PTR_ERR(pipe);
 			goto err_cleanup;
 		}
+
+		/* D71 HW doesn't update shadow registers when display output
+		 * is turning off, so when we disable all pipeline components
+		 * together with display output disable by one flush or one
+		 * operation, the disable operation updated registers will not
+		 * be flush to or valid in HW, which may leads problem.
+		 * To workaround this problem, introduce a two phase disable.
+		 * Phase1: Disabling components with display is on to make sure
+		 *	   the disable can be flushed to HW.
+		 * Phase2: Only turn-off display output.
+		 */
+		value = KOMEDA_PIPELINE_IMPROCS |
+			BIT(KOMEDA_COMPONENT_TIMING_CTRLR);
+
+		pipe->standalone_disabled_comps = value;
+
 		d71->pipes[i] = to_d71_pipeline(pipe);
 	}
 
@@ -561,17 +577,18 @@ static int d71_disconnect_iommu(struct komeda_dev *mdev)
 }
 
 static const struct komeda_dev_funcs d71_chip_funcs = {
-	.init_format_table = d71_init_fmt_tbl,
-	.enum_resources	= d71_enum_resources,
-	.cleanup	= d71_cleanup,
-	.irq_handler	= d71_irq_handler,
-	.enable_irq	= d71_enable_irq,
-	.disable_irq	= d71_disable_irq,
-	.on_off_vblank	= d71_on_off_vblank,
-	.change_opmode	= d71_change_opmode,
-	.flush		= d71_flush,
-	.connect_iommu	= d71_connect_iommu,
-	.disconnect_iommu = d71_disconnect_iommu,
+	.init_format_table	= d71_init_fmt_tbl,
+	.enum_resources		= d71_enum_resources,
+	.cleanup		= d71_cleanup,
+	.irq_handler		= d71_irq_handler,
+	.enable_irq		= d71_enable_irq,
+	.disable_irq		= d71_disable_irq,
+	.on_off_vblank		= d71_on_off_vblank,
+	.change_opmode		= d71_change_opmode,
+	.flush			= d71_flush,
+	.connect_iommu		= d71_connect_iommu,
+	.disconnect_iommu	= d71_disconnect_iommu,
+	.dump_register		= d71_dump,
 };
 
 const struct komeda_dev_funcs *
diff --git a/drivers/gpu/drm/arm/display/komeda/d71/d71_dev.h b/drivers/gpu/drm/arm/display/komeda/d71/d71_dev.h
index 84f1878b647d..c7357c2b9e62 100644
--- a/drivers/gpu/drm/arm/display/komeda/d71/d71_dev.h
+++ b/drivers/gpu/drm/arm/display/komeda/d71/d71_dev.h
@@ -49,4 +49,6 @@ int d71_probe_block(struct d71_dev *d71,
 		    struct block_header *blk, u32 __iomem *reg);
 void d71_read_block_header(u32 __iomem *reg, struct block_header *blk);
 
+void d71_dump(struct komeda_dev *mdev, struct seq_file *sf);
+
 #endif /* !_D71_DEV_H_ */
diff --git a/drivers/gpu/drm/arm/display/komeda/d71/d71_regs.h b/drivers/gpu/drm/arm/display/komeda/d71/d71_regs.h
index 2d5e6d00b42c..1727dc993909 100644
--- a/drivers/gpu/drm/arm/display/komeda/d71/d71_regs.h
+++ b/drivers/gpu/drm/arm/display/komeda/d71/d71_regs.h
@@ -10,6 +10,7 @@
 /* Common block registers offset */
 #define BLK_BLOCK_INFO		0x000
 #define BLK_PIPELINE_INFO	0x004
+#define BLK_MAX_LINE_SIZE	0x008
 #define BLK_VALID_INPUT_ID0	0x020
 #define BLK_OUTPUT_ID0		0x060
 #define BLK_INPUT_ID0		0x080
@@ -321,6 +322,7 @@
 #define L_INFO_RF		BIT(0)
 #define L_INFO_CM		BIT(1)
 #define L_INFO_ABUF_SIZE(x)	(((x) >> 4) & 0x7)
+#define L_INFO_YUV_MAX_LINESZ(x)	(((x) >> 16) & 0xFFFF)
 
 /* Scaler registers */
 #define SC_COEFFTAB		0x0DC
@@ -494,13 +496,6 @@ enum d71_blk_type {
 #define D71_DEFAULT_PREPRETCH_LINE	5
 #define D71_BUS_WIDTH_16_BYTES		16
 
-#define D71_MIN_LINE_SIZE		64
-#define D71_MIN_VERTICAL_SIZE		64
-#define D71_SC_MIN_LIN_SIZE		4
-#define D71_SC_MIN_VERTICAL_SIZE	4
-#define D71_SC_MAX_LIN_SIZE		2048
-#define D71_SC_MAX_VERTICAL_SIZE	4096
-
 #define D71_SC_MAX_UPSCALING		64
 #define D71_SC_MAX_DOWNSCALING		6
 #define D71_SC_SPLIT_OVERLAP		8
diff --git a/drivers/gpu/drm/arm/display/komeda/komeda_crtc.c b/drivers/gpu/drm/arm/display/komeda/komeda_crtc.c
index 624d257da20f..252015210fbc 100644
--- a/drivers/gpu/drm/arm/display/komeda/komeda_crtc.c
+++ b/drivers/gpu/drm/arm/display/komeda/komeda_crtc.c
@@ -5,7 +5,6 @@
  *
  */
 #include <linux/clk.h>
-#include <linux/pm_runtime.h>
 #include <linux/spinlock.h>
 
 #include <drm/drm_atomic.h>
@@ -18,6 +17,33 @@
 #include "komeda_dev.h"
 #include "komeda_kms.h"
 
+void komeda_crtc_get_color_config(struct drm_crtc_state *crtc_st,
+				  u32 *color_depths, u32 *color_formats)
+{
+	struct drm_connector *conn;
+	struct drm_connector_state *conn_st;
+	u32 conn_color_formats = ~0u;
+	int i, min_bpc = 31, conn_bpc = 0;
+
+	for_each_new_connector_in_state(crtc_st->state, conn, conn_st, i) {
+		if (conn_st->crtc != crtc_st->crtc)
+			continue;
+
+		conn_bpc = conn->display_info.bpc ? conn->display_info.bpc : 8;
+		conn_color_formats &= conn->display_info.color_formats;
+
+		if (conn_bpc < min_bpc)
+			min_bpc = conn_bpc;
+	}
+
+	/* connector doesn't config any color_format, use RGB444 as default */
+	if (!conn_color_formats)
+		conn_color_formats = DRM_COLOR_FORMAT_RGB444;
+
+	*color_depths = GENMASK(min_bpc, 0);
+	*color_formats = conn_color_formats;
+}
+
 static void komeda_crtc_update_clock_ratio(struct komeda_crtc_state *kcrtc_st)
 {
 	u64 pxlclk, aclk;
@@ -250,23 +276,57 @@ komeda_crtc_atomic_enable(struct drm_crtc *crtc,
 {
 	komeda_crtc_prepare(to_kcrtc(crtc));
 	drm_crtc_vblank_on(crtc);
+	WARN_ON(drm_crtc_vblank_get(crtc));
 	komeda_crtc_do_flush(crtc, old);
 }
 
 static void
+komeda_crtc_flush_and_wait_for_flip_done(struct komeda_crtc *kcrtc,
+					 struct completion *input_flip_done)
+{
+	struct drm_device *drm = kcrtc->base.dev;
+	struct komeda_dev *mdev = kcrtc->master->mdev;
+	struct completion *flip_done;
+	struct completion temp;
+	int timeout;
+
+	/* if caller doesn't send a flip_done, use a private flip_done */
+	if (input_flip_done) {
+		flip_done = input_flip_done;
+	} else {
+		init_completion(&temp);
+		kcrtc->disable_done = &temp;
+		flip_done = &temp;
+	}
+
+	mdev->funcs->flush(mdev, kcrtc->master->id, 0);
+
+	/* wait the flip take affect.*/
+	timeout = wait_for_completion_timeout(flip_done, HZ);
+	if (timeout == 0) {
+		DRM_ERROR("wait pipe%d flip done timeout\n", kcrtc->master->id);
+		if (!input_flip_done) {
+			unsigned long flags;
+
+			spin_lock_irqsave(&drm->event_lock, flags);
+			kcrtc->disable_done = NULL;
+			spin_unlock_irqrestore(&drm->event_lock, flags);
+		}
+	}
+}
+
+static void
 komeda_crtc_atomic_disable(struct drm_crtc *crtc,
 			   struct drm_crtc_state *old)
 {
 	struct komeda_crtc *kcrtc = to_kcrtc(crtc);
 	struct komeda_crtc_state *old_st = to_kcrtc_st(old);
-	struct komeda_dev *mdev = crtc->dev->dev_private;
 	struct komeda_pipeline *master = kcrtc->master;
 	struct komeda_pipeline *slave  = kcrtc->slave;
-	struct completion *disable_done = &crtc->state->commit->flip_done;
-	struct completion temp;
-	int timeout;
+	struct completion *disable_done;
+	bool needs_phase2 = false;
 
-	DRM_DEBUG_ATOMIC("CRTC%d_DISABLE: active_pipes: 0x%x, affected: 0x%x.\n",
+	DRM_DEBUG_ATOMIC("CRTC%d_DISABLE: active_pipes: 0x%x, affected: 0x%x\n",
 			 drm_crtc_index(crtc),
 			 old_st->active_pipes, old_st->affected_pipes);
 
@@ -274,7 +334,7 @@ komeda_crtc_atomic_disable(struct drm_crtc *crtc,
 		komeda_pipeline_disable(slave, old->state);
 
 	if (has_bit(master->id, old_st->active_pipes))
-		komeda_pipeline_disable(master, old->state);
+		needs_phase2 = komeda_pipeline_disable(master, old->state);
 
 	/* crtc_disable has two scenarios according to the state->active switch.
 	 * 1. active -> inactive
@@ -293,32 +353,23 @@ komeda_crtc_atomic_disable(struct drm_crtc *crtc,
 	 *    That's also the reason why skip modeset commit in
 	 *    komeda_crtc_atomic_flush()
 	 */
-	if (crtc->state->active) {
-		struct komeda_pipeline_state *pipe_st;
-		/* clear the old active_comps to zero */
-		pipe_st = komeda_pipeline_get_old_state(master, old->state);
-		pipe_st->active_comps = 0;
+	disable_done = (needs_phase2 || crtc->state->active) ?
+		       NULL : &crtc->state->commit->flip_done;
 
-		init_completion(&temp);
-		kcrtc->disable_done = &temp;
-		disable_done = &temp;
-	}
+	/* wait phase 1 disable done */
+	komeda_crtc_flush_and_wait_for_flip_done(kcrtc, disable_done);
 
-	mdev->funcs->flush(mdev, master->id, 0);
+	/* phase 2 */
+	if (needs_phase2) {
+		komeda_pipeline_disable(kcrtc->master, old->state);
 
-	/* wait the disable take affect.*/
-	timeout = wait_for_completion_timeout(disable_done, HZ);
-	if (timeout == 0) {
-		DRM_ERROR("disable pipeline%d timeout.\n", kcrtc->master->id);
-		if (crtc->state->active) {
-			unsigned long flags;
+		disable_done = crtc->state->active ?
+			       NULL : &crtc->state->commit->flip_done;
 
-			spin_lock_irqsave(&crtc->dev->event_lock, flags);
-			kcrtc->disable_done = NULL;
-			spin_unlock_irqrestore(&crtc->dev->event_lock, flags);
-		}
+		komeda_crtc_flush_and_wait_for_flip_done(kcrtc, disable_done);
 	}
 
+	drm_crtc_vblank_put(crtc);
 	drm_crtc_vblank_off(crtc);
 	komeda_crtc_unprepare(kcrtc);
 }
diff --git a/drivers/gpu/drm/arm/display/komeda/komeda_dev.c b/drivers/gpu/drm/arm/display/komeda/komeda_dev.c
index ca64a129c594..937a6d4c4865 100644
--- a/drivers/gpu/drm/arm/display/komeda/komeda_dev.c
+++ b/drivers/gpu/drm/arm/display/komeda/komeda_dev.c
@@ -25,6 +25,8 @@ static int komeda_register_show(struct seq_file *sf, void *x)
 	struct komeda_dev *mdev = sf->private;
 	int i;
 
+	seq_puts(sf, "\n====== Komeda register dump =========\n");
+
 	if (mdev->funcs->dump_register)
 		mdev->funcs->dump_register(mdev, sf);
 
@@ -91,9 +93,19 @@ config_id_show(struct device *dev, struct device_attribute *attr, char *buf)
 }
 static DEVICE_ATTR_RO(config_id);
 
+static ssize_t
+aclk_hz_show(struct device *dev, struct device_attribute *attr, char *buf)
+{
+	struct komeda_dev *mdev = dev_to_mdev(dev);
+
+	return snprintf(buf, PAGE_SIZE, "%lu\n", clk_get_rate(mdev->aclk));
+}
+static DEVICE_ATTR_RO(aclk_hz);
+
 static struct attribute *komeda_sysfs_entries[] = {
 	&dev_attr_core_id.attr,
 	&dev_attr_config_id.attr,
+	&dev_attr_aclk_hz.attr,
 	NULL,
 };
 
@@ -216,7 +228,7 @@ struct komeda_dev *komeda_dev_create(struct device *dev)
 			  product->product_id,
 			  MALIDP_CORE_ID_PRODUCT_ID(mdev->chip.core_id));
 		err = -ENODEV;
-		goto err_cleanup;
+		goto disable_clk;
 	}
 
 	DRM_INFO("Found ARM Mali-D%x version r%dp%d\n",
@@ -229,19 +241,19 @@ struct komeda_dev *komeda_dev_create(struct device *dev)
 	err = mdev->funcs->enum_resources(mdev);
 	if (err) {
 		DRM_ERROR("enumerate display resource failed.\n");
-		goto err_cleanup;
+		goto disable_clk;
 	}
 
 	err = komeda_parse_dt(dev, mdev);
 	if (err) {
 		DRM_ERROR("parse device tree failed.\n");
-		goto err_cleanup;
+		goto disable_clk;
 	}
 
 	err = komeda_assemble_pipelines(mdev);
 	if (err) {
 		DRM_ERROR("assemble display pipelines failed.\n");
-		goto err_cleanup;
+		goto disable_clk;
 	}
 
 	dev->dma_parms = &mdev->dma_parms;
@@ -254,11 +266,14 @@ struct komeda_dev *komeda_dev_create(struct device *dev)
 	if (mdev->iommu && mdev->funcs->connect_iommu) {
 		err = mdev->funcs->connect_iommu(mdev);
 		if (err) {
+			DRM_ERROR("connect iommu failed.\n");
 			mdev->iommu = NULL;
-			goto err_cleanup;
+			goto disable_clk;
 		}
 	}
 
+	clk_disable_unprepare(mdev->aclk);
+
 	err = sysfs_create_group(&dev->kobj, &komeda_sysfs_attr_group);
 	if (err) {
 		DRM_ERROR("create sysfs group failed.\n");
@@ -271,6 +286,8 @@ struct komeda_dev *komeda_dev_create(struct device *dev)
 
 	return mdev;
 
+disable_clk:
+	clk_disable_unprepare(mdev->aclk);
 err_cleanup:
 	komeda_dev_destroy(mdev);
 	return ERR_PTR(err);
@@ -288,8 +305,12 @@ void komeda_dev_destroy(struct komeda_dev *mdev)
 	debugfs_remove_recursive(mdev->debugfs_root);
 #endif
 
+	if (mdev->aclk)
+		clk_prepare_enable(mdev->aclk);
+
 	if (mdev->iommu && mdev->funcs->disconnect_iommu)
-		mdev->funcs->disconnect_iommu(mdev);
+		if (mdev->funcs->disconnect_iommu(mdev))
+			DRM_ERROR("disconnect iommu failed.\n");
 	mdev->iommu = NULL;
 
 	for (i = 0; i < mdev->n_pipelines; i++) {
@@ -317,3 +338,47 @@ void komeda_dev_destroy(struct komeda_dev *mdev)
 
 	devm_kfree(dev, mdev);
 }
+
+int komeda_dev_resume(struct komeda_dev *mdev)
+{
+	int ret = 0;
+
+	clk_prepare_enable(mdev->aclk);
+
+	if (mdev->iommu && mdev->funcs->connect_iommu) {
+		ret = mdev->funcs->connect_iommu(mdev);
+		if (ret < 0) {
+			DRM_ERROR("connect iommu failed.\n");
+			goto disable_clk;
+		}
+	}
+
+	ret = mdev->funcs->enable_irq(mdev);
+
+disable_clk:
+	clk_disable_unprepare(mdev->aclk);
+
+	return ret;
+}
+
+int komeda_dev_suspend(struct komeda_dev *mdev)
+{
+	int ret = 0;
+
+	clk_prepare_enable(mdev->aclk);
+
+	if (mdev->iommu && mdev->funcs->disconnect_iommu) {
+		ret = mdev->funcs->disconnect_iommu(mdev);
+		if (ret < 0) {
+			DRM_ERROR("disconnect iommu failed.\n");
+			goto disable_clk;
+		}
+	}
+
+	ret = mdev->funcs->disable_irq(mdev);
+
+disable_clk:
+	clk_disable_unprepare(mdev->aclk);
+
+	return ret;
+}
diff --git a/drivers/gpu/drm/arm/display/komeda/komeda_dev.h b/drivers/gpu/drm/arm/display/komeda/komeda_dev.h
index d1c86b6174c8..414200233b64 100644
--- a/drivers/gpu/drm/arm/display/komeda/komeda_dev.h
+++ b/drivers/gpu/drm/arm/display/komeda/komeda_dev.h
@@ -40,6 +40,17 @@
 #define KOMEDA_ERR_TTNG			BIT_ULL(30)
 #define KOMEDA_ERR_TTF			BIT_ULL(31)
 
+#define KOMEDA_ERR_EVENTS	\
+	(KOMEDA_EVENT_URUN	| KOMEDA_EVENT_IBSY	| KOMEDA_EVENT_OVR |\
+	KOMEDA_ERR_TETO		| KOMEDA_ERR_TEMR	| KOMEDA_ERR_TITR |\
+	KOMEDA_ERR_CPE		| KOMEDA_ERR_CFGE	| KOMEDA_ERR_AXIE |\
+	KOMEDA_ERR_ACE0		| KOMEDA_ERR_ACE1	| KOMEDA_ERR_ACE2 |\
+	KOMEDA_ERR_ACE3		| KOMEDA_ERR_DRIFTTO	| KOMEDA_ERR_FRAMETO |\
+	KOMEDA_ERR_ZME		| KOMEDA_ERR_MERR	| KOMEDA_ERR_TCF |\
+	KOMEDA_ERR_TTNG		| KOMEDA_ERR_TTF)
+
+#define KOMEDA_WARN_EVENTS	KOMEDA_ERR_CSCE
+
 /* malidp device id */
 enum {
 	MALI_D71 = 0,
@@ -207,4 +218,13 @@ void komeda_dev_destroy(struct komeda_dev *mdev);
 
 struct komeda_dev *dev_to_mdev(struct device *dev);
 
+#ifdef CONFIG_DRM_KOMEDA_ERROR_PRINT
+void komeda_print_events(struct komeda_events *evts);
+#else
+static inline void komeda_print_events(struct komeda_events *evts) {}
+#endif
+
+int komeda_dev_resume(struct komeda_dev *mdev);
+int komeda_dev_suspend(struct komeda_dev *mdev);
+
 #endif /*_KOMEDA_DEV_H_*/
diff --git a/drivers/gpu/drm/arm/display/komeda/komeda_drv.c b/drivers/gpu/drm/arm/display/komeda/komeda_drv.c
index 69ace6f9055d..d6c2222c5d33 100644
--- a/drivers/gpu/drm/arm/display/komeda/komeda_drv.c
+++ b/drivers/gpu/drm/arm/display/komeda/komeda_drv.c
@@ -8,6 +8,7 @@
 #include <linux/kernel.h>
 #include <linux/platform_device.h>
 #include <linux/component.h>
+#include <linux/pm_runtime.h>
 #include <drm/drm_of.h>
 #include "komeda_dev.h"
 #include "komeda_kms.h"
@@ -136,13 +137,40 @@ static const struct of_device_id komeda_of_match[] = {
 
 MODULE_DEVICE_TABLE(of, komeda_of_match);
 
+static int __maybe_unused komeda_pm_suspend(struct device *dev)
+{
+	struct komeda_drv *mdrv = dev_get_drvdata(dev);
+	struct drm_device *drm = &mdrv->kms->base;
+	int res;
+
+	res = drm_mode_config_helper_suspend(drm);
+
+	komeda_dev_suspend(mdrv->mdev);
+
+	return res;
+}
+
+static int __maybe_unused komeda_pm_resume(struct device *dev)
+{
+	struct komeda_drv *mdrv = dev_get_drvdata(dev);
+	struct drm_device *drm = &mdrv->kms->base;
+
+	komeda_dev_resume(mdrv->mdev);
+
+	return drm_mode_config_helper_resume(drm);
+}
+
+static const struct dev_pm_ops komeda_pm_ops = {
+	SET_SYSTEM_SLEEP_PM_OPS(komeda_pm_suspend, komeda_pm_resume)
+};
+
 static struct platform_driver komeda_platform_driver = {
 	.probe	= komeda_platform_probe,
 	.remove	= komeda_platform_remove,
 	.driver	= {
 		.name = "komeda",
 		.of_match_table	= komeda_of_match,
-		.pm = NULL,
+		.pm = &komeda_pm_ops,
 	},
 };
 
diff --git a/drivers/gpu/drm/arm/display/komeda/komeda_event.c b/drivers/gpu/drm/arm/display/komeda/komeda_event.c
new file mode 100644
index 000000000000..a36fb86cc054
--- /dev/null
+++ b/drivers/gpu/drm/arm/display/komeda/komeda_event.c
@@ -0,0 +1,140 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * (C) COPYRIGHT 2019 ARM Limited. All rights reserved.
+ * Author: James.Qian.Wang <james.qian.wang@arm.com>
+ *
+ */
+#include <drm/drm_print.h>
+
+#include "komeda_dev.h"
+
+struct komeda_str {
+	char *str;
+	u32 sz;
+	u32 len;
+};
+
+/* return 0 on success,  < 0 on no space.
+ */
+static int komeda_sprintf(struct komeda_str *str, const char *fmt, ...)
+{
+	va_list args;
+	int num, free_sz;
+	int err;
+
+	free_sz = str->sz - str->len - 1;
+	if (free_sz <= 0)
+		return -ENOSPC;
+
+	va_start(args, fmt);
+
+	num = vsnprintf(str->str + str->len, free_sz, fmt, args);
+
+	va_end(args);
+
+	if (num < free_sz) {
+		str->len += num;
+		err = 0;
+	} else {
+		str->len = str->sz - 1;
+		err = -ENOSPC;
+	}
+
+	return err;
+}
+
+static void evt_sprintf(struct komeda_str *str, u64 evt, const char *msg)
+{
+	if (evt)
+		komeda_sprintf(str, msg);
+}
+
+static void evt_str(struct komeda_str *str, u64 events)
+{
+	if (events == 0ULL) {
+		komeda_sprintf(str, "None");
+		return;
+	}
+
+	evt_sprintf(str, events & KOMEDA_EVENT_VSYNC, "VSYNC|");
+	evt_sprintf(str, events & KOMEDA_EVENT_FLIP, "FLIP|");
+	evt_sprintf(str, events & KOMEDA_EVENT_EOW, "EOW|");
+	evt_sprintf(str, events & KOMEDA_EVENT_MODE, "OP-MODE|");
+
+	evt_sprintf(str, events & KOMEDA_EVENT_URUN, "UNDERRUN|");
+	evt_sprintf(str, events & KOMEDA_EVENT_OVR, "OVERRUN|");
+
+	/* GLB error */
+	evt_sprintf(str, events & KOMEDA_ERR_MERR, "MERR|");
+	evt_sprintf(str, events & KOMEDA_ERR_FRAMETO, "FRAMETO|");
+
+	/* DOU error */
+	evt_sprintf(str, events & KOMEDA_ERR_DRIFTTO, "DRIFTTO|");
+	evt_sprintf(str, events & KOMEDA_ERR_FRAMETO, "FRAMETO|");
+	evt_sprintf(str, events & KOMEDA_ERR_TETO, "TETO|");
+	evt_sprintf(str, events & KOMEDA_ERR_CSCE, "CSCE|");
+
+	/* LPU errors or events */
+	evt_sprintf(str, events & KOMEDA_EVENT_IBSY, "IBSY|");
+	evt_sprintf(str, events & KOMEDA_ERR_AXIE, "AXIE|");
+	evt_sprintf(str, events & KOMEDA_ERR_ACE0, "ACE0|");
+	evt_sprintf(str, events & KOMEDA_ERR_ACE1, "ACE1|");
+	evt_sprintf(str, events & KOMEDA_ERR_ACE2, "ACE2|");
+	evt_sprintf(str, events & KOMEDA_ERR_ACE3, "ACE3|");
+
+	/* LPU TBU errors*/
+	evt_sprintf(str, events & KOMEDA_ERR_TCF, "TCF|");
+	evt_sprintf(str, events & KOMEDA_ERR_TTNG, "TTNG|");
+	evt_sprintf(str, events & KOMEDA_ERR_TITR, "TITR|");
+	evt_sprintf(str, events & KOMEDA_ERR_TEMR, "TEMR|");
+	evt_sprintf(str, events & KOMEDA_ERR_TTF, "TTF|");
+
+	/* CU errors*/
+	evt_sprintf(str, events & KOMEDA_ERR_CPE, "COPROC|");
+	evt_sprintf(str, events & KOMEDA_ERR_ZME, "ZME|");
+	evt_sprintf(str, events & KOMEDA_ERR_CFGE, "CFGE|");
+	evt_sprintf(str, events & KOMEDA_ERR_TEMR, "TEMR|");
+
+	if (str->len > 0 && (str->str[str->len - 1] == '|')) {
+		str->str[str->len - 1] = 0;
+		str->len--;
+	}
+}
+
+static bool is_new_frame(struct komeda_events *a)
+{
+	return (a->pipes[0] | a->pipes[1]) &
+	       (KOMEDA_EVENT_FLIP | KOMEDA_EVENT_EOW);
+}
+
+void komeda_print_events(struct komeda_events *evts)
+{
+	u64 print_evts = KOMEDA_ERR_EVENTS;
+	static bool en_print = true;
+
+	/* reduce the same msg print, only print the first evt for one frame */
+	if (evts->global || is_new_frame(evts))
+		en_print = true;
+	if (!en_print)
+		return;
+
+	if ((evts->global | evts->pipes[0] | evts->pipes[1]) & print_evts) {
+		char msg[256];
+		struct komeda_str str;
+
+		str.str = msg;
+		str.sz  = sizeof(msg);
+		str.len = 0;
+
+		komeda_sprintf(&str, "gcu: ");
+		evt_str(&str, evts->global);
+		komeda_sprintf(&str, ", pipes[0]: ");
+		evt_str(&str, evts->pipes[0]);
+		komeda_sprintf(&str, ", pipes[1]: ");
+		evt_str(&str, evts->pipes[1]);
+
+		DRM_ERROR("err detect: %s\n", msg);
+
+		en_print = false;
+	}
+}
diff --git a/drivers/gpu/drm/arm/display/komeda/komeda_kms.c b/drivers/gpu/drm/arm/display/komeda/komeda_kms.c
index 8820ce15ce37..52648b4008bc 100644
--- a/drivers/gpu/drm/arm/display/komeda/komeda_kms.c
+++ b/drivers/gpu/drm/arm/display/komeda/komeda_kms.c
@@ -48,6 +48,8 @@ static irqreturn_t komeda_kms_irq_handler(int irq, void *data)
 	memset(&evts, 0, sizeof(evts));
 	status = mdev->funcs->irq_handler(mdev, &evts);
 
+	komeda_print_events(&evts);
+
 	/* Notify the crtc to handle the events */
 	for (i = 0; i < kms->n_crtcs; i++)
 		komeda_crtc_handle_event(&kms->crtcs[i], &evts);
@@ -82,7 +84,8 @@ static void komeda_kms_commit_tail(struct drm_atomic_state *old_state)
 
 	drm_atomic_helper_commit_modeset_disables(dev, old_state);
 
-	drm_atomic_helper_commit_planes(dev, old_state, 0);
+	drm_atomic_helper_commit_planes(dev, old_state,
+					DRM_PLANE_COMMIT_ACTIVE_ONLY);
 
 	drm_atomic_helper_commit_modeset_enables(dev, old_state);
 
diff --git a/drivers/gpu/drm/arm/display/komeda/komeda_kms.h b/drivers/gpu/drm/arm/display/komeda/komeda_kms.h
index 45c498e15e7a..456f3c435719 100644
--- a/drivers/gpu/drm/arm/display/komeda/komeda_kms.h
+++ b/drivers/gpu/drm/arm/display/komeda/komeda_kms.h
@@ -166,6 +166,8 @@ static inline bool has_flip_h(u32 rot)
 		return !!(rotation & DRM_MODE_REFLECT_X);
 }
 
+void komeda_crtc_get_color_config(struct drm_crtc_state *crtc_st,
+				  u32 *color_depths, u32 *color_formats);
 unsigned long komeda_crtc_get_aclk(struct komeda_crtc_state *kcrtc_st);
 
 int komeda_kms_setup_crtcs(struct komeda_kms_dev *kms, struct komeda_dev *mdev);
diff --git a/drivers/gpu/drm/arm/display/komeda/komeda_pipeline.h b/drivers/gpu/drm/arm/display/komeda/komeda_pipeline.h
index cf5bea578ad9..bd6ca7c87037 100644
--- a/drivers/gpu/drm/arm/display/komeda/komeda_pipeline.h
+++ b/drivers/gpu/drm/arm/display/komeda/komeda_pipeline.h
@@ -227,6 +227,8 @@ struct komeda_layer {
 	/* accepted h/v input range before rotation */
 	struct malidp_range hsize_in, vsize_in;
 	u32 layer_type; /* RICH, SIMPLE or WB */
+	u32 line_sz;
+	u32 yuv_line_sz; /* maximum line size for YUV422 and YUV420 */
 	u32 supported_rots;
 	/* komeda supports layer split which splits a whole image to two parts
 	 * left and right and handle them by two individual layer processors
@@ -323,6 +325,7 @@ struct komeda_improc {
 
 struct komeda_improc_state {
 	struct komeda_component_state base;
+	u8 color_format, color_depth;
 	u16 hsize, vsize;
 };
 
@@ -389,6 +392,18 @@ struct komeda_pipeline {
 	int id;
 	/** @avail_comps: available components mask of pipeline */
 	u32 avail_comps;
+	/**
+	 * @standalone_disabled_comps:
+	 *
+	 * When disable the pipeline, some components can not be disabled
+	 * together with others, but need a sparated and standalone disable.
+	 * The standalone_disabled_comps are the components which need to be
+	 * disabled standalone, and this concept also introduce concept of
+	 * two phase.
+	 * phase 1: for disabling the common components.
+	 * phase 2: for disabling the standalong_disabled_comps.
+	 */
+	u32 standalone_disabled_comps;
 	/** @n_layers: the number of layer on @layers */
 	int n_layers;
 	/** @layers: the pipeline layers */
@@ -535,7 +550,7 @@ int komeda_release_unclaimed_resources(struct komeda_pipeline *pipe,
 struct komeda_pipeline_state *
 komeda_pipeline_get_old_state(struct komeda_pipeline *pipe,
 			      struct drm_atomic_state *state);
-void komeda_pipeline_disable(struct komeda_pipeline *pipe,
+bool komeda_pipeline_disable(struct komeda_pipeline *pipe,
 			     struct drm_atomic_state *old_state);
 void komeda_pipeline_update(struct komeda_pipeline *pipe,
 			    struct drm_atomic_state *old_state);
diff --git a/drivers/gpu/drm/arm/display/komeda/komeda_pipeline_state.c b/drivers/gpu/drm/arm/display/komeda/komeda_pipeline_state.c
index ea26bc9c2d00..52750116aa19 100644
--- a/drivers/gpu/drm/arm/display/komeda/komeda_pipeline_state.c
+++ b/drivers/gpu/drm/arm/display/komeda/komeda_pipeline_state.c
@@ -285,6 +285,7 @@ komeda_layer_check_cfg(struct komeda_layer *layer,
 		       struct komeda_data_flow_cfg *dflow)
 {
 	u32 src_x, src_y, src_w, src_h;
+	u32 line_sz, max_line_sz;
 
 	if (!komeda_fb_is_layer_supported(kfb, layer->layer_type, dflow->rot))
 		return -EINVAL;
@@ -314,6 +315,22 @@ komeda_layer_check_cfg(struct komeda_layer *layer,
 		return -EINVAL;
 	}
 
+	if (drm_rotation_90_or_270(dflow->rot))
+		line_sz = dflow->in_h;
+	else
+		line_sz = dflow->in_w;
+
+	if (kfb->base.format->hsub > 1)
+		max_line_sz = layer->yuv_line_sz;
+	else
+		max_line_sz = layer->line_sz;
+
+	if (line_sz > max_line_sz) {
+		DRM_DEBUG_ATOMIC("Required line_sz: %d exceeds the max size %d\n",
+				 line_sz, max_line_sz);
+		return -EINVAL;
+	}
+
 	return 0;
 }
 
@@ -564,8 +581,8 @@ komeda_splitter_validate(struct komeda_splitter *splitter,
 	}
 
 	if (!in_range(&splitter->vsize, dflow->in_h)) {
-		DRM_DEBUG_ATOMIC("split in_in: %d exceed the acceptable range.\n",
-				 dflow->in_w);
+		DRM_DEBUG_ATOMIC("split in_h: %d exceeds the acceptable range.\n",
+				 dflow->in_h);
 		return -EINVAL;
 	}
 
@@ -743,6 +760,7 @@ komeda_improc_validate(struct komeda_improc *improc,
 		       struct komeda_data_flow_cfg *dflow)
 {
 	struct drm_crtc *crtc = kcrtc_st->base.crtc;
+	struct drm_crtc_state *crtc_st = &kcrtc_st->base;
 	struct komeda_component_state *c_st;
 	struct komeda_improc_state *st;
 
@@ -756,6 +774,34 @@ komeda_improc_validate(struct komeda_improc *improc,
 	st->hsize = dflow->in_w;
 	st->vsize = dflow->in_h;
 
+	if (drm_atomic_crtc_needs_modeset(crtc_st)) {
+		u32 output_depths, output_formats;
+		u32 avail_depths, avail_formats;
+
+		komeda_crtc_get_color_config(crtc_st, &output_depths,
+					     &output_formats);
+
+		avail_depths = output_depths & improc->supported_color_depths;
+		if (avail_depths == 0) {
+			DRM_DEBUG_ATOMIC("No available color depths, conn depths: 0x%x & display: 0x%x\n",
+					 output_depths,
+					 improc->supported_color_depths);
+			return -EINVAL;
+		}
+
+		avail_formats = output_formats &
+				improc->supported_color_formats;
+		if (!avail_formats) {
+			DRM_DEBUG_ATOMIC("No available color_formats, conn formats 0x%x & display: 0x%x\n",
+					 output_formats,
+					 improc->supported_color_formats);
+			return -EINVAL;
+		}
+
+		st->color_depth = __fls(avail_depths);
+		st->color_format = BIT(__ffs(avail_formats));
+	}
+
 	komeda_component_add_input(&st->base, &dflow->input, 0);
 	komeda_component_set_output(&dflow->input, &improc->base, 0);
 
@@ -1218,7 +1264,17 @@ int komeda_release_unclaimed_resources(struct komeda_pipeline *pipe,
 	return 0;
 }
 
-void komeda_pipeline_disable(struct komeda_pipeline *pipe,
+/* Since standalong disabled components must be disabled separately and in the
+ * last, So a complete disable operation may needs to call pipeline_disable
+ * twice (two phase disabling).
+ * Phase 1: disable the common components, flush it.
+ * Phase 2: disable the standalone disabled components, flush it.
+ *
+ * RETURNS:
+ * true: disable is not complete, needs a phase 2 disable.
+ * false: disable is complete.
+ */
+bool komeda_pipeline_disable(struct komeda_pipeline *pipe,
 			     struct drm_atomic_state *old_state)
 {
 	struct komeda_pipeline_state *old;
@@ -1228,9 +1284,14 @@ void komeda_pipeline_disable(struct komeda_pipeline *pipe,
 
 	old = komeda_pipeline_get_old_state(pipe, old_state);
 
-	disabling_comps = old->active_comps;
-	DRM_DEBUG_ATOMIC("PIPE%d: disabling_comps: 0x%x.\n",
-			 pipe->id, disabling_comps);
+	disabling_comps = old->active_comps &
+			  (~pipe->standalone_disabled_comps);
+	if (!disabling_comps)
+		disabling_comps = old->active_comps &
+				  pipe->standalone_disabled_comps;
+
+	DRM_DEBUG_ATOMIC("PIPE%d: active_comps: 0x%x, disabling_comps: 0x%x.\n",
+			 pipe->id, old->active_comps, disabling_comps);
 
 	dp_for_each_set_bit(id, disabling_comps) {
 		c = komeda_pipeline_get_component(pipe, id);
@@ -1248,6 +1309,13 @@ void komeda_pipeline_disable(struct komeda_pipeline *pipe,
 
 		c->funcs->disable(c);
 	}
+
+	/* Update the pipeline state, if there are components that are still
+	 * active, return true for calling the phase 2 disable.
+	 */
+	old->active_comps &= ~disabling_comps;
+
+	return old->active_comps ? true : false;
 }
 
 void komeda_pipeline_update(struct komeda_pipeline *pipe,
diff --git a/drivers/gpu/drm/arm/display/komeda/komeda_wb_connector.c b/drivers/gpu/drm/arm/display/komeda/komeda_wb_connector.c
index 2851cac94d86..e465cc4879c9 100644
--- a/drivers/gpu/drm/arm/display/komeda/komeda_wb_connector.c
+++ b/drivers/gpu/drm/arm/display/komeda/komeda_wb_connector.c
@@ -43,9 +43,8 @@ komeda_wb_encoder_atomic_check(struct drm_encoder *encoder,
 	struct komeda_data_flow_cfg dflow;
 	int err;
 
-	if (!writeback_job || !writeback_job->fb) {
+	if (!writeback_job)
 		return 0;
-	}
 
 	if (!crtc_st->active) {
 		DRM_DEBUG_ATOMIC("Cannot write the composition result out on a inactive CRTC.\n");
@@ -142,6 +141,7 @@ static int komeda_wb_connector_add(struct komeda_kms_dev *kms,
 	struct komeda_dev *mdev = kms->base.dev_private;
 	struct komeda_wb_connector *kwb_conn;
 	struct drm_writeback_connector *wb_conn;
+	struct drm_display_info *info;
 	u32 *formats, n_formats = 0;
 	int err;
 
@@ -166,11 +166,17 @@ static int komeda_wb_connector_add(struct komeda_kms_dev *kms,
 					   &komeda_wb_encoder_helper_funcs,
 					   formats, n_formats);
 	komeda_put_fourcc_list(formats);
-	if (err)
+	if (err) {
+		kfree(kwb_conn);
 		return err;
+	}
 
 	drm_connector_helper_add(&wb_conn->base, &komeda_wb_conn_helper_funcs);
 
+	info = &kwb_conn->base.base.display_info;
+	info->bpc = __fls(kcrtc->master->improc->supported_color_depths);
+	info->color_formats = kcrtc->master->improc->supported_color_formats;
+
 	kcrtc->wb_conn = kwb_conn;
 
 	return 0;
diff --git a/drivers/gpu/drm/arm/malidp_drv.c b/drivers/gpu/drm/arm/malidp_drv.c
index 333b88a5efb0..37d92a06318e 100644
--- a/drivers/gpu/drm/arm/malidp_drv.c
+++ b/drivers/gpu/drm/arm/malidp_drv.c
@@ -368,7 +368,7 @@ malidp_verify_afbc_framebuffer(struct drm_device *dev, struct drm_file *file,
 	return false;
 }
 
-struct drm_framebuffer *
+static struct drm_framebuffer *
 malidp_fb_create(struct drm_device *dev, struct drm_file *file,
 		 const struct drm_mode_fb_cmd2 *mode_cmd)
 {
@@ -491,9 +491,9 @@ void malidp_error(struct malidp_drm *malidp,
 	spin_unlock_irqrestore(&malidp->errors_lock, irqflags);
 }
 
-void malidp_error_stats_dump(const char *prefix,
-			     struct malidp_error_stats error_stats,
-			     struct seq_file *m)
+static void malidp_error_stats_dump(const char *prefix,
+				    struct malidp_error_stats error_stats,
+				    struct seq_file *m)
 {
 	seq_printf(m, "[%s] num_errors : %d\n", prefix,
 		   error_stats.num_errors);
@@ -665,7 +665,7 @@ static ssize_t core_id_show(struct device *dev, struct device_attribute *attr,
 	return snprintf(buf, PAGE_SIZE, "%08x\n", malidp->core_id);
 }
 
-DEVICE_ATTR_RO(core_id);
+static DEVICE_ATTR_RO(core_id);
 
 static int malidp_init_sysfs(struct device *dev)
 {
@@ -817,6 +817,12 @@ static int malidp_bind(struct device *dev)
 
 	malidp->core_id = version;
 
+	ret = of_property_read_u32(dev->of_node,
+					"arm,malidp-arqos-value",
+					&hwdev->arqos_value);
+	if (ret)
+		hwdev->arqos_value = 0x0;
+
 	/* set the number of lines used for output of RGB data */
 	ret = of_property_read_u8_array(dev->of_node,
 					"arm,malidp-output-port-lines",
diff --git a/drivers/gpu/drm/arm/malidp_hw.c b/drivers/gpu/drm/arm/malidp_hw.c
index bd8265f02e0b..ca570b135478 100644
--- a/drivers/gpu/drm/arm/malidp_hw.c
+++ b/drivers/gpu/drm/arm/malidp_hw.c
@@ -379,6 +379,15 @@ static void malidp500_modeset(struct malidp_hw_device *hwdev, struct videomode *
 		malidp_hw_setbits(hwdev, MALIDP_DISP_FUNC_ILACED, MALIDP_DE_DISPLAY_FUNC);
 	else
 		malidp_hw_clearbits(hwdev, MALIDP_DISP_FUNC_ILACED, MALIDP_DE_DISPLAY_FUNC);
+
+	/*
+	 * Program the RQoS register to avoid high resolutions flicker
+	 * issue on the LS1028A.
+	 */
+	if (hwdev->arqos_value) {
+		val = hwdev->arqos_value;
+		malidp_hw_setbits(hwdev, val, MALIDP500_RQOS_QUALITY);
+	}
 }
 
 int malidp_format_get_bpp(u32 fmt)
diff --git a/drivers/gpu/drm/arm/malidp_hw.h b/drivers/gpu/drm/arm/malidp_hw.h
index 968a65eed371..e4c36bc90bda 100644
--- a/drivers/gpu/drm/arm/malidp_hw.h
+++ b/drivers/gpu/drm/arm/malidp_hw.h
@@ -251,6 +251,9 @@ struct malidp_hw_device {
 
 	/* size of memory used for rotating layers, up to two banks available */
 	u32 rotation_memory[2];
+
+	/* priority level of RQOS register used for driven the ARQOS signal */
+	u32 arqos_value;
 };
 
 static inline u32 malidp_hw_read(struct malidp_hw_device *hwdev, u32 reg)
diff --git a/drivers/gpu/drm/arm/malidp_mw.c b/drivers/gpu/drm/arm/malidp_mw.c
index 22c0847986df..875a3a9eabfa 100644
--- a/drivers/gpu/drm/arm/malidp_mw.c
+++ b/drivers/gpu/drm/arm/malidp_mw.c
@@ -131,7 +131,7 @@ malidp_mw_encoder_atomic_check(struct drm_encoder *encoder,
 	struct drm_framebuffer *fb;
 	int i, n_planes;
 
-	if (!conn_state->writeback_job || !conn_state->writeback_job->fb)
+	if (!conn_state->writeback_job)
 		return 0;
 
 	fb = conn_state->writeback_job->fb;
@@ -248,7 +248,7 @@ void malidp_mw_atomic_commit(struct drm_device *drm,
 
 	mw_state = to_mw_state(conn_state);
 
-	if (conn_state->writeback_job && conn_state->writeback_job->fb) {
+	if (conn_state->writeback_job) {
 		struct drm_framebuffer *fb = conn_state->writeback_job->fb;
 
 		DRM_DEV_DEBUG_DRIVER(drm->dev,
diff --git a/drivers/gpu/drm/arm/malidp_regs.h b/drivers/gpu/drm/arm/malidp_regs.h
index 993031542fa1..514c50dcb74d 100644
--- a/drivers/gpu/drm/arm/malidp_regs.h
+++ b/drivers/gpu/drm/arm/malidp_regs.h
@@ -210,6 +210,16 @@
 #define MALIDP500_CONFIG_VALID		0x00f00
 #define MALIDP500_CONFIG_ID		0x00fd4
 
+/*
+ * The quality of service (QoS) register on the DP500. RQOS register values
+ * are driven by the ARQOS signal, using AXI transacations, dependent on the
+ * FIFO input level.
+ * The RQOS register can also set QoS levels for:
+ *    - RED_ARQOS   @ A 4-bit signal value for close to underflow conditions
+ *    - GREEN_ARQOS @ A 4-bit signal value for normal conditions
+ */
+#define MALIDP500_RQOS_QUALITY          0x00500
+
 /* register offsets and bits specific to DP550/DP650 */
 #define MALIDP550_ADDR_SPACE_SIZE	0x10000
 #define MALIDP550_DE_CONTROL		0x00010
diff --git a/drivers/gpu/drm/ast/Kconfig b/drivers/gpu/drm/ast/Kconfig
index 829620d5326c..fbcf2f45cef5 100644
--- a/drivers/gpu/drm/ast/Kconfig
+++ b/drivers/gpu/drm/ast/Kconfig
@@ -4,6 +4,8 @@ config DRM_AST
 	depends on DRM && PCI && MMU
 	select DRM_KMS_HELPER
 	select DRM_VRAM_HELPER
+	select DRM_TTM
+	select DRM_TTM_HELPER
 	help
 	 Say yes for experimental AST GPU driver. Do not enable
 	 this driver without having a working -modesetting,
diff --git a/drivers/gpu/drm/ast/ast_drv.c b/drivers/gpu/drm/ast/ast_drv.c
index 6ed6ff49efc0..1f17794b0890 100644
--- a/drivers/gpu/drm/ast/ast_drv.c
+++ b/drivers/gpu/drm/ast/ast_drv.c
@@ -35,7 +35,6 @@
 #include <drm/drm_gem_vram_helper.h>
 #include <drm/drm_pci.h>
 #include <drm/drm_probe_helper.h>
-#include <drm/drm_vram_mm_helper.h>
 
 #include "ast_drv.h"
 
@@ -201,10 +200,7 @@ static struct pci_driver ast_pci_driver = {
 	.driver.pm = &ast_pm_ops,
 };
 
-static const struct file_operations ast_fops = {
-	.owner = THIS_MODULE,
-	DRM_VRAM_MM_FILE_OPERATIONS
-};
+DEFINE_DRM_GEM_FOPS(ast_fops);
 
 static struct drm_driver driver = {
 	.driver_features = DRIVER_MODESET | DRIVER_GEM,
diff --git a/drivers/gpu/drm/ast/ast_drv.h b/drivers/gpu/drm/ast/ast_drv.h
index 244cc7c382af..ff161bd622f3 100644
--- a/drivers/gpu/drm/ast/ast_drv.h
+++ b/drivers/gpu/drm/ast/ast_drv.h
@@ -82,6 +82,25 @@ enum ast_tx_chip {
 #define AST_DRAM_4Gx16   7
 #define AST_DRAM_8Gx16   8
 
+
+#define AST_MAX_HWC_WIDTH	64
+#define AST_MAX_HWC_HEIGHT	64
+
+#define AST_HWC_SIZE		(AST_MAX_HWC_WIDTH * AST_MAX_HWC_HEIGHT * 2)
+#define AST_HWC_SIGNATURE_SIZE	32
+
+#define AST_DEFAULT_HWC_NUM	2
+
+/* define for signature structure */
+#define AST_HWC_SIGNATURE_CHECKSUM	0x00
+#define AST_HWC_SIGNATURE_SizeX		0x04
+#define AST_HWC_SIGNATURE_SizeY		0x08
+#define AST_HWC_SIGNATURE_X		0x0C
+#define AST_HWC_SIGNATURE_Y		0x10
+#define AST_HWC_SIGNATURE_HOTSPOTX	0x14
+#define AST_HWC_SIGNATURE_HOTSPOTY	0x18
+
+
 struct ast_private {
 	struct drm_device *dev;
 
@@ -97,8 +116,11 @@ struct ast_private {
 
 	int fb_mtrr;
 
-	struct drm_gem_object *cursor_cache;
-	int next_cursor;
+	struct {
+		struct drm_gem_vram_object *gbo[AST_DEFAULT_HWC_NUM];
+		unsigned int next_index;
+	} cursor;
+
 	bool support_wide_screen;
 	enum {
 		ast_use_p2a,
@@ -199,23 +221,6 @@ static inline void ast_open_key(struct ast_private *ast)
 
 #define AST_VIDMEM_DEFAULT_SIZE AST_VIDMEM_SIZE_8M
 
-#define AST_MAX_HWC_WIDTH 64
-#define AST_MAX_HWC_HEIGHT 64
-
-#define AST_HWC_SIZE                (AST_MAX_HWC_WIDTH*AST_MAX_HWC_HEIGHT*2)
-#define AST_HWC_SIGNATURE_SIZE      32
-
-#define AST_DEFAULT_HWC_NUM 2
-/* define for signature structure */
-#define AST_HWC_SIGNATURE_CHECKSUM  0x00
-#define AST_HWC_SIGNATURE_SizeX     0x04
-#define AST_HWC_SIGNATURE_SizeY     0x08
-#define AST_HWC_SIGNATURE_X         0x0C
-#define AST_HWC_SIGNATURE_Y         0x10
-#define AST_HWC_SIGNATURE_HOTSPOTX  0x14
-#define AST_HWC_SIGNATURE_HOTSPOTY  0x18
-
-
 struct ast_i2c_chan {
 	struct i2c_adapter adapter;
 	struct drm_device *dev;
diff --git a/drivers/gpu/drm/ast/ast_main.c b/drivers/gpu/drm/ast/ast_main.c
index 50de8e47659c..21715d6a9b56 100644
--- a/drivers/gpu/drm/ast/ast_main.c
+++ b/drivers/gpu/drm/ast/ast_main.c
@@ -33,7 +33,6 @@
 #include <drm/drm_gem.h>
 #include <drm/drm_gem_framebuffer_helper.h>
 #include <drm/drm_gem_vram_helper.h>
-#include <drm/drm_vram_mm_helper.h>
 
 #include "ast_drv.h"
 
diff --git a/drivers/gpu/drm/ast/ast_mode.c b/drivers/gpu/drm/ast/ast_mode.c
index d349c721501c..b13eaa2619ab 100644
--- a/drivers/gpu/drm/ast/ast_mode.c
+++ b/drivers/gpu/drm/ast/ast_mode.c
@@ -687,17 +687,6 @@ static void ast_encoder_destroy(struct drm_encoder *encoder)
 	kfree(encoder);
 }
 
-
-static struct drm_encoder *ast_best_single_encoder(struct drm_connector *connector)
-{
-	int enc_id = connector->encoder_ids[0];
-	/* pick the encoder ids */
-	if (enc_id)
-		return drm_encoder_find(connector->dev, NULL, enc_id);
-	return NULL;
-}
-
-
 static const struct drm_encoder_funcs ast_enc_funcs = {
 	.destroy = ast_encoder_destroy,
 };
@@ -847,7 +836,6 @@ static void ast_connector_destroy(struct drm_connector *connector)
 static const struct drm_connector_helper_funcs ast_connector_helper_funcs = {
 	.mode_valid = ast_mode_valid,
 	.get_modes = ast_get_modes,
-	.best_encoder = ast_best_single_encoder,
 };
 
 static const struct drm_connector_funcs ast_connector_funcs = {
@@ -895,50 +883,53 @@ static int ast_connector_init(struct drm_device *dev)
 static int ast_cursor_init(struct drm_device *dev)
 {
 	struct ast_private *ast = dev->dev_private;
-	int size;
-	int ret;
-	struct drm_gem_object *obj;
+	size_t size, i;
 	struct drm_gem_vram_object *gbo;
-	s64 gpu_addr;
-	void *base;
+	int ret;
 
-	size = (AST_HWC_SIZE + AST_HWC_SIGNATURE_SIZE) * AST_DEFAULT_HWC_NUM;
+	size = roundup(AST_HWC_SIZE + AST_HWC_SIGNATURE_SIZE, PAGE_SIZE);
 
-	ret = ast_gem_create(dev, size, true, &obj);
-	if (ret)
-		return ret;
-	gbo = drm_gem_vram_of_gem(obj);
-	ret = drm_gem_vram_pin(gbo, DRM_GEM_VRAM_PL_FLAG_VRAM);
-	if (ret)
-		goto fail;
-	gpu_addr = drm_gem_vram_offset(gbo);
-	if (gpu_addr < 0) {
-		drm_gem_vram_unpin(gbo);
-		ret = (int)gpu_addr;
-		goto fail;
-	}
+	for (i = 0; i < ARRAY_SIZE(ast->cursor.gbo); ++i) {
+		gbo = drm_gem_vram_create(dev, &dev->vram_mm->bdev,
+					  size, 0, false);
+		if (IS_ERR(gbo)) {
+			ret = PTR_ERR(gbo);
+			goto err_drm_gem_vram_put;
+		}
+		ret = drm_gem_vram_pin(gbo, DRM_GEM_VRAM_PL_FLAG_VRAM |
+					    DRM_GEM_VRAM_PL_FLAG_TOPDOWN);
+		if (ret) {
+			drm_gem_vram_put(gbo);
+			goto err_drm_gem_vram_put;
+		}
 
-	/* kmap the object */
-	base = drm_gem_vram_kmap(gbo, true, NULL);
-	if (IS_ERR(base)) {
-		ret = PTR_ERR(base);
-		goto fail;
+		ast->cursor.gbo[i] = gbo;
 	}
 
-	ast->cursor_cache = obj;
 	return 0;
-fail:
+
+err_drm_gem_vram_put:
+	while (i) {
+		--i;
+		gbo = ast->cursor.gbo[i];
+		drm_gem_vram_unpin(gbo);
+		drm_gem_vram_put(gbo);
+		ast->cursor.gbo[i] = NULL;
+	}
 	return ret;
 }
 
 static void ast_cursor_fini(struct drm_device *dev)
 {
 	struct ast_private *ast = dev->dev_private;
-	struct drm_gem_vram_object *gbo =
-		drm_gem_vram_of_gem(ast->cursor_cache);
-	drm_gem_vram_kunmap(gbo);
-	drm_gem_vram_unpin(gbo);
-	drm_gem_object_put_unlocked(ast->cursor_cache);
+	size_t i;
+	struct drm_gem_vram_object *gbo;
+
+	for (i = 0; i < ARRAY_SIZE(ast->cursor.gbo); ++i) {
+		gbo = ast->cursor.gbo[i];
+		drm_gem_vram_unpin(gbo);
+		drm_gem_vram_put(gbo);
+	}
 }
 
 int ast_mode_init(struct drm_device *dev)
@@ -1076,23 +1067,6 @@ static void ast_i2c_destroy(struct ast_i2c_chan *i2c)
 	kfree(i2c);
 }
 
-static void ast_show_cursor(struct drm_crtc *crtc)
-{
-	struct ast_private *ast = crtc->dev->dev_private;
-	u8 jreg;
-
-	jreg = 0x2;
-	/* enable ARGB cursor */
-	jreg |= 1;
-	ast_set_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xcb, 0xfc, jreg);
-}
-
-static void ast_hide_cursor(struct drm_crtc *crtc)
-{
-	struct ast_private *ast = crtc->dev->dev_private;
-	ast_set_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xcb, 0xfc, 0x00);
-}
-
 static u32 copy_cursor_image(u8 *src, u8 *dst, int width, int height)
 {
 	union {
@@ -1149,21 +1123,99 @@ static u32 copy_cursor_image(u8 *src, u8 *dst, int width, int height)
 	return csum;
 }
 
+static int ast_cursor_update(void *dst, void *src, unsigned int width,
+			     unsigned int height)
+{
+	u32 csum;
+
+	/* do data transfer to cursor cache */
+	csum = copy_cursor_image(src, dst, width, height);
+
+	/* write checksum + signature */
+	dst += AST_HWC_SIZE;
+	writel(csum, dst);
+	writel(width, dst + AST_HWC_SIGNATURE_SizeX);
+	writel(height, dst + AST_HWC_SIGNATURE_SizeY);
+	writel(0, dst + AST_HWC_SIGNATURE_HOTSPOTX);
+	writel(0, dst + AST_HWC_SIGNATURE_HOTSPOTY);
+
+	return 0;
+}
+
+static void ast_cursor_set_base(struct ast_private *ast, u64 address)
+{
+	u8 addr0 = (address >> 3) & 0xff;
+	u8 addr1 = (address >> 11) & 0xff;
+	u8 addr2 = (address >> 19) & 0xff;
+
+	ast_set_index_reg(ast, AST_IO_CRTC_PORT, 0xc8, addr0);
+	ast_set_index_reg(ast, AST_IO_CRTC_PORT, 0xc9, addr1);
+	ast_set_index_reg(ast, AST_IO_CRTC_PORT, 0xca, addr2);
+}
+
+static int ast_show_cursor(struct drm_crtc *crtc, void *src,
+			   unsigned int width, unsigned int height)
+{
+	struct ast_private *ast = crtc->dev->dev_private;
+	struct ast_crtc *ast_crtc = to_ast_crtc(crtc);
+	struct drm_gem_vram_object *gbo;
+	void *dst;
+	s64 off;
+	int ret;
+	u8 jreg;
+
+	gbo = ast->cursor.gbo[ast->cursor.next_index];
+	dst = drm_gem_vram_vmap(gbo);
+	if (IS_ERR(dst))
+		return PTR_ERR(dst);
+	off = drm_gem_vram_offset(gbo);
+	if (off < 0) {
+		ret = (int)off;
+		goto err_drm_gem_vram_vunmap;
+	}
+
+	ret = ast_cursor_update(dst, src, width, height);
+	if (ret)
+		goto err_drm_gem_vram_vunmap;
+	ast_cursor_set_base(ast, off);
+
+	ast_crtc->offset_x = AST_MAX_HWC_WIDTH - width;
+	ast_crtc->offset_y = AST_MAX_HWC_WIDTH - height;
+
+	jreg = 0x2;
+	/* enable ARGB cursor */
+	jreg |= 1;
+	ast_set_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xcb, 0xfc, jreg);
+
+	++ast->cursor.next_index;
+	ast->cursor.next_index %= ARRAY_SIZE(ast->cursor.gbo);
+
+	drm_gem_vram_vunmap(gbo, dst);
+
+	return 0;
+
+err_drm_gem_vram_vunmap:
+	drm_gem_vram_vunmap(gbo, dst);
+	return ret;
+}
+
+static void ast_hide_cursor(struct drm_crtc *crtc)
+{
+	struct ast_private *ast = crtc->dev->dev_private;
+
+	ast_set_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xcb, 0xfc, 0x00);
+}
+
 static int ast_cursor_set(struct drm_crtc *crtc,
 			  struct drm_file *file_priv,
 			  uint32_t handle,
 			  uint32_t width,
 			  uint32_t height)
 {
-	struct ast_private *ast = crtc->dev->dev_private;
-	struct ast_crtc *ast_crtc = to_ast_crtc(crtc);
 	struct drm_gem_object *obj;
 	struct drm_gem_vram_object *gbo;
-	s64 dst_gpu;
-	u64 gpu_addr;
-	u32 csum;
+	u8 *src;
 	int ret;
-	u8 *src, *dst;
 
 	if (!handle) {
 		ast_hide_cursor(crtc);
@@ -1179,70 +1231,23 @@ static int ast_cursor_set(struct drm_crtc *crtc,
 		return -ENOENT;
 	}
 	gbo = drm_gem_vram_of_gem(obj);
-
-	ret = drm_gem_vram_pin(gbo, 0);
-	if (ret)
-		goto err_drm_gem_object_put_unlocked;
-	src = drm_gem_vram_kmap(gbo, true, NULL);
+	src = drm_gem_vram_vmap(gbo);
 	if (IS_ERR(src)) {
 		ret = PTR_ERR(src);
-		goto err_drm_gem_vram_unpin;
-	}
-
-	dst = drm_gem_vram_kmap(drm_gem_vram_of_gem(ast->cursor_cache),
-				false, NULL);
-	if (IS_ERR(dst)) {
-		ret = PTR_ERR(dst);
-		goto err_drm_gem_vram_kunmap;
-	}
-	dst_gpu = drm_gem_vram_offset(drm_gem_vram_of_gem(ast->cursor_cache));
-	if (dst_gpu < 0) {
-		ret = (int)dst_gpu;
-		goto err_drm_gem_vram_kunmap;
-	}
-
-	dst += (AST_HWC_SIZE + AST_HWC_SIGNATURE_SIZE)*ast->next_cursor;
-
-	/* do data transfer to cursor cache */
-	csum = copy_cursor_image(src, dst, width, height);
-
-	/* write checksum + signature */
-	{
-		struct drm_gem_vram_object *dst_gbo =
-			drm_gem_vram_of_gem(ast->cursor_cache);
-		u8 *dst = drm_gem_vram_kmap(dst_gbo, false, NULL);
-		dst += (AST_HWC_SIZE + AST_HWC_SIGNATURE_SIZE)*ast->next_cursor + AST_HWC_SIZE;
-		writel(csum, dst);
-		writel(width, dst + AST_HWC_SIGNATURE_SizeX);
-		writel(height, dst + AST_HWC_SIGNATURE_SizeY);
-		writel(0, dst + AST_HWC_SIGNATURE_HOTSPOTX);
-		writel(0, dst + AST_HWC_SIGNATURE_HOTSPOTY);
-
-		/* set pattern offset */
-		gpu_addr = (u64)dst_gpu;
-		gpu_addr += (AST_HWC_SIZE + AST_HWC_SIGNATURE_SIZE)*ast->next_cursor;
-		gpu_addr >>= 3;
-		ast_set_index_reg(ast, AST_IO_CRTC_PORT, 0xc8, gpu_addr & 0xff);
-		ast_set_index_reg(ast, AST_IO_CRTC_PORT, 0xc9, (gpu_addr >> 8) & 0xff);
-		ast_set_index_reg(ast, AST_IO_CRTC_PORT, 0xca, (gpu_addr >> 16) & 0xff);
+		goto err_drm_gem_object_put_unlocked;
 	}
-	ast_crtc->offset_x = AST_MAX_HWC_WIDTH - width;
-	ast_crtc->offset_y = AST_MAX_HWC_WIDTH - height;
-
-	ast->next_cursor = (ast->next_cursor + 1) % AST_DEFAULT_HWC_NUM;
 
-	ast_show_cursor(crtc);
+	ret = ast_show_cursor(crtc, src, width, height);
+	if (ret)
+		goto err_drm_gem_vram_vunmap;
 
-	drm_gem_vram_kunmap(gbo);
-	drm_gem_vram_unpin(gbo);
+	drm_gem_vram_vunmap(gbo, src);
 	drm_gem_object_put_unlocked(obj);
 
 	return 0;
 
-err_drm_gem_vram_kunmap:
-	drm_gem_vram_kunmap(gbo);
-err_drm_gem_vram_unpin:
-	drm_gem_vram_unpin(gbo);
+err_drm_gem_vram_vunmap:
+	drm_gem_vram_vunmap(gbo, src);
 err_drm_gem_object_put_unlocked:
 	drm_gem_object_put_unlocked(obj);
 	return ret;
@@ -1253,12 +1258,17 @@ static int ast_cursor_move(struct drm_crtc *crtc,
 {
 	struct ast_crtc *ast_crtc = to_ast_crtc(crtc);
 	struct ast_private *ast = crtc->dev->dev_private;
+	struct drm_gem_vram_object *gbo;
 	int x_offset, y_offset;
-	u8 *sig;
+	u8 *dst, *sig;
+	u8 jreg;
 
-	sig = drm_gem_vram_kmap(drm_gem_vram_of_gem(ast->cursor_cache),
-				false, NULL);
-	sig += (AST_HWC_SIZE + AST_HWC_SIGNATURE_SIZE)*ast->next_cursor + AST_HWC_SIZE;
+	gbo = ast->cursor.gbo[ast->cursor.next_index];
+	dst = drm_gem_vram_vmap(gbo);
+	if (IS_ERR(dst))
+		return PTR_ERR(dst);
+
+	sig = dst + AST_HWC_SIZE;
 	writel(x, sig + AST_HWC_SIGNATURE_X);
 	writel(y, sig + AST_HWC_SIGNATURE_Y);
 
@@ -1281,7 +1291,11 @@ static int ast_cursor_move(struct drm_crtc *crtc,
 	ast_set_index_reg(ast, AST_IO_CRTC_PORT, 0xc7, ((y >> 8) & 0x07));
 
 	/* dummy write to fire HWC */
-	ast_show_cursor(crtc);
+	jreg = 0x02 |
+	       0x01; /* enable ARGB4444 cursor */
+	ast_set_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xcb, 0xfc, jreg);
+
+	drm_gem_vram_vunmap(gbo, dst);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/ast/ast_ttm.c b/drivers/gpu/drm/ast/ast_ttm.c
index c52d92294171..fad34106083a 100644
--- a/drivers/gpu/drm/ast/ast_ttm.c
+++ b/drivers/gpu/drm/ast/ast_ttm.c
@@ -30,7 +30,6 @@
 
 #include <drm/drm_print.h>
 #include <drm/drm_gem_vram_helper.h>
-#include <drm/drm_vram_mm_helper.h>
 
 #include "ast_drv.h"
 
@@ -42,7 +41,7 @@ int ast_mm_init(struct ast_private *ast)
 
 	vmm = drm_vram_helper_alloc_mm(
 		dev, pci_resource_start(dev->pdev, 0),
-		ast->vram_size, &drm_gem_vram_mm_funcs);
+		ast->vram_size);
 	if (IS_ERR(vmm)) {
 		ret = PTR_ERR(vmm);
 		DRM_ERROR("Error initializing VRAM MM; %d\n", ret);
diff --git a/drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_output.c b/drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_output.c
index 375fa84c548b..121b62682d80 100644
--- a/drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_output.c
+++ b/drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_output.c
@@ -107,7 +107,8 @@ static int atmel_hlcdc_attach_endpoint(struct drm_device *dev, int endpoint)
 	output->encoder.possible_crtcs = 0x1;
 
 	if (panel) {
-		bridge = drm_panel_bridge_add(panel, DRM_MODE_CONNECTOR_Unknown);
+		bridge = drm_panel_bridge_add_typed(panel,
+						    DRM_MODE_CONNECTOR_Unknown);
 		if (IS_ERR(bridge))
 			return PTR_ERR(bridge);
 	}
diff --git a/drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_plane.c b/drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_plane.c
index 89f5a756fa37..034f202dfe8f 100644
--- a/drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_plane.c
+++ b/drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_plane.c
@@ -601,7 +601,6 @@ static int atmel_hlcdc_plane_atomic_check(struct drm_plane *p,
 	struct drm_framebuffer *fb = state->base.fb;
 	const struct drm_display_mode *mode;
 	struct drm_crtc_state *crtc_state;
-	unsigned int tmp;
 	int ret;
 	int i;
 
@@ -694,9 +693,7 @@ static int atmel_hlcdc_plane_atomic_check(struct drm_plane *p,
 	 * Swap width and size in case of 90 or 270 degrees rotation
 	 */
 	if (drm_rotation_90_or_270(state->base.rotation)) {
-		tmp = state->src_w;
-		state->src_w = state->src_h;
-		state->src_h = tmp;
+		swap(state->src_w, state->src_h);
 	}
 
 	if (!desc->layout.size &&
diff --git a/drivers/gpu/drm/bochs/Kconfig b/drivers/gpu/drm/bochs/Kconfig
index 32b043abb668..7bcdf294fed8 100644
--- a/drivers/gpu/drm/bochs/Kconfig
+++ b/drivers/gpu/drm/bochs/Kconfig
@@ -4,6 +4,8 @@ config DRM_BOCHS
 	depends on DRM && PCI && MMU
 	select DRM_KMS_HELPER
 	select DRM_VRAM_HELPER
+	select DRM_TTM
+	select DRM_TTM_HELPER
 	help
 	  Choose this option for qemu.
 	  If M is selected the module will be called bochs-drm.
diff --git a/drivers/gpu/drm/bochs/bochs.h b/drivers/gpu/drm/bochs/bochs.h
index 68483a2fc12c..917767173ee6 100644
--- a/drivers/gpu/drm/bochs/bochs.h
+++ b/drivers/gpu/drm/bochs/bochs.h
@@ -10,7 +10,6 @@
 #include <drm/drm_gem.h>
 #include <drm/drm_gem_vram_helper.h>
 #include <drm/drm_simple_kms_helper.h>
-#include <drm/drm_vram_mm_helper.h>
 
 /* ---------------------------------------------------------------------- */
 
diff --git a/drivers/gpu/drm/bochs/bochs_drv.c b/drivers/gpu/drm/bochs/bochs_drv.c
index 770e1625d05e..10460878414e 100644
--- a/drivers/gpu/drm/bochs/bochs_drv.c
+++ b/drivers/gpu/drm/bochs/bochs_drv.c
@@ -58,10 +58,7 @@ err:
 	return ret;
 }
 
-static const struct file_operations bochs_fops = {
-	.owner		= THIS_MODULE,
-	DRM_VRAM_MM_FILE_OPERATIONS
-};
+DEFINE_DRM_GEM_FOPS(bochs_fops);
 
 static struct drm_driver bochs_driver = {
 	.driver_features	= DRIVER_GEM | DRIVER_MODESET | DRIVER_ATOMIC,
@@ -114,7 +111,7 @@ static int bochs_pci_probe(struct pci_dev *pdev,
 		return -ENOMEM;
 	}
 
-	ret = drm_fb_helper_remove_conflicting_pci_framebuffers(pdev, 0, "bochsdrmfb");
+	ret = drm_fb_helper_remove_conflicting_pci_framebuffers(pdev, "bochsdrmfb");
 	if (ret)
 		return ret;
 
diff --git a/drivers/gpu/drm/bochs/bochs_kms.c b/drivers/gpu/drm/bochs/bochs_kms.c
index 02a9c1ed165b..3f0006c2470d 100644
--- a/drivers/gpu/drm/bochs/bochs_kms.c
+++ b/drivers/gpu/drm/bochs/bochs_kms.c
@@ -69,33 +69,11 @@ static void bochs_pipe_update(struct drm_simple_display_pipe *pipe,
 	}
 }
 
-static int bochs_pipe_prepare_fb(struct drm_simple_display_pipe *pipe,
-				 struct drm_plane_state *new_state)
-{
-	struct drm_gem_vram_object *gbo;
-
-	if (!new_state->fb)
-		return 0;
-	gbo = drm_gem_vram_of_gem(new_state->fb->obj[0]);
-	return drm_gem_vram_pin(gbo, DRM_GEM_VRAM_PL_FLAG_VRAM);
-}
-
-static void bochs_pipe_cleanup_fb(struct drm_simple_display_pipe *pipe,
-				  struct drm_plane_state *old_state)
-{
-	struct drm_gem_vram_object *gbo;
-
-	if (!old_state->fb)
-		return;
-	gbo = drm_gem_vram_of_gem(old_state->fb->obj[0]);
-	drm_gem_vram_unpin(gbo);
-}
-
 static const struct drm_simple_display_pipe_funcs bochs_pipe_funcs = {
 	.enable	    = bochs_pipe_enable,
 	.update	    = bochs_pipe_update,
-	.prepare_fb = bochs_pipe_prepare_fb,
-	.cleanup_fb = bochs_pipe_cleanup_fb,
+	.prepare_fb = drm_gem_vram_simple_display_pipe_prepare_fb,
+	.cleanup_fb = drm_gem_vram_simple_display_pipe_cleanup_fb,
 };
 
 static int bochs_connector_get_modes(struct drm_connector *connector)
diff --git a/drivers/gpu/drm/bochs/bochs_mm.c b/drivers/gpu/drm/bochs/bochs_mm.c
index 8f9bb886f7ad..1b74f530b07c 100644
--- a/drivers/gpu/drm/bochs/bochs_mm.c
+++ b/drivers/gpu/drm/bochs/bochs_mm.c
@@ -11,8 +11,7 @@ int bochs_mm_init(struct bochs_device *bochs)
 	struct drm_vram_mm *vmm;
 
 	vmm = drm_vram_helper_alloc_mm(bochs->dev, bochs->fb_base,
-				       bochs->fb_size,
-				       &drm_gem_vram_mm_funcs);
+				       bochs->fb_size);
 	return PTR_ERR_OR_ZERO(vmm);
 }
 
diff --git a/drivers/gpu/drm/bridge/Kconfig b/drivers/gpu/drm/bridge/Kconfig
index 1cc9f502c1f2..34362976cd6f 100644
--- a/drivers/gpu/drm/bridge/Kconfig
+++ b/drivers/gpu/drm/bridge/Kconfig
@@ -87,8 +87,7 @@ config DRM_SIL_SII8620
 	depends on OF
 	select DRM_KMS_HELPER
 	imply EXTCON
-	select INPUT
-	select RC_CORE
+	depends on RC_CORE || !RC_CORE
 	help
 	  Silicon Image SII8620 HDMI/MHL bridge chip driver.
 
diff --git a/drivers/gpu/drm/bridge/analogix-anx78xx.c b/drivers/gpu/drm/bridge/analogix-anx78xx.c
index 3c7cc5af735c..274989f96a91 100644
--- a/drivers/gpu/drm/bridge/analogix-anx78xx.c
+++ b/drivers/gpu/drm/bridge/analogix-anx78xx.c
@@ -19,6 +19,7 @@
 #include <linux/types.h>
 
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_crtc.h>
 #include <drm/drm_dp_helper.h>
 #include <drm/drm_edid.h>
@@ -38,12 +39,20 @@
 #define AUX_CH_BUFFER_SIZE	16
 #define AUX_WAIT_TIMEOUT_MS	15
 
-static const u8 anx78xx_i2c_addresses[] = {
-	[I2C_IDX_TX_P0] = TX_P0,
-	[I2C_IDX_TX_P1] = TX_P1,
-	[I2C_IDX_TX_P2] = TX_P2,
-	[I2C_IDX_RX_P0] = RX_P0,
-	[I2C_IDX_RX_P1] = RX_P1,
+static const u8 anx7808_i2c_addresses[] = {
+	[I2C_IDX_TX_P0] = 0x78,
+	[I2C_IDX_TX_P1] = 0x7a,
+	[I2C_IDX_TX_P2] = 0x72,
+	[I2C_IDX_RX_P0] = 0x7e,
+	[I2C_IDX_RX_P1] = 0x80,
+};
+
+static const u8 anx781x_i2c_addresses[] = {
+	[I2C_IDX_TX_P0] = 0x70,
+	[I2C_IDX_TX_P1] = 0x7a,
+	[I2C_IDX_TX_P2] = 0x72,
+	[I2C_IDX_RX_P0] = 0x7e,
+	[I2C_IDX_RX_P1] = 0x80,
 };
 
 struct anx78xx_platform_data {
@@ -62,7 +71,6 @@ struct anx78xx {
 	struct i2c_client *client;
 	struct edid *edid;
 	struct drm_connector connector;
-	struct drm_dp_link link;
 	struct anx78xx_platform_data pdata;
 	struct mutex lock;
 
@@ -715,7 +723,9 @@ static int anx78xx_init_pdata(struct anx78xx *anx78xx)
 	/* 1.0V digital core power regulator  */
 	pdata->dvdd10 = devm_regulator_get(dev, "dvdd10");
 	if (IS_ERR(pdata->dvdd10)) {
-		DRM_ERROR("DVDD10 regulator not found\n");
+		if (PTR_ERR(pdata->dvdd10) != -EPROBE_DEFER)
+			DRM_ERROR("DVDD10 regulator not found\n");
+
 		return PTR_ERR(pdata->dvdd10);
 	}
 
@@ -737,7 +747,7 @@ static int anx78xx_init_pdata(struct anx78xx *anx78xx)
 
 static int anx78xx_dp_link_training(struct anx78xx *anx78xx)
 {
-	u8 dp_bw, value;
+	u8 dp_bw, dpcd[2];
 	int err;
 
 	err = regmap_write(anx78xx->map[I2C_IDX_RX_P0], SP_HDMI_MUTE_CTRL_REG,
@@ -790,18 +800,34 @@ static int anx78xx_dp_link_training(struct anx78xx *anx78xx)
 	if (err)
 		return err;
 
-	/* Check link capabilities */
-	err = drm_dp_link_probe(&anx78xx->aux, &anx78xx->link);
-	if (err < 0) {
-		DRM_ERROR("Failed to probe link capabilities: %d\n", err);
-		return err;
-	}
+	/*
+	 * Power up the sink (DP_SET_POWER register is only available on DPCD
+	 * v1.1 and later).
+	 */
+	if (anx78xx->dpcd[DP_DPCD_REV] >= 0x11) {
+		err = drm_dp_dpcd_readb(&anx78xx->aux, DP_SET_POWER, &dpcd[0]);
+		if (err < 0) {
+			DRM_ERROR("Failed to read DP_SET_POWER register: %d\n",
+				  err);
+			return err;
+		}
 
-	/* Power up the sink */
-	err = drm_dp_link_power_up(&anx78xx->aux, &anx78xx->link);
-	if (err < 0) {
-		DRM_ERROR("Failed to power up DisplayPort link: %d\n", err);
-		return err;
+		dpcd[0] &= ~DP_SET_POWER_MASK;
+		dpcd[0] |= DP_SET_POWER_D0;
+
+		err = drm_dp_dpcd_writeb(&anx78xx->aux, DP_SET_POWER, dpcd[0]);
+		if (err < 0) {
+			DRM_ERROR("Failed to power up DisplayPort link: %d\n",
+				  err);
+			return err;
+		}
+
+		/*
+		 * According to the DP 1.1 specification, a "Sink Device must
+		 * exit the power saving state within 1 ms" (Section 2.5.3.1,
+		 * Table 5-52, "Sink Control Field" (register 0x600).
+		 */
+		usleep_range(1000, 2000);
 	}
 
 	/* Possibly enable downspread on the sink */
@@ -840,15 +866,22 @@ static int anx78xx_dp_link_training(struct anx78xx *anx78xx)
 	if (err)
 		return err;
 
-	value = drm_dp_link_rate_to_bw_code(anx78xx->link.rate);
+	dpcd[0] = drm_dp_max_link_rate(anx78xx->dpcd);
+	dpcd[0] = drm_dp_link_rate_to_bw_code(dpcd[0]);
 	err = regmap_write(anx78xx->map[I2C_IDX_TX_P0],
-			   SP_DP_MAIN_LINK_BW_SET_REG, value);
+			   SP_DP_MAIN_LINK_BW_SET_REG, dpcd[0]);
 	if (err)
 		return err;
 
-	err = drm_dp_link_configure(&anx78xx->aux, &anx78xx->link);
+	dpcd[1] = drm_dp_max_lane_count(anx78xx->dpcd);
+
+	if (drm_dp_enhanced_frame_cap(anx78xx->dpcd))
+		dpcd[1] |= DP_LANE_COUNT_ENHANCED_FRAME_EN;
+
+	err = drm_dp_dpcd_write(&anx78xx->aux, DP_LINK_BW_SET, dpcd,
+				sizeof(dpcd));
 	if (err < 0) {
-		DRM_ERROR("Failed to configure DisplayPort link: %d\n", err);
+		DRM_ERROR("Failed to configure link: %d\n", err);
 		return err;
 	}
 
@@ -1301,6 +1334,7 @@ static const struct regmap_config anx78xx_regmap_config = {
 };
 
 static const u16 anx78xx_chipid_list[] = {
+	0x7808,
 	0x7812,
 	0x7814,
 	0x7818,
@@ -1312,6 +1346,7 @@ static int anx78xx_i2c_probe(struct i2c_client *client,
 	struct anx78xx *anx78xx;
 	struct anx78xx_platform_data *pdata;
 	unsigned int i, idl, idh, version;
+	const u8 *i2c_addresses;
 	bool found = false;
 	int err;
 
@@ -1332,7 +1367,9 @@ static int anx78xx_i2c_probe(struct i2c_client *client,
 
 	err = anx78xx_init_pdata(anx78xx);
 	if (err) {
-		DRM_ERROR("Failed to initialize pdata: %d\n", err);
+		if (err != -EPROBE_DEFER)
+			DRM_ERROR("Failed to initialize pdata: %d\n", err);
+
 		return err;
 	}
 
@@ -1349,22 +1386,26 @@ static int anx78xx_i2c_probe(struct i2c_client *client,
 	}
 
 	/* Map slave addresses of ANX7814 */
+	i2c_addresses = device_get_match_data(&client->dev);
 	for (i = 0; i < I2C_NUM_ADDRESSES; i++) {
-		anx78xx->i2c_dummy[i] = i2c_new_dummy(client->adapter,
-						anx78xx_i2c_addresses[i] >> 1);
-		if (!anx78xx->i2c_dummy[i]) {
-			err = -ENOMEM;
-			DRM_ERROR("Failed to reserve I2C bus %02x\n",
-				  anx78xx_i2c_addresses[i]);
+		struct i2c_client *i2c_dummy;
+
+		i2c_dummy = i2c_new_dummy_device(client->adapter,
+						 i2c_addresses[i] >> 1);
+		if (IS_ERR(i2c_dummy)) {
+			err = PTR_ERR(i2c_dummy);
+			DRM_ERROR("Failed to reserve I2C bus %02x: %d\n",
+				  i2c_addresses[i], err);
 			goto err_unregister_i2c;
 		}
 
+		anx78xx->i2c_dummy[i] = i2c_dummy;
 		anx78xx->map[i] = devm_regmap_init_i2c(anx78xx->i2c_dummy[i],
 						       &anx78xx_regmap_config);
 		if (IS_ERR(anx78xx->map[i])) {
 			err = PTR_ERR(anx78xx->map[i]);
 			DRM_ERROR("Failed regmap initialization %02x\n",
-				  anx78xx_i2c_addresses[i]);
+				  i2c_addresses[i]);
 			goto err_unregister_i2c;
 		}
 	}
@@ -1463,7 +1504,10 @@ MODULE_DEVICE_TABLE(i2c, anx78xx_id);
 
 #if IS_ENABLED(CONFIG_OF)
 static const struct of_device_id anx78xx_match_table[] = {
-	{ .compatible = "analogix,anx7814", },
+	{ .compatible = "analogix,anx7808", .data = anx7808_i2c_addresses },
+	{ .compatible = "analogix,anx7812", .data = anx781x_i2c_addresses },
+	{ .compatible = "analogix,anx7814", .data = anx781x_i2c_addresses },
+	{ .compatible = "analogix,anx7818", .data = anx781x_i2c_addresses },
 	{ /* sentinel */ },
 };
 MODULE_DEVICE_TABLE(of, anx78xx_match_table);
diff --git a/drivers/gpu/drm/bridge/analogix-anx78xx.h b/drivers/gpu/drm/bridge/analogix-anx78xx.h
index 25e063bcecbc..55d6c2109740 100644
--- a/drivers/gpu/drm/bridge/analogix-anx78xx.h
+++ b/drivers/gpu/drm/bridge/analogix-anx78xx.h
@@ -6,15 +6,8 @@
 #ifndef __ANX78xx_H
 #define __ANX78xx_H
 
-#define TX_P0				0x70
-#define TX_P1				0x7a
-#define TX_P2				0x72
-
-#define RX_P0				0x7e
-#define RX_P1				0x80
-
 /***************************************************************/
-/* Register definition of device address 0x7e                  */
+/* Register definitions for RX_PO                              */
 /***************************************************************/
 
 /*
@@ -171,7 +164,7 @@
 #define SP_VSI_RCVD			BIT(1)
 
 /***************************************************************/
-/* Register definition of device address 0x80                  */
+/* Register definitions for RX_P1                              */
 /***************************************************************/
 
 /* HDCP BCAPS Shadow Register */
@@ -217,7 +210,7 @@
 #define SP_SET_AVMUTE			BIT(0)
 
 /***************************************************************/
-/* Register definition of device address 0x70                  */
+/* Register definitions for TX_P0                              */
 /***************************************************************/
 
 /* HDCP Status Register */
@@ -451,7 +444,7 @@
 #define SP_DP_BUF_DATA0_REG		0xf0
 
 /***************************************************************/
-/* Register definition of device address 0x72                  */
+/* Register definitions for TX_P2                              */
 /***************************************************************/
 
 /*
@@ -674,7 +667,7 @@
 #define SP_INT_CTRL_REG			0xff
 
 /***************************************************************/
-/* Register definition of device address 0x7a                  */
+/* Register definitions for TX_P1                              */
 /***************************************************************/
 
 /* DP TX Link Training Control Register */
diff --git a/drivers/gpu/drm/bridge/analogix/analogix_dp_core.c b/drivers/gpu/drm/bridge/analogix/analogix_dp_core.c
index 22885dceaa17..bb411fe52ae8 100644
--- a/drivers/gpu/drm/bridge/analogix/analogix_dp_core.c
+++ b/drivers/gpu/drm/bridge/analogix/analogix_dp_core.c
@@ -21,6 +21,7 @@
 #include <drm/bridge/analogix_dp.h>
 #include <drm/drm_atomic.h>
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_crtc.h>
 #include <drm/drm_device.h>
 #include <drm/drm_panel.h>
diff --git a/drivers/gpu/drm/bridge/cdns-dsi.c b/drivers/gpu/drm/bridge/cdns-dsi.c
index 6166dca6be81..3a5bd4e7fd1e 100644
--- a/drivers/gpu/drm/bridge/cdns-dsi.c
+++ b/drivers/gpu/drm/bridge/cdns-dsi.c
@@ -956,7 +956,8 @@ static int cdns_dsi_attach(struct mipi_dsi_host *host,
 
 	panel = of_drm_find_panel(np);
 	if (!IS_ERR(panel)) {
-		bridge = drm_panel_bridge_add(panel, DRM_MODE_CONNECTOR_DSI);
+		bridge = drm_panel_bridge_add_typed(panel,
+						    DRM_MODE_CONNECTOR_DSI);
 	} else {
 		bridge = of_drm_find_bridge(dev->dev.of_node);
 		if (!bridge)
diff --git a/drivers/gpu/drm/bridge/dumb-vga-dac.c b/drivers/gpu/drm/bridge/dumb-vga-dac.c
index 7aa789c35882..cc33dc411b9e 100644
--- a/drivers/gpu/drm/bridge/dumb-vga-dac.c
+++ b/drivers/gpu/drm/bridge/dumb-vga-dac.c
@@ -12,6 +12,7 @@
 #include <linux/regulator/consumer.h>
 
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_crtc.h>
 #include <drm/drm_print.h>
 #include <drm/drm_probe_helper.h>
diff --git a/drivers/gpu/drm/bridge/lvds-encoder.c b/drivers/gpu/drm/bridge/lvds-encoder.c
index 2ab2c234f26c..e2132a8d5106 100644
--- a/drivers/gpu/drm/bridge/lvds-encoder.c
+++ b/drivers/gpu/drm/bridge/lvds-encoder.c
@@ -106,7 +106,8 @@ static int lvds_encoder_probe(struct platform_device *pdev)
 	}
 
 	lvds_encoder->panel_bridge =
-		devm_drm_panel_bridge_add(dev, panel, DRM_MODE_CONNECTOR_LVDS);
+		devm_drm_panel_bridge_add_typed(dev, panel,
+						DRM_MODE_CONNECTOR_LVDS);
 	if (IS_ERR(lvds_encoder->panel_bridge))
 		return PTR_ERR(lvds_encoder->panel_bridge);
 
diff --git a/drivers/gpu/drm/bridge/megachips-stdpxxxx-ge-b850v3-fw.c b/drivers/gpu/drm/bridge/megachips-stdpxxxx-ge-b850v3-fw.c
index 6e81e5db57f2..e8a49f6146c6 100644
--- a/drivers/gpu/drm/bridge/megachips-stdpxxxx-ge-b850v3-fw.c
+++ b/drivers/gpu/drm/bridge/megachips-stdpxxxx-ge-b850v3-fw.c
@@ -25,6 +25,7 @@
 
 #include <drm/drm_atomic.h>
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_edid.h>
 #include <drm/drm_print.h>
 #include <drm/drm_probe_helper.h>
diff --git a/drivers/gpu/drm/bridge/nxp-ptn3460.c b/drivers/gpu/drm/bridge/nxp-ptn3460.c
index d4a1cc5052c3..57ff01339559 100644
--- a/drivers/gpu/drm/bridge/nxp-ptn3460.c
+++ b/drivers/gpu/drm/bridge/nxp-ptn3460.c
@@ -11,6 +11,7 @@
 #include <linux/module.h>
 #include <linux/of.h>
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_crtc.h>
 #include <drm/drm_edid.h>
 #include <drm/drm_of.h>
diff --git a/drivers/gpu/drm/bridge/panel.c b/drivers/gpu/drm/bridge/panel.c
index b12ae3a4c5f1..f4e293e7cf64 100644
--- a/drivers/gpu/drm/bridge/panel.c
+++ b/drivers/gpu/drm/bridge/panel.c
@@ -5,6 +5,7 @@
  */
 
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_connector.h>
 #include <drm/drm_encoder.h>
 #include <drm/drm_modeset_helper_vtables.h>
@@ -133,8 +134,6 @@ static const struct drm_bridge_funcs panel_bridge_bridge_funcs = {
  * just calls the appropriate functions from &drm_panel.
  *
  * @panel: The drm_panel being wrapped.  Must be non-NULL.
- * @connector_type: The DRM_MODE_CONNECTOR_* for the connector to be
- * created.
  *
  * For drivers converting from directly using drm_panel: The expected
  * usage pattern is that during either encoder module probe or DSI
@@ -148,11 +147,37 @@ static const struct drm_bridge_funcs panel_bridge_bridge_funcs = {
  * drm_mode_config_cleanup() if the bridge has already been attached), then
  * drm_panel_bridge_remove() to free it.
  *
+ * The connector type is set to @panel->connector_type, which must be set to a
+ * known type. Calling this function with a panel whose connector type is
+ * DRM_MODE_CONNECTOR_Unknown will return NULL.
+ *
  * See devm_drm_panel_bridge_add() for an automatically manged version of this
  * function.
  */
-struct drm_bridge *drm_panel_bridge_add(struct drm_panel *panel,
-					u32 connector_type)
+struct drm_bridge *drm_panel_bridge_add(struct drm_panel *panel)
+{
+	if (WARN_ON(panel->connector_type == DRM_MODE_CONNECTOR_Unknown))
+		return NULL;
+
+	return drm_panel_bridge_add_typed(panel, panel->connector_type);
+}
+EXPORT_SYMBOL(drm_panel_bridge_add);
+
+/**
+ * drm_panel_bridge_add_typed - Creates a &drm_bridge and &drm_connector with
+ * an explicit connector type.
+ * @panel: The drm_panel being wrapped.  Must be non-NULL.
+ * @connector_type: The connector type (DRM_MODE_CONNECTOR_*)
+ *
+ * This is just like drm_panel_bridge_add(), but forces the connector type to
+ * @connector_type instead of infering it from the panel.
+ *
+ * This function is deprecated and should not be used in new drivers. Use
+ * drm_panel_bridge_add() instead, and fix panel drivers as necessary if they
+ * don't report a connector type.
+ */
+struct drm_bridge *drm_panel_bridge_add_typed(struct drm_panel *panel,
+					      u32 connector_type)
 {
 	struct panel_bridge *panel_bridge;
 
@@ -176,7 +201,7 @@ struct drm_bridge *drm_panel_bridge_add(struct drm_panel *panel,
 
 	return &panel_bridge->bridge;
 }
-EXPORT_SYMBOL(drm_panel_bridge_add);
+EXPORT_SYMBOL(drm_panel_bridge_add_typed);
 
 /**
  * drm_panel_bridge_remove - Unregisters and frees a drm_bridge
@@ -213,15 +238,38 @@ static void devm_drm_panel_bridge_release(struct device *dev, void *res)
  * that just calls the appropriate functions from &drm_panel.
  * @dev: device to tie the bridge lifetime to
  * @panel: The drm_panel being wrapped.  Must be non-NULL.
- * @connector_type: The DRM_MODE_CONNECTOR_* for the connector to be
- * created.
  *
  * This is the managed version of drm_panel_bridge_add() which automatically
  * calls drm_panel_bridge_remove() when @dev is unbound.
  */
 struct drm_bridge *devm_drm_panel_bridge_add(struct device *dev,
-					     struct drm_panel *panel,
-					     u32 connector_type)
+					     struct drm_panel *panel)
+{
+	if (WARN_ON(panel->connector_type == DRM_MODE_CONNECTOR_Unknown))
+		return NULL;
+
+	return devm_drm_panel_bridge_add_typed(dev, panel,
+					       panel->connector_type);
+}
+EXPORT_SYMBOL(devm_drm_panel_bridge_add);
+
+/**
+ * devm_drm_panel_bridge_add_typed - Creates a managed &drm_bridge and
+ * &drm_connector with an explicit connector type.
+ * @dev: device to tie the bridge lifetime to
+ * @panel: The drm_panel being wrapped.  Must be non-NULL.
+ * @connector_type: The connector type (DRM_MODE_CONNECTOR_*)
+ *
+ * This is just like devm_drm_panel_bridge_add(), but forces the connector type
+ * to @connector_type instead of infering it from the panel.
+ *
+ * This function is deprecated and should not be used in new drivers. Use
+ * devm_drm_panel_bridge_add() instead, and fix panel drivers as necessary if
+ * they don't report a connector type.
+ */
+struct drm_bridge *devm_drm_panel_bridge_add_typed(struct device *dev,
+						   struct drm_panel *panel,
+						   u32 connector_type)
 {
 	struct drm_bridge **ptr, *bridge;
 
@@ -230,7 +278,7 @@ struct drm_bridge *devm_drm_panel_bridge_add(struct device *dev,
 	if (!ptr)
 		return ERR_PTR(-ENOMEM);
 
-	bridge = drm_panel_bridge_add(panel, connector_type);
+	bridge = drm_panel_bridge_add_typed(panel, connector_type);
 	if (!IS_ERR(bridge)) {
 		*ptr = bridge;
 		devres_add(dev, ptr);
@@ -240,4 +288,4 @@ struct drm_bridge *devm_drm_panel_bridge_add(struct device *dev,
 
 	return bridge;
 }
-EXPORT_SYMBOL(devm_drm_panel_bridge_add);
+EXPORT_SYMBOL(devm_drm_panel_bridge_add_typed);
diff --git a/drivers/gpu/drm/bridge/parade-ps8622.c b/drivers/gpu/drm/bridge/parade-ps8622.c
index 93c68e2e9484..b7a72dfdcac3 100644
--- a/drivers/gpu/drm/bridge/parade-ps8622.c
+++ b/drivers/gpu/drm/bridge/parade-ps8622.c
@@ -17,6 +17,7 @@
 #include <linux/regulator/consumer.h>
 
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_crtc.h>
 #include <drm/drm_of.h>
 #include <drm/drm_panel.h>
diff --git a/drivers/gpu/drm/bridge/sii902x.c b/drivers/gpu/drm/bridge/sii902x.c
index 38f75ac580df..b70e8c5cf2e1 100644
--- a/drivers/gpu/drm/bridge/sii902x.c
+++ b/drivers/gpu/drm/bridge/sii902x.c
@@ -20,6 +20,7 @@
 #include <linux/clk.h>
 
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_drv.h>
 #include <drm/drm_edid.h>
 #include <drm/drm_print.h>
diff --git a/drivers/gpu/drm/bridge/sii9234.c b/drivers/gpu/drm/bridge/sii9234.c
index 25d4ad8c7ad6..f81f81b7051f 100644
--- a/drivers/gpu/drm/bridge/sii9234.c
+++ b/drivers/gpu/drm/bridge/sii9234.c
@@ -13,6 +13,7 @@
  *    Dharam Kumar <dharam.kr@samsung.com>
  */
 #include <drm/bridge/mhl.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_crtc.h>
 #include <drm/drm_edid.h>
 
@@ -841,39 +842,28 @@ static int sii9234_init_resources(struct sii9234 *ctx,
 
 	ctx->client[I2C_MHL] = client;
 
-	ctx->client[I2C_TPI] = i2c_new_dummy(adapter, I2C_TPI_ADDR);
-	if (!ctx->client[I2C_TPI]) {
+	ctx->client[I2C_TPI] = devm_i2c_new_dummy_device(&client->dev, adapter,
+							 I2C_TPI_ADDR);
+	if (IS_ERR(ctx->client[I2C_TPI])) {
 		dev_err(ctx->dev, "failed to create TPI client\n");
-		return -ENODEV;
+		return PTR_ERR(ctx->client[I2C_TPI]);
 	}
 
-	ctx->client[I2C_HDMI] = i2c_new_dummy(adapter, I2C_HDMI_ADDR);
-	if (!ctx->client[I2C_HDMI]) {
+	ctx->client[I2C_HDMI] = devm_i2c_new_dummy_device(&client->dev, adapter,
+							  I2C_HDMI_ADDR);
+	if (IS_ERR(ctx->client[I2C_HDMI])) {
 		dev_err(ctx->dev, "failed to create HDMI RX client\n");
-		goto fail_tpi;
+		return PTR_ERR(ctx->client[I2C_HDMI]);
 	}
 
-	ctx->client[I2C_CBUS] = i2c_new_dummy(adapter, I2C_CBUS_ADDR);
-	if (!ctx->client[I2C_CBUS]) {
+	ctx->client[I2C_CBUS] = devm_i2c_new_dummy_device(&client->dev, adapter,
+							  I2C_CBUS_ADDR);
+	if (IS_ERR(ctx->client[I2C_CBUS])) {
 		dev_err(ctx->dev, "failed to create CBUS client\n");
-		goto fail_hdmi;
+		return PTR_ERR(ctx->client[I2C_CBUS]);
 	}
 
 	return 0;
-
-fail_hdmi:
-	i2c_unregister_device(ctx->client[I2C_HDMI]);
-fail_tpi:
-	i2c_unregister_device(ctx->client[I2C_TPI]);
-
-	return -ENODEV;
-}
-
-static void sii9234_deinit_resources(struct sii9234 *ctx)
-{
-	i2c_unregister_device(ctx->client[I2C_CBUS]);
-	i2c_unregister_device(ctx->client[I2C_HDMI]);
-	i2c_unregister_device(ctx->client[I2C_TPI]);
 }
 
 static inline struct sii9234 *bridge_to_sii9234(struct drm_bridge *bridge)
@@ -950,7 +940,6 @@ static int sii9234_remove(struct i2c_client *client)
 
 	sii9234_cable_out(ctx);
 	drm_bridge_remove(&ctx->bridge);
-	sii9234_deinit_resources(ctx);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/bridge/sil-sii8620.c b/drivers/gpu/drm/bridge/sil-sii8620.c
index bd3165ee5354..4c0eef406eb1 100644
--- a/drivers/gpu/drm/bridge/sil-sii8620.c
+++ b/drivers/gpu/drm/bridge/sil-sii8620.c
@@ -9,6 +9,7 @@
 #include <asm/unaligned.h>
 
 #include <drm/bridge/mhl.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_crtc.h>
 #include <drm/drm_edid.h>
 #include <drm/drm_encoder.h>
@@ -1759,10 +1760,8 @@ static bool sii8620_rcp_consume(struct sii8620 *ctx, u8 scancode)
 
 	scancode &= MHL_RCP_KEY_ID_MASK;
 
-	if (!ctx->rc_dev) {
-		dev_dbg(ctx->dev, "RCP input device not initialized\n");
+	if (!IS_ENABLED(CONFIG_RC_CORE) || !ctx->rc_dev)
 		return false;
-	}
 
 	if (pressed)
 		rc_keydown(ctx->rc_dev, RC_PROTO_CEC, scancode, 0);
@@ -2099,6 +2098,9 @@ static void sii8620_init_rcp_input_dev(struct sii8620 *ctx)
 	struct rc_dev *rc_dev;
 	int ret;
 
+	if (!IS_ENABLED(CONFIG_RC_CORE))
+		return;
+
 	rc_dev = rc_allocate_device(RC_DRIVER_SCANCODE);
 	if (!rc_dev) {
 		dev_err(ctx->dev, "Failed to allocate RC device\n");
@@ -2213,6 +2215,9 @@ static void sii8620_detach(struct drm_bridge *bridge)
 {
 	struct sii8620 *ctx = bridge_to_sii8620(bridge);
 
+	if (!IS_ENABLED(CONFIG_RC_CORE))
+		return;
+
 	rc_unregister_device(ctx->rc_dev);
 }
 
diff --git a/drivers/gpu/drm/bridge/synopsys/dw-hdmi-cec.c b/drivers/gpu/drm/bridge/synopsys/dw-hdmi-cec.c
index ac1e001d0882..70ab4fbdc23e 100644
--- a/drivers/gpu/drm/bridge/synopsys/dw-hdmi-cec.c
+++ b/drivers/gpu/drm/bridge/synopsys/dw-hdmi-cec.c
@@ -285,7 +285,7 @@ static int dw_hdmi_cec_probe(struct platform_device *pdev)
 
 	ret = cec_register_adapter(cec->adap, pdev->dev.parent);
 	if (ret < 0) {
-		cec_notifier_cec_adap_unregister(cec->notify);
+		cec_notifier_cec_adap_unregister(cec->notify, cec->adap);
 		return ret;
 	}
 
@@ -302,7 +302,7 @@ static int dw_hdmi_cec_remove(struct platform_device *pdev)
 {
 	struct dw_hdmi_cec *cec = platform_get_drvdata(pdev);
 
-	cec_notifier_cec_adap_unregister(cec->notify);
+	cec_notifier_cec_adap_unregister(cec->notify, cec->adap);
 	cec_unregister_adapter(cec->adap);
 
 	return 0;
diff --git a/drivers/gpu/drm/bridge/synopsys/dw-hdmi-i2s-audio.c b/drivers/gpu/drm/bridge/synopsys/dw-hdmi-i2s-audio.c
index 1d15cf9b6821..d7e65c869415 100644
--- a/drivers/gpu/drm/bridge/synopsys/dw-hdmi-i2s-audio.c
+++ b/drivers/gpu/drm/bridge/synopsys/dw-hdmi-i2s-audio.c
@@ -102,6 +102,7 @@ static int dw_hdmi_i2s_hw_params(struct device *dev, void *data,
 	}
 
 	dw_hdmi_set_sample_rate(hdmi, hparms->sample_rate);
+	dw_hdmi_set_channel_status(hdmi, hparms->iec.status);
 	dw_hdmi_set_channel_count(hdmi, hparms->channels);
 	dw_hdmi_set_channel_allocation(hdmi, hparms->cea.channel_allocation);
 
@@ -109,6 +110,14 @@ static int dw_hdmi_i2s_hw_params(struct device *dev, void *data,
 	hdmi_write(audio, conf0, HDMI_AUD_CONF0);
 	hdmi_write(audio, conf1, HDMI_AUD_CONF1);
 
+	return 0;
+}
+
+static int dw_hdmi_i2s_audio_startup(struct device *dev, void *data)
+{
+	struct dw_hdmi_i2s_audio_data *audio = data;
+	struct dw_hdmi *hdmi = audio->hdmi;
+
 	dw_hdmi_audio_enable(hdmi);
 
 	return 0;
@@ -151,11 +160,23 @@ static int dw_hdmi_i2s_get_dai_id(struct snd_soc_component *component,
 	return -EINVAL;
 }
 
+static int dw_hdmi_i2s_hook_plugged_cb(struct device *dev, void *data,
+				       hdmi_codec_plugged_cb fn,
+				       struct device *codec_dev)
+{
+	struct dw_hdmi_i2s_audio_data *audio = data;
+	struct dw_hdmi *hdmi = audio->hdmi;
+
+	return dw_hdmi_set_plugged_cb(hdmi, fn, codec_dev);
+}
+
 static struct hdmi_codec_ops dw_hdmi_i2s_ops = {
 	.hw_params	= dw_hdmi_i2s_hw_params,
+	.audio_startup  = dw_hdmi_i2s_audio_startup,
 	.audio_shutdown	= dw_hdmi_i2s_audio_shutdown,
 	.get_eld	= dw_hdmi_i2s_get_eld,
 	.get_dai_id	= dw_hdmi_i2s_get_dai_id,
+	.hook_plugged_cb = dw_hdmi_i2s_hook_plugged_cb,
 };
 
 static int snd_dw_hdmi_probe(struct platform_device *pdev)
diff --git a/drivers/gpu/drm/bridge/synopsys/dw-hdmi.c b/drivers/gpu/drm/bridge/synopsys/dw-hdmi.c
index 521d689413c8..67fca439bbfb 100644
--- a/drivers/gpu/drm/bridge/synopsys/dw-hdmi.c
+++ b/drivers/gpu/drm/bridge/synopsys/dw-hdmi.c
@@ -25,7 +25,9 @@
 #include <uapi/linux/videodev2.h>
 
 #include <drm/bridge/dw_hdmi.h>
+#include <drm/drm_atomic.h>
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_edid.h>
 #include <drm/drm_of.h>
 #include <drm/drm_print.h>
@@ -36,6 +38,7 @@
 #include "dw-hdmi-cec.h"
 #include "dw-hdmi.h"
 
+#define DDC_CI_ADDR		0x37
 #define DDC_SEGMENT_ADDR	0x30
 
 #define HDMI_EDID_LEN		512
@@ -191,6 +194,10 @@ struct dw_hdmi {
 
 	struct mutex cec_notifier_mutex;
 	struct cec_notifier *cec_notifier;
+
+	hdmi_codec_plugged_cb plugged_cb;
+	struct device *codec_dev;
+	enum drm_connector_status last_connector_result;
 };
 
 #define HDMI_IH_PHY_STAT0_RX_SENSE \
@@ -215,6 +222,28 @@ static inline u8 hdmi_readb(struct dw_hdmi *hdmi, int offset)
 	return val;
 }
 
+static void handle_plugged_change(struct dw_hdmi *hdmi, bool plugged)
+{
+	if (hdmi->plugged_cb && hdmi->codec_dev)
+		hdmi->plugged_cb(hdmi->codec_dev, plugged);
+}
+
+int dw_hdmi_set_plugged_cb(struct dw_hdmi *hdmi, hdmi_codec_plugged_cb fn,
+			   struct device *codec_dev)
+{
+	bool plugged;
+
+	mutex_lock(&hdmi->mutex);
+	hdmi->plugged_cb = fn;
+	hdmi->codec_dev = codec_dev;
+	plugged = hdmi->last_connector_result == connector_status_connected;
+	handle_plugged_change(hdmi, plugged);
+	mutex_unlock(&hdmi->mutex);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(dw_hdmi_set_plugged_cb);
+
 static void hdmi_modb(struct dw_hdmi *hdmi, u8 data, u8 mask, unsigned reg)
 {
 	regmap_update_bits(hdmi->regm, reg << hdmi->reg_shift, mask, data);
@@ -398,6 +427,15 @@ static int dw_hdmi_i2c_xfer(struct i2c_adapter *adap,
 	u8 addr = msgs[0].addr;
 	int i, ret = 0;
 
+	if (addr == DDC_CI_ADDR)
+		/*
+		 * The internal I2C controller does not support the multi-byte
+		 * read and write operations needed for DDC/CI.
+		 * TOFIX: Blacklist the DDC/CI address until we filter out
+		 * unsupported I2C operations.
+		 */
+		return -EOPNOTSUPP;
+
 	dev_dbg(hdmi->dev, "xfer: num: %d, addr: %#x\n", num, addr);
 
 	for (i = 0; i < num; i++) {
@@ -580,6 +618,26 @@ static unsigned int hdmi_compute_n(unsigned int freq, unsigned long pixel_clk)
 	return n;
 }
 
+/*
+ * When transmitting IEC60958 linear PCM audio, these registers allow to
+ * configure the channel status information of all the channel status
+ * bits in the IEC60958 frame. For the moment this configuration is only
+ * used when the I2S audio interface, General Purpose Audio (GPA),
+ * or AHB audio DMA (AHBAUDDMA) interface is active
+ * (for S/PDIF interface this information comes from the stream).
+ */
+void dw_hdmi_set_channel_status(struct dw_hdmi *hdmi,
+				u8 *channel_status)
+{
+	/*
+	 * Set channel status register for frequency and word length.
+	 * Use default values for other registers.
+	 */
+	hdmi_writeb(hdmi, channel_status[3], HDMI_FC_AUDSCHNLS7);
+	hdmi_writeb(hdmi, channel_status[4], HDMI_FC_AUDSCHNLS8);
+}
+EXPORT_SYMBOL_GPL(dw_hdmi_set_channel_status);
+
 static void hdmi_set_clk_regenerator(struct dw_hdmi *hdmi,
 	unsigned long pixel_clk, unsigned int sample_rate)
 {
@@ -1712,6 +1770,41 @@ static void hdmi_config_vendor_specific_infoframe(struct dw_hdmi *hdmi,
 			HDMI_FC_DATAUTO0_VSD_MASK);
 }
 
+static void hdmi_config_drm_infoframe(struct dw_hdmi *hdmi)
+{
+	const struct drm_connector_state *conn_state = hdmi->connector.state;
+	struct hdmi_drm_infoframe frame;
+	u8 buffer[30];
+	ssize_t err;
+	int i;
+
+	if (!hdmi->plat_data->use_drm_infoframe)
+		return;
+
+	hdmi_modb(hdmi, HDMI_FC_PACKET_TX_EN_DRM_DISABLE,
+		  HDMI_FC_PACKET_TX_EN_DRM_MASK, HDMI_FC_PACKET_TX_EN);
+
+	err = drm_hdmi_infoframe_set_hdr_metadata(&frame, conn_state);
+	if (err < 0)
+		return;
+
+	err = hdmi_drm_infoframe_pack(&frame, buffer, sizeof(buffer));
+	if (err < 0) {
+		dev_err(hdmi->dev, "Failed to pack drm infoframe: %zd\n", err);
+		return;
+	}
+
+	hdmi_writeb(hdmi, frame.version, HDMI_FC_DRM_HB0);
+	hdmi_writeb(hdmi, frame.length, HDMI_FC_DRM_HB1);
+
+	for (i = 0; i < frame.length; i++)
+		hdmi_writeb(hdmi, buffer[4 + i], HDMI_FC_DRM_PB0 + i);
+
+	hdmi_writeb(hdmi, 1, HDMI_FC_DRM_UP);
+	hdmi_modb(hdmi, HDMI_FC_PACKET_TX_EN_DRM_ENABLE,
+		  HDMI_FC_PACKET_TX_EN_DRM_MASK, HDMI_FC_PACKET_TX_EN);
+}
+
 static void hdmi_av_composer(struct dw_hdmi *hdmi,
 			     const struct drm_display_mode *mode)
 {
@@ -2023,7 +2116,7 @@ static int dw_hdmi_setup(struct dw_hdmi *hdmi, struct drm_display_mode *mode)
 
 		/* HDMI Initialization Step E - Configure audio */
 		hdmi_clk_regenerator_update_pixel_clock(hdmi);
-		hdmi_enable_audio_clk(hdmi, true);
+		hdmi_enable_audio_clk(hdmi, hdmi->audio_enable);
 	}
 
 	/* not for DVI mode */
@@ -2033,6 +2126,7 @@ static int dw_hdmi_setup(struct dw_hdmi *hdmi, struct drm_display_mode *mode)
 		/* HDMI Initialization Step F - Configure AVI InfoFrame */
 		hdmi_config_AVI(hdmi, mode);
 		hdmi_config_vendor_specific_infoframe(hdmi, mode);
+		hdmi_config_drm_infoframe(hdmi);
 	} else {
 		dev_dbg(hdmi->dev, "%s DVI mode\n", __func__);
 	}
@@ -2161,6 +2255,7 @@ dw_hdmi_connector_detect(struct drm_connector *connector, bool force)
 {
 	struct dw_hdmi *hdmi = container_of(connector, struct dw_hdmi,
 					     connector);
+	enum drm_connector_status result;
 
 	mutex_lock(&hdmi->mutex);
 	hdmi->force = DRM_FORCE_UNSPECIFIED;
@@ -2168,7 +2263,18 @@ dw_hdmi_connector_detect(struct drm_connector *connector, bool force)
 	dw_hdmi_update_phy_mask(hdmi);
 	mutex_unlock(&hdmi->mutex);
 
-	return hdmi->phy.ops->read_hpd(hdmi, hdmi->phy.data);
+	result = hdmi->phy.ops->read_hpd(hdmi, hdmi->phy.data);
+
+	mutex_lock(&hdmi->mutex);
+	if (result != hdmi->last_connector_result) {
+		dev_dbg(hdmi->dev, "read_hpd result: %d", result);
+		handle_plugged_change(hdmi,
+				      result == connector_status_connected);
+		hdmi->last_connector_result = result;
+	}
+	mutex_unlock(&hdmi->mutex);
+
+	return result;
 }
 
 static int dw_hdmi_connector_get_modes(struct drm_connector *connector)
@@ -2199,6 +2305,45 @@ static int dw_hdmi_connector_get_modes(struct drm_connector *connector)
 	return ret;
 }
 
+static bool hdr_metadata_equal(const struct drm_connector_state *old_state,
+			       const struct drm_connector_state *new_state)
+{
+	struct drm_property_blob *old_blob = old_state->hdr_output_metadata;
+	struct drm_property_blob *new_blob = new_state->hdr_output_metadata;
+
+	if (!old_blob || !new_blob)
+		return old_blob == new_blob;
+
+	if (old_blob->length != new_blob->length)
+		return false;
+
+	return !memcmp(old_blob->data, new_blob->data, old_blob->length);
+}
+
+static int dw_hdmi_connector_atomic_check(struct drm_connector *connector,
+					  struct drm_atomic_state *state)
+{
+	struct drm_connector_state *old_state =
+		drm_atomic_get_old_connector_state(state, connector);
+	struct drm_connector_state *new_state =
+		drm_atomic_get_new_connector_state(state, connector);
+	struct drm_crtc *crtc = new_state->crtc;
+	struct drm_crtc_state *crtc_state;
+
+	if (!crtc)
+		return 0;
+
+	if (!hdr_metadata_equal(old_state, new_state)) {
+		crtc_state = drm_atomic_get_crtc_state(state, crtc);
+		if (IS_ERR(crtc_state))
+			return PTR_ERR(crtc_state);
+
+		crtc_state->mode_changed = true;
+	}
+
+	return 0;
+}
+
 static void dw_hdmi_connector_force(struct drm_connector *connector)
 {
 	struct dw_hdmi *hdmi = container_of(connector, struct dw_hdmi,
@@ -2223,6 +2368,7 @@ static const struct drm_connector_funcs dw_hdmi_connector_funcs = {
 
 static const struct drm_connector_helper_funcs dw_hdmi_connector_helper_funcs = {
 	.get_modes = dw_hdmi_connector_get_modes,
+	.atomic_check = dw_hdmi_connector_atomic_check,
 };
 
 static int dw_hdmi_bridge_attach(struct drm_bridge *bridge)
@@ -2243,6 +2389,10 @@ static int dw_hdmi_bridge_attach(struct drm_bridge *bridge)
 				    DRM_MODE_CONNECTOR_HDMIA,
 				    hdmi->ddc);
 
+	if (hdmi->version >= 0x200a && hdmi->plat_data->use_drm_infoframe)
+		drm_object_attach_property(&connector->base,
+			connector->dev->mode_config.hdr_output_metadata_property, 0);
+
 	drm_connector_attach_encoder(connector, encoder);
 
 	cec_fill_conn_info_from_drm(&conn_info, connector);
@@ -2619,6 +2769,7 @@ __dw_hdmi_probe(struct platform_device *pdev,
 	hdmi->rxsense = true;
 	hdmi->phy_mask = (u8)~(HDMI_PHY_HPD | HDMI_PHY_RX_SENSE);
 	hdmi->mc_clkdis = 0x7f;
+	hdmi->last_connector_result = connector_status_disconnected;
 
 	mutex_init(&hdmi->mutex);
 	mutex_init(&hdmi->audio_mutex);
diff --git a/drivers/gpu/drm/bridge/synopsys/dw-hdmi.h b/drivers/gpu/drm/bridge/synopsys/dw-hdmi.h
index 6988f12d89d9..1999db05bc3b 100644
--- a/drivers/gpu/drm/bridge/synopsys/dw-hdmi.h
+++ b/drivers/gpu/drm/bridge/synopsys/dw-hdmi.h
@@ -158,6 +158,8 @@
 #define HDMI_FC_SPDDEVICEINF                    0x1062
 #define HDMI_FC_AUDSCONF                        0x1063
 #define HDMI_FC_AUDSSTAT                        0x1064
+#define HDMI_FC_AUDSCHNLS7                      0x106e
+#define HDMI_FC_AUDSCHNLS8                      0x106f
 #define HDMI_FC_DATACH0FILL                     0x1070
 #define HDMI_FC_DATACH1FILL                     0x1071
 #define HDMI_FC_DATACH2FILL                     0x1072
@@ -252,6 +254,7 @@
 #define HDMI_FC_POL2                            0x10DB
 #define HDMI_FC_PRCONF                          0x10E0
 #define HDMI_FC_SCRAMBLER_CTRL                  0x10E1
+#define HDMI_FC_PACKET_TX_EN                    0x10E3
 
 #define HDMI_FC_GMD_STAT                        0x1100
 #define HDMI_FC_GMD_EN                          0x1101
@@ -287,6 +290,37 @@
 #define HDMI_FC_GMD_PB26                        0x111F
 #define HDMI_FC_GMD_PB27                        0x1120
 
+#define HDMI_FC_DRM_UP                          0x1167
+#define HDMI_FC_DRM_HB0                         0x1168
+#define HDMI_FC_DRM_HB1                         0x1169
+#define HDMI_FC_DRM_PB0                         0x116A
+#define HDMI_FC_DRM_PB1                         0x116B
+#define HDMI_FC_DRM_PB2                         0x116C
+#define HDMI_FC_DRM_PB3                         0x116D
+#define HDMI_FC_DRM_PB4                         0x116E
+#define HDMI_FC_DRM_PB5                         0x116F
+#define HDMI_FC_DRM_PB6                         0x1170
+#define HDMI_FC_DRM_PB7                         0x1171
+#define HDMI_FC_DRM_PB8                         0x1172
+#define HDMI_FC_DRM_PB9                         0x1173
+#define HDMI_FC_DRM_PB10                        0x1174
+#define HDMI_FC_DRM_PB11                        0x1175
+#define HDMI_FC_DRM_PB12                        0x1176
+#define HDMI_FC_DRM_PB13                        0x1177
+#define HDMI_FC_DRM_PB14                        0x1178
+#define HDMI_FC_DRM_PB15                        0x1179
+#define HDMI_FC_DRM_PB16                        0x117A
+#define HDMI_FC_DRM_PB17                        0x117B
+#define HDMI_FC_DRM_PB18                        0x117C
+#define HDMI_FC_DRM_PB19                        0x117D
+#define HDMI_FC_DRM_PB20                        0x117E
+#define HDMI_FC_DRM_PB21                        0x117F
+#define HDMI_FC_DRM_PB22                        0x1180
+#define HDMI_FC_DRM_PB23                        0x1181
+#define HDMI_FC_DRM_PB24                        0x1182
+#define HDMI_FC_DRM_PB25                        0x1183
+#define HDMI_FC_DRM_PB26                        0x1184
+
 #define HDMI_FC_DBGFORCE                        0x1200
 #define HDMI_FC_DBGAUD0CH0                      0x1201
 #define HDMI_FC_DBGAUD1CH0                      0x1202
@@ -742,6 +776,11 @@ enum {
 	HDMI_FC_PRCONF_OUTPUT_PR_FACTOR_MASK = 0x0F,
 	HDMI_FC_PRCONF_OUTPUT_PR_FACTOR_OFFSET = 0,
 
+/* FC_PACKET_TX_EN field values */
+	HDMI_FC_PACKET_TX_EN_DRM_MASK = 0x80,
+	HDMI_FC_PACKET_TX_EN_DRM_ENABLE = 0x80,
+	HDMI_FC_PACKET_TX_EN_DRM_DISABLE = 0x00,
+
 /* FC_AVICONF0-FC_AVICONF3 field values */
 	HDMI_FC_AVICONF0_PIX_FMT_MASK = 0x03,
 	HDMI_FC_AVICONF0_PIX_FMT_RGB = 0x00,
diff --git a/drivers/gpu/drm/bridge/synopsys/dw-mipi-dsi.c b/drivers/gpu/drm/bridge/synopsys/dw-mipi-dsi.c
index 675442bfc1bd..b6e793bb653c 100644
--- a/drivers/gpu/drm/bridge/synopsys/dw-mipi-dsi.c
+++ b/drivers/gpu/drm/bridge/synopsys/dw-mipi-dsi.c
@@ -316,7 +316,8 @@ static int dw_mipi_dsi_host_attach(struct mipi_dsi_host *host,
 		return ret;
 
 	if (panel) {
-		bridge = drm_panel_bridge_add(panel, DRM_MODE_CONNECTOR_DSI);
+		bridge = drm_panel_bridge_add_typed(panel,
+						    DRM_MODE_CONNECTOR_DSI);
 		if (IS_ERR(bridge))
 			return PTR_ERR(bridge);
 	}
@@ -981,7 +982,6 @@ __dw_mipi_dsi_probe(struct platform_device *pdev,
 	struct device *dev = &pdev->dev;
 	struct reset_control *apb_rst;
 	struct dw_mipi_dsi *dsi;
-	struct resource *res;
 	int ret;
 
 	dsi = devm_kzalloc(dev, sizeof(*dsi), GFP_KERNEL);
@@ -997,11 +997,7 @@ __dw_mipi_dsi_probe(struct platform_device *pdev,
 	}
 
 	if (!plat_data->base) {
-		res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
-		if (!res)
-			return ERR_PTR(-ENODEV);
-
-		dsi->base = devm_ioremap_resource(dev, res);
+		dsi->base = devm_platform_ioremap_resource(pdev, 0);
 		if (IS_ERR(dsi->base))
 			return ERR_PTR(-ENODEV);
 
diff --git a/drivers/gpu/drm/bridge/tc358764.c b/drivers/gpu/drm/bridge/tc358764.c
index 170f162ffa55..db298f550a5a 100644
--- a/drivers/gpu/drm/bridge/tc358764.c
+++ b/drivers/gpu/drm/bridge/tc358764.c
@@ -16,6 +16,7 @@
 #include <video/mipi_display.h>
 
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_crtc.h>
 #include <drm/drm_fb_helper.h>
 #include <drm/drm_mipi_dsi.h>
diff --git a/drivers/gpu/drm/bridge/tc358767.c b/drivers/gpu/drm/bridge/tc358767.c
index cebc8e620820..8029478ffebb 100644
--- a/drivers/gpu/drm/bridge/tc358767.c
+++ b/drivers/gpu/drm/bridge/tc358767.c
@@ -26,6 +26,7 @@
 #include <linux/slab.h>
 
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_dp_helper.h>
 #include <drm/drm_edid.h>
 #include <drm/drm_of.h>
@@ -228,7 +229,9 @@ static bool tc_test_pattern;
 module_param_named(test, tc_test_pattern, bool, 0644);
 
 struct tc_edp_link {
-	struct drm_dp_link	base;
+	u8			dpcd[DP_RECEIVER_CAP_SIZE];
+	unsigned int		rate;
+	u8			num_lanes;
 	u8			assr;
 	bool			scrambler_dis;
 	bool			spread;
@@ -437,9 +440,9 @@ static u32 tc_srcctrl(struct tc_data *tc)
 		reg |= DP0_SRCCTRL_SCRMBLDIS;	/* Scrambler Disabled */
 	if (tc->link.spread)
 		reg |= DP0_SRCCTRL_SSCG;	/* Spread Spectrum Enable */
-	if (tc->link.base.num_lanes == 2)
+	if (tc->link.num_lanes == 2)
 		reg |= DP0_SRCCTRL_LANES_2;	/* Two Main Channel Lanes */
-	if (tc->link.base.rate != 162000)
+	if (tc->link.rate != 162000)
 		reg |= DP0_SRCCTRL_BW27;	/* 2.7 Gbps link */
 	return reg;
 }
@@ -662,23 +665,35 @@ err:
 
 static int tc_get_display_props(struct tc_data *tc)
 {
+	u8 revision, num_lanes;
+	unsigned int rate;
 	int ret;
 	u8 reg;
 
 	/* Read DP Rx Link Capability */
-	ret = drm_dp_link_probe(&tc->aux, &tc->link.base);
+	ret = drm_dp_dpcd_read(&tc->aux, DP_DPCD_REV, tc->link.dpcd,
+			       DP_RECEIVER_CAP_SIZE);
 	if (ret < 0)
 		goto err_dpcd_read;
-	if (tc->link.base.rate != 162000 && tc->link.base.rate != 270000) {
+
+	revision = tc->link.dpcd[DP_DPCD_REV];
+	rate = drm_dp_max_link_rate(tc->link.dpcd);
+	num_lanes = drm_dp_max_lane_count(tc->link.dpcd);
+
+	if (rate != 162000 && rate != 270000) {
 		dev_dbg(tc->dev, "Falling to 2.7 Gbps rate\n");
-		tc->link.base.rate = 270000;
+		rate = 270000;
 	}
 
-	if (tc->link.base.num_lanes > 2) {
+	tc->link.rate = rate;
+
+	if (num_lanes > 2) {
 		dev_dbg(tc->dev, "Falling to 2 lanes\n");
-		tc->link.base.num_lanes = 2;
+		num_lanes = 2;
 	}
 
+	tc->link.num_lanes = num_lanes;
+
 	ret = drm_dp_dpcd_readb(&tc->aux, DP_MAX_DOWNSPREAD, &reg);
 	if (ret < 0)
 		goto err_dpcd_read;
@@ -696,11 +711,11 @@ static int tc_get_display_props(struct tc_data *tc)
 	tc->link.assr = reg & DP_ALTERNATE_SCRAMBLER_RESET_ENABLE;
 
 	dev_dbg(tc->dev, "DPCD rev: %d.%d, rate: %s, lanes: %d, framing: %s\n",
-		tc->link.base.revision >> 4, tc->link.base.revision & 0x0f,
-		(tc->link.base.rate == 162000) ? "1.62Gbps" : "2.7Gbps",
-		tc->link.base.num_lanes,
-		(tc->link.base.capabilities & DP_LINK_CAP_ENHANCED_FRAMING) ?
-		"enhanced" : "non-enhanced");
+		revision >> 4, revision & 0x0f,
+		(tc->link.rate == 162000) ? "1.62Gbps" : "2.7Gbps",
+		tc->link.num_lanes,
+		drm_dp_enhanced_frame_cap(tc->link.dpcd) ?
+		"enhanced" : "default");
 	dev_dbg(tc->dev, "Downspread: %s, scrambler: %s\n",
 		tc->link.spread ? "0.5%" : "0.0%",
 		tc->link.scrambler_dis ? "disabled" : "enabled");
@@ -728,6 +743,8 @@ static int tc_set_video_mode(struct tc_data *tc,
 	int lower_margin = mode->vsync_start - mode->vdisplay;
 	int vsync_len = mode->vsync_end - mode->vsync_start;
 	u32 dp0_syncval;
+	u32 bits_per_pixel = 24;
+	u32 in_bw, out_bw;
 
 	/*
 	 * Recommended maximum number of symbols transferred in a transfer unit:
@@ -735,7 +752,10 @@ static int tc_set_video_mode(struct tc_data *tc,
 	 *              (output active video bandwidth in bytes))
 	 * Must be less than tu_size.
 	 */
-	max_tu_symbol = TU_SIZE_RECOMMENDED - 1;
+
+	in_bw = mode->clock * bits_per_pixel / 8;
+	out_bw = tc->link.num_lanes * tc->link.rate;
+	max_tu_symbol = DIV_ROUND_UP(in_bw * TU_SIZE_RECOMMENDED, out_bw);
 
 	dev_dbg(tc->dev, "set mode %dx%d\n",
 		mode->hdisplay, mode->vdisplay);
@@ -896,7 +916,7 @@ static int tc_main_link_enable(struct tc_data *tc)
 	/* SSCG and BW27 on DP1 must be set to the same as on DP0 */
 	ret = regmap_write(tc->regmap, DP1_SRCCTRL,
 		 (tc->link.spread ? DP0_SRCCTRL_SSCG : 0) |
-		 ((tc->link.base.rate != 162000) ? DP0_SRCCTRL_BW27 : 0));
+		 ((tc->link.rate != 162000) ? DP0_SRCCTRL_BW27 : 0));
 	if (ret)
 		return ret;
 
@@ -906,7 +926,7 @@ static int tc_main_link_enable(struct tc_data *tc)
 
 	/* Setup Main Link */
 	dp_phy_ctrl = BGREN | PWR_SW_EN | PHY_A0_EN | PHY_M0_EN;
-	if (tc->link.base.num_lanes == 2)
+	if (tc->link.num_lanes == 2)
 		dp_phy_ctrl |= PHY_2LANE;
 
 	ret = regmap_write(tc->regmap, DP_PHY_CTRL, dp_phy_ctrl);
@@ -969,7 +989,13 @@ static int tc_main_link_enable(struct tc_data *tc)
 	}
 
 	/* Setup Link & DPRx Config for Training */
-	ret = drm_dp_link_configure(aux, &tc->link.base);
+	tmp[0] = drm_dp_link_rate_to_bw_code(tc->link.rate);
+	tmp[1] = tc->link.num_lanes;
+
+	if (drm_dp_enhanced_frame_cap(tc->link.dpcd))
+		tmp[1] |= DP_LANE_COUNT_ENHANCED_FRAME_EN;
+
+	ret = drm_dp_dpcd_write(aux, DP_LINK_BW_SET, tmp, 2);
 	if (ret < 0)
 		goto err_dpcd_write;
 
@@ -1013,9 +1039,8 @@ static int tc_main_link_enable(struct tc_data *tc)
 
 	/* Enable DP0 to start Link Training */
 	ret = regmap_write(tc->regmap, DP0CTL,
-			   ((tc->link.base.capabilities &
-			     DP_LINK_CAP_ENHANCED_FRAMING) ? EF_EN : 0) |
-			   DP_EN);
+			   (drm_dp_enhanced_frame_cap(tc->link.dpcd) ?
+				EF_EN : 0) | DP_EN);
 	if (ret)
 		return ret;
 
@@ -1094,7 +1119,7 @@ static int tc_main_link_enable(struct tc_data *tc)
 		ret = -ENODEV;
 	}
 
-	if (tc->link.base.num_lanes == 2) {
+	if (tc->link.num_lanes == 2) {
 		value = (tmp[0] >> 4) & DP_CHANNEL_EQ_BITS;
 
 		if (value != DP_CHANNEL_EQ_BITS) {
@@ -1165,7 +1190,7 @@ static int tc_stream_enable(struct tc_data *tc)
 		return ret;
 
 	value = VID_MN_GEN | DP_EN;
-	if (tc->link.base.capabilities & DP_LINK_CAP_ENHANCED_FRAMING)
+	if (drm_dp_enhanced_frame_cap(tc->link.dpcd))
 		value |= EF_EN;
 	ret = regmap_write(tc->regmap, DP0CTL, value);
 	if (ret)
@@ -1291,7 +1316,7 @@ static enum drm_mode_status tc_mode_valid(struct drm_bridge *bridge,
 		return MODE_CLOCK_HIGH;
 
 	req = mode->clock * bits_per_pixel / 8;
-	avail = tc->link.base.num_lanes * tc->link.base.rate;
+	avail = tc->link.num_lanes * tc->link.rate;
 
 	if (req > avail)
 		return MODE_BAD;
diff --git a/drivers/gpu/drm/bridge/ti-sn65dsi86.c b/drivers/gpu/drm/bridge/ti-sn65dsi86.c
index 0a580957c8cf..43abf01ebd4c 100644
--- a/drivers/gpu/drm/bridge/ti-sn65dsi86.c
+++ b/drivers/gpu/drm/bridge/ti-sn65dsi86.c
@@ -17,6 +17,7 @@
 
 #include <drm/drm_atomic.h>
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_dp_helper.h>
 #include <drm/drm_mipi_dsi.h>
 #include <drm/drm_of.h>
diff --git a/drivers/gpu/drm/bridge/ti-tfp410.c b/drivers/gpu/drm/bridge/ti-tfp410.c
index 61cc2354ef1b..6f6d6d1e60ae 100644
--- a/drivers/gpu/drm/bridge/ti-tfp410.c
+++ b/drivers/gpu/drm/bridge/ti-tfp410.c
@@ -14,6 +14,7 @@
 #include <linux/platform_device.h>
 
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_crtc.h>
 #include <drm/drm_print.h>
 #include <drm/drm_probe_helper.h>
@@ -284,8 +285,8 @@ static int tfp410_get_connector_properties(struct tfp410 *dvi)
 	else
 		dvi->connector_type = DRM_MODE_CONNECTOR_DVID;
 
-	dvi->hpd = fwnode_get_named_gpiod(&connector_node->fwnode,
-					"hpd-gpios", 0, GPIOD_IN, "hpd");
+	dvi->hpd = fwnode_gpiod_get_index(&connector_node->fwnode,
+					  "hpd", 0, GPIOD_IN, "hpd");
 	if (IS_ERR(dvi->hpd)) {
 		ret = PTR_ERR(dvi->hpd);
 		dvi->hpd = NULL;
diff --git a/drivers/gpu/drm/cirrus/cirrus.c b/drivers/gpu/drm/cirrus/cirrus.c
index 36a69aec8a4b..248c9f765c45 100644
--- a/drivers/gpu/drm/cirrus/cirrus.c
+++ b/drivers/gpu/drm/cirrus/cirrus.c
@@ -390,7 +390,7 @@ static int cirrus_conn_init(struct cirrus_device *cirrus)
 /* ------------------------------------------------------------------ */
 /* cirrus (simple) display pipe					      */
 
-static enum drm_mode_status cirrus_pipe_mode_valid(struct drm_crtc *crtc,
+static enum drm_mode_status cirrus_pipe_mode_valid(struct drm_simple_display_pipe *pipe,
 						   const struct drm_display_mode *mode)
 {
 	if (cirrus_check_size(mode->hdisplay, mode->vdisplay, NULL) < 0)
@@ -510,7 +510,7 @@ static void cirrus_mode_config_init(struct cirrus_device *cirrus)
 
 /* ------------------------------------------------------------------ */
 
-DEFINE_DRM_GEM_SHMEM_FOPS(cirrus_fops);
+DEFINE_DRM_GEM_FOPS(cirrus_fops);
 
 static struct drm_driver cirrus_driver = {
 	.driver_features = DRIVER_MODESET | DRIVER_GEM | DRIVER_ATOMIC,
@@ -532,7 +532,7 @@ static int cirrus_pci_probe(struct pci_dev *pdev,
 	struct cirrus_device *cirrus;
 	int ret;
 
-	ret = drm_fb_helper_remove_conflicting_pci_framebuffers(pdev, 0, "cirrusdrmfb");
+	ret = drm_fb_helper_remove_conflicting_pci_framebuffers(pdev, "cirrusdrmfb");
 	if (ret)
 		return ret;
 
diff --git a/drivers/gpu/drm/cirrus/cirrus_drv.h b/drivers/gpu/drm/cirrus/cirrus_drv.h
deleted file mode 100644
index 1f73916e528e..000000000000
--- a/drivers/gpu/drm/cirrus/cirrus_drv.h
+++ /dev/null
@@ -1,247 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0-only */
-/*
- * Copyright 2012 Red Hat
- *
- * Authors: Matthew Garrett
- *          Dave Airlie
- */
-#ifndef __CIRRUS_DRV_H__
-#define __CIRRUS_DRV_H__
-
-#include <video/vga.h>
-
-#include <drm/drm_encoder.h>
-#include <drm/drm_fb_helper.h>
-
-#include <drm/ttm/ttm_bo_api.h>
-#include <drm/ttm/ttm_bo_driver.h>
-#include <drm/ttm/ttm_placement.h>
-#include <drm/ttm/ttm_memory.h>
-#include <drm/ttm/ttm_module.h>
-
-#include <drm/drm_gem.h>
-
-#define DRIVER_AUTHOR		"Matthew Garrett"
-
-#define DRIVER_NAME		"cirrus"
-#define DRIVER_DESC		"qemu Cirrus emulation"
-#define DRIVER_DATE		"20110418"
-
-#define DRIVER_MAJOR		1
-#define DRIVER_MINOR		0
-#define DRIVER_PATCHLEVEL	0
-
-#define CIRRUSFB_CONN_LIMIT 1
-
-#define RREG8(reg) ioread8(((void __iomem *)cdev->rmmio) + (reg))
-#define WREG8(reg, v) iowrite8(v, ((void __iomem *)cdev->rmmio) + (reg))
-#define RREG32(reg) ioread32(((void __iomem *)cdev->rmmio) + (reg))
-#define WREG32(reg, v) iowrite32(v, ((void __iomem *)cdev->rmmio) + (reg))
-
-#define SEQ_INDEX 4
-#define SEQ_DATA 5
-
-#define WREG_SEQ(reg, v)					\
-	do {							\
-		WREG8(SEQ_INDEX, reg);				\
-		WREG8(SEQ_DATA, v);				\
-	} while (0)						\
-
-#define CRT_INDEX 0x14
-#define CRT_DATA 0x15
-
-#define WREG_CRT(reg, v)					\
-	do {							\
-		WREG8(CRT_INDEX, reg);				\
-		WREG8(CRT_DATA, v);				\
-	} while (0)						\
-
-#define GFX_INDEX 0xe
-#define GFX_DATA 0xf
-
-#define WREG_GFX(reg, v)					\
-	do {							\
-		WREG8(GFX_INDEX, reg);				\
-		WREG8(GFX_DATA, v);				\
-	} while (0)						\
-
-/*
- * Cirrus has a "hidden" DAC register that can be accessed by writing to
- * the pixel mask register to reset the state, then reading from the register
- * four times. The next write will then pass to the DAC
- */
-#define VGA_DAC_MASK 0x6
-
-#define WREG_HDR(v)						\
-	do {							\
-		RREG8(VGA_DAC_MASK);					\
-		RREG8(VGA_DAC_MASK);					\
-		RREG8(VGA_DAC_MASK);					\
-		RREG8(VGA_DAC_MASK);					\
-		WREG8(VGA_DAC_MASK, v);					\
-	} while (0)						\
-
-
-#define CIRRUS_MAX_FB_HEIGHT 4096
-#define CIRRUS_MAX_FB_WIDTH 4096
-
-#define CIRRUS_DPMS_CLEARED (-1)
-
-#define to_cirrus_crtc(x) container_of(x, struct cirrus_crtc, base)
-#define to_cirrus_encoder(x) container_of(x, struct cirrus_encoder, base)
-
-struct cirrus_crtc {
-	struct drm_crtc			base;
-	int				last_dpms;
-	bool				enabled;
-};
-
-struct cirrus_fbdev;
-struct cirrus_mode_info {
-	struct cirrus_crtc		*crtc;
-	/* pointer to fbdev info structure */
-	struct cirrus_fbdev		*gfbdev;
-};
-
-struct cirrus_encoder {
-	struct drm_encoder		base;
-	int				last_dpms;
-};
-
-struct cirrus_connector {
-	struct drm_connector		base;
-};
-
-struct cirrus_mc {
-	resource_size_t			vram_size;
-	resource_size_t			vram_base;
-};
-
-struct cirrus_device {
-	struct drm_device		*dev;
-	unsigned long			flags;
-
-	resource_size_t			rmmio_base;
-	resource_size_t			rmmio_size;
-	void __iomem			*rmmio;
-
-	struct cirrus_mc			mc;
-	struct cirrus_mode_info		mode_info;
-
-	int				num_crtc;
-	int fb_mtrr;
-
-	struct {
-		struct ttm_bo_device bdev;
-	} ttm;
-	bool mm_inited;
-};
-
-
-struct cirrus_fbdev {
-	struct drm_fb_helper helper; /* must be first */
-	struct drm_framebuffer *gfb;
-	void *sysram;
-	int size;
-	int x1, y1, x2, y2; /* dirty rect */
-	spinlock_t dirty_lock;
-};
-
-struct cirrus_bo {
-	struct ttm_buffer_object bo;
-	struct ttm_placement placement;
-	struct ttm_bo_kmap_obj kmap;
-	struct drm_gem_object gem;
-	struct ttm_place placements[3];
-	int pin_count;
-};
-#define gem_to_cirrus_bo(gobj) container_of((gobj), struct cirrus_bo, gem)
-
-static inline struct cirrus_bo *
-cirrus_bo(struct ttm_buffer_object *bo)
-{
-	return container_of(bo, struct cirrus_bo, bo);
-}
-
-
-#define to_cirrus_obj(x) container_of(x, struct cirrus_gem_object, base)
-
-				/* cirrus_main.c */
-int cirrus_device_init(struct cirrus_device *cdev,
-		      struct drm_device *ddev,
-		      struct pci_dev *pdev,
-		      uint32_t flags);
-void cirrus_device_fini(struct cirrus_device *cdev);
-void cirrus_gem_free_object(struct drm_gem_object *obj);
-int cirrus_dumb_mmap_offset(struct drm_file *file,
-			    struct drm_device *dev,
-			    uint32_t handle,
-			    uint64_t *offset);
-int cirrus_gem_create(struct drm_device *dev,
-		   u32 size, bool iskernel,
-		      struct drm_gem_object **obj);
-int cirrus_dumb_create(struct drm_file *file,
-		    struct drm_device *dev,
-		       struct drm_mode_create_dumb *args);
-
-int cirrus_framebuffer_init(struct drm_device *dev,
-			    struct drm_framebuffer *gfb,
-			    const struct drm_mode_fb_cmd2 *mode_cmd,
-			    struct drm_gem_object *obj);
-
-bool cirrus_check_framebuffer(struct cirrus_device *cdev, int width, int height,
-			      int bpp, int pitch);
-
-				/* cirrus_display.c */
-int cirrus_modeset_init(struct cirrus_device *cdev);
-void cirrus_modeset_fini(struct cirrus_device *cdev);
-
-				/* cirrus_fbdev.c */
-int cirrus_fbdev_init(struct cirrus_device *cdev);
-void cirrus_fbdev_fini(struct cirrus_device *cdev);
-
-
-
-				/* cirrus_irq.c */
-void cirrus_driver_irq_preinstall(struct drm_device *dev);
-int cirrus_driver_irq_postinstall(struct drm_device *dev);
-void cirrus_driver_irq_uninstall(struct drm_device *dev);
-irqreturn_t cirrus_driver_irq_handler(int irq, void *arg);
-
-				/* cirrus_kms.c */
-int cirrus_driver_load(struct drm_device *dev, unsigned long flags);
-void cirrus_driver_unload(struct drm_device *dev);
-extern struct drm_ioctl_desc cirrus_ioctls[];
-extern int cirrus_max_ioctl;
-
-int cirrus_mm_init(struct cirrus_device *cirrus);
-void cirrus_mm_fini(struct cirrus_device *cirrus);
-void cirrus_ttm_placement(struct cirrus_bo *bo, int domain);
-int cirrus_bo_create(struct drm_device *dev, int size, int align,
-		     uint32_t flags, struct cirrus_bo **pcirrusbo);
-int cirrus_mmap(struct file *filp, struct vm_area_struct *vma);
-
-static inline int cirrus_bo_reserve(struct cirrus_bo *bo, bool no_wait)
-{
-	int ret;
-
-	ret = ttm_bo_reserve(&bo->bo, true, no_wait, NULL);
-	if (ret) {
-		if (ret != -ERESTARTSYS && ret != -EBUSY)
-			DRM_ERROR("reserve failed %p\n", bo);
-		return ret;
-	}
-	return 0;
-}
-
-static inline void cirrus_bo_unreserve(struct cirrus_bo *bo)
-{
-	ttm_bo_unreserve(&bo->bo);
-}
-
-int cirrus_bo_push_sysram(struct cirrus_bo *bo);
-int cirrus_bo_pin(struct cirrus_bo *bo, u32 pl_flag, u64 *gpu_addr);
-
-extern int cirrus_bpp;
-
-#endif				/* __CIRRUS_DRV_H__ */
diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
index 419381abbdd1..14aeaf736321 100644
--- a/drivers/gpu/drm/drm_atomic.c
+++ b/drivers/gpu/drm/drm_atomic.c
@@ -430,10 +430,15 @@ static int drm_atomic_connector_check(struct drm_connector *connector,
 		return -EINVAL;
 	}
 
-	if (writeback_job->out_fence && !writeback_job->fb) {
-		DRM_DEBUG_ATOMIC("[CONNECTOR:%d:%s] requesting out-fence without framebuffer\n",
-				 connector->base.id, connector->name);
-		return -EINVAL;
+	if (!writeback_job->fb) {
+		if (writeback_job->out_fence) {
+			DRM_DEBUG_ATOMIC("[CONNECTOR:%d:%s] requesting out-fence without framebuffer\n",
+					 connector->base.id, connector->name);
+			return -EINVAL;
+		}
+
+		drm_writeback_cleanup_job(writeback_job);
+		state->writeback_job = NULL;
 	}
 
 	return 0;
diff --git a/drivers/gpu/drm/drm_atomic_helper.c b/drivers/gpu/drm/drm_atomic_helper.c
index 3ef2ac52ce94..b191d39c071d 100644
--- a/drivers/gpu/drm/drm_atomic_helper.c
+++ b/drivers/gpu/drm/drm_atomic_helper.c
@@ -31,6 +31,7 @@
 #include <drm/drm_atomic.h>
 #include <drm/drm_atomic_helper.h>
 #include <drm/drm_atomic_uapi.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_damage_helper.h>
 #include <drm/drm_device.h>
 #include <drm/drm_plane_helper.h>
@@ -97,17 +98,6 @@ drm_atomic_helper_plane_changed(struct drm_atomic_state *state,
 	}
 }
 
-/*
- * For connectors that support multiple encoders, either the
- * .atomic_best_encoder() or .best_encoder() operation must be implemented.
- */
-static struct drm_encoder *
-pick_single_encoder_for_connector(struct drm_connector *connector)
-{
-	WARN_ON(connector->encoder_ids[1]);
-	return drm_encoder_find(connector->dev, NULL, connector->encoder_ids[0]);
-}
-
 static int handle_conflicting_encoders(struct drm_atomic_state *state,
 				       bool disable_conflicting_encoders)
 {
@@ -135,7 +125,7 @@ static int handle_conflicting_encoders(struct drm_atomic_state *state,
 		else if (funcs->best_encoder)
 			new_encoder = funcs->best_encoder(connector);
 		else
-			new_encoder = pick_single_encoder_for_connector(connector);
+			new_encoder = drm_connector_get_single_encoder(connector);
 
 		if (new_encoder) {
 			if (encoder_mask & drm_encoder_mask(new_encoder)) {
@@ -359,7 +349,7 @@ update_connector_routing(struct drm_atomic_state *state,
 	else if (funcs->best_encoder)
 		new_encoder = funcs->best_encoder(connector);
 	else
-		new_encoder = pick_single_encoder_for_connector(connector);
+		new_encoder = drm_connector_get_single_encoder(connector);
 
 	if (!new_encoder) {
 		DRM_DEBUG_ATOMIC("No suitable encoder found for [CONNECTOR:%d:%s]\n",
@@ -482,7 +472,7 @@ mode_fixup(struct drm_atomic_state *state)
 			continue;
 
 		funcs = crtc->helper_private;
-		if (!funcs->mode_fixup)
+		if (!funcs || !funcs->mode_fixup)
 			continue;
 
 		ret = funcs->mode_fixup(crtc, &new_crtc_state->mode,
@@ -1581,8 +1571,11 @@ static void commit_tail(struct drm_atomic_state *old_state)
 {
 	struct drm_device *dev = old_state->dev;
 	const struct drm_mode_config_helper_funcs *funcs;
+	struct drm_crtc_state *new_crtc_state;
+	struct drm_crtc *crtc;
 	ktime_t start;
 	s64 commit_time_ms;
+	unsigned int i, new_self_refresh_mask = 0;
 
 	funcs = dev->mode_config.helper_private;
 
@@ -1602,6 +1595,15 @@ static void commit_tail(struct drm_atomic_state *old_state)
 
 	drm_atomic_helper_wait_for_dependencies(old_state);
 
+	/*
+	 * We cannot safely access new_crtc_state after
+	 * drm_atomic_helper_commit_hw_done() so figure out which crtc's have
+	 * self-refresh active beforehand:
+	 */
+	for_each_new_crtc_in_state(old_state, crtc, new_crtc_state, i)
+		if (new_crtc_state->self_refresh_active)
+			new_self_refresh_mask |= BIT(i);
+
 	if (funcs && funcs->atomic_commit_tail)
 		funcs->atomic_commit_tail(old_state);
 	else
@@ -1610,7 +1612,8 @@ static void commit_tail(struct drm_atomic_state *old_state)
 	commit_time_ms = ktime_ms_delta(ktime_get(), start);
 	if (commit_time_ms > 0)
 		drm_self_refresh_helper_update_avg_times(old_state,
-						 (unsigned long)commit_time_ms);
+						 (unsigned long)commit_time_ms,
+						 new_self_refresh_mask);
 
 	drm_atomic_helper_commit_cleanup_done(old_state);
 
diff --git a/drivers/gpu/drm/drm_atomic_uapi.c b/drivers/gpu/drm/drm_atomic_uapi.c
index 7a26bfb5329c..0d466d3b0809 100644
--- a/drivers/gpu/drm/drm_atomic_uapi.c
+++ b/drivers/gpu/drm/drm_atomic_uapi.c
@@ -1405,7 +1405,7 @@ retry:
 	} else if (arg->flags & DRM_MODE_ATOMIC_NONBLOCK) {
 		ret = drm_atomic_nonblocking_commit(state);
 	} else {
-		if (unlikely(drm_debug & DRM_UT_STATE))
+		if (drm_debug_enabled(DRM_UT_STATE))
 			drm_atomic_print_state(state);
 
 		ret = drm_atomic_commit(state);
diff --git a/drivers/gpu/drm/drm_blend.c b/drivers/gpu/drm/drm_blend.c
index 37ac168fcb60..121481f6aa71 100644
--- a/drivers/gpu/drm/drm_blend.c
+++ b/drivers/gpu/drm/drm_blend.c
@@ -130,7 +130,12 @@
  *	Z position is set up with drm_plane_create_zpos_immutable_property() and
  *	drm_plane_create_zpos_property(). It controls the visibility of overlapping
  *	planes. Without this property the primary plane is always below the cursor
- *	plane, and ordering between all other planes is undefined.
+ *	plane, and ordering between all other planes is undefined. The positive
+ *	Z axis points towards the user, i.e. planes with lower Z position values
+ *	are underneath planes with higher Z position values. Two planes with the
+ *	same Z position value have undefined ordering. Note that the Z position
+ *	value can also be immutable, to inform userspace about the hard-coded
+ *	stacking of planes, see drm_plane_create_zpos_immutable_property().
  *
  * pixel blend mode:
  *	Pixel blend mode is set up with drm_plane_create_blend_mode_property().
diff --git a/drivers/gpu/drm/drm_cache.c b/drivers/gpu/drm/drm_cache.c
index 3bd76e918b5d..03e01b000f7a 100644
--- a/drivers/gpu/drm/drm_cache.c
+++ b/drivers/gpu/drm/drm_cache.c
@@ -62,10 +62,10 @@ static void drm_cache_flush_clflush(struct page *pages[],
 {
 	unsigned long i;
 
-	mb();
+	mb(); /*Full memory barrier used before so that CLFLUSH is ordered*/
 	for (i = 0; i < num_pages; i++)
 		drm_clflush_page(*pages++);
-	mb();
+	mb(); /*Also used after CLFLUSH so that all cache is flushed*/
 }
 #endif
 
@@ -92,6 +92,7 @@ drm_clflush_pages(struct page *pages[], unsigned long num_pages)
 
 #elif defined(__powerpc__)
 	unsigned long i;
+
 	for (i = 0; i < num_pages; i++) {
 		struct page *page = pages[i];
 		void *page_virtual;
@@ -125,10 +126,10 @@ drm_clflush_sg(struct sg_table *st)
 	if (static_cpu_has(X86_FEATURE_CLFLUSH)) {
 		struct sg_page_iter sg_iter;
 
-		mb();
+		mb(); /*CLFLUSH is ordered only by using memory barriers*/
 		for_each_sg_page(st->sgl, &sg_iter, st->nents, 0)
 			drm_clflush_page(sg_page_iter_page(&sg_iter));
-		mb();
+		mb(); /*Make sure that all cache line entry is flushed*/
 
 		return;
 	}
@@ -157,12 +158,13 @@ drm_clflush_virt_range(void *addr, unsigned long length)
 	if (static_cpu_has(X86_FEATURE_CLFLUSH)) {
 		const int size = boot_cpu_data.x86_clflush_size;
 		void *end = addr + length;
+
 		addr = (void *)(((unsigned long)addr) & -size);
-		mb();
+		mb(); /*CLFLUSH is only ordered with a full memory barrier*/
 		for (; addr < end; addr += size)
 			clflushopt(addr);
 		clflushopt(end - 1); /* force serialisation */
-		mb();
+		mb(); /*Ensure that evry data cache line entry is flushed*/
 		return;
 	}
 
diff --git a/drivers/gpu/drm/drm_client_modeset.c b/drivers/gpu/drm/drm_client_modeset.c
index c8922b7cac09..895b73f23079 100644
--- a/drivers/gpu/drm/drm_client_modeset.c
+++ b/drivers/gpu/drm/drm_client_modeset.c
@@ -415,9 +415,8 @@ static bool connector_has_possible_crtc(struct drm_connector *connector,
 					struct drm_crtc *crtc)
 {
 	struct drm_encoder *encoder;
-	int i;
 
-	drm_connector_for_each_possible_encoder(connector, encoder, i) {
+	drm_connector_for_each_possible_encoder(connector, encoder) {
 		if (encoder->possible_crtcs & drm_crtc_mask(crtc))
 			return true;
 	}
diff --git a/drivers/gpu/drm/drm_connector.c b/drivers/gpu/drm/drm_connector.c
index 4c766624b20d..2166000ed057 100644
--- a/drivers/gpu/drm/drm_connector.c
+++ b/drivers/gpu/drm/drm_connector.c
@@ -365,8 +365,6 @@ EXPORT_SYMBOL(drm_connector_attach_edid_property);
 int drm_connector_attach_encoder(struct drm_connector *connector,
 				 struct drm_encoder *encoder)
 {
-	int i;
-
 	/*
 	 * In the past, drivers have attempted to model the static association
 	 * of connector to encoder in simple connector/encoder devices using a
@@ -381,18 +379,15 @@ int drm_connector_attach_encoder(struct drm_connector *connector,
 	if (WARN_ON(connector->encoder))
 		return -EINVAL;
 
-	for (i = 0; i < ARRAY_SIZE(connector->encoder_ids); i++) {
-		if (connector->encoder_ids[i] == 0) {
-			connector->encoder_ids[i] = encoder->base.id;
-			return 0;
-		}
-	}
-	return -ENOMEM;
+	connector->possible_encoders |= drm_encoder_mask(encoder);
+
+	return 0;
 }
 EXPORT_SYMBOL(drm_connector_attach_encoder);
 
 /**
- * drm_connector_has_possible_encoder - check if the connector and encoder are assosicated with each other
+ * drm_connector_has_possible_encoder - check if the connector and encoder are
+ * associated with each other
  * @connector: the connector
  * @encoder: the encoder
  *
@@ -402,15 +397,7 @@ EXPORT_SYMBOL(drm_connector_attach_encoder);
 bool drm_connector_has_possible_encoder(struct drm_connector *connector,
 					struct drm_encoder *encoder)
 {
-	struct drm_encoder *enc;
-	int i;
-
-	drm_connector_for_each_possible_encoder(connector, enc, i) {
-		if (enc == encoder)
-			return true;
-	}
-
-	return false;
+	return connector->possible_encoders & drm_encoder_mask(encoder);
 }
 EXPORT_SYMBOL(drm_connector_has_possible_encoder);
 
@@ -480,7 +467,10 @@ EXPORT_SYMBOL(drm_connector_cleanup);
  * drm_connector_register - register a connector
  * @connector: the connector to register
  *
- * Register userspace interfaces for a connector
+ * Register userspace interfaces for a connector. Only call this for connectors
+ * which can be hotplugged after drm_dev_register() has been called already,
+ * e.g. DP MST connectors. All other connectors will be registered automatically
+ * when calling drm_dev_register().
  *
  * Returns:
  * Zero on success, error code on failure.
@@ -526,7 +516,10 @@ EXPORT_SYMBOL(drm_connector_register);
  * drm_connector_unregister - unregister a connector
  * @connector: the connector to unregister
  *
- * Unregister userspace interfaces for a connector
+ * Unregister userspace interfaces for a connector. Only call this for
+ * connectors which have registered explicitly by calling drm_dev_register(),
+ * since connectors are unregistered automatically when drm_dev_unregister() is
+ * called.
  */
 void drm_connector_unregister(struct drm_connector *connector)
 {
@@ -719,7 +712,7 @@ void drm_connector_list_iter_end(struct drm_connector_list_iter *iter)
 		__drm_connector_put_safe(iter->conn);
 		spin_unlock_irqrestore(&config->connector_list_lock, flags);
 	}
-	lock_release(&connector_list_iter_dep_map, 0, _RET_IP_);
+	lock_release(&connector_list_iter_dep_map, _RET_IP_);
 }
 EXPORT_SYMBOL(drm_connector_list_iter_end);
 
@@ -882,6 +875,38 @@ static const struct drm_prop_enum_list hdmi_colorspaces[] = {
 	{ DRM_MODE_COLORIMETRY_DCI_P3_RGB_THEATER, "DCI-P3_RGB_Theater" },
 };
 
+/*
+ * As per DP 1.4a spec, 2.2.5.7.5 VSC SDP Payload for Pixel Encoding/Colorimetry
+ * Format Table 2-120
+ */
+static const struct drm_prop_enum_list dp_colorspaces[] = {
+	/* For Default case, driver will set the colorspace */
+	{ DRM_MODE_COLORIMETRY_DEFAULT, "Default" },
+	{ DRM_MODE_COLORIMETRY_RGB_WIDE_FIXED, "RGB_Wide_Gamut_Fixed_Point" },
+	/* Colorimetry based on scRGB (IEC 61966-2-2) */
+	{ DRM_MODE_COLORIMETRY_RGB_WIDE_FLOAT, "RGB_Wide_Gamut_Floating_Point" },
+	/* Colorimetry based on IEC 61966-2-5 */
+	{ DRM_MODE_COLORIMETRY_OPRGB, "opRGB" },
+	/* Colorimetry based on SMPTE RP 431-2 */
+	{ DRM_MODE_COLORIMETRY_DCI_P3_RGB_D65, "DCI-P3_RGB_D65" },
+	/* Colorimetry based on ITU-R BT.2020 */
+	{ DRM_MODE_COLORIMETRY_BT2020_RGB, "BT2020_RGB" },
+	{ DRM_MODE_COLORIMETRY_BT601_YCC, "BT601_YCC" },
+	{ DRM_MODE_COLORIMETRY_BT709_YCC, "BT709_YCC" },
+	/* Standard Definition Colorimetry based on IEC 61966-2-4 */
+	{ DRM_MODE_COLORIMETRY_XVYCC_601, "XVYCC_601" },
+	/* High Definition Colorimetry based on IEC 61966-2-4 */
+	{ DRM_MODE_COLORIMETRY_XVYCC_709, "XVYCC_709" },
+	/* Colorimetry based on IEC 61966-2-1/Amendment 1 */
+	{ DRM_MODE_COLORIMETRY_SYCC_601, "SYCC_601" },
+	/* Colorimetry based on IEC 61966-2-5 [33] */
+	{ DRM_MODE_COLORIMETRY_OPYCC_601, "opYCC_601" },
+	/* Colorimetry based on ITU-R BT.2020 */
+	{ DRM_MODE_COLORIMETRY_BT2020_CYCC, "BT2020_CYCC" },
+	/* Colorimetry based on ITU-R BT.2020 */
+	{ DRM_MODE_COLORIMETRY_BT2020_YCC, "BT2020_YCC" },
+};
+
 /**
  * DOC: standard connector properties
  *
@@ -1674,7 +1699,6 @@ EXPORT_SYMBOL(drm_mode_create_aspect_ratio_property);
  * DOC: standard connector properties
  *
  * Colorspace:
- *     drm_mode_create_colorspace_property - create colorspace property
  *     This property helps select a suitable colorspace based on the sink
  *     capability. Modern sink devices support wider gamut like BT2020.
  *     This helps switch to BT2020 mode if the BT2020 encoded video stream
@@ -1694,32 +1718,68 @@ EXPORT_SYMBOL(drm_mode_create_aspect_ratio_property);
  *      - This property is just to inform sink what colorspace
  *        source is trying to drive.
  *
+ * Because between HDMI and DP have different colorspaces,
+ * drm_mode_create_hdmi_colorspace_property() is used for HDMI connector and
+ * drm_mode_create_dp_colorspace_property() is used for DP connector.
+ */
+
+/**
+ * drm_mode_create_hdmi_colorspace_property - create hdmi colorspace property
+ * @connector: connector to create the Colorspace property on.
+ *
  * Called by a driver the first time it's needed, must be attached to desired
- * connectors.
+ * HDMI connectors.
+ *
+ * Returns:
+ * Zero on success, negative errono on failure.
  */
-int drm_mode_create_colorspace_property(struct drm_connector *connector)
+int drm_mode_create_hdmi_colorspace_property(struct drm_connector *connector)
 {
 	struct drm_device *dev = connector->dev;
-	struct drm_property *prop;
 
-	if (connector->connector_type == DRM_MODE_CONNECTOR_HDMIA ||
-	    connector->connector_type == DRM_MODE_CONNECTOR_HDMIB) {
-		prop = drm_property_create_enum(dev, DRM_MODE_PROP_ENUM,
-						"Colorspace",
-						hdmi_colorspaces,
-						ARRAY_SIZE(hdmi_colorspaces));
-		if (!prop)
-			return -ENOMEM;
-	} else {
-		DRM_DEBUG_KMS("Colorspace property not supported\n");
+	if (connector->colorspace_property)
 		return 0;
-	}
 
-	connector->colorspace_property = prop;
+	connector->colorspace_property =
+		drm_property_create_enum(dev, DRM_MODE_PROP_ENUM, "Colorspace",
+					 hdmi_colorspaces,
+					 ARRAY_SIZE(hdmi_colorspaces));
+
+	if (!connector->colorspace_property)
+		return -ENOMEM;
+
+	return 0;
+}
+EXPORT_SYMBOL(drm_mode_create_hdmi_colorspace_property);
+
+/**
+ * drm_mode_create_dp_colorspace_property - create dp colorspace property
+ * @connector: connector to create the Colorspace property on.
+ *
+ * Called by a driver the first time it's needed, must be attached to desired
+ * DP connectors.
+ *
+ * Returns:
+ * Zero on success, negative errono on failure.
+ */
+int drm_mode_create_dp_colorspace_property(struct drm_connector *connector)
+{
+	struct drm_device *dev = connector->dev;
+
+	if (connector->colorspace_property)
+		return 0;
+
+	connector->colorspace_property =
+		drm_property_create_enum(dev, DRM_MODE_PROP_ENUM, "Colorspace",
+					 dp_colorspaces,
+					 ARRAY_SIZE(dp_colorspaces));
+
+	if (!connector->colorspace_property)
+		return -ENOMEM;
 
 	return 0;
 }
-EXPORT_SYMBOL(drm_mode_create_colorspace_property);
+EXPORT_SYMBOL(drm_mode_create_dp_colorspace_property);
 
 /**
  * drm_mode_create_content_type_property - create content type property
@@ -2121,7 +2181,6 @@ int drm_mode_getconnector(struct drm_device *dev, void *data,
 	int encoders_count = 0;
 	int ret = 0;
 	int copied = 0;
-	int i;
 	struct drm_mode_modeinfo u_mode;
 	struct drm_mode_modeinfo __user *mode_ptr;
 	uint32_t __user *encoder_ptr;
@@ -2136,14 +2195,13 @@ int drm_mode_getconnector(struct drm_device *dev, void *data,
 	if (!connector)
 		return -ENOENT;
 
-	drm_connector_for_each_possible_encoder(connector, encoder, i)
-		encoders_count++;
+	encoders_count = hweight32(connector->possible_encoders);
 
 	if ((out_resp->count_encoders >= encoders_count) && encoders_count) {
 		copied = 0;
 		encoder_ptr = (uint32_t __user *)(unsigned long)(out_resp->encoders_ptr);
 
-		drm_connector_for_each_possible_encoder(connector, encoder, i) {
+		drm_connector_for_each_possible_encoder(connector, encoder) {
 			if (put_user(encoder->base.id, encoder_ptr + copied)) {
 				ret = -EFAULT;
 				goto out;
diff --git a/drivers/gpu/drm/drm_crtc_helper.c b/drivers/gpu/drm/drm_crtc_helper.c
index 80ddf13ad996..499b05aaccfc 100644
--- a/drivers/gpu/drm/drm_crtc_helper.c
+++ b/drivers/gpu/drm/drm_crtc_helper.c
@@ -36,6 +36,7 @@
 #include <drm/drm_atomic.h>
 #include <drm/drm_atomic_helper.h>
 #include <drm/drm_atomic_uapi.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_crtc.h>
 #include <drm/drm_crtc_helper.h>
 #include <drm/drm_drv.h>
@@ -459,6 +460,22 @@ drm_crtc_helper_disable(struct drm_crtc *crtc)
 	__drm_helper_disable_unused_functions(dev);
 }
 
+/*
+ * For connectors that support multiple encoders, either the
+ * .atomic_best_encoder() or .best_encoder() operation must be implemented.
+ */
+struct drm_encoder *
+drm_connector_get_single_encoder(struct drm_connector *connector)
+{
+	struct drm_encoder *encoder;
+
+	WARN_ON(hweight32(connector->possible_encoders) > 1);
+	drm_connector_for_each_possible_encoder(connector, encoder)
+		return encoder;
+
+	return NULL;
+}
+
 /**
  * drm_crtc_helper_set_config - set a new config from userspace
  * @set: mode set configuration
@@ -624,7 +641,11 @@ int drm_crtc_helper_set_config(struct drm_mode_set *set,
 		new_encoder = connector->encoder;
 		for (ro = 0; ro < set->num_connectors; ro++) {
 			if (set->connectors[ro] == connector) {
-				new_encoder = connector_funcs->best_encoder(connector);
+				if (connector_funcs->best_encoder)
+					new_encoder = connector_funcs->best_encoder(connector);
+				else
+					new_encoder = drm_connector_get_single_encoder(connector);
+
 				/* if we can't get an encoder for a connector
 				   we are setting now - then fail */
 				if (new_encoder == NULL)
diff --git a/drivers/gpu/drm/drm_crtc_helper_internal.h b/drivers/gpu/drm/drm_crtc_helper_internal.h
index b5ac1581e623..f0a66ef47e5a 100644
--- a/drivers/gpu/drm/drm_crtc_helper_internal.h
+++ b/drivers/gpu/drm/drm_crtc_helper_internal.h
@@ -75,3 +75,6 @@ enum drm_mode_status drm_encoder_mode_valid(struct drm_encoder *encoder,
 					    const struct drm_display_mode *mode);
 enum drm_mode_status drm_connector_mode_valid(struct drm_connector *connector,
 					      struct drm_display_mode *mode);
+
+struct drm_encoder *
+drm_connector_get_single_encoder(struct drm_connector *connector);
diff --git a/drivers/gpu/drm/drm_damage_helper.c b/drivers/gpu/drm/drm_damage_helper.c
index 8230dac01a89..3a4126dc2520 100644
--- a/drivers/gpu/drm/drm_damage_helper.c
+++ b/drivers/gpu/drm/drm_damage_helper.c
@@ -212,8 +212,14 @@ retry:
 	drm_for_each_plane(plane, fb->dev) {
 		struct drm_plane_state *plane_state;
 
-		if (plane->state->fb != fb)
+		ret = drm_modeset_lock(&plane->mutex, state->acquire_ctx);
+		if (ret)
+			goto out;
+
+		if (plane->state->fb != fb) {
+			drm_modeset_unlock(&plane->mutex);
 			continue;
+		}
 
 		plane_state = drm_atomic_get_plane_state(state, plane);
 		if (IS_ERR(plane_state)) {
diff --git a/drivers/gpu/drm/drm_debugfs_crc.c b/drivers/gpu/drm/drm_debugfs_crc.c
index be1b7ba92ffe..ca3c55c6b815 100644
--- a/drivers/gpu/drm/drm_debugfs_crc.c
+++ b/drivers/gpu/drm/drm_debugfs_crc.c
@@ -334,19 +334,17 @@ static ssize_t crtc_crc_read(struct file *filep, char __user *user_buf,
 	return LINE_LEN(crc->values_cnt);
 }
 
-static unsigned int crtc_crc_poll(struct file *file, poll_table *wait)
+static __poll_t crtc_crc_poll(struct file *file, poll_table *wait)
 {
 	struct drm_crtc *crtc = file->f_inode->i_private;
 	struct drm_crtc_crc *crc = &crtc->crc;
-	unsigned ret;
+	__poll_t ret = 0;
 
 	poll_wait(file, &crc->wq, wait);
 
 	spin_lock_irq(&crc->lock);
 	if (crc->source && crtc_crc_data_count(crc))
-		ret = POLLIN | POLLRDNORM;
-	else
-		ret = 0;
+		ret |= EPOLLIN | EPOLLRDNORM;
 	spin_unlock_irq(&crc->lock);
 
 	return ret;
diff --git a/drivers/gpu/drm/drm_dp_cec.c b/drivers/gpu/drm/drm_dp_cec.c
index b15cee85b702..3ab2609f9ec7 100644
--- a/drivers/gpu/drm/drm_dp_cec.c
+++ b/drivers/gpu/drm/drm_dp_cec.c
@@ -8,9 +8,13 @@
 #include <linux/kernel.h>
 #include <linux/module.h>
 #include <linux/slab.h>
-#include <drm/drm_dp_helper.h>
+
 #include <media/cec.h>
 
+#include <drm/drm_connector.h>
+#include <drm/drm_device.h>
+#include <drm/drm_dp_helper.h>
+
 /*
  * Unfortunately it turns out that we have a chicken-and-egg situation
  * here. Quite a few active (mini-)DP-to-HDMI or USB-C-to-HDMI adapters
@@ -295,7 +299,10 @@ static void drm_dp_cec_unregister_work(struct work_struct *work)
  */
 void drm_dp_cec_set_edid(struct drm_dp_aux *aux, const struct edid *edid)
 {
-	u32 cec_caps = CEC_CAP_DEFAULTS | CEC_CAP_NEEDS_HPD;
+	struct drm_connector *connector = aux->cec.connector;
+	u32 cec_caps = CEC_CAP_DEFAULTS | CEC_CAP_NEEDS_HPD |
+		       CEC_CAP_CONNECTOR_INFO;
+	struct cec_connector_info conn_info;
 	unsigned int num_las = 1;
 	u8 cap;
 
@@ -344,13 +351,17 @@ void drm_dp_cec_set_edid(struct drm_dp_aux *aux, const struct edid *edid)
 
 	/* Create a new adapter */
 	aux->cec.adap = cec_allocate_adapter(&drm_dp_cec_adap_ops,
-					     aux, aux->cec.name, cec_caps,
+					     aux, connector->name, cec_caps,
 					     num_las);
 	if (IS_ERR(aux->cec.adap)) {
 		aux->cec.adap = NULL;
 		goto unlock;
 	}
-	if (cec_register_adapter(aux->cec.adap, aux->cec.parent)) {
+
+	cec_fill_conn_info_from_drm(&conn_info, connector);
+	cec_s_conn_info(aux->cec.adap, &conn_info);
+
+	if (cec_register_adapter(aux->cec.adap, connector->dev->dev)) {
 		cec_delete_adapter(aux->cec.adap);
 		aux->cec.adap = NULL;
 	} else {
@@ -406,22 +417,20 @@ EXPORT_SYMBOL(drm_dp_cec_unset_edid);
 /**
  * drm_dp_cec_register_connector() - register a new connector
  * @aux: DisplayPort AUX channel
- * @name: name of the CEC device
- * @parent: parent device
+ * @connector: drm connector
  *
  * A new connector was registered with associated CEC adapter name and
  * CEC adapter parent device. After registering the name and parent
  * drm_dp_cec_set_edid() is called to check if the connector supports
  * CEC and to register a CEC adapter if that is the case.
  */
-void drm_dp_cec_register_connector(struct drm_dp_aux *aux, const char *name,
-				   struct device *parent)
+void drm_dp_cec_register_connector(struct drm_dp_aux *aux,
+				   struct drm_connector *connector)
 {
 	WARN_ON(aux->cec.adap);
 	if (WARN_ON(!aux->transfer))
 		return;
-	aux->cec.name = name;
-	aux->cec.parent = parent;
+	aux->cec.connector = connector;
 	INIT_DELAYED_WORK(&aux->cec.unregister_work,
 			  drm_dp_cec_unregister_work);
 }
diff --git a/drivers/gpu/drm/drm_dp_helper.c b/drivers/gpu/drm/drm_dp_helper.c
index ffc68d305afe..2c7870aef469 100644
--- a/drivers/gpu/drm/drm_dp_helper.c
+++ b/drivers/gpu/drm/drm_dp_helper.c
@@ -120,33 +120,49 @@ u8 drm_dp_get_adjust_request_pre_emphasis(const u8 link_status[DP_LINK_STATUS_SI
 }
 EXPORT_SYMBOL(drm_dp_get_adjust_request_pre_emphasis);
 
-void drm_dp_link_train_clock_recovery_delay(const u8 dpcd[DP_RECEIVER_CAP_SIZE]) {
-	int rd_interval = dpcd[DP_TRAINING_AUX_RD_INTERVAL] &
-			  DP_TRAINING_AUX_RD_MASK;
+u8 drm_dp_get_adjust_request_post_cursor(const u8 link_status[DP_LINK_STATUS_SIZE],
+					 unsigned int lane)
+{
+	unsigned int offset = DP_ADJUST_REQUEST_POST_CURSOR2;
+	u8 value = dp_link_status(link_status, offset);
+
+	return (value >> (lane << 1)) & 0x3;
+}
+EXPORT_SYMBOL(drm_dp_get_adjust_request_post_cursor);
+
+void drm_dp_link_train_clock_recovery_delay(const u8 dpcd[DP_RECEIVER_CAP_SIZE])
+{
+	unsigned long rd_interval = dpcd[DP_TRAINING_AUX_RD_INTERVAL] &
+					 DP_TRAINING_AUX_RD_MASK;
 
 	if (rd_interval > 4)
-		DRM_DEBUG_KMS("AUX interval %d, out of range (max 4)\n",
+		DRM_DEBUG_KMS("AUX interval %lu, out of range (max 4)\n",
 			      rd_interval);
 
 	if (rd_interval == 0 || dpcd[DP_DPCD_REV] >= DP_DPCD_REV_14)
-		udelay(100);
+		rd_interval = 100;
 	else
-		mdelay(rd_interval * 4);
+		rd_interval *= 4 * USEC_PER_MSEC;
+
+	usleep_range(rd_interval, rd_interval * 2);
 }
 EXPORT_SYMBOL(drm_dp_link_train_clock_recovery_delay);
 
-void drm_dp_link_train_channel_eq_delay(const u8 dpcd[DP_RECEIVER_CAP_SIZE]) {
-	int rd_interval = dpcd[DP_TRAINING_AUX_RD_INTERVAL] &
-			  DP_TRAINING_AUX_RD_MASK;
+void drm_dp_link_train_channel_eq_delay(const u8 dpcd[DP_RECEIVER_CAP_SIZE])
+{
+	unsigned long rd_interval = dpcd[DP_TRAINING_AUX_RD_INTERVAL] &
+					 DP_TRAINING_AUX_RD_MASK;
 
 	if (rd_interval > 4)
-		DRM_DEBUG_KMS("AUX interval %d, out of range (max 4)\n",
+		DRM_DEBUG_KMS("AUX interval %lu, out of range (max 4)\n",
 			      rd_interval);
 
 	if (rd_interval == 0)
-		udelay(400);
+		rd_interval = 400;
 	else
-		mdelay(rd_interval * 4);
+		rd_interval *= 4 * USEC_PER_MSEC;
+
+	usleep_range(rd_interval, rd_interval * 2);
 }
 EXPORT_SYMBOL(drm_dp_link_train_channel_eq_delay);
 
@@ -220,7 +236,6 @@ static int drm_dp_dpcd_access(struct drm_dp_aux *aux, u8 request,
 		}
 
 		ret = aux->transfer(aux, &msg);
-
 		if (ret >= 0) {
 			native_reply = msg.reply & DP_AUX_NATIVE_REPLY_MASK;
 			if (native_reply == DP_AUX_NATIVE_REPLY_ACK) {
@@ -337,134 +352,6 @@ int drm_dp_dpcd_read_link_status(struct drm_dp_aux *aux,
 EXPORT_SYMBOL(drm_dp_dpcd_read_link_status);
 
 /**
- * drm_dp_link_probe() - probe a DisplayPort link for capabilities
- * @aux: DisplayPort AUX channel
- * @link: pointer to structure in which to return link capabilities
- *
- * The structure filled in by this function can usually be passed directly
- * into drm_dp_link_power_up() and drm_dp_link_configure() to power up and
- * configure the link based on the link's capabilities.
- *
- * Returns 0 on success or a negative error code on failure.
- */
-int drm_dp_link_probe(struct drm_dp_aux *aux, struct drm_dp_link *link)
-{
-	u8 values[3];
-	int err;
-
-	memset(link, 0, sizeof(*link));
-
-	err = drm_dp_dpcd_read(aux, DP_DPCD_REV, values, sizeof(values));
-	if (err < 0)
-		return err;
-
-	link->revision = values[0];
-	link->rate = drm_dp_bw_code_to_link_rate(values[1]);
-	link->num_lanes = values[2] & DP_MAX_LANE_COUNT_MASK;
-
-	if (values[2] & DP_ENHANCED_FRAME_CAP)
-		link->capabilities |= DP_LINK_CAP_ENHANCED_FRAMING;
-
-	return 0;
-}
-EXPORT_SYMBOL(drm_dp_link_probe);
-
-/**
- * drm_dp_link_power_up() - power up a DisplayPort link
- * @aux: DisplayPort AUX channel
- * @link: pointer to a structure containing the link configuration
- *
- * Returns 0 on success or a negative error code on failure.
- */
-int drm_dp_link_power_up(struct drm_dp_aux *aux, struct drm_dp_link *link)
-{
-	u8 value;
-	int err;
-
-	/* DP_SET_POWER register is only available on DPCD v1.1 and later */
-	if (link->revision < 0x11)
-		return 0;
-
-	err = drm_dp_dpcd_readb(aux, DP_SET_POWER, &value);
-	if (err < 0)
-		return err;
-
-	value &= ~DP_SET_POWER_MASK;
-	value |= DP_SET_POWER_D0;
-
-	err = drm_dp_dpcd_writeb(aux, DP_SET_POWER, value);
-	if (err < 0)
-		return err;
-
-	/*
-	 * According to the DP 1.1 specification, a "Sink Device must exit the
-	 * power saving state within 1 ms" (Section 2.5.3.1, Table 5-52, "Sink
-	 * Control Field" (register 0x600).
-	 */
-	usleep_range(1000, 2000);
-
-	return 0;
-}
-EXPORT_SYMBOL(drm_dp_link_power_up);
-
-/**
- * drm_dp_link_power_down() - power down a DisplayPort link
- * @aux: DisplayPort AUX channel
- * @link: pointer to a structure containing the link configuration
- *
- * Returns 0 on success or a negative error code on failure.
- */
-int drm_dp_link_power_down(struct drm_dp_aux *aux, struct drm_dp_link *link)
-{
-	u8 value;
-	int err;
-
-	/* DP_SET_POWER register is only available on DPCD v1.1 and later */
-	if (link->revision < 0x11)
-		return 0;
-
-	err = drm_dp_dpcd_readb(aux, DP_SET_POWER, &value);
-	if (err < 0)
-		return err;
-
-	value &= ~DP_SET_POWER_MASK;
-	value |= DP_SET_POWER_D3;
-
-	err = drm_dp_dpcd_writeb(aux, DP_SET_POWER, value);
-	if (err < 0)
-		return err;
-
-	return 0;
-}
-EXPORT_SYMBOL(drm_dp_link_power_down);
-
-/**
- * drm_dp_link_configure() - configure a DisplayPort link
- * @aux: DisplayPort AUX channel
- * @link: pointer to a structure containing the link configuration
- *
- * Returns 0 on success or a negative error code on failure.
- */
-int drm_dp_link_configure(struct drm_dp_aux *aux, struct drm_dp_link *link)
-{
-	u8 values[2];
-	int err;
-
-	values[0] = drm_dp_link_rate_to_bw_code(link->rate);
-	values[1] = link->num_lanes;
-
-	if (link->capabilities & DP_LINK_CAP_ENHANCED_FRAMING)
-		values[1] |= DP_LANE_COUNT_ENHANCED_FRAME_EN;
-
-	err = drm_dp_dpcd_write(aux, DP_LINK_BW_SET, values, sizeof(values));
-	if (err < 0)
-		return err;
-
-	return 0;
-}
-EXPORT_SYMBOL(drm_dp_link_configure);
-
-/**
  * drm_dp_downstream_max_clock() - extract branch device max
  *                                 pixel rate for legacy VGA
  *                                 converter or max TMDS clock
@@ -1109,6 +996,14 @@ EXPORT_SYMBOL(drm_dp_aux_init);
  * @aux: DisplayPort AUX channel
  *
  * Automatically calls drm_dp_aux_init() if this hasn't been done yet.
+ * This should only be called when the underlying &struct drm_connector is
+ * initialiazed already. Therefore the best place to call this is from
+ * &drm_connector_funcs.late_register. Not that drivers which don't follow this
+ * will Oops when CONFIG_DRM_DP_AUX_CHARDEV is enabled.
+ *
+ * Drivers which need to use the aux channel before that point (e.g. at driver
+ * load time, before drm_dev_register() has been called) need to call
+ * drm_dp_aux_init().
  *
  * Returns 0 on success or a negative error code on failure.
  */
diff --git a/drivers/gpu/drm/drm_dp_mst_topology.c b/drivers/gpu/drm/drm_dp_mst_topology.c
index 82add736e17d..ae5809a1f19a 100644
--- a/drivers/gpu/drm/drm_dp_mst_topology.c
+++ b/drivers/gpu/drm/drm_dp_mst_topology.c
@@ -28,15 +28,22 @@
 #include <linux/sched.h>
 #include <linux/seq_file.h>
 
+#if IS_ENABLED(CONFIG_DRM_DEBUG_DP_MST_TOPOLOGY_REFS)
+#include <linux/stacktrace.h>
+#include <linux/sort.h>
+#include <linux/timekeeping.h>
+#include <linux/math64.h>
+#endif
+
 #include <drm/drm_atomic.h>
 #include <drm/drm_atomic_helper.h>
 #include <drm/drm_dp_mst_helper.h>
 #include <drm/drm_drv.h>
-#include <drm/drm_fixed.h>
 #include <drm/drm_print.h>
 #include <drm/drm_probe_helper.h>
 
 #include "drm_crtc_helper_internal.h"
+#include "drm_dp_mst_topology_internal.h"
 
 /**
  * DOC: dp mst helper
@@ -45,9 +52,14 @@
  * protocol. The helpers contain a topology manager and bandwidth manager.
  * The helpers encapsulate the sending and received of sideband msgs.
  */
+struct drm_dp_pending_up_req {
+	struct drm_dp_sideband_msg_hdr hdr;
+	struct drm_dp_sideband_msg_req_body msg;
+	struct list_head next;
+};
+
 static bool dump_dp_payload_table(struct drm_dp_mst_topology_mgr *mgr,
 				  char *buf);
-static int test_calc_pbn_mode(void);
 
 static void drm_dp_mst_topology_put_port(struct drm_dp_mst_port *port);
 
@@ -62,8 +74,8 @@ static int drm_dp_send_dpcd_write(struct drm_dp_mst_topology_mgr *mgr,
 				  struct drm_dp_mst_port *port,
 				  int offset, int size, u8 *bytes);
 
-static void drm_dp_send_link_address(struct drm_dp_mst_topology_mgr *mgr,
-				     struct drm_dp_mst_branch *mstb);
+static int drm_dp_send_link_address(struct drm_dp_mst_topology_mgr *mgr,
+				    struct drm_dp_mst_branch *mstb);
 static int drm_dp_send_enum_path_resources(struct drm_dp_mst_topology_mgr *mgr,
 					   struct drm_dp_mst_branch *mstb,
 					   struct drm_dp_mst_port *port);
@@ -74,6 +86,8 @@ static int drm_dp_mst_register_i2c_bus(struct drm_dp_aux *aux);
 static void drm_dp_mst_unregister_i2c_bus(struct drm_dp_aux *aux);
 static void drm_dp_mst_kick_tx(struct drm_dp_mst_topology_mgr *mgr);
 
+#define DBG_PREFIX "[dp_mst]"
+
 #define DP_STR(x) [DP_ ## x] = #x
 
 static const char *drm_dp_mst_req_type_str(u8 req_type)
@@ -130,6 +144,43 @@ static const char *drm_dp_mst_nak_reason_str(u8 nak_reason)
 }
 
 #undef DP_STR
+#define DP_STR(x) [DRM_DP_SIDEBAND_TX_ ## x] = #x
+
+static const char *drm_dp_mst_sideband_tx_state_str(int state)
+{
+	static const char * const sideband_reason_str[] = {
+		DP_STR(QUEUED),
+		DP_STR(START_SEND),
+		DP_STR(SENT),
+		DP_STR(RX),
+		DP_STR(TIMEOUT),
+	};
+
+	if (state >= ARRAY_SIZE(sideband_reason_str) ||
+	    !sideband_reason_str[state])
+		return "unknown";
+
+	return sideband_reason_str[state];
+}
+
+static int
+drm_dp_mst_rad_to_str(const u8 rad[8], u8 lct, char *out, size_t len)
+{
+	int i;
+	u8 unpacked_rad[16];
+
+	for (i = 0; i < lct; i++) {
+		if (i % 2)
+			unpacked_rad[i] = rad[i / 2] >> 4;
+		else
+			unpacked_rad[i] = rad[i / 2] & BIT_MASK(4);
+	}
+
+	/* TODO: Eventually add something to printk so we can format the rad
+	 * like this: 1.2.3
+	 */
+	return snprintf(out, len, "%*phC", lct, unpacked_rad);
+}
 
 /* sideband msg handling */
 static u8 drm_dp_msg_header_crc4(const uint8_t *data, size_t num_nibbles)
@@ -262,8 +313,9 @@ static bool drm_dp_decode_sideband_msg_hdr(struct drm_dp_sideband_msg_hdr *hdr,
 	return true;
 }
 
-static void drm_dp_encode_sideband_req(struct drm_dp_sideband_msg_req_body *req,
-				       struct drm_dp_sideband_msg_tx *raw)
+void
+drm_dp_encode_sideband_req(const struct drm_dp_sideband_msg_req_body *req,
+			   struct drm_dp_sideband_msg_tx *raw)
 {
 	int idx = 0;
 	int i;
@@ -272,6 +324,8 @@ static void drm_dp_encode_sideband_req(struct drm_dp_sideband_msg_req_body *req,
 
 	switch (req->req_type) {
 	case DP_ENUM_PATH_RESOURCES:
+	case DP_POWER_DOWN_PHY:
+	case DP_POWER_UP_PHY:
 		buf[idx] = (req->u.port_num.port_number & 0xf) << 4;
 		idx++;
 		break;
@@ -359,14 +413,253 @@ static void drm_dp_encode_sideband_req(struct drm_dp_sideband_msg_req_body *req,
 		memcpy(&buf[idx], req->u.i2c_write.bytes, req->u.i2c_write.num_bytes);
 		idx += req->u.i2c_write.num_bytes;
 		break;
+	}
+	raw->cur_len = idx;
+}
+EXPORT_SYMBOL_FOR_TESTS_ONLY(drm_dp_encode_sideband_req);
+
+/* Decode a sideband request we've encoded, mainly used for debugging */
+int
+drm_dp_decode_sideband_req(const struct drm_dp_sideband_msg_tx *raw,
+			   struct drm_dp_sideband_msg_req_body *req)
+{
+	const u8 *buf = raw->msg;
+	int i, idx = 0;
 
+	req->req_type = buf[idx++] & 0x7f;
+	switch (req->req_type) {
+	case DP_ENUM_PATH_RESOURCES:
 	case DP_POWER_DOWN_PHY:
 	case DP_POWER_UP_PHY:
-		buf[idx] = (req->u.port_num.port_number & 0xf) << 4;
-		idx++;
+		req->u.port_num.port_number = (buf[idx] >> 4) & 0xf;
+		break;
+	case DP_ALLOCATE_PAYLOAD:
+		{
+			struct drm_dp_allocate_payload *a =
+				&req->u.allocate_payload;
+
+			a->number_sdp_streams = buf[idx] & 0xf;
+			a->port_number = (buf[idx] >> 4) & 0xf;
+
+			WARN_ON(buf[++idx] & 0x80);
+			a->vcpi = buf[idx] & 0x7f;
+
+			a->pbn = buf[++idx] << 8;
+			a->pbn |= buf[++idx];
+
+			idx++;
+			for (i = 0; i < a->number_sdp_streams; i++) {
+				a->sdp_stream_sink[i] =
+					(buf[idx + (i / 2)] >> ((i % 2) ? 0 : 4)) & 0xf;
+			}
+		}
+		break;
+	case DP_QUERY_PAYLOAD:
+		req->u.query_payload.port_number = (buf[idx] >> 4) & 0xf;
+		WARN_ON(buf[++idx] & 0x80);
+		req->u.query_payload.vcpi = buf[idx] & 0x7f;
+		break;
+	case DP_REMOTE_DPCD_READ:
+		{
+			struct drm_dp_remote_dpcd_read *r = &req->u.dpcd_read;
+
+			r->port_number = (buf[idx] >> 4) & 0xf;
+
+			r->dpcd_address = (buf[idx] << 16) & 0xf0000;
+			r->dpcd_address |= (buf[++idx] << 8) & 0xff00;
+			r->dpcd_address |= buf[++idx] & 0xff;
+
+			r->num_bytes = buf[++idx];
+		}
+		break;
+	case DP_REMOTE_DPCD_WRITE:
+		{
+			struct drm_dp_remote_dpcd_write *w =
+				&req->u.dpcd_write;
+
+			w->port_number = (buf[idx] >> 4) & 0xf;
+
+			w->dpcd_address = (buf[idx] << 16) & 0xf0000;
+			w->dpcd_address |= (buf[++idx] << 8) & 0xff00;
+			w->dpcd_address |= buf[++idx] & 0xff;
+
+			w->num_bytes = buf[++idx];
+
+			w->bytes = kmemdup(&buf[++idx], w->num_bytes,
+					   GFP_KERNEL);
+			if (!w->bytes)
+				return -ENOMEM;
+		}
+		break;
+	case DP_REMOTE_I2C_READ:
+		{
+			struct drm_dp_remote_i2c_read *r = &req->u.i2c_read;
+			struct drm_dp_remote_i2c_read_tx *tx;
+			bool failed = false;
+
+			r->num_transactions = buf[idx] & 0x3;
+			r->port_number = (buf[idx] >> 4) & 0xf;
+			for (i = 0; i < r->num_transactions; i++) {
+				tx = &r->transactions[i];
+
+				tx->i2c_dev_id = buf[++idx] & 0x7f;
+				tx->num_bytes = buf[++idx];
+				tx->bytes = kmemdup(&buf[++idx],
+						    tx->num_bytes,
+						    GFP_KERNEL);
+				if (!tx->bytes) {
+					failed = true;
+					break;
+				}
+				idx += tx->num_bytes;
+				tx->no_stop_bit = (buf[idx] >> 5) & 0x1;
+				tx->i2c_transaction_delay = buf[idx] & 0xf;
+			}
+
+			if (failed) {
+				for (i = 0; i < r->num_transactions; i++)
+					kfree(tx->bytes);
+				return -ENOMEM;
+			}
+
+			r->read_i2c_device_id = buf[++idx] & 0x7f;
+			r->num_bytes_read = buf[++idx];
+		}
+		break;
+	case DP_REMOTE_I2C_WRITE:
+		{
+			struct drm_dp_remote_i2c_write *w = &req->u.i2c_write;
+
+			w->port_number = (buf[idx] >> 4) & 0xf;
+			w->write_i2c_device_id = buf[++idx] & 0x7f;
+			w->num_bytes = buf[++idx];
+			w->bytes = kmemdup(&buf[++idx], w->num_bytes,
+					   GFP_KERNEL);
+			if (!w->bytes)
+				return -ENOMEM;
+		}
+		break;
+	}
+
+	return 0;
+}
+EXPORT_SYMBOL_FOR_TESTS_ONLY(drm_dp_decode_sideband_req);
+
+void
+drm_dp_dump_sideband_msg_req_body(const struct drm_dp_sideband_msg_req_body *req,
+				  int indent, struct drm_printer *printer)
+{
+	int i;
+
+#define P(f, ...) drm_printf_indent(printer, indent, f, ##__VA_ARGS__)
+	if (req->req_type == DP_LINK_ADDRESS) {
+		/* No contents to print */
+		P("type=%s\n", drm_dp_mst_req_type_str(req->req_type));
+		return;
+	}
+
+	P("type=%s contents:\n", drm_dp_mst_req_type_str(req->req_type));
+	indent++;
+
+	switch (req->req_type) {
+	case DP_ENUM_PATH_RESOURCES:
+	case DP_POWER_DOWN_PHY:
+	case DP_POWER_UP_PHY:
+		P("port=%d\n", req->u.port_num.port_number);
+		break;
+	case DP_ALLOCATE_PAYLOAD:
+		P("port=%d vcpi=%d pbn=%d sdp_streams=%d %*ph\n",
+		  req->u.allocate_payload.port_number,
+		  req->u.allocate_payload.vcpi, req->u.allocate_payload.pbn,
+		  req->u.allocate_payload.number_sdp_streams,
+		  req->u.allocate_payload.number_sdp_streams,
+		  req->u.allocate_payload.sdp_stream_sink);
+		break;
+	case DP_QUERY_PAYLOAD:
+		P("port=%d vcpi=%d\n",
+		  req->u.query_payload.port_number,
+		  req->u.query_payload.vcpi);
+		break;
+	case DP_REMOTE_DPCD_READ:
+		P("port=%d dpcd_addr=%05x len=%d\n",
+		  req->u.dpcd_read.port_number, req->u.dpcd_read.dpcd_address,
+		  req->u.dpcd_read.num_bytes);
+		break;
+	case DP_REMOTE_DPCD_WRITE:
+		P("port=%d addr=%05x len=%d: %*ph\n",
+		  req->u.dpcd_write.port_number,
+		  req->u.dpcd_write.dpcd_address,
+		  req->u.dpcd_write.num_bytes, req->u.dpcd_write.num_bytes,
+		  req->u.dpcd_write.bytes);
+		break;
+	case DP_REMOTE_I2C_READ:
+		P("port=%d num_tx=%d id=%d size=%d:\n",
+		  req->u.i2c_read.port_number,
+		  req->u.i2c_read.num_transactions,
+		  req->u.i2c_read.read_i2c_device_id,
+		  req->u.i2c_read.num_bytes_read);
+
+		indent++;
+		for (i = 0; i < req->u.i2c_read.num_transactions; i++) {
+			const struct drm_dp_remote_i2c_read_tx *rtx =
+				&req->u.i2c_read.transactions[i];
+
+			P("%d: id=%03d size=%03d no_stop_bit=%d tx_delay=%03d: %*ph\n",
+			  i, rtx->i2c_dev_id, rtx->num_bytes,
+			  rtx->no_stop_bit, rtx->i2c_transaction_delay,
+			  rtx->num_bytes, rtx->bytes);
+		}
+		break;
+	case DP_REMOTE_I2C_WRITE:
+		P("port=%d id=%d size=%d: %*ph\n",
+		  req->u.i2c_write.port_number,
+		  req->u.i2c_write.write_i2c_device_id,
+		  req->u.i2c_write.num_bytes, req->u.i2c_write.num_bytes,
+		  req->u.i2c_write.bytes);
+		break;
+	default:
+		P("???\n");
+		break;
+	}
+#undef P
+}
+EXPORT_SYMBOL_FOR_TESTS_ONLY(drm_dp_dump_sideband_msg_req_body);
+
+static inline void
+drm_dp_mst_dump_sideband_msg_tx(struct drm_printer *p,
+				const struct drm_dp_sideband_msg_tx *txmsg)
+{
+	struct drm_dp_sideband_msg_req_body req;
+	char buf[64];
+	int ret;
+	int i;
+
+	drm_dp_mst_rad_to_str(txmsg->dst->rad, txmsg->dst->lct, buf,
+			      sizeof(buf));
+	drm_printf(p, "txmsg cur_offset=%x cur_len=%x seqno=%x state=%s path_msg=%d dst=%s\n",
+		   txmsg->cur_offset, txmsg->cur_len, txmsg->seqno,
+		   drm_dp_mst_sideband_tx_state_str(txmsg->state),
+		   txmsg->path_msg, buf);
+
+	ret = drm_dp_decode_sideband_req(txmsg, &req);
+	if (ret) {
+		drm_printf(p, "<failed to decode sideband req: %d>\n", ret);
+		return;
+	}
+	drm_dp_dump_sideband_msg_req_body(&req, 1, p);
+
+	switch (req.req_type) {
+	case DP_REMOTE_DPCD_WRITE:
+		kfree(req.u.dpcd_write.bytes);
+		break;
+	case DP_REMOTE_I2C_READ:
+		for (i = 0; i < req.u.i2c_read.num_transactions; i++)
+			kfree(req.u.i2c_read.transactions[i].bytes);
+		break;
+	case DP_REMOTE_I2C_WRITE:
+		kfree(req.u.i2c_write.bytes);
 		break;
 	}
-	raw->cur_len = idx;
 }
 
 static void drm_dp_crc_sideband_chunk_req(u8 *msg, u8 len)
@@ -842,11 +1135,11 @@ static void drm_dp_mst_put_payload_id(struct drm_dp_mst_topology_mgr *mgr,
 	clear_bit(vcpi - 1, &mgr->vcpi_mask);
 
 	for (i = 0; i < mgr->max_payloads; i++) {
-		if (mgr->proposed_vcpis[i])
-			if (mgr->proposed_vcpis[i]->vcpi == vcpi) {
-				mgr->proposed_vcpis[i] = NULL;
-				clear_bit(i + 1, &mgr->payload_mask);
-			}
+		if (mgr->proposed_vcpis[i] &&
+		    mgr->proposed_vcpis[i]->vcpi == vcpi) {
+			mgr->proposed_vcpis[i] = NULL;
+			clear_bit(i + 1, &mgr->payload_mask);
+		}
 	}
 	mutex_unlock(&mgr->payload_lock);
 }
@@ -899,6 +1192,11 @@ static int drm_dp_mst_wait_tx_reply(struct drm_dp_mst_branch *mstb,
 		}
 	}
 out:
+	if (unlikely(ret == -EIO) && drm_debug_enabled(DRM_UT_DP)) {
+		struct drm_printer p = drm_debug_printer(DBG_PREFIX);
+
+		drm_dp_mst_dump_sideband_msg_tx(&p, txmsg);
+	}
 	mutex_unlock(&mgr->qlock);
 
 	return ret;
@@ -1108,39 +1406,194 @@ drm_dp_mst_put_port_malloc(struct drm_dp_mst_port *port)
 }
 EXPORT_SYMBOL(drm_dp_mst_put_port_malloc);
 
-static void drm_dp_destroy_mst_branch_device(struct kref *kref)
+#if IS_ENABLED(CONFIG_DRM_DEBUG_DP_MST_TOPOLOGY_REFS)
+
+#define STACK_DEPTH 8
+
+static noinline void
+__topology_ref_save(struct drm_dp_mst_topology_mgr *mgr,
+		    struct drm_dp_mst_topology_ref_history *history,
+		    enum drm_dp_mst_topology_ref_type type)
 {
-	struct drm_dp_mst_branch *mstb =
-		container_of(kref, struct drm_dp_mst_branch, topology_kref);
-	struct drm_dp_mst_topology_mgr *mgr = mstb->mgr;
-	struct drm_dp_mst_port *port, *tmp;
-	bool wake_tx = false;
+	struct drm_dp_mst_topology_ref_entry *entry = NULL;
+	depot_stack_handle_t backtrace;
+	ulong stack_entries[STACK_DEPTH];
+	uint n;
+	int i;
 
-	mutex_lock(&mgr->lock);
-	list_for_each_entry_safe(port, tmp, &mstb->ports, next) {
-		list_del(&port->next);
-		drm_dp_mst_topology_put_port(port);
+	n = stack_trace_save(stack_entries, ARRAY_SIZE(stack_entries), 1);
+	backtrace = stack_depot_save(stack_entries, n, GFP_KERNEL);
+	if (!backtrace)
+		return;
+
+	/* Try to find an existing entry for this backtrace */
+	for (i = 0; i < history->len; i++) {
+		if (history->entries[i].backtrace == backtrace) {
+			entry = &history->entries[i];
+			break;
+		}
 	}
-	mutex_unlock(&mgr->lock);
 
-	/* drop any tx slots msg */
-	mutex_lock(&mstb->mgr->qlock);
-	if (mstb->tx_slots[0]) {
-		mstb->tx_slots[0]->state = DRM_DP_SIDEBAND_TX_TIMEOUT;
-		mstb->tx_slots[0] = NULL;
-		wake_tx = true;
+	/* Otherwise add one */
+	if (!entry) {
+		struct drm_dp_mst_topology_ref_entry *new;
+		int new_len = history->len + 1;
+
+		new = krealloc(history->entries, sizeof(*new) * new_len,
+			       GFP_KERNEL);
+		if (!new)
+			return;
+
+		entry = &new[history->len];
+		history->len = new_len;
+		history->entries = new;
+
+		entry->backtrace = backtrace;
+		entry->type = type;
+		entry->count = 0;
 	}
-	if (mstb->tx_slots[1]) {
-		mstb->tx_slots[1]->state = DRM_DP_SIDEBAND_TX_TIMEOUT;
-		mstb->tx_slots[1] = NULL;
-		wake_tx = true;
+	entry->count++;
+	entry->ts_nsec = ktime_get_ns();
+}
+
+static int
+topology_ref_history_cmp(const void *a, const void *b)
+{
+	const struct drm_dp_mst_topology_ref_entry *entry_a = a, *entry_b = b;
+
+	if (entry_a->ts_nsec > entry_b->ts_nsec)
+		return 1;
+	else if (entry_a->ts_nsec < entry_b->ts_nsec)
+		return -1;
+	else
+		return 0;
+}
+
+static inline const char *
+topology_ref_type_to_str(enum drm_dp_mst_topology_ref_type type)
+{
+	if (type == DRM_DP_MST_TOPOLOGY_REF_GET)
+		return "get";
+	else
+		return "put";
+}
+
+static void
+__dump_topology_ref_history(struct drm_dp_mst_topology_ref_history *history,
+			    void *ptr, const char *type_str)
+{
+	struct drm_printer p = drm_debug_printer(DBG_PREFIX);
+	char *buf = kzalloc(PAGE_SIZE, GFP_KERNEL);
+	int i;
+
+	if (!buf)
+		return;
+
+	if (!history->len)
+		goto out;
+
+	/* First, sort the list so that it goes from oldest to newest
+	 * reference entry
+	 */
+	sort(history->entries, history->len, sizeof(*history->entries),
+	     topology_ref_history_cmp, NULL);
+
+	drm_printf(&p, "%s (%p) topology count reached 0, dumping history:\n",
+		   type_str, ptr);
+
+	for (i = 0; i < history->len; i++) {
+		const struct drm_dp_mst_topology_ref_entry *entry =
+			&history->entries[i];
+		ulong *entries;
+		uint nr_entries;
+		u64 ts_nsec = entry->ts_nsec;
+		u32 rem_nsec = do_div(ts_nsec, 1000000000);
+
+		nr_entries = stack_depot_fetch(entry->backtrace, &entries);
+		stack_trace_snprint(buf, PAGE_SIZE, entries, nr_entries, 4);
+
+		drm_printf(&p, "  %d %ss (last at %5llu.%06u):\n%s",
+			   entry->count,
+			   topology_ref_type_to_str(entry->type),
+			   ts_nsec, rem_nsec / 1000, buf);
 	}
-	mutex_unlock(&mstb->mgr->qlock);
 
-	if (wake_tx)
-		wake_up_all(&mstb->mgr->tx_waitq);
+	/* Now free the history, since this is the only time we expose it */
+	kfree(history->entries);
+out:
+	kfree(buf);
+}
 
-	drm_dp_mst_put_mstb_malloc(mstb);
+static __always_inline void
+drm_dp_mst_dump_mstb_topology_history(struct drm_dp_mst_branch *mstb)
+{
+	__dump_topology_ref_history(&mstb->topology_ref_history, mstb,
+				    "MSTB");
+}
+
+static __always_inline void
+drm_dp_mst_dump_port_topology_history(struct drm_dp_mst_port *port)
+{
+	__dump_topology_ref_history(&port->topology_ref_history, port,
+				    "Port");
+}
+
+static __always_inline void
+save_mstb_topology_ref(struct drm_dp_mst_branch *mstb,
+		       enum drm_dp_mst_topology_ref_type type)
+{
+	__topology_ref_save(mstb->mgr, &mstb->topology_ref_history, type);
+}
+
+static __always_inline void
+save_port_topology_ref(struct drm_dp_mst_port *port,
+		       enum drm_dp_mst_topology_ref_type type)
+{
+	__topology_ref_save(port->mgr, &port->topology_ref_history, type);
+}
+
+static inline void
+topology_ref_history_lock(struct drm_dp_mst_topology_mgr *mgr)
+{
+	mutex_lock(&mgr->topology_ref_history_lock);
+}
+
+static inline void
+topology_ref_history_unlock(struct drm_dp_mst_topology_mgr *mgr)
+{
+	mutex_unlock(&mgr->topology_ref_history_lock);
+}
+#else
+static inline void
+topology_ref_history_lock(struct drm_dp_mst_topology_mgr *mgr) {}
+static inline void
+topology_ref_history_unlock(struct drm_dp_mst_topology_mgr *mgr) {}
+static inline void
+drm_dp_mst_dump_mstb_topology_history(struct drm_dp_mst_branch *mstb) {}
+static inline void
+drm_dp_mst_dump_port_topology_history(struct drm_dp_mst_port *port) {}
+#define save_mstb_topology_ref(mstb, type)
+#define save_port_topology_ref(port, type)
+#endif
+
+static void drm_dp_destroy_mst_branch_device(struct kref *kref)
+{
+	struct drm_dp_mst_branch *mstb =
+		container_of(kref, struct drm_dp_mst_branch, topology_kref);
+	struct drm_dp_mst_topology_mgr *mgr = mstb->mgr;
+
+	drm_dp_mst_dump_mstb_topology_history(mstb);
+
+	INIT_LIST_HEAD(&mstb->destroy_next);
+
+	/*
+	 * This can get called under mgr->mutex, so we need to perform the
+	 * actual destruction of the mstb in another worker
+	 */
+	mutex_lock(&mgr->delayed_destroy_lock);
+	list_add(&mstb->destroy_next, &mgr->destroy_branch_device_list);
+	mutex_unlock(&mgr->delayed_destroy_lock);
+	schedule_work(&mgr->delayed_destroy_work);
 }
 
 /**
@@ -1168,11 +1621,17 @@ static void drm_dp_destroy_mst_branch_device(struct kref *kref)
 static int __must_check
 drm_dp_mst_topology_try_get_mstb(struct drm_dp_mst_branch *mstb)
 {
-	int ret = kref_get_unless_zero(&mstb->topology_kref);
+	int ret;
 
-	if (ret)
-		DRM_DEBUG("mstb %p (%d)\n", mstb,
-			  kref_read(&mstb->topology_kref));
+	topology_ref_history_lock(mstb->mgr);
+	ret = kref_get_unless_zero(&mstb->topology_kref);
+	if (ret) {
+		DRM_DEBUG("mstb %p (%d)\n",
+			  mstb, kref_read(&mstb->topology_kref));
+		save_mstb_topology_ref(mstb, DRM_DP_MST_TOPOLOGY_REF_GET);
+	}
+
+	topology_ref_history_unlock(mstb->mgr);
 
 	return ret;
 }
@@ -1193,9 +1652,14 @@ drm_dp_mst_topology_try_get_mstb(struct drm_dp_mst_branch *mstb)
  */
 static void drm_dp_mst_topology_get_mstb(struct drm_dp_mst_branch *mstb)
 {
+	topology_ref_history_lock(mstb->mgr);
+
+	save_mstb_topology_ref(mstb, DRM_DP_MST_TOPOLOGY_REF_GET);
 	WARN_ON(kref_read(&mstb->topology_kref) == 0);
 	kref_get(&mstb->topology_kref);
 	DRM_DEBUG("mstb %p (%d)\n", mstb, kref_read(&mstb->topology_kref));
+
+	topology_ref_history_unlock(mstb->mgr);
 }
 
 /**
@@ -1213,27 +1677,14 @@ static void drm_dp_mst_topology_get_mstb(struct drm_dp_mst_branch *mstb)
 static void
 drm_dp_mst_topology_put_mstb(struct drm_dp_mst_branch *mstb)
 {
+	topology_ref_history_lock(mstb->mgr);
+
 	DRM_DEBUG("mstb %p (%d)\n",
 		  mstb, kref_read(&mstb->topology_kref) - 1);
-	kref_put(&mstb->topology_kref, drm_dp_destroy_mst_branch_device);
-}
+	save_mstb_topology_ref(mstb, DRM_DP_MST_TOPOLOGY_REF_PUT);
 
-static void drm_dp_port_teardown_pdt(struct drm_dp_mst_port *port, int old_pdt)
-{
-	struct drm_dp_mst_branch *mstb;
-
-	switch (old_pdt) {
-	case DP_PEER_DEVICE_DP_LEGACY_CONV:
-	case DP_PEER_DEVICE_SST_SINK:
-		/* remove i2c over sideband */
-		drm_dp_mst_unregister_i2c_bus(&port->aux);
-		break;
-	case DP_PEER_DEVICE_MST_BRANCHING:
-		mstb = port->mstb;
-		port->mstb = NULL;
-		drm_dp_mst_topology_put_mstb(mstb);
-		break;
-	}
+	topology_ref_history_unlock(mstb->mgr);
+	kref_put(&mstb->topology_kref, drm_dp_destroy_mst_branch_device);
 }
 
 static void drm_dp_destroy_port(struct kref *kref)
@@ -1242,31 +1693,24 @@ static void drm_dp_destroy_port(struct kref *kref)
 		container_of(kref, struct drm_dp_mst_port, topology_kref);
 	struct drm_dp_mst_topology_mgr *mgr = port->mgr;
 
-	if (!port->input) {
-		kfree(port->cached_edid);
+	drm_dp_mst_dump_port_topology_history(port);
 
-		/*
-		 * The only time we don't have a connector
-		 * on an output port is if the connector init
-		 * fails.
-		 */
-		if (port->connector) {
-			/* we can't destroy the connector here, as
-			 * we might be holding the mode_config.mutex
-			 * from an EDID retrieval */
-
-			mutex_lock(&mgr->destroy_connector_lock);
-			list_add(&port->next, &mgr->destroy_connector_list);
-			mutex_unlock(&mgr->destroy_connector_lock);
-			schedule_work(&mgr->destroy_connector_work);
-			return;
-		}
-		/* no need to clean up vcpi
-		 * as if we have no connector we never setup a vcpi */
-		drm_dp_port_teardown_pdt(port, port->pdt);
-		port->pdt = DP_PEER_DEVICE_NONE;
+	/* There's nothing that needs locking to destroy an input port yet */
+	if (port->input) {
+		drm_dp_mst_put_port_malloc(port);
+		return;
 	}
-	drm_dp_mst_put_port_malloc(port);
+
+	kfree(port->cached_edid);
+
+	/*
+	 * we can't destroy the connector here, as we might be holding the
+	 * mode_config.mutex from an EDID retrieval
+	 */
+	mutex_lock(&mgr->delayed_destroy_lock);
+	list_add(&port->next, &mgr->destroy_port_list);
+	mutex_unlock(&mgr->delayed_destroy_lock);
+	schedule_work(&mgr->delayed_destroy_work);
 }
 
 /**
@@ -1294,12 +1738,17 @@ static void drm_dp_destroy_port(struct kref *kref)
 static int __must_check
 drm_dp_mst_topology_try_get_port(struct drm_dp_mst_port *port)
 {
-	int ret = kref_get_unless_zero(&port->topology_kref);
+	int ret;
 
-	if (ret)
-		DRM_DEBUG("port %p (%d)\n", port,
-			  kref_read(&port->topology_kref));
+	topology_ref_history_lock(port->mgr);
+	ret = kref_get_unless_zero(&port->topology_kref);
+	if (ret) {
+		DRM_DEBUG("port %p (%d)\n",
+			  port, kref_read(&port->topology_kref));
+		save_port_topology_ref(port, DRM_DP_MST_TOPOLOGY_REF_GET);
+	}
 
+	topology_ref_history_unlock(port->mgr);
 	return ret;
 }
 
@@ -1318,9 +1767,14 @@ drm_dp_mst_topology_try_get_port(struct drm_dp_mst_port *port)
  */
 static void drm_dp_mst_topology_get_port(struct drm_dp_mst_port *port)
 {
+	topology_ref_history_lock(port->mgr);
+
 	WARN_ON(kref_read(&port->topology_kref) == 0);
 	kref_get(&port->topology_kref);
 	DRM_DEBUG("port %p (%d)\n", port, kref_read(&port->topology_kref));
+	save_port_topology_ref(port, DRM_DP_MST_TOPOLOGY_REF_GET);
+
+	topology_ref_history_unlock(port->mgr);
 }
 
 /**
@@ -1336,8 +1790,13 @@ static void drm_dp_mst_topology_get_port(struct drm_dp_mst_port *port)
  */
 static void drm_dp_mst_topology_put_port(struct drm_dp_mst_port *port)
 {
+	topology_ref_history_lock(port->mgr);
+
 	DRM_DEBUG("port %p (%d)\n",
 		  port, kref_read(&port->topology_kref) - 1);
+	save_port_topology_ref(port, DRM_DP_MST_TOPOLOGY_REF_PUT);
+
+	topology_ref_history_unlock(port->mgr);
 	kref_put(&port->topology_kref, drm_dp_destroy_port);
 }
 
@@ -1454,38 +1913,79 @@ static u8 drm_dp_calculate_rad(struct drm_dp_mst_port *port,
 	return parent_lct + 1;
 }
 
-/*
- * return sends link address for new mstb
- */
-static bool drm_dp_port_setup_pdt(struct drm_dp_mst_port *port)
+static int drm_dp_port_set_pdt(struct drm_dp_mst_port *port, u8 new_pdt)
 {
-	int ret;
-	u8 rad[6], lct;
-	bool send_link = false;
+	struct drm_dp_mst_topology_mgr *mgr = port->mgr;
+	struct drm_dp_mst_branch *mstb;
+	u8 rad[8], lct;
+	int ret = 0;
+
+	if (port->pdt == new_pdt)
+		return 0;
+
+	/* Teardown the old pdt, if there is one */
+	switch (port->pdt) {
+	case DP_PEER_DEVICE_DP_LEGACY_CONV:
+	case DP_PEER_DEVICE_SST_SINK:
+		/*
+		 * If the new PDT would also have an i2c bus, don't bother
+		 * with reregistering it
+		 */
+		if (new_pdt == DP_PEER_DEVICE_DP_LEGACY_CONV ||
+		    new_pdt == DP_PEER_DEVICE_SST_SINK) {
+			port->pdt = new_pdt;
+			return 0;
+		}
+
+		/* remove i2c over sideband */
+		drm_dp_mst_unregister_i2c_bus(&port->aux);
+		break;
+	case DP_PEER_DEVICE_MST_BRANCHING:
+		mutex_lock(&mgr->lock);
+		drm_dp_mst_topology_put_mstb(port->mstb);
+		port->mstb = NULL;
+		mutex_unlock(&mgr->lock);
+		break;
+	}
+
+	port->pdt = new_pdt;
 	switch (port->pdt) {
 	case DP_PEER_DEVICE_DP_LEGACY_CONV:
 	case DP_PEER_DEVICE_SST_SINK:
 		/* add i2c over sideband */
 		ret = drm_dp_mst_register_i2c_bus(&port->aux);
 		break;
+
 	case DP_PEER_DEVICE_MST_BRANCHING:
 		lct = drm_dp_calculate_rad(port, rad);
+		mstb = drm_dp_add_mst_branch_device(lct, rad);
+		if (!mstb) {
+			ret = -ENOMEM;
+			DRM_ERROR("Failed to create MSTB for port %p", port);
+			goto out;
+		}
 
-		port->mstb = drm_dp_add_mst_branch_device(lct, rad);
-		if (port->mstb) {
-			port->mstb->mgr = port->mgr;
-			port->mstb->port_parent = port;
-			/*
-			 * Make sure this port's memory allocation stays
-			 * around until its child MSTB releases it
-			 */
-			drm_dp_mst_get_port_malloc(port);
+		mutex_lock(&mgr->lock);
+		port->mstb = mstb;
+		mstb->mgr = port->mgr;
+		mstb->port_parent = port;
 
-			send_link = true;
-		}
+		/*
+		 * Make sure this port's memory allocation stays
+		 * around until its child MSTB releases it
+		 */
+		drm_dp_mst_get_port_malloc(port);
+		mutex_unlock(&mgr->lock);
+
+		/* And make sure we send a link address for this */
+		ret = 1;
 		break;
 	}
-	return send_link;
+
+out:
+	if (ret < 0)
+		port->pdt = DP_PEER_DEVICE_NONE;
+	return ret;
 }
 
 /**
@@ -1617,44 +2117,131 @@ void drm_dp_mst_connector_early_unregister(struct drm_connector *connector,
 }
 EXPORT_SYMBOL(drm_dp_mst_connector_early_unregister);
 
-static void drm_dp_add_port(struct drm_dp_mst_branch *mstb,
-			    struct drm_device *dev,
-			    struct drm_dp_link_addr_reply_port *port_msg)
+static void
+drm_dp_mst_port_add_connector(struct drm_dp_mst_branch *mstb,
+			      struct drm_dp_mst_port *port)
+{
+	struct drm_dp_mst_topology_mgr *mgr = port->mgr;
+	char proppath[255];
+	int ret;
+
+	build_mst_prop_path(mstb, port->port_num, proppath, sizeof(proppath));
+	port->connector = mgr->cbs->add_connector(mgr, port, proppath);
+	if (!port->connector) {
+		ret = -ENOMEM;
+		goto error;
+	}
+
+	if ((port->pdt == DP_PEER_DEVICE_DP_LEGACY_CONV ||
+	     port->pdt == DP_PEER_DEVICE_SST_SINK) &&
+	    port->port_num >= DP_MST_LOGICAL_PORT_0) {
+		port->cached_edid = drm_get_edid(port->connector,
+						 &port->aux.ddc);
+		drm_connector_set_tile_property(port->connector);
+	}
+
+	mgr->cbs->register_connector(port->connector);
+	return;
+
+error:
+	DRM_ERROR("Failed to create connector for port %p: %d\n", port, ret);
+}
+
+/*
+ * Drop a topology reference, and unlink the port from the in-memory topology
+ * layout
+ */
+static void
+drm_dp_mst_topology_unlink_port(struct drm_dp_mst_topology_mgr *mgr,
+				struct drm_dp_mst_port *port)
+{
+	mutex_lock(&mgr->lock);
+	list_del(&port->next);
+	mutex_unlock(&mgr->lock);
+	drm_dp_mst_topology_put_port(port);
+}
+
+static struct drm_dp_mst_port *
+drm_dp_mst_add_port(struct drm_device *dev,
+		    struct drm_dp_mst_topology_mgr *mgr,
+		    struct drm_dp_mst_branch *mstb, u8 port_number)
+{
+	struct drm_dp_mst_port *port = kzalloc(sizeof(*port), GFP_KERNEL);
+
+	if (!port)
+		return NULL;
+
+	kref_init(&port->topology_kref);
+	kref_init(&port->malloc_kref);
+	port->parent = mstb;
+	port->port_num = port_number;
+	port->mgr = mgr;
+	port->aux.name = "DPMST";
+	port->aux.dev = dev->dev;
+	port->aux.is_remote = true;
+
+	/*
+	 * Make sure the memory allocation for our parent branch stays
+	 * around until our own memory allocation is released
+	 */
+	drm_dp_mst_get_mstb_malloc(mstb);
+
+	return port;
+}
+
+static int
+drm_dp_mst_handle_link_address_port(struct drm_dp_mst_branch *mstb,
+				    struct drm_device *dev,
+				    struct drm_dp_link_addr_reply_port *port_msg)
 {
+	struct drm_dp_mst_topology_mgr *mgr = mstb->mgr;
 	struct drm_dp_mst_port *port;
-	bool ret;
-	bool created = false;
-	int old_pdt = 0;
-	int old_ddps = 0;
+	int old_ddps = 0, ret;
+	u8 new_pdt = DP_PEER_DEVICE_NONE;
+	bool created = false, send_link_addr = false, changed = false;
 
 	port = drm_dp_get_port(mstb, port_msg->port_number);
 	if (!port) {
-		port = kzalloc(sizeof(*port), GFP_KERNEL);
+		port = drm_dp_mst_add_port(dev, mgr, mstb,
+					   port_msg->port_number);
 		if (!port)
-			return;
-		kref_init(&port->topology_kref);
-		kref_init(&port->malloc_kref);
-		port->parent = mstb;
-		port->port_num = port_msg->port_number;
-		port->mgr = mstb->mgr;
-		port->aux.name = "DPMST";
-		port->aux.dev = dev->dev;
-		port->aux.is_remote = true;
-
-		/*
-		 * Make sure the memory allocation for our parent branch stays
-		 * around until our own memory allocation is released
+			return -ENOMEM;
+		created = true;
+		changed = true;
+	} else if (!port->input && port_msg->input_port && port->connector) {
+		/* Since port->connector can't be changed here, we create a
+		 * new port if input_port changes from 0 to 1
 		 */
-		drm_dp_mst_get_mstb_malloc(mstb);
-
+		drm_dp_mst_topology_unlink_port(mgr, port);
+		drm_dp_mst_topology_put_port(port);
+		port = drm_dp_mst_add_port(dev, mgr, mstb,
+					   port_msg->port_number);
+		if (!port)
+			return -ENOMEM;
+		changed = true;
 		created = true;
-	} else {
-		old_pdt = port->pdt;
+	} else if (port->input && !port_msg->input_port) {
+		changed = true;
+	} else if (port->connector) {
+		/* We're updating a port that's exposed to userspace, so do it
+		 * under lock
+		 */
+		drm_modeset_lock(&mgr->base.lock, NULL);
+
 		old_ddps = port->ddps;
+		changed = port->ddps != port_msg->ddps ||
+			(port->ddps &&
+			 (port->ldps != port_msg->legacy_device_plug_status ||
+			  port->dpcd_rev != port_msg->dpcd_revision ||
+			  port->mcs != port_msg->mcs ||
+			  port->pdt != port_msg->peer_device_type ||
+			  port->num_sdp_stream_sinks !=
+			  port_msg->num_sdp_stream_sinks));
 	}
 
-	port->pdt = port_msg->peer_device_type;
 	port->input = port_msg->input_port;
+	if (!port->input)
+		new_pdt = port_msg->peer_device_type;
 	port->mcs = port_msg->mcs;
 	port->ddps = port_msg->ddps;
 	port->ldps = port_msg->legacy_device_plug_status;
@@ -1665,77 +2252,104 @@ static void drm_dp_add_port(struct drm_dp_mst_branch *mstb,
 	/* manage mstb port lists with mgr lock - take a reference
 	   for this list */
 	if (created) {
-		mutex_lock(&mstb->mgr->lock);
+		mutex_lock(&mgr->lock);
 		drm_dp_mst_topology_get_port(port);
 		list_add(&port->next, &mstb->ports);
-		mutex_unlock(&mstb->mgr->lock);
+		mutex_unlock(&mgr->lock);
 	}
 
 	if (old_ddps != port->ddps) {
 		if (port->ddps) {
 			if (!port->input) {
-				drm_dp_send_enum_path_resources(mstb->mgr,
-								mstb, port);
+				drm_dp_send_enum_path_resources(mgr, mstb,
+								port);
 			}
 		} else {
 			port->available_pbn = 0;
 		}
 	}
 
-	if (old_pdt != port->pdt && !port->input) {
-		drm_dp_port_teardown_pdt(port, old_pdt);
-
-		ret = drm_dp_port_setup_pdt(port);
-		if (ret == true)
-			drm_dp_send_link_address(mstb->mgr, port->mstb);
+	ret = drm_dp_port_set_pdt(port, new_pdt);
+	if (ret == 1) {
+		send_link_addr = true;
+	} else if (ret < 0) {
+		DRM_ERROR("Failed to change PDT on port %p: %d\n",
+			  port, ret);
+		goto fail;
 	}
 
-	if (created && !port->input) {
-		char proppath[255];
-
-		build_mst_prop_path(mstb, port->port_num, proppath,
-				    sizeof(proppath));
-		port->connector = (*mstb->mgr->cbs->add_connector)(mstb->mgr,
-								   port,
-								   proppath);
-		if (!port->connector) {
-			/* remove it from the port list */
-			mutex_lock(&mstb->mgr->lock);
-			list_del(&port->next);
-			mutex_unlock(&mstb->mgr->lock);
-			/* drop port list reference */
-			drm_dp_mst_topology_put_port(port);
-			goto out;
-		}
-		if ((port->pdt == DP_PEER_DEVICE_DP_LEGACY_CONV ||
-		     port->pdt == DP_PEER_DEVICE_SST_SINK) &&
-		    port->port_num >= DP_MST_LOGICAL_PORT_0) {
-			port->cached_edid = drm_get_edid(port->connector,
-							 &port->aux.ddc);
-			drm_connector_set_tile_property(port->connector);
-		}
-		(*mstb->mgr->cbs->register_connector)(port->connector);
+	/*
+	 * If this port wasn't just created, then we're reprobing because
+	 * we're coming out of suspend. In this case, always resend the link
+	 * address if there's an MSTB on this port
+	 */
+	if (!created && port->pdt == DP_PEER_DEVICE_MST_BRANCHING)
+		send_link_addr = true;
+
+	if (port->connector)
+		drm_modeset_unlock(&mgr->base.lock);
+	else if (!port->input)
+		drm_dp_mst_port_add_connector(mstb, port);
+
+	if (send_link_addr && port->mstb) {
+		ret = drm_dp_send_link_address(mgr, port->mstb);
+		if (ret == 1) /* MSTB below us changed */
+			changed = true;
+		else if (ret < 0)
+			goto fail_put;
 	}
 
-out:
 	/* put reference to this port */
 	drm_dp_mst_topology_put_port(port);
+	return changed;
+
+fail:
+	drm_dp_mst_topology_unlink_port(mgr, port);
+	if (port->connector)
+		drm_modeset_unlock(&mgr->base.lock);
+fail_put:
+	drm_dp_mst_topology_put_port(port);
+	return ret;
 }
 
-static void drm_dp_update_port(struct drm_dp_mst_branch *mstb,
-			       struct drm_dp_connection_status_notify *conn_stat)
+static void
+drm_dp_mst_handle_conn_stat(struct drm_dp_mst_branch *mstb,
+			    struct drm_dp_connection_status_notify *conn_stat)
 {
+	struct drm_dp_mst_topology_mgr *mgr = mstb->mgr;
 	struct drm_dp_mst_port *port;
-	int old_pdt;
-	int old_ddps;
-	bool dowork = false;
+	int old_ddps, ret;
+	u8 new_pdt;
+	bool dowork = false, create_connector = false;
+
 	port = drm_dp_get_port(mstb, conn_stat->port_number);
 	if (!port)
 		return;
 
+	if (port->connector) {
+		if (!port->input && conn_stat->input_port) {
+			/*
+			 * We can't remove a connector from an already exposed
+			 * port, so just throw the port out and make sure we
+			 * reprobe the link address of it's parent MSTB
+			 */
+			drm_dp_mst_topology_unlink_port(mgr, port);
+			mstb->link_address_sent = false;
+			dowork = true;
+			goto out;
+		}
+
+		/* Locking is only needed if the port's exposed to userspace */
+		drm_modeset_lock(&mgr->base.lock, NULL);
+	} else if (port->input && !conn_stat->input_port) {
+		create_connector = true;
+		/* Reprobe link address so we get num_sdp_streams */
+		mstb->link_address_sent = false;
+		dowork = true;
+	}
+
 	old_ddps = port->ddps;
-	old_pdt = port->pdt;
-	port->pdt = conn_stat->peer_device_type;
+	port->input = conn_stat->input_port;
 	port->mcs = conn_stat->message_capability_status;
 	port->ldps = conn_stat->legacy_device_plug_status;
 	port->ddps = conn_stat->displayport_device_plug_status;
@@ -1747,17 +2361,27 @@ static void drm_dp_update_port(struct drm_dp_mst_branch *mstb,
 			port->available_pbn = 0;
 		}
 	}
-	if (old_pdt != port->pdt && !port->input) {
-		drm_dp_port_teardown_pdt(port, old_pdt);
 
-		if (drm_dp_port_setup_pdt(port))
-			dowork = true;
+	new_pdt = port->input ? DP_PEER_DEVICE_NONE : conn_stat->peer_device_type;
+
+	ret = drm_dp_port_set_pdt(port, new_pdt);
+	if (ret == 1) {
+		dowork = true;
+	} else if (ret < 0) {
+		DRM_ERROR("Failed to change PDT for port %p: %d\n",
+			  port, ret);
+		dowork = false;
 	}
 
+	if (port->connector)
+		drm_modeset_unlock(&mgr->base.lock);
+	else if (create_connector)
+		drm_dp_mst_port_add_connector(mstb, port);
+
+out:
 	drm_dp_mst_topology_put_port(port);
 	if (dowork)
 		queue_work(system_long_wq, &mstb->mgr->work);
-
 }
 
 static struct drm_dp_mst_branch *drm_dp_get_mst_branch_device(struct drm_dp_mst_topology_mgr *mgr,
@@ -1800,7 +2424,7 @@ out:
 
 static struct drm_dp_mst_branch *get_mst_branch_device_by_guid_helper(
 	struct drm_dp_mst_branch *mstb,
-	uint8_t *guid)
+	const uint8_t *guid)
 {
 	struct drm_dp_mst_branch *found_mstb;
 	struct drm_dp_mst_port *port;
@@ -1824,7 +2448,7 @@ static struct drm_dp_mst_branch *get_mst_branch_device_by_guid_helper(
 
 static struct drm_dp_mst_branch *
 drm_dp_get_mst_branch_device_by_guid(struct drm_dp_mst_topology_mgr *mgr,
-				     uint8_t *guid)
+				     const uint8_t *guid)
 {
 	struct drm_dp_mst_branch *mstb;
 	int ret;
@@ -1843,41 +2467,62 @@ drm_dp_get_mst_branch_device_by_guid(struct drm_dp_mst_topology_mgr *mgr,
 	return mstb;
 }
 
-static void drm_dp_check_and_send_link_address(struct drm_dp_mst_topology_mgr *mgr,
+static int drm_dp_check_and_send_link_address(struct drm_dp_mst_topology_mgr *mgr,
 					       struct drm_dp_mst_branch *mstb)
 {
 	struct drm_dp_mst_port *port;
-	struct drm_dp_mst_branch *mstb_child;
-	if (!mstb->link_address_sent)
-		drm_dp_send_link_address(mgr, mstb);
+	int ret;
+	bool changed = false;
+
+	if (!mstb->link_address_sent) {
+		ret = drm_dp_send_link_address(mgr, mstb);
+		if (ret == 1)
+			changed = true;
+		else if (ret < 0)
+			return ret;
+	}
 
 	list_for_each_entry(port, &mstb->ports, next) {
-		if (port->input)
-			continue;
+		struct drm_dp_mst_branch *mstb_child = NULL;
 
-		if (!port->ddps)
+		if (port->input || !port->ddps)
 			continue;
 
-		if (!port->available_pbn)
+		if (!port->available_pbn) {
+			drm_modeset_lock(&mgr->base.lock, NULL);
 			drm_dp_send_enum_path_resources(mgr, mstb, port);
+			drm_modeset_unlock(&mgr->base.lock);
+			changed = true;
+		}
 
-		if (port->mstb) {
+		if (port->mstb)
 			mstb_child = drm_dp_mst_topology_get_mstb_validated(
 			    mgr, port->mstb);
-			if (mstb_child) {
-				drm_dp_check_and_send_link_address(mgr, mstb_child);
-				drm_dp_mst_topology_put_mstb(mstb_child);
-			}
+
+		if (mstb_child) {
+			ret = drm_dp_check_and_send_link_address(mgr,
+								 mstb_child);
+			drm_dp_mst_topology_put_mstb(mstb_child);
+			if (ret == 1)
+				changed = true;
+			else if (ret < 0)
+				return ret;
 		}
 	}
+
+	return changed;
 }
 
 static void drm_dp_mst_link_probe_work(struct work_struct *work)
 {
-	struct drm_dp_mst_topology_mgr *mgr = container_of(work, struct drm_dp_mst_topology_mgr, work);
+	struct drm_dp_mst_topology_mgr *mgr =
+		container_of(work, struct drm_dp_mst_topology_mgr, work);
+	struct drm_device *dev = mgr->dev;
 	struct drm_dp_mst_branch *mstb;
 	int ret;
 
+	mutex_lock(&mgr->probe_lock);
+
 	mutex_lock(&mgr->lock);
 	mstb = mgr->mst_primary;
 	if (mstb) {
@@ -1886,10 +2531,17 @@ static void drm_dp_mst_link_probe_work(struct work_struct *work)
 			mstb = NULL;
 	}
 	mutex_unlock(&mgr->lock);
-	if (mstb) {
-		drm_dp_check_and_send_link_address(mgr, mstb);
-		drm_dp_mst_topology_put_mstb(mstb);
+	if (!mstb) {
+		mutex_unlock(&mgr->probe_lock);
+		return;
 	}
+
+	ret = drm_dp_check_and_send_link_address(mgr, mstb);
+	drm_dp_mst_topology_put_mstb(mstb);
+
+	mutex_unlock(&mgr->probe_lock);
+	if (ret)
+		drm_kms_helper_hotplug_event(dev);
 }
 
 static bool drm_dp_validate_guid(struct drm_dp_mst_topology_mgr *mgr,
@@ -2035,8 +2687,11 @@ static int process_single_tx_qlock(struct drm_dp_mst_topology_mgr *mgr,
 	idx += tosend + 1;
 
 	ret = drm_dp_send_sideband_msg(mgr, up, chunk, idx);
-	if (ret) {
-		DRM_DEBUG_KMS("sideband msg failed to send\n");
+	if (unlikely(ret) && drm_debug_enabled(DRM_UT_DP)) {
+		struct drm_printer p = drm_debug_printer(DBG_PREFIX);
+
+		drm_printf(&p, "sideband msg failed to send\n");
+		drm_dp_mst_dump_sideband_msg_tx(&p, txmsg);
 		return ret;
 	}
 
@@ -2098,21 +2753,52 @@ static void drm_dp_queue_down_tx(struct drm_dp_mst_topology_mgr *mgr,
 {
 	mutex_lock(&mgr->qlock);
 	list_add_tail(&txmsg->next, &mgr->tx_msg_downq);
+
+	if (drm_debug_enabled(DRM_UT_DP)) {
+		struct drm_printer p = drm_debug_printer(DBG_PREFIX);
+
+		drm_dp_mst_dump_sideband_msg_tx(&p, txmsg);
+	}
+
 	if (list_is_singular(&mgr->tx_msg_downq))
 		process_single_down_tx_qlock(mgr);
 	mutex_unlock(&mgr->qlock);
 }
 
-static void drm_dp_send_link_address(struct drm_dp_mst_topology_mgr *mgr,
+static void
+drm_dp_dump_link_address(struct drm_dp_link_address_ack_reply *reply)
+{
+	struct drm_dp_link_addr_reply_port *port_reply;
+	int i;
+
+	for (i = 0; i < reply->nports; i++) {
+		port_reply = &reply->ports[i];
+		DRM_DEBUG_KMS("port %d: input %d, pdt: %d, pn: %d, dpcd_rev: %02x, mcs: %d, ddps: %d, ldps %d, sdp %d/%d\n",
+			      i,
+			      port_reply->input_port,
+			      port_reply->peer_device_type,
+			      port_reply->port_number,
+			      port_reply->dpcd_revision,
+			      port_reply->mcs,
+			      port_reply->ddps,
+			      port_reply->legacy_device_plug_status,
+			      port_reply->num_sdp_streams,
+			      port_reply->num_sdp_stream_sinks);
+	}
+}
+
+static int drm_dp_send_link_address(struct drm_dp_mst_topology_mgr *mgr,
 				     struct drm_dp_mst_branch *mstb)
 {
-	int len;
 	struct drm_dp_sideband_msg_tx *txmsg;
-	int ret;
+	struct drm_dp_link_address_ack_reply *reply;
+	struct drm_dp_mst_port *port, *tmp;
+	int i, len, ret, port_mask = 0;
+	bool changed = false;
 
 	txmsg = kzalloc(sizeof(*txmsg), GFP_KERNEL);
 	if (!txmsg)
-		return;
+		return -ENOMEM;
 
 	txmsg->dst = mstb;
 	len = build_link_address(txmsg);
@@ -2120,48 +2806,67 @@ static void drm_dp_send_link_address(struct drm_dp_mst_topology_mgr *mgr,
 	mstb->link_address_sent = true;
 	drm_dp_queue_down_tx(mgr, txmsg);
 
+	/* FIXME: Actually do some real error handling here */
 	ret = drm_dp_mst_wait_tx_reply(mstb, txmsg);
-	if (ret > 0) {
-		int i;
+	if (ret <= 0) {
+		DRM_ERROR("Sending link address failed with %d\n", ret);
+		goto out;
+	}
+	if (txmsg->reply.reply_type == DP_SIDEBAND_REPLY_NAK) {
+		DRM_ERROR("link address NAK received\n");
+		ret = -EIO;
+		goto out;
+	}
 
-		if (txmsg->reply.reply_type == DP_SIDEBAND_REPLY_NAK) {
-			DRM_DEBUG_KMS("link address nak received\n");
-		} else {
-			DRM_DEBUG_KMS("link address reply: %d\n", txmsg->reply.u.link_addr.nports);
-			for (i = 0; i < txmsg->reply.u.link_addr.nports; i++) {
-				DRM_DEBUG_KMS("port %d: input %d, pdt: %d, pn: %d, dpcd_rev: %02x, mcs: %d, ddps: %d, ldps %d, sdp %d/%d\n", i,
-				       txmsg->reply.u.link_addr.ports[i].input_port,
-				       txmsg->reply.u.link_addr.ports[i].peer_device_type,
-				       txmsg->reply.u.link_addr.ports[i].port_number,
-				       txmsg->reply.u.link_addr.ports[i].dpcd_revision,
-				       txmsg->reply.u.link_addr.ports[i].mcs,
-				       txmsg->reply.u.link_addr.ports[i].ddps,
-				       txmsg->reply.u.link_addr.ports[i].legacy_device_plug_status,
-				       txmsg->reply.u.link_addr.ports[i].num_sdp_streams,
-				       txmsg->reply.u.link_addr.ports[i].num_sdp_stream_sinks);
-			}
+	reply = &txmsg->reply.u.link_addr;
+	DRM_DEBUG_KMS("link address reply: %d\n", reply->nports);
+	drm_dp_dump_link_address(reply);
 
-			drm_dp_check_mstb_guid(mstb, txmsg->reply.u.link_addr.guid);
+	drm_dp_check_mstb_guid(mstb, reply->guid);
 
-			for (i = 0; i < txmsg->reply.u.link_addr.nports; i++) {
-				drm_dp_add_port(mstb, mgr->dev, &txmsg->reply.u.link_addr.ports[i]);
-			}
-			drm_kms_helper_hotplug_event(mgr->dev);
-		}
-	} else {
-		mstb->link_address_sent = false;
-		DRM_DEBUG_KMS("link address failed %d\n", ret);
+	for (i = 0; i < reply->nports; i++) {
+		port_mask |= BIT(reply->ports[i].port_number);
+		ret = drm_dp_mst_handle_link_address_port(mstb, mgr->dev,
+							  &reply->ports[i]);
+		if (ret == 1)
+			changed = true;
+		else if (ret < 0)
+			goto out;
 	}
 
+	/* Prune any ports that are currently a part of mstb in our in-memory
+	 * topology, but were not seen in this link address. Usually this
+	 * means that they were removed while the topology was out of sync,
+	 * e.g. during suspend/resume
+	 */
+	mutex_lock(&mgr->lock);
+	list_for_each_entry_safe(port, tmp, &mstb->ports, next) {
+		if (port_mask & BIT(port->port_num))
+			continue;
+
+		DRM_DEBUG_KMS("port %d was not in link address, removing\n",
+			      port->port_num);
+		list_del(&port->next);
+		drm_dp_mst_topology_put_port(port);
+		changed = true;
+	}
+	mutex_unlock(&mgr->lock);
+
+out:
+	if (ret <= 0)
+		mstb->link_address_sent = false;
 	kfree(txmsg);
+	return ret < 0 ? ret : changed;
 }
 
-static int drm_dp_send_enum_path_resources(struct drm_dp_mst_topology_mgr *mgr,
-					   struct drm_dp_mst_branch *mstb,
-					   struct drm_dp_mst_port *port)
+static int
+drm_dp_send_enum_path_resources(struct drm_dp_mst_topology_mgr *mgr,
+				struct drm_dp_mst_branch *mstb,
+				struct drm_dp_mst_port *port)
 {
-	int len;
+	struct drm_dp_enum_path_resources_ack_reply *path_res;
 	struct drm_dp_sideband_msg_tx *txmsg;
+	int len;
 	int ret;
 
 	txmsg = kzalloc(sizeof(*txmsg), GFP_KERNEL);
@@ -2175,14 +2880,20 @@ static int drm_dp_send_enum_path_resources(struct drm_dp_mst_topology_mgr *mgr,
 
 	ret = drm_dp_mst_wait_tx_reply(mstb, txmsg);
 	if (ret > 0) {
+		path_res = &txmsg->reply.u.path_resources;
+
 		if (txmsg->reply.reply_type == DP_SIDEBAND_REPLY_NAK) {
 			DRM_DEBUG_KMS("enum path resources nak received\n");
 		} else {
-			if (port->port_num != txmsg->reply.u.path_resources.port_number)
+			if (port->port_num != path_res->port_number)
 				DRM_ERROR("got incorrect port in response\n");
-			DRM_DEBUG_KMS("enum path resources %d: %d %d\n", txmsg->reply.u.path_resources.port_number, txmsg->reply.u.path_resources.full_payload_bw_number,
-			       txmsg->reply.u.path_resources.avail_payload_bw_number);
-			port->available_pbn = txmsg->reply.u.path_resources.avail_payload_bw_number;
+
+			DRM_DEBUG_KMS("enum path resources %d: %d %d\n",
+				      path_res->port_number,
+				      path_res->full_payload_bw_number,
+				      path_res->avail_payload_bw_number);
+			port->available_pbn =
+				path_res->avail_payload_bw_number;
 		}
 	}
 
@@ -2655,30 +3366,13 @@ static int drm_dp_send_up_ack_reply(struct drm_dp_mst_topology_mgr *mgr,
 	return 0;
 }
 
-static bool drm_dp_get_vc_payload_bw(int dp_link_bw,
-				     int dp_link_count,
-				     int *out)
+static int drm_dp_get_vc_payload_bw(u8 dp_link_bw, u8  dp_link_count)
 {
-	switch (dp_link_bw) {
-	default:
+	if (dp_link_bw == 0 || dp_link_count == 0)
 		DRM_DEBUG_KMS("invalid link bandwidth in DPCD: %x (link count: %d)\n",
 			      dp_link_bw, dp_link_count);
-		return false;
 
-	case DP_LINK_BW_1_62:
-		*out = 3 * dp_link_count;
-		break;
-	case DP_LINK_BW_2_7:
-		*out = 5 * dp_link_count;
-		break;
-	case DP_LINK_BW_5_4:
-		*out = 10 * dp_link_count;
-		break;
-	case DP_LINK_BW_8_1:
-		*out = 15 * dp_link_count;
-		break;
-	}
-	return true;
+	return dp_link_bw * dp_link_count / 2;
 }
 
 /**
@@ -2710,9 +3404,9 @@ int drm_dp_mst_topology_mgr_set_mst(struct drm_dp_mst_topology_mgr *mgr, bool ms
 			goto out_unlock;
 		}
 
-		if (!drm_dp_get_vc_payload_bw(mgr->dpcd[1],
-					      mgr->dpcd[2] & DP_MAX_LANE_COUNT_MASK,
-					      &mgr->pbn_div)) {
+		mgr->pbn_div = drm_dp_get_vc_payload_bw(mgr->dpcd[1],
+							mgr->dpcd[2] & DP_MAX_LANE_COUNT_MASK);
+		if (mgr->pbn_div == 0) {
 			ret = -EINVAL;
 			goto out_unlock;
 		}
@@ -2767,6 +3461,23 @@ out_unlock:
 }
 EXPORT_SYMBOL(drm_dp_mst_topology_mgr_set_mst);
 
+static void
+drm_dp_mst_topology_mgr_invalidate_mstb(struct drm_dp_mst_branch *mstb)
+{
+	struct drm_dp_mst_port *port;
+
+	/* The link address will need to be re-sent on resume */
+	mstb->link_address_sent = false;
+
+	list_for_each_entry(port, &mstb->ports, next) {
+		/* The PBN for each port will also need to be re-probed */
+		port->available_pbn = 0;
+
+		if (port->mstb)
+			drm_dp_mst_topology_mgr_invalidate_mstb(port->mstb);
+	}
+}
+
 /**
  * drm_dp_mst_topology_mgr_suspend() - suspend the MST manager
  * @mgr: manager to suspend
@@ -2780,62 +3491,89 @@ void drm_dp_mst_topology_mgr_suspend(struct drm_dp_mst_topology_mgr *mgr)
 	drm_dp_dpcd_writeb(mgr->aux, DP_MSTM_CTRL,
 			   DP_MST_EN | DP_UPSTREAM_IS_SRC);
 	mutex_unlock(&mgr->lock);
+	flush_work(&mgr->up_req_work);
 	flush_work(&mgr->work);
-	flush_work(&mgr->destroy_connector_work);
+	flush_work(&mgr->delayed_destroy_work);
+
+	mutex_lock(&mgr->lock);
+	if (mgr->mst_state && mgr->mst_primary)
+		drm_dp_mst_topology_mgr_invalidate_mstb(mgr->mst_primary);
+	mutex_unlock(&mgr->lock);
 }
 EXPORT_SYMBOL(drm_dp_mst_topology_mgr_suspend);
 
 /**
  * drm_dp_mst_topology_mgr_resume() - resume the MST manager
  * @mgr: manager to resume
+ * @sync: whether or not to perform topology reprobing synchronously
  *
  * This will fetch DPCD and see if the device is still there,
  * if it is, it will rewrite the MSTM control bits, and return.
  *
- * if the device fails this returns -1, and the driver should do
+ * If the device fails this returns -1, and the driver should do
  * a full MST reprobe, in case we were undocked.
+ *
+ * During system resume (where it is assumed that the driver will be calling
+ * drm_atomic_helper_resume()) this function should be called beforehand with
+ * @sync set to true. In contexts like runtime resume where the driver is not
+ * expected to be calling drm_atomic_helper_resume(), this function should be
+ * called with @sync set to false in order to avoid deadlocking.
+ *
+ * Returns: -1 if the MST topology was removed while we were suspended, 0
+ * otherwise.
  */
-int drm_dp_mst_topology_mgr_resume(struct drm_dp_mst_topology_mgr *mgr)
+int drm_dp_mst_topology_mgr_resume(struct drm_dp_mst_topology_mgr *mgr,
+				   bool sync)
 {
-	int ret = 0;
+	int ret;
+	u8 guid[16];
 
 	mutex_lock(&mgr->lock);
+	if (!mgr->mst_primary)
+		goto out_fail;
 
-	if (mgr->mst_primary) {
-		int sret;
-		u8 guid[16];
+	ret = drm_dp_dpcd_read(mgr->aux, DP_DPCD_REV, mgr->dpcd,
+			       DP_RECEIVER_CAP_SIZE);
+	if (ret != DP_RECEIVER_CAP_SIZE) {
+		DRM_DEBUG_KMS("dpcd read failed - undocked during suspend?\n");
+		goto out_fail;
+	}
 
-		sret = drm_dp_dpcd_read(mgr->aux, DP_DPCD_REV, mgr->dpcd, DP_RECEIVER_CAP_SIZE);
-		if (sret != DP_RECEIVER_CAP_SIZE) {
-			DRM_DEBUG_KMS("dpcd read failed - undocked during suspend?\n");
-			ret = -1;
-			goto out_unlock;
-		}
+	ret = drm_dp_dpcd_writeb(mgr->aux, DP_MSTM_CTRL,
+				 DP_MST_EN |
+				 DP_UP_REQ_EN |
+				 DP_UPSTREAM_IS_SRC);
+	if (ret < 0) {
+		DRM_DEBUG_KMS("mst write failed - undocked during suspend?\n");
+		goto out_fail;
+	}
 
-		ret = drm_dp_dpcd_writeb(mgr->aux, DP_MSTM_CTRL,
-					 DP_MST_EN | DP_UP_REQ_EN | DP_UPSTREAM_IS_SRC);
-		if (ret < 0) {
-			DRM_DEBUG_KMS("mst write failed - undocked during suspend?\n");
-			ret = -1;
-			goto out_unlock;
-		}
+	/* Some hubs forget their guids after they resume */
+	ret = drm_dp_dpcd_read(mgr->aux, DP_GUID, guid, 16);
+	if (ret != 16) {
+		DRM_DEBUG_KMS("dpcd read failed - undocked during suspend?\n");
+		goto out_fail;
+	}
+	drm_dp_check_mstb_guid(mgr->mst_primary, guid);
 
-		/* Some hubs forget their guids after they resume */
-		sret = drm_dp_dpcd_read(mgr->aux, DP_GUID, guid, 16);
-		if (sret != 16) {
-			DRM_DEBUG_KMS("dpcd read failed - undocked during suspend?\n");
-			ret = -1;
-			goto out_unlock;
-		}
-		drm_dp_check_mstb_guid(mgr->mst_primary, guid);
+	/*
+	 * For the final step of resuming the topology, we need to bring the
+	 * state of our in-memory topology back into sync with reality. So,
+	 * restart the probing process as if we're probing a new hub
+	 */
+	queue_work(system_long_wq, &mgr->work);
+	mutex_unlock(&mgr->lock);
 
-		ret = 0;
-	} else
-		ret = -1;
+	if (sync) {
+		DRM_DEBUG_KMS("Waiting for link probe work to finish re-syncing topology...\n");
+		flush_work(&mgr->work);
+	}
 
-out_unlock:
+	return 0;
+
+out_fail:
 	mutex_unlock(&mgr->lock);
-	return ret;
+	return -1;
 }
 EXPORT_SYMBOL(drm_dp_mst_topology_mgr_resume);
 
@@ -2890,136 +3628,198 @@ static bool drm_dp_get_one_sb_msg(struct drm_dp_mst_topology_mgr *mgr, bool up)
 
 static int drm_dp_mst_handle_down_rep(struct drm_dp_mst_topology_mgr *mgr)
 {
-	int ret = 0;
+	struct drm_dp_sideband_msg_tx *txmsg;
+	struct drm_dp_mst_branch *mstb;
+	struct drm_dp_sideband_msg_hdr *hdr = &mgr->down_rep_recv.initial_hdr;
+	int slot = -1;
+
+	if (!drm_dp_get_one_sb_msg(mgr, false))
+		goto clear_down_rep_recv;
 
-	if (!drm_dp_get_one_sb_msg(mgr, false)) {
-		memset(&mgr->down_rep_recv, 0,
-		       sizeof(struct drm_dp_sideband_msg_rx));
+	if (!mgr->down_rep_recv.have_eomt)
 		return 0;
+
+	mstb = drm_dp_get_mst_branch_device(mgr, hdr->lct, hdr->rad);
+	if (!mstb) {
+		DRM_DEBUG_KMS("Got MST reply from unknown device %d\n",
+			      hdr->lct);
+		goto clear_down_rep_recv;
+	}
+
+	/* find the message */
+	slot = hdr->seqno;
+	mutex_lock(&mgr->qlock);
+	txmsg = mstb->tx_slots[slot];
+	/* remove from slots */
+	mutex_unlock(&mgr->qlock);
+
+	if (!txmsg) {
+		DRM_DEBUG_KMS("Got MST reply with no msg %p %d %d %02x %02x\n",
+			      mstb, hdr->seqno, hdr->lct, hdr->rad[0],
+			      mgr->down_rep_recv.msg[0]);
+		goto no_msg;
 	}
 
-	if (mgr->down_rep_recv.have_eomt) {
-		struct drm_dp_sideband_msg_tx *txmsg;
-		struct drm_dp_mst_branch *mstb;
-		int slot = -1;
-		mstb = drm_dp_get_mst_branch_device(mgr,
-						    mgr->down_rep_recv.initial_hdr.lct,
-						    mgr->down_rep_recv.initial_hdr.rad);
+	drm_dp_sideband_parse_reply(&mgr->down_rep_recv, &txmsg->reply);
 
-		if (!mstb) {
-			DRM_DEBUG_KMS("Got MST reply from unknown device %d\n", mgr->down_rep_recv.initial_hdr.lct);
-			memset(&mgr->down_rep_recv, 0, sizeof(struct drm_dp_sideband_msg_rx));
-			return 0;
-		}
+	if (txmsg->reply.reply_type == DP_SIDEBAND_REPLY_NAK)
+		DRM_DEBUG_KMS("Got NAK reply: req 0x%02x (%s), reason 0x%02x (%s), nak data 0x%02x\n",
+			      txmsg->reply.req_type,
+			      drm_dp_mst_req_type_str(txmsg->reply.req_type),
+			      txmsg->reply.u.nak.reason,
+			      drm_dp_mst_nak_reason_str(txmsg->reply.u.nak.reason),
+			      txmsg->reply.u.nak.nak_data);
 
-		/* find the message */
-		slot = mgr->down_rep_recv.initial_hdr.seqno;
-		mutex_lock(&mgr->qlock);
-		txmsg = mstb->tx_slots[slot];
-		/* remove from slots */
-		mutex_unlock(&mgr->qlock);
-
-		if (!txmsg) {
-			DRM_DEBUG_KMS("Got MST reply with no msg %p %d %d %02x %02x\n",
-			       mstb,
-			       mgr->down_rep_recv.initial_hdr.seqno,
-			       mgr->down_rep_recv.initial_hdr.lct,
-				      mgr->down_rep_recv.initial_hdr.rad[0],
-				      mgr->down_rep_recv.msg[0]);
-			drm_dp_mst_topology_put_mstb(mstb);
-			memset(&mgr->down_rep_recv, 0, sizeof(struct drm_dp_sideband_msg_rx));
-			return 0;
-		}
+	memset(&mgr->down_rep_recv, 0, sizeof(struct drm_dp_sideband_msg_rx));
+	drm_dp_mst_topology_put_mstb(mstb);
 
-		drm_dp_sideband_parse_reply(&mgr->down_rep_recv, &txmsg->reply);
+	mutex_lock(&mgr->qlock);
+	txmsg->state = DRM_DP_SIDEBAND_TX_RX;
+	mstb->tx_slots[slot] = NULL;
+	mutex_unlock(&mgr->qlock);
 
-		if (txmsg->reply.reply_type == DP_SIDEBAND_REPLY_NAK)
-			DRM_DEBUG_KMS("Got NAK reply: req 0x%02x (%s), reason 0x%02x (%s), nak data 0x%02x\n",
-				      txmsg->reply.req_type,
-				      drm_dp_mst_req_type_str(txmsg->reply.req_type),
-				      txmsg->reply.u.nak.reason,
-				      drm_dp_mst_nak_reason_str(txmsg->reply.u.nak.reason),
-				      txmsg->reply.u.nak.nak_data);
-
-		memset(&mgr->down_rep_recv, 0, sizeof(struct drm_dp_sideband_msg_rx));
-		drm_dp_mst_topology_put_mstb(mstb);
+	wake_up_all(&mgr->tx_waitq);
+
+	return 0;
 
-		mutex_lock(&mgr->qlock);
-		txmsg->state = DRM_DP_SIDEBAND_TX_RX;
-		mstb->tx_slots[slot] = NULL;
-		mutex_unlock(&mgr->qlock);
+no_msg:
+	drm_dp_mst_topology_put_mstb(mstb);
+clear_down_rep_recv:
+	memset(&mgr->down_rep_recv, 0, sizeof(struct drm_dp_sideband_msg_rx));
 
-		wake_up_all(&mgr->tx_waitq);
-	}
-	return ret;
+	return 0;
 }
 
-static int drm_dp_mst_handle_up_req(struct drm_dp_mst_topology_mgr *mgr)
+static inline bool
+drm_dp_mst_process_up_req(struct drm_dp_mst_topology_mgr *mgr,
+			  struct drm_dp_pending_up_req *up_req)
 {
-	int ret = 0;
+	struct drm_dp_mst_branch *mstb = NULL;
+	struct drm_dp_sideband_msg_req_body *msg = &up_req->msg;
+	struct drm_dp_sideband_msg_hdr *hdr = &up_req->hdr;
+	bool hotplug = false;
 
-	if (!drm_dp_get_one_sb_msg(mgr, true)) {
-		memset(&mgr->up_req_recv, 0,
-		       sizeof(struct drm_dp_sideband_msg_rx));
-		return 0;
+	if (hdr->broadcast) {
+		const u8 *guid = NULL;
+
+		if (msg->req_type == DP_CONNECTION_STATUS_NOTIFY)
+			guid = msg->u.conn_stat.guid;
+		else if (msg->req_type == DP_RESOURCE_STATUS_NOTIFY)
+			guid = msg->u.resource_stat.guid;
+
+		mstb = drm_dp_get_mst_branch_device_by_guid(mgr, guid);
+	} else {
+		mstb = drm_dp_get_mst_branch_device(mgr, hdr->lct, hdr->rad);
 	}
 
-	if (mgr->up_req_recv.have_eomt) {
-		struct drm_dp_sideband_msg_req_body msg;
-		struct drm_dp_mst_branch *mstb = NULL;
-		bool seqno;
+	if (!mstb) {
+		DRM_DEBUG_KMS("Got MST reply from unknown device %d\n",
+			      hdr->lct);
+		return false;
+	}
 
-		if (!mgr->up_req_recv.initial_hdr.broadcast) {
-			mstb = drm_dp_get_mst_branch_device(mgr,
-							    mgr->up_req_recv.initial_hdr.lct,
-							    mgr->up_req_recv.initial_hdr.rad);
-			if (!mstb) {
-				DRM_DEBUG_KMS("Got MST reply from unknown device %d\n", mgr->up_req_recv.initial_hdr.lct);
-				memset(&mgr->up_req_recv, 0, sizeof(struct drm_dp_sideband_msg_rx));
-				return 0;
-			}
-		}
+	/* TODO: Add missing handler for DP_RESOURCE_STATUS_NOTIFY events */
+	if (msg->req_type == DP_CONNECTION_STATUS_NOTIFY) {
+		drm_dp_mst_handle_conn_stat(mstb, &msg->u.conn_stat);
+		hotplug = true;
+	}
 
-		seqno = mgr->up_req_recv.initial_hdr.seqno;
-		drm_dp_sideband_parse_req(&mgr->up_req_recv, &msg);
+	drm_dp_mst_topology_put_mstb(mstb);
+	return hotplug;
+}
 
-		if (msg.req_type == DP_CONNECTION_STATUS_NOTIFY) {
-			drm_dp_send_up_ack_reply(mgr, mgr->mst_primary, msg.req_type, seqno, false);
+static void drm_dp_mst_up_req_work(struct work_struct *work)
+{
+	struct drm_dp_mst_topology_mgr *mgr =
+		container_of(work, struct drm_dp_mst_topology_mgr,
+			     up_req_work);
+	struct drm_dp_pending_up_req *up_req;
+	bool send_hotplug = false;
 
-			if (!mstb)
-				mstb = drm_dp_get_mst_branch_device_by_guid(mgr, msg.u.conn_stat.guid);
+	mutex_lock(&mgr->probe_lock);
+	while (true) {
+		mutex_lock(&mgr->up_req_lock);
+		up_req = list_first_entry_or_null(&mgr->up_req_list,
+						  struct drm_dp_pending_up_req,
+						  next);
+		if (up_req)
+			list_del(&up_req->next);
+		mutex_unlock(&mgr->up_req_lock);
+
+		if (!up_req)
+			break;
 
-			if (!mstb) {
-				DRM_DEBUG_KMS("Got MST reply from unknown device %d\n", mgr->up_req_recv.initial_hdr.lct);
-				memset(&mgr->up_req_recv, 0, sizeof(struct drm_dp_sideband_msg_rx));
-				return 0;
-			}
+		send_hotplug |= drm_dp_mst_process_up_req(mgr, up_req);
+		kfree(up_req);
+	}
+	mutex_unlock(&mgr->probe_lock);
 
-			drm_dp_update_port(mstb, &msg.u.conn_stat);
+	if (send_hotplug)
+		drm_kms_helper_hotplug_event(mgr->dev);
+}
 
-			DRM_DEBUG_KMS("Got CSN: pn: %d ldps:%d ddps: %d mcs: %d ip: %d pdt: %d\n", msg.u.conn_stat.port_number, msg.u.conn_stat.legacy_device_plug_status, msg.u.conn_stat.displayport_device_plug_status, msg.u.conn_stat.message_capability_status, msg.u.conn_stat.input_port, msg.u.conn_stat.peer_device_type);
-			drm_kms_helper_hotplug_event(mgr->dev);
+static int drm_dp_mst_handle_up_req(struct drm_dp_mst_topology_mgr *mgr)
+{
+	struct drm_dp_sideband_msg_hdr *hdr = &mgr->up_req_recv.initial_hdr;
+	struct drm_dp_pending_up_req *up_req;
+	bool seqno;
 
-		} else if (msg.req_type == DP_RESOURCE_STATUS_NOTIFY) {
-			drm_dp_send_up_ack_reply(mgr, mgr->mst_primary, msg.req_type, seqno, false);
-			if (!mstb)
-				mstb = drm_dp_get_mst_branch_device_by_guid(mgr, msg.u.resource_stat.guid);
+	if (!drm_dp_get_one_sb_msg(mgr, true))
+		goto out;
 
-			if (!mstb) {
-				DRM_DEBUG_KMS("Got MST reply from unknown device %d\n", mgr->up_req_recv.initial_hdr.lct);
-				memset(&mgr->up_req_recv, 0, sizeof(struct drm_dp_sideband_msg_rx));
-				return 0;
-			}
+	if (!mgr->up_req_recv.have_eomt)
+		return 0;
 
-			DRM_DEBUG_KMS("Got RSN: pn: %d avail_pbn %d\n", msg.u.resource_stat.port_number, msg.u.resource_stat.available_pbn);
-		}
+	up_req = kzalloc(sizeof(*up_req), GFP_KERNEL);
+	if (!up_req) {
+		DRM_ERROR("Not enough memory to process MST up req\n");
+		return -ENOMEM;
+	}
+	INIT_LIST_HEAD(&up_req->next);
 
-		if (mstb)
-			drm_dp_mst_topology_put_mstb(mstb);
+	seqno = hdr->seqno;
+	drm_dp_sideband_parse_req(&mgr->up_req_recv, &up_req->msg);
 
-		memset(&mgr->up_req_recv, 0, sizeof(struct drm_dp_sideband_msg_rx));
+	if (up_req->msg.req_type != DP_CONNECTION_STATUS_NOTIFY &&
+	    up_req->msg.req_type != DP_RESOURCE_STATUS_NOTIFY) {
+		DRM_DEBUG_KMS("Received unknown up req type, ignoring: %x\n",
+			      up_req->msg.req_type);
+		kfree(up_req);
+		goto out;
 	}
-	return ret;
+
+	drm_dp_send_up_ack_reply(mgr, mgr->mst_primary, up_req->msg.req_type,
+				 seqno, false);
+
+	if (up_req->msg.req_type == DP_CONNECTION_STATUS_NOTIFY) {
+		const struct drm_dp_connection_status_notify *conn_stat =
+			&up_req->msg.u.conn_stat;
+
+		DRM_DEBUG_KMS("Got CSN: pn: %d ldps:%d ddps: %d mcs: %d ip: %d pdt: %d\n",
+			      conn_stat->port_number,
+			      conn_stat->legacy_device_plug_status,
+			      conn_stat->displayport_device_plug_status,
+			      conn_stat->message_capability_status,
+			      conn_stat->input_port,
+			      conn_stat->peer_device_type);
+	} else if (up_req->msg.req_type == DP_RESOURCE_STATUS_NOTIFY) {
+		const struct drm_dp_resource_status_notify *res_stat =
+			&up_req->msg.u.resource_stat;
+
+		DRM_DEBUG_KMS("Got RSN: pn: %d avail_pbn %d\n",
+			      res_stat->port_number,
+			      res_stat->available_pbn);
+	}
+
+	up_req->hdr = *hdr;
+	mutex_lock(&mgr->up_req_lock);
+	list_add_tail(&up_req->next, &mgr->up_req_list);
+	mutex_unlock(&mgr->up_req_lock);
+	queue_work(system_long_wq, &mgr->up_req_work);
+
+out:
+	memset(&mgr->up_req_recv, 0, sizeof(struct drm_dp_sideband_msg_rx));
+	return 0;
 }
 
 /**
@@ -3063,22 +3863,31 @@ EXPORT_SYMBOL(drm_dp_mst_hpd_irq);
 /**
  * drm_dp_mst_detect_port() - get connection status for an MST port
  * @connector: DRM connector for this port
+ * @ctx: The acquisition context to use for grabbing locks
  * @mgr: manager for this port
- * @port: unverified pointer to a port
+ * @port: pointer to a port
  *
- * This returns the current connection state for a port. It validates the
- * port pointer still exists so the caller doesn't require a reference
+ * This returns the current connection state for a port.
  */
-enum drm_connector_status drm_dp_mst_detect_port(struct drm_connector *connector,
-						 struct drm_dp_mst_topology_mgr *mgr, struct drm_dp_mst_port *port)
+int
+drm_dp_mst_detect_port(struct drm_connector *connector,
+		       struct drm_modeset_acquire_ctx *ctx,
+		       struct drm_dp_mst_topology_mgr *mgr,
+		       struct drm_dp_mst_port *port)
 {
-	enum drm_connector_status status = connector_status_disconnected;
+	int ret;
 
 	/* we need to search for the port in the mgr in case it's gone */
 	port = drm_dp_mst_topology_get_port_validated(mgr, port);
 	if (!port)
 		return connector_status_disconnected;
 
+	ret = drm_modeset_lock(&mgr->base.lock, ctx);
+	if (ret)
+		goto out;
+
+	ret = connector_status_disconnected;
+
 	if (!port->ddps)
 		goto out;
 
@@ -3088,7 +3897,7 @@ enum drm_connector_status drm_dp_mst_detect_port(struct drm_connector *connector
 		break;
 
 	case DP_PEER_DEVICE_SST_SINK:
-		status = connector_status_connected;
+		ret = connector_status_connected;
 		/* for logical ports - cache the EDID */
 		if (port->port_num >= 8 && !port->cached_edid) {
 			port->cached_edid = drm_get_edid(connector, &port->aux.ddc);
@@ -3096,12 +3905,12 @@ enum drm_connector_status drm_dp_mst_detect_port(struct drm_connector *connector
 		break;
 	case DP_PEER_DEVICE_DP_LEGACY_CONV:
 		if (port->ldps)
-			status = connector_status_connected;
+			ret = connector_status_connected;
 		break;
 	}
 out:
 	drm_dp_mst_topology_put_port(port);
-	return status;
+	return ret;
 }
 EXPORT_SYMBOL(drm_dp_mst_detect_port);
 
@@ -3237,7 +4046,7 @@ int drm_dp_atomic_find_vcpi_slots(struct drm_atomic_state *state,
 {
 	struct drm_dp_mst_topology_state *topology_state;
 	struct drm_dp_vcpi_allocation *pos, *vcpi = NULL;
-	int prev_slots, req_slots, ret;
+	int prev_slots, req_slots;
 
 	topology_state = drm_atomic_get_mst_topology_state(state, mgr);
 	if (IS_ERR(topology_state))
@@ -3284,8 +4093,7 @@ int drm_dp_atomic_find_vcpi_slots(struct drm_atomic_state *state,
 	}
 	vcpi->vcpi = req_slots;
 
-	ret = req_slots;
-	return ret;
+	return req_slots;
 }
 EXPORT_SYMBOL(drm_dp_atomic_find_vcpi_slots);
 
@@ -3539,13 +4347,6 @@ EXPORT_SYMBOL(drm_dp_check_act_status);
  */
 int drm_dp_calc_pbn_mode(int clock, int bpp)
 {
-	u64 kbps;
-	s64 peak_kbps;
-	u32 numerator;
-	u32 denominator;
-
-	kbps = clock * bpp;
-
 	/*
 	 * margin 5300ppm + 300ppm ~ 0.6% as per spec, factor is 1.006
 	 * The unit of 54/64Mbytes/sec is an arbitrary unit chosen based on
@@ -3556,41 +4357,11 @@ int drm_dp_calc_pbn_mode(int clock, int bpp)
 	 * peak_kbps *= (64/54)
 	 * peak_kbps *= 8    convert to bytes
 	 */
-
-	numerator = 64 * 1006;
-	denominator = 54 * 8 * 1000 * 1000;
-
-	kbps *= numerator;
-	peak_kbps = drm_fixp_from_fraction(kbps, denominator);
-
-	return drm_fixp2int_ceil(peak_kbps);
+	return DIV_ROUND_UP_ULL(mul_u32_u32(clock * bpp, 64 * 1006),
+				8 * 54 * 1000 * 1000);
 }
 EXPORT_SYMBOL(drm_dp_calc_pbn_mode);
 
-static int test_calc_pbn_mode(void)
-{
-	int ret;
-	ret = drm_dp_calc_pbn_mode(154000, 30);
-	if (ret != 689) {
-		DRM_ERROR("PBN calculation test failed - clock %d, bpp %d, expected PBN %d, actual PBN %d.\n",
-				154000, 30, 689, ret);
-		return -EINVAL;
-	}
-	ret = drm_dp_calc_pbn_mode(234000, 30);
-	if (ret != 1047) {
-		DRM_ERROR("PBN calculation test failed - clock %d, bpp %d, expected PBN %d, actual PBN %d.\n",
-				234000, 30, 1047, ret);
-		return -EINVAL;
-	}
-	ret = drm_dp_calc_pbn_mode(297000, 24);
-	if (ret != 1063) {
-		DRM_ERROR("PBN calculation test failed - clock %d, bpp %d, expected PBN %d, actual PBN %d.\n",
-				297000, 24, 1063, ret);
-		return -EINVAL;
-	}
-	return 0;
-}
-
 /* we want to kick the TX after we've ack the up/down IRQs. */
 static void drm_dp_mst_kick_tx(struct drm_dp_mst_topology_mgr *mgr)
 {
@@ -3729,36 +4500,103 @@ static void drm_dp_tx_work(struct work_struct *work)
 	mutex_unlock(&mgr->qlock);
 }
 
-static void drm_dp_destroy_connector_work(struct work_struct *work)
+static inline void
+drm_dp_delayed_destroy_port(struct drm_dp_mst_port *port)
 {
-	struct drm_dp_mst_topology_mgr *mgr = container_of(work, struct drm_dp_mst_topology_mgr, destroy_connector_work);
-	struct drm_dp_mst_port *port;
-	bool send_hotplug = false;
+	if (port->connector)
+		port->mgr->cbs->destroy_connector(port->mgr, port->connector);
+
+	drm_dp_port_set_pdt(port, DP_PEER_DEVICE_NONE);
+	drm_dp_mst_put_port_malloc(port);
+}
+
+static inline void
+drm_dp_delayed_destroy_mstb(struct drm_dp_mst_branch *mstb)
+{
+	struct drm_dp_mst_topology_mgr *mgr = mstb->mgr;
+	struct drm_dp_mst_port *port, *tmp;
+	bool wake_tx = false;
+
+	mutex_lock(&mgr->lock);
+	list_for_each_entry_safe(port, tmp, &mstb->ports, next) {
+		list_del(&port->next);
+		drm_dp_mst_topology_put_port(port);
+	}
+	mutex_unlock(&mgr->lock);
+
+	/* drop any tx slots msg */
+	mutex_lock(&mstb->mgr->qlock);
+	if (mstb->tx_slots[0]) {
+		mstb->tx_slots[0]->state = DRM_DP_SIDEBAND_TX_TIMEOUT;
+		mstb->tx_slots[0] = NULL;
+		wake_tx = true;
+	}
+	if (mstb->tx_slots[1]) {
+		mstb->tx_slots[1]->state = DRM_DP_SIDEBAND_TX_TIMEOUT;
+		mstb->tx_slots[1] = NULL;
+		wake_tx = true;
+	}
+	mutex_unlock(&mstb->mgr->qlock);
+
+	if (wake_tx)
+		wake_up_all(&mstb->mgr->tx_waitq);
+
+	drm_dp_mst_put_mstb_malloc(mstb);
+}
+
+static void drm_dp_delayed_destroy_work(struct work_struct *work)
+{
+	struct drm_dp_mst_topology_mgr *mgr =
+		container_of(work, struct drm_dp_mst_topology_mgr,
+			     delayed_destroy_work);
+	bool send_hotplug = false, go_again;
+
 	/*
 	 * Not a regular list traverse as we have to drop the destroy
-	 * connector lock before destroying the connector, to avoid AB->BA
+	 * connector lock before destroying the mstb/port, to avoid AB->BA
 	 * ordering between this lock and the config mutex.
 	 */
-	for (;;) {
-		mutex_lock(&mgr->destroy_connector_lock);
-		port = list_first_entry_or_null(&mgr->destroy_connector_list, struct drm_dp_mst_port, next);
-		if (!port) {
-			mutex_unlock(&mgr->destroy_connector_lock);
-			break;
+	do {
+		go_again = false;
+
+		for (;;) {
+			struct drm_dp_mst_branch *mstb;
+
+			mutex_lock(&mgr->delayed_destroy_lock);
+			mstb = list_first_entry_or_null(&mgr->destroy_branch_device_list,
+							struct drm_dp_mst_branch,
+							destroy_next);
+			if (mstb)
+				list_del(&mstb->destroy_next);
+			mutex_unlock(&mgr->delayed_destroy_lock);
+
+			if (!mstb)
+				break;
+
+			drm_dp_delayed_destroy_mstb(mstb);
+			go_again = true;
 		}
-		list_del(&port->next);
-		mutex_unlock(&mgr->destroy_connector_lock);
 
-		INIT_LIST_HEAD(&port->next);
+		for (;;) {
+			struct drm_dp_mst_port *port;
 
-		mgr->cbs->destroy_connector(mgr, port->connector);
+			mutex_lock(&mgr->delayed_destroy_lock);
+			port = list_first_entry_or_null(&mgr->destroy_port_list,
+							struct drm_dp_mst_port,
+							next);
+			if (port)
+				list_del(&port->next);
+			mutex_unlock(&mgr->delayed_destroy_lock);
 
-		drm_dp_port_teardown_pdt(port, port->pdt);
-		port->pdt = DP_PEER_DEVICE_NONE;
+			if (!port)
+				break;
+
+			drm_dp_delayed_destroy_port(port);
+			send_hotplug = true;
+			go_again = true;
+		}
+	} while (go_again);
 
-		drm_dp_mst_put_port_malloc(port);
-		send_hotplug = true;
-	}
 	if (send_hotplug)
 		drm_kms_helper_hotplug_event(mgr->dev);
 }
@@ -3920,9 +4758,6 @@ EXPORT_SYMBOL(drm_dp_mst_topology_state_funcs);
 struct drm_dp_mst_topology_state *drm_atomic_get_mst_topology_state(struct drm_atomic_state *state,
 								    struct drm_dp_mst_topology_mgr *mgr)
 {
-	struct drm_device *dev = mgr->dev;
-
-	WARN_ON(!drm_modeset_is_locked(&dev->mode_config.connection_mutex));
 	return to_dp_mst_topology_state(drm_atomic_get_private_obj_state(state, &mgr->base));
 }
 EXPORT_SYMBOL(drm_atomic_get_mst_topology_state);
@@ -3948,12 +4783,20 @@ int drm_dp_mst_topology_mgr_init(struct drm_dp_mst_topology_mgr *mgr,
 	mutex_init(&mgr->lock);
 	mutex_init(&mgr->qlock);
 	mutex_init(&mgr->payload_lock);
-	mutex_init(&mgr->destroy_connector_lock);
+	mutex_init(&mgr->delayed_destroy_lock);
+	mutex_init(&mgr->up_req_lock);
+	mutex_init(&mgr->probe_lock);
+#if IS_ENABLED(CONFIG_DRM_DEBUG_DP_MST_TOPOLOGY_REFS)
+	mutex_init(&mgr->topology_ref_history_lock);
+#endif
 	INIT_LIST_HEAD(&mgr->tx_msg_downq);
-	INIT_LIST_HEAD(&mgr->destroy_connector_list);
+	INIT_LIST_HEAD(&mgr->destroy_port_list);
+	INIT_LIST_HEAD(&mgr->destroy_branch_device_list);
+	INIT_LIST_HEAD(&mgr->up_req_list);
 	INIT_WORK(&mgr->work, drm_dp_mst_link_probe_work);
 	INIT_WORK(&mgr->tx_work, drm_dp_tx_work);
-	INIT_WORK(&mgr->destroy_connector_work, drm_dp_destroy_connector_work);
+	INIT_WORK(&mgr->delayed_destroy_work, drm_dp_delayed_destroy_work);
+	INIT_WORK(&mgr->up_req_work, drm_dp_mst_up_req_work);
 	init_waitqueue_head(&mgr->tx_waitq);
 	mgr->dev = dev;
 	mgr->aux = aux;
@@ -3970,8 +4813,6 @@ int drm_dp_mst_topology_mgr_init(struct drm_dp_mst_topology_mgr *mgr,
 	if (!mgr->proposed_vcpis)
 		return -ENOMEM;
 	set_bit(0, &mgr->payload_mask);
-	if (test_calc_pbn_mode() < 0)
-		DRM_ERROR("MST PBN self-test failed\n");
 
 	mst_state = kzalloc(sizeof(*mst_state), GFP_KERNEL);
 	if (mst_state == NULL)
@@ -3996,7 +4837,7 @@ void drm_dp_mst_topology_mgr_destroy(struct drm_dp_mst_topology_mgr *mgr)
 {
 	drm_dp_mst_topology_mgr_set_mst(mgr, false);
 	flush_work(&mgr->work);
-	flush_work(&mgr->destroy_connector_work);
+	cancel_work_sync(&mgr->delayed_destroy_work);
 	mutex_lock(&mgr->payload_lock);
 	kfree(mgr->payloads);
 	mgr->payloads = NULL;
@@ -4007,6 +4848,16 @@ void drm_dp_mst_topology_mgr_destroy(struct drm_dp_mst_topology_mgr *mgr)
 	mgr->aux = NULL;
 	drm_atomic_private_obj_fini(&mgr->base);
 	mgr->funcs = NULL;
+
+	mutex_destroy(&mgr->delayed_destroy_lock);
+	mutex_destroy(&mgr->payload_lock);
+	mutex_destroy(&mgr->qlock);
+	mutex_destroy(&mgr->lock);
+	mutex_destroy(&mgr->up_req_lock);
+	mutex_destroy(&mgr->probe_lock);
+#if IS_ENABLED(CONFIG_DRM_DEBUG_DP_MST_TOPOLOGY_REFS)
+	mutex_destroy(&mgr->topology_ref_history_lock);
+#endif
 }
 EXPORT_SYMBOL(drm_dp_mst_topology_mgr_destroy);
 
diff --git a/drivers/gpu/drm/drm_dp_mst_topology_internal.h b/drivers/gpu/drm/drm_dp_mst_topology_internal.h
new file mode 100644
index 000000000000..eeda9a61c657
--- /dev/null
+++ b/drivers/gpu/drm/drm_dp_mst_topology_internal.h
@@ -0,0 +1,24 @@
+/* SPDX-License-Identifier: GPL-2.0-only
+ *
+ * Declarations for DP MST related functions which are only used in selftests
+ *
+ * Copyright © 2018 Red Hat
+ * Authors:
+ *     Lyude Paul <lyude@redhat.com>
+ */
+
+#ifndef _DRM_DP_MST_HELPER_INTERNAL_H_
+#define _DRM_DP_MST_HELPER_INTERNAL_H_
+
+#include <drm/drm_dp_mst_helper.h>
+
+void
+drm_dp_encode_sideband_req(const struct drm_dp_sideband_msg_req_body *req,
+			   struct drm_dp_sideband_msg_tx *raw);
+int drm_dp_decode_sideband_req(const struct drm_dp_sideband_msg_tx *raw,
+			       struct drm_dp_sideband_msg_req_body *req);
+void
+drm_dp_dump_sideband_msg_req_body(const struct drm_dp_sideband_msg_req_body *req,
+				  int indent, struct drm_printer *printer);
+
+#endif /* !_DRM_DP_MST_HELPER_INTERNAL_H_ */
diff --git a/drivers/gpu/drm/drm_drv.c b/drivers/gpu/drm/drm_drv.c
index 769feefeeeef..1b9b40a1c7c9 100644
--- a/drivers/gpu/drm/drm_drv.c
+++ b/drivers/gpu/drm/drm_drv.c
@@ -46,26 +46,9 @@
 #include "drm_internal.h"
 #include "drm_legacy.h"
 
-/*
- * drm_debug: Enable debug output.
- * Bitmask of DRM_UT_x. See include/drm/drm_print.h for details.
- */
-unsigned int drm_debug = 0;
-EXPORT_SYMBOL(drm_debug);
-
 MODULE_AUTHOR("Gareth Hughes, Leif Delgass, José Fonseca, Jon Smirl");
 MODULE_DESCRIPTION("DRM shared core routines");
 MODULE_LICENSE("GPL and additional rights");
-MODULE_PARM_DESC(debug, "Enable debug output, where each bit enables a debug category.\n"
-"\t\tBit 0 (0x01)  will enable CORE messages (drm core code)\n"
-"\t\tBit 1 (0x02)  will enable DRIVER messages (drm controller code)\n"
-"\t\tBit 2 (0x04)  will enable KMS messages (modesetting code)\n"
-"\t\tBit 3 (0x08)  will enable PRIME messages (prime code)\n"
-"\t\tBit 4 (0x10)  will enable ATOMIC messages (atomic code)\n"
-"\t\tBit 5 (0x20)  will enable VBL messages (vblank code)\n"
-"\t\tBit 7 (0x80)  will enable LEASE messages (leasing code)\n"
-"\t\tBit 8 (0x100) will enable DP messages (displayport code)");
-module_param_named(debug, drm_debug, int, 0600);
 
 static DEFINE_SPINLOCK(drm_minor_lock);
 static struct idr drm_minors_idr;
diff --git a/drivers/gpu/drm/drm_dsc.c b/drivers/gpu/drm/drm_dsc.c
index 77f4e5ae4197..4a475d9696ff 100644
--- a/drivers/gpu/drm/drm_dsc.c
+++ b/drivers/gpu/drm/drm_dsc.c
@@ -216,13 +216,11 @@ void drm_dsc_pps_payload_pack(struct drm_dsc_picture_parameter_set *pps_payload,
 	 */
 	for (i = 0; i < DSC_NUM_BUF_RANGES; i++) {
 		pps_payload->rc_range_parameters[i] =
-			((dsc_cfg->rc_range_params[i].range_min_qp <<
-			  DSC_PPS_RC_RANGE_MINQP_SHIFT) |
-			 (dsc_cfg->rc_range_params[i].range_max_qp <<
-			  DSC_PPS_RC_RANGE_MAXQP_SHIFT) |
-			 (dsc_cfg->rc_range_params[i].range_bpg_offset));
-		pps_payload->rc_range_parameters[i] =
-			cpu_to_be16(pps_payload->rc_range_parameters[i]);
+			cpu_to_be16((dsc_cfg->rc_range_params[i].range_min_qp <<
+				     DSC_PPS_RC_RANGE_MINQP_SHIFT) |
+				    (dsc_cfg->rc_range_params[i].range_max_qp <<
+				     DSC_PPS_RC_RANGE_MAXQP_SHIFT) |
+				    (dsc_cfg->rc_range_params[i].range_bpg_offset));
 	}
 
 	/* PPS 88 */
@@ -336,12 +334,6 @@ int drm_dsc_compute_rc_parameters(struct drm_dsc_config *vdsc_cfg)
 	else
 		vdsc_cfg->nfl_bpg_offset = 0;
 
-	/* 2^16 - 1 */
-	if (vdsc_cfg->nfl_bpg_offset > 65535) {
-		DRM_DEBUG_KMS("NflBpgOffset is too large for this slice height\n");
-		return -ERANGE;
-	}
-
 	/* Number of groups used to code the entire slice */
 	groups_total = groups_per_line * vdsc_cfg->slice_height;
 
@@ -371,11 +363,6 @@ int drm_dsc_compute_rc_parameters(struct drm_dsc_config *vdsc_cfg)
 		vdsc_cfg->scale_increment_interval = 0;
 	}
 
-	if (vdsc_cfg->scale_increment_interval > 65535) {
-		DRM_DEBUG_KMS("ScaleIncrementInterval is large for slice height\n");
-		return -ERANGE;
-	}
-
 	/*
 	 * DSC spec mentions that bits_per_pixel specifies the target
 	 * bits/pixel (bpp) rate that is used by the encoder,
diff --git a/drivers/gpu/drm/drm_edid.c b/drivers/gpu/drm/drm_edid.c
index 82a4ceed3fcf..474ac04d5600 100644
--- a/drivers/gpu/drm/drm_edid.c
+++ b/drivers/gpu/drm/drm_edid.c
@@ -159,6 +159,9 @@ static const struct edid_quirk {
 	/* Medion MD 30217 PG */
 	{ "MED", 0x7b8, EDID_QUIRK_PREFER_LARGE_75 },
 
+	/* Lenovo G50 */
+	{ "SDC", 18514, EDID_QUIRK_FORCE_6BPC },
+
 	/* Panel in Samsung NP700G7A-S01PL notebook reports 6bpc */
 	{ "SEC", 0xd033, EDID_QUIRK_FORCE_8BPC },
 
@@ -1275,6 +1278,106 @@ static const struct drm_display_mode edid_cea_modes[] = {
 		   4104, 4400, 0, 2160, 2168, 2178, 2250, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
+	/* 108 - 1280x720@48Hz 16:9 */
+	{ DRM_MODE("1280x720", DRM_MODE_TYPE_DRIVER, 90000, 1280, 2240,
+		   2280, 2500, 0, 720, 725, 730, 750, 0,
+		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
+	  .vrefresh = 48, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
+	/* 109 - 1280x720@48Hz 64:27 */
+	{ DRM_MODE("1280x720", DRM_MODE_TYPE_DRIVER, 90000, 1280, 2240,
+		   2280, 2500, 0, 720, 725, 730, 750, 0,
+		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
+	  .vrefresh = 48, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
+	/* 110 - 1680x720@48Hz 64:27 */
+	{ DRM_MODE("1680x720", DRM_MODE_TYPE_DRIVER, 99000, 1680, 2490,
+		   2530, 2750, 0, 720, 725, 730, 750, 0,
+		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
+	  .vrefresh = 48, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
+	/* 111 - 1920x1080@48Hz 16:9 */
+	{ DRM_MODE("1920x1080", DRM_MODE_TYPE_DRIVER, 148500, 1920, 2558,
+		   2602, 2750, 0, 1080, 1084, 1089, 1125, 0,
+		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
+	  .vrefresh = 48, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
+	/* 112 - 1920x1080@48Hz 64:27 */
+	{ DRM_MODE("1920x1080", DRM_MODE_TYPE_DRIVER, 148500, 1920, 2558,
+		   2602, 2750, 0, 1080, 1084, 1089, 1125, 0,
+		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
+	  .vrefresh = 48, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
+	/* 113 - 2560x1080@48Hz 64:27 */
+	{ DRM_MODE("2560x1080", DRM_MODE_TYPE_DRIVER, 198000, 2560, 3558,
+		   3602, 3750, 0, 1080, 1084, 1089, 1100, 0,
+		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
+	  .vrefresh = 48, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
+	/* 114 - 3840x2160@48Hz 16:9 */
+	{ DRM_MODE("3840x2160", DRM_MODE_TYPE_DRIVER, 594000, 3840, 5116,
+		   5204, 5500, 0, 2160, 2168, 2178, 2250, 0,
+		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
+	  .vrefresh = 48, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
+	/* 115 - 4096x2160@48Hz 256:135 */
+	{ DRM_MODE("4096x2160", DRM_MODE_TYPE_DRIVER, 594000, 4096, 5116,
+		   5204, 5500, 0, 2160, 2168, 2178, 2250, 0,
+		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
+	  .vrefresh = 48, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_256_135, },
+	/* 116 - 3840x2160@48Hz 64:27 */
+	{ DRM_MODE("3840x2160", DRM_MODE_TYPE_DRIVER, 594000, 3840, 5116,
+		   5204, 5500, 0, 2160, 2168, 2178, 2250, 0,
+		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
+	  .vrefresh = 48, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
+	/* 117 - 3840x2160@100Hz 16:9 */
+	{ DRM_MODE("3840x2160", DRM_MODE_TYPE_DRIVER, 1188000, 3840, 4896,
+		   4984, 5280, 0, 2160, 2168, 2178, 2250, 0,
+		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
+	  .vrefresh = 100, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
+	/* 118 - 3840x2160@120Hz 16:9 */
+	{ DRM_MODE("3840x2160", DRM_MODE_TYPE_DRIVER, 1188000, 3840, 4016,
+		   4104, 4400, 0, 2160, 2168, 2178, 2250, 0,
+		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
+	  .vrefresh = 120, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
+	/* 119 - 3840x2160@100Hz 64:27 */
+	{ DRM_MODE("3840x2160", DRM_MODE_TYPE_DRIVER, 1188000, 3840, 4896,
+		   4984, 5280, 0, 2160, 2168, 2178, 2250, 0,
+		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
+	  .vrefresh = 100, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
+	/* 120 - 3840x2160@120Hz 64:27 */
+	{ DRM_MODE("3840x2160", DRM_MODE_TYPE_DRIVER, 1188000, 3840, 4016,
+		   4104, 4400, 0, 2160, 2168, 2178, 2250, 0,
+		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
+	  .vrefresh = 120, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
+	/* 121 - 5120x2160@24Hz 64:27 */
+	{ DRM_MODE("5120x2160", DRM_MODE_TYPE_DRIVER, 396000, 5120, 7116,
+		   7204, 7500, 0, 2160, 2168, 2178, 2200, 0,
+		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
+	  .vrefresh = 24, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
+	/* 122 - 5120x2160@25Hz 64:27 */
+	{ DRM_MODE("5120x2160", DRM_MODE_TYPE_DRIVER, 396000, 5120, 6816,
+		   6904, 7200, 0, 2160, 2168, 2178, 2200, 0,
+		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
+	  .vrefresh = 25, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
+	/* 123 - 5120x2160@30Hz 64:27 */
+	{ DRM_MODE("5120x2160", DRM_MODE_TYPE_DRIVER, 396000, 5120, 5784,
+		   5872, 6000, 0, 2160, 2168, 2178, 2200, 0,
+		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
+	  .vrefresh = 30, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
+	/* 124 - 5120x2160@48Hz 64:27 */
+	{ DRM_MODE("5120x2160", DRM_MODE_TYPE_DRIVER, 742500, 5120, 5866,
+		   5954, 6250, 0, 2160, 2168, 2178, 2475, 0,
+		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
+	  .vrefresh = 48, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
+	/* 125 - 5120x2160@50Hz 64:27 */
+	{ DRM_MODE("5120x2160", DRM_MODE_TYPE_DRIVER, 742500, 5120, 6216,
+		   6304, 6600, 0, 2160, 2168, 2178, 2250, 0,
+		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
+	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
+	/* 126 - 5120x2160@60Hz 64:27 */
+	{ DRM_MODE("5120x2160", DRM_MODE_TYPE_DRIVER, 742500, 5120, 5284,
+		   5372, 5500, 0, 2160, 2168, 2178, 2250, 0,
+		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
+	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
+	/* 127 - 5120x2160@100Hz 64:27 */
+	{ DRM_MODE("5120x2160", DRM_MODE_TYPE_DRIVER, 1485000, 5120, 6216,
+		   6304, 6600, 0, 2160, 2168, 2178, 2250, 0,
+		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
+	  .vrefresh = 100, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
 };
 
 /*
@@ -1551,7 +1654,7 @@ static void connector_bad_edid(struct drm_connector *connector,
 {
 	int i;
 
-	if (connector->bad_edid_counter++ && !(drm_debug & DRM_UT_KMS))
+	if (connector->bad_edid_counter++ && !drm_debug_enabled(DRM_UT_KMS))
 		return;
 
 	dev_warn(connector->dev->dev,
@@ -2089,7 +2192,8 @@ static int standard_timing_level(struct edid *edid)
 			return LEVEL_CVT;
 		if (drm_gtf2_hbreak(edid))
 			return LEVEL_GTF2;
-		return LEVEL_GTF;
+		if (edid->features & DRM_EDID_FEATURE_DEFAULT_GTF)
+			return LEVEL_GTF;
 	}
 	return LEVEL_DMT;
 }
@@ -3105,18 +3209,10 @@ static bool drm_valid_cea_vic(u8 vic)
 	return vic > 0 && vic < ARRAY_SIZE(edid_cea_modes);
 }
 
-/**
- * drm_get_cea_aspect_ratio - get the picture aspect ratio corresponding to
- * the input VIC from the CEA mode list
- * @video_code: ID given to each of the CEA modes
- *
- * Returns picture aspect ratio
- */
-enum hdmi_picture_aspect drm_get_cea_aspect_ratio(const u8 video_code)
+static enum hdmi_picture_aspect drm_get_cea_aspect_ratio(const u8 video_code)
 {
 	return edid_cea_modes[video_code].picture_aspect_ratio;
 }
-EXPORT_SYMBOL(drm_get_cea_aspect_ratio);
 
 /*
  * Calculate the alternate clock for HDMI modes (those from the HDMI vendor
@@ -3719,7 +3815,7 @@ cea_db_offsets(const u8 *cea, int *start, int *end)
 		if (*end < 4 || *end > 127)
 			return -ERANGE;
 	} else {
-		return -ENOTSUPP;
+		return -EOPNOTSUPP;
 	}
 
 	return 0;
@@ -4188,7 +4284,7 @@ int drm_edid_to_sad(struct edid *edid, struct cea_sad **sads)
 
 	if (cea_revision(cea) < 3) {
 		DRM_DEBUG_KMS("SAD: wrong CEA revision\n");
-		return -ENOTSUPP;
+		return -EOPNOTSUPP;
 	}
 
 	if (cea_db_offsets(cea, &start, &end)) {
@@ -4249,7 +4345,7 @@ int drm_edid_to_speaker_allocation(struct edid *edid, u8 **sadb)
 
 	if (cea_revision(cea) < 3) {
 		DRM_DEBUG_KMS("SAD: wrong CEA revision\n");
-		return -ENOTSUPP;
+		return -EOPNOTSUPP;
 	}
 
 	if (cea_db_offsets(cea, &start, &end)) {
@@ -5068,6 +5164,49 @@ drm_hdmi_infoframe_set_hdr_metadata(struct hdmi_drm_infoframe *frame,
 }
 EXPORT_SYMBOL(drm_hdmi_infoframe_set_hdr_metadata);
 
+static u8 drm_mode_hdmi_vic(struct drm_connector *connector,
+			    const struct drm_display_mode *mode)
+{
+	bool has_hdmi_infoframe = connector ?
+		connector->display_info.has_hdmi_infoframe : false;
+
+	if (!has_hdmi_infoframe)
+		return 0;
+
+	/* No HDMI VIC when signalling 3D video format */
+	if (mode->flags & DRM_MODE_FLAG_3D_MASK)
+		return 0;
+
+	return drm_match_hdmi_mode(mode);
+}
+
+static u8 drm_mode_cea_vic(struct drm_connector *connector,
+			   const struct drm_display_mode *mode)
+{
+	u8 vic;
+
+	/*
+	 * HDMI spec says if a mode is found in HDMI 1.4b 4K modes
+	 * we should send its VIC in vendor infoframes, else send the
+	 * VIC in AVI infoframes. Lets check if this mode is present in
+	 * HDMI 1.4b 4K modes
+	 */
+	if (drm_mode_hdmi_vic(connector, mode))
+		return 0;
+
+	vic = drm_match_cea_mode(mode);
+
+	/*
+	 * HDMI 1.4 VIC range: 1 <= VIC <= 64 (CEA-861-D) but
+	 * HDMI 2.0 VIC range: 1 <= VIC <= 107 (CEA-861-F). So we
+	 * have to make sure we dont break HDMI 1.4 sinks.
+	 */
+	if (!is_hdmi2_sink(connector) && vic > 64)
+		return 0;
+
+	return vic;
+}
+
 /**
  * drm_hdmi_avi_infoframe_from_display_mode() - fill an HDMI AVI infoframe with
  *                                              data from a DRM display mode
@@ -5095,29 +5234,7 @@ drm_hdmi_avi_infoframe_from_display_mode(struct hdmi_avi_infoframe *frame,
 	if (mode->flags & DRM_MODE_FLAG_DBLCLK)
 		frame->pixel_repeat = 1;
 
-	frame->video_code = drm_match_cea_mode(mode);
-
-	/*
-	 * HDMI 1.4 VIC range: 1 <= VIC <= 64 (CEA-861-D) but
-	 * HDMI 2.0 VIC range: 1 <= VIC <= 107 (CEA-861-F). So we
-	 * have to make sure we dont break HDMI 1.4 sinks.
-	 */
-	if (!is_hdmi2_sink(connector) && frame->video_code > 64)
-		frame->video_code = 0;
-
-	/*
-	 * HDMI spec says if a mode is found in HDMI 1.4b 4K modes
-	 * we should send its VIC in vendor infoframes, else send the
-	 * VIC in AVI infoframes. Lets check if this mode is present in
-	 * HDMI 1.4b 4K modes
-	 */
-	if (frame->video_code) {
-		u8 vendor_if_vic = drm_match_hdmi_mode(mode);
-		bool is_s3d = mode->flags & DRM_MODE_FLAG_3D_MASK;
-
-		if (drm_valid_hdmi_vic(vendor_if_vic) && !is_s3d)
-			frame->video_code = 0;
-	}
+	frame->video_code = drm_mode_cea_vic(connector, mode);
 
 	frame->picture_aspect = HDMI_PICTURE_ASPECT_NONE;
 
@@ -5282,6 +5399,23 @@ drm_hdmi_avi_infoframe_quant_range(struct hdmi_avi_infoframe *frame,
 }
 EXPORT_SYMBOL(drm_hdmi_avi_infoframe_quant_range);
 
+/**
+ * drm_hdmi_avi_infoframe_bars() - fill the HDMI AVI infoframe
+ *                                 bar information
+ * @frame: HDMI AVI infoframe
+ * @conn_state: connector state
+ */
+void
+drm_hdmi_avi_infoframe_bars(struct hdmi_avi_infoframe *frame,
+			    const struct drm_connector_state *conn_state)
+{
+	frame->right_bar = conn_state->tv.margins.right;
+	frame->left_bar = conn_state->tv.margins.left;
+	frame->top_bar = conn_state->tv.margins.top;
+	frame->bottom_bar = conn_state->tv.margins.bottom;
+}
+EXPORT_SYMBOL(drm_hdmi_avi_infoframe_bars);
+
 static enum hdmi_3d_structure
 s3d_structure_from_display_mode(const struct drm_display_mode *mode)
 {
@@ -5334,8 +5468,6 @@ drm_hdmi_vendor_infoframe_from_display_mode(struct hdmi_vendor_infoframe *frame,
 	bool has_hdmi_infoframe = connector ?
 		connector->display_info.has_hdmi_infoframe : false;
 	int err;
-	u32 s3d_flags;
-	u8 vic;
 
 	if (!frame || !mode)
 		return -EINVAL;
@@ -5343,8 +5475,9 @@ drm_hdmi_vendor_infoframe_from_display_mode(struct hdmi_vendor_infoframe *frame,
 	if (!has_hdmi_infoframe)
 		return -EINVAL;
 
-	vic = drm_match_hdmi_mode(mode);
-	s3d_flags = mode->flags & DRM_MODE_FLAG_3D_MASK;
+	err = hdmi_vendor_infoframe_init(frame);
+	if (err < 0)
+		return err;
 
 	/*
 	 * Even if it's not absolutely necessary to send the infoframe
@@ -5355,15 +5488,7 @@ drm_hdmi_vendor_infoframe_from_display_mode(struct hdmi_vendor_infoframe *frame,
 	 * mode if the source simply stops sending the infoframe when
 	 * it wants to switch from 3D to 2D.
 	 */
-
-	if (vic && s3d_flags)
-		return -EINVAL;
-
-	err = hdmi_vendor_infoframe_init(frame);
-	if (err < 0)
-		return err;
-
-	frame->vic = vic;
+	frame->vic = drm_mode_hdmi_vic(connector, mode);
 	frame->s3d_struct = s3d_structure_from_display_mode(mode);
 
 	return 0;
diff --git a/drivers/gpu/drm/drm_edid_load.c b/drivers/gpu/drm/drm_edid_load.c
index d38b3b255926..37d8ba3ddb46 100644
--- a/drivers/gpu/drm/drm_edid_load.c
+++ b/drivers/gpu/drm/drm_edid_load.c
@@ -175,7 +175,7 @@ static void *edid_load(struct drm_connector *connector, const char *name,
 	u8 *edid;
 	int fwsize, builtin;
 	int i, valid_extensions = 0;
-	bool print_bad_edid = !connector->bad_edid_counter || (drm_debug & DRM_UT_KMS);
+	bool print_bad_edid = !connector->bad_edid_counter || drm_debug_enabled(DRM_UT_KMS);
 
 	builtin = match_string(generic_edid_name, GENERIC_EDIDS, name);
 	if (builtin >= 0) {
diff --git a/drivers/gpu/drm/drm_encoder.c b/drivers/gpu/drm/drm_encoder.c
index 7fb47b7b8b44..80d88a55302e 100644
--- a/drivers/gpu/drm/drm_encoder.c
+++ b/drivers/gpu/drm/drm_encoder.c
@@ -22,6 +22,7 @@
 
 #include <linux/export.h>
 
+#include <drm/drm_bridge.h>
 #include <drm/drm_device.h>
 #include <drm/drm_drv.h>
 #include <drm/drm_encoder.h>
diff --git a/drivers/gpu/drm/drm_fb_helper.c b/drivers/gpu/drm/drm_fb_helper.c
index a7ba5b4902d6..8ebeccdeed23 100644
--- a/drivers/gpu/drm/drm_fb_helper.c
+++ b/drivers/gpu/drm/drm_fb_helper.c
@@ -46,6 +46,7 @@
 #include <drm/drm_print.h>
 #include <drm/drm_vblank.h>
 
+#include "drm_crtc_helper_internal.h"
 #include "drm_internal.h"
 
 static bool drm_fbdev_emulation = true;
@@ -91,9 +92,12 @@ static DEFINE_MUTEX(kernel_fb_helper_lock);
  *
  * Drivers that support a dumb buffer with a virtual address and mmap support,
  * should try out the generic fbdev emulation using drm_fbdev_generic_setup().
+ * It will automatically set up deferred I/O if the driver requires a shadow
+ * buffer.
  *
- * Setup fbdev emulation by calling drm_fb_helper_fbdev_setup() and tear it
- * down by calling drm_fb_helper_fbdev_teardown().
+ * For other drivers, setup fbdev emulation by calling
+ * drm_fb_helper_fbdev_setup() and tear it down by calling
+ * drm_fb_helper_fbdev_teardown().
  *
  * At runtime drivers should restore the fbdev console by using
  * drm_fb_helper_lastclose() as their &drm_driver.lastclose callback.
@@ -126,8 +130,10 @@ static DEFINE_MUTEX(kernel_fb_helper_lock);
  * always run in process context since the fb_*() function could be running in
  * atomic context. If drm_fb_helper_deferred_io() is used as the deferred_io
  * callback it will also schedule dirty_work with the damage collected from the
- * mmap page writes. Drivers can use drm_fb_helper_defio_init() to setup
- * deferred I/O (coupled with drm_fb_helper_fbdev_teardown()).
+ * mmap page writes.
+ *
+ * Deferred I/O is not compatible with SHMEM. Such drivers should request an
+ * fbdev shadow buffer and call drm_fbdev_generic_setup() instead.
  */
 
 static void drm_fb_helper_restore_lut_atomic(struct drm_crtc *crtc)
@@ -679,49 +685,6 @@ void drm_fb_helper_deferred_io(struct fb_info *info,
 EXPORT_SYMBOL(drm_fb_helper_deferred_io);
 
 /**
- * drm_fb_helper_defio_init - fbdev deferred I/O initialization
- * @fb_helper: driver-allocated fbdev helper
- *
- * This function allocates &fb_deferred_io, sets callback to
- * drm_fb_helper_deferred_io(), delay to 50ms and calls fb_deferred_io_init().
- * It should be called from the &drm_fb_helper_funcs->fb_probe callback.
- * drm_fb_helper_fbdev_teardown() cleans up deferred I/O.
- *
- * NOTE: A copy of &fb_ops is made and assigned to &info->fbops. This is done
- * because fb_deferred_io_cleanup() clears &fbops->fb_mmap and would thereby
- * affect other instances of that &fb_ops.
- *
- * Returns:
- * 0 on success or a negative error code on failure.
- */
-int drm_fb_helper_defio_init(struct drm_fb_helper *fb_helper)
-{
-	struct fb_info *info = fb_helper->fbdev;
-	struct fb_deferred_io *fbdefio;
-	struct fb_ops *fbops;
-
-	fbdefio = kzalloc(sizeof(*fbdefio), GFP_KERNEL);
-	fbops = kzalloc(sizeof(*fbops), GFP_KERNEL);
-	if (!fbdefio || !fbops) {
-		kfree(fbdefio);
-		kfree(fbops);
-		return -ENOMEM;
-	}
-
-	info->fbdefio = fbdefio;
-	fbdefio->delay = msecs_to_jiffies(50);
-	fbdefio->deferred_io = drm_fb_helper_deferred_io;
-
-	*fbops = *info->fbops;
-	info->fbops = fbops;
-
-	fb_deferred_io_init(info);
-
-	return 0;
-}
-EXPORT_SYMBOL(drm_fb_helper_defio_init);
-
-/**
  * drm_fb_helper_sys_read - wrapper around fb_sys_read
  * @info: fb_info struct pointer
  * @buf: userspace buffer to read from framebuffer memory
@@ -2355,7 +2318,10 @@ static const struct drm_client_funcs drm_fbdev_client_funcs = {
  *
  * Drivers that set the dirty callback on their framebuffer will get a shadow
  * fbdev buffer that is blitted onto the real buffer. This is done in order to
- * make deferred I/O work with all kinds of buffers.
+ * make deferred I/O work with all kinds of buffers. A shadow buffer can be
+ * requested explicitly by setting struct drm_mode_config.prefer_shadow or
+ * struct drm_mode_config.prefer_shadow_fbdev to true beforehand. This is
+ * required to use generic fbdev emulation with SHMEM helpers.
  *
  * This function is safe to call even when there are no connectors present.
  * Setup will be retried on the next hotplug event.
diff --git a/drivers/gpu/drm/drm_gem.c b/drivers/gpu/drm/drm_gem.c
index 6854f5867d51..000fa4a1899f 100644
--- a/drivers/gpu/drm/drm_gem.c
+++ b/drivers/gpu/drm/drm_gem.c
@@ -1099,23 +1099,12 @@ int drm_gem_mmap_obj(struct drm_gem_object *obj, unsigned long obj_size,
 		     struct vm_area_struct *vma)
 {
 	struct drm_device *dev = obj->dev;
+	int ret;
 
 	/* Check for valid size. */
 	if (obj_size < vma->vm_end - vma->vm_start)
 		return -EINVAL;
 
-	if (obj->funcs && obj->funcs->vm_ops)
-		vma->vm_ops = obj->funcs->vm_ops;
-	else if (dev->driver->gem_vm_ops)
-		vma->vm_ops = dev->driver->gem_vm_ops;
-	else
-		return -EINVAL;
-
-	vma->vm_flags |= VM_IO | VM_PFNMAP | VM_DONTEXPAND | VM_DONTDUMP;
-	vma->vm_private_data = obj;
-	vma->vm_page_prot = pgprot_writecombine(vm_get_page_prot(vma->vm_flags));
-	vma->vm_page_prot = pgprot_decrypted(vma->vm_page_prot);
-
 	/* Take a ref for this mapping of the object, so that the fault
 	 * handler can dereference the mmap offset's pointer to the object.
 	 * This reference is cleaned up by the corresponding vm_close
@@ -1124,6 +1113,33 @@ int drm_gem_mmap_obj(struct drm_gem_object *obj, unsigned long obj_size,
 	 */
 	drm_gem_object_get(obj);
 
+	if (obj->funcs && obj->funcs->mmap) {
+		/* Remove the fake offset */
+		vma->vm_pgoff -= drm_vma_node_start(&obj->vma_node);
+
+		ret = obj->funcs->mmap(obj, vma);
+		if (ret) {
+			drm_gem_object_put_unlocked(obj);
+			return ret;
+		}
+		WARN_ON(!(vma->vm_flags & VM_DONTEXPAND));
+	} else {
+		if (obj->funcs && obj->funcs->vm_ops)
+			vma->vm_ops = obj->funcs->vm_ops;
+		else if (dev->driver->gem_vm_ops)
+			vma->vm_ops = dev->driver->gem_vm_ops;
+		else {
+			drm_gem_object_put_unlocked(obj);
+			return -EINVAL;
+		}
+
+		vma->vm_flags |= VM_IO | VM_PFNMAP | VM_DONTEXPAND | VM_DONTDUMP;
+		vma->vm_page_prot = pgprot_writecombine(vm_get_page_prot(vma->vm_flags));
+		vma->vm_page_prot = pgprot_decrypted(vma->vm_page_prot);
+	}
+
+	vma->vm_private_data = obj;
+
 	return 0;
 }
 EXPORT_SYMBOL(drm_gem_mmap_obj);
diff --git a/drivers/gpu/drm/drm_gem_shmem_helper.c b/drivers/gpu/drm/drm_gem_shmem_helper.c
index f5918707672f..0810d3ef6961 100644
--- a/drivers/gpu/drm/drm_gem_shmem_helper.c
+++ b/drivers/gpu/drm/drm_gem_shmem_helper.c
@@ -32,7 +32,7 @@ static const struct drm_gem_object_funcs drm_gem_shmem_funcs = {
 	.get_sg_table = drm_gem_shmem_get_sg_table,
 	.vmap = drm_gem_shmem_vmap,
 	.vunmap = drm_gem_shmem_vunmap,
-	.vm_ops = &drm_gem_shmem_vm_ops,
+	.mmap = drm_gem_shmem_mmap,
 };
 
 /**
@@ -505,39 +505,30 @@ static void drm_gem_shmem_vm_close(struct vm_area_struct *vma)
 	drm_gem_vm_close(vma);
 }
 
-const struct vm_operations_struct drm_gem_shmem_vm_ops = {
+static const struct vm_operations_struct drm_gem_shmem_vm_ops = {
 	.fault = drm_gem_shmem_fault,
 	.open = drm_gem_shmem_vm_open,
 	.close = drm_gem_shmem_vm_close,
 };
-EXPORT_SYMBOL_GPL(drm_gem_shmem_vm_ops);
 
 /**
  * drm_gem_shmem_mmap - Memory-map a shmem GEM object
- * @filp: File object
+ * @obj: gem object
  * @vma: VMA for the area to be mapped
  *
  * This function implements an augmented version of the GEM DRM file mmap
  * operation for shmem objects. Drivers which employ the shmem helpers should
- * use this function as their &file_operations.mmap handler in the DRM device file's
- * file_operations structure.
- *
- * Instead of directly referencing this function, drivers should use the
- * DEFINE_DRM_GEM_SHMEM_FOPS() macro.
+ * use this function as their &drm_gem_object_funcs.mmap handler.
  *
  * Returns:
  * 0 on success or a negative error code on failure.
  */
-int drm_gem_shmem_mmap(struct file *filp, struct vm_area_struct *vma)
+int drm_gem_shmem_mmap(struct drm_gem_object *obj, struct vm_area_struct *vma)
 {
 	struct drm_gem_shmem_object *shmem;
 	int ret;
 
-	ret = drm_gem_mmap(filp, vma);
-	if (ret)
-		return ret;
-
-	shmem = to_drm_gem_shmem_obj(vma->vm_private_data);
+	shmem = to_drm_gem_shmem_obj(obj);
 
 	ret = drm_gem_shmem_get_pages(shmem);
 	if (ret) {
@@ -545,12 +536,10 @@ int drm_gem_shmem_mmap(struct file *filp, struct vm_area_struct *vma)
 		return ret;
 	}
 
-	/* VM_PFNMAP was set by drm_gem_mmap() */
-	vma->vm_flags &= ~VM_PFNMAP;
-	vma->vm_flags |= VM_MIXEDMAP;
-
-	/* Remove the fake offset */
-	vma->vm_pgoff -= drm_vma_node_start(&shmem->base.vma_node);
+	vma->vm_flags |= VM_MIXEDMAP | VM_DONTEXPAND;
+	vma->vm_page_prot = pgprot_writecombine(vm_get_page_prot(vma->vm_flags));
+	vma->vm_page_prot = pgprot_decrypted(vma->vm_page_prot);
+	vma->vm_ops = &drm_gem_shmem_vm_ops;
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/drm_gem_ttm_helper.c b/drivers/gpu/drm/drm_gem_ttm_helper.c
new file mode 100644
index 000000000000..605a8a3da7f9
--- /dev/null
+++ b/drivers/gpu/drm/drm_gem_ttm_helper.c
@@ -0,0 +1,84 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+
+#include <linux/module.h>
+
+#include <drm/drm_gem_ttm_helper.h>
+
+/**
+ * DOC: overview
+ *
+ * This library provides helper functions for gem objects backed by
+ * ttm.
+ */
+
+/**
+ * drm_gem_ttm_print_info() - Print &ttm_buffer_object info for debugfs
+ * @p: DRM printer
+ * @indent: Tab indentation level
+ * @gem: GEM object
+ *
+ * This function can be used as &drm_gem_object_funcs.print_info
+ * callback.
+ */
+void drm_gem_ttm_print_info(struct drm_printer *p, unsigned int indent,
+			    const struct drm_gem_object *gem)
+{
+	static const char * const plname[] = {
+		[ TTM_PL_SYSTEM ] = "system",
+		[ TTM_PL_TT     ] = "tt",
+		[ TTM_PL_VRAM   ] = "vram",
+		[ TTM_PL_PRIV   ] = "priv",
+
+		[ 16 ]            = "cached",
+		[ 17 ]            = "uncached",
+		[ 18 ]            = "wc",
+		[ 19 ]            = "contig",
+
+		[ 21 ]            = "pinned", /* NO_EVICT */
+		[ 22 ]            = "topdown",
+	};
+	const struct ttm_buffer_object *bo = drm_gem_ttm_of_gem(gem);
+
+	drm_printf_indent(p, indent, "placement=");
+	drm_print_bits(p, bo->mem.placement, plname, ARRAY_SIZE(plname));
+	drm_printf(p, "\n");
+
+	if (bo->mem.bus.is_iomem) {
+		drm_printf_indent(p, indent, "bus.base=%lx\n",
+				  (unsigned long)bo->mem.bus.base);
+		drm_printf_indent(p, indent, "bus.offset=%lx\n",
+				  (unsigned long)bo->mem.bus.offset);
+	}
+}
+EXPORT_SYMBOL(drm_gem_ttm_print_info);
+
+/**
+ * drm_gem_ttm_mmap() - mmap &ttm_buffer_object
+ * @gem: GEM object.
+ * @vma: vm area.
+ *
+ * This function can be used as &drm_gem_object_funcs.mmap
+ * callback.
+ */
+int drm_gem_ttm_mmap(struct drm_gem_object *gem,
+		     struct vm_area_struct *vma)
+{
+	struct ttm_buffer_object *bo = drm_gem_ttm_of_gem(gem);
+	int ret;
+
+	ret = ttm_bo_mmap_obj(vma, bo);
+	if (ret < 0)
+		return ret;
+
+	/*
+	 * ttm has its own object refcounting, so drop gem reference
+	 * to avoid double accounting counting.
+	 */
+	drm_gem_object_put_unlocked(gem);
+
+	return 0;
+}
+EXPORT_SYMBOL(drm_gem_ttm_mmap);
+
+MODULE_DESCRIPTION("DRM gem ttm helpers");
+MODULE_LICENSE("GPL");
diff --git a/drivers/gpu/drm/drm_gem_vram_helper.c b/drivers/gpu/drm/drm_gem_vram_helper.c
index fd751078bae1..666cb4c22bb9 100644
--- a/drivers/gpu/drm/drm_gem_vram_helper.c
+++ b/drivers/gpu/drm/drm_gem_vram_helper.c
@@ -1,10 +1,15 @@
 // SPDX-License-Identifier: GPL-2.0-or-later
 
-#include <drm/drm_gem_vram_helper.h>
+#include <drm/drm_debugfs.h>
 #include <drm/drm_device.h>
+#include <drm/drm_file.h>
+#include <drm/drm_framebuffer.h>
+#include <drm/drm_gem_ttm_helper.h>
+#include <drm/drm_gem_vram_helper.h>
 #include <drm/drm_mode.h>
+#include <drm/drm_plane.h>
 #include <drm/drm_prime.h>
-#include <drm/drm_vram_mm_helper.h>
+#include <drm/drm_simple_kms_helper.h>
 #include <drm/ttm/ttm_page_alloc.h>
 
 static const struct drm_gem_object_funcs drm_gem_vram_object_funcs;
@@ -14,6 +19,11 @@ static const struct drm_gem_object_funcs drm_gem_vram_object_funcs;
  *
  * This library provides a GEM buffer object that is backed by video RAM
  * (VRAM). It can be used for framebuffer devices with dedicated memory.
+ *
+ * The data structure &struct drm_vram_mm and its helpers implement a memory
+ * manager for simple framebuffer devices with dedicated video memory. Buffer
+ * objects are either placed in video RAM or evicted to system memory. The rsp.
+ * buffer object is provided by &struct drm_gem_vram_object.
  */
 
 /*
@@ -26,6 +36,10 @@ static void drm_gem_vram_cleanup(struct drm_gem_vram_object *gbo)
 	 * TTM buffer object in 'bo' has already been cleaned
 	 * up; only release the GEM object.
 	 */
+
+	WARN_ON(gbo->kmap_use_count);
+	WARN_ON(gbo->kmap.virtual);
+
 	drm_gem_object_release(&gbo->bo.base);
 }
 
@@ -47,6 +61,7 @@ static void drm_gem_vram_placement(struct drm_gem_vram_object *gbo,
 {
 	unsigned int i;
 	unsigned int c = 0;
+	u32 invariant_flags = pl_flag & TTM_PL_FLAG_TOPDOWN;
 
 	gbo->placement.placement = gbo->placements;
 	gbo->placement.busy_placement = gbo->placements;
@@ -54,15 +69,18 @@ static void drm_gem_vram_placement(struct drm_gem_vram_object *gbo,
 	if (pl_flag & TTM_PL_FLAG_VRAM)
 		gbo->placements[c++].flags = TTM_PL_FLAG_WC |
 					     TTM_PL_FLAG_UNCACHED |
-					     TTM_PL_FLAG_VRAM;
+					     TTM_PL_FLAG_VRAM |
+					     invariant_flags;
 
 	if (pl_flag & TTM_PL_FLAG_SYSTEM)
 		gbo->placements[c++].flags = TTM_PL_MASK_CACHING |
-					     TTM_PL_FLAG_SYSTEM;
+					     TTM_PL_FLAG_SYSTEM |
+					     invariant_flags;
 
 	if (!c)
 		gbo->placements[c++].flags = TTM_PL_MASK_CACHING |
-					     TTM_PL_FLAG_SYSTEM;
+					     TTM_PL_FLAG_SYSTEM |
+					     invariant_flags;
 
 	gbo->placement.num_placement = c;
 	gbo->placement.num_busy_placement = c;
@@ -82,8 +100,7 @@ static int drm_gem_vram_init(struct drm_device *dev,
 	int ret;
 	size_t acc_size;
 
-	if (!gbo->bo.base.funcs)
-		gbo->bo.base.funcs = &drm_gem_vram_object_funcs;
+	gbo->bo.base.funcs = &drm_gem_vram_object_funcs;
 
 	ret = drm_gem_object_init(dev, &gbo->bo.base, size);
 	if (ret)
@@ -192,30 +209,12 @@ s64 drm_gem_vram_offset(struct drm_gem_vram_object *gbo)
 }
 EXPORT_SYMBOL(drm_gem_vram_offset);
 
-/**
- * drm_gem_vram_pin() - Pins a GEM VRAM object in a region.
- * @gbo:	the GEM VRAM object
- * @pl_flag:	a bitmask of possible memory regions
- *
- * Pinning a buffer object ensures that it is not evicted from
- * a memory region. A pinned buffer object has to be unpinned before
- * it can be pinned to another region. If the pl_flag argument is 0,
- * the buffer is pinned at its current location (video RAM or system
- * memory).
- *
- * Returns:
- * 0 on success, or
- * a negative error code otherwise.
- */
-int drm_gem_vram_pin(struct drm_gem_vram_object *gbo, unsigned long pl_flag)
+static int drm_gem_vram_pin_locked(struct drm_gem_vram_object *gbo,
+				   unsigned long pl_flag)
 {
 	int i, ret;
 	struct ttm_operation_ctx ctx = { false, false };
 
-	ret = ttm_bo_reserve(&gbo->bo, true, false, NULL);
-	if (ret < 0)
-		return ret;
-
 	if (gbo->pin_count)
 		goto out;
 
@@ -227,62 +226,123 @@ int drm_gem_vram_pin(struct drm_gem_vram_object *gbo, unsigned long pl_flag)
 
 	ret = ttm_bo_validate(&gbo->bo, &gbo->placement, &ctx);
 	if (ret < 0)
-		goto err_ttm_bo_unreserve;
+		return ret;
 
 out:
 	++gbo->pin_count;
-	ttm_bo_unreserve(&gbo->bo);
 
 	return 0;
-
-err_ttm_bo_unreserve:
-	ttm_bo_unreserve(&gbo->bo);
-	return ret;
 }
-EXPORT_SYMBOL(drm_gem_vram_pin);
 
 /**
- * drm_gem_vram_unpin() - Unpins a GEM VRAM object
+ * drm_gem_vram_pin() - Pins a GEM VRAM object in a region.
  * @gbo:	the GEM VRAM object
+ * @pl_flag:	a bitmask of possible memory regions
+ *
+ * Pinning a buffer object ensures that it is not evicted from
+ * a memory region. A pinned buffer object has to be unpinned before
+ * it can be pinned to another region. If the pl_flag argument is 0,
+ * the buffer is pinned at its current location (video RAM or system
+ * memory).
+ *
+ * Small buffer objects, such as cursor images, can lead to memory
+ * fragmentation if they are pinned in the middle of video RAM. This
+ * is especially a problem on devices with only a small amount of
+ * video RAM. Fragmentation can prevent the primary framebuffer from
+ * fitting in, even though there's enough memory overall. The modifier
+ * DRM_GEM_VRAM_PL_FLAG_TOPDOWN marks the buffer object to be pinned
+ * at the high end of the memory region to avoid fragmentation.
  *
  * Returns:
  * 0 on success, or
  * a negative error code otherwise.
  */
-int drm_gem_vram_unpin(struct drm_gem_vram_object *gbo)
+int drm_gem_vram_pin(struct drm_gem_vram_object *gbo, unsigned long pl_flag)
 {
-	int i, ret;
-	struct ttm_operation_ctx ctx = { false, false };
+	int ret;
 
 	ret = ttm_bo_reserve(&gbo->bo, true, false, NULL);
-	if (ret < 0)
+	if (ret)
 		return ret;
+	ret = drm_gem_vram_pin_locked(gbo, pl_flag);
+	ttm_bo_unreserve(&gbo->bo);
+
+	return ret;
+}
+EXPORT_SYMBOL(drm_gem_vram_pin);
+
+static int drm_gem_vram_unpin_locked(struct drm_gem_vram_object *gbo)
+{
+	int i, ret;
+	struct ttm_operation_ctx ctx = { false, false };
 
 	if (WARN_ON_ONCE(!gbo->pin_count))
-		goto out;
+		return 0;
 
 	--gbo->pin_count;
 	if (gbo->pin_count)
-		goto out;
+		return 0;
 
 	for (i = 0; i < gbo->placement.num_placement ; ++i)
 		gbo->placements[i].flags &= ~TTM_PL_FLAG_NO_EVICT;
 
 	ret = ttm_bo_validate(&gbo->bo, &gbo->placement, &ctx);
 	if (ret < 0)
-		goto err_ttm_bo_unreserve;
-
-out:
-	ttm_bo_unreserve(&gbo->bo);
+		return ret;
 
 	return 0;
+}
 
-err_ttm_bo_unreserve:
+/**
+ * drm_gem_vram_unpin() - Unpins a GEM VRAM object
+ * @gbo:	the GEM VRAM object
+ *
+ * Returns:
+ * 0 on success, or
+ * a negative error code otherwise.
+ */
+int drm_gem_vram_unpin(struct drm_gem_vram_object *gbo)
+{
+	int ret;
+
+	ret = ttm_bo_reserve(&gbo->bo, true, false, NULL);
+	if (ret)
+		return ret;
+	ret = drm_gem_vram_unpin_locked(gbo);
 	ttm_bo_unreserve(&gbo->bo);
+
 	return ret;
 }
 EXPORT_SYMBOL(drm_gem_vram_unpin);
 
+static void *drm_gem_vram_kmap_locked(struct drm_gem_vram_object *gbo,
+				      bool map, bool *is_iomem)
+{
+	int ret;
+	struct ttm_bo_kmap_obj *kmap = &gbo->kmap;
+
+	if (gbo->kmap_use_count > 0)
+		goto out;
+
+	if (kmap->virtual || !map)
+		goto out;
+
+	ret = ttm_bo_kmap(&gbo->bo, 0, gbo->bo.num_pages, kmap);
+	if (ret)
+		return ERR_PTR(ret);
+
+out:
+	if (!kmap->virtual) {
+		if (is_iomem)
+			*is_iomem = false;
+		return NULL; /* not mapped; don't increment ref */
+	}
+	++gbo->kmap_use_count;
+	if (is_iomem)
+		return ttm_kmap_obj_virtual(kmap, is_iomem);
+	return kmap->virtual;
+}
+
 /**
  * drm_gem_vram_kmap() - Maps a GEM VRAM object into kernel address space
  * @gbo:	the GEM VRAM object
@@ -304,43 +364,121 @@ void *drm_gem_vram_kmap(struct drm_gem_vram_object *gbo, bool map,
 			bool *is_iomem)
 {
 	int ret;
-	struct ttm_bo_kmap_obj *kmap = &gbo->kmap;
-
-	if (kmap->virtual || !map)
-		goto out;
+	void *virtual;
 
-	ret = ttm_bo_kmap(&gbo->bo, 0, gbo->bo.num_pages, kmap);
+	ret = ttm_bo_reserve(&gbo->bo, true, false, NULL);
 	if (ret)
 		return ERR_PTR(ret);
+	virtual = drm_gem_vram_kmap_locked(gbo, map, is_iomem);
+	ttm_bo_unreserve(&gbo->bo);
 
-out:
-	if (!is_iomem)
-		return kmap->virtual;
-	if (!kmap->virtual) {
-		*is_iomem = false;
-		return NULL;
-	}
-	return ttm_kmap_obj_virtual(kmap, is_iomem);
+	return virtual;
 }
 EXPORT_SYMBOL(drm_gem_vram_kmap);
 
+static void drm_gem_vram_kunmap_locked(struct drm_gem_vram_object *gbo)
+{
+	if (WARN_ON_ONCE(!gbo->kmap_use_count))
+		return;
+	if (--gbo->kmap_use_count > 0)
+		return;
+
+	/*
+	 * Permanently mapping and unmapping buffers adds overhead from
+	 * updating the page tables and creates debugging output. Therefore,
+	 * we delay the actual unmap operation until the BO gets evicted
+	 * from memory. See drm_gem_vram_bo_driver_move_notify().
+	 */
+}
+
 /**
  * drm_gem_vram_kunmap() - Unmaps a GEM VRAM object
  * @gbo:	the GEM VRAM object
  */
 void drm_gem_vram_kunmap(struct drm_gem_vram_object *gbo)
 {
-	struct ttm_bo_kmap_obj *kmap = &gbo->kmap;
+	int ret;
 
-	if (!kmap->virtual)
+	ret = ttm_bo_reserve(&gbo->bo, false, false, NULL);
+	if (WARN_ONCE(ret, "ttm_bo_reserve_failed(): ret=%d\n", ret))
 		return;
-
-	ttm_bo_kunmap(kmap);
-	kmap->virtual = NULL;
+	drm_gem_vram_kunmap_locked(gbo);
+	ttm_bo_unreserve(&gbo->bo);
 }
 EXPORT_SYMBOL(drm_gem_vram_kunmap);
 
 /**
+ * drm_gem_vram_vmap() - Pins and maps a GEM VRAM object into kernel address
+ *                       space
+ * @gbo:	The GEM VRAM object to map
+ *
+ * The vmap function pins a GEM VRAM object to its current location, either
+ * system or video memory, and maps its buffer into kernel address space.
+ * As pinned object cannot be relocated, you should avoid pinning objects
+ * permanently. Call drm_gem_vram_vunmap() with the returned address to
+ * unmap and unpin the GEM VRAM object.
+ *
+ * If you have special requirements for the pinning or mapping operations,
+ * call drm_gem_vram_pin() and drm_gem_vram_kmap() directly.
+ *
+ * Returns:
+ * The buffer's virtual address on success, or
+ * an ERR_PTR()-encoded error code otherwise.
+ */
+void *drm_gem_vram_vmap(struct drm_gem_vram_object *gbo)
+{
+	int ret;
+	void *base;
+
+	ret = ttm_bo_reserve(&gbo->bo, true, false, NULL);
+	if (ret)
+		return ERR_PTR(ret);
+
+	ret = drm_gem_vram_pin_locked(gbo, 0);
+	if (ret)
+		goto err_ttm_bo_unreserve;
+	base = drm_gem_vram_kmap_locked(gbo, true, NULL);
+	if (IS_ERR(base)) {
+		ret = PTR_ERR(base);
+		goto err_drm_gem_vram_unpin_locked;
+	}
+
+	ttm_bo_unreserve(&gbo->bo);
+
+	return base;
+
+err_drm_gem_vram_unpin_locked:
+	drm_gem_vram_unpin_locked(gbo);
+err_ttm_bo_unreserve:
+	ttm_bo_unreserve(&gbo->bo);
+	return ERR_PTR(ret);
+}
+EXPORT_SYMBOL(drm_gem_vram_vmap);
+
+/**
+ * drm_gem_vram_vunmap() - Unmaps and unpins a GEM VRAM object
+ * @gbo:	The GEM VRAM object to unmap
+ * @vaddr:	The mapping's base address as returned by drm_gem_vram_vmap()
+ *
+ * A call to drm_gem_vram_vunmap() unmaps and unpins a GEM VRAM buffer. See
+ * the documentation for drm_gem_vram_vmap() for more information.
+ */
+void drm_gem_vram_vunmap(struct drm_gem_vram_object *gbo, void *vaddr)
+{
+	int ret;
+
+	ret = ttm_bo_reserve(&gbo->bo, false, false, NULL);
+	if (WARN_ONCE(ret, "ttm_bo_reserve_failed(): ret=%d\n", ret))
+		return;
+
+	drm_gem_vram_kunmap_locked(gbo);
+	drm_gem_vram_unpin_locked(gbo);
+
+	ttm_bo_unreserve(&gbo->bo);
+}
+EXPORT_SYMBOL(drm_gem_vram_vunmap);
+
+/**
  * drm_gem_vram_fill_create_dumb() - \
 	Helper for implementing &struct drm_driver.dumb_create
  * @file:		the DRM file
@@ -410,59 +548,27 @@ static bool drm_is_gem_vram(struct ttm_buffer_object *bo)
 	return (bo->destroy == ttm_buffer_object_destroy);
 }
 
-/**
- * drm_gem_vram_bo_driver_evict_flags() - \
-	Implements &struct ttm_bo_driver.evict_flags
- * @bo:	TTM buffer object. Refers to &struct drm_gem_vram_object.bo
- * @pl:	TTM placement information.
- */
-void drm_gem_vram_bo_driver_evict_flags(struct ttm_buffer_object *bo,
-					struct ttm_placement *pl)
+static void drm_gem_vram_bo_driver_evict_flags(struct drm_gem_vram_object *gbo,
+					       struct ttm_placement *pl)
 {
-	struct drm_gem_vram_object *gbo;
-
-	/* TTM may pass BOs that are not GEM VRAM BOs. */
-	if (!drm_is_gem_vram(bo))
-		return;
-
-	gbo = drm_gem_vram_of_bo(bo);
 	drm_gem_vram_placement(gbo, TTM_PL_FLAG_SYSTEM);
 	*pl = gbo->placement;
 }
-EXPORT_SYMBOL(drm_gem_vram_bo_driver_evict_flags);
 
-/**
- * drm_gem_vram_bo_driver_verify_access() - \
-	Implements &struct ttm_bo_driver.verify_access
- * @bo:		TTM buffer object. Refers to &struct drm_gem_vram_object.bo
- * @filp:	File pointer.
- *
- * Returns:
- * 0 on success, or
- * a negative errno code otherwise.
- */
-int drm_gem_vram_bo_driver_verify_access(struct ttm_buffer_object *bo,
-					 struct file *filp)
+static void drm_gem_vram_bo_driver_move_notify(struct drm_gem_vram_object *gbo,
+					       bool evict,
+					       struct ttm_mem_reg *new_mem)
 {
-	struct drm_gem_vram_object *gbo = drm_gem_vram_of_bo(bo);
+	struct ttm_bo_kmap_obj *kmap = &gbo->kmap;
 
-	return drm_vma_node_verify_access(&gbo->bo.base.vma_node,
-					  filp->private_data);
-}
-EXPORT_SYMBOL(drm_gem_vram_bo_driver_verify_access);
+	if (WARN_ON_ONCE(gbo->kmap_use_count))
+		return;
 
-/*
- * drm_gem_vram_mm_funcs - Functions for &struct drm_vram_mm
- *
- * Most users of @struct drm_gem_vram_object will also use
- * @struct drm_vram_mm. This instance of &struct drm_vram_mm_funcs
- * can be used to connect both.
- */
-const struct drm_vram_mm_funcs drm_gem_vram_mm_funcs = {
-	.evict_flags = drm_gem_vram_bo_driver_evict_flags,
-	.verify_access = drm_gem_vram_bo_driver_verify_access
-};
-EXPORT_SYMBOL(drm_gem_vram_mm_funcs);
+	if (!kmap->virtual)
+		return;
+	ttm_bo_kunmap(kmap);
+	kmap->virtual = NULL;
+}
 
 /*
  * Helpers for struct drm_gem_object_funcs
@@ -544,6 +650,129 @@ int drm_gem_vram_driver_dumb_mmap_offset(struct drm_file *file,
 EXPORT_SYMBOL(drm_gem_vram_driver_dumb_mmap_offset);
 
 /*
+ * Helpers for struct drm_plane_helper_funcs
+ */
+
+/**
+ * drm_gem_vram_plane_helper_prepare_fb() - \
+ *	Implements &struct drm_plane_helper_funcs.prepare_fb
+ * @plane:	a DRM plane
+ * @new_state:	the plane's new state
+ *
+ * During plane updates, this function pins the GEM VRAM
+ * objects of the plane's new framebuffer to VRAM. Call
+ * drm_gem_vram_plane_helper_cleanup_fb() to unpin them.
+ *
+ * Returns:
+ *	0 on success, or
+ *	a negative errno code otherwise.
+ */
+int
+drm_gem_vram_plane_helper_prepare_fb(struct drm_plane *plane,
+				     struct drm_plane_state *new_state)
+{
+	size_t i;
+	struct drm_gem_vram_object *gbo;
+	int ret;
+
+	if (!new_state->fb)
+		return 0;
+
+	for (i = 0; i < ARRAY_SIZE(new_state->fb->obj); ++i) {
+		if (!new_state->fb->obj[i])
+			continue;
+		gbo = drm_gem_vram_of_gem(new_state->fb->obj[i]);
+		ret = drm_gem_vram_pin(gbo, DRM_GEM_VRAM_PL_FLAG_VRAM);
+		if (ret)
+			goto err_drm_gem_vram_unpin;
+	}
+
+	return 0;
+
+err_drm_gem_vram_unpin:
+	while (i) {
+		--i;
+		gbo = drm_gem_vram_of_gem(new_state->fb->obj[i]);
+		drm_gem_vram_unpin(gbo);
+	}
+	return ret;
+}
+EXPORT_SYMBOL(drm_gem_vram_plane_helper_prepare_fb);
+
+/**
+ * drm_gem_vram_plane_helper_cleanup_fb() - \
+ *	Implements &struct drm_plane_helper_funcs.cleanup_fb
+ * @plane:	a DRM plane
+ * @old_state:	the plane's old state
+ *
+ * During plane updates, this function unpins the GEM VRAM
+ * objects of the plane's old framebuffer from VRAM. Complements
+ * drm_gem_vram_plane_helper_prepare_fb().
+ */
+void
+drm_gem_vram_plane_helper_cleanup_fb(struct drm_plane *plane,
+				     struct drm_plane_state *old_state)
+{
+	size_t i;
+	struct drm_gem_vram_object *gbo;
+
+	if (!old_state->fb)
+		return;
+
+	for (i = 0; i < ARRAY_SIZE(old_state->fb->obj); ++i) {
+		if (!old_state->fb->obj[i])
+			continue;
+		gbo = drm_gem_vram_of_gem(old_state->fb->obj[i]);
+		drm_gem_vram_unpin(gbo);
+	}
+}
+EXPORT_SYMBOL(drm_gem_vram_plane_helper_cleanup_fb);
+
+/*
+ * Helpers for struct drm_simple_display_pipe_funcs
+ */
+
+/**
+ * drm_gem_vram_simple_display_pipe_prepare_fb() - \
+ *	Implements &struct drm_simple_display_pipe_funcs.prepare_fb
+ * @pipe:	a simple display pipe
+ * @new_state:	the plane's new state
+ *
+ * During plane updates, this function pins the GEM VRAM
+ * objects of the plane's new framebuffer to VRAM. Call
+ * drm_gem_vram_simple_display_pipe_cleanup_fb() to unpin them.
+ *
+ * Returns:
+ *	0 on success, or
+ *	a negative errno code otherwise.
+ */
+int drm_gem_vram_simple_display_pipe_prepare_fb(
+	struct drm_simple_display_pipe *pipe,
+	struct drm_plane_state *new_state)
+{
+	return drm_gem_vram_plane_helper_prepare_fb(&pipe->plane, new_state);
+}
+EXPORT_SYMBOL(drm_gem_vram_simple_display_pipe_prepare_fb);
+
+/**
+ * drm_gem_vram_simple_display_pipe_cleanup_fb() - \
+ *	Implements &struct drm_simple_display_pipe_funcs.cleanup_fb
+ * @pipe:	a simple display pipe
+ * @old_state:	the plane's old state
+ *
+ * During plane updates, this function unpins the GEM VRAM
+ * objects of the plane's old framebuffer from VRAM. Complements
+ * drm_gem_vram_simple_display_pipe_prepare_fb().
+ */
+void drm_gem_vram_simple_display_pipe_cleanup_fb(
+	struct drm_simple_display_pipe *pipe,
+	struct drm_plane_state *old_state)
+{
+	drm_gem_vram_plane_helper_cleanup_fb(&pipe->plane, old_state);
+}
+EXPORT_SYMBOL(drm_gem_vram_simple_display_pipe_cleanup_fb);
+
+/*
  * PRIME helpers
  */
 
@@ -595,17 +824,11 @@ static void drm_gem_vram_object_unpin(struct drm_gem_object *gem)
 static void *drm_gem_vram_object_vmap(struct drm_gem_object *gem)
 {
 	struct drm_gem_vram_object *gbo = drm_gem_vram_of_gem(gem);
-	int ret;
 	void *base;
 
-	ret = drm_gem_vram_pin(gbo, 0);
-	if (ret)
-		return NULL;
-	base = drm_gem_vram_kmap(gbo, true, NULL);
-	if (IS_ERR(base)) {
-		drm_gem_vram_unpin(gbo);
+	base = drm_gem_vram_vmap(gbo);
+	if (IS_ERR(base))
 		return NULL;
-	}
 	return base;
 }
 
@@ -620,8 +843,7 @@ static void drm_gem_vram_object_vunmap(struct drm_gem_object *gem,
 {
 	struct drm_gem_vram_object *gbo = drm_gem_vram_of_gem(gem);
 
-	drm_gem_vram_kunmap(gbo);
-	drm_gem_vram_unpin(gbo);
+	drm_gem_vram_vunmap(gbo, vaddr);
 }
 
 /*
@@ -633,5 +855,278 @@ static const struct drm_gem_object_funcs drm_gem_vram_object_funcs = {
 	.pin	= drm_gem_vram_object_pin,
 	.unpin	= drm_gem_vram_object_unpin,
 	.vmap	= drm_gem_vram_object_vmap,
-	.vunmap	= drm_gem_vram_object_vunmap
+	.vunmap	= drm_gem_vram_object_vunmap,
+	.mmap   = drm_gem_ttm_mmap,
+	.print_info = drm_gem_ttm_print_info,
+};
+
+/*
+ * VRAM memory manager
+ */
+
+/*
+ * TTM TT
+ */
+
+static void backend_func_destroy(struct ttm_tt *tt)
+{
+	ttm_tt_fini(tt);
+	kfree(tt);
+}
+
+static struct ttm_backend_func backend_func = {
+	.destroy = backend_func_destroy
+};
+
+/*
+ * TTM BO device
+ */
+
+static struct ttm_tt *bo_driver_ttm_tt_create(struct ttm_buffer_object *bo,
+					      uint32_t page_flags)
+{
+	struct ttm_tt *tt;
+	int ret;
+
+	tt = kzalloc(sizeof(*tt), GFP_KERNEL);
+	if (!tt)
+		return NULL;
+
+	tt->func = &backend_func;
+
+	ret = ttm_tt_init(tt, bo, page_flags);
+	if (ret < 0)
+		goto err_ttm_tt_init;
+
+	return tt;
+
+err_ttm_tt_init:
+	kfree(tt);
+	return NULL;
+}
+
+static int bo_driver_init_mem_type(struct ttm_bo_device *bdev, uint32_t type,
+				   struct ttm_mem_type_manager *man)
+{
+	switch (type) {
+	case TTM_PL_SYSTEM:
+		man->flags = TTM_MEMTYPE_FLAG_MAPPABLE;
+		man->available_caching = TTM_PL_MASK_CACHING;
+		man->default_caching = TTM_PL_FLAG_CACHED;
+		break;
+	case TTM_PL_VRAM:
+		man->func = &ttm_bo_manager_func;
+		man->flags = TTM_MEMTYPE_FLAG_FIXED |
+			     TTM_MEMTYPE_FLAG_MAPPABLE;
+		man->available_caching = TTM_PL_FLAG_UNCACHED |
+					 TTM_PL_FLAG_WC;
+		man->default_caching = TTM_PL_FLAG_WC;
+		break;
+	default:
+		return -EINVAL;
+	}
+	return 0;
+}
+
+static void bo_driver_evict_flags(struct ttm_buffer_object *bo,
+				  struct ttm_placement *placement)
+{
+	struct drm_gem_vram_object *gbo;
+
+	/* TTM may pass BOs that are not GEM VRAM BOs. */
+	if (!drm_is_gem_vram(bo))
+		return;
+
+	gbo = drm_gem_vram_of_bo(bo);
+
+	drm_gem_vram_bo_driver_evict_flags(gbo, placement);
+}
+
+static void bo_driver_move_notify(struct ttm_buffer_object *bo,
+				  bool evict,
+				  struct ttm_mem_reg *new_mem)
+{
+	struct drm_gem_vram_object *gbo;
+
+	/* TTM may pass BOs that are not GEM VRAM BOs. */
+	if (!drm_is_gem_vram(bo))
+		return;
+
+	gbo = drm_gem_vram_of_bo(bo);
+
+	drm_gem_vram_bo_driver_move_notify(gbo, evict, new_mem);
+}
+
+static int bo_driver_io_mem_reserve(struct ttm_bo_device *bdev,
+				    struct ttm_mem_reg *mem)
+{
+	struct ttm_mem_type_manager *man = bdev->man + mem->mem_type;
+	struct drm_vram_mm *vmm = drm_vram_mm_of_bdev(bdev);
+
+	if (!(man->flags & TTM_MEMTYPE_FLAG_MAPPABLE))
+		return -EINVAL;
+
+	mem->bus.addr = NULL;
+	mem->bus.size = mem->num_pages << PAGE_SHIFT;
+
+	switch (mem->mem_type) {
+	case TTM_PL_SYSTEM:	/* nothing to do */
+		mem->bus.offset = 0;
+		mem->bus.base = 0;
+		mem->bus.is_iomem = false;
+		break;
+	case TTM_PL_VRAM:
+		mem->bus.offset = mem->start << PAGE_SHIFT;
+		mem->bus.base = vmm->vram_base;
+		mem->bus.is_iomem = true;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static void bo_driver_io_mem_free(struct ttm_bo_device *bdev,
+				  struct ttm_mem_reg *mem)
+{ }
+
+static struct ttm_bo_driver bo_driver = {
+	.ttm_tt_create = bo_driver_ttm_tt_create,
+	.ttm_tt_populate = ttm_pool_populate,
+	.ttm_tt_unpopulate = ttm_pool_unpopulate,
+	.init_mem_type = bo_driver_init_mem_type,
+	.eviction_valuable = ttm_bo_eviction_valuable,
+	.evict_flags = bo_driver_evict_flags,
+	.move_notify = bo_driver_move_notify,
+	.io_mem_reserve = bo_driver_io_mem_reserve,
+	.io_mem_free = bo_driver_io_mem_free,
+};
+
+/*
+ * struct drm_vram_mm
+ */
+
+#if defined(CONFIG_DEBUG_FS)
+static int drm_vram_mm_debugfs(struct seq_file *m, void *data)
+{
+	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_vram_mm *vmm = node->minor->dev->vram_mm;
+	struct drm_mm *mm = vmm->bdev.man[TTM_PL_VRAM].priv;
+	struct drm_printer p = drm_seq_file_printer(m);
+
+	spin_lock(&ttm_bo_glob.lru_lock);
+	drm_mm_print(mm, &p);
+	spin_unlock(&ttm_bo_glob.lru_lock);
+	return 0;
+}
+
+static const struct drm_info_list drm_vram_mm_debugfs_list[] = {
+	{ "vram-mm", drm_vram_mm_debugfs, 0, NULL },
 };
+#endif
+
+/**
+ * drm_vram_mm_debugfs_init() - Register VRAM MM debugfs file.
+ *
+ * @minor: drm minor device.
+ *
+ * Returns:
+ * 0 on success, or
+ * a negative error code otherwise.
+ */
+int drm_vram_mm_debugfs_init(struct drm_minor *minor)
+{
+	int ret = 0;
+
+#if defined(CONFIG_DEBUG_FS)
+	ret = drm_debugfs_create_files(drm_vram_mm_debugfs_list,
+				       ARRAY_SIZE(drm_vram_mm_debugfs_list),
+				       minor->debugfs_root, minor);
+#endif
+	return ret;
+}
+EXPORT_SYMBOL(drm_vram_mm_debugfs_init);
+
+static int drm_vram_mm_init(struct drm_vram_mm *vmm, struct drm_device *dev,
+			    uint64_t vram_base, size_t vram_size)
+{
+	int ret;
+
+	vmm->vram_base = vram_base;
+	vmm->vram_size = vram_size;
+
+	ret = ttm_bo_device_init(&vmm->bdev, &bo_driver,
+				 dev->anon_inode->i_mapping,
+				 dev->vma_offset_manager,
+				 true);
+	if (ret)
+		return ret;
+
+	ret = ttm_bo_init_mm(&vmm->bdev, TTM_PL_VRAM, vram_size >> PAGE_SHIFT);
+	if (ret)
+		return ret;
+
+	return 0;
+}
+
+static void drm_vram_mm_cleanup(struct drm_vram_mm *vmm)
+{
+	ttm_bo_device_release(&vmm->bdev);
+}
+
+/*
+ * Helpers for integration with struct drm_device
+ */
+
+/**
+ * drm_vram_helper_alloc_mm - Allocates a device's instance of \
+	&struct drm_vram_mm
+ * @dev:	the DRM device
+ * @vram_base:	the base address of the video memory
+ * @vram_size:	the size of the video memory in bytes
+ *
+ * Returns:
+ * The new instance of &struct drm_vram_mm on success, or
+ * an ERR_PTR()-encoded errno code otherwise.
+ */
+struct drm_vram_mm *drm_vram_helper_alloc_mm(
+	struct drm_device *dev, uint64_t vram_base, size_t vram_size)
+{
+	int ret;
+
+	if (WARN_ON(dev->vram_mm))
+		return dev->vram_mm;
+
+	dev->vram_mm = kzalloc(sizeof(*dev->vram_mm), GFP_KERNEL);
+	if (!dev->vram_mm)
+		return ERR_PTR(-ENOMEM);
+
+	ret = drm_vram_mm_init(dev->vram_mm, dev, vram_base, vram_size);
+	if (ret)
+		goto err_kfree;
+
+	return dev->vram_mm;
+
+err_kfree:
+	kfree(dev->vram_mm);
+	dev->vram_mm = NULL;
+	return ERR_PTR(ret);
+}
+EXPORT_SYMBOL(drm_vram_helper_alloc_mm);
+
+/**
+ * drm_vram_helper_release_mm - Releases a device's instance of \
+	&struct drm_vram_mm
+ * @dev:	the DRM device
+ */
+void drm_vram_helper_release_mm(struct drm_device *dev)
+{
+	if (!dev->vram_mm)
+		return;
+
+	drm_vram_mm_cleanup(dev->vram_mm);
+	kfree(dev->vram_mm);
+	dev->vram_mm = NULL;
+}
+EXPORT_SYMBOL(drm_vram_helper_release_mm);
diff --git a/drivers/gpu/drm/drm_memory.c b/drivers/gpu/drm/drm_memory.c
index 0bec6dbb0142..fbea69d6f909 100644
--- a/drivers/gpu/drm/drm_memory.c
+++ b/drivers/gpu/drm/drm_memory.c
@@ -40,6 +40,7 @@
 #include <xen/xen.h>
 
 #include <drm/drm_agpsupport.h>
+#include <drm/drm_cache.h>
 #include <drm/drm_device.h>
 
 #include "drm_legacy.h"
diff --git a/drivers/gpu/drm/drm_mipi_dbi.c b/drivers/gpu/drm/drm_mipi_dbi.c
index 1961f713aaab..e34058c721be 100644
--- a/drivers/gpu/drm/drm_mipi_dbi.c
+++ b/drivers/gpu/drm/drm_mipi_dbi.c
@@ -783,7 +783,7 @@ static int mipi_dbi_spi1e_transfer(struct mipi_dbi *dbi, int dc,
 	int i, ret;
 	u8 *dst;
 
-	if (drm_debug & DRM_UT_DRIVER)
+	if (drm_debug_enabled(DRM_UT_DRIVER))
 		pr_debug("[drm:%s] dc=%d, max_chunk=%zu, transfers:\n",
 			 __func__, dc, max_chunk);
 
@@ -907,7 +907,7 @@ static int mipi_dbi_spi1_transfer(struct mipi_dbi *dbi, int dc,
 	max_chunk = dbi->tx_buf9_len;
 	dst16 = dbi->tx_buf9;
 
-	if (drm_debug & DRM_UT_DRIVER)
+	if (drm_debug_enabled(DRM_UT_DRIVER))
 		pr_debug("[drm:%s] dc=%d, max_chunk=%zu, transfers:\n",
 			 __func__, dc, max_chunk);
 
@@ -955,7 +955,7 @@ static int mipi_dbi_typec1_command(struct mipi_dbi *dbi, u8 *cmd,
 	int ret;
 
 	if (mipi_dbi_command_is_read(dbi, *cmd))
-		return -ENOTSUPP;
+		return -EOPNOTSUPP;
 
 	MIPI_DBI_DEBUG_COMMAND(*cmd, parameters, num);
 
@@ -1021,7 +1021,7 @@ static int mipi_dbi_typec3_command_read(struct mipi_dbi *dbi, u8 *cmd,
 		unsigned int i;
 
 		for (i = 0; i < len; i++)
-			data[i] = (buf[i] << 1) | !!(buf[i + 1] & BIT(7));
+			data[i] = (buf[i] << 1) | (buf[i + 1] >> 7);
 	}
 
 	MIPI_DBI_DEBUG_COMMAND(*cmd, data, len);
@@ -1187,8 +1187,7 @@ static ssize_t mipi_dbi_debugfs_command_write(struct file *file,
 	struct mipi_dbi_dev *dbidev = m->private;
 	u8 val, cmd = 0, parameters[64];
 	char *buf, *pos, *token;
-	unsigned int i;
-	int ret, idx;
+	int i, ret, idx;
 
 	if (!drm_dev_enter(&dbidev->drm, &idx))
 		return -ENODEV;
diff --git a/drivers/gpu/drm/drm_mm.c b/drivers/gpu/drm/drm_mm.c
index 4581c5387372..2a6e34663146 100644
--- a/drivers/gpu/drm/drm_mm.c
+++ b/drivers/gpu/drm/drm_mm.c
@@ -174,7 +174,7 @@ static void drm_mm_interval_tree_add_node(struct drm_mm_node *hole_node,
 
 	node->__subtree_last = LAST(node);
 
-	if (hole_node->allocated) {
+	if (drm_mm_node_allocated(hole_node)) {
 		rb = &hole_node->rb;
 		while (rb) {
 			parent = rb_entry(rb, struct drm_mm_node, rb);
@@ -424,9 +424,9 @@ int drm_mm_reserve_node(struct drm_mm *mm, struct drm_mm_node *node)
 
 	node->mm = mm;
 
+	__set_bit(DRM_MM_NODE_ALLOCATED_BIT, &node->flags);
 	list_add(&node->node_list, &hole->node_list);
 	drm_mm_interval_tree_add_node(hole, node);
-	node->allocated = true;
 	node->hole_size = 0;
 
 	rm_hole(hole);
@@ -543,9 +543,9 @@ int drm_mm_insert_node_in_range(struct drm_mm * const mm,
 		node->color = color;
 		node->hole_size = 0;
 
+		__set_bit(DRM_MM_NODE_ALLOCATED_BIT, &node->flags);
 		list_add(&node->node_list, &hole->node_list);
 		drm_mm_interval_tree_add_node(hole, node);
-		node->allocated = true;
 
 		rm_hole(hole);
 		if (adj_start > hole_start)
@@ -561,6 +561,11 @@ int drm_mm_insert_node_in_range(struct drm_mm * const mm,
 }
 EXPORT_SYMBOL(drm_mm_insert_node_in_range);
 
+static inline bool drm_mm_node_scanned_block(const struct drm_mm_node *node)
+{
+	return test_bit(DRM_MM_NODE_SCANNED_BIT, &node->flags);
+}
+
 /**
  * drm_mm_remove_node - Remove a memory node from the allocator.
  * @node: drm_mm_node to remove
@@ -574,8 +579,8 @@ void drm_mm_remove_node(struct drm_mm_node *node)
 	struct drm_mm *mm = node->mm;
 	struct drm_mm_node *prev_node;
 
-	DRM_MM_BUG_ON(!node->allocated);
-	DRM_MM_BUG_ON(node->scanned_block);
+	DRM_MM_BUG_ON(!drm_mm_node_allocated(node));
+	DRM_MM_BUG_ON(drm_mm_node_scanned_block(node));
 
 	prev_node = list_prev_entry(node, node_list);
 
@@ -584,11 +589,12 @@ void drm_mm_remove_node(struct drm_mm_node *node)
 
 	drm_mm_interval_tree_remove(node, &mm->interval_tree);
 	list_del(&node->node_list);
-	node->allocated = false;
 
 	if (drm_mm_hole_follows(prev_node))
 		rm_hole(prev_node);
 	add_hole(prev_node);
+
+	clear_bit_unlock(DRM_MM_NODE_ALLOCATED_BIT, &node->flags);
 }
 EXPORT_SYMBOL(drm_mm_remove_node);
 
@@ -605,10 +611,11 @@ void drm_mm_replace_node(struct drm_mm_node *old, struct drm_mm_node *new)
 {
 	struct drm_mm *mm = old->mm;
 
-	DRM_MM_BUG_ON(!old->allocated);
+	DRM_MM_BUG_ON(!drm_mm_node_allocated(old));
 
 	*new = *old;
 
+	__set_bit(DRM_MM_NODE_ALLOCATED_BIT, &new->flags);
 	list_replace(&old->node_list, &new->node_list);
 	rb_replace_node_cached(&old->rb, &new->rb, &mm->interval_tree);
 
@@ -622,8 +629,7 @@ void drm_mm_replace_node(struct drm_mm_node *old, struct drm_mm_node *new)
 				&mm->holes_addr);
 	}
 
-	old->allocated = false;
-	new->allocated = true;
+	clear_bit_unlock(DRM_MM_NODE_ALLOCATED_BIT, &old->flags);
 }
 EXPORT_SYMBOL(drm_mm_replace_node);
 
@@ -731,9 +737,9 @@ bool drm_mm_scan_add_block(struct drm_mm_scan *scan,
 	u64 adj_start, adj_end;
 
 	DRM_MM_BUG_ON(node->mm != mm);
-	DRM_MM_BUG_ON(!node->allocated);
-	DRM_MM_BUG_ON(node->scanned_block);
-	node->scanned_block = true;
+	DRM_MM_BUG_ON(!drm_mm_node_allocated(node));
+	DRM_MM_BUG_ON(drm_mm_node_scanned_block(node));
+	__set_bit(DRM_MM_NODE_SCANNED_BIT, &node->flags);
 	mm->scan_active++;
 
 	/* Remove this block from the node_list so that we enlarge the hole
@@ -818,8 +824,8 @@ bool drm_mm_scan_remove_block(struct drm_mm_scan *scan,
 	struct drm_mm_node *prev_node;
 
 	DRM_MM_BUG_ON(node->mm != scan->mm);
-	DRM_MM_BUG_ON(!node->scanned_block);
-	node->scanned_block = false;
+	DRM_MM_BUG_ON(!drm_mm_node_scanned_block(node));
+	__clear_bit(DRM_MM_NODE_SCANNED_BIT, &node->flags);
 
 	DRM_MM_BUG_ON(!node->mm->scan_active);
 	node->mm->scan_active--;
@@ -917,7 +923,7 @@ void drm_mm_init(struct drm_mm *mm, u64 start, u64 size)
 
 	/* Clever trick to avoid a special case in the free hole tracking. */
 	INIT_LIST_HEAD(&mm->head_node.node_list);
-	mm->head_node.allocated = false;
+	mm->head_node.flags = 0;
 	mm->head_node.mm = mm;
 	mm->head_node.start = start + size;
 	mm->head_node.size = -size;
diff --git a/drivers/gpu/drm/drm_mode_config.c b/drivers/gpu/drm/drm_mode_config.c
index 7bc03c3c154f..3b570a404933 100644
--- a/drivers/gpu/drm/drm_mode_config.c
+++ b/drivers/gpu/drm/drm_mode_config.c
@@ -428,8 +428,6 @@ EXPORT_SYMBOL(drm_mode_config_init);
  * Note that since this /should/ happen single-threaded at driver/device
  * teardown time, no locking is required. It's the driver's job to ensure that
  * this guarantee actually holds true.
- *
- * FIXME: cleanup any dangling user buffer objects too
  */
 void drm_mode_config_cleanup(struct drm_device *dev)
 {
diff --git a/drivers/gpu/drm/drm_of.c b/drivers/gpu/drm/drm_of.c
index 43d89dd59c6b..0ca58803ba46 100644
--- a/drivers/gpu/drm/drm_of.c
+++ b/drivers/gpu/drm/drm_of.c
@@ -250,11 +250,6 @@ int drm_of_find_panel_or_bridge(const struct device_node *np,
 	if (!remote)
 		return -ENODEV;
 
-	if (!of_device_is_available(remote)) {
-		of_node_put(remote);
-		return -ENODEV;
-	}
-
 	if (panel) {
 		*panel = of_drm_find_panel(remote);
 		if (!IS_ERR(*panel))
diff --git a/drivers/gpu/drm/drm_panel.c b/drivers/gpu/drm/drm_panel.c
index 6b0bf42039cf..ed7985c0535a 100644
--- a/drivers/gpu/drm/drm_panel.c
+++ b/drivers/gpu/drm/drm_panel.c
@@ -44,13 +44,21 @@ static LIST_HEAD(panel_list);
 /**
  * drm_panel_init - initialize a panel
  * @panel: DRM panel
+ * @dev: parent device of the panel
+ * @funcs: panel operations
+ * @connector_type: the connector type (DRM_MODE_CONNECTOR_*) corresponding to
+ *	the panel interface
  *
- * Sets up internal fields of the panel so that it can subsequently be added
- * to the registry.
+ * Initialize the panel structure for subsequent registration with
+ * drm_panel_add().
  */
-void drm_panel_init(struct drm_panel *panel)
+void drm_panel_init(struct drm_panel *panel, struct device *dev,
+		    const struct drm_panel_funcs *funcs, int connector_type)
 {
 	INIT_LIST_HEAD(&panel->list);
+	panel->dev = dev;
+	panel->funcs = funcs;
+	panel->connector_type = connector_type;
 }
 EXPORT_SYMBOL(drm_panel_init);
 
diff --git a/drivers/gpu/drm/drm_prime.c b/drivers/gpu/drm/drm_prime.c
index 0a2316e0e812..0814211b0f3f 100644
--- a/drivers/gpu/drm/drm_prime.c
+++ b/drivers/gpu/drm/drm_prime.c
@@ -713,6 +713,15 @@ int drm_gem_prime_mmap(struct drm_gem_object *obj, struct vm_area_struct *vma)
 	struct file *fil;
 	int ret;
 
+	if (obj->funcs && obj->funcs->mmap) {
+		ret = obj->funcs->mmap(obj, vma);
+		if (ret)
+			return ret;
+		vma->vm_private_data = obj;
+		drm_gem_object_get(obj);
+		return 0;
+	}
+
 	priv = kzalloc(sizeof(*priv), GFP_KERNEL);
 	fil = kzalloc(sizeof(*fil), GFP_KERNEL);
 	if (!priv || !fil) {
diff --git a/drivers/gpu/drm/drm_print.c b/drivers/gpu/drm/drm_print.c
index a17c8a14dba4..9a25d73c155c 100644
--- a/drivers/gpu/drm/drm_print.c
+++ b/drivers/gpu/drm/drm_print.c
@@ -28,6 +28,7 @@
 #include <stdarg.h>
 
 #include <linux/io.h>
+#include <linux/moduleparam.h>
 #include <linux/seq_file.h>
 #include <linux/slab.h>
 
@@ -35,6 +36,24 @@
 #include <drm/drm_drv.h>
 #include <drm/drm_print.h>
 
+/*
+ * drm_debug: Enable debug output.
+ * Bitmask of DRM_UT_x. See include/drm/drm_print.h for details.
+ */
+unsigned int drm_debug;
+EXPORT_SYMBOL(drm_debug);
+
+MODULE_PARM_DESC(debug, "Enable debug output, where each bit enables a debug category.\n"
+"\t\tBit 0 (0x01)  will enable CORE messages (drm core code)\n"
+"\t\tBit 1 (0x02)  will enable DRIVER messages (drm controller code)\n"
+"\t\tBit 2 (0x04)  will enable KMS messages (modesetting code)\n"
+"\t\tBit 3 (0x08)  will enable PRIME messages (prime code)\n"
+"\t\tBit 4 (0x10)  will enable ATOMIC messages (atomic code)\n"
+"\t\tBit 5 (0x20)  will enable VBL messages (vblank code)\n"
+"\t\tBit 7 (0x80)  will enable LEASE messages (leasing code)\n"
+"\t\tBit 8 (0x100) will enable DP messages (displayport code)");
+module_param_named(debug, drm_debug, int, 0600);
+
 void __drm_puts_coredump(struct drm_printer *p, const char *str)
 {
 	struct drm_print_iterator *iterator = p->arg;
@@ -147,6 +166,12 @@ void __drm_printfn_debug(struct drm_printer *p, struct va_format *vaf)
 }
 EXPORT_SYMBOL(__drm_printfn_debug);
 
+void __drm_printfn_err(struct drm_printer *p, struct va_format *vaf)
+{
+	pr_err("*ERROR* %s %pV", p->prefix, vaf);
+}
+EXPORT_SYMBOL(__drm_printfn_err);
+
 /**
  * drm_puts - print a const string to a &drm_printer stream
  * @p: the &drm printer
@@ -179,6 +204,37 @@ void drm_printf(struct drm_printer *p, const char *f, ...)
 }
 EXPORT_SYMBOL(drm_printf);
 
+/**
+ * drm_print_bits - print bits to a &drm_printer stream
+ *
+ * Print bits (in flag fields for example) in human readable form.
+ *
+ * @p: the &drm_printer
+ * @value: field value.
+ * @bits: Array with bit names.
+ * @nbits: Size of bit names array.
+ */
+void drm_print_bits(struct drm_printer *p, unsigned long value,
+		    const char * const bits[], unsigned int nbits)
+{
+	bool first = true;
+	unsigned int i;
+
+	if (WARN_ON_ONCE(nbits > BITS_PER_TYPE(value)))
+		nbits = BITS_PER_TYPE(value);
+
+	for_each_set_bit(i, &value, nbits) {
+		if (WARN_ON_ONCE(!bits[i]))
+			continue;
+		drm_printf(p, "%s%s", first ? "" : ",",
+			   bits[i]);
+		first = false;
+	}
+	if (first)
+		drm_printf(p, "(none)");
+}
+EXPORT_SYMBOL(drm_print_bits);
+
 void drm_dev_printk(const struct device *dev, const char *level,
 		    const char *format, ...)
 {
@@ -206,7 +262,7 @@ void drm_dev_dbg(const struct device *dev, unsigned int category,
 	struct va_format vaf;
 	va_list args;
 
-	if (!(drm_debug & category))
+	if (!drm_debug_enabled(category))
 		return;
 
 	va_start(args, format);
@@ -229,7 +285,7 @@ void drm_dbg(unsigned int category, const char *format, ...)
 	struct va_format vaf;
 	va_list args;
 
-	if (!(drm_debug & category))
+	if (!drm_debug_enabled(category))
 		return;
 
 	va_start(args, format);
diff --git a/drivers/gpu/drm/drm_probe_helper.c b/drivers/gpu/drm/drm_probe_helper.c
index ef2c468205a2..a7c87abe88d0 100644
--- a/drivers/gpu/drm/drm_probe_helper.c
+++ b/drivers/gpu/drm/drm_probe_helper.c
@@ -32,6 +32,7 @@
 #include <linux/export.h>
 #include <linux/moduleparam.h>
 
+#include <drm/drm_bridge.h>
 #include <drm/drm_client.h>
 #include <drm/drm_crtc.h>
 #include <drm/drm_edid.h>
@@ -92,7 +93,6 @@ drm_mode_validate_pipeline(struct drm_display_mode *mode,
 	struct drm_device *dev = connector->dev;
 	enum drm_mode_status ret = MODE_OK;
 	struct drm_encoder *encoder;
-	int i;
 
 	/* Step 1: Validate against connector */
 	ret = drm_connector_mode_valid(connector, mode);
@@ -100,7 +100,7 @@ drm_mode_validate_pipeline(struct drm_display_mode *mode,
 		return ret;
 
 	/* Step 2: Validate against encoders and crtcs */
-	drm_connector_for_each_possible_encoder(connector, encoder, i) {
+	drm_connector_for_each_possible_encoder(connector, encoder) {
 		struct drm_crtc *crtc;
 
 		ret = drm_encoder_mode_valid(encoder, mode);
diff --git a/drivers/gpu/drm/drm_self_refresh_helper.c b/drivers/gpu/drm/drm_self_refresh_helper.c
index 68f4765a5896..dd33fec5aabd 100644
--- a/drivers/gpu/drm/drm_self_refresh_helper.c
+++ b/drivers/gpu/drm/drm_self_refresh_helper.c
@@ -133,29 +133,33 @@ out_drop_locks:
  * drm_self_refresh_helper_update_avg_times - Updates a crtc's SR time averages
  * @state: the state which has just been applied to hardware
  * @commit_time_ms: the amount of time in ms that this commit took to complete
+ * @new_self_refresh_mask: bitmask of crtc's that have self_refresh_active in
+ *    new state
  *
  * Called after &drm_mode_config_funcs.atomic_commit_tail, this function will
  * update the average entry/exit self refresh times on self refresh transitions.
  * These averages will be used when calculating how long to delay before
  * entering self refresh mode after activity.
  */
-void drm_self_refresh_helper_update_avg_times(struct drm_atomic_state *state,
-					      unsigned int commit_time_ms)
+void
+drm_self_refresh_helper_update_avg_times(struct drm_atomic_state *state,
+					 unsigned int commit_time_ms,
+					 unsigned int new_self_refresh_mask)
 {
 	struct drm_crtc *crtc;
-	struct drm_crtc_state *old_crtc_state, *new_crtc_state;
+	struct drm_crtc_state *old_crtc_state;
 	int i;
 
-	for_each_oldnew_crtc_in_state(state, crtc, old_crtc_state,
-				      new_crtc_state, i) {
+	for_each_old_crtc_in_state(state, crtc, old_crtc_state, i) {
+		bool new_self_refresh_active = new_self_refresh_mask & BIT(i);
 		struct drm_self_refresh_data *sr_data = crtc->self_refresh_data;
 		struct ewma_psr_time *time;
 
 		if (old_crtc_state->self_refresh_active ==
-		    new_crtc_state->self_refresh_active)
+		    new_self_refresh_active)
 			continue;
 
-		if (new_crtc_state->self_refresh_active)
+		if (new_self_refresh_active)
 			time = &sr_data->entry_avg_ms;
 		else
 			time = &sr_data->exit_avg_ms;
diff --git a/drivers/gpu/drm/drm_simple_kms_helper.c b/drivers/gpu/drm/drm_simple_kms_helper.c
index b11910f14c46..15fb516ae2d8 100644
--- a/drivers/gpu/drm/drm_simple_kms_helper.c
+++ b/drivers/gpu/drm/drm_simple_kms_helper.c
@@ -8,6 +8,7 @@
 
 #include <drm/drm_atomic.h>
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_plane_helper.h>
 #include <drm/drm_probe_helper.h>
 #include <drm/drm_simple_kms_helper.h>
@@ -42,7 +43,7 @@ drm_simple_kms_crtc_mode_valid(struct drm_crtc *crtc,
 		/* Anything goes */
 		return MODE_OK;
 
-	return pipe->funcs->mode_valid(crtc, mode);
+	return pipe->funcs->mode_valid(pipe, mode);
 }
 
 static int drm_simple_kms_crtc_check(struct drm_crtc *crtc,
diff --git a/drivers/gpu/drm/drm_syncobj.c b/drivers/gpu/drm/drm_syncobj.c
index 4b5c7b0ed714..669c93fe2500 100644
--- a/drivers/gpu/drm/drm_syncobj.c
+++ b/drivers/gpu/drm/drm_syncobj.c
@@ -135,6 +135,7 @@
 #include <drm/drm_gem.h>
 #include <drm/drm_print.h>
 #include <drm/drm_syncobj.h>
+#include <drm/drm_utils.h>
 
 #include "drm_internal.h"
 
@@ -1279,7 +1280,7 @@ drm_syncobj_timeline_signal_ioctl(struct drm_device *dev, void *data,
 	if (!drm_core_check_feature(dev, DRIVER_SYNCOBJ_TIMELINE))
 		return -EOPNOTSUPP;
 
-	if (args->pad != 0)
+	if (args->flags != 0)
 		return -EINVAL;
 
 	if (args->count_handles == 0)
@@ -1350,7 +1351,7 @@ int drm_syncobj_query_ioctl(struct drm_device *dev, void *data,
 	if (!drm_core_check_feature(dev, DRIVER_SYNCOBJ_TIMELINE))
 		return -EOPNOTSUPP;
 
-	if (args->pad != 0)
+	if (args->flags & ~DRM_SYNCOBJ_QUERY_FLAGS_LAST_SUBMITTED)
 		return -EINVAL;
 
 	if (args->count_handles == 0)
@@ -1371,25 +1372,32 @@ int drm_syncobj_query_ioctl(struct drm_device *dev, void *data,
 		fence = drm_syncobj_fence_get(syncobjs[i]);
 		chain = to_dma_fence_chain(fence);
 		if (chain) {
-			struct dma_fence *iter, *last_signaled = NULL;
-
-			dma_fence_chain_for_each(iter, fence) {
-				if (iter->context != fence->context) {
-					dma_fence_put(iter);
-					/* It is most likely that timeline has
-					 * unorder points. */
-					break;
+			struct dma_fence *iter, *last_signaled =
+				dma_fence_get(fence);
+
+			if (args->flags &
+			    DRM_SYNCOBJ_QUERY_FLAGS_LAST_SUBMITTED) {
+				point = fence->seqno;
+			} else {
+				dma_fence_chain_for_each(iter, fence) {
+					if (iter->context != fence->context) {
+						dma_fence_put(iter);
+						/* It is most likely that timeline has
+						* unorder points. */
+						break;
+					}
+					dma_fence_put(last_signaled);
+					last_signaled = dma_fence_get(iter);
 				}
-				dma_fence_put(last_signaled);
-				last_signaled = dma_fence_get(iter);
+				point = dma_fence_is_signaled(last_signaled) ?
+					last_signaled->seqno :
+					to_dma_fence_chain(last_signaled)->prev_seqno;
 			}
-			point = dma_fence_is_signaled(last_signaled) ?
-				last_signaled->seqno :
-				to_dma_fence_chain(last_signaled)->prev_seqno;
 			dma_fence_put(last_signaled);
 		} else {
 			point = 0;
 		}
+		dma_fence_put(fence);
 		ret = copy_to_user(&points[i], &point, sizeof(uint64_t));
 		ret = ret ? -EFAULT : 0;
 		if (ret)
diff --git a/drivers/gpu/drm/drm_trace.h b/drivers/gpu/drm/drm_trace.h
index 471eb927474b..11c6dd577e8e 100644
--- a/drivers/gpu/drm/drm_trace.h
+++ b/drivers/gpu/drm/drm_trace.h
@@ -13,17 +13,23 @@ struct drm_file;
 #define TRACE_INCLUDE_FILE drm_trace
 
 TRACE_EVENT(drm_vblank_event,
-	    TP_PROTO(int crtc, unsigned int seq),
-	    TP_ARGS(crtc, seq),
+	    TP_PROTO(int crtc, unsigned int seq, ktime_t time, bool high_prec),
+	    TP_ARGS(crtc, seq, time, high_prec),
 	    TP_STRUCT__entry(
 		    __field(int, crtc)
 		    __field(unsigned int, seq)
+		    __field(ktime_t, time)
+		    __field(bool, high_prec)
 		    ),
 	    TP_fast_assign(
 		    __entry->crtc = crtc;
 		    __entry->seq = seq;
-		    ),
-	    TP_printk("crtc=%d, seq=%u", __entry->crtc, __entry->seq)
+		    __entry->time = time;
+		    __entry->high_prec = high_prec;
+			),
+	    TP_printk("crtc=%d, seq=%u, time=%lld, high-prec=%s",
+			__entry->crtc, __entry->seq, __entry->time,
+			__entry->high_prec ? "true" : "false")
 );
 
 TRACE_EVENT(drm_vblank_event_queued,
diff --git a/drivers/gpu/drm/drm_vblank.c b/drivers/gpu/drm/drm_vblank.c
index fd1fbc77871f..1659b13b178c 100644
--- a/drivers/gpu/drm/drm_vblank.c
+++ b/drivers/gpu/drm/drm_vblank.c
@@ -106,7 +106,7 @@ static void store_vblank(struct drm_device *dev, unsigned int pipe,
 
 	write_seqlock(&vblank->seqlock);
 	vblank->time = t_vblank;
-	vblank->count += vblank_count_inc;
+	atomic64_add(vblank_count_inc, &vblank->count);
 	write_sequnlock(&vblank->seqlock);
 }
 
@@ -272,7 +272,8 @@ static void drm_update_vblank_count(struct drm_device *dev, unsigned int pipe,
 
 	DRM_DEBUG_VBL("updating vblank count on crtc %u:"
 		      " current=%llu, diff=%u, hw=%u hw_last=%u\n",
-		      pipe, vblank->count, diff, cur_vblank, vblank->last);
+		      pipe, atomic64_read(&vblank->count), diff,
+		      cur_vblank, vblank->last);
 
 	if (diff == 0) {
 		WARN_ON_ONCE(cur_vblank != vblank->last);
@@ -294,11 +295,23 @@ static void drm_update_vblank_count(struct drm_device *dev, unsigned int pipe,
 static u64 drm_vblank_count(struct drm_device *dev, unsigned int pipe)
 {
 	struct drm_vblank_crtc *vblank = &dev->vblank[pipe];
+	u64 count;
 
 	if (WARN_ON(pipe >= dev->num_crtcs))
 		return 0;
 
-	return vblank->count;
+	count = atomic64_read(&vblank->count);
+
+	/*
+	 * This read barrier corresponds to the implicit write barrier of the
+	 * write seqlock in store_vblank(). Note that this is the only place
+	 * where we need an explicit barrier, since all other access goes
+	 * through drm_vblank_count_and_time(), which already has the required
+	 * read barrier curtesy of the read seqlock.
+	 */
+	smp_rmb();
+
+	return count;
 }
 
 /**
@@ -319,7 +332,7 @@ u64 drm_crtc_accurate_vblank_count(struct drm_crtc *crtc)
 	u64 vblank;
 	unsigned long flags;
 
-	WARN_ONCE(drm_debug & DRM_UT_VBL && !dev->driver->get_vblank_timestamp,
+	WARN_ONCE(drm_debug_enabled(DRM_UT_VBL) && !dev->driver->get_vblank_timestamp,
 		  "This function requires support for accurate vblank timestamps.");
 
 	spin_lock_irqsave(&dev->vblank_time_lock, flags);
@@ -693,7 +706,7 @@ bool drm_calc_vbltimestamp_from_scanoutpos(struct drm_device *dev,
 	 */
 	*vblank_time = ktime_sub_ns(etime, delta_ns);
 
-	if ((drm_debug & DRM_UT_VBL) == 0)
+	if (!drm_debug_enabled(DRM_UT_VBL))
 		return true;
 
 	ts_etime = ktime_to_timespec64(etime);
@@ -763,6 +776,14 @@ drm_get_last_vbltimestamp(struct drm_device *dev, unsigned int pipe,
  * vblank interrupt (since it only reports the software vblank counter), see
  * drm_crtc_accurate_vblank_count() for such use-cases.
  *
+ * Note that for a given vblank counter value drm_crtc_handle_vblank()
+ * and drm_crtc_vblank_count() or drm_crtc_vblank_count_and_time()
+ * provide a barrier: Any writes done before calling
+ * drm_crtc_handle_vblank() will be visible to callers of the later
+ * functions, iff the vblank count is the same or a later one.
+ *
+ * See also &drm_vblank_crtc.count.
+ *
  * Returns:
  * The software vblank counter.
  */
@@ -800,7 +821,7 @@ static u64 drm_vblank_count_and_time(struct drm_device *dev, unsigned int pipe,
 
 	do {
 		seq = read_seqbegin(&vblank->seqlock);
-		vblank_count = vblank->count;
+		vblank_count = atomic64_read(&vblank->count);
 		*vblanktime = vblank->time;
 	} while (read_seqretry(&vblank->seqlock, seq));
 
@@ -817,6 +838,14 @@ static u64 drm_vblank_count_and_time(struct drm_device *dev, unsigned int pipe,
  * vblank events since the system was booted, including lost events due to
  * modesetting activity. Returns corresponding system timestamp of the time
  * of the vblank interval that corresponds to the current vblank counter value.
+ *
+ * Note that for a given vblank counter value drm_crtc_handle_vblank()
+ * and drm_crtc_vblank_count() or drm_crtc_vblank_count_and_time()
+ * provide a barrier: Any writes done before calling
+ * drm_crtc_handle_vblank() will be visible to callers of the later
+ * functions, iff the vblank count is the same or a later one.
+ *
+ * See also &drm_vblank_crtc.count.
  */
 u64 drm_crtc_vblank_count_and_time(struct drm_crtc *crtc,
 				   ktime_t *vblanktime)
@@ -1323,7 +1352,7 @@ void drm_vblank_restore(struct drm_device *dev, unsigned int pipe)
 	assert_spin_locked(&dev->vblank_time_lock);
 
 	vblank = &dev->vblank[pipe];
-	WARN_ONCE((drm_debug & DRM_UT_VBL) && !vblank->framedur_ns,
+	WARN_ONCE(drm_debug_enabled(DRM_UT_VBL) && !vblank->framedur_ns,
 		  "Cannot compute missed vblanks without frame duration\n");
 	framedur_ns = vblank->framedur_ns;
 
@@ -1581,7 +1610,7 @@ int drm_wait_vblank_ioctl(struct drm_device *dev, void *data,
 	unsigned int flags, pipe, high_pipe;
 
 	if (!dev->irq_enabled)
-		return -EINVAL;
+		return -EOPNOTSUPP;
 
 	if (vblwait->request.type & _DRM_VBLANK_SIGNAL)
 		return -EINVAL;
@@ -1731,7 +1760,8 @@ static void drm_handle_vblank_events(struct drm_device *dev, unsigned int pipe)
 		send_vblank_event(dev, e, seq, now);
 	}
 
-	trace_drm_vblank_event(pipe, seq);
+	trace_drm_vblank_event(pipe, seq, now,
+			dev->driver->get_vblank_timestamp != NULL);
 }
 
 /**
@@ -1806,6 +1836,14 @@ EXPORT_SYMBOL(drm_handle_vblank);
  *
  * This is the native KMS version of drm_handle_vblank().
  *
+ * Note that for a given vblank counter value drm_crtc_handle_vblank()
+ * and drm_crtc_vblank_count() or drm_crtc_vblank_count_and_time()
+ * provide a barrier: Any writes done before calling
+ * drm_crtc_handle_vblank() will be visible to callers of the later
+ * functions, iff the vblank count is the same or a later one.
+ *
+ * See also &drm_vblank_crtc.count.
+ *
  * Returns:
  * True if the event was successfully handled, false on failure.
  */
@@ -1838,7 +1876,7 @@ int drm_crtc_get_sequence_ioctl(struct drm_device *dev, void *data,
 		return -EOPNOTSUPP;
 
 	if (!dev->irq_enabled)
-		return -EINVAL;
+		return -EOPNOTSUPP;
 
 	crtc = drm_crtc_find(dev, file_priv, get_seq->crtc_id);
 	if (!crtc)
@@ -1896,7 +1934,7 @@ int drm_crtc_queue_sequence_ioctl(struct drm_device *dev, void *data,
 		return -EOPNOTSUPP;
 
 	if (!dev->irq_enabled)
-		return -EINVAL;
+		return -EOPNOTSUPP;
 
 	crtc = drm_crtc_find(dev, file_priv, queue_seq->crtc_id);
 	if (!crtc)
diff --git a/drivers/gpu/drm/drm_vram_helper_common.c b/drivers/gpu/drm/drm_vram_helper_common.c
index e9c9f9a80ba3..2000d9b33fd5 100644
--- a/drivers/gpu/drm/drm_vram_helper_common.c
+++ b/drivers/gpu/drm/drm_vram_helper_common.c
@@ -7,9 +7,8 @@
  *
  * This library provides &struct drm_gem_vram_object (GEM VRAM), a GEM
  * buffer object that is backed by video RAM. It can be used for
- * framebuffer devices with dedicated memory. The video RAM can be
- * managed with &struct drm_vram_mm (VRAM MM). Both data structures are
- * supposed to be used together, but can also be used individually.
+ * framebuffer devices with dedicated memory. The video RAM is managed
+ * by &struct drm_vram_mm (VRAM MM).
  *
  * With the GEM interface userspace applications create, manage and destroy
  * graphics buffers, such as an on-screen framebuffer. GEM does not provide
@@ -50,8 +49,7 @@
  *		// setup device, vram base and size
  *		// ...
  *
- *		ret = drm_vram_helper_alloc_mm(dev, vram_base, vram_size,
- *					       &drm_gem_vram_mm_funcs);
+ *		ret = drm_vram_helper_alloc_mm(dev, vram_base, vram_size);
  *		if (ret)
  *			return ret;
  *		return 0;
diff --git a/drivers/gpu/drm/drm_vram_mm_helper.c b/drivers/gpu/drm/drm_vram_mm_helper.c
deleted file mode 100644
index c911781d6728..000000000000
--- a/drivers/gpu/drm/drm_vram_mm_helper.c
+++ /dev/null
@@ -1,297 +0,0 @@
-// SPDX-License-Identifier: GPL-2.0-or-later
-
-#include <drm/drm_device.h>
-#include <drm/drm_file.h>
-#include <drm/drm_vram_mm_helper.h>
-
-#include <drm/ttm/ttm_page_alloc.h>
-
-/**
- * DOC: overview
- *
- * The data structure &struct drm_vram_mm and its helpers implement a memory
- * manager for simple framebuffer devices with dedicated video memory. Buffer
- * objects are either placed in video RAM or evicted to system memory. These
- * helper functions work well with &struct drm_gem_vram_object.
- */
-
-/*
- * TTM TT
- */
-
-static void backend_func_destroy(struct ttm_tt *tt)
-{
-	ttm_tt_fini(tt);
-	kfree(tt);
-}
-
-static struct ttm_backend_func backend_func = {
-	.destroy = backend_func_destroy
-};
-
-/*
- * TTM BO device
- */
-
-static struct ttm_tt *bo_driver_ttm_tt_create(struct ttm_buffer_object *bo,
-					      uint32_t page_flags)
-{
-	struct ttm_tt *tt;
-	int ret;
-
-	tt = kzalloc(sizeof(*tt), GFP_KERNEL);
-	if (!tt)
-		return NULL;
-
-	tt->func = &backend_func;
-
-	ret = ttm_tt_init(tt, bo, page_flags);
-	if (ret < 0)
-		goto err_ttm_tt_init;
-
-	return tt;
-
-err_ttm_tt_init:
-	kfree(tt);
-	return NULL;
-}
-
-static int bo_driver_init_mem_type(struct ttm_bo_device *bdev, uint32_t type,
-				   struct ttm_mem_type_manager *man)
-{
-	switch (type) {
-	case TTM_PL_SYSTEM:
-		man->flags = TTM_MEMTYPE_FLAG_MAPPABLE;
-		man->available_caching = TTM_PL_MASK_CACHING;
-		man->default_caching = TTM_PL_FLAG_CACHED;
-		break;
-	case TTM_PL_VRAM:
-		man->func = &ttm_bo_manager_func;
-		man->flags = TTM_MEMTYPE_FLAG_FIXED |
-			     TTM_MEMTYPE_FLAG_MAPPABLE;
-		man->available_caching = TTM_PL_FLAG_UNCACHED |
-					 TTM_PL_FLAG_WC;
-		man->default_caching = TTM_PL_FLAG_WC;
-		break;
-	default:
-		return -EINVAL;
-	}
-	return 0;
-}
-
-static void bo_driver_evict_flags(struct ttm_buffer_object *bo,
-				  struct ttm_placement *placement)
-{
-	struct drm_vram_mm *vmm = drm_vram_mm_of_bdev(bo->bdev);
-
-	if (vmm->funcs && vmm->funcs->evict_flags)
-		vmm->funcs->evict_flags(bo, placement);
-}
-
-static int bo_driver_verify_access(struct ttm_buffer_object *bo,
-				   struct file *filp)
-{
-	struct drm_vram_mm *vmm = drm_vram_mm_of_bdev(bo->bdev);
-
-	if (!vmm->funcs || !vmm->funcs->verify_access)
-		return 0;
-	return vmm->funcs->verify_access(bo, filp);
-}
-
-static int bo_driver_io_mem_reserve(struct ttm_bo_device *bdev,
-				    struct ttm_mem_reg *mem)
-{
-	struct ttm_mem_type_manager *man = bdev->man + mem->mem_type;
-	struct drm_vram_mm *vmm = drm_vram_mm_of_bdev(bdev);
-
-	if (!(man->flags & TTM_MEMTYPE_FLAG_MAPPABLE))
-		return -EINVAL;
-
-	mem->bus.addr = NULL;
-	mem->bus.size = mem->num_pages << PAGE_SHIFT;
-
-	switch (mem->mem_type) {
-	case TTM_PL_SYSTEM:	/* nothing to do */
-		mem->bus.offset = 0;
-		mem->bus.base = 0;
-		mem->bus.is_iomem = false;
-		break;
-	case TTM_PL_VRAM:
-		mem->bus.offset = mem->start << PAGE_SHIFT;
-		mem->bus.base = vmm->vram_base;
-		mem->bus.is_iomem = true;
-		break;
-	default:
-		return -EINVAL;
-	}
-
-	return 0;
-}
-
-static void bo_driver_io_mem_free(struct ttm_bo_device *bdev,
-				  struct ttm_mem_reg *mem)
-{ }
-
-static struct ttm_bo_driver bo_driver = {
-	.ttm_tt_create = bo_driver_ttm_tt_create,
-	.ttm_tt_populate = ttm_pool_populate,
-	.ttm_tt_unpopulate = ttm_pool_unpopulate,
-	.init_mem_type = bo_driver_init_mem_type,
-	.eviction_valuable = ttm_bo_eviction_valuable,
-	.evict_flags = bo_driver_evict_flags,
-	.verify_access = bo_driver_verify_access,
-	.io_mem_reserve = bo_driver_io_mem_reserve,
-	.io_mem_free = bo_driver_io_mem_free,
-};
-
-/*
- * struct drm_vram_mm
- */
-
-/**
- * drm_vram_mm_init() - Initialize an instance of VRAM MM.
- * @vmm:	the VRAM MM instance to initialize
- * @dev:	the DRM device
- * @vram_base:	the base address of the video memory
- * @vram_size:	the size of the video memory in bytes
- * @funcs:	callback functions for buffer objects
- *
- * Returns:
- * 0 on success, or
- * a negative error code otherwise.
- */
-int drm_vram_mm_init(struct drm_vram_mm *vmm, struct drm_device *dev,
-		     uint64_t vram_base, size_t vram_size,
-		     const struct drm_vram_mm_funcs *funcs)
-{
-	int ret;
-
-	vmm->vram_base = vram_base;
-	vmm->vram_size = vram_size;
-	vmm->funcs = funcs;
-
-	ret = ttm_bo_device_init(&vmm->bdev, &bo_driver,
-				 dev->anon_inode->i_mapping,
-				 true);
-	if (ret)
-		return ret;
-
-	ret = ttm_bo_init_mm(&vmm->bdev, TTM_PL_VRAM, vram_size >> PAGE_SHIFT);
-	if (ret)
-		return ret;
-
-	return 0;
-}
-EXPORT_SYMBOL(drm_vram_mm_init);
-
-/**
- * drm_vram_mm_cleanup() - Cleans up an initialized instance of VRAM MM.
- * @vmm:	the VRAM MM instance to clean up
- */
-void drm_vram_mm_cleanup(struct drm_vram_mm *vmm)
-{
-	ttm_bo_device_release(&vmm->bdev);
-}
-EXPORT_SYMBOL(drm_vram_mm_cleanup);
-
-/**
- * drm_vram_mm_mmap() - Helper for implementing &struct file_operations.mmap()
- * @filp:	the mapping's file structure
- * @vma:	the mapping's memory area
- * @vmm:	the VRAM MM instance
- *
- * Returns:
- * 0 on success, or
- * a negative error code otherwise.
- */
-int drm_vram_mm_mmap(struct file *filp, struct vm_area_struct *vma,
-		     struct drm_vram_mm *vmm)
-{
-	return ttm_bo_mmap(filp, vma, &vmm->bdev);
-}
-EXPORT_SYMBOL(drm_vram_mm_mmap);
-
-/*
- * Helpers for integration with struct drm_device
- */
-
-/**
- * drm_vram_helper_alloc_mm - Allocates a device's instance of \
-	&struct drm_vram_mm
- * @dev:	the DRM device
- * @vram_base:	the base address of the video memory
- * @vram_size:	the size of the video memory in bytes
- * @funcs:	callback functions for buffer objects
- *
- * Returns:
- * The new instance of &struct drm_vram_mm on success, or
- * an ERR_PTR()-encoded errno code otherwise.
- */
-struct drm_vram_mm *drm_vram_helper_alloc_mm(
-	struct drm_device *dev, uint64_t vram_base, size_t vram_size,
-	const struct drm_vram_mm_funcs *funcs)
-{
-	int ret;
-
-	if (WARN_ON(dev->vram_mm))
-		return dev->vram_mm;
-
-	dev->vram_mm = kzalloc(sizeof(*dev->vram_mm), GFP_KERNEL);
-	if (!dev->vram_mm)
-		return ERR_PTR(-ENOMEM);
-
-	ret = drm_vram_mm_init(dev->vram_mm, dev, vram_base, vram_size, funcs);
-	if (ret)
-		goto err_kfree;
-
-	return dev->vram_mm;
-
-err_kfree:
-	kfree(dev->vram_mm);
-	dev->vram_mm = NULL;
-	return ERR_PTR(ret);
-}
-EXPORT_SYMBOL(drm_vram_helper_alloc_mm);
-
-/**
- * drm_vram_helper_release_mm - Releases a device's instance of \
-	&struct drm_vram_mm
- * @dev:	the DRM device
- */
-void drm_vram_helper_release_mm(struct drm_device *dev)
-{
-	if (!dev->vram_mm)
-		return;
-
-	drm_vram_mm_cleanup(dev->vram_mm);
-	kfree(dev->vram_mm);
-	dev->vram_mm = NULL;
-}
-EXPORT_SYMBOL(drm_vram_helper_release_mm);
-
-/*
- * Helpers for &struct file_operations
- */
-
-/**
- * drm_vram_mm_file_operations_mmap() - \
-	Implements &struct file_operations.mmap()
- * @filp:	the mapping's file structure
- * @vma:	the mapping's memory area
- *
- * Returns:
- * 0 on success, or
- * a negative error code otherwise.
- */
-int drm_vram_mm_file_operations_mmap(
-	struct file *filp, struct vm_area_struct *vma)
-{
-	struct drm_file *file_priv = filp->private_data;
-	struct drm_device *dev = file_priv->minor->dev;
-
-	if (WARN_ONCE(!dev->vram_mm, "VRAM MM not initialized"))
-		return -EINVAL;
-
-	return drm_vram_mm_mmap(filp, vma, dev->vram_mm);
-}
-EXPORT_SYMBOL(drm_vram_mm_file_operations_mmap);
diff --git a/drivers/gpu/drm/drm_writeback.c b/drivers/gpu/drm/drm_writeback.c
index ff138b6ec48b..43d9e3bb3a94 100644
--- a/drivers/gpu/drm/drm_writeback.c
+++ b/drivers/gpu/drm/drm_writeback.c
@@ -324,6 +324,9 @@ void drm_writeback_cleanup_job(struct drm_writeback_job *job)
 	if (job->fb)
 		drm_framebuffer_put(job->fb);
 
+	if (job->out_fence)
+		dma_fence_put(job->out_fence);
+
 	kfree(job);
 }
 EXPORT_SYMBOL(drm_writeback_cleanup_job);
@@ -366,25 +369,29 @@ drm_writeback_signal_completion(struct drm_writeback_connector *wb_connector,
 {
 	unsigned long flags;
 	struct drm_writeback_job *job;
+	struct dma_fence *out_fence;
 
 	spin_lock_irqsave(&wb_connector->job_lock, flags);
 	job = list_first_entry_or_null(&wb_connector->job_queue,
 				       struct drm_writeback_job,
 				       list_entry);
-	if (job) {
+	if (job)
 		list_del(&job->list_entry);
-		if (job->out_fence) {
-			if (status)
-				dma_fence_set_error(job->out_fence, status);
-			dma_fence_signal(job->out_fence);
-			dma_fence_put(job->out_fence);
-		}
-	}
+
 	spin_unlock_irqrestore(&wb_connector->job_lock, flags);
 
 	if (WARN_ON(!job))
 		return;
 
+	out_fence = job->out_fence;
+	if (out_fence) {
+		if (status)
+			dma_fence_set_error(out_fence, status);
+		dma_fence_signal(out_fence);
+		dma_fence_put(out_fence);
+		job->out_fence = NULL;
+	}
+
 	INIT_WORK(&job->cleanup_work, cleanup_work);
 	queue_work(system_long_wq, &job->cleanup_work);
 }
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_buffer.c b/drivers/gpu/drm/etnaviv/etnaviv_buffer.c
index 7e4e2959bf4f..32d9fac587f9 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_buffer.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_buffer.c
@@ -326,7 +326,7 @@ void etnaviv_buffer_queue(struct etnaviv_gpu *gpu, u32 exec_state,
 
 	lockdep_assert_held(&gpu->lock);
 
-	if (drm_debug & DRM_UT_DRIVER)
+	if (drm_debug_enabled(DRM_UT_DRIVER))
 		etnaviv_buffer_dump(gpu, buffer, 0, 0x50);
 
 	link_target = etnaviv_cmdbuf_get_va(cmdbuf,
@@ -459,13 +459,13 @@ void etnaviv_buffer_queue(struct etnaviv_gpu *gpu, u32 exec_state,
 		 etnaviv_cmdbuf_get_va(buffer, &gpu->mmu_context->cmdbuf_mapping)
 		 + buffer->user_size - 4);
 
-	if (drm_debug & DRM_UT_DRIVER)
+	if (drm_debug_enabled(DRM_UT_DRIVER))
 		pr_info("stream link to 0x%08x @ 0x%08x %p\n",
 			return_target,
 			etnaviv_cmdbuf_get_va(cmdbuf, &gpu->mmu_context->cmdbuf_mapping),
 			cmdbuf->vaddr);
 
-	if (drm_debug & DRM_UT_DRIVER) {
+	if (drm_debug_enabled(DRM_UT_DRIVER)) {
 		print_hex_dump(KERN_INFO, "cmd ", DUMP_PREFIX_OFFSET, 16, 4,
 			       cmdbuf->vaddr, cmdbuf->size, 0);
 
@@ -484,6 +484,6 @@ void etnaviv_buffer_queue(struct etnaviv_gpu *gpu, u32 exec_state,
 				    VIV_FE_LINK_HEADER_PREFETCH(link_dwords),
 				    link_target);
 
-	if (drm_debug & DRM_UT_DRIVER)
+	if (drm_debug_enabled(DRM_UT_DRIVER))
 		etnaviv_buffer_dump(gpu, buffer, 0, 0x50);
 }
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_dump.c b/drivers/gpu/drm/etnaviv/etnaviv_dump.c
index 698db540972c..648cf0207309 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_dump.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_dump.c
@@ -180,6 +180,8 @@ void etnaviv_core_dump(struct etnaviv_gem_submit *submit)
 			      etnaviv_cmdbuf_get_va(&submit->cmdbuf,
 					&gpu->mmu_context->cmdbuf_mapping));
 
+	mutex_unlock(&gpu->mmu_context->lock);
+
 	/* Reserve space for the bomap */
 	if (n_bomap_pages) {
 		bomap_start = bomap = iter.data;
@@ -221,8 +223,6 @@ void etnaviv_core_dump(struct etnaviv_gem_submit *submit)
 					 obj->base.size);
 	}
 
-	mutex_unlock(&gpu->mmu_context->lock);
-
 	etnaviv_core_dump_header(&iter, ETDUMP_BUF_END, iter.data);
 
 	dev_coredumpv(gpu->dev, iter.start, iter.data - iter.start, GFP_KERNEL);
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_iommu_v2.c b/drivers/gpu/drm/etnaviv/etnaviv_iommu_v2.c
index 043111a1d60c..f8bf488e9d71 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_iommu_v2.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_iommu_v2.c
@@ -155,9 +155,11 @@ static void etnaviv_iommuv2_dump(struct etnaviv_iommu_context *context, void *bu
 
 	memcpy(buf, v2_context->mtlb_cpu, SZ_4K);
 	buf += SZ_4K;
-	for (i = 0; i < MMUv2_MAX_STLB_ENTRIES; i++, buf += SZ_4K)
-		if (v2_context->mtlb_cpu[i] & MMUv2_PTE_PRESENT)
+	for (i = 0; i < MMUv2_MAX_STLB_ENTRIES; i++)
+		if (v2_context->mtlb_cpu[i] & MMUv2_PTE_PRESENT) {
 			memcpy(buf, v2_context->stlb_cpu[i], SZ_4K);
+			buf += SZ_4K;
+		}
 }
 
 static void etnaviv_iommuv2_restore_nonsec(struct etnaviv_gpu *gpu,
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_mmu.c b/drivers/gpu/drm/etnaviv/etnaviv_mmu.c
index 35ebae6a1be7..3607d348c298 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_mmu.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_mmu.c
@@ -328,12 +328,23 @@ etnaviv_iommu_context_init(struct etnaviv_iommu_global *global,
 
 	ret = etnaviv_cmdbuf_suballoc_map(suballoc, ctx, &ctx->cmdbuf_mapping,
 					  global->memory_base);
-	if (ret) {
-		global->ops->free(ctx);
-		return NULL;
+	if (ret)
+		goto out_free;
+
+	if (global->version == ETNAVIV_IOMMU_V1 &&
+	    ctx->cmdbuf_mapping.iova > 0x80000000) {
+		dev_err(global->dev,
+		        "command buffer outside valid memory window\n");
+		goto out_unmap;
 	}
 
 	return ctx;
+
+out_unmap:
+	etnaviv_cmdbuf_suballoc_unmap(ctx, &ctx->cmdbuf_mapping);
+out_free:
+	global->ops->free(ctx);
+	return NULL;
 }
 
 void etnaviv_iommu_restore(struct etnaviv_gpu *gpu,
diff --git a/drivers/gpu/drm/exynos/exynos_dp.c b/drivers/gpu/drm/exynos/exynos_dp.c
index 3a0f0ba8c63a..1e6aa24bf45e 100644
--- a/drivers/gpu/drm/exynos/exynos_dp.c
+++ b/drivers/gpu/drm/exynos/exynos_dp.c
@@ -19,6 +19,7 @@
 
 #include <drm/bridge/analogix_dp.h>
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_crtc.h>
 #include <drm/drm_of.h>
 #include <drm/drm_panel.h>
diff --git a/drivers/gpu/drm/exynos/exynos_drm_dsi.c b/drivers/gpu/drm/exynos/exynos_drm_dsi.c
index 6926cee91b36..72726f2c7a9f 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_dsi.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_dsi.c
@@ -24,6 +24,7 @@
 #include <video/videomode.h>
 
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_fb_helper.h>
 #include <drm/drm_mipi_dsi.h>
 #include <drm/drm_panel.h>
diff --git a/drivers/gpu/drm/exynos/exynos_drm_mic.c b/drivers/gpu/drm/exynos/exynos_drm_mic.c
index b78e8c5ba553..f41d75923557 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_mic.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_mic.c
@@ -21,6 +21,7 @@
 #include <video/of_videomode.h>
 #include <video/videomode.h>
 
+#include <drm/drm_bridge.h>
 #include <drm/drm_encoder.h>
 #include <drm/drm_print.h>
 
diff --git a/drivers/gpu/drm/exynos/exynos_hdmi.c b/drivers/gpu/drm/exynos/exynos_hdmi.c
index bc1565f1822a..48159d5d2214 100644
--- a/drivers/gpu/drm/exynos/exynos_hdmi.c
+++ b/drivers/gpu/drm/exynos/exynos_hdmi.c
@@ -34,6 +34,7 @@
 #include <media/cec-notifier.h>
 
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_edid.h>
 #include <drm/drm_print.h>
 #include <drm/drm_probe_helper.h>
@@ -852,6 +853,10 @@ static enum drm_connector_status hdmi_detect(struct drm_connector *connector,
 
 static void hdmi_connector_destroy(struct drm_connector *connector)
 {
+	struct hdmi_context *hdata = connector_to_hdmi(connector);
+
+	cec_notifier_conn_unregister(hdata->notifier);
+
 	drm_connector_unregister(connector);
 	drm_connector_cleanup(connector);
 }
@@ -935,6 +940,7 @@ static int hdmi_create_connector(struct drm_encoder *encoder)
 {
 	struct hdmi_context *hdata = encoder_to_hdmi(encoder);
 	struct drm_connector *connector = &hdata->connector;
+	struct cec_connector_info conn_info;
 	int ret;
 
 	connector->interlace_allowed = true;
@@ -957,6 +963,15 @@ static int hdmi_create_connector(struct drm_encoder *encoder)
 			DRM_DEV_ERROR(hdata->dev, "Failed to attach bridge\n");
 	}
 
+	cec_fill_conn_info_from_drm(&conn_info, connector);
+
+	hdata->notifier = cec_notifier_conn_register(hdata->dev, NULL,
+						     &conn_info);
+	if (!hdata->notifier) {
+		ret = -ENOMEM;
+		DRM_DEV_ERROR(hdata->dev, "Failed to allocate CEC notifier\n");
+	}
+
 	return ret;
 }
 
@@ -1528,8 +1543,8 @@ static void hdmi_disable(struct drm_encoder *encoder)
 		 */
 		mutex_unlock(&hdata->mutex);
 		cancel_delayed_work(&hdata->hotplug_work);
-		cec_notifier_set_phys_addr(hdata->notifier,
-					   CEC_PHYS_ADDR_INVALID);
+		if (hdata->notifier)
+			cec_notifier_phys_addr_invalidate(hdata->notifier);
 		return;
 	}
 
@@ -2006,12 +2021,6 @@ static int hdmi_probe(struct platform_device *pdev)
 		}
 	}
 
-	hdata->notifier = cec_notifier_get(&pdev->dev);
-	if (hdata->notifier == NULL) {
-		ret = -ENOMEM;
-		goto err_hdmiphy;
-	}
-
 	pm_runtime_enable(dev);
 
 	audio_infoframe = &hdata->audio.infoframe;
@@ -2023,7 +2032,7 @@ static int hdmi_probe(struct platform_device *pdev)
 
 	ret = hdmi_register_audio_device(hdata);
 	if (ret)
-		goto err_notifier_put;
+		goto err_rpm_disable;
 
 	ret = component_add(&pdev->dev, &hdmi_component_ops);
 	if (ret)
@@ -2034,8 +2043,7 @@ static int hdmi_probe(struct platform_device *pdev)
 err_unregister_audio:
 	platform_device_unregister(hdata->audio.pdev);
 
-err_notifier_put:
-	cec_notifier_put(hdata->notifier);
+err_rpm_disable:
 	pm_runtime_disable(dev);
 
 err_hdmiphy:
@@ -2054,12 +2062,10 @@ static int hdmi_remove(struct platform_device *pdev)
 	struct hdmi_context *hdata = platform_get_drvdata(pdev);
 
 	cancel_delayed_work_sync(&hdata->hotplug_work);
-	cec_notifier_set_phys_addr(hdata->notifier, CEC_PHYS_ADDR_INVALID);
 
 	component_del(&pdev->dev, &hdmi_component_ops);
 	platform_device_unregister(hdata->audio.pdev);
 
-	cec_notifier_put(hdata->notifier);
 	pm_runtime_disable(&pdev->dev);
 
 	if (!IS_ERR(hdata->reg_hdmi_en))
diff --git a/drivers/gpu/drm/exynos/exynos_mixer.c b/drivers/gpu/drm/exynos/exynos_mixer.c
index 7b24338fad3c..6cfdb95fef2f 100644
--- a/drivers/gpu/drm/exynos/exynos_mixer.c
+++ b/drivers/gpu/drm/exynos/exynos_mixer.c
@@ -1069,9 +1069,9 @@ static bool mixer_mode_fixup(struct exynos_drm_crtc *crtc,
 	struct mixer_context *ctx = crtc->ctx;
 	int width = mode->hdisplay, height = mode->vdisplay, i;
 
-	struct {
+	static const struct {
 		int hdisplay, vdisplay, htotal, vtotal, scan_val;
-	} static const modes[] = {
+	} modes[] = {
 		{ 720, 480, 858, 525, MXR_CFG_SCAN_NTSC | MXR_CFG_SCAN_SD },
 		{ 720, 576, 864, 625, MXR_CFG_SCAN_PAL | MXR_CFG_SCAN_SD },
 		{ 1280, 720, 1650, 750, MXR_CFG_SCAN_HD_720 | MXR_CFG_SCAN_HD },
diff --git a/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_rgb.c b/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_rgb.c
index a92fd6c70b09..82c972e9c024 100644
--- a/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_rgb.c
+++ b/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_rgb.c
@@ -9,6 +9,7 @@
 #include <linux/of_graph.h>
 
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_of.h>
 #include <drm/drm_panel.h>
 #include <drm/drm_probe_helper.h>
diff --git a/drivers/gpu/drm/gma500/cdv_intel_display.c b/drivers/gpu/drm/gma500/cdv_intel_display.c
index f56852a503e8..8b784947ed3b 100644
--- a/drivers/gpu/drm/gma500/cdv_intel_display.c
+++ b/drivers/gpu/drm/gma500/cdv_intel_display.c
@@ -405,6 +405,8 @@ static bool cdv_intel_find_dp_pll(const struct gma_limit_t *limit,
 	struct gma_crtc *gma_crtc = to_gma_crtc(crtc);
 	struct gma_clock_t clock;
 
+	memset(&clock, 0, sizeof(clock));
+
 	switch (refclk) {
 	case 27000:
 		if (target < 200000) {
diff --git a/drivers/gpu/drm/gma500/mdfld_dsi_output.c b/drivers/gpu/drm/gma500/mdfld_dsi_output.c
index 03023fa0fb6f..f350ac1ead18 100644
--- a/drivers/gpu/drm/gma500/mdfld_dsi_output.c
+++ b/drivers/gpu/drm/gma500/mdfld_dsi_output.c
@@ -498,7 +498,7 @@ void mdfld_dsi_output_init(struct drm_device *dev,
 		return;
 	}
 
-	/*create a new connetor*/
+	/*create a new connector*/
 	dsi_connector = kzalloc(sizeof(struct mdfld_dsi_connector), GFP_KERNEL);
 	if (!dsi_connector) {
 		DRM_ERROR("No memory");
diff --git a/drivers/gpu/drm/gma500/oaktrail_crtc.c b/drivers/gpu/drm/gma500/oaktrail_crtc.c
index 167c10767dd4..900e5499249d 100644
--- a/drivers/gpu/drm/gma500/oaktrail_crtc.c
+++ b/drivers/gpu/drm/gma500/oaktrail_crtc.c
@@ -129,6 +129,7 @@ static bool mrst_sdvo_find_best_pll(const struct gma_limit_t *limit,
 	s32 freq_error, min_error = 100000;
 
 	memset(best_clock, 0, sizeof(*best_clock));
+	memset(&clock, 0, sizeof(clock));
 
 	for (clock.m = limit->m.min; clock.m <= limit->m.max; clock.m++) {
 		for (clock.n = limit->n.min; clock.n <= limit->n.max;
@@ -185,6 +186,7 @@ static bool mrst_lvds_find_best_pll(const struct gma_limit_t *limit,
 	int err = target;
 
 	memset(best_clock, 0, sizeof(*best_clock));
+	memset(&clock, 0, sizeof(clock));
 
 	for (clock.m = limit->m.min; clock.m <= limit->m.max; clock.m++) {
 		for (clock.p1 = limit->p1.min; clock.p1 <= limit->p1.max;
diff --git a/drivers/gpu/drm/hisilicon/hibmc/Kconfig b/drivers/gpu/drm/hisilicon/hibmc/Kconfig
index 35a3c5f0c38c..dfc5aef62f7b 100644
--- a/drivers/gpu/drm/hisilicon/hibmc/Kconfig
+++ b/drivers/gpu/drm/hisilicon/hibmc/Kconfig
@@ -4,7 +4,8 @@ config DRM_HISI_HIBMC
 	depends on DRM && PCI && MMU && ARM64
 	select DRM_KMS_HELPER
 	select DRM_VRAM_HELPER
-
+	select DRM_TTM
+	select DRM_TTM_HELPER
 	help
 	  Choose this option if you have a Hisilicon Hibmc soc chipset.
 	  If M is selected the module will be called hibmc-drm.
diff --git a/drivers/gpu/drm/hisilicon/hibmc/hibmc_drm_de.c b/drivers/gpu/drm/hisilicon/hibmc/hibmc_drm_de.c
index cc4c41748cfb..6527a97f68a3 100644
--- a/drivers/gpu/drm/hisilicon/hibmc/hibmc_drm_de.c
+++ b/drivers/gpu/drm/hisilicon/hibmc/hibmc_drm_de.c
@@ -96,7 +96,6 @@ static void hibmc_plane_atomic_update(struct drm_plane *plane,
 {
 	struct drm_plane_state	*state	= plane->state;
 	u32 reg;
-	int ret;
 	s64 gpu_addr = 0;
 	unsigned int line_l;
 	struct hibmc_drm_private *priv = plane->dev->dev_private;
@@ -109,16 +108,9 @@ static void hibmc_plane_atomic_update(struct drm_plane *plane,
 	hibmc_fb = to_hibmc_framebuffer(state->fb);
 	gbo = drm_gem_vram_of_gem(hibmc_fb->obj);
 
-	ret = drm_gem_vram_pin(gbo, DRM_GEM_VRAM_PL_FLAG_VRAM);
-	if (ret) {
-		DRM_ERROR("failed to pin bo: %d", ret);
-		return;
-	}
 	gpu_addr = drm_gem_vram_offset(gbo);
-	if (gpu_addr < 0) {
-		drm_gem_vram_unpin(gbo);
-		return;
-	}
+	if (WARN_ON_ONCE(gpu_addr < 0))
+		return; /* Bug: we didn't pin the BO to VRAM in prepare_fb. */
 
 	writel(gpu_addr, priv->mmio + HIBMC_CRT_FB_ADDRESS);
 
@@ -157,6 +149,8 @@ static struct drm_plane_funcs hibmc_plane_funcs = {
 };
 
 static const struct drm_plane_helper_funcs hibmc_plane_helper_funcs = {
+	.prepare_fb	= drm_gem_vram_plane_helper_prepare_fb,
+	.cleanup_fb	= drm_gem_vram_plane_helper_cleanup_fb,
 	.atomic_check = hibmc_plane_atomic_check,
 	.atomic_update = hibmc_plane_atomic_update,
 };
diff --git a/drivers/gpu/drm/hisilicon/hibmc/hibmc_drm_drv.c b/drivers/gpu/drm/hisilicon/hibmc/hibmc_drm_drv.c
index c103005b0a33..2fd4ca91a62d 100644
--- a/drivers/gpu/drm/hisilicon/hibmc/hibmc_drm_drv.c
+++ b/drivers/gpu/drm/hisilicon/hibmc/hibmc_drm_drv.c
@@ -22,15 +22,11 @@
 #include <drm/drm_print.h>
 #include <drm/drm_probe_helper.h>
 #include <drm/drm_vblank.h>
-#include <drm/drm_vram_mm_helper.h>
 
 #include "hibmc_drm_drv.h"
 #include "hibmc_drm_regs.h"
 
-static const struct file_operations hibmc_fops = {
-	.owner		= THIS_MODULE,
-	DRM_VRAM_MM_FILE_OPERATIONS
-};
+DEFINE_DRM_GEM_FOPS(hibmc_fops);
 
 static irqreturn_t hibmc_drm_interrupt(int irq, void *arg)
 {
diff --git a/drivers/gpu/drm/hisilicon/hibmc/hibmc_ttm.c b/drivers/gpu/drm/hisilicon/hibmc/hibmc_ttm.c
index 9f6e473e6295..21b684eab5c9 100644
--- a/drivers/gpu/drm/hisilicon/hibmc/hibmc_ttm.c
+++ b/drivers/gpu/drm/hisilicon/hibmc/hibmc_ttm.c
@@ -17,7 +17,6 @@
 #include <drm/drm_gem.h>
 #include <drm/drm_gem_vram_helper.h>
 #include <drm/drm_print.h>
-#include <drm/drm_vram_mm_helper.h>
 
 #include "hibmc_drm_drv.h"
 
@@ -29,7 +28,7 @@ int hibmc_mm_init(struct hibmc_drm_private *hibmc)
 
 	vmm = drm_vram_helper_alloc_mm(dev,
 				       pci_resource_start(dev->pdev, 0),
-				       hibmc->fb_size, &drm_gem_vram_mm_funcs);
+				       hibmc->fb_size);
 	if (IS_ERR(vmm)) {
 		ret = PTR_ERR(vmm);
 		DRM_ERROR("Error initializing VRAM MM; %d\n", ret);
diff --git a/drivers/gpu/drm/hisilicon/kirin/dw_drm_dsi.c b/drivers/gpu/drm/hisilicon/kirin/dw_drm_dsi.c
index 5bf8138941de..bdcf9c6ae9e9 100644
--- a/drivers/gpu/drm/hisilicon/kirin/dw_drm_dsi.c
+++ b/drivers/gpu/drm/hisilicon/kirin/dw_drm_dsi.c
@@ -18,6 +18,7 @@
 #include <linux/platform_device.h>
 
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_device.h>
 #include <drm/drm_encoder_slave.h>
 #include <drm/drm_mipi_dsi.h>
diff --git a/drivers/gpu/drm/i2c/sil164_drv.c b/drivers/gpu/drm/i2c/sil164_drv.c
index 8bcf0d199145..a839f78a4c8a 100644
--- a/drivers/gpu/drm/i2c/sil164_drv.c
+++ b/drivers/gpu/drm/i2c/sil164_drv.c
@@ -44,7 +44,7 @@ struct sil164_priv {
 	((struct sil164_priv *)to_encoder_slave(x)->slave_priv)
 
 #define sil164_dbg(client, format, ...) do {				\
-		if (drm_debug & DRM_UT_KMS)				\
+		if (drm_debug_enabled(DRM_UT_KMS))			\
 			dev_printk(KERN_DEBUG, &client->dev,		\
 				   "%s: " format, __func__, ## __VA_ARGS__); \
 	} while (0)
diff --git a/drivers/gpu/drm/i2c/tda9950.c b/drivers/gpu/drm/i2c/tda9950.c
index 8039fc0d83db..5b03fdd1eaa4 100644
--- a/drivers/gpu/drm/i2c/tda9950.c
+++ b/drivers/gpu/drm/i2c/tda9950.c
@@ -420,7 +420,8 @@ static int tda9950_probe(struct i2c_client *client,
 		priv->hdmi = glue->parent;
 
 	priv->adap = cec_allocate_adapter(&tda9950_cec_ops, priv, "tda9950",
-					  CEC_CAP_DEFAULTS,
+					  CEC_CAP_DEFAULTS |
+					  CEC_CAP_CONNECTOR_INFO,
 					  CEC_MAX_LOG_ADDRS);
 	if (IS_ERR(priv->adap))
 		return PTR_ERR(priv->adap);
@@ -457,13 +458,14 @@ static int tda9950_probe(struct i2c_client *client,
 	if (ret < 0)
 		return ret;
 
-	priv->notify = cec_notifier_get(priv->hdmi);
+	priv->notify = cec_notifier_cec_adap_register(priv->hdmi, NULL,
+						      priv->adap);
 	if (!priv->notify)
 		return -ENOMEM;
 
 	ret = cec_register_adapter(priv->adap, priv->hdmi);
 	if (ret < 0) {
-		cec_notifier_put(priv->notify);
+		cec_notifier_cec_adap_unregister(priv->notify, priv->adap);
 		return ret;
 	}
 
@@ -473,8 +475,6 @@ static int tda9950_probe(struct i2c_client *client,
 	 */
 	devm_remove_action(dev, tda9950_cec_del, priv);
 
-	cec_register_cec_notifier(priv->adap, priv->notify);
-
 	return 0;
 }
 
@@ -482,8 +482,8 @@ static int tda9950_remove(struct i2c_client *client)
 {
 	struct tda9950_priv *priv = i2c_get_clientdata(client);
 
+	cec_notifier_cec_adap_unregister(priv->notify, priv->adap);
 	cec_unregister_adapter(priv->adap);
-	cec_notifier_put(priv->notify);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/i2c/tda998x_drv.c b/drivers/gpu/drm/i2c/tda998x_drv.c
index 84c6d4c91c65..a63790d32d75 100644
--- a/drivers/gpu/drm/i2c/tda998x_drv.c
+++ b/drivers/gpu/drm/i2c/tda998x_drv.c
@@ -14,6 +14,7 @@
 #include <sound/hdmi-codec.h>
 
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_edid.h>
 #include <drm/drm_of.h>
 #include <drm/drm_print.h>
@@ -805,8 +806,8 @@ static irqreturn_t tda998x_irq_thread(int irq, void *data)
 				tda998x_edid_delay_start(priv);
 			} else {
 				schedule_work(&priv->detect_work);
-				cec_notifier_set_phys_addr(priv->cec_notify,
-						   CEC_PHYS_ADDR_INVALID);
+				cec_notifier_phys_addr_invalidate(
+						priv->cec_notify);
 			}
 
 			handled = true;
@@ -1790,8 +1791,7 @@ static void tda998x_destroy(struct device *dev)
 
 	i2c_unregister_device(priv->cec);
 
-	if (priv->cec_notify)
-		cec_notifier_put(priv->cec_notify);
+	cec_notifier_conn_unregister(priv->cec_notify);
 }
 
 static int tda998x_create(struct device *dev)
@@ -1916,7 +1916,7 @@ static int tda998x_create(struct device *dev)
 		cec_write(priv, REG_CEC_RXSHPDINTENA, CEC_RXSHPDLEV_HPD);
 	}
 
-	priv->cec_notify = cec_notifier_get(dev);
+	priv->cec_notify = cec_notifier_conn_register(dev, NULL, NULL);
 	if (!priv->cec_notify) {
 		ret = -ENOMEM;
 		goto fail;
diff --git a/drivers/gpu/drm/i810/i810_dma.c b/drivers/gpu/drm/i810/i810_dma.c
index 2a77823b8e9a..e66c38332df4 100644
--- a/drivers/gpu/drm/i810/i810_dma.c
+++ b/drivers/gpu/drm/i810/i810_dma.c
@@ -728,7 +728,7 @@ static void i810_dma_dispatch_vertex(struct drm_device *dev,
 	if (nbox > I810_NR_SAREA_CLIPRECTS)
 		nbox = I810_NR_SAREA_CLIPRECTS;
 
-	if (used > 4 * 1024)
+	if (used < 0 || used > 4 * 1024)
 		used = 0;
 
 	if (sarea_priv->dirty)
@@ -1048,7 +1048,7 @@ static void i810_dma_dispatch_mc(struct drm_device *dev, struct drm_buf *buf, in
 	if (u != I810_BUF_CLIENT)
 		DRM_DEBUG("MC found buffer that isn't mine!\n");
 
-	if (used > 4 * 1024)
+	if (used < 0 || used > 4 * 1024)
 		used = 0;
 
 	sarea_priv->dirty = 0x7f;
diff --git a/drivers/gpu/drm/i915/Kconfig b/drivers/gpu/drm/i915/Kconfig
index 0d21402945ab..ba9595960bbe 100644
--- a/drivers/gpu/drm/i915/Kconfig
+++ b/drivers/gpu/drm/i915/Kconfig
@@ -76,7 +76,7 @@ config DRM_I915_CAPTURE_ERROR
 	  This option enables capturing the GPU state when a hang is detected.
 	  This information is vital for triaging hangs and assists in debugging.
 	  Please report any hang to
-            https://bugs.freedesktop.org/enter_bug.cgi?product=DRI
+	    https://bugs.freedesktop.org/enter_bug.cgi?product=DRI
 	  for triaging.
 
 	  If in doubt, say "Y".
@@ -105,11 +105,11 @@ config DRM_I915_USERPTR
 	  If in doubt, say "Y".
 
 config DRM_I915_GVT
-        bool "Enable Intel GVT-g graphics virtualization host support"
-        depends on DRM_I915
-        depends on 64BIT
-        default n
-        help
+	bool "Enable Intel GVT-g graphics virtualization host support"
+	depends on DRM_I915
+	depends on 64BIT
+	default n
+	help
 	  Choose this option if you want to enable Intel GVT-g graphics
 	  virtualization technology host support with integrated graphics.
 	  With GVT-g, it's possible to have one integrated graphics
@@ -148,3 +148,9 @@ menu "drm/i915 Profile Guided Optimisation"
 	depends on DRM_I915
 	source "drivers/gpu/drm/i915/Kconfig.profile"
 endmenu
+
+menu "drm/i915 Unstable Evolution"
+	visible if EXPERT && STAGING && BROKEN
+	depends on DRM_I915
+	source "drivers/gpu/drm/i915/Kconfig.unstable"
+endmenu
diff --git a/drivers/gpu/drm/i915/Kconfig.debug b/drivers/gpu/drm/i915/Kconfig.debug
index 00786a142ff0..438040ff0179 100644
--- a/drivers/gpu/drm/i915/Kconfig.debug
+++ b/drivers/gpu/drm/i915/Kconfig.debug
@@ -1,34 +1,32 @@
 # SPDX-License-Identifier: GPL-2.0-only
 config DRM_I915_WERROR
-        bool "Force GCC to throw an error instead of a warning when compiling"
-        # As this may inadvertently break the build, only allow the user
-        # to shoot oneself in the foot iff they aim really hard
-        depends on EXPERT
-        # We use the dependency on !COMPILE_TEST to not be enabled in
-        # allmodconfig or allyesconfig configurations
-        depends on !COMPILE_TEST
-	select HEADER_TEST
-        default n
-        help
-          Add -Werror to the build flags for (and only for) i915.ko.
-          Do not enable this unless you are writing code for the i915.ko module.
-
-          Recommended for driver developers only.
-
-          If in doubt, say "N".
+	bool "Force GCC to throw an error instead of a warning when compiling"
+	# As this may inadvertently break the build, only allow the user
+	# to shoot oneself in the foot iff they aim really hard
+	depends on EXPERT
+	# We use the dependency on !COMPILE_TEST to not be enabled in
+	# allmodconfig or allyesconfig configurations
+	depends on !COMPILE_TEST
+	default n
+	help
+	  Add -Werror to the build flags for (and only for) i915.ko.
+	  Do not enable this unless you are writing code for the i915.ko module.
+
+	  Recommended for driver developers only.
+
+	  If in doubt, say "N".
 
 config DRM_I915_DEBUG
-        bool "Enable additional driver debugging"
-        depends on DRM_I915
-        select DEBUG_FS
-        select PREEMPT_COUNT
-        select REFCOUNT_FULL
-        select I2C_CHARDEV
-        select STACKDEPOT
-        select DRM_DP_AUX_CHARDEV
-        select X86_MSR # used by igt/pm_rpm
-        select DRM_VGEM # used by igt/prime_vgem (dmabuf interop checks)
-        select DRM_DEBUG_MM if DRM=y
+	bool "Enable additional driver debugging"
+	depends on DRM_I915
+	select DEBUG_FS
+	select PREEMPT_COUNT
+	select I2C_CHARDEV
+	select STACKDEPOT
+	select DRM_DP_AUX_CHARDEV
+	select X86_MSR # used by igt/pm_rpm
+	select DRM_VGEM # used by igt/prime_vgem (dmabuf interop checks)
+	select DRM_DEBUG_MM if DRM=y
 	select DRM_DEBUG_SELFTEST
 	select DMABUF_SELFTESTS
 	select SW_SYNC # signaling validation framework (igt/syncobj*)
@@ -36,14 +34,14 @@ config DRM_I915_DEBUG
 	select DRM_I915_SELFTEST
 	select DRM_I915_DEBUG_RUNTIME_PM
 	select DRM_I915_DEBUG_MMIO
-        default n
-        help
-          Choose this option to turn on extra driver debugging that may affect
-          performance but will catch some internal issues.
+	default n
+	help
+	  Choose this option to turn on extra driver debugging that may affect
+	  performance but will catch some internal issues.
 
-          Recommended for driver developers only.
+	  Recommended for driver developers only.
 
-          If in doubt, say "N".
+	  If in doubt, say "N".
 
 config DRM_I915_DEBUG_MMIO
 	bool "Always insert extra checks around mmio access by default"
@@ -59,16 +57,16 @@ config DRM_I915_DEBUG_MMIO
 	  If in doubt, say "N".
 
 config DRM_I915_DEBUG_GEM
-        bool "Insert extra checks into the GEM internals"
-        default n
-        depends on DRM_I915_WERROR
-        help
-          Enable extra sanity checks (including BUGs) along the GEM driver
-          paths that may slow the system down and if hit hang the machine.
+	bool "Insert extra checks into the GEM internals"
+	default n
+	depends on DRM_I915_WERROR
+	help
+	  Enable extra sanity checks (including BUGs) along the GEM driver
+	  paths that may slow the system down and if hit hang the machine.
 
-          Recommended for driver developers only.
+	  Recommended for driver developers only.
 
-          If in doubt, say "N".
+	  If in doubt, say "N".
 
 config DRM_I915_ERRLOG_GEM
 	bool "Insert extra logging (very verbose) for common GEM errors"
@@ -111,41 +109,41 @@ config DRM_I915_TRACE_GTT
 	  If in doubt, say "N".
 
 config DRM_I915_SW_FENCE_DEBUG_OBJECTS
-        bool "Enable additional driver debugging for fence objects"
-        depends on DRM_I915
-        select DEBUG_OBJECTS
-        default n
-        help
-          Choose this option to turn on extra driver debugging that may affect
-          performance but will catch some internal issues.
+	bool "Enable additional driver debugging for fence objects"
+	depends on DRM_I915
+	select DEBUG_OBJECTS
+	default n
+	help
+	  Choose this option to turn on extra driver debugging that may affect
+	  performance but will catch some internal issues.
 
-          Recommended for driver developers only.
+	  Recommended for driver developers only.
 
-          If in doubt, say "N".
+	  If in doubt, say "N".
 
 config DRM_I915_SW_FENCE_CHECK_DAG
-        bool "Enable additional driver debugging for detecting dependency cycles"
-        depends on DRM_I915
-        default n
-        help
-          Choose this option to turn on extra driver debugging that may affect
-          performance but will catch some internal issues.
+	bool "Enable additional driver debugging for detecting dependency cycles"
+	depends on DRM_I915
+	default n
+	help
+	  Choose this option to turn on extra driver debugging that may affect
+	  performance but will catch some internal issues.
 
-          Recommended for driver developers only.
+	  Recommended for driver developers only.
 
-          If in doubt, say "N".
+	  If in doubt, say "N".
 
 config DRM_I915_DEBUG_GUC
-        bool "Enable additional driver debugging for GuC"
-        depends on DRM_I915
-        default n
-        help
-          Choose this option to turn on extra driver debugging that may affect
-          performance but will help resolve GuC related issues.
+	bool "Enable additional driver debugging for GuC"
+	depends on DRM_I915
+	default n
+	help
+	  Choose this option to turn on extra driver debugging that may affect
+	  performance but will help resolve GuC related issues.
 
-          Recommended for driver developers only.
+	  Recommended for driver developers only.
 
-          If in doubt, say "N".
+	  If in doubt, say "N".
 
 config DRM_I915_SELFTEST
 	bool "Enable selftests upon driver load"
@@ -178,15 +176,15 @@ config DRM_I915_SELFTEST_BROKEN
 	  If in doubt, say "N".
 
 config DRM_I915_LOW_LEVEL_TRACEPOINTS
-        bool "Enable low level request tracing events"
-        depends on DRM_I915
-        default n
-        help
-          Choose this option to turn on low level request tracing events.
-          This provides the ability to precisely monitor engine utilisation
-          and also analyze the request dependency resolving timeline.
-
-          If in doubt, say "N".
+	bool "Enable low level request tracing events"
+	depends on DRM_I915
+	default n
+	help
+	  Choose this option to turn on low level request tracing events.
+	  This provides the ability to precisely monitor engine utilisation
+	  and also analyze the request dependency resolving timeline.
+
+	  If in doubt, say "N".
 
 config DRM_I915_DEBUG_VBLANK_EVADE
 	bool "Enable extra debug warnings for vblank evasion"
diff --git a/drivers/gpu/drm/i915/Kconfig.profile b/drivers/gpu/drm/i915/Kconfig.profile
index 48df8889a88a..1799537a3228 100644
--- a/drivers/gpu/drm/i915/Kconfig.profile
+++ b/drivers/gpu/drm/i915/Kconfig.profile
@@ -12,6 +12,29 @@ config DRM_I915_USERFAULT_AUTOSUSPEND
 	  May be 0 to disable the extra delay and solely use the device level
 	  runtime pm autosuspend delay tunable.
 
+config DRM_I915_HEARTBEAT_INTERVAL
+	int "Interval between heartbeat pulses (ms)"
+	default 2500 # milliseconds
+	help
+	  The driver sends a periodic heartbeat down all active engines to
+	  check the health of the GPU and undertake regular house-keeping of
+	  internal driver state.
+
+	  May be 0 to disable heartbeats and therefore disable automatic GPU
+	  hang detection.
+
+config DRM_I915_PREEMPT_TIMEOUT
+	int "Preempt timeout (ms, jiffy granularity)"
+	default 100 # milliseconds
+	help
+	  How long to wait (in milliseconds) for a preemption event to occur
+	  when submitting a new context via execlists. If the current context
+	  does not hit an arbitration point and yield to HW before the timer
+	  expires, the HW will be reset to allow the more important context
+	  to execute.
+
+	  May be 0 to disable the timeout.
+
 config DRM_I915_SPIN_REQUEST
 	int "Busywait for request completion (us)"
 	default 5 # microseconds
@@ -25,3 +48,29 @@ config DRM_I915_SPIN_REQUEST
 	  May be 0 to disable the initial spin. In practice, we estimate
 	  the cost of enabling the interrupt (if currently disabled) to be
 	  a few microseconds.
+
+config DRM_I915_STOP_TIMEOUT
+	int "How long to wait for an engine to quiesce gracefully before reset (ms)"
+	default 100 # milliseconds
+	help
+	  By stopping submission and sleeping for a short time before resetting
+	  the GPU, we allow the innocent contexts also on the system to quiesce.
+	  It is then less likely for a hanging context to cause collateral
+	  damage as the system is reset in order to recover. The corollary is
+	  that the reset itself may take longer and so be more disruptive to
+	  interactive or low latency workloads.
+
+config DRM_I915_TIMESLICE_DURATION
+	int "Scheduling quantum for userspace batches (ms, jiffy granularity)"
+	default 1 # milliseconds
+	help
+	  When two user batches of equal priority are executing, we will
+	  alternate execution of each batch to ensure forward progress of
+	  all users. This is necessary in some cases where there may be
+	  an implicit dependency between those batches that requires
+	  concurrent execution in order for them to proceed, e.g. they
+	  interact with each other via userspace semaphores. Each context
+	  is scheduled for execution for the timeslice duration, before
+	  switching to the next context.
+
+	  May be 0 to disable timeslicing.
diff --git a/drivers/gpu/drm/i915/Kconfig.unstable b/drivers/gpu/drm/i915/Kconfig.unstable
new file mode 100644
index 000000000000..0c2276155c2b
--- /dev/null
+++ b/drivers/gpu/drm/i915/Kconfig.unstable
@@ -0,0 +1,29 @@
+# SPDX-License-Identifier: GPL-2.0-only
+config DRM_I915_UNSTABLE
+	bool "Enable unstable API for early prototype development"
+	depends on EXPERT
+	depends on STAGING
+	depends on BROKEN # should never be enabled by distros!
+	# We use the dependency on !COMPILE_TEST to not be enabled in
+	# allmodconfig or allyesconfig configurations
+	depends on !COMPILE_TEST
+	default n
+	help
+	  Enable prototype uAPI under general discussion before they are
+	  finalized. Such prototypes may be withdrawn or substantially
+	  changed before release. They are only enabled here so that a wide
+	  number of interested parties (userspace driver developers) can
+	  verify that the uAPI meet their expectations. These uAPI should
+	  never be used in production.
+
+	  Recommended for driver developers _only_.
+
+	  If in the slightest bit of doubt, say "N".
+
+config DRM_I915_UNSTABLE_FAKE_LMEM
+	bool "Enable the experimental fake lmem"
+	depends on DRM_I915_UNSTABLE
+	default n
+	help
+	  Convert some system memory into a fake local memory region for
+	  testing.
diff --git a/drivers/gpu/drm/i915/Makefile b/drivers/gpu/drm/i915/Makefile
index 2587ea834f06..90dcf09f52cc 100644
--- a/drivers/gpu/drm/i915/Makefile
+++ b/drivers/gpu/drm/i915/Makefile
@@ -46,10 +46,12 @@ i915-y += i915_drv.o \
 	  i915_pci.o \
 	  i915_scatterlist.o \
 	  i915_suspend.o \
+	  i915_switcheroo.o \
 	  i915_sysfs.o \
 	  i915_utils.o \
 	  intel_csr.o \
 	  intel_device_info.o \
+	  intel_memory_region.o \
 	  intel_pch.o \
 	  intel_pm.o \
 	  intel_runtime_pm.o \
@@ -76,19 +78,24 @@ gt-y += \
 	gt/intel_breadcrumbs.o \
 	gt/intel_context.o \
 	gt/intel_engine_cs.o \
-	gt/intel_engine_pool.o \
+	gt/intel_engine_heartbeat.o \
 	gt/intel_engine_pm.o \
+	gt/intel_engine_pool.o \
 	gt/intel_engine_user.o \
 	gt/intel_gt.o \
 	gt/intel_gt_irq.o \
 	gt/intel_gt_pm.o \
 	gt/intel_gt_pm_irq.o \
-	gt/intel_hangcheck.o \
+	gt/intel_gt_requests.o \
+	gt/intel_llc.o \
 	gt/intel_lrc.o \
+	gt/intel_mocs.o \
+	gt/intel_rc6.o \
 	gt/intel_renderstate.o \
 	gt/intel_reset.o \
-	gt/intel_ringbuffer.o \
-	gt/intel_mocs.o \
+	gt/intel_ring.o \
+	gt/intel_ring_submission.o \
+	gt/intel_rps.o \
 	gt/intel_sseu.o \
 	gt/intel_timeline.o \
 	gt/intel_workarounds.o
@@ -114,10 +121,12 @@ gem-y += \
 	gem/i915_gem_internal.o \
 	gem/i915_gem_object.o \
 	gem/i915_gem_object_blt.o \
+	gem/i915_gem_lmem.o \
 	gem/i915_gem_mman.o \
 	gem/i915_gem_pages.o \
 	gem/i915_gem_phys.o \
 	gem/i915_gem_pm.o \
+	gem/i915_gem_region.o \
 	gem/i915_gem_shmem.o \
 	gem/i915_gem_shrinker.o \
 	gem/i915_gem_stolen.o \
@@ -141,6 +150,7 @@ i915-y += \
 	  i915_scheduler.o \
 	  i915_trace_points.o \
 	  i915_vma.o \
+	  intel_region_lmem.o \
 	  intel_wopcm.o
 
 # general-purpose microcontroller (GuC) support
@@ -172,6 +182,7 @@ i915-y += \
 	display/intel_display_power.o \
 	display/intel_dpio_phy.o \
 	display/intel_dpll_mgr.o \
+	display/intel_dsb.o \
 	display/intel_fbc.o \
 	display/intel_fifo_underrun.o \
 	display/intel_frontbuffer.o \
@@ -182,7 +193,8 @@ i915-y += \
 	display/intel_psr.o \
 	display/intel_quirks.o \
 	display/intel_sprite.o \
-	display/intel_tc.o
+	display/intel_tc.o \
+	display/intel_vga.o
 i915-$(CONFIG_ACPI) += \
 	display/intel_acpi.o \
 	display/intel_opregion.o
@@ -235,7 +247,8 @@ i915-y += \
 	oa/i915_oa_cflgt2.o \
 	oa/i915_oa_cflgt3.o \
 	oa/i915_oa_cnl.o \
-	oa/i915_oa_icl.o
+	oa/i915_oa_icl.o \
+	oa/i915_oa_tgl.o
 i915-y += i915_perf.o
 
 # Post-mortem debug and GPU hang state capture
diff --git a/drivers/gpu/drm/i915/display/icl_dsi.c b/drivers/gpu/drm/i915/display/icl_dsi.c
index 6e398c33a524..325df29b0447 100644
--- a/drivers/gpu/drm/i915/display/icl_dsi.c
+++ b/drivers/gpu/drm/i915/display/icl_dsi.c
@@ -1584,7 +1584,7 @@ void icl_dsi_init(struct drm_i915_private *dev_priv)
 	encoder->get_hw_state = gen11_dsi_get_hw_state;
 	encoder->type = INTEL_OUTPUT_DSI;
 	encoder->cloneable = 0;
-	encoder->crtc_mask = BIT(PIPE_A) | BIT(PIPE_B) | BIT(PIPE_C);
+	encoder->pipe_mask = ~0;
 	encoder->power_domain = POWER_DOMAIN_PORT_DSI;
 	encoder->get_power_domains = gen11_dsi_get_power_domains;
 
diff --git a/drivers/gpu/drm/i915/display/intel_atomic.c b/drivers/gpu/drm/i915/display/intel_atomic.c
index d3fb75bb9eb1..c2875b10adf9 100644
--- a/drivers/gpu/drm/i915/display/intel_atomic.c
+++ b/drivers/gpu/drm/i915/display/intel_atomic.c
@@ -199,8 +199,8 @@ intel_crtc_duplicate_state(struct drm_crtc *crtc)
 	crtc_state->disable_cxsr = false;
 	crtc_state->update_wm_pre = false;
 	crtc_state->update_wm_post = false;
-	crtc_state->fb_changed = false;
 	crtc_state->fifo_changed = false;
+	crtc_state->preload_luts = false;
 	crtc_state->wm.need_postvbl_update = false;
 	crtc_state->fb_bits = 0;
 	crtc_state->update_planes = 0;
@@ -264,10 +264,13 @@ static void intel_atomic_setup_scaler(struct intel_crtc_scaler_state *scaler_sta
 			 */
 			mode = PS_SCALER_MODE_NORMAL;
 		} else {
+			struct intel_plane *linked =
+				plane_state->planar_linked_plane;
+
 			mode = PS_SCALER_MODE_PLANAR;
 
-			if (plane_state->linked_plane)
-				mode |= PS_PLANE_Y_SEL(plane_state->linked_plane->id);
+			if (linked)
+				mode |= PS_PLANE_Y_SEL(linked->id);
 		}
 	} else if (INTEL_GEN(dev_priv) > 9 || IS_GEMINILAKE(dev_priv)) {
 		mode = PS_SCALER_MODE_NORMAL;
@@ -371,6 +374,15 @@ int intel_atomic_setup_scalers(struct drm_i915_private *dev_priv,
 			 */
 			if (!plane) {
 				struct drm_plane_state *state;
+
+				/*
+				 * GLK+ scalers don't have a HQ mode so it
+				 * isn't necessary to change between HQ and dyn mode
+				 * on those platforms.
+				 */
+				if (INTEL_GEN(dev_priv) >= 10 || IS_GEMINILAKE(dev_priv))
+					continue;
+
 				plane = drm_plane_from_index(&dev_priv->drm, i);
 				state = drm_atomic_get_plane_state(drm_state, plane);
 				if (IS_ERR(state)) {
@@ -378,13 +390,6 @@ int intel_atomic_setup_scalers(struct drm_i915_private *dev_priv,
 						plane->base.id);
 					return PTR_ERR(state);
 				}
-
-				/*
-				 * the plane is added after plane checks are run,
-				 * but since this plane is unchanged just do the
-				 * minimum required validation.
-				 */
-				crtc_state->base.planes_changed = true;
 			}
 
 			intel_plane = to_intel_plane(plane);
@@ -425,6 +430,13 @@ void intel_atomic_state_clear(struct drm_atomic_state *s)
 	struct intel_atomic_state *state = to_intel_atomic_state(s);
 	drm_atomic_state_default_clear(&state->base);
 	state->dpll_set = state->modeset = false;
+	state->global_state_changed = false;
+	state->active_pipes = 0;
+	memset(&state->min_cdclk, 0, sizeof(state->min_cdclk));
+	memset(&state->min_voltage_level, 0, sizeof(state->min_voltage_level));
+	memset(&state->cdclk.logical, 0, sizeof(state->cdclk.logical));
+	memset(&state->cdclk.actual, 0, sizeof(state->cdclk.actual));
+	state->cdclk.pipe = INVALID_PIPE;
 }
 
 struct intel_crtc_state *
@@ -438,3 +450,40 @@ intel_atomic_get_crtc_state(struct drm_atomic_state *state,
 
 	return to_intel_crtc_state(crtc_state);
 }
+
+int intel_atomic_lock_global_state(struct intel_atomic_state *state)
+{
+	struct drm_i915_private *dev_priv = to_i915(state->base.dev);
+	struct intel_crtc *crtc;
+
+	state->global_state_changed = true;
+
+	for_each_intel_crtc(&dev_priv->drm, crtc) {
+		int ret;
+
+		ret = drm_modeset_lock(&crtc->base.mutex,
+				       state->base.acquire_ctx);
+		if (ret)
+			return ret;
+	}
+
+	return 0;
+}
+
+int intel_atomic_serialize_global_state(struct intel_atomic_state *state)
+{
+	struct drm_i915_private *dev_priv = to_i915(state->base.dev);
+	struct intel_crtc *crtc;
+
+	state->global_state_changed = true;
+
+	for_each_intel_crtc(&dev_priv->drm, crtc) {
+		struct intel_crtc_state *crtc_state;
+
+		crtc_state = intel_atomic_get_crtc_state(&state->base, crtc);
+		if (IS_ERR(crtc_state))
+			return PTR_ERR(crtc_state);
+	}
+
+	return 0;
+}
diff --git a/drivers/gpu/drm/i915/display/intel_atomic.h b/drivers/gpu/drm/i915/display/intel_atomic.h
index 58065d3161a3..49d5cb1b9e0a 100644
--- a/drivers/gpu/drm/i915/display/intel_atomic.h
+++ b/drivers/gpu/drm/i915/display/intel_atomic.h
@@ -16,6 +16,7 @@ struct drm_crtc_state;
 struct drm_device;
 struct drm_i915_private;
 struct drm_property;
+struct intel_atomic_state;
 struct intel_crtc;
 struct intel_crtc_state;
 
@@ -46,4 +47,8 @@ int intel_atomic_setup_scalers(struct drm_i915_private *dev_priv,
 			       struct intel_crtc *intel_crtc,
 			       struct intel_crtc_state *crtc_state);
 
+int intel_atomic_lock_global_state(struct intel_atomic_state *state);
+
+int intel_atomic_serialize_global_state(struct intel_atomic_state *state);
+
 #endif /* __INTEL_ATOMIC_H__ */
diff --git a/drivers/gpu/drm/i915/display/intel_atomic_plane.c b/drivers/gpu/drm/i915/display/intel_atomic_plane.c
index d1fcdf206da4..98f557a9f8ee 100644
--- a/drivers/gpu/drm/i915/display/intel_atomic_plane.c
+++ b/drivers/gpu/drm/i915/display/intel_atomic_plane.c
@@ -138,18 +138,58 @@ unsigned int intel_plane_data_rate(const struct intel_crtc_state *crtc_state,
 	return cpp * crtc_state->pixel_rate;
 }
 
+bool intel_plane_calc_min_cdclk(struct intel_atomic_state *state,
+				struct intel_plane *plane)
+{
+	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
+	const struct intel_plane_state *plane_state =
+		intel_atomic_get_new_plane_state(state, plane);
+	struct intel_crtc *crtc = to_intel_crtc(plane_state->base.crtc);
+	struct intel_crtc_state *crtc_state;
+
+	if (!plane_state->base.visible || !plane->min_cdclk)
+		return false;
+
+	crtc_state = intel_atomic_get_new_crtc_state(state, crtc);
+
+	crtc_state->min_cdclk[plane->id] =
+		plane->min_cdclk(crtc_state, plane_state);
+
+	/*
+	 * Does the cdclk need to be bumbed up?
+	 *
+	 * Note: we obviously need to be called before the new
+	 * cdclk frequency is calculated so state->cdclk.logical
+	 * hasn't been populated yet. Hence we look at the old
+	 * cdclk state under dev_priv->cdclk.logical. This is
+	 * safe as long we hold at least one crtc mutex (which
+	 * must be true since we have crtc_state).
+	 */
+	if (crtc_state->min_cdclk[plane->id] > dev_priv->cdclk.logical.cdclk) {
+		DRM_DEBUG_KMS("[PLANE:%d:%s] min_cdclk (%d kHz) > logical cdclk (%d kHz)\n",
+			      plane->base.base.id, plane->base.name,
+			      crtc_state->min_cdclk[plane->id],
+			      dev_priv->cdclk.logical.cdclk);
+		return true;
+	}
+
+	return false;
+}
+
 int intel_plane_atomic_check_with_state(const struct intel_crtc_state *old_crtc_state,
 					struct intel_crtc_state *new_crtc_state,
 					const struct intel_plane_state *old_plane_state,
 					struct intel_plane_state *new_plane_state)
 {
 	struct intel_plane *plane = to_intel_plane(new_plane_state->base.plane);
+	const struct drm_framebuffer *fb = new_plane_state->base.fb;
 	int ret;
 
 	new_crtc_state->active_planes &= ~BIT(plane->id);
 	new_crtc_state->nv12_planes &= ~BIT(plane->id);
 	new_crtc_state->c8_planes &= ~BIT(plane->id);
 	new_crtc_state->data_rate[plane->id] = 0;
+	new_crtc_state->min_cdclk[plane->id] = 0;
 	new_plane_state->base.visible = false;
 
 	if (!new_plane_state->base.crtc && !old_plane_state->base.crtc)
@@ -164,11 +204,11 @@ int intel_plane_atomic_check_with_state(const struct intel_crtc_state *old_crtc_
 		new_crtc_state->active_planes |= BIT(plane->id);
 
 	if (new_plane_state->base.visible &&
-	    is_planar_yuv_format(new_plane_state->base.fb->format->format))
+	    drm_format_info_is_yuv_semiplanar(fb->format))
 		new_crtc_state->nv12_planes |= BIT(plane->id);
 
 	if (new_plane_state->base.visible &&
-	    new_plane_state->base.fb->format->format == DRM_FORMAT_C8)
+	    fb->format->format == DRM_FORMAT_C8)
 		new_crtc_state->c8_planes |= BIT(plane->id);
 
 	if (new_plane_state->base.visible || old_plane_state->base.visible)
@@ -194,14 +234,11 @@ get_crtc_from_states(const struct intel_plane_state *old_plane_state,
 	return NULL;
 }
 
-static int intel_plane_atomic_check(struct drm_plane *_plane,
-				    struct drm_plane_state *_new_plane_state)
+int intel_plane_atomic_check(struct intel_atomic_state *state,
+			     struct intel_plane *plane)
 {
-	struct intel_plane *plane = to_intel_plane(_plane);
-	struct intel_atomic_state *state =
-		to_intel_atomic_state(_new_plane_state->state);
 	struct intel_plane_state *new_plane_state =
-		to_intel_plane_state(_new_plane_state);
+		intel_atomic_get_new_plane_state(state, plane);
 	const struct intel_plane_state *old_plane_state =
 		intel_atomic_get_old_plane_state(state, plane);
 	struct intel_crtc *crtc =
@@ -320,9 +357,9 @@ void skl_update_planes_on_crtc(struct intel_atomic_state *state,
 
 		if (new_plane_state->base.visible) {
 			intel_update_plane(plane, new_crtc_state, new_plane_state);
-		} else if (new_plane_state->slave) {
+		} else if (new_plane_state->planar_slave) {
 			struct intel_plane *master =
-				new_plane_state->linked_plane;
+				new_plane_state->planar_linked_plane;
 
 			/*
 			 * We update the slave plane from this function because
@@ -368,5 +405,4 @@ void i9xx_update_planes_on_crtc(struct intel_atomic_state *state,
 const struct drm_plane_helper_funcs intel_plane_helper_funcs = {
 	.prepare_fb = intel_prepare_plane_fb,
 	.cleanup_fb = intel_cleanup_plane_fb,
-	.atomic_check = intel_plane_atomic_check,
 };
diff --git a/drivers/gpu/drm/i915/display/intel_atomic_plane.h b/drivers/gpu/drm/i915/display/intel_atomic_plane.h
index cb7ef4f9eafd..e61e9a82aadf 100644
--- a/drivers/gpu/drm/i915/display/intel_atomic_plane.h
+++ b/drivers/gpu/drm/i915/display/intel_atomic_plane.h
@@ -41,9 +41,13 @@ int intel_plane_atomic_check_with_state(const struct intel_crtc_state *old_crtc_
 					struct intel_crtc_state *crtc_state,
 					const struct intel_plane_state *old_plane_state,
 					struct intel_plane_state *intel_state);
+int intel_plane_atomic_check(struct intel_atomic_state *state,
+			     struct intel_plane *plane);
 int intel_plane_atomic_calc_changes(const struct intel_crtc_state *old_crtc_state,
 				    struct intel_crtc_state *crtc_state,
 				    const struct intel_plane_state *old_plane_state,
 				    struct intel_plane_state *plane_state);
+bool intel_plane_calc_min_cdclk(struct intel_atomic_state *state,
+				struct intel_plane *plane);
 
 #endif /* __INTEL_ATOMIC_PLANE_H__ */
diff --git a/drivers/gpu/drm/i915/display/intel_audio.c b/drivers/gpu/drm/i915/display/intel_audio.c
index ddcccf4408c3..85e6b2bbb34f 100644
--- a/drivers/gpu/drm/i915/display/intel_audio.c
+++ b/drivers/gpu/drm/i915/display/intel_audio.c
@@ -28,6 +28,7 @@
 #include <drm/i915_component.h>
 
 #include "i915_drv.h"
+#include "intel_atomic.h"
 #include "intel_audio.h"
 #include "intel_display_types.h"
 #include "intel_lpe_audio.h"
@@ -560,8 +561,9 @@ static void ilk_audio_codec_disable(struct intel_encoder *encoder,
 	u32 tmp, eldv;
 	i915_reg_t aud_config, aud_cntrl_st2;
 
-	DRM_DEBUG_KMS("Disable audio codec on port %c, pipe %c\n",
-		      port_name(port), pipe_name(pipe));
+	DRM_DEBUG_KMS("Disable audio codec on [ENCODER:%d:%s], pipe %c\n",
+		      encoder->base.base.id, encoder->base.name,
+		      pipe_name(pipe));
 
 	if (WARN_ON(port == PORT_A))
 		return;
@@ -609,8 +611,9 @@ static void ilk_audio_codec_enable(struct intel_encoder *encoder,
 	int len, i;
 	i915_reg_t hdmiw_hdmiedid, aud_config, aud_cntl_st, aud_cntrl_st2;
 
-	DRM_DEBUG_KMS("Enable audio codec on port %c, pipe %c, %u bytes ELD\n",
-		      port_name(port), pipe_name(pipe), drm_eld_size(eld));
+	DRM_DEBUG_KMS("Enable audio codec on [ENCODER:%d:%s], pipe %c, %u bytes ELD\n",
+		      encoder->base.base.id, encoder->base.name,
+		      pipe_name(pipe), drm_eld_size(eld));
 
 	if (WARN_ON(port == PORT_A))
 		return;
@@ -816,13 +819,8 @@ retry:
 	to_intel_atomic_state(state)->cdclk.force_min_cdclk =
 		enable ? 2 * 96000 : 0;
 
-	/*
-	 * Protects dev_priv->cdclk.force_min_cdclk
-	 * Need to lock this here in case we have no active pipes
-	 * and thus wouldn't lock it during the commit otherwise.
-	 */
-	ret = drm_modeset_lock(&dev_priv->drm.mode_config.connection_mutex,
-			       &ctx);
+	/* Protects dev_priv->cdclk.force_min_cdclk */
+	ret = intel_atomic_lock_global_state(to_intel_atomic_state(state));
 	if (!ret)
 		ret = drm_atomic_commit(state);
 
@@ -850,11 +848,23 @@ static unsigned long i915_audio_component_get_power(struct device *kdev)
 
 	ret = intel_display_power_get(dev_priv, POWER_DOMAIN_AUDIO);
 
-	/* Force CDCLK to 2*BCLK as long as we need audio to be powered. */
-	if (dev_priv->audio_power_refcount++ == 0)
-		if (IS_CANNONLAKE(dev_priv) || IS_GEMINILAKE(dev_priv))
+	if (dev_priv->audio_power_refcount++ == 0) {
+		if (IS_TIGERLAKE(dev_priv) || IS_ICELAKE(dev_priv)) {
+			I915_WRITE(AUD_FREQ_CNTRL, dev_priv->audio_freq_cntrl);
+			DRM_DEBUG_KMS("restored AUD_FREQ_CNTRL to 0x%x\n",
+				      dev_priv->audio_freq_cntrl);
+		}
+
+		/* Force CDCLK to 2*BCLK as long as we need audio powered. */
+		if (INTEL_GEN(dev_priv) >= 10 || IS_GEMINILAKE(dev_priv))
 			glk_force_audio_cdclk(dev_priv, true);
 
+		if (INTEL_GEN(dev_priv) >= 10 || IS_GEMINILAKE(dev_priv))
+			I915_WRITE(AUD_PIN_BUF_CTL,
+				   (I915_READ(AUD_PIN_BUF_CTL) |
+				    AUD_PIN_BUF_ENABLE));
+	}
+
 	return ret;
 }
 
@@ -865,7 +875,7 @@ static void i915_audio_component_put_power(struct device *kdev,
 
 	/* Stop forcing CDCLK to 2*BCLK if no need for audio to be powered. */
 	if (--dev_priv->audio_power_refcount == 0)
-		if (IS_CANNONLAKE(dev_priv) || IS_GEMINILAKE(dev_priv))
+		if (INTEL_GEN(dev_priv) >= 10 || IS_GEMINILAKE(dev_priv))
 			glk_force_audio_cdclk(dev_priv, false);
 
 	intel_display_power_put(dev_priv, POWER_DOMAIN_AUDIO, cookie);
@@ -1114,6 +1124,12 @@ static void i915_audio_component_init(struct drm_i915_private *dev_priv)
 		return;
 	}
 
+	if (IS_TIGERLAKE(dev_priv) || IS_ICELAKE(dev_priv)) {
+		dev_priv->audio_freq_cntrl = I915_READ(AUD_FREQ_CNTRL);
+		DRM_DEBUG_KMS("init value of AUD_FREQ_CNTRL of 0x%x\n",
+			      dev_priv->audio_freq_cntrl);
+	}
+
 	dev_priv->audio_component_registered = true;
 }
 
diff --git a/drivers/gpu/drm/i915/display/intel_bios.c b/drivers/gpu/drm/i915/display/intel_bios.c
index efb39f350b19..63c1bd4c2954 100644
--- a/drivers/gpu/drm/i915/display/intel_bios.c
+++ b/drivers/gpu/drm/i915/display/intel_bios.c
@@ -1270,7 +1270,7 @@ static void sanitize_ddc_pin(struct drm_i915_private *dev_priv,
 		DRM_DEBUG_KMS("port %c trying to use the same DDC pin (0x%x) as port %c, "
 			      "disabling port %c DVI/HDMI support\n",
 			      port_name(port), info->alternate_ddc_pin,
-			      port_name(p), port_name(port));
+			      port_name(p), port_name(p));
 
 		/*
 		 * If we have multiple ports supposedly sharing the
@@ -1278,9 +1278,14 @@ static void sanitize_ddc_pin(struct drm_i915_private *dev_priv,
 		 * port. Otherwise they share the same ddc bin and
 		 * system couldn't communicate with them separately.
 		 *
-		 * Give child device order the priority, first come first
-		 * served.
+		 * Give inverse child device order the priority,
+		 * last one wins. Yes, there are real machines
+		 * (eg. Asrock B250M-HDV) where VBT has both
+		 * port A and port E with the same AUX ch and
+		 * we must pick port E :(
 		 */
+		info = &dev_priv->vbt.ddi_port_info[p];
+
 		info->supports_dvi = false;
 		info->supports_hdmi = false;
 		info->alternate_ddc_pin = 0;
@@ -1316,7 +1321,7 @@ static void sanitize_aux_ch(struct drm_i915_private *dev_priv,
 		DRM_DEBUG_KMS("port %c trying to use the same AUX CH (0x%x) as port %c, "
 			      "disabling port %c DP support\n",
 			      port_name(port), info->alternate_aux_channel,
-			      port_name(p), port_name(port));
+			      port_name(p), port_name(p));
 
 		/*
 		 * If we have multiple ports supposedlt sharing the
@@ -1324,9 +1329,14 @@ static void sanitize_aux_ch(struct drm_i915_private *dev_priv,
 		 * port. Otherwise they share the same aux channel
 		 * and system couldn't communicate with them separately.
 		 *
-		 * Give child device order the priority, first come first
-		 * served.
+		 * Give inverse child device order the priority,
+		 * last one wins. Yes, there are real machines
+		 * (eg. Asrock B250M-HDV) where VBT has both
+		 * port A and port E with the same AUX ch and
+		 * we must pick port E :(
 		 */
+		info = &dev_priv->vbt.ddi_port_info[p];
+
 		info->supports_dp = false;
 		info->alternate_aux_channel = 0;
 	}
@@ -1389,6 +1399,7 @@ static enum port dvo_port_to_port(u8 dvo_port)
 		[PORT_D] = { DVO_PORT_HDMID, DVO_PORT_DPD, -1},
 		[PORT_E] = { DVO_PORT_CRT, DVO_PORT_HDMIE, DVO_PORT_DPE},
 		[PORT_F] = { DVO_PORT_HDMIF, DVO_PORT_DPF, -1},
+		[PORT_G] = { DVO_PORT_HDMIG, DVO_PORT_DPG, -1},
 	};
 	enum port port;
 	int i;
@@ -1615,7 +1626,7 @@ parse_general_definitions(struct drm_i915_private *dev_priv,
 		expected_size = 37;
 	} else if (bdb->version <= 215) {
 		expected_size = 38;
-	} else if (bdb->version <= 216) {
+	} else if (bdb->version <= 229) {
 		expected_size = 39;
 	} else {
 		expected_size = sizeof(*child);
@@ -1833,7 +1844,7 @@ void intel_bios_init(struct drm_i915_private *dev_priv)
 	const struct bdb_header *bdb;
 	u8 __iomem *bios = NULL;
 
-	if (!HAS_DISPLAY(dev_priv)) {
+	if (!HAS_DISPLAY(dev_priv) || !INTEL_DISPLAY_ENABLED(dev_priv)) {
 		DRM_DEBUG_KMS("Skipping VBT init due to disabled display.\n");
 		return;
 	}
@@ -2248,6 +2259,9 @@ enum aux_ch intel_bios_port_aux_ch(struct drm_i915_private *dev_priv,
 	case DP_AUX_F:
 		aux_ch = AUX_CH_F;
 		break;
+	case DP_AUX_G:
+		aux_ch = AUX_CH_G;
+		break;
 	default:
 		MISSING_CASE(info->alternate_aux_channel);
 		aux_ch = AUX_CH_A;
diff --git a/drivers/gpu/drm/i915/display/intel_bios.h b/drivers/gpu/drm/i915/display/intel_bios.h
index 4969189e620f..98f064828a57 100644
--- a/drivers/gpu/drm/i915/display/intel_bios.h
+++ b/drivers/gpu/drm/i915/display/intel_bios.h
@@ -1,5 +1,5 @@
 /*
- * Copyright © 2016 Intel Corporation
+ * Copyright © 2016-2019 Intel Corporation
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the "Software"),
@@ -35,6 +35,7 @@
 #include <drm/i915_drm.h>
 
 struct drm_i915_private;
+enum port;
 
 enum intel_backlight_type {
 	INTEL_BACKLIGHT_PMIC,
diff --git a/drivers/gpu/drm/i915/display/intel_bw.c b/drivers/gpu/drm/i915/display/intel_bw.c
index 688858ebe4d0..22e83f857de8 100644
--- a/drivers/gpu/drm/i915/display/intel_bw.c
+++ b/drivers/gpu/drm/i915/display/intel_bw.c
@@ -35,28 +35,54 @@ static int icl_pcode_read_mem_global_info(struct drm_i915_private *dev_priv,
 	if (ret)
 		return ret;
 
-	switch (val & 0xf) {
-	case 0:
-		qi->dram_type = INTEL_DRAM_DDR4;
-		break;
-	case 1:
-		qi->dram_type = INTEL_DRAM_DDR3;
-		break;
-	case 2:
-		qi->dram_type = INTEL_DRAM_LPDDR3;
-		break;
-	case 3:
-		qi->dram_type = INTEL_DRAM_LPDDR3;
-		break;
-	default:
-		MISSING_CASE(val & 0xf);
-		break;
+	if (IS_GEN(dev_priv, 12)) {
+		switch (val & 0xf) {
+		case 0:
+			qi->dram_type = INTEL_DRAM_DDR4;
+			break;
+		case 3:
+			qi->dram_type = INTEL_DRAM_LPDDR4;
+			break;
+		case 4:
+			qi->dram_type = INTEL_DRAM_DDR3;
+			break;
+		case 5:
+			qi->dram_type = INTEL_DRAM_LPDDR3;
+			break;
+		default:
+			MISSING_CASE(val & 0xf);
+			break;
+		}
+	} else if (IS_GEN(dev_priv, 11)) {
+		switch (val & 0xf) {
+		case 0:
+			qi->dram_type = INTEL_DRAM_DDR4;
+			break;
+		case 1:
+			qi->dram_type = INTEL_DRAM_DDR3;
+			break;
+		case 2:
+			qi->dram_type = INTEL_DRAM_LPDDR3;
+			break;
+		case 3:
+			qi->dram_type = INTEL_DRAM_LPDDR4;
+			break;
+		default:
+			MISSING_CASE(val & 0xf);
+			break;
+		}
+	} else {
+		MISSING_CASE(INTEL_GEN(dev_priv));
+		qi->dram_type = INTEL_DRAM_LPDDR3; /* Conservative default */
 	}
 
 	qi->num_channels = (val & 0xf0) >> 4;
 	qi->num_points = (val & 0xf00) >> 8;
 
-	qi->t_bl = qi->dram_type == INTEL_DRAM_DDR4 ? 4 : 8;
+	if (IS_GEN(dev_priv, 12))
+		qi->t_bl = qi->dram_type == INTEL_DRAM_DDR4 ? 4 : 16;
+	else if (IS_GEN(dev_priv, 11))
+		qi->t_bl = qi->dram_type == INTEL_DRAM_DDR4 ? 4 : 8;
 
 	return 0;
 }
@@ -132,20 +158,25 @@ static int icl_sagv_max_dclk(const struct intel_qgv_info *qi)
 }
 
 struct intel_sa_info {
-	u8 deburst, mpagesize, deprogbwlimit, displayrtids;
+	u16 displayrtids;
+	u8 deburst, deprogbwlimit;
 };
 
 static const struct intel_sa_info icl_sa_info = {
 	.deburst = 8,
-	.mpagesize = 16,
 	.deprogbwlimit = 25, /* GB/s */
 	.displayrtids = 128,
 };
 
-static int icl_get_bw_info(struct drm_i915_private *dev_priv)
+static const struct intel_sa_info tgl_sa_info = {
+	.deburst = 16,
+	.deprogbwlimit = 34, /* GB/s */
+	.displayrtids = 256,
+};
+
+static int icl_get_bw_info(struct drm_i915_private *dev_priv, const struct intel_sa_info *sa)
 {
 	struct intel_qgv_info qi = {};
-	const struct intel_sa_info *sa = &icl_sa_info;
 	bool is_y_tile = true; /* assume y tile may be used */
 	int num_channels;
 	int deinterleave;
@@ -233,14 +264,16 @@ static unsigned int icl_max_bw(struct drm_i915_private *dev_priv,
 
 void intel_bw_init_hw(struct drm_i915_private *dev_priv)
 {
-	if (IS_GEN(dev_priv, 11))
-		icl_get_bw_info(dev_priv);
+	if (IS_GEN(dev_priv, 12))
+		icl_get_bw_info(dev_priv, &tgl_sa_info);
+	else if (IS_GEN(dev_priv, 11))
+		icl_get_bw_info(dev_priv, &icl_sa_info);
 }
 
 static unsigned int intel_max_data_rate(struct drm_i915_private *dev_priv,
 					int num_planes)
 {
-	if (IS_GEN(dev_priv, 11))
+	if (INTEL_GEN(dev_priv) >= 11)
 		/*
 		 * FIXME with SAGV disabled maybe we can assume
 		 * point 1 will always be used? Seems to match
diff --git a/drivers/gpu/drm/i915/display/intel_cdclk.c b/drivers/gpu/drm/i915/display/intel_cdclk.c
index d0bc42e5039c..0caef2592a7e 100644
--- a/drivers/gpu/drm/i915/display/intel_cdclk.c
+++ b/drivers/gpu/drm/i915/display/intel_cdclk.c
@@ -21,6 +21,7 @@
  * DEALINGS IN THE SOFTWARE.
  */
 
+#include "intel_atomic.h"
 #include "intel_cdclk.h"
 #include "intel_display_types.h"
 #include "intel_sideband.h"
@@ -1161,28 +1162,88 @@ static void skl_uninit_cdclk(struct drm_i915_private *dev_priv)
 	skl_set_cdclk(dev_priv, &cdclk_state, INVALID_PIPE);
 }
 
-static int bxt_calc_cdclk(int min_cdclk)
-{
-	if (min_cdclk > 576000)
-		return 624000;
-	else if (min_cdclk > 384000)
-		return 576000;
-	else if (min_cdclk > 288000)
-		return 384000;
-	else if (min_cdclk > 144000)
-		return 288000;
-	else
-		return 144000;
+static const struct intel_cdclk_vals bxt_cdclk_table[] = {
+	{ .refclk = 19200, .cdclk = 144000, .divider = 8, .ratio = 60 },
+	{ .refclk = 19200, .cdclk = 288000, .divider = 4, .ratio = 60 },
+	{ .refclk = 19200, .cdclk = 384000, .divider = 3, .ratio = 60 },
+	{ .refclk = 19200, .cdclk = 576000, .divider = 2, .ratio = 60 },
+	{ .refclk = 19200, .cdclk = 624000, .divider = 2, .ratio = 65 },
+	{}
+};
+
+static const struct intel_cdclk_vals glk_cdclk_table[] = {
+	{ .refclk = 19200, .cdclk =  79200, .divider = 8, .ratio = 33 },
+	{ .refclk = 19200, .cdclk = 158400, .divider = 4, .ratio = 33 },
+	{ .refclk = 19200, .cdclk = 316800, .divider = 2, .ratio = 33 },
+	{}
+};
+
+static const struct intel_cdclk_vals cnl_cdclk_table[] = {
+	{ .refclk = 19200, .cdclk = 168000, .divider = 4, .ratio = 35 },
+	{ .refclk = 19200, .cdclk = 336000, .divider = 2, .ratio = 35 },
+	{ .refclk = 19200, .cdclk = 528000, .divider = 2, .ratio = 55 },
+
+	{ .refclk = 24000, .cdclk = 168000, .divider = 4, .ratio = 28 },
+	{ .refclk = 24000, .cdclk = 336000, .divider = 2, .ratio = 28 },
+	{ .refclk = 24000, .cdclk = 528000, .divider = 2, .ratio = 44 },
+	{}
+};
+
+static const struct intel_cdclk_vals icl_cdclk_table[] = {
+	{ .refclk = 19200, .cdclk = 172800, .divider = 2, .ratio = 18 },
+	{ .refclk = 19200, .cdclk = 192000, .divider = 2, .ratio = 20 },
+	{ .refclk = 19200, .cdclk = 307200, .divider = 2, .ratio = 32 },
+	{ .refclk = 19200, .cdclk = 326400, .divider = 4, .ratio = 68 },
+	{ .refclk = 19200, .cdclk = 556800, .divider = 2, .ratio = 58 },
+	{ .refclk = 19200, .cdclk = 652800, .divider = 2, .ratio = 68 },
+
+	{ .refclk = 24000, .cdclk = 180000, .divider = 2, .ratio = 15 },
+	{ .refclk = 24000, .cdclk = 192000, .divider = 2, .ratio = 16 },
+	{ .refclk = 24000, .cdclk = 312000, .divider = 2, .ratio = 26 },
+	{ .refclk = 24000, .cdclk = 324000, .divider = 4, .ratio = 54 },
+	{ .refclk = 24000, .cdclk = 552000, .divider = 2, .ratio = 46 },
+	{ .refclk = 24000, .cdclk = 648000, .divider = 2, .ratio = 54 },
+
+	{ .refclk = 38400, .cdclk = 172800, .divider = 2, .ratio =  9 },
+	{ .refclk = 38400, .cdclk = 192000, .divider = 2, .ratio = 10 },
+	{ .refclk = 38400, .cdclk = 307200, .divider = 2, .ratio = 16 },
+	{ .refclk = 38400, .cdclk = 326400, .divider = 4, .ratio = 34 },
+	{ .refclk = 38400, .cdclk = 556800, .divider = 2, .ratio = 29 },
+	{ .refclk = 38400, .cdclk = 652800, .divider = 2, .ratio = 34 },
+	{}
+};
+
+static int bxt_calc_cdclk(struct drm_i915_private *dev_priv, int min_cdclk)
+{
+	const struct intel_cdclk_vals *table = dev_priv->cdclk.table;
+	int i;
+
+	for (i = 0; table[i].refclk; i++)
+		if (table[i].refclk == dev_priv->cdclk.hw.ref &&
+		    table[i].cdclk >= min_cdclk)
+			return table[i].cdclk;
+
+	WARN(1, "Cannot satisfy minimum cdclk %d with refclk %u\n",
+	     min_cdclk, dev_priv->cdclk.hw.ref);
+	return 0;
 }
 
-static int glk_calc_cdclk(int min_cdclk)
+static int bxt_calc_cdclk_pll_vco(struct drm_i915_private *dev_priv, int cdclk)
 {
-	if (min_cdclk > 158400)
-		return 316800;
-	else if (min_cdclk > 79200)
-		return 158400;
-	else
-		return 79200;
+	const struct intel_cdclk_vals *table = dev_priv->cdclk.table;
+	int i;
+
+	if (cdclk == dev_priv->cdclk.hw.bypass)
+		return 0;
+
+	for (i = 0; table[i].refclk; i++)
+		if (table[i].refclk == dev_priv->cdclk.hw.ref &&
+		    table[i].cdclk == cdclk)
+			return dev_priv->cdclk.hw.ref * table[i].ratio;
+
+	WARN(1, "cdclk %d not valid for refclk %u\n",
+	     cdclk, dev_priv->cdclk.hw.ref);
+	return 0;
 }
 
 static u8 bxt_calc_voltage_level(int cdclk)
@@ -1190,69 +1251,99 @@ static u8 bxt_calc_voltage_level(int cdclk)
 	return DIV_ROUND_UP(cdclk, 25000);
 }
 
-static int bxt_de_pll_vco(struct drm_i915_private *dev_priv, int cdclk)
+static u8 cnl_calc_voltage_level(int cdclk)
 {
-	int ratio;
-
-	if (cdclk == dev_priv->cdclk.hw.bypass)
+	if (cdclk > 336000)
+		return 2;
+	else if (cdclk > 168000)
+		return 1;
+	else
 		return 0;
+}
 
-	switch (cdclk) {
-	default:
-		MISSING_CASE(cdclk);
-		/* fall through */
-	case 144000:
-	case 288000:
-	case 384000:
-	case 576000:
-		ratio = 60;
-		break;
-	case 624000:
-		ratio = 65;
-		break;
-	}
+static u8 icl_calc_voltage_level(int cdclk)
+{
+	if (cdclk > 556800)
+		return 2;
+	else if (cdclk > 312000)
+		return 1;
+	else
+		return 0;
+}
 
-	return dev_priv->cdclk.hw.ref * ratio;
+static u8 ehl_calc_voltage_level(int cdclk)
+{
+	if (cdclk > 312000)
+		return 2;
+	else if (cdclk > 180000)
+		return 1;
+	else
+		return 0;
 }
 
-static int glk_de_pll_vco(struct drm_i915_private *dev_priv, int cdclk)
+static void cnl_readout_refclk(struct drm_i915_private *dev_priv,
+			       struct intel_cdclk_state *cdclk_state)
 {
-	int ratio;
+	if (I915_READ(SKL_DSSM) & CNL_DSSM_CDCLK_PLL_REFCLK_24MHz)
+		cdclk_state->ref = 24000;
+	else
+		cdclk_state->ref = 19200;
+}
 
-	if (cdclk == dev_priv->cdclk.hw.bypass)
-		return 0;
+static void icl_readout_refclk(struct drm_i915_private *dev_priv,
+			       struct intel_cdclk_state *cdclk_state)
+{
+	u32 dssm = I915_READ(SKL_DSSM) & ICL_DSSM_CDCLK_PLL_REFCLK_MASK;
 
-	switch (cdclk) {
+	switch (dssm) {
 	default:
-		MISSING_CASE(cdclk);
+		MISSING_CASE(dssm);
 		/* fall through */
-	case  79200:
-	case 158400:
-	case 316800:
-		ratio = 33;
+	case ICL_DSSM_CDCLK_PLL_REFCLK_24MHz:
+		cdclk_state->ref = 24000;
+		break;
+	case ICL_DSSM_CDCLK_PLL_REFCLK_19_2MHz:
+		cdclk_state->ref = 19200;
+		break;
+	case ICL_DSSM_CDCLK_PLL_REFCLK_38_4MHz:
+		cdclk_state->ref = 38400;
 		break;
 	}
-
-	return dev_priv->cdclk.hw.ref * ratio;
 }
 
-static void bxt_de_pll_update(struct drm_i915_private *dev_priv,
-			      struct intel_cdclk_state *cdclk_state)
+static void bxt_de_pll_readout(struct drm_i915_private *dev_priv,
+			       struct intel_cdclk_state *cdclk_state)
 {
-	u32 val;
+	u32 val, ratio;
 
-	cdclk_state->ref = 19200;
-	cdclk_state->vco = 0;
+	if (INTEL_GEN(dev_priv) >= 11)
+		icl_readout_refclk(dev_priv, cdclk_state);
+	else if (IS_CANNONLAKE(dev_priv))
+		cnl_readout_refclk(dev_priv, cdclk_state);
+	else
+		cdclk_state->ref = 19200;
 
 	val = I915_READ(BXT_DE_PLL_ENABLE);
-	if ((val & BXT_DE_PLL_PLL_ENABLE) == 0)
+	if ((val & BXT_DE_PLL_PLL_ENABLE) == 0 ||
+	    (val & BXT_DE_PLL_LOCK) == 0) {
+		/*
+		 * CDCLK PLL is disabled, the VCO/ratio doesn't matter, but
+		 * setting it to zero is a way to signal that.
+		 */
+		cdclk_state->vco = 0;
 		return;
+	}
 
-	if (WARN_ON((val & BXT_DE_PLL_LOCK) == 0))
-		return;
+	/*
+	 * CNL+ have the ratio directly in the PLL enable register, gen9lp had
+	 * it in a separate PLL control register.
+	 */
+	if (INTEL_GEN(dev_priv) >= 10)
+		ratio = val & CNL_CDCLK_PLL_RATIO_MASK;
+	else
+		ratio = I915_READ(BXT_DE_PLL_CTL) & BXT_DE_PLL_RATIO_MASK;
 
-	val = I915_READ(BXT_DE_PLL_CTL);
-	cdclk_state->vco = (val & BXT_DE_PLL_RATIO_MASK) * cdclk_state->ref;
+	cdclk_state->vco = ratio * cdclk_state->ref;
 }
 
 static void bxt_get_cdclk(struct drm_i915_private *dev_priv,
@@ -1261,12 +1352,19 @@ static void bxt_get_cdclk(struct drm_i915_private *dev_priv,
 	u32 divider;
 	int div;
 
-	bxt_de_pll_update(dev_priv, cdclk_state);
+	bxt_de_pll_readout(dev_priv, cdclk_state);
 
-	cdclk_state->cdclk = cdclk_state->bypass = cdclk_state->ref;
+	if (INTEL_GEN(dev_priv) >= 12)
+		cdclk_state->bypass = cdclk_state->ref / 2;
+	else if (INTEL_GEN(dev_priv) >= 11)
+		cdclk_state->bypass = 50000;
+	else
+		cdclk_state->bypass = cdclk_state->ref;
 
-	if (cdclk_state->vco == 0)
+	if (cdclk_state->vco == 0) {
+		cdclk_state->cdclk = cdclk_state->bypass;
 		goto out;
+	}
 
 	divider = I915_READ(CDCLK_CTL) & BXT_CDCLK_CD2X_DIV_SEL_MASK;
 
@@ -1275,13 +1373,15 @@ static void bxt_get_cdclk(struct drm_i915_private *dev_priv,
 		div = 2;
 		break;
 	case BXT_CDCLK_CD2X_DIV_SEL_1_5:
-		WARN(IS_GEMINILAKE(dev_priv), "Unsupported divider\n");
+		WARN(IS_GEMINILAKE(dev_priv) || INTEL_GEN(dev_priv) >= 10,
+		     "Unsupported divider\n");
 		div = 3;
 		break;
 	case BXT_CDCLK_CD2X_DIV_SEL_2:
 		div = 4;
 		break;
 	case BXT_CDCLK_CD2X_DIV_SEL_4:
+		WARN(INTEL_GEN(dev_priv) >= 10, "Unsupported divider\n");
 		div = 8;
 		break;
 	default:
@@ -1297,7 +1397,7 @@ static void bxt_get_cdclk(struct drm_i915_private *dev_priv,
 	 * at least what the CDCLK frequency requires.
 	 */
 	cdclk_state->voltage_level =
-		bxt_calc_voltage_level(cdclk_state->cdclk);
+		dev_priv->display.calc_voltage_level(cdclk_state->cdclk);
 }
 
 static void bxt_de_pll_disable(struct drm_i915_private *dev_priv)
@@ -1332,259 +1432,6 @@ static void bxt_de_pll_enable(struct drm_i915_private *dev_priv, int vco)
 	dev_priv->cdclk.hw.vco = vco;
 }
 
-static void bxt_set_cdclk(struct drm_i915_private *dev_priv,
-			  const struct intel_cdclk_state *cdclk_state,
-			  enum pipe pipe)
-{
-	int cdclk = cdclk_state->cdclk;
-	int vco = cdclk_state->vco;
-	u32 val, divider;
-	int ret;
-
-	/* cdclk = vco / 2 / div{1,1.5,2,4} */
-	switch (DIV_ROUND_CLOSEST(vco, cdclk)) {
-	default:
-		WARN_ON(cdclk != dev_priv->cdclk.hw.bypass);
-		WARN_ON(vco != 0);
-		/* fall through */
-	case 2:
-		divider = BXT_CDCLK_CD2X_DIV_SEL_1;
-		break;
-	case 3:
-		WARN(IS_GEMINILAKE(dev_priv), "Unsupported divider\n");
-		divider = BXT_CDCLK_CD2X_DIV_SEL_1_5;
-		break;
-	case 4:
-		divider = BXT_CDCLK_CD2X_DIV_SEL_2;
-		break;
-	case 8:
-		divider = BXT_CDCLK_CD2X_DIV_SEL_4;
-		break;
-	}
-
-	/*
-	 * Inform power controller of upcoming frequency change. BSpec
-	 * requires us to wait up to 150usec, but that leads to timeouts;
-	 * the 2ms used here is based on experiment.
-	 */
-	ret = sandybridge_pcode_write_timeout(dev_priv,
-					      HSW_PCODE_DE_WRITE_FREQ_REQ,
-					      0x80000000, 150, 2);
-	if (ret) {
-		DRM_ERROR("PCode CDCLK freq change notify failed (err %d, freq %d)\n",
-			  ret, cdclk);
-		return;
-	}
-
-	if (dev_priv->cdclk.hw.vco != 0 &&
-	    dev_priv->cdclk.hw.vco != vco)
-		bxt_de_pll_disable(dev_priv);
-
-	if (dev_priv->cdclk.hw.vco != vco)
-		bxt_de_pll_enable(dev_priv, vco);
-
-	val = divider | skl_cdclk_decimal(cdclk);
-	if (pipe == INVALID_PIPE)
-		val |= BXT_CDCLK_CD2X_PIPE_NONE;
-	else
-		val |= BXT_CDCLK_CD2X_PIPE(pipe);
-	/*
-	 * Disable SSA Precharge when CD clock frequency < 500 MHz,
-	 * enable otherwise.
-	 */
-	if (cdclk >= 500000)
-		val |= BXT_CDCLK_SSA_PRECHARGE_ENABLE;
-	I915_WRITE(CDCLK_CTL, val);
-
-	if (pipe != INVALID_PIPE)
-		intel_wait_for_vblank(dev_priv, pipe);
-
-	/*
-	 * The timeout isn't specified, the 2ms used here is based on
-	 * experiment.
-	 * FIXME: Waiting for the request completion could be delayed until
-	 * the next PCODE request based on BSpec.
-	 */
-	ret = sandybridge_pcode_write_timeout(dev_priv,
-					      HSW_PCODE_DE_WRITE_FREQ_REQ,
-					      cdclk_state->voltage_level, 150, 2);
-	if (ret) {
-		DRM_ERROR("PCode CDCLK freq set failed, (err %d, freq %d)\n",
-			  ret, cdclk);
-		return;
-	}
-
-	intel_update_cdclk(dev_priv);
-}
-
-static void bxt_sanitize_cdclk(struct drm_i915_private *dev_priv)
-{
-	u32 cdctl, expected;
-
-	intel_update_cdclk(dev_priv);
-	intel_dump_cdclk_state(&dev_priv->cdclk.hw, "Current CDCLK");
-
-	if (dev_priv->cdclk.hw.vco == 0 ||
-	    dev_priv->cdclk.hw.cdclk == dev_priv->cdclk.hw.bypass)
-		goto sanitize;
-
-	/* DPLL okay; verify the cdclock
-	 *
-	 * Some BIOS versions leave an incorrect decimal frequency value and
-	 * set reserved MBZ bits in CDCLK_CTL at least during exiting from S4,
-	 * so sanitize this register.
-	 */
-	cdctl = I915_READ(CDCLK_CTL);
-	/*
-	 * Let's ignore the pipe field, since BIOS could have configured the
-	 * dividers both synching to an active pipe, or asynchronously
-	 * (PIPE_NONE).
-	 */
-	cdctl &= ~BXT_CDCLK_CD2X_PIPE_NONE;
-
-	expected = (cdctl & BXT_CDCLK_CD2X_DIV_SEL_MASK) |
-		skl_cdclk_decimal(dev_priv->cdclk.hw.cdclk);
-	/*
-	 * Disable SSA Precharge when CD clock frequency < 500 MHz,
-	 * enable otherwise.
-	 */
-	if (dev_priv->cdclk.hw.cdclk >= 500000)
-		expected |= BXT_CDCLK_SSA_PRECHARGE_ENABLE;
-
-	if (cdctl == expected)
-		/* All well; nothing to sanitize */
-		return;
-
-sanitize:
-	DRM_DEBUG_KMS("Sanitizing cdclk programmed by pre-os\n");
-
-	/* force cdclk programming */
-	dev_priv->cdclk.hw.cdclk = 0;
-
-	/* force full PLL disable + enable */
-	dev_priv->cdclk.hw.vco = -1;
-}
-
-static void bxt_init_cdclk(struct drm_i915_private *dev_priv)
-{
-	struct intel_cdclk_state cdclk_state;
-
-	bxt_sanitize_cdclk(dev_priv);
-
-	if (dev_priv->cdclk.hw.cdclk != 0 &&
-	    dev_priv->cdclk.hw.vco != 0)
-		return;
-
-	cdclk_state = dev_priv->cdclk.hw;
-
-	/*
-	 * FIXME:
-	 * - The initial CDCLK needs to be read from VBT.
-	 *   Need to make this change after VBT has changes for BXT.
-	 */
-	if (IS_GEMINILAKE(dev_priv)) {
-		cdclk_state.cdclk = glk_calc_cdclk(0);
-		cdclk_state.vco = glk_de_pll_vco(dev_priv, cdclk_state.cdclk);
-	} else {
-		cdclk_state.cdclk = bxt_calc_cdclk(0);
-		cdclk_state.vco = bxt_de_pll_vco(dev_priv, cdclk_state.cdclk);
-	}
-	cdclk_state.voltage_level = bxt_calc_voltage_level(cdclk_state.cdclk);
-
-	bxt_set_cdclk(dev_priv, &cdclk_state, INVALID_PIPE);
-}
-
-static void bxt_uninit_cdclk(struct drm_i915_private *dev_priv)
-{
-	struct intel_cdclk_state cdclk_state = dev_priv->cdclk.hw;
-
-	cdclk_state.cdclk = cdclk_state.bypass;
-	cdclk_state.vco = 0;
-	cdclk_state.voltage_level = bxt_calc_voltage_level(cdclk_state.cdclk);
-
-	bxt_set_cdclk(dev_priv, &cdclk_state, INVALID_PIPE);
-}
-
-static int cnl_calc_cdclk(int min_cdclk)
-{
-	if (min_cdclk > 336000)
-		return 528000;
-	else if (min_cdclk > 168000)
-		return 336000;
-	else
-		return 168000;
-}
-
-static u8 cnl_calc_voltage_level(int cdclk)
-{
-	if (cdclk > 336000)
-		return 2;
-	else if (cdclk > 168000)
-		return 1;
-	else
-		return 0;
-}
-
-static void cnl_cdclk_pll_update(struct drm_i915_private *dev_priv,
-				 struct intel_cdclk_state *cdclk_state)
-{
-	u32 val;
-
-	if (I915_READ(SKL_DSSM) & CNL_DSSM_CDCLK_PLL_REFCLK_24MHz)
-		cdclk_state->ref = 24000;
-	else
-		cdclk_state->ref = 19200;
-
-	cdclk_state->vco = 0;
-
-	val = I915_READ(BXT_DE_PLL_ENABLE);
-	if ((val & BXT_DE_PLL_PLL_ENABLE) == 0)
-		return;
-
-	if (WARN_ON((val & BXT_DE_PLL_LOCK) == 0))
-		return;
-
-	cdclk_state->vco = (val & CNL_CDCLK_PLL_RATIO_MASK) * cdclk_state->ref;
-}
-
-static void cnl_get_cdclk(struct drm_i915_private *dev_priv,
-			 struct intel_cdclk_state *cdclk_state)
-{
-	u32 divider;
-	int div;
-
-	cnl_cdclk_pll_update(dev_priv, cdclk_state);
-
-	cdclk_state->cdclk = cdclk_state->bypass = cdclk_state->ref;
-
-	if (cdclk_state->vco == 0)
-		goto out;
-
-	divider = I915_READ(CDCLK_CTL) & BXT_CDCLK_CD2X_DIV_SEL_MASK;
-
-	switch (divider) {
-	case BXT_CDCLK_CD2X_DIV_SEL_1:
-		div = 2;
-		break;
-	case BXT_CDCLK_CD2X_DIV_SEL_2:
-		div = 4;
-		break;
-	default:
-		MISSING_CASE(divider);
-		return;
-	}
-
-	cdclk_state->cdclk = DIV_ROUND_CLOSEST(cdclk_state->vco, div);
-
- out:
-	/*
-	 * Can't read this out :( Let's assume it's
-	 * at least what the CDCLK frequency requires.
-	 */
-	cdclk_state->voltage_level =
-		cnl_calc_voltage_level(cdclk_state->cdclk);
-}
-
 static void cnl_cdclk_pll_disable(struct drm_i915_private *dev_priv)
 {
 	u32 val;
@@ -1618,7 +1465,27 @@ static void cnl_cdclk_pll_enable(struct drm_i915_private *dev_priv, int vco)
 	dev_priv->cdclk.hw.vco = vco;
 }
 
-static void cnl_set_cdclk(struct drm_i915_private *dev_priv,
+static u32 bxt_cdclk_cd2x_pipe(struct drm_i915_private *dev_priv, enum pipe pipe)
+{
+	if (INTEL_GEN(dev_priv) >= 12) {
+		if (pipe == INVALID_PIPE)
+			return TGL_CDCLK_CD2X_PIPE_NONE;
+		else
+			return TGL_CDCLK_CD2X_PIPE(pipe);
+	} else if (INTEL_GEN(dev_priv) >= 11) {
+		if (pipe == INVALID_PIPE)
+			return ICL_CDCLK_CD2X_PIPE_NONE;
+		else
+			return ICL_CDCLK_CD2X_PIPE(pipe);
+	} else {
+		if (pipe == INVALID_PIPE)
+			return BXT_CDCLK_CD2X_PIPE_NONE;
+		else
+			return BXT_CDCLK_CD2X_PIPE(pipe);
+	}
+}
+
+static void bxt_set_cdclk(struct drm_i915_private *dev_priv,
 			  const struct intel_cdclk_state *cdclk_state,
 			  enum pipe pipe)
 {
@@ -1627,17 +1494,28 @@ static void cnl_set_cdclk(struct drm_i915_private *dev_priv,
 	u32 val, divider;
 	int ret;
 
-	ret = skl_pcode_request(dev_priv, SKL_PCODE_CDCLK_CONTROL,
-				SKL_CDCLK_PREPARE_FOR_CHANGE,
-				SKL_CDCLK_READY_FOR_CHANGE,
-				SKL_CDCLK_READY_FOR_CHANGE, 3);
+	/* Inform power controller of upcoming frequency change. */
+	if (INTEL_GEN(dev_priv) >= 10)
+		ret = skl_pcode_request(dev_priv, SKL_PCODE_CDCLK_CONTROL,
+					SKL_CDCLK_PREPARE_FOR_CHANGE,
+					SKL_CDCLK_READY_FOR_CHANGE,
+					SKL_CDCLK_READY_FOR_CHANGE, 3);
+	else
+		/*
+		 * BSpec requires us to wait up to 150usec, but that leads to
+		 * timeouts; the 2ms used here is based on experiment.
+		 */
+		ret = sandybridge_pcode_write_timeout(dev_priv,
+						      HSW_PCODE_DE_WRITE_FREQ_REQ,
+						      0x80000000, 150, 2);
+
 	if (ret) {
-		DRM_ERROR("Failed to inform PCU about cdclk change (%d)\n",
-			  ret);
+		DRM_ERROR("Failed to inform PCU about cdclk change (err %d, freq %d)\n",
+			  ret, cdclk);
 		return;
 	}
 
-	/* cdclk = vco / 2 / div{1,2} */
+	/* cdclk = vco / 2 / div{1,1.5,2,4} */
 	switch (DIV_ROUND_CLOSEST(vco, cdclk)) {
 	default:
 		WARN_ON(cdclk != dev_priv->cdclk.hw.bypass);
@@ -1646,67 +1524,87 @@ static void cnl_set_cdclk(struct drm_i915_private *dev_priv,
 	case 2:
 		divider = BXT_CDCLK_CD2X_DIV_SEL_1;
 		break;
+	case 3:
+		WARN(IS_GEMINILAKE(dev_priv) || INTEL_GEN(dev_priv) >= 10,
+		     "Unsupported divider\n");
+		divider = BXT_CDCLK_CD2X_DIV_SEL_1_5;
+		break;
 	case 4:
 		divider = BXT_CDCLK_CD2X_DIV_SEL_2;
 		break;
+	case 8:
+		WARN(INTEL_GEN(dev_priv) >= 10, "Unsupported divider\n");
+		divider = BXT_CDCLK_CD2X_DIV_SEL_4;
+		break;
 	}
 
-	if (dev_priv->cdclk.hw.vco != 0 &&
-	    dev_priv->cdclk.hw.vco != vco)
-		cnl_cdclk_pll_disable(dev_priv);
-
-	if (dev_priv->cdclk.hw.vco != vco)
-		cnl_cdclk_pll_enable(dev_priv, vco);
+	if (INTEL_GEN(dev_priv) >= 10) {
+		if (dev_priv->cdclk.hw.vco != 0 &&
+		    dev_priv->cdclk.hw.vco != vco)
+			cnl_cdclk_pll_disable(dev_priv);
 
-	val = divider | skl_cdclk_decimal(cdclk);
-	if (pipe == INVALID_PIPE)
-		val |= BXT_CDCLK_CD2X_PIPE_NONE;
-	else
-		val |= BXT_CDCLK_CD2X_PIPE(pipe);
-	I915_WRITE(CDCLK_CTL, val);
+		if (dev_priv->cdclk.hw.vco != vco)
+			cnl_cdclk_pll_enable(dev_priv, vco);
 
-	if (pipe != INVALID_PIPE)
-		intel_wait_for_vblank(dev_priv, pipe);
+	} else {
+		if (dev_priv->cdclk.hw.vco != 0 &&
+		    dev_priv->cdclk.hw.vco != vco)
+			bxt_de_pll_disable(dev_priv);
 
-	/* inform PCU of the change */
-	sandybridge_pcode_write(dev_priv, SKL_PCODE_CDCLK_CONTROL,
-				cdclk_state->voltage_level);
+		if (dev_priv->cdclk.hw.vco != vco)
+			bxt_de_pll_enable(dev_priv, vco);
+	}
 
-	intel_update_cdclk(dev_priv);
+	val = divider | skl_cdclk_decimal(cdclk) |
+		bxt_cdclk_cd2x_pipe(dev_priv, pipe);
 
 	/*
-	 * Can't read out the voltage level :(
-	 * Let's just assume everything is as expected.
+	 * Disable SSA Precharge when CD clock frequency < 500 MHz,
+	 * enable otherwise.
 	 */
-	dev_priv->cdclk.hw.voltage_level = cdclk_state->voltage_level;
-}
+	if (IS_GEN9_LP(dev_priv) && cdclk >= 500000)
+		val |= BXT_CDCLK_SSA_PRECHARGE_ENABLE;
+	I915_WRITE(CDCLK_CTL, val);
 
-static int cnl_cdclk_pll_vco(struct drm_i915_private *dev_priv, int cdclk)
-{
-	int ratio;
+	if (pipe != INVALID_PIPE)
+		intel_wait_for_vblank(dev_priv, pipe);
 
-	if (cdclk == dev_priv->cdclk.hw.bypass)
-		return 0;
+	if (INTEL_GEN(dev_priv) >= 10) {
+		ret = sandybridge_pcode_write(dev_priv, SKL_PCODE_CDCLK_CONTROL,
+					      cdclk_state->voltage_level);
+	} else {
+		/*
+		 * The timeout isn't specified, the 2ms used here is based on
+		 * experiment.
+		 * FIXME: Waiting for the request completion could be delayed
+		 * until the next PCODE request based on BSpec.
+		 */
+		ret = sandybridge_pcode_write_timeout(dev_priv,
+						      HSW_PCODE_DE_WRITE_FREQ_REQ,
+						      cdclk_state->voltage_level,
+						      150, 2);
+	}
 
-	switch (cdclk) {
-	default:
-		MISSING_CASE(cdclk);
-		/* fall through */
-	case 168000:
-	case 336000:
-		ratio = dev_priv->cdclk.hw.ref == 19200 ? 35 : 28;
-		break;
-	case 528000:
-		ratio = dev_priv->cdclk.hw.ref == 19200 ? 55 : 44;
-		break;
+	if (ret) {
+		DRM_ERROR("PCode CDCLK freq set failed, (err %d, freq %d)\n",
+			  ret, cdclk);
+		return;
 	}
 
-	return dev_priv->cdclk.hw.ref * ratio;
+	intel_update_cdclk(dev_priv);
+
+	if (INTEL_GEN(dev_priv) >= 10)
+		/*
+		 * Can't read out the voltage level :(
+		 * Let's just assume everything is as expected.
+		 */
+		dev_priv->cdclk.hw.voltage_level = cdclk_state->voltage_level;
 }
 
-static void cnl_sanitize_cdclk(struct drm_i915_private *dev_priv)
+static void bxt_sanitize_cdclk(struct drm_i915_private *dev_priv)
 {
 	u32 cdctl, expected;
+	int cdclk, vco;
 
 	intel_update_cdclk(dev_priv);
 	intel_dump_cdclk_state(&dev_priv->cdclk.hw, "Current CDCLK");
@@ -1727,262 +1625,65 @@ static void cnl_sanitize_cdclk(struct drm_i915_private *dev_priv)
 	 * dividers both synching to an active pipe, or asynchronously
 	 * (PIPE_NONE).
 	 */
-	cdctl &= ~BXT_CDCLK_CD2X_PIPE_NONE;
-
-	expected = (cdctl & BXT_CDCLK_CD2X_DIV_SEL_MASK) |
-		   skl_cdclk_decimal(dev_priv->cdclk.hw.cdclk);
-
-	if (cdctl == expected)
-		/* All well; nothing to sanitize */
-		return;
-
-sanitize:
-	DRM_DEBUG_KMS("Sanitizing cdclk programmed by pre-os\n");
+	cdctl &= ~bxt_cdclk_cd2x_pipe(dev_priv, INVALID_PIPE);
 
-	/* force cdclk programming */
-	dev_priv->cdclk.hw.cdclk = 0;
+	/* Make sure this is a legal cdclk value for the platform */
+	cdclk = bxt_calc_cdclk(dev_priv, dev_priv->cdclk.hw.cdclk);
+	if (cdclk != dev_priv->cdclk.hw.cdclk)
+		goto sanitize;
 
-	/* force full PLL disable + enable */
-	dev_priv->cdclk.hw.vco = -1;
-}
+	/* Make sure the VCO is correct for the cdclk */
+	vco = bxt_calc_cdclk_pll_vco(dev_priv, cdclk);
+	if (vco != dev_priv->cdclk.hw.vco)
+		goto sanitize;
 
-static int icl_calc_cdclk(int min_cdclk, unsigned int ref)
-{
-	static const int ranges_24[] = { 180000, 192000, 312000, 552000, 648000 };
-	static const int ranges_19_38[] = { 172800, 192000, 307200, 556800, 652800 };
-	const int *ranges;
-	int len, i;
+	expected = skl_cdclk_decimal(cdclk);
 
-	switch (ref) {
-	default:
-		MISSING_CASE(ref);
-		/* fall through */
-	case 24000:
-		ranges = ranges_24;
-		len = ARRAY_SIZE(ranges_24);
-		break;
-	case 19200:
-	case 38400:
-		ranges = ranges_19_38;
-		len = ARRAY_SIZE(ranges_19_38);
+	/* Figure out what CD2X divider we should be using for this cdclk */
+	switch (DIV_ROUND_CLOSEST(dev_priv->cdclk.hw.vco,
+				  dev_priv->cdclk.hw.cdclk)) {
+	case 2:
+		expected |= BXT_CDCLK_CD2X_DIV_SEL_1;
 		break;
-	}
-
-	for (i = 0; i < len; i++) {
-		if (min_cdclk <= ranges[i])
-			return ranges[i];
-	}
-
-	WARN_ON(min_cdclk > ranges[len - 1]);
-	return ranges[len - 1];
-}
-
-static int icl_calc_cdclk_pll_vco(struct drm_i915_private *dev_priv, int cdclk)
-{
-	int ratio;
-
-	if (cdclk == dev_priv->cdclk.hw.bypass)
-		return 0;
-
-	switch (cdclk) {
-	default:
-		MISSING_CASE(cdclk);
-		/* fall through */
-	case 172800:
-	case 307200:
-	case 556800:
-	case 652800:
-		WARN_ON(dev_priv->cdclk.hw.ref != 19200 &&
-			dev_priv->cdclk.hw.ref != 38400);
+	case 3:
+		expected |= BXT_CDCLK_CD2X_DIV_SEL_1_5;
 		break;
-	case 180000:
-	case 312000:
-	case 552000:
-	case 648000:
-		WARN_ON(dev_priv->cdclk.hw.ref != 24000);
+	case 4:
+		expected |= BXT_CDCLK_CD2X_DIV_SEL_2;
 		break;
-	case 192000:
-		WARN_ON(dev_priv->cdclk.hw.ref != 19200 &&
-			dev_priv->cdclk.hw.ref != 38400 &&
-			dev_priv->cdclk.hw.ref != 24000);
+	case 8:
+		expected |= BXT_CDCLK_CD2X_DIV_SEL_4;
 		break;
-	}
-
-	ratio = cdclk / (dev_priv->cdclk.hw.ref / 2);
-
-	return dev_priv->cdclk.hw.ref * ratio;
-}
-
-static void icl_set_cdclk(struct drm_i915_private *dev_priv,
-			  const struct intel_cdclk_state *cdclk_state,
-			  enum pipe pipe)
-{
-	unsigned int cdclk = cdclk_state->cdclk;
-	unsigned int vco = cdclk_state->vco;
-	int ret;
-
-	ret = skl_pcode_request(dev_priv, SKL_PCODE_CDCLK_CONTROL,
-				SKL_CDCLK_PREPARE_FOR_CHANGE,
-				SKL_CDCLK_READY_FOR_CHANGE,
-				SKL_CDCLK_READY_FOR_CHANGE, 3);
-	if (ret) {
-		DRM_ERROR("Failed to inform PCU about cdclk change (%d)\n",
-			  ret);
-		return;
-	}
-
-	if (dev_priv->cdclk.hw.vco != 0 &&
-	    dev_priv->cdclk.hw.vco != vco)
-		cnl_cdclk_pll_disable(dev_priv);
-
-	if (dev_priv->cdclk.hw.vco != vco)
-		cnl_cdclk_pll_enable(dev_priv, vco);
-
-	/*
-	 * On ICL CD2X_DIV can only be 1, so we'll never end up changing the
-	 * divider here synchronized to a pipe while CDCLK is on, nor will we
-	 * need the corresponding vblank wait.
-	 */
-	I915_WRITE(CDCLK_CTL, ICL_CDCLK_CD2X_PIPE_NONE |
-			      skl_cdclk_decimal(cdclk));
-
-	sandybridge_pcode_write(dev_priv, SKL_PCODE_CDCLK_CONTROL,
-				cdclk_state->voltage_level);
-
-	intel_update_cdclk(dev_priv);
-
-	/*
-	 * Can't read out the voltage level :(
-	 * Let's just assume everything is as expected.
-	 */
-	dev_priv->cdclk.hw.voltage_level = cdclk_state->voltage_level;
-}
-
-static u8 icl_calc_voltage_level(struct drm_i915_private *dev_priv, int cdclk)
-{
-	if (IS_ELKHARTLAKE(dev_priv)) {
-		if (cdclk > 312000)
-			return 2;
-		else if (cdclk > 180000)
-			return 1;
-		else
-			return 0;
-	} else {
-		if (cdclk > 556800)
-			return 2;
-		else if (cdclk > 312000)
-			return 1;
-		else
-			return 0;
-	}
-}
-
-static void icl_get_cdclk(struct drm_i915_private *dev_priv,
-			  struct intel_cdclk_state *cdclk_state)
-{
-	u32 val;
-
-	cdclk_state->bypass = 50000;
-
-	val = I915_READ(SKL_DSSM);
-	switch (val & ICL_DSSM_CDCLK_PLL_REFCLK_MASK) {
 	default:
-		MISSING_CASE(val);
-		/* fall through */
-	case ICL_DSSM_CDCLK_PLL_REFCLK_24MHz:
-		cdclk_state->ref = 24000;
-		break;
-	case ICL_DSSM_CDCLK_PLL_REFCLK_19_2MHz:
-		cdclk_state->ref = 19200;
-		break;
-	case ICL_DSSM_CDCLK_PLL_REFCLK_38_4MHz:
-		cdclk_state->ref = 38400;
-		break;
-	}
-
-	val = I915_READ(BXT_DE_PLL_ENABLE);
-	if ((val & BXT_DE_PLL_PLL_ENABLE) == 0 ||
-	    (val & BXT_DE_PLL_LOCK) == 0) {
-		/*
-		 * CDCLK PLL is disabled, the VCO/ratio doesn't matter, but
-		 * setting it to zero is a way to signal that.
-		 */
-		cdclk_state->vco = 0;
-		cdclk_state->cdclk = cdclk_state->bypass;
-		goto out;
+		goto sanitize;
 	}
 
-	cdclk_state->vco = (val & BXT_DE_PLL_RATIO_MASK) * cdclk_state->ref;
-
-	val = I915_READ(CDCLK_CTL);
-	WARN_ON((val & BXT_CDCLK_CD2X_DIV_SEL_MASK) != 0);
-
-	cdclk_state->cdclk = cdclk_state->vco / 2;
-
-out:
 	/*
-	 * Can't read this out :( Let's assume it's
-	 * at least what the CDCLK frequency requires.
+	 * Disable SSA Precharge when CD clock frequency < 500 MHz,
+	 * enable otherwise.
 	 */
-	cdclk_state->voltage_level =
-		icl_calc_voltage_level(dev_priv, cdclk_state->cdclk);
-}
-
-static void icl_init_cdclk(struct drm_i915_private *dev_priv)
-{
-	struct intel_cdclk_state sanitized_state;
-	u32 val;
-
-	/* This sets dev_priv->cdclk.hw. */
-	intel_update_cdclk(dev_priv);
-	intel_dump_cdclk_state(&dev_priv->cdclk.hw, "Current CDCLK");
-
-	/* This means CDCLK disabled. */
-	if (dev_priv->cdclk.hw.cdclk == dev_priv->cdclk.hw.bypass)
-		goto sanitize;
-
-	val = I915_READ(CDCLK_CTL);
-
-	if ((val & BXT_CDCLK_CD2X_DIV_SEL_MASK) != 0)
-		goto sanitize;
-
-	if ((val & CDCLK_FREQ_DECIMAL_MASK) !=
-	    skl_cdclk_decimal(dev_priv->cdclk.hw.cdclk))
-		goto sanitize;
+	if (IS_GEN9_LP(dev_priv) && dev_priv->cdclk.hw.cdclk >= 500000)
+		expected |= BXT_CDCLK_SSA_PRECHARGE_ENABLE;
 
-	return;
+	if (cdctl == expected)
+		/* All well; nothing to sanitize */
+		return;
 
 sanitize:
 	DRM_DEBUG_KMS("Sanitizing cdclk programmed by pre-os\n");
 
-	sanitized_state.ref = dev_priv->cdclk.hw.ref;
-	sanitized_state.cdclk = icl_calc_cdclk(0, sanitized_state.ref);
-	sanitized_state.vco = icl_calc_cdclk_pll_vco(dev_priv,
-						     sanitized_state.cdclk);
-	sanitized_state.voltage_level =
-				icl_calc_voltage_level(dev_priv,
-						       sanitized_state.cdclk);
-
-	icl_set_cdclk(dev_priv, &sanitized_state, INVALID_PIPE);
-}
-
-static void icl_uninit_cdclk(struct drm_i915_private *dev_priv)
-{
-	struct intel_cdclk_state cdclk_state = dev_priv->cdclk.hw;
-
-	cdclk_state.cdclk = cdclk_state.bypass;
-	cdclk_state.vco = 0;
-	cdclk_state.voltage_level = icl_calc_voltage_level(dev_priv,
-							   cdclk_state.cdclk);
+	/* force cdclk programming */
+	dev_priv->cdclk.hw.cdclk = 0;
 
-	icl_set_cdclk(dev_priv, &cdclk_state, INVALID_PIPE);
+	/* force full PLL disable + enable */
+	dev_priv->cdclk.hw.vco = -1;
 }
 
-static void cnl_init_cdclk(struct drm_i915_private *dev_priv)
+static void bxt_init_cdclk(struct drm_i915_private *dev_priv)
 {
 	struct intel_cdclk_state cdclk_state;
 
-	cnl_sanitize_cdclk(dev_priv);
+	bxt_sanitize_cdclk(dev_priv);
 
 	if (dev_priv->cdclk.hw.cdclk != 0 &&
 	    dev_priv->cdclk.hw.vco != 0)
@@ -1990,22 +1691,29 @@ static void cnl_init_cdclk(struct drm_i915_private *dev_priv)
 
 	cdclk_state = dev_priv->cdclk.hw;
 
-	cdclk_state.cdclk = cnl_calc_cdclk(0);
-	cdclk_state.vco = cnl_cdclk_pll_vco(dev_priv, cdclk_state.cdclk);
-	cdclk_state.voltage_level = cnl_calc_voltage_level(cdclk_state.cdclk);
+	/*
+	 * FIXME:
+	 * - The initial CDCLK needs to be read from VBT.
+	 *   Need to make this change after VBT has changes for BXT.
+	 */
+	cdclk_state.cdclk = bxt_calc_cdclk(dev_priv, 0);
+	cdclk_state.vco = bxt_calc_cdclk_pll_vco(dev_priv, cdclk_state.cdclk);
+	cdclk_state.voltage_level =
+		dev_priv->display.calc_voltage_level(cdclk_state.cdclk);
 
-	cnl_set_cdclk(dev_priv, &cdclk_state, INVALID_PIPE);
+	bxt_set_cdclk(dev_priv, &cdclk_state, INVALID_PIPE);
 }
 
-static void cnl_uninit_cdclk(struct drm_i915_private *dev_priv)
+static void bxt_uninit_cdclk(struct drm_i915_private *dev_priv)
 {
 	struct intel_cdclk_state cdclk_state = dev_priv->cdclk.hw;
 
 	cdclk_state.cdclk = cdclk_state.bypass;
 	cdclk_state.vco = 0;
-	cdclk_state.voltage_level = cnl_calc_voltage_level(cdclk_state.cdclk);
+	cdclk_state.voltage_level =
+		dev_priv->display.calc_voltage_level(cdclk_state.cdclk);
 
-	cnl_set_cdclk(dev_priv, &cdclk_state, INVALID_PIPE);
+	bxt_set_cdclk(dev_priv, &cdclk_state, INVALID_PIPE);
 }
 
 /**
@@ -2019,14 +1727,10 @@ static void cnl_uninit_cdclk(struct drm_i915_private *dev_priv)
  */
 void intel_cdclk_init(struct drm_i915_private *i915)
 {
-	if (INTEL_GEN(i915) >= 11)
-		icl_init_cdclk(i915);
-	else if (IS_CANNONLAKE(i915))
-		cnl_init_cdclk(i915);
+	if (IS_GEN9_LP(i915) || INTEL_GEN(i915) >= 10)
+		bxt_init_cdclk(i915);
 	else if (IS_GEN9_BC(i915))
 		skl_init_cdclk(i915);
-	else if (IS_GEN9_LP(i915))
-		bxt_init_cdclk(i915);
 }
 
 /**
@@ -2038,14 +1742,10 @@ void intel_cdclk_init(struct drm_i915_private *i915)
  */
 void intel_cdclk_uninit(struct drm_i915_private *i915)
 {
-	if (INTEL_GEN(i915) >= 11)
-		icl_uninit_cdclk(i915);
-	else if (IS_CANNONLAKE(i915))
-		cnl_uninit_cdclk(i915);
+	if (INTEL_GEN(i915) >= 10 || IS_GEN9_LP(i915))
+		bxt_uninit_cdclk(i915);
 	else if (IS_GEN9_BC(i915))
 		skl_uninit_cdclk(i915);
-	else if (IS_GEN9_LP(i915))
-		bxt_uninit_cdclk(i915);
 }
 
 /**
@@ -2073,9 +1773,9 @@ bool intel_cdclk_needs_modeset(const struct intel_cdclk_state *a,
  * Returns:
  * True if the CDCLK states require just a cd2x divider update, false if not.
  */
-bool intel_cdclk_needs_cd2x_update(struct drm_i915_private *dev_priv,
-				   const struct intel_cdclk_state *a,
-				   const struct intel_cdclk_state *b)
+static bool intel_cdclk_needs_cd2x_update(struct drm_i915_private *dev_priv,
+					  const struct intel_cdclk_state *a,
+					  const struct intel_cdclk_state *b)
 {
 	/* Older hw doesn't have the capability */
 	if (INTEL_GEN(dev_priv) < 10 && !IS_GEN9_LP(dev_priv))
@@ -2094,8 +1794,8 @@ bool intel_cdclk_needs_cd2x_update(struct drm_i915_private *dev_priv,
  * Returns:
  * True if the CDCLK states don't match, false if they do.
  */
-bool intel_cdclk_changed(const struct intel_cdclk_state *a,
-			 const struct intel_cdclk_state *b)
+static bool intel_cdclk_changed(const struct intel_cdclk_state *a,
+				const struct intel_cdclk_state *b)
 {
 	return intel_cdclk_needs_modeset(a, b) ||
 		a->voltage_level != b->voltage_level;
@@ -2200,9 +1900,11 @@ intel_set_cdclk_post_plane_update(struct drm_i915_private *dev_priv,
 		intel_set_cdclk(dev_priv, new_state, pipe);
 }
 
-static int intel_pixel_rate_to_cdclk(struct drm_i915_private *dev_priv,
-				     int pixel_rate)
+static int intel_pixel_rate_to_cdclk(const struct intel_crtc_state *crtc_state)
 {
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	int pixel_rate = crtc_state->pixel_rate;
+
 	if (INTEL_GEN(dev_priv) >= 10 || IS_GEMINILAKE(dev_priv))
 		return DIV_ROUND_UP(pixel_rate, 2);
 	else if (IS_GEN(dev_priv, 9) ||
@@ -2210,10 +1912,25 @@ static int intel_pixel_rate_to_cdclk(struct drm_i915_private *dev_priv,
 		return pixel_rate;
 	else if (IS_CHERRYVIEW(dev_priv))
 		return DIV_ROUND_UP(pixel_rate * 100, 95);
+	else if (crtc_state->double_wide)
+		return DIV_ROUND_UP(pixel_rate * 100, 90 * 2);
 	else
 		return DIV_ROUND_UP(pixel_rate * 100, 90);
 }
 
+static int intel_planes_min_cdclk(const struct intel_crtc_state *crtc_state)
+{
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	struct intel_plane *plane;
+	int min_cdclk = 0;
+
+	for_each_intel_plane_on_crtc(&dev_priv->drm, crtc, plane)
+		min_cdclk = max(crtc_state->min_cdclk[plane->id], min_cdclk);
+
+	return min_cdclk;
+}
+
 int intel_crtc_compute_min_cdclk(const struct intel_crtc_state *crtc_state)
 {
 	struct drm_i915_private *dev_priv =
@@ -2223,7 +1940,7 @@ int intel_crtc_compute_min_cdclk(const struct intel_crtc_state *crtc_state)
 	if (!crtc_state->base.enable)
 		return 0;
 
-	min_cdclk = intel_pixel_rate_to_cdclk(dev_priv, crtc_state->pixel_rate);
+	min_cdclk = intel_pixel_rate_to_cdclk(crtc_state);
 
 	/* pixel rate mustn't exceed 95% of cdclk with IPS on BDW */
 	if (IS_BROADWELL(dev_priv) && hsw_crtc_state_ips_capable(crtc_state))
@@ -2282,6 +1999,9 @@ int intel_crtc_compute_min_cdclk(const struct intel_crtc_state *crtc_state)
 	    IS_GEMINILAKE(dev_priv))
 		min_cdclk = max(158400, min_cdclk);
 
+	/* Account for additional needs from the planes */
+	min_cdclk = max(intel_planes_min_cdclk(crtc_state), min_cdclk);
+
 	if (min_cdclk > dev_priv->max_cdclk_freq) {
 		DRM_DEBUG_KMS("required cdclk (%d kHz) exceeds max (%d kHz)\n",
 			      min_cdclk, dev_priv->max_cdclk_freq);
@@ -2303,11 +2023,20 @@ static int intel_compute_min_cdclk(struct intel_atomic_state *state)
 	       sizeof(state->min_cdclk));
 
 	for_each_new_intel_crtc_in_state(state, crtc, crtc_state, i) {
+		int ret;
+
 		min_cdclk = intel_crtc_compute_min_cdclk(crtc_state);
 		if (min_cdclk < 0)
 			return min_cdclk;
 
+		if (state->min_cdclk[i] == min_cdclk)
+			continue;
+
 		state->min_cdclk[i] = min_cdclk;
+
+		ret = intel_atomic_lock_global_state(state);
+		if (ret)
+			return ret;
 	}
 
 	min_cdclk = state->cdclk.force_min_cdclk;
@@ -2318,6 +2047,10 @@ static int intel_compute_min_cdclk(struct intel_atomic_state *state)
 }
 
 /*
+ * Account for port clock min voltage level requirements.
+ * This only really does something on CNL+ but can be
+ * called on earlier platforms as well.
+ *
  * Note that this functions assumes that 0 is
  * the lowest voltage value, and higher values
  * correspond to increasingly higher voltages.
@@ -2326,7 +2059,7 @@ static int intel_compute_min_cdclk(struct intel_atomic_state *state)
  * future platforms this code will need to be
  * adjusted.
  */
-static u8 cnl_compute_min_voltage_level(struct intel_atomic_state *state)
+static int bxt_compute_min_voltage_level(struct intel_atomic_state *state)
 {
 	struct drm_i915_private *dev_priv = to_i915(state->base.dev);
 	struct intel_crtc *crtc;
@@ -2339,11 +2072,21 @@ static u8 cnl_compute_min_voltage_level(struct intel_atomic_state *state)
 	       sizeof(state->min_voltage_level));
 
 	for_each_new_intel_crtc_in_state(state, crtc, crtc_state, i) {
+		int ret;
+
 		if (crtc_state->base.enable)
-			state->min_voltage_level[i] =
-				crtc_state->min_voltage_level;
+			min_voltage_level = crtc_state->min_voltage_level;
 		else
-			state->min_voltage_level[i] = 0;
+			min_voltage_level = 0;
+
+		if (state->min_voltage_level[i] == min_voltage_level)
+			continue;
+
+		state->min_voltage_level[i] = min_voltage_level;
+
+		ret = intel_atomic_lock_global_state(state);
+		if (ret)
+			return ret;
 	}
 
 	min_voltage_level = 0;
@@ -2369,7 +2112,7 @@ static int vlv_modeset_calc_cdclk(struct intel_atomic_state *state)
 	state->cdclk.logical.voltage_level =
 		vlv_calc_voltage_level(dev_priv, cdclk);
 
-	if (!state->active_crtcs) {
+	if (!state->active_pipes) {
 		cdclk = vlv_calc_cdclk(dev_priv, state->cdclk.force_min_cdclk);
 
 		state->cdclk.actual.cdclk = cdclk;
@@ -2400,7 +2143,7 @@ static int bdw_modeset_calc_cdclk(struct intel_atomic_state *state)
 	state->cdclk.logical.voltage_level =
 		bdw_calc_voltage_level(cdclk);
 
-	if (!state->active_crtcs) {
+	if (!state->active_pipes) {
 		cdclk = bdw_calc_cdclk(state->cdclk.force_min_cdclk);
 
 		state->cdclk.actual.cdclk = cdclk;
@@ -2470,7 +2213,7 @@ static int skl_modeset_calc_cdclk(struct intel_atomic_state *state)
 	state->cdclk.logical.voltage_level =
 		skl_calc_voltage_level(cdclk);
 
-	if (!state->active_crtcs) {
+	if (!state->active_pipes) {
 		cdclk = skl_calc_cdclk(state->cdclk.force_min_cdclk, vco);
 
 		state->cdclk.actual.vco = vco;
@@ -2487,38 +2230,33 @@ static int skl_modeset_calc_cdclk(struct intel_atomic_state *state)
 static int bxt_modeset_calc_cdclk(struct intel_atomic_state *state)
 {
 	struct drm_i915_private *dev_priv = to_i915(state->base.dev);
-	int min_cdclk, cdclk, vco;
+	int min_cdclk, min_voltage_level, cdclk, vco;
 
 	min_cdclk = intel_compute_min_cdclk(state);
 	if (min_cdclk < 0)
 		return min_cdclk;
 
-	if (IS_GEMINILAKE(dev_priv)) {
-		cdclk = glk_calc_cdclk(min_cdclk);
-		vco = glk_de_pll_vco(dev_priv, cdclk);
-	} else {
-		cdclk = bxt_calc_cdclk(min_cdclk);
-		vco = bxt_de_pll_vco(dev_priv, cdclk);
-	}
+	min_voltage_level = bxt_compute_min_voltage_level(state);
+	if (min_voltage_level < 0)
+		return min_voltage_level;
+
+	cdclk = bxt_calc_cdclk(dev_priv, min_cdclk);
+	vco = bxt_calc_cdclk_pll_vco(dev_priv, cdclk);
 
 	state->cdclk.logical.vco = vco;
 	state->cdclk.logical.cdclk = cdclk;
 	state->cdclk.logical.voltage_level =
-		bxt_calc_voltage_level(cdclk);
-
-	if (!state->active_crtcs) {
-		if (IS_GEMINILAKE(dev_priv)) {
-			cdclk = glk_calc_cdclk(state->cdclk.force_min_cdclk);
-			vco = glk_de_pll_vco(dev_priv, cdclk);
-		} else {
-			cdclk = bxt_calc_cdclk(state->cdclk.force_min_cdclk);
-			vco = bxt_de_pll_vco(dev_priv, cdclk);
-		}
+		max_t(int, min_voltage_level,
+		      dev_priv->display.calc_voltage_level(cdclk));
+
+	if (!state->active_pipes) {
+		cdclk = bxt_calc_cdclk(dev_priv, state->cdclk.force_min_cdclk);
+		vco = bxt_calc_cdclk_pll_vco(dev_priv, cdclk);
 
 		state->cdclk.actual.vco = vco;
 		state->cdclk.actual.cdclk = cdclk;
 		state->cdclk.actual.voltage_level =
-			bxt_calc_voltage_level(cdclk);
+			dev_priv->display.calc_voltage_level(cdclk);
 	} else {
 		state->cdclk.actual = state->cdclk.logical;
 	}
@@ -2526,70 +2264,138 @@ static int bxt_modeset_calc_cdclk(struct intel_atomic_state *state)
 	return 0;
 }
 
-static int cnl_modeset_calc_cdclk(struct intel_atomic_state *state)
+static int intel_modeset_all_pipes(struct intel_atomic_state *state)
 {
 	struct drm_i915_private *dev_priv = to_i915(state->base.dev);
-	int min_cdclk, cdclk, vco;
+	struct intel_crtc *crtc;
 
-	min_cdclk = intel_compute_min_cdclk(state);
-	if (min_cdclk < 0)
-		return min_cdclk;
+	/*
+	 * Add all pipes to the state, and force
+	 * a modeset on all the active ones.
+	 */
+	for_each_intel_crtc(&dev_priv->drm, crtc) {
+		struct intel_crtc_state *crtc_state;
+		int ret;
 
-	cdclk = cnl_calc_cdclk(min_cdclk);
-	vco = cnl_cdclk_pll_vco(dev_priv, cdclk);
+		crtc_state = intel_atomic_get_crtc_state(&state->base, crtc);
+		if (IS_ERR(crtc_state))
+			return PTR_ERR(crtc_state);
 
-	state->cdclk.logical.vco = vco;
-	state->cdclk.logical.cdclk = cdclk;
-	state->cdclk.logical.voltage_level =
-		max(cnl_calc_voltage_level(cdclk),
-		    cnl_compute_min_voltage_level(state));
+		if (!crtc_state->base.active ||
+		    drm_atomic_crtc_needs_modeset(&crtc_state->base))
+			continue;
 
-	if (!state->active_crtcs) {
-		cdclk = cnl_calc_cdclk(state->cdclk.force_min_cdclk);
-		vco = cnl_cdclk_pll_vco(dev_priv, cdclk);
+		crtc_state->base.mode_changed = true;
 
-		state->cdclk.actual.vco = vco;
-		state->cdclk.actual.cdclk = cdclk;
-		state->cdclk.actual.voltage_level =
-			cnl_calc_voltage_level(cdclk);
-	} else {
-		state->cdclk.actual = state->cdclk.logical;
+		ret = drm_atomic_add_affected_connectors(&state->base,
+							 &crtc->base);
+		if (ret)
+			return ret;
+
+		ret = drm_atomic_add_affected_planes(&state->base,
+						     &crtc->base);
+		if (ret)
+			return ret;
+
+		crtc_state->update_planes |= crtc_state->active_planes;
 	}
 
 	return 0;
 }
 
-static int icl_modeset_calc_cdclk(struct intel_atomic_state *state)
+static int fixed_modeset_calc_cdclk(struct intel_atomic_state *state)
 {
-	struct drm_i915_private *dev_priv = to_i915(state->base.dev);
-	unsigned int ref = state->cdclk.logical.ref;
-	int min_cdclk, cdclk, vco;
+	int min_cdclk;
 
+	/*
+	 * We can't change the cdclk frequency, but we still want to
+	 * check that the required minimum frequency doesn't exceed
+	 * the actual cdclk frequency.
+	 */
 	min_cdclk = intel_compute_min_cdclk(state);
 	if (min_cdclk < 0)
 		return min_cdclk;
 
-	cdclk = icl_calc_cdclk(min_cdclk, ref);
-	vco = icl_calc_cdclk_pll_vco(dev_priv, cdclk);
+	return 0;
+}
 
-	state->cdclk.logical.vco = vco;
-	state->cdclk.logical.cdclk = cdclk;
-	state->cdclk.logical.voltage_level =
-		max(icl_calc_voltage_level(dev_priv, cdclk),
-		    cnl_compute_min_voltage_level(state));
+int intel_modeset_calc_cdclk(struct intel_atomic_state *state)
+{
+	struct drm_i915_private *dev_priv = to_i915(state->base.dev);
+	enum pipe pipe;
+	int ret;
 
-	if (!state->active_crtcs) {
-		cdclk = icl_calc_cdclk(state->cdclk.force_min_cdclk, ref);
-		vco = icl_calc_cdclk_pll_vco(dev_priv, cdclk);
+	ret = dev_priv->display.modeset_calc_cdclk(state);
+	if (ret)
+		return ret;
 
-		state->cdclk.actual.vco = vco;
-		state->cdclk.actual.cdclk = cdclk;
-		state->cdclk.actual.voltage_level =
-			icl_calc_voltage_level(dev_priv, cdclk);
+	/*
+	 * Writes to dev_priv->cdclk.{actual,logical} must protected
+	 * by holding all the crtc mutexes even if we don't end up
+	 * touching the hardware
+	 */
+	if (intel_cdclk_changed(&dev_priv->cdclk.actual,
+				&state->cdclk.actual)) {
+		/*
+		 * Also serialize commits across all crtcs
+		 * if the actual hw needs to be poked.
+		 */
+		ret = intel_atomic_serialize_global_state(state);
+		if (ret)
+			return ret;
+	} else if (intel_cdclk_changed(&dev_priv->cdclk.logical,
+				       &state->cdclk.logical)) {
+		ret = intel_atomic_lock_global_state(state);
+		if (ret)
+			return ret;
 	} else {
-		state->cdclk.actual = state->cdclk.logical;
+		return 0;
+	}
+
+	if (is_power_of_2(state->active_pipes) &&
+	    intel_cdclk_needs_cd2x_update(dev_priv,
+					  &dev_priv->cdclk.actual,
+					  &state->cdclk.actual)) {
+		struct intel_crtc *crtc;
+		struct intel_crtc_state *crtc_state;
+
+		pipe = ilog2(state->active_pipes);
+		crtc = intel_get_crtc_for_pipe(dev_priv, pipe);
+
+		crtc_state = intel_atomic_get_crtc_state(&state->base, crtc);
+		if (IS_ERR(crtc_state))
+			return PTR_ERR(crtc_state);
+
+		if (drm_atomic_crtc_needs_modeset(&crtc_state->base))
+			pipe = INVALID_PIPE;
+	} else {
+		pipe = INVALID_PIPE;
+	}
+
+	if (pipe != INVALID_PIPE) {
+		state->cdclk.pipe = pipe;
+
+		DRM_DEBUG_KMS("Can change cdclk with pipe %c active\n",
+			      pipe_name(pipe));
+	} else if (intel_cdclk_needs_modeset(&dev_priv->cdclk.actual,
+					     &state->cdclk.actual)) {
+		/* All pipes must be switched off while we change the cdclk. */
+		ret = intel_modeset_all_pipes(state);
+		if (ret)
+			return ret;
+
+		state->cdclk.pipe = INVALID_PIPE;
+
+		DRM_DEBUG_KMS("Modeset required for cdclk change\n");
 	}
 
+	DRM_DEBUG_KMS("New cdclk calculated to be logical %u kHz, actual %u kHz\n",
+		      state->cdclk.logical.cdclk,
+		      state->cdclk.actual.cdclk);
+	DRM_DEBUG_KMS("New voltage level calculated to be logical %u, actual %u\n",
+		      state->cdclk.logical.voltage_level,
+		      state->cdclk.actual.voltage_level);
+
 	return 0;
 }
 
@@ -2809,15 +2615,29 @@ void intel_update_rawclk(struct drm_i915_private *dev_priv)
  */
 void intel_init_cdclk_hooks(struct drm_i915_private *dev_priv)
 {
-	if (INTEL_GEN(dev_priv) >= 11) {
-		dev_priv->display.set_cdclk = icl_set_cdclk;
-		dev_priv->display.modeset_calc_cdclk = icl_modeset_calc_cdclk;
+	if (IS_ELKHARTLAKE(dev_priv)) {
+		dev_priv->display.set_cdclk = bxt_set_cdclk;
+		dev_priv->display.modeset_calc_cdclk = bxt_modeset_calc_cdclk;
+		dev_priv->display.calc_voltage_level = ehl_calc_voltage_level;
+		dev_priv->cdclk.table = icl_cdclk_table;
+	} else if (INTEL_GEN(dev_priv) >= 11) {
+		dev_priv->display.set_cdclk = bxt_set_cdclk;
+		dev_priv->display.modeset_calc_cdclk = bxt_modeset_calc_cdclk;
+		dev_priv->display.calc_voltage_level = icl_calc_voltage_level;
+		dev_priv->cdclk.table = icl_cdclk_table;
 	} else if (IS_CANNONLAKE(dev_priv)) {
-		dev_priv->display.set_cdclk = cnl_set_cdclk;
-		dev_priv->display.modeset_calc_cdclk = cnl_modeset_calc_cdclk;
+		dev_priv->display.set_cdclk = bxt_set_cdclk;
+		dev_priv->display.modeset_calc_cdclk = bxt_modeset_calc_cdclk;
+		dev_priv->display.calc_voltage_level = cnl_calc_voltage_level;
+		dev_priv->cdclk.table = cnl_cdclk_table;
 	} else if (IS_GEN9_LP(dev_priv)) {
 		dev_priv->display.set_cdclk = bxt_set_cdclk;
 		dev_priv->display.modeset_calc_cdclk = bxt_modeset_calc_cdclk;
+		dev_priv->display.calc_voltage_level = bxt_calc_voltage_level;
+		if (IS_GEMINILAKE(dev_priv))
+			dev_priv->cdclk.table = glk_cdclk_table;
+		else
+			dev_priv->cdclk.table = bxt_cdclk_table;
 	} else if (IS_GEN9_BC(dev_priv)) {
 		dev_priv->display.set_cdclk = skl_set_cdclk;
 		dev_priv->display.modeset_calc_cdclk = skl_modeset_calc_cdclk;
@@ -2830,13 +2650,11 @@ void intel_init_cdclk_hooks(struct drm_i915_private *dev_priv)
 	} else if (IS_VALLEYVIEW(dev_priv)) {
 		dev_priv->display.set_cdclk = vlv_set_cdclk;
 		dev_priv->display.modeset_calc_cdclk = vlv_modeset_calc_cdclk;
+	} else {
+		dev_priv->display.modeset_calc_cdclk = fixed_modeset_calc_cdclk;
 	}
 
-	if (INTEL_GEN(dev_priv) >= 11)
-		dev_priv->display.get_cdclk = icl_get_cdclk;
-	else if (IS_CANNONLAKE(dev_priv))
-		dev_priv->display.get_cdclk = cnl_get_cdclk;
-	else if (IS_GEN9_LP(dev_priv))
+	if (INTEL_GEN(dev_priv) >= 10 || IS_GEN9_LP(dev_priv))
 		dev_priv->display.get_cdclk = bxt_get_cdclk;
 	else if (IS_GEN9_BC(dev_priv))
 		dev_priv->display.get_cdclk = skl_get_cdclk;
diff --git a/drivers/gpu/drm/i915/display/intel_cdclk.h b/drivers/gpu/drm/i915/display/intel_cdclk.h
index 4d6f7f5f8930..cf71394cc79c 100644
--- a/drivers/gpu/drm/i915/display/intel_cdclk.h
+++ b/drivers/gpu/drm/i915/display/intel_cdclk.h
@@ -15,6 +15,13 @@ struct intel_atomic_state;
 struct intel_cdclk_state;
 struct intel_crtc_state;
 
+struct intel_cdclk_vals {
+	u16 refclk;
+	u32 cdclk;
+	u8 divider;	/* CD2X divider * 2 */
+	u8 ratio;
+};
+
 int intel_crtc_compute_min_cdclk(const struct intel_crtc_state *crtc_state);
 void intel_cdclk_init(struct drm_i915_private *i915);
 void intel_cdclk_uninit(struct drm_i915_private *i915);
@@ -22,13 +29,8 @@ void intel_init_cdclk_hooks(struct drm_i915_private *dev_priv);
 void intel_update_max_cdclk(struct drm_i915_private *dev_priv);
 void intel_update_cdclk(struct drm_i915_private *dev_priv);
 void intel_update_rawclk(struct drm_i915_private *dev_priv);
-bool intel_cdclk_needs_cd2x_update(struct drm_i915_private *dev_priv,
-				   const struct intel_cdclk_state *a,
-				   const struct intel_cdclk_state *b);
 bool intel_cdclk_needs_modeset(const struct intel_cdclk_state *a,
 			       const struct intel_cdclk_state *b);
-bool intel_cdclk_changed(const struct intel_cdclk_state *a,
-			 const struct intel_cdclk_state *b);
 void intel_cdclk_swap_state(struct intel_atomic_state *state);
 void
 intel_set_cdclk_pre_plane_update(struct drm_i915_private *dev_priv,
@@ -42,5 +44,6 @@ intel_set_cdclk_post_plane_update(struct drm_i915_private *dev_priv,
 				  enum pipe pipe);
 void intel_dump_cdclk_state(const struct intel_cdclk_state *cdclk_state,
 			    const char *context);
+int intel_modeset_calc_cdclk(struct intel_atomic_state *state);
 
 #endif /* __INTEL_CDCLK_H__ */
diff --git a/drivers/gpu/drm/i915/display/intel_color.c b/drivers/gpu/drm/i915/display/intel_color.c
index 71a0201437a9..aa3a063549c3 100644
--- a/drivers/gpu/drm/i915/display/intel_color.c
+++ b/drivers/gpu/drm/i915/display/intel_color.c
@@ -43,6 +43,21 @@
 #define LEGACY_LUT_LENGTH		256
 
 /*
+ * ILK+ csc matrix:
+ *
+ * |R/Cr|   | c0 c1 c2 |   ( |R/Cr|   |preoff0| )   |postoff0|
+ * |G/Y | = | c3 c4 c5 | x ( |G/Y | + |preoff1| ) + |postoff1|
+ * |B/Cb|   | c6 c7 c8 |   ( |B/Cb|   |preoff2| )   |postoff2|
+ *
+ * ILK/SNB don't have explicit post offsets, and instead
+ * CSC_MODE_YUV_TO_RGB and CSC_BLACK_SCREEN_OFFSET are used:
+ *  CSC_MODE_YUV_TO_RGB=0 + CSC_BLACK_SCREEN_OFFSET=0 -> 1/2, 0, 1/2
+ *  CSC_MODE_YUV_TO_RGB=0 + CSC_BLACK_SCREEN_OFFSET=1 -> 1/2, 1/16, 1/2
+ *  CSC_MODE_YUV_TO_RGB=1 + CSC_BLACK_SCREEN_OFFSET=0 -> 0, 0, 0
+ *  CSC_MODE_YUV_TO_RGB=1 + CSC_BLACK_SCREEN_OFFSET=1 -> 1/16, 1/16, 1/16
+ */
+
+/*
  * Extract the CSC coefficient from a CTM coefficient (in U32.32 fixed point
  * format). This macro takes the coefficient we want transformed and the
  * number of fractional bits.
@@ -59,37 +74,38 @@
 
 #define ILK_CSC_POSTOFF_LIMITED_RANGE (16 * (1 << 12) / 255)
 
+/* Nop pre/post offsets */
 static const u16 ilk_csc_off_zero[3] = {};
 
+/* Identity matrix */
 static const u16 ilk_csc_coeff_identity[9] = {
 	ILK_CSC_COEFF_1_0, 0, 0,
 	0, ILK_CSC_COEFF_1_0, 0,
 	0, 0, ILK_CSC_COEFF_1_0,
 };
 
+/* Limited range RGB post offsets */
 static const u16 ilk_csc_postoff_limited_range[3] = {
 	ILK_CSC_POSTOFF_LIMITED_RANGE,
 	ILK_CSC_POSTOFF_LIMITED_RANGE,
 	ILK_CSC_POSTOFF_LIMITED_RANGE,
 };
 
+/* Full range RGB -> limited range RGB matrix */
 static const u16 ilk_csc_coeff_limited_range[9] = {
 	ILK_CSC_COEFF_LIMITED_RANGE, 0, 0,
 	0, ILK_CSC_COEFF_LIMITED_RANGE, 0,
 	0, 0, ILK_CSC_COEFF_LIMITED_RANGE,
 };
 
-/*
- * These values are direct register values specified in the Bspec,
- * for RGB->YUV conversion matrix (colorspace BT709)
- */
+/* BT.709 full range RGB -> limited range YCbCr matrix */
 static const u16 ilk_csc_coeff_rgb_to_ycbcr[9] = {
 	0x1e08, 0x9cc0, 0xb528,
 	0x2ba8, 0x09d8, 0x37e8,
 	0xbce8, 0x9ad8, 0x1e08,
 };
 
-/* Post offset values for RGB->YCBCR conversion */
+/* Limited range YCbCr post offsets */
 static const u16 ilk_csc_postoff_rgb_to_ycbcr[3] = {
 	0x0800, 0x0100, 0x0800,
 };
@@ -611,12 +627,13 @@ static void bdw_load_lut_10(struct intel_crtc *crtc,
 static void ivb_load_lut_ext_max(struct intel_crtc *crtc)
 {
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	struct intel_dsb *dsb = intel_dsb_get(crtc);
 	enum pipe pipe = crtc->pipe;
 
 	/* Program the max register to clamp values > 1.0. */
-	I915_WRITE(PREC_PAL_EXT_GC_MAX(pipe, 0), 1 << 16);
-	I915_WRITE(PREC_PAL_EXT_GC_MAX(pipe, 1), 1 << 16);
-	I915_WRITE(PREC_PAL_EXT_GC_MAX(pipe, 2), 1 << 16);
+	intel_dsb_reg_write(dsb, PREC_PAL_EXT_GC_MAX(pipe, 0), 1 << 16);
+	intel_dsb_reg_write(dsb, PREC_PAL_EXT_GC_MAX(pipe, 1), 1 << 16);
+	intel_dsb_reg_write(dsb, PREC_PAL_EXT_GC_MAX(pipe, 2), 1 << 16);
 
 	/*
 	 * Program the gc max 2 register to clamp values > 1.0.
@@ -624,10 +641,15 @@ static void ivb_load_lut_ext_max(struct intel_crtc *crtc)
 	 * from 3.0 to 7.0
 	 */
 	if (INTEL_GEN(dev_priv) >= 10 || IS_GEMINILAKE(dev_priv)) {
-		I915_WRITE(PREC_PAL_EXT2_GC_MAX(pipe, 0), 1 << 16);
-		I915_WRITE(PREC_PAL_EXT2_GC_MAX(pipe, 1), 1 << 16);
-		I915_WRITE(PREC_PAL_EXT2_GC_MAX(pipe, 2), 1 << 16);
+		intel_dsb_reg_write(dsb, PREC_PAL_EXT2_GC_MAX(pipe, 0),
+				    1 << 16);
+		intel_dsb_reg_write(dsb, PREC_PAL_EXT2_GC_MAX(pipe, 1),
+				    1 << 16);
+		intel_dsb_reg_write(dsb, PREC_PAL_EXT2_GC_MAX(pipe, 2),
+				    1 << 16);
 	}
+
+	intel_dsb_put(dsb);
 }
 
 static void ivb_load_luts(const struct intel_crtc_state *crtc_state)
@@ -787,78 +809,83 @@ icl_load_gcmax(const struct intel_crtc_state *crtc_state,
 	       const struct drm_color_lut *color)
 {
 	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
-	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	struct intel_dsb *dsb = intel_dsb_get(crtc);
 	enum pipe pipe = crtc->pipe;
 
 	/* Fixme: LUT entries are 16 bit only, so we can prog 0xFFFF max */
-	I915_WRITE(PREC_PAL_GC_MAX(pipe, 0), color->red);
-	I915_WRITE(PREC_PAL_GC_MAX(pipe, 1), color->green);
-	I915_WRITE(PREC_PAL_GC_MAX(pipe, 2), color->blue);
+	intel_dsb_reg_write(dsb, PREC_PAL_GC_MAX(pipe, 0), color->red);
+	intel_dsb_reg_write(dsb, PREC_PAL_GC_MAX(pipe, 1), color->green);
+	intel_dsb_reg_write(dsb, PREC_PAL_GC_MAX(pipe, 2), color->blue);
+	intel_dsb_put(dsb);
 }
 
 static void
 icl_program_gamma_superfine_segment(const struct intel_crtc_state *crtc_state)
 {
 	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
-	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	const struct drm_property_blob *blob = crtc_state->base.gamma_lut;
 	const struct drm_color_lut *lut = blob->data;
+	struct intel_dsb *dsb = intel_dsb_get(crtc);
 	enum pipe pipe = crtc->pipe;
 	u32 i;
 
 	/*
-	 * Every entry in the multi-segment LUT is corresponding to a superfine
-	 * segment step which is 1/(8 * 128 * 256).
+	 * Program Super Fine segment (let's call it seg1)...
 	 *
-	 * Superfine segment has 9 entries, corresponding to values
-	 * 0, 1/(8 * 128 * 256), 2/(8 * 128 * 256) .... 8/(8 * 128 * 256).
+	 * Super Fine segment's step is 1/(8 * 128 * 256) and it has
+	 * 9 entries, corresponding to values 0, 1/(8 * 128 * 256),
+	 * 2/(8 * 128 * 256) ... 8/(8 * 128 * 256).
 	 */
-	I915_WRITE(PREC_PAL_MULTI_SEG_INDEX(pipe), PAL_PREC_AUTO_INCREMENT);
+	intel_dsb_reg_write(dsb, PREC_PAL_MULTI_SEG_INDEX(pipe),
+			    PAL_PREC_AUTO_INCREMENT);
 
 	for (i = 0; i < 9; i++) {
 		const struct drm_color_lut *entry = &lut[i];
 
-		I915_WRITE(PREC_PAL_MULTI_SEG_DATA(pipe),
-			   ilk_lut_12p4_ldw(entry));
-		I915_WRITE(PREC_PAL_MULTI_SEG_DATA(pipe),
-			   ilk_lut_12p4_udw(entry));
+		intel_dsb_indexed_reg_write(dsb, PREC_PAL_MULTI_SEG_DATA(pipe),
+					    ilk_lut_12p4_ldw(entry));
+		intel_dsb_indexed_reg_write(dsb, PREC_PAL_MULTI_SEG_DATA(pipe),
+					    ilk_lut_12p4_udw(entry));
 	}
+
+	intel_dsb_put(dsb);
 }
 
 static void
 icl_program_gamma_multi_segment(const struct intel_crtc_state *crtc_state)
 {
 	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
-	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	const struct drm_property_blob *blob = crtc_state->base.gamma_lut;
 	const struct drm_color_lut *lut = blob->data;
 	const struct drm_color_lut *entry;
+	struct intel_dsb *dsb = intel_dsb_get(crtc);
 	enum pipe pipe = crtc->pipe;
 	u32 i;
 
 	/*
-	 *
 	 * Program Fine segment (let's call it seg2)...
 	 *
-	 * Fine segment's step is 1/(128 * 256) ie 1/(128 * 256),  2/(128*256)
-	 * ... 256/(128*256). So in order to program fine segment of LUT we
-	 * need to pick every 8'th entry in LUT, and program 256 indexes.
+	 * Fine segment's step is 1/(128 * 256) i.e. 1/(128 * 256), 2/(128 * 256)
+	 * ... 256/(128 * 256). So in order to program fine segment of LUT we
+	 * need to pick every 8th entry in the LUT, and program 256 indexes.
 	 *
 	 * PAL_PREC_INDEX[0] and PAL_PREC_INDEX[1] map to seg2[1],
-	 * with seg2[0] being unused by the hardware.
+	 * seg2[0] being unused by the hardware.
 	 */
-	I915_WRITE(PREC_PAL_INDEX(pipe), PAL_PREC_AUTO_INCREMENT);
+	intel_dsb_reg_write(dsb, PREC_PAL_INDEX(pipe), PAL_PREC_AUTO_INCREMENT);
 	for (i = 1; i < 257; i++) {
 		entry = &lut[i * 8];
-		I915_WRITE(PREC_PAL_DATA(pipe), ilk_lut_12p4_ldw(entry));
-		I915_WRITE(PREC_PAL_DATA(pipe), ilk_lut_12p4_udw(entry));
+		intel_dsb_indexed_reg_write(dsb, PREC_PAL_DATA(pipe),
+					    ilk_lut_12p4_ldw(entry));
+		intel_dsb_indexed_reg_write(dsb, PREC_PAL_DATA(pipe),
+					    ilk_lut_12p4_udw(entry));
 	}
 
 	/*
 	 * Program Coarse segment (let's call it seg3)...
 	 *
-	 * Coarse segment's starts from index 0 and it's step is 1/256 ie 0,
-	 * 1/256, 2/256 ...256/256. As per the description of each entry in LUT
+	 * Coarse segment starts from index 0 and it's step is 1/256 ie 0,
+	 * 1/256, 2/256 ... 256/256. As per the description of each entry in LUT
 	 * above, we need to pick every (8 * 128)th entry in LUT, and
 	 * program 256 of those.
 	 *
@@ -868,20 +895,24 @@ icl_program_gamma_multi_segment(const struct intel_crtc_state *crtc_state)
 	 */
 	for (i = 0; i < 256; i++) {
 		entry = &lut[i * 8 * 128];
-		I915_WRITE(PREC_PAL_DATA(pipe), ilk_lut_12p4_ldw(entry));
-		I915_WRITE(PREC_PAL_DATA(pipe), ilk_lut_12p4_udw(entry));
+		intel_dsb_indexed_reg_write(dsb, PREC_PAL_DATA(pipe),
+					    ilk_lut_12p4_ldw(entry));
+		intel_dsb_indexed_reg_write(dsb, PREC_PAL_DATA(pipe),
+					    ilk_lut_12p4_udw(entry));
 	}
 
 	/* The last entry in the LUT is to be programmed in GCMAX */
 	entry = &lut[256 * 8 * 128];
 	icl_load_gcmax(crtc_state, entry);
 	ivb_load_lut_ext_max(crtc);
+	intel_dsb_put(dsb);
 }
 
 static void icl_load_luts(const struct intel_crtc_state *crtc_state)
 {
 	const struct drm_property_blob *gamma_lut = crtc_state->base.gamma_lut;
 	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_dsb *dsb = intel_dsb_get(crtc);
 
 	if (crtc_state->base.degamma_lut)
 		glk_load_degamma_lut(crtc_state);
@@ -890,16 +921,17 @@ static void icl_load_luts(const struct intel_crtc_state *crtc_state)
 	case GAMMA_MODE_MODE_8BIT:
 		i9xx_load_luts(crtc_state);
 		break;
-
 	case GAMMA_MODE_MODE_12BIT_MULTI_SEGMENTED:
 		icl_program_gamma_superfine_segment(crtc_state);
 		icl_program_gamma_multi_segment(crtc_state);
 		break;
-
 	default:
 		bdw_load_lut_10(crtc, gamma_lut, PAL_PREC_INDEX_VALUE(0));
 		ivb_load_lut_ext_max(crtc);
 	}
+
+	intel_dsb_commit(dsb);
+	intel_dsb_put(dsb);
 }
 
 static u32 chv_cgm_degamma_ldw(const struct drm_color_lut *color)
@@ -990,6 +1022,55 @@ void intel_color_commit(const struct intel_crtc_state *crtc_state)
 	dev_priv->display.color_commit(crtc_state);
 }
 
+static bool intel_can_preload_luts(const struct intel_crtc_state *new_crtc_state)
+{
+	struct intel_crtc *crtc = to_intel_crtc(new_crtc_state->base.crtc);
+	struct intel_atomic_state *state =
+		to_intel_atomic_state(new_crtc_state->base.state);
+	const struct intel_crtc_state *old_crtc_state =
+		intel_atomic_get_old_crtc_state(state, crtc);
+
+	return !old_crtc_state->base.gamma_lut &&
+		!old_crtc_state->base.degamma_lut;
+}
+
+static bool chv_can_preload_luts(const struct intel_crtc_state *new_crtc_state)
+{
+	struct intel_crtc *crtc = to_intel_crtc(new_crtc_state->base.crtc);
+	struct intel_atomic_state *state =
+		to_intel_atomic_state(new_crtc_state->base.state);
+	const struct intel_crtc_state *old_crtc_state =
+		intel_atomic_get_old_crtc_state(state, crtc);
+
+	/*
+	 * CGM_PIPE_MODE is itself single buffered. We'd have to
+	 * somehow split it out from chv_load_luts() if we wanted
+	 * the ability to preload the CGM LUTs/CSC without tearing.
+	 */
+	if (old_crtc_state->cgm_mode || new_crtc_state->cgm_mode)
+		return false;
+
+	return !old_crtc_state->base.gamma_lut;
+}
+
+static bool glk_can_preload_luts(const struct intel_crtc_state *new_crtc_state)
+{
+	struct intel_crtc *crtc = to_intel_crtc(new_crtc_state->base.crtc);
+	struct intel_atomic_state *state =
+		to_intel_atomic_state(new_crtc_state->base.state);
+	const struct intel_crtc_state *old_crtc_state =
+		intel_atomic_get_old_crtc_state(state, crtc);
+
+	/*
+	 * The hardware degamma is active whenever the pipe
+	 * CSC is active. Thus even if the old state has no
+	 * software degamma we need to avoid clobbering the
+	 * linear hardware degamma mid scanout.
+	 */
+	return !old_crtc_state->csc_enable &&
+		!old_crtc_state->base.gamma_lut;
+}
+
 int intel_color_check(struct intel_crtc_state *crtc_state)
 {
 	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
@@ -1133,6 +1214,8 @@ static int i9xx_color_check(struct intel_crtc_state *crtc_state)
 	if (ret)
 		return ret;
 
+	crtc_state->preload_luts = intel_can_preload_luts(crtc_state);
+
 	return 0;
 }
 
@@ -1185,6 +1268,8 @@ static int chv_color_check(struct intel_crtc_state *crtc_state)
 	if (ret)
 		return ret;
 
+	crtc_state->preload_luts = chv_can_preload_luts(crtc_state);
+
 	return 0;
 }
 
@@ -1197,6 +1282,21 @@ static u32 ilk_gamma_mode(const struct intel_crtc_state *crtc_state)
 		return GAMMA_MODE_MODE_10BIT;
 }
 
+static u32 ilk_csc_mode(const struct intel_crtc_state *crtc_state)
+{
+	/*
+	 * CSC comes after the LUT in RGB->YCbCr mode.
+	 * RGB->YCbCr needs the limited range offsets added to
+	 * the output. RGB limited range output is handled by
+	 * the hw automagically elsewhere.
+	 */
+	if (crtc_state->output_format != INTEL_OUTPUT_FORMAT_RGB)
+		return CSC_BLACK_SCREEN_OFFSET;
+
+	return CSC_MODE_YUV_TO_RGB |
+		CSC_POSITION_BEFORE_GAMMA;
+}
+
 static int ilk_color_check(struct intel_crtc_state *crtc_state)
 {
 	int ret;
@@ -1210,20 +1310,22 @@ static int ilk_color_check(struct intel_crtc_state *crtc_state)
 		!crtc_state->c8_planes;
 
 	/*
-	 * We don't expose the ctm on ilk/snb currently,
-	 * nor do we enable YCbCr output. Also RGB limited
-	 * range output is handled by the hw automagically.
+	 * We don't expose the ctm on ilk/snb currently, also RGB
+	 * limited range output is handled by the hw automagically.
 	 */
-	crtc_state->csc_enable = false;
+	crtc_state->csc_enable =
+		crtc_state->output_format != INTEL_OUTPUT_FORMAT_RGB;
 
 	crtc_state->gamma_mode = ilk_gamma_mode(crtc_state);
 
-	crtc_state->csc_mode = 0;
+	crtc_state->csc_mode = ilk_csc_mode(crtc_state);
 
 	ret = intel_color_add_affected_planes(crtc_state);
 	if (ret)
 		return ret;
 
+	crtc_state->preload_luts = intel_can_preload_luts(crtc_state);
+
 	return 0;
 }
 
@@ -1281,6 +1383,8 @@ static int ivb_color_check(struct intel_crtc_state *crtc_state)
 	if (ret)
 		return ret;
 
+	crtc_state->preload_luts = intel_can_preload_luts(crtc_state);
+
 	return 0;
 }
 
@@ -1319,6 +1423,8 @@ static int glk_color_check(struct intel_crtc_state *crtc_state)
 	if (ret)
 		return ret;
 
+	crtc_state->preload_luts = glk_can_preload_luts(crtc_state);
+
 	return 0;
 }
 
@@ -1368,9 +1474,408 @@ static int icl_color_check(struct intel_crtc_state *crtc_state)
 
 	crtc_state->csc_mode = icl_csc_mode(crtc_state);
 
+	crtc_state->preload_luts = intel_can_preload_luts(crtc_state);
+
 	return 0;
 }
 
+static int i9xx_gamma_precision(const struct intel_crtc_state *crtc_state)
+{
+	if (!crtc_state->gamma_enable)
+		return 0;
+
+	switch (crtc_state->gamma_mode) {
+	case GAMMA_MODE_MODE_8BIT:
+		return 8;
+	case GAMMA_MODE_MODE_10BIT:
+		return 16;
+	default:
+		MISSING_CASE(crtc_state->gamma_mode);
+		return 0;
+	}
+}
+
+static int ilk_gamma_precision(const struct intel_crtc_state *crtc_state)
+{
+	if (!crtc_state->gamma_enable)
+		return 0;
+
+	if ((crtc_state->csc_mode & CSC_POSITION_BEFORE_GAMMA) == 0)
+		return 0;
+
+	switch (crtc_state->gamma_mode) {
+	case GAMMA_MODE_MODE_8BIT:
+		return 8;
+	case GAMMA_MODE_MODE_10BIT:
+		return 10;
+	default:
+		MISSING_CASE(crtc_state->gamma_mode);
+		return 0;
+	}
+}
+
+static int chv_gamma_precision(const struct intel_crtc_state *crtc_state)
+{
+	if (crtc_state->cgm_mode & CGM_PIPE_MODE_GAMMA)
+		return 10;
+	else
+		return i9xx_gamma_precision(crtc_state);
+}
+
+static int glk_gamma_precision(const struct intel_crtc_state *crtc_state)
+{
+	if (!crtc_state->gamma_enable)
+		return 0;
+
+	switch (crtc_state->gamma_mode) {
+	case GAMMA_MODE_MODE_8BIT:
+		return 8;
+	case GAMMA_MODE_MODE_10BIT:
+		return 10;
+	default:
+		MISSING_CASE(crtc_state->gamma_mode);
+		return 0;
+	}
+}
+
+int intel_color_get_gamma_bit_precision(const struct intel_crtc_state *crtc_state)
+{
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+
+	if (HAS_GMCH(dev_priv)) {
+		if (IS_CHERRYVIEW(dev_priv))
+			return chv_gamma_precision(crtc_state);
+		else
+			return i9xx_gamma_precision(crtc_state);
+	} else {
+		if (IS_CANNONLAKE(dev_priv) || IS_GEMINILAKE(dev_priv))
+			return glk_gamma_precision(crtc_state);
+		else if (IS_IRONLAKE(dev_priv))
+			return ilk_gamma_precision(crtc_state);
+	}
+
+	return 0;
+}
+
+static bool err_check(struct drm_color_lut *lut1,
+		      struct drm_color_lut *lut2, u32 err)
+{
+	return ((abs((long)lut2->red - lut1->red)) <= err) &&
+		((abs((long)lut2->blue - lut1->blue)) <= err) &&
+		((abs((long)lut2->green - lut1->green)) <= err);
+}
+
+static bool intel_color_lut_entry_equal(struct drm_color_lut *lut1,
+					struct drm_color_lut *lut2,
+					int lut_size, u32 err)
+{
+	int i;
+
+	for (i = 0; i < lut_size; i++) {
+		if (!err_check(&lut1[i], &lut2[i], err))
+			return false;
+	}
+
+	return true;
+}
+
+bool intel_color_lut_equal(struct drm_property_blob *blob1,
+			   struct drm_property_blob *blob2,
+			   u32 gamma_mode, u32 bit_precision)
+{
+	struct drm_color_lut *lut1, *lut2;
+	int lut_size1, lut_size2;
+	u32 err;
+
+	if (!blob1 != !blob2)
+		return false;
+
+	if (!blob1)
+		return true;
+
+	lut_size1 = drm_color_lut_size(blob1);
+	lut_size2 = drm_color_lut_size(blob2);
+
+	/* check sw and hw lut size */
+	switch (gamma_mode) {
+	case GAMMA_MODE_MODE_8BIT:
+	case GAMMA_MODE_MODE_10BIT:
+		if (lut_size1 != lut_size2)
+			return false;
+		break;
+	default:
+		MISSING_CASE(gamma_mode);
+			return false;
+	}
+
+	lut1 = blob1->data;
+	lut2 = blob2->data;
+
+	err = 0xffff >> bit_precision;
+
+	/* check sw and hw lut entry to be equal */
+	switch (gamma_mode) {
+	case GAMMA_MODE_MODE_8BIT:
+	case GAMMA_MODE_MODE_10BIT:
+		if (!intel_color_lut_entry_equal(lut1, lut2,
+						 lut_size2, err))
+			return false;
+		break;
+	default:
+		MISSING_CASE(gamma_mode);
+			return false;
+	}
+
+	return true;
+}
+
+/* convert hw value with given bit_precision to lut property val */
+static u32 intel_color_lut_pack(u32 val, u32 bit_precision)
+{
+	u32 max = 0xffff >> (16 - bit_precision);
+
+	val = clamp_val(val, 0, max);
+
+	if (bit_precision < 16)
+		val <<= 16 - bit_precision;
+
+	return val;
+}
+
+static struct drm_property_blob *
+i9xx_read_lut_8(const struct intel_crtc_state *crtc_state)
+{
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	enum pipe pipe = crtc->pipe;
+	struct drm_property_blob *blob;
+	struct drm_color_lut *blob_data;
+	u32 i, val;
+
+	blob = drm_property_create_blob(&dev_priv->drm,
+					sizeof(struct drm_color_lut) * LEGACY_LUT_LENGTH,
+					NULL);
+	if (IS_ERR(blob))
+		return NULL;
+
+	blob_data = blob->data;
+
+	for (i = 0; i < LEGACY_LUT_LENGTH; i++) {
+		if (HAS_GMCH(dev_priv))
+			val = I915_READ(PALETTE(pipe, i));
+		else
+			val = I915_READ(LGC_PALETTE(pipe, i));
+
+		blob_data[i].red = intel_color_lut_pack(REG_FIELD_GET(
+							LGC_PALETTE_RED_MASK, val), 8);
+		blob_data[i].green = intel_color_lut_pack(REG_FIELD_GET(
+							  LGC_PALETTE_GREEN_MASK, val), 8);
+		blob_data[i].blue = intel_color_lut_pack(REG_FIELD_GET(
+							 LGC_PALETTE_BLUE_MASK, val), 8);
+	}
+
+	return blob;
+}
+
+static void i9xx_read_luts(struct intel_crtc_state *crtc_state)
+{
+	if (!crtc_state->gamma_enable)
+		return;
+
+	crtc_state->base.gamma_lut = i9xx_read_lut_8(crtc_state);
+}
+
+static struct drm_property_blob *
+i965_read_lut_10p6(const struct intel_crtc_state *crtc_state)
+{
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	u32 lut_size = INTEL_INFO(dev_priv)->color.gamma_lut_size;
+	enum pipe pipe = crtc->pipe;
+	struct drm_property_blob *blob;
+	struct drm_color_lut *blob_data;
+	u32 i, val1, val2;
+
+	blob = drm_property_create_blob(&dev_priv->drm,
+					sizeof(struct drm_color_lut) * lut_size,
+					NULL);
+	if (IS_ERR(blob))
+		return NULL;
+
+	blob_data = blob->data;
+
+	for (i = 0; i < lut_size - 1; i++) {
+		val1 = I915_READ(PALETTE(pipe, 2 * i + 0));
+		val2 = I915_READ(PALETTE(pipe, 2 * i + 1));
+
+		blob_data[i].red = REG_FIELD_GET(PALETTE_RED_MASK, val2) << 8 |
+						 REG_FIELD_GET(PALETTE_RED_MASK, val1);
+		blob_data[i].green = REG_FIELD_GET(PALETTE_GREEN_MASK, val2) << 8 |
+						   REG_FIELD_GET(PALETTE_GREEN_MASK, val1);
+		blob_data[i].blue = REG_FIELD_GET(PALETTE_BLUE_MASK, val2) << 8 |
+						  REG_FIELD_GET(PALETTE_BLUE_MASK, val1);
+	}
+
+	blob_data[i].red = REG_FIELD_GET(PIPEGCMAX_RGB_MASK,
+					 I915_READ(PIPEGCMAX(pipe, 0)));
+	blob_data[i].green = REG_FIELD_GET(PIPEGCMAX_RGB_MASK,
+					   I915_READ(PIPEGCMAX(pipe, 1)));
+	blob_data[i].blue = REG_FIELD_GET(PIPEGCMAX_RGB_MASK,
+					  I915_READ(PIPEGCMAX(pipe, 2)));
+
+	return blob;
+}
+
+static void i965_read_luts(struct intel_crtc_state *crtc_state)
+{
+	if (!crtc_state->gamma_enable)
+		return;
+
+	if (crtc_state->gamma_mode == GAMMA_MODE_MODE_8BIT)
+		crtc_state->base.gamma_lut = i9xx_read_lut_8(crtc_state);
+	else
+		crtc_state->base.gamma_lut = i965_read_lut_10p6(crtc_state);
+}
+
+static struct drm_property_blob *
+chv_read_cgm_lut(const struct intel_crtc_state *crtc_state)
+{
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	u32 lut_size = INTEL_INFO(dev_priv)->color.gamma_lut_size;
+	enum pipe pipe = crtc->pipe;
+	struct drm_property_blob *blob;
+	struct drm_color_lut *blob_data;
+	u32 i, val;
+
+	blob = drm_property_create_blob(&dev_priv->drm,
+					sizeof(struct drm_color_lut) * lut_size,
+					NULL);
+	if (IS_ERR(blob))
+		return NULL;
+
+	blob_data = blob->data;
+
+	for (i = 0; i < lut_size; i++) {
+		val = I915_READ(CGM_PIPE_GAMMA(pipe, i, 0));
+		blob_data[i].green = intel_color_lut_pack(REG_FIELD_GET(
+							  CGM_PIPE_GAMMA_GREEN_MASK, val), 10);
+		blob_data[i].blue = intel_color_lut_pack(REG_FIELD_GET(
+							 CGM_PIPE_GAMMA_BLUE_MASK, val), 10);
+
+		val = I915_READ(CGM_PIPE_GAMMA(pipe, i, 1));
+		blob_data[i].red = intel_color_lut_pack(REG_FIELD_GET(
+							CGM_PIPE_GAMMA_RED_MASK, val), 10);
+	}
+
+	return blob;
+}
+
+static void chv_read_luts(struct intel_crtc_state *crtc_state)
+{
+	if (crtc_state->cgm_mode & CGM_PIPE_MODE_GAMMA)
+		crtc_state->base.gamma_lut = chv_read_cgm_lut(crtc_state);
+	else
+		i965_read_luts(crtc_state);
+}
+
+static struct drm_property_blob *
+ilk_read_lut_10(const struct intel_crtc_state *crtc_state)
+{
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	u32 lut_size = INTEL_INFO(dev_priv)->color.gamma_lut_size;
+	enum pipe pipe = crtc->pipe;
+	struct drm_property_blob *blob;
+	struct drm_color_lut *blob_data;
+	u32 i, val;
+
+	blob = drm_property_create_blob(&dev_priv->drm,
+					sizeof(struct drm_color_lut) * lut_size,
+					NULL);
+	if (IS_ERR(blob))
+		return NULL;
+
+	blob_data = blob->data;
+
+	for (i = 0; i < lut_size; i++) {
+		val = I915_READ(PREC_PALETTE(pipe, i));
+
+		blob_data[i].red = intel_color_lut_pack(REG_FIELD_GET(
+							PREC_PALETTE_RED_MASK, val), 10);
+		blob_data[i].green = intel_color_lut_pack(REG_FIELD_GET(
+							  PREC_PALETTE_GREEN_MASK, val), 10);
+		blob_data[i].blue = intel_color_lut_pack(REG_FIELD_GET(
+							 PREC_PALETTE_BLUE_MASK, val), 10);
+	}
+
+	return blob;
+}
+
+static void ilk_read_luts(struct intel_crtc_state *crtc_state)
+{
+	if (!crtc_state->gamma_enable)
+		return;
+
+	if ((crtc_state->csc_mode & CSC_POSITION_BEFORE_GAMMA) == 0)
+		return;
+
+	if (crtc_state->gamma_mode == GAMMA_MODE_MODE_8BIT)
+		crtc_state->base.gamma_lut = i9xx_read_lut_8(crtc_state);
+	else
+		crtc_state->base.gamma_lut = ilk_read_lut_10(crtc_state);
+}
+
+static struct drm_property_blob *
+glk_read_lut_10(const struct intel_crtc_state *crtc_state, u32 prec_index)
+{
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	int hw_lut_size = ivb_lut_10_size(prec_index);
+	enum pipe pipe = crtc->pipe;
+	struct drm_property_blob *blob;
+	struct drm_color_lut *blob_data;
+	u32 i, val;
+
+	blob = drm_property_create_blob(&dev_priv->drm,
+					sizeof(struct drm_color_lut) * hw_lut_size,
+					NULL);
+	if (IS_ERR(blob))
+		return NULL;
+
+	blob_data = blob->data;
+
+	I915_WRITE(PREC_PAL_INDEX(pipe), prec_index |
+		   PAL_PREC_AUTO_INCREMENT);
+
+	for (i = 0; i < hw_lut_size; i++) {
+		val = I915_READ(PREC_PAL_DATA(pipe));
+
+		blob_data[i].red = intel_color_lut_pack(REG_FIELD_GET(
+							PREC_PAL_DATA_RED_MASK, val), 10);
+		blob_data[i].green = intel_color_lut_pack(REG_FIELD_GET(
+							PREC_PAL_DATA_GREEN_MASK, val), 10);
+		blob_data[i].blue = intel_color_lut_pack(REG_FIELD_GET(
+							PREC_PAL_DATA_BLUE_MASK, val), 10);
+	}
+
+	I915_WRITE(PREC_PAL_INDEX(pipe), 0);
+
+	return blob;
+}
+
+static void glk_read_luts(struct intel_crtc_state *crtc_state)
+{
+	if (!crtc_state->gamma_enable)
+		return;
+
+	if (crtc_state->gamma_mode == GAMMA_MODE_MODE_8BIT)
+		crtc_state->base.gamma_lut = i9xx_read_lut_8(crtc_state);
+	else
+		crtc_state->base.gamma_lut = glk_read_lut_10(crtc_state, PAL_PREC_INDEX_VALUE(0));
+}
+
 void intel_color_init(struct intel_crtc *crtc)
 {
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
@@ -1383,14 +1888,17 @@ void intel_color_init(struct intel_crtc *crtc)
 			dev_priv->display.color_check = chv_color_check;
 			dev_priv->display.color_commit = i9xx_color_commit;
 			dev_priv->display.load_luts = chv_load_luts;
+			dev_priv->display.read_luts = chv_read_luts;
 		} else if (INTEL_GEN(dev_priv) >= 4) {
 			dev_priv->display.color_check = i9xx_color_check;
 			dev_priv->display.color_commit = i9xx_color_commit;
 			dev_priv->display.load_luts = i965_load_luts;
+			dev_priv->display.read_luts = i965_read_luts;
 		} else {
 			dev_priv->display.color_check = i9xx_color_check;
 			dev_priv->display.color_commit = i9xx_color_commit;
 			dev_priv->display.load_luts = i9xx_load_luts;
+			dev_priv->display.read_luts = i9xx_read_luts;
 		}
 	} else {
 		if (INTEL_GEN(dev_priv) >= 11)
@@ -1409,16 +1917,19 @@ void intel_color_init(struct intel_crtc *crtc)
 		else
 			dev_priv->display.color_commit = ilk_color_commit;
 
-		if (INTEL_GEN(dev_priv) >= 11)
+		if (INTEL_GEN(dev_priv) >= 11) {
 			dev_priv->display.load_luts = icl_load_luts;
-		else if (IS_CANNONLAKE(dev_priv) || IS_GEMINILAKE(dev_priv))
+		} else if (IS_CANNONLAKE(dev_priv) || IS_GEMINILAKE(dev_priv)) {
 			dev_priv->display.load_luts = glk_load_luts;
-		else if (INTEL_GEN(dev_priv) >= 8)
+			dev_priv->display.read_luts = glk_read_luts;
+		} else if (INTEL_GEN(dev_priv) >= 8) {
 			dev_priv->display.load_luts = bdw_load_luts;
-		else if (INTEL_GEN(dev_priv) >= 7)
+		} else if (INTEL_GEN(dev_priv) >= 7) {
 			dev_priv->display.load_luts = ivb_load_luts;
-		else
+		} else {
 			dev_priv->display.load_luts = ilk_load_luts;
+			dev_priv->display.read_luts = ilk_read_luts;
+		}
 	}
 
 	drm_crtc_enable_color_mgmt(&crtc->base,
diff --git a/drivers/gpu/drm/i915/display/intel_color.h b/drivers/gpu/drm/i915/display/intel_color.h
index 057e8ac63555..173727aaa24d 100644
--- a/drivers/gpu/drm/i915/display/intel_color.h
+++ b/drivers/gpu/drm/i915/display/intel_color.h
@@ -6,13 +6,20 @@
 #ifndef __INTEL_COLOR_H__
 #define __INTEL_COLOR_H__
 
+#include <linux/types.h>
+
 struct intel_crtc_state;
 struct intel_crtc;
+struct drm_property_blob;
 
 void intel_color_init(struct intel_crtc *crtc);
 int intel_color_check(struct intel_crtc_state *crtc_state);
 void intel_color_commit(const struct intel_crtc_state *crtc_state);
 void intel_color_load_luts(const struct intel_crtc_state *crtc_state);
 void intel_color_get_config(struct intel_crtc_state *crtc_state);
+int intel_color_get_gamma_bit_precision(const struct intel_crtc_state *crtc_state);
+bool intel_color_lut_equal(struct drm_property_blob *blob1,
+			   struct drm_property_blob *blob2,
+			   u32 gamma_mode, u32 bit_precision);
 
 #endif /* __INTEL_COLOR_H__ */
diff --git a/drivers/gpu/drm/i915/display/intel_connector.c b/drivers/gpu/drm/i915/display/intel_connector.c
index 308ec63207ee..1133c4e97bb4 100644
--- a/drivers/gpu/drm/i915/display/intel_connector.c
+++ b/drivers/gpu/drm/i915/display/intel_connector.c
@@ -277,7 +277,22 @@ intel_attach_aspect_ratio_property(struct drm_connector *connector)
 void
 intel_attach_colorspace_property(struct drm_connector *connector)
 {
-	if (!drm_mode_create_colorspace_property(connector))
-		drm_object_attach_property(&connector->base,
-					   connector->colorspace_property, 0);
+	switch (connector->connector_type) {
+	case DRM_MODE_CONNECTOR_HDMIA:
+	case DRM_MODE_CONNECTOR_HDMIB:
+		if (drm_mode_create_hdmi_colorspace_property(connector))
+			return;
+		break;
+	case DRM_MODE_CONNECTOR_DisplayPort:
+	case DRM_MODE_CONNECTOR_eDP:
+		if (drm_mode_create_dp_colorspace_property(connector))
+			return;
+		break;
+	default:
+		DRM_DEBUG_KMS("Colorspace property not supported\n");
+		return;
+	}
+
+	drm_object_attach_property(&connector->base,
+				   connector->colorspace_property, 0);
 }
diff --git a/drivers/gpu/drm/i915/display/intel_crt.c b/drivers/gpu/drm/i915/display/intel_crt.c
index e6e8d4a82044..39cc6d79dc85 100644
--- a/drivers/gpu/drm/i915/display/intel_crt.c
+++ b/drivers/gpu/drm/i915/display/intel_crt.c
@@ -844,7 +844,7 @@ load_detect:
 	}
 
 	/* for pre-945g platforms use load detect */
-	ret = intel_get_load_detect_pipe(connector, NULL, &tmp, ctx);
+	ret = intel_get_load_detect_pipe(connector, &tmp, ctx);
 	if (ret > 0) {
 		if (intel_crt_detect_ddc(connector))
 			status = connector_status_connected;
@@ -864,6 +864,13 @@ load_detect:
 
 out:
 	intel_display_power_put(dev_priv, intel_encoder->power_domain, wakeref);
+
+	/*
+	 * Make sure the refs for power wells enabled during detect are
+	 * dropped to avoid a new detect cycle triggered by HPD polling.
+	 */
+	intel_display_power_flush_work(dev_priv);
+
 	return status;
 }
 
@@ -994,9 +1001,9 @@ void intel_crt_init(struct drm_i915_private *dev_priv)
 	crt->base.type = INTEL_OUTPUT_ANALOG;
 	crt->base.cloneable = (1 << INTEL_OUTPUT_DVO) | (1 << INTEL_OUTPUT_HDMI);
 	if (IS_I830(dev_priv))
-		crt->base.crtc_mask = (1 << 0);
+		crt->base.pipe_mask = BIT(PIPE_A);
 	else
-		crt->base.crtc_mask = (1 << 0) | (1 << 1) | (1 << 2);
+		crt->base.pipe_mask = ~0;
 
 	if (IS_GEN(dev_priv, 2))
 		connector->interlace_allowed = 0;
diff --git a/drivers/gpu/drm/i915/display/intel_ddi.c b/drivers/gpu/drm/i915/display/intel_ddi.c
index 8eb2b3ec01ed..0d6e494b4508 100644
--- a/drivers/gpu/drm/i915/display/intel_ddi.c
+++ b/drivers/gpu/drm/i915/display/intel_ddi.c
@@ -45,6 +45,7 @@
 #include "intel_lspcon.h"
 #include "intel_panel.h"
 #include "intel_psr.h"
+#include "intel_sprite.h"
 #include "intel_tc.h"
 #include "intel_vdsc.h"
 
@@ -586,6 +587,26 @@ static const struct icl_mg_phy_ddi_buf_trans icl_mg_phy_ddi_translations[] = {
 	{ 0x0, 0x00, 0x00 },	/* 3              0   */
 };
 
+struct tgl_dkl_phy_ddi_buf_trans {
+	u32 dkl_vswing_control;
+	u32 dkl_preshoot_control;
+	u32 dkl_de_emphasis_control;
+};
+
+static const struct tgl_dkl_phy_ddi_buf_trans tgl_dkl_phy_ddi_translations[] = {
+				/* VS	pre-emp	Non-trans mV	Pre-emph dB */
+	{ 0x7, 0x0, 0x00 },	/* 0	0	400mV		0 dB */
+	{ 0x5, 0x0, 0x03 },	/* 0	1	400mV		3.5 dB */
+	{ 0x2, 0x0, 0x0b },	/* 0	2	400mV		6 dB */
+	{ 0x0, 0x0, 0x19 },	/* 0	3	400mV		9.5 dB */
+	{ 0x5, 0x0, 0x00 },	/* 1	0	600mV		0 dB */
+	{ 0x2, 0x0, 0x03 },	/* 1	1	600mV		3.5 dB */
+	{ 0x0, 0x0, 0x14 },	/* 1	2	600mV		6 dB */
+	{ 0x2, 0x0, 0x00 },	/* 2	0	800mV		0 dB */
+	{ 0x0, 0x0, 0x0B },	/* 2	1	800mV		3.5 dB */
+	{ 0x0, 0x0, 0x00 },	/* 3	0	1200mV		0 dB HDMI default */
+};
+
 static const struct ddi_buf_trans *
 bdw_get_buf_trans_edp(struct drm_i915_private *dev_priv, int *n_entries)
 {
@@ -872,7 +893,14 @@ static int intel_ddi_hdmi_level(struct drm_i915_private *dev_priv, enum port por
 
 	level = dev_priv->vbt.ddi_port_info[port].hdmi_level_shift;
 
-	if (INTEL_GEN(dev_priv) >= 11) {
+	if (INTEL_GEN(dev_priv) >= 12) {
+		if (intel_phy_is_combo(dev_priv, phy))
+			icl_get_combo_buf_trans(dev_priv, INTEL_OUTPUT_HDMI,
+						0, &n_entries);
+		else
+			n_entries = ARRAY_SIZE(tgl_dkl_phy_ddi_translations);
+		default_entry = n_entries - 1;
+	} else if (INTEL_GEN(dev_priv) == 11) {
 		if (intel_phy_is_combo(dev_priv, phy))
 			icl_get_combo_buf_trans(dev_priv, INTEL_OUTPUT_HDMI,
 						0, &n_entries);
@@ -1049,6 +1077,8 @@ static u32 icl_pll_to_ddi_clk_sel(struct intel_encoder *encoder,
 	case DPLL_ID_ICL_MGPLL2:
 	case DPLL_ID_ICL_MGPLL3:
 	case DPLL_ID_ICL_MGPLL4:
+	case DPLL_ID_TGL_MGPLL5:
+	case DPLL_ID_TGL_MGPLL6:
 		return DDI_CLK_SEL_MG;
 	}
 }
@@ -1413,11 +1443,30 @@ static int icl_calc_mg_pll_link(struct drm_i915_private *dev_priv,
 
 	ref_clock = dev_priv->cdclk.hw.ref;
 
-	m1 = pll_state->mg_pll_div1 & MG_PLL_DIV1_FBPREDIV_MASK;
-	m2_int = pll_state->mg_pll_div0 & MG_PLL_DIV0_FBDIV_INT_MASK;
-	m2_frac = (pll_state->mg_pll_div0 & MG_PLL_DIV0_FRACNEN_H) ?
-		(pll_state->mg_pll_div0 & MG_PLL_DIV0_FBDIV_FRAC_MASK) >>
-		MG_PLL_DIV0_FBDIV_FRAC_SHIFT : 0;
+	if (INTEL_GEN(dev_priv) >= 12) {
+		m1 = pll_state->mg_pll_div0 & DKL_PLL_DIV0_FBPREDIV_MASK;
+		m1 = m1 >> DKL_PLL_DIV0_FBPREDIV_SHIFT;
+		m2_int = pll_state->mg_pll_div0 & DKL_PLL_DIV0_FBDIV_INT_MASK;
+
+		if (pll_state->mg_pll_bias & DKL_PLL_BIAS_FRAC_EN_H) {
+			m2_frac = pll_state->mg_pll_bias &
+				  DKL_PLL_BIAS_FBDIV_FRAC_MASK;
+			m2_frac = m2_frac >> DKL_PLL_BIAS_FBDIV_SHIFT;
+		} else {
+			m2_frac = 0;
+		}
+	} else {
+		m1 = pll_state->mg_pll_div1 & MG_PLL_DIV1_FBPREDIV_MASK;
+		m2_int = pll_state->mg_pll_div0 & MG_PLL_DIV0_FBDIV_INT_MASK;
+
+		if (pll_state->mg_pll_div0 & MG_PLL_DIV0_FRACNEN_H) {
+			m2_frac = pll_state->mg_pll_div0 &
+				  MG_PLL_DIV0_FBDIV_FRAC_MASK;
+			m2_frac = m2_frac >> MG_PLL_DIV0_FBDIV_FRAC_SHIFT;
+		} else {
+			m2_frac = 0;
+		}
+	}
 
 	switch (pll_state->mg_clktop2_hsclkctl &
 		MG_CLKTOP2_HSCLKCTL_HSDIV_RATIO_MASK) {
@@ -1692,7 +1741,8 @@ static void intel_ddi_clock_get(struct intel_encoder *encoder,
 		hsw_ddi_clock_get(encoder, pipe_config);
 }
 
-void intel_ddi_set_pipe_settings(const struct intel_crtc_state *crtc_state)
+void intel_ddi_set_dp_msa(const struct intel_crtc_state *crtc_state,
+			  const struct drm_connector_state *conn_state)
 {
 	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
@@ -1704,44 +1754,50 @@ void intel_ddi_set_pipe_settings(const struct intel_crtc_state *crtc_state)
 
 	WARN_ON(transcoder_is_dsi(cpu_transcoder));
 
-	temp = TRANS_MSA_SYNC_CLK;
-
-	if (crtc_state->limited_color_range)
-		temp |= TRANS_MSA_CEA_RANGE;
+	temp = DP_MSA_MISC_SYNC_CLOCK;
 
 	switch (crtc_state->pipe_bpp) {
 	case 18:
-		temp |= TRANS_MSA_6_BPC;
+		temp |= DP_MSA_MISC_6_BPC;
 		break;
 	case 24:
-		temp |= TRANS_MSA_8_BPC;
+		temp |= DP_MSA_MISC_8_BPC;
 		break;
 	case 30:
-		temp |= TRANS_MSA_10_BPC;
+		temp |= DP_MSA_MISC_10_BPC;
 		break;
 	case 36:
-		temp |= TRANS_MSA_12_BPC;
+		temp |= DP_MSA_MISC_12_BPC;
 		break;
 	default:
 		MISSING_CASE(crtc_state->pipe_bpp);
 		break;
 	}
 
+	/* nonsense combination */
+	WARN_ON(crtc_state->limited_color_range &&
+		crtc_state->output_format != INTEL_OUTPUT_FORMAT_RGB);
+
+	if (crtc_state->limited_color_range)
+		temp |= DP_MSA_MISC_COLOR_CEA_RGB;
+
 	/*
 	 * As per DP 1.2 spec section 2.3.4.3 while sending
 	 * YCBCR 444 signals we should program MSA MISC1/0 fields with
-	 * colorspace information. The output colorspace encoding is BT601.
+	 * colorspace information.
 	 */
 	if (crtc_state->output_format == INTEL_OUTPUT_FORMAT_YCBCR444)
-		temp |= TRANS_MSA_SAMPLING_444 | TRANS_MSA_CLRSP_YCBCR;
+		temp |= DP_MSA_MISC_COLOR_YCBCR_444_BT709;
+
 	/*
 	 * As per DP 1.4a spec section 2.2.4.3 [MSA Field for Indication
 	 * of Color Encoding Format and Content Color Gamut] while sending
-	 * YCBCR 420 signals we should program MSA MISC1 fields which
-	 * indicate VSC SDP for the Pixel Encoding/Colorimetry Format.
+	 * YCBCR 420, HDR BT.2020 signals we should program MSA MISC1 fields
+	 * which indicate VSC SDP for the Pixel Encoding/Colorimetry Format.
 	 */
-	if (crtc_state->output_format == INTEL_OUTPUT_FORMAT_YCBCR420)
-		temp |= TRANS_MSA_USE_VSC_SDP;
+	if (intel_dp_needs_vsc_sdp(crtc_state, conn_state))
+		temp |= DP_MSA_MISC_COLOR_VSC_SDP;
+
 	I915_WRITE(TRANS_MSA_MISC(cpu_transcoder), temp);
 }
 
@@ -1761,7 +1817,14 @@ void intel_ddi_set_vc_payload_alloc(const struct intel_crtc_state *crtc_state,
 	I915_WRITE(TRANS_DDI_FUNC_CTL(cpu_transcoder), temp);
 }
 
-void intel_ddi_enable_transcoder_func(const struct intel_crtc_state *crtc_state)
+/*
+ * Returns the TRANS_DDI_FUNC_CTL value based on CRTC state.
+ *
+ * Only intended to be used by intel_ddi_enable_transcoder_func() and
+ * intel_ddi_config_transcoder_func().
+ */
+static u32
+intel_ddi_transcoder_func_reg_val_get(const struct intel_crtc_state *crtc_state)
 {
 	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
 	struct intel_encoder *encoder = intel_ddi_get_crtc_encoder(crtc);
@@ -1840,11 +1903,42 @@ void intel_ddi_enable_transcoder_func(const struct intel_crtc_state *crtc_state)
 	} else if (intel_crtc_has_type(crtc_state, INTEL_OUTPUT_DP_MST)) {
 		temp |= TRANS_DDI_MODE_SELECT_DP_MST;
 		temp |= DDI_PORT_WIDTH(crtc_state->lane_count);
+
+		if (INTEL_GEN(dev_priv) >= 12)
+			temp |= TRANS_DDI_MST_TRANSPORT_SELECT(crtc_state->cpu_transcoder);
 	} else {
 		temp |= TRANS_DDI_MODE_SELECT_DP_SST;
 		temp |= DDI_PORT_WIDTH(crtc_state->lane_count);
 	}
 
+	return temp;
+}
+
+void intel_ddi_enable_transcoder_func(const struct intel_crtc_state *crtc_state)
+{
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	enum transcoder cpu_transcoder = crtc_state->cpu_transcoder;
+	u32 temp;
+
+	temp = intel_ddi_transcoder_func_reg_val_get(crtc_state);
+	I915_WRITE(TRANS_DDI_FUNC_CTL(cpu_transcoder), temp);
+}
+
+/*
+ * Same as intel_ddi_enable_transcoder_func(), but it does not set the enable
+ * bit.
+ */
+static void
+intel_ddi_config_transcoder_func(const struct intel_crtc_state *crtc_state)
+{
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	enum transcoder cpu_transcoder = crtc_state->cpu_transcoder;
+	u32 temp;
+
+	temp = intel_ddi_transcoder_func_reg_val_get(crtc_state);
+	temp &= ~TRANS_DDI_FUNC_ENABLE;
 	I915_WRITE(TRANS_DDI_FUNC_CTL(cpu_transcoder), temp);
 }
 
@@ -2045,18 +2139,20 @@ static void intel_ddi_get_encoder_pipes(struct intel_encoder *encoder,
 	}
 
 	if (!*pipe_mask)
-		DRM_DEBUG_KMS("No pipe for ddi port %c found\n",
-			      port_name(port));
+		DRM_DEBUG_KMS("No pipe for [ENCODER:%d:%s] found\n",
+			      encoder->base.base.id, encoder->base.name);
 
 	if (!mst_pipe_mask && hweight8(*pipe_mask) > 1) {
-		DRM_DEBUG_KMS("Multiple pipes for non DP-MST port %c (pipe_mask %02x)\n",
-			      port_name(port), *pipe_mask);
+		DRM_DEBUG_KMS("Multiple pipes for [ENCODER:%d:%s] (pipe_mask %02x)\n",
+			      encoder->base.base.id, encoder->base.name,
+			      *pipe_mask);
 		*pipe_mask = BIT(ffs(*pipe_mask) - 1);
 	}
 
 	if (mst_pipe_mask && mst_pipe_mask != *pipe_mask)
-		DRM_DEBUG_KMS("Conflicting MST and non-MST encoders for port %c (pipe_mask %02x mst_pipe_mask %02x)\n",
-			      port_name(port), *pipe_mask, mst_pipe_mask);
+		DRM_DEBUG_KMS("Conflicting MST and non-MST state for [ENCODER:%d:%s] (pipe_mask %02x mst_pipe_mask %02x)\n",
+			      encoder->base.base.id, encoder->base.name,
+			      *pipe_mask, mst_pipe_mask);
 	else
 		*is_dp_mst = mst_pipe_mask;
 
@@ -2066,8 +2162,9 @@ out:
 		if ((tmp & (BXT_PHY_CMNLANE_POWERDOWN_ACK |
 			    BXT_PHY_LANE_POWERDOWN_ACK |
 			    BXT_PHY_LANE_ENABLED)) != BXT_PHY_LANE_ENABLED)
-			DRM_ERROR("Port %c enabled but PHY powered down? "
-				  "(PHY_CTL %08x)\n", port_name(port), tmp);
+			DRM_ERROR("[ENCODER:%d:%s] enabled but PHY powered down? "
+				  "(PHY_CTL %08x)\n", encoder->base.base.id,
+				  encoder->base.name, tmp);
 	}
 
 	intel_display_power_put(dev_priv, encoder->power_domain, wakeref);
@@ -2138,7 +2235,7 @@ static void intel_ddi_get_power_domains(struct intel_encoder *encoder,
 	/*
 	 * VDSC power is needed when DSC is enabled
 	 */
-	if (crtc_state->dsc_params.compression_enable)
+	if (crtc_state->dsc.compression_enable)
 		intel_display_power_get(dev_priv,
 					intel_dsc_power_domain(crtc_state));
 }
@@ -2269,7 +2366,13 @@ u8 intel_ddi_dp_voltage_max(struct intel_encoder *encoder)
 	enum phy phy = intel_port_to_phy(dev_priv, port);
 	int n_entries;
 
-	if (INTEL_GEN(dev_priv) >= 11) {
+	if (INTEL_GEN(dev_priv) >= 12) {
+		if (intel_phy_is_combo(dev_priv, phy))
+			icl_get_combo_buf_trans(dev_priv, encoder->type,
+						intel_dp->link_rate, &n_entries);
+		else
+			n_entries = ARRAY_SIZE(tgl_dkl_phy_ddi_translations);
+	} else if (INTEL_GEN(dev_priv) == 11) {
 		if (intel_phy_is_combo(dev_priv, phy))
 			icl_get_combo_buf_trans(dev_priv, encoder->type,
 						intel_dp->link_rate, &n_entries);
@@ -2583,7 +2686,7 @@ static void icl_mg_phy_ddi_vswing_sequence(struct intel_encoder *encoder,
 					   u32 level)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	enum port port = encoder->port;
+	enum tc_port tc_port = intel_port_to_tc(dev_priv, encoder->port);
 	const struct icl_mg_phy_ddi_buf_trans *ddi_translations;
 	u32 n_entries, val;
 	int ln;
@@ -2599,33 +2702,33 @@ static void icl_mg_phy_ddi_vswing_sequence(struct intel_encoder *encoder,
 
 	/* Set MG_TX_LINK_PARAMS cri_use_fs32 to 0. */
 	for (ln = 0; ln < 2; ln++) {
-		val = I915_READ(MG_TX1_LINK_PARAMS(ln, port));
+		val = I915_READ(MG_TX1_LINK_PARAMS(ln, tc_port));
 		val &= ~CRI_USE_FS32;
-		I915_WRITE(MG_TX1_LINK_PARAMS(ln, port), val);
+		I915_WRITE(MG_TX1_LINK_PARAMS(ln, tc_port), val);
 
-		val = I915_READ(MG_TX2_LINK_PARAMS(ln, port));
+		val = I915_READ(MG_TX2_LINK_PARAMS(ln, tc_port));
 		val &= ~CRI_USE_FS32;
-		I915_WRITE(MG_TX2_LINK_PARAMS(ln, port), val);
+		I915_WRITE(MG_TX2_LINK_PARAMS(ln, tc_port), val);
 	}
 
 	/* Program MG_TX_SWINGCTRL with values from vswing table */
 	for (ln = 0; ln < 2; ln++) {
-		val = I915_READ(MG_TX1_SWINGCTRL(ln, port));
+		val = I915_READ(MG_TX1_SWINGCTRL(ln, tc_port));
 		val &= ~CRI_TXDEEMPH_OVERRIDE_17_12_MASK;
 		val |= CRI_TXDEEMPH_OVERRIDE_17_12(
 			ddi_translations[level].cri_txdeemph_override_17_12);
-		I915_WRITE(MG_TX1_SWINGCTRL(ln, port), val);
+		I915_WRITE(MG_TX1_SWINGCTRL(ln, tc_port), val);
 
-		val = I915_READ(MG_TX2_SWINGCTRL(ln, port));
+		val = I915_READ(MG_TX2_SWINGCTRL(ln, tc_port));
 		val &= ~CRI_TXDEEMPH_OVERRIDE_17_12_MASK;
 		val |= CRI_TXDEEMPH_OVERRIDE_17_12(
 			ddi_translations[level].cri_txdeemph_override_17_12);
-		I915_WRITE(MG_TX2_SWINGCTRL(ln, port), val);
+		I915_WRITE(MG_TX2_SWINGCTRL(ln, tc_port), val);
 	}
 
 	/* Program MG_TX_DRVCTRL with values from vswing table */
 	for (ln = 0; ln < 2; ln++) {
-		val = I915_READ(MG_TX1_DRVCTRL(ln, port));
+		val = I915_READ(MG_TX1_DRVCTRL(ln, tc_port));
 		val &= ~(CRI_TXDEEMPH_OVERRIDE_11_6_MASK |
 			 CRI_TXDEEMPH_OVERRIDE_5_0_MASK);
 		val |= CRI_TXDEEMPH_OVERRIDE_5_0(
@@ -2633,9 +2736,9 @@ static void icl_mg_phy_ddi_vswing_sequence(struct intel_encoder *encoder,
 			CRI_TXDEEMPH_OVERRIDE_11_6(
 				ddi_translations[level].cri_txdeemph_override_11_6) |
 			CRI_TXDEEMPH_OVERRIDE_EN;
-		I915_WRITE(MG_TX1_DRVCTRL(ln, port), val);
+		I915_WRITE(MG_TX1_DRVCTRL(ln, tc_port), val);
 
-		val = I915_READ(MG_TX2_DRVCTRL(ln, port));
+		val = I915_READ(MG_TX2_DRVCTRL(ln, tc_port));
 		val &= ~(CRI_TXDEEMPH_OVERRIDE_11_6_MASK |
 			 CRI_TXDEEMPH_OVERRIDE_5_0_MASK);
 		val |= CRI_TXDEEMPH_OVERRIDE_5_0(
@@ -2643,7 +2746,7 @@ static void icl_mg_phy_ddi_vswing_sequence(struct intel_encoder *encoder,
 			CRI_TXDEEMPH_OVERRIDE_11_6(
 				ddi_translations[level].cri_txdeemph_override_11_6) |
 			CRI_TXDEEMPH_OVERRIDE_EN;
-		I915_WRITE(MG_TX2_DRVCTRL(ln, port), val);
+		I915_WRITE(MG_TX2_DRVCTRL(ln, tc_port), val);
 
 		/* FIXME: Program CRI_LOADGEN_SEL after the spec is updated */
 	}
@@ -2654,17 +2757,17 @@ static void icl_mg_phy_ddi_vswing_sequence(struct intel_encoder *encoder,
 	 * values from table for which TX1 and TX2 enabled.
 	 */
 	for (ln = 0; ln < 2; ln++) {
-		val = I915_READ(MG_CLKHUB(ln, port));
+		val = I915_READ(MG_CLKHUB(ln, tc_port));
 		if (link_clock < 300000)
 			val |= CFG_LOW_RATE_LKREN_EN;
 		else
 			val &= ~CFG_LOW_RATE_LKREN_EN;
-		I915_WRITE(MG_CLKHUB(ln, port), val);
+		I915_WRITE(MG_CLKHUB(ln, tc_port), val);
 	}
 
 	/* Program the MG_TX_DCC<LN, port being used> based on the link frequency */
 	for (ln = 0; ln < 2; ln++) {
-		val = I915_READ(MG_TX1_DCC(ln, port));
+		val = I915_READ(MG_TX1_DCC(ln, tc_port));
 		val &= ~CFG_AMI_CK_DIV_OVERRIDE_VAL_MASK;
 		if (link_clock <= 500000) {
 			val &= ~CFG_AMI_CK_DIV_OVERRIDE_EN;
@@ -2672,9 +2775,9 @@ static void icl_mg_phy_ddi_vswing_sequence(struct intel_encoder *encoder,
 			val |= CFG_AMI_CK_DIV_OVERRIDE_EN |
 				CFG_AMI_CK_DIV_OVERRIDE_VAL(1);
 		}
-		I915_WRITE(MG_TX1_DCC(ln, port), val);
+		I915_WRITE(MG_TX1_DCC(ln, tc_port), val);
 
-		val = I915_READ(MG_TX2_DCC(ln, port));
+		val = I915_READ(MG_TX2_DCC(ln, tc_port));
 		val &= ~CFG_AMI_CK_DIV_OVERRIDE_VAL_MASK;
 		if (link_clock <= 500000) {
 			val &= ~CFG_AMI_CK_DIV_OVERRIDE_EN;
@@ -2682,18 +2785,18 @@ static void icl_mg_phy_ddi_vswing_sequence(struct intel_encoder *encoder,
 			val |= CFG_AMI_CK_DIV_OVERRIDE_EN |
 				CFG_AMI_CK_DIV_OVERRIDE_VAL(1);
 		}
-		I915_WRITE(MG_TX2_DCC(ln, port), val);
+		I915_WRITE(MG_TX2_DCC(ln, tc_port), val);
 	}
 
 	/* Program MG_TX_PISO_READLOAD with values from vswing table */
 	for (ln = 0; ln < 2; ln++) {
-		val = I915_READ(MG_TX1_PISO_READLOAD(ln, port));
+		val = I915_READ(MG_TX1_PISO_READLOAD(ln, tc_port));
 		val |= CRI_CALCINIT;
-		I915_WRITE(MG_TX1_PISO_READLOAD(ln, port), val);
+		I915_WRITE(MG_TX1_PISO_READLOAD(ln, tc_port), val);
 
-		val = I915_READ(MG_TX2_PISO_READLOAD(ln, port));
+		val = I915_READ(MG_TX2_PISO_READLOAD(ln, tc_port));
 		val |= CRI_CALCINIT;
-		I915_WRITE(MG_TX2_PISO_READLOAD(ln, port), val);
+		I915_WRITE(MG_TX2_PISO_READLOAD(ln, tc_port), val);
 	}
 }
 
@@ -2711,6 +2814,64 @@ static void icl_ddi_vswing_sequence(struct intel_encoder *encoder,
 		icl_mg_phy_ddi_vswing_sequence(encoder, link_clock, level);
 }
 
+static void
+tgl_dkl_phy_ddi_vswing_sequence(struct intel_encoder *encoder, int link_clock,
+				u32 level)
+{
+	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
+	enum tc_port tc_port = intel_port_to_tc(dev_priv, encoder->port);
+	const struct tgl_dkl_phy_ddi_buf_trans *ddi_translations;
+	u32 n_entries, val, ln, dpcnt_mask, dpcnt_val;
+
+	n_entries = ARRAY_SIZE(tgl_dkl_phy_ddi_translations);
+	ddi_translations = tgl_dkl_phy_ddi_translations;
+
+	if (level >= n_entries)
+		level = n_entries - 1;
+
+	dpcnt_mask = (DKL_TX_PRESHOOT_COEFF_MASK |
+		      DKL_TX_DE_EMPAHSIS_COEFF_MASK |
+		      DKL_TX_VSWING_CONTROL_MASK);
+	dpcnt_val = DKL_TX_VSWING_CONTROL(ddi_translations[level].dkl_vswing_control);
+	dpcnt_val |= DKL_TX_DE_EMPHASIS_COEFF(ddi_translations[level].dkl_de_emphasis_control);
+	dpcnt_val |= DKL_TX_PRESHOOT_COEFF(ddi_translations[level].dkl_preshoot_control);
+
+	for (ln = 0; ln < 2; ln++) {
+		I915_WRITE(HIP_INDEX_REG(tc_port), HIP_INDEX_VAL(tc_port, ln));
+
+		I915_WRITE(DKL_TX_PMD_LANE_SUS(tc_port), 0);
+
+		/* All the registers are RMW */
+		val = I915_READ(DKL_TX_DPCNTL0(tc_port));
+		val &= ~dpcnt_mask;
+		val |= dpcnt_val;
+		I915_WRITE(DKL_TX_DPCNTL0(tc_port), val);
+
+		val = I915_READ(DKL_TX_DPCNTL1(tc_port));
+		val &= ~dpcnt_mask;
+		val |= dpcnt_val;
+		I915_WRITE(DKL_TX_DPCNTL1(tc_port), val);
+
+		val = I915_READ(DKL_TX_DPCNTL2(tc_port));
+		val &= ~DKL_TX_DP20BITMODE;
+		I915_WRITE(DKL_TX_DPCNTL2(tc_port), val);
+	}
+}
+
+static void tgl_ddi_vswing_sequence(struct intel_encoder *encoder,
+				    int link_clock,
+				    u32 level,
+				    enum intel_output_type type)
+{
+	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
+	enum phy phy = intel_port_to_phy(dev_priv, encoder->port);
+
+	if (intel_phy_is_combo(dev_priv, phy))
+		icl_combo_phy_ddi_vswing_sequence(encoder, level, type);
+	else
+		tgl_dkl_phy_ddi_vswing_sequence(encoder, link_clock, level);
+}
+
 static u32 translate_signal_level(int signal_levels)
 {
 	int i;
@@ -2742,7 +2903,10 @@ u32 bxt_signal_levels(struct intel_dp *intel_dp)
 	struct intel_encoder *encoder = &dport->base;
 	int level = intel_ddi_dp_level(intel_dp);
 
-	if (INTEL_GEN(dev_priv) >= 11)
+	if (INTEL_GEN(dev_priv) >= 12)
+		tgl_ddi_vswing_sequence(encoder, intel_dp->link_rate,
+					level, encoder->type);
+	else if (INTEL_GEN(dev_priv) >= 11)
 		icl_ddi_vswing_sequence(encoder, intel_dp->link_rate,
 					level, encoder->type);
 	else if (IS_CANNONLAKE(dev_priv))
@@ -2989,130 +3153,141 @@ static void intel_ddi_clk_disable(struct intel_encoder *encoder)
 	}
 }
 
-static void icl_enable_phy_clock_gating(struct intel_digital_port *dig_port)
+static void
+icl_phy_set_clock_gating(struct intel_digital_port *dig_port, bool enable)
 {
 	struct drm_i915_private *dev_priv = to_i915(dig_port->base.base.dev);
-	enum port port = dig_port->base.port;
-	enum tc_port tc_port = intel_port_to_tc(dev_priv, port);
-	u32 val;
+	enum tc_port tc_port = intel_port_to_tc(dev_priv, dig_port->base.port);
+	u32 val, bits;
 	int ln;
 
 	if (tc_port == PORT_TC_NONE)
 		return;
 
-	for (ln = 0; ln < 2; ln++) {
-		val = I915_READ(MG_DP_MODE(ln, port));
-		val |= MG_DP_MODE_CFG_TR2PWR_GATING |
-		       MG_DP_MODE_CFG_TRPWR_GATING |
-		       MG_DP_MODE_CFG_CLNPWR_GATING |
-		       MG_DP_MODE_CFG_DIGPWR_GATING |
-		       MG_DP_MODE_CFG_GAONPWR_GATING;
-		I915_WRITE(MG_DP_MODE(ln, port), val);
-	}
+	bits = MG_DP_MODE_CFG_TR2PWR_GATING | MG_DP_MODE_CFG_TRPWR_GATING |
+	       MG_DP_MODE_CFG_CLNPWR_GATING | MG_DP_MODE_CFG_DIGPWR_GATING |
+	       MG_DP_MODE_CFG_GAONPWR_GATING;
 
-	val = I915_READ(MG_MISC_SUS0(tc_port));
-	val |= MG_MISC_SUS0_SUSCLK_DYNCLKGATE_MODE(3) |
-	       MG_MISC_SUS0_CFG_TR2PWR_GATING |
-	       MG_MISC_SUS0_CFG_CL2PWR_GATING |
-	       MG_MISC_SUS0_CFG_GAONPWR_GATING |
-	       MG_MISC_SUS0_CFG_TRPWR_GATING |
-	       MG_MISC_SUS0_CFG_CL1PWR_GATING |
-	       MG_MISC_SUS0_CFG_DGPWR_GATING;
-	I915_WRITE(MG_MISC_SUS0(tc_port), val);
-}
-
-static void icl_disable_phy_clock_gating(struct intel_digital_port *dig_port)
-{
-	struct drm_i915_private *dev_priv = to_i915(dig_port->base.base.dev);
-	enum port port = dig_port->base.port;
-	enum tc_port tc_port = intel_port_to_tc(dev_priv, port);
-	u32 val;
-	int ln;
+	for (ln = 0; ln < 2; ln++) {
+		if (INTEL_GEN(dev_priv) >= 12) {
+			I915_WRITE(HIP_INDEX_REG(tc_port), HIP_INDEX_VAL(tc_port, ln));
+			val = I915_READ(DKL_DP_MODE(tc_port));
+		} else {
+			val = I915_READ(MG_DP_MODE(ln, tc_port));
+		}
 
-	if (tc_port == PORT_TC_NONE)
-		return;
+		if (enable)
+			val |= bits;
+		else
+			val &= ~bits;
 
-	for (ln = 0; ln < 2; ln++) {
-		val = I915_READ(MG_DP_MODE(ln, port));
-		val &= ~(MG_DP_MODE_CFG_TR2PWR_GATING |
-			 MG_DP_MODE_CFG_TRPWR_GATING |
-			 MG_DP_MODE_CFG_CLNPWR_GATING |
-			 MG_DP_MODE_CFG_DIGPWR_GATING |
-			 MG_DP_MODE_CFG_GAONPWR_GATING);
-		I915_WRITE(MG_DP_MODE(ln, port), val);
+		if (INTEL_GEN(dev_priv) >= 12)
+			I915_WRITE(DKL_DP_MODE(tc_port), val);
+		else
+			I915_WRITE(MG_DP_MODE(ln, tc_port), val);
 	}
 
-	val = I915_READ(MG_MISC_SUS0(tc_port));
-	val &= ~(MG_MISC_SUS0_SUSCLK_DYNCLKGATE_MODE_MASK |
-		 MG_MISC_SUS0_CFG_TR2PWR_GATING |
-		 MG_MISC_SUS0_CFG_CL2PWR_GATING |
-		 MG_MISC_SUS0_CFG_GAONPWR_GATING |
-		 MG_MISC_SUS0_CFG_TRPWR_GATING |
-		 MG_MISC_SUS0_CFG_CL1PWR_GATING |
-		 MG_MISC_SUS0_CFG_DGPWR_GATING);
-	I915_WRITE(MG_MISC_SUS0(tc_port), val);
+	if (INTEL_GEN(dev_priv) == 11) {
+		bits = MG_MISC_SUS0_CFG_TR2PWR_GATING |
+		       MG_MISC_SUS0_CFG_CL2PWR_GATING |
+		       MG_MISC_SUS0_CFG_GAONPWR_GATING |
+		       MG_MISC_SUS0_CFG_TRPWR_GATING |
+		       MG_MISC_SUS0_CFG_CL1PWR_GATING |
+		       MG_MISC_SUS0_CFG_DGPWR_GATING;
+
+		val = I915_READ(MG_MISC_SUS0(tc_port));
+		if (enable)
+			val |= (bits | MG_MISC_SUS0_SUSCLK_DYNCLKGATE_MODE(3));
+		else
+			val &= ~(bits | MG_MISC_SUS0_SUSCLK_DYNCLKGATE_MODE_MASK);
+		I915_WRITE(MG_MISC_SUS0(tc_port), val);
+	}
 }
 
-static void icl_program_mg_dp_mode(struct intel_digital_port *intel_dig_port)
+static void
+icl_program_mg_dp_mode(struct intel_digital_port *intel_dig_port,
+		       const struct intel_crtc_state *crtc_state)
 {
 	struct drm_i915_private *dev_priv = to_i915(intel_dig_port->base.base.dev);
-	enum port port = intel_dig_port->base.port;
-	u32 ln0, ln1, lane_mask;
+	enum tc_port tc_port = intel_port_to_tc(dev_priv, intel_dig_port->base.port);
+	u32 ln0, ln1, pin_assignment;
+	u8 width;
 
 	if (intel_dig_port->tc_mode == TC_PORT_TBT_ALT)
 		return;
 
-	ln0 = I915_READ(MG_DP_MODE(0, port));
-	ln1 = I915_READ(MG_DP_MODE(1, port));
+	if (INTEL_GEN(dev_priv) >= 12) {
+		I915_WRITE(HIP_INDEX_REG(tc_port), HIP_INDEX_VAL(tc_port, 0x0));
+		ln0 = I915_READ(DKL_DP_MODE(tc_port));
+		I915_WRITE(HIP_INDEX_REG(tc_port), HIP_INDEX_VAL(tc_port, 0x1));
+		ln1 = I915_READ(DKL_DP_MODE(tc_port));
+	} else {
+		ln0 = I915_READ(MG_DP_MODE(0, tc_port));
+		ln1 = I915_READ(MG_DP_MODE(1, tc_port));
+	}
 
-	switch (intel_dig_port->tc_mode) {
-	case TC_PORT_DP_ALT:
-		ln0 &= ~(MG_DP_MODE_CFG_DP_X1_MODE | MG_DP_MODE_CFG_DP_X2_MODE);
-		ln1 &= ~(MG_DP_MODE_CFG_DP_X1_MODE | MG_DP_MODE_CFG_DP_X2_MODE);
+	ln0 &= ~(MG_DP_MODE_CFG_DP_X1_MODE | MG_DP_MODE_CFG_DP_X1_MODE);
+	ln1 &= ~(MG_DP_MODE_CFG_DP_X1_MODE | MG_DP_MODE_CFG_DP_X2_MODE);
 
-		lane_mask = intel_tc_port_get_lane_mask(intel_dig_port);
+	/* DPPATC */
+	pin_assignment = intel_tc_port_get_pin_assignment_mask(intel_dig_port);
+	width = crtc_state->lane_count;
 
-		switch (lane_mask) {
-		case 0x1:
-		case 0x4:
-			break;
-		case 0x2:
+	switch (pin_assignment) {
+	case 0x0:
+		WARN_ON(intel_dig_port->tc_mode != TC_PORT_LEGACY);
+		if (width == 1) {
+			ln1 |= MG_DP_MODE_CFG_DP_X1_MODE;
+		} else {
+			ln0 |= MG_DP_MODE_CFG_DP_X2_MODE;
+			ln1 |= MG_DP_MODE_CFG_DP_X2_MODE;
+		}
+		break;
+	case 0x1:
+		if (width == 4) {
+			ln0 |= MG_DP_MODE_CFG_DP_X2_MODE;
+			ln1 |= MG_DP_MODE_CFG_DP_X2_MODE;
+		}
+		break;
+	case 0x2:
+		if (width == 2) {
+			ln0 |= MG_DP_MODE_CFG_DP_X2_MODE;
+			ln1 |= MG_DP_MODE_CFG_DP_X2_MODE;
+		}
+		break;
+	case 0x3:
+	case 0x5:
+		if (width == 1) {
 			ln0 |= MG_DP_MODE_CFG_DP_X1_MODE;
-			break;
-		case 0x3:
-			ln0 |= MG_DP_MODE_CFG_DP_X1_MODE |
-			       MG_DP_MODE_CFG_DP_X2_MODE;
-			break;
-		case 0x8:
 			ln1 |= MG_DP_MODE_CFG_DP_X1_MODE;
-			break;
-		case 0xC:
-			ln1 |= MG_DP_MODE_CFG_DP_X1_MODE |
-			       MG_DP_MODE_CFG_DP_X2_MODE;
-			break;
-		case 0xF:
-			ln0 |= MG_DP_MODE_CFG_DP_X1_MODE |
-			       MG_DP_MODE_CFG_DP_X2_MODE;
-			ln1 |= MG_DP_MODE_CFG_DP_X1_MODE |
-			       MG_DP_MODE_CFG_DP_X2_MODE;
-			break;
-		default:
-			MISSING_CASE(lane_mask);
+		} else {
+			ln0 |= MG_DP_MODE_CFG_DP_X2_MODE;
+			ln1 |= MG_DP_MODE_CFG_DP_X2_MODE;
 		}
 		break;
-
-	case TC_PORT_LEGACY:
-		ln0 |= MG_DP_MODE_CFG_DP_X1_MODE | MG_DP_MODE_CFG_DP_X2_MODE;
-		ln1 |= MG_DP_MODE_CFG_DP_X1_MODE | MG_DP_MODE_CFG_DP_X2_MODE;
+	case 0x4:
+	case 0x6:
+		if (width == 1) {
+			ln0 |= MG_DP_MODE_CFG_DP_X1_MODE;
+			ln1 |= MG_DP_MODE_CFG_DP_X1_MODE;
+		} else {
+			ln0 |= MG_DP_MODE_CFG_DP_X2_MODE;
+			ln1 |= MG_DP_MODE_CFG_DP_X2_MODE;
+		}
 		break;
-
 	default:
-		MISSING_CASE(intel_dig_port->tc_mode);
-		return;
+		MISSING_CASE(pin_assignment);
 	}
 
-	I915_WRITE(MG_DP_MODE(0, port), ln0);
-	I915_WRITE(MG_DP_MODE(1, port), ln1);
+	if (INTEL_GEN(dev_priv) >= 12) {
+		I915_WRITE(HIP_INDEX_REG(tc_port), HIP_INDEX_VAL(tc_port, 0x0));
+		I915_WRITE(DKL_DP_MODE(tc_port), ln0);
+		I915_WRITE(HIP_INDEX_REG(tc_port), HIP_INDEX_VAL(tc_port, 0x1));
+		I915_WRITE(DKL_DP_MODE(tc_port), ln1);
+	} else {
+		I915_WRITE(MG_DP_MODE(0, tc_port), ln0);
+		I915_WRITE(MG_DP_MODE(1, tc_port), ln1);
+	}
 }
 
 static void intel_dp_sink_set_fec_ready(struct intel_dp *intel_dp,
@@ -3129,17 +3304,18 @@ static void intel_ddi_enable_fec(struct intel_encoder *encoder,
 				 const struct intel_crtc_state *crtc_state)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	enum port port = encoder->port;
+	struct intel_dp *intel_dp;
 	u32 val;
 
 	if (!crtc_state->fec_enable)
 		return;
 
-	val = I915_READ(DP_TP_CTL(port));
+	intel_dp = enc_to_intel_dp(&encoder->base);
+	val = I915_READ(intel_dp->regs.dp_tp_ctl);
 	val |= DP_TP_CTL_FEC_ENABLE;
-	I915_WRITE(DP_TP_CTL(port), val);
+	I915_WRITE(intel_dp->regs.dp_tp_ctl, val);
 
-	if (intel_de_wait_for_set(dev_priv, DP_TP_STATUS(port),
+	if (intel_de_wait_for_set(dev_priv, intel_dp->regs.dp_tp_status,
 				  DP_TP_STATUS_FEC_ENABLE_LIVE, 1))
 		DRM_ERROR("Timed out waiting for FEC Enable Status\n");
 }
@@ -3148,21 +3324,205 @@ static void intel_ddi_disable_fec_state(struct intel_encoder *encoder,
 					const struct intel_crtc_state *crtc_state)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	enum port port = encoder->port;
+	struct intel_dp *intel_dp;
 	u32 val;
 
 	if (!crtc_state->fec_enable)
 		return;
 
-	val = I915_READ(DP_TP_CTL(port));
+	intel_dp = enc_to_intel_dp(&encoder->base);
+	val = I915_READ(intel_dp->regs.dp_tp_ctl);
 	val &= ~DP_TP_CTL_FEC_ENABLE;
-	I915_WRITE(DP_TP_CTL(port), val);
-	POSTING_READ(DP_TP_CTL(port));
+	I915_WRITE(intel_dp->regs.dp_tp_ctl, val);
+	POSTING_READ(intel_dp->regs.dp_tp_ctl);
 }
 
-static void intel_ddi_pre_enable_dp(struct intel_encoder *encoder,
-				    const struct intel_crtc_state *crtc_state,
-				    const struct drm_connector_state *conn_state)
+static void
+tgl_clear_psr2_transcoder_exitline(const struct intel_crtc_state *cstate)
+{
+	struct drm_i915_private *dev_priv = to_i915(cstate->base.crtc->dev);
+	u32 val;
+
+	if (!cstate->dc3co_exitline)
+		return;
+
+	val = I915_READ(EXITLINE(cstate->cpu_transcoder));
+	val &= ~(EXITLINE_MASK | EXITLINE_ENABLE);
+	I915_WRITE(EXITLINE(cstate->cpu_transcoder), val);
+}
+
+static void
+tgl_set_psr2_transcoder_exitline(const struct intel_crtc_state *cstate)
+{
+	u32 val, exit_scanlines;
+	struct drm_i915_private *dev_priv = to_i915(cstate->base.crtc->dev);
+
+	if (!cstate->dc3co_exitline)
+		return;
+
+	exit_scanlines = cstate->dc3co_exitline;
+	exit_scanlines <<= EXITLINE_SHIFT;
+	val = I915_READ(EXITLINE(cstate->cpu_transcoder));
+	val &= ~(EXITLINE_MASK | EXITLINE_ENABLE);
+	val |= exit_scanlines;
+	val |= EXITLINE_ENABLE;
+	I915_WRITE(EXITLINE(cstate->cpu_transcoder), val);
+}
+
+static void tgl_dc3co_exitline_compute_config(struct intel_encoder *encoder,
+					      struct intel_crtc_state *cstate)
+{
+	u32 exit_scanlines;
+	struct drm_i915_private *dev_priv = to_i915(cstate->base.crtc->dev);
+	u32 crtc_vdisplay = cstate->base.adjusted_mode.crtc_vdisplay;
+
+	cstate->dc3co_exitline = 0;
+
+	if (!(dev_priv->csr.allowed_dc_mask & DC_STATE_EN_DC3CO))
+		return;
+
+	/* B.Specs:49196 DC3CO only works with pipeA and DDIA.*/
+	if (to_intel_crtc(cstate->base.crtc)->pipe != PIPE_A ||
+	    encoder->port != PORT_A)
+		return;
+
+	if (!cstate->has_psr2 || !cstate->base.active)
+		return;
+
+	/*
+	 * DC3CO Exit time 200us B.Spec 49196
+	 * PSR2 transcoder Early Exit scanlines = ROUNDUP(200 / line time) + 1
+	 */
+	exit_scanlines =
+		intel_usecs_to_scanlines(&cstate->base.adjusted_mode, 200) + 1;
+
+	if (WARN_ON(exit_scanlines > crtc_vdisplay))
+		return;
+
+	cstate->dc3co_exitline = crtc_vdisplay - exit_scanlines;
+	DRM_DEBUG_KMS("DC3CO exit scanlines %d\n", cstate->dc3co_exitline);
+}
+
+static void tgl_dc3co_exitline_get_config(struct intel_crtc_state *crtc_state)
+{
+	u32 val;
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+
+	if (INTEL_GEN(dev_priv) < 12)
+		return;
+
+	val = I915_READ(EXITLINE(crtc_state->cpu_transcoder));
+
+	if (val & EXITLINE_ENABLE)
+		crtc_state->dc3co_exitline = val & EXITLINE_MASK;
+}
+
+static void tgl_ddi_pre_enable_dp(struct intel_encoder *encoder,
+				  const struct intel_crtc_state *crtc_state,
+				  const struct drm_connector_state *conn_state)
+{
+	struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
+	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
+	enum phy phy = intel_port_to_phy(dev_priv, encoder->port);
+	struct intel_digital_port *dig_port = enc_to_dig_port(&encoder->base);
+	bool is_mst = intel_crtc_has_type(crtc_state, INTEL_OUTPUT_DP_MST);
+	int level = intel_ddi_dp_level(intel_dp);
+	enum transcoder transcoder = crtc_state->cpu_transcoder;
+
+	tgl_set_psr2_transcoder_exitline(crtc_state);
+	intel_dp_set_link_params(intel_dp, crtc_state->port_clock,
+				 crtc_state->lane_count, is_mst);
+
+	intel_dp->regs.dp_tp_ctl = TGL_DP_TP_CTL(transcoder);
+	intel_dp->regs.dp_tp_status = TGL_DP_TP_STATUS(transcoder);
+
+	/* 1.a got on intel_atomic_commit_tail() */
+
+	/* 2. */
+	intel_edp_panel_on(intel_dp);
+
+	/*
+	 * 1.b, 3. and 4.a is done before tgl_ddi_pre_enable_dp() by:
+	 * haswell_crtc_enable()->intel_encoders_pre_pll_enable() and
+	 * haswell_crtc_enable()->intel_enable_shared_dpll()
+	 */
+
+	/* 4.b */
+	intel_ddi_clk_select(encoder, crtc_state);
+
+	/* 5. */
+	if (!intel_phy_is_tc(dev_priv, phy) ||
+	    dig_port->tc_mode != TC_PORT_TBT_ALT)
+		intel_display_power_get(dev_priv,
+					dig_port->ddi_io_power_domain);
+
+	/* 6. */
+	icl_program_mg_dp_mode(dig_port, crtc_state);
+
+	/*
+	 * 7.a - Steps in this function should only be executed over MST
+	 * master, what will be taken in care by MST hook
+	 * intel_mst_pre_enable_dp()
+	 */
+	intel_ddi_enable_pipe_clock(crtc_state);
+
+	/* 7.b */
+	intel_ddi_config_transcoder_func(crtc_state);
+
+	/* 7.d */
+	icl_phy_set_clock_gating(dig_port, false);
+
+	/* 7.e */
+	tgl_ddi_vswing_sequence(encoder, crtc_state->port_clock, level,
+				encoder->type);
+
+	/* 7.f */
+	if (intel_phy_is_combo(dev_priv, phy)) {
+		bool lane_reversal =
+			dig_port->saved_port_bits & DDI_BUF_PORT_REVERSAL;
+
+		intel_combo_phy_power_up_lanes(dev_priv, phy, false,
+					       crtc_state->lane_count,
+					       lane_reversal);
+	}
+
+	/* 7.g */
+	intel_ddi_init_dp_buf_reg(encoder);
+
+	if (!is_mst)
+		intel_dp_sink_dpms(intel_dp, DRM_MODE_DPMS_ON);
+
+	intel_dp_sink_set_decompression_state(intel_dp, crtc_state, true);
+	/*
+	 * DDI FEC: "anticipates enabling FEC encoding sets the FEC_READY bit
+	 * in the FEC_CONFIGURATION register to 1 before initiating link
+	 * training
+	 */
+	intel_dp_sink_set_fec_ready(intel_dp, crtc_state);
+	/* 7.c, 7.h, 7.i, 7.j */
+	intel_dp_start_link_train(intel_dp);
+
+	/* 7.k */
+	if (!is_trans_port_sync_mode(crtc_state))
+		intel_dp_stop_link_train(intel_dp);
+
+	/*
+	 * TODO: enable clock gating
+	 *
+	 * It is not written in DP enabling sequence but "PHY Clockgating
+	 * programming" states that clock gating should be enabled after the
+	 * link training but doing so causes all the following trainings to fail
+	 * so not enabling it for now.
+	 */
+
+	/* 7.l */
+	intel_ddi_enable_fec(encoder, crtc_state);
+	intel_dsc_enable(encoder, crtc_state);
+}
+
+static void hsw_ddi_pre_enable_dp(struct intel_encoder *encoder,
+				  const struct intel_crtc_state *crtc_state,
+				  const struct drm_connector_state *conn_state)
 {
 	struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
@@ -3177,6 +3537,9 @@ static void intel_ddi_pre_enable_dp(struct intel_encoder *encoder,
 	intel_dp_set_link_params(intel_dp, crtc_state->port_clock,
 				 crtc_state->lane_count, is_mst);
 
+	intel_dp->regs.dp_tp_ctl = DP_TP_CTL(port);
+	intel_dp->regs.dp_tp_status = DP_TP_STATUS(port);
+
 	intel_edp_panel_on(intel_dp);
 
 	intel_ddi_clk_select(encoder, crtc_state);
@@ -3186,8 +3549,8 @@ static void intel_ddi_pre_enable_dp(struct intel_encoder *encoder,
 		intel_display_power_get(dev_priv,
 					dig_port->ddi_io_power_domain);
 
-	icl_program_mg_dp_mode(dig_port);
-	icl_disable_phy_clock_gating(dig_port);
+	icl_program_mg_dp_mode(dig_port, crtc_state);
+	icl_phy_set_clock_gating(dig_port, false);
 
 	if (INTEL_GEN(dev_priv) >= 11)
 		icl_ddi_vswing_sequence(encoder, crtc_state->port_clock,
@@ -3215,12 +3578,13 @@ static void intel_ddi_pre_enable_dp(struct intel_encoder *encoder,
 					      true);
 	intel_dp_sink_set_fec_ready(intel_dp, crtc_state);
 	intel_dp_start_link_train(intel_dp);
-	if (port != PORT_A || INTEL_GEN(dev_priv) >= 9)
+	if ((port != PORT_A || INTEL_GEN(dev_priv) >= 9) &&
+	    !is_trans_port_sync_mode(crtc_state))
 		intel_dp_stop_link_train(intel_dp);
 
 	intel_ddi_enable_fec(encoder, crtc_state);
 
-	icl_enable_phy_clock_gating(dig_port);
+	icl_phy_set_clock_gating(dig_port, true);
 
 	if (!is_mst)
 		intel_ddi_enable_pipe_clock(crtc_state);
@@ -3228,6 +3592,24 @@ static void intel_ddi_pre_enable_dp(struct intel_encoder *encoder,
 	intel_dsc_enable(encoder, crtc_state);
 }
 
+static void intel_ddi_pre_enable_dp(struct intel_encoder *encoder,
+				    const struct intel_crtc_state *crtc_state,
+				    const struct drm_connector_state *conn_state)
+{
+	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
+
+	if (INTEL_GEN(dev_priv) >= 12)
+		tgl_ddi_pre_enable_dp(encoder, crtc_state, conn_state);
+	else
+		hsw_ddi_pre_enable_dp(encoder, crtc_state, conn_state);
+
+	/* MST will call a setting of MSA after an allocating of Virtual Channel
+	 * from MST encoder pre_enable callback.
+	 */
+	if (!intel_crtc_has_type(crtc_state, INTEL_OUTPUT_DP_MST))
+		intel_ddi_set_dp_msa(crtc_state, conn_state);
+}
+
 static void intel_ddi_pre_enable_hdmi(struct intel_encoder *encoder,
 				      const struct intel_crtc_state *crtc_state,
 				      const struct drm_connector_state *conn_state)
@@ -3244,10 +3626,13 @@ static void intel_ddi_pre_enable_hdmi(struct intel_encoder *encoder,
 
 	intel_display_power_get(dev_priv, dig_port->ddi_io_power_domain);
 
-	icl_program_mg_dp_mode(dig_port);
-	icl_disable_phy_clock_gating(dig_port);
+	icl_program_mg_dp_mode(dig_port, crtc_state);
+	icl_phy_set_clock_gating(dig_port, false);
 
-	if (INTEL_GEN(dev_priv) >= 11)
+	if (INTEL_GEN(dev_priv) >= 12)
+		tgl_ddi_vswing_sequence(encoder, crtc_state->port_clock,
+					level, INTEL_OUTPUT_HDMI);
+	else if (INTEL_GEN(dev_priv) == 11)
 		icl_ddi_vswing_sequence(encoder, crtc_state->port_clock,
 					level, INTEL_OUTPUT_HDMI);
 	else if (IS_CANNONLAKE(dev_priv))
@@ -3257,7 +3642,7 @@ static void intel_ddi_pre_enable_hdmi(struct intel_encoder *encoder,
 	else
 		intel_prepare_hdmi_ddi_buffers(encoder, level);
 
-	icl_enable_phy_clock_gating(dig_port);
+	icl_phy_set_clock_gating(dig_port, true);
 
 	if (IS_GEN9_BC(dev_priv))
 		skl_ddi_set_iboost(encoder, level, INTEL_OUTPUT_HDMI);
@@ -3330,10 +3715,14 @@ static void intel_disable_ddi_buf(struct intel_encoder *encoder,
 		wait = true;
 	}
 
-	val = I915_READ(DP_TP_CTL(port));
-	val &= ~(DP_TP_CTL_ENABLE | DP_TP_CTL_LINK_TRAIN_MASK);
-	val |= DP_TP_CTL_LINK_TRAIN_PAT1;
-	I915_WRITE(DP_TP_CTL(port), val);
+	if (intel_crtc_has_dp_encoder(crtc_state)) {
+		struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
+
+		val = I915_READ(intel_dp->regs.dp_tp_ctl);
+		val &= ~(DP_TP_CTL_ENABLE | DP_TP_CTL_LINK_TRAIN_MASK);
+		val |= DP_TP_CTL_LINK_TRAIN_PAT1;
+		I915_WRITE(intel_dp->regs.dp_tp_ctl, val);
+	}
 
 	/* Disable FEC in DP Sink */
 	intel_ddi_disable_fec_state(encoder, crtc_state);
@@ -3373,6 +3762,7 @@ static void intel_ddi_post_disable_dp(struct intel_encoder *encoder,
 						  dig_port->ddi_io_power_domain);
 
 	intel_ddi_clk_disable(encoder);
+	tgl_clear_psr2_transcoder_exitline(old_crtc_state);
 }
 
 static void intel_ddi_post_disable_hdmi(struct intel_encoder *encoder,
@@ -3475,7 +3865,8 @@ static void intel_enable_ddi_dp(struct intel_encoder *encoder,
 
 	intel_edp_backlight_on(crtc_state, conn_state);
 	intel_psr_enable(intel_dp, crtc_state);
-	intel_dp_ycbcr_420_enable(intel_dp, crtc_state);
+	intel_dp_vsc_enable(intel_dp, crtc_state, conn_state);
+	intel_dp_hdr_metadata_enable(intel_dp, crtc_state, conn_state);
 	intel_edp_drrs_enable(intel_dp, crtc_state);
 
 	if (crtc_state->has_audio)
@@ -3486,12 +3877,12 @@ static i915_reg_t
 gen9_chicken_trans_reg_by_port(struct drm_i915_private *dev_priv,
 			       enum port port)
 {
-	static const i915_reg_t regs[] = {
-		[PORT_A] = CHICKEN_TRANS_EDP,
-		[PORT_B] = CHICKEN_TRANS_A,
-		[PORT_C] = CHICKEN_TRANS_B,
-		[PORT_D] = CHICKEN_TRANS_C,
-		[PORT_E] = CHICKEN_TRANS_A,
+	static const enum transcoder trans[] = {
+		[PORT_A] = TRANSCODER_EDP,
+		[PORT_B] = TRANSCODER_A,
+		[PORT_C] = TRANSCODER_B,
+		[PORT_D] = TRANSCODER_C,
+		[PORT_E] = TRANSCODER_A,
 	};
 
 	WARN_ON(INTEL_GEN(dev_priv) < 9);
@@ -3499,7 +3890,7 @@ gen9_chicken_trans_reg_by_port(struct drm_i915_private *dev_priv,
 	if (WARN_ON(port < PORT_A || port > PORT_E))
 		port = PORT_A;
 
-	return regs[port];
+	return CHICKEN_TRANS(trans[port]);
 }
 
 static void intel_enable_ddi_hdmi(struct intel_encoder *encoder,
@@ -3633,7 +4024,7 @@ static void intel_ddi_update_pipe_dp(struct intel_encoder *encoder,
 {
 	struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
 
-	intel_ddi_set_pipe_settings(crtc_state);
+	intel_ddi_set_dp_msa(crtc_state, conn_state);
 
 	intel_psr_update(intel_dp, crtc_state);
 	intel_edp_drrs_enable(intel_dp, crtc_state);
@@ -3761,7 +4152,7 @@ static void intel_ddi_prepare_link_retrain(struct intel_dp *intel_dp)
 	u32 val;
 	bool wait = false;
 
-	if (I915_READ(DP_TP_CTL(port)) & DP_TP_CTL_ENABLE) {
+	if (I915_READ(intel_dp->regs.dp_tp_ctl) & DP_TP_CTL_ENABLE) {
 		val = I915_READ(DDI_BUF_CTL(port));
 		if (val & DDI_BUF_CTL_ENABLE) {
 			val &= ~DDI_BUF_CTL_ENABLE;
@@ -3769,11 +4160,11 @@ static void intel_ddi_prepare_link_retrain(struct intel_dp *intel_dp)
 			wait = true;
 		}
 
-		val = I915_READ(DP_TP_CTL(port));
+		val = I915_READ(intel_dp->regs.dp_tp_ctl);
 		val &= ~(DP_TP_CTL_ENABLE | DP_TP_CTL_LINK_TRAIN_MASK);
 		val |= DP_TP_CTL_LINK_TRAIN_PAT1;
-		I915_WRITE(DP_TP_CTL(port), val);
-		POSTING_READ(DP_TP_CTL(port));
+		I915_WRITE(intel_dp->regs.dp_tp_ctl, val);
+		POSTING_READ(intel_dp->regs.dp_tp_ctl);
 
 		if (wait)
 			intel_wait_ddi_buf_idle(dev_priv, port);
@@ -3788,8 +4179,8 @@ static void intel_ddi_prepare_link_retrain(struct intel_dp *intel_dp)
 		if (drm_dp_enhanced_frame_cap(intel_dp->dpcd))
 			val |= DP_TP_CTL_ENHANCED_FRAME_ENABLE;
 	}
-	I915_WRITE(DP_TP_CTL(port), val);
-	POSTING_READ(DP_TP_CTL(port));
+	I915_WRITE(intel_dp->regs.dp_tp_ctl, val);
+	POSTING_READ(intel_dp->regs.dp_tp_ctl);
 
 	intel_dp->DP |= DDI_BUF_CTL_ENABLE;
 	I915_WRITE(DDI_BUF_CTL(port), intel_dp->DP);
@@ -3891,6 +4282,23 @@ void intel_ddi_get_config(struct intel_encoder *encoder,
 		pipe_config->lane_count =
 			((temp & DDI_PORT_WIDTH_MASK) >> DDI_PORT_WIDTH_SHIFT) + 1;
 		intel_dp_get_m_n(intel_crtc, pipe_config);
+
+		if (INTEL_GEN(dev_priv) >= 11) {
+			i915_reg_t dp_tp_ctl;
+
+			if (IS_GEN(dev_priv, 11))
+				dp_tp_ctl = DP_TP_CTL(encoder->port);
+			else
+				dp_tp_ctl = TGL_DP_TP_CTL(pipe_config->cpu_transcoder);
+
+			pipe_config->fec_enable =
+				I915_READ(dp_tp_ctl) & DP_TP_CTL_FEC_ENABLE;
+
+			DRM_DEBUG_KMS("[ENCODER:%d:%s] Fec status: %u\n",
+				      encoder->base.base.id, encoder->base.name,
+				      pipe_config->fec_enable);
+		}
+
 		break;
 	case TRANS_DDI_MODE_SELECT_DP_MST:
 		pipe_config->output_types |= BIT(INTEL_OUTPUT_DP_MST);
@@ -3902,6 +4310,9 @@ void intel_ddi_get_config(struct intel_encoder *encoder,
 		break;
 	}
 
+	if (encoder->type == INTEL_OUTPUT_EDP)
+		tgl_dc3co_exitline_get_config(pipe_config);
+
 	pipe_config->has_audio =
 		intel_ddi_is_audio_enabled(dev_priv, cpu_transcoder);
 
@@ -3979,10 +4390,13 @@ static int intel_ddi_compute_config(struct intel_encoder *encoder,
 	if (HAS_TRANSCODER_EDP(dev_priv) && port == PORT_A)
 		pipe_config->cpu_transcoder = TRANSCODER_EDP;
 
-	if (intel_crtc_has_type(pipe_config, INTEL_OUTPUT_HDMI))
+	if (intel_crtc_has_type(pipe_config, INTEL_OUTPUT_HDMI)) {
 		ret = intel_hdmi_compute_config(encoder, pipe_config, conn_state);
-	else
+	} else {
 		ret = intel_dp_compute_config(encoder, pipe_config, conn_state);
+		tgl_dc3co_exitline_compute_config(encoder, pipe_config);
+	}
+
 	if (ret)
 		return ret;
 
@@ -4276,7 +4690,6 @@ void intel_ddi_init(struct drm_i915_private *dev_priv, enum port port)
 	struct intel_encoder *intel_encoder;
 	struct drm_encoder *encoder;
 	bool init_hdmi, init_dp, init_lspcon = false;
-	enum pipe pipe;
 	enum phy phy = intel_port_to_phy(dev_priv, port);
 
 	init_hdmi = port_info->supports_dvi || port_info->supports_hdmi;
@@ -4328,8 +4741,7 @@ void intel_ddi_init(struct drm_i915_private *dev_priv, enum port port)
 	intel_encoder->power_domain = intel_port_to_power_domain(port);
 	intel_encoder->port = port;
 	intel_encoder->cloneable = 0;
-	for_each_pipe(dev_priv, pipe)
-		intel_encoder->crtc_mask |= BIT(pipe);
+	intel_encoder->pipe_mask = ~0;
 
 	if (INTEL_GEN(dev_priv) >= 11)
 		intel_dig_port->saved_port_bits = I915_READ(DDI_BUF_CTL(port)) &
@@ -4351,46 +4763,9 @@ void intel_ddi_init(struct drm_i915_private *dev_priv, enum port port)
 		intel_encoder->update_complete = intel_ddi_update_complete;
 	}
 
-	switch (port) {
-	case PORT_A:
-		intel_dig_port->ddi_io_power_domain =
-			POWER_DOMAIN_PORT_DDI_A_IO;
-		break;
-	case PORT_B:
-		intel_dig_port->ddi_io_power_domain =
-			POWER_DOMAIN_PORT_DDI_B_IO;
-		break;
-	case PORT_C:
-		intel_dig_port->ddi_io_power_domain =
-			POWER_DOMAIN_PORT_DDI_C_IO;
-		break;
-	case PORT_D:
-		intel_dig_port->ddi_io_power_domain =
-			POWER_DOMAIN_PORT_DDI_D_IO;
-		break;
-	case PORT_E:
-		intel_dig_port->ddi_io_power_domain =
-			POWER_DOMAIN_PORT_DDI_E_IO;
-		break;
-	case PORT_F:
-		intel_dig_port->ddi_io_power_domain =
-			POWER_DOMAIN_PORT_DDI_F_IO;
-		break;
-	case PORT_G:
-		intel_dig_port->ddi_io_power_domain =
-			POWER_DOMAIN_PORT_DDI_G_IO;
-		break;
-	case PORT_H:
-		intel_dig_port->ddi_io_power_domain =
-			POWER_DOMAIN_PORT_DDI_H_IO;
-		break;
-	case PORT_I:
-		intel_dig_port->ddi_io_power_domain =
-			POWER_DOMAIN_PORT_DDI_I_IO;
-		break;
-	default:
-		MISSING_CASE(port);
-	}
+	WARN_ON(port > PORT_I);
+	intel_dig_port->ddi_io_power_domain = POWER_DOMAIN_PORT_DDI_A_IO +
+					      port - PORT_A;
 
 	if (init_dp) {
 		if (!intel_ddi_init_dp_connector(intel_dig_port))
diff --git a/drivers/gpu/drm/i915/display/intel_ddi.h b/drivers/gpu/drm/i915/display/intel_ddi.h
index a08365da2643..19aeab1246ee 100644
--- a/drivers/gpu/drm/i915/display/intel_ddi.h
+++ b/drivers/gpu/drm/i915/display/intel_ddi.h
@@ -30,7 +30,8 @@ void intel_ddi_enable_transcoder_func(const struct intel_crtc_state *crtc_state)
 void intel_ddi_disable_transcoder_func(const struct intel_crtc_state *crtc_state);
 void intel_ddi_enable_pipe_clock(const struct intel_crtc_state *crtc_state);
 void intel_ddi_disable_pipe_clock(const  struct intel_crtc_state *crtc_state);
-void intel_ddi_set_pipe_settings(const struct intel_crtc_state *crtc_state);
+void intel_ddi_set_dp_msa(const struct intel_crtc_state *crtc_state,
+			  const struct drm_connector_state *conn_state);
 bool intel_ddi_connector_get_hw_state(struct intel_connector *intel_connector);
 void intel_ddi_get_config(struct intel_encoder *encoder,
 			  struct intel_crtc_state *pipe_config);
diff --git a/drivers/gpu/drm/i915/display/intel_display.c b/drivers/gpu/drm/i915/display/intel_display.c
index b51d1ceb8739..6f5e3bd13ad1 100644
--- a/drivers/gpu/drm/i915/display/intel_display.c
+++ b/drivers/gpu/drm/i915/display/intel_display.c
@@ -31,7 +31,6 @@
 #include <linux/module.h>
 #include <linux/dma-resv.h>
 #include <linux/slab.h>
-#include <linux/vgaarb.h>
 
 #include <drm/drm_atomic.h>
 #include <drm/drm_atomic_helper.h>
@@ -56,6 +55,8 @@
 #include "display/intel_tv.h"
 #include "display/intel_vdsc.h"
 
+#include "gt/intel_rps.h"
+
 #include "i915_drv.h"
 #include "i915_trace.h"
 #include "intel_acpi.h"
@@ -65,6 +66,7 @@
 #include "intel_cdclk.h"
 #include "intel_color.h"
 #include "intel_display_types.h"
+#include "intel_dp_link_training.h"
 #include "intel_fbc.h"
 #include "intel_fbdev.h"
 #include "intel_fifo_underrun.h"
@@ -79,6 +81,7 @@
 #include "intel_sideband.h"
 #include "intel_sprite.h"
 #include "intel_tc.h"
+#include "intel_vga.h"
 
 /* Primary plane formats for gen <= 3 */
 static const u32 i8xx_primary_formats[] = {
@@ -88,7 +91,17 @@ static const u32 i8xx_primary_formats[] = {
 	DRM_FORMAT_XRGB8888,
 };
 
-/* Primary plane formats for gen >= 4 */
+/* Primary plane formats for ivb (no fp16 due to hw issue) */
+static const u32 ivb_primary_formats[] = {
+	DRM_FORMAT_C8,
+	DRM_FORMAT_RGB565,
+	DRM_FORMAT_XRGB8888,
+	DRM_FORMAT_XBGR8888,
+	DRM_FORMAT_XRGB2101010,
+	DRM_FORMAT_XBGR2101010,
+};
+
+/* Primary plane formats for gen >= 4, except ivb */
 static const u32 i965_primary_formats[] = {
 	DRM_FORMAT_C8,
 	DRM_FORMAT_RGB565,
@@ -96,6 +109,7 @@ static const u32 i965_primary_formats[] = {
 	DRM_FORMAT_XBGR8888,
 	DRM_FORMAT_XRGB2101010,
 	DRM_FORMAT_XBGR2101010,
+	DRM_FORMAT_XBGR16161616F,
 };
 
 static const u64 i9xx_format_modifiers[] = {
@@ -135,8 +149,6 @@ static void vlv_prepare_pll(struct intel_crtc *crtc,
 			    const struct intel_crtc_state *pipe_config);
 static void chv_prepare_pll(struct intel_crtc *crtc,
 			    const struct intel_crtc_state *pipe_config);
-static void intel_begin_crtc_commit(struct intel_atomic_state *, struct intel_crtc *);
-static void intel_finish_crtc_commit(struct intel_atomic_state *, struct intel_crtc *);
 static void intel_crtc_init_scalers(struct intel_crtc *crtc,
 				    struct intel_crtc_state *crtc_state);
 static void skylake_pfit_enable(const struct intel_crtc_state *crtc_state);
@@ -490,7 +502,7 @@ static const struct intel_limit intel_limits_bxt = {
 
 /* WA Display #0827: Gen9:all */
 static void
-skl_wa_827(struct drm_i915_private *dev_priv, int pipe, bool enable)
+skl_wa_827(struct drm_i915_private *dev_priv, enum pipe pipe, bool enable)
 {
 	if (enable)
 		I915_WRITE(CLKGATE_DIS_PSL(pipe),
@@ -521,6 +533,20 @@ needs_modeset(const struct intel_crtc_state *state)
 	return drm_atomic_crtc_needs_modeset(&state->base);
 }
 
+bool
+is_trans_port_sync_mode(const struct intel_crtc_state *crtc_state)
+{
+	return (crtc_state->master_transcoder != INVALID_TRANSCODER ||
+		crtc_state->sync_mode_slaves_mask);
+}
+
+static bool
+is_trans_port_sync_master(const struct intel_crtc_state *crtc_state)
+{
+	return (crtc_state->master_transcoder == INVALID_TRANSCODER &&
+		crtc_state->sync_mode_slaves_mask);
+}
+
 /*
  * Platform specific helpers to calculate the port PLL loopback- (clock.m),
  * and post-divider (clock.p) values, pre- (clock.vco) and post-divided fast
@@ -1612,8 +1638,8 @@ void vlv_wait_port_ready(struct drm_i915_private *dev_priv,
 
 	if (intel_de_wait_for_register(dev_priv, dpll_reg,
 				       port_mask, expected_mask, 1000))
-		WARN(1, "timed out waiting for port %c ready: got 0x%x, expected 0x%x\n",
-		     port_name(dport->base.port),
+		WARN(1, "timed out waiting for [ENCODER:%d:%s] port ready: got 0x%x, expected 0x%x\n",
+		     dport->base.base.base.id, dport->base.base.name,
 		     I915_READ(dpll_reg) & port_mask, expected_mask);
 }
 
@@ -2079,7 +2105,8 @@ intel_pin_and_fence_fb_obj(struct drm_framebuffer *fb,
 	unsigned int pinctl;
 	u32 alignment;
 
-	WARN_ON(!mutex_is_locked(&dev->struct_mutex));
+	if (WARN_ON(!i915_gem_object_is_framebuffer(obj)))
+		return ERR_PTR(-EINVAL);
 
 	alignment = intel_surf_alignment(fb, 0);
 
@@ -2161,8 +2188,6 @@ err:
 
 void intel_unpin_fb_vma(struct i915_vma *vma, unsigned long flags)
 {
-	lockdep_assert_held(&vma->vm->i915->drm.struct_mutex);
-
 	i915_gem_object_lock(vma->obj);
 	if (flags & PLANE_HAS_FENCE)
 		i915_vma_unpin_fence(vma);
@@ -2504,6 +2529,9 @@ u32 intel_plane_fb_max_stride(struct drm_i915_private *dev_priv,
 	 * the highest stride limits of them all.
 	 */
 	crtc = intel_get_crtc_for_pipe(dev_priv, PIPE_A);
+	if (!crtc)
+		return 0;
+
 	plane = to_intel_plane(crtc->base.primary);
 
 	return plane->max_stride(plane, pixel_format, modifier,
@@ -2736,10 +2764,7 @@ intel_fill_fb_info(struct drm_i915_private *dev_priv,
 				size++;
 
 			/* rotate the x/y offsets to match the GTT view */
-			r.x1 = x;
-			r.y1 = y;
-			r.x2 = x + width;
-			r.y2 = y + height;
+			drm_rect_init(&r, x, y, width, height);
 			drm_rect_rotate(&r,
 					rot_info->plane[i].width * tile_width,
 					rot_info->plane[i].height * tile_height,
@@ -2861,10 +2886,7 @@ intel_plane_remap_gtt(struct intel_plane_state *plane_state)
 			struct drm_rect r;
 
 			/* rotate the x/y offsets to match the GTT view */
-			r.x1 = x;
-			r.y1 = y;
-			r.x2 = x + width;
-			r.y2 = y + height;
+			drm_rect_init(&r, x, y, width, height);
 			drm_rect_rotate(&r,
 					info->plane[i].width * tile_width,
 					info->plane[i].height * tile_height,
@@ -2966,6 +2988,8 @@ static int i9xx_format_to_fourcc(int format)
 		return DRM_FORMAT_XRGB2101010;
 	case DISPPLANE_RGBX101010:
 		return DRM_FORMAT_XBGR2101010;
+	case DISPPLANE_RGBX161616:
+		return DRM_FORMAT_XBGR16161616F;
 	}
 }
 
@@ -3063,13 +3087,11 @@ intel_alloc_initial_plane_obj(struct intel_crtc *crtc,
 		return false;
 	}
 
-	mutex_lock(&dev->struct_mutex);
 	obj = i915_gem_object_create_stolen_for_preallocated(dev_priv,
 							     base_aligned,
 							     base_aligned,
 							     size_aligned);
-	mutex_unlock(&dev->struct_mutex);
-	if (!obj)
+	if (IS_ERR(obj))
 		return false;
 
 	switch (plane_config->tiling) {
@@ -3151,6 +3173,7 @@ static void intel_plane_disable_noatomic(struct intel_crtc *crtc,
 	intel_set_plane_visible(crtc_state, plane_state, false);
 	fixup_active_planes(crtc_state);
 	crtc_state->data_rate[plane->id] = 0;
+	crtc_state->min_cdclk[plane->id] = 0;
 
 	if (plane->id == PLANE_PRIMARY)
 		intel_pre_disable_primary_noatomic(&crtc->base);
@@ -3230,13 +3253,11 @@ valid_fb:
 	intel_state->color_plane[0].stride =
 		intel_fb_pitch(fb, 0, intel_state->base.rotation);
 
-	mutex_lock(&dev->struct_mutex);
 	intel_state->vma =
 		intel_pin_and_fence_fb_obj(fb,
 					   &intel_state->view,
 					   intel_plane_uses_fence(intel_state),
 					   &intel_state->flags);
-	mutex_unlock(&dev->struct_mutex);
 	if (IS_ERR(intel_state->vma)) {
 		DRM_ERROR("failed to pin boot fb on pipe %d: %li\n",
 			  intel_crtc->pipe, PTR_ERR(intel_state->vma));
@@ -3280,7 +3301,20 @@ static int skl_max_plane_width(const struct drm_framebuffer *fb,
 	switch (fb->modifier) {
 	case DRM_FORMAT_MOD_LINEAR:
 	case I915_FORMAT_MOD_X_TILED:
-		return 4096;
+		/*
+		 * Validated limit is 4k, but has 5k should
+		 * work apart from the following features:
+		 * - Ytile (already limited to 4k)
+		 * - FP16 (already limited to 4k)
+		 * - render compression (already limited to 4k)
+		 * - KVMR sprite and cursor (don't care)
+		 * - horizontal panning (TODO verify this)
+		 * - pipe and plane scaling (TODO verify this)
+		 */
+		if (cpp == 8)
+			return 4096;
+		else
+			return 5120;
 	case I915_FORMAT_MOD_Y_TILED_CCS:
 	case I915_FORMAT_MOD_Yf_TILED_CCS:
 		/* FIXME AUX plane? */
@@ -3331,6 +3365,16 @@ static int icl_max_plane_width(const struct drm_framebuffer *fb,
 	return 5120;
 }
 
+static int skl_max_plane_height(void)
+{
+	return 4096;
+}
+
+static int icl_max_plane_height(void)
+{
+	return 4320;
+}
+
 static bool skl_check_main_ccs_coordinates(struct intel_plane_state *plane_state,
 					   int main_x, int main_y, u32 main_offset)
 {
@@ -3379,7 +3423,7 @@ static int skl_check_main_surface(struct intel_plane_state *plane_state)
 	int w = drm_rect_width(&plane_state->base.src) >> 16;
 	int h = drm_rect_height(&plane_state->base.src) >> 16;
 	int max_width;
-	int max_height = 4096;
+	int max_height;
 	u32 alignment, offset, aux_offset = plane_state->color_plane[1].offset;
 
 	if (INTEL_GEN(dev_priv) >= 11)
@@ -3389,6 +3433,11 @@ static int skl_check_main_surface(struct intel_plane_state *plane_state)
 	else
 		max_width = skl_max_plane_width(fb, 0, rotation);
 
+	if (INTEL_GEN(dev_priv) >= 11)
+		max_height = icl_max_plane_height();
+	else
+		max_height = skl_max_plane_height();
+
 	if (w > max_width || h > max_height) {
 		DRM_DEBUG_KMS("requested Y/RGB source size %dx%d too big (limit %dx%d)\n",
 			      w, h, max_width, max_height);
@@ -3455,9 +3504,8 @@ static int skl_check_main_surface(struct intel_plane_state *plane_state)
 	 * Put the final coordinates back so that the src
 	 * coordinate checks will see the right values.
 	 */
-	drm_rect_translate(&plane_state->base.src,
-			   (x << 16) - plane_state->base.src.x1,
-			   (y << 16) - plane_state->base.src.y1);
+	drm_rect_translate_to(&plane_state->base.src,
+			      x << 16, y << 16);
 
 	return 0;
 }
@@ -3528,7 +3576,7 @@ int skl_check_plane_surface(struct intel_plane_state *plane_state)
 	 * Handle the AUX surface first since
 	 * the main surface setup depends on it.
 	 */
-	if (is_planar_yuv_format(fb->format->format)) {
+	if (drm_format_info_is_yuv_semiplanar(fb->format)) {
 		ret = skl_check_nv12_aux_surface(plane_state);
 		if (ret)
 			return ret;
@@ -3549,6 +3597,53 @@ int skl_check_plane_surface(struct intel_plane_state *plane_state)
 	return 0;
 }
 
+static void i9xx_plane_ratio(const struct intel_crtc_state *crtc_state,
+			     const struct intel_plane_state *plane_state,
+			     unsigned int *num, unsigned int *den)
+{
+	const struct drm_framebuffer *fb = plane_state->base.fb;
+	unsigned int cpp = fb->format->cpp[0];
+
+	/*
+	 * g4x bspec says 64bpp pixel rate can't exceed 80%
+	 * of cdclk when the sprite plane is enabled on the
+	 * same pipe. ilk/snb bspec says 64bpp pixel rate is
+	 * never allowed to exceed 80% of cdclk. Let's just go
+	 * with the ilk/snb limit always.
+	 */
+	if (cpp == 8) {
+		*num = 10;
+		*den = 8;
+	} else {
+		*num = 1;
+		*den = 1;
+	}
+}
+
+static int i9xx_plane_min_cdclk(const struct intel_crtc_state *crtc_state,
+				const struct intel_plane_state *plane_state)
+{
+	unsigned int pixel_rate;
+	unsigned int num, den;
+
+	/*
+	 * Note that crtc_state->pixel_rate accounts for both
+	 * horizontal and vertical panel fitter downscaling factors.
+	 * Pre-HSW bspec tells us to only consider the horizontal
+	 * downscaling factor here. We ignore that and just consider
+	 * both for simplicity.
+	 */
+	pixel_rate = crtc_state->pixel_rate;
+
+	i9xx_plane_ratio(crtc_state, plane_state, &num, &den);
+
+	/* two pixels per clock with double wide pipe */
+	if (crtc_state->double_wide)
+		den *= 2;
+
+	return DIV_ROUND_UP(pixel_rate * num, den);
+}
+
 unsigned int
 i9xx_plane_max_stride(struct intel_plane *plane,
 		      u32 pixel_format, u64 modifier,
@@ -3631,6 +3726,9 @@ static u32 i9xx_plane_ctl(const struct intel_crtc_state *crtc_state,
 	case DRM_FORMAT_XBGR2101010:
 		dspcntr |= DISPPLANE_RGBX101010;
 		break;
+	case DRM_FORMAT_XBGR16161616F:
+		dspcntr |= DISPPLANE_RGBX161616;
+		break;
 	default:
 		MISSING_CASE(fb->format->format);
 		return 0;
@@ -3653,7 +3751,8 @@ int i9xx_check_plane_surface(struct intel_plane_state *plane_state)
 {
 	struct drm_i915_private *dev_priv =
 		to_i915(plane_state->base.plane->dev);
-	int src_x, src_y;
+	const struct drm_framebuffer *fb = plane_state->base.fb;
+	int src_x, src_y, src_w;
 	u32 offset;
 	int ret;
 
@@ -3664,9 +3763,14 @@ int i9xx_check_plane_surface(struct intel_plane_state *plane_state)
 	if (!plane_state->base.visible)
 		return 0;
 
+	src_w = drm_rect_width(&plane_state->base.src) >> 16;
 	src_x = plane_state->base.src.x1 >> 16;
 	src_y = plane_state->base.src.y1 >> 16;
 
+	/* Undocumented hardware limit on i965/g4x/vlv/chv */
+	if (HAS_GMCH(dev_priv) && fb->format->cpp[0] == 8 && src_w > 2048)
+		return -EINVAL;
+
 	intel_add_fb_offsets(&src_x, &src_y, plane_state, 0);
 
 	if (INTEL_GEN(dev_priv) >= 4)
@@ -3679,9 +3783,8 @@ int i9xx_check_plane_surface(struct intel_plane_state *plane_state)
 	 * Put the final coordinates back so that the src
 	 * coordinate checks will see the right values.
 	 */
-	drm_rect_translate(&plane_state->base.src,
-			   (src_x << 16) - plane_state->base.src.x1,
-			   (src_y << 16) - plane_state->base.src.y1);
+	drm_rect_translate_to(&plane_state->base.src,
+			      src_x << 16, src_y << 16);
 
 	/* HSW/BDW do this automagically in hardware */
 	if (!IS_HASWELL(dev_priv) && !IS_BROADWELL(dev_priv)) {
@@ -4211,7 +4314,7 @@ __intel_display_resume(struct drm_device *dev,
 	int i, ret;
 
 	intel_modeset_setup_hw_state(dev, ctx);
-	i915_redisable_vga(to_i915(dev));
+	intel_vga_redisable(to_i915(dev));
 
 	if (!state)
 		return 0;
@@ -4243,7 +4346,7 @@ __intel_display_resume(struct drm_device *dev,
 static bool gpu_reset_clobbers_display(struct drm_i915_private *dev_priv)
 {
 	return (INTEL_INFO(dev_priv)->gpu_reset_clobbers_display &&
-		intel_has_gpu_reset(dev_priv));
+		intel_has_gpu_reset(&dev_priv->gt));
 }
 
 void intel_prepare_reset(struct drm_i915_private *dev_priv)
@@ -4330,7 +4433,7 @@ void intel_finish_reset(struct drm_i915_private *dev_priv)
 		 * so need a full re-initialization.
 		 */
 		intel_pps_unlock_regs_wa(dev_priv);
-		intel_modeset_init_hw(dev);
+		intel_modeset_init_hw(dev_priv);
 		intel_init_clock_gating(dev_priv);
 
 		spin_lock_irq(&dev_priv->irq_lock);
@@ -4378,50 +4481,60 @@ static void icl_set_pipe_chicken(struct intel_crtc *crtc)
 	I915_WRITE(PIPE_CHICKEN(pipe), tmp);
 }
 
-static void intel_update_pipe_config(const struct intel_crtc_state *old_crtc_state,
-				     const struct intel_crtc_state *new_crtc_state)
+static void icl_enable_trans_port_sync(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(new_crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
-
-	/* drm_atomic_helper_update_legacy_modeset_state might not be called. */
-	crtc->base.mode = new_crtc_state->base.mode;
+	u32 trans_ddi_func_ctl2_val;
+	u8 master_select;
 
 	/*
-	 * Update pipe size and adjust fitter if needed: the reason for this is
-	 * that in compute_mode_changes we check the native mode (not the pfit
-	 * mode) to see if we can flip rather than do a full mode set. In the
-	 * fastboot case, we'll flip, but if we don't update the pipesrc and
-	 * pfit state, we'll end up with a big fb scanned out into the wrong
-	 * sized surface.
+	 * Configure the master select and enable Transcoder Port Sync for
+	 * Slave CRTCs transcoder.
 	 */
+	if (crtc_state->master_transcoder == INVALID_TRANSCODER)
+		return;
 
-	I915_WRITE(PIPESRC(crtc->pipe),
-		   ((new_crtc_state->pipe_src_w - 1) << 16) |
-		   (new_crtc_state->pipe_src_h - 1));
+	if (crtc_state->master_transcoder == TRANSCODER_EDP)
+		master_select = 0;
+	else
+		master_select = crtc_state->master_transcoder + 1;
 
-	/* on skylake this is done by detaching scalers */
-	if (INTEL_GEN(dev_priv) >= 9) {
-		skl_detach_scalers(new_crtc_state);
+	/* Set the master select bits for Tranascoder Port Sync */
+	trans_ddi_func_ctl2_val = (PORT_SYNC_MODE_MASTER_SELECT(master_select) &
+				   PORT_SYNC_MODE_MASTER_SELECT_MASK) <<
+		PORT_SYNC_MODE_MASTER_SELECT_SHIFT;
+	/* Enable Transcoder Port Sync */
+	trans_ddi_func_ctl2_val |= PORT_SYNC_MODE_ENABLE;
 
-		if (new_crtc_state->pch_pfit.enabled)
-			skylake_pfit_enable(new_crtc_state);
-	} else if (HAS_PCH_SPLIT(dev_priv)) {
-		if (new_crtc_state->pch_pfit.enabled)
-			ironlake_pfit_enable(new_crtc_state);
-		else if (old_crtc_state->pch_pfit.enabled)
-			ironlake_pfit_disable(old_crtc_state);
-	}
+	I915_WRITE(TRANS_DDI_FUNC_CTL2(crtc_state->cpu_transcoder),
+		   trans_ddi_func_ctl2_val);
+}
 
-	if (INTEL_GEN(dev_priv) >= 11)
-		icl_set_pipe_chicken(crtc);
+static void icl_disable_transcoder_port_sync(const struct intel_crtc_state *old_crtc_state)
+{
+	struct intel_crtc *crtc = to_intel_crtc(old_crtc_state->base.crtc);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	i915_reg_t reg;
+	u32 trans_ddi_func_ctl2_val;
+
+	if (old_crtc_state->master_transcoder == INVALID_TRANSCODER)
+		return;
+
+	DRM_DEBUG_KMS("Disabling Transcoder Port Sync on Slave Transcoder %s\n",
+		      transcoder_name(old_crtc_state->cpu_transcoder));
+
+	reg = TRANS_DDI_FUNC_CTL2(old_crtc_state->cpu_transcoder);
+	trans_ddi_func_ctl2_val = ~(PORT_SYNC_MODE_ENABLE |
+				    PORT_SYNC_MODE_MASTER_SELECT_MASK);
+	I915_WRITE(reg, trans_ddi_func_ctl2_val);
 }
 
 static void intel_fdi_normal_train(struct intel_crtc *crtc)
 {
 	struct drm_device *dev = crtc->base.dev;
 	struct drm_i915_private *dev_priv = to_i915(dev);
-	int pipe = crtc->pipe;
+	enum pipe pipe = crtc->pipe;
 	i915_reg_t reg;
 	u32 temp;
 
@@ -4464,7 +4577,7 @@ static void ironlake_fdi_link_train(struct intel_crtc *crtc,
 {
 	struct drm_device *dev = crtc->base.dev;
 	struct drm_i915_private *dev_priv = to_i915(dev);
-	int pipe = crtc->pipe;
+	enum pipe pipe = crtc->pipe;
 	i915_reg_t reg;
 	u32 temp, tries;
 
@@ -4565,7 +4678,7 @@ static void gen6_fdi_link_train(struct intel_crtc *crtc,
 {
 	struct drm_device *dev = crtc->base.dev;
 	struct drm_i915_private *dev_priv = to_i915(dev);
-	int pipe = crtc->pipe;
+	enum pipe pipe = crtc->pipe;
 	i915_reg_t reg;
 	u32 temp, i, retry;
 
@@ -4698,7 +4811,7 @@ static void ivb_manual_fdi_link_train(struct intel_crtc *crtc,
 {
 	struct drm_device *dev = crtc->base.dev;
 	struct drm_i915_private *dev_priv = to_i915(dev);
-	int pipe = crtc->pipe;
+	enum pipe pipe = crtc->pipe;
 	i915_reg_t reg;
 	u32 temp, i, j;
 
@@ -4816,7 +4929,7 @@ static void ironlake_fdi_pll_enable(const struct intel_crtc_state *crtc_state)
 {
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc_state->base.crtc);
 	struct drm_i915_private *dev_priv = to_i915(intel_crtc->base.dev);
-	int pipe = intel_crtc->pipe;
+	enum pipe pipe = intel_crtc->pipe;
 	i915_reg_t reg;
 	u32 temp;
 
@@ -4853,7 +4966,7 @@ static void ironlake_fdi_pll_disable(struct intel_crtc *intel_crtc)
 {
 	struct drm_device *dev = intel_crtc->base.dev;
 	struct drm_i915_private *dev_priv = to_i915(dev);
-	int pipe = intel_crtc->pipe;
+	enum pipe pipe = intel_crtc->pipe;
 	i915_reg_t reg;
 	u32 temp;
 
@@ -4884,7 +4997,7 @@ static void ironlake_fdi_disable(struct drm_crtc *crtc)
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = to_i915(dev);
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	int pipe = intel_crtc->pipe;
+	enum pipe pipe = intel_crtc->pipe;
 	i915_reg_t reg;
 	u32 temp;
 
@@ -5199,7 +5312,7 @@ static void ironlake_pch_enable(const struct intel_atomic_state *state,
 	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
 	struct drm_device *dev = crtc->base.dev;
 	struct drm_i915_private *dev_priv = to_i915(dev);
-	int pipe = crtc->pipe;
+	enum pipe pipe = crtc->pipe;
 	u32 temp;
 
 	assert_pch_transcoder_disabled(dev_priv, pipe);
@@ -5294,7 +5407,7 @@ static void lpt_pch_enable(const struct intel_atomic_state *state,
 	lpt_enable_pch_transcoder(dev_priv, cpu_transcoder);
 }
 
-static void cpt_verify_modeset(struct drm_device *dev, int pipe)
+static void cpt_verify_modeset(struct drm_device *dev, enum pipe pipe)
 {
 	struct drm_i915_private *dev_priv = to_i915(dev);
 	i915_reg_t dslreg = PIPEDSL(pipe);
@@ -5446,7 +5559,7 @@ skl_update_scaler(struct intel_crtc_state *crtc_state, bool force_detach,
 		return 0;
 	}
 
-	if (format && is_planar_yuv_format(format->format) &&
+	if (format && drm_format_info_is_yuv_semiplanar(format) &&
 	    (src_h < SKL_MIN_YUV_420_SRC_H || src_w < SKL_MIN_YUV_420_SRC_W)) {
 		DRM_DEBUG_KMS("Planar YUV: src dimensions not met\n");
 		return -EINVAL;
@@ -5523,7 +5636,7 @@ static int skl_update_scaler_plane(struct intel_crtc_state *crtc_state,
 
 	/* Pre-gen11 and SDR planes always need a scaler for planar formats. */
 	if (!icl_is_hdr_plane(dev_priv, intel_plane->id) &&
-	    fb && is_planar_yuv_format(fb->format->format))
+	    fb && drm_format_info_is_yuv_semiplanar(fb->format))
 		need_scaler = true;
 
 	ret = skl_update_scaler(crtc_state, force_detach,
@@ -5555,10 +5668,6 @@ static int skl_update_scaler_plane(struct intel_crtc_state *crtc_state,
 	case DRM_FORMAT_ARGB8888:
 	case DRM_FORMAT_XRGB2101010:
 	case DRM_FORMAT_XBGR2101010:
-	case DRM_FORMAT_XBGR16161616F:
-	case DRM_FORMAT_ABGR16161616F:
-	case DRM_FORMAT_XRGB16161616F:
-	case DRM_FORMAT_ARGB16161616F:
 	case DRM_FORMAT_YUYV:
 	case DRM_FORMAT_YVYU:
 	case DRM_FORMAT_UYVY:
@@ -5574,6 +5683,13 @@ static int skl_update_scaler_plane(struct intel_crtc_state *crtc_state,
 	case DRM_FORMAT_XVYU12_16161616:
 	case DRM_FORMAT_XVYU16161616:
 		break;
+	case DRM_FORMAT_XBGR16161616F:
+	case DRM_FORMAT_ABGR16161616F:
+	case DRM_FORMAT_XRGB16161616F:
+	case DRM_FORMAT_ARGB16161616F:
+		if (INTEL_GEN(dev_priv) >= 11)
+			break;
+		/* fall through */
 	default:
 		DRM_DEBUG_KMS("[PLANE:%d:%s] FB:%d unsupported scaling format 0x%x\n",
 			      intel_plane->base.base.id, intel_plane->base.name,
@@ -5633,7 +5749,7 @@ static void ironlake_pfit_enable(const struct intel_crtc_state *crtc_state)
 {
 	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
-	int pipe = crtc->pipe;
+	enum pipe pipe = crtc->pipe;
 
 	if (crtc_state->pch_pfit.enabled) {
 		/* Force use of hard-coded filter coefficients
@@ -5715,13 +5831,8 @@ void hsw_disable_ips(const struct intel_crtc_state *crtc_state)
 
 static void intel_crtc_dpms_overlay_disable(struct intel_crtc *intel_crtc)
 {
-	if (intel_crtc->overlay) {
-		struct drm_device *dev = intel_crtc->base.dev;
-
-		mutex_lock(&dev->struct_mutex);
+	if (intel_crtc->overlay)
 		(void) intel_overlay_switch_off(intel_crtc->overlay);
-		mutex_unlock(&dev->struct_mutex);
-	}
 
 	/* Let userspace switch the overlay on again. In most cases userspace
 	 * has to recompute where to put it anyway.
@@ -5746,7 +5857,7 @@ intel_post_enable_primary(struct drm_crtc *crtc,
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = to_i915(dev);
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	int pipe = intel_crtc->pipe;
+	enum pipe pipe = intel_crtc->pipe;
 
 	/*
 	 * Gen2 reports pipe underruns whenever all planes are disabled.
@@ -5770,7 +5881,7 @@ intel_pre_disable_primary_noatomic(struct drm_crtc *crtc)
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = to_i915(dev);
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	int pipe = intel_crtc->pipe;
+	enum pipe pipe = intel_crtc->pipe;
 
 	/*
 	 * Gen2 reports pipe underruns whenever all planes are disabled.
@@ -6293,7 +6404,7 @@ static void ironlake_crtc_enable(struct intel_crtc_state *pipe_config,
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = to_i915(dev);
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	int pipe = intel_crtc->pipe;
+	enum pipe pipe = intel_crtc->pipe;
 
 	if (WARN_ON(intel_crtc->active))
 		return;
@@ -6426,7 +6537,7 @@ static void haswell_crtc_enable(struct intel_crtc_state *pipe_config,
 	struct drm_crtc *crtc = pipe_config->base.crtc;
 	struct drm_i915_private *dev_priv = to_i915(crtc->dev);
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	int pipe = intel_crtc->pipe, hsw_workaround_pipe;
+	enum pipe pipe = intel_crtc->pipe, hsw_workaround_pipe;
 	enum transcoder cpu_transcoder = pipe_config->cpu_transcoder;
 	bool psl_clkgate_wa;
 
@@ -6446,6 +6557,9 @@ static void haswell_crtc_enable(struct intel_crtc_state *pipe_config,
 	if (!transcoder_is_dsi(cpu_transcoder))
 		intel_set_pipe_timings(pipe_config);
 
+	if (INTEL_GEN(dev_priv) >= 11)
+		icl_enable_trans_port_sync(pipe_config);
+
 	intel_set_pipe_src_size(pipe_config);
 
 	if (cpu_transcoder != TRANSCODER_EDP &&
@@ -6491,7 +6605,6 @@ static void haswell_crtc_enable(struct intel_crtc_state *pipe_config,
 	if (INTEL_GEN(dev_priv) >= 11)
 		icl_set_pipe_chicken(intel_crtc);
 
-	intel_ddi_set_pipe_settings(pipe_config);
 	if (!transcoder_is_dsi(cpu_transcoder))
 		intel_ddi_enable_transcoder_func(pipe_config);
 
@@ -6552,7 +6665,7 @@ static void ironlake_crtc_disable(struct intel_crtc_state *old_crtc_state,
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = to_i915(dev);
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	int pipe = intel_crtc->pipe;
+	enum pipe pipe = intel_crtc->pipe;
 
 	/*
 	 * Sometimes spurious CPU pipe underruns happen when the
@@ -6624,6 +6737,9 @@ static void haswell_crtc_disable(struct intel_crtc_state *old_crtc_state,
 	if (intel_crtc_has_type(old_crtc_state, INTEL_OUTPUT_DP_MST))
 		intel_ddi_set_vc_payload_alloc(old_crtc_state, false);
 
+	if (INTEL_GEN(dev_priv) >= 11)
+		icl_disable_transcoder_port_sync(old_crtc_state);
+
 	if (!transcoder_is_dsi(cpu_transcoder))
 		intel_ddi_disable_transcoder_func(old_crtc_state);
 
@@ -6721,6 +6837,8 @@ enum intel_display_power_domain intel_port_to_power_domain(enum port port)
 		return POWER_DOMAIN_PORT_DDI_E_LANES;
 	case PORT_F:
 		return POWER_DOMAIN_PORT_DDI_F_LANES;
+	case PORT_G:
+		return POWER_DOMAIN_PORT_DDI_G_LANES;
 	default:
 		MISSING_CASE(port);
 		return POWER_DOMAIN_PORT_OTHER;
@@ -6737,16 +6855,18 @@ intel_aux_power_domain(struct intel_digital_port *dig_port)
 	    dig_port->tc_mode == TC_PORT_TBT_ALT) {
 		switch (dig_port->aux_ch) {
 		case AUX_CH_C:
-			return POWER_DOMAIN_AUX_TBT1;
+			return POWER_DOMAIN_AUX_C_TBT;
 		case AUX_CH_D:
-			return POWER_DOMAIN_AUX_TBT2;
+			return POWER_DOMAIN_AUX_D_TBT;
 		case AUX_CH_E:
-			return POWER_DOMAIN_AUX_TBT3;
+			return POWER_DOMAIN_AUX_E_TBT;
 		case AUX_CH_F:
-			return POWER_DOMAIN_AUX_TBT4;
+			return POWER_DOMAIN_AUX_F_TBT;
+		case AUX_CH_G:
+			return POWER_DOMAIN_AUX_G_TBT;
 		default:
 			MISSING_CASE(dig_port->aux_ch);
-			return POWER_DOMAIN_AUX_TBT1;
+			return POWER_DOMAIN_AUX_C_TBT;
 		}
 	}
 
@@ -6763,6 +6883,8 @@ intel_aux_power_domain(struct intel_digital_port *dig_port)
 		return POWER_DOMAIN_AUX_E;
 	case AUX_CH_F:
 		return POWER_DOMAIN_AUX_F;
+	case AUX_CH_G:
+		return POWER_DOMAIN_AUX_G;
 	default:
 		MISSING_CASE(dig_port->aux_ch);
 		return POWER_DOMAIN_AUX_A;
@@ -6839,7 +6961,7 @@ static void valleyview_crtc_enable(struct intel_crtc_state *pipe_config,
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = to_i915(dev);
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	int pipe = intel_crtc->pipe;
+	enum pipe pipe = intel_crtc->pipe;
 
 	if (WARN_ON(intel_crtc->active))
 		return;
@@ -6971,7 +7093,7 @@ static void i9xx_crtc_disable(struct intel_crtc_state *old_crtc_state,
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = to_i915(dev);
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	int pipe = intel_crtc->pipe;
+	enum pipe pipe = intel_crtc->pipe;
 
 	/*
 	 * On gen2 planes are double buffered but the pipe isn't, so we must
@@ -7080,7 +7202,7 @@ static void intel_crtc_disable_noatomic(struct drm_crtc *crtc,
 		intel_display_power_put_unchecked(dev_priv, domain);
 	intel_crtc->enabled_power_domains = 0;
 
-	dev_priv->active_crtcs &= ~(1 << intel_crtc->pipe);
+	dev_priv->active_pipes &= ~BIT(intel_crtc->pipe);
 	dev_priv->min_cdclk[intel_crtc->pipe] = 0;
 	dev_priv->min_voltage_level[intel_crtc->pipe] = 0;
 
@@ -7188,7 +7310,7 @@ static int ironlake_check_fdi_lanes(struct drm_device *dev, enum pipe pipe,
 		}
 	}
 
-	if (INTEL_INFO(dev_priv)->num_pipes == 2)
+	if (INTEL_NUM_PIPES(dev_priv) == 2)
 		return 0;
 
 	/* Ivybridge 3 pipe is really complicated */
@@ -7261,7 +7383,7 @@ retry:
 	pipe_config->fdi_lanes = lane;
 
 	intel_link_compute_m_n(pipe_config->pipe_bpp, lane, fdi_dotclock,
-			       link_bw, &pipe_config->fdi_m_n, false);
+			       link_bw, &pipe_config->fdi_m_n, false, false);
 
 	ret = ironlake_check_fdi_lanes(dev, intel_crtc->pipe, pipe_config);
 	if (ret == -EDEADLK)
@@ -7508,11 +7630,15 @@ void
 intel_link_compute_m_n(u16 bits_per_pixel, int nlanes,
 		       int pixel_clock, int link_clock,
 		       struct intel_link_m_n *m_n,
-		       bool constant_n)
+		       bool constant_n, bool fec_enable)
 {
-	m_n->tu = 64;
+	u32 data_clock = bits_per_pixel * pixel_clock;
+
+	if (fec_enable)
+		data_clock = intel_dp_mode_to_fec_clock(data_clock);
 
-	compute_m_n(bits_per_pixel * pixel_clock,
+	m_n->tu = 64;
+	compute_m_n(data_clock,
 		    link_clock * nlanes * 8,
 		    &m_n->gmch_m, &m_n->gmch_n,
 		    constant_n);
@@ -7522,6 +7648,27 @@ intel_link_compute_m_n(u16 bits_per_pixel, int nlanes,
 		    constant_n);
 }
 
+static void intel_panel_sanitize_ssc(struct drm_i915_private *dev_priv)
+{
+	/*
+	 * There may be no VBT; and if the BIOS enabled SSC we can
+	 * just keep using it to avoid unnecessary flicker.  Whereas if the
+	 * BIOS isn't using it, don't assume it will work even if the VBT
+	 * indicates as much.
+	 */
+	if (HAS_PCH_IBX(dev_priv) || HAS_PCH_CPT(dev_priv)) {
+		bool bios_lvds_use_ssc = I915_READ(PCH_DREF_CONTROL) &
+			DREF_SSC1_ENABLE;
+
+		if (dev_priv->vbt.lvds_use_ssc != bios_lvds_use_ssc) {
+			DRM_DEBUG_KMS("SSC %s by BIOS, overriding VBT which says %s\n",
+				      enableddisabled(bios_lvds_use_ssc),
+				      enableddisabled(dev_priv->vbt.lvds_use_ssc));
+			dev_priv->vbt.lvds_use_ssc = bios_lvds_use_ssc;
+		}
+	}
+}
+
 static inline bool intel_panel_use_ssc(struct drm_i915_private *dev_priv)
 {
 	if (i915_modparams.panel_use_ssc >= 0)
@@ -8173,6 +8320,21 @@ static void intel_set_pipe_src_size(const struct intel_crtc_state *crtc_state)
 		   (crtc_state->pipe_src_h - 1));
 }
 
+static bool intel_pipe_is_interlaced(const struct intel_crtc_state *crtc_state)
+{
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	enum transcoder cpu_transcoder = crtc_state->cpu_transcoder;
+
+	if (IS_GEN(dev_priv, 2))
+		return false;
+
+	if (INTEL_GEN(dev_priv) >= 9 ||
+	    IS_BROADWELL(dev_priv) || IS_HASWELL(dev_priv))
+		return I915_READ(PIPECONF(cpu_transcoder)) & PIPECONF_INTERLACE_MASK_HSW;
+	else
+		return I915_READ(PIPECONF(cpu_transcoder)) & PIPECONF_INTERLACE_MASK;
+}
+
 static void intel_get_pipe_timings(struct intel_crtc *crtc,
 				   struct intel_crtc_state *pipe_config)
 {
@@ -8211,7 +8373,7 @@ static void intel_get_pipe_timings(struct intel_crtc *crtc,
 	pipe_config->base.adjusted_mode.crtc_vsync_start = (tmp & 0xffff) + 1;
 	pipe_config->base.adjusted_mode.crtc_vsync_end = ((tmp >> 16) & 0xffff) + 1;
 
-	if (I915_READ(PIPECONF(cpu_transcoder)) & PIPECONF_INTERLACE_MASK) {
+	if (intel_pipe_is_interlaced(pipe_config)) {
 		pipe_config->base.adjusted_mode.flags |= DRM_MODE_FLAG_INTERLACE;
 		pipe_config->base.adjusted_mode.crtc_vtotal += 1;
 		pipe_config->base.adjusted_mode.crtc_vblank_end += 1;
@@ -8543,7 +8705,7 @@ static void vlv_crtc_clock_get(struct intel_crtc *crtc,
 {
 	struct drm_device *dev = crtc->base.dev;
 	struct drm_i915_private *dev_priv = to_i915(dev);
-	int pipe = pipe_config->cpu_transcoder;
+	enum pipe pipe = crtc->pipe;
 	struct dpll clock;
 	u32 mdiv;
 	int refclk = 100000;
@@ -8653,7 +8815,7 @@ static void chv_crtc_clock_get(struct intel_crtc *crtc,
 {
 	struct drm_device *dev = crtc->base.dev;
 	struct drm_i915_private *dev_priv = to_i915(dev);
-	int pipe = pipe_config->cpu_transcoder;
+	enum pipe pipe = crtc->pipe;
 	enum dpio_channel port = vlv_pipe_to_channel(pipe);
 	struct dpll clock;
 	u32 cmn_dw13, pll_dw0, pll_dw1, pll_dw2, pll_dw3;
@@ -8682,47 +8844,24 @@ static void chv_crtc_clock_get(struct intel_crtc *crtc,
 	pipe_config->port_clock = chv_calc_dpll_params(refclk, &clock);
 }
 
-static void intel_get_crtc_ycbcr_config(struct intel_crtc *crtc,
-					struct intel_crtc_state *pipe_config)
+static enum intel_output_format
+bdw_get_pipemisc_output_format(struct intel_crtc *crtc)
 {
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
-	enum intel_output_format output = INTEL_OUTPUT_FORMAT_RGB;
-
-	pipe_config->lspcon_downsampling = false;
+	u32 tmp;
 
-	if (IS_BROADWELL(dev_priv) || INTEL_GEN(dev_priv) >= 9) {
-		u32 tmp = I915_READ(PIPEMISC(crtc->pipe));
+	tmp = I915_READ(PIPEMISC(crtc->pipe));
 
-		if (tmp & PIPEMISC_OUTPUT_COLORSPACE_YUV) {
-			bool ycbcr420_enabled = tmp & PIPEMISC_YUV420_ENABLE;
-			bool blend = tmp & PIPEMISC_YUV420_MODE_FULL_BLEND;
+	if (tmp & PIPEMISC_YUV420_ENABLE) {
+		/* We support 4:2:0 in full blend mode only */
+		WARN_ON((tmp & PIPEMISC_YUV420_MODE_FULL_BLEND) == 0);
 
-			if (ycbcr420_enabled) {
-				/* We support 4:2:0 in full blend mode only */
-				if (!blend)
-					output = INTEL_OUTPUT_FORMAT_INVALID;
-				else if (!(IS_GEMINILAKE(dev_priv) ||
-					   INTEL_GEN(dev_priv) >= 10))
-					output = INTEL_OUTPUT_FORMAT_INVALID;
-				else
-					output = INTEL_OUTPUT_FORMAT_YCBCR420;
-			} else {
-				/*
-				 * Currently there is no interface defined to
-				 * check user preference between RGB/YCBCR444
-				 * or YCBCR420. So the only possible case for
-				 * YCBCR444 usage is driving YCBCR420 output
-				 * with LSPCON, when pipe is configured for
-				 * YCBCR444 output and LSPCON takes care of
-				 * downsampling it.
-				 */
-				pipe_config->lspcon_downsampling = true;
-				output = INTEL_OUTPUT_FORMAT_YCBCR444;
-			}
-		}
+		return INTEL_OUTPUT_FORMAT_YCBCR420;
+	} else if (tmp & PIPEMISC_OUTPUT_COLORSPACE_YUV) {
+		return INTEL_OUTPUT_FORMAT_YCBCR444;
+	} else {
+		return INTEL_OUTPUT_FORMAT_RGB;
 	}
-
-	pipe_config->output_format = output;
 }
 
 static void i9xx_get_pipe_color_config(struct intel_crtc_state *crtc_state)
@@ -8760,6 +8899,7 @@ static bool i9xx_get_pipe_config(struct intel_crtc *crtc,
 	pipe_config->output_format = INTEL_OUTPUT_FORMAT_RGB;
 	pipe_config->cpu_transcoder = (enum transcoder) crtc->pipe;
 	pipe_config->shared_dpll = NULL;
+	pipe_config->master_transcoder = INVALID_TRANSCODER;
 
 	ret = false;
 
@@ -9298,7 +9438,6 @@ static bool wrpll_uses_pch_ssc(struct drm_i915_private *dev_priv,
 static void lpt_init_pch_refclk(struct drm_i915_private *dev_priv)
 {
 	struct intel_encoder *encoder;
-	bool pch_ssc_in_use = false;
 	bool has_fdi = false;
 
 	for_each_intel_encoder(&dev_priv->drm, encoder) {
@@ -9326,22 +9465,24 @@ static void lpt_init_pch_refclk(struct drm_i915_private *dev_priv)
 	 * clock hierarchy. That would also allow us to do
 	 * clock bending finally.
 	 */
+	dev_priv->pch_ssc_use = 0;
+
 	if (spll_uses_pch_ssc(dev_priv)) {
 		DRM_DEBUG_KMS("SPLL using PCH SSC\n");
-		pch_ssc_in_use = true;
+		dev_priv->pch_ssc_use |= BIT(DPLL_ID_SPLL);
 	}
 
 	if (wrpll_uses_pch_ssc(dev_priv, DPLL_ID_WRPLL1)) {
 		DRM_DEBUG_KMS("WRPLL1 using PCH SSC\n");
-		pch_ssc_in_use = true;
+		dev_priv->pch_ssc_use |= BIT(DPLL_ID_WRPLL1);
 	}
 
 	if (wrpll_uses_pch_ssc(dev_priv, DPLL_ID_WRPLL2)) {
 		DRM_DEBUG_KMS("WRPLL2 using PCH SSC\n");
-		pch_ssc_in_use = true;
+		dev_priv->pch_ssc_use |= BIT(DPLL_ID_WRPLL2);
 	}
 
-	if (pch_ssc_in_use)
+	if (dev_priv->pch_ssc_use)
 		return;
 
 	if (has_fdi) {
@@ -9398,9 +9539,19 @@ static void ironlake_set_pipeconf(const struct intel_crtc_state *crtc_state)
 	else
 		val |= PIPECONF_PROGRESSIVE;
 
+	/*
+	 * This would end up with an odd purple hue over
+	 * the entire display. Make sure we don't do it.
+	 */
+	WARN_ON(crtc_state->limited_color_range &&
+		crtc_state->output_format != INTEL_OUTPUT_FORMAT_RGB);
+
 	if (crtc_state->limited_color_range)
 		val |= PIPECONF_COLOR_RANGE_SELECT;
 
+	if (crtc_state->output_format != INTEL_OUTPUT_FORMAT_RGB)
+		val |= PIPECONF_OUTPUT_COLORSPACE_YUV709;
+
 	val |= PIPECONF_GAMMA_MODE(crtc_state->gamma_mode);
 
 	I915_WRITE(PIPECONF(pipe), val);
@@ -9422,6 +9573,10 @@ static void haswell_set_pipeconf(const struct intel_crtc_state *crtc_state)
 	else
 		val |= PIPECONF_PROGRESSIVE;
 
+	if (IS_HASWELL(dev_priv) &&
+	    crtc_state->output_format != INTEL_OUTPUT_FORMAT_RGB)
+		val |= PIPECONF_OUTPUT_COLORSPACE_YUV_HSW;
+
 	I915_WRITE(PIPECONF(cpu_transcoder), val);
 	POSTING_READ(PIPECONF(cpu_transcoder));
 }
@@ -9572,7 +9727,7 @@ static void ironlake_compute_dpll(struct intel_crtc *crtc,
 	 * clear if it''s a win or loss power wise. No point in doing
 	 * this on ILK at all since it has a fixed DPLL<->pipe mapping.
 	 */
-	if (INTEL_INFO(dev_priv)->num_pipes == 3 &&
+	if (INTEL_NUM_PIPES(dev_priv) == 3 &&
 	    intel_crtc_has_type(crtc_state, INTEL_OUTPUT_ANALOG))
 		dpll |= DPLL_SDVO_HIGH_SPEED;
 
@@ -9871,8 +10026,8 @@ skylake_get_initial_plane_config(struct intel_crtc *crtc,
 	offset = I915_READ(PLANE_OFFSET(pipe, plane_id));
 
 	val = I915_READ(PLANE_SIZE(pipe, plane_id));
-	fb->height = ((val >> 16) & 0xfff) + 1;
-	fb->width = ((val >> 0) & 0x1fff) + 1;
+	fb->height = ((val >> 16) & 0xffff) + 1;
+	fb->width = ((val >> 0) & 0xffff) + 1;
 
 	val = I915_READ(PLANE_STRIDE(pipe, plane_id));
 	stride_mult = skl_plane_stride_mult(fb, 0, DRM_MODE_ROTATE_0);
@@ -9933,9 +10088,9 @@ static bool ironlake_get_pipe_config(struct intel_crtc *crtc,
 	if (!wakeref)
 		return false;
 
-	pipe_config->output_format = INTEL_OUTPUT_FORMAT_RGB;
 	pipe_config->cpu_transcoder = (enum transcoder) crtc->pipe;
 	pipe_config->shared_dpll = NULL;
+	pipe_config->master_transcoder = INVALID_TRANSCODER;
 
 	ret = false;
 	tmp = I915_READ(PIPECONF(crtc->pipe));
@@ -9962,6 +10117,16 @@ static bool ironlake_get_pipe_config(struct intel_crtc *crtc,
 	if (tmp & PIPECONF_COLOR_RANGE_SELECT)
 		pipe_config->limited_color_range = true;
 
+	switch (tmp & PIPECONF_OUTPUT_COLORSPACE_MASK) {
+	case PIPECONF_OUTPUT_COLORSPACE_YUV601:
+	case PIPECONF_OUTPUT_COLORSPACE_YUV709:
+		pipe_config->output_format = INTEL_OUTPUT_FORMAT_YCBCR444;
+		break;
+	default:
+		pipe_config->output_format = INTEL_OUTPUT_FORMAT_RGB;
+		break;
+	}
+
 	pipe_config->gamma_mode = (tmp & PIPECONF_GAMMA_MODE_MASK_ILK) >>
 		PIPECONF_GAMMA_MODE_SHIFT;
 
@@ -10376,6 +10541,59 @@ static void haswell_get_ddi_port_state(struct intel_crtc *crtc,
 	}
 }
 
+static enum transcoder transcoder_master_readout(struct drm_i915_private *dev_priv,
+						 enum transcoder cpu_transcoder)
+{
+	u32 trans_port_sync, master_select;
+
+	trans_port_sync = I915_READ(TRANS_DDI_FUNC_CTL2(cpu_transcoder));
+
+	if ((trans_port_sync & PORT_SYNC_MODE_ENABLE) == 0)
+		return INVALID_TRANSCODER;
+
+	master_select = trans_port_sync &
+			PORT_SYNC_MODE_MASTER_SELECT_MASK;
+	if (master_select == 0)
+		return TRANSCODER_EDP;
+	else
+		return master_select - 1;
+}
+
+static void icelake_get_trans_port_sync_config(struct intel_crtc_state *crtc_state)
+{
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	u32 transcoders;
+	enum transcoder cpu_transcoder;
+
+	crtc_state->master_transcoder = transcoder_master_readout(dev_priv,
+								  crtc_state->cpu_transcoder);
+
+	transcoders = BIT(TRANSCODER_A) |
+		BIT(TRANSCODER_B) |
+		BIT(TRANSCODER_C) |
+		BIT(TRANSCODER_D);
+	for_each_cpu_transcoder_masked(dev_priv, cpu_transcoder, transcoders) {
+		enum intel_display_power_domain power_domain;
+		intel_wakeref_t trans_wakeref;
+
+		power_domain = POWER_DOMAIN_TRANSCODER(cpu_transcoder);
+		trans_wakeref = intel_display_power_get_if_enabled(dev_priv,
+								   power_domain);
+
+		if (!trans_wakeref)
+			continue;
+
+		if (transcoder_master_readout(dev_priv, cpu_transcoder) ==
+		    crtc_state->cpu_transcoder)
+			crtc_state->sync_mode_slaves_mask |= BIT(cpu_transcoder);
+
+		intel_display_power_put(dev_priv, power_domain, trans_wakeref);
+	}
+
+	WARN_ON(crtc_state->master_transcoder != INVALID_TRANSCODER &&
+		crtc_state->sync_mode_slaves_mask);
+}
+
 static bool haswell_get_pipe_config(struct intel_crtc *crtc,
 				    struct intel_crtc_state *pipe_config)
 {
@@ -10387,6 +10605,8 @@ static bool haswell_get_pipe_config(struct intel_crtc *crtc,
 
 	intel_crtc_init_scalers(crtc, pipe_config);
 
+	pipe_config->master_transcoder = INVALID_TRANSCODER;
+
 	power_domain = POWER_DOMAIN_PIPE(crtc->pipe);
 	wf = intel_display_power_get_if_enabled(dev_priv, power_domain);
 	if (!wf)
@@ -10417,7 +10637,30 @@ static bool haswell_get_pipe_config(struct intel_crtc *crtc,
 	}
 
 	intel_get_pipe_src_size(crtc, pipe_config);
-	intel_get_crtc_ycbcr_config(crtc, pipe_config);
+
+	if (IS_HASWELL(dev_priv)) {
+		u32 tmp = I915_READ(PIPECONF(pipe_config->cpu_transcoder));
+
+		if (tmp & PIPECONF_OUTPUT_COLORSPACE_YUV_HSW)
+			pipe_config->output_format = INTEL_OUTPUT_FORMAT_YCBCR444;
+		else
+			pipe_config->output_format = INTEL_OUTPUT_FORMAT_RGB;
+	} else {
+		pipe_config->output_format =
+			bdw_get_pipemisc_output_format(crtc);
+
+		/*
+		 * Currently there is no interface defined to
+		 * check user preference between RGB/YCBCR444
+		 * or YCBCR420. So the only possible case for
+		 * YCBCR444 usage is driving YCBCR420 output
+		 * with LSPCON, when pipe is configured for
+		 * YCBCR444 output and LSPCON takes care of
+		 * downsampling it.
+		 */
+		pipe_config->lspcon_downsampling =
+			pipe_config->output_format == INTEL_OUTPUT_FORMAT_YCBCR444;
+	}
 
 	pipe_config->gamma_mode = I915_READ(GAMMA_MODE(crtc->pipe));
 
@@ -10472,6 +10715,10 @@ static bool haswell_get_pipe_config(struct intel_crtc *crtc,
 		pipe_config->pixel_multiplier = 1;
 	}
 
+	if (INTEL_GEN(dev_priv) >= 11 &&
+	    !transcoder_is_dsi(pipe_config->cpu_transcoder))
+		icelake_get_trans_port_sync_config(pipe_config);
+
 out:
 	for_each_power_domain(power_domain, power_domain_mask)
 		intel_display_power_put(dev_priv,
@@ -10493,21 +10740,13 @@ static u32 intel_cursor_base(const struct intel_plane_state *plane_state)
 	else
 		base = intel_plane_ggtt_offset(plane_state);
 
-	base += plane_state->color_plane[0].offset;
-
-	/* ILK+ do this automagically */
-	if (HAS_GMCH(dev_priv) &&
-	    plane_state->base.rotation & DRM_MODE_ROTATE_180)
-		base += (plane_state->base.crtc_h *
-			 plane_state->base.crtc_w - 1) * fb->format->cpp[0];
-
-	return base;
+	return base + plane_state->color_plane[0].offset;
 }
 
 static u32 intel_cursor_position(const struct intel_plane_state *plane_state)
 {
-	int x = plane_state->base.crtc_x;
-	int y = plane_state->base.crtc_y;
+	int x = plane_state->base.dst.x1;
+	int y = plane_state->base.dst.y1;
 	u32 pos = 0;
 
 	if (x < 0) {
@@ -10529,8 +10768,8 @@ static bool intel_cursor_size_ok(const struct intel_plane_state *plane_state)
 {
 	const struct drm_mode_config *config =
 		&plane_state->base.plane->dev->mode_config;
-	int width = plane_state->base.crtc_w;
-	int height = plane_state->base.crtc_h;
+	int width = drm_rect_width(&plane_state->base.dst);
+	int height = drm_rect_height(&plane_state->base.dst);
 
 	return width > 0 && width <= config->cursor_width &&
 		height > 0 && height <= config->cursor_height;
@@ -10538,6 +10777,9 @@ static bool intel_cursor_size_ok(const struct intel_plane_state *plane_state)
 
 static int intel_cursor_check_surface(struct intel_plane_state *plane_state)
 {
+	struct drm_i915_private *dev_priv =
+		to_i915(plane_state->base.plane->dev);
+	unsigned int rotation = plane_state->base.rotation;
 	int src_x, src_y;
 	u32 offset;
 	int ret;
@@ -10549,8 +10791,8 @@ static int intel_cursor_check_surface(struct intel_plane_state *plane_state)
 	if (!plane_state->base.visible)
 		return 0;
 
-	src_x = plane_state->base.src_x >> 16;
-	src_y = plane_state->base.src_y >> 16;
+	src_x = plane_state->base.src.x1 >> 16;
+	src_y = plane_state->base.src.y1 >> 16;
 
 	intel_add_fb_offsets(&src_x, &src_y, plane_state, 0);
 	offset = intel_plane_compute_aligned_offset(&src_x, &src_y,
@@ -10561,7 +10803,25 @@ static int intel_cursor_check_surface(struct intel_plane_state *plane_state)
 		return -EINVAL;
 	}
 
+	/*
+	 * Put the final coordinates back so that the src
+	 * coordinate checks will see the right values.
+	 */
+	drm_rect_translate_to(&plane_state->base.src,
+			      src_x << 16, src_y << 16);
+
+	/* ILK+ do this automagically in hardware */
+	if (HAS_GMCH(dev_priv) && rotation & DRM_MODE_ROTATE_180) {
+		const struct drm_framebuffer *fb = plane_state->base.fb;
+		int src_w = drm_rect_width(&plane_state->base.src) >> 16;
+		int src_h = drm_rect_height(&plane_state->base.src) >> 16;
+
+		offset += (src_h * src_w - 1) * fb->format->cpp[0];
+	}
+
 	plane_state->color_plane[0].offset = offset;
+	plane_state->color_plane[0].x = src_x;
+	plane_state->color_plane[0].y = src_y;
 
 	return 0;
 }
@@ -10585,6 +10845,10 @@ static int intel_check_cursor(struct intel_crtc_state *crtc_state,
 	if (ret)
 		return ret;
 
+	/* Use the unclipped src/dst rectangles, which we program to hw */
+	plane_state->base.src = drm_plane_state_src(&plane_state->base);
+	plane_state->base.dst = drm_plane_state_dest(&plane_state->base);
+
 	ret = intel_cursor_check_surface(plane_state);
 	if (ret)
 		return ret;
@@ -10627,7 +10891,7 @@ static u32 i845_cursor_ctl(const struct intel_crtc_state *crtc_state,
 
 static bool i845_cursor_size_ok(const struct intel_plane_state *plane_state)
 {
-	int width = plane_state->base.crtc_w;
+	int width = drm_rect_width(&plane_state->base.dst);
 
 	/*
 	 * 845g/865g are only limited by the width of their cursors,
@@ -10653,8 +10917,8 @@ static int i845_check_cursor(struct intel_crtc_state *crtc_state,
 	/* Check for which cursor types we support */
 	if (!i845_cursor_size_ok(plane_state)) {
 		DRM_DEBUG("Cursor dimension %dx%d not supported\n",
-			  plane_state->base.crtc_w,
-			  plane_state->base.crtc_h);
+			  drm_rect_width(&plane_state->base.dst),
+			  drm_rect_height(&plane_state->base.dst));
 		return -EINVAL;
 	}
 
@@ -10687,8 +10951,8 @@ static void i845_update_cursor(struct intel_plane *plane,
 	unsigned long irqflags;
 
 	if (plane_state && plane_state->base.visible) {
-		unsigned int width = plane_state->base.crtc_w;
-		unsigned int height = plane_state->base.crtc_h;
+		unsigned int width = drm_rect_width(&plane_state->base.dst);
+		unsigned int height = drm_rect_height(&plane_state->base.dst);
 
 		cntl = plane_state->ctl |
 			i845_cursor_ctl_crtc(crtc_state);
@@ -10790,7 +11054,7 @@ static u32 i9xx_cursor_ctl(const struct intel_crtc_state *crtc_state,
 	if (IS_GEN(dev_priv, 6) || IS_IVYBRIDGE(dev_priv))
 		cntl |= MCURSOR_TRICKLE_FEED_DISABLE;
 
-	switch (plane_state->base.crtc_w) {
+	switch (drm_rect_width(&plane_state->base.dst)) {
 	case 64:
 		cntl |= MCURSOR_MODE_64_ARGB_AX;
 		break;
@@ -10801,7 +11065,7 @@ static u32 i9xx_cursor_ctl(const struct intel_crtc_state *crtc_state,
 		cntl |= MCURSOR_MODE_256_ARGB_AX;
 		break;
 	default:
-		MISSING_CASE(plane_state->base.crtc_w);
+		MISSING_CASE(drm_rect_width(&plane_state->base.dst));
 		return 0;
 	}
 
@@ -10815,8 +11079,8 @@ static bool i9xx_cursor_size_ok(const struct intel_plane_state *plane_state)
 {
 	struct drm_i915_private *dev_priv =
 		to_i915(plane_state->base.plane->dev);
-	int width = plane_state->base.crtc_w;
-	int height = plane_state->base.crtc_h;
+	int width = drm_rect_width(&plane_state->base.dst);
+	int height = drm_rect_height(&plane_state->base.dst);
 
 	if (!intel_cursor_size_ok(plane_state))
 		return false;
@@ -10869,17 +11133,19 @@ static int i9xx_check_cursor(struct intel_crtc_state *crtc_state,
 	/* Check for which cursor types we support */
 	if (!i9xx_cursor_size_ok(plane_state)) {
 		DRM_DEBUG("Cursor dimension %dx%d not supported\n",
-			  plane_state->base.crtc_w,
-			  plane_state->base.crtc_h);
+			  drm_rect_width(&plane_state->base.dst),
+			  drm_rect_height(&plane_state->base.dst));
 		return -EINVAL;
 	}
 
 	WARN_ON(plane_state->base.visible &&
 		plane_state->color_plane[0].stride != fb->pitches[0]);
 
-	if (fb->pitches[0] != plane_state->base.crtc_w * fb->format->cpp[0]) {
+	if (fb->pitches[0] !=
+	    drm_rect_width(&plane_state->base.dst) * fb->format->cpp[0]) {
 		DRM_DEBUG_KMS("Invalid cursor stride (%u) (cursor width %d)\n",
-			      fb->pitches[0], plane_state->base.crtc_w);
+			      fb->pitches[0],
+			      drm_rect_width(&plane_state->base.dst));
 		return -EINVAL;
 	}
 
@@ -10894,7 +11160,7 @@ static int i9xx_check_cursor(struct intel_crtc_state *crtc_state,
 	 * Refuse the put the cursor into that compromised position.
 	 */
 	if (IS_CHERRYVIEW(dev_priv) && pipe == PIPE_C &&
-	    plane_state->base.visible && plane_state->base.crtc_x < 0) {
+	    plane_state->base.visible && plane_state->base.dst.x1 < 0) {
 		DRM_DEBUG_KMS("CHV cursor C not allowed to straddle the left screen edge\n");
 		return -EINVAL;
 	}
@@ -10914,11 +11180,14 @@ static void i9xx_update_cursor(struct intel_plane *plane,
 	unsigned long irqflags;
 
 	if (plane_state && plane_state->base.visible) {
+		unsigned width = drm_rect_width(&plane_state->base.dst);
+		unsigned height = drm_rect_height(&plane_state->base.dst);
+
 		cntl = plane_state->ctl |
 			i9xx_cursor_ctl_crtc(crtc_state);
 
-		if (plane_state->base.crtc_h != plane_state->base.crtc_w)
-			fbc_ctl = CUR_FBC_CTL_EN | (plane_state->base.crtc_h - 1);
+		if (width != height)
+			fbc_ctl = CUR_FBC_CTL_EN | (height - 1);
 
 		base = intel_cursor_base(plane_state);
 		pos = intel_cursor_position(plane_state);
@@ -11063,7 +11332,6 @@ static int intel_modeset_disable_planes(struct drm_atomic_state *state,
 }
 
 int intel_get_load_detect_pipe(struct drm_connector *connector,
-			       const struct drm_display_mode *mode,
 			       struct intel_load_detect_pipe *old,
 			       struct drm_modeset_acquire_ctx *ctx)
 {
@@ -11170,10 +11438,8 @@ found:
 
 	crtc_state->base.active = crtc_state->base.enable = true;
 
-	if (!mode)
-		mode = &load_detect_mode;
-
-	ret = drm_atomic_set_mode_for_crtc(&crtc_state->base, mode);
+	ret = drm_atomic_set_mode_for_crtc(&crtc_state->base,
+					   &load_detect_mode);
 	if (ret)
 		goto fail;
 
@@ -11265,7 +11531,7 @@ static void i9xx_crtc_clock_get(struct intel_crtc *crtc,
 {
 	struct drm_device *dev = crtc->base.dev;
 	struct drm_i915_private *dev_priv = to_i915(dev);
-	int pipe = pipe_config->cpu_transcoder;
+	enum pipe pipe = crtc->pipe;
 	u32 dpll = pipe_config->dpll_hw_state.dpll;
 	u32 fp;
 	struct dpll clock;
@@ -11489,7 +11755,6 @@ int intel_plane_atomic_calc_changes(const struct intel_crtc_state *old_crtc_stat
 	bool was_crtc_enabled = old_crtc_state->base.active;
 	bool is_crtc_enabled = crtc_state->base.active;
 	bool turn_off, turn_on, visible, was_visible;
-	struct drm_framebuffer *fb = plane_state->base.fb;
 	int ret;
 
 	if (INTEL_GEN(dev_priv) >= 9 && plane->id != PLANE_CURSOR) {
@@ -11518,24 +11783,18 @@ int intel_plane_atomic_calc_changes(const struct intel_crtc_state *old_crtc_stat
 		plane_state->base.visible = visible = false;
 		crtc_state->active_planes &= ~BIT(plane->id);
 		crtc_state->data_rate[plane->id] = 0;
+		crtc_state->min_cdclk[plane->id] = 0;
 	}
 
 	if (!was_visible && !visible)
 		return 0;
 
-	if (fb != old_plane_state->base.fb)
-		crtc_state->fb_changed = true;
-
 	turn_off = was_visible && (!visible || mode_changed);
 	turn_on = visible && (!was_visible || mode_changed);
 
-	DRM_DEBUG_ATOMIC("[CRTC:%d:%s] has [PLANE:%d:%s] with fb %i\n",
+	DRM_DEBUG_ATOMIC("[CRTC:%d:%s] with [PLANE:%d:%s] visible %i -> %i, off %i, on %i, ms %i\n",
 			 crtc->base.base.id, crtc->base.name,
 			 plane->base.base.id, plane->base.name,
-			 fb ? fb->base.id : -1);
-
-	DRM_DEBUG_ATOMIC("[PLANE:%d:%s] visible %i -> %i, off %i, on %i, ms %i\n",
-			 plane->base.base.id, plane->base.name,
 			 was_visible, visible,
 			 turn_off, turn_on, mode_changed);
 
@@ -11644,7 +11903,7 @@ static int icl_add_linked_planes(struct intel_atomic_state *state)
 	int i;
 
 	for_each_new_intel_plane_in_state(state, plane, plane_state, i) {
-		linked = plane_state->linked_plane;
+		linked = plane_state->planar_linked_plane;
 
 		if (!linked)
 			continue;
@@ -11653,8 +11912,8 @@ static int icl_add_linked_planes(struct intel_atomic_state *state)
 		if (IS_ERR(linked_plane_state))
 			return PTR_ERR(linked_plane_state);
 
-		WARN_ON(linked_plane_state->linked_plane != plane);
-		WARN_ON(linked_plane_state->slave == plane_state->slave);
+		WARN_ON(linked_plane_state->planar_linked_plane != plane);
+		WARN_ON(linked_plane_state->planar_slave == plane_state->planar_slave);
 	}
 
 	return 0;
@@ -11677,16 +11936,16 @@ static int icl_check_nv12_planes(struct intel_crtc_state *crtc_state)
 	 * in the crtc_state->active_planes mask.
 	 */
 	for_each_new_intel_plane_in_state(state, plane, plane_state, i) {
-		if (plane->pipe != crtc->pipe || !plane_state->linked_plane)
+		if (plane->pipe != crtc->pipe || !plane_state->planar_linked_plane)
 			continue;
 
-		plane_state->linked_plane = NULL;
-		if (plane_state->slave && !plane_state->base.visible) {
+		plane_state->planar_linked_plane = NULL;
+		if (plane_state->planar_slave && !plane_state->base.visible) {
 			crtc_state->active_planes &= ~BIT(plane->id);
 			crtc_state->update_planes |= BIT(plane->id);
 		}
 
-		plane_state->slave = false;
+		plane_state->planar_slave = false;
 	}
 
 	if (!crtc_state->nv12_planes)
@@ -11720,10 +11979,10 @@ static int icl_check_nv12_planes(struct intel_crtc_state *crtc_state)
 			return -EINVAL;
 		}
 
-		plane_state->linked_plane = linked;
+		plane_state->planar_linked_plane = linked;
 
-		linked_state->slave = true;
-		linked_state->linked_plane = plane;
+		linked_state->planar_slave = true;
+		linked_state->planar_linked_plane = plane;
 		crtc_state->active_planes |= BIT(linked->id);
 		crtc_state->update_planes |= BIT(linked->id);
 		DRM_DEBUG_KMS("Using %s as Y plane for %s\n", linked->base.name, plane->base.name);
@@ -11743,25 +12002,108 @@ static bool c8_planes_changed(const struct intel_crtc_state *new_crtc_state)
 	return !old_crtc_state->c8_planes != !new_crtc_state->c8_planes;
 }
 
-static int intel_crtc_atomic_check(struct drm_crtc *crtc,
-				   struct drm_crtc_state *crtc_state)
+static int icl_add_sync_mode_crtcs(struct intel_crtc_state *crtc_state)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc->dev);
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	struct intel_crtc_state *pipe_config =
-		to_intel_crtc_state(crtc_state);
+	struct drm_crtc *crtc = crtc_state->base.crtc;
+	struct intel_atomic_state *state = to_intel_atomic_state(crtc_state->base.state);
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	struct drm_connector *master_connector, *connector;
+	struct drm_connector_state *connector_state;
+	struct drm_connector_list_iter conn_iter;
+	struct drm_crtc *master_crtc = NULL;
+	struct drm_crtc_state *master_crtc_state;
+	struct intel_crtc_state *master_pipe_config;
+	int i, tile_group_id;
+
+	if (INTEL_GEN(dev_priv) < 11)
+		return 0;
+
+	/*
+	 * In case of tiled displays there could be one or more slaves but there is
+	 * only one master. Lets make the CRTC used by the connector corresponding
+	 * to the last horizonal and last vertical tile a master/genlock CRTC.
+	 * All the other CRTCs corresponding to other tiles of the same Tile group
+	 * are the slave CRTCs and hold a pointer to their genlock CRTC.
+	 */
+	for_each_new_connector_in_state(&state->base, connector, connector_state, i) {
+		if (connector_state->crtc != crtc)
+			continue;
+		if (!connector->has_tile)
+			continue;
+		if (crtc_state->base.mode.hdisplay != connector->tile_h_size ||
+		    crtc_state->base.mode.vdisplay != connector->tile_v_size)
+			return 0;
+		if (connector->tile_h_loc == connector->num_h_tile - 1 &&
+		    connector->tile_v_loc == connector->num_v_tile - 1)
+			continue;
+		crtc_state->sync_mode_slaves_mask = 0;
+		tile_group_id = connector->tile_group->id;
+		drm_connector_list_iter_begin(&dev_priv->drm, &conn_iter);
+		drm_for_each_connector_iter(master_connector, &conn_iter) {
+			struct drm_connector_state *master_conn_state = NULL;
+
+			if (!master_connector->has_tile)
+				continue;
+			if (master_connector->tile_h_loc != master_connector->num_h_tile - 1 ||
+			    master_connector->tile_v_loc != master_connector->num_v_tile - 1)
+				continue;
+			if (master_connector->tile_group->id != tile_group_id)
+				continue;
+
+			master_conn_state = drm_atomic_get_connector_state(&state->base,
+									   master_connector);
+			if (IS_ERR(master_conn_state)) {
+				drm_connector_list_iter_end(&conn_iter);
+				return PTR_ERR(master_conn_state);
+			}
+			if (master_conn_state->crtc) {
+				master_crtc = master_conn_state->crtc;
+				break;
+			}
+		}
+		drm_connector_list_iter_end(&conn_iter);
+
+		if (!master_crtc) {
+			DRM_DEBUG_KMS("Could not find Master CRTC for Slave CRTC %d\n",
+				      connector_state->crtc->base.id);
+			return -EINVAL;
+		}
+
+		master_crtc_state = drm_atomic_get_crtc_state(&state->base,
+							      master_crtc);
+		if (IS_ERR(master_crtc_state))
+			return PTR_ERR(master_crtc_state);
+
+		master_pipe_config = to_intel_crtc_state(master_crtc_state);
+		crtc_state->master_transcoder = master_pipe_config->cpu_transcoder;
+		master_pipe_config->sync_mode_slaves_mask |=
+			BIT(crtc_state->cpu_transcoder);
+		DRM_DEBUG_KMS("Master Transcoder = %s added for Slave CRTC = %d, slave transcoder bitmask = %d\n",
+			      transcoder_name(crtc_state->master_transcoder),
+			      crtc_state->base.crtc->base.id,
+			      master_pipe_config->sync_mode_slaves_mask);
+	}
+
+	return 0;
+}
+
+static int intel_crtc_atomic_check(struct intel_atomic_state *state,
+				   struct intel_crtc *crtc)
+{
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	struct intel_crtc_state *crtc_state =
+		intel_atomic_get_new_crtc_state(state, crtc);
+	bool mode_changed = needs_modeset(crtc_state);
 	int ret;
-	bool mode_changed = needs_modeset(pipe_config);
 
 	if (INTEL_GEN(dev_priv) < 5 && !IS_G4X(dev_priv) &&
-	    mode_changed && !crtc_state->active)
-		pipe_config->update_wm_post = true;
+	    mode_changed && !crtc_state->base.active)
+		crtc_state->update_wm_post = true;
 
-	if (mode_changed && crtc_state->enable &&
+	if (mode_changed && crtc_state->base.enable &&
 	    dev_priv->display.crtc_compute_clock &&
-	    !WARN_ON(pipe_config->shared_dpll)) {
-		ret = dev_priv->display.crtc_compute_clock(intel_crtc,
-							   pipe_config);
+	    !WARN_ON(crtc_state->shared_dpll)) {
+		ret = dev_priv->display.crtc_compute_clock(crtc, crtc_state);
 		if (ret)
 			return ret;
 	}
@@ -11770,19 +12112,19 @@ static int intel_crtc_atomic_check(struct drm_crtc *crtc,
 	 * May need to update pipe gamma enable bits
 	 * when C8 planes are getting enabled/disabled.
 	 */
-	if (c8_planes_changed(pipe_config))
-		crtc_state->color_mgmt_changed = true;
+	if (c8_planes_changed(crtc_state))
+		crtc_state->base.color_mgmt_changed = true;
 
-	if (mode_changed || pipe_config->update_pipe ||
-	    crtc_state->color_mgmt_changed) {
-		ret = intel_color_check(pipe_config);
+	if (mode_changed || crtc_state->update_pipe ||
+	    crtc_state->base.color_mgmt_changed) {
+		ret = intel_color_check(crtc_state);
 		if (ret)
 			return ret;
 	}
 
 	ret = 0;
 	if (dev_priv->display.compute_pipe_wm) {
-		ret = dev_priv->display.compute_pipe_wm(pipe_config);
+		ret = dev_priv->display.compute_pipe_wm(crtc_state);
 		if (ret) {
 			DRM_DEBUG_KMS("Target pipe watermarks are invalid\n");
 			return ret;
@@ -11798,7 +12140,7 @@ static int intel_crtc_atomic_check(struct drm_crtc *crtc,
 		 * old state and the new state.  We can program these
 		 * immediately.
 		 */
-		ret = dev_priv->display.compute_intermediate_wm(pipe_config);
+		ret = dev_priv->display.compute_intermediate_wm(crtc_state);
 		if (ret) {
 			DRM_DEBUG_KMS("No valid intermediate pipe watermarks are possible\n");
 			return ret;
@@ -11806,29 +12148,19 @@ static int intel_crtc_atomic_check(struct drm_crtc *crtc,
 	}
 
 	if (INTEL_GEN(dev_priv) >= 9) {
-		if (mode_changed || pipe_config->update_pipe)
-			ret = skl_update_scaler_crtc(pipe_config);
-
-		if (!ret)
-			ret = icl_check_nv12_planes(pipe_config);
-		if (!ret)
-			ret = skl_check_pipe_max_pixel_rate(intel_crtc,
-							    pipe_config);
+		if (mode_changed || crtc_state->update_pipe)
+			ret = skl_update_scaler_crtc(crtc_state);
 		if (!ret)
-			ret = intel_atomic_setup_scalers(dev_priv, intel_crtc,
-							 pipe_config);
+			ret = intel_atomic_setup_scalers(dev_priv, crtc,
+							 crtc_state);
 	}
 
 	if (HAS_IPS(dev_priv))
-		pipe_config->ips_enabled = hsw_compute_ips_config(pipe_config);
+		crtc_state->ips_enabled = hsw_compute_ips_config(crtc_state);
 
 	return ret;
 }
 
-static const struct drm_crtc_helper_funcs intel_helper_funcs = {
-	.atomic_check = intel_crtc_atomic_check,
-};
-
 static void intel_modeset_update_connector_atomic_state(struct drm_device *dev)
 {
 	struct intel_connector *connector;
@@ -12138,6 +12470,15 @@ static void intel_dump_pipe_config(const struct intel_crtc_state *pipe_config,
 
 	intel_dpll_dump_hw_state(dev_priv, &pipe_config->dpll_hw_state);
 
+	if (IS_CHERRYVIEW(dev_priv))
+		DRM_DEBUG_KMS("cgm_mode: 0x%x gamma_mode: 0x%x gamma_enable: %d csc_enable: %d\n",
+			      pipe_config->cgm_mode, pipe_config->gamma_mode,
+			      pipe_config->gamma_enable, pipe_config->csc_enable);
+	else
+		DRM_DEBUG_KMS("csc_mode: 0x%x gamma_mode: 0x%x gamma_enable: %d csc_enable: %d\n",
+			      pipe_config->csc_mode, pipe_config->gamma_mode,
+			      pipe_config->gamma_enable, pipe_config->csc_enable);
+
 dump_planes:
 	if (!state)
 		return;
@@ -12158,6 +12499,12 @@ static bool check_digital_port_conflicts(struct intel_atomic_state *state)
 	bool ret = true;
 
 	/*
+	 * We're going to peek into connector->state,
+	 * hence connection_mutex must be held.
+	 */
+	drm_modeset_lock_assert_held(&dev->mode_config.connection_mutex);
+
+	/*
 	 * Walk the connector list instead of the encoder
 	 * list to detect the problem on ddi platforms
 	 * where there's just one encoder per digital port.
@@ -12239,6 +12586,13 @@ clear_intel_crtc_state(struct intel_crtc_state *crtc_state)
 	if (IS_G4X(dev_priv) ||
 	    IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv))
 		saved_state->wm = crtc_state->wm;
+	/*
+	 * Save the slave bitmask which gets filled for master crtc state during
+	 * slave atomic check call.
+	 */
+	if (is_trans_port_sync_master(crtc_state))
+		saved_state->sync_mode_slaves_mask =
+			crtc_state->sync_mode_slaves_mask;
 
 	/* Keep base drm_crtc_state intact, only clear our extended struct */
 	BUILD_BUG_ON(offsetof(struct intel_crtc_state, base));
@@ -12332,6 +12686,15 @@ encoder_retry:
 	drm_mode_set_crtcinfo(&pipe_config->base.adjusted_mode,
 			      CRTC_STEREO_DOUBLE);
 
+	/* Set the crtc_state defaults for trans_port_sync */
+	pipe_config->master_transcoder = INVALID_TRANSCODER;
+	ret = icl_add_sync_mode_crtcs(pipe_config);
+	if (ret) {
+		DRM_DEBUG_KMS("Cannot assign Sync Mode CRTCs: %d\n",
+			      ret);
+		return ret;
+	}
+
 	/* Pass our mode to the connectors and the CRTC to give them a chance to
 	 * adjust it according to limitations or connector properties, and also
 	 * a chance to reject the mode entirely.
@@ -12464,22 +12827,23 @@ pipe_config_infoframe_mismatch(struct drm_i915_private *dev_priv,
 		if ((drm_debug & DRM_UT_KMS) == 0)
 			return;
 
-		drm_dbg(DRM_UT_KMS, "fastset mismatch in %s infoframe", name);
-		drm_dbg(DRM_UT_KMS, "expected:");
+		DRM_DEBUG_KMS("fastset mismatch in %s infoframe\n", name);
+		DRM_DEBUG_KMS("expected:\n");
 		hdmi_infoframe_log(KERN_DEBUG, dev_priv->drm.dev, a);
-		drm_dbg(DRM_UT_KMS, "found");
+		DRM_DEBUG_KMS("found:\n");
 		hdmi_infoframe_log(KERN_DEBUG, dev_priv->drm.dev, b);
 	} else {
-		drm_err("mismatch in %s infoframe", name);
-		drm_err("expected:");
+		DRM_ERROR("mismatch in %s infoframe\n", name);
+		DRM_ERROR("expected:\n");
 		hdmi_infoframe_log(KERN_ERR, dev_priv->drm.dev, a);
-		drm_err("found");
+		DRM_ERROR("found:\n");
 		hdmi_infoframe_log(KERN_ERR, dev_priv->drm.dev, b);
 	}
 }
 
-static void __printf(3, 4)
-pipe_config_mismatch(bool fastset, const char *name, const char *format, ...)
+static void __printf(4, 5)
+pipe_config_mismatch(bool fastset, const struct intel_crtc *crtc,
+		     const char *name, const char *format, ...)
 {
 	struct va_format vaf;
 	va_list args;
@@ -12489,9 +12853,11 @@ pipe_config_mismatch(bool fastset, const char *name, const char *format, ...)
 	vaf.va = &args;
 
 	if (fastset)
-		drm_dbg(DRM_UT_KMS, "fastset mismatch in %s %pV", name, &vaf);
+		DRM_DEBUG_KMS("[CRTC:%d:%s] fastset mismatch in %s %pV\n",
+			      crtc->base.base.id, crtc->base.name, name, &vaf);
 	else
-		drm_err("mismatch in %s %pV", name, &vaf);
+		DRM_ERROR("[CRTC:%d:%s] mismatch in %s %pV\n",
+			  crtc->base.base.id, crtc->base.name, name, &vaf);
 
 	va_end(args);
 }
@@ -12519,7 +12885,9 @@ intel_pipe_config_compare(const struct intel_crtc_state *current_config,
 			  bool fastset)
 {
 	struct drm_i915_private *dev_priv = to_i915(current_config->base.crtc->dev);
+	struct intel_crtc *crtc = to_intel_crtc(pipe_config->base.crtc);
 	bool ret = true;
+	u32 bp_gamma = 0;
 	bool fixup_inherited = fastset &&
 		(current_config->base.mode.private_flags & I915_MODE_FLAG_INHERITED) &&
 		!(pipe_config->base.mode.private_flags & I915_MODE_FLAG_INHERITED);
@@ -12531,8 +12899,8 @@ intel_pipe_config_compare(const struct intel_crtc_state *current_config,
 
 #define PIPE_CONF_CHECK_X(name) do { \
 	if (current_config->name != pipe_config->name) { \
-		pipe_config_mismatch(fastset, __stringify(name), \
-				     "(expected 0x%08x, found 0x%08x)\n", \
+		pipe_config_mismatch(fastset, crtc, __stringify(name), \
+				     "(expected 0x%08x, found 0x%08x)", \
 				     current_config->name, \
 				     pipe_config->name); \
 		ret = false; \
@@ -12541,8 +12909,8 @@ intel_pipe_config_compare(const struct intel_crtc_state *current_config,
 
 #define PIPE_CONF_CHECK_I(name) do { \
 	if (current_config->name != pipe_config->name) { \
-		pipe_config_mismatch(fastset, __stringify(name), \
-				     "(expected %i, found %i)\n", \
+		pipe_config_mismatch(fastset, crtc, __stringify(name), \
+				     "(expected %i, found %i)", \
 				     current_config->name, \
 				     pipe_config->name); \
 		ret = false; \
@@ -12551,8 +12919,8 @@ intel_pipe_config_compare(const struct intel_crtc_state *current_config,
 
 #define PIPE_CONF_CHECK_BOOL(name) do { \
 	if (current_config->name != pipe_config->name) { \
-		pipe_config_mismatch(fastset, __stringify(name), \
-				     "(expected %s, found %s)\n", \
+		pipe_config_mismatch(fastset, crtc,  __stringify(name), \
+				     "(expected %s, found %s)", \
 				     yesno(current_config->name), \
 				     yesno(pipe_config->name)); \
 		ret = false; \
@@ -12568,8 +12936,8 @@ intel_pipe_config_compare(const struct intel_crtc_state *current_config,
 	if (!fixup_inherited || (!current_config->name && !pipe_config->name)) { \
 		PIPE_CONF_CHECK_BOOL(name); \
 	} else { \
-		pipe_config_mismatch(fastset, __stringify(name), \
-				     "unable to verify whether state matches exactly, forcing modeset (expected %s, found %s)\n", \
+		pipe_config_mismatch(fastset, crtc, __stringify(name), \
+				     "unable to verify whether state matches exactly, forcing modeset (expected %s, found %s)", \
 				     yesno(current_config->name), \
 				     yesno(pipe_config->name)); \
 		ret = false; \
@@ -12578,8 +12946,8 @@ intel_pipe_config_compare(const struct intel_crtc_state *current_config,
 
 #define PIPE_CONF_CHECK_P(name) do { \
 	if (current_config->name != pipe_config->name) { \
-		pipe_config_mismatch(fastset, __stringify(name), \
-				     "(expected %p, found %p)\n", \
+		pipe_config_mismatch(fastset, crtc, __stringify(name), \
+				     "(expected %p, found %p)", \
 				     current_config->name, \
 				     pipe_config->name); \
 		ret = false; \
@@ -12590,9 +12958,9 @@ intel_pipe_config_compare(const struct intel_crtc_state *current_config,
 	if (!intel_compare_link_m_n(&current_config->name, \
 				    &pipe_config->name,\
 				    !fastset)) { \
-		pipe_config_mismatch(fastset, __stringify(name), \
+		pipe_config_mismatch(fastset, crtc, __stringify(name), \
 				     "(expected tu %i gmch %i/%i link %i/%i, " \
-				     "found tu %i, gmch %i/%i link %i/%i)\n", \
+				     "found tu %i, gmch %i/%i link %i/%i)", \
 				     current_config->name.tu, \
 				     current_config->name.gmch_m, \
 				     current_config->name.gmch_n, \
@@ -12617,10 +12985,10 @@ intel_pipe_config_compare(const struct intel_crtc_state *current_config,
 				    &pipe_config->name, !fastset) && \
 	    !intel_compare_link_m_n(&current_config->alt_name, \
 				    &pipe_config->name, !fastset)) { \
-		pipe_config_mismatch(fastset, __stringify(name), \
+		pipe_config_mismatch(fastset, crtc, __stringify(name), \
 				     "(expected tu %i gmch %i/%i link %i/%i, " \
 				     "or tu %i gmch %i/%i link %i/%i, " \
-				     "found tu %i, gmch %i/%i link %i/%i)\n", \
+				     "found tu %i, gmch %i/%i link %i/%i)", \
 				     current_config->name.tu, \
 				     current_config->name.gmch_m, \
 				     current_config->name.gmch_n, \
@@ -12642,8 +13010,8 @@ intel_pipe_config_compare(const struct intel_crtc_state *current_config,
 
 #define PIPE_CONF_CHECK_FLAGS(name, mask) do { \
 	if ((current_config->name ^ pipe_config->name) & (mask)) { \
-		pipe_config_mismatch(fastset, __stringify(name), \
-				     "(%x) (expected %i, found %i)\n", \
+		pipe_config_mismatch(fastset, crtc, __stringify(name), \
+				     "(%x) (expected %i, found %i)", \
 				     (mask), \
 				     current_config->name & (mask), \
 				     pipe_config->name & (mask)); \
@@ -12653,8 +13021,8 @@ intel_pipe_config_compare(const struct intel_crtc_state *current_config,
 
 #define PIPE_CONF_CHECK_CLOCK_FUZZY(name) do { \
 	if (!intel_fuzzy_clock_check(current_config->name, pipe_config->name)) { \
-		pipe_config_mismatch(fastset, __stringify(name), \
-				     "(expected %i, found %i)\n", \
+		pipe_config_mismatch(fastset, crtc, __stringify(name), \
+				     "(expected %i, found %i)", \
 				     current_config->name, \
 				     pipe_config->name); \
 		ret = false; \
@@ -12671,6 +13039,24 @@ intel_pipe_config_compare(const struct intel_crtc_state *current_config,
 	} \
 } while (0)
 
+#define PIPE_CONF_CHECK_COLOR_LUT(name1, name2, bit_precision) do { \
+	if (current_config->name1 != pipe_config->name1) { \
+		pipe_config_mismatch(fastset, crtc, __stringify(name1), \
+				"(expected %i, found %i, won't compare lut values)", \
+				current_config->name1, \
+				pipe_config->name1); \
+		ret = false;\
+	} else { \
+		if (!intel_color_lut_equal(current_config->name2, \
+					pipe_config->name2, pipe_config->name1, \
+					bit_precision)) { \
+			pipe_config_mismatch(fastset, crtc, __stringify(name2), \
+					"hw_state doesn't match sw_state"); \
+			ret = false; \
+		} \
+	} \
+} while (0)
+
 #define PIPE_CONF_QUIRK(quirk) \
 	((current_config->quirks | pipe_config->quirks) & (quirk))
 
@@ -12709,6 +13095,7 @@ intel_pipe_config_compare(const struct intel_crtc_state *current_config,
 
 	PIPE_CONF_CHECK_I(pixel_multiplier);
 	PIPE_CONF_CHECK_I(output_format);
+	PIPE_CONF_CHECK_I(dc3co_exitline);
 	PIPE_CONF_CHECK_BOOL(has_hdmi_sink);
 	if ((INTEL_GEN(dev_priv) < 8 && !IS_HASWELL(dev_priv)) ||
 	    IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv))
@@ -12717,6 +13104,7 @@ intel_pipe_config_compare(const struct intel_crtc_state *current_config,
 	PIPE_CONF_CHECK_BOOL(hdmi_scrambling);
 	PIPE_CONF_CHECK_BOOL(hdmi_high_tmds_clock_ratio);
 	PIPE_CONF_CHECK_BOOL(has_infoframe);
+	PIPE_CONF_CHECK_BOOL(fec_enable);
 
 	PIPE_CONF_CHECK_BOOL_INCOMPLETE(has_audio);
 
@@ -12766,6 +13154,11 @@ intel_pipe_config_compare(const struct intel_crtc_state *current_config,
 			PIPE_CONF_CHECK_X(csc_mode);
 		PIPE_CONF_CHECK_BOOL(gamma_enable);
 		PIPE_CONF_CHECK_BOOL(csc_enable);
+
+		bp_gamma = intel_color_get_gamma_bit_precision(pipe_config);
+		if (bp_gamma)
+			PIPE_CONF_CHECK_COLOR_LUT(gamma_mode, base.gamma_lut, bp_gamma);
+
 	}
 
 	PIPE_CONF_CHECK_BOOL(double_wide);
@@ -12821,6 +13214,9 @@ intel_pipe_config_compare(const struct intel_crtc_state *current_config,
 	PIPE_CONF_CHECK_INFOFRAME(hdmi);
 	PIPE_CONF_CHECK_INFOFRAME(drm);
 
+	PIPE_CONF_CHECK_I(sync_mode_slaves_mask);
+	PIPE_CONF_CHECK_I(master_transcoder);
+
 #undef PIPE_CONF_CHECK_X
 #undef PIPE_CONF_CHECK_I
 #undef PIPE_CONF_CHECK_BOOL
@@ -12828,6 +13224,7 @@ intel_pipe_config_compare(const struct intel_crtc_state *current_config,
 #undef PIPE_CONF_CHECK_P
 #undef PIPE_CONF_CHECK_FLAGS
 #undef PIPE_CONF_CHECK_CLOCK_FUZZY
+#undef PIPE_CONF_CHECK_COLOR_LUT
 #undef PIPE_CONF_QUIRK
 
 	return ret;
@@ -13139,7 +13536,7 @@ intel_verify_planes(struct intel_atomic_state *state)
 
 	for_each_new_intel_plane_in_state(state, plane,
 					  plane_state, i)
-		assert_plane(plane, plane_state->slave ||
+		assert_plane(plane, plane_state->planar_slave ||
 			     plane_state->base.visible);
 }
 
@@ -13255,10 +13652,15 @@ intel_modeset_verify_disabled(struct drm_i915_private *dev_priv,
 	verify_disabled_dpll_state(dev_priv);
 }
 
-static void update_scanline_offset(const struct intel_crtc_state *crtc_state)
+static void
+intel_crtc_update_active_timings(const struct intel_crtc_state *crtc_state)
 {
 	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	const struct drm_display_mode *adjusted_mode =
+		&crtc_state->base.adjusted_mode;
+
+	drm_calc_timestamping_constants(&crtc->base, adjusted_mode);
 
 	/*
 	 * The scanline counter increments at the leading edge of hsync.
@@ -13288,7 +13690,6 @@ static void update_scanline_offset(const struct intel_crtc_state *crtc_state)
 	 * answer that's slightly in the future.
 	 */
 	if (IS_GEN(dev_priv, 2)) {
-		const struct drm_display_mode *adjusted_mode = &crtc_state->base.adjusted_mode;
 		int vtotal;
 
 		vtotal = adjusted_mode->crtc_vtotal;
@@ -13299,8 +13700,9 @@ static void update_scanline_offset(const struct intel_crtc_state *crtc_state)
 	} else if (HAS_DDI(dev_priv) &&
 		   intel_crtc_has_type(crtc_state, INTEL_OUTPUT_HDMI)) {
 		crtc->scanline_offset = 2;
-	} else
+	} else {
 		crtc->scanline_offset = 1;
+	}
 }
 
 static void intel_modeset_clear_plls(struct intel_atomic_state *state)
@@ -13382,158 +13784,43 @@ static int haswell_mode_set_planes_workaround(struct intel_atomic_state *state)
 	return 0;
 }
 
-static int intel_lock_all_pipes(struct intel_atomic_state *state)
-{
-	struct drm_i915_private *dev_priv = to_i915(state->base.dev);
-	struct intel_crtc *crtc;
-
-	/* Add all pipes to the state */
-	for_each_intel_crtc(&dev_priv->drm, crtc) {
-		struct intel_crtc_state *crtc_state;
-
-		crtc_state = intel_atomic_get_crtc_state(&state->base, crtc);
-		if (IS_ERR(crtc_state))
-			return PTR_ERR(crtc_state);
-	}
-
-	return 0;
-}
-
-static int intel_modeset_all_pipes(struct intel_atomic_state *state)
-{
-	struct drm_i915_private *dev_priv = to_i915(state->base.dev);
-	struct intel_crtc *crtc;
-
-	/*
-	 * Add all pipes to the state, and force
-	 * a modeset on all the active ones.
-	 */
-	for_each_intel_crtc(&dev_priv->drm, crtc) {
-		struct intel_crtc_state *crtc_state;
-		int ret;
-
-		crtc_state = intel_atomic_get_crtc_state(&state->base, crtc);
-		if (IS_ERR(crtc_state))
-			return PTR_ERR(crtc_state);
-
-		if (!crtc_state->base.active || needs_modeset(crtc_state))
-			continue;
-
-		crtc_state->base.mode_changed = true;
-
-		ret = drm_atomic_add_affected_connectors(&state->base,
-							 &crtc->base);
-		if (ret)
-			return ret;
-
-		ret = drm_atomic_add_affected_planes(&state->base,
-						     &crtc->base);
-		if (ret)
-			return ret;
-	}
-
-	return 0;
-}
-
 static int intel_modeset_checks(struct intel_atomic_state *state)
 {
 	struct drm_i915_private *dev_priv = to_i915(state->base.dev);
 	struct intel_crtc_state *old_crtc_state, *new_crtc_state;
 	struct intel_crtc *crtc;
-	int ret = 0, i;
-
-	if (!check_digital_port_conflicts(state)) {
-		DRM_DEBUG_KMS("rejecting conflicting digital port configuration\n");
-		return -EINVAL;
-	}
+	int ret, i;
 
 	/* keep the current setting */
 	if (!state->cdclk.force_min_cdclk_changed)
 		state->cdclk.force_min_cdclk = dev_priv->cdclk.force_min_cdclk;
 
 	state->modeset = true;
-	state->active_crtcs = dev_priv->active_crtcs;
+	state->active_pipes = dev_priv->active_pipes;
 	state->cdclk.logical = dev_priv->cdclk.logical;
 	state->cdclk.actual = dev_priv->cdclk.actual;
-	state->cdclk.pipe = INVALID_PIPE;
 
 	for_each_oldnew_intel_crtc_in_state(state, crtc, old_crtc_state,
 					    new_crtc_state, i) {
 		if (new_crtc_state->base.active)
-			state->active_crtcs |= 1 << i;
+			state->active_pipes |= BIT(crtc->pipe);
 		else
-			state->active_crtcs &= ~(1 << i);
+			state->active_pipes &= ~BIT(crtc->pipe);
 
 		if (old_crtc_state->base.active != new_crtc_state->base.active)
-			state->active_pipe_changes |= drm_crtc_mask(&crtc->base);
+			state->active_pipe_changes |= BIT(crtc->pipe);
 	}
 
-	/*
-	 * See if the config requires any additional preparation, e.g.
-	 * to adjust global state with pipes off.  We need to do this
-	 * here so we can get the modeset_pipe updated config for the new
-	 * mode set on this crtc.  For other crtcs we need to use the
-	 * adjusted_mode bits in the crtc directly.
-	 */
-	if (dev_priv->display.modeset_calc_cdclk) {
-		enum pipe pipe;
-
-		ret = dev_priv->display.modeset_calc_cdclk(state);
-		if (ret < 0)
+	if (state->active_pipe_changes) {
+		ret = intel_atomic_lock_global_state(state);
+		if (ret)
 			return ret;
-
-		/*
-		 * Writes to dev_priv->cdclk.logical must protected by
-		 * holding all the crtc locks, even if we don't end up
-		 * touching the hardware
-		 */
-		if (intel_cdclk_changed(&dev_priv->cdclk.logical,
-					&state->cdclk.logical)) {
-			ret = intel_lock_all_pipes(state);
-			if (ret < 0)
-				return ret;
-		}
-
-		if (is_power_of_2(state->active_crtcs)) {
-			struct intel_crtc *crtc;
-			struct intel_crtc_state *crtc_state;
-
-			pipe = ilog2(state->active_crtcs);
-			crtc = intel_get_crtc_for_pipe(dev_priv, pipe);
-			crtc_state = intel_atomic_get_new_crtc_state(state, crtc);
-			if (crtc_state && needs_modeset(crtc_state))
-				pipe = INVALID_PIPE;
-		} else {
-			pipe = INVALID_PIPE;
-		}
-
-		/* All pipes must be switched off while we change the cdclk. */
-		if (pipe != INVALID_PIPE &&
-		    intel_cdclk_needs_cd2x_update(dev_priv,
-						  &dev_priv->cdclk.actual,
-						  &state->cdclk.actual)) {
-			ret = intel_lock_all_pipes(state);
-			if (ret < 0)
-				return ret;
-
-			state->cdclk.pipe = pipe;
-		} else if (intel_cdclk_needs_modeset(&dev_priv->cdclk.actual,
-						     &state->cdclk.actual)) {
-			ret = intel_modeset_all_pipes(state);
-			if (ret < 0)
-				return ret;
-
-			state->cdclk.pipe = INVALID_PIPE;
-		}
-
-		DRM_DEBUG_KMS("New cdclk calculated to be logical %u kHz, actual %u kHz\n",
-			      state->cdclk.logical.cdclk,
-			      state->cdclk.actual.cdclk);
-		DRM_DEBUG_KMS("New voltage level calculated to be logical %u, actual %u\n",
-			      state->cdclk.logical.voltage_level,
-			      state->cdclk.actual.voltage_level);
 	}
 
+	ret = intel_modeset_calc_cdclk(state);
+	if (ret)
+		return ret;
+
 	intel_modeset_clear_plls(state);
 
 	if (IS_HASWELL(dev_priv))
@@ -13582,6 +13869,114 @@ static void intel_crtc_check_fastset(const struct intel_crtc_state *old_crtc_sta
 	new_crtc_state->has_drrs = old_crtc_state->has_drrs;
 }
 
+static int intel_crtc_add_planes_to_state(struct intel_atomic_state *state,
+					  struct intel_crtc *crtc,
+					  u8 plane_ids_mask)
+{
+	struct drm_i915_private *dev_priv = to_i915(state->base.dev);
+	struct intel_plane *plane;
+
+	for_each_intel_plane_on_crtc(&dev_priv->drm, crtc, plane) {
+		struct intel_plane_state *plane_state;
+
+		if ((plane_ids_mask & BIT(plane->id)) == 0)
+			continue;
+
+		plane_state = intel_atomic_get_plane_state(state, plane);
+		if (IS_ERR(plane_state))
+			return PTR_ERR(plane_state);
+	}
+
+	return 0;
+}
+
+static bool active_planes_affects_min_cdclk(struct drm_i915_private *dev_priv)
+{
+	/* See {hsw,vlv,ivb}_plane_ratio() */
+	return IS_BROADWELL(dev_priv) || IS_HASWELL(dev_priv) ||
+		IS_CHERRYVIEW(dev_priv) || IS_VALLEYVIEW(dev_priv) ||
+		IS_IVYBRIDGE(dev_priv);
+}
+
+static int intel_atomic_check_planes(struct intel_atomic_state *state,
+				     bool *need_modeset)
+{
+	struct drm_i915_private *dev_priv = to_i915(state->base.dev);
+	struct intel_crtc_state *old_crtc_state, *new_crtc_state;
+	struct intel_plane_state *plane_state;
+	struct intel_plane *plane;
+	struct intel_crtc *crtc;
+	int i, ret;
+
+	ret = icl_add_linked_planes(state);
+	if (ret)
+		return ret;
+
+	for_each_new_intel_plane_in_state(state, plane, plane_state, i) {
+		ret = intel_plane_atomic_check(state, plane);
+		if (ret) {
+			DRM_DEBUG_ATOMIC("[PLANE:%d:%s] atomic driver check failed\n",
+					 plane->base.base.id, plane->base.name);
+			return ret;
+		}
+	}
+
+	for_each_oldnew_intel_crtc_in_state(state, crtc, old_crtc_state,
+					    new_crtc_state, i) {
+		u8 old_active_planes, new_active_planes;
+
+		ret = icl_check_nv12_planes(new_crtc_state);
+		if (ret)
+			return ret;
+
+		/*
+		 * On some platforms the number of active planes affects
+		 * the planes' minimum cdclk calculation. Add such planes
+		 * to the state before we compute the minimum cdclk.
+		 */
+		if (!active_planes_affects_min_cdclk(dev_priv))
+			continue;
+
+		old_active_planes = old_crtc_state->active_planes & ~BIT(PLANE_CURSOR);
+		new_active_planes = new_crtc_state->active_planes & ~BIT(PLANE_CURSOR);
+
+		if (hweight8(old_active_planes) == hweight8(new_active_planes))
+			continue;
+
+		ret = intel_crtc_add_planes_to_state(state, crtc, new_active_planes);
+		if (ret)
+			return ret;
+	}
+
+	/*
+	 * active_planes bitmask has been updated, and potentially
+	 * affected planes are part of the state. We can now
+	 * compute the minimum cdclk for each plane.
+	 */
+	for_each_new_intel_plane_in_state(state, plane, plane_state, i)
+		*need_modeset |= intel_plane_calc_min_cdclk(state, plane);
+
+	return 0;
+}
+
+static int intel_atomic_check_crtcs(struct intel_atomic_state *state)
+{
+	struct intel_crtc_state *crtc_state;
+	struct intel_crtc *crtc;
+	int i;
+
+	for_each_new_intel_crtc_in_state(state, crtc, crtc_state, i) {
+		int ret = intel_crtc_atomic_check(state, crtc);
+		if (ret) {
+			DRM_DEBUG_ATOMIC("[CRTC:%d:%s] atomic driver check failed\n",
+					 crtc->base.base.id, crtc->base.name);
+			return ret;
+		}
+	}
+
+	return 0;
+}
+
 /**
  * intel_atomic_check - validate state object
  * @dev: drm device
@@ -13595,7 +13990,7 @@ static int intel_atomic_check(struct drm_device *dev,
 	struct intel_crtc_state *old_crtc_state, *new_crtc_state;
 	struct intel_crtc *crtc;
 	int ret, i;
-	bool any_ms = state->cdclk.force_min_cdclk_changed;
+	bool any_ms = false;
 
 	/* Catch I915_MODE_FLAG_INHERITED */
 	for_each_oldnew_intel_crtc_in_state(state, crtc, old_crtc_state,
@@ -13629,10 +14024,22 @@ static int intel_atomic_check(struct drm_device *dev,
 			any_ms = true;
 	}
 
+	if (any_ms && !check_digital_port_conflicts(state)) {
+		DRM_DEBUG_KMS("rejecting conflicting digital port configuration\n");
+		ret = EINVAL;
+		goto fail;
+	}
+
 	ret = drm_dp_mst_atomic_check(&state->base);
 	if (ret)
 		goto fail;
 
+	any_ms |= state->cdclk.force_min_cdclk_changed;
+
+	ret = intel_atomic_check_planes(state, &any_ms);
+	if (ret)
+		goto fail;
+
 	if (any_ms) {
 		ret = intel_modeset_checks(state);
 		if (ret)
@@ -13641,11 +14048,7 @@ static int intel_atomic_check(struct drm_device *dev,
 		state->cdclk.logical = dev_priv->cdclk.logical;
 	}
 
-	ret = icl_add_linked_planes(state);
-	if (ret)
-		goto fail;
-
-	ret = drm_atomic_helper_check_planes(dev, &state->base);
+	ret = intel_atomic_check_crtcs(state);
 	if (ret)
 		goto fail;
 
@@ -13703,25 +14106,110 @@ u32 intel_crtc_get_vblank_counter(struct intel_crtc *crtc)
 	return crtc->base.funcs->get_vblank_counter(&crtc->base);
 }
 
+void intel_crtc_arm_fifo_underrun(struct intel_crtc *crtc,
+				  struct intel_crtc_state *crtc_state)
+{
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+
+	if (!IS_GEN(dev_priv, 2))
+		intel_set_cpu_fifo_underrun_reporting(dev_priv, crtc->pipe, true);
+
+	if (crtc_state->has_pch_encoder) {
+		enum pipe pch_transcoder =
+			intel_crtc_pch_transcoder(crtc);
+
+		intel_set_pch_fifo_underrun_reporting(dev_priv, pch_transcoder, true);
+	}
+}
+
+static void intel_pipe_fastset(const struct intel_crtc_state *old_crtc_state,
+			       const struct intel_crtc_state *new_crtc_state)
+{
+	struct intel_crtc *crtc = to_intel_crtc(new_crtc_state->base.crtc);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+
+	/*
+	 * Update pipe size and adjust fitter if needed: the reason for this is
+	 * that in compute_mode_changes we check the native mode (not the pfit
+	 * mode) to see if we can flip rather than do a full mode set. In the
+	 * fastboot case, we'll flip, but if we don't update the pipesrc and
+	 * pfit state, we'll end up with a big fb scanned out into the wrong
+	 * sized surface.
+	 */
+	intel_set_pipe_src_size(new_crtc_state);
+
+	/* on skylake this is done by detaching scalers */
+	if (INTEL_GEN(dev_priv) >= 9) {
+		skl_detach_scalers(new_crtc_state);
+
+		if (new_crtc_state->pch_pfit.enabled)
+			skylake_pfit_enable(new_crtc_state);
+	} else if (HAS_PCH_SPLIT(dev_priv)) {
+		if (new_crtc_state->pch_pfit.enabled)
+			ironlake_pfit_enable(new_crtc_state);
+		else if (old_crtc_state->pch_pfit.enabled)
+			ironlake_pfit_disable(old_crtc_state);
+	}
+
+	if (INTEL_GEN(dev_priv) >= 11)
+		icl_set_pipe_chicken(crtc);
+}
+
+static void commit_pipe_config(struct intel_atomic_state *state,
+			       struct intel_crtc_state *old_crtc_state,
+			       struct intel_crtc_state *new_crtc_state)
+{
+	struct drm_i915_private *dev_priv = to_i915(state->base.dev);
+	bool modeset = needs_modeset(new_crtc_state);
+
+	/*
+	 * During modesets pipe configuration was programmed as the
+	 * CRTC was enabled.
+	 */
+	if (!modeset) {
+		if (new_crtc_state->base.color_mgmt_changed ||
+		    new_crtc_state->update_pipe)
+			intel_color_commit(new_crtc_state);
+
+		if (INTEL_GEN(dev_priv) >= 9)
+			skl_detach_scalers(new_crtc_state);
+
+		if (INTEL_GEN(dev_priv) >= 9 || IS_BROADWELL(dev_priv))
+			bdw_set_pipemisc(new_crtc_state);
+
+		if (new_crtc_state->update_pipe)
+			intel_pipe_fastset(old_crtc_state, new_crtc_state);
+	}
+
+	if (dev_priv->display.atomic_update_watermarks)
+		dev_priv->display.atomic_update_watermarks(state,
+							   new_crtc_state);
+}
+
 static void intel_update_crtc(struct intel_crtc *crtc,
 			      struct intel_atomic_state *state,
 			      struct intel_crtc_state *old_crtc_state,
 			      struct intel_crtc_state *new_crtc_state)
 {
-	struct drm_device *dev = state->base.dev;
-	struct drm_i915_private *dev_priv = to_i915(dev);
+	struct drm_i915_private *dev_priv = to_i915(state->base.dev);
 	bool modeset = needs_modeset(new_crtc_state);
 	struct intel_plane_state *new_plane_state =
 		intel_atomic_get_new_plane_state(state,
 						 to_intel_plane(crtc->base.primary));
 
 	if (modeset) {
-		update_scanline_offset(new_crtc_state);
+		intel_crtc_update_active_timings(new_crtc_state);
+
 		dev_priv->display.crtc_enable(new_crtc_state, state);
 
 		/* vblanks work again, re-enable pipe CRC. */
 		intel_crtc_enable_pipe_crc(crtc);
 	} else {
+		if (new_crtc_state->preload_luts &&
+		    (new_crtc_state->base.color_mgmt_changed ||
+		     new_crtc_state->update_pipe))
+			intel_color_load_luts(new_crtc_state);
+
 		intel_pre_plane_update(old_crtc_state, new_crtc_state);
 
 		if (new_crtc_state->update_pipe)
@@ -13733,17 +14221,151 @@ static void intel_update_crtc(struct intel_crtc *crtc,
 	else if (new_plane_state)
 		intel_fbc_enable(crtc, new_crtc_state, new_plane_state);
 
-	intel_begin_crtc_commit(state, crtc);
+	/* Perform vblank evasion around commit operation */
+	intel_pipe_update_start(new_crtc_state);
+
+	commit_pipe_config(state, old_crtc_state, new_crtc_state);
 
 	if (INTEL_GEN(dev_priv) >= 9)
 		skl_update_planes_on_crtc(state, crtc);
 	else
 		i9xx_update_planes_on_crtc(state, crtc);
 
-	intel_finish_crtc_commit(state, crtc);
+	intel_pipe_update_end(new_crtc_state);
+
+	/*
+	 * We usually enable FIFO underrun interrupts as part of the
+	 * CRTC enable sequence during modesets.  But when we inherit a
+	 * valid pipe configuration from the BIOS we need to take care
+	 * of enabling them on the CRTC's first fastset.
+	 */
+	if (new_crtc_state->update_pipe && !modeset &&
+	    old_crtc_state->base.mode.private_flags & I915_MODE_FLAG_INHERITED)
+		intel_crtc_arm_fifo_underrun(crtc, new_crtc_state);
+}
+
+static struct intel_crtc *intel_get_slave_crtc(const struct intel_crtc_state *new_crtc_state)
+{
+	struct drm_i915_private *dev_priv = to_i915(new_crtc_state->base.crtc->dev);
+	enum transcoder slave_transcoder;
+
+	WARN_ON(!is_power_of_2(new_crtc_state->sync_mode_slaves_mask));
+
+	slave_transcoder = ffs(new_crtc_state->sync_mode_slaves_mask) - 1;
+	return intel_get_crtc_for_pipe(dev_priv,
+				       (enum pipe)slave_transcoder);
+}
+
+static void intel_old_crtc_state_disables(struct intel_atomic_state *state,
+					  struct intel_crtc_state *old_crtc_state,
+					  struct intel_crtc_state *new_crtc_state,
+					  struct intel_crtc *crtc)
+{
+	struct drm_i915_private *dev_priv = to_i915(state->base.dev);
+
+	intel_crtc_disable_planes(state, crtc);
+
+	/*
+	 * We need to disable pipe CRC before disabling the pipe,
+	 * or we race against vblank off.
+	 */
+	intel_crtc_disable_pipe_crc(crtc);
+
+	dev_priv->display.crtc_disable(old_crtc_state, state);
+	crtc->active = false;
+	intel_fbc_disable(crtc);
+	intel_disable_shared_dpll(old_crtc_state);
+
+	/*
+	 * Underruns don't always raise interrupts,
+	 * so check manually.
+	 */
+	intel_check_cpu_fifo_underruns(dev_priv);
+	intel_check_pch_fifo_underruns(dev_priv);
+
+	/* FIXME unify this for all platforms */
+	if (!new_crtc_state->base.active &&
+	    !HAS_GMCH(dev_priv) &&
+	    dev_priv->display.initial_watermarks)
+		dev_priv->display.initial_watermarks(state,
+						     new_crtc_state);
+}
+
+static void intel_trans_port_sync_modeset_disables(struct intel_atomic_state *state,
+						   struct intel_crtc *crtc,
+						   struct intel_crtc_state *old_crtc_state,
+						   struct intel_crtc_state *new_crtc_state)
+{
+	struct intel_crtc *slave_crtc = intel_get_slave_crtc(new_crtc_state);
+	struct intel_crtc_state *new_slave_crtc_state =
+		intel_atomic_get_new_crtc_state(state, slave_crtc);
+	struct intel_crtc_state *old_slave_crtc_state =
+		intel_atomic_get_old_crtc_state(state, slave_crtc);
+
+	WARN_ON(!slave_crtc || !new_slave_crtc_state ||
+		!old_slave_crtc_state);
+
+	/* Disable Slave first */
+	intel_pre_plane_update(old_slave_crtc_state, new_slave_crtc_state);
+	if (old_slave_crtc_state->base.active)
+		intel_old_crtc_state_disables(state,
+					      old_slave_crtc_state,
+					      new_slave_crtc_state,
+					      slave_crtc);
+
+	/* Disable Master */
+	intel_pre_plane_update(old_crtc_state, new_crtc_state);
+	if (old_crtc_state->base.active)
+		intel_old_crtc_state_disables(state,
+					      old_crtc_state,
+					      new_crtc_state,
+					      crtc);
 }
 
-static void intel_update_crtcs(struct intel_atomic_state *state)
+static void intel_commit_modeset_disables(struct intel_atomic_state *state)
+{
+	struct intel_crtc_state *new_crtc_state, *old_crtc_state;
+	struct intel_crtc *crtc;
+	int i;
+
+	/*
+	 * Disable CRTC/pipes in reverse order because some features(MST in
+	 * TGL+) requires master and slave relationship between pipes, so it
+	 * should always pick the lowest pipe as master as it will be enabled
+	 * first and disable in the reverse order so the master will be the
+	 * last one to be disabled.
+	 */
+	for_each_oldnew_intel_crtc_in_state_reverse(state, crtc, old_crtc_state,
+						    new_crtc_state, i) {
+		if (!needs_modeset(new_crtc_state))
+			continue;
+
+		/* In case of Transcoder port Sync master slave CRTCs can be
+		 * assigned in any order and we need to make sure that
+		 * slave CRTCs are disabled first and then master CRTC since
+		 * Slave vblanks are masked till Master Vblanks.
+		 */
+		if (is_trans_port_sync_mode(new_crtc_state)) {
+			if (is_trans_port_sync_master(new_crtc_state))
+				intel_trans_port_sync_modeset_disables(state,
+								       crtc,
+								       old_crtc_state,
+								       new_crtc_state);
+			else
+				continue;
+		} else {
+			intel_pre_plane_update(old_crtc_state, new_crtc_state);
+
+			if (old_crtc_state->base.active)
+				intel_old_crtc_state_disables(state,
+							      old_crtc_state,
+							      new_crtc_state,
+							      crtc);
+		}
+	}
+}
+
+static void intel_commit_modeset_enables(struct intel_atomic_state *state)
 {
 	struct intel_crtc *crtc;
 	struct intel_crtc_state *old_crtc_state, *new_crtc_state;
@@ -13758,14 +14380,120 @@ static void intel_update_crtcs(struct intel_atomic_state *state)
 	}
 }
 
-static void skl_update_crtcs(struct intel_atomic_state *state)
+static void intel_crtc_enable_trans_port_sync(struct intel_crtc *crtc,
+					      struct intel_atomic_state *state,
+					      struct intel_crtc_state *new_crtc_state)
+{
+	struct drm_i915_private *dev_priv = to_i915(state->base.dev);
+
+	intel_crtc_update_active_timings(new_crtc_state);
+	dev_priv->display.crtc_enable(new_crtc_state, state);
+	intel_crtc_enable_pipe_crc(crtc);
+}
+
+static void intel_set_dp_tp_ctl_normal(struct intel_crtc *crtc,
+				       struct intel_atomic_state *state)
+{
+	struct drm_connector *uninitialized_var(conn);
+	struct drm_connector_state *conn_state;
+	struct intel_dp *intel_dp;
+	int i;
+
+	for_each_new_connector_in_state(&state->base, conn, conn_state, i) {
+		if (conn_state->crtc == &crtc->base)
+			break;
+	}
+	intel_dp = enc_to_intel_dp(&intel_attached_encoder(conn)->base);
+	intel_dp_stop_link_train(intel_dp);
+}
+
+static void intel_post_crtc_enable_updates(struct intel_crtc *crtc,
+					   struct intel_atomic_state *state)
+{
+	struct intel_crtc_state *new_crtc_state =
+		intel_atomic_get_new_crtc_state(state, crtc);
+	struct intel_crtc_state *old_crtc_state =
+		intel_atomic_get_old_crtc_state(state, crtc);
+	struct intel_plane_state *new_plane_state =
+		intel_atomic_get_new_plane_state(state,
+						 to_intel_plane(crtc->base.primary));
+	bool modeset = needs_modeset(new_crtc_state);
+
+	if (new_crtc_state->update_pipe && !new_crtc_state->enable_fbc)
+		intel_fbc_disable(crtc);
+	else if (new_plane_state)
+		intel_fbc_enable(crtc, new_crtc_state, new_plane_state);
+
+	/* Perform vblank evasion around commit operation */
+	intel_pipe_update_start(new_crtc_state);
+	commit_pipe_config(state, old_crtc_state, new_crtc_state);
+	skl_update_planes_on_crtc(state, crtc);
+	intel_pipe_update_end(new_crtc_state);
+
+	/*
+	 * We usually enable FIFO underrun interrupts as part of the
+	 * CRTC enable sequence during modesets.  But when we inherit a
+	 * valid pipe configuration from the BIOS we need to take care
+	 * of enabling them on the CRTC's first fastset.
+	 */
+	if (new_crtc_state->update_pipe && !modeset &&
+	    old_crtc_state->base.mode.private_flags & I915_MODE_FLAG_INHERITED)
+		intel_crtc_arm_fifo_underrun(crtc, new_crtc_state);
+}
+
+static void intel_update_trans_port_sync_crtcs(struct intel_crtc *crtc,
+					       struct intel_atomic_state *state,
+					       struct intel_crtc_state *old_crtc_state,
+					       struct intel_crtc_state *new_crtc_state)
+{
+	struct intel_crtc *slave_crtc = intel_get_slave_crtc(new_crtc_state);
+	struct intel_crtc_state *new_slave_crtc_state =
+		intel_atomic_get_new_crtc_state(state, slave_crtc);
+	struct intel_crtc_state *old_slave_crtc_state =
+		intel_atomic_get_old_crtc_state(state, slave_crtc);
+
+	WARN_ON(!slave_crtc || !new_slave_crtc_state ||
+		!old_slave_crtc_state);
+
+	DRM_DEBUG_KMS("Updating Transcoder Port Sync Master CRTC = %d %s and Slave CRTC %d %s\n",
+		      crtc->base.base.id, crtc->base.name, slave_crtc->base.base.id,
+		      slave_crtc->base.name);
+
+	/* Enable seq for slave with with DP_TP_CTL left Idle until the
+	 * master is ready
+	 */
+	intel_crtc_enable_trans_port_sync(slave_crtc,
+					  state,
+					  new_slave_crtc_state);
+
+	/* Enable seq for master with with DP_TP_CTL left Idle */
+	intel_crtc_enable_trans_port_sync(crtc,
+					  state,
+					  new_crtc_state);
+
+	/* Set Slave's DP_TP_CTL to Normal */
+	intel_set_dp_tp_ctl_normal(slave_crtc,
+				   state);
+
+	/* Set Master's DP_TP_CTL To Normal */
+	usleep_range(200, 400);
+	intel_set_dp_tp_ctl_normal(crtc,
+				   state);
+
+	/* Now do the post crtc enable for all master and slaves */
+	intel_post_crtc_enable_updates(slave_crtc,
+				       state);
+	intel_post_crtc_enable_updates(crtc,
+				       state);
+}
+
+static void skl_commit_modeset_enables(struct intel_atomic_state *state)
 {
 	struct drm_i915_private *dev_priv = to_i915(state->base.dev);
 	struct intel_crtc *crtc;
 	struct intel_crtc_state *old_crtc_state, *new_crtc_state;
 	unsigned int updated = 0;
 	bool progress;
-	enum pipe pipe;
 	int i;
 	u8 hw_enabled_slices = dev_priv->wm.skl_hw.ddb.enabled_slices;
 	u8 required_slices = state->wm_results.ddb.enabled_slices;
@@ -13790,20 +14518,19 @@ static void skl_update_crtcs(struct intel_atomic_state *state)
 		progress = false;
 
 		for_each_oldnew_intel_crtc_in_state(state, crtc, old_crtc_state, new_crtc_state, i) {
+			enum pipe pipe = crtc->pipe;
 			bool vbl_wait = false;
-			unsigned int cmask = drm_crtc_mask(&crtc->base);
-
-			pipe = crtc->pipe;
+			bool modeset = needs_modeset(new_crtc_state);
 
-			if (updated & cmask || !new_crtc_state->base.active)
+			if (updated & BIT(crtc->pipe) || !new_crtc_state->base.active)
 				continue;
 
 			if (skl_ddb_allocation_overlaps(&new_crtc_state->wm.skl.ddb,
 							entries,
-							INTEL_INFO(dev_priv)->num_pipes, i))
+							INTEL_NUM_PIPES(dev_priv), i))
 				continue;
 
-			updated |= cmask;
+			updated |= BIT(pipe);
 			entries[i] = new_crtc_state->wm.skl.ddb;
 
 			/*
@@ -13814,12 +14541,22 @@ static void skl_update_crtcs(struct intel_atomic_state *state)
 			 */
 			if (!skl_ddb_entry_equal(&new_crtc_state->wm.skl.ddb,
 						 &old_crtc_state->wm.skl.ddb) &&
-			    !new_crtc_state->base.active_changed &&
+			    !modeset &&
 			    state->wm_results.dirty_pipes != updated)
 				vbl_wait = true;
 
-			intel_update_crtc(crtc, state, old_crtc_state,
-					  new_crtc_state);
+			if (modeset && is_trans_port_sync_mode(new_crtc_state)) {
+				if (is_trans_port_sync_master(new_crtc_state))
+					intel_update_trans_port_sync_crtcs(crtc,
+									   state,
+									   old_crtc_state,
+									   new_crtc_state);
+				else
+					continue;
+			} else {
+				intel_update_crtc(crtc, state, old_crtc_state,
+						  new_crtc_state);
+			}
 
 			if (vbl_wait)
 				intel_wait_for_vblank(dev_priv, pipe);
@@ -13908,49 +14645,18 @@ static void intel_atomic_commit_tail(struct intel_atomic_state *state)
 	if (state->modeset)
 		wakeref = intel_display_power_get(dev_priv, POWER_DOMAIN_MODESET);
 
-	for_each_oldnew_intel_crtc_in_state(state, crtc, old_crtc_state, new_crtc_state, i) {
+	for_each_oldnew_intel_crtc_in_state(state, crtc, old_crtc_state,
+					    new_crtc_state, i) {
 		if (needs_modeset(new_crtc_state) ||
 		    new_crtc_state->update_pipe) {
 
 			put_domains[crtc->pipe] =
 				modeset_get_crtc_power_domains(new_crtc_state);
 		}
-
-		if (!needs_modeset(new_crtc_state))
-			continue;
-
-		intel_pre_plane_update(old_crtc_state, new_crtc_state);
-
-		if (old_crtc_state->base.active) {
-			intel_crtc_disable_planes(state, crtc);
-
-			/*
-			 * We need to disable pipe CRC before disabling the pipe,
-			 * or we race against vblank off.
-			 */
-			intel_crtc_disable_pipe_crc(crtc);
-
-			dev_priv->display.crtc_disable(old_crtc_state, state);
-			crtc->active = false;
-			intel_fbc_disable(crtc);
-			intel_disable_shared_dpll(old_crtc_state);
-
-			/*
-			 * Underruns don't always raise
-			 * interrupts, so check manually.
-			 */
-			intel_check_cpu_fifo_underruns(dev_priv);
-			intel_check_pch_fifo_underruns(dev_priv);
-
-			/* FIXME unify this for all platforms */
-			if (!new_crtc_state->base.active &&
-			    !HAS_GMCH(dev_priv) &&
-			    dev_priv->display.initial_watermarks)
-				dev_priv->display.initial_watermarks(state,
-								     new_crtc_state);
-		}
 	}
 
+	intel_commit_modeset_disables(state);
+
 	/* FIXME: Eventually get rid of our crtc->config pointer */
 	for_each_new_intel_crtc_in_state(state, crtc, new_crtc_state, i)
 		crtc->config = new_crtc_state;
@@ -13991,7 +14697,7 @@ static void intel_atomic_commit_tail(struct intel_atomic_state *state)
 		intel_encoders_update_prepare(state);
 
 	/* Now enable the clocks, plane, pipe, and connectors that we set up. */
-	dev_priv->display.update_crtcs(state);
+	dev_priv->display.commit_modeset_enables(state);
 
 	if (state->modeset) {
 		intel_encoders_update_complete(state);
@@ -14016,6 +14722,7 @@ static void intel_atomic_commit_tail(struct intel_atomic_state *state)
 	for_each_new_intel_crtc_in_state(state, crtc, new_crtc_state, i) {
 		if (new_crtc_state->base.active &&
 		    !needs_modeset(new_crtc_state) &&
+		    !new_crtc_state->preload_luts &&
 		    (new_crtc_state->base.color_mgmt_changed ||
 		     new_crtc_state->update_pipe))
 			intel_color_load_luts(new_crtc_state);
@@ -14121,6 +14828,14 @@ static void intel_atomic_track_fbs(struct intel_atomic_state *state)
 					plane->frontbuffer_bit);
 }
 
+static void assert_global_state_locked(struct drm_i915_private *dev_priv)
+{
+	struct intel_crtc *crtc;
+
+	for_each_intel_crtc(&dev_priv->drm, crtc)
+		drm_modeset_lock_assert_held(&crtc->base.mutex);
+}
+
 static int intel_atomic_commit(struct drm_device *dev,
 			       struct drm_atomic_state *_state,
 			       bool nonblock)
@@ -14186,12 +14901,14 @@ static int intel_atomic_commit(struct drm_device *dev,
 	intel_shared_dpll_swap_state(state);
 	intel_atomic_track_fbs(state);
 
-	if (state->modeset) {
+	if (state->global_state_changed) {
+		assert_global_state_locked(dev_priv);
+
 		memcpy(dev_priv->min_cdclk, state->min_cdclk,
 		       sizeof(state->min_cdclk));
 		memcpy(dev_priv->min_voltage_level, state->min_voltage_level,
 		       sizeof(state->min_voltage_level));
-		dev_priv->active_crtcs = state->active_crtcs;
+		dev_priv->active_pipes = state->active_pipes;
 		dev_priv->cdclk.force_min_cdclk = state->cdclk.force_min_cdclk;
 
 		intel_cdclk_swap_state(state);
@@ -14204,7 +14921,7 @@ static int intel_atomic_commit(struct drm_device *dev,
 	if (nonblock && state->modeset) {
 		queue_work(dev_priv->modeset_wq, &state->base.commit_work);
 	} else if (nonblock) {
-		queue_work(system_unbound_wq, &state->base.commit_work);
+		queue_work(dev_priv->flip_wq, &state->base.commit_work);
 	} else {
 		if (state->modeset)
 			flush_workqueue(dev_priv->modeset_wq);
@@ -14233,7 +14950,7 @@ static int do_rps_boost(struct wait_queue_entry *_wait,
 	 * vblank without our intervention, so leave RPS alone.
 	 */
 	if (!i915_request_started(rq))
-		gen6_rps_boost(rq);
+		intel_rps_boost(rq);
 	i915_request_put(rq);
 
 	drm_crtc_vblank_put(wait->crtc);
@@ -14314,7 +15031,7 @@ static void intel_plane_unpin_fb(struct intel_plane_state *old_plane_state)
 static void fb_obj_bump_render_priority(struct drm_i915_gem_object *obj)
 {
 	struct i915_sched_attr attr = {
-		.priority = I915_PRIORITY_DISPLAY,
+		.priority = I915_USER_PRIORITY(I915_PRIORITY_DISPLAY),
 	};
 
 	i915_gem_object_wait_priority(obj, 0, &attr);
@@ -14323,25 +15040,25 @@ static void fb_obj_bump_render_priority(struct drm_i915_gem_object *obj)
 /**
  * intel_prepare_plane_fb - Prepare fb for usage on plane
  * @plane: drm plane to prepare for
- * @new_state: the plane state being prepared
+ * @_new_plane_state: the plane state being prepared
  *
  * Prepares a framebuffer for usage on a display plane.  Generally this
  * involves pinning the underlying object and updating the frontbuffer tracking
  * bits.  Some older platforms need special physical address handling for
  * cursor planes.
  *
- * Must be called with struct_mutex held.
- *
  * Returns 0 on success, negative error code on failure.
  */
 int
 intel_prepare_plane_fb(struct drm_plane *plane,
-		       struct drm_plane_state *new_state)
+		       struct drm_plane_state *_new_plane_state)
 {
+	struct intel_plane_state *new_plane_state =
+		to_intel_plane_state(_new_plane_state);
 	struct intel_atomic_state *intel_state =
-		to_intel_atomic_state(new_state->state);
+		to_intel_atomic_state(new_plane_state->base.state);
 	struct drm_i915_private *dev_priv = to_i915(plane->dev);
-	struct drm_framebuffer *fb = new_state->fb;
+	struct drm_framebuffer *fb = new_plane_state->base.fb;
 	struct drm_i915_gem_object *obj = intel_fb_obj(fb);
 	struct drm_i915_gem_object *old_obj = intel_fb_obj(plane->state->fb);
 	int ret;
@@ -14372,9 +15089,9 @@ intel_prepare_plane_fb(struct drm_plane *plane,
 		}
 	}
 
-	if (new_state->fence) { /* explicit fencing */
+	if (new_plane_state->base.fence) { /* explicit fencing */
 		ret = i915_sw_fence_await_dma_fence(&intel_state->commit_ready,
-						    new_state->fence,
+						    new_plane_state->base.fence,
 						    I915_FENCE_TIMEOUT,
 						    GFP_KERNEL);
 		if (ret < 0)
@@ -14388,15 +15105,8 @@ intel_prepare_plane_fb(struct drm_plane *plane,
 	if (ret)
 		return ret;
 
-	ret = mutex_lock_interruptible(&dev_priv->drm.struct_mutex);
-	if (ret) {
-		i915_gem_object_unpin_pages(obj);
-		return ret;
-	}
-
-	ret = intel_plane_pin_fb(to_intel_plane_state(new_state));
+	ret = intel_plane_pin_fb(new_plane_state);
 
-	mutex_unlock(&dev_priv->drm.struct_mutex);
 	i915_gem_object_unpin_pages(obj);
 	if (ret)
 		return ret;
@@ -14404,7 +15114,7 @@ intel_prepare_plane_fb(struct drm_plane *plane,
 	fb_obj_bump_render_priority(obj);
 	intel_frontbuffer_flush(obj->frontbuffer, ORIGIN_DIRTYFB);
 
-	if (!new_state->fence) { /* implicit fencing */
+	if (!new_plane_state->base.fence) { /* implicit fencing */
 		struct dma_fence *fence;
 
 		ret = i915_sw_fence_await_reservation(&intel_state->commit_ready,
@@ -14416,11 +15126,13 @@ intel_prepare_plane_fb(struct drm_plane *plane,
 
 		fence = dma_resv_get_excl_rcu(obj->base.resv);
 		if (fence) {
-			add_rps_boost_after_vblank(new_state->crtc, fence);
+			add_rps_boost_after_vblank(new_plane_state->base.crtc,
+						   fence);
 			dma_fence_put(fence);
 		}
 	} else {
-		add_rps_boost_after_vblank(new_state->crtc, new_state->fence);
+		add_rps_boost_after_vblank(new_plane_state->base.crtc,
+					   new_plane_state->base.fence);
 	}
 
 	/*
@@ -14432,7 +15144,7 @@ intel_prepare_plane_fb(struct drm_plane *plane,
 	 * maximum clocks following a vblank miss (see do_rps_boost()).
 	 */
 	if (!intel_state->rps_interactive) {
-		intel_rps_mark_interactive(dev_priv, true);
+		intel_rps_mark_interactive(&dev_priv->gt.rps, true);
 		intel_state->rps_interactive = true;
 	}
 
@@ -14442,130 +15154,27 @@ intel_prepare_plane_fb(struct drm_plane *plane,
 /**
  * intel_cleanup_plane_fb - Cleans up an fb after plane use
  * @plane: drm plane to clean up for
- * @old_state: the state from the previous modeset
+ * @_old_plane_state: the state from the previous modeset
  *
  * Cleans up a framebuffer that has just been removed from a plane.
- *
- * Must be called with struct_mutex held.
  */
 void
 intel_cleanup_plane_fb(struct drm_plane *plane,
-		       struct drm_plane_state *old_state)
+		       struct drm_plane_state *_old_plane_state)
 {
+	struct intel_plane_state *old_plane_state =
+		to_intel_plane_state(_old_plane_state);
 	struct intel_atomic_state *intel_state =
-		to_intel_atomic_state(old_state->state);
+		to_intel_atomic_state(old_plane_state->base.state);
 	struct drm_i915_private *dev_priv = to_i915(plane->dev);
 
 	if (intel_state->rps_interactive) {
-		intel_rps_mark_interactive(dev_priv, false);
+		intel_rps_mark_interactive(&dev_priv->gt.rps, false);
 		intel_state->rps_interactive = false;
 	}
 
 	/* Should only be called after a successful intel_prepare_plane_fb()! */
-	mutex_lock(&dev_priv->drm.struct_mutex);
-	intel_plane_unpin_fb(to_intel_plane_state(old_state));
-	mutex_unlock(&dev_priv->drm.struct_mutex);
-}
-
-int
-skl_max_scale(const struct intel_crtc_state *crtc_state,
-	      u32 pixel_format)
-{
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
-	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
-	int max_scale, mult;
-	int crtc_clock, max_dotclk, tmpclk1, tmpclk2;
-
-	if (!crtc_state->base.enable)
-		return DRM_PLANE_HELPER_NO_SCALING;
-
-	crtc_clock = crtc_state->base.adjusted_mode.crtc_clock;
-	max_dotclk = to_intel_atomic_state(crtc_state->base.state)->cdclk.logical.cdclk;
-
-	if (IS_GEMINILAKE(dev_priv) || INTEL_GEN(dev_priv) >= 10)
-		max_dotclk *= 2;
-
-	if (WARN_ON_ONCE(!crtc_clock || max_dotclk < crtc_clock))
-		return DRM_PLANE_HELPER_NO_SCALING;
-
-	/*
-	 * skl max scale is lower of:
-	 *    close to 3 but not 3, -1 is for that purpose
-	 *            or
-	 *    cdclk/crtc_clock
-	 */
-	mult = is_planar_yuv_format(pixel_format) ? 2 : 3;
-	tmpclk1 = (1 << 16) * mult - 1;
-	tmpclk2 = (1 << 8) * ((max_dotclk << 8) / crtc_clock);
-	max_scale = min(tmpclk1, tmpclk2);
-
-	return max_scale;
-}
-
-static void intel_begin_crtc_commit(struct intel_atomic_state *state,
-				    struct intel_crtc *crtc)
-{
-	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
-	struct intel_crtc_state *old_crtc_state =
-		intel_atomic_get_old_crtc_state(state, crtc);
-	struct intel_crtc_state *new_crtc_state =
-		intel_atomic_get_new_crtc_state(state, crtc);
-	bool modeset = needs_modeset(new_crtc_state);
-
-	/* Perform vblank evasion around commit operation */
-	intel_pipe_update_start(new_crtc_state);
-
-	if (modeset)
-		goto out;
-
-	if (new_crtc_state->base.color_mgmt_changed ||
-	    new_crtc_state->update_pipe)
-		intel_color_commit(new_crtc_state);
-
-	if (new_crtc_state->update_pipe)
-		intel_update_pipe_config(old_crtc_state, new_crtc_state);
-	else if (INTEL_GEN(dev_priv) >= 9)
-		skl_detach_scalers(new_crtc_state);
-
-	if (INTEL_GEN(dev_priv) >= 9 || IS_BROADWELL(dev_priv))
-		bdw_set_pipemisc(new_crtc_state);
-
-out:
-	if (dev_priv->display.atomic_update_watermarks)
-		dev_priv->display.atomic_update_watermarks(state,
-							   new_crtc_state);
-}
-
-void intel_crtc_arm_fifo_underrun(struct intel_crtc *crtc,
-				  struct intel_crtc_state *crtc_state)
-{
-	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
-
-	if (!IS_GEN(dev_priv, 2))
-		intel_set_cpu_fifo_underrun_reporting(dev_priv, crtc->pipe, true);
-
-	if (crtc_state->has_pch_encoder) {
-		enum pipe pch_transcoder =
-			intel_crtc_pch_transcoder(crtc);
-
-		intel_set_pch_fifo_underrun_reporting(dev_priv, pch_transcoder, true);
-	}
-}
-
-static void intel_finish_crtc_commit(struct intel_atomic_state *state,
-				     struct intel_crtc *crtc)
-{
-	struct intel_crtc_state *old_crtc_state =
-		intel_atomic_get_old_crtc_state(state, crtc);
-	struct intel_crtc_state *new_crtc_state =
-		intel_atomic_get_new_crtc_state(state, crtc);
-
-	intel_pipe_update_end(new_crtc_state);
-
-	if (new_crtc_state->update_pipe &&
-	    !needs_modeset(new_crtc_state) &&
-	    old_crtc_state->base.mode.private_flags & I915_MODE_FLAG_INHERITED)
-		intel_crtc_arm_fifo_underrun(crtc, new_crtc_state);
+	intel_plane_unpin_fb(old_plane_state);
 }
 
 /**
@@ -14622,6 +15231,7 @@ static bool i965_plane_format_mod_supported(struct drm_plane *_plane,
 	case DRM_FORMAT_XBGR8888:
 	case DRM_FORMAT_XRGB2101010:
 	case DRM_FORMAT_XBGR2101010:
+	case DRM_FORMAT_XBGR16161616F:
 		return modifier == DRM_FORMAT_MOD_LINEAR ||
 			modifier == I915_FORMAT_MOD_X_TILED;
 	default:
@@ -14655,8 +15265,8 @@ static const struct drm_plane_funcs i8xx_plane_funcs = {
 };
 
 static int
-intel_legacy_cursor_update(struct drm_plane *plane,
-			   struct drm_crtc *crtc,
+intel_legacy_cursor_update(struct drm_plane *_plane,
+			   struct drm_crtc *_crtc,
 			   struct drm_framebuffer *fb,
 			   int crtc_x, int crtc_y,
 			   unsigned int crtc_w, unsigned int crtc_h,
@@ -14664,11 +15274,13 @@ intel_legacy_cursor_update(struct drm_plane *plane,
 			   u32 src_w, u32 src_h,
 			   struct drm_modeset_acquire_ctx *ctx)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc->dev);
-	struct drm_plane_state *old_plane_state, *new_plane_state;
-	struct intel_plane *intel_plane = to_intel_plane(plane);
+	struct intel_plane *plane = to_intel_plane(_plane);
+	struct intel_crtc *crtc = to_intel_crtc(_crtc);
+	struct intel_plane_state *old_plane_state =
+		to_intel_plane_state(plane->base.state);
+	struct intel_plane_state *new_plane_state;
 	struct intel_crtc_state *crtc_state =
-		to_intel_crtc_state(crtc->state);
+		to_intel_crtc_state(crtc->base.state);
 	struct intel_crtc_state *new_crtc_state;
 	int ret;
 
@@ -14680,14 +15292,13 @@ intel_legacy_cursor_update(struct drm_plane *plane,
 	    crtc_state->update_pipe)
 		goto slow;
 
-	old_plane_state = plane->state;
 	/*
 	 * Don't do an async update if there is an outstanding commit modifying
 	 * the plane.  This prevents our async update's changes from getting
 	 * overridden by a previous synchronous update's state.
 	 */
-	if (old_plane_state->commit &&
-	    !try_wait_for_completion(&old_plane_state->commit->hw_done))
+	if (old_plane_state->base.commit &&
+	    !try_wait_for_completion(&old_plane_state->base.commit->hw_done))
 		goto slow;
 
 	/*
@@ -14695,56 +15306,51 @@ intel_legacy_cursor_update(struct drm_plane *plane,
 	 * take the slowpath. Only changing fb or position should be
 	 * in the fastpath.
 	 */
-	if (old_plane_state->crtc != crtc ||
-	    old_plane_state->src_w != src_w ||
-	    old_plane_state->src_h != src_h ||
-	    old_plane_state->crtc_w != crtc_w ||
-	    old_plane_state->crtc_h != crtc_h ||
-	    !old_plane_state->fb != !fb)
+	if (old_plane_state->base.crtc != &crtc->base ||
+	    old_plane_state->base.src_w != src_w ||
+	    old_plane_state->base.src_h != src_h ||
+	    old_plane_state->base.crtc_w != crtc_w ||
+	    old_plane_state->base.crtc_h != crtc_h ||
+	    !old_plane_state->base.fb != !fb)
 		goto slow;
 
-	new_plane_state = intel_plane_duplicate_state(plane);
+	new_plane_state = to_intel_plane_state(intel_plane_duplicate_state(&plane->base));
 	if (!new_plane_state)
 		return -ENOMEM;
 
-	new_crtc_state = to_intel_crtc_state(intel_crtc_duplicate_state(crtc));
+	new_crtc_state = to_intel_crtc_state(intel_crtc_duplicate_state(&crtc->base));
 	if (!new_crtc_state) {
 		ret = -ENOMEM;
 		goto out_free;
 	}
 
-	drm_atomic_set_fb_for_plane(new_plane_state, fb);
+	drm_atomic_set_fb_for_plane(&new_plane_state->base, fb);
 
-	new_plane_state->src_x = src_x;
-	new_plane_state->src_y = src_y;
-	new_plane_state->src_w = src_w;
-	new_plane_state->src_h = src_h;
-	new_plane_state->crtc_x = crtc_x;
-	new_plane_state->crtc_y = crtc_y;
-	new_plane_state->crtc_w = crtc_w;
-	new_plane_state->crtc_h = crtc_h;
+	new_plane_state->base.src_x = src_x;
+	new_plane_state->base.src_y = src_y;
+	new_plane_state->base.src_w = src_w;
+	new_plane_state->base.src_h = src_h;
+	new_plane_state->base.crtc_x = crtc_x;
+	new_plane_state->base.crtc_y = crtc_y;
+	new_plane_state->base.crtc_w = crtc_w;
+	new_plane_state->base.crtc_h = crtc_h;
 
 	ret = intel_plane_atomic_check_with_state(crtc_state, new_crtc_state,
-						  to_intel_plane_state(old_plane_state),
-						  to_intel_plane_state(new_plane_state));
+						  old_plane_state, new_plane_state);
 	if (ret)
 		goto out_free;
 
-	ret = mutex_lock_interruptible(&dev_priv->drm.struct_mutex);
+	ret = intel_plane_pin_fb(new_plane_state);
 	if (ret)
 		goto out_free;
 
-	ret = intel_plane_pin_fb(to_intel_plane_state(new_plane_state));
-	if (ret)
-		goto out_unlock;
-
-	intel_frontbuffer_flush(to_intel_frontbuffer(fb), ORIGIN_FLIP);
-	intel_frontbuffer_track(to_intel_frontbuffer(old_plane_state->fb),
-				to_intel_frontbuffer(fb),
-				intel_plane->frontbuffer_bit);
+	intel_frontbuffer_flush(to_intel_frontbuffer(new_plane_state->base.fb), ORIGIN_FLIP);
+	intel_frontbuffer_track(to_intel_frontbuffer(old_plane_state->base.fb),
+				to_intel_frontbuffer(new_plane_state->base.fb),
+				plane->frontbuffer_bit);
 
 	/* Swap plane state */
-	plane->state = new_plane_state;
+	plane->base.state = &new_plane_state->base;
 
 	/*
 	 * We cannot swap crtc_state as it may be in use by an atomic commit or
@@ -14758,27 +15364,24 @@ intel_legacy_cursor_update(struct drm_plane *plane,
 	 */
 	crtc_state->active_planes = new_crtc_state->active_planes;
 
-	if (plane->state->visible)
-		intel_update_plane(intel_plane, crtc_state,
-				   to_intel_plane_state(plane->state));
+	if (new_plane_state->base.visible)
+		intel_update_plane(plane, crtc_state, new_plane_state);
 	else
-		intel_disable_plane(intel_plane, crtc_state);
+		intel_disable_plane(plane, crtc_state);
 
-	intel_plane_unpin_fb(to_intel_plane_state(old_plane_state));
+	intel_plane_unpin_fb(old_plane_state);
 
-out_unlock:
-	mutex_unlock(&dev_priv->drm.struct_mutex);
 out_free:
 	if (new_crtc_state)
-		intel_crtc_destroy_state(crtc, &new_crtc_state->base);
+		intel_crtc_destroy_state(&crtc->base, &new_crtc_state->base);
 	if (ret)
-		intel_plane_destroy_state(plane, new_plane_state);
+		intel_plane_destroy_state(&plane->base, &new_plane_state->base);
 	else
-		intel_plane_destroy_state(plane, old_plane_state);
+		intel_plane_destroy_state(&plane->base, &old_plane_state->base);
 	return ret;
 
 slow:
-	return drm_atomic_helper_update_plane(plane, crtc, fb,
+	return drm_atomic_helper_update_plane(&plane->base, &crtc->base, fb,
 					      crtc_x, crtc_y, crtc_w, crtc_h,
 					      src_x, src_y, src_w, src_h, ctx);
 }
@@ -14819,7 +15422,7 @@ intel_primary_plane_create(struct drm_i915_private *dev_priv, enum pipe pipe)
 	const u64 *modifiers;
 	const u32 *formats;
 	int num_formats;
-	int ret;
+	int ret, zpos;
 
 	if (INTEL_GEN(dev_priv) >= 9)
 		return skl_universal_plane_create(dev_priv, pipe,
@@ -14849,8 +15452,26 @@ intel_primary_plane_create(struct drm_i915_private *dev_priv, enum pipe pipe)
 	}
 
 	if (INTEL_GEN(dev_priv) >= 4) {
-		formats = i965_primary_formats;
-		num_formats = ARRAY_SIZE(i965_primary_formats);
+		/*
+		 * WaFP16GammaEnabling:ivb
+		 * "Workaround : When using the 64-bit format, the plane
+		 *  output on each color channel has one quarter amplitude.
+		 *  It can be brought up to full amplitude by using pipe
+		 *  gamma correction or pipe color space conversion to
+		 *  multiply the plane output by four."
+		 *
+		 * There is no dedicated plane gamma for the primary plane,
+		 * and using the pipe gamma/csc could conflict with other
+		 * planes, so we choose not to expose fp16 on IVB primary
+		 * planes. HSW primary planes no longer have this problem.
+		 */
+		if (IS_IVYBRIDGE(dev_priv)) {
+			formats = ivb_primary_formats;
+			num_formats = ARRAY_SIZE(ivb_primary_formats);
+		} else {
+			formats = i965_primary_formats;
+			num_formats = ARRAY_SIZE(i965_primary_formats);
+		}
 		modifiers = i9xx_format_modifiers;
 
 		plane->max_stride = i9xx_plane_max_stride;
@@ -14859,6 +15480,15 @@ intel_primary_plane_create(struct drm_i915_private *dev_priv, enum pipe pipe)
 		plane->get_hw_state = i9xx_plane_get_hw_state;
 		plane->check_plane = i9xx_plane_check;
 
+		if (IS_BROADWELL(dev_priv) || IS_HASWELL(dev_priv))
+			plane->min_cdclk = hsw_plane_min_cdclk;
+		else if (IS_IVYBRIDGE(dev_priv))
+			plane->min_cdclk = ivb_plane_min_cdclk;
+		else if (IS_CHERRYVIEW(dev_priv) || IS_VALLEYVIEW(dev_priv))
+			plane->min_cdclk = vlv_plane_min_cdclk;
+		else
+			plane->min_cdclk = i9xx_plane_min_cdclk;
+
 		plane_funcs = &i965_plane_funcs;
 	} else {
 		formats = i8xx_primary_formats;
@@ -14870,6 +15500,7 @@ intel_primary_plane_create(struct drm_i915_private *dev_priv, enum pipe pipe)
 		plane->disable_plane = i9xx_disable_plane;
 		plane->get_hw_state = i9xx_plane_get_hw_state;
 		plane->check_plane = i9xx_plane_check;
+		plane->min_cdclk = i9xx_plane_min_cdclk;
 
 		plane_funcs = &i8xx_plane_funcs;
 	}
@@ -14908,6 +15539,9 @@ intel_primary_plane_create(struct drm_i915_private *dev_priv, enum pipe pipe)
 						   DRM_MODE_ROTATE_0,
 						   supported_rotations);
 
+	zpos = 0;
+	drm_plane_create_zpos_immutable_property(&plane->base, zpos);
+
 	drm_plane_helper_add(&plane->base, &intel_plane_helper_funcs);
 
 	return plane;
@@ -14924,7 +15558,7 @@ intel_cursor_plane_create(struct drm_i915_private *dev_priv,
 {
 	unsigned int possible_crtcs;
 	struct intel_plane *cursor;
-	int ret;
+	int ret, zpos;
 
 	cursor = intel_plane_alloc();
 	if (IS_ERR(cursor))
@@ -14973,6 +15607,9 @@ intel_cursor_plane_create(struct drm_i915_private *dev_priv,
 						   DRM_MODE_ROTATE_0 |
 						   DRM_MODE_ROTATE_180);
 
+	zpos = RUNTIME_INFO(dev_priv)->num_sprites[pipe] + 1;
+	drm_plane_create_zpos_immutable_property(&cursor->base, zpos);
+
 	drm_plane_helper_add(&cursor->base, &intel_plane_helper_funcs);
 
 	return cursor;
@@ -15048,12 +15685,12 @@ static const struct drm_crtc_funcs i965_crtc_funcs = {
 	.disable_vblank = i965_disable_vblank,
 };
 
-static const struct drm_crtc_funcs i945gm_crtc_funcs = {
+static const struct drm_crtc_funcs i915gm_crtc_funcs = {
 	INTEL_CRTC_FUNCS,
 
 	.get_vblank_counter = i915_get_vblank_counter,
-	.enable_vblank = i945gm_enable_vblank,
-	.disable_vblank = i945gm_disable_vblank,
+	.enable_vblank = i915gm_enable_vblank,
+	.disable_vblank = i915gm_disable_vblank,
 };
 
 static const struct drm_crtc_funcs i915_crtc_funcs = {
@@ -15124,8 +15761,8 @@ static int intel_crtc_init(struct drm_i915_private *dev_priv, enum pipe pipe)
 			funcs = &g4x_crtc_funcs;
 		else if (IS_GEN(dev_priv, 4))
 			funcs = &i965_crtc_funcs;
-		else if (IS_I945GM(dev_priv))
-			funcs = &i945gm_crtc_funcs;
+		else if (IS_I945GM(dev_priv) || IS_I915GM(dev_priv))
+			funcs = &i915gm_crtc_funcs;
 		else if (IS_GEN(dev_priv, 3))
 			funcs = &i915_crtc_funcs;
 		else
@@ -15160,8 +15797,6 @@ static int intel_crtc_init(struct drm_i915_private *dev_priv, enum pipe pipe)
 		dev_priv->plane_to_crtc_mapping[i9xx_plane] = intel_crtc;
 	}
 
-	drm_crtc_helper_add(&intel_crtc->base, &intel_helper_funcs);
-
 	intel_color_init(intel_crtc);
 
 	WARN_ON(drm_crtc_index(&intel_crtc->base) != intel_crtc->pipe);
@@ -15196,21 +15831,32 @@ int intel_get_pipe_from_crtc_id_ioctl(struct drm_device *dev, void *data,
 	return 0;
 }
 
-static int intel_encoder_clones(struct intel_encoder *encoder)
+static u32 intel_encoder_possible_clones(struct intel_encoder *encoder)
 {
 	struct drm_device *dev = encoder->base.dev;
 	struct intel_encoder *source_encoder;
-	int index_mask = 0;
-	int entry = 0;
+	u32 possible_clones = 0;
 
 	for_each_intel_encoder(dev, source_encoder) {
 		if (encoders_cloneable(encoder, source_encoder))
-			index_mask |= (1 << entry);
+			possible_clones |= drm_encoder_mask(&source_encoder->base);
+	}
+
+	return possible_clones;
+}
+
+static u32 intel_encoder_possible_crtcs(struct intel_encoder *encoder)
+{
+	struct drm_device *dev = encoder->base.dev;
+	struct intel_crtc *crtc;
+	u32 possible_crtcs = 0;
 
-		entry++;
+	for_each_intel_crtc(dev, crtc) {
+		if (encoder->pipe_mask & BIT(crtc->pipe))
+			possible_crtcs |= drm_crtc_mask(&crtc->base);
 	}
 
-	return index_mask;
+	return possible_crtcs;
 }
 
 static bool ilk_has_edp_a(struct drm_i915_private *dev_priv)
@@ -15292,13 +15938,18 @@ static void intel_setup_outputs(struct drm_i915_private *dev_priv)
 
 	intel_pps_init(dev_priv);
 
-	if (!HAS_DISPLAY(dev_priv))
+	if (!HAS_DISPLAY(dev_priv) || !INTEL_DISPLAY_ENABLED(dev_priv))
 		return;
 
 	if (INTEL_GEN(dev_priv) >= 12) {
-		/* TODO: initialize TC ports as well */
 		intel_ddi_init(dev_priv, PORT_A);
 		intel_ddi_init(dev_priv, PORT_B);
+		intel_ddi_init(dev_priv, PORT_D);
+		intel_ddi_init(dev_priv, PORT_E);
+		intel_ddi_init(dev_priv, PORT_F);
+		intel_ddi_init(dev_priv, PORT_G);
+		intel_ddi_init(dev_priv, PORT_H);
+		intel_ddi_init(dev_priv, PORT_I);
 		icl_dsi_init(dev_priv);
 	} else if (IS_ELKHARTLAKE(dev_priv)) {
 		intel_ddi_init(dev_priv, PORT_A);
@@ -15508,9 +16159,10 @@ static void intel_setup_outputs(struct drm_i915_private *dev_priv)
 	intel_psr_init(dev_priv);
 
 	for_each_intel_encoder(&dev_priv->drm, encoder) {
-		encoder->base.possible_crtcs = encoder->crtc_mask;
+		encoder->base.possible_crtcs =
+			intel_encoder_possible_crtcs(encoder);
 		encoder->base.possible_clones =
-			intel_encoder_clones(encoder);
+			intel_encoder_possible_clones(encoder);
 	}
 
 	intel_init_pch_refclk(dev_priv);
@@ -15765,8 +16417,14 @@ intel_mode_valid(struct drm_device *dev,
 			   DRM_MODE_FLAG_CLKDIV2))
 		return MODE_BAD;
 
-	if (INTEL_GEN(dev_priv) >= 9 ||
-	    IS_BROADWELL(dev_priv) || IS_HASWELL(dev_priv)) {
+	/* Transcoder timing limits */
+	if (INTEL_GEN(dev_priv) >= 11) {
+		hdisplay_max = 16384;
+		vdisplay_max = 8192;
+		htotal_max = 16384;
+		vtotal_max = 8192;
+	} else if (INTEL_GEN(dev_priv) >= 9 ||
+		   IS_BROADWELL(dev_priv) || IS_HASWELL(dev_priv)) {
 		hdisplay_max = 8192; /* FDI max 4096 handled elsewhere */
 		vdisplay_max = 4096;
 		htotal_max = 8192;
@@ -15795,6 +16453,56 @@ intel_mode_valid(struct drm_device *dev,
 	    mode->vtotal > vtotal_max)
 		return MODE_V_ILLEGAL;
 
+	if (INTEL_GEN(dev_priv) >= 5) {
+		if (mode->hdisplay < 64 ||
+		    mode->htotal - mode->hdisplay < 32)
+			return MODE_H_ILLEGAL;
+
+		if (mode->vtotal - mode->vdisplay < 5)
+			return MODE_V_ILLEGAL;
+	} else {
+		if (mode->htotal - mode->hdisplay < 32)
+			return MODE_H_ILLEGAL;
+
+		if (mode->vtotal - mode->vdisplay < 3)
+			return MODE_V_ILLEGAL;
+	}
+
+	return MODE_OK;
+}
+
+enum drm_mode_status
+intel_mode_valid_max_plane_size(struct drm_i915_private *dev_priv,
+				const struct drm_display_mode *mode)
+{
+	int plane_width_max, plane_height_max;
+
+	/*
+	 * intel_mode_valid() should be
+	 * sufficient on older platforms.
+	 */
+	if (INTEL_GEN(dev_priv) < 9)
+		return MODE_OK;
+
+	/*
+	 * Most people will probably want a fullscreen
+	 * plane so let's not advertize modes that are
+	 * too big for that.
+	 */
+	if (INTEL_GEN(dev_priv) >= 11) {
+		plane_width_max = 5120;
+		plane_height_max = 4320;
+	} else {
+		plane_width_max = 5120;
+		plane_height_max = 4096;
+	}
+
+	if (mode->hdisplay > plane_width_max)
+		return MODE_H_ILLEGAL;
+
+	if (mode->vdisplay > plane_height_max)
+		return MODE_V_ILLEGAL;
+
 	return MODE_OK;
 }
 
@@ -15898,47 +16606,17 @@ void intel_init_display_hooks(struct drm_i915_private *dev_priv)
 	}
 
 	if (INTEL_GEN(dev_priv) >= 9)
-		dev_priv->display.update_crtcs = skl_update_crtcs;
+		dev_priv->display.commit_modeset_enables = skl_commit_modeset_enables;
 	else
-		dev_priv->display.update_crtcs = intel_update_crtcs;
-}
+		dev_priv->display.commit_modeset_enables = intel_commit_modeset_enables;
 
-static i915_reg_t i915_vgacntrl_reg(struct drm_i915_private *dev_priv)
-{
-	if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv))
-		return VLV_VGACNTRL;
-	else if (INTEL_GEN(dev_priv) >= 5)
-		return CPU_VGACNTRL;
-	else
-		return VGACNTRL;
 }
 
-/* Disable the VGA plane that we never use */
-static void i915_disable_vga(struct drm_i915_private *dev_priv)
+void intel_modeset_init_hw(struct drm_i915_private *i915)
 {
-	struct pci_dev *pdev = dev_priv->drm.pdev;
-	u8 sr1;
-	i915_reg_t vga_reg = i915_vgacntrl_reg(dev_priv);
-
-	/* WaEnableVGAAccessThroughIOPort:ctg,elk,ilk,snb,ivb,vlv,hsw */
-	vga_get_uninterruptible(pdev, VGA_RSRC_LEGACY_IO);
-	outb(SR01, VGA_SR_INDEX);
-	sr1 = inb(VGA_SR_DATA);
-	outb(sr1 | 1<<5, VGA_SR_DATA);
-	vga_put(pdev, VGA_RSRC_LEGACY_IO);
-	udelay(300);
-
-	I915_WRITE(vga_reg, VGA_DISP_DISABLE);
-	POSTING_READ(vga_reg);
-}
-
-void intel_modeset_init_hw(struct drm_device *dev)
-{
-	struct drm_i915_private *dev_priv = to_i915(dev);
-
-	intel_update_cdclk(dev_priv);
-	intel_dump_cdclk_state(&dev_priv->cdclk.hw, "Current CDCLK");
-	dev_priv->cdclk.logical = dev_priv->cdclk.actual = dev_priv->cdclk.hw;
+	intel_update_cdclk(i915);
+	intel_dump_cdclk_state(&i915->cdclk.hw, "Current CDCLK");
+	i915->cdclk.logical = i915->cdclk.actual = i915->cdclk.hw;
 }
 
 /*
@@ -16098,114 +16776,111 @@ out:
 	return ret;
 }
 
-int intel_modeset_init(struct drm_device *dev)
+static void intel_mode_config_init(struct drm_i915_private *i915)
 {
-	struct drm_i915_private *dev_priv = to_i915(dev);
-	enum pipe pipe;
-	struct intel_crtc *crtc;
-	int ret;
-
-	dev_priv->modeset_wq = alloc_ordered_workqueue("i915_modeset", 0);
-
-	drm_mode_config_init(dev);
-
-	ret = intel_bw_init(dev_priv);
-	if (ret)
-		return ret;
-
-	dev->mode_config.min_width = 0;
-	dev->mode_config.min_height = 0;
+	struct drm_mode_config *mode_config = &i915->drm.mode_config;
 
-	dev->mode_config.preferred_depth = 24;
-	dev->mode_config.prefer_shadow = 1;
+	drm_mode_config_init(&i915->drm);
 
-	dev->mode_config.allow_fb_modifiers = true;
+	mode_config->min_width = 0;
+	mode_config->min_height = 0;
 
-	dev->mode_config.funcs = &intel_mode_funcs;
+	mode_config->preferred_depth = 24;
+	mode_config->prefer_shadow = 1;
 
-	init_llist_head(&dev_priv->atomic_helper.free_list);
-	INIT_WORK(&dev_priv->atomic_helper.free_work,
-		  intel_atomic_helper_free_state_worker);
-
-	intel_init_quirks(dev_priv);
+	mode_config->allow_fb_modifiers = true;
 
-	intel_fbc_init(dev_priv);
-
-	intel_init_pm(dev_priv);
-
-	/*
-	 * There may be no VBT; and if the BIOS enabled SSC we can
-	 * just keep using it to avoid unnecessary flicker.  Whereas if the
-	 * BIOS isn't using it, don't assume it will work even if the VBT
-	 * indicates as much.
-	 */
-	if (HAS_PCH_IBX(dev_priv) || HAS_PCH_CPT(dev_priv)) {
-		bool bios_lvds_use_ssc = !!(I915_READ(PCH_DREF_CONTROL) &
-					    DREF_SSC1_ENABLE);
-
-		if (dev_priv->vbt.lvds_use_ssc != bios_lvds_use_ssc) {
-			DRM_DEBUG_KMS("SSC %sabled by BIOS, overriding VBT which says %sabled\n",
-				     bios_lvds_use_ssc ? "en" : "dis",
-				     dev_priv->vbt.lvds_use_ssc ? "en" : "dis");
-			dev_priv->vbt.lvds_use_ssc = bios_lvds_use_ssc;
-		}
-	}
+	mode_config->funcs = &intel_mode_funcs;
 
 	/*
 	 * Maximum framebuffer dimensions, chosen to match
 	 * the maximum render engine surface size on gen4+.
 	 */
-	if (INTEL_GEN(dev_priv) >= 7) {
-		dev->mode_config.max_width = 16384;
-		dev->mode_config.max_height = 16384;
-	} else if (INTEL_GEN(dev_priv) >= 4) {
-		dev->mode_config.max_width = 8192;
-		dev->mode_config.max_height = 8192;
-	} else if (IS_GEN(dev_priv, 3)) {
-		dev->mode_config.max_width = 4096;
-		dev->mode_config.max_height = 4096;
+	if (INTEL_GEN(i915) >= 7) {
+		mode_config->max_width = 16384;
+		mode_config->max_height = 16384;
+	} else if (INTEL_GEN(i915) >= 4) {
+		mode_config->max_width = 8192;
+		mode_config->max_height = 8192;
+	} else if (IS_GEN(i915, 3)) {
+		mode_config->max_width = 4096;
+		mode_config->max_height = 4096;
 	} else {
-		dev->mode_config.max_width = 2048;
-		dev->mode_config.max_height = 2048;
+		mode_config->max_width = 2048;
+		mode_config->max_height = 2048;
 	}
 
-	if (IS_I845G(dev_priv) || IS_I865G(dev_priv)) {
-		dev->mode_config.cursor_width = IS_I845G(dev_priv) ? 64 : 512;
-		dev->mode_config.cursor_height = 1023;
-	} else if (IS_GEN(dev_priv, 2)) {
-		dev->mode_config.cursor_width = 64;
-		dev->mode_config.cursor_height = 64;
+	if (IS_I845G(i915) || IS_I865G(i915)) {
+		mode_config->cursor_width = IS_I845G(i915) ? 64 : 512;
+		mode_config->cursor_height = 1023;
+	} else if (IS_GEN(i915, 2)) {
+		mode_config->cursor_width = 64;
+		mode_config->cursor_height = 64;
 	} else {
-		dev->mode_config.cursor_width = 256;
-		dev->mode_config.cursor_height = 256;
+		mode_config->cursor_width = 256;
+		mode_config->cursor_height = 256;
 	}
+}
 
-	DRM_DEBUG_KMS("%d display pipe%s available.\n",
-		      INTEL_INFO(dev_priv)->num_pipes,
-		      INTEL_INFO(dev_priv)->num_pipes > 1 ? "s" : "");
+int intel_modeset_init(struct drm_i915_private *i915)
+{
+	struct drm_device *dev = &i915->drm;
+	enum pipe pipe;
+	struct intel_crtc *crtc;
+	int ret;
 
-	for_each_pipe(dev_priv, pipe) {
-		ret = intel_crtc_init(dev_priv, pipe);
-		if (ret) {
-			drm_mode_config_cleanup(dev);
-			return ret;
+	i915->modeset_wq = alloc_ordered_workqueue("i915_modeset", 0);
+	i915->flip_wq = alloc_workqueue("i915_flip", WQ_HIGHPRI |
+					WQ_UNBOUND, WQ_UNBOUND_MAX_ACTIVE);
+
+	intel_mode_config_init(i915);
+
+	ret = intel_bw_init(i915);
+	if (ret)
+		return ret;
+
+	init_llist_head(&i915->atomic_helper.free_list);
+	INIT_WORK(&i915->atomic_helper.free_work,
+		  intel_atomic_helper_free_state_worker);
+
+	intel_init_quirks(i915);
+
+	intel_fbc_init(i915);
+
+	intel_init_pm(i915);
+
+	intel_panel_sanitize_ssc(i915);
+
+	intel_gmbus_setup(i915);
+
+	DRM_DEBUG_KMS("%d display pipe%s available.\n",
+		      INTEL_NUM_PIPES(i915),
+		      INTEL_NUM_PIPES(i915) > 1 ? "s" : "");
+
+	if (HAS_DISPLAY(i915) && INTEL_DISPLAY_ENABLED(i915)) {
+		for_each_pipe(i915, pipe) {
+			ret = intel_crtc_init(i915, pipe);
+			if (ret) {
+				drm_mode_config_cleanup(dev);
+				return ret;
+			}
 		}
 	}
 
 	intel_shared_dpll_init(dev);
-	intel_update_fdi_pll_freq(dev_priv);
+	intel_update_fdi_pll_freq(i915);
 
-	intel_update_czclk(dev_priv);
-	intel_modeset_init_hw(dev);
+	intel_update_czclk(i915);
+	intel_modeset_init_hw(i915);
 
-	intel_hdcp_component_init(dev_priv);
+	intel_hdcp_component_init(i915);
 
-	if (dev_priv->max_cdclk_freq == 0)
-		intel_update_max_cdclk(dev_priv);
+	if (i915->max_cdclk_freq == 0)
+		intel_update_max_cdclk(i915);
 
 	/* Just disable it once at startup */
-	i915_disable_vga(dev_priv);
-	intel_setup_outputs(dev_priv);
+	intel_vga_disable(i915);
+	intel_setup_outputs(i915);
 
 	drm_modeset_lock_all(dev);
 	intel_modeset_setup_hw_state(dev, dev->mode_config.acquire_ctx);
@@ -16224,8 +16899,7 @@ int intel_modeset_init(struct drm_device *dev)
 		 * can even allow for smooth boot transitions if the BIOS
 		 * fb is large enough for the active pipe configuration.
 		 */
-		dev_priv->display.get_initial_plane_config(crtc,
-							   &plane_config);
+		i915->display.get_initial_plane_config(crtc, &plane_config);
 
 		/*
 		 * If the fb is shared between multiple heads, we'll
@@ -16239,7 +16913,7 @@ int intel_modeset_init(struct drm_device *dev)
 	 * Note that we need to do this after reconstructing the BIOS fb's
 	 * since the watermark calculation done here will use pstate->fb.
 	 */
-	if (!HAS_GMCH(dev_priv))
+	if (!HAS_GMCH(i915))
 		sanitize_watermarks(dev);
 
 	/*
@@ -16564,39 +17238,6 @@ static void intel_sanitize_encoder(struct intel_encoder *encoder)
 		icl_sanitize_encoder_pll_mapping(encoder);
 }
 
-void i915_redisable_vga_power_on(struct drm_i915_private *dev_priv)
-{
-	i915_reg_t vga_reg = i915_vgacntrl_reg(dev_priv);
-
-	if (!(I915_READ(vga_reg) & VGA_DISP_DISABLE)) {
-		DRM_DEBUG_KMS("Something enabled VGA plane, disabling it\n");
-		i915_disable_vga(dev_priv);
-	}
-}
-
-void i915_redisable_vga(struct drm_i915_private *dev_priv)
-{
-	intel_wakeref_t wakeref;
-
-	/*
-	 * This function can be called both from intel_modeset_setup_hw_state or
-	 * at a very early point in our resume sequence, where the power well
-	 * structures are not yet restored. Since this function is at a very
-	 * paranoid "someone might have enabled VGA while we were not looking"
-	 * level, just check if the power well is enabled instead of trying to
-	 * follow the "don't touch the power well if we don't need it" policy
-	 * the rest of the driver uses.
-	 */
-	wakeref = intel_display_power_get_if_enabled(dev_priv,
-						     POWER_DOMAIN_VGA);
-	if (!wakeref)
-		return;
-
-	i915_redisable_vga_power_on(dev_priv);
-
-	intel_display_power_put(dev_priv, POWER_DOMAIN_VGA, wakeref);
-}
-
 /* FIXME read out full plane state for all planes */
 static void readout_plane_state(struct drm_i915_private *dev_priv)
 {
@@ -16640,7 +17281,7 @@ static void intel_modeset_readout_hw_state(struct drm_device *dev)
 	struct drm_connector_list_iter conn_iter;
 	int i;
 
-	dev_priv->active_crtcs = 0;
+	dev_priv->active_pipes = 0;
 
 	for_each_intel_crtc(dev, crtc) {
 		struct intel_crtc_state *crtc_state =
@@ -16657,7 +17298,7 @@ static void intel_modeset_readout_hw_state(struct drm_device *dev)
 		crtc->active = crtc_state->base.active;
 
 		if (crtc_state->base.active)
-			dev_priv->active_crtcs |= 1 << crtc->pipe;
+			dev_priv->active_pipes |= BIT(crtc->pipe);
 
 		DRM_DEBUG_KMS("[CRTC:%d:%s] hw state readout: %s\n",
 			      crtc->base.base.id, crtc->base.name,
@@ -16717,24 +17358,28 @@ static void intel_modeset_readout_hw_state(struct drm_device *dev)
 	drm_connector_list_iter_begin(dev, &conn_iter);
 	for_each_intel_connector_iter(connector, &conn_iter) {
 		if (connector->get_hw_state(connector)) {
+			struct intel_crtc_state *crtc_state;
+			struct intel_crtc *crtc;
+
 			connector->base.dpms = DRM_MODE_DPMS_ON;
 
 			encoder = connector->encoder;
 			connector->base.encoder = &encoder->base;
 
-			if (encoder->base.crtc &&
-			    encoder->base.crtc->state->active) {
+			crtc = to_intel_crtc(encoder->base.crtc);
+			crtc_state = crtc ? to_intel_crtc_state(crtc->base.state) : NULL;
+
+			if (crtc_state && crtc_state->base.active) {
 				/*
 				 * This has to be done during hardware readout
 				 * because anything calling .crtc_disable may
 				 * rely on the connector_mask being accurate.
 				 */
-				encoder->base.crtc->state->connector_mask |=
+				crtc_state->base.connector_mask |=
 					drm_connector_mask(&connector->base);
-				encoder->base.crtc->state->encoder_mask |=
+				crtc_state->base.encoder_mask |=
 					drm_encoder_mask(&encoder->base);
 			}
-
 		} else {
 			connector->base.dpms = DRM_MODE_DPMS_OFF;
 			connector->base.encoder = NULL;
@@ -16753,13 +17398,16 @@ static void intel_modeset_readout_hw_state(struct drm_device *dev)
 		struct intel_plane *plane;
 		int min_cdclk = 0;
 
-		memset(&crtc->base.mode, 0, sizeof(crtc->base.mode));
 		if (crtc_state->base.active) {
-			intel_mode_from_pipe_config(&crtc->base.mode, crtc_state);
-			crtc->base.mode.hdisplay = crtc_state->pipe_src_w;
-			crtc->base.mode.vdisplay = crtc_state->pipe_src_h;
-			intel_mode_from_pipe_config(&crtc_state->base.adjusted_mode, crtc_state);
-			WARN_ON(drm_atomic_set_mode_for_crtc(crtc->base.state, &crtc->base.mode));
+			struct drm_display_mode mode;
+
+			intel_mode_from_pipe_config(&crtc_state->base.adjusted_mode,
+						    crtc_state);
+
+			mode = crtc_state->base.adjusted_mode;
+			mode.hdisplay = crtc_state->pipe_src_w;
+			mode.vdisplay = crtc_state->pipe_src_h;
+			WARN_ON(drm_atomic_set_mode_for_crtc(&crtc_state->base, &mode));
 
 			/*
 			 * The initial mode needs to be set in order to keep
@@ -16774,21 +17422,9 @@ static void intel_modeset_readout_hw_state(struct drm_device *dev)
 
 			intel_crtc_compute_pixel_rate(crtc_state);
 
-			if (dev_priv->display.modeset_calc_cdclk) {
-				min_cdclk = intel_crtc_compute_min_cdclk(crtc_state);
-				if (WARN_ON(min_cdclk < 0))
-					min_cdclk = 0;
-			}
-
-			drm_calc_timestamping_constants(&crtc->base,
-							&crtc_state->base.adjusted_mode);
-			update_scanline_offset(crtc_state);
+			intel_crtc_update_active_timings(crtc_state);
 		}
 
-		dev_priv->min_cdclk[crtc->pipe] = min_cdclk;
-		dev_priv->min_voltage_level[crtc->pipe] =
-			crtc_state->min_voltage_level;
-
 		for_each_intel_plane_on_crtc(&dev_priv->drm, crtc, plane) {
 			const struct intel_plane_state *plane_state =
 				to_intel_plane_state(plane->base.state);
@@ -16800,8 +17436,34 @@ static void intel_modeset_readout_hw_state(struct drm_device *dev)
 			if (plane_state->base.visible)
 				crtc_state->data_rate[plane->id] =
 					4 * crtc_state->pixel_rate;
+			/*
+			 * FIXME don't have the fb yet, so can't
+			 * use plane->min_cdclk() :(
+			 */
+			if (plane_state->base.visible && plane->min_cdclk) {
+				if (crtc_state->double_wide ||
+				    INTEL_GEN(dev_priv) >= 10 || IS_GEMINILAKE(dev_priv))
+					crtc_state->min_cdclk[plane->id] =
+						DIV_ROUND_UP(crtc_state->pixel_rate, 2);
+				else
+					crtc_state->min_cdclk[plane->id] =
+						crtc_state->pixel_rate;
+			}
+			DRM_DEBUG_KMS("[PLANE:%d:%s] min_cdclk %d kHz\n",
+				      plane->base.base.id, plane->base.name,
+				      crtc_state->min_cdclk[plane->id]);
+		}
+
+		if (crtc_state->base.active) {
+			min_cdclk = intel_crtc_compute_min_cdclk(crtc_state);
+			if (WARN_ON(min_cdclk < 0))
+				min_cdclk = 0;
 		}
 
+		dev_priv->min_cdclk[crtc->pipe] = min_cdclk;
+		dev_priv->min_voltage_level[crtc->pipe] =
+			crtc_state->min_voltage_level;
+
 		intel_bw_crtc_update(bw_state, crtc_state);
 
 		intel_pipe_config_sanity_check(dev_priv, crtc_state);
@@ -17042,13 +17704,13 @@ void intel_display_resume(struct drm_device *dev)
 		drm_atomic_state_put(state);
 }
 
-static void intel_hpd_poll_fini(struct drm_device *dev)
+static void intel_hpd_poll_fini(struct drm_i915_private *i915)
 {
 	struct intel_connector *connector;
 	struct drm_connector_list_iter conn_iter;
 
 	/* Kill all the work that may have been queued by hpd. */
-	drm_connector_list_iter_begin(dev, &conn_iter);
+	drm_connector_list_iter_begin(&i915->drm, &conn_iter);
 	for_each_intel_connector_iter(connector, &conn_iter) {
 		if (connector->modeset_retry_work.func)
 			cancel_work_sync(&connector->modeset_retry_work);
@@ -17060,78 +17722,49 @@ static void intel_hpd_poll_fini(struct drm_device *dev)
 	drm_connector_list_iter_end(&conn_iter);
 }
 
-void intel_modeset_driver_remove(struct drm_device *dev)
+void intel_modeset_driver_remove(struct drm_i915_private *i915)
 {
-	struct drm_i915_private *dev_priv = to_i915(dev);
+	flush_workqueue(i915->flip_wq);
+	flush_workqueue(i915->modeset_wq);
 
-	flush_workqueue(dev_priv->modeset_wq);
-
-	flush_work(&dev_priv->atomic_helper.free_work);
-	WARN_ON(!llist_empty(&dev_priv->atomic_helper.free_list));
+	flush_work(&i915->atomic_helper.free_work);
+	WARN_ON(!llist_empty(&i915->atomic_helper.free_list));
 
 	/*
 	 * Interrupts and polling as the first thing to avoid creating havoc.
 	 * Too much stuff here (turning of connectors, ...) would
 	 * experience fancy races otherwise.
 	 */
-	intel_irq_uninstall(dev_priv);
+	intel_irq_uninstall(i915);
 
 	/*
 	 * Due to the hpd irq storm handling the hotplug work can re-arm the
 	 * poll handlers. Hence disable polling after hpd handling is shut down.
 	 */
-	intel_hpd_poll_fini(dev);
+	intel_hpd_poll_fini(i915);
 
 	/* poll work can call into fbdev, hence clean that up afterwards */
-	intel_fbdev_fini(dev_priv);
+	intel_fbdev_fini(i915);
 
 	intel_unregister_dsm_handler();
 
-	intel_fbc_global_disable(dev_priv);
+	intel_fbc_global_disable(i915);
 
 	/* flush any delayed tasks or pending work */
 	flush_scheduled_work();
 
-	intel_hdcp_component_fini(dev_priv);
-
-	drm_mode_config_cleanup(dev);
+	intel_hdcp_component_fini(i915);
 
-	intel_overlay_cleanup(dev_priv);
+	drm_mode_config_cleanup(&i915->drm);
 
-	intel_gmbus_teardown(dev_priv);
+	intel_overlay_cleanup(i915);
 
-	destroy_workqueue(dev_priv->modeset_wq);
+	intel_gmbus_teardown(i915);
 
-	intel_fbc_cleanup_cfb(dev_priv);
-}
+	destroy_workqueue(i915->flip_wq);
+	destroy_workqueue(i915->modeset_wq);
 
-/*
- * set vga decode state - true == enable VGA decode
- */
-int intel_modeset_vga_set_state(struct drm_i915_private *dev_priv, bool state)
-{
-	unsigned reg = INTEL_GEN(dev_priv) >= 6 ? SNB_GMCH_CTRL : INTEL_GMCH_CTRL;
-	u16 gmch_ctrl;
-
-	if (pci_read_config_word(dev_priv->bridge_dev, reg, &gmch_ctrl)) {
-		DRM_ERROR("failed to read control word\n");
-		return -EIO;
-	}
-
-	if (!!(gmch_ctrl & INTEL_GMCH_VGA_DISABLE) == !state)
-		return 0;
-
-	if (state)
-		gmch_ctrl &= ~INTEL_GMCH_VGA_DISABLE;
-	else
-		gmch_ctrl |= INTEL_GMCH_VGA_DISABLE;
-
-	if (pci_write_config_word(dev_priv->bridge_dev, reg, gmch_ctrl)) {
-		DRM_ERROR("failed to write control word\n");
-		return -EIO;
-	}
-
-	return 0;
+	intel_fbc_cleanup_cfb(i915);
 }
 
 #if IS_ENABLED(CONFIG_DRM_I915_CAPTURE_ERROR)
@@ -17194,7 +17827,7 @@ intel_display_capture_error_state(struct drm_i915_private *dev_priv)
 
 	BUILD_BUG_ON(ARRAY_SIZE(transcoders) != ARRAY_SIZE(error->transcoder));
 
-	if (!HAS_DISPLAY(dev_priv))
+	if (!HAS_DISPLAY(dev_priv) || !INTEL_DISPLAY_ENABLED(dev_priv))
 		return NULL;
 
 	error = kzalloc(sizeof(*error), GFP_ATOMIC);
@@ -17273,7 +17906,7 @@ intel_display_print_error_state(struct drm_i915_error_state_buf *m,
 	if (!error)
 		return;
 
-	err_printf(m, "Num Pipes: %d\n", INTEL_INFO(dev_priv)->num_pipes);
+	err_printf(m, "Num Pipes: %d\n", INTEL_NUM_PIPES(dev_priv));
 	if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv))
 		err_printf(m, "PWR_WELL_CTL2: %08x\n",
 			   error->power_well_driver);
diff --git a/drivers/gpu/drm/i915/display/intel_display.h b/drivers/gpu/drm/i915/display/intel_display.h
index e57e6969051d..f417e0948001 100644
--- a/drivers/gpu/drm/i915/display/intel_display.h
+++ b/drivers/gpu/drm/i915/display/intel_display.h
@@ -1,5 +1,5 @@
 /*
- * Copyright © 2006-2017 Intel Corporation
+ * Copyright © 2006-2019 Intel Corporation
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the "Software"),
@@ -32,8 +32,10 @@ enum link_m_n_set;
 struct dpll;
 struct drm_connector;
 struct drm_device;
+struct drm_display_mode;
 struct drm_encoder;
 struct drm_file;
+struct drm_format_info;
 struct drm_framebuffer;
 struct drm_i915_error_state_buf;
 struct drm_i915_gem_object;
@@ -52,6 +54,7 @@ struct intel_plane;
 struct intel_plane_state;
 struct intel_remapped_info;
 struct intel_rotation_info;
+struct intel_crtc_state;
 
 enum i915_gpio {
 	GPIOA,
@@ -91,6 +94,7 @@ enum pipe {
 #define pipe_name(p) ((p) + 'A')
 
 enum transcoder {
+	INVALID_TRANSCODER = -1,
 	/*
 	 * The following transcoders have a 1:1 transcoder -> pipe mapping,
 	 * keep their values fixed: the code assumes that TRANSCODER_A=0, the
@@ -182,6 +186,24 @@ enum plane_id {
 	for ((__p) = PLANE_PRIMARY; (__p) < I915_MAX_PLANES; (__p)++) \
 		for_each_if((__crtc)->plane_ids_mask & BIT(__p))
 
+enum port {
+	PORT_NONE = -1,
+
+	PORT_A = 0,
+	PORT_B,
+	PORT_C,
+	PORT_D,
+	PORT_E,
+	PORT_F,
+	PORT_G,
+	PORT_H,
+	PORT_I,
+
+	I915_MAX_PORTS
+};
+
+#define port_name(p) ((p) + 'A')
+
 /*
  * Ports identifier referenced from other drivers.
  * Expected to remain stable over time
@@ -251,6 +273,7 @@ enum aux_ch {
 	AUX_CH_D,
 	AUX_CH_E, /* ICL+ */
 	AUX_CH_F,
+	AUX_CH_G,
 };
 
 #define aux_ch_name(a) ((a) + 'A')
@@ -289,10 +312,10 @@ enum phy_fia {
 };
 
 #define for_each_pipe(__dev_priv, __p) \
-	for ((__p) = 0; (__p) < INTEL_INFO(__dev_priv)->num_pipes; (__p)++)
+	for ((__p) = 0; (__p) < INTEL_NUM_PIPES(__dev_priv); (__p)++)
 
 #define for_each_pipe_masked(__dev_priv, __p, __mask) \
-	for ((__p) = 0; (__p) < INTEL_INFO(__dev_priv)->num_pipes; (__p)++) \
+	for ((__p) = 0; (__p) < INTEL_NUM_PIPES(__dev_priv); (__p)++) \
 		for_each_if((__mask) & BIT(__p))
 
 #define for_each_cpu_transcoder_masked(__dev_priv, __t, __mask) \
@@ -330,7 +353,7 @@ enum phy_fia {
 			    &(dev)->mode_config.plane_list,		\
 			    base.head)					\
 		for_each_if((plane_mask) &				\
-			    drm_plane_mask(&intel_plane->base)))
+			    drm_plane_mask(&intel_plane->base))
 
 #define for_each_intel_plane_on_crtc(dev, intel_crtc, intel_plane)	\
 	list_for_each_entry(intel_plane,				\
@@ -411,16 +434,37 @@ enum phy_fia {
 	     (__i)++) \
 		for_each_if(crtc)
 
+#define for_each_oldnew_intel_crtc_in_state_reverse(__state, crtc, old_crtc_state, new_crtc_state, __i) \
+	for ((__i) = (__state)->base.dev->mode_config.num_crtc - 1; \
+	     (__i) >= 0  && \
+	     ((crtc) = to_intel_crtc((__state)->base.crtcs[__i].ptr), \
+	      (old_crtc_state) = to_intel_crtc_state((__state)->base.crtcs[__i].old_state), \
+	      (new_crtc_state) = to_intel_crtc_state((__state)->base.crtcs[__i].new_state), 1); \
+	     (__i)--) \
+		for_each_if(crtc)
+
+#define intel_atomic_crtc_state_for_each_plane_state( \
+		  plane, plane_state, \
+		  crtc_state) \
+	for_each_intel_plane_mask(((crtc_state)->base.state->dev), (plane), \
+				((crtc_state)->base.plane_mask)) \
+		for_each_if ((plane_state = \
+			      to_intel_plane_state(__drm_atomic_get_current_plane_state((crtc_state)->base.state, &plane->base))))
+
 void intel_link_compute_m_n(u16 bpp, int nlanes,
 			    int pixel_clock, int link_clock,
 			    struct intel_link_m_n *m_n,
-			    bool constant_n);
+			    bool constant_n, bool fec_enable);
 bool is_ccs_modifier(u64 modifier);
 void lpt_disable_clkout_dp(struct drm_i915_private *dev_priv);
 u32 intel_plane_fb_max_stride(struct drm_i915_private *dev_priv,
 			      u32 pixel_format, u64 modifier);
 bool intel_plane_can_remap(const struct intel_plane_state *plane_state);
+enum drm_mode_status
+intel_mode_valid_max_plane_size(struct drm_i915_private *dev_priv,
+				const struct drm_display_mode *mode);
 enum phy intel_port_to_phy(struct drm_i915_private *i915, enum port port);
+bool is_trans_port_sync_mode(const struct intel_crtc_state *state);
 
 void intel_plane_destroy(struct drm_plane *plane);
 void i830_enable_pipe(struct drm_i915_private *dev_priv, enum pipe pipe);
@@ -464,7 +508,6 @@ void vlv_wait_port_ready(struct drm_i915_private *dev_priv,
 			 struct intel_digital_port *dport,
 			 unsigned int expected_mask);
 int intel_get_load_detect_pipe(struct drm_connector *connector,
-			       const struct drm_display_mode *mode,
 			       struct intel_load_detect_pipe *old,
 			       struct drm_modeset_acquire_ctx *ctx);
 void intel_release_load_detect_pipe(struct drm_connector *connector,
@@ -499,8 +542,6 @@ void intel_dp_get_m_n(struct intel_crtc *crtc,
 		      struct intel_crtc_state *pipe_config);
 void intel_dp_set_m_n(const struct intel_crtc_state *crtc_state,
 		      enum link_m_n_set m_n);
-void intel_dp_ycbcr_420_enable(struct intel_dp *intel_dp,
-			       const struct intel_crtc_state *crtc_state);
 int intel_dotclock_calculate(int link_freq, const struct intel_link_m_n *m_n);
 bool bxt_find_best_dpll(struct intel_crtc_state *crtc_state,
 			struct dpll *best_clock);
@@ -520,8 +561,6 @@ void intel_crtc_arm_fifo_underrun(struct intel_crtc *crtc,
 
 u16 skl_scaler_calc_phase(int sub, int scale, bool chroma_center);
 int skl_update_scaler_crtc(struct intel_crtc_state *crtc_state);
-int skl_max_scale(const struct intel_crtc_state *crtc_state,
-		  u32 pixel_format);
 u32 glk_plane_color_ctl(const struct intel_crtc_state *crtc_state,
 			const struct intel_plane_state *plane_state);
 u32 glk_plane_color_ctl_crtc(const struct intel_crtc_state *crtc_state);
@@ -544,13 +583,10 @@ void intel_display_print_error_state(struct drm_i915_error_state_buf *e,
 				     struct intel_display_error_state *error);
 
 /* modesetting */
-void intel_modeset_init_hw(struct drm_device *dev);
-int intel_modeset_init(struct drm_device *dev);
-void intel_modeset_driver_remove(struct drm_device *dev);
-int intel_modeset_vga_set_state(struct drm_i915_private *dev_priv, bool state);
+void intel_modeset_init_hw(struct drm_i915_private *i915);
+int intel_modeset_init(struct drm_i915_private *i915);
+void intel_modeset_driver_remove(struct drm_i915_private *i915);
 void intel_display_resume(struct drm_device *dev);
-void i915_redisable_vga(struct drm_i915_private *dev_priv);
-void i915_redisable_vga_power_on(struct drm_i915_private *dev_priv);
 void intel_init_pch_refclk(struct drm_i915_private *dev_priv);
 
 /* modesetting asserts */
diff --git a/drivers/gpu/drm/i915/display/intel_display_power.c b/drivers/gpu/drm/i915/display/intel_display_power.c
index 12099760d99e..ce1b64f4dd44 100644
--- a/drivers/gpu/drm/i915/display/intel_display_power.c
+++ b/drivers/gpu/drm/i915/display/intel_display_power.c
@@ -3,8 +3,6 @@
  * Copyright © 2019 Intel Corporation
  */
 
-#include <linux/vgaarb.h>
-
 #include "display/intel_crt.h"
 #include "display/intel_dp.h"
 
@@ -19,16 +17,14 @@
 #include "intel_hotplug.h"
 #include "intel_sideband.h"
 #include "intel_tc.h"
+#include "intel_vga.h"
 
 bool intel_display_power_well_is_enabled(struct drm_i915_private *dev_priv,
 					 enum i915_power_well_id power_well_id);
 
 const char *
-intel_display_power_domain_str(struct drm_i915_private *i915,
-			       enum intel_display_power_domain domain)
+intel_display_power_domain_str(enum intel_display_power_domain domain)
 {
-	bool ddi_tc_ports = IS_GEN(i915, 12);
-
 	switch (domain) {
 	case POWER_DOMAIN_DISPLAY_CORE:
 		return "DISPLAY_CORE";
@@ -71,23 +67,17 @@ intel_display_power_domain_str(struct drm_i915_private *i915,
 	case POWER_DOMAIN_PORT_DDI_C_LANES:
 		return "PORT_DDI_C_LANES";
 	case POWER_DOMAIN_PORT_DDI_D_LANES:
-		BUILD_BUG_ON(POWER_DOMAIN_PORT_DDI_D_LANES !=
-			     POWER_DOMAIN_PORT_DDI_TC1_LANES);
-		return ddi_tc_ports ? "PORT_DDI_TC1_LANES" : "PORT_DDI_D_LANES";
+		return "PORT_DDI_D_LANES";
 	case POWER_DOMAIN_PORT_DDI_E_LANES:
-		BUILD_BUG_ON(POWER_DOMAIN_PORT_DDI_E_LANES !=
-			     POWER_DOMAIN_PORT_DDI_TC2_LANES);
-		return ddi_tc_ports ? "PORT_DDI_TC2_LANES" : "PORT_DDI_E_LANES";
+		return "PORT_DDI_E_LANES";
 	case POWER_DOMAIN_PORT_DDI_F_LANES:
-		BUILD_BUG_ON(POWER_DOMAIN_PORT_DDI_F_LANES !=
-			     POWER_DOMAIN_PORT_DDI_TC3_LANES);
-		return ddi_tc_ports ? "PORT_DDI_TC3_LANES" : "PORT_DDI_F_LANES";
-	case POWER_DOMAIN_PORT_DDI_TC4_LANES:
-		return "PORT_DDI_TC4_LANES";
-	case POWER_DOMAIN_PORT_DDI_TC5_LANES:
-		return "PORT_DDI_TC5_LANES";
-	case POWER_DOMAIN_PORT_DDI_TC6_LANES:
-		return "PORT_DDI_TC6_LANES";
+		return "PORT_DDI_F_LANES";
+	case POWER_DOMAIN_PORT_DDI_G_LANES:
+		return "PORT_DDI_G_LANES";
+	case POWER_DOMAIN_PORT_DDI_H_LANES:
+		return "PORT_DDI_H_LANES";
+	case POWER_DOMAIN_PORT_DDI_I_LANES:
+		return "PORT_DDI_I_LANES";
 	case POWER_DOMAIN_PORT_DDI_A_IO:
 		return "PORT_DDI_A_IO";
 	case POWER_DOMAIN_PORT_DDI_B_IO:
@@ -95,23 +85,17 @@ intel_display_power_domain_str(struct drm_i915_private *i915,
 	case POWER_DOMAIN_PORT_DDI_C_IO:
 		return "PORT_DDI_C_IO";
 	case POWER_DOMAIN_PORT_DDI_D_IO:
-		BUILD_BUG_ON(POWER_DOMAIN_PORT_DDI_D_IO !=
-			     POWER_DOMAIN_PORT_DDI_TC1_IO);
-		return ddi_tc_ports ? "PORT_DDI_TC1_IO" : "PORT_DDI_D_IO";
+		return "PORT_DDI_D_IO";
 	case POWER_DOMAIN_PORT_DDI_E_IO:
-		BUILD_BUG_ON(POWER_DOMAIN_PORT_DDI_E_IO !=
-			     POWER_DOMAIN_PORT_DDI_TC2_IO);
-		return ddi_tc_ports ? "PORT_DDI_TC2_IO" : "PORT_DDI_E_IO";
+		return "PORT_DDI_E_IO";
 	case POWER_DOMAIN_PORT_DDI_F_IO:
-		BUILD_BUG_ON(POWER_DOMAIN_PORT_DDI_F_IO !=
-			     POWER_DOMAIN_PORT_DDI_TC3_IO);
-		return ddi_tc_ports ? "PORT_DDI_TC3_IO" : "PORT_DDI_F_IO";
-	case POWER_DOMAIN_PORT_DDI_TC4_IO:
-		return "PORT_DDI_TC4_IO";
-	case POWER_DOMAIN_PORT_DDI_TC5_IO:
-		return "PORT_DDI_TC5_IO";
-	case POWER_DOMAIN_PORT_DDI_TC6_IO:
-		return "PORT_DDI_TC6_IO";
+		return "PORT_DDI_F_IO";
+	case POWER_DOMAIN_PORT_DDI_G_IO:
+		return "PORT_DDI_G_IO";
+	case POWER_DOMAIN_PORT_DDI_H_IO:
+		return "PORT_DDI_H_IO";
+	case POWER_DOMAIN_PORT_DDI_I_IO:
+		return "PORT_DDI_I_IO";
 	case POWER_DOMAIN_PORT_DSI:
 		return "PORT_DSI";
 	case POWER_DOMAIN_PORT_CRT:
@@ -129,34 +113,33 @@ intel_display_power_domain_str(struct drm_i915_private *i915,
 	case POWER_DOMAIN_AUX_C:
 		return "AUX_C";
 	case POWER_DOMAIN_AUX_D:
-		BUILD_BUG_ON(POWER_DOMAIN_AUX_D != POWER_DOMAIN_AUX_TC1);
-		return ddi_tc_ports ? "AUX_TC1" : "AUX_D";
+		return "AUX_D";
 	case POWER_DOMAIN_AUX_E:
-		BUILD_BUG_ON(POWER_DOMAIN_AUX_E != POWER_DOMAIN_AUX_TC2);
-		return ddi_tc_ports ? "AUX_TC2" : "AUX_E";
+		return "AUX_E";
 	case POWER_DOMAIN_AUX_F:
-		BUILD_BUG_ON(POWER_DOMAIN_AUX_F != POWER_DOMAIN_AUX_TC3);
-		return ddi_tc_ports ? "AUX_TC3" : "AUX_F";
-	case POWER_DOMAIN_AUX_TC4:
-		return "AUX_TC4";
-	case POWER_DOMAIN_AUX_TC5:
-		return "AUX_TC5";
-	case POWER_DOMAIN_AUX_TC6:
-		return "AUX_TC6";
+		return "AUX_F";
+	case POWER_DOMAIN_AUX_G:
+		return "AUX_G";
+	case POWER_DOMAIN_AUX_H:
+		return "AUX_H";
+	case POWER_DOMAIN_AUX_I:
+		return "AUX_I";
 	case POWER_DOMAIN_AUX_IO_A:
 		return "AUX_IO_A";
-	case POWER_DOMAIN_AUX_TBT1:
-		return "AUX_TBT1";
-	case POWER_DOMAIN_AUX_TBT2:
-		return "AUX_TBT2";
-	case POWER_DOMAIN_AUX_TBT3:
-		return "AUX_TBT3";
-	case POWER_DOMAIN_AUX_TBT4:
-		return "AUX_TBT4";
-	case POWER_DOMAIN_AUX_TBT5:
-		return "AUX_TBT5";
-	case POWER_DOMAIN_AUX_TBT6:
-		return "AUX_TBT6";
+	case POWER_DOMAIN_AUX_C_TBT:
+		return "AUX_C_TBT";
+	case POWER_DOMAIN_AUX_D_TBT:
+		return "AUX_D_TBT";
+	case POWER_DOMAIN_AUX_E_TBT:
+		return "AUX_E_TBT";
+	case POWER_DOMAIN_AUX_F_TBT:
+		return "AUX_F_TBT";
+	case POWER_DOMAIN_AUX_G_TBT:
+		return "AUX_G_TBT";
+	case POWER_DOMAIN_AUX_H_TBT:
+		return "AUX_H_TBT";
+	case POWER_DOMAIN_AUX_I_TBT:
+		return "AUX_I_TBT";
 	case POWER_DOMAIN_GMBUS:
 		return "GMBUS";
 	case POWER_DOMAIN_INIT:
@@ -283,23 +266,8 @@ bool intel_display_power_is_enabled(struct drm_i915_private *dev_priv,
 static void hsw_power_well_post_enable(struct drm_i915_private *dev_priv,
 				       u8 irq_pipe_mask, bool has_vga)
 {
-	struct pci_dev *pdev = dev_priv->drm.pdev;
-
-	/*
-	 * After we re-enable the power well, if we touch VGA register 0x3d5
-	 * we'll get unclaimed register interrupts. This stops after we write
-	 * anything to the VGA MSR register. The vgacon module uses this
-	 * register all the time, so if we unbind our driver and, as a
-	 * consequence, bind vgacon, we'll get stuck in an infinite loop at
-	 * console_unlock(). So make here we touch the VGA MSR register, making
-	 * sure vgacon can keep working normally without triggering interrupts
-	 * and error messages.
-	 */
-	if (has_vga) {
-		vga_get_uninterruptible(pdev, VGA_RSRC_LEGACY_IO);
-		outb(inb(VGA_MSR_READ), VGA_MSR_WRITE);
-		vga_put(pdev, VGA_RSRC_LEGACY_IO);
-	}
+	if (has_vga)
+		intel_vga_reset_io_mem(dev_priv);
 
 	if (irq_pipe_mask)
 		gen8_irq_power_well_post_enable(dev_priv, irq_pipe_mask);
@@ -578,6 +546,8 @@ static void icl_tc_port_assert_ref_held(struct drm_i915_private *dev_priv,
 
 #endif
 
+#define TGL_AUX_PW_TO_TC_PORT(pw_idx)	((pw_idx) - TGL_PW_CTL_IDX_AUX_TC1)
+
 static void
 icl_tc_phy_aux_power_well_enable(struct drm_i915_private *dev_priv,
 				 struct i915_power_well *power_well)
@@ -594,6 +564,17 @@ icl_tc_phy_aux_power_well_enable(struct drm_i915_private *dev_priv,
 	I915_WRITE(DP_AUX_CH_CTL(aux_ch), val);
 
 	hsw_power_well_enable(dev_priv, power_well);
+
+	if (INTEL_GEN(dev_priv) >= 12 && !power_well->desc->hsw.is_tc_tbt) {
+		enum tc_port tc_port;
+
+		tc_port = TGL_AUX_PW_TO_TC_PORT(power_well->desc->hsw.idx);
+		I915_WRITE(HIP_INDEX_REG(tc_port), HIP_INDEX_VAL(tc_port, 0x2));
+
+		if (intel_de_wait_for_set(dev_priv, DKL_CMN_UC_DW_27(tc_port),
+					  DKL_CMN_UC_DW27_UC_HEALTH, 1))
+			DRM_WARN("Timeout waiting TC uC health\n");
+	}
 }
 
 static void
@@ -714,7 +695,11 @@ static u32 gen9_dc_mask(struct drm_i915_private *dev_priv)
 	u32 mask;
 
 	mask = DC_STATE_EN_UPTO_DC5;
-	if (INTEL_GEN(dev_priv) >= 11)
+
+	if (INTEL_GEN(dev_priv) >= 12)
+		mask |= DC_STATE_EN_DC3CO | DC_STATE_EN_UPTO_DC6
+					  | DC_STATE_EN_DC9;
+	else if (IS_GEN(dev_priv, 11))
 		mask |= DC_STATE_EN_UPTO_DC6 | DC_STATE_EN_DC9;
 	else if (IS_GEN9_LP(dev_priv))
 		mask |= DC_STATE_EN_DC9;
@@ -784,6 +769,52 @@ static void gen9_set_dc_state(struct drm_i915_private *dev_priv, u32 state)
 	dev_priv->csr.dc_state = val & mask;
 }
 
+static u32
+sanitize_target_dc_state(struct drm_i915_private *dev_priv,
+			 u32 target_dc_state)
+{
+	u32 states[] = {
+		DC_STATE_EN_UPTO_DC6,
+		DC_STATE_EN_UPTO_DC5,
+		DC_STATE_EN_DC3CO,
+		DC_STATE_DISABLE,
+	};
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(states) - 1; i++) {
+		if (target_dc_state != states[i])
+			continue;
+
+		if (dev_priv->csr.allowed_dc_mask & target_dc_state)
+			break;
+
+		target_dc_state = states[i + 1];
+	}
+
+	return target_dc_state;
+}
+
+static void tgl_enable_dc3co(struct drm_i915_private *dev_priv)
+{
+	DRM_DEBUG_KMS("Enabling DC3CO\n");
+	gen9_set_dc_state(dev_priv, DC_STATE_EN_DC3CO);
+}
+
+static void tgl_disable_dc3co(struct drm_i915_private *dev_priv)
+{
+	u32 val;
+
+	DRM_DEBUG_KMS("Disabling DC3CO\n");
+	val = I915_READ(DC_STATE_EN);
+	val &= ~DC_STATE_DC3CO_STATUS;
+	I915_WRITE(DC_STATE_EN, val);
+	gen9_set_dc_state(dev_priv, DC_STATE_DISABLE);
+	/*
+	 * Delay of 200us DC3CO Exit time B.Spec 49196
+	 */
+	usleep_range(200, 210);
+}
+
 static void bxt_enable_dc9(struct drm_i915_private *dev_priv)
 {
 	assert_can_enable_dc9(dev_priv);
@@ -839,6 +870,51 @@ lookup_power_well(struct drm_i915_private *dev_priv,
 	return &dev_priv->power_domains.power_wells[0];
 }
 
+/**
+ * intel_display_power_set_target_dc_state - Set target dc state.
+ * @dev_priv: i915 device
+ * @state: state which needs to be set as target_dc_state.
+ *
+ * This function set the "DC off" power well target_dc_state,
+ * based upon this target_dc_stste, "DC off" power well will
+ * enable desired DC state.
+ */
+void intel_display_power_set_target_dc_state(struct drm_i915_private *dev_priv,
+					     u32 state)
+{
+	struct i915_power_well *power_well;
+	bool dc_off_enabled;
+	struct i915_power_domains *power_domains = &dev_priv->power_domains;
+
+	mutex_lock(&power_domains->lock);
+	power_well = lookup_power_well(dev_priv, SKL_DISP_DC_OFF);
+
+	if (WARN_ON(!power_well))
+		goto unlock;
+
+	state = sanitize_target_dc_state(dev_priv, state);
+
+	if (state == dev_priv->csr.target_dc_state)
+		goto unlock;
+
+	dc_off_enabled = power_well->desc->ops->is_enabled(dev_priv,
+							   power_well);
+	/*
+	 * If DC off power well is disabled, need to enable and disable the
+	 * DC off power well to effect target DC state.
+	 */
+	if (!dc_off_enabled)
+		power_well->desc->ops->enable(dev_priv, power_well);
+
+	dev_priv->csr.target_dc_state = state;
+
+	if (!dc_off_enabled)
+		power_well->desc->ops->disable(dev_priv, power_well);
+
+unlock:
+	mutex_unlock(&power_domains->lock);
+}
+
 static void assert_can_enable_dc5(struct drm_i915_private *dev_priv)
 {
 	bool pg2_enabled = intel_display_power_well_is_enabled(dev_priv,
@@ -951,7 +1027,8 @@ static void bxt_verify_ddi_phy_power_wells(struct drm_i915_private *dev_priv)
 static bool gen9_dc_off_power_well_enabled(struct drm_i915_private *dev_priv,
 					   struct i915_power_well *power_well)
 {
-	return (I915_READ(DC_STATE_EN) & DC_STATE_EN_UPTO_DC5_DC6_MASK) == 0;
+	return ((I915_READ(DC_STATE_EN) & DC_STATE_EN_DC3CO) == 0 &&
+		(I915_READ(DC_STATE_EN) & DC_STATE_EN_UPTO_DC5_DC6_MASK) == 0);
 }
 
 static void gen9_assert_dbuf_enabled(struct drm_i915_private *dev_priv)
@@ -967,6 +1044,11 @@ static void gen9_disable_dc_states(struct drm_i915_private *dev_priv)
 {
 	struct intel_cdclk_state cdclk_state = {};
 
+	if (dev_priv->csr.target_dc_state == DC_STATE_EN_DC3CO) {
+		tgl_disable_dc3co(dev_priv);
+		return;
+	}
+
 	gen9_set_dc_state(dev_priv, DC_STATE_DISABLE);
 
 	dev_priv->display.get_cdclk(dev_priv, &cdclk_state);
@@ -999,10 +1081,17 @@ static void gen9_dc_off_power_well_disable(struct drm_i915_private *dev_priv,
 	if (!dev_priv->csr.dmc_payload)
 		return;
 
-	if (dev_priv->csr.allowed_dc_mask & DC_STATE_EN_UPTO_DC6)
+	switch (dev_priv->csr.target_dc_state) {
+	case DC_STATE_EN_DC3CO:
+		tgl_enable_dc3co(dev_priv);
+		break;
+	case DC_STATE_EN_UPTO_DC6:
 		skl_enable_dc6(dev_priv);
-	else if (dev_priv->csr.allowed_dc_mask & DC_STATE_EN_UPTO_DC5)
+		break;
+	case DC_STATE_EN_UPTO_DC5:
 		gen9_enable_dc5(dev_priv);
+		break;
+	}
 }
 
 static void i9xx_power_well_sync_hw_noop(struct drm_i915_private *dev_priv,
@@ -1208,7 +1297,7 @@ static void vlv_display_power_well_init(struct drm_i915_private *dev_priv)
 			intel_crt_reset(&encoder->base);
 	}
 
-	i915_redisable_vga_power_on(dev_priv);
+	intel_vga_redisable_power_on(dev_priv);
 
 	intel_pps_unlock_regs_wa(dev_priv);
 }
@@ -1718,15 +1807,12 @@ __async_put_domains_state_ok(struct i915_power_domains *power_domains)
 static void print_power_domains(struct i915_power_domains *power_domains,
 				const char *prefix, u64 mask)
 {
-	struct drm_i915_private *i915 =
-		container_of(power_domains, struct drm_i915_private,
-			     power_domains);
 	enum intel_display_power_domain domain;
 
 	DRM_DEBUG_DRIVER("%s (%lu):\n", prefix, hweight64(mask));
 	for_each_power_domain(domain, mask)
 		DRM_DEBUG_DRIVER("%s use_count %d\n",
-				 intel_display_power_domain_str(i915, domain),
+				 intel_display_power_domain_str(domain),
 				 power_domains->domain_use_count[domain]);
 }
 
@@ -1896,7 +1982,7 @@ __intel_display_power_put_domain(struct drm_i915_private *dev_priv,
 {
 	struct i915_power_domains *power_domains;
 	struct i915_power_well *power_well;
-	const char *name = intel_display_power_domain_str(dev_priv, domain);
+	const char *name = intel_display_power_domain_str(domain);
 
 	power_domains = &dev_priv->power_domains;
 
@@ -2487,10 +2573,10 @@ void intel_display_power_put(struct drm_i915_private *dev_priv,
 	BIT_ULL(POWER_DOMAIN_AUX_D) |			\
 	BIT_ULL(POWER_DOMAIN_AUX_E) |			\
 	BIT_ULL(POWER_DOMAIN_AUX_F) |			\
-	BIT_ULL(POWER_DOMAIN_AUX_TBT1) |		\
-	BIT_ULL(POWER_DOMAIN_AUX_TBT2) |		\
-	BIT_ULL(POWER_DOMAIN_AUX_TBT3) |		\
-	BIT_ULL(POWER_DOMAIN_AUX_TBT4) |		\
+	BIT_ULL(POWER_DOMAIN_AUX_C_TBT) |		\
+	BIT_ULL(POWER_DOMAIN_AUX_D_TBT) |		\
+	BIT_ULL(POWER_DOMAIN_AUX_E_TBT) |		\
+	BIT_ULL(POWER_DOMAIN_AUX_F_TBT) |		\
 	BIT_ULL(POWER_DOMAIN_VGA) |			\
 	BIT_ULL(POWER_DOMAIN_AUDIO) |			\
 	BIT_ULL(POWER_DOMAIN_INIT))
@@ -2530,22 +2616,22 @@ void intel_display_power_put(struct drm_i915_private *dev_priv,
 	BIT_ULL(POWER_DOMAIN_AUX_A))
 #define ICL_AUX_B_IO_POWER_DOMAINS (			\
 	BIT_ULL(POWER_DOMAIN_AUX_B))
-#define ICL_AUX_C_IO_POWER_DOMAINS (			\
+#define ICL_AUX_C_TC1_IO_POWER_DOMAINS (		\
 	BIT_ULL(POWER_DOMAIN_AUX_C))
-#define ICL_AUX_D_IO_POWER_DOMAINS (			\
+#define ICL_AUX_D_TC2_IO_POWER_DOMAINS (		\
 	BIT_ULL(POWER_DOMAIN_AUX_D))
-#define ICL_AUX_E_IO_POWER_DOMAINS (			\
+#define ICL_AUX_E_TC3_IO_POWER_DOMAINS (		\
 	BIT_ULL(POWER_DOMAIN_AUX_E))
-#define ICL_AUX_F_IO_POWER_DOMAINS (			\
+#define ICL_AUX_F_TC4_IO_POWER_DOMAINS (		\
 	BIT_ULL(POWER_DOMAIN_AUX_F))
-#define ICL_AUX_TBT1_IO_POWER_DOMAINS (			\
-	BIT_ULL(POWER_DOMAIN_AUX_TBT1))
-#define ICL_AUX_TBT2_IO_POWER_DOMAINS (			\
-	BIT_ULL(POWER_DOMAIN_AUX_TBT2))
-#define ICL_AUX_TBT3_IO_POWER_DOMAINS (			\
-	BIT_ULL(POWER_DOMAIN_AUX_TBT3))
-#define ICL_AUX_TBT4_IO_POWER_DOMAINS (			\
-	BIT_ULL(POWER_DOMAIN_AUX_TBT4))
+#define ICL_AUX_C_TBT1_IO_POWER_DOMAINS (		\
+	BIT_ULL(POWER_DOMAIN_AUX_C_TBT))
+#define ICL_AUX_D_TBT2_IO_POWER_DOMAINS (		\
+	BIT_ULL(POWER_DOMAIN_AUX_D_TBT))
+#define ICL_AUX_E_TBT3_IO_POWER_DOMAINS (		\
+	BIT_ULL(POWER_DOMAIN_AUX_E_TBT))
+#define ICL_AUX_F_TBT4_IO_POWER_DOMAINS (		\
+	BIT_ULL(POWER_DOMAIN_AUX_F_TBT))
 
 #define TGL_PW_5_POWER_DOMAINS (			\
 	BIT_ULL(POWER_DOMAIN_PIPE_D) |			\
@@ -2565,24 +2651,24 @@ void intel_display_power_put(struct drm_i915_private *dev_priv,
 	BIT_ULL(POWER_DOMAIN_PIPE_B) |			\
 	BIT_ULL(POWER_DOMAIN_TRANSCODER_B) |		\
 	BIT_ULL(POWER_DOMAIN_PIPE_B_PANEL_FITTER) |	\
-	BIT_ULL(POWER_DOMAIN_PORT_DDI_TC1_LANES) |	\
-	BIT_ULL(POWER_DOMAIN_PORT_DDI_TC2_LANES) |	\
-	BIT_ULL(POWER_DOMAIN_PORT_DDI_TC3_LANES) |	\
-	BIT_ULL(POWER_DOMAIN_PORT_DDI_TC4_LANES) |	\
-	BIT_ULL(POWER_DOMAIN_PORT_DDI_TC5_LANES) |	\
-	BIT_ULL(POWER_DOMAIN_PORT_DDI_TC6_LANES) |	\
-	BIT_ULL(POWER_DOMAIN_AUX_TC1) |		\
-	BIT_ULL(POWER_DOMAIN_AUX_TC2) |		\
-	BIT_ULL(POWER_DOMAIN_AUX_TC3) |		\
-	BIT_ULL(POWER_DOMAIN_AUX_TC4) |		\
-	BIT_ULL(POWER_DOMAIN_AUX_TC5) |		\
-	BIT_ULL(POWER_DOMAIN_AUX_TC6) |		\
-	BIT_ULL(POWER_DOMAIN_AUX_TBT1) |		\
-	BIT_ULL(POWER_DOMAIN_AUX_TBT2) |		\
-	BIT_ULL(POWER_DOMAIN_AUX_TBT3) |		\
-	BIT_ULL(POWER_DOMAIN_AUX_TBT4) |		\
-	BIT_ULL(POWER_DOMAIN_AUX_TBT5) |		\
-	BIT_ULL(POWER_DOMAIN_AUX_TBT6) |		\
+	BIT_ULL(POWER_DOMAIN_PORT_DDI_D_LANES) |	\
+	BIT_ULL(POWER_DOMAIN_PORT_DDI_E_LANES) |	\
+	BIT_ULL(POWER_DOMAIN_PORT_DDI_F_LANES) |	\
+	BIT_ULL(POWER_DOMAIN_PORT_DDI_G_LANES) |	\
+	BIT_ULL(POWER_DOMAIN_PORT_DDI_H_LANES) |	\
+	BIT_ULL(POWER_DOMAIN_PORT_DDI_I_LANES) |	\
+	BIT_ULL(POWER_DOMAIN_AUX_D) |			\
+	BIT_ULL(POWER_DOMAIN_AUX_E) |			\
+	BIT_ULL(POWER_DOMAIN_AUX_F) |			\
+	BIT_ULL(POWER_DOMAIN_AUX_G) |			\
+	BIT_ULL(POWER_DOMAIN_AUX_H) |			\
+	BIT_ULL(POWER_DOMAIN_AUX_I) |			\
+	BIT_ULL(POWER_DOMAIN_AUX_D_TBT) |		\
+	BIT_ULL(POWER_DOMAIN_AUX_E_TBT) |		\
+	BIT_ULL(POWER_DOMAIN_AUX_F_TBT) |		\
+	BIT_ULL(POWER_DOMAIN_AUX_G_TBT) |		\
+	BIT_ULL(POWER_DOMAIN_AUX_H_TBT) |		\
+	BIT_ULL(POWER_DOMAIN_AUX_I_TBT) |		\
 	BIT_ULL(POWER_DOMAIN_VGA) |			\
 	BIT_ULL(POWER_DOMAIN_AUDIO) |			\
 	BIT_ULL(POWER_DOMAIN_INIT))
@@ -2596,37 +2682,54 @@ void intel_display_power_put(struct drm_i915_private *dev_priv,
 	TGL_PW_2_POWER_DOMAINS |			\
 	BIT_ULL(POWER_DOMAIN_MODESET) |			\
 	BIT_ULL(POWER_DOMAIN_AUX_A) |			\
+	BIT_ULL(POWER_DOMAIN_AUX_B) |			\
+	BIT_ULL(POWER_DOMAIN_AUX_C) |			\
 	BIT_ULL(POWER_DOMAIN_INIT))
 
-#define TGL_DDI_IO_TC1_POWER_DOMAINS (		\
-	BIT_ULL(POWER_DOMAIN_PORT_DDI_TC1_IO))
-#define TGL_DDI_IO_TC2_POWER_DOMAINS (		\
-	BIT_ULL(POWER_DOMAIN_PORT_DDI_TC2_IO))
-#define TGL_DDI_IO_TC3_POWER_DOMAINS (		\
-	BIT_ULL(POWER_DOMAIN_PORT_DDI_TC3_IO))
-#define TGL_DDI_IO_TC4_POWER_DOMAINS (		\
-	BIT_ULL(POWER_DOMAIN_PORT_DDI_TC4_IO))
-#define TGL_DDI_IO_TC5_POWER_DOMAINS (		\
-	BIT_ULL(POWER_DOMAIN_PORT_DDI_TC5_IO))
-#define TGL_DDI_IO_TC6_POWER_DOMAINS (		\
-	BIT_ULL(POWER_DOMAIN_PORT_DDI_TC6_IO))
-
-#define TGL_AUX_TC1_IO_POWER_DOMAINS (		\
-	BIT_ULL(POWER_DOMAIN_AUX_TC1))
-#define TGL_AUX_TC2_IO_POWER_DOMAINS (		\
-	BIT_ULL(POWER_DOMAIN_AUX_TC2))
-#define TGL_AUX_TC3_IO_POWER_DOMAINS (		\
-	BIT_ULL(POWER_DOMAIN_AUX_TC3))
-#define TGL_AUX_TC4_IO_POWER_DOMAINS (		\
-	BIT_ULL(POWER_DOMAIN_AUX_TC4))
-#define TGL_AUX_TC5_IO_POWER_DOMAINS (		\
-	BIT_ULL(POWER_DOMAIN_AUX_TC5))
-#define TGL_AUX_TC6_IO_POWER_DOMAINS (		\
-	BIT_ULL(POWER_DOMAIN_AUX_TC6))
-#define TGL_AUX_TBT5_IO_POWER_DOMAINS (		\
-	BIT_ULL(POWER_DOMAIN_AUX_TBT5))
-#define TGL_AUX_TBT6_IO_POWER_DOMAINS (		\
-	BIT_ULL(POWER_DOMAIN_AUX_TBT6))
+#define TGL_DDI_IO_D_TC1_POWER_DOMAINS (	\
+	BIT_ULL(POWER_DOMAIN_PORT_DDI_D_IO))
+#define TGL_DDI_IO_E_TC2_POWER_DOMAINS (	\
+	BIT_ULL(POWER_DOMAIN_PORT_DDI_E_IO))
+#define TGL_DDI_IO_F_TC3_POWER_DOMAINS (	\
+	BIT_ULL(POWER_DOMAIN_PORT_DDI_F_IO))
+#define TGL_DDI_IO_G_TC4_POWER_DOMAINS (	\
+	BIT_ULL(POWER_DOMAIN_PORT_DDI_G_IO))
+#define TGL_DDI_IO_H_TC5_POWER_DOMAINS (	\
+	BIT_ULL(POWER_DOMAIN_PORT_DDI_H_IO))
+#define TGL_DDI_IO_I_TC6_POWER_DOMAINS (	\
+	BIT_ULL(POWER_DOMAIN_PORT_DDI_I_IO))
+
+#define TGL_AUX_A_IO_POWER_DOMAINS (		\
+	BIT_ULL(POWER_DOMAIN_AUX_IO_A) |	\
+	BIT_ULL(POWER_DOMAIN_AUX_A))
+#define TGL_AUX_B_IO_POWER_DOMAINS (		\
+	BIT_ULL(POWER_DOMAIN_AUX_B))
+#define TGL_AUX_C_IO_POWER_DOMAINS (		\
+	BIT_ULL(POWER_DOMAIN_AUX_C))
+#define TGL_AUX_D_TC1_IO_POWER_DOMAINS (	\
+	BIT_ULL(POWER_DOMAIN_AUX_D))
+#define TGL_AUX_E_TC2_IO_POWER_DOMAINS (	\
+	BIT_ULL(POWER_DOMAIN_AUX_E))
+#define TGL_AUX_F_TC3_IO_POWER_DOMAINS (	\
+	BIT_ULL(POWER_DOMAIN_AUX_F))
+#define TGL_AUX_G_TC4_IO_POWER_DOMAINS (	\
+	BIT_ULL(POWER_DOMAIN_AUX_G))
+#define TGL_AUX_H_TC5_IO_POWER_DOMAINS (	\
+	BIT_ULL(POWER_DOMAIN_AUX_H))
+#define TGL_AUX_I_TC6_IO_POWER_DOMAINS (	\
+	BIT_ULL(POWER_DOMAIN_AUX_I))
+#define TGL_AUX_D_TBT1_IO_POWER_DOMAINS (	\
+	BIT_ULL(POWER_DOMAIN_AUX_D_TBT))
+#define TGL_AUX_E_TBT2_IO_POWER_DOMAINS (	\
+	BIT_ULL(POWER_DOMAIN_AUX_E_TBT))
+#define TGL_AUX_F_TBT3_IO_POWER_DOMAINS (	\
+	BIT_ULL(POWER_DOMAIN_AUX_F_TBT))
+#define TGL_AUX_G_TBT4_IO_POWER_DOMAINS (	\
+	BIT_ULL(POWER_DOMAIN_AUX_G_TBT))
+#define TGL_AUX_H_TBT5_IO_POWER_DOMAINS (	\
+	BIT_ULL(POWER_DOMAIN_AUX_H_TBT))
+#define TGL_AUX_I_TBT6_IO_POWER_DOMAINS (	\
+	BIT_ULL(POWER_DOMAIN_AUX_I_TBT))
 
 static const struct i915_power_well_ops i9xx_always_on_power_well_ops = {
 	.sync_hw = i9xx_power_well_sync_hw_noop,
@@ -2938,7 +3041,7 @@ static const struct i915_power_well_desc skl_power_wells[] = {
 		.name = "DC off",
 		.domains = SKL_DISPLAY_DC_OFF_POWER_DOMAINS,
 		.ops = &gen9_dc_off_power_well_ops,
-		.id = DISP_PW_ID_NONE,
+		.id = SKL_DISP_DC_OFF,
 	},
 	{
 		.name = "power well 2",
@@ -3020,7 +3123,7 @@ static const struct i915_power_well_desc bxt_power_wells[] = {
 		.name = "DC off",
 		.domains = BXT_DISPLAY_DC_OFF_POWER_DOMAINS,
 		.ops = &gen9_dc_off_power_well_ops,
-		.id = DISP_PW_ID_NONE,
+		.id = SKL_DISP_DC_OFF,
 	},
 	{
 		.name = "power well 2",
@@ -3080,7 +3183,7 @@ static const struct i915_power_well_desc glk_power_wells[] = {
 		.name = "DC off",
 		.domains = GLK_DISPLAY_DC_OFF_POWER_DOMAINS,
 		.ops = &gen9_dc_off_power_well_ops,
-		.id = DISP_PW_ID_NONE,
+		.id = SKL_DISP_DC_OFF,
 	},
 	{
 		.name = "power well 2",
@@ -3249,7 +3352,7 @@ static const struct i915_power_well_desc cnl_power_wells[] = {
 		.name = "DC off",
 		.domains = CNL_DISPLAY_DC_OFF_POWER_DOMAINS,
 		.ops = &gen9_dc_off_power_well_ops,
-		.id = DISP_PW_ID_NONE,
+		.id = SKL_DISP_DC_OFF,
 	},
 	{
 		.name = "power well 2",
@@ -3377,7 +3480,7 @@ static const struct i915_power_well_desc icl_power_wells[] = {
 		.name = "DC off",
 		.domains = ICL_DISPLAY_DC_OFF_POWER_DOMAINS,
 		.ops = &gen9_dc_off_power_well_ops,
-		.id = DISP_PW_ID_NONE,
+		.id = SKL_DISP_DC_OFF,
 	},
 	{
 		.name = "power well 2",
@@ -3484,8 +3587,8 @@ static const struct i915_power_well_desc icl_power_wells[] = {
 		},
 	},
 	{
-		.name = "AUX C",
-		.domains = ICL_AUX_C_IO_POWER_DOMAINS,
+		.name = "AUX C TC1",
+		.domains = ICL_AUX_C_TC1_IO_POWER_DOMAINS,
 		.ops = &icl_tc_phy_aux_power_well_ops,
 		.id = DISP_PW_ID_NONE,
 		{
@@ -3495,8 +3598,8 @@ static const struct i915_power_well_desc icl_power_wells[] = {
 		},
 	},
 	{
-		.name = "AUX D",
-		.domains = ICL_AUX_D_IO_POWER_DOMAINS,
+		.name = "AUX D TC2",
+		.domains = ICL_AUX_D_TC2_IO_POWER_DOMAINS,
 		.ops = &icl_tc_phy_aux_power_well_ops,
 		.id = DISP_PW_ID_NONE,
 		{
@@ -3506,8 +3609,8 @@ static const struct i915_power_well_desc icl_power_wells[] = {
 		},
 	},
 	{
-		.name = "AUX E",
-		.domains = ICL_AUX_E_IO_POWER_DOMAINS,
+		.name = "AUX E TC3",
+		.domains = ICL_AUX_E_TC3_IO_POWER_DOMAINS,
 		.ops = &icl_tc_phy_aux_power_well_ops,
 		.id = DISP_PW_ID_NONE,
 		{
@@ -3517,8 +3620,8 @@ static const struct i915_power_well_desc icl_power_wells[] = {
 		},
 	},
 	{
-		.name = "AUX F",
-		.domains = ICL_AUX_F_IO_POWER_DOMAINS,
+		.name = "AUX F TC4",
+		.domains = ICL_AUX_F_TC4_IO_POWER_DOMAINS,
 		.ops = &icl_tc_phy_aux_power_well_ops,
 		.id = DISP_PW_ID_NONE,
 		{
@@ -3528,8 +3631,8 @@ static const struct i915_power_well_desc icl_power_wells[] = {
 		},
 	},
 	{
-		.name = "AUX TBT1",
-		.domains = ICL_AUX_TBT1_IO_POWER_DOMAINS,
+		.name = "AUX C TBT1",
+		.domains = ICL_AUX_C_TBT1_IO_POWER_DOMAINS,
 		.ops = &icl_tc_phy_aux_power_well_ops,
 		.id = DISP_PW_ID_NONE,
 		{
@@ -3539,8 +3642,8 @@ static const struct i915_power_well_desc icl_power_wells[] = {
 		},
 	},
 	{
-		.name = "AUX TBT2",
-		.domains = ICL_AUX_TBT2_IO_POWER_DOMAINS,
+		.name = "AUX D TBT2",
+		.domains = ICL_AUX_D_TBT2_IO_POWER_DOMAINS,
 		.ops = &icl_tc_phy_aux_power_well_ops,
 		.id = DISP_PW_ID_NONE,
 		{
@@ -3550,8 +3653,8 @@ static const struct i915_power_well_desc icl_power_wells[] = {
 		},
 	},
 	{
-		.name = "AUX TBT3",
-		.domains = ICL_AUX_TBT3_IO_POWER_DOMAINS,
+		.name = "AUX E TBT3",
+		.domains = ICL_AUX_E_TBT3_IO_POWER_DOMAINS,
 		.ops = &icl_tc_phy_aux_power_well_ops,
 		.id = DISP_PW_ID_NONE,
 		{
@@ -3561,8 +3664,8 @@ static const struct i915_power_well_desc icl_power_wells[] = {
 		},
 	},
 	{
-		.name = "AUX TBT4",
-		.domains = ICL_AUX_TBT4_IO_POWER_DOMAINS,
+		.name = "AUX F TBT4",
+		.domains = ICL_AUX_F_TBT4_IO_POWER_DOMAINS,
 		.ops = &icl_tc_phy_aux_power_well_ops,
 		.id = DISP_PW_ID_NONE,
 		{
@@ -3610,7 +3713,7 @@ static const struct i915_power_well_desc tgl_power_wells[] = {
 		.name = "DC off",
 		.domains = TGL_DISPLAY_DC_OFF_POWER_DOMAINS,
 		.ops = &gen9_dc_off_power_well_ops,
-		.id = DISP_PW_ID_NONE,
+		.id = SKL_DISP_DC_OFF,
 	},
 	{
 		.name = "power well 2",
@@ -3667,8 +3770,8 @@ static const struct i915_power_well_desc tgl_power_wells[] = {
 		}
 	},
 	{
-		.name = "DDI TC1 IO",
-		.domains = TGL_DDI_IO_TC1_POWER_DOMAINS,
+		.name = "DDI D TC1 IO",
+		.domains = TGL_DDI_IO_D_TC1_POWER_DOMAINS,
 		.ops = &hsw_power_well_ops,
 		.id = DISP_PW_ID_NONE,
 		{
@@ -3677,8 +3780,8 @@ static const struct i915_power_well_desc tgl_power_wells[] = {
 		},
 	},
 	{
-		.name = "DDI TC2 IO",
-		.domains = TGL_DDI_IO_TC2_POWER_DOMAINS,
+		.name = "DDI E TC2 IO",
+		.domains = TGL_DDI_IO_E_TC2_POWER_DOMAINS,
 		.ops = &hsw_power_well_ops,
 		.id = DISP_PW_ID_NONE,
 		{
@@ -3687,8 +3790,8 @@ static const struct i915_power_well_desc tgl_power_wells[] = {
 		},
 	},
 	{
-		.name = "DDI TC3 IO",
-		.domains = TGL_DDI_IO_TC3_POWER_DOMAINS,
+		.name = "DDI F TC3 IO",
+		.domains = TGL_DDI_IO_F_TC3_POWER_DOMAINS,
 		.ops = &hsw_power_well_ops,
 		.id = DISP_PW_ID_NONE,
 		{
@@ -3697,8 +3800,8 @@ static const struct i915_power_well_desc tgl_power_wells[] = {
 		},
 	},
 	{
-		.name = "DDI TC4 IO",
-		.domains = TGL_DDI_IO_TC4_POWER_DOMAINS,
+		.name = "DDI G TC4 IO",
+		.domains = TGL_DDI_IO_G_TC4_POWER_DOMAINS,
 		.ops = &hsw_power_well_ops,
 		.id = DISP_PW_ID_NONE,
 		{
@@ -3707,8 +3810,8 @@ static const struct i915_power_well_desc tgl_power_wells[] = {
 		},
 	},
 	{
-		.name = "DDI TC5 IO",
-		.domains = TGL_DDI_IO_TC5_POWER_DOMAINS,
+		.name = "DDI H TC5 IO",
+		.domains = TGL_DDI_IO_H_TC5_POWER_DOMAINS,
 		.ops = &hsw_power_well_ops,
 		.id = DISP_PW_ID_NONE,
 		{
@@ -3717,8 +3820,8 @@ static const struct i915_power_well_desc tgl_power_wells[] = {
 		},
 	},
 	{
-		.name = "DDI TC6 IO",
-		.domains = TGL_DDI_IO_TC6_POWER_DOMAINS,
+		.name = "DDI I TC6 IO",
+		.domains = TGL_DDI_IO_I_TC6_POWER_DOMAINS,
 		.ops = &hsw_power_well_ops,
 		.id = DISP_PW_ID_NONE,
 		{
@@ -3728,7 +3831,7 @@ static const struct i915_power_well_desc tgl_power_wells[] = {
 	},
 	{
 		.name = "AUX A",
-		.domains = ICL_AUX_A_IO_POWER_DOMAINS,
+		.domains = TGL_AUX_A_IO_POWER_DOMAINS,
 		.ops = &icl_combo_phy_aux_power_well_ops,
 		.id = DISP_PW_ID_NONE,
 		{
@@ -3738,7 +3841,7 @@ static const struct i915_power_well_desc tgl_power_wells[] = {
 	},
 	{
 		.name = "AUX B",
-		.domains = ICL_AUX_B_IO_POWER_DOMAINS,
+		.domains = TGL_AUX_B_IO_POWER_DOMAINS,
 		.ops = &icl_combo_phy_aux_power_well_ops,
 		.id = DISP_PW_ID_NONE,
 		{
@@ -3748,7 +3851,7 @@ static const struct i915_power_well_desc tgl_power_wells[] = {
 	},
 	{
 		.name = "AUX C",
-		.domains = ICL_AUX_C_IO_POWER_DOMAINS,
+		.domains = TGL_AUX_C_IO_POWER_DOMAINS,
 		.ops = &icl_combo_phy_aux_power_well_ops,
 		.id = DISP_PW_ID_NONE,
 		{
@@ -3757,8 +3860,8 @@ static const struct i915_power_well_desc tgl_power_wells[] = {
 		},
 	},
 	{
-		.name = "AUX TC1",
-		.domains = TGL_AUX_TC1_IO_POWER_DOMAINS,
+		.name = "AUX D TC1",
+		.domains = TGL_AUX_D_TC1_IO_POWER_DOMAINS,
 		.ops = &icl_tc_phy_aux_power_well_ops,
 		.id = DISP_PW_ID_NONE,
 		{
@@ -3768,8 +3871,8 @@ static const struct i915_power_well_desc tgl_power_wells[] = {
 		},
 	},
 	{
-		.name = "AUX TC2",
-		.domains = TGL_AUX_TC2_IO_POWER_DOMAINS,
+		.name = "AUX E TC2",
+		.domains = TGL_AUX_E_TC2_IO_POWER_DOMAINS,
 		.ops = &icl_tc_phy_aux_power_well_ops,
 		.id = DISP_PW_ID_NONE,
 		{
@@ -3779,8 +3882,8 @@ static const struct i915_power_well_desc tgl_power_wells[] = {
 		},
 	},
 	{
-		.name = "AUX TC3",
-		.domains = TGL_AUX_TC3_IO_POWER_DOMAINS,
+		.name = "AUX F TC3",
+		.domains = TGL_AUX_F_TC3_IO_POWER_DOMAINS,
 		.ops = &icl_tc_phy_aux_power_well_ops,
 		.id = DISP_PW_ID_NONE,
 		{
@@ -3790,8 +3893,8 @@ static const struct i915_power_well_desc tgl_power_wells[] = {
 		},
 	},
 	{
-		.name = "AUX TC4",
-		.domains = TGL_AUX_TC4_IO_POWER_DOMAINS,
+		.name = "AUX G TC4",
+		.domains = TGL_AUX_G_TC4_IO_POWER_DOMAINS,
 		.ops = &icl_tc_phy_aux_power_well_ops,
 		.id = DISP_PW_ID_NONE,
 		{
@@ -3801,8 +3904,8 @@ static const struct i915_power_well_desc tgl_power_wells[] = {
 		},
 	},
 	{
-		.name = "AUX TC5",
-		.domains = TGL_AUX_TC5_IO_POWER_DOMAINS,
+		.name = "AUX H TC5",
+		.domains = TGL_AUX_H_TC5_IO_POWER_DOMAINS,
 		.ops = &icl_tc_phy_aux_power_well_ops,
 		.id = DISP_PW_ID_NONE,
 		{
@@ -3812,8 +3915,8 @@ static const struct i915_power_well_desc tgl_power_wells[] = {
 		},
 	},
 	{
-		.name = "AUX TC6",
-		.domains = TGL_AUX_TC6_IO_POWER_DOMAINS,
+		.name = "AUX I TC6",
+		.domains = TGL_AUX_I_TC6_IO_POWER_DOMAINS,
 		.ops = &icl_tc_phy_aux_power_well_ops,
 		.id = DISP_PW_ID_NONE,
 		{
@@ -3823,8 +3926,8 @@ static const struct i915_power_well_desc tgl_power_wells[] = {
 		},
 	},
 	{
-		.name = "AUX TBT1",
-		.domains = ICL_AUX_TBT1_IO_POWER_DOMAINS,
+		.name = "AUX D TBT1",
+		.domains = TGL_AUX_D_TBT1_IO_POWER_DOMAINS,
 		.ops = &hsw_power_well_ops,
 		.id = DISP_PW_ID_NONE,
 		{
@@ -3834,8 +3937,8 @@ static const struct i915_power_well_desc tgl_power_wells[] = {
 		},
 	},
 	{
-		.name = "AUX TBT2",
-		.domains = ICL_AUX_TBT2_IO_POWER_DOMAINS,
+		.name = "AUX E TBT2",
+		.domains = TGL_AUX_E_TBT2_IO_POWER_DOMAINS,
 		.ops = &hsw_power_well_ops,
 		.id = DISP_PW_ID_NONE,
 		{
@@ -3845,8 +3948,8 @@ static const struct i915_power_well_desc tgl_power_wells[] = {
 		},
 	},
 	{
-		.name = "AUX TBT3",
-		.domains = ICL_AUX_TBT3_IO_POWER_DOMAINS,
+		.name = "AUX F TBT3",
+		.domains = TGL_AUX_F_TBT3_IO_POWER_DOMAINS,
 		.ops = &hsw_power_well_ops,
 		.id = DISP_PW_ID_NONE,
 		{
@@ -3856,8 +3959,8 @@ static const struct i915_power_well_desc tgl_power_wells[] = {
 		},
 	},
 	{
-		.name = "AUX TBT4",
-		.domains = ICL_AUX_TBT4_IO_POWER_DOMAINS,
+		.name = "AUX G TBT4",
+		.domains = TGL_AUX_G_TBT4_IO_POWER_DOMAINS,
 		.ops = &hsw_power_well_ops,
 		.id = DISP_PW_ID_NONE,
 		{
@@ -3867,8 +3970,8 @@ static const struct i915_power_well_desc tgl_power_wells[] = {
 		},
 	},
 	{
-		.name = "AUX TBT5",
-		.domains = TGL_AUX_TBT5_IO_POWER_DOMAINS,
+		.name = "AUX H TBT5",
+		.domains = TGL_AUX_H_TBT5_IO_POWER_DOMAINS,
 		.ops = &hsw_power_well_ops,
 		.id = DISP_PW_ID_NONE,
 		{
@@ -3878,8 +3981,8 @@ static const struct i915_power_well_desc tgl_power_wells[] = {
 		},
 	},
 	{
-		.name = "AUX TBT6",
-		.domains = TGL_AUX_TBT6_IO_POWER_DOMAINS,
+		.name = "AUX I TBT6",
+		.domains = TGL_AUX_I_TBT6_IO_POWER_DOMAINS,
 		.ops = &hsw_power_well_ops,
 		.id = DISP_PW_ID_NONE,
 		{
@@ -3931,14 +4034,17 @@ static u32 get_allowed_dc_mask(const struct drm_i915_private *dev_priv,
 	int requested_dc;
 	int max_dc;
 
-	if (INTEL_GEN(dev_priv) >= 11) {
-		max_dc = 2;
+	if (INTEL_GEN(dev_priv) >= 12) {
+		max_dc = 4;
 		/*
 		 * DC9 has a separate HW flow from the rest of the DC states,
 		 * not depending on the DMC firmware. It's needed by system
 		 * suspend/resume, so allow it unconditionally.
 		 */
 		mask = DC_STATE_EN_DC9;
+	} else if (IS_GEN(dev_priv, 11)) {
+		max_dc = 2;
+		mask = DC_STATE_EN_DC9;
 	} else if (IS_GEN(dev_priv, 10) || IS_GEN9_BC(dev_priv)) {
 		max_dc = 2;
 		mask = 0;
@@ -3957,7 +4063,7 @@ static u32 get_allowed_dc_mask(const struct drm_i915_private *dev_priv,
 		requested_dc = enable_dc;
 	} else if (enable_dc == -1) {
 		requested_dc = max_dc;
-	} else if (enable_dc > max_dc && enable_dc <= 2) {
+	} else if (enable_dc > max_dc && enable_dc <= 4) {
 		DRM_DEBUG_KMS("Adjusting requested max DC state (%d->%d)\n",
 			      enable_dc, max_dc);
 		requested_dc = max_dc;
@@ -3966,10 +4072,20 @@ static u32 get_allowed_dc_mask(const struct drm_i915_private *dev_priv,
 		requested_dc = max_dc;
 	}
 
-	if (requested_dc > 1)
+	switch (requested_dc) {
+	case 4:
+		mask |= DC_STATE_EN_DC3CO | DC_STATE_EN_UPTO_DC6;
+		break;
+	case 3:
+		mask |= DC_STATE_EN_DC3CO | DC_STATE_EN_UPTO_DC5;
+		break;
+	case 2:
 		mask |= DC_STATE_EN_UPTO_DC6;
-	if (requested_dc > 0)
+		break;
+	case 1:
 		mask |= DC_STATE_EN_UPTO_DC5;
+		break;
+	}
 
 	DRM_DEBUG_KMS("Allowed DC state mask %02x\n", mask);
 
@@ -4030,6 +4146,9 @@ int intel_power_domains_init(struct drm_i915_private *dev_priv)
 	dev_priv->csr.allowed_dc_mask =
 		get_allowed_dc_mask(dev_priv, i915_modparams.enable_dc);
 
+	dev_priv->csr.target_dc_state =
+		sanitize_target_dc_state(dev_priv, DC_STATE_EN_UPTO_DC6);
+
 	BUILD_BUG_ON(POWER_DOMAIN_NUM > 64);
 
 	mutex_init(&power_domains->lock);
@@ -4896,6 +5015,9 @@ void intel_power_domains_init_hw(struct drm_i915_private *i915, bool resume)
 
 	power_domains->initializing = true;
 
+	/* Must happen before power domain init on VLV/CHV */
+	intel_update_rawclk(i915);
+
 	if (INTEL_GEN(i915) >= 11) {
 		icl_display_core_init(i915, resume);
 	} else if (IS_CANNONLAKE(i915)) {
@@ -5104,8 +5226,7 @@ static void intel_power_domains_dump_info(struct drm_i915_private *i915)
 
 		for_each_power_domain(domain, power_well->desc->domains)
 			DRM_DEBUG_DRIVER("  %-23s %d\n",
-					 intel_display_power_domain_str(i915,
-									domain),
+					 intel_display_power_domain_str(domain),
 					 power_domains->domain_use_count[domain]);
 	}
 }
diff --git a/drivers/gpu/drm/i915/display/intel_display_power.h b/drivers/gpu/drm/i915/display/intel_display_power.h
index a50605b8b1ad..1da04f3e0fb3 100644
--- a/drivers/gpu/drm/i915/display/intel_display_power.h
+++ b/drivers/gpu/drm/i915/display/intel_display_power.h
@@ -36,29 +36,20 @@ enum intel_display_power_domain {
 	POWER_DOMAIN_PORT_DDI_B_LANES,
 	POWER_DOMAIN_PORT_DDI_C_LANES,
 	POWER_DOMAIN_PORT_DDI_D_LANES,
-	POWER_DOMAIN_PORT_DDI_TC1_LANES = POWER_DOMAIN_PORT_DDI_D_LANES,
 	POWER_DOMAIN_PORT_DDI_E_LANES,
-	POWER_DOMAIN_PORT_DDI_TC2_LANES = POWER_DOMAIN_PORT_DDI_E_LANES,
 	POWER_DOMAIN_PORT_DDI_F_LANES,
-	POWER_DOMAIN_PORT_DDI_TC3_LANES = POWER_DOMAIN_PORT_DDI_F_LANES,
-	POWER_DOMAIN_PORT_DDI_TC4_LANES,
-	POWER_DOMAIN_PORT_DDI_TC5_LANES,
-	POWER_DOMAIN_PORT_DDI_TC6_LANES,
+	POWER_DOMAIN_PORT_DDI_G_LANES,
+	POWER_DOMAIN_PORT_DDI_H_LANES,
+	POWER_DOMAIN_PORT_DDI_I_LANES,
 	POWER_DOMAIN_PORT_DDI_A_IO,
 	POWER_DOMAIN_PORT_DDI_B_IO,
 	POWER_DOMAIN_PORT_DDI_C_IO,
 	POWER_DOMAIN_PORT_DDI_D_IO,
-	POWER_DOMAIN_PORT_DDI_TC1_IO = POWER_DOMAIN_PORT_DDI_D_IO,
 	POWER_DOMAIN_PORT_DDI_E_IO,
-	POWER_DOMAIN_PORT_DDI_TC2_IO = POWER_DOMAIN_PORT_DDI_E_IO,
 	POWER_DOMAIN_PORT_DDI_F_IO,
-	POWER_DOMAIN_PORT_DDI_TC3_IO = POWER_DOMAIN_PORT_DDI_F_IO,
 	POWER_DOMAIN_PORT_DDI_G_IO,
-	POWER_DOMAIN_PORT_DDI_TC4_IO = POWER_DOMAIN_PORT_DDI_G_IO,
 	POWER_DOMAIN_PORT_DDI_H_IO,
-	POWER_DOMAIN_PORT_DDI_TC5_IO = POWER_DOMAIN_PORT_DDI_H_IO,
 	POWER_DOMAIN_PORT_DDI_I_IO,
-	POWER_DOMAIN_PORT_DDI_TC6_IO = POWER_DOMAIN_PORT_DDI_I_IO,
 	POWER_DOMAIN_PORT_DSI,
 	POWER_DOMAIN_PORT_CRT,
 	POWER_DOMAIN_PORT_OTHER,
@@ -68,21 +59,19 @@ enum intel_display_power_domain {
 	POWER_DOMAIN_AUX_B,
 	POWER_DOMAIN_AUX_C,
 	POWER_DOMAIN_AUX_D,
-	POWER_DOMAIN_AUX_TC1 = POWER_DOMAIN_AUX_D,
 	POWER_DOMAIN_AUX_E,
-	POWER_DOMAIN_AUX_TC2 = POWER_DOMAIN_AUX_E,
 	POWER_DOMAIN_AUX_F,
-	POWER_DOMAIN_AUX_TC3 = POWER_DOMAIN_AUX_F,
-	POWER_DOMAIN_AUX_TC4,
-	POWER_DOMAIN_AUX_TC5,
-	POWER_DOMAIN_AUX_TC6,
+	POWER_DOMAIN_AUX_G,
+	POWER_DOMAIN_AUX_H,
+	POWER_DOMAIN_AUX_I,
 	POWER_DOMAIN_AUX_IO_A,
-	POWER_DOMAIN_AUX_TBT1,
-	POWER_DOMAIN_AUX_TBT2,
-	POWER_DOMAIN_AUX_TBT3,
-	POWER_DOMAIN_AUX_TBT4,
-	POWER_DOMAIN_AUX_TBT5,
-	POWER_DOMAIN_AUX_TBT6,
+	POWER_DOMAIN_AUX_C_TBT,
+	POWER_DOMAIN_AUX_D_TBT,
+	POWER_DOMAIN_AUX_E_TBT,
+	POWER_DOMAIN_AUX_F_TBT,
+	POWER_DOMAIN_AUX_G_TBT,
+	POWER_DOMAIN_AUX_H_TBT,
+	POWER_DOMAIN_AUX_I_TBT,
 	POWER_DOMAIN_GMBUS,
 	POWER_DOMAIN_MODESET,
 	POWER_DOMAIN_GT_IRQ,
@@ -111,6 +100,7 @@ enum i915_power_well_id {
 	SKL_DISP_PW_MISC_IO,
 	SKL_DISP_PW_1,
 	SKL_DISP_PW_2,
+	SKL_DISP_DC_OFF,
 };
 
 #define POWER_DOMAIN_PIPE(pipe) ((pipe) + POWER_DOMAIN_PIPE_A)
@@ -267,10 +257,11 @@ void intel_display_power_suspend_late(struct drm_i915_private *i915);
 void intel_display_power_resume_early(struct drm_i915_private *i915);
 void intel_display_power_suspend(struct drm_i915_private *i915);
 void intel_display_power_resume(struct drm_i915_private *i915);
+void intel_display_power_set_target_dc_state(struct drm_i915_private *dev_priv,
+					     u32 state);
 
 const char *
-intel_display_power_domain_str(struct drm_i915_private *i915,
-			       enum intel_display_power_domain domain);
+intel_display_power_domain_str(enum intel_display_power_domain domain);
 
 bool intel_display_power_is_enabled(struct drm_i915_private *dev_priv,
 				    enum intel_display_power_domain domain);
diff --git a/drivers/gpu/drm/i915/display/intel_display_types.h b/drivers/gpu/drm/i915/display/intel_display_types.h
index 449abaea619f..1a7334dbe802 100644
--- a/drivers/gpu/drm/i915/display/intel_display_types.h
+++ b/drivers/gpu/drm/i915/display/intel_display_types.h
@@ -128,7 +128,8 @@ struct intel_encoder {
 
 	enum intel_output_type type;
 	enum port port;
-	unsigned int cloneable;
+	u16 cloneable;
+	u8 pipe_mask;
 	enum intel_hotplug_state (*hotplug)(struct intel_encoder *encoder,
 					    struct intel_connector *connector,
 					    bool irq_received);
@@ -187,7 +188,6 @@ struct intel_encoder {
 	 * device interrupts are disabled.
 	 */
 	void (*suspend)(struct intel_encoder *);
-	int crtc_mask;
 	enum hpd_pin hpd_pin;
 	enum intel_display_power_domain power_domain;
 	/* for communication with audio component; protected by av_mutex */
@@ -388,6 +388,13 @@ struct intel_hdcp {
 	wait_queue_head_t cp_irq_queue;
 	atomic_t cp_irq_count;
 	int cp_irq_count_cached;
+
+	/*
+	 * HDCP register access for gen12+ need the transcoder associated.
+	 * Transcoder attached to the connector could be changed at modeset.
+	 * Hence caching the transcoder here.
+	 */
+	enum transcoder cpu_transcoder;
 };
 
 struct intel_connector {
@@ -481,9 +488,9 @@ struct intel_atomic_state {
 	 * but the converse is not necessarily true; simply changing a mode may
 	 * not flip the final active status of any CRTC's
 	 */
-	unsigned int active_pipe_changes;
+	u8 active_pipe_changes;
 
-	unsigned int active_crtcs;
+	u8 active_pipes;
 	/* minimum acceptable cdclk for each pipe */
 	int min_cdclk[I915_MAX_PIPES];
 	/* minimum acceptable voltage level for each pipe */
@@ -499,6 +506,14 @@ struct intel_atomic_state {
 
 	bool rps_interactive;
 
+	/*
+	 * active_pipes
+	 * min_cdclk[]
+	 * min_voltage_level[]
+	 * cdclk.*
+	 */
+	bool global_state_changed;
+
 	/* Gen9+ only */
 	struct skl_ddb_values wm_results;
 
@@ -552,24 +567,24 @@ struct intel_plane_state {
 	int scaler_id;
 
 	/*
-	 * linked_plane:
+	 * planar_linked_plane:
 	 *
 	 * ICL planar formats require 2 planes that are updated as pairs.
 	 * This member is used to make sure the other plane is also updated
 	 * when required, and for update_slave() to find the correct
 	 * plane_state to pass as argument.
 	 */
-	struct intel_plane *linked_plane;
+	struct intel_plane *planar_linked_plane;
 
 	/*
-	 * slave:
+	 * planar_slave:
 	 * If set don't update use the linked plane's state for updating
 	 * this plane during atomic commit with the update_slave() callback.
 	 *
 	 * It's also used by the watermark code to ignore wm calculations on
 	 * this plane. They're calculated by the linked plane's wm code.
 	 */
-	u32 slave;
+	u32 planar_slave;
 
 	struct drm_intel_sprite_colorkey ckey;
 };
@@ -759,8 +774,8 @@ struct intel_crtc_state {
 	bool update_pipe; /* can a fast modeset be performed? */
 	bool disable_cxsr;
 	bool update_wm_pre, update_wm_post; /* watermarks are updated */
-	bool fb_changed; /* fb on any of the planes is changed */
 	bool fifo_changed; /* FIFO split is changed */
+	bool preload_luts;
 
 	/* Pipe source size (ie. panel fitter input size)
 	 * All planes will be positioned inside this space,
@@ -864,6 +879,7 @@ struct intel_crtc_state {
 
 	bool has_psr;
 	bool has_psr2;
+	u32 dc3co_exitline;
 
 	/*
 	 * Frequence the dpll for the port should run at. Differs from the
@@ -925,6 +941,8 @@ struct intel_crtc_state {
 
 	struct intel_crtc_wm_state wm;
 
+	int min_cdclk[I915_MAX_PLANES];
+
 	u32 data_rate[I915_MAX_PLANES];
 
 	/* Gamma mode programmed on the pipe */
@@ -979,11 +997,17 @@ struct intel_crtc_state {
 		bool dsc_split;
 		u16 compressed_bpp;
 		u8 slice_count;
-	} dsc_params;
-	struct drm_dsc_config dp_dsc_cfg;
+		struct drm_dsc_config config;
+	} dsc;
 
 	/* Forward Error correction State */
 	bool fec_enable;
+
+	/* Pointer to master transcoder in case of tiled displays */
+	enum transcoder master_transcoder;
+
+	/* Bitmask to indicate slaves attached */
+	u8 sync_mode_slaves_mask;
 };
 
 struct intel_crtc {
@@ -1026,6 +1050,9 @@ struct intel_crtc {
 
 	/* scalers available on this crtc */
 	int num_scalers;
+
+	/* per pipe DSB related info */
+	struct intel_dsb dsb;
 };
 
 struct intel_plane {
@@ -1061,6 +1088,8 @@ struct intel_plane {
 	bool (*get_hw_state)(struct intel_plane *plane, enum pipe *pipe);
 	int (*check_plane)(struct intel_crtc_state *crtc_state,
 			   struct intel_plane_state *plane_state);
+	int (*min_cdclk)(const struct intel_crtc_state *crtc_state,
+			 const struct intel_plane_state *plane_state);
 };
 
 struct intel_watermark_params {
@@ -1176,6 +1205,7 @@ struct intel_dp {
 	/* sink or branch descriptor */
 	struct drm_dp_desc desc;
 	struct drm_dp_aux aux;
+	u32 aux_busy_last_status;
 	u8 train_set[4];
 	int panel_power_up_delay;
 	int panel_power_down_delay;
@@ -1211,6 +1241,15 @@ struct intel_dp {
 	bool can_mst; /* this port supports mst */
 	bool is_mst;
 	int active_mst_links;
+
+	/*
+	 * DP_TP_* registers may be either on port or transcoder register space.
+	 */
+	struct {
+		i915_reg_t dp_tp_ctl;
+		i915_reg_t dp_tp_status;
+	} regs;
+
 	/* connector directly attached - won't be use for modeset in mst world */
 	struct intel_connector *attached_connector;
 
@@ -1269,6 +1308,7 @@ struct intel_digital_port {
 	char tc_port_name[8];
 	enum tc_port_mode tc_mode;
 	enum phy_fia tc_phy_fia;
+	u8 tc_phy_fia_idx;
 
 	void (*write_infoframe)(struct intel_encoder *encoder,
 				const struct intel_crtc_state *crtc_state,
@@ -1509,7 +1549,7 @@ intel_wait_for_vblank(struct drm_i915_private *dev_priv, enum pipe pipe)
 	drm_wait_one_vblank(&dev_priv->drm, pipe);
 }
 static inline void
-intel_wait_for_vblank_if_active(struct drm_i915_private *dev_priv, int pipe)
+intel_wait_for_vblank_if_active(struct drm_i915_private *dev_priv, enum pipe pipe)
 {
 	const struct intel_crtc *crtc = intel_get_crtc_for_pipe(dev_priv, pipe);
 
diff --git a/drivers/gpu/drm/i915/display/intel_dp.c b/drivers/gpu/drm/i915/display/intel_dp.c
index 921ad0a2f7ba..c61ac0c3acb5 100644
--- a/drivers/gpu/drm/i915/display/intel_dp.c
+++ b/drivers/gpu/drm/i915/display/intel_dp.c
@@ -68,18 +68,13 @@
 
 #define DP_DPRX_ESI_LEN 14
 
-/* DP DSC small joiner has 2 FIFOs each of 640 x 6 bytes */
-#define DP_DSC_MAX_SMALL_JOINER_RAM_BUFFER	61440
-#define DP_DSC_MIN_SUPPORTED_BPC		8
-#define DP_DSC_MAX_SUPPORTED_BPC		10
-
 /* DP DSC throughput values used for slice count calculations KPixels/s */
 #define DP_DSC_PEAK_PIXEL_RATE			2720000
 #define DP_DSC_MAX_ENC_THROUGHPUT_0		340000
 #define DP_DSC_MAX_ENC_THROUGHPUT_1		400000
 
-/* DP DSC FEC Overhead factor = (100 - 2.4)/100 */
-#define DP_DSC_FEC_OVERHEAD_FACTOR		976
+/* DP DSC FEC Overhead factor = 1/(0.972261) */
+#define DP_DSC_FEC_OVERHEAD_FACTOR		972261
 
 /* Compliance test status bits  */
 #define INTEL_DP_RESOLUTION_SHIFT_MASK	0
@@ -494,6 +489,127 @@ int intel_dp_get_link_train_fallback_values(struct intel_dp *intel_dp,
 	return 0;
 }
 
+u32 intel_dp_mode_to_fec_clock(u32 mode_clock)
+{
+	return div_u64(mul_u32_u32(mode_clock, 1000000U),
+		       DP_DSC_FEC_OVERHEAD_FACTOR);
+}
+
+static int
+small_joiner_ram_size_bits(struct drm_i915_private *i915)
+{
+	if (INTEL_GEN(i915) >= 11)
+		return 7680 * 8;
+	else
+		return 6144 * 8;
+}
+
+static u16 intel_dp_dsc_get_output_bpp(struct drm_i915_private *i915,
+				       u32 link_clock, u32 lane_count,
+				       u32 mode_clock, u32 mode_hdisplay)
+{
+	u32 bits_per_pixel, max_bpp_small_joiner_ram;
+	int i;
+
+	/*
+	 * Available Link Bandwidth(Kbits/sec) = (NumberOfLanes)*
+	 * (LinkSymbolClock)* 8 * (TimeSlotsPerMTP)
+	 * for SST -> TimeSlotsPerMTP is 1,
+	 * for MST -> TimeSlotsPerMTP has to be calculated
+	 */
+	bits_per_pixel = (link_clock * lane_count * 8) /
+			 intel_dp_mode_to_fec_clock(mode_clock);
+	DRM_DEBUG_KMS("Max link bpp: %u\n", bits_per_pixel);
+
+	/* Small Joiner Check: output bpp <= joiner RAM (bits) / Horiz. width */
+	max_bpp_small_joiner_ram = small_joiner_ram_size_bits(i915) /
+		mode_hdisplay;
+	DRM_DEBUG_KMS("Max small joiner bpp: %u\n", max_bpp_small_joiner_ram);
+
+	/*
+	 * Greatest allowed DSC BPP = MIN (output BPP from available Link BW
+	 * check, output bpp from small joiner RAM check)
+	 */
+	bits_per_pixel = min(bits_per_pixel, max_bpp_small_joiner_ram);
+
+	/* Error out if the max bpp is less than smallest allowed valid bpp */
+	if (bits_per_pixel < valid_dsc_bpp[0]) {
+		DRM_DEBUG_KMS("Unsupported BPP %u, min %u\n",
+			      bits_per_pixel, valid_dsc_bpp[0]);
+		return 0;
+	}
+
+	/* Find the nearest match in the array of known BPPs from VESA */
+	for (i = 0; i < ARRAY_SIZE(valid_dsc_bpp) - 1; i++) {
+		if (bits_per_pixel < valid_dsc_bpp[i + 1])
+			break;
+	}
+	bits_per_pixel = valid_dsc_bpp[i];
+
+	/*
+	 * Compressed BPP in U6.4 format so multiply by 16, for Gen 11,
+	 * fractional part is 0
+	 */
+	return bits_per_pixel << 4;
+}
+
+static u8 intel_dp_dsc_get_slice_count(struct intel_dp *intel_dp,
+				       int mode_clock, int mode_hdisplay)
+{
+	u8 min_slice_count, i;
+	int max_slice_width;
+
+	if (mode_clock <= DP_DSC_PEAK_PIXEL_RATE)
+		min_slice_count = DIV_ROUND_UP(mode_clock,
+					       DP_DSC_MAX_ENC_THROUGHPUT_0);
+	else
+		min_slice_count = DIV_ROUND_UP(mode_clock,
+					       DP_DSC_MAX_ENC_THROUGHPUT_1);
+
+	max_slice_width = drm_dp_dsc_sink_max_slice_width(intel_dp->dsc_dpcd);
+	if (max_slice_width < DP_DSC_MIN_SLICE_WIDTH_VALUE) {
+		DRM_DEBUG_KMS("Unsupported slice width %d by DP DSC Sink device\n",
+			      max_slice_width);
+		return 0;
+	}
+	/* Also take into account max slice width */
+	min_slice_count = min_t(u8, min_slice_count,
+				DIV_ROUND_UP(mode_hdisplay,
+					     max_slice_width));
+
+	/* Find the closest match to the valid slice count values */
+	for (i = 0; i < ARRAY_SIZE(valid_dsc_slicecount); i++) {
+		if (valid_dsc_slicecount[i] >
+		    drm_dp_dsc_sink_max_slice_count(intel_dp->dsc_dpcd,
+						    false))
+			break;
+		if (min_slice_count  <= valid_dsc_slicecount[i])
+			return valid_dsc_slicecount[i];
+	}
+
+	DRM_DEBUG_KMS("Unsupported Slice Count %d\n", min_slice_count);
+	return 0;
+}
+
+static bool intel_dp_hdisplay_bad(struct drm_i915_private *dev_priv,
+				  int hdisplay)
+{
+	/*
+	 * Older platforms don't like hdisplay==4096 with DP.
+	 *
+	 * On ILK/SNB/IVB the pipe seems to be somewhat running (scanline
+	 * and frame counter increment), but we don't get vblank interrupts,
+	 * and the pipe underruns immediately. The link also doesn't seem
+	 * to get trained properly.
+	 *
+	 * On CHV the vblank interrupts don't seem to disappear but
+	 * otherwise the symptoms are similar.
+	 *
+	 * TODO: confirm the behaviour on HSW+
+	 */
+	return hdisplay == 4096 && !HAS_DDI(dev_priv);
+}
+
 static enum drm_mode_status
 intel_dp_mode_valid(struct drm_connector *connector,
 		    struct drm_display_mode *mode)
@@ -529,6 +645,9 @@ intel_dp_mode_valid(struct drm_connector *connector,
 	max_rate = intel_dp_max_data_rate(max_link_clock, max_lanes);
 	mode_rate = intel_dp_link_required(target_clock, 18);
 
+	if (intel_dp_hdisplay_bad(dev_priv, mode->hdisplay))
+		return MODE_H_ILLEGAL;
+
 	/*
 	 * Output bpp is stored in 6.4 format so right shift by 4 to get the
 	 * integer value since we support only integer values of bpp.
@@ -543,7 +662,8 @@ intel_dp_mode_valid(struct drm_connector *connector,
 								true);
 		} else if (drm_dp_sink_supports_fec(intel_dp->fec_capable)) {
 			dsc_max_output_bpp =
-				intel_dp_dsc_get_output_bpp(max_link_clock,
+				intel_dp_dsc_get_output_bpp(dev_priv,
+							    max_link_clock,
 							    max_lanes,
 							    target_clock,
 							    mode->hdisplay) >> 4;
@@ -564,7 +684,7 @@ intel_dp_mode_valid(struct drm_connector *connector,
 	if (mode->flags & DRM_MODE_FLAG_DBLCLK)
 		return MODE_H_ILLEGAL;
 
-	return MODE_OK;
+	return intel_mode_valid_max_plane_size(dev_priv, mode);
 }
 
 u32 intel_dp_pack_aux(const u8 *src, int src_bytes)
@@ -641,12 +761,14 @@ vlv_power_sequencer_kick(struct intel_dp *intel_dp)
 	u32 DP;
 
 	if (WARN(I915_READ(intel_dp->output_reg) & DP_PORT_EN,
-		 "skipping pipe %c power sequencer kick due to port %c being active\n",
-		 pipe_name(pipe), port_name(intel_dig_port->base.port)))
+		 "skipping pipe %c power sequencer kick due to [ENCODER:%d:%s] being active\n",
+		 pipe_name(pipe), intel_dig_port->base.base.base.id,
+		 intel_dig_port->base.base.name))
 		return;
 
-	DRM_DEBUG_KMS("kicking pipe %c power sequencer for port %c\n",
-		      pipe_name(pipe), port_name(intel_dig_port->base.port));
+	DRM_DEBUG_KMS("kicking pipe %c power sequencer for [ENCODER:%d:%s]\n",
+		      pipe_name(pipe), intel_dig_port->base.base.base.id,
+		      intel_dig_port->base.base.name);
 
 	/* Preserve the BIOS-computed detected bit. This is
 	 * supposed to be read-only.
@@ -764,9 +886,10 @@ vlv_power_sequencer_pipe(struct intel_dp *intel_dp)
 	vlv_steal_power_sequencer(dev_priv, pipe);
 	intel_dp->pps_pipe = pipe;
 
-	DRM_DEBUG_KMS("picked pipe %c power sequencer for port %c\n",
+	DRM_DEBUG_KMS("picked pipe %c power sequencer for [ENCODER:%d:%s]\n",
 		      pipe_name(intel_dp->pps_pipe),
-		      port_name(intel_dig_port->base.port));
+		      intel_dig_port->base.base.base.id,
+		      intel_dig_port->base.base.name);
 
 	/* init power sequencer on this pipe and port */
 	intel_dp_init_panel_power_sequencer(intel_dp);
@@ -874,13 +997,16 @@ vlv_initial_power_sequencer_setup(struct intel_dp *intel_dp)
 
 	/* didn't find one? just let vlv_power_sequencer_pipe() pick one when needed */
 	if (intel_dp->pps_pipe == INVALID_PIPE) {
-		DRM_DEBUG_KMS("no initial power sequencer for port %c\n",
-			      port_name(port));
+		DRM_DEBUG_KMS("no initial power sequencer for [ENCODER:%d:%s]\n",
+			      intel_dig_port->base.base.base.id,
+			      intel_dig_port->base.base.name);
 		return;
 	}
 
-	DRM_DEBUG_KMS("initial power sequencer for port %c: pipe %c\n",
-		      port_name(port), pipe_name(intel_dp->pps_pipe));
+	DRM_DEBUG_KMS("initial power sequencer for [ENCODER:%d:%s]: pipe %c\n",
+		      intel_dig_port->base.base.base.id,
+		      intel_dig_port->base.base.name,
+		      pipe_name(intel_dp->pps_pipe));
 
 	intel_dp_init_panel_power_sequencer(intel_dp);
 	intel_dp_init_panel_power_sequencer_registers(intel_dp, false);
@@ -1053,18 +1179,20 @@ intel_dp_aux_wait_done(struct intel_dp *intel_dp)
 {
 	struct drm_i915_private *i915 = dp_to_i915(intel_dp);
 	i915_reg_t ch_ctl = intel_dp->aux_ch_ctl_reg(intel_dp);
+	const unsigned int timeout_ms = 10;
 	u32 status;
 	bool done;
 
 #define C (((status = intel_uncore_read_notrace(&i915->uncore, ch_ctl)) & DP_AUX_CH_CTL_SEND_BUSY) == 0)
 	done = wait_event_timeout(i915->gmbus_wait_queue, C,
-				  msecs_to_jiffies_timeout(10));
+				  msecs_to_jiffies_timeout(timeout_ms));
 
 	/* just trace the final value */
 	trace_i915_reg_rw(false, ch_ctl, status, sizeof(status), true);
 
 	if (!done)
-		DRM_ERROR("dp aux hw did not signal timeout!\n");
+		DRM_ERROR("%s did not complete or timeout within %ums (status 0x%08x)\n",
+			  intel_dp->aux.name, timeout_ms, status);
 #undef C
 
 	return status;
@@ -1165,6 +1293,9 @@ static u32 skl_get_aux_send_ctl(struct intel_dp *intel_dp,
 				u32 unused)
 {
 	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
+	struct drm_i915_private *i915 =
+			to_i915(intel_dig_port->base.base.dev);
+	enum phy phy = intel_port_to_phy(i915, intel_dig_port->base.port);
 	u32 ret;
 
 	ret = DP_AUX_CH_CTL_SEND_BUSY |
@@ -1177,7 +1308,8 @@ static u32 skl_get_aux_send_ctl(struct intel_dp *intel_dp,
 	      DP_AUX_CH_CTL_FW_SYNC_PULSE_SKL(32) |
 	      DP_AUX_CH_CTL_SYNC_PULSE_SKL(32);
 
-	if (intel_dig_port->tc_mode == TC_PORT_TBT_ALT)
+	if (intel_phy_is_tc(i915, phy) &&
+	    intel_dig_port->tc_mode == TC_PORT_TBT_ALT)
 		ret |= DP_AUX_CH_CTL_TBT_IO;
 
 	return ret;
@@ -1243,13 +1375,12 @@ intel_dp_aux_xfer(struct intel_dp *intel_dp,
 	trace_i915_reg_rw(false, ch_ctl, status, sizeof(status), true);
 
 	if (try == 3) {
-		static u32 last_status = -1;
 		const u32 status = intel_uncore_read(uncore, ch_ctl);
 
-		if (status != last_status) {
+		if (status != intel_dp->aux_busy_last_status) {
 			WARN(1, "dp_aux_ch not started status 0x%08x\n",
 			     status);
-			last_status = status;
+			intel_dp->aux_busy_last_status = status;
 		}
 
 		ret = -EBUSY;
@@ -1541,6 +1672,7 @@ static i915_reg_t skl_aux_ctl_reg(struct intel_dp *intel_dp)
 	case AUX_CH_D:
 	case AUX_CH_E:
 	case AUX_CH_F:
+	case AUX_CH_G:
 		return DP_AUX_CH_CTL(aux_ch);
 	default:
 		MISSING_CASE(aux_ch);
@@ -1561,6 +1693,7 @@ static i915_reg_t skl_aux_data_reg(struct intel_dp *intel_dp, int index)
 	case AUX_CH_D:
 	case AUX_CH_E:
 	case AUX_CH_F:
+	case AUX_CH_G:
 		return DP_AUX_CH_DATA(aux_ch, index);
 	default:
 		MISSING_CASE(aux_ch);
@@ -1739,8 +1872,14 @@ static bool intel_dp_source_supports_fec(struct intel_dp *intel_dp,
 {
 	struct drm_i915_private *dev_priv = dp_to_i915(intel_dp);
 
-	return INTEL_GEN(dev_priv) >= 11 &&
-		pipe_config->cpu_transcoder != TRANSCODER_A;
+	/* On TGL, FEC is supported on all Pipes */
+	if (INTEL_GEN(dev_priv) >= 12)
+		return true;
+
+	if (IS_GEN(dev_priv, 11) && pipe_config->cpu_transcoder != TRANSCODER_A)
+		return true;
+
+	return false;
 }
 
 static bool intel_dp_supports_fec(struct intel_dp *intel_dp,
@@ -1755,8 +1894,18 @@ static bool intel_dp_source_supports_dsc(struct intel_dp *intel_dp,
 {
 	struct drm_i915_private *dev_priv = dp_to_i915(intel_dp);
 
-	return INTEL_GEN(dev_priv) >= 10 &&
-		pipe_config->cpu_transcoder != TRANSCODER_A;
+	if (!INTEL_INFO(dev_priv)->display.has_dsc)
+		return false;
+
+	/* On TGL, DSC is supported on all Pipes */
+	if (INTEL_GEN(dev_priv) >= 12)
+		return true;
+
+	if (INTEL_GEN(dev_priv) >= 10 &&
+	    pipe_config->cpu_transcoder != TRANSCODER_A)
+		return true;
+
+	return false;
 }
 
 static bool intel_dp_supports_dsc(struct intel_dp *intel_dp,
@@ -1915,11 +2064,17 @@ static int intel_dp_dsc_compute_config(struct intel_dp *intel_dp,
 	if (!intel_dp_supports_dsc(intel_dp, pipe_config))
 		return -EINVAL;
 
-	dsc_max_bpc = min_t(u8, DP_DSC_MAX_SUPPORTED_BPC,
-			    conn_state->max_requested_bpc);
+	/* Max DSC Input BPC for ICL is 10 and for TGL+ is 12 */
+	if (INTEL_GEN(dev_priv) >= 12)
+		dsc_max_bpc = min_t(u8, 12, conn_state->max_requested_bpc);
+	else
+		dsc_max_bpc = min_t(u8, 10,
+				    conn_state->max_requested_bpc);
 
 	pipe_bpp = intel_dp_dsc_compute_bpp(intel_dp, dsc_max_bpc);
-	if (pipe_bpp < DP_DSC_MIN_SUPPORTED_BPC * 3) {
+
+	/* Min Input BPC for ICL+ is 8 */
+	if (pipe_bpp < 8 * 3) {
 		DRM_DEBUG_KMS("No DSC support for less than 8bpc\n");
 		return -EINVAL;
 	}
@@ -1934,10 +2089,10 @@ static int intel_dp_dsc_compute_config(struct intel_dp *intel_dp,
 	pipe_config->lane_count = limits->max_lane_count;
 
 	if (intel_dp_is_edp(intel_dp)) {
-		pipe_config->dsc_params.compressed_bpp =
+		pipe_config->dsc.compressed_bpp =
 			min_t(u16, drm_edp_dsc_sink_output_bpp(intel_dp->dsc_dpcd) >> 4,
 			      pipe_config->pipe_bpp);
-		pipe_config->dsc_params.slice_count =
+		pipe_config->dsc.slice_count =
 			drm_dp_dsc_sink_max_slice_count(intel_dp->dsc_dpcd,
 							true);
 	} else {
@@ -1945,7 +2100,8 @@ static int intel_dp_dsc_compute_config(struct intel_dp *intel_dp,
 		u8 dsc_dp_slice_count;
 
 		dsc_max_output_bpp =
-			intel_dp_dsc_get_output_bpp(pipe_config->port_clock,
+			intel_dp_dsc_get_output_bpp(dev_priv,
+						    pipe_config->port_clock,
 						    pipe_config->lane_count,
 						    adjusted_mode->crtc_clock,
 						    adjusted_mode->crtc_hdisplay);
@@ -1957,10 +2113,10 @@ static int intel_dp_dsc_compute_config(struct intel_dp *intel_dp,
 			DRM_DEBUG_KMS("Compressed BPP/Slice Count not supported\n");
 			return -EINVAL;
 		}
-		pipe_config->dsc_params.compressed_bpp = min_t(u16,
+		pipe_config->dsc.compressed_bpp = min_t(u16,
 							       dsc_max_output_bpp >> 4,
 							       pipe_config->pipe_bpp);
-		pipe_config->dsc_params.slice_count = dsc_dp_slice_count;
+		pipe_config->dsc.slice_count = dsc_dp_slice_count;
 	}
 	/*
 	 * VDSC engine operates at 1 Pixel per clock, so if peak pixel rate
@@ -1968,8 +2124,8 @@ static int intel_dp_dsc_compute_config(struct intel_dp *intel_dp,
 	 * then we need to use 2 VDSC instances.
 	 */
 	if (adjusted_mode->crtc_clock > dev_priv->max_cdclk_freq) {
-		if (pipe_config->dsc_params.slice_count > 1) {
-			pipe_config->dsc_params.dsc_split = true;
+		if (pipe_config->dsc.slice_count > 1) {
+			pipe_config->dsc.dsc_split = true;
 		} else {
 			DRM_DEBUG_KMS("Cannot split stream to use 2 VDSC instances\n");
 			return -EINVAL;
@@ -1981,16 +2137,16 @@ static int intel_dp_dsc_compute_config(struct intel_dp *intel_dp,
 		DRM_DEBUG_KMS("Cannot compute valid DSC parameters for Input Bpp = %d "
 			      "Compressed BPP = %d\n",
 			      pipe_config->pipe_bpp,
-			      pipe_config->dsc_params.compressed_bpp);
+			      pipe_config->dsc.compressed_bpp);
 		return ret;
 	}
 
-	pipe_config->dsc_params.compression_enable = true;
+	pipe_config->dsc.compression_enable = true;
 	DRM_DEBUG_KMS("DP DSC computed with Input Bpp = %d "
 		      "Compressed Bpp = %d Slice Count = %d\n",
 		      pipe_config->pipe_bpp,
-		      pipe_config->dsc_params.compressed_bpp,
-		      pipe_config->dsc_params.slice_count);
+		      pipe_config->dsc.compressed_bpp,
+		      pipe_config->dsc.slice_count);
 
 	return 0;
 }
@@ -2064,15 +2220,15 @@ intel_dp_compute_link_config(struct intel_encoder *encoder,
 			return ret;
 	}
 
-	if (pipe_config->dsc_params.compression_enable) {
+	if (pipe_config->dsc.compression_enable) {
 		DRM_DEBUG_KMS("DP lane count %d clock %d Input bpp %d Compressed bpp %d\n",
 			      pipe_config->lane_count, pipe_config->port_clock,
 			      pipe_config->pipe_bpp,
-			      pipe_config->dsc_params.compressed_bpp);
+			      pipe_config->dsc.compressed_bpp);
 
 		DRM_DEBUG_KMS("DP link rate required %i available %i\n",
 			      intel_dp_link_required(adjusted_mode->crtc_clock,
-						     pipe_config->dsc_params.compressed_bpp),
+						     pipe_config->dsc.compressed_bpp),
 			      intel_dp_max_data_rate(pipe_config->port_clock,
 						     pipe_config->lane_count));
 	} else {
@@ -2127,6 +2283,16 @@ bool intel_dp_limited_color_range(const struct intel_crtc_state *crtc_state,
 	const struct drm_display_mode *adjusted_mode =
 		&crtc_state->base.adjusted_mode;
 
+	/*
+	 * Our YCbCr output is always limited range.
+	 * crtc_state->limited_color_range only applies to RGB,
+	 * and it must never be set for YCbCr or we risk setting
+	 * some conflicting bits in PIPECONF which will mess up
+	 * the colors on the monitor.
+	 */
+	if (crtc_state->output_format != INTEL_OUTPUT_FORMAT_RGB)
+		return false;
+
 	if (intel_conn_state->broadcast_rgb == INTEL_BROADCAST_RGB_AUTO) {
 		/*
 		 * See:
@@ -2164,6 +2330,7 @@ intel_dp_compute_config(struct intel_encoder *encoder,
 		pipe_config->has_pch_encoder = true;
 
 	pipe_config->output_format = INTEL_OUTPUT_FORMAT_RGB;
+
 	if (lspcon->active)
 		lspcon_ycbcr420_config(&intel_connector->base, pipe_config);
 	else
@@ -2209,6 +2376,9 @@ intel_dp_compute_config(struct intel_encoder *encoder,
 	if (adjusted_mode->flags & DRM_MODE_FLAG_DBLCLK)
 		return -EINVAL;
 
+	if (intel_dp_hdisplay_bad(dev_priv, adjusted_mode->crtc_hdisplay))
+		return -EINVAL;
+
 	ret = intel_dp_compute_link_config(encoder, pipe_config, conn_state);
 	if (ret < 0)
 		return ret;
@@ -2216,8 +2386,8 @@ intel_dp_compute_config(struct intel_encoder *encoder,
 	pipe_config->limited_color_range =
 		intel_dp_limited_color_range(pipe_config, conn_state);
 
-	if (pipe_config->dsc_params.compression_enable)
-		output_bpp = pipe_config->dsc_params.compressed_bpp;
+	if (pipe_config->dsc.compression_enable)
+		output_bpp = pipe_config->dsc.compressed_bpp;
 	else
 		output_bpp = intel_dp_output_bpp(pipe_config, pipe_config->pipe_bpp);
 
@@ -2226,7 +2396,7 @@ intel_dp_compute_config(struct intel_encoder *encoder,
 			       adjusted_mode->crtc_clock,
 			       pipe_config->port_clock,
 			       &pipe_config->dp_m_n,
-			       constant_n);
+			       constant_n, pipe_config->fec_enable);
 
 	if (intel_connector->panel.downclock_mode != NULL &&
 		dev_priv->drrs.type == SEAMLESS_DRRS_SUPPORT) {
@@ -2236,7 +2406,7 @@ intel_dp_compute_config(struct intel_encoder *encoder,
 					       intel_connector->panel.downclock_mode->clock,
 					       pipe_config->port_clock,
 					       &pipe_config->dp_m2_n2,
-					       constant_n);
+					       constant_n, pipe_config->fec_enable);
 	}
 
 	if (!HAS_DDI(dev_priv))
@@ -2244,6 +2414,9 @@ intel_dp_compute_config(struct intel_encoder *encoder,
 
 	intel_psr_compute_config(intel_dp, pipe_config);
 
+	intel_hdcp_transcoder_config(intel_connector,
+				     pipe_config->cpu_transcoder);
+
 	return 0;
 }
 
@@ -2271,6 +2444,9 @@ static void intel_dp_prepare(struct intel_encoder *encoder,
 				 intel_crtc_has_type(pipe_config,
 						     INTEL_OUTPUT_DP_MST));
 
+	intel_dp->regs.dp_tp_ctl = DP_TP_CTL(port);
+	intel_dp->regs.dp_tp_status = DP_TP_STATUS(port);
+
 	/*
 	 * There are four kinds of DP registers:
 	 *
@@ -2472,8 +2648,9 @@ static bool edp_panel_vdd_on(struct intel_dp *intel_dp)
 	intel_display_power_get(dev_priv,
 				intel_aux_power_domain(intel_dig_port));
 
-	DRM_DEBUG_KMS("Turning eDP port %c VDD on\n",
-		      port_name(intel_dig_port->base.port));
+	DRM_DEBUG_KMS("Turning [ENCODER:%d:%s] VDD on\n",
+		      intel_dig_port->base.base.base.id,
+		      intel_dig_port->base.base.name);
 
 	if (!edp_have_panel_power(intel_dp))
 		wait_panel_power_cycle(intel_dp);
@@ -2492,8 +2669,9 @@ static bool edp_panel_vdd_on(struct intel_dp *intel_dp)
 	 * If the panel wasn't on, delay before accessing aux channel
 	 */
 	if (!edp_have_panel_power(intel_dp)) {
-		DRM_DEBUG_KMS("eDP port %c panel power wasn't enabled\n",
-			      port_name(intel_dig_port->base.port));
+		DRM_DEBUG_KMS("[ENCODER:%d:%s] panel power wasn't enabled\n",
+			      intel_dig_port->base.base.base.id,
+			      intel_dig_port->base.base.name);
 		msleep(intel_dp->panel_power_up_delay);
 	}
 
@@ -2518,8 +2696,9 @@ void intel_edp_panel_vdd_on(struct intel_dp *intel_dp)
 	vdd = false;
 	with_pps_lock(intel_dp, wakeref)
 		vdd = edp_panel_vdd_on(intel_dp);
-	I915_STATE_WARN(!vdd, "eDP port %c VDD already requested on\n",
-	     port_name(dp_to_dig_port(intel_dp)->base.port));
+	I915_STATE_WARN(!vdd, "[ENCODER:%d:%s] VDD already requested on\n",
+			dp_to_dig_port(intel_dp)->base.base.base.id,
+			dp_to_dig_port(intel_dp)->base.base.name);
 }
 
 static void edp_panel_vdd_off_sync(struct intel_dp *intel_dp)
@@ -2537,8 +2716,9 @@ static void edp_panel_vdd_off_sync(struct intel_dp *intel_dp)
 	if (!edp_have_panel_vdd(intel_dp))
 		return;
 
-	DRM_DEBUG_KMS("Turning eDP port %c VDD off\n",
-		      port_name(intel_dig_port->base.port));
+	DRM_DEBUG_KMS("Turning [ENCODER:%d:%s] VDD off\n",
+		      intel_dig_port->base.base.base.id,
+		      intel_dig_port->base.base.name);
 
 	pp = ironlake_get_pp_control(intel_dp);
 	pp &= ~EDP_FORCE_VDD;
@@ -2600,8 +2780,9 @@ static void edp_panel_vdd_off(struct intel_dp *intel_dp, bool sync)
 	if (!intel_dp_is_edp(intel_dp))
 		return;
 
-	I915_STATE_WARN(!intel_dp->want_panel_vdd, "eDP port %c VDD not forced on",
-	     port_name(dp_to_dig_port(intel_dp)->base.port));
+	I915_STATE_WARN(!intel_dp->want_panel_vdd, "[ENCODER:%d:%s] VDD not forced on",
+			dp_to_dig_port(intel_dp)->base.base.base.id,
+			dp_to_dig_port(intel_dp)->base.base.name);
 
 	intel_dp->want_panel_vdd = false;
 
@@ -2622,12 +2803,14 @@ static void edp_panel_on(struct intel_dp *intel_dp)
 	if (!intel_dp_is_edp(intel_dp))
 		return;
 
-	DRM_DEBUG_KMS("Turn eDP port %c panel power on\n",
-		      port_name(dp_to_dig_port(intel_dp)->base.port));
+	DRM_DEBUG_KMS("Turn [ENCODER:%d:%s] panel power on\n",
+		      dp_to_dig_port(intel_dp)->base.base.base.id,
+		      dp_to_dig_port(intel_dp)->base.base.name);
 
 	if (WARN(edp_have_panel_power(intel_dp),
-		 "eDP port %c panel power already on\n",
-		 port_name(dp_to_dig_port(intel_dp)->base.port)))
+		 "[ENCODER:%d:%s] panel power already on\n",
+		 dp_to_dig_port(intel_dp)->base.base.base.id,
+		 dp_to_dig_port(intel_dp)->base.base.name))
 		return;
 
 	wait_panel_power_cycle(intel_dp);
@@ -2682,11 +2865,11 @@ static void edp_panel_off(struct intel_dp *intel_dp)
 	if (!intel_dp_is_edp(intel_dp))
 		return;
 
-	DRM_DEBUG_KMS("Turn eDP port %c panel power off\n",
-		      port_name(dig_port->base.port));
+	DRM_DEBUG_KMS("Turn [ENCODER:%d:%s] panel power off\n",
+		      dig_port->base.base.base.id, dig_port->base.base.name);
 
-	WARN(!intel_dp->want_panel_vdd, "Need eDP port %c VDD to turn off panel\n",
-	     port_name(dig_port->base.port));
+	WARN(!intel_dp->want_panel_vdd, "Need [ENCODER:%d:%s] VDD to turn off panel\n",
+	     dig_port->base.base.base.id, dig_port->base.base.name);
 
 	pp = ironlake_get_pp_control(intel_dp);
 	/* We need to switch off panel power _and_ force vdd, for otherwise some
@@ -2831,8 +3014,8 @@ static void assert_dp_port(struct intel_dp *intel_dp, bool state)
 	bool cur_state = I915_READ(intel_dp->output_reg) & DP_PORT_EN;
 
 	I915_STATE_WARN(cur_state != state,
-			"DP port %c state assertion failure (expected %s, current %s)\n",
-			port_name(dig_port->base.port),
+			"[ENCODER:%d:%s] state assertion failure (expected %s, current %s)\n",
+			dig_port->base.base.base.id, dig_port->base.base.name,
 			onoff(state), onoff(cur_state));
 }
 #define assert_dp_port_disabled(d) assert_dp_port((d), false)
@@ -2928,7 +3111,7 @@ void intel_dp_sink_set_decompression_state(struct intel_dp *intel_dp,
 {
 	int ret;
 
-	if (!crtc_state->dsc_params.compression_enable)
+	if (!crtc_state->dsc.compression_enable)
 		return;
 
 	ret = drm_dp_dpcd_writeb(&intel_dp->aux, DP_DSC_ENABLE,
@@ -3220,7 +3403,7 @@ _intel_dp_set_link_train(struct intel_dp *intel_dp,
 			      dp_train_pat & train_pat_mask);
 
 	if (HAS_DDI(dev_priv)) {
-		u32 temp = I915_READ(DP_TP_CTL(port));
+		u32 temp = I915_READ(intel_dp->regs.dp_tp_ctl);
 
 		if (dp_train_pat & DP_LINK_SCRAMBLING_DISABLE)
 			temp |= DP_TP_CTL_SCRAMBLE_DISABLE;
@@ -3246,7 +3429,7 @@ _intel_dp_set_link_train(struct intel_dp *intel_dp,
 			temp |= DP_TP_CTL_LINK_TRAIN_PAT4;
 			break;
 		}
-		I915_WRITE(DP_TP_CTL(port), temp);
+		I915_WRITE(intel_dp->regs.dp_tp_ctl, temp);
 
 	} else if ((IS_IVYBRIDGE(dev_priv) && port == PORT_A) ||
 		   (HAS_PCH_CPT(dev_priv) && port != PORT_A)) {
@@ -3410,8 +3593,9 @@ static void vlv_detach_power_sequencer(struct intel_dp *intel_dp)
 	 * port select always when logically disconnecting a power sequencer
 	 * from a port.
 	 */
-	DRM_DEBUG_KMS("detaching pipe %c power sequencer from port %c\n",
-		      pipe_name(pipe), port_name(intel_dig_port->base.port));
+	DRM_DEBUG_KMS("detaching pipe %c power sequencer from [ENCODER:%d:%s]\n",
+		      pipe_name(pipe), intel_dig_port->base.base.base.id,
+		      intel_dig_port->base.base.name);
 	I915_WRITE(pp_on_reg, 0);
 	POSTING_READ(pp_on_reg);
 
@@ -3427,17 +3611,18 @@ static void vlv_steal_power_sequencer(struct drm_i915_private *dev_priv,
 
 	for_each_intel_dp(&dev_priv->drm, encoder) {
 		struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
-		enum port port = encoder->port;
 
 		WARN(intel_dp->active_pipe == pipe,
-		     "stealing pipe %c power sequencer from active (e)DP port %c\n",
-		     pipe_name(pipe), port_name(port));
+		     "stealing pipe %c power sequencer from active [ENCODER:%d:%s]\n",
+		     pipe_name(pipe), encoder->base.base.id,
+		     encoder->base.name);
 
 		if (intel_dp->pps_pipe != pipe)
 			continue;
 
-		DRM_DEBUG_KMS("stealing pipe %c power sequencer from port %c\n",
-			      pipe_name(pipe), port_name(port));
+		DRM_DEBUG_KMS("stealing pipe %c power sequencer from [ENCODER:%d:%s]\n",
+			      pipe_name(pipe), encoder->base.base.id,
+			      encoder->base.name);
 
 		/* make sure vdd is off before we steal it */
 		vlv_detach_power_sequencer(intel_dp);
@@ -3479,8 +3664,9 @@ static void vlv_init_panel_power_sequencer(struct intel_encoder *encoder,
 	/* now it's all ours */
 	intel_dp->pps_pipe = crtc->pipe;
 
-	DRM_DEBUG_KMS("initializing pipe %c power sequencer for port %c\n",
-		      pipe_name(intel_dp->pps_pipe), port_name(encoder->port));
+	DRM_DEBUG_KMS("initializing pipe %c power sequencer for [ENCODER:%d:%s]\n",
+		      pipe_name(intel_dp->pps_pipe), encoder->base.base.id,
+		      encoder->base.name);
 
 	/* init power sequencer on this pipe and port */
 	intel_dp_init_panel_power_sequencer(intel_dp);
@@ -3944,22 +4130,22 @@ void intel_dp_set_idle_link_train(struct intel_dp *intel_dp)
 	if (!HAS_DDI(dev_priv))
 		return;
 
-	val = I915_READ(DP_TP_CTL(port));
+	val = I915_READ(intel_dp->regs.dp_tp_ctl);
 	val &= ~DP_TP_CTL_LINK_TRAIN_MASK;
 	val |= DP_TP_CTL_LINK_TRAIN_IDLE;
-	I915_WRITE(DP_TP_CTL(port), val);
+	I915_WRITE(intel_dp->regs.dp_tp_ctl, val);
 
 	/*
-	 * On PORT_A we can have only eDP in SST mode. There the only reason
-	 * we need to set idle transmission mode is to work around a HW issue
-	 * where we enable the pipe while not in idle link-training mode.
+	 * Until TGL on PORT_A we can have only eDP in SST mode. There the only
+	 * reason we need to set idle transmission mode is to work around a HW
+	 * issue where we enable the pipe while not in idle link-training mode.
 	 * In this case there is requirement to wait for a minimum number of
 	 * idle patterns to be sent.
 	 */
-	if (port == PORT_A)
+	if (port == PORT_A && INTEL_GEN(dev_priv) < 12)
 		return;
 
-	if (intel_de_wait_for_set(dev_priv, DP_TP_STATUS(port),
+	if (intel_de_wait_for_set(dev_priv, intel_dp->regs.dp_tp_status,
 				  DP_TP_STATUS_IDLE_DONE, 1))
 		DRM_ERROR("Timed out waiting for DP idle patterns\n");
 }
@@ -4301,9 +4487,10 @@ intel_dp_configure_mst(struct intel_dp *intel_dp)
 		&dp_to_dig_port(intel_dp)->base;
 	bool sink_can_mst = intel_dp_sink_can_mst(intel_dp);
 
-	DRM_DEBUG_KMS("MST support? port %c: %s, sink: %s, modparam: %s\n",
-		      port_name(encoder->port), yesno(intel_dp->can_mst),
-		      yesno(sink_can_mst), yesno(i915_modparams.enable_dp_mst));
+	DRM_DEBUG_KMS("[ENCODER:%d:%s] MST support: port: %s, sink: %s, modparam: %s\n",
+		      encoder->base.base.id, encoder->base.name,
+		      yesno(intel_dp->can_mst), yesno(sink_can_mst),
+		      yesno(i915_modparams.enable_dp_mst));
 
 	if (!intel_dp->can_mst)
 		return;
@@ -4323,94 +4510,36 @@ intel_dp_get_sink_irq_esi(struct intel_dp *intel_dp, u8 *sink_irq_vector)
 		DP_DPRX_ESI_LEN;
 }
 
-u16 intel_dp_dsc_get_output_bpp(int link_clock, u8 lane_count,
-				int mode_clock, int mode_hdisplay)
+bool
+intel_dp_needs_vsc_sdp(const struct intel_crtc_state *crtc_state,
+		       const struct drm_connector_state *conn_state)
 {
-	u16 bits_per_pixel, max_bpp_small_joiner_ram;
-	int i;
-
 	/*
-	 * Available Link Bandwidth(Kbits/sec) = (NumberOfLanes)*
-	 * (LinkSymbolClock)* 8 * ((100-FECOverhead)/100)*(TimeSlotsPerMTP)
-	 * FECOverhead = 2.4%, for SST -> TimeSlotsPerMTP is 1,
-	 * for MST -> TimeSlotsPerMTP has to be calculated
+	 * As per DP 1.4a spec section 2.2.4.3 [MSA Field for Indication
+	 * of Color Encoding Format and Content Color Gamut], in order to
+	 * sending YCBCR 420 or HDR BT.2020 signals we should use DP VSC SDP.
 	 */
-	bits_per_pixel = (link_clock * lane_count * 8 *
-			  DP_DSC_FEC_OVERHEAD_FACTOR) /
-		mode_clock;
-
-	/* Small Joiner Check: output bpp <= joiner RAM (bits) / Horiz. width */
-	max_bpp_small_joiner_ram = DP_DSC_MAX_SMALL_JOINER_RAM_BUFFER /
-		mode_hdisplay;
-
-	/*
-	 * Greatest allowed DSC BPP = MIN (output BPP from avaialble Link BW
-	 * check, output bpp from small joiner RAM check)
-	 */
-	bits_per_pixel = min(bits_per_pixel, max_bpp_small_joiner_ram);
-
-	/* Error out if the max bpp is less than smallest allowed valid bpp */
-	if (bits_per_pixel < valid_dsc_bpp[0]) {
-		DRM_DEBUG_KMS("Unsupported BPP %d\n", bits_per_pixel);
-		return 0;
-	}
-
-	/* Find the nearest match in the array of known BPPs from VESA */
-	for (i = 0; i < ARRAY_SIZE(valid_dsc_bpp) - 1; i++) {
-		if (bits_per_pixel < valid_dsc_bpp[i + 1])
-			break;
-	}
-	bits_per_pixel = valid_dsc_bpp[i];
-
-	/*
-	 * Compressed BPP in U6.4 format so multiply by 16, for Gen 11,
-	 * fractional part is 0
-	 */
-	return bits_per_pixel << 4;
-}
-
-u8 intel_dp_dsc_get_slice_count(struct intel_dp *intel_dp,
-				int mode_clock,
-				int mode_hdisplay)
-{
-	u8 min_slice_count, i;
-	int max_slice_width;
-
-	if (mode_clock <= DP_DSC_PEAK_PIXEL_RATE)
-		min_slice_count = DIV_ROUND_UP(mode_clock,
-					       DP_DSC_MAX_ENC_THROUGHPUT_0);
-	else
-		min_slice_count = DIV_ROUND_UP(mode_clock,
-					       DP_DSC_MAX_ENC_THROUGHPUT_1);
-
-	max_slice_width = drm_dp_dsc_sink_max_slice_width(intel_dp->dsc_dpcd);
-	if (max_slice_width < DP_DSC_MIN_SLICE_WIDTH_VALUE) {
-		DRM_DEBUG_KMS("Unsupported slice width %d by DP DSC Sink device\n",
-			      max_slice_width);
-		return 0;
-	}
-	/* Also take into account max slice width */
-	min_slice_count = min_t(u8, min_slice_count,
-				DIV_ROUND_UP(mode_hdisplay,
-					     max_slice_width));
+	if (crtc_state->output_format == INTEL_OUTPUT_FORMAT_YCBCR420)
+		return true;
 
-	/* Find the closest match to the valid slice count values */
-	for (i = 0; i < ARRAY_SIZE(valid_dsc_slicecount); i++) {
-		if (valid_dsc_slicecount[i] >
-		    drm_dp_dsc_sink_max_slice_count(intel_dp->dsc_dpcd,
-						    false))
-			break;
-		if (min_slice_count  <= valid_dsc_slicecount[i])
-			return valid_dsc_slicecount[i];
+	switch (conn_state->colorspace) {
+	case DRM_MODE_COLORIMETRY_SYCC_601:
+	case DRM_MODE_COLORIMETRY_OPYCC_601:
+	case DRM_MODE_COLORIMETRY_BT2020_YCC:
+	case DRM_MODE_COLORIMETRY_BT2020_RGB:
+	case DRM_MODE_COLORIMETRY_BT2020_CYCC:
+		return true;
+	default:
+		break;
 	}
 
-	DRM_DEBUG_KMS("Unsupported Slice Count %d\n", min_slice_count);
-	return 0;
+	return false;
 }
 
 static void
-intel_pixel_encoding_setup_vsc(struct intel_dp *intel_dp,
-			       const struct intel_crtc_state *crtc_state)
+intel_dp_setup_vsc_sdp(struct intel_dp *intel_dp,
+		       const struct intel_crtc_state *crtc_state,
+		       const struct drm_connector_state *conn_state)
 {
 	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
 	struct dp_sdp vsc_sdp = {};
@@ -4431,13 +4560,55 @@ intel_pixel_encoding_setup_vsc(struct intel_dp *intel_dp,
 	 */
 	vsc_sdp.sdp_header.HB3 = 0x13;
 
-	/*
-	 * YCbCr 420 = 3h DB16[7:4] ITU-R BT.601 = 0h, ITU-R BT.709 = 1h
-	 * DB16[3:0] DP 1.4a spec, Table 2-120
-	 */
-	vsc_sdp.db[16] = 0x3 << 4; /* 0x3 << 4 , YCbCr 420*/
-	/* RGB->YCBCR color conversion uses the BT.709 color space. */
-	vsc_sdp.db[16] |= 0x1; /* 0x1, ITU-R BT.709 */
+	/* DP 1.4a spec, Table 2-120 */
+	switch (crtc_state->output_format) {
+	case INTEL_OUTPUT_FORMAT_YCBCR444:
+		vsc_sdp.db[16] = 0x1 << 4; /* YCbCr 444 : DB16[7:4] = 1h */
+		break;
+	case INTEL_OUTPUT_FORMAT_YCBCR420:
+		vsc_sdp.db[16] = 0x3 << 4; /* YCbCr 420 : DB16[7:4] = 3h */
+		break;
+	case INTEL_OUTPUT_FORMAT_RGB:
+	default:
+		/* RGB: DB16[7:4] = 0h */
+		break;
+	}
+
+	switch (conn_state->colorspace) {
+	case DRM_MODE_COLORIMETRY_BT709_YCC:
+		vsc_sdp.db[16] |= 0x1;
+		break;
+	case DRM_MODE_COLORIMETRY_XVYCC_601:
+		vsc_sdp.db[16] |= 0x2;
+		break;
+	case DRM_MODE_COLORIMETRY_XVYCC_709:
+		vsc_sdp.db[16] |= 0x3;
+		break;
+	case DRM_MODE_COLORIMETRY_SYCC_601:
+		vsc_sdp.db[16] |= 0x4;
+		break;
+	case DRM_MODE_COLORIMETRY_OPYCC_601:
+		vsc_sdp.db[16] |= 0x5;
+		break;
+	case DRM_MODE_COLORIMETRY_BT2020_CYCC:
+	case DRM_MODE_COLORIMETRY_BT2020_RGB:
+		vsc_sdp.db[16] |= 0x6;
+		break;
+	case DRM_MODE_COLORIMETRY_BT2020_YCC:
+		vsc_sdp.db[16] |= 0x7;
+		break;
+	case DRM_MODE_COLORIMETRY_DCI_P3_RGB_D65:
+	case DRM_MODE_COLORIMETRY_DCI_P3_RGB_THEATER:
+		vsc_sdp.db[16] |= 0x4; /* DCI-P3 (SMPTE RP 431-2) */
+		break;
+	default:
+		/* sRGB (IEC 61966-2-1) / ITU-R BT.601: DB16[0:3] = 0h */
+
+		/* RGB->YCBCR color conversion uses the BT.709 color space. */
+		if (crtc_state->output_format == INTEL_OUTPUT_FORMAT_YCBCR420)
+			vsc_sdp.db[16] |= 0x1; /* 0x1, ITU-R BT.709 */
+		break;
+	}
 
 	/*
 	 * For pixel encoding formats YCbCr444, YCbCr422, YCbCr420, and Y Only,
@@ -4489,13 +4660,106 @@ intel_pixel_encoding_setup_vsc(struct intel_dp *intel_dp,
 			crtc_state, DP_SDP_VSC, &vsc_sdp, sizeof(vsc_sdp));
 }
 
-void intel_dp_ycbcr_420_enable(struct intel_dp *intel_dp,
-			       const struct intel_crtc_state *crtc_state)
+static void
+intel_dp_setup_hdr_metadata_infoframe_sdp(struct intel_dp *intel_dp,
+					  const struct intel_crtc_state *crtc_state,
+					  const struct drm_connector_state *conn_state)
+{
+	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
+	struct dp_sdp infoframe_sdp = {};
+	struct hdmi_drm_infoframe drm_infoframe = {};
+	const int infoframe_size = HDMI_INFOFRAME_HEADER_SIZE + HDMI_DRM_INFOFRAME_SIZE;
+	unsigned char buf[HDMI_INFOFRAME_HEADER_SIZE + HDMI_DRM_INFOFRAME_SIZE];
+	ssize_t len;
+	int ret;
+
+	ret = drm_hdmi_infoframe_set_hdr_metadata(&drm_infoframe, conn_state);
+	if (ret) {
+		DRM_DEBUG_KMS("couldn't set HDR metadata in infoframe\n");
+		return;
+	}
+
+	len = hdmi_drm_infoframe_pack_only(&drm_infoframe, buf, sizeof(buf));
+	if (len < 0) {
+		DRM_DEBUG_KMS("buffer size is smaller than hdr metadata infoframe\n");
+		return;
+	}
+
+	if (len != infoframe_size) {
+		DRM_DEBUG_KMS("wrong static hdr metadata size\n");
+		return;
+	}
+
+	/*
+	 * Set up the infoframe sdp packet for HDR static metadata.
+	 * Prepare VSC Header for SU as per DP 1.4a spec,
+	 * Table 2-100 and Table 2-101
+	 */
+
+	/* Packet ID, 00h for non-Audio INFOFRAME */
+	infoframe_sdp.sdp_header.HB0 = 0;
+	/*
+	 * Packet Type 80h + Non-audio INFOFRAME Type value
+	 * HDMI_INFOFRAME_TYPE_DRM: 0x87,
+	 */
+	infoframe_sdp.sdp_header.HB1 = drm_infoframe.type;
+	/*
+	 * Least Significant Eight Bits of (Data Byte Count – 1)
+	 * infoframe_size - 1,
+	 */
+	infoframe_sdp.sdp_header.HB2 = 0x1D;
+	/* INFOFRAME SDP Version Number */
+	infoframe_sdp.sdp_header.HB3 = (0x13 << 2);
+	/* CTA Header Byte 2 (INFOFRAME Version Number) */
+	infoframe_sdp.db[0] = drm_infoframe.version;
+	/* CTA Header Byte 3 (Length of INFOFRAME): HDMI_DRM_INFOFRAME_SIZE */
+	infoframe_sdp.db[1] = drm_infoframe.length;
+	/*
+	 * Copy HDMI_DRM_INFOFRAME_SIZE size from a buffer after
+	 * HDMI_INFOFRAME_HEADER_SIZE
+	 */
+	BUILD_BUG_ON(sizeof(infoframe_sdp.db) < HDMI_DRM_INFOFRAME_SIZE + 2);
+	memcpy(&infoframe_sdp.db[2], &buf[HDMI_INFOFRAME_HEADER_SIZE],
+	       HDMI_DRM_INFOFRAME_SIZE);
+
+	/*
+	 * Size of DP infoframe sdp packet for HDR static metadata is consist of
+	 * - DP SDP Header(struct dp_sdp_header): 4 bytes
+	 * - Two Data Blocks: 2 bytes
+	 *    CTA Header Byte2 (INFOFRAME Version Number)
+	 *    CTA Header Byte3 (Length of INFOFRAME)
+	 * - HDMI_DRM_INFOFRAME_SIZE: 26 bytes
+	 *
+	 * Prior to GEN11's GMP register size is identical to DP HDR static metadata
+	 * infoframe size. But GEN11+ has larger than that size, write_infoframe
+	 * will pad rest of the size.
+	 */
+	intel_dig_port->write_infoframe(&intel_dig_port->base, crtc_state,
+					HDMI_PACKET_TYPE_GAMUT_METADATA,
+					&infoframe_sdp,
+					sizeof(struct dp_sdp_header) + 2 + HDMI_DRM_INFOFRAME_SIZE);
+}
+
+void intel_dp_vsc_enable(struct intel_dp *intel_dp,
+			 const struct intel_crtc_state *crtc_state,
+			 const struct drm_connector_state *conn_state)
 {
-	if (crtc_state->output_format != INTEL_OUTPUT_FORMAT_YCBCR420)
+	if (!intel_dp_needs_vsc_sdp(crtc_state, conn_state))
 		return;
 
-	intel_pixel_encoding_setup_vsc(intel_dp, crtc_state);
+	intel_dp_setup_vsc_sdp(intel_dp, crtc_state, conn_state);
+}
+
+void intel_dp_hdr_metadata_enable(struct intel_dp *intel_dp,
+				  const struct intel_crtc_state *crtc_state,
+				  const struct drm_connector_state *conn_state)
+{
+	if (!conn_state->hdr_output_metadata)
+		return;
+
+	intel_dp_setup_hdr_metadata_infoframe_sdp(intel_dp,
+						  crtc_state,
+						  conn_state);
 }
 
 static u8 intel_dp_autotest_link_training(struct intel_dp *intel_dp)
@@ -5217,6 +5481,9 @@ static bool icl_combo_port_connected(struct drm_i915_private *dev_priv,
 {
 	enum port port = intel_dig_port->base.port;
 
+	if (HAS_PCH_MCC(dev_priv) && port == PORT_C)
+		return I915_READ(SDEISR) & SDE_TC_HOTPLUG_ICP(PORT_TC1);
+
 	return I915_READ(SDEISR) & SDE_DDI_HOTPLUG_ICP(port);
 }
 
@@ -5430,6 +5697,12 @@ out:
 	if (status != connector_status_connected && !intel_dp->is_mst)
 		intel_dp_unset_edid(intel_dp);
 
+	/*
+	 * Make sure the refs for power wells enabled during detect are
+	 * dropped to avoid a new detect cycle triggered by HPD polling.
+	 */
+	intel_display_power_flush_work(dev_priv);
+
 	return status;
 }
 
@@ -5490,7 +5763,6 @@ static int
 intel_dp_connector_register(struct drm_connector *connector)
 {
 	struct intel_dp *intel_dp = intel_attached_dp(connector);
-	struct drm_device *dev = connector->dev;
 	int ret;
 
 	ret = intel_connector_register(connector);
@@ -5505,8 +5777,7 @@ intel_dp_connector_register(struct drm_connector *connector)
 	intel_dp->aux.dev = connector->kdev;
 	ret = drm_dp_aux_register(&intel_dp->aux);
 	if (!ret)
-		drm_dp_cec_register_connector(&intel_dp->aux,
-					      connector->name, dev->dev);
+		drm_dp_cec_register_connector(&intel_dp->aux, connector);
 	return ret;
 }
 
@@ -6264,13 +6535,15 @@ intel_dp_hpd_pulse(struct intel_digital_port *intel_dig_port, bool long_hpd)
 		 * would end up in an endless cycle of
 		 * "vdd off -> long hpd -> vdd on -> detect -> vdd off -> ..."
 		 */
-		DRM_DEBUG_KMS("ignoring long hpd on eDP port %c\n",
-			      port_name(intel_dig_port->base.port));
+		DRM_DEBUG_KMS("ignoring long hpd on eDP [ENCODER:%d:%s]\n",
+			      intel_dig_port->base.base.base.id,
+			      intel_dig_port->base.base.name);
 		return IRQ_HANDLED;
 	}
 
-	DRM_DEBUG_KMS("got hpd irq on port %c - %s\n",
-		      port_name(intel_dig_port->base.port),
+	DRM_DEBUG_KMS("got hpd irq on [ENCODER:%d:%s] - %s\n",
+		      intel_dig_port->base.base.base.id,
+		      intel_dig_port->base.base.name,
 		      long_hpd ? "long" : "short");
 
 	if (long_hpd) {
@@ -6337,6 +6610,13 @@ intel_dp_add_properties(struct intel_dp *intel_dp, struct drm_connector *connect
 	else if (INTEL_GEN(dev_priv) >= 5)
 		drm_connector_attach_max_bpc_property(connector, 6, 12);
 
+	intel_attach_colorspace_property(connector);
+
+	if (IS_GEMINILAKE(dev_priv) || INTEL_GEN(dev_priv) >= 11)
+		drm_object_attach_property(&connector->base,
+					   connector->dev->mode_config.hdr_output_metadata_property,
+					   0);
+
 	if (intel_dp_is_edp(intel_dp)) {
 		u32 allowed_scalers;
 
@@ -7134,8 +7414,9 @@ intel_dp_init_connector(struct intel_digital_port *intel_dig_port,
 		  intel_dp_modeset_retry_work_fn);
 
 	if (WARN(intel_dig_port->max_lanes < 1,
-		 "Not enough lanes (%d) for DP on port %c\n",
-		 intel_dig_port->max_lanes, port_name(port)))
+		 "Not enough lanes (%d) for DP on [ENCODER:%d:%s]\n",
+		 intel_dig_port->max_lanes, intel_encoder->base.base.id,
+		 intel_encoder->base.name))
 		return false;
 
 	intel_dp_set_source_rates(intel_dp);
@@ -7176,9 +7457,9 @@ intel_dp_init_connector(struct intel_digital_port *intel_dig_port,
 		    port != PORT_B && port != PORT_C))
 		return false;
 
-	DRM_DEBUG_KMS("Adding %s connector on port %c\n",
-			type == DRM_MODE_CONNECTOR_eDP ? "eDP" : "DP",
-			port_name(port));
+	DRM_DEBUG_KMS("Adding %s connector on [ENCODER:%d:%s]\n",
+		      type == DRM_MODE_CONNECTOR_eDP ? "eDP" : "DP",
+		      intel_encoder->base.base.id, intel_encoder->base.name);
 
 	drm_connector_init(dev, connector, &intel_dp_connector_funcs, type);
 	drm_connector_helper_add(connector, &intel_dp_connector_helper_funcs);
@@ -7202,11 +7483,8 @@ intel_dp_init_connector(struct intel_digital_port *intel_dig_port,
 		intel_connector->get_hw_state = intel_connector_get_hw_state;
 
 	/* init MST on ports that can support it */
-	if (HAS_DP_MST(dev_priv) && !intel_dp_is_edp(intel_dp) &&
-	    (port == PORT_B || port == PORT_C ||
-	     port == PORT_D || port == PORT_F))
-		intel_dp_mst_encoder_init(intel_dig_port,
-					  intel_connector->base.base.id);
+	intel_dp_mst_encoder_init(intel_dig_port,
+				  intel_connector->base.base.id);
 
 	if (!intel_edp_init_connector(intel_dp, intel_connector)) {
 		intel_dp_aux_fini(intel_dp);
@@ -7297,11 +7575,11 @@ bool intel_dp_init(struct drm_i915_private *dev_priv,
 	intel_encoder->power_domain = intel_port_to_power_domain(port);
 	if (IS_CHERRYVIEW(dev_priv)) {
 		if (port == PORT_D)
-			intel_encoder->crtc_mask = 1 << 2;
+			intel_encoder->pipe_mask = BIT(PIPE_C);
 		else
-			intel_encoder->crtc_mask = (1 << 0) | (1 << 1);
+			intel_encoder->pipe_mask = BIT(PIPE_A) | BIT(PIPE_B);
 	} else {
-		intel_encoder->crtc_mask = (1 << 0) | (1 << 1) | (1 << 2);
+		intel_encoder->pipe_mask = ~0;
 	}
 	intel_encoder->cloneable = 0;
 	intel_encoder->port = port;
@@ -7362,7 +7640,8 @@ void intel_dp_mst_resume(struct drm_i915_private *dev_priv)
 		if (!intel_dp->can_mst)
 			continue;
 
-		ret = drm_dp_mst_topology_mgr_resume(&intel_dp->mst_mgr);
+		ret = drm_dp_mst_topology_mgr_resume(&intel_dp->mst_mgr,
+						     true);
 		if (ret) {
 			intel_dp->is_mst = false;
 			drm_dp_mst_topology_mgr_set_mst(&intel_dp->mst_mgr,
diff --git a/drivers/gpu/drm/i915/display/intel_dp.h b/drivers/gpu/drm/i915/display/intel_dp.h
index 657bbb1f5ed0..3da166054788 100644
--- a/drivers/gpu/drm/i915/display/intel_dp.h
+++ b/drivers/gpu/drm/i915/display/intel_dp.h
@@ -13,6 +13,7 @@
 #include "i915_reg.h"
 
 enum pipe;
+enum port;
 struct drm_connector_state;
 struct drm_encoder;
 struct drm_i915_private;
@@ -102,15 +103,19 @@ bool intel_dp_source_supports_hbr2(struct intel_dp *intel_dp);
 bool intel_dp_source_supports_hbr3(struct intel_dp *intel_dp);
 bool
 intel_dp_get_link_status(struct intel_dp *intel_dp, u8 *link_status);
-u16 intel_dp_dsc_get_output_bpp(int link_clock, u8 lane_count,
-				int mode_clock, int mode_hdisplay);
-u8 intel_dp_dsc_get_slice_count(struct intel_dp *intel_dp, int mode_clock,
-				int mode_hdisplay);
 
 bool intel_dp_read_dpcd(struct intel_dp *intel_dp);
 bool intel_dp_get_colorimetry_status(struct intel_dp *intel_dp);
 int intel_dp_link_required(int pixel_clock, int bpp);
 int intel_dp_max_data_rate(int max_link_clock, int max_lanes);
+bool intel_dp_needs_vsc_sdp(const struct intel_crtc_state *crtc_state,
+			    const struct drm_connector_state *conn_state);
+void intel_dp_vsc_enable(struct intel_dp *intel_dp,
+			 const struct intel_crtc_state *crtc_state,
+			 const struct drm_connector_state *conn_state);
+void intel_dp_hdr_metadata_enable(struct intel_dp *intel_dp,
+				  const struct intel_crtc_state *crtc_state,
+				  const struct drm_connector_state *conn_state);
 bool intel_digital_port_connected(struct intel_encoder *encoder);
 
 static inline unsigned int intel_dp_unused_lane_mask(int lane_count)
@@ -118,4 +123,6 @@ static inline unsigned int intel_dp_unused_lane_mask(int lane_count)
 	return ~((1 << lane_count) - 1) & 0xf;
 }
 
+u32 intel_dp_mode_to_fec_clock(u32 mode_clock);
+
 #endif /* __INTEL_DP_H__ */
diff --git a/drivers/gpu/drm/i915/display/intel_dp_mst.c b/drivers/gpu/drm/i915/display/intel_dp_mst.c
index 6df240a01b8c..03d1cba0b696 100644
--- a/drivers/gpu/drm/i915/display/intel_dp_mst.c
+++ b/drivers/gpu/drm/i915/display/intel_dp_mst.c
@@ -81,7 +81,7 @@ static int intel_dp_mst_compute_link_config(struct intel_encoder *encoder,
 			       adjusted_mode->crtc_clock,
 			       crtc_state->port_clock,
 			       &crtc_state->dp_m_n,
-			       constant_n);
+			       constant_n, crtc_state->fec_enable);
 	crtc_state->dp_m_n.tu = slots;
 
 	return 0;
@@ -215,7 +215,7 @@ static void intel_mst_disable_dp(struct intel_encoder *encoder,
 
 	ret = drm_dp_update_payload_part1(&intel_dp->mst_mgr);
 	if (ret) {
-		DRM_ERROR("failed to update payload %d\n", ret);
+		DRM_DEBUG_KMS("failed to update payload %d\n", ret);
 	}
 	if (old_crtc_state->has_audio)
 		intel_audio_codec_disable(encoder,
@@ -295,7 +295,6 @@ static void intel_mst_pre_enable_dp(struct intel_encoder *encoder,
 	struct intel_digital_port *intel_dig_port = intel_mst->primary;
 	struct intel_dp *intel_dp = &intel_dig_port->dp;
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	enum port port = intel_dig_port->base.port;
 	struct intel_connector *connector =
 		to_intel_connector(conn_state->connector);
 	int ret;
@@ -326,12 +325,14 @@ static void intel_mst_pre_enable_dp(struct intel_encoder *encoder,
 		DRM_ERROR("failed to allocate vcpi\n");
 
 	intel_dp->active_mst_links++;
-	temp = I915_READ(DP_TP_STATUS(port));
-	I915_WRITE(DP_TP_STATUS(port), temp);
+	temp = I915_READ(intel_dp->regs.dp_tp_status);
+	I915_WRITE(intel_dp->regs.dp_tp_status, temp);
 
 	ret = drm_dp_update_payload_part1(&intel_dp->mst_mgr);
 
 	intel_ddi_enable_pipe_clock(pipe_config);
+
+	intel_ddi_set_dp_msa(pipe_config, conn_state);
 }
 
 static void intel_mst_enable_dp(struct intel_encoder *encoder,
@@ -342,11 +343,10 @@ static void intel_mst_enable_dp(struct intel_encoder *encoder,
 	struct intel_digital_port *intel_dig_port = intel_mst->primary;
 	struct intel_dp *intel_dp = &intel_dig_port->dp;
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	enum port port = intel_dig_port->base.port;
 
 	DRM_DEBUG_KMS("active links %d\n", intel_dp->active_mst_links);
 
-	if (intel_de_wait_for_set(dev_priv, DP_TP_STATUS(port),
+	if (intel_de_wait_for_set(dev_priv, intel_dp->regs.dp_tp_status,
 				  DP_TP_STATUS_ACT_SENT, 1))
 		DRM_ERROR("Timed out waiting for ACT sent\n");
 
@@ -393,20 +393,7 @@ static int intel_dp_mst_get_ddc_modes(struct drm_connector *connector)
 	return ret;
 }
 
-static enum drm_connector_status
-intel_dp_mst_detect(struct drm_connector *connector, bool force)
-{
-	struct intel_connector *intel_connector = to_intel_connector(connector);
-	struct intel_dp *intel_dp = intel_connector->mst_port;
-
-	if (drm_connector_is_unregistered(connector))
-		return connector_status_disconnected;
-	return drm_dp_mst_detect_port(connector, &intel_dp->mst_mgr,
-				      intel_connector->port);
-}
-
 static const struct drm_connector_funcs intel_dp_mst_connector_funcs = {
-	.detect = intel_dp_mst_detect,
 	.fill_modes = drm_helper_probe_single_connector_modes,
 	.atomic_get_property = intel_digital_connector_atomic_get_property,
 	.atomic_set_property = intel_digital_connector_atomic_set_property,
@@ -426,6 +413,7 @@ static enum drm_mode_status
 intel_dp_mst_mode_valid(struct drm_connector *connector,
 			struct drm_display_mode *mode)
 {
+	struct drm_i915_private *dev_priv = to_i915(connector->dev);
 	struct intel_connector *intel_connector = to_intel_connector(connector);
 	struct intel_dp *intel_dp = intel_connector->mst_port;
 	int max_dotclk = to_i915(connector->dev)->max_dotclk_freq;
@@ -453,7 +441,7 @@ intel_dp_mst_mode_valid(struct drm_connector *connector,
 	if (mode_rate > max_rate || mode->clock > max_dotclk)
 		return MODE_CLOCK_HIGH;
 
-	return MODE_OK;
+	return intel_mode_valid_max_plane_size(dev_priv, mode);
 }
 
 static struct drm_encoder *intel_mst_atomic_best_encoder(struct drm_connector *connector,
@@ -466,11 +454,26 @@ static struct drm_encoder *intel_mst_atomic_best_encoder(struct drm_connector *c
 	return &intel_dp->mst_encoders[crtc->pipe]->base.base;
 }
 
+static int
+intel_dp_mst_detect(struct drm_connector *connector,
+		    struct drm_modeset_acquire_ctx *ctx, bool force)
+{
+	struct intel_connector *intel_connector = to_intel_connector(connector);
+	struct intel_dp *intel_dp = intel_connector->mst_port;
+
+	if (drm_connector_is_unregistered(connector))
+		return connector_status_disconnected;
+
+	return drm_dp_mst_detect_port(connector, ctx, &intel_dp->mst_mgr,
+				      intel_connector->port);
+}
+
 static const struct drm_connector_helper_funcs intel_dp_mst_connector_helper_funcs = {
 	.get_modes = intel_dp_mst_get_modes,
 	.mode_valid = intel_dp_mst_mode_valid,
 	.atomic_best_encoder = intel_mst_atomic_best_encoder,
 	.atomic_check = intel_dp_mst_atomic_check,
+	.detect_ctx = intel_dp_mst_detect,
 };
 
 static void intel_dp_mst_encoder_destroy(struct drm_encoder *encoder)
@@ -615,8 +618,16 @@ intel_dp_create_fake_mst_encoder(struct intel_digital_port *intel_dig_port, enum
 	intel_encoder->type = INTEL_OUTPUT_DP_MST;
 	intel_encoder->power_domain = intel_dig_port->base.power_domain;
 	intel_encoder->port = intel_dig_port->base.port;
-	intel_encoder->crtc_mask = BIT(pipe);
 	intel_encoder->cloneable = 0;
+	/*
+	 * This is wrong, but broken userspace uses the intersection
+	 * of possible_crtcs of all the encoders of a given connector
+	 * to figure out which crtcs can drive said connector. What
+	 * should be used instead is the union of possible_crtcs.
+	 * To keep such userspace functioning we must misconfigure
+	 * this to make sure the intersection is not empty :(
+	 */
+	intel_encoder->pipe_mask = ~0;
 
 	intel_encoder->compute_config = intel_dp_mst_compute_config;
 	intel_encoder->disable = intel_mst_disable_dp;
@@ -653,21 +664,31 @@ intel_dp_mst_encoder_active_links(struct intel_digital_port *intel_dig_port)
 int
 intel_dp_mst_encoder_init(struct intel_digital_port *intel_dig_port, int conn_base_id)
 {
+	struct drm_i915_private *i915 = to_i915(intel_dig_port->base.base.dev);
 	struct intel_dp *intel_dp = &intel_dig_port->dp;
-	struct drm_device *dev = intel_dig_port->base.base.dev;
+	enum port port = intel_dig_port->base.port;
 	int ret;
 
-	intel_dp->can_mst = true;
+	if (!HAS_DP_MST(i915) || intel_dp_is_edp(intel_dp))
+		return 0;
+
+	if (INTEL_GEN(i915) < 12 && port == PORT_A)
+		return 0;
+
+	if (INTEL_GEN(i915) < 11 && port == PORT_E)
+		return 0;
+
 	intel_dp->mst_mgr.cbs = &mst_cbs;
 
 	/* create encoders */
 	intel_dp_create_fake_mst_encoders(intel_dig_port);
-	ret = drm_dp_mst_topology_mgr_init(&intel_dp->mst_mgr, dev,
+	ret = drm_dp_mst_topology_mgr_init(&intel_dp->mst_mgr, &i915->drm,
 					   &intel_dp->aux, 16, 3, conn_base_id);
-	if (ret) {
-		intel_dp->can_mst = false;
+	if (ret)
 		return ret;
-	}
+
+	intel_dp->can_mst = true;
+
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/i915/display/intel_dpll_mgr.c b/drivers/gpu/drm/i915/display/intel_dpll_mgr.c
index b8148f838354..3ce0a023eee0 100644
--- a/drivers/gpu/drm/i915/display/intel_dpll_mgr.c
+++ b/drivers/gpu/drm/i915/display/intel_dpll_mgr.c
@@ -247,8 +247,7 @@ static struct intel_shared_dpll *
 intel_find_shared_dpll(struct intel_atomic_state *state,
 		       const struct intel_crtc *crtc,
 		       const struct intel_dpll_hw_state *pll_state,
-		       enum intel_dpll_id range_min,
-		       enum intel_dpll_id range_max)
+		       unsigned long dpll_mask)
 {
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	struct intel_shared_dpll *pll, *unused_pll = NULL;
@@ -257,7 +256,9 @@ intel_find_shared_dpll(struct intel_atomic_state *state,
 
 	shared_dpll = intel_atomic_get_shared_dpll_state(&state->base);
 
-	for (i = range_min; i <= range_max; i++) {
+	WARN_ON(dpll_mask & ~(BIT(I915_NUM_PLLS) - 1));
+
+	for_each_set_bit(i, &dpll_mask, I915_NUM_PLLS) {
 		pll = &dev_priv->shared_dplls[i];
 
 		/* Only want to check enabled timings first */
@@ -464,8 +465,8 @@ static bool ibx_get_dpll(struct intel_atomic_state *state,
 	} else {
 		pll = intel_find_shared_dpll(state, crtc,
 					     &crtc_state->dpll_hw_state,
-					     DPLL_ID_PCH_PLL_A,
-					     DPLL_ID_PCH_PLL_B);
+					     BIT(DPLL_ID_PCH_PLL_B) |
+					     BIT(DPLL_ID_PCH_PLL_A));
 	}
 
 	if (!pll)
@@ -525,16 +526,31 @@ static void hsw_ddi_wrpll_disable(struct drm_i915_private *dev_priv,
 	val = I915_READ(WRPLL_CTL(id));
 	I915_WRITE(WRPLL_CTL(id), val & ~WRPLL_PLL_ENABLE);
 	POSTING_READ(WRPLL_CTL(id));
+
+	/*
+	 * Try to set up the PCH reference clock once all DPLLs
+	 * that depend on it have been shut down.
+	 */
+	if (dev_priv->pch_ssc_use & BIT(id))
+		intel_init_pch_refclk(dev_priv);
 }
 
 static void hsw_ddi_spll_disable(struct drm_i915_private *dev_priv,
 				 struct intel_shared_dpll *pll)
 {
+	enum intel_dpll_id id = pll->info->id;
 	u32 val;
 
 	val = I915_READ(SPLL_CTL);
 	I915_WRITE(SPLL_CTL, val & ~SPLL_PLL_ENABLE);
 	POSTING_READ(SPLL_CTL);
+
+	/*
+	 * Try to set up the PCH reference clock once all DPLLs
+	 * that depend on it have been shut down.
+	 */
+	if (dev_priv->pch_ssc_use & BIT(id))
+		intel_init_pch_refclk(dev_priv);
 }
 
 static bool hsw_ddi_wrpll_get_hw_state(struct drm_i915_private *dev_priv,
@@ -814,7 +830,8 @@ hsw_ddi_hdmi_get_dpll(struct intel_atomic_state *state,
 
 	pll = intel_find_shared_dpll(state, crtc,
 				     &crtc_state->dpll_hw_state,
-				     DPLL_ID_WRPLL1, DPLL_ID_WRPLL2);
+				     BIT(DPLL_ID_WRPLL2) |
+				     BIT(DPLL_ID_WRPLL1));
 
 	if (!pll)
 		return NULL;
@@ -877,7 +894,7 @@ static bool hsw_get_dpll(struct intel_atomic_state *state,
 
 		pll = intel_find_shared_dpll(state, crtc,
 					     &crtc_state->dpll_hw_state,
-					     DPLL_ID_SPLL, DPLL_ID_SPLL);
+					     BIT(DPLL_ID_SPLL));
 	} else {
 		return false;
 	}
@@ -1447,13 +1464,13 @@ static bool skl_get_dpll(struct intel_atomic_state *state,
 	if (intel_crtc_has_type(crtc_state, INTEL_OUTPUT_EDP))
 		pll = intel_find_shared_dpll(state, crtc,
 					     &crtc_state->dpll_hw_state,
-					     DPLL_ID_SKL_DPLL0,
-					     DPLL_ID_SKL_DPLL0);
+					     BIT(DPLL_ID_SKL_DPLL0));
 	else
 		pll = intel_find_shared_dpll(state, crtc,
 					     &crtc_state->dpll_hw_state,
-					     DPLL_ID_SKL_DPLL1,
-					     DPLL_ID_SKL_DPLL3);
+					     BIT(DPLL_ID_SKL_DPLL3) |
+					     BIT(DPLL_ID_SKL_DPLL2) |
+					     BIT(DPLL_ID_SKL_DPLL1));
 	if (!pll)
 		return false;
 
@@ -2401,8 +2418,9 @@ static bool cnl_get_dpll(struct intel_atomic_state *state,
 
 	pll = intel_find_shared_dpll(state, crtc,
 				     &crtc_state->dpll_hw_state,
-				     DPLL_ID_SKL_DPLL0,
-				     DPLL_ID_SKL_DPLL2);
+				     BIT(DPLL_ID_SKL_DPLL2) |
+				     BIT(DPLL_ID_SKL_DPLL1) |
+				     BIT(DPLL_ID_SKL_DPLL0));
 	if (!pll) {
 		DRM_DEBUG_KMS("No PLL selected\n");
 		return false;
@@ -2520,6 +2538,18 @@ static const struct skl_wrpll_params icl_tbt_pll_19_2MHz_values = {
 	.pdiv = 0x4 /* 5 */, .kdiv = 1, .qdiv_mode = 0, .qdiv_ratio = 0,
 };
 
+static const struct skl_wrpll_params tgl_tbt_pll_19_2MHz_values = {
+	.dco_integer = 0x54, .dco_fraction = 0x3000,
+	/* the following params are unused */
+	.pdiv = 0, .kdiv = 0, .qdiv_mode = 0, .qdiv_ratio = 0,
+};
+
+static const struct skl_wrpll_params tgl_tbt_pll_24MHz_values = {
+	.dco_integer = 0x43, .dco_fraction = 0x4000,
+	/* the following params are unused */
+	.pdiv = 0, .kdiv = 0, .qdiv_mode = 0, .qdiv_ratio = 0,
+};
+
 static bool icl_calc_dp_combo_pll(struct intel_crtc_state *crtc_state,
 				  struct skl_wrpll_params *pll_params)
 {
@@ -2547,8 +2577,34 @@ static bool icl_calc_tbt_pll(struct intel_crtc_state *crtc_state,
 {
 	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
 
-	*pll_params = dev_priv->cdclk.hw.ref == 24000 ?
-			icl_tbt_pll_24MHz_values : icl_tbt_pll_19_2MHz_values;
+	if (INTEL_GEN(dev_priv) >= 12) {
+		switch (dev_priv->cdclk.hw.ref) {
+		default:
+			MISSING_CASE(dev_priv->cdclk.hw.ref);
+			/* fall-through */
+		case 19200:
+		case 38400:
+			*pll_params = tgl_tbt_pll_19_2MHz_values;
+			break;
+		case 24000:
+			*pll_params = tgl_tbt_pll_24MHz_values;
+			break;
+		}
+	} else {
+		switch (dev_priv->cdclk.hw.ref) {
+		default:
+			MISSING_CASE(dev_priv->cdclk.hw.ref);
+			/* fall-through */
+		case 19200:
+		case 38400:
+			*pll_params = icl_tbt_pll_19_2MHz_values;
+			break;
+		case 24000:
+			*pll_params = icl_tbt_pll_24MHz_values;
+			break;
+		}
+	}
+
 	return true;
 }
 
@@ -2607,7 +2663,8 @@ enum intel_dpll_id icl_tc_port_to_pll_id(enum tc_port tc_port)
 
 static bool icl_mg_pll_find_divisors(int clock_khz, bool is_dp, bool use_ssc,
 				     u32 *target_dco_khz,
-				     struct intel_dpll_hw_state *state)
+				     struct intel_dpll_hw_state *state,
+				     bool is_dkl)
 {
 	u32 dco_min_freq, dco_max_freq;
 	int div1_vals[] = {7, 5, 3, 2};
@@ -2629,8 +2686,13 @@ static bool icl_mg_pll_find_divisors(int clock_khz, bool is_dp, bool use_ssc,
 				continue;
 
 			if (div2 >= 2) {
+				/*
+				 * Note: a_divratio not matching TGL BSpec
+				 * algorithm but matching hardcoded values and
+				 * working on HW for DP alt-mode at least
+				 */
 				a_divratio = is_dp ? 10 : 5;
-				tlinedrv = 2;
+				tlinedrv = is_dkl ? 1 : 2;
 			} else {
 				a_divratio = 5;
 				tlinedrv = 0;
@@ -2693,11 +2755,12 @@ static bool icl_calc_mg_pll_state(struct intel_crtc_state *crtc_state,
 	u64 tmp;
 	bool use_ssc = false;
 	bool is_dp = !intel_crtc_has_type(crtc_state, INTEL_OUTPUT_HDMI);
+	bool is_dkl = INTEL_GEN(dev_priv) >= 12;
 
 	memset(pll_state, 0, sizeof(*pll_state));
 
 	if (!icl_mg_pll_find_divisors(clock, is_dp, use_ssc, &dco_khz,
-				      pll_state)) {
+				      pll_state, is_dkl)) {
 		DRM_DEBUG_KMS("Failed to find divisors for clock %d\n", clock);
 		return false;
 	}
@@ -2705,8 +2768,11 @@ static bool icl_calc_mg_pll_state(struct intel_crtc_state *crtc_state,
 	m1div = 2;
 	m2div_int = dco_khz / (refclk_khz * m1div);
 	if (m2div_int > 255) {
-		m1div = 4;
-		m2div_int = dco_khz / (refclk_khz * m1div);
+		if (!is_dkl) {
+			m1div = 4;
+			m2div_int = dco_khz / (refclk_khz * m1div);
+		}
+
 		if (m2div_int > 255) {
 			DRM_DEBUG_KMS("Failed to find mdiv for clock %d\n",
 				      clock);
@@ -2786,60 +2852,94 @@ static bool icl_calc_mg_pll_state(struct intel_crtc_state *crtc_state,
 	}
 	ssc_steplog = 4;
 
-	pll_state->mg_pll_div0 = (m2div_rem > 0 ? MG_PLL_DIV0_FRACNEN_H : 0) |
-				  MG_PLL_DIV0_FBDIV_FRAC(m2div_frac) |
-				  MG_PLL_DIV0_FBDIV_INT(m2div_int);
-
-	pll_state->mg_pll_div1 = MG_PLL_DIV1_IREF_NDIVRATIO(iref_ndiv) |
-				 MG_PLL_DIV1_DITHER_DIV_2 |
-				 MG_PLL_DIV1_NDIVRATIO(1) |
-				 MG_PLL_DIV1_FBPREDIV(m1div);
-
-	pll_state->mg_pll_lf = MG_PLL_LF_TDCTARGETCNT(tdc_targetcnt) |
-			       MG_PLL_LF_AFCCNTSEL_512 |
-			       MG_PLL_LF_GAINCTRL(1) |
-			       MG_PLL_LF_INT_COEFF(int_coeff) |
-			       MG_PLL_LF_PROP_COEFF(prop_coeff);
-
-	pll_state->mg_pll_frac_lock = MG_PLL_FRAC_LOCK_TRUELOCK_CRIT_32 |
-				      MG_PLL_FRAC_LOCK_EARLYLOCK_CRIT_32 |
-				      MG_PLL_FRAC_LOCK_LOCKTHRESH(10) |
-				      MG_PLL_FRAC_LOCK_DCODITHEREN |
-				      MG_PLL_FRAC_LOCK_FEEDFWRDGAIN(feedfwgain);
-	if (use_ssc || m2div_rem > 0)
-		pll_state->mg_pll_frac_lock |= MG_PLL_FRAC_LOCK_FEEDFWRDCAL_EN;
-
-	pll_state->mg_pll_ssc = (use_ssc ? MG_PLL_SSC_EN : 0) |
-				MG_PLL_SSC_TYPE(2) |
-				MG_PLL_SSC_STEPLENGTH(ssc_steplen) |
-				MG_PLL_SSC_STEPNUM(ssc_steplog) |
-				MG_PLL_SSC_FLLEN |
-				MG_PLL_SSC_STEPSIZE(ssc_stepsize);
-
-	pll_state->mg_pll_tdc_coldst_bias = MG_PLL_TDC_COLDST_COLDSTART |
-					    MG_PLL_TDC_COLDST_IREFINT_EN |
-					    MG_PLL_TDC_COLDST_REFBIAS_START_PULSE_W(iref_pulse_w) |
-					    MG_PLL_TDC_TDCOVCCORR_EN |
-					    MG_PLL_TDC_TDCSEL(3);
-
-	pll_state->mg_pll_bias = MG_PLL_BIAS_BIAS_GB_SEL(3) |
-				 MG_PLL_BIAS_INIT_DCOAMP(0x3F) |
-				 MG_PLL_BIAS_BIAS_BONUS(10) |
-				 MG_PLL_BIAS_BIASCAL_EN |
-				 MG_PLL_BIAS_CTRIM(12) |
-				 MG_PLL_BIAS_VREF_RDAC(4) |
-				 MG_PLL_BIAS_IREFTRIM(iref_trim);
-
-	if (refclk_khz == 38400) {
-		pll_state->mg_pll_tdc_coldst_bias_mask = MG_PLL_TDC_COLDST_COLDSTART;
-		pll_state->mg_pll_bias_mask = 0;
+	/* write pll_state calculations */
+	if (is_dkl) {
+		pll_state->mg_pll_div0 = DKL_PLL_DIV0_INTEG_COEFF(int_coeff) |
+					 DKL_PLL_DIV0_PROP_COEFF(prop_coeff) |
+					 DKL_PLL_DIV0_FBPREDIV(m1div) |
+					 DKL_PLL_DIV0_FBDIV_INT(m2div_int);
+
+		pll_state->mg_pll_div1 = DKL_PLL_DIV1_IREF_TRIM(iref_trim) |
+					 DKL_PLL_DIV1_TDC_TARGET_CNT(tdc_targetcnt);
+
+		pll_state->mg_pll_ssc = DKL_PLL_SSC_IREF_NDIV_RATIO(iref_ndiv) |
+					DKL_PLL_SSC_STEP_LEN(ssc_steplen) |
+					DKL_PLL_SSC_STEP_NUM(ssc_steplog) |
+					(use_ssc ? DKL_PLL_SSC_EN : 0);
+
+		pll_state->mg_pll_bias = (m2div_frac ? DKL_PLL_BIAS_FRAC_EN_H : 0) |
+					  DKL_PLL_BIAS_FBDIV_FRAC(m2div_frac);
+
+		pll_state->mg_pll_tdc_coldst_bias =
+				DKL_PLL_TDC_SSC_STEP_SIZE(ssc_stepsize) |
+				DKL_PLL_TDC_FEED_FWD_GAIN(feedfwgain);
+
 	} else {
-		pll_state->mg_pll_tdc_coldst_bias_mask = -1U;
-		pll_state->mg_pll_bias_mask = -1U;
-	}
+		pll_state->mg_pll_div0 =
+			(m2div_rem > 0 ? MG_PLL_DIV0_FRACNEN_H : 0) |
+			MG_PLL_DIV0_FBDIV_FRAC(m2div_frac) |
+			MG_PLL_DIV0_FBDIV_INT(m2div_int);
+
+		pll_state->mg_pll_div1 =
+			MG_PLL_DIV1_IREF_NDIVRATIO(iref_ndiv) |
+			MG_PLL_DIV1_DITHER_DIV_2 |
+			MG_PLL_DIV1_NDIVRATIO(1) |
+			MG_PLL_DIV1_FBPREDIV(m1div);
+
+		pll_state->mg_pll_lf =
+			MG_PLL_LF_TDCTARGETCNT(tdc_targetcnt) |
+			MG_PLL_LF_AFCCNTSEL_512 |
+			MG_PLL_LF_GAINCTRL(1) |
+			MG_PLL_LF_INT_COEFF(int_coeff) |
+			MG_PLL_LF_PROP_COEFF(prop_coeff);
+
+		pll_state->mg_pll_frac_lock =
+			MG_PLL_FRAC_LOCK_TRUELOCK_CRIT_32 |
+			MG_PLL_FRAC_LOCK_EARLYLOCK_CRIT_32 |
+			MG_PLL_FRAC_LOCK_LOCKTHRESH(10) |
+			MG_PLL_FRAC_LOCK_DCODITHEREN |
+			MG_PLL_FRAC_LOCK_FEEDFWRDGAIN(feedfwgain);
+		if (use_ssc || m2div_rem > 0)
+			pll_state->mg_pll_frac_lock |=
+				MG_PLL_FRAC_LOCK_FEEDFWRDCAL_EN;
+
+		pll_state->mg_pll_ssc =
+			(use_ssc ? MG_PLL_SSC_EN : 0) |
+			MG_PLL_SSC_TYPE(2) |
+			MG_PLL_SSC_STEPLENGTH(ssc_steplen) |
+			MG_PLL_SSC_STEPNUM(ssc_steplog) |
+			MG_PLL_SSC_FLLEN |
+			MG_PLL_SSC_STEPSIZE(ssc_stepsize);
+
+		pll_state->mg_pll_tdc_coldst_bias =
+			MG_PLL_TDC_COLDST_COLDSTART |
+			MG_PLL_TDC_COLDST_IREFINT_EN |
+			MG_PLL_TDC_COLDST_REFBIAS_START_PULSE_W(iref_pulse_w) |
+			MG_PLL_TDC_TDCOVCCORR_EN |
+			MG_PLL_TDC_TDCSEL(3);
+
+		pll_state->mg_pll_bias =
+			MG_PLL_BIAS_BIAS_GB_SEL(3) |
+			MG_PLL_BIAS_INIT_DCOAMP(0x3F) |
+			MG_PLL_BIAS_BIAS_BONUS(10) |
+			MG_PLL_BIAS_BIASCAL_EN |
+			MG_PLL_BIAS_CTRIM(12) |
+			MG_PLL_BIAS_VREF_RDAC(4) |
+			MG_PLL_BIAS_IREFTRIM(iref_trim);
+
+		if (refclk_khz == 38400) {
+			pll_state->mg_pll_tdc_coldst_bias_mask =
+				MG_PLL_TDC_COLDST_COLDSTART;
+			pll_state->mg_pll_bias_mask = 0;
+		} else {
+			pll_state->mg_pll_tdc_coldst_bias_mask = -1U;
+			pll_state->mg_pll_bias_mask = -1U;
+		}
 
-	pll_state->mg_pll_tdc_coldst_bias &= pll_state->mg_pll_tdc_coldst_bias_mask;
-	pll_state->mg_pll_bias &= pll_state->mg_pll_bias_mask;
+		pll_state->mg_pll_tdc_coldst_bias &=
+			pll_state->mg_pll_tdc_coldst_bias_mask;
+		pll_state->mg_pll_bias &= pll_state->mg_pll_bias_mask;
+	}
 
 	return true;
 }
@@ -2893,7 +2993,7 @@ static bool icl_get_combo_phy_dpll(struct intel_atomic_state *state,
 		&crtc_state->icl_port_dplls[ICL_PORT_DPLL_DEFAULT];
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	enum port port = encoder->port;
-	bool has_dpll4 = false;
+	unsigned long dpll_mask;
 
 	if (!icl_calc_dpll_state(crtc_state, encoder, &port_dpll->hw_state)) {
 		DRM_DEBUG_KMS("Could not calculate combo PHY PLL state.\n");
@@ -2902,16 +3002,19 @@ static bool icl_get_combo_phy_dpll(struct intel_atomic_state *state,
 	}
 
 	if (IS_ELKHARTLAKE(dev_priv) && port != PORT_A)
-		has_dpll4 = true;
+		dpll_mask =
+			BIT(DPLL_ID_EHL_DPLL4) |
+			BIT(DPLL_ID_ICL_DPLL1) |
+			BIT(DPLL_ID_ICL_DPLL0);
+	else
+		dpll_mask = BIT(DPLL_ID_ICL_DPLL1) | BIT(DPLL_ID_ICL_DPLL0);
 
 	port_dpll->pll = intel_find_shared_dpll(state, crtc,
 						&port_dpll->hw_state,
-						DPLL_ID_ICL_DPLL0,
-						has_dpll4 ? DPLL_ID_EHL_DPLL4
-							  : DPLL_ID_ICL_DPLL1);
+						dpll_mask);
 	if (!port_dpll->pll) {
-		DRM_DEBUG_KMS("No combo PHY PLL found for port %c\n",
-			      port_name(encoder->port));
+		DRM_DEBUG_KMS("No combo PHY PLL found for [ENCODER:%d:%s]\n",
+			      encoder->base.base.id, encoder->base.name);
 		return false;
 	}
 
@@ -2941,8 +3044,7 @@ static bool icl_get_tc_phy_dplls(struct intel_atomic_state *state,
 
 	port_dpll->pll = intel_find_shared_dpll(state, crtc,
 						&port_dpll->hw_state,
-						DPLL_ID_ICL_TBTPLL,
-						DPLL_ID_ICL_TBTPLL);
+						BIT(DPLL_ID_ICL_TBTPLL));
 	if (!port_dpll->pll) {
 		DRM_DEBUG_KMS("No TBT-ALT PLL found\n");
 		return false;
@@ -2961,8 +3063,7 @@ static bool icl_get_tc_phy_dplls(struct intel_atomic_state *state,
 							 encoder->port));
 	port_dpll->pll = intel_find_shared_dpll(state, crtc,
 						&port_dpll->hw_state,
-						dpll_id,
-						dpll_id);
+						BIT(dpll_id));
 	if (!port_dpll->pll) {
 		DRM_DEBUG_KMS("No MG PHY PLL found\n");
 		goto err_unreference_tbt_pll;
@@ -3086,6 +3187,78 @@ out:
 	return ret;
 }
 
+static bool dkl_pll_get_hw_state(struct drm_i915_private *dev_priv,
+				 struct intel_shared_dpll *pll,
+				 struct intel_dpll_hw_state *hw_state)
+{
+	const enum intel_dpll_id id = pll->info->id;
+	enum tc_port tc_port = icl_pll_id_to_tc_port(id);
+	intel_wakeref_t wakeref;
+	bool ret = false;
+	u32 val;
+
+	wakeref = intel_display_power_get_if_enabled(dev_priv,
+						     POWER_DOMAIN_DISPLAY_CORE);
+	if (!wakeref)
+		return false;
+
+	val = I915_READ(MG_PLL_ENABLE(tc_port));
+	if (!(val & PLL_ENABLE))
+		goto out;
+
+	/*
+	 * All registers read here have the same HIP_INDEX_REG even though
+	 * they are on different building blocks
+	 */
+	I915_WRITE(HIP_INDEX_REG(tc_port), HIP_INDEX_VAL(tc_port, 0x2));
+
+	hw_state->mg_refclkin_ctl = I915_READ(DKL_REFCLKIN_CTL(tc_port));
+	hw_state->mg_refclkin_ctl &= MG_REFCLKIN_CTL_OD_2_MUX_MASK;
+
+	hw_state->mg_clktop2_hsclkctl =
+		I915_READ(DKL_CLKTOP2_HSCLKCTL(tc_port));
+	hw_state->mg_clktop2_hsclkctl &=
+		MG_CLKTOP2_HSCLKCTL_TLINEDRV_CLKSEL_MASK |
+		MG_CLKTOP2_HSCLKCTL_CORE_INPUTSEL_MASK |
+		MG_CLKTOP2_HSCLKCTL_HSDIV_RATIO_MASK |
+		MG_CLKTOP2_HSCLKCTL_DSDIV_RATIO_MASK;
+
+	hw_state->mg_clktop2_coreclkctl1 =
+		I915_READ(DKL_CLKTOP2_CORECLKCTL1(tc_port));
+	hw_state->mg_clktop2_coreclkctl1 &=
+		MG_CLKTOP2_CORECLKCTL1_A_DIVRATIO_MASK;
+
+	hw_state->mg_pll_div0 = I915_READ(DKL_PLL_DIV0(tc_port));
+	hw_state->mg_pll_div0 &= (DKL_PLL_DIV0_INTEG_COEFF_MASK |
+				  DKL_PLL_DIV0_PROP_COEFF_MASK |
+				  DKL_PLL_DIV0_FBPREDIV_MASK |
+				  DKL_PLL_DIV0_FBDIV_INT_MASK);
+
+	hw_state->mg_pll_div1 = I915_READ(DKL_PLL_DIV1(tc_port));
+	hw_state->mg_pll_div1 &= (DKL_PLL_DIV1_IREF_TRIM_MASK |
+				  DKL_PLL_DIV1_TDC_TARGET_CNT_MASK);
+
+	hw_state->mg_pll_ssc = I915_READ(DKL_PLL_SSC(tc_port));
+	hw_state->mg_pll_ssc &= (DKL_PLL_SSC_IREF_NDIV_RATIO_MASK |
+				 DKL_PLL_SSC_STEP_LEN_MASK |
+				 DKL_PLL_SSC_STEP_NUM_MASK |
+				 DKL_PLL_SSC_EN);
+
+	hw_state->mg_pll_bias = I915_READ(DKL_PLL_BIAS(tc_port));
+	hw_state->mg_pll_bias &= (DKL_PLL_BIAS_FRAC_EN_H |
+				  DKL_PLL_BIAS_FBDIV_FRAC_MASK);
+
+	hw_state->mg_pll_tdc_coldst_bias =
+		I915_READ(DKL_PLL_TDC_COLDST_BIAS(tc_port));
+	hw_state->mg_pll_tdc_coldst_bias &= (DKL_PLL_TDC_SSC_STEP_SIZE_MASK |
+					     DKL_PLL_TDC_FEED_FWD_GAIN_MASK);
+
+	ret = true;
+out:
+	intel_display_power_put(dev_priv, POWER_DOMAIN_DISPLAY_CORE, wakeref);
+	return ret;
+}
+
 static bool icl_pll_get_hw_state(struct drm_i915_private *dev_priv,
 				 struct intel_shared_dpll *pll,
 				 struct intel_dpll_hw_state *hw_state,
@@ -3220,6 +3393,75 @@ static void icl_mg_pll_write(struct drm_i915_private *dev_priv,
 	POSTING_READ(MG_PLL_TDC_COLDST_BIAS(tc_port));
 }
 
+static void dkl_pll_write(struct drm_i915_private *dev_priv,
+			  struct intel_shared_dpll *pll)
+{
+	struct intel_dpll_hw_state *hw_state = &pll->state.hw_state;
+	enum tc_port tc_port = icl_pll_id_to_tc_port(pll->info->id);
+	u32 val;
+
+	/*
+	 * All registers programmed here have the same HIP_INDEX_REG even
+	 * though on different building block
+	 */
+	I915_WRITE(HIP_INDEX_REG(tc_port), HIP_INDEX_VAL(tc_port, 0x2));
+
+	/* All the registers are RMW */
+	val = I915_READ(DKL_REFCLKIN_CTL(tc_port));
+	val &= ~MG_REFCLKIN_CTL_OD_2_MUX_MASK;
+	val |= hw_state->mg_refclkin_ctl;
+	I915_WRITE(DKL_REFCLKIN_CTL(tc_port), val);
+
+	val = I915_READ(DKL_CLKTOP2_CORECLKCTL1(tc_port));
+	val &= ~MG_CLKTOP2_CORECLKCTL1_A_DIVRATIO_MASK;
+	val |= hw_state->mg_clktop2_coreclkctl1;
+	I915_WRITE(DKL_CLKTOP2_CORECLKCTL1(tc_port), val);
+
+	val = I915_READ(DKL_CLKTOP2_HSCLKCTL(tc_port));
+	val &= ~(MG_CLKTOP2_HSCLKCTL_TLINEDRV_CLKSEL_MASK |
+		 MG_CLKTOP2_HSCLKCTL_CORE_INPUTSEL_MASK |
+		 MG_CLKTOP2_HSCLKCTL_HSDIV_RATIO_MASK |
+		 MG_CLKTOP2_HSCLKCTL_DSDIV_RATIO_MASK);
+	val |= hw_state->mg_clktop2_hsclkctl;
+	I915_WRITE(DKL_CLKTOP2_HSCLKCTL(tc_port), val);
+
+	val = I915_READ(DKL_PLL_DIV0(tc_port));
+	val &= ~(DKL_PLL_DIV0_INTEG_COEFF_MASK |
+		 DKL_PLL_DIV0_PROP_COEFF_MASK |
+		 DKL_PLL_DIV0_FBPREDIV_MASK |
+		 DKL_PLL_DIV0_FBDIV_INT_MASK);
+	val |= hw_state->mg_pll_div0;
+	I915_WRITE(DKL_PLL_DIV0(tc_port), val);
+
+	val = I915_READ(DKL_PLL_DIV1(tc_port));
+	val &= ~(DKL_PLL_DIV1_IREF_TRIM_MASK |
+		 DKL_PLL_DIV1_TDC_TARGET_CNT_MASK);
+	val |= hw_state->mg_pll_div1;
+	I915_WRITE(DKL_PLL_DIV1(tc_port), val);
+
+	val = I915_READ(DKL_PLL_SSC(tc_port));
+	val &= ~(DKL_PLL_SSC_IREF_NDIV_RATIO_MASK |
+		 DKL_PLL_SSC_STEP_LEN_MASK |
+		 DKL_PLL_SSC_STEP_NUM_MASK |
+		 DKL_PLL_SSC_EN);
+	val |= hw_state->mg_pll_ssc;
+	I915_WRITE(DKL_PLL_SSC(tc_port), val);
+
+	val = I915_READ(DKL_PLL_BIAS(tc_port));
+	val &= ~(DKL_PLL_BIAS_FRAC_EN_H |
+		 DKL_PLL_BIAS_FBDIV_FRAC_MASK);
+	val |= hw_state->mg_pll_bias;
+	I915_WRITE(DKL_PLL_BIAS(tc_port), val);
+
+	val = I915_READ(DKL_PLL_TDC_COLDST_BIAS(tc_port));
+	val &= ~(DKL_PLL_TDC_SSC_STEP_SIZE_MASK |
+		 DKL_PLL_TDC_FEED_FWD_GAIN_MASK);
+	val |= hw_state->mg_pll_tdc_coldst_bias;
+	I915_WRITE(DKL_PLL_TDC_COLDST_BIAS(tc_port), val);
+
+	POSTING_READ(DKL_PLL_TDC_COLDST_BIAS(tc_port));
+}
+
 static void icl_pll_power_enable(struct drm_i915_private *dev_priv,
 				 struct intel_shared_dpll *pll,
 				 i915_reg_t enable_reg)
@@ -3312,7 +3554,10 @@ static void mg_pll_enable(struct drm_i915_private *dev_priv,
 
 	icl_pll_power_enable(dev_priv, pll, enable_reg);
 
-	icl_mg_pll_write(dev_priv, pll);
+	if (INTEL_GEN(dev_priv) >= 12)
+		dkl_pll_write(dev_priv, pll);
+	else
+		icl_mg_pll_write(dev_priv, pll);
 
 	/*
 	 * DVFS pre sequence would be here, but in our driver the cdclk code
@@ -3467,11 +3712,22 @@ static const struct intel_dpll_mgr ehl_pll_mgr = {
 	.dump_hw_state = icl_dump_hw_state,
 };
 
+static const struct intel_shared_dpll_funcs dkl_pll_funcs = {
+	.enable = mg_pll_enable,
+	.disable = mg_pll_disable,
+	.get_hw_state = dkl_pll_get_hw_state,
+};
+
 static const struct dpll_info tgl_plls[] = {
 	{ "DPLL 0", &combo_pll_funcs, DPLL_ID_ICL_DPLL0,  0 },
 	{ "DPLL 1", &combo_pll_funcs, DPLL_ID_ICL_DPLL1,  0 },
 	{ "TBT PLL",  &tbt_pll_funcs, DPLL_ID_ICL_TBTPLL, 0 },
-	/* TODO: Add typeC plls */
+	{ "TC PLL 1", &dkl_pll_funcs, DPLL_ID_ICL_MGPLL1, 0 },
+	{ "TC PLL 2", &dkl_pll_funcs, DPLL_ID_ICL_MGPLL2, 0 },
+	{ "TC PLL 3", &dkl_pll_funcs, DPLL_ID_ICL_MGPLL3, 0 },
+	{ "TC PLL 4", &dkl_pll_funcs, DPLL_ID_ICL_MGPLL4, 0 },
+	{ "TC PLL 5", &dkl_pll_funcs, DPLL_ID_TGL_MGPLL5, 0 },
+	{ "TC PLL 6", &dkl_pll_funcs, DPLL_ID_TGL_MGPLL6, 0 },
 	{ },
 };
 
@@ -3479,6 +3735,7 @@ static const struct intel_dpll_mgr tgl_pll_mgr = {
 	.dpll_info = tgl_plls,
 	.get_dplls = icl_get_dplls,
 	.put_dplls = icl_put_dplls,
+	.update_active_dpll = icl_update_active_dpll,
 	.dump_hw_state = icl_dump_hw_state,
 };
 
diff --git a/drivers/gpu/drm/i915/display/intel_dpll_mgr.h b/drivers/gpu/drm/i915/display/intel_dpll_mgr.h
index e7588799fce5..2a104c64291d 100644
--- a/drivers/gpu/drm/i915/display/intel_dpll_mgr.h
+++ b/drivers/gpu/drm/i915/display/intel_dpll_mgr.h
@@ -147,11 +147,11 @@ enum intel_dpll_id {
 	 */
 	DPLL_ID_ICL_MGPLL4 = 6,
 	/**
-	 * @DPLL_ID_TGL_TCPLL5: TGL TC PLL port 5 (TC5)
+	 * @DPLL_ID_TGL_MGPLL5: TGL TC PLL port 5 (TC5)
 	 */
 	DPLL_ID_TGL_MGPLL5 = 7,
 	/**
-	 * @DPLL_ID_TGL_TCPLL6: TGL TC PLL port 6 (TC6)
+	 * @DPLL_ID_TGL_MGPLL6: TGL TC PLL port 6 (TC6)
 	 */
 	DPLL_ID_TGL_MGPLL6 = 8,
 };
@@ -337,6 +337,11 @@ struct intel_shared_dpll {
 	 * @info: platform specific info
 	 */
 	const struct dpll_info *info;
+
+	/**
+	 * @wakeref: In some platforms a device-level runtime pm reference may
+	 * need to be grabbed to disable DC states while this DPLL is enabled
+	 */
 	intel_wakeref_t wakeref;
 };
 
diff --git a/drivers/gpu/drm/i915/display/intel_dsb.c b/drivers/gpu/drm/i915/display/intel_dsb.c
new file mode 100644
index 000000000000..bb5a0e91b370
--- /dev/null
+++ b/drivers/gpu/drm/i915/display/intel_dsb.c
@@ -0,0 +1,332 @@
+// SPDX-License-Identifier: MIT
+/*
+ * Copyright © 2019 Intel Corporation
+ *
+ */
+
+#include "i915_drv.h"
+#include "intel_display_types.h"
+
+#define DSB_BUF_SIZE    (2 * PAGE_SIZE)
+
+/**
+ * DOC: DSB
+ *
+ * A DSB (Display State Buffer) is a queue of MMIO instructions in the memory
+ * which can be offloaded to DSB HW in Display Controller. DSB HW is a DMA
+ * engine that can be programmed to download the DSB from memory.
+ * It allows driver to batch submit display HW programming. This helps to
+ * reduce loading time and CPU activity, thereby making the context switch
+ * faster. DSB Support added from Gen12 Intel graphics based platform.
+ *
+ * DSB's can access only the pipe, plane, and transcoder Data Island Packet
+ * registers.
+ *
+ * DSB HW can support only register writes (both indexed and direct MMIO
+ * writes). There are no registers reads possible with DSB HW engine.
+ */
+
+/* DSB opcodes. */
+#define DSB_OPCODE_SHIFT		24
+#define DSB_OPCODE_MMIO_WRITE		0x1
+#define DSB_OPCODE_INDEXED_WRITE	0x9
+#define DSB_BYTE_EN			0xF
+#define DSB_BYTE_EN_SHIFT		20
+#define DSB_REG_VALUE_MASK		0xfffff
+
+static inline bool is_dsb_busy(struct intel_dsb *dsb)
+{
+	struct intel_crtc *crtc = container_of(dsb, typeof(*crtc), dsb);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	enum pipe pipe = crtc->pipe;
+
+	return DSB_STATUS & I915_READ(DSB_CTRL(pipe, dsb->id));
+}
+
+static inline bool intel_dsb_enable_engine(struct intel_dsb *dsb)
+{
+	struct intel_crtc *crtc = container_of(dsb, typeof(*crtc), dsb);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	enum pipe pipe = crtc->pipe;
+	u32 dsb_ctrl;
+
+	dsb_ctrl = I915_READ(DSB_CTRL(pipe, dsb->id));
+	if (DSB_STATUS & dsb_ctrl) {
+		DRM_DEBUG_KMS("DSB engine is busy.\n");
+		return false;
+	}
+
+	dsb_ctrl |= DSB_ENABLE;
+	I915_WRITE(DSB_CTRL(pipe, dsb->id), dsb_ctrl);
+
+	POSTING_READ(DSB_CTRL(pipe, dsb->id));
+	return true;
+}
+
+static inline bool intel_dsb_disable_engine(struct intel_dsb *dsb)
+{
+	struct intel_crtc *crtc = container_of(dsb, typeof(*crtc), dsb);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	enum pipe pipe = crtc->pipe;
+	u32 dsb_ctrl;
+
+	dsb_ctrl = I915_READ(DSB_CTRL(pipe, dsb->id));
+	if (DSB_STATUS & dsb_ctrl) {
+		DRM_DEBUG_KMS("DSB engine is busy.\n");
+		return false;
+	}
+
+	dsb_ctrl &= ~DSB_ENABLE;
+	I915_WRITE(DSB_CTRL(pipe, dsb->id), dsb_ctrl);
+
+	POSTING_READ(DSB_CTRL(pipe, dsb->id));
+	return true;
+}
+
+/**
+ * intel_dsb_get() - Allocate DSB context and return a DSB instance.
+ * @crtc: intel_crtc structure to get pipe info.
+ *
+ * This function provides handle of a DSB instance, for the further DSB
+ * operations.
+ *
+ * Returns: address of Intel_dsb instance requested for.
+ * Failure: Returns the same DSB instance, but without a command buffer.
+ */
+
+struct intel_dsb *
+intel_dsb_get(struct intel_crtc *crtc)
+{
+	struct drm_device *dev = crtc->base.dev;
+	struct drm_i915_private *i915 = to_i915(dev);
+	struct intel_dsb *dsb = &crtc->dsb;
+	struct drm_i915_gem_object *obj;
+	struct i915_vma *vma;
+	intel_wakeref_t wakeref;
+
+	if (!HAS_DSB(i915))
+		return dsb;
+
+	if (atomic_add_return(1, &dsb->refcount) != 1)
+		return dsb;
+
+	dsb->id = DSB1;
+	wakeref = intel_runtime_pm_get(&i915->runtime_pm);
+
+	obj = i915_gem_object_create_internal(i915, DSB_BUF_SIZE);
+	if (IS_ERR(obj)) {
+		DRM_ERROR("Gem object creation failed\n");
+		goto err;
+	}
+
+	vma = i915_gem_object_ggtt_pin(obj, NULL, 0, 0, PIN_MAPPABLE);
+	if (IS_ERR(vma)) {
+		DRM_ERROR("Vma creation failed\n");
+		i915_gem_object_put(obj);
+		atomic_dec(&dsb->refcount);
+		goto err;
+	}
+
+	dsb->cmd_buf = i915_gem_object_pin_map(vma->obj, I915_MAP_WC);
+	if (IS_ERR(dsb->cmd_buf)) {
+		DRM_ERROR("Command buffer creation failed\n");
+		i915_vma_unpin_and_release(&vma, 0);
+		dsb->cmd_buf = NULL;
+		atomic_dec(&dsb->refcount);
+		goto err;
+	}
+	dsb->vma = vma;
+
+err:
+	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
+	return dsb;
+}
+
+/**
+ * intel_dsb_put() - To destroy DSB context.
+ * @dsb: intel_dsb structure.
+ *
+ * This function destroys the DSB context allocated by a dsb_get(), by
+ * unpinning and releasing the VMA object associated with it.
+ */
+
+void intel_dsb_put(struct intel_dsb *dsb)
+{
+	struct intel_crtc *crtc = container_of(dsb, typeof(*crtc), dsb);
+	struct drm_i915_private *i915 = to_i915(crtc->base.dev);
+
+	if (!HAS_DSB(i915))
+		return;
+
+	if (WARN_ON(atomic_read(&dsb->refcount) == 0))
+		return;
+
+	if (atomic_dec_and_test(&dsb->refcount)) {
+		i915_vma_unpin_and_release(&dsb->vma, I915_VMA_RELEASE_MAP);
+		dsb->cmd_buf = NULL;
+		dsb->free_pos = 0;
+		dsb->ins_start_offset = 0;
+	}
+}
+
+/**
+ * intel_dsb_indexed_reg_write() -Write to the DSB context for auto
+ * increment register.
+ * @dsb: intel_dsb structure.
+ * @reg: register address.
+ * @val: value.
+ *
+ * This function is used for writing register-value pair in command
+ * buffer of DSB for auto-increment register. During command buffer overflow,
+ * a warning is thrown and rest all erroneous condition register programming
+ * is done through mmio write.
+ */
+
+void intel_dsb_indexed_reg_write(struct intel_dsb *dsb, i915_reg_t reg,
+				 u32 val)
+{
+	struct intel_crtc *crtc = container_of(dsb, typeof(*crtc), dsb);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	u32 *buf = dsb->cmd_buf;
+	u32 reg_val;
+
+	if (!buf) {
+		I915_WRITE(reg, val);
+		return;
+	}
+
+	if (WARN_ON(dsb->free_pos >= DSB_BUF_SIZE)) {
+		DRM_DEBUG_KMS("DSB buffer overflow\n");
+		return;
+	}
+
+	/*
+	 * For example the buffer will look like below for 3 dwords for auto
+	 * increment register:
+	 * +--------------------------------------------------------+
+	 * | size = 3 | offset &| value1 | value2 | value3 | zero   |
+	 * |          | opcode  |        |        |        |        |
+	 * +--------------------------------------------------------+
+	 * +          +         +        +        +        +        +
+	 * 0          4         8        12       16       20       24
+	 * Byte
+	 *
+	 * As every instruction is 8 byte aligned the index of dsb instruction
+	 * will start always from even number while dealing with u32 array. If
+	 * we are writing odd no of dwords, Zeros will be added in the end for
+	 * padding.
+	 */
+	reg_val = buf[dsb->ins_start_offset + 1] & DSB_REG_VALUE_MASK;
+	if (reg_val != i915_mmio_reg_offset(reg)) {
+		/* Every instruction should be 8 byte aligned. */
+		dsb->free_pos = ALIGN(dsb->free_pos, 2);
+
+		dsb->ins_start_offset = dsb->free_pos;
+
+		/* Update the size. */
+		buf[dsb->free_pos++] = 1;
+
+		/* Update the opcode and reg. */
+		buf[dsb->free_pos++] = (DSB_OPCODE_INDEXED_WRITE  <<
+					DSB_OPCODE_SHIFT) |
+					i915_mmio_reg_offset(reg);
+
+		/* Update the value. */
+		buf[dsb->free_pos++] = val;
+	} else {
+		/* Update the new value. */
+		buf[dsb->free_pos++] = val;
+
+		/* Update the size. */
+		buf[dsb->ins_start_offset]++;
+	}
+
+	/* if number of data words is odd, then the last dword should be 0.*/
+	if (dsb->free_pos & 0x1)
+		buf[dsb->free_pos] = 0;
+}
+
+/**
+ * intel_dsb_reg_write() -Write to the DSB context for normal
+ * register.
+ * @dsb: intel_dsb structure.
+ * @reg: register address.
+ * @val: value.
+ *
+ * This function is used for writing register-value pair in command
+ * buffer of DSB. During command buffer overflow, a warning  is thrown
+ * and rest all erroneous condition register programming is done
+ * through mmio write.
+ */
+void intel_dsb_reg_write(struct intel_dsb *dsb, i915_reg_t reg, u32 val)
+{
+	struct intel_crtc *crtc = container_of(dsb, typeof(*crtc), dsb);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	u32 *buf = dsb->cmd_buf;
+
+	if (!buf) {
+		I915_WRITE(reg, val);
+		return;
+	}
+
+	if (WARN_ON(dsb->free_pos >= DSB_BUF_SIZE)) {
+		DRM_DEBUG_KMS("DSB buffer overflow\n");
+		return;
+	}
+
+	dsb->ins_start_offset = dsb->free_pos;
+	buf[dsb->free_pos++] = val;
+	buf[dsb->free_pos++] = (DSB_OPCODE_MMIO_WRITE  << DSB_OPCODE_SHIFT) |
+			       (DSB_BYTE_EN << DSB_BYTE_EN_SHIFT) |
+			       i915_mmio_reg_offset(reg);
+}
+
+/**
+ * intel_dsb_commit() - Trigger workload execution of DSB.
+ * @dsb: intel_dsb structure.
+ *
+ * This function is used to do actual write to hardware using DSB.
+ * On errors, fall back to MMIO. Also this function help to reset the context.
+ */
+void intel_dsb_commit(struct intel_dsb *dsb)
+{
+	struct intel_crtc *crtc = container_of(dsb, typeof(*crtc), dsb);
+	struct drm_device *dev = crtc->base.dev;
+	struct drm_i915_private *dev_priv = to_i915(dev);
+	enum pipe pipe = crtc->pipe;
+	u32 tail;
+
+	if (!dsb->free_pos)
+		return;
+
+	if (!intel_dsb_enable_engine(dsb))
+		goto reset;
+
+	if (is_dsb_busy(dsb)) {
+		DRM_ERROR("HEAD_PTR write failed - dsb engine is busy.\n");
+		goto reset;
+	}
+	I915_WRITE(DSB_HEAD(pipe, dsb->id), i915_ggtt_offset(dsb->vma));
+
+	tail = ALIGN(dsb->free_pos * 4, CACHELINE_BYTES);
+	if (tail > dsb->free_pos * 4)
+		memset(&dsb->cmd_buf[dsb->free_pos], 0,
+		       (tail - dsb->free_pos * 4));
+
+	if (is_dsb_busy(dsb)) {
+		DRM_ERROR("TAIL_PTR write failed - dsb engine is busy.\n");
+		goto reset;
+	}
+	DRM_DEBUG_KMS("DSB execution started - head 0x%x, tail 0x%x\n",
+		      i915_ggtt_offset(dsb->vma), tail);
+	I915_WRITE(DSB_TAIL(pipe, dsb->id), i915_ggtt_offset(dsb->vma) + tail);
+	if (wait_for(!is_dsb_busy(dsb), 1)) {
+		DRM_ERROR("Timed out waiting for DSB workload completion.\n");
+		goto reset;
+	}
+
+reset:
+	dsb->free_pos = 0;
+	dsb->ins_start_offset = 0;
+	intel_dsb_disable_engine(dsb);
+}
diff --git a/drivers/gpu/drm/i915/display/intel_dsb.h b/drivers/gpu/drm/i915/display/intel_dsb.h
new file mode 100644
index 000000000000..6f95c8e909e6
--- /dev/null
+++ b/drivers/gpu/drm/i915/display/intel_dsb.h
@@ -0,0 +1,52 @@
+/* SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2019 Intel Corporation
+ */
+
+#ifndef _INTEL_DSB_H
+#define _INTEL_DSB_H
+
+#include <linux/types.h>
+
+#include "i915_reg.h"
+
+struct intel_crtc;
+struct i915_vma;
+
+enum dsb_id {
+	INVALID_DSB = -1,
+	DSB1,
+	DSB2,
+	DSB3,
+	MAX_DSB_PER_PIPE
+};
+
+struct intel_dsb {
+	atomic_t refcount;
+	enum dsb_id id;
+	u32 *cmd_buf;
+	struct i915_vma *vma;
+
+	/*
+	 * free_pos will point the first free entry position
+	 * and help in calculating tail of command buffer.
+	 */
+	int free_pos;
+
+	/*
+	 * ins_start_offset will help to store start address of the dsb
+	 * instuction and help in identifying the batch of auto-increment
+	 * register.
+	 */
+	u32 ins_start_offset;
+};
+
+struct intel_dsb *
+intel_dsb_get(struct intel_crtc *crtc);
+void intel_dsb_put(struct intel_dsb *dsb);
+void intel_dsb_reg_write(struct intel_dsb *dsb, i915_reg_t reg, u32 val);
+void intel_dsb_indexed_reg_write(struct intel_dsb *dsb, i915_reg_t reg,
+				 u32 val);
+void intel_dsb_commit(struct intel_dsb *dsb);
+
+#endif
diff --git a/drivers/gpu/drm/i915/display/intel_dsi.c b/drivers/gpu/drm/i915/display/intel_dsi.c
index 5fec02aceaed..a2a937109a5a 100644
--- a/drivers/gpu/drm/i915/display/intel_dsi.c
+++ b/drivers/gpu/drm/i915/display/intel_dsi.c
@@ -55,6 +55,7 @@ int intel_dsi_get_modes(struct drm_connector *connector)
 enum drm_mode_status intel_dsi_mode_valid(struct drm_connector *connector,
 					  struct drm_display_mode *mode)
 {
+	struct drm_i915_private *dev_priv = to_i915(connector->dev);
 	struct intel_connector *intel_connector = to_intel_connector(connector);
 	const struct drm_display_mode *fixed_mode = intel_connector->panel.fixed_mode;
 	int max_dotclk = to_i915(connector->dev)->max_dotclk_freq;
@@ -73,7 +74,7 @@ enum drm_mode_status intel_dsi_mode_valid(struct drm_connector *connector,
 			return MODE_CLOCK_HIGH;
 	}
 
-	return MODE_OK;
+	return intel_mode_valid_max_plane_size(dev_priv, mode);
 }
 
 struct intel_dsi_host *intel_dsi_host_init(struct intel_dsi *intel_dsi,
diff --git a/drivers/gpu/drm/i915/display/intel_dvo.c b/drivers/gpu/drm/i915/display/intel_dvo.c
index 93baf366692e..bcfbcb743e7d 100644
--- a/drivers/gpu/drm/i915/display/intel_dvo.c
+++ b/drivers/gpu/drm/i915/display/intel_dvo.c
@@ -280,7 +280,7 @@ static void intel_dvo_pre_enable(struct intel_encoder *encoder,
 	struct intel_crtc *crtc = to_intel_crtc(pipe_config->base.crtc);
 	const struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode;
 	struct intel_dvo *intel_dvo = enc_to_dvo(encoder);
-	int pipe = crtc->pipe;
+	enum pipe pipe = crtc->pipe;
 	u32 dvo_val;
 	i915_reg_t dvo_reg = intel_dvo->dev.dvo_reg;
 	i915_reg_t dvo_srcdim_reg = intel_dvo->dev.dvo_srcdim_reg;
@@ -505,7 +505,7 @@ void intel_dvo_init(struct drm_i915_private *dev_priv)
 		intel_encoder->type = INTEL_OUTPUT_DVO;
 		intel_encoder->power_domain = POWER_DOMAIN_PORT_OTHER;
 		intel_encoder->port = port;
-		intel_encoder->crtc_mask = (1 << 0) | (1 << 1);
+		intel_encoder->pipe_mask = ~0;
 
 		switch (dvo->type) {
 		case INTEL_DVO_CHIP_TMDS:
diff --git a/drivers/gpu/drm/i915/display/intel_fbc.c b/drivers/gpu/drm/i915/display/intel_fbc.c
index 16ed44bfd734..3111ecaeabd0 100644
--- a/drivers/gpu/drm/i915/display/intel_fbc.c
+++ b/drivers/gpu/drm/i915/display/intel_fbc.c
@@ -343,8 +343,8 @@ static void gen7_fbc_activate(struct drm_i915_private *dev_priv)
 			   HSW_FBCQ_DIS);
 	}
 
-	if (IS_GEN(dev_priv, 11))
-		/* Wa_1409120013:icl,ehl */
+	if (INTEL_GEN(dev_priv) >= 11)
+		/* Wa_1409120013:icl,ehl,tgl */
 		I915_WRITE(ILK_DPFC_CHICKEN, ILK_DPFC_CHICKEN_COMP_DUMMY_PIXEL);
 
 	I915_WRITE(ILK_DPFC_CONTROL, dpfc_ctl | DPFC_CTL_EN);
@@ -1320,6 +1320,9 @@ void intel_fbc_init(struct drm_i915_private *dev_priv)
 	fbc->enabled = false;
 	fbc->active = false;
 
+	if (!drm_mm_initialized(&dev_priv->mm.stolen))
+		mkwrite_device_info(dev_priv)->display.has_fbc = false;
+
 	if (need_fbc_vtd_wa(dev_priv))
 		mkwrite_device_info(dev_priv)->display.has_fbc = false;
 
diff --git a/drivers/gpu/drm/i915/display/intel_fbdev.c b/drivers/gpu/drm/i915/display/intel_fbdev.c
index d59eee5c5d9c..48c960ca12fb 100644
--- a/drivers/gpu/drm/i915/display/intel_fbdev.c
+++ b/drivers/gpu/drm/i915/display/intel_fbdev.c
@@ -141,10 +141,10 @@ static int intelfb_alloc(struct drm_fb_helper *helper,
 	/* If the FB is too big, just don't use it since fbdev is not very
 	 * important and we should probably use that space with FBC or other
 	 * features. */
-	obj = NULL;
+	obj = ERR_PTR(-ENODEV);
 	if (size * 2 < dev_priv->stolen_usable_size)
 		obj = i915_gem_object_create_stolen(dev_priv, size);
-	if (obj == NULL)
+	if (IS_ERR(obj))
 		obj = i915_gem_object_create_shmem(dev_priv, size);
 	if (IS_ERR(obj)) {
 		DRM_ERROR("failed to allocate framebuffer\n");
@@ -204,7 +204,6 @@ static int intelfb_create(struct drm_fb_helper *helper,
 		sizes->fb_height = intel_fb->base.height;
 	}
 
-	mutex_lock(&dev->struct_mutex);
 	wakeref = intel_runtime_pm_get(&dev_priv->runtime_pm);
 
 	/* Pin the GGTT vma for our access via info->screen_base.
@@ -235,6 +234,11 @@ static int intelfb_create(struct drm_fb_helper *helper,
 	info->apertures->ranges[0].base = ggtt->gmadr.start;
 	info->apertures->ranges[0].size = ggtt->mappable_end;
 
+	/* Our framebuffer is the entirety of fbdev's system memory */
+	info->fix.smem_start =
+		(unsigned long)(ggtt->gmadr.start + vma->node.start);
+	info->fix.smem_len = vma->node.size;
+
 	vaddr = i915_vma_pin_iomap(vma);
 	if (IS_ERR(vaddr)) {
 		DRM_ERROR("Failed to remap framebuffer into virtual memory\n");
@@ -244,10 +248,6 @@ static int intelfb_create(struct drm_fb_helper *helper,
 	info->screen_base = vaddr;
 	info->screen_size = vma->node.size;
 
-	/* Our framebuffer is the entirety of fbdev's system memory */
-	info->fix.smem_start = (unsigned long)info->screen_base;
-	info->fix.smem_len = info->screen_size;
-
 	drm_fb_helper_fill_info(info, &ifbdev->helper, sizes);
 
 	/* If the object is shmemfs backed, it will have given us zeroed pages.
@@ -266,7 +266,6 @@ static int intelfb_create(struct drm_fb_helper *helper,
 	ifbdev->vma_flags = flags;
 
 	intel_runtime_pm_put(&dev_priv->runtime_pm, wakeref);
-	mutex_unlock(&dev->struct_mutex);
 	vga_switcheroo_client_fb_set(pdev, info);
 	return 0;
 
@@ -274,7 +273,6 @@ out_unpin:
 	intel_unpin_fb_vma(vma, flags);
 out_unlock:
 	intel_runtime_pm_put(&dev_priv->runtime_pm, wakeref);
-	mutex_unlock(&dev->struct_mutex);
 	return ret;
 }
 
@@ -291,11 +289,8 @@ static void intel_fbdev_destroy(struct intel_fbdev *ifbdev)
 
 	drm_fb_helper_fini(&ifbdev->helper);
 
-	if (ifbdev->vma) {
-		mutex_lock(&ifbdev->helper.dev->struct_mutex);
+	if (ifbdev->vma)
 		intel_unpin_fb_vma(ifbdev->vma, ifbdev->vma_flags);
-		mutex_unlock(&ifbdev->helper.dev->struct_mutex);
-	}
 
 	if (ifbdev->fb)
 		drm_framebuffer_remove(&ifbdev->fb->base);
@@ -444,7 +439,7 @@ int intel_fbdev_init(struct drm_device *dev)
 	struct intel_fbdev *ifbdev;
 	int ret;
 
-	if (WARN_ON(!HAS_DISPLAY(dev_priv)))
+	if (WARN_ON(!HAS_DISPLAY(dev_priv) || !INTEL_DISPLAY_ENABLED(dev_priv)))
 		return -ENODEV;
 
 	ifbdev = kzalloc(sizeof(struct intel_fbdev), GFP_KERNEL);
diff --git a/drivers/gpu/drm/i915/display/intel_frontbuffer.c b/drivers/gpu/drm/i915/display/intel_frontbuffer.c
index 719379774fa5..84b164f31895 100644
--- a/drivers/gpu/drm/i915/display/intel_frontbuffer.c
+++ b/drivers/gpu/drm/i915/display/intel_frontbuffer.c
@@ -206,6 +206,7 @@ static int frontbuffer_active(struct i915_active *ref)
 	return 0;
 }
 
+__i915_active_call
 static void frontbuffer_retire(struct i915_active *ref)
 {
 	struct intel_frontbuffer *front =
@@ -220,11 +221,18 @@ static void frontbuffer_release(struct kref *ref)
 {
 	struct intel_frontbuffer *front =
 		container_of(ref, typeof(*front), ref);
+	struct drm_i915_gem_object *obj = front->obj;
+	struct i915_vma *vma;
 
-	front->obj->frontbuffer = NULL;
-	spin_unlock(&to_i915(front->obj->base.dev)->fb_tracking.lock);
+	spin_lock(&obj->vma.lock);
+	for_each_ggtt_vma(vma, obj)
+		vma->display_alignment = I915_GTT_MIN_ALIGNMENT;
+	spin_unlock(&obj->vma.lock);
 
-	i915_gem_object_put(front->obj);
+	obj->frontbuffer = NULL;
+	spin_unlock(&to_i915(obj->base.dev)->fb_tracking.lock);
+
+	i915_gem_object_put(obj);
 	kfree(front);
 }
 
@@ -249,8 +257,9 @@ intel_frontbuffer_get(struct drm_i915_gem_object *obj)
 	front->obj = obj;
 	kref_init(&front->ref);
 	atomic_set(&front->bits, 0);
-	i915_active_init(i915, &front->write,
-			 frontbuffer_active, frontbuffer_retire);
+	i915_active_init(&front->write,
+			 frontbuffer_active,
+			 i915_active_may_sleep(frontbuffer_retire));
 
 	spin_lock(&i915->fb_tracking.lock);
 	if (obj->frontbuffer) {
diff --git a/drivers/gpu/drm/i915/display/intel_gmbus.c b/drivers/gpu/drm/i915/display/intel_gmbus.c
index d6775a005726..3d4d19ac1d14 100644
--- a/drivers/gpu/drm/i915/display/intel_gmbus.c
+++ b/drivers/gpu/drm/i915/display/intel_gmbus.c
@@ -836,7 +836,7 @@ int intel_gmbus_setup(struct drm_i915_private *dev_priv)
 	unsigned int pin;
 	int ret;
 
-	if (!HAS_DISPLAY(dev_priv))
+	if (!HAS_DISPLAY(dev_priv) || !INTEL_DISPLAY_ENABLED(dev_priv))
 		return 0;
 
 	if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv))
diff --git a/drivers/gpu/drm/i915/display/intel_hdcp.c b/drivers/gpu/drm/i915/display/intel_hdcp.c
index 6ec5ceeab601..f1f41ca8402b 100644
--- a/drivers/gpu/drm/i915/display/intel_hdcp.c
+++ b/drivers/gpu/drm/i915/display/intel_hdcp.c
@@ -1,9 +1,11 @@
 /* SPDX-License-Identifier: MIT */
 /*
  * Copyright (C) 2017 Google, Inc.
+ * Copyright _ 2017-2019, Intel Corporation.
  *
  * Authors:
  * Sean Paul <seanpaul@chromium.org>
+ * Ramalingam C <ramalingam.c@intel.com>
  */
 
 #include <linux/component.h>
@@ -18,6 +20,7 @@
 #include "intel_display_types.h"
 #include "intel_hdcp.h"
 #include "intel_sideband.h"
+#include "intel_connector.h"
 
 #define KEY_LOAD_TRIES	5
 #define ENCRYPT_STATUS_CHANGE_TIMEOUT_MS	50
@@ -105,24 +108,20 @@ bool intel_hdcp2_capable(struct intel_connector *connector)
 	return capable;
 }
 
-static inline bool intel_hdcp_in_use(struct intel_connector *connector)
+static inline
+bool intel_hdcp_in_use(struct drm_i915_private *dev_priv,
+		       enum transcoder cpu_transcoder, enum port port)
 {
-	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
-	enum port port = connector->encoder->port;
-	u32 reg;
-
-	reg = I915_READ(PORT_HDCP_STATUS(port));
-	return reg & HDCP_STATUS_ENC;
+	return I915_READ(HDCP_STATUS(dev_priv, cpu_transcoder, port)) &
+	       HDCP_STATUS_ENC;
 }
 
-static inline bool intel_hdcp2_in_use(struct intel_connector *connector)
+static inline
+bool intel_hdcp2_in_use(struct drm_i915_private *dev_priv,
+			enum transcoder cpu_transcoder, enum port port)
 {
-	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
-	enum port port = connector->encoder->port;
-	u32 reg;
-
-	reg = I915_READ(HDCP2_STATUS_DDI(port));
-	return reg & LINK_ENCRYPTION_STATUS;
+	return I915_READ(HDCP2_STATUS(dev_priv, cpu_transcoder, port)) &
+	       LINK_ENCRYPTION_STATUS;
 }
 
 static int intel_hdcp_poll_ksv_fifo(struct intel_digital_port *intel_dig_port,
@@ -253,9 +252,29 @@ static int intel_write_sha_text(struct drm_i915_private *dev_priv, u32 sha_text)
 }
 
 static
-u32 intel_hdcp_get_repeater_ctl(struct intel_digital_port *intel_dig_port)
+u32 intel_hdcp_get_repeater_ctl(struct drm_i915_private *dev_priv,
+				enum transcoder cpu_transcoder, enum port port)
 {
-	enum port port = intel_dig_port->base.port;
+	if (INTEL_GEN(dev_priv) >= 12) {
+		switch (cpu_transcoder) {
+		case TRANSCODER_A:
+			return HDCP_TRANSA_REP_PRESENT |
+			       HDCP_TRANSA_SHA1_M0;
+		case TRANSCODER_B:
+			return HDCP_TRANSB_REP_PRESENT |
+			       HDCP_TRANSB_SHA1_M0;
+		case TRANSCODER_C:
+			return HDCP_TRANSC_REP_PRESENT |
+			       HDCP_TRANSC_SHA1_M0;
+		case TRANSCODER_D:
+			return HDCP_TRANSD_REP_PRESENT |
+			       HDCP_TRANSD_SHA1_M0;
+		default:
+			DRM_ERROR("Unknown transcoder %d\n", cpu_transcoder);
+			return -EINVAL;
+		}
+	}
+
 	switch (port) {
 	case PORT_A:
 		return HDCP_DDIA_REP_PRESENT | HDCP_DDIA_SHA1_M0;
@@ -268,18 +287,20 @@ u32 intel_hdcp_get_repeater_ctl(struct intel_digital_port *intel_dig_port)
 	case PORT_E:
 		return HDCP_DDIE_REP_PRESENT | HDCP_DDIE_SHA1_M0;
 	default:
-		break;
+		DRM_ERROR("Unknown port %d\n", port);
+		return -EINVAL;
 	}
-	DRM_ERROR("Unknown port %d\n", port);
-	return -EINVAL;
 }
 
 static
-int intel_hdcp_validate_v_prime(struct intel_digital_port *intel_dig_port,
+int intel_hdcp_validate_v_prime(struct intel_connector *connector,
 				const struct intel_hdcp_shim *shim,
 				u8 *ksv_fifo, u8 num_downstream, u8 *bstatus)
 {
+	struct intel_digital_port *intel_dig_port = conn_to_dig_port(connector);
 	struct drm_i915_private *dev_priv;
+	enum transcoder cpu_transcoder = connector->hdcp.cpu_transcoder;
+	enum port port = intel_dig_port->base.port;
 	u32 vprime, sha_text, sha_leftovers, rep_ctl;
 	int ret, i, j, sha_idx;
 
@@ -306,7 +327,7 @@ int intel_hdcp_validate_v_prime(struct intel_digital_port *intel_dig_port,
 	sha_idx = 0;
 	sha_text = 0;
 	sha_leftovers = 0;
-	rep_ctl = intel_hdcp_get_repeater_ctl(intel_dig_port);
+	rep_ctl = intel_hdcp_get_repeater_ctl(dev_priv, cpu_transcoder, port);
 	I915_WRITE(HDCP_REP_CTL, rep_ctl | HDCP_SHA1_TEXT_32);
 	for (i = 0; i < num_downstream; i++) {
 		unsigned int sha_empty;
@@ -548,7 +569,7 @@ int intel_hdcp_auth_downstream(struct intel_connector *connector)
 	 * V prime atleast twice.
 	 */
 	for (i = 0; i < tries; i++) {
-		ret = intel_hdcp_validate_v_prime(intel_dig_port, shim,
+		ret = intel_hdcp_validate_v_prime(connector, shim,
 						  ksv_fifo, num_downstream,
 						  bstatus);
 		if (!ret)
@@ -576,6 +597,7 @@ static int intel_hdcp_auth(struct intel_connector *connector)
 	struct drm_device *dev = connector->base.dev;
 	const struct intel_hdcp_shim *shim = hdcp->shim;
 	struct drm_i915_private *dev_priv;
+	enum transcoder cpu_transcoder = connector->hdcp.cpu_transcoder;
 	enum port port;
 	unsigned long r0_prime_gen_start;
 	int ret, i, tries = 2;
@@ -615,18 +637,21 @@ static int intel_hdcp_auth(struct intel_connector *connector)
 
 	/* Initialize An with 2 random values and acquire it */
 	for (i = 0; i < 2; i++)
-		I915_WRITE(PORT_HDCP_ANINIT(port), get_random_u32());
-	I915_WRITE(PORT_HDCP_CONF(port), HDCP_CONF_CAPTURE_AN);
+		I915_WRITE(HDCP_ANINIT(dev_priv, cpu_transcoder, port),
+			   get_random_u32());
+	I915_WRITE(HDCP_CONF(dev_priv, cpu_transcoder, port),
+		   HDCP_CONF_CAPTURE_AN);
 
 	/* Wait for An to be acquired */
-	if (intel_de_wait_for_set(dev_priv, PORT_HDCP_STATUS(port),
+	if (intel_de_wait_for_set(dev_priv,
+				  HDCP_STATUS(dev_priv, cpu_transcoder, port),
 				  HDCP_STATUS_AN_READY, 1)) {
 		DRM_ERROR("Timed out waiting for An\n");
 		return -ETIMEDOUT;
 	}
 
-	an.reg[0] = I915_READ(PORT_HDCP_ANLO(port));
-	an.reg[1] = I915_READ(PORT_HDCP_ANHI(port));
+	an.reg[0] = I915_READ(HDCP_ANLO(dev_priv, cpu_transcoder, port));
+	an.reg[1] = I915_READ(HDCP_ANHI(dev_priv, cpu_transcoder, port));
 	ret = shim->write_an_aksv(intel_dig_port, an.shim);
 	if (ret)
 		return ret;
@@ -644,24 +669,26 @@ static int intel_hdcp_auth(struct intel_connector *connector)
 		return -EPERM;
 	}
 
-	I915_WRITE(PORT_HDCP_BKSVLO(port), bksv.reg[0]);
-	I915_WRITE(PORT_HDCP_BKSVHI(port), bksv.reg[1]);
+	I915_WRITE(HDCP_BKSVLO(dev_priv, cpu_transcoder, port), bksv.reg[0]);
+	I915_WRITE(HDCP_BKSVHI(dev_priv, cpu_transcoder, port), bksv.reg[1]);
 
 	ret = shim->repeater_present(intel_dig_port, &repeater_present);
 	if (ret)
 		return ret;
 	if (repeater_present)
 		I915_WRITE(HDCP_REP_CTL,
-			   intel_hdcp_get_repeater_ctl(intel_dig_port));
+			   intel_hdcp_get_repeater_ctl(dev_priv, cpu_transcoder,
+						       port));
 
 	ret = shim->toggle_signalling(intel_dig_port, true);
 	if (ret)
 		return ret;
 
-	I915_WRITE(PORT_HDCP_CONF(port), HDCP_CONF_AUTH_AND_ENC);
+	I915_WRITE(HDCP_CONF(dev_priv, cpu_transcoder, port),
+		   HDCP_CONF_AUTH_AND_ENC);
 
 	/* Wait for R0 ready */
-	if (wait_for(I915_READ(PORT_HDCP_STATUS(port)) &
+	if (wait_for(I915_READ(HDCP_STATUS(dev_priv, cpu_transcoder, port)) &
 		     (HDCP_STATUS_R0_READY | HDCP_STATUS_ENC), 1)) {
 		DRM_ERROR("Timed out waiting for R0 ready\n");
 		return -ETIMEDOUT;
@@ -689,22 +716,25 @@ static int intel_hdcp_auth(struct intel_connector *connector)
 		ret = shim->read_ri_prime(intel_dig_port, ri.shim);
 		if (ret)
 			return ret;
-		I915_WRITE(PORT_HDCP_RPRIME(port), ri.reg);
+		I915_WRITE(HDCP_RPRIME(dev_priv, cpu_transcoder, port), ri.reg);
 
 		/* Wait for Ri prime match */
-		if (!wait_for(I915_READ(PORT_HDCP_STATUS(port)) &
+		if (!wait_for(I915_READ(HDCP_STATUS(dev_priv, cpu_transcoder,
+						    port)) &
 		    (HDCP_STATUS_RI_MATCH | HDCP_STATUS_ENC), 1))
 			break;
 	}
 
 	if (i == tries) {
 		DRM_DEBUG_KMS("Timed out waiting for Ri prime match (%x)\n",
-			      I915_READ(PORT_HDCP_STATUS(port)));
+			      I915_READ(HDCP_STATUS(dev_priv, cpu_transcoder,
+						    port)));
 		return -ETIMEDOUT;
 	}
 
 	/* Wait for encryption confirmation */
-	if (intel_de_wait_for_set(dev_priv, PORT_HDCP_STATUS(port),
+	if (intel_de_wait_for_set(dev_priv,
+				  HDCP_STATUS(dev_priv, cpu_transcoder, port),
 				  HDCP_STATUS_ENC,
 				  ENCRYPT_STATUS_CHANGE_TIMEOUT_MS)) {
 		DRM_ERROR("Timed out waiting for encryption\n");
@@ -729,15 +759,17 @@ static int _intel_hdcp_disable(struct intel_connector *connector)
 	struct drm_i915_private *dev_priv = connector->base.dev->dev_private;
 	struct intel_digital_port *intel_dig_port = conn_to_dig_port(connector);
 	enum port port = intel_dig_port->base.port;
+	enum transcoder cpu_transcoder = hdcp->cpu_transcoder;
 	int ret;
 
 	DRM_DEBUG_KMS("[%s:%d] HDCP is being disabled...\n",
 		      connector->base.name, connector->base.base.id);
 
 	hdcp->hdcp_encrypted = false;
-	I915_WRITE(PORT_HDCP_CONF(port), 0);
-	if (intel_de_wait_for_clear(dev_priv, PORT_HDCP_STATUS(port), ~0,
-				    ENCRYPT_STATUS_CHANGE_TIMEOUT_MS)) {
+	I915_WRITE(HDCP_CONF(dev_priv, cpu_transcoder, port), 0);
+	if (intel_de_wait_for_clear(dev_priv,
+				    HDCP_STATUS(dev_priv, cpu_transcoder, port),
+				    ~0, ENCRYPT_STATUS_CHANGE_TIMEOUT_MS)) {
 		DRM_ERROR("Failed to disable HDCP, timeout clearing status\n");
 		return -ETIMEDOUT;
 	}
@@ -808,9 +840,11 @@ static int intel_hdcp_check_link(struct intel_connector *connector)
 	struct drm_i915_private *dev_priv = connector->base.dev->dev_private;
 	struct intel_digital_port *intel_dig_port = conn_to_dig_port(connector);
 	enum port port = intel_dig_port->base.port;
+	enum transcoder cpu_transcoder;
 	int ret = 0;
 
 	mutex_lock(&hdcp->mutex);
+	cpu_transcoder = hdcp->cpu_transcoder;
 
 	/* Check_link valid only when HDCP1.4 is enabled */
 	if (hdcp->value != DRM_MODE_CONTENT_PROTECTION_ENABLED ||
@@ -819,10 +853,11 @@ static int intel_hdcp_check_link(struct intel_connector *connector)
 		goto out;
 	}
 
-	if (WARN_ON(!intel_hdcp_in_use(connector))) {
+	if (WARN_ON(!intel_hdcp_in_use(dev_priv, cpu_transcoder, port))) {
 		DRM_ERROR("%s:%d HDCP link stopped encryption,%x\n",
 			  connector->base.name, connector->base.base.id,
-			  I915_READ(PORT_HDCP_STATUS(port)));
+			  I915_READ(HDCP_STATUS(dev_priv, cpu_transcoder,
+						port)));
 		ret = -ENXIO;
 		hdcp->value = DRM_MODE_CONTENT_PROTECTION_DESIRED;
 		schedule_work(&hdcp->prop_work);
@@ -887,7 +922,7 @@ static void intel_hdcp_prop_work(struct work_struct *work)
 bool is_hdcp_supported(struct drm_i915_private *dev_priv, enum port port)
 {
 	/* PORT E doesn't have HDCP, and PORT F is disabled */
-	return INTEL_GEN(dev_priv) >= 9 && port < PORT_E;
+	return INTEL_INFO(dev_priv)->display.has_hdcp && port < PORT_E;
 }
 
 static int
@@ -1493,10 +1528,11 @@ static int hdcp2_enable_encryption(struct intel_connector *connector)
 	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
 	struct intel_hdcp *hdcp = &connector->hdcp;
 	enum port port = connector->encoder->port;
+	enum transcoder cpu_transcoder = hdcp->cpu_transcoder;
 	int ret;
 
-	WARN_ON(I915_READ(HDCP2_STATUS_DDI(port)) & LINK_ENCRYPTION_STATUS);
-
+	WARN_ON(I915_READ(HDCP2_STATUS(dev_priv, cpu_transcoder, port)) &
+		LINK_ENCRYPTION_STATUS);
 	if (hdcp->shim->toggle_signalling) {
 		ret = hdcp->shim->toggle_signalling(intel_dig_port, true);
 		if (ret) {
@@ -1506,14 +1542,18 @@ static int hdcp2_enable_encryption(struct intel_connector *connector)
 		}
 	}
 
-	if (I915_READ(HDCP2_STATUS_DDI(port)) & LINK_AUTH_STATUS) {
+	if (I915_READ(HDCP2_STATUS(dev_priv, cpu_transcoder, port)) &
+	    LINK_AUTH_STATUS) {
 		/* Link is Authenticated. Now set for Encryption */
-		I915_WRITE(HDCP2_CTL_DDI(port),
-			   I915_READ(HDCP2_CTL_DDI(port)) |
+		I915_WRITE(HDCP2_CTL(dev_priv, cpu_transcoder, port),
+			   I915_READ(HDCP2_CTL(dev_priv, cpu_transcoder,
+					       port)) |
 			   CTL_LINK_ENCRYPTION_REQ);
 	}
 
-	ret = intel_de_wait_for_set(dev_priv, HDCP2_STATUS_DDI(port),
+	ret = intel_de_wait_for_set(dev_priv,
+				    HDCP2_STATUS(dev_priv, cpu_transcoder,
+						 port),
 				    LINK_ENCRYPTION_STATUS,
 				    ENCRYPT_STATUS_CHANGE_TIMEOUT_MS);
 
@@ -1526,14 +1566,19 @@ static int hdcp2_disable_encryption(struct intel_connector *connector)
 	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
 	struct intel_hdcp *hdcp = &connector->hdcp;
 	enum port port = connector->encoder->port;
+	enum transcoder cpu_transcoder = hdcp->cpu_transcoder;
 	int ret;
 
-	WARN_ON(!(I915_READ(HDCP2_STATUS_DDI(port)) & LINK_ENCRYPTION_STATUS));
+	WARN_ON(!(I915_READ(HDCP2_STATUS(dev_priv, cpu_transcoder, port)) &
+			    LINK_ENCRYPTION_STATUS));
 
-	I915_WRITE(HDCP2_CTL_DDI(port),
-		   I915_READ(HDCP2_CTL_DDI(port)) & ~CTL_LINK_ENCRYPTION_REQ);
+	I915_WRITE(HDCP2_CTL(dev_priv, cpu_transcoder, port),
+		   I915_READ(HDCP2_CTL(dev_priv, cpu_transcoder, port)) &
+		   ~CTL_LINK_ENCRYPTION_REQ);
 
-	ret = intel_de_wait_for_clear(dev_priv, HDCP2_STATUS_DDI(port),
+	ret = intel_de_wait_for_clear(dev_priv,
+				      HDCP2_STATUS(dev_priv, cpu_transcoder,
+						   port),
 				      LINK_ENCRYPTION_STATUS,
 				      ENCRYPT_STATUS_CHANGE_TIMEOUT_MS);
 	if (ret == -ETIMEDOUT)
@@ -1632,9 +1677,11 @@ static int intel_hdcp2_check_link(struct intel_connector *connector)
 	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
 	struct intel_hdcp *hdcp = &connector->hdcp;
 	enum port port = connector->encoder->port;
+	enum transcoder cpu_transcoder;
 	int ret = 0;
 
 	mutex_lock(&hdcp->mutex);
+	cpu_transcoder = hdcp->cpu_transcoder;
 
 	/* hdcp2_check_link is expected only when HDCP2.2 is Enabled */
 	if (hdcp->value != DRM_MODE_CONTENT_PROTECTION_ENABLED ||
@@ -1643,9 +1690,10 @@ static int intel_hdcp2_check_link(struct intel_connector *connector)
 		goto out;
 	}
 
-	if (WARN_ON(!intel_hdcp2_in_use(connector))) {
+	if (WARN_ON(!intel_hdcp2_in_use(dev_priv, cpu_transcoder, port))) {
 		DRM_ERROR("HDCP2.2 link stopped the encryption, %x\n",
-			  I915_READ(HDCP2_STATUS_DDI(port)));
+			  I915_READ(HDCP2_STATUS(dev_priv, cpu_transcoder,
+						 port)));
 		ret = -ENXIO;
 		hdcp->value = DRM_MODE_CONTENT_PROTECTION_DESIRED;
 		schedule_work(&hdcp->prop_work);
@@ -1749,13 +1797,71 @@ static const struct component_ops i915_hdcp_component_ops = {
 	.unbind = i915_hdcp_component_unbind,
 };
 
+static inline
+enum mei_fw_ddi intel_get_mei_fw_ddi_index(enum port port)
+{
+	switch (port) {
+	case PORT_A:
+		return MEI_DDI_A;
+	case PORT_B ... PORT_F:
+		return (enum mei_fw_ddi)port;
+	default:
+		return MEI_DDI_INVALID_PORT;
+	}
+}
+
+static inline
+enum mei_fw_tc intel_get_mei_fw_tc(enum transcoder cpu_transcoder)
+{
+	switch (cpu_transcoder) {
+	case TRANSCODER_A ... TRANSCODER_D:
+		return (enum mei_fw_tc)(cpu_transcoder | 0x10);
+	default: /* eDP, DSI TRANSCODERS are non HDCP capable */
+		return MEI_INVALID_TRANSCODER;
+	}
+}
+
+void intel_hdcp_transcoder_config(struct intel_connector *connector,
+				  enum transcoder cpu_transcoder)
+{
+	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
+	struct intel_hdcp *hdcp = &connector->hdcp;
+
+	if (!hdcp->shim)
+		return;
+
+	if (INTEL_GEN(dev_priv) >= 12) {
+		mutex_lock(&hdcp->mutex);
+		hdcp->cpu_transcoder = cpu_transcoder;
+		hdcp->port_data.fw_tc = intel_get_mei_fw_tc(cpu_transcoder);
+		mutex_unlock(&hdcp->mutex);
+	}
+}
+
 static inline int initialize_hdcp_port_data(struct intel_connector *connector,
 					    const struct intel_hdcp_shim *shim)
 {
+	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
 	struct intel_hdcp *hdcp = &connector->hdcp;
 	struct hdcp_port_data *data = &hdcp->port_data;
 
-	data->port = connector->encoder->port;
+	if (INTEL_GEN(dev_priv) < 12)
+		data->fw_ddi =
+			intel_get_mei_fw_ddi_index(connector->encoder->port);
+	else
+		/*
+		 * As per ME FW API expectation, for GEN 12+, fw_ddi is filled
+		 * with zero(INVALID PORT index).
+		 */
+		data->fw_ddi = MEI_DDI_INVALID_PORT;
+
+	/*
+	 * As associated transcoder is set and modified at modeset, here fw_tc
+	 * is initialized to zero (invalid transcoder index). This will be
+	 * retained for <Gen12 forever.
+	 */
+	data->fw_tc = MEI_INVALID_TRANSCODER;
+
 	data->port_type = (u8)HDCP_PORT_TYPE_INTEGRATED;
 	data->protocol = (u8)shim->protocol;
 
diff --git a/drivers/gpu/drm/i915/display/intel_hdcp.h b/drivers/gpu/drm/i915/display/intel_hdcp.h
index 13555b054930..41c1053d9e38 100644
--- a/drivers/gpu/drm/i915/display/intel_hdcp.h
+++ b/drivers/gpu/drm/i915/display/intel_hdcp.h
@@ -15,10 +15,14 @@ struct drm_connector_state;
 struct drm_i915_private;
 struct intel_connector;
 struct intel_hdcp_shim;
+enum port;
+enum transcoder;
 
 void intel_hdcp_atomic_check(struct drm_connector *connector,
 			     struct drm_connector_state *old_state,
 			     struct drm_connector_state *new_state);
+void intel_hdcp_transcoder_config(struct intel_connector *connector,
+				  enum transcoder cpu_transcoder);
 int intel_hdcp_init(struct intel_connector *connector,
 		    const struct intel_hdcp_shim *hdcp_shim);
 int intel_hdcp_enable(struct intel_connector *connector, u8 content_type);
diff --git a/drivers/gpu/drm/i915/display/intel_hdmi.c b/drivers/gpu/drm/i915/display/intel_hdmi.c
index e02f0faecf02..f6f5312205c4 100644
--- a/drivers/gpu/drm/i915/display/intel_hdmi.c
+++ b/drivers/gpu/drm/i915/display/intel_hdmi.c
@@ -189,13 +189,19 @@ hsw_dip_data_reg(struct drm_i915_private *dev_priv,
 	}
 }
 
-static int hsw_dip_data_size(unsigned int type)
+static int hsw_dip_data_size(struct drm_i915_private *dev_priv,
+			     unsigned int type)
 {
 	switch (type) {
 	case DP_SDP_VSC:
 		return VIDEO_DIP_VSC_DATA_SIZE;
 	case DP_SDP_PPS:
 		return VIDEO_DIP_PPS_DATA_SIZE;
+	case HDMI_PACKET_TYPE_GAMUT_METADATA:
+		if (INTEL_GEN(dev_priv) >= 11)
+			return VIDEO_DIP_GMP_DATA_SIZE;
+		else
+			return VIDEO_DIP_DATA_SIZE;
 	default:
 		return VIDEO_DIP_DATA_SIZE;
 	}
@@ -514,7 +520,9 @@ static void hsw_write_infoframe(struct intel_encoder *encoder,
 	int i;
 	u32 val = I915_READ(ctl_reg);
 
-	data_size = hsw_dip_data_size(type);
+	data_size = hsw_dip_data_size(dev_priv, type);
+
+	WARN_ON(len > data_size);
 
 	val &= ~hsw_infoframe_enable(type);
 	I915_WRITE(ctl_reg, val);
@@ -724,11 +732,20 @@ intel_hdmi_compute_avi_infoframe(struct intel_encoder *encoder,
 
 	drm_hdmi_avi_infoframe_colorspace(frame, conn_state);
 
-	drm_hdmi_avi_infoframe_quant_range(frame, connector,
-					   adjusted_mode,
-					   crtc_state->limited_color_range ?
-					   HDMI_QUANTIZATION_RANGE_LIMITED :
-					   HDMI_QUANTIZATION_RANGE_FULL);
+	/* nonsense combination */
+	WARN_ON(crtc_state->limited_color_range &&
+		crtc_state->output_format != INTEL_OUTPUT_FORMAT_RGB);
+
+	if (crtc_state->output_format == INTEL_OUTPUT_FORMAT_RGB) {
+		drm_hdmi_avi_infoframe_quant_range(frame, connector,
+						   adjusted_mode,
+						   crtc_state->limited_color_range ?
+						   HDMI_QUANTIZATION_RANGE_LIMITED :
+						   HDMI_QUANTIZATION_RANGE_FULL);
+	} else {
+		frame->quantization_range = HDMI_QUANTIZATION_RANGE_DEFAULT;
+		frame->ycc_quantization_range = HDMI_YCC_QUANTIZATION_RANGE_LIMITED;
+	}
 
 	drm_hdmi_avi_infoframe_content_type(frame, conn_state);
 
@@ -1491,7 +1508,10 @@ bool intel_hdmi_hdcp_check_link(struct intel_digital_port *intel_dig_port)
 {
 	struct drm_i915_private *dev_priv =
 		intel_dig_port->base.base.dev->dev_private;
+	struct intel_connector *connector =
+		intel_dig_port->hdmi.attached_connector;
 	enum port port = intel_dig_port->base.port;
+	enum transcoder cpu_transcoder = connector->hdcp.cpu_transcoder;
 	int ret;
 	union {
 		u32 reg;
@@ -1502,39 +1522,30 @@ bool intel_hdmi_hdcp_check_link(struct intel_digital_port *intel_dig_port)
 	if (ret)
 		return false;
 
-	I915_WRITE(PORT_HDCP_RPRIME(port), ri.reg);
+	I915_WRITE(HDCP_RPRIME(dev_priv, cpu_transcoder, port), ri.reg);
 
 	/* Wait for Ri prime match */
-	if (wait_for(I915_READ(PORT_HDCP_STATUS(port)) &
+	if (wait_for(I915_READ(HDCP_STATUS(dev_priv, cpu_transcoder, port)) &
 		     (HDCP_STATUS_RI_MATCH | HDCP_STATUS_ENC), 1)) {
 		DRM_ERROR("Ri' mismatch detected, link check failed (%x)\n",
-			  I915_READ(PORT_HDCP_STATUS(port)));
+			  I915_READ(HDCP_STATUS(dev_priv, cpu_transcoder,
+						port)));
 		return false;
 	}
 	return true;
 }
 
-struct hdcp2_hdmi_msg_data {
+struct hdcp2_hdmi_msg_timeout {
 	u8 msg_id;
-	u32 timeout;
-	u32 timeout2;
+	u16 timeout;
 };
 
-static const struct hdcp2_hdmi_msg_data hdcp2_msg_data[] = {
-	{ HDCP_2_2_AKE_INIT, 0, 0 },
-	{ HDCP_2_2_AKE_SEND_CERT, HDCP_2_2_CERT_TIMEOUT_MS, 0 },
-	{ HDCP_2_2_AKE_NO_STORED_KM, 0, 0 },
-	{ HDCP_2_2_AKE_STORED_KM, 0, 0 },
-	{ HDCP_2_2_AKE_SEND_HPRIME, HDCP_2_2_HPRIME_PAIRED_TIMEOUT_MS,
-	  HDCP_2_2_HPRIME_NO_PAIRED_TIMEOUT_MS },
-	{ HDCP_2_2_AKE_SEND_PAIRING_INFO, HDCP_2_2_PAIRING_TIMEOUT_MS, 0 },
-	{ HDCP_2_2_LC_INIT, 0, 0 },
-	{ HDCP_2_2_LC_SEND_LPRIME, HDCP_2_2_HDMI_LPRIME_TIMEOUT_MS, 0 },
-	{ HDCP_2_2_SKE_SEND_EKS, 0, 0 },
-	{ HDCP_2_2_REP_SEND_RECVID_LIST, HDCP_2_2_RECVID_LIST_TIMEOUT_MS, 0 },
-	{ HDCP_2_2_REP_SEND_ACK, 0, 0 },
-	{ HDCP_2_2_REP_STREAM_MANAGE, 0, 0 },
-	{ HDCP_2_2_REP_STREAM_READY, HDCP_2_2_STREAM_READY_TIMEOUT_MS, 0 },
+static const struct hdcp2_hdmi_msg_timeout hdcp2_msg_timeout[] = {
+	{ HDCP_2_2_AKE_SEND_CERT, HDCP_2_2_CERT_TIMEOUT_MS, },
+	{ HDCP_2_2_AKE_SEND_PAIRING_INFO, HDCP_2_2_PAIRING_TIMEOUT_MS, },
+	{ HDCP_2_2_LC_SEND_LPRIME, HDCP_2_2_HDMI_LPRIME_TIMEOUT_MS, },
+	{ HDCP_2_2_REP_SEND_RECVID_LIST, HDCP_2_2_RECVID_LIST_TIMEOUT_MS, },
+	{ HDCP_2_2_REP_STREAM_READY, HDCP_2_2_STREAM_READY_TIMEOUT_MS, },
 };
 
 static
@@ -1551,12 +1562,17 @@ static int get_hdcp2_msg_timeout(u8 msg_id, bool is_paired)
 {
 	int i;
 
-	for (i = 0; i < ARRAY_SIZE(hdcp2_msg_data); i++)
-		if (hdcp2_msg_data[i].msg_id == msg_id &&
-		    (msg_id != HDCP_2_2_AKE_SEND_HPRIME || is_paired))
-			return hdcp2_msg_data[i].timeout;
-		else if (hdcp2_msg_data[i].msg_id == msg_id)
-			return hdcp2_msg_data[i].timeout2;
+	if (msg_id == HDCP_2_2_AKE_SEND_HPRIME) {
+		if (is_paired)
+			return HDCP_2_2_HPRIME_PAIRED_TIMEOUT_MS;
+		else
+			return HDCP_2_2_HPRIME_NO_PAIRED_TIMEOUT_MS;
+	}
+
+	for (i = 0; i < ARRAY_SIZE(hdcp2_msg_timeout); i++) {
+		if (hdcp2_msg_timeout[i].msg_id == msg_id)
+			return hdcp2_msg_timeout[i].timeout;
+	}
 
 	return -EINVAL;
 }
@@ -2184,8 +2200,10 @@ intel_hdmi_mode_valid(struct drm_connector *connector,
 			status = hdmi_port_clock_valid(hdmi, clock * 5 / 4,
 						       true, force_dvi);
 	}
+	if (status != MODE_OK)
+		return status;
 
-	return status;
+	return intel_mode_valid_max_plane_size(dev_priv, mode);
 }
 
 static bool hdmi_deep_color_possible(const struct intel_crtc_state *crtc_state,
@@ -2261,9 +2279,7 @@ static bool hdmi_deep_color_possible(const struct intel_crtc_state *crtc_state,
 
 static bool
 intel_hdmi_ycbcr420_config(struct drm_connector *connector,
-			   struct intel_crtc_state *config,
-			   int *clock_12bpc, int *clock_10bpc,
-			   int *clock_8bpc)
+			   struct intel_crtc_state *config)
 {
 	struct intel_crtc *intel_crtc = to_intel_crtc(config->base.crtc);
 
@@ -2272,11 +2288,6 @@ intel_hdmi_ycbcr420_config(struct drm_connector *connector,
 		return false;
 	}
 
-	/* YCBCR420 TMDS rate requirement is half the pixel clock */
-	config->port_clock /= 2;
-	*clock_12bpc /= 2;
-	*clock_10bpc /= 2;
-	*clock_8bpc /= 2;
 	config->output_format = INTEL_OUTPUT_FORMAT_YCBCR420;
 
 	/* YCBCR 420 output conversion needs a scaler */
@@ -2291,6 +2302,104 @@ intel_hdmi_ycbcr420_config(struct drm_connector *connector,
 	return true;
 }
 
+static int intel_hdmi_port_clock(int clock, int bpc)
+{
+	/*
+	 * Need to adjust the port link by:
+	 *  1.5x for 12bpc
+	 *  1.25x for 10bpc
+	 */
+	return clock * bpc / 8;
+}
+
+static int intel_hdmi_compute_bpc(struct intel_encoder *encoder,
+				  struct intel_crtc_state *crtc_state,
+				  int clock, bool force_dvi)
+{
+	struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(&encoder->base);
+	int bpc;
+
+	for (bpc = 12; bpc >= 10; bpc -= 2) {
+		if (hdmi_deep_color_possible(crtc_state, bpc) &&
+		    hdmi_port_clock_valid(intel_hdmi,
+					  intel_hdmi_port_clock(clock, bpc),
+					  true, force_dvi) == MODE_OK)
+			return bpc;
+	}
+
+	return 8;
+}
+
+static int intel_hdmi_compute_clock(struct intel_encoder *encoder,
+				    struct intel_crtc_state *crtc_state,
+				    bool force_dvi)
+{
+	struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(&encoder->base);
+	const struct drm_display_mode *adjusted_mode =
+		&crtc_state->base.adjusted_mode;
+	int bpc, clock = adjusted_mode->crtc_clock;
+
+	if (adjusted_mode->flags & DRM_MODE_FLAG_DBLCLK)
+		clock *= 2;
+
+	/* YCBCR420 TMDS rate requirement is half the pixel clock */
+	if (crtc_state->output_format == INTEL_OUTPUT_FORMAT_YCBCR420)
+		clock /= 2;
+
+	bpc = intel_hdmi_compute_bpc(encoder, crtc_state,
+				     clock, force_dvi);
+
+	crtc_state->port_clock = intel_hdmi_port_clock(clock, bpc);
+
+	/*
+	 * pipe_bpp could already be below 8bpc due to
+	 * FDI bandwidth constraints. We shouldn't bump it
+	 * back up to 8bpc in that case.
+	 */
+	if (crtc_state->pipe_bpp > bpc * 3)
+		crtc_state->pipe_bpp = bpc * 3;
+
+	DRM_DEBUG_KMS("picking %d bpc for HDMI output (pipe bpp: %d)\n",
+		      bpc, crtc_state->pipe_bpp);
+
+	if (hdmi_port_clock_valid(intel_hdmi, crtc_state->port_clock,
+				  false, force_dvi) != MODE_OK) {
+		DRM_DEBUG_KMS("unsupported HDMI clock (%d kHz), rejecting mode\n",
+			      crtc_state->port_clock);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static bool intel_hdmi_limited_color_range(const struct intel_crtc_state *crtc_state,
+					   const struct drm_connector_state *conn_state)
+{
+	const struct intel_digital_connector_state *intel_conn_state =
+		to_intel_digital_connector_state(conn_state);
+	const struct drm_display_mode *adjusted_mode =
+		&crtc_state->base.adjusted_mode;
+
+	/*
+	 * Our YCbCr output is always limited range.
+	 * crtc_state->limited_color_range only applies to RGB,
+	 * and it must never be set for YCbCr or we risk setting
+	 * some conflicting bits in PIPECONF which will mess up
+	 * the colors on the monitor.
+	 */
+	if (crtc_state->output_format != INTEL_OUTPUT_FORMAT_RGB)
+		return false;
+
+	if (intel_conn_state->broadcast_rgb == INTEL_BROADCAST_RGB_AUTO) {
+		/* See CEA-861-E - 5.1 Default Encoding Parameters */
+		return crtc_state->has_hdmi_sink &&
+			drm_default_rgb_quant_range(adjusted_mode) ==
+			HDMI_QUANTIZATION_RANGE_LIMITED;
+	} else {
+		return intel_conn_state->broadcast_rgb == INTEL_BROADCAST_RGB_LIMITED;
+	}
+}
+
 int intel_hdmi_compute_config(struct intel_encoder *encoder,
 			      struct intel_crtc_state *pipe_config,
 			      struct drm_connector_state *conn_state)
@@ -2302,11 +2411,8 @@ int intel_hdmi_compute_config(struct intel_encoder *encoder,
 	struct drm_scdc *scdc = &connector->display_info.hdmi.scdc;
 	struct intel_digital_connector_state *intel_conn_state =
 		to_intel_digital_connector_state(conn_state);
-	int clock_8bpc = pipe_config->base.adjusted_mode.crtc_clock;
-	int clock_10bpc = clock_8bpc * 5 / 4;
-	int clock_12bpc = clock_8bpc * 3 / 2;
-	int desired_bpp;
 	bool force_dvi = intel_conn_state->force_audio == HDMI_AUDIO_OFF_DVI;
+	int ret;
 
 	if (adjusted_mode->flags & DRM_MODE_FLAG_DBLSCAN)
 		return -EINVAL;
@@ -2317,33 +2423,19 @@ int intel_hdmi_compute_config(struct intel_encoder *encoder,
 	if (pipe_config->has_hdmi_sink)
 		pipe_config->has_infoframe = true;
 
-	if (intel_conn_state->broadcast_rgb == INTEL_BROADCAST_RGB_AUTO) {
-		/* See CEA-861-E - 5.1 Default Encoding Parameters */
-		pipe_config->limited_color_range =
-			pipe_config->has_hdmi_sink &&
-			drm_default_rgb_quant_range(adjusted_mode) ==
-			HDMI_QUANTIZATION_RANGE_LIMITED;
-	} else {
-		pipe_config->limited_color_range =
-			intel_conn_state->broadcast_rgb == INTEL_BROADCAST_RGB_LIMITED;
-	}
-
-	if (adjusted_mode->flags & DRM_MODE_FLAG_DBLCLK) {
+	if (adjusted_mode->flags & DRM_MODE_FLAG_DBLCLK)
 		pipe_config->pixel_multiplier = 2;
-		clock_8bpc *= 2;
-		clock_10bpc *= 2;
-		clock_12bpc *= 2;
-	}
 
 	if (drm_mode_is_420_only(&connector->display_info, adjusted_mode)) {
-		if (!intel_hdmi_ycbcr420_config(connector, pipe_config,
-						&clock_12bpc, &clock_10bpc,
-						&clock_8bpc)) {
+		if (!intel_hdmi_ycbcr420_config(connector, pipe_config)) {
 			DRM_ERROR("Can't support YCBCR420 output\n");
 			return -EINVAL;
 		}
 	}
 
+	pipe_config->limited_color_range =
+		intel_hdmi_limited_color_range(pipe_config, conn_state);
+
 	if (HAS_PCH_SPLIT(dev_priv) && !HAS_DDI(dev_priv))
 		pipe_config->has_pch_encoder = true;
 
@@ -2355,43 +2447,9 @@ int intel_hdmi_compute_config(struct intel_encoder *encoder,
 				intel_conn_state->force_audio == HDMI_AUDIO_ON;
 	}
 
-	/*
-	 * Note that g4x/vlv don't support 12bpc hdmi outputs. We also need
-	 * to check that the higher clock still fits within limits.
-	 */
-	if (hdmi_deep_color_possible(pipe_config, 12) &&
-	    hdmi_port_clock_valid(intel_hdmi, clock_12bpc,
-				  true, force_dvi) == MODE_OK) {
-		DRM_DEBUG_KMS("picking bpc to 12 for HDMI output\n");
-		desired_bpp = 12*3;
-
-		/* Need to adjust the port link by 1.5x for 12bpc. */
-		pipe_config->port_clock = clock_12bpc;
-	} else if (hdmi_deep_color_possible(pipe_config, 10) &&
-		   hdmi_port_clock_valid(intel_hdmi, clock_10bpc,
-					 true, force_dvi) == MODE_OK) {
-		DRM_DEBUG_KMS("picking bpc to 10 for HDMI output\n");
-		desired_bpp = 10 * 3;
-
-		/* Need to adjust the port link by 1.25x for 10bpc. */
-		pipe_config->port_clock = clock_10bpc;
-	} else {
-		DRM_DEBUG_KMS("picking bpc to 8 for HDMI output\n");
-		desired_bpp = 8*3;
-
-		pipe_config->port_clock = clock_8bpc;
-	}
-
-	if (!pipe_config->bw_constrained) {
-		DRM_DEBUG_KMS("forcing pipe bpp to %i for HDMI\n", desired_bpp);
-		pipe_config->pipe_bpp = desired_bpp;
-	}
-
-	if (hdmi_port_clock_valid(intel_hdmi, pipe_config->port_clock,
-				  false, force_dvi) != MODE_OK) {
-		DRM_DEBUG_KMS("unsupported HDMI clock, rejecting mode\n");
-		return -EINVAL;
-	}
+	ret = intel_hdmi_compute_clock(encoder, pipe_config, force_dvi);
+	if (ret)
+		return ret;
 
 	/* Set user selected PAR to incoming mode's member */
 	adjusted_mode->picture_aspect_ratio = conn_state->picture_aspect_ratio;
@@ -2431,6 +2489,9 @@ int intel_hdmi_compute_config(struct intel_encoder *encoder,
 		return -EINVAL;
 	}
 
+	intel_hdcp_transcoder_config(intel_hdmi->attached_connector,
+				     pipe_config->cpu_transcoder);
+
 	return 0;
 }
 
@@ -2565,6 +2626,12 @@ out:
 	if (status != connector_status_connected)
 		cec_notifier_phys_addr_invalidate(intel_hdmi->cec_notifier);
 
+	/*
+	 * Make sure the refs for power wells enabled during detect are
+	 * dropped to avoid a new detect cycle triggered by HPD polling.
+	 */
+	intel_display_power_flush_work(dev_priv);
+
 	return status;
 }
 
@@ -2751,8 +2818,9 @@ intel_hdmi_connector_register(struct drm_connector *connector)
 
 static void intel_hdmi_destroy(struct drm_connector *connector)
 {
-	if (intel_attached_hdmi(connector)->cec_notifier)
-		cec_notifier_put(intel_attached_hdmi(connector)->cec_notifier);
+	struct cec_notifier *n = intel_attached_hdmi(connector)->cec_notifier;
+
+	cec_notifier_conn_unregister(n);
 
 	intel_connector_destroy(connector);
 }
@@ -3001,7 +3069,7 @@ static u8 intel_hdmi_ddc_pin(struct drm_i915_private *dev_priv,
 
 	if (HAS_PCH_MCC(dev_priv))
 		ddc_pin = mcc_port_to_ddc_pin(dev_priv, port);
-	else if (HAS_PCH_TGP(dev_priv) || HAS_PCH_ICP(dev_priv))
+	else if (INTEL_PCH_TYPE(dev_priv) >= PCH_ICP)
 		ddc_pin = icl_port_to_ddc_pin(dev_priv, port);
 	else if (HAS_PCH_CNP(dev_priv))
 		ddc_pin = cnp_port_to_ddc_pin(dev_priv, port);
@@ -3067,13 +3135,15 @@ void intel_hdmi_init_connector(struct intel_digital_port *intel_dig_port,
 	struct drm_device *dev = intel_encoder->base.dev;
 	struct drm_i915_private *dev_priv = to_i915(dev);
 	enum port port = intel_encoder->port;
+	struct cec_connector_info conn_info;
 
-	DRM_DEBUG_KMS("Adding HDMI connector on port %c\n",
-		      port_name(port));
+	DRM_DEBUG_KMS("Adding HDMI connector on [ENCODER:%d:%s]\n",
+		      intel_encoder->base.base.id, intel_encoder->base.name);
 
 	if (WARN(intel_dig_port->max_lanes < 4,
-		 "Not enough lanes (%d) for HDMI on port %c\n",
-		 intel_dig_port->max_lanes, port_name(port)))
+		 "Not enough lanes (%d) for HDMI on [ENCODER:%d:%s]\n",
+		 intel_dig_port->max_lanes, intel_encoder->base.base.id,
+		 intel_encoder->base.name))
 		return;
 
 	drm_connector_init(dev, connector, &intel_hdmi_connector_funcs,
@@ -3119,8 +3189,11 @@ void intel_hdmi_init_connector(struct intel_digital_port *intel_dig_port,
 		I915_WRITE(PEG_BAND_GAP_DATA, (temp & ~0xf) | 0xd);
 	}
 
-	intel_hdmi->cec_notifier = cec_notifier_get_conn(dev->dev,
-							 port_identifier(port));
+	cec_fill_conn_info_from_drm(&conn_info, connector);
+
+	intel_hdmi->cec_notifier =
+		cec_notifier_conn_register(dev->dev, port_identifier(port),
+					   &conn_info);
 	if (!intel_hdmi->cec_notifier)
 		DRM_DEBUG_KMS("CEC notifier get failed\n");
 }
@@ -3210,11 +3283,11 @@ void intel_hdmi_init(struct drm_i915_private *dev_priv,
 	intel_encoder->port = port;
 	if (IS_CHERRYVIEW(dev_priv)) {
 		if (port == PORT_D)
-			intel_encoder->crtc_mask = 1 << 2;
+			intel_encoder->pipe_mask = BIT(PIPE_C);
 		else
-			intel_encoder->crtc_mask = (1 << 0) | (1 << 1);
+			intel_encoder->pipe_mask = BIT(PIPE_A) | BIT(PIPE_B);
 	} else {
-		intel_encoder->crtc_mask = (1 << 0) | (1 << 1) | (1 << 2);
+		intel_encoder->pipe_mask = ~0;
 	}
 	intel_encoder->cloneable = 1 << INTEL_OUTPUT_ANALOG;
 	/*
diff --git a/drivers/gpu/drm/i915/display/intel_hdmi.h b/drivers/gpu/drm/i915/display/intel_hdmi.h
index 106c2e0bc3c9..cf1ea5427639 100644
--- a/drivers/gpu/drm/i915/display/intel_hdmi.h
+++ b/drivers/gpu/drm/i915/display/intel_hdmi.h
@@ -23,6 +23,7 @@ struct intel_crtc_state;
 struct intel_hdmi;
 struct drm_connector_state;
 union hdmi_infoframe;
+enum port;
 
 void intel_hdmi_init(struct drm_i915_private *dev_priv, i915_reg_t hdmi_reg,
 		     enum port port);
diff --git a/drivers/gpu/drm/i915/display/intel_hotplug.c b/drivers/gpu/drm/i915/display/intel_hotplug.c
index 56be20f6f47e..fc29046d48ea 100644
--- a/drivers/gpu/drm/i915/display/intel_hotplug.c
+++ b/drivers/gpu/drm/i915/display/intel_hotplug.c
@@ -481,7 +481,8 @@ void intel_hpd_irq_handler(struct drm_i915_private *dev_priv,
 
 		long_hpd = long_mask & BIT(pin);
 
-		DRM_DEBUG_DRIVER("digital hpd port %c - %s\n", port_name(port),
+		DRM_DEBUG_DRIVER("digital hpd on [ENCODER:%d:%s] - %s\n",
+				 encoder->base.base.id, encoder->base.name,
 				 long_hpd ? "long" : "short");
 		queue_dig = true;
 
diff --git a/drivers/gpu/drm/i915/display/intel_hotplug.h b/drivers/gpu/drm/i915/display/intel_hotplug.h
index b0cd447b7fbc..087b5f57b321 100644
--- a/drivers/gpu/drm/i915/display/intel_hotplug.h
+++ b/drivers/gpu/drm/i915/display/intel_hotplug.h
@@ -13,6 +13,7 @@
 struct drm_i915_private;
 struct intel_connector;
 struct intel_encoder;
+enum port;
 
 void intel_hpd_poll_init(struct drm_i915_private *dev_priv);
 enum intel_hotplug_state intel_encoder_hotplug(struct intel_encoder *encoder,
diff --git a/drivers/gpu/drm/i915/display/intel_lpe_audio.c b/drivers/gpu/drm/i915/display/intel_lpe_audio.c
index b19800b58442..0b67f7887cd0 100644
--- a/drivers/gpu/drm/i915/display/intel_lpe_audio.c
+++ b/drivers/gpu/drm/i915/display/intel_lpe_audio.c
@@ -114,7 +114,7 @@ lpe_audio_platdev_create(struct drm_i915_private *dev_priv)
 	pinfo.size_data = sizeof(*pdata);
 	pinfo.dma_mask = DMA_BIT_MASK(32);
 
-	pdata->num_pipes = INTEL_INFO(dev_priv)->num_pipes;
+	pdata->num_pipes = INTEL_NUM_PIPES(dev_priv);
 	pdata->num_ports = IS_CHERRYVIEW(dev_priv) ? 3 : 2; /* B,C,D or B,C */
 	pdata->port[0].pipe = -1;
 	pdata->port[1].pipe = -1;
diff --git a/drivers/gpu/drm/i915/display/intel_lvds.c b/drivers/gpu/drm/i915/display/intel_lvds.c
index b7c459a8931c..b1bc78623647 100644
--- a/drivers/gpu/drm/i915/display/intel_lvds.c
+++ b/drivers/gpu/drm/i915/display/intel_lvds.c
@@ -232,7 +232,7 @@ static void intel_pre_enable_lvds(struct intel_encoder *encoder,
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
 	struct intel_crtc *crtc = to_intel_crtc(pipe_config->base.crtc);
 	const struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode;
-	int pipe = crtc->pipe;
+	enum pipe pipe = crtc->pipe;
 	u32 temp;
 
 	if (HAS_PCH_SPLIT(dev_priv)) {
@@ -899,12 +899,10 @@ void intel_lvds_init(struct drm_i915_private *dev_priv)
 	intel_encoder->power_domain = POWER_DOMAIN_PORT_OTHER;
 	intel_encoder->port = PORT_NONE;
 	intel_encoder->cloneable = 0;
-	if (HAS_PCH_SPLIT(dev_priv))
-		intel_encoder->crtc_mask = (1 << 0) | (1 << 1) | (1 << 2);
-	else if (IS_GEN(dev_priv, 4))
-		intel_encoder->crtc_mask = (1 << 0) | (1 << 1);
+	if (INTEL_GEN(dev_priv) < 4)
+		intel_encoder->pipe_mask = BIT(PIPE_B);
 	else
-		intel_encoder->crtc_mask = (1 << 1);
+		intel_encoder->pipe_mask = ~0;
 
 	drm_connector_helper_add(connector, &intel_lvds_connector_helper_funcs);
 	connector->display_info.subpixel_order = SubPixelHorizontalRGB;
diff --git a/drivers/gpu/drm/i915/display/intel_overlay.c b/drivers/gpu/drm/i915/display/intel_overlay.c
index 29edfc343716..848ce07a8ec2 100644
--- a/drivers/gpu/drm/i915/display/intel_overlay.c
+++ b/drivers/gpu/drm/i915/display/intel_overlay.c
@@ -30,6 +30,7 @@
 #include <drm/i915_drm.h>
 
 #include "gem/i915_gem_pm.h"
+#include "gt/intel_ring.h"
 
 #include "i915_drv.h"
 #include "i915_reg.h"
@@ -230,7 +231,7 @@ alloc_request(struct intel_overlay *overlay, void (*fn)(struct intel_overlay *))
 	if (IS_ERR(rq))
 		return rq;
 
-	err = i915_active_ref(&overlay->last_flip, rq->timeline, rq);
+	err = i915_active_add_request(&overlay->last_flip, rq);
 	if (err) {
 		i915_request_add(rq);
 		return ERR_PTR(err);
@@ -439,8 +440,6 @@ static int intel_overlay_release_old_vid(struct intel_overlay *overlay)
 	struct i915_request *rq;
 	u32 *cs;
 
-	lockdep_assert_held(&dev_priv->drm.struct_mutex);
-
 	/*
 	 * Only wait if there is actually an old frame to release to
 	 * guarantee forward progress.
@@ -751,7 +750,6 @@ static int intel_overlay_do_put_image(struct intel_overlay *overlay,
 	struct i915_vma *vma;
 	int ret, tmp_width;
 
-	lockdep_assert_held(&dev_priv->drm.struct_mutex);
 	WARN_ON(!drm_modeset_is_locked(&dev_priv->drm.mode_config.connection_mutex));
 
 	ret = intel_overlay_release_old_vid(overlay);
@@ -852,7 +850,6 @@ int intel_overlay_switch_off(struct intel_overlay *overlay)
 	struct drm_i915_private *dev_priv = overlay->i915;
 	int ret;
 
-	lockdep_assert_held(&dev_priv->drm.struct_mutex);
 	WARN_ON(!drm_modeset_is_locked(&dev_priv->drm.mode_config.connection_mutex));
 
 	ret = intel_overlay_recover_from_interrupt(overlay);
@@ -1068,11 +1065,7 @@ int intel_overlay_put_image_ioctl(struct drm_device *dev, void *data,
 
 	if (!(params->flags & I915_OVERLAY_ENABLE)) {
 		drm_modeset_lock_all(dev);
-		mutex_lock(&dev->struct_mutex);
-
 		ret = intel_overlay_switch_off(overlay);
-
-		mutex_unlock(&dev->struct_mutex);
 		drm_modeset_unlock_all(dev);
 
 		return ret;
@@ -1088,7 +1081,6 @@ int intel_overlay_put_image_ioctl(struct drm_device *dev, void *data,
 		return -ENOENT;
 
 	drm_modeset_lock_all(dev);
-	mutex_lock(&dev->struct_mutex);
 
 	if (i915_gem_object_is_tiled(new_bo)) {
 		DRM_DEBUG_KMS("buffer used for overlay image can not be tiled\n");
@@ -1152,14 +1144,12 @@ int intel_overlay_put_image_ioctl(struct drm_device *dev, void *data,
 	if (ret != 0)
 		goto out_unlock;
 
-	mutex_unlock(&dev->struct_mutex);
 	drm_modeset_unlock_all(dev);
 	i915_gem_object_put(new_bo);
 
 	return 0;
 
 out_unlock:
-	mutex_unlock(&dev->struct_mutex);
 	drm_modeset_unlock_all(dev);
 	i915_gem_object_put(new_bo);
 
@@ -1233,7 +1223,6 @@ int intel_overlay_attrs_ioctl(struct drm_device *dev, void *data,
 	}
 
 	drm_modeset_lock_all(dev);
-	mutex_lock(&dev->struct_mutex);
 
 	ret = -EINVAL;
 	if (!(attrs->flags & I915_OVERLAY_UPDATE_ATTRS)) {
@@ -1290,7 +1279,6 @@ int intel_overlay_attrs_ioctl(struct drm_device *dev, void *data,
 
 	ret = 0;
 out_unlock:
-	mutex_unlock(&dev->struct_mutex);
 	drm_modeset_unlock_all(dev);
 
 	return ret;
@@ -1303,15 +1291,11 @@ static int get_registers(struct intel_overlay *overlay, bool use_phys)
 	struct i915_vma *vma;
 	int err;
 
-	mutex_lock(&i915->drm.struct_mutex);
-
 	obj = i915_gem_object_create_stolen(i915, PAGE_SIZE);
-	if (obj == NULL)
+	if (IS_ERR(obj))
 		obj = i915_gem_object_create_internal(i915, PAGE_SIZE);
-	if (IS_ERR(obj)) {
-		err = PTR_ERR(obj);
-		goto err_unlock;
-	}
+	if (IS_ERR(obj))
+		return PTR_ERR(obj);
 
 	vma = i915_gem_object_ggtt_pin(obj, NULL, 0, 0, PIN_MAPPABLE);
 	if (IS_ERR(vma)) {
@@ -1332,13 +1316,10 @@ static int get_registers(struct intel_overlay *overlay, bool use_phys)
 	}
 
 	overlay->reg_bo = obj;
-	mutex_unlock(&i915->drm.struct_mutex);
 	return 0;
 
 err_put_bo:
 	i915_gem_object_put(obj);
-err_unlock:
-	mutex_unlock(&i915->drm.struct_mutex);
 	return err;
 }
 
@@ -1367,8 +1348,7 @@ void intel_overlay_setup(struct drm_i915_private *dev_priv)
 	overlay->contrast = 75;
 	overlay->saturation = 146;
 
-	i915_active_init(dev_priv,
-			 &overlay->last_flip,
+	i915_active_init(&overlay->last_flip,
 			 NULL, intel_overlay_last_flip_retire);
 
 	ret = get_registers(overlay, OVERLAY_NEEDS_PHYSICAL(dev_priv));
diff --git a/drivers/gpu/drm/i915/display/intel_psr.c b/drivers/gpu/drm/i915/display/intel_psr.c
index 3bfb720560c2..6a9f322d3fca 100644
--- a/drivers/gpu/drm/i915/display/intel_psr.c
+++ b/drivers/gpu/drm/i915/display/intel_psr.c
@@ -76,7 +76,7 @@ static bool intel_psr2_enabled(struct drm_i915_private *dev_priv,
 			       const struct intel_crtc_state *crtc_state)
 {
 	/* Cannot enable DSC and PSR2 simultaneously */
-	WARN_ON(crtc_state->dsc_params.compression_enable &&
+	WARN_ON(crtc_state->dsc.compression_enable &&
 		crtc_state->has_psr2);
 
 	switch (dev_priv->psr.debug & I915_PSR_DEBUG_MODE_MASK) {
@@ -88,48 +88,35 @@ static bool intel_psr2_enabled(struct drm_i915_private *dev_priv,
 	}
 }
 
-static int edp_psr_shift(enum transcoder cpu_transcoder)
+static void psr_irq_control(struct drm_i915_private *dev_priv)
 {
-	switch (cpu_transcoder) {
-	case TRANSCODER_A:
-		return EDP_PSR_TRANSCODER_A_SHIFT;
-	case TRANSCODER_B:
-		return EDP_PSR_TRANSCODER_B_SHIFT;
-	case TRANSCODER_C:
-		return EDP_PSR_TRANSCODER_C_SHIFT;
-	default:
-		MISSING_CASE(cpu_transcoder);
-		/* fallthrough */
-	case TRANSCODER_EDP:
-		return EDP_PSR_TRANSCODER_EDP_SHIFT;
-	}
-}
-
-void intel_psr_irq_control(struct drm_i915_private *dev_priv, u32 debug)
-{
-	u32 debug_mask, mask;
-	enum transcoder cpu_transcoder;
-	u32 transcoders = BIT(TRANSCODER_EDP);
+	enum transcoder trans_shift;
+	u32 mask, val;
+	i915_reg_t imr_reg;
 
-	if (INTEL_GEN(dev_priv) >= 8)
-		transcoders |= BIT(TRANSCODER_A) |
-			       BIT(TRANSCODER_B) |
-			       BIT(TRANSCODER_C);
-
-	debug_mask = 0;
-	mask = 0;
-	for_each_cpu_transcoder_masked(dev_priv, cpu_transcoder, transcoders) {
-		int shift = edp_psr_shift(cpu_transcoder);
-
-		mask |= EDP_PSR_ERROR(shift);
-		debug_mask |= EDP_PSR_POST_EXIT(shift) |
-			      EDP_PSR_PRE_ENTRY(shift);
+	/*
+	 * gen12+ has registers relative to transcoder and one per transcoder
+	 * using the same bit definition: handle it as TRANSCODER_EDP to force
+	 * 0 shift in bit definition
+	 */
+	if (INTEL_GEN(dev_priv) >= 12) {
+		trans_shift = 0;
+		imr_reg = TRANS_PSR_IMR(dev_priv->psr.transcoder);
+	} else {
+		trans_shift = dev_priv->psr.transcoder;
+		imr_reg = EDP_PSR_IMR;
 	}
 
-	if (debug & I915_PSR_DEBUG_IRQ)
-		mask |= debug_mask;
+	mask = EDP_PSR_ERROR(trans_shift);
+	if (dev_priv->psr.debug & I915_PSR_DEBUG_IRQ)
+		mask |= EDP_PSR_POST_EXIT(trans_shift) |
+			EDP_PSR_PRE_ENTRY(trans_shift);
 
-	I915_WRITE(EDP_PSR_IMR, ~mask);
+	/* Warning: it is masking/setting reserved bits too */
+	val = I915_READ(imr_reg);
+	val &= ~EDP_PSR_TRANS_MASK(trans_shift);
+	val |= ~mask;
+	I915_WRITE(imr_reg, val);
 }
 
 static void psr_event_print(u32 val, bool psr2_enabled)
@@ -171,60 +158,58 @@ static void psr_event_print(u32 val, bool psr2_enabled)
 
 void intel_psr_irq_handler(struct drm_i915_private *dev_priv, u32 psr_iir)
 {
-	u32 transcoders = BIT(TRANSCODER_EDP);
-	enum transcoder cpu_transcoder;
+	enum transcoder cpu_transcoder = dev_priv->psr.transcoder;
+	enum transcoder trans_shift;
+	i915_reg_t imr_reg;
 	ktime_t time_ns =  ktime_get();
-	u32 mask = 0;
 
-	if (INTEL_GEN(dev_priv) >= 8)
-		transcoders |= BIT(TRANSCODER_A) |
-			       BIT(TRANSCODER_B) |
-			       BIT(TRANSCODER_C);
+	if (INTEL_GEN(dev_priv) >= 12) {
+		trans_shift = 0;
+		imr_reg = TRANS_PSR_IMR(dev_priv->psr.transcoder);
+	} else {
+		trans_shift = dev_priv->psr.transcoder;
+		imr_reg = EDP_PSR_IMR;
+	}
 
-	for_each_cpu_transcoder_masked(dev_priv, cpu_transcoder, transcoders) {
-		int shift = edp_psr_shift(cpu_transcoder);
+	if (psr_iir & EDP_PSR_PRE_ENTRY(trans_shift)) {
+		dev_priv->psr.last_entry_attempt = time_ns;
+		DRM_DEBUG_KMS("[transcoder %s] PSR entry attempt in 2 vblanks\n",
+			      transcoder_name(cpu_transcoder));
+	}
 
-		if (psr_iir & EDP_PSR_ERROR(shift)) {
-			DRM_WARN("[transcoder %s] PSR aux error\n",
-				 transcoder_name(cpu_transcoder));
+	if (psr_iir & EDP_PSR_POST_EXIT(trans_shift)) {
+		dev_priv->psr.last_exit = time_ns;
+		DRM_DEBUG_KMS("[transcoder %s] PSR exit completed\n",
+			      transcoder_name(cpu_transcoder));
 
-			dev_priv->psr.irq_aux_error = true;
+		if (INTEL_GEN(dev_priv) >= 9) {
+			u32 val = I915_READ(PSR_EVENT(cpu_transcoder));
+			bool psr2_enabled = dev_priv->psr.psr2_enabled;
 
-			/*
-			 * If this interruption is not masked it will keep
-			 * interrupting so fast that it prevents the scheduled
-			 * work to run.
-			 * Also after a PSR error, we don't want to arm PSR
-			 * again so we don't care about unmask the interruption
-			 * or unset irq_aux_error.
-			 */
-			mask |= EDP_PSR_ERROR(shift);
-		}
-
-		if (psr_iir & EDP_PSR_PRE_ENTRY(shift)) {
-			dev_priv->psr.last_entry_attempt = time_ns;
-			DRM_DEBUG_KMS("[transcoder %s] PSR entry attempt in 2 vblanks\n",
-				      transcoder_name(cpu_transcoder));
+			I915_WRITE(PSR_EVENT(cpu_transcoder), val);
+			psr_event_print(val, psr2_enabled);
 		}
+	}
 
-		if (psr_iir & EDP_PSR_POST_EXIT(shift)) {
-			dev_priv->psr.last_exit = time_ns;
-			DRM_DEBUG_KMS("[transcoder %s] PSR exit completed\n",
-				      transcoder_name(cpu_transcoder));
+	if (psr_iir & EDP_PSR_ERROR(trans_shift)) {
+		u32 val;
 
-			if (INTEL_GEN(dev_priv) >= 9) {
-				u32 val = I915_READ(PSR_EVENT(cpu_transcoder));
-				bool psr2_enabled = dev_priv->psr.psr2_enabled;
+		DRM_WARN("[transcoder %s] PSR aux error\n",
+			 transcoder_name(cpu_transcoder));
 
-				I915_WRITE(PSR_EVENT(cpu_transcoder), val);
-				psr_event_print(val, psr2_enabled);
-			}
-		}
-	}
+		dev_priv->psr.irq_aux_error = true;
 
-	if (mask) {
-		mask |= I915_READ(EDP_PSR_IMR);
-		I915_WRITE(EDP_PSR_IMR, mask);
+		/*
+		 * If this interruption is not masked it will keep
+		 * interrupting so fast that it prevents the scheduled
+		 * work to run.
+		 * Also after a PSR error, we don't want to arm PSR
+		 * again so we don't care about unmask the interruption
+		 * or unset irq_aux_error.
+		 */
+		val = I915_READ(imr_reg);
+		val |= EDP_PSR_ERROR(trans_shift);
+		I915_WRITE(imr_reg, val);
 
 		schedule_work(&dev_priv->psr.work);
 	}
@@ -283,6 +268,11 @@ void intel_psr_init_dpcd(struct intel_dp *intel_dp)
 	struct drm_i915_private *dev_priv =
 		to_i915(dp_to_dig_port(intel_dp)->base.base.dev);
 
+	if (dev_priv->psr.dp) {
+		DRM_WARN("More than one eDP panel found, PSR support should be extended\n");
+		return;
+	}
+
 	drm_dp_dpcd_read(&intel_dp->aux, DP_PSR_SUPPORT, intel_dp->psr_dpcd,
 			 sizeof(intel_dp->psr_dpcd));
 
@@ -305,7 +295,6 @@ void intel_psr_init_dpcd(struct intel_dp *intel_dp)
 	dev_priv->psr.sink_sync_latency =
 		intel_dp_get_sink_sync_latency(intel_dp);
 
-	WARN_ON(dev_priv->psr.dp);
 	dev_priv->psr.dp = intel_dp;
 
 	if (INTEL_GEN(dev_priv) >= 9 &&
@@ -390,7 +379,7 @@ static void hsw_psr_setup_aux(struct intel_dp *intel_dp)
 
 	BUILD_BUG_ON(sizeof(aux_msg) > 20);
 	for (i = 0; i < sizeof(aux_msg); i += 4)
-		I915_WRITE(EDP_PSR_AUX_DATA(i >> 2),
+		I915_WRITE(EDP_PSR_AUX_DATA(dev_priv->psr.transcoder, i >> 2),
 			   intel_dp_pack_aux(&aux_msg[i], sizeof(aux_msg) - i));
 
 	aux_clock_divider = intel_dp->get_aux_clock_divider(intel_dp, 0);
@@ -401,7 +390,7 @@ static void hsw_psr_setup_aux(struct intel_dp *intel_dp)
 
 	/* Select only valid bits for SRD_AUX_CTL */
 	aux_ctl &= psr_aux_mask;
-	I915_WRITE(EDP_PSR_AUX_CTL, aux_ctl);
+	I915_WRITE(EDP_PSR_AUX_CTL(dev_priv->psr.transcoder), aux_ctl);
 }
 
 static void intel_psr_enable_sink(struct intel_dp *intel_dp)
@@ -491,8 +480,9 @@ static void hsw_activate_psr1(struct intel_dp *intel_dp)
 	if (INTEL_GEN(dev_priv) >= 8)
 		val |= EDP_PSR_CRC_ENABLE;
 
-	val |= I915_READ(EDP_PSR_CTL) & EDP_PSR_RESTORE_PSR_ACTIVE_CTX_MASK;
-	I915_WRITE(EDP_PSR_CTL, val);
+	val |= (I915_READ(EDP_PSR_CTL(dev_priv->psr.transcoder)) &
+		EDP_PSR_RESTORE_PSR_ACTIVE_CTX_MASK);
+	I915_WRITE(EDP_PSR_CTL(dev_priv->psr.transcoder), val);
 }
 
 static void hsw_activate_psr2(struct intel_dp *intel_dp)
@@ -528,9 +518,87 @@ static void hsw_activate_psr2(struct intel_dp *intel_dp)
 	 * PSR2 HW is incorrectly using EDP_PSR_TP1_TP3_SEL and BSpec is
 	 * recommending keep this bit unset while PSR2 is enabled.
 	 */
-	I915_WRITE(EDP_PSR_CTL, 0);
+	I915_WRITE(EDP_PSR_CTL(dev_priv->psr.transcoder), 0);
+
+	I915_WRITE(EDP_PSR2_CTL(dev_priv->psr.transcoder), val);
+}
 
-	I915_WRITE(EDP_PSR2_CTL, val);
+static bool
+transcoder_has_psr2(struct drm_i915_private *dev_priv, enum transcoder trans)
+{
+	if (INTEL_GEN(dev_priv) < 9)
+		return false;
+	else if (INTEL_GEN(dev_priv) >= 12)
+		return trans == TRANSCODER_A;
+	else
+		return trans == TRANSCODER_EDP;
+}
+
+static u32 intel_get_frame_time_us(const struct intel_crtc_state *cstate)
+{
+	if (!cstate || !cstate->base.active)
+		return 0;
+
+	return DIV_ROUND_UP(1000 * 1000,
+			    drm_mode_vrefresh(&cstate->base.adjusted_mode));
+}
+
+static void psr2_program_idle_frames(struct drm_i915_private *dev_priv,
+				     u32 idle_frames)
+{
+	u32 val;
+
+	idle_frames <<=  EDP_PSR2_IDLE_FRAME_SHIFT;
+	val = I915_READ(EDP_PSR2_CTL(dev_priv->psr.transcoder));
+	val &= ~EDP_PSR2_IDLE_FRAME_MASK;
+	val |= idle_frames;
+	I915_WRITE(EDP_PSR2_CTL(dev_priv->psr.transcoder), val);
+}
+
+static void tgl_psr2_enable_dc3co(struct drm_i915_private *dev_priv)
+{
+	psr2_program_idle_frames(dev_priv, 0);
+	intel_display_power_set_target_dc_state(dev_priv, DC_STATE_EN_DC3CO);
+}
+
+static void tgl_psr2_disable_dc3co(struct drm_i915_private *dev_priv)
+{
+	int idle_frames;
+
+	intel_display_power_set_target_dc_state(dev_priv, DC_STATE_EN_UPTO_DC6);
+	/*
+	 * Restore PSR2 idle frame let's use 6 as the minimum to cover all known
+	 * cases including the off-by-one issue that HW has in some cases.
+	 */
+	idle_frames = max(6, dev_priv->vbt.psr.idle_frames);
+	idle_frames = max(idle_frames, dev_priv->psr.sink_sync_latency + 1);
+	psr2_program_idle_frames(dev_priv, idle_frames);
+}
+
+static void tgl_dc5_idle_thread(struct work_struct *work)
+{
+	struct drm_i915_private *dev_priv =
+		container_of(work, typeof(*dev_priv), psr.idle_work.work);
+
+	mutex_lock(&dev_priv->psr.lock);
+	/* If delayed work is pending, it is not idle */
+	if (delayed_work_pending(&dev_priv->psr.idle_work))
+		goto unlock;
+
+	DRM_DEBUG_KMS("DC5/6 idle thread\n");
+	tgl_psr2_disable_dc3co(dev_priv);
+unlock:
+	mutex_unlock(&dev_priv->psr.lock);
+}
+
+static void tgl_disallow_dc3co_on_psr2_exit(struct drm_i915_private *dev_priv)
+{
+	if (!dev_priv->psr.dc3co_enabled)
+		return;
+
+	cancel_delayed_work(&dev_priv->psr.idle_work);
+	/* Before PSR2 exit disallow dc3co*/
+	tgl_psr2_disable_dc3co(dev_priv);
 }
 
 static bool intel_psr2_config_valid(struct intel_dp *intel_dp,
@@ -544,17 +612,26 @@ static bool intel_psr2_config_valid(struct intel_dp *intel_dp,
 	if (!dev_priv->psr.sink_psr2_support)
 		return false;
 
+	if (!transcoder_has_psr2(dev_priv, crtc_state->cpu_transcoder)) {
+		DRM_DEBUG_KMS("PSR2 not supported in transcoder %s\n",
+			      transcoder_name(crtc_state->cpu_transcoder));
+		return false;
+	}
+
 	/*
 	 * DSC and PSR2 cannot be enabled simultaneously. If a requested
 	 * resolution requires DSC to be enabled, priority is given to DSC
 	 * over PSR2.
 	 */
-	if (crtc_state->dsc_params.compression_enable) {
+	if (crtc_state->dsc.compression_enable) {
 		DRM_DEBUG_KMS("PSR2 cannot be enabled since DSC is enabled\n");
 		return false;
 	}
 
-	if (INTEL_GEN(dev_priv) >= 10 || IS_GEMINILAKE(dev_priv)) {
+	if (INTEL_GEN(dev_priv) >= 12) {
+		psr_max_h = 5120;
+		psr_max_v = 3200;
+	} else if (INTEL_GEN(dev_priv) >= 10 || IS_GEMINILAKE(dev_priv)) {
 		psr_max_h = 4096;
 		psr_max_v = 2304;
 	} else if (IS_GEN(dev_priv, 9)) {
@@ -606,10 +683,9 @@ void intel_psr_compute_config(struct intel_dp *intel_dp,
 
 	/*
 	 * HSW spec explicitly says PSR is tied to port A.
-	 * BDW+ platforms with DDI implementation of PSR have different
-	 * PSR registers per transcoder and we only implement transcoder EDP
-	 * ones. Since by Display design transcoder EDP is tied to port A
-	 * we can safely escape based on the port A.
+	 * BDW+ platforms have a instance of PSR registers per transcoder but
+	 * for now it only supports one instance of PSR, so lets keep it
+	 * hardcoded to PORT_A
 	 */
 	if (dig_port->base.port != PORT_A) {
 		DRM_DEBUG_KMS("PSR condition failed: Port not supported\n");
@@ -648,9 +724,10 @@ static void intel_psr_activate(struct intel_dp *intel_dp)
 {
 	struct drm_i915_private *dev_priv = dp_to_i915(intel_dp);
 
-	if (INTEL_GEN(dev_priv) >= 9)
-		WARN_ON(I915_READ(EDP_PSR2_CTL) & EDP_PSR2_ENABLE);
-	WARN_ON(I915_READ(EDP_PSR_CTL) & EDP_PSR_ENABLE);
+	if (transcoder_has_psr2(dev_priv, dev_priv->psr.transcoder))
+		WARN_ON(I915_READ(EDP_PSR2_CTL(dev_priv->psr.transcoder)) & EDP_PSR2_ENABLE);
+
+	WARN_ON(I915_READ(EDP_PSR_CTL(dev_priv->psr.transcoder)) & EDP_PSR_ENABLE);
 	WARN_ON(dev_priv->psr.active);
 	lockdep_assert_held(&dev_priv->psr.lock);
 
@@ -663,25 +740,6 @@ static void intel_psr_activate(struct intel_dp *intel_dp)
 	dev_priv->psr.active = true;
 }
 
-static i915_reg_t gen9_chicken_trans_reg(struct drm_i915_private *dev_priv,
-					 enum transcoder cpu_transcoder)
-{
-	static const i915_reg_t regs[] = {
-		[TRANSCODER_A] = CHICKEN_TRANS_A,
-		[TRANSCODER_B] = CHICKEN_TRANS_B,
-		[TRANSCODER_C] = CHICKEN_TRANS_C,
-		[TRANSCODER_EDP] = CHICKEN_TRANS_EDP,
-	};
-
-	WARN_ON(INTEL_GEN(dev_priv) < 9);
-
-	if (WARN_ON(cpu_transcoder >= ARRAY_SIZE(regs) ||
-		    !regs[cpu_transcoder].reg))
-		cpu_transcoder = TRANSCODER_A;
-
-	return regs[cpu_transcoder];
-}
-
 static void intel_psr_enable_source(struct intel_dp *intel_dp,
 				    const struct intel_crtc_state *crtc_state)
 {
@@ -697,8 +755,7 @@ static void intel_psr_enable_source(struct intel_dp *intel_dp,
 
 	if (dev_priv->psr.psr2_enabled && (IS_GEN(dev_priv, 9) &&
 					   !IS_GEMINILAKE(dev_priv))) {
-		i915_reg_t reg = gen9_chicken_trans_reg(dev_priv,
-							cpu_transcoder);
+		i915_reg_t reg = CHICKEN_TRANS(cpu_transcoder);
 		u32 chicken = I915_READ(reg);
 
 		chicken |= PSR2_VSC_ENABLE_PROG_HEADER |
@@ -720,19 +777,46 @@ static void intel_psr_enable_source(struct intel_dp *intel_dp,
 	if (INTEL_GEN(dev_priv) < 11)
 		mask |= EDP_PSR_DEBUG_MASK_DISP_REG_WRITE;
 
-	I915_WRITE(EDP_PSR_DEBUG, mask);
+	I915_WRITE(EDP_PSR_DEBUG(dev_priv->psr.transcoder), mask);
+
+	psr_irq_control(dev_priv);
 }
 
 static void intel_psr_enable_locked(struct drm_i915_private *dev_priv,
 				    const struct intel_crtc_state *crtc_state)
 {
 	struct intel_dp *intel_dp = dev_priv->psr.dp;
+	u32 val;
 
 	WARN_ON(dev_priv->psr.enabled);
 
 	dev_priv->psr.psr2_enabled = intel_psr2_enabled(dev_priv, crtc_state);
 	dev_priv->psr.busy_frontbuffer_bits = 0;
 	dev_priv->psr.pipe = to_intel_crtc(crtc_state->base.crtc)->pipe;
+	dev_priv->psr.dc3co_enabled = !!crtc_state->dc3co_exitline;
+	dev_priv->psr.dc3co_exit_delay = intel_get_frame_time_us(crtc_state);
+	dev_priv->psr.transcoder = crtc_state->cpu_transcoder;
+
+	/*
+	 * If a PSR error happened and the driver is reloaded, the EDP_PSR_IIR
+	 * will still keep the error set even after the reset done in the
+	 * irq_preinstall and irq_uninstall hooks.
+	 * And enabling in this situation cause the screen to freeze in the
+	 * first time that PSR HW tries to activate so lets keep PSR disabled
+	 * to avoid any rendering problems.
+	 */
+	if (INTEL_GEN(dev_priv) >= 12) {
+		val = I915_READ(TRANS_PSR_IIR(dev_priv->psr.transcoder));
+		val &= EDP_PSR_ERROR(0);
+	} else {
+		val = I915_READ(EDP_PSR_IIR);
+		val &= EDP_PSR_ERROR(dev_priv->psr.transcoder);
+	}
+	if (val) {
+		dev_priv->psr.sink_not_reliable = true;
+		DRM_DEBUG_KMS("PSR interruption error set, not enabling PSR\n");
+		return;
+	}
 
 	DRM_DEBUG_KMS("Enabling PSR%s\n",
 		      dev_priv->psr.psr2_enabled ? "2" : "1");
@@ -782,20 +866,28 @@ static void intel_psr_exit(struct drm_i915_private *dev_priv)
 	u32 val;
 
 	if (!dev_priv->psr.active) {
-		if (INTEL_GEN(dev_priv) >= 9)
-			WARN_ON(I915_READ(EDP_PSR2_CTL) & EDP_PSR2_ENABLE);
-		WARN_ON(I915_READ(EDP_PSR_CTL) & EDP_PSR_ENABLE);
+		if (transcoder_has_psr2(dev_priv, dev_priv->psr.transcoder)) {
+			val = I915_READ(EDP_PSR2_CTL(dev_priv->psr.transcoder));
+			WARN_ON(val & EDP_PSR2_ENABLE);
+		}
+
+		val = I915_READ(EDP_PSR_CTL(dev_priv->psr.transcoder));
+		WARN_ON(val & EDP_PSR_ENABLE);
+
 		return;
 	}
 
 	if (dev_priv->psr.psr2_enabled) {
-		val = I915_READ(EDP_PSR2_CTL);
+		tgl_disallow_dc3co_on_psr2_exit(dev_priv);
+		val = I915_READ(EDP_PSR2_CTL(dev_priv->psr.transcoder));
 		WARN_ON(!(val & EDP_PSR2_ENABLE));
-		I915_WRITE(EDP_PSR2_CTL, val & ~EDP_PSR2_ENABLE);
+		val &= ~EDP_PSR2_ENABLE;
+		I915_WRITE(EDP_PSR2_CTL(dev_priv->psr.transcoder), val);
 	} else {
-		val = I915_READ(EDP_PSR_CTL);
+		val = I915_READ(EDP_PSR_CTL(dev_priv->psr.transcoder));
 		WARN_ON(!(val & EDP_PSR_ENABLE));
-		I915_WRITE(EDP_PSR_CTL, val & ~EDP_PSR_ENABLE);
+		val &= ~EDP_PSR_ENABLE;
+		I915_WRITE(EDP_PSR_CTL(dev_priv->psr.transcoder), val);
 	}
 	dev_priv->psr.active = false;
 }
@@ -817,10 +909,10 @@ static void intel_psr_disable_locked(struct intel_dp *intel_dp)
 	intel_psr_exit(dev_priv);
 
 	if (dev_priv->psr.psr2_enabled) {
-		psr_status = EDP_PSR2_STATUS;
+		psr_status = EDP_PSR2_STATUS(dev_priv->psr.transcoder);
 		psr_status_mask = EDP_PSR2_STATUS_STATE_MASK;
 	} else {
-		psr_status = EDP_PSR_STATUS;
+		psr_status = EDP_PSR_STATUS(dev_priv->psr.transcoder);
 		psr_status_mask = EDP_PSR_STATUS_STATE_MASK;
 	}
 
@@ -859,6 +951,7 @@ void intel_psr_disable(struct intel_dp *intel_dp,
 
 	mutex_unlock(&dev_priv->psr.lock);
 	cancel_work_sync(&dev_priv->psr.work);
+	cancel_delayed_work_sync(&dev_priv->psr.idle_work);
 }
 
 static void psr_force_hw_tracking_exit(struct drm_i915_private *dev_priv)
@@ -963,7 +1056,8 @@ int intel_psr_wait_for_idle(const struct intel_crtc_state *new_crtc_state,
 	 * defensive enough to cover everything.
 	 */
 
-	return __intel_wait_for_register(&dev_priv->uncore, EDP_PSR_STATUS,
+	return __intel_wait_for_register(&dev_priv->uncore,
+					 EDP_PSR_STATUS(dev_priv->psr.transcoder),
 					 EDP_PSR_STATUS_STATE_MASK,
 					 EDP_PSR_STATUS_STATE_IDLE, 2, 50,
 					 out_value);
@@ -979,10 +1073,10 @@ static bool __psr_wait_for_idle_locked(struct drm_i915_private *dev_priv)
 		return false;
 
 	if (dev_priv->psr.psr2_enabled) {
-		reg = EDP_PSR2_STATUS;
+		reg = EDP_PSR2_STATUS(dev_priv->psr.transcoder);
 		mask = EDP_PSR2_STATUS_STATE_MASK;
 	} else {
-		reg = EDP_PSR_STATUS;
+		reg = EDP_PSR_STATUS(dev_priv->psr.transcoder);
 		mask = EDP_PSR_STATUS_STATE_MASK;
 	}
 
@@ -1067,7 +1161,13 @@ int intel_psr_debug_set(struct drm_i915_private *dev_priv, u64 val)
 
 	old_mode = dev_priv->psr.debug & I915_PSR_DEBUG_MODE_MASK;
 	dev_priv->psr.debug = val;
-	intel_psr_irq_control(dev_priv, dev_priv->psr.debug);
+
+	/*
+	 * Do it right away if it's already enabled, otherwise it will be done
+	 * when enabling the source.
+	 */
+	if (dev_priv->psr.enabled)
+		psr_irq_control(dev_priv);
 
 	mutex_unlock(&dev_priv->psr.lock);
 
@@ -1159,6 +1259,44 @@ void intel_psr_invalidate(struct drm_i915_private *dev_priv,
 	mutex_unlock(&dev_priv->psr.lock);
 }
 
+/*
+ * When we will be completely rely on PSR2 S/W tracking in future,
+ * intel_psr_flush() will invalidate and flush the PSR for ORIGIN_FLIP
+ * event also therefore tgl_dc3co_flush() require to be changed
+ * accrodingly in future.
+ */
+static void
+tgl_dc3co_flush(struct drm_i915_private *dev_priv,
+		unsigned int frontbuffer_bits, enum fb_op_origin origin)
+{
+	u32 delay;
+
+	mutex_lock(&dev_priv->psr.lock);
+
+	if (!dev_priv->psr.dc3co_enabled)
+		goto unlock;
+
+	if (!dev_priv->psr.psr2_enabled || !dev_priv->psr.active)
+		goto unlock;
+
+	/*
+	 * At every frontbuffer flush flip event modified delay of delayed work,
+	 * when delayed work schedules that means display has been idle.
+	 */
+	if (!(frontbuffer_bits &
+	    INTEL_FRONTBUFFER_ALL_MASK(dev_priv->psr.pipe)))
+		goto unlock;
+
+	tgl_psr2_enable_dc3co(dev_priv);
+	/* DC5/DC6 required idle frames = 6 */
+	delay = 6 * dev_priv->psr.dc3co_exit_delay;
+	mod_delayed_work(system_wq, &dev_priv->psr.idle_work,
+			 usecs_to_jiffies(delay));
+
+unlock:
+	mutex_unlock(&dev_priv->psr.lock);
+}
+
 /**
  * intel_psr_flush - Flush PSR
  * @dev_priv: i915 device
@@ -1178,8 +1316,10 @@ void intel_psr_flush(struct drm_i915_private *dev_priv,
 	if (!CAN_PSR(dev_priv))
 		return;
 
-	if (origin == ORIGIN_FLIP)
+	if (origin == ORIGIN_FLIP) {
+		tgl_dc3co_flush(dev_priv, frontbuffer_bits, origin);
 		return;
+	}
 
 	mutex_lock(&dev_priv->psr.lock);
 	if (!dev_priv->psr.enabled) {
@@ -1208,45 +1348,34 @@ void intel_psr_flush(struct drm_i915_private *dev_priv,
  */
 void intel_psr_init(struct drm_i915_private *dev_priv)
 {
-	u32 val;
-
 	if (!HAS_PSR(dev_priv))
 		return;
 
-	dev_priv->psr_mmio_base = IS_HASWELL(dev_priv) ?
-		HSW_EDP_PSR_BASE : BDW_EDP_PSR_BASE;
-
 	if (!dev_priv->psr.sink_support)
 		return;
 
+	if (IS_HASWELL(dev_priv))
+		/*
+		 * HSW don't have PSR registers on the same space as transcoder
+		 * so set this to a value that when subtract to the register
+		 * in transcoder space results in the right offset for HSW
+		 */
+		dev_priv->hsw_psr_mmio_adjust = _SRD_CTL_EDP - _HSW_EDP_PSR_BASE;
+
 	if (i915_modparams.enable_psr == -1)
 		if (INTEL_GEN(dev_priv) < 9 || !dev_priv->vbt.psr.enable)
 			i915_modparams.enable_psr = 0;
 
-	/*
-	 * If a PSR error happened and the driver is reloaded, the EDP_PSR_IIR
-	 * will still keep the error set even after the reset done in the
-	 * irq_preinstall and irq_uninstall hooks.
-	 * And enabling in this situation cause the screen to freeze in the
-	 * first time that PSR HW tries to activate so lets keep PSR disabled
-	 * to avoid any rendering problems.
-	 */
-	val = I915_READ(EDP_PSR_IIR);
-	val &= EDP_PSR_ERROR(edp_psr_shift(TRANSCODER_EDP));
-	if (val) {
-		DRM_DEBUG_KMS("PSR interruption error set\n");
-		dev_priv->psr.sink_not_reliable = true;
-	}
-
 	/* Set link_standby x link_off defaults */
 	if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv))
 		/* HSW and BDW require workarounds that we don't implement. */
 		dev_priv->psr.link_standby = false;
-	else
-		/* For new platforms let's respect VBT back again */
+	else if (INTEL_GEN(dev_priv) < 12)
+		/* For new platforms up to TGL let's respect VBT back again */
 		dev_priv->psr.link_standby = dev_priv->vbt.psr.full_link;
 
 	INIT_WORK(&dev_priv->psr.work, intel_psr_work);
+	INIT_DELAYED_WORK(&dev_priv->psr.idle_work, tgl_dc5_idle_thread);
 	mutex_init(&dev_priv->psr.lock);
 }
 
@@ -1288,7 +1417,7 @@ void intel_psr_short_pulse(struct intel_dp *intel_dp)
 	if (val & DP_PSR_VSC_SDP_UNCORRECTABLE_ERROR)
 		DRM_DEBUG_KMS("PSR VSC SDP uncorrectable error, disabling PSR\n");
 	if (val & DP_PSR_LINK_CRC_ERROR)
-		DRM_ERROR("PSR Link CRC error, disabling PSR\n");
+		DRM_DEBUG_KMS("PSR Link CRC error, disabling PSR\n");
 
 	if (val & ~errors)
 		DRM_ERROR("PSR_ERROR_STATUS unhandled errors %x\n",
diff --git a/drivers/gpu/drm/i915/display/intel_psr.h b/drivers/gpu/drm/i915/display/intel_psr.h
index dc818826f36d..46e4de8b8cd5 100644
--- a/drivers/gpu/drm/i915/display/intel_psr.h
+++ b/drivers/gpu/drm/i915/display/intel_psr.h
@@ -30,7 +30,6 @@ void intel_psr_flush(struct drm_i915_private *dev_priv,
 void intel_psr_init(struct drm_i915_private *dev_priv);
 void intel_psr_compute_config(struct intel_dp *intel_dp,
 			      struct intel_crtc_state *crtc_state);
-void intel_psr_irq_control(struct drm_i915_private *dev_priv, u32 debug);
 void intel_psr_irq_handler(struct drm_i915_private *dev_priv, u32 psr_iir);
 void intel_psr_short_pulse(struct intel_dp *intel_dp);
 int intel_psr_wait_for_idle(const struct intel_crtc_state *new_crtc_state,
diff --git a/drivers/gpu/drm/i915/display/intel_sdvo.c b/drivers/gpu/drm/i915/display/intel_sdvo.c
index adeb1c840976..5b7f4baf7348 100644
--- a/drivers/gpu/drm/i915/display/intel_sdvo.c
+++ b/drivers/gpu/drm/i915/display/intel_sdvo.c
@@ -2921,7 +2921,7 @@ intel_sdvo_output_setup(struct intel_sdvo *intel_sdvo, u16 flags)
 			      bytes[0], bytes[1]);
 		return false;
 	}
-	intel_sdvo->base.crtc_mask = (1 << 0) | (1 << 1) | (1 << 2);
+	intel_sdvo->base.pipe_mask = ~0;
 
 	return true;
 }
diff --git a/drivers/gpu/drm/i915/display/intel_sdvo.h b/drivers/gpu/drm/i915/display/intel_sdvo.h
index c9e05bcdd141..a66f224aa17d 100644
--- a/drivers/gpu/drm/i915/display/intel_sdvo.h
+++ b/drivers/gpu/drm/i915/display/intel_sdvo.h
@@ -14,6 +14,7 @@
 
 struct drm_i915_private;
 enum pipe;
+enum port;
 
 bool intel_sdvo_port_enabled(struct drm_i915_private *dev_priv,
 			     i915_reg_t sdvo_reg, enum pipe *pipe);
diff --git a/drivers/gpu/drm/i915/display/intel_sprite.c b/drivers/gpu/drm/i915/display/intel_sprite.c
index dea63be1964f..72fda0430062 100644
--- a/drivers/gpu/drm/i915/display/intel_sprite.c
+++ b/drivers/gpu/drm/i915/display/intel_sprite.c
@@ -48,19 +48,6 @@
 #include "intel_psr.h"
 #include "intel_sprite.h"
 
-bool is_planar_yuv_format(u32 pixelformat)
-{
-	switch (pixelformat) {
-	case DRM_FORMAT_NV12:
-	case DRM_FORMAT_P010:
-	case DRM_FORMAT_P012:
-	case DRM_FORMAT_P016:
-		return true;
-	default:
-		return false;
-	}
-}
-
 int intel_usecs_to_scanlines(const struct drm_display_mode *adjusted_mode,
 			     int usecs)
 {
@@ -300,10 +287,8 @@ int intel_plane_check_src_coordinates(struct intel_plane_state *plane_state)
 	src_y = src->y1 >> 16;
 	src_h = drm_rect_height(src) >> 16;
 
-	src->x1 = src_x << 16;
-	src->x2 = (src_x + src_w) << 16;
-	src->y1 = src_y << 16;
-	src->y2 = (src_y + src_h) << 16;
+	drm_rect_init(src, src_x << 16, src_y << 16,
+		      src_w << 16, src_h << 16);
 
 	if (!fb->format->is_yuv)
 		return 0;
@@ -337,6 +322,55 @@ bool icl_is_hdr_plane(struct drm_i915_private *dev_priv, enum plane_id plane_id)
 		icl_hdr_plane_mask() & BIT(plane_id);
 }
 
+static void
+skl_plane_ratio(const struct intel_crtc_state *crtc_state,
+		const struct intel_plane_state *plane_state,
+		unsigned int *num, unsigned int *den)
+{
+	struct drm_i915_private *dev_priv = to_i915(plane_state->base.plane->dev);
+	const struct drm_framebuffer *fb = plane_state->base.fb;
+
+	if (fb->format->cpp[0] == 8) {
+		if (INTEL_GEN(dev_priv) >= 10 || IS_GEMINILAKE(dev_priv)) {
+			*num = 10;
+			*den = 8;
+		} else {
+			*num = 9;
+			*den = 8;
+		}
+	} else {
+		*num = 1;
+		*den = 1;
+	}
+}
+
+static int skl_plane_min_cdclk(const struct intel_crtc_state *crtc_state,
+			       const struct intel_plane_state *plane_state)
+{
+	struct drm_i915_private *dev_priv = to_i915(plane_state->base.plane->dev);
+	unsigned int pixel_rate = crtc_state->pixel_rate;
+	unsigned int src_w, src_h, dst_w, dst_h;
+	unsigned int num, den;
+
+	skl_plane_ratio(crtc_state, plane_state, &num, &den);
+
+	/* two pixels per clock on glk+ */
+	if (INTEL_GEN(dev_priv) >= 10 || IS_GEMINILAKE(dev_priv))
+		den *= 2;
+
+	src_w = drm_rect_width(&plane_state->base.src) >> 16;
+	src_h = drm_rect_height(&plane_state->base.src) >> 16;
+	dst_w = drm_rect_width(&plane_state->base.dst);
+	dst_h = drm_rect_height(&plane_state->base.dst);
+
+	/* Downscaling limits the maximum pixel rate */
+	dst_w = min(src_w, dst_w);
+	dst_h = min(src_h, dst_h);
+
+	return DIV64_U64_ROUND_UP(mul_u32_u32(pixel_rate * num, src_w * src_h),
+				  mul_u32_u32(den, dst_w * dst_h));
+}
+
 static unsigned int
 skl_plane_max_stride(struct intel_plane *plane,
 		     u32 pixel_format, u64 modifier,
@@ -361,6 +395,7 @@ skl_program_scaler(struct intel_plane *plane,
 		   const struct intel_plane_state *plane_state)
 {
 	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
+	const struct drm_framebuffer *fb = plane_state->base.fb;
 	enum pipe pipe = plane->pipe;
 	int scaler_id = plane_state->scaler_id;
 	const struct intel_scaler *scaler =
@@ -381,7 +416,7 @@ skl_program_scaler(struct intel_plane *plane,
 				      0, INT_MAX);
 
 	/* TODO: handle sub-pixel coordinates */
-	if (is_planar_yuv_format(plane_state->base.fb->format->format) &&
+	if (drm_format_info_is_yuv_semiplanar(fb->format) &&
 	    !icl_is_hdr_plane(dev_priv, plane->id)) {
 		y_hphase = skl_scaler_calc_phase(1, hscale, false);
 		y_vphase = skl_scaler_calc_phase(1, vscale, false);
@@ -554,7 +589,7 @@ skl_program_plane(struct intel_plane *plane,
 	u32 y = plane_state->color_plane[color_plane].y;
 	u32 src_w = drm_rect_width(&plane_state->base.src) >> 16;
 	u32 src_h = drm_rect_height(&plane_state->base.src) >> 16;
-	struct intel_plane *linked = plane_state->linked_plane;
+	struct intel_plane *linked = plane_state->planar_linked_plane;
 	const struct drm_framebuffer *fb = plane_state->base.fb;
 	u8 alpha = plane_state->base.alpha >> 8;
 	u32 plane_color_ctl = 0;
@@ -653,7 +688,7 @@ skl_update_plane(struct intel_plane *plane,
 {
 	int color_plane = 0;
 
-	if (plane_state->linked_plane) {
+	if (plane_state->planar_linked_plane) {
 		/* Program the UV plane */
 		color_plane = 1;
 	}
@@ -825,6 +860,85 @@ vlv_update_clrc(const struct intel_plane_state *plane_state)
 		      SP_SH_SIN(sh_sin) | SP_SH_COS(sh_cos));
 }
 
+static void
+vlv_plane_ratio(const struct intel_crtc_state *crtc_state,
+		const struct intel_plane_state *plane_state,
+		unsigned int *num, unsigned int *den)
+{
+	u8 active_planes = crtc_state->active_planes & ~BIT(PLANE_CURSOR);
+	const struct drm_framebuffer *fb = plane_state->base.fb;
+	unsigned int cpp = fb->format->cpp[0];
+
+	/*
+	 * VLV bspec only considers cases where all three planes are
+	 * enabled, and cases where the primary and one sprite is enabled.
+	 * Let's assume the case with just two sprites enabled also
+	 * maps to the latter case.
+	 */
+	if (hweight8(active_planes) == 3) {
+		switch (cpp) {
+		case 8:
+			*num = 11;
+			*den = 8;
+			break;
+		case 4:
+			*num = 18;
+			*den = 16;
+			break;
+		default:
+			*num = 1;
+			*den = 1;
+			break;
+		}
+	} else if (hweight8(active_planes) == 2) {
+		switch (cpp) {
+		case 8:
+			*num = 10;
+			*den = 8;
+			break;
+		case 4:
+			*num = 17;
+			*den = 16;
+			break;
+		default:
+			*num = 1;
+			*den = 1;
+			break;
+		}
+	} else {
+		switch (cpp) {
+		case 8:
+			*num = 10;
+			*den = 8;
+			break;
+		default:
+			*num = 1;
+			*den = 1;
+			break;
+		}
+	}
+}
+
+int vlv_plane_min_cdclk(const struct intel_crtc_state *crtc_state,
+			const struct intel_plane_state *plane_state)
+{
+	unsigned int pixel_rate;
+	unsigned int num, den;
+
+	/*
+	 * Note that crtc_state->pixel_rate accounts for both
+	 * horizontal and vertical panel fitter downscaling factors.
+	 * Pre-HSW bspec tells us to only consider the horizontal
+	 * downscaling factor here. We ignore that and just consider
+	 * both for simplicity.
+	 */
+	pixel_rate = crtc_state->pixel_rate;
+
+	vlv_plane_ratio(crtc_state, plane_state, &num, &den);
+
+	return DIV_ROUND_UP(pixel_rate * num, den);
+}
+
 static u32 vlv_sprite_ctl_crtc(const struct intel_crtc_state *crtc_state)
 {
 	u32 sprctl = 0;
@@ -1031,6 +1145,164 @@ vlv_plane_get_hw_state(struct intel_plane *plane,
 	return ret;
 }
 
+static void ivb_plane_ratio(const struct intel_crtc_state *crtc_state,
+			    const struct intel_plane_state *plane_state,
+			    unsigned int *num, unsigned int *den)
+{
+	u8 active_planes = crtc_state->active_planes & ~BIT(PLANE_CURSOR);
+	const struct drm_framebuffer *fb = plane_state->base.fb;
+	unsigned int cpp = fb->format->cpp[0];
+
+	if (hweight8(active_planes) == 2) {
+		switch (cpp) {
+		case 8:
+			*num = 10;
+			*den = 8;
+			break;
+		case 4:
+			*num = 17;
+			*den = 16;
+			break;
+		default:
+			*num = 1;
+			*den = 1;
+			break;
+		}
+	} else {
+		switch (cpp) {
+		case 8:
+			*num = 9;
+			*den = 8;
+			break;
+		default:
+			*num = 1;
+			*den = 1;
+			break;
+		}
+	}
+}
+
+static void ivb_plane_ratio_scaling(const struct intel_crtc_state *crtc_state,
+				    const struct intel_plane_state *plane_state,
+				    unsigned int *num, unsigned int *den)
+{
+	const struct drm_framebuffer *fb = plane_state->base.fb;
+	unsigned int cpp = fb->format->cpp[0];
+
+	switch (cpp) {
+	case 8:
+		*num = 12;
+		*den = 8;
+		break;
+	case 4:
+		*num = 19;
+		*den = 16;
+		break;
+	case 2:
+		*num = 33;
+		*den = 32;
+		break;
+	default:
+		*num = 1;
+		*den = 1;
+		break;
+	}
+}
+
+int ivb_plane_min_cdclk(const struct intel_crtc_state *crtc_state,
+			const struct intel_plane_state *plane_state)
+{
+	unsigned int pixel_rate;
+	unsigned int num, den;
+
+	/*
+	 * Note that crtc_state->pixel_rate accounts for both
+	 * horizontal and vertical panel fitter downscaling factors.
+	 * Pre-HSW bspec tells us to only consider the horizontal
+	 * downscaling factor here. We ignore that and just consider
+	 * both for simplicity.
+	 */
+	pixel_rate = crtc_state->pixel_rate;
+
+	ivb_plane_ratio(crtc_state, plane_state, &num, &den);
+
+	return DIV_ROUND_UP(pixel_rate * num, den);
+}
+
+static int ivb_sprite_min_cdclk(const struct intel_crtc_state *crtc_state,
+				const struct intel_plane_state *plane_state)
+{
+	unsigned int src_w, dst_w, pixel_rate;
+	unsigned int num, den;
+
+	/*
+	 * Note that crtc_state->pixel_rate accounts for both
+	 * horizontal and vertical panel fitter downscaling factors.
+	 * Pre-HSW bspec tells us to only consider the horizontal
+	 * downscaling factor here. We ignore that and just consider
+	 * both for simplicity.
+	 */
+	pixel_rate = crtc_state->pixel_rate;
+
+	src_w = drm_rect_width(&plane_state->base.src) >> 16;
+	dst_w = drm_rect_width(&plane_state->base.dst);
+
+	if (src_w != dst_w)
+		ivb_plane_ratio_scaling(crtc_state, plane_state, &num, &den);
+	else
+		ivb_plane_ratio(crtc_state, plane_state, &num, &den);
+
+	/* Horizontal downscaling limits the maximum pixel rate */
+	dst_w = min(src_w, dst_w);
+
+	return DIV_ROUND_UP_ULL(mul_u32_u32(pixel_rate, num * src_w),
+				den * dst_w);
+}
+
+static void hsw_plane_ratio(const struct intel_crtc_state *crtc_state,
+			    const struct intel_plane_state *plane_state,
+			    unsigned int *num, unsigned int *den)
+{
+	u8 active_planes = crtc_state->active_planes & ~BIT(PLANE_CURSOR);
+	const struct drm_framebuffer *fb = plane_state->base.fb;
+	unsigned int cpp = fb->format->cpp[0];
+
+	if (hweight8(active_planes) == 2) {
+		switch (cpp) {
+		case 8:
+			*num = 10;
+			*den = 8;
+			break;
+		default:
+			*num = 1;
+			*den = 1;
+			break;
+		}
+	} else {
+		switch (cpp) {
+		case 8:
+			*num = 9;
+			*den = 8;
+			break;
+		default:
+			*num = 1;
+			*den = 1;
+			break;
+		}
+	}
+}
+
+int hsw_plane_min_cdclk(const struct intel_crtc_state *crtc_state,
+			const struct intel_plane_state *plane_state)
+{
+	unsigned int pixel_rate = crtc_state->pixel_rate;
+	unsigned int num, den;
+
+	hsw_plane_ratio(crtc_state, plane_state, &num, &den);
+
+	return DIV_ROUND_UP(pixel_rate * num, den);
+}
+
 static u32 ivb_sprite_ctl_crtc(const struct intel_crtc_state *crtc_state)
 {
 	u32 sprctl = 0;
@@ -1044,6 +1316,16 @@ static u32 ivb_sprite_ctl_crtc(const struct intel_crtc_state *crtc_state)
 	return sprctl;
 }
 
+static bool ivb_need_sprite_gamma(const struct intel_plane_state *plane_state)
+{
+	struct drm_i915_private *dev_priv =
+		to_i915(plane_state->base.plane->dev);
+	const struct drm_framebuffer *fb = plane_state->base.fb;
+
+	return fb->format->cpp[0] == 8 &&
+		(IS_IVYBRIDGE(dev_priv) || IS_HASWELL(dev_priv));
+}
+
 static u32 ivb_sprite_ctl(const struct intel_crtc_state *crtc_state,
 			  const struct intel_plane_state *plane_state)
 {
@@ -1066,6 +1348,12 @@ static u32 ivb_sprite_ctl(const struct intel_crtc_state *crtc_state,
 	case DRM_FORMAT_XRGB8888:
 		sprctl |= SPRITE_FORMAT_RGBX888;
 		break;
+	case DRM_FORMAT_XBGR16161616F:
+		sprctl |= SPRITE_FORMAT_RGBX161616 | SPRITE_RGB_ORDER_RGBX;
+		break;
+	case DRM_FORMAT_XRGB16161616F:
+		sprctl |= SPRITE_FORMAT_RGBX161616;
+		break;
 	case DRM_FORMAT_YUYV:
 		sprctl |= SPRITE_FORMAT_YUV422 | SPRITE_YUV_ORDER_YUYV;
 		break;
@@ -1083,7 +1371,8 @@ static u32 ivb_sprite_ctl(const struct intel_crtc_state *crtc_state,
 		return 0;
 	}
 
-	sprctl |= SPRITE_INT_GAMMA_DISABLE;
+	if (!ivb_need_sprite_gamma(plane_state))
+		sprctl |= SPRITE_INT_GAMMA_DISABLE;
 
 	if (plane_state->base.color_encoding == DRM_COLOR_YCBCR_BT709)
 		sprctl |= SPRITE_YUV_TO_RGB_CSC_FORMAT_BT709;
@@ -1105,12 +1394,26 @@ static u32 ivb_sprite_ctl(const struct intel_crtc_state *crtc_state,
 	return sprctl;
 }
 
-static void ivb_sprite_linear_gamma(u16 gamma[18])
+static void ivb_sprite_linear_gamma(const struct intel_plane_state *plane_state,
+				    u16 gamma[18])
 {
-	int i;
+	int scale, i;
 
-	for (i = 0; i < 17; i++)
-		gamma[i] = (i << 10) / 16;
+	/*
+	 * WaFP16GammaEnabling:ivb,hsw
+	 * "Workaround : When using the 64-bit format, the sprite output
+	 *  on each color channel has one quarter amplitude. It can be
+	 *  brought up to full amplitude by using sprite internal gamma
+	 *  correction, pipe gamma correction, or pipe color space
+	 *  conversion to multiply the sprite output by four."
+	 */
+	scale = 4;
+
+	for (i = 0; i < 16; i++)
+		gamma[i] = min((scale * i << 10) / 16, (1 << 10) - 1);
+
+	gamma[i] = min((scale * i << 10) / 16, 1 << 10);
+	i++;
 
 	gamma[i] = 3 << 10;
 	i++;
@@ -1124,7 +1427,10 @@ static void ivb_update_gamma(const struct intel_plane_state *plane_state)
 	u16 gamma[18];
 	int i;
 
-	ivb_sprite_linear_gamma(gamma);
+	if (!ivb_need_sprite_gamma(plane_state))
+		return;
+
+	ivb_sprite_linear_gamma(plane_state, gamma);
 
 	/* FIXME these register are single buffered :( */
 	for (i = 0; i < 16; i++)
@@ -1257,6 +1563,53 @@ ivb_plane_get_hw_state(struct intel_plane *plane,
 	return ret;
 }
 
+static int g4x_sprite_min_cdclk(const struct intel_crtc_state *crtc_state,
+				const struct intel_plane_state *plane_state)
+{
+	const struct drm_framebuffer *fb = plane_state->base.fb;
+	unsigned int hscale, pixel_rate;
+	unsigned int limit, decimate;
+
+	/*
+	 * Note that crtc_state->pixel_rate accounts for both
+	 * horizontal and vertical panel fitter downscaling factors.
+	 * Pre-HSW bspec tells us to only consider the horizontal
+	 * downscaling factor here. We ignore that and just consider
+	 * both for simplicity.
+	 */
+	pixel_rate = crtc_state->pixel_rate;
+
+	/* Horizontal downscaling limits the maximum pixel rate */
+	hscale = drm_rect_calc_hscale(&plane_state->base.src,
+				      &plane_state->base.dst,
+				      0, INT_MAX);
+	if (hscale < 0x10000)
+		return pixel_rate;
+
+	/* Decimation steps at 2x,4x,8x,16x */
+	decimate = ilog2(hscale >> 16);
+	hscale >>= decimate;
+
+	/* Starting limit is 90% of cdclk */
+	limit = 9;
+
+	/* -10% per decimation step */
+	limit -= decimate;
+
+	/* -10% for RGB */
+	if (fb->format->cpp[0] >= 4)
+		limit--; /* -10% for RGB */
+
+	/*
+	 * We should also do -10% if sprite scaling is enabled
+	 * on the other pipe, but we can't really check for that,
+	 * so we ignore it.
+	 */
+
+	return DIV_ROUND_UP_ULL(mul_u32_u32(pixel_rate, 10 * hscale),
+				limit << 16);
+}
+
 static unsigned int
 g4x_sprite_max_stride(struct intel_plane *plane,
 		      u32 pixel_format, u64 modifier,
@@ -1300,6 +1653,12 @@ static u32 g4x_sprite_ctl(const struct intel_crtc_state *crtc_state,
 	case DRM_FORMAT_XRGB8888:
 		dvscntr |= DVS_FORMAT_RGBX888;
 		break;
+	case DRM_FORMAT_XBGR16161616F:
+		dvscntr |= DVS_FORMAT_RGBX161616 | DVS_RGB_ORDER_XBGR;
+		break;
+	case DRM_FORMAT_XRGB16161616F:
+		dvscntr |= DVS_FORMAT_RGBX161616;
+		break;
 	case DRM_FORMAT_YUYV:
 		dvscntr |= DVS_FORMAT_YUV422 | DVS_YUV_ORDER_YUYV;
 		break;
@@ -1513,6 +1872,11 @@ static bool intel_fb_scalable(const struct drm_framebuffer *fb)
 	switch (fb->format->format) {
 	case DRM_FORMAT_C8:
 		return false;
+	case DRM_FORMAT_XRGB16161616F:
+	case DRM_FORMAT_ARGB16161616F:
+	case DRM_FORMAT_XBGR16161616F:
+	case DRM_FORMAT_ABGR16161616F:
+		return INTEL_GEN(to_i915(fb->dev)) >= 11;
 	default:
 		return true;
 	}
@@ -1528,6 +1892,7 @@ g4x_sprite_check_scaling(struct intel_crtc_state *crtc_state,
 	int src_x, src_w, src_h, crtc_w, crtc_h;
 	const struct drm_display_mode *adjusted_mode =
 		&crtc_state->base.adjusted_mode;
+	unsigned int stride = plane_state->color_plane[0].stride;
 	unsigned int cpp = fb->format->cpp[0];
 	unsigned int width_bytes;
 	int min_width, min_height;
@@ -1569,9 +1934,9 @@ g4x_sprite_check_scaling(struct intel_crtc_state *crtc_state,
 		return -EINVAL;
 	}
 
-	if (width_bytes > 4096 || fb->pitches[0] > 4096) {
+	if (stride > 4096) {
 		DRM_DEBUG_KMS("Stride (%u) exceeds hardware max with scaling (%u)\n",
-			      fb->pitches[0], 4096);
+			      stride, 4096);
 		return -EINVAL;
 	}
 
@@ -1790,7 +2155,7 @@ static int skl_plane_check_nv12_rotation(const struct intel_plane_state *plane_s
 	int src_w = drm_rect_width(&plane_state->base.src) >> 16;
 
 	/* Display WA #1106 */
-	if (is_planar_yuv_format(fb->format->format) && src_w & 3 &&
+	if (drm_format_info_is_yuv_semiplanar(fb->format) && src_w & 3 &&
 	    (rotation == DRM_MODE_ROTATE_270 ||
 	     rotation == (DRM_MODE_REFLECT_X | DRM_MODE_ROTATE_90))) {
 		DRM_DEBUG_KMS("src width must be multiple of 4 for rotated planar YUV\n");
@@ -1800,6 +2165,22 @@ static int skl_plane_check_nv12_rotation(const struct intel_plane_state *plane_s
 	return 0;
 }
 
+static int skl_plane_max_scale(struct drm_i915_private *dev_priv,
+			       const struct drm_framebuffer *fb)
+{
+	/*
+	 * We don't yet know the final source width nor
+	 * whether we can use the HQ scaler mode. Assume
+	 * the best case.
+	 * FIXME need to properly check this later.
+	 */
+	if (INTEL_GEN(dev_priv) >= 10 || IS_GEMINILAKE(dev_priv) ||
+	    !drm_format_info_is_yuv_semiplanar(fb->format))
+		return 0x30000 - 1;
+	else
+		return 0x20000 - 1;
+}
+
 static int skl_plane_check(struct intel_crtc_state *crtc_state,
 			   struct intel_plane_state *plane_state)
 {
@@ -1817,7 +2198,7 @@ static int skl_plane_check(struct intel_crtc_state *crtc_state,
 	/* use scaler when colorkey is not required */
 	if (!plane_state->ckey.flags && intel_fb_scalable(fb)) {
 		min_scale = 1;
-		max_scale = skl_max_scale(crtc_state, fb->format->format);
+		max_scale = skl_plane_max_scale(dev_priv, fb);
 	}
 
 	ret = drm_atomic_helper_check_plane_state(&plane_state->base,
@@ -1992,8 +2373,10 @@ static const u64 i9xx_plane_format_modifiers[] = {
 };
 
 static const u32 snb_plane_formats[] = {
-	DRM_FORMAT_XBGR8888,
 	DRM_FORMAT_XRGB8888,
+	DRM_FORMAT_XBGR8888,
+	DRM_FORMAT_XRGB16161616F,
+	DRM_FORMAT_XBGR16161616F,
 	DRM_FORMAT_YUYV,
 	DRM_FORMAT_YVYU,
 	DRM_FORMAT_UYVY,
@@ -2023,6 +2406,8 @@ static const u32 skl_plane_formats[] = {
 	DRM_FORMAT_ABGR8888,
 	DRM_FORMAT_XRGB2101010,
 	DRM_FORMAT_XBGR2101010,
+	DRM_FORMAT_XRGB16161616F,
+	DRM_FORMAT_XBGR16161616F,
 	DRM_FORMAT_YUYV,
 	DRM_FORMAT_YVYU,
 	DRM_FORMAT_UYVY,
@@ -2038,6 +2423,8 @@ static const u32 skl_planar_formats[] = {
 	DRM_FORMAT_ABGR8888,
 	DRM_FORMAT_XRGB2101010,
 	DRM_FORMAT_XBGR2101010,
+	DRM_FORMAT_XRGB16161616F,
+	DRM_FORMAT_XBGR16161616F,
 	DRM_FORMAT_YUYV,
 	DRM_FORMAT_YVYU,
 	DRM_FORMAT_UYVY,
@@ -2054,6 +2441,8 @@ static const u32 glk_planar_formats[] = {
 	DRM_FORMAT_ABGR8888,
 	DRM_FORMAT_XRGB2101010,
 	DRM_FORMAT_XBGR2101010,
+	DRM_FORMAT_XRGB16161616F,
+	DRM_FORMAT_XBGR16161616F,
 	DRM_FORMAT_YUYV,
 	DRM_FORMAT_YVYU,
 	DRM_FORMAT_UYVY,
@@ -2157,6 +2546,13 @@ static const u64 skl_plane_format_modifiers_ccs[] = {
 	DRM_FORMAT_MOD_INVALID
 };
 
+static const u64 gen12_plane_format_modifiers_noccs[] = {
+	I915_FORMAT_MOD_Y_TILED,
+	I915_FORMAT_MOD_X_TILED,
+	DRM_FORMAT_MOD_LINEAR,
+	DRM_FORMAT_MOD_INVALID
+};
+
 static bool g4x_sprite_format_mod_supported(struct drm_plane *_plane,
 					    u32 format, u64 modifier)
 {
@@ -2197,6 +2593,8 @@ static bool snb_sprite_format_mod_supported(struct drm_plane *_plane,
 	switch (format) {
 	case DRM_FORMAT_XRGB8888:
 	case DRM_FORMAT_XBGR8888:
+	case DRM_FORMAT_XRGB16161616F:
+	case DRM_FORMAT_XBGR16161616F:
 	case DRM_FORMAT_YUYV:
 	case DRM_FORMAT_YVYU:
 	case DRM_FORMAT_UYVY:
@@ -2305,6 +2703,55 @@ static bool skl_plane_format_mod_supported(struct drm_plane *_plane,
 	}
 }
 
+static bool gen12_plane_format_mod_supported(struct drm_plane *_plane,
+					     u32 format, u64 modifier)
+{
+	switch (modifier) {
+	case DRM_FORMAT_MOD_LINEAR:
+	case I915_FORMAT_MOD_X_TILED:
+	case I915_FORMAT_MOD_Y_TILED:
+		break;
+	default:
+		return false;
+	}
+
+	switch (format) {
+	case DRM_FORMAT_XRGB8888:
+	case DRM_FORMAT_XBGR8888:
+	case DRM_FORMAT_ARGB8888:
+	case DRM_FORMAT_ABGR8888:
+	case DRM_FORMAT_RGB565:
+	case DRM_FORMAT_XRGB2101010:
+	case DRM_FORMAT_XBGR2101010:
+	case DRM_FORMAT_YUYV:
+	case DRM_FORMAT_YVYU:
+	case DRM_FORMAT_UYVY:
+	case DRM_FORMAT_VYUY:
+	case DRM_FORMAT_NV12:
+	case DRM_FORMAT_P010:
+	case DRM_FORMAT_P012:
+	case DRM_FORMAT_P016:
+	case DRM_FORMAT_XVYU2101010:
+	case DRM_FORMAT_C8:
+	case DRM_FORMAT_XBGR16161616F:
+	case DRM_FORMAT_ABGR16161616F:
+	case DRM_FORMAT_XRGB16161616F:
+	case DRM_FORMAT_ARGB16161616F:
+	case DRM_FORMAT_Y210:
+	case DRM_FORMAT_Y212:
+	case DRM_FORMAT_Y216:
+	case DRM_FORMAT_XVYU12_16161616:
+	case DRM_FORMAT_XVYU16161616:
+		if (modifier == DRM_FORMAT_MOD_LINEAR ||
+		    modifier == I915_FORMAT_MOD_X_TILED ||
+		    modifier == I915_FORMAT_MOD_Y_TILED)
+			return true;
+		/* fall through */
+	default:
+		return false;
+	}
+}
+
 static const struct drm_plane_funcs g4x_sprite_funcs = {
 	.update_plane = drm_atomic_helper_update_plane,
 	.disable_plane = drm_atomic_helper_disable_plane,
@@ -2341,6 +2788,15 @@ static const struct drm_plane_funcs skl_plane_funcs = {
 	.format_mod_supported = skl_plane_format_mod_supported,
 };
 
+static const struct drm_plane_funcs gen12_plane_funcs = {
+	.update_plane = drm_atomic_helper_update_plane,
+	.disable_plane = drm_atomic_helper_disable_plane,
+	.destroy = intel_plane_destroy,
+	.atomic_duplicate_state = intel_plane_duplicate_state,
+	.atomic_destroy_state = intel_plane_destroy_state,
+	.format_mod_supported = gen12_plane_format_mod_supported,
+};
+
 static bool skl_plane_has_fbc(struct drm_i915_private *dev_priv,
 			      enum pipe pipe, enum plane_id plane_id)
 {
@@ -2429,6 +2885,7 @@ struct intel_plane *
 skl_universal_plane_create(struct drm_i915_private *dev_priv,
 			   enum pipe pipe, enum plane_id plane_id)
 {
+	const struct drm_plane_funcs *plane_funcs;
 	struct intel_plane *plane;
 	enum drm_plane_type plane_type;
 	unsigned int supported_rotations;
@@ -2458,6 +2915,7 @@ skl_universal_plane_create(struct drm_i915_private *dev_priv,
 	plane->disable_plane = skl_disable_plane;
 	plane->get_hw_state = skl_plane_get_hw_state;
 	plane->check_plane = skl_plane_check;
+	plane->min_cdclk = skl_plane_min_cdclk;
 	if (icl_is_nv12_y_plane(plane_id))
 		plane->update_slave = icl_update_slave;
 
@@ -2471,11 +2929,19 @@ skl_universal_plane_create(struct drm_i915_private *dev_priv,
 		formats = skl_get_plane_formats(dev_priv, pipe,
 						plane_id, &num_formats);
 
-	plane->has_ccs = skl_plane_has_ccs(dev_priv, pipe, plane_id);
-	if (plane->has_ccs)
-		modifiers = skl_plane_format_modifiers_ccs;
-	else
-		modifiers = skl_plane_format_modifiers_noccs;
+	if (INTEL_GEN(dev_priv) >= 12) {
+		/* TODO: Implement support for gen-12 CCS modifiers */
+		plane->has_ccs = false;
+		modifiers = gen12_plane_format_modifiers_noccs;
+		plane_funcs = &gen12_plane_funcs;
+	} else {
+		plane->has_ccs = skl_plane_has_ccs(dev_priv, pipe, plane_id);
+		if (plane->has_ccs)
+			modifiers = skl_plane_format_modifiers_ccs;
+		else
+			modifiers = skl_plane_format_modifiers_noccs;
+		plane_funcs = &skl_plane_funcs;
+	}
 
 	if (plane_id == PLANE_PRIMARY)
 		plane_type = DRM_PLANE_TYPE_PRIMARY;
@@ -2485,7 +2951,7 @@ skl_universal_plane_create(struct drm_i915_private *dev_priv,
 	possible_crtcs = BIT(pipe);
 
 	ret = drm_universal_plane_init(&dev_priv->drm, &plane->base,
-				       possible_crtcs, &skl_plane_funcs,
+				       possible_crtcs, plane_funcs,
 				       formats, num_formats, modifiers,
 				       plane_type,
 				       "plane %d%c", plane_id + 1,
@@ -2518,6 +2984,8 @@ skl_universal_plane_create(struct drm_i915_private *dev_priv,
 					     BIT(DRM_MODE_BLEND_PREMULTI) |
 					     BIT(DRM_MODE_BLEND_COVERAGE));
 
+	drm_plane_create_zpos_immutable_property(&plane->base, plane_id);
+
 	drm_plane_helper_add(&plane->base, &intel_plane_helper_funcs);
 
 	return plane;
@@ -2539,7 +3007,7 @@ intel_sprite_plane_create(struct drm_i915_private *dev_priv,
 	const u64 *modifiers;
 	const u32 *formats;
 	int num_formats;
-	int ret;
+	int ret, zpos;
 
 	if (INTEL_GEN(dev_priv) >= 9)
 		return skl_universal_plane_create(dev_priv, pipe,
@@ -2555,6 +3023,7 @@ intel_sprite_plane_create(struct drm_i915_private *dev_priv,
 		plane->disable_plane = vlv_disable_plane;
 		plane->get_hw_state = vlv_plane_get_hw_state;
 		plane->check_plane = vlv_sprite_check;
+		plane->min_cdclk = vlv_plane_min_cdclk;
 
 		formats = vlv_plane_formats;
 		num_formats = ARRAY_SIZE(vlv_plane_formats);
@@ -2568,6 +3037,11 @@ intel_sprite_plane_create(struct drm_i915_private *dev_priv,
 		plane->get_hw_state = ivb_plane_get_hw_state;
 		plane->check_plane = g4x_sprite_check;
 
+		if (IS_BROADWELL(dev_priv) || IS_HASWELL(dev_priv))
+			plane->min_cdclk = hsw_plane_min_cdclk;
+		else
+			plane->min_cdclk = ivb_sprite_min_cdclk;
+
 		formats = snb_plane_formats;
 		num_formats = ARRAY_SIZE(snb_plane_formats);
 		modifiers = i9xx_plane_format_modifiers;
@@ -2579,6 +3053,7 @@ intel_sprite_plane_create(struct drm_i915_private *dev_priv,
 		plane->disable_plane = g4x_disable_plane;
 		plane->get_hw_state = g4x_plane_get_hw_state;
 		plane->check_plane = g4x_sprite_check;
+		plane->min_cdclk = g4x_sprite_min_cdclk;
 
 		modifiers = i9xx_plane_format_modifiers;
 		if (IS_GEN(dev_priv, 6)) {
@@ -2629,6 +3104,9 @@ intel_sprite_plane_create(struct drm_i915_private *dev_priv,
 					  DRM_COLOR_YCBCR_BT709,
 					  DRM_COLOR_YCBCR_LIMITED_RANGE);
 
+	zpos = sprite + 1;
+	drm_plane_create_zpos_immutable_property(&plane->base, zpos);
+
 	drm_plane_helper_add(&plane->base, &intel_plane_helper_funcs);
 
 	return plane;
diff --git a/drivers/gpu/drm/i915/display/intel_sprite.h b/drivers/gpu/drm/i915/display/intel_sprite.h
index 093a2d156f1e..5eeaa92420d1 100644
--- a/drivers/gpu/drm/i915/display/intel_sprite.h
+++ b/drivers/gpu/drm/i915/display/intel_sprite.h
@@ -17,7 +17,6 @@ struct drm_i915_private;
 struct intel_crtc_state;
 struct intel_plane_state;
 
-bool is_planar_yuv_format(u32 pixelformat);
 int intel_usecs_to_scanlines(const struct drm_display_mode *adjusted_mode,
 			     int usecs);
 struct intel_plane *intel_sprite_plane_create(struct drm_i915_private *dev_priv,
@@ -50,4 +49,11 @@ static inline u8 icl_hdr_plane_mask(void)
 
 bool icl_is_hdr_plane(struct drm_i915_private *dev_priv, enum plane_id plane_id);
 
+int ivb_plane_min_cdclk(const struct intel_crtc_state *crtc_state,
+			const struct intel_plane_state *plane_state);
+int hsw_plane_min_cdclk(const struct intel_crtc_state *crtc_state,
+			const struct intel_plane_state *plane_state);
+int vlv_plane_min_cdclk(const struct intel_crtc_state *crtc_state,
+			const struct intel_plane_state *plane_state);
+
 #endif /* __INTEL_SPRITE_H__ */
diff --git a/drivers/gpu/drm/i915/display/intel_tc.c b/drivers/gpu/drm/i915/display/intel_tc.c
index 85743a43bee2..7773169b7331 100644
--- a/drivers/gpu/drm/i915/display/intel_tc.c
+++ b/drivers/gpu/drm/i915/display/intel_tc.c
@@ -23,32 +23,38 @@ static const char *tc_port_mode_name(enum tc_port_mode mode)
 	return names[mode];
 }
 
-static bool has_modular_fia(struct drm_i915_private *i915)
-{
-	if (!INTEL_INFO(i915)->display.has_modular_fia)
-		return false;
-
-	return intel_uncore_read(&i915->uncore,
-				 PORT_TX_DFLEXDPSP(FIA1)) & MODULAR_FIA_MASK;
-}
-
-static enum phy_fia tc_port_to_fia(struct drm_i915_private *i915,
-				   enum tc_port tc_port)
+static void
+tc_port_load_fia_params(struct drm_i915_private *i915,
+			struct intel_digital_port *dig_port)
 {
-	if (!has_modular_fia(i915))
-		return FIA1;
+	enum port port = dig_port->base.port;
+	enum tc_port tc_port = intel_port_to_tc(i915, port);
+	u32 modular_fia;
+
+	if (INTEL_INFO(i915)->display.has_modular_fia) {
+		modular_fia = intel_uncore_read(&i915->uncore,
+						PORT_TX_DFLEXDPSP(FIA1));
+		modular_fia &= MODULAR_FIA_MASK;
+	} else {
+		modular_fia = 0;
+	}
 
 	/*
 	 * Each Modular FIA instance houses 2 TC ports. In SOC that has more
 	 * than two TC ports, there are multiple instances of Modular FIA.
 	 */
-	return tc_port / 2;
+	if (modular_fia) {
+		dig_port->tc_phy_fia = tc_port / 2;
+		dig_port->tc_phy_fia_idx = tc_port % 2;
+	} else {
+		dig_port->tc_phy_fia = FIA1;
+		dig_port->tc_phy_fia_idx = tc_port;
+	}
 }
 
 u32 intel_tc_port_get_lane_mask(struct intel_digital_port *dig_port)
 {
 	struct drm_i915_private *i915 = to_i915(dig_port->base.base.dev);
-	enum tc_port tc_port = intel_port_to_tc(i915, dig_port->base.port);
 	struct intel_uncore *uncore = &i915->uncore;
 	u32 lane_mask;
 
@@ -57,8 +63,23 @@ u32 intel_tc_port_get_lane_mask(struct intel_digital_port *dig_port)
 
 	WARN_ON(lane_mask == 0xffffffff);
 
-	return (lane_mask & DP_LANE_ASSIGNMENT_MASK(tc_port)) >>
-	       DP_LANE_ASSIGNMENT_SHIFT(tc_port);
+	lane_mask &= DP_LANE_ASSIGNMENT_MASK(dig_port->tc_phy_fia_idx);
+	return lane_mask >> DP_LANE_ASSIGNMENT_SHIFT(dig_port->tc_phy_fia_idx);
+}
+
+u32 intel_tc_port_get_pin_assignment_mask(struct intel_digital_port *dig_port)
+{
+	struct drm_i915_private *i915 = to_i915(dig_port->base.base.dev);
+	struct intel_uncore *uncore = &i915->uncore;
+	u32 pin_mask;
+
+	pin_mask = intel_uncore_read(uncore,
+				     PORT_TX_DFLEXPA1(dig_port->tc_phy_fia));
+
+	WARN_ON(pin_mask == 0xffffffff);
+
+	return (pin_mask & DP_PIN_ASSIGNMENT_MASK(dig_port->tc_phy_fia_idx)) >>
+	       DP_PIN_ASSIGNMENT_SHIFT(dig_port->tc_phy_fia_idx);
 }
 
 int intel_tc_port_fia_max_lane_count(struct intel_digital_port *dig_port)
@@ -95,7 +116,6 @@ void intel_tc_port_set_fia_lane_count(struct intel_digital_port *dig_port,
 				      int required_lanes)
 {
 	struct drm_i915_private *i915 = to_i915(dig_port->base.base.dev);
-	enum tc_port tc_port = intel_port_to_tc(i915, dig_port->base.port);
 	bool lane_reversal = dig_port->saved_port_bits & DDI_BUF_PORT_REVERSAL;
 	struct intel_uncore *uncore = &i915->uncore;
 	u32 val;
@@ -104,19 +124,21 @@ void intel_tc_port_set_fia_lane_count(struct intel_digital_port *dig_port,
 
 	val = intel_uncore_read(uncore,
 				PORT_TX_DFLEXDPMLE1(dig_port->tc_phy_fia));
-	val &= ~DFLEXDPMLE1_DPMLETC_MASK(tc_port);
+	val &= ~DFLEXDPMLE1_DPMLETC_MASK(dig_port->tc_phy_fia_idx);
 
 	switch (required_lanes) {
 	case 1:
-		val |= lane_reversal ? DFLEXDPMLE1_DPMLETC_ML3(tc_port) :
-			DFLEXDPMLE1_DPMLETC_ML0(tc_port);
+		val |= lane_reversal ?
+			DFLEXDPMLE1_DPMLETC_ML3(dig_port->tc_phy_fia_idx) :
+			DFLEXDPMLE1_DPMLETC_ML0(dig_port->tc_phy_fia_idx);
 		break;
 	case 2:
-		val |= lane_reversal ? DFLEXDPMLE1_DPMLETC_ML3_2(tc_port) :
-			DFLEXDPMLE1_DPMLETC_ML1_0(tc_port);
+		val |= lane_reversal ?
+			DFLEXDPMLE1_DPMLETC_ML3_2(dig_port->tc_phy_fia_idx) :
+			DFLEXDPMLE1_DPMLETC_ML1_0(dig_port->tc_phy_fia_idx);
 		break;
 	case 4:
-		val |= DFLEXDPMLE1_DPMLETC_ML3_0(tc_port);
+		val |= DFLEXDPMLE1_DPMLETC_ML3_0(dig_port->tc_phy_fia_idx);
 		break;
 	default:
 		MISSING_CASE(required_lanes);
@@ -164,9 +186,9 @@ static u32 tc_port_live_status_mask(struct intel_digital_port *dig_port)
 		return mask;
 	}
 
-	if (val & TC_LIVE_STATE_TBT(tc_port))
+	if (val & TC_LIVE_STATE_TBT(dig_port->tc_phy_fia_idx))
 		mask |= BIT(TC_PORT_TBT_ALT);
-	if (val & TC_LIVE_STATE_TC(tc_port))
+	if (val & TC_LIVE_STATE_TC(dig_port->tc_phy_fia_idx))
 		mask |= BIT(TC_PORT_DP_ALT);
 
 	if (intel_uncore_read(uncore, SDEISR) & SDE_TC_HOTPLUG_ICP(tc_port))
@@ -182,7 +204,6 @@ static u32 tc_port_live_status_mask(struct intel_digital_port *dig_port)
 static bool icl_tc_phy_status_complete(struct intel_digital_port *dig_port)
 {
 	struct drm_i915_private *i915 = to_i915(dig_port->base.base.dev);
-	enum tc_port tc_port = intel_port_to_tc(i915, dig_port->base.port);
 	struct intel_uncore *uncore = &i915->uncore;
 	u32 val;
 
@@ -194,14 +215,13 @@ static bool icl_tc_phy_status_complete(struct intel_digital_port *dig_port)
 		return false;
 	}
 
-	return val & DP_PHY_MODE_STATUS_COMPLETED(tc_port);
+	return val & DP_PHY_MODE_STATUS_COMPLETED(dig_port->tc_phy_fia_idx);
 }
 
 static bool icl_tc_phy_set_safe_mode(struct intel_digital_port *dig_port,
 				     bool enable)
 {
 	struct drm_i915_private *i915 = to_i915(dig_port->base.base.dev);
-	enum tc_port tc_port = intel_port_to_tc(i915, dig_port->base.port);
 	struct intel_uncore *uncore = &i915->uncore;
 	u32 val;
 
@@ -215,9 +235,9 @@ static bool icl_tc_phy_set_safe_mode(struct intel_digital_port *dig_port,
 		return false;
 	}
 
-	val &= ~DP_PHY_MODE_STATUS_NOT_SAFE(tc_port);
+	val &= ~DP_PHY_MODE_STATUS_NOT_SAFE(dig_port->tc_phy_fia_idx);
 	if (!enable)
-		val |= DP_PHY_MODE_STATUS_NOT_SAFE(tc_port);
+		val |= DP_PHY_MODE_STATUS_NOT_SAFE(dig_port->tc_phy_fia_idx);
 
 	intel_uncore_write(uncore,
 			   PORT_TX_DFLEXDPCSSS(dig_port->tc_phy_fia), val);
@@ -232,7 +252,6 @@ static bool icl_tc_phy_set_safe_mode(struct intel_digital_port *dig_port,
 static bool icl_tc_phy_is_in_safe_mode(struct intel_digital_port *dig_port)
 {
 	struct drm_i915_private *i915 = to_i915(dig_port->base.base.dev);
-	enum tc_port tc_port = intel_port_to_tc(i915, dig_port->base.port);
 	struct intel_uncore *uncore = &i915->uncore;
 	u32 val;
 
@@ -244,7 +263,7 @@ static bool icl_tc_phy_is_in_safe_mode(struct intel_digital_port *dig_port)
 		return true;
 	}
 
-	return !(val & DP_PHY_MODE_STATUS_NOT_SAFE(tc_port));
+	return !(val & DP_PHY_MODE_STATUS_NOT_SAFE(dig_port->tc_phy_fia_idx));
 }
 
 /*
@@ -540,5 +559,5 @@ void intel_tc_port_init(struct intel_digital_port *dig_port, bool is_legacy)
 	mutex_init(&dig_port->tc_lock);
 	dig_port->tc_legacy_port = is_legacy;
 	dig_port->tc_link_refcount = 0;
-	dig_port->tc_phy_fia = tc_port_to_fia(i915, tc_port);
+	tc_port_load_fia_params(i915, dig_port);
 }
diff --git a/drivers/gpu/drm/i915/display/intel_tc.h b/drivers/gpu/drm/i915/display/intel_tc.h
index 783d75531435..463f1b3c836f 100644
--- a/drivers/gpu/drm/i915/display/intel_tc.h
+++ b/drivers/gpu/drm/i915/display/intel_tc.h
@@ -13,6 +13,7 @@ struct intel_digital_port;
 
 bool intel_tc_port_connected(struct intel_digital_port *dig_port);
 u32 intel_tc_port_get_lane_mask(struct intel_digital_port *dig_port);
+u32 intel_tc_port_get_pin_assignment_mask(struct intel_digital_port *dig_port);
 int intel_tc_port_fia_max_lane_count(struct intel_digital_port *dig_port);
 void intel_tc_port_set_fia_lane_count(struct intel_digital_port *dig_port,
 				      int required_lanes);
diff --git a/drivers/gpu/drm/i915/display/intel_tv.c b/drivers/gpu/drm/i915/display/intel_tv.c
index b70221f5112a..9983fadf6c28 100644
--- a/drivers/gpu/drm/i915/display/intel_tv.c
+++ b/drivers/gpu/drm/i915/display/intel_tv.c
@@ -961,11 +961,10 @@ intel_tv_mode_valid(struct drm_connector *connector,
 		return MODE_CLOCK_HIGH;
 
 	/* Ensure TV refresh is close to desired refresh */
-	if (tv_mode && abs(tv_mode->refresh - drm_mode_vrefresh(mode) * 1000)
-				< 1000)
-		return MODE_OK;
+	if (abs(tv_mode->refresh - drm_mode_vrefresh(mode) * 1000) >= 1000)
+		return MODE_CLOCK_RANGE;
 
-	return MODE_CLOCK_RANGE;
+	return MODE_OK;
 }
 
 static int
@@ -1702,7 +1701,7 @@ intel_tv_detect(struct drm_connector *connector,
 		struct intel_load_detect_pipe tmp;
 		int ret;
 
-		ret = intel_get_load_detect_pipe(connector, NULL, &tmp, ctx);
+		ret = intel_get_load_detect_pipe(connector, &tmp, ctx);
 		if (ret < 0)
 			return ret;
 
@@ -1948,9 +1947,8 @@ intel_tv_init(struct drm_i915_private *dev_priv)
 	intel_encoder->type = INTEL_OUTPUT_TVOUT;
 	intel_encoder->power_domain = POWER_DOMAIN_PORT_OTHER;
 	intel_encoder->port = PORT_NONE;
-	intel_encoder->crtc_mask = (1 << 0) | (1 << 1);
+	intel_encoder->pipe_mask = ~0;
 	intel_encoder->cloneable = 0;
-	intel_encoder->base.possible_crtcs = ((1 << 0) | (1 << 1));
 	intel_tv->type = DRM_MODE_CONNECTOR_Unknown;
 
 	/* BIOS margin values */
diff --git a/drivers/gpu/drm/i915/display/intel_vbt_defs.h b/drivers/gpu/drm/i915/display/intel_vbt_defs.h
index dfcd156b5094..69a7cb1fa121 100644
--- a/drivers/gpu/drm/i915/display/intel_vbt_defs.h
+++ b/drivers/gpu/drm/i915/display/intel_vbt_defs.h
@@ -114,6 +114,7 @@ enum bdb_block_id {
 	BDB_LVDS_POWER			= 44,
 	BDB_MIPI_CONFIG			= 52,
 	BDB_MIPI_SEQUENCE		= 53,
+	BDB_COMPRESSION_PARAMETERS	= 56,
 	BDB_SKIP			= 254, /* VBIOS private block, ignore */
 };
 
@@ -291,6 +292,8 @@ struct bdb_general_features {
 #define DVO_PORT_HDMIE		12				/* 193 */
 #define DVO_PORT_DPF		13				/* N/A */
 #define DVO_PORT_HDMIF		14				/* N/A */
+#define DVO_PORT_DPG		15
+#define DVO_PORT_HDMIG		16
 #define DVO_PORT_MIPIA		21				/* 171 */
 #define DVO_PORT_MIPIB		22				/* 171 */
 #define DVO_PORT_MIPIC		23				/* 171 */
@@ -325,6 +328,7 @@ enum vbt_gmbus_ddi {
 #define DP_AUX_D 0x30
 #define DP_AUX_E 0x50
 #define DP_AUX_F 0x60
+#define DP_AUX_G 0x70
 
 #define VBT_DP_MAX_LINK_RATE_HBR3	0
 #define VBT_DP_MAX_LINK_RATE_HBR2	1
@@ -808,4 +812,55 @@ struct bdb_mipi_sequence {
 	u8 data[0]; /* up to 6 variable length blocks */
 } __packed;
 
+/*
+ * Block 56 - Compression Parameters
+ */
+
+#define VBT_RC_BUFFER_BLOCK_SIZE_1KB	0
+#define VBT_RC_BUFFER_BLOCK_SIZE_4KB	1
+#define VBT_RC_BUFFER_BLOCK_SIZE_16KB	2
+#define VBT_RC_BUFFER_BLOCK_SIZE_64KB	3
+
+#define VBT_DSC_LINE_BUFFER_DEPTH(vbt_value)	((vbt_value) + 8) /* bits */
+#define VBT_DSC_MAX_BPP(vbt_value)		(6 + (vbt_value) * 2)
+
+struct dsc_compression_parameters_entry {
+	u8 version_major:4;
+	u8 version_minor:4;
+
+	u8 rc_buffer_block_size:2;
+	u8 reserved1:6;
+
+	/*
+	 * Buffer size in bytes:
+	 *
+	 * 4 ^ rc_buffer_block_size * 1024 * (rc_buffer_size + 1) bytes
+	 */
+	u8 rc_buffer_size;
+	u32 slices_per_line;
+
+	u8 line_buffer_depth:4;
+	u8 reserved2:4;
+
+	/* Flag Bits 1 */
+	u8 block_prediction_enable:1;
+	u8 reserved3:7;
+
+	u8 max_bpp; /* mapping */
+
+	/* Color depth capabilities */
+	u8 reserved4:1;
+	u8 support_8bpc:1;
+	u8 support_10bpc:1;
+	u8 support_12bpc:1;
+	u8 reserved5:4;
+
+	u16 slice_height;
+} __packed;
+
+struct bdb_compression_parameters {
+	u16 entry_size;
+	struct dsc_compression_parameters_entry data[16];
+} __packed;
+
 #endif /* _INTEL_VBT_DEFS_H_ */
diff --git a/drivers/gpu/drm/i915/display/intel_vdsc.c b/drivers/gpu/drm/i915/display/intel_vdsc.c
index d4fb7f16f9f6..896b0c334f5e 100644
--- a/drivers/gpu/drm/i915/display/intel_vdsc.c
+++ b/drivers/gpu/drm/i915/display/intel_vdsc.c
@@ -322,8 +322,8 @@ static int get_column_index_for_rc_params(u8 bits_per_component)
 int intel_dp_compute_dsc_params(struct intel_dp *intel_dp,
 				struct intel_crtc_state *pipe_config)
 {
-	struct drm_dsc_config *vdsc_cfg = &pipe_config->dp_dsc_cfg;
-	u16 compressed_bpp = pipe_config->dsc_params.compressed_bpp;
+	struct drm_dsc_config *vdsc_cfg = &pipe_config->dsc.config;
+	u16 compressed_bpp = pipe_config->dsc.compressed_bpp;
 	u8 i = 0;
 	int row_index = 0;
 	int column_index = 0;
@@ -332,7 +332,7 @@ int intel_dp_compute_dsc_params(struct intel_dp *intel_dp,
 	vdsc_cfg->pic_width = pipe_config->base.adjusted_mode.crtc_hdisplay;
 	vdsc_cfg->pic_height = pipe_config->base.adjusted_mode.crtc_vdisplay;
 	vdsc_cfg->slice_width = DIV_ROUND_UP(vdsc_cfg->pic_width,
-					     pipe_config->dsc_params.slice_count);
+					     pipe_config->dsc.slice_count);
 	/*
 	 * Slice Height of 8 works for all currently available panels. So start
 	 * with that if pic_height is an integral multiple of 8.
@@ -485,13 +485,13 @@ static void intel_configure_pps_for_dsc_encoder(struct intel_encoder *encoder,
 {
 	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	const struct drm_dsc_config *vdsc_cfg = &crtc_state->dp_dsc_cfg;
+	const struct drm_dsc_config *vdsc_cfg = &crtc_state->dsc.config;
 	enum pipe pipe = crtc->pipe;
 	enum transcoder cpu_transcoder = crtc_state->cpu_transcoder;
 	u32 pps_val = 0;
 	u32 rc_buf_thresh_dword[4];
 	u32 rc_range_params_dword[8];
-	u8 num_vdsc_instances = (crtc_state->dsc_params.dsc_split) ? 2 : 1;
+	u8 num_vdsc_instances = (crtc_state->dsc.dsc_split) ? 2 : 1;
 	int i = 0;
 
 	/* Populate PICTURE_PARAMETER_SET_0 registers */
@@ -514,11 +514,11 @@ static void intel_configure_pps_for_dsc_encoder(struct intel_encoder *encoder,
 		 * If 2 VDSC instances are needed, configure PPS for second
 		 * VDSC
 		 */
-		if (crtc_state->dsc_params.dsc_split)
+		if (crtc_state->dsc.dsc_split)
 			I915_WRITE(DSCC_PICTURE_PARAMETER_SET_0, pps_val);
 	} else {
 		I915_WRITE(ICL_DSC0_PICTURE_PARAMETER_SET_0(pipe), pps_val);
-		if (crtc_state->dsc_params.dsc_split)
+		if (crtc_state->dsc.dsc_split)
 			I915_WRITE(ICL_DSC1_PICTURE_PARAMETER_SET_0(pipe),
 				   pps_val);
 	}
@@ -533,11 +533,11 @@ static void intel_configure_pps_for_dsc_encoder(struct intel_encoder *encoder,
 		 * If 2 VDSC instances are needed, configure PPS for second
 		 * VDSC
 		 */
-		if (crtc_state->dsc_params.dsc_split)
+		if (crtc_state->dsc.dsc_split)
 			I915_WRITE(DSCC_PICTURE_PARAMETER_SET_1, pps_val);
 	} else {
 		I915_WRITE(ICL_DSC0_PICTURE_PARAMETER_SET_1(pipe), pps_val);
-		if (crtc_state->dsc_params.dsc_split)
+		if (crtc_state->dsc.dsc_split)
 			I915_WRITE(ICL_DSC1_PICTURE_PARAMETER_SET_1(pipe),
 				   pps_val);
 	}
@@ -553,11 +553,11 @@ static void intel_configure_pps_for_dsc_encoder(struct intel_encoder *encoder,
 		 * If 2 VDSC instances are needed, configure PPS for second
 		 * VDSC
 		 */
-		if (crtc_state->dsc_params.dsc_split)
+		if (crtc_state->dsc.dsc_split)
 			I915_WRITE(DSCC_PICTURE_PARAMETER_SET_2, pps_val);
 	} else {
 		I915_WRITE(ICL_DSC0_PICTURE_PARAMETER_SET_2(pipe), pps_val);
-		if (crtc_state->dsc_params.dsc_split)
+		if (crtc_state->dsc.dsc_split)
 			I915_WRITE(ICL_DSC1_PICTURE_PARAMETER_SET_2(pipe),
 				   pps_val);
 	}
@@ -573,11 +573,11 @@ static void intel_configure_pps_for_dsc_encoder(struct intel_encoder *encoder,
 		 * If 2 VDSC instances are needed, configure PPS for second
 		 * VDSC
 		 */
-		if (crtc_state->dsc_params.dsc_split)
+		if (crtc_state->dsc.dsc_split)
 			I915_WRITE(DSCC_PICTURE_PARAMETER_SET_3, pps_val);
 	} else {
 		I915_WRITE(ICL_DSC0_PICTURE_PARAMETER_SET_3(pipe), pps_val);
-		if (crtc_state->dsc_params.dsc_split)
+		if (crtc_state->dsc.dsc_split)
 			I915_WRITE(ICL_DSC1_PICTURE_PARAMETER_SET_3(pipe),
 				   pps_val);
 	}
@@ -593,11 +593,11 @@ static void intel_configure_pps_for_dsc_encoder(struct intel_encoder *encoder,
 		 * If 2 VDSC instances are needed, configure PPS for second
 		 * VDSC
 		 */
-		if (crtc_state->dsc_params.dsc_split)
+		if (crtc_state->dsc.dsc_split)
 			I915_WRITE(DSCC_PICTURE_PARAMETER_SET_4, pps_val);
 	} else {
 		I915_WRITE(ICL_DSC0_PICTURE_PARAMETER_SET_4(pipe), pps_val);
-		if (crtc_state->dsc_params.dsc_split)
+		if (crtc_state->dsc.dsc_split)
 			I915_WRITE(ICL_DSC1_PICTURE_PARAMETER_SET_4(pipe),
 				   pps_val);
 	}
@@ -613,11 +613,11 @@ static void intel_configure_pps_for_dsc_encoder(struct intel_encoder *encoder,
 		 * If 2 VDSC instances are needed, configure PPS for second
 		 * VDSC
 		 */
-		if (crtc_state->dsc_params.dsc_split)
+		if (crtc_state->dsc.dsc_split)
 			I915_WRITE(DSCC_PICTURE_PARAMETER_SET_5, pps_val);
 	} else {
 		I915_WRITE(ICL_DSC0_PICTURE_PARAMETER_SET_5(pipe), pps_val);
-		if (crtc_state->dsc_params.dsc_split)
+		if (crtc_state->dsc.dsc_split)
 			I915_WRITE(ICL_DSC1_PICTURE_PARAMETER_SET_5(pipe),
 				   pps_val);
 	}
@@ -635,11 +635,11 @@ static void intel_configure_pps_for_dsc_encoder(struct intel_encoder *encoder,
 		 * If 2 VDSC instances are needed, configure PPS for second
 		 * VDSC
 		 */
-		if (crtc_state->dsc_params.dsc_split)
+		if (crtc_state->dsc.dsc_split)
 			I915_WRITE(DSCC_PICTURE_PARAMETER_SET_6, pps_val);
 	} else {
 		I915_WRITE(ICL_DSC0_PICTURE_PARAMETER_SET_6(pipe), pps_val);
-		if (crtc_state->dsc_params.dsc_split)
+		if (crtc_state->dsc.dsc_split)
 			I915_WRITE(ICL_DSC1_PICTURE_PARAMETER_SET_6(pipe),
 				   pps_val);
 	}
@@ -655,11 +655,11 @@ static void intel_configure_pps_for_dsc_encoder(struct intel_encoder *encoder,
 		 * If 2 VDSC instances are needed, configure PPS for second
 		 * VDSC
 		 */
-		if (crtc_state->dsc_params.dsc_split)
+		if (crtc_state->dsc.dsc_split)
 			I915_WRITE(DSCC_PICTURE_PARAMETER_SET_7, pps_val);
 	} else {
 		I915_WRITE(ICL_DSC0_PICTURE_PARAMETER_SET_7(pipe), pps_val);
-		if (crtc_state->dsc_params.dsc_split)
+		if (crtc_state->dsc.dsc_split)
 			I915_WRITE(ICL_DSC1_PICTURE_PARAMETER_SET_7(pipe),
 				   pps_val);
 	}
@@ -675,11 +675,11 @@ static void intel_configure_pps_for_dsc_encoder(struct intel_encoder *encoder,
 		 * If 2 VDSC instances are needed, configure PPS for second
 		 * VDSC
 		 */
-		if (crtc_state->dsc_params.dsc_split)
+		if (crtc_state->dsc.dsc_split)
 			I915_WRITE(DSCC_PICTURE_PARAMETER_SET_8, pps_val);
 	} else {
 		I915_WRITE(ICL_DSC0_PICTURE_PARAMETER_SET_8(pipe), pps_val);
-		if (crtc_state->dsc_params.dsc_split)
+		if (crtc_state->dsc.dsc_split)
 			I915_WRITE(ICL_DSC1_PICTURE_PARAMETER_SET_8(pipe),
 				   pps_val);
 	}
@@ -695,11 +695,11 @@ static void intel_configure_pps_for_dsc_encoder(struct intel_encoder *encoder,
 		 * If 2 VDSC instances are needed, configure PPS for second
 		 * VDSC
 		 */
-		if (crtc_state->dsc_params.dsc_split)
+		if (crtc_state->dsc.dsc_split)
 			I915_WRITE(DSCC_PICTURE_PARAMETER_SET_9, pps_val);
 	} else {
 		I915_WRITE(ICL_DSC0_PICTURE_PARAMETER_SET_9(pipe), pps_val);
-		if (crtc_state->dsc_params.dsc_split)
+		if (crtc_state->dsc.dsc_split)
 			I915_WRITE(ICL_DSC1_PICTURE_PARAMETER_SET_9(pipe),
 				   pps_val);
 	}
@@ -717,11 +717,11 @@ static void intel_configure_pps_for_dsc_encoder(struct intel_encoder *encoder,
 		 * If 2 VDSC instances are needed, configure PPS for second
 		 * VDSC
 		 */
-		if (crtc_state->dsc_params.dsc_split)
+		if (crtc_state->dsc.dsc_split)
 			I915_WRITE(DSCC_PICTURE_PARAMETER_SET_10, pps_val);
 	} else {
 		I915_WRITE(ICL_DSC0_PICTURE_PARAMETER_SET_10(pipe), pps_val);
-		if (crtc_state->dsc_params.dsc_split)
+		if (crtc_state->dsc.dsc_split)
 			I915_WRITE(ICL_DSC1_PICTURE_PARAMETER_SET_10(pipe),
 				   pps_val);
 	}
@@ -740,11 +740,11 @@ static void intel_configure_pps_for_dsc_encoder(struct intel_encoder *encoder,
 		 * If 2 VDSC instances are needed, configure PPS for second
 		 * VDSC
 		 */
-		if (crtc_state->dsc_params.dsc_split)
+		if (crtc_state->dsc.dsc_split)
 			I915_WRITE(DSCC_PICTURE_PARAMETER_SET_16, pps_val);
 	} else {
 		I915_WRITE(ICL_DSC0_PICTURE_PARAMETER_SET_16(pipe), pps_val);
-		if (crtc_state->dsc_params.dsc_split)
+		if (crtc_state->dsc.dsc_split)
 			I915_WRITE(ICL_DSC1_PICTURE_PARAMETER_SET_16(pipe),
 				   pps_val);
 	}
@@ -763,7 +763,7 @@ static void intel_configure_pps_for_dsc_encoder(struct intel_encoder *encoder,
 		I915_WRITE(DSCA_RC_BUF_THRESH_0_UDW, rc_buf_thresh_dword[1]);
 		I915_WRITE(DSCA_RC_BUF_THRESH_1, rc_buf_thresh_dword[2]);
 		I915_WRITE(DSCA_RC_BUF_THRESH_1_UDW, rc_buf_thresh_dword[3]);
-		if (crtc_state->dsc_params.dsc_split) {
+		if (crtc_state->dsc.dsc_split) {
 			I915_WRITE(DSCC_RC_BUF_THRESH_0,
 				   rc_buf_thresh_dword[0]);
 			I915_WRITE(DSCC_RC_BUF_THRESH_0_UDW,
@@ -782,7 +782,7 @@ static void intel_configure_pps_for_dsc_encoder(struct intel_encoder *encoder,
 			   rc_buf_thresh_dword[2]);
 		I915_WRITE(ICL_DSC0_RC_BUF_THRESH_1_UDW(pipe),
 			   rc_buf_thresh_dword[3]);
-		if (crtc_state->dsc_params.dsc_split) {
+		if (crtc_state->dsc.dsc_split) {
 			I915_WRITE(ICL_DSC1_RC_BUF_THRESH_0(pipe),
 				   rc_buf_thresh_dword[0]);
 			I915_WRITE(ICL_DSC1_RC_BUF_THRESH_0_UDW(pipe),
@@ -824,7 +824,7 @@ static void intel_configure_pps_for_dsc_encoder(struct intel_encoder *encoder,
 			   rc_range_params_dword[6]);
 		I915_WRITE(DSCA_RC_RANGE_PARAMETERS_3_UDW,
 			   rc_range_params_dword[7]);
-		if (crtc_state->dsc_params.dsc_split) {
+		if (crtc_state->dsc.dsc_split) {
 			I915_WRITE(DSCC_RC_RANGE_PARAMETERS_0,
 				   rc_range_params_dword[0]);
 			I915_WRITE(DSCC_RC_RANGE_PARAMETERS_0_UDW,
@@ -859,7 +859,7 @@ static void intel_configure_pps_for_dsc_encoder(struct intel_encoder *encoder,
 			   rc_range_params_dword[6]);
 		I915_WRITE(ICL_DSC0_RC_RANGE_PARAMETERS_3_UDW(pipe),
 			   rc_range_params_dword[7]);
-		if (crtc_state->dsc_params.dsc_split) {
+		if (crtc_state->dsc.dsc_split) {
 			I915_WRITE(ICL_DSC1_RC_RANGE_PARAMETERS_0(pipe),
 				   rc_range_params_dword[0]);
 			I915_WRITE(ICL_DSC1_RC_RANGE_PARAMETERS_0_UDW(pipe),
@@ -885,7 +885,7 @@ static void intel_dp_write_dsc_pps_sdp(struct intel_encoder *encoder,
 {
 	struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
 	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
-	const struct drm_dsc_config *vdsc_cfg = &crtc_state->dp_dsc_cfg;
+	const struct drm_dsc_config *vdsc_cfg = &crtc_state->dsc.config;
 	struct drm_dsc_pps_infoframe dp_dsc_pps_sdp;
 
 	/* Prepare DP SDP PPS header as per DP 1.4 spec, Table 2-123 */
@@ -909,7 +909,7 @@ void intel_dsc_enable(struct intel_encoder *encoder,
 	u32 dss_ctl1_val = 0;
 	u32 dss_ctl2_val = 0;
 
-	if (!crtc_state->dsc_params.compression_enable)
+	if (!crtc_state->dsc.compression_enable)
 		return;
 
 	/* Enable Power wells for VDSC/joining */
@@ -928,7 +928,7 @@ void intel_dsc_enable(struct intel_encoder *encoder,
 		dss_ctl2_reg = ICL_PIPE_DSS_CTL2(pipe);
 	}
 	dss_ctl2_val |= LEFT_BRANCH_VDSC_ENABLE;
-	if (crtc_state->dsc_params.dsc_split) {
+	if (crtc_state->dsc.dsc_split) {
 		dss_ctl2_val |= RIGHT_BRANCH_VDSC_ENABLE;
 		dss_ctl1_val |= JOINER_ENABLE;
 	}
@@ -944,7 +944,7 @@ void intel_dsc_disable(const struct intel_crtc_state *old_crtc_state)
 	i915_reg_t dss_ctl1_reg, dss_ctl2_reg;
 	u32 dss_ctl1_val = 0, dss_ctl2_val = 0;
 
-	if (!old_crtc_state->dsc_params.compression_enable)
+	if (!old_crtc_state->dsc.compression_enable)
 		return;
 
 	if (old_crtc_state->cpu_transcoder == TRANSCODER_EDP) {
diff --git a/drivers/gpu/drm/i915/display/intel_vga.c b/drivers/gpu/drm/i915/display/intel_vga.c
new file mode 100644
index 000000000000..2ff7293986d4
--- /dev/null
+++ b/drivers/gpu/drm/i915/display/intel_vga.c
@@ -0,0 +1,160 @@
+// SPDX-License-Identifier: MIT
+/*
+ * Copyright © 2019 Intel Corporation
+ */
+
+#include <linux/pci.h>
+#include <linux/vgaarb.h>
+
+#include <drm/i915_drm.h>
+
+#include "i915_drv.h"
+#include "intel_vga.h"
+
+static i915_reg_t intel_vga_cntrl_reg(struct drm_i915_private *i915)
+{
+	if (IS_VALLEYVIEW(i915) || IS_CHERRYVIEW(i915))
+		return VLV_VGACNTRL;
+	else if (INTEL_GEN(i915) >= 5)
+		return CPU_VGACNTRL;
+	else
+		return VGACNTRL;
+}
+
+/* Disable the VGA plane that we never use */
+void intel_vga_disable(struct drm_i915_private *dev_priv)
+{
+	struct pci_dev *pdev = dev_priv->drm.pdev;
+	i915_reg_t vga_reg = intel_vga_cntrl_reg(dev_priv);
+	u8 sr1;
+
+	/* WaEnableVGAAccessThroughIOPort:ctg,elk,ilk,snb,ivb,vlv,hsw */
+	vga_get_uninterruptible(pdev, VGA_RSRC_LEGACY_IO);
+	outb(SR01, VGA_SR_INDEX);
+	sr1 = inb(VGA_SR_DATA);
+	outb(sr1 | 1 << 5, VGA_SR_DATA);
+	vga_put(pdev, VGA_RSRC_LEGACY_IO);
+	udelay(300);
+
+	I915_WRITE(vga_reg, VGA_DISP_DISABLE);
+	POSTING_READ(vga_reg);
+}
+
+void intel_vga_redisable_power_on(struct drm_i915_private *dev_priv)
+{
+	i915_reg_t vga_reg = intel_vga_cntrl_reg(dev_priv);
+
+	if (!(I915_READ(vga_reg) & VGA_DISP_DISABLE)) {
+		DRM_DEBUG_KMS("Something enabled VGA plane, disabling it\n");
+		intel_vga_disable(dev_priv);
+	}
+}
+
+void intel_vga_redisable(struct drm_i915_private *i915)
+{
+	intel_wakeref_t wakeref;
+
+	/*
+	 * This function can be called both from intel_modeset_setup_hw_state or
+	 * at a very early point in our resume sequence, where the power well
+	 * structures are not yet restored. Since this function is at a very
+	 * paranoid "someone might have enabled VGA while we were not looking"
+	 * level, just check if the power well is enabled instead of trying to
+	 * follow the "don't touch the power well if we don't need it" policy
+	 * the rest of the driver uses.
+	 */
+	wakeref = intel_display_power_get_if_enabled(i915, POWER_DOMAIN_VGA);
+	if (!wakeref)
+		return;
+
+	intel_vga_redisable_power_on(i915);
+
+	intel_display_power_put(i915, POWER_DOMAIN_VGA, wakeref);
+}
+
+void intel_vga_reset_io_mem(struct drm_i915_private *i915)
+{
+	struct pci_dev *pdev = i915->drm.pdev;
+
+	/*
+	 * After we re-enable the power well, if we touch VGA register 0x3d5
+	 * we'll get unclaimed register interrupts. This stops after we write
+	 * anything to the VGA MSR register. The vgacon module uses this
+	 * register all the time, so if we unbind our driver and, as a
+	 * consequence, bind vgacon, we'll get stuck in an infinite loop at
+	 * console_unlock(). So make here we touch the VGA MSR register, making
+	 * sure vgacon can keep working normally without triggering interrupts
+	 * and error messages.
+	 */
+	vga_get_uninterruptible(pdev, VGA_RSRC_LEGACY_IO);
+	outb(inb(VGA_MSR_READ), VGA_MSR_WRITE);
+	vga_put(pdev, VGA_RSRC_LEGACY_IO);
+}
+
+static int
+intel_vga_set_state(struct drm_i915_private *i915, bool enable_decode)
+{
+	unsigned int reg = INTEL_GEN(i915) >= 6 ? SNB_GMCH_CTRL : INTEL_GMCH_CTRL;
+	u16 gmch_ctrl;
+
+	if (pci_read_config_word(i915->bridge_dev, reg, &gmch_ctrl)) {
+		DRM_ERROR("failed to read control word\n");
+		return -EIO;
+	}
+
+	if (!!(gmch_ctrl & INTEL_GMCH_VGA_DISABLE) == !enable_decode)
+		return 0;
+
+	if (enable_decode)
+		gmch_ctrl &= ~INTEL_GMCH_VGA_DISABLE;
+	else
+		gmch_ctrl |= INTEL_GMCH_VGA_DISABLE;
+
+	if (pci_write_config_word(i915->bridge_dev, reg, gmch_ctrl)) {
+		DRM_ERROR("failed to write control word\n");
+		return -EIO;
+	}
+
+	return 0;
+}
+
+static unsigned int
+intel_vga_set_decode(void *cookie, bool enable_decode)
+{
+	struct drm_i915_private *i915 = cookie;
+
+	intel_vga_set_state(i915, enable_decode);
+
+	if (enable_decode)
+		return VGA_RSRC_LEGACY_IO | VGA_RSRC_LEGACY_MEM |
+		       VGA_RSRC_NORMAL_IO | VGA_RSRC_NORMAL_MEM;
+	else
+		return VGA_RSRC_NORMAL_IO | VGA_RSRC_NORMAL_MEM;
+}
+
+int intel_vga_register(struct drm_i915_private *i915)
+{
+	struct pci_dev *pdev = i915->drm.pdev;
+	int ret;
+
+	/*
+	 * If we have > 1 VGA cards, then we need to arbitrate access to the
+	 * common VGA resources.
+	 *
+	 * If we are a secondary display controller (!PCI_DISPLAY_CLASS_VGA),
+	 * then we do not take part in VGA arbitration and the
+	 * vga_client_register() fails with -ENODEV.
+	 */
+	ret = vga_client_register(pdev, i915, NULL, intel_vga_set_decode);
+	if (ret && ret != -ENODEV)
+		return ret;
+
+	return 0;
+}
+
+void intel_vga_unregister(struct drm_i915_private *i915)
+{
+	struct pci_dev *pdev = i915->drm.pdev;
+
+	vga_client_register(pdev, NULL, NULL, NULL);
+}
diff --git a/drivers/gpu/drm/i915/display/intel_vga.h b/drivers/gpu/drm/i915/display/intel_vga.h
new file mode 100644
index 000000000000..ba5b55b917f0
--- /dev/null
+++ b/drivers/gpu/drm/i915/display/intel_vga.h
@@ -0,0 +1,18 @@
+/* SPDX-License-Identifier: MIT */
+/*
+ * Copyright © 2019 Intel Corporation
+ */
+
+#ifndef __INTEL_VGA_H__
+#define __INTEL_VGA_H__
+
+struct drm_i915_private;
+
+void intel_vga_reset_io_mem(struct drm_i915_private *i915);
+void intel_vga_disable(struct drm_i915_private *i915);
+void intel_vga_redisable(struct drm_i915_private *i915);
+void intel_vga_redisable_power_on(struct drm_i915_private *i915);
+int intel_vga_register(struct drm_i915_private *i915);
+void intel_vga_unregister(struct drm_i915_private *i915);
+
+#endif /* __INTEL_VGA_H__ */
diff --git a/drivers/gpu/drm/i915/display/vlv_dsi.c b/drivers/gpu/drm/i915/display/vlv_dsi.c
index a71b22bdd95b..0ca49b1604c6 100644
--- a/drivers/gpu/drm/i915/display/vlv_dsi.c
+++ b/drivers/gpu/drm/i915/display/vlv_dsi.c
@@ -749,7 +749,7 @@ static void intel_dsi_pre_enable(struct intel_encoder *encoder,
 	struct drm_crtc *crtc = pipe_config->base.crtc;
 	struct drm_i915_private *dev_priv = to_i915(crtc->dev);
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	int pipe = intel_crtc->pipe;
+	enum pipe pipe = intel_crtc->pipe;
 	enum port port;
 	u32 val;
 	bool glk_cold_boot = false;
@@ -1870,11 +1870,11 @@ void vlv_dsi_init(struct drm_i915_private *dev_priv)
 	 * port C. BXT isn't limited like this.
 	 */
 	if (IS_GEN9_LP(dev_priv))
-		intel_encoder->crtc_mask = BIT(PIPE_A) | BIT(PIPE_B) | BIT(PIPE_C);
+		intel_encoder->pipe_mask = ~0;
 	else if (port == PORT_A)
-		intel_encoder->crtc_mask = BIT(PIPE_A);
+		intel_encoder->pipe_mask = BIT(PIPE_A);
 	else
-		intel_encoder->crtc_mask = BIT(PIPE_B);
+		intel_encoder->pipe_mask = BIT(PIPE_B);
 
 	if (dev_priv->vbt.dsi.config->dual_link)
 		intel_dsi->ports = BIT(PORT_A) | BIT(PORT_C);
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_client_blt.c b/drivers/gpu/drm/i915/gem/i915_gem_client_blt.c
index f99920652751..81366aa4812b 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_client_blt.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_client_blt.c
@@ -155,7 +155,6 @@ static void clear_pages_dma_fence_cb(struct dma_fence *fence,
 static void clear_pages_worker(struct work_struct *work)
 {
 	struct clear_pages_work *w = container_of(work, typeof(*w), work);
-	struct drm_i915_private *i915 = w->ce->engine->i915;
 	struct drm_i915_gem_object *obj = w->sleeve->vma->obj;
 	struct i915_vma *vma = w->sleeve->vma;
 	struct i915_request *rq;
@@ -173,11 +172,9 @@ static void clear_pages_worker(struct work_struct *work)
 	obj->read_domains = I915_GEM_GPU_DOMAINS;
 	obj->write_domain = 0;
 
-	/* XXX: we need to kill this */
-	mutex_lock(&i915->drm.struct_mutex);
 	err = i915_vma_pin(vma, 0, 0, PIN_USER);
 	if (unlikely(err))
-		goto out_unlock;
+		goto out_signal;
 
 	batch = intel_emit_vma_fill_blt(w->ce, vma, w->value);
 	if (IS_ERR(batch)) {
@@ -211,7 +208,7 @@ static void clear_pages_worker(struct work_struct *work)
 	 * keep track of the GPU activity within this vma/request, and
 	 * propagate the signal from the request to w->dma.
 	 */
-	err = i915_active_ref(&vma->active, rq->timeline, rq);
+	err = __i915_vma_move_to_active(vma, rq);
 	if (err)
 		goto out_request;
 
@@ -229,8 +226,6 @@ out_batch:
 	intel_emit_vma_release(w->ce, batch);
 out_unpin:
 	i915_vma_unpin(vma);
-out_unlock:
-	mutex_unlock(&i915->drm.struct_mutex);
 out_signal:
 	if (unlikely(err)) {
 		dma_fence_set_error(&w->dma, err);
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_context.c b/drivers/gpu/drm/i915/gem/i915_gem_context.c
index 1cdfe05514c3..e553ca8d98eb 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_context.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_context.c
@@ -69,8 +69,10 @@
 
 #include <drm/i915_drm.h>
 
-#include "gt/intel_lrc_reg.h"
+#include "gt/intel_engine_heartbeat.h"
 #include "gt/intel_engine_user.h"
+#include "gt/intel_lrc_reg.h"
+#include "gt/intel_ring.h"
 
 #include "i915_gem_context.h"
 #include "i915_globals.h"
@@ -167,97 +169,6 @@ lookup_user_engine(struct i915_gem_context *ctx,
 	return i915_gem_context_get_engine(ctx, idx);
 }
 
-static inline int new_hw_id(struct drm_i915_private *i915, gfp_t gfp)
-{
-	unsigned int max;
-
-	lockdep_assert_held(&i915->contexts.mutex);
-
-	if (INTEL_GEN(i915) >= 12)
-		max = GEN12_MAX_CONTEXT_HW_ID;
-	else if (INTEL_GEN(i915) >= 11)
-		max = GEN11_MAX_CONTEXT_HW_ID;
-	else if (USES_GUC_SUBMISSION(i915))
-		/*
-		 * When using GuC in proxy submission, GuC consumes the
-		 * highest bit in the context id to indicate proxy submission.
-		 */
-		max = MAX_GUC_CONTEXT_HW_ID;
-	else
-		max = MAX_CONTEXT_HW_ID;
-
-	return ida_simple_get(&i915->contexts.hw_ida, 0, max, gfp);
-}
-
-static int steal_hw_id(struct drm_i915_private *i915)
-{
-	struct i915_gem_context *ctx, *cn;
-	LIST_HEAD(pinned);
-	int id = -ENOSPC;
-
-	lockdep_assert_held(&i915->contexts.mutex);
-
-	list_for_each_entry_safe(ctx, cn,
-				 &i915->contexts.hw_id_list, hw_id_link) {
-		if (atomic_read(&ctx->hw_id_pin_count)) {
-			list_move_tail(&ctx->hw_id_link, &pinned);
-			continue;
-		}
-
-		GEM_BUG_ON(!ctx->hw_id); /* perma-pinned kernel context */
-		list_del_init(&ctx->hw_id_link);
-		id = ctx->hw_id;
-		break;
-	}
-
-	/*
-	 * Remember how far we got up on the last repossesion scan, so the
-	 * list is kept in a "least recently scanned" order.
-	 */
-	list_splice_tail(&pinned, &i915->contexts.hw_id_list);
-	return id;
-}
-
-static int assign_hw_id(struct drm_i915_private *i915, unsigned int *out)
-{
-	int ret;
-
-	lockdep_assert_held(&i915->contexts.mutex);
-
-	/*
-	 * We prefer to steal/stall ourselves and our users over that of the
-	 * entire system. That may be a little unfair to our users, and
-	 * even hurt high priority clients. The choice is whether to oomkill
-	 * something else, or steal a context id.
-	 */
-	ret = new_hw_id(i915, GFP_KERNEL | __GFP_RETRY_MAYFAIL | __GFP_NOWARN);
-	if (unlikely(ret < 0)) {
-		ret = steal_hw_id(i915);
-		if (ret < 0) /* once again for the correct errno code */
-			ret = new_hw_id(i915, GFP_KERNEL);
-		if (ret < 0)
-			return ret;
-	}
-
-	*out = ret;
-	return 0;
-}
-
-static void release_hw_id(struct i915_gem_context *ctx)
-{
-	struct drm_i915_private *i915 = ctx->i915;
-
-	if (list_empty(&ctx->hw_id_link))
-		return;
-
-	mutex_lock(&i915->contexts.mutex);
-	if (!list_empty(&ctx->hw_id_link)) {
-		ida_simple_remove(&i915->contexts.hw_ida, ctx->hw_id);
-		list_del_init(&ctx->hw_id_link);
-	}
-	mutex_unlock(&i915->contexts.mutex);
-}
-
 static void __free_engines(struct i915_gem_engines *e, unsigned int count)
 {
 	while (count--) {
@@ -294,101 +205,241 @@ static struct i915_gem_engines *default_engines(struct i915_gem_context *ctx)
 	for_each_engine(engine, gt, id) {
 		struct intel_context *ce;
 
+		if (engine->legacy_idx == INVALID_ENGINE)
+			continue;
+
+		GEM_BUG_ON(engine->legacy_idx >= I915_NUM_ENGINES);
+		GEM_BUG_ON(e->engines[engine->legacy_idx]);
+
 		ce = intel_context_create(ctx, engine);
 		if (IS_ERR(ce)) {
-			__free_engines(e, id);
+			__free_engines(e, e->num_engines + 1);
 			return ERR_CAST(ce);
 		}
 
-		e->engines[id] = ce;
-		e->num_engines = id + 1;
+		e->engines[engine->legacy_idx] = ce;
+		e->num_engines = max(e->num_engines, engine->legacy_idx);
 	}
+	e->num_engines++;
 
 	return e;
 }
 
 static void i915_gem_context_free(struct i915_gem_context *ctx)
 {
-	lockdep_assert_held(&ctx->i915->drm.struct_mutex);
 	GEM_BUG_ON(!i915_gem_context_is_closed(ctx));
 
-	release_hw_id(ctx);
-	if (ctx->vm)
-		i915_vm_put(ctx->vm);
+	spin_lock(&ctx->i915->gem.contexts.lock);
+	list_del(&ctx->link);
+	spin_unlock(&ctx->i915->gem.contexts.lock);
 
 	free_engines(rcu_access_pointer(ctx->engines));
 	mutex_destroy(&ctx->engines_mutex);
 
+	kfree(ctx->jump_whitelist);
+
 	if (ctx->timeline)
 		intel_timeline_put(ctx->timeline);
 
 	kfree(ctx->name);
 	put_pid(ctx->pid);
 
-	list_del(&ctx->link);
 	mutex_destroy(&ctx->mutex);
 
 	kfree_rcu(ctx, rcu);
 }
 
-static void contexts_free(struct drm_i915_private *i915)
+static void contexts_free_all(struct llist_node *list)
 {
-	struct llist_node *freed = llist_del_all(&i915->contexts.free_list);
 	struct i915_gem_context *ctx, *cn;
 
-	lockdep_assert_held(&i915->drm.struct_mutex);
-
-	llist_for_each_entry_safe(ctx, cn, freed, free_link)
+	llist_for_each_entry_safe(ctx, cn, list, free_link)
 		i915_gem_context_free(ctx);
 }
 
-static void contexts_free_first(struct drm_i915_private *i915)
+static void contexts_flush_free(struct i915_gem_contexts *gc)
 {
-	struct i915_gem_context *ctx;
-	struct llist_node *freed;
-
-	lockdep_assert_held(&i915->drm.struct_mutex);
-
-	freed = llist_del_first(&i915->contexts.free_list);
-	if (!freed)
-		return;
-
-	ctx = container_of(freed, typeof(*ctx), free_link);
-	i915_gem_context_free(ctx);
+	contexts_free_all(llist_del_all(&gc->free_list));
 }
 
 static void contexts_free_worker(struct work_struct *work)
 {
-	struct drm_i915_private *i915 =
-		container_of(work, typeof(*i915), contexts.free_work);
+	struct i915_gem_contexts *gc =
+		container_of(work, typeof(*gc), free_work);
 
-	mutex_lock(&i915->drm.struct_mutex);
-	contexts_free(i915);
-	mutex_unlock(&i915->drm.struct_mutex);
+	contexts_flush_free(gc);
 }
 
 void i915_gem_context_release(struct kref *ref)
 {
 	struct i915_gem_context *ctx = container_of(ref, typeof(*ctx), ref);
-	struct drm_i915_private *i915 = ctx->i915;
+	struct i915_gem_contexts *gc = &ctx->i915->gem.contexts;
 
 	trace_i915_context_free(ctx);
-	if (llist_add(&ctx->free_link, &i915->contexts.free_list))
-		queue_work(i915->wq, &i915->contexts.free_work);
+	if (llist_add(&ctx->free_link, &gc->free_list))
+		schedule_work(&gc->free_work);
 }
 
-static void context_close(struct i915_gem_context *ctx)
+static inline struct i915_gem_engines *
+__context_engines_static(const struct i915_gem_context *ctx)
 {
-	mutex_lock(&ctx->mutex);
+	return rcu_dereference_protected(ctx->engines, true);
+}
 
-	i915_gem_context_set_closed(ctx);
-	ctx->file_priv = ERR_PTR(-EBADF);
+static bool __reset_engine(struct intel_engine_cs *engine)
+{
+	struct intel_gt *gt = engine->gt;
+	bool success = false;
+
+	if (!intel_has_reset_engine(gt))
+		return false;
+
+	if (!test_and_set_bit(I915_RESET_ENGINE + engine->id,
+			      &gt->reset.flags)) {
+		success = intel_engine_reset(engine, NULL) == 0;
+		clear_and_wake_up_bit(I915_RESET_ENGINE + engine->id,
+				      &gt->reset.flags);
+	}
+
+	return success;
+}
+
+static void __reset_context(struct i915_gem_context *ctx,
+			    struct intel_engine_cs *engine)
+{
+	intel_gt_handle_error(engine->gt, engine->mask, 0,
+			      "context closure in %s", ctx->name);
+}
+
+static bool __cancel_engine(struct intel_engine_cs *engine)
+{
+	/*
+	 * Send a "high priority pulse" down the engine to cause the
+	 * current request to be momentarily preempted. (If it fails to
+	 * be preempted, it will be reset). As we have marked our context
+	 * as banned, any incomplete request, including any running, will
+	 * be skipped following the preemption.
+	 *
+	 * If there is no hangchecking (one of the reasons why we try to
+	 * cancel the context) and no forced preemption, there may be no
+	 * means by which we reset the GPU and evict the persistent hog.
+	 * Ergo if we are unable to inject a preemptive pulse that can
+	 * kill the banned context, we fallback to doing a local reset
+	 * instead.
+	 */
+	if (IS_ACTIVE(CONFIG_DRM_I915_PREEMPT_TIMEOUT) &&
+	    !intel_engine_pulse(engine))
+		return true;
+
+	/* If we are unable to send a pulse, try resetting this engine. */
+	return __reset_engine(engine);
+}
+
+static struct intel_engine_cs *__active_engine(struct i915_request *rq)
+{
+	struct intel_engine_cs *engine, *locked;
+
+	/*
+	 * Serialise with __i915_request_submit() so that it sees
+	 * is-banned?, or we know the request is already inflight.
+	 */
+	locked = READ_ONCE(rq->engine);
+	spin_lock_irq(&locked->active.lock);
+	while (unlikely(locked != (engine = READ_ONCE(rq->engine)))) {
+		spin_unlock(&locked->active.lock);
+		spin_lock(&engine->active.lock);
+		locked = engine;
+	}
+
+	engine = NULL;
+	if (i915_request_is_active(rq) && !rq->fence.error)
+		engine = rq->engine;
+
+	spin_unlock_irq(&locked->active.lock);
+
+	return engine;
+}
+
+static struct intel_engine_cs *active_engine(struct intel_context *ce)
+{
+	struct intel_engine_cs *engine = NULL;
+	struct i915_request *rq;
+
+	if (!ce->timeline)
+		return NULL;
+
+	rcu_read_lock();
+	list_for_each_entry_reverse(rq, &ce->timeline->requests, link) {
+		if (i915_request_completed(rq))
+			break;
+
+		/* Check with the backend if the request is inflight */
+		engine = __active_engine(rq);
+		if (engine)
+			break;
+	}
+	rcu_read_unlock();
+
+	return engine;
+}
+
+static void kill_context(struct i915_gem_context *ctx)
+{
+	struct i915_gem_engines_iter it;
+	struct intel_context *ce;
+
+	/*
+	 * If we are already banned, it was due to a guilty request causing
+	 * a reset and the entire context being evicted from the GPU.
+	 */
+	if (i915_gem_context_is_banned(ctx))
+		return;
+
+	i915_gem_context_set_banned(ctx);
 
 	/*
-	 * This context will never again be assinged to HW, so we can
-	 * reuse its ID for the next context.
+	 * Map the user's engine back to the actual engines; one virtual
+	 * engine will be mapped to multiple engines, and using ctx->engine[]
+	 * the same engine may be have multiple instances in the user's map.
+	 * However, we only care about pending requests, so only include
+	 * engines on which there are incomplete requests.
 	 */
-	release_hw_id(ctx);
+	for_each_gem_engine(ce, __context_engines_static(ctx), it) {
+		struct intel_engine_cs *engine;
+
+		/*
+		 * Check the current active state of this context; if we
+		 * are currently executing on the GPU we need to evict
+		 * ourselves. On the other hand, if we haven't yet been
+		 * submitted to the GPU or if everything is complete,
+		 * we have nothing to do.
+		 */
+		engine = active_engine(ce);
+
+		/* First attempt to gracefully cancel the context */
+		if (engine && !__cancel_engine(engine))
+			/*
+			 * If we are unable to send a preemptive pulse to bump
+			 * the context from the GPU, we have to resort to a full
+			 * reset. We hope the collateral damage is worth it.
+			 */
+			__reset_context(ctx, engine);
+	}
+}
+
+static void context_close(struct i915_gem_context *ctx)
+{
+	struct i915_address_space *vm;
+
+	i915_gem_context_set_closed(ctx);
+
+	mutex_lock(&ctx->mutex);
+
+	vm = i915_gem_context_vm(ctx);
+	if (vm)
+		i915_vm_close(vm);
+
+	ctx->file_priv = ERR_PTR(-EBADF);
 
 	/*
 	 * The LUT uses the VMA as a backpointer to unref the object,
@@ -398,9 +449,47 @@ static void context_close(struct i915_gem_context *ctx)
 	lut_close(ctx);
 
 	mutex_unlock(&ctx->mutex);
+
+	/*
+	 * If the user has disabled hangchecking, we can not be sure that
+	 * the batches will ever complete after the context is closed,
+	 * keeping the context and all resources pinned forever. So in this
+	 * case we opt to forcibly kill off all remaining requests on
+	 * context close.
+	 */
+	if (!i915_gem_context_is_persistent(ctx) ||
+	    !i915_modparams.enable_hangcheck)
+		kill_context(ctx);
+
 	i915_gem_context_put(ctx);
 }
 
+static int __context_set_persistence(struct i915_gem_context *ctx, bool state)
+{
+	if (i915_gem_context_is_persistent(ctx) == state)
+		return 0;
+
+	if (state) {
+		/*
+		 * Only contexts that are short-lived [that will expire or be
+		 * reset] are allowed to survive past termination. We require
+		 * hangcheck to ensure that the persistent requests are healthy.
+		 */
+		if (!i915_modparams.enable_hangcheck)
+			return -EINVAL;
+
+		i915_gem_context_set_persistence(ctx);
+	} else {
+		/* To cancel a context we use "preempt-to-idle" */
+		if (!(ctx->i915->caps.scheduler & I915_SCHEDULER_CAP_PREEMPTION))
+			return -ENODEV;
+
+		i915_gem_context_clear_persistence(ctx);
+	}
+
+	return 0;
+}
+
 static struct i915_gem_context *
 __create_context(struct drm_i915_private *i915)
 {
@@ -414,7 +503,6 @@ __create_context(struct drm_i915_private *i915)
 		return ERR_PTR(-ENOMEM);
 
 	kref_init(&ctx->ref);
-	list_add_tail(&ctx->link, &i915->contexts.list);
 	ctx->i915 = i915;
 	ctx->sched.priority = I915_USER_PRIORITY(I915_PRIORITY_NORMAL);
 	mutex_init(&ctx->mutex);
@@ -428,7 +516,6 @@ __create_context(struct drm_i915_private *i915)
 	RCU_INIT_POINTER(ctx->engines, e);
 
 	INIT_RADIX_TREE(&ctx->handles_vma, GFP_KERNEL);
-	INIT_LIST_HEAD(&ctx->hw_id_link);
 
 	/* NB: Mark all slices as needing a remap so that when the context first
 	 * loads it will restore whatever remap state already exists. If there
@@ -437,10 +524,18 @@ __create_context(struct drm_i915_private *i915)
 
 	i915_gem_context_set_bannable(ctx);
 	i915_gem_context_set_recoverable(ctx);
+	__context_set_persistence(ctx, true /* cgroup hook? */);
 
 	for (i = 0; i < ARRAY_SIZE(ctx->hang_timestamp); i++)
 		ctx->hang_timestamp[i] = jiffies - CONTEXT_FAST_HANG_JIFFIES;
 
+	ctx->jump_whitelist = NULL;
+	ctx->jump_whitelist_cmds = 0;
+
+	spin_lock(&i915->gem.contexts.lock);
+	list_add_tail(&ctx->link, &i915->gem.contexts.list);
+	spin_unlock(&i915->gem.contexts.lock);
+
 	return ctx;
 
 err_free:
@@ -470,11 +565,11 @@ static void __apply_ppgtt(struct intel_context *ce, void *vm)
 static struct i915_address_space *
 __set_ppgtt(struct i915_gem_context *ctx, struct i915_address_space *vm)
 {
-	struct i915_address_space *old = ctx->vm;
+	struct i915_address_space *old = i915_gem_context_vm(ctx);
 
 	GEM_BUG_ON(old && i915_vm_is_4lvl(vm) != i915_vm_is_4lvl(old));
 
-	ctx->vm = i915_vm_get(vm);
+	rcu_assign_pointer(ctx->vm, i915_vm_open(vm));
 	context_apply_all(ctx, __apply_ppgtt, vm);
 
 	return old;
@@ -483,12 +578,12 @@ __set_ppgtt(struct i915_gem_context *ctx, struct i915_address_space *vm)
 static void __assign_ppgtt(struct i915_gem_context *ctx,
 			   struct i915_address_space *vm)
 {
-	if (vm == ctx->vm)
+	if (vm == rcu_access_pointer(ctx->vm))
 		return;
 
 	vm = __set_ppgtt(ctx, vm);
 	if (vm)
-		i915_vm_put(vm);
+		i915_vm_close(vm);
 }
 
 static void __set_timeline(struct intel_timeline **dst,
@@ -515,27 +610,25 @@ static void __assign_timeline(struct i915_gem_context *ctx,
 }
 
 static struct i915_gem_context *
-i915_gem_create_context(struct drm_i915_private *dev_priv, unsigned int flags)
+i915_gem_create_context(struct drm_i915_private *i915, unsigned int flags)
 {
 	struct i915_gem_context *ctx;
 
-	lockdep_assert_held(&dev_priv->drm.struct_mutex);
-
 	if (flags & I915_CONTEXT_CREATE_FLAGS_SINGLE_TIMELINE &&
-	    !HAS_EXECLISTS(dev_priv))
+	    !HAS_EXECLISTS(i915))
 		return ERR_PTR(-EINVAL);
 
-	/* Reap the most stale context */
-	contexts_free_first(dev_priv);
+	/* Reap the stale contexts */
+	contexts_flush_free(&i915->gem.contexts);
 
-	ctx = __create_context(dev_priv);
+	ctx = __create_context(i915);
 	if (IS_ERR(ctx))
 		return ctx;
 
-	if (HAS_FULL_PPGTT(dev_priv)) {
+	if (HAS_FULL_PPGTT(i915)) {
 		struct i915_ppgtt *ppgtt;
 
-		ppgtt = i915_ppgtt_create(dev_priv);
+		ppgtt = i915_ppgtt_create(i915);
 		if (IS_ERR(ppgtt)) {
 			DRM_DEBUG_DRIVER("PPGTT setup failed (%ld)\n",
 					 PTR_ERR(ppgtt));
@@ -543,14 +636,17 @@ i915_gem_create_context(struct drm_i915_private *dev_priv, unsigned int flags)
 			return ERR_CAST(ppgtt);
 		}
 
+		mutex_lock(&ctx->mutex);
 		__assign_ppgtt(ctx, &ppgtt->vm);
+		mutex_unlock(&ctx->mutex);
+
 		i915_vm_put(&ppgtt->vm);
 	}
 
 	if (flags & I915_CONTEXT_CREATE_FLAGS_SINGLE_TIMELINE) {
 		struct intel_timeline *timeline;
 
-		timeline = intel_timeline_create(&dev_priv->gt, NULL);
+		timeline = intel_timeline_create(&i915->gt, NULL);
 		if (IS_ERR(timeline)) {
 			context_close(ctx);
 			return ERR_CAST(timeline);
@@ -582,19 +678,13 @@ struct i915_gem_context *
 i915_gem_context_create_kernel(struct drm_i915_private *i915, int prio)
 {
 	struct i915_gem_context *ctx;
-	int err;
 
 	ctx = i915_gem_create_context(i915, 0);
 	if (IS_ERR(ctx))
 		return ctx;
 
-	err = i915_gem_context_pin_hw_id(ctx);
-	if (err) {
-		destroy_kernel_context(&ctx);
-		return ERR_PTR(err);
-	}
-
 	i915_gem_context_clear_bannable(ctx);
+	i915_gem_context_set_persistence(ctx);
 	ctx->sched.priority = I915_USER_PRIORITY(prio);
 
 	GEM_BUG_ON(!i915_gem_context_is_kernel(ctx));
@@ -602,62 +692,42 @@ i915_gem_context_create_kernel(struct drm_i915_private *i915, int prio)
 	return ctx;
 }
 
-static void init_contexts(struct drm_i915_private *i915)
+static void init_contexts(struct i915_gem_contexts *gc)
 {
-	mutex_init(&i915->contexts.mutex);
-	INIT_LIST_HEAD(&i915->contexts.list);
-
-	/* Using the simple ida interface, the max is limited by sizeof(int) */
-	BUILD_BUG_ON(MAX_CONTEXT_HW_ID > INT_MAX);
-	BUILD_BUG_ON(GEN11_MAX_CONTEXT_HW_ID > INT_MAX);
-	ida_init(&i915->contexts.hw_ida);
-	INIT_LIST_HEAD(&i915->contexts.hw_id_list);
+	spin_lock_init(&gc->lock);
+	INIT_LIST_HEAD(&gc->list);
 
-	INIT_WORK(&i915->contexts.free_work, contexts_free_worker);
-	init_llist_head(&i915->contexts.free_list);
+	INIT_WORK(&gc->free_work, contexts_free_worker);
+	init_llist_head(&gc->free_list);
 }
 
-int i915_gem_contexts_init(struct drm_i915_private *dev_priv)
+int i915_gem_init_contexts(struct drm_i915_private *i915)
 {
 	struct i915_gem_context *ctx;
 
 	/* Reassure ourselves we are only called once */
-	GEM_BUG_ON(dev_priv->kernel_context);
+	GEM_BUG_ON(i915->kernel_context);
 
-	init_contexts(dev_priv);
+	init_contexts(&i915->gem.contexts);
 
 	/* lowest priority; idle task */
-	ctx = i915_gem_context_create_kernel(dev_priv, I915_PRIORITY_MIN);
+	ctx = i915_gem_context_create_kernel(i915, I915_PRIORITY_MIN);
 	if (IS_ERR(ctx)) {
 		DRM_ERROR("Failed to create default global context\n");
 		return PTR_ERR(ctx);
 	}
-	/*
-	 * For easy recognisablity, we want the kernel context to be 0 and then
-	 * all user contexts will have non-zero hw_id. Kernel contexts are
-	 * permanently pinned, so that we never suffer a stall and can
-	 * use them from any allocation context (e.g. for evicting other
-	 * contexts and from inside the shrinker).
-	 */
-	GEM_BUG_ON(ctx->hw_id);
-	GEM_BUG_ON(!atomic_read(&ctx->hw_id_pin_count));
-	dev_priv->kernel_context = ctx;
+	i915->kernel_context = ctx;
 
 	DRM_DEBUG_DRIVER("%s context support initialized\n",
-			 DRIVER_CAPS(dev_priv)->has_logical_contexts ?
+			 DRIVER_CAPS(i915)->has_logical_contexts ?
 			 "logical" : "fake");
 	return 0;
 }
 
-void i915_gem_contexts_fini(struct drm_i915_private *i915)
+void i915_gem_driver_release__contexts(struct drm_i915_private *i915)
 {
-	lockdep_assert_held(&i915->drm.struct_mutex);
-
 	destroy_kernel_context(&i915->kernel_context);
-
-	/* Must free all deferred contexts (via flush_workqueue) first */
-	GEM_BUG_ON(!list_empty(&i915->contexts.hw_id_list));
-	ida_destroy(&i915->contexts.hw_ida);
+	flush_work(&i915->gem.contexts.free_work);
 }
 
 static int context_idr_cleanup(int id, void *p, void *data)
@@ -675,11 +745,16 @@ static int vm_idr_cleanup(int id, void *p, void *data)
 static int gem_context_register(struct i915_gem_context *ctx,
 				struct drm_i915_file_private *fpriv)
 {
+	struct i915_address_space *vm;
 	int ret;
 
 	ctx->file_priv = fpriv;
-	if (ctx->vm)
-		ctx->vm->file = fpriv;
+
+	mutex_lock(&ctx->mutex);
+	vm = i915_gem_context_vm(ctx);
+	if (vm)
+		WRITE_ONCE(vm->file, fpriv); /* XXX */
+	mutex_unlock(&ctx->mutex);
 
 	ctx->pid = get_task_pid(current, PIDTYPE_PID);
 	ctx->name = kasprintf(GFP_KERNEL, "%s[%d]",
@@ -716,9 +791,7 @@ int i915_gem_context_open(struct drm_i915_private *i915,
 	idr_init(&file_priv->context_idr);
 	idr_init_base(&file_priv->vm_idr, 1);
 
-	mutex_lock(&i915->drm.struct_mutex);
 	ctx = i915_gem_create_context(i915, 0);
-	mutex_unlock(&i915->drm.struct_mutex);
 	if (IS_ERR(ctx)) {
 		err = PTR_ERR(ctx);
 		goto err;
@@ -746,6 +819,7 @@ err:
 void i915_gem_context_close(struct drm_file *file)
 {
 	struct drm_i915_file_private *file_priv = file->driver_priv;
+	struct drm_i915_private *i915 = file_priv->dev_priv;
 
 	idr_for_each(&file_priv->context_idr, context_idr_cleanup, NULL);
 	idr_destroy(&file_priv->context_idr);
@@ -754,6 +828,8 @@ void i915_gem_context_close(struct drm_file *file)
 	idr_for_each(&file_priv->vm_idr, vm_idr_cleanup, NULL);
 	idr_destroy(&file_priv->vm_idr);
 	mutex_destroy(&file_priv->vm_idr_lock);
+
+	contexts_flush_free(&i915->gem.contexts);
 }
 
 int i915_gem_vm_create_ioctl(struct drm_device *dev, void *data,
@@ -846,6 +922,7 @@ struct context_barrier_task {
 	void *data;
 };
 
+__i915_active_call
 static void cb_retire(struct i915_active *base)
 {
 	struct context_barrier_task *cb = container_of(base, typeof(*cb), base);
@@ -865,20 +942,18 @@ static int context_barrier_task(struct i915_gem_context *ctx,
 				void (*task)(void *data),
 				void *data)
 {
-	struct drm_i915_private *i915 = ctx->i915;
 	struct context_barrier_task *cb;
 	struct i915_gem_engines_iter it;
 	struct intel_context *ce;
 	int err = 0;
 
-	lockdep_assert_held(&i915->drm.struct_mutex);
 	GEM_BUG_ON(!task);
 
 	cb = kmalloc(sizeof(*cb), GFP_KERNEL);
 	if (!cb)
 		return -ENOMEM;
 
-	i915_active_init(i915, &cb->base, NULL, cb_retire);
+	i915_active_init(&cb->base, NULL, cb_retire);
 	err = i915_active_acquire(&cb->base);
 	if (err) {
 		kfree(cb);
@@ -910,7 +985,7 @@ static int context_barrier_task(struct i915_gem_context *ctx,
 		if (emit)
 			err = emit(rq, data);
 		if (err == 0)
-			err = i915_active_ref(&cb->base, rq->timeline, rq);
+			err = i915_active_add_request(&cb->base, rq);
 
 		i915_request_add(rq);
 		if (err)
@@ -933,16 +1008,12 @@ static int get_ppgtt(struct drm_i915_file_private *file_priv,
 	struct i915_address_space *vm;
 	int ret;
 
-	if (!ctx->vm)
+	if (!rcu_access_pointer(ctx->vm))
 		return -ENODEV;
 
-	/* XXX rcu acquire? */
-	ret = mutex_lock_interruptible(&ctx->i915->drm.struct_mutex);
-	if (ret)
-		return ret;
-
+	rcu_read_lock();
 	vm = i915_vm_get(ctx->vm);
-	mutex_unlock(&ctx->i915->drm.struct_mutex);
+	rcu_read_unlock();
 
 	ret = mutex_lock_interruptible(&file_priv->vm_idr_lock);
 	if (ret)
@@ -953,7 +1024,7 @@ static int get_ppgtt(struct drm_i915_file_private *file_priv,
 	if (ret < 0)
 		goto err_unlock;
 
-	i915_vm_get(vm);
+	i915_vm_open(vm);
 
 	args->size = 0;
 	args->value = ret;
@@ -973,7 +1044,7 @@ static void set_ppgtt_barrier(void *data)
 	if (INTEL_GEN(old->i915) < 8)
 		gen6_ppgtt_unpin_all(i915_vm_to_ppgtt(old));
 
-	i915_vm_put(old);
+	i915_vm_close(old);
 }
 
 static int emit_ppgtt_update(struct i915_request *rq, void *data)
@@ -1003,12 +1074,18 @@ static int emit_ppgtt_update(struct i915_request *rq, void *data)
 		intel_ring_advance(rq, cs);
 	} else if (HAS_LOGICAL_RING_CONTEXTS(engine->i915)) {
 		struct i915_ppgtt *ppgtt = i915_vm_to_ppgtt(vm);
+		int err;
+
+		/* Magic required to prevent forcewake errors! */
+		err = engine->emit_flush(rq, EMIT_INVALIDATE);
+		if (err)
+			return err;
 
 		cs = intel_ring_begin(rq, 4 * GEN8_3LVL_PDPES + 2);
 		if (IS_ERR(cs))
 			return PTR_ERR(cs);
 
-		*cs++ = MI_LOAD_REGISTER_IMM(2 * GEN8_3LVL_PDPES);
+		*cs++ = MI_LOAD_REGISTER_IMM(2 * GEN8_3LVL_PDPES) | MI_LRI_FORCE_POSTED;
 		for (i = GEN8_3LVL_PDPES; i--; ) {
 			const dma_addr_t pd_daddr = i915_page_dir_dma_addr(ppgtt, i);
 
@@ -1045,34 +1122,34 @@ static int set_ppgtt(struct drm_i915_file_private *file_priv,
 	if (args->size)
 		return -EINVAL;
 
-	if (!ctx->vm)
+	if (!rcu_access_pointer(ctx->vm))
 		return -ENODEV;
 
 	if (upper_32_bits(args->value))
 		return -ENOENT;
 
-	err = mutex_lock_interruptible(&file_priv->vm_idr_lock);
-	if (err)
-		return err;
-
+	rcu_read_lock();
 	vm = idr_find(&file_priv->vm_idr, args->value);
-	if (vm)
-		i915_vm_get(vm);
-	mutex_unlock(&file_priv->vm_idr_lock);
+	if (vm && !kref_get_unless_zero(&vm->ref))
+		vm = NULL;
+	rcu_read_unlock();
 	if (!vm)
 		return -ENOENT;
 
-	err = mutex_lock_interruptible(&ctx->i915->drm.struct_mutex);
+	err = mutex_lock_interruptible(&ctx->mutex);
 	if (err)
 		goto out;
 
-	if (vm == ctx->vm)
+	if (i915_gem_context_is_closed(ctx)) {
+		err = -ENOENT;
+		goto unlock;
+	}
+
+	if (vm == rcu_access_pointer(ctx->vm))
 		goto unlock;
 
 	/* Teardown the existing obj:vma cache, it will have to be rebuilt. */
-	mutex_lock(&ctx->mutex);
 	lut_close(ctx);
-	mutex_unlock(&ctx->mutex);
 
 	old = __set_ppgtt(ctx, vm);
 
@@ -1087,13 +1164,12 @@ static int set_ppgtt(struct drm_i915_file_private *file_priv,
 				   set_ppgtt_barrier,
 				   old);
 	if (err) {
-		i915_vm_put(__set_ppgtt(ctx, old));
-		i915_vm_put(old);
+		i915_vm_close(__set_ppgtt(ctx, old));
+		i915_vm_close(old);
 	}
 
 unlock:
-	mutex_unlock(&ctx->i915->drm.struct_mutex);
-
+	mutex_unlock(&ctx->mutex);
 out:
 	i915_vm_put(vm);
 	return err;
@@ -1112,7 +1188,7 @@ static int gen8_emit_rpcs_config(struct i915_request *rq,
 
 	offset = i915_ggtt_offset(ce->state) +
 		 LRC_STATE_PN * PAGE_SIZE +
-		 (CTX_R_PWR_CLK_STATE + 1) * 4;
+		 CTX_R_PWR_CLK_STATE * 4;
 
 	*cs++ = MI_STORE_DWORD_IMM_GEN4 | MI_USE_GGTT;
 	*cs++ = lower_32_bits(offset);
@@ -1155,8 +1231,7 @@ gen8_modify_rpcs(struct intel_context *ce, struct intel_sseu sseu)
 }
 
 static int
-__intel_context_reconfigure_sseu(struct intel_context *ce,
-				 struct intel_sseu sseu)
+intel_context_reconfigure_sseu(struct intel_context *ce, struct intel_sseu sseu)
 {
 	int ret;
 
@@ -1180,23 +1255,6 @@ unlock:
 }
 
 static int
-intel_context_reconfigure_sseu(struct intel_context *ce, struct intel_sseu sseu)
-{
-	struct drm_i915_private *i915 = ce->engine->i915;
-	int ret;
-
-	ret = mutex_lock_interruptible(&i915->drm.struct_mutex);
-	if (ret)
-		return ret;
-
-	ret = __intel_context_reconfigure_sseu(ce, sseu);
-
-	mutex_unlock(&i915->drm.struct_mutex);
-
-	return ret;
-}
-
-static int
 user_to_context_sseu(struct drm_i915_private *i915,
 		     const struct drm_i915_gem_context_param_sseu *user,
 		     struct intel_sseu *context)
@@ -1629,7 +1687,7 @@ replace:
 		i915_gem_context_set_user_engines(ctx);
 	else
 		i915_gem_context_clear_user_engines(ctx);
-	rcu_swap_protected(ctx->engines, set.engines, 1);
+	set.engines = rcu_replace_pointer(ctx->engines, set.engines, 1);
 	mutex_unlock(&ctx->engines_mutex);
 
 	call_rcu(&set.engines->rcu, free_engines_rcu);
@@ -1738,6 +1796,16 @@ err_free:
 	return err;
 }
 
+static int
+set_persistence(struct i915_gem_context *ctx,
+		const struct drm_i915_gem_context_param *args)
+{
+	if (args->size)
+		return -EINVAL;
+
+	return __context_set_persistence(ctx, args->value);
+}
+
 static int ctx_setparam(struct drm_i915_file_private *fpriv,
 			struct i915_gem_context *ctx,
 			struct drm_i915_gem_context_param *args)
@@ -1815,6 +1883,10 @@ static int ctx_setparam(struct drm_i915_file_private *fpriv,
 		ret = set_engines(ctx, args);
 		break;
 
+	case I915_CONTEXT_PARAM_PERSISTENCE:
+		ret = set_persistence(ctx, args);
+		break;
+
 	case I915_CONTEXT_PARAM_BAN_PERIOD:
 	default:
 		ret = -EINVAL;
@@ -1967,10 +2039,11 @@ static int clone_vm(struct i915_gem_context *dst,
 		    struct i915_gem_context *src)
 {
 	struct i915_address_space *vm;
+	int err = 0;
 
 	rcu_read_lock();
 	do {
-		vm = READ_ONCE(src->vm);
+		vm = rcu_dereference(src->vm);
 		if (!vm)
 			break;
 
@@ -1992,7 +2065,7 @@ static int clone_vm(struct i915_gem_context *dst,
 		 * it cannot be reallocated elsewhere.
 		 */
 
-		if (vm == READ_ONCE(src->vm))
+		if (vm == rcu_access_pointer(src->vm))
 			break;
 
 		i915_vm_put(vm);
@@ -2000,11 +2073,16 @@ static int clone_vm(struct i915_gem_context *dst,
 	rcu_read_unlock();
 
 	if (vm) {
-		__assign_ppgtt(dst, vm);
+		if (!mutex_lock_interruptible(&dst->mutex)) {
+			__assign_ppgtt(dst, vm);
+			mutex_unlock(&dst->mutex);
+		} else {
+			err = -EINTR;
+		}
 		i915_vm_put(vm);
 	}
 
-	return 0;
+	return err;
 }
 
 static int create_clone(struct i915_user_extension __user *ext, void *data)
@@ -2094,12 +2172,7 @@ int i915_gem_context_create_ioctl(struct drm_device *dev, void *data,
 		return -EIO;
 	}
 
-	ret = i915_mutex_lock_interruptible(dev);
-	if (ret)
-		return ret;
-
 	ext_data.ctx = i915_gem_create_context(i915, args->flags);
-	mutex_unlock(&dev->struct_mutex);
 	if (IS_ERR(ext_data.ctx))
 		return PTR_ERR(ext_data.ctx);
 
@@ -2226,12 +2299,12 @@ int i915_gem_context_getparam_ioctl(struct drm_device *dev, void *data,
 
 	case I915_CONTEXT_PARAM_GTT_SIZE:
 		args->size = 0;
-		if (ctx->vm)
-			args->value = ctx->vm->total;
-		else if (to_i915(dev)->ggtt.alias)
-			args->value = to_i915(dev)->ggtt.alias->vm.total;
+		rcu_read_lock();
+		if (rcu_access_pointer(ctx->vm))
+			args->value = rcu_dereference(ctx->vm)->total;
 		else
 			args->value = to_i915(dev)->ggtt.vm.total;
+		rcu_read_unlock();
 		break;
 
 	case I915_CONTEXT_PARAM_NO_ERROR_CAPTURE:
@@ -2266,6 +2339,11 @@ int i915_gem_context_getparam_ioctl(struct drm_device *dev, void *data,
 		ret = get_engines(ctx, args);
 		break;
 
+	case I915_CONTEXT_PARAM_PERSISTENCE:
+		args->size = 0;
+		args->value = i915_gem_context_is_persistent(ctx);
+		break;
+
 	case I915_CONTEXT_PARAM_BAN_PERIOD:
 	default:
 		ret = -EINVAL;
@@ -2297,7 +2375,7 @@ int i915_gem_context_setparam_ioctl(struct drm_device *dev, void *data,
 int i915_gem_context_reset_stats_ioctl(struct drm_device *dev,
 				       void *data, struct drm_file *file)
 {
-	struct drm_i915_private *dev_priv = to_i915(dev);
+	struct drm_i915_private *i915 = to_i915(dev);
 	struct drm_i915_reset_stats *args = data;
 	struct i915_gem_context *ctx;
 	int ret;
@@ -2319,7 +2397,7 @@ int i915_gem_context_reset_stats_ioctl(struct drm_device *dev,
 	 */
 
 	if (capable(CAP_SYS_ADMIN))
-		args->reset_count = i915_reset_count(&dev_priv->gpu_error);
+		args->reset_count = i915_reset_count(&i915->gpu_error);
 	else
 		args->reset_count = 0;
 
@@ -2332,33 +2410,6 @@ out:
 	return ret;
 }
 
-int __i915_gem_context_pin_hw_id(struct i915_gem_context *ctx)
-{
-	struct drm_i915_private *i915 = ctx->i915;
-	int err = 0;
-
-	mutex_lock(&i915->contexts.mutex);
-
-	GEM_BUG_ON(i915_gem_context_is_closed(ctx));
-
-	if (list_empty(&ctx->hw_id_link)) {
-		GEM_BUG_ON(atomic_read(&ctx->hw_id_pin_count));
-
-		err = assign_hw_id(i915, &ctx->hw_id);
-		if (err)
-			goto out_unlock;
-
-		list_add_tail(&ctx->hw_id_link, &i915->contexts.hw_id_list);
-	}
-
-	GEM_BUG_ON(atomic_read(&ctx->hw_id_pin_count) == ~0u);
-	atomic_inc(&ctx->hw_id_pin_count);
-
-out_unlock:
-	mutex_unlock(&i915->contexts.mutex);
-	return err;
-}
-
 /* GEM context-engines iterator: for_each_gem_engine() */
 struct intel_context *
 i915_gem_engines_iter_next(struct i915_gem_engines_iter *it)
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_context.h b/drivers/gpu/drm/i915/gem/i915_gem_context.h
index 176978608b6f..18e50a769a6e 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_context.h
+++ b/drivers/gpu/drm/i915/gem/i915_gem_context.h
@@ -11,7 +11,9 @@
 
 #include "gt/intel_context.h"
 
+#include "i915_drv.h"
 #include "i915_gem.h"
+#include "i915_gem_gtt.h"
 #include "i915_scheduler.h"
 #include "intel_device_info.h"
 
@@ -74,6 +76,21 @@ static inline void i915_gem_context_clear_recoverable(struct i915_gem_context *c
 	clear_bit(UCONTEXT_RECOVERABLE, &ctx->user_flags);
 }
 
+static inline bool i915_gem_context_is_persistent(const struct i915_gem_context *ctx)
+{
+	return test_bit(UCONTEXT_PERSISTENCE, &ctx->user_flags);
+}
+
+static inline void i915_gem_context_set_persistence(struct i915_gem_context *ctx)
+{
+	set_bit(UCONTEXT_PERSISTENCE, &ctx->user_flags);
+}
+
+static inline void i915_gem_context_clear_persistence(struct i915_gem_context *ctx)
+{
+	clear_bit(UCONTEXT_PERSISTENCE, &ctx->user_flags);
+}
+
 static inline bool i915_gem_context_is_banned(const struct i915_gem_context *ctx)
 {
 	return test_bit(CONTEXT_BANNED, &ctx->flags);
@@ -112,19 +129,22 @@ i915_gem_context_clear_user_engines(struct i915_gem_context *ctx)
 	clear_bit(CONTEXT_USER_ENGINES, &ctx->flags);
 }
 
-int __i915_gem_context_pin_hw_id(struct i915_gem_context *ctx);
-static inline int i915_gem_context_pin_hw_id(struct i915_gem_context *ctx)
+static inline bool
+i915_gem_context_nopreempt(const struct i915_gem_context *ctx)
 {
-	if (atomic_inc_not_zero(&ctx->hw_id_pin_count))
-		return 0;
+	return test_bit(CONTEXT_NOPREEMPT, &ctx->flags);
+}
 
-	return __i915_gem_context_pin_hw_id(ctx);
+static inline void
+i915_gem_context_set_nopreempt(struct i915_gem_context *ctx)
+{
+	set_bit(CONTEXT_NOPREEMPT, &ctx->flags);
 }
 
-static inline void i915_gem_context_unpin_hw_id(struct i915_gem_context *ctx)
+static inline void
+i915_gem_context_clear_nopreempt(struct i915_gem_context *ctx)
 {
-	GEM_BUG_ON(atomic_read(&ctx->hw_id_pin_count) == 0u);
-	atomic_dec(&ctx->hw_id_pin_count);
+	clear_bit(CONTEXT_NOPREEMPT, &ctx->flags);
 }
 
 static inline bool i915_gem_context_is_kernel(struct i915_gem_context *ctx)
@@ -133,8 +153,8 @@ static inline bool i915_gem_context_is_kernel(struct i915_gem_context *ctx)
 }
 
 /* i915_gem_context.c */
-int __must_check i915_gem_contexts_init(struct drm_i915_private *dev_priv);
-void i915_gem_contexts_fini(struct drm_i915_private *dev_priv);
+int __must_check i915_gem_init_contexts(struct drm_i915_private *i915);
+void i915_gem_driver_release__contexts(struct drm_i915_private *i915);
 
 int i915_gem_context_open(struct drm_i915_private *i915,
 			  struct drm_file *file);
@@ -173,6 +193,27 @@ static inline void i915_gem_context_put(struct i915_gem_context *ctx)
 	kref_put(&ctx->ref, i915_gem_context_release);
 }
 
+static inline struct i915_address_space *
+i915_gem_context_vm(struct i915_gem_context *ctx)
+{
+	return rcu_dereference_protected(ctx->vm, lockdep_is_held(&ctx->mutex));
+}
+
+static inline struct i915_address_space *
+i915_gem_context_get_vm_rcu(struct i915_gem_context *ctx)
+{
+	struct i915_address_space *vm;
+
+	rcu_read_lock();
+	vm = rcu_dereference(ctx->vm);
+	if (!vm)
+		vm = &ctx->i915->ggtt.vm;
+	vm = i915_vm_get(vm);
+	rcu_read_unlock();
+
+	return vm;
+}
+
 static inline struct i915_gem_engines *
 i915_gem_context_engines(struct i915_gem_context *ctx)
 {
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_context_types.h b/drivers/gpu/drm/i915/gem/i915_gem_context_types.h
index 260d59cc3de8..3870dd5daaa0 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_context_types.h
+++ b/drivers/gpu/drm/i915/gem/i915_gem_context_types.h
@@ -88,7 +88,7 @@ struct i915_gem_context {
 	 * In other modes, this is a NULL pointer with the expectation that
 	 * the caller uses the shared global GTT.
 	 */
-	struct i915_address_space *vm;
+	struct i915_address_space __rcu *vm;
 
 	/**
 	 * @pid: process id of creator
@@ -137,6 +137,7 @@ struct i915_gem_context {
 #define UCONTEXT_NO_ERROR_CAPTURE	1
 #define UCONTEXT_BANNABLE		2
 #define UCONTEXT_RECOVERABLE		3
+#define UCONTEXT_PERSISTENCE		4
 
 	/**
 	 * @flags: small set of booleans
@@ -146,24 +147,7 @@ struct i915_gem_context {
 #define CONTEXT_CLOSED			1
 #define CONTEXT_FORCE_SINGLE_SUBMISSION	2
 #define CONTEXT_USER_ENGINES		3
-
-	/**
-	 * @hw_id: - unique identifier for the context
-	 *
-	 * The hardware needs to uniquely identify the context for a few
-	 * functions like fault reporting, PASID, scheduling. The
-	 * &drm_i915_private.context_hw_ida is used to assign a unqiue
-	 * id for the lifetime of the context.
-	 *
-	 * @hw_id_pin_count: - number of times this context had been pinned
-	 * for use (should be, at most, once per engine).
-	 *
-	 * @hw_id_link: - all contexts with an assigned id are tracked
-	 * for possible repossession.
-	 */
-	unsigned int hw_id;
-	atomic_t hw_id_pin_count;
-	struct list_head hw_id_link;
+#define CONTEXT_NOPREEMPT		4
 
 	struct mutex mutex;
 
@@ -192,6 +176,13 @@ struct i915_gem_context {
 	 * per vm, which may be one per context or shared with the global GTT)
 	 */
 	struct radix_tree_root handles_vma;
+
+	/** jump_whitelist: Bit array for tracking cmds during cmdparsing
+	 *  Guarded by struct_mutex
+	 */
+	unsigned long *jump_whitelist;
+	/** jump_whitelist_cmds: No of cmd slots available */
+	u32 jump_whitelist_cmds;
 };
 
 #endif /* __I915_GEM_CONTEXT_TYPES_H__ */
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_dmabuf.c b/drivers/gpu/drm/i915/gem/i915_gem_dmabuf.c
index 96ce95c8ac5a..eaea49d08eb5 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_dmabuf.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_dmabuf.c
@@ -256,6 +256,7 @@ static const struct drm_i915_gem_object_ops i915_gem_object_dmabuf_ops = {
 struct drm_gem_object *i915_gem_prime_import(struct drm_device *dev,
 					     struct dma_buf *dma_buf)
 {
+	static struct lock_class_key lock_class;
 	struct dma_buf_attachment *attach;
 	struct drm_i915_gem_object *obj;
 	int ret;
@@ -287,7 +288,7 @@ struct drm_gem_object *i915_gem_prime_import(struct drm_device *dev,
 	}
 
 	drm_gem_private_object_init(dev, &obj->base, dma_buf->size);
-	i915_gem_object_init(obj, &i915_gem_object_dmabuf_ops);
+	i915_gem_object_init(obj, &i915_gem_object_dmabuf_ops, &lock_class);
 	obj->base.import_attach = attach;
 	obj->base.resv = dma_buf->resv;
 
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_domain.c b/drivers/gpu/drm/i915/gem/i915_gem_domain.c
index 9c58e8fac1d9..9937b4c341f1 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_domain.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_domain.c
@@ -27,7 +27,7 @@ static void __i915_gem_object_flush_for_display(struct drm_i915_gem_object *obj)
 
 void i915_gem_object_flush_if_display(struct drm_i915_gem_object *obj)
 {
-	if (!READ_ONCE(obj->pin_global))
+	if (!i915_gem_object_is_framebuffer(obj))
 		return;
 
 	i915_gem_object_lock(obj);
@@ -288,14 +288,21 @@ restart:
 			if (!drm_mm_node_allocated(&vma->node))
 				continue;
 
-			ret = i915_vma_bind(vma, cache_level, PIN_UPDATE);
+			/* Wait for an earlier async bind, need to rewrite it */
+			ret = i915_vma_sync(vma);
+			if (ret)
+				return ret;
+
+			ret = i915_vma_bind(vma, cache_level, PIN_UPDATE, NULL);
 			if (ret)
 				return ret;
 		}
 	}
 
-	list_for_each_entry(vma, &obj->vma.list, obj_link)
-		vma->node.color = cache_level;
+	list_for_each_entry(vma, &obj->vma.list, obj_link) {
+		if (i915_vm_has_cache_coloring(vma->vm))
+			vma->node.color = cache_level;
+	}
 	i915_gem_object_set_cache_coherency(obj, cache_level);
 	obj->cache_dirty = true; /* Always invalidate stale cachelines */
 
@@ -389,16 +396,11 @@ int i915_gem_set_caching_ioctl(struct drm_device *dev, void *data,
 	if (ret)
 		goto out;
 
-	ret = mutex_lock_interruptible(&i915->drm.struct_mutex);
-	if (ret)
-		goto out;
-
 	ret = i915_gem_object_lock_interruptible(obj);
 	if (ret == 0) {
 		ret = i915_gem_object_set_cache_level(obj, level);
 		i915_gem_object_unlock(obj);
 	}
-	mutex_unlock(&i915->drm.struct_mutex);
 
 out:
 	i915_gem_object_put(obj);
@@ -422,12 +424,8 @@ i915_gem_object_pin_to_display_plane(struct drm_i915_gem_object *obj,
 
 	assert_object_held(obj);
 
-	/* Mark the global pin early so that we account for the
-	 * display coherency whilst setting up the cache domains.
-	 */
-	obj->pin_global++;
-
-	/* The display engine is not coherent with the LLC cache on gen6.  As
+	/*
+	 * The display engine is not coherent with the LLC cache on gen6.  As
 	 * a result, we make sure that the pinning that is about to occur is
 	 * done with uncached PTEs. This is lowest common denominator for all
 	 * chipsets.
@@ -439,12 +437,11 @@ i915_gem_object_pin_to_display_plane(struct drm_i915_gem_object *obj,
 	ret = i915_gem_object_set_cache_level(obj,
 					      HAS_WT(to_i915(obj->base.dev)) ?
 					      I915_CACHE_WT : I915_CACHE_NONE);
-	if (ret) {
-		vma = ERR_PTR(ret);
-		goto err_unpin_global;
-	}
+	if (ret)
+		return ERR_PTR(ret);
 
-	/* As the user may map the buffer once pinned in the display plane
+	/*
+	 * As the user may map the buffer once pinned in the display plane
 	 * (e.g. libkms for the bootup splash), we have to ensure that we
 	 * always use map_and_fenceable for all scanout buffers. However,
 	 * it may simply be too big to fit into mappable, in which case
@@ -461,22 +458,19 @@ i915_gem_object_pin_to_display_plane(struct drm_i915_gem_object *obj,
 	if (IS_ERR(vma))
 		vma = i915_gem_object_ggtt_pin(obj, view, 0, alignment, flags);
 	if (IS_ERR(vma))
-		goto err_unpin_global;
+		return vma;
 
 	vma->display_alignment = max_t(u64, vma->display_alignment, alignment);
 
 	__i915_gem_object_flush_for_display(obj);
 
-	/* It should now be out of any other write domains, and we can update
+	/*
+	 * It should now be out of any other write domains, and we can update
 	 * the domain values for our changes.
 	 */
 	obj->read_domains |= I915_GEM_DOMAIN_GTT;
 
 	return vma;
-
-err_unpin_global:
-	obj->pin_global--;
-	return vma;
 }
 
 static void i915_gem_object_bump_inactive_ggtt(struct drm_i915_gem_object *obj)
@@ -491,6 +485,7 @@ static void i915_gem_object_bump_inactive_ggtt(struct drm_i915_gem_object *obj)
 		if (!drm_mm_node_allocated(&vma->node))
 			continue;
 
+		GEM_BUG_ON(vma->vm != &i915->ggtt.vm);
 		list_move_tail(&vma->vm_link, &vma->vm->bound_list);
 	}
 	mutex_unlock(&i915->ggtt.vm.mutex);
@@ -500,7 +495,8 @@ static void i915_gem_object_bump_inactive_ggtt(struct drm_i915_gem_object *obj)
 
 		spin_lock_irqsave(&i915->mm.obj_lock, flags);
 
-		if (obj->mm.madv == I915_MADV_WILLNEED)
+		if (obj->mm.madv == I915_MADV_WILLNEED &&
+		    !atomic_read(&obj->mm.shrink_pin))
 			list_move_tail(&obj->mm.link, &i915->mm.shrink_list);
 
 		spin_unlock_irqrestore(&i915->mm.obj_lock, flags);
@@ -514,12 +510,6 @@ i915_gem_object_unpin_from_display_plane(struct i915_vma *vma)
 
 	assert_object_held(obj);
 
-	if (WARN_ON(obj->pin_global == 0))
-		return;
-
-	if (--obj->pin_global == 0)
-		vma->display_alignment = I915_GTT_MIN_ALIGNMENT;
-
 	/* Bump the LRU to try and avoid premature eviction whilst flipping  */
 	i915_gem_object_bump_inactive_ggtt(obj);
 
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/gem/i915_gem_execbuffer.c
index b5f6937369ea..f0998f1225af 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_execbuffer.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_execbuffer.c
@@ -19,6 +19,7 @@
 #include "gt/intel_engine_pool.h"
 #include "gt/intel_gt.h"
 #include "gt/intel_gt_pm.h"
+#include "gt/intel_ring.h"
 
 #include "i915_drv.h"
 #include "i915_gem_clflush.h"
@@ -252,6 +253,7 @@ struct i915_execbuffer {
 		bool has_fence : 1;
 		bool needs_unfenced : 1;
 
+		struct intel_context *ce;
 		struct i915_request *rq;
 		u32 *rq_cmd;
 		unsigned int rq_size;
@@ -296,7 +298,9 @@ static inline u64 gen8_noncanonical_addr(u64 address)
 
 static inline bool eb_use_cmdparser(const struct i915_execbuffer *eb)
 {
-	return intel_engine_needs_cmd_parser(eb->engine) && eb->batch_len;
+	return intel_engine_requires_cmd_parser(eb->engine) ||
+		(intel_engine_using_cmd_parser(eb->engine) &&
+		 eb->args->batch_len);
 }
 
 static int eb_create(struct i915_execbuffer *eb)
@@ -697,7 +701,9 @@ static int eb_reserve(struct i915_execbuffer *eb)
 
 		case 1:
 			/* Too fragmented, unbind everything and retry */
+			mutex_lock(&eb->context->vm->mutex);
 			err = i915_gem_evict_vm(eb->context->vm);
+			mutex_unlock(&eb->context->vm->mutex);
 			if (err)
 				return err;
 			break;
@@ -725,7 +731,7 @@ static int eb_select_context(struct i915_execbuffer *eb)
 		return -ENOENT;
 
 	eb->gem_context = ctx;
-	if (ctx->vm)
+	if (rcu_access_pointer(ctx->vm))
 		eb->invalid_flags |= EXEC_OBJECT_NEEDS_GTT;
 
 	eb->context_flags = 0;
@@ -880,6 +886,9 @@ static void eb_destroy(const struct i915_execbuffer *eb)
 {
 	GEM_BUG_ON(eb->reloc_cache.rq);
 
+	if (eb->reloc_cache.ce)
+		intel_context_put(eb->reloc_cache.ce);
+
 	if (eb->lut_size > 0)
 		kfree(eb->buckets);
 }
@@ -902,7 +911,8 @@ static void reloc_cache_init(struct reloc_cache *cache,
 	cache->use_64bit_reloc = HAS_64BIT_RELOC(i915);
 	cache->has_fence = cache->gen < 4;
 	cache->needs_unfenced = INTEL_INFO(i915)->unfenced_needs_alignment;
-	cache->node.allocated = false;
+	cache->node.flags = 0;
+	cache->ce = NULL;
 	cache->rq = NULL;
 	cache->rq_size = 0;
 }
@@ -963,11 +973,13 @@ static void reloc_cache_reset(struct reloc_cache *cache)
 		intel_gt_flush_ggtt_writes(ggtt->vm.gt);
 		io_mapping_unmap_atomic((void __iomem *)vaddr);
 
-		if (cache->node.allocated) {
+		if (drm_mm_node_allocated(&cache->node)) {
 			ggtt->vm.clear_range(&ggtt->vm,
 					     cache->node.start,
 					     cache->node.size);
+			mutex_lock(&ggtt->vm.mutex);
 			drm_mm_remove_node(&cache->node);
+			mutex_unlock(&ggtt->vm.mutex);
 		} else {
 			i915_vma_unpin((struct i915_vma *)cache->node.mm);
 		}
@@ -1042,11 +1054,13 @@ static void *reloc_iomap(struct drm_i915_gem_object *obj,
 					       PIN_NOEVICT);
 		if (IS_ERR(vma)) {
 			memset(&cache->node, 0, sizeof(cache->node));
+			mutex_lock(&ggtt->vm.mutex);
 			err = drm_mm_insert_node_in_range
 				(&ggtt->vm.mm, &cache->node,
 				 PAGE_SIZE, 0, I915_COLOR_UNEVICTABLE,
 				 0, ggtt->mappable_end,
 				 DRM_MM_INSERT_LOW);
+			mutex_unlock(&ggtt->vm.mutex);
 			if (err) /* no inactive aperture space, use cpu reloc */
 				return NULL;
 		} else {
@@ -1056,7 +1070,7 @@ static void *reloc_iomap(struct drm_i915_gem_object *obj,
 	}
 
 	offset = cache->node.start;
-	if (cache->node.allocated) {
+	if (drm_mm_node_allocated(&cache->node)) {
 		ggtt->vm.insert_page(&ggtt->vm,
 				     i915_gem_object_get_dma_address(obj, page),
 				     offset, I915_CACHE_NONE, 0);
@@ -1145,7 +1159,7 @@ static int __reloc_gpu_alloc(struct i915_execbuffer *eb,
 	u32 *cmd;
 	int err;
 
-	pool = intel_engine_pool_get(&eb->engine->pool, PAGE_SIZE);
+	pool = intel_engine_get_pool(eb->engine, PAGE_SIZE);
 	if (IS_ERR(pool))
 		return PTR_ERR(pool);
 
@@ -1168,7 +1182,7 @@ static int __reloc_gpu_alloc(struct i915_execbuffer *eb,
 	if (err)
 		goto err_unmap;
 
-	rq = i915_request_create(eb->context);
+	rq = intel_context_create_request(cache->ce);
 	if (IS_ERR(rq)) {
 		err = PTR_ERR(rq);
 		goto err_unpin;
@@ -1239,6 +1253,29 @@ static u32 *reloc_gpu(struct i915_execbuffer *eb,
 		if (!intel_engine_can_store_dword(eb->engine))
 			return ERR_PTR(-ENODEV);
 
+		if (!cache->ce) {
+			struct intel_context *ce;
+
+			/*
+			 * The CS pre-parser can pre-fetch commands across
+			 * memory sync points and starting gen12 it is able to
+			 * pre-fetch across BB_START and BB_END boundaries
+			 * (within the same context). We therefore use a
+			 * separate context gen12+ to guarantee that the reloc
+			 * writes land before the parser gets to the target
+			 * memory location.
+			 */
+			if (cache->gen >= 12)
+				ce = intel_context_create(eb->context->gem_context,
+							  eb->engine);
+			else
+				ce = intel_context_get(eb->context);
+			if (IS_ERR(ce))
+				return ERR_CAST(ce);
+
+			cache->ce = ce;
+		}
+
 		err = __reloc_gpu_alloc(eb, vma, len);
 		if (unlikely(err))
 			return ERR_PTR(err);
@@ -1388,7 +1425,7 @@ eb_relocate_entry(struct i915_execbuffer *eb,
 		if (reloc->write_domain == I915_GEM_DOMAIN_INSTRUCTION &&
 		    IS_GEN(eb->i915, 6)) {
 			err = i915_vma_bind(target, target->obj->cache_level,
-					    PIN_GLOBAL);
+					    PIN_GLOBAL, NULL);
 			if (WARN_ONCE(err,
 				      "Unexpected failure to bind target VMA!"))
 				return err;
@@ -1955,40 +1992,94 @@ static int i915_reset_gen7_sol_offsets(struct i915_request *rq)
 	return 0;
 }
 
-static struct i915_vma *eb_parse(struct i915_execbuffer *eb, bool is_master)
+static struct i915_vma *
+shadow_batch_pin(struct i915_execbuffer *eb, struct drm_i915_gem_object *obj)
+{
+	struct drm_i915_private *dev_priv = eb->i915;
+	struct i915_vma * const vma = *eb->vma;
+	struct i915_address_space *vm;
+	u64 flags;
+
+	/*
+	 * PPGTT backed shadow buffers must be mapped RO, to prevent
+	 * post-scan tampering
+	 */
+	if (CMDPARSER_USES_GGTT(dev_priv)) {
+		flags = PIN_GLOBAL;
+		vm = &dev_priv->ggtt.vm;
+	} else if (vma->vm->has_read_only) {
+		flags = PIN_USER;
+		vm = vma->vm;
+		i915_gem_object_set_readonly(obj);
+	} else {
+		DRM_DEBUG("Cannot prevent post-scan tampering without RO capable vm\n");
+		return ERR_PTR(-EINVAL);
+	}
+
+	return i915_gem_object_pin(obj, vm, NULL, 0, 0, flags);
+}
+
+static struct i915_vma *eb_parse(struct i915_execbuffer *eb)
 {
 	struct intel_engine_pool_node *pool;
 	struct i915_vma *vma;
+	u64 batch_start;
+	u64 shadow_batch_start;
 	int err;
 
-	pool = intel_engine_pool_get(&eb->engine->pool, eb->batch_len);
+	pool = intel_engine_get_pool(eb->engine, eb->batch_len);
 	if (IS_ERR(pool))
 		return ERR_CAST(pool);
 
-	err = intel_engine_cmd_parser(eb->engine,
+	vma = shadow_batch_pin(eb, pool->obj);
+	if (IS_ERR(vma))
+		goto err;
+
+	batch_start = gen8_canonical_addr(eb->batch->node.start) +
+		      eb->batch_start_offset;
+
+	shadow_batch_start = gen8_canonical_addr(vma->node.start);
+
+	err = intel_engine_cmd_parser(eb->gem_context,
+				      eb->engine,
 				      eb->batch->obj,
-				      pool->obj,
+				      batch_start,
 				      eb->batch_start_offset,
 				      eb->batch_len,
-				      is_master);
+				      pool->obj,
+				      shadow_batch_start);
+
 	if (err) {
-		if (err == -EACCES) /* unhandled chained batch */
+		i915_vma_unpin(vma);
+
+		/*
+		 * Unsafe GGTT-backed buffers can still be submitted safely
+		 * as non-secure.
+		 * For PPGTT backing however, we have no choice but to forcibly
+		 * reject unsafe buffers
+		 */
+		if (CMDPARSER_USES_GGTT(eb->i915) && (err == -EACCES))
+			/* Execute original buffer non-secure */
 			vma = NULL;
 		else
 			vma = ERR_PTR(err);
 		goto err;
 	}
 
-	vma = i915_gem_object_ggtt_pin(pool->obj, NULL, 0, 0, 0);
-	if (IS_ERR(vma))
-		goto err;
-
 	eb->vma[eb->buffer_count] = i915_vma_get(vma);
 	eb->flags[eb->buffer_count] =
 		__EXEC_OBJECT_HAS_PIN | __EXEC_OBJECT_HAS_REF;
 	vma->exec_flags = &eb->flags[eb->buffer_count];
 	eb->buffer_count++;
 
+	eb->batch_start_offset = 0;
+	eb->batch = vma;
+
+	if (CMDPARSER_USES_GGTT(eb->i915))
+		eb->batch_flags |= I915_DISPATCH_SECURE;
+
+	/* eb->batch_len unchanged */
+
 	vma->private = pool;
 	return vma;
 
@@ -2043,6 +2134,9 @@ static int eb_submit(struct i915_execbuffer *eb)
 	if (err)
 		return err;
 
+	if (i915_gem_context_nopreempt(eb->gem_context))
+		eb->request->flags |= I915_REQUEST_NOPREEMPT;
+
 	return 0;
 }
 
@@ -2112,35 +2206,6 @@ static struct i915_request *eb_throttle(struct intel_context *ce)
 	return i915_request_get(rq);
 }
 
-static int
-__eb_pin_context(struct i915_execbuffer *eb, struct intel_context *ce)
-{
-	int err;
-
-	if (likely(atomic_inc_not_zero(&ce->pin_count)))
-		return 0;
-
-	err = mutex_lock_interruptible(&eb->i915->drm.struct_mutex);
-	if (err)
-		return err;
-
-	err = __intel_context_do_pin(ce);
-	mutex_unlock(&eb->i915->drm.struct_mutex);
-
-	return err;
-}
-
-static void
-__eb_unpin_context(struct i915_execbuffer *eb, struct intel_context *ce)
-{
-	if (likely(atomic_add_unless(&ce->pin_count, -1, 1)))
-		return;
-
-	mutex_lock(&eb->i915->drm.struct_mutex);
-	intel_context_unpin(ce);
-	mutex_unlock(&eb->i915->drm.struct_mutex);
-}
-
 static int __eb_pin_engine(struct i915_execbuffer *eb, struct intel_context *ce)
 {
 	struct intel_timeline *tl;
@@ -2160,7 +2225,7 @@ static int __eb_pin_engine(struct i915_execbuffer *eb, struct intel_context *ce)
 	 * GGTT space, so do this first before we reserve a seqno for
 	 * ourselves.
 	 */
-	err = __eb_pin_context(eb, ce);
+	err = intel_context_pin(ce);
 	if (err)
 		return err;
 
@@ -2204,7 +2269,7 @@ err_exit:
 	intel_context_exit(ce);
 	intel_context_timeline_unlock(tl);
 err_unpin:
-	__eb_unpin_context(eb, ce);
+	intel_context_unpin(ce);
 	return err;
 }
 
@@ -2217,7 +2282,7 @@ static void eb_unpin_engine(struct i915_execbuffer *eb)
 	intel_context_exit(ce);
 	mutex_unlock(&tl->mutex);
 
-	__eb_unpin_context(eb, ce);
+	intel_context_unpin(ce);
 }
 
 static unsigned int
@@ -2421,6 +2486,7 @@ i915_gem_do_execbuffer(struct drm_device *dev,
 		       struct drm_i915_gem_exec_object2 *exec,
 		       struct drm_syncobj **fences)
 {
+	struct drm_i915_private *i915 = to_i915(dev);
 	struct i915_execbuffer eb;
 	struct dma_fence *in_fence = NULL;
 	struct dma_fence *exec_fence = NULL;
@@ -2432,7 +2498,7 @@ i915_gem_do_execbuffer(struct drm_device *dev,
 	BUILD_BUG_ON(__EXEC_OBJECT_INTERNAL_FLAGS &
 		     ~__EXEC_OBJECT_UNKNOWN_FLAGS);
 
-	eb.i915 = to_i915(dev);
+	eb.i915 = i915;
 	eb.file = file;
 	eb.args = args;
 	if (DBG_FORCE_RELOC || !(args->flags & I915_EXEC_NO_RELOC))
@@ -2452,8 +2518,15 @@ i915_gem_do_execbuffer(struct drm_device *dev,
 
 	eb.batch_flags = 0;
 	if (args->flags & I915_EXEC_SECURE) {
+		if (INTEL_GEN(i915) >= 11)
+			return -ENODEV;
+
+		/* Return -EPERM to trigger fallback code on old binaries. */
+		if (!HAS_SECURE_BATCHES(i915))
+			return -EPERM;
+
 		if (!drm_is_current_master(file) || !capable(CAP_SYS_ADMIN))
-		    return -EPERM;
+			return -EPERM;
 
 		eb.batch_flags |= I915_DISPATCH_SECURE;
 	}
@@ -2530,34 +2603,19 @@ i915_gem_do_execbuffer(struct drm_device *dev,
 		goto err_vma;
 	}
 
+	if (eb.batch_len == 0)
+		eb.batch_len = eb.batch->size - eb.batch_start_offset;
+
 	if (eb_use_cmdparser(&eb)) {
 		struct i915_vma *vma;
 
-		vma = eb_parse(&eb, drm_is_current_master(file));
+		vma = eb_parse(&eb);
 		if (IS_ERR(vma)) {
 			err = PTR_ERR(vma);
 			goto err_vma;
 		}
-
-		if (vma) {
-			/*
-			 * Batch parsed and accepted:
-			 *
-			 * Set the DISPATCH_SECURE bit to remove the NON_SECURE
-			 * bit from MI_BATCH_BUFFER_START commands issued in
-			 * the dispatch_execbuffer implementations. We
-			 * specifically don't want that set on batches the
-			 * command parser has accepted.
-			 */
-			eb.batch_flags |= I915_DISPATCH_SECURE;
-			eb.batch_start_offset = 0;
-			eb.batch = vma;
-		}
 	}
 
-	if (eb.batch_len == 0)
-		eb.batch_len = eb.batch->size - eb.batch_start_offset;
-
 	/*
 	 * snb/ivb/vlv conflate the "batch in ppgtt" bit with the "non-secure
 	 * batch" bit. Hence we need to pin secure batches into the global gtt.
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_internal.c b/drivers/gpu/drm/i915/gem/i915_gem_internal.c
index 0c41e04ab8fa..9cfb0e41ff06 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_internal.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_internal.c
@@ -117,13 +117,6 @@ create_st:
 		goto err;
 	}
 
-	/* Mark the pages as dontneed whilst they are still pinned. As soon
-	 * as they are unpinned they are allowed to be reaped by the shrinker,
-	 * and the caller is expected to repopulate - the contents of this
-	 * object are only valid whilst active and pinned.
-	 */
-	obj->mm.madv = I915_MADV_DONTNEED;
-
 	__i915_gem_object_set_pages(obj, st, sg_page_sizes);
 
 	return 0;
@@ -143,7 +136,6 @@ static void i915_gem_object_put_pages_internal(struct drm_i915_gem_object *obj,
 	internal_free_pages(pages);
 
 	obj->mm.dirty = false;
-	obj->mm.madv = I915_MADV_WILLNEED;
 }
 
 static const struct drm_i915_gem_object_ops i915_gem_object_internal_ops = {
@@ -172,6 +164,7 @@ struct drm_i915_gem_object *
 i915_gem_object_create_internal(struct drm_i915_private *i915,
 				phys_addr_t size)
 {
+	static struct lock_class_key lock_class;
 	struct drm_i915_gem_object *obj;
 	unsigned int cache_level;
 
@@ -186,7 +179,16 @@ i915_gem_object_create_internal(struct drm_i915_private *i915,
 		return ERR_PTR(-ENOMEM);
 
 	drm_gem_private_object_init(&i915->drm, &obj->base, size);
-	i915_gem_object_init(obj, &i915_gem_object_internal_ops);
+	i915_gem_object_init(obj, &i915_gem_object_internal_ops, &lock_class);
+
+	/*
+	 * Mark the object as volatile, such that the pages are marked as
+	 * dontneed whilst they are still pinned. As soon as they are unpinned
+	 * they are allowed to be reaped by the shrinker, and the caller is
+	 * expected to repopulate - the contents of this object are only valid
+	 * whilst active and pinned.
+	 */
+	i915_gem_object_set_volatile(obj);
 
 	obj->read_domains = I915_GEM_DOMAIN_CPU;
 	obj->write_domain = I915_GEM_DOMAIN_CPU;
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_lmem.c b/drivers/gpu/drm/i915/gem/i915_gem_lmem.c
new file mode 100644
index 000000000000..0e2bf6b7e143
--- /dev/null
+++ b/drivers/gpu/drm/i915/gem/i915_gem_lmem.c
@@ -0,0 +1,99 @@
+// SPDX-License-Identifier: MIT
+/*
+ * Copyright © 2019 Intel Corporation
+ */
+
+#include "intel_memory_region.h"
+#include "gem/i915_gem_region.h"
+#include "gem/i915_gem_lmem.h"
+#include "i915_drv.h"
+
+const struct drm_i915_gem_object_ops i915_gem_lmem_obj_ops = {
+	.flags = I915_GEM_OBJECT_HAS_IOMEM,
+
+	.get_pages = i915_gem_object_get_pages_buddy,
+	.put_pages = i915_gem_object_put_pages_buddy,
+	.release = i915_gem_object_release_memory_region,
+};
+
+/* XXX: Time to vfunc your life up? */
+void __iomem *
+i915_gem_object_lmem_io_map_page(struct drm_i915_gem_object *obj,
+				 unsigned long n)
+{
+	resource_size_t offset;
+
+	offset = i915_gem_object_get_dma_address(obj, n);
+	offset -= obj->mm.region->region.start;
+
+	return io_mapping_map_wc(&obj->mm.region->iomap, offset, PAGE_SIZE);
+}
+
+void __iomem *
+i915_gem_object_lmem_io_map_page_atomic(struct drm_i915_gem_object *obj,
+					unsigned long n)
+{
+	resource_size_t offset;
+
+	offset = i915_gem_object_get_dma_address(obj, n);
+	offset -= obj->mm.region->region.start;
+
+	return io_mapping_map_atomic_wc(&obj->mm.region->iomap, offset);
+}
+
+void __iomem *
+i915_gem_object_lmem_io_map(struct drm_i915_gem_object *obj,
+			    unsigned long n,
+			    unsigned long size)
+{
+	resource_size_t offset;
+
+	GEM_BUG_ON(!i915_gem_object_is_contiguous(obj));
+
+	offset = i915_gem_object_get_dma_address(obj, n);
+	offset -= obj->mm.region->region.start;
+
+	return io_mapping_map_wc(&obj->mm.region->iomap, offset, size);
+}
+
+bool i915_gem_object_is_lmem(struct drm_i915_gem_object *obj)
+{
+	return obj->ops == &i915_gem_lmem_obj_ops;
+}
+
+struct drm_i915_gem_object *
+i915_gem_object_create_lmem(struct drm_i915_private *i915,
+			    resource_size_t size,
+			    unsigned int flags)
+{
+	return i915_gem_object_create_region(i915->mm.regions[INTEL_REGION_LMEM],
+					     size, flags);
+}
+
+struct drm_i915_gem_object *
+__i915_gem_lmem_object_create(struct intel_memory_region *mem,
+			      resource_size_t size,
+			      unsigned int flags)
+{
+	static struct lock_class_key lock_class;
+	struct drm_i915_private *i915 = mem->i915;
+	struct drm_i915_gem_object *obj;
+
+	if (size > BIT(mem->mm.max_order) * mem->mm.chunk_size)
+		return ERR_PTR(-E2BIG);
+
+	obj = i915_gem_object_alloc();
+	if (!obj)
+		return ERR_PTR(-ENOMEM);
+
+	drm_gem_private_object_init(&i915->drm, &obj->base, size);
+	i915_gem_object_init(obj, &i915_gem_lmem_obj_ops, &lock_class);
+
+	obj->read_domains = I915_GEM_DOMAIN_WC | I915_GEM_DOMAIN_GTT;
+
+	i915_gem_object_set_cache_coherency(obj, I915_CACHE_NONE);
+
+	i915_gem_object_init_memory_region(obj, mem, flags);
+
+	return obj;
+}
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_lmem.h b/drivers/gpu/drm/i915/gem/i915_gem_lmem.h
new file mode 100644
index 000000000000..7c176b8b7d2f
--- /dev/null
+++ b/drivers/gpu/drm/i915/gem/i915_gem_lmem.h
@@ -0,0 +1,37 @@
+/* SPDX-License-Identifier: MIT */
+/*
+ * Copyright © 2019 Intel Corporation
+ */
+
+#ifndef __I915_GEM_LMEM_H
+#define __I915_GEM_LMEM_H
+
+#include <linux/types.h>
+
+struct drm_i915_private;
+struct drm_i915_gem_object;
+struct intel_memory_region;
+
+extern const struct drm_i915_gem_object_ops i915_gem_lmem_obj_ops;
+
+void __iomem *i915_gem_object_lmem_io_map(struct drm_i915_gem_object *obj,
+					  unsigned long n, unsigned long size);
+void __iomem *i915_gem_object_lmem_io_map_page(struct drm_i915_gem_object *obj,
+					       unsigned long n);
+void __iomem *
+i915_gem_object_lmem_io_map_page_atomic(struct drm_i915_gem_object *obj,
+					unsigned long n);
+
+bool i915_gem_object_is_lmem(struct drm_i915_gem_object *obj);
+
+struct drm_i915_gem_object *
+i915_gem_object_create_lmem(struct drm_i915_private *i915,
+			    resource_size_t size,
+			    unsigned int flags);
+
+struct drm_i915_gem_object *
+__i915_gem_lmem_object_create(struct intel_memory_region *mem,
+			      resource_size_t size,
+			      unsigned int flags);
+
+#endif /* !__I915_GEM_LMEM_H */
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_mman.c b/drivers/gpu/drm/i915/gem/i915_gem_mman.c
index 261c9bd83f51..e3002849844b 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_mman.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_mman.c
@@ -8,6 +8,7 @@
 #include <linux/sizes.h>
 
 #include "gt/intel_gt.h"
+#include "gt/intel_gt_requests.h"
 
 #include "i915_drv.h"
 #include "i915_gem_gtt.h"
@@ -245,21 +246,9 @@ vm_fault_t i915_gem_fault(struct vm_fault *vmf)
 
 	wakeref = intel_runtime_pm_get(rpm);
 
-	srcu = intel_gt_reset_trylock(ggtt->vm.gt);
-	if (srcu < 0) {
-		ret = srcu;
-		goto err_rpm;
-	}
-
-	ret = i915_mutex_lock_interruptible(dev);
+	ret = intel_gt_reset_trylock(ggtt->vm.gt, &srcu);
 	if (ret)
-		goto err_reset;
-
-	/* Access to snoopable pages through the GTT is incoherent. */
-	if (obj->cache_level != I915_CACHE_NONE && !HAS_LLC(i915)) {
-		ret = -EFAULT;
-		goto err_unlock;
-	}
+		goto err_rpm;
 
 	/* Now pin it into the GTT as needed */
 	vma = i915_gem_object_ggtt_pin(obj, NULL, 0, 0,
@@ -287,10 +276,19 @@ vm_fault_t i915_gem_fault(struct vm_fault *vmf)
 			view.type = I915_GGTT_VIEW_PARTIAL;
 			vma = i915_gem_object_ggtt_pin(obj, &view, 0, 0, flags);
 		}
+
+		/* The entire mappable GGTT is pinned? Unexpected! */
+		GEM_BUG_ON(vma == ERR_PTR(-ENOSPC));
 	}
 	if (IS_ERR(vma)) {
 		ret = PTR_ERR(vma);
-		goto err_unlock;
+		goto err_reset;
+	}
+
+	/* Access to snoopable pages through the GTT is incoherent. */
+	if (obj->cache_level != I915_CACHE_NONE && !HAS_LLC(i915)) {
+		ret = -EFAULT;
+		goto err_unpin;
 	}
 
 	ret = i915_vma_pin_fence(vma);
@@ -314,18 +312,20 @@ vm_fault_t i915_gem_fault(struct vm_fault *vmf)
 		list_add(&obj->userfault_link, &i915->ggtt.userfault_list);
 	mutex_unlock(&i915->ggtt.vm.mutex);
 
-	if (CONFIG_DRM_I915_USERFAULT_AUTOSUSPEND)
+	if (IS_ACTIVE(CONFIG_DRM_I915_USERFAULT_AUTOSUSPEND))
 		intel_wakeref_auto(&i915->ggtt.userfault_wakeref,
 				   msecs_to_jiffies_timeout(CONFIG_DRM_I915_USERFAULT_AUTOSUSPEND));
 
-	i915_vma_set_ggtt_write(vma);
+	if (write) {
+		GEM_BUG_ON(!i915_gem_object_has_pinned_pages(obj));
+		i915_vma_set_ggtt_write(vma);
+		obj->mm.dirty = true;
+	}
 
 err_fence:
 	i915_vma_unpin_fence(vma);
 err_unpin:
 	__i915_vma_unpin(vma);
-err_unlock:
-	mutex_unlock(&dev->struct_mutex);
 err_reset:
 	intel_gt_reset_unlock(ggtt->vm.gt, srcu);
 err_rpm:
@@ -333,23 +333,20 @@ err_rpm:
 	i915_gem_object_unpin_pages(obj);
 err:
 	switch (ret) {
-	case -EIO:
-		/*
-		 * We eat errors when the gpu is terminally wedged to avoid
-		 * userspace unduly crashing (gl has no provisions for mmaps to
-		 * fail). But any other -EIO isn't ours (e.g. swap in failure)
-		 * and so needs to be reported.
-		 */
-		if (!intel_gt_is_wedged(ggtt->vm.gt))
-			return VM_FAULT_SIGBUS;
-		/* else, fall through */
-	case -EAGAIN:
-		/*
-		 * EAGAIN means the gpu is hung and we'll wait for the error
-		 * handler to reset everything when re-faulting in
-		 * i915_mutex_lock_interruptible.
-		 */
+	default:
+		WARN_ONCE(ret, "unhandled error in %s: %i\n", __func__, ret);
+		/* fallthrough */
+	case -EIO: /* shmemfs failure from swap device */
+	case -EFAULT: /* purged object */
+	case -ENODEV: /* bad object, how did you get here! */
+		return VM_FAULT_SIGBUS;
+
+	case -ENOSPC: /* shmemfs allocation failure */
+	case -ENOMEM: /* our allocation failure */
+		return VM_FAULT_OOM;
+
 	case 0:
+	case -EAGAIN:
 	case -ERESTARTSYS:
 	case -EINTR:
 	case -EBUSY:
@@ -358,14 +355,6 @@ err:
 		 * already did the job.
 		 */
 		return VM_FAULT_NOPAGE;
-	case -ENOMEM:
-		return VM_FAULT_OOM;
-	case -ENOSPC:
-	case -EFAULT:
-		return VM_FAULT_SIGBUS;
-	default:
-		WARN_ONCE(ret, "unhandled error in %s: %i\n", __func__, ret);
-		return VM_FAULT_SIGBUS;
 	}
 }
 
@@ -436,6 +425,7 @@ out:
 static int create_mmap_offset(struct drm_i915_gem_object *obj)
 {
 	struct drm_i915_private *i915 = to_i915(obj->base.dev);
+	struct intel_gt *gt = &i915->gt;
 	int err;
 
 	err = drm_gem_create_mmap_offset(&obj->base);
@@ -443,21 +433,12 @@ static int create_mmap_offset(struct drm_i915_gem_object *obj)
 		return 0;
 
 	/* Attempt to reap some mmap space from dead objects */
-	do {
-		err = i915_gem_wait_for_idle(i915,
-					     I915_WAIT_INTERRUPTIBLE,
-					     MAX_SCHEDULE_TIMEOUT);
-		if (err)
-			break;
-
-		i915_gem_drain_freed_objects(i915);
-		err = drm_gem_create_mmap_offset(&obj->base);
-		if (!err)
-			break;
+	err = intel_gt_retire_requests_timeout(gt, MAX_SCHEDULE_TIMEOUT);
+	if (err)
+		return err;
 
-	} while (flush_delayed_work(&i915->gem.retire_work));
-
-	return err;
+	i915_gem_drain_freed_objects(i915);
+	return drm_gem_create_mmap_offset(&obj->base);
 }
 
 int
@@ -473,10 +454,16 @@ i915_gem_mmap_gtt(struct drm_file *file,
 	if (!obj)
 		return -ENOENT;
 
+	if (i915_gem_object_never_bind_ggtt(obj)) {
+		ret = -ENODEV;
+		goto out;
+	}
+
 	ret = create_mmap_offset(obj);
 	if (ret == 0)
 		*offset = drm_vma_node_offset_addr(&obj->base.vma_node);
 
+out:
 	i915_gem_object_put(obj);
 	return ret;
 }
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_object.c b/drivers/gpu/drm/i915/gem/i915_gem_object.c
index d7855dc5a5c5..a50296cce0d8 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_object.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_object.c
@@ -47,9 +47,10 @@ void i915_gem_object_free(struct drm_i915_gem_object *obj)
 }
 
 void i915_gem_object_init(struct drm_i915_gem_object *obj,
-			  const struct drm_i915_gem_object_ops *ops)
+			  const struct drm_i915_gem_object_ops *ops,
+			  struct lock_class_key *key)
 {
-	mutex_init(&obj->mm.lock);
+	__mutex_init(&obj->mm.lock, "obj->mm.lock", key);
 
 	spin_lock_init(&obj->vma.lock);
 	INIT_LIST_HEAD(&obj->vma.list);
@@ -155,21 +156,30 @@ static void __i915_gem_free_objects(struct drm_i915_private *i915,
 
 	wakeref = intel_runtime_pm_get(&i915->runtime_pm);
 	llist_for_each_entry_safe(obj, on, freed, freed) {
-		struct i915_vma *vma, *vn;
-
 		trace_i915_gem_object_destroy(obj);
 
-		mutex_lock(&i915->drm.struct_mutex);
-
-		list_for_each_entry_safe(vma, vn, &obj->vma.list, obj_link) {
-			GEM_BUG_ON(i915_vma_is_active(vma));
-			vma->flags &= ~I915_VMA_PIN_MASK;
-			i915_vma_destroy(vma);
+		if (!list_empty(&obj->vma.list)) {
+			struct i915_vma *vma;
+
+			/*
+			 * Note that the vma keeps an object reference while
+			 * it is active, so it *should* not sleep while we
+			 * destroy it. Our debug code errs insits it *might*.
+			 * For the moment, play along.
+			 */
+			spin_lock(&obj->vma.lock);
+			while ((vma = list_first_entry_or_null(&obj->vma.list,
+							       struct i915_vma,
+							       obj_link))) {
+				GEM_BUG_ON(vma->obj != obj);
+				spin_unlock(&obj->vma.lock);
+
+				i915_vma_destroy(vma);
+
+				spin_lock(&obj->vma.lock);
+			}
+			spin_unlock(&obj->vma.lock);
 		}
-		GEM_BUG_ON(!list_empty(&obj->vma.list));
-		GEM_BUG_ON(!RB_EMPTY_ROOT(&obj->vma.tree));
-
-		mutex_unlock(&i915->drm.struct_mutex);
 
 		GEM_BUG_ON(atomic_read(&obj->bind_count));
 		GEM_BUG_ON(obj->userfault_count);
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_object.h b/drivers/gpu/drm/i915/gem/i915_gem_object.h
index 5efb9936e05b..458cd51331f1 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_object.h
+++ b/drivers/gpu/drm/i915/gem/i915_gem_object.h
@@ -23,12 +23,14 @@ struct drm_i915_gem_object *i915_gem_object_alloc(void);
 void i915_gem_object_free(struct drm_i915_gem_object *obj);
 
 void i915_gem_object_init(struct drm_i915_gem_object *obj,
-			  const struct drm_i915_gem_object_ops *ops);
+			  const struct drm_i915_gem_object_ops *ops,
+			  struct lock_class_key *key);
 struct drm_i915_gem_object *
-i915_gem_object_create_shmem(struct drm_i915_private *i915, u64 size);
+i915_gem_object_create_shmem(struct drm_i915_private *i915,
+			     resource_size_t size);
 struct drm_i915_gem_object *
 i915_gem_object_create_shmem_from_data(struct drm_i915_private *i915,
-				       const void *data, size_t size);
+				       const void *data, resource_size_t size);
 
 extern const struct drm_i915_gem_object_ops i915_gem_shmem_ops;
 void __i915_gem_object_release_shmem(struct drm_i915_gem_object *obj,
@@ -106,6 +108,11 @@ static inline void i915_gem_object_lock(struct drm_i915_gem_object *obj)
 	dma_resv_lock(obj->base.resv, NULL);
 }
 
+static inline bool i915_gem_object_trylock(struct drm_i915_gem_object *obj)
+{
+	return dma_resv_trylock(obj->base.resv);
+}
+
 static inline int
 i915_gem_object_lock_interruptible(struct drm_i915_gem_object *obj)
 {
@@ -135,27 +142,58 @@ i915_gem_object_is_readonly(const struct drm_i915_gem_object *obj)
 }
 
 static inline bool
+i915_gem_object_is_contiguous(const struct drm_i915_gem_object *obj)
+{
+	return obj->flags & I915_BO_ALLOC_CONTIGUOUS;
+}
+
+static inline bool
+i915_gem_object_is_volatile(const struct drm_i915_gem_object *obj)
+{
+	return obj->flags & I915_BO_ALLOC_VOLATILE;
+}
+
+static inline void
+i915_gem_object_set_volatile(struct drm_i915_gem_object *obj)
+{
+	obj->flags |= I915_BO_ALLOC_VOLATILE;
+}
+
+static inline bool
+i915_gem_object_type_has(const struct drm_i915_gem_object *obj,
+			 unsigned long flags)
+{
+	return obj->ops->flags & flags;
+}
+
+static inline bool
 i915_gem_object_has_struct_page(const struct drm_i915_gem_object *obj)
 {
-	return obj->ops->flags & I915_GEM_OBJECT_HAS_STRUCT_PAGE;
+	return i915_gem_object_type_has(obj, I915_GEM_OBJECT_HAS_STRUCT_PAGE);
 }
 
 static inline bool
 i915_gem_object_is_shrinkable(const struct drm_i915_gem_object *obj)
 {
-	return obj->ops->flags & I915_GEM_OBJECT_IS_SHRINKABLE;
+	return i915_gem_object_type_has(obj, I915_GEM_OBJECT_IS_SHRINKABLE);
 }
 
 static inline bool
 i915_gem_object_is_proxy(const struct drm_i915_gem_object *obj)
 {
-	return obj->ops->flags & I915_GEM_OBJECT_IS_PROXY;
+	return i915_gem_object_type_has(obj, I915_GEM_OBJECT_IS_PROXY);
+}
+
+static inline bool
+i915_gem_object_never_bind_ggtt(const struct drm_i915_gem_object *obj)
+{
+	return i915_gem_object_type_has(obj, I915_GEM_OBJECT_NO_GGTT);
 }
 
 static inline bool
 i915_gem_object_needs_async_cancel(const struct drm_i915_gem_object *obj)
 {
-	return obj->ops->flags & I915_GEM_OBJECT_ASYNC_CANCEL;
+	return i915_gem_object_type_has(obj, I915_GEM_OBJECT_ASYNC_CANCEL);
 }
 
 static inline bool
@@ -406,7 +444,8 @@ static inline bool cpu_write_needs_clflush(struct drm_i915_gem_object *obj)
 	if (!(obj->cache_coherent & I915_BO_CACHE_COHERENT_FOR_WRITE))
 		return true;
 
-	return obj->pin_global; /* currently in use by HW, keep flushed */
+	/* Currently in use by HW (display engine)? Keep flushed. */
+	return i915_gem_object_is_framebuffer(obj);
 }
 
 static inline void __start_cpu_write(struct drm_i915_gem_object *obj)
@@ -423,6 +462,5 @@ int i915_gem_object_wait(struct drm_i915_gem_object *obj,
 int i915_gem_object_wait_priority(struct drm_i915_gem_object *obj,
 				  unsigned int flags,
 				  const struct i915_sched_attr *attr);
-#define I915_PRIORITY_DISPLAY I915_USER_PRIORITY(I915_PRIORITY_MAX)
 
 #endif
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_object_blt.c b/drivers/gpu/drm/i915/gem/i915_gem_object_blt.c
index 6415f9a17e2d..70809d8897cd 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_object_blt.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_object_blt.c
@@ -8,6 +8,7 @@
 #include "gt/intel_engine_pm.h"
 #include "gt/intel_engine_pool.h"
 #include "gt/intel_gt.h"
+#include "gt/intel_ring.h"
 #include "i915_gem_clflush.h"
 #include "i915_gem_object_blt.h"
 
@@ -16,7 +17,7 @@ struct i915_vma *intel_emit_vma_fill_blt(struct intel_context *ce,
 					 u32 value)
 {
 	struct drm_i915_private *i915 = ce->vm->i915;
-	const u32 block_size = S16_MAX * PAGE_SIZE;
+	const u32 block_size = SZ_8M; /* ~1ms at 8GiB/s preemption delay */
 	struct intel_engine_pool_node *pool;
 	struct i915_vma *batch;
 	u64 offset;
@@ -29,10 +30,10 @@ struct i915_vma *intel_emit_vma_fill_blt(struct intel_context *ce,
 	GEM_BUG_ON(intel_engine_is_virtual(ce->engine));
 	intel_engine_pm_get(ce->engine);
 
-	count = div_u64(vma->size, block_size);
+	count = div_u64(round_up(vma->size, block_size), block_size);
 	size = (1 + 8 * count) * sizeof(u32);
 	size = round_up(size, PAGE_SIZE);
-	pool = intel_engine_pool_get(&ce->engine->pool, size);
+	pool = intel_engine_get_pool(ce->engine, size);
 	if (IS_ERR(pool)) {
 		err = PTR_ERR(pool);
 		goto out_pm;
@@ -200,7 +201,7 @@ struct i915_vma *intel_emit_vma_copy_blt(struct intel_context *ce,
 					 struct i915_vma *dst)
 {
 	struct drm_i915_private *i915 = ce->vm->i915;
-	const u32 block_size = S16_MAX * PAGE_SIZE;
+	const u32 block_size = SZ_8M; /* ~1ms at 8GiB/s preemption delay */
 	struct intel_engine_pool_node *pool;
 	struct i915_vma *batch;
 	u64 src_offset, dst_offset;
@@ -213,10 +214,10 @@ struct i915_vma *intel_emit_vma_copy_blt(struct intel_context *ce,
 	GEM_BUG_ON(intel_engine_is_virtual(ce->engine));
 	intel_engine_pm_get(ce->engine);
 
-	count = div_u64(dst->size, block_size);
+	count = div_u64(round_up(dst->size, block_size), block_size);
 	size = (1 + 11 * count) * sizeof(u32);
 	size = round_up(size, PAGE_SIZE);
-	pool = intel_engine_pool_get(&ce->engine->pool, size);
+	pool = intel_engine_get_pool(ce->engine, size);
 	if (IS_ERR(pool)) {
 		err = PTR_ERR(pool);
 		goto out_pm;
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_object_types.h b/drivers/gpu/drm/i915/gem/i915_gem_object_types.h
index ede0eb4218a8..96008374a412 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_object_types.h
+++ b/drivers/gpu/drm/i915/gem/i915_gem_object_types.h
@@ -8,6 +8,7 @@
 #define __I915_GEM_OBJECT_TYPES_H__
 
 #include <drm/drm_gem.h>
+#include <uapi/drm/i915_drm.h>
 
 #include "i915_active.h"
 #include "i915_selftest.h"
@@ -30,9 +31,11 @@ struct i915_lut_handle {
 struct drm_i915_gem_object_ops {
 	unsigned int flags;
 #define I915_GEM_OBJECT_HAS_STRUCT_PAGE	BIT(0)
-#define I915_GEM_OBJECT_IS_SHRINKABLE	BIT(1)
-#define I915_GEM_OBJECT_IS_PROXY	BIT(2)
-#define I915_GEM_OBJECT_ASYNC_CANCEL	BIT(3)
+#define I915_GEM_OBJECT_HAS_IOMEM	BIT(1)
+#define I915_GEM_OBJECT_IS_SHRINKABLE	BIT(2)
+#define I915_GEM_OBJECT_IS_PROXY	BIT(3)
+#define I915_GEM_OBJECT_NO_GGTT		BIT(4)
+#define I915_GEM_OBJECT_ASYNC_CANCEL	BIT(5)
 
 	/* Interface between the GEM object and its backing storage.
 	 * get_pages() is called once prior to the use of the associated set
@@ -117,6 +120,11 @@ struct drm_i915_gem_object {
 
 	I915_SELFTEST_DECLARE(struct list_head st_link);
 
+	unsigned long flags;
+#define I915_BO_ALLOC_CONTIGUOUS BIT(0)
+#define I915_BO_ALLOC_VOLATILE   BIT(1)
+#define I915_BO_ALLOC_FLAGS (I915_BO_ALLOC_CONTIGUOUS | I915_BO_ALLOC_VOLATILE)
+
 	/*
 	 * Is the object to be mapped as read-only to the GPU
 	 * Only honoured if hardware has relevant pte bit
@@ -152,17 +160,30 @@ struct drm_i915_gem_object {
 
 	/** Count of VMA actually bound by this object */
 	atomic_t bind_count;
-	/** Count of how many global VMA are currently pinned for use by HW */
-	unsigned int pin_global;
 
 	struct {
 		struct mutex lock; /* protects the pages and their use */
 		atomic_t pages_pin_count;
+		atomic_t shrink_pin;
+
+		/**
+		 * Memory region for this object.
+		 */
+		struct intel_memory_region *region;
+		/**
+		 * List of memory region blocks allocated for this object.
+		 */
+		struct list_head blocks;
+		/**
+		 * Element within memory_region->objects or region->purgeable
+		 * if the object is marked as DONTNEED. Access is protected by
+		 * region->obj_lock.
+		 */
+		struct list_head region_link;
 
 		struct sg_table *pages;
 		void *mapping;
 
-		/* TODO: whack some of this into the error state */
 		struct i915_page_sizes {
 			/**
 			 * The sg mask of the pages sg_table. i.e the mask of
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_pages.c b/drivers/gpu/drm/i915/gem/i915_gem_pages.c
index 18f0ce0135c1..29f4c2850745 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_pages.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_pages.c
@@ -7,6 +7,7 @@
 #include "i915_drv.h"
 #include "i915_gem_object.h"
 #include "i915_scatterlist.h"
+#include "i915_gem_lmem.h"
 
 void __i915_gem_object_set_pages(struct drm_i915_gem_object *obj,
 				 struct sg_table *pages,
@@ -18,6 +19,9 @@ void __i915_gem_object_set_pages(struct drm_i915_gem_object *obj,
 
 	lockdep_assert_held(&obj->mm.lock);
 
+	if (i915_gem_object_is_volatile(obj))
+		obj->mm.madv = I915_MADV_DONTNEED;
+
 	/* Make the pages coherent with the GPU (flushing any swapin). */
 	if (obj->cache_dirty) {
 		obj->write_domain = 0;
@@ -71,6 +75,7 @@ void __i915_gem_object_set_pages(struct drm_i915_gem_object *obj,
 			list = &i915->mm.shrink_list;
 		list_add_tail(&obj->mm.link, list);
 
+		atomic_set(&obj->mm.shrink_pin, 0);
 		spin_unlock_irqrestore(&i915->mm.obj_lock, flags);
 	}
 }
@@ -150,6 +155,16 @@ static void __i915_gem_object_reset_page_iter(struct drm_i915_gem_object *obj)
 	rcu_read_unlock();
 }
 
+static void unmap_object(struct drm_i915_gem_object *obj, void *ptr)
+{
+	if (i915_gem_object_is_lmem(obj))
+		io_mapping_unmap((void __force __iomem *)ptr);
+	else if (is_vmalloc_addr(ptr))
+		vunmap(ptr);
+	else
+		kunmap(kmap_to_page(ptr));
+}
+
 struct sg_table *
 __i915_gem_object_unset_pages(struct drm_i915_gem_object *obj)
 {
@@ -159,17 +174,13 @@ __i915_gem_object_unset_pages(struct drm_i915_gem_object *obj)
 	if (IS_ERR_OR_NULL(pages))
 		return pages;
 
+	if (i915_gem_object_is_volatile(obj))
+		obj->mm.madv = I915_MADV_WILLNEED;
+
 	i915_gem_object_make_unshrinkable(obj);
 
 	if (obj->mm.mapping) {
-		void *ptr;
-
-		ptr = page_mask_bits(obj->mm.mapping);
-		if (is_vmalloc_addr(ptr))
-			vunmap(ptr);
-		else
-			kunmap(kmap_to_page(ptr));
-
+		unmap_object(obj, page_mask_bits(obj->mm.mapping));
 		obj->mm.mapping = NULL;
 	}
 
@@ -224,7 +235,7 @@ unlock:
 }
 
 /* The 'mapping' part of i915_gem_object_pin_map() below */
-static void *i915_gem_object_map(const struct drm_i915_gem_object *obj,
+static void *i915_gem_object_map(struct drm_i915_gem_object *obj,
 				 enum i915_map_type type)
 {
 	unsigned long n_pages = obj->base.size >> PAGE_SHIFT;
@@ -237,6 +248,16 @@ static void *i915_gem_object_map(const struct drm_i915_gem_object *obj,
 	pgprot_t pgprot;
 	void *addr;
 
+	if (i915_gem_object_is_lmem(obj)) {
+		void __iomem *io;
+
+		if (type != I915_MAP_WC)
+			return NULL;
+
+		io = i915_gem_object_lmem_io_map(obj, 0, obj->base.size);
+		return (void __force *)io;
+	}
+
 	/* A single page can always be kmapped */
 	if (n_pages == 1 && type == I915_MAP_WB)
 		return kmap(sg_page(sgt->sgl));
@@ -278,11 +299,13 @@ void *i915_gem_object_pin_map(struct drm_i915_gem_object *obj,
 			      enum i915_map_type type)
 {
 	enum i915_map_type has_type;
+	unsigned int flags;
 	bool pinned;
 	void *ptr;
 	int err;
 
-	if (unlikely(!i915_gem_object_has_struct_page(obj)))
+	flags = I915_GEM_OBJECT_HAS_STRUCT_PAGE | I915_GEM_OBJECT_HAS_IOMEM;
+	if (!i915_gem_object_type_has(obj, flags))
 		return ERR_PTR(-ENXIO);
 
 	err = mutex_lock_interruptible(&obj->mm.lock);
@@ -314,10 +337,7 @@ void *i915_gem_object_pin_map(struct drm_i915_gem_object *obj,
 			goto err_unpin;
 		}
 
-		if (is_vmalloc_addr(ptr))
-			vunmap(ptr);
-		else
-			kunmap(kmap_to_page(ptr));
+		unmap_object(obj, ptr);
 
 		ptr = obj->mm.mapping = NULL;
 	}
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_phys.c b/drivers/gpu/drm/i915/gem/i915_gem_phys.c
index 768356908160..8043ff63d73f 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_phys.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_phys.c
@@ -16,6 +16,7 @@
 #include "gt/intel_gt.h"
 #include "i915_drv.h"
 #include "i915_gem_object.h"
+#include "i915_gem_region.h"
 #include "i915_scatterlist.h"
 
 static int i915_gem_object_get_pages_phys(struct drm_i915_gem_object *obj)
@@ -191,8 +192,10 @@ int i915_gem_object_attach_phys(struct drm_i915_gem_object *obj, int align)
 	/* Perma-pin (until release) the physical set of pages */
 	__i915_gem_object_pin_pages(obj);
 
-	if (!IS_ERR_OR_NULL(pages))
+	if (!IS_ERR_OR_NULL(pages)) {
 		i915_gem_shmem_ops.put_pages(obj, pages);
+		i915_gem_object_release_memory_region(obj);
+	}
 	mutex_unlock(&obj->mm.lock);
 	return 0;
 
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_pm.c b/drivers/gpu/drm/i915/gem/i915_gem_pm.c
index 92e53c25424c..f88ee1317bb4 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_pm.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_pm.c
@@ -7,138 +7,9 @@
 #include "gem/i915_gem_pm.h"
 #include "gt/intel_gt.h"
 #include "gt/intel_gt_pm.h"
+#include "gt/intel_gt_requests.h"
 
 #include "i915_drv.h"
-#include "i915_globals.h"
-
-static void call_idle_barriers(struct intel_engine_cs *engine)
-{
-	struct llist_node *node, *next;
-
-	llist_for_each_safe(node, next, llist_del_all(&engine->barrier_tasks)) {
-		struct i915_active_request *active =
-			container_of((struct list_head *)node,
-				     typeof(*active), link);
-
-		INIT_LIST_HEAD(&active->link);
-		RCU_INIT_POINTER(active->request, NULL);
-
-		active->retire(active, NULL);
-	}
-}
-
-static void i915_gem_park(struct drm_i915_private *i915)
-{
-	struct intel_engine_cs *engine;
-	enum intel_engine_id id;
-
-	lockdep_assert_held(&i915->drm.struct_mutex);
-
-	for_each_engine(engine, i915, id)
-		call_idle_barriers(engine); /* cleanup after wedging */
-
-	i915_vma_parked(i915);
-
-	i915_globals_park();
-}
-
-static void idle_work_handler(struct work_struct *work)
-{
-	struct drm_i915_private *i915 =
-		container_of(work, typeof(*i915), gem.idle_work);
-	bool park;
-
-	cancel_delayed_work_sync(&i915->gem.retire_work);
-	mutex_lock(&i915->drm.struct_mutex);
-
-	intel_wakeref_lock(&i915->gt.wakeref);
-	park = (!intel_wakeref_is_active(&i915->gt.wakeref) &&
-		!work_pending(work));
-	intel_wakeref_unlock(&i915->gt.wakeref);
-	if (park)
-		i915_gem_park(i915);
-	else
-		queue_delayed_work(i915->wq,
-				   &i915->gem.retire_work,
-				   round_jiffies_up_relative(HZ));
-
-	mutex_unlock(&i915->drm.struct_mutex);
-}
-
-static void retire_work_handler(struct work_struct *work)
-{
-	struct drm_i915_private *i915 =
-		container_of(work, typeof(*i915), gem.retire_work.work);
-
-	/* Come back later if the device is busy... */
-	if (mutex_trylock(&i915->drm.struct_mutex)) {
-		i915_retire_requests(i915);
-		mutex_unlock(&i915->drm.struct_mutex);
-	}
-
-	queue_delayed_work(i915->wq,
-			   &i915->gem.retire_work,
-			   round_jiffies_up_relative(HZ));
-}
-
-static int pm_notifier(struct notifier_block *nb,
-		       unsigned long action,
-		       void *data)
-{
-	struct drm_i915_private *i915 =
-		container_of(nb, typeof(*i915), gem.pm_notifier);
-
-	switch (action) {
-	case INTEL_GT_UNPARK:
-		i915_globals_unpark();
-		queue_delayed_work(i915->wq,
-				   &i915->gem.retire_work,
-				   round_jiffies_up_relative(HZ));
-		break;
-
-	case INTEL_GT_PARK:
-		queue_work(i915->wq, &i915->gem.idle_work);
-		break;
-	}
-
-	return NOTIFY_OK;
-}
-
-static bool switch_to_kernel_context_sync(struct intel_gt *gt)
-{
-	bool result = !intel_gt_is_wedged(gt);
-
-	do {
-		if (i915_gem_wait_for_idle(gt->i915,
-					   I915_WAIT_LOCKED |
-					   I915_WAIT_FOR_IDLE_BOOST,
-					   I915_GEM_IDLE_TIMEOUT) == -ETIME) {
-			/* XXX hide warning from gem_eio */
-			if (i915_modparams.reset) {
-				dev_err(gt->i915->drm.dev,
-					"Failed to idle engines, declaring wedged!\n");
-				GEM_TRACE_DUMP();
-			}
-
-			/*
-			 * Forcibly cancel outstanding work and leave
-			 * the gpu quiet.
-			 */
-			intel_gt_set_wedged(gt);
-			result = false;
-		}
-	} while (i915_retire_requests(gt->i915) && result);
-
-	if (intel_gt_pm_wait_for_idle(gt))
-		result = false;
-
-	return result;
-}
-
-bool i915_gem_load_power_context(struct drm_i915_private *i915)
-{
-	return switch_to_kernel_context_sync(&i915->gt);
-}
 
 void i915_gem_suspend(struct drm_i915_private *i915)
 {
@@ -147,8 +18,6 @@ void i915_gem_suspend(struct drm_i915_private *i915)
 	intel_wakeref_auto(&i915->ggtt.userfault_wakeref, 0);
 	flush_workqueue(i915->wq);
 
-	mutex_lock(&i915->drm.struct_mutex);
-
 	/*
 	 * We have to flush all the executing contexts to main memory so
 	 * that they can saved in the hibernation image. To ensure the last
@@ -158,15 +27,9 @@ void i915_gem_suspend(struct drm_i915_private *i915)
 	 * state. Fortunately, the kernel_context is disposable and we do
 	 * not rely on its state.
 	 */
-	switch_to_kernel_context_sync(&i915->gt);
-
-	mutex_unlock(&i915->drm.struct_mutex);
-
-	cancel_delayed_work_sync(&i915->gt.hangcheck.work);
+	intel_gt_suspend_prepare(&i915->gt);
 
 	i915_gem_drain_freed_objects(i915);
-
-	intel_uc_suspend(&i915->gt.uc);
 }
 
 static struct drm_i915_gem_object *first_mm_object(struct list_head *list)
@@ -206,6 +69,8 @@ void i915_gem_suspend_late(struct drm_i915_private *i915)
 	 * machine in an unusable condition.
 	 */
 
+	intel_gt_suspend_late(&i915->gt);
+
 	spin_lock_irqsave(&i915->mm.obj_lock, flags);
 	for (phase = phases; *phase; phase++) {
 		LIST_HEAD(keep);
@@ -230,21 +95,15 @@ void i915_gem_suspend_late(struct drm_i915_private *i915)
 		list_splice_tail(&keep, *phase);
 	}
 	spin_unlock_irqrestore(&i915->mm.obj_lock, flags);
-
-	i915_gem_sanitize(i915);
 }
 
 void i915_gem_resume(struct drm_i915_private *i915)
 {
 	GEM_TRACE("\n");
 
-	mutex_lock(&i915->drm.struct_mutex);
 	intel_uncore_forcewake_get(&i915->uncore, FORCEWAKE_ALL);
 
-	i915_gem_restore_gtt_mappings(i915);
-	i915_gem_restore_fences(i915);
-
-	if (i915_gem_init_hw(i915))
+	if (intel_gt_init_hw(&i915->gt))
 		goto err_wedged;
 
 	/*
@@ -255,15 +114,8 @@ void i915_gem_resume(struct drm_i915_private *i915)
 	if (intel_gt_resume(&i915->gt))
 		goto err_wedged;
 
-	intel_uc_resume(&i915->gt.uc);
-
-	/* Always reload a context for powersaving. */
-	if (!i915_gem_load_power_context(i915))
-		goto err_wedged;
-
 out_unlock:
 	intel_uncore_forcewake_put(&i915->uncore, FORCEWAKE_ALL);
-	mutex_unlock(&i915->drm.struct_mutex);
 	return;
 
 err_wedged:
@@ -274,13 +126,3 @@ err_wedged:
 	}
 	goto out_unlock;
 }
-
-void i915_gem_init__pm(struct drm_i915_private *i915)
-{
-	INIT_WORK(&i915->gem.idle_work, idle_work_handler);
-	INIT_DELAYED_WORK(&i915->gem.retire_work, retire_work_handler);
-
-	i915->gem.pm_notifier.notifier_call = pm_notifier;
-	blocking_notifier_chain_register(&i915->gt.pm_notifications,
-					 &i915->gem.pm_notifier);
-}
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_pm.h b/drivers/gpu/drm/i915/gem/i915_gem_pm.h
index 6f7d5d11ac3b..26b78dbdc225 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_pm.h
+++ b/drivers/gpu/drm/i915/gem/i915_gem_pm.h
@@ -12,9 +12,6 @@
 struct drm_i915_private;
 struct work_struct;
 
-void i915_gem_init__pm(struct drm_i915_private *i915);
-
-bool i915_gem_load_power_context(struct drm_i915_private *i915);
 void i915_gem_resume(struct drm_i915_private *i915);
 
 void i915_gem_idle_work_handler(struct work_struct *work);
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_region.c b/drivers/gpu/drm/i915/gem/i915_gem_region.c
new file mode 100644
index 000000000000..2f7bcfb9c964
--- /dev/null
+++ b/drivers/gpu/drm/i915/gem/i915_gem_region.c
@@ -0,0 +1,174 @@
+// SPDX-License-Identifier: MIT
+/*
+ * Copyright © 2019 Intel Corporation
+ */
+
+#include "intel_memory_region.h"
+#include "i915_gem_region.h"
+#include "i915_drv.h"
+#include "i915_trace.h"
+
+void
+i915_gem_object_put_pages_buddy(struct drm_i915_gem_object *obj,
+				struct sg_table *pages)
+{
+	__intel_memory_region_put_pages_buddy(obj->mm.region, &obj->mm.blocks);
+
+	obj->mm.dirty = false;
+	sg_free_table(pages);
+	kfree(pages);
+}
+
+int
+i915_gem_object_get_pages_buddy(struct drm_i915_gem_object *obj)
+{
+	struct intel_memory_region *mem = obj->mm.region;
+	struct list_head *blocks = &obj->mm.blocks;
+	resource_size_t size = obj->base.size;
+	resource_size_t prev_end;
+	struct i915_buddy_block *block;
+	unsigned int flags;
+	struct sg_table *st;
+	struct scatterlist *sg;
+	unsigned int sg_page_sizes;
+	int ret;
+
+	st = kmalloc(sizeof(*st), GFP_KERNEL);
+	if (!st)
+		return -ENOMEM;
+
+	if (sg_alloc_table(st, size >> ilog2(mem->mm.chunk_size), GFP_KERNEL)) {
+		kfree(st);
+		return -ENOMEM;
+	}
+
+	flags = I915_ALLOC_MIN_PAGE_SIZE;
+	if (obj->flags & I915_BO_ALLOC_CONTIGUOUS)
+		flags |= I915_ALLOC_CONTIGUOUS;
+
+	ret = __intel_memory_region_get_pages_buddy(mem, size, flags, blocks);
+	if (ret)
+		goto err_free_sg;
+
+	GEM_BUG_ON(list_empty(blocks));
+
+	sg = st->sgl;
+	st->nents = 0;
+	sg_page_sizes = 0;
+	prev_end = (resource_size_t)-1;
+
+	list_for_each_entry(block, blocks, link) {
+		u64 block_size, offset;
+
+		block_size = min_t(u64, size,
+				   i915_buddy_block_size(&mem->mm, block));
+		offset = i915_buddy_block_offset(block);
+
+		GEM_BUG_ON(overflows_type(block_size, sg->length));
+
+		if (offset != prev_end ||
+		    add_overflows_t(typeof(sg->length), sg->length, block_size)) {
+			if (st->nents) {
+				sg_page_sizes |= sg->length;
+				sg = __sg_next(sg);
+			}
+
+			sg_dma_address(sg) = mem->region.start + offset;
+			sg_dma_len(sg) = block_size;
+
+			sg->length = block_size;
+
+			st->nents++;
+		} else {
+			sg->length += block_size;
+			sg_dma_len(sg) += block_size;
+		}
+
+		prev_end = offset + block_size;
+	};
+
+	sg_page_sizes |= sg->length;
+	sg_mark_end(sg);
+	i915_sg_trim(st);
+
+	__i915_gem_object_set_pages(obj, st, sg_page_sizes);
+
+	return 0;
+
+err_free_sg:
+	sg_free_table(st);
+	kfree(st);
+	return ret;
+}
+
+void i915_gem_object_init_memory_region(struct drm_i915_gem_object *obj,
+					struct intel_memory_region *mem,
+					unsigned long flags)
+{
+	INIT_LIST_HEAD(&obj->mm.blocks);
+	obj->mm.region = intel_memory_region_get(mem);
+	obj->flags |= flags;
+
+	mutex_lock(&mem->objects.lock);
+
+	if (obj->flags & I915_BO_ALLOC_VOLATILE)
+		list_add(&obj->mm.region_link, &mem->objects.purgeable);
+	else
+		list_add(&obj->mm.region_link, &mem->objects.list);
+
+	mutex_unlock(&mem->objects.lock);
+}
+
+void i915_gem_object_release_memory_region(struct drm_i915_gem_object *obj)
+{
+	struct intel_memory_region *mem = obj->mm.region;
+
+	mutex_lock(&mem->objects.lock);
+	list_del(&obj->mm.region_link);
+	mutex_unlock(&mem->objects.lock);
+
+	intel_memory_region_put(mem);
+}
+
+struct drm_i915_gem_object *
+i915_gem_object_create_region(struct intel_memory_region *mem,
+			      resource_size_t size,
+			      unsigned int flags)
+{
+	struct drm_i915_gem_object *obj;
+
+	/*
+	 * NB: Our use of resource_size_t for the size stems from using struct
+	 * resource for the mem->region. We might need to revisit this in the
+	 * future.
+	 */
+
+	GEM_BUG_ON(flags & ~I915_BO_ALLOC_FLAGS);
+
+	if (!mem)
+		return ERR_PTR(-ENODEV);
+
+	size = round_up(size, mem->min_page_size);
+
+	GEM_BUG_ON(!size);
+	GEM_BUG_ON(!IS_ALIGNED(size, I915_GTT_MIN_ALIGNMENT));
+
+	/*
+	 * XXX: There is a prevalence of the assumption that we fit the
+	 * object's page count inside a 32bit _signed_ variable. Let's document
+	 * this and catch if we ever need to fix it. In the meantime, if you do
+	 * spot such a local variable, please consider fixing!
+	 */
+
+	if (size >> PAGE_SHIFT > INT_MAX)
+		return ERR_PTR(-E2BIG);
+
+	if (overflows_type(size, obj->base.size))
+		return ERR_PTR(-E2BIG);
+
+	obj = mem->ops->create_object(mem, size, flags);
+	if (!IS_ERR(obj))
+		trace_i915_gem_object_create(obj);
+
+	return obj;
+}
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_region.h b/drivers/gpu/drm/i915/gem/i915_gem_region.h
new file mode 100644
index 000000000000..f2ff6f8bff74
--- /dev/null
+++ b/drivers/gpu/drm/i915/gem/i915_gem_region.h
@@ -0,0 +1,29 @@
+/* SPDX-License-Identifier: MIT */
+/*
+ * Copyright © 2019 Intel Corporation
+ */
+
+#ifndef __I915_GEM_REGION_H__
+#define __I915_GEM_REGION_H__
+
+#include <linux/types.h>
+
+struct intel_memory_region;
+struct drm_i915_gem_object;
+struct sg_table;
+
+int i915_gem_object_get_pages_buddy(struct drm_i915_gem_object *obj);
+void i915_gem_object_put_pages_buddy(struct drm_i915_gem_object *obj,
+				     struct sg_table *pages);
+
+void i915_gem_object_init_memory_region(struct drm_i915_gem_object *obj,
+					struct intel_memory_region *mem,
+					unsigned long flags);
+void i915_gem_object_release_memory_region(struct drm_i915_gem_object *obj);
+
+struct drm_i915_gem_object *
+i915_gem_object_create_region(struct intel_memory_region *mem,
+			      resource_size_t size,
+			      unsigned int flags);
+
+#endif
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_shmem.c b/drivers/gpu/drm/i915/gem/i915_gem_shmem.c
index 4c4954e8ce0a..4d69c3fc3439 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_shmem.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_shmem.c
@@ -7,7 +7,9 @@
 #include <linux/pagevec.h>
 #include <linux/swap.h>
 
+#include "gem/i915_gem_region.h"
 #include "i915_drv.h"
+#include "i915_gemfs.h"
 #include "i915_gem_object.h"
 #include "i915_scatterlist.h"
 #include "i915_trace.h"
@@ -26,6 +28,7 @@ static void check_release_pagevec(struct pagevec *pvec)
 static int shmem_get_pages(struct drm_i915_gem_object *obj)
 {
 	struct drm_i915_private *i915 = to_i915(obj->base.dev);
+	struct intel_memory_region *mem = obj->mm.region;
 	const unsigned long page_count = obj->base.size / PAGE_SIZE;
 	unsigned long i;
 	struct address_space *mapping;
@@ -52,7 +55,7 @@ static int shmem_get_pages(struct drm_i915_gem_object *obj)
 	 * If there's no chance of allocating enough pages for the whole
 	 * object, bail early.
 	 */
-	if (page_count > totalram_pages())
+	if (obj->base.size > resource_size(&mem->region))
 		return -ENOMEM;
 
 	st = kmalloc(sizeof(*st), GFP_KERNEL);
@@ -417,6 +420,8 @@ shmem_pwrite(struct drm_i915_gem_object *obj,
 
 static void shmem_release(struct drm_i915_gem_object *obj)
 {
+	i915_gem_object_release_memory_region(obj);
+
 	fput(obj->base.filp);
 }
 
@@ -434,9 +439,9 @@ const struct drm_i915_gem_object_ops i915_gem_shmem_ops = {
 	.release = shmem_release,
 };
 
-static int create_shmem(struct drm_i915_private *i915,
-			struct drm_gem_object *obj,
-			size_t size)
+static int __create_shmem(struct drm_i915_private *i915,
+			  struct drm_gem_object *obj,
+			  resource_size_t size)
 {
 	unsigned long flags = VM_NORESERVE;
 	struct file *filp;
@@ -455,31 +460,24 @@ static int create_shmem(struct drm_i915_private *i915,
 	return 0;
 }
 
-struct drm_i915_gem_object *
-i915_gem_object_create_shmem(struct drm_i915_private *i915, u64 size)
+static struct drm_i915_gem_object *
+create_shmem(struct intel_memory_region *mem,
+	     resource_size_t size,
+	     unsigned int flags)
 {
+	static struct lock_class_key lock_class;
+	struct drm_i915_private *i915 = mem->i915;
 	struct drm_i915_gem_object *obj;
 	struct address_space *mapping;
 	unsigned int cache_level;
 	gfp_t mask;
 	int ret;
 
-	/* There is a prevalence of the assumption that we fit the object's
-	 * page count inside a 32bit _signed_ variable. Let's document this and
-	 * catch if we ever need to fix it. In the meantime, if you do spot
-	 * such a local variable, please consider fixing!
-	 */
-	if (size >> PAGE_SHIFT > INT_MAX)
-		return ERR_PTR(-E2BIG);
-
-	if (overflows_type(size, obj->base.size))
-		return ERR_PTR(-E2BIG);
-
 	obj = i915_gem_object_alloc();
 	if (!obj)
 		return ERR_PTR(-ENOMEM);
 
-	ret = create_shmem(i915, &obj->base, size);
+	ret = __create_shmem(i915, &obj->base, size);
 	if (ret)
 		goto fail;
 
@@ -494,7 +492,7 @@ i915_gem_object_create_shmem(struct drm_i915_private *i915, u64 size)
 	mapping_set_gfp_mask(mapping, mask);
 	GEM_BUG_ON(!(mapping_gfp_mask(mapping) & __GFP_RECLAIM));
 
-	i915_gem_object_init(obj, &i915_gem_shmem_ops);
+	i915_gem_object_init(obj, &i915_gem_shmem_ops, &lock_class);
 
 	obj->write_domain = I915_GEM_DOMAIN_CPU;
 	obj->read_domains = I915_GEM_DOMAIN_CPU;
@@ -518,7 +516,7 @@ i915_gem_object_create_shmem(struct drm_i915_private *i915, u64 size)
 
 	i915_gem_object_set_cache_coherency(obj, cache_level);
 
-	trace_i915_gem_object_create(obj);
+	i915_gem_object_init_memory_region(obj, mem, 0);
 
 	return obj;
 
@@ -527,14 +525,22 @@ fail:
 	return ERR_PTR(ret);
 }
 
+struct drm_i915_gem_object *
+i915_gem_object_create_shmem(struct drm_i915_private *i915,
+			     resource_size_t size)
+{
+	return i915_gem_object_create_region(i915->mm.regions[INTEL_REGION_SMEM],
+					     size, 0);
+}
+
 /* Allocate a new GEM object and fill it with the supplied data */
 struct drm_i915_gem_object *
 i915_gem_object_create_shmem_from_data(struct drm_i915_private *dev_priv,
-				       const void *data, size_t size)
+				       const void *data, resource_size_t size)
 {
 	struct drm_i915_gem_object *obj;
 	struct file *file;
-	size_t offset;
+	resource_size_t offset;
 	int err;
 
 	obj = i915_gem_object_create_shmem(dev_priv, round_up(size, PAGE_SIZE));
@@ -577,3 +583,35 @@ fail:
 	i915_gem_object_put(obj);
 	return ERR_PTR(err);
 }
+
+static int init_shmem(struct intel_memory_region *mem)
+{
+	int err;
+
+	err = i915_gemfs_init(mem->i915);
+	if (err) {
+		DRM_NOTE("Unable to create a private tmpfs mount, hugepage support will be disabled(%d).\n",
+			 err);
+	}
+
+	return 0; /* Don't error, we can simply fallback to the kernel mnt */
+}
+
+static void release_shmem(struct intel_memory_region *mem)
+{
+	i915_gemfs_fini(mem->i915);
+}
+
+static const struct intel_memory_region_ops shmem_region_ops = {
+	.init = init_shmem,
+	.release = release_shmem,
+	.create_object = create_shmem,
+};
+
+struct intel_memory_region *i915_gem_shmem_setup(struct drm_i915_private *i915)
+{
+	return intel_memory_region_create(i915, 0,
+					  totalram_pages() << PAGE_SHIFT,
+					  PAGE_SIZE, 0,
+					  &shmem_region_ops);
+}
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_shrinker.c b/drivers/gpu/drm/i915/gem/i915_gem_shrinker.c
index edd21d14e64f..f2418a1cfe68 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_shrinker.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_shrinker.c
@@ -16,40 +16,6 @@
 
 #include "i915_trace.h"
 
-static bool shrinker_lock(struct drm_i915_private *i915,
-			  unsigned int flags,
-			  bool *unlock)
-{
-	struct mutex *m = &i915->drm.struct_mutex;
-
-	switch (mutex_trylock_recursive(m)) {
-	case MUTEX_TRYLOCK_RECURSIVE:
-		*unlock = false;
-		return true;
-
-	case MUTEX_TRYLOCK_FAILED:
-		*unlock = false;
-		if (flags & I915_SHRINK_ACTIVE &&
-		    mutex_lock_killable_nested(m, I915_MM_SHRINKER) == 0)
-			*unlock = true;
-		return *unlock;
-
-	case MUTEX_TRYLOCK_SUCCESS:
-		*unlock = true;
-		return true;
-	}
-
-	BUG();
-}
-
-static void shrinker_unlock(struct drm_i915_private *i915, bool unlock)
-{
-	if (!unlock)
-		return;
-
-	mutex_unlock(&i915->drm.struct_mutex);
-}
-
 static bool swap_available(void)
 {
 	return get_nr_swap_pages() > 0;
@@ -61,7 +27,8 @@ static bool can_release_pages(struct drm_i915_gem_object *obj)
 	if (!i915_gem_object_is_shrinkable(obj))
 		return false;
 
-	/* Only report true if by unbinding the object and putting its pages
+	/*
+	 * Only report true if by unbinding the object and putting its pages
 	 * we can actually make forward progress towards freeing physical
 	 * pages.
 	 *
@@ -72,16 +39,8 @@ static bool can_release_pages(struct drm_i915_gem_object *obj)
 	if (atomic_read(&obj->mm.pages_pin_count) > atomic_read(&obj->bind_count))
 		return false;
 
-	/* If any vma are "permanently" pinned, it will prevent us from
-	 * reclaiming the obj->mm.pages. We only allow scanout objects to claim
-	 * a permanent pin, along with a few others like the context objects.
-	 * To simplify the scan, and to avoid walking the list of vma under the
-	 * object, we just check the count of its permanently pinned.
-	 */
-	if (READ_ONCE(obj->pin_global))
-		return false;
-
-	/* We can only return physical pages to the system if we can either
+	/*
+	 * We can only return physical pages to the system if we can either
 	 * discard the contents (because the user has marked them as being
 	 * purgeable) or if we can move their contents out to swap.
 	 */
@@ -162,10 +121,6 @@ i915_gem_shrink(struct drm_i915_private *i915,
 	intel_wakeref_t wakeref = 0;
 	unsigned long count = 0;
 	unsigned long scanned = 0;
-	bool unlock;
-
-	if (!shrinker_lock(i915, shrink, &unlock))
-		return 0;
 
 	/*
 	 * When shrinking the active list, we should also consider active
@@ -275,8 +230,6 @@ i915_gem_shrink(struct drm_i915_private *i915,
 	if (shrink & I915_SHRINK_BOUND)
 		intel_runtime_pm_put(&i915->runtime_pm, wakeref);
 
-	shrinker_unlock(i915, unlock);
-
 	if (nr_scanned)
 		*nr_scanned += scanned;
 	return count;
@@ -346,19 +299,14 @@ i915_gem_shrinker_scan(struct shrinker *shrinker, struct shrink_control *sc)
 	struct drm_i915_private *i915 =
 		container_of(shrinker, struct drm_i915_private, mm.shrinker);
 	unsigned long freed;
-	bool unlock;
 
 	sc->nr_scanned = 0;
 
-	if (!shrinker_lock(i915, 0, &unlock))
-		return SHRINK_STOP;
-
 	freed = i915_gem_shrink(i915,
 				sc->nr_to_scan,
 				&sc->nr_scanned,
 				I915_SHRINK_BOUND |
-				I915_SHRINK_UNBOUND |
-				I915_SHRINK_WRITEBACK);
+				I915_SHRINK_UNBOUND);
 	if (sc->nr_scanned < sc->nr_to_scan && current_is_kswapd()) {
 		intel_wakeref_t wakeref;
 
@@ -373,8 +321,6 @@ i915_gem_shrinker_scan(struct shrinker *shrinker, struct shrink_control *sc)
 		}
 	}
 
-	shrinker_unlock(i915, unlock);
-
 	return sc->nr_scanned ? freed : SHRINK_STOP;
 }
 
@@ -391,6 +337,7 @@ i915_gem_shrinker_oom(struct notifier_block *nb, unsigned long event, void *ptr)
 	freed_pages = 0;
 	with_intel_runtime_pm(&i915->runtime_pm, wakeref)
 		freed_pages += i915_gem_shrink(i915, -1UL, NULL,
+					       I915_SHRINK_ACTIVE |
 					       I915_SHRINK_BOUND |
 					       I915_SHRINK_UNBOUND |
 					       I915_SHRINK_WRITEBACK);
@@ -426,10 +373,6 @@ i915_gem_shrinker_vmap(struct notifier_block *nb, unsigned long event, void *ptr
 	struct i915_vma *vma, *next;
 	unsigned long freed_pages = 0;
 	intel_wakeref_t wakeref;
-	bool unlock;
-
-	if (!shrinker_lock(i915, 0, &unlock))
-		return NOTIFY_DONE;
 
 	with_intel_runtime_pm(&i915->runtime_pm, wakeref)
 		freed_pages += i915_gem_shrink(i915, -1UL, NULL,
@@ -446,15 +389,11 @@ i915_gem_shrinker_vmap(struct notifier_block *nb, unsigned long event, void *ptr
 		if (!vma->iomap || i915_vma_is_active(vma))
 			continue;
 
-		mutex_unlock(&i915->ggtt.vm.mutex);
-		if (i915_vma_unbind(vma) == 0)
+		if (__i915_vma_unbind(vma) == 0)
 			freed_pages += count;
-		mutex_lock(&i915->ggtt.vm.mutex);
 	}
 	mutex_unlock(&i915->ggtt.vm.mutex);
 
-	shrinker_unlock(i915, unlock);
-
 	*(unsigned long *)ptr += freed_pages;
 	return NOTIFY_DONE;
 }
@@ -497,72 +436,65 @@ void i915_gem_shrinker_taints_mutex(struct drm_i915_private *i915,
 
 	fs_reclaim_acquire(GFP_KERNEL);
 
-	/*
-	 * As we invariably rely on the struct_mutex within the shrinker,
-	 * but have a complicated recursion dance, taint all the mutexes used
-	 * within the shrinker with the struct_mutex. For completeness, we
-	 * taint with all subclass of struct_mutex, even though we should
-	 * only need tainting by I915_MM_NORMAL to catch possible ABBA
-	 * deadlocks from using struct_mutex inside @mutex.
-	 */
-	mutex_acquire(&i915->drm.struct_mutex.dep_map,
-		      I915_MM_SHRINKER, 0, _RET_IP_);
-
 	mutex_acquire(&mutex->dep_map, 0, 0, _RET_IP_);
-	mutex_release(&mutex->dep_map, 0, _RET_IP_);
-
-	mutex_release(&i915->drm.struct_mutex.dep_map, 0, _RET_IP_);
+	mutex_release(&mutex->dep_map, _RET_IP_);
 
 	fs_reclaim_release(GFP_KERNEL);
 
 	if (unlock)
-		mutex_release(&i915->drm.struct_mutex.dep_map, 0, _RET_IP_);
+		mutex_release(&i915->drm.struct_mutex.dep_map, _RET_IP_);
 }
 
 #define obj_to_i915(obj__) to_i915((obj__)->base.dev)
 
 void i915_gem_object_make_unshrinkable(struct drm_i915_gem_object *obj)
 {
+	struct drm_i915_private *i915 = obj_to_i915(obj);
+	unsigned long flags;
+
 	/*
 	 * We can only be called while the pages are pinned or when
 	 * the pages are released. If pinned, we should only be called
 	 * from a single caller under controlled conditions; and on release
 	 * only one caller may release us. Neither the two may cross.
 	 */
-	if (!list_empty(&obj->mm.link)) { /* pinned by caller */
-		struct drm_i915_private *i915 = obj_to_i915(obj);
-		unsigned long flags;
-
-		spin_lock_irqsave(&i915->mm.obj_lock, flags);
-		GEM_BUG_ON(list_empty(&obj->mm.link));
+	if (atomic_add_unless(&obj->mm.shrink_pin, 1, 0))
+		return;
 
+	spin_lock_irqsave(&i915->mm.obj_lock, flags);
+	if (!atomic_fetch_inc(&obj->mm.shrink_pin) &&
+	    !list_empty(&obj->mm.link)) {
 		list_del_init(&obj->mm.link);
 		i915->mm.shrink_count--;
 		i915->mm.shrink_memory -= obj->base.size;
-
-		spin_unlock_irqrestore(&i915->mm.obj_lock, flags);
 	}
+	spin_unlock_irqrestore(&i915->mm.obj_lock, flags);
 }
 
 static void __i915_gem_object_make_shrinkable(struct drm_i915_gem_object *obj,
 					      struct list_head *head)
 {
+	struct drm_i915_private *i915 = obj_to_i915(obj);
+	unsigned long flags;
+
 	GEM_BUG_ON(!i915_gem_object_has_pages(obj));
-	GEM_BUG_ON(!list_empty(&obj->mm.link));
+	if (!i915_gem_object_is_shrinkable(obj))
+		return;
 
-	if (i915_gem_object_is_shrinkable(obj)) {
-		struct drm_i915_private *i915 = obj_to_i915(obj);
-		unsigned long flags;
+	if (atomic_add_unless(&obj->mm.shrink_pin, -1, 1))
+		return;
 
-		spin_lock_irqsave(&i915->mm.obj_lock, flags);
-		GEM_BUG_ON(!kref_read(&obj->base.refcount));
+	spin_lock_irqsave(&i915->mm.obj_lock, flags);
+	GEM_BUG_ON(!kref_read(&obj->base.refcount));
+	if (atomic_dec_and_test(&obj->mm.shrink_pin)) {
+		GEM_BUG_ON(!list_empty(&obj->mm.link));
 
 		list_add_tail(&obj->mm.link, head);
 		i915->mm.shrink_count++;
 		i915->mm.shrink_memory += obj->base.size;
 
-		spin_unlock_irqrestore(&i915->mm.obj_lock, flags);
 	}
+	spin_unlock_irqrestore(&i915->mm.obj_lock, flags);
 }
 
 void i915_gem_object_make_shrinkable(struct drm_i915_gem_object *obj)
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_stolen.c b/drivers/gpu/drm/i915/gem/i915_gem_stolen.c
index aa533b4ab5f5..a2d49c04e6a4 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_stolen.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_stolen.c
@@ -10,6 +10,7 @@
 #include <drm/drm_mm.h>
 #include <drm/i915_drm.h>
 
+#include "gem/i915_gem_region.h"
 #include "i915_drv.h"
 #include "i915_gem_stolen.h"
 
@@ -150,7 +151,7 @@ static int i915_adjust_stolen(struct drm_i915_private *dev_priv,
 	return 0;
 }
 
-void i915_gem_cleanup_stolen(struct drm_i915_private *dev_priv)
+static void i915_gem_cleanup_stolen(struct drm_i915_private *dev_priv)
 {
 	if (!drm_mm_initialized(&dev_priv->mm.stolen))
 		return;
@@ -355,7 +356,7 @@ static void icl_get_stolen_reserved(struct drm_i915_private *i915,
 	}
 }
 
-int i915_gem_init_stolen(struct drm_i915_private *dev_priv)
+static int i915_gem_init_stolen(struct drm_i915_private *dev_priv)
 {
 	resource_size_t reserved_base, stolen_top;
 	resource_size_t reserved_total, reserved_size;
@@ -425,8 +426,11 @@ int i915_gem_init_stolen(struct drm_i915_private *dev_priv)
 			bdw_get_stolen_reserved(dev_priv,
 						&reserved_base, &reserved_size);
 		break;
-	case 11:
 	default:
+		MISSING_CASE(INTEL_GEN(dev_priv));
+		/* fall-through */
+	case 11:
+	case 12:
 		icl_get_stolen_reserved(dev_priv, &reserved_base,
 					&reserved_size);
 		break;
@@ -536,6 +540,9 @@ i915_gem_object_release_stolen(struct drm_i915_gem_object *obj)
 
 	i915_gem_stolen_remove_node(dev_priv, stolen);
 	kfree(stolen);
+
+	if (obj->mm.region)
+		i915_gem_object_release_memory_region(obj);
 }
 
 static const struct drm_i915_gem_object_ops i915_gem_object_stolen_ops = {
@@ -545,65 +552,116 @@ static const struct drm_i915_gem_object_ops i915_gem_object_stolen_ops = {
 };
 
 static struct drm_i915_gem_object *
-_i915_gem_object_create_stolen(struct drm_i915_private *dev_priv,
-			       struct drm_mm_node *stolen)
+__i915_gem_object_create_stolen(struct drm_i915_private *dev_priv,
+				struct drm_mm_node *stolen,
+				struct intel_memory_region *mem)
 {
+	static struct lock_class_key lock_class;
 	struct drm_i915_gem_object *obj;
 	unsigned int cache_level;
+	int err = -ENOMEM;
 
 	obj = i915_gem_object_alloc();
-	if (obj == NULL)
-		return NULL;
+	if (!obj)
+		goto err;
 
 	drm_gem_private_object_init(&dev_priv->drm, &obj->base, stolen->size);
-	i915_gem_object_init(obj, &i915_gem_object_stolen_ops);
+	i915_gem_object_init(obj, &i915_gem_object_stolen_ops, &lock_class);
 
 	obj->stolen = stolen;
 	obj->read_domains = I915_GEM_DOMAIN_CPU | I915_GEM_DOMAIN_GTT;
 	cache_level = HAS_LLC(dev_priv) ? I915_CACHE_LLC : I915_CACHE_NONE;
 	i915_gem_object_set_cache_coherency(obj, cache_level);
 
-	if (i915_gem_object_pin_pages(obj))
+	err = i915_gem_object_pin_pages(obj);
+	if (err)
 		goto cleanup;
 
+	if (mem)
+		i915_gem_object_init_memory_region(obj, mem, 0);
+
 	return obj;
 
 cleanup:
 	i915_gem_object_free(obj);
-	return NULL;
+err:
+	return ERR_PTR(err);
 }
 
-struct drm_i915_gem_object *
-i915_gem_object_create_stolen(struct drm_i915_private *dev_priv,
-			      resource_size_t size)
+static struct drm_i915_gem_object *
+_i915_gem_object_create_stolen(struct intel_memory_region *mem,
+			       resource_size_t size,
+			       unsigned int flags)
 {
+	struct drm_i915_private *dev_priv = mem->i915;
 	struct drm_i915_gem_object *obj;
 	struct drm_mm_node *stolen;
 	int ret;
 
 	if (!drm_mm_initialized(&dev_priv->mm.stolen))
-		return NULL;
+		return ERR_PTR(-ENODEV);
 
 	if (size == 0)
-		return NULL;
+		return ERR_PTR(-EINVAL);
 
 	stolen = kzalloc(sizeof(*stolen), GFP_KERNEL);
 	if (!stolen)
-		return NULL;
+		return ERR_PTR(-ENOMEM);
 
 	ret = i915_gem_stolen_insert_node(dev_priv, stolen, size, 4096);
 	if (ret) {
-		kfree(stolen);
-		return NULL;
+		obj = ERR_PTR(ret);
+		goto err_free;
 	}
 
-	obj = _i915_gem_object_create_stolen(dev_priv, stolen);
-	if (obj)
-		return obj;
+	obj = __i915_gem_object_create_stolen(dev_priv, stolen, mem);
+	if (IS_ERR(obj))
+		goto err_remove;
 
+	return obj;
+
+err_remove:
 	i915_gem_stolen_remove_node(dev_priv, stolen);
+err_free:
 	kfree(stolen);
-	return NULL;
+	return obj;
+}
+
+struct drm_i915_gem_object *
+i915_gem_object_create_stolen(struct drm_i915_private *dev_priv,
+			      resource_size_t size)
+{
+	return i915_gem_object_create_region(dev_priv->mm.regions[INTEL_REGION_STOLEN],
+					     size, I915_BO_ALLOC_CONTIGUOUS);
+}
+
+static int init_stolen(struct intel_memory_region *mem)
+{
+	/*
+	 * Initialise stolen early so that we may reserve preallocated
+	 * objects for the BIOS to KMS transition.
+	 */
+	return i915_gem_init_stolen(mem->i915);
+}
+
+static void release_stolen(struct intel_memory_region *mem)
+{
+	i915_gem_cleanup_stolen(mem->i915);
+}
+
+static const struct intel_memory_region_ops i915_region_stolen_ops = {
+	.init = init_stolen,
+	.release = release_stolen,
+	.create_object = _i915_gem_object_create_stolen,
+};
+
+struct intel_memory_region *i915_gem_stolen_setup(struct drm_i915_private *i915)
+{
+	return intel_memory_region_create(i915,
+					  intel_graphics_stolen_res.start,
+					  resource_size(&intel_graphics_stolen_res),
+					  PAGE_SIZE, 0,
+					  &i915_region_stolen_ops);
 }
 
 struct drm_i915_gem_object *
@@ -619,9 +677,7 @@ i915_gem_object_create_stolen_for_preallocated(struct drm_i915_private *dev_priv
 	int ret;
 
 	if (!drm_mm_initialized(&dev_priv->mm.stolen))
-		return NULL;
-
-	lockdep_assert_held(&dev_priv->drm.struct_mutex);
+		return ERR_PTR(-ENODEV);
 
 	DRM_DEBUG_DRIVER("creating preallocated stolen object: stolen_offset=%pa, gtt_offset=%pa, size=%pa\n",
 			 &stolen_offset, &gtt_offset, &size);
@@ -630,11 +686,11 @@ i915_gem_object_create_stolen_for_preallocated(struct drm_i915_private *dev_priv
 	if (WARN_ON(size == 0) ||
 	    WARN_ON(!IS_ALIGNED(size, I915_GTT_PAGE_SIZE)) ||
 	    WARN_ON(!IS_ALIGNED(stolen_offset, I915_GTT_MIN_ALIGNMENT)))
-		return NULL;
+		return ERR_PTR(-EINVAL);
 
 	stolen = kzalloc(sizeof(*stolen), GFP_KERNEL);
 	if (!stolen)
-		return NULL;
+		return ERR_PTR(-ENOMEM);
 
 	stolen->start = stolen_offset;
 	stolen->size = size;
@@ -644,15 +700,15 @@ i915_gem_object_create_stolen_for_preallocated(struct drm_i915_private *dev_priv
 	if (ret) {
 		DRM_DEBUG_DRIVER("failed to allocate stolen space\n");
 		kfree(stolen);
-		return NULL;
+		return ERR_PTR(ret);
 	}
 
-	obj = _i915_gem_object_create_stolen(dev_priv, stolen);
-	if (obj == NULL) {
+	obj = __i915_gem_object_create_stolen(dev_priv, stolen, NULL);
+	if (IS_ERR(obj)) {
 		DRM_DEBUG_DRIVER("failed to allocate stolen object\n");
 		i915_gem_stolen_remove_node(dev_priv, stolen);
 		kfree(stolen);
-		return NULL;
+		return obj;
 	}
 
 	/* Some objects just need physical mem from stolen space */
@@ -674,22 +730,26 @@ i915_gem_object_create_stolen_for_preallocated(struct drm_i915_private *dev_priv
 	 * setting up the GTT space. The actual reservation will occur
 	 * later.
 	 */
+	mutex_lock(&ggtt->vm.mutex);
 	ret = i915_gem_gtt_reserve(&ggtt->vm, &vma->node,
 				   size, gtt_offset, obj->cache_level,
 				   0);
 	if (ret) {
 		DRM_DEBUG_DRIVER("failed to allocate stolen GTT space\n");
+		mutex_unlock(&ggtt->vm.mutex);
 		goto err_pages;
 	}
 
 	GEM_BUG_ON(!drm_mm_node_allocated(&vma->node));
 
+	GEM_BUG_ON(vma->pages);
 	vma->pages = obj->mm.pages;
-	vma->flags |= I915_VMA_GLOBAL_BIND;
+	atomic_set(&vma->pages_count, I915_VMA_PAGES_ACTIVE);
+
+	set_bit(I915_VMA_GLOBAL_BIND_BIT, __i915_vma_flags(vma));
 	__i915_vma_set_map_and_fenceable(vma);
 
-	mutex_lock(&ggtt->vm.mutex);
-	list_move_tail(&vma->vm_link, &ggtt->vm.bound_list);
+	list_add_tail(&vma->vm_link, &ggtt->vm.bound_list);
 	mutex_unlock(&ggtt->vm.mutex);
 
 	GEM_BUG_ON(i915_gem_object_is_shrinkable(obj));
@@ -701,5 +761,5 @@ err_pages:
 	i915_gem_object_unpin_pages(obj);
 err:
 	i915_gem_object_put(obj);
-	return NULL;
+	return ERR_PTR(ret);
 }
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_stolen.h b/drivers/gpu/drm/i915/gem/i915_gem_stolen.h
index 2289644d8604..c1040627fbf3 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_stolen.h
+++ b/drivers/gpu/drm/i915/gem/i915_gem_stolen.h
@@ -21,8 +21,7 @@ int i915_gem_stolen_insert_node_in_range(struct drm_i915_private *dev_priv,
 					 u64 end);
 void i915_gem_stolen_remove_node(struct drm_i915_private *dev_priv,
 				 struct drm_mm_node *node);
-int i915_gem_init_stolen(struct drm_i915_private *dev_priv);
-void i915_gem_cleanup_stolen(struct drm_i915_private *dev_priv);
+struct intel_memory_region *i915_gem_stolen_setup(struct drm_i915_private *i915);
 struct drm_i915_gem_object *
 i915_gem_object_create_stolen(struct drm_i915_private *dev_priv,
 			      resource_size_t size);
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_throttle.c b/drivers/gpu/drm/i915/gem/i915_gem_throttle.c
index 1e372420771b..540ef0551789 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_throttle.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_throttle.c
@@ -50,10 +50,8 @@ i915_gem_throttle_ioctl(struct drm_device *dev, void *data,
 		if (time_after_eq(request->emitted_jiffies, recent_enough))
 			break;
 
-		if (target) {
+		if (target && xchg(&target->file_priv, NULL))
 			list_del(&target->client_link);
-			target->file_priv = NULL;
-		}
 
 		target = request;
 	}
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_tiling.c b/drivers/gpu/drm/i915/gem/i915_gem_tiling.c
index ca0c2f451742..1fa592d82af5 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_tiling.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_tiling.c
@@ -181,22 +181,25 @@ static int
 i915_gem_object_fence_prepare(struct drm_i915_gem_object *obj,
 			      int tiling_mode, unsigned int stride)
 {
+	struct i915_ggtt *ggtt = &to_i915(obj->base.dev)->ggtt;
 	struct i915_vma *vma;
-	int ret;
+	int ret = 0;
 
 	if (tiling_mode == I915_TILING_NONE)
 		return 0;
 
+	mutex_lock(&ggtt->vm.mutex);
 	for_each_ggtt_vma(vma, obj) {
 		if (i915_vma_fence_prepare(vma, tiling_mode, stride))
 			continue;
 
-		ret = i915_vma_unbind(vma);
+		ret = __i915_vma_unbind(vma);
 		if (ret)
-			return ret;
+			break;
 	}
+	mutex_unlock(&ggtt->vm.mutex);
 
-	return 0;
+	return ret;
 }
 
 int
@@ -212,7 +215,6 @@ i915_gem_object_set_tiling(struct drm_i915_gem_object *obj,
 
 	GEM_BUG_ON(!i915_tiling_ok(obj, tiling, stride));
 	GEM_BUG_ON(!stride ^ (tiling == I915_TILING_NONE));
-	lockdep_assert_held(&i915->drm.struct_mutex);
 
 	if ((tiling | stride) == obj->tiling_and_stride)
 		return 0;
@@ -233,16 +235,18 @@ i915_gem_object_set_tiling(struct drm_i915_gem_object *obj,
 	 * whilst executing a fenced command for an untiled object.
 	 */
 
-	err = i915_gem_object_fence_prepare(obj, tiling, stride);
-	if (err)
-		return err;
-
 	i915_gem_object_lock(obj);
 	if (i915_gem_object_is_framebuffer(obj)) {
 		i915_gem_object_unlock(obj);
 		return -EBUSY;
 	}
 
+	err = i915_gem_object_fence_prepare(obj, tiling, stride);
+	if (err) {
+		i915_gem_object_unlock(obj);
+		return err;
+	}
+
 	/* If the memory has unknown (i.e. varying) swizzling, we pin the
 	 * pages to prevent them being swapped out and causing corruption
 	 * due to the change in swizzling.
@@ -313,10 +317,14 @@ int
 i915_gem_set_tiling_ioctl(struct drm_device *dev, void *data,
 			  struct drm_file *file)
 {
+	struct drm_i915_private *dev_priv = to_i915(dev);
 	struct drm_i915_gem_set_tiling *args = data;
 	struct drm_i915_gem_object *obj;
 	int err;
 
+	if (!dev_priv->ggtt.num_fences)
+		return -EOPNOTSUPP;
+
 	obj = i915_gem_object_lookup(file, args->handle);
 	if (!obj)
 		return -ENOENT;
@@ -340,9 +348,9 @@ i915_gem_set_tiling_ioctl(struct drm_device *dev, void *data,
 		args->stride = 0;
 	} else {
 		if (args->tiling_mode == I915_TILING_X)
-			args->swizzle_mode = to_i915(dev)->mm.bit_6_swizzle_x;
+			args->swizzle_mode = to_i915(dev)->ggtt.bit_6_swizzle_x;
 		else
-			args->swizzle_mode = to_i915(dev)->mm.bit_6_swizzle_y;
+			args->swizzle_mode = to_i915(dev)->ggtt.bit_6_swizzle_y;
 
 		/* Hide bit 17 swizzling from the user.  This prevents old Mesa
 		 * from aborting the application on sw fallbacks to bit 17,
@@ -364,12 +372,7 @@ i915_gem_set_tiling_ioctl(struct drm_device *dev, void *data,
 		}
 	}
 
-	err = mutex_lock_interruptible(&dev->struct_mutex);
-	if (err)
-		goto err;
-
 	err = i915_gem_object_set_tiling(obj, args->tiling_mode, args->stride);
-	mutex_unlock(&dev->struct_mutex);
 
 	/* We have to maintain this existing ABI... */
 	args->stride = i915_gem_object_get_stride(obj);
@@ -402,6 +405,9 @@ i915_gem_get_tiling_ioctl(struct drm_device *dev, void *data,
 	struct drm_i915_gem_object *obj;
 	int err = -ENOENT;
 
+	if (!dev_priv->ggtt.num_fences)
+		return -EOPNOTSUPP;
+
 	rcu_read_lock();
 	obj = i915_gem_object_lookup_rcu(file, args->handle);
 	if (obj) {
@@ -415,10 +421,10 @@ i915_gem_get_tiling_ioctl(struct drm_device *dev, void *data,
 
 	switch (args->tiling_mode) {
 	case I915_TILING_X:
-		args->swizzle_mode = dev_priv->mm.bit_6_swizzle_x;
+		args->swizzle_mode = dev_priv->ggtt.bit_6_swizzle_x;
 		break;
 	case I915_TILING_Y:
-		args->swizzle_mode = dev_priv->mm.bit_6_swizzle_y;
+		args->swizzle_mode = dev_priv->ggtt.bit_6_swizzle_y;
 		break;
 	default:
 	case I915_TILING_NONE:
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_userptr.c b/drivers/gpu/drm/i915/gem/i915_gem_userptr.c
index 11b231c187c5..4c72d74d6576 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_userptr.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_userptr.c
@@ -92,7 +92,6 @@ userptr_mn_invalidate_range_start(struct mmu_notifier *_mn,
 	struct i915_mmu_notifier *mn =
 		container_of(_mn, struct i915_mmu_notifier, mn);
 	struct interval_tree_node *it;
-	struct mutex *unlock = NULL;
 	unsigned long end;
 	int ret = 0;
 
@@ -129,33 +128,13 @@ userptr_mn_invalidate_range_start(struct mmu_notifier *_mn,
 		}
 		spin_unlock(&mn->lock);
 
-		if (!unlock) {
-			unlock = &mn->mm->i915->drm.struct_mutex;
-
-			switch (mutex_trylock_recursive(unlock)) {
-			default:
-			case MUTEX_TRYLOCK_FAILED:
-				if (mutex_lock_killable_nested(unlock, I915_MM_SHRINKER)) {
-					i915_gem_object_put(obj);
-					return -EINTR;
-				}
-				/* fall through */
-			case MUTEX_TRYLOCK_SUCCESS:
-				break;
-
-			case MUTEX_TRYLOCK_RECURSIVE:
-				unlock = ERR_PTR(-EEXIST);
-				break;
-			}
-		}
-
 		ret = i915_gem_object_unbind(obj,
 					     I915_GEM_OBJECT_UNBIND_ACTIVE);
 		if (ret == 0)
 			ret = __i915_gem_object_put_pages(obj, I915_MM_SHRINKER);
 		i915_gem_object_put(obj);
 		if (ret)
-			goto unlock;
+			return ret;
 
 		spin_lock(&mn->lock);
 
@@ -168,10 +147,6 @@ userptr_mn_invalidate_range_start(struct mmu_notifier *_mn,
 	}
 	spin_unlock(&mn->lock);
 
-unlock:
-	if (!IS_ERR_OR_NULL(unlock))
-		mutex_unlock(unlock);
-
 	return ret;
 
 }
@@ -671,8 +646,28 @@ i915_gem_userptr_put_pages(struct drm_i915_gem_object *obj,
 		obj->mm.dirty = false;
 
 	for_each_sgt_page(page, sgt_iter, pages) {
-		if (obj->mm.dirty)
+		if (obj->mm.dirty && trylock_page(page)) {
+			/*
+			 * As this may not be anonymous memory (e.g. shmem)
+			 * but exist on a real mapping, we have to lock
+			 * the page in order to dirty it -- holding
+			 * the page reference is not sufficient to
+			 * prevent the inode from being truncated.
+			 * Play safe and take the lock.
+			 *
+			 * However...!
+			 *
+			 * The mmu-notifier can be invalidated for a
+			 * migrate_page, that is alreadying holding the lock
+			 * on the page. Such a try_to_unmap() will result
+			 * in us calling put_pages() and so recursively try
+			 * to lock the page. We avoid that deadlock with
+			 * a trylock_page() and in exchange we risk missing
+			 * some page dirtying.
+			 */
 			set_page_dirty(page);
+			unlock_page(page);
+		}
 
 		mark_page_accessed(page);
 		put_page(page);
@@ -702,6 +697,7 @@ i915_gem_userptr_dmabuf_export(struct drm_i915_gem_object *obj)
 static const struct drm_i915_gem_object_ops i915_gem_userptr_ops = {
 	.flags = I915_GEM_OBJECT_HAS_STRUCT_PAGE |
 		 I915_GEM_OBJECT_IS_SHRINKABLE |
+		 I915_GEM_OBJECT_NO_GGTT |
 		 I915_GEM_OBJECT_ASYNC_CANCEL,
 	.get_pages = i915_gem_userptr_get_pages,
 	.put_pages = i915_gem_userptr_put_pages,
@@ -749,6 +745,7 @@ i915_gem_userptr_ioctl(struct drm_device *dev,
 		       void *data,
 		       struct drm_file *file)
 {
+	static struct lock_class_key lock_class;
 	struct drm_i915_private *dev_priv = to_i915(dev);
 	struct drm_i915_gem_userptr *args = data;
 	struct drm_i915_gem_object *obj;
@@ -782,7 +779,8 @@ i915_gem_userptr_ioctl(struct drm_device *dev,
 		 * On almost all of the older hw, we cannot tell the GPU that
 		 * a page is readonly.
 		 */
-		vm = dev_priv->kernel_context->vm;
+		vm = rcu_dereference_protected(dev_priv->kernel_context->vm,
+					       true); /* static vm */
 		if (!vm || !vm->has_read_only)
 			return -ENODEV;
 	}
@@ -792,7 +790,7 @@ i915_gem_userptr_ioctl(struct drm_device *dev,
 		return -ENOMEM;
 
 	drm_gem_private_object_init(dev, &obj->base, args->user_size);
-	i915_gem_object_init(obj, &i915_gem_userptr_ops);
+	i915_gem_object_init(obj, &i915_gem_userptr_ops, &lock_class);
 	obj->read_domains = I915_GEM_DOMAIN_CPU;
 	obj->write_domain = I915_GEM_DOMAIN_CPU;
 	i915_gem_object_set_cache_coherency(obj, I915_CACHE_LLC);
diff --git a/drivers/gpu/drm/i915/gem/selftests/huge_gem_object.c b/drivers/gpu/drm/i915/gem/selftests/huge_gem_object.c
index 3c5d17b2b670..892d12db6c49 100644
--- a/drivers/gpu/drm/i915/gem/selftests/huge_gem_object.c
+++ b/drivers/gpu/drm/i915/gem/selftests/huge_gem_object.c
@@ -96,6 +96,7 @@ huge_gem_object(struct drm_i915_private *i915,
 		phys_addr_t phys_size,
 		dma_addr_t dma_size)
 {
+	static struct lock_class_key lock_class;
 	struct drm_i915_gem_object *obj;
 	unsigned int cache_level;
 
@@ -111,7 +112,7 @@ huge_gem_object(struct drm_i915_private *i915,
 		return ERR_PTR(-ENOMEM);
 
 	drm_gem_private_object_init(&i915->drm, &obj->base, dma_size);
-	i915_gem_object_init(obj, &huge_ops);
+	i915_gem_object_init(obj, &huge_ops, &lock_class);
 
 	obj->read_domains = I915_GEM_DOMAIN_CPU;
 	obj->write_domain = I915_GEM_DOMAIN_CPU;
diff --git a/drivers/gpu/drm/i915/gem/selftests/huge_pages.c b/drivers/gpu/drm/i915/gem/selftests/huge_pages.c
index 8de83c6d81f5..688c49a24f32 100644
--- a/drivers/gpu/drm/i915/gem/selftests/huge_pages.c
+++ b/drivers/gpu/drm/i915/gem/selftests/huge_pages.c
@@ -8,6 +8,8 @@
 
 #include "i915_selftest.h"
 
+#include "gem/i915_gem_region.h"
+#include "gem/i915_gem_lmem.h"
 #include "gem/i915_gem_pm.h"
 
 #include "gt/intel_gt.h"
@@ -17,6 +19,7 @@
 
 #include "selftests/mock_drm.h"
 #include "selftests/mock_gem_device.h"
+#include "selftests/mock_region.h"
 #include "selftests/i915_random.h"
 
 static const unsigned int page_sizes[] = {
@@ -113,8 +116,6 @@ static int get_huge_pages(struct drm_i915_gem_object *obj)
 	if (i915_gem_gtt_prepare_pages(obj, st))
 		goto err;
 
-	obj->mm.madv = I915_MADV_DONTNEED;
-
 	GEM_BUG_ON(sg_page_sizes != obj->mm.page_mask);
 	__i915_gem_object_set_pages(obj, st, sg_page_sizes);
 
@@ -135,7 +136,6 @@ static void put_huge_pages(struct drm_i915_gem_object *obj,
 	huge_pages_free_pages(pages);
 
 	obj->mm.dirty = false;
-	obj->mm.madv = I915_MADV_WILLNEED;
 }
 
 static const struct drm_i915_gem_object_ops huge_page_ops = {
@@ -150,6 +150,7 @@ huge_pages_object(struct drm_i915_private *i915,
 		  u64 size,
 		  unsigned int page_mask)
 {
+	static struct lock_class_key lock_class;
 	struct drm_i915_gem_object *obj;
 
 	GEM_BUG_ON(!size);
@@ -166,7 +167,9 @@ huge_pages_object(struct drm_i915_private *i915,
 		return ERR_PTR(-ENOMEM);
 
 	drm_gem_private_object_init(&i915->drm, &obj->base, size);
-	i915_gem_object_init(obj, &huge_page_ops);
+	i915_gem_object_init(obj, &huge_page_ops, &lock_class);
+
+	i915_gem_object_set_volatile(obj);
 
 	obj->write_domain = I915_GEM_DOMAIN_CPU;
 	obj->read_domains = I915_GEM_DOMAIN_CPU;
@@ -227,8 +230,6 @@ static int fake_get_huge_pages(struct drm_i915_gem_object *obj)
 
 	i915_sg_trim(st);
 
-	obj->mm.madv = I915_MADV_DONTNEED;
-
 	__i915_gem_object_set_pages(obj, st, sg_page_sizes);
 
 	return 0;
@@ -261,8 +262,6 @@ static int fake_get_huge_pages_single(struct drm_i915_gem_object *obj)
 	sg_dma_len(sg) = obj->base.size;
 	sg_dma_address(sg) = page_size;
 
-	obj->mm.madv = I915_MADV_DONTNEED;
-
 	__i915_gem_object_set_pages(obj, st, sg->length);
 
 	return 0;
@@ -281,7 +280,6 @@ static void fake_put_huge_pages(struct drm_i915_gem_object *obj,
 {
 	fake_free_huge_pages(obj, pages);
 	obj->mm.dirty = false;
-	obj->mm.madv = I915_MADV_WILLNEED;
 }
 
 static const struct drm_i915_gem_object_ops fake_ops = {
@@ -299,6 +297,7 @@ static const struct drm_i915_gem_object_ops fake_ops_single = {
 static struct drm_i915_gem_object *
 fake_huge_pages_object(struct drm_i915_private *i915, u64 size, bool single)
 {
+	static struct lock_class_key lock_class;
 	struct drm_i915_gem_object *obj;
 
 	GEM_BUG_ON(!size);
@@ -317,9 +316,11 @@ fake_huge_pages_object(struct drm_i915_private *i915, u64 size, bool single)
 	drm_gem_private_object_init(&i915->drm, &obj->base, size);
 
 	if (single)
-		i915_gem_object_init(obj, &fake_ops_single);
+		i915_gem_object_init(obj, &fake_ops_single, &lock_class);
 	else
-		i915_gem_object_init(obj, &fake_ops);
+		i915_gem_object_init(obj, &fake_ops, &lock_class);
+
+	i915_gem_object_set_volatile(obj);
 
 	obj->write_domain = I915_GEM_DOMAIN_CPU;
 	obj->read_domains = I915_GEM_DOMAIN_CPU;
@@ -333,7 +334,12 @@ static int igt_check_page_sizes(struct i915_vma *vma)
 	struct drm_i915_private *i915 = vma->vm->i915;
 	unsigned int supported = INTEL_INFO(i915)->page_sizes;
 	struct drm_i915_gem_object *obj = vma->obj;
-	int err = 0;
+	int err;
+
+	/* We have to wait for the async bind to complete before our asserts */
+	err = i915_vma_sync(vma);
+	if (err)
+		return err;
 
 	if (!HAS_PAGE_SIZES(i915, vma->page_sizes.sg)) {
 		pr_err("unsupported page_sizes.sg=%u, supported=%u\n",
@@ -447,6 +453,88 @@ out_device:
 	return err;
 }
 
+static int igt_mock_memory_region_huge_pages(void *arg)
+{
+	const unsigned int flags[] = { 0, I915_BO_ALLOC_CONTIGUOUS };
+	struct i915_ppgtt *ppgtt = arg;
+	struct drm_i915_private *i915 = ppgtt->vm.i915;
+	unsigned long supported = INTEL_INFO(i915)->page_sizes;
+	struct intel_memory_region *mem;
+	struct drm_i915_gem_object *obj;
+	struct i915_vma *vma;
+	int bit;
+	int err = 0;
+
+	mem = mock_region_create(i915, 0, SZ_2G, I915_GTT_PAGE_SIZE_4K, 0);
+	if (IS_ERR(mem)) {
+		pr_err("%s failed to create memory region\n", __func__);
+		return PTR_ERR(mem);
+	}
+
+	for_each_set_bit(bit, &supported, ilog2(I915_GTT_MAX_PAGE_SIZE) + 1) {
+		unsigned int page_size = BIT(bit);
+		resource_size_t phys;
+		int i;
+
+		for (i = 0; i < ARRAY_SIZE(flags); ++i) {
+			obj = i915_gem_object_create_region(mem, page_size,
+							    flags[i]);
+			if (IS_ERR(obj)) {
+				err = PTR_ERR(obj);
+				goto out_region;
+			}
+
+			vma = i915_vma_instance(obj, &ppgtt->vm, NULL);
+			if (IS_ERR(vma)) {
+				err = PTR_ERR(vma);
+				goto out_put;
+			}
+
+			err = i915_vma_pin(vma, 0, 0, PIN_USER);
+			if (err)
+				goto out_close;
+
+			err = igt_check_page_sizes(vma);
+			if (err)
+				goto out_unpin;
+
+			phys = i915_gem_object_get_dma_address(obj, 0);
+			if (!IS_ALIGNED(phys, page_size)) {
+				pr_err("%s addr misaligned(%pa) page_size=%u\n",
+				       __func__, &phys, page_size);
+				err = -EINVAL;
+				goto out_unpin;
+			}
+
+			if (vma->page_sizes.gtt != page_size) {
+				pr_err("%s page_sizes.gtt=%u, expected=%u\n",
+				       __func__, vma->page_sizes.gtt,
+				       page_size);
+				err = -EINVAL;
+				goto out_unpin;
+			}
+
+			i915_vma_unpin(vma);
+			i915_vma_close(vma);
+
+			__i915_gem_object_put_pages(obj, I915_MM_NORMAL);
+			i915_gem_object_put(obj);
+		}
+	}
+
+	goto out_region;
+
+out_unpin:
+	i915_vma_unpin(vma);
+out_close:
+	i915_vma_close(vma);
+out_put:
+	i915_gem_object_put(obj);
+out_region:
+	intel_memory_region_put(mem);
+	return err;
+}
+
 static int igt_mock_ppgtt_misaligned_dma(void *arg)
 {
 	struct i915_ppgtt *ppgtt = arg;
@@ -879,9 +967,8 @@ out_object_put:
 	return err;
 }
 
-static int gpu_write(struct i915_vma *vma,
-		     struct i915_gem_context *ctx,
-		     struct intel_engine_cs *engine,
+static int gpu_write(struct intel_context *ce,
+		     struct i915_vma *vma,
 		     u32 dw,
 		     u32 val)
 {
@@ -893,11 +980,12 @@ static int gpu_write(struct i915_vma *vma,
 	if (err)
 		return err;
 
-	return igt_gpu_fill_dw(vma, ctx, engine, dw * sizeof(u32),
+	return igt_gpu_fill_dw(ce, vma, dw * sizeof(u32),
 			       vma->size >> PAGE_SHIFT, val);
 }
 
-static int cpu_check(struct drm_i915_gem_object *obj, u32 dword, u32 val)
+static int
+__cpu_check_shmem(struct drm_i915_gem_object *obj, u32 dword, u32 val)
 {
 	unsigned int needs_flush;
 	unsigned long n;
@@ -929,18 +1017,61 @@ static int cpu_check(struct drm_i915_gem_object *obj, u32 dword, u32 val)
 	return err;
 }
 
-static int __igt_write_huge(struct i915_gem_context *ctx,
-			    struct intel_engine_cs *engine,
+static int __cpu_check_lmem(struct drm_i915_gem_object *obj, u32 dword, u32 val)
+{
+	unsigned long n;
+	int err;
+
+	i915_gem_object_lock(obj);
+	err = i915_gem_object_set_to_wc_domain(obj, false);
+	i915_gem_object_unlock(obj);
+	if (err)
+		return err;
+
+	err = i915_gem_object_pin_pages(obj);
+	if (err)
+		return err;
+
+	for (n = 0; n < obj->base.size >> PAGE_SHIFT; ++n) {
+		u32 __iomem *base;
+		u32 read_val;
+
+		base = i915_gem_object_lmem_io_map_page_atomic(obj, n);
+
+		read_val = ioread32(base + dword);
+		io_mapping_unmap_atomic(base);
+		if (read_val != val) {
+			pr_err("n=%lu base[%u]=%u, val=%u\n",
+			       n, dword, read_val, val);
+			err = -EINVAL;
+			break;
+		}
+	}
+
+	i915_gem_object_unpin_pages(obj);
+	return err;
+}
+
+static int cpu_check(struct drm_i915_gem_object *obj, u32 dword, u32 val)
+{
+	if (i915_gem_object_has_struct_page(obj))
+		return __cpu_check_shmem(obj, dword, val);
+	else if (i915_gem_object_is_lmem(obj))
+		return __cpu_check_lmem(obj, dword, val);
+
+	return -ENODEV;
+}
+
+static int __igt_write_huge(struct intel_context *ce,
 			    struct drm_i915_gem_object *obj,
 			    u64 size, u64 offset,
 			    u32 dword, u32 val)
 {
-	struct i915_address_space *vm = ctx->vm ?: &engine->gt->ggtt->vm;
 	unsigned int flags = PIN_USER | PIN_OFFSET_FIXED;
 	struct i915_vma *vma;
 	int err;
 
-	vma = i915_vma_instance(obj, vm, NULL);
+	vma = i915_vma_instance(obj, ce->vm, NULL);
 	if (IS_ERR(vma))
 		return PTR_ERR(vma);
 
@@ -954,7 +1085,7 @@ static int __igt_write_huge(struct i915_gem_context *ctx,
 		 * The ggtt may have some pages reserved so
 		 * refrain from erroring out.
 		 */
-		if (err == -ENOSPC && i915_is_ggtt(vm))
+		if (err == -ENOSPC && i915_is_ggtt(ce->vm))
 			err = 0;
 
 		goto out_vma_close;
@@ -964,7 +1095,7 @@ static int __igt_write_huge(struct i915_gem_context *ctx,
 	if (err)
 		goto out_vma_unpin;
 
-	err = gpu_write(vma, ctx, engine, dword, val);
+	err = gpu_write(ce, vma, dword, val);
 	if (err) {
 		pr_err("gpu-write failed at offset=%llx\n", offset);
 		goto out_vma_unpin;
@@ -987,14 +1118,13 @@ out_vma_close:
 static int igt_write_huge(struct i915_gem_context *ctx,
 			  struct drm_i915_gem_object *obj)
 {
-	struct drm_i915_private *i915 = to_i915(obj->base.dev);
-	struct i915_address_space *vm = ctx->vm ?: &i915->ggtt.vm;
-	static struct intel_engine_cs *engines[I915_NUM_ENGINES];
-	struct intel_engine_cs *engine;
+	struct i915_gem_engines *engines;
+	struct i915_gem_engines_iter it;
+	struct intel_context *ce;
 	I915_RND_STATE(prng);
 	IGT_TIMEOUT(end_time);
 	unsigned int max_page_size;
-	unsigned int id;
+	unsigned int count;
 	u64 max;
 	u64 num;
 	u64 size;
@@ -1008,19 +1138,18 @@ static int igt_write_huge(struct i915_gem_context *ctx,
 	if (obj->mm.page_sizes.sg & I915_GTT_PAGE_SIZE_64K)
 		size = round_up(size, I915_GTT_PAGE_SIZE_2M);
 
-	max_page_size = rounddown_pow_of_two(obj->mm.page_sizes.sg);
-	max = div_u64((vm->total - size), max_page_size);
-
 	n = 0;
-	for_each_engine(engine, i915, id) {
-		if (!intel_engine_can_store_dword(engine)) {
-			pr_info("store-dword-imm not supported on engine=%u\n",
-				id);
+	count = 0;
+	max = U64_MAX;
+	for_each_gem_engine(ce, i915_gem_context_lock_engines(ctx), it) {
+		count++;
+		if (!intel_engine_can_store_dword(ce->engine))
 			continue;
-		}
-		engines[n++] = engine;
-	}
 
+		max = min(max, ce->vm->total);
+		n++;
+	}
+	i915_gem_context_unlock_engines(ctx);
 	if (!n)
 		return 0;
 
@@ -1029,23 +1158,30 @@ static int igt_write_huge(struct i915_gem_context *ctx,
 	 * randomized order, lets also make feeding to the same engine a few
 	 * times in succession a possibility by enlarging the permutation array.
 	 */
-	order = i915_random_order(n * I915_NUM_ENGINES, &prng);
+	order = i915_random_order(count * count, &prng);
 	if (!order)
 		return -ENOMEM;
 
+	max_page_size = rounddown_pow_of_two(obj->mm.page_sizes.sg);
+	max = div_u64(max - size, max_page_size);
+
 	/*
 	 * Try various offsets in an ascending/descending fashion until we
 	 * timeout -- we want to avoid issues hidden by effectively always using
 	 * offset = 0.
 	 */
 	i = 0;
+	engines = i915_gem_context_lock_engines(ctx);
 	for_each_prime_number_from(num, 0, max) {
 		u64 offset_low = num * max_page_size;
 		u64 offset_high = (max - num) * max_page_size;
 		u32 dword = offset_in_page(num) / 4;
+		struct intel_context *ce;
 
-		engine = engines[order[i] % n];
-		i = (i + 1) % (n * I915_NUM_ENGINES);
+		ce = engines->engines[order[i] % engines->num_engines];
+		i = (i + 1) % (count * count);
+		if (!ce || !intel_engine_can_store_dword(ce->engine))
+			continue;
 
 		/*
 		 * In order to utilize 64K pages we need to both pad the vma
@@ -1057,22 +1193,23 @@ static int igt_write_huge(struct i915_gem_context *ctx,
 			offset_low = round_down(offset_low,
 						I915_GTT_PAGE_SIZE_2M);
 
-		err = __igt_write_huge(ctx, engine, obj, size, offset_low,
+		err = __igt_write_huge(ce, obj, size, offset_low,
 				       dword, num + 1);
 		if (err)
 			break;
 
-		err = __igt_write_huge(ctx, engine, obj, size, offset_high,
+		err = __igt_write_huge(ce, obj, size, offset_high,
 				       dword, num + 1);
 		if (err)
 			break;
 
 		if (igt_timeout(end_time,
-				"%s timed out on engine=%u, offset_low=%llx offset_high=%llx, max_page_size=%x\n",
-				__func__, engine->id, offset_low, offset_high,
+				"%s timed out on %s, offset_low=%llx offset_high=%llx, max_page_size=%x\n",
+				__func__, ce->engine->name, offset_low, offset_high,
 				max_page_size))
 			break;
 	}
+	i915_gem_context_unlock_engines(ctx);
 
 	kfree(order);
 
@@ -1180,131 +1317,235 @@ out_device:
 	return err;
 }
 
-static int igt_ppgtt_internal_huge(void *arg)
+typedef struct drm_i915_gem_object *
+(*igt_create_fn)(struct drm_i915_private *i915, u32 size, u32 flags);
+
+static inline bool igt_can_allocate_thp(struct drm_i915_private *i915)
+{
+	return i915->mm.gemfs && has_transparent_hugepage();
+}
+
+static struct drm_i915_gem_object *
+igt_create_shmem(struct drm_i915_private *i915, u32 size, u32 flags)
+{
+	if (!igt_can_allocate_thp(i915)) {
+		pr_info("%s missing THP support, skipping\n", __func__);
+		return ERR_PTR(-ENODEV);
+	}
+
+	return i915_gem_object_create_shmem(i915, size);
+}
+
+static struct drm_i915_gem_object *
+igt_create_internal(struct drm_i915_private *i915, u32 size, u32 flags)
+{
+	return i915_gem_object_create_internal(i915, size);
+}
+
+static struct drm_i915_gem_object *
+igt_create_system(struct drm_i915_private *i915, u32 size, u32 flags)
+{
+	return huge_pages_object(i915, size, size);
+}
+
+static struct drm_i915_gem_object *
+igt_create_local(struct drm_i915_private *i915, u32 size, u32 flags)
+{
+	return i915_gem_object_create_lmem(i915, size, flags);
+}
+
+static u32 igt_random_size(struct rnd_state *prng,
+			   u32 min_page_size,
+			   u32 max_page_size)
+{
+	u64 mask;
+	u32 size;
+
+	GEM_BUG_ON(!is_power_of_2(min_page_size));
+	GEM_BUG_ON(!is_power_of_2(max_page_size));
+	GEM_BUG_ON(min_page_size < PAGE_SIZE);
+	GEM_BUG_ON(min_page_size > max_page_size);
+
+	mask = ((max_page_size << 1ULL) - 1) & PAGE_MASK;
+	size = prandom_u32_state(prng) & mask;
+	if (size < min_page_size)
+		size |= min_page_size;
+
+	return size;
+}
+
+static int igt_ppgtt_smoke_huge(void *arg)
 {
 	struct i915_gem_context *ctx = arg;
 	struct drm_i915_private *i915 = ctx->i915;
 	struct drm_i915_gem_object *obj;
-	static const unsigned int sizes[] = {
-		SZ_64K,
-		SZ_128K,
-		SZ_256K,
-		SZ_512K,
-		SZ_1M,
-		SZ_2M,
+	I915_RND_STATE(prng);
+	struct {
+		igt_create_fn fn;
+		u32 min;
+		u32 max;
+	} backends[] = {
+		{ igt_create_internal, SZ_64K, SZ_2M,  },
+		{ igt_create_shmem,    SZ_64K, SZ_32M, },
+		{ igt_create_local,    SZ_64K, SZ_1G,  },
 	};
-	int i;
 	int err;
+	int i;
 
 	/*
-	 * Sanity check that the HW uses huge pages correctly through internal
-	 * -- ensure that our writes land in the right place.
+	 * Sanity check that the HW uses huge pages correctly through our
+	 * various backends -- ensure that our writes land in the right place.
 	 */
 
-	for (i = 0; i < ARRAY_SIZE(sizes); ++i) {
-		unsigned int size = sizes[i];
+	for (i = 0; i < ARRAY_SIZE(backends); ++i) {
+		u32 min = backends[i].min;
+		u32 max = backends[i].max;
+		u32 size = max;
+try_again:
+		size = igt_random_size(&prng, min, rounddown_pow_of_two(size));
 
-		obj = i915_gem_object_create_internal(i915, size);
-		if (IS_ERR(obj))
-			return PTR_ERR(obj);
+		obj = backends[i].fn(i915, size, 0);
+		if (IS_ERR(obj)) {
+			err = PTR_ERR(obj);
+			if (err == -E2BIG) {
+				size >>= 1;
+				goto try_again;
+			} else if (err == -ENODEV) {
+				err = 0;
+				continue;
+			}
+
+			return err;
+		}
 
 		err = i915_gem_object_pin_pages(obj);
-		if (err)
+		if (err) {
+			if (err == -ENXIO) {
+				i915_gem_object_put(obj);
+				size >>= 1;
+				goto try_again;
+			}
 			goto out_put;
+		}
 
-		if (obj->mm.page_sizes.phys < I915_GTT_PAGE_SIZE_64K) {
-			pr_info("internal unable to allocate huge-page(s) with size=%u\n",
-				size);
+		if (obj->mm.page_sizes.phys < min) {
+			pr_info("%s unable to allocate huge-page(s) with size=%u, i=%d\n",
+				__func__, size, i);
+			err = -ENOMEM;
 			goto out_unpin;
 		}
 
 		err = igt_write_huge(ctx, obj);
 		if (err) {
-			pr_err("internal write-huge failed with size=%u\n",
-			       size);
-			goto out_unpin;
+			pr_err("%s write-huge failed with size=%u, i=%d\n",
+			       __func__, size, i);
 		}
-
+out_unpin:
 		i915_gem_object_unpin_pages(obj);
 		__i915_gem_object_put_pages(obj, I915_MM_NORMAL);
+out_put:
 		i915_gem_object_put(obj);
-	}
 
-	return 0;
+		if (err == -ENOMEM || err == -ENXIO)
+			err = 0;
 
-out_unpin:
-	i915_gem_object_unpin_pages(obj);
-out_put:
-	i915_gem_object_put(obj);
+		if (err)
+			break;
 
-	return err;
-}
+		cond_resched();
+	}
 
-static inline bool igt_can_allocate_thp(struct drm_i915_private *i915)
-{
-	return i915->mm.gemfs && has_transparent_hugepage();
+	return err;
 }
 
-static int igt_ppgtt_gemfs_huge(void *arg)
+static int igt_ppgtt_sanity_check(void *arg)
 {
 	struct i915_gem_context *ctx = arg;
 	struct drm_i915_private *i915 = ctx->i915;
-	struct drm_i915_gem_object *obj;
-	static const unsigned int sizes[] = {
-		SZ_2M,
-		SZ_4M,
-		SZ_8M,
-		SZ_16M,
-		SZ_32M,
+	unsigned int supported = INTEL_INFO(i915)->page_sizes;
+	struct {
+		igt_create_fn fn;
+		unsigned int flags;
+	} backends[] = {
+		{ igt_create_system, 0,                        },
+		{ igt_create_local,  I915_BO_ALLOC_CONTIGUOUS, },
 	};
-	int i;
+	struct {
+		u32 size;
+		u32 pages;
+	} combos[] = {
+		{ SZ_64K,		SZ_64K		},
+		{ SZ_2M,		SZ_2M		},
+		{ SZ_2M,		SZ_64K		},
+		{ SZ_2M - SZ_64K,	SZ_64K		},
+		{ SZ_2M - SZ_4K,	SZ_64K | SZ_4K	},
+		{ SZ_2M + SZ_4K,	SZ_64K | SZ_4K	},
+		{ SZ_2M + SZ_4K,	SZ_2M  | SZ_4K	},
+		{ SZ_2M + SZ_64K,	SZ_2M  | SZ_64K },
+	};
+	int i, j;
 	int err;
 
+	if (supported == I915_GTT_PAGE_SIZE_4K)
+		return 0;
+
 	/*
-	 * Sanity check that the HW uses huge pages correctly through gemfs --
-	 * ensure that our writes land in the right place.
+	 * Sanity check that the HW behaves with a limited set of combinations.
+	 * We already have a bunch of randomised testing, which should give us
+	 * a decent amount of variation between runs, however we should keep
+	 * this to limit the chances of introducing a temporary regression, by
+	 * testing the most obvious cases that might make something blow up.
 	 */
 
-	if (!igt_can_allocate_thp(i915)) {
-		pr_info("missing THP support, skipping\n");
-		return 0;
-	}
+	for (i = 0; i < ARRAY_SIZE(backends); ++i) {
+		for (j = 0; j < ARRAY_SIZE(combos); ++j) {
+			struct drm_i915_gem_object *obj;
+			u32 size = combos[j].size;
+			u32 pages = combos[j].pages;
 
-	for (i = 0; i < ARRAY_SIZE(sizes); ++i) {
-		unsigned int size = sizes[i];
+			obj = backends[i].fn(i915, size, backends[i].flags);
+			if (IS_ERR(obj)) {
+				err = PTR_ERR(obj);
+				if (err == -ENODEV) {
+					pr_info("Device lacks local memory, skipping\n");
+					err = 0;
+					break;
+				}
 
-		obj = i915_gem_object_create_shmem(i915, size);
-		if (IS_ERR(obj))
-			return PTR_ERR(obj);
+				return err;
+			}
 
-		err = i915_gem_object_pin_pages(obj);
-		if (err)
-			goto out_put;
+			err = i915_gem_object_pin_pages(obj);
+			if (err) {
+				i915_gem_object_put(obj);
+				goto out;
+			}
 
-		if (obj->mm.page_sizes.phys < I915_GTT_PAGE_SIZE_2M) {
-			pr_info("finishing test early, gemfs unable to allocate huge-page(s) with size=%u\n",
-				size);
-			goto out_unpin;
-		}
+			GEM_BUG_ON(pages > obj->base.size);
+			pages = pages & supported;
 
-		err = igt_write_huge(ctx, obj);
-		if (err) {
-			pr_err("gemfs write-huge failed with size=%u\n",
-			       size);
-			goto out_unpin;
+			if (pages)
+				obj->mm.page_sizes.sg = pages;
+
+			err = igt_write_huge(ctx, obj);
+
+			i915_gem_object_unpin_pages(obj);
+			__i915_gem_object_put_pages(obj, I915_MM_NORMAL);
+			i915_gem_object_put(obj);
+
+			if (err) {
+				pr_err("%s write-huge failed with size=%u pages=%u i=%d, j=%d\n",
+				       __func__, size, pages, i, j);
+				goto out;
+			}
 		}
 
-		i915_gem_object_unpin_pages(obj);
-		__i915_gem_object_put_pages(obj, I915_MM_NORMAL);
-		i915_gem_object_put(obj);
+		cond_resched();
 	}
 
-	return 0;
-
-out_unpin:
-	i915_gem_object_unpin_pages(obj);
-out_put:
-	i915_gem_object_put(obj);
+out:
+	if (err == -ENOMEM)
+		err = 0;
 
 	return err;
 }
@@ -1314,15 +1555,15 @@ static int igt_ppgtt_pin_update(void *arg)
 	struct i915_gem_context *ctx = arg;
 	struct drm_i915_private *dev_priv = ctx->i915;
 	unsigned long supported = INTEL_INFO(dev_priv)->page_sizes;
-	struct i915_address_space *vm = ctx->vm;
 	struct drm_i915_gem_object *obj;
+	struct i915_gem_engines_iter it;
+	struct i915_address_space *vm;
+	struct intel_context *ce;
 	struct i915_vma *vma;
 	unsigned int flags = PIN_USER | PIN_OFFSET_FIXED;
-	struct intel_engine_cs *engine;
-	enum intel_engine_id id;
 	unsigned int n;
 	int first, last;
-	int err;
+	int err = 0;
 
 	/*
 	 * Make sure there's no funny business when doing a PIN_UPDATE -- in the
@@ -1332,9 +1573,10 @@ static int igt_ppgtt_pin_update(void *arg)
 	 * huge-gtt-pages.
 	 */
 
-	if (!vm || !i915_vm_is_4lvl(vm)) {
+	vm = i915_gem_context_get_vm_rcu(ctx);
+	if (!i915_vm_is_4lvl(vm)) {
 		pr_info("48b PPGTT not supported, skipping\n");
-		return 0;
+		goto out_vm;
 	}
 
 	first = ilog2(I915_GTT_PAGE_SIZE_64K);
@@ -1387,7 +1629,7 @@ static int igt_ppgtt_pin_update(void *arg)
 			goto out_unpin;
 		}
 
-		err = i915_vma_bind(vma, I915_CACHE_NONE, PIN_UPDATE);
+		err = i915_vma_bind(vma, I915_CACHE_NONE, PIN_UPDATE, NULL);
 		if (err)
 			goto out_unpin;
 
@@ -1419,14 +1661,18 @@ static int igt_ppgtt_pin_update(void *arg)
 	 */
 
 	n = 0;
-	for_each_engine(engine, dev_priv, id) {
-		if (!intel_engine_can_store_dword(engine))
+	for_each_gem_engine(ce, i915_gem_context_lock_engines(ctx), it) {
+		if (!intel_engine_can_store_dword(ce->engine))
 			continue;
 
-		err = gpu_write(vma, ctx, engine, n++, 0xdeadbeaf);
+		err = gpu_write(ce, vma, n++, 0xdeadbeaf);
 		if (err)
-			goto out_unpin;
+			break;
 	}
+	i915_gem_context_unlock_engines(ctx);
+	if (err)
+		goto out_unpin;
+
 	while (n--) {
 		err = cpu_check(obj, n, 0xdeadbeaf);
 		if (err)
@@ -1439,6 +1685,8 @@ out_close:
 	i915_vma_close(vma);
 out_put:
 	i915_gem_object_put(obj);
+out_vm:
+	i915_vm_put(vm);
 
 	return err;
 }
@@ -1448,7 +1696,7 @@ static int igt_tmpfs_fallback(void *arg)
 	struct i915_gem_context *ctx = arg;
 	struct drm_i915_private *i915 = ctx->i915;
 	struct vfsmount *gemfs = i915->mm.gemfs;
-	struct i915_address_space *vm = ctx->vm ?: &i915->ggtt.vm;
+	struct i915_address_space *vm = i915_gem_context_get_vm_rcu(ctx);
 	struct drm_i915_gem_object *obj;
 	struct i915_vma *vma;
 	u32 *vaddr;
@@ -1498,6 +1746,7 @@ out_put:
 out_restore:
 	i915->mm.gemfs = gemfs;
 
+	i915_vm_put(vm);
 	return err;
 }
 
@@ -1505,14 +1754,14 @@ static int igt_shrink_thp(void *arg)
 {
 	struct i915_gem_context *ctx = arg;
 	struct drm_i915_private *i915 = ctx->i915;
-	struct i915_address_space *vm = ctx->vm ?: &i915->ggtt.vm;
+	struct i915_address_space *vm = i915_gem_context_get_vm_rcu(ctx);
 	struct drm_i915_gem_object *obj;
-	struct intel_engine_cs *engine;
-	enum intel_engine_id id;
+	struct i915_gem_engines_iter it;
+	struct intel_context *ce;
 	struct i915_vma *vma;
 	unsigned int flags = PIN_USER;
 	unsigned int n;
-	int err;
+	int err = 0;
 
 	/*
 	 * Sanity check shrinking huge-paged object -- make sure nothing blows
@@ -1521,12 +1770,14 @@ static int igt_shrink_thp(void *arg)
 
 	if (!igt_can_allocate_thp(i915)) {
 		pr_info("missing THP support, skipping\n");
-		return 0;
+		goto out_vm;
 	}
 
 	obj = i915_gem_object_create_shmem(i915, SZ_2M);
-	if (IS_ERR(obj))
-		return PTR_ERR(obj);
+	if (IS_ERR(obj)) {
+		err = PTR_ERR(obj);
+		goto out_vm;
+	}
 
 	vma = i915_vma_instance(obj, vm, NULL);
 	if (IS_ERR(vma)) {
@@ -1548,16 +1799,19 @@ static int igt_shrink_thp(void *arg)
 		goto out_unpin;
 
 	n = 0;
-	for_each_engine(engine, i915, id) {
-		if (!intel_engine_can_store_dword(engine))
+
+	for_each_gem_engine(ce, i915_gem_context_lock_engines(ctx), it) {
+		if (!intel_engine_can_store_dword(ce->engine))
 			continue;
 
-		err = gpu_write(vma, ctx, engine, n++, 0xdeadbeaf);
+		err = gpu_write(ce, vma, n++, 0xdeadbeaf);
 		if (err)
-			goto out_unpin;
+			break;
 	}
-
+	i915_gem_context_unlock_engines(ctx);
 	i915_vma_unpin(vma);
+	if (err)
+		goto out_close;
 
 	/*
 	 * Now that the pages are *unpinned* shrink-all should invoke
@@ -1583,16 +1837,17 @@ static int igt_shrink_thp(void *arg)
 	while (n--) {
 		err = cpu_check(obj, n, 0xdeadbeaf);
 		if (err)
-			goto out_unpin;
+			break;
 	}
 
-
 out_unpin:
 	i915_vma_unpin(vma);
 out_close:
 	i915_vma_close(vma);
 out_put:
 	i915_gem_object_put(obj);
+out_vm:
+	i915_vm_put(vm);
 
 	return err;
 }
@@ -1601,6 +1856,7 @@ int i915_gem_huge_page_mock_selftests(void)
 {
 	static const struct i915_subtest tests[] = {
 		SUBTEST(igt_mock_exhaust_device_supported_pages),
+		SUBTEST(igt_mock_memory_region_huge_pages),
 		SUBTEST(igt_mock_ppgtt_misaligned_dma),
 		SUBTEST(igt_mock_ppgtt_huge_fill),
 		SUBTEST(igt_mock_ppgtt_64K),
@@ -1617,7 +1873,6 @@ int i915_gem_huge_page_mock_selftests(void)
 	mkwrite_device_info(dev_priv)->ppgtt_type = INTEL_PPGTT_FULL;
 	mkwrite_device_info(dev_priv)->ppgtt_size = 48;
 
-	mutex_lock(&dev_priv->drm.struct_mutex);
 	ppgtt = i915_ppgtt_create(dev_priv);
 	if (IS_ERR(ppgtt)) {
 		err = PTR_ERR(ppgtt);
@@ -1643,9 +1898,7 @@ out_close:
 	i915_vm_put(&ppgtt->vm);
 
 out_unlock:
-	mutex_unlock(&dev_priv->drm.struct_mutex);
 	drm_dev_put(&dev_priv->drm);
-
 	return err;
 }
 
@@ -1656,12 +1909,12 @@ int i915_gem_huge_page_live_selftests(struct drm_i915_private *i915)
 		SUBTEST(igt_ppgtt_pin_update),
 		SUBTEST(igt_tmpfs_fallback),
 		SUBTEST(igt_ppgtt_exhaust_huge),
-		SUBTEST(igt_ppgtt_gemfs_huge),
-		SUBTEST(igt_ppgtt_internal_huge),
+		SUBTEST(igt_ppgtt_smoke_huge),
+		SUBTEST(igt_ppgtt_sanity_check),
 	};
 	struct drm_file *file;
 	struct i915_gem_context *ctx;
-	intel_wakeref_t wakeref;
+	struct i915_address_space *vm;
 	int err;
 
 	if (!HAS_PPGTT(i915)) {
@@ -1676,25 +1929,21 @@ int i915_gem_huge_page_live_selftests(struct drm_i915_private *i915)
 	if (IS_ERR(file))
 		return PTR_ERR(file);
 
-	mutex_lock(&i915->drm.struct_mutex);
-	wakeref = intel_runtime_pm_get(&i915->runtime_pm);
-
 	ctx = live_context(i915, file);
 	if (IS_ERR(ctx)) {
 		err = PTR_ERR(ctx);
-		goto out_unlock;
+		goto out_file;
 	}
 
-	if (ctx->vm)
-		ctx->vm->scrub_64K = true;
+	mutex_lock(&ctx->mutex);
+	vm = i915_gem_context_vm(ctx);
+	if (vm)
+		WRITE_ONCE(vm->scrub_64K, true);
+	mutex_unlock(&ctx->mutex);
 
 	err = i915_subtests(tests, ctx);
 
-out_unlock:
-	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
-	mutex_unlock(&i915->drm.struct_mutex);
-
+out_file:
 	mock_file_free(i915, file);
-
 	return err;
 }
diff --git a/drivers/gpu/drm/i915/gem/selftests/i915_gem_client_blt.c b/drivers/gpu/drm/i915/gem/selftests/i915_gem_client_blt.c
index d8804a847945..da8edee4fe0a 100644
--- a/drivers/gpu/drm/i915/gem/selftests/i915_gem_client_blt.c
+++ b/drivers/gpu/drm/i915/gem/selftests/i915_gem_client_blt.c
@@ -5,6 +5,7 @@
 
 #include "i915_selftest.h"
 
+#include "gt/intel_engine_user.h"
 #include "gt/intel_gt.h"
 
 #include "selftests/igt_flush_test.h"
@@ -12,10 +13,9 @@
 #include "huge_gem_object.h"
 #include "mock_context.h"
 
-static int igt_client_fill(void *arg)
+static int __igt_client_fill(struct intel_engine_cs *engine)
 {
-	struct drm_i915_private *i915 = arg;
-	struct intel_context *ce = i915->engine[BCS0]->kernel_context;
+	struct intel_context *ce = engine->kernel_context;
 	struct drm_i915_gem_object *obj;
 	struct rnd_state prng;
 	IGT_TIMEOUT(end);
@@ -37,7 +37,7 @@ static int igt_client_fill(void *arg)
 		pr_debug("%s with phys_sz= %x, sz=%x, val=%x\n", __func__,
 			 phys_sz, sz, val);
 
-		obj = huge_gem_object(i915, phys_sz, sz);
+		obj = huge_gem_object(engine->i915, phys_sz, sz);
 		if (IS_ERR(obj)) {
 			err = PTR_ERR(obj);
 			goto err_flush;
@@ -103,6 +103,28 @@ err_flush:
 	return err;
 }
 
+static int igt_client_fill(void *arg)
+{
+	int inst = 0;
+
+	do {
+		struct intel_engine_cs *engine;
+		int err;
+
+		engine = intel_engine_lookup_user(arg,
+						  I915_ENGINE_CLASS_COPY,
+						  inst++);
+		if (!engine)
+			return 0;
+
+		err = __igt_client_fill(engine);
+		if (err == -ENOMEM)
+			err = 0;
+		if (err)
+			return err;
+	} while (1);
+}
+
 int i915_gem_client_blt_live_selftests(struct drm_i915_private *i915)
 {
 	static const struct i915_subtest tests[] = {
diff --git a/drivers/gpu/drm/i915/gem/selftests/i915_gem_coherency.c b/drivers/gpu/drm/i915/gem/selftests/i915_gem_coherency.c
index 0ff7a89aadca..2b29f6b4e1dd 100644
--- a/drivers/gpu/drm/i915/gem/selftests/i915_gem_coherency.c
+++ b/drivers/gpu/drm/i915/gem/selftests/i915_gem_coherency.c
@@ -7,13 +7,18 @@
 #include <linux/prime_numbers.h>
 
 #include "gt/intel_gt.h"
+#include "gt/intel_gt_pm.h"
+#include "gt/intel_ring.h"
 
 #include "i915_selftest.h"
 #include "selftests/i915_random.h"
 
-static int cpu_set(struct drm_i915_gem_object *obj,
-		   unsigned long offset,
-		   u32 v)
+struct context {
+	struct drm_i915_gem_object *obj;
+	struct intel_engine_cs *engine;
+};
+
+static int cpu_set(struct context *ctx, unsigned long offset, u32 v)
 {
 	unsigned int needs_clflush;
 	struct page *page;
@@ -21,11 +26,11 @@ static int cpu_set(struct drm_i915_gem_object *obj,
 	u32 *cpu;
 	int err;
 
-	err = i915_gem_object_prepare_write(obj, &needs_clflush);
+	err = i915_gem_object_prepare_write(ctx->obj, &needs_clflush);
 	if (err)
 		return err;
 
-	page = i915_gem_object_get_page(obj, offset >> PAGE_SHIFT);
+	page = i915_gem_object_get_page(ctx->obj, offset >> PAGE_SHIFT);
 	map = kmap_atomic(page);
 	cpu = map + offset_in_page(offset);
 
@@ -38,14 +43,12 @@ static int cpu_set(struct drm_i915_gem_object *obj,
 		drm_clflush_virt_range(cpu, sizeof(*cpu));
 
 	kunmap_atomic(map);
-	i915_gem_object_finish_access(obj);
+	i915_gem_object_finish_access(ctx->obj);
 
 	return 0;
 }
 
-static int cpu_get(struct drm_i915_gem_object *obj,
-		   unsigned long offset,
-		   u32 *v)
+static int cpu_get(struct context *ctx, unsigned long offset, u32 *v)
 {
 	unsigned int needs_clflush;
 	struct page *page;
@@ -53,11 +56,11 @@ static int cpu_get(struct drm_i915_gem_object *obj,
 	u32 *cpu;
 	int err;
 
-	err = i915_gem_object_prepare_read(obj, &needs_clflush);
+	err = i915_gem_object_prepare_read(ctx->obj, &needs_clflush);
 	if (err)
 		return err;
 
-	page = i915_gem_object_get_page(obj, offset >> PAGE_SHIFT);
+	page = i915_gem_object_get_page(ctx->obj, offset >> PAGE_SHIFT);
 	map = kmap_atomic(page);
 	cpu = map + offset_in_page(offset);
 
@@ -67,136 +70,137 @@ static int cpu_get(struct drm_i915_gem_object *obj,
 	*v = *cpu;
 
 	kunmap_atomic(map);
-	i915_gem_object_finish_access(obj);
+	i915_gem_object_finish_access(ctx->obj);
 
 	return 0;
 }
 
-static int gtt_set(struct drm_i915_gem_object *obj,
-		   unsigned long offset,
-		   u32 v)
+static int gtt_set(struct context *ctx, unsigned long offset, u32 v)
 {
 	struct i915_vma *vma;
 	u32 __iomem *map;
-	int err;
+	int err = 0;
 
-	i915_gem_object_lock(obj);
-	err = i915_gem_object_set_to_gtt_domain(obj, true);
-	i915_gem_object_unlock(obj);
+	i915_gem_object_lock(ctx->obj);
+	err = i915_gem_object_set_to_gtt_domain(ctx->obj, true);
+	i915_gem_object_unlock(ctx->obj);
 	if (err)
 		return err;
 
-	vma = i915_gem_object_ggtt_pin(obj, NULL, 0, 0, PIN_MAPPABLE);
+	vma = i915_gem_object_ggtt_pin(ctx->obj, NULL, 0, 0, PIN_MAPPABLE);
 	if (IS_ERR(vma))
 		return PTR_ERR(vma);
 
+	intel_gt_pm_get(vma->vm->gt);
+
 	map = i915_vma_pin_iomap(vma);
 	i915_vma_unpin(vma);
-	if (IS_ERR(map))
-		return PTR_ERR(map);
+	if (IS_ERR(map)) {
+		err = PTR_ERR(map);
+		goto out_rpm;
+	}
 
 	iowrite32(v, &map[offset / sizeof(*map)]);
 	i915_vma_unpin_iomap(vma);
 
-	return 0;
+out_rpm:
+	intel_gt_pm_put(vma->vm->gt);
+	return err;
 }
 
-static int gtt_get(struct drm_i915_gem_object *obj,
-		   unsigned long offset,
-		   u32 *v)
+static int gtt_get(struct context *ctx, unsigned long offset, u32 *v)
 {
 	struct i915_vma *vma;
 	u32 __iomem *map;
-	int err;
+	int err = 0;
 
-	i915_gem_object_lock(obj);
-	err = i915_gem_object_set_to_gtt_domain(obj, false);
-	i915_gem_object_unlock(obj);
+	i915_gem_object_lock(ctx->obj);
+	err = i915_gem_object_set_to_gtt_domain(ctx->obj, false);
+	i915_gem_object_unlock(ctx->obj);
 	if (err)
 		return err;
 
-	vma = i915_gem_object_ggtt_pin(obj, NULL, 0, 0, PIN_MAPPABLE);
+	vma = i915_gem_object_ggtt_pin(ctx->obj, NULL, 0, 0, PIN_MAPPABLE);
 	if (IS_ERR(vma))
 		return PTR_ERR(vma);
 
+	intel_gt_pm_get(vma->vm->gt);
+
 	map = i915_vma_pin_iomap(vma);
 	i915_vma_unpin(vma);
-	if (IS_ERR(map))
-		return PTR_ERR(map);
+	if (IS_ERR(map)) {
+		err = PTR_ERR(map);
+		goto out_rpm;
+	}
 
 	*v = ioread32(&map[offset / sizeof(*map)]);
 	i915_vma_unpin_iomap(vma);
 
-	return 0;
+out_rpm:
+	intel_gt_pm_put(vma->vm->gt);
+	return err;
 }
 
-static int wc_set(struct drm_i915_gem_object *obj,
-		  unsigned long offset,
-		  u32 v)
+static int wc_set(struct context *ctx, unsigned long offset, u32 v)
 {
 	u32 *map;
 	int err;
 
-	i915_gem_object_lock(obj);
-	err = i915_gem_object_set_to_wc_domain(obj, true);
-	i915_gem_object_unlock(obj);
+	i915_gem_object_lock(ctx->obj);
+	err = i915_gem_object_set_to_wc_domain(ctx->obj, true);
+	i915_gem_object_unlock(ctx->obj);
 	if (err)
 		return err;
 
-	map = i915_gem_object_pin_map(obj, I915_MAP_WC);
+	map = i915_gem_object_pin_map(ctx->obj, I915_MAP_WC);
 	if (IS_ERR(map))
 		return PTR_ERR(map);
 
 	map[offset / sizeof(*map)] = v;
-	i915_gem_object_unpin_map(obj);
+	i915_gem_object_unpin_map(ctx->obj);
 
 	return 0;
 }
 
-static int wc_get(struct drm_i915_gem_object *obj,
-		  unsigned long offset,
-		  u32 *v)
+static int wc_get(struct context *ctx, unsigned long offset, u32 *v)
 {
 	u32 *map;
 	int err;
 
-	i915_gem_object_lock(obj);
-	err = i915_gem_object_set_to_wc_domain(obj, false);
-	i915_gem_object_unlock(obj);
+	i915_gem_object_lock(ctx->obj);
+	err = i915_gem_object_set_to_wc_domain(ctx->obj, false);
+	i915_gem_object_unlock(ctx->obj);
 	if (err)
 		return err;
 
-	map = i915_gem_object_pin_map(obj, I915_MAP_WC);
+	map = i915_gem_object_pin_map(ctx->obj, I915_MAP_WC);
 	if (IS_ERR(map))
 		return PTR_ERR(map);
 
 	*v = map[offset / sizeof(*map)];
-	i915_gem_object_unpin_map(obj);
+	i915_gem_object_unpin_map(ctx->obj);
 
 	return 0;
 }
 
-static int gpu_set(struct drm_i915_gem_object *obj,
-		   unsigned long offset,
-		   u32 v)
+static int gpu_set(struct context *ctx, unsigned long offset, u32 v)
 {
-	struct drm_i915_private *i915 = to_i915(obj->base.dev);
 	struct i915_request *rq;
 	struct i915_vma *vma;
 	u32 *cs;
 	int err;
 
-	i915_gem_object_lock(obj);
-	err = i915_gem_object_set_to_gtt_domain(obj, true);
-	i915_gem_object_unlock(obj);
+	i915_gem_object_lock(ctx->obj);
+	err = i915_gem_object_set_to_gtt_domain(ctx->obj, true);
+	i915_gem_object_unlock(ctx->obj);
 	if (err)
 		return err;
 
-	vma = i915_gem_object_ggtt_pin(obj, NULL, 0, 0, 0);
+	vma = i915_gem_object_ggtt_pin(ctx->obj, NULL, 0, 0, 0);
 	if (IS_ERR(vma))
 		return PTR_ERR(vma);
 
-	rq = i915_request_create(i915->engine[RCS0]->kernel_context);
+	rq = i915_request_create(ctx->engine->kernel_context);
 	if (IS_ERR(rq)) {
 		i915_vma_unpin(vma);
 		return PTR_ERR(rq);
@@ -209,12 +213,12 @@ static int gpu_set(struct drm_i915_gem_object *obj,
 		return PTR_ERR(cs);
 	}
 
-	if (INTEL_GEN(i915) >= 8) {
+	if (INTEL_GEN(ctx->engine->i915) >= 8) {
 		*cs++ = MI_STORE_DWORD_IMM_GEN4 | 1 << 22;
 		*cs++ = lower_32_bits(i915_ggtt_offset(vma) + offset);
 		*cs++ = upper_32_bits(i915_ggtt_offset(vma) + offset);
 		*cs++ = v;
-	} else if (INTEL_GEN(i915) >= 4) {
+	} else if (INTEL_GEN(ctx->engine->i915) >= 4) {
 		*cs++ = MI_STORE_DWORD_IMM_GEN4 | MI_USE_GGTT;
 		*cs++ = 0;
 		*cs++ = i915_ggtt_offset(vma) + offset;
@@ -239,32 +243,34 @@ static int gpu_set(struct drm_i915_gem_object *obj,
 	return err;
 }
 
-static bool always_valid(struct drm_i915_private *i915)
+static bool always_valid(struct context *ctx)
 {
 	return true;
 }
 
-static bool needs_fence_registers(struct drm_i915_private *i915)
+static bool needs_fence_registers(struct context *ctx)
 {
-	return !intel_gt_is_wedged(&i915->gt);
-}
+	struct intel_gt *gt = ctx->engine->gt;
 
-static bool needs_mi_store_dword(struct drm_i915_private *i915)
-{
-	if (intel_gt_is_wedged(&i915->gt))
+	if (intel_gt_is_wedged(gt))
 		return false;
 
-	if (!HAS_ENGINE(i915, RCS0))
+	return gt->ggtt->num_fences;
+}
+
+static bool needs_mi_store_dword(struct context *ctx)
+{
+	if (intel_gt_is_wedged(ctx->engine->gt))
 		return false;
 
-	return intel_engine_can_store_dword(i915->engine[RCS0]);
+	return intel_engine_can_store_dword(ctx->engine);
 }
 
 static const struct igt_coherency_mode {
 	const char *name;
-	int (*set)(struct drm_i915_gem_object *, unsigned long offset, u32 v);
-	int (*get)(struct drm_i915_gem_object *, unsigned long offset, u32 *v);
-	bool (*valid)(struct drm_i915_private *i915);
+	int (*set)(struct context *ctx, unsigned long offset, u32 v);
+	int (*get)(struct context *ctx, unsigned long offset, u32 *v);
+	bool (*valid)(struct context *ctx);
 } igt_coherency_mode[] = {
 	{ "cpu", cpu_set, cpu_get, always_valid },
 	{ "gtt", gtt_set, gtt_get, needs_fence_registers },
@@ -273,19 +279,37 @@ static const struct igt_coherency_mode {
 	{ },
 };
 
+static struct intel_engine_cs *
+random_engine(struct drm_i915_private *i915, struct rnd_state *prng)
+{
+	struct intel_engine_cs *engine;
+	unsigned int count;
+
+	count = 0;
+	for_each_uabi_engine(engine, i915)
+		count++;
+
+	count = i915_prandom_u32_max_state(count, prng);
+	for_each_uabi_engine(engine, i915)
+		if (count-- == 0)
+			return engine;
+
+	return NULL;
+}
+
 static int igt_gem_coherency(void *arg)
 {
 	const unsigned int ncachelines = PAGE_SIZE/64;
-	I915_RND_STATE(prng);
 	struct drm_i915_private *i915 = arg;
 	const struct igt_coherency_mode *read, *write, *over;
-	struct drm_i915_gem_object *obj;
-	intel_wakeref_t wakeref;
 	unsigned long count, n;
 	u32 *offsets, *values;
+	I915_RND_STATE(prng);
+	struct context ctx;
 	int err = 0;
 
-	/* We repeatedly write, overwrite and read from a sequence of
+	/*
+	 * We repeatedly write, overwrite and read from a sequence of
 	 * cachelines in order to try and detect incoherency (unflushed writes
 	 * from either the CPU or GPU). Each setter/getter uses our cache
 	 * domain API which should prevent incoherency.
@@ -299,34 +323,36 @@ static int igt_gem_coherency(void *arg)
 
 	values = offsets + ncachelines;
 
-	mutex_lock(&i915->drm.struct_mutex);
-	wakeref = intel_runtime_pm_get(&i915->runtime_pm);
+	ctx.engine = random_engine(i915, &prng);
+	GEM_BUG_ON(!ctx.engine);
+	pr_info("%s: using %s\n", __func__, ctx.engine->name);
+
 	for (over = igt_coherency_mode; over->name; over++) {
 		if (!over->set)
 			continue;
 
-		if (!over->valid(i915))
+		if (!over->valid(&ctx))
 			continue;
 
 		for (write = igt_coherency_mode; write->name; write++) {
 			if (!write->set)
 				continue;
 
-			if (!write->valid(i915))
+			if (!write->valid(&ctx))
 				continue;
 
 			for (read = igt_coherency_mode; read->name; read++) {
 				if (!read->get)
 					continue;
 
-				if (!read->valid(i915))
+				if (!read->valid(&ctx))
 					continue;
 
 				for_each_prime_number_from(count, 1, ncachelines) {
-					obj = i915_gem_object_create_internal(i915, PAGE_SIZE);
-					if (IS_ERR(obj)) {
-						err = PTR_ERR(obj);
-						goto unlock;
+					ctx.obj = i915_gem_object_create_internal(i915, PAGE_SIZE);
+					if (IS_ERR(ctx.obj)) {
+						err = PTR_ERR(ctx.obj);
+						goto free;
 					}
 
 					i915_random_reorder(offsets, ncachelines, &prng);
@@ -334,7 +360,7 @@ static int igt_gem_coherency(void *arg)
 						values[n] = prandom_u32_state(&prng);
 
 					for (n = 0; n < count; n++) {
-						err = over->set(obj, offsets[n], ~values[n]);
+						err = over->set(&ctx, offsets[n], ~values[n]);
 						if (err) {
 							pr_err("Failed to set stale value[%ld/%ld] in object using %s, err=%d\n",
 							       n, count, over->name, err);
@@ -343,7 +369,7 @@ static int igt_gem_coherency(void *arg)
 					}
 
 					for (n = 0; n < count; n++) {
-						err = write->set(obj, offsets[n], values[n]);
+						err = write->set(&ctx, offsets[n], values[n]);
 						if (err) {
 							pr_err("Failed to set value[%ld/%ld] in object using %s, err=%d\n",
 							       n, count, write->name, err);
@@ -354,7 +380,7 @@ static int igt_gem_coherency(void *arg)
 					for (n = 0; n < count; n++) {
 						u32 found;
 
-						err = read->get(obj, offsets[n], &found);
+						err = read->get(&ctx, offsets[n], &found);
 						if (err) {
 							pr_err("Failed to get value[%ld/%ld] in object using %s, err=%d\n",
 							       n, count, read->name, err);
@@ -372,20 +398,18 @@ static int igt_gem_coherency(void *arg)
 						}
 					}
 
-					i915_gem_object_put(obj);
+					i915_gem_object_put(ctx.obj);
 				}
 			}
 		}
 	}
-unlock:
-	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
-	mutex_unlock(&i915->drm.struct_mutex);
+free:
 	kfree(offsets);
 	return err;
 
 put_object:
-	i915_gem_object_put(obj);
-	goto unlock;
+	i915_gem_object_put(ctx.obj);
+	goto free;
 }
 
 int i915_gem_coherency_live_selftests(struct drm_i915_private *i915)
diff --git a/drivers/gpu/drm/i915/gem/selftests/i915_gem_context.c b/drivers/gpu/drm/i915/gem/selftests/i915_gem_context.c
index 3e6f4a65d356..62fabc023a83 100644
--- a/drivers/gpu/drm/i915/gem/selftests/i915_gem_context.c
+++ b/drivers/gpu/drm/i915/gem/selftests/i915_gem_context.c
@@ -8,6 +8,7 @@
 
 #include "gem/i915_gem_pm.h"
 #include "gt/intel_gt.h"
+#include "gt/intel_gt_requests.h"
 #include "gt/intel_reset.h"
 #include "i915_selftest.h"
 
@@ -31,7 +32,6 @@ static int live_nop_switch(void *arg)
 	struct drm_i915_private *i915 = arg;
 	struct intel_engine_cs *engine;
 	struct i915_gem_context **ctx;
-	enum intel_engine_id id;
 	struct igt_live_test t;
 	struct drm_file *file;
 	unsigned long n;
@@ -52,23 +52,21 @@ static int live_nop_switch(void *arg)
 	if (IS_ERR(file))
 		return PTR_ERR(file);
 
-	mutex_lock(&i915->drm.struct_mutex);
-
 	ctx = kcalloc(nctx, sizeof(*ctx), GFP_KERNEL);
 	if (!ctx) {
 		err = -ENOMEM;
-		goto out_unlock;
+		goto out_file;
 	}
 
 	for (n = 0; n < nctx; n++) {
 		ctx[n] = live_context(i915, file);
 		if (IS_ERR(ctx[n])) {
 			err = PTR_ERR(ctx[n]);
-			goto out_unlock;
+			goto out_file;
 		}
 	}
 
-	for_each_engine(engine, i915, id) {
+	for_each_uabi_engine(engine, i915) {
 		struct i915_request *rq;
 		unsigned long end_time, prime;
 		ktime_t times[2] = {};
@@ -78,7 +76,7 @@ static int live_nop_switch(void *arg)
 			rq = igt_request_alloc(ctx[n], engine);
 			if (IS_ERR(rq)) {
 				err = PTR_ERR(rq);
-				goto out_unlock;
+				goto out_file;
 			}
 			i915_request_add(rq);
 		}
@@ -86,7 +84,7 @@ static int live_nop_switch(void *arg)
 			pr_err("Failed to populated %d contexts\n", nctx);
 			intel_gt_set_wedged(&i915->gt);
 			err = -EIO;
-			goto out_unlock;
+			goto out_file;
 		}
 
 		times[1] = ktime_get_raw();
@@ -96,7 +94,7 @@ static int live_nop_switch(void *arg)
 
 		err = igt_live_test_begin(&t, i915, __func__, engine->name);
 		if (err)
-			goto out_unlock;
+			goto out_file;
 
 		end_time = jiffies + i915_selftest.timeout_jiffies;
 		for_each_prime_number_from(prime, 2, 8192) {
@@ -106,7 +104,7 @@ static int live_nop_switch(void *arg)
 				rq = igt_request_alloc(ctx[n % nctx], engine);
 				if (IS_ERR(rq)) {
 					err = PTR_ERR(rq);
-					goto out_unlock;
+					goto out_file;
 				}
 
 				/*
@@ -142,7 +140,7 @@ static int live_nop_switch(void *arg)
 
 		err = igt_live_test_end(&t);
 		if (err)
-			goto out_unlock;
+			goto out_file;
 
 		pr_info("Switch latencies on %s: 1 = %lluns, %lu = %lluns\n",
 			engine->name,
@@ -150,8 +148,235 @@ static int live_nop_switch(void *arg)
 			prime - 1, div64_u64(ktime_to_ns(times[1]), prime - 1));
 	}
 
-out_unlock:
-	mutex_unlock(&i915->drm.struct_mutex);
+out_file:
+	mock_file_free(i915, file);
+	return err;
+}
+
+struct parallel_switch {
+	struct task_struct *tsk;
+	struct intel_context *ce[2];
+};
+
+static int __live_parallel_switch1(void *data)
+{
+	struct parallel_switch *arg = data;
+	IGT_TIMEOUT(end_time);
+	unsigned long count;
+
+	count = 0;
+	do {
+		struct i915_request *rq = NULL;
+		int err, n;
+
+		err = 0;
+		for (n = 0; !err && n < ARRAY_SIZE(arg->ce); n++) {
+			struct i915_request *prev = rq;
+
+			rq = i915_request_create(arg->ce[n]);
+			if (IS_ERR(rq)) {
+				i915_request_put(prev);
+				return PTR_ERR(rq);
+			}
+
+			i915_request_get(rq);
+			if (prev) {
+				err = i915_request_await_dma_fence(rq, &prev->fence);
+				i915_request_put(prev);
+			}
+
+			i915_request_add(rq);
+		}
+		if (i915_request_wait(rq, 0, HZ / 5) < 0)
+			err = -ETIME;
+		i915_request_put(rq);
+		if (err)
+			return err;
+
+		count++;
+	} while (!__igt_timeout(end_time, NULL));
+
+	pr_info("%s: %lu switches (sync)\n", arg->ce[0]->engine->name, count);
+	return 0;
+}
+
+static int __live_parallel_switchN(void *data)
+{
+	struct parallel_switch *arg = data;
+	struct i915_request *rq = NULL;
+	IGT_TIMEOUT(end_time);
+	unsigned long count;
+	int n;
+
+	count = 0;
+	do {
+		for (n = 0; n < ARRAY_SIZE(arg->ce); n++) {
+			struct i915_request *prev = rq;
+			int err = 0;
+
+			rq = i915_request_create(arg->ce[n]);
+			if (IS_ERR(rq)) {
+				i915_request_put(prev);
+				return PTR_ERR(rq);
+			}
+
+			i915_request_get(rq);
+			if (prev) {
+				err = i915_request_await_dma_fence(rq, &prev->fence);
+				i915_request_put(prev);
+			}
+
+			i915_request_add(rq);
+			if (err) {
+				i915_request_put(rq);
+				return err;
+			}
+		}
+
+		count++;
+	} while (!__igt_timeout(end_time, NULL));
+	i915_request_put(rq);
+
+	pr_info("%s: %lu switches (many)\n", arg->ce[0]->engine->name, count);
+	return 0;
+}
+
+static int live_parallel_switch(void *arg)
+{
+	struct drm_i915_private *i915 = arg;
+	static int (* const func[])(void *arg) = {
+		__live_parallel_switch1,
+		__live_parallel_switchN,
+		NULL,
+	};
+	struct parallel_switch *data = NULL;
+	struct i915_gem_engines *engines;
+	struct i915_gem_engines_iter it;
+	int (* const *fn)(void *arg);
+	struct i915_gem_context *ctx;
+	struct intel_context *ce;
+	struct drm_file *file;
+	int n, m, count;
+	int err = 0;
+
+	/*
+	 * Check we can process switches on all engines simultaneously.
+	 */
+
+	if (!DRIVER_CAPS(i915)->has_logical_contexts)
+		return 0;
+
+	file = mock_file(i915);
+	if (IS_ERR(file))
+		return PTR_ERR(file);
+
+	ctx = live_context(i915, file);
+	if (IS_ERR(ctx)) {
+		err = PTR_ERR(ctx);
+		goto out_file;
+	}
+
+	engines = i915_gem_context_lock_engines(ctx);
+	count = engines->num_engines;
+
+	data = kcalloc(count, sizeof(*data), GFP_KERNEL);
+	if (!data) {
+		i915_gem_context_unlock_engines(ctx);
+		err = -ENOMEM;
+		goto out_file;
+	}
+
+	m = 0; /* Use the first context as our template for the engines */
+	for_each_gem_engine(ce, engines, it) {
+		err = intel_context_pin(ce);
+		if (err) {
+			i915_gem_context_unlock_engines(ctx);
+			goto out;
+		}
+		data[m++].ce[0] = intel_context_get(ce);
+	}
+	i915_gem_context_unlock_engines(ctx);
+
+	/* Clone the same set of engines into the other contexts */
+	for (n = 1; n < ARRAY_SIZE(data->ce); n++) {
+		ctx = live_context(i915, file);
+		if (IS_ERR(ctx)) {
+			err = PTR_ERR(ctx);
+			goto out;
+		}
+
+		for (m = 0; m < count; m++) {
+			if (!data[m].ce[0])
+				continue;
+
+			ce = intel_context_create(ctx, data[m].ce[0]->engine);
+			if (IS_ERR(ce))
+				goto out;
+
+			err = intel_context_pin(ce);
+			if (err) {
+				intel_context_put(ce);
+				goto out;
+			}
+
+			data[m].ce[n] = ce;
+		}
+	}
+
+	for (fn = func; !err && *fn; fn++) {
+		struct igt_live_test t;
+		int n;
+
+		err = igt_live_test_begin(&t, i915, __func__, "");
+		if (err)
+			break;
+
+		for (n = 0; n < count; n++) {
+			if (!data[n].ce[0])
+				continue;
+
+			data[n].tsk = kthread_run(*fn, &data[n],
+						  "igt/parallel:%s",
+						  data[n].ce[0]->engine->name);
+			if (IS_ERR(data[n].tsk)) {
+				err = PTR_ERR(data[n].tsk);
+				break;
+			}
+			get_task_struct(data[n].tsk);
+		}
+
+		yield(); /* start all threads before we kthread_stop() */
+
+		for (n = 0; n < count; n++) {
+			int status;
+
+			if (IS_ERR_OR_NULL(data[n].tsk))
+				continue;
+
+			status = kthread_stop(data[n].tsk);
+			if (status && !err)
+				err = status;
+
+			put_task_struct(data[n].tsk);
+			data[n].tsk = NULL;
+		}
+
+		if (igt_live_test_end(&t))
+			err = -EIO;
+	}
+
+out:
+	for (n = 0; n < count; n++) {
+		for (m = 0; m < ARRAY_SIZE(data->ce); m++) {
+			if (!data[n].ce[m])
+				continue;
+
+			intel_context_unpin(data[n].ce[m]);
+			intel_context_put(data[n].ce[m]);
+		}
+	}
+	kfree(data);
+out_file:
 	mock_file_free(i915, file);
 	return err;
 }
@@ -166,28 +391,20 @@ static unsigned long fake_page_count(struct drm_i915_gem_object *obj)
 	return huge_gem_object_dma_size(obj) >> PAGE_SHIFT;
 }
 
-static int gpu_fill(struct drm_i915_gem_object *obj,
-		    struct i915_gem_context *ctx,
-		    struct intel_engine_cs *engine,
+static int gpu_fill(struct intel_context *ce,
+		    struct drm_i915_gem_object *obj,
 		    unsigned int dw)
 {
-	struct i915_address_space *vm = ctx->vm ?: &engine->gt->ggtt->vm;
 	struct i915_vma *vma;
 	int err;
 
-	GEM_BUG_ON(obj->base.size > vm->total);
-	GEM_BUG_ON(!intel_engine_can_store_dword(engine));
+	GEM_BUG_ON(obj->base.size > ce->vm->total);
+	GEM_BUG_ON(!intel_engine_can_store_dword(ce->engine));
 
-	vma = i915_vma_instance(obj, vm, NULL);
+	vma = i915_vma_instance(obj, ce->vm, NULL);
 	if (IS_ERR(vma))
 		return PTR_ERR(vma);
 
-	i915_gem_object_lock(obj);
-	err = i915_gem_object_set_to_gtt_domain(obj, true);
-	i915_gem_object_unlock(obj);
-	if (err)
-		return err;
-
 	err = i915_vma_pin(vma, 0, 0, PIN_HIGH | PIN_USER);
 	if (err)
 		return err;
@@ -200,9 +417,7 @@ static int gpu_fill(struct drm_i915_gem_object *obj,
 	 * whilst checking that each context provides a unique view
 	 * into the object.
 	 */
-	err = igt_gpu_fill_dw(vma,
-			      ctx,
-			      engine,
+	err = igt_gpu_fill_dw(ce, vma,
 			      (dw * real_page_count(obj)) << PAGE_SHIFT |
 			      (dw * sizeof(u32)),
 			      real_page_count(obj),
@@ -305,22 +520,21 @@ static int file_add_object(struct drm_file *file,
 }
 
 static struct drm_i915_gem_object *
-create_test_object(struct i915_gem_context *ctx,
+create_test_object(struct i915_address_space *vm,
 		   struct drm_file *file,
 		   struct list_head *objects)
 {
 	struct drm_i915_gem_object *obj;
-	struct i915_address_space *vm = ctx->vm ?: &ctx->i915->ggtt.vm;
 	u64 size;
 	int err;
 
 	/* Keep in GEM's good graces */
-	i915_retire_requests(ctx->i915);
+	intel_gt_retire_requests(vm->gt);
 
 	size = min(vm->total / 2, 1024ull * DW_PER_PAGE * PAGE_SIZE);
 	size = round_down(size, DW_PER_PAGE * PAGE_SIZE);
 
-	obj = huge_gem_object(ctx->i915, DW_PER_PAGE * PAGE_SIZE, size);
+	obj = huge_gem_object(vm->i915, DW_PER_PAGE * PAGE_SIZE, size);
 	if (IS_ERR(obj))
 		return obj;
 
@@ -348,11 +562,49 @@ static unsigned long max_dwords(struct drm_i915_gem_object *obj)
 	return npages / DW_PER_PAGE;
 }
 
+static void throttle_release(struct i915_request **q, int count)
+{
+	int i;
+
+	for (i = 0; i < count; i++) {
+		if (IS_ERR_OR_NULL(q[i]))
+			continue;
+
+		i915_request_put(fetch_and_zero(&q[i]));
+	}
+}
+
+static int throttle(struct intel_context *ce,
+		    struct i915_request **q, int count)
+{
+	int i;
+
+	if (!IS_ERR_OR_NULL(q[0])) {
+		if (i915_request_wait(q[0],
+				      I915_WAIT_INTERRUPTIBLE,
+				      MAX_SCHEDULE_TIMEOUT) < 0)
+			return -EINTR;
+
+		i915_request_put(q[0]);
+	}
+
+	for (i = 0; i < count - 1; i++)
+		q[i] = q[i + 1];
+
+	q[i] = intel_context_create_request(ce);
+	if (IS_ERR(q[i]))
+		return PTR_ERR(q[i]);
+
+	i915_request_get(q[i]);
+	i915_request_add(q[i]);
+
+	return 0;
+}
+
 static int igt_ctx_exec(void *arg)
 {
 	struct drm_i915_private *i915 = arg;
 	struct intel_engine_cs *engine;
-	enum intel_engine_id id;
 	int err = -ENODEV;
 
 	/*
@@ -364,9 +616,10 @@ static int igt_ctx_exec(void *arg)
 	if (!DRIVER_CAPS(i915)->has_logical_contexts)
 		return 0;
 
-	for_each_engine(engine, i915, id) {
+	for_each_uabi_engine(engine, i915) {
 		struct drm_i915_gem_object *obj = NULL;
 		unsigned long ncontexts, ndwords, dw;
+		struct i915_request *tq[5] = {};
 		struct igt_live_test t;
 		struct drm_file *file;
 		IGT_TIMEOUT(end_time);
@@ -382,39 +635,53 @@ static int igt_ctx_exec(void *arg)
 		if (IS_ERR(file))
 			return PTR_ERR(file);
 
-		mutex_lock(&i915->drm.struct_mutex);
-
 		err = igt_live_test_begin(&t, i915, __func__, engine->name);
 		if (err)
-			goto out_unlock;
+			goto out_file;
 
 		ncontexts = 0;
 		ndwords = 0;
 		dw = 0;
 		while (!time_after(jiffies, end_time)) {
 			struct i915_gem_context *ctx;
+			struct intel_context *ce;
 
-			ctx = live_context(i915, file);
+			ctx = kernel_context(i915);
 			if (IS_ERR(ctx)) {
 				err = PTR_ERR(ctx);
-				goto out_unlock;
+				goto out_file;
 			}
 
+			ce = i915_gem_context_get_engine(ctx, engine->legacy_idx);
+			GEM_BUG_ON(IS_ERR(ce));
+
 			if (!obj) {
-				obj = create_test_object(ctx, file, &objects);
+				obj = create_test_object(ce->vm, file, &objects);
 				if (IS_ERR(obj)) {
 					err = PTR_ERR(obj);
-					goto out_unlock;
+					intel_context_put(ce);
+					kernel_context_close(ctx);
+					goto out_file;
 				}
 			}
 
-			err = gpu_fill(obj, ctx, engine, dw);
+			err = gpu_fill(ce, obj, dw);
 			if (err) {
-				pr_err("Failed to fill dword %lu [%lu/%lu] with gpu (%s) in ctx %u [full-ppgtt? %s], err=%d\n",
+				pr_err("Failed to fill dword %lu [%lu/%lu] with gpu (%s) [full-ppgtt? %s], err=%d\n",
 				       ndwords, dw, max_dwords(obj),
-				       engine->name, ctx->hw_id,
-				       yesno(!!ctx->vm), err);
-				goto out_unlock;
+				       engine->name,
+				       yesno(!!rcu_access_pointer(ctx->vm)),
+				       err);
+				intel_context_put(ce);
+				kernel_context_close(ctx);
+				goto out_file;
+			}
+
+			err = throttle(ce, tq, ARRAY_SIZE(tq));
+			if (err) {
+				intel_context_put(ce);
+				kernel_context_close(ctx);
+				goto out_file;
 			}
 
 			if (++dw == max_dwords(obj)) {
@@ -424,6 +691,9 @@ static int igt_ctx_exec(void *arg)
 
 			ndwords++;
 			ncontexts++;
+
+			intel_context_put(ce);
+			kernel_context_close(ctx);
 		}
 
 		pr_info("Submitted %lu contexts to %s, filling %lu dwords\n",
@@ -441,10 +711,10 @@ static int igt_ctx_exec(void *arg)
 			dw += rem;
 		}
 
-out_unlock:
+out_file:
+		throttle_release(tq, ARRAY_SIZE(tq));
 		if (igt_live_test_end(&t))
 			err = -EIO;
-		mutex_unlock(&i915->drm.struct_mutex);
 
 		mock_file_free(i915, file);
 		if (err)
@@ -459,9 +729,9 @@ out_unlock:
 static int igt_shared_ctx_exec(void *arg)
 {
 	struct drm_i915_private *i915 = arg;
+	struct i915_request *tq[5] = {};
 	struct i915_gem_context *parent;
 	struct intel_engine_cs *engine;
-	enum intel_engine_id id;
 	struct igt_live_test t;
 	struct drm_file *file;
 	int err = 0;
@@ -478,24 +748,22 @@ static int igt_shared_ctx_exec(void *arg)
 	if (IS_ERR(file))
 		return PTR_ERR(file);
 
-	mutex_lock(&i915->drm.struct_mutex);
-
 	parent = live_context(i915, file);
 	if (IS_ERR(parent)) {
 		err = PTR_ERR(parent);
-		goto out_unlock;
+		goto out_file;
 	}
 
 	if (!parent->vm) { /* not full-ppgtt; nothing to share */
 		err = 0;
-		goto out_unlock;
+		goto out_file;
 	}
 
 	err = igt_live_test_begin(&t, i915, __func__, "");
 	if (err)
-		goto out_unlock;
+		goto out_file;
 
-	for_each_engine(engine, i915, id) {
+	for_each_uabi_engine(engine, i915) {
 		unsigned long ncontexts, ndwords, dw;
 		struct drm_i915_gem_object *obj = NULL;
 		IGT_TIMEOUT(end_time);
@@ -509,6 +777,7 @@ static int igt_shared_ctx_exec(void *arg)
 		ncontexts = 0;
 		while (!time_after(jiffies, end_time)) {
 			struct i915_gem_context *ctx;
+			struct intel_context *ce;
 
 			ctx = kernel_context(i915);
 			if (IS_ERR(ctx)) {
@@ -516,23 +785,38 @@ static int igt_shared_ctx_exec(void *arg)
 				goto out_test;
 			}
 
+			mutex_lock(&ctx->mutex);
 			__assign_ppgtt(ctx, parent->vm);
+			mutex_unlock(&ctx->mutex);
+
+			ce = i915_gem_context_get_engine(ctx, engine->legacy_idx);
+			GEM_BUG_ON(IS_ERR(ce));
 
 			if (!obj) {
-				obj = create_test_object(parent, file, &objects);
+				obj = create_test_object(parent->vm, file, &objects);
 				if (IS_ERR(obj)) {
 					err = PTR_ERR(obj);
+					intel_context_put(ce);
 					kernel_context_close(ctx);
 					goto out_test;
 				}
 			}
 
-			err = gpu_fill(obj, ctx, engine, dw);
+			err = gpu_fill(ce, obj, dw);
 			if (err) {
-				pr_err("Failed to fill dword %lu [%lu/%lu] with gpu (%s) in ctx %u [full-ppgtt? %s], err=%d\n",
+				pr_err("Failed to fill dword %lu [%lu/%lu] with gpu (%s) [full-ppgtt? %s], err=%d\n",
 				       ndwords, dw, max_dwords(obj),
-				       engine->name, ctx->hw_id,
-				       yesno(!!ctx->vm), err);
+				       engine->name,
+				       yesno(!!rcu_access_pointer(ctx->vm)),
+				       err);
+				intel_context_put(ce);
+				kernel_context_close(ctx);
+				goto out_test;
+			}
+
+			err = throttle(ce, tq, ARRAY_SIZE(tq));
+			if (err) {
+				intel_context_put(ce);
 				kernel_context_close(ctx);
 				goto out_test;
 			}
@@ -545,6 +829,7 @@ static int igt_shared_ctx_exec(void *arg)
 			ndwords++;
 			ncontexts++;
 
+			intel_context_put(ce);
 			kernel_context_close(ctx);
 		}
 		pr_info("Submitted %lu contexts to %s, filling %lu dwords\n",
@@ -562,16 +847,13 @@ static int igt_shared_ctx_exec(void *arg)
 			dw += rem;
 		}
 
-		mutex_unlock(&i915->drm.struct_mutex);
 		i915_gem_drain_freed_objects(i915);
-		mutex_lock(&i915->drm.struct_mutex);
 	}
 out_test:
+	throttle_release(tq, ARRAY_SIZE(tq));
 	if (igt_live_test_end(&t))
 		err = -EIO;
-out_unlock:
-	mutex_unlock(&i915->drm.struct_mutex);
-
+out_file:
 	mock_file_free(i915, file);
 	return err;
 }
@@ -604,6 +886,8 @@ static struct i915_vma *rpcs_query_batch(struct i915_vma *vma)
 	__i915_gem_object_flush_map(obj, 0, 64);
 	i915_gem_object_unpin_map(obj);
 
+	intel_gt_chipset_flush(vma->vm->gt);
+
 	vma = i915_vma_instance(obj, vma->vm, NULL);
 	if (IS_ERR(vma)) {
 		err = PTR_ERR(vma);
@@ -681,10 +965,7 @@ emit_rpcs_query(struct drm_i915_gem_object *obj,
 	if (err)
 		goto skip_request;
 
-	i915_vma_unpin(batch);
-	i915_vma_close(batch);
-	i915_vma_put(batch);
-
+	i915_vma_unpin_and_release(&batch, 0);
 	i915_vma_unpin(vma);
 
 	*rq_out = i915_request_get(rq);
@@ -698,8 +979,7 @@ skip_request:
 err_request:
 	i915_request_add(rq);
 err_batch:
-	i915_vma_unpin(batch);
-	i915_vma_put(batch);
+	i915_vma_unpin_and_release(&batch, 0);
 err_vma:
 	i915_vma_unpin(vma);
 
@@ -860,8 +1140,8 @@ out:
 		igt_spinner_end(spin);
 
 	if ((flags & TEST_IDLE) && ret == 0) {
-		ret = i915_gem_wait_for_idle(ce->engine->i915,
-					     0, MAX_SCHEDULE_TIMEOUT);
+		ret = intel_gt_wait_for_idle(ce->engine->gt,
+					     MAX_SCHEDULE_TIMEOUT);
 		if (ret)
 			return ret;
 
@@ -887,7 +1167,7 @@ __sseu_test(const char *name,
 	if (ret)
 		return ret;
 
-	ret = __intel_context_reconfigure_sseu(ce, sseu);
+	ret = intel_context_reconfigure_sseu(ce, sseu);
 	if (ret)
 		goto out_spin;
 
@@ -908,106 +1188,97 @@ __igt_ctx_sseu(struct drm_i915_private *i915,
 	       const char *name,
 	       unsigned int flags)
 {
-	struct intel_engine_cs *engine = i915->engine[RCS0];
 	struct drm_i915_gem_object *obj;
-	struct i915_gem_context *ctx;
-	struct intel_context *ce;
-	struct intel_sseu pg_sseu;
-	struct drm_file *file;
-	int ret;
-
-	if (INTEL_GEN(i915) < 9 || !engine)
-		return 0;
-
-	if (!RUNTIME_INFO(i915)->sseu.has_slice_pg)
-		return 0;
+	int inst = 0;
+	int ret = 0;
 
-	if (hweight32(engine->sseu.slice_mask) < 2)
+	if (INTEL_GEN(i915) < 9 || !RUNTIME_INFO(i915)->sseu.has_slice_pg)
 		return 0;
 
-	/*
-	 * Gen11 VME friendly power-gated configuration with half enabled
-	 * sub-slices.
-	 */
-	pg_sseu = engine->sseu;
-	pg_sseu.slice_mask = 1;
-	pg_sseu.subslice_mask =
-		~(~0 << (hweight32(engine->sseu.subslice_mask) / 2));
-
-	pr_info("SSEU subtest '%s', flags=%x, def_slices=%u, pg_slices=%u\n",
-		name, flags, hweight32(engine->sseu.slice_mask),
-		hweight32(pg_sseu.slice_mask));
-
-	file = mock_file(i915);
-	if (IS_ERR(file))
-		return PTR_ERR(file);
-
 	if (flags & TEST_RESET)
 		igt_global_reset_lock(&i915->gt);
 
-	mutex_lock(&i915->drm.struct_mutex);
-
-	ctx = live_context(i915, file);
-	if (IS_ERR(ctx)) {
-		ret = PTR_ERR(ctx);
-		goto out_unlock;
-	}
-	i915_gem_context_clear_bannable(ctx); /* to reset and beyond! */
-
 	obj = i915_gem_object_create_internal(i915, PAGE_SIZE);
 	if (IS_ERR(obj)) {
 		ret = PTR_ERR(obj);
 		goto out_unlock;
 	}
 
-	ce = i915_gem_context_get_engine(ctx, RCS0);
-	if (IS_ERR(ce)) {
-		ret = PTR_ERR(ce);
-		goto out_put;
-	}
+	do {
+		struct intel_engine_cs *engine;
+		struct intel_context *ce;
+		struct intel_sseu pg_sseu;
 
-	ret = intel_context_pin(ce);
-	if (ret)
-		goto out_context;
+		engine = intel_engine_lookup_user(i915,
+						  I915_ENGINE_CLASS_RENDER,
+						  inst++);
+		if (!engine)
+			break;
 
-	/* First set the default mask. */
-	ret = __sseu_test(name, flags, ce, obj, engine->sseu);
-	if (ret)
-		goto out_fail;
+		if (hweight32(engine->sseu.slice_mask) < 2)
+			continue;
 
-	/* Then set a power-gated configuration. */
-	ret = __sseu_test(name, flags, ce, obj, pg_sseu);
-	if (ret)
-		goto out_fail;
+		/*
+		 * Gen11 VME friendly power-gated configuration with
+		 * half enabled sub-slices.
+		 */
+		pg_sseu = engine->sseu;
+		pg_sseu.slice_mask = 1;
+		pg_sseu.subslice_mask =
+			~(~0 << (hweight32(engine->sseu.subslice_mask) / 2));
+
+		pr_info("%s: SSEU subtest '%s', flags=%x, def_slices=%u, pg_slices=%u\n",
+			engine->name, name, flags,
+			hweight32(engine->sseu.slice_mask),
+			hweight32(pg_sseu.slice_mask));
+
+		ce = intel_context_create(engine->kernel_context->gem_context,
+					  engine);
+		if (IS_ERR(ce)) {
+			ret = PTR_ERR(ce);
+			goto out_put;
+		}
 
-	/* Back to defaults. */
-	ret = __sseu_test(name, flags, ce, obj, engine->sseu);
-	if (ret)
-		goto out_fail;
+		ret = intel_context_pin(ce);
+		if (ret)
+			goto out_ce;
 
-	/* One last power-gated configuration for the road. */
-	ret = __sseu_test(name, flags, ce, obj, pg_sseu);
-	if (ret)
-		goto out_fail;
+		/* First set the default mask. */
+		ret = __sseu_test(name, flags, ce, obj, engine->sseu);
+		if (ret)
+			goto out_unpin;
+
+		/* Then set a power-gated configuration. */
+		ret = __sseu_test(name, flags, ce, obj, pg_sseu);
+		if (ret)
+			goto out_unpin;
+
+		/* Back to defaults. */
+		ret = __sseu_test(name, flags, ce, obj, engine->sseu);
+		if (ret)
+			goto out_unpin;
 
-out_fail:
-	if (igt_flush_test(i915, I915_WAIT_LOCKED))
+		/* One last power-gated configuration for the road. */
+		ret = __sseu_test(name, flags, ce, obj, pg_sseu);
+		if (ret)
+			goto out_unpin;
+
+out_unpin:
+		intel_context_unpin(ce);
+out_ce:
+		intel_context_put(ce);
+	} while (!ret);
+
+	if (igt_flush_test(i915))
 		ret = -EIO;
 
-	intel_context_unpin(ce);
-out_context:
-	intel_context_put(ce);
 out_put:
 	i915_gem_object_put(obj);
 
 out_unlock:
-	mutex_unlock(&i915->drm.struct_mutex);
-
 	if (flags & TEST_RESET)
 		igt_global_reset_unlock(&i915->gt);
 
-	mock_file_free(i915, file);
-
 	if (ret)
 		pr_err("%s: Failed with %d!\n", name, ret);
 
@@ -1041,6 +1312,7 @@ static int igt_ctx_readonly(void *arg)
 {
 	struct drm_i915_private *i915 = arg;
 	struct drm_i915_gem_object *obj = NULL;
+	struct i915_request *tq[5] = {};
 	struct i915_address_space *vm;
 	struct i915_gem_context *ctx;
 	unsigned long idx, ndwords, dw;
@@ -1061,52 +1333,63 @@ static int igt_ctx_readonly(void *arg)
 	if (IS_ERR(file))
 		return PTR_ERR(file);
 
-	mutex_lock(&i915->drm.struct_mutex);
-
 	err = igt_live_test_begin(&t, i915, __func__, "");
 	if (err)
-		goto out_unlock;
+		goto out_file;
 
 	ctx = live_context(i915, file);
 	if (IS_ERR(ctx)) {
 		err = PTR_ERR(ctx);
-		goto out_unlock;
+		goto out_file;
 	}
 
-	vm = ctx->vm ?: &i915->ggtt.alias->vm;
+	rcu_read_lock();
+	vm = rcu_dereference(ctx->vm) ?: &i915->ggtt.alias->vm;
 	if (!vm || !vm->has_read_only) {
+		rcu_read_unlock();
 		err = 0;
-		goto out_unlock;
+		goto out_file;
 	}
+	rcu_read_unlock();
 
 	ndwords = 0;
 	dw = 0;
 	while (!time_after(jiffies, end_time)) {
-		struct intel_engine_cs *engine;
-		unsigned int id;
+		struct i915_gem_engines_iter it;
+		struct intel_context *ce;
 
-		for_each_engine(engine, i915, id) {
-			if (!intel_engine_can_store_dword(engine))
+		for_each_gem_engine(ce,
+				    i915_gem_context_lock_engines(ctx), it) {
+			if (!intel_engine_can_store_dword(ce->engine))
 				continue;
 
 			if (!obj) {
-				obj = create_test_object(ctx, file, &objects);
+				obj = create_test_object(ce->vm, file, &objects);
 				if (IS_ERR(obj)) {
 					err = PTR_ERR(obj);
-					goto out_unlock;
+					i915_gem_context_unlock_engines(ctx);
+					goto out_file;
 				}
 
 				if (prandom_u32_state(&prng) & 1)
 					i915_gem_object_set_readonly(obj);
 			}
 
-			err = gpu_fill(obj, ctx, engine, dw);
+			err = gpu_fill(ce, obj, dw);
 			if (err) {
-				pr_err("Failed to fill dword %lu [%lu/%lu] with gpu (%s) in ctx %u [full-ppgtt? %s], err=%d\n",
+				pr_err("Failed to fill dword %lu [%lu/%lu] with gpu (%s) [full-ppgtt? %s], err=%d\n",
 				       ndwords, dw, max_dwords(obj),
-				       engine->name, ctx->hw_id,
-				       yesno(!!ctx->vm), err);
-				goto out_unlock;
+				       ce->engine->name,
+				       yesno(!!rcu_access_pointer(ctx->vm)),
+				       err);
+				i915_gem_context_unlock_engines(ctx);
+				goto out_file;
+			}
+
+			err = throttle(ce, tq, ARRAY_SIZE(tq));
+			if (err) {
+				i915_gem_context_unlock_engines(ctx);
+				goto out_file;
 			}
 
 			if (++dw == max_dwords(obj)) {
@@ -1115,6 +1398,7 @@ static int igt_ctx_readonly(void *arg)
 			}
 			ndwords++;
 		}
+		i915_gem_context_unlock_engines(ctx);
 	}
 	pr_info("Submitted %lu dwords (across %u engines)\n",
 		ndwords, RUNTIME_INFO(i915)->num_engines);
@@ -1137,19 +1421,19 @@ static int igt_ctx_readonly(void *arg)
 		dw += rem;
 	}
 
-out_unlock:
+out_file:
+	throttle_release(tq, ARRAY_SIZE(tq));
 	if (igt_live_test_end(&t))
 		err = -EIO;
-	mutex_unlock(&i915->drm.struct_mutex);
 
 	mock_file_free(i915, file);
 	return err;
 }
 
-static int check_scratch(struct i915_gem_context *ctx, u64 offset)
+static int check_scratch(struct i915_address_space *vm, u64 offset)
 {
 	struct drm_mm_node *node =
-		__drm_mm_interval_first(&ctx->vm->mm,
+		__drm_mm_interval_first(&vm->mm,
 					offset, offset + sizeof(u32) - 1);
 	if (!node || node->start > offset)
 		return 0;
@@ -1167,6 +1451,7 @@ static int write_to_scratch(struct i915_gem_context *ctx,
 {
 	struct drm_i915_private *i915 = ctx->i915;
 	struct drm_i915_gem_object *obj;
+	struct i915_address_space *vm;
 	struct i915_request *rq;
 	struct i915_vma *vma;
 	u32 *cmd;
@@ -1197,17 +1482,20 @@ static int write_to_scratch(struct i915_gem_context *ctx,
 	__i915_gem_object_flush_map(obj, 0, 64);
 	i915_gem_object_unpin_map(obj);
 
-	vma = i915_vma_instance(obj, ctx->vm, NULL);
+	intel_gt_chipset_flush(engine->gt);
+
+	vm = i915_gem_context_get_vm_rcu(ctx);
+	vma = i915_vma_instance(obj, vm, NULL);
 	if (IS_ERR(vma)) {
 		err = PTR_ERR(vma);
-		goto err;
+		goto err_vm;
 	}
 
 	err = i915_vma_pin(vma, 0, 0, PIN_USER | PIN_OFFSET_FIXED);
 	if (err)
-		goto err;
+		goto err_vm;
 
-	err = check_scratch(ctx, offset);
+	err = check_scratch(vm, offset);
 	if (err)
 		goto err_unpin;
 
@@ -1229,12 +1517,11 @@ static int write_to_scratch(struct i915_gem_context *ctx,
 	if (err)
 		goto skip_request;
 
-	i915_vma_unpin(vma);
-	i915_vma_close(vma);
-	i915_vma_put(vma);
+	i915_vma_unpin_and_release(&vma, 0);
 
 	i915_request_add(rq);
 
+	i915_vm_put(vm);
 	return 0;
 
 skip_request:
@@ -1243,6 +1530,8 @@ err_request:
 	i915_request_add(rq);
 err_unpin:
 	i915_vma_unpin(vma);
+err_vm:
+	i915_vm_put(vm);
 err:
 	i915_gem_object_put(obj);
 	return err;
@@ -1254,6 +1543,7 @@ static int read_from_scratch(struct i915_gem_context *ctx,
 {
 	struct drm_i915_private *i915 = ctx->i915;
 	struct drm_i915_gem_object *obj;
+	struct i915_address_space *vm;
 	const u32 RCS_GPR0 = 0x2600; /* not all engines have their own GPR! */
 	const u32 result = 0x100;
 	struct i915_request *rq;
@@ -1296,17 +1586,20 @@ static int read_from_scratch(struct i915_gem_context *ctx,
 	i915_gem_object_flush_map(obj);
 	i915_gem_object_unpin_map(obj);
 
-	vma = i915_vma_instance(obj, ctx->vm, NULL);
+	intel_gt_chipset_flush(engine->gt);
+
+	vm = i915_gem_context_get_vm_rcu(ctx);
+	vma = i915_vma_instance(obj, vm, NULL);
 	if (IS_ERR(vma)) {
 		err = PTR_ERR(vma);
-		goto err;
+		goto err_vm;
 	}
 
 	err = i915_vma_pin(vma, 0, 0, PIN_USER | PIN_OFFSET_FIXED);
 	if (err)
-		goto err;
+		goto err_vm;
 
-	err = check_scratch(ctx, offset);
+	err = check_scratch(vm, offset);
 	if (err)
 		goto err_unpin;
 
@@ -1337,12 +1630,12 @@ static int read_from_scratch(struct i915_gem_context *ctx,
 	err = i915_gem_object_set_to_cpu_domain(obj, false);
 	i915_gem_object_unlock(obj);
 	if (err)
-		goto err;
+		goto err_vm;
 
 	cmd = i915_gem_object_pin_map(obj, I915_MAP_WB);
 	if (IS_ERR(cmd)) {
 		err = PTR_ERR(cmd);
-		goto err;
+		goto err_vm;
 	}
 
 	*value = cmd[result / sizeof(*cmd)];
@@ -1357,6 +1650,8 @@ err_request:
 	i915_request_add(rq);
 err_unpin:
 	i915_vma_unpin(vma);
+err_vm:
+	i915_vm_put(vm);
 err:
 	i915_gem_object_put(obj);
 	return err;
@@ -1371,7 +1666,6 @@ static int igt_vm_isolation(void *arg)
 	struct drm_file *file;
 	I915_RND_STATE(prng);
 	unsigned long count;
-	unsigned int id;
 	u64 vm_total;
 	int err;
 
@@ -1387,34 +1681,32 @@ static int igt_vm_isolation(void *arg)
 	if (IS_ERR(file))
 		return PTR_ERR(file);
 
-	mutex_lock(&i915->drm.struct_mutex);
-
 	err = igt_live_test_begin(&t, i915, __func__, "");
 	if (err)
-		goto out_unlock;
+		goto out_file;
 
 	ctx_a = live_context(i915, file);
 	if (IS_ERR(ctx_a)) {
 		err = PTR_ERR(ctx_a);
-		goto out_unlock;
+		goto out_file;
 	}
 
 	ctx_b = live_context(i915, file);
 	if (IS_ERR(ctx_b)) {
 		err = PTR_ERR(ctx_b);
-		goto out_unlock;
+		goto out_file;
 	}
 
 	/* We can only test vm isolation, if the vm are distinct */
 	if (ctx_a->vm == ctx_b->vm)
-		goto out_unlock;
+		goto out_file;
 
 	vm_total = ctx_a->vm->total;
 	GEM_BUG_ON(ctx_b->vm->total != vm_total);
 	vm_total -= I915_GTT_PAGE_SIZE;
 
 	count = 0;
-	for_each_engine(engine, i915, id) {
+	for_each_uabi_engine(engine, i915) {
 		IGT_TIMEOUT(end_time);
 		unsigned long this = 0;
 
@@ -1436,7 +1728,7 @@ static int igt_vm_isolation(void *arg)
 				err = read_from_scratch(ctx_b, engine,
 							offset, &value);
 			if (err)
-				goto out_unlock;
+				goto out_file;
 
 			if (value) {
 				pr_err("%s: Read %08x from scratch (offset 0x%08x_%08x), after %lu reads!\n",
@@ -1445,7 +1737,7 @@ static int igt_vm_isolation(void *arg)
 				       lower_32_bits(offset),
 				       this);
 				err = -EINVAL;
-				goto out_unlock;
+				goto out_file;
 			}
 
 			this++;
@@ -1455,30 +1747,13 @@ static int igt_vm_isolation(void *arg)
 	pr_info("Checked %lu scratch offsets across %d engines\n",
 		count, RUNTIME_INFO(i915)->num_engines);
 
-out_unlock:
+out_file:
 	if (igt_live_test_end(&t))
 		err = -EIO;
-	mutex_unlock(&i915->drm.struct_mutex);
-
 	mock_file_free(i915, file);
 	return err;
 }
 
-static __maybe_unused const char *
-__engine_name(struct drm_i915_private *i915, intel_engine_mask_t engines)
-{
-	struct intel_engine_cs *engine;
-	intel_engine_mask_t tmp;
-
-	if (engines == ALL_ENGINES)
-		return "all";
-
-	for_each_engine_masked(engine, i915, engines, tmp)
-		return engine->name;
-
-	return "none";
-}
-
 static bool skip_unused_engines(struct intel_context *ce, void *data)
 {
 	return !ce->state;
@@ -1506,13 +1781,9 @@ static int mock_context_barrier(void *arg)
 	 * a request; useful for retiring old state after loading new.
 	 */
 
-	mutex_lock(&i915->drm.struct_mutex);
-
 	ctx = mock_context(i915, "mock");
-	if (!ctx) {
-		err = -ENOMEM;
-		goto unlock;
-	}
+	if (!ctx)
+		return -ENOMEM;
 
 	counter = 0;
 	err = context_barrier_task(ctx, 0,
@@ -1585,8 +1856,6 @@ static int mock_context_barrier(void *arg)
 
 out:
 	mock_context_close(ctx);
-unlock:
-	mutex_unlock(&i915->drm.struct_mutex);
 	return err;
 #undef pr_fmt
 #define pr_fmt(x) x
@@ -1614,6 +1883,7 @@ int i915_gem_context_live_selftests(struct drm_i915_private *i915)
 {
 	static const struct i915_subtest tests[] = {
 		SUBTEST(live_nop_switch),
+		SUBTEST(live_parallel_switch),
 		SUBTEST(igt_ctx_exec),
 		SUBTEST(igt_ctx_readonly),
 		SUBTEST(igt_ctx_sseu),
diff --git a/drivers/gpu/drm/i915/gem/selftests/i915_gem_mman.c b/drivers/gpu/drm/i915/gem/selftests/i915_gem_mman.c
index 1d27babff0ce..29b2077b73d2 100644
--- a/drivers/gpu/drm/i915/gem/selftests/i915_gem_mman.c
+++ b/drivers/gpu/drm/i915/gem/selftests/i915_gem_mman.c
@@ -10,6 +10,7 @@
 #include "gt/intel_gt_pm.h"
 #include "huge_gem_object.h"
 #include "i915_selftest.h"
+#include "selftests/i915_random.h"
 #include "selftests/igt_flush_test.h"
 
 struct tile {
@@ -76,18 +77,103 @@ static u64 tiled_offset(const struct tile *tile, u64 v)
 
 static int check_partial_mapping(struct drm_i915_gem_object *obj,
 				 const struct tile *tile,
-				 unsigned long end_time)
+				 struct rnd_state *prng)
 {
-	const unsigned int nreal = obj->scratch / PAGE_SIZE;
 	const unsigned long npages = obj->base.size / PAGE_SIZE;
+	struct i915_ggtt_view view;
 	struct i915_vma *vma;
 	unsigned long page;
+	u32 __iomem *io;
+	struct page *p;
+	unsigned int n;
+	u64 offset;
+	u32 *cpu;
 	int err;
 
-	if (igt_timeout(end_time,
-			"%s: timed out before tiling=%d stride=%d\n",
-			__func__, tile->tiling, tile->stride))
-		return -EINTR;
+	err = i915_gem_object_set_tiling(obj, tile->tiling, tile->stride);
+	if (err) {
+		pr_err("Failed to set tiling mode=%u, stride=%u, err=%d\n",
+		       tile->tiling, tile->stride, err);
+		return err;
+	}
+
+	GEM_BUG_ON(i915_gem_object_get_tiling(obj) != tile->tiling);
+	GEM_BUG_ON(i915_gem_object_get_stride(obj) != tile->stride);
+
+	i915_gem_object_lock(obj);
+	err = i915_gem_object_set_to_gtt_domain(obj, true);
+	i915_gem_object_unlock(obj);
+	if (err) {
+		pr_err("Failed to flush to GTT write domain; err=%d\n", err);
+		return err;
+	}
+
+	page = i915_prandom_u32_max_state(npages, prng);
+	view = compute_partial_view(obj, page, MIN_CHUNK_PAGES);
+
+	vma = i915_gem_object_ggtt_pin(obj, &view, 0, 0, PIN_MAPPABLE);
+	if (IS_ERR(vma)) {
+		pr_err("Failed to pin partial view: offset=%lu; err=%d\n",
+		       page, (int)PTR_ERR(vma));
+		return PTR_ERR(vma);
+	}
+
+	n = page - view.partial.offset;
+	GEM_BUG_ON(n >= view.partial.size);
+
+	io = i915_vma_pin_iomap(vma);
+	i915_vma_unpin(vma);
+	if (IS_ERR(io)) {
+		pr_err("Failed to iomap partial view: offset=%lu; err=%d\n",
+		       page, (int)PTR_ERR(io));
+		err = PTR_ERR(io);
+		goto out;
+	}
+
+	iowrite32(page, io + n * PAGE_SIZE / sizeof(*io));
+	i915_vma_unpin_iomap(vma);
+
+	offset = tiled_offset(tile, page << PAGE_SHIFT);
+	if (offset >= obj->base.size)
+		goto out;
+
+	intel_gt_flush_ggtt_writes(&to_i915(obj->base.dev)->gt);
+
+	p = i915_gem_object_get_page(obj, offset >> PAGE_SHIFT);
+	cpu = kmap(p) + offset_in_page(offset);
+	drm_clflush_virt_range(cpu, sizeof(*cpu));
+	if (*cpu != (u32)page) {
+		pr_err("Partial view for %lu [%u] (offset=%llu, size=%u [%llu, row size %u], fence=%d, tiling=%d, stride=%d) misalignment, expected write to page (%llu + %u [0x%llx]) of 0x%x, found 0x%x\n",
+		       page, n,
+		       view.partial.offset,
+		       view.partial.size,
+		       vma->size >> PAGE_SHIFT,
+		       tile->tiling ? tile_row_pages(obj) : 0,
+		       vma->fence ? vma->fence->id : -1, tile->tiling, tile->stride,
+		       offset >> PAGE_SHIFT,
+		       (unsigned int)offset_in_page(offset),
+		       offset,
+		       (u32)page, *cpu);
+		err = -EINVAL;
+	}
+	*cpu = 0;
+	drm_clflush_virt_range(cpu, sizeof(*cpu));
+	kunmap(p);
+
+out:
+	i915_vma_destroy(vma);
+	return err;
+}
+
+static int check_partial_mappings(struct drm_i915_gem_object *obj,
+				  const struct tile *tile,
+				  unsigned long end_time)
+{
+	const unsigned int nreal = obj->scratch / PAGE_SIZE;
+	const unsigned long npages = obj->base.size / PAGE_SIZE;
+	struct i915_vma *vma;
+	unsigned long page;
+	int err;
 
 	err = i915_gem_object_set_tiling(obj, tile->tiling, tile->stride);
 	if (err) {
@@ -170,11 +256,42 @@ static int check_partial_mapping(struct drm_i915_gem_object *obj,
 			return err;
 
 		i915_vma_destroy(vma);
+
+		if (igt_timeout(end_time,
+				"%s: timed out after tiling=%d stride=%d\n",
+				__func__, tile->tiling, tile->stride))
+			return -EINTR;
 	}
 
 	return 0;
 }
 
+static unsigned int
+setup_tile_size(struct tile *tile, struct drm_i915_private *i915)
+{
+	if (INTEL_GEN(i915) <= 2) {
+		tile->height = 16;
+		tile->width = 128;
+		tile->size = 11;
+	} else if (tile->tiling == I915_TILING_Y &&
+		   HAS_128_BYTE_Y_TILING(i915)) {
+		tile->height = 32;
+		tile->width = 128;
+		tile->size = 12;
+	} else {
+		tile->height = 8;
+		tile->width = 512;
+		tile->size = 12;
+	}
+
+	if (INTEL_GEN(i915) < 4)
+		return 8192 / tile->width;
+	else if (INTEL_GEN(i915) < 7)
+		return 128 * I965_FENCE_MAX_PITCH_VAL / tile->width;
+	else
+		return 128 * GEN7_FENCE_MAX_PITCH_VAL / tile->width;
+}
+
 static int igt_partial_tiling(void *arg)
 {
 	const unsigned int nreal = 1 << 12; /* largest tile row x2 */
@@ -184,6 +301,9 @@ static int igt_partial_tiling(void *arg)
 	int tiling;
 	int err;
 
+	if (!i915_ggtt_has_aperture(&i915->ggtt))
+		return 0;
+
 	/* We want to check the page mapping and fencing of a large object
 	 * mmapped through the GTT. The object we create is larger than can
 	 * possibly be mmaped as a whole, and so we must use partial GGTT vma.
@@ -205,7 +325,6 @@ static int igt_partial_tiling(void *arg)
 		goto out;
 	}
 
-	mutex_lock(&i915->drm.struct_mutex);
 	wakeref = intel_runtime_pm_get(&i915->runtime_pm);
 
 	if (1) {
@@ -219,7 +338,7 @@ static int igt_partial_tiling(void *arg)
 		tile.swizzle = I915_BIT_6_SWIZZLE_NONE;
 		tile.tiling = I915_TILING_NONE;
 
-		err = check_partial_mapping(obj, &tile, end);
+		err = check_partial_mappings(obj, &tile, end);
 		if (err && err != -EINTR)
 			goto out_unlock;
 	}
@@ -241,10 +360,10 @@ static int igt_partial_tiling(void *arg)
 		tile.tiling = tiling;
 		switch (tiling) {
 		case I915_TILING_X:
-			tile.swizzle = i915->mm.bit_6_swizzle_x;
+			tile.swizzle = i915->ggtt.bit_6_swizzle_x;
 			break;
 		case I915_TILING_Y:
-			tile.swizzle = i915->mm.bit_6_swizzle_y;
+			tile.swizzle = i915->ggtt.bit_6_swizzle_y;
 			break;
 		}
 
@@ -253,31 +372,11 @@ static int igt_partial_tiling(void *arg)
 		    tile.swizzle == I915_BIT_6_SWIZZLE_9_10_17)
 			continue;
 
-		if (INTEL_GEN(i915) <= 2) {
-			tile.height = 16;
-			tile.width = 128;
-			tile.size = 11;
-		} else if (tile.tiling == I915_TILING_Y &&
-			   HAS_128_BYTE_Y_TILING(i915)) {
-			tile.height = 32;
-			tile.width = 128;
-			tile.size = 12;
-		} else {
-			tile.height = 8;
-			tile.width = 512;
-			tile.size = 12;
-		}
-
-		if (INTEL_GEN(i915) < 4)
-			max_pitch = 8192 / tile.width;
-		else if (INTEL_GEN(i915) < 7)
-			max_pitch = 128 * I965_FENCE_MAX_PITCH_VAL / tile.width;
-		else
-			max_pitch = 128 * GEN7_FENCE_MAX_PITCH_VAL / tile.width;
+		max_pitch = setup_tile_size(&tile, i915);
 
 		for (pitch = max_pitch; pitch; pitch >>= 1) {
 			tile.stride = tile.width * pitch;
-			err = check_partial_mapping(obj, &tile, end);
+			err = check_partial_mappings(obj, &tile, end);
 			if (err == -EINTR)
 				goto next_tiling;
 			if (err)
@@ -285,7 +384,7 @@ static int igt_partial_tiling(void *arg)
 
 			if (pitch > 2 && INTEL_GEN(i915) >= 4) {
 				tile.stride = tile.width * (pitch - 1);
-				err = check_partial_mapping(obj, &tile, end);
+				err = check_partial_mappings(obj, &tile, end);
 				if (err == -EINTR)
 					goto next_tiling;
 				if (err)
@@ -294,7 +393,7 @@ static int igt_partial_tiling(void *arg)
 
 			if (pitch < max_pitch && INTEL_GEN(i915) >= 4) {
 				tile.stride = tile.width * (pitch + 1);
-				err = check_partial_mapping(obj, &tile, end);
+				err = check_partial_mappings(obj, &tile, end);
 				if (err == -EINTR)
 					goto next_tiling;
 				if (err)
@@ -305,7 +404,7 @@ static int igt_partial_tiling(void *arg)
 		if (INTEL_GEN(i915) >= 4) {
 			for_each_prime_number(pitch, max_pitch) {
 				tile.stride = tile.width * pitch;
-				err = check_partial_mapping(obj, &tile, end);
+				err = check_partial_mappings(obj, &tile, end);
 				if (err == -EINTR)
 					goto next_tiling;
 				if (err)
@@ -318,7 +417,100 @@ next_tiling: ;
 
 out_unlock:
 	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
-	mutex_unlock(&i915->drm.struct_mutex);
+	i915_gem_object_unpin_pages(obj);
+out:
+	i915_gem_object_put(obj);
+	return err;
+}
+
+static int igt_smoke_tiling(void *arg)
+{
+	const unsigned int nreal = 1 << 12; /* largest tile row x2 */
+	struct drm_i915_private *i915 = arg;
+	struct drm_i915_gem_object *obj;
+	intel_wakeref_t wakeref;
+	I915_RND_STATE(prng);
+	unsigned long count;
+	IGT_TIMEOUT(end);
+	int err;
+
+	if (!i915_ggtt_has_aperture(&i915->ggtt))
+		return 0;
+
+	/*
+	 * igt_partial_tiling() does an exhastive check of partial tiling
+	 * chunking, but will undoubtably run out of time. Here, we do a
+	 * randomised search and hope over many runs of 1s with different
+	 * seeds we will do a thorough check.
+	 *
+	 * Remember to look at the st_seed if we see a flip-flop in BAT!
+	 */
+
+	if (i915->quirks & QUIRK_PIN_SWIZZLED_PAGES)
+		return 0;
+
+	obj = huge_gem_object(i915,
+			      nreal << PAGE_SHIFT,
+			      (1 + next_prime_number(i915->ggtt.vm.total >> PAGE_SHIFT)) << PAGE_SHIFT);
+	if (IS_ERR(obj))
+		return PTR_ERR(obj);
+
+	err = i915_gem_object_pin_pages(obj);
+	if (err) {
+		pr_err("Failed to allocate %u pages (%lu total), err=%d\n",
+		       nreal, obj->base.size / PAGE_SIZE, err);
+		goto out;
+	}
+
+	wakeref = intel_runtime_pm_get(&i915->runtime_pm);
+
+	count = 0;
+	do {
+		struct tile tile;
+
+		tile.tiling =
+			i915_prandom_u32_max_state(I915_TILING_Y + 1, &prng);
+		switch (tile.tiling) {
+		case I915_TILING_NONE:
+			tile.height = 1;
+			tile.width = 1;
+			tile.size = 0;
+			tile.stride = 0;
+			tile.swizzle = I915_BIT_6_SWIZZLE_NONE;
+			break;
+
+		case I915_TILING_X:
+			tile.swizzle = i915->ggtt.bit_6_swizzle_x;
+			break;
+		case I915_TILING_Y:
+			tile.swizzle = i915->ggtt.bit_6_swizzle_y;
+			break;
+		}
+
+		if (tile.swizzle == I915_BIT_6_SWIZZLE_9_17 ||
+		    tile.swizzle == I915_BIT_6_SWIZZLE_9_10_17)
+			continue;
+
+		if (tile.tiling != I915_TILING_NONE) {
+			unsigned int max_pitch = setup_tile_size(&tile, i915);
+
+			tile.stride =
+				i915_prandom_u32_max_state(max_pitch, &prng);
+			tile.stride = (1 + tile.stride) * tile.width;
+			if (INTEL_GEN(i915) < 4)
+				tile.stride = rounddown_pow_of_two(tile.stride);
+		}
+
+		err = check_partial_mapping(obj, &tile, &prng);
+		if (err)
+			break;
+
+		count++;
+	} while (!__igt_timeout(end, NULL));
+
+	pr_info("%s: Completed %lu trials\n", __func__, count);
+
+	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
 	i915_gem_object_unpin_pages(obj);
 out:
 	i915_gem_object_put(obj);
@@ -329,20 +521,19 @@ static int make_obj_busy(struct drm_i915_gem_object *obj)
 {
 	struct drm_i915_private *i915 = to_i915(obj->base.dev);
 	struct intel_engine_cs *engine;
-	enum intel_engine_id id;
-	struct i915_vma *vma;
-	int err;
 
-	vma = i915_vma_instance(obj, &i915->ggtt.vm, NULL);
-	if (IS_ERR(vma))
-		return PTR_ERR(vma);
+	for_each_uabi_engine(engine, i915) {
+		struct i915_request *rq;
+		struct i915_vma *vma;
+		int err;
 
-	err = i915_vma_pin(vma, 0, 0, PIN_USER);
-	if (err)
-		return err;
+		vma = i915_vma_instance(obj, &engine->gt->ggtt->vm, NULL);
+		if (IS_ERR(vma))
+			return PTR_ERR(vma);
 
-	for_each_engine(engine, i915, id) {
-		struct i915_request *rq;
+		err = i915_vma_pin(vma, 0, 0, PIN_USER);
+		if (err)
+			return err;
 
 		rq = i915_request_create(engine->kernel_context);
 		if (IS_ERR(rq)) {
@@ -358,12 +549,13 @@ static int make_obj_busy(struct drm_i915_gem_object *obj)
 		i915_vma_unlock(vma);
 
 		i915_request_add(rq);
+		i915_vma_unpin(vma);
+		if (err)
+			return err;
 	}
 
-	i915_vma_unpin(vma);
 	i915_gem_object_put(obj); /* leave it only alive via its active ref */
-
-	return err;
+	return 0;
 }
 
 static bool assert_mmap_offset(struct drm_i915_private *i915,
@@ -386,21 +578,14 @@ static bool assert_mmap_offset(struct drm_i915_private *i915,
 static void disable_retire_worker(struct drm_i915_private *i915)
 {
 	i915_gem_driver_unregister__shrinker(i915);
-
 	intel_gt_pm_get(&i915->gt);
-
-	cancel_delayed_work_sync(&i915->gem.retire_work);
-	flush_work(&i915->gem.idle_work);
+	cancel_delayed_work_sync(&i915->gt.requests.retire_work);
 }
 
 static void restore_retire_worker(struct drm_i915_private *i915)
 {
+	igt_flush_test(i915);
 	intel_gt_pm_put(&i915->gt);
-
-	mutex_lock(&i915->drm.struct_mutex);
-	igt_flush_test(i915, I915_WAIT_LOCKED);
-	mutex_unlock(&i915->drm.struct_mutex);
-
 	i915_gem_driver_register__shrinker(i915);
 }
 
@@ -490,9 +675,7 @@ static int igt_mmap_offset_exhaustion(void *arg)
 			goto out;
 		}
 
-		mutex_lock(&i915->drm.struct_mutex);
 		err = make_obj_busy(obj);
-		mutex_unlock(&i915->drm.struct_mutex);
 		if (err) {
 			pr_err("[loop %d] Failed to busy the object\n", loop);
 			goto err_obj;
@@ -515,6 +698,7 @@ int i915_gem_mman_live_selftests(struct drm_i915_private *i915)
 {
 	static const struct i915_subtest tests[] = {
 		SUBTEST(igt_partial_tiling),
+		SUBTEST(igt_smoke_tiling),
 		SUBTEST(igt_mmap_offset_exhaustion),
 	};
 
diff --git a/drivers/gpu/drm/i915/gem/selftests/i915_gem_object_blt.c b/drivers/gpu/drm/i915/gem/selftests/i915_gem_object_blt.c
index c21d747e7d05..e8132aca0bb6 100644
--- a/drivers/gpu/drm/i915/gem/selftests/i915_gem_object_blt.c
+++ b/drivers/gpu/drm/i915/gem/selftests/i915_gem_object_blt.c
@@ -3,40 +3,241 @@
  * Copyright © 2019 Intel Corporation
  */
 
+#include <linux/sort.h>
+
 #include "gt/intel_gt.h"
+#include "gt/intel_engine_user.h"
 
 #include "i915_selftest.h"
 
+#include "gem/i915_gem_context.h"
 #include "selftests/igt_flush_test.h"
+#include "selftests/i915_random.h"
 #include "selftests/mock_drm.h"
 #include "huge_gem_object.h"
 #include "mock_context.h"
 
-static int igt_fill_blt(void *arg)
+static int wrap_ktime_compare(const void *A, const void *B)
+{
+	const ktime_t *a = A, *b = B;
+
+	return ktime_compare(*a, *b);
+}
+
+static int __perf_fill_blt(struct drm_i915_gem_object *obj)
+{
+	struct drm_i915_private *i915 = to_i915(obj->base.dev);
+	int inst = 0;
+
+	do {
+		struct intel_engine_cs *engine;
+		ktime_t t[5];
+		int pass;
+		int err;
+
+		engine = intel_engine_lookup_user(i915,
+						  I915_ENGINE_CLASS_COPY,
+						  inst++);
+		if (!engine)
+			return 0;
+
+		for (pass = 0; pass < ARRAY_SIZE(t); pass++) {
+			struct intel_context *ce = engine->kernel_context;
+			ktime_t t0, t1;
+
+			t0 = ktime_get();
+
+			err = i915_gem_object_fill_blt(obj, ce, 0);
+			if (err)
+				return err;
+
+			err = i915_gem_object_wait(obj,
+						   I915_WAIT_ALL,
+						   MAX_SCHEDULE_TIMEOUT);
+			if (err)
+				return err;
+
+			t1 = ktime_get();
+			t[pass] = ktime_sub(t1, t0);
+		}
+
+		sort(t, ARRAY_SIZE(t), sizeof(*t), wrap_ktime_compare, NULL);
+		pr_info("%s: blt %zd KiB fill: %lld MiB/s\n",
+			engine->name,
+			obj->base.size >> 10,
+			div64_u64(mul_u32_u32(4 * obj->base.size,
+					      1000 * 1000 * 1000),
+				  t[1] + 2 * t[2] + t[3]) >> 20);
+	} while (1);
+}
+
+static int perf_fill_blt(void *arg)
 {
 	struct drm_i915_private *i915 = arg;
-	struct intel_context *ce = i915->engine[BCS0]->kernel_context;
-	struct drm_i915_gem_object *obj;
+	static const unsigned long sizes[] = {
+		SZ_4K,
+		SZ_64K,
+		SZ_2M,
+		SZ_64M
+	};
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(sizes); i++) {
+		struct drm_i915_gem_object *obj;
+		int err;
+
+		obj = i915_gem_object_create_internal(i915, sizes[i]);
+		if (IS_ERR(obj))
+			return PTR_ERR(obj);
+
+		err = __perf_fill_blt(obj);
+		i915_gem_object_put(obj);
+		if (err)
+			return err;
+	}
+
+	return 0;
+}
+
+static int __perf_copy_blt(struct drm_i915_gem_object *src,
+			   struct drm_i915_gem_object *dst)
+{
+	struct drm_i915_private *i915 = to_i915(src->base.dev);
+	int inst = 0;
+
+	do {
+		struct intel_engine_cs *engine;
+		ktime_t t[5];
+		int pass;
+
+		engine = intel_engine_lookup_user(i915,
+						  I915_ENGINE_CLASS_COPY,
+						  inst++);
+		if (!engine)
+			return 0;
+
+		for (pass = 0; pass < ARRAY_SIZE(t); pass++) {
+			struct intel_context *ce = engine->kernel_context;
+			ktime_t t0, t1;
+			int err;
+
+			t0 = ktime_get();
+
+			err = i915_gem_object_copy_blt(src, dst, ce);
+			if (err)
+				return err;
+
+			err = i915_gem_object_wait(dst,
+						   I915_WAIT_ALL,
+						   MAX_SCHEDULE_TIMEOUT);
+			if (err)
+				return err;
+
+			t1 = ktime_get();
+			t[pass] = ktime_sub(t1, t0);
+		}
+
+		sort(t, ARRAY_SIZE(t), sizeof(*t), wrap_ktime_compare, NULL);
+		pr_info("%s: blt %zd KiB copy: %lld MiB/s\n",
+			engine->name,
+			src->base.size >> 10,
+			div64_u64(mul_u32_u32(4 * src->base.size,
+					      1000 * 1000 * 1000),
+				  t[1] + 2 * t[2] + t[3]) >> 20);
+	} while (1);
+}
+
+static int perf_copy_blt(void *arg)
+{
+	struct drm_i915_private *i915 = arg;
+	static const unsigned long sizes[] = {
+		SZ_4K,
+		SZ_64K,
+		SZ_2M,
+		SZ_64M
+	};
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(sizes); i++) {
+		struct drm_i915_gem_object *src, *dst;
+		int err;
+
+		src = i915_gem_object_create_internal(i915, sizes[i]);
+		if (IS_ERR(src))
+			return PTR_ERR(src);
+
+		dst = i915_gem_object_create_internal(i915, sizes[i]);
+		if (IS_ERR(dst)) {
+			err = PTR_ERR(dst);
+			goto err_src;
+		}
+
+		err = __perf_copy_blt(src, dst);
+
+		i915_gem_object_put(dst);
+err_src:
+		i915_gem_object_put(src);
+		if (err)
+			return err;
+	}
+
+	return 0;
+}
+
+struct igt_thread_arg {
+	struct drm_i915_private *i915;
 	struct rnd_state prng;
+	unsigned int n_cpus;
+};
+
+static int igt_fill_blt_thread(void *arg)
+{
+	struct igt_thread_arg *thread = arg;
+	struct drm_i915_private *i915 = thread->i915;
+	struct rnd_state *prng = &thread->prng;
+	struct drm_i915_gem_object *obj;
+	struct i915_gem_context *ctx;
+	struct intel_context *ce;
+	struct drm_file *file;
+	unsigned int prio;
 	IGT_TIMEOUT(end);
-	u32 *vaddr;
-	int err = 0;
+	int err;
+
+	file = mock_file(i915);
+	if (IS_ERR(file))
+		return PTR_ERR(file);
+
+	ctx = live_context(i915, file);
+	if (IS_ERR(ctx)) {
+		err = PTR_ERR(ctx);
+		goto out_file;
+	}
 
-	prandom_seed_state(&prng, i915_selftest.random_seed);
+	prio = i915_prandom_u32_max_state(I915_PRIORITY_MAX, prng);
+	ctx->sched.priority = I915_USER_PRIORITY(prio);
 
-	/*
-	 * XXX: needs some threads to scale all these tests, also maybe throw
-	 * in submission from higher priority context to see if we are
-	 * preempted for very large objects...
-	 */
+	ce = i915_gem_context_get_engine(ctx, BCS0);
+	GEM_BUG_ON(IS_ERR(ce));
 
 	do {
 		const u32 max_block_size = S16_MAX * PAGE_SIZE;
-		u32 sz = min_t(u64, ce->vm->total >> 4, prandom_u32_state(&prng));
-		u32 phys_sz = sz % (max_block_size + 1);
-		u32 val = prandom_u32_state(&prng);
+		u32 val = prandom_u32_state(prng);
+		u64 total = ce->vm->total;
+		u32 phys_sz;
+		u32 sz;
+		u32 *vaddr;
 		u32 i;
 
+		/*
+		 * If we have a tiny shared address space, like for the GGTT
+		 * then we can't be too greedy.
+		 */
+		if (i915_is_ggtt(ce->vm))
+			total = div64_u64(total, thread->n_cpus);
+
+		sz = min_t(u64, total >> 4, prandom_u32_state(prng));
+		phys_sz = sz % (max_block_size + 1);
+
 		sz = round_up(sz, PAGE_SIZE);
 		phys_sz = round_up(phys_sz, PAGE_SIZE);
 
@@ -65,9 +266,7 @@ static int igt_fill_blt(void *arg)
 		if (!(obj->cache_coherent & I915_BO_CACHE_COHERENT_FOR_WRITE))
 			obj->cache_dirty = true;
 
-		mutex_lock(&i915->drm.struct_mutex);
 		err = i915_gem_object_fill_blt(obj, ce, val);
-		mutex_unlock(&i915->drm.struct_mutex);
 		if (err)
 			goto err_unpin;
 
@@ -100,28 +299,56 @@ err_flush:
 	if (err == -ENOMEM)
 		err = 0;
 
+	intel_context_put(ce);
+out_file:
+	mock_file_free(i915, file);
 	return err;
 }
 
-static int igt_copy_blt(void *arg)
+static int igt_copy_blt_thread(void *arg)
 {
-	struct drm_i915_private *i915 = arg;
-	struct intel_context *ce = i915->engine[BCS0]->kernel_context;
+	struct igt_thread_arg *thread = arg;
+	struct drm_i915_private *i915 = thread->i915;
+	struct rnd_state *prng = &thread->prng;
 	struct drm_i915_gem_object *src, *dst;
-	struct rnd_state prng;
+	struct i915_gem_context *ctx;
+	struct intel_context *ce;
+	struct drm_file *file;
+	unsigned int prio;
 	IGT_TIMEOUT(end);
-	u32 *vaddr;
-	int err = 0;
+	int err;
+
+	file = mock_file(i915);
+	if (IS_ERR(file))
+		return PTR_ERR(file);
 
-	prandom_seed_state(&prng, i915_selftest.random_seed);
+	ctx = live_context(i915, file);
+	if (IS_ERR(ctx)) {
+		err = PTR_ERR(ctx);
+		goto out_file;
+	}
+
+	prio = i915_prandom_u32_max_state(I915_PRIORITY_MAX, prng);
+	ctx->sched.priority = I915_USER_PRIORITY(prio);
+
+	ce = i915_gem_context_get_engine(ctx, BCS0);
+	GEM_BUG_ON(IS_ERR(ce));
 
 	do {
 		const u32 max_block_size = S16_MAX * PAGE_SIZE;
-		u32 sz = min_t(u64, ce->vm->total >> 4, prandom_u32_state(&prng));
-		u32 phys_sz = sz % (max_block_size + 1);
-		u32 val = prandom_u32_state(&prng);
+		u32 val = prandom_u32_state(prng);
+		u64 total = ce->vm->total;
+		u32 phys_sz;
+		u32 sz;
+		u32 *vaddr;
 		u32 i;
 
+		if (i915_is_ggtt(ce->vm))
+			total = div64_u64(total, thread->n_cpus);
+
+		sz = min_t(u64, total >> 4, prandom_u32_state(prng));
+		phys_sz = sz % (max_block_size + 1);
+
 		sz = round_up(sz, PAGE_SIZE);
 		phys_sz = round_up(phys_sz, PAGE_SIZE);
 
@@ -166,9 +393,7 @@ static int igt_copy_blt(void *arg)
 		if (!(dst->cache_coherent & I915_BO_CACHE_COHERENT_FOR_WRITE))
 			dst->cache_dirty = true;
 
-		mutex_lock(&i915->drm.struct_mutex);
 		err = i915_gem_object_copy_blt(src, dst, ce);
-		mutex_unlock(&i915->drm.struct_mutex);
 		if (err)
 			goto err_unpin;
 
@@ -205,12 +430,85 @@ err_flush:
 	if (err == -ENOMEM)
 		err = 0;
 
+	intel_context_put(ce);
+out_file:
+	mock_file_free(i915, file);
+	return err;
+}
+
+static int igt_threaded_blt(struct drm_i915_private *i915,
+			    int (*blt_fn)(void *arg))
+{
+	struct igt_thread_arg *thread;
+	struct task_struct **tsk;
+	I915_RND_STATE(prng);
+	unsigned int n_cpus;
+	unsigned int i;
+	int err = 0;
+
+	n_cpus = num_online_cpus() + 1;
+
+	tsk = kcalloc(n_cpus, sizeof(struct task_struct *), GFP_KERNEL);
+	if (!tsk)
+		return 0;
+
+	thread = kcalloc(n_cpus, sizeof(struct igt_thread_arg), GFP_KERNEL);
+	if (!thread) {
+		kfree(tsk);
+		return 0;
+	}
+
+	for (i = 0; i < n_cpus; ++i) {
+		thread[i].i915 = i915;
+		thread[i].n_cpus = n_cpus;
+		thread[i].prng =
+			I915_RND_STATE_INITIALIZER(prandom_u32_state(&prng));
+
+		tsk[i] = kthread_run(blt_fn, &thread[i], "igt/blt-%d", i);
+		if (IS_ERR(tsk[i])) {
+			err = PTR_ERR(tsk[i]);
+			break;
+		}
+
+		get_task_struct(tsk[i]);
+	}
+
+	yield(); /* start all threads before we kthread_stop() */
+
+	for (i = 0; i < n_cpus; ++i) {
+		int status;
+
+		if (IS_ERR_OR_NULL(tsk[i]))
+			continue;
+
+		status = kthread_stop(tsk[i]);
+		if (status && !err)
+			err = status;
+
+		put_task_struct(tsk[i]);
+	}
+
+	kfree(tsk);
+	kfree(thread);
+
 	return err;
 }
 
+static int igt_fill_blt(void *arg)
+{
+	return igt_threaded_blt(arg, igt_fill_blt_thread);
+}
+
+static int igt_copy_blt(void *arg)
+{
+	return igt_threaded_blt(arg, igt_copy_blt_thread);
+}
+
 int i915_gem_object_blt_live_selftests(struct drm_i915_private *i915)
 {
 	static const struct i915_subtest tests[] = {
+		SUBTEST(perf_fill_blt),
+		SUBTEST(perf_copy_blt),
 		SUBTEST(igt_fill_blt),
 		SUBTEST(igt_copy_blt),
 	};
diff --git a/drivers/gpu/drm/i915/gem/selftests/i915_gem_phys.c b/drivers/gpu/drm/i915/gem/selftests/i915_gem_phys.c
index 94a15e3f6db8..34932871b3a5 100644
--- a/drivers/gpu/drm/i915/gem/selftests/i915_gem_phys.c
+++ b/drivers/gpu/drm/i915/gem/selftests/i915_gem_phys.c
@@ -25,9 +25,7 @@ static int mock_phys_object(void *arg)
 		goto out;
 	}
 
-	mutex_lock(&i915->drm.struct_mutex);
 	err = i915_gem_object_attach_phys(obj, PAGE_SIZE);
-	mutex_unlock(&i915->drm.struct_mutex);
 	if (err) {
 		pr_err("i915_gem_object_attach_phys failed, err=%d\n", err);
 		goto out_obj;
diff --git a/drivers/gpu/drm/i915/gem/selftests/igt_gem_utils.c b/drivers/gpu/drm/i915/gem/selftests/igt_gem_utils.c
index 57ece53c1075..6718da20f35d 100644
--- a/drivers/gpu/drm/i915/gem/selftests/igt_gem_utils.c
+++ b/drivers/gpu/drm/i915/gem/selftests/igt_gem_utils.c
@@ -9,6 +9,7 @@
 #include "gem/i915_gem_context.h"
 #include "gem/i915_gem_pm.h"
 #include "gt/intel_context.h"
+#include "gt/intel_gt.h"
 #include "i915_vma.h"
 #include "i915_drv.h"
 
@@ -84,6 +85,8 @@ igt_emit_store_dw(struct i915_vma *vma,
 	*cmd = MI_BATCH_BUFFER_END;
 	i915_gem_object_unpin_map(obj);
 
+	intel_gt_chipset_flush(vma->vm->gt);
+
 	vma = i915_vma_instance(obj, vma->vm, NULL);
 	if (IS_ERR(vma)) {
 		err = PTR_ERR(vma);
@@ -101,40 +104,35 @@ err:
 	return ERR_PTR(err);
 }
 
-int igt_gpu_fill_dw(struct i915_vma *vma,
-		    struct i915_gem_context *ctx,
-		    struct intel_engine_cs *engine,
-		    u64 offset,
-		    unsigned long count,
-		    u32 val)
+int igt_gpu_fill_dw(struct intel_context *ce,
+		    struct i915_vma *vma, u64 offset,
+		    unsigned long count, u32 val)
 {
-	struct i915_address_space *vm = ctx->vm ?: &engine->gt->ggtt->vm;
 	struct i915_request *rq;
 	struct i915_vma *batch;
 	unsigned int flags;
 	int err;
 
-	GEM_BUG_ON(vma->size > vm->total);
-	GEM_BUG_ON(!intel_engine_can_store_dword(engine));
+	GEM_BUG_ON(!intel_engine_can_store_dword(ce->engine));
 	GEM_BUG_ON(!i915_vma_is_pinned(vma));
 
 	batch = igt_emit_store_dw(vma, offset, count, val);
 	if (IS_ERR(batch))
 		return PTR_ERR(batch);
 
-	rq = igt_request_alloc(ctx, engine);
+	rq = intel_context_create_request(ce);
 	if (IS_ERR(rq)) {
 		err = PTR_ERR(rq);
 		goto err_batch;
 	}
 
 	flags = 0;
-	if (INTEL_GEN(vm->i915) <= 5)
+	if (INTEL_GEN(ce->vm->i915) <= 5)
 		flags |= I915_DISPATCH_SECURE;
 
-	err = engine->emit_bb_start(rq,
-				    batch->node.start, batch->node.size,
-				    flags);
+	err = rq->engine->emit_bb_start(rq,
+					batch->node.start, batch->node.size,
+					flags);
 	if (err)
 		goto err_request;
 
@@ -156,9 +154,7 @@ int igt_gpu_fill_dw(struct i915_vma *vma,
 
 	i915_request_add(rq);
 
-	i915_vma_unpin(batch);
-	i915_vma_close(batch);
-	i915_vma_put(batch);
+	i915_vma_unpin_and_release(&batch, 0);
 
 	return 0;
 
@@ -167,7 +163,6 @@ skip_request:
 err_request:
 	i915_request_add(rq);
 err_batch:
-	i915_vma_unpin(batch);
-	i915_vma_put(batch);
+	i915_vma_unpin_and_release(&batch, 0);
 	return err;
 }
diff --git a/drivers/gpu/drm/i915/gem/selftests/igt_gem_utils.h b/drivers/gpu/drm/i915/gem/selftests/igt_gem_utils.h
index 361a7ef866b0..4221cf84d175 100644
--- a/drivers/gpu/drm/i915/gem/selftests/igt_gem_utils.h
+++ b/drivers/gpu/drm/i915/gem/selftests/igt_gem_utils.h
@@ -11,9 +11,11 @@
 
 struct i915_request;
 struct i915_gem_context;
-struct intel_engine_cs;
 struct i915_vma;
 
+struct intel_context;
+struct intel_engine_cs;
+
 struct i915_request *
 igt_request_alloc(struct i915_gem_context *ctx, struct intel_engine_cs *engine);
 
@@ -23,11 +25,8 @@ igt_emit_store_dw(struct i915_vma *vma,
 		  unsigned long count,
 		  u32 val);
 
-int igt_gpu_fill_dw(struct i915_vma *vma,
-		    struct i915_gem_context *ctx,
-		    struct intel_engine_cs *engine,
-		    u64 offset,
-		    unsigned long count,
-		    u32 val);
+int igt_gpu_fill_dw(struct intel_context *ce,
+		    struct i915_vma *vma, u64 offset,
+		    unsigned long count, u32 val);
 
 #endif /* __IGT_GEM_UTILS_H__ */
diff --git a/drivers/gpu/drm/i915/gem/selftests/mock_context.c b/drivers/gpu/drm/i915/gem/selftests/mock_context.c
index be8974ccff24..29b8984f0e47 100644
--- a/drivers/gpu/drm/i915/gem/selftests/mock_context.c
+++ b/drivers/gpu/drm/i915/gem/selftests/mock_context.c
@@ -13,7 +13,6 @@ mock_context(struct drm_i915_private *i915,
 {
 	struct i915_gem_context *ctx;
 	struct i915_gem_engines *e;
-	int ret;
 
 	ctx = kzalloc(sizeof(*ctx), GFP_KERNEL);
 	if (!ctx)
@@ -23,6 +22,8 @@ mock_context(struct drm_i915_private *i915,
 	INIT_LIST_HEAD(&ctx->link);
 	ctx->i915 = i915;
 
+	i915_gem_context_set_persistence(ctx);
+
 	mutex_init(&ctx->engines_mutex);
 	e = default_engines(ctx);
 	if (IS_ERR(e))
@@ -30,13 +31,8 @@ mock_context(struct drm_i915_private *i915,
 	RCU_INIT_POINTER(ctx->engines, e);
 
 	INIT_RADIX_TREE(&ctx->handles_vma, GFP_KERNEL);
-	INIT_LIST_HEAD(&ctx->hw_id_link);
 	mutex_init(&ctx->mutex);
 
-	ret = i915_gem_context_pin_hw_id(ctx);
-	if (ret < 0)
-		goto err_engines;
-
 	if (name) {
 		struct i915_ppgtt *ppgtt;
 
@@ -48,14 +44,15 @@ mock_context(struct drm_i915_private *i915,
 		if (!ppgtt)
 			goto err_put;
 
+		mutex_lock(&ctx->mutex);
 		__set_ppgtt(ctx, &ppgtt->vm);
+		mutex_unlock(&ctx->mutex);
+
 		i915_vm_put(&ppgtt->vm);
 	}
 
 	return ctx;
 
-err_engines:
-	free_engines(rcu_access_pointer(ctx->engines));
 err_free:
 	kfree(ctx);
 	return NULL;
@@ -73,7 +70,7 @@ void mock_context_close(struct i915_gem_context *ctx)
 
 void mock_init_contexts(struct drm_i915_private *i915)
 {
-	init_contexts(i915);
+	init_contexts(&i915->gem.contexts);
 }
 
 struct i915_gem_context *
@@ -82,8 +79,6 @@ live_context(struct drm_i915_private *i915, struct drm_file *file)
 	struct i915_gem_context *ctx;
 	int err;
 
-	lockdep_assert_held(&i915->drm.struct_mutex);
-
 	ctx = i915_gem_create_context(i915, 0);
 	if (IS_ERR(ctx))
 		return ctx;
diff --git a/drivers/gpu/drm/i915/gt/intel_breadcrumbs.c b/drivers/gpu/drm/i915/gt/intel_breadcrumbs.c
index 09c68dda2098..55317081d48b 100644
--- a/drivers/gpu/drm/i915/gt/intel_breadcrumbs.c
+++ b/drivers/gpu/drm/i915/gt/intel_breadcrumbs.c
@@ -120,7 +120,6 @@ __dma_fence_signal__notify(struct dma_fence *fence,
 	struct dma_fence_cb *cur, *tmp;
 
 	lockdep_assert_held(fence->lock);
-	lockdep_assert_irqs_disabled();
 
 	list_for_each_entry_safe(cur, tmp, list, node) {
 		INIT_LIST_HEAD(&cur->node);
@@ -134,9 +133,10 @@ void intel_engine_breadcrumbs_irq(struct intel_engine_cs *engine)
 	const ktime_t timestamp = ktime_get();
 	struct intel_context *ce, *cn;
 	struct list_head *pos, *next;
+	unsigned long flags;
 	LIST_HEAD(signal);
 
-	spin_lock(&b->irq_lock);
+	spin_lock_irqsave(&b->irq_lock, flags);
 
 	if (b->irq_armed && list_empty(&b->signalers))
 		__intel_breadcrumbs_disarm_irq(b);
@@ -182,30 +182,23 @@ void intel_engine_breadcrumbs_irq(struct intel_engine_cs *engine)
 		}
 	}
 
-	spin_unlock(&b->irq_lock);
+	spin_unlock_irqrestore(&b->irq_lock, flags);
 
 	list_for_each_safe(pos, next, &signal) {
 		struct i915_request *rq =
 			list_entry(pos, typeof(*rq), signal_link);
 		struct list_head cb_list;
 
-		spin_lock(&rq->lock);
+		spin_lock_irqsave(&rq->lock, flags);
 		list_replace(&rq->fence.cb_list, &cb_list);
 		__dma_fence_signal__timestamp(&rq->fence, timestamp);
 		__dma_fence_signal__notify(&rq->fence, &cb_list);
-		spin_unlock(&rq->lock);
+		spin_unlock_irqrestore(&rq->lock, flags);
 
 		i915_request_put(rq);
 	}
 }
 
-void intel_engine_signal_breadcrumbs(struct intel_engine_cs *engine)
-{
-	local_irq_disable();
-	intel_engine_breadcrumbs_irq(engine);
-	local_irq_enable();
-}
-
 static void signal_irq_work(struct irq_work *work)
 {
 	struct intel_engine_cs *engine =
@@ -275,7 +268,6 @@ void intel_engine_fini_breadcrumbs(struct intel_engine_cs *engine)
 bool i915_request_enable_breadcrumb(struct i915_request *rq)
 {
 	lockdep_assert_held(&rq->lock);
-	lockdep_assert_irqs_disabled();
 
 	if (test_bit(I915_FENCE_FLAG_ACTIVE, &rq->fence.flags)) {
 		struct intel_breadcrumbs *b = &rq->engine->breadcrumbs;
@@ -325,7 +317,6 @@ void i915_request_cancel_breadcrumb(struct i915_request *rq)
 	struct intel_breadcrumbs *b = &rq->engine->breadcrumbs;
 
 	lockdep_assert_held(&rq->lock);
-	lockdep_assert_irqs_disabled();
 
 	/*
 	 * We must wait for b->irq_lock so that we know the interrupt handler
diff --git a/drivers/gpu/drm/i915/gt/intel_context.c b/drivers/gpu/drm/i915/gt/intel_context.c
index f55691d151ae..ee9d2bcd2c13 100644
--- a/drivers/gpu/drm/i915/gt/intel_context.c
+++ b/drivers/gpu/drm/i915/gt/intel_context.c
@@ -13,6 +13,7 @@
 #include "intel_context.h"
 #include "intel_engine.h"
 #include "intel_engine_pm.h"
+#include "intel_ring.h"
 
 static struct i915_global_context {
 	struct i915_global base;
@@ -62,7 +63,7 @@ int __intel_context_do_pin(struct intel_context *ce)
 		}
 
 		err = 0;
-		with_intel_runtime_pm(&ce->engine->i915->runtime_pm, wakeref)
+		with_intel_runtime_pm(ce->engine->uncore->rpm, wakeref)
 			err = ce->ops->pin(ce);
 		if (err)
 			goto err;
@@ -134,10 +135,11 @@ static int __context_pin_state(struct i915_vma *vma)
 
 static void __context_unpin_state(struct i915_vma *vma)
 {
-	__i915_vma_unpin(vma);
 	i915_vma_make_shrinkable(vma);
+	__i915_vma_unpin(vma);
 }
 
+__i915_active_call
 static void __intel_context_retire(struct i915_active *active)
 {
 	struct intel_context *ce = container_of(active, typeof(*ce), active);
@@ -150,6 +152,7 @@ static void __intel_context_retire(struct i915_active *active)
 
 	intel_timeline_unpin(ce->timeline);
 	intel_ring_unpin(ce->ring);
+
 	intel_context_put(ce);
 }
 
@@ -219,12 +222,20 @@ intel_context_init(struct intel_context *ce,
 		   struct i915_gem_context *ctx,
 		   struct intel_engine_cs *engine)
 {
+	struct i915_address_space *vm;
+
 	GEM_BUG_ON(!engine->cops);
 
 	kref_init(&ce->ref);
 
 	ce->gem_context = ctx;
-	ce->vm = i915_vm_get(ctx->vm ?: &engine->gt->ggtt->vm);
+	rcu_read_lock();
+	vm = rcu_dereference(ctx->vm);
+	if (vm)
+		ce->vm = i915_vm_get(vm);
+	else
+		ce->vm = i915_vm_get(&engine->gt->ggtt->vm);
+	rcu_read_unlock();
 	if (ctx->timeline)
 		ce->timeline = intel_timeline_get(ctx->timeline);
 
@@ -238,7 +249,7 @@ intel_context_init(struct intel_context *ce,
 
 	mutex_init(&ce->pin_mutex);
 
-	i915_active_init(ctx->i915, &ce->active,
+	i915_active_init(&ce->active,
 			 __intel_context_active, __intel_context_retire);
 }
 
@@ -298,14 +309,14 @@ int intel_context_prepare_remote_request(struct intel_context *ce,
 	/* Only suitable for use in remotely modifying this context */
 	GEM_BUG_ON(rq->hw_context == ce);
 
-	if (rq->timeline != tl) { /* beware timeline sharing */
+	if (rcu_access_pointer(rq->timeline) != tl) { /* timeline sharing! */
 		err = mutex_lock_interruptible_nested(&tl->mutex,
 						      SINGLE_DEPTH_NESTING);
 		if (err)
 			return err;
 
 		/* Queue this switch after current activity by this context. */
-		err = i915_active_request_set(&tl->last_request, rq);
+		err = i915_active_fence_set(&tl->last_request, rq);
 		mutex_unlock(&tl->mutex);
 		if (err)
 			return err;
@@ -319,7 +330,7 @@ int intel_context_prepare_remote_request(struct intel_context *ce,
 	 * words transfer the pinned ce object to tracked active request.
 	 */
 	GEM_BUG_ON(i915_active_is_idle(&ce->active));
-	return i915_active_ref(&ce->active, rq->timeline, rq);
+	return i915_active_add_request(&ce->active, rq);
 }
 
 struct i915_request *intel_context_create_request(struct intel_context *ce)
diff --git a/drivers/gpu/drm/i915/gt/intel_context.h b/drivers/gpu/drm/i915/gt/intel_context.h
index dd742ac2fbdb..68b3d317d959 100644
--- a/drivers/gpu/drm/i915/gt/intel_context.h
+++ b/drivers/gpu/drm/i915/gt/intel_context.h
@@ -12,6 +12,7 @@
 #include "i915_active.h"
 #include "intel_context_types.h"
 #include "intel_engine_types.h"
+#include "intel_ring_types.h"
 #include "intel_timeline_types.h"
 
 void intel_context_init(struct intel_context *ce,
diff --git a/drivers/gpu/drm/i915/gt/intel_context_types.h b/drivers/gpu/drm/i915/gt/intel_context_types.h
index bf9cedfccbf0..6959b05ae5f8 100644
--- a/drivers/gpu/drm/i915/gt/intel_context_types.h
+++ b/drivers/gpu/drm/i915/gt/intel_context_types.h
@@ -58,6 +58,7 @@ struct intel_context {
 
 	u32 *lrc_reg_state;
 	u64 lrc_desc;
+	u32 tag; /* cookie passed to HW to track this context on submission */
 
 	unsigned int active_count; /* protected by timeline->mutex */
 
diff --git a/drivers/gpu/drm/i915/gt/intel_engine.h b/drivers/gpu/drm/i915/gt/intel_engine.h
index d3c6993f4f46..bc3b72bfa9e3 100644
--- a/drivers/gpu/drm/i915/gt/intel_engine.h
+++ b/drivers/gpu/drm/i915/gt/intel_engine.h
@@ -19,6 +19,7 @@
 #include "intel_workarounds.h"
 
 struct drm_printer;
+struct intel_gt;
 
 /* Early gen2 devices have a cacheline of just 32 bytes, using 64 is overkill,
  * but keeps the logic simple. Indeed, the whole purpose of this macro is just
@@ -89,38 +90,6 @@ struct drm_printer;
 /* seqno size is actually only a uint32, but since we plan to use MI_FLUSH_DW to
  * do the writes, and that must have qw aligned offsets, simply pretend it's 8b.
  */
-enum intel_engine_hangcheck_action {
-	ENGINE_IDLE = 0,
-	ENGINE_WAIT,
-	ENGINE_ACTIVE_SEQNO,
-	ENGINE_ACTIVE_HEAD,
-	ENGINE_ACTIVE_SUBUNITS,
-	ENGINE_WAIT_KICK,
-	ENGINE_DEAD,
-};
-
-static inline const char *
-hangcheck_action_to_str(const enum intel_engine_hangcheck_action a)
-{
-	switch (a) {
-	case ENGINE_IDLE:
-		return "idle";
-	case ENGINE_WAIT:
-		return "wait";
-	case ENGINE_ACTIVE_SEQNO:
-		return "active seqno";
-	case ENGINE_ACTIVE_HEAD:
-		return "active head";
-	case ENGINE_ACTIVE_SUBUNITS:
-		return "active subunits";
-	case ENGINE_WAIT_KICK:
-		return "wait kick";
-	case ENGINE_DEAD:
-		return "dead";
-	}
-
-	return "unknown";
-}
 
 static inline unsigned int
 execlists_num_ports(const struct intel_engine_execlists * const execlists)
@@ -136,6 +105,20 @@ execlists_active(const struct intel_engine_execlists *execlists)
 	return READ_ONCE(*execlists->active);
 }
 
+static inline void
+execlists_active_lock_bh(struct intel_engine_execlists *execlists)
+{
+	local_bh_disable(); /* prevent local softirq and lock recursion */
+	tasklet_lock(&execlists->tasklet);
+}
+
+static inline void
+execlists_active_unlock_bh(struct intel_engine_execlists *execlists)
+{
+	tasklet_unlock(&execlists->tasklet);
+	local_bh_enable(); /* restore softirq, and kick ksoftirqd! */
+}
+
 struct i915_request *
 execlists_unwind_incomplete_requests(struct intel_engine_execlists *execlists);
 
@@ -192,126 +175,13 @@ intel_write_status_page(struct intel_engine_cs *engine, int reg, u32 value)
 #define I915_HWS_CSB_WRITE_INDEX	0x1f
 #define CNL_HWS_CSB_WRITE_INDEX		0x2f
 
-struct intel_ring *
-intel_engine_create_ring(struct intel_engine_cs *engine, int size);
-int intel_ring_pin(struct intel_ring *ring);
-void intel_ring_reset(struct intel_ring *ring, u32 tail);
-unsigned int intel_ring_update_space(struct intel_ring *ring);
-void intel_ring_unpin(struct intel_ring *ring);
-void intel_ring_free(struct kref *ref);
-
-static inline struct intel_ring *intel_ring_get(struct intel_ring *ring)
-{
-	kref_get(&ring->ref);
-	return ring;
-}
-
-static inline void intel_ring_put(struct intel_ring *ring)
-{
-	kref_put(&ring->ref, intel_ring_free);
-}
-
 void intel_engine_stop(struct intel_engine_cs *engine);
 void intel_engine_cleanup(struct intel_engine_cs *engine);
 
-int __must_check intel_ring_cacheline_align(struct i915_request *rq);
-
-u32 __must_check *intel_ring_begin(struct i915_request *rq, unsigned int n);
-
-static inline void intel_ring_advance(struct i915_request *rq, u32 *cs)
-{
-	/* Dummy function.
-	 *
-	 * This serves as a placeholder in the code so that the reader
-	 * can compare against the preceding intel_ring_begin() and
-	 * check that the number of dwords emitted matches the space
-	 * reserved for the command packet (i.e. the value passed to
-	 * intel_ring_begin()).
-	 */
-	GEM_BUG_ON((rq->ring->vaddr + rq->ring->emit) != cs);
-}
-
-static inline u32 intel_ring_wrap(const struct intel_ring *ring, u32 pos)
-{
-	return pos & (ring->size - 1);
-}
-
-static inline bool
-intel_ring_offset_valid(const struct intel_ring *ring,
-			unsigned int pos)
-{
-	if (pos & -ring->size) /* must be strictly within the ring */
-		return false;
-
-	if (!IS_ALIGNED(pos, 8)) /* must be qword aligned */
-		return false;
-
-	return true;
-}
-
-static inline u32 intel_ring_offset(const struct i915_request *rq, void *addr)
-{
-	/* Don't write ring->size (equivalent to 0) as that hangs some GPUs. */
-	u32 offset = addr - rq->ring->vaddr;
-	GEM_BUG_ON(offset > rq->ring->size);
-	return intel_ring_wrap(rq->ring, offset);
-}
-
-static inline void
-assert_ring_tail_valid(const struct intel_ring *ring, unsigned int tail)
-{
-	GEM_BUG_ON(!intel_ring_offset_valid(ring, tail));
-
-	/*
-	 * "Ring Buffer Use"
-	 *	Gen2 BSpec "1. Programming Environment" / 1.4.4.6
-	 *	Gen3 BSpec "1c Memory Interface Functions" / 2.3.4.5
-	 *	Gen4+ BSpec "1c Memory Interface and Command Stream" / 5.3.4.5
-	 * "If the Ring Buffer Head Pointer and the Tail Pointer are on the
-	 * same cacheline, the Head Pointer must not be greater than the Tail
-	 * Pointer."
-	 *
-	 * We use ring->head as the last known location of the actual RING_HEAD,
-	 * it may have advanced but in the worst case it is equally the same
-	 * as ring->head and so we should never program RING_TAIL to advance
-	 * into the same cacheline as ring->head.
-	 */
-#define cacheline(a) round_down(a, CACHELINE_BYTES)
-	GEM_BUG_ON(cacheline(tail) == cacheline(ring->head) &&
-		   tail < ring->head);
-#undef cacheline
-}
-
-static inline unsigned int
-intel_ring_set_tail(struct intel_ring *ring, unsigned int tail)
-{
-	/* Whilst writes to the tail are strictly order, there is no
-	 * serialisation between readers and the writers. The tail may be
-	 * read by i915_request_retire() just as it is being updated
-	 * by execlists, as although the breadcrumb is complete, the context
-	 * switch hasn't been seen.
-	 */
-	assert_ring_tail_valid(ring, tail);
-	ring->tail = tail;
-	return tail;
-}
-
-static inline unsigned int
-__intel_ring_space(unsigned int head, unsigned int tail, unsigned int size)
-{
-	/*
-	 * "If the Ring Buffer Head Pointer and the Tail Pointer are on the
-	 * same cacheline, the Head Pointer must not be greater than the Tail
-	 * Pointer."
-	 */
-	GEM_BUG_ON(!is_power_of_2(size));
-	return (head - tail - CACHELINE_BYTES) & (size - 1);
-}
-
-int intel_engines_init_mmio(struct drm_i915_private *i915);
-int intel_engines_setup(struct drm_i915_private *i915);
-int intel_engines_init(struct drm_i915_private *i915);
-void intel_engines_cleanup(struct drm_i915_private *i915);
+int intel_engines_init_mmio(struct intel_gt *gt);
+int intel_engines_setup(struct intel_gt *gt);
+int intel_engines_init(struct intel_gt *gt);
+void intel_engines_cleanup(struct intel_gt *gt);
 
 int intel_engine_init_common(struct intel_engine_cs *engine);
 void intel_engine_cleanup_common(struct intel_engine_cs *engine);
@@ -335,7 +205,6 @@ void intel_engine_init_execlists(struct intel_engine_cs *engine);
 void intel_engine_init_breadcrumbs(struct intel_engine_cs *engine);
 void intel_engine_fini_breadcrumbs(struct intel_engine_cs *engine);
 
-void intel_engine_signal_breadcrumbs(struct intel_engine_cs *engine);
 void intel_engine_disarm_breadcrumbs(struct intel_engine_cs *engine);
 
 static inline void
@@ -408,8 +277,9 @@ static inline void __intel_engine_reset(struct intel_engine_cs *engine,
 	engine->serial++; /* contexts lost */
 }
 
-bool intel_engine_is_idle(struct intel_engine_cs *engine);
 bool intel_engines_are_idle(struct intel_gt *gt);
+bool intel_engine_is_idle(struct intel_engine_cs *engine);
+void intel_engine_flush_submission(struct intel_engine_cs *engine);
 
 void intel_engines_reset_default_submission(struct intel_gt *gt);
 
@@ -420,61 +290,6 @@ void intel_engine_dump(struct intel_engine_cs *engine,
 		       struct drm_printer *m,
 		       const char *header, ...);
 
-static inline void intel_engine_context_in(struct intel_engine_cs *engine)
-{
-	unsigned long flags;
-
-	if (READ_ONCE(engine->stats.enabled) == 0)
-		return;
-
-	write_seqlock_irqsave(&engine->stats.lock, flags);
-
-	if (engine->stats.enabled > 0) {
-		if (engine->stats.active++ == 0)
-			engine->stats.start = ktime_get();
-		GEM_BUG_ON(engine->stats.active == 0);
-	}
-
-	write_sequnlock_irqrestore(&engine->stats.lock, flags);
-}
-
-static inline void intel_engine_context_out(struct intel_engine_cs *engine)
-{
-	unsigned long flags;
-
-	if (READ_ONCE(engine->stats.enabled) == 0)
-		return;
-
-	write_seqlock_irqsave(&engine->stats.lock, flags);
-
-	if (engine->stats.enabled > 0) {
-		ktime_t last;
-
-		if (engine->stats.active && --engine->stats.active == 0) {
-			/*
-			 * Decrement the active context count and in case GPU
-			 * is now idle add up to the running total.
-			 */
-			last = ktime_sub(ktime_get(), engine->stats.start);
-
-			engine->stats.total = ktime_add(engine->stats.total,
-							last);
-		} else if (engine->stats.active == 0) {
-			/*
-			 * After turning on engine stats, context out might be
-			 * the first event in which case we account from the
-			 * time stats gathering was turned on.
-			 */
-			last = ktime_sub(ktime_get(), engine->stats.enabled_at);
-
-			engine->stats.total = ktime_add(engine->stats.total,
-							last);
-		}
-	}
-
-	write_sequnlock_irqrestore(&engine->stats.lock, flags);
-}
-
 int intel_enable_engine_stats(struct intel_engine_cs *engine);
 void intel_disable_engine_stats(struct intel_engine_cs *engine);
 
@@ -511,4 +326,22 @@ void intel_engine_init_active(struct intel_engine_cs *engine,
 #define ENGINE_MOCK	1
 #define ENGINE_VIRTUAL	2
 
+static inline bool
+intel_engine_has_preempt_reset(const struct intel_engine_cs *engine)
+{
+	if (!IS_ACTIVE(CONFIG_DRM_I915_PREEMPT_TIMEOUT))
+		return false;
+
+	return intel_engine_has_preemption(engine);
+}
+
+static inline bool
+intel_engine_has_timeslices(const struct intel_engine_cs *engine)
+{
+	if (!IS_ACTIVE(CONFIG_DRM_I915_TIMESLICE_DURATION))
+		return false;
+
+	return intel_engine_has_semaphores(engine);
+}
+
 #endif /* _INTEL_RINGBUFFER_H_ */
diff --git a/drivers/gpu/drm/i915/gt/intel_engine_cs.c b/drivers/gpu/drm/i915/gt/intel_engine_cs.c
index 82630db0394b..5ca3ec911e50 100644
--- a/drivers/gpu/drm/i915/gt/intel_engine_cs.c
+++ b/drivers/gpu/drm/i915/gt/intel_engine_cs.c
@@ -37,6 +37,7 @@
 #include "intel_context.h"
 #include "intel_lrc.h"
 #include "intel_reset.h"
+#include "intel_ring.h"
 
 /* Haswell does have the CXT_SIZE register however it does not appear to be
  * valid. Now, docs explain in dwords what is in the context object. The full
@@ -277,6 +278,9 @@ static int intel_engine_setup(struct intel_gt *gt, enum intel_engine_id id)
 	BUILD_BUG_ON(MAX_ENGINE_CLASS >= BIT(GEN11_ENGINE_CLASS_WIDTH));
 	BUILD_BUG_ON(MAX_ENGINE_INSTANCE >= BIT(GEN11_ENGINE_INSTANCE_WIDTH));
 
+	if (GEM_DEBUG_WARN_ON(id >= ARRAY_SIZE(gt->engine)))
+		return -EINVAL;
+
 	if (GEM_DEBUG_WARN_ON(info->class > MAX_ENGINE_CLASS))
 		return -EINVAL;
 
@@ -293,6 +297,7 @@ static int intel_engine_setup(struct intel_gt *gt, enum intel_engine_id id)
 	BUILD_BUG_ON(BITS_PER_TYPE(engine->mask) < I915_NUM_ENGINES);
 
 	engine->id = id;
+	engine->legacy_idx = INVALID_ENGINE;
 	engine->mask = BIT(id);
 	engine->i915 = gt->i915;
 	engine->gt = gt;
@@ -304,6 +309,15 @@ static int intel_engine_setup(struct intel_gt *gt, enum intel_engine_id id)
 	engine->instance = info->instance;
 	__sprint_engine_name(engine);
 
+	engine->props.heartbeat_interval_ms =
+		CONFIG_DRM_I915_HEARTBEAT_INTERVAL;
+	engine->props.preempt_timeout_ms =
+		CONFIG_DRM_I915_PREEMPT_TIMEOUT;
+	engine->props.stop_timeout_ms =
+		CONFIG_DRM_I915_STOP_TIMEOUT;
+	engine->props.timeslice_duration_ms =
+		CONFIG_DRM_I915_TIMESLICE_DURATION;
+
 	/*
 	 * To be overridden by the backend on setup. However to facilitate
 	 * cleanup on error during setup, we always provide the destroy vfunc.
@@ -328,6 +342,7 @@ static int intel_engine_setup(struct intel_gt *gt, enum intel_engine_id id)
 	intel_engine_sanitize_mmio(engine);
 
 	gt->engine_class[info->class][info->instance] = engine;
+	gt->engine[id] = engine;
 
 	intel_engine_add_user(engine);
 	gt->i915->engine[id] = engine;
@@ -365,38 +380,40 @@ static void __setup_engine_capabilities(struct intel_engine_cs *engine)
 	}
 }
 
-static void intel_setup_engine_capabilities(struct drm_i915_private *i915)
+static void intel_setup_engine_capabilities(struct intel_gt *gt)
 {
 	struct intel_engine_cs *engine;
 	enum intel_engine_id id;
 
-	for_each_engine(engine, i915, id)
+	for_each_engine(engine, gt, id)
 		__setup_engine_capabilities(engine);
 }
 
 /**
  * intel_engines_cleanup() - free the resources allocated for Command Streamers
- * @i915: the i915 devic
+ * @gt: pointer to struct intel_gt
  */
-void intel_engines_cleanup(struct drm_i915_private *i915)
+void intel_engines_cleanup(struct intel_gt *gt)
 {
 	struct intel_engine_cs *engine;
 	enum intel_engine_id id;
 
-	for_each_engine(engine, i915, id) {
+	for_each_engine(engine, gt, id) {
 		engine->destroy(engine);
-		i915->engine[id] = NULL;
+		gt->engine[id] = NULL;
+		gt->i915->engine[id] = NULL;
 	}
 }
 
 /**
  * intel_engines_init_mmio() - allocate and prepare the Engine Command Streamers
- * @i915: the i915 device
+ * @gt: pointer to struct intel_gt
  *
  * Return: non-zero if the initialization failed.
  */
-int intel_engines_init_mmio(struct drm_i915_private *i915)
+int intel_engines_init_mmio(struct intel_gt *gt)
 {
+	struct drm_i915_private *i915 = gt->i915;
 	struct intel_device_info *device_info = mkwrite_device_info(i915);
 	const unsigned int engine_mask = INTEL_INFO(i915)->engine_mask;
 	unsigned int mask = 0;
@@ -414,7 +431,7 @@ int intel_engines_init_mmio(struct drm_i915_private *i915)
 		if (!HAS_ENGINE(i915, i))
 			continue;
 
-		err = intel_engine_setup(&i915->gt, i);
+		err = intel_engine_setup(gt, i);
 		if (err)
 			goto cleanup;
 
@@ -431,36 +448,36 @@ int intel_engines_init_mmio(struct drm_i915_private *i915)
 
 	RUNTIME_INFO(i915)->num_engines = hweight32(mask);
 
-	intel_gt_check_and_clear_faults(&i915->gt);
+	intel_gt_check_and_clear_faults(gt);
 
-	intel_setup_engine_capabilities(i915);
+	intel_setup_engine_capabilities(gt);
 
 	return 0;
 
 cleanup:
-	intel_engines_cleanup(i915);
+	intel_engines_cleanup(gt);
 	return err;
 }
 
 /**
  * intel_engines_init() - init the Engine Command Streamers
- * @i915: i915 device private
+ * @gt: pointer to struct intel_gt
  *
  * Return: non-zero if the initialization failed.
  */
-int intel_engines_init(struct drm_i915_private *i915)
+int intel_engines_init(struct intel_gt *gt)
 {
 	int (*init)(struct intel_engine_cs *engine);
 	struct intel_engine_cs *engine;
 	enum intel_engine_id id;
 	int err;
 
-	if (HAS_EXECLISTS(i915))
+	if (HAS_EXECLISTS(gt->i915))
 		init = intel_execlists_submission_init;
 	else
 		init = intel_ring_submission_init;
 
-	for_each_engine(engine, i915, id) {
+	for_each_engine(engine, gt, id) {
 		err = init(engine);
 		if (err)
 			goto cleanup;
@@ -469,7 +486,7 @@ int intel_engines_init(struct drm_i915_private *i915)
 	return 0;
 
 cleanup:
-	intel_engines_cleanup(i915);
+	intel_engines_cleanup(gt);
 	return err;
 }
 
@@ -513,7 +530,7 @@ static int pin_ggtt_status_page(struct intel_engine_cs *engine,
 	unsigned int flags;
 
 	flags = PIN_GLOBAL;
-	if (!HAS_LLC(engine->i915))
+	if (!HAS_LLC(engine->i915) && i915_ggtt_has_aperture(engine->gt->ggtt))
 		/*
 		 * On g33, we cannot place HWS above 256MiB, so
 		 * restrict its pinning to the low mappable arena.
@@ -597,7 +614,6 @@ static int intel_engine_setup_common(struct intel_engine_cs *engine)
 	intel_engine_init_active(engine, ENGINE_PHYSICAL);
 	intel_engine_init_breadcrumbs(engine);
 	intel_engine_init_execlists(engine);
-	intel_engine_init_hangcheck(engine);
 	intel_engine_init_cmd_parser(engine);
 	intel_engine_init__pm(engine);
 
@@ -616,26 +632,26 @@ static int intel_engine_setup_common(struct intel_engine_cs *engine)
 
 /**
  * intel_engines_setup- setup engine state not requiring hw access
- * @i915: Device to setup.
+ * @gt: pointer to struct intel_gt
  *
  * Initializes engine structure members shared between legacy and execlists
  * submission modes which do not require hardware access.
  *
  * Typically done early in the submission mode specific engine setup stage.
  */
-int intel_engines_setup(struct drm_i915_private *i915)
+int intel_engines_setup(struct intel_gt *gt)
 {
 	int (*setup)(struct intel_engine_cs *engine);
 	struct intel_engine_cs *engine;
 	enum intel_engine_id id;
 	int err;
 
-	if (HAS_EXECLISTS(i915))
+	if (HAS_EXECLISTS(gt->i915))
 		setup = intel_execlists_submission_setup;
 	else
 		setup = intel_ring_submission_setup;
 
-	for_each_engine(engine, i915, id) {
+	for_each_engine(engine, gt, id) {
 		err = intel_engine_setup_common(engine);
 		if (err)
 			goto cleanup;
@@ -653,7 +669,7 @@ int intel_engines_setup(struct drm_i915_private *i915)
 	return 0;
 
 cleanup:
-	intel_engines_cleanup(i915);
+	intel_engines_cleanup(gt);
 	return err;
 }
 
@@ -680,6 +696,8 @@ static int measure_breadcrumb_dw(struct intel_engine_cs *engine)
 				engine->status_page.vma))
 		goto out_frame;
 
+	mutex_lock(&frame->timeline.mutex);
+
 	frame->ring.vaddr = frame->cs;
 	frame->ring.size = sizeof(frame->cs);
 	frame->ring.effective_size = frame->ring.size;
@@ -688,18 +706,22 @@ static int measure_breadcrumb_dw(struct intel_engine_cs *engine)
 	frame->rq.i915 = engine->i915;
 	frame->rq.engine = engine;
 	frame->rq.ring = &frame->ring;
-	frame->rq.timeline = &frame->timeline;
+	rcu_assign_pointer(frame->rq.timeline, &frame->timeline);
 
 	dw = intel_timeline_pin(&frame->timeline);
 	if (dw < 0)
 		goto out_timeline;
 
+	spin_lock_irq(&engine->active.lock);
 	dw = engine->emit_fini_breadcrumb(&frame->rq, frame->cs) - frame->cs;
+	spin_unlock_irq(&engine->active.lock);
+
 	GEM_BUG_ON(dw & 1); /* RING_TAIL must be qword aligned */
 
 	intel_timeline_unpin(&frame->timeline);
 
 out_timeline:
+	mutex_unlock(&frame->timeline.mutex);
 	intel_timeline_fini(&frame->timeline);
 out_frame:
 	kfree(frame);
@@ -730,6 +752,7 @@ intel_engine_init_active(struct intel_engine_cs *engine, unsigned int subclass)
 static struct intel_context *
 create_kernel_context(struct intel_engine_cs *engine)
 {
+	static struct lock_class_key kernel;
 	struct intel_context *ce;
 	int err;
 
@@ -745,6 +768,14 @@ create_kernel_context(struct intel_engine_cs *engine)
 		return ERR_PTR(err);
 	}
 
+	/*
+	 * Give our perma-pinned kernel timelines a separate lockdep class,
+	 * so that we can use them from within the normal user timelines
+	 * should we need to inject GPU operations during their request
+	 * construction.
+	 */
+	lockdep_set_class(&ce->timeline->mutex, &kernel);
+
 	return ce;
 }
 
@@ -814,8 +845,10 @@ void intel_engine_cleanup_common(struct intel_engine_cs *engine)
 	if (engine->default_state)
 		i915_gem_object_put(engine->default_state);
 
-	intel_context_unpin(engine->kernel_context);
-	intel_context_put(engine->kernel_context);
+	if (engine->kernel_context) {
+		intel_context_unpin(engine->kernel_context);
+		intel_context_put(engine->kernel_context);
+	}
 	GEM_BUG_ON(!llist_empty(&engine->barrier_tasks));
 
 	intel_wa_list_free(&engine->ctx_wa_list);
@@ -851,6 +884,21 @@ u64 intel_engine_get_last_batch_head(const struct intel_engine_cs *engine)
 	return bbaddr;
 }
 
+static unsigned long stop_timeout(const struct intel_engine_cs *engine)
+{
+	if (in_atomic() || irqs_disabled()) /* inside atomic preempt-reset? */
+		return 0;
+
+	/*
+	 * If we are doing a normal GPU reset, we can take our time and allow
+	 * the engine to quiesce. We've stopped submission to the engine, and
+	 * if we wait long enough an innocent context should complete and
+	 * leave the engine idle. So they should not be caught unaware by
+	 * the forthcoming GPU reset (which usually follows the stop_cs)!
+	 */
+	return READ_ONCE(engine->props.stop_timeout_ms);
+}
+
 int intel_engine_stop_cs(struct intel_engine_cs *engine)
 {
 	struct intel_uncore *uncore = engine->uncore;
@@ -868,7 +916,7 @@ int intel_engine_stop_cs(struct intel_engine_cs *engine)
 	err = 0;
 	if (__intel_wait_for_register_fw(uncore,
 					 mode, MODE_IDLE, MODE_IDLE,
-					 1000, 0,
+					 1000, stop_timeout(engine),
 					 NULL)) {
 		GEM_TRACE("%s: timed out on STOP_RING -> IDLE\n", engine->name);
 		err = -ETIMEDOUT;
@@ -948,6 +996,7 @@ void intel_engine_get_instdone(struct intel_engine_cs *engine,
 			       struct intel_instdone *instdone)
 {
 	struct drm_i915_private *i915 = engine->i915;
+	const struct sseu_dev_info *sseu = &RUNTIME_INFO(i915)->sseu;
 	struct intel_uncore *uncore = engine->uncore;
 	u32 mmio_base = engine->mmio_base;
 	int slice;
@@ -965,7 +1014,7 @@ void intel_engine_get_instdone(struct intel_engine_cs *engine,
 
 		instdone->slice_common =
 			intel_uncore_read(uncore, GEN7_SC_INSTDONE);
-		for_each_instdone_slice_subslice(i915, slice, subslice) {
+		for_each_instdone_slice_subslice(i915, sseu, slice, subslice) {
 			instdone->sampler[slice][subslice] =
 				read_subslice_reg(engine, slice, subslice,
 						  GEN7_SAMPLER_INSTDONE);
@@ -1031,6 +1080,25 @@ static bool ring_is_idle(struct intel_engine_cs *engine)
 	return idle;
 }
 
+void intel_engine_flush_submission(struct intel_engine_cs *engine)
+{
+	struct tasklet_struct *t = &engine->execlists.tasklet;
+
+	if (__tasklet_is_scheduled(t)) {
+		local_bh_disable();
+		if (tasklet_trylock(t)) {
+			/* Must wait for any GPU reset in progress. */
+			if (__tasklet_is_enabled(t))
+				t->func(t->data);
+			tasklet_unlock(t);
+		}
+		local_bh_enable();
+	}
+
+	/* Otherwise flush the tasklet if it was running on another cpu */
+	tasklet_unlock_wait(t);
+}
+
 /**
  * intel_engine_is_idle() - Report if the engine has finished process all work
  * @engine: the intel_engine_cs
@@ -1049,21 +1117,9 @@ bool intel_engine_is_idle(struct intel_engine_cs *engine)
 
 	/* Waiting to drain ELSP? */
 	if (execlists_active(&engine->execlists)) {
-		struct tasklet_struct *t = &engine->execlists.tasklet;
-
 		synchronize_hardirq(engine->i915->drm.pdev->irq);
 
-		local_bh_disable();
-		if (tasklet_trylock(t)) {
-			/* Must wait for any GPU reset in progress. */
-			if (__tasklet_is_enabled(t))
-				t->func(t->data);
-			tasklet_unlock(t);
-		}
-		local_bh_enable();
-
-		/* Otherwise flush the tasklet if it was on another cpu */
-		tasklet_unlock_wait(t);
+		intel_engine_flush_submission(engine);
 
 		if (execlists_active(&engine->execlists))
 			return false;
@@ -1093,7 +1149,7 @@ bool intel_engines_are_idle(struct intel_gt *gt)
 	if (!READ_ONCE(gt->awake))
 		return true;
 
-	for_each_engine(engine, gt->i915, id) {
+	for_each_engine(engine, gt, id) {
 		if (!intel_engine_is_idle(engine))
 			return false;
 	}
@@ -1106,7 +1162,7 @@ void intel_engines_reset_default_submission(struct intel_gt *gt)
 	struct intel_engine_cs *engine;
 	enum intel_engine_id id;
 
-	for_each_engine(engine, gt->i915, id)
+	for_each_engine(engine, gt, id)
 		engine->set_default_submission(engine);
 }
 
@@ -1118,6 +1174,8 @@ bool intel_engine_can_store_dword(struct intel_engine_cs *engine)
 	case 3:
 		/* maybe only uses physical not virtual addresses */
 		return !(IS_I915G(engine->i915) || IS_I915GM(engine->i915));
+	case 4:
+		return !IS_I965G(engine->i915); /* who knows! */
 	case 6:
 		return engine->class != VIDEO_DECODE_CLASS; /* b0rked */
 	default:
@@ -1193,13 +1251,43 @@ static void hexdump(struct drm_printer *m, const void *buf, size_t len)
 	}
 }
 
+static struct intel_timeline *get_timeline(struct i915_request *rq)
+{
+	struct intel_timeline *tl;
+
+	/*
+	 * Even though we are holding the engine->active.lock here, there
+	 * is no control over the submission queue per-se and we are
+	 * inspecting the active state at a random point in time, with an
+	 * unknown queue. Play safe and make sure the timeline remains valid.
+	 * (Only being used for pretty printing, one extra kref shouldn't
+	 * cause a camel stampede!)
+	 */
+	rcu_read_lock();
+	tl = rcu_dereference(rq->timeline);
+	if (!kref_get_unless_zero(&tl->kref))
+		tl = NULL;
+	rcu_read_unlock();
+
+	return tl;
+}
+
+static const char *repr_timer(const struct timer_list *t)
+{
+	if (!READ_ONCE(t->expires))
+		return "inactive";
+
+	if (timer_pending(t))
+		return "active";
+
+	return "expired";
+}
+
 static void intel_engine_print_registers(struct intel_engine_cs *engine,
 					 struct drm_printer *m)
 {
 	struct drm_i915_private *dev_priv = engine->i915;
-	const struct intel_engine_execlists * const execlists =
-		&engine->execlists;
-	unsigned long flags;
+	struct intel_engine_execlists * const execlists = &engine->execlists;
 	u64 addr;
 
 	if (engine->id == RENDER_CLASS && IS_GEN_RANGE(dev_priv, 4, 7))
@@ -1256,19 +1344,21 @@ static void intel_engine_print_registers(struct intel_engine_cs *engine,
 		unsigned int idx;
 		u8 read, write;
 
-		drm_printf(m, "\tExeclist status: 0x%08x %08x, entries %u\n",
-			   ENGINE_READ(engine, RING_EXECLIST_STATUS_LO),
-			   ENGINE_READ(engine, RING_EXECLIST_STATUS_HI),
-			   num_entries);
+		drm_printf(m, "\tExeclist tasklet queued? %s (%s), preempt? %s, timeslice? %s\n",
+			   yesno(test_bit(TASKLET_STATE_SCHED,
+					  &engine->execlists.tasklet.state)),
+			   enableddisabled(!atomic_read(&engine->execlists.tasklet.count)),
+			   repr_timer(&engine->execlists.preempt),
+			   repr_timer(&engine->execlists.timer));
 
 		read = execlists->csb_head;
 		write = READ_ONCE(*execlists->csb_write);
 
-		drm_printf(m, "\tExeclist CSB read %d, write %d, tasklet queued? %s (%s)\n",
-			   read, write,
-			   yesno(test_bit(TASKLET_STATE_SCHED,
-					  &engine->execlists.tasklet.state)),
-			   enableddisabled(!atomic_read(&engine->execlists.tasklet.count)));
+		drm_printf(m, "\tExeclist status: 0x%08x %08x; CSB read:%d, write:%d, entries:%d\n",
+			   ENGINE_READ(engine, RING_EXECLIST_STATUS_LO),
+			   ENGINE_READ(engine, RING_EXECLIST_STATUS_HI),
+			   read, write, num_entries);
+
 		if (read >= num_entries)
 			read = 0;
 		if (write >= num_entries)
@@ -1281,35 +1371,47 @@ static void intel_engine_print_registers(struct intel_engine_cs *engine,
 				   idx, hws[idx * 2], hws[idx * 2 + 1]);
 		}
 
-		spin_lock_irqsave(&engine->active.lock, flags);
+		execlists_active_lock_bh(execlists);
+		rcu_read_lock();
 		for (port = execlists->active; (rq = *port); port++) {
 			char hdr[80];
 			int len;
 
 			len = snprintf(hdr, sizeof(hdr),
-				       "\t\tActive[%d: ",
+				       "\t\tActive[%d]: ",
 				       (int)(port - execlists->active));
-			if (!i915_request_signaled(rq))
+			if (!i915_request_signaled(rq)) {
+				struct intel_timeline *tl = get_timeline(rq);
+
 				len += snprintf(hdr + len, sizeof(hdr) - len,
 						"ring:{start:%08x, hwsp:%08x, seqno:%08x}, ",
 						i915_ggtt_offset(rq->ring->vma),
-						rq->timeline->hwsp_offset,
+						tl ? tl->hwsp_offset : 0,
 						hwsp_seqno(rq));
+
+				if (tl)
+					intel_timeline_put(tl);
+			}
 			snprintf(hdr + len, sizeof(hdr) - len, "rq: ");
 			print_request(m, rq, hdr);
 		}
 		for (port = execlists->pending; (rq = *port); port++) {
+			struct intel_timeline *tl = get_timeline(rq);
 			char hdr[80];
 
 			snprintf(hdr, sizeof(hdr),
 				 "\t\tPending[%d] ring:{start:%08x, hwsp:%08x, seqno:%08x}, rq: ",
 				 (int)(port - execlists->pending),
 				 i915_ggtt_offset(rq->ring->vma),
-				 rq->timeline->hwsp_offset,
+				 tl ? tl->hwsp_offset : 0,
 				 hwsp_seqno(rq));
 			print_request(m, rq, hdr);
+
+			if (tl)
+				intel_timeline_put(tl);
 		}
-		spin_unlock_irqrestore(&engine->active.lock, flags);
+		rcu_read_unlock();
+		execlists_active_unlock_bh(execlists);
 	} else if (INTEL_GEN(dev_priv) > 6) {
 		drm_printf(m, "\tPP_DIR_BASE: 0x%08x\n",
 			   ENGINE_READ(engine, RING_PP_DIR_BASE));
@@ -1374,8 +1476,13 @@ void intel_engine_dump(struct intel_engine_cs *engine,
 		drm_printf(m, "*** WEDGED ***\n");
 
 	drm_printf(m, "\tAwake? %d\n", atomic_read(&engine->wakeref.count));
-	drm_printf(m, "\tHangcheck: %d ms ago\n",
-		   jiffies_to_msecs(jiffies - engine->hangcheck.action_timestamp));
+
+	rcu_read_lock();
+	rq = READ_ONCE(engine->heartbeat.systole);
+	if (rq)
+		drm_printf(m, "\tHeartbeat: %d ms ago\n",
+			   jiffies_to_msecs(jiffies - rq->emitted_jiffies));
+	rcu_read_unlock();
 	drm_printf(m, "\tReset count: %d (global %d)\n",
 		   i915_reset_engine_count(error, engine),
 		   i915_reset_count(error));
@@ -1385,6 +1492,8 @@ void intel_engine_dump(struct intel_engine_cs *engine,
 	spin_lock_irqsave(&engine->active.lock, flags);
 	rq = intel_engine_find_active_request(engine);
 	if (rq) {
+		struct intel_timeline *tl = get_timeline(rq);
+
 		print_request(m, rq, "\t\tactive ");
 
 		drm_printf(m, "\t\tring->start:  0x%08x\n",
@@ -1397,18 +1506,27 @@ void intel_engine_dump(struct intel_engine_cs *engine,
 			   rq->ring->emit);
 		drm_printf(m, "\t\tring->space:  0x%08x\n",
 			   rq->ring->space);
-		drm_printf(m, "\t\tring->hwsp:   0x%08x\n",
-			   rq->timeline->hwsp_offset);
+
+		if (tl) {
+			drm_printf(m, "\t\tring->hwsp:   0x%08x\n",
+				   tl->hwsp_offset);
+			intel_timeline_put(tl);
+		}
 
 		print_request_ring(m, rq);
+
+		if (rq->hw_context->lrc_reg_state) {
+			drm_printf(m, "Logical Ring Context:\n");
+			hexdump(m, rq->hw_context->lrc_reg_state, PAGE_SIZE);
+		}
 	}
 	spin_unlock_irqrestore(&engine->active.lock, flags);
 
 	drm_printf(m, "\tMMIO base:  0x%08x\n", engine->mmio_base);
-	wakeref = intel_runtime_pm_get_if_in_use(&engine->i915->runtime_pm);
+	wakeref = intel_runtime_pm_get_if_in_use(engine->uncore->rpm);
 	if (wakeref) {
 		intel_engine_print_registers(engine, m);
-		intel_runtime_pm_put(&engine->i915->runtime_pm, wakeref);
+		intel_runtime_pm_put(engine->uncore->rpm, wakeref);
 	} else {
 		drm_printf(m, "\tDevice is asleep; skipping register dump\n");
 	}
@@ -1440,8 +1558,8 @@ int intel_enable_engine_stats(struct intel_engine_cs *engine)
 	if (!intel_engine_supports_stats(engine))
 		return -ENODEV;
 
-	spin_lock_irqsave(&engine->active.lock, flags);
-	write_seqlock(&engine->stats.lock);
+	execlists_active_lock_bh(execlists);
+	write_seqlock_irqsave(&engine->stats.lock, flags);
 
 	if (unlikely(engine->stats.enabled == ~0)) {
 		err = -EBUSY;
@@ -1469,8 +1587,8 @@ int intel_enable_engine_stats(struct intel_engine_cs *engine)
 	}
 
 unlock:
-	write_sequnlock(&engine->stats.lock);
-	spin_unlock_irqrestore(&engine->active.lock, flags);
+	write_sequnlock_irqrestore(&engine->stats.lock, flags);
+	execlists_active_unlock_bh(execlists);
 
 	return err;
 }
diff --git a/drivers/gpu/drm/i915/gt/intel_engine_heartbeat.c b/drivers/gpu/drm/i915/gt/intel_engine_heartbeat.c
new file mode 100644
index 000000000000..06aa14c7aa8c
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/intel_engine_heartbeat.c
@@ -0,0 +1,234 @@
+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2019 Intel Corporation
+ */
+
+#include "i915_request.h"
+
+#include "intel_context.h"
+#include "intel_engine_heartbeat.h"
+#include "intel_engine_pm.h"
+#include "intel_engine.h"
+#include "intel_gt.h"
+#include "intel_reset.h"
+
+/*
+ * While the engine is active, we send a periodic pulse along the engine
+ * to check on its health and to flush any idle-barriers. If that request
+ * is stuck, and we fail to preempt it, we declare the engine hung and
+ * issue a reset -- in the hope that restores progress.
+ */
+
+static bool next_heartbeat(struct intel_engine_cs *engine)
+{
+	long delay;
+
+	delay = READ_ONCE(engine->props.heartbeat_interval_ms);
+	if (!delay)
+		return false;
+
+	delay = msecs_to_jiffies_timeout(delay);
+	if (delay >= HZ)
+		delay = round_jiffies_up_relative(delay);
+	schedule_delayed_work(&engine->heartbeat.work, delay);
+
+	return true;
+}
+
+static void idle_pulse(struct intel_engine_cs *engine, struct i915_request *rq)
+{
+	engine->wakeref_serial = READ_ONCE(engine->serial) + 1;
+	i915_request_add_active_barriers(rq);
+}
+
+static void show_heartbeat(const struct i915_request *rq,
+			   struct intel_engine_cs *engine)
+{
+	struct drm_printer p = drm_debug_printer("heartbeat");
+
+	intel_engine_dump(engine, &p,
+			  "%s heartbeat {prio:%d} not ticking\n",
+			  engine->name,
+			  rq->sched.attr.priority);
+}
+
+static void heartbeat(struct work_struct *wrk)
+{
+	struct i915_sched_attr attr = {
+		.priority = I915_USER_PRIORITY(I915_PRIORITY_MIN),
+	};
+	struct intel_engine_cs *engine =
+		container_of(wrk, typeof(*engine), heartbeat.work.work);
+	struct intel_context *ce = engine->kernel_context;
+	struct i915_request *rq;
+
+	if (!intel_engine_pm_get_if_awake(engine))
+		return;
+
+	rq = engine->heartbeat.systole;
+	if (rq && i915_request_completed(rq)) {
+		i915_request_put(rq);
+		engine->heartbeat.systole = NULL;
+	}
+
+	if (intel_gt_is_wedged(engine->gt))
+		goto out;
+
+	if (engine->heartbeat.systole) {
+		if (engine->schedule &&
+		    rq->sched.attr.priority < I915_PRIORITY_BARRIER) {
+			/*
+			 * Gradually raise the priority of the heartbeat to
+			 * give high priority work [which presumably desires
+			 * low latency and no jitter] the chance to naturally
+			 * complete before being preempted.
+			 */
+			attr.priority = I915_PRIORITY_MASK;
+			if (rq->sched.attr.priority >= attr.priority)
+				attr.priority |= I915_USER_PRIORITY(I915_PRIORITY_HEARTBEAT);
+			if (rq->sched.attr.priority >= attr.priority)
+				attr.priority = I915_PRIORITY_BARRIER;
+
+			local_bh_disable();
+			engine->schedule(rq, &attr);
+			local_bh_enable();
+		} else {
+			if (IS_ENABLED(CONFIG_DRM_I915_DEBUG_GEM))
+				show_heartbeat(rq, engine);
+
+			intel_gt_handle_error(engine->gt, engine->mask,
+					      I915_ERROR_CAPTURE,
+					      "stopped heartbeat on %s",
+					      engine->name);
+		}
+		goto out;
+	}
+
+	if (engine->wakeref_serial == engine->serial)
+		goto out;
+
+	mutex_lock(&ce->timeline->mutex);
+
+	intel_context_enter(ce);
+	rq = __i915_request_create(ce, GFP_NOWAIT | __GFP_NOWARN);
+	intel_context_exit(ce);
+	if (IS_ERR(rq))
+		goto unlock;
+
+	idle_pulse(engine, rq);
+	if (i915_modparams.enable_hangcheck)
+		engine->heartbeat.systole = i915_request_get(rq);
+
+	__i915_request_commit(rq);
+	__i915_request_queue(rq, &attr);
+
+unlock:
+	mutex_unlock(&ce->timeline->mutex);
+out:
+	if (!next_heartbeat(engine))
+		i915_request_put(fetch_and_zero(&engine->heartbeat.systole));
+	intel_engine_pm_put(engine);
+}
+
+void intel_engine_unpark_heartbeat(struct intel_engine_cs *engine)
+{
+	if (!IS_ACTIVE(CONFIG_DRM_I915_HEARTBEAT_INTERVAL))
+		return;
+
+	next_heartbeat(engine);
+}
+
+void intel_engine_park_heartbeat(struct intel_engine_cs *engine)
+{
+	if (cancel_delayed_work(&engine->heartbeat.work))
+		i915_request_put(fetch_and_zero(&engine->heartbeat.systole));
+}
+
+void intel_engine_init_heartbeat(struct intel_engine_cs *engine)
+{
+	INIT_DELAYED_WORK(&engine->heartbeat.work, heartbeat);
+}
+
+int intel_engine_set_heartbeat(struct intel_engine_cs *engine,
+			       unsigned long delay)
+{
+	int err;
+
+	/* Send one last pulse before to cleanup persistent hogs */
+	if (!delay && IS_ACTIVE(CONFIG_DRM_I915_PREEMPT_TIMEOUT)) {
+		err = intel_engine_pulse(engine);
+		if (err)
+			return err;
+	}
+
+	WRITE_ONCE(engine->props.heartbeat_interval_ms, delay);
+
+	if (intel_engine_pm_get_if_awake(engine)) {
+		if (delay)
+			intel_engine_unpark_heartbeat(engine);
+		else
+			intel_engine_park_heartbeat(engine);
+		intel_engine_pm_put(engine);
+	}
+
+	return 0;
+}
+
+int intel_engine_pulse(struct intel_engine_cs *engine)
+{
+	struct i915_sched_attr attr = { .priority = I915_PRIORITY_BARRIER };
+	struct intel_context *ce = engine->kernel_context;
+	struct i915_request *rq;
+	int err = 0;
+
+	if (!intel_engine_has_preemption(engine))
+		return -ENODEV;
+
+	if (!intel_engine_pm_get_if_awake(engine))
+		return 0;
+
+	if (mutex_lock_interruptible(&ce->timeline->mutex))
+		goto out_rpm;
+
+	intel_context_enter(ce);
+	rq = __i915_request_create(ce, GFP_NOWAIT | __GFP_NOWARN);
+	intel_context_exit(ce);
+	if (IS_ERR(rq)) {
+		err = PTR_ERR(rq);
+		goto out_unlock;
+	}
+
+	rq->flags |= I915_REQUEST_SENTINEL;
+	idle_pulse(engine, rq);
+
+	__i915_request_commit(rq);
+	__i915_request_queue(rq, &attr);
+
+out_unlock:
+	mutex_unlock(&ce->timeline->mutex);
+out_rpm:
+	intel_engine_pm_put(engine);
+	return err;
+}
+
+int intel_engine_flush_barriers(struct intel_engine_cs *engine)
+{
+	struct i915_request *rq;
+
+	if (llist_empty(&engine->barrier_tasks))
+		return 0;
+
+	rq = i915_request_create(engine->kernel_context);
+	if (IS_ERR(rq))
+		return PTR_ERR(rq);
+
+	idle_pulse(engine, rq);
+	i915_request_add(rq);
+
+	return 0;
+}
+
+#if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
+#include "selftest_engine_heartbeat.c"
+#endif
diff --git a/drivers/gpu/drm/i915/gt/intel_engine_heartbeat.h b/drivers/gpu/drm/i915/gt/intel_engine_heartbeat.h
new file mode 100644
index 000000000000..a7b8c0f9e005
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/intel_engine_heartbeat.h
@@ -0,0 +1,23 @@
+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2019 Intel Corporation
+ */
+
+#ifndef INTEL_ENGINE_HEARTBEAT_H
+#define INTEL_ENGINE_HEARTBEAT_H
+
+struct intel_engine_cs;
+
+void intel_engine_init_heartbeat(struct intel_engine_cs *engine);
+
+int intel_engine_set_heartbeat(struct intel_engine_cs *engine,
+			       unsigned long delay);
+
+void intel_engine_park_heartbeat(struct intel_engine_cs *engine);
+void intel_engine_unpark_heartbeat(struct intel_engine_cs *engine);
+
+int intel_engine_pulse(struct intel_engine_cs *engine);
+int intel_engine_flush_barriers(struct intel_engine_cs *engine);
+
+#endif /* INTEL_ENGINE_HEARTBEAT_H */
diff --git a/drivers/gpu/drm/i915/gt/intel_engine_pm.c b/drivers/gpu/drm/i915/gt/intel_engine_pm.c
index 65b5ca74b394..874d82677179 100644
--- a/drivers/gpu/drm/i915/gt/intel_engine_pm.c
+++ b/drivers/gpu/drm/i915/gt/intel_engine_pm.c
@@ -7,10 +7,13 @@
 #include "i915_drv.h"
 
 #include "intel_engine.h"
+#include "intel_engine_heartbeat.h"
 #include "intel_engine_pm.h"
 #include "intel_engine_pool.h"
 #include "intel_gt.h"
 #include "intel_gt_pm.h"
+#include "intel_rc6.h"
+#include "intel_ring.h"
 
 static int __engine_unpark(struct intel_wakeref *wf)
 {
@@ -33,7 +36,7 @@ static int __engine_unpark(struct intel_wakeref *wf)
 	if (engine->unpark)
 		engine->unpark(engine);
 
-	intel_engine_init_hangcheck(engine);
+	intel_engine_unpark_heartbeat(engine);
 	return 0;
 }
 
@@ -52,7 +55,7 @@ static inline unsigned long __timeline_mark_lock(struct intel_context *ce)
 static inline void __timeline_mark_unlock(struct intel_context *ce,
 					  unsigned long flags)
 {
-	mutex_release(&ce->timeline->mutex.dep_map, 0, _THIS_IP_);
+	mutex_release(&ce->timeline->mutex.dep_map, _THIS_IP_);
 	local_irq_restore(flags);
 }
 
@@ -103,14 +106,14 @@ static bool switch_to_kernel_context(struct intel_engine_cs *engine)
 		/* Context switch failed, hope for the best! Maybe reset? */
 		goto out_unlock;
 
-	intel_timeline_enter(rq->timeline);
+	intel_timeline_enter(i915_request_timeline(rq));
 
 	/* Check again on the next retirement. */
 	engine->wakeref_serial = engine->serial + 1;
 	i915_request_add_active_barriers(rq);
 
 	/* Install ourselves as a preemption barrier */
-	rq->sched.attr.priority = I915_PRIORITY_UNPREEMPTABLE;
+	rq->sched.attr.priority = I915_PRIORITY_BARRIER;
 	__i915_request_commit(rq);
 
 	/* Release our exclusive hold on the engine */
@@ -123,6 +126,19 @@ out_unlock:
 	return result;
 }
 
+static void call_idle_barriers(struct intel_engine_cs *engine)
+{
+	struct llist_node *node, *next;
+
+	llist_for_each_safe(node, next, llist_del_all(&engine->barrier_tasks)) {
+		struct dma_fence_cb *cb =
+			container_of((struct list_head *)node,
+				     typeof(*cb), node);
+
+		cb->func(NULL, cb);
+	}
+}
+
 static int __engine_park(struct intel_wakeref *wf)
 {
 	struct intel_engine_cs *engine =
@@ -142,6 +158,9 @@ static int __engine_park(struct intel_wakeref *wf)
 
 	GEM_TRACE("%s\n", engine->name);
 
+	call_idle_barriers(engine); /* cleanup after wedging */
+
+	intel_engine_park_heartbeat(engine);
 	intel_engine_disarm_breadcrumbs(engine);
 	intel_engine_pool_park(&engine->pool);
 
@@ -169,9 +188,10 @@ static const struct intel_wakeref_ops wf_ops = {
 
 void intel_engine_init__pm(struct intel_engine_cs *engine)
 {
-	struct intel_runtime_pm *rpm = &engine->i915->runtime_pm;
+	struct intel_runtime_pm *rpm = engine->uncore->rpm;
 
 	intel_wakeref_init(&engine->wakeref, rpm, &wf_ops);
+	intel_engine_init_heartbeat(engine);
 }
 
 #if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
diff --git a/drivers/gpu/drm/i915/gt/intel_engine_pool.c b/drivers/gpu/drm/i915/gt/intel_engine_pool.c
index 4cd54c569911..397186818305 100644
--- a/drivers/gpu/drm/i915/gt/intel_engine_pool.c
+++ b/drivers/gpu/drm/i915/gt/intel_engine_pool.c
@@ -61,6 +61,7 @@ static int pool_active(struct i915_active *ref)
 	return 0;
 }
 
+__i915_active_call
 static void pool_retire(struct i915_active *ref)
 {
 	struct intel_engine_pool_node *node =
@@ -94,7 +95,7 @@ node_create(struct intel_engine_pool *pool, size_t sz)
 		return ERR_PTR(-ENOMEM);
 
 	node->pool = pool;
-	i915_active_init(engine->i915, &node->active, pool_active, pool_retire);
+	i915_active_init(&node->active, pool_active, pool_retire);
 
 	obj = i915_gem_object_create_internal(engine->i915, sz);
 	if (IS_ERR(obj)) {
@@ -103,13 +104,25 @@ node_create(struct intel_engine_pool *pool, size_t sz)
 		return ERR_CAST(obj);
 	}
 
+	i915_gem_object_set_readonly(obj);
+
 	node->obj = obj;
 	return node;
 }
 
+static struct intel_engine_pool *lookup_pool(struct intel_engine_cs *engine)
+{
+	if (intel_engine_is_virtual(engine))
+		engine = intel_virtual_engine_get_sibling(engine, 0);
+
+	GEM_BUG_ON(!engine);
+	return &engine->pool;
+}
+
 struct intel_engine_pool_node *
-intel_engine_pool_get(struct intel_engine_pool *pool, size_t size)
+intel_engine_get_pool(struct intel_engine_cs *engine, size_t size)
 {
+	struct intel_engine_pool *pool = lookup_pool(engine);
 	struct intel_engine_pool_node *node;
 	struct list_head *list;
 	unsigned long flags;
diff --git a/drivers/gpu/drm/i915/gt/intel_engine_pool.h b/drivers/gpu/drm/i915/gt/intel_engine_pool.h
index 8d069efd9457..1bd89cadc3b7 100644
--- a/drivers/gpu/drm/i915/gt/intel_engine_pool.h
+++ b/drivers/gpu/drm/i915/gt/intel_engine_pool.h
@@ -12,13 +12,13 @@
 #include "i915_request.h"
 
 struct intel_engine_pool_node *
-intel_engine_pool_get(struct intel_engine_pool *pool, size_t size);
+intel_engine_get_pool(struct intel_engine_cs *engine, size_t size);
 
 static inline int
 intel_engine_pool_mark_active(struct intel_engine_pool_node *node,
 			      struct i915_request *rq)
 {
-	return i915_active_ref(&node->active, rq->timeline, rq);
+	return i915_active_add_request(&node->active, rq);
 }
 
 static inline void
diff --git a/drivers/gpu/drm/i915/gt/intel_engine_types.h b/drivers/gpu/drm/i915/gt/intel_engine_types.h
index a82cea95c2f2..758f0e8ec672 100644
--- a/drivers/gpu/drm/i915/gt/intel_engine_types.h
+++ b/drivers/gpu/drm/i915/gt/intel_engine_types.h
@@ -15,6 +15,7 @@
 #include <linux/rbtree.h>
 #include <linux/timer.h>
 #include <linux/types.h>
+#include <linux/workqueue.h>
 
 #include "i915_gem.h"
 #include "i915_pmu.h"
@@ -58,6 +59,7 @@ struct i915_gem_context;
 struct i915_request;
 struct i915_sched_attr;
 struct intel_gt;
+struct intel_ring;
 struct intel_uncore;
 
 typedef u8 intel_engine_mask_t;
@@ -76,40 +78,6 @@ struct intel_instdone {
 	u32 row[I915_MAX_SLICES][I915_MAX_SUBSLICES];
 };
 
-struct intel_engine_hangcheck {
-	u64 acthd;
-	u32 last_ring;
-	u32 last_head;
-	unsigned long action_timestamp;
-	struct intel_instdone instdone;
-};
-
-struct intel_ring {
-	struct kref ref;
-	struct i915_vma *vma;
-	void *vaddr;
-
-	/*
-	 * As we have two types of rings, one global to the engine used
-	 * by ringbuffer submission and those that are exclusive to a
-	 * context used by execlists, we have to play safe and allow
-	 * atomic updates to the pin_count. However, the actual pinning
-	 * of the context is either done during initialisation for
-	 * ringbuffer submission or serialised as part of the context
-	 * pinning for execlists, and so we do not need a mutex ourselves
-	 * to serialise intel_ring_pin/intel_ring_unpin.
-	 */
-	atomic_t pin_count;
-
-	u32 head;
-	u32 tail;
-	u32 emit;
-
-	u32 space;
-	u32 size;
-	u32 effective_size;
-};
-
 /*
  * we use a single page to load ctx workarounds so all of these
  * values are referred in terms of dwords
@@ -148,6 +116,7 @@ enum intel_engine_id {
 	VECS1,
 #define _VECS(n) (VECS0 + (n))
 	I915_NUM_ENGINES
+#define INVALID_ENGINE ((enum intel_engine_id)-1)
 };
 
 struct st_preempt_hang {
@@ -174,6 +143,11 @@ struct intel_engine_execlists {
 	struct timer_list timer;
 
 	/**
+	 * @preempt: reset the current context if it fails to give way
+	 */
+	struct timer_list preempt;
+
+	/**
 	 * @default_priolist: priority list for I915_PRIORITY_NORMAL
 	 */
 	struct i915_priolist default_priolist;
@@ -303,10 +277,12 @@ struct intel_engine_cs {
 	u8 uabi_class;
 	u8 uabi_instance;
 
+	u32 uabi_capabilities;
 	u32 context_size;
 	u32 mmio_base;
 
-	u32 uabi_capabilities;
+	unsigned int context_tag;
+#define NUM_CONTEXT_TAG roundup_pow_of_two(2 * EXECLIST_MAX_PORTS)
 
 	struct rb_node uabi_node;
 
@@ -323,6 +299,11 @@ struct intel_engine_cs {
 
 	intel_engine_mask_t saturated; /* submitting semaphores too late? */
 
+	struct {
+		struct delayed_work work;
+		struct i915_request *systole;
+	} heartbeat;
+
 	unsigned long serial;
 
 	unsigned long wakeref_serial;
@@ -473,14 +454,14 @@ struct intel_engine_cs {
 	/* status_notifier: list of callbacks for context-switch changes */
 	struct atomic_notifier_head context_status_notifier;
 
-	struct intel_engine_hangcheck hangcheck;
-
-#define I915_ENGINE_NEEDS_CMD_PARSER BIT(0)
+#define I915_ENGINE_USING_CMD_PARSER BIT(0)
 #define I915_ENGINE_SUPPORTS_STATS   BIT(1)
 #define I915_ENGINE_HAS_PREEMPTION   BIT(2)
 #define I915_ENGINE_HAS_SEMAPHORES   BIT(3)
 #define I915_ENGINE_NEEDS_BREADCRUMB_TASKLET BIT(4)
 #define I915_ENGINE_IS_VIRTUAL       BIT(5)
+#define I915_ENGINE_HAS_RELATIVE_MMIO BIT(6)
+#define I915_ENGINE_REQUIRES_CMD_PARSER BIT(7)
 	unsigned int flags;
 
 	/*
@@ -538,12 +519,25 @@ struct intel_engine_cs {
 		 */
 		ktime_t total;
 	} stats;
+
+	struct {
+		unsigned long heartbeat_interval_ms;
+		unsigned long preempt_timeout_ms;
+		unsigned long stop_timeout_ms;
+		unsigned long timeslice_duration_ms;
+	} props;
 };
 
 static inline bool
-intel_engine_needs_cmd_parser(const struct intel_engine_cs *engine)
+intel_engine_using_cmd_parser(const struct intel_engine_cs *engine)
 {
-	return engine->flags & I915_ENGINE_NEEDS_CMD_PARSER;
+	return engine->flags & I915_ENGINE_USING_CMD_PARSER;
+}
+
+static inline bool
+intel_engine_requires_cmd_parser(const struct intel_engine_cs *engine)
+{
+	return engine->flags & I915_ENGINE_REQUIRES_CMD_PARSER;
 }
 
 static inline bool
@@ -576,20 +570,24 @@ intel_engine_is_virtual(const struct intel_engine_cs *engine)
 	return engine->flags & I915_ENGINE_IS_VIRTUAL;
 }
 
-#define instdone_slice_mask(dev_priv__) \
-	(IS_GEN(dev_priv__, 7) ? \
-	 1 : RUNTIME_INFO(dev_priv__)->sseu.slice_mask)
+static inline bool
+intel_engine_has_relative_mmio(const struct intel_engine_cs * const engine)
+{
+	return engine->flags & I915_ENGINE_HAS_RELATIVE_MMIO;
+}
 
-#define instdone_subslice_mask(dev_priv__) \
-	(IS_GEN(dev_priv__, 7) ? \
-	 1 : RUNTIME_INFO(dev_priv__)->sseu.subslice_mask[0])
+#define instdone_has_slice(dev_priv___, sseu___, slice___) \
+	((IS_GEN(dev_priv___, 7) ? 1 : ((sseu___)->slice_mask)) & BIT(slice___))
 
-#define for_each_instdone_slice_subslice(dev_priv__, slice__, subslice__) \
-	for ((slice__) = 0, (subslice__) = 0; \
-	     (slice__) < I915_MAX_SLICES; \
-	     (subslice__) = ((subslice__) + 1) < I915_MAX_SUBSLICES ? (subslice__) + 1 : 0, \
-	       (slice__) += ((subslice__) == 0)) \
-		for_each_if((BIT(slice__) & instdone_slice_mask(dev_priv__)) && \
-			    (BIT(subslice__) & instdone_subslice_mask(dev_priv__)))
+#define instdone_has_subslice(dev_priv__, sseu__, slice__, subslice__) \
+	(IS_GEN(dev_priv__, 7) ? (1 & BIT(subslice__)) : \
+	 intel_sseu_has_subslice(sseu__, 0, subslice__))
 
+#define for_each_instdone_slice_subslice(dev_priv_, sseu_, slice_, subslice_) \
+	for ((slice_) = 0, (subslice_) = 0; (slice_) < I915_MAX_SLICES; \
+	     (subslice_) = ((subslice_) + 1) % I915_MAX_SUBSLICES, \
+	     (slice_) += ((subslice_) == 0)) \
+		for_each_if((instdone_has_slice(dev_priv_, sseu_, slice_)) && \
+			    (instdone_has_subslice(dev_priv_, sseu_, slice_, \
+						    subslice_)))
 #endif /* __INTEL_ENGINE_TYPES_H__ */
diff --git a/drivers/gpu/drm/i915/gt/intel_engine_user.c b/drivers/gpu/drm/i915/gt/intel_engine_user.c
index 77cd5de83930..7f7150a733f4 100644
--- a/drivers/gpu/drm/i915/gt/intel_engine_user.c
+++ b/drivers/gpu/drm/i915/gt/intel_engine_user.c
@@ -160,10 +160,10 @@ static int legacy_ring_idx(const struct legacy_ring *ring)
 	};
 
 	if (GEM_DEBUG_WARN_ON(ring->class >= ARRAY_SIZE(map)))
-		return -1;
+		return INVALID_ENGINE;
 
 	if (GEM_DEBUG_WARN_ON(ring->instance >= map[ring->class].max))
-		return -1;
+		return INVALID_ENGINE;
 
 	return map[ring->class].base + ring->instance;
 }
@@ -171,23 +171,15 @@ static int legacy_ring_idx(const struct legacy_ring *ring)
 static void add_legacy_ring(struct legacy_ring *ring,
 			    struct intel_engine_cs *engine)
 {
-	int idx;
-
 	if (engine->gt != ring->gt || engine->class != ring->class) {
 		ring->gt = engine->gt;
 		ring->class = engine->class;
 		ring->instance = 0;
 	}
 
-	idx = legacy_ring_idx(ring);
-	if (unlikely(idx == -1))
-		return;
-
-	GEM_BUG_ON(idx >= ARRAY_SIZE(ring->gt->engine));
-	ring->gt->engine[idx] = engine;
-	ring->instance++;
-
-	engine->legacy_idx = idx;
+	engine->legacy_idx = legacy_ring_idx(ring);
+	if (engine->legacy_idx != INVALID_ENGINE)
+		ring->instance++;
 }
 
 void intel_engines_driver_register(struct drm_i915_private *i915)
diff --git a/drivers/gpu/drm/i915/gt/intel_gpu_commands.h b/drivers/gpu/drm/i915/gt/intel_gpu_commands.h
index 86e00a2db8a4..4294f146f13c 100644
--- a/drivers/gpu/drm/i915/gt/intel_gpu_commands.h
+++ b/drivers/gpu/drm/i915/gt/intel_gpu_commands.h
@@ -112,6 +112,7 @@
 #define MI_SEMAPHORE_SIGNAL	MI_INSTR(0x1b, 0) /* GEN8+ */
 #define   MI_SEMAPHORE_TARGET(engine)	((engine)<<15)
 #define MI_SEMAPHORE_WAIT	MI_INSTR(0x1c, 2) /* GEN8+ */
+#define MI_SEMAPHORE_WAIT_TOKEN	MI_INSTR(0x1c, 3) /* GEN12+ */
 #define   MI_SEMAPHORE_POLL		(1 << 15)
 #define   MI_SEMAPHORE_SAD_GT_SDD	(0 << 12)
 #define   MI_SEMAPHORE_SAD_GTE_SDD	(1 << 12)
@@ -119,6 +120,8 @@
 #define   MI_SEMAPHORE_SAD_LTE_SDD	(3 << 12)
 #define   MI_SEMAPHORE_SAD_EQ_SDD	(4 << 12)
 #define   MI_SEMAPHORE_SAD_NEQ_SDD	(5 << 12)
+#define   MI_SEMAPHORE_TOKEN_MASK	REG_GENMASK(9, 5)
+#define   MI_SEMAPHORE_TOKEN_SHIFT	5
 #define MI_STORE_DWORD_IMM	MI_INSTR(0x20, 1)
 #define MI_STORE_DWORD_IMM_GEN4	MI_INSTR(0x20, 2)
 #define   MI_MEM_VIRTUAL	(1 << 22) /* 945,g33,965 */
@@ -132,7 +135,10 @@
  *   address/value pairs. Don't overdue it, though, x <= 2^4 must hold!
  */
 #define MI_LOAD_REGISTER_IMM(x)	MI_INSTR(0x22, 2*(x)-1)
+/* Gen11+. addr = base + (ctx_restore ? offset & GENMASK(12,2) : offset) */
+#define   MI_LRI_CS_MMIO		(1<<19)
 #define   MI_LRI_FORCE_POSTED		(1<<12)
+#define MI_LOAD_REGISTER_IMM_MAX_REGS (126)
 #define MI_STORE_REGISTER_MEM        MI_INSTR(0x24, 1)
 #define MI_STORE_REGISTER_MEM_GEN8   MI_INSTR(0x24, 2)
 #define   MI_SRM_LRM_GLOBAL_GTT		(1<<22)
@@ -147,6 +153,7 @@
 #define   MI_FLUSH_DW_USE_PPGTT		(0<<2)
 #define MI_LOAD_REGISTER_MEM	   MI_INSTR(0x29, 1)
 #define MI_LOAD_REGISTER_MEM_GEN8  MI_INSTR(0x29, 2)
+#define MI_LOAD_REGISTER_REG    MI_INSTR(0x2A, 1)
 #define MI_BATCH_BUFFER		MI_INSTR(0x30, 1)
 #define   MI_BATCH_NON_SECURE		(1)
 /* for snb/ivb/vlv this also means "batch in ppgtt" when ppgtt is enabled. */
@@ -156,7 +163,8 @@
 #define MI_BATCH_BUFFER_START	MI_INSTR(0x31, 0)
 #define   MI_BATCH_GTT		    (2<<6) /* aliased with (1<<7) on gen4 */
 #define MI_BATCH_BUFFER_START_GEN8	MI_INSTR(0x31, 1)
-#define   MI_BATCH_RESOURCE_STREAMER (1<<10)
+#define   MI_BATCH_RESOURCE_STREAMER REG_BIT(10)
+#define   MI_BATCH_PREDICATE         REG_BIT(15) /* HSW+ on RCS only*/
 
 /*
  * 3D instructions used by the kernel
@@ -217,6 +225,7 @@
 #define   PIPE_CONTROL_CS_STALL				(1<<20)
 #define   PIPE_CONTROL_TLB_INVALIDATE			(1<<18)
 #define   PIPE_CONTROL_MEDIA_STATE_CLEAR		(1<<16)
+#define   PIPE_CONTROL_WRITE_TIMESTAMP			(3<<14)
 #define   PIPE_CONTROL_QW_WRITE				(1<<14)
 #define   PIPE_CONTROL_POST_SYNC_OP_MASK                (3<<14)
 #define   PIPE_CONTROL_DEPTH_STALL			(1<<13)
@@ -224,7 +233,9 @@
 #define   PIPE_CONTROL_RENDER_TARGET_CACHE_FLUSH	(1<<12) /* gen6+ */
 #define   PIPE_CONTROL_INSTRUCTION_CACHE_INVALIDATE	(1<<11) /* MBZ on ILK */
 #define   PIPE_CONTROL_TEXTURE_CACHE_INVALIDATE		(1<<10) /* GM45+ only */
+#define   PIPE_CONTROL_L3_RO_CACHE_INVALIDATE		REG_BIT(10) /* gen12 */
 #define   PIPE_CONTROL_INDIRECT_STATE_DISABLE		(1<<9)
+#define   PIPE_CONTROL_HDC_PIPELINE_FLUSH		REG_BIT(9)  /* gen12 */
 #define   PIPE_CONTROL_NOTIFY				(1<<8)
 #define   PIPE_CONTROL_FLUSH_ENABLE			(1<<7) /* gen7+ */
 #define   PIPE_CONTROL_DC_FLUSH_ENABLE			(1<<5)
@@ -235,6 +246,29 @@
 #define   PIPE_CONTROL_DEPTH_CACHE_FLUSH		(1<<0)
 #define   PIPE_CONTROL_GLOBAL_GTT (1<<2) /* in addr dword */
 
+#define MI_MATH(x)			MI_INSTR(0x1a, (x) - 1)
+#define MI_MATH_INSTR(opcode, op1, op2) ((opcode) << 20 | (op1) << 10 | (op2))
+/* Opcodes for MI_MATH_INSTR */
+#define   MI_MATH_NOOP			MI_MATH_INSTR(0x000, 0x0, 0x0)
+#define   MI_MATH_LOAD(op1, op2)	MI_MATH_INSTR(0x080, op1, op2)
+#define   MI_MATH_LOADINV(op1, op2)	MI_MATH_INSTR(0x480, op1, op2)
+#define   MI_MATH_LOAD0(op1)		MI_MATH_INSTR(0x081, op1)
+#define   MI_MATH_LOAD1(op1)		MI_MATH_INSTR(0x481, op1)
+#define   MI_MATH_ADD			MI_MATH_INSTR(0x100, 0x0, 0x0)
+#define   MI_MATH_SUB			MI_MATH_INSTR(0x101, 0x0, 0x0)
+#define   MI_MATH_AND			MI_MATH_INSTR(0x102, 0x0, 0x0)
+#define   MI_MATH_OR			MI_MATH_INSTR(0x103, 0x0, 0x0)
+#define   MI_MATH_XOR			MI_MATH_INSTR(0x104, 0x0, 0x0)
+#define   MI_MATH_STORE(op1, op2)	MI_MATH_INSTR(0x180, op1, op2)
+#define   MI_MATH_STOREINV(op1, op2)	MI_MATH_INSTR(0x580, op1, op2)
+/* Registers used as operands in MI_MATH_INSTR */
+#define   MI_MATH_REG(x)		(x)
+#define   MI_MATH_REG_SRCA		0x20
+#define   MI_MATH_REG_SRCB		0x21
+#define   MI_MATH_REG_ACCU		0x31
+#define   MI_MATH_REG_ZF		0x32
+#define   MI_MATH_REG_CF		0x33
+
 /*
  * Commands used only by the command parser
  */
@@ -251,7 +285,6 @@
 #define MI_CLFLUSH              MI_INSTR(0x27, 0)
 #define MI_REPORT_PERF_COUNT    MI_INSTR(0x28, 0)
 #define   MI_REPORT_PERF_COUNT_GGTT (1<<0)
-#define MI_LOAD_REGISTER_REG    MI_INSTR(0x2A, 0)
 #define MI_RS_STORE_DATA_IMM    MI_INSTR(0x2B, 0)
 #define MI_LOAD_URB_MEM         MI_INSTR(0x2C, 0)
 #define MI_STORE_URB_MEM        MI_INSTR(0x2D, 0)
diff --git a/drivers/gpu/drm/i915/gt/intel_gt.c b/drivers/gpu/drm/i915/gt/intel_gt.c
index d48ec9a76ed1..4c26daf7ee46 100644
--- a/drivers/gpu/drm/i915/gt/intel_gt.c
+++ b/drivers/gpu/drm/i915/gt/intel_gt.c
@@ -6,7 +6,12 @@
 #include "i915_drv.h"
 #include "intel_gt.h"
 #include "intel_gt_pm.h"
+#include "intel_gt_requests.h"
+#include "intel_mocs.h"
+#include "intel_rc6.h"
+#include "intel_rps.h"
 #include "intel_uncore.h"
+#include "intel_pm.h"
 
 void intel_gt_init_early(struct intel_gt *gt, struct drm_i915_private *i915)
 {
@@ -18,15 +23,108 @@ void intel_gt_init_early(struct intel_gt *gt, struct drm_i915_private *i915)
 	INIT_LIST_HEAD(&gt->closed_vma);
 	spin_lock_init(&gt->closed_lock);
 
-	intel_gt_init_hangcheck(gt);
 	intel_gt_init_reset(gt);
+	intel_gt_init_requests(gt);
 	intel_gt_pm_init_early(gt);
+
+	intel_rps_init_early(&gt->rps);
 	intel_uc_init_early(&gt->uc);
 }
 
-void intel_gt_init_hw(struct drm_i915_private *i915)
+void intel_gt_init_hw_early(struct intel_gt *gt, struct i915_ggtt *ggtt)
+{
+	gt->ggtt = ggtt;
+
+	intel_gt_sanitize(gt, false);
+}
+
+static void init_unused_ring(struct intel_gt *gt, u32 base)
 {
-	i915->gt.ggtt = &i915->ggtt;
+	struct intel_uncore *uncore = gt->uncore;
+
+	intel_uncore_write(uncore, RING_CTL(base), 0);
+	intel_uncore_write(uncore, RING_HEAD(base), 0);
+	intel_uncore_write(uncore, RING_TAIL(base), 0);
+	intel_uncore_write(uncore, RING_START(base), 0);
+}
+
+static void init_unused_rings(struct intel_gt *gt)
+{
+	struct drm_i915_private *i915 = gt->i915;
+
+	if (IS_I830(i915)) {
+		init_unused_ring(gt, PRB1_BASE);
+		init_unused_ring(gt, SRB0_BASE);
+		init_unused_ring(gt, SRB1_BASE);
+		init_unused_ring(gt, SRB2_BASE);
+		init_unused_ring(gt, SRB3_BASE);
+	} else if (IS_GEN(i915, 2)) {
+		init_unused_ring(gt, SRB0_BASE);
+		init_unused_ring(gt, SRB1_BASE);
+	} else if (IS_GEN(i915, 3)) {
+		init_unused_ring(gt, PRB1_BASE);
+		init_unused_ring(gt, PRB2_BASE);
+	}
+}
+
+int intel_gt_init_hw(struct intel_gt *gt)
+{
+	struct drm_i915_private *i915 = gt->i915;
+	struct intel_uncore *uncore = gt->uncore;
+	int ret;
+
+	BUG_ON(!i915->kernel_context);
+	ret = intel_gt_terminally_wedged(gt);
+	if (ret)
+		return ret;
+
+	gt->last_init_time = ktime_get();
+
+	/* Double layer security blanket, see i915_gem_init() */
+	intel_uncore_forcewake_get(uncore, FORCEWAKE_ALL);
+
+	if (HAS_EDRAM(i915) && INTEL_GEN(i915) < 9)
+		intel_uncore_rmw(uncore, HSW_IDICR, 0, IDIHASHMSK(0xf));
+
+	if (IS_HASWELL(i915))
+		intel_uncore_write(uncore,
+				   MI_PREDICATE_RESULT_2,
+				   IS_HSW_GT3(i915) ?
+				   LOWER_SLICE_ENABLED : LOWER_SLICE_DISABLED);
+
+	/* Apply the GT workarounds... */
+	intel_gt_apply_workarounds(gt);
+	/* ...and determine whether they are sticking. */
+	intel_gt_verify_workarounds(gt, "init");
+
+	intel_gt_init_swizzling(gt);
+
+	/*
+	 * At least 830 can leave some of the unused rings
+	 * "active" (ie. head != tail) after resume which
+	 * will prevent c3 entry. Makes sure all unused rings
+	 * are totally idle.
+	 */
+	init_unused_rings(gt);
+
+	ret = i915_ppgtt_init_hw(gt);
+	if (ret) {
+		DRM_ERROR("Enabling PPGTT failed (%d)\n", ret);
+		goto out;
+	}
+
+	/* We can't enable contexts until all firmware is loaded */
+	ret = intel_uc_init_hw(&gt->uc);
+	if (ret) {
+		i915_probe_error(i915, "Enabling uc failed (%d)\n", ret);
+		goto out;
+	}
+
+	intel_mocs_init(gt);
+
+out:
+	intel_uncore_forcewake_put(uncore, FORCEWAKE_ALL);
+	return ret;
 }
 
 static void rmw_set(struct intel_uncore *uncore, i915_reg_t reg, u32 set)
@@ -89,7 +187,7 @@ intel_gt_clear_error_registers(struct intel_gt *gt,
 		struct intel_engine_cs *engine;
 		enum intel_engine_id id;
 
-		for_each_engine_masked(engine, i915, engine_mask, id)
+		for_each_engine_masked(engine, gt, engine_mask, id)
 			gen8_clear_engine_error_register(engine);
 	}
 }
@@ -100,7 +198,7 @@ static void gen6_check_faults(struct intel_gt *gt)
 	enum intel_engine_id id;
 	u32 fault;
 
-	for_each_engine(engine, gt->i915, id) {
+	for_each_engine(engine, gt, id) {
 		fault = GEN6_RING_FAULT_REG_READ(engine);
 		if (fault & RING_FAULT_VALID) {
 			DRM_DEBUG_DRIVER("Unexpected fault\n"
@@ -176,7 +274,7 @@ void intel_gt_check_and_clear_faults(struct intel_gt *gt)
 
 void intel_gt_flush_ggtt_writes(struct intel_gt *gt)
 {
-	struct drm_i915_private *i915 = gt->i915;
+	struct intel_uncore *uncore = gt->uncore;
 	intel_wakeref_t wakeref;
 
 	/*
@@ -200,18 +298,18 @@ void intel_gt_flush_ggtt_writes(struct intel_gt *gt)
 
 	wmb();
 
-	if (INTEL_INFO(i915)->has_coherent_ggtt)
+	if (INTEL_INFO(gt->i915)->has_coherent_ggtt)
 		return;
 
 	intel_gt_chipset_flush(gt);
 
-	with_intel_runtime_pm(&i915->runtime_pm, wakeref) {
-		struct intel_uncore *uncore = gt->uncore;
+	with_intel_runtime_pm(uncore->rpm, wakeref) {
+		unsigned long flags;
 
-		spin_lock_irq(&uncore->lock);
+		spin_lock_irqsave(&uncore->lock, flags);
 		intel_uncore_posting_read_fw(uncore,
 					     RING_HEAD(RENDER_RING_BASE));
-		spin_unlock_irq(&uncore->lock);
+		spin_unlock_irqrestore(&uncore->lock, flags);
 	}
 }
 
@@ -222,7 +320,12 @@ void intel_gt_chipset_flush(struct intel_gt *gt)
 		intel_gtt_chipset_flush();
 }
 
-int intel_gt_init_scratch(struct intel_gt *gt, unsigned int size)
+void intel_gt_driver_register(struct intel_gt *gt)
+{
+	intel_rps_driver_register(&gt->rps);
+}
+
+static int intel_gt_init_scratch(struct intel_gt *gt, unsigned int size)
 {
 	struct drm_i915_private *i915 = gt->i915;
 	struct drm_i915_gem_object *obj;
@@ -230,7 +333,7 @@ int intel_gt_init_scratch(struct intel_gt *gt, unsigned int size)
 	int ret;
 
 	obj = i915_gem_object_create_stolen(i915, size);
-	if (!obj)
+	if (IS_ERR(obj))
 		obj = i915_gem_object_create_internal(i915, size);
 	if (IS_ERR(obj)) {
 		DRM_ERROR("Failed to allocate scratch page\n");
@@ -256,11 +359,40 @@ err_unref:
 	return ret;
 }
 
-void intel_gt_fini_scratch(struct intel_gt *gt)
+static void intel_gt_fini_scratch(struct intel_gt *gt)
 {
 	i915_vma_unpin_and_release(&gt->scratch, 0);
 }
 
+int intel_gt_init(struct intel_gt *gt)
+{
+	int err;
+
+	err = intel_gt_init_scratch(gt, IS_GEN(gt->i915, 2) ? SZ_256K : SZ_4K);
+	if (err)
+		return err;
+
+	intel_gt_pm_init(gt);
+
+	return 0;
+}
+
+void intel_gt_driver_remove(struct intel_gt *gt)
+{
+	GEM_BUG_ON(gt->awake);
+}
+
+void intel_gt_driver_unregister(struct intel_gt *gt)
+{
+	intel_rps_driver_unregister(&gt->rps);
+}
+
+void intel_gt_driver_release(struct intel_gt *gt)
+{
+	intel_gt_pm_fini(gt);
+	intel_gt_fini_scratch(gt);
+}
+
 void intel_gt_driver_late_release(struct intel_gt *gt)
 {
 	intel_uc_driver_late_release(&gt->uc);
diff --git a/drivers/gpu/drm/i915/gt/intel_gt.h b/drivers/gpu/drm/i915/gt/intel_gt.h
index 4920cb351f10..5436f8c30708 100644
--- a/drivers/gpu/drm/i915/gt/intel_gt.h
+++ b/drivers/gpu/drm/i915/gt/intel_gt.h
@@ -28,7 +28,14 @@ static inline struct intel_gt *huc_to_gt(struct intel_huc *huc)
 }
 
 void intel_gt_init_early(struct intel_gt *gt, struct drm_i915_private *i915);
-void intel_gt_init_hw(struct drm_i915_private *i915);
+void intel_gt_init_hw_early(struct intel_gt *gt, struct i915_ggtt *ggtt);
+int __must_check intel_gt_init_hw(struct intel_gt *gt);
+int intel_gt_init(struct intel_gt *gt);
+void intel_gt_driver_register(struct intel_gt *gt);
+
+void intel_gt_driver_unregister(struct intel_gt *gt);
+void intel_gt_driver_remove(struct intel_gt *gt);
+void intel_gt_driver_release(struct intel_gt *gt);
 
 void intel_gt_driver_late_release(struct intel_gt *gt);
 
@@ -39,11 +46,6 @@ void intel_gt_clear_error_registers(struct intel_gt *gt,
 void intel_gt_flush_ggtt_writes(struct intel_gt *gt);
 void intel_gt_chipset_flush(struct intel_gt *gt);
 
-void intel_gt_init_hangcheck(struct intel_gt *gt);
-
-int intel_gt_init_scratch(struct intel_gt *gt, unsigned int size);
-void intel_gt_fini_scratch(struct intel_gt *gt);
-
 static inline u32 intel_gt_scratch_offset(const struct intel_gt *gt,
 					  enum intel_gt_scratch_field field)
 {
@@ -55,6 +57,4 @@ static inline bool intel_gt_is_wedged(struct intel_gt *gt)
 	return __intel_reset_failed(&gt->reset);
 }
 
-void intel_gt_queue_hangcheck(struct intel_gt *gt);
-
 #endif /* __INTEL_GT_H__ */
diff --git a/drivers/gpu/drm/i915/gt/intel_gt_irq.c b/drivers/gpu/drm/i915/gt/intel_gt_irq.c
index 34a4fb624bf7..973ee7eded64 100644
--- a/drivers/gpu/drm/i915/gt/intel_gt_irq.c
+++ b/drivers/gpu/drm/i915/gt/intel_gt_irq.c
@@ -11,6 +11,7 @@
 #include "intel_gt.h"
 #include "intel_gt_irq.h"
 #include "intel_uncore.h"
+#include "intel_rps.h"
 
 static void guc_irq_handler(struct intel_guc *guc, u16 iir)
 {
@@ -77,7 +78,7 @@ gen11_other_irq_handler(struct intel_gt *gt, const u8 instance,
 		return guc_irq_handler(&gt->uc.guc, iir);
 
 	if (instance == OTHER_GTPM_INSTANCE)
-		return gen11_rps_irq_handler(gt, iir);
+		return gen11_rps_irq_handler(&gt->rps, iir);
 
 	WARN_ONCE(1, "unhandled other interrupt instance=0x%x, iir=0x%x\n",
 		  instance, iir);
@@ -336,7 +337,7 @@ void gen8_gt_irq_handler(struct intel_gt *gt, u32 master_ctl, u32 gt_iir[4])
 	}
 
 	if (master_ctl & (GEN8_GT_PM_IRQ | GEN8_GT_GUC_IRQ)) {
-		gen6_rps_irq_handler(gt->i915, gt_iir[2]);
+		gen6_rps_irq_handler(&gt->rps, gt_iir[2]);
 		guc_irq_handler(&gt->uc.guc, gt_iir[2] >> 16);
 	}
 }
diff --git a/drivers/gpu/drm/i915/gt/intel_gt_pm.c b/drivers/gpu/drm/i915/gt/intel_gt_pm.c
index 1363e069ec83..6187cdd06646 100644
--- a/drivers/gpu/drm/i915/gt/intel_gt_pm.c
+++ b/drivers/gpu/drm/i915/gt/intel_gt_pm.c
@@ -4,17 +4,38 @@
  * Copyright © 2019 Intel Corporation
  */
 
+#include <linux/suspend.h>
+
 #include "i915_drv.h"
+#include "i915_globals.h"
 #include "i915_params.h"
+#include "intel_context.h"
 #include "intel_engine_pm.h"
 #include "intel_gt.h"
 #include "intel_gt_pm.h"
+#include "intel_gt_requests.h"
+#include "intel_llc.h"
 #include "intel_pm.h"
+#include "intel_rc6.h"
+#include "intel_rps.h"
 #include "intel_wakeref.h"
 
-static void pm_notify(struct drm_i915_private *i915, int state)
+static void user_forcewake(struct intel_gt *gt, bool suspend)
 {
-	blocking_notifier_call_chain(&i915->gt.pm_notifications, state, i915);
+	int count = atomic_read(&gt->user_wakeref);
+
+	/* Inside suspend/resume so single threaded, no races to worry about. */
+	if (likely(!count))
+		return;
+
+	intel_gt_pm_get(gt);
+	if (suspend) {
+		GEM_BUG_ON(count > atomic_read(&gt->wakeref.count));
+		atomic_sub(count, &gt->wakeref.count);
+	} else {
+		atomic_add(count, &gt->wakeref.count);
+	}
+	intel_gt_pm_put(gt);
 }
 
 static int __gt_unpark(struct intel_wakeref *wf)
@@ -24,6 +45,8 @@ static int __gt_unpark(struct intel_wakeref *wf)
 
 	GEM_TRACE("\n");
 
+	i915_globals_unpark();
+
 	/*
 	 * It seems that the DMC likes to transition between the DC states a lot
 	 * when there are no connected displays (no active power domains) during
@@ -38,41 +61,44 @@ static int __gt_unpark(struct intel_wakeref *wf)
 	gt->awake = intel_display_power_get(i915, POWER_DOMAIN_GT_IRQ);
 	GEM_BUG_ON(!gt->awake);
 
-	intel_enable_gt_powersave(i915);
-
-	i915_update_gfx_val(i915);
-	if (INTEL_GEN(i915) >= 6)
-		gen6_rps_busy(i915);
+	if (NEEDS_RC6_CTX_CORRUPTION_WA(i915))
+		intel_uncore_forcewake_get(&i915->uncore, FORCEWAKE_ALL);
 
+	intel_rps_unpark(&gt->rps);
 	i915_pmu_gt_unparked(i915);
 
-	intel_gt_queue_hangcheck(gt);
-
-	pm_notify(i915, INTEL_GT_UNPARK);
+	intel_gt_unpark_requests(gt);
 
 	return 0;
 }
 
 static int __gt_park(struct intel_wakeref *wf)
 {
-	struct drm_i915_private *i915 =
-		container_of(wf, typeof(*i915), gt.wakeref);
-	intel_wakeref_t wakeref = fetch_and_zero(&i915->gt.awake);
+	struct intel_gt *gt = container_of(wf, typeof(*gt), wakeref);
+	intel_wakeref_t wakeref = fetch_and_zero(&gt->awake);
+	struct drm_i915_private *i915 = gt->i915;
 
 	GEM_TRACE("\n");
 
-	pm_notify(i915, INTEL_GT_PARK);
+	intel_gt_park_requests(gt);
 
+	i915_vma_parked(gt);
 	i915_pmu_gt_parked(i915);
-	if (INTEL_GEN(i915) >= 6)
-		gen6_rps_idle(i915);
+	intel_rps_park(&gt->rps);
 
 	/* Everything switched off, flush any residual interrupt just in case */
 	intel_synchronize_irq(i915);
 
+	if (NEEDS_RC6_CTX_CORRUPTION_WA(i915)) {
+		intel_rc6_ctx_wa_check(&i915->gt.rc6);
+		intel_uncore_forcewake_put(&i915->uncore, FORCEWAKE_ALL);
+	}
+
 	GEM_BUG_ON(!wakeref);
 	intel_display_power_put(i915, POWER_DOMAIN_GT_IRQ, wakeref);
 
+	i915_globals_park();
+
 	return 0;
 }
 
@@ -84,9 +110,18 @@ static const struct intel_wakeref_ops wf_ops = {
 
 void intel_gt_pm_init_early(struct intel_gt *gt)
 {
-	intel_wakeref_init(&gt->wakeref, &gt->i915->runtime_pm, &wf_ops);
+	intel_wakeref_init(&gt->wakeref, gt->uncore->rpm, &wf_ops);
+}
 
-	BLOCKING_INIT_NOTIFIER_HEAD(&gt->pm_notifications);
+void intel_gt_pm_init(struct intel_gt *gt)
+{
+	/*
+	 * Enabling power-management should be "self-healing". If we cannot
+	 * enable a feature, simply leave it disabled with a notice to the
+	 * user.
+	 */
+	intel_rc6_init(&gt->rc6);
+	intel_rps_init(&gt->rps);
 }
 
 static bool reset_engines(struct intel_gt *gt)
@@ -111,16 +146,47 @@ void intel_gt_sanitize(struct intel_gt *gt, bool force)
 {
 	struct intel_engine_cs *engine;
 	enum intel_engine_id id;
+	intel_wakeref_t wakeref;
 
-	GEM_TRACE("\n");
+	GEM_TRACE("force:%s\n", yesno(force));
+
+	/* Use a raw wakeref to avoid calling intel_display_power_get early */
+	wakeref = intel_runtime_pm_get(gt->uncore->rpm);
+	intel_uncore_forcewake_get(gt->uncore, FORCEWAKE_ALL);
+
+	/*
+	 * As we have just resumed the machine and woken the device up from
+	 * deep PCI sleep (presumably D3_cold), assume the HW has been reset
+	 * back to defaults, recovering from whatever wedged state we left it
+	 * in and so worth trying to use the device once more.
+	 */
+	if (intel_gt_is_wedged(gt))
+		intel_gt_unset_wedged(gt);
 
 	intel_uc_sanitize(&gt->uc);
 
-	if (!reset_engines(gt) && !force)
-		return;
+	for_each_engine(engine, gt, id)
+		if (engine->reset.prepare)
+			engine->reset.prepare(engine);
+
+	intel_uc_reset_prepare(&gt->uc);
+
+	if (reset_engines(gt) || force) {
+		for_each_engine(engine, gt, id)
+			__intel_engine_reset(engine, false);
+	}
+
+	for_each_engine(engine, gt, id)
+		if (engine->reset.finish)
+			engine->reset.finish(engine);
 
-	for_each_engine(engine, gt->i915, id)
-		__intel_engine_reset(engine, false);
+	intel_uncore_forcewake_put(gt->uncore, FORCEWAKE_ALL);
+	intel_runtime_pm_put(gt->uncore->rpm, wakeref);
+}
+
+void intel_gt_pm_fini(struct intel_gt *gt)
+{
+	intel_rc6_fini(&gt->rc6);
 }
 
 int intel_gt_resume(struct intel_gt *gt)
@@ -129,6 +195,8 @@ int intel_gt_resume(struct intel_gt *gt)
 	enum intel_engine_id id;
 	int err = 0;
 
+	GEM_TRACE("\n");
+
 	/*
 	 * After resume, we may need to poke into the pinned kernel
 	 * contexts to paper over any damage caused by the sudden suspend.
@@ -136,14 +204,23 @@ int intel_gt_resume(struct intel_gt *gt)
 	 * allowing us to fixup the user contexts on their first pin.
 	 */
 	intel_gt_pm_get(gt);
-	for_each_engine(engine, gt->i915, id) {
+
+	intel_uncore_forcewake_get(gt->uncore, FORCEWAKE_ALL);
+	intel_rc6_sanitize(&gt->rc6);
+
+	intel_rps_enable(&gt->rps);
+	intel_llc_enable(&gt->llc);
+
+	for_each_engine(engine, gt, id) {
 		struct intel_context *ce;
 
 		intel_engine_pm_get(engine);
 
 		ce = engine->kernel_context;
-		if (ce)
+		if (ce) {
+			GEM_BUG_ON(!intel_context_is_pinned(ce));
 			ce->ops->reset(ce);
+		}
 
 		engine->serial++; /* kernel context lost */
 		err = engine->resume(engine);
@@ -156,19 +233,99 @@ int intel_gt_resume(struct intel_gt *gt)
 			break;
 		}
 	}
+
+	intel_rc6_enable(&gt->rc6);
+
+	intel_uc_resume(&gt->uc);
+
+	user_forcewake(gt, false);
+
+	intel_uncore_forcewake_put(gt->uncore, FORCEWAKE_ALL);
 	intel_gt_pm_put(gt);
 
 	return err;
 }
 
+static void wait_for_suspend(struct intel_gt *gt)
+{
+	if (!intel_gt_pm_is_awake(gt))
+		return;
+
+	if (intel_gt_wait_for_idle(gt, I915_GEM_IDLE_TIMEOUT) == -ETIME) {
+		/*
+		 * Forcibly cancel outstanding work and leave
+		 * the gpu quiet.
+		 */
+		intel_gt_set_wedged(gt);
+	}
+
+	intel_gt_pm_wait_for_idle(gt);
+}
+
+void intel_gt_suspend_prepare(struct intel_gt *gt)
+{
+	user_forcewake(gt, true);
+	wait_for_suspend(gt);
+
+	intel_uc_suspend(&gt->uc);
+}
+
+static suspend_state_t pm_suspend_target(void)
+{
+#if IS_ENABLED(CONFIG_PM_SLEEP)
+	return pm_suspend_target_state;
+#else
+	return PM_SUSPEND_TO_IDLE;
+#endif
+}
+
+void intel_gt_suspend_late(struct intel_gt *gt)
+{
+	intel_wakeref_t wakeref;
+
+	/* We expect to be idle already; but also want to be independent */
+	wait_for_suspend(gt);
+
+	/*
+	 * On disabling the device, we want to turn off HW access to memory
+	 * that we no longer own.
+	 *
+	 * However, not all suspend-states disable the device. S0 (s2idle)
+	 * is effectively runtime-suspend, the device is left powered on
+	 * but needs to be put into a low power state. We need to keep
+	 * powermanagement enabled, but we also retain system state and so
+	 * it remains safe to keep on using our allocated memory.
+	 */
+	if (pm_suspend_target() == PM_SUSPEND_TO_IDLE)
+		return;
+
+	with_intel_runtime_pm(gt->uncore->rpm, wakeref) {
+		intel_rps_disable(&gt->rps);
+		intel_rc6_disable(&gt->rc6);
+		intel_llc_disable(&gt->llc);
+	}
+
+	intel_gt_sanitize(gt, false);
+
+	GEM_TRACE("\n");
+}
+
 void intel_gt_runtime_suspend(struct intel_gt *gt)
 {
 	intel_uc_runtime_suspend(&gt->uc);
+
+	GEM_TRACE("\n");
 }
 
 int intel_gt_runtime_resume(struct intel_gt *gt)
 {
+	GEM_TRACE("\n");
+
 	intel_gt_init_swizzling(gt);
 
 	return intel_uc_runtime_resume(&gt->uc);
 }
+
+#if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
+#include "selftest_gt_pm.c"
+#endif
diff --git a/drivers/gpu/drm/i915/gt/intel_gt_pm.h b/drivers/gpu/drm/i915/gt/intel_gt_pm.h
index fb39d99cd6ee..b3e17399be9b 100644
--- a/drivers/gpu/drm/i915/gt/intel_gt_pm.h
+++ b/drivers/gpu/drm/i915/gt/intel_gt_pm.h
@@ -12,11 +12,6 @@
 #include "intel_gt_types.h"
 #include "intel_wakeref.h"
 
-enum {
-	INTEL_GT_UNPARK,
-	INTEL_GT_PARK,
-};
-
 static inline bool intel_gt_pm_is_awake(const struct intel_gt *gt)
 {
 	return intel_wakeref_is_active(&gt->wakeref);
@@ -43,10 +38,21 @@ static inline int intel_gt_pm_wait_for_idle(struct intel_gt *gt)
 }
 
 void intel_gt_pm_init_early(struct intel_gt *gt);
+void intel_gt_pm_init(struct intel_gt *gt);
+void intel_gt_pm_fini(struct intel_gt *gt);
 
 void intel_gt_sanitize(struct intel_gt *gt, bool force);
+
+void intel_gt_suspend_prepare(struct intel_gt *gt);
+void intel_gt_suspend_late(struct intel_gt *gt);
 int intel_gt_resume(struct intel_gt *gt);
+
 void intel_gt_runtime_suspend(struct intel_gt *gt);
 int intel_gt_runtime_resume(struct intel_gt *gt);
 
+static inline bool is_mock_gt(const struct intel_gt *gt)
+{
+	return I915_SELFTEST_ONLY(gt->awake == -ENODEV);
+}
+
 #endif /* INTEL_GT_PM_H */
diff --git a/drivers/gpu/drm/i915/gt/intel_gt_requests.c b/drivers/gpu/drm/i915/gt/intel_gt_requests.c
new file mode 100644
index 000000000000..353809ac2754
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/intel_gt_requests.c
@@ -0,0 +1,137 @@
+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2019 Intel Corporation
+ */
+
+#include "i915_drv.h" /* for_each_engine() */
+#include "i915_request.h"
+#include "intel_gt.h"
+#include "intel_gt_pm.h"
+#include "intel_gt_requests.h"
+#include "intel_timeline.h"
+
+static void retire_requests(struct intel_timeline *tl)
+{
+	struct i915_request *rq, *rn;
+
+	list_for_each_entry_safe(rq, rn, &tl->requests, link)
+		if (!i915_request_retire(rq))
+			break;
+}
+
+static void flush_submission(struct intel_gt *gt)
+{
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
+
+	for_each_engine(engine, gt, id)
+		intel_engine_flush_submission(engine);
+}
+
+long intel_gt_retire_requests_timeout(struct intel_gt *gt, long timeout)
+{
+	struct intel_gt_timelines *timelines = &gt->timelines;
+	struct intel_timeline *tl, *tn;
+	unsigned long active_count = 0;
+	unsigned long flags;
+	bool interruptible;
+	LIST_HEAD(free);
+
+	interruptible = true;
+	if (unlikely(timeout < 0))
+		timeout = -timeout, interruptible = false;
+
+	flush_submission(gt); /* kick the ksoftirqd tasklets */
+
+	spin_lock_irqsave(&timelines->lock, flags);
+	list_for_each_entry_safe(tl, tn, &timelines->active_list, link) {
+		if (!mutex_trylock(&tl->mutex)) {
+			active_count++; /* report busy to caller, try again? */
+			continue;
+		}
+
+		intel_timeline_get(tl);
+		GEM_BUG_ON(!tl->active_count);
+		tl->active_count++; /* pin the list element */
+		spin_unlock_irqrestore(&timelines->lock, flags);
+
+		if (timeout > 0) {
+			struct dma_fence *fence;
+
+			fence = i915_active_fence_get(&tl->last_request);
+			if (fence) {
+				timeout = dma_fence_wait_timeout(fence,
+								 interruptible,
+								 timeout);
+				dma_fence_put(fence);
+			}
+		}
+
+		retire_requests(tl);
+
+		spin_lock_irqsave(&timelines->lock, flags);
+
+		/* Resume iteration after dropping lock */
+		list_safe_reset_next(tl, tn, link);
+		if (!--tl->active_count)
+			list_del(&tl->link);
+		else
+			active_count += !!rcu_access_pointer(tl->last_request.fence);
+
+		mutex_unlock(&tl->mutex);
+
+		/* Defer the final release to after the spinlock */
+		if (refcount_dec_and_test(&tl->kref.refcount)) {
+			GEM_BUG_ON(tl->active_count);
+			list_add(&tl->link, &free);
+		}
+	}
+	spin_unlock_irqrestore(&timelines->lock, flags);
+
+	list_for_each_entry_safe(tl, tn, &free, link)
+		__intel_timeline_free(&tl->kref);
+
+	return active_count ? timeout : 0;
+}
+
+int intel_gt_wait_for_idle(struct intel_gt *gt, long timeout)
+{
+	/* If the device is asleep, we have no requests outstanding */
+	if (!intel_gt_pm_is_awake(gt))
+		return 0;
+
+	while ((timeout = intel_gt_retire_requests_timeout(gt, timeout)) > 0) {
+		cond_resched();
+		if (signal_pending(current))
+			return -EINTR;
+	}
+
+	return timeout;
+}
+
+static void retire_work_handler(struct work_struct *work)
+{
+	struct intel_gt *gt =
+		container_of(work, typeof(*gt), requests.retire_work.work);
+
+	intel_gt_retire_requests(gt);
+	schedule_delayed_work(&gt->requests.retire_work,
+			      round_jiffies_up_relative(HZ));
+}
+
+void intel_gt_init_requests(struct intel_gt *gt)
+{
+	INIT_DELAYED_WORK(&gt->requests.retire_work, retire_work_handler);
+}
+
+void intel_gt_park_requests(struct intel_gt *gt)
+{
+	cancel_delayed_work(&gt->requests.retire_work);
+}
+
+void intel_gt_unpark_requests(struct intel_gt *gt)
+{
+	schedule_delayed_work(&gt->requests.retire_work,
+			      round_jiffies_up_relative(HZ));
+}
diff --git a/drivers/gpu/drm/i915/gt/intel_gt_requests.h b/drivers/gpu/drm/i915/gt/intel_gt_requests.h
new file mode 100644
index 000000000000..bd31cbce47e0
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/intel_gt_requests.h
@@ -0,0 +1,24 @@
+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2019 Intel Corporation
+ */
+
+#ifndef INTEL_GT_REQUESTS_H
+#define INTEL_GT_REQUESTS_H
+
+struct intel_gt;
+
+long intel_gt_retire_requests_timeout(struct intel_gt *gt, long timeout);
+static inline void intel_gt_retire_requests(struct intel_gt *gt)
+{
+	intel_gt_retire_requests_timeout(gt, 0);
+}
+
+int intel_gt_wait_for_idle(struct intel_gt *gt, long timeout);
+
+void intel_gt_init_requests(struct intel_gt *gt);
+void intel_gt_park_requests(struct intel_gt *gt);
+void intel_gt_unpark_requests(struct intel_gt *gt);
+
+#endif /* INTEL_GT_REQUESTS_H */
diff --git a/drivers/gpu/drm/i915/gt/intel_gt_types.h b/drivers/gpu/drm/i915/gt/intel_gt_types.h
index dc295c196d11..d4e14dbd172e 100644
--- a/drivers/gpu/drm/i915/gt/intel_gt_types.h
+++ b/drivers/gpu/drm/i915/gt/intel_gt_types.h
@@ -17,7 +17,10 @@
 
 #include "i915_vma.h"
 #include "intel_engine_types.h"
+#include "intel_llc_types.h"
 #include "intel_reset_types.h"
+#include "intel_rc6_types.h"
+#include "intel_rps_types.h"
 #include "intel_wakeref.h"
 
 struct drm_i915_private;
@@ -25,14 +28,6 @@ struct i915_ggtt;
 struct intel_engine_cs;
 struct intel_uncore;
 
-struct intel_hangcheck {
-	/* For hangcheck timer */
-#define DRM_I915_HANGCHECK_PERIOD 1500 /* in ms */
-#define DRM_I915_HANGCHECK_JIFFIES msecs_to_jiffies(DRM_I915_HANGCHECK_PERIOD)
-
-	struct delayed_work work;
-};
-
 struct intel_gt {
 	struct drm_i915_private *i915;
 	struct intel_uncore *uncore;
@@ -49,12 +44,23 @@ struct intel_gt {
 		struct list_head hwsp_free_list;
 	} timelines;
 
+	struct intel_gt_requests {
+		/**
+		 * We leave the user IRQ off as much as possible,
+		 * but this means that requests will finish and never
+		 * be retired once the system goes idle. Set a timer to
+		 * fire periodically while the ring is running. When it
+		 * fires, go retire requests.
+		 */
+		struct delayed_work retire_work;
+	} requests;
+
 	struct intel_wakeref wakeref;
+	atomic_t user_wakeref;
 
 	struct list_head closed_vma;
 	spinlock_t closed_lock; /* guards the list of closed_vma */
 
-	struct intel_hangcheck hangcheck;
 	struct intel_reset reset;
 
 	/**
@@ -66,7 +72,9 @@ struct intel_gt {
 	 */
 	intel_wakeref_t awake;
 
-	struct blocking_notifier_head pm_notifications;
+	struct intel_llc llc;
+	struct intel_rc6 rc6;
+	struct intel_rps rps;
 
 	ktime_t last_init_time;
 
@@ -89,14 +97,16 @@ enum intel_gt_scratch_field {
 	INTEL_GT_SCRATCH_FIELD_DEFAULT = 0,
 
 	/* 8 bytes */
-	INTEL_GT_SCRATCH_FIELD_CLEAR_SLM_WA = 128,
-
-	/* 8 bytes */
 	INTEL_GT_SCRATCH_FIELD_RENDER_FLUSH = 128,
 
 	/* 8 bytes */
 	INTEL_GT_SCRATCH_FIELD_COHERENTL3_WA = 256,
 
+	/* 6 * 8 bytes */
+	INTEL_GT_SCRATCH_FIELD_PERF_CS_GPR = 2048,
+
+	/* 4 bytes */
+	INTEL_GT_SCRATCH_FIELD_PERF_PREDICATE_RESULT_1 = 2096,
 };
 
 #endif /* __INTEL_GT_TYPES_H__ */
diff --git a/drivers/gpu/drm/i915/gt/intel_hangcheck.c b/drivers/gpu/drm/i915/gt/intel_hangcheck.c
deleted file mode 100644
index 05d042cdefe2..000000000000
--- a/drivers/gpu/drm/i915/gt/intel_hangcheck.c
+++ /dev/null
@@ -1,360 +0,0 @@
-/*
- * Copyright © 2016 Intel Corporation
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice (including the next
- * paragraph) shall be included in all copies or substantial portions of the
- * Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
- * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
- * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
- * IN THE SOFTWARE.
- *
- */
-
-#include "i915_drv.h"
-#include "intel_engine.h"
-#include "intel_gt.h"
-#include "intel_reset.h"
-
-struct hangcheck {
-	u64 acthd;
-	u32 ring;
-	u32 head;
-	enum intel_engine_hangcheck_action action;
-	unsigned long action_timestamp;
-	int deadlock;
-	struct intel_instdone instdone;
-	bool wedged:1;
-	bool stalled:1;
-};
-
-static bool instdone_unchanged(u32 current_instdone, u32 *old_instdone)
-{
-	u32 tmp = current_instdone | *old_instdone;
-	bool unchanged;
-
-	unchanged = tmp == *old_instdone;
-	*old_instdone |= tmp;
-
-	return unchanged;
-}
-
-static bool subunits_stuck(struct intel_engine_cs *engine)
-{
-	struct drm_i915_private *dev_priv = engine->i915;
-	struct intel_instdone instdone;
-	struct intel_instdone *accu_instdone = &engine->hangcheck.instdone;
-	bool stuck;
-	int slice;
-	int subslice;
-
-	intel_engine_get_instdone(engine, &instdone);
-
-	/* There might be unstable subunit states even when
-	 * actual head is not moving. Filter out the unstable ones by
-	 * accumulating the undone -> done transitions and only
-	 * consider those as progress.
-	 */
-	stuck = instdone_unchanged(instdone.instdone,
-				   &accu_instdone->instdone);
-	stuck &= instdone_unchanged(instdone.slice_common,
-				    &accu_instdone->slice_common);
-
-	for_each_instdone_slice_subslice(dev_priv, slice, subslice) {
-		stuck &= instdone_unchanged(instdone.sampler[slice][subslice],
-					    &accu_instdone->sampler[slice][subslice]);
-		stuck &= instdone_unchanged(instdone.row[slice][subslice],
-					    &accu_instdone->row[slice][subslice]);
-	}
-
-	return stuck;
-}
-
-static enum intel_engine_hangcheck_action
-head_stuck(struct intel_engine_cs *engine, u64 acthd)
-{
-	if (acthd != engine->hangcheck.acthd) {
-
-		/* Clear subunit states on head movement */
-		memset(&engine->hangcheck.instdone, 0,
-		       sizeof(engine->hangcheck.instdone));
-
-		return ENGINE_ACTIVE_HEAD;
-	}
-
-	if (!subunits_stuck(engine))
-		return ENGINE_ACTIVE_SUBUNITS;
-
-	return ENGINE_DEAD;
-}
-
-static enum intel_engine_hangcheck_action
-engine_stuck(struct intel_engine_cs *engine, u64 acthd)
-{
-	enum intel_engine_hangcheck_action ha;
-	u32 tmp;
-
-	ha = head_stuck(engine, acthd);
-	if (ha != ENGINE_DEAD)
-		return ha;
-
-	if (IS_GEN(engine->i915, 2))
-		return ENGINE_DEAD;
-
-	/* Is the chip hanging on a WAIT_FOR_EVENT?
-	 * If so we can simply poke the RB_WAIT bit
-	 * and break the hang. This should work on
-	 * all but the second generation chipsets.
-	 */
-	tmp = ENGINE_READ(engine, RING_CTL);
-	if (tmp & RING_WAIT) {
-		intel_gt_handle_error(engine->gt, engine->mask, 0,
-				      "stuck wait on %s", engine->name);
-		ENGINE_WRITE(engine, RING_CTL, tmp);
-		return ENGINE_WAIT_KICK;
-	}
-
-	return ENGINE_DEAD;
-}
-
-static void hangcheck_load_sample(struct intel_engine_cs *engine,
-				  struct hangcheck *hc)
-{
-	hc->acthd = intel_engine_get_active_head(engine);
-	hc->ring = ENGINE_READ(engine, RING_START);
-	hc->head = ENGINE_READ(engine, RING_HEAD);
-}
-
-static void hangcheck_store_sample(struct intel_engine_cs *engine,
-				   const struct hangcheck *hc)
-{
-	engine->hangcheck.acthd = hc->acthd;
-	engine->hangcheck.last_ring = hc->ring;
-	engine->hangcheck.last_head = hc->head;
-}
-
-static enum intel_engine_hangcheck_action
-hangcheck_get_action(struct intel_engine_cs *engine,
-		     const struct hangcheck *hc)
-{
-	if (intel_engine_is_idle(engine))
-		return ENGINE_IDLE;
-
-	if (engine->hangcheck.last_ring != hc->ring)
-		return ENGINE_ACTIVE_SEQNO;
-
-	if (engine->hangcheck.last_head != hc->head)
-		return ENGINE_ACTIVE_SEQNO;
-
-	return engine_stuck(engine, hc->acthd);
-}
-
-static void hangcheck_accumulate_sample(struct intel_engine_cs *engine,
-					struct hangcheck *hc)
-{
-	unsigned long timeout = I915_ENGINE_DEAD_TIMEOUT;
-
-	hc->action = hangcheck_get_action(engine, hc);
-
-	/* We always increment the progress
-	 * if the engine is busy and still processing
-	 * the same request, so that no single request
-	 * can run indefinitely (such as a chain of
-	 * batches). The only time we do not increment
-	 * the hangcheck score on this ring, if this
-	 * engine is in a legitimate wait for another
-	 * engine. In that case the waiting engine is a
-	 * victim and we want to be sure we catch the
-	 * right culprit. Then every time we do kick
-	 * the ring, make it as a progress as the seqno
-	 * advancement might ensure and if not, it
-	 * will catch the hanging engine.
-	 */
-
-	switch (hc->action) {
-	case ENGINE_IDLE:
-	case ENGINE_ACTIVE_SEQNO:
-		/* Clear head and subunit states on seqno movement */
-		hc->acthd = 0;
-
-		memset(&engine->hangcheck.instdone, 0,
-		       sizeof(engine->hangcheck.instdone));
-
-		/* Intentional fall through */
-	case ENGINE_WAIT_KICK:
-	case ENGINE_WAIT:
-		engine->hangcheck.action_timestamp = jiffies;
-		break;
-
-	case ENGINE_ACTIVE_HEAD:
-	case ENGINE_ACTIVE_SUBUNITS:
-		/*
-		 * Seqno stuck with still active engine gets leeway,
-		 * in hopes that it is just a long shader.
-		 */
-		timeout = I915_SEQNO_DEAD_TIMEOUT;
-		break;
-
-	case ENGINE_DEAD:
-		break;
-
-	default:
-		MISSING_CASE(hc->action);
-	}
-
-	hc->stalled = time_after(jiffies,
-				 engine->hangcheck.action_timestamp + timeout);
-	hc->wedged = time_after(jiffies,
-				 engine->hangcheck.action_timestamp +
-				 I915_ENGINE_WEDGED_TIMEOUT);
-}
-
-static void hangcheck_declare_hang(struct intel_gt *gt,
-				   intel_engine_mask_t hung,
-				   intel_engine_mask_t stuck)
-{
-	struct intel_engine_cs *engine;
-	intel_engine_mask_t tmp;
-	char msg[80];
-	int len;
-
-	/* If some rings hung but others were still busy, only
-	 * blame the hanging rings in the synopsis.
-	 */
-	if (stuck != hung)
-		hung &= ~stuck;
-	len = scnprintf(msg, sizeof(msg),
-			"%s on ", stuck == hung ? "no progress" : "hang");
-	for_each_engine_masked(engine, gt->i915, hung, tmp)
-		len += scnprintf(msg + len, sizeof(msg) - len,
-				 "%s, ", engine->name);
-	msg[len-2] = '\0';
-
-	return intel_gt_handle_error(gt, hung, I915_ERROR_CAPTURE, "%s", msg);
-}
-
-/*
- * This is called when the chip hasn't reported back with completed
- * batchbuffers in a long time. We keep track per ring seqno progress and
- * if there are no progress, hangcheck score for that ring is increased.
- * Further, acthd is inspected to see if the ring is stuck. On stuck case
- * we kick the ring. If we see no progress on three subsequent calls
- * we assume chip is wedged and try to fix it by resetting the chip.
- */
-static void hangcheck_elapsed(struct work_struct *work)
-{
-	struct intel_gt *gt =
-		container_of(work, typeof(*gt), hangcheck.work.work);
-	intel_engine_mask_t hung = 0, stuck = 0, wedged = 0;
-	struct intel_engine_cs *engine;
-	enum intel_engine_id id;
-	intel_wakeref_t wakeref;
-
-	if (!i915_modparams.enable_hangcheck)
-		return;
-
-	if (!READ_ONCE(gt->awake))
-		return;
-
-	if (intel_gt_is_wedged(gt))
-		return;
-
-	wakeref = intel_runtime_pm_get_if_in_use(&gt->i915->runtime_pm);
-	if (!wakeref)
-		return;
-
-	/* As enabling the GPU requires fairly extensive mmio access,
-	 * periodically arm the mmio checker to see if we are triggering
-	 * any invalid access.
-	 */
-	intel_uncore_arm_unclaimed_mmio_detection(gt->uncore);
-
-	for_each_engine(engine, gt->i915, id) {
-		struct hangcheck hc;
-
-		intel_engine_signal_breadcrumbs(engine);
-
-		hangcheck_load_sample(engine, &hc);
-		hangcheck_accumulate_sample(engine, &hc);
-		hangcheck_store_sample(engine, &hc);
-
-		if (hc.stalled) {
-			hung |= engine->mask;
-			if (hc.action != ENGINE_DEAD)
-				stuck |= engine->mask;
-		}
-
-		if (hc.wedged)
-			wedged |= engine->mask;
-	}
-
-	if (GEM_SHOW_DEBUG() && (hung | stuck)) {
-		struct drm_printer p = drm_debug_printer("hangcheck");
-
-		for_each_engine(engine, gt->i915, id) {
-			if (intel_engine_is_idle(engine))
-				continue;
-
-			intel_engine_dump(engine, &p, "%s\n", engine->name);
-		}
-	}
-
-	if (wedged) {
-		dev_err(gt->i915->drm.dev,
-			"GPU recovery timed out,"
-			" cancelling all in-flight rendering.\n");
-		GEM_TRACE_DUMP();
-		intel_gt_set_wedged(gt);
-	}
-
-	if (hung)
-		hangcheck_declare_hang(gt, hung, stuck);
-
-	intel_runtime_pm_put(&gt->i915->runtime_pm, wakeref);
-
-	/* Reset timer in case GPU hangs without another request being added */
-	intel_gt_queue_hangcheck(gt);
-}
-
-void intel_gt_queue_hangcheck(struct intel_gt *gt)
-{
-	unsigned long delay;
-
-	if (unlikely(!i915_modparams.enable_hangcheck))
-		return;
-
-	/*
-	 * Don't continually defer the hangcheck so that it is always run at
-	 * least once after work has been scheduled on any ring. Otherwise,
-	 * we will ignore a hung ring if a second ring is kept busy.
-	 */
-
-	delay = round_jiffies_up_relative(DRM_I915_HANGCHECK_JIFFIES);
-	queue_delayed_work(system_long_wq, &gt->hangcheck.work, delay);
-}
-
-void intel_engine_init_hangcheck(struct intel_engine_cs *engine)
-{
-	memset(&engine->hangcheck, 0, sizeof(engine->hangcheck));
-	engine->hangcheck.action_timestamp = jiffies;
-}
-
-void intel_gt_init_hangcheck(struct intel_gt *gt)
-{
-	INIT_DELAYED_WORK(&gt->hangcheck.work, hangcheck_elapsed);
-}
-
-#if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
-#include "selftest_hangcheck.c"
-#endif
diff --git a/drivers/gpu/drm/i915/gt/intel_llc.c b/drivers/gpu/drm/i915/gt/intel_llc.c
new file mode 100644
index 000000000000..ceb785b75c25
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/intel_llc.c
@@ -0,0 +1,161 @@
+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2019 Intel Corporation
+ */
+
+#include <linux/cpufreq.h>
+
+#include "i915_drv.h"
+#include "intel_gt.h"
+#include "intel_llc.h"
+#include "intel_sideband.h"
+
+struct ia_constants {
+	unsigned int min_gpu_freq;
+	unsigned int max_gpu_freq;
+
+	unsigned int min_ring_freq;
+	unsigned int max_ia_freq;
+};
+
+static struct intel_gt *llc_to_gt(struct intel_llc *llc)
+{
+	return container_of(llc, struct intel_gt, llc);
+}
+
+static unsigned int cpu_max_MHz(void)
+{
+	struct cpufreq_policy *policy;
+	unsigned int max_khz;
+
+	policy = cpufreq_cpu_get(0);
+	if (policy) {
+		max_khz = policy->cpuinfo.max_freq;
+		cpufreq_cpu_put(policy);
+	} else {
+		/*
+		 * Default to measured freq if none found, PCU will ensure we
+		 * don't go over
+		 */
+		max_khz = tsc_khz;
+	}
+
+	return max_khz / 1000;
+}
+
+static bool get_ia_constants(struct intel_llc *llc,
+			     struct ia_constants *consts)
+{
+	struct drm_i915_private *i915 = llc_to_gt(llc)->i915;
+	struct intel_rps *rps = &llc_to_gt(llc)->rps;
+
+	if (rps->max_freq <= rps->min_freq)
+		return false;
+
+	consts->max_ia_freq = cpu_max_MHz();
+
+	consts->min_ring_freq =
+		intel_uncore_read(llc_to_gt(llc)->uncore, DCLK) & 0xf;
+	/* convert DDR frequency from units of 266.6MHz to bandwidth */
+	consts->min_ring_freq = mult_frac(consts->min_ring_freq, 8, 3);
+
+	consts->min_gpu_freq = rps->min_freq;
+	consts->max_gpu_freq = rps->max_freq;
+	if (INTEL_GEN(i915) >= 9) {
+		/* Convert GT frequency to 50 HZ units */
+		consts->min_gpu_freq /= GEN9_FREQ_SCALER;
+		consts->max_gpu_freq /= GEN9_FREQ_SCALER;
+	}
+
+	return true;
+}
+
+static void calc_ia_freq(struct intel_llc *llc,
+			 unsigned int gpu_freq,
+			 const struct ia_constants *consts,
+			 unsigned int *out_ia_freq,
+			 unsigned int *out_ring_freq)
+{
+	struct drm_i915_private *i915 = llc_to_gt(llc)->i915;
+	const int diff = consts->max_gpu_freq - gpu_freq;
+	unsigned int ia_freq = 0, ring_freq = 0;
+
+	if (INTEL_GEN(i915) >= 9) {
+		/*
+		 * ring_freq = 2 * GT. ring_freq is in 100MHz units
+		 * No floor required for ring frequency on SKL.
+		 */
+		ring_freq = gpu_freq;
+	} else if (INTEL_GEN(i915) >= 8) {
+		/* max(2 * GT, DDR). NB: GT is 50MHz units */
+		ring_freq = max(consts->min_ring_freq, gpu_freq);
+	} else if (IS_HASWELL(i915)) {
+		ring_freq = mult_frac(gpu_freq, 5, 4);
+		ring_freq = max(consts->min_ring_freq, ring_freq);
+		/* leave ia_freq as the default, chosen by cpufreq */
+	} else {
+		const int min_freq = 15;
+		const int scale = 180;
+
+		/*
+		 * On older processors, there is no separate ring
+		 * clock domain, so in order to boost the bandwidth
+		 * of the ring, we need to upclock the CPU (ia_freq).
+		 *
+		 * For GPU frequencies less than 750MHz,
+		 * just use the lowest ring freq.
+		 */
+		if (gpu_freq < min_freq)
+			ia_freq = 800;
+		else
+			ia_freq = consts->max_ia_freq - diff * scale / 2;
+		ia_freq = DIV_ROUND_CLOSEST(ia_freq, 100);
+	}
+
+	*out_ia_freq = ia_freq;
+	*out_ring_freq = ring_freq;
+}
+
+static void gen6_update_ring_freq(struct intel_llc *llc)
+{
+	struct drm_i915_private *i915 = llc_to_gt(llc)->i915;
+	struct ia_constants consts;
+	unsigned int gpu_freq;
+
+	if (!get_ia_constants(llc, &consts))
+		return;
+
+	/*
+	 * For each potential GPU frequency, load a ring frequency we'd like
+	 * to use for memory access.  We do this by specifying the IA frequency
+	 * the PCU should use as a reference to determine the ring frequency.
+	 */
+	for (gpu_freq = consts.max_gpu_freq;
+	     gpu_freq >= consts.min_gpu_freq;
+	     gpu_freq--) {
+		unsigned int ia_freq, ring_freq;
+
+		calc_ia_freq(llc, gpu_freq, &consts, &ia_freq, &ring_freq);
+		sandybridge_pcode_write(i915,
+					GEN6_PCODE_WRITE_MIN_FREQ_TABLE,
+					ia_freq << GEN6_PCODE_FREQ_IA_RATIO_SHIFT |
+					ring_freq << GEN6_PCODE_FREQ_RING_RATIO_SHIFT |
+					gpu_freq);
+	}
+}
+
+void intel_llc_enable(struct intel_llc *llc)
+{
+	if (HAS_LLC(llc_to_gt(llc)->i915))
+		gen6_update_ring_freq(llc);
+}
+
+void intel_llc_disable(struct intel_llc *llc)
+{
+	/* Currently there is no HW configuration to be done to disable. */
+}
+
+#if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
+#include "selftest_llc.c"
+#endif
diff --git a/drivers/gpu/drm/i915/gt/intel_llc.h b/drivers/gpu/drm/i915/gt/intel_llc.h
new file mode 100644
index 000000000000..ef09a890d2b7
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/intel_llc.h
@@ -0,0 +1,15 @@
+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2019 Intel Corporation
+ */
+
+#ifndef INTEL_LLC_H
+#define INTEL_LLC_H
+
+struct intel_llc;
+
+void intel_llc_enable(struct intel_llc *llc);
+void intel_llc_disable(struct intel_llc *llc);
+
+#endif /* INTEL_LLC_H */
diff --git a/drivers/gpu/drm/i915/gt/intel_llc_types.h b/drivers/gpu/drm/i915/gt/intel_llc_types.h
new file mode 100644
index 000000000000..ecad4687b930
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/intel_llc_types.h
@@ -0,0 +1,13 @@
+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2019 Intel Corporation
+ */
+
+#ifndef INTEL_LLC_TYPES_H
+#define INTEL_LLC_TYPES_H
+
+struct intel_llc {
+};
+
+#endif /* INTEL_LLC_TYPES_H */
diff --git a/drivers/gpu/drm/i915/gt/intel_lrc.c b/drivers/gpu/drm/i915/gt/intel_lrc.c
index d42584439f51..0ac3b26674ad 100644
--- a/drivers/gpu/drm/i915/gt/intel_lrc.c
+++ b/drivers/gpu/drm/i915/gt/intel_lrc.c
@@ -145,6 +145,7 @@
 #include "intel_lrc_reg.h"
 #include "intel_mocs.h"
 #include "intel_reset.h"
+#include "intel_ring.h"
 #include "intel_workarounds.h"
 
 #define RING_EXECLIST_QFULL		(1 << 0x2)
@@ -230,9 +231,41 @@ static int __execlists_context_alloc(struct intel_context *ce,
 				     struct intel_engine_cs *engine);
 
 static void execlists_init_reg_state(u32 *reg_state,
-				     struct intel_context *ce,
-				     struct intel_engine_cs *engine,
-				     struct intel_ring *ring);
+				     const struct intel_context *ce,
+				     const struct intel_engine_cs *engine,
+				     const struct intel_ring *ring,
+				     bool close);
+static void
+__execlists_update_reg_state(const struct intel_context *ce,
+			     const struct intel_engine_cs *engine);
+
+static void mark_eio(struct i915_request *rq)
+{
+	if (i915_request_completed(rq))
+		return;
+
+	GEM_BUG_ON(i915_request_signaled(rq));
+
+	dma_fence_set_error(&rq->fence, -EIO);
+	i915_request_mark_complete(rq);
+}
+
+static struct i915_request *
+active_request(const struct intel_timeline * const tl, struct i915_request *rq)
+{
+	struct i915_request *active = rq;
+
+	rcu_read_lock();
+	list_for_each_entry_continue_reverse(rq, &tl->requests, link) {
+		if (i915_request_completed(rq))
+			break;
+
+		active = rq;
+	}
+	rcu_read_unlock();
+
+	return active;
+}
 
 static inline u32 intel_hws_preempt_address(struct intel_engine_cs *engine)
 {
@@ -330,10 +363,15 @@ static inline bool need_preempt(const struct intel_engine_cs *engine,
 	 * However, the priority hint is a mere hint that we may need to
 	 * preempt. If that hint is stale or we may be trying to preempt
 	 * ourselves, ignore the request.
+	 *
+	 * More naturally we would write
+	 *      prio >= max(0, last);
+	 * except that we wish to prevent triggering preemption at the same
+	 * priority level: the task that is running should remain running
+	 * to preserve FIFO ordering of dependencies.
 	 */
-	last_prio = effective_prio(rq);
-	if (!i915_scheduler_need_preempt(engine->execlists.queue_priority_hint,
-					 last_prio))
+	last_prio = max(effective_prio(rq), I915_PRIORITY_NORMAL - 1);
+	if (engine->execlists.queue_priority_hint <= last_prio)
 		return false;
 
 	/*
@@ -422,12 +460,8 @@ assert_priority_queue(const struct i915_request *prev,
 static u64
 lrc_descriptor(struct intel_context *ce, struct intel_engine_cs *engine)
 {
-	struct i915_gem_context *ctx = ce->gem_context;
 	u64 desc;
 
-	BUILD_BUG_ON(MAX_CONTEXT_HW_ID > (BIT(GEN8_CTX_ID_WIDTH)));
-	BUILD_BUG_ON(GEN11_MAX_CONTEXT_HW_ID > (BIT(GEN11_SW_CTX_ID_WIDTH)));
-
 	desc = INTEL_LEGACY_32B_CONTEXT;
 	if (i915_vm_is_4lvl(ce->vm))
 		desc = INTEL_LEGACY_64B_CONTEXT;
@@ -437,33 +471,379 @@ lrc_descriptor(struct intel_context *ce, struct intel_engine_cs *engine)
 	if (IS_GEN(engine->i915, 8))
 		desc |= GEN8_CTX_L3LLC_COHERENT;
 
-	desc |= i915_ggtt_offset(ce->state) + LRC_HEADER_PAGES * PAGE_SIZE;
-								/* bits 12-31 */
+	desc |= i915_ggtt_offset(ce->state); /* bits 12-31 */
 	/*
 	 * The following 32bits are copied into the OA reports (dword 2).
 	 * Consider updating oa_get_render_ctx_id in i915_perf.c when changing
 	 * anything below.
 	 */
 	if (INTEL_GEN(engine->i915) >= 11) {
-		GEM_BUG_ON(ctx->hw_id >= BIT(GEN11_SW_CTX_ID_WIDTH));
-		desc |= (u64)ctx->hw_id << GEN11_SW_CTX_ID_SHIFT;
-								/* bits 37-47 */
-
 		desc |= (u64)engine->instance << GEN11_ENGINE_INSTANCE_SHIFT;
 								/* bits 48-53 */
 
-		/* TODO: decide what to do with SW counter (bits 55-60) */
-
 		desc |= (u64)engine->class << GEN11_ENGINE_CLASS_SHIFT;
 								/* bits 61-63 */
-	} else {
-		GEM_BUG_ON(ctx->hw_id >= BIT(GEN8_CTX_ID_WIDTH));
-		desc |= (u64)ctx->hw_id << GEN8_CTX_ID_SHIFT;	/* bits 32-52 */
 	}
 
 	return desc;
 }
 
+static u32 *set_offsets(u32 *regs,
+			const u8 *data,
+			const struct intel_engine_cs *engine)
+#define NOP(x) (BIT(7) | (x))
+#define LRI(count, flags) ((flags) << 6 | (count))
+#define POSTED BIT(0)
+#define REG(x) (((x) >> 2) | BUILD_BUG_ON_ZERO(x >= 0x200))
+#define REG16(x) \
+	(((x) >> 9) | BIT(7) | BUILD_BUG_ON_ZERO(x >= 0x10000)), \
+	(((x) >> 2) & 0x7f)
+#define END() 0
+{
+	const u32 base = engine->mmio_base;
+
+	while (*data) {
+		u8 count, flags;
+
+		if (*data & BIT(7)) { /* skip */
+			regs += *data++ & ~BIT(7);
+			continue;
+		}
+
+		count = *data & 0x3f;
+		flags = *data >> 6;
+		data++;
+
+		*regs = MI_LOAD_REGISTER_IMM(count);
+		if (flags & POSTED)
+			*regs |= MI_LRI_FORCE_POSTED;
+		if (INTEL_GEN(engine->i915) >= 11)
+			*regs |= MI_LRI_CS_MMIO;
+		regs++;
+
+		GEM_BUG_ON(!count);
+		do {
+			u32 offset = 0;
+			u8 v;
+
+			do {
+				v = *data++;
+				offset <<= 7;
+				offset |= v & ~BIT(7);
+			} while (v & BIT(7));
+
+			*regs = base + (offset << 2);
+			regs += 2;
+		} while (--count);
+	}
+
+	return regs;
+}
+
+static const u8 gen8_xcs_offsets[] = {
+	NOP(1),
+	LRI(11, 0),
+	REG16(0x244),
+	REG(0x034),
+	REG(0x030),
+	REG(0x038),
+	REG(0x03c),
+	REG(0x168),
+	REG(0x140),
+	REG(0x110),
+	REG(0x11c),
+	REG(0x114),
+	REG(0x118),
+
+	NOP(9),
+	LRI(9, 0),
+	REG16(0x3a8),
+	REG16(0x28c),
+	REG16(0x288),
+	REG16(0x284),
+	REG16(0x280),
+	REG16(0x27c),
+	REG16(0x278),
+	REG16(0x274),
+	REG16(0x270),
+
+	NOP(13),
+	LRI(2, 0),
+	REG16(0x200),
+	REG(0x028),
+
+	END(),
+};
+
+static const u8 gen9_xcs_offsets[] = {
+	NOP(1),
+	LRI(14, POSTED),
+	REG16(0x244),
+	REG(0x034),
+	REG(0x030),
+	REG(0x038),
+	REG(0x03c),
+	REG(0x168),
+	REG(0x140),
+	REG(0x110),
+	REG(0x11c),
+	REG(0x114),
+	REG(0x118),
+	REG(0x1c0),
+	REG(0x1c4),
+	REG(0x1c8),
+
+	NOP(3),
+	LRI(9, POSTED),
+	REG16(0x3a8),
+	REG16(0x28c),
+	REG16(0x288),
+	REG16(0x284),
+	REG16(0x280),
+	REG16(0x27c),
+	REG16(0x278),
+	REG16(0x274),
+	REG16(0x270),
+
+	NOP(13),
+	LRI(1, POSTED),
+	REG16(0x200),
+
+	NOP(13),
+	LRI(44, POSTED),
+	REG(0x028),
+	REG(0x09c),
+	REG(0x0c0),
+	REG(0x178),
+	REG(0x17c),
+	REG16(0x358),
+	REG(0x170),
+	REG(0x150),
+	REG(0x154),
+	REG(0x158),
+	REG16(0x41c),
+	REG16(0x600),
+	REG16(0x604),
+	REG16(0x608),
+	REG16(0x60c),
+	REG16(0x610),
+	REG16(0x614),
+	REG16(0x618),
+	REG16(0x61c),
+	REG16(0x620),
+	REG16(0x624),
+	REG16(0x628),
+	REG16(0x62c),
+	REG16(0x630),
+	REG16(0x634),
+	REG16(0x638),
+	REG16(0x63c),
+	REG16(0x640),
+	REG16(0x644),
+	REG16(0x648),
+	REG16(0x64c),
+	REG16(0x650),
+	REG16(0x654),
+	REG16(0x658),
+	REG16(0x65c),
+	REG16(0x660),
+	REG16(0x664),
+	REG16(0x668),
+	REG16(0x66c),
+	REG16(0x670),
+	REG16(0x674),
+	REG16(0x678),
+	REG16(0x67c),
+	REG(0x068),
+
+	END(),
+};
+
+static const u8 gen12_xcs_offsets[] = {
+	NOP(1),
+	LRI(13, POSTED),
+	REG16(0x244),
+	REG(0x034),
+	REG(0x030),
+	REG(0x038),
+	REG(0x03c),
+	REG(0x168),
+	REG(0x140),
+	REG(0x110),
+	REG(0x1c0),
+	REG(0x1c4),
+	REG(0x1c8),
+	REG(0x180),
+	REG16(0x2b4),
+
+	NOP(5),
+	LRI(9, POSTED),
+	REG16(0x3a8),
+	REG16(0x28c),
+	REG16(0x288),
+	REG16(0x284),
+	REG16(0x280),
+	REG16(0x27c),
+	REG16(0x278),
+	REG16(0x274),
+	REG16(0x270),
+
+	END(),
+};
+
+static const u8 gen8_rcs_offsets[] = {
+	NOP(1),
+	LRI(14, POSTED),
+	REG16(0x244),
+	REG(0x034),
+	REG(0x030),
+	REG(0x038),
+	REG(0x03c),
+	REG(0x168),
+	REG(0x140),
+	REG(0x110),
+	REG(0x11c),
+	REG(0x114),
+	REG(0x118),
+	REG(0x1c0),
+	REG(0x1c4),
+	REG(0x1c8),
+
+	NOP(3),
+	LRI(9, POSTED),
+	REG16(0x3a8),
+	REG16(0x28c),
+	REG16(0x288),
+	REG16(0x284),
+	REG16(0x280),
+	REG16(0x27c),
+	REG16(0x278),
+	REG16(0x274),
+	REG16(0x270),
+
+	NOP(13),
+	LRI(1, 0),
+	REG(0x0c8),
+
+	END(),
+};
+
+static const u8 gen11_rcs_offsets[] = {
+	NOP(1),
+	LRI(15, POSTED),
+	REG16(0x244),
+	REG(0x034),
+	REG(0x030),
+	REG(0x038),
+	REG(0x03c),
+	REG(0x168),
+	REG(0x140),
+	REG(0x110),
+	REG(0x11c),
+	REG(0x114),
+	REG(0x118),
+	REG(0x1c0),
+	REG(0x1c4),
+	REG(0x1c8),
+	REG(0x180),
+
+	NOP(1),
+	LRI(9, POSTED),
+	REG16(0x3a8),
+	REG16(0x28c),
+	REG16(0x288),
+	REG16(0x284),
+	REG16(0x280),
+	REG16(0x27c),
+	REG16(0x278),
+	REG16(0x274),
+	REG16(0x270),
+
+	LRI(1, POSTED),
+	REG(0x1b0),
+
+	NOP(10),
+	LRI(1, 0),
+	REG(0x0c8),
+
+	END(),
+};
+
+static const u8 gen12_rcs_offsets[] = {
+	NOP(1),
+	LRI(13, POSTED),
+	REG16(0x244),
+	REG(0x034),
+	REG(0x030),
+	REG(0x038),
+	REG(0x03c),
+	REG(0x168),
+	REG(0x140),
+	REG(0x110),
+	REG(0x1c0),
+	REG(0x1c4),
+	REG(0x1c8),
+	REG(0x180),
+	REG16(0x2b4),
+
+	NOP(5),
+	LRI(9, POSTED),
+	REG16(0x3a8),
+	REG16(0x28c),
+	REG16(0x288),
+	REG16(0x284),
+	REG16(0x280),
+	REG16(0x27c),
+	REG16(0x278),
+	REG16(0x274),
+	REG16(0x270),
+
+	LRI(3, POSTED),
+	REG(0x1b0),
+	REG16(0x5a8),
+	REG16(0x5ac),
+
+	NOP(6),
+	LRI(1, 0),
+	REG(0x0c8),
+
+	END(),
+};
+
+#undef END
+#undef REG16
+#undef REG
+#undef LRI
+#undef NOP
+
+static const u8 *reg_offsets(const struct intel_engine_cs *engine)
+{
+	/*
+	 * The gen12+ lists only have the registers we program in the basic
+	 * default state. We rely on the context image using relative
+	 * addressing to automatic fixup the register state between the
+	 * physical engines for virtual engine.
+	 */
+	GEM_BUG_ON(INTEL_GEN(engine->i915) >= 12 &&
+		   !intel_engine_has_relative_mmio(engine));
+
+	if (engine->class == RENDER_CLASS) {
+		if (INTEL_GEN(engine->i915) >= 12)
+			return gen12_rcs_offsets;
+		else if (INTEL_GEN(engine->i915) >= 11)
+			return gen11_rcs_offsets;
+		else
+			return gen8_rcs_offsets;
+	} else {
+		if (INTEL_GEN(engine->i915) >= 12)
+			return gen12_xcs_offsets;
+		else if (INTEL_GEN(engine->i915) >= 9)
+			return gen9_xcs_offsets;
+		else
+			return gen8_xcs_offsets;
+	}
+}
+
 static void unwind_wa_tail(struct i915_request *rq)
 {
 	rq->tail = intel_ring_wrap(rq->ring, rq->wa_tail - WA_TAIL_BYTES);
@@ -482,7 +862,6 @@ __unwind_incomplete_requests(struct intel_engine_cs *engine)
 	list_for_each_entry_safe_reverse(rq, rn,
 					 &engine->active.requests,
 					 sched.link) {
-		struct intel_engine_cs *owner;
 
 		if (i915_request_completed(rq))
 			continue; /* XXX */
@@ -497,8 +876,7 @@ __unwind_incomplete_requests(struct intel_engine_cs *engine)
 		 * engine so that it can be moved across onto another physical
 		 * engine as load dictates.
 		 */
-		owner = rq->hw_context->engine;
-		if (likely(owner == engine)) {
+		if (likely(rq->execution_mask == engine->mask)) {
 			GEM_BUG_ON(rq_prio(rq) == I915_PRIORITY_INVALID);
 			if (rq_prio(rq) != prio) {
 				prio = rq_prio(rq);
@@ -509,6 +887,8 @@ __unwind_incomplete_requests(struct intel_engine_cs *engine)
 			list_move(&rq->sched.link, pl);
 			active = rq;
 		} else {
+			struct intel_engine_cs *owner = rq->hw_context->engine;
+
 			/*
 			 * Decouple the virtual breadcrumb before moving it
 			 * back to the virtual engine -- we don't want the
@@ -518,7 +898,8 @@ __unwind_incomplete_requests(struct intel_engine_cs *engine)
 			 */
 			if (test_bit(DMA_FENCE_FLAG_ENABLE_SIGNAL_BIT,
 				     &rq->fence.flags)) {
-				spin_lock(&rq->lock);
+				spin_lock_nested(&rq->lock,
+						 SINGLE_DEPTH_NESTING);
 				i915_request_cancel_breadcrumb(rq);
 				spin_unlock(&rq->lock);
 			}
@@ -554,6 +935,114 @@ execlists_context_status_change(struct i915_request *rq, unsigned long status)
 				   status, rq);
 }
 
+static void intel_engine_context_in(struct intel_engine_cs *engine)
+{
+	unsigned long flags;
+
+	if (READ_ONCE(engine->stats.enabled) == 0)
+		return;
+
+	write_seqlock_irqsave(&engine->stats.lock, flags);
+
+	if (engine->stats.enabled > 0) {
+		if (engine->stats.active++ == 0)
+			engine->stats.start = ktime_get();
+		GEM_BUG_ON(engine->stats.active == 0);
+	}
+
+	write_sequnlock_irqrestore(&engine->stats.lock, flags);
+}
+
+static void intel_engine_context_out(struct intel_engine_cs *engine)
+{
+	unsigned long flags;
+
+	if (READ_ONCE(engine->stats.enabled) == 0)
+		return;
+
+	write_seqlock_irqsave(&engine->stats.lock, flags);
+
+	if (engine->stats.enabled > 0) {
+		ktime_t last;
+
+		if (engine->stats.active && --engine->stats.active == 0) {
+			/*
+			 * Decrement the active context count and in case GPU
+			 * is now idle add up to the running total.
+			 */
+			last = ktime_sub(ktime_get(), engine->stats.start);
+
+			engine->stats.total = ktime_add(engine->stats.total,
+							last);
+		} else if (engine->stats.active == 0) {
+			/*
+			 * After turning on engine stats, context out might be
+			 * the first event in which case we account from the
+			 * time stats gathering was turned on.
+			 */
+			last = ktime_sub(ktime_get(), engine->stats.enabled_at);
+
+			engine->stats.total = ktime_add(engine->stats.total,
+							last);
+		}
+	}
+
+	write_sequnlock_irqrestore(&engine->stats.lock, flags);
+}
+
+static void restore_default_state(struct intel_context *ce,
+				  struct intel_engine_cs *engine)
+{
+	u32 *regs = ce->lrc_reg_state;
+
+	if (engine->pinned_default_state)
+		memcpy(regs, /* skip restoring the vanilla PPHWSP */
+		       engine->pinned_default_state + LRC_STATE_PN * PAGE_SIZE,
+		       engine->context_size - PAGE_SIZE);
+
+	execlists_init_reg_state(regs, ce, engine, ce->ring, false);
+}
+
+static void reset_active(struct i915_request *rq,
+			 struct intel_engine_cs *engine)
+{
+	struct intel_context * const ce = rq->hw_context;
+	u32 head;
+
+	/*
+	 * The executing context has been cancelled. We want to prevent
+	 * further execution along this context and propagate the error on
+	 * to anything depending on its results.
+	 *
+	 * In __i915_request_submit(), we apply the -EIO and remove the
+	 * requests' payloads for any banned requests. But first, we must
+	 * rewind the context back to the start of the incomplete request so
+	 * that we do not jump back into the middle of the batch.
+	 *
+	 * We preserve the breadcrumbs and semaphores of the incomplete
+	 * requests so that inter-timeline dependencies (i.e other timelines)
+	 * remain correctly ordered. And we defer to __i915_request_submit()
+	 * so that all asynchronous waits are correctly handled.
+	 */
+	GEM_TRACE("%s(%s): { rq=%llx:%lld }\n",
+		  __func__, engine->name, rq->fence.context, rq->fence.seqno);
+
+	/* On resubmission of the active request, payload will be scrubbed */
+	if (i915_request_completed(rq))
+		head = rq->tail;
+	else
+		head = active_request(ce->timeline, rq)->head;
+	ce->ring->head = intel_ring_wrap(ce->ring, head);
+	intel_ring_update_space(ce->ring);
+
+	/* Scrub the context image to prevent replaying the previous batch */
+	restore_default_state(ce, engine);
+	__execlists_update_reg_state(ce, engine);
+
+	/* We've switched away, so this should be a no-op, but intent matters */
+	ce->lrc_desc |= CTX_DESC_FORCE_RESTORE;
+}
+
 static inline struct intel_engine_cs *
 __execlists_schedule_in(struct i915_request *rq)
 {
@@ -562,6 +1051,21 @@ __execlists_schedule_in(struct i915_request *rq)
 
 	intel_context_get(ce);
 
+	if (unlikely(i915_gem_context_is_banned(ce->gem_context)))
+		reset_active(rq, engine);
+
+	if (ce->tag) {
+		/* Use a fixed tag for OA and friends */
+		ce->lrc_desc |= (u64)ce->tag << 32;
+	} else {
+		/* We don't need a strict matching tag, just different values */
+		ce->lrc_desc &= ~GENMASK_ULL(47, 37);
+		ce->lrc_desc |=
+			(u64)(engine->context_tag++ % NUM_CONTEXT_TAG) <<
+			GEN11_SW_CTX_ID_SHIFT;
+		BUILD_BUG_ON(NUM_CONTEXT_TAG > GEN12_MAX_CONTEXT_HW_ID);
+	}
+
 	intel_gt_pm_get(engine->gt);
 	execlists_context_status_change(rq, INTEL_CONTEXT_SCHEDULE_IN);
 	intel_engine_context_in(engine);
@@ -605,6 +1109,12 @@ __execlists_schedule_out(struct i915_request *rq,
 {
 	struct intel_context * const ce = rq->hw_context;
 
+	/*
+	 * NB process_csb() is not under the engine->active.lock and hence
+	 * schedule_out can race with schedule_in meaning that we should
+	 * refrain from doing non-trivial work here.
+	 */
+
 	intel_engine_context_out(engine);
 	execlists_context_status_change(rq, INTEL_CONTEXT_SCHEDULE_OUT);
 	intel_gt_pm_put(engine->gt);
@@ -631,7 +1141,6 @@ execlists_schedule_out(struct i915_request *rq)
 	struct intel_engine_cs *cur, *old;
 
 	trace_i915_request_out(rq);
-	GEM_BUG_ON(intel_context_inflight(ce) != rq->engine);
 
 	old = READ_ONCE(ce->inflight);
 	do
@@ -648,7 +1157,7 @@ static u64 execlists_update_context(const struct i915_request *rq)
 	struct intel_context *ce = rq->hw_context;
 	u64 desc;
 
-	ce->lrc_reg_state[CTX_RING_TAIL + 1] =
+	ce->lrc_reg_state[CTX_RING_TAIL] =
 		intel_ring_set_tail(rq->ring, rq->tail);
 
 	/*
@@ -671,6 +1180,10 @@ static u64 execlists_update_context(const struct i915_request *rq)
 	desc = ce->lrc_desc;
 	ce->lrc_desc &= ~CTX_DESC_FORCE_RESTORE;
 
+	/* Wa_1607138340:tgl */
+	if (IS_TGL_REVID(rq->i915, TGL_REVID_A0, TGL_REVID_A0))
+		desc |= CTX_DESC_FORCE_RESTORE;
+
 	return desc;
 }
 
@@ -693,6 +1206,9 @@ trace_ports(const struct intel_engine_execlists *execlists,
 	const struct intel_engine_cs *engine =
 		container_of(execlists, typeof(*engine), execlists);
 
+	if (!ports[0])
+		return;
+
 	GEM_TRACE("%s: %s { %llx:%lld%s, %llx:%lld }\n",
 		  engine->name, msg,
 		  ports[0]->fence.context,
@@ -713,25 +1229,45 @@ assert_pending_valid(const struct intel_engine_execlists *execlists,
 
 	trace_ports(execlists, msg, execlists->pending);
 
-	if (!execlists->pending[0])
+	if (!execlists->pending[0]) {
+		GEM_TRACE_ERR("Nothing pending for promotion!\n");
 		return false;
+	}
 
-	if (execlists->pending[execlists_num_ports(execlists)])
+	if (execlists->pending[execlists_num_ports(execlists)]) {
+		GEM_TRACE_ERR("Excess pending[%d] for promotion!\n",
+			      execlists_num_ports(execlists));
 		return false;
+	}
 
 	for (port = execlists->pending; (rq = *port); port++) {
-		if (ce == rq->hw_context)
+		if (ce == rq->hw_context) {
+			GEM_TRACE_ERR("Duplicate context in pending[%zd]\n",
+				      port - execlists->pending);
 			return false;
+		}
 
 		ce = rq->hw_context;
 		if (i915_request_completed(rq))
 			continue;
 
-		if (i915_active_is_idle(&ce->active))
+		if (i915_active_is_idle(&ce->active)) {
+			GEM_TRACE_ERR("Inactive context in pending[%zd]\n",
+				      port - execlists->pending);
+			return false;
+		}
+
+		if (!i915_vma_is_pinned(ce->state)) {
+			GEM_TRACE_ERR("Unpinned context in pending[%zd]\n",
+				      port - execlists->pending);
 			return false;
+		}
 
-		if (!i915_vma_is_pinned(ce->state))
+		if (!i915_vma_is_pinned(ce->ring->vma)) {
+			GEM_TRACE_ERR("Unpinned ringbuffer in pending[%zd]\n",
+				      port - execlists->pending);
 			return false;
+		}
 	}
 
 	return ce;
@@ -797,6 +1333,21 @@ static bool can_merge_rq(const struct i915_request *prev,
 	GEM_BUG_ON(prev == next);
 	GEM_BUG_ON(!assert_priority_queue(prev, next));
 
+	/*
+	 * We do not submit known completed requests. Therefore if the next
+	 * request is already completed, we can pretend to merge it in
+	 * with the previous context (and we will skip updating the ELSP
+	 * and tracking). Thus hopefully keeping the ELSP full with active
+	 * contexts, despite the best efforts of preempt-to-busy to confuse
+	 * us.
+	 */
+	if (i915_request_completed(next))
+		return true;
+
+	if (unlikely((prev->flags ^ next->flags) &
+		     (I915_REQUEST_NOPREEMPT | I915_REQUEST_SENTINEL)))
+		return false;
+
 	if (!can_merge_ctx(prev->hw_context, next->hw_context))
 		return false;
 
@@ -806,47 +1357,7 @@ static bool can_merge_rq(const struct i915_request *prev,
 static void virtual_update_register_offsets(u32 *regs,
 					    struct intel_engine_cs *engine)
 {
-	u32 base = engine->mmio_base;
-
-	/* Must match execlists_init_reg_state()! */
-
-	regs[CTX_CONTEXT_CONTROL] =
-		i915_mmio_reg_offset(RING_CONTEXT_CONTROL(base));
-	regs[CTX_RING_HEAD] = i915_mmio_reg_offset(RING_HEAD(base));
-	regs[CTX_RING_TAIL] = i915_mmio_reg_offset(RING_TAIL(base));
-	regs[CTX_RING_BUFFER_START] = i915_mmio_reg_offset(RING_START(base));
-	regs[CTX_RING_BUFFER_CONTROL] = i915_mmio_reg_offset(RING_CTL(base));
-
-	regs[CTX_BB_HEAD_U] = i915_mmio_reg_offset(RING_BBADDR_UDW(base));
-	regs[CTX_BB_HEAD_L] = i915_mmio_reg_offset(RING_BBADDR(base));
-	regs[CTX_BB_STATE] = i915_mmio_reg_offset(RING_BBSTATE(base));
-	regs[CTX_SECOND_BB_HEAD_U] =
-		i915_mmio_reg_offset(RING_SBBADDR_UDW(base));
-	regs[CTX_SECOND_BB_HEAD_L] = i915_mmio_reg_offset(RING_SBBADDR(base));
-	regs[CTX_SECOND_BB_STATE] = i915_mmio_reg_offset(RING_SBBSTATE(base));
-
-	regs[CTX_CTX_TIMESTAMP] =
-		i915_mmio_reg_offset(RING_CTX_TIMESTAMP(base));
-	regs[CTX_PDP3_UDW] = i915_mmio_reg_offset(GEN8_RING_PDP_UDW(base, 3));
-	regs[CTX_PDP3_LDW] = i915_mmio_reg_offset(GEN8_RING_PDP_LDW(base, 3));
-	regs[CTX_PDP2_UDW] = i915_mmio_reg_offset(GEN8_RING_PDP_UDW(base, 2));
-	regs[CTX_PDP2_LDW] = i915_mmio_reg_offset(GEN8_RING_PDP_LDW(base, 2));
-	regs[CTX_PDP1_UDW] = i915_mmio_reg_offset(GEN8_RING_PDP_UDW(base, 1));
-	regs[CTX_PDP1_LDW] = i915_mmio_reg_offset(GEN8_RING_PDP_LDW(base, 1));
-	regs[CTX_PDP0_UDW] = i915_mmio_reg_offset(GEN8_RING_PDP_UDW(base, 0));
-	regs[CTX_PDP0_LDW] = i915_mmio_reg_offset(GEN8_RING_PDP_LDW(base, 0));
-
-	if (engine->class == RENDER_CLASS) {
-		regs[CTX_RCS_INDIRECT_CTX] =
-			i915_mmio_reg_offset(RING_INDIRECT_CTX(base));
-		regs[CTX_RCS_INDIRECT_CTX_OFFSET] =
-			i915_mmio_reg_offset(RING_INDIRECT_CTX_OFFSET(base));
-		regs[CTX_BB_PER_CTX_PTR] =
-			i915_mmio_reg_offset(RING_BB_PER_CTX_PTR(base));
-
-		regs[CTX_R_PWR_CLK_STATE] =
-			i915_mmio_reg_offset(GEN8_R_PWR_CLK_STATE);
-	}
+	set_offsets(regs, reg_offsets(engine), engine);
 }
 
 static bool virtual_matches(const struct virtual_engine *ve,
@@ -893,7 +1404,7 @@ static void virtual_xfer_breadcrumbs(struct virtual_engine *ve,
 static struct i915_request *
 last_active(const struct intel_engine_execlists *execlists)
 {
-	struct i915_request * const *last = execlists->active;
+	struct i915_request * const *last = READ_ONCE(execlists->active);
 
 	while (*last && i915_request_completed(*last))
 		last++;
@@ -961,7 +1472,7 @@ need_timeslice(struct intel_engine_cs *engine, const struct i915_request *rq)
 {
 	int hint;
 
-	if (!intel_engine_has_semaphores(engine))
+	if (!intel_engine_has_timeslices(engine))
 		return false;
 
 	if (list_is_last(&rq->sched.link, &engine->active.requests))
@@ -982,15 +1493,32 @@ switch_prio(struct intel_engine_cs *engine, const struct i915_request *rq)
 	return rq_prio(list_next_entry(rq, sched.link));
 }
 
-static bool
-enable_timeslice(const struct intel_engine_execlists *execlists)
+static inline unsigned long
+timeslice(const struct intel_engine_cs *engine)
 {
-	const struct i915_request *rq = *execlists->active;
+	return READ_ONCE(engine->props.timeslice_duration_ms);
+}
+
+static unsigned long
+active_timeslice(const struct intel_engine_cs *engine)
+{
+	const struct i915_request *rq = *engine->execlists.active;
 
 	if (i915_request_completed(rq))
-		return false;
+		return 0;
 
-	return execlists->switch_priority_hint >= effective_prio(rq);
+	if (engine->execlists.switch_priority_hint < effective_prio(rq))
+		return 0;
+
+	return timeslice(engine);
+}
+
+static void set_timeslice(struct intel_engine_cs *engine)
+{
+	if (!intel_engine_has_timeslices(engine))
+		return;
+
+	set_timer_ms(&engine->execlists.timer, active_timeslice(engine));
 }
 
 static void record_preemption(struct intel_engine_execlists *execlists)
@@ -998,6 +1526,30 @@ static void record_preemption(struct intel_engine_execlists *execlists)
 	(void)I915_SELFTEST_ONLY(execlists->preempt_hang.count++);
 }
 
+static unsigned long active_preempt_timeout(struct intel_engine_cs *engine)
+{
+	struct i915_request *rq;
+
+	rq = last_active(&engine->execlists);
+	if (!rq)
+		return 0;
+
+	/* Force a fast reset for terminated contexts (ignoring sysfs!) */
+	if (unlikely(i915_gem_context_is_banned(rq->gem_context)))
+		return 1;
+
+	return READ_ONCE(engine->props.preempt_timeout_ms);
+}
+
+static void set_preempt_timeout(struct intel_engine_cs *engine)
+{
+	if (!intel_engine_has_preempt_reset(engine))
+		return;
+
+	set_timer_ms(&engine->execlists.preempt,
+		     active_preempt_timeout(engine));
+}
+
 static void execlists_dequeue(struct intel_engine_cs *engine)
 {
 	struct intel_engine_execlists * const execlists = &engine->execlists;
@@ -1094,7 +1646,7 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 			last->hw_context->lrc_desc |= CTX_DESC_FORCE_RESTORE;
 			last = NULL;
 		} else if (need_timeslice(engine, last) &&
-			   !timer_pending(&engine->execlists.timer)) {
+			   timer_expired(&engine->execlists.timer)) {
 			GEM_TRACE("%s: expired last=%llx:%lld, prio=%d, hint=%d\n",
 				  engine->name,
 				  last->fence.context,
@@ -1130,8 +1682,18 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 			 * submission.
 			 */
 			if (!list_is_last(&last->sched.link,
-					  &engine->active.requests))
+					  &engine->active.requests)) {
+				/*
+				 * Even if ELSP[1] is occupied and not worthy
+				 * of timeslices, our queue might be.
+				 */
+				if (!execlists->timer.expires &&
+				    need_timeslice(engine, last))
+					set_timer_ms(&execlists->timer,
+						     timeslice(engine));
+
 				return;
+			}
 
 			/*
 			 * WaIdleLiteRestore:bdw,skl
@@ -1172,21 +1734,6 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 				continue;
 			}
 
-			if (i915_request_completed(rq)) {
-				ve->request = NULL;
-				ve->base.execlists.queue_priority_hint = INT_MIN;
-				rb_erase_cached(rb, &execlists->virtual);
-				RB_CLEAR_NODE(rb);
-
-				rq->engine = engine;
-				__i915_request_submit(rq);
-
-				spin_unlock(&ve->base.active.lock);
-
-				rb = rb_first_cached(&execlists->virtual);
-				continue;
-			}
-
 			if (last && !can_merge_rq(last, rq)) {
 				spin_unlock(&ve->base.active.lock);
 				return; /* leave this for another */
@@ -1214,7 +1761,10 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 				unsigned int n;
 
 				GEM_BUG_ON(READ_ONCE(ve->context.inflight));
-				virtual_update_register_offsets(regs, engine);
+
+				if (!intel_engine_has_relative_mmio(engine))
+					virtual_update_register_offsets(regs,
+									engine);
 
 				if (!list_empty(&ve->context.signals))
 					virtual_xfer_breadcrumbs(ve, engine);
@@ -1237,11 +1787,24 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 				GEM_BUG_ON(ve->siblings[0] != engine);
 			}
 
-			__i915_request_submit(rq);
-			if (!i915_request_completed(rq)) {
+			if (__i915_request_submit(rq)) {
 				submit = true;
 				last = rq;
 			}
+			i915_request_put(rq);
+
+			/*
+			 * Hmm, we have a bunch of virtual engine requests,
+			 * but the first one was already completed (thanks
+			 * preempt-to-busy!). Keep looking at the veng queue
+			 * until we have no more relevant requests (i.e.
+			 * the normal submit queue has higher priority).
+			 */
+			if (!submit) {
+				spin_unlock(&ve->base.active.lock);
+				rb = rb_first_cached(&execlists->virtual);
+				continue;
+			}
 		}
 
 		spin_unlock(&ve->base.active.lock);
@@ -1254,8 +1817,7 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 		int i;
 
 		priolist_for_each_request_consume(rq, rn, p, i) {
-			if (i915_request_completed(rq))
-				goto skip;
+			bool merge = true;
 
 			/*
 			 * Can we combine this request with the current port?
@@ -1285,6 +1847,9 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 				if (last->hw_context == rq->hw_context)
 					goto done;
 
+				if (i915_request_has_sentinel(last))
+					goto done;
+
 				/*
 				 * If GVT overrides us we only ever submit
 				 * port[0], leaving port[1] empty. Note that we
@@ -1296,14 +1861,23 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 				    ctx_single_port_submission(rq->hw_context))
 					goto done;
 
-				*port = execlists_schedule_in(last, port - execlists->pending);
-				port++;
+				merge = false;
 			}
 
-			last = rq;
-			submit = true;
-skip:
-			__i915_request_submit(rq);
+			if (__i915_request_submit(rq)) {
+				if (!merge) {
+					*port = execlists_schedule_in(last, port - execlists->pending);
+					port++;
+					last = NULL;
+				}
+
+				GEM_BUG_ON(last &&
+					   !can_merge_ctx(last->hw_context,
+							  rq->hw_context));
+
+				submit = true;
+				last = rq;
+			}
 		}
 
 		rb_erase_cached(&p->node, &execlists->queue);
@@ -1334,11 +1908,28 @@ done:
 
 	if (submit) {
 		*port = execlists_schedule_in(last, port - execlists->pending);
-		memset(port + 1, 0, (last_port - port) * sizeof(*port));
 		execlists->switch_priority_hint =
 			switch_prio(engine, *execlists->pending);
+
+		/*
+		 * Skip if we ended up with exactly the same set of requests,
+		 * e.g. trying to timeslice a pair of ordered contexts
+		 */
+		if (!memcmp(execlists->active, execlists->pending,
+			    (port - execlists->pending + 1) * sizeof(*port))) {
+			do
+				execlists_schedule_out(fetch_and_zero(port));
+			while (port-- != execlists->pending);
+
+			goto skip_submit;
+		}
+
+		memset(port + 1, 0, (last_port - port) * sizeof(*port));
 		execlists_submit_ports(engine);
+
+		set_preempt_timeout(engine);
 	} else {
+skip_submit:
 		ring_set_paused(engine, 0);
 	}
 }
@@ -1371,13 +1962,6 @@ reset_in_progress(const struct intel_engine_execlists *execlists)
 	return unlikely(!__tasklet_is_enabled(&execlists->tasklet));
 }
 
-enum csb_step {
-	CSB_NOP,
-	CSB_PROMOTE,
-	CSB_PREEMPT,
-	CSB_COMPLETE,
-};
-
 /*
  * Starting with Gen12, the status has a new format:
  *
@@ -1404,7 +1988,7 @@ enum csb_step {
  *     bits 47-57: sw context id of the lrc the GT switched away from
  *     bits 58-63: sw counter of the lrc the GT switched away from
  */
-static inline enum csb_step
+static inline bool
 gen12_csb_parse(const struct intel_engine_execlists *execlists, const u32 *csb)
 {
 	u32 lower_dw = csb[0];
@@ -1413,9 +1997,6 @@ gen12_csb_parse(const struct intel_engine_execlists *execlists, const u32 *csb)
 	bool ctx_away_valid = GEN12_CSB_CTX_VALID(upper_dw);
 	bool new_queue = lower_dw & GEN12_CTX_STATUS_SWITCHED_TO_NEW_QUEUE;
 
-	if (!ctx_away_valid && ctx_to_valid)
-		return CSB_PROMOTE;
-
 	/*
 	 * The context switch detail is not guaranteed to be 5 when a preemption
 	 * occurs, so we can't just check for that. The check below works for
@@ -1423,8 +2004,10 @@ gen12_csb_parse(const struct intel_engine_execlists *execlists, const u32 *csb)
 	 * instructions and lite-restore. Preempt-to-idle via the CTRL register
 	 * would require some extra handling, but we don't support that.
 	 */
-	if (new_queue && ctx_away_valid)
-		return CSB_PREEMPT;
+	if (!ctx_away_valid || new_queue) {
+		GEM_BUG_ON(!ctx_to_valid);
+		return true;
+	}
 
 	/*
 	 * switch detail = 5 is covered by the case above and we do not expect a
@@ -1432,30 +2015,13 @@ gen12_csb_parse(const struct intel_engine_execlists *execlists, const u32 *csb)
 	 * use polling mode.
 	 */
 	GEM_BUG_ON(GEN12_CTX_SWITCH_DETAIL(upper_dw));
-
-	if (*execlists->active) {
-		GEM_BUG_ON(!ctx_away_valid);
-		return CSB_COMPLETE;
-	}
-
-	return CSB_NOP;
+	return false;
 }
 
-static inline enum csb_step
+static inline bool
 gen8_csb_parse(const struct intel_engine_execlists *execlists, const u32 *csb)
 {
-	unsigned int status = *csb;
-
-	if (status & GEN8_CTX_STATUS_IDLE_ACTIVE)
-		return CSB_PROMOTE;
-
-	if (status & GEN8_CTX_STATUS_PREEMPTED)
-		return CSB_PREEMPT;
-
-	if (*execlists->active)
-		return CSB_COMPLETE;
-
-	return CSB_NOP;
+	return *csb & (GEN8_CTX_STATUS_IDLE_ACTIVE | GEN8_CTX_STATUS_PREEMPTED);
 }
 
 static void process_csb(struct intel_engine_cs *engine)
@@ -1465,7 +2031,14 @@ static void process_csb(struct intel_engine_cs *engine)
 	const u8 num_entries = execlists->csb_size;
 	u8 head, tail;
 
-	GEM_BUG_ON(USES_GUC_SUBMISSION(engine->i915));
+	/*
+	 * As we modify our execlists state tracking we require exclusive
+	 * access. Either we are inside the tasklet, or the tasklet is disabled
+	 * and we assume that is only inside the reset paths and so serialised.
+	 */
+	GEM_BUG_ON(!tasklet_is_locked(&execlists->tasklet) &&
+		   !reset_in_progress(execlists));
+	GEM_BUG_ON(!intel_engine_in_execlists_submission_mode(engine));
 
 	/*
 	 * Note that csb_write, csb_status may be either in HWSP or mmio.
@@ -1494,7 +2067,7 @@ static void process_csb(struct intel_engine_cs *engine)
 	rmb();
 
 	do {
-		enum csb_step csb_step;
+		bool promote;
 
 		if (++head == num_entries)
 			head = 0;
@@ -1522,20 +2095,19 @@ static void process_csb(struct intel_engine_cs *engine)
 			  buf[2 * head + 0], buf[2 * head + 1]);
 
 		if (INTEL_GEN(engine->i915) >= 12)
-			csb_step = gen12_csb_parse(execlists, buf + 2 * head);
+			promote = gen12_csb_parse(execlists, buf + 2 * head);
 		else
-			csb_step = gen8_csb_parse(execlists, buf + 2 * head);
+			promote = gen8_csb_parse(execlists, buf + 2 * head);
+		if (promote) {
+			if (!inject_preempt_hang(execlists))
+				ring_set_paused(engine, 0);
 
-		switch (csb_step) {
-		case CSB_PREEMPT: /* cancel old inflight, prepare for switch */
+			/* cancel old inflight, prepare for switch */
 			trace_ports(execlists, "preempted", execlists->active);
-
 			while (*execlists->active)
 				execlists_schedule_out(*execlists->active++);
 
-			/* fallthrough */
-		case CSB_PROMOTE: /* switch pending to inflight */
-			GEM_BUG_ON(*execlists->active);
+			/* switch pending to inflight */
 			GEM_BUG_ON(!assert_pending_valid(execlists, "promote"));
 			execlists->active =
 				memcpy(execlists->inflight,
@@ -1543,16 +2115,13 @@ static void process_csb(struct intel_engine_cs *engine)
 				       execlists_num_ports(execlists) *
 				       sizeof(*execlists->pending));
 
-			if (enable_timeslice(execlists))
-				mod_timer(&execlists->timer, jiffies + 1);
-
-			if (!inject_preempt_hang(execlists))
-				ring_set_paused(engine, 0);
+			set_timeslice(engine);
 
 			WRITE_ONCE(execlists->pending[0], NULL);
-			break;
+		} else {
+			GEM_BUG_ON(!*execlists->active);
 
-		case CSB_COMPLETE: /* port0 completed, advanced to port1 */
+			/* port0 completed, advanced to port1 */
 			trace_ports(execlists, "completed", execlists->active);
 
 			/*
@@ -1567,10 +2136,6 @@ static void process_csb(struct intel_engine_cs *engine)
 
 			GEM_BUG_ON(execlists->active - execlists->inflight >
 				   execlists_num_ports(execlists));
-			break;
-
-		case CSB_NOP:
-			break;
 		}
 	} while (head != tail);
 
@@ -1593,8 +2158,48 @@ static void process_csb(struct intel_engine_cs *engine)
 static void __execlists_submission_tasklet(struct intel_engine_cs *const engine)
 {
 	lockdep_assert_held(&engine->active.lock);
-	if (!engine->execlists.pending[0])
+	if (!engine->execlists.pending[0]) {
+		rcu_read_lock(); /* protect peeking at execlists->active */
 		execlists_dequeue(engine);
+		rcu_read_unlock();
+	}
+}
+
+static noinline void preempt_reset(struct intel_engine_cs *engine)
+{
+	const unsigned int bit = I915_RESET_ENGINE + engine->id;
+	unsigned long *lock = &engine->gt->reset.flags;
+
+	if (i915_modparams.reset < 3)
+		return;
+
+	if (test_and_set_bit(bit, lock))
+		return;
+
+	/* Mark this tasklet as disabled to avoid waiting for it to complete */
+	tasklet_disable_nosync(&engine->execlists.tasklet);
+
+	GEM_TRACE("%s: preempt timeout %lu+%ums\n",
+		  engine->name,
+		  READ_ONCE(engine->props.preempt_timeout_ms),
+		  jiffies_to_msecs(jiffies - engine->execlists.preempt.expires));
+	intel_engine_reset(engine, "preemption time out");
+
+	tasklet_enable(&engine->execlists.tasklet);
+	clear_and_wake_up_bit(bit, lock);
+}
+
+static bool preempt_timeout(const struct intel_engine_cs *const engine)
+{
+	const struct timer_list *t = &engine->execlists.preempt;
+
+	if (!CONFIG_DRM_I915_PREEMPT_TIMEOUT)
+		return false;
+
+	if (!timer_expired(t))
+		return false;
+
+	return READ_ONCE(engine->execlists.pending[0]);
 }
 
 /*
@@ -1604,23 +2209,39 @@ static void __execlists_submission_tasklet(struct intel_engine_cs *const engine)
 static void execlists_submission_tasklet(unsigned long data)
 {
 	struct intel_engine_cs * const engine = (struct intel_engine_cs *)data;
-	unsigned long flags;
+	bool timeout = preempt_timeout(engine);
 
 	process_csb(engine);
-	if (!READ_ONCE(engine->execlists.pending[0])) {
+	if (!READ_ONCE(engine->execlists.pending[0]) || timeout) {
+		unsigned long flags;
+
 		spin_lock_irqsave(&engine->active.lock, flags);
 		__execlists_submission_tasklet(engine);
 		spin_unlock_irqrestore(&engine->active.lock, flags);
+
+		/* Recheck after serialising with direct-submission */
+		if (timeout && preempt_timeout(engine))
+			preempt_reset(engine);
 	}
 }
 
-static void execlists_submission_timer(struct timer_list *timer)
+static void __execlists_kick(struct intel_engine_execlists *execlists)
 {
-	struct intel_engine_cs *engine =
-		from_timer(engine, timer, execlists.timer);
-
 	/* Kick the tasklet for some interrupt coalescing and reset handling */
-	tasklet_hi_schedule(&engine->execlists.tasklet);
+	tasklet_hi_schedule(&execlists->tasklet);
+}
+
+#define execlists_kick(t, member) \
+	__execlists_kick(container_of(t, struct intel_engine_execlists, member))
+
+static void execlists_timeslice(struct timer_list *timer)
+{
+	execlists_kick(timer, timer);
+}
+
+static void execlists_preempt(struct timer_list *timer)
+{
+	execlists_kick(timer, preempt);
 }
 
 static void queue_request(struct intel_engine_cs *engine,
@@ -1700,7 +2321,6 @@ set_redzone(void *vaddr, const struct intel_engine_cs *engine)
 	if (!IS_ENABLED(CONFIG_DRM_I915_DEBUG_GEM))
 		return;
 
-	vaddr += LRC_HEADER_PAGES * PAGE_SIZE;
 	vaddr += engine->context_size;
 
 	memset(vaddr, POISON_INUSE, I915_GTT_PAGE_SIZE);
@@ -1712,7 +2332,6 @@ check_redzone(const void *vaddr, const struct intel_engine_cs *engine)
 	if (!IS_ENABLED(CONFIG_DRM_I915_DEBUG_GEM))
 		return;
 
-	vaddr += LRC_HEADER_PAGES * PAGE_SIZE;
 	vaddr += engine->context_size;
 
 	if (memchr_inv(vaddr, POISON_INUSE, I915_GTT_PAGE_SIZE))
@@ -1726,14 +2345,13 @@ static void execlists_context_unpin(struct intel_context *ce)
 	check_redzone((void *)ce->lrc_reg_state - LRC_STATE_PN * PAGE_SIZE,
 		      ce->engine);
 
-	i915_gem_context_unpin_hw_id(ce->gem_context);
 	i915_gem_object_unpin_map(ce->state->obj);
 	intel_ring_reset(ce->ring, ce->ring->tail);
 }
 
 static void
-__execlists_update_reg_state(struct intel_context *ce,
-			     struct intel_engine_cs *engine)
+__execlists_update_reg_state(const struct intel_context *ce,
+			     const struct intel_engine_cs *engine)
 {
 	struct intel_ring *ring = ce->ring;
 	u32 *regs = ce->lrc_reg_state;
@@ -1741,16 +2359,16 @@ __execlists_update_reg_state(struct intel_context *ce,
 	GEM_BUG_ON(!intel_ring_offset_valid(ring, ring->head));
 	GEM_BUG_ON(!intel_ring_offset_valid(ring, ring->tail));
 
-	regs[CTX_RING_BUFFER_START + 1] = i915_ggtt_offset(ring->vma);
-	regs[CTX_RING_HEAD + 1] = ring->head;
-	regs[CTX_RING_TAIL + 1] = ring->tail;
+	regs[CTX_RING_BUFFER_START] = i915_ggtt_offset(ring->vma);
+	regs[CTX_RING_HEAD] = ring->head;
+	regs[CTX_RING_TAIL] = ring->tail;
 
 	/* RPCS */
 	if (engine->class == RENDER_CLASS) {
-		regs[CTX_R_PWR_CLK_STATE + 1] =
+		regs[CTX_R_PWR_CLK_STATE] =
 			intel_sseu_make_rpcs(engine->i915, &ce->sseu);
 
-		i915_oa_init_reg_state(engine, ce, regs);
+		i915_oa_init_reg_state(ce, engine);
 	}
 }
 
@@ -1776,18 +2394,12 @@ __execlists_context_pin(struct intel_context *ce,
 		goto unpin_active;
 	}
 
-	ret = i915_gem_context_pin_hw_id(ce->gem_context);
-	if (ret)
-		goto unpin_map;
-
 	ce->lrc_desc = lrc_descriptor(ce, engine);
 	ce->lrc_reg_state = vaddr + LRC_STATE_PN * PAGE_SIZE;
 	__execlists_update_reg_state(ce, engine);
 
 	return 0;
 
-unpin_map:
-	i915_gem_object_unpin_map(ce->state->obj);
 unpin_active:
 	intel_context_active_release(ce);
 err:
@@ -1843,7 +2455,7 @@ static int gen8_emit_init_breadcrumb(struct i915_request *rq)
 {
 	u32 *cs;
 
-	GEM_BUG_ON(!rq->timeline->has_initial_breadcrumb);
+	GEM_BUG_ON(!i915_request_timeline(rq)->has_initial_breadcrumb);
 
 	cs = intel_ring_begin(rq, 6);
 	if (IS_ERR(cs))
@@ -1859,7 +2471,7 @@ static int gen8_emit_init_breadcrumb(struct i915_request *rq)
 	*cs++ = MI_NOOP;
 
 	*cs++ = MI_STORE_DWORD_IMM_GEN4 | MI_USE_GGTT;
-	*cs++ = rq->timeline->hwsp_offset;
+	*cs++ = i915_request_timeline(rq)->hwsp_offset;
 	*cs++ = 0;
 	*cs++ = rq->fence.seqno - 1;
 
@@ -1871,60 +2483,6 @@ static int gen8_emit_init_breadcrumb(struct i915_request *rq)
 	return 0;
 }
 
-static int emit_pdps(struct i915_request *rq)
-{
-	const struct intel_engine_cs * const engine = rq->engine;
-	struct i915_ppgtt * const ppgtt = i915_vm_to_ppgtt(rq->hw_context->vm);
-	int err, i;
-	u32 *cs;
-
-	GEM_BUG_ON(intel_vgpu_active(rq->i915));
-
-	/*
-	 * Beware ye of the dragons, this sequence is magic!
-	 *
-	 * Small changes to this sequence can cause anything from
-	 * GPU hangs to forcewake errors and machine lockups!
-	 */
-
-	/* Flush any residual operations from the context load */
-	err = engine->emit_flush(rq, EMIT_FLUSH);
-	if (err)
-		return err;
-
-	/* Magic required to prevent forcewake errors! */
-	err = engine->emit_flush(rq, EMIT_INVALIDATE);
-	if (err)
-		return err;
-
-	cs = intel_ring_begin(rq, 4 * GEN8_3LVL_PDPES + 2);
-	if (IS_ERR(cs))
-		return PTR_ERR(cs);
-
-	/* Ensure the LRI have landed before we invalidate & continue */
-	*cs++ = MI_LOAD_REGISTER_IMM(2 * GEN8_3LVL_PDPES) | MI_LRI_FORCE_POSTED;
-	for (i = GEN8_3LVL_PDPES; i--; ) {
-		const dma_addr_t pd_daddr = i915_page_dir_dma_addr(ppgtt, i);
-		u32 base = engine->mmio_base;
-
-		*cs++ = i915_mmio_reg_offset(GEN8_RING_PDP_UDW(base, i));
-		*cs++ = upper_32_bits(pd_daddr);
-		*cs++ = i915_mmio_reg_offset(GEN8_RING_PDP_LDW(base, i));
-		*cs++ = lower_32_bits(pd_daddr);
-	}
-	*cs++ = MI_NOOP;
-
-	intel_ring_advance(rq, cs);
-
-	/* Be doubly sure the LRI have landed before proceeding */
-	err = engine->emit_flush(rq, EMIT_FLUSH);
-	if (err)
-		return err;
-
-	/* Re-invalidate the TLB for luck */
-	return engine->emit_flush(rq, EMIT_INVALIDATE);
-}
-
 static int execlists_request_alloc(struct i915_request *request)
 {
 	int ret;
@@ -1947,10 +2505,7 @@ static int execlists_request_alloc(struct i915_request *request)
 	 */
 
 	/* Unconditionally invalidate GPU caches and TLBs. */
-	if (i915_vm_is_4lvl(request->hw_context->vm))
-		ret = request->engine->emit_flush(request, EMIT_INVALIDATE);
-	else
-		ret = emit_pdps(request);
+	ret = request->engine->emit_flush(request, EMIT_INVALIDATE);
 	if (ret)
 		return ret;
 
@@ -2002,12 +2557,6 @@ gen8_emit_flush_coherentl3_wa(struct intel_engine_cs *engine, u32 *batch)
 	return batch;
 }
 
-static u32 slm_offset(struct intel_engine_cs *engine)
-{
-	return intel_gt_scratch_offset(engine->gt,
-				       INTEL_GT_SCRATCH_FIELD_CLEAR_SLM_WA);
-}
-
 /*
  * Typically we only have one indirect_ctx and per_ctx batch buffer which are
  * initialized at the beginning and shared across all contexts but this field
@@ -2036,10 +2585,10 @@ static u32 *gen8_init_indirectctx_bb(struct intel_engine_cs *engine, u32 *batch)
 	/* Actual scratch location is at 128 bytes offset */
 	batch = gen8_emit_pipe_control(batch,
 				       PIPE_CONTROL_FLUSH_L3 |
-				       PIPE_CONTROL_GLOBAL_GTT_IVB |
+				       PIPE_CONTROL_STORE_DATA_INDEX |
 				       PIPE_CONTROL_CS_STALL |
 				       PIPE_CONTROL_QW_WRITE,
-				       slm_offset(engine));
+				       LRC_PPHWSP_SCRATCH_ADDR);
 
 	*batch++ = MI_ARB_ON_OFF | MI_ARB_ENABLE;
 
@@ -2397,23 +2946,29 @@ static void reset_csb_pointers(struct intel_engine_cs *engine)
 			       &execlists->csb_status[reset_value]);
 }
 
-static struct i915_request *active_request(struct i915_request *rq)
+static int lrc_ring_mi_mode(const struct intel_engine_cs *engine)
 {
-	const struct list_head * const list = &rq->timeline->requests;
-	const struct intel_context * const ce = rq->hw_context;
-	struct i915_request *active = NULL;
-
-	list_for_each_entry_from_reverse(rq, list, link) {
-		if (i915_request_completed(rq))
-			break;
+	if (INTEL_GEN(engine->i915) >= 12)
+		return 0x60;
+	else if (INTEL_GEN(engine->i915) >= 9)
+		return 0x54;
+	else if (engine->class == RENDER_CLASS)
+		return 0x58;
+	else
+		return -1;
+}
 
-		if (rq->hw_context != ce)
-			break;
+static void __execlists_reset_reg_state(const struct intel_context *ce,
+					const struct intel_engine_cs *engine)
+{
+	u32 *regs = ce->lrc_reg_state;
+	int x;
 
-		active = rq;
+	x = lrc_ring_mi_mode(engine);
+	if (x != -1) {
+		regs[x + 1] &= ~STOP_RING;
+		regs[x + 1] |= STOP_RING << 16;
 	}
-
-	return active;
 }
 
 static void __execlists_reset(struct intel_engine_cs *engine, bool stalled)
@@ -2421,7 +2976,10 @@ static void __execlists_reset(struct intel_engine_cs *engine, bool stalled)
 	struct intel_engine_execlists * const execlists = &engine->execlists;
 	struct intel_context *ce;
 	struct i915_request *rq;
-	u32 *regs;
+
+	mb(); /* paranoia: read the CSB pointers from after the reset */
+	clflush(execlists->csb_write);
+	mb();
 
 	process_csb(engine); /* drain preemption events */
 
@@ -2437,16 +2995,23 @@ static void __execlists_reset(struct intel_engine_cs *engine, bool stalled)
 	if (!rq)
 		goto unwind;
 
+	/* We still have requests in-flight; the engine should be active */
+	GEM_BUG_ON(!intel_engine_pm_is_awake(engine));
+
 	ce = rq->hw_context;
-	GEM_BUG_ON(i915_active_is_idle(&ce->active));
 	GEM_BUG_ON(!i915_vma_is_pinned(ce->state));
-	rq = active_request(rq);
-	if (!rq) {
-		ce->ring->head = ce->ring->tail;
+
+	if (i915_request_completed(rq)) {
+		/* Idle context; tidy up the ring so we can restart afresh */
+		ce->ring->head = intel_ring_wrap(ce->ring, rq->tail);
 		goto out_replay;
 	}
 
+	/* Context has requests still in-flight; it should not be idle! */
+	GEM_BUG_ON(i915_active_is_idle(&ce->active));
+	rq = active_request(ce->timeline, rq);
 	ce->ring->head = intel_ring_wrap(ce->ring, rq->head);
+	GEM_BUG_ON(ce->ring->head == ce->ring->tail);
 
 	/*
 	 * If this request hasn't started yet, e.g. it is waiting on a
@@ -2486,19 +3051,16 @@ static void __execlists_reset(struct intel_engine_cs *engine, bool stalled)
 	 * future request will be after userspace has had the opportunity
 	 * to recreate its own state.
 	 */
-	regs = ce->lrc_reg_state;
-	if (engine->pinned_default_state) {
-		memcpy(regs, /* skip restoring the vanilla PPHWSP */
-		       engine->pinned_default_state + LRC_STATE_PN * PAGE_SIZE,
-		       engine->context_size - PAGE_SIZE);
-	}
-	execlists_init_reg_state(regs, ce, engine, ce->ring);
+	GEM_BUG_ON(!intel_context_is_pinned(ce));
+	restore_default_state(ce, engine);
 
 out_replay:
-	GEM_TRACE("%s replay {head:%04x, tail:%04x\n",
+	GEM_TRACE("%s replay {head:%04x, tail:%04x}\n",
 		  engine->name, ce->ring->head, ce->ring->tail);
 	intel_ring_update_space(ce->ring);
+	__execlists_reset_reg_state(ce, engine);
 	__execlists_update_reg_state(ce, engine);
+	ce->lrc_desc |= CTX_DESC_FORCE_RESTORE; /* paranoid: GPU was reset! */
 
 unwind:
 	/* Push back any incomplete requests for replay after the reset. */
@@ -2552,12 +3114,8 @@ static void execlists_cancel_requests(struct intel_engine_cs *engine)
 	__execlists_reset(engine, true);
 
 	/* Mark all executing requests as skipped. */
-	list_for_each_entry(rq, &engine->active.requests, sched.link) {
-		if (!i915_request_signaled(rq))
-			dma_fence_set_error(&rq->fence, -EIO);
-
-		i915_request_mark_complete(rq);
-	}
+	list_for_each_entry(rq, &engine->active.requests, sched.link)
+		mark_eio(rq);
 
 	/* Flush the queued requests to the timeline list (for retiring). */
 	while ((rb = rb_first_cached(&execlists->queue))) {
@@ -2565,10 +3123,8 @@ static void execlists_cancel_requests(struct intel_engine_cs *engine)
 		int i;
 
 		priolist_for_each_request_consume(rq, rn, p, i) {
-			list_del_init(&rq->sched.link);
+			mark_eio(rq);
 			__i915_request_submit(rq);
-			dma_fence_set_error(&rq->fence, -EIO);
-			i915_request_mark_complete(rq);
 		}
 
 		rb_erase_cached(&p->node, &execlists->queue);
@@ -2584,13 +3140,15 @@ static void execlists_cancel_requests(struct intel_engine_cs *engine)
 		RB_CLEAR_NODE(rb);
 
 		spin_lock(&ve->base.active.lock);
-		if (ve->request) {
-			ve->request->engine = engine;
-			__i915_request_submit(ve->request);
-			dma_fence_set_error(&ve->request->fence, -EIO);
-			i915_request_mark_complete(ve->request);
+		rq = fetch_and_zero(&ve->request);
+		if (rq) {
+			mark_eio(rq);
+
+			rq->engine = engine;
+			__i915_request_submit(rq);
+			i915_request_put(rq);
+
 			ve->base.execlists.queue_priority_hint = INT_MIN;
-			ve->request = NULL;
 		}
 		spin_unlock(&ve->base.active.lock);
 	}
@@ -2723,7 +3281,7 @@ static int gen8_emit_flush(struct i915_request *request, u32 mode)
 	}
 
 	*cs++ = cmd;
-	*cs++ = I915_GEM_HWS_SCRATCH_ADDR | MI_FLUSH_DW_USE_GTT;
+	*cs++ = LRC_PPHWSP_SCRATCH_ADDR;
 	*cs++ = 0; /* upper addr */
 	*cs++ = 0; /* value */
 	intel_ring_advance(request, cs);
@@ -2734,10 +3292,6 @@ static int gen8_emit_flush(struct i915_request *request, u32 mode)
 static int gen8_emit_flush_render(struct i915_request *request,
 				  u32 mode)
 {
-	struct intel_engine_cs *engine = request->engine;
-	u32 scratch_addr =
-		intel_gt_scratch_offset(engine->gt,
-					INTEL_GT_SCRATCH_FIELD_RENDER_FLUSH);
 	bool vf_flush_wa = false, dc_flush_wa = false;
 	u32 *cs, flags = 0;
 	int len;
@@ -2759,7 +3313,7 @@ static int gen8_emit_flush_render(struct i915_request *request,
 		flags |= PIPE_CONTROL_CONST_CACHE_INVALIDATE;
 		flags |= PIPE_CONTROL_STATE_CACHE_INVALIDATE;
 		flags |= PIPE_CONTROL_QW_WRITE;
-		flags |= PIPE_CONTROL_GLOBAL_GTT_IVB;
+		flags |= PIPE_CONTROL_STORE_DATA_INDEX;
 
 		/*
 		 * On GEN9: before VF_CACHE_INVALIDATE we need to emit a NULL
@@ -2792,7 +3346,7 @@ static int gen8_emit_flush_render(struct i915_request *request,
 		cs = gen8_emit_pipe_control(cs, PIPE_CONTROL_DC_FLUSH_ENABLE,
 					    0);
 
-	cs = gen8_emit_pipe_control(cs, flags, scratch_addr);
+	cs = gen8_emit_pipe_control(cs, flags, LRC_PPHWSP_SCRATCH_ADDR);
 
 	if (dc_flush_wa)
 		cs = gen8_emit_pipe_control(cs, PIPE_CONTROL_CS_STALL, 0);
@@ -2805,11 +3359,6 @@ static int gen8_emit_flush_render(struct i915_request *request,
 static int gen11_emit_flush_render(struct i915_request *request,
 				   u32 mode)
 {
-	struct intel_engine_cs *engine = request->engine;
-	const u32 scratch_addr =
-		intel_gt_scratch_offset(engine->gt,
-					INTEL_GT_SCRATCH_FIELD_RENDER_FLUSH);
-
 	if (mode & EMIT_FLUSH) {
 		u32 *cs;
 		u32 flags = 0;
@@ -2822,13 +3371,13 @@ static int gen11_emit_flush_render(struct i915_request *request,
 		flags |= PIPE_CONTROL_DC_FLUSH_ENABLE;
 		flags |= PIPE_CONTROL_FLUSH_ENABLE;
 		flags |= PIPE_CONTROL_QW_WRITE;
-		flags |= PIPE_CONTROL_GLOBAL_GTT_IVB;
+		flags |= PIPE_CONTROL_STORE_DATA_INDEX;
 
 		cs = intel_ring_begin(request, 6);
 		if (IS_ERR(cs))
 			return PTR_ERR(cs);
 
-		cs = gen8_emit_pipe_control(cs, flags, scratch_addr);
+		cs = gen8_emit_pipe_control(cs, flags, LRC_PPHWSP_SCRATCH_ADDR);
 		intel_ring_advance(request, cs);
 	}
 
@@ -2846,19 +3395,111 @@ static int gen11_emit_flush_render(struct i915_request *request,
 		flags |= PIPE_CONTROL_CONST_CACHE_INVALIDATE;
 		flags |= PIPE_CONTROL_STATE_CACHE_INVALIDATE;
 		flags |= PIPE_CONTROL_QW_WRITE;
-		flags |= PIPE_CONTROL_GLOBAL_GTT_IVB;
+		flags |= PIPE_CONTROL_STORE_DATA_INDEX;
 
 		cs = intel_ring_begin(request, 6);
 		if (IS_ERR(cs))
 			return PTR_ERR(cs);
 
-		cs = gen8_emit_pipe_control(cs, flags, scratch_addr);
+		cs = gen8_emit_pipe_control(cs, flags, LRC_PPHWSP_SCRATCH_ADDR);
 		intel_ring_advance(request, cs);
 	}
 
 	return 0;
 }
 
+static u32 preparser_disable(bool state)
+{
+	return MI_ARB_CHECK | 1 << 8 | state;
+}
+
+static int gen12_emit_flush_render(struct i915_request *request,
+				   u32 mode)
+{
+	if (mode & EMIT_FLUSH) {
+		u32 flags = 0;
+		u32 *cs;
+
+		flags |= PIPE_CONTROL_TILE_CACHE_FLUSH;
+		flags |= PIPE_CONTROL_RENDER_TARGET_CACHE_FLUSH;
+		flags |= PIPE_CONTROL_DEPTH_CACHE_FLUSH;
+		/* Wa_1409600907:tgl */
+		flags |= PIPE_CONTROL_DEPTH_STALL;
+		flags |= PIPE_CONTROL_DC_FLUSH_ENABLE;
+		flags |= PIPE_CONTROL_FLUSH_ENABLE;
+		flags |= PIPE_CONTROL_HDC_PIPELINE_FLUSH;
+
+		flags |= PIPE_CONTROL_STORE_DATA_INDEX;
+		flags |= PIPE_CONTROL_QW_WRITE;
+
+		flags |= PIPE_CONTROL_CS_STALL;
+
+		cs = intel_ring_begin(request, 6);
+		if (IS_ERR(cs))
+			return PTR_ERR(cs);
+
+		cs = gen8_emit_pipe_control(cs, flags, LRC_PPHWSP_SCRATCH_ADDR);
+		intel_ring_advance(request, cs);
+	}
+
+	if (mode & EMIT_INVALIDATE) {
+		u32 flags = 0;
+		u32 *cs;
+
+		flags |= PIPE_CONTROL_COMMAND_CACHE_INVALIDATE;
+		flags |= PIPE_CONTROL_TLB_INVALIDATE;
+		flags |= PIPE_CONTROL_INSTRUCTION_CACHE_INVALIDATE;
+		flags |= PIPE_CONTROL_TEXTURE_CACHE_INVALIDATE;
+		flags |= PIPE_CONTROL_VF_CACHE_INVALIDATE;
+		flags |= PIPE_CONTROL_CONST_CACHE_INVALIDATE;
+		flags |= PIPE_CONTROL_STATE_CACHE_INVALIDATE;
+		flags |= PIPE_CONTROL_L3_RO_CACHE_INVALIDATE;
+
+		flags |= PIPE_CONTROL_STORE_DATA_INDEX;
+		flags |= PIPE_CONTROL_QW_WRITE;
+
+		flags |= PIPE_CONTROL_CS_STALL;
+
+		cs = intel_ring_begin(request, 8);
+		if (IS_ERR(cs))
+			return PTR_ERR(cs);
+
+		/*
+		 * Prevent the pre-parser from skipping past the TLB
+		 * invalidate and loading a stale page for the batch
+		 * buffer / request payload.
+		 */
+		*cs++ = preparser_disable(true);
+
+		cs = gen8_emit_pipe_control(cs, flags, LRC_PPHWSP_SCRATCH_ADDR);
+
+		*cs++ = preparser_disable(false);
+		intel_ring_advance(request, cs);
+
+		/*
+		 * Wa_1604544889:tgl
+		 */
+		if (IS_TGL_REVID(request->i915, TGL_REVID_A0, TGL_REVID_A0)) {
+			flags = 0;
+			flags |= PIPE_CONTROL_CS_STALL;
+			flags |= PIPE_CONTROL_HDC_PIPELINE_FLUSH;
+
+			flags |= PIPE_CONTROL_STORE_DATA_INDEX;
+			flags |= PIPE_CONTROL_QW_WRITE;
+
+			cs = intel_ring_begin(request, 6);
+			if (IS_ERR(cs))
+				return PTR_ERR(cs);
+
+			cs = gen8_emit_pipe_control(cs, flags,
+						    LRC_PPHWSP_SCRATCH_ADDR);
+			intel_ring_advance(request, cs);
+		}
+	}
+
+	return 0;
+}
+
 /*
  * Reserve space for 2 NOOPs at the end of each request to be
  * used as a workaround for not being allowed to do lite
@@ -2907,7 +3548,7 @@ static u32 *gen8_emit_fini_breadcrumb(struct i915_request *request, u32 *cs)
 {
 	cs = gen8_emit_ggtt_write(cs,
 				  request->fence.seqno,
-				  request->timeline->hwsp_offset,
+				  i915_request_active_timeline(request)->hwsp_offset,
 				  0);
 
 	return gen8_emit_fini_breadcrumb_footer(request, cs);
@@ -2915,28 +3556,28 @@ static u32 *gen8_emit_fini_breadcrumb(struct i915_request *request, u32 *cs)
 
 static u32 *gen8_emit_fini_breadcrumb_rcs(struct i915_request *request, u32 *cs)
 {
-	cs = gen8_emit_ggtt_write_rcs(cs,
-				      request->fence.seqno,
-				      request->timeline->hwsp_offset,
-				      PIPE_CONTROL_RENDER_TARGET_CACHE_FLUSH |
-				      PIPE_CONTROL_DEPTH_CACHE_FLUSH |
-				      PIPE_CONTROL_DC_FLUSH_ENABLE);
-
-	/* XXX flush+write+CS_STALL all in one upsets gem_concurrent_blt:kbl */
 	cs = gen8_emit_pipe_control(cs,
-				    PIPE_CONTROL_FLUSH_ENABLE |
-				    PIPE_CONTROL_CS_STALL,
+				    PIPE_CONTROL_RENDER_TARGET_CACHE_FLUSH |
+				    PIPE_CONTROL_DEPTH_CACHE_FLUSH |
+				    PIPE_CONTROL_DC_FLUSH_ENABLE,
 				    0);
 
+	/* XXX flush+write+CS_STALL all in one upsets gem_concurrent_blt:kbl */
+	cs = gen8_emit_ggtt_write_rcs(cs,
+				      request->fence.seqno,
+				      i915_request_active_timeline(request)->hwsp_offset,
+				      PIPE_CONTROL_FLUSH_ENABLE |
+				      PIPE_CONTROL_CS_STALL);
+
 	return gen8_emit_fini_breadcrumb_footer(request, cs);
 }
 
-static u32 *gen11_emit_fini_breadcrumb_rcs(struct i915_request *request,
-					   u32 *cs)
+static u32 *
+gen11_emit_fini_breadcrumb_rcs(struct i915_request *request, u32 *cs)
 {
 	cs = gen8_emit_ggtt_write_rcs(cs,
 				      request->fence.seqno,
-				      request->timeline->hwsp_offset,
+				      i915_request_active_timeline(request)->hwsp_offset,
 				      PIPE_CONTROL_CS_STALL |
 				      PIPE_CONTROL_TILE_CACHE_FLUSH |
 				      PIPE_CONTROL_RENDER_TARGET_CACHE_FLUSH |
@@ -2947,9 +3588,88 @@ static u32 *gen11_emit_fini_breadcrumb_rcs(struct i915_request *request,
 	return gen8_emit_fini_breadcrumb_footer(request, cs);
 }
 
+/*
+ * Note that the CS instruction pre-parser will not stall on the breadcrumb
+ * flush and will continue pre-fetching the instructions after it before the
+ * memory sync is completed. On pre-gen12 HW, the pre-parser will stop at
+ * BB_START/END instructions, so, even though we might pre-fetch the pre-amble
+ * of the next request before the memory has been flushed, we're guaranteed that
+ * we won't access the batch itself too early.
+ * However, on gen12+ the parser can pre-fetch across the BB_START/END commands,
+ * so, if the current request is modifying an instruction in the next request on
+ * the same intel_context, we might pre-fetch and then execute the pre-update
+ * instruction. To avoid this, the users of self-modifying code should either
+ * disable the parser around the code emitting the memory writes, via a new flag
+ * added to MI_ARB_CHECK, or emit the writes from a different intel_context. For
+ * the in-kernel use-cases we've opted to use a separate context, see
+ * reloc_gpu() as an example.
+ * All the above applies only to the instructions themselves. Non-inline data
+ * used by the instructions is not pre-fetched.
+ */
+
+static u32 *gen12_emit_preempt_busywait(struct i915_request *request, u32 *cs)
+{
+	*cs++ = MI_SEMAPHORE_WAIT_TOKEN |
+		MI_SEMAPHORE_GLOBAL_GTT |
+		MI_SEMAPHORE_POLL |
+		MI_SEMAPHORE_SAD_EQ_SDD;
+	*cs++ = 0;
+	*cs++ = intel_hws_preempt_address(request->engine);
+	*cs++ = 0;
+	*cs++ = 0;
+	*cs++ = MI_NOOP;
+
+	return cs;
+}
+
+static __always_inline u32*
+gen12_emit_fini_breadcrumb_footer(struct i915_request *request, u32 *cs)
+{
+	*cs++ = MI_USER_INTERRUPT;
+
+	*cs++ = MI_ARB_ON_OFF | MI_ARB_ENABLE;
+	if (intel_engine_has_semaphores(request->engine))
+		cs = gen12_emit_preempt_busywait(request, cs);
+
+	request->tail = intel_ring_offset(request, cs);
+	assert_ring_tail_valid(request->ring, request->tail);
+
+	return gen8_emit_wa_tail(request, cs);
+}
+
+static u32 *gen12_emit_fini_breadcrumb(struct i915_request *request, u32 *cs)
+{
+	cs = gen8_emit_ggtt_write(cs,
+				  request->fence.seqno,
+				  i915_request_active_timeline(request)->hwsp_offset,
+				  0);
+
+	return gen12_emit_fini_breadcrumb_footer(request, cs);
+}
+
+static u32 *
+gen12_emit_fini_breadcrumb_rcs(struct i915_request *request, u32 *cs)
+{
+	cs = gen8_emit_ggtt_write_rcs(cs,
+				      request->fence.seqno,
+				      i915_request_active_timeline(request)->hwsp_offset,
+				      PIPE_CONTROL_CS_STALL |
+				      PIPE_CONTROL_TILE_CACHE_FLUSH |
+				      PIPE_CONTROL_RENDER_TARGET_CACHE_FLUSH |
+				      PIPE_CONTROL_DEPTH_CACHE_FLUSH |
+				      /* Wa_1409600907:tgl */
+				      PIPE_CONTROL_DEPTH_STALL |
+				      PIPE_CONTROL_DC_FLUSH_ENABLE |
+				      PIPE_CONTROL_FLUSH_ENABLE |
+				      PIPE_CONTROL_HDC_PIPELINE_FLUSH);
+
+	return gen12_emit_fini_breadcrumb_footer(request, cs);
+}
+
 static void execlists_park(struct intel_engine_cs *engine)
 {
-	del_timer(&engine->execlists.timer);
+	cancel_timer(&engine->execlists.timer);
+	cancel_timer(&engine->execlists.preempt);
 }
 
 void intel_execlists_set_default_submission(struct intel_engine_cs *engine)
@@ -2972,6 +3692,9 @@ void intel_execlists_set_default_submission(struct intel_engine_cs *engine)
 		if (HAS_LOGICAL_RING_PREEMPTION(engine->i915))
 			engine->flags |= I915_ENGINE_HAS_PREEMPTION;
 	}
+
+	if (INTEL_GEN(engine->i915) >= 12)
+		engine->flags |= I915_ENGINE_HAS_RELATIVE_MMIO;
 }
 
 static void execlists_destroy(struct intel_engine_cs *engine)
@@ -2999,6 +3722,8 @@ logical_ring_default_vfuncs(struct intel_engine_cs *engine)
 	engine->emit_flush = gen8_emit_flush;
 	engine->emit_init_breadcrumb = gen8_emit_init_breadcrumb;
 	engine->emit_fini_breadcrumb = gen8_emit_fini_breadcrumb;
+	if (INTEL_GEN(engine->i915) >= 12)
+		engine->emit_fini_breadcrumb = gen12_emit_fini_breadcrumb;
 
 	engine->set_default_submission = intel_execlists_set_default_submission;
 
@@ -3044,6 +3769,9 @@ static void rcs_submission_override(struct intel_engine_cs *engine)
 {
 	switch (INTEL_GEN(engine->i915)) {
 	case 12:
+		engine->emit_flush = gen12_emit_flush_render;
+		engine->emit_fini_breadcrumb = gen12_emit_fini_breadcrumb_rcs;
+		break;
 	case 11:
 		engine->emit_flush = gen11_emit_flush_render;
 		engine->emit_fini_breadcrumb = gen11_emit_fini_breadcrumb_rcs;
@@ -3059,7 +3787,8 @@ int intel_execlists_submission_setup(struct intel_engine_cs *engine)
 {
 	tasklet_init(&engine->execlists.tasklet,
 		     execlists_submission_tasklet, (unsigned long)engine);
-	timer_setup(&engine->execlists.timer, execlists_submission_timer, 0);
+	timer_setup(&engine->execlists.timer, execlists_timeslice, 0);
+	timer_setup(&engine->execlists.preempt, execlists_preempt, 0);
 
 	logical_ring_default_vfuncs(engine);
 	logical_ring_default_irqs(engine);
@@ -3116,7 +3845,7 @@ int intel_execlists_submission_init(struct intel_engine_cs *engine)
 	return 0;
 }
 
-static u32 intel_lr_indirect_ctx_offset(struct intel_engine_cs *engine)
+static u32 intel_lr_indirect_ctx_offset(const struct intel_engine_cs *engine)
 {
 	u32 indirect_ctx_offset;
 
@@ -3149,86 +3878,50 @@ static u32 intel_lr_indirect_ctx_offset(struct intel_engine_cs *engine)
 	return indirect_ctx_offset;
 }
 
-static void execlists_init_reg_state(u32 *regs,
-				     struct intel_context *ce,
-				     struct intel_engine_cs *engine,
-				     struct intel_ring *ring)
-{
-	struct i915_ppgtt *ppgtt = i915_vm_to_ppgtt(ce->vm);
-	bool rcs = engine->class == RENDER_CLASS;
-	u32 base = engine->mmio_base;
-
-	/*
-	 * A context is actually a big batch buffer with several
-	 * MI_LOAD_REGISTER_IMM commands followed by (reg, value) pairs. The
-	 * values we are setting here are only for the first context restore:
-	 * on a subsequent save, the GPU will recreate this batchbuffer with new
-	 * values (including all the missing MI_LOAD_REGISTER_IMM commands that
-	 * we are not initializing here).
-	 *
-	 * Must keep consistent with virtual_update_register_offsets().
-	 */
-	regs[CTX_LRI_HEADER_0] = MI_LOAD_REGISTER_IMM(rcs ? 14 : 11) |
-				 MI_LRI_FORCE_POSTED;
 
-	CTX_REG(regs, CTX_CONTEXT_CONTROL, RING_CONTEXT_CONTROL(base),
+static void init_common_reg_state(u32 * const regs,
+				  const struct intel_engine_cs *engine,
+				  const struct intel_ring *ring)
+{
+	regs[CTX_CONTEXT_CONTROL] =
 		_MASKED_BIT_DISABLE(CTX_CTRL_ENGINE_CTX_RESTORE_INHIBIT) |
-		_MASKED_BIT_ENABLE(CTX_CTRL_INHIBIT_SYN_CTX_SWITCH));
-	if (INTEL_GEN(engine->i915) < 11) {
-		regs[CTX_CONTEXT_CONTROL + 1] |=
+		_MASKED_BIT_ENABLE(CTX_CTRL_INHIBIT_SYN_CTX_SWITCH);
+	if (INTEL_GEN(engine->i915) < 11)
+		regs[CTX_CONTEXT_CONTROL] |=
 			_MASKED_BIT_DISABLE(CTX_CTRL_ENGINE_CTX_SAVE_INHIBIT |
 					    CTX_CTRL_RS_CTX_ENABLE);
-	}
-	CTX_REG(regs, CTX_RING_HEAD, RING_HEAD(base), 0);
-	CTX_REG(regs, CTX_RING_TAIL, RING_TAIL(base), 0);
-	CTX_REG(regs, CTX_RING_BUFFER_START, RING_START(base), 0);
-	CTX_REG(regs, CTX_RING_BUFFER_CONTROL, RING_CTL(base),
-		RING_CTL_SIZE(ring->size) | RING_VALID);
-	CTX_REG(regs, CTX_BB_HEAD_U, RING_BBADDR_UDW(base), 0);
-	CTX_REG(regs, CTX_BB_HEAD_L, RING_BBADDR(base), 0);
-	CTX_REG(regs, CTX_BB_STATE, RING_BBSTATE(base), RING_BB_PPGTT);
-	CTX_REG(regs, CTX_SECOND_BB_HEAD_U, RING_SBBADDR_UDW(base), 0);
-	CTX_REG(regs, CTX_SECOND_BB_HEAD_L, RING_SBBADDR(base), 0);
-	CTX_REG(regs, CTX_SECOND_BB_STATE, RING_SBBSTATE(base), 0);
-	if (rcs) {
-		struct i915_ctx_workarounds *wa_ctx = &engine->wa_ctx;
-
-		CTX_REG(regs, CTX_RCS_INDIRECT_CTX, RING_INDIRECT_CTX(base), 0);
-		CTX_REG(regs, CTX_RCS_INDIRECT_CTX_OFFSET,
-			RING_INDIRECT_CTX_OFFSET(base), 0);
-		if (wa_ctx->indirect_ctx.size) {
-			u32 ggtt_offset = i915_ggtt_offset(wa_ctx->vma);
-
-			regs[CTX_RCS_INDIRECT_CTX + 1] =
-				(ggtt_offset + wa_ctx->indirect_ctx.offset) |
-				(wa_ctx->indirect_ctx.size / CACHELINE_BYTES);
-
-			regs[CTX_RCS_INDIRECT_CTX_OFFSET + 1] =
-				intel_lr_indirect_ctx_offset(engine) << 6;
-		}
 
-		CTX_REG(regs, CTX_BB_PER_CTX_PTR, RING_BB_PER_CTX_PTR(base), 0);
-		if (wa_ctx->per_ctx.size) {
-			u32 ggtt_offset = i915_ggtt_offset(wa_ctx->vma);
+	regs[CTX_RING_BUFFER_CONTROL] = RING_CTL_SIZE(ring->size) | RING_VALID;
+	regs[CTX_BB_STATE] = RING_BB_PPGTT;
+}
 
-			regs[CTX_BB_PER_CTX_PTR + 1] =
-				(ggtt_offset + wa_ctx->per_ctx.offset) | 0x01;
-		}
+static void init_wa_bb_reg_state(u32 * const regs,
+				 const struct intel_engine_cs *engine,
+				 u32 pos_bb_per_ctx)
+{
+	const struct i915_ctx_workarounds * const wa_ctx = &engine->wa_ctx;
+
+	if (wa_ctx->per_ctx.size) {
+		const u32 ggtt_offset = i915_ggtt_offset(wa_ctx->vma);
+
+		regs[pos_bb_per_ctx] =
+			(ggtt_offset + wa_ctx->per_ctx.offset) | 0x01;
 	}
 
-	regs[CTX_LRI_HEADER_1] = MI_LOAD_REGISTER_IMM(9) | MI_LRI_FORCE_POSTED;
+	if (wa_ctx->indirect_ctx.size) {
+		const u32 ggtt_offset = i915_ggtt_offset(wa_ctx->vma);
 
-	CTX_REG(regs, CTX_CTX_TIMESTAMP, RING_CTX_TIMESTAMP(base), 0);
-	/* PDP values well be assigned later if needed */
-	CTX_REG(regs, CTX_PDP3_UDW, GEN8_RING_PDP_UDW(base, 3), 0);
-	CTX_REG(regs, CTX_PDP3_LDW, GEN8_RING_PDP_LDW(base, 3), 0);
-	CTX_REG(regs, CTX_PDP2_UDW, GEN8_RING_PDP_UDW(base, 2), 0);
-	CTX_REG(regs, CTX_PDP2_LDW, GEN8_RING_PDP_LDW(base, 2), 0);
-	CTX_REG(regs, CTX_PDP1_UDW, GEN8_RING_PDP_UDW(base, 1), 0);
-	CTX_REG(regs, CTX_PDP1_LDW, GEN8_RING_PDP_LDW(base, 1), 0);
-	CTX_REG(regs, CTX_PDP0_UDW, GEN8_RING_PDP_UDW(base, 0), 0);
-	CTX_REG(regs, CTX_PDP0_LDW, GEN8_RING_PDP_LDW(base, 0), 0);
+		regs[pos_bb_per_ctx + 2] =
+			(ggtt_offset + wa_ctx->indirect_ctx.offset) |
+			(wa_ctx->indirect_ctx.size / CACHELINE_BYTES);
 
+		regs[pos_bb_per_ctx + 4] =
+			intel_lr_indirect_ctx_offset(engine) << 6;
+	}
+}
+
+static void init_ppgtt_reg_state(u32 *regs, const struct i915_ppgtt *ppgtt)
+{
 	if (i915_vm_is_4lvl(&ppgtt->vm)) {
 		/* 64b PPGTT (48bit canonical)
 		 * PDP0_DESCRIPTOR contains the base address to PML4 and
@@ -3241,15 +3934,47 @@ static void execlists_init_reg_state(u32 *regs,
 		ASSIGN_CTX_PDP(ppgtt, regs, 1);
 		ASSIGN_CTX_PDP(ppgtt, regs, 0);
 	}
+}
+
+static struct i915_ppgtt *vm_alias(struct i915_address_space *vm)
+{
+	if (i915_is_ggtt(vm))
+		return i915_vm_to_ggtt(vm)->alias;
+	else
+		return i915_vm_to_ppgtt(vm);
+}
 
-	if (rcs) {
-		regs[CTX_LRI_HEADER_2] = MI_LOAD_REGISTER_IMM(1);
-		CTX_REG(regs, CTX_R_PWR_CLK_STATE, GEN8_R_PWR_CLK_STATE, 0);
+static void execlists_init_reg_state(u32 *regs,
+				     const struct intel_context *ce,
+				     const struct intel_engine_cs *engine,
+				     const struct intel_ring *ring,
+				     bool close)
+{
+	/*
+	 * A context is actually a big batch buffer with several
+	 * MI_LOAD_REGISTER_IMM commands followed by (reg, value) pairs. The
+	 * values we are setting here are only for the first context restore:
+	 * on a subsequent save, the GPU will recreate this batchbuffer with new
+	 * values (including all the missing MI_LOAD_REGISTER_IMM commands that
+	 * we are not initializing here).
+	 *
+	 * Must keep consistent with virtual_update_register_offsets().
+	 */
+	u32 *bbe = set_offsets(regs, reg_offsets(engine), engine);
+
+	if (close) { /* Close the batch; used mainly by live_lrc_layout() */
+		*bbe = MI_BATCH_BUFFER_END;
+		if (INTEL_GEN(engine->i915) >= 10)
+			*bbe |= BIT(0);
 	}
 
-	regs[CTX_END] = MI_BATCH_BUFFER_END;
-	if (INTEL_GEN(engine->i915) >= 10)
-		regs[CTX_END] |= BIT(0);
+	init_common_reg_state(regs, engine, ring);
+	init_ppgtt_reg_state(regs, vm_alias(ce->vm));
+
+	init_wa_bb_reg_state(regs, engine,
+			     INTEL_GEN(engine->i915) >= 12 ?
+			     GEN12_CTX_BB_PER_CTX_PTR :
+			     CTX_BB_PER_CTX_PTR);
 }
 
 static int
@@ -3258,6 +3983,7 @@ populate_lr_context(struct intel_context *ce,
 		    struct intel_engine_cs *engine,
 		    struct intel_ring *ring)
 {
+	bool inhibit = true;
 	void *vaddr;
 	u32 *regs;
 	int ret;
@@ -3272,12 +3998,6 @@ populate_lr_context(struct intel_context *ce,
 	set_redzone(vaddr, engine);
 
 	if (engine->default_state) {
-		/*
-		 * We only want to copy over the template context state;
-		 * skipping over the headers reserved for GuC communication,
-		 * leaving those as zero.
-		 */
-		const unsigned long start = LRC_HEADER_PAGES * PAGE_SIZE;
 		void *defaults;
 
 		defaults = i915_gem_object_pin_map(engine->default_state,
@@ -3287,23 +4007,22 @@ populate_lr_context(struct intel_context *ce,
 			goto err_unpin_ctx;
 		}
 
-		memcpy(vaddr + start, defaults + start, engine->context_size);
+		memcpy(vaddr, defaults, engine->context_size);
 		i915_gem_object_unpin_map(engine->default_state);
+		inhibit = false;
 	}
 
 	/* The second page of the context object contains some fields which must
 	 * be set up prior to the first execution. */
 	regs = vaddr + LRC_STATE_PN * PAGE_SIZE;
-	execlists_init_reg_state(regs, ce, engine, ring);
-	if (!engine->default_state)
-		regs[CTX_CONTEXT_CONTROL + 1] |=
+	execlists_init_reg_state(regs, ce, engine, ring, inhibit);
+	if (inhibit)
+		regs[CTX_CONTEXT_CONTROL] |=
 			_MASKED_BIT_ENABLE(CTX_CTRL_ENGINE_CTX_RESTORE_INHIBIT);
 
 	ret = 0;
 err_unpin_ctx:
-	__i915_gem_object_flush_map(ctx_obj,
-				    LRC_HEADER_PAGES * PAGE_SIZE,
-				    engine->context_size);
+	__i915_gem_object_flush_map(ctx_obj, 0, engine->context_size);
 	i915_gem_object_unpin_map(ctx_obj);
 	return ret;
 }
@@ -3320,11 +4039,6 @@ static int __execlists_context_alloc(struct intel_context *ce,
 	GEM_BUG_ON(ce->state);
 	context_size = round_up(engine->context_size, I915_GTT_PAGE_SIZE);
 
-	/*
-	 * Before the actual start of the context image, we insert a few pages
-	 * for our own use and for sharing with the GuC.
-	 */
-	context_size += LRC_HEADER_PAGES * PAGE_SIZE;
 	if (IS_ENABLED(CONFIG_DRM_I915_DEBUG_GEM))
 		context_size += I915_GTT_PAGE_SIZE; /* for redzone */
 
@@ -3436,8 +4150,9 @@ static void virtual_engine_initial_hint(struct virtual_engine *ve)
 		return;
 
 	swap(ve->siblings[swp], ve->siblings[0]);
-	virtual_update_register_offsets(ve->context.lrc_reg_state,
-					ve->siblings[0]);
+	if (!intel_engine_has_relative_mmio(ve->siblings[0]))
+		virtual_update_register_offsets(ve->context.lrc_reg_state,
+						ve->siblings[0]);
 }
 
 static int virtual_context_pin(struct intel_context *ce)
@@ -3594,6 +4309,8 @@ submit_engine:
 static void virtual_submit_request(struct i915_request *rq)
 {
 	struct virtual_engine *ve = to_virtual_engine(rq->engine);
+	struct i915_request *old;
+	unsigned long flags;
 
 	GEM_TRACE("%s: rq=%llx:%lld\n",
 		  ve->base.name,
@@ -3602,15 +4319,31 @@ static void virtual_submit_request(struct i915_request *rq)
 
 	GEM_BUG_ON(ve->base.submit_request != virtual_submit_request);
 
-	GEM_BUG_ON(ve->request);
-	GEM_BUG_ON(!list_empty(virtual_queue(ve)));
+	spin_lock_irqsave(&ve->base.active.lock, flags);
 
-	ve->base.execlists.queue_priority_hint = rq_prio(rq);
-	WRITE_ONCE(ve->request, rq);
+	old = ve->request;
+	if (old) { /* background completion event from preempt-to-busy */
+		GEM_BUG_ON(!i915_request_completed(old));
+		__i915_request_submit(old);
+		i915_request_put(old);
+	}
+
+	if (i915_request_completed(rq)) {
+		__i915_request_submit(rq);
+
+		ve->base.execlists.queue_priority_hint = INT_MIN;
+		ve->request = NULL;
+	} else {
+		ve->base.execlists.queue_priority_hint = rq_prio(rq);
+		ve->request = i915_request_get(rq);
 
-	list_move_tail(&rq->sched.link, virtual_queue(ve));
+		GEM_BUG_ON(!list_empty(virtual_queue(ve)));
+		list_move_tail(&rq->sched.link, virtual_queue(ve));
 
-	tasklet_schedule(&ve->base.execlists.tasklet);
+		tasklet_schedule(&ve->base.execlists.tasklet);
+	}
+
+	spin_unlock_irqrestore(&ve->base.active.lock, flags);
 }
 
 static struct ve_bond *
@@ -3631,18 +4364,22 @@ static void
 virtual_bond_execute(struct i915_request *rq, struct dma_fence *signal)
 {
 	struct virtual_engine *ve = to_virtual_engine(rq->engine);
+	intel_engine_mask_t allowed, exec;
 	struct ve_bond *bond;
 
+	allowed = ~to_request(signal)->engine->mask;
+
 	bond = virtual_find_bond(ve, to_request(signal)->engine);
-	if (bond) {
-		intel_engine_mask_t old, new, cmp;
+	if (bond)
+		allowed &= bond->sibling_mask;
 
-		cmp = READ_ONCE(rq->execution_mask);
-		do {
-			old = cmp;
-			new = cmp & bond->sibling_mask;
-		} while ((cmp = cmpxchg(&rq->execution_mask, old, new)) != old);
-	}
+	/* Restrict the bonded request to run on only the available engines */
+	exec = READ_ONCE(rq->execution_mask);
+	while (!try_cmpxchg(&rq->execution_mask, &exec, exec & allowed))
+		;
+
+	/* Prevent the master from being re-run on the bonded engines */
+	to_request(signal)->execution_mask &= ~allowed;
 }
 
 struct intel_context *
@@ -3666,6 +4403,7 @@ intel_execlists_create_virtual(struct i915_gem_context *ctx,
 
 	ve->base.i915 = ctx->i915;
 	ve->base.gt = siblings[0]->gt;
+	ve->base.uncore = siblings[0]->uncore;
 	ve->base.id = -1;
 	ve->base.class = OTHER_CLASS;
 	ve->base.uabi_class = I915_ENGINE_CLASS_INVALID;
@@ -3689,6 +4427,7 @@ intel_execlists_create_virtual(struct i915_gem_context *ctx,
 	snprintf(ve->base.name, sizeof(ve->base.name), "virtual");
 
 	intel_engine_init_active(&ve->base, ENGINE_VIRTUAL);
+	intel_engine_init_breadcrumbs(&ve->base);
 
 	intel_engine_init_execlists(&ve->base);
 
@@ -3851,6 +4590,18 @@ int intel_virtual_engine_attach_bond(struct intel_engine_cs *engine,
 	return 0;
 }
 
+struct intel_engine_cs *
+intel_virtual_engine_get_sibling(struct intel_engine_cs *engine,
+				 unsigned int sibling)
+{
+	struct virtual_engine *ve = to_virtual_engine(engine);
+
+	if (sibling >= ve->num_siblings)
+		return NULL;
+
+	return ve->siblings[sibling];
+}
+
 void intel_execlists_show_requests(struct intel_engine_cs *engine,
 				   struct drm_printer *m,
 				   void (*show_request)(struct drm_printer *m,
@@ -3939,6 +4690,8 @@ void intel_lr_context_reset(struct intel_engine_cs *engine,
 			    u32 head,
 			    bool scrub)
 {
+	GEM_BUG_ON(!intel_context_is_pinned(ce));
+
 	/*
 	 * We want a simple context + ring to execute the breadcrumb update.
 	 * We cannot rely on the context being intact across the GPU hang,
@@ -3947,16 +4700,8 @@ void intel_lr_context_reset(struct intel_engine_cs *engine,
 	 * future request will be after userspace has had the opportunity
 	 * to recreate its own state.
 	 */
-	if (scrub) {
-		u32 *regs = ce->lrc_reg_state;
-
-		if (engine->pinned_default_state) {
-			memcpy(regs, /* skip restoring the vanilla PPHWSP */
-			       engine->pinned_default_state + LRC_STATE_PN * PAGE_SIZE,
-			       engine->context_size - PAGE_SIZE);
-		}
-		execlists_init_reg_state(regs, ce, engine, ce->ring);
-	}
+	if (scrub)
+		restore_default_state(ce, engine);
 
 	/* Rerun the request; its payload has been neutered (if guilty). */
 	ce->ring->head = head;
@@ -3965,6 +4710,13 @@ void intel_lr_context_reset(struct intel_engine_cs *engine,
 	__execlists_update_reg_state(ce, engine);
 }
 
+bool
+intel_engine_in_execlists_submission_mode(const struct intel_engine_cs *engine)
+{
+	return engine->set_default_submission ==
+	       intel_execlists_set_default_submission;
+}
+
 #if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
 #include "selftest_lrc.c"
 #endif
diff --git a/drivers/gpu/drm/i915/gt/intel_lrc.h b/drivers/gpu/drm/i915/gt/intel_lrc.h
index c2bba82bcc16..04511d8ebdc1 100644
--- a/drivers/gpu/drm/i915/gt/intel_lrc.h
+++ b/drivers/gpu/drm/i915/gt/intel_lrc.h
@@ -43,6 +43,7 @@ struct intel_engine_cs;
 #define	  CTX_CTRL_ENGINE_CTX_RESTORE_INHIBIT	(1 << 0)
 #define   CTX_CTRL_RS_CTX_ENABLE		(1 << 1)
 #define	  CTX_CTRL_ENGINE_CTX_SAVE_INHIBIT	(1 << 2)
+#define	  GEN12_CTX_CTRL_OAR_CONTEXT_ENABLE	(1 << 8)
 #define RING_CONTEXT_STATUS_PTR(base)		_MMIO((base) + 0x3a0)
 #define RING_EXECLIST_SQ_CONTENTS(base)		_MMIO((base) + 0x510)
 #define RING_EXECLIST_CONTROL(base)		_MMIO((base) + 0x550)
@@ -66,6 +67,12 @@ struct intel_engine_cs;
 #define GEN11_CSB_READ_PTR_MASK (GEN11_CSB_PTR_MASK << 8)
 #define GEN11_CSB_WRITE_PTR_MASK (GEN11_CSB_PTR_MASK << 0)
 
+#define MAX_CONTEXT_HW_ID (1<<21) /* exclusive */
+#define MAX_GUC_CONTEXT_HW_ID (1 << 20) /* exclusive */
+#define GEN11_MAX_CONTEXT_HW_ID (1<<11) /* exclusive */
+/* in Gen12 ID 0x7FF is reserved to indicate idle */
+#define GEN12_MAX_CONTEXT_HW_ID	(GEN11_MAX_CONTEXT_HW_ID - 1)
+
 enum {
 	INTEL_CONTEXT_SCHEDULE_IN = 0,
 	INTEL_CONTEXT_SCHEDULE_OUT,
@@ -79,30 +86,15 @@ int intel_execlists_submission_setup(struct intel_engine_cs *engine);
 int intel_execlists_submission_init(struct intel_engine_cs *engine);
 
 /* Logical Ring Contexts */
-
-/*
- * We allocate a header at the start of the context image for our own
- * use, therefore the actual location of the logical state is offset
- * from the start of the VMA. The layout is
- *
- * | [guc]          | [hwsp] [logical state] |
- * |<- our header ->|<- context image      ->|
- *
- */
-/* The first page is used for sharing data with the GuC */
-#define LRC_GUCSHR_PN	(0)
-#define LRC_GUCSHR_SZ	(1)
 /* At the start of the context image is its per-process HWS page */
-#define LRC_PPHWSP_PN	(LRC_GUCSHR_PN + LRC_GUCSHR_SZ)
+#define LRC_PPHWSP_PN	(0)
 #define LRC_PPHWSP_SZ	(1)
-/* Finally we have the logical state for the context */
+/* After the PPHWSP we have the logical state for the context */
 #define LRC_STATE_PN	(LRC_PPHWSP_PN + LRC_PPHWSP_SZ)
 
-/*
- * Currently we include the PPHWSP in __intel_engine_context_size() so
- * the size of the header is synonymous with the start of the PPHWSP.
- */
-#define LRC_HEADER_PAGES LRC_PPHWSP_PN
+/* Space within PPHWSP reserved to be used as scratch */
+#define LRC_PPHWSP_SCRATCH		0x34
+#define LRC_PPHWSP_SCRATCH_ADDR		(LRC_PPHWSP_SCRATCH * sizeof(u32))
 
 void intel_execlists_set_default_submission(struct intel_engine_cs *engine);
 
@@ -131,4 +123,11 @@ int intel_virtual_engine_attach_bond(struct intel_engine_cs *engine,
 				     const struct intel_engine_cs *master,
 				     const struct intel_engine_cs *sibling);
 
+struct intel_engine_cs *
+intel_virtual_engine_get_sibling(struct intel_engine_cs *engine,
+				 unsigned int sibling);
+
+bool
+intel_engine_in_execlists_submission_mode(const struct intel_engine_cs *engine);
+
 #endif /* _INTEL_LRC_H_ */
diff --git a/drivers/gpu/drm/i915/gt/intel_lrc_reg.h b/drivers/gpu/drm/i915/gt/intel_lrc_reg.h
index b8f20ad71169..06ab0276e10e 100644
--- a/drivers/gpu/drm/i915/gt/intel_lrc_reg.h
+++ b/drivers/gpu/drm/i915/gt/intel_lrc_reg.h
@@ -9,55 +9,41 @@
 
 #include <linux/types.h>
 
-/* GEN8+ Reg State Context */
-#define CTX_LRI_HEADER_0		0x01
-#define CTX_CONTEXT_CONTROL		0x02
-#define CTX_RING_HEAD			0x04
-#define CTX_RING_TAIL			0x06
-#define CTX_RING_BUFFER_START		0x08
-#define CTX_RING_BUFFER_CONTROL		0x0a
-#define CTX_BB_HEAD_U			0x0c
-#define CTX_BB_HEAD_L			0x0e
-#define CTX_BB_STATE			0x10
-#define CTX_SECOND_BB_HEAD_U		0x12
-#define CTX_SECOND_BB_HEAD_L		0x14
-#define CTX_SECOND_BB_STATE		0x16
-#define CTX_BB_PER_CTX_PTR		0x18
-#define CTX_RCS_INDIRECT_CTX		0x1a
-#define CTX_RCS_INDIRECT_CTX_OFFSET	0x1c
-#define CTX_LRI_HEADER_1		0x21
-#define CTX_CTX_TIMESTAMP		0x22
-#define CTX_PDP3_UDW			0x24
-#define CTX_PDP3_LDW			0x26
-#define CTX_PDP2_UDW			0x28
-#define CTX_PDP2_LDW			0x2a
-#define CTX_PDP1_UDW			0x2c
-#define CTX_PDP1_LDW			0x2e
-#define CTX_PDP0_UDW			0x30
-#define CTX_PDP0_LDW			0x32
-#define CTX_LRI_HEADER_2		0x41
-#define CTX_R_PWR_CLK_STATE		0x42
-#define CTX_END				0x44
-
-#define CTX_REG(reg_state, pos, reg, val) do { \
-	u32 *reg_state__ = (reg_state); \
-	const u32 pos__ = (pos); \
-	(reg_state__)[(pos__) + 0] = i915_mmio_reg_offset(reg); \
-	(reg_state__)[(pos__) + 1] = (val); \
-} while (0)
+/* GEN8 to GEN11 Reg State Context */
+#define CTX_CONTEXT_CONTROL		(0x02 + 1)
+#define CTX_RING_HEAD			(0x04 + 1)
+#define CTX_RING_TAIL			(0x06 + 1)
+#define CTX_RING_BUFFER_START		(0x08 + 1)
+#define CTX_RING_BUFFER_CONTROL		(0x0a + 1)
+#define CTX_BB_STATE			(0x10 + 1)
+#define CTX_BB_PER_CTX_PTR		(0x18 + 1)
+#define CTX_PDP3_UDW			(0x24 + 1)
+#define CTX_PDP3_LDW			(0x26 + 1)
+#define CTX_PDP2_UDW			(0x28 + 1)
+#define CTX_PDP2_LDW			(0x2a + 1)
+#define CTX_PDP1_UDW			(0x2c + 1)
+#define CTX_PDP1_LDW			(0x2e + 1)
+#define CTX_PDP0_UDW			(0x30 + 1)
+#define CTX_PDP0_LDW			(0x32 + 1)
+#define CTX_R_PWR_CLK_STATE		(0x42 + 1)
+
+#define GEN9_CTX_RING_MI_MODE		0x54
+
+/* GEN12+ Reg State Context */
+#define GEN12_CTX_BB_PER_CTX_PTR		(0x12 + 1)
 
 #define ASSIGN_CTX_PDP(ppgtt, reg_state, n) do { \
 	u32 *reg_state__ = (reg_state); \
 	const u64 addr__ = i915_page_dir_dma_addr((ppgtt), (n)); \
-	(reg_state__)[CTX_PDP ## n ## _UDW + 1] = upper_32_bits(addr__); \
-	(reg_state__)[CTX_PDP ## n ## _LDW + 1] = lower_32_bits(addr__); \
+	(reg_state__)[CTX_PDP ## n ## _UDW] = upper_32_bits(addr__); \
+	(reg_state__)[CTX_PDP ## n ## _LDW] = lower_32_bits(addr__); \
 } while (0)
 
 #define ASSIGN_CTX_PML4(ppgtt, reg_state) do { \
 	u32 *reg_state__ = (reg_state); \
 	const u64 addr__ = px_dma(ppgtt->pd); \
-	(reg_state__)[CTX_PDP0_UDW + 1] = upper_32_bits(addr__); \
-	(reg_state__)[CTX_PDP0_LDW + 1] = lower_32_bits(addr__); \
+	(reg_state__)[CTX_PDP0_UDW] = upper_32_bits(addr__); \
+	(reg_state__)[CTX_PDP0_LDW] = lower_32_bits(addr__); \
 } while (0)
 
 #define GEN8_CTX_RCS_INDIRECT_CTX_OFFSET_DEFAULT	0x17
diff --git a/drivers/gpu/drm/i915/gt/intel_mocs.c b/drivers/gpu/drm/i915/gt/intel_mocs.c
index 728704bbbe18..2b977991b785 100644
--- a/drivers/gpu/drm/i915/gt/intel_mocs.c
+++ b/drivers/gpu/drm/i915/gt/intel_mocs.c
@@ -26,6 +26,7 @@
 #include "intel_gt.h"
 #include "intel_mocs.h"
 #include "intel_lrc.h"
+#include "intel_ring.h"
 
 /* structures required */
 struct drm_i915_mocs_entry {
@@ -199,14 +200,6 @@ static const struct drm_i915_mocs_entry broxton_mocs_table[] = {
 	MOCS_ENTRY(15, \
 		   LE_3_WB | LE_TC_1_LLC | LE_LRUM(2) | LE_AOM(1), \
 		   L3_3_WB), \
-	/* Bypass LLC - Uncached (EHL+) */ \
-	MOCS_ENTRY(16, \
-		   LE_1_UC | LE_TC_1_LLC | LE_SCF(1), \
-		   L3_1_UC), \
-	/* Bypass LLC - L3 (Read-Only) (EHL+) */ \
-	MOCS_ENTRY(17, \
-		   LE_1_UC | LE_TC_1_LLC | LE_SCF(1), \
-		   L3_3_WB), \
 	/* Self-Snoop - L3 + LLC */ \
 	MOCS_ENTRY(18, \
 		   LE_3_WB | LE_TC_1_LLC | LE_LRUM(3) | LE_SSE(3), \
@@ -270,7 +263,7 @@ static const struct drm_i915_mocs_entry tigerlake_mocs_table[] = {
 		   L3_1_UC),
 	/* HW Special Case (Displayable) */
 	MOCS_ENTRY(61,
-		   LE_1_UC | LE_TC_1_LLC | LE_SCF(1),
+		   LE_1_UC | LE_TC_1_LLC,
 		   L3_3_WB),
 };
 
@@ -287,10 +280,9 @@ static const struct drm_i915_mocs_entry icelake_mocs_table[] = {
 	GEN11_MOCS_ENTRIES
 };
 
-static bool get_mocs_settings(struct intel_gt *gt,
+static bool get_mocs_settings(const struct drm_i915_private *i915,
 			      struct drm_i915_mocs_table *table)
 {
-	struct drm_i915_private *i915 = gt->i915;
 	bool result = false;
 
 	if (INTEL_GEN(i915) >= 12) {
@@ -331,9 +323,9 @@ static bool get_mocs_settings(struct intel_gt *gt,
 	return result;
 }
 
-static i915_reg_t mocs_register(enum intel_engine_id engine_id, int index)
+static i915_reg_t mocs_register(const struct intel_engine_cs *engine, int index)
 {
-	switch (engine_id) {
+	switch (engine->id) {
 	case RCS0:
 		return GEN9_GFX_MOCS(index);
 	case VCS0:
@@ -347,7 +339,7 @@ static i915_reg_t mocs_register(enum intel_engine_id engine_id, int index)
 	case VCS2:
 		return GEN11_MFX2_MOCS(index);
 	default:
-		MISSING_CASE(engine_id);
+		MISSING_CASE(engine->id);
 		return INVALID_MMIO_REG;
 	}
 }
@@ -365,118 +357,25 @@ static u32 get_entry_control(const struct drm_i915_mocs_table *table,
 	return table->table[I915_MOCS_PTE].control_value;
 }
 
-/**
- * intel_mocs_init_engine() - emit the mocs control table
- * @engine:	The engine for whom to emit the registers.
- *
- * This function simply emits a MI_LOAD_REGISTER_IMM command for the
- * given table starting at the given address.
- */
-void intel_mocs_init_engine(struct intel_engine_cs *engine)
+static void init_mocs_table(struct intel_engine_cs *engine,
+			    const struct drm_i915_mocs_table *table)
 {
-	struct intel_gt *gt = engine->gt;
-	struct intel_uncore *uncore = gt->uncore;
-	struct drm_i915_mocs_table table;
-	unsigned int index;
-	u32 unused_value;
-
-	/* Platforms with global MOCS do not need per-engine initialization. */
-	if (HAS_GLOBAL_MOCS_REGISTERS(gt->i915))
-		return;
-
-	/* Called under a blanket forcewake */
-	assert_forcewakes_active(uncore, FORCEWAKE_ALL);
-
-	if (!get_mocs_settings(gt, &table))
-		return;
-
-	/* Set unused values to PTE */
-	unused_value = table.table[I915_MOCS_PTE].control_value;
-
-	for (index = 0; index < table.size; index++) {
-		u32 value = get_entry_control(&table, index);
+	struct intel_uncore *uncore = engine->uncore;
+	u32 unused_value = table->table[I915_MOCS_PTE].control_value;
+	unsigned int i;
 
+	for (i = 0; i < table->size; i++)
 		intel_uncore_write_fw(uncore,
-				      mocs_register(engine->id, index),
-				      value);
-	}
+				      mocs_register(engine, i),
+				      get_entry_control(table, i));
 
-	/* All remaining entries are also unused */
-	for (; index < table.n_entries; index++)
+	/* All remaining entries are unused */
+	for (; i < table->n_entries; i++)
 		intel_uncore_write_fw(uncore,
-				      mocs_register(engine->id, index),
+				      mocs_register(engine, i),
 				      unused_value);
 }
 
-static void intel_mocs_init_global(struct intel_gt *gt)
-{
-	struct intel_uncore *uncore = gt->uncore;
-	struct drm_i915_mocs_table table;
-	unsigned int index;
-
-	GEM_BUG_ON(!HAS_GLOBAL_MOCS_REGISTERS(gt->i915));
-
-	if (!get_mocs_settings(gt, &table))
-		return;
-
-	if (GEM_DEBUG_WARN_ON(table.size > table.n_entries))
-		return;
-
-	for (index = 0; index < table.size; index++)
-		intel_uncore_write(uncore,
-				   GEN12_GLOBAL_MOCS(index),
-				   table.table[index].control_value);
-
-	/*
-	 * Ok, now set the unused entries to the invalid entry (index 0). These
-	 * entries are officially undefined and no contract for the contents and
-	 * settings is given for these entries.
-	 */
-	for (; index < table.n_entries; index++)
-		intel_uncore_write(uncore,
-				   GEN12_GLOBAL_MOCS(index),
-				   table.table[0].control_value);
-}
-
-static int emit_mocs_control_table(struct i915_request *rq,
-				   const struct drm_i915_mocs_table *table)
-{
-	enum intel_engine_id engine = rq->engine->id;
-	unsigned int index;
-	u32 unused_value;
-	u32 *cs;
-
-	if (GEM_WARN_ON(table->size > table->n_entries))
-		return -ENODEV;
-
-	/* Set unused values to PTE */
-	unused_value = table->table[I915_MOCS_PTE].control_value;
-
-	cs = intel_ring_begin(rq, 2 + 2 * table->n_entries);
-	if (IS_ERR(cs))
-		return PTR_ERR(cs);
-
-	*cs++ = MI_LOAD_REGISTER_IMM(table->n_entries);
-
-	for (index = 0; index < table->size; index++) {
-		u32 value = get_entry_control(table, index);
-
-		*cs++ = i915_mmio_reg_offset(mocs_register(engine, index));
-		*cs++ = value;
-	}
-
-	/* All remaining entries are also unused */
-	for (; index < table->n_entries; index++) {
-		*cs++ = i915_mmio_reg_offset(mocs_register(engine, index));
-		*cs++ = unused_value;
-	}
-
-	*cs++ = MI_NOOP;
-	intel_ring_advance(rq, cs);
-
-	return 0;
-}
-
 /*
  * Get l3cc_value from MOCS entry taking into account when it's not used:
  * I915_MOCS_PTE's value is returned in this case.
@@ -494,141 +393,99 @@ static inline u32 l3cc_combine(const struct drm_i915_mocs_table *table,
 			       u16 low,
 			       u16 high)
 {
-	return low | high << 16;
+	return low | (u32)high << 16;
 }
 
-static int emit_mocs_l3cc_table(struct i915_request *rq,
-				const struct drm_i915_mocs_table *table)
+static void init_l3cc_table(struct intel_engine_cs *engine,
+			    const struct drm_i915_mocs_table *table)
 {
-	u16 unused_value;
+	struct intel_uncore *uncore = engine->uncore;
+	u16 unused_value = table->table[I915_MOCS_PTE].l3cc_value;
 	unsigned int i;
-	u32 *cs;
-
-	if (GEM_WARN_ON(table->size > table->n_entries))
-		return -ENODEV;
-
-	/* Set unused values to PTE */
-	unused_value = table->table[I915_MOCS_PTE].l3cc_value;
-
-	cs = intel_ring_begin(rq, 2 + table->n_entries);
-	if (IS_ERR(cs))
-		return PTR_ERR(cs);
-
-	*cs++ = MI_LOAD_REGISTER_IMM(table->n_entries / 2);
 
 	for (i = 0; i < table->size / 2; i++) {
 		u16 low = get_entry_l3cc(table, 2 * i);
 		u16 high = get_entry_l3cc(table, 2 * i + 1);
 
-		*cs++ = i915_mmio_reg_offset(GEN9_LNCFCMOCS(i));
-		*cs++ = l3cc_combine(table, low, high);
+		intel_uncore_write(uncore,
+				   GEN9_LNCFCMOCS(i),
+				   l3cc_combine(table, low, high));
 	}
 
 	/* Odd table size - 1 left over */
-	if (table->size & 0x01) {
+	if (table->size & 1) {
 		u16 low = get_entry_l3cc(table, 2 * i);
 
-		*cs++ = i915_mmio_reg_offset(GEN9_LNCFCMOCS(i));
-		*cs++ = l3cc_combine(table, low, unused_value);
+		intel_uncore_write(uncore,
+				   GEN9_LNCFCMOCS(i),
+				   l3cc_combine(table, low, unused_value));
 		i++;
 	}
 
 	/* All remaining entries are also unused */
-	for (; i < table->n_entries / 2; i++) {
-		*cs++ = i915_mmio_reg_offset(GEN9_LNCFCMOCS(i));
-		*cs++ = l3cc_combine(table, unused_value, unused_value);
-	}
+	for (; i < table->n_entries / 2; i++)
+		intel_uncore_write(uncore,
+				   GEN9_LNCFCMOCS(i),
+				   l3cc_combine(table, unused_value,
+						unused_value));
+}
 
-	*cs++ = MI_NOOP;
-	intel_ring_advance(rq, cs);
+void intel_mocs_init_engine(struct intel_engine_cs *engine)
+{
+	struct drm_i915_mocs_table table;
+
+	/* Called under a blanket forcewake */
+	assert_forcewakes_active(engine->uncore, FORCEWAKE_ALL);
+
+	if (!get_mocs_settings(engine->i915, &table))
+		return;
+
+	/* Platforms with global MOCS do not need per-engine initialization. */
+	if (!HAS_GLOBAL_MOCS_REGISTERS(engine->i915))
+		init_mocs_table(engine, &table);
 
-	return 0;
+	if (engine->class == RENDER_CLASS)
+		init_l3cc_table(engine, &table);
 }
 
-static void intel_mocs_init_l3cc_table(struct intel_gt *gt)
+static void intel_mocs_init_global(struct intel_gt *gt)
 {
 	struct intel_uncore *uncore = gt->uncore;
 	struct drm_i915_mocs_table table;
-	unsigned int i;
-	u16 unused_value;
+	unsigned int index;
 
-	if (!get_mocs_settings(gt, &table))
+	/*
+	 * LLC and eDRAM control values are not applicable to dgfx
+	 */
+	if (IS_DGFX(gt->i915))
 		return;
 
-	/* Set unused values to PTE */
-	unused_value = table.table[I915_MOCS_PTE].l3cc_value;
-
-	for (i = 0; i < table.size / 2; i++) {
-		u16 low = get_entry_l3cc(&table, 2 * i);
-		u16 high = get_entry_l3cc(&table, 2 * i + 1);
+	GEM_BUG_ON(!HAS_GLOBAL_MOCS_REGISTERS(gt->i915));
 
-		intel_uncore_write(uncore,
-				   GEN9_LNCFCMOCS(i),
-				   l3cc_combine(&table, low, high));
-	}
+	if (!get_mocs_settings(gt->i915, &table))
+		return;
 
-	/* Odd table size - 1 left over */
-	if (table.size & 0x01) {
-		u16 low = get_entry_l3cc(&table, 2 * i);
+	if (GEM_DEBUG_WARN_ON(table.size > table.n_entries))
+		return;
 
+	for (index = 0; index < table.size; index++)
 		intel_uncore_write(uncore,
-				   GEN9_LNCFCMOCS(i),
-				   l3cc_combine(&table, low, unused_value));
-		i++;
-	}
+				   GEN12_GLOBAL_MOCS(index),
+				   table.table[index].control_value);
 
-	/* All remaining entries are also unused */
-	for (; i < table.n_entries / 2; i++)
+	/*
+	 * Ok, now set the unused entries to the invalid entry (index 0). These
+	 * entries are officially undefined and no contract for the contents and
+	 * settings is given for these entries.
+	 */
+	for (; index < table.n_entries; index++)
 		intel_uncore_write(uncore,
-				   GEN9_LNCFCMOCS(i),
-				   l3cc_combine(&table, unused_value,
-						unused_value));
-}
-
-/**
- * intel_mocs_emit() - program the MOCS register.
- * @rq:	Request to use to set up the MOCS tables.
- *
- * This function will emit a batch buffer with the values required for
- * programming the MOCS register values for all the currently supported
- * rings.
- *
- * These registers are partially stored in the RCS context, so they are
- * emitted at the same time so that when a context is created these registers
- * are set up. These registers have to be emitted into the start of the
- * context as setting the ELSP will re-init some of these registers back
- * to the hw values.
- *
- * Return: 0 on success, otherwise the error status.
- */
-int intel_mocs_emit(struct i915_request *rq)
-{
-	struct drm_i915_mocs_table t;
-	int ret;
-
-	if (HAS_GLOBAL_MOCS_REGISTERS(rq->i915) ||
-	    rq->engine->class != RENDER_CLASS)
-		return 0;
-
-	if (get_mocs_settings(rq->engine->gt, &t)) {
-		/* Program the RCS control registers */
-		ret = emit_mocs_control_table(rq, &t);
-		if (ret)
-			return ret;
-
-		/* Now program the l3cc registers */
-		ret = emit_mocs_l3cc_table(rq, &t);
-		if (ret)
-			return ret;
-	}
-
-	return 0;
+				   GEN12_GLOBAL_MOCS(index),
+				   table.table[0].control_value);
 }
 
 void intel_mocs_init(struct intel_gt *gt)
 {
-	intel_mocs_init_l3cc_table(gt);
-
 	if (HAS_GLOBAL_MOCS_REGISTERS(gt->i915))
 		intel_mocs_init_global(gt);
 }
diff --git a/drivers/gpu/drm/i915/gt/intel_mocs.h b/drivers/gpu/drm/i915/gt/intel_mocs.h
index 2ae816b7ca19..83371f3e6ba1 100644
--- a/drivers/gpu/drm/i915/gt/intel_mocs.h
+++ b/drivers/gpu/drm/i915/gt/intel_mocs.h
@@ -49,13 +49,10 @@
  * context handling keep the MOCS in step.
  */
 
-struct i915_request;
 struct intel_engine_cs;
 struct intel_gt;
 
 void intel_mocs_init(struct intel_gt *gt);
 void intel_mocs_init_engine(struct intel_engine_cs *engine);
 
-int intel_mocs_emit(struct i915_request *rq);
-
 #endif
diff --git a/drivers/gpu/drm/i915/gt/intel_rc6.c b/drivers/gpu/drm/i915/gt/intel_rc6.c
new file mode 100644
index 000000000000..700104b90163
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/intel_rc6.c
@@ -0,0 +1,787 @@
+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2019 Intel Corporation
+ */
+
+#include <linux/pm_runtime.h>
+
+#include "i915_drv.h"
+#include "intel_gt.h"
+#include "intel_gt_pm.h"
+#include "intel_rc6.h"
+#include "intel_sideband.h"
+
+/**
+ * DOC: RC6
+ *
+ * RC6 is a special power stage which allows the GPU to enter an very
+ * low-voltage mode when idle, using down to 0V while at this stage.  This
+ * stage is entered automatically when the GPU is idle when RC6 support is
+ * enabled, and as soon as new workload arises GPU wakes up automatically as
+ * well.
+ *
+ * There are different RC6 modes available in Intel GPU, which differentiate
+ * among each other with the latency required to enter and leave RC6 and
+ * voltage consumed by the GPU in different states.
+ *
+ * The combination of the following flags define which states GPU is allowed
+ * to enter, while RC6 is the normal RC6 state, RC6p is the deep RC6, and
+ * RC6pp is deepest RC6. Their support by hardware varies according to the
+ * GPU, BIOS, chipset and platform. RC6 is usually the safest one and the one
+ * which brings the most power savings; deeper states save more power, but
+ * require higher latency to switch to and wake up.
+ */
+
+static struct intel_gt *rc6_to_gt(struct intel_rc6 *rc6)
+{
+	return container_of(rc6, struct intel_gt, rc6);
+}
+
+static struct intel_uncore *rc6_to_uncore(struct intel_rc6 *rc)
+{
+	return rc6_to_gt(rc)->uncore;
+}
+
+static struct drm_i915_private *rc6_to_i915(struct intel_rc6 *rc)
+{
+	return rc6_to_gt(rc)->i915;
+}
+
+static inline void set(struct intel_uncore *uncore, i915_reg_t reg, u32 val)
+{
+	intel_uncore_write_fw(uncore, reg, val);
+}
+
+static void gen11_rc6_enable(struct intel_rc6 *rc6)
+{
+	struct intel_uncore *uncore = rc6_to_uncore(rc6);
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
+
+	/* 2b: Program RC6 thresholds.*/
+	set(uncore, GEN6_RC6_WAKE_RATE_LIMIT, 54 << 16 | 85);
+	set(uncore, GEN10_MEDIA_WAKE_RATE_LIMIT, 150);
+
+	set(uncore, GEN6_RC_EVALUATION_INTERVAL, 125000); /* 12500 * 1280ns */
+	set(uncore, GEN6_RC_IDLE_HYSTERSIS, 25); /* 25 * 1280ns */
+	for_each_engine(engine, rc6_to_gt(rc6), id)
+		set(uncore, RING_MAX_IDLE(engine->mmio_base), 10);
+
+	set(uncore, GUC_MAX_IDLE_COUNT, 0xA);
+
+	set(uncore, GEN6_RC_SLEEP, 0);
+
+	set(uncore, GEN6_RC6_THRESHOLD, 50000); /* 50/125ms per EI */
+
+	/*
+	 * 2c: Program Coarse Power Gating Policies.
+	 *
+	 * Bspec's guidance is to use 25us (really 25 * 1280ns) here. What we
+	 * use instead is a more conservative estimate for the maximum time
+	 * it takes us to service a CS interrupt and submit a new ELSP - that
+	 * is the time which the GPU is idle waiting for the CPU to select the
+	 * next request to execute. If the idle hysteresis is less than that
+	 * interrupt service latency, the hardware will automatically gate
+	 * the power well and we will then incur the wake up cost on top of
+	 * the service latency. A similar guide from plane_state is that we
+	 * do not want the enable hysteresis to less than the wakeup latency.
+	 *
+	 * igt/gem_exec_nop/sequential provides a rough estimate for the
+	 * service latency, and puts it around 10us for Broadwell (and other
+	 * big core) and around 40us for Broxton (and other low power cores).
+	 * [Note that for legacy ringbuffer submission, this is less than 1us!]
+	 * However, the wakeup latency on Broxton is closer to 100us. To be
+	 * conservative, we have to factor in a context switch on top (due
+	 * to ksoftirqd).
+	 */
+	set(uncore, GEN9_MEDIA_PG_IDLE_HYSTERESIS, 250);
+	set(uncore, GEN9_RENDER_PG_IDLE_HYSTERESIS, 250);
+
+	/* 3a: Enable RC6 */
+	set(uncore, GEN6_RC_CONTROL,
+	    GEN6_RC_CTL_HW_ENABLE |
+	    GEN6_RC_CTL_RC6_ENABLE |
+	    GEN6_RC_CTL_EI_MODE(1));
+
+	set(uncore, GEN9_PG_ENABLE,
+	    GEN9_RENDER_PG_ENABLE |
+	    GEN9_MEDIA_PG_ENABLE |
+	    GEN11_MEDIA_SAMPLER_PG_ENABLE);
+}
+
+static void gen9_rc6_enable(struct intel_rc6 *rc6)
+{
+	struct intel_uncore *uncore = rc6_to_uncore(rc6);
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
+	u32 rc6_mode;
+
+	/* 2b: Program RC6 thresholds.*/
+	if (INTEL_GEN(rc6_to_i915(rc6)) >= 10) {
+		set(uncore, GEN6_RC6_WAKE_RATE_LIMIT, 54 << 16 | 85);
+		set(uncore, GEN10_MEDIA_WAKE_RATE_LIMIT, 150);
+	} else if (IS_SKYLAKE(rc6_to_i915(rc6))) {
+		/*
+		 * WaRsDoubleRc6WrlWithCoarsePowerGating:skl Doubling WRL only
+		 * when CPG is enabled
+		 */
+		set(uncore, GEN6_RC6_WAKE_RATE_LIMIT, 108 << 16);
+	} else {
+		set(uncore, GEN6_RC6_WAKE_RATE_LIMIT, 54 << 16);
+	}
+
+	set(uncore, GEN6_RC_EVALUATION_INTERVAL, 125000); /* 12500 * 1280ns */
+	set(uncore, GEN6_RC_IDLE_HYSTERSIS, 25); /* 25 * 1280ns */
+	for_each_engine(engine, rc6_to_gt(rc6), id)
+		set(uncore, RING_MAX_IDLE(engine->mmio_base), 10);
+
+	set(uncore, GUC_MAX_IDLE_COUNT, 0xA);
+
+	set(uncore, GEN6_RC_SLEEP, 0);
+
+	/*
+	 * 2c: Program Coarse Power Gating Policies.
+	 *
+	 * Bspec's guidance is to use 25us (really 25 * 1280ns) here. What we
+	 * use instead is a more conservative estimate for the maximum time
+	 * it takes us to service a CS interrupt and submit a new ELSP - that
+	 * is the time which the GPU is idle waiting for the CPU to select the
+	 * next request to execute. If the idle hysteresis is less than that
+	 * interrupt service latency, the hardware will automatically gate
+	 * the power well and we will then incur the wake up cost on top of
+	 * the service latency. A similar guide from plane_state is that we
+	 * do not want the enable hysteresis to less than the wakeup latency.
+	 *
+	 * igt/gem_exec_nop/sequential provides a rough estimate for the
+	 * service latency, and puts it around 10us for Broadwell (and other
+	 * big core) and around 40us for Broxton (and other low power cores).
+	 * [Note that for legacy ringbuffer submission, this is less than 1us!]
+	 * However, the wakeup latency on Broxton is closer to 100us. To be
+	 * conservative, we have to factor in a context switch on top (due
+	 * to ksoftirqd).
+	 */
+	set(uncore, GEN9_MEDIA_PG_IDLE_HYSTERESIS, 250);
+	set(uncore, GEN9_RENDER_PG_IDLE_HYSTERESIS, 250);
+
+	/* 3a: Enable RC6 */
+	set(uncore, GEN6_RC6_THRESHOLD, 37500); /* 37.5/125ms per EI */
+
+	/* WaRsUseTimeoutMode:cnl (pre-prod) */
+	if (IS_CNL_REVID(rc6_to_i915(rc6), CNL_REVID_A0, CNL_REVID_C0))
+		rc6_mode = GEN7_RC_CTL_TO_MODE;
+	else
+		rc6_mode = GEN6_RC_CTL_EI_MODE(1);
+
+	set(uncore, GEN6_RC_CONTROL,
+	    GEN6_RC_CTL_HW_ENABLE |
+	    GEN6_RC_CTL_RC6_ENABLE |
+	    rc6_mode);
+
+	/*
+	 * WaRsDisableCoarsePowerGating:skl,cnl
+	 *   - Render/Media PG need to be disabled with RC6.
+	 */
+	if (!NEEDS_WaRsDisableCoarsePowerGating(rc6_to_i915(rc6)))
+		set(uncore, GEN9_PG_ENABLE,
+		    GEN9_RENDER_PG_ENABLE | GEN9_MEDIA_PG_ENABLE);
+}
+
+static void gen8_rc6_enable(struct intel_rc6 *rc6)
+{
+	struct intel_uncore *uncore = rc6_to_uncore(rc6);
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
+
+	/* 2b: Program RC6 thresholds.*/
+	set(uncore, GEN6_RC6_WAKE_RATE_LIMIT, 40 << 16);
+	set(uncore, GEN6_RC_EVALUATION_INTERVAL, 125000); /* 12500 * 1280ns */
+	set(uncore, GEN6_RC_IDLE_HYSTERSIS, 25); /* 25 * 1280ns */
+	for_each_engine(engine, rc6_to_gt(rc6), id)
+		set(uncore, RING_MAX_IDLE(engine->mmio_base), 10);
+	set(uncore, GEN6_RC_SLEEP, 0);
+	set(uncore, GEN6_RC6_THRESHOLD, 625); /* 800us/1.28 for TO */
+
+	/* 3: Enable RC6 */
+	set(uncore, GEN6_RC_CONTROL,
+	    GEN6_RC_CTL_HW_ENABLE |
+	    GEN7_RC_CTL_TO_MODE |
+	    GEN6_RC_CTL_RC6_ENABLE);
+}
+
+static void gen6_rc6_enable(struct intel_rc6 *rc6)
+{
+	struct intel_uncore *uncore = rc6_to_uncore(rc6);
+	struct drm_i915_private *i915 = rc6_to_i915(rc6);
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
+	u32 rc6vids, rc6_mask;
+	int ret;
+
+	set(uncore, GEN6_RC1_WAKE_RATE_LIMIT, 1000 << 16);
+	set(uncore, GEN6_RC6_WAKE_RATE_LIMIT, 40 << 16 | 30);
+	set(uncore, GEN6_RC6pp_WAKE_RATE_LIMIT, 30);
+	set(uncore, GEN6_RC_EVALUATION_INTERVAL, 125000);
+	set(uncore, GEN6_RC_IDLE_HYSTERSIS, 25);
+
+	for_each_engine(engine, rc6_to_gt(rc6), id)
+		set(uncore, RING_MAX_IDLE(engine->mmio_base), 10);
+
+	set(uncore, GEN6_RC_SLEEP, 0);
+	set(uncore, GEN6_RC1e_THRESHOLD, 1000);
+	if (IS_IVYBRIDGE(i915))
+		set(uncore, GEN6_RC6_THRESHOLD, 125000);
+	else
+		set(uncore, GEN6_RC6_THRESHOLD, 50000);
+	set(uncore, GEN6_RC6p_THRESHOLD, 150000);
+	set(uncore, GEN6_RC6pp_THRESHOLD, 64000); /* unused */
+
+	/* We don't use those on Haswell */
+	rc6_mask = GEN6_RC_CTL_RC6_ENABLE;
+	if (HAS_RC6p(i915))
+		rc6_mask |= GEN6_RC_CTL_RC6p_ENABLE;
+	if (HAS_RC6pp(i915))
+		rc6_mask |= GEN6_RC_CTL_RC6pp_ENABLE;
+	set(uncore, GEN6_RC_CONTROL,
+	    rc6_mask |
+	    GEN6_RC_CTL_EI_MODE(1) |
+	    GEN6_RC_CTL_HW_ENABLE);
+
+	rc6vids = 0;
+	ret = sandybridge_pcode_read(i915, GEN6_PCODE_READ_RC6VIDS,
+				     &rc6vids, NULL);
+	if (IS_GEN(i915, 6) && ret) {
+		DRM_DEBUG_DRIVER("Couldn't check for BIOS workaround\n");
+	} else if (IS_GEN(i915, 6) &&
+		   (GEN6_DECODE_RC6_VID(rc6vids & 0xff) < 450)) {
+		DRM_DEBUG_DRIVER("You should update your BIOS. Correcting minimum rc6 voltage (%dmV->%dmV)\n",
+				 GEN6_DECODE_RC6_VID(rc6vids & 0xff), 450);
+		rc6vids &= 0xffff00;
+		rc6vids |= GEN6_ENCODE_RC6_VID(450);
+		ret = sandybridge_pcode_write(i915, GEN6_PCODE_WRITE_RC6VIDS, rc6vids);
+		if (ret)
+			DRM_ERROR("Couldn't fix incorrect rc6 voltage\n");
+	}
+}
+
+/* Check that the pcbr address is not empty. */
+static int chv_rc6_init(struct intel_rc6 *rc6)
+{
+	struct intel_uncore *uncore = rc6_to_uncore(rc6);
+	resource_size_t pctx_paddr, paddr;
+	resource_size_t pctx_size = 32 * SZ_1K;
+	u32 pcbr;
+
+	pcbr = intel_uncore_read(uncore, VLV_PCBR);
+	if ((pcbr >> VLV_PCBR_ADDR_SHIFT) == 0) {
+		DRM_DEBUG_DRIVER("BIOS didn't set up PCBR, fixing up\n");
+		paddr = rc6_to_i915(rc6)->dsm.end + 1 - pctx_size;
+		GEM_BUG_ON(paddr > U32_MAX);
+
+		pctx_paddr = (paddr & ~4095);
+		intel_uncore_write(uncore, VLV_PCBR, pctx_paddr);
+	}
+
+	return 0;
+}
+
+static int vlv_rc6_init(struct intel_rc6 *rc6)
+{
+	struct drm_i915_private *i915 = rc6_to_i915(rc6);
+	struct intel_uncore *uncore = rc6_to_uncore(rc6);
+	struct drm_i915_gem_object *pctx;
+	resource_size_t pctx_paddr;
+	resource_size_t pctx_size = 24 * SZ_1K;
+	u32 pcbr;
+
+	pcbr = intel_uncore_read(uncore, VLV_PCBR);
+	if (pcbr) {
+		/* BIOS set it up already, grab the pre-alloc'd space */
+		resource_size_t pcbr_offset;
+
+		pcbr_offset = (pcbr & ~4095) - i915->dsm.start;
+		pctx = i915_gem_object_create_stolen_for_preallocated(i915,
+								      pcbr_offset,
+								      I915_GTT_OFFSET_NONE,
+								      pctx_size);
+		if (IS_ERR(pctx))
+			return PTR_ERR(pctx);
+
+		goto out;
+	}
+
+	DRM_DEBUG_DRIVER("BIOS didn't set up PCBR, fixing up\n");
+
+	/*
+	 * From the Gunit register HAS:
+	 * The Gfx driver is expected to program this register and ensure
+	 * proper allocation within Gfx stolen memory.  For example, this
+	 * register should be programmed such than the PCBR range does not
+	 * overlap with other ranges, such as the frame buffer, protected
+	 * memory, or any other relevant ranges.
+	 */
+	pctx = i915_gem_object_create_stolen(i915, pctx_size);
+	if (IS_ERR(pctx)) {
+		DRM_DEBUG("not enough stolen space for PCTX, disabling\n");
+		return PTR_ERR(pctx);
+	}
+
+	GEM_BUG_ON(range_overflows_t(u64,
+				     i915->dsm.start,
+				     pctx->stolen->start,
+				     U32_MAX));
+	pctx_paddr = i915->dsm.start + pctx->stolen->start;
+	intel_uncore_write(uncore, VLV_PCBR, pctx_paddr);
+
+out:
+	rc6->pctx = pctx;
+	return 0;
+}
+
+static void chv_rc6_enable(struct intel_rc6 *rc6)
+{
+	struct intel_uncore *uncore = rc6_to_uncore(rc6);
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
+
+	/* 2a: Program RC6 thresholds.*/
+	set(uncore, GEN6_RC6_WAKE_RATE_LIMIT, 40 << 16);
+	set(uncore, GEN6_RC_EVALUATION_INTERVAL, 125000); /* 12500 * 1280ns */
+	set(uncore, GEN6_RC_IDLE_HYSTERSIS, 25); /* 25 * 1280ns */
+
+	for_each_engine(engine, rc6_to_gt(rc6), id)
+		set(uncore, RING_MAX_IDLE(engine->mmio_base), 10);
+	set(uncore, GEN6_RC_SLEEP, 0);
+
+	/* TO threshold set to 500 us (0x186 * 1.28 us) */
+	set(uncore, GEN6_RC6_THRESHOLD, 0x186);
+
+	/* Allows RC6 residency counter to work */
+	set(uncore, VLV_COUNTER_CONTROL,
+	    _MASKED_BIT_ENABLE(VLV_COUNT_RANGE_HIGH |
+			       VLV_MEDIA_RC6_COUNT_EN |
+			       VLV_RENDER_RC6_COUNT_EN));
+
+	/* 3: Enable RC6 */
+	set(uncore, GEN6_RC_CONTROL, GEN7_RC_CTL_TO_MODE);
+}
+
+static void vlv_rc6_enable(struct intel_rc6 *rc6)
+{
+	struct intel_uncore *uncore = rc6_to_uncore(rc6);
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
+
+	set(uncore, GEN6_RC6_WAKE_RATE_LIMIT, 0x00280000);
+	set(uncore, GEN6_RC_EVALUATION_INTERVAL, 125000);
+	set(uncore, GEN6_RC_IDLE_HYSTERSIS, 25);
+
+	for_each_engine(engine, rc6_to_gt(rc6), id)
+		set(uncore, RING_MAX_IDLE(engine->mmio_base), 10);
+
+	set(uncore, GEN6_RC6_THRESHOLD, 0x557);
+
+	/* Allows RC6 residency counter to work */
+	set(uncore, VLV_COUNTER_CONTROL,
+	    _MASKED_BIT_ENABLE(VLV_COUNT_RANGE_HIGH |
+			       VLV_MEDIA_RC0_COUNT_EN |
+			       VLV_RENDER_RC0_COUNT_EN |
+			       VLV_MEDIA_RC6_COUNT_EN |
+			       VLV_RENDER_RC6_COUNT_EN));
+
+	set(uncore, GEN6_RC_CONTROL,
+	    GEN7_RC_CTL_TO_MODE | VLV_RC_CTL_CTX_RST_PARALLEL);
+}
+
+static bool bxt_check_bios_rc6_setup(struct intel_rc6 *rc6)
+{
+	struct intel_uncore *uncore = rc6_to_uncore(rc6);
+	struct drm_i915_private *i915 = rc6_to_i915(rc6);
+	u32 rc6_ctx_base, rc_ctl, rc_sw_target;
+	bool enable_rc6 = true;
+
+	rc_ctl = intel_uncore_read(uncore, GEN6_RC_CONTROL);
+	rc_sw_target = intel_uncore_read(uncore, GEN6_RC_STATE);
+	rc_sw_target &= RC_SW_TARGET_STATE_MASK;
+	rc_sw_target >>= RC_SW_TARGET_STATE_SHIFT;
+	DRM_DEBUG_DRIVER("BIOS enabled RC states: "
+			 "HW_CTRL %s HW_RC6 %s SW_TARGET_STATE %x\n",
+			 onoff(rc_ctl & GEN6_RC_CTL_HW_ENABLE),
+			 onoff(rc_ctl & GEN6_RC_CTL_RC6_ENABLE),
+			 rc_sw_target);
+
+	if (!(intel_uncore_read(uncore, RC6_LOCATION) & RC6_CTX_IN_DRAM)) {
+		DRM_DEBUG_DRIVER("RC6 Base location not set properly.\n");
+		enable_rc6 = false;
+	}
+
+	/*
+	 * The exact context size is not known for BXT, so assume a page size
+	 * for this check.
+	 */
+	rc6_ctx_base =
+		intel_uncore_read(uncore, RC6_CTX_BASE) & RC6_CTX_BASE_MASK;
+	if (!(rc6_ctx_base >= i915->dsm_reserved.start &&
+	      rc6_ctx_base + PAGE_SIZE < i915->dsm_reserved.end)) {
+		DRM_DEBUG_DRIVER("RC6 Base address not as expected.\n");
+		enable_rc6 = false;
+	}
+
+	if (!((intel_uncore_read(uncore, PWRCTX_MAXCNT_RCSUNIT) & IDLE_TIME_MASK) > 1 &&
+	      (intel_uncore_read(uncore, PWRCTX_MAXCNT_VCSUNIT0) & IDLE_TIME_MASK) > 1 &&
+	      (intel_uncore_read(uncore, PWRCTX_MAXCNT_BCSUNIT) & IDLE_TIME_MASK) > 1 &&
+	      (intel_uncore_read(uncore, PWRCTX_MAXCNT_VECSUNIT) & IDLE_TIME_MASK) > 1)) {
+		DRM_DEBUG_DRIVER("Engine Idle wait time not set properly.\n");
+		enable_rc6 = false;
+	}
+
+	if (!intel_uncore_read(uncore, GEN8_PUSHBUS_CONTROL) ||
+	    !intel_uncore_read(uncore, GEN8_PUSHBUS_ENABLE) ||
+	    !intel_uncore_read(uncore, GEN8_PUSHBUS_SHIFT)) {
+		DRM_DEBUG_DRIVER("Pushbus not setup properly.\n");
+		enable_rc6 = false;
+	}
+
+	if (!intel_uncore_read(uncore, GEN6_GFXPAUSE)) {
+		DRM_DEBUG_DRIVER("GFX pause not setup properly.\n");
+		enable_rc6 = false;
+	}
+
+	if (!intel_uncore_read(uncore, GEN8_MISC_CTRL0)) {
+		DRM_DEBUG_DRIVER("GPM control not setup properly.\n");
+		enable_rc6 = false;
+	}
+
+	return enable_rc6;
+}
+
+static bool rc6_supported(struct intel_rc6 *rc6)
+{
+	struct drm_i915_private *i915 = rc6_to_i915(rc6);
+
+	if (!HAS_RC6(i915))
+		return false;
+
+	if (intel_vgpu_active(i915))
+		return false;
+
+	if (is_mock_gt(rc6_to_gt(rc6)))
+		return false;
+
+	if (IS_GEN9_LP(i915) && !bxt_check_bios_rc6_setup(rc6)) {
+		dev_notice(i915->drm.dev,
+			   "RC6 and powersaving disabled by BIOS\n");
+		return false;
+	}
+
+	return true;
+}
+
+static void rpm_get(struct intel_rc6 *rc6)
+{
+	GEM_BUG_ON(rc6->wakeref);
+	pm_runtime_get_sync(&rc6_to_i915(rc6)->drm.pdev->dev);
+	rc6->wakeref = true;
+}
+
+static void rpm_put(struct intel_rc6 *rc6)
+{
+	GEM_BUG_ON(!rc6->wakeref);
+	pm_runtime_put(&rc6_to_i915(rc6)->drm.pdev->dev);
+	rc6->wakeref = false;
+}
+
+static bool intel_rc6_ctx_corrupted(struct intel_rc6 *rc6)
+{
+	return !intel_uncore_read(rc6_to_uncore(rc6), GEN8_RC6_CTX_INFO);
+}
+
+static void intel_rc6_ctx_wa_init(struct intel_rc6 *rc6)
+{
+	struct drm_i915_private *i915 = rc6_to_i915(rc6);
+
+	if (!NEEDS_RC6_CTX_CORRUPTION_WA(i915))
+		return;
+
+	if (intel_rc6_ctx_corrupted(rc6)) {
+		DRM_INFO("RC6 context corrupted, disabling runtime power management\n");
+		rc6->ctx_corrupted = true;
+	}
+}
+
+/**
+ * intel_rc6_ctx_wa_resume - system resume sequence for the RC6 CTX WA
+ * @rc6: rc6 state
+ *
+ * Perform any steps needed to re-init the RC6 CTX WA after system resume.
+ */
+void intel_rc6_ctx_wa_resume(struct intel_rc6 *rc6)
+{
+	if (rc6->ctx_corrupted && !intel_rc6_ctx_corrupted(rc6)) {
+		DRM_INFO("RC6 context restored, re-enabling runtime power management\n");
+		rc6->ctx_corrupted = false;
+	}
+}
+
+/**
+ * intel_rc6_ctx_wa_check - check for a new RC6 CTX corruption
+ * @rc6: rc6 state
+ *
+ * Check if an RC6 CTX corruption has happened since the last check and if so
+ * disable RC6 and runtime power management.
+*/
+void intel_rc6_ctx_wa_check(struct intel_rc6 *rc6)
+{
+	struct drm_i915_private *i915 = rc6_to_i915(rc6);
+
+	if (!NEEDS_RC6_CTX_CORRUPTION_WA(i915))
+		return;
+
+	if (rc6->ctx_corrupted)
+		return;
+
+	if (!intel_rc6_ctx_corrupted(rc6))
+		return;
+
+	DRM_NOTE("RC6 context corruption, disabling runtime power management\n");
+
+	intel_rc6_disable(rc6);
+	rc6->ctx_corrupted = true;
+
+	return;
+}
+
+static void __intel_rc6_disable(struct intel_rc6 *rc6)
+{
+	struct drm_i915_private *i915 = rc6_to_i915(rc6);
+	struct intel_uncore *uncore = rc6_to_uncore(rc6);
+
+	intel_uncore_forcewake_get(uncore, FORCEWAKE_ALL);
+	if (INTEL_GEN(i915) >= 9)
+		set(uncore, GEN9_PG_ENABLE, 0);
+	set(uncore, GEN6_RC_CONTROL, 0);
+	set(uncore, GEN6_RC_STATE, 0);
+	intel_uncore_forcewake_put(uncore, FORCEWAKE_ALL);
+}
+
+void intel_rc6_init(struct intel_rc6 *rc6)
+{
+	struct drm_i915_private *i915 = rc6_to_i915(rc6);
+	int err;
+
+	/* Disable runtime-pm until we can save the GPU state with rc6 pctx */
+	rpm_get(rc6);
+
+	if (!rc6_supported(rc6))
+		return;
+
+	intel_rc6_ctx_wa_init(rc6);
+
+	if (IS_CHERRYVIEW(i915))
+		err = chv_rc6_init(rc6);
+	else if (IS_VALLEYVIEW(i915))
+		err = vlv_rc6_init(rc6);
+	else
+		err = 0;
+
+	/* Sanitize rc6, ensure it is disabled before we are ready. */
+	__intel_rc6_disable(rc6);
+
+	rc6->supported = err == 0;
+}
+
+void intel_rc6_sanitize(struct intel_rc6 *rc6)
+{
+	if (rc6->enabled) { /* unbalanced suspend/resume */
+		rpm_get(rc6);
+		rc6->enabled = false;
+	}
+
+	if (rc6->supported)
+		__intel_rc6_disable(rc6);
+}
+
+void intel_rc6_enable(struct intel_rc6 *rc6)
+{
+	struct drm_i915_private *i915 = rc6_to_i915(rc6);
+	struct intel_uncore *uncore = rc6_to_uncore(rc6);
+
+	if (!rc6->supported)
+		return;
+
+	GEM_BUG_ON(rc6->enabled);
+
+	if (rc6->ctx_corrupted)
+		return;
+
+	intel_uncore_forcewake_get(uncore, FORCEWAKE_ALL);
+
+	if (IS_CHERRYVIEW(i915))
+		chv_rc6_enable(rc6);
+	else if (IS_VALLEYVIEW(i915))
+		vlv_rc6_enable(rc6);
+	else if (INTEL_GEN(i915) >= 11)
+		gen11_rc6_enable(rc6);
+	else if (INTEL_GEN(i915) >= 9)
+		gen9_rc6_enable(rc6);
+	else if (IS_BROADWELL(i915))
+		gen8_rc6_enable(rc6);
+	else if (INTEL_GEN(i915) >= 6)
+		gen6_rc6_enable(rc6);
+
+	intel_uncore_forcewake_put(uncore, FORCEWAKE_ALL);
+
+	/* rc6 is ready, runtime-pm is go! */
+	rpm_put(rc6);
+	rc6->enabled = true;
+}
+
+void intel_rc6_disable(struct intel_rc6 *rc6)
+{
+	if (!rc6->enabled)
+		return;
+
+	rpm_get(rc6);
+	rc6->enabled = false;
+
+	__intel_rc6_disable(rc6);
+}
+
+void intel_rc6_fini(struct intel_rc6 *rc6)
+{
+	struct drm_i915_gem_object *pctx;
+
+	intel_rc6_disable(rc6);
+
+	pctx = fetch_and_zero(&rc6->pctx);
+	if (pctx)
+		i915_gem_object_put(pctx);
+
+	if (rc6->wakeref)
+		rpm_put(rc6);
+}
+
+static u64 vlv_residency_raw(struct intel_uncore *uncore, const i915_reg_t reg)
+{
+	u32 lower, upper, tmp;
+	int loop = 2;
+
+	/*
+	 * The register accessed do not need forcewake. We borrow
+	 * uncore lock to prevent concurrent access to range reg.
+	 */
+	lockdep_assert_held(&uncore->lock);
+
+	/*
+	 * vlv and chv residency counters are 40 bits in width.
+	 * With a control bit, we can choose between upper or lower
+	 * 32bit window into this counter.
+	 *
+	 * Although we always use the counter in high-range mode elsewhere,
+	 * userspace may attempt to read the value before rc6 is initialised,
+	 * before we have set the default VLV_COUNTER_CONTROL value. So always
+	 * set the high bit to be safe.
+	 */
+	set(uncore, VLV_COUNTER_CONTROL,
+	    _MASKED_BIT_ENABLE(VLV_COUNT_RANGE_HIGH));
+	upper = intel_uncore_read_fw(uncore, reg);
+	do {
+		tmp = upper;
+
+		set(uncore, VLV_COUNTER_CONTROL,
+		    _MASKED_BIT_DISABLE(VLV_COUNT_RANGE_HIGH));
+		lower = intel_uncore_read_fw(uncore, reg);
+
+		set(uncore, VLV_COUNTER_CONTROL,
+		    _MASKED_BIT_ENABLE(VLV_COUNT_RANGE_HIGH));
+		upper = intel_uncore_read_fw(uncore, reg);
+	} while (upper != tmp && --loop);
+
+	/*
+	 * Everywhere else we always use VLV_COUNTER_CONTROL with the
+	 * VLV_COUNT_RANGE_HIGH bit set - so it is safe to leave it set
+	 * now.
+	 */
+
+	return lower | (u64)upper << 8;
+}
+
+u64 intel_rc6_residency_ns(struct intel_rc6 *rc6, const i915_reg_t reg)
+{
+	struct drm_i915_private *i915 = rc6_to_i915(rc6);
+	struct intel_uncore *uncore = rc6_to_uncore(rc6);
+	u64 time_hw, prev_hw, overflow_hw;
+	unsigned int fw_domains;
+	unsigned long flags;
+	unsigned int i;
+	u32 mul, div;
+
+	if (!rc6->supported)
+		return 0;
+
+	/*
+	 * Store previous hw counter values for counter wrap-around handling.
+	 *
+	 * There are only four interesting registers and they live next to each
+	 * other so we can use the relative address, compared to the smallest
+	 * one as the index into driver storage.
+	 */
+	i = (i915_mmio_reg_offset(reg) -
+	     i915_mmio_reg_offset(GEN6_GT_GFX_RC6_LOCKED)) / sizeof(u32);
+	if (WARN_ON_ONCE(i >= ARRAY_SIZE(rc6->cur_residency)))
+		return 0;
+
+	fw_domains = intel_uncore_forcewake_for_reg(uncore, reg, FW_REG_READ);
+
+	spin_lock_irqsave(&uncore->lock, flags);
+	intel_uncore_forcewake_get__locked(uncore, fw_domains);
+
+	/* On VLV and CHV, residency time is in CZ units rather than 1.28us */
+	if (IS_VALLEYVIEW(i915) || IS_CHERRYVIEW(i915)) {
+		mul = 1000000;
+		div = i915->czclk_freq;
+		overflow_hw = BIT_ULL(40);
+		time_hw = vlv_residency_raw(uncore, reg);
+	} else {
+		/* 833.33ns units on Gen9LP, 1.28us elsewhere. */
+		if (IS_GEN9_LP(i915)) {
+			mul = 10000;
+			div = 12;
+		} else {
+			mul = 1280;
+			div = 1;
+		}
+
+		overflow_hw = BIT_ULL(32);
+		time_hw = intel_uncore_read_fw(uncore, reg);
+	}
+
+	/*
+	 * Counter wrap handling.
+	 *
+	 * But relying on a sufficient frequency of queries otherwise counters
+	 * can still wrap.
+	 */
+	prev_hw = rc6->prev_hw_residency[i];
+	rc6->prev_hw_residency[i] = time_hw;
+
+	/* RC6 delta from last sample. */
+	if (time_hw >= prev_hw)
+		time_hw -= prev_hw;
+	else
+		time_hw += overflow_hw - prev_hw;
+
+	/* Add delta to RC6 extended raw driver copy. */
+	time_hw += rc6->cur_residency[i];
+	rc6->cur_residency[i] = time_hw;
+
+	intel_uncore_forcewake_put__locked(uncore, fw_domains);
+	spin_unlock_irqrestore(&uncore->lock, flags);
+
+	return mul_u64_u32_div(time_hw, mul, div);
+}
+
+u64 intel_rc6_residency_us(struct intel_rc6 *rc6, i915_reg_t reg)
+{
+	return DIV_ROUND_UP_ULL(intel_rc6_residency_ns(rc6, reg), 1000);
+}
diff --git a/drivers/gpu/drm/i915/gt/intel_rc6.h b/drivers/gpu/drm/i915/gt/intel_rc6.h
new file mode 100644
index 000000000000..1370f6834a4c
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/intel_rc6.h
@@ -0,0 +1,28 @@
+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2019 Intel Corporation
+ */
+
+#ifndef INTEL_RC6_H
+#define INTEL_RC6_H
+
+#include "i915_reg.h"
+
+struct intel_engine_cs;
+struct intel_rc6;
+
+void intel_rc6_init(struct intel_rc6 *rc6);
+void intel_rc6_fini(struct intel_rc6 *rc6);
+
+void intel_rc6_sanitize(struct intel_rc6 *rc6);
+void intel_rc6_enable(struct intel_rc6 *rc6);
+void intel_rc6_disable(struct intel_rc6 *rc6);
+
+u64 intel_rc6_residency_ns(struct intel_rc6 *rc6, i915_reg_t reg);
+u64 intel_rc6_residency_us(struct intel_rc6 *rc6, i915_reg_t reg);
+
+void intel_rc6_ctx_wa_check(struct intel_rc6 *rc6);
+void intel_rc6_ctx_wa_resume(struct intel_rc6 *rc6);
+
+#endif /* INTEL_RC6_H */
diff --git a/drivers/gpu/drm/i915/gt/intel_rc6_types.h b/drivers/gpu/drm/i915/gt/intel_rc6_types.h
new file mode 100644
index 000000000000..89ad5697a8d4
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/intel_rc6_types.h
@@ -0,0 +1,29 @@
+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2019 Intel Corporation
+ */
+
+#ifndef INTEL_RC6_TYPES_H
+#define INTEL_RC6_TYPES_H
+
+#include <linux/spinlock.h>
+#include <linux/types.h>
+
+#include "intel_engine_types.h"
+
+struct drm_i915_gem_object;
+
+struct intel_rc6 {
+	u64 prev_hw_residency[4];
+	u64 cur_residency[4];
+
+	struct drm_i915_gem_object *pctx;
+
+	bool supported : 1;
+	bool enabled : 1;
+	bool wakeref : 1;
+	bool ctx_corrupted : 1;
+};
+
+#endif /* INTEL_RC6_TYPES_H */
diff --git a/drivers/gpu/drm/i915/gt/intel_renderstate.c b/drivers/gpu/drm/i915/gt/intel_renderstate.c
index 6d05f9c64178..c4edc35e7d89 100644
--- a/drivers/gpu/drm/i915/gt/intel_renderstate.c
+++ b/drivers/gpu/drm/i915/gt/intel_renderstate.c
@@ -27,6 +27,7 @@
 
 #include "i915_drv.h"
 #include "intel_renderstate.h"
+#include "intel_ring.h"
 
 struct intel_renderstate {
 	const struct intel_renderstate_rodata *rodata;
diff --git a/drivers/gpu/drm/i915/gt/intel_reset.c b/drivers/gpu/drm/i915/gt/intel_reset.c
index b9d84d52e986..f03e000051c1 100644
--- a/drivers/gpu/drm/i915/gt/intel_reset.c
+++ b/drivers/gpu/drm/i915/gt/intel_reset.c
@@ -42,11 +42,10 @@ static void engine_skip_context(struct i915_request *rq)
 	struct intel_engine_cs *engine = rq->engine;
 	struct i915_gem_context *hung_ctx = rq->gem_context;
 
-	lockdep_assert_held(&engine->active.lock);
-
 	if (!i915_request_is_active(rq))
 		return;
 
+	lockdep_assert_held(&engine->active.lock);
 	list_for_each_entry_continue(rq, &engine->active.requests, sched.link)
 		if (rq->gem_context == hung_ctx)
 			i915_request_skip(rq, -EIO);
@@ -123,7 +122,6 @@ void __i915_request_reset(struct i915_request *rq, bool guilty)
 		  rq->fence.seqno,
 		  yesno(guilty));
 
-	lockdep_assert_held(&rq->engine->active.lock);
 	GEM_BUG_ON(i915_request_completed(rq));
 
 	if (guilty) {
@@ -284,14 +282,14 @@ static int gen6_reset_engines(struct intel_gt *gt,
 			      intel_engine_mask_t engine_mask,
 			      unsigned int retry)
 {
-	struct intel_engine_cs *engine;
-	const u32 hw_engine_mask[] = {
+	static const u32 hw_engine_mask[] = {
 		[RCS0]  = GEN6_GRDOM_RENDER,
 		[BCS0]  = GEN6_GRDOM_BLT,
 		[VCS0]  = GEN6_GRDOM_MEDIA,
 		[VCS1]  = GEN8_GRDOM_MEDIA2,
 		[VECS0] = GEN6_GRDOM_VECS,
 	};
+	struct intel_engine_cs *engine;
 	u32 hw_mask;
 
 	if (engine_mask == ALL_ENGINES) {
@@ -300,7 +298,7 @@ static int gen6_reset_engines(struct intel_gt *gt,
 		intel_engine_mask_t tmp;
 
 		hw_mask = 0;
-		for_each_engine_masked(engine, gt->i915, engine_mask, tmp) {
+		for_each_engine_masked(engine, gt, engine_mask, tmp) {
 			GEM_BUG_ON(engine->id >= ARRAY_SIZE(hw_engine_mask));
 			hw_mask |= hw_engine_mask[engine->id];
 		}
@@ -309,7 +307,7 @@ static int gen6_reset_engines(struct intel_gt *gt,
 	return gen6_hw_domain_reset(gt, hw_mask);
 }
 
-static u32 gen11_lock_sfc(struct intel_engine_cs *engine)
+static int gen11_lock_sfc(struct intel_engine_cs *engine, u32 *hw_mask)
 {
 	struct intel_uncore *uncore = engine->uncore;
 	u8 vdbox_sfc_access = RUNTIME_INFO(engine->i915)->vdbox_sfc_access;
@@ -318,6 +316,7 @@ static u32 gen11_lock_sfc(struct intel_engine_cs *engine)
 	i915_reg_t sfc_usage;
 	u32 sfc_usage_bit;
 	u32 sfc_reset_bit;
+	int ret;
 
 	switch (engine->class) {
 	case VIDEO_DECODE_CLASS:
@@ -352,27 +351,33 @@ static u32 gen11_lock_sfc(struct intel_engine_cs *engine)
 	}
 
 	/*
-	 * Tell the engine that a software reset is going to happen. The engine
-	 * will then try to force lock the SFC (if currently locked, it will
-	 * remain so until we tell the engine it is safe to unlock; if currently
-	 * unlocked, it will ignore this and all new lock requests). If SFC
-	 * ends up being locked to the engine we want to reset, we have to reset
-	 * it as well (we will unlock it once the reset sequence is completed).
+	 * If the engine is using a SFC, tell the engine that a software reset
+	 * is going to happen. The engine will then try to force lock the SFC.
+	 * If SFC ends up being locked to the engine we want to reset, we have
+	 * to reset it as well (we will unlock it once the reset sequence is
+	 * completed).
 	 */
+	if (!(intel_uncore_read_fw(uncore, sfc_usage) & sfc_usage_bit))
+		return 0;
+
 	rmw_set_fw(uncore, sfc_forced_lock, sfc_forced_lock_bit);
 
-	if (__intel_wait_for_register_fw(uncore,
-					 sfc_forced_lock_ack,
-					 sfc_forced_lock_ack_bit,
-					 sfc_forced_lock_ack_bit,
-					 1000, 0, NULL)) {
-		DRM_DEBUG_DRIVER("Wait for SFC forced lock ack failed\n");
+	ret = __intel_wait_for_register_fw(uncore,
+					   sfc_forced_lock_ack,
+					   sfc_forced_lock_ack_bit,
+					   sfc_forced_lock_ack_bit,
+					   1000, 0, NULL);
+
+	/* Was the SFC released while we were trying to lock it? */
+	if (!(intel_uncore_read_fw(uncore, sfc_usage) & sfc_usage_bit))
 		return 0;
-	}
 
-	if (intel_uncore_read_fw(uncore, sfc_usage) & sfc_usage_bit)
-		return sfc_reset_bit;
+	if (ret) {
+		DRM_DEBUG_DRIVER("Wait for SFC forced lock ack failed\n");
+		return ret;
+	}
 
+	*hw_mask |= sfc_reset_bit;
 	return 0;
 }
 
@@ -408,7 +413,7 @@ static int gen11_reset_engines(struct intel_gt *gt,
 			       intel_engine_mask_t engine_mask,
 			       unsigned int retry)
 {
-	const u32 hw_engine_mask[] = {
+	static const u32 hw_engine_mask[] = {
 		[RCS0]  = GEN11_GRDOM_RENDER,
 		[BCS0]  = GEN11_GRDOM_BLT,
 		[VCS0]  = GEN11_GRDOM_MEDIA,
@@ -427,17 +432,26 @@ static int gen11_reset_engines(struct intel_gt *gt,
 		hw_mask = GEN11_GRDOM_FULL;
 	} else {
 		hw_mask = 0;
-		for_each_engine_masked(engine, gt->i915, engine_mask, tmp) {
+		for_each_engine_masked(engine, gt, engine_mask, tmp) {
 			GEM_BUG_ON(engine->id >= ARRAY_SIZE(hw_engine_mask));
 			hw_mask |= hw_engine_mask[engine->id];
-			hw_mask |= gen11_lock_sfc(engine);
+			ret = gen11_lock_sfc(engine, &hw_mask);
+			if (ret)
+				goto sfc_unlock;
 		}
 	}
 
 	ret = gen6_hw_domain_reset(gt, hw_mask);
 
+sfc_unlock:
+	/*
+	 * We unlock the SFC based on the lock status and not the result of
+	 * gen11_lock_sfc to make sure that we clean properly if something
+	 * wrong happened during the lock (e.g. lock acquired after timeout
+	 * expiration).
+	 */
 	if (engine_mask != ALL_ENGINES)
-		for_each_engine_masked(engine, gt->i915, engine_mask, tmp)
+		for_each_engine_masked(engine, gt, engine_mask, tmp)
 			gen11_unlock_sfc(engine);
 
 	return ret;
@@ -496,7 +510,7 @@ static int gen8_reset_engines(struct intel_gt *gt,
 	intel_engine_mask_t tmp;
 	int ret;
 
-	for_each_engine_masked(engine, gt->i915, engine_mask, tmp) {
+	for_each_engine_masked(engine, gt, engine_mask, tmp) {
 		ret = gen8_engine_reset_prepare(engine);
 		if (ret && !reset_non_ready)
 			goto skip_reset;
@@ -522,19 +536,30 @@ static int gen8_reset_engines(struct intel_gt *gt,
 		ret = gen6_reset_engines(gt, engine_mask, retry);
 
 skip_reset:
-	for_each_engine_masked(engine, gt->i915, engine_mask, tmp)
+	for_each_engine_masked(engine, gt, engine_mask, tmp)
 		gen8_engine_reset_cancel(engine);
 
 	return ret;
 }
 
+static int mock_reset(struct intel_gt *gt,
+		      intel_engine_mask_t mask,
+		      unsigned int retry)
+{
+	return 0;
+}
+
 typedef int (*reset_func)(struct intel_gt *,
 			  intel_engine_mask_t engine_mask,
 			  unsigned int retry);
 
-static reset_func intel_get_gpu_reset(struct drm_i915_private *i915)
+static reset_func intel_get_gpu_reset(const struct intel_gt *gt)
 {
-	if (INTEL_GEN(i915) >= 8)
+	struct drm_i915_private *i915 = gt->i915;
+
+	if (is_mock_gt(gt))
+		return mock_reset;
+	else if (INTEL_GEN(i915) >= 8)
 		return gen8_reset_engines;
 	else if (INTEL_GEN(i915) >= 6)
 		return gen6_reset_engines;
@@ -557,7 +582,7 @@ int __intel_gt_reset(struct intel_gt *gt, intel_engine_mask_t engine_mask)
 	int ret = -ETIMEDOUT;
 	int retry;
 
-	reset = intel_get_gpu_reset(gt->i915);
+	reset = intel_get_gpu_reset(gt);
 	if (!reset)
 		return -ENODEV;
 
@@ -577,17 +602,20 @@ int __intel_gt_reset(struct intel_gt *gt, intel_engine_mask_t engine_mask)
 	return ret;
 }
 
-bool intel_has_gpu_reset(struct drm_i915_private *i915)
+bool intel_has_gpu_reset(const struct intel_gt *gt)
 {
 	if (!i915_modparams.reset)
 		return NULL;
 
-	return intel_get_gpu_reset(i915);
+	return intel_get_gpu_reset(gt);
 }
 
-bool intel_has_reset_engine(struct drm_i915_private *i915)
+bool intel_has_reset_engine(const struct intel_gt *gt)
 {
-	return INTEL_INFO(i915)->has_reset_engine && i915_modparams.reset >= 2;
+	if (i915_modparams.reset < 2)
+		return false;
+
+	return INTEL_INFO(gt->i915)->has_reset_engine;
 }
 
 int intel_reset_guc(struct intel_gt *gt)
@@ -654,7 +682,7 @@ static intel_engine_mask_t reset_prepare(struct intel_gt *gt)
 	intel_engine_mask_t awake = 0;
 	enum intel_engine_id id;
 
-	for_each_engine(engine, gt->i915, id) {
+	for_each_engine(engine, gt, id) {
 		if (intel_engine_pm_get_if_awake(engine))
 			awake |= engine->mask;
 		reset_prepare_engine(engine);
@@ -684,10 +712,10 @@ static int gt_reset(struct intel_gt *gt, intel_engine_mask_t stalled_mask)
 	if (err)
 		return err;
 
-	for_each_engine(engine, gt->i915, id)
+	for_each_engine(engine, gt, id)
 		__intel_engine_reset(engine, stalled_mask & engine->mask);
 
-	i915_gem_restore_fences(gt->i915);
+	i915_gem_restore_fences(gt->ggtt);
 
 	return err;
 }
@@ -697,7 +725,7 @@ static void reset_finish_engine(struct intel_engine_cs *engine)
 	engine->reset.finish(engine);
 	intel_uncore_forcewake_put(engine->uncore, FORCEWAKE_ALL);
 
-	intel_engine_signal_breadcrumbs(engine);
+	intel_engine_breadcrumbs_irq(engine);
 }
 
 static void reset_finish(struct intel_gt *gt, intel_engine_mask_t awake)
@@ -705,7 +733,7 @@ static void reset_finish(struct intel_gt *gt, intel_engine_mask_t awake)
 	struct intel_engine_cs *engine;
 	enum intel_engine_id id;
 
-	for_each_engine(engine, gt->i915, id) {
+	for_each_engine(engine, gt, id) {
 		reset_finish_engine(engine);
 		if (awake & engine->mask)
 			intel_engine_pm_put(engine);
@@ -741,7 +769,7 @@ static void __intel_gt_set_wedged(struct intel_gt *gt)
 	if (GEM_SHOW_DEBUG() && !intel_engines_are_idle(gt)) {
 		struct drm_printer p = drm_debug_printer(__func__);
 
-		for_each_engine(engine, gt->i915, id)
+		for_each_engine(engine, gt, id)
 			intel_engine_dump(engine, &p, "%s\n", engine->name);
 	}
 
@@ -758,7 +786,7 @@ static void __intel_gt_set_wedged(struct intel_gt *gt)
 	if (!INTEL_INFO(gt->i915)->gpu_reset_clobbers_display)
 		__intel_gt_reset(gt, ALL_ENGINES);
 
-	for_each_engine(engine, gt->i915, id)
+	for_each_engine(engine, gt, id)
 		engine->submit_request = nop_submit_request;
 
 	/*
@@ -770,7 +798,7 @@ static void __intel_gt_set_wedged(struct intel_gt *gt)
 	set_bit(I915_WEDGED, &gt->reset.flags);
 
 	/* Mark all executing requests as skipped */
-	for_each_engine(engine, gt->i915, id)
+	for_each_engine(engine, gt, id)
 		engine->cancel_requests(engine);
 
 	reset_finish(gt, awake);
@@ -783,7 +811,7 @@ void intel_gt_set_wedged(struct intel_gt *gt)
 	intel_wakeref_t wakeref;
 
 	mutex_lock(&gt->reset.mutex);
-	with_intel_runtime_pm(&gt->i915->runtime_pm, wakeref)
+	with_intel_runtime_pm(gt->uncore->rpm, wakeref)
 		__intel_gt_set_wedged(gt);
 	mutex_unlock(&gt->reset.mutex);
 }
@@ -793,11 +821,13 @@ static bool __intel_gt_unset_wedged(struct intel_gt *gt)
 	struct intel_gt_timelines *timelines = &gt->timelines;
 	struct intel_timeline *tl;
 	unsigned long flags;
+	bool ok;
 
 	if (!test_bit(I915_WEDGED, &gt->reset.flags))
 		return true;
 
-	if (!gt->scratch) /* Never full initialised, recovery impossible */
+	/* Never fully initialised, recovery impossible */
+	if (test_bit(I915_WEDGED_ON_INIT, &gt->reset.flags))
 		return false;
 
 	GEM_TRACE("start\n");
@@ -814,10 +844,10 @@ static bool __intel_gt_unset_wedged(struct intel_gt *gt)
 	 */
 	spin_lock_irqsave(&timelines->lock, flags);
 	list_for_each_entry(tl, &timelines->active_list, link) {
-		struct i915_request *rq;
+		struct dma_fence *fence;
 
-		rq = i915_active_request_get_unlocked(&tl->last_request);
-		if (!rq)
+		fence = i915_active_fence_get(&tl->last_request);
+		if (!fence)
 			continue;
 
 		spin_unlock_irqrestore(&timelines->lock, flags);
@@ -829,8 +859,8 @@ static bool __intel_gt_unset_wedged(struct intel_gt *gt)
 		 * (I915_FENCE_TIMEOUT) so this wait should not be unbounded
 		 * in the worst case.
 		 */
-		dma_fence_default_wait(&rq->fence, false, MAX_SCHEDULE_TIMEOUT);
-		i915_request_put(rq);
+		dma_fence_default_wait(fence, false, MAX_SCHEDULE_TIMEOUT);
+		dma_fence_put(fence);
 
 		/* Restart iteration after droping lock */
 		spin_lock_irqsave(&timelines->lock, flags);
@@ -838,7 +868,18 @@ static bool __intel_gt_unset_wedged(struct intel_gt *gt)
 	}
 	spin_unlock_irqrestore(&timelines->lock, flags);
 
-	intel_gt_sanitize(gt, false);
+	/* We must reset pending GPU events before restoring our submission */
+	ok = !HAS_EXECLISTS(gt->i915); /* XXX better agnosticism desired */
+	if (!INTEL_INFO(gt->i915)->gpu_reset_clobbers_display)
+		ok = __intel_gt_reset(gt, ALL_ENGINES) == 0;
+	if (!ok) {
+		/*
+		 * Warn CI about the unrecoverable wedged condition.
+		 * Time for a reboot.
+		 */
+		add_taint_for_CI(TAINT_WARN);
+		return false;
+	}
 
 	/*
 	 * Undo nop_submit_request. We prevent all new i915 requests from
@@ -893,7 +934,7 @@ static int resume(struct intel_gt *gt)
 	enum intel_engine_id id;
 	int ret;
 
-	for_each_engine(engine, gt->i915, id) {
+	for_each_engine(engine, gt, id) {
 		ret = engine->resume(engine);
 		if (ret)
 			return ret;
@@ -943,7 +984,7 @@ void intel_gt_reset(struct intel_gt *gt,
 
 	awake = reset_prepare(gt);
 
-	if (!intel_has_gpu_reset(gt->i915)) {
+	if (!intel_has_gpu_reset(gt)) {
 		if (i915_modparams.reset)
 			dev_err(gt->i915->drm.dev, "GPU reset not supported\n");
 		else
@@ -972,7 +1013,7 @@ void intel_gt_reset(struct intel_gt *gt,
 	 * was running at the time of the reset (i.e. we weren't VT
 	 * switched away).
 	 */
-	ret = i915_gem_init_hw(gt->i915);
+	ret = intel_gt_init_hw(gt);
 	if (ret) {
 		DRM_ERROR("Failed to initialise HW following reset (%d)\n",
 			  ret);
@@ -983,8 +1024,6 @@ void intel_gt_reset(struct intel_gt *gt,
 	if (ret)
 		goto taint;
 
-	intel_gt_queue_hangcheck(gt);
-
 finish:
 	reset_finish(gt, awake);
 unlock:
@@ -1151,7 +1190,7 @@ void intel_gt_handle_error(struct intel_gt *gt,
 	 * isn't the case at least when we get here by doing a
 	 * simulated reset via debugfs, so get an RPM reference.
 	 */
-	wakeref = intel_runtime_pm_get(&gt->i915->runtime_pm);
+	wakeref = intel_runtime_pm_get(gt->uncore->rpm);
 
 	engine_mask &= INTEL_INFO(gt->i915)->engine_mask;
 
@@ -1164,8 +1203,8 @@ void intel_gt_handle_error(struct intel_gt *gt,
 	 * Try engine reset when available. We fall back to full reset if
 	 * single reset fails.
 	 */
-	if (intel_has_reset_engine(gt->i915) && !intel_gt_is_wedged(gt)) {
-		for_each_engine_masked(engine, gt->i915, engine_mask, tmp) {
+	if (intel_has_reset_engine(gt) && !intel_gt_is_wedged(gt)) {
+		for_each_engine_masked(engine, gt, engine_mask, tmp) {
 			BUILD_BUG_ON(I915_RESET_MODESET >= I915_RESET_ENGINE);
 			if (test_and_set_bit(I915_RESET_ENGINE + engine->id,
 					     &gt->reset.flags))
@@ -1193,7 +1232,7 @@ void intel_gt_handle_error(struct intel_gt *gt,
 	synchronize_rcu_expedited();
 
 	/* Prevent any other reset-engine attempt. */
-	for_each_engine(engine, gt->i915, tmp) {
+	for_each_engine(engine, gt, tmp) {
 		while (test_and_set_bit(I915_RESET_ENGINE + engine->id,
 					&gt->reset.flags))
 			wait_on_bit(&gt->reset.flags,
@@ -1203,7 +1242,7 @@ void intel_gt_handle_error(struct intel_gt *gt,
 
 	intel_gt_reset_global(gt, engine_mask, msg);
 
-	for_each_engine(engine, gt->i915, tmp)
+	for_each_engine(engine, gt, tmp)
 		clear_bit_unlock(I915_RESET_ENGINE + engine->id,
 				 &gt->reset.flags);
 	clear_bit_unlock(I915_RESET_BACKOFF, &gt->reset.flags);
@@ -1211,13 +1250,11 @@ void intel_gt_handle_error(struct intel_gt *gt,
 	wake_up_all(&gt->reset.queue);
 
 out:
-	intel_runtime_pm_put(&gt->i915->runtime_pm, wakeref);
+	intel_runtime_pm_put(gt->uncore->rpm, wakeref);
 }
 
-int intel_gt_reset_trylock(struct intel_gt *gt)
+int intel_gt_reset_trylock(struct intel_gt *gt, int *srcu)
 {
-	int srcu;
-
 	might_lock(&gt->reset.backoff_srcu);
 	might_sleep();
 
@@ -1232,10 +1269,10 @@ int intel_gt_reset_trylock(struct intel_gt *gt)
 
 		rcu_read_lock();
 	}
-	srcu = srcu_read_lock(&gt->reset.backoff_srcu);
+	*srcu = srcu_read_lock(&gt->reset.backoff_srcu);
 	rcu_read_unlock();
 
-	return srcu;
+	return 0;
 }
 
 void intel_gt_reset_unlock(struct intel_gt *gt, int tag)
@@ -1255,10 +1292,6 @@ int intel_gt_terminally_wedged(struct intel_gt *gt)
 	if (!test_bit(I915_RESET_BACKOFF, &gt->reset.flags))
 		return -EIO;
 
-	/* XXX intel_reset_finish() still takes struct_mutex!!! */
-	if (mutex_is_locked(&gt->i915->drm.struct_mutex))
-		return -EAGAIN;
-
 	if (wait_event_interruptible(gt->reset.queue,
 				     !test_bit(I915_RESET_BACKOFF,
 					       &gt->reset.flags)))
@@ -1267,6 +1300,14 @@ int intel_gt_terminally_wedged(struct intel_gt *gt)
 	return intel_gt_is_wedged(gt) ? -EIO : 0;
 }
 
+void intel_gt_set_wedged_on_init(struct intel_gt *gt)
+{
+	BUILD_BUG_ON(I915_RESET_ENGINE + I915_NUM_ENGINES >
+		     I915_WEDGED_ON_INIT);
+	intel_gt_set_wedged(gt);
+	set_bit(I915_WEDGED_ON_INIT, &gt->reset.flags);
+}
+
 void intel_gt_init_reset(struct intel_gt *gt)
 {
 	init_waitqueue_head(&gt->reset.queue);
@@ -1310,4 +1351,5 @@ void __intel_fini_wedge(struct intel_wedge_me *w)
 
 #if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
 #include "selftest_reset.c"
+#include "selftest_hangcheck.c"
 #endif
diff --git a/drivers/gpu/drm/i915/gt/intel_reset.h b/drivers/gpu/drm/i915/gt/intel_reset.h
index 37a987b17108..8e8d5f761166 100644
--- a/drivers/gpu/drm/i915/gt/intel_reset.h
+++ b/drivers/gpu/drm/i915/gt/intel_reset.h
@@ -14,7 +14,6 @@
 #include "intel_engine_types.h"
 #include "intel_reset_types.h"
 
-struct drm_i915_private;
 struct i915_request;
 struct intel_engine_cs;
 struct intel_gt;
@@ -38,13 +37,19 @@ int intel_engine_reset(struct intel_engine_cs *engine,
 
 void __i915_request_reset(struct i915_request *rq, bool guilty);
 
-int __must_check intel_gt_reset_trylock(struct intel_gt *gt);
+int __must_check intel_gt_reset_trylock(struct intel_gt *gt, int *srcu);
 void intel_gt_reset_unlock(struct intel_gt *gt, int tag);
 
 void intel_gt_set_wedged(struct intel_gt *gt);
 bool intel_gt_unset_wedged(struct intel_gt *gt);
 int intel_gt_terminally_wedged(struct intel_gt *gt);
 
+/*
+ * There's no unset_wedged_on_init paired with this one.
+ * Once we're wedged on init, there's no going back.
+ */
+void intel_gt_set_wedged_on_init(struct intel_gt *gt);
+
 int __intel_gt_reset(struct intel_gt *gt, intel_engine_mask_t engine_mask);
 
 int intel_reset_guc(struct intel_gt *gt);
@@ -68,10 +73,13 @@ void __intel_fini_wedge(struct intel_wedge_me *w);
 
 static inline bool __intel_reset_failed(const struct intel_reset *reset)
 {
+	GEM_BUG_ON(test_bit(I915_WEDGED_ON_INIT, &reset->flags) ?
+		   !test_bit(I915_WEDGED, &reset->flags) : false);
+
 	return unlikely(test_bit(I915_WEDGED, &reset->flags));
 }
 
-bool intel_has_gpu_reset(struct drm_i915_private *i915);
-bool intel_has_reset_engine(struct drm_i915_private *i915);
+bool intel_has_gpu_reset(const struct intel_gt *gt);
+bool intel_has_reset_engine(const struct intel_gt *gt);
 
 #endif /* I915_RESET_H */
diff --git a/drivers/gpu/drm/i915/gt/intel_reset_types.h b/drivers/gpu/drm/i915/gt/intel_reset_types.h
index 31968356e0c0..f43bc3a0fe4f 100644
--- a/drivers/gpu/drm/i915/gt/intel_reset_types.h
+++ b/drivers/gpu/drm/i915/gt/intel_reset_types.h
@@ -29,11 +29,17 @@ struct intel_reset {
 	 * we set the #I915_WEDGED bit. Prior to command submission, e.g.
 	 * i915_request_alloc(), this bit is checked and the sequence
 	 * aborted (with -EIO reported to userspace) if set.
+	 *
+	 * #I915_WEDGED_ON_INIT - If we fail to initialize the GPU we can no
+	 * longer use the GPU - similar to #I915_WEDGED bit. The difference in
+	 * in the way we're handling "forced" unwedged (e.g. through debugfs),
+	 * which is not allowed in case we failed to initialize.
 	 */
 	unsigned long flags;
 #define I915_RESET_BACKOFF	0
 #define I915_RESET_MODESET	1
 #define I915_RESET_ENGINE	2
+#define I915_WEDGED_ON_INIT	(BITS_PER_LONG - 2)
 #define I915_WEDGED		(BITS_PER_LONG - 1)
 
 	struct mutex mutex; /* serialises wedging/unwedging */
diff --git a/drivers/gpu/drm/i915/gt/intel_ring.c b/drivers/gpu/drm/i915/gt/intel_ring.c
new file mode 100644
index 000000000000..ece20504d240
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/intel_ring.c
@@ -0,0 +1,323 @@
+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2019 Intel Corporation
+ */
+
+#include "gem/i915_gem_object.h"
+#include "i915_drv.h"
+#include "i915_vma.h"
+#include "intel_engine.h"
+#include "intel_ring.h"
+#include "intel_timeline.h"
+
+unsigned int intel_ring_update_space(struct intel_ring *ring)
+{
+	unsigned int space;
+
+	space = __intel_ring_space(ring->head, ring->emit, ring->size);
+
+	ring->space = space;
+	return space;
+}
+
+int intel_ring_pin(struct intel_ring *ring)
+{
+	struct i915_vma *vma = ring->vma;
+	unsigned int flags;
+	void *addr;
+	int ret;
+
+	if (atomic_fetch_inc(&ring->pin_count))
+		return 0;
+
+	flags = PIN_GLOBAL;
+
+	/* Ring wraparound at offset 0 sometimes hangs. No idea why. */
+	flags |= PIN_OFFSET_BIAS | i915_ggtt_pin_bias(vma);
+
+	if (vma->obj->stolen)
+		flags |= PIN_MAPPABLE;
+	else
+		flags |= PIN_HIGH;
+
+	ret = i915_vma_pin(vma, 0, 0, flags);
+	if (unlikely(ret))
+		goto err_unpin;
+
+	if (i915_vma_is_map_and_fenceable(vma))
+		addr = (void __force *)i915_vma_pin_iomap(vma);
+	else
+		addr = i915_gem_object_pin_map(vma->obj,
+					       i915_coherent_map_type(vma->vm->i915));
+	if (IS_ERR(addr)) {
+		ret = PTR_ERR(addr);
+		goto err_ring;
+	}
+
+	i915_vma_make_unshrinkable(vma);
+
+	GEM_BUG_ON(ring->vaddr);
+	ring->vaddr = addr;
+
+	return 0;
+
+err_ring:
+	i915_vma_unpin(vma);
+err_unpin:
+	atomic_dec(&ring->pin_count);
+	return ret;
+}
+
+void intel_ring_reset(struct intel_ring *ring, u32 tail)
+{
+	tail = intel_ring_wrap(ring, tail);
+	ring->tail = tail;
+	ring->head = tail;
+	ring->emit = tail;
+	intel_ring_update_space(ring);
+}
+
+void intel_ring_unpin(struct intel_ring *ring)
+{
+	struct i915_vma *vma = ring->vma;
+
+	if (!atomic_dec_and_test(&ring->pin_count))
+		return;
+
+	/* Discard any unused bytes beyond that submitted to hw. */
+	intel_ring_reset(ring, ring->emit);
+
+	i915_vma_unset_ggtt_write(vma);
+	if (i915_vma_is_map_and_fenceable(vma))
+		i915_vma_unpin_iomap(vma);
+	else
+		i915_gem_object_unpin_map(vma->obj);
+
+	GEM_BUG_ON(!ring->vaddr);
+	ring->vaddr = NULL;
+
+	i915_vma_unpin(vma);
+	i915_vma_make_purgeable(vma);
+}
+
+static struct i915_vma *create_ring_vma(struct i915_ggtt *ggtt, int size)
+{
+	struct i915_address_space *vm = &ggtt->vm;
+	struct drm_i915_private *i915 = vm->i915;
+	struct drm_i915_gem_object *obj;
+	struct i915_vma *vma;
+
+	obj = ERR_PTR(-ENODEV);
+	if (i915_ggtt_has_aperture(ggtt))
+		obj = i915_gem_object_create_stolen(i915, size);
+	if (IS_ERR(obj))
+		obj = i915_gem_object_create_internal(i915, size);
+	if (IS_ERR(obj))
+		return ERR_CAST(obj);
+
+	/*
+	 * Mark ring buffers as read-only from GPU side (so no stray overwrites)
+	 * if supported by the platform's GGTT.
+	 */
+	if (vm->has_read_only)
+		i915_gem_object_set_readonly(obj);
+
+	vma = i915_vma_instance(obj, vm, NULL);
+	if (IS_ERR(vma))
+		goto err;
+
+	return vma;
+
+err:
+	i915_gem_object_put(obj);
+	return vma;
+}
+
+struct intel_ring *
+intel_engine_create_ring(struct intel_engine_cs *engine, int size)
+{
+	struct drm_i915_private *i915 = engine->i915;
+	struct intel_ring *ring;
+	struct i915_vma *vma;
+
+	GEM_BUG_ON(!is_power_of_2(size));
+	GEM_BUG_ON(RING_CTL_SIZE(size) & ~RING_NR_PAGES);
+
+	ring = kzalloc(sizeof(*ring), GFP_KERNEL);
+	if (!ring)
+		return ERR_PTR(-ENOMEM);
+
+	kref_init(&ring->ref);
+	ring->size = size;
+
+	/*
+	 * Workaround an erratum on the i830 which causes a hang if
+	 * the TAIL pointer points to within the last 2 cachelines
+	 * of the buffer.
+	 */
+	ring->effective_size = size;
+	if (IS_I830(i915) || IS_I845G(i915))
+		ring->effective_size -= 2 * CACHELINE_BYTES;
+
+	intel_ring_update_space(ring);
+
+	vma = create_ring_vma(engine->gt->ggtt, size);
+	if (IS_ERR(vma)) {
+		kfree(ring);
+		return ERR_CAST(vma);
+	}
+	ring->vma = vma;
+
+	return ring;
+}
+
+void intel_ring_free(struct kref *ref)
+{
+	struct intel_ring *ring = container_of(ref, typeof(*ring), ref);
+
+	i915_vma_put(ring->vma);
+	kfree(ring);
+}
+
+static noinline int
+wait_for_space(struct intel_ring *ring,
+	       struct intel_timeline *tl,
+	       unsigned int bytes)
+{
+	struct i915_request *target;
+	long timeout;
+
+	if (intel_ring_update_space(ring) >= bytes)
+		return 0;
+
+	GEM_BUG_ON(list_empty(&tl->requests));
+	list_for_each_entry(target, &tl->requests, link) {
+		if (target->ring != ring)
+			continue;
+
+		/* Would completion of this request free enough space? */
+		if (bytes <= __intel_ring_space(target->postfix,
+						ring->emit, ring->size))
+			break;
+	}
+
+	if (GEM_WARN_ON(&target->link == &tl->requests))
+		return -ENOSPC;
+
+	timeout = i915_request_wait(target,
+				    I915_WAIT_INTERRUPTIBLE,
+				    MAX_SCHEDULE_TIMEOUT);
+	if (timeout < 0)
+		return timeout;
+
+	i915_request_retire_upto(target);
+
+	intel_ring_update_space(ring);
+	GEM_BUG_ON(ring->space < bytes);
+	return 0;
+}
+
+u32 *intel_ring_begin(struct i915_request *rq, unsigned int num_dwords)
+{
+	struct intel_ring *ring = rq->ring;
+	const unsigned int remain_usable = ring->effective_size - ring->emit;
+	const unsigned int bytes = num_dwords * sizeof(u32);
+	unsigned int need_wrap = 0;
+	unsigned int total_bytes;
+	u32 *cs;
+
+	/* Packets must be qword aligned. */
+	GEM_BUG_ON(num_dwords & 1);
+
+	total_bytes = bytes + rq->reserved_space;
+	GEM_BUG_ON(total_bytes > ring->effective_size);
+
+	if (unlikely(total_bytes > remain_usable)) {
+		const int remain_actual = ring->size - ring->emit;
+
+		if (bytes > remain_usable) {
+			/*
+			 * Not enough space for the basic request. So need to
+			 * flush out the remainder and then wait for
+			 * base + reserved.
+			 */
+			total_bytes += remain_actual;
+			need_wrap = remain_actual | 1;
+		} else  {
+			/*
+			 * The base request will fit but the reserved space
+			 * falls off the end. So we don't need an immediate
+			 * wrap and only need to effectively wait for the
+			 * reserved size from the start of ringbuffer.
+			 */
+			total_bytes = rq->reserved_space + remain_actual;
+		}
+	}
+
+	if (unlikely(total_bytes > ring->space)) {
+		int ret;
+
+		/*
+		 * Space is reserved in the ringbuffer for finalising the
+		 * request, as that cannot be allowed to fail. During request
+		 * finalisation, reserved_space is set to 0 to stop the
+		 * overallocation and the assumption is that then we never need
+		 * to wait (which has the risk of failing with EINTR).
+		 *
+		 * See also i915_request_alloc() and i915_request_add().
+		 */
+		GEM_BUG_ON(!rq->reserved_space);
+
+		ret = wait_for_space(ring,
+				     i915_request_timeline(rq),
+				     total_bytes);
+		if (unlikely(ret))
+			return ERR_PTR(ret);
+	}
+
+	if (unlikely(need_wrap)) {
+		need_wrap &= ~1;
+		GEM_BUG_ON(need_wrap > ring->space);
+		GEM_BUG_ON(ring->emit + need_wrap > ring->size);
+		GEM_BUG_ON(!IS_ALIGNED(need_wrap, sizeof(u64)));
+
+		/* Fill the tail with MI_NOOP */
+		memset64(ring->vaddr + ring->emit, 0, need_wrap / sizeof(u64));
+		ring->space -= need_wrap;
+		ring->emit = 0;
+	}
+
+	GEM_BUG_ON(ring->emit > ring->size - bytes);
+	GEM_BUG_ON(ring->space < bytes);
+	cs = ring->vaddr + ring->emit;
+	GEM_DEBUG_EXEC(memset32(cs, POISON_INUSE, bytes / sizeof(*cs)));
+	ring->emit += bytes;
+	ring->space -= bytes;
+
+	return cs;
+}
+
+/* Align the ring tail to a cacheline boundary */
+int intel_ring_cacheline_align(struct i915_request *rq)
+{
+	int num_dwords;
+	void *cs;
+
+	num_dwords = (rq->ring->emit & (CACHELINE_BYTES - 1)) / sizeof(u32);
+	if (num_dwords == 0)
+		return 0;
+
+	num_dwords = CACHELINE_DWORDS - num_dwords;
+	GEM_BUG_ON(num_dwords & 1);
+
+	cs = intel_ring_begin(rq, num_dwords);
+	if (IS_ERR(cs))
+		return PTR_ERR(cs);
+
+	memset64(cs, (u64)MI_NOOP << 32 | MI_NOOP, num_dwords / 2);
+	intel_ring_advance(rq, cs + num_dwords);
+
+	GEM_BUG_ON(rq->ring->emit & (CACHELINE_BYTES - 1));
+	return 0;
+}
diff --git a/drivers/gpu/drm/i915/gt/intel_ring.h b/drivers/gpu/drm/i915/gt/intel_ring.h
new file mode 100644
index 000000000000..ea2839d9e044
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/intel_ring.h
@@ -0,0 +1,131 @@
+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2019 Intel Corporation
+ */
+
+#ifndef INTEL_RING_H
+#define INTEL_RING_H
+
+#include "i915_gem.h" /* GEM_BUG_ON */
+#include "i915_request.h"
+#include "intel_ring_types.h"
+
+struct intel_engine_cs;
+
+struct intel_ring *
+intel_engine_create_ring(struct intel_engine_cs *engine, int size);
+
+u32 *intel_ring_begin(struct i915_request *rq, unsigned int num_dwords);
+int intel_ring_cacheline_align(struct i915_request *rq);
+
+unsigned int intel_ring_update_space(struct intel_ring *ring);
+
+int intel_ring_pin(struct intel_ring *ring);
+void intel_ring_unpin(struct intel_ring *ring);
+void intel_ring_reset(struct intel_ring *ring, u32 tail);
+
+void intel_ring_free(struct kref *ref);
+
+static inline struct intel_ring *intel_ring_get(struct intel_ring *ring)
+{
+	kref_get(&ring->ref);
+	return ring;
+}
+
+static inline void intel_ring_put(struct intel_ring *ring)
+{
+	kref_put(&ring->ref, intel_ring_free);
+}
+
+static inline void intel_ring_advance(struct i915_request *rq, u32 *cs)
+{
+	/* Dummy function.
+	 *
+	 * This serves as a placeholder in the code so that the reader
+	 * can compare against the preceding intel_ring_begin() and
+	 * check that the number of dwords emitted matches the space
+	 * reserved for the command packet (i.e. the value passed to
+	 * intel_ring_begin()).
+	 */
+	GEM_BUG_ON((rq->ring->vaddr + rq->ring->emit) != cs);
+}
+
+static inline u32 intel_ring_wrap(const struct intel_ring *ring, u32 pos)
+{
+	return pos & (ring->size - 1);
+}
+
+static inline bool
+intel_ring_offset_valid(const struct intel_ring *ring,
+			unsigned int pos)
+{
+	if (pos & -ring->size) /* must be strictly within the ring */
+		return false;
+
+	if (!IS_ALIGNED(pos, 8)) /* must be qword aligned */
+		return false;
+
+	return true;
+}
+
+static inline u32 intel_ring_offset(const struct i915_request *rq, void *addr)
+{
+	/* Don't write ring->size (equivalent to 0) as that hangs some GPUs. */
+	u32 offset = addr - rq->ring->vaddr;
+	GEM_BUG_ON(offset > rq->ring->size);
+	return intel_ring_wrap(rq->ring, offset);
+}
+
+static inline void
+assert_ring_tail_valid(const struct intel_ring *ring, unsigned int tail)
+{
+	GEM_BUG_ON(!intel_ring_offset_valid(ring, tail));
+
+	/*
+	 * "Ring Buffer Use"
+	 *	Gen2 BSpec "1. Programming Environment" / 1.4.4.6
+	 *	Gen3 BSpec "1c Memory Interface Functions" / 2.3.4.5
+	 *	Gen4+ BSpec "1c Memory Interface and Command Stream" / 5.3.4.5
+	 * "If the Ring Buffer Head Pointer and the Tail Pointer are on the
+	 * same cacheline, the Head Pointer must not be greater than the Tail
+	 * Pointer."
+	 *
+	 * We use ring->head as the last known location of the actual RING_HEAD,
+	 * it may have advanced but in the worst case it is equally the same
+	 * as ring->head and so we should never program RING_TAIL to advance
+	 * into the same cacheline as ring->head.
+	 */
+#define cacheline(a) round_down(a, CACHELINE_BYTES)
+	GEM_BUG_ON(cacheline(tail) == cacheline(ring->head) &&
+		   tail < ring->head);
+#undef cacheline
+}
+
+static inline unsigned int
+intel_ring_set_tail(struct intel_ring *ring, unsigned int tail)
+{
+	/* Whilst writes to the tail are strictly order, there is no
+	 * serialisation between readers and the writers. The tail may be
+	 * read by i915_request_retire() just as it is being updated
+	 * by execlists, as although the breadcrumb is complete, the context
+	 * switch hasn't been seen.
+	 */
+	assert_ring_tail_valid(ring, tail);
+	ring->tail = tail;
+	return tail;
+}
+
+static inline unsigned int
+__intel_ring_space(unsigned int head, unsigned int tail, unsigned int size)
+{
+	/*
+	 * "If the Ring Buffer Head Pointer and the Tail Pointer are on the
+	 * same cacheline, the Head Pointer must not be greater than the Tail
+	 * Pointer."
+	 */
+	GEM_BUG_ON(!is_power_of_2(size));
+	return (head - tail - CACHELINE_BYTES) & (size - 1);
+}
+
+#endif /* INTEL_RING_H */
diff --git a/drivers/gpu/drm/i915/gt/intel_ringbuffer.c b/drivers/gpu/drm/i915/gt/intel_ring_submission.c
index 601c16239fdf..a47d5a7c32c9 100644
--- a/drivers/gpu/drm/i915/gt/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/gt/intel_ring_submission.c
@@ -40,6 +40,7 @@
 #include "intel_gt_irq.h"
 #include "intel_gt_pm_irq.h"
 #include "intel_reset.h"
+#include "intel_ring.h"
 #include "intel_workarounds.h"
 
 /* Rough estimate of the typical request size, performing a flush,
@@ -47,16 +48,6 @@
  */
 #define LEGACY_REQUEST_SIZE 200
 
-unsigned int intel_ring_update_space(struct intel_ring *ring)
-{
-	unsigned int space;
-
-	space = __intel_ring_space(ring->head, ring->emit, ring->size);
-
-	ring->space = space;
-	return space;
-}
-
 static int
 gen2_render_ring_flush(struct i915_request *rq, u32 mode)
 {
@@ -322,7 +313,8 @@ static u32 *gen6_rcs_emit_breadcrumb(struct i915_request *rq, u32 *cs)
 		 PIPE_CONTROL_DC_FLUSH_ENABLE |
 		 PIPE_CONTROL_QW_WRITE |
 		 PIPE_CONTROL_CS_STALL);
-	*cs++ = rq->timeline->hwsp_offset | PIPE_CONTROL_GLOBAL_GTT;
+	*cs++ = i915_request_active_timeline(rq)->hwsp_offset |
+		PIPE_CONTROL_GLOBAL_GTT;
 	*cs++ = rq->fence.seqno;
 
 	*cs++ = MI_USER_INTERRUPT;
@@ -425,7 +417,7 @@ static u32 *gen7_rcs_emit_breadcrumb(struct i915_request *rq, u32 *cs)
 		 PIPE_CONTROL_QW_WRITE |
 		 PIPE_CONTROL_GLOBAL_GTT_IVB |
 		 PIPE_CONTROL_CS_STALL);
-	*cs++ = rq->timeline->hwsp_offset;
+	*cs++ = i915_request_active_timeline(rq)->hwsp_offset;
 	*cs++ = rq->fence.seqno;
 
 	*cs++ = MI_USER_INTERRUPT;
@@ -439,8 +431,8 @@ static u32 *gen7_rcs_emit_breadcrumb(struct i915_request *rq, u32 *cs)
 
 static u32 *gen6_xcs_emit_breadcrumb(struct i915_request *rq, u32 *cs)
 {
-	GEM_BUG_ON(rq->timeline->hwsp_ggtt != rq->engine->status_page.vma);
-	GEM_BUG_ON(offset_in_page(rq->timeline->hwsp_offset) != I915_GEM_HWS_SEQNO_ADDR);
+	GEM_BUG_ON(i915_request_active_timeline(rq)->hwsp_ggtt != rq->engine->status_page.vma);
+	GEM_BUG_ON(offset_in_page(i915_request_active_timeline(rq)->hwsp_offset) != I915_GEM_HWS_SEQNO_ADDR);
 
 	*cs++ = MI_FLUSH_DW | MI_FLUSH_DW_OP_STOREDW | MI_FLUSH_DW_STORE_INDEX;
 	*cs++ = I915_GEM_HWS_SEQNO_ADDR | MI_FLUSH_DW_USE_GTT;
@@ -459,8 +451,8 @@ static u32 *gen7_xcs_emit_breadcrumb(struct i915_request *rq, u32 *cs)
 {
 	int i;
 
-	GEM_BUG_ON(rq->timeline->hwsp_ggtt != rq->engine->status_page.vma);
-	GEM_BUG_ON(offset_in_page(rq->timeline->hwsp_offset) != I915_GEM_HWS_SEQNO_ADDR);
+	GEM_BUG_ON(i915_request_active_timeline(rq)->hwsp_ggtt != rq->engine->status_page.vma);
+	GEM_BUG_ON(offset_in_page(i915_request_active_timeline(rq)->hwsp_offset) != I915_GEM_HWS_SEQNO_ADDR);
 
 	*cs++ = MI_FLUSH_DW | MI_FLUSH_DW_OP_STOREDW | MI_FLUSH_DW_STORE_INDEX;
 	*cs++ = I915_GEM_HWS_SEQNO_ADDR | MI_FLUSH_DW_USE_GTT;
@@ -930,6 +922,7 @@ static void cancel_requests(struct intel_engine_cs *engine)
 static void i9xx_submit_request(struct i915_request *request)
 {
 	i915_request_submit(request);
+	wmb(); /* paranoid flush writes out of the WCB before mmio */
 
 	ENGINE_WRITE(request->engine, RING_TAIL,
 		     intel_ring_set_tail(request->ring, request->tail));
@@ -937,8 +930,8 @@ static void i9xx_submit_request(struct i915_request *request)
 
 static u32 *i9xx_emit_breadcrumb(struct i915_request *rq, u32 *cs)
 {
-	GEM_BUG_ON(rq->timeline->hwsp_ggtt != rq->engine->status_page.vma);
-	GEM_BUG_ON(offset_in_page(rq->timeline->hwsp_offset) != I915_GEM_HWS_SEQNO_ADDR);
+	GEM_BUG_ON(i915_request_active_timeline(rq)->hwsp_ggtt != rq->engine->status_page.vma);
+	GEM_BUG_ON(offset_in_page(i915_request_active_timeline(rq)->hwsp_offset) != I915_GEM_HWS_SEQNO_ADDR);
 
 	*cs++ = MI_FLUSH;
 
@@ -960,8 +953,8 @@ static u32 *gen5_emit_breadcrumb(struct i915_request *rq, u32 *cs)
 {
 	int i;
 
-	GEM_BUG_ON(rq->timeline->hwsp_ggtt != rq->engine->status_page.vma);
-	GEM_BUG_ON(offset_in_page(rq->timeline->hwsp_offset) != I915_GEM_HWS_SEQNO_ADDR);
+	GEM_BUG_ON(i915_request_active_timeline(rq)->hwsp_ggtt != rq->engine->status_page.vma);
+	GEM_BUG_ON(offset_in_page(i915_request_active_timeline(rq)->hwsp_offset) != I915_GEM_HWS_SEQNO_ADDR);
 
 	*cs++ = MI_FLUSH;
 
@@ -1184,167 +1177,9 @@ i915_emit_bb_start(struct i915_request *rq,
 	return 0;
 }
 
-int intel_ring_pin(struct intel_ring *ring)
-{
-	struct i915_vma *vma = ring->vma;
-	unsigned int flags;
-	void *addr;
-	int ret;
-
-	if (atomic_fetch_inc(&ring->pin_count))
-		return 0;
-
-	flags = PIN_GLOBAL;
-
-	/* Ring wraparound at offset 0 sometimes hangs. No idea why. */
-	flags |= PIN_OFFSET_BIAS | i915_ggtt_pin_bias(vma);
-
-	if (vma->obj->stolen)
-		flags |= PIN_MAPPABLE;
-	else
-		flags |= PIN_HIGH;
-
-	ret = i915_vma_pin(vma, 0, 0, flags);
-	if (unlikely(ret))
-		goto err_unpin;
-
-	if (i915_vma_is_map_and_fenceable(vma))
-		addr = (void __force *)i915_vma_pin_iomap(vma);
-	else
-		addr = i915_gem_object_pin_map(vma->obj,
-					       i915_coherent_map_type(vma->vm->i915));
-	if (IS_ERR(addr)) {
-		ret = PTR_ERR(addr);
-		goto err_ring;
-	}
-
-	i915_vma_make_unshrinkable(vma);
-
-	GEM_BUG_ON(ring->vaddr);
-	ring->vaddr = addr;
-
-	return 0;
-
-err_ring:
-	i915_vma_unpin(vma);
-err_unpin:
-	atomic_dec(&ring->pin_count);
-	return ret;
-}
-
-void intel_ring_reset(struct intel_ring *ring, u32 tail)
-{
-	tail = intel_ring_wrap(ring, tail);
-	ring->tail = tail;
-	ring->head = tail;
-	ring->emit = tail;
-	intel_ring_update_space(ring);
-}
-
-void intel_ring_unpin(struct intel_ring *ring)
-{
-	struct i915_vma *vma = ring->vma;
-
-	if (!atomic_dec_and_test(&ring->pin_count))
-		return;
-
-	/* Discard any unused bytes beyond that submitted to hw. */
-	intel_ring_reset(ring, ring->emit);
-
-	i915_vma_unset_ggtt_write(vma);
-	if (i915_vma_is_map_and_fenceable(vma))
-		i915_vma_unpin_iomap(vma);
-	else
-		i915_gem_object_unpin_map(vma->obj);
-
-	GEM_BUG_ON(!ring->vaddr);
-	ring->vaddr = NULL;
-
-	i915_vma_unpin(vma);
-	i915_vma_make_purgeable(vma);
-}
-
-static struct i915_vma *create_ring_vma(struct i915_ggtt *ggtt, int size)
-{
-	struct i915_address_space *vm = &ggtt->vm;
-	struct drm_i915_private *i915 = vm->i915;
-	struct drm_i915_gem_object *obj;
-	struct i915_vma *vma;
-
-	obj = i915_gem_object_create_stolen(i915, size);
-	if (!obj)
-		obj = i915_gem_object_create_internal(i915, size);
-	if (IS_ERR(obj))
-		return ERR_CAST(obj);
-
-	/*
-	 * Mark ring buffers as read-only from GPU side (so no stray overwrites)
-	 * if supported by the platform's GGTT.
-	 */
-	if (vm->has_read_only)
-		i915_gem_object_set_readonly(obj);
-
-	vma = i915_vma_instance(obj, vm, NULL);
-	if (IS_ERR(vma))
-		goto err;
-
-	return vma;
-
-err:
-	i915_gem_object_put(obj);
-	return vma;
-}
-
-struct intel_ring *
-intel_engine_create_ring(struct intel_engine_cs *engine, int size)
-{
-	struct drm_i915_private *i915 = engine->i915;
-	struct intel_ring *ring;
-	struct i915_vma *vma;
-
-	GEM_BUG_ON(!is_power_of_2(size));
-	GEM_BUG_ON(RING_CTL_SIZE(size) & ~RING_NR_PAGES);
-
-	ring = kzalloc(sizeof(*ring), GFP_KERNEL);
-	if (!ring)
-		return ERR_PTR(-ENOMEM);
-
-	kref_init(&ring->ref);
-
-	ring->size = size;
-	/* Workaround an erratum on the i830 which causes a hang if
-	 * the TAIL pointer points to within the last 2 cachelines
-	 * of the buffer.
-	 */
-	ring->effective_size = size;
-	if (IS_I830(i915) || IS_I845G(i915))
-		ring->effective_size -= 2 * CACHELINE_BYTES;
-
-	intel_ring_update_space(ring);
-
-	vma = create_ring_vma(engine->gt->ggtt, size);
-	if (IS_ERR(vma)) {
-		kfree(ring);
-		return ERR_CAST(vma);
-	}
-	ring->vma = vma;
-
-	return ring;
-}
-
-void intel_ring_free(struct kref *ref)
-{
-	struct intel_ring *ring = container_of(ref, typeof(*ring), ref);
-
-	i915_vma_close(ring->vma);
-	i915_vma_put(ring->vma);
-
-	kfree(ring);
-}
-
 static void __ring_context_fini(struct intel_context *ce)
 {
-	i915_gem_object_put(ce->state->obj);
+	i915_vma_put(ce->state);
 }
 
 static void ring_context_destroy(struct kref *ref)
@@ -1573,7 +1408,7 @@ static inline int mi_set_context(struct i915_request *rq, u32 flags)
 	struct intel_engine_cs *engine = rq->engine;
 	enum intel_engine_id id;
 	const int num_engines =
-		IS_HSW_GT1(i915) ? RUNTIME_INFO(i915)->num_engines - 1 : 0;
+		IS_HASWELL(i915) ? RUNTIME_INFO(i915)->num_engines - 1 : 0;
 	bool force_restore = false;
 	int len;
 	u32 *cs;
@@ -1609,7 +1444,7 @@ static inline int mi_set_context(struct i915_request *rq, u32 flags)
 			struct intel_engine_cs *signaller;
 
 			*cs++ = MI_LOAD_REGISTER_IMM(num_engines);
-			for_each_engine(signaller, i915, id) {
+			for_each_engine(signaller, engine->gt, id) {
 				if (signaller == engine)
 					continue;
 
@@ -1663,7 +1498,7 @@ static inline int mi_set_context(struct i915_request *rq, u32 flags)
 			i915_reg_t last_reg = {}; /* keep gcc quiet */
 
 			*cs++ = MI_LOAD_REGISTER_IMM(num_engines);
-			for_each_engine(signaller, i915, id) {
+			for_each_engine(signaller, engine->gt, id) {
 				if (signaller == engine)
 					continue;
 
@@ -1676,7 +1511,7 @@ static inline int mi_set_context(struct i915_request *rq, u32 flags)
 			/* Insert a delay before the next switch! */
 			*cs++ = MI_STORE_REGISTER_MEM | MI_SRM_LRM_GLOBAL_GTT;
 			*cs++ = i915_mmio_reg_offset(last_reg);
-			*cs++ = intel_gt_scratch_offset(rq->engine->gt,
+			*cs++ = intel_gt_scratch_offset(engine->gt,
 							INTEL_GT_SCRATCH_FIELD_DEFAULT);
 			*cs++ = MI_NOOP;
 		}
@@ -1741,46 +1576,22 @@ static int remap_l3(struct i915_request *rq)
 
 static int switch_context(struct i915_request *rq)
 {
-	struct intel_engine_cs *engine = rq->engine;
-	struct i915_address_space *vm = vm_alias(rq->hw_context);
-	unsigned int unwind_mm = 0;
-	u32 hw_flags = 0;
+	struct intel_context *ce = rq->hw_context;
+	struct i915_address_space *vm = vm_alias(ce);
 	int ret;
 
 	GEM_BUG_ON(HAS_EXECLISTS(rq->i915));
 
 	if (vm) {
-		struct i915_ppgtt *ppgtt = i915_vm_to_ppgtt(vm);
-		int loops;
-
-		/*
-		 * Baytail takes a little more convincing that it really needs
-		 * to reload the PD between contexts. It is not just a little
-		 * longer, as adding more stalls after the load_pd_dir (i.e.
-		 * adding a long loop around flush_pd_dir) is not as effective
-		 * as reloading the PD umpteen times. 32 is derived from
-		 * experimentation (gem_exec_parallel/fds) and has no good
-		 * explanation.
-		 */
-		loops = 1;
-		if (engine->id == BCS0 && IS_VALLEYVIEW(engine->i915))
-			loops = 32;
-
-		do {
-			ret = load_pd_dir(rq, ppgtt);
-			if (ret)
-				goto err;
-		} while (--loops);
-
-		if (ppgtt->pd_dirty_engines & engine->mask) {
-			unwind_mm = engine->mask;
-			ppgtt->pd_dirty_engines &= ~unwind_mm;
-			hw_flags = MI_FORCE_RESTORE;
-		}
+		ret = load_pd_dir(rq, i915_vm_to_ppgtt(vm));
+		if (ret)
+			return ret;
 	}
 
-	if (rq->hw_context->state) {
-		GEM_BUG_ON(engine->id != RCS0);
+	if (ce->state) {
+		u32 hw_flags;
+
+		GEM_BUG_ON(rq->engine->id != RCS0);
 
 		/*
 		 * The kernel context(s) is treated as pure scratch and is not
@@ -1789,22 +1600,25 @@ static int switch_context(struct i915_request *rq)
 		 * as nothing actually executes using the kernel context; it
 		 * is purely used for flushing user contexts.
 		 */
+		hw_flags = 0;
 		if (i915_gem_context_is_kernel(rq->gem_context))
 			hw_flags = MI_RESTORE_INHIBIT;
 
 		ret = mi_set_context(rq, hw_flags);
 		if (ret)
-			goto err_mm;
+			return ret;
 	}
 
 	if (vm) {
+		struct intel_engine_cs *engine = rq->engine;
+
 		ret = engine->emit_flush(rq, EMIT_INVALIDATE);
 		if (ret)
-			goto err_mm;
+			return ret;
 
 		ret = flush_pd_dir(rq);
 		if (ret)
-			goto err_mm;
+			return ret;
 
 		/*
 		 * Not only do we need a full barrier (post-sync write) after
@@ -1816,24 +1630,18 @@ static int switch_context(struct i915_request *rq)
 		 */
 		ret = engine->emit_flush(rq, EMIT_INVALIDATE);
 		if (ret)
-			goto err_mm;
+			return ret;
 
 		ret = engine->emit_flush(rq, EMIT_FLUSH);
 		if (ret)
-			goto err_mm;
+			return ret;
 	}
 
 	ret = remap_l3(rq);
 	if (ret)
-		goto err_mm;
+		return ret;
 
 	return 0;
-
-err_mm:
-	if (unwind_mm)
-		i915_vm_to_ppgtt(vm)->pd_dirty_engines |= unwind_mm;
-err:
-	return ret;
 }
 
 static int ring_request_alloc(struct i915_request *request)
@@ -1841,7 +1649,7 @@ static int ring_request_alloc(struct i915_request *request)
 	int ret;
 
 	GEM_BUG_ON(!intel_context_is_pinned(request->hw_context));
-	GEM_BUG_ON(request->timeline->has_initial_breadcrumb);
+	GEM_BUG_ON(i915_request_timeline(request)->has_initial_breadcrumb);
 
 	/*
 	 * Flush enough space to reduce the likelihood of waiting after
@@ -1863,146 +1671,6 @@ static int ring_request_alloc(struct i915_request *request)
 	return 0;
 }
 
-static noinline int
-wait_for_space(struct intel_ring *ring,
-	       struct intel_timeline *tl,
-	       unsigned int bytes)
-{
-	struct i915_request *target;
-	long timeout;
-
-	if (intel_ring_update_space(ring) >= bytes)
-		return 0;
-
-	GEM_BUG_ON(list_empty(&tl->requests));
-	list_for_each_entry(target, &tl->requests, link) {
-		if (target->ring != ring)
-			continue;
-
-		/* Would completion of this request free enough space? */
-		if (bytes <= __intel_ring_space(target->postfix,
-						ring->emit, ring->size))
-			break;
-	}
-
-	if (GEM_WARN_ON(&target->link == &tl->requests))
-		return -ENOSPC;
-
-	timeout = i915_request_wait(target,
-				    I915_WAIT_INTERRUPTIBLE,
-				    MAX_SCHEDULE_TIMEOUT);
-	if (timeout < 0)
-		return timeout;
-
-	i915_request_retire_upto(target);
-
-	intel_ring_update_space(ring);
-	GEM_BUG_ON(ring->space < bytes);
-	return 0;
-}
-
-u32 *intel_ring_begin(struct i915_request *rq, unsigned int num_dwords)
-{
-	struct intel_ring *ring = rq->ring;
-	const unsigned int remain_usable = ring->effective_size - ring->emit;
-	const unsigned int bytes = num_dwords * sizeof(u32);
-	unsigned int need_wrap = 0;
-	unsigned int total_bytes;
-	u32 *cs;
-
-	/* Packets must be qword aligned. */
-	GEM_BUG_ON(num_dwords & 1);
-
-	total_bytes = bytes + rq->reserved_space;
-	GEM_BUG_ON(total_bytes > ring->effective_size);
-
-	if (unlikely(total_bytes > remain_usable)) {
-		const int remain_actual = ring->size - ring->emit;
-
-		if (bytes > remain_usable) {
-			/*
-			 * Not enough space for the basic request. So need to
-			 * flush out the remainder and then wait for
-			 * base + reserved.
-			 */
-			total_bytes += remain_actual;
-			need_wrap = remain_actual | 1;
-		} else  {
-			/*
-			 * The base request will fit but the reserved space
-			 * falls off the end. So we don't need an immediate
-			 * wrap and only need to effectively wait for the
-			 * reserved size from the start of ringbuffer.
-			 */
-			total_bytes = rq->reserved_space + remain_actual;
-		}
-	}
-
-	if (unlikely(total_bytes > ring->space)) {
-		int ret;
-
-		/*
-		 * Space is reserved in the ringbuffer for finalising the
-		 * request, as that cannot be allowed to fail. During request
-		 * finalisation, reserved_space is set to 0 to stop the
-		 * overallocation and the assumption is that then we never need
-		 * to wait (which has the risk of failing with EINTR).
-		 *
-		 * See also i915_request_alloc() and i915_request_add().
-		 */
-		GEM_BUG_ON(!rq->reserved_space);
-
-		ret = wait_for_space(ring, rq->timeline, total_bytes);
-		if (unlikely(ret))
-			return ERR_PTR(ret);
-	}
-
-	if (unlikely(need_wrap)) {
-		need_wrap &= ~1;
-		GEM_BUG_ON(need_wrap > ring->space);
-		GEM_BUG_ON(ring->emit + need_wrap > ring->size);
-		GEM_BUG_ON(!IS_ALIGNED(need_wrap, sizeof(u64)));
-
-		/* Fill the tail with MI_NOOP */
-		memset64(ring->vaddr + ring->emit, 0, need_wrap / sizeof(u64));
-		ring->space -= need_wrap;
-		ring->emit = 0;
-	}
-
-	GEM_BUG_ON(ring->emit > ring->size - bytes);
-	GEM_BUG_ON(ring->space < bytes);
-	cs = ring->vaddr + ring->emit;
-	GEM_DEBUG_EXEC(memset32(cs, POISON_INUSE, bytes / sizeof(*cs)));
-	ring->emit += bytes;
-	ring->space -= bytes;
-
-	return cs;
-}
-
-/* Align the ring tail to a cacheline boundary */
-int intel_ring_cacheline_align(struct i915_request *rq)
-{
-	int num_dwords;
-	void *cs;
-
-	num_dwords = (rq->ring->emit & (CACHELINE_BYTES - 1)) / sizeof(u32);
-	if (num_dwords == 0)
-		return 0;
-
-	num_dwords = CACHELINE_DWORDS - num_dwords;
-	GEM_BUG_ON(num_dwords & 1);
-
-	cs = intel_ring_begin(rq, num_dwords);
-	if (IS_ERR(cs))
-		return PTR_ERR(cs);
-
-	memset64(cs, (u64)MI_NOOP << 32 | MI_NOOP, num_dwords / 2);
-	intel_ring_advance(rq, cs);
-
-	GEM_BUG_ON(rq->ring->emit & (CACHELINE_BYTES - 1));
-	return 0;
-}
-
 static void gen6_bsd_submit_request(struct i915_request *request)
 {
 	struct intel_uncore *uncore = request->engine->uncore;
diff --git a/drivers/gpu/drm/i915/gt/intel_ring_types.h b/drivers/gpu/drm/i915/gt/intel_ring_types.h
new file mode 100644
index 000000000000..d9f17f38e0cc
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/intel_ring_types.h
@@ -0,0 +1,51 @@
+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2019 Intel Corporation
+ */
+
+#ifndef INTEL_RING_TYPES_H
+#define INTEL_RING_TYPES_H
+
+#include <linux/atomic.h>
+#include <linux/kref.h>
+#include <linux/types.h>
+
+/*
+ * Early gen2 devices have a cacheline of just 32 bytes, using 64 is overkill,
+ * but keeps the logic simple. Indeed, the whole purpose of this macro is just
+ * to give some inclination as to some of the magic values used in the various
+ * workarounds!
+ */
+#define CACHELINE_BYTES 64
+#define CACHELINE_DWORDS (CACHELINE_BYTES / sizeof(u32))
+
+struct i915_vma;
+
+struct intel_ring {
+	struct kref ref;
+	struct i915_vma *vma;
+	void *vaddr;
+
+	/*
+	 * As we have two types of rings, one global to the engine used
+	 * by ringbuffer submission and those that are exclusive to a
+	 * context used by execlists, we have to play safe and allow
+	 * atomic updates to the pin_count. However, the actual pinning
+	 * of the context is either done during initialisation for
+	 * ringbuffer submission or serialised as part of the context
+	 * pinning for execlists, and so we do not need a mutex ourselves
+	 * to serialise intel_ring_pin/intel_ring_unpin.
+	 */
+	atomic_t pin_count;
+
+	u32 head;
+	u32 tail;
+	u32 emit;
+
+	u32 space;
+	u32 size;
+	u32 effective_size;
+};
+
+#endif /* INTEL_RING_TYPES_H */
diff --git a/drivers/gpu/drm/i915/gt/intel_rps.c b/drivers/gpu/drm/i915/gt/intel_rps.c
new file mode 100644
index 000000000000..20d6ee148afc
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/intel_rps.c
@@ -0,0 +1,1872 @@
+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2019 Intel Corporation
+ */
+
+#include "i915_drv.h"
+#include "intel_gt.h"
+#include "intel_gt_irq.h"
+#include "intel_gt_pm_irq.h"
+#include "intel_rps.h"
+#include "intel_sideband.h"
+#include "../../../platform/x86/intel_ips.h"
+
+/*
+ * Lock protecting IPS related data structures
+ */
+static DEFINE_SPINLOCK(mchdev_lock);
+
+static struct intel_gt *rps_to_gt(struct intel_rps *rps)
+{
+	return container_of(rps, struct intel_gt, rps);
+}
+
+static struct drm_i915_private *rps_to_i915(struct intel_rps *rps)
+{
+	return rps_to_gt(rps)->i915;
+}
+
+static struct intel_uncore *rps_to_uncore(struct intel_rps *rps)
+{
+	return rps_to_gt(rps)->uncore;
+}
+
+static u32 rps_pm_sanitize_mask(struct intel_rps *rps, u32 mask)
+{
+	return mask & ~rps->pm_intrmsk_mbz;
+}
+
+static u32 rps_pm_mask(struct intel_rps *rps, u8 val)
+{
+	u32 mask = 0;
+
+	/* We use UP_EI_EXPIRED interrupts for both up/down in manual mode */
+	if (val > rps->min_freq_softlimit)
+		mask |= (GEN6_PM_RP_UP_EI_EXPIRED |
+			 GEN6_PM_RP_DOWN_THRESHOLD |
+			 GEN6_PM_RP_DOWN_TIMEOUT);
+
+	if (val < rps->max_freq_softlimit)
+		mask |= GEN6_PM_RP_UP_EI_EXPIRED | GEN6_PM_RP_UP_THRESHOLD;
+
+	mask &= rps->pm_events;
+
+	return rps_pm_sanitize_mask(rps, ~mask);
+}
+
+static void rps_reset_ei(struct intel_rps *rps)
+{
+	memset(&rps->ei, 0, sizeof(rps->ei));
+}
+
+static void rps_enable_interrupts(struct intel_rps *rps)
+{
+	struct intel_gt *gt = rps_to_gt(rps);
+
+	rps_reset_ei(rps);
+
+	if (IS_VALLEYVIEW(gt->i915))
+		/* WaGsvRC0ResidencyMethod:vlv */
+		rps->pm_events = GEN6_PM_RP_UP_EI_EXPIRED;
+	else
+		rps->pm_events = (GEN6_PM_RP_UP_THRESHOLD |
+				  GEN6_PM_RP_DOWN_THRESHOLD |
+				  GEN6_PM_RP_DOWN_TIMEOUT);
+
+	spin_lock_irq(&gt->irq_lock);
+	gen6_gt_pm_enable_irq(gt, rps->pm_events);
+	spin_unlock_irq(&gt->irq_lock);
+
+	intel_uncore_write(gt->uncore, GEN6_PMINTRMSK,
+			   rps_pm_mask(rps, rps->cur_freq));
+}
+
+static void gen6_rps_reset_interrupts(struct intel_rps *rps)
+{
+	gen6_gt_pm_reset_iir(rps_to_gt(rps), GEN6_PM_RPS_EVENTS);
+}
+
+static void gen11_rps_reset_interrupts(struct intel_rps *rps)
+{
+	while (gen11_gt_reset_one_iir(rps_to_gt(rps), 0, GEN11_GTPM))
+		;
+}
+
+static void rps_reset_interrupts(struct intel_rps *rps)
+{
+	struct intel_gt *gt = rps_to_gt(rps);
+
+	spin_lock_irq(&gt->irq_lock);
+	if (INTEL_GEN(gt->i915) >= 11)
+		gen11_rps_reset_interrupts(rps);
+	else
+		gen6_rps_reset_interrupts(rps);
+
+	rps->pm_iir = 0;
+	spin_unlock_irq(&gt->irq_lock);
+}
+
+static void rps_disable_interrupts(struct intel_rps *rps)
+{
+	struct intel_gt *gt = rps_to_gt(rps);
+
+	rps->pm_events = 0;
+
+	intel_uncore_write(gt->uncore, GEN6_PMINTRMSK,
+			   rps_pm_sanitize_mask(rps, ~0u));
+
+	spin_lock_irq(&gt->irq_lock);
+	gen6_gt_pm_disable_irq(gt, GEN6_PM_RPS_EVENTS);
+	spin_unlock_irq(&gt->irq_lock);
+
+	intel_synchronize_irq(gt->i915);
+
+	/*
+	 * Now that we will not be generating any more work, flush any
+	 * outstanding tasks. As we are called on the RPS idle path,
+	 * we will reset the GPU to minimum frequencies, so the current
+	 * state of the worker can be discarded.
+	 */
+	cancel_work_sync(&rps->work);
+
+	rps_reset_interrupts(rps);
+}
+
+static const struct cparams {
+	u16 i;
+	u16 t;
+	u16 m;
+	u16 c;
+} cparams[] = {
+	{ 1, 1333, 301, 28664 },
+	{ 1, 1066, 294, 24460 },
+	{ 1, 800, 294, 25192 },
+	{ 0, 1333, 276, 27605 },
+	{ 0, 1066, 276, 27605 },
+	{ 0, 800, 231, 23784 },
+};
+
+static void gen5_rps_init(struct intel_rps *rps)
+{
+	struct drm_i915_private *i915 = rps_to_i915(rps);
+	struct intel_uncore *uncore = rps_to_uncore(rps);
+	u8 fmax, fmin, fstart;
+	u32 rgvmodectl;
+	int c_m, i;
+
+	if (i915->fsb_freq <= 3200)
+		c_m = 0;
+	else if (i915->fsb_freq <= 4800)
+		c_m = 1;
+	else
+		c_m = 2;
+
+	for (i = 0; i < ARRAY_SIZE(cparams); i++) {
+		if (cparams[i].i == c_m && cparams[i].t == i915->mem_freq) {
+			rps->ips.m = cparams[i].m;
+			rps->ips.c = cparams[i].c;
+			break;
+		}
+	}
+
+	rgvmodectl = intel_uncore_read(uncore, MEMMODECTL);
+
+	/* Set up min, max, and cur for interrupt handling */
+	fmax = (rgvmodectl & MEMMODE_FMAX_MASK) >> MEMMODE_FMAX_SHIFT;
+	fmin = (rgvmodectl & MEMMODE_FMIN_MASK);
+	fstart = (rgvmodectl & MEMMODE_FSTART_MASK) >>
+		MEMMODE_FSTART_SHIFT;
+	DRM_DEBUG_DRIVER("fmax: %d, fmin: %d, fstart: %d\n",
+			 fmax, fmin, fstart);
+
+	rps->min_freq = fmax;
+	rps->max_freq = fmin;
+
+	rps->idle_freq = rps->min_freq;
+	rps->cur_freq = rps->idle_freq;
+}
+
+static unsigned long
+__ips_chipset_val(struct intel_ips *ips)
+{
+	struct intel_uncore *uncore =
+		rps_to_uncore(container_of(ips, struct intel_rps, ips));
+	unsigned long now = jiffies_to_msecs(jiffies), dt;
+	unsigned long result;
+	u64 total, delta;
+
+	lockdep_assert_held(&mchdev_lock);
+
+	/*
+	 * Prevent division-by-zero if we are asking too fast.
+	 * Also, we don't get interesting results if we are polling
+	 * faster than once in 10ms, so just return the saved value
+	 * in such cases.
+	 */
+	dt = now - ips->last_time1;
+	if (dt <= 10)
+		return ips->chipset_power;
+
+	/* FIXME: handle per-counter overflow */
+	total = intel_uncore_read(uncore, DMIEC);
+	total += intel_uncore_read(uncore, DDREC);
+	total += intel_uncore_read(uncore, CSIEC);
+
+	delta = total - ips->last_count1;
+
+	result = div_u64(div_u64(ips->m * delta, dt) + ips->c, 10);
+
+	ips->last_count1 = total;
+	ips->last_time1 = now;
+
+	ips->chipset_power = result;
+
+	return result;
+}
+
+static unsigned long ips_mch_val(struct intel_uncore *uncore)
+{
+	unsigned int m, x, b;
+	u32 tsfs;
+
+	tsfs = intel_uncore_read(uncore, TSFS);
+	x = intel_uncore_read8(uncore, TR1);
+
+	b = tsfs & TSFS_INTR_MASK;
+	m = (tsfs & TSFS_SLOPE_MASK) >> TSFS_SLOPE_SHIFT;
+
+	return m * x / 127 - b;
+}
+
+static int _pxvid_to_vd(u8 pxvid)
+{
+	if (pxvid == 0)
+		return 0;
+
+	if (pxvid >= 8 && pxvid < 31)
+		pxvid = 31;
+
+	return (pxvid + 2) * 125;
+}
+
+static u32 pvid_to_extvid(struct drm_i915_private *i915, u8 pxvid)
+{
+	const int vd = _pxvid_to_vd(pxvid);
+
+	if (INTEL_INFO(i915)->is_mobile)
+		return max(vd - 1125, 0);
+
+	return vd;
+}
+
+static void __gen5_ips_update(struct intel_ips *ips)
+{
+	struct intel_uncore *uncore =
+		rps_to_uncore(container_of(ips, struct intel_rps, ips));
+	u64 now, delta, dt;
+	u32 count;
+
+	lockdep_assert_held(&mchdev_lock);
+
+	now = ktime_get_raw_ns();
+	dt = now - ips->last_time2;
+	do_div(dt, NSEC_PER_MSEC);
+
+	/* Don't divide by 0 */
+	if (dt <= 10)
+		return;
+
+	count = intel_uncore_read(uncore, GFXEC);
+	delta = count - ips->last_count2;
+
+	ips->last_count2 = count;
+	ips->last_time2 = now;
+
+	/* More magic constants... */
+	ips->gfx_power = div_u64(delta * 1181, dt * 10);
+}
+
+static void gen5_rps_update(struct intel_rps *rps)
+{
+	spin_lock_irq(&mchdev_lock);
+	__gen5_ips_update(&rps->ips);
+	spin_unlock_irq(&mchdev_lock);
+}
+
+static bool gen5_rps_set(struct intel_rps *rps, u8 val)
+{
+	struct intel_uncore *uncore = rps_to_uncore(rps);
+	u16 rgvswctl;
+
+	lockdep_assert_held(&mchdev_lock);
+
+	rgvswctl = intel_uncore_read16(uncore, MEMSWCTL);
+	if (rgvswctl & MEMCTL_CMD_STS) {
+		DRM_DEBUG("gpu busy, RCS change rejected\n");
+		return false; /* still busy with another command */
+	}
+
+	/* Invert the frequency bin into an ips delay */
+	val = rps->max_freq - val;
+	val = rps->min_freq + val;
+
+	rgvswctl =
+		(MEMCTL_CMD_CHFREQ << MEMCTL_CMD_SHIFT) |
+		(val << MEMCTL_FREQ_SHIFT) |
+		MEMCTL_SFCAVM;
+	intel_uncore_write16(uncore, MEMSWCTL, rgvswctl);
+	intel_uncore_posting_read16(uncore, MEMSWCTL);
+
+	rgvswctl |= MEMCTL_CMD_STS;
+	intel_uncore_write16(uncore, MEMSWCTL, rgvswctl);
+
+	return true;
+}
+
+static unsigned long intel_pxfreq(u32 vidfreq)
+{
+	int div = (vidfreq & 0x3f0000) >> 16;
+	int post = (vidfreq & 0x3000) >> 12;
+	int pre = (vidfreq & 0x7);
+
+	if (!pre)
+		return 0;
+
+	return div * 133333 / (pre << post);
+}
+
+static unsigned int init_emon(struct intel_uncore *uncore)
+{
+	u8 pxw[16];
+	int i;
+
+	/* Disable to program */
+	intel_uncore_write(uncore, ECR, 0);
+	intel_uncore_posting_read(uncore, ECR);
+
+	/* Program energy weights for various events */
+	intel_uncore_write(uncore, SDEW, 0x15040d00);
+	intel_uncore_write(uncore, CSIEW0, 0x007f0000);
+	intel_uncore_write(uncore, CSIEW1, 0x1e220004);
+	intel_uncore_write(uncore, CSIEW2, 0x04000004);
+
+	for (i = 0; i < 5; i++)
+		intel_uncore_write(uncore, PEW(i), 0);
+	for (i = 0; i < 3; i++)
+		intel_uncore_write(uncore, DEW(i), 0);
+
+	/* Program P-state weights to account for frequency power adjustment */
+	for (i = 0; i < 16; i++) {
+		u32 pxvidfreq = intel_uncore_read(uncore, PXVFREQ(i));
+		unsigned int freq = intel_pxfreq(pxvidfreq);
+		unsigned int vid =
+			(pxvidfreq & PXVFREQ_PX_MASK) >> PXVFREQ_PX_SHIFT;
+		unsigned int val;
+
+		val = vid * vid * freq / 1000 * 255;
+		val /= 127 * 127 * 900;
+
+		pxw[i] = val;
+	}
+	/* Render standby states get 0 weight */
+	pxw[14] = 0;
+	pxw[15] = 0;
+
+	for (i = 0; i < 4; i++) {
+		intel_uncore_write(uncore, PXW(i),
+				   pxw[i * 4 + 0] << 24 |
+				   pxw[i * 4 + 1] << 16 |
+				   pxw[i * 4 + 2] <<  8 |
+				   pxw[i * 4 + 3] <<  0);
+	}
+
+	/* Adjust magic regs to magic values (more experimental results) */
+	intel_uncore_write(uncore, OGW0, 0);
+	intel_uncore_write(uncore, OGW1, 0);
+	intel_uncore_write(uncore, EG0, 0x00007f00);
+	intel_uncore_write(uncore, EG1, 0x0000000e);
+	intel_uncore_write(uncore, EG2, 0x000e0000);
+	intel_uncore_write(uncore, EG3, 0x68000300);
+	intel_uncore_write(uncore, EG4, 0x42000000);
+	intel_uncore_write(uncore, EG5, 0x00140031);
+	intel_uncore_write(uncore, EG6, 0);
+	intel_uncore_write(uncore, EG7, 0);
+
+	for (i = 0; i < 8; i++)
+		intel_uncore_write(uncore, PXWL(i), 0);
+
+	/* Enable PMON + select events */
+	intel_uncore_write(uncore, ECR, 0x80000019);
+
+	return intel_uncore_read(uncore, LCFUSE02) & LCFUSE_HIV_MASK;
+}
+
+static bool gen5_rps_enable(struct intel_rps *rps)
+{
+	struct intel_uncore *uncore = rps_to_uncore(rps);
+	u8 fstart, vstart;
+	u32 rgvmodectl;
+
+	spin_lock_irq(&mchdev_lock);
+
+	rgvmodectl = intel_uncore_read(uncore, MEMMODECTL);
+
+	/* Enable temp reporting */
+	intel_uncore_write16(uncore, PMMISC,
+			     intel_uncore_read16(uncore, PMMISC) | MCPPCE_EN);
+	intel_uncore_write16(uncore, TSC1,
+			     intel_uncore_read16(uncore, TSC1) | TSE);
+
+	/* 100ms RC evaluation intervals */
+	intel_uncore_write(uncore, RCUPEI, 100000);
+	intel_uncore_write(uncore, RCDNEI, 100000);
+
+	/* Set max/min thresholds to 90ms and 80ms respectively */
+	intel_uncore_write(uncore, RCBMAXAVG, 90000);
+	intel_uncore_write(uncore, RCBMINAVG, 80000);
+
+	intel_uncore_write(uncore, MEMIHYST, 1);
+
+	/* Set up min, max, and cur for interrupt handling */
+	fstart = (rgvmodectl & MEMMODE_FSTART_MASK) >>
+		MEMMODE_FSTART_SHIFT;
+
+	vstart = (intel_uncore_read(uncore, PXVFREQ(fstart)) &
+		  PXVFREQ_PX_MASK) >> PXVFREQ_PX_SHIFT;
+
+	intel_uncore_write(uncore,
+			   MEMINTREN,
+			   MEMINT_CX_SUPR_EN | MEMINT_EVAL_CHG_EN);
+
+	intel_uncore_write(uncore, VIDSTART, vstart);
+	intel_uncore_posting_read(uncore, VIDSTART);
+
+	rgvmodectl |= MEMMODE_SWMODE_EN;
+	intel_uncore_write(uncore, MEMMODECTL, rgvmodectl);
+
+	if (wait_for_atomic((intel_uncore_read(uncore, MEMSWCTL) &
+			     MEMCTL_CMD_STS) == 0, 10))
+		DRM_ERROR("stuck trying to change perf mode\n");
+	mdelay(1);
+
+	gen5_rps_set(rps, rps->cur_freq);
+
+	rps->ips.last_count1 = intel_uncore_read(uncore, DMIEC);
+	rps->ips.last_count1 += intel_uncore_read(uncore, DDREC);
+	rps->ips.last_count1 += intel_uncore_read(uncore, CSIEC);
+	rps->ips.last_time1 = jiffies_to_msecs(jiffies);
+
+	rps->ips.last_count2 = intel_uncore_read(uncore, GFXEC);
+	rps->ips.last_time2 = ktime_get_raw_ns();
+
+	spin_unlock_irq(&mchdev_lock);
+
+	rps->ips.corr = init_emon(uncore);
+
+	return true;
+}
+
+static void gen5_rps_disable(struct intel_rps *rps)
+{
+	struct intel_uncore *uncore = rps_to_uncore(rps);
+	u16 rgvswctl;
+
+	spin_lock_irq(&mchdev_lock);
+
+	rgvswctl = intel_uncore_read16(uncore, MEMSWCTL);
+
+	/* Ack interrupts, disable EFC interrupt */
+	intel_uncore_write(uncore, MEMINTREN,
+			   intel_uncore_read(uncore, MEMINTREN) &
+			   ~MEMINT_EVAL_CHG_EN);
+	intel_uncore_write(uncore, MEMINTRSTS, MEMINT_EVAL_CHG);
+	intel_uncore_write(uncore, DEIER,
+			   intel_uncore_read(uncore, DEIER) & ~DE_PCU_EVENT);
+	intel_uncore_write(uncore, DEIIR, DE_PCU_EVENT);
+	intel_uncore_write(uncore, DEIMR,
+			   intel_uncore_read(uncore, DEIMR) | DE_PCU_EVENT);
+
+	/* Go back to the starting frequency */
+	gen5_rps_set(rps, rps->idle_freq);
+	mdelay(1);
+	rgvswctl |= MEMCTL_CMD_STS;
+	intel_uncore_write(uncore, MEMSWCTL, rgvswctl);
+	mdelay(1);
+
+	spin_unlock_irq(&mchdev_lock);
+}
+
+static u32 rps_limits(struct intel_rps *rps, u8 val)
+{
+	u32 limits;
+
+	/*
+	 * Only set the down limit when we've reached the lowest level to avoid
+	 * getting more interrupts, otherwise leave this clear. This prevents a
+	 * race in the hw when coming out of rc6: There's a tiny window where
+	 * the hw runs at the minimal clock before selecting the desired
+	 * frequency, if the down threshold expires in that window we will not
+	 * receive a down interrupt.
+	 */
+	if (INTEL_GEN(rps_to_i915(rps)) >= 9) {
+		limits = rps->max_freq_softlimit << 23;
+		if (val <= rps->min_freq_softlimit)
+			limits |= rps->min_freq_softlimit << 14;
+	} else {
+		limits = rps->max_freq_softlimit << 24;
+		if (val <= rps->min_freq_softlimit)
+			limits |= rps->min_freq_softlimit << 16;
+	}
+
+	return limits;
+}
+
+static void rps_set_power(struct intel_rps *rps, int new_power)
+{
+	struct intel_uncore *uncore = rps_to_uncore(rps);
+	struct drm_i915_private *i915 = rps_to_i915(rps);
+	u32 threshold_up = 0, threshold_down = 0; /* in % */
+	u32 ei_up = 0, ei_down = 0;
+
+	lockdep_assert_held(&rps->power.mutex);
+
+	if (new_power == rps->power.mode)
+		return;
+
+	/* Note the units here are not exactly 1us, but 1280ns. */
+	switch (new_power) {
+	case LOW_POWER:
+		/* Upclock if more than 95% busy over 16ms */
+		ei_up = 16000;
+		threshold_up = 95;
+
+		/* Downclock if less than 85% busy over 32ms */
+		ei_down = 32000;
+		threshold_down = 85;
+		break;
+
+	case BETWEEN:
+		/* Upclock if more than 90% busy over 13ms */
+		ei_up = 13000;
+		threshold_up = 90;
+
+		/* Downclock if less than 75% busy over 32ms */
+		ei_down = 32000;
+		threshold_down = 75;
+		break;
+
+	case HIGH_POWER:
+		/* Upclock if more than 85% busy over 10ms */
+		ei_up = 10000;
+		threshold_up = 85;
+
+		/* Downclock if less than 60% busy over 32ms */
+		ei_down = 32000;
+		threshold_down = 60;
+		break;
+	}
+
+	/* When byt can survive without system hang with dynamic
+	 * sw freq adjustments, this restriction can be lifted.
+	 */
+	if (IS_VALLEYVIEW(i915))
+		goto skip_hw_write;
+
+	intel_uncore_write(uncore, GEN6_RP_UP_EI,
+			   GT_INTERVAL_FROM_US(i915, ei_up));
+	intel_uncore_write(uncore, GEN6_RP_UP_THRESHOLD,
+			   GT_INTERVAL_FROM_US(i915,
+					       ei_up * threshold_up / 100));
+
+	intel_uncore_write(uncore, GEN6_RP_DOWN_EI,
+			   GT_INTERVAL_FROM_US(i915, ei_down));
+	intel_uncore_write(uncore, GEN6_RP_DOWN_THRESHOLD,
+			   GT_INTERVAL_FROM_US(i915,
+					       ei_down * threshold_down / 100));
+
+	intel_uncore_write(uncore, GEN6_RP_CONTROL,
+			   (INTEL_GEN(i915) > 9 ? 0 : GEN6_RP_MEDIA_TURBO) |
+			   GEN6_RP_MEDIA_HW_NORMAL_MODE |
+			   GEN6_RP_MEDIA_IS_GFX |
+			   GEN6_RP_ENABLE |
+			   GEN6_RP_UP_BUSY_AVG |
+			   GEN6_RP_DOWN_IDLE_AVG);
+
+skip_hw_write:
+	rps->power.mode = new_power;
+	rps->power.up_threshold = threshold_up;
+	rps->power.down_threshold = threshold_down;
+}
+
+static void gen6_rps_set_thresholds(struct intel_rps *rps, u8 val)
+{
+	int new_power;
+
+	new_power = rps->power.mode;
+	switch (rps->power.mode) {
+	case LOW_POWER:
+		if (val > rps->efficient_freq + 1 &&
+		    val > rps->cur_freq)
+			new_power = BETWEEN;
+		break;
+
+	case BETWEEN:
+		if (val <= rps->efficient_freq &&
+		    val < rps->cur_freq)
+			new_power = LOW_POWER;
+		else if (val >= rps->rp0_freq &&
+			 val > rps->cur_freq)
+			new_power = HIGH_POWER;
+		break;
+
+	case HIGH_POWER:
+		if (val < (rps->rp1_freq + rps->rp0_freq) >> 1 &&
+		    val < rps->cur_freq)
+			new_power = BETWEEN;
+		break;
+	}
+	/* Max/min bins are special */
+	if (val <= rps->min_freq_softlimit)
+		new_power = LOW_POWER;
+	if (val >= rps->max_freq_softlimit)
+		new_power = HIGH_POWER;
+
+	mutex_lock(&rps->power.mutex);
+	if (rps->power.interactive)
+		new_power = HIGH_POWER;
+	rps_set_power(rps, new_power);
+	mutex_unlock(&rps->power.mutex);
+}
+
+void intel_rps_mark_interactive(struct intel_rps *rps, bool interactive)
+{
+	mutex_lock(&rps->power.mutex);
+	if (interactive) {
+		if (!rps->power.interactive++ && rps->active)
+			rps_set_power(rps, HIGH_POWER);
+	} else {
+		GEM_BUG_ON(!rps->power.interactive);
+		rps->power.interactive--;
+	}
+	mutex_unlock(&rps->power.mutex);
+}
+
+static int gen6_rps_set(struct intel_rps *rps, u8 val)
+{
+	struct intel_uncore *uncore = rps_to_uncore(rps);
+	struct drm_i915_private *i915 = rps_to_i915(rps);
+	u32 swreq;
+
+	if (INTEL_GEN(i915) >= 9)
+		swreq = GEN9_FREQUENCY(val);
+	else if (IS_HASWELL(i915) || IS_BROADWELL(i915))
+		swreq = HSW_FREQUENCY(val);
+	else
+		swreq = (GEN6_FREQUENCY(val) |
+			 GEN6_OFFSET(0) |
+			 GEN6_AGGRESSIVE_TURBO);
+	intel_uncore_write(uncore, GEN6_RPNSWREQ, swreq);
+
+	return 0;
+}
+
+static int vlv_rps_set(struct intel_rps *rps, u8 val)
+{
+	struct drm_i915_private *i915 = rps_to_i915(rps);
+	int err;
+
+	vlv_punit_get(i915);
+	err = vlv_punit_write(i915, PUNIT_REG_GPU_FREQ_REQ, val);
+	vlv_punit_put(i915);
+
+	return err;
+}
+
+static int rps_set(struct intel_rps *rps, u8 val)
+{
+	struct drm_i915_private *i915 = rps_to_i915(rps);
+	int err;
+
+	if (INTEL_GEN(i915) < 6)
+		return 0;
+
+	if (val == rps->last_freq)
+		return 0;
+
+	if (IS_VALLEYVIEW(i915) || IS_CHERRYVIEW(i915))
+		err = vlv_rps_set(rps, val);
+	else
+		err = gen6_rps_set(rps, val);
+	if (err)
+		return err;
+
+	gen6_rps_set_thresholds(rps, val);
+	rps->last_freq = val;
+
+	return 0;
+}
+
+void intel_rps_unpark(struct intel_rps *rps)
+{
+	u8 freq;
+
+	if (!rps->enabled)
+		return;
+
+	/*
+	 * Use the user's desired frequency as a guide, but for better
+	 * performance, jump directly to RPe as our starting frequency.
+	 */
+	mutex_lock(&rps->lock);
+	rps->active = true;
+	freq = max(rps->cur_freq, rps->efficient_freq),
+	freq = clamp(freq, rps->min_freq_softlimit, rps->max_freq_softlimit);
+	intel_rps_set(rps, freq);
+	rps->last_adj = 0;
+	mutex_unlock(&rps->lock);
+
+	if (INTEL_GEN(rps_to_i915(rps)) >= 6)
+		rps_enable_interrupts(rps);
+
+	if (IS_GEN(rps_to_i915(rps), 5))
+		gen5_rps_update(rps);
+}
+
+void intel_rps_park(struct intel_rps *rps)
+{
+	struct drm_i915_private *i915 = rps_to_i915(rps);
+
+	if (!rps->enabled)
+		return;
+
+	if (INTEL_GEN(i915) >= 6)
+		rps_disable_interrupts(rps);
+
+	rps->active = false;
+	if (rps->last_freq <= rps->idle_freq)
+		return;
+
+	/*
+	 * The punit delays the write of the frequency and voltage until it
+	 * determines the GPU is awake. During normal usage we don't want to
+	 * waste power changing the frequency if the GPU is sleeping (rc6).
+	 * However, the GPU and driver is now idle and we do not want to delay
+	 * switching to minimum voltage (reducing power whilst idle) as we do
+	 * not expect to be woken in the near future and so must flush the
+	 * change by waking the device.
+	 *
+	 * We choose to take the media powerwell (either would do to trick the
+	 * punit into committing the voltage change) as that takes a lot less
+	 * power than the render powerwell.
+	 */
+	intel_uncore_forcewake_get(rps_to_uncore(rps), FORCEWAKE_MEDIA);
+	rps_set(rps, rps->idle_freq);
+	intel_uncore_forcewake_put(rps_to_uncore(rps), FORCEWAKE_MEDIA);
+}
+
+void intel_rps_boost(struct i915_request *rq)
+{
+	struct intel_rps *rps = &rq->engine->gt->rps;
+	unsigned long flags;
+
+	if (i915_request_signaled(rq) || !rps->active)
+		return;
+
+	/* Serializes with i915_request_retire() */
+	spin_lock_irqsave(&rq->lock, flags);
+	if (!i915_request_has_waitboost(rq) &&
+	    !dma_fence_is_signaled_locked(&rq->fence)) {
+		rq->flags |= I915_REQUEST_WAITBOOST;
+
+		if (!atomic_fetch_inc(&rps->num_waiters) &&
+		    READ_ONCE(rps->cur_freq) < rps->boost_freq)
+			schedule_work(&rps->work);
+
+		atomic_inc(&rps->boosts);
+	}
+	spin_unlock_irqrestore(&rq->lock, flags);
+}
+
+int intel_rps_set(struct intel_rps *rps, u8 val)
+{
+	int err = 0;
+
+	lockdep_assert_held(&rps->lock);
+	GEM_BUG_ON(val > rps->max_freq);
+	GEM_BUG_ON(val < rps->min_freq);
+
+	if (rps->active) {
+		err = rps_set(rps, val);
+
+		/*
+		 * Make sure we continue to get interrupts
+		 * until we hit the minimum or maximum frequencies.
+		 */
+		if (INTEL_GEN(rps_to_i915(rps)) >= 6) {
+			struct intel_uncore *uncore = rps_to_uncore(rps);
+
+			intel_uncore_write(uncore, GEN6_RP_INTERRUPT_LIMITS,
+					   rps_limits(rps, val));
+
+			intel_uncore_write(uncore, GEN6_PMINTRMSK,
+					   rps_pm_mask(rps, val));
+		}
+	}
+
+	if (err == 0)
+		rps->cur_freq = val;
+
+	return err;
+}
+
+static void gen6_rps_init(struct intel_rps *rps)
+{
+	struct drm_i915_private *i915 = rps_to_i915(rps);
+	struct intel_uncore *uncore = rps_to_uncore(rps);
+
+	/* All of these values are in units of 50MHz */
+
+	/* static values from HW: RP0 > RP1 > RPn (min_freq) */
+	if (IS_GEN9_LP(i915)) {
+		u32 rp_state_cap = intel_uncore_read(uncore, BXT_RP_STATE_CAP);
+
+		rps->rp0_freq = (rp_state_cap >> 16) & 0xff;
+		rps->rp1_freq = (rp_state_cap >>  8) & 0xff;
+		rps->min_freq = (rp_state_cap >>  0) & 0xff;
+	} else {
+		u32 rp_state_cap = intel_uncore_read(uncore, GEN6_RP_STATE_CAP);
+
+		rps->rp0_freq = (rp_state_cap >>  0) & 0xff;
+		rps->rp1_freq = (rp_state_cap >>  8) & 0xff;
+		rps->min_freq = (rp_state_cap >> 16) & 0xff;
+	}
+
+	/* hw_max = RP0 until we check for overclocking */
+	rps->max_freq = rps->rp0_freq;
+
+	rps->efficient_freq = rps->rp1_freq;
+	if (IS_HASWELL(i915) || IS_BROADWELL(i915) ||
+	    IS_GEN9_BC(i915) || INTEL_GEN(i915) >= 10) {
+		u32 ddcc_status = 0;
+
+		if (sandybridge_pcode_read(i915,
+					   HSW_PCODE_DYNAMIC_DUTY_CYCLE_CONTROL,
+					   &ddcc_status, NULL) == 0)
+			rps->efficient_freq =
+				clamp_t(u8,
+					(ddcc_status >> 8) & 0xff,
+					rps->min_freq,
+					rps->max_freq);
+	}
+
+	if (IS_GEN9_BC(i915) || INTEL_GEN(i915) >= 10) {
+		/* Store the frequency values in 16.66 MHZ units, which is
+		 * the natural hardware unit for SKL
+		 */
+		rps->rp0_freq *= GEN9_FREQ_SCALER;
+		rps->rp1_freq *= GEN9_FREQ_SCALER;
+		rps->min_freq *= GEN9_FREQ_SCALER;
+		rps->max_freq *= GEN9_FREQ_SCALER;
+		rps->efficient_freq *= GEN9_FREQ_SCALER;
+	}
+}
+
+static bool rps_reset(struct intel_rps *rps)
+{
+	/* force a reset */
+	rps->power.mode = -1;
+	rps->last_freq = -1;
+
+	if (rps_set(rps, rps->min_freq)) {
+		DRM_ERROR("Failed to reset RPS to initial values\n");
+		return false;
+	}
+
+	rps->cur_freq = rps->min_freq;
+	return true;
+}
+
+/* See the Gen9_GT_PM_Programming_Guide doc for the below */
+static bool gen9_rps_enable(struct intel_rps *rps)
+{
+	struct drm_i915_private *i915 = rps_to_i915(rps);
+	struct intel_uncore *uncore = rps_to_uncore(rps);
+
+	/* Program defaults and thresholds for RPS */
+	if (IS_GEN(i915, 9))
+		intel_uncore_write_fw(uncore, GEN6_RC_VIDEO_FREQ,
+				      GEN9_FREQUENCY(rps->rp1_freq));
+
+	/* 1 second timeout */
+	intel_uncore_write_fw(uncore, GEN6_RP_DOWN_TIMEOUT,
+			      GT_INTERVAL_FROM_US(i915, 1000000));
+
+	intel_uncore_write_fw(uncore, GEN6_RP_IDLE_HYSTERSIS, 0xa);
+
+	return rps_reset(rps);
+}
+
+static bool gen8_rps_enable(struct intel_rps *rps)
+{
+	struct intel_uncore *uncore = rps_to_uncore(rps);
+
+	intel_uncore_write_fw(uncore, GEN6_RC_VIDEO_FREQ,
+			      HSW_FREQUENCY(rps->rp1_freq));
+
+	/* NB: Docs say 1s, and 1000000 - which aren't equivalent */
+	intel_uncore_write_fw(uncore, GEN6_RP_DOWN_TIMEOUT,
+			      100000000 / 128); /* 1 second timeout */
+
+	intel_uncore_write_fw(uncore, GEN6_RP_IDLE_HYSTERSIS, 10);
+
+	return rps_reset(rps);
+}
+
+static bool gen6_rps_enable(struct intel_rps *rps)
+{
+	struct intel_uncore *uncore = rps_to_uncore(rps);
+
+	/* Power down if completely idle for over 50ms */
+	intel_uncore_write_fw(uncore, GEN6_RP_DOWN_TIMEOUT, 50000);
+	intel_uncore_write_fw(uncore, GEN6_RP_IDLE_HYSTERSIS, 10);
+
+	return rps_reset(rps);
+}
+
+static int chv_rps_max_freq(struct intel_rps *rps)
+{
+	struct drm_i915_private *i915 = rps_to_i915(rps);
+	u32 val;
+
+	val = vlv_punit_read(i915, FB_GFX_FMAX_AT_VMAX_FUSE);
+
+	switch (RUNTIME_INFO(i915)->sseu.eu_total) {
+	case 8:
+		/* (2 * 4) config */
+		val >>= FB_GFX_FMAX_AT_VMAX_2SS4EU_FUSE_SHIFT;
+		break;
+	case 12:
+		/* (2 * 6) config */
+		val >>= FB_GFX_FMAX_AT_VMAX_2SS6EU_FUSE_SHIFT;
+		break;
+	case 16:
+		/* (2 * 8) config */
+	default:
+		/* Setting (2 * 8) Min RP0 for any other combination */
+		val >>= FB_GFX_FMAX_AT_VMAX_2SS8EU_FUSE_SHIFT;
+		break;
+	}
+
+	return val & FB_GFX_FREQ_FUSE_MASK;
+}
+
+static int chv_rps_rpe_freq(struct intel_rps *rps)
+{
+	struct drm_i915_private *i915 = rps_to_i915(rps);
+	u32 val;
+
+	val = vlv_punit_read(i915, PUNIT_GPU_DUTYCYCLE_REG);
+	val >>= PUNIT_GPU_DUTYCYCLE_RPE_FREQ_SHIFT;
+
+	return val & PUNIT_GPU_DUTYCYCLE_RPE_FREQ_MASK;
+}
+
+static int chv_rps_guar_freq(struct intel_rps *rps)
+{
+	struct drm_i915_private *i915 = rps_to_i915(rps);
+	u32 val;
+
+	val = vlv_punit_read(i915, FB_GFX_FMAX_AT_VMAX_FUSE);
+
+	return val & FB_GFX_FREQ_FUSE_MASK;
+}
+
+static u32 chv_rps_min_freq(struct intel_rps *rps)
+{
+	struct drm_i915_private *i915 = rps_to_i915(rps);
+	u32 val;
+
+	val = vlv_punit_read(i915, FB_GFX_FMIN_AT_VMIN_FUSE);
+	val >>= FB_GFX_FMIN_AT_VMIN_FUSE_SHIFT;
+
+	return val & FB_GFX_FREQ_FUSE_MASK;
+}
+
+static bool chv_rps_enable(struct intel_rps *rps)
+{
+	struct intel_uncore *uncore = rps_to_uncore(rps);
+	struct drm_i915_private *i915 = rps_to_i915(rps);
+	u32 val;
+
+	/* 1: Program defaults and thresholds for RPS*/
+	intel_uncore_write_fw(uncore, GEN6_RP_DOWN_TIMEOUT, 1000000);
+	intel_uncore_write_fw(uncore, GEN6_RP_UP_THRESHOLD, 59400);
+	intel_uncore_write_fw(uncore, GEN6_RP_DOWN_THRESHOLD, 245000);
+	intel_uncore_write_fw(uncore, GEN6_RP_UP_EI, 66000);
+	intel_uncore_write_fw(uncore, GEN6_RP_DOWN_EI, 350000);
+
+	intel_uncore_write_fw(uncore, GEN6_RP_IDLE_HYSTERSIS, 10);
+
+	/* 2: Enable RPS */
+	intel_uncore_write_fw(uncore, GEN6_RP_CONTROL,
+			      GEN6_RP_MEDIA_HW_NORMAL_MODE |
+			      GEN6_RP_MEDIA_IS_GFX |
+			      GEN6_RP_ENABLE |
+			      GEN6_RP_UP_BUSY_AVG |
+			      GEN6_RP_DOWN_IDLE_AVG);
+
+	/* Setting Fixed Bias */
+	vlv_punit_get(i915);
+
+	val = VLV_OVERRIDE_EN | VLV_SOC_TDP_EN | CHV_BIAS_CPU_50_SOC_50;
+	vlv_punit_write(i915, VLV_TURBO_SOC_OVERRIDE, val);
+
+	val = vlv_punit_read(i915, PUNIT_REG_GPU_FREQ_STS);
+
+	vlv_punit_put(i915);
+
+	/* RPS code assumes GPLL is used */
+	WARN_ONCE((val & GPLLENABLE) == 0, "GPLL not enabled\n");
+
+	DRM_DEBUG_DRIVER("GPLL enabled? %s\n", yesno(val & GPLLENABLE));
+	DRM_DEBUG_DRIVER("GPU status: 0x%08x\n", val);
+
+	return rps_reset(rps);
+}
+
+static int vlv_rps_guar_freq(struct intel_rps *rps)
+{
+	struct drm_i915_private *i915 = rps_to_i915(rps);
+	u32 val, rp1;
+
+	val = vlv_nc_read(i915, IOSF_NC_FB_GFX_FREQ_FUSE);
+
+	rp1 = val & FB_GFX_FGUARANTEED_FREQ_FUSE_MASK;
+	rp1 >>= FB_GFX_FGUARANTEED_FREQ_FUSE_SHIFT;
+
+	return rp1;
+}
+
+static int vlv_rps_max_freq(struct intel_rps *rps)
+{
+	struct drm_i915_private *i915 = rps_to_i915(rps);
+	u32 val, rp0;
+
+	val = vlv_nc_read(i915, IOSF_NC_FB_GFX_FREQ_FUSE);
+
+	rp0 = (val & FB_GFX_MAX_FREQ_FUSE_MASK) >> FB_GFX_MAX_FREQ_FUSE_SHIFT;
+	/* Clamp to max */
+	rp0 = min_t(u32, rp0, 0xea);
+
+	return rp0;
+}
+
+static int vlv_rps_rpe_freq(struct intel_rps *rps)
+{
+	struct drm_i915_private *i915 = rps_to_i915(rps);
+	u32 val, rpe;
+
+	val = vlv_nc_read(i915, IOSF_NC_FB_GFX_FMAX_FUSE_LO);
+	rpe = (val & FB_FMAX_VMIN_FREQ_LO_MASK) >> FB_FMAX_VMIN_FREQ_LO_SHIFT;
+	val = vlv_nc_read(i915, IOSF_NC_FB_GFX_FMAX_FUSE_HI);
+	rpe |= (val & FB_FMAX_VMIN_FREQ_HI_MASK) << 5;
+
+	return rpe;
+}
+
+static int vlv_rps_min_freq(struct intel_rps *rps)
+{
+	struct drm_i915_private *i915 = rps_to_i915(rps);
+	u32 val;
+
+	val = vlv_punit_read(i915, PUNIT_REG_GPU_LFM) & 0xff;
+	/*
+	 * According to the BYT Punit GPU turbo HAS 1.1.6.3 the minimum value
+	 * for the minimum frequency in GPLL mode is 0xc1. Contrary to this on
+	 * a BYT-M B0 the above register contains 0xbf. Moreover when setting
+	 * a frequency Punit will not allow values below 0xc0. Clamp it 0xc0
+	 * to make sure it matches what Punit accepts.
+	 */
+	return max_t(u32, val, 0xc0);
+}
+
+static bool vlv_rps_enable(struct intel_rps *rps)
+{
+	struct intel_uncore *uncore = rps_to_uncore(rps);
+	struct drm_i915_private *i915 = rps_to_i915(rps);
+	u32 val;
+
+	intel_uncore_write_fw(uncore, GEN6_RP_DOWN_TIMEOUT, 1000000);
+	intel_uncore_write_fw(uncore, GEN6_RP_UP_THRESHOLD, 59400);
+	intel_uncore_write_fw(uncore, GEN6_RP_DOWN_THRESHOLD, 245000);
+	intel_uncore_write_fw(uncore, GEN6_RP_UP_EI, 66000);
+	intel_uncore_write_fw(uncore, GEN6_RP_DOWN_EI, 350000);
+
+	intel_uncore_write_fw(uncore, GEN6_RP_IDLE_HYSTERSIS, 10);
+
+	intel_uncore_write_fw(uncore, GEN6_RP_CONTROL,
+			      GEN6_RP_MEDIA_TURBO |
+			      GEN6_RP_MEDIA_HW_NORMAL_MODE |
+			      GEN6_RP_MEDIA_IS_GFX |
+			      GEN6_RP_ENABLE |
+			      GEN6_RP_UP_BUSY_AVG |
+			      GEN6_RP_DOWN_IDLE_CONT);
+
+	vlv_punit_get(i915);
+
+	/* Setting Fixed Bias */
+	val = VLV_OVERRIDE_EN | VLV_SOC_TDP_EN | VLV_BIAS_CPU_125_SOC_875;
+	vlv_punit_write(i915, VLV_TURBO_SOC_OVERRIDE, val);
+
+	val = vlv_punit_read(i915, PUNIT_REG_GPU_FREQ_STS);
+
+	vlv_punit_put(i915);
+
+	/* RPS code assumes GPLL is used */
+	WARN_ONCE((val & GPLLENABLE) == 0, "GPLL not enabled\n");
+
+	DRM_DEBUG_DRIVER("GPLL enabled? %s\n", yesno(val & GPLLENABLE));
+	DRM_DEBUG_DRIVER("GPU status: 0x%08x\n", val);
+
+	return rps_reset(rps);
+}
+
+static unsigned long __ips_gfx_val(struct intel_ips *ips)
+{
+	struct intel_rps *rps = container_of(ips, typeof(*rps), ips);
+	struct intel_uncore *uncore = rps_to_uncore(rps);
+	unsigned long t, corr, state1, corr2, state2;
+	u32 pxvid, ext_v;
+
+	lockdep_assert_held(&mchdev_lock);
+
+	pxvid = intel_uncore_read(uncore, PXVFREQ(rps->cur_freq));
+	pxvid = (pxvid >> 24) & 0x7f;
+	ext_v = pvid_to_extvid(rps_to_i915(rps), pxvid);
+
+	state1 = ext_v;
+
+	/* Revel in the empirically derived constants */
+
+	/* Correction factor in 1/100000 units */
+	t = ips_mch_val(uncore);
+	if (t > 80)
+		corr = t * 2349 + 135940;
+	else if (t >= 50)
+		corr = t * 964 + 29317;
+	else /* < 50 */
+		corr = t * 301 + 1004;
+
+	corr = corr * 150142 * state1 / 10000 - 78642;
+	corr /= 100000;
+	corr2 = corr * ips->corr;
+
+	state2 = corr2 * state1 / 10000;
+	state2 /= 100; /* convert to mW */
+
+	__gen5_ips_update(ips);
+
+	return ips->gfx_power + state2;
+}
+
+void intel_rps_enable(struct intel_rps *rps)
+{
+	struct drm_i915_private *i915 = rps_to_i915(rps);
+	struct intel_uncore *uncore = rps_to_uncore(rps);
+
+	intel_uncore_forcewake_get(uncore, FORCEWAKE_ALL);
+	if (IS_CHERRYVIEW(i915))
+		rps->enabled = chv_rps_enable(rps);
+	else if (IS_VALLEYVIEW(i915))
+		rps->enabled = vlv_rps_enable(rps);
+	else if (INTEL_GEN(i915) >= 9)
+		rps->enabled = gen9_rps_enable(rps);
+	else if (INTEL_GEN(i915) >= 8)
+		rps->enabled = gen8_rps_enable(rps);
+	else if (INTEL_GEN(i915) >= 6)
+		rps->enabled = gen6_rps_enable(rps);
+	else if (IS_IRONLAKE_M(i915))
+		rps->enabled = gen5_rps_enable(rps);
+	intel_uncore_forcewake_put(uncore, FORCEWAKE_ALL);
+	if (!rps->enabled)
+		return;
+
+	WARN_ON(rps->max_freq < rps->min_freq);
+	WARN_ON(rps->idle_freq > rps->max_freq);
+
+	WARN_ON(rps->efficient_freq < rps->min_freq);
+	WARN_ON(rps->efficient_freq > rps->max_freq);
+}
+
+static void gen6_rps_disable(struct intel_rps *rps)
+{
+	intel_uncore_write(rps_to_uncore(rps), GEN6_RP_CONTROL, 0);
+}
+
+void intel_rps_disable(struct intel_rps *rps)
+{
+	struct drm_i915_private *i915 = rps_to_i915(rps);
+
+	rps->enabled = false;
+
+	if (INTEL_GEN(i915) >= 6)
+		gen6_rps_disable(rps);
+	else if (IS_IRONLAKE_M(i915))
+		gen5_rps_disable(rps);
+}
+
+static int byt_gpu_freq(struct intel_rps *rps, int val)
+{
+	/*
+	 * N = val - 0xb7
+	 * Slow = Fast = GPLL ref * N
+	 */
+	return DIV_ROUND_CLOSEST(rps->gpll_ref_freq * (val - 0xb7), 1000);
+}
+
+static int byt_freq_opcode(struct intel_rps *rps, int val)
+{
+	return DIV_ROUND_CLOSEST(1000 * val, rps->gpll_ref_freq) + 0xb7;
+}
+
+static int chv_gpu_freq(struct intel_rps *rps, int val)
+{
+	/*
+	 * N = val / 2
+	 * CU (slow) = CU2x (fast) / 2 = GPLL ref * N / 2
+	 */
+	return DIV_ROUND_CLOSEST(rps->gpll_ref_freq * val, 2 * 2 * 1000);
+}
+
+static int chv_freq_opcode(struct intel_rps *rps, int val)
+{
+	/* CHV needs even values */
+	return DIV_ROUND_CLOSEST(2 * 1000 * val, rps->gpll_ref_freq) * 2;
+}
+
+int intel_gpu_freq(struct intel_rps *rps, int val)
+{
+	struct drm_i915_private *i915 = rps_to_i915(rps);
+
+	if (INTEL_GEN(i915) >= 9)
+		return DIV_ROUND_CLOSEST(val * GT_FREQUENCY_MULTIPLIER,
+					 GEN9_FREQ_SCALER);
+	else if (IS_CHERRYVIEW(i915))
+		return chv_gpu_freq(rps, val);
+	else if (IS_VALLEYVIEW(i915))
+		return byt_gpu_freq(rps, val);
+	else
+		return val * GT_FREQUENCY_MULTIPLIER;
+}
+
+int intel_freq_opcode(struct intel_rps *rps, int val)
+{
+	struct drm_i915_private *i915 = rps_to_i915(rps);
+
+	if (INTEL_GEN(i915) >= 9)
+		return DIV_ROUND_CLOSEST(val * GEN9_FREQ_SCALER,
+					 GT_FREQUENCY_MULTIPLIER);
+	else if (IS_CHERRYVIEW(i915))
+		return chv_freq_opcode(rps, val);
+	else if (IS_VALLEYVIEW(i915))
+		return byt_freq_opcode(rps, val);
+	else
+		return DIV_ROUND_CLOSEST(val, GT_FREQUENCY_MULTIPLIER);
+}
+
+static void vlv_init_gpll_ref_freq(struct intel_rps *rps)
+{
+	struct drm_i915_private *i915 = rps_to_i915(rps);
+
+	rps->gpll_ref_freq =
+		vlv_get_cck_clock(i915, "GPLL ref",
+				  CCK_GPLL_CLOCK_CONTROL,
+				  i915->czclk_freq);
+
+	DRM_DEBUG_DRIVER("GPLL reference freq: %d kHz\n", rps->gpll_ref_freq);
+}
+
+static void vlv_rps_init(struct intel_rps *rps)
+{
+	struct drm_i915_private *i915 = rps_to_i915(rps);
+	u32 val;
+
+	vlv_iosf_sb_get(i915,
+			BIT(VLV_IOSF_SB_PUNIT) |
+			BIT(VLV_IOSF_SB_NC) |
+			BIT(VLV_IOSF_SB_CCK));
+
+	vlv_init_gpll_ref_freq(rps);
+
+	val = vlv_punit_read(i915, PUNIT_REG_GPU_FREQ_STS);
+	switch ((val >> 6) & 3) {
+	case 0:
+	case 1:
+		i915->mem_freq = 800;
+		break;
+	case 2:
+		i915->mem_freq = 1066;
+		break;
+	case 3:
+		i915->mem_freq = 1333;
+		break;
+	}
+	DRM_DEBUG_DRIVER("DDR speed: %d MHz\n", i915->mem_freq);
+
+	rps->max_freq = vlv_rps_max_freq(rps);
+	rps->rp0_freq = rps->max_freq;
+	DRM_DEBUG_DRIVER("max GPU freq: %d MHz (%u)\n",
+			 intel_gpu_freq(rps, rps->max_freq),
+			 rps->max_freq);
+
+	rps->efficient_freq = vlv_rps_rpe_freq(rps);
+	DRM_DEBUG_DRIVER("RPe GPU freq: %d MHz (%u)\n",
+			 intel_gpu_freq(rps, rps->efficient_freq),
+			 rps->efficient_freq);
+
+	rps->rp1_freq = vlv_rps_guar_freq(rps);
+	DRM_DEBUG_DRIVER("RP1(Guar Freq) GPU freq: %d MHz (%u)\n",
+			 intel_gpu_freq(rps, rps->rp1_freq),
+			 rps->rp1_freq);
+
+	rps->min_freq = vlv_rps_min_freq(rps);
+	DRM_DEBUG_DRIVER("min GPU freq: %d MHz (%u)\n",
+			 intel_gpu_freq(rps, rps->min_freq),
+			 rps->min_freq);
+
+	vlv_iosf_sb_put(i915,
+			BIT(VLV_IOSF_SB_PUNIT) |
+			BIT(VLV_IOSF_SB_NC) |
+			BIT(VLV_IOSF_SB_CCK));
+}
+
+static void chv_rps_init(struct intel_rps *rps)
+{
+	struct drm_i915_private *i915 = rps_to_i915(rps);
+	u32 val;
+
+	vlv_iosf_sb_get(i915,
+			BIT(VLV_IOSF_SB_PUNIT) |
+			BIT(VLV_IOSF_SB_NC) |
+			BIT(VLV_IOSF_SB_CCK));
+
+	vlv_init_gpll_ref_freq(rps);
+
+	val = vlv_cck_read(i915, CCK_FUSE_REG);
+
+	switch ((val >> 2) & 0x7) {
+	case 3:
+		i915->mem_freq = 2000;
+		break;
+	default:
+		i915->mem_freq = 1600;
+		break;
+	}
+	DRM_DEBUG_DRIVER("DDR speed: %d MHz\n", i915->mem_freq);
+
+	rps->max_freq = chv_rps_max_freq(rps);
+	rps->rp0_freq = rps->max_freq;
+	DRM_DEBUG_DRIVER("max GPU freq: %d MHz (%u)\n",
+			 intel_gpu_freq(rps, rps->max_freq),
+			 rps->max_freq);
+
+	rps->efficient_freq = chv_rps_rpe_freq(rps);
+	DRM_DEBUG_DRIVER("RPe GPU freq: %d MHz (%u)\n",
+			 intel_gpu_freq(rps, rps->efficient_freq),
+			 rps->efficient_freq);
+
+	rps->rp1_freq = chv_rps_guar_freq(rps);
+	DRM_DEBUG_DRIVER("RP1(Guar) GPU freq: %d MHz (%u)\n",
+			 intel_gpu_freq(rps, rps->rp1_freq),
+			 rps->rp1_freq);
+
+	rps->min_freq = chv_rps_min_freq(rps);
+	DRM_DEBUG_DRIVER("min GPU freq: %d MHz (%u)\n",
+			 intel_gpu_freq(rps, rps->min_freq),
+			 rps->min_freq);
+
+	vlv_iosf_sb_put(i915,
+			BIT(VLV_IOSF_SB_PUNIT) |
+			BIT(VLV_IOSF_SB_NC) |
+			BIT(VLV_IOSF_SB_CCK));
+
+	WARN_ONCE((rps->max_freq | rps->efficient_freq | rps->rp1_freq |
+		   rps->min_freq) & 1,
+		  "Odd GPU freq values\n");
+}
+
+static void vlv_c0_read(struct intel_uncore *uncore, struct intel_rps_ei *ei)
+{
+	ei->ktime = ktime_get_raw();
+	ei->render_c0 = intel_uncore_read(uncore, VLV_RENDER_C0_COUNT);
+	ei->media_c0 = intel_uncore_read(uncore, VLV_MEDIA_C0_COUNT);
+}
+
+static u32 vlv_wa_c0_ei(struct intel_rps *rps, u32 pm_iir)
+{
+	struct intel_uncore *uncore = rps_to_uncore(rps);
+	const struct intel_rps_ei *prev = &rps->ei;
+	struct intel_rps_ei now;
+	u32 events = 0;
+
+	if ((pm_iir & GEN6_PM_RP_UP_EI_EXPIRED) == 0)
+		return 0;
+
+	vlv_c0_read(uncore, &now);
+
+	if (prev->ktime) {
+		u64 time, c0;
+		u32 render, media;
+
+		time = ktime_us_delta(now.ktime, prev->ktime);
+
+		time *= rps_to_i915(rps)->czclk_freq;
+
+		/* Workload can be split between render + media,
+		 * e.g. SwapBuffers being blitted in X after being rendered in
+		 * mesa. To account for this we need to combine both engines
+		 * into our activity counter.
+		 */
+		render = now.render_c0 - prev->render_c0;
+		media = now.media_c0 - prev->media_c0;
+		c0 = max(render, media);
+		c0 *= 1000 * 100 << 8; /* to usecs and scale to threshold% */
+
+		if (c0 > time * rps->power.up_threshold)
+			events = GEN6_PM_RP_UP_THRESHOLD;
+		else if (c0 < time * rps->power.down_threshold)
+			events = GEN6_PM_RP_DOWN_THRESHOLD;
+	}
+
+	rps->ei = now;
+	return events;
+}
+
+static void rps_work(struct work_struct *work)
+{
+	struct intel_rps *rps = container_of(work, typeof(*rps), work);
+	struct intel_gt *gt = rps_to_gt(rps);
+	bool client_boost = false;
+	int new_freq, adj, min, max;
+	u32 pm_iir = 0;
+
+	spin_lock_irq(&gt->irq_lock);
+	pm_iir = fetch_and_zero(&rps->pm_iir);
+	client_boost = atomic_read(&rps->num_waiters);
+	spin_unlock_irq(&gt->irq_lock);
+
+	/* Make sure we didn't queue anything we're not going to process. */
+	if ((pm_iir & rps->pm_events) == 0 && !client_boost)
+		goto out;
+
+	mutex_lock(&rps->lock);
+
+	pm_iir |= vlv_wa_c0_ei(rps, pm_iir);
+
+	adj = rps->last_adj;
+	new_freq = rps->cur_freq;
+	min = rps->min_freq_softlimit;
+	max = rps->max_freq_softlimit;
+	if (client_boost)
+		max = rps->max_freq;
+	if (client_boost && new_freq < rps->boost_freq) {
+		new_freq = rps->boost_freq;
+		adj = 0;
+	} else if (pm_iir & GEN6_PM_RP_UP_THRESHOLD) {
+		if (adj > 0)
+			adj *= 2;
+		else /* CHV needs even encode values */
+			adj = IS_CHERRYVIEW(gt->i915) ? 2 : 1;
+
+		if (new_freq >= rps->max_freq_softlimit)
+			adj = 0;
+	} else if (client_boost) {
+		adj = 0;
+	} else if (pm_iir & GEN6_PM_RP_DOWN_TIMEOUT) {
+		if (rps->cur_freq > rps->efficient_freq)
+			new_freq = rps->efficient_freq;
+		else if (rps->cur_freq > rps->min_freq_softlimit)
+			new_freq = rps->min_freq_softlimit;
+		adj = 0;
+	} else if (pm_iir & GEN6_PM_RP_DOWN_THRESHOLD) {
+		if (adj < 0)
+			adj *= 2;
+		else /* CHV needs even encode values */
+			adj = IS_CHERRYVIEW(gt->i915) ? -2 : -1;
+
+		if (new_freq <= rps->min_freq_softlimit)
+			adj = 0;
+	} else { /* unknown event */
+		adj = 0;
+	}
+
+	rps->last_adj = adj;
+
+	/*
+	 * Limit deboosting and boosting to keep ourselves at the extremes
+	 * when in the respective power modes (i.e. slowly decrease frequencies
+	 * while in the HIGH_POWER zone and slowly increase frequencies while
+	 * in the LOW_POWER zone). On idle, we will hit the timeout and drop
+	 * to the next level quickly, and conversely if busy we expect to
+	 * hit a waitboost and rapidly switch into max power.
+	 */
+	if ((adj < 0 && rps->power.mode == HIGH_POWER) ||
+	    (adj > 0 && rps->power.mode == LOW_POWER))
+		rps->last_adj = 0;
+
+	/* sysfs frequency interfaces may have snuck in while servicing the
+	 * interrupt
+	 */
+	new_freq += adj;
+	new_freq = clamp_t(int, new_freq, min, max);
+
+	if (intel_rps_set(rps, new_freq)) {
+		DRM_DEBUG_DRIVER("Failed to set new GPU frequency\n");
+		rps->last_adj = 0;
+	}
+
+	mutex_unlock(&rps->lock);
+
+out:
+	spin_lock_irq(&gt->irq_lock);
+	gen6_gt_pm_unmask_irq(gt, rps->pm_events);
+	spin_unlock_irq(&gt->irq_lock);
+}
+
+void gen11_rps_irq_handler(struct intel_rps *rps, u32 pm_iir)
+{
+	struct intel_gt *gt = rps_to_gt(rps);
+	const u32 events = rps->pm_events & pm_iir;
+
+	lockdep_assert_held(&gt->irq_lock);
+
+	if (unlikely(!events))
+		return;
+
+	gen6_gt_pm_mask_irq(gt, events);
+
+	rps->pm_iir |= events;
+	schedule_work(&rps->work);
+}
+
+void gen6_rps_irq_handler(struct intel_rps *rps, u32 pm_iir)
+{
+	struct intel_gt *gt = rps_to_gt(rps);
+
+	if (pm_iir & rps->pm_events) {
+		spin_lock(&gt->irq_lock);
+		gen6_gt_pm_mask_irq(gt, pm_iir & rps->pm_events);
+		rps->pm_iir |= pm_iir & rps->pm_events;
+		schedule_work(&rps->work);
+		spin_unlock(&gt->irq_lock);
+	}
+
+	if (INTEL_GEN(gt->i915) >= 8)
+		return;
+
+	if (pm_iir & PM_VEBOX_USER_INTERRUPT)
+		intel_engine_breadcrumbs_irq(gt->engine[VECS0]);
+
+	if (pm_iir & PM_VEBOX_CS_ERROR_INTERRUPT)
+		DRM_DEBUG("Command parser error, pm_iir 0x%08x\n", pm_iir);
+}
+
+void gen5_rps_irq_handler(struct intel_rps *rps)
+{
+	struct intel_uncore *uncore = rps_to_uncore(rps);
+	u32 busy_up, busy_down, max_avg, min_avg;
+	u8 new_freq;
+
+	spin_lock(&mchdev_lock);
+
+	intel_uncore_write16(uncore,
+			     MEMINTRSTS,
+			     intel_uncore_read(uncore, MEMINTRSTS));
+
+	intel_uncore_write16(uncore, MEMINTRSTS, MEMINT_EVAL_CHG);
+	busy_up = intel_uncore_read(uncore, RCPREVBSYTUPAVG);
+	busy_down = intel_uncore_read(uncore, RCPREVBSYTDNAVG);
+	max_avg = intel_uncore_read(uncore, RCBMAXAVG);
+	min_avg = intel_uncore_read(uncore, RCBMINAVG);
+
+	/* Handle RCS change request from hw */
+	new_freq = rps->cur_freq;
+	if (busy_up > max_avg)
+		new_freq++;
+	else if (busy_down < min_avg)
+		new_freq--;
+	new_freq = clamp(new_freq,
+			 rps->min_freq_softlimit,
+			 rps->max_freq_softlimit);
+
+	if (new_freq != rps->cur_freq && gen5_rps_set(rps, new_freq))
+		rps->cur_freq = new_freq;
+
+	spin_unlock(&mchdev_lock);
+}
+
+void intel_rps_init_early(struct intel_rps *rps)
+{
+	mutex_init(&rps->lock);
+	mutex_init(&rps->power.mutex);
+
+	INIT_WORK(&rps->work, rps_work);
+
+	atomic_set(&rps->num_waiters, 0);
+}
+
+void intel_rps_init(struct intel_rps *rps)
+{
+	struct drm_i915_private *i915 = rps_to_i915(rps);
+
+	if (IS_CHERRYVIEW(i915))
+		chv_rps_init(rps);
+	else if (IS_VALLEYVIEW(i915))
+		vlv_rps_init(rps);
+	else if (INTEL_GEN(i915) >= 6)
+		gen6_rps_init(rps);
+	else if (IS_IRONLAKE_M(i915))
+		gen5_rps_init(rps);
+
+	/* Derive initial user preferences/limits from the hardware limits */
+	rps->max_freq_softlimit = rps->max_freq;
+	rps->min_freq_softlimit = rps->min_freq;
+
+	/* After setting max-softlimit, find the overclock max freq */
+	if (IS_GEN(i915, 6) || IS_IVYBRIDGE(i915) || IS_HASWELL(i915)) {
+		u32 params = 0;
+
+		sandybridge_pcode_read(i915, GEN6_READ_OC_PARAMS,
+				       &params, NULL);
+		if (params & BIT(31)) { /* OC supported */
+			DRM_DEBUG_DRIVER("Overclocking supported, max: %dMHz, overclock: %dMHz\n",
+					 (rps->max_freq & 0xff) * 50,
+					 (params & 0xff) * 50);
+			rps->max_freq = params & 0xff;
+		}
+	}
+
+	/* Finally allow us to boost to max by default */
+	rps->boost_freq = rps->max_freq;
+	rps->idle_freq = rps->min_freq;
+	rps->cur_freq = rps->idle_freq;
+
+	rps->pm_intrmsk_mbz = 0;
+
+	/*
+	 * SNB,IVB,HSW can while VLV,CHV may hard hang on looping batchbuffer
+	 * if GEN6_PM_UP_EI_EXPIRED is masked.
+	 *
+	 * TODO: verify if this can be reproduced on VLV,CHV.
+	 */
+	if (INTEL_GEN(i915) <= 7)
+		rps->pm_intrmsk_mbz |= GEN6_PM_RP_UP_EI_EXPIRED;
+
+	if (INTEL_GEN(i915) >= 8)
+		rps->pm_intrmsk_mbz |= GEN8_PMINTR_DISABLE_REDIRECT_TO_GUC;
+}
+
+u32 intel_get_cagf(struct intel_rps *rps, u32 rpstat)
+{
+	struct drm_i915_private *i915 = rps_to_i915(rps);
+	u32 cagf;
+
+	if (INTEL_GEN(i915) >= 9)
+		cagf = (rpstat & GEN9_CAGF_MASK) >> GEN9_CAGF_SHIFT;
+	else if (IS_HASWELL(i915) || IS_BROADWELL(i915))
+		cagf = (rpstat & HSW_CAGF_MASK) >> HSW_CAGF_SHIFT;
+	else
+		cagf = (rpstat & GEN6_CAGF_MASK) >> GEN6_CAGF_SHIFT;
+
+	return  cagf;
+}
+
+/* External interface for intel_ips.ko */
+
+static struct drm_i915_private __rcu *ips_mchdev;
+
+/**
+ * Tells the intel_ips driver that the i915 driver is now loaded, if
+ * IPS got loaded first.
+ *
+ * This awkward dance is so that neither module has to depend on the
+ * other in order for IPS to do the appropriate communication of
+ * GPU turbo limits to i915.
+ */
+static void
+ips_ping_for_i915_load(void)
+{
+	void (*link)(void);
+
+	link = symbol_get(ips_link_to_i915_driver);
+	if (link) {
+		link();
+		symbol_put(ips_link_to_i915_driver);
+	}
+}
+
+void intel_rps_driver_register(struct intel_rps *rps)
+{
+	struct intel_gt *gt = rps_to_gt(rps);
+
+	/*
+	 * We only register the i915 ips part with intel-ips once everything is
+	 * set up, to avoid intel-ips sneaking in and reading bogus values.
+	 */
+	if (IS_GEN(gt->i915, 5)) {
+		rcu_assign_pointer(ips_mchdev, gt->i915);
+		ips_ping_for_i915_load();
+	}
+}
+
+void intel_rps_driver_unregister(struct intel_rps *rps)
+{
+	rcu_assign_pointer(ips_mchdev, NULL);
+}
+
+static struct drm_i915_private *mchdev_get(void)
+{
+	struct drm_i915_private *i915;
+
+	rcu_read_lock();
+	i915 = rcu_dereference(ips_mchdev);
+	if (!kref_get_unless_zero(&i915->drm.ref))
+		i915 = NULL;
+	rcu_read_unlock();
+
+	return i915;
+}
+
+/**
+ * i915_read_mch_val - return value for IPS use
+ *
+ * Calculate and return a value for the IPS driver to use when deciding whether
+ * we have thermal and power headroom to increase CPU or GPU power budget.
+ */
+unsigned long i915_read_mch_val(void)
+{
+	struct drm_i915_private *i915;
+	unsigned long chipset_val = 0;
+	unsigned long graphics_val = 0;
+	intel_wakeref_t wakeref;
+
+	i915 = mchdev_get();
+	if (!i915)
+		return 0;
+
+	with_intel_runtime_pm(&i915->runtime_pm, wakeref) {
+		struct intel_ips *ips = &i915->gt.rps.ips;
+
+		spin_lock_irq(&mchdev_lock);
+		chipset_val = __ips_chipset_val(ips);
+		graphics_val = __ips_gfx_val(ips);
+		spin_unlock_irq(&mchdev_lock);
+	}
+
+	drm_dev_put(&i915->drm);
+	return chipset_val + graphics_val;
+}
+EXPORT_SYMBOL_GPL(i915_read_mch_val);
+
+/**
+ * i915_gpu_raise - raise GPU frequency limit
+ *
+ * Raise the limit; IPS indicates we have thermal headroom.
+ */
+bool i915_gpu_raise(void)
+{
+	struct drm_i915_private *i915;
+	struct intel_rps *rps;
+
+	i915 = mchdev_get();
+	if (!i915)
+		return false;
+
+	rps = &i915->gt.rps;
+
+	spin_lock_irq(&mchdev_lock);
+	if (rps->max_freq_softlimit < rps->max_freq)
+		rps->max_freq_softlimit++;
+	spin_unlock_irq(&mchdev_lock);
+
+	drm_dev_put(&i915->drm);
+	return true;
+}
+EXPORT_SYMBOL_GPL(i915_gpu_raise);
+
+/**
+ * i915_gpu_lower - lower GPU frequency limit
+ *
+ * IPS indicates we're close to a thermal limit, so throttle back the GPU
+ * frequency maximum.
+ */
+bool i915_gpu_lower(void)
+{
+	struct drm_i915_private *i915;
+	struct intel_rps *rps;
+
+	i915 = mchdev_get();
+	if (!i915)
+		return false;
+
+	rps = &i915->gt.rps;
+
+	spin_lock_irq(&mchdev_lock);
+	if (rps->max_freq_softlimit > rps->min_freq)
+		rps->max_freq_softlimit--;
+	spin_unlock_irq(&mchdev_lock);
+
+	drm_dev_put(&i915->drm);
+	return true;
+}
+EXPORT_SYMBOL_GPL(i915_gpu_lower);
+
+/**
+ * i915_gpu_busy - indicate GPU business to IPS
+ *
+ * Tell the IPS driver whether or not the GPU is busy.
+ */
+bool i915_gpu_busy(void)
+{
+	struct drm_i915_private *i915;
+	bool ret;
+
+	i915 = mchdev_get();
+	if (!i915)
+		return false;
+
+	ret = i915->gt.awake;
+
+	drm_dev_put(&i915->drm);
+	return ret;
+}
+EXPORT_SYMBOL_GPL(i915_gpu_busy);
+
+/**
+ * i915_gpu_turbo_disable - disable graphics turbo
+ *
+ * Disable graphics turbo by resetting the max frequency and setting the
+ * current frequency to the default.
+ */
+bool i915_gpu_turbo_disable(void)
+{
+	struct drm_i915_private *i915;
+	struct intel_rps *rps;
+	bool ret;
+
+	i915 = mchdev_get();
+	if (!i915)
+		return false;
+
+	rps = &i915->gt.rps;
+
+	spin_lock_irq(&mchdev_lock);
+	rps->max_freq_softlimit = rps->min_freq;
+	ret = gen5_rps_set(&i915->gt.rps, rps->min_freq);
+	spin_unlock_irq(&mchdev_lock);
+
+	drm_dev_put(&i915->drm);
+	return ret;
+}
+EXPORT_SYMBOL_GPL(i915_gpu_turbo_disable);
diff --git a/drivers/gpu/drm/i915/gt/intel_rps.h b/drivers/gpu/drm/i915/gt/intel_rps.h
new file mode 100644
index 000000000000..9518c66c9792
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/intel_rps.h
@@ -0,0 +1,38 @@
+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2019 Intel Corporation
+ */
+
+#ifndef INTEL_RPS_H
+#define INTEL_RPS_H
+
+#include "intel_rps_types.h"
+
+struct i915_request;
+
+void intel_rps_init_early(struct intel_rps *rps);
+void intel_rps_init(struct intel_rps *rps);
+
+void intel_rps_driver_register(struct intel_rps *rps);
+void intel_rps_driver_unregister(struct intel_rps *rps);
+
+void intel_rps_enable(struct intel_rps *rps);
+void intel_rps_disable(struct intel_rps *rps);
+
+void intel_rps_park(struct intel_rps *rps);
+void intel_rps_unpark(struct intel_rps *rps);
+void intel_rps_boost(struct i915_request *rq);
+
+int intel_rps_set(struct intel_rps *rps, u8 val);
+void intel_rps_mark_interactive(struct intel_rps *rps, bool interactive);
+
+int intel_gpu_freq(struct intel_rps *rps, int val);
+int intel_freq_opcode(struct intel_rps *rps, int val);
+u32 intel_get_cagf(struct intel_rps *rps, u32 rpstat1);
+
+void gen5_rps_irq_handler(struct intel_rps *rps);
+void gen6_rps_irq_handler(struct intel_rps *rps, u32 pm_iir);
+void gen11_rps_irq_handler(struct intel_rps *rps, u32 pm_iir);
+
+#endif /* INTEL_RPS_H */
diff --git a/drivers/gpu/drm/i915/gt/intel_rps_types.h b/drivers/gpu/drm/i915/gt/intel_rps_types.h
new file mode 100644
index 000000000000..c2e279154bd5
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/intel_rps_types.h
@@ -0,0 +1,93 @@
+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2019 Intel Corporation
+ */
+
+#ifndef INTEL_RPS_TYPES_H
+#define INTEL_RPS_TYPES_H
+
+#include <linux/atomic.h>
+#include <linux/ktime.h>
+#include <linux/mutex.h>
+#include <linux/types.h>
+#include <linux/workqueue.h>
+
+struct intel_ips {
+	u64 last_count1;
+	unsigned long last_time1;
+	unsigned long chipset_power;
+	u64 last_count2;
+	u64 last_time2;
+	unsigned long gfx_power;
+	u8 corr;
+
+	int c, m;
+};
+
+struct intel_rps_ei {
+	ktime_t ktime;
+	u32 render_c0;
+	u32 media_c0;
+};
+
+struct intel_rps {
+	struct mutex lock; /* protects enabling and the worker */
+
+	/*
+	 * work, interrupts_enabled and pm_iir are protected by
+	 * dev_priv->irq_lock
+	 */
+	struct work_struct work;
+	bool enabled;
+	bool active;
+	u32 pm_iir;
+
+	/* PM interrupt bits that should never be masked */
+	u32 pm_intrmsk_mbz;
+	u32 pm_events;
+
+	/* Frequencies are stored in potentially platform dependent multiples.
+	 * In other words, *_freq needs to be multiplied by X to be interesting.
+	 * Soft limits are those which are used for the dynamic reclocking done
+	 * by the driver (raise frequencies under heavy loads, and lower for
+	 * lighter loads). Hard limits are those imposed by the hardware.
+	 *
+	 * A distinction is made for overclocking, which is never enabled by
+	 * default, and is considered to be above the hard limit if it's
+	 * possible at all.
+	 */
+	u8 cur_freq;		/* Current frequency (cached, may not == HW) */
+	u8 last_freq;		/* Last SWREQ frequency */
+	u8 min_freq_softlimit;	/* Minimum frequency permitted by the driver */
+	u8 max_freq_softlimit;	/* Max frequency permitted by the driver */
+	u8 max_freq;		/* Maximum frequency, RP0 if not overclocking */
+	u8 min_freq;		/* AKA RPn. Minimum frequency */
+	u8 boost_freq;		/* Frequency to request when wait boosting */
+	u8 idle_freq;		/* Frequency to request when we are idle */
+	u8 efficient_freq;	/* AKA RPe. Pre-determined balanced frequency */
+	u8 rp1_freq;		/* "less than" RP0 power/freqency */
+	u8 rp0_freq;		/* Non-overclocked max frequency. */
+	u16 gpll_ref_freq;	/* vlv/chv GPLL reference frequency */
+
+	int last_adj;
+
+	struct {
+		struct mutex mutex;
+
+		enum { LOW_POWER, BETWEEN, HIGH_POWER } mode;
+		unsigned int interactive;
+
+		u8 up_threshold; /* Current %busy required to uplock */
+		u8 down_threshold; /* Current %busy required to downclock */
+	} power;
+
+	atomic_t num_waiters;
+	atomic_t boosts;
+
+	/* manual wa residency calculations */
+	struct intel_rps_ei ei;
+	struct intel_ips ips;
+};
+
+#endif /* INTEL_RPS_TYPES_H */
diff --git a/drivers/gpu/drm/i915/gt/intel_sseu.c b/drivers/gpu/drm/i915/gt/intel_sseu.c
index 6bf2d87da109..74f793423231 100644
--- a/drivers/gpu/drm/i915/gt/intel_sseu.c
+++ b/drivers/gpu/drm/i915/gt/intel_sseu.c
@@ -8,6 +8,19 @@
 #include "intel_lrc_reg.h"
 #include "intel_sseu.h"
 
+void intel_sseu_set_info(struct sseu_dev_info *sseu, u8 max_slices,
+			 u8 max_subslices, u8 max_eus_per_subslice)
+{
+	sseu->max_slices = max_slices;
+	sseu->max_subslices = max_subslices;
+	sseu->max_eus_per_subslice = max_eus_per_subslice;
+
+	sseu->ss_stride = GEN_SSEU_STRIDE(sseu->max_subslices);
+	GEM_BUG_ON(sseu->ss_stride > GEN_MAX_SUBSLICE_STRIDE);
+	sseu->eu_stride = GEN_SSEU_STRIDE(sseu->max_eus_per_subslice);
+	GEM_BUG_ON(sseu->eu_stride > GEN_MAX_EU_STRIDE);
+}
+
 unsigned int
 intel_sseu_subslice_total(const struct sseu_dev_info *sseu)
 {
@@ -19,10 +32,32 @@ intel_sseu_subslice_total(const struct sseu_dev_info *sseu)
 	return total;
 }
 
+u32 intel_sseu_get_subslices(const struct sseu_dev_info *sseu, u8 slice)
+{
+	int i, offset = slice * sseu->ss_stride;
+	u32 mask = 0;
+
+	GEM_BUG_ON(slice >= sseu->max_slices);
+
+	for (i = 0; i < sseu->ss_stride; i++)
+		mask |= (u32)sseu->subslice_mask[offset + i] <<
+			i * BITS_PER_BYTE;
+
+	return mask;
+}
+
+void intel_sseu_set_subslices(struct sseu_dev_info *sseu, int slice,
+			      u32 ss_mask)
+{
+	int offset = slice * sseu->ss_stride;
+
+	memcpy(&sseu->subslice_mask[offset], &ss_mask, sseu->ss_stride);
+}
+
 unsigned int
 intel_sseu_subslices_per_slice(const struct sseu_dev_info *sseu, u8 slice)
 {
-	return hweight8(sseu->subslice_mask[slice]);
+	return hweight32(intel_sseu_get_subslices(sseu, slice));
 }
 
 u32 intel_sseu_make_rpcs(struct drm_i915_private *i915,
diff --git a/drivers/gpu/drm/i915/gt/intel_sseu.h b/drivers/gpu/drm/i915/gt/intel_sseu.h
index b50d0401a4e2..d1d225204f09 100644
--- a/drivers/gpu/drm/i915/gt/intel_sseu.h
+++ b/drivers/gpu/drm/i915/gt/intel_sseu.h
@@ -10,15 +10,21 @@
 #include <linux/types.h>
 #include <linux/kernel.h>
 
+#include "i915_gem.h"
+
 struct drm_i915_private;
 
 #define GEN_MAX_SLICES		(6) /* CNL upper bound */
 #define GEN_MAX_SUBSLICES	(8) /* ICL upper bound */
 #define GEN_SSEU_STRIDE(max_entries) DIV_ROUND_UP(max_entries, BITS_PER_BYTE)
+#define GEN_MAX_SUBSLICE_STRIDE GEN_SSEU_STRIDE(GEN_MAX_SUBSLICES)
+#define GEN_MAX_EUS		(16) /* TGL upper bound */
+#define GEN_MAX_EU_STRIDE GEN_SSEU_STRIDE(GEN_MAX_EUS)
 
 struct sseu_dev_info {
 	u8 slice_mask;
-	u8 subslice_mask[GEN_MAX_SLICES];
+	u8 subslice_mask[GEN_MAX_SLICES * GEN_MAX_SUBSLICE_STRIDE];
+	u8 eu_mask[GEN_MAX_SLICES * GEN_MAX_SUBSLICES * GEN_MAX_EU_STRIDE];
 	u16 eu_total;
 	u8 eu_per_subslice;
 	u8 min_eu_in_pool;
@@ -33,11 +39,8 @@ struct sseu_dev_info {
 	u8 max_subslices;
 	u8 max_eus_per_subslice;
 
-	/* We don't have more than 8 eus per subslice at the moment and as we
-	 * store eus enabled using bits, no need to multiply by eus per
-	 * subslice.
-	 */
-	u8 eu_mask[GEN_MAX_SLICES * GEN_MAX_SUBSLICES];
+	u8 ss_stride;
+	u8 eu_stride;
 };
 
 /*
@@ -63,12 +66,34 @@ intel_sseu_from_device_info(const struct sseu_dev_info *sseu)
 	return value;
 }
 
+static inline bool
+intel_sseu_has_subslice(const struct sseu_dev_info *sseu, int slice,
+			int subslice)
+{
+	u8 mask;
+	int ss_idx = subslice / BITS_PER_BYTE;
+
+	GEM_BUG_ON(ss_idx >= sseu->ss_stride);
+
+	mask = sseu->subslice_mask[slice * sseu->ss_stride + ss_idx];
+
+	return mask & BIT(subslice % BITS_PER_BYTE);
+}
+
+void intel_sseu_set_info(struct sseu_dev_info *sseu, u8 max_slices,
+			 u8 max_subslices, u8 max_eus_per_subslice);
+
 unsigned int
 intel_sseu_subslice_total(const struct sseu_dev_info *sseu);
 
 unsigned int
 intel_sseu_subslices_per_slice(const struct sseu_dev_info *sseu, u8 slice);
 
+u32  intel_sseu_get_subslices(const struct sseu_dev_info *sseu, u8 slice);
+
+void intel_sseu_set_subslices(struct sseu_dev_info *sseu, int slice,
+			      u32 ss_mask);
+
 u32 intel_sseu_make_rpcs(struct drm_i915_private *i915,
 			 const struct intel_sseu *req_sseu);
 
diff --git a/drivers/gpu/drm/i915/gt/intel_timeline.c b/drivers/gpu/drm/i915/gt/intel_timeline.c
index 9cb01d9828f1..14ad10acd548 100644
--- a/drivers/gpu/drm/i915/gt/intel_timeline.c
+++ b/drivers/gpu/drm/i915/gt/intel_timeline.c
@@ -4,13 +4,13 @@
  * Copyright © 2016-2018 Intel Corporation
  */
 
-#include "gt/intel_gt_types.h"
-
 #include "i915_drv.h"
 
 #include "i915_active.h"
 #include "i915_syncmap.h"
-#include "gt/intel_timeline.h"
+#include "intel_gt.h"
+#include "intel_ring.h"
+#include "intel_timeline.h"
 
 #define ptr_set_bit(ptr, bit) ((typeof(ptr))((unsigned long)(ptr) | BIT(bit)))
 #define ptr_test_bit(ptr, bit) ((unsigned long)(ptr) & BIT(bit))
@@ -136,6 +136,7 @@ static void __idle_cacheline_free(struct intel_timeline_cacheline *cl)
 	kfree(cl);
 }
 
+__i915_active_call
 static void __cacheline_retire(struct i915_active *active)
 {
 	struct intel_timeline_cacheline *cl =
@@ -177,8 +178,7 @@ cacheline_alloc(struct intel_timeline_hwsp *hwsp, unsigned int cacheline)
 	cl->hwsp = hwsp;
 	cl->vaddr = page_pack_bits(vaddr, cacheline);
 
-	i915_active_init(hwsp->gt->i915, &cl->active,
-			 __cacheline_active, __cacheline_retire);
+	i915_active_init(&cl->active, __cacheline_active, __cacheline_retire);
 
 	return cl;
 }
@@ -254,7 +254,7 @@ int intel_timeline_init(struct intel_timeline *timeline,
 
 	mutex_init(&timeline->mutex);
 
-	INIT_ACTIVE_REQUEST(&timeline->last_request, &timeline->mutex);
+	INIT_ACTIVE_FENCE(&timeline->last_request, &timeline->mutex);
 	INIT_LIST_HEAD(&timeline->requests);
 
 	i915_syncmap_init(&timeline->sync);
@@ -442,7 +442,7 @@ __intel_timeline_get_seqno(struct intel_timeline *tl,
 	 * free it after the current request is retired, which ensures that
 	 * all writes into the cacheline from previous requests are complete.
 	 */
-	err = i915_active_ref(&tl->hwsp_cacheline->active, tl, rq);
+	err = i915_active_ref(&tl->hwsp_cacheline->active, tl, &rq->fence);
 	if (err)
 		goto err_cacheline;
 
@@ -493,24 +493,39 @@ int intel_timeline_get_seqno(struct intel_timeline *tl,
 static int cacheline_ref(struct intel_timeline_cacheline *cl,
 			 struct i915_request *rq)
 {
-	return i915_active_ref(&cl->active, rq->timeline, rq);
+	return i915_active_add_request(&cl->active, rq);
 }
 
 int intel_timeline_read_hwsp(struct i915_request *from,
 			     struct i915_request *to,
 			     u32 *hwsp)
 {
-	struct intel_timeline_cacheline *cl = from->hwsp_cacheline;
-	struct intel_timeline *tl = from->timeline;
+	struct intel_timeline *tl;
 	int err;
 
-	GEM_BUG_ON(to->timeline == tl);
+	rcu_read_lock();
+	tl = rcu_dereference(from->timeline);
+	if (i915_request_completed(from) || !kref_get_unless_zero(&tl->kref))
+		tl = NULL;
+	rcu_read_unlock();
+	if (!tl) /* already completed */
+		return 1;
+
+	GEM_BUG_ON(rcu_access_pointer(to->timeline) == tl);
+
+	err = -EBUSY;
+	if (mutex_trylock(&tl->mutex)) {
+		struct intel_timeline_cacheline *cl = from->hwsp_cacheline;
+
+		if (i915_request_completed(from)) {
+			err = 1;
+			goto unlock;
+		}
 
-	mutex_lock_nested(&tl->mutex, SINGLE_DEPTH_NESTING);
-	err = i915_request_completed(from);
-	if (!err)
 		err = cacheline_ref(cl, to);
-	if (!err) {
+		if (err)
+			goto unlock;
+
 		if (likely(cl == tl->hwsp_cacheline)) {
 			*hwsp = tl->hwsp_offset;
 		} else { /* across a seqno wrap, recover the original offset */
@@ -518,8 +533,11 @@ int intel_timeline_read_hwsp(struct i915_request *from,
 				ptr_unmask_bits(cl->vaddr, CACHELINE_BITS) *
 				CACHELINE_BYTES;
 		}
+
+unlock:
+		mutex_unlock(&tl->mutex);
 	}
-	mutex_unlock(&tl->mutex);
+	intel_timeline_put(tl);
 
 	return err;
 }
@@ -541,7 +559,7 @@ void __intel_timeline_free(struct kref *kref)
 		container_of(kref, typeof(*timeline), kref);
 
 	intel_timeline_fini(timeline);
-	kfree(timeline);
+	kfree_rcu(timeline, rcu);
 }
 
 static void timelines_fini(struct intel_gt *gt)
diff --git a/drivers/gpu/drm/i915/gt/intel_timeline_types.h b/drivers/gpu/drm/i915/gt/intel_timeline_types.h
index 2b1baf2fcc8e..98d9ee166379 100644
--- a/drivers/gpu/drm/i915/gt/intel_timeline_types.h
+++ b/drivers/gpu/drm/i915/gt/intel_timeline_types.h
@@ -58,12 +58,13 @@ struct intel_timeline {
 	 */
 	struct list_head requests;
 
-	/* Contains an RCU guarded pointer to the last request. No reference is
+	/*
+	 * Contains an RCU guarded pointer to the last request. No reference is
 	 * held to the request, users must carefully acquire a reference to
-	 * the request using i915_active_request_get_request_rcu(), or hold the
-	 * struct_mutex.
+	 * the request using i915_active_fence_get(), or manage the RCU
+	 * protection themselves (cf the i915_active_fence API).
 	 */
-	struct i915_active_request last_request;
+	struct i915_active_fence last_request;
 
 	/**
 	 * We track the most recent seqno that we wait on in every context so
@@ -80,6 +81,7 @@ struct intel_timeline {
 	struct intel_gt *gt;
 
 	struct kref kref;
+	struct rcu_head rcu;
 };
 
 #endif /* __I915_TIMELINE_TYPES_H__ */
diff --git a/drivers/gpu/drm/i915/gt/intel_workarounds.c b/drivers/gpu/drm/i915/gt/intel_workarounds.c
index 45481eb1fa3c..e4bccc14602f 100644
--- a/drivers/gpu/drm/i915/gt/intel_workarounds.c
+++ b/drivers/gpu/drm/i915/gt/intel_workarounds.c
@@ -7,6 +7,7 @@
 #include "i915_drv.h"
 #include "intel_context.h"
 #include "intel_gt.h"
+#include "intel_ring.h"
 #include "intel_workarounds.h"
 
 /**
@@ -567,6 +568,9 @@ static void icl_ctx_workarounds_init(struct intel_engine_cs *engine,
 static void tgl_ctx_workarounds_init(struct intel_engine_cs *engine,
 				     struct i915_wa_list *wal)
 {
+	/* Wa_1409142259:tgl */
+	WA_SET_BIT_MASKED(GEN11_COMMON_SLICE_CHICKEN3,
+			  GEN12_DISABLE_CPS_AWARE_COLOR_PIPE);
 }
 
 static void
@@ -796,11 +800,10 @@ wa_init_mcr(struct drm_i915_private *i915, struct i915_wa_list *wal)
 	}
 
 	slice = fls(sseu->slice_mask) - 1;
-	GEM_BUG_ON(slice >= ARRAY_SIZE(sseu->subslice_mask));
-	subslice = fls(l3_en & sseu->subslice_mask[slice]);
+	subslice = fls(l3_en & intel_sseu_get_subslices(sseu, slice));
 	if (!subslice) {
 		DRM_WARN("No common index found between subslice mask %x and L3 bank mask %x!\n",
-			 sseu->subslice_mask[slice], l3_en);
+			 intel_sseu_get_subslices(sseu, slice), l3_en);
 		subslice = fls(l3_en);
 		WARN_ON(!subslice);
 	}
@@ -890,11 +893,27 @@ icl_gt_workarounds_init(struct drm_i915_private *i915, struct i915_wa_list *wal)
 	wa_write_or(wal,
 		    GAMT_CHKN_BIT_REG,
 		    GAMT_CHKN_DISABLE_L3_COH_PIPE);
+
+	/* Wa_1607087056:icl */
+	wa_write_or(wal,
+		    SLICE_UNIT_LEVEL_CLKGATE,
+		    L3_CLKGATE_DIS | L3_CR2X_CLKGATE_DIS);
 }
 
 static void
 tgl_gt_workarounds_init(struct drm_i915_private *i915, struct i915_wa_list *wal)
 {
+	/* Wa_1409420604:tgl */
+	if (IS_TGL_REVID(i915, TGL_REVID_A0, TGL_REVID_A0))
+		wa_write_or(wal,
+			    SUBSLICE_UNIT_LEVEL_CLKGATE2,
+			    CPSSUNIT_CLKGATE_DIS);
+
+	/* Wa_1409180338:tgl */
+	if (IS_TGL_REVID(i915, TGL_REVID_A0, TGL_REVID_A0))
+		wa_write_or(wal,
+			    SLICE_UNIT_LEVEL_CLKGATE,
+			    L3_CLKGATE_DIS | L3_CR2X_CLKGATE_DIS);
 }
 
 static void
@@ -1063,6 +1082,9 @@ static void gen9_whitelist_build(struct i915_wa_list *w)
 
 	/* WaAllowUMDToModifyHDCChicken1:skl,bxt,kbl,glk,cfl */
 	whitelist_reg(w, GEN8_HDC_CHICKEN1);
+
+	/* WaSendPushConstantsFromMMIO:skl,bxt */
+	whitelist_reg(w, COMMON_SLICE_CHICKEN2);
 }
 
 static void skl_whitelist_build(struct intel_engine_cs *engine)
@@ -1194,6 +1216,26 @@ static void icl_whitelist_build(struct intel_engine_cs *engine)
 
 static void tgl_whitelist_build(struct intel_engine_cs *engine)
 {
+	struct i915_wa_list *w = &engine->whitelist;
+
+	switch (engine->class) {
+	case RENDER_CLASS:
+		/*
+		 * WaAllowPMDepthAndInvocationCountAccessFromUMD:tgl
+		 *
+		 * This covers 4 registers which are next to one another :
+		 *   - PS_INVOCATION_COUNT
+		 *   - PS_INVOCATION_COUNT_UDW
+		 *   - PS_DEPTH_COUNT
+		 *   - PS_DEPTH_COUNT_UDW
+		 */
+		whitelist_reg_ext(w, PS_INVOCATION_COUNT,
+				  RING_FORCE_TO_NONPRIV_ACCESS_RD |
+				  RING_FORCE_TO_NONPRIV_RANGE_4);
+		break;
+	default:
+		break;
+	}
 }
 
 void intel_engine_init_whitelist(struct intel_engine_cs *engine)
@@ -1255,6 +1297,26 @@ rcs_engine_wa_init(struct intel_engine_cs *engine, struct i915_wa_list *wal)
 {
 	struct drm_i915_private *i915 = engine->i915;
 
+	if (IS_TGL_REVID(i915, TGL_REVID_A0, TGL_REVID_A0)) {
+		/* Wa_1606700617:tgl */
+		wa_masked_en(wal,
+			     GEN9_CS_DEBUG_MODE1,
+			     FF_DOP_CLOCK_GATE_DISABLE);
+
+		/* Wa_1607138336:tgl */
+		wa_write_or(wal,
+			    GEN9_CTX_PREEMPT_REG,
+			    GEN12_DISABLE_POSH_BUSY_FF_DOP_CG);
+
+		/* Wa_1607030317:tgl */
+		/* Wa_1607186500:tgl */
+		/* Wa_1607297627:tgl */
+		wa_masked_en(wal,
+			     GEN6_RC_SLEEP_PSMI_CONTROL,
+			     GEN12_WAIT_FOR_EVENT_POWER_DOWN_DISABLE |
+			     GEN8_RC_SEMA_IDLE_MSG_DISABLE);
+	}
+
 	if (IS_GEN(i915, 11)) {
 		/* This is not an Wa. Enable for better image quality */
 		wa_masked_en(wal,
@@ -1449,7 +1511,7 @@ static bool mcr_range(struct drm_i915_private *i915, u32 offset)
 	 * which only controls CPU initiated MMIO. Routing does not
 	 * work for CS access so we cannot verify them on this path.
 	 */
-	if (INTEL_GEN(i915) >= 8 && (offset >= 0xb100 && offset <= 0xb3ff))
+	if (INTEL_GEN(i915) >= 8 && (offset >= 0xb000 && offset <= 0xb4ff))
 		return true;
 
 	return false;
diff --git a/drivers/gpu/drm/i915/gt/mock_engine.c b/drivers/gpu/drm/i915/gt/mock_engine.c
index 5d43cbc3f345..83f549d203a0 100644
--- a/drivers/gpu/drm/i915/gt/mock_engine.c
+++ b/drivers/gpu/drm/i915/gt/mock_engine.c
@@ -23,6 +23,7 @@
  */
 
 #include "gem/i915_gem_context.h"
+#include "gt/intel_ring.h"
 
 #include "i915_drv.h"
 #include "intel_context.h"
@@ -240,6 +241,7 @@ struct intel_engine_cs *mock_engine(struct drm_i915_private *i915,
 	struct mock_engine *engine;
 
 	GEM_BUG_ON(id >= I915_NUM_ENGINES);
+	GEM_BUG_ON(!i915->gt.uncore);
 
 	engine = kzalloc(sizeof(*engine) + PAGE_SIZE, GFP_KERNEL);
 	if (!engine)
@@ -248,9 +250,11 @@ struct intel_engine_cs *mock_engine(struct drm_i915_private *i915,
 	/* minimal engine setup for requests */
 	engine->base.i915 = i915;
 	engine->base.gt = &i915->gt;
+	engine->base.uncore = i915->gt.uncore;
 	snprintf(engine->base.name, sizeof(engine->base.name), "%s", name);
 	engine->base.id = id;
 	engine->base.mask = BIT(id);
+	engine->base.legacy_idx = INVALID_ENGINE;
 	engine->base.instance = id;
 	engine->base.status_page.addr = (void *)(engine + 1);
 
@@ -265,6 +269,9 @@ struct intel_engine_cs *mock_engine(struct drm_i915_private *i915,
 	engine->base.reset.finish = mock_reset_finish;
 	engine->base.cancel_requests = mock_cancel_requests;
 
+	i915->gt.engine[id] = &engine->base;
+	i915->gt.engine_class[0][id] = &engine->base;
+
 	/* fake hw queue */
 	spin_lock_init(&engine->hw_lock);
 	timer_setup(&engine->hw_delay, hw_delay_complete, 0);
diff --git a/drivers/gpu/drm/i915/gt/selftest_context.c b/drivers/gpu/drm/i915/gt/selftest_context.c
index 9d1ea26c7a2d..bc720defc6b8 100644
--- a/drivers/gpu/drm/i915/gt/selftest_context.c
+++ b/drivers/gpu/drm/i915/gt/selftest_context.c
@@ -14,22 +14,28 @@
 
 static int request_sync(struct i915_request *rq)
 {
+	struct intel_timeline *tl = i915_request_timeline(rq);
 	long timeout;
 	int err = 0;
 
+	intel_timeline_get(tl);
 	i915_request_get(rq);
 
-	i915_request_add(rq);
+	/* Opencode i915_request_add() so we can keep the timeline locked. */
+	__i915_request_commit(rq);
+	__i915_request_queue(rq, NULL);
+
 	timeout = i915_request_wait(rq, 0, HZ / 10);
-	if (timeout < 0) {
+	if (timeout < 0)
 		err = timeout;
-	} else {
-		mutex_lock(&rq->timeline->mutex);
+	else
 		i915_request_retire_upto(rq);
-		mutex_unlock(&rq->timeline->mutex);
-	}
+
+	lockdep_unpin_lock(&tl->mutex, rq->cookie);
+	mutex_unlock(&tl->mutex);
 
 	i915_request_put(rq);
+	intel_timeline_put(tl);
 
 	return err;
 }
@@ -41,24 +47,20 @@ static int context_sync(struct intel_context *ce)
 
 	mutex_lock(&tl->mutex);
 	do {
-		struct i915_request *rq;
+		struct dma_fence *fence;
 		long timeout;
 
-		rcu_read_lock();
-		rq = rcu_dereference(tl->last_request.request);
-		if (rq)
-			rq = i915_request_get_rcu(rq);
-		rcu_read_unlock();
-		if (!rq)
+		fence = i915_active_fence_get(&tl->last_request);
+		if (!fence)
 			break;
 
-		timeout = i915_request_wait(rq, 0, HZ / 10);
+		timeout = dma_fence_wait_timeout(fence, false, HZ / 10);
 		if (timeout < 0)
 			err = timeout;
 		else
-			i915_request_retire_upto(rq);
+			i915_request_retire_upto(to_request(fence));
 
-		i915_request_put(rq);
+		dma_fence_put(fence);
 	} while (!err);
 	mutex_unlock(&tl->mutex);
 
@@ -101,9 +103,6 @@ static int __live_context_size(struct intel_engine_cs *engine,
 	 *
 	 * TLDR; this overlaps with the execlists redzone.
 	 */
-	if (HAS_EXECLISTS(engine->i915))
-		vaddr += LRC_HEADER_PAGES * PAGE_SIZE;
-
 	vaddr += engine->context_size - I915_GTT_PAGE_SIZE;
 	memset(vaddr, POISON_INUSE, I915_GTT_PAGE_SIZE);
 
@@ -153,15 +152,11 @@ static int live_context_size(void *arg)
 	 * HW tries to write past the end of one.
 	 */
 
-	mutex_lock(&gt->i915->drm.struct_mutex);
-
 	fixme = kernel_context(gt->i915);
-	if (IS_ERR(fixme)) {
-		err = PTR_ERR(fixme);
-		goto unlock;
-	}
+	if (IS_ERR(fixme))
+		return PTR_ERR(fixme);
 
-	for_each_engine(engine, gt->i915, id) {
+	for_each_engine(engine, gt, id) {
 		struct {
 			struct drm_i915_gem_object *state;
 			void *pinned;
@@ -199,8 +194,6 @@ static int live_context_size(void *arg)
 	}
 
 	kernel_context_close(fixme);
-unlock:
-	mutex_unlock(&gt->i915->drm.struct_mutex);
 	return err;
 }
 
@@ -303,26 +296,23 @@ static int live_active_context(void *arg)
 	if (IS_ERR(file))
 		return PTR_ERR(file);
 
-	mutex_lock(&gt->i915->drm.struct_mutex);
-
 	fixme = live_context(gt->i915, file);
 	if (IS_ERR(fixme)) {
 		err = PTR_ERR(fixme);
-		goto unlock;
+		goto out_file;
 	}
 
-	for_each_engine(engine, gt->i915, id) {
+	for_each_engine(engine, gt, id) {
 		err = __live_active_context(engine, fixme);
 		if (err)
 			break;
 
-		err = igt_flush_test(gt->i915, I915_WAIT_LOCKED);
+		err = igt_flush_test(gt->i915);
 		if (err)
 			break;
 	}
 
-unlock:
-	mutex_unlock(&gt->i915->drm.struct_mutex);
+out_file:
 	mock_file_free(gt->i915, file);
 	return err;
 }
@@ -416,26 +406,23 @@ static int live_remote_context(void *arg)
 	if (IS_ERR(file))
 		return PTR_ERR(file);
 
-	mutex_lock(&gt->i915->drm.struct_mutex);
-
 	fixme = live_context(gt->i915, file);
 	if (IS_ERR(fixme)) {
 		err = PTR_ERR(fixme);
-		goto unlock;
+		goto out_file;
 	}
 
-	for_each_engine(engine, gt->i915, id) {
+	for_each_engine(engine, gt, id) {
 		err = __live_remote_context(engine, fixme);
 		if (err)
 			break;
 
-		err = igt_flush_test(gt->i915, I915_WAIT_LOCKED);
+		err = igt_flush_test(gt->i915);
 		if (err)
 			break;
 	}
 
-unlock:
-	mutex_unlock(&gt->i915->drm.struct_mutex);
+out_file:
 	mock_file_free(gt->i915, file);
 	return err;
 }
diff --git a/drivers/gpu/drm/i915/gt/selftest_engine_heartbeat.c b/drivers/gpu/drm/i915/gt/selftest_engine_heartbeat.c
new file mode 100644
index 000000000000..e864406bd2d9
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/selftest_engine_heartbeat.c
@@ -0,0 +1,350 @@
+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2018 Intel Corporation
+ */
+
+#include <linux/sort.h>
+
+#include "i915_drv.h"
+
+#include "intel_gt_requests.h"
+#include "i915_selftest.h"
+
+struct pulse {
+	struct i915_active active;
+	struct kref kref;
+};
+
+static int pulse_active(struct i915_active *active)
+{
+	kref_get(&container_of(active, struct pulse, active)->kref);
+	return 0;
+}
+
+static void pulse_free(struct kref *kref)
+{
+	kfree(container_of(kref, struct pulse, kref));
+}
+
+static void pulse_put(struct pulse *p)
+{
+	kref_put(&p->kref, pulse_free);
+}
+
+static void pulse_retire(struct i915_active *active)
+{
+	pulse_put(container_of(active, struct pulse, active));
+}
+
+static struct pulse *pulse_create(void)
+{
+	struct pulse *p;
+
+	p = kmalloc(sizeof(*p), GFP_KERNEL);
+	if (!p)
+		return p;
+
+	kref_init(&p->kref);
+	i915_active_init(&p->active, pulse_active, pulse_retire);
+
+	return p;
+}
+
+static void pulse_unlock_wait(struct pulse *p)
+{
+	mutex_lock(&p->active.mutex);
+	mutex_unlock(&p->active.mutex);
+	flush_work(&p->active.work);
+}
+
+static int __live_idle_pulse(struct intel_engine_cs *engine,
+			     int (*fn)(struct intel_engine_cs *cs))
+{
+	struct pulse *p;
+	int err;
+
+	GEM_BUG_ON(!intel_engine_pm_is_awake(engine));
+
+	p = pulse_create();
+	if (!p)
+		return -ENOMEM;
+
+	err = i915_active_acquire(&p->active);
+	if (err)
+		goto out;
+
+	err = i915_active_acquire_preallocate_barrier(&p->active, engine);
+	if (err) {
+		i915_active_release(&p->active);
+		goto out;
+	}
+
+	i915_active_acquire_barrier(&p->active);
+	i915_active_release(&p->active);
+
+	GEM_BUG_ON(i915_active_is_idle(&p->active));
+	GEM_BUG_ON(llist_empty(&engine->barrier_tasks));
+
+	err = fn(engine);
+	if (err)
+		goto out;
+
+	GEM_BUG_ON(!llist_empty(&engine->barrier_tasks));
+
+	if (intel_gt_retire_requests_timeout(engine->gt, HZ / 5)) {
+		err = -ETIME;
+		goto out;
+	}
+
+	GEM_BUG_ON(READ_ONCE(engine->serial) != engine->wakeref_serial);
+
+	pulse_unlock_wait(p); /* synchronize with the retirement callback */
+
+	if (!i915_active_is_idle(&p->active)) {
+		struct drm_printer m = drm_err_printer("pulse");
+
+		pr_err("%s: heartbeat pulse did not flush idle tasks\n",
+		       engine->name);
+		i915_active_print(&p->active, &m);
+
+		err = -EINVAL;
+		goto out;
+	}
+
+out:
+	pulse_put(p);
+	return err;
+}
+
+static int live_idle_flush(void *arg)
+{
+	struct intel_gt *gt = arg;
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
+	int err = 0;
+
+	/* Check that we can flush the idle barriers */
+
+	for_each_engine(engine, gt, id) {
+		intel_engine_pm_get(engine);
+		err = __live_idle_pulse(engine, intel_engine_flush_barriers);
+		intel_engine_pm_put(engine);
+		if (err)
+			break;
+	}
+
+	return err;
+}
+
+static int live_idle_pulse(void *arg)
+{
+	struct intel_gt *gt = arg;
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
+	int err = 0;
+
+	/* Check that heartbeat pulses flush the idle barriers */
+
+	for_each_engine(engine, gt, id) {
+		intel_engine_pm_get(engine);
+		err = __live_idle_pulse(engine, intel_engine_pulse);
+		intel_engine_pm_put(engine);
+		if (err && err != -ENODEV)
+			break;
+
+		err = 0;
+	}
+
+	return err;
+}
+
+static int cmp_u32(const void *_a, const void *_b)
+{
+	const u32 *a = _a, *b = _b;
+
+	return *a - *b;
+}
+
+static int __live_heartbeat_fast(struct intel_engine_cs *engine)
+{
+	struct intel_context *ce;
+	struct i915_request *rq;
+	ktime_t t0, t1;
+	u32 times[5];
+	int err;
+	int i;
+
+	ce = intel_context_create(engine->kernel_context->gem_context,
+				  engine);
+	if (IS_ERR(ce))
+		return PTR_ERR(ce);
+
+	intel_engine_pm_get(engine);
+
+	err = intel_engine_set_heartbeat(engine, 1);
+	if (err)
+		goto err_pm;
+
+	for (i = 0; i < ARRAY_SIZE(times); i++) {
+		/* Manufacture a tick */
+		do {
+			while (READ_ONCE(engine->heartbeat.systole))
+				flush_delayed_work(&engine->heartbeat.work);
+
+			engine->serial++; /* quick, pretend we are not idle! */
+			flush_delayed_work(&engine->heartbeat.work);
+			if (!delayed_work_pending(&engine->heartbeat.work)) {
+				pr_err("%s: heartbeat did not start\n",
+				       engine->name);
+				err = -EINVAL;
+				goto err_pm;
+			}
+
+			rcu_read_lock();
+			rq = READ_ONCE(engine->heartbeat.systole);
+			if (rq)
+				rq = i915_request_get_rcu(rq);
+			rcu_read_unlock();
+		} while (!rq);
+
+		t0 = ktime_get();
+		while (rq == READ_ONCE(engine->heartbeat.systole))
+			yield(); /* work is on the local cpu! */
+		t1 = ktime_get();
+
+		i915_request_put(rq);
+		times[i] = ktime_us_delta(t1, t0);
+	}
+
+	sort(times, ARRAY_SIZE(times), sizeof(times[0]), cmp_u32, NULL);
+
+	pr_info("%s: Heartbeat delay: %uus [%u, %u]\n",
+		engine->name,
+		times[ARRAY_SIZE(times) / 2],
+		times[0],
+		times[ARRAY_SIZE(times) - 1]);
+
+	/* Min work delay is 2 * 2 (worst), +1 for scheduling, +1 for slack */
+	if (times[ARRAY_SIZE(times) / 2] > jiffies_to_usecs(6)) {
+		pr_err("%s: Heartbeat delay was %uus, expected less than %dus\n",
+		       engine->name,
+		       times[ARRAY_SIZE(times) / 2],
+		       jiffies_to_usecs(6));
+		err = -EINVAL;
+	}
+
+	intel_engine_set_heartbeat(engine, CONFIG_DRM_I915_HEARTBEAT_INTERVAL);
+err_pm:
+	intel_engine_pm_put(engine);
+	intel_context_put(ce);
+	return err;
+}
+
+static int live_heartbeat_fast(void *arg)
+{
+	struct intel_gt *gt = arg;
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
+	int err = 0;
+
+	/* Check that the heartbeat ticks at the desired rate. */
+	if (!CONFIG_DRM_I915_HEARTBEAT_INTERVAL)
+		return 0;
+
+	for_each_engine(engine, gt, id) {
+		err = __live_heartbeat_fast(engine);
+		if (err)
+			break;
+	}
+
+	return err;
+}
+
+static int __live_heartbeat_off(struct intel_engine_cs *engine)
+{
+	int err;
+
+	intel_engine_pm_get(engine);
+
+	engine->serial++;
+	flush_delayed_work(&engine->heartbeat.work);
+	if (!delayed_work_pending(&engine->heartbeat.work)) {
+		pr_err("%s: heartbeat not running\n",
+		       engine->name);
+		err = -EINVAL;
+		goto err_pm;
+	}
+
+	err = intel_engine_set_heartbeat(engine, 0);
+	if (err)
+		goto err_pm;
+
+	engine->serial++;
+	flush_delayed_work(&engine->heartbeat.work);
+	if (delayed_work_pending(&engine->heartbeat.work)) {
+		pr_err("%s: heartbeat still running\n",
+		       engine->name);
+		err = -EINVAL;
+		goto err_beat;
+	}
+
+	if (READ_ONCE(engine->heartbeat.systole)) {
+		pr_err("%s: heartbeat still allocated\n",
+		       engine->name);
+		err = -EINVAL;
+		goto err_beat;
+	}
+
+err_beat:
+	intel_engine_set_heartbeat(engine, CONFIG_DRM_I915_HEARTBEAT_INTERVAL);
+err_pm:
+	intel_engine_pm_put(engine);
+	return err;
+}
+
+static int live_heartbeat_off(void *arg)
+{
+	struct intel_gt *gt = arg;
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
+	int err = 0;
+
+	/* Check that we can turn off heartbeat and not interrupt VIP */
+	if (!CONFIG_DRM_I915_HEARTBEAT_INTERVAL)
+		return 0;
+
+	for_each_engine(engine, gt, id) {
+		if (!intel_engine_has_preemption(engine))
+			continue;
+
+		err = __live_heartbeat_off(engine);
+		if (err)
+			break;
+	}
+
+	return err;
+}
+
+int intel_heartbeat_live_selftests(struct drm_i915_private *i915)
+{
+	static const struct i915_subtest tests[] = {
+		SUBTEST(live_idle_flush),
+		SUBTEST(live_idle_pulse),
+		SUBTEST(live_heartbeat_fast),
+		SUBTEST(live_heartbeat_off),
+	};
+	int saved_hangcheck;
+	int err;
+
+	if (intel_gt_is_wedged(&i915->gt))
+		return 0;
+
+	saved_hangcheck = i915_modparams.enable_hangcheck;
+	i915_modparams.enable_hangcheck = INT_MAX;
+
+	err = intel_gt_live_subtests(tests, &i915->gt);
+
+	i915_modparams.enable_hangcheck = saved_hangcheck;
+	return err;
+}
diff --git a/drivers/gpu/drm/i915/gt/selftest_engine_pm.c b/drivers/gpu/drm/i915/gt/selftest_engine_pm.c
index 3a1419376912..20b9c83f43ad 100644
--- a/drivers/gpu/drm/i915/gt/selftest_engine_pm.c
+++ b/drivers/gpu/drm/i915/gt/selftest_engine_pm.c
@@ -25,7 +25,7 @@ static int live_engine_pm(void *arg)
 	}
 
 	GEM_BUG_ON(intel_gt_pm_is_awake(gt));
-	for_each_engine(engine, gt->i915, id) {
+	for_each_engine(engine, gt, id) {
 		const typeof(*igt_atomic_phases) *p;
 
 		for (p = igt_atomic_phases; p->name; p++) {
diff --git a/drivers/gpu/drm/i915/gt/selftest_gt_pm.c b/drivers/gpu/drm/i915/gt/selftest_gt_pm.c
new file mode 100644
index 000000000000..d1752f15702a
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/selftest_gt_pm.c
@@ -0,0 +1,60 @@
+
+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2019 Intel Corporation
+ */
+
+#include "selftest_llc.h"
+
+static int live_gt_resume(void *arg)
+{
+	struct intel_gt *gt = arg;
+	IGT_TIMEOUT(end_time);
+	int err;
+
+	/* Do several suspend/resume cycles to check we don't explode! */
+	do {
+		intel_gt_suspend_prepare(gt);
+		intel_gt_suspend_late(gt);
+
+		if (gt->rc6.enabled) {
+			pr_err("rc6 still enabled after suspend!\n");
+			intel_gt_set_wedged_on_init(gt);
+			err = -EINVAL;
+			break;
+		}
+
+		err = intel_gt_resume(gt);
+		if (err)
+			break;
+
+		if (gt->rc6.supported && !gt->rc6.enabled) {
+			pr_err("rc6 not enabled upon resume!\n");
+			intel_gt_set_wedged_on_init(gt);
+			err = -EINVAL;
+			break;
+		}
+
+		err = st_llc_verify(&gt->llc);
+		if (err) {
+			pr_err("llc state not restored upon resume!\n");
+			intel_gt_set_wedged_on_init(gt);
+			break;
+		}
+	} while (!__igt_timeout(end_time, NULL));
+
+	return err;
+}
+
+int intel_gt_pm_live_selftests(struct drm_i915_private *i915)
+{
+	static const struct i915_subtest tests[] = {
+		SUBTEST(live_gt_resume),
+	};
+
+	if (intel_gt_is_wedged(&i915->gt))
+		return 0;
+
+	return intel_gt_live_subtests(tests, &i915->gt);
+}
diff --git a/drivers/gpu/drm/i915/gt/selftest_hangcheck.c b/drivers/gpu/drm/i915/gt/selftest_hangcheck.c
index a0098fc35921..85e9ccf5c304 100644
--- a/drivers/gpu/drm/i915/gt/selftest_hangcheck.c
+++ b/drivers/gpu/drm/i915/gt/selftest_hangcheck.c
@@ -131,7 +131,7 @@ static struct i915_request *
 hang_create_request(struct hang *h, struct intel_engine_cs *engine)
 {
 	struct intel_gt *gt = h->gt;
-	struct i915_address_space *vm = h->ctx->vm ?: &engine->gt->ggtt->vm;
+	struct i915_address_space *vm = i915_gem_context_get_vm_rcu(h->ctx);
 	struct drm_i915_gem_object *obj;
 	struct i915_request *rq = NULL;
 	struct i915_vma *hws, *vma;
@@ -141,12 +141,15 @@ hang_create_request(struct hang *h, struct intel_engine_cs *engine)
 	int err;
 
 	obj = i915_gem_object_create_internal(gt->i915, PAGE_SIZE);
-	if (IS_ERR(obj))
+	if (IS_ERR(obj)) {
+		i915_vm_put(vm);
 		return ERR_CAST(obj);
+	}
 
 	vaddr = i915_gem_object_pin_map(obj, i915_coherent_map_type(gt->i915));
 	if (IS_ERR(vaddr)) {
 		i915_gem_object_put(obj);
+		i915_vm_put(vm);
 		return ERR_CAST(vaddr);
 	}
 
@@ -157,16 +160,22 @@ hang_create_request(struct hang *h, struct intel_engine_cs *engine)
 	h->batch = vaddr;
 
 	vma = i915_vma_instance(h->obj, vm, NULL);
-	if (IS_ERR(vma))
+	if (IS_ERR(vma)) {
+		i915_vm_put(vm);
 		return ERR_CAST(vma);
+	}
 
 	hws = i915_vma_instance(h->hws, vm, NULL);
-	if (IS_ERR(hws))
+	if (IS_ERR(hws)) {
+		i915_vm_put(vm);
 		return ERR_CAST(hws);
+	}
 
 	err = i915_vma_pin(vma, 0, 0, PIN_USER);
-	if (err)
+	if (err) {
+		i915_vm_put(vm);
 		return ERR_PTR(err);
+	}
 
 	err = i915_vma_pin(hws, 0, 0, PIN_USER);
 	if (err)
@@ -264,6 +273,7 @@ unpin_hws:
 	i915_vma_unpin(hws);
 unpin_vma:
 	i915_vma_unpin(vma);
+	i915_vm_put(vm);
 	return err ? ERR_PTR(err) : rq;
 }
 
@@ -285,7 +295,7 @@ static void hang_fini(struct hang *h)
 
 	kernel_context_close(h->ctx);
 
-	igt_flush_test(h->gt->i915, I915_WAIT_LOCKED);
+	igt_flush_test(h->gt->i915);
 }
 
 static bool wait_until_running(struct hang *h, struct i915_request *rq)
@@ -309,12 +319,11 @@ static int igt_hang_sanitycheck(void *arg)
 
 	/* Basic check that we can execute our hanging batch */
 
-	mutex_lock(&gt->i915->drm.struct_mutex);
 	err = hang_init(&h, gt);
 	if (err)
-		goto unlock;
+		return err;
 
-	for_each_engine(engine, gt->i915, id) {
+	for_each_engine(engine, gt, id) {
 		struct intel_wedge_me w;
 		long timeout;
 
@@ -355,8 +364,6 @@ static int igt_hang_sanitycheck(void *arg)
 
 fini:
 	hang_fini(&h);
-unlock:
-	mutex_unlock(&gt->i915->drm.struct_mutex);
 	return err;
 }
 
@@ -383,9 +390,7 @@ static int igt_reset_nop(void *arg)
 	if (IS_ERR(file))
 		return PTR_ERR(file);
 
-	mutex_lock(&gt->i915->drm.struct_mutex);
 	ctx = live_context(gt->i915, file);
-	mutex_unlock(&gt->i915->drm.struct_mutex);
 	if (IS_ERR(ctx)) {
 		err = PTR_ERR(ctx);
 		goto out;
@@ -395,9 +400,7 @@ static int igt_reset_nop(void *arg)
 	reset_count = i915_reset_count(global);
 	count = 0;
 	do {
-		mutex_lock(&gt->i915->drm.struct_mutex);
-
-		for_each_engine(engine, gt->i915, id) {
+		for_each_engine(engine, gt, id) {
 			int i;
 
 			for (i = 0; i < 16; i++) {
@@ -417,7 +420,6 @@ static int igt_reset_nop(void *arg)
 		intel_gt_reset(gt, ALL_ENGINES, NULL);
 		igt_global_reset_unlock(gt);
 
-		mutex_unlock(&gt->i915->drm.struct_mutex);
 		if (intel_gt_is_wedged(gt)) {
 			err = -EIO;
 			break;
@@ -429,16 +431,13 @@ static int igt_reset_nop(void *arg)
 			break;
 		}
 
-		err = igt_flush_test(gt->i915, 0);
+		err = igt_flush_test(gt->i915);
 		if (err)
 			break;
 	} while (time_before(jiffies, end_time));
 	pr_info("%s: %d resets\n", __func__, count);
 
-	mutex_lock(&gt->i915->drm.struct_mutex);
-	err = igt_flush_test(gt->i915, I915_WAIT_LOCKED);
-	mutex_unlock(&gt->i915->drm.struct_mutex);
-
+	err = igt_flush_test(gt->i915);
 out:
 	mock_file_free(gt->i915, file);
 	if (intel_gt_is_wedged(gt))
@@ -458,23 +457,21 @@ static int igt_reset_nop_engine(void *arg)
 
 	/* Check that we can engine-reset during non-user portions */
 
-	if (!intel_has_reset_engine(gt->i915))
+	if (!intel_has_reset_engine(gt))
 		return 0;
 
 	file = mock_file(gt->i915);
 	if (IS_ERR(file))
 		return PTR_ERR(file);
 
-	mutex_lock(&gt->i915->drm.struct_mutex);
 	ctx = live_context(gt->i915, file);
-	mutex_unlock(&gt->i915->drm.struct_mutex);
 	if (IS_ERR(ctx)) {
 		err = PTR_ERR(ctx);
 		goto out;
 	}
 
 	i915_gem_context_clear_bannable(ctx);
-	for_each_engine(engine, gt->i915, id) {
+	for_each_engine(engine, gt, id) {
 		unsigned int reset_count, reset_engine_count;
 		unsigned int count;
 		IGT_TIMEOUT(end_time);
@@ -494,7 +491,6 @@ static int igt_reset_nop_engine(void *arg)
 				break;
 			}
 
-			mutex_lock(&gt->i915->drm.struct_mutex);
 			for (i = 0; i < 16; i++) {
 				struct i915_request *rq;
 
@@ -507,7 +503,6 @@ static int igt_reset_nop_engine(void *arg)
 				i915_request_add(rq);
 			}
 			err = intel_engine_reset(engine, NULL);
-			mutex_unlock(&gt->i915->drm.struct_mutex);
 			if (err) {
 				pr_err("i915_reset_engine failed\n");
 				break;
@@ -533,15 +528,12 @@ static int igt_reset_nop_engine(void *arg)
 		if (err)
 			break;
 
-		err = igt_flush_test(gt->i915, 0);
+		err = igt_flush_test(gt->i915);
 		if (err)
 			break;
 	}
 
-	mutex_lock(&gt->i915->drm.struct_mutex);
-	err = igt_flush_test(gt->i915, I915_WAIT_LOCKED);
-	mutex_unlock(&gt->i915->drm.struct_mutex);
-
+	err = igt_flush_test(gt->i915);
 out:
 	mock_file_free(gt->i915, file);
 	if (intel_gt_is_wedged(gt))
@@ -559,18 +551,16 @@ static int __igt_reset_engine(struct intel_gt *gt, bool active)
 
 	/* Check that we can issue an engine reset on an idle engine (no-op) */
 
-	if (!intel_has_reset_engine(gt->i915))
+	if (!intel_has_reset_engine(gt))
 		return 0;
 
 	if (active) {
-		mutex_lock(&gt->i915->drm.struct_mutex);
 		err = hang_init(&h, gt);
-		mutex_unlock(&gt->i915->drm.struct_mutex);
 		if (err)
 			return err;
 	}
 
-	for_each_engine(engine, gt->i915, id) {
+	for_each_engine(engine, gt, id) {
 		unsigned int reset_count, reset_engine_count;
 		IGT_TIMEOUT(end_time);
 
@@ -593,17 +583,14 @@ static int __igt_reset_engine(struct intel_gt *gt, bool active)
 			if (active) {
 				struct i915_request *rq;
 
-				mutex_lock(&gt->i915->drm.struct_mutex);
 				rq = hang_create_request(&h, engine);
 				if (IS_ERR(rq)) {
 					err = PTR_ERR(rq);
-					mutex_unlock(&gt->i915->drm.struct_mutex);
 					break;
 				}
 
 				i915_request_get(rq);
 				i915_request_add(rq);
-				mutex_unlock(&gt->i915->drm.struct_mutex);
 
 				if (!wait_until_running(&h, rq)) {
 					struct drm_printer p = drm_info_printer(gt->i915->drm.dev);
@@ -647,7 +634,7 @@ static int __igt_reset_engine(struct intel_gt *gt, bool active)
 		if (err)
 			break;
 
-		err = igt_flush_test(gt->i915, 0);
+		err = igt_flush_test(gt->i915);
 		if (err)
 			break;
 	}
@@ -655,11 +642,8 @@ static int __igt_reset_engine(struct intel_gt *gt, bool active)
 	if (intel_gt_is_wedged(gt))
 		err = -EIO;
 
-	if (active) {
-		mutex_lock(&gt->i915->drm.struct_mutex);
+	if (active)
 		hang_fini(&h);
-		mutex_unlock(&gt->i915->drm.struct_mutex);
-	}
 
 	return err;
 }
@@ -725,9 +709,7 @@ static int active_engine(void *data)
 		return PTR_ERR(file);
 
 	for (count = 0; count < ARRAY_SIZE(ctx); count++) {
-		mutex_lock(&engine->i915->drm.struct_mutex);
 		ctx[count] = live_context(engine->i915, file);
-		mutex_unlock(&engine->i915->drm.struct_mutex);
 		if (IS_ERR(ctx[count])) {
 			err = PTR_ERR(ctx[count]);
 			while (--count)
@@ -741,10 +723,8 @@ static int active_engine(void *data)
 		struct i915_request *old = rq[idx];
 		struct i915_request *new;
 
-		mutex_lock(&engine->i915->drm.struct_mutex);
 		new = igt_request_alloc(ctx[idx], engine);
 		if (IS_ERR(new)) {
-			mutex_unlock(&engine->i915->drm.struct_mutex);
 			err = PTR_ERR(new);
 			break;
 		}
@@ -755,7 +735,6 @@ static int active_engine(void *data)
 
 		rq[idx] = i915_request_get(new);
 		i915_request_add(new);
-		mutex_unlock(&engine->i915->drm.struct_mutex);
 
 		err = active_request_put(old);
 		if (err)
@@ -791,13 +770,11 @@ static int __igt_reset_engines(struct intel_gt *gt,
 	 * with any other engine.
 	 */
 
-	if (!intel_has_reset_engine(gt->i915))
+	if (!intel_has_reset_engine(gt))
 		return 0;
 
 	if (flags & TEST_ACTIVE) {
-		mutex_lock(&gt->i915->drm.struct_mutex);
 		err = hang_init(&h, gt);
-		mutex_unlock(&gt->i915->drm.struct_mutex);
 		if (err)
 			return err;
 
@@ -805,7 +782,7 @@ static int __igt_reset_engines(struct intel_gt *gt,
 			h.ctx->sched.priority = 1024;
 	}
 
-	for_each_engine(engine, gt->i915, id) {
+	for_each_engine(engine, gt, id) {
 		struct active_engine threads[I915_NUM_ENGINES] = {};
 		unsigned long device = i915_reset_count(global);
 		unsigned long count = 0, reported;
@@ -823,7 +800,7 @@ static int __igt_reset_engines(struct intel_gt *gt,
 		}
 
 		memset(threads, 0, sizeof(threads));
-		for_each_engine(other, gt->i915, tmp) {
+		for_each_engine(other, gt, tmp) {
 			struct task_struct *tsk;
 
 			threads[tmp].resets =
@@ -849,23 +826,22 @@ static int __igt_reset_engines(struct intel_gt *gt,
 			get_task_struct(tsk);
 		}
 
+		yield(); /* start all threads before we begin */
+
 		intel_engine_pm_get(engine);
 		set_bit(I915_RESET_ENGINE + id, &gt->reset.flags);
 		do {
 			struct i915_request *rq = NULL;
 
 			if (flags & TEST_ACTIVE) {
-				mutex_lock(&gt->i915->drm.struct_mutex);
 				rq = hang_create_request(&h, engine);
 				if (IS_ERR(rq)) {
 					err = PTR_ERR(rq);
-					mutex_unlock(&gt->i915->drm.struct_mutex);
 					break;
 				}
 
 				i915_request_get(rq);
 				i915_request_add(rq);
-				mutex_unlock(&gt->i915->drm.struct_mutex);
 
 				if (!wait_until_running(&h, rq)) {
 					struct drm_printer p = drm_info_printer(gt->i915->drm.dev);
@@ -940,7 +916,7 @@ static int __igt_reset_engines(struct intel_gt *gt,
 		}
 
 unwind:
-		for_each_engine(other, gt->i915, tmp) {
+		for_each_engine(other, gt, tmp) {
 			int ret;
 
 			if (!threads[tmp].task)
@@ -977,9 +953,7 @@ unwind:
 		if (err)
 			break;
 
-		mutex_lock(&gt->i915->drm.struct_mutex);
-		err = igt_flush_test(gt->i915, I915_WAIT_LOCKED);
-		mutex_unlock(&gt->i915->drm.struct_mutex);
+		err = igt_flush_test(gt->i915);
 		if (err)
 			break;
 	}
@@ -987,11 +961,8 @@ unwind:
 	if (intel_gt_is_wedged(gt))
 		err = -EIO;
 
-	if (flags & TEST_ACTIVE) {
-		mutex_lock(&gt->i915->drm.struct_mutex);
+	if (flags & TEST_ACTIVE)
 		hang_fini(&h);
-		mutex_unlock(&gt->i915->drm.struct_mutex);
-	}
 
 	return err;
 }
@@ -1047,7 +1018,7 @@ static int igt_reset_wait(void *arg)
 {
 	struct intel_gt *gt = arg;
 	struct i915_gpu_error *global = &gt->i915->gpu_error;
-	struct intel_engine_cs *engine = gt->i915->engine[RCS0];
+	struct intel_engine_cs *engine = gt->engine[RCS0];
 	struct i915_request *rq;
 	unsigned int reset_count;
 	struct hang h;
@@ -1061,7 +1032,6 @@ static int igt_reset_wait(void *arg)
 
 	igt_global_reset_lock(gt);
 
-	mutex_lock(&gt->i915->drm.struct_mutex);
 	err = hang_init(&h, gt);
 	if (err)
 		goto unlock;
@@ -1109,7 +1079,6 @@ out_rq:
 fini:
 	hang_fini(&h);
 unlock:
-	mutex_unlock(&gt->i915->drm.struct_mutex);
 	igt_global_reset_unlock(gt);
 
 	if (intel_gt_is_wedged(gt))
@@ -1127,15 +1096,14 @@ static int evict_vma(void *data)
 {
 	struct evict_vma *arg = data;
 	struct i915_address_space *vm = arg->vma->vm;
-	struct drm_i915_private *i915 = vm->i915;
 	struct drm_mm_node evict = arg->vma->node;
 	int err;
 
 	complete(&arg->completion);
 
-	mutex_lock(&i915->drm.struct_mutex);
+	mutex_lock(&vm->mutex);
 	err = i915_gem_evict_for_node(vm, &evict, 0);
-	mutex_unlock(&i915->drm.struct_mutex);
+	mutex_unlock(&vm->mutex);
 
 	return err;
 }
@@ -1143,39 +1111,33 @@ static int evict_vma(void *data)
 static int evict_fence(void *data)
 {
 	struct evict_vma *arg = data;
-	struct drm_i915_private *i915 = arg->vma->vm->i915;
 	int err;
 
 	complete(&arg->completion);
 
-	mutex_lock(&i915->drm.struct_mutex);
-
 	/* Mark the fence register as dirty to force the mmio update. */
 	err = i915_gem_object_set_tiling(arg->vma->obj, I915_TILING_Y, 512);
 	if (err) {
 		pr_err("Invalid Y-tiling settings; err:%d\n", err);
-		goto out_unlock;
+		return err;
 	}
 
 	err = i915_vma_pin(arg->vma, 0, 0, PIN_GLOBAL | PIN_MAPPABLE);
 	if (err) {
 		pr_err("Unable to pin vma for Y-tiled fence; err:%d\n", err);
-		goto out_unlock;
+		return err;
 	}
 
 	err = i915_vma_pin_fence(arg->vma);
 	i915_vma_unpin(arg->vma);
 	if (err) {
 		pr_err("Unable to pin Y-tiled fence; err:%d\n", err);
-		goto out_unlock;
+		return err;
 	}
 
 	i915_vma_unpin_fence(arg->vma);
 
-out_unlock:
-	mutex_unlock(&i915->drm.struct_mutex);
-
-	return err;
+	return 0;
 }
 
 static int __igt_reset_evict_vma(struct intel_gt *gt,
@@ -1183,23 +1145,26 @@ static int __igt_reset_evict_vma(struct intel_gt *gt,
 				 int (*fn)(void *),
 				 unsigned int flags)
 {
-	struct intel_engine_cs *engine = gt->i915->engine[RCS0];
+	struct intel_engine_cs *engine = gt->engine[RCS0];
 	struct drm_i915_gem_object *obj;
 	struct task_struct *tsk = NULL;
 	struct i915_request *rq;
 	struct evict_vma arg;
 	struct hang h;
+	unsigned int pin_flags;
 	int err;
 
+	if (!gt->ggtt->num_fences && flags & EXEC_OBJECT_NEEDS_FENCE)
+		return 0;
+
 	if (!engine || !intel_engine_can_store_dword(engine))
 		return 0;
 
 	/* Check that we can recover an unbind stuck on a hanging request */
 
-	mutex_lock(&gt->i915->drm.struct_mutex);
 	err = hang_init(&h, gt);
 	if (err)
-		goto unlock;
+		return err;
 
 	obj = i915_gem_object_create_internal(gt->i915, SZ_1M);
 	if (IS_ERR(obj)) {
@@ -1227,10 +1192,12 @@ static int __igt_reset_evict_vma(struct intel_gt *gt,
 		goto out_obj;
 	}
 
-	err = i915_vma_pin(arg.vma, 0, 0,
-			   i915_vma_is_ggtt(arg.vma) ?
-			   PIN_GLOBAL | PIN_MAPPABLE :
-			   PIN_USER);
+	pin_flags = i915_vma_is_ggtt(arg.vma) ? PIN_GLOBAL : PIN_USER;
+
+	if (flags & EXEC_OBJECT_NEEDS_FENCE)
+		pin_flags |= PIN_MAPPABLE;
+
+	err = i915_vma_pin(arg.vma, 0, 0, pin_flags);
 	if (err) {
 		i915_request_add(rq);
 		goto out_obj;
@@ -1262,8 +1229,6 @@ static int __igt_reset_evict_vma(struct intel_gt *gt,
 	if (err)
 		goto out_rq;
 
-	mutex_unlock(&gt->i915->drm.struct_mutex);
-
 	if (!wait_until_running(&h, rq)) {
 		struct drm_printer p = drm_info_printer(gt->i915->drm.dev);
 
@@ -1312,16 +1277,12 @@ out_reset:
 		put_task_struct(tsk);
 	}
 
-	mutex_lock(&gt->i915->drm.struct_mutex);
 out_rq:
 	i915_request_put(rq);
 out_obj:
 	i915_gem_object_put(obj);
 fini:
 	hang_fini(&h);
-unlock:
-	mutex_unlock(&gt->i915->drm.struct_mutex);
-
 	if (intel_gt_is_wedged(gt))
 		return -EIO;
 
@@ -1340,6 +1301,7 @@ static int igt_reset_evict_ppgtt(void *arg)
 {
 	struct intel_gt *gt = arg;
 	struct i915_gem_context *ctx;
+	struct i915_address_space *vm;
 	struct drm_file *file;
 	int err;
 
@@ -1347,18 +1309,20 @@ static int igt_reset_evict_ppgtt(void *arg)
 	if (IS_ERR(file))
 		return PTR_ERR(file);
 
-	mutex_lock(&gt->i915->drm.struct_mutex);
 	ctx = live_context(gt->i915, file);
-	mutex_unlock(&gt->i915->drm.struct_mutex);
 	if (IS_ERR(ctx)) {
 		err = PTR_ERR(ctx);
 		goto out;
 	}
 
 	err = 0;
-	if (ctx->vm) /* aliasing == global gtt locking, covered above */
-		err = __igt_reset_evict_vma(gt, ctx->vm,
+	vm = i915_gem_context_get_vm_rcu(ctx);
+	if (!i915_is_ggtt(vm)) {
+		/* aliasing == global gtt locking, covered above */
+		err = __igt_reset_evict_vma(gt, vm,
 					    evict_vma, EXEC_OBJECT_WRITE);
+	}
+	i915_vm_put(vm);
 
 out:
 	mock_file_free(gt->i915, file);
@@ -1379,7 +1343,7 @@ static int wait_for_others(struct intel_gt *gt,
 	struct intel_engine_cs *engine;
 	enum intel_engine_id id;
 
-	for_each_engine(engine, gt->i915, id) {
+	for_each_engine(engine, gt, id) {
 		if (engine == exclude)
 			continue;
 
@@ -1403,12 +1367,11 @@ static int igt_reset_queue(void *arg)
 
 	igt_global_reset_lock(gt);
 
-	mutex_lock(&gt->i915->drm.struct_mutex);
 	err = hang_init(&h, gt);
 	if (err)
 		goto unlock;
 
-	for_each_engine(engine, gt->i915, id) {
+	for_each_engine(engine, gt, id) {
 		struct i915_request *prev;
 		IGT_TIMEOUT(end_time);
 		unsigned int count;
@@ -1518,7 +1481,7 @@ static int igt_reset_queue(void *arg)
 
 		i915_request_put(prev);
 
-		err = igt_flush_test(gt->i915, I915_WAIT_LOCKED);
+		err = igt_flush_test(gt->i915);
 		if (err)
 			break;
 	}
@@ -1526,7 +1489,6 @@ static int igt_reset_queue(void *arg)
 fini:
 	hang_fini(&h);
 unlock:
-	mutex_unlock(&gt->i915->drm.struct_mutex);
 	igt_global_reset_unlock(gt);
 
 	if (intel_gt_is_wedged(gt))
@@ -1539,7 +1501,7 @@ static int igt_handle_error(void *arg)
 {
 	struct intel_gt *gt = arg;
 	struct i915_gpu_error *global = &gt->i915->gpu_error;
-	struct intel_engine_cs *engine = gt->i915->engine[RCS0];
+	struct intel_engine_cs *engine = gt->engine[RCS0];
 	struct hang h;
 	struct i915_request *rq;
 	struct i915_gpu_state *error;
@@ -1547,17 +1509,15 @@ static int igt_handle_error(void *arg)
 
 	/* Check that we can issue a global GPU and engine reset */
 
-	if (!intel_has_reset_engine(gt->i915))
+	if (!intel_has_reset_engine(gt))
 		return 0;
 
 	if (!engine || !intel_engine_can_store_dword(engine))
 		return 0;
 
-	mutex_lock(&gt->i915->drm.struct_mutex);
-
 	err = hang_init(&h, gt);
 	if (err)
-		goto err_unlock;
+		return err;
 
 	rq = hang_create_request(&h, engine);
 	if (IS_ERR(rq)) {
@@ -1581,8 +1541,6 @@ static int igt_handle_error(void *arg)
 		goto err_request;
 	}
 
-	mutex_unlock(&gt->i915->drm.struct_mutex);
-
 	/* Temporarily disable error capture */
 	error = xchg(&global->first_error, (void *)-1);
 
@@ -1590,8 +1548,6 @@ static int igt_handle_error(void *arg)
 
 	xchg(&global->first_error, error);
 
-	mutex_lock(&gt->i915->drm.struct_mutex);
-
 	if (rq->fence.error != -EIO) {
 		pr_err("Guilty request not identified!\n");
 		err = -EINVAL;
@@ -1602,8 +1558,6 @@ err_request:
 	i915_request_put(rq);
 err_fini:
 	hang_fini(&h);
-err_unlock:
-	mutex_unlock(&gt->i915->drm.struct_mutex);
 	return err;
 }
 
@@ -1617,7 +1571,7 @@ static int __igt_atomic_reset_engine(struct intel_engine_cs *engine,
 	GEM_TRACE("i915_reset_engine(%s:%s) under %s\n",
 		  engine->name, mode, p->name);
 
-	tasklet_disable_nosync(t);
+	tasklet_disable(t);
 	p->critical_section_begin();
 
 	err = intel_engine_reset(engine, NULL);
@@ -1689,14 +1643,13 @@ static int igt_reset_engines_atomic(void *arg)
 
 	/* Check that the engines resets are usable from atomic context */
 
-	if (!intel_has_reset_engine(gt->i915))
+	if (!intel_has_reset_engine(gt))
 		return 0;
 
 	if (USES_GUC_SUBMISSION(gt->i915))
 		return 0;
 
 	igt_global_reset_lock(gt);
-	mutex_lock(&gt->i915->drm.struct_mutex);
 
 	/* Flush any requests before we get started and check basics */
 	if (!igt_force_reset(gt))
@@ -1706,7 +1659,7 @@ static int igt_reset_engines_atomic(void *arg)
 		struct intel_engine_cs *engine;
 		enum intel_engine_id id;
 
-		for_each_engine(engine, gt->i915, id) {
+		for_each_engine(engine, gt, id) {
 			err = igt_atomic_reset_engine(engine, p);
 			if (err)
 				goto out;
@@ -1716,9 +1669,7 @@ static int igt_reset_engines_atomic(void *arg)
 out:
 	/* As we poke around the guts, do a full reset before continuing. */
 	igt_force_reset(gt);
-
 unlock:
-	mutex_unlock(&gt->i915->drm.struct_mutex);
 	igt_global_reset_unlock(gt);
 
 	return err;
@@ -1743,27 +1694,19 @@ int intel_hangcheck_live_selftests(struct drm_i915_private *i915)
 	};
 	struct intel_gt *gt = &i915->gt;
 	intel_wakeref_t wakeref;
-	bool saved_hangcheck;
 	int err;
 
-	if (!intel_has_gpu_reset(gt->i915))
+	if (!intel_has_gpu_reset(gt))
 		return 0;
 
 	if (intel_gt_is_wedged(gt))
 		return -EIO; /* we're long past hope of a successful reset */
 
-	wakeref = intel_runtime_pm_get(&gt->i915->runtime_pm);
-	saved_hangcheck = fetch_and_zero(&i915_modparams.enable_hangcheck);
-	drain_delayed_work(&gt->hangcheck.work); /* flush param */
+	wakeref = intel_runtime_pm_get(gt->uncore->rpm);
 
 	err = intel_gt_live_subtests(tests, gt);
 
-	mutex_lock(&gt->i915->drm.struct_mutex);
-	igt_flush_test(gt->i915, I915_WAIT_LOCKED);
-	mutex_unlock(&gt->i915->drm.struct_mutex);
-
-	i915_modparams.enable_hangcheck = saved_hangcheck;
-	intel_runtime_pm_put(&gt->i915->runtime_pm, wakeref);
+	intel_runtime_pm_put(gt->uncore->rpm, wakeref);
 
 	return err;
 }
diff --git a/drivers/gpu/drm/i915/gt/selftest_llc.c b/drivers/gpu/drm/i915/gt/selftest_llc.c
new file mode 100644
index 000000000000..fd3770e48ac7
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/selftest_llc.c
@@ -0,0 +1,80 @@
+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2019 Intel Corporation
+ */
+
+#include "intel_pm.h" /* intel_gpu_freq() */
+#include "selftest_llc.h"
+#include "intel_rps.h"
+
+static int gen6_verify_ring_freq(struct intel_llc *llc)
+{
+	struct drm_i915_private *i915 = llc_to_gt(llc)->i915;
+	struct ia_constants consts;
+	intel_wakeref_t wakeref;
+	unsigned int gpu_freq;
+	int err = 0;
+
+	wakeref = intel_runtime_pm_get(llc_to_gt(llc)->uncore->rpm);
+
+	if (!get_ia_constants(llc, &consts)) {
+		err = -ENODEV;
+		goto out_rpm;
+	}
+
+	for (gpu_freq = consts.min_gpu_freq;
+	     gpu_freq <= consts.max_gpu_freq;
+	     gpu_freq++) {
+		struct intel_rps *rps = &llc_to_gt(llc)->rps;
+
+		unsigned int ia_freq, ring_freq, found;
+		u32 val;
+
+		calc_ia_freq(llc, gpu_freq, &consts, &ia_freq, &ring_freq);
+
+		val = gpu_freq;
+		if (sandybridge_pcode_read(i915,
+					   GEN6_PCODE_READ_MIN_FREQ_TABLE,
+					   &val, NULL)) {
+			pr_err("Failed to read freq table[%d], range [%d, %d]\n",
+			       gpu_freq, consts.min_gpu_freq, consts.max_gpu_freq);
+			err = -ENXIO;
+			break;
+		}
+
+		found = (val >> 0) & 0xff;
+		if (found != ia_freq) {
+			pr_err("Min freq table(%d/[%d, %d]):%dMHz did not match expected CPU freq, found %d, expected %d\n",
+			       gpu_freq, consts.min_gpu_freq, consts.max_gpu_freq,
+			       intel_gpu_freq(rps, gpu_freq * (INTEL_GEN(i915) >= 9 ? GEN9_FREQ_SCALER : 1)),
+			       found, ia_freq);
+			err = -EINVAL;
+			break;
+		}
+
+		found = (val >> 8) & 0xff;
+		if (found != ring_freq) {
+			pr_err("Min freq table(%d/[%d, %d]):%dMHz did not match expected ring freq, found %d, expected %d\n",
+			       gpu_freq, consts.min_gpu_freq, consts.max_gpu_freq,
+			       intel_gpu_freq(rps, gpu_freq * (INTEL_GEN(i915) >= 9 ? GEN9_FREQ_SCALER : 1)),
+			       found, ring_freq);
+			err = -EINVAL;
+			break;
+		}
+	}
+
+out_rpm:
+	intel_runtime_pm_put(llc_to_gt(llc)->uncore->rpm, wakeref);
+	return err;
+}
+
+int st_llc_verify(struct intel_llc *llc)
+{
+	int err = 0;
+
+	if (HAS_LLC(llc_to_gt(llc)->i915))
+		err = gen6_verify_ring_freq(llc);
+
+	return err;
+}
diff --git a/drivers/gpu/drm/i915/gt/selftest_llc.h b/drivers/gpu/drm/i915/gt/selftest_llc.h
new file mode 100644
index 000000000000..873f896e72f2
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/selftest_llc.h
@@ -0,0 +1,14 @@
+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2019 Intel Corporation
+ */
+
+#ifndef SELFTEST_LLC_H
+#define SELFTEST_LLC_H
+
+struct intel_llc;
+
+int st_llc_verify(struct intel_llc *llc);
+
+#endif /* SELFTEST_LLC_H */
diff --git a/drivers/gpu/drm/i915/gt/selftest_lrc.c b/drivers/gpu/drm/i915/gt/selftest_lrc.c
index d791158988d6..eb71ac2f992c 100644
--- a/drivers/gpu/drm/i915/gt/selftest_lrc.c
+++ b/drivers/gpu/drm/i915/gt/selftest_lrc.c
@@ -7,6 +7,7 @@
 #include <linux/prime_numbers.h>
 
 #include "gem/i915_gem_pm.h"
+#include "gt/intel_engine_heartbeat.h"
 #include "gt/intel_reset.h"
 
 #include "i915_selftest.h"
@@ -19,26 +20,52 @@
 #include "gem/selftests/igt_gem_utils.h"
 #include "gem/selftests/mock_context.h"
 
+#define CS_GPR(engine, n) ((engine)->mmio_base + 0x600 + (n) * 4)
+#define NUM_GPR_DW (16 * 2) /* each GPR is 2 dwords */
+
+static struct i915_vma *create_scratch(struct intel_gt *gt)
+{
+	struct drm_i915_gem_object *obj;
+	struct i915_vma *vma;
+	int err;
+
+	obj = i915_gem_object_create_internal(gt->i915, PAGE_SIZE);
+	if (IS_ERR(obj))
+		return ERR_CAST(obj);
+
+	i915_gem_object_set_cache_coherency(obj, I915_CACHING_CACHED);
+
+	vma = i915_vma_instance(obj, &gt->ggtt->vm, NULL);
+	if (IS_ERR(vma)) {
+		i915_gem_object_put(obj);
+		return vma;
+	}
+
+	err = i915_vma_pin(vma, 0, 0, PIN_GLOBAL);
+	if (err) {
+		i915_gem_object_put(obj);
+		return ERR_PTR(err);
+	}
+
+	return vma;
+}
+
 static int live_sanitycheck(void *arg)
 {
-	struct drm_i915_private *i915 = arg;
+	struct intel_gt *gt = arg;
 	struct i915_gem_engines_iter it;
 	struct i915_gem_context *ctx;
 	struct intel_context *ce;
 	struct igt_spinner spin;
-	intel_wakeref_t wakeref;
 	int err = -ENOMEM;
 
-	if (!HAS_LOGICAL_RING_CONTEXTS(i915))
+	if (!HAS_LOGICAL_RING_CONTEXTS(gt->i915))
 		return 0;
 
-	mutex_lock(&i915->drm.struct_mutex);
-	wakeref = intel_runtime_pm_get(&i915->runtime_pm);
-
-	if (igt_spinner_init(&spin, &i915->gt))
-		goto err_unlock;
+	if (igt_spinner_init(&spin, gt))
+		return -ENOMEM;
 
-	ctx = kernel_context(i915);
+	ctx = kernel_context(gt->i915);
 	if (!ctx)
 		goto err_spin;
 
@@ -55,13 +82,13 @@ static int live_sanitycheck(void *arg)
 		if (!igt_wait_for_spinner(&spin, rq)) {
 			GEM_TRACE("spinner failed to start\n");
 			GEM_TRACE_DUMP();
-			intel_gt_set_wedged(&i915->gt);
+			intel_gt_set_wedged(gt);
 			err = -EIO;
 			goto err_ctx;
 		}
 
 		igt_spinner_end(&spin);
-		if (igt_flush_test(i915, I915_WAIT_LOCKED)) {
+		if (igt_flush_test(gt->i915)) {
 			err = -EIO;
 			goto err_ctx;
 		}
@@ -73,12 +100,175 @@ err_ctx:
 	kernel_context_close(ctx);
 err_spin:
 	igt_spinner_fini(&spin);
-err_unlock:
-	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
-	mutex_unlock(&i915->drm.struct_mutex);
 	return err;
 }
 
+static int live_unlite_restore(struct intel_gt *gt, int prio)
+{
+	struct intel_engine_cs *engine;
+	struct i915_gem_context *ctx;
+	enum intel_engine_id id;
+	struct igt_spinner spin;
+	int err = -ENOMEM;
+
+	/*
+	 * Check that we can correctly context switch between 2 instances
+	 * on the same engine from the same parent context.
+	 */
+
+	if (igt_spinner_init(&spin, gt))
+		return err;
+
+	ctx = kernel_context(gt->i915);
+	if (!ctx)
+		goto err_spin;
+
+	err = 0;
+	for_each_engine(engine, gt, id) {
+		struct intel_context *ce[2] = {};
+		struct i915_request *rq[2];
+		struct igt_live_test t;
+		int n;
+
+		if (prio && !intel_engine_has_preemption(engine))
+			continue;
+
+		if (!intel_engine_can_store_dword(engine))
+			continue;
+
+		if (igt_live_test_begin(&t, gt->i915, __func__, engine->name)) {
+			err = -EIO;
+			break;
+		}
+
+		for (n = 0; n < ARRAY_SIZE(ce); n++) {
+			struct intel_context *tmp;
+
+			tmp = intel_context_create(ctx, engine);
+			if (IS_ERR(tmp)) {
+				err = PTR_ERR(tmp);
+				goto err_ce;
+			}
+
+			err = intel_context_pin(tmp);
+			if (err) {
+				intel_context_put(tmp);
+				goto err_ce;
+			}
+
+			/*
+			 * Setup the pair of contexts such that if we
+			 * lite-restore using the RING_TAIL from ce[1] it
+			 * will execute garbage from ce[0]->ring.
+			 */
+			memset(tmp->ring->vaddr,
+			       POISON_INUSE, /* IPEHR: 0x5a5a5a5a [hung!] */
+			       tmp->ring->vma->size);
+
+			ce[n] = tmp;
+		}
+		GEM_BUG_ON(!ce[1]->ring->size);
+		intel_ring_reset(ce[1]->ring, ce[1]->ring->size / 2);
+		__execlists_update_reg_state(ce[1], engine);
+
+		rq[0] = igt_spinner_create_request(&spin, ce[0], MI_ARB_CHECK);
+		if (IS_ERR(rq[0])) {
+			err = PTR_ERR(rq[0]);
+			goto err_ce;
+		}
+
+		i915_request_get(rq[0]);
+		i915_request_add(rq[0]);
+		GEM_BUG_ON(rq[0]->postfix > ce[1]->ring->emit);
+
+		if (!igt_wait_for_spinner(&spin, rq[0])) {
+			i915_request_put(rq[0]);
+			goto err_ce;
+		}
+
+		rq[1] = i915_request_create(ce[1]);
+		if (IS_ERR(rq[1])) {
+			err = PTR_ERR(rq[1]);
+			i915_request_put(rq[0]);
+			goto err_ce;
+		}
+
+		if (!prio) {
+			/*
+			 * Ensure we do the switch to ce[1] on completion.
+			 *
+			 * rq[0] is already submitted, so this should reduce
+			 * to a no-op (a wait on a request on the same engine
+			 * uses the submit fence, not the completion fence),
+			 * but it will install a dependency on rq[1] for rq[0]
+			 * that will prevent the pair being reordered by
+			 * timeslicing.
+			 */
+			i915_request_await_dma_fence(rq[1], &rq[0]->fence);
+		}
+
+		i915_request_get(rq[1]);
+		i915_request_add(rq[1]);
+		GEM_BUG_ON(rq[1]->postfix <= rq[0]->postfix);
+		i915_request_put(rq[0]);
+
+		if (prio) {
+			struct i915_sched_attr attr = {
+				.priority = prio,
+			};
+
+			/* Alternatively preempt the spinner with ce[1] */
+			engine->schedule(rq[1], &attr);
+		}
+
+		/* And switch back to ce[0] for good measure */
+		rq[0] = i915_request_create(ce[0]);
+		if (IS_ERR(rq[0])) {
+			err = PTR_ERR(rq[0]);
+			i915_request_put(rq[1]);
+			goto err_ce;
+		}
+
+		i915_request_await_dma_fence(rq[0], &rq[1]->fence);
+		i915_request_get(rq[0]);
+		i915_request_add(rq[0]);
+		GEM_BUG_ON(rq[0]->postfix > rq[1]->postfix);
+		i915_request_put(rq[1]);
+		i915_request_put(rq[0]);
+
+err_ce:
+		tasklet_kill(&engine->execlists.tasklet); /* flush submission */
+		igt_spinner_end(&spin);
+		for (n = 0; n < ARRAY_SIZE(ce); n++) {
+			if (IS_ERR_OR_NULL(ce[n]))
+				break;
+
+			intel_context_unpin(ce[n]);
+			intel_context_put(ce[n]);
+		}
+
+		if (igt_live_test_end(&t))
+			err = -EIO;
+		if (err)
+			break;
+	}
+
+	kernel_context_close(ctx);
+err_spin:
+	igt_spinner_fini(&spin);
+	return err;
+}
+
+static int live_unlite_switch(void *arg)
+{
+	return live_unlite_restore(arg, 0);
+}
+
+static int live_unlite_preempt(void *arg)
+{
+	return live_unlite_restore(arg, I915_USER_PRIORITY(I915_PRIORITY_MAX));
+}
+
 static int
 emit_semaphore_chain(struct i915_request *rq, struct i915_vma *vma, int idx)
 {
@@ -131,7 +321,13 @@ semaphore_queue(struct intel_engine_cs *engine, struct i915_vma *vma, int idx)
 	if (IS_ERR(rq))
 		goto out_ctx;
 
-	err = emit_semaphore_chain(rq, vma, idx);
+	err = 0;
+	if (rq->engine->emit_init_breadcrumb)
+		err = rq->engine->emit_init_breadcrumb(rq);
+	if (err == 0)
+		err = emit_semaphore_chain(rq, vma, idx);
+	if (err == 0)
+		i915_request_get(rq);
 	i915_request_add(rq);
 	if (err)
 		rq = ERR_PTR(err);
@@ -144,10 +340,10 @@ out_ctx:
 static int
 release_queue(struct intel_engine_cs *engine,
 	      struct i915_vma *vma,
-	      int idx)
+	      int idx, int prio)
 {
 	struct i915_sched_attr attr = {
-		.priority = I915_USER_PRIORITY(I915_PRIORITY_MAX),
+		.priority = prio,
 	};
 	struct i915_request *rq;
 	u32 *cs;
@@ -168,9 +364,15 @@ release_queue(struct intel_engine_cs *engine,
 	*cs++ = 1;
 
 	intel_ring_advance(rq, cs);
+
+	i915_request_get(rq);
 	i915_request_add(rq);
 
+	local_bh_disable();
 	engine->schedule(rq, &attr);
+	local_bh_enable(); /* kick tasklet */
+
+	i915_request_put(rq);
 
 	return 0;
 }
@@ -189,8 +391,7 @@ slice_semaphore_queue(struct intel_engine_cs *outer,
 	if (IS_ERR(head))
 		return PTR_ERR(head);
 
-	i915_request_get(head);
-	for_each_engine(engine, outer->i915, id) {
+	for_each_engine(engine, outer->gt, id) {
 		for (i = 0; i < count; i++) {
 			struct i915_request *rq;
 
@@ -199,15 +400,16 @@ slice_semaphore_queue(struct intel_engine_cs *outer,
 				err = PTR_ERR(rq);
 				goto out;
 			}
+
+			i915_request_put(rq);
 		}
 	}
 
-	err = release_queue(outer, vma, n);
+	err = release_queue(outer, vma, n, INT_MAX);
 	if (err)
 		goto out;
 
-	if (i915_request_wait(head,
-			      I915_WAIT_LOCKED,
+	if (i915_request_wait(head, 0,
 			      2 * RUNTIME_INFO(outer->i915)->num_engines * (count + 2) * (count + 3)) < 0) {
 		pr_err("Failed to slice along semaphore chain of length (%d, %d)!\n",
 		       count, n);
@@ -223,9 +425,8 @@ out:
 
 static int live_timeslice_preempt(void *arg)
 {
-	struct drm_i915_private *i915 = arg;
+	struct intel_gt *gt = arg;
 	struct drm_i915_gem_object *obj;
-	intel_wakeref_t wakeref;
 	struct i915_vma *vma;
 	void *vaddr;
 	int err = 0;
@@ -239,17 +440,14 @@ static int live_timeslice_preempt(void *arg)
 	 * need to preempt the current task and replace it with another
 	 * ready task.
 	 */
+	if (!IS_ACTIVE(CONFIG_DRM_I915_TIMESLICE_DURATION))
+		return 0;
 
-	mutex_lock(&i915->drm.struct_mutex);
-	wakeref = intel_runtime_pm_get(&i915->runtime_pm);
-
-	obj = i915_gem_object_create_internal(i915, PAGE_SIZE);
-	if (IS_ERR(obj)) {
-		err = PTR_ERR(obj);
-		goto err_unlock;
-	}
+	obj = i915_gem_object_create_internal(gt->i915, PAGE_SIZE);
+	if (IS_ERR(obj))
+		return PTR_ERR(obj);
 
-	vma = i915_vma_instance(obj, &i915->ggtt.vm, NULL);
+	vma = i915_vma_instance(obj, &gt->ggtt->vm, NULL);
 	if (IS_ERR(vma)) {
 		err = PTR_ERR(vma);
 		goto err_obj;
@@ -269,7 +467,7 @@ static int live_timeslice_preempt(void *arg)
 		struct intel_engine_cs *engine;
 		enum intel_engine_id id;
 
-		for_each_engine(engine, i915, id) {
+		for_each_engine(engine, gt, id) {
 			if (!intel_engine_has_preemption(engine))
 				continue;
 
@@ -279,7 +477,7 @@ static int live_timeslice_preempt(void *arg)
 			if (err)
 				goto err_pin;
 
-			if (igt_flush_test(i915, I915_WAIT_LOCKED)) {
+			if (igt_flush_test(gt->i915)) {
 				err = -EIO;
 				goto err_pin;
 			}
@@ -292,22 +490,168 @@ err_map:
 	i915_gem_object_unpin_map(obj);
 err_obj:
 	i915_gem_object_put(obj);
-err_unlock:
-	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
-	mutex_unlock(&i915->drm.struct_mutex);
+	return err;
+}
+
+static struct i915_request *nop_request(struct intel_engine_cs *engine)
+{
+	struct i915_request *rq;
+
+	rq = i915_request_create(engine->kernel_context);
+	if (IS_ERR(rq))
+		return rq;
+
+	i915_request_get(rq);
+	i915_request_add(rq);
 
+	return rq;
+}
+
+static void wait_for_submit(struct intel_engine_cs *engine,
+			    struct i915_request *rq)
+{
+	do {
+		cond_resched();
+		intel_engine_flush_submission(engine);
+	} while (!i915_request_is_active(rq));
+}
+
+static long timeslice_threshold(const struct intel_engine_cs *engine)
+{
+	return 2 * msecs_to_jiffies_timeout(timeslice(engine)) + 1;
+}
+
+static int live_timeslice_queue(void *arg)
+{
+	struct intel_gt *gt = arg;
+	struct drm_i915_gem_object *obj;
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
+	struct i915_vma *vma;
+	void *vaddr;
+	int err = 0;
+
+	/*
+	 * Make sure that even if ELSP[0] and ELSP[1] are filled with
+	 * timeslicing between them disabled, we *do* enable timeslicing
+	 * if the queue demands it. (Normally, we do not submit if
+	 * ELSP[1] is already occupied, so must rely on timeslicing to
+	 * eject ELSP[0] in favour of the queue.)
+	 */
+	if (!IS_ACTIVE(CONFIG_DRM_I915_TIMESLICE_DURATION))
+		return 0;
+
+	obj = i915_gem_object_create_internal(gt->i915, PAGE_SIZE);
+	if (IS_ERR(obj))
+		return PTR_ERR(obj);
+
+	vma = i915_vma_instance(obj, &gt->ggtt->vm, NULL);
+	if (IS_ERR(vma)) {
+		err = PTR_ERR(vma);
+		goto err_obj;
+	}
+
+	vaddr = i915_gem_object_pin_map(obj, I915_MAP_WC);
+	if (IS_ERR(vaddr)) {
+		err = PTR_ERR(vaddr);
+		goto err_obj;
+	}
+
+	err = i915_vma_pin(vma, 0, 0, PIN_GLOBAL);
+	if (err)
+		goto err_map;
+
+	for_each_engine(engine, gt, id) {
+		struct i915_sched_attr attr = {
+			.priority = I915_USER_PRIORITY(I915_PRIORITY_MAX),
+		};
+		struct i915_request *rq, *nop;
+
+		if (!intel_engine_has_preemption(engine))
+			continue;
+
+		memset(vaddr, 0, PAGE_SIZE);
+
+		/* ELSP[0]: semaphore wait */
+		rq = semaphore_queue(engine, vma, 0);
+		if (IS_ERR(rq)) {
+			err = PTR_ERR(rq);
+			goto err_pin;
+		}
+		engine->schedule(rq, &attr);
+		wait_for_submit(engine, rq);
+
+		/* ELSP[1]: nop request */
+		nop = nop_request(engine);
+		if (IS_ERR(nop)) {
+			err = PTR_ERR(nop);
+			i915_request_put(rq);
+			goto err_pin;
+		}
+		wait_for_submit(engine, nop);
+		i915_request_put(nop);
+
+		GEM_BUG_ON(i915_request_completed(rq));
+		GEM_BUG_ON(execlists_active(&engine->execlists) != rq);
+
+		/* Queue: semaphore signal, matching priority as semaphore */
+		err = release_queue(engine, vma, 1, effective_prio(rq));
+		if (err) {
+			i915_request_put(rq);
+			goto err_pin;
+		}
+
+		intel_engine_flush_submission(engine);
+		if (!READ_ONCE(engine->execlists.timer.expires) &&
+		    !i915_request_completed(rq)) {
+			struct drm_printer p =
+				drm_info_printer(gt->i915->drm.dev);
+
+			GEM_TRACE_ERR("%s: Failed to enable timeslicing!\n",
+				      engine->name);
+			intel_engine_dump(engine, &p,
+					  "%s\n", engine->name);
+			GEM_TRACE_DUMP();
+
+			memset(vaddr, 0xff, PAGE_SIZE);
+			err = -EINVAL;
+		}
+
+		/* Timeslice every jiffy, so within 2 we should signal */
+		if (i915_request_wait(rq, 0, timeslice_threshold(engine)) < 0) {
+			struct drm_printer p =
+				drm_info_printer(gt->i915->drm.dev);
+
+			pr_err("%s: Failed to timeslice into queue\n",
+			       engine->name);
+			intel_engine_dump(engine, &p,
+					  "%s\n", engine->name);
+
+			memset(vaddr, 0xff, PAGE_SIZE);
+			err = -EIO;
+		}
+		i915_request_put(rq);
+		if (err)
+			break;
+	}
+
+err_pin:
+	i915_vma_unpin(vma);
+err_map:
+	i915_gem_object_unpin_map(obj);
+err_obj:
+	i915_gem_object_put(obj);
 	return err;
 }
 
 static int live_busywait_preempt(void *arg)
 {
-	struct drm_i915_private *i915 = arg;
+	struct intel_gt *gt = arg;
 	struct i915_gem_context *ctx_hi, *ctx_lo;
 	struct intel_engine_cs *engine;
 	struct drm_i915_gem_object *obj;
 	struct i915_vma *vma;
 	enum intel_engine_id id;
-	intel_wakeref_t wakeref;
 	int err = -ENOMEM;
 	u32 *map;
 
@@ -316,22 +660,19 @@ static int live_busywait_preempt(void *arg)
 	 * preempt the busywaits used to synchronise between rings.
 	 */
 
-	mutex_lock(&i915->drm.struct_mutex);
-	wakeref = intel_runtime_pm_get(&i915->runtime_pm);
-
-	ctx_hi = kernel_context(i915);
+	ctx_hi = kernel_context(gt->i915);
 	if (!ctx_hi)
-		goto err_unlock;
+		return -ENOMEM;
 	ctx_hi->sched.priority =
 		I915_USER_PRIORITY(I915_CONTEXT_MAX_USER_PRIORITY);
 
-	ctx_lo = kernel_context(i915);
+	ctx_lo = kernel_context(gt->i915);
 	if (!ctx_lo)
 		goto err_ctx_hi;
 	ctx_lo->sched.priority =
 		I915_USER_PRIORITY(I915_CONTEXT_MIN_USER_PRIORITY);
 
-	obj = i915_gem_object_create_internal(i915, PAGE_SIZE);
+	obj = i915_gem_object_create_internal(gt->i915, PAGE_SIZE);
 	if (IS_ERR(obj)) {
 		err = PTR_ERR(obj);
 		goto err_ctx_lo;
@@ -343,7 +684,7 @@ static int live_busywait_preempt(void *arg)
 		goto err_obj;
 	}
 
-	vma = i915_vma_instance(obj, &i915->ggtt.vm, NULL);
+	vma = i915_vma_instance(obj, &gt->ggtt->vm, NULL);
 	if (IS_ERR(vma)) {
 		err = PTR_ERR(vma);
 		goto err_map;
@@ -353,7 +694,7 @@ static int live_busywait_preempt(void *arg)
 	if (err)
 		goto err_map;
 
-	for_each_engine(engine, i915, id) {
+	for_each_engine(engine, gt, id) {
 		struct i915_request *lo, *hi;
 		struct igt_live_test t;
 		u32 *cs;
@@ -364,7 +705,7 @@ static int live_busywait_preempt(void *arg)
 		if (!intel_engine_can_store_dword(engine))
 			continue;
 
-		if (igt_live_test_begin(&t, i915, __func__, engine->name)) {
+		if (igt_live_test_begin(&t, gt->i915, __func__, engine->name)) {
 			err = -EIO;
 			goto err_vma;
 		}
@@ -444,7 +785,7 @@ static int live_busywait_preempt(void *arg)
 		i915_request_add(hi);
 
 		if (i915_request_wait(lo, 0, HZ / 5) < 0) {
-			struct drm_printer p = drm_info_printer(i915->drm.dev);
+			struct drm_printer p = drm_info_printer(gt->i915->drm.dev);
 
 			pr_err("%s: Failed to preempt semaphore busywait!\n",
 			       engine->name);
@@ -452,7 +793,7 @@ static int live_busywait_preempt(void *arg)
 			intel_engine_dump(engine, &p, "%s\n", engine->name);
 			GEM_TRACE_DUMP();
 
-			intel_gt_set_wedged(&i915->gt);
+			intel_gt_set_wedged(gt);
 			err = -EIO;
 			goto err_vma;
 		}
@@ -475,9 +816,6 @@ err_ctx_lo:
 	kernel_context_close(ctx_lo);
 err_ctx_hi:
 	kernel_context_close(ctx_hi);
-err_unlock:
-	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
-	mutex_unlock(&i915->drm.struct_mutex);
 	return err;
 }
 
@@ -501,49 +839,45 @@ spinner_create_request(struct igt_spinner *spin,
 
 static int live_preempt(void *arg)
 {
-	struct drm_i915_private *i915 = arg;
+	struct intel_gt *gt = arg;
 	struct i915_gem_context *ctx_hi, *ctx_lo;
 	struct igt_spinner spin_hi, spin_lo;
 	struct intel_engine_cs *engine;
 	enum intel_engine_id id;
-	intel_wakeref_t wakeref;
 	int err = -ENOMEM;
 
-	if (!HAS_LOGICAL_RING_PREEMPTION(i915))
+	if (!HAS_LOGICAL_RING_PREEMPTION(gt->i915))
 		return 0;
 
-	if (!(i915->caps.scheduler & I915_SCHEDULER_CAP_PREEMPTION))
+	if (!(gt->i915->caps.scheduler & I915_SCHEDULER_CAP_PREEMPTION))
 		pr_err("Logical preemption supported, but not exposed\n");
 
-	mutex_lock(&i915->drm.struct_mutex);
-	wakeref = intel_runtime_pm_get(&i915->runtime_pm);
-
-	if (igt_spinner_init(&spin_hi, &i915->gt))
-		goto err_unlock;
+	if (igt_spinner_init(&spin_hi, gt))
+		return -ENOMEM;
 
-	if (igt_spinner_init(&spin_lo, &i915->gt))
+	if (igt_spinner_init(&spin_lo, gt))
 		goto err_spin_hi;
 
-	ctx_hi = kernel_context(i915);
+	ctx_hi = kernel_context(gt->i915);
 	if (!ctx_hi)
 		goto err_spin_lo;
 	ctx_hi->sched.priority =
 		I915_USER_PRIORITY(I915_CONTEXT_MAX_USER_PRIORITY);
 
-	ctx_lo = kernel_context(i915);
+	ctx_lo = kernel_context(gt->i915);
 	if (!ctx_lo)
 		goto err_ctx_hi;
 	ctx_lo->sched.priority =
 		I915_USER_PRIORITY(I915_CONTEXT_MIN_USER_PRIORITY);
 
-	for_each_engine(engine, i915, id) {
+	for_each_engine(engine, gt, id) {
 		struct igt_live_test t;
 		struct i915_request *rq;
 
 		if (!intel_engine_has_preemption(engine))
 			continue;
 
-		if (igt_live_test_begin(&t, i915, __func__, engine->name)) {
+		if (igt_live_test_begin(&t, gt->i915, __func__, engine->name)) {
 			err = -EIO;
 			goto err_ctx_lo;
 		}
@@ -559,7 +893,7 @@ static int live_preempt(void *arg)
 		if (!igt_wait_for_spinner(&spin_lo, rq)) {
 			GEM_TRACE("lo spinner failed to start\n");
 			GEM_TRACE_DUMP();
-			intel_gt_set_wedged(&i915->gt);
+			intel_gt_set_wedged(gt);
 			err = -EIO;
 			goto err_ctx_lo;
 		}
@@ -576,7 +910,7 @@ static int live_preempt(void *arg)
 		if (!igt_wait_for_spinner(&spin_hi, rq)) {
 			GEM_TRACE("hi spinner failed to start\n");
 			GEM_TRACE_DUMP();
-			intel_gt_set_wedged(&i915->gt);
+			intel_gt_set_wedged(gt);
 			err = -EIO;
 			goto err_ctx_lo;
 		}
@@ -599,54 +933,47 @@ err_spin_lo:
 	igt_spinner_fini(&spin_lo);
 err_spin_hi:
 	igt_spinner_fini(&spin_hi);
-err_unlock:
-	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
-	mutex_unlock(&i915->drm.struct_mutex);
 	return err;
 }
 
 static int live_late_preempt(void *arg)
 {
-	struct drm_i915_private *i915 = arg;
+	struct intel_gt *gt = arg;
 	struct i915_gem_context *ctx_hi, *ctx_lo;
 	struct igt_spinner spin_hi, spin_lo;
 	struct intel_engine_cs *engine;
 	struct i915_sched_attr attr = {};
 	enum intel_engine_id id;
-	intel_wakeref_t wakeref;
 	int err = -ENOMEM;
 
-	if (!HAS_LOGICAL_RING_PREEMPTION(i915))
+	if (!HAS_LOGICAL_RING_PREEMPTION(gt->i915))
 		return 0;
 
-	mutex_lock(&i915->drm.struct_mutex);
-	wakeref = intel_runtime_pm_get(&i915->runtime_pm);
-
-	if (igt_spinner_init(&spin_hi, &i915->gt))
-		goto err_unlock;
+	if (igt_spinner_init(&spin_hi, gt))
+		return -ENOMEM;
 
-	if (igt_spinner_init(&spin_lo, &i915->gt))
+	if (igt_spinner_init(&spin_lo, gt))
 		goto err_spin_hi;
 
-	ctx_hi = kernel_context(i915);
+	ctx_hi = kernel_context(gt->i915);
 	if (!ctx_hi)
 		goto err_spin_lo;
 
-	ctx_lo = kernel_context(i915);
+	ctx_lo = kernel_context(gt->i915);
 	if (!ctx_lo)
 		goto err_ctx_hi;
 
 	/* Make sure ctx_lo stays before ctx_hi until we trigger preemption. */
 	ctx_lo->sched.priority = I915_USER_PRIORITY(1);
 
-	for_each_engine(engine, i915, id) {
+	for_each_engine(engine, gt, id) {
 		struct igt_live_test t;
 		struct i915_request *rq;
 
 		if (!intel_engine_has_preemption(engine))
 			continue;
 
-		if (igt_live_test_begin(&t, i915, __func__, engine->name)) {
+		if (igt_live_test_begin(&t, gt->i915, __func__, engine->name)) {
 			err = -EIO;
 			goto err_ctx_lo;
 		}
@@ -705,15 +1032,12 @@ err_spin_lo:
 	igt_spinner_fini(&spin_lo);
 err_spin_hi:
 	igt_spinner_fini(&spin_hi);
-err_unlock:
-	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
-	mutex_unlock(&i915->drm.struct_mutex);
 	return err;
 
 err_wedged:
 	igt_spinner_end(&spin_hi);
 	igt_spinner_end(&spin_lo);
-	intel_gt_set_wedged(&i915->gt);
+	intel_gt_set_wedged(gt);
 	err = -EIO;
 	goto err_ctx_lo;
 }
@@ -723,14 +1047,13 @@ struct preempt_client {
 	struct i915_gem_context *ctx;
 };
 
-static int preempt_client_init(struct drm_i915_private *i915,
-			       struct preempt_client *c)
+static int preempt_client_init(struct intel_gt *gt, struct preempt_client *c)
 {
-	c->ctx = kernel_context(i915);
+	c->ctx = kernel_context(gt->i915);
 	if (!c->ctx)
 		return -ENOMEM;
 
-	if (igt_spinner_init(&c->spin, &i915->gt))
+	if (igt_spinner_init(&c->spin, gt))
 		goto err_ctx;
 
 	return 0;
@@ -748,11 +1071,10 @@ static void preempt_client_fini(struct preempt_client *c)
 
 static int live_nopreempt(void *arg)
 {
-	struct drm_i915_private *i915 = arg;
+	struct intel_gt *gt = arg;
 	struct intel_engine_cs *engine;
 	struct preempt_client a, b;
 	enum intel_engine_id id;
-	intel_wakeref_t wakeref;
 	int err = -ENOMEM;
 
 	/*
@@ -760,19 +1082,16 @@ static int live_nopreempt(void *arg)
 	 * that may be being observed and not want to be interrupted.
 	 */
 
-	if (!HAS_LOGICAL_RING_PREEMPTION(i915))
+	if (!HAS_LOGICAL_RING_PREEMPTION(gt->i915))
 		return 0;
 
-	mutex_lock(&i915->drm.struct_mutex);
-	wakeref = intel_runtime_pm_get(&i915->runtime_pm);
-
-	if (preempt_client_init(i915, &a))
-		goto err_unlock;
-	if (preempt_client_init(i915, &b))
+	if (preempt_client_init(gt, &a))
+		return -ENOMEM;
+	if (preempt_client_init(gt, &b))
 		goto err_client_a;
 	b.ctx->sched.priority = I915_USER_PRIORITY(I915_PRIORITY_MAX);
 
-	for_each_engine(engine, i915, id) {
+	for_each_engine(engine, gt, id) {
 		struct i915_request *rq_a, *rq_b;
 
 		if (!intel_engine_has_preemption(engine))
@@ -832,7 +1151,7 @@ static int live_nopreempt(void *arg)
 			goto err_wedged;
 		}
 
-		if (igt_flush_test(i915, I915_WAIT_LOCKED))
+		if (igt_flush_test(gt->i915))
 			goto err_wedged;
 	}
 
@@ -841,29 +1160,344 @@ err_client_b:
 	preempt_client_fini(&b);
 err_client_a:
 	preempt_client_fini(&a);
-err_unlock:
-	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
-	mutex_unlock(&i915->drm.struct_mutex);
 	return err;
 
 err_wedged:
 	igt_spinner_end(&b.spin);
 	igt_spinner_end(&a.spin);
-	intel_gt_set_wedged(&i915->gt);
+	intel_gt_set_wedged(gt);
 	err = -EIO;
 	goto err_client_b;
 }
 
+struct live_preempt_cancel {
+	struct intel_engine_cs *engine;
+	struct preempt_client a, b;
+};
+
+static int __cancel_active0(struct live_preempt_cancel *arg)
+{
+	struct i915_request *rq;
+	struct igt_live_test t;
+	int err;
+
+	/* Preempt cancel of ELSP0 */
+	GEM_TRACE("%s(%s)\n", __func__, arg->engine->name);
+	if (igt_live_test_begin(&t, arg->engine->i915,
+				__func__, arg->engine->name))
+		return -EIO;
+
+	clear_bit(CONTEXT_BANNED, &arg->a.ctx->flags);
+	rq = spinner_create_request(&arg->a.spin,
+				    arg->a.ctx, arg->engine,
+				    MI_ARB_CHECK);
+	if (IS_ERR(rq))
+		return PTR_ERR(rq);
+
+	i915_request_get(rq);
+	i915_request_add(rq);
+	if (!igt_wait_for_spinner(&arg->a.spin, rq)) {
+		err = -EIO;
+		goto out;
+	}
+
+	i915_gem_context_set_banned(arg->a.ctx);
+	err = intel_engine_pulse(arg->engine);
+	if (err)
+		goto out;
+
+	if (i915_request_wait(rq, 0, HZ / 5) < 0) {
+		err = -EIO;
+		goto out;
+	}
+
+	if (rq->fence.error != -EIO) {
+		pr_err("Cancelled inflight0 request did not report -EIO\n");
+		err = -EINVAL;
+		goto out;
+	}
+
+out:
+	i915_request_put(rq);
+	if (igt_live_test_end(&t))
+		err = -EIO;
+	return err;
+}
+
+static int __cancel_active1(struct live_preempt_cancel *arg)
+{
+	struct i915_request *rq[2] = {};
+	struct igt_live_test t;
+	int err;
+
+	/* Preempt cancel of ELSP1 */
+	GEM_TRACE("%s(%s)\n", __func__, arg->engine->name);
+	if (igt_live_test_begin(&t, arg->engine->i915,
+				__func__, arg->engine->name))
+		return -EIO;
+
+	clear_bit(CONTEXT_BANNED, &arg->a.ctx->flags);
+	rq[0] = spinner_create_request(&arg->a.spin,
+				       arg->a.ctx, arg->engine,
+				       MI_NOOP); /* no preemption */
+	if (IS_ERR(rq[0]))
+		return PTR_ERR(rq[0]);
+
+	i915_request_get(rq[0]);
+	i915_request_add(rq[0]);
+	if (!igt_wait_for_spinner(&arg->a.spin, rq[0])) {
+		err = -EIO;
+		goto out;
+	}
+
+	clear_bit(CONTEXT_BANNED, &arg->b.ctx->flags);
+	rq[1] = spinner_create_request(&arg->b.spin,
+				       arg->b.ctx, arg->engine,
+				       MI_ARB_CHECK);
+	if (IS_ERR(rq[1])) {
+		err = PTR_ERR(rq[1]);
+		goto out;
+	}
+
+	i915_request_get(rq[1]);
+	err = i915_request_await_dma_fence(rq[1], &rq[0]->fence);
+	i915_request_add(rq[1]);
+	if (err)
+		goto out;
+
+	i915_gem_context_set_banned(arg->b.ctx);
+	err = intel_engine_pulse(arg->engine);
+	if (err)
+		goto out;
+
+	igt_spinner_end(&arg->a.spin);
+	if (i915_request_wait(rq[1], 0, HZ / 5) < 0) {
+		err = -EIO;
+		goto out;
+	}
+
+	if (rq[0]->fence.error != 0) {
+		pr_err("Normal inflight0 request did not complete\n");
+		err = -EINVAL;
+		goto out;
+	}
+
+	if (rq[1]->fence.error != -EIO) {
+		pr_err("Cancelled inflight1 request did not report -EIO\n");
+		err = -EINVAL;
+		goto out;
+	}
+
+out:
+	i915_request_put(rq[1]);
+	i915_request_put(rq[0]);
+	if (igt_live_test_end(&t))
+		err = -EIO;
+	return err;
+}
+
+static int __cancel_queued(struct live_preempt_cancel *arg)
+{
+	struct i915_request *rq[3] = {};
+	struct igt_live_test t;
+	int err;
+
+	/* Full ELSP and one in the wings */
+	GEM_TRACE("%s(%s)\n", __func__, arg->engine->name);
+	if (igt_live_test_begin(&t, arg->engine->i915,
+				__func__, arg->engine->name))
+		return -EIO;
+
+	clear_bit(CONTEXT_BANNED, &arg->a.ctx->flags);
+	rq[0] = spinner_create_request(&arg->a.spin,
+				       arg->a.ctx, arg->engine,
+				       MI_ARB_CHECK);
+	if (IS_ERR(rq[0]))
+		return PTR_ERR(rq[0]);
+
+	i915_request_get(rq[0]);
+	i915_request_add(rq[0]);
+	if (!igt_wait_for_spinner(&arg->a.spin, rq[0])) {
+		err = -EIO;
+		goto out;
+	}
+
+	clear_bit(CONTEXT_BANNED, &arg->b.ctx->flags);
+	rq[1] = igt_request_alloc(arg->b.ctx, arg->engine);
+	if (IS_ERR(rq[1])) {
+		err = PTR_ERR(rq[1]);
+		goto out;
+	}
+
+	i915_request_get(rq[1]);
+	err = i915_request_await_dma_fence(rq[1], &rq[0]->fence);
+	i915_request_add(rq[1]);
+	if (err)
+		goto out;
+
+	rq[2] = spinner_create_request(&arg->b.spin,
+				       arg->a.ctx, arg->engine,
+				       MI_ARB_CHECK);
+	if (IS_ERR(rq[2])) {
+		err = PTR_ERR(rq[2]);
+		goto out;
+	}
+
+	i915_request_get(rq[2]);
+	err = i915_request_await_dma_fence(rq[2], &rq[1]->fence);
+	i915_request_add(rq[2]);
+	if (err)
+		goto out;
+
+	i915_gem_context_set_banned(arg->a.ctx);
+	err = intel_engine_pulse(arg->engine);
+	if (err)
+		goto out;
+
+	if (i915_request_wait(rq[2], 0, HZ / 5) < 0) {
+		err = -EIO;
+		goto out;
+	}
+
+	if (rq[0]->fence.error != -EIO) {
+		pr_err("Cancelled inflight0 request did not report -EIO\n");
+		err = -EINVAL;
+		goto out;
+	}
+
+	if (rq[1]->fence.error != 0) {
+		pr_err("Normal inflight1 request did not complete\n");
+		err = -EINVAL;
+		goto out;
+	}
+
+	if (rq[2]->fence.error != -EIO) {
+		pr_err("Cancelled queued request did not report -EIO\n");
+		err = -EINVAL;
+		goto out;
+	}
+
+out:
+	i915_request_put(rq[2]);
+	i915_request_put(rq[1]);
+	i915_request_put(rq[0]);
+	if (igt_live_test_end(&t))
+		err = -EIO;
+	return err;
+}
+
+static int __cancel_hostile(struct live_preempt_cancel *arg)
+{
+	struct i915_request *rq;
+	int err;
+
+	/* Preempt cancel non-preemptible spinner in ELSP0 */
+	if (!IS_ACTIVE(CONFIG_DRM_I915_PREEMPT_TIMEOUT))
+		return 0;
+
+	GEM_TRACE("%s(%s)\n", __func__, arg->engine->name);
+	clear_bit(CONTEXT_BANNED, &arg->a.ctx->flags);
+	rq = spinner_create_request(&arg->a.spin,
+				    arg->a.ctx, arg->engine,
+				    MI_NOOP); /* preemption disabled */
+	if (IS_ERR(rq))
+		return PTR_ERR(rq);
+
+	i915_request_get(rq);
+	i915_request_add(rq);
+	if (!igt_wait_for_spinner(&arg->a.spin, rq)) {
+		err = -EIO;
+		goto out;
+	}
+
+	i915_gem_context_set_banned(arg->a.ctx);
+	err = intel_engine_pulse(arg->engine); /* force reset */
+	if (err)
+		goto out;
+
+	if (i915_request_wait(rq, 0, HZ / 5) < 0) {
+		err = -EIO;
+		goto out;
+	}
+
+	if (rq->fence.error != -EIO) {
+		pr_err("Cancelled inflight0 request did not report -EIO\n");
+		err = -EINVAL;
+		goto out;
+	}
+
+out:
+	i915_request_put(rq);
+	if (igt_flush_test(arg->engine->i915))
+		err = -EIO;
+	return err;
+}
+
+static int live_preempt_cancel(void *arg)
+{
+	struct intel_gt *gt = arg;
+	struct live_preempt_cancel data;
+	enum intel_engine_id id;
+	int err = -ENOMEM;
+
+	/*
+	 * To cancel an inflight context, we need to first remove it from the
+	 * GPU. That sounds like preemption! Plus a little bit of bookkeeping.
+	 */
+
+	if (!HAS_LOGICAL_RING_PREEMPTION(gt->i915))
+		return 0;
+
+	if (preempt_client_init(gt, &data.a))
+		return -ENOMEM;
+	if (preempt_client_init(gt, &data.b))
+		goto err_client_a;
+
+	for_each_engine(data.engine, gt, id) {
+		if (!intel_engine_has_preemption(data.engine))
+			continue;
+
+		err = __cancel_active0(&data);
+		if (err)
+			goto err_wedged;
+
+		err = __cancel_active1(&data);
+		if (err)
+			goto err_wedged;
+
+		err = __cancel_queued(&data);
+		if (err)
+			goto err_wedged;
+
+		err = __cancel_hostile(&data);
+		if (err)
+			goto err_wedged;
+	}
+
+	err = 0;
+err_client_b:
+	preempt_client_fini(&data.b);
+err_client_a:
+	preempt_client_fini(&data.a);
+	return err;
+
+err_wedged:
+	GEM_TRACE_DUMP();
+	igt_spinner_end(&data.b.spin);
+	igt_spinner_end(&data.a.spin);
+	intel_gt_set_wedged(gt);
+	goto err_client_b;
+}
+
 static int live_suppress_self_preempt(void *arg)
 {
-	struct drm_i915_private *i915 = arg;
+	struct intel_gt *gt = arg;
 	struct intel_engine_cs *engine;
 	struct i915_sched_attr attr = {
 		.priority = I915_USER_PRIORITY(I915_PRIORITY_MAX)
 	};
 	struct preempt_client a, b;
 	enum intel_engine_id id;
-	intel_wakeref_t wakeref;
 	int err = -ENOMEM;
 
 	/*
@@ -873,30 +1507,31 @@ static int live_suppress_self_preempt(void *arg)
 	 * completion event.
 	 */
 
-	if (!HAS_LOGICAL_RING_PREEMPTION(i915))
+	if (!HAS_LOGICAL_RING_PREEMPTION(gt->i915))
 		return 0;
 
-	if (USES_GUC_SUBMISSION(i915))
+	if (USES_GUC_SUBMISSION(gt->i915))
 		return 0; /* presume black blox */
 
-	if (intel_vgpu_active(i915))
+	if (intel_vgpu_active(gt->i915))
 		return 0; /* GVT forces single port & request submission */
 
-	mutex_lock(&i915->drm.struct_mutex);
-	wakeref = intel_runtime_pm_get(&i915->runtime_pm);
-
-	if (preempt_client_init(i915, &a))
-		goto err_unlock;
-	if (preempt_client_init(i915, &b))
+	if (preempt_client_init(gt, &a))
+		return -ENOMEM;
+	if (preempt_client_init(gt, &b))
 		goto err_client_a;
 
-	for_each_engine(engine, i915, id) {
+	for_each_engine(engine, gt, id) {
 		struct i915_request *rq_a, *rq_b;
 		int depth;
 
 		if (!intel_engine_has_preemption(engine))
 			continue;
 
+		if (igt_flush_test(gt->i915))
+			goto err_wedged;
+
+		intel_engine_pm_get(engine);
 		engine->execlists.preempt_hang.count = 0;
 
 		rq_a = spinner_create_request(&a.spin,
@@ -904,12 +1539,14 @@ static int live_suppress_self_preempt(void *arg)
 					      MI_NOOP);
 		if (IS_ERR(rq_a)) {
 			err = PTR_ERR(rq_a);
+			intel_engine_pm_put(engine);
 			goto err_client_b;
 		}
 
 		i915_request_add(rq_a);
 		if (!igt_wait_for_spinner(&a.spin, rq_a)) {
 			pr_err("First client failed to start\n");
+			intel_engine_pm_put(engine);
 			goto err_wedged;
 		}
 
@@ -921,6 +1558,7 @@ static int live_suppress_self_preempt(void *arg)
 						      MI_NOOP);
 			if (IS_ERR(rq_b)) {
 				err = PTR_ERR(rq_b);
+				intel_engine_pm_put(engine);
 				goto err_client_b;
 			}
 			i915_request_add(rq_b);
@@ -931,6 +1569,7 @@ static int live_suppress_self_preempt(void *arg)
 
 			if (!igt_wait_for_spinner(&b.spin, rq_b)) {
 				pr_err("Second client failed to start\n");
+				intel_engine_pm_put(engine);
 				goto err_wedged;
 			}
 
@@ -944,11 +1583,13 @@ static int live_suppress_self_preempt(void *arg)
 			       engine->name,
 			       engine->execlists.preempt_hang.count,
 			       depth);
+			intel_engine_pm_put(engine);
 			err = -EINVAL;
 			goto err_client_b;
 		}
 
-		if (igt_flush_test(i915, I915_WAIT_LOCKED))
+		intel_engine_pm_put(engine);
+		if (igt_flush_test(gt->i915))
 			goto err_wedged;
 	}
 
@@ -957,15 +1598,12 @@ err_client_b:
 	preempt_client_fini(&b);
 err_client_a:
 	preempt_client_fini(&a);
-err_unlock:
-	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
-	mutex_unlock(&i915->drm.struct_mutex);
 	return err;
 
 err_wedged:
 	igt_spinner_end(&b.spin);
 	igt_spinner_end(&a.spin);
-	intel_gt_set_wedged(&i915->gt);
+	intel_gt_set_wedged(gt);
 	err = -EIO;
 	goto err_client_b;
 }
@@ -984,9 +1622,13 @@ static struct i915_request *dummy_request(struct intel_engine_cs *engine)
 	if (!rq)
 		return NULL;
 
-	INIT_LIST_HEAD(&rq->active_list);
 	rq->engine = engine;
 
+	spin_lock_init(&rq->lock);
+	INIT_LIST_HEAD(&rq->fence.cb_list);
+	rq->fence.lock = &rq->lock;
+	rq->fence.ops = &i915_fence_ops;
+
 	i915_sched_node_init(&rq->sched);
 
 	/* mark this request as permanently incomplete */
@@ -1021,11 +1663,10 @@ static void dummy_request_free(struct i915_request *dummy)
 
 static int live_suppress_wait_preempt(void *arg)
 {
-	struct drm_i915_private *i915 = arg;
+	struct intel_gt *gt = arg;
 	struct preempt_client client[4];
 	struct intel_engine_cs *engine;
 	enum intel_engine_id id;
-	intel_wakeref_t wakeref;
 	int err = -ENOMEM;
 	int i;
 
@@ -1035,22 +1676,19 @@ static int live_suppress_wait_preempt(void *arg)
 	 * not needlessly generate preempt-to-idle cycles.
 	 */
 
-	if (!HAS_LOGICAL_RING_PREEMPTION(i915))
+	if (!HAS_LOGICAL_RING_PREEMPTION(gt->i915))
 		return 0;
 
-	mutex_lock(&i915->drm.struct_mutex);
-	wakeref = intel_runtime_pm_get(&i915->runtime_pm);
-
-	if (preempt_client_init(i915, &client[0])) /* ELSP[0] */
-		goto err_unlock;
-	if (preempt_client_init(i915, &client[1])) /* ELSP[1] */
+	if (preempt_client_init(gt, &client[0])) /* ELSP[0] */
+		return -ENOMEM;
+	if (preempt_client_init(gt, &client[1])) /* ELSP[1] */
 		goto err_client_0;
-	if (preempt_client_init(i915, &client[2])) /* head of queue */
+	if (preempt_client_init(gt, &client[2])) /* head of queue */
 		goto err_client_1;
-	if (preempt_client_init(i915, &client[3])) /* bystander */
+	if (preempt_client_init(gt, &client[3])) /* bystander */
 		goto err_client_2;
 
-	for_each_engine(engine, i915, id) {
+	for_each_engine(engine, gt, id) {
 		int depth;
 
 		if (!intel_engine_has_preemption(engine))
@@ -1079,8 +1717,8 @@ static int live_suppress_wait_preempt(void *arg)
 				}
 
 				/* Disable NEWCLIENT promotion */
-				__i915_active_request_set(&rq[i]->timeline->last_request,
-							  dummy);
+				__i915_active_fence_set(&i915_request_timeline(rq[i])->last_request,
+							&dummy->fence);
 				i915_request_add(rq[i]);
 			}
 
@@ -1105,7 +1743,7 @@ static int live_suppress_wait_preempt(void *arg)
 			for (i = 0; i < ARRAY_SIZE(client); i++)
 				igt_spinner_end(&client[i].spin);
 
-			if (igt_flush_test(i915, I915_WAIT_LOCKED))
+			if (igt_flush_test(gt->i915))
 				goto err_wedged;
 
 			if (engine->execlists.preempt_hang.count) {
@@ -1128,26 +1766,22 @@ err_client_1:
 	preempt_client_fini(&client[1]);
 err_client_0:
 	preempt_client_fini(&client[0]);
-err_unlock:
-	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
-	mutex_unlock(&i915->drm.struct_mutex);
 	return err;
 
 err_wedged:
 	for (i = 0; i < ARRAY_SIZE(client); i++)
 		igt_spinner_end(&client[i].spin);
-	intel_gt_set_wedged(&i915->gt);
+	intel_gt_set_wedged(gt);
 	err = -EIO;
 	goto err_client_3;
 }
 
 static int live_chain_preempt(void *arg)
 {
-	struct drm_i915_private *i915 = arg;
+	struct intel_gt *gt = arg;
 	struct intel_engine_cs *engine;
 	struct preempt_client hi, lo;
 	enum intel_engine_id id;
-	intel_wakeref_t wakeref;
 	int err = -ENOMEM;
 
 	/*
@@ -1156,19 +1790,16 @@ static int live_chain_preempt(void *arg)
 	 * the previously submitted spinner in B.
 	 */
 
-	if (!HAS_LOGICAL_RING_PREEMPTION(i915))
+	if (!HAS_LOGICAL_RING_PREEMPTION(gt->i915))
 		return 0;
 
-	mutex_lock(&i915->drm.struct_mutex);
-	wakeref = intel_runtime_pm_get(&i915->runtime_pm);
-
-	if (preempt_client_init(i915, &hi))
-		goto err_unlock;
+	if (preempt_client_init(gt, &hi))
+		return -ENOMEM;
 
-	if (preempt_client_init(i915, &lo))
+	if (preempt_client_init(gt, &lo))
 		goto err_client_hi;
 
-	for_each_engine(engine, i915, id) {
+	for_each_engine(engine, gt, id) {
 		struct i915_sched_attr attr = {
 			.priority = I915_USER_PRIORITY(I915_PRIORITY_MAX),
 		};
@@ -1199,7 +1830,7 @@ static int live_chain_preempt(void *arg)
 			goto err_wedged;
 		}
 
-		if (igt_live_test_begin(&t, i915, __func__, engine->name)) {
+		if (igt_live_test_begin(&t, gt->i915, __func__, engine->name)) {
 			err = -EIO;
 			goto err_wedged;
 		}
@@ -1237,7 +1868,7 @@ static int live_chain_preempt(void *arg)
 			igt_spinner_end(&hi.spin);
 			if (i915_request_wait(rq, 0, HZ / 5) < 0) {
 				struct drm_printer p =
-					drm_info_printer(i915->drm.dev);
+					drm_info_printer(gt->i915->drm.dev);
 
 				pr_err("Failed to preempt over chain of %d\n",
 				       count);
@@ -1253,7 +1884,7 @@ static int live_chain_preempt(void *arg)
 			i915_request_add(rq);
 			if (i915_request_wait(rq, 0, HZ / 5) < 0) {
 				struct drm_printer p =
-					drm_info_printer(i915->drm.dev);
+					drm_info_printer(gt->i915->drm.dev);
 
 				pr_err("Failed to flush low priority chain of %d requests\n",
 				       count);
@@ -1274,57 +1905,50 @@ err_client_lo:
 	preempt_client_fini(&lo);
 err_client_hi:
 	preempt_client_fini(&hi);
-err_unlock:
-	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
-	mutex_unlock(&i915->drm.struct_mutex);
 	return err;
 
 err_wedged:
 	igt_spinner_end(&hi.spin);
 	igt_spinner_end(&lo.spin);
-	intel_gt_set_wedged(&i915->gt);
+	intel_gt_set_wedged(gt);
 	err = -EIO;
 	goto err_client_lo;
 }
 
 static int live_preempt_hang(void *arg)
 {
-	struct drm_i915_private *i915 = arg;
+	struct intel_gt *gt = arg;
 	struct i915_gem_context *ctx_hi, *ctx_lo;
 	struct igt_spinner spin_hi, spin_lo;
 	struct intel_engine_cs *engine;
 	enum intel_engine_id id;
-	intel_wakeref_t wakeref;
 	int err = -ENOMEM;
 
-	if (!HAS_LOGICAL_RING_PREEMPTION(i915))
+	if (!HAS_LOGICAL_RING_PREEMPTION(gt->i915))
 		return 0;
 
-	if (!intel_has_reset_engine(i915))
+	if (!intel_has_reset_engine(gt))
 		return 0;
 
-	mutex_lock(&i915->drm.struct_mutex);
-	wakeref = intel_runtime_pm_get(&i915->runtime_pm);
-
-	if (igt_spinner_init(&spin_hi, &i915->gt))
-		goto err_unlock;
+	if (igt_spinner_init(&spin_hi, gt))
+		return -ENOMEM;
 
-	if (igt_spinner_init(&spin_lo, &i915->gt))
+	if (igt_spinner_init(&spin_lo, gt))
 		goto err_spin_hi;
 
-	ctx_hi = kernel_context(i915);
+	ctx_hi = kernel_context(gt->i915);
 	if (!ctx_hi)
 		goto err_spin_lo;
 	ctx_hi->sched.priority =
 		I915_USER_PRIORITY(I915_CONTEXT_MAX_USER_PRIORITY);
 
-	ctx_lo = kernel_context(i915);
+	ctx_lo = kernel_context(gt->i915);
 	if (!ctx_lo)
 		goto err_ctx_hi;
 	ctx_lo->sched.priority =
 		I915_USER_PRIORITY(I915_CONTEXT_MIN_USER_PRIORITY);
 
-	for_each_engine(engine, i915, id) {
+	for_each_engine(engine, gt, id) {
 		struct i915_request *rq;
 
 		if (!intel_engine_has_preemption(engine))
@@ -1341,7 +1965,7 @@ static int live_preempt_hang(void *arg)
 		if (!igt_wait_for_spinner(&spin_lo, rq)) {
 			GEM_TRACE("lo spinner failed to start\n");
 			GEM_TRACE_DUMP();
-			intel_gt_set_wedged(&i915->gt);
+			intel_gt_set_wedged(gt);
 			err = -EIO;
 			goto err_ctx_lo;
 		}
@@ -1363,28 +1987,28 @@ static int live_preempt_hang(void *arg)
 						 HZ / 10)) {
 			pr_err("Preemption did not occur within timeout!");
 			GEM_TRACE_DUMP();
-			intel_gt_set_wedged(&i915->gt);
+			intel_gt_set_wedged(gt);
 			err = -EIO;
 			goto err_ctx_lo;
 		}
 
-		set_bit(I915_RESET_ENGINE + id, &i915->gt.reset.flags);
+		set_bit(I915_RESET_ENGINE + id, &gt->reset.flags);
 		intel_engine_reset(engine, NULL);
-		clear_bit(I915_RESET_ENGINE + id, &i915->gt.reset.flags);
+		clear_bit(I915_RESET_ENGINE + id, &gt->reset.flags);
 
 		engine->execlists.preempt_hang.inject_hang = false;
 
 		if (!igt_wait_for_spinner(&spin_hi, rq)) {
 			GEM_TRACE("hi spinner failed to start\n");
 			GEM_TRACE_DUMP();
-			intel_gt_set_wedged(&i915->gt);
+			intel_gt_set_wedged(gt);
 			err = -EIO;
 			goto err_ctx_lo;
 		}
 
 		igt_spinner_end(&spin_hi);
 		igt_spinner_end(&spin_lo);
-		if (igt_flush_test(i915, I915_WAIT_LOCKED)) {
+		if (igt_flush_test(gt->i915)) {
 			err = -EIO;
 			goto err_ctx_lo;
 		}
@@ -1399,9 +2023,105 @@ err_spin_lo:
 	igt_spinner_fini(&spin_lo);
 err_spin_hi:
 	igt_spinner_fini(&spin_hi);
-err_unlock:
-	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
-	mutex_unlock(&i915->drm.struct_mutex);
+	return err;
+}
+
+static int live_preempt_timeout(void *arg)
+{
+	struct intel_gt *gt = arg;
+	struct i915_gem_context *ctx_hi, *ctx_lo;
+	struct igt_spinner spin_lo;
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
+	int err = -ENOMEM;
+
+	/*
+	 * Check that we force preemption to occur by cancelling the previous
+	 * context if it refuses to yield the GPU.
+	 */
+	if (!IS_ACTIVE(CONFIG_DRM_I915_PREEMPT_TIMEOUT))
+		return 0;
+
+	if (!HAS_LOGICAL_RING_PREEMPTION(gt->i915))
+		return 0;
+
+	if (!intel_has_reset_engine(gt))
+		return 0;
+
+	if (igt_spinner_init(&spin_lo, gt))
+		return -ENOMEM;
+
+	ctx_hi = kernel_context(gt->i915);
+	if (!ctx_hi)
+		goto err_spin_lo;
+	ctx_hi->sched.priority =
+		I915_USER_PRIORITY(I915_CONTEXT_MAX_USER_PRIORITY);
+
+	ctx_lo = kernel_context(gt->i915);
+	if (!ctx_lo)
+		goto err_ctx_hi;
+	ctx_lo->sched.priority =
+		I915_USER_PRIORITY(I915_CONTEXT_MIN_USER_PRIORITY);
+
+	for_each_engine(engine, gt, id) {
+		unsigned long saved_timeout;
+		struct i915_request *rq;
+
+		if (!intel_engine_has_preemption(engine))
+			continue;
+
+		rq = spinner_create_request(&spin_lo, ctx_lo, engine,
+					    MI_NOOP); /* preemption disabled */
+		if (IS_ERR(rq)) {
+			err = PTR_ERR(rq);
+			goto err_ctx_lo;
+		}
+
+		i915_request_add(rq);
+		if (!igt_wait_for_spinner(&spin_lo, rq)) {
+			intel_gt_set_wedged(gt);
+			err = -EIO;
+			goto err_ctx_lo;
+		}
+
+		rq = igt_request_alloc(ctx_hi, engine);
+		if (IS_ERR(rq)) {
+			igt_spinner_end(&spin_lo);
+			err = PTR_ERR(rq);
+			goto err_ctx_lo;
+		}
+
+		/* Flush the previous CS ack before changing timeouts */
+		while (READ_ONCE(engine->execlists.pending[0]))
+			cpu_relax();
+
+		saved_timeout = engine->props.preempt_timeout_ms;
+		engine->props.preempt_timeout_ms = 1; /* in ms, -> 1 jiffie */
+
+		i915_request_get(rq);
+		i915_request_add(rq);
+
+		intel_engine_flush_submission(engine);
+		engine->props.preempt_timeout_ms = saved_timeout;
+
+		if (i915_request_wait(rq, 0, HZ / 10) < 0) {
+			intel_gt_set_wedged(gt);
+			i915_request_put(rq);
+			err = -ETIME;
+			goto err_ctx_lo;
+		}
+
+		igt_spinner_end(&spin_lo);
+		i915_request_put(rq);
+	}
+
+	err = 0;
+err_ctx_lo:
+	kernel_context_close(ctx_lo);
+err_ctx_hi:
+	kernel_context_close(ctx_hi);
+err_spin_lo:
+	igt_spinner_fini(&spin_lo);
 	return err;
 }
 
@@ -1416,7 +2136,7 @@ static int random_priority(struct rnd_state *rnd)
 }
 
 struct preempt_smoke {
-	struct drm_i915_private *i915;
+	struct intel_gt *gt;
 	struct i915_gem_context **contexts;
 	struct intel_engine_cs *engine;
 	struct drm_i915_gem_object *batch;
@@ -1440,7 +2160,11 @@ static int smoke_submit(struct preempt_smoke *smoke,
 	int err = 0;
 
 	if (batch) {
-		vma = i915_vma_instance(batch, ctx->vm, NULL);
+		struct i915_address_space *vm;
+
+		vm = i915_gem_context_get_vm_rcu(ctx);
+		vma = i915_vma_instance(batch, vm, NULL);
+		i915_vm_put(vm);
 		if (IS_ERR(vma))
 			return PTR_ERR(vma);
 
@@ -1489,11 +2213,9 @@ static int smoke_crescendo_thread(void *arg)
 		struct i915_gem_context *ctx = smoke_context(smoke);
 		int err;
 
-		mutex_lock(&smoke->i915->drm.struct_mutex);
 		err = smoke_submit(smoke,
 				   ctx, count % I915_PRIORITY_MAX,
 				   smoke->batch);
-		mutex_unlock(&smoke->i915->drm.struct_mutex);
 		if (err)
 			return err;
 
@@ -1514,9 +2236,7 @@ static int smoke_crescendo(struct preempt_smoke *smoke, unsigned int flags)
 	unsigned long count;
 	int err = 0;
 
-	mutex_unlock(&smoke->i915->drm.struct_mutex);
-
-	for_each_engine(engine, smoke->i915, id) {
+	for_each_engine(engine, smoke->gt, id) {
 		arg[id] = *smoke;
 		arg[id].engine = engine;
 		if (!(flags & BATCH))
@@ -1532,8 +2252,10 @@ static int smoke_crescendo(struct preempt_smoke *smoke, unsigned int flags)
 		get_task_struct(tsk[id]);
 	}
 
+	yield(); /* start all threads before we kthread_stop() */
+
 	count = 0;
-	for_each_engine(engine, smoke->i915, id) {
+	for_each_engine(engine, smoke->gt, id) {
 		int status;
 
 		if (IS_ERR_OR_NULL(tsk[id]))
@@ -1548,11 +2270,9 @@ static int smoke_crescendo(struct preempt_smoke *smoke, unsigned int flags)
 		put_task_struct(tsk[id]);
 	}
 
-	mutex_lock(&smoke->i915->drm.struct_mutex);
-
 	pr_info("Submitted %lu crescendo:%x requests across %d engines and %d contexts\n",
 		count, flags,
-		RUNTIME_INFO(smoke->i915)->num_engines, smoke->ncontext);
+		RUNTIME_INFO(smoke->gt->i915)->num_engines, smoke->ncontext);
 	return 0;
 }
 
@@ -1564,7 +2284,7 @@ static int smoke_random(struct preempt_smoke *smoke, unsigned int flags)
 
 	count = 0;
 	do {
-		for_each_engine(smoke->engine, smoke->i915, id) {
+		for_each_engine(smoke->engine, smoke->gt, id) {
 			struct i915_gem_context *ctx = smoke_context(smoke);
 			int err;
 
@@ -1580,25 +2300,24 @@ static int smoke_random(struct preempt_smoke *smoke, unsigned int flags)
 
 	pr_info("Submitted %lu random:%x requests across %d engines and %d contexts\n",
 		count, flags,
-		RUNTIME_INFO(smoke->i915)->num_engines, smoke->ncontext);
+		RUNTIME_INFO(smoke->gt->i915)->num_engines, smoke->ncontext);
 	return 0;
 }
 
 static int live_preempt_smoke(void *arg)
 {
 	struct preempt_smoke smoke = {
-		.i915 = arg,
+		.gt = arg,
 		.prng = I915_RND_STATE_INITIALIZER(i915_selftest.random_seed),
 		.ncontext = 1024,
 	};
 	const unsigned int phase[] = { 0, BATCH };
-	intel_wakeref_t wakeref;
 	struct igt_live_test t;
 	int err = -ENOMEM;
 	u32 *cs;
 	int n;
 
-	if (!HAS_LOGICAL_RING_PREEMPTION(smoke.i915))
+	if (!HAS_LOGICAL_RING_PREEMPTION(smoke.gt->i915))
 		return 0;
 
 	smoke.contexts = kmalloc_array(smoke.ncontext,
@@ -1607,13 +2326,11 @@ static int live_preempt_smoke(void *arg)
 	if (!smoke.contexts)
 		return -ENOMEM;
 
-	mutex_lock(&smoke.i915->drm.struct_mutex);
-	wakeref = intel_runtime_pm_get(&smoke.i915->runtime_pm);
-
-	smoke.batch = i915_gem_object_create_internal(smoke.i915, PAGE_SIZE);
+	smoke.batch =
+		i915_gem_object_create_internal(smoke.gt->i915, PAGE_SIZE);
 	if (IS_ERR(smoke.batch)) {
 		err = PTR_ERR(smoke.batch);
-		goto err_unlock;
+		goto err_free;
 	}
 
 	cs = i915_gem_object_pin_map(smoke.batch, I915_MAP_WB);
@@ -1627,13 +2344,13 @@ static int live_preempt_smoke(void *arg)
 	i915_gem_object_flush_map(smoke.batch);
 	i915_gem_object_unpin_map(smoke.batch);
 
-	if (igt_live_test_begin(&t, smoke.i915, __func__, "all")) {
+	if (igt_live_test_begin(&t, smoke.gt->i915, __func__, "all")) {
 		err = -EIO;
 		goto err_batch;
 	}
 
 	for (n = 0; n < smoke.ncontext; n++) {
-		smoke.contexts[n] = kernel_context(smoke.i915);
+		smoke.contexts[n] = kernel_context(smoke.gt->i915);
 		if (!smoke.contexts[n])
 			goto err_ctx;
 	}
@@ -1660,15 +2377,13 @@ err_ctx:
 
 err_batch:
 	i915_gem_object_put(smoke.batch);
-err_unlock:
-	intel_runtime_pm_put(&smoke.i915->runtime_pm, wakeref);
-	mutex_unlock(&smoke.i915->drm.struct_mutex);
+err_free:
 	kfree(smoke.contexts);
 
 	return err;
 }
 
-static int nop_virtual_engine(struct drm_i915_private *i915,
+static int nop_virtual_engine(struct intel_gt *gt,
 			      struct intel_engine_cs **siblings,
 			      unsigned int nsibling,
 			      unsigned int nctx,
@@ -1687,7 +2402,7 @@ static int nop_virtual_engine(struct drm_i915_private *i915,
 	GEM_BUG_ON(!nctx || nctx > ARRAY_SIZE(ctx));
 
 	for (n = 0; n < nctx; n++) {
-		ctx[n] = kernel_context(i915);
+		ctx[n] = kernel_context(gt->i915);
 		if (!ctx[n]) {
 			err = -ENOMEM;
 			nctx = n;
@@ -1712,7 +2427,7 @@ static int nop_virtual_engine(struct drm_i915_private *i915,
 		}
 	}
 
-	err = igt_live_test_begin(&t, i915, __func__, ve[0]->engine->name);
+	err = igt_live_test_begin(&t, gt->i915, __func__, ve[0]->engine->name);
 	if (err)
 		goto out;
 
@@ -1759,7 +2474,7 @@ static int nop_virtual_engine(struct drm_i915_private *i915,
 					  request[nc]->fence.context,
 					  request[nc]->fence.seqno);
 				GEM_TRACE_DUMP();
-				intel_gt_set_wedged(&i915->gt);
+				intel_gt_set_wedged(gt);
 				break;
 			}
 		}
@@ -1781,7 +2496,7 @@ static int nop_virtual_engine(struct drm_i915_private *i915,
 		prime, div64_u64(ktime_to_ns(times[1]), prime));
 
 out:
-	if (igt_flush_test(i915, I915_WAIT_LOCKED))
+	if (igt_flush_test(gt->i915))
 		err = -EIO;
 
 	for (nc = 0; nc < nctx; nc++) {
@@ -1794,25 +2509,22 @@ out:
 
 static int live_virtual_engine(void *arg)
 {
-	struct drm_i915_private *i915 = arg;
+	struct intel_gt *gt = arg;
 	struct intel_engine_cs *siblings[MAX_ENGINE_INSTANCE + 1];
 	struct intel_engine_cs *engine;
-	struct intel_gt *gt = &i915->gt;
 	enum intel_engine_id id;
 	unsigned int class, inst;
-	int err = -ENODEV;
+	int err;
 
-	if (USES_GUC_SUBMISSION(i915))
+	if (USES_GUC_SUBMISSION(gt->i915))
 		return 0;
 
-	mutex_lock(&i915->drm.struct_mutex);
-
-	for_each_engine(engine, i915, id) {
-		err = nop_virtual_engine(i915, &engine, 1, 1, 0);
+	for_each_engine(engine, gt, id) {
+		err = nop_virtual_engine(gt, &engine, 1, 1, 0);
 		if (err) {
 			pr_err("Failed to wrap engine %s: err=%d\n",
 			       engine->name, err);
-			goto out_unlock;
+			return err;
 		}
 	}
 
@@ -1830,23 +2542,21 @@ static int live_virtual_engine(void *arg)
 			continue;
 
 		for (n = 1; n <= nsibling + 1; n++) {
-			err = nop_virtual_engine(i915, siblings, nsibling,
+			err = nop_virtual_engine(gt, siblings, nsibling,
 						 n, 0);
 			if (err)
-				goto out_unlock;
+				return err;
 		}
 
-		err = nop_virtual_engine(i915, siblings, nsibling, n, CHAIN);
+		err = nop_virtual_engine(gt, siblings, nsibling, n, CHAIN);
 		if (err)
-			goto out_unlock;
+			return err;
 	}
 
-out_unlock:
-	mutex_unlock(&i915->drm.struct_mutex);
-	return err;
+	return 0;
 }
 
-static int mask_virtual_engine(struct drm_i915_private *i915,
+static int mask_virtual_engine(struct intel_gt *gt,
 			       struct intel_engine_cs **siblings,
 			       unsigned int nsibling)
 {
@@ -1862,7 +2572,7 @@ static int mask_virtual_engine(struct drm_i915_private *i915,
 	 * restrict it to our desired engine within the virtual engine.
 	 */
 
-	ctx = kernel_context(i915);
+	ctx = kernel_context(gt->i915);
 	if (!ctx)
 		return -ENOMEM;
 
@@ -1876,7 +2586,7 @@ static int mask_virtual_engine(struct drm_i915_private *i915,
 	if (err)
 		goto out_put;
 
-	err = igt_live_test_begin(&t, i915, __func__, ve->engine->name);
+	err = igt_live_test_begin(&t, gt->i915, __func__, ve->engine->name);
 	if (err)
 		goto out_unpin;
 
@@ -1907,7 +2617,7 @@ static int mask_virtual_engine(struct drm_i915_private *i915,
 				  request[n]->fence.context,
 				  request[n]->fence.seqno);
 			GEM_TRACE_DUMP();
-			intel_gt_set_wedged(&i915->gt);
+			intel_gt_set_wedged(gt);
 			err = -EIO;
 			goto out;
 		}
@@ -1922,11 +2632,8 @@ static int mask_virtual_engine(struct drm_i915_private *i915,
 	}
 
 	err = igt_live_test_end(&t);
-	if (err)
-		goto out;
-
 out:
-	if (igt_flush_test(i915, I915_WAIT_LOCKED))
+	if (igt_flush_test(gt->i915))
 		err = -EIO;
 
 	for (n = 0; n < nsibling; n++)
@@ -1943,17 +2650,14 @@ out_close:
 
 static int live_virtual_mask(void *arg)
 {
-	struct drm_i915_private *i915 = arg;
+	struct intel_gt *gt = arg;
 	struct intel_engine_cs *siblings[MAX_ENGINE_INSTANCE + 1];
-	struct intel_gt *gt = &i915->gt;
 	unsigned int class, inst;
-	int err = 0;
+	int err;
 
-	if (USES_GUC_SUBMISSION(i915))
+	if (USES_GUC_SUBMISSION(gt->i915))
 		return 0;
 
-	mutex_lock(&i915->drm.struct_mutex);
-
 	for (class = 0; class <= MAX_ENGINE_CLASS; class++) {
 		unsigned int nsibling;
 
@@ -1967,17 +2671,166 @@ static int live_virtual_mask(void *arg)
 		if (nsibling < 2)
 			continue;
 
-		err = mask_virtual_engine(i915, siblings, nsibling);
+		err = mask_virtual_engine(gt, siblings, nsibling);
 		if (err)
-			goto out_unlock;
+			return err;
+	}
+
+	return 0;
+}
+
+static int preserved_virtual_engine(struct intel_gt *gt,
+				    struct intel_engine_cs **siblings,
+				    unsigned int nsibling)
+{
+	struct i915_request *last = NULL;
+	struct i915_gem_context *ctx;
+	struct intel_context *ve;
+	struct i915_vma *scratch;
+	struct igt_live_test t;
+	unsigned int n;
+	int err = 0;
+	u32 *cs;
+
+	ctx = kernel_context(gt->i915);
+	if (!ctx)
+		return -ENOMEM;
+
+	scratch = create_scratch(siblings[0]->gt);
+	if (IS_ERR(scratch)) {
+		err = PTR_ERR(scratch);
+		goto out_close;
 	}
 
-out_unlock:
-	mutex_unlock(&i915->drm.struct_mutex);
+	ve = intel_execlists_create_virtual(ctx, siblings, nsibling);
+	if (IS_ERR(ve)) {
+		err = PTR_ERR(ve);
+		goto out_scratch;
+	}
+
+	err = intel_context_pin(ve);
+	if (err)
+		goto out_put;
+
+	err = igt_live_test_begin(&t, gt->i915, __func__, ve->engine->name);
+	if (err)
+		goto out_unpin;
+
+	for (n = 0; n < NUM_GPR_DW; n++) {
+		struct intel_engine_cs *engine = siblings[n % nsibling];
+		struct i915_request *rq;
+
+		rq = i915_request_create(ve);
+		if (IS_ERR(rq)) {
+			err = PTR_ERR(rq);
+			goto out_end;
+		}
+
+		i915_request_put(last);
+		last = i915_request_get(rq);
+
+		cs = intel_ring_begin(rq, 8);
+		if (IS_ERR(cs)) {
+			i915_request_add(rq);
+			err = PTR_ERR(cs);
+			goto out_end;
+		}
+
+		*cs++ = MI_STORE_REGISTER_MEM_GEN8 | MI_USE_GGTT;
+		*cs++ = CS_GPR(engine, n);
+		*cs++ = i915_ggtt_offset(scratch) + n * sizeof(u32);
+		*cs++ = 0;
+
+		*cs++ = MI_LOAD_REGISTER_IMM(1);
+		*cs++ = CS_GPR(engine, (n + 1) % NUM_GPR_DW);
+		*cs++ = n + 1;
+
+		*cs++ = MI_NOOP;
+		intel_ring_advance(rq, cs);
+
+		/* Restrict this request to run on a particular engine */
+		rq->execution_mask = engine->mask;
+		i915_request_add(rq);
+	}
+
+	if (i915_request_wait(last, 0, HZ / 5) < 0) {
+		err = -ETIME;
+		goto out_end;
+	}
+
+	cs = i915_gem_object_pin_map(scratch->obj, I915_MAP_WB);
+	if (IS_ERR(cs)) {
+		err = PTR_ERR(cs);
+		goto out_end;
+	}
+
+	for (n = 0; n < NUM_GPR_DW; n++) {
+		if (cs[n] != n) {
+			pr_err("Incorrect value[%d] found for GPR[%d]\n",
+			       cs[n], n);
+			err = -EINVAL;
+			break;
+		}
+	}
+
+	i915_gem_object_unpin_map(scratch->obj);
+
+out_end:
+	if (igt_live_test_end(&t))
+		err = -EIO;
+	i915_request_put(last);
+out_unpin:
+	intel_context_unpin(ve);
+out_put:
+	intel_context_put(ve);
+out_scratch:
+	i915_vma_unpin_and_release(&scratch, 0);
+out_close:
+	kernel_context_close(ctx);
 	return err;
 }
 
-static int bond_virtual_engine(struct drm_i915_private *i915,
+static int live_virtual_preserved(void *arg)
+{
+	struct intel_gt *gt = arg;
+	struct intel_engine_cs *siblings[MAX_ENGINE_INSTANCE + 1];
+	unsigned int class, inst;
+
+	/*
+	 * Check that the context image retains non-privileged (user) registers
+	 * from one engine to the next. For this we check that the CS_GPR
+	 * are preserved.
+	 */
+
+	if (USES_GUC_SUBMISSION(gt->i915))
+		return 0;
+
+	/* As we use CS_GPR we cannot run before they existed on all engines. */
+	if (INTEL_GEN(gt->i915) < 9)
+		return 0;
+
+	for (class = 0; class <= MAX_ENGINE_CLASS; class++) {
+		int nsibling, err;
+
+		nsibling = 0;
+		for (inst = 0; inst <= MAX_ENGINE_INSTANCE; inst++) {
+			if (!gt->engine_class[class][inst])
+				continue;
+
+			siblings[nsibling++] = gt->engine_class[class][inst];
+		}
+		if (nsibling < 2)
+			continue;
+
+		err = preserved_virtual_engine(gt, siblings, nsibling);
+		if (err)
+			return err;
+	}
+
+	return 0;
+}
+
+static int bond_virtual_engine(struct intel_gt *gt,
 			       unsigned int class,
 			       struct intel_engine_cs **siblings,
 			       unsigned int nsibling,
@@ -1993,13 +2846,13 @@ static int bond_virtual_engine(struct drm_i915_private *i915,
 
 	GEM_BUG_ON(nsibling >= ARRAY_SIZE(rq) - 1);
 
-	ctx = kernel_context(i915);
+	ctx = kernel_context(gt->i915);
 	if (!ctx)
 		return -ENOMEM;
 
 	err = 0;
 	rq[0] = ERR_PTR(-ENOMEM);
-	for_each_engine(master, i915, id) {
+	for_each_engine(master, gt, id) {
 		struct i915_sw_fence fence = {};
 
 		if (master->class == class)
@@ -2104,7 +2957,7 @@ static int bond_virtual_engine(struct drm_i915_private *i915,
 out:
 	for (n = 0; !IS_ERR(rq[n]); n++)
 		i915_request_put(rq[n]);
-	if (igt_flush_test(i915, I915_WAIT_LOCKED))
+	if (igt_flush_test(gt->i915))
 		err = -EIO;
 
 	kernel_context_close(ctx);
@@ -2121,17 +2974,14 @@ static int live_virtual_bond(void *arg)
 		{ "schedule", BOND_SCHEDULE },
 		{ },
 	};
-	struct drm_i915_private *i915 = arg;
+	struct intel_gt *gt = arg;
 	struct intel_engine_cs *siblings[MAX_ENGINE_INSTANCE + 1];
-	struct intel_gt *gt = &i915->gt;
 	unsigned int class, inst;
-	int err = 0;
+	int err;
 
-	if (USES_GUC_SUBMISSION(i915))
+	if (USES_GUC_SUBMISSION(gt->i915))
 		return 0;
 
-	mutex_lock(&i915->drm.struct_mutex);
-
 	for (class = 0; class <= MAX_ENGINE_CLASS; class++) {
 		const struct phase *p;
 		int nsibling;
@@ -2148,38 +2998,42 @@ static int live_virtual_bond(void *arg)
 			continue;
 
 		for (p = phases; p->name; p++) {
-			err = bond_virtual_engine(i915,
+			err = bond_virtual_engine(gt,
 						  class, siblings, nsibling,
 						  p->flags);
 			if (err) {
 				pr_err("%s(%s): failed class=%d, nsibling=%d, err=%d\n",
 				       __func__, p->name, class, nsibling, err);
-				goto out_unlock;
+				return err;
 			}
 		}
 	}
 
-out_unlock:
-	mutex_unlock(&i915->drm.struct_mutex);
-	return err;
+	return 0;
 }
 
 int intel_execlists_live_selftests(struct drm_i915_private *i915)
 {
 	static const struct i915_subtest tests[] = {
 		SUBTEST(live_sanitycheck),
+		SUBTEST(live_unlite_switch),
+		SUBTEST(live_unlite_preempt),
 		SUBTEST(live_timeslice_preempt),
+		SUBTEST(live_timeslice_queue),
 		SUBTEST(live_busywait_preempt),
 		SUBTEST(live_preempt),
 		SUBTEST(live_late_preempt),
 		SUBTEST(live_nopreempt),
+		SUBTEST(live_preempt_cancel),
 		SUBTEST(live_suppress_self_preempt),
 		SUBTEST(live_suppress_wait_preempt),
 		SUBTEST(live_chain_preempt),
 		SUBTEST(live_preempt_hang),
+		SUBTEST(live_preempt_timeout),
 		SUBTEST(live_preempt_smoke),
 		SUBTEST(live_virtual_engine),
 		SUBTEST(live_virtual_mask),
+		SUBTEST(live_virtual_preserved),
 		SUBTEST(live_virtual_bond),
 	};
 
@@ -2189,5 +3043,512 @@ int intel_execlists_live_selftests(struct drm_i915_private *i915)
 	if (intel_gt_is_wedged(&i915->gt))
 		return 0;
 
-	return i915_live_subtests(tests, i915);
+	return intel_gt_live_subtests(tests, &i915->gt);
+}
+
+static void hexdump(const void *buf, size_t len)
+{
+	const size_t rowsize = 8 * sizeof(u32);
+	const void *prev = NULL;
+	bool skip = false;
+	size_t pos;
+
+	for (pos = 0; pos < len; pos += rowsize) {
+		char line[128];
+
+		if (prev && !memcmp(prev, buf + pos, rowsize)) {
+			if (!skip) {
+				pr_info("*\n");
+				skip = true;
+			}
+			continue;
+		}
+
+		WARN_ON_ONCE(hex_dump_to_buffer(buf + pos, len - pos,
+						rowsize, sizeof(u32),
+						line, sizeof(line),
+						false) >= sizeof(line));
+		pr_info("[%04zx] %s\n", pos, line);
+
+		prev = buf + pos;
+		skip = false;
+	}
+}
+
+static int live_lrc_layout(void *arg)
+{
+	struct intel_gt *gt = arg;
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
+	u32 *mem;
+	int err;
+
+	/*
+	 * Check the registers offsets we use to create the initial reg state
+	 * match the layout saved by HW.
+	 */
+
+	mem = kmalloc(PAGE_SIZE, GFP_KERNEL);
+	if (!mem)
+		return -ENOMEM;
+
+	err = 0;
+	for_each_engine(engine, gt, id) {
+		u32 *hw, *lrc;
+		int dw;
+
+		if (!engine->default_state)
+			continue;
+
+		hw = i915_gem_object_pin_map(engine->default_state,
+					     I915_MAP_WB);
+		if (IS_ERR(hw)) {
+			err = PTR_ERR(hw);
+			break;
+		}
+		hw += LRC_STATE_PN * PAGE_SIZE / sizeof(*hw);
+
+		lrc = memset(mem, 0, PAGE_SIZE);
+		execlists_init_reg_state(lrc,
+					 engine->kernel_context,
+					 engine,
+					 engine->kernel_context->ring,
+					 true);
+
+		dw = 0;
+		do {
+			u32 lri = hw[dw];
+
+			if (lri == 0) {
+				dw++;
+				continue;
+			}
+
+			if ((lri & GENMASK(31, 23)) != MI_INSTR(0x22, 0)) {
+				pr_err("%s: Expected LRI command at dword %d, found %08x\n",
+				       engine->name, dw, lri);
+				err = -EINVAL;
+				break;
+			}
+
+			if (lrc[dw] != lri) {
+				pr_err("%s: LRI command mismatch at dword %d, expected %08x found %08x\n",
+				       engine->name, dw, lri, lrc[dw]);
+				err = -EINVAL;
+				break;
+			}
+
+			lri &= 0x7f;
+			lri++;
+			dw++;
+
+			while (lri) {
+				if (hw[dw] != lrc[dw]) {
+					pr_err("%s: Different registers found at dword %d, expected %x, found %x\n",
+					       engine->name, dw, hw[dw], lrc[dw]);
+					err = -EINVAL;
+					break;
+				}
+
+				/*
+				 * Skip over the actual register value as we
+				 * expect that to differ.
+				 */
+				dw += 2;
+				lri -= 2;
+			}
+		} while ((lrc[dw] & ~BIT(0)) != MI_BATCH_BUFFER_END);
+
+		if (err) {
+			pr_info("%s: HW register image:\n", engine->name);
+			hexdump(hw, PAGE_SIZE);
+
+			pr_info("%s: SW register image:\n", engine->name);
+			hexdump(lrc, PAGE_SIZE);
+		}
+
+		i915_gem_object_unpin_map(engine->default_state);
+		if (err)
+			break;
+	}
+
+	kfree(mem);
+	return err;
+}
+
+static int find_offset(const u32 *lri, u32 offset)
+{
+	int i;
+
+	for (i = 0; i < PAGE_SIZE / sizeof(u32); i++)
+		if (lri[i] == offset)
+			return i;
+
+	return -1;
+}
+
+static int live_lrc_fixed(void *arg)
+{
+	struct intel_gt *gt = arg;
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
+	int err = 0;
+
+	/*
+	 * Check the assumed register offsets match the actual locations in
+	 * the context image.
+	 */
+
+	for_each_engine(engine, gt, id) {
+		const struct {
+			u32 reg;
+			u32 offset;
+			const char *name;
+		} tbl[] = {
+			{
+				i915_mmio_reg_offset(RING_START(engine->mmio_base)),
+				CTX_RING_BUFFER_START - 1,
+				"RING_START"
+			},
+			{
+				i915_mmio_reg_offset(RING_CTL(engine->mmio_base)),
+				CTX_RING_BUFFER_CONTROL - 1,
+				"RING_CTL"
+			},
+			{
+				i915_mmio_reg_offset(RING_HEAD(engine->mmio_base)),
+				CTX_RING_HEAD - 1,
+				"RING_HEAD"
+			},
+			{
+				i915_mmio_reg_offset(RING_TAIL(engine->mmio_base)),
+				CTX_RING_TAIL - 1,
+				"RING_TAIL"
+			},
+			{
+				i915_mmio_reg_offset(RING_MI_MODE(engine->mmio_base)),
+				lrc_ring_mi_mode(engine),
+				"RING_MI_MODE"
+			},
+			{
+				engine->mmio_base + 0x110,
+				CTX_BB_STATE - 1,
+				"BB_STATE"
+			},
+			{ },
+		}, *t;
+		u32 *hw;
+
+		if (!engine->default_state)
+			continue;
+
+		hw = i915_gem_object_pin_map(engine->default_state,
+					     I915_MAP_WB);
+		if (IS_ERR(hw)) {
+			err = PTR_ERR(hw);
+			break;
+		}
+		hw += LRC_STATE_PN * PAGE_SIZE / sizeof(*hw);
+
+		for (t = tbl; t->name; t++) {
+			int dw = find_offset(hw, t->reg);
+
+			if (dw != t->offset) {
+				pr_err("%s: Offset for %s [0x%x] mismatch, found %x, expected %x\n",
+				       engine->name,
+				       t->name,
+				       t->reg,
+				       dw,
+				       t->offset);
+				err = -EINVAL;
+			}
+		}
+
+		i915_gem_object_unpin_map(engine->default_state);
+	}
+
+	return err;
+}
+
+static int __live_lrc_state(struct i915_gem_context *fixme,
+			    struct intel_engine_cs *engine,
+			    struct i915_vma *scratch)
+{
+	struct intel_context *ce;
+	struct i915_request *rq;
+	enum {
+		RING_START_IDX = 0,
+		RING_TAIL_IDX,
+		MAX_IDX
+	};
+	u32 expected[MAX_IDX];
+	u32 *cs;
+	int err;
+	int n;
+
+	ce = intel_context_create(fixme, engine);
+	if (IS_ERR(ce))
+		return PTR_ERR(ce);
+
+	err = intel_context_pin(ce);
+	if (err)
+		goto err_put;
+
+	rq = i915_request_create(ce);
+	if (IS_ERR(rq)) {
+		err = PTR_ERR(rq);
+		goto err_unpin;
+	}
+
+	cs = intel_ring_begin(rq, 4 * MAX_IDX);
+	if (IS_ERR(cs)) {
+		err = PTR_ERR(cs);
+		i915_request_add(rq);
+		goto err_unpin;
+	}
+
+	*cs++ = MI_STORE_REGISTER_MEM_GEN8 | MI_USE_GGTT;
+	*cs++ = i915_mmio_reg_offset(RING_START(engine->mmio_base));
+	*cs++ = i915_ggtt_offset(scratch) + RING_START_IDX * sizeof(u32);
+	*cs++ = 0;
+
+	expected[RING_START_IDX] = i915_ggtt_offset(ce->ring->vma);
+
+	*cs++ = MI_STORE_REGISTER_MEM_GEN8 | MI_USE_GGTT;
+	*cs++ = i915_mmio_reg_offset(RING_TAIL(engine->mmio_base));
+	*cs++ = i915_ggtt_offset(scratch) + RING_TAIL_IDX * sizeof(u32);
+	*cs++ = 0;
+
+	i915_request_get(rq);
+	i915_request_add(rq);
+
+	intel_engine_flush_submission(engine);
+	expected[RING_TAIL_IDX] = ce->ring->tail;
+
+	if (i915_request_wait(rq, 0, HZ / 5) < 0) {
+		err = -ETIME;
+		goto err_rq;
+	}
+
+	cs = i915_gem_object_pin_map(scratch->obj, I915_MAP_WB);
+	if (IS_ERR(cs)) {
+		err = PTR_ERR(cs);
+		goto err_rq;
+	}
+
+	for (n = 0; n < MAX_IDX; n++) {
+		if (cs[n] != expected[n]) {
+			pr_err("%s: Stored register[%d] value[0x%x] did not match expected[0x%x]\n",
+			       engine->name, n, cs[n], expected[n]);
+			err = -EINVAL;
+			break;
+		}
+	}
+
+	i915_gem_object_unpin_map(scratch->obj);
+
+err_rq:
+	i915_request_put(rq);
+err_unpin:
+	intel_context_unpin(ce);
+err_put:
+	intel_context_put(ce);
+	return err;
+}
+
+static int live_lrc_state(void *arg)
+{
+	struct intel_gt *gt = arg;
+	struct intel_engine_cs *engine;
+	struct i915_gem_context *fixme;
+	struct i915_vma *scratch;
+	enum intel_engine_id id;
+	int err = 0;
+
+	/*
+	 * Check the live register state matches what we expect for this
+	 * intel_context.
+	 */
+
+	fixme = kernel_context(gt->i915);
+	if (!fixme)
+		return -ENOMEM;
+
+	scratch = create_scratch(gt);
+	if (IS_ERR(scratch)) {
+		err = PTR_ERR(scratch);
+		goto out_close;
+	}
+
+	for_each_engine(engine, gt, id) {
+		err = __live_lrc_state(fixme, engine, scratch);
+		if (err)
+			break;
+	}
+
+	if (igt_flush_test(gt->i915))
+		err = -EIO;
+
+	i915_vma_unpin_and_release(&scratch, 0);
+out_close:
+	kernel_context_close(fixme);
+	return err;
+}
+
+static int gpr_make_dirty(struct intel_engine_cs *engine)
+{
+	struct i915_request *rq;
+	u32 *cs;
+	int n;
+
+	rq = i915_request_create(engine->kernel_context);
+	if (IS_ERR(rq))
+		return PTR_ERR(rq);
+
+	cs = intel_ring_begin(rq, 2 * NUM_GPR_DW + 2);
+	if (IS_ERR(cs)) {
+		i915_request_add(rq);
+		return PTR_ERR(cs);
+	}
+
+	*cs++ = MI_LOAD_REGISTER_IMM(NUM_GPR_DW);
+	for (n = 0; n < NUM_GPR_DW; n++) {
+		*cs++ = CS_GPR(engine, n);
+		*cs++ = STACK_MAGIC;
+	}
+	*cs++ = MI_NOOP;
+
+	intel_ring_advance(rq, cs);
+	i915_request_add(rq);
+
+	return 0;
+}
+
+static int __live_gpr_clear(struct i915_gem_context *fixme,
+			    struct intel_engine_cs *engine,
+			    struct i915_vma *scratch)
+{
+	struct intel_context *ce;
+	struct i915_request *rq;
+	u32 *cs;
+	int err;
+	int n;
+
+	if (INTEL_GEN(engine->i915) < 9 && engine->class != RENDER_CLASS)
+		return 0; /* GPR only on rcs0 for gen8 */
+
+	err = gpr_make_dirty(engine);
+	if (err)
+		return err;
+
+	ce = intel_context_create(fixme, engine);
+	if (IS_ERR(ce))
+		return PTR_ERR(ce);
+
+	rq = intel_context_create_request(ce);
+	if (IS_ERR(rq)) {
+		err = PTR_ERR(rq);
+		goto err_put;
+	}
+
+	cs = intel_ring_begin(rq, 4 * NUM_GPR_DW);
+	if (IS_ERR(cs)) {
+		err = PTR_ERR(cs);
+		i915_request_add(rq);
+		goto err_put;
+	}
+
+	for (n = 0; n < NUM_GPR_DW; n++) {
+		*cs++ = MI_STORE_REGISTER_MEM_GEN8 | MI_USE_GGTT;
+		*cs++ = CS_GPR(engine, n);
+		*cs++ = i915_ggtt_offset(scratch) + n * sizeof(u32);
+		*cs++ = 0;
+	}
+
+	i915_request_get(rq);
+	i915_request_add(rq);
+
+	if (i915_request_wait(rq, 0, HZ / 5) < 0) {
+		err = -ETIME;
+		goto err_rq;
+	}
+
+	cs = i915_gem_object_pin_map(scratch->obj, I915_MAP_WB);
+	if (IS_ERR(cs)) {
+		err = PTR_ERR(cs);
+		goto err_rq;
+	}
+
+	for (n = 0; n < NUM_GPR_DW; n++) {
+		if (cs[n]) {
+			pr_err("%s: GPR[%d].%s was not zero, found 0x%08x!\n",
+			       engine->name,
+			       n / 2, n & 1 ? "udw" : "ldw",
+			       cs[n]);
+			err = -EINVAL;
+			break;
+		}
+	}
+
+	i915_gem_object_unpin_map(scratch->obj);
+
+err_rq:
+	i915_request_put(rq);
+err_put:
+	intel_context_put(ce);
+	return err;
+}
+
+static int live_gpr_clear(void *arg)
+{
+	struct intel_gt *gt = arg;
+	struct intel_engine_cs *engine;
+	struct i915_gem_context *fixme;
+	struct i915_vma *scratch;
+	enum intel_engine_id id;
+	int err = 0;
+
+	/*
+	 * Check that GPR registers are cleared in new contexts as we need
+	 * to avoid leaking any information from previous contexts.
+	 */
+
+	fixme = kernel_context(gt->i915);
+	if (!fixme)
+		return -ENOMEM;
+
+	scratch = create_scratch(gt);
+	if (IS_ERR(scratch)) {
+		err = PTR_ERR(scratch);
+		goto out_close;
+	}
+
+	for_each_engine(engine, gt, id) {
+		err = __live_gpr_clear(fixme, engine, scratch);
+		if (err)
+			break;
+	}
+
+	if (igt_flush_test(gt->i915))
+		err = -EIO;
+
+	i915_vma_unpin_and_release(&scratch, 0);
+out_close:
+	kernel_context_close(fixme);
+	return err;
+}
+
+int intel_lrc_live_selftests(struct drm_i915_private *i915)
+{
+	static const struct i915_subtest tests[] = {
+		SUBTEST(live_lrc_layout),
+		SUBTEST(live_lrc_fixed),
+		SUBTEST(live_lrc_state),
+		SUBTEST(live_gpr_clear),
+	};
+
+	if (!HAS_LOGICAL_RING_CONTEXTS(i915))
+		return 0;
+
+	return intel_gt_live_subtests(tests, &i915->gt);
 }
diff --git a/drivers/gpu/drm/i915/gt/selftest_reset.c b/drivers/gpu/drm/i915/gt/selftest_reset.c
index 00a4f60cdfd5..6ad6aca315f6 100644
--- a/drivers/gpu/drm/i915/gt/selftest_reset.c
+++ b/drivers/gpu/drm/i915/gt/selftest_reset.c
@@ -17,7 +17,7 @@ static int igt_global_reset(void *arg)
 	/* Check that we can issue a global GPU reset */
 
 	igt_global_reset_lock(gt);
-	wakeref = intel_runtime_pm_get(&gt->i915->runtime_pm);
+	wakeref = intel_runtime_pm_get(gt->uncore->rpm);
 
 	reset_count = i915_reset_count(&gt->i915->gpu_error);
 
@@ -28,7 +28,7 @@ static int igt_global_reset(void *arg)
 		err = -EINVAL;
 	}
 
-	intel_runtime_pm_put(&gt->i915->runtime_pm, wakeref);
+	intel_runtime_pm_put(gt->uncore->rpm, wakeref);
 	igt_global_reset_unlock(gt);
 
 	if (intel_gt_is_wedged(gt))
@@ -45,14 +45,14 @@ static int igt_wedged_reset(void *arg)
 	/* Check that we can recover a wedged device with a GPU reset */
 
 	igt_global_reset_lock(gt);
-	wakeref = intel_runtime_pm_get(&gt->i915->runtime_pm);
+	wakeref = intel_runtime_pm_get(gt->uncore->rpm);
 
 	intel_gt_set_wedged(gt);
 
 	GEM_BUG_ON(!intel_gt_is_wedged(gt));
 	intel_gt_reset(gt, ALL_ENGINES, NULL);
 
-	intel_runtime_pm_put(&gt->i915->runtime_pm, wakeref);
+	intel_runtime_pm_put(gt->uncore->rpm, wakeref);
 	igt_global_reset_unlock(gt);
 
 	return intel_gt_is_wedged(gt) ? -EIO : 0;
@@ -112,7 +112,7 @@ static int igt_atomic_engine_reset(void *arg)
 
 	/* Check that the resets are usable from atomic context */
 
-	if (!intel_has_reset_engine(gt->i915))
+	if (!intel_has_reset_engine(gt))
 		return 0;
 
 	if (USES_GUC_SUBMISSION(gt->i915))
@@ -125,8 +125,8 @@ static int igt_atomic_engine_reset(void *arg)
 	if (!igt_force_reset(gt))
 		goto out_unlock;
 
-	for_each_engine(engine, gt->i915, id) {
-		tasklet_disable_nosync(&engine->execlists.tasklet);
+	for_each_engine(engine, gt, id) {
+		tasklet_disable(&engine->execlists.tasklet);
 		intel_engine_pm_get(engine);
 
 		for (p = igt_atomic_phases; p->name; p++) {
@@ -170,7 +170,7 @@ int intel_reset_live_selftests(struct drm_i915_private *i915)
 	};
 	struct intel_gt *gt = &i915->gt;
 
-	if (!intel_has_gpu_reset(gt->i915))
+	if (!intel_has_gpu_reset(gt))
 		return 0;
 
 	if (intel_gt_is_wedged(gt))
diff --git a/drivers/gpu/drm/i915/gt/selftest_timeline.c b/drivers/gpu/drm/i915/gt/selftest_timeline.c
index 321481403165..f04a59fe5d2c 100644
--- a/drivers/gpu/drm/i915/gt/selftest_timeline.c
+++ b/drivers/gpu/drm/i915/gt/selftest_timeline.c
@@ -6,8 +6,10 @@
 
 #include <linux/prime_numbers.h>
 
-#include "gem/i915_gem_pm.h"
+#include "intel_engine_pm.h"
 #include "intel_gt.h"
+#include "intel_gt_requests.h"
+#include "intel_ring.h"
 
 #include "../selftests/i915_random.h"
 #include "../i915_selftest.h"
@@ -34,7 +36,7 @@ static unsigned long hwsp_cacheline(struct intel_timeline *tl)
 #define CACHELINES_PER_PAGE (PAGE_SIZE / CACHELINE_BYTES)
 
 struct mock_hwsp_freelist {
-	struct drm_i915_private *i915;
+	struct intel_gt *gt;
 	struct radix_tree_root cachelines;
 	struct intel_timeline **history;
 	unsigned long count, max;
@@ -67,7 +69,7 @@ static int __mock_hwsp_timeline(struct mock_hwsp_freelist *state,
 		unsigned long cacheline;
 		int err;
 
-		tl = intel_timeline_create(&state->i915->gt, NULL);
+		tl = intel_timeline_create(state->gt, NULL);
 		if (IS_ERR(tl))
 			return PTR_ERR(tl);
 
@@ -105,6 +107,7 @@ static int __mock_hwsp_timeline(struct mock_hwsp_freelist *state,
 static int mock_hwsp_freelist(void *arg)
 {
 	struct mock_hwsp_freelist state;
+	struct drm_i915_private *i915;
 	const struct {
 		const char *name;
 		unsigned int flags;
@@ -116,12 +119,14 @@ static int mock_hwsp_freelist(void *arg)
 	unsigned int na;
 	int err = 0;
 
+	i915 = mock_gem_device();
+	if (!i915)
+		return -ENOMEM;
+
 	INIT_RADIX_TREE(&state.cachelines, GFP_KERNEL);
 	state.prng = I915_RND_STATE_INITIALIZER(i915_selftest.random_seed);
 
-	state.i915 = mock_gem_device();
-	if (!state.i915)
-		return -ENOMEM;
+	state.gt = &i915->gt;
 
 	/*
 	 * Create a bunch of timelines and check that their HWSP do not overlap.
@@ -136,7 +141,6 @@ static int mock_hwsp_freelist(void *arg)
 		goto err_put;
 	}
 
-	mutex_lock(&state.i915->drm.struct_mutex);
 	for (p = phases; p->name; p++) {
 		pr_debug("%s(%s)\n", __func__, p->name);
 		for_each_prime_number_from(na, 1, 2 * CACHELINES_PER_PAGE) {
@@ -149,10 +153,9 @@ static int mock_hwsp_freelist(void *arg)
 out:
 	for (na = 0; na < state.max; na++)
 		__mock_hwsp_record(&state, na, NULL);
-	mutex_unlock(&state.i915->drm.struct_mutex);
 	kfree(state.history);
 err_put:
-	drm_dev_put(&state.i915->drm);
+	drm_dev_put(&i915->drm);
 	return err;
 }
 
@@ -449,8 +452,6 @@ tl_write(struct intel_timeline *tl, struct intel_engine_cs *engine, u32 value)
 	struct i915_request *rq;
 	int err;
 
-	lockdep_assert_held(&tl->gt->i915->drm.struct_mutex); /* lazy rq refs */
-
 	err = intel_timeline_pin(tl);
 	if (err) {
 		rq = ERR_PTR(err);
@@ -461,10 +462,14 @@ tl_write(struct intel_timeline *tl, struct intel_engine_cs *engine, u32 value)
 	if (IS_ERR(rq))
 		goto out_unpin;
 
+	i915_request_get(rq);
+
 	err = emit_ggtt_store_dw(rq, tl->hwsp_offset, value);
 	i915_request_add(rq);
-	if (err)
+	if (err) {
+		i915_request_put(rq);
 		rq = ERR_PTR(err);
+	}
 
 out_unpin:
 	intel_timeline_unpin(tl);
@@ -475,11 +480,11 @@ out:
 }
 
 static struct intel_timeline *
-checked_intel_timeline_create(struct drm_i915_private *i915)
+checked_intel_timeline_create(struct intel_gt *gt)
 {
 	struct intel_timeline *tl;
 
-	tl = intel_timeline_create(&i915->gt, NULL);
+	tl = intel_timeline_create(gt, NULL);
 	if (IS_ERR(tl))
 		return tl;
 
@@ -496,11 +501,10 @@ checked_intel_timeline_create(struct drm_i915_private *i915)
 static int live_hwsp_engine(void *arg)
 {
 #define NUM_TIMELINES 4096
-	struct drm_i915_private *i915 = arg;
+	struct intel_gt *gt = arg;
 	struct intel_timeline **timelines;
 	struct intel_engine_cs *engine;
 	enum intel_engine_id id;
-	intel_wakeref_t wakeref;
 	unsigned long count, n;
 	int err = 0;
 
@@ -515,37 +519,40 @@ static int live_hwsp_engine(void *arg)
 	if (!timelines)
 		return -ENOMEM;
 
-	mutex_lock(&i915->drm.struct_mutex);
-	wakeref = intel_runtime_pm_get(&i915->runtime_pm);
-
 	count = 0;
-	for_each_engine(engine, i915, id) {
+	for_each_engine(engine, gt, id) {
 		if (!intel_engine_can_store_dword(engine))
 			continue;
 
+		intel_engine_pm_get(engine);
+
 		for (n = 0; n < NUM_TIMELINES; n++) {
 			struct intel_timeline *tl;
 			struct i915_request *rq;
 
-			tl = checked_intel_timeline_create(i915);
+			tl = checked_intel_timeline_create(gt);
 			if (IS_ERR(tl)) {
 				err = PTR_ERR(tl);
-				goto out;
+				break;
 			}
 
 			rq = tl_write(tl, engine, count);
 			if (IS_ERR(rq)) {
 				intel_timeline_put(tl);
 				err = PTR_ERR(rq);
-				goto out;
+				break;
 			}
 
 			timelines[count++] = tl;
+			i915_request_put(rq);
 		}
+
+		intel_engine_pm_put(engine);
+		if (err)
+			break;
 	}
 
-out:
-	if (igt_flush_test(i915, I915_WAIT_LOCKED))
+	if (igt_flush_test(gt->i915))
 		err = -EIO;
 
 	for (n = 0; n < count; n++) {
@@ -559,11 +566,7 @@ out:
 		intel_timeline_put(tl);
 	}
 
-	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
-	mutex_unlock(&i915->drm.struct_mutex);
-
 	kvfree(timelines);
-
 	return err;
 #undef NUM_TIMELINES
 }
@@ -571,11 +574,10 @@ out:
 static int live_hwsp_alternate(void *arg)
 {
 #define NUM_TIMELINES 4096
-	struct drm_i915_private *i915 = arg;
+	struct intel_gt *gt = arg;
 	struct intel_timeline **timelines;
 	struct intel_engine_cs *engine;
 	enum intel_engine_id id;
-	intel_wakeref_t wakeref;
 	unsigned long count, n;
 	int err = 0;
 
@@ -591,25 +593,25 @@ static int live_hwsp_alternate(void *arg)
 	if (!timelines)
 		return -ENOMEM;
 
-	mutex_lock(&i915->drm.struct_mutex);
-	wakeref = intel_runtime_pm_get(&i915->runtime_pm);
-
 	count = 0;
 	for (n = 0; n < NUM_TIMELINES; n++) {
-		for_each_engine(engine, i915, id) {
+		for_each_engine(engine, gt, id) {
 			struct intel_timeline *tl;
 			struct i915_request *rq;
 
 			if (!intel_engine_can_store_dword(engine))
 				continue;
 
-			tl = checked_intel_timeline_create(i915);
+			tl = checked_intel_timeline_create(gt);
 			if (IS_ERR(tl)) {
+				intel_engine_pm_put(engine);
 				err = PTR_ERR(tl);
 				goto out;
 			}
 
+			intel_engine_pm_get(engine);
 			rq = tl_write(tl, engine, count);
+			intel_engine_pm_put(engine);
 			if (IS_ERR(rq)) {
 				intel_timeline_put(tl);
 				err = PTR_ERR(rq);
@@ -617,11 +619,12 @@ static int live_hwsp_alternate(void *arg)
 			}
 
 			timelines[count++] = tl;
+			i915_request_put(rq);
 		}
 	}
 
 out:
-	if (igt_flush_test(i915, I915_WAIT_LOCKED))
+	if (igt_flush_test(gt->i915))
 		err = -EIO;
 
 	for (n = 0; n < count; n++) {
@@ -635,22 +638,17 @@ out:
 		intel_timeline_put(tl);
 	}
 
-	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
-	mutex_unlock(&i915->drm.struct_mutex);
-
 	kvfree(timelines);
-
 	return err;
 #undef NUM_TIMELINES
 }
 
 static int live_hwsp_wrap(void *arg)
 {
-	struct drm_i915_private *i915 = arg;
+	struct intel_gt *gt = arg;
 	struct intel_engine_cs *engine;
 	struct intel_timeline *tl;
 	enum intel_engine_id id;
-	intel_wakeref_t wakeref;
 	int err = 0;
 
 	/*
@@ -658,14 +656,10 @@ static int live_hwsp_wrap(void *arg)
 	 * foreign GPU references.
 	 */
 
-	mutex_lock(&i915->drm.struct_mutex);
-	wakeref = intel_runtime_pm_get(&i915->runtime_pm);
+	tl = intel_timeline_create(gt, NULL);
+	if (IS_ERR(tl))
+		return PTR_ERR(tl);
 
-	tl = intel_timeline_create(&i915->gt, NULL);
-	if (IS_ERR(tl)) {
-		err = PTR_ERR(tl);
-		goto out_rpm;
-	}
 	if (!tl->has_initial_breadcrumb || !tl->hwsp_cacheline)
 		goto out_free;
 
@@ -673,7 +667,7 @@ static int live_hwsp_wrap(void *arg)
 	if (err)
 		goto out_free;
 
-	for_each_engine(engine, i915, id) {
+	for_each_engine(engine, gt, id) {
 		const u32 *hwsp_seqno[2];
 		struct i915_request *rq;
 		u32 seqno[2];
@@ -681,7 +675,9 @@ static int live_hwsp_wrap(void *arg)
 		if (!intel_engine_can_store_dword(engine))
 			continue;
 
+		intel_engine_pm_get(engine);
 		rq = i915_request_create(engine->kernel_context);
+		intel_engine_pm_put(engine);
 		if (IS_ERR(rq)) {
 			err = PTR_ERR(rq);
 			goto out;
@@ -743,29 +739,24 @@ static int live_hwsp_wrap(void *arg)
 			goto out;
 		}
 
-		i915_retire_requests(i915); /* recycle HWSP */
+		intel_gt_retire_requests(gt); /* recycle HWSP */
 	}
 
 out:
-	if (igt_flush_test(i915, I915_WAIT_LOCKED))
+	if (igt_flush_test(gt->i915))
 		err = -EIO;
 
 	intel_timeline_unpin(tl);
 out_free:
 	intel_timeline_put(tl);
-out_rpm:
-	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
-	mutex_unlock(&i915->drm.struct_mutex);
-
 	return err;
 }
 
 static int live_hwsp_recycle(void *arg)
 {
-	struct drm_i915_private *i915 = arg;
+	struct intel_gt *gt = arg;
 	struct intel_engine_cs *engine;
 	enum intel_engine_id id;
-	intel_wakeref_t wakeref;
 	unsigned long count;
 	int err = 0;
 
@@ -775,38 +766,38 @@ static int live_hwsp_recycle(void *arg)
 	 * want to confuse ourselves or the GPU.
 	 */
 
-	mutex_lock(&i915->drm.struct_mutex);
-	wakeref = intel_runtime_pm_get(&i915->runtime_pm);
-
 	count = 0;
-	for_each_engine(engine, i915, id) {
+	for_each_engine(engine, gt, id) {
 		IGT_TIMEOUT(end_time);
 
 		if (!intel_engine_can_store_dword(engine))
 			continue;
 
+		intel_engine_pm_get(engine);
+
 		do {
 			struct intel_timeline *tl;
 			struct i915_request *rq;
 
-			tl = checked_intel_timeline_create(i915);
+			tl = checked_intel_timeline_create(gt);
 			if (IS_ERR(tl)) {
 				err = PTR_ERR(tl);
-				goto out;
+				break;
 			}
 
 			rq = tl_write(tl, engine, count);
 			if (IS_ERR(rq)) {
 				intel_timeline_put(tl);
 				err = PTR_ERR(rq);
-				goto out;
+				break;
 			}
 
 			if (i915_request_wait(rq, 0, HZ / 5) < 0) {
 				pr_err("Wait for timeline writes timed out!\n");
+				i915_request_put(rq);
 				intel_timeline_put(tl);
 				err = -EIO;
-				goto out;
+				break;
 			}
 
 			if (*tl->hwsp_seqno != count) {
@@ -815,17 +806,18 @@ static int live_hwsp_recycle(void *arg)
 				err = -EINVAL;
 			}
 
+			i915_request_put(rq);
 			intel_timeline_put(tl);
 			count++;
 
 			if (err)
-				goto out;
+				break;
 		} while (!__igt_timeout(end_time, NULL));
-	}
 
-out:
-	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
-	mutex_unlock(&i915->drm.struct_mutex);
+		intel_engine_pm_put(engine);
+		if (err)
+			break;
+	}
 
 	return err;
 }
@@ -842,5 +834,5 @@ int intel_timeline_live_selftests(struct drm_i915_private *i915)
 	if (intel_gt_is_wedged(&i915->gt))
 		return 0;
 
-	return i915_live_subtests(tests, i915);
+	return intel_gt_live_subtests(tests, &i915->gt);
 }
diff --git a/drivers/gpu/drm/i915/gt/selftest_workarounds.c b/drivers/gpu/drm/i915/gt/selftest_workarounds.c
index d06d68ac2a3b..abce6e4ec9c0 100644
--- a/drivers/gpu/drm/i915/gt/selftest_workarounds.c
+++ b/drivers/gpu/drm/i915/gt/selftest_workarounds.c
@@ -33,8 +33,32 @@ struct wa_lists {
 	} engine[I915_NUM_ENGINES];
 };
 
+static int request_add_sync(struct i915_request *rq, int err)
+{
+	i915_request_get(rq);
+	i915_request_add(rq);
+	if (i915_request_wait(rq, 0, HZ / 5) < 0)
+		err = -EIO;
+	i915_request_put(rq);
+
+	return err;
+}
+
+static int request_add_spin(struct i915_request *rq, struct igt_spinner *spin)
+{
+	int err = 0;
+
+	i915_request_get(rq);
+	i915_request_add(rq);
+	if (spin && !igt_wait_for_spinner(spin, rq))
+		err = -ETIMEDOUT;
+	i915_request_put(rq);
+
+	return err;
+}
+
 static void
-reference_lists_init(struct drm_i915_private *i915, struct wa_lists *lists)
+reference_lists_init(struct intel_gt *gt, struct wa_lists *lists)
 {
 	struct intel_engine_cs *engine;
 	enum intel_engine_id id;
@@ -42,10 +66,10 @@ reference_lists_init(struct drm_i915_private *i915, struct wa_lists *lists)
 	memset(lists, 0, sizeof(*lists));
 
 	wa_init_start(&lists->gt_wa_list, "GT_REF", "global");
-	gt_init_workarounds(i915, &lists->gt_wa_list);
+	gt_init_workarounds(gt->i915, &lists->gt_wa_list);
 	wa_init_finish(&lists->gt_wa_list);
 
-	for_each_engine(engine, i915, id) {
+	for_each_engine(engine, gt, id) {
 		struct i915_wa_list *wal = &lists->engine[id].wa_list;
 
 		wa_init_start(wal, "REF", engine->name);
@@ -59,12 +83,12 @@ reference_lists_init(struct drm_i915_private *i915, struct wa_lists *lists)
 }
 
 static void
-reference_lists_fini(struct drm_i915_private *i915, struct wa_lists *lists)
+reference_lists_fini(struct intel_gt *gt, struct wa_lists *lists)
 {
 	struct intel_engine_cs *engine;
 	enum intel_engine_id id;
 
-	for_each_engine(engine, i915, id)
+	for_each_engine(engine, gt, id)
 		intel_wa_list_free(&lists->engine[id].wa_list);
 
 	intel_wa_list_free(&lists->gt_wa_list);
@@ -191,10 +215,10 @@ static int check_whitelist(struct i915_gem_context *ctx,
 
 	err = 0;
 	i915_gem_object_lock(results);
-	intel_wedge_on_timeout(&wedge, &ctx->i915->gt, HZ / 5) /* safety net! */
+	intel_wedge_on_timeout(&wedge, engine->gt, HZ / 5) /* safety net! */
 		err = i915_gem_object_set_to_cpu_domain(results, false);
 	i915_gem_object_unlock(results);
-	if (intel_gt_is_wedged(&ctx->i915->gt))
+	if (intel_gt_is_wedged(engine->gt))
 		err = -EIO;
 	if (err)
 		goto out_put;
@@ -243,7 +267,6 @@ switch_to_scratch_context(struct intel_engine_cs *engine,
 	struct i915_gem_context *ctx;
 	struct intel_context *ce;
 	struct i915_request *rq;
-	intel_wakeref_t wakeref;
 	int err = 0;
 
 	ctx = kernel_context(engine->i915);
@@ -255,12 +278,9 @@ switch_to_scratch_context(struct intel_engine_cs *engine,
 	ce = i915_gem_context_get_engine(ctx, engine->legacy_idx);
 	GEM_BUG_ON(IS_ERR(ce));
 
-	rq = ERR_PTR(-ENODEV);
-	with_intel_runtime_pm(&engine->i915->runtime_pm, wakeref)
-		rq = igt_spinner_create_request(spin, ce, MI_NOOP);
+	rq = igt_spinner_create_request(spin, ce, MI_NOOP);
 
 	intel_context_put(ce);
-	kernel_context_close(ctx);
 
 	if (IS_ERR(rq)) {
 		spin = NULL;
@@ -268,17 +288,12 @@ switch_to_scratch_context(struct intel_engine_cs *engine,
 		goto err;
 	}
 
-	i915_request_add(rq);
-
-	if (spin && !igt_wait_for_spinner(spin, rq)) {
-		pr_err("Spinner failed to start\n");
-		err = -ETIMEDOUT;
-	}
-
+	err = request_add_spin(rq, spin);
 err:
 	if (err && spin)
 		igt_spinner_end(spin);
 
+	kernel_context_close(ctx);
 	return err;
 }
 
@@ -313,7 +328,7 @@ static int check_whitelist_across_reset(struct intel_engine_cs *engine,
 	if (err)
 		goto out_spin;
 
-	with_intel_runtime_pm(&i915->runtime_pm, wakeref)
+	with_intel_runtime_pm(engine->uncore->rpm, wakeref)
 		err = reset(engine);
 
 	igt_spinner_end(&spin);
@@ -355,6 +370,7 @@ out_ctx:
 static struct i915_vma *create_batch(struct i915_gem_context *ctx)
 {
 	struct drm_i915_gem_object *obj;
+	struct i915_address_space *vm;
 	struct i915_vma *vma;
 	int err;
 
@@ -362,7 +378,9 @@ static struct i915_vma *create_batch(struct i915_gem_context *ctx)
 	if (IS_ERR(obj))
 		return ERR_CAST(obj);
 
-	vma = i915_vma_instance(obj, ctx->vm, NULL);
+	vm = i915_gem_context_get_vm_rcu(ctx);
+	vma = i915_vma_instance(obj, vm, NULL);
+	i915_vm_put(vm);
 	if (IS_ERR(vma)) {
 		err = PTR_ERR(vma);
 		goto err_obj;
@@ -463,12 +481,15 @@ static int check_dirty_whitelist(struct i915_gem_context *ctx,
 		0xffff00ff,
 		0xffffffff,
 	};
+	struct i915_address_space *vm;
 	struct i915_vma *scratch;
 	struct i915_vma *batch;
 	int err = 0, i, v;
 	u32 *cs, *results;
 
-	scratch = create_scratch(ctx->vm, 2 * ARRAY_SIZE(values) + 1);
+	vm = i915_gem_context_get_vm_rcu(ctx);
+	scratch = create_scratch(vm, 2 * ARRAY_SIZE(values) + 1);
+	i915_vm_put(vm);
 	if (IS_ERR(scratch))
 		return PTR_ERR(scratch);
 
@@ -492,6 +513,9 @@ static int check_dirty_whitelist(struct i915_gem_context *ctx,
 
 		ro_reg = ro_register(reg);
 
+		/* Clear non priv flags */
+		reg &= RING_FORCE_TO_NONPRIV_ADDRESS_MASK;
+
 		srm = MI_STORE_REGISTER_MEM;
 		lrm = MI_LOAD_REGISTER_MEM;
 		if (INTEL_GEN(ctx->i915) >= 8)
@@ -565,6 +589,14 @@ static int check_dirty_whitelist(struct i915_gem_context *ctx,
 				goto err_request;
 		}
 
+		i915_vma_lock(batch);
+		err = i915_request_await_object(rq, batch->obj, false);
+		if (err == 0)
+			err = i915_vma_move_to_active(batch, rq, 0);
+		i915_vma_unlock(batch);
+		if (err)
+			goto err_request;
+
 		err = engine->emit_bb_start(rq,
 					    batch->node.start, PAGE_SIZE,
 					    0);
@@ -572,15 +604,11 @@ static int check_dirty_whitelist(struct i915_gem_context *ctx,
 			goto err_request;
 
 err_request:
-		i915_request_add(rq);
-		if (err)
-			goto out_batch;
-
-		if (i915_request_wait(rq, 0, HZ / 5) < 0) {
+		err = request_add_sync(rq, err);
+		if (err) {
 			pr_err("%s: Futzing %x timedout; cancelling test\n",
 			       engine->name, reg);
-			intel_gt_set_wedged(&ctx->i915->gt);
-			err = -EIO;
+			intel_gt_set_wedged(engine->gt);
 			goto out_batch;
 		}
 
@@ -668,7 +696,7 @@ out_unpin:
 			break;
 	}
 
-	if (igt_flush_test(ctx->i915, I915_WAIT_LOCKED))
+	if (igt_flush_test(ctx->i915))
 		err = -EIO;
 out_batch:
 	i915_vma_unpin_and_release(&batch, 0);
@@ -679,36 +707,29 @@ out_scratch:
 
 static int live_dirty_whitelist(void *arg)
 {
-	struct drm_i915_private *i915 = arg;
+	struct intel_gt *gt = arg;
 	struct intel_engine_cs *engine;
 	struct i915_gem_context *ctx;
 	enum intel_engine_id id;
-	intel_wakeref_t wakeref;
 	struct drm_file *file;
 	int err = 0;
 
 	/* Can the user write to the whitelisted registers? */
 
-	if (INTEL_GEN(i915) < 7) /* minimum requirement for LRI, SRM, LRM */
+	if (INTEL_GEN(gt->i915) < 7) /* minimum requirement for LRI, SRM, LRM */
 		return 0;
 
-	wakeref = intel_runtime_pm_get(&i915->runtime_pm);
+	file = mock_file(gt->i915);
+	if (IS_ERR(file))
+		return PTR_ERR(file);
 
-	mutex_unlock(&i915->drm.struct_mutex);
-	file = mock_file(i915);
-	mutex_lock(&i915->drm.struct_mutex);
-	if (IS_ERR(file)) {
-		err = PTR_ERR(file);
-		goto out_rpm;
-	}
-
-	ctx = live_context(i915, file);
+	ctx = live_context(gt->i915, file);
 	if (IS_ERR(ctx)) {
 		err = PTR_ERR(ctx);
 		goto out_file;
 	}
 
-	for_each_engine(engine, i915, id) {
+	for_each_engine(engine, gt, id) {
 		if (engine->whitelist.count == 0)
 			continue;
 
@@ -718,45 +739,43 @@ static int live_dirty_whitelist(void *arg)
 	}
 
 out_file:
-	mutex_unlock(&i915->drm.struct_mutex);
-	mock_file_free(i915, file);
-	mutex_lock(&i915->drm.struct_mutex);
-out_rpm:
-	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
+	mock_file_free(gt->i915, file);
 	return err;
 }
 
 static int live_reset_whitelist(void *arg)
 {
-	struct drm_i915_private *i915 = arg;
-	struct intel_engine_cs *engine = i915->engine[RCS0];
+	struct intel_gt *gt = arg;
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
 	int err = 0;
 
 	/* If we reset the gpu, we should not lose the RING_NONPRIV */
+	igt_global_reset_lock(gt);
 
-	if (!engine || engine->whitelist.count == 0)
-		return 0;
-
-	igt_global_reset_lock(&i915->gt);
+	for_each_engine(engine, gt, id) {
+		if (engine->whitelist.count == 0)
+			continue;
 
-	if (intel_has_reset_engine(i915)) {
-		err = check_whitelist_across_reset(engine,
-						   do_engine_reset,
-						   "engine");
-		if (err)
-			goto out;
-	}
+		if (intel_has_reset_engine(gt)) {
+			err = check_whitelist_across_reset(engine,
+							   do_engine_reset,
+							   "engine");
+			if (err)
+				goto out;
+		}
 
-	if (intel_has_gpu_reset(i915)) {
-		err = check_whitelist_across_reset(engine,
-						   do_device_reset,
-						   "device");
-		if (err)
-			goto out;
+		if (intel_has_gpu_reset(gt)) {
+			err = check_whitelist_across_reset(engine,
+							   do_device_reset,
+							   "device");
+			if (err)
+				goto out;
+		}
 	}
 
 out:
-	igt_global_reset_unlock(&i915->gt);
+	igt_global_reset_unlock(gt);
 	return err;
 }
 
@@ -772,6 +791,14 @@ static int read_whitelisted_registers(struct i915_gem_context *ctx,
 	if (IS_ERR(rq))
 		return PTR_ERR(rq);
 
+	i915_vma_lock(results);
+	err = i915_request_await_object(rq, results->obj, true);
+	if (err == 0)
+		err = i915_vma_move_to_active(results, rq, EXEC_OBJECT_WRITE);
+	i915_vma_unlock(results);
+	if (err)
+		goto err_req;
+
 	srm = MI_STORE_REGISTER_MEM;
 	if (INTEL_GEN(ctx->i915) >= 8)
 		srm++;
@@ -786,8 +813,8 @@ static int read_whitelisted_registers(struct i915_gem_context *ctx,
 		u64 offset = results->node.start + sizeof(u32) * i;
 		u32 reg = i915_mmio_reg_offset(engine->whitelist.list[i].reg);
 
-		/* Clear access permission field */
-		reg &= ~RING_FORCE_TO_NONPRIV_ACCESS_MASK;
+		/* Clear non priv flags */
+		reg &= RING_FORCE_TO_NONPRIV_ADDRESS_MASK;
 
 		*cs++ = srm;
 		*cs++ = reg;
@@ -797,12 +824,7 @@ static int read_whitelisted_registers(struct i915_gem_context *ctx,
 	intel_ring_advance(rq, cs);
 
 err_req:
-	i915_request_add(rq);
-
-	if (i915_request_wait(rq, 0, HZ / 5) < 0)
-		err = -EIO;
-
-	return err;
+	return request_add_sync(rq, err);
 }
 
 static int scrub_whitelisted_registers(struct i915_gem_context *ctx,
@@ -830,6 +852,9 @@ static int scrub_whitelisted_registers(struct i915_gem_context *ctx,
 		if (ro_register(reg))
 			continue;
 
+		/* Clear non priv flags */
+		reg &= RING_FORCE_TO_NONPRIV_ADDRESS_MASK;
+
 		*cs++ = reg;
 		*cs++ = 0xffffffff;
 	}
@@ -850,13 +875,19 @@ static int scrub_whitelisted_registers(struct i915_gem_context *ctx,
 			goto err_request;
 	}
 
+	i915_vma_lock(batch);
+	err = i915_request_await_object(rq, batch->obj, false);
+	if (err == 0)
+		err = i915_vma_move_to_active(batch, rq, 0);
+	i915_vma_unlock(batch);
+	if (err)
+		goto err_request;
+
 	/* Perform the writes from an unprivileged "user" batch */
 	err = engine->emit_bb_start(rq, batch->node.start, 0, 0);
 
 err_request:
-	i915_request_add(rq);
-	if (i915_request_wait(rq, 0, HZ / 5) < 0)
-		err = -EIO;
+	err = request_add_sync(rq, err);
 
 err_unpin:
 	i915_gem_object_unpin_map(batch->obj);
@@ -973,7 +1004,7 @@ err_a:
 
 static int live_isolated_whitelist(void *arg)
 {
-	struct drm_i915_private *i915 = arg;
+	struct intel_gt *gt = arg;
 	struct {
 		struct i915_gem_context *ctx;
 		struct i915_vma *scratch[2];
@@ -987,40 +1018,46 @@ static int live_isolated_whitelist(void *arg)
 	 * invisible to a second context.
 	 */
 
-	if (!intel_engines_has_context_isolation(i915))
-		return 0;
-
-	if (!i915->kernel_context->vm)
+	if (!intel_engines_has_context_isolation(gt->i915))
 		return 0;
 
 	for (i = 0; i < ARRAY_SIZE(client); i++) {
+		struct i915_address_space *vm;
 		struct i915_gem_context *c;
 
-		c = kernel_context(i915);
+		c = kernel_context(gt->i915);
 		if (IS_ERR(c)) {
 			err = PTR_ERR(c);
 			goto err;
 		}
 
-		client[i].scratch[0] = create_scratch(c->vm, 1024);
+		vm = i915_gem_context_get_vm_rcu(c);
+
+		client[i].scratch[0] = create_scratch(vm, 1024);
 		if (IS_ERR(client[i].scratch[0])) {
 			err = PTR_ERR(client[i].scratch[0]);
+			i915_vm_put(vm);
 			kernel_context_close(c);
 			goto err;
 		}
 
-		client[i].scratch[1] = create_scratch(c->vm, 1024);
+		client[i].scratch[1] = create_scratch(vm, 1024);
 		if (IS_ERR(client[i].scratch[1])) {
 			err = PTR_ERR(client[i].scratch[1]);
 			i915_vma_unpin_and_release(&client[i].scratch[0], 0);
+			i915_vm_put(vm);
 			kernel_context_close(c);
 			goto err;
 		}
 
 		client[i].ctx = c;
+		i915_vm_put(vm);
 	}
 
-	for_each_engine(engine, i915, id) {
+	for_each_engine(engine, gt, id) {
+		if (!engine->kernel_context->vm)
+			continue;
+
 		if (!whitelist_writable_count(engine))
 			continue;
 
@@ -1074,7 +1111,7 @@ err:
 		kernel_context_close(client[i].ctx);
 	}
 
-	if (igt_flush_test(i915, I915_WAIT_LOCKED))
+	if (igt_flush_test(gt->i915))
 		err = -EIO;
 
 	return err;
@@ -1109,16 +1146,16 @@ verify_wa_lists(struct i915_gem_context *ctx, struct wa_lists *lists,
 static int
 live_gpu_reset_workarounds(void *arg)
 {
-	struct drm_i915_private *i915 = arg;
+	struct intel_gt *gt = arg;
 	struct i915_gem_context *ctx;
 	intel_wakeref_t wakeref;
 	struct wa_lists lists;
 	bool ok;
 
-	if (!intel_has_gpu_reset(i915))
+	if (!intel_has_gpu_reset(gt))
 		return 0;
 
-	ctx = kernel_context(i915);
+	ctx = kernel_context(gt->i915);
 	if (IS_ERR(ctx))
 		return PTR_ERR(ctx);
 
@@ -1126,25 +1163,25 @@ live_gpu_reset_workarounds(void *arg)
 
 	pr_info("Verifying after GPU reset...\n");
 
-	igt_global_reset_lock(&i915->gt);
-	wakeref = intel_runtime_pm_get(&i915->runtime_pm);
+	igt_global_reset_lock(gt);
+	wakeref = intel_runtime_pm_get(gt->uncore->rpm);
 
-	reference_lists_init(i915, &lists);
+	reference_lists_init(gt, &lists);
 
 	ok = verify_wa_lists(ctx, &lists, "before reset");
 	if (!ok)
 		goto out;
 
-	intel_gt_reset(&i915->gt, ALL_ENGINES, "live_workarounds");
+	intel_gt_reset(gt, ALL_ENGINES, "live_workarounds");
 
 	ok = verify_wa_lists(ctx, &lists, "after reset");
 
 out:
 	i915_gem_context_unlock_engines(ctx);
 	kernel_context_close(ctx);
-	reference_lists_fini(i915, &lists);
-	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
-	igt_global_reset_unlock(&i915->gt);
+	reference_lists_fini(gt, &lists);
+	intel_runtime_pm_put(gt->uncore->rpm, wakeref);
+	igt_global_reset_unlock(gt);
 
 	return ok ? 0 : -ESRCH;
 }
@@ -1152,7 +1189,7 @@ out:
 static int
 live_engine_reset_workarounds(void *arg)
 {
-	struct drm_i915_private *i915 = arg;
+	struct intel_gt *gt = arg;
 	struct i915_gem_engines_iter it;
 	struct i915_gem_context *ctx;
 	struct intel_context *ce;
@@ -1162,17 +1199,17 @@ live_engine_reset_workarounds(void *arg)
 	struct wa_lists lists;
 	int ret = 0;
 
-	if (!intel_has_reset_engine(i915))
+	if (!intel_has_reset_engine(gt))
 		return 0;
 
-	ctx = kernel_context(i915);
+	ctx = kernel_context(gt->i915);
 	if (IS_ERR(ctx))
 		return PTR_ERR(ctx);
 
-	igt_global_reset_lock(&i915->gt);
-	wakeref = intel_runtime_pm_get(&i915->runtime_pm);
+	igt_global_reset_lock(gt);
+	wakeref = intel_runtime_pm_get(gt->uncore->rpm);
 
-	reference_lists_init(i915, &lists);
+	reference_lists_init(gt, &lists);
 
 	for_each_gem_engine(ce, i915_gem_context_lock_engines(ctx), it) {
 		struct intel_engine_cs *engine = ce->engine;
@@ -1205,12 +1242,10 @@ live_engine_reset_workarounds(void *arg)
 			goto err;
 		}
 
-		i915_request_add(rq);
-
-		if (!igt_wait_for_spinner(&spin, rq)) {
+		ret = request_add_spin(rq, &spin);
+		if (ret) {
 			pr_err("Spinner failed to start\n");
 			igt_spinner_fini(&spin);
-			ret = -ETIMEDOUT;
 			goto err;
 		}
 
@@ -1227,12 +1262,12 @@ live_engine_reset_workarounds(void *arg)
 	}
 err:
 	i915_gem_context_unlock_engines(ctx);
-	reference_lists_fini(i915, &lists);
-	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
-	igt_global_reset_unlock(&i915->gt);
+	reference_lists_fini(gt, &lists);
+	intel_runtime_pm_put(gt->uncore->rpm, wakeref);
+	igt_global_reset_unlock(gt);
 	kernel_context_close(ctx);
 
-	igt_flush_test(i915, I915_WAIT_LOCKED);
+	igt_flush_test(gt->i915);
 
 	return ret;
 }
@@ -1246,14 +1281,9 @@ int intel_workarounds_live_selftests(struct drm_i915_private *i915)
 		SUBTEST(live_gpu_reset_workarounds),
 		SUBTEST(live_engine_reset_workarounds),
 	};
-	int err;
 
 	if (intel_gt_is_wedged(&i915->gt))
 		return 0;
 
-	mutex_lock(&i915->drm.struct_mutex);
-	err = i915_subtests(tests, i915);
-	mutex_unlock(&i915->drm.struct_mutex);
-
-	return err;
+	return intel_gt_live_subtests(tests, &i915->gt);
 }
diff --git a/drivers/gpu/drm/i915/gt/selftests/mock_timeline.c b/drivers/gpu/drm/i915/gt/selftests/mock_timeline.c
index 598170efcaf6..2a77c051f36a 100644
--- a/drivers/gpu/drm/i915/gt/selftests/mock_timeline.c
+++ b/drivers/gpu/drm/i915/gt/selftests/mock_timeline.c
@@ -15,7 +15,7 @@ void mock_timeline_init(struct intel_timeline *timeline, u64 context)
 
 	mutex_init(&timeline->mutex);
 
-	INIT_ACTIVE_REQUEST(&timeline->last_request, &timeline->mutex);
+	INIT_ACTIVE_FENCE(&timeline->last_request, &timeline->mutex);
 	INIT_LIST_HEAD(&timeline->requests);
 
 	i915_syncmap_init(&timeline->sync);
diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc.c b/drivers/gpu/drm/i915/gt/uc/intel_guc.c
index 249c747e9756..3ee4a4e7689d 100644
--- a/drivers/gpu/drm/i915/gt/uc/intel_guc.c
+++ b/drivers/gpu/drm/i915/gt/uc/intel_guc.c
@@ -4,11 +4,34 @@
  */
 
 #include "gt/intel_gt.h"
+#include "gt/intel_gt_irq.h"
+#include "gt/intel_gt_pm_irq.h"
 #include "intel_guc.h"
 #include "intel_guc_ads.h"
 #include "intel_guc_submission.h"
 #include "i915_drv.h"
 
+/**
+ * DOC: GuC
+ *
+ * The GuC is a microcontroller inside the GT HW, introduced in gen9. The GuC is
+ * designed to offload some of the functionality usually performed by the host
+ * driver; currently the main operations it can take care of are:
+ *
+ * - Authentication of the HuC, which is required to fully enable HuC usage.
+ * - Low latency graphics context scheduling (a.k.a. GuC submission).
+ * - GT Power management.
+ *
+ * The enable_guc module parameter can be used to select which of those
+ * operations to enable within GuC. Note that not all the operations are
+ * supported on all gen9+ platforms.
+ *
+ * Enabling the GuC is not mandatory and therefore the firmware is only loaded
+ * if at least one of the operations is selected. However, not loading the GuC
+ * might result in the loss of some features that do require the GuC (currently
+ * just the HuC, but more are expected to land in the future).
+ */
+
 static void gen8_guc_raise_irq(struct intel_guc *guc)
 {
 	struct intel_gt *gt = guc_to_gt(guc);
@@ -56,6 +79,93 @@ void intel_guc_init_send_regs(struct intel_guc *guc)
 	guc->send_regs.fw_domains = fw_domains;
 }
 
+static void gen9_reset_guc_interrupts(struct intel_guc *guc)
+{
+	struct intel_gt *gt = guc_to_gt(guc);
+
+	assert_rpm_wakelock_held(&gt->i915->runtime_pm);
+
+	spin_lock_irq(&gt->irq_lock);
+	gen6_gt_pm_reset_iir(gt, gt->pm_guc_events);
+	spin_unlock_irq(&gt->irq_lock);
+}
+
+static void gen9_enable_guc_interrupts(struct intel_guc *guc)
+{
+	struct intel_gt *gt = guc_to_gt(guc);
+
+	assert_rpm_wakelock_held(&gt->i915->runtime_pm);
+
+	spin_lock_irq(&gt->irq_lock);
+	if (!guc->interrupts.enabled) {
+		WARN_ON_ONCE(intel_uncore_read(gt->uncore, GEN8_GT_IIR(2)) &
+			     gt->pm_guc_events);
+		guc->interrupts.enabled = true;
+		gen6_gt_pm_enable_irq(gt, gt->pm_guc_events);
+	}
+	spin_unlock_irq(&gt->irq_lock);
+}
+
+static void gen9_disable_guc_interrupts(struct intel_guc *guc)
+{
+	struct intel_gt *gt = guc_to_gt(guc);
+
+	assert_rpm_wakelock_held(&gt->i915->runtime_pm);
+
+	spin_lock_irq(&gt->irq_lock);
+	guc->interrupts.enabled = false;
+
+	gen6_gt_pm_disable_irq(gt, gt->pm_guc_events);
+
+	spin_unlock_irq(&gt->irq_lock);
+	intel_synchronize_irq(gt->i915);
+
+	gen9_reset_guc_interrupts(guc);
+}
+
+static void gen11_reset_guc_interrupts(struct intel_guc *guc)
+{
+	struct intel_gt *gt = guc_to_gt(guc);
+
+	spin_lock_irq(&gt->irq_lock);
+	gen11_gt_reset_one_iir(gt, 0, GEN11_GUC);
+	spin_unlock_irq(&gt->irq_lock);
+}
+
+static void gen11_enable_guc_interrupts(struct intel_guc *guc)
+{
+	struct intel_gt *gt = guc_to_gt(guc);
+
+	spin_lock_irq(&gt->irq_lock);
+	if (!guc->interrupts.enabled) {
+		u32 events = REG_FIELD_PREP(ENGINE1_MASK, GUC_INTR_GUC2HOST);
+
+		WARN_ON_ONCE(gen11_gt_reset_one_iir(gt, 0, GEN11_GUC));
+		intel_uncore_write(gt->uncore,
+				   GEN11_GUC_SG_INTR_ENABLE, events);
+		intel_uncore_write(gt->uncore,
+				   GEN11_GUC_SG_INTR_MASK, ~events);
+		guc->interrupts.enabled = true;
+	}
+	spin_unlock_irq(&gt->irq_lock);
+}
+
+static void gen11_disable_guc_interrupts(struct intel_guc *guc)
+{
+	struct intel_gt *gt = guc_to_gt(guc);
+
+	spin_lock_irq(&gt->irq_lock);
+	guc->interrupts.enabled = false;
+
+	intel_uncore_write(gt->uncore, GEN11_GUC_SG_INTR_MASK, ~0);
+	intel_uncore_write(gt->uncore, GEN11_GUC_SG_INTR_ENABLE, 0);
+
+	spin_unlock_irq(&gt->irq_lock);
+	intel_synchronize_irq(gt->i915);
+
+	gen11_reset_guc_interrupts(guc);
+}
+
 void intel_guc_init_early(struct intel_guc *guc)
 {
 	struct drm_i915_private *i915 = guc_to_gt(guc)->i915;
@@ -82,32 +192,6 @@ void intel_guc_init_early(struct intel_guc *guc)
 	}
 }
 
-static int guc_shared_data_create(struct intel_guc *guc)
-{
-	struct i915_vma *vma;
-	void *vaddr;
-
-	vma = intel_guc_allocate_vma(guc, PAGE_SIZE);
-	if (IS_ERR(vma))
-		return PTR_ERR(vma);
-
-	vaddr = i915_gem_object_pin_map(vma->obj, I915_MAP_WB);
-	if (IS_ERR(vaddr)) {
-		i915_vma_unpin_and_release(&vma, 0);
-		return PTR_ERR(vaddr);
-	}
-
-	guc->shared_data = vma;
-	guc->shared_data_vaddr = vaddr;
-
-	return 0;
-}
-
-static void guc_shared_data_destroy(struct intel_guc *guc)
-{
-	i915_vma_unpin_and_release(&guc->shared_data, I915_VMA_RELEASE_MAP);
-}
-
 static u32 guc_ctl_debug_flags(struct intel_guc *guc)
 {
 	u32 level = intel_guc_log_get_level(&guc->log);
@@ -254,14 +338,9 @@ int intel_guc_init(struct intel_guc *guc)
 	if (ret)
 		goto err_fetch;
 
-	ret = guc_shared_data_create(guc);
-	if (ret)
-		goto err_fw;
-	GEM_BUG_ON(!guc->shared_data);
-
 	ret = intel_guc_log_create(&guc->log);
 	if (ret)
-		goto err_shared;
+		goto err_fw;
 
 	ret = intel_guc_ads_create(guc);
 	if (ret)
@@ -296,8 +375,6 @@ err_ads:
 	intel_guc_ads_destroy(guc);
 err_log:
 	intel_guc_log_destroy(&guc->log);
-err_shared:
-	guc_shared_data_destroy(guc);
 err_fw:
 	intel_uc_fw_fini(&guc->fw);
 err_fetch:
@@ -322,7 +399,6 @@ void intel_guc_fini(struct intel_guc *guc)
 
 	intel_guc_ads_destroy(guc);
 	intel_guc_log_destroy(&guc->log);
-	guc_shared_data_destroy(guc);
 	intel_uc_fw_fini(&guc->fw);
 	intel_uc_fw_cleanup_fetch(&guc->fw);
 }
@@ -478,6 +554,13 @@ int intel_guc_suspend(struct intel_guc *guc)
 	};
 
 	/*
+	 * If GuC communication is enabled but submission is not supported,
+	 * we do not need to suspend the GuC.
+	 */
+	if (!intel_guc_submission_is_enabled(guc))
+		return 0;
+
+	/*
 	 * The ENTER_S_STATE action queues the save/restore operation in GuC FW
 	 * and then returns, so waiting on the H2G is not enough to guarantee
 	 * GuC is done. When all the processing is done, GuC writes
@@ -518,19 +601,9 @@ int intel_guc_suspend(struct intel_guc *guc)
 int intel_guc_reset_engine(struct intel_guc *guc,
 			   struct intel_engine_cs *engine)
 {
-	u32 data[7];
-
-	GEM_BUG_ON(!guc->execbuf_client);
-
-	data[0] = INTEL_GUC_ACTION_REQUEST_ENGINE_RESET;
-	data[1] = engine->guc_id;
-	data[2] = 0;
-	data[3] = 0;
-	data[4] = 0;
-	data[5] = guc->execbuf_client->stage_id;
-	data[6] = intel_guc_ggtt_offset(guc, guc->shared_data);
+	/* XXX: to be implemented with submission interface rework */
 
-	return intel_guc_send(guc, data, ARRAY_SIZE(data));
+	return -ENODEV;
 }
 
 /**
@@ -544,13 +617,27 @@ int intel_guc_resume(struct intel_guc *guc)
 		GUC_POWER_D0,
 	};
 
+	/*
+	 * If GuC communication is enabled but submission is not supported,
+	 * we do not need to resume the GuC but we do need to enable the
+	 * GuC communication on resume (above).
+	 */
+	if (!intel_guc_submission_is_enabled(guc))
+		return 0;
+
 	return intel_guc_send(guc, action, ARRAY_SIZE(action));
 }
 
 /**
- * DOC: GuC Address Space
+ * DOC: GuC Memory Management
  *
- * The layout of GuC address space is shown below:
+ * GuC can't allocate any memory for its own usage, so all the allocations must
+ * be handled by the host driver. GuC accesses the memory via the GGTT, with the
+ * exception of the top and bottom parts of the 4GB address space, which are
+ * instead re-mapped by the GuC HW to memory location of the FW itself (WOPCM)
+ * or other parts of the HW. The driver must take care not to place objects that
+ * the GuC is going to access in these reserved ranges. The layout of the GuC
+ * address space is shown below:
  *
  * ::
  *
diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc.h b/drivers/gpu/drm/i915/gt/uc/intel_guc.h
index 2b2f046d3cc3..e6400204a2bd 100644
--- a/drivers/gpu/drm/i915/gt/uc/intel_guc.h
+++ b/drivers/gpu/drm/i915/gt/uc/intel_guc.h
@@ -47,8 +47,6 @@ struct intel_guc {
 	struct i915_vma *stage_desc_pool;
 	void *stage_desc_pool_vaddr;
 	struct ida stage_ids;
-	struct i915_vma *shared_data;
-	void *shared_data_vaddr;
 
 	struct intel_guc_client *execbuf_client;
 
diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc_fwif.h b/drivers/gpu/drm/i915/gt/uc/intel_guc_fwif.h
index 1d3cdd67ca2f..a26a85d50209 100644
--- a/drivers/gpu/drm/i915/gt/uc/intel_guc_fwif.h
+++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_fwif.h
@@ -548,6 +548,7 @@ enum intel_guc_action {
 	INTEL_GUC_ACTION_ALLOCATE_DOORBELL = 0x10,
 	INTEL_GUC_ACTION_DEALLOCATE_DOORBELL = 0x20,
 	INTEL_GUC_ACTION_LOG_BUFFER_FILE_FLUSH_COMPLETE = 0x30,
+	INTEL_GUC_ACTION_UK_LOG_ENABLE_LOGGING = 0x40,
 	INTEL_GUC_ACTION_FORCE_LOG_BUFFER_FLUSH = 0x302,
 	INTEL_GUC_ACTION_ENTER_S_STATE = 0x501,
 	INTEL_GUC_ACTION_EXIT_S_STATE = 0x502,
@@ -556,7 +557,6 @@ enum intel_guc_action {
 	INTEL_GUC_ACTION_AUTHENTICATE_HUC = 0x4000,
 	INTEL_GUC_ACTION_REGISTER_COMMAND_TRANSPORT_BUFFER = 0x4505,
 	INTEL_GUC_ACTION_DEREGISTER_COMMAND_TRANSPORT_BUFFER = 0x4506,
-	INTEL_GUC_ACTION_UK_LOG_ENABLE_LOGGING = 0x0E000,
 	INTEL_GUC_ACTION_LIMIT
 };
 
diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc_log.c b/drivers/gpu/drm/i915/gt/uc/intel_guc_log.c
index 36332064de9c..caed0d57e704 100644
--- a/drivers/gpu/drm/i915/gt/uc/intel_guc_log.c
+++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_log.c
@@ -226,7 +226,7 @@ static void guc_read_update_log_buffer(struct intel_guc_log *log)
 
 	mutex_lock(&log->relay.lock);
 
-	if (WARN_ON(!intel_guc_log_relay_enabled(log)))
+	if (WARN_ON(!intel_guc_log_relay_created(log)))
 		goto out_unlock;
 
 	/* Get the pointer to shared GuC log buffer */
@@ -361,6 +361,7 @@ void intel_guc_log_init_early(struct intel_guc_log *log)
 {
 	mutex_init(&log->relay.lock);
 	INIT_WORK(&log->relay.flush_work, capture_logs_work);
+	log->relay.started = false;
 }
 
 static int guc_log_relay_create(struct intel_guc_log *log)
@@ -546,7 +547,7 @@ out_unlock:
 	return ret;
 }
 
-bool intel_guc_log_relay_enabled(const struct intel_guc_log *log)
+bool intel_guc_log_relay_created(const struct intel_guc_log *log)
 {
 	return log->relay.buf_addr;
 }
@@ -560,7 +561,7 @@ int intel_guc_log_relay_open(struct intel_guc_log *log)
 
 	mutex_lock(&log->relay.lock);
 
-	if (intel_guc_log_relay_enabled(log)) {
+	if (intel_guc_log_relay_created(log)) {
 		ret = -EEXIST;
 		goto out_unlock;
 	}
@@ -585,6 +586,21 @@ int intel_guc_log_relay_open(struct intel_guc_log *log)
 
 	mutex_unlock(&log->relay.lock);
 
+	return 0;
+
+out_relay:
+	guc_log_relay_destroy(log);
+out_unlock:
+	mutex_unlock(&log->relay.lock);
+
+	return ret;
+}
+
+int intel_guc_log_relay_start(struct intel_guc_log *log)
+{
+	if (log->relay.started)
+		return -EEXIST;
+
 	guc_log_enable_flush_events(log);
 
 	/*
@@ -594,47 +610,59 @@ int intel_guc_log_relay_open(struct intel_guc_log *log)
 	 */
 	queue_work(system_highpri_wq, &log->relay.flush_work);
 
-	return 0;
+	log->relay.started = true;
 
-out_relay:
-	guc_log_relay_destroy(log);
-out_unlock:
-	mutex_unlock(&log->relay.lock);
-
-	return ret;
+	return 0;
 }
 
 void intel_guc_log_relay_flush(struct intel_guc_log *log)
 {
 	struct intel_guc *guc = log_to_guc(log);
-	struct drm_i915_private *i915 = guc_to_gt(guc)->i915;
 	intel_wakeref_t wakeref;
 
+	if (!log->relay.started)
+		return;
+
 	/*
 	 * Before initiating the forceful flush, wait for any pending/ongoing
 	 * flush to complete otherwise forceful flush may not actually happen.
 	 */
 	flush_work(&log->relay.flush_work);
 
-	with_intel_runtime_pm(&i915->runtime_pm, wakeref)
+	with_intel_runtime_pm(guc_to_gt(guc)->uncore->rpm, wakeref)
 		guc_action_flush_log(guc);
 
 	/* GuC would have updated log buffer by now, so capture it */
 	guc_log_capture_logs(log);
 }
 
-void intel_guc_log_relay_close(struct intel_guc_log *log)
+/*
+ * Stops the relay log. Called from intel_guc_log_relay_close(), so no
+ * possibility of race with start/flush since relay_write cannot race
+ * relay_close.
+ */
+static void guc_log_relay_stop(struct intel_guc_log *log)
 {
 	struct intel_guc *guc = log_to_guc(log);
 	struct drm_i915_private *i915 = guc_to_gt(guc)->i915;
 
+	if (!log->relay.started)
+		return;
+
 	guc_log_disable_flush_events(log);
 	intel_synchronize_irq(i915);
 
 	flush_work(&log->relay.flush_work);
 
+	log->relay.started = false;
+}
+
+void intel_guc_log_relay_close(struct intel_guc_log *log)
+{
+	guc_log_relay_stop(log);
+
 	mutex_lock(&log->relay.lock);
-	GEM_BUG_ON(!intel_guc_log_relay_enabled(log));
+	GEM_BUG_ON(!intel_guc_log_relay_created(log));
 	guc_log_unmap(log);
 	guc_log_relay_destroy(log);
 	mutex_unlock(&log->relay.lock);
diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc_log.h b/drivers/gpu/drm/i915/gt/uc/intel_guc_log.h
index 6f764879acb1..c252c022c5fc 100644
--- a/drivers/gpu/drm/i915/gt/uc/intel_guc_log.h
+++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_log.h
@@ -47,6 +47,7 @@ struct intel_guc_log {
 	struct i915_vma *vma;
 	struct {
 		void *buf_addr;
+		bool started;
 		struct work_struct flush_work;
 		struct rchan *channel;
 		struct mutex lock;
@@ -65,8 +66,9 @@ int intel_guc_log_create(struct intel_guc_log *log);
 void intel_guc_log_destroy(struct intel_guc_log *log);
 
 int intel_guc_log_set_level(struct intel_guc_log *log, u32 level);
-bool intel_guc_log_relay_enabled(const struct intel_guc_log *log);
+bool intel_guc_log_relay_created(const struct intel_guc_log *log);
 int intel_guc_log_relay_open(struct intel_guc_log *log);
+int intel_guc_log_relay_start(struct intel_guc_log *log);
 void intel_guc_log_relay_flush(struct intel_guc_log *log);
 void intel_guc_log_relay_close(struct intel_guc_log *log);
 
diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc_reg.h b/drivers/gpu/drm/i915/gt/uc/intel_guc_reg.h
index edf194d23c6b..1949346e714e 100644
--- a/drivers/gpu/drm/i915/gt/uc/intel_guc_reg.h
+++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_reg.h
@@ -83,6 +83,9 @@
 #define GEN8_GTCR			_MMIO(0x4274)
 #define   GEN8_GTCR_INVALIDATE		  (1<<0)
 
+#define GEN12_GUC_TLB_INV_CR		_MMIO(0xcee8)
+#define   GEN12_GUC_TLB_INV_CR_INVALIDATE	(1 << 0)
+
 #define GUC_ARAT_C6DIS			_MMIO(0xA178)
 
 #define GUC_SHIM_CONTROL		_MMIO(0xc064)
diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c b/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c
index f325d3dd564f..2498c55e0ea5 100644
--- a/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c
+++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c
@@ -6,12 +6,13 @@
 #include <linux/circ_buf.h>
 
 #include "gem/i915_gem_context.h"
-
 #include "gt/intel_context.h"
 #include "gt/intel_engine_pm.h"
 #include "gt/intel_gt.h"
 #include "gt/intel_gt_pm.h"
 #include "gt/intel_lrc_reg.h"
+#include "gt/intel_ring.h"
+
 #include "intel_guc_submission.h"
 
 #include "i915_drv.h"
@@ -29,6 +30,12 @@ enum {
 /**
  * DOC: GuC-based command submission
  *
+ * IMPORTANT NOTE: GuC submission is currently not supported in i915. The GuC
+ * firmware is moving to an updated submission interface and we plan to
+ * turn submission back on when that lands. The below documentation (and related
+ * code) matches the old submission model and will be updated as part of the
+ * upgrade to the new flow.
+ *
  * GuC client:
  * A intel_guc_client refers to a submission path through GuC. Currently, there
  * is only one client, which is charged with all submissions to the GuC. This
@@ -1004,7 +1011,7 @@ void intel_guc_submission_fini(struct intel_guc *guc)
 
 static void guc_interrupts_capture(struct intel_gt *gt)
 {
-	struct intel_rps *rps = &gt->i915->gt_pm.rps;
+	struct intel_rps *rps = &gt->rps;
 	struct intel_uncore *uncore = gt->uncore;
 	struct intel_engine_cs *engine;
 	enum intel_engine_id id;
@@ -1014,7 +1021,7 @@ static void guc_interrupts_capture(struct intel_gt *gt)
 	 * to GuC
 	 */
 	irqs = _MASKED_BIT_ENABLE(GFX_INTERRUPT_STEERING);
-	for_each_engine(engine, gt->i915, id)
+	for_each_engine(engine, gt, id)
 		ENGINE_WRITE(engine, RING_MODE_GEN7, irqs);
 
 	/* route USER_INTERRUPT to Host, all others are sent to GuC. */
@@ -1050,7 +1057,7 @@ static void guc_interrupts_capture(struct intel_gt *gt)
 
 static void guc_interrupts_release(struct intel_gt *gt)
 {
-	struct intel_rps *rps = &gt->i915->gt_pm.rps;
+	struct intel_rps *rps = &gt->rps;
 	struct intel_uncore *uncore = gt->uncore;
 	struct intel_engine_cs *engine;
 	enum intel_engine_id id;
@@ -1062,7 +1069,7 @@ static void guc_interrupts_release(struct intel_gt *gt)
 	 */
 	irqs = _MASKED_FIELD(GFX_FORWARD_VBLANK_MASK, GFX_FORWARD_VBLANK_NEVER);
 	irqs |= _MASKED_BIT_DISABLE(GFX_INTERRUPT_STEERING);
-	for_each_engine(engine, gt->i915, id)
+	for_each_engine(engine, gt, id)
 		ENGINE_WRITE(engine, RING_MODE_GEN7, irqs);
 
 	/* route all GT interrupts to the host */
@@ -1119,7 +1126,7 @@ int intel_guc_submission_enable(struct intel_guc *guc)
 	enum intel_engine_id id;
 	int err;
 
-	err = i915_inject_load_error(gt->i915, -ENXIO);
+	err = i915_inject_probe_error(gt->i915, -ENXIO);
 	if (err)
 		return err;
 
@@ -1145,7 +1152,7 @@ int intel_guc_submission_enable(struct intel_guc *guc)
 	/* Take over from manual control of ELSP (execlists) */
 	guc_interrupts_capture(gt);
 
-	for_each_engine(engine, gt->i915, id) {
+	for_each_engine(engine, gt, id) {
 		engine->set_default_submission = guc_set_default_submission;
 		engine->set_default_submission(engine);
 	}
diff --git a/drivers/gpu/drm/i915/gt/uc/intel_huc.c b/drivers/gpu/drm/i915/gt/uc/intel_huc.c
index d4625c97b4f9..32a069841c14 100644
--- a/drivers/gpu/drm/i915/gt/uc/intel_huc.c
+++ b/drivers/gpu/drm/i915/gt/uc/intel_huc.c
@@ -9,6 +9,34 @@
 #include "intel_huc.h"
 #include "i915_drv.h"
 
+/**
+ * DOC: HuC
+ *
+ * The HuC is a dedicated microcontroller for usage in media HEVC (High
+ * Efficiency Video Coding) operations. Userspace can directly use the firmware
+ * capabilities by adding HuC specific commands to batch buffers.
+ *
+ * The kernel driver is only responsible for loading the HuC firmware and
+ * triggering its security authentication, which is performed by the GuC. For
+ * The GuC to correctly perform the authentication, the HuC binary must be
+ * loaded before the GuC one. Loading the HuC is optional; however, not using
+ * the HuC might negatively impact power usage and/or performance of media
+ * workloads, depending on the use-cases.
+ *
+ * See https://github.com/intel/media-driver for the latest details on HuC
+ * functionality.
+ */
+
+/**
+ * DOC: HuC Memory Management
+ *
+ * Similarly to the GuC, the HuC can't do any memory allocations on its own,
+ * with the difference being that the allocations for HuC usage are handled by
+ * the userspace driver instead of the kernel one. The HuC accesses the memory
+ * via the PPGTT belonging to the context loaded on the VCS executing the
+ * HuC-specific commands.
+ */
+
 void intel_huc_init_early(struct intel_huc *huc)
 {
 	struct drm_i915_private *i915 = huc_to_gt(huc)->i915;
@@ -35,7 +63,7 @@ static int intel_huc_rsa_data_create(struct intel_huc *huc)
 	void *vaddr;
 	int err;
 
-	err = i915_inject_load_error(gt->i915, -ENXIO);
+	err = i915_inject_probe_error(gt->i915, -ENXIO);
 	if (err)
 		return err;
 
@@ -118,10 +146,9 @@ void intel_huc_fini(struct intel_huc *huc)
  *
  * Called after HuC and GuC firmware loading during intel_uc_init_hw().
  *
- * This function pins HuC firmware image object into GGTT.
- * Then it invokes GuC action to authenticate passing the offset to RSA
- * signature through intel_guc_auth_huc(). It then waits for 50ms for
- * firmware verification ACK and unpins the object.
+ * This function invokes the GuC action to authenticate the HuC firmware,
+ * passing the offset of the RSA signature to intel_guc_auth_huc(). It then
+ * waits for up to 50ms for firmware verification ACK.
  */
 int intel_huc_auth(struct intel_huc *huc)
 {
@@ -134,7 +161,7 @@ int intel_huc_auth(struct intel_huc *huc)
 	if (!intel_uc_fw_is_loaded(&huc->fw))
 		return -ENOEXEC;
 
-	ret = i915_inject_load_error(gt->i915, -ENXIO);
+	ret = i915_inject_probe_error(gt->i915, -ENXIO);
 	if (ret)
 		goto fail;
 
@@ -185,7 +212,7 @@ int intel_huc_check_status(struct intel_huc *huc)
 	if (!intel_huc_is_supported(huc))
 		return -ENODEV;
 
-	with_intel_runtime_pm(&gt->i915->runtime_pm, wakeref)
+	with_intel_runtime_pm(gt->uncore->rpm, wakeref)
 		status = intel_uncore_read(gt->uncore, huc->status.reg);
 
 	return (status & huc->status.mask) == huc->status.value;
diff --git a/drivers/gpu/drm/i915/gt/uc/intel_huc_fw.c b/drivers/gpu/drm/i915/gt/uc/intel_huc_fw.c
index 74602487ed67..d654340d4d03 100644
--- a/drivers/gpu/drm/i915/gt/uc/intel_huc_fw.c
+++ b/drivers/gpu/drm/i915/gt/uc/intel_huc_fw.c
@@ -8,21 +8,6 @@
 #include "i915_drv.h"
 
 /**
- * DOC: HuC Firmware
- *
- * Motivation:
- * GEN9 introduces a new dedicated firmware for usage in media HEVC (High
- * Efficiency Video Coding) operations. Userspace can use the firmware
- * capabilities by adding HuC specific commands to batch buffers.
- *
- * Implementation:
- * The same firmware loader is used as the GuC. However, the actual
- * loading to HW is deferred until GEM initialization is done.
- *
- * Note that HuC firmware loading must be done before GuC loading.
- */
-
-/**
  * intel_huc_fw_init_early() - initializes HuC firmware struct
  * @huc: intel_huc struct
  *
diff --git a/drivers/gpu/drm/i915/gt/uc/intel_uc.c b/drivers/gpu/drm/i915/gt/uc/intel_uc.c
index 71ee7ab035cc..629b19377a29 100644
--- a/drivers/gpu/drm/i915/gt/uc/intel_uc.c
+++ b/drivers/gpu/drm/i915/gt/uc/intel_uc.c
@@ -20,7 +20,7 @@ static int __intel_uc_reset_hw(struct intel_uc *uc)
 	int ret;
 	u32 guc_status;
 
-	ret = i915_inject_load_error(gt->i915, -ENXIO);
+	ret = i915_inject_probe_error(gt->i915, -ENXIO);
 	if (ret)
 		return ret;
 
@@ -197,7 +197,7 @@ static int guc_enable_communication(struct intel_guc *guc)
 
 	GEM_BUG_ON(guc_communication_enabled(guc));
 
-	ret = i915_inject_load_error(i915, -ENXIO);
+	ret = i915_inject_probe_error(i915, -ENXIO);
 	if (ret)
 		return ret;
 
@@ -224,17 +224,7 @@ static int guc_enable_communication(struct intel_guc *guc)
 	return 0;
 }
 
-static void guc_stop_communication(struct intel_guc *guc)
-{
-	intel_guc_ct_stop(&guc->ct);
-
-	guc->send = intel_guc_send_nop;
-	guc->handler = intel_guc_to_host_event_handler_nop;
-
-	guc_clear_mmio_msg(guc);
-}
-
-static void guc_disable_communication(struct intel_guc *guc)
+static void __guc_stop_communication(struct intel_guc *guc)
 {
 	/*
 	 * Events generated during or after CT disable are logged by guc in
@@ -247,6 +237,20 @@ static void guc_disable_communication(struct intel_guc *guc)
 
 	guc->send = intel_guc_send_nop;
 	guc->handler = intel_guc_to_host_event_handler_nop;
+}
+
+static void guc_stop_communication(struct intel_guc *guc)
+{
+	intel_guc_ct_stop(&guc->ct);
+
+	__guc_stop_communication(guc);
+
+	DRM_INFO("GuC communication stopped\n");
+}
+
+static void guc_disable_communication(struct intel_guc *guc)
+{
+	__guc_stop_communication(guc);
 
 	intel_guc_ct_disable(&guc->ct);
 
@@ -368,7 +372,7 @@ static int uc_init_wopcm(struct intel_uc *uc)
 	GEM_BUG_ON(!(size & GUC_WOPCM_SIZE_MASK));
 	GEM_BUG_ON(size & ~GUC_WOPCM_SIZE_MASK);
 
-	err = i915_inject_load_error(gt->i915, -ENXIO);
+	err = i915_inject_probe_error(gt->i915, -ENXIO);
 	if (err)
 		return err;
 
@@ -537,7 +541,9 @@ void intel_uc_fini_hw(struct intel_uc *uc)
 	if (intel_uc_supports_guc_submission(uc))
 		intel_guc_submission_disable(guc);
 
-	guc_disable_communication(guc);
+	if (guc_communication_enabled(guc))
+		guc_disable_communication(guc);
+
 	__uc_sanitize(uc);
 }
 
@@ -581,7 +587,7 @@ void intel_uc_suspend(struct intel_uc *uc)
 	if (!intel_guc_is_running(guc))
 		return;
 
-	with_intel_runtime_pm(&uc_to_gt(uc)->i915->runtime_pm, wakeref)
+	with_intel_runtime_pm(uc_to_gt(uc)->uncore->rpm, wakeref)
 		intel_uc_runtime_suspend(uc);
 }
 
diff --git a/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.c b/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.c
index bd22bf11adad..66a30ab7044a 100644
--- a/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.c
+++ b/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.c
@@ -37,27 +37,34 @@ void intel_uc_fw_change_status(struct intel_uc_fw *uc_fw,
 /*
  * List of required GuC and HuC binaries per-platform.
  * Must be ordered based on platform + revid, from newer to older.
+ *
+ * TGL 35.2 is interface-compatible with 33.0 for previous Gens. The deltas
+ * between 33.0 and 35.2 are only related to new additions to support new Gen12
+ * features.
  */
 #define INTEL_UC_FIRMWARE_DEFS(fw_def, guc_def, huc_def) \
-	fw_def(ICELAKE,    0, guc_def(icl, 33, 0, 0), huc_def(icl,  8,  4, 3238)) \
-	fw_def(COFFEELAKE, 0, guc_def(kbl, 33, 0, 0), huc_def(kbl, 02, 00, 1810)) \
-	fw_def(GEMINILAKE, 0, guc_def(glk, 33, 0, 0), huc_def(glk, 03, 01, 2893)) \
-	fw_def(KABYLAKE,   0, guc_def(kbl, 33, 0, 0), huc_def(kbl, 02, 00, 1810)) \
-	fw_def(BROXTON,    0, guc_def(bxt, 33, 0, 0), huc_def(bxt, 01,  8, 2893)) \
-	fw_def(SKYLAKE,    0, guc_def(skl, 33, 0, 0), huc_def(skl, 01, 07, 1398))
-
-#define __MAKE_UC_FW_PATH(prefix_, name_, separator_, major_, minor_, patch_) \
+	fw_def(TIGERLAKE,   0, guc_def(tgl, 35, 2, 0), huc_def(tgl,  7, 0, 3)) \
+	fw_def(ELKHARTLAKE, 0, guc_def(ehl, 33, 0, 4), huc_def(ehl,  9, 0, 0)) \
+	fw_def(ICELAKE,     0, guc_def(icl, 33, 0, 0), huc_def(icl,  9, 0, 0)) \
+	fw_def(COFFEELAKE,  5, guc_def(cml, 33, 0, 0), huc_def(cml,  4, 0, 0)) \
+	fw_def(COFFEELAKE,  0, guc_def(kbl, 33, 0, 0), huc_def(kbl,  4, 0, 0)) \
+	fw_def(GEMINILAKE,  0, guc_def(glk, 33, 0, 0), huc_def(glk,  4, 0, 0)) \
+	fw_def(KABYLAKE,    0, guc_def(kbl, 33, 0, 0), huc_def(kbl,  4, 0, 0)) \
+	fw_def(BROXTON,     0, guc_def(bxt, 33, 0, 0), huc_def(bxt,  2, 0, 0)) \
+	fw_def(SKYLAKE,     0, guc_def(skl, 33, 0, 0), huc_def(skl,  2, 0, 0))
+
+#define __MAKE_UC_FW_PATH(prefix_, name_, major_, minor_, patch_) \
 	"i915/" \
 	__stringify(prefix_) name_ \
-	__stringify(major_) separator_ \
-	__stringify(minor_) separator_ \
+	__stringify(major_) "." \
+	__stringify(minor_) "." \
 	__stringify(patch_) ".bin"
 
 #define MAKE_GUC_FW_PATH(prefix_, major_, minor_, patch_) \
-	__MAKE_UC_FW_PATH(prefix_, "_guc_", ".", major_, minor_, patch_)
+	__MAKE_UC_FW_PATH(prefix_, "_guc_", major_, minor_, patch_)
 
 #define MAKE_HUC_FW_PATH(prefix_, major_, minor_, bld_num_) \
-	__MAKE_UC_FW_PATH(prefix_, "_huc_ver", "_", major_, minor_, bld_num_)
+	__MAKE_UC_FW_PATH(prefix_, "_huc_", major_, minor_, bld_num_)
 
 /* All blobs need to be declared via MODULE_FIRMWARE() */
 #define INTEL_UC_MODULE_FW(platform_, revid_, guc_, huc_) \
@@ -218,29 +225,31 @@ static void __force_fw_fetch_failures(struct intel_uc_fw *uc_fw,
 {
 	bool user = e == -EINVAL;
 
-	if (i915_inject_load_error(i915, e)) {
+	if (i915_inject_probe_error(i915, e)) {
 		/* non-existing blob */
 		uc_fw->path = "<invalid>";
 		uc_fw->user_overridden = user;
-	} else if (i915_inject_load_error(i915, e)) {
+	} else if (i915_inject_probe_error(i915, e)) {
 		/* require next major version */
 		uc_fw->major_ver_wanted += 1;
 		uc_fw->minor_ver_wanted = 0;
 		uc_fw->user_overridden = user;
-	} else if (i915_inject_load_error(i915, e)) {
+	} else if (i915_inject_probe_error(i915, e)) {
 		/* require next minor version */
 		uc_fw->minor_ver_wanted += 1;
 		uc_fw->user_overridden = user;
-	} else if (uc_fw->major_ver_wanted && i915_inject_load_error(i915, e)) {
+	} else if (uc_fw->major_ver_wanted &&
+		   i915_inject_probe_error(i915, e)) {
 		/* require prev major version */
 		uc_fw->major_ver_wanted -= 1;
 		uc_fw->minor_ver_wanted = 0;
 		uc_fw->user_overridden = user;
-	} else if (uc_fw->minor_ver_wanted && i915_inject_load_error(i915, e)) {
+	} else if (uc_fw->minor_ver_wanted &&
+		   i915_inject_probe_error(i915, e)) {
 		/* require prev minor version - hey, this should work! */
 		uc_fw->minor_ver_wanted -= 1;
 		uc_fw->user_overridden = user;
-	} else if (user && i915_inject_load_error(i915, e)) {
+	} else if (user && i915_inject_probe_error(i915, e)) {
 		/* officially unsupported platform */
 		uc_fw->major_ver_wanted = 0;
 		uc_fw->minor_ver_wanted = 0;
@@ -269,7 +278,7 @@ int intel_uc_fw_fetch(struct intel_uc_fw *uc_fw, struct drm_i915_private *i915)
 	GEM_BUG_ON(!i915->wopcm.size);
 	GEM_BUG_ON(!intel_uc_fw_is_enabled(uc_fw));
 
-	err = i915_inject_load_error(i915, -ENXIO);
+	err = i915_inject_probe_error(i915, -ENXIO);
 	if (err)
 		return err;
 
@@ -337,25 +346,10 @@ int intel_uc_fw_fetch(struct intel_uc_fw *uc_fw, struct drm_i915_private *i915)
 	}
 
 	/* Get version numbers from the CSS header */
-	switch (uc_fw->type) {
-	case INTEL_UC_FW_TYPE_GUC:
-		uc_fw->major_ver_found = FIELD_GET(CSS_SW_VERSION_GUC_MAJOR,
-						   css->sw_version);
-		uc_fw->minor_ver_found = FIELD_GET(CSS_SW_VERSION_GUC_MINOR,
-						   css->sw_version);
-		break;
-
-	case INTEL_UC_FW_TYPE_HUC:
-		uc_fw->major_ver_found = FIELD_GET(CSS_SW_VERSION_HUC_MAJOR,
-						   css->sw_version);
-		uc_fw->minor_ver_found = FIELD_GET(CSS_SW_VERSION_HUC_MINOR,
-						   css->sw_version);
-		break;
-
-	default:
-		MISSING_CASE(uc_fw->type);
-		break;
-	}
+	uc_fw->major_ver_found = FIELD_GET(CSS_SW_VERSION_UC_MAJOR,
+					   css->sw_version);
+	uc_fw->minor_ver_found = FIELD_GET(CSS_SW_VERSION_UC_MINOR,
+					   css->sw_version);
 
 	if (uc_fw->major_ver_found != uc_fw->major_ver_wanted ||
 	    uc_fw->minor_ver_found < uc_fw->minor_ver_wanted) {
@@ -400,7 +394,7 @@ static u32 uc_fw_ggtt_offset(struct intel_uc_fw *uc_fw, struct i915_ggtt *ggtt)
 {
 	struct drm_mm_node *node = &ggtt->uc_fw;
 
-	GEM_BUG_ON(!node->allocated);
+	GEM_BUG_ON(!drm_mm_node_allocated(node));
 	GEM_BUG_ON(upper_32_bits(node->start));
 	GEM_BUG_ON(upper_32_bits(node->start + node->size - 1));
 
@@ -445,7 +439,7 @@ static int uc_fw_xfer(struct intel_uc_fw *uc_fw, struct intel_gt *gt,
 	u64 offset;
 	int ret;
 
-	ret = i915_inject_load_error(gt->i915, -ETIMEDOUT);
+	ret = i915_inject_probe_error(gt->i915, -ETIMEDOUT);
 	if (ret)
 		return ret;
 
@@ -506,7 +500,7 @@ int intel_uc_fw_upload(struct intel_uc_fw *uc_fw, struct intel_gt *gt,
 	/* make sure the status was cleared the last time we reset the uc */
 	GEM_BUG_ON(intel_uc_fw_is_loaded(uc_fw));
 
-	err = i915_inject_load_error(gt->i915, -ENOEXEC);
+	err = i915_inject_probe_error(gt->i915, -ENOEXEC);
 	if (err)
 		return err;
 
diff --git a/drivers/gpu/drm/i915/gt/uc/intel_uc_fw_abi.h b/drivers/gpu/drm/i915/gt/uc/intel_uc_fw_abi.h
index ae58e8a8c53b..029214cdedd5 100644
--- a/drivers/gpu/drm/i915/gt/uc/intel_uc_fw_abi.h
+++ b/drivers/gpu/drm/i915/gt/uc/intel_uc_fw_abi.h
@@ -39,9 +39,6 @@
  * 3. Length info of each component can be found in header, in dwords.
  * 4. Modulus and exponent key are not required by driver. They may not appear
  *    in fw. So driver will load a truncated firmware in this case.
- *
- * The only difference between GuC and HuC firmwares is how the version
- * information is saved.
  */
 
 struct uc_css_header {
@@ -69,11 +66,9 @@ struct uc_css_header {
 	char username[8];
 	char buildnumber[12];
 	u32 sw_version;
-#define CSS_SW_VERSION_GUC_MAJOR	(0xFF << 16)
-#define CSS_SW_VERSION_GUC_MINOR	(0xFF << 8)
-#define CSS_SW_VERSION_GUC_PATCH	(0xFF << 0)
-#define CSS_SW_VERSION_HUC_MAJOR	(0xFFFF << 16)
-#define CSS_SW_VERSION_HUC_MINOR	(0xFFFF << 0)
+#define CSS_SW_VERSION_UC_MAJOR		(0xFF << 16)
+#define CSS_SW_VERSION_UC_MINOR		(0xFF << 8)
+#define CSS_SW_VERSION_UC_PATCH		(0xFF << 0)
 	u32 reserved[14];
 	u32 header_info;
 } __packed;
diff --git a/drivers/gpu/drm/i915/gt/uc/selftest_guc.c b/drivers/gpu/drm/i915/gt/uc/selftest_guc.c
index bba0eafe1cdb..d8a80388bd31 100644
--- a/drivers/gpu/drm/i915/gt/uc/selftest_guc.c
+++ b/drivers/gpu/drm/i915/gt/uc/selftest_guc.c
@@ -108,23 +108,15 @@ static bool client_doorbell_in_sync(struct intel_guc_client *client)
  * validating that the doorbells status expected by the driver matches what the
  * GuC/HW have.
  */
-static int igt_guc_clients(void *args)
+static int igt_guc_clients(void *arg)
 {
-	struct drm_i915_private *dev_priv = args;
+	struct intel_gt *gt = arg;
+	struct intel_guc *guc = &gt->uc.guc;
 	intel_wakeref_t wakeref;
-	struct intel_guc *guc;
 	int err = 0;
 
-	GEM_BUG_ON(!HAS_GT_UC(dev_priv));
-	mutex_lock(&dev_priv->drm.struct_mutex);
-	wakeref = intel_runtime_pm_get(&dev_priv->runtime_pm);
-
-	guc = &dev_priv->gt.uc.guc;
-	if (!guc) {
-		pr_err("No guc object!\n");
-		err = -EINVAL;
-		goto unlock;
-	}
+	GEM_BUG_ON(!HAS_GT_UC(gt->i915));
+	wakeref = intel_runtime_pm_get(gt->uncore->rpm);
 
 	err = check_all_doorbells(guc);
 	if (err)
@@ -189,8 +181,7 @@ out:
 	guc_clients_create(guc);
 	guc_clients_enable(guc);
 unlock:
-	intel_runtime_pm_put(&dev_priv->runtime_pm, wakeref);
-	mutex_unlock(&dev_priv->drm.struct_mutex);
+	intel_runtime_pm_put(gt->uncore->rpm, wakeref);
 	return err;
 }
 
@@ -201,22 +192,14 @@ unlock:
  */
 static int igt_guc_doorbells(void *arg)
 {
-	struct drm_i915_private *dev_priv = arg;
+	struct intel_gt *gt = arg;
+	struct intel_guc *guc = &gt->uc.guc;
 	intel_wakeref_t wakeref;
-	struct intel_guc *guc;
 	int i, err = 0;
 	u16 db_id;
 
-	GEM_BUG_ON(!HAS_GT_UC(dev_priv));
-	mutex_lock(&dev_priv->drm.struct_mutex);
-	wakeref = intel_runtime_pm_get(&dev_priv->runtime_pm);
-
-	guc = &dev_priv->gt.uc.guc;
-	if (!guc) {
-		pr_err("No guc object!\n");
-		err = -EINVAL;
-		goto unlock;
-	}
+	GEM_BUG_ON(!HAS_GT_UC(gt->i915));
+	wakeref = intel_runtime_pm_get(gt->uncore->rpm);
 
 	err = check_all_doorbells(guc);
 	if (err)
@@ -298,20 +281,19 @@ out:
 			guc_client_free(clients[i]);
 		}
 unlock:
-	intel_runtime_pm_put(&dev_priv->runtime_pm, wakeref);
-	mutex_unlock(&dev_priv->drm.struct_mutex);
+	intel_runtime_pm_put(gt->uncore->rpm, wakeref);
 	return err;
 }
 
-int intel_guc_live_selftest(struct drm_i915_private *dev_priv)
+int intel_guc_live_selftest(struct drm_i915_private *i915)
 {
 	static const struct i915_subtest tests[] = {
 		SUBTEST(igt_guc_clients),
 		SUBTEST(igt_guc_doorbells),
 	};
 
-	if (!USES_GUC_SUBMISSION(dev_priv))
+	if (!USES_GUC_SUBMISSION(i915))
 		return 0;
 
-	return i915_subtests(tests, dev_priv);
+	return intel_gt_live_subtests(tests, &i915->gt);
 }
diff --git a/drivers/gpu/drm/i915/gvt/aperture_gm.c b/drivers/gpu/drm/i915/gvt/aperture_gm.c
index 5ff2437b2998..771420453f82 100644
--- a/drivers/gpu/drm/i915/gvt/aperture_gm.c
+++ b/drivers/gpu/drm/i915/gvt/aperture_gm.c
@@ -61,14 +61,14 @@ static int alloc_gm(struct intel_vgpu *vgpu, bool high_gm)
 		flags = PIN_MAPPABLE;
 	}
 
-	mutex_lock(&dev_priv->drm.struct_mutex);
+	mutex_lock(&dev_priv->ggtt.vm.mutex);
 	mmio_hw_access_pre(dev_priv);
 	ret = i915_gem_gtt_insert(&dev_priv->ggtt.vm, node,
 				  size, I915_GTT_PAGE_SIZE,
 				  I915_COLOR_UNEVICTABLE,
 				  start, end, flags);
 	mmio_hw_access_post(dev_priv);
-	mutex_unlock(&dev_priv->drm.struct_mutex);
+	mutex_unlock(&dev_priv->ggtt.vm.mutex);
 	if (ret)
 		gvt_err("fail to alloc %s gm space from host\n",
 			high_gm ? "high" : "low");
@@ -98,9 +98,9 @@ static int alloc_vgpu_gm(struct intel_vgpu *vgpu)
 
 	return 0;
 out_free_aperture:
-	mutex_lock(&dev_priv->drm.struct_mutex);
+	mutex_lock(&dev_priv->ggtt.vm.mutex);
 	drm_mm_remove_node(&vgpu->gm.low_gm_node);
-	mutex_unlock(&dev_priv->drm.struct_mutex);
+	mutex_unlock(&dev_priv->ggtt.vm.mutex);
 	return ret;
 }
 
@@ -108,10 +108,10 @@ static void free_vgpu_gm(struct intel_vgpu *vgpu)
 {
 	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
 
-	mutex_lock(&dev_priv->drm.struct_mutex);
+	mutex_lock(&dev_priv->ggtt.vm.mutex);
 	drm_mm_remove_node(&vgpu->gm.low_gm_node);
 	drm_mm_remove_node(&vgpu->gm.high_gm_node);
-	mutex_unlock(&dev_priv->drm.struct_mutex);
+	mutex_unlock(&dev_priv->ggtt.vm.mutex);
 }
 
 /**
@@ -198,7 +198,7 @@ static int alloc_vgpu_fence(struct intel_vgpu *vgpu)
 	mutex_lock(&dev_priv->ggtt.vm.mutex);
 
 	for (i = 0; i < vgpu_fence_sz(vgpu); i++) {
-		reg = i915_reserve_fence(dev_priv);
+		reg = i915_reserve_fence(&dev_priv->ggtt);
 		if (IS_ERR(reg))
 			goto out_free_fence;
 
diff --git a/drivers/gpu/drm/i915/gvt/cmd_parser.c b/drivers/gpu/drm/i915/gvt/cmd_parser.c
index e753b1e706e2..6a3ac8cde95d 100644
--- a/drivers/gpu/drm/i915/gvt/cmd_parser.c
+++ b/drivers/gpu/drm/i915/gvt/cmd_parser.c
@@ -35,7 +35,9 @@
  */
 
 #include <linux/slab.h>
+
 #include "i915_drv.h"
+#include "gt/intel_ring.h"
 #include "gvt.h"
 #include "i915_pvinfo.h"
 #include "trace.h"
diff --git a/drivers/gpu/drm/i915/gvt/dmabuf.c b/drivers/gpu/drm/i915/gvt/dmabuf.c
index 13044c027f27..e451298d11c3 100644
--- a/drivers/gpu/drm/i915/gvt/dmabuf.c
+++ b/drivers/gpu/drm/i915/gvt/dmabuf.c
@@ -152,6 +152,7 @@ static const struct drm_i915_gem_object_ops intel_vgpu_gem_ops = {
 static struct drm_i915_gem_object *vgpu_create_gem(struct drm_device *dev,
 		struct intel_vgpu_fb_info *info)
 {
+	static struct lock_class_key lock_class;
 	struct drm_i915_private *dev_priv = to_i915(dev);
 	struct drm_i915_gem_object *obj;
 
@@ -161,7 +162,7 @@ static struct drm_i915_gem_object *vgpu_create_gem(struct drm_device *dev,
 
 	drm_gem_private_object_init(dev, &obj->base,
 		roundup(info->size, PAGE_SIZE));
-	i915_gem_object_init(obj, &intel_vgpu_gem_ops);
+	i915_gem_object_init(obj, &intel_vgpu_gem_ops, &lock_class);
 
 	obj->read_domains = I915_GEM_DOMAIN_GTT;
 	obj->write_domain = 0;
@@ -498,8 +499,6 @@ int intel_vgpu_get_dmabuf(struct intel_vgpu *vgpu, unsigned int dmabuf_id)
 		goto out_free_gem;
 	}
 
-	i915_gem_object_put(obj);
-
 	ret = dma_buf_fd(dmabuf, DRM_CLOEXEC | DRM_RDWR);
 	if (ret < 0) {
 		gvt_vgpu_err("create dma-buf fd failed ret:%d\n", ret);
@@ -524,6 +523,8 @@ int intel_vgpu_get_dmabuf(struct intel_vgpu *vgpu, unsigned int dmabuf_id)
 		    file_count(dmabuf->file),
 		    kref_read(&obj->base.refcount));
 
+	i915_gem_object_put(obj);
+
 	return dmabuf_fd;
 
 out_free_dmabuf:
diff --git a/drivers/gpu/drm/i915/gvt/execlist.c b/drivers/gpu/drm/i915/gvt/execlist.c
index f21b8fb5b37e..d6e7a1189bad 100644
--- a/drivers/gpu/drm/i915/gvt/execlist.c
+++ b/drivers/gpu/drm/i915/gvt/execlist.c
@@ -534,7 +534,7 @@ static void clean_execlist(struct intel_vgpu *vgpu,
 	struct intel_vgpu_submission *s = &vgpu->submission;
 	intel_engine_mask_t tmp;
 
-	for_each_engine_masked(engine, dev_priv, engine_mask, tmp) {
+	for_each_engine_masked(engine, &dev_priv->gt, engine_mask, tmp) {
 		kfree(s->ring_scan_buffer[engine->id]);
 		s->ring_scan_buffer[engine->id] = NULL;
 		s->ring_scan_buffer_size[engine->id] = 0;
@@ -548,7 +548,7 @@ static void reset_execlist(struct intel_vgpu *vgpu,
 	struct intel_engine_cs *engine;
 	intel_engine_mask_t tmp;
 
-	for_each_engine_masked(engine, dev_priv, engine_mask, tmp)
+	for_each_engine_masked(engine, &dev_priv->gt, engine_mask, tmp)
 		init_vgpu_execlist(vgpu, engine->id);
 }
 
diff --git a/drivers/gpu/drm/i915/gvt/handlers.c b/drivers/gpu/drm/i915/gvt/handlers.c
index 25f78196b964..bd12af349123 100644
--- a/drivers/gpu/drm/i915/gvt/handlers.c
+++ b/drivers/gpu/drm/i915/gvt/handlers.c
@@ -819,13 +819,16 @@ static int trigger_aux_channel_interrupt(struct intel_vgpu *vgpu,
 	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
 	enum intel_gvt_event_type event;
 
-	if (reg == _DPA_AUX_CH_CTL)
+	if (reg == i915_mmio_reg_offset(DP_AUX_CH_CTL(AUX_CH_A)))
 		event = AUX_CHANNEL_A;
-	else if (reg == _PCH_DPB_AUX_CH_CTL || reg == _DPB_AUX_CH_CTL)
+	else if (reg == _PCH_DPB_AUX_CH_CTL ||
+		 reg == i915_mmio_reg_offset(DP_AUX_CH_CTL(AUX_CH_B)))
 		event = AUX_CHANNEL_B;
-	else if (reg == _PCH_DPC_AUX_CH_CTL || reg == _DPC_AUX_CH_CTL)
+	else if (reg == _PCH_DPC_AUX_CH_CTL ||
+		 reg == i915_mmio_reg_offset(DP_AUX_CH_CTL(AUX_CH_C)))
 		event = AUX_CHANNEL_C;
-	else if (reg == _PCH_DPD_AUX_CH_CTL || reg == _DPD_AUX_CH_CTL)
+	else if (reg == _PCH_DPD_AUX_CH_CTL ||
+		 reg == i915_mmio_reg_offset(DP_AUX_CH_CTL(AUX_CH_D)))
 		event = AUX_CHANNEL_D;
 	else {
 		WARN_ON(true);
@@ -2796,7 +2799,7 @@ static int init_broadwell_mmio_info(struct intel_gvt *gvt)
 	MMIO_D(CHICKEN_PIPESL_1(PIPE_C), D_BDW_PLUS);
 
 	MMIO_D(WM_MISC, D_BDW);
-	MMIO_D(_MMIO(BDW_EDP_PSR_BASE), D_BDW);
+	MMIO_D(_MMIO(_SRD_CTL_EDP), D_BDW);
 
 	MMIO_D(_MMIO(0x6671c), D_BDW_PLUS);
 	MMIO_D(_MMIO(0x66c00), D_BDW_PLUS);
@@ -2872,11 +2875,11 @@ static int init_skl_mmio_info(struct intel_gvt *gvt)
 	MMIO_DH(FORCEWAKE_MEDIA_GEN9, D_SKL_PLUS, NULL, mul_force_wake_write);
 	MMIO_DH(FORCEWAKE_ACK_MEDIA_GEN9, D_SKL_PLUS, NULL, NULL);
 
-	MMIO_F(_MMIO(_DPB_AUX_CH_CTL), 6 * 4, 0, 0, 0, D_SKL_PLUS, NULL,
+	MMIO_F(DP_AUX_CH_CTL(AUX_CH_B), 6 * 4, 0, 0, 0, D_SKL_PLUS, NULL,
 						dp_aux_ch_ctl_mmio_write);
-	MMIO_F(_MMIO(_DPC_AUX_CH_CTL), 6 * 4, 0, 0, 0, D_SKL_PLUS, NULL,
+	MMIO_F(DP_AUX_CH_CTL(AUX_CH_C), 6 * 4, 0, 0, 0, D_SKL_PLUS, NULL,
 						dp_aux_ch_ctl_mmio_write);
-	MMIO_F(_MMIO(_DPD_AUX_CH_CTL), 6 * 4, 0, 0, 0, D_SKL_PLUS, NULL,
+	MMIO_F(DP_AUX_CH_CTL(AUX_CH_D), 6 * 4, 0, 0, 0, D_SKL_PLUS, NULL,
 						dp_aux_ch_ctl_mmio_write);
 
 	MMIO_D(HSW_PWR_WELL_CTL1, D_SKL_PLUS);
@@ -3417,6 +3420,10 @@ int intel_gvt_for_each_tracked_mmio(struct intel_gvt *gvt,
 	}
 
 	for (i = 0; i < gvt->mmio.num_mmio_block; i++, block++) {
+		/* pvinfo data doesn't come from hw mmio */
+		if (i915_mmio_reg_offset(block->offset) == VGT_PVINFO_PAGE)
+			continue;
+
 		for (j = 0; j < block->size; j += 4) {
 			ret = handler(gvt,
 				      i915_mmio_reg_offset(block->offset) + j,
diff --git a/drivers/gpu/drm/i915/gvt/kvmgt.c b/drivers/gpu/drm/i915/gvt/kvmgt.c
index 343d79c1cb7e..04a5a0d90823 100644
--- a/drivers/gpu/drm/i915/gvt/kvmgt.c
+++ b/drivers/gpu/drm/i915/gvt/kvmgt.c
@@ -1564,27 +1564,10 @@ vgpu_id_show(struct device *dev, struct device_attribute *attr,
 	return sprintf(buf, "\n");
 }
 
-static ssize_t
-hw_id_show(struct device *dev, struct device_attribute *attr,
-	   char *buf)
-{
-	struct mdev_device *mdev = mdev_from_dev(dev);
-
-	if (mdev) {
-		struct intel_vgpu *vgpu = (struct intel_vgpu *)
-			mdev_get_drvdata(mdev);
-		return sprintf(buf, "%u\n",
-			       vgpu->submission.shadow[0]->gem_context->hw_id);
-	}
-	return sprintf(buf, "\n");
-}
-
 static DEVICE_ATTR_RO(vgpu_id);
-static DEVICE_ATTR_RO(hw_id);
 
 static struct attribute *intel_vgpu_attrs[] = {
 	&dev_attr_vgpu_id.attr,
-	&dev_attr_hw_id.attr,
 	NULL
 };
 
diff --git a/drivers/gpu/drm/i915/gvt/mmio_context.c b/drivers/gpu/drm/i915/gvt/mmio_context.c
index 4208e40445b1..aaf15916d29a 100644
--- a/drivers/gpu/drm/i915/gvt/mmio_context.c
+++ b/drivers/gpu/drm/i915/gvt/mmio_context.c
@@ -35,6 +35,7 @@
 
 #include "i915_drv.h"
 #include "gt/intel_context.h"
+#include "gt/intel_ring.h"
 #include "gvt.h"
 #include "trace.h"
 
diff --git a/drivers/gpu/drm/i915/gvt/scheduler.c b/drivers/gpu/drm/i915/gvt/scheduler.c
index 6c79d16b381e..5b2a7d072ec9 100644
--- a/drivers/gpu/drm/i915/gvt/scheduler.c
+++ b/drivers/gpu/drm/i915/gvt/scheduler.c
@@ -38,6 +38,7 @@
 #include "gem/i915_gem_context.h"
 #include "gem/i915_gem_pm.h"
 #include "gt/intel_context.h"
+#include "gt/intel_ring.h"
 
 #include "i915_drv.h"
 #include "gvt.h"
@@ -194,7 +195,7 @@ static int populate_shadow_context(struct intel_vgpu_workload *workload)
 			return -EFAULT;
 		}
 
-		page = i915_gem_object_get_page(ctx_obj, LRC_HEADER_PAGES + i);
+		page = i915_gem_object_get_page(ctx_obj, i);
 		dst = kmap(page);
 		intel_gvt_hypervisor_read_gpa(vgpu, context_gpa, dst,
 				I915_GTT_PAGE_SIZE);
@@ -365,7 +366,8 @@ static void set_context_ppgtt_from_shadow(struct intel_vgpu_workload *workload,
 					  struct i915_gem_context *ctx)
 {
 	struct intel_vgpu_mm *mm = workload->shadow_mm;
-	struct i915_ppgtt *ppgtt = i915_vm_to_ppgtt(ctx->vm);
+	struct i915_ppgtt *ppgtt =
+		i915_vm_to_ppgtt(i915_gem_context_get_vm_rcu(ctx));
 	int i = 0;
 
 	if (mm->ppgtt_mm.root_entry_type == GTT_TYPE_PPGTT_ROOT_L4_ENTRY) {
@@ -378,6 +380,8 @@ static void set_context_ppgtt_from_shadow(struct intel_vgpu_workload *workload,
 			px_dma(pd) = mm->ppgtt_mm.shadow_pdps[i];
 		}
 	}
+
+	i915_vm_put(&ppgtt->vm);
 }
 
 static int
@@ -385,11 +389,8 @@ intel_gvt_workload_req_alloc(struct intel_vgpu_workload *workload)
 {
 	struct intel_vgpu *vgpu = workload->vgpu;
 	struct intel_vgpu_submission *s = &vgpu->submission;
-	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
 	struct i915_request *rq;
 
-	lockdep_assert_held(&dev_priv->drm.struct_mutex);
-
 	if (workload->req)
 		return 0;
 
@@ -415,10 +416,9 @@ int intel_gvt_scan_and_shadow_workload(struct intel_vgpu_workload *workload)
 {
 	struct intel_vgpu *vgpu = workload->vgpu;
 	struct intel_vgpu_submission *s = &vgpu->submission;
-	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
 	int ret;
 
-	lockdep_assert_held(&dev_priv->drm.struct_mutex);
+	lockdep_assert_held(&vgpu->vgpu_lock);
 
 	if (workload->shadow)
 		return 0;
@@ -580,8 +580,6 @@ static void update_vreg_in_ctx(struct intel_vgpu_workload *workload)
 
 static void release_shadow_batch_buffer(struct intel_vgpu_workload *workload)
 {
-	struct intel_vgpu *vgpu = workload->vgpu;
-	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
 	struct intel_vgpu_shadow_bb *bb, *pos;
 
 	if (list_empty(&workload->shadow_bb))
@@ -590,8 +588,6 @@ static void release_shadow_batch_buffer(struct intel_vgpu_workload *workload)
 	bb = list_first_entry(&workload->shadow_bb,
 			struct intel_vgpu_shadow_bb, list);
 
-	mutex_lock(&dev_priv->drm.struct_mutex);
-
 	list_for_each_entry_safe(bb, pos, &workload->shadow_bb, list) {
 		if (bb->obj) {
 			if (bb->accessing)
@@ -609,8 +605,6 @@ static void release_shadow_batch_buffer(struct intel_vgpu_workload *workload)
 		list_del(&bb->list);
 		kfree(bb);
 	}
-
-	mutex_unlock(&dev_priv->drm.struct_mutex);
 }
 
 static int prepare_workload(struct intel_vgpu_workload *workload)
@@ -685,7 +679,6 @@ err_unpin_mm:
 static int dispatch_workload(struct intel_vgpu_workload *workload)
 {
 	struct intel_vgpu *vgpu = workload->vgpu;
-	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
 	struct i915_request *rq;
 	int ring_id = workload->ring_id;
 	int ret;
@@ -694,7 +687,6 @@ static int dispatch_workload(struct intel_vgpu_workload *workload)
 		ring_id, workload);
 
 	mutex_lock(&vgpu->vgpu_lock);
-	mutex_lock(&dev_priv->drm.struct_mutex);
 
 	ret = intel_gvt_workload_req_alloc(workload);
 	if (ret)
@@ -729,7 +721,6 @@ out:
 err_req:
 	if (ret)
 		workload->status = ret;
-	mutex_unlock(&dev_priv->drm.struct_mutex);
 	mutex_unlock(&vgpu->vgpu_lock);
 	return ret;
 }
@@ -844,7 +835,7 @@ static void update_guest_context(struct intel_vgpu_workload *workload)
 			return;
 		}
 
-		page = i915_gem_object_get_page(ctx_obj, LRC_HEADER_PAGES + i);
+		page = i915_gem_object_get_page(ctx_obj, i);
 		src = kmap(page);
 		intel_gvt_hypervisor_write_gpa(vgpu, context_gpa, src,
 				I915_GTT_PAGE_SIZE);
@@ -887,7 +878,7 @@ void intel_vgpu_clean_workloads(struct intel_vgpu *vgpu,
 	intel_engine_mask_t tmp;
 
 	/* free the unsubmited workloads in the queues. */
-	for_each_engine_masked(engine, dev_priv, engine_mask, tmp) {
+	for_each_engine_masked(engine, &dev_priv->gt, engine_mask, tmp) {
 		list_for_each_entry_safe(pos, n,
 			&s->workload_q_head[engine->id], list) {
 			list_del_init(&pos->list);
@@ -1233,20 +1224,18 @@ int intel_vgpu_setup_submission(struct intel_vgpu *vgpu)
 	struct intel_vgpu_submission *s = &vgpu->submission;
 	struct intel_engine_cs *engine;
 	struct i915_gem_context *ctx;
+	struct i915_ppgtt *ppgtt;
 	enum intel_engine_id i;
 	int ret;
 
-	mutex_lock(&i915->drm.struct_mutex);
-
 	ctx = i915_gem_context_create_kernel(i915, I915_PRIORITY_MAX);
-	if (IS_ERR(ctx)) {
-		ret = PTR_ERR(ctx);
-		goto out_unlock;
-	}
+	if (IS_ERR(ctx))
+		return PTR_ERR(ctx);
 
 	i915_gem_context_set_force_single_submission(ctx);
 
-	i915_context_ppgtt_root_save(s, i915_vm_to_ppgtt(ctx->vm));
+	ppgtt = i915_vm_to_ppgtt(i915_gem_context_get_vm_rcu(ctx));
+	i915_context_ppgtt_root_save(s, ppgtt);
 
 	for_each_engine(engine, i915, i) {
 		struct intel_context *ce;
@@ -1291,12 +1280,12 @@ int intel_vgpu_setup_submission(struct intel_vgpu *vgpu)
 	atomic_set(&s->running_workload_num, 0);
 	bitmap_zero(s->tlb_handle_pending, I915_NUM_ENGINES);
 
+	i915_vm_put(&ppgtt->vm);
 	i915_gem_context_put(ctx);
-	mutex_unlock(&i915->drm.struct_mutex);
 	return 0;
 
 out_shadow_ctx:
-	i915_context_ppgtt_root_restore(s, i915_vm_to_ppgtt(ctx->vm));
+	i915_context_ppgtt_root_restore(s, ppgtt);
 	for_each_engine(engine, i915, i) {
 		if (IS_ERR(s->shadow[i]))
 			break;
@@ -1304,9 +1293,8 @@ out_shadow_ctx:
 		intel_context_unpin(s->shadow[i]);
 		intel_context_put(s->shadow[i]);
 	}
+	i915_vm_put(&ppgtt->vm);
 	i915_gem_context_put(ctx);
-out_unlock:
-	mutex_unlock(&i915->drm.struct_mutex);
 	return ret;
 }
 
@@ -1597,9 +1585,7 @@ intel_vgpu_create_workload(struct intel_vgpu *vgpu, int ring_id,
 	 */
 	if (list_empty(workload_q_head(vgpu, ring_id))) {
 		intel_runtime_pm_get(&dev_priv->runtime_pm);
-		mutex_lock(&dev_priv->drm.struct_mutex);
 		ret = intel_gvt_scan_and_shadow_workload(workload);
-		mutex_unlock(&dev_priv->drm.struct_mutex);
 		intel_runtime_pm_put_unchecked(&dev_priv->runtime_pm);
 	}
 
diff --git a/drivers/gpu/drm/i915/i915_active.c b/drivers/gpu/drm/i915/i915_active.c
index 48e16ad93bbd..3c424cb90702 100644
--- a/drivers/gpu/drm/i915/i915_active.c
+++ b/drivers/gpu/drm/i915/i915_active.c
@@ -7,13 +7,12 @@
 #include <linux/debugobjects.h>
 
 #include "gt/intel_engine_pm.h"
+#include "gt/intel_ring.h"
 
 #include "i915_drv.h"
 #include "i915_active.h"
 #include "i915_globals.h"
 
-#define BKL(ref) (&(ref)->i915->drm.struct_mutex)
-
 /*
  * Active refs memory management
  *
@@ -27,35 +26,35 @@ static struct i915_global_active {
 } global;
 
 struct active_node {
-	struct i915_active_request base;
+	struct i915_active_fence base;
 	struct i915_active *ref;
 	struct rb_node node;
 	u64 timeline;
 };
 
 static inline struct active_node *
-node_from_active(struct i915_active_request *active)
+node_from_active(struct i915_active_fence *active)
 {
 	return container_of(active, struct active_node, base);
 }
 
 #define take_preallocated_barriers(x) llist_del_all(&(x)->preallocated_barriers)
 
-static inline bool is_barrier(const struct i915_active_request *active)
+static inline bool is_barrier(const struct i915_active_fence *active)
 {
-	return IS_ERR(rcu_access_pointer(active->request));
+	return IS_ERR(rcu_access_pointer(active->fence));
 }
 
 static inline struct llist_node *barrier_to_ll(struct active_node *node)
 {
 	GEM_BUG_ON(!is_barrier(&node->base));
-	return (struct llist_node *)&node->base.link;
+	return (struct llist_node *)&node->base.cb.node;
 }
 
 static inline struct intel_engine_cs *
 __barrier_to_engine(struct active_node *node)
 {
-	return (struct intel_engine_cs *)READ_ONCE(node->base.link.prev);
+	return (struct intel_engine_cs *)READ_ONCE(node->base.cb.node.prev);
 }
 
 static inline struct intel_engine_cs *
@@ -68,7 +67,7 @@ barrier_to_engine(struct active_node *node)
 static inline struct active_node *barrier_from_ll(struct llist_node *x)
 {
 	return container_of((struct list_head *)x,
-			    struct active_node, base.link);
+			    struct active_node, base.cb.node);
 }
 
 #if IS_ENABLED(CONFIG_DRM_I915_DEBUG_GEM) && IS_ENABLED(CONFIG_DEBUG_OBJECTS)
@@ -92,12 +91,17 @@ static void debug_active_init(struct i915_active *ref)
 
 static void debug_active_activate(struct i915_active *ref)
 {
-	debug_object_activate(ref, &active_debug_desc);
+	spin_lock_irq(&ref->tree_lock);
+	if (!atomic_read(&ref->count)) /* before the first inc */
+		debug_object_activate(ref, &active_debug_desc);
+	spin_unlock_irq(&ref->tree_lock);
 }
 
 static void debug_active_deactivate(struct i915_active *ref)
 {
-	debug_object_deactivate(ref, &active_debug_desc);
+	lockdep_assert_held(&ref->tree_lock);
+	if (!atomic_read(&ref->count)) /* after the last dec */
+		debug_object_deactivate(ref, &active_debug_desc);
 }
 
 static void debug_active_fini(struct i915_active *ref)
@@ -125,31 +129,46 @@ __active_retire(struct i915_active *ref)
 {
 	struct active_node *it, *n;
 	struct rb_root root;
-	bool retire = false;
+	unsigned long flags;
 
-	lockdep_assert_held(&ref->mutex);
+	GEM_BUG_ON(i915_active_is_idle(ref));
 
 	/* return the unused nodes to our slabcache -- flushing the allocator */
-	if (atomic_dec_and_test(&ref->count)) {
-		debug_active_deactivate(ref);
-		root = ref->tree;
-		ref->tree = RB_ROOT;
-		ref->cache = NULL;
-		retire = true;
-	}
-
-	mutex_unlock(&ref->mutex);
-	if (!retire)
+	if (!atomic_dec_and_lock_irqsave(&ref->count, &ref->tree_lock, flags))
 		return;
 
-	rbtree_postorder_for_each_entry_safe(it, n, &root, node) {
-		GEM_BUG_ON(i915_active_request_isset(&it->base));
-		kmem_cache_free(global.slab_cache, it);
-	}
+	GEM_BUG_ON(rcu_access_pointer(ref->excl.fence));
+	debug_active_deactivate(ref);
+
+	root = ref->tree;
+	ref->tree = RB_ROOT;
+	ref->cache = NULL;
+
+	spin_unlock_irqrestore(&ref->tree_lock, flags);
 
 	/* After the final retire, the entire struct may be freed */
 	if (ref->retire)
 		ref->retire(ref);
+
+	/* ... except if you wait on it, you must manage your own references! */
+	wake_up_var(ref);
+
+	rbtree_postorder_for_each_entry_safe(it, n, &root, node) {
+		GEM_BUG_ON(i915_active_fence_isset(&it->base));
+		kmem_cache_free(global.slab_cache, it);
+	}
+}
+
+static void
+active_work(struct work_struct *wrk)
+{
+	struct i915_active *ref = container_of(wrk, typeof(*ref), work);
+
+	GEM_BUG_ON(!atomic_read(&ref->count));
+	if (atomic_add_unless(&ref->count, -1, 1))
+		return;
+
+	__active_retire(ref);
 }
 
 static void
@@ -159,18 +178,29 @@ active_retire(struct i915_active *ref)
 	if (atomic_add_unless(&ref->count, -1, 1))
 		return;
 
-	/* One active may be flushed from inside the acquire of another */
-	mutex_lock_nested(&ref->mutex, SINGLE_DEPTH_NESTING);
+	if (ref->flags & I915_ACTIVE_RETIRE_SLEEPS) {
+		queue_work(system_unbound_wq, &ref->work);
+		return;
+	}
+
 	__active_retire(ref);
 }
 
 static void
-node_retire(struct i915_active_request *base, struct i915_request *rq)
+node_retire(struct dma_fence *fence, struct dma_fence_cb *cb)
 {
-	active_retire(node_from_active(base)->ref);
+	i915_active_fence_cb(fence, cb);
+	active_retire(container_of(cb, struct active_node, base.cb)->ref);
 }
 
-static struct i915_active_request *
+static void
+excl_retire(struct dma_fence *fence, struct dma_fence_cb *cb)
+{
+	i915_active_fence_cb(fence, cb);
+	active_retire(container_of(cb, struct i915_active, excl.cb));
+}
+
+static struct i915_active_fence *
 active_instance(struct i915_active *ref, struct intel_timeline *tl)
 {
 	struct active_node *node, *prealloc;
@@ -193,7 +223,7 @@ active_instance(struct i915_active *ref, struct intel_timeline *tl)
 	if (!prealloc)
 		return NULL;
 
-	mutex_lock(&ref->mutex);
+	spin_lock_irq(&ref->tree_lock);
 	GEM_BUG_ON(i915_active_is_idle(ref));
 
 	parent = NULL;
@@ -214,7 +244,7 @@ active_instance(struct i915_active *ref, struct intel_timeline *tl)
 	}
 
 	node = prealloc;
-	i915_active_request_init(&node->base, &tl->mutex, NULL, node_retire);
+	__i915_active_fence_init(&node->base, &tl->mutex, NULL, node_retire);
 	node->ref = ref;
 	node->timeline = idx;
 
@@ -223,29 +253,36 @@ active_instance(struct i915_active *ref, struct intel_timeline *tl)
 
 out:
 	ref->cache = node;
-	mutex_unlock(&ref->mutex);
+	spin_unlock_irq(&ref->tree_lock);
 
 	BUILD_BUG_ON(offsetof(typeof(*node), base));
 	return &node->base;
 }
 
-void __i915_active_init(struct drm_i915_private *i915,
-			struct i915_active *ref,
+void __i915_active_init(struct i915_active *ref,
 			int (*active)(struct i915_active *ref),
 			void (*retire)(struct i915_active *ref),
 			struct lock_class_key *key)
 {
+	unsigned long bits;
+
 	debug_active_init(ref);
 
-	ref->i915 = i915;
 	ref->flags = 0;
 	ref->active = active;
-	ref->retire = retire;
+	ref->retire = ptr_unpack_bits(retire, &bits, 2);
+	if (bits & I915_ACTIVE_MAY_SLEEP)
+		ref->flags |= I915_ACTIVE_RETIRE_SLEEPS;
+
+	spin_lock_init(&ref->tree_lock);
 	ref->tree = RB_ROOT;
 	ref->cache = NULL;
+
 	init_llist_head(&ref->preallocated_barriers);
 	atomic_set(&ref->count, 0);
 	__mutex_init(&ref->mutex, "i915_active", key);
+	__i915_active_fence_init(&ref->excl, &ref->mutex, NULL, excl_retire);
+	INIT_WORK(&ref->work, active_work);
 }
 
 static bool ____active_del_barrier(struct i915_active *ref,
@@ -298,9 +335,9 @@ __active_del_barrier(struct i915_active *ref, struct active_node *node)
 
 int i915_active_ref(struct i915_active *ref,
 		    struct intel_timeline *tl,
-		    struct i915_request *rq)
+		    struct dma_fence *fence)
 {
-	struct i915_active_request *active;
+	struct i915_active_fence *active;
 	int err;
 
 	lockdep_assert_held(&tl->mutex);
@@ -323,26 +360,44 @@ int i915_active_ref(struct i915_active *ref,
 		 * request that we want to emit on the kernel_context.
 		 */
 		__active_del_barrier(ref, node_from_active(active));
-		RCU_INIT_POINTER(active->request, NULL);
-		INIT_LIST_HEAD(&active->link);
-	} else {
-		if (!i915_active_request_isset(active))
-			atomic_inc(&ref->count);
+		RCU_INIT_POINTER(active->fence, NULL);
+		atomic_dec(&ref->count);
 	}
-	GEM_BUG_ON(!atomic_read(&ref->count));
-	__i915_active_request_set(active, rq);
+	if (!__i915_active_fence_set(active, fence))
+		atomic_inc(&ref->count);
 
 out:
 	i915_active_release(ref);
 	return err;
 }
 
+void i915_active_set_exclusive(struct i915_active *ref, struct dma_fence *f)
+{
+	/* We expect the caller to manage the exclusive timeline ordering */
+	GEM_BUG_ON(i915_active_is_idle(ref));
+
+	/*
+	 * As we don't know which mutex the caller is using, we told a small
+	 * lie to the debug code that it is using the i915_active.mutex;
+	 * and now we must stick to that lie.
+	 */
+	mutex_acquire(&ref->mutex.dep_map, 0, 0, _THIS_IP_);
+	if (!__i915_active_fence_set(&ref->excl, f))
+		atomic_inc(&ref->count);
+	mutex_release(&ref->mutex.dep_map, _THIS_IP_);
+}
+
+bool i915_active_acquire_if_busy(struct i915_active *ref)
+{
+	debug_active_assert(ref);
+	return atomic_add_unless(&ref->count, 1, 0);
+}
+
 int i915_active_acquire(struct i915_active *ref)
 {
 	int err;
 
-	debug_active_assert(ref);
-	if (atomic_add_unless(&ref->count, 1, 0))
+	if (i915_active_acquire_if_busy(ref))
 		return 0;
 
 	err = mutex_lock_interruptible(&ref->mutex);
@@ -367,109 +422,66 @@ void i915_active_release(struct i915_active *ref)
 	active_retire(ref);
 }
 
-static void __active_ungrab(struct i915_active *ref)
-{
-	clear_and_wake_up_bit(I915_ACTIVE_GRAB_BIT, &ref->flags);
-}
-
-bool i915_active_trygrab(struct i915_active *ref)
+static void enable_signaling(struct i915_active_fence *active)
 {
-	debug_active_assert(ref);
-
-	if (test_and_set_bit(I915_ACTIVE_GRAB_BIT, &ref->flags))
-		return false;
+	struct dma_fence *fence;
 
-	if (!atomic_add_unless(&ref->count, 1, 0)) {
-		__active_ungrab(ref);
-		return false;
-	}
+	fence = i915_active_fence_get(active);
+	if (!fence)
+		return;
 
-	return true;
-}
-
-void i915_active_ungrab(struct i915_active *ref)
-{
-	GEM_BUG_ON(!test_bit(I915_ACTIVE_GRAB_BIT, &ref->flags));
-
-	active_retire(ref);
-	__active_ungrab(ref);
+	dma_fence_enable_sw_signaling(fence);
+	dma_fence_put(fence);
 }
 
 int i915_active_wait(struct i915_active *ref)
 {
 	struct active_node *it, *n;
-	int err;
+	int err = 0;
 
 	might_sleep();
-	might_lock(&ref->mutex);
 
-	if (i915_active_is_idle(ref))
+	if (!i915_active_acquire_if_busy(ref))
 		return 0;
 
-	err = mutex_lock_interruptible(&ref->mutex);
-	if (err)
-		return err;
-
-	if (!atomic_add_unless(&ref->count, 1, 0)) {
-		mutex_unlock(&ref->mutex);
-		return 0;
-	}
-
+	/* Flush lazy signals */
+	enable_signaling(&ref->excl);
 	rbtree_postorder_for_each_entry_safe(it, n, &ref->tree, node) {
-		if (is_barrier(&it->base)) { /* unconnected idle-barrier */
-			err = -EBUSY;
-			break;
-		}
+		if (is_barrier(&it->base)) /* unconnected idle barrier */
+			continue;
 
-		err = i915_active_request_retire(&it->base, BKL(ref));
-		if (err)
-			break;
+		enable_signaling(&it->base);
 	}
+	/* Any fence added after the wait begins will not be auto-signaled */
 
-	__active_retire(ref);
+	i915_active_release(ref);
 	if (err)
 		return err;
 
-	if (wait_on_bit(&ref->flags, I915_ACTIVE_GRAB_BIT, TASK_KILLABLE))
+	if (wait_var_event_interruptible(ref, i915_active_is_idle(ref)))
 		return -EINTR;
 
-	if (!i915_active_is_idle(ref))
-		return -EBUSY;
-
 	return 0;
 }
 
-int i915_request_await_active_request(struct i915_request *rq,
-				      struct i915_active_request *active)
-{
-	struct i915_request *barrier =
-		i915_active_request_raw(active, &rq->i915->drm.struct_mutex);
-
-	return barrier ? i915_request_await_dma_fence(rq, &barrier->fence) : 0;
-}
-
 int i915_request_await_active(struct i915_request *rq, struct i915_active *ref)
 {
-	struct active_node *it, *n;
-	int err;
-
-	if (RB_EMPTY_ROOT(&ref->tree))
-		return 0;
+	int err = 0;
 
-	/* await allocates and so we need to avoid hitting the shrinker */
-	err = i915_active_acquire(ref);
-	if (err)
-		return err;
+	if (rcu_access_pointer(ref->excl.fence)) {
+		struct dma_fence *fence;
 
-	mutex_lock(&ref->mutex);
-	rbtree_postorder_for_each_entry_safe(it, n, &ref->tree, node) {
-		err = i915_request_await_active_request(rq, &it->base);
-		if (err)
-			break;
+		rcu_read_lock();
+		fence = dma_fence_get_rcu_safe(&ref->excl.fence);
+		rcu_read_unlock();
+		if (fence) {
+			err = i915_request_await_dma_fence(rq, fence);
+			dma_fence_put(fence);
+		}
 	}
-	mutex_unlock(&ref->mutex);
 
-	i915_active_release(ref);
+	/* In the future we may choose to await on all fences */
+
 	return err;
 }
 
@@ -477,15 +489,16 @@ int i915_request_await_active(struct i915_request *rq, struct i915_active *ref)
 void i915_active_fini(struct i915_active *ref)
 {
 	debug_active_fini(ref);
-	GEM_BUG_ON(!RB_EMPTY_ROOT(&ref->tree));
 	GEM_BUG_ON(atomic_read(&ref->count));
+	GEM_BUG_ON(work_pending(&ref->work));
+	GEM_BUG_ON(!RB_EMPTY_ROOT(&ref->tree));
 	mutex_destroy(&ref->mutex);
 }
 #endif
 
 static inline bool is_idle_barrier(struct active_node *node, u64 idx)
 {
-	return node->timeline == idx && !i915_active_request_isset(&node->base);
+	return node->timeline == idx && !i915_active_fence_isset(&node->base);
 }
 
 static struct active_node *reuse_idle_barrier(struct i915_active *ref, u64 idx)
@@ -495,7 +508,7 @@ static struct active_node *reuse_idle_barrier(struct i915_active *ref, u64 idx)
 	if (RB_EMPTY_ROOT(&ref->tree))
 		return NULL;
 
-	mutex_lock(&ref->mutex);
+	spin_lock_irq(&ref->tree_lock);
 	GEM_BUG_ON(i915_active_is_idle(ref));
 
 	/*
@@ -560,7 +573,7 @@ static struct active_node *reuse_idle_barrier(struct i915_active *ref, u64 idx)
 			goto match;
 	}
 
-	mutex_unlock(&ref->mutex);
+	spin_unlock_irq(&ref->tree_lock);
 
 	return NULL;
 
@@ -568,7 +581,7 @@ match:
 	rb_erase(p, &ref->tree); /* Hide from waits and sibling allocations */
 	if (p == &ref->cache->node)
 		ref->cache = NULL;
-	mutex_unlock(&ref->mutex);
+	spin_unlock_irq(&ref->tree_lock);
 
 	return rb_entry(p, struct active_node, node);
 }
@@ -576,11 +589,12 @@ match:
 int i915_active_acquire_preallocate_barrier(struct i915_active *ref,
 					    struct intel_engine_cs *engine)
 {
-	struct drm_i915_private *i915 = engine->i915;
 	intel_engine_mask_t tmp, mask = engine->mask;
+	struct intel_gt *gt = engine->gt;
 	struct llist_node *pos, *next;
 	int err;
 
+	GEM_BUG_ON(i915_active_is_idle(ref));
 	GEM_BUG_ON(!llist_empty(&ref->preallocated_barriers));
 
 	/*
@@ -589,7 +603,7 @@ int i915_active_acquire_preallocate_barrier(struct i915_active *ref,
 	 * We can then use the preallocated nodes in
 	 * i915_active_acquire_barrier()
 	 */
-	for_each_engine_masked(engine, i915, mask, tmp) {
+	for_each_engine_masked(engine, gt, mask, tmp) {
 		u64 idx = engine->kernel_context->timeline->fence_context;
 		struct active_node *node;
 
@@ -605,13 +619,13 @@ int i915_active_acquire_preallocate_barrier(struct i915_active *ref,
 			node->base.lock =
 				&engine->kernel_context->timeline->mutex;
 #endif
-			RCU_INIT_POINTER(node->base.request, NULL);
-			node->base.retire = node_retire;
+			RCU_INIT_POINTER(node->base.fence, NULL);
+			node->base.cb.func = node_retire;
 			node->timeline = idx;
 			node->ref = ref;
 		}
 
-		if (!i915_active_request_isset(&node->base)) {
+		if (!i915_active_fence_isset(&node->base)) {
 			/*
 			 * Mark this as being *our* unconnected proto-node.
 			 *
@@ -621,8 +635,8 @@ int i915_active_acquire_preallocate_barrier(struct i915_active *ref,
 			 * and then we can use the rb_node and list pointers
 			 * for our tracking of the pending barrier.
 			 */
-			RCU_INIT_POINTER(node->base.request, ERR_PTR(-EAGAIN));
-			node->base.link.prev = (void *)engine;
+			RCU_INIT_POINTER(node->base.fence, ERR_PTR(-EAGAIN));
+			node->base.cb.node.prev = (void *)engine;
 			atomic_inc(&ref->count);
 		}
 
@@ -648,6 +662,7 @@ unwind:
 void i915_active_acquire_barrier(struct i915_active *ref)
 {
 	struct llist_node *pos, *next;
+	unsigned long flags;
 
 	GEM_BUG_ON(i915_active_is_idle(ref));
 
@@ -657,7 +672,7 @@ void i915_active_acquire_barrier(struct i915_active *ref)
 	 * populated by i915_request_add_active_barriers() to point to the
 	 * request that will eventually release them.
 	 */
-	mutex_lock_nested(&ref->mutex, SINGLE_DEPTH_NESTING);
+	spin_lock_irqsave_nested(&ref->tree_lock, flags, SINGLE_DEPTH_NESTING);
 	llist_for_each_safe(pos, next, take_preallocated_barriers(ref)) {
 		struct active_node *node = barrier_from_ll(pos);
 		struct intel_engine_cs *engine = barrier_to_engine(node);
@@ -679,54 +694,124 @@ void i915_active_acquire_barrier(struct i915_active *ref)
 		rb_link_node(&node->node, parent, p);
 		rb_insert_color(&node->node, &ref->tree);
 
+		GEM_BUG_ON(!intel_engine_pm_is_awake(engine));
 		llist_add(barrier_to_ll(node), &engine->barrier_tasks);
 		intel_engine_pm_put(engine);
 	}
-	mutex_unlock(&ref->mutex);
+	spin_unlock_irqrestore(&ref->tree_lock, flags);
 }
 
 void i915_request_add_active_barriers(struct i915_request *rq)
 {
 	struct intel_engine_cs *engine = rq->engine;
 	struct llist_node *node, *next;
+	unsigned long flags;
 
 	GEM_BUG_ON(intel_engine_is_virtual(engine));
-	GEM_BUG_ON(rq->timeline != engine->kernel_context->timeline);
+	GEM_BUG_ON(i915_request_timeline(rq) != engine->kernel_context->timeline);
 
+	node = llist_del_all(&engine->barrier_tasks);
+	if (!node)
+		return;
 	/*
 	 * Attach the list of proto-fences to the in-flight request such
 	 * that the parent i915_active will be released when this request
 	 * is retired.
 	 */
-	llist_for_each_safe(node, next, llist_del_all(&engine->barrier_tasks)) {
-		RCU_INIT_POINTER(barrier_from_ll(node)->base.request, rq);
+	spin_lock_irqsave(&rq->lock, flags);
+	llist_for_each_safe(node, next, node) {
+		RCU_INIT_POINTER(barrier_from_ll(node)->base.fence, &rq->fence);
 		smp_wmb(); /* serialise with reuse_idle_barrier */
-		list_add_tail((struct list_head *)node, &rq->active_list);
+		list_add_tail((struct list_head *)node, &rq->fence.cb_list);
 	}
+	spin_unlock_irqrestore(&rq->lock, flags);
 }
 
-int i915_active_request_set(struct i915_active_request *active,
-			    struct i915_request *rq)
+#if IS_ENABLED(CONFIG_DRM_I915_DEBUG_GEM)
+#define active_is_held(active) lockdep_is_held((active)->lock)
+#else
+#define active_is_held(active) true
+#endif
+
+/*
+ * __i915_active_fence_set: Update the last active fence along its timeline
+ * @active: the active tracker
+ * @fence: the new fence (under construction)
+ *
+ * Records the new @fence as the last active fence along its timeline in
+ * this active tracker, moving the tracking callbacks from the previous
+ * fence onto this one. Returns the previous fence (if not already completed),
+ * which the caller must ensure is executed before the new fence. To ensure
+ * that the order of fences within the timeline of the i915_active_fence is
+ * maintained, it must be locked by the caller.
+ */
+struct dma_fence *
+__i915_active_fence_set(struct i915_active_fence *active,
+			struct dma_fence *fence)
 {
-	int err;
+	struct dma_fence *prev;
+	unsigned long flags;
+
+	/* NB: must be serialised by an outer timeline mutex (active->lock) */
+	spin_lock_irqsave(fence->lock, flags);
+	GEM_BUG_ON(test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &fence->flags));
+
+	prev = rcu_dereference_protected(active->fence, active_is_held(active));
+	if (prev) {
+		GEM_BUG_ON(prev == fence);
+		spin_lock_nested(prev->lock, SINGLE_DEPTH_NESTING);
+		__list_del_entry(&active->cb.node);
+		spin_unlock(prev->lock); /* serialise with prev->cb_list */
+
+		/*
+		 * active->fence is reset by the callback from inside
+		 * interrupt context. We need to serialise our list
+		 * manipulation with the fence->lock to prevent the prev
+		 * being lost inside an interrupt (it can't be replaced as
+		 * no other caller is allowed to enter __i915_active_fence_set
+		 * as we hold the timeline lock). After serialising with
+		 * the callback, we need to double check which ran first,
+		 * our list_del() [decoupling prev from the callback] or
+		 * the callback...
+		 */
+		prev = rcu_access_pointer(active->fence);
+	}
+
+	rcu_assign_pointer(active->fence, fence);
+	list_add_tail(&active->cb.node, &fence->cb_list);
+
+	spin_unlock_irqrestore(fence->lock, flags);
+
+	return prev;
+}
+
+int i915_active_fence_set(struct i915_active_fence *active,
+			  struct i915_request *rq)
+{
+	struct dma_fence *fence;
+	int err = 0;
 
 #if IS_ENABLED(CONFIG_DRM_I915_DEBUG_GEM)
 	lockdep_assert_held(active->lock);
 #endif
 
-	/* Must maintain ordering wrt previous active requests */
-	err = i915_request_await_active_request(rq, active);
-	if (err)
-		return err;
+	/* Must maintain timeline ordering wrt previous active requests */
+	rcu_read_lock();
+	fence = __i915_active_fence_set(active, &rq->fence);
+	if (fence) /* but the previous fence may not belong to that timeline! */
+		fence = dma_fence_get_rcu(fence);
+	rcu_read_unlock();
+	if (fence) {
+		err = i915_request_await_dma_fence(rq, fence);
+		dma_fence_put(fence);
+	}
 
-	__i915_active_request_set(active, rq);
-	return 0;
+	return err;
 }
 
-void i915_active_retire_noop(struct i915_active_request *active,
-			     struct i915_request *request)
+void i915_active_noop(struct dma_fence *fence, struct dma_fence_cb *cb)
 {
-	/* Space left intentionally blank */
+	i915_active_fence_cb(fence, cb);
 }
 
 #if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
diff --git a/drivers/gpu/drm/i915/i915_active.h b/drivers/gpu/drm/i915/i915_active.h
index f95058f99057..44859356ce97 100644
--- a/drivers/gpu/drm/i915/i915_active.h
+++ b/drivers/gpu/drm/i915/i915_active.h
@@ -12,6 +12,10 @@
 #include "i915_active_types.h"
 #include "i915_request.h"
 
+struct i915_request;
+struct intel_engine_cs;
+struct intel_timeline;
+
 /*
  * We treat requests as fences. This is not be to confused with our
  * "fence registers" but pipeline synchronisation objects ala GL_ARB_sync.
@@ -28,308 +32,108 @@
  * write access so that we can perform concurrent read operations between
  * the CPU and GPU engines, as well as waiting for all rendering to
  * complete, or waiting for the last GPU user of a "fence register". The
- * object then embeds a #i915_active_request to track the most recent (in
+ * object then embeds a #i915_active_fence to track the most recent (in
  * retirement order) request relevant for the desired mode of access.
- * The #i915_active_request is updated with i915_active_request_set() to
+ * The #i915_active_fence is updated with i915_active_fence_set() to
  * track the most recent fence request, typically this is done as part of
  * i915_vma_move_to_active().
  *
- * When the #i915_active_request completes (is retired), it will
+ * When the #i915_active_fence completes (is retired), it will
  * signal its completion to the owner through a callback as well as mark
- * itself as idle (i915_active_request.request == NULL). The owner
+ * itself as idle (i915_active_fence.request == NULL). The owner
  * can then perform any action, such as delayed freeing of an active
  * resource including itself.
  */
 
-void i915_active_retire_noop(struct i915_active_request *active,
-			     struct i915_request *request);
+void i915_active_noop(struct dma_fence *fence, struct dma_fence_cb *cb);
 
 /**
- * i915_active_request_init - prepares the activity tracker for use
+ * __i915_active_fence_init - prepares the activity tracker for use
  * @active - the active tracker
- * @rq - initial request to track, can be NULL
+ * @fence - initial fence to track, can be NULL
  * @func - a callback when then the tracker is retired (becomes idle),
  *         can be NULL
  *
- * i915_active_request_init() prepares the embedded @active struct for use as
- * an activity tracker, that is for tracking the last known active request
- * associated with it. When the last request becomes idle, when it is retired
+ * i915_active_fence_init() prepares the embedded @active struct for use as
+ * an activity tracker, that is for tracking the last known active fence
+ * associated with it. When the last fence becomes idle, when it is retired
  * after completion, the optional callback @func is invoked.
  */
 static inline void
-i915_active_request_init(struct i915_active_request *active,
+__i915_active_fence_init(struct i915_active_fence *active,
 			 struct mutex *lock,
-			 struct i915_request *rq,
-			 i915_active_retire_fn retire)
+			 void *fence,
+			 dma_fence_func_t fn)
 {
-	RCU_INIT_POINTER(active->request, rq);
-	INIT_LIST_HEAD(&active->link);
-	active->retire = retire ?: i915_active_retire_noop;
+	RCU_INIT_POINTER(active->fence, fence);
+	active->cb.func = fn ?: i915_active_noop;
 #if IS_ENABLED(CONFIG_DRM_I915_DEBUG_GEM)
 	active->lock = lock;
 #endif
 }
 
-#define INIT_ACTIVE_REQUEST(name, lock) \
-	i915_active_request_init((name), (lock), NULL, NULL)
-
-/**
- * i915_active_request_set - updates the tracker to watch the current request
- * @active - the active tracker
- * @request - the request to watch
- *
- * __i915_active_request_set() watches the given @request for completion. Whilst
- * that @request is busy, the @active reports busy. When that @request is
- * retired, the @active tracker is updated to report idle.
- */
-static inline void
-__i915_active_request_set(struct i915_active_request *active,
-			  struct i915_request *request)
-{
-#if IS_ENABLED(CONFIG_DRM_I915_DEBUG_GEM)
-	lockdep_assert_held(active->lock);
-#endif
-	list_move(&active->link, &request->active_list);
-	rcu_assign_pointer(active->request, request);
-}
+#define INIT_ACTIVE_FENCE(A, LOCK) \
+	__i915_active_fence_init((A), (LOCK), NULL, NULL)
 
-int __must_check
-i915_active_request_set(struct i915_active_request *active,
-			struct i915_request *rq);
+struct dma_fence *
+__i915_active_fence_set(struct i915_active_fence *active,
+			struct dma_fence *fence);
 
 /**
- * i915_active_request_raw - return the active request
+ * i915_active_fence_set - updates the tracker to watch the current fence
  * @active - the active tracker
+ * @rq - the request to watch
  *
- * i915_active_request_raw() returns the current request being tracked, or NULL.
- * It does not obtain a reference on the request for the caller, so the caller
- * must hold struct_mutex.
+ * i915_active_fence_set() watches the given @rq for completion. While
+ * that @rq is busy, the @active reports busy. When that @rq is signaled
+ * (or else retired) the @active tracker is updated to report idle.
  */
-static inline struct i915_request *
-i915_active_request_raw(const struct i915_active_request *active,
-			struct mutex *mutex)
-{
-	return rcu_dereference_protected(active->request,
-					 lockdep_is_held(mutex));
-}
-
-/**
- * i915_active_request_peek - report the active request being monitored
- * @active - the active tracker
- *
- * i915_active_request_peek() returns the current request being tracked if
- * still active, or NULL. It does not obtain a reference on the request
- * for the caller, so the caller must hold struct_mutex.
- */
-static inline struct i915_request *
-i915_active_request_peek(const struct i915_active_request *active,
-			 struct mutex *mutex)
-{
-	struct i915_request *request;
-
-	request = i915_active_request_raw(active, mutex);
-	if (!request || i915_request_completed(request))
-		return NULL;
-
-	return request;
-}
-
-/**
- * i915_active_request_get - return a reference to the active request
- * @active - the active tracker
- *
- * i915_active_request_get() returns a reference to the active request, or NULL
- * if the active tracker is idle. The caller must hold struct_mutex.
- */
-static inline struct i915_request *
-i915_active_request_get(const struct i915_active_request *active,
-			struct mutex *mutex)
-{
-	return i915_request_get(i915_active_request_peek(active, mutex));
-}
-
-/**
- * __i915_active_request_get_rcu - return a reference to the active request
- * @active - the active tracker
- *
- * __i915_active_request_get() returns a reference to the active request,
- * or NULL if the active tracker is idle. The caller must hold the RCU read
- * lock, but the returned pointer is safe to use outside of RCU.
- */
-static inline struct i915_request *
-__i915_active_request_get_rcu(const struct i915_active_request *active)
-{
-	/*
-	 * Performing a lockless retrieval of the active request is super
-	 * tricky. SLAB_TYPESAFE_BY_RCU merely guarantees that the backing
-	 * slab of request objects will not be freed whilst we hold the
-	 * RCU read lock. It does not guarantee that the request itself
-	 * will not be freed and then *reused*. Viz,
-	 *
-	 * Thread A			Thread B
-	 *
-	 * rq = active.request
-	 *				retire(rq) -> free(rq);
-	 *				(rq is now first on the slab freelist)
-	 *				active.request = NULL
-	 *
-	 *				rq = new submission on a new object
-	 * ref(rq)
-	 *
-	 * To prevent the request from being reused whilst the caller
-	 * uses it, we take a reference like normal. Whilst acquiring
-	 * the reference we check that it is not in a destroyed state
-	 * (refcnt == 0). That prevents the request being reallocated
-	 * whilst the caller holds on to it. To check that the request
-	 * was not reallocated as we acquired the reference we have to
-	 * check that our request remains the active request across
-	 * the lookup, in the same manner as a seqlock. The visibility
-	 * of the pointer versus the reference counting is controlled
-	 * by using RCU barriers (rcu_dereference and rcu_assign_pointer).
-	 *
-	 * In the middle of all that, we inspect whether the request is
-	 * complete. Retiring is lazy so the request may be completed long
-	 * before the active tracker is updated. Querying whether the
-	 * request is complete is far cheaper (as it involves no locked
-	 * instructions setting cachelines to exclusive) than acquiring
-	 * the reference, so we do it first. The RCU read lock ensures the
-	 * pointer dereference is valid, but does not ensure that the
-	 * seqno nor HWS is the right one! However, if the request was
-	 * reallocated, that means the active tracker's request was complete.
-	 * If the new request is also complete, then both are and we can
-	 * just report the active tracker is idle. If the new request is
-	 * incomplete, then we acquire a reference on it and check that
-	 * it remained the active request.
-	 *
-	 * It is then imperative that we do not zero the request on
-	 * reallocation, so that we can chase the dangling pointers!
-	 * See i915_request_alloc().
-	 */
-	do {
-		struct i915_request *request;
-
-		request = rcu_dereference(active->request);
-		if (!request || i915_request_completed(request))
-			return NULL;
-
-		/*
-		 * An especially silly compiler could decide to recompute the
-		 * result of i915_request_completed, more specifically
-		 * re-emit the load for request->fence.seqno. A race would catch
-		 * a later seqno value, which could flip the result from true to
-		 * false. Which means part of the instructions below might not
-		 * be executed, while later on instructions are executed. Due to
-		 * barriers within the refcounting the inconsistency can't reach
-		 * past the call to i915_request_get_rcu, but not executing
-		 * that while still executing i915_request_put() creates
-		 * havoc enough.  Prevent this with a compiler barrier.
-		 */
-		barrier();
-
-		request = i915_request_get_rcu(request);
-
-		/*
-		 * What stops the following rcu_access_pointer() from occurring
-		 * before the above i915_request_get_rcu()? If we were
-		 * to read the value before pausing to get the reference to
-		 * the request, we may not notice a change in the active
-		 * tracker.
-		 *
-		 * The rcu_access_pointer() is a mere compiler barrier, which
-		 * means both the CPU and compiler are free to perform the
-		 * memory read without constraint. The compiler only has to
-		 * ensure that any operations after the rcu_access_pointer()
-		 * occur afterwards in program order. This means the read may
-		 * be performed earlier by an out-of-order CPU, or adventurous
-		 * compiler.
-		 *
-		 * The atomic operation at the heart of
-		 * i915_request_get_rcu(), see dma_fence_get_rcu(), is
-		 * atomic_inc_not_zero() which is only a full memory barrier
-		 * when successful. That is, if i915_request_get_rcu()
-		 * returns the request (and so with the reference counted
-		 * incremented) then the following read for rcu_access_pointer()
-		 * must occur after the atomic operation and so confirm
-		 * that this request is the one currently being tracked.
-		 *
-		 * The corresponding write barrier is part of
-		 * rcu_assign_pointer().
-		 */
-		if (!request || request == rcu_access_pointer(active->request))
-			return rcu_pointer_handoff(request);
-
-		i915_request_put(request);
-	} while (1);
-}
-
+int __must_check
+i915_active_fence_set(struct i915_active_fence *active,
+		      struct i915_request *rq);
 /**
- * i915_active_request_get_unlocked - return a reference to the active request
+ * i915_active_fence_get - return a reference to the active fence
  * @active - the active tracker
  *
- * i915_active_request_get_unlocked() returns a reference to the active request,
+ * i915_active_fence_get() returns a reference to the active fence,
  * or NULL if the active tracker is idle. The reference is obtained under RCU,
  * so no locking is required by the caller.
  *
- * The reference should be freed with i915_request_put().
+ * The reference should be freed with dma_fence_put().
  */
-static inline struct i915_request *
-i915_active_request_get_unlocked(const struct i915_active_request *active)
+static inline struct dma_fence *
+i915_active_fence_get(struct i915_active_fence *active)
 {
-	struct i915_request *request;
+	struct dma_fence *fence;
 
 	rcu_read_lock();
-	request = __i915_active_request_get_rcu(active);
+	fence = dma_fence_get_rcu_safe(&active->fence);
 	rcu_read_unlock();
 
-	return request;
+	return fence;
 }
 
 /**
- * i915_active_request_isset - report whether the active tracker is assigned
+ * i915_active_fence_isset - report whether the active tracker is assigned
  * @active - the active tracker
  *
- * i915_active_request_isset() returns true if the active tracker is currently
- * assigned to a request. Due to the lazy retiring, that request may be idle
+ * i915_active_fence_isset() returns true if the active tracker is currently
+ * assigned to a fence. Due to the lazy retiring, that fence may be idle
  * and this may report stale information.
  */
 static inline bool
-i915_active_request_isset(const struct i915_active_request *active)
+i915_active_fence_isset(const struct i915_active_fence *active)
 {
-	return rcu_access_pointer(active->request);
+	return rcu_access_pointer(active->fence);
 }
 
-/**
- * i915_active_request_retire - waits until the request is retired
- * @active - the active request on which to wait
- *
- * i915_active_request_retire() waits until the request is completed,
- * and then ensures that at least the retirement handler for this
- * @active tracker is called before returning. If the @active
- * tracker is idle, the function returns immediately.
- */
-static inline int __must_check
-i915_active_request_retire(struct i915_active_request *active,
-			   struct mutex *mutex)
+static inline void
+i915_active_fence_cb(struct dma_fence *fence, struct dma_fence_cb *cb)
 {
-	struct i915_request *request;
-	long ret;
-
-	request = i915_active_request_raw(active, mutex);
-	if (!request)
-		return 0;
-
-	ret = i915_request_wait(request,
-				I915_WAIT_INTERRUPTIBLE,
-				MAX_SCHEDULE_TIMEOUT);
-	if (ret < 0)
-		return ret;
+	struct i915_active_fence *active =
+		container_of(cb, typeof(*active), cb);
 
-	list_del_init(&active->link);
-	RCU_INIT_POINTER(active->request, NULL);
-
-	active->retire(active, request);
-
-	return 0;
+	RCU_INIT_POINTER(active->fence, NULL);
 }
 
 /*
@@ -358,34 +162,40 @@ i915_active_request_retire(struct i915_active_request *active,
  * synchronisation.
  */
 
-void __i915_active_init(struct drm_i915_private *i915,
-			struct i915_active *ref,
+void __i915_active_init(struct i915_active *ref,
 			int (*active)(struct i915_active *ref),
 			void (*retire)(struct i915_active *ref),
 			struct lock_class_key *key);
-#define i915_active_init(i915, ref, active, retire) do {		\
+#define i915_active_init(ref, active, retire) do {		\
 	static struct lock_class_key __key;				\
 									\
-	__i915_active_init(i915, ref, active, retire, &__key);		\
+	__i915_active_init(ref, active, retire, &__key);		\
 } while (0)
 
 int i915_active_ref(struct i915_active *ref,
 		    struct intel_timeline *tl,
-		    struct i915_request *rq);
+		    struct dma_fence *fence);
+
+static inline int
+i915_active_add_request(struct i915_active *ref, struct i915_request *rq)
+{
+	return i915_active_ref(ref, i915_request_timeline(rq), &rq->fence);
+}
+
+void i915_active_set_exclusive(struct i915_active *ref, struct dma_fence *f);
+
+static inline bool i915_active_has_exclusive(struct i915_active *ref)
+{
+	return rcu_access_pointer(ref->excl.fence);
+}
 
 int i915_active_wait(struct i915_active *ref);
 
-int i915_request_await_active(struct i915_request *rq,
-			      struct i915_active *ref);
-int i915_request_await_active_request(struct i915_request *rq,
-				      struct i915_active_request *active);
+int i915_request_await_active(struct i915_request *rq, struct i915_active *ref);
 
 int i915_active_acquire(struct i915_active *ref);
+bool i915_active_acquire_if_busy(struct i915_active *ref);
 void i915_active_release(struct i915_active *ref);
-void __i915_active_release_nested(struct i915_active *ref, int subclass);
-
-bool i915_active_trygrab(struct i915_active *ref);
-void i915_active_ungrab(struct i915_active *ref);
 
 static inline bool
 i915_active_is_idle(const struct i915_active *ref)
@@ -404,4 +214,6 @@ int i915_active_acquire_preallocate_barrier(struct i915_active *ref,
 void i915_active_acquire_barrier(struct i915_active *ref);
 void i915_request_add_active_barriers(struct i915_request *rq);
 
+void i915_active_print(struct i915_active *ref, struct drm_printer *m);
+
 #endif /* _I915_ACTIVE_H_ */
diff --git a/drivers/gpu/drm/i915/i915_active_types.h b/drivers/gpu/drm/i915/i915_active_types.h
index 1854e7d168c1..96aed0ee700a 100644
--- a/drivers/gpu/drm/i915/i915_active_types.h
+++ b/drivers/gpu/drm/i915/i915_active_types.h
@@ -8,22 +8,18 @@
 #define _I915_ACTIVE_TYPES_H_
 
 #include <linux/atomic.h>
+#include <linux/dma-fence.h>
 #include <linux/llist.h>
 #include <linux/mutex.h>
 #include <linux/rbtree.h>
 #include <linux/rcupdate.h>
+#include <linux/workqueue.h>
 
-struct drm_i915_private;
-struct i915_active_request;
-struct i915_request;
+#include "i915_utils.h"
 
-typedef void (*i915_active_retire_fn)(struct i915_active_request *,
-				      struct i915_request *);
-
-struct i915_active_request {
-	struct i915_request __rcu *request;
-	struct list_head link;
-	i915_active_retire_fn retire;
+struct i915_active_fence {
+	struct dma_fence __rcu *fence;
+	struct dma_fence_cb cb;
 #if IS_ENABLED(CONFIG_DRM_I915_DEBUG_GEM)
 	/*
 	 * Incorporeal!
@@ -43,20 +39,30 @@ struct i915_active_request {
 
 struct active_node;
 
+#define I915_ACTIVE_MAY_SLEEP BIT(0)
+
+#define __i915_active_call __aligned(4)
+#define i915_active_may_sleep(fn) ptr_pack_bits(&(fn), I915_ACTIVE_MAY_SLEEP, 2)
+
 struct i915_active {
-	struct drm_i915_private *i915;
+	atomic_t count;
+	struct mutex mutex;
 
+	spinlock_t tree_lock;
 	struct active_node *cache;
 	struct rb_root tree;
-	struct mutex mutex;
-	atomic_t count;
+
+	/* Preallocated "exclusive" node */
+	struct i915_active_fence excl;
 
 	unsigned long flags;
-#define I915_ACTIVE_GRAB_BIT 0
+#define I915_ACTIVE_RETIRE_SLEEPS BIT(0)
 
 	int (*active)(struct i915_active *ref);
 	void (*retire)(struct i915_active *ref);
 
+	struct work_struct work;
+
 	struct llist_head preallocated_barriers;
 };
 
diff --git a/drivers/gpu/drm/i915/i915_buddy.c b/drivers/gpu/drm/i915/i915_buddy.c
index fe1871d7c126..e9d4200ce3bc 100644
--- a/drivers/gpu/drm/i915/i915_buddy.c
+++ b/drivers/gpu/drm/i915/i915_buddy.c
@@ -38,6 +38,7 @@ int __init i915_global_buddy_init(void)
 	if (!global.slab_blocks)
 		return -ENOMEM;
 
+	i915_global_register(&global.base);
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/i915/i915_cmd_parser.c b/drivers/gpu/drm/i915/i915_cmd_parser.c
index 24555102e198..f24096e27bef 100644
--- a/drivers/gpu/drm/i915/i915_cmd_parser.c
+++ b/drivers/gpu/drm/i915/i915_cmd_parser.c
@@ -53,13 +53,11 @@
  * granting userspace undue privileges. There are three categories of privilege.
  *
  * First, commands which are explicitly defined as privileged or which should
- * only be used by the kernel driver. The parser generally rejects such
- * commands, though it may allow some from the drm master process.
+ * only be used by the kernel driver. The parser rejects such commands
  *
  * Second, commands which access registers. To support correct/enhanced
  * userspace functionality, particularly certain OpenGL extensions, the parser
- * provides a whitelist of registers which userspace may safely access (for both
- * normal and drm master processes).
+ * provides a whitelist of registers which userspace may safely access
  *
  * Third, commands which access privileged memory (i.e. GGTT, HWS page, etc).
  * The parser always rejects such commands.
@@ -84,9 +82,9 @@
  * in the per-engine command tables.
  *
  * Other command table entries map fairly directly to high level categories
- * mentioned above: rejected, master-only, register whitelist. The parser
- * implements a number of checks, including the privileged memory checks, via a
- * general bitmasking mechanism.
+ * mentioned above: rejected, register whitelist. The parser implements a number
+ * of checks, including the privileged memory checks, via a general bitmasking
+ * mechanism.
  */
 
 /*
@@ -104,8 +102,6 @@ struct drm_i915_cmd_descriptor {
 	 * CMD_DESC_REJECT: The command is never allowed
 	 * CMD_DESC_REGISTER: The command should be checked against the
 	 *                    register whitelist for the appropriate ring
-	 * CMD_DESC_MASTER: The command is allowed if the submitting process
-	 *                  is the DRM master
 	 */
 	u32 flags;
 #define CMD_DESC_FIXED    (1<<0)
@@ -113,7 +109,6 @@ struct drm_i915_cmd_descriptor {
 #define CMD_DESC_REJECT   (1<<2)
 #define CMD_DESC_REGISTER (1<<3)
 #define CMD_DESC_BITMASK  (1<<4)
-#define CMD_DESC_MASTER   (1<<5)
 
 	/*
 	 * The command's unique identification bits and the bitmask to get them.
@@ -194,7 +189,7 @@ struct drm_i915_cmd_table {
 #define CMD(op, opm, f, lm, fl, ...)				\
 	{							\
 		.flags = (fl) | ((f) ? CMD_DESC_FIXED : 0),	\
-		.cmd = { (op), ~0u << (opm) },			\
+		.cmd = { (op & ~0u << (opm)), ~0u << (opm) },	\
 		.length = { (lm) },				\
 		__VA_ARGS__					\
 	}
@@ -209,14 +204,13 @@ struct drm_i915_cmd_table {
 #define R CMD_DESC_REJECT
 #define W CMD_DESC_REGISTER
 #define B CMD_DESC_BITMASK
-#define M CMD_DESC_MASTER
 
 /*            Command                          Mask   Fixed Len   Action
 	      ---------------------------------------------------------- */
-static const struct drm_i915_cmd_descriptor common_cmds[] = {
+static const struct drm_i915_cmd_descriptor gen7_common_cmds[] = {
 	CMD(  MI_NOOP,                          SMI,    F,  1,      S  ),
 	CMD(  MI_USER_INTERRUPT,                SMI,    F,  1,      R  ),
-	CMD(  MI_WAIT_FOR_EVENT,                SMI,    F,  1,      M  ),
+	CMD(  MI_WAIT_FOR_EVENT,                SMI,    F,  1,      R  ),
 	CMD(  MI_ARB_CHECK,                     SMI,    F,  1,      S  ),
 	CMD(  MI_REPORT_HEAD,                   SMI,    F,  1,      S  ),
 	CMD(  MI_SUSPEND_FLUSH,                 SMI,    F,  1,      S  ),
@@ -246,7 +240,7 @@ static const struct drm_i915_cmd_descriptor common_cmds[] = {
 	CMD(  MI_BATCH_BUFFER_START,            SMI,   !F,  0xFF,   S  ),
 };
 
-static const struct drm_i915_cmd_descriptor render_cmds[] = {
+static const struct drm_i915_cmd_descriptor gen7_render_cmds[] = {
 	CMD(  MI_FLUSH,                         SMI,    F,  1,      S  ),
 	CMD(  MI_ARB_ON_OFF,                    SMI,    F,  1,      R  ),
 	CMD(  MI_PREDICATE,                     SMI,    F,  1,      S  ),
@@ -313,7 +307,7 @@ static const struct drm_i915_cmd_descriptor hsw_render_cmds[] = {
 	CMD(  MI_URB_ATOMIC_ALLOC,              SMI,    F,  1,      S  ),
 	CMD(  MI_SET_APPID,                     SMI,    F,  1,      S  ),
 	CMD(  MI_RS_CONTEXT,                    SMI,    F,  1,      S  ),
-	CMD(  MI_LOAD_SCAN_LINES_INCL,          SMI,   !F,  0x3F,   M  ),
+	CMD(  MI_LOAD_SCAN_LINES_INCL,          SMI,   !F,  0x3F,   R  ),
 	CMD(  MI_LOAD_SCAN_LINES_EXCL,          SMI,   !F,  0x3F,   R  ),
 	CMD(  MI_LOAD_REGISTER_REG,             SMI,   !F,  0xFF,   W,
 	      .reg = { .offset = 1, .mask = 0x007FFFFC, .step = 1 }    ),
@@ -330,7 +324,7 @@ static const struct drm_i915_cmd_descriptor hsw_render_cmds[] = {
 	CMD(  GFX_OP_3DSTATE_BINDING_TABLE_EDIT_PS,  S3D,   !F,  0x1FF,  S  ),
 };
 
-static const struct drm_i915_cmd_descriptor video_cmds[] = {
+static const struct drm_i915_cmd_descriptor gen7_video_cmds[] = {
 	CMD(  MI_ARB_ON_OFF,                    SMI,    F,  1,      R  ),
 	CMD(  MI_SET_APPID,                     SMI,    F,  1,      S  ),
 	CMD(  MI_STORE_DWORD_IMM,               SMI,   !F,  0xFF,   B,
@@ -374,7 +368,7 @@ static const struct drm_i915_cmd_descriptor video_cmds[] = {
 	CMD(  MFX_WAIT,                         SMFX,   F,  1,      S  ),
 };
 
-static const struct drm_i915_cmd_descriptor vecs_cmds[] = {
+static const struct drm_i915_cmd_descriptor gen7_vecs_cmds[] = {
 	CMD(  MI_ARB_ON_OFF,                    SMI,    F,  1,      R  ),
 	CMD(  MI_SET_APPID,                     SMI,    F,  1,      S  ),
 	CMD(  MI_STORE_DWORD_IMM,               SMI,   !F,  0xFF,   B,
@@ -412,7 +406,7 @@ static const struct drm_i915_cmd_descriptor vecs_cmds[] = {
 	      }},						       ),
 };
 
-static const struct drm_i915_cmd_descriptor blt_cmds[] = {
+static const struct drm_i915_cmd_descriptor gen7_blt_cmds[] = {
 	CMD(  MI_DISPLAY_FLIP,                  SMI,   !F,  0xFF,   R  ),
 	CMD(  MI_STORE_DWORD_IMM,               SMI,   !F,  0x3FF,  B,
 	      .bits = {{
@@ -446,10 +440,64 @@ static const struct drm_i915_cmd_descriptor blt_cmds[] = {
 };
 
 static const struct drm_i915_cmd_descriptor hsw_blt_cmds[] = {
-	CMD(  MI_LOAD_SCAN_LINES_INCL,          SMI,   !F,  0x3F,   M  ),
+	CMD(  MI_LOAD_SCAN_LINES_INCL,          SMI,   !F,  0x3F,   R  ),
 	CMD(  MI_LOAD_SCAN_LINES_EXCL,          SMI,   !F,  0x3F,   R  ),
 };
 
+/*
+ * For Gen9 we can still rely on the h/w to enforce cmd security, and only
+ * need to re-enforce the register access checks. We therefore only need to
+ * teach the cmdparser how to find the end of each command, and identify
+ * register accesses. The table doesn't need to reject any commands, and so
+ * the only commands listed here are:
+ *   1) Those that touch registers
+ *   2) Those that do not have the default 8-bit length
+ *
+ * Note that the default MI length mask chosen for this table is 0xFF, not
+ * the 0x3F used on older devices. This is because the vast majority of MI
+ * cmds on Gen9 use a standard 8-bit Length field.
+ * All the Gen9 blitter instructions are standard 0xFF length mask, and
+ * none allow access to non-general registers, so in fact no BLT cmds are
+ * included in the table at all.
+ *
+ */
+static const struct drm_i915_cmd_descriptor gen9_blt_cmds[] = {
+	CMD(  MI_NOOP,                          SMI,    F,  1,      S  ),
+	CMD(  MI_USER_INTERRUPT,                SMI,    F,  1,      S  ),
+	CMD(  MI_WAIT_FOR_EVENT,                SMI,    F,  1,      S  ),
+	CMD(  MI_FLUSH,                         SMI,    F,  1,      S  ),
+	CMD(  MI_ARB_CHECK,                     SMI,    F,  1,      S  ),
+	CMD(  MI_REPORT_HEAD,                   SMI,    F,  1,      S  ),
+	CMD(  MI_ARB_ON_OFF,                    SMI,    F,  1,      S  ),
+	CMD(  MI_SUSPEND_FLUSH,                 SMI,    F,  1,      S  ),
+	CMD(  MI_LOAD_SCAN_LINES_INCL,          SMI,   !F,  0x3F,   S  ),
+	CMD(  MI_LOAD_SCAN_LINES_EXCL,          SMI,   !F,  0x3F,   S  ),
+	CMD(  MI_STORE_DWORD_IMM,               SMI,   !F,  0x3FF,  S  ),
+	CMD(  MI_LOAD_REGISTER_IMM(1),          SMI,   !F,  0xFF,   W,
+	      .reg = { .offset = 1, .mask = 0x007FFFFC, .step = 2 }    ),
+	CMD(  MI_UPDATE_GTT,                    SMI,   !F,  0x3FF,  S  ),
+	CMD(  MI_STORE_REGISTER_MEM_GEN8,       SMI,    F,  4,      W,
+	      .reg = { .offset = 1, .mask = 0x007FFFFC }               ),
+	CMD(  MI_FLUSH_DW,                      SMI,   !F,  0x3F,   S  ),
+	CMD(  MI_LOAD_REGISTER_MEM_GEN8,        SMI,    F,  4,      W,
+	      .reg = { .offset = 1, .mask = 0x007FFFFC }               ),
+	CMD(  MI_LOAD_REGISTER_REG,             SMI,    !F,  0xFF,  W,
+	      .reg = { .offset = 1, .mask = 0x007FFFFC, .step = 1 }    ),
+
+	/*
+	 * We allow BB_START but apply further checks. We just sanitize the
+	 * basic fields here.
+	 */
+#define MI_BB_START_OPERAND_MASK   GENMASK(SMI-1, 0)
+#define MI_BB_START_OPERAND_EXPECT (MI_BATCH_PPGTT_HSW | 1)
+	CMD(  MI_BATCH_BUFFER_START_GEN8,       SMI,    !F,  0xFF,  B,
+	      .bits = {{
+			.offset = 0,
+			.mask = MI_BB_START_OPERAND_MASK,
+			.expected = MI_BB_START_OPERAND_EXPECT,
+	      }},						       ),
+};
+
 static const struct drm_i915_cmd_descriptor noop_desc =
 	CMD(MI_NOOP, SMI, F, 1, S);
 
@@ -463,40 +511,44 @@ static const struct drm_i915_cmd_descriptor noop_desc =
 #undef R
 #undef W
 #undef B
-#undef M
 
-static const struct drm_i915_cmd_table gen7_render_cmds[] = {
-	{ common_cmds, ARRAY_SIZE(common_cmds) },
-	{ render_cmds, ARRAY_SIZE(render_cmds) },
+static const struct drm_i915_cmd_table gen7_render_cmd_table[] = {
+	{ gen7_common_cmds, ARRAY_SIZE(gen7_common_cmds) },
+	{ gen7_render_cmds, ARRAY_SIZE(gen7_render_cmds) },
 };
 
-static const struct drm_i915_cmd_table hsw_render_ring_cmds[] = {
-	{ common_cmds, ARRAY_SIZE(common_cmds) },
-	{ render_cmds, ARRAY_SIZE(render_cmds) },
+static const struct drm_i915_cmd_table hsw_render_ring_cmd_table[] = {
+	{ gen7_common_cmds, ARRAY_SIZE(gen7_common_cmds) },
+	{ gen7_render_cmds, ARRAY_SIZE(gen7_render_cmds) },
 	{ hsw_render_cmds, ARRAY_SIZE(hsw_render_cmds) },
 };
 
-static const struct drm_i915_cmd_table gen7_video_cmds[] = {
-	{ common_cmds, ARRAY_SIZE(common_cmds) },
-	{ video_cmds, ARRAY_SIZE(video_cmds) },
+static const struct drm_i915_cmd_table gen7_video_cmd_table[] = {
+	{ gen7_common_cmds, ARRAY_SIZE(gen7_common_cmds) },
+	{ gen7_video_cmds, ARRAY_SIZE(gen7_video_cmds) },
 };
 
-static const struct drm_i915_cmd_table hsw_vebox_cmds[] = {
-	{ common_cmds, ARRAY_SIZE(common_cmds) },
-	{ vecs_cmds, ARRAY_SIZE(vecs_cmds) },
+static const struct drm_i915_cmd_table hsw_vebox_cmd_table[] = {
+	{ gen7_common_cmds, ARRAY_SIZE(gen7_common_cmds) },
+	{ gen7_vecs_cmds, ARRAY_SIZE(gen7_vecs_cmds) },
 };
 
-static const struct drm_i915_cmd_table gen7_blt_cmds[] = {
-	{ common_cmds, ARRAY_SIZE(common_cmds) },
-	{ blt_cmds, ARRAY_SIZE(blt_cmds) },
+static const struct drm_i915_cmd_table gen7_blt_cmd_table[] = {
+	{ gen7_common_cmds, ARRAY_SIZE(gen7_common_cmds) },
+	{ gen7_blt_cmds, ARRAY_SIZE(gen7_blt_cmds) },
 };
 
-static const struct drm_i915_cmd_table hsw_blt_ring_cmds[] = {
-	{ common_cmds, ARRAY_SIZE(common_cmds) },
-	{ blt_cmds, ARRAY_SIZE(blt_cmds) },
+static const struct drm_i915_cmd_table hsw_blt_ring_cmd_table[] = {
+	{ gen7_common_cmds, ARRAY_SIZE(gen7_common_cmds) },
+	{ gen7_blt_cmds, ARRAY_SIZE(gen7_blt_cmds) },
 	{ hsw_blt_cmds, ARRAY_SIZE(hsw_blt_cmds) },
 };
 
+static const struct drm_i915_cmd_table gen9_blt_cmd_table[] = {
+	{ gen9_blt_cmds, ARRAY_SIZE(gen9_blt_cmds) },
+};
+
+
 /*
  * Register whitelists, sorted by increasing register offset.
  */
@@ -612,17 +664,27 @@ static const struct drm_i915_reg_descriptor gen7_blt_regs[] = {
 	REG64_IDX(RING_TIMESTAMP, BLT_RING_BASE),
 };
 
-static const struct drm_i915_reg_descriptor ivb_master_regs[] = {
-	REG32(FORCEWAKE_MT),
-	REG32(DERRMR),
-	REG32(GEN7_PIPE_DE_LOAD_SL(PIPE_A)),
-	REG32(GEN7_PIPE_DE_LOAD_SL(PIPE_B)),
-	REG32(GEN7_PIPE_DE_LOAD_SL(PIPE_C)),
-};
-
-static const struct drm_i915_reg_descriptor hsw_master_regs[] = {
-	REG32(FORCEWAKE_MT),
-	REG32(DERRMR),
+static const struct drm_i915_reg_descriptor gen9_blt_regs[] = {
+	REG64_IDX(RING_TIMESTAMP, RENDER_RING_BASE),
+	REG64_IDX(RING_TIMESTAMP, BSD_RING_BASE),
+	REG32(BCS_SWCTRL),
+	REG64_IDX(RING_TIMESTAMP, BLT_RING_BASE),
+	REG64_IDX(BCS_GPR, 0),
+	REG64_IDX(BCS_GPR, 1),
+	REG64_IDX(BCS_GPR, 2),
+	REG64_IDX(BCS_GPR, 3),
+	REG64_IDX(BCS_GPR, 4),
+	REG64_IDX(BCS_GPR, 5),
+	REG64_IDX(BCS_GPR, 6),
+	REG64_IDX(BCS_GPR, 7),
+	REG64_IDX(BCS_GPR, 8),
+	REG64_IDX(BCS_GPR, 9),
+	REG64_IDX(BCS_GPR, 10),
+	REG64_IDX(BCS_GPR, 11),
+	REG64_IDX(BCS_GPR, 12),
+	REG64_IDX(BCS_GPR, 13),
+	REG64_IDX(BCS_GPR, 14),
+	REG64_IDX(BCS_GPR, 15),
 };
 
 #undef REG64
@@ -631,28 +693,27 @@ static const struct drm_i915_reg_descriptor hsw_master_regs[] = {
 struct drm_i915_reg_table {
 	const struct drm_i915_reg_descriptor *regs;
 	int num_regs;
-	bool master;
 };
 
 static const struct drm_i915_reg_table ivb_render_reg_tables[] = {
-	{ gen7_render_regs, ARRAY_SIZE(gen7_render_regs), false },
-	{ ivb_master_regs, ARRAY_SIZE(ivb_master_regs), true },
+	{ gen7_render_regs, ARRAY_SIZE(gen7_render_regs) },
 };
 
 static const struct drm_i915_reg_table ivb_blt_reg_tables[] = {
-	{ gen7_blt_regs, ARRAY_SIZE(gen7_blt_regs), false },
-	{ ivb_master_regs, ARRAY_SIZE(ivb_master_regs), true },
+	{ gen7_blt_regs, ARRAY_SIZE(gen7_blt_regs) },
 };
 
 static const struct drm_i915_reg_table hsw_render_reg_tables[] = {
-	{ gen7_render_regs, ARRAY_SIZE(gen7_render_regs), false },
-	{ hsw_render_regs, ARRAY_SIZE(hsw_render_regs), false },
-	{ hsw_master_regs, ARRAY_SIZE(hsw_master_regs), true },
+	{ gen7_render_regs, ARRAY_SIZE(gen7_render_regs) },
+	{ hsw_render_regs, ARRAY_SIZE(hsw_render_regs) },
 };
 
 static const struct drm_i915_reg_table hsw_blt_reg_tables[] = {
-	{ gen7_blt_regs, ARRAY_SIZE(gen7_blt_regs), false },
-	{ hsw_master_regs, ARRAY_SIZE(hsw_master_regs), true },
+	{ gen7_blt_regs, ARRAY_SIZE(gen7_blt_regs) },
+};
+
+static const struct drm_i915_reg_table gen9_blt_reg_tables[] = {
+	{ gen9_blt_regs, ARRAY_SIZE(gen9_blt_regs) },
 };
 
 static u32 gen7_render_get_cmd_length_mask(u32 cmd_header)
@@ -710,6 +771,17 @@ static u32 gen7_blt_get_cmd_length_mask(u32 cmd_header)
 	return 0;
 }
 
+static u32 gen9_blt_get_cmd_length_mask(u32 cmd_header)
+{
+	u32 client = cmd_header >> INSTR_CLIENT_SHIFT;
+
+	if (client == INSTR_MI_CLIENT || client == INSTR_BC_CLIENT)
+		return 0xFF;
+
+	DRM_DEBUG_DRIVER("CMD: Abnormal blt cmd length! 0x%08X\n", cmd_header);
+	return 0;
+}
+
 static bool validate_cmds_sorted(const struct intel_engine_cs *engine,
 				 const struct drm_i915_cmd_table *cmd_tables,
 				 int cmd_table_count)
@@ -867,18 +939,19 @@ void intel_engine_init_cmd_parser(struct intel_engine_cs *engine)
 	int cmd_table_count;
 	int ret;
 
-	if (!IS_GEN(engine->i915, 7))
+	if (!IS_GEN(engine->i915, 7) && !(IS_GEN(engine->i915, 9) &&
+					  engine->class == COPY_ENGINE_CLASS))
 		return;
 
 	switch (engine->class) {
 	case RENDER_CLASS:
 		if (IS_HASWELL(engine->i915)) {
-			cmd_tables = hsw_render_ring_cmds;
+			cmd_tables = hsw_render_ring_cmd_table;
 			cmd_table_count =
-				ARRAY_SIZE(hsw_render_ring_cmds);
+				ARRAY_SIZE(hsw_render_ring_cmd_table);
 		} else {
-			cmd_tables = gen7_render_cmds;
-			cmd_table_count = ARRAY_SIZE(gen7_render_cmds);
+			cmd_tables = gen7_render_cmd_table;
+			cmd_table_count = ARRAY_SIZE(gen7_render_cmd_table);
 		}
 
 		if (IS_HASWELL(engine->i915)) {
@@ -888,36 +961,46 @@ void intel_engine_init_cmd_parser(struct intel_engine_cs *engine)
 			engine->reg_tables = ivb_render_reg_tables;
 			engine->reg_table_count = ARRAY_SIZE(ivb_render_reg_tables);
 		}
-
 		engine->get_cmd_length_mask = gen7_render_get_cmd_length_mask;
 		break;
 	case VIDEO_DECODE_CLASS:
-		cmd_tables = gen7_video_cmds;
-		cmd_table_count = ARRAY_SIZE(gen7_video_cmds);
+		cmd_tables = gen7_video_cmd_table;
+		cmd_table_count = ARRAY_SIZE(gen7_video_cmd_table);
 		engine->get_cmd_length_mask = gen7_bsd_get_cmd_length_mask;
 		break;
 	case COPY_ENGINE_CLASS:
-		if (IS_HASWELL(engine->i915)) {
-			cmd_tables = hsw_blt_ring_cmds;
-			cmd_table_count = ARRAY_SIZE(hsw_blt_ring_cmds);
+		engine->get_cmd_length_mask = gen7_blt_get_cmd_length_mask;
+		if (IS_GEN(engine->i915, 9)) {
+			cmd_tables = gen9_blt_cmd_table;
+			cmd_table_count = ARRAY_SIZE(gen9_blt_cmd_table);
+			engine->get_cmd_length_mask =
+				gen9_blt_get_cmd_length_mask;
+
+			/* BCS Engine unsafe without parser */
+			engine->flags |= I915_ENGINE_REQUIRES_CMD_PARSER;
+		} else if (IS_HASWELL(engine->i915)) {
+			cmd_tables = hsw_blt_ring_cmd_table;
+			cmd_table_count = ARRAY_SIZE(hsw_blt_ring_cmd_table);
 		} else {
-			cmd_tables = gen7_blt_cmds;
-			cmd_table_count = ARRAY_SIZE(gen7_blt_cmds);
+			cmd_tables = gen7_blt_cmd_table;
+			cmd_table_count = ARRAY_SIZE(gen7_blt_cmd_table);
 		}
 
-		if (IS_HASWELL(engine->i915)) {
+		if (IS_GEN(engine->i915, 9)) {
+			engine->reg_tables = gen9_blt_reg_tables;
+			engine->reg_table_count =
+				ARRAY_SIZE(gen9_blt_reg_tables);
+		} else if (IS_HASWELL(engine->i915)) {
 			engine->reg_tables = hsw_blt_reg_tables;
 			engine->reg_table_count = ARRAY_SIZE(hsw_blt_reg_tables);
 		} else {
 			engine->reg_tables = ivb_blt_reg_tables;
 			engine->reg_table_count = ARRAY_SIZE(ivb_blt_reg_tables);
 		}
-
-		engine->get_cmd_length_mask = gen7_blt_get_cmd_length_mask;
 		break;
 	case VIDEO_ENHANCEMENT_CLASS:
-		cmd_tables = hsw_vebox_cmds;
-		cmd_table_count = ARRAY_SIZE(hsw_vebox_cmds);
+		cmd_tables = hsw_vebox_cmd_table;
+		cmd_table_count = ARRAY_SIZE(hsw_vebox_cmd_table);
 		/* VECS can use the same length_mask function as VCS */
 		engine->get_cmd_length_mask = gen7_bsd_get_cmd_length_mask;
 		break;
@@ -943,7 +1026,7 @@ void intel_engine_init_cmd_parser(struct intel_engine_cs *engine)
 		return;
 	}
 
-	engine->flags |= I915_ENGINE_NEEDS_CMD_PARSER;
+	engine->flags |= I915_ENGINE_USING_CMD_PARSER;
 }
 
 /**
@@ -955,7 +1038,7 @@ void intel_engine_init_cmd_parser(struct intel_engine_cs *engine)
  */
 void intel_engine_cleanup_cmd_parser(struct intel_engine_cs *engine)
 {
-	if (!intel_engine_needs_cmd_parser(engine))
+	if (!intel_engine_using_cmd_parser(engine))
 		return;
 
 	fini_hash_table(engine);
@@ -1029,22 +1112,16 @@ __find_reg(const struct drm_i915_reg_descriptor *table, int count, u32 addr)
 }
 
 static const struct drm_i915_reg_descriptor *
-find_reg(const struct intel_engine_cs *engine, bool is_master, u32 addr)
+find_reg(const struct intel_engine_cs *engine, u32 addr)
 {
 	const struct drm_i915_reg_table *table = engine->reg_tables;
+	const struct drm_i915_reg_descriptor *reg = NULL;
 	int count = engine->reg_table_count;
 
-	for (; count > 0; ++table, --count) {
-		if (!table->master || is_master) {
-			const struct drm_i915_reg_descriptor *reg;
+	for (; !reg && (count > 0); ++table, --count)
+		reg = __find_reg(table->regs, table->num_regs, addr);
 
-			reg = __find_reg(table->regs, table->num_regs, addr);
-			if (reg != NULL)
-				return reg;
-		}
-	}
-
-	return NULL;
+	return reg;
 }
 
 /* Returns a vmap'd pointer to dst_obj, which the caller must unmap */
@@ -1128,8 +1205,7 @@ static u32 *copy_batch(struct drm_i915_gem_object *dst_obj,
 
 static bool check_cmd(const struct intel_engine_cs *engine,
 		      const struct drm_i915_cmd_descriptor *desc,
-		      const u32 *cmd, u32 length,
-		      const bool is_master)
+		      const u32 *cmd, u32 length)
 {
 	if (desc->flags & CMD_DESC_SKIP)
 		return true;
@@ -1139,12 +1215,6 @@ static bool check_cmd(const struct intel_engine_cs *engine,
 		return false;
 	}
 
-	if ((desc->flags & CMD_DESC_MASTER) && !is_master) {
-		DRM_DEBUG_DRIVER("CMD: Rejected master-only command: 0x%08X\n",
-				 *cmd);
-		return false;
-	}
-
 	if (desc->flags & CMD_DESC_REGISTER) {
 		/*
 		 * Get the distance between individual register offset
@@ -1158,7 +1228,7 @@ static bool check_cmd(const struct intel_engine_cs *engine,
 		     offset += step) {
 			const u32 reg_addr = cmd[offset] & desc->reg.mask;
 			const struct drm_i915_reg_descriptor *reg =
-				find_reg(engine, is_master, reg_addr);
+				find_reg(engine, reg_addr);
 
 			if (!reg) {
 				DRM_DEBUG_DRIVER("CMD: Rejected register 0x%08X in command: 0x%08X (%s)\n",
@@ -1236,16 +1306,112 @@ static bool check_cmd(const struct intel_engine_cs *engine,
 	return true;
 }
 
+static int check_bbstart(const struct i915_gem_context *ctx,
+			 u32 *cmd, u32 offset, u32 length,
+			 u32 batch_len,
+			 u64 batch_start,
+			 u64 shadow_batch_start)
+{
+	u64 jump_offset, jump_target;
+	u32 target_cmd_offset, target_cmd_index;
+
+	/* For igt compatibility on older platforms */
+	if (CMDPARSER_USES_GGTT(ctx->i915)) {
+		DRM_DEBUG("CMD: Rejecting BB_START for ggtt based submission\n");
+		return -EACCES;
+	}
+
+	if (length != 3) {
+		DRM_DEBUG("CMD: Recursive BB_START with bad length(%u)\n",
+			  length);
+		return -EINVAL;
+	}
+
+	jump_target = *(u64*)(cmd+1);
+	jump_offset = jump_target - batch_start;
+
+	/*
+	 * Any underflow of jump_target is guaranteed to be outside the range
+	 * of a u32, so >= test catches both too large and too small
+	 */
+	if (jump_offset >= batch_len) {
+		DRM_DEBUG("CMD: BB_START to 0x%llx jumps out of BB\n",
+			  jump_target);
+		return -EINVAL;
+	}
+
+	/*
+	 * This cannot overflow a u32 because we already checked jump_offset
+	 * is within the BB, and the batch_len is a u32
+	 */
+	target_cmd_offset = lower_32_bits(jump_offset);
+	target_cmd_index = target_cmd_offset / sizeof(u32);
+
+	*(u64*)(cmd + 1) = shadow_batch_start + target_cmd_offset;
+
+	if (target_cmd_index == offset)
+		return 0;
+
+	if (ctx->jump_whitelist_cmds <= target_cmd_index) {
+		DRM_DEBUG("CMD: Rejecting BB_START - truncated whitelist array\n");
+		return -EINVAL;
+	} else if (!test_bit(target_cmd_index, ctx->jump_whitelist)) {
+		DRM_DEBUG("CMD: BB_START to 0x%llx not a previously executed cmd\n",
+			  jump_target);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static void init_whitelist(struct i915_gem_context *ctx, u32 batch_len)
+{
+	const u32 batch_cmds = DIV_ROUND_UP(batch_len, sizeof(u32));
+	const u32 exact_size = BITS_TO_LONGS(batch_cmds);
+	u32 next_size = BITS_TO_LONGS(roundup_pow_of_two(batch_cmds));
+	unsigned long *next_whitelist;
+
+	if (CMDPARSER_USES_GGTT(ctx->i915))
+		return;
+
+	if (batch_cmds <= ctx->jump_whitelist_cmds) {
+		bitmap_zero(ctx->jump_whitelist, batch_cmds);
+		return;
+	}
+
+again:
+	next_whitelist = kcalloc(next_size, sizeof(long), GFP_KERNEL);
+	if (next_whitelist) {
+		kfree(ctx->jump_whitelist);
+		ctx->jump_whitelist = next_whitelist;
+		ctx->jump_whitelist_cmds =
+			next_size * BITS_PER_BYTE * sizeof(long);
+		return;
+	}
+
+	if (next_size > exact_size) {
+		next_size = exact_size;
+		goto again;
+	}
+
+	DRM_DEBUG("CMD: Failed to extend whitelist. BB_START may be disallowed\n");
+	bitmap_zero(ctx->jump_whitelist, ctx->jump_whitelist_cmds);
+
+	return;
+}
+
 #define LENGTH_BIAS 2
 
 /**
  * i915_parse_cmds() - parse a submitted batch buffer for privilege violations
+ * @ctx: the context in which the batch is to execute
  * @engine: the engine on which the batch is to execute
  * @batch_obj: the batch buffer in question
- * @shadow_batch_obj: copy of the batch buffer in question
+ * @batch_start: Canonical base address of batch
  * @batch_start_offset: byte offset in the batch at which execution starts
  * @batch_len: length of the commands in batch_obj
- * @is_master: is the submitting process the drm master?
+ * @shadow_batch_obj: copy of the batch buffer in question
+ * @shadow_batch_start: Canonical base address of shadow_batch_obj
  *
  * Parses the specified batch buffer looking for privilege violations as
  * described in the overview.
@@ -1253,14 +1419,17 @@ static bool check_cmd(const struct intel_engine_cs *engine,
  * Return: non-zero if the parser finds violations or otherwise fails; -EACCES
  * if the batch appears legal but should use hardware parsing
  */
-int intel_engine_cmd_parser(struct intel_engine_cs *engine,
+
+int intel_engine_cmd_parser(struct i915_gem_context *ctx,
+			    struct intel_engine_cs *engine,
 			    struct drm_i915_gem_object *batch_obj,
-			    struct drm_i915_gem_object *shadow_batch_obj,
+			    u64 batch_start,
 			    u32 batch_start_offset,
 			    u32 batch_len,
-			    bool is_master)
+			    struct drm_i915_gem_object *shadow_batch_obj,
+			    u64 shadow_batch_start)
 {
-	u32 *cmd, *batch_end;
+	u32 *cmd, *batch_end, offset = 0;
 	struct drm_i915_cmd_descriptor default_desc = noop_desc;
 	const struct drm_i915_cmd_descriptor *desc = &default_desc;
 	bool needs_clflush_after = false;
@@ -1274,6 +1443,8 @@ int intel_engine_cmd_parser(struct intel_engine_cs *engine,
 		return PTR_ERR(cmd);
 	}
 
+	init_whitelist(ctx, batch_len);
+
 	/*
 	 * We use the batch length as size because the shadow object is as
 	 * large or larger and copy_batch() will write MI_NOPs to the extra
@@ -1283,31 +1454,15 @@ int intel_engine_cmd_parser(struct intel_engine_cs *engine,
 	do {
 		u32 length;
 
-		if (*cmd == MI_BATCH_BUFFER_END) {
-			if (needs_clflush_after) {
-				void *ptr = page_mask_bits(shadow_batch_obj->mm.mapping);
-				drm_clflush_virt_range(ptr,
-						       (void *)(cmd + 1) - ptr);
-			}
+		if (*cmd == MI_BATCH_BUFFER_END)
 			break;
-		}
 
 		desc = find_cmd(engine, *cmd, desc, &default_desc);
 		if (!desc) {
 			DRM_DEBUG_DRIVER("CMD: Unrecognized command: 0x%08X\n",
 					 *cmd);
 			ret = -EINVAL;
-			break;
-		}
-
-		/*
-		 * If the batch buffer contains a chained batch, return an
-		 * error that tells the caller to abort and dispatch the
-		 * workload as a non-secure batch.
-		 */
-		if (desc->cmd.value == MI_BATCH_BUFFER_START) {
-			ret = -EACCES;
-			break;
+			goto err;
 		}
 
 		if (desc->flags & CMD_DESC_FIXED)
@@ -1321,22 +1476,43 @@ int intel_engine_cmd_parser(struct intel_engine_cs *engine,
 					 length,
 					 batch_end - cmd);
 			ret = -EINVAL;
-			break;
+			goto err;
 		}
 
-		if (!check_cmd(engine, desc, cmd, length, is_master)) {
+		if (!check_cmd(engine, desc, cmd, length)) {
 			ret = -EACCES;
+			goto err;
+		}
+
+		if (desc->cmd.value == MI_BATCH_BUFFER_START) {
+			ret = check_bbstart(ctx, cmd, offset, length,
+					    batch_len, batch_start,
+					    shadow_batch_start);
+
+			if (ret)
+				goto err;
 			break;
 		}
 
+		if (ctx->jump_whitelist_cmds > offset)
+			set_bit(offset, ctx->jump_whitelist);
+
 		cmd += length;
+		offset += length;
 		if  (cmd >= batch_end) {
 			DRM_DEBUG_DRIVER("CMD: Got to the end of the buffer w/o a BBE cmd!\n");
 			ret = -EINVAL;
-			break;
+			goto err;
 		}
 	} while (1);
 
+	if (needs_clflush_after) {
+		void *ptr = page_mask_bits(shadow_batch_obj->mm.mapping);
+
+		drm_clflush_virt_range(ptr, (void *)(cmd + 1) - ptr);
+	}
+
+err:
 	i915_gem_object_unpin_map(shadow_batch_obj);
 	return ret;
 }
@@ -1357,7 +1533,7 @@ int i915_cmd_parser_get_version(struct drm_i915_private *dev_priv)
 
 	/* If the command parser is not enabled, report 0 - unsupported */
 	for_each_uabi_engine(engine, dev_priv) {
-		if (intel_engine_needs_cmd_parser(engine)) {
+		if (intel_engine_using_cmd_parser(engine)) {
 			active = true;
 			break;
 		}
@@ -1382,6 +1558,7 @@ int i915_cmd_parser_get_version(struct drm_i915_private *dev_priv)
 	 *    the parser enabled.
 	 * 9. Don't whitelist or handle oacontrol specially, as ownership
 	 *    for oacontrol state is moving to i915-perf.
+	 * 10. Support for Gen9 BCS Parsing
 	 */
-	return 9;
+	return 10;
 }
diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c
index b0f51591f2e4..8016484ebcd3 100644
--- a/drivers/gpu/drm/i915/i915_debugfs.c
+++ b/drivers/gpu/drm/i915/i915_debugfs.c
@@ -41,7 +41,10 @@
 
 #include "gem/i915_gem_context.h"
 #include "gt/intel_gt_pm.h"
+#include "gt/intel_gt_requests.h"
 #include "gt/intel_reset.h"
+#include "gt/intel_rc6.h"
+#include "gt/intel_rps.h"
 #include "gt/uc/intel_guc_submission.h"
 
 #include "i915_debugfs.h"
@@ -61,11 +64,18 @@ static int i915_capabilities(struct seq_file *m, void *data)
 	struct drm_i915_private *dev_priv = node_to_i915(m->private);
 	const struct intel_device_info *info = INTEL_INFO(dev_priv);
 	struct drm_printer p = drm_seq_file_printer(m);
+	const char *msg;
 
 	seq_printf(m, "gen: %d\n", INTEL_GEN(dev_priv));
 	seq_printf(m, "platform: %s\n", intel_platform_name(info->platform));
 	seq_printf(m, "pch: %d\n", INTEL_PCH_TYPE(dev_priv));
 
+	msg = "n/a";
+#ifdef CONFIG_INTEL_IOMMU
+	msg = enableddisabled(intel_iommu_gfx_mapped);
+#endif
+	seq_printf(m, "iommu: %s\n", msg);
+
 	intel_device_info_dump_flags(info, &p);
 	intel_device_info_dump_runtime(RUNTIME_INFO(dev_priv), &p);
 	intel_driver_caps_print(&dev_priv->caps, &p);
@@ -77,11 +87,6 @@ static int i915_capabilities(struct seq_file *m, void *data)
 	return 0;
 }
 
-static char get_pin_flag(struct drm_i915_gem_object *obj)
-{
-	return obj->pin_global ? 'p' : ' ';
-}
-
 static char get_tiling_flag(struct drm_i915_gem_object *obj)
 {
 	switch (i915_gem_object_get_tiling(obj)) {
@@ -140,9 +145,8 @@ describe_obj(struct seq_file *m, struct drm_i915_gem_object *obj)
 	struct i915_vma *vma;
 	int pin_count = 0;
 
-	seq_printf(m, "%pK: %c%c%c%c %8zdKiB %02x %02x %s%s%s",
+	seq_printf(m, "%pK: %c%c%c %8zdKiB %02x %02x %s%s%s",
 		   &obj->base,
-		   get_pin_flag(obj),
 		   get_tiling_flag(obj),
 		   get_global_flag(obj),
 		   get_pin_mapped_flag(obj),
@@ -221,8 +225,8 @@ describe_obj(struct seq_file *m, struct drm_i915_gem_object *obj)
 	seq_printf(m, " (pinned x %d)", pin_count);
 	if (obj->stolen)
 		seq_printf(m, " (stolen: %08llx)", obj->stolen->start);
-	if (obj->pin_global)
-		seq_printf(m, " (global)");
+	if (i915_gem_object_is_framebuffer(obj))
+		seq_printf(m, " (fb)");
 
 	engine = i915_gem_object_last_write_engine(obj);
 	if (engine)
@@ -243,6 +247,9 @@ static int per_file_stats(int id, void *ptr, void *data)
 	struct file_stats *stats = data;
 	struct i915_vma *vma;
 
+	if (!kref_get_unless_zero(&obj->base.refcount))
+		return 0;
+
 	stats->count++;
 	stats->total += obj->base.size;
 	if (!atomic_read(&obj->bind_count))
@@ -290,6 +297,7 @@ static int per_file_stats(int id, void *ptr, void *data)
 	}
 	spin_unlock(&obj->vma.lock);
 
+	i915_gem_object_put(obj);
 	return 0;
 }
 
@@ -309,34 +317,44 @@ static void print_context_stats(struct seq_file *m,
 				struct drm_i915_private *i915)
 {
 	struct file_stats kstats = {};
-	struct i915_gem_context *ctx;
+	struct i915_gem_context *ctx, *cn;
 
-	list_for_each_entry(ctx, &i915->contexts.list, link) {
+	spin_lock(&i915->gem.contexts.lock);
+	list_for_each_entry_safe(ctx, cn, &i915->gem.contexts.list, link) {
 		struct i915_gem_engines_iter it;
 		struct intel_context *ce;
 
+		if (!kref_get_unless_zero(&ctx->ref))
+			continue;
+
+		spin_unlock(&i915->gem.contexts.lock);
+
 		for_each_gem_engine(ce,
 				    i915_gem_context_lock_engines(ctx), it) {
 			intel_context_lock_pinned(ce);
 			if (intel_context_is_pinned(ce)) {
+				rcu_read_lock();
 				if (ce->state)
 					per_file_stats(0,
 						       ce->state->obj, &kstats);
 				per_file_stats(0, ce->ring->vma->obj, &kstats);
+				rcu_read_unlock();
 			}
 			intel_context_unlock_pinned(ce);
 		}
 		i915_gem_context_unlock_engines(ctx);
 
 		if (!IS_ERR_OR_NULL(ctx->file_priv)) {
-			struct file_stats stats = { .vm = ctx->vm, };
+			struct file_stats stats = {
+				.vm = rcu_access_pointer(ctx->vm),
+			};
 			struct drm_file *file = ctx->file_priv->file;
 			struct task_struct *task;
 			char name[80];
 
-			spin_lock(&file->table_lock);
+			rcu_read_lock();
 			idr_for_each(&file->object_idr, per_file_stats, &stats);
-			spin_unlock(&file->table_lock);
+			rcu_read_unlock();
 
 			rcu_read_lock();
 			task = pid_task(ctx->pid ?: file->pid, PIDTYPE_PID);
@@ -346,7 +364,12 @@ static void print_context_stats(struct seq_file *m,
 
 			print_file_stats(m, name, stats);
 		}
+
+		spin_lock(&i915->gem.contexts.lock);
+		list_safe_reset_next(ctx, cn, link);
+		i915_gem_context_put(ctx);
 	}
+	spin_unlock(&i915->gem.contexts.lock);
 
 	print_file_stats(m, "[k]contexts", kstats);
 }
@@ -354,7 +377,6 @@ static void print_context_stats(struct seq_file *m,
 static int i915_gem_object_info(struct seq_file *m, void *data)
 {
 	struct drm_i915_private *i915 = node_to_i915(m->private);
-	int ret;
 
 	seq_printf(m, "%u shrinkable [%u free] objects, %llu bytes\n",
 		   i915->mm.shrink_count,
@@ -363,12 +385,7 @@ static int i915_gem_object_info(struct seq_file *m, void *data)
 
 	seq_putc(m, '\n');
 
-	ret = mutex_lock_interruptible(&i915->drm.struct_mutex);
-	if (ret)
-		return ret;
-
 	print_context_stats(m, i915);
-	mutex_unlock(&i915->drm.struct_mutex);
 
 	return 0;
 }
@@ -376,7 +393,7 @@ static int i915_gem_object_info(struct seq_file *m, void *data)
 static void gen8_display_interrupt_info(struct seq_file *m)
 {
 	struct drm_i915_private *dev_priv = node_to_i915(m->private);
-	int pipe;
+	enum pipe pipe;
 
 	for_each_pipe(dev_priv, pipe) {
 		enum intel_display_power_domain power_domain;
@@ -527,6 +544,8 @@ static int i915_interrupt_info(struct seq_file *m, void *data)
 
 		gen8_display_interrupt_info(m);
 	} else if (IS_VALLEYVIEW(dev_priv)) {
+		intel_wakeref_t pref;
+
 		seq_printf(m, "Display IER:\t%08x\n",
 			   I915_READ(VLV_IER));
 		seq_printf(m, "Display IIR:\t%08x\n",
@@ -537,7 +556,6 @@ static int i915_interrupt_info(struct seq_file *m, void *data)
 			   I915_READ(VLV_IMR));
 		for_each_pipe(dev_priv, pipe) {
 			enum intel_display_power_domain power_domain;
-			intel_wakeref_t pref;
 
 			power_domain = POWER_DOMAIN_PIPE(pipe);
 			pref = intel_display_power_get_if_enabled(dev_priv,
@@ -571,12 +589,14 @@ static int i915_interrupt_info(struct seq_file *m, void *data)
 		seq_printf(m, "PM IMR:\t\t%08x\n",
 			   I915_READ(GEN6_PMIMR));
 
+		pref = intel_display_power_get(dev_priv, POWER_DOMAIN_INIT);
 		seq_printf(m, "Port hotplug:\t%08x\n",
 			   I915_READ(PORT_HOTPLUG_EN));
 		seq_printf(m, "DPFLIPSTAT:\t%08x\n",
 			   I915_READ(VLV_DPFLIPSTAT));
 		seq_printf(m, "DPINVGTT:\t%08x\n",
 			   I915_READ(DPINVGTT));
+		intel_display_power_put(dev_priv, POWER_DOMAIN_INIT, pref);
 
 	} else if (!HAS_PCH_SPLIT(dev_priv)) {
 		seq_printf(m, "Interrupt enable:    %08x\n",
@@ -772,7 +792,7 @@ static int i915_frequency_info(struct seq_file *m, void *unused)
 {
 	struct drm_i915_private *dev_priv = node_to_i915(m->private);
 	struct intel_uncore *uncore = &dev_priv->uncore;
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
+	struct intel_rps *rps = &dev_priv->gt.rps;
 	intel_wakeref_t wakeref;
 	int ret = 0;
 
@@ -808,23 +828,23 @@ static int i915_frequency_info(struct seq_file *m, void *unused)
 		seq_printf(m, "DDR freq: %d MHz\n", dev_priv->mem_freq);
 
 		seq_printf(m, "actual GPU freq: %d MHz\n",
-			   intel_gpu_freq(dev_priv, (freq_sts >> 8) & 0xff));
+			   intel_gpu_freq(rps, (freq_sts >> 8) & 0xff));
 
 		seq_printf(m, "current GPU freq: %d MHz\n",
-			   intel_gpu_freq(dev_priv, rps->cur_freq));
+			   intel_gpu_freq(rps, rps->cur_freq));
 
 		seq_printf(m, "max GPU freq: %d MHz\n",
-			   intel_gpu_freq(dev_priv, rps->max_freq));
+			   intel_gpu_freq(rps, rps->max_freq));
 
 		seq_printf(m, "min GPU freq: %d MHz\n",
-			   intel_gpu_freq(dev_priv, rps->min_freq));
+			   intel_gpu_freq(rps, rps->min_freq));
 
 		seq_printf(m, "idle GPU freq: %d MHz\n",
-			   intel_gpu_freq(dev_priv, rps->idle_freq));
+			   intel_gpu_freq(rps, rps->idle_freq));
 
 		seq_printf(m,
 			   "efficient (RPe) frequency: %d MHz\n",
-			   intel_gpu_freq(dev_priv, rps->efficient_freq));
+			   intel_gpu_freq(rps, rps->efficient_freq));
 	} else if (INTEL_GEN(dev_priv) >= 6) {
 		u32 rp_state_limits;
 		u32 gt_perf_status;
@@ -858,7 +878,7 @@ static int i915_frequency_info(struct seq_file *m, void *unused)
 			else
 				reqf >>= 25;
 		}
-		reqf = intel_gpu_freq(dev_priv, reqf);
+		reqf = intel_gpu_freq(rps, reqf);
 
 		rpmodectl = I915_READ(GEN6_RP_CONTROL);
 		rpinclimit = I915_READ(GEN6_RP_UP_THRESHOLD);
@@ -871,8 +891,7 @@ static int i915_frequency_info(struct seq_file *m, void *unused)
 		rpdownei = I915_READ(GEN6_RP_CUR_DOWN_EI) & GEN6_CURIAVG_MASK;
 		rpcurdown = I915_READ(GEN6_RP_CUR_DOWN) & GEN6_CURBSYTAVG_MASK;
 		rpprevdown = I915_READ(GEN6_RP_PREV_DOWN) & GEN6_CURBSYTAVG_MASK;
-		cagf = intel_gpu_freq(dev_priv,
-				      intel_get_cagf(dev_priv, rpstat));
+		cagf = intel_gpu_freq(rps, intel_get_cagf(rps, rpstat));
 
 		intel_uncore_forcewake_put(&dev_priv->uncore, FORCEWAKE_ALL);
 
@@ -949,37 +968,37 @@ static int i915_frequency_info(struct seq_file *m, void *unused)
 		max_freq *= (IS_GEN9_BC(dev_priv) ||
 			     INTEL_GEN(dev_priv) >= 10 ? GEN9_FREQ_SCALER : 1);
 		seq_printf(m, "Lowest (RPN) frequency: %dMHz\n",
-			   intel_gpu_freq(dev_priv, max_freq));
+			   intel_gpu_freq(rps, max_freq));
 
 		max_freq = (rp_state_cap & 0xff00) >> 8;
 		max_freq *= (IS_GEN9_BC(dev_priv) ||
 			     INTEL_GEN(dev_priv) >= 10 ? GEN9_FREQ_SCALER : 1);
 		seq_printf(m, "Nominal (RP1) frequency: %dMHz\n",
-			   intel_gpu_freq(dev_priv, max_freq));
+			   intel_gpu_freq(rps, max_freq));
 
 		max_freq = (IS_GEN9_LP(dev_priv) ? rp_state_cap >> 16 :
 			    rp_state_cap >> 0) & 0xff;
 		max_freq *= (IS_GEN9_BC(dev_priv) ||
 			     INTEL_GEN(dev_priv) >= 10 ? GEN9_FREQ_SCALER : 1);
 		seq_printf(m, "Max non-overclocked (RP0) frequency: %dMHz\n",
-			   intel_gpu_freq(dev_priv, max_freq));
+			   intel_gpu_freq(rps, max_freq));
 		seq_printf(m, "Max overclocked frequency: %dMHz\n",
-			   intel_gpu_freq(dev_priv, rps->max_freq));
+			   intel_gpu_freq(rps, rps->max_freq));
 
 		seq_printf(m, "Current freq: %d MHz\n",
-			   intel_gpu_freq(dev_priv, rps->cur_freq));
+			   intel_gpu_freq(rps, rps->cur_freq));
 		seq_printf(m, "Actual freq: %d MHz\n", cagf);
 		seq_printf(m, "Idle freq: %d MHz\n",
-			   intel_gpu_freq(dev_priv, rps->idle_freq));
+			   intel_gpu_freq(rps, rps->idle_freq));
 		seq_printf(m, "Min freq: %d MHz\n",
-			   intel_gpu_freq(dev_priv, rps->min_freq));
+			   intel_gpu_freq(rps, rps->min_freq));
 		seq_printf(m, "Boost freq: %d MHz\n",
-			   intel_gpu_freq(dev_priv, rps->boost_freq));
+			   intel_gpu_freq(rps, rps->boost_freq));
 		seq_printf(m, "Max freq: %d MHz\n",
-			   intel_gpu_freq(dev_priv, rps->max_freq));
+			   intel_gpu_freq(rps, rps->max_freq));
 		seq_printf(m,
 			   "efficient (RPe) frequency: %d MHz\n",
-			   intel_gpu_freq(dev_priv, rps->efficient_freq));
+			   intel_gpu_freq(rps, rps->efficient_freq));
 	} else {
 		seq_puts(m, "no P-state info available\n");
 	}
@@ -992,91 +1011,6 @@ static int i915_frequency_info(struct seq_file *m, void *unused)
 	return ret;
 }
 
-static void i915_instdone_info(struct drm_i915_private *dev_priv,
-			       struct seq_file *m,
-			       struct intel_instdone *instdone)
-{
-	int slice;
-	int subslice;
-
-	seq_printf(m, "\t\tINSTDONE: 0x%08x\n",
-		   instdone->instdone);
-
-	if (INTEL_GEN(dev_priv) <= 3)
-		return;
-
-	seq_printf(m, "\t\tSC_INSTDONE: 0x%08x\n",
-		   instdone->slice_common);
-
-	if (INTEL_GEN(dev_priv) <= 6)
-		return;
-
-	for_each_instdone_slice_subslice(dev_priv, slice, subslice)
-		seq_printf(m, "\t\tSAMPLER_INSTDONE[%d][%d]: 0x%08x\n",
-			   slice, subslice, instdone->sampler[slice][subslice]);
-
-	for_each_instdone_slice_subslice(dev_priv, slice, subslice)
-		seq_printf(m, "\t\tROW_INSTDONE[%d][%d]: 0x%08x\n",
-			   slice, subslice, instdone->row[slice][subslice]);
-}
-
-static int i915_hangcheck_info(struct seq_file *m, void *unused)
-{
-	struct drm_i915_private *i915 = node_to_i915(m->private);
-	struct intel_gt *gt = &i915->gt;
-	struct intel_engine_cs *engine;
-	intel_wakeref_t wakeref;
-	enum intel_engine_id id;
-
-	seq_printf(m, "Reset flags: %lx\n", gt->reset.flags);
-	if (test_bit(I915_WEDGED, &gt->reset.flags))
-		seq_puts(m, "\tWedged\n");
-	if (test_bit(I915_RESET_BACKOFF, &gt->reset.flags))
-		seq_puts(m, "\tDevice (global) reset in progress\n");
-
-	if (!i915_modparams.enable_hangcheck) {
-		seq_puts(m, "Hangcheck disabled\n");
-		return 0;
-	}
-
-	if (timer_pending(&gt->hangcheck.work.timer))
-		seq_printf(m, "Hangcheck active, timer fires in %dms\n",
-			   jiffies_to_msecs(gt->hangcheck.work.timer.expires -
-					    jiffies));
-	else if (delayed_work_pending(&gt->hangcheck.work))
-		seq_puts(m, "Hangcheck active, work pending\n");
-	else
-		seq_puts(m, "Hangcheck inactive\n");
-
-	seq_printf(m, "GT active? %s\n", yesno(gt->awake));
-
-	with_intel_runtime_pm(&i915->runtime_pm, wakeref) {
-		for_each_engine(engine, i915, id) {
-			struct intel_instdone instdone;
-
-			seq_printf(m, "%s: %d ms ago\n",
-				   engine->name,
-				   jiffies_to_msecs(jiffies -
-						    engine->hangcheck.action_timestamp));
-
-			seq_printf(m, "\tACTHD = 0x%08llx [current 0x%08llx]\n",
-				   (long long)engine->hangcheck.acthd,
-				   intel_engine_get_active_head(engine));
-
-			intel_engine_get_instdone(engine, &instdone);
-
-			seq_puts(m, "\tinstdone read =\n");
-			i915_instdone_info(i915, m, &instdone);
-
-			seq_puts(m, "\tinstdone accu =\n");
-			i915_instdone_info(i915, m,
-					   &engine->hangcheck.instdone);
-		}
-	}
-
-	return 0;
-}
-
 static int ironlake_drpc_info(struct seq_file *m)
 {
 	struct drm_i915_private *i915 = node_to_i915(m->private);
@@ -1157,11 +1091,13 @@ static void print_rc6_res(struct seq_file *m,
 			  const char *title,
 			  const i915_reg_t reg)
 {
-	struct drm_i915_private *dev_priv = node_to_i915(m->private);
+	struct drm_i915_private *i915 = node_to_i915(m->private);
+	intel_wakeref_t wakeref;
 
-	seq_printf(m, "%s %u (%llu us)\n",
-		   title, I915_READ(reg),
-		   intel_rc6_residency_us(dev_priv, reg));
+	with_intel_runtime_pm(&i915->runtime_pm, wakeref)
+		seq_printf(m, "%s %u (%llu us)\n", title,
+			   intel_uncore_read(&i915->uncore, reg),
+			   intel_rc6_residency_us(&i915->gt.rc6, reg));
 }
 
 static int vlv_drpc_info(struct seq_file *m)
@@ -1439,7 +1375,7 @@ static int i915_sr_status(struct seq_file *m, void *unused)
 static int i915_ring_freq_table(struct seq_file *m, void *unused)
 {
 	struct drm_i915_private *dev_priv = node_to_i915(m->private);
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
+	struct intel_rps *rps = &dev_priv->gt.rps;
 	unsigned int max_gpu_freq, min_gpu_freq;
 	intel_wakeref_t wakeref;
 	int gpu_freq, ia_freq;
@@ -1464,10 +1400,11 @@ static int i915_ring_freq_table(struct seq_file *m, void *unused)
 				       GEN6_PCODE_READ_MIN_FREQ_TABLE,
 				       &ia_freq, NULL);
 		seq_printf(m, "%d\t\t%d\t\t\t\t%d\n",
-			   intel_gpu_freq(dev_priv, (gpu_freq *
-						     (IS_GEN9_BC(dev_priv) ||
-						      INTEL_GEN(dev_priv) >= 10 ?
-						      GEN9_FREQ_SCALER : 1))),
+			   intel_gpu_freq(rps,
+					  (gpu_freq *
+					   (IS_GEN9_BC(dev_priv) ||
+					    INTEL_GEN(dev_priv) >= 10 ?
+					    GEN9_FREQ_SCALER : 1))),
 			   ((ia_freq >> 0) & 0xff) * 100,
 			   ((ia_freq >> 8) & 0xff) * 100);
 	}
@@ -1478,21 +1415,11 @@ static int i915_ring_freq_table(struct seq_file *m, void *unused)
 
 static int i915_opregion(struct seq_file *m, void *unused)
 {
-	struct drm_i915_private *dev_priv = node_to_i915(m->private);
-	struct drm_device *dev = &dev_priv->drm;
-	struct intel_opregion *opregion = &dev_priv->opregion;
-	int ret;
-
-	ret = mutex_lock_interruptible(&dev->struct_mutex);
-	if (ret)
-		goto out;
+	struct intel_opregion *opregion = &node_to_i915(m->private)->opregion;
 
 	if (opregion->header)
 		seq_write(m, opregion->header, OPREGION_SIZE);
 
-	mutex_unlock(&dev->struct_mutex);
-
-out:
 	return 0;
 }
 
@@ -1512,11 +1439,6 @@ static int i915_gem_framebuffer_info(struct seq_file *m, void *data)
 	struct drm_device *dev = &dev_priv->drm;
 	struct intel_framebuffer *fbdev_fb = NULL;
 	struct drm_framebuffer *drm_fb;
-	int ret;
-
-	ret = mutex_lock_interruptible(&dev->struct_mutex);
-	if (ret)
-		return ret;
 
 #ifdef CONFIG_DRM_FBDEV_EMULATION
 	if (dev_priv->fbdev && dev_priv->fbdev->helper.fb) {
@@ -1551,7 +1473,6 @@ static int i915_gem_framebuffer_info(struct seq_file *m, void *data)
 		seq_putc(m, '\n');
 	}
 	mutex_unlock(&dev->mode_config.fb_lock);
-	mutex_unlock(&dev->struct_mutex);
 
 	return 0;
 }
@@ -1564,23 +1485,20 @@ static void describe_ctx_ring(struct seq_file *m, struct intel_ring *ring)
 
 static int i915_context_status(struct seq_file *m, void *unused)
 {
-	struct drm_i915_private *dev_priv = node_to_i915(m->private);
-	struct drm_device *dev = &dev_priv->drm;
-	struct i915_gem_context *ctx;
-	int ret;
-
-	ret = mutex_lock_interruptible(&dev->struct_mutex);
-	if (ret)
-		return ret;
+	struct drm_i915_private *i915 = node_to_i915(m->private);
+	struct i915_gem_context *ctx, *cn;
 
-	list_for_each_entry(ctx, &dev_priv->contexts.list, link) {
+	spin_lock(&i915->gem.contexts.lock);
+	list_for_each_entry_safe(ctx, cn, &i915->gem.contexts.list, link) {
 		struct i915_gem_engines_iter it;
 		struct intel_context *ce;
 
+		if (!kref_get_unless_zero(&ctx->ref))
+			continue;
+
+		spin_unlock(&i915->gem.contexts.lock);
+
 		seq_puts(m, "HW context ");
-		if (!list_empty(&ctx->hw_id_link))
-			seq_printf(m, "%x [pin %u]", ctx->hw_id,
-				   atomic_read(&ctx->hw_id_pin_count));
 		if (ctx->pid) {
 			struct task_struct *task;
 
@@ -1614,9 +1532,12 @@ static int i915_context_status(struct seq_file *m, void *unused)
 		i915_gem_context_unlock_engines(ctx);
 
 		seq_putc(m, '\n');
-	}
 
-	mutex_unlock(&dev->struct_mutex);
+		spin_lock(&i915->gem.contexts.lock);
+		list_safe_reset_next(ctx, cn, link);
+		i915_gem_context_put(ctx);
+	}
+	spin_unlock(&i915->gem.contexts.lock);
 
 	return 0;
 }
@@ -1654,9 +1575,9 @@ static int i915_swizzle_info(struct seq_file *m, void *data)
 	wakeref = intel_runtime_pm_get(&dev_priv->runtime_pm);
 
 	seq_printf(m, "bit6 swizzle for X-tiling = %s\n",
-		   swizzle_string(dev_priv->mm.bit_6_swizzle_x));
+		   swizzle_string(dev_priv->ggtt.bit_6_swizzle_x));
 	seq_printf(m, "bit6 swizzle for Y-tiling = %s\n",
-		   swizzle_string(dev_priv->mm.bit_6_swizzle_y));
+		   swizzle_string(dev_priv->ggtt.bit_6_swizzle_y));
 
 	if (IS_GEN_RANGE(dev_priv, 3, 4)) {
 		seq_printf(m, "DDC = 0x%08x\n",
@@ -1711,7 +1632,7 @@ static const char *rps_power_to_str(unsigned int power)
 static int i915_rps_boost_info(struct seq_file *m, void *data)
 {
 	struct drm_i915_private *dev_priv = node_to_i915(m->private);
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
+	struct intel_rps *rps = &dev_priv->gt.rps;
 	u32 act_freq = rps->cur_freq;
 	intel_wakeref_t wakeref;
 
@@ -1723,7 +1644,7 @@ static int i915_rps_boost_info(struct seq_file *m, void *data)
 			vlv_punit_put(dev_priv);
 			act_freq = (act_freq >> 8) & 0xff;
 		} else {
-			act_freq = intel_get_cagf(dev_priv,
+			act_freq = intel_get_cagf(rps,
 						  I915_READ(GEN6_RPSTAT1));
 		}
 	}
@@ -1734,17 +1655,17 @@ static int i915_rps_boost_info(struct seq_file *m, void *data)
 		   atomic_read(&rps->num_waiters));
 	seq_printf(m, "Interactive? %d\n", READ_ONCE(rps->power.interactive));
 	seq_printf(m, "Frequency requested %d, actual %d\n",
-		   intel_gpu_freq(dev_priv, rps->cur_freq),
-		   intel_gpu_freq(dev_priv, act_freq));
+		   intel_gpu_freq(rps, rps->cur_freq),
+		   intel_gpu_freq(rps, act_freq));
 	seq_printf(m, "  min hard:%d, soft:%d; max soft:%d, hard:%d\n",
-		   intel_gpu_freq(dev_priv, rps->min_freq),
-		   intel_gpu_freq(dev_priv, rps->min_freq_softlimit),
-		   intel_gpu_freq(dev_priv, rps->max_freq_softlimit),
-		   intel_gpu_freq(dev_priv, rps->max_freq));
+		   intel_gpu_freq(rps, rps->min_freq),
+		   intel_gpu_freq(rps, rps->min_freq_softlimit),
+		   intel_gpu_freq(rps, rps->max_freq_softlimit),
+		   intel_gpu_freq(rps, rps->max_freq));
 	seq_printf(m, "  idle:%d, efficient:%d, boost:%d\n",
-		   intel_gpu_freq(dev_priv, rps->idle_freq),
-		   intel_gpu_freq(dev_priv, rps->efficient_freq),
-		   intel_gpu_freq(dev_priv, rps->boost_freq));
+		   intel_gpu_freq(rps, rps->idle_freq),
+		   intel_gpu_freq(rps, rps->efficient_freq),
+		   intel_gpu_freq(rps, rps->boost_freq));
 
 	seq_printf(m, "Wait boosts: %d\n", atomic_read(&rps->boosts));
 
@@ -1860,8 +1781,8 @@ static void i915_guc_log_info(struct seq_file *m,
 	struct intel_guc_log *log = &dev_priv->gt.uc.guc.log;
 	enum guc_log_buffer_type type;
 
-	if (!intel_guc_log_relay_enabled(log)) {
-		seq_puts(m, "GuC log relay disabled\n");
+	if (!intel_guc_log_relay_created(log)) {
+		seq_puts(m, "GuC log relay not created\n");
 		return;
 	}
 
@@ -2048,9 +1969,23 @@ i915_guc_log_relay_write(struct file *filp,
 			 loff_t *ppos)
 {
 	struct intel_guc_log *log = filp->private_data;
+	int val;
+	int ret;
 
-	intel_guc_log_relay_flush(log);
-	return cnt;
+	ret = kstrtoint_from_user(ubuf, cnt, 0, &val);
+	if (ret < 0)
+		return ret;
+
+	/*
+	 * Enable and start the guc log relay on value of 1.
+	 * Flush log relay for any other value.
+	 */
+	if (val == 1)
+		ret = intel_guc_log_relay_start(log);
+	else
+		intel_guc_log_relay_flush(log);
+
+	return ret ?: cnt;
 }
 
 static int i915_guc_log_relay_release(struct inode *inode, struct file *file)
@@ -2133,7 +2068,7 @@ psr_source_status(struct drm_i915_private *dev_priv, struct seq_file *m)
 			"BUF_ON",
 			"TG_ON"
 		};
-		val = I915_READ(EDP_PSR2_STATUS);
+		val = I915_READ(EDP_PSR2_STATUS(dev_priv->psr.transcoder));
 		status_val = (val & EDP_PSR2_STATUS_STATE_MASK) >>
 			      EDP_PSR2_STATUS_STATE_SHIFT;
 		if (status_val < ARRAY_SIZE(live_status))
@@ -2149,7 +2084,7 @@ psr_source_status(struct drm_i915_private *dev_priv, struct seq_file *m)
 			"SRDOFFACK",
 			"SRDENT_ON",
 		};
-		val = I915_READ(EDP_PSR_STATUS);
+		val = I915_READ(EDP_PSR_STATUS(dev_priv->psr.transcoder));
 		status_val = (val & EDP_PSR_STATUS_STATE_MASK) >>
 			      EDP_PSR_STATUS_STATE_SHIFT;
 		if (status_val < ARRAY_SIZE(live_status))
@@ -2188,14 +2123,18 @@ static int i915_edp_psr_status(struct seq_file *m, void *data)
 		status = "disabled";
 	seq_printf(m, "PSR mode: %s\n", status);
 
-	if (!psr->enabled)
+	if (!psr->enabled) {
+		seq_printf(m, "PSR sink not reliable: %s\n",
+			   yesno(psr->sink_not_reliable));
+
 		goto unlock;
+	}
 
 	if (psr->psr2_enabled) {
-		val = I915_READ(EDP_PSR2_CTL);
+		val = I915_READ(EDP_PSR2_CTL(dev_priv->psr.transcoder));
 		enabled = val & EDP_PSR2_ENABLE;
 	} else {
-		val = I915_READ(EDP_PSR_CTL);
+		val = I915_READ(EDP_PSR_CTL(dev_priv->psr.transcoder));
 		enabled = val & EDP_PSR_ENABLE;
 	}
 	seq_printf(m, "Source PSR ctl: %s [0x%08x]\n",
@@ -2208,7 +2147,8 @@ static int i915_edp_psr_status(struct seq_file *m, void *data)
 	 * SKL+ Perf counter is reset to 0 everytime DC state is entered
 	 */
 	if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv)) {
-		val = I915_READ(EDP_PSR_PERF_CNT) & EDP_PSR_PERF_CNT_MASK;
+		val = I915_READ(EDP_PSR_PERF_CNT(dev_priv->psr.transcoder));
+		val &= EDP_PSR_PERF_CNT_MASK;
 		seq_printf(m, "Performance counter: %u\n", val);
 	}
 
@@ -2226,8 +2166,11 @@ static int i915_edp_psr_status(struct seq_file *m, void *data)
 		 * Reading all 3 registers before hand to minimize crossing a
 		 * frame boundary between register reads
 		 */
-		for (frame = 0; frame < PSR2_SU_STATUS_FRAMES; frame += 3)
-			su_frames_val[frame / 3] = I915_READ(PSR2_SU_STATUS(frame));
+		for (frame = 0; frame < PSR2_SU_STATUS_FRAMES; frame += 3) {
+			val = I915_READ(PSR2_SU_STATUS(dev_priv->psr.transcoder,
+						       frame));
+			su_frames_val[frame / 3] = val;
+		}
 
 		seq_puts(m, "Frame:\tPSR2 SU blocks:\n");
 
@@ -2360,8 +2303,7 @@ static int i915_power_domain_info(struct seq_file *m, void *unused)
 
 		for_each_power_domain(power_domain, power_well->desc->domains)
 			seq_printf(m, "  %-23s %d\n",
-				 intel_display_power_domain_str(dev_priv,
-								power_domain),
+				 intel_display_power_domain_str(power_domain),
 				 power_domains->domain_use_count[power_domain]);
 	}
 
@@ -2396,6 +2338,13 @@ static int i915_dmc_info(struct seq_file *m, void *unused)
 	if (INTEL_GEN(dev_priv) >= 12) {
 		dc5_reg = TGL_DMC_DEBUG_DC5_COUNT;
 		dc6_reg = TGL_DMC_DEBUG_DC6_COUNT;
+		/*
+		 * NOTE: DMC_DEBUG3 is a general purpose reg.
+		 * According to B.Specs:49196 DMC f/w reuses DC5/6 counter
+		 * reg for DC3CO debugging and validation,
+		 * but TGL DMC f/w is using DMC_DEBUG3 reg for DC3CO counter.
+		 */
+		seq_printf(m, "DC3CO count: %d\n", I915_READ(DMC_DEBUG3));
 	} else {
 		dc5_reg = IS_BROXTON(dev_priv) ? BXT_CSR_DC3_DC5_COUNT :
 						 SKL_CSR_DC3_DC5_COUNT;
@@ -3110,8 +3059,9 @@ static int i915_dp_mst_info(struct seq_file *m, void *unused)
 		if (!intel_dig_port->dp.can_mst)
 			continue;
 
-		seq_printf(m, "MST Source Port %c\n",
-			   port_name(intel_dig_port->base.port));
+		seq_printf(m, "MST Source Port [ENCODER:%d:%s]\n",
+			   intel_dig_port->base.base.base.id,
+			   intel_dig_port->base.base.name);
 		drm_dp_mst_dump_topology(m, &intel_dig_port->dp.mst_mgr);
 	}
 	drm_connector_list_iter_end(&conn_iter);
@@ -3573,6 +3523,37 @@ DEFINE_SIMPLE_ATTRIBUTE(i915_wedged_fops,
 			i915_wedged_get, i915_wedged_set,
 			"%llu\n");
 
+static int
+i915_perf_noa_delay_set(void *data, u64 val)
+{
+	struct drm_i915_private *i915 = data;
+	const u32 clk = RUNTIME_INFO(i915)->cs_timestamp_frequency_khz;
+
+	/*
+	 * This would lead to infinite waits as we're doing timestamp
+	 * difference on the CS with only 32bits.
+	 */
+	if (val > mul_u32_u32(U32_MAX, clk))
+		return -EINVAL;
+
+	atomic64_set(&i915->perf.noa_programming_delay, val);
+	return 0;
+}
+
+static int
+i915_perf_noa_delay_get(void *data, u64 *val)
+{
+	struct drm_i915_private *i915 = data;
+
+	*val = atomic64_read(&i915->perf.noa_programming_delay);
+	return 0;
+}
+
+DEFINE_SIMPLE_ATTRIBUTE(i915_perf_noa_delay_fops,
+			i915_perf_noa_delay_get,
+			i915_perf_noa_delay_set,
+			"%llu\n");
+
 #define DROP_UNBOUND	BIT(0)
 #define DROP_BOUND	BIT(1)
 #define DROP_RETIRE	BIT(2)
@@ -3582,6 +3563,7 @@ DEFINE_SIMPLE_ATTRIBUTE(i915_wedged_fops,
 #define DROP_IDLE	BIT(6)
 #define DROP_RESET_ACTIVE	BIT(7)
 #define DROP_RESET_SEQNO	BIT(8)
+#define DROP_RCU	BIT(9)
 #define DROP_ALL (DROP_UNBOUND	| \
 		  DROP_BOUND	| \
 		  DROP_RETIRE	| \
@@ -3590,7 +3572,8 @@ DEFINE_SIMPLE_ATTRIBUTE(i915_wedged_fops,
 		  DROP_SHRINK_ALL |\
 		  DROP_IDLE	| \
 		  DROP_RESET_ACTIVE | \
-		  DROP_RESET_SEQNO)
+		  DROP_RESET_SEQNO | \
+		  DROP_RCU)
 static int
 i915_drop_caches_get(void *data, u64 *val)
 {
@@ -3598,58 +3581,48 @@ i915_drop_caches_get(void *data, u64 *val)
 
 	return 0;
 }
-
 static int
-i915_drop_caches_set(void *data, u64 val)
+gt_drop_caches(struct intel_gt *gt, u64 val)
 {
-	struct drm_i915_private *i915 = data;
-
-	DRM_DEBUG("Dropping caches: 0x%08llx [0x%08llx]\n",
-		  val, val & DROP_ALL);
+	int ret;
 
 	if (val & DROP_RESET_ACTIVE &&
-	    wait_for(intel_engines_are_idle(&i915->gt),
-		     I915_IDLE_ENGINES_TIMEOUT))
-		intel_gt_set_wedged(&i915->gt);
+	    wait_for(intel_engines_are_idle(gt), I915_IDLE_ENGINES_TIMEOUT))
+		intel_gt_set_wedged(gt);
 
-	/* No need to check and wait for gpu resets, only libdrm auto-restarts
-	 * on ioctls on -EAGAIN. */
-	if (val & (DROP_ACTIVE | DROP_IDLE | DROP_RETIRE | DROP_RESET_SEQNO)) {
-		int ret;
+	if (val & DROP_RETIRE)
+		intel_gt_retire_requests(gt);
 
-		ret = mutex_lock_interruptible(&i915->drm.struct_mutex);
+	if (val & (DROP_IDLE | DROP_ACTIVE)) {
+		ret = intel_gt_wait_for_idle(gt, MAX_SCHEDULE_TIMEOUT);
 		if (ret)
 			return ret;
+	}
 
-		/*
-		 * To finish the flush of the idle_worker, we must complete
-		 * the switch-to-kernel-context, which requires a double
-		 * pass through wait_for_idle: first queues the switch,
-		 * second waits for the switch.
-		 */
-		if (ret == 0 && val & (DROP_IDLE | DROP_ACTIVE))
-			ret = i915_gem_wait_for_idle(i915,
-						     I915_WAIT_INTERRUPTIBLE |
-						     I915_WAIT_LOCKED,
-						     MAX_SCHEDULE_TIMEOUT);
+	if (val & DROP_IDLE) {
+		ret = intel_gt_pm_wait_for_idle(gt);
+		if (ret)
+			return ret;
+	}
 
-		if (ret == 0 && val & DROP_IDLE)
-			ret = i915_gem_wait_for_idle(i915,
-						     I915_WAIT_INTERRUPTIBLE |
-						     I915_WAIT_LOCKED,
-						     MAX_SCHEDULE_TIMEOUT);
+	if (val & DROP_RESET_ACTIVE && intel_gt_terminally_wedged(gt))
+		intel_gt_handle_error(gt, ALL_ENGINES, 0, NULL);
 
-		if (val & DROP_RETIRE)
-			i915_retire_requests(i915);
+	return 0;
+}
 
-		mutex_unlock(&i915->drm.struct_mutex);
+static int
+i915_drop_caches_set(void *data, u64 val)
+{
+	struct drm_i915_private *i915 = data;
+	int ret;
 
-		if (ret == 0 && val & DROP_IDLE)
-			ret = intel_gt_pm_wait_for_idle(&i915->gt);
-	}
+	DRM_DEBUG("Dropping caches: 0x%08llx [0x%08llx]\n",
+		  val, val & DROP_ALL);
 
-	if (val & DROP_RESET_ACTIVE && intel_gt_terminally_wedged(&i915->gt))
-		intel_gt_handle_error(&i915->gt, ALL_ENGINES, 0, NULL);
+	ret = gt_drop_caches(&i915->gt, val);
+	if (ret)
+		return ret;
 
 	fs_reclaim_acquire(GFP_KERNEL);
 	if (val & DROP_BOUND)
@@ -3662,10 +3635,8 @@ i915_drop_caches_set(void *data, u64 val)
 		i915_gem_shrink_all(i915);
 	fs_reclaim_release(GFP_KERNEL);
 
-	if (val & DROP_IDLE) {
-		flush_delayed_work(&i915->gem.retire_work);
-		flush_work(&i915->gem.idle_work);
-	}
+	if (val & DROP_RCU)
+		rcu_barrier();
 
 	if (val & DROP_FREED)
 		i915_gem_drain_freed_objects(i915);
@@ -3721,6 +3692,15 @@ i915_cache_sharing_set(void *data, u64 val)
 	return 0;
 }
 
+static void
+intel_sseu_copy_subslices(const struct sseu_dev_info *sseu, int slice,
+			  u8 *to_mask)
+{
+	int offset = slice * sseu->ss_stride;
+
+	memcpy(&to_mask[offset], &sseu->subslice_mask[offset], sseu->ss_stride);
+}
+
 DEFINE_SIMPLE_ATTRIBUTE(i915_cache_sharing_fops,
 			i915_cache_sharing_get, i915_cache_sharing_set,
 			"%llu\n");
@@ -3794,12 +3774,13 @@ static void gen10_sseu_device_status(struct drm_i915_private *dev_priv,
 			continue;
 
 		sseu->slice_mask |= BIT(s);
-		sseu->subslice_mask[s] = info->sseu.subslice_mask[s];
+		intel_sseu_copy_subslices(&info->sseu, s, sseu->subslice_mask);
 
 		for (ss = 0; ss < info->sseu.max_subslices; ss++) {
 			unsigned int eu_cnt;
 
-			if (!(s_reg[s] & (GEN9_PGCTL_SS_ACK(ss))))
+			if (info->sseu.has_subslice_pg &&
+			    !(s_reg[s] & (GEN9_PGCTL_SS_ACK(ss))))
 				/* skip disabled subslice */
 				continue;
 
@@ -3845,18 +3826,21 @@ static void gen9_sseu_device_status(struct drm_i915_private *dev_priv,
 		sseu->slice_mask |= BIT(s);
 
 		if (IS_GEN9_BC(dev_priv))
-			sseu->subslice_mask[s] =
-				RUNTIME_INFO(dev_priv)->sseu.subslice_mask[s];
+			intel_sseu_copy_subslices(&info->sseu, s,
+						  sseu->subslice_mask);
 
 		for (ss = 0; ss < info->sseu.max_subslices; ss++) {
 			unsigned int eu_cnt;
+			u8 ss_idx = s * info->sseu.ss_stride +
+				    ss / BITS_PER_BYTE;
 
 			if (IS_GEN9_LP(dev_priv)) {
 				if (!(s_reg[s] & (GEN9_PGCTL_SS_ACK(ss))))
 					/* skip disabled subslice */
 					continue;
 
-				sseu->subslice_mask[s] |= BIT(ss);
+				sseu->subslice_mask[ss_idx] |=
+					BIT(ss % BITS_PER_BYTE);
 			}
 
 			eu_cnt = 2 * hweight32(eu_reg[2*s + ss/2] &
@@ -3873,25 +3857,23 @@ static void gen9_sseu_device_status(struct drm_i915_private *dev_priv,
 static void broadwell_sseu_device_status(struct drm_i915_private *dev_priv,
 					 struct sseu_dev_info *sseu)
 {
+	const struct intel_runtime_info *info = RUNTIME_INFO(dev_priv);
 	u32 slice_info = I915_READ(GEN8_GT_SLICE_INFO);
 	int s;
 
 	sseu->slice_mask = slice_info & GEN8_LSLICESTAT_MASK;
 
 	if (sseu->slice_mask) {
-		sseu->eu_per_subslice =
-			RUNTIME_INFO(dev_priv)->sseu.eu_per_subslice;
-		for (s = 0; s < fls(sseu->slice_mask); s++) {
-			sseu->subslice_mask[s] =
-				RUNTIME_INFO(dev_priv)->sseu.subslice_mask[s];
-		}
+		sseu->eu_per_subslice = info->sseu.eu_per_subslice;
+		for (s = 0; s < fls(sseu->slice_mask); s++)
+			intel_sseu_copy_subslices(&info->sseu, s,
+						  sseu->subslice_mask);
 		sseu->eu_total = sseu->eu_per_subslice *
 				 intel_sseu_subslice_total(sseu);
 
 		/* subtract fused off EU(s) from enabled slice(s) */
 		for (s = 0; s < fls(sseu->slice_mask); s++) {
-			u8 subslice_7eu =
-				RUNTIME_INFO(dev_priv)->sseu.subslice_7eu[s];
+			u8 subslice_7eu = info->sseu.subslice_7eu[s];
 
 			sseu->eu_total -= hweight8(subslice_7eu);
 		}
@@ -3938,6 +3920,7 @@ static void i915_print_sseu_info(struct seq_file *m, bool is_available_info,
 static int i915_sseu_status(struct seq_file *m, void *unused)
 {
 	struct drm_i915_private *dev_priv = node_to_i915(m->private);
+	const struct intel_runtime_info *info = RUNTIME_INFO(dev_priv);
 	struct sseu_dev_info sseu;
 	intel_wakeref_t wakeref;
 
@@ -3945,14 +3928,13 @@ static int i915_sseu_status(struct seq_file *m, void *unused)
 		return -ENODEV;
 
 	seq_puts(m, "SSEU Device Info\n");
-	i915_print_sseu_info(m, true, &RUNTIME_INFO(dev_priv)->sseu);
+	i915_print_sseu_info(m, true, &info->sseu);
 
 	seq_puts(m, "SSEU Device Status\n");
 	memset(&sseu, 0, sizeof(sseu));
-	sseu.max_slices = RUNTIME_INFO(dev_priv)->sseu.max_slices;
-	sseu.max_subslices = RUNTIME_INFO(dev_priv)->sseu.max_subslices;
-	sseu.max_eus_per_subslice =
-		RUNTIME_INFO(dev_priv)->sseu.max_eus_per_subslice;
+	intel_sseu_set_info(&sseu, info->sseu.max_slices,
+			    info->sseu.max_subslices,
+			    info->sseu.max_eus_per_subslice);
 
 	with_intel_runtime_pm(&dev_priv->runtime_pm, wakeref) {
 		if (IS_CHERRYVIEW(dev_priv))
@@ -3973,13 +3955,12 @@ static int i915_sseu_status(struct seq_file *m, void *unused)
 static int i915_forcewake_open(struct inode *inode, struct file *file)
 {
 	struct drm_i915_private *i915 = inode->i_private;
+	struct intel_gt *gt = &i915->gt;
 
-	if (INTEL_GEN(i915) < 6)
-		return 0;
-
-	file->private_data =
-		(void *)(uintptr_t)intel_runtime_pm_get(&i915->runtime_pm);
-	intel_uncore_forcewake_user_get(&i915->uncore);
+	atomic_inc(&gt->user_wakeref);
+	intel_gt_pm_get(gt);
+	if (INTEL_GEN(i915) >= 6)
+		intel_uncore_forcewake_user_get(gt->uncore);
 
 	return 0;
 }
@@ -3987,13 +3968,12 @@ static int i915_forcewake_open(struct inode *inode, struct file *file)
 static int i915_forcewake_release(struct inode *inode, struct file *file)
 {
 	struct drm_i915_private *i915 = inode->i_private;
+	struct intel_gt *gt = &i915->gt;
 
-	if (INTEL_GEN(i915) < 6)
-		return 0;
-
-	intel_uncore_forcewake_user_put(&i915->uncore);
-	intel_runtime_pm_put(&i915->runtime_pm,
-			     (intel_wakeref_t)(uintptr_t)file->private_data);
+	if (INTEL_GEN(i915) >= 6)
+		intel_uncore_forcewake_user_put(&i915->uncore);
+	intel_gt_pm_put(gt);
+	atomic_dec(&gt->user_wakeref);
 
 	return 0;
 }
@@ -4302,7 +4282,6 @@ static const struct drm_info_list i915_debugfs_list[] = {
 	{"i915_guc_stage_pool", i915_guc_stage_pool, 0},
 	{"i915_huc_load_status", i915_huc_load_status_info, 0},
 	{"i915_frequency_info", i915_frequency_info, 0},
-	{"i915_hangcheck_info", i915_hangcheck_info, 0},
 	{"i915_drpc_info", i915_drpc_info, 0},
 	{"i915_ring_freq_table", i915_ring_freq_table, 0},
 	{"i915_frontbuffer_tracking", i915_frontbuffer_tracking, 0},
@@ -4339,6 +4318,7 @@ static const struct i915_debugfs_files {
 	const char *name;
 	const struct file_operations *fops;
 } i915_debugfs_files[] = {
+	{"i915_perf_noa_delay", &i915_perf_noa_delay_fops},
 	{"i915_wedged", &i915_wedged_fops},
 	{"i915_cache_sharing", &i915_cache_sharing_fops},
 	{"i915_gem_drop_caches", &i915_drop_caches_fops},
@@ -4528,7 +4508,7 @@ static int i915_dsc_fec_support_show(struct seq_file *m, void *data)
 		intel_dp = enc_to_intel_dp(&intel_attached_encoder(connector)->base);
 		crtc_state = to_intel_crtc_state(crtc->state);
 		seq_printf(m, "DSC_Enabled: %s\n",
-			   yesno(crtc_state->dsc_params.compression_enable));
+			   yesno(crtc_state->dsc.compression_enable));
 		seq_printf(m, "DSC_Sink_Support: %s\n",
 			   yesno(drm_dp_sink_supports_dsc(intel_dp->dsc_dpcd)));
 		seq_printf(m, "Force_DSC_Enable: %s\n",
diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c
index 020696726f9e..3c512c571e60 100644
--- a/drivers/gpu/drm/i915/i915_drv.c
+++ b/drivers/gpu/drm/i915/i915_drv.c
@@ -36,7 +36,6 @@
 #include <linux/pm_runtime.h>
 #include <linux/pnp.h>
 #include <linux/slab.h>
-#include <linux/vgaarb.h>
 #include <linux/vga_switcheroo.h>
 #include <linux/vt.h>
 #include <acpi/video.h>
@@ -54,16 +53,17 @@
 #include "display/intel_display_types.h"
 #include "display/intel_dp.h"
 #include "display/intel_fbdev.h"
-#include "display/intel_gmbus.h"
 #include "display/intel_hotplug.h"
 #include "display/intel_overlay.h"
 #include "display/intel_pipe_crc.h"
 #include "display/intel_sprite.h"
+#include "display/intel_vga.h"
 
 #include "gem/i915_gem_context.h"
 #include "gem/i915_gem_ioctls.h"
 #include "gt/intel_gt.h"
 #include "gt/intel_gt_pm.h"
+#include "gt/intel_rc6.h"
 
 #include "i915_debugfs.h"
 #include "i915_drv.h"
@@ -72,10 +72,12 @@
 #include "i915_perf.h"
 #include "i915_query.h"
 #include "i915_suspend.h"
+#include "i915_switcheroo.h"
 #include "i915_sysfs.h"
 #include "i915_trace.h"
 #include "i915_vgpu.h"
 #include "intel_csr.h"
+#include "intel_memory_region.h"
 #include "intel_pm.h"
 
 static struct drm_driver driver;
@@ -269,182 +271,97 @@ intel_teardown_mchbar(struct drm_i915_private *dev_priv)
 		release_resource(&dev_priv->mch_res);
 }
 
-/* true = enable decode, false = disable decoder */
-static unsigned int i915_vga_set_decode(void *cookie, bool state)
+static int i915_driver_modeset_probe(struct drm_i915_private *i915)
 {
-	struct drm_i915_private *dev_priv = cookie;
-
-	intel_modeset_vga_set_state(dev_priv, state);
-	if (state)
-		return VGA_RSRC_LEGACY_IO | VGA_RSRC_LEGACY_MEM |
-		       VGA_RSRC_NORMAL_IO | VGA_RSRC_NORMAL_MEM;
-	else
-		return VGA_RSRC_NORMAL_IO | VGA_RSRC_NORMAL_MEM;
-}
-
-static int i915_resume_switcheroo(struct drm_i915_private *i915);
-static int i915_suspend_switcheroo(struct drm_i915_private *i915,
-				   pm_message_t state);
-
-static void i915_switcheroo_set_state(struct pci_dev *pdev, enum vga_switcheroo_state state)
-{
-	struct drm_i915_private *i915 = pdev_to_i915(pdev);
-	pm_message_t pmm = { .event = PM_EVENT_SUSPEND };
-
-	if (!i915) {
-		dev_err(&pdev->dev, "DRM not initialized, aborting switch.\n");
-		return;
-	}
-
-	if (state == VGA_SWITCHEROO_ON) {
-		pr_info("switched on\n");
-		i915->drm.switch_power_state = DRM_SWITCH_POWER_CHANGING;
-		/* i915 resume handler doesn't set to D0 */
-		pci_set_power_state(pdev, PCI_D0);
-		i915_resume_switcheroo(i915);
-		i915->drm.switch_power_state = DRM_SWITCH_POWER_ON;
-	} else {
-		pr_info("switched off\n");
-		i915->drm.switch_power_state = DRM_SWITCH_POWER_CHANGING;
-		i915_suspend_switcheroo(i915, pmm);
-		i915->drm.switch_power_state = DRM_SWITCH_POWER_OFF;
-	}
-}
-
-static bool i915_switcheroo_can_switch(struct pci_dev *pdev)
-{
-	struct drm_i915_private *i915 = pdev_to_i915(pdev);
-
-	/*
-	 * FIXME: open_count is protected by drm_global_mutex but that would lead to
-	 * locking inversion with the driver load path. And the access here is
-	 * completely racy anyway. So don't bother with locking for now.
-	 */
-	return i915 && i915->drm.open_count == 0;
-}
-
-static const struct vga_switcheroo_client_ops i915_switcheroo_ops = {
-	.set_gpu_state = i915_switcheroo_set_state,
-	.reprobe = NULL,
-	.can_switch = i915_switcheroo_can_switch,
-};
-
-static int i915_driver_modeset_probe(struct drm_device *dev)
-{
-	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct pci_dev *pdev = dev_priv->drm.pdev;
 	int ret;
 
-	if (i915_inject_probe_failure(dev_priv))
+	if (i915_inject_probe_failure(i915))
 		return -ENODEV;
 
-	if (HAS_DISPLAY(dev_priv)) {
-		ret = drm_vblank_init(&dev_priv->drm,
-				      INTEL_INFO(dev_priv)->num_pipes);
+	if (HAS_DISPLAY(i915) && INTEL_DISPLAY_ENABLED(i915)) {
+		ret = drm_vblank_init(&i915->drm,
+				      INTEL_NUM_PIPES(i915));
 		if (ret)
 			goto out;
 	}
 
-	intel_bios_init(dev_priv);
+	intel_bios_init(i915);
 
-	/* If we have > 1 VGA cards, then we need to arbitrate access
-	 * to the common VGA resources.
-	 *
-	 * If we are a secondary display controller (!PCI_DISPLAY_CLASS_VGA),
-	 * then we do not take part in VGA arbitration and the
-	 * vga_client_register() fails with -ENODEV.
-	 */
-	ret = vga_client_register(pdev, dev_priv, NULL, i915_vga_set_decode);
-	if (ret && ret != -ENODEV)
+	ret = intel_vga_register(i915);
+	if (ret)
 		goto out;
 
 	intel_register_dsm_handler();
 
-	ret = vga_switcheroo_register_client(pdev, &i915_switcheroo_ops, false);
+	ret = i915_switcheroo_register(i915);
 	if (ret)
 		goto cleanup_vga_client;
 
-	/* must happen before intel_power_domains_init_hw() on VLV/CHV */
-	intel_update_rawclk(dev_priv);
-
-	intel_power_domains_init_hw(dev_priv, false);
+	intel_power_domains_init_hw(i915, false);
 
-	intel_csr_ucode_init(dev_priv);
+	intel_csr_ucode_init(i915);
 
-	ret = intel_irq_install(dev_priv);
+	ret = intel_irq_install(i915);
 	if (ret)
 		goto cleanup_csr;
 
-	intel_gmbus_setup(dev_priv);
-
 	/* Important: The output setup functions called by modeset_init need
 	 * working irqs for e.g. gmbus and dp aux transfers. */
-	ret = intel_modeset_init(dev);
+	ret = intel_modeset_init(i915);
 	if (ret)
 		goto cleanup_irq;
 
-	ret = i915_gem_init(dev_priv);
+	ret = i915_gem_init(i915);
 	if (ret)
 		goto cleanup_modeset;
 
-	intel_overlay_setup(dev_priv);
+	intel_overlay_setup(i915);
 
-	if (!HAS_DISPLAY(dev_priv))
+	if (!HAS_DISPLAY(i915) || !INTEL_DISPLAY_ENABLED(i915))
 		return 0;
 
-	ret = intel_fbdev_init(dev);
+	ret = intel_fbdev_init(&i915->drm);
 	if (ret)
 		goto cleanup_gem;
 
 	/* Only enable hotplug handling once the fbdev is fully set up. */
-	intel_hpd_init(dev_priv);
+	intel_hpd_init(i915);
 
-	intel_init_ipc(dev_priv);
+	intel_init_ipc(i915);
 
 	return 0;
 
 cleanup_gem:
-	i915_gem_suspend(dev_priv);
-	i915_gem_driver_remove(dev_priv);
-	i915_gem_driver_release(dev_priv);
+	i915_gem_suspend(i915);
+	i915_gem_driver_remove(i915);
+	i915_gem_driver_release(i915);
 cleanup_modeset:
-	intel_modeset_driver_remove(dev);
+	intel_modeset_driver_remove(i915);
 cleanup_irq:
-	intel_irq_uninstall(dev_priv);
-	intel_gmbus_teardown(dev_priv);
+	intel_irq_uninstall(i915);
 cleanup_csr:
-	intel_csr_ucode_fini(dev_priv);
-	intel_power_domains_driver_remove(dev_priv);
-	vga_switcheroo_unregister_client(pdev);
+	intel_csr_ucode_fini(i915);
+	intel_power_domains_driver_remove(i915);
+	i915_switcheroo_unregister(i915);
 cleanup_vga_client:
-	vga_client_register(pdev, NULL, NULL, NULL);
+	intel_vga_unregister(i915);
 out:
 	return ret;
 }
 
-static int i915_kick_out_firmware_fb(struct drm_i915_private *dev_priv)
+static void i915_driver_modeset_remove(struct drm_i915_private *i915)
 {
-	struct apertures_struct *ap;
-	struct pci_dev *pdev = dev_priv->drm.pdev;
-	struct i915_ggtt *ggtt = &dev_priv->ggtt;
-	bool primary;
-	int ret;
+	intel_modeset_driver_remove(i915);
 
-	ap = alloc_apertures(1);
-	if (!ap)
-		return -ENOMEM;
+	intel_irq_uninstall(i915);
 
-	ap->ranges[0].base = ggtt->gmadr.start;
-	ap->ranges[0].size = ggtt->mappable_end;
-
-	primary =
-		pdev->resource[PCI_ROM_RESOURCE].flags & IORESOURCE_ROM_SHADOW;
+	intel_bios_driver_remove(i915);
 
-	ret = drm_fb_helper_remove_conflicting_framebuffers(ap, "inteldrmfb", primary);
+	i915_switcheroo_unregister(i915);
 
-	kfree(ap);
+	intel_vga_unregister(i915);
 
-	return ret;
+	intel_csr_ucode_fini(i915);
 }
 
 static void intel_init_dpio(struct drm_i915_private *dev_priv)
@@ -601,9 +518,7 @@ static int i915_driver_early_probe(struct drm_i915_private *dev_priv)
 
 	intel_gt_init_early(&dev_priv->gt, dev_priv);
 
-	ret = i915_gem_init_early(dev_priv);
-	if (ret < 0)
-		goto err_gt;
+	i915_gem_init_early(dev_priv);
 
 	/* This must be called before any calls to HAS_PCH_* */
 	intel_detect_pch(dev_priv);
@@ -625,7 +540,6 @@ static int i915_driver_early_probe(struct drm_i915_private *dev_priv)
 
 err_gem:
 	i915_gem_cleanup_early(dev_priv);
-err_gt:
 	intel_gt_driver_late_release(&dev_priv->gt);
 	vlv_free_s0ix_state(dev_priv);
 err_workqueues:
@@ -683,12 +597,10 @@ static int i915_driver_mmio_probe(struct drm_i915_private *dev_priv)
 
 	intel_uc_init_mmio(&dev_priv->gt.uc);
 
-	ret = intel_engines_init_mmio(dev_priv);
+	ret = intel_engines_init_mmio(&dev_priv->gt);
 	if (ret)
 		goto err_uncore;
 
-	i915_gem_init_mmio(dev_priv);
-
 	return 0;
 
 err_uncore:
@@ -706,7 +618,7 @@ err_bridge:
  */
 static void i915_driver_mmio_release(struct drm_i915_private *dev_priv)
 {
-	intel_engines_cleanup(dev_priv);
+	intel_engines_cleanup(&dev_priv->gt);
 	intel_teardown_mchbar(dev_priv);
 	intel_uncore_fini_mmio(&dev_priv->uncore);
 	pci_dev_put(dev_priv->bridge_dev);
@@ -1160,8 +1072,8 @@ intel_get_dram_info(struct drm_i915_private *dev_priv)
 
 static u32 gen9_edram_size_mb(struct drm_i915_private *dev_priv, u32 cap)
 {
-	const unsigned int ways[8] = { 4, 8, 12, 16, 16, 16, 16, 16 };
-	const unsigned int sets[4] = { 1, 1, 2, 2 };
+	static const u8 ways[8] = { 4, 8, 12, 16, 16, 16, 16, 16 };
+	static const u8 sets[4] = { 1, 1, 2, 2 };
 
 	return EDRAM_NUM_BANKS(cap) *
 		ways[EDRAM_WAYS_IDX(cap)] *
@@ -1249,32 +1161,24 @@ static int i915_driver_hw_probe(struct drm_i915_private *dev_priv)
 	if (ret)
 		goto err_perf;
 
-	/*
-	 * WARNING: Apparently we must kick fbdev drivers before vgacon,
-	 * otherwise the vga fbdev driver falls over.
-	 */
-	ret = i915_kick_out_firmware_fb(dev_priv);
-	if (ret) {
-		DRM_ERROR("failed to remove conflicting framebuffer drivers\n");
+	ret = drm_fb_helper_remove_conflicting_pci_framebuffers(pdev, "inteldrmfb");
+	if (ret)
 		goto err_ggtt;
-	}
 
-	ret = vga_remove_vgacon(pdev);
-	if (ret) {
-		DRM_ERROR("failed to remove conflicting VGA console\n");
+	ret = i915_ggtt_init_hw(dev_priv);
+	if (ret)
 		goto err_ggtt;
-	}
 
-	ret = i915_ggtt_init_hw(dev_priv);
+	ret = intel_memory_regions_hw_probe(dev_priv);
 	if (ret)
 		goto err_ggtt;
 
-	intel_gt_init_hw(dev_priv);
+	intel_gt_init_hw_early(&dev_priv->gt, &dev_priv->ggtt);
 
 	ret = i915_ggtt_enable_hw(dev_priv);
 	if (ret) {
 		DRM_ERROR("failed to enable GGTT\n");
-		goto err_ggtt;
+		goto err_mem_regions;
 	}
 
 	pci_set_master(pdev);
@@ -1291,7 +1195,7 @@ static int i915_driver_hw_probe(struct drm_i915_private *dev_priv)
 		if (ret) {
 			DRM_ERROR("failed to set DMA mask\n");
 
-			goto err_ggtt;
+			goto err_mem_regions;
 		}
 	}
 
@@ -1309,16 +1213,13 @@ static int i915_driver_hw_probe(struct drm_i915_private *dev_priv)
 		if (ret) {
 			DRM_ERROR("failed to set DMA mask\n");
 
-			goto err_ggtt;
+			goto err_mem_regions;
 		}
 	}
 
 	pm_qos_add_request(&dev_priv->pm_qos, PM_QOS_CPU_DMA_LATENCY,
 			   PM_QOS_DEFAULT_VALUE);
 
-	/* BIOS often leaves RC6 enabled, but disable it for hw init */
-	intel_sanitize_gt_powersave(dev_priv);
-
 	intel_gt_init_workarounds(dev_priv);
 
 	/* On the 945G/GM, the chipset reports the MSI capability on the
@@ -1364,6 +1265,8 @@ err_msi:
 	if (pdev->msi_enabled)
 		pci_disable_msi(pdev);
 	pm_qos_remove_request(&dev_priv->pm_qos);
+err_mem_regions:
+	intel_memory_regions_driver_release(dev_priv);
 err_ggtt:
 	i915_ggtt_driver_release(dev_priv);
 err_perf:
@@ -1418,14 +1321,13 @@ static void i915_driver_register(struct drm_i915_private *dev_priv)
 	} else
 		DRM_ERROR("Failed to register driver for userspace access!\n");
 
-	if (HAS_DISPLAY(dev_priv)) {
+	if (HAS_DISPLAY(dev_priv) && INTEL_DISPLAY_ENABLED(dev_priv)) {
 		/* Must be done after probing outputs */
 		intel_opregion_register(dev_priv);
 		acpi_video_register();
 	}
 
-	if (IS_GEN(dev_priv, 5))
-		intel_gpu_ips_init(dev_priv);
+	intel_gt_driver_register(&dev_priv->gt);
 
 	intel_audio_init(dev_priv);
 
@@ -1442,7 +1344,7 @@ static void i915_driver_register(struct drm_i915_private *dev_priv)
 	 * We need to coordinate the hotplugs with the asynchronous fbdev
 	 * configuration, for which we use the fbdev->async_cookie.
 	 */
-	if (HAS_DISPLAY(dev_priv))
+	if (HAS_DISPLAY(dev_priv) && INTEL_DISPLAY_ENABLED(dev_priv))
 		drm_kms_helper_poll_init(dev);
 
 	intel_power_domains_enable(dev_priv);
@@ -1468,7 +1370,7 @@ static void i915_driver_unregister(struct drm_i915_private *dev_priv)
 	 */
 	drm_kms_helper_poll_fini(&dev_priv->drm);
 
-	intel_gpu_ips_teardown();
+	intel_gt_driver_unregister(&dev_priv->gt);
 	acpi_video_unregister();
 	intel_opregion_unregister(dev_priv);
 
@@ -1577,6 +1479,23 @@ int i915_driver_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 	if (!i915_modparams.nuclear_pageflip && match_info->gen < 5)
 		dev_priv->drm.driver_features &= ~DRIVER_ATOMIC;
 
+	/*
+	 * Check if we support fake LMEM -- for now we only unleash this for
+	 * the live selftests(test-and-exit).
+	 */
+#if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
+	if (IS_ENABLED(CONFIG_DRM_I915_UNSTABLE_FAKE_LMEM)) {
+		if (INTEL_GEN(dev_priv) >= 9 && i915_selftest.live < 0 &&
+		    i915_modparams.fake_lmem_start) {
+			mkwrite_device_info(dev_priv)->memory_regions =
+				REGION_SMEM | REGION_LMEM | REGION_STOLEN;
+			mkwrite_device_info(dev_priv)->is_dgfx = true;
+			GEM_BUG_ON(!HAS_LMEM(dev_priv));
+			GEM_BUG_ON(!IS_DGFX(dev_priv));
+		}
+	}
+#endif
+
 	ret = pci_enable_device(pdev);
 	if (ret)
 		goto out_fini;
@@ -1597,7 +1516,7 @@ int i915_driver_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 	if (ret < 0)
 		goto out_cleanup_mmio;
 
-	ret = i915_driver_modeset_probe(&dev_priv->drm);
+	ret = i915_driver_modeset_probe(dev_priv);
 	if (ret < 0)
 		goto out_cleanup_hw;
 
@@ -1611,10 +1530,8 @@ int i915_driver_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 
 out_cleanup_hw:
 	i915_driver_hw_remove(dev_priv);
+	intel_memory_regions_driver_release(dev_priv);
 	i915_ggtt_driver_release(dev_priv);
-
-	/* Paranoia: make sure we have disabled everything before we exit. */
-	intel_sanitize_gt_powersave(dev_priv);
 out_cleanup_mmio:
 	i915_driver_mmio_release(dev_priv);
 out_runtime_pm_put:
@@ -1630,8 +1547,6 @@ out_fini:
 
 void i915_driver_remove(struct drm_i915_private *i915)
 {
-	struct pci_dev *pdev = i915->drm.pdev;
-
 	disable_rpm_wakeref_asserts(&i915->runtime_pm);
 
 	i915_driver_unregister(i915);
@@ -1652,19 +1567,9 @@ void i915_driver_remove(struct drm_i915_private *i915)
 
 	intel_gvt_driver_remove(i915);
 
-	intel_modeset_driver_remove(&i915->drm);
+	i915_driver_modeset_remove(i915);
 
-	intel_bios_driver_remove(i915);
-
-	vga_switcheroo_unregister_client(pdev);
-	vga_client_register(pdev, NULL, NULL, NULL);
-
-	intel_csr_ucode_fini(i915);
-
-	/* Free error state after interrupts are fully disabled. */
-	cancel_delayed_work_sync(&i915->gt.hangcheck.work);
 	i915_reset_error_state(i915);
-
 	i915_gem_driver_remove(i915);
 
 	intel_power_domains_driver_remove(i915);
@@ -1683,11 +1588,9 @@ static void i915_driver_release(struct drm_device *dev)
 
 	i915_gem_driver_release(dev_priv);
 
+	intel_memory_regions_driver_release(dev_priv);
 	i915_ggtt_driver_release(dev_priv);
 
-	/* Paranoia: make sure we have disabled everything before we exit. */
-	intel_sanitize_gt_powersave(dev_priv);
-
 	i915_driver_mmio_release(dev_priv);
 
 	enable_rpm_wakeref_asserts(rpm);
@@ -1731,12 +1634,10 @@ static void i915_driver_postclose(struct drm_device *dev, struct drm_file *file)
 {
 	struct drm_i915_file_private *file_priv = file->driver_priv;
 
-	mutex_lock(&dev->struct_mutex);
 	i915_gem_context_close(file);
 	i915_gem_release(dev, file);
-	mutex_unlock(&dev->struct_mutex);
 
-	kfree(file_priv);
+	kfree_rcu(file_priv, rcu);
 
 	/* Catch up with all the deferred frees from "this" client */
 	i915_gem_flush_free_objects(to_i915(dev));
@@ -1891,8 +1792,7 @@ out:
 	return ret;
 }
 
-static int
-i915_suspend_switcheroo(struct drm_i915_private *i915, pm_message_t state)
+int i915_suspend_switcheroo(struct drm_i915_private *i915, pm_message_t state)
 {
 	int error;
 
@@ -1916,14 +1816,18 @@ static int i915_drm_resume(struct drm_device *dev)
 	int ret;
 
 	disable_rpm_wakeref_asserts(&dev_priv->runtime_pm);
-	intel_sanitize_gt_powersave(dev_priv);
 
-	i915_gem_sanitize(dev_priv);
+	intel_rc6_ctx_wa_resume(&dev_priv->gt.rc6);
+
+	intel_gt_sanitize(&dev_priv->gt, true);
 
 	ret = i915_ggtt_enable_hw(dev_priv);
 	if (ret)
 		DRM_ERROR("failed to re-enable GGTT\n");
 
+	i915_gem_restore_gtt_mappings(dev_priv);
+	i915_gem_restore_fences(&dev_priv->ggtt);
+
 	intel_csr_ucode_resume(dev_priv);
 
 	i915_restore_state(dev_priv);
@@ -1947,7 +1851,7 @@ static int i915_drm_resume(struct drm_device *dev)
 
 	i915_gem_resume(dev_priv);
 
-	intel_modeset_init_hw(dev);
+	intel_modeset_init_hw(dev_priv);
 	intel_init_clock_gating(dev_priv);
 
 	spin_lock_irq(&dev_priv->irq_lock);
@@ -2044,18 +1948,14 @@ static int i915_drm_resume_early(struct drm_device *dev)
 
 	intel_display_power_resume_early(dev_priv);
 
-	intel_sanitize_gt_powersave(dev_priv);
-
 	intel_power_domains_resume(dev_priv);
 
-	intel_gt_sanitize(&dev_priv->gt, true);
-
 	enable_rpm_wakeref_asserts(&dev_priv->runtime_pm);
 
 	return ret;
 }
 
-static int i915_resume_switcheroo(struct drm_i915_private *i915)
+int i915_resume_switcheroo(struct drm_i915_private *i915)
 {
 	int ret;
 
@@ -2588,9 +2488,6 @@ static int intel_runtime_suspend(struct device *kdev)
 	struct intel_runtime_pm *rpm = &dev_priv->runtime_pm;
 	int ret = 0;
 
-	if (WARN_ON_ONCE(!(dev_priv->gt_pm.rc6.enabled && HAS_RC6(dev_priv))))
-		return -ENODEV;
-
 	if (WARN_ON_ONCE(!HAS_RUNTIME_PM(dev_priv)))
 		return -ENODEV;
 
@@ -2623,7 +2520,7 @@ static int intel_runtime_suspend(struct device *kdev)
 
 		intel_gt_runtime_resume(&dev_priv->gt);
 
-		i915_gem_restore_fences(dev_priv);
+		i915_gem_restore_fences(&dev_priv->ggtt);
 
 		enable_rpm_wakeref_asserts(rpm);
 
@@ -2703,7 +2600,7 @@ static int intel_runtime_resume(struct device *kdev)
 	 * we can do is to hope that things will still work (and disable RPM).
 	 */
 	intel_gt_runtime_resume(&dev_priv->gt);
-	i915_gem_restore_fences(dev_priv);
+	i915_gem_restore_fences(&dev_priv->ggtt);
 
 	/*
 	 * On VLV/CHV display interrupts are part of the display
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 772154e4073e..e29bc137e7ba 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -67,6 +67,7 @@
 #include "display/intel_display.h"
 #include "display/intel_display_power.h"
 #include "display/intel_dpll_mgr.h"
+#include "display/intel_dsb.h"
 #include "display/intel_frontbuffer.h"
 #include "display/intel_gmbus.h"
 #include "display/intel_opregion.h"
@@ -84,6 +85,7 @@
 #include "intel_device_info.h"
 #include "intel_pch.h"
 #include "intel_runtime_pm.h"
+#include "intel_memory_region.h"
 #include "intel_uncore.h"
 #include "intel_wakeref.h"
 #include "intel_wopcm.h"
@@ -92,12 +94,15 @@
 #include "i915_gem_fence_reg.h"
 #include "i915_gem_gtt.h"
 #include "i915_gpu_error.h"
+#include "i915_perf_types.h"
 #include "i915_request.h"
 #include "i915_scheduler.h"
 #include "gt/intel_timeline.h"
 #include "i915_vma.h"
 #include "i915_irq.h"
 
+#include "intel_region_lmem.h"
+
 #include "intel_gvt.h"
 
 /* General customization:
@@ -105,8 +110,8 @@
 
 #define DRIVER_NAME		"i915"
 #define DRIVER_DESC		"Intel Graphics"
-#define DRIVER_DATE		"20190822"
-#define DRIVER_TIMESTAMP	1566477988
+#define DRIVER_DATE		"20191101"
+#define DRIVER_TIMESTAMP	1572604873
 
 struct drm_i915_gem_object;
 
@@ -185,7 +190,11 @@ struct i915_mmu_object;
 
 struct drm_i915_file_private {
 	struct drm_i915_private *dev_priv;
-	struct drm_file *file;
+
+	union {
+		struct drm_file *file;
+		struct rcu_head rcu;
+	};
 
 	struct {
 		spinlock_t lock;
@@ -272,6 +281,7 @@ struct drm_i915_display_funcs {
 	int (*compute_global_watermarks)(struct intel_atomic_state *state);
 	void (*update_wm)(struct intel_crtc *crtc);
 	int (*modeset_calc_cdclk)(struct intel_atomic_state *state);
+	u8 (*calc_voltage_level)(int cdclk);
 	/* Returns the active state of the crtc, and if the crtc is active,
 	 * fills out the pipe-config with the hw state. */
 	bool (*get_pipe_config)(struct intel_crtc *,
@@ -284,7 +294,8 @@ struct drm_i915_display_funcs {
 			    struct intel_atomic_state *old_state);
 	void (*crtc_disable)(struct intel_crtc_state *old_crtc_state,
 			     struct intel_atomic_state *old_state);
-	void (*update_crtcs)(struct intel_atomic_state *state);
+	void (*commit_modeset_enables)(struct intel_atomic_state *state);
+	void (*commit_modeset_disables)(struct intel_atomic_state *state);
 	void (*audio_codec_enable)(struct intel_encoder *encoder,
 				   const struct intel_crtc_state *crtc_state,
 				   const struct drm_connector_state *conn_state);
@@ -331,6 +342,7 @@ struct intel_csr {
 	i915_reg_t mmioaddr[20];
 	u32 mmiodata[20];
 	u32 dc_state;
+	u32 target_dc_state;
 	u32 allowed_dc_mask;
 	intel_wakeref_t wakeref;
 };
@@ -479,6 +491,7 @@ struct i915_psr {
 	bool enabled;
 	struct intel_dp *dp;
 	enum pipe pipe;
+	enum transcoder transcoder;
 	bool active;
 	struct work_struct work;
 	unsigned busy_frontbuffer_bits;
@@ -492,6 +505,9 @@ struct i915_psr {
 	bool sink_not_reliable;
 	bool irq_aux_error;
 	u16 su_x_granularity;
+	bool dc3co_enabled;
+	u32 dc3co_exit_delay;
+	struct delayed_work idle_work;
 };
 
 #define QUIRK_LVDS_SSC_DISABLE (1<<1)
@@ -529,106 +545,6 @@ struct i915_suspend_saved_registers {
 
 struct vlv_s0ix_state;
 
-struct intel_rps_ei {
-	ktime_t ktime;
-	u32 render_c0;
-	u32 media_c0;
-};
-
-struct intel_rps {
-	struct mutex lock; /* protects enabling and the worker */
-
-	/*
-	 * work, interrupts_enabled and pm_iir are protected by
-	 * dev_priv->irq_lock
-	 */
-	struct work_struct work;
-	bool interrupts_enabled;
-	u32 pm_iir;
-
-	/* PM interrupt bits that should never be masked */
-	u32 pm_intrmsk_mbz;
-
-	/* Frequencies are stored in potentially platform dependent multiples.
-	 * In other words, *_freq needs to be multiplied by X to be interesting.
-	 * Soft limits are those which are used for the dynamic reclocking done
-	 * by the driver (raise frequencies under heavy loads, and lower for
-	 * lighter loads). Hard limits are those imposed by the hardware.
-	 *
-	 * A distinction is made for overclocking, which is never enabled by
-	 * default, and is considered to be above the hard limit if it's
-	 * possible at all.
-	 */
-	u8 cur_freq;		/* Current frequency (cached, may not == HW) */
-	u8 min_freq_softlimit;	/* Minimum frequency permitted by the driver */
-	u8 max_freq_softlimit;	/* Max frequency permitted by the driver */
-	u8 max_freq;		/* Maximum frequency, RP0 if not overclocking */
-	u8 min_freq;		/* AKA RPn. Minimum frequency */
-	u8 boost_freq;		/* Frequency to request when wait boosting */
-	u8 idle_freq;		/* Frequency to request when we are idle */
-	u8 efficient_freq;	/* AKA RPe. Pre-determined balanced frequency */
-	u8 rp1_freq;		/* "less than" RP0 power/freqency */
-	u8 rp0_freq;		/* Non-overclocked max frequency. */
-	u16 gpll_ref_freq;	/* vlv/chv GPLL reference frequency */
-
-	int last_adj;
-
-	struct {
-		struct mutex mutex;
-
-		enum { LOW_POWER, BETWEEN, HIGH_POWER } mode;
-		unsigned int interactive;
-
-		u8 up_threshold; /* Current %busy required to uplock */
-		u8 down_threshold; /* Current %busy required to downclock */
-	} power;
-
-	bool enabled;
-	atomic_t num_waiters;
-	atomic_t boosts;
-
-	/* manual wa residency calculations */
-	struct intel_rps_ei ei;
-};
-
-struct intel_rc6 {
-	bool enabled;
-	u64 prev_hw_residency[4];
-	u64 cur_residency[4];
-};
-
-struct intel_llc_pstate {
-	bool enabled;
-};
-
-struct intel_gen6_power_mgmt {
-	struct intel_rps rps;
-	struct intel_rc6 rc6;
-	struct intel_llc_pstate llc_pstate;
-};
-
-/* defined intel_pm.c */
-extern spinlock_t mchdev_lock;
-
-struct intel_ilk_power_mgmt {
-	u8 cur_delay;
-	u8 min_delay;
-	u8 max_delay;
-	u8 fmax;
-	u8 fstart;
-
-	u64 last_count1;
-	unsigned long last_time1;
-	unsigned long chipset_power;
-	u64 last_count2;
-	u64 last_time2;
-	unsigned long gfx_power;
-	u8 corr;
-
-	int c_m;
-	int r_t;
-};
-
 #define MAX_L3_SLICES 2
 struct intel_l3_parity {
 	u32 *remap_info[MAX_L3_SLICES];
@@ -677,6 +593,8 @@ struct i915_gem_mm {
 	 */
 	struct vfsmount *gemfs;
 
+	struct intel_memory_region *regions[INTEL_REGION_UNKNOWN];
+
 	struct notifier_block oom_notifier;
 	struct notifier_block vmap_notifier;
 	struct shrinker shrinker;
@@ -688,11 +606,6 @@ struct i915_gem_mm {
 	 */
 	struct workqueue_struct *userptr_wq;
 
-	/** Bit 6 swizzling required for X tiling */
-	u32 bit_6_swizzle_x;
-	/** Bit 6 swizzling required for Y tiling */
-	u32 bit_6_swizzle_y;
-
 	/* shrinker accounting, also useful for userland debugging */
 	u64 shrink_memory;
 	u32 shrink_count;
@@ -973,305 +886,6 @@ struct intel_wm_config {
 	bool sprites_scaled;
 };
 
-struct i915_oa_format {
-	u32 format;
-	int size;
-};
-
-struct i915_oa_reg {
-	i915_reg_t addr;
-	u32 value;
-};
-
-struct i915_oa_config {
-	char uuid[UUID_STRING_LEN + 1];
-	int id;
-
-	const struct i915_oa_reg *mux_regs;
-	u32 mux_regs_len;
-	const struct i915_oa_reg *b_counter_regs;
-	u32 b_counter_regs_len;
-	const struct i915_oa_reg *flex_regs;
-	u32 flex_regs_len;
-
-	struct attribute_group sysfs_metric;
-	struct attribute *attrs[2];
-	struct device_attribute sysfs_metric_id;
-
-	atomic_t ref_count;
-};
-
-struct i915_perf_stream;
-
-/**
- * struct i915_perf_stream_ops - the OPs to support a specific stream type
- */
-struct i915_perf_stream_ops {
-	/**
-	 * @enable: Enables the collection of HW samples, either in response to
-	 * `I915_PERF_IOCTL_ENABLE` or implicitly called when stream is opened
-	 * without `I915_PERF_FLAG_DISABLED`.
-	 */
-	void (*enable)(struct i915_perf_stream *stream);
-
-	/**
-	 * @disable: Disables the collection of HW samples, either in response
-	 * to `I915_PERF_IOCTL_DISABLE` or implicitly called before destroying
-	 * the stream.
-	 */
-	void (*disable)(struct i915_perf_stream *stream);
-
-	/**
-	 * @poll_wait: Call poll_wait, passing a wait queue that will be woken
-	 * once there is something ready to read() for the stream
-	 */
-	void (*poll_wait)(struct i915_perf_stream *stream,
-			  struct file *file,
-			  poll_table *wait);
-
-	/**
-	 * @wait_unlocked: For handling a blocking read, wait until there is
-	 * something to ready to read() for the stream. E.g. wait on the same
-	 * wait queue that would be passed to poll_wait().
-	 */
-	int (*wait_unlocked)(struct i915_perf_stream *stream);
-
-	/**
-	 * @read: Copy buffered metrics as records to userspace
-	 * **buf**: the userspace, destination buffer
-	 * **count**: the number of bytes to copy, requested by userspace
-	 * **offset**: zero at the start of the read, updated as the read
-	 * proceeds, it represents how many bytes have been copied so far and
-	 * the buffer offset for copying the next record.
-	 *
-	 * Copy as many buffered i915 perf samples and records for this stream
-	 * to userspace as will fit in the given buffer.
-	 *
-	 * Only write complete records; returning -%ENOSPC if there isn't room
-	 * for a complete record.
-	 *
-	 * Return any error condition that results in a short read such as
-	 * -%ENOSPC or -%EFAULT, even though these may be squashed before
-	 * returning to userspace.
-	 */
-	int (*read)(struct i915_perf_stream *stream,
-		    char __user *buf,
-		    size_t count,
-		    size_t *offset);
-
-	/**
-	 * @destroy: Cleanup any stream specific resources.
-	 *
-	 * The stream will always be disabled before this is called.
-	 */
-	void (*destroy)(struct i915_perf_stream *stream);
-};
-
-/**
- * struct i915_perf_stream - state for a single open stream FD
- */
-struct i915_perf_stream {
-	/**
-	 * @dev_priv: i915 drm device
-	 */
-	struct drm_i915_private *dev_priv;
-
-	/**
-	 * @link: Links the stream into ``&drm_i915_private->streams``
-	 */
-	struct list_head link;
-
-	/**
-	 * @wakeref: As we keep the device awake while the perf stream is
-	 * active, we track our runtime pm reference for later release.
-	 */
-	intel_wakeref_t wakeref;
-
-	/**
-	 * @sample_flags: Flags representing the `DRM_I915_PERF_PROP_SAMPLE_*`
-	 * properties given when opening a stream, representing the contents
-	 * of a single sample as read() by userspace.
-	 */
-	u32 sample_flags;
-
-	/**
-	 * @sample_size: Considering the configured contents of a sample
-	 * combined with the required header size, this is the total size
-	 * of a single sample record.
-	 */
-	int sample_size;
-
-	/**
-	 * @ctx: %NULL if measuring system-wide across all contexts or a
-	 * specific context that is being monitored.
-	 */
-	struct i915_gem_context *ctx;
-
-	/**
-	 * @enabled: Whether the stream is currently enabled, considering
-	 * whether the stream was opened in a disabled state and based
-	 * on `I915_PERF_IOCTL_ENABLE` and `I915_PERF_IOCTL_DISABLE` calls.
-	 */
-	bool enabled;
-
-	/**
-	 * @ops: The callbacks providing the implementation of this specific
-	 * type of configured stream.
-	 */
-	const struct i915_perf_stream_ops *ops;
-
-	/**
-	 * @oa_config: The OA configuration used by the stream.
-	 */
-	struct i915_oa_config *oa_config;
-
-	/**
-	 * The OA context specific information.
-	 */
-	struct intel_context *pinned_ctx;
-	u32 specific_ctx_id;
-	u32 specific_ctx_id_mask;
-
-	struct hrtimer poll_check_timer;
-	wait_queue_head_t poll_wq;
-	bool pollin;
-
-	bool periodic;
-	int period_exponent;
-
-	/**
-	 * State of the OA buffer.
-	 */
-	struct {
-		struct i915_vma *vma;
-		u8 *vaddr;
-		u32 last_ctx_id;
-		int format;
-		int format_size;
-		int size_exponent;
-
-		/**
-		 * Locks reads and writes to all head/tail state
-		 *
-		 * Consider: the head and tail pointer state needs to be read
-		 * consistently from a hrtimer callback (atomic context) and
-		 * read() fop (user context) with tail pointer updates happening
-		 * in atomic context and head updates in user context and the
-		 * (unlikely) possibility of read() errors needing to reset all
-		 * head/tail state.
-		 *
-		 * Note: Contention/performance aren't currently a significant
-		 * concern here considering the relatively low frequency of
-		 * hrtimer callbacks (5ms period) and that reads typically only
-		 * happen in response to a hrtimer event and likely complete
-		 * before the next callback.
-		 *
-		 * Note: This lock is not held *while* reading and copying data
-		 * to userspace so the value of head observed in htrimer
-		 * callbacks won't represent any partial consumption of data.
-		 */
-		spinlock_t ptr_lock;
-
-		/**
-		 * One 'aging' tail pointer and one 'aged' tail pointer ready to
-		 * used for reading.
-		 *
-		 * Initial values of 0xffffffff are invalid and imply that an
-		 * update is required (and should be ignored by an attempted
-		 * read)
-		 */
-		struct {
-			u32 offset;
-		} tails[2];
-
-		/**
-		 * Index for the aged tail ready to read() data up to.
-		 */
-		unsigned int aged_tail_idx;
-
-		/**
-		 * A monotonic timestamp for when the current aging tail pointer
-		 * was read; used to determine when it is old enough to trust.
-		 */
-		u64 aging_timestamp;
-
-		/**
-		 * Although we can always read back the head pointer register,
-		 * we prefer to avoid trusting the HW state, just to avoid any
-		 * risk that some hardware condition could * somehow bump the
-		 * head pointer unpredictably and cause us to forward the wrong
-		 * OA buffer data to userspace.
-		 */
-		u32 head;
-	} oa_buffer;
-};
-
-/**
- * struct i915_oa_ops - Gen specific implementation of an OA unit stream
- */
-struct i915_oa_ops {
-	/**
-	 * @is_valid_b_counter_reg: Validates register's address for
-	 * programming boolean counters for a particular platform.
-	 */
-	bool (*is_valid_b_counter_reg)(struct drm_i915_private *dev_priv,
-				       u32 addr);
-
-	/**
-	 * @is_valid_mux_reg: Validates register's address for programming mux
-	 * for a particular platform.
-	 */
-	bool (*is_valid_mux_reg)(struct drm_i915_private *dev_priv, u32 addr);
-
-	/**
-	 * @is_valid_flex_reg: Validates register's address for programming
-	 * flex EU filtering for a particular platform.
-	 */
-	bool (*is_valid_flex_reg)(struct drm_i915_private *dev_priv, u32 addr);
-
-	/**
-	 * @enable_metric_set: Selects and applies any MUX configuration to set
-	 * up the Boolean and Custom (B/C) counters that are part of the
-	 * counter reports being sampled. May apply system constraints such as
-	 * disabling EU clock gating as required.
-	 */
-	int (*enable_metric_set)(struct i915_perf_stream *stream);
-
-	/**
-	 * @disable_metric_set: Remove system constraints associated with using
-	 * the OA unit.
-	 */
-	void (*disable_metric_set)(struct i915_perf_stream *stream);
-
-	/**
-	 * @oa_enable: Enable periodic sampling
-	 */
-	void (*oa_enable)(struct i915_perf_stream *stream);
-
-	/**
-	 * @oa_disable: Disable periodic sampling
-	 */
-	void (*oa_disable)(struct i915_perf_stream *stream);
-
-	/**
-	 * @read: Copy data from the circular OA buffer into a given userspace
-	 * buffer.
-	 */
-	int (*read)(struct i915_perf_stream *stream,
-		    char __user *buf,
-		    size_t count,
-		    size_t *offset);
-
-	/**
-	 * @oa_hw_tail_read: read the OA tail pointer register
-	 *
-	 * In particular this enables us to share all the fiddly code for
-	 * handling the OA unit tail pointer race that affects multiple
-	 * generations.
-	 */
-	u32 (*oa_hw_tail_read)(struct i915_perf_stream *stream);
-};
-
 struct intel_cdclk_state {
 	unsigned int cdclk, vco, ref, bypass;
 	u8 voltage_level;
@@ -1331,11 +945,11 @@ struct drm_i915_private {
 	 */
 	u32 gpio_mmio_base;
 
+	u32 hsw_psr_mmio_adjust;
+
 	/* MMIO base address for MIPI regs */
 	u32 mipi_mmio_base;
 
-	u32 psr_mmio_base;
-
 	u32 pps_mmio_base;
 
 	wait_queue_head_t gmbus_wait_queue;
@@ -1367,7 +981,6 @@ struct drm_i915_private {
 		u32 irq_mask;
 		u32 de_irq_mask[I915_MAX_PIPES];
 	};
-	u32 pm_rps_events;
 	u32 pipestat_irq_mask[I915_MAX_PIPES];
 
 	struct i915_hotplug hotplug;
@@ -1397,13 +1010,14 @@ struct drm_i915_private {
 	unsigned int fdi_pll_freq;
 	unsigned int czclk_freq;
 
+	/*
+	 * For reading holding any crtc lock is sufficient,
+	 * for writing must hold all of them.
+	 */
 	struct {
 		/*
 		 * The current logical cdclk state.
 		 * See intel_atomic_state.cdclk.logical
-		 *
-		 * For reading holding any crtc lock is sufficient,
-		 * for writing must hold all of them.
 		 */
 		struct intel_cdclk_state logical;
 		/*
@@ -1414,6 +1028,9 @@ struct drm_i915_private {
 		/* The current hardware cdclk state */
 		struct intel_cdclk_state hw;
 
+		/* cdclk, divider, and ratio table from bspec */
+		const struct intel_cdclk_vals *table;
+
 		int force_min_cdclk;
 	} cdclk;
 
@@ -1428,6 +1045,8 @@ struct drm_i915_private {
 
 	/* ordered wq for modesets */
 	struct workqueue_struct *modeset_wq;
+	/* unbound hipri wq for page flips/plane updates */
+	struct workqueue_struct *flip_wq;
 
 	/* Display functions */
 	struct drm_i915_display_funcs display;
@@ -1468,7 +1087,11 @@ struct drm_i915_private {
 	 */
 	struct mutex dpll_lock;
 
-	unsigned int active_crtcs;
+	/*
+	 * For reading active_pipes, min_cdclk, min_voltage_level holding
+	 * any crtc lock is sufficient, for writing must hold all of them.
+	 */
+	u8 active_pipes;
 	/* minimum acceptable cdclk for each pipe */
 	int min_cdclk[I915_MAX_PIPES];
 	/* minimum acceptable voltage level for each pipe */
@@ -1497,13 +1120,6 @@ struct drm_i915_private {
 	 */
 	u32 edram_size_mb;
 
-	/* gen6+ GT PM state */
-	struct intel_gen6_power_mgmt gt_pm;
-
-	/* ilk-only ips/rps state. Everything in here is protected by the global
-	 * mchdev_lock in intel_pm.c */
-	struct intel_ilk_power_mgmt ips;
-
 	struct i915_power_domains power_domains;
 
 	struct i915_psr psr;
@@ -1528,25 +1144,7 @@ struct drm_i915_private {
 	 */
 	struct mutex av_mutex;
 	int audio_power_refcount;
-
-	struct {
-		struct mutex mutex;
-		struct list_head list;
-		struct llist_head free_list;
-		struct work_struct free_work;
-
-		/* The hw wants to have a stable context identifier for the
-		 * lifetime of the context (for OA, PASID, faults, etc).
-		 * This is limited in execlists to 21 bits.
-		 */
-		struct ida hw_ida;
-#define MAX_CONTEXT_HW_ID (1<<21) /* exclusive */
-#define MAX_GUC_CONTEXT_HW_ID (1 << 20) /* exclusive */
-#define GEN11_MAX_CONTEXT_HW_ID (1<<11) /* exclusive */
-/* in Gen12 ID 0x7FF is reserved to indicate idle */
-#define GEN12_MAX_CONTEXT_HW_ID	(GEN11_MAX_CONTEXT_HW_ID - 1)
-		struct list_head hw_id_list;
-	} contexts;
+	u32 audio_freq_cntrl;
 
 	u32 fdi_rx_config;
 
@@ -1572,6 +1170,8 @@ struct drm_i915_private {
 		I915_SAGV_NOT_CONTROLLED
 	} sagv_status;
 
+	u32 sagv_block_time_us;
+
 	struct {
 		/*
 		 * Raw watermark latency values:
@@ -1642,61 +1242,7 @@ struct drm_i915_private {
 
 	struct intel_runtime_pm runtime_pm;
 
-	struct {
-		bool initialized;
-
-		struct kobject *metrics_kobj;
-		struct ctl_table_header *sysctl_header;
-
-		/*
-		 * Lock associated with adding/modifying/removing OA configs
-		 * in dev_priv->perf.metrics_idr.
-		 */
-		struct mutex metrics_lock;
-
-		/*
-		 * List of dynamic configurations, you need to hold
-		 * dev_priv->perf.metrics_lock to access it.
-		 */
-		struct idr metrics_idr;
-
-		/*
-		 * Lock associated with anything below within this structure
-		 * except exclusive_stream.
-		 */
-		struct mutex lock;
-		struct list_head streams;
-
-		/*
-		 * The stream currently using the OA unit. If accessed
-		 * outside a syscall associated to its file
-		 * descriptor, you need to hold
-		 * dev_priv->drm.struct_mutex.
-		 */
-		struct i915_perf_stream *exclusive_stream;
-
-		/**
-		 * For rate limiting any notifications of spurious
-		 * invalid OA reports
-		 */
-		struct ratelimit_state spurious_report_rs;
-
-		struct i915_oa_config test_config;
-
-		u32 gen7_latched_oastatus1;
-		u32 ctx_oactxctrl_offset;
-		u32 ctx_flexeu0_offset;
-
-		/**
-		 * The RPT_ID/reason field for Gen8+ includes a bit
-		 * to determine if the CTX ID in the report is valid
-		 * but the specific bit differs between Gen 8 and 9
-		 */
-		u32 gen8_valid_ctx_bit;
-
-		struct i915_oa_ops ops;
-		const struct i915_oa_format *oa_formats;
-	} perf;
+	struct i915_perf perf;
 
 	/* Abstract the submission mechanism (legacy ringbuffer or execlists) away */
 	struct intel_gt gt;
@@ -1704,32 +1250,19 @@ struct drm_i915_private {
 	struct {
 		struct notifier_block pm_notifier;
 
-		/**
-		 * We leave the user IRQ off as much as possible,
-		 * but this means that requests will finish and never
-		 * be retired once the system goes idle. Set a timer to
-		 * fire periodically while the ring is running. When it
-		 * fires, go retire requests.
-		 */
-		struct delayed_work retire_work;
-
-		/**
-		 * When we detect an idle GPU, we want to turn on
-		 * powersaving features. So once we see that there
-		 * are no more requests outstanding and no more
-		 * arrive within a small period of time, we fire
-		 * off the idle_work.
-		 */
-		struct work_struct idle_work;
+		struct i915_gem_contexts {
+			spinlock_t lock; /* locks list */
+			struct list_head list;
+
+			struct llist_head free_list;
+			struct work_struct free_work;
+		} contexts;
 	} gem;
 
-	/* For i945gm vblank irq vs. C3 workaround */
-	struct {
-		struct work_struct work;
-		struct pm_qos_request pm_qos;
-		u8 c3_disable_latency;
-		u8 enabled;
-	} i945gm_vblank;
+	u8 pch_ssc_use;
+
+	/* For i915gm/i945gm vblank irq workaround */
+	u8 vblank_enabled;
 
 	/* perform PHY state sanity checks? */
 	bool chv_phy_assert[2];
@@ -1792,10 +1325,10 @@ static inline struct drm_i915_private *pdev_to_i915(struct pci_dev *pdev)
 		for_each_if ((engine__) = (dev_priv__)->engine[(id__)])
 
 /* Iterator over subset of engines selected by mask */
-#define for_each_engine_masked(engine__, dev_priv__, mask__, tmp__) \
-	for ((tmp__) = (mask__) & INTEL_INFO(dev_priv__)->engine_mask; \
+#define for_each_engine_masked(engine__, gt__, mask__, tmp__) \
+	for ((tmp__) = (mask__) & INTEL_INFO((gt__)->i915)->engine_mask; \
 	     (tmp__) ? \
-	     ((engine__) = (dev_priv__)->engine[__mask_next_bit(tmp__)]), 1 : \
+	     ((engine__) = (gt__)->engine[__mask_next_bit(tmp__)]), 1 : \
 	     0;)
 
 #define rb_to_uabi_engine(rb) \
@@ -1851,6 +1384,8 @@ static inline struct drm_i915_private *pdev_to_i915(struct pci_dev *pdev)
 	(BUILD_BUG_ON_ZERO(!__builtin_constant_p(n)) + \
 	 INTEL_INFO(dev_priv)->gen == (n))
 
+#define HAS_DSB(dev_priv)	(INTEL_INFO(dev_priv)->display.has_dsb)
+
 /*
  * Return true if revision is in range [since,until] inclusive.
  *
@@ -1922,6 +1457,7 @@ IS_SUBPLATFORM(const struct drm_i915_private *i915,
 }
 
 #define IS_MOBILE(dev_priv)	(INTEL_INFO(dev_priv)->is_mobile)
+#define IS_DGFX(dev_priv)   (INTEL_INFO(dev_priv)->is_dgfx)
 
 #define IS_I830(dev_priv)	IS_PLATFORM(dev_priv, INTEL_I830)
 #define IS_I845G(dev_priv)	IS_PLATFORM(dev_priv, INTEL_I845G)
@@ -2056,6 +1592,11 @@ IS_SUBPLATFORM(const struct drm_i915_private *i915,
 #define IS_ICL_REVID(p, since, until) \
 	(IS_ICELAKE(p) && IS_REVID(p, since, until))
 
+#define TGL_REVID_A0		0x0
+
+#define IS_TGL_REVID(p, since, until) \
+	(IS_TIGERLAKE(p) && IS_REVID(p, since, until))
+
 #define IS_LP(dev_priv)	(INTEL_INFO(dev_priv)->is_lp)
 #define IS_GEN9_LP(dev_priv)	(IS_GEN(dev_priv, 9) && IS_LP(dev_priv))
 #define IS_GEN9_BC(dev_priv)	(IS_GEN(dev_priv, 9) && !IS_LP(dev_priv))
@@ -2073,9 +1614,16 @@ IS_SUBPLATFORM(const struct drm_i915_private *i915,
 #define VEBOX_MASK(dev_priv) \
 	ENGINE_INSTANCES_MASK(dev_priv, VECS0, I915_MAX_VECS)
 
+/*
+ * The Gen7 cmdparser copies the scanned buffer to the ggtt for execution
+ * All later gens can run the final buffer from the ppgtt
+ */
+#define CMDPARSER_USES_GGTT(dev_priv) IS_GEN(dev_priv, 7)
+
 #define HAS_LLC(dev_priv)	(INTEL_INFO(dev_priv)->has_llc)
 #define HAS_SNOOP(dev_priv)	(INTEL_INFO(dev_priv)->has_snoop)
 #define HAS_EDRAM(dev_priv)	((dev_priv)->edram_size_mb)
+#define HAS_SECURE_BATCHES(dev_priv) (INTEL_GEN(dev_priv) < 6)
 #define HAS_WT(dev_priv)	((IS_HASWELL(dev_priv) || \
 				 IS_BROADWELL(dev_priv)) && HAS_EDRAM(dev_priv))
 
@@ -2108,10 +1656,12 @@ IS_SUBPLATFORM(const struct drm_i915_private *i915,
 /* Early gen2 have a totally busted CS tlb and require pinned batches. */
 #define HAS_BROKEN_CS_TLB(dev_priv)	(IS_I830(dev_priv) || IS_I845G(dev_priv))
 
+#define NEEDS_RC6_CTX_CORRUPTION_WA(dev_priv)	\
+	(IS_BROADWELL(dev_priv) || IS_GEN(dev_priv, 9))
+
 /* WaRsDisableCoarsePowerGating:skl,cnl */
 #define NEEDS_WaRsDisableCoarsePowerGating(dev_priv) \
-	(IS_CANNONLAKE(dev_priv) || \
-	 IS_SKL_GT3(dev_priv) || IS_SKL_GT4(dev_priv))
+	(IS_CANNONLAKE(dev_priv) || IS_GEN(dev_priv, 9))
 
 #define HAS_GMBUS_IRQ(dev_priv) (INTEL_GEN(dev_priv) >= 4)
 #define HAS_GMBUS_BURST_READ(dev_priv) (INTEL_GEN(dev_priv) >= 10 || \
@@ -2153,6 +1703,9 @@ IS_SUBPLATFORM(const struct drm_i915_private *i915,
 
 #define HAS_IPC(dev_priv)		 (INTEL_INFO(dev_priv)->display.has_ipc)
 
+#define HAS_REGION(i915, i) (INTEL_INFO(i915)->memory_regions & (i))
+#define HAS_LMEM(i915) HAS_REGION(i915, REGION_LMEM)
+
 #define HAS_GT_UC(dev_priv)	(INTEL_INFO(dev_priv)->has_gt_uc)
 
 /* Having GuC is not the same as using GuC */
@@ -2176,7 +1729,12 @@ IS_SUBPLATFORM(const struct drm_i915_private *i915,
 #define GT_FREQUENCY_MULTIPLIER 50
 #define GEN9_FREQ_SCALER 3
 
-#define HAS_DISPLAY(dev_priv) (INTEL_INFO(dev_priv)->num_pipes > 0)
+#define INTEL_NUM_PIPES(dev_priv) (hweight8(INTEL_INFO(dev_priv)->pipe_mask))
+
+#define HAS_DISPLAY(dev_priv) (INTEL_INFO(dev_priv)->pipe_mask != 0)
+
+/* Only valid when HAS_DISPLAY() is true */
+#define INTEL_DISPLAY_ENABLED(dev_priv) (WARN_ON(!HAS_DISPLAY(dev_priv)), !i915_modparams.disable_display)
 
 static inline bool intel_vtd_active(void)
 {
@@ -2209,7 +1767,9 @@ extern const struct dev_pm_ops i915_pm_ops;
 int i915_driver_probe(struct pci_dev *pdev, const struct pci_device_id *ent);
 void i915_driver_remove(struct drm_i915_private *i915);
 
-void intel_engine_init_hangcheck(struct intel_engine_cs *engine);
+int i915_resume_switcheroo(struct drm_i915_private *i915);
+int i915_suspend_switcheroo(struct drm_i915_private *i915, pm_message_t state);
+
 int vlv_force_gfx_clock(struct drm_i915_private *dev_priv, bool on);
 
 static inline bool intel_gvt_active(struct drm_i915_private *dev_priv)
@@ -2228,12 +1788,13 @@ int i915_getparam_ioctl(struct drm_device *dev, void *data,
 /* i915_gem.c */
 int i915_gem_init_userptr(struct drm_i915_private *dev_priv);
 void i915_gem_cleanup_userptr(struct drm_i915_private *dev_priv);
-void i915_gem_sanitize(struct drm_i915_private *i915);
-int i915_gem_init_early(struct drm_i915_private *dev_priv);
+void i915_gem_init_early(struct drm_i915_private *dev_priv);
 void i915_gem_cleanup_early(struct drm_i915_private *dev_priv);
 int i915_gem_freeze(struct drm_i915_private *dev_priv);
 int i915_gem_freeze_late(struct drm_i915_private *dev_priv);
 
+struct intel_memory_region *i915_gem_shmem_setup(struct drm_i915_private *i915);
+
 static inline void i915_gem_drain_freed_objects(struct drm_i915_private *i915)
 {
 	/*
@@ -2282,6 +1843,14 @@ int i915_gem_object_unbind(struct drm_i915_gem_object *obj,
 			   unsigned long flags);
 #define I915_GEM_OBJECT_UNBIND_ACTIVE BIT(0)
 
+struct i915_vma * __must_check
+i915_gem_object_pin(struct drm_i915_gem_object *obj,
+		    struct i915_address_space *vm,
+		    const struct i915_ggtt_view *view,
+		    u64 size,
+		    u64 alignment,
+		    u64 flags);
+
 void i915_gem_runtime_suspend(struct drm_i915_private *dev_priv);
 
 static inline int __must_check
@@ -2310,15 +1879,11 @@ static inline u32 i915_reset_engine_count(struct i915_gpu_error *error,
 	return atomic_read(&error->reset_engine_count[engine->uabi_class]);
 }
 
-void i915_gem_init_mmio(struct drm_i915_private *i915);
 int __must_check i915_gem_init(struct drm_i915_private *dev_priv);
-int __must_check i915_gem_init_hw(struct drm_i915_private *dev_priv);
 void i915_gem_driver_register(struct drm_i915_private *i915);
 void i915_gem_driver_unregister(struct drm_i915_private *i915);
 void i915_gem_driver_remove(struct drm_i915_private *dev_priv);
 void i915_gem_driver_release(struct drm_i915_private *dev_priv);
-int i915_gem_wait_for_idle(struct drm_i915_private *dev_priv,
-			   unsigned int flags, long timeout);
 void i915_gem_suspend(struct drm_i915_private *dev_priv);
 void i915_gem_suspend_late(struct drm_i915_private *dev_priv);
 void i915_gem_resume(struct drm_i915_private *dev_priv);
@@ -2358,7 +1923,7 @@ i915_gem_context_lookup(struct drm_i915_file_private *file_priv, u32 id)
 /* i915_gem_evict.c */
 int __must_check i915_gem_evict_something(struct i915_address_space *vm,
 					  u64 min_size, u64 alignment,
-					  unsigned cache_level,
+					  unsigned long color,
 					  u64 start, u64 end,
 					  unsigned flags);
 int __must_check i915_gem_evict_for_node(struct i915_address_space *vm,
@@ -2374,9 +1939,9 @@ i915_gem_object_create_internal(struct drm_i915_private *dev_priv,
 /* i915_gem_tiling.c */
 static inline bool i915_gem_object_needs_bit17_swizzle(struct drm_i915_gem_object *obj)
 {
-	struct drm_i915_private *dev_priv = to_i915(obj->base.dev);
+	struct drm_i915_private *i915 = to_i915(obj->base.dev);
 
-	return dev_priv->mm.bit_6_swizzle_x == I915_BIT_6_SWIZZLE_9_10_17 &&
+	return i915->ggtt.bit_6_swizzle_x == I915_BIT_6_SWIZZLE_9_10_17 &&
 		i915_gem_object_is_tiled(obj);
 }
 
@@ -2391,12 +1956,14 @@ const char *i915_cache_level_str(struct drm_i915_private *i915, int type);
 int i915_cmd_parser_get_version(struct drm_i915_private *dev_priv);
 void intel_engine_init_cmd_parser(struct intel_engine_cs *engine);
 void intel_engine_cleanup_cmd_parser(struct intel_engine_cs *engine);
-int intel_engine_cmd_parser(struct intel_engine_cs *engine,
+int intel_engine_cmd_parser(struct i915_gem_context *cxt,
+			    struct intel_engine_cs *engine,
 			    struct drm_i915_gem_object *batch_obj,
-			    struct drm_i915_gem_object *shadow_batch_obj,
+			    u64 user_batch_start,
 			    u32 batch_start_offset,
 			    u32 batch_len,
-			    bool is_master);
+			    struct drm_i915_gem_object *shadow_batch_obj,
+			    u64 shadow_batch_start);
 
 /* intel_device_info.c */
 static inline struct intel_device_info *
@@ -2478,4 +2045,10 @@ i915_coherent_map_type(struct drm_i915_private *i915)
 	return HAS_LLC(i915) ? I915_MAP_WB : I915_MAP_WC;
 }
 
+static inline bool intel_guc_submission_is_enabled(struct intel_guc *guc)
+{
+	return intel_guc_is_submission_supported(guc) &&
+		intel_guc_is_running(guc);
+}
+
 #endif
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 95e7c52cf8ed..b9eb6b3149b7 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -45,13 +45,14 @@
 #include "gem/i915_gem_context.h"
 #include "gem/i915_gem_ioctls.h"
 #include "gem/i915_gem_pm.h"
-#include "gem/i915_gemfs.h"
 #include "gt/intel_engine_user.h"
 #include "gt/intel_gt.h"
 #include "gt/intel_gt_pm.h"
+#include "gt/intel_gt_requests.h"
 #include "gt/intel_mocs.h"
 #include "gt/intel_reset.h"
 #include "gt/intel_renderstate.h"
+#include "gt/intel_rps.h"
 #include "gt/intel_workarounds.h"
 
 #include "i915_drv.h"
@@ -62,20 +63,31 @@
 #include "intel_pm.h"
 
 static int
-insert_mappable_node(struct i915_ggtt *ggtt,
-                     struct drm_mm_node *node, u32 size)
+insert_mappable_node(struct i915_ggtt *ggtt, struct drm_mm_node *node, u32 size)
 {
+	int err;
+
+	err = mutex_lock_interruptible(&ggtt->vm.mutex);
+	if (err)
+		return err;
+
 	memset(node, 0, sizeof(*node));
-	return drm_mm_insert_node_in_range(&ggtt->vm.mm, node,
-					   size, 0, I915_COLOR_UNEVICTABLE,
-					   0, ggtt->mappable_end,
-					   DRM_MM_INSERT_LOW);
+	err = drm_mm_insert_node_in_range(&ggtt->vm.mm, node,
+					  size, 0, I915_COLOR_UNEVICTABLE,
+					  0, ggtt->mappable_end,
+					  DRM_MM_INSERT_LOW);
+
+	mutex_unlock(&ggtt->vm.mutex);
+
+	return err;
 }
 
 static void
-remove_mappable_node(struct drm_mm_node *node)
+remove_mappable_node(struct i915_ggtt *ggtt, struct drm_mm_node *node)
 {
+	mutex_lock(&ggtt->vm.mutex);
 	drm_mm_remove_node(node);
+	mutex_unlock(&ggtt->vm.mutex);
 }
 
 int
@@ -87,7 +99,8 @@ i915_gem_get_aperture_ioctl(struct drm_device *dev, void *data,
 	struct i915_vma *vma;
 	u64 pinned;
 
-	mutex_lock(&ggtt->vm.mutex);
+	if (mutex_lock_interruptible(&ggtt->vm.mutex))
+		return -EINTR;
 
 	pinned = ggtt->vm.reserved;
 	list_for_each_entry(vma, &ggtt->vm.bound_list, vm_link)
@@ -109,20 +122,24 @@ int i915_gem_object_unbind(struct drm_i915_gem_object *obj,
 	LIST_HEAD(still_in_list);
 	int ret = 0;
 
-	lockdep_assert_held(&obj->base.dev->struct_mutex);
-
 	spin_lock(&obj->vma.lock);
 	while (!ret && (vma = list_first_entry_or_null(&obj->vma.list,
 						       struct i915_vma,
 						       obj_link))) {
+		struct i915_address_space *vm = vma->vm;
+
+		ret = -EBUSY;
+		if (!i915_vm_tryopen(vm))
+			break;
+
 		list_move_tail(&vma->obj_link, &still_in_list);
 		spin_unlock(&obj->vma.lock);
 
-		ret = -EBUSY;
 		if (flags & I915_GEM_OBJECT_UNBIND_ACTIVE ||
 		    !i915_vma_is_active(vma))
 			ret = i915_vma_unbind(vma);
 
+		i915_vm_close(vm);
 		spin_lock(&obj->vma.lock);
 	}
 	list_splice(&still_in_list, &obj->vma.list);
@@ -338,10 +355,6 @@ i915_gem_gtt_pread(struct drm_i915_gem_object *obj,
 	u64 remain, offset;
 	int ret;
 
-	ret = mutex_lock_interruptible(&i915->drm.struct_mutex);
-	if (ret)
-		return ret;
-
 	wakeref = intel_runtime_pm_get(&i915->runtime_pm);
 	vma = ERR_PTR(-ENODEV);
 	if (!i915_gem_object_is_tiled(obj))
@@ -351,16 +364,14 @@ i915_gem_gtt_pread(struct drm_i915_gem_object *obj,
 					       PIN_NOEVICT);
 	if (!IS_ERR(vma)) {
 		node.start = i915_ggtt_offset(vma);
-		node.allocated = false;
+		node.flags = 0;
 	} else {
 		ret = insert_mappable_node(ggtt, &node, PAGE_SIZE);
 		if (ret)
-			goto out_unlock;
-		GEM_BUG_ON(!node.allocated);
+			goto out_rpm;
+		GEM_BUG_ON(!drm_mm_node_allocated(&node));
 	}
 
-	mutex_unlock(&i915->drm.struct_mutex);
-
 	ret = i915_gem_object_lock_interruptible(obj);
 	if (ret)
 		goto out_unpin;
@@ -393,7 +404,7 @@ i915_gem_gtt_pread(struct drm_i915_gem_object *obj,
 		unsigned page_offset = offset_in_page(offset);
 		unsigned page_length = PAGE_SIZE - page_offset;
 		page_length = remain < page_length ? remain : page_length;
-		if (node.allocated) {
+		if (drm_mm_node_allocated(&node)) {
 			ggtt->vm.insert_page(&ggtt->vm,
 					     i915_gem_object_get_dma_address(obj, offset >> PAGE_SHIFT),
 					     node.start, I915_CACHE_NONE, 0);
@@ -414,17 +425,14 @@ i915_gem_gtt_pread(struct drm_i915_gem_object *obj,
 
 	i915_gem_object_unlock_fence(obj, fence);
 out_unpin:
-	mutex_lock(&i915->drm.struct_mutex);
-	if (node.allocated) {
+	if (drm_mm_node_allocated(&node)) {
 		ggtt->vm.clear_range(&ggtt->vm, node.start, node.size);
-		remove_mappable_node(&node);
+		remove_mappable_node(ggtt, &node);
 	} else {
 		i915_vma_unpin(vma);
 	}
-out_unlock:
+out_rpm:
 	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
-	mutex_unlock(&i915->drm.struct_mutex);
-
 	return ret;
 }
 
@@ -531,10 +539,6 @@ i915_gem_gtt_pwrite_fast(struct drm_i915_gem_object *obj,
 	void __user *user_data;
 	int ret;
 
-	ret = mutex_lock_interruptible(&i915->drm.struct_mutex);
-	if (ret)
-		return ret;
-
 	if (i915_gem_object_has_struct_page(obj)) {
 		/*
 		 * Avoid waking the device up if we can fallback, as
@@ -544,10 +548,8 @@ i915_gem_gtt_pwrite_fast(struct drm_i915_gem_object *obj,
 		 * using the cache bypass of indirect GGTT access.
 		 */
 		wakeref = intel_runtime_pm_get_if_in_use(rpm);
-		if (!wakeref) {
-			ret = -EFAULT;
-			goto out_unlock;
-		}
+		if (!wakeref)
+			return -EFAULT;
 	} else {
 		/* No backing pages, no fallback, we must force GGTT access */
 		wakeref = intel_runtime_pm_get(rpm);
@@ -561,16 +563,14 @@ i915_gem_gtt_pwrite_fast(struct drm_i915_gem_object *obj,
 					       PIN_NOEVICT);
 	if (!IS_ERR(vma)) {
 		node.start = i915_ggtt_offset(vma);
-		node.allocated = false;
+		node.flags = 0;
 	} else {
 		ret = insert_mappable_node(ggtt, &node, PAGE_SIZE);
 		if (ret)
 			goto out_rpm;
-		GEM_BUG_ON(!node.allocated);
+		GEM_BUG_ON(!drm_mm_node_allocated(&node));
 	}
 
-	mutex_unlock(&i915->drm.struct_mutex);
-
 	ret = i915_gem_object_lock_interruptible(obj);
 	if (ret)
 		goto out_unpin;
@@ -604,7 +604,7 @@ i915_gem_gtt_pwrite_fast(struct drm_i915_gem_object *obj,
 		unsigned int page_offset = offset_in_page(offset);
 		unsigned int page_length = PAGE_SIZE - page_offset;
 		page_length = remain < page_length ? remain : page_length;
-		if (node.allocated) {
+		if (drm_mm_node_allocated(&node)) {
 			/* flush the write before we modify the GGTT */
 			intel_gt_flush_ggtt_writes(ggtt->vm.gt);
 			ggtt->vm.insert_page(&ggtt->vm,
@@ -634,18 +634,15 @@ i915_gem_gtt_pwrite_fast(struct drm_i915_gem_object *obj,
 
 	i915_gem_object_unlock_fence(obj, fence);
 out_unpin:
-	mutex_lock(&i915->drm.struct_mutex);
 	intel_gt_flush_ggtt_writes(ggtt->vm.gt);
-	if (node.allocated) {
+	if (drm_mm_node_allocated(&node)) {
 		ggtt->vm.clear_range(&ggtt->vm, node.start, node.size);
-		remove_mappable_node(&node);
+		remove_mappable_node(ggtt, &node);
 	} else {
 		i915_vma_unpin(vma);
 	}
 out_rpm:
 	intel_runtime_pm_put(rpm, wakeref);
-out_unlock:
-	mutex_unlock(&i915->drm.struct_mutex);
 	return ret;
 }
 
@@ -887,74 +884,6 @@ void i915_gem_runtime_suspend(struct drm_i915_private *i915)
 	}
 }
 
-static long
-wait_for_timelines(struct drm_i915_private *i915,
-		   unsigned int wait, long timeout)
-{
-	struct intel_gt_timelines *timelines = &i915->gt.timelines;
-	struct intel_timeline *tl;
-	unsigned long flags;
-
-	spin_lock_irqsave(&timelines->lock, flags);
-	list_for_each_entry(tl, &timelines->active_list, link) {
-		struct i915_request *rq;
-
-		rq = i915_active_request_get_unlocked(&tl->last_request);
-		if (!rq)
-			continue;
-
-		spin_unlock_irqrestore(&timelines->lock, flags);
-
-		/*
-		 * "Race-to-idle".
-		 *
-		 * Switching to the kernel context is often used a synchronous
-		 * step prior to idling, e.g. in suspend for flushing all
-		 * current operations to memory before sleeping. These we
-		 * want to complete as quickly as possible to avoid prolonged
-		 * stalls, so allow the gpu to boost to maximum clocks.
-		 */
-		if (wait & I915_WAIT_FOR_IDLE_BOOST)
-			gen6_rps_boost(rq);
-
-		timeout = i915_request_wait(rq, wait, timeout);
-		i915_request_put(rq);
-		if (timeout < 0)
-			return timeout;
-
-		/* restart after reacquiring the lock */
-		spin_lock_irqsave(&timelines->lock, flags);
-		tl = list_entry(&timelines->active_list, typeof(*tl), link);
-	}
-	spin_unlock_irqrestore(&timelines->lock, flags);
-
-	return timeout;
-}
-
-int i915_gem_wait_for_idle(struct drm_i915_private *i915,
-			   unsigned int flags, long timeout)
-{
-	/* If the device is asleep, we have no requests outstanding */
-	if (!intel_gt_pm_is_awake(&i915->gt))
-		return 0;
-
-	GEM_TRACE("flags=%x (%s), timeout=%ld%s\n",
-		  flags, flags & I915_WAIT_LOCKED ? "locked" : "unlocked",
-		  timeout, timeout == MAX_SCHEDULE_TIMEOUT ? " (forever)" : "");
-
-	timeout = wait_for_timelines(i915, flags, timeout);
-	if (timeout < 0)
-		return timeout;
-
-	if (flags & I915_WAIT_LOCKED) {
-		lockdep_assert_held(&i915->drm.struct_mutex);
-
-		i915_retire_requests(i915);
-	}
-
-	return 0;
-}
-
 struct i915_vma *
 i915_gem_object_ggtt_pin(struct drm_i915_gem_object *obj,
 			 const struct i915_ggtt_view *view,
@@ -964,10 +893,25 @@ i915_gem_object_ggtt_pin(struct drm_i915_gem_object *obj,
 {
 	struct drm_i915_private *dev_priv = to_i915(obj->base.dev);
 	struct i915_address_space *vm = &dev_priv->ggtt.vm;
+
+	return i915_gem_object_pin(obj, vm, view, size, alignment,
+				   flags | PIN_GLOBAL);
+}
+
+struct i915_vma *
+i915_gem_object_pin(struct drm_i915_gem_object *obj,
+		    struct i915_address_space *vm,
+		    const struct i915_ggtt_view *view,
+		    u64 size,
+		    u64 alignment,
+		    u64 flags)
+{
+	struct drm_i915_private *dev_priv = to_i915(obj->base.dev);
 	struct i915_vma *vma;
 	int ret;
 
-	lockdep_assert_held(&obj->base.dev->struct_mutex);
+	if (i915_gem_object_never_bind_ggtt(obj))
+		return ERR_PTR(-ENODEV);
 
 	if (flags & PIN_MAPPABLE &&
 	    (!view || view->type == I915_GGTT_VIEW_NORMAL)) {
@@ -1015,13 +959,6 @@ i915_gem_object_ggtt_pin(struct drm_i915_gem_object *obj,
 				return ERR_PTR(-ENOSPC);
 		}
 
-		WARN(i915_vma_is_pinned(vma),
-		     "bo is already pinned in ggtt with incorrect alignment:"
-		     " offset=%08x, req.alignment=%llx,"
-		     " req.map_and_fenceable=%d, vma->map_and_fenceable=%d\n",
-		     i915_ggtt_offset(vma), alignment,
-		     !!(flags & PIN_MAPPABLE),
-		     i915_vma_is_map_and_fenceable(vma));
 		ret = i915_vma_unbind(vma);
 		if (ret)
 			return ERR_PTR(ret);
@@ -1035,7 +972,7 @@ i915_gem_object_ggtt_pin(struct drm_i915_gem_object *obj,
 			return ERR_PTR(ret);
 	}
 
-	ret = i915_vma_pin(vma, size, alignment, flags | PIN_GLOBAL);
+	ret = i915_vma_pin(vma, size, alignment, flags);
 	if (ret)
 		return ERR_PTR(ret);
 
@@ -1116,128 +1053,7 @@ out:
 	return err;
 }
 
-void i915_gem_sanitize(struct drm_i915_private *i915)
-{
-	intel_wakeref_t wakeref;
-
-	GEM_TRACE("\n");
-
-	wakeref = intel_runtime_pm_get(&i915->runtime_pm);
-	intel_uncore_forcewake_get(&i915->uncore, FORCEWAKE_ALL);
-
-	/*
-	 * As we have just resumed the machine and woken the device up from
-	 * deep PCI sleep (presumably D3_cold), assume the HW has been reset
-	 * back to defaults, recovering from whatever wedged state we left it
-	 * in and so worth trying to use the device once more.
-	 */
-	if (intel_gt_is_wedged(&i915->gt))
-		intel_gt_unset_wedged(&i915->gt);
-
-	/*
-	 * If we inherit context state from the BIOS or earlier occupants
-	 * of the GPU, the GPU may be in an inconsistent state when we
-	 * try to take over. The only way to remove the earlier state
-	 * is by resetting. However, resetting on earlier gen is tricky as
-	 * it may impact the display and we are uncertain about the stability
-	 * of the reset, so this could be applied to even earlier gen.
-	 */
-	intel_gt_sanitize(&i915->gt, false);
-
-	intel_uncore_forcewake_put(&i915->uncore, FORCEWAKE_ALL);
-	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
-}
-
-static void init_unused_ring(struct intel_gt *gt, u32 base)
-{
-	struct intel_uncore *uncore = gt->uncore;
-
-	intel_uncore_write(uncore, RING_CTL(base), 0);
-	intel_uncore_write(uncore, RING_HEAD(base), 0);
-	intel_uncore_write(uncore, RING_TAIL(base), 0);
-	intel_uncore_write(uncore, RING_START(base), 0);
-}
-
-static void init_unused_rings(struct intel_gt *gt)
-{
-	struct drm_i915_private *i915 = gt->i915;
-
-	if (IS_I830(i915)) {
-		init_unused_ring(gt, PRB1_BASE);
-		init_unused_ring(gt, SRB0_BASE);
-		init_unused_ring(gt, SRB1_BASE);
-		init_unused_ring(gt, SRB2_BASE);
-		init_unused_ring(gt, SRB3_BASE);
-	} else if (IS_GEN(i915, 2)) {
-		init_unused_ring(gt, SRB0_BASE);
-		init_unused_ring(gt, SRB1_BASE);
-	} else if (IS_GEN(i915, 3)) {
-		init_unused_ring(gt, PRB1_BASE);
-		init_unused_ring(gt, PRB2_BASE);
-	}
-}
-
-int i915_gem_init_hw(struct drm_i915_private *i915)
-{
-	struct intel_uncore *uncore = &i915->uncore;
-	struct intel_gt *gt = &i915->gt;
-	int ret;
-
-	BUG_ON(!i915->kernel_context);
-	ret = intel_gt_terminally_wedged(gt);
-	if (ret)
-		return ret;
-
-	gt->last_init_time = ktime_get();
-
-	/* Double layer security blanket, see i915_gem_init() */
-	intel_uncore_forcewake_get(uncore, FORCEWAKE_ALL);
-
-	if (HAS_EDRAM(i915) && INTEL_GEN(i915) < 9)
-		intel_uncore_rmw(uncore, HSW_IDICR, 0, IDIHASHMSK(0xf));
-
-	if (IS_HASWELL(i915))
-		intel_uncore_write(uncore,
-				   MI_PREDICATE_RESULT_2,
-				   IS_HSW_GT3(i915) ?
-				   LOWER_SLICE_ENABLED : LOWER_SLICE_DISABLED);
-
-	/* Apply the GT workarounds... */
-	intel_gt_apply_workarounds(gt);
-	/* ...and determine whether they are sticking. */
-	intel_gt_verify_workarounds(gt, "init");
-
-	intel_gt_init_swizzling(gt);
-
-	/*
-	 * At least 830 can leave some of the unused rings
-	 * "active" (ie. head != tail) after resume which
-	 * will prevent c3 entry. Makes sure all unused rings
-	 * are totally idle.
-	 */
-	init_unused_rings(gt);
-
-	ret = i915_ppgtt_init_hw(gt);
-	if (ret) {
-		DRM_ERROR("Enabling PPGTT failed (%d)\n", ret);
-		goto out;
-	}
-
-	/* We can't enable contexts until all firmware is loaded */
-	ret = intel_uc_init_hw(&gt->uc);
-	if (ret) {
-		i915_probe_error(i915, "Enabling uc failed (%d)\n", ret);
-		goto out;
-	}
-
-	intel_mocs_init(gt);
-
-out:
-	intel_uncore_forcewake_put(uncore, FORCEWAKE_ALL);
-	return ret;
-}
-
-static int __intel_engines_record_defaults(struct drm_i915_private *i915)
+static int __intel_engines_record_defaults(struct intel_gt *gt)
 {
 	struct i915_request *requests[I915_NUM_ENGINES] = {};
 	struct intel_engine_cs *engine;
@@ -1253,7 +1069,7 @@ static int __intel_engines_record_defaults(struct drm_i915_private *i915)
 	 * from the same default HW values.
 	 */
 
-	for_each_engine(engine, i915, id) {
+	for_each_engine(engine, gt, id) {
 		struct intel_context *ce;
 		struct i915_request *rq;
 
@@ -1261,7 +1077,8 @@ static int __intel_engines_record_defaults(struct drm_i915_private *i915)
 		GEM_BUG_ON(!engine->kernel_context);
 		engine->serial++; /* force the kernel context switch */
 
-		ce = intel_context_create(i915->kernel_context, engine);
+		ce = intel_context_create(engine->kernel_context->gem_context,
+					  engine);
 		if (IS_ERR(ce)) {
 			err = PTR_ERR(ce);
 			goto out;
@@ -1278,15 +1095,6 @@ static int __intel_engines_record_defaults(struct drm_i915_private *i915)
 		if (err)
 			goto err_rq;
 
-		/*
-		 * Failing to program the MOCS is non-fatal.The system will not
-		 * run at peak performance. So warn the user and carry on.
-		 */
-		err = intel_mocs_emit(rq);
-		if (err)
-			dev_notice(i915->drm.dev,
-				   "Failed to program MOCS registers; expect performance issues.\n");
-
 		err = intel_renderstate_emit(rq);
 		if (err)
 			goto err_rq;
@@ -1299,7 +1107,7 @@ err_rq:
 	}
 
 	/* Flush the default context image to memory, and enable powersaving. */
-	if (!i915_gem_load_power_context(i915)) {
+	if (intel_gt_wait_for_idle(gt, I915_GEM_IDLE_TIMEOUT) == -ETIME) {
 		err = -EIO;
 		goto out;
 	}
@@ -1358,7 +1166,7 @@ out:
 	 * this is by declaring ourselves wedged.
 	 */
 	if (err)
-		intel_gt_set_wedged(&i915->gt);
+		intel_gt_set_wedged(gt);
 
 	for (id = 0; id < ARRAY_SIZE(requests); id++) {
 		struct intel_context *ce;
@@ -1375,18 +1183,7 @@ out:
 	return err;
 }
 
-static int
-i915_gem_init_scratch(struct drm_i915_private *i915, unsigned int size)
-{
-	return intel_gt_init_scratch(&i915->gt, size);
-}
-
-static void i915_gem_fini_scratch(struct drm_i915_private *i915)
-{
-	intel_gt_fini_scratch(&i915->gt);
-}
-
-static int intel_engines_verify_workarounds(struct drm_i915_private *i915)
+static int intel_engines_verify_workarounds(struct intel_gt *gt)
 {
 	struct intel_engine_cs *engine;
 	enum intel_engine_id id;
@@ -1395,7 +1192,7 @@ static int intel_engines_verify_workarounds(struct drm_i915_private *i915)
 	if (!IS_ENABLED(CONFIG_DRM_I915_DEBUG_GEM))
 		return 0;
 
-	for_each_engine(engine, i915, id) {
+	for_each_engine(engine, gt, id) {
 		if (intel_engine_verify_workarounds(engine, "load"))
 			err = -EIO;
 	}
@@ -1427,7 +1224,6 @@ int i915_gem_init(struct drm_i915_private *dev_priv)
 	 * we hold the forcewake during initialisation these problems
 	 * just magically go away.
 	 */
-	mutex_lock(&dev_priv->drm.struct_mutex);
 	intel_uncore_forcewake_get(&dev_priv->uncore, FORCEWAKE_ALL);
 
 	ret = i915_init_ggtt(dev_priv);
@@ -1436,36 +1232,29 @@ int i915_gem_init(struct drm_i915_private *dev_priv)
 		goto err_unlock;
 	}
 
-	ret = i915_gem_init_scratch(dev_priv,
-				    IS_GEN(dev_priv, 2) ? SZ_256K : PAGE_SIZE);
-	if (ret) {
-		GEM_BUG_ON(ret == -EIO);
-		goto err_ggtt;
-	}
+	intel_gt_init(&dev_priv->gt);
 
-	ret = intel_engines_setup(dev_priv);
+	ret = intel_engines_setup(&dev_priv->gt);
 	if (ret) {
 		GEM_BUG_ON(ret == -EIO);
 		goto err_unlock;
 	}
 
-	ret = i915_gem_contexts_init(dev_priv);
+	ret = i915_gem_init_contexts(dev_priv);
 	if (ret) {
 		GEM_BUG_ON(ret == -EIO);
 		goto err_scratch;
 	}
 
-	ret = intel_engines_init(dev_priv);
+	ret = intel_engines_init(&dev_priv->gt);
 	if (ret) {
 		GEM_BUG_ON(ret == -EIO);
 		goto err_context;
 	}
 
-	intel_init_gt_powersave(dev_priv);
-
 	intel_uc_init(&dev_priv->gt.uc);
 
-	ret = i915_gem_init_hw(dev_priv);
+	ret = intel_gt_init_hw(&dev_priv->gt);
 	if (ret)
 		goto err_uc_init;
 
@@ -1485,24 +1274,23 @@ int i915_gem_init(struct drm_i915_private *dev_priv)
 	 */
 	intel_init_clock_gating(dev_priv);
 
-	ret = intel_engines_verify_workarounds(dev_priv);
+	ret = intel_engines_verify_workarounds(&dev_priv->gt);
 	if (ret)
 		goto err_gt;
 
-	ret = __intel_engines_record_defaults(dev_priv);
+	ret = __intel_engines_record_defaults(&dev_priv->gt);
 	if (ret)
 		goto err_gt;
 
-	ret = i915_inject_load_error(dev_priv, -ENODEV);
+	ret = i915_inject_probe_error(dev_priv, -ENODEV);
 	if (ret)
 		goto err_gt;
 
-	ret = i915_inject_load_error(dev_priv, -EIO);
+	ret = i915_inject_probe_error(dev_priv, -EIO);
 	if (ret)
 		goto err_gt;
 
 	intel_uncore_forcewake_put(&dev_priv->uncore, FORCEWAKE_ALL);
-	mutex_unlock(&dev_priv->drm.struct_mutex);
 
 	return 0;
 
@@ -1513,32 +1301,25 @@ int i915_gem_init(struct drm_i915_private *dev_priv)
 	 * driver doesn't explode during runtime.
 	 */
 err_gt:
-	mutex_unlock(&dev_priv->drm.struct_mutex);
-
-	intel_gt_set_wedged(&dev_priv->gt);
+	intel_gt_set_wedged_on_init(&dev_priv->gt);
 	i915_gem_suspend(dev_priv);
 	i915_gem_suspend_late(dev_priv);
 
 	i915_gem_drain_workqueue(dev_priv);
-
-	mutex_lock(&dev_priv->drm.struct_mutex);
 err_init_hw:
 	intel_uc_fini_hw(&dev_priv->gt.uc);
 err_uc_init:
 	if (ret != -EIO) {
 		intel_uc_fini(&dev_priv->gt.uc);
-		intel_cleanup_gt_powersave(dev_priv);
-		intel_engines_cleanup(dev_priv);
+		intel_engines_cleanup(&dev_priv->gt);
 	}
 err_context:
 	if (ret != -EIO)
-		i915_gem_contexts_fini(dev_priv);
+		i915_gem_driver_release__contexts(dev_priv);
 err_scratch:
-	i915_gem_fini_scratch(dev_priv);
-err_ggtt:
+	intel_gt_driver_release(&dev_priv->gt);
 err_unlock:
 	intel_uncore_forcewake_put(&dev_priv->uncore, FORCEWAKE_ALL);
-	mutex_unlock(&dev_priv->drm.struct_mutex);
 
 	if (ret != -EIO) {
 		intel_uc_cleanup_firmwares(&dev_priv->gt.uc);
@@ -1547,8 +1328,6 @@ err_unlock:
 	}
 
 	if (ret == -EIO) {
-		mutex_lock(&dev_priv->drm.struct_mutex);
-
 		/*
 		 * Allow engines or uC initialisation to fail by marking the GPU
 		 * as wedged. But we only want to do this when the GPU is angry,
@@ -1563,10 +1342,8 @@ err_unlock:
 		/* Minimal basic recovery for KMS */
 		ret = i915_ggtt_enable_hw(dev_priv);
 		i915_gem_restore_gtt_mappings(dev_priv);
-		i915_gem_restore_fences(dev_priv);
+		i915_gem_restore_fences(&dev_priv->ggtt);
 		intel_init_clock_gating(dev_priv);
-
-		mutex_unlock(&dev_priv->drm.struct_mutex);
 	}
 
 	i915_gem_drain_freed_objects(dev_priv);
@@ -1587,48 +1364,35 @@ void i915_gem_driver_unregister(struct drm_i915_private *i915)
 
 void i915_gem_driver_remove(struct drm_i915_private *dev_priv)
 {
-	GEM_BUG_ON(dev_priv->gt.awake);
-
 	intel_wakeref_auto_fini(&dev_priv->ggtt.userfault_wakeref);
 
 	i915_gem_suspend_late(dev_priv);
-	intel_disable_gt_powersave(dev_priv);
+	intel_gt_driver_remove(&dev_priv->gt);
 
 	/* Flush any outstanding unpin_work. */
 	i915_gem_drain_workqueue(dev_priv);
 
-	mutex_lock(&dev_priv->drm.struct_mutex);
 	intel_uc_fini_hw(&dev_priv->gt.uc);
 	intel_uc_fini(&dev_priv->gt.uc);
-	mutex_unlock(&dev_priv->drm.struct_mutex);
 
 	i915_gem_drain_freed_objects(dev_priv);
 }
 
 void i915_gem_driver_release(struct drm_i915_private *dev_priv)
 {
-	mutex_lock(&dev_priv->drm.struct_mutex);
-	intel_engines_cleanup(dev_priv);
-	i915_gem_contexts_fini(dev_priv);
-	i915_gem_fini_scratch(dev_priv);
-	mutex_unlock(&dev_priv->drm.struct_mutex);
+	intel_engines_cleanup(&dev_priv->gt);
+	i915_gem_driver_release__contexts(dev_priv);
+	intel_gt_driver_release(&dev_priv->gt);
 
 	intel_wa_list_free(&dev_priv->gt_wa_list);
 
-	intel_cleanup_gt_powersave(dev_priv);
-
 	intel_uc_cleanup_firmwares(&dev_priv->gt.uc);
 	i915_gem_cleanup_userptr(dev_priv);
 	intel_timelines_fini(dev_priv);
 
 	i915_gem_drain_freed_objects(dev_priv);
 
-	WARN_ON(!list_empty(&dev_priv->contexts.list));
-}
-
-void i915_gem_init_mmio(struct drm_i915_private *i915)
-{
-	i915_gem_sanitize(i915);
+	WARN_ON(!list_empty(&dev_priv->gem.contexts.list));
 }
 
 static void i915_gem_init__mm(struct drm_i915_private *i915)
@@ -1643,20 +1407,11 @@ static void i915_gem_init__mm(struct drm_i915_private *i915)
 	i915_gem_init__objects(i915);
 }
 
-int i915_gem_init_early(struct drm_i915_private *dev_priv)
+void i915_gem_init_early(struct drm_i915_private *dev_priv)
 {
-	int err;
-
 	i915_gem_init__mm(dev_priv);
-	i915_gem_init__pm(dev_priv);
 
 	spin_lock_init(&dev_priv->fb_tracking.lock);
-
-	err = i915_gemfs_init(dev_priv);
-	if (err)
-		DRM_NOTE("Unable to create a private tmpfs mount, hugepage support will be disabled(%d).\n", err);
-
-	return 0;
 }
 
 void i915_gem_cleanup_early(struct drm_i915_private *dev_priv)
@@ -1665,8 +1420,6 @@ void i915_gem_cleanup_early(struct drm_i915_private *dev_priv)
 	GEM_BUG_ON(!llist_empty(&dev_priv->mm.free_list));
 	GEM_BUG_ON(atomic_read(&dev_priv->mm.free_count));
 	WARN_ON(dev_priv->mm.shrink_count);
-
-	i915_gemfs_fini(dev_priv);
 }
 
 int i915_gem_freeze(struct drm_i915_private *dev_priv)
diff --git a/drivers/gpu/drm/i915/i915_gem.h b/drivers/gpu/drm/i915/i915_gem.h
index 167a7b56ed5b..f6f9675848b8 100644
--- a/drivers/gpu/drm/i915/i915_gem.h
+++ b/drivers/gpu/drm/i915/i915_gem.h
@@ -37,10 +37,8 @@ struct drm_i915_private;
 #define GEM_SHOW_DEBUG() (drm_debug & DRM_UT_DRIVER)
 
 #define GEM_BUG_ON(condition) do { if (unlikely((condition))) {	\
-		pr_err("%s:%d GEM_BUG_ON(%s)\n", \
-		       __func__, __LINE__, __stringify(condition)); \
-		GEM_TRACE("%s:%d GEM_BUG_ON(%s)\n", \
-			  __func__, __LINE__, __stringify(condition)); \
+		GEM_TRACE_ERR("%s:%d GEM_BUG_ON(%s)\n", \
+			      __func__, __LINE__, __stringify(condition)); \
 		BUG(); \
 		} \
 	} while(0)
@@ -66,17 +64,33 @@ struct drm_i915_private;
 
 #if IS_ENABLED(CONFIG_DRM_I915_TRACE_GEM)
 #define GEM_TRACE(...) trace_printk(__VA_ARGS__)
+#define GEM_TRACE_ERR(...) do {						\
+	pr_err(__VA_ARGS__);						\
+	trace_printk(__VA_ARGS__);					\
+} while (0)
 #define GEM_TRACE_DUMP() ftrace_dump(DUMP_ALL)
 #define GEM_TRACE_DUMP_ON(expr) \
 	do { if (expr) ftrace_dump(DUMP_ALL); } while (0)
 #else
 #define GEM_TRACE(...) do { } while (0)
+#define GEM_TRACE_ERR(...) do { } while (0)
 #define GEM_TRACE_DUMP() do { } while (0)
 #define GEM_TRACE_DUMP_ON(expr) BUILD_BUG_ON_INVALID(expr)
 #endif
 
 #define I915_GEM_IDLE_TIMEOUT (HZ / 5)
 
+static inline void tasklet_lock(struct tasklet_struct *t)
+{
+	while (!tasklet_trylock(t))
+		cpu_relax();
+}
+
+static inline bool tasklet_is_locked(const struct tasklet_struct *t)
+{
+	return test_bit(TASKLET_STATE_RUN, &t->state);
+}
+
 static inline void __tasklet_disable_sync_once(struct tasklet_struct *t)
 {
 	if (!atomic_fetch_inc(&t->count))
diff --git a/drivers/gpu/drm/i915/i915_gem_evict.c b/drivers/gpu/drm/i915/i915_gem_evict.c
index 52c86c6e0673..7e62c310290f 100644
--- a/drivers/gpu/drm/i915/i915_gem_evict.c
+++ b/drivers/gpu/drm/i915/i915_gem_evict.c
@@ -29,6 +29,7 @@
 #include <drm/i915_drm.h>
 
 #include "gem/i915_gem_context.h"
+#include "gt/intel_gt_requests.h"
 
 #include "i915_drv.h"
 #include "i915_trace.h"
@@ -37,7 +38,7 @@ I915_SELFTEST_DECLARE(static struct igt_evict_ctl {
 	bool fail_if_busy:1;
 } igt_evict_ctl;)
 
-static int ggtt_flush(struct drm_i915_private *i915)
+static int ggtt_flush(struct intel_gt *gt)
 {
 	/*
 	 * Not everything in the GGTT is tracked via vma (otherwise we
@@ -46,10 +47,7 @@ static int ggtt_flush(struct drm_i915_private *i915)
 	 * the hopes that we can then remove contexts and the like only
 	 * bound by their active reference.
 	 */
-	return i915_gem_wait_for_idle(i915,
-				      I915_WAIT_INTERRUPTIBLE |
-				      I915_WAIT_LOCKED,
-				      MAX_SCHEDULE_TIMEOUT);
+	return intel_gt_wait_for_idle(gt, MAX_SCHEDULE_TIMEOUT);
 }
 
 static bool
@@ -70,7 +68,7 @@ mark_free(struct drm_mm_scan *scan,
  * @vm: address space to evict from
  * @min_size: size of the desired free space
  * @alignment: alignment constraint of the desired free space
- * @cache_level: cache_level for the desired space
+ * @color: color for the desired space
  * @start: start (inclusive) of the range from which to evict objects
  * @end: end (exclusive) of the range from which to evict objects
  * @flags: additional flags to control the eviction algorithm
@@ -91,11 +89,10 @@ mark_free(struct drm_mm_scan *scan,
 int
 i915_gem_evict_something(struct i915_address_space *vm,
 			 u64 min_size, u64 alignment,
-			 unsigned cache_level,
+			 unsigned long color,
 			 u64 start, u64 end,
 			 unsigned flags)
 {
-	struct drm_i915_private *dev_priv = vm->i915;
 	struct drm_mm_scan scan;
 	struct list_head eviction_list;
 	struct i915_vma *vma, *next;
@@ -104,7 +101,7 @@ i915_gem_evict_something(struct i915_address_space *vm,
 	struct i915_vma *active;
 	int ret;
 
-	lockdep_assert_held(&vm->i915->drm.struct_mutex);
+	lockdep_assert_held(&vm->mutex);
 	trace_i915_gem_evict(vm, min_size, alignment, flags);
 
 	/*
@@ -124,17 +121,10 @@ i915_gem_evict_something(struct i915_address_space *vm,
 	if (flags & PIN_MAPPABLE)
 		mode = DRM_MM_INSERT_LOW;
 	drm_mm_scan_init_with_range(&scan, &vm->mm,
-				    min_size, alignment, cache_level,
+				    min_size, alignment, color,
 				    start, end, mode);
 
-	/*
-	 * Retire before we search the active list. Although we have
-	 * reasonable accuracy in our retirement lists, we may have
-	 * a stray pin (preventing eviction) that can only be resolved by
-	 * retiring.
-	 */
-	if (!(flags & PIN_NONBLOCK))
-		i915_retire_requests(dev_priv);
+	intel_gt_retire_requests(vm->gt);
 
 search_again:
 	active = NULL;
@@ -207,7 +197,7 @@ search_again:
 	if (I915_SELFTEST_ONLY(igt_evict_ctl.fail_if_busy))
 		return -EBUSY;
 
-	ret = ggtt_flush(dev_priv);
+	ret = ggtt_flush(vm->gt);
 	if (ret)
 		return ret;
 
@@ -235,12 +225,12 @@ found:
 	list_for_each_entry_safe(vma, next, &eviction_list, evict_link) {
 		__i915_vma_unpin(vma);
 		if (ret == 0)
-			ret = i915_vma_unbind(vma);
+			ret = __i915_vma_unbind(vma);
 	}
 
 	while (ret == 0 && (node = drm_mm_scan_color_evict(&scan))) {
 		vma = container_of(node, struct i915_vma, node);
-		ret = i915_vma_unbind(vma);
+		ret = __i915_vma_unbind(vma);
 	}
 
 	return ret;
@@ -266,25 +256,23 @@ int i915_gem_evict_for_node(struct i915_address_space *vm,
 	u64 start = target->start;
 	u64 end = start + target->size;
 	struct i915_vma *vma, *next;
-	bool check_color;
 	int ret = 0;
 
-	lockdep_assert_held(&vm->i915->drm.struct_mutex);
+	lockdep_assert_held(&vm->mutex);
 	GEM_BUG_ON(!IS_ALIGNED(start, I915_GTT_PAGE_SIZE));
 	GEM_BUG_ON(!IS_ALIGNED(end, I915_GTT_PAGE_SIZE));
 
 	trace_i915_gem_evict_node(vm, target, flags);
 
-	/* Retire before we search the active list. Although we have
+	/*
+	 * Retire before we search the active list. Although we have
 	 * reasonable accuracy in our retirement lists, we may have
 	 * a stray pin (preventing eviction) that can only be resolved by
 	 * retiring.
 	 */
-	if (!(flags & PIN_NONBLOCK))
-		i915_retire_requests(vm->i915);
+	intel_gt_retire_requests(vm->gt);
 
-	check_color = vm->mm.color_adjust;
-	if (check_color) {
+	if (i915_vm_has_cache_coloring(vm)) {
 		/* Expand search to cover neighbouring guard pages (or lack!) */
 		if (start)
 			start -= I915_GTT_PAGE_SIZE;
@@ -301,7 +289,7 @@ int i915_gem_evict_for_node(struct i915_address_space *vm,
 			break;
 		}
 
-		GEM_BUG_ON(!node->allocated);
+		GEM_BUG_ON(!drm_mm_node_allocated(node));
 		vma = container_of(node, typeof(*vma), node);
 
 		/* If we are using coloring to insert guard pages between
@@ -310,7 +298,7 @@ int i915_gem_evict_for_node(struct i915_address_space *vm,
 		 * abutt and conflict. If they are in conflict, then we evict
 		 * those as well to make room for our guard pages.
 		 */
-		if (check_color) {
+		if (i915_vm_has_cache_coloring(vm)) {
 			if (node->start + node->size == target->start) {
 				if (node->color == target->color)
 					continue;
@@ -351,7 +339,7 @@ int i915_gem_evict_for_node(struct i915_address_space *vm,
 	list_for_each_entry_safe(vma, next, &eviction_list, evict_link) {
 		__i915_vma_unpin(vma);
 		if (ret == 0)
-			ret = i915_vma_unbind(vma);
+			ret = __i915_vma_unbind(vma);
 	}
 
 	return ret;
@@ -375,7 +363,7 @@ int i915_gem_evict_vm(struct i915_address_space *vm)
 	struct i915_vma *vma, *next;
 	int ret;
 
-	lockdep_assert_held(&vm->i915->drm.struct_mutex);
+	lockdep_assert_held(&vm->mutex);
 	trace_i915_gem_evict_vm(vm);
 
 	/* Switch back to the default context in order to unpin
@@ -384,13 +372,12 @@ int i915_gem_evict_vm(struct i915_address_space *vm)
 	 * switch otherwise is ineffective.
 	 */
 	if (i915_is_ggtt(vm)) {
-		ret = ggtt_flush(vm->i915);
+		ret = ggtt_flush(vm->gt);
 		if (ret)
 			return ret;
 	}
 
 	INIT_LIST_HEAD(&eviction_list);
-	mutex_lock(&vm->mutex);
 	list_for_each_entry(vma, &vm->bound_list, vm_link) {
 		if (i915_vma_is_pinned(vma))
 			continue;
@@ -398,13 +385,12 @@ int i915_gem_evict_vm(struct i915_address_space *vm)
 		__i915_vma_pin(vma);
 		list_add(&vma->evict_link, &eviction_list);
 	}
-	mutex_unlock(&vm->mutex);
 
 	ret = 0;
 	list_for_each_entry_safe(vma, next, &eviction_list, evict_link) {
 		__i915_vma_unpin(vma);
 		if (ret == 0)
-			ret = i915_vma_unbind(vma);
+			ret = __i915_vma_unbind(vma);
 	}
 	return ret;
 }
diff --git a/drivers/gpu/drm/i915/i915_gem_fence_reg.c b/drivers/gpu/drm/i915/i915_gem_fence_reg.c
index 615a9f4ef30c..71efccfde122 100644
--- a/drivers/gpu/drm/i915/i915_gem_fence_reg.c
+++ b/drivers/gpu/drm/i915/i915_gem_fence_reg.c
@@ -59,6 +59,16 @@
 
 #define pipelined 0
 
+static struct drm_i915_private *fence_to_i915(struct i915_fence_reg *fence)
+{
+	return fence->ggtt->vm.i915;
+}
+
+static struct intel_uncore *fence_to_uncore(struct i915_fence_reg *fence)
+{
+	return fence->ggtt->vm.gt->uncore;
+}
+
 static void i965_write_fence_reg(struct i915_fence_reg *fence,
 				 struct i915_vma *vma)
 {
@@ -66,7 +76,7 @@ static void i965_write_fence_reg(struct i915_fence_reg *fence,
 	int fence_pitch_shift;
 	u64 val;
 
-	if (INTEL_GEN(fence->i915) >= 6) {
+	if (INTEL_GEN(fence_to_i915(fence)) >= 6) {
 		fence_reg_lo = FENCE_REG_GEN6_LO(fence->id);
 		fence_reg_hi = FENCE_REG_GEN6_HI(fence->id);
 		fence_pitch_shift = GEN6_FENCE_PITCH_SHIFT;
@@ -95,7 +105,7 @@ static void i965_write_fence_reg(struct i915_fence_reg *fence,
 	}
 
 	if (!pipelined) {
-		struct intel_uncore *uncore = &fence->i915->uncore;
+		struct intel_uncore *uncore = fence_to_uncore(fence);
 
 		/*
 		 * To w/a incoherency with non-atomic 64-bit register updates,
@@ -132,7 +142,7 @@ static void i915_write_fence_reg(struct i915_fence_reg *fence,
 		GEM_BUG_ON(!is_power_of_2(vma->fence_size));
 		GEM_BUG_ON(!IS_ALIGNED(vma->node.start, vma->fence_size));
 
-		if (is_y_tiled && HAS_128_BYTE_Y_TILING(fence->i915))
+		if (is_y_tiled && HAS_128_BYTE_Y_TILING(fence_to_i915(fence)))
 			stride /= 128;
 		else
 			stride /= 512;
@@ -148,7 +158,7 @@ static void i915_write_fence_reg(struct i915_fence_reg *fence,
 	}
 
 	if (!pipelined) {
-		struct intel_uncore *uncore = &fence->i915->uncore;
+		struct intel_uncore *uncore = fence_to_uncore(fence);
 		i915_reg_t reg = FENCE_REG(fence->id);
 
 		intel_uncore_write_fw(uncore, reg, val);
@@ -180,7 +190,7 @@ static void i830_write_fence_reg(struct i915_fence_reg *fence,
 	}
 
 	if (!pipelined) {
-		struct intel_uncore *uncore = &fence->i915->uncore;
+		struct intel_uncore *uncore = fence_to_uncore(fence);
 		i915_reg_t reg = FENCE_REG(fence->id);
 
 		intel_uncore_write_fw(uncore, reg, val);
@@ -191,15 +201,17 @@ static void i830_write_fence_reg(struct i915_fence_reg *fence,
 static void fence_write(struct i915_fence_reg *fence,
 			struct i915_vma *vma)
 {
+	struct drm_i915_private *i915 = fence_to_i915(fence);
+
 	/*
 	 * Previous access through the fence register is marshalled by
 	 * the mb() inside the fault handlers (i915_gem_release_mmaps)
 	 * and explicitly managed for internal users.
 	 */
 
-	if (IS_GEN(fence->i915, 2))
+	if (IS_GEN(i915, 2))
 		i830_write_fence_reg(fence, vma);
-	else if (IS_GEN(fence->i915, 3))
+	else if (IS_GEN(i915, 3))
 		i915_write_fence_reg(fence, vma);
 	else
 		i965_write_fence_reg(fence, vma);
@@ -215,6 +227,8 @@ static void fence_write(struct i915_fence_reg *fence,
 static int fence_update(struct i915_fence_reg *fence,
 			struct i915_vma *vma)
 {
+	struct i915_ggtt *ggtt = fence->ggtt;
+	struct intel_uncore *uncore = fence_to_uncore(fence);
 	intel_wakeref_t wakeref;
 	struct i915_vma *old;
 	int ret;
@@ -230,14 +244,15 @@ static int fence_update(struct i915_fence_reg *fence,
 			 i915_gem_object_get_tiling(vma->obj)))
 			return -EINVAL;
 
-		ret = i915_active_wait(&vma->active);
+		ret = i915_vma_sync(vma);
 		if (ret)
 			return ret;
 	}
 
 	old = xchg(&fence->vma, NULL);
 	if (old) {
-		ret = i915_active_wait(&old->active);
+		/* XXX Ideally we would move the waiting to outside the mutex */
+		ret = i915_vma_sync(old);
 		if (ret) {
 			fence->vma = old;
 			return ret;
@@ -255,7 +270,7 @@ static int fence_update(struct i915_fence_reg *fence,
 			old->fence = NULL;
 		}
 
-		list_move(&fence->link, &fence->i915->ggtt.fence_list);
+		list_move(&fence->link, &ggtt->fence_list);
 	}
 
 	/*
@@ -268,7 +283,7 @@ static int fence_update(struct i915_fence_reg *fence,
 	 * be cleared before we can use any other fences to ensure that
 	 * the new fences do not overlap the elided clears, confusing HW.
 	 */
-	wakeref = intel_runtime_pm_get_if_in_use(&fence->i915->runtime_pm);
+	wakeref = intel_runtime_pm_get_if_in_use(uncore->rpm);
 	if (!wakeref) {
 		GEM_BUG_ON(vma);
 		return 0;
@@ -279,10 +294,10 @@ static int fence_update(struct i915_fence_reg *fence,
 
 	if (vma) {
 		vma->fence = fence;
-		list_move_tail(&fence->link, &fence->i915->ggtt.fence_list);
+		list_move_tail(&fence->link, &ggtt->fence_list);
 	}
 
-	intel_runtime_pm_put(&fence->i915->runtime_pm, wakeref);
+	intel_runtime_pm_put(uncore->rpm, wakeref);
 	return 0;
 }
 
@@ -311,11 +326,11 @@ int i915_vma_revoke_fence(struct i915_vma *vma)
 	return fence_update(fence, NULL);
 }
 
-static struct i915_fence_reg *fence_find(struct drm_i915_private *i915)
+static struct i915_fence_reg *fence_find(struct i915_ggtt *ggtt)
 {
 	struct i915_fence_reg *fence;
 
-	list_for_each_entry(fence, &i915->ggtt.fence_list, link) {
+	list_for_each_entry(fence, &ggtt->fence_list, link) {
 		GEM_BUG_ON(fence->vma && fence->vma->fence != fence);
 
 		if (atomic_read(&fence->pin_count))
@@ -325,19 +340,21 @@ static struct i915_fence_reg *fence_find(struct drm_i915_private *i915)
 	}
 
 	/* Wait for completion of pending flips which consume fences */
-	if (intel_has_pending_fb_unpin(i915))
+	if (intel_has_pending_fb_unpin(ggtt->vm.i915))
 		return ERR_PTR(-EAGAIN);
 
 	return ERR_PTR(-EDEADLK);
 }
 
-static int __i915_vma_pin_fence(struct i915_vma *vma)
+int __i915_vma_pin_fence(struct i915_vma *vma)
 {
 	struct i915_ggtt *ggtt = i915_vm_to_ggtt(vma->vm);
 	struct i915_fence_reg *fence;
 	struct i915_vma *set = i915_gem_object_is_tiled(vma->obj) ? vma : NULL;
 	int err;
 
+	lockdep_assert_held(&vma->vm->mutex);
+
 	/* Just update our place in the LRU if our fence is getting reused. */
 	if (vma->fence) {
 		fence = vma->fence;
@@ -348,7 +365,7 @@ static int __i915_vma_pin_fence(struct i915_vma *vma)
 			return 0;
 		}
 	} else if (set) {
-		fence = fence_find(vma->vm->i915);
+		fence = fence_find(ggtt);
 		if (IS_ERR(fence))
 			return PTR_ERR(fence);
 
@@ -399,7 +416,7 @@ int i915_vma_pin_fence(struct i915_vma *vma)
 	 * Note that we revoke fences on runtime suspend. Therefore the user
 	 * must keep the device awake whilst using the fence.
 	 */
-	assert_rpm_wakelock_held(&vma->vm->i915->runtime_pm);
+	assert_rpm_wakelock_held(vma->vm->gt->uncore->rpm);
 	GEM_BUG_ON(!i915_vma_is_pinned(vma));
 	GEM_BUG_ON(!i915_vma_is_ggtt(vma));
 
@@ -415,14 +432,13 @@ int i915_vma_pin_fence(struct i915_vma *vma)
 
 /**
  * i915_reserve_fence - Reserve a fence for vGPU
- * @i915: i915 device private
+ * @ggtt: Global GTT
  *
  * This function walks the fence regs looking for a free one and remove
  * it from the fence_list. It is used to reserve fence for vGPU to use.
  */
-struct i915_fence_reg *i915_reserve_fence(struct drm_i915_private *i915)
+struct i915_fence_reg *i915_reserve_fence(struct i915_ggtt *ggtt)
 {
-	struct i915_ggtt *ggtt = &i915->ggtt;
 	struct i915_fence_reg *fence;
 	int count;
 	int ret;
@@ -436,7 +452,7 @@ struct i915_fence_reg *i915_reserve_fence(struct drm_i915_private *i915)
 	if (count <= 1)
 		return ERR_PTR(-ENOSPC);
 
-	fence = fence_find(i915);
+	fence = fence_find(ggtt);
 	if (IS_ERR(fence))
 		return fence;
 
@@ -460,7 +476,7 @@ struct i915_fence_reg *i915_reserve_fence(struct drm_i915_private *i915)
  */
 void i915_unreserve_fence(struct i915_fence_reg *fence)
 {
-	struct i915_ggtt *ggtt = &fence->i915->ggtt;
+	struct i915_ggtt *ggtt = fence->ggtt;
 
 	lockdep_assert_held(&ggtt->vm.mutex);
 
@@ -469,19 +485,19 @@ void i915_unreserve_fence(struct i915_fence_reg *fence)
 
 /**
  * i915_gem_restore_fences - restore fence state
- * @i915: i915 device private
+ * @ggtt: Global GTT
  *
  * Restore the hw fence state to match the software tracking again, to be called
  * after a gpu reset and on resume. Note that on runtime suspend we only cancel
  * the fences, to be reacquired by the user later.
  */
-void i915_gem_restore_fences(struct drm_i915_private *i915)
+void i915_gem_restore_fences(struct i915_ggtt *ggtt)
 {
 	int i;
 
 	rcu_read_lock(); /* keep obj alive as we dereference */
-	for (i = 0; i < i915->ggtt.num_fences; i++) {
-		struct i915_fence_reg *reg = &i915->ggtt.fence_regs[i];
+	for (i = 0; i < ggtt->num_fences; i++) {
+		struct i915_fence_reg *reg = &ggtt->fence_regs[i];
 		struct i915_vma *vma = READ_ONCE(reg->vma);
 
 		GEM_BUG_ON(vma && vma->fence != reg);
@@ -547,15 +563,16 @@ void i915_gem_restore_fences(struct drm_i915_private *i915)
  */
 
 /**
- * i915_gem_detect_bit_6_swizzle - detect bit 6 swizzling pattern
- * @i915: i915 device private
+ * detect_bit_6_swizzle - detect bit 6 swizzling pattern
+ * @ggtt: Global GGTT
  *
  * Detects bit 6 swizzling of address lookup between IGD access and CPU
  * access through main memory.
  */
-static void detect_bit_6_swizzle(struct drm_i915_private *i915)
+static void detect_bit_6_swizzle(struct i915_ggtt *ggtt)
 {
-	struct intel_uncore *uncore = &i915->uncore;
+	struct intel_uncore *uncore = ggtt->vm.gt->uncore;
+	struct drm_i915_private *i915 = ggtt->vm.i915;
 	u32 swizzle_x = I915_BIT_6_SWIZZLE_UNKNOWN;
 	u32 swizzle_y = I915_BIT_6_SWIZZLE_UNKNOWN;
 
@@ -717,8 +734,8 @@ static void detect_bit_6_swizzle(struct drm_i915_private *i915)
 		swizzle_y = I915_BIT_6_SWIZZLE_NONE;
 	}
 
-	i915->mm.bit_6_swizzle_x = swizzle_x;
-	i915->mm.bit_6_swizzle_y = swizzle_y;
+	i915->ggtt.bit_6_swizzle_x = swizzle_x;
+	i915->ggtt.bit_6_swizzle_y = swizzle_y;
 }
 
 /*
@@ -819,17 +836,20 @@ i915_gem_object_save_bit_17_swizzle(struct drm_i915_gem_object *obj,
 void i915_ggtt_init_fences(struct i915_ggtt *ggtt)
 {
 	struct drm_i915_private *i915 = ggtt->vm.i915;
+	struct intel_uncore *uncore = ggtt->vm.gt->uncore;
 	int num_fences;
 	int i;
 
 	INIT_LIST_HEAD(&ggtt->fence_list);
 	INIT_LIST_HEAD(&ggtt->userfault_list);
-	intel_wakeref_auto_init(&ggtt->userfault_wakeref, &i915->runtime_pm);
+	intel_wakeref_auto_init(&ggtt->userfault_wakeref, uncore->rpm);
 
-	detect_bit_6_swizzle(i915);
+	detect_bit_6_swizzle(ggtt);
 
-	if (INTEL_GEN(i915) >= 7 &&
-	    !(IS_VALLEYVIEW(i915) || IS_CHERRYVIEW(i915)))
+	if (!i915_ggtt_has_aperture(ggtt))
+		num_fences = 0;
+	else if (INTEL_GEN(i915) >= 7 &&
+		 !(IS_VALLEYVIEW(i915) || IS_CHERRYVIEW(i915)))
 		num_fences = 32;
 	else if (INTEL_GEN(i915) >= 4 ||
 		 IS_I945G(i915) || IS_I945GM(i915) ||
@@ -839,20 +859,20 @@ void i915_ggtt_init_fences(struct i915_ggtt *ggtt)
 		num_fences = 8;
 
 	if (intel_vgpu_active(i915))
-		num_fences = intel_uncore_read(&i915->uncore,
+		num_fences = intel_uncore_read(uncore,
 					       vgtif_reg(avail_rs.fence_num));
 
 	/* Initialize fence registers to zero */
 	for (i = 0; i < num_fences; i++) {
 		struct i915_fence_reg *fence = &ggtt->fence_regs[i];
 
-		fence->i915 = i915;
+		fence->ggtt = ggtt;
 		fence->id = i;
 		list_add_tail(&fence->link, &ggtt->fence_list);
 	}
 	ggtt->num_fences = num_fences;
 
-	i915_gem_restore_fences(i915);
+	i915_gem_restore_fences(ggtt);
 }
 
 void intel_gt_init_swizzling(struct intel_gt *gt)
@@ -861,7 +881,7 @@ void intel_gt_init_swizzling(struct intel_gt *gt)
 	struct intel_uncore *uncore = gt->uncore;
 
 	if (INTEL_GEN(i915) < 5 ||
-	    i915->mm.bit_6_swizzle_x == I915_BIT_6_SWIZZLE_NONE)
+	    i915->ggtt.bit_6_swizzle_x == I915_BIT_6_SWIZZLE_NONE)
 		return;
 
 	intel_uncore_rmw(uncore, DISP_ARB_CTL, 0, DISP_TILE_SURFACE_SWIZZLING);
diff --git a/drivers/gpu/drm/i915/i915_gem_fence_reg.h b/drivers/gpu/drm/i915/i915_gem_fence_reg.h
index 99866fb9d94f..7bd521cd7cd7 100644
--- a/drivers/gpu/drm/i915/i915_gem_fence_reg.h
+++ b/drivers/gpu/drm/i915/i915_gem_fence_reg.h
@@ -29,7 +29,6 @@
 #include <linux/types.h>
 
 struct drm_i915_gem_object;
-struct drm_i915_private;
 struct i915_ggtt;
 struct i915_vma;
 struct intel_gt;
@@ -39,7 +38,7 @@ struct sg_table;
 
 struct i915_fence_reg {
 	struct list_head link;
-	struct drm_i915_private *i915;
+	struct i915_ggtt *ggtt;
 	struct i915_vma *vma;
 	atomic_t pin_count;
 	int id;
@@ -55,10 +54,10 @@ struct i915_fence_reg {
 };
 
 /* i915_gem_fence_reg.c */
-struct i915_fence_reg *i915_reserve_fence(struct drm_i915_private *i915);
+struct i915_fence_reg *i915_reserve_fence(struct i915_ggtt *ggtt);
 void i915_unreserve_fence(struct i915_fence_reg *fence);
 
-void i915_gem_restore_fences(struct drm_i915_private *i915);
+void i915_gem_restore_fences(struct i915_ggtt *ggtt);
 
 void i915_gem_object_do_bit_17_swizzle(struct drm_i915_gem_object *obj,
 				       struct sg_table *pages);
diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c
index b1a7a8b9b46a..6239a9adbf14 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.c
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.c
@@ -38,6 +38,7 @@
 
 #include "display/intel_frontbuffer.h"
 #include "gt/intel_gt.h"
+#include "gt/intel_gt_requests.h"
 
 #include "i915_drv.h"
 #include "i915_scatterlist.h"
@@ -132,9 +133,15 @@ static void gen6_ggtt_invalidate(struct i915_ggtt *ggtt)
 static void guc_ggtt_invalidate(struct i915_ggtt *ggtt)
 {
 	struct intel_uncore *uncore = ggtt->vm.gt->uncore;
+	struct drm_i915_private *i915 = ggtt->vm.i915;
 
 	gen6_ggtt_invalidate(ggtt);
-	intel_uncore_write_fw(uncore, GEN8_GTCR, GEN8_GTCR_INVALIDATE);
+
+	if (INTEL_GEN(i915) >= 12)
+		intel_uncore_write_fw(uncore, GEN12_GUC_TLB_INV_CR,
+				      GEN12_GUC_TLB_INV_CR_INVALIDATE);
+	else
+		intel_uncore_write_fw(uncore, GEN8_GTCR, GEN8_GTCR_INVALIDATE);
 }
 
 static void gmch_ggtt_invalidate(struct i915_ggtt *ggtt)
@@ -144,16 +151,18 @@ static void gmch_ggtt_invalidate(struct i915_ggtt *ggtt)
 
 static int ppgtt_bind_vma(struct i915_vma *vma,
 			  enum i915_cache_level cache_level,
-			  u32 unused)
+			  u32 flags)
 {
 	u32 pte_flags;
 	int err;
 
-	if (!(vma->flags & I915_VMA_LOCAL_BIND)) {
+	if (flags & I915_VMA_ALLOC) {
 		err = vma->vm->allocate_va_range(vma->vm,
 						 vma->node.start, vma->size);
 		if (err)
 			return err;
+
+		set_bit(I915_VMA_ALLOC_BIT, __i915_vma_flags(vma));
 	}
 
 	/* Applicable to VLV, and gen8+ */
@@ -161,14 +170,17 @@ static int ppgtt_bind_vma(struct i915_vma *vma,
 	if (i915_gem_object_is_readonly(vma->obj))
 		pte_flags |= PTE_READ_ONLY;
 
+	GEM_BUG_ON(!test_bit(I915_VMA_ALLOC_BIT, __i915_vma_flags(vma)));
 	vma->vm->insert_entries(vma->vm, vma, cache_level, pte_flags);
+	wmb();
 
 	return 0;
 }
 
 static void ppgtt_unbind_vma(struct i915_vma *vma)
 {
-	vma->vm->clear_range(vma->vm, vma->node.start, vma->size);
+	if (test_and_clear_bit(I915_VMA_ALLOC_BIT, __i915_vma_flags(vma)))
+		vma->vm->clear_range(vma->vm, vma->node.start, vma->size);
 }
 
 static int ppgtt_set_pages(struct i915_vma *vma)
@@ -496,22 +508,26 @@ static void i915_address_space_fini(struct i915_address_space *vm)
 	mutex_destroy(&vm->mutex);
 }
 
-static void ppgtt_destroy_vma(struct i915_address_space *vm)
+void __i915_vm_close(struct i915_address_space *vm)
 {
-	struct list_head *phases[] = {
-		&vm->bound_list,
-		&vm->unbound_list,
-		NULL,
-	}, **phase;
+	struct i915_vma *vma, *vn;
+
+	mutex_lock(&vm->mutex);
+	list_for_each_entry_safe(vma, vn, &vm->bound_list, vm_link) {
+		struct drm_i915_gem_object *obj = vma->obj;
 
-	mutex_lock(&vm->i915->drm.struct_mutex);
-	for (phase = phases; *phase; phase++) {
-		struct i915_vma *vma, *vn;
+		/* Keep the obj (and hence the vma) alive as _we_ destroy it */
+		if (!kref_get_unless_zero(&obj->base.refcount))
+			continue;
+
+		atomic_and(~I915_VMA_PIN_MASK, &vma->flags);
+		WARN_ON(__i915_vma_unbind(vma));
+		i915_vma_destroy(vma);
 
-		list_for_each_entry_safe(vma, vn, *phase, vm_link)
-			i915_vma_destroy(vma);
+		i915_gem_object_put(obj);
 	}
-	mutex_unlock(&vm->i915->drm.struct_mutex);
+	GEM_BUG_ON(!list_empty(&vm->bound_list));
+	mutex_unlock(&vm->mutex);
 }
 
 static void __i915_vm_release(struct work_struct *work)
@@ -519,11 +535,6 @@ static void __i915_vm_release(struct work_struct *work)
 	struct i915_address_space *vm =
 		container_of(work, struct i915_address_space, rcu.work);
 
-	ppgtt_destroy_vma(vm);
-
-	GEM_BUG_ON(!list_empty(&vm->bound_list));
-	GEM_BUG_ON(!list_empty(&vm->unbound_list));
-
 	vm->cleanup(vm);
 	i915_address_space_fini(vm);
 
@@ -538,7 +549,6 @@ void i915_vm_release(struct kref *kref)
 	GEM_BUG_ON(i915_is_ggtt(vm));
 	trace_i915_ppgtt_release(vm);
 
-	vm->closed = true;
 	queue_rcu_work(vm->i915->wq, &vm->rcu);
 }
 
@@ -546,6 +556,7 @@ static void i915_address_space_init(struct i915_address_space *vm, int subclass)
 {
 	kref_init(&vm->ref);
 	INIT_RCU_WORK(&vm->rcu, __i915_vm_release);
+	atomic_set(&vm->open, 1);
 
 	/*
 	 * The vm->mutex must be reclaim safe (for use in the shrinker).
@@ -562,7 +573,6 @@ static void i915_address_space_init(struct i915_address_space *vm, int subclass)
 
 	stash_init(&vm->free_pages);
 
-	INIT_LIST_HEAD(&vm->unbound_list);
 	INIT_LIST_HEAD(&vm->bound_list);
 }
 
@@ -816,17 +826,6 @@ release_pd_entry(struct i915_page_directory * const pd,
 	return free;
 }
 
-/*
- * PDE TLBs are a pain to invalidate on GEN8+. When we modify
- * the page table structures, we mark them dirty so that
- * context switching/execlist queuing code takes extra steps
- * to ensure that tlbs are flushed.
- */
-static void mark_tlbs_dirty(struct i915_ppgtt *ppgtt)
-{
-	ppgtt->pd_dirty_engines = ALL_ENGINES;
-}
-
 static void gen8_ppgtt_notify_vgt(struct i915_ppgtt *ppgtt, bool create)
 {
 	struct drm_i915_private *dev_priv = ppgtt->vm.i915;
@@ -1367,7 +1366,9 @@ static int gen8_init_scratch(struct i915_address_space *vm)
 	if (vm->has_read_only &&
 	    vm->i915->kernel_context &&
 	    vm->i915->kernel_context->vm) {
-		struct i915_address_space *clone = vm->i915->kernel_context->vm;
+		struct i915_address_space *clone =
+			rcu_dereference_protected(vm->i915->kernel_context->vm,
+						  true); /* static */
 
 		GEM_BUG_ON(!clone->has_read_only);
 
@@ -1422,6 +1423,7 @@ static int gen8_preallocate_top_level_pdp(struct i915_ppgtt *ppgtt)
 		set_pd_entry(pd, idx, pde);
 		atomic_inc(px_used(pde)); /* keep pinned */
 	}
+	wmb();
 
 	return 0;
 }
@@ -1489,8 +1491,10 @@ static struct i915_ppgtt *gen8_ppgtt_create(struct drm_i915_private *i915)
 	 *
 	 * Gen11 has HSDES#:1807136187 unresolved. Disable ro support
 	 * for now.
+	 *
+	 * Gen12 has inherited the same read-only fault issue from gen11.
 	 */
-	ppgtt->vm.has_read_only = INTEL_GEN(i915) != 11;
+	ppgtt->vm.has_read_only = !IS_GEN_RANGE(i915, 11, 12);
 
 	/* There are only few exceptions for gen >=6. chv and bxt.
 	 * And we are not sure about the latter so play safe for now.
@@ -1509,13 +1513,12 @@ static struct i915_ppgtt *gen8_ppgtt_create(struct drm_i915_private *i915)
 	}
 
 	if (!i915_vm_is_4lvl(&ppgtt->vm)) {
-		if (intel_vgpu_active(i915)) {
-			err = gen8_preallocate_top_level_pdp(ppgtt);
-			if (err)
-				goto err_free_pd;
-		}
+		err = gen8_preallocate_top_level_pdp(ppgtt);
+		if (err)
+			goto err_free_pd;
 	}
 
+	ppgtt->vm.bind_async_flags = I915_VMA_LOCAL_BIND;
 	ppgtt->vm.insert_entries = gen8_ppgtt_insert;
 	ppgtt->vm.allocate_va_range = gen8_ppgtt_alloc;
 	ppgtt->vm.clear_range = gen8_ppgtt_clear;
@@ -1566,7 +1569,7 @@ static void gen7_ppgtt_enable(struct intel_gt *gt)
 	}
 	intel_uncore_write(uncore, GAM_ECOCHK, ecochk);
 
-	for_each_engine(engine, i915, id) {
+	for_each_engine(engine, gt, id) {
 		/* GFX_MODE is per-ring on gen7+ */
 		ENGINE_WRITE(engine,
 			     RING_MODE_GEN7,
@@ -1729,10 +1732,8 @@ static int gen6_alloc_va_range(struct i915_address_space *vm,
 	}
 	spin_unlock(&pd->lock);
 
-	if (flush) {
-		mark_tlbs_dirty(&ppgtt->base);
+	if (flush)
 		gen6_ggtt_invalidate(vm->gt->ggtt);
-	}
 
 	goto out;
 
@@ -1786,15 +1787,13 @@ static void gen6_ppgtt_free_pd(struct gen6_ppgtt *ppgtt)
 static void gen6_ppgtt_cleanup(struct i915_address_space *vm)
 {
 	struct gen6_ppgtt *ppgtt = to_gen6_ppgtt(i915_vm_to_ppgtt(vm));
-	struct drm_i915_private *i915 = vm->i915;
 
-	/* FIXME remove the struct_mutex to bring the locking under control */
-	mutex_lock(&i915->drm.struct_mutex);
 	i915_vma_destroy(ppgtt->vma);
-	mutex_unlock(&i915->drm.struct_mutex);
 
 	gen6_ppgtt_free_pd(ppgtt);
 	free_scratch(vm);
+
+	mutex_destroy(&ppgtt->pin_mutex);
 	kfree(ppgtt->base.pd);
 }
 
@@ -1827,7 +1826,6 @@ static int pd_vma_bind(struct i915_vma *vma,
 	gen6_for_all_pdes(pt, ppgtt->base.pd, pde)
 		gen6_write_pde(ppgtt, pde, pt);
 
-	mark_tlbs_dirty(&ppgtt->base);
 	gen6_ggtt_invalidate(ggtt);
 
 	return 0;
@@ -1866,7 +1864,6 @@ static const struct i915_vma_ops pd_vma_ops = {
 
 static struct i915_vma *pd_vma_create(struct gen6_ppgtt *ppgtt, int size)
 {
-	struct drm_i915_private *i915 = ppgtt->base.vm.i915;
 	struct i915_ggtt *ggtt = ppgtt->base.vm.gt->ggtt;
 	struct i915_vma *vma;
 
@@ -1877,33 +1874,30 @@ static struct i915_vma *pd_vma_create(struct gen6_ppgtt *ppgtt, int size)
 	if (!vma)
 		return ERR_PTR(-ENOMEM);
 
-	i915_active_init(i915, &vma->active, NULL, NULL);
+	i915_active_init(&vma->active, NULL, NULL);
 
-	vma->vm = &ggtt->vm;
+	mutex_init(&vma->pages_mutex);
+	vma->vm = i915_vm_get(&ggtt->vm);
 	vma->ops = &pd_vma_ops;
 	vma->private = ppgtt;
 
 	vma->size = size;
 	vma->fence_size = size;
-	vma->flags = I915_VMA_GGTT;
+	atomic_set(&vma->flags, I915_VMA_GGTT);
 	vma->ggtt_view.type = I915_GGTT_VIEW_ROTATED; /* prevent fencing */
 
 	INIT_LIST_HEAD(&vma->obj_link);
 	INIT_LIST_HEAD(&vma->closed_link);
 
-	mutex_lock(&vma->vm->mutex);
-	list_add(&vma->vm_link, &vma->vm->unbound_list);
-	mutex_unlock(&vma->vm->mutex);
-
 	return vma;
 }
 
 int gen6_ppgtt_pin(struct i915_ppgtt *base)
 {
 	struct gen6_ppgtt *ppgtt = to_gen6_ppgtt(base);
-	int err;
+	int err = 0;
 
-	GEM_BUG_ON(ppgtt->base.vm.closed);
+	GEM_BUG_ON(!atomic_read(&ppgtt->base.vm.open));
 
 	/*
 	 * Workaround the limited maximum vma->pin_count and the aliasing_ppgtt
@@ -1911,24 +1905,26 @@ int gen6_ppgtt_pin(struct i915_ppgtt *base)
 	 * (When vma->pin_count becomes atomic, I expect we will naturally
 	 * need a larger, unpacked, type and kill this redundancy.)
 	 */
-	if (ppgtt->pin_count++)
+	if (atomic_add_unless(&ppgtt->pin_count, 1, 0))
 		return 0;
 
+	if (mutex_lock_interruptible(&ppgtt->pin_mutex))
+		return -EINTR;
+
 	/*
 	 * PPGTT PDEs reside in the GGTT and consists of 512 entries. The
 	 * allocator works in address space sizes, so it's multiplied by page
 	 * size. We allocate at the top of the GTT to avoid fragmentation.
 	 */
-	err = i915_vma_pin(ppgtt->vma,
-			   0, GEN6_PD_ALIGN,
-			   PIN_GLOBAL | PIN_HIGH);
-	if (err)
-		goto unpin;
-
-	return 0;
+	if (!atomic_read(&ppgtt->pin_count)) {
+		err = i915_vma_pin(ppgtt->vma,
+				   0, GEN6_PD_ALIGN,
+				   PIN_GLOBAL | PIN_HIGH);
+	}
+	if (!err)
+		atomic_inc(&ppgtt->pin_count);
+	mutex_unlock(&ppgtt->pin_mutex);
 
-unpin:
-	ppgtt->pin_count = 0;
 	return err;
 }
 
@@ -1936,22 +1932,20 @@ void gen6_ppgtt_unpin(struct i915_ppgtt *base)
 {
 	struct gen6_ppgtt *ppgtt = to_gen6_ppgtt(base);
 
-	GEM_BUG_ON(!ppgtt->pin_count);
-	if (--ppgtt->pin_count)
-		return;
-
-	i915_vma_unpin(ppgtt->vma);
+	GEM_BUG_ON(!atomic_read(&ppgtt->pin_count));
+	if (atomic_dec_and_test(&ppgtt->pin_count))
+		i915_vma_unpin(ppgtt->vma);
 }
 
 void gen6_ppgtt_unpin_all(struct i915_ppgtt *base)
 {
 	struct gen6_ppgtt *ppgtt = to_gen6_ppgtt(base);
 
-	if (!ppgtt->pin_count)
+	if (!atomic_read(&ppgtt->pin_count))
 		return;
 
-	ppgtt->pin_count = 0;
 	i915_vma_unpin(ppgtt->vma);
+	atomic_set(&ppgtt->pin_count, 0);
 }
 
 static struct i915_ppgtt *gen6_ppgtt_create(struct drm_i915_private *i915)
@@ -1964,9 +1958,12 @@ static struct i915_ppgtt *gen6_ppgtt_create(struct drm_i915_private *i915)
 	if (!ppgtt)
 		return ERR_PTR(-ENOMEM);
 
+	mutex_init(&ppgtt->pin_mutex);
+
 	ppgtt_init(&ppgtt->base, &i915->gt);
 	ppgtt->base.vm.top = 1;
 
+	ppgtt->base.vm.bind_async_flags = I915_VMA_LOCAL_BIND;
 	ppgtt->base.vm.allocate_va_range = gen6_alloc_va_range;
 	ppgtt->base.vm.clear_range = gen6_ppgtt_clear_range;
 	ppgtt->base.vm.insert_entries = gen6_ppgtt_insert_entries;
@@ -2023,7 +2020,7 @@ static void gtt_write_workarounds(struct intel_gt *gt)
 		intel_uncore_write(uncore,
 				   GEN8_L3_LRA_1_GPGPU,
 				   GEN9_L3_LRA_1_GPGPU_DEFAULT_VALUE_BXT);
-	else if (INTEL_GEN(i915) >= 9)
+	else if (INTEL_GEN(i915) >= 9 && INTEL_GEN(i915) <= 11)
 		intel_uncore_write(uncore,
 				   GEN8_L3_LRA_1_GPGPU,
 				   GEN9_L3_LRA_1_GPGPU_DEFAULT_VALUE_SKL);
@@ -2202,7 +2199,7 @@ static void gen8_ggtt_insert_entries(struct i915_address_space *vm,
 
 	gtt_entries = (gen8_pte_t __iomem *)ggtt->gsm;
 	gtt_entries += vma->node.start / I915_GTT_PAGE_SIZE;
-	for_each_sgt_dma(addr, sgt_iter, vma->pages)
+	for_each_sgt_daddr(addr, sgt_iter, vma->pages)
 		gen8_set_pte(gtt_entries++, pte_encode | addr);
 
 	/*
@@ -2243,7 +2240,7 @@ static void gen6_ggtt_insert_entries(struct i915_address_space *vm,
 	unsigned int i = vma->node.start / I915_GTT_PAGE_SIZE;
 	struct sgt_iter iter;
 	dma_addr_t addr;
-	for_each_sgt_dma(addr, iter, vma->pages)
+	for_each_sgt_daddr(addr, iter, vma->pages)
 		iowrite32(vm->pte_encode(addr, level, flags), &entries[i++]);
 
 	/*
@@ -2448,7 +2445,7 @@ static int ggtt_bind_vma(struct i915_vma *vma,
 	 * GLOBAL/LOCAL_BIND, it's all the same ptes. Hence unconditionally
 	 * upgrade to both bound if we bind either to avoid double-binding.
 	 */
-	vma->flags |= I915_VMA_GLOBAL_BIND | I915_VMA_LOCAL_BIND;
+	atomic_or(I915_VMA_GLOBAL_BIND | I915_VMA_LOCAL_BIND, &vma->flags);
 
 	return 0;
 }
@@ -2478,14 +2475,18 @@ static int aliasing_gtt_bind_vma(struct i915_vma *vma,
 	if (flags & I915_VMA_LOCAL_BIND) {
 		struct i915_ppgtt *alias = i915_vm_to_ggtt(vma->vm)->alias;
 
-		if (!(vma->flags & I915_VMA_LOCAL_BIND)) {
+		if (flags & I915_VMA_ALLOC) {
 			ret = alias->vm.allocate_va_range(&alias->vm,
 							  vma->node.start,
 							  vma->size);
 			if (ret)
 				return ret;
+
+			set_bit(I915_VMA_ALLOC_BIT, __i915_vma_flags(vma));
 		}
 
+		GEM_BUG_ON(!test_bit(I915_VMA_ALLOC_BIT,
+				     __i915_vma_flags(vma)));
 		alias->vm.insert_entries(&alias->vm, vma,
 					 cache_level, pte_flags);
 	}
@@ -2506,7 +2507,7 @@ static void aliasing_gtt_unbind_vma(struct i915_vma *vma)
 {
 	struct drm_i915_private *i915 = vma->vm->i915;
 
-	if (vma->flags & I915_VMA_GLOBAL_BIND) {
+	if (i915_vma_is_bound(vma, I915_VMA_GLOBAL_BIND)) {
 		struct i915_address_space *vm = vma->vm;
 		intel_wakeref_t wakeref;
 
@@ -2514,7 +2515,7 @@ static void aliasing_gtt_unbind_vma(struct i915_vma *vma)
 			vm->clear_range(vm, vma->node.start, vma->size);
 	}
 
-	if (vma->flags & I915_VMA_LOCAL_BIND) {
+	if (test_and_clear_bit(I915_VMA_ALLOC_BIT, __i915_vma_flags(vma))) {
 		struct i915_address_space *vm =
 			&i915_vm_to_ggtt(vma->vm)->alias->vm;
 
@@ -2530,7 +2531,9 @@ void i915_gem_gtt_finish_pages(struct drm_i915_gem_object *obj,
 	struct i915_ggtt *ggtt = &dev_priv->ggtt;
 
 	if (unlikely(ggtt->do_idle_maps)) {
-		if (i915_gem_wait_for_idle(dev_priv, 0, MAX_SCHEDULE_TIMEOUT)) {
+		/* XXX This does not prevent more requests being submitted! */
+		if (intel_gt_retire_requests_timeout(ggtt->vm.gt,
+						     -MAX_SCHEDULE_TIMEOUT)) {
 			DRM_ERROR("Failed to wait for idle; VT'd may hang.\n");
 			/* Wait a bit, in hopes it avoids the hang */
 			udelay(10);
@@ -2555,12 +2558,12 @@ static int ggtt_set_pages(struct i915_vma *vma)
 	return 0;
 }
 
-static void i915_gtt_color_adjust(const struct drm_mm_node *node,
-				  unsigned long color,
-				  u64 *start,
-				  u64 *end)
+static void i915_ggtt_color_adjust(const struct drm_mm_node *node,
+				   unsigned long color,
+				   u64 *start,
+				   u64 *end)
 {
-	if (node->allocated && node->color != color)
+	if (i915_node_color_differs(node, color))
 		*start += I915_GTT_PAGE_SIZE;
 
 	/* Also leave a space between the unallocated reserved node after the
@@ -2598,6 +2601,7 @@ static int init_aliasing_ppgtt(struct i915_ggtt *ggtt)
 		goto err_ppgtt;
 
 	ggtt->alias = ppgtt;
+	ggtt->vm.bind_async_flags |= ppgtt->vm.bind_async_flags;
 
 	GEM_BUG_ON(ggtt->vm.vma_ops.bind_vma != ggtt_bind_vma);
 	ggtt->vm.vma_ops.bind_vma = aliasing_gtt_bind_vma;
@@ -2614,22 +2618,16 @@ err_ppgtt:
 
 static void fini_aliasing_ppgtt(struct i915_ggtt *ggtt)
 {
-	struct drm_i915_private *i915 = ggtt->vm.i915;
 	struct i915_ppgtt *ppgtt;
 
-	mutex_lock(&i915->drm.struct_mutex);
-
 	ppgtt = fetch_and_zero(&ggtt->alias);
 	if (!ppgtt)
-		goto out;
+		return;
 
 	i915_vm_put(&ppgtt->vm);
 
 	ggtt->vm.vma_ops.bind_vma   = ggtt_bind_vma;
 	ggtt->vm.vma_ops.unbind_vma = ggtt_unbind_vma;
-
-out:
-	mutex_unlock(&i915->drm.struct_mutex);
 }
 
 static int ggtt_reserve_guc_top(struct i915_ggtt *ggtt)
@@ -2661,7 +2659,8 @@ static void ggtt_release_guc_top(struct i915_ggtt *ggtt)
 static void cleanup_init_ggtt(struct i915_ggtt *ggtt)
 {
 	ggtt_release_guc_top(ggtt);
-	drm_mm_remove_node(&ggtt->error_capture);
+	if (drm_mm_node_allocated(&ggtt->error_capture))
+		drm_mm_remove_node(&ggtt->error_capture);
 }
 
 static int init_ggtt(struct i915_ggtt *ggtt)
@@ -2692,13 +2691,15 @@ static int init_ggtt(struct i915_ggtt *ggtt)
 	if (ret)
 		return ret;
 
-	/* Reserve a mappable slot for our lockless error capture */
-	ret = drm_mm_insert_node_in_range(&ggtt->vm.mm, &ggtt->error_capture,
-					  PAGE_SIZE, 0, I915_COLOR_UNEVICTABLE,
-					  0, ggtt->mappable_end,
-					  DRM_MM_INSERT_LOW);
-	if (ret)
-		return ret;
+	if (ggtt->mappable_end) {
+		/* Reserve a mappable slot for our lockless error capture */
+		ret = drm_mm_insert_node_in_range(&ggtt->vm.mm, &ggtt->error_capture,
+						  PAGE_SIZE, 0, I915_COLOR_UNEVICTABLE,
+						  0, ggtt->mappable_end,
+						  DRM_MM_INSERT_LOW);
+		if (ret)
+			return ret;
+	}
 
 	/*
 	 * The upper portion of the GuC address space has a sizeable hole
@@ -2746,35 +2747,33 @@ int i915_init_ggtt(struct drm_i915_private *i915)
 
 static void ggtt_cleanup_hw(struct i915_ggtt *ggtt)
 {
-	struct drm_i915_private *i915 = ggtt->vm.i915;
 	struct i915_vma *vma, *vn;
 
-	ggtt->vm.closed = true;
+	atomic_set(&ggtt->vm.open, 0);
 
 	rcu_barrier(); /* flush the RCU'ed__i915_vm_release */
-	flush_workqueue(i915->wq);
+	flush_workqueue(ggtt->vm.i915->wq);
 
-	mutex_lock(&i915->drm.struct_mutex);
+	mutex_lock(&ggtt->vm.mutex);
 
 	list_for_each_entry_safe(vma, vn, &ggtt->vm.bound_list, vm_link)
-		WARN_ON(i915_vma_unbind(vma));
+		WARN_ON(__i915_vma_unbind(vma));
 
 	if (drm_mm_node_allocated(&ggtt->error_capture))
 		drm_mm_remove_node(&ggtt->error_capture);
 
 	ggtt_release_guc_top(ggtt);
-
-	if (drm_mm_initialized(&ggtt->vm.mm)) {
-		intel_vgt_deballoon(ggtt);
-		i915_address_space_fini(&ggtt->vm);
-	}
+	intel_vgt_deballoon(ggtt);
 
 	ggtt->vm.cleanup(&ggtt->vm);
 
-	mutex_unlock(&i915->drm.struct_mutex);
+	mutex_unlock(&ggtt->vm.mutex);
+	i915_address_space_fini(&ggtt->vm);
 
 	arch_phys_wc_del(ggtt->mtrr);
-	io_mapping_fini(&ggtt->iomap);
+
+	if (ggtt->iomap.size)
+		io_mapping_fini(&ggtt->iomap);
 }
 
 /**
@@ -2794,8 +2793,6 @@ void i915_ggtt_driver_release(struct drm_i915_private *i915)
 		set_pages_array_wb(pvec->pages, pvec->nr);
 		__pagevec_release(pvec);
 	}
-
-	i915_gem_cleanup_stolen(i915);
 }
 
 static unsigned int gen6_get_total_gtt_size(u16 snb_gmch_ctl)
@@ -2873,35 +2870,51 @@ static int ggtt_probe_common(struct i915_ggtt *ggtt, u64 size)
 	return 0;
 }
 
-static void tgl_setup_private_ppat(struct drm_i915_private *dev_priv)
+static void tgl_setup_private_ppat(struct intel_uncore *uncore)
 {
 	/* TGL doesn't support LLC or AGE settings */
-	I915_WRITE(GEN12_PAT_INDEX(0), GEN8_PPAT_WB);
-	I915_WRITE(GEN12_PAT_INDEX(1), GEN8_PPAT_WC);
-	I915_WRITE(GEN12_PAT_INDEX(2), GEN8_PPAT_WT);
-	I915_WRITE(GEN12_PAT_INDEX(3), GEN8_PPAT_UC);
-	I915_WRITE(GEN12_PAT_INDEX(4), GEN8_PPAT_WB);
-	I915_WRITE(GEN12_PAT_INDEX(5), GEN8_PPAT_WB);
-	I915_WRITE(GEN12_PAT_INDEX(6), GEN8_PPAT_WB);
-	I915_WRITE(GEN12_PAT_INDEX(7), GEN8_PPAT_WB);
-}
-
-static void cnl_setup_private_ppat(struct drm_i915_private *dev_priv)
-{
-	I915_WRITE(GEN10_PAT_INDEX(0), GEN8_PPAT_WB | GEN8_PPAT_LLC);
-	I915_WRITE(GEN10_PAT_INDEX(1), GEN8_PPAT_WC | GEN8_PPAT_LLCELLC);
-	I915_WRITE(GEN10_PAT_INDEX(2), GEN8_PPAT_WT | GEN8_PPAT_LLCELLC);
-	I915_WRITE(GEN10_PAT_INDEX(3), GEN8_PPAT_UC);
-	I915_WRITE(GEN10_PAT_INDEX(4), GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(0));
-	I915_WRITE(GEN10_PAT_INDEX(5), GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(1));
-	I915_WRITE(GEN10_PAT_INDEX(6), GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(2));
-	I915_WRITE(GEN10_PAT_INDEX(7), GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(3));
+	intel_uncore_write(uncore, GEN12_PAT_INDEX(0), GEN8_PPAT_WB);
+	intel_uncore_write(uncore, GEN12_PAT_INDEX(1), GEN8_PPAT_WC);
+	intel_uncore_write(uncore, GEN12_PAT_INDEX(2), GEN8_PPAT_WT);
+	intel_uncore_write(uncore, GEN12_PAT_INDEX(3), GEN8_PPAT_UC);
+	intel_uncore_write(uncore, GEN12_PAT_INDEX(4), GEN8_PPAT_WB);
+	intel_uncore_write(uncore, GEN12_PAT_INDEX(5), GEN8_PPAT_WB);
+	intel_uncore_write(uncore, GEN12_PAT_INDEX(6), GEN8_PPAT_WB);
+	intel_uncore_write(uncore, GEN12_PAT_INDEX(7), GEN8_PPAT_WB);
+}
+
+static void cnl_setup_private_ppat(struct intel_uncore *uncore)
+{
+	intel_uncore_write(uncore,
+			   GEN10_PAT_INDEX(0),
+			   GEN8_PPAT_WB | GEN8_PPAT_LLC);
+	intel_uncore_write(uncore,
+			   GEN10_PAT_INDEX(1),
+			   GEN8_PPAT_WC | GEN8_PPAT_LLCELLC);
+	intel_uncore_write(uncore,
+			   GEN10_PAT_INDEX(2),
+			   GEN8_PPAT_WT | GEN8_PPAT_LLCELLC);
+	intel_uncore_write(uncore,
+			   GEN10_PAT_INDEX(3),
+			   GEN8_PPAT_UC);
+	intel_uncore_write(uncore,
+			   GEN10_PAT_INDEX(4),
+			   GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(0));
+	intel_uncore_write(uncore,
+			   GEN10_PAT_INDEX(5),
+			   GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(1));
+	intel_uncore_write(uncore,
+			   GEN10_PAT_INDEX(6),
+			   GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(2));
+	intel_uncore_write(uncore,
+			   GEN10_PAT_INDEX(7),
+			   GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(3));
 }
 
 /* The GGTT and PPGTT need a private PPAT setup in order to handle cacheability
  * bits. When using advanced contexts each context stores its own PAT, but
  * writing this data shouldn't be harmful even in those cases. */
-static void bdw_setup_private_ppat(struct drm_i915_private *dev_priv)
+static void bdw_setup_private_ppat(struct intel_uncore *uncore)
 {
 	u64 pat;
 
@@ -2914,11 +2927,11 @@ static void bdw_setup_private_ppat(struct drm_i915_private *dev_priv)
 	      GEN8_PPAT(6, GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(2)) |
 	      GEN8_PPAT(7, GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(3));
 
-	I915_WRITE(GEN8_PRIVATE_PAT_LO, lower_32_bits(pat));
-	I915_WRITE(GEN8_PRIVATE_PAT_HI, upper_32_bits(pat));
+	intel_uncore_write(uncore, GEN8_PRIVATE_PAT_LO, lower_32_bits(pat));
+	intel_uncore_write(uncore, GEN8_PRIVATE_PAT_HI, upper_32_bits(pat));
 }
 
-static void chv_setup_private_ppat(struct drm_i915_private *dev_priv)
+static void chv_setup_private_ppat(struct intel_uncore *uncore)
 {
 	u64 pat;
 
@@ -2950,8 +2963,8 @@ static void chv_setup_private_ppat(struct drm_i915_private *dev_priv)
 	      GEN8_PPAT(6, CHV_PPAT_SNOOP) |
 	      GEN8_PPAT(7, CHV_PPAT_SNOOP);
 
-	I915_WRITE(GEN8_PRIVATE_PAT_LO, lower_32_bits(pat));
-	I915_WRITE(GEN8_PRIVATE_PAT_HI, upper_32_bits(pat));
+	intel_uncore_write(uncore, GEN8_PRIVATE_PAT_LO, lower_32_bits(pat));
+	intel_uncore_write(uncore, GEN8_PRIVATE_PAT_HI, upper_32_bits(pat));
 }
 
 static void gen6_gmch_remove(struct i915_address_space *vm)
@@ -2962,18 +2975,26 @@ static void gen6_gmch_remove(struct i915_address_space *vm)
 	cleanup_scratch_page(vm);
 }
 
-static void setup_private_pat(struct drm_i915_private *dev_priv)
+static void setup_private_pat(struct intel_uncore *uncore)
 {
-	GEM_BUG_ON(INTEL_GEN(dev_priv) < 8);
+	struct drm_i915_private *i915 = uncore->i915;
 
-	if (INTEL_GEN(dev_priv) >= 12)
-		tgl_setup_private_ppat(dev_priv);
-	else if (INTEL_GEN(dev_priv) >= 10)
-		cnl_setup_private_ppat(dev_priv);
-	else if (IS_CHERRYVIEW(dev_priv) || IS_GEN9_LP(dev_priv))
-		chv_setup_private_ppat(dev_priv);
+	GEM_BUG_ON(INTEL_GEN(i915) < 8);
+
+	if (INTEL_GEN(i915) >= 12)
+		tgl_setup_private_ppat(uncore);
+	else if (INTEL_GEN(i915) >= 10)
+		cnl_setup_private_ppat(uncore);
+	else if (IS_CHERRYVIEW(i915) || IS_GEN9_LP(i915))
+		chv_setup_private_ppat(uncore);
 	else
-		bdw_setup_private_ppat(dev_priv);
+		bdw_setup_private_ppat(uncore);
+}
+
+static struct resource pci_resource(struct pci_dev *pdev, int bar)
+{
+	return (struct resource)DEFINE_RES_MEM(pci_resource_start(pdev, bar),
+					       pci_resource_len(pdev, bar));
 }
 
 static int gen8_gmch_probe(struct i915_ggtt *ggtt)
@@ -2985,10 +3006,10 @@ static int gen8_gmch_probe(struct i915_ggtt *ggtt)
 	int err;
 
 	/* TODO: We're not aware of mappable constraints on gen8 yet */
-	ggtt->gmadr =
-		(struct resource) DEFINE_RES_MEM(pci_resource_start(pdev, 2),
-						 pci_resource_len(pdev, 2));
-	ggtt->mappable_end = resource_size(&ggtt->gmadr);
+	if (!IS_DGFX(dev_priv)) {
+		ggtt->gmadr = pci_resource(pdev, 2);
+		ggtt->mappable_end = resource_size(&ggtt->gmadr);
+	}
 
 	err = pci_set_dma_mask(pdev, DMA_BIT_MASK(39));
 	if (!err)
@@ -3029,7 +3050,7 @@ static int gen8_gmch_probe(struct i915_ggtt *ggtt)
 
 	ggtt->vm.pte_encode = gen8_pte_encode;
 
-	setup_private_pat(dev_priv);
+	setup_private_pat(ggtt->vm.gt->uncore);
 
 	return ggtt_probe_common(ggtt, size);
 }
@@ -3200,9 +3221,6 @@ int i915_ggtt_probe_hw(struct drm_i915_private *i915)
 static int ggtt_init_hw(struct i915_ggtt *ggtt)
 {
 	struct drm_i915_private *i915 = ggtt->vm.i915;
-	int ret = 0;
-
-	mutex_lock(&i915->drm.struct_mutex);
 
 	i915_address_space_init(&ggtt->vm, VM_CLASS_GGTT);
 
@@ -3212,24 +3230,23 @@ static int ggtt_init_hw(struct i915_ggtt *ggtt)
 	ggtt->vm.has_read_only = IS_VALLEYVIEW(i915);
 
 	if (!HAS_LLC(i915) && !HAS_PPGTT(i915))
-		ggtt->vm.mm.color_adjust = i915_gtt_color_adjust;
-
-	if (!io_mapping_init_wc(&ggtt->iomap,
-				ggtt->gmadr.start,
-				ggtt->mappable_end)) {
-		ggtt->vm.cleanup(&ggtt->vm);
-		ret = -EIO;
-		goto out;
-	}
+		ggtt->vm.mm.color_adjust = i915_ggtt_color_adjust;
+
+	if (ggtt->mappable_end) {
+		if (!io_mapping_init_wc(&ggtt->iomap,
+					ggtt->gmadr.start,
+					ggtt->mappable_end)) {
+			ggtt->vm.cleanup(&ggtt->vm);
+			return -EIO;
+		}
 
-	ggtt->mtrr = arch_phys_wc_add(ggtt->gmadr.start, ggtt->mappable_end);
+		ggtt->mtrr = arch_phys_wc_add(ggtt->gmadr.start,
+					      ggtt->mappable_end);
+	}
 
 	i915_ggtt_init_fences(ggtt);
 
-out:
-	mutex_unlock(&i915->drm.struct_mutex);
-
-	return ret;
+	return 0;
 }
 
 /**
@@ -3251,19 +3268,7 @@ int i915_ggtt_init_hw(struct drm_i915_private *dev_priv)
 	if (ret)
 		return ret;
 
-	/*
-	 * Initialise stolen early so that we may reserve preallocated
-	 * objects for the BIOS to KMS transition.
-	 */
-	ret = i915_gem_init_stolen(dev_priv);
-	if (ret)
-		goto out_gtt_cleanup;
-
 	return 0;
-
-out_gtt_cleanup:
-	dev_priv->ggtt.vm.cleanup(&dev_priv->ggtt.vm);
-	return ret;
 }
 
 int i915_ggtt_enable_hw(struct drm_i915_private *dev_priv)
@@ -3301,6 +3306,7 @@ static void ggtt_restore_mappings(struct i915_ggtt *ggtt)
 {
 	struct i915_vma *vma, *vn;
 	bool flush = false;
+	int open;
 
 	intel_gt_check_and_clear_faults(ggtt->vm.gt);
 
@@ -3308,33 +3314,31 @@ static void ggtt_restore_mappings(struct i915_ggtt *ggtt)
 
 	/* First fill our portion of the GTT with scratch pages */
 	ggtt->vm.clear_range(&ggtt->vm, 0, ggtt->vm.total);
-	ggtt->vm.closed = true; /* skip rewriting PTE on VMA unbind */
+
+	/* Skip rewriting PTE on VMA unbind. */
+	open = atomic_xchg(&ggtt->vm.open, 0);
 
 	/* clflush objects bound into the GGTT and rebind them. */
 	list_for_each_entry_safe(vma, vn, &ggtt->vm.bound_list, vm_link) {
 		struct drm_i915_gem_object *obj = vma->obj;
 
-		if (!(vma->flags & I915_VMA_GLOBAL_BIND))
+		if (!i915_vma_is_bound(vma, I915_VMA_GLOBAL_BIND))
 			continue;
 
-		mutex_unlock(&ggtt->vm.mutex);
-
-		if (!i915_vma_unbind(vma))
-			goto lock;
+		if (!__i915_vma_unbind(vma))
+			continue;
 
+		clear_bit(I915_VMA_GLOBAL_BIND_BIT, __i915_vma_flags(vma));
 		WARN_ON(i915_vma_bind(vma,
 				      obj ? obj->cache_level : 0,
-				      PIN_UPDATE));
+				      PIN_GLOBAL, NULL));
 		if (obj) { /* only used during resume => exclusive access */
 			flush |= fetch_and_zero(&obj->write_domain);
 			obj->read_domains |= I915_GEM_DOMAIN_GTT;
 		}
-
-lock:
-		mutex_lock(&ggtt->vm.mutex);
 	}
 
-	ggtt->vm.closed = false;
+	atomic_set(&ggtt->vm.open, open);
 	ggtt->invalidate(ggtt);
 
 	mutex_unlock(&ggtt->vm.mutex);
@@ -3345,10 +3349,12 @@ lock:
 
 void i915_gem_restore_gtt_mappings(struct drm_i915_private *i915)
 {
-	ggtt_restore_mappings(&i915->ggtt);
+	struct i915_ggtt *ggtt = &i915->ggtt;
+
+	ggtt_restore_mappings(ggtt);
 
 	if (INTEL_GEN(i915) >= 8)
-		setup_private_pat(i915);
+		setup_private_pat(ggtt->vm.gt->uncore);
 }
 
 static struct scatterlist *
@@ -3726,7 +3732,8 @@ int i915_gem_gtt_insert(struct i915_address_space *vm,
 	u64 offset;
 	int err;
 
-	lockdep_assert_held(&vm->i915->drm.struct_mutex);
+	lockdep_assert_held(&vm->mutex);
+
 	GEM_BUG_ON(!size);
 	GEM_BUG_ON(!IS_ALIGNED(size, I915_GTT_PAGE_SIZE));
 	GEM_BUG_ON(alignment && !is_power_of_2(alignment));
diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.h b/drivers/gpu/drm/i915/i915_gem_gtt.h
index b97a47fc7a68..402283ce2864 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.h
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.h
@@ -148,8 +148,8 @@ typedef u64 gen8_pte_t;
 #define GEN8_PDE_IPS_64K BIT(11)
 #define GEN8_PDE_PS_2M   BIT(7)
 
-#define for_each_sgt_dma(__dmap, __iter, __sgt) \
-	__for_each_sgt_dma(__dmap, __iter, __sgt, I915_GTT_PAGE_SIZE)
+#define for_each_sgt_daddr(__dp, __iter, __sgt) \
+	__for_each_sgt_daddr(__dp, __iter, __sgt, I915_GTT_PAGE_SIZE)
 
 struct intel_remapped_plane_info {
 	/* in gtt pages */
@@ -305,7 +305,16 @@ struct i915_address_space {
 	u64 total;		/* size addr space maps (ex. 2GB for ggtt) */
 	u64 reserved;		/* size addr space reserved */
 
-	bool closed;
+	unsigned int bind_async_flags;
+
+	/*
+	 * Each active user context has its own address space (in full-ppgtt).
+	 * Since the vm may be shared between multiple contexts, we count how
+	 * many contexts keep us "open". Once open hits zero, we are closed
+	 * and do not allow any new attachments, and proceed to shutdown our
+	 * vma and page directories.
+	 */
+	atomic_t open;
 
 	struct mutex mutex; /* protects vma and our lists */
 #define VM_CLASS_GGTT 0
@@ -320,11 +329,6 @@ struct i915_address_space {
 	 */
 	struct list_head bound_list;
 
-	/**
-	 * List of vma that are not unbound.
-	 */
-	struct list_head unbound_list;
-
 	struct pagestash free_pages;
 
 	/* Global GTT */
@@ -376,6 +380,12 @@ i915_vm_has_scratch_64K(struct i915_address_space *vm)
 	return vm->scratch_order == get_order(I915_GTT_PAGE_SIZE_64K);
 }
 
+static inline bool
+i915_vm_has_cache_coloring(struct i915_address_space *vm)
+{
+	return i915_is_ggtt(vm) && vm->mm.color_adjust;
+}
+
 /* The Graphics Translation Table is the way in which GEN hardware translates a
  * Graphics Virtual Address into a Physical Address. In addition to the normal
  * collateral associated with any va->pa translations GEN hardware also has a
@@ -401,6 +411,11 @@ struct i915_ggtt {
 
 	int mtrr;
 
+	/** Bit 6 swizzling required for X tiling */
+	u32 bit_6_swizzle_x;
+	/** Bit 6 swizzling required for Y tiling */
+	u32 bit_6_swizzle_y;
+
 	u32 pin_bias;
 
 	unsigned int num_fences;
@@ -422,7 +437,6 @@ struct i915_ggtt {
 struct i915_ppgtt {
 	struct i915_address_space vm;
 
-	intel_engine_mask_t pd_dirty_engines;
 	struct i915_page_directory *pd;
 };
 
@@ -432,7 +446,9 @@ struct gen6_ppgtt {
 	struct i915_vma *vma;
 	gen6_pte_t __iomem *pd_addr;
 
-	unsigned int pin_count;
+	atomic_t pin_count;
+	struct mutex pin_mutex;
+
 	bool scan_for_unused_pt;
 };
 
@@ -559,6 +575,11 @@ void i915_ggtt_disable_guc(struct i915_ggtt *ggtt);
 int i915_init_ggtt(struct drm_i915_private *dev_priv);
 void i915_ggtt_driver_release(struct drm_i915_private *dev_priv);
 
+static inline bool i915_ggtt_has_aperture(const struct i915_ggtt *ggtt)
+{
+	return ggtt->mappable_end > 0;
+}
+
 int i915_ppgtt_init_hw(struct intel_gt *gt);
 
 struct i915_ppgtt *i915_ppgtt_create(struct drm_i915_private *dev_priv);
@@ -577,6 +598,35 @@ static inline void i915_vm_put(struct i915_address_space *vm)
 	kref_put(&vm->ref, i915_vm_release);
 }
 
+static inline struct i915_address_space *
+i915_vm_open(struct i915_address_space *vm)
+{
+	GEM_BUG_ON(!atomic_read(&vm->open));
+	atomic_inc(&vm->open);
+	return i915_vm_get(vm);
+}
+
+static inline bool
+i915_vm_tryopen(struct i915_address_space *vm)
+{
+	if (atomic_add_unless(&vm->open, 1, 0))
+		return i915_vm_get(vm);
+
+	return false;
+}
+
+void __i915_vm_close(struct i915_address_space *vm);
+
+static inline void
+i915_vm_close(struct i915_address_space *vm)
+{
+	GEM_BUG_ON(!atomic_read(&vm->open));
+	if (atomic_dec_and_test(&vm->open))
+		__i915_vm_close(vm);
+
+	i915_vm_put(vm);
+}
+
 int gen6_ppgtt_pin(struct i915_ppgtt *base);
 void gen6_ppgtt_unpin(struct i915_ppgtt *base);
 void gen6_ppgtt_unpin_all(struct i915_ppgtt *base);
@@ -609,10 +659,9 @@ int i915_gem_gtt_insert(struct i915_address_space *vm,
 #define PIN_OFFSET_BIAS		BIT_ULL(6)
 #define PIN_OFFSET_FIXED	BIT_ULL(7)
 
-#define PIN_MBZ			BIT_ULL(8) /* I915_VMA_PIN_OVERFLOW */
-#define PIN_GLOBAL		BIT_ULL(9) /* I915_VMA_GLOBAL_BIND */
-#define PIN_USER		BIT_ULL(10) /* I915_VMA_LOCAL_BIND */
-#define PIN_UPDATE		BIT_ULL(11)
+#define PIN_UPDATE		BIT_ULL(9)
+#define PIN_GLOBAL		BIT_ULL(10) /* I915_VMA_GLOBAL_BIND */
+#define PIN_USER		BIT_ULL(11) /* I915_VMA_LOCAL_BIND */
 
 #define PIN_OFFSET_MASK		(-I915_GTT_PAGE_SIZE)
 
diff --git a/drivers/gpu/drm/i915/i915_getparam.c b/drivers/gpu/drm/i915/i915_getparam.c
index 5d9101376a3d..cf8a8c3ef047 100644
--- a/drivers/gpu/drm/i915/i915_getparam.c
+++ b/drivers/gpu/drm/i915/i915_getparam.c
@@ -5,6 +5,7 @@
 #include "gt/intel_engine_user.h"
 
 #include "i915_drv.h"
+#include "i915_perf.h"
 
 int i915_getparam_ioctl(struct drm_device *dev, void *data,
 			struct drm_file *file_priv)
@@ -62,7 +63,7 @@ int i915_getparam_ioctl(struct drm_device *dev, void *data,
 		value = !!(i915->caps.scheduler & I915_SCHEDULER_CAP_SEMAPHORES);
 		break;
 	case I915_PARAM_HAS_SECURE_BATCHES:
-		value = capable(CAP_SYS_ADMIN);
+		value = HAS_SECURE_BATCHES(i915) && capable(CAP_SYS_ADMIN);
 		break;
 	case I915_PARAM_CMD_PARSER_VERSION:
 		value = i915_cmd_parser_get_version(i915);
@@ -79,8 +80,8 @@ int i915_getparam_ioctl(struct drm_device *dev, void *data,
 		break;
 	case I915_PARAM_HAS_GPU_RESET:
 		value = i915_modparams.enable_hangcheck &&
-			intel_has_gpu_reset(i915);
-		if (value && intel_has_reset_engine(i915))
+			intel_has_gpu_reset(&i915->gt);
+		if (value && intel_has_reset_engine(&i915->gt))
 			value = 2;
 		break;
 	case I915_PARAM_HAS_RESOURCE_STREAMER:
@@ -156,6 +157,9 @@ int i915_getparam_ioctl(struct drm_device *dev, void *data,
 	case I915_PARAM_MMAP_GTT_COHERENT:
 		value = INTEL_INFO(i915)->has_coherent_ggtt;
 		break;
+	case I915_PARAM_PERF_REVISION:
+		value = i915_perf_ioctl_version();
+		break;
 	default:
 		DRM_DEBUG("Unknown parameter %d\n", param->param);
 		return -EINVAL;
diff --git a/drivers/gpu/drm/i915/i915_gpu_error.c b/drivers/gpu/drm/i915/i915_gpu_error.c
index e284bd76fa86..3c85cb0ee99f 100644
--- a/drivers/gpu/drm/i915/i915_gpu_error.c
+++ b/drivers/gpu/drm/i915/i915_gpu_error.c
@@ -40,6 +40,7 @@
 #include "display/intel_overlay.h"
 
 #include "gem/i915_gem_context.h"
+#include "gem/i915_gem_lmem.h"
 
 #include "i915_drv.h"
 #include "i915_gpu_error.h"
@@ -235,6 +236,7 @@ struct compress {
 	struct pagevec pool;
 	struct z_stream_s zstream;
 	void *tmp;
+	bool wc;
 };
 
 static bool compress_init(struct compress *c)
@@ -292,7 +294,7 @@ static int compress_page(struct compress *c,
 	struct z_stream_s *zstream = &c->zstream;
 
 	zstream->next_in = src;
-	if (c->tmp && i915_memcpy_from_wc(c->tmp, src, PAGE_SIZE))
+	if (c->wc && c->tmp && i915_memcpy_from_wc(c->tmp, src, PAGE_SIZE))
 		zstream->next_in = c->tmp;
 	zstream->avail_in = PAGE_SIZE;
 
@@ -367,6 +369,7 @@ static void err_compression_marker(struct drm_i915_error_state_buf *m)
 
 struct compress {
 	struct pagevec pool;
+	bool wc;
 };
 
 static bool compress_init(struct compress *c)
@@ -389,7 +392,7 @@ static int compress_page(struct compress *c,
 	if (!ptr)
 		return -ENOMEM;
 
-	if (!i915_memcpy_from_wc(ptr, src, PAGE_SIZE))
+	if (!(c->wc && i915_memcpy_from_wc(ptr, src, PAGE_SIZE)))
 		memcpy(ptr, src, PAGE_SIZE);
 	dst->pages[dst->page_count++] = ptr;
 
@@ -421,6 +424,7 @@ static void err_compression_marker(struct drm_i915_error_state_buf *m)
 static void error_print_instdone(struct drm_i915_error_state_buf *m,
 				 const struct drm_i915_error_engine *ee)
 {
+	const struct sseu_dev_info *sseu = &RUNTIME_INFO(m->i915)->sseu;
 	int slice;
 	int subslice;
 
@@ -436,12 +440,12 @@ static void error_print_instdone(struct drm_i915_error_state_buf *m,
 	if (INTEL_GEN(m->i915) <= 6)
 		return;
 
-	for_each_instdone_slice_subslice(m->i915, slice, subslice)
+	for_each_instdone_slice_subslice(m->i915, sseu, slice, subslice)
 		err_printf(m, "  SAMPLER_INSTDONE[%d][%d]: 0x%08x\n",
 			   slice, subslice,
 			   ee->instdone.sampler[slice][subslice]);
 
-	for_each_instdone_slice_subslice(m->i915, slice, subslice)
+	for_each_instdone_slice_subslice(m->i915, sseu, slice, subslice)
 		err_printf(m, "  ROW_INSTDONE[%d][%d]: 0x%08x\n",
 			   slice, subslice,
 			   ee->instdone.row[slice][subslice]);
@@ -470,9 +474,9 @@ static void error_print_context(struct drm_i915_error_state_buf *m,
 				const char *header,
 				const struct drm_i915_error_context *ctx)
 {
-	err_printf(m, "%s%s[%d] hw_id %d, prio %d, guilty %d active %d\n",
-		   header, ctx->comm, ctx->pid, ctx->hw_id,
-		   ctx->sched_attr.priority, ctx->guilty, ctx->active);
+	err_printf(m, "%s%s[%d] prio %d, guilty %d active %d\n",
+		   header, ctx->comm, ctx->pid, ctx->sched_attr.priority,
+		   ctx->guilty, ctx->active);
 }
 
 static void error_print_engine(struct drm_i915_error_state_buf *m,
@@ -533,10 +537,6 @@ static void error_print_engine(struct drm_i915_error_state_buf *m,
 	}
 	err_printf(m, "  ring->head: 0x%08x\n", ee->cpu_ring_head);
 	err_printf(m, "  ring->tail: 0x%08x\n", ee->cpu_ring_tail);
-	err_printf(m, "  hangcheck timestamp: %dms (%lu%s)\n",
-		   jiffies_to_msecs(ee->hangcheck_timestamp - epoch),
-		   ee->hangcheck_timestamp,
-		   ee->hangcheck_timestamp == epoch ? "; epoch" : "");
 	err_printf(m, "  engine reset count: %u\n", ee->reset_count);
 
 	for (n = 0; n < ee->num_ports; n++) {
@@ -574,6 +574,9 @@ static void print_error_obj(struct drm_i915_error_state_buf *m,
 			   lower_32_bits(obj->gtt_offset));
 	}
 
+	if (obj->gtt_page_sizes > I915_GTT_PAGE_SIZE_4K)
+		err_printf(m, "gtt_page_sizes = 0x%08x\n", obj->gtt_page_sizes);
+
 	err_compression_marker(m);
 	for (page = 0; page < obj->page_count; page++) {
 		int i, len;
@@ -675,11 +678,8 @@ static void __err_print_to_sgl(struct drm_i915_error_state_buf *m,
 	ts = ktime_to_timespec64(error->uptime);
 	err_printf(m, "Uptime: %lld s %ld us\n",
 		   (s64)ts.tv_sec, ts.tv_nsec / NSEC_PER_USEC);
-	err_printf(m, "Epoch: %lu jiffies (%u HZ)\n", error->epoch, HZ);
-	err_printf(m, "Capture: %lu jiffies; %d ms ago, %d ms after epoch\n",
-		   error->capture,
-		   jiffies_to_msecs(jiffies - error->capture),
-		   jiffies_to_msecs(error->capture - error->epoch));
+	err_printf(m, "Capture: %lu jiffies; %d ms ago\n",
+		   error->capture, jiffies_to_msecs(jiffies - error->capture));
 
 	for (ee = error->engine; ee; ee = ee->next)
 		err_printf(m, "Active process (on ring %s): %s [%d]\n",
@@ -734,8 +734,24 @@ static void __err_print_to_sgl(struct drm_i915_error_state_buf *m,
 	if (IS_GEN(m->i915, 7))
 		err_printf(m, "ERR_INT: 0x%08x\n", error->err_int);
 
+	if (IS_GEN_RANGE(m->i915, 8, 11))
+		err_printf(m, "GTT_CACHE_EN: 0x%08x\n", error->gtt_cache);
+
+	if (IS_GEN(m->i915, 12))
+		err_printf(m, "AUX_ERR_DBG: 0x%08x\n", error->aux_err);
+
+	if (INTEL_GEN(m->i915) >= 12) {
+		int i;
+
+		for (i = 0; i < GEN12_SFC_DONE_MAX; i++)
+			err_printf(m, "  SFC_DONE[%d]: 0x%08x\n", i,
+				   error->sfc_done[i]);
+
+		err_printf(m, "  GAM_DONE: 0x%08x\n", error->gam_done);
+	}
+
 	for (ee = error->engine; ee; ee = ee->next)
-		error_print_engine(m, ee, error->epoch);
+		error_print_engine(m, ee, error->capture);
 
 	for (ee = error->engine; ee; ee = ee->next) {
 		const struct drm_i915_error_object *obj;
@@ -763,7 +779,7 @@ static void __err_print_to_sgl(struct drm_i915_error_state_buf *m,
 			for (j = 0; j < ee->num_requests; j++)
 				error_print_request(m, " ",
 						    &ee->requests[j],
-						    error->epoch);
+						    error->capture);
 		}
 
 		print_error_obj(m, ee->engine, "ringbuffer", ee->ringbuffer);
@@ -963,7 +979,6 @@ i915_error_object_create(struct drm_i915_private *i915,
 	struct drm_i915_error_object *dst;
 	unsigned long num_pages;
 	struct sgt_iter iter;
-	dma_addr_t dma;
 	int ret;
 
 	might_sleep();
@@ -984,21 +999,59 @@ i915_error_object_create(struct drm_i915_private *i915,
 
 	dst->gtt_offset = vma->node.start;
 	dst->gtt_size = vma->node.size;
+	dst->gtt_page_sizes = vma->page_sizes.gtt;
 	dst->num_pages = num_pages;
 	dst->page_count = 0;
 	dst->unused = 0;
 
+	compress->wc = i915_gem_object_is_lmem(vma->obj) ||
+		       drm_mm_node_allocated(&ggtt->error_capture);
+
 	ret = -EINVAL;
-	for_each_sgt_dma(dma, iter, vma->pages) {
+	if (drm_mm_node_allocated(&ggtt->error_capture)) {
 		void __iomem *s;
+		dma_addr_t dma;
 
-		ggtt->vm.insert_page(&ggtt->vm, dma, slot, I915_CACHE_NONE, 0);
+		for_each_sgt_daddr(dma, iter, vma->pages) {
+			ggtt->vm.insert_page(&ggtt->vm, dma, slot,
+					     I915_CACHE_NONE, 0);
 
-		s = io_mapping_map_wc(&ggtt->iomap, slot, PAGE_SIZE);
-		ret = compress_page(compress, (void  __force *)s, dst);
-		io_mapping_unmap(s);
-		if (ret)
-			break;
+			s = io_mapping_map_wc(&ggtt->iomap, slot, PAGE_SIZE);
+			ret = compress_page(compress, (void  __force *)s, dst);
+			io_mapping_unmap(s);
+			if (ret)
+				break;
+		}
+	} else if (i915_gem_object_is_lmem(vma->obj)) {
+		struct intel_memory_region *mem = vma->obj->mm.region;
+		dma_addr_t dma;
+
+		for_each_sgt_daddr(dma, iter, vma->pages) {
+			void __iomem *s;
+
+			s = io_mapping_map_wc(&mem->iomap, dma, PAGE_SIZE);
+			ret = compress_page(compress, (void __force *)s, dst);
+			io_mapping_unmap(s);
+			if (ret)
+				break;
+		}
+	} else {
+		struct page *page;
+
+		for_each_sgt_page(page, iter, vma->pages) {
+			void *s;
+
+			drm_clflush_pages(&page, 1);
+
+			s = kmap(page);
+			ret = compress_page(compress, s, dst);
+			kunmap(s);
+
+			drm_clflush_pages(&page, 1);
+
+			if (ret)
+				break;
+		}
 	}
 
 	if (ret || compress_flush(compress, dst)) {
@@ -1136,8 +1189,6 @@ static void error_record_engine_registers(struct i915_gpu_state *error,
 	}
 
 	ee->idle = intel_engine_is_idle(engine);
-	if (!ee->idle)
-		ee->hangcheck_timestamp = engine->hangcheck.action_timestamp;
 	ee->reset_count = i915_reset_engine_count(&dev_priv->gpu_error,
 						  engine);
 
@@ -1263,7 +1314,6 @@ static bool record_context(struct drm_i915_error_context *e,
 		rcu_read_unlock();
 	}
 
-	e->hw_id = ctx->hw_id;
 	e->sched_attr = ctx->sched;
 	e->guilty = atomic_read(&ctx->guilty_count);
 	e->active = atomic_read(&ctx->active_count);
@@ -1291,7 +1341,7 @@ capture_vma(struct capture_vma *next,
 	if (!c)
 		return next;
 
-	if (!i915_active_trygrab(&vma->active)) {
+	if (!i915_active_acquire_if_busy(&vma->active)) {
 		kfree(c);
 		return next;
 	}
@@ -1431,7 +1481,7 @@ gem_record_rings(struct i915_gpu_state *error, struct compress *compress)
 			*this->slot =
 				i915_error_object_create(i915, vma, compress);
 
-			i915_active_ungrab(&vma->active);
+			i915_active_release(&vma->active);
 			i915_vma_put(vma);
 
 			capture = this->next;
@@ -1553,6 +1603,21 @@ static void capture_reg_state(struct i915_gpu_state *error)
 		error->gac_eco = intel_uncore_read(uncore, GAC_ECO_BITS);
 	}
 
+	if (IS_GEN_RANGE(i915, 8, 11))
+		error->gtt_cache = intel_uncore_read(uncore, HSW_GTT_CACHE_EN);
+
+	if (IS_GEN(i915, 12))
+		error->aux_err = intel_uncore_read(uncore, GEN12_AUX_ERR_DBG);
+
+	if (INTEL_GEN(i915) >= 12) {
+		for (i = 0; i < GEN12_SFC_DONE_MAX; i++) {
+			error->sfc_done[i] =
+				intel_uncore_read(uncore, GEN12_SFC_DONE(i));
+		}
+
+		error->gam_done = intel_uncore_read(uncore, GEN12_GAM_DONE);
+	}
+
 	/* 4: Everything else */
 	if (INTEL_GEN(i915) >= 11) {
 		error->ier = intel_uncore_read(uncore, GEN8_DE_MISC_IER);
@@ -1647,26 +1712,15 @@ static void capture_params(struct i915_gpu_state *error)
 	i915_params_copy(&error->params, &i915_modparams);
 }
 
-static unsigned long capture_find_epoch(const struct i915_gpu_state *error)
-{
-	const struct drm_i915_error_engine *ee;
-	unsigned long epoch = error->capture;
-
-	for (ee = error->engine; ee; ee = ee->next) {
-		if (ee->hangcheck_timestamp &&
-		    time_before(ee->hangcheck_timestamp, epoch))
-			epoch = ee->hangcheck_timestamp;
-	}
-
-	return epoch;
-}
-
 static void capture_finish(struct i915_gpu_state *error)
 {
 	struct i915_ggtt *ggtt = &error->i915->ggtt;
-	const u64 slot = ggtt->error_capture.start;
 
-	ggtt->vm.clear_range(&ggtt->vm, slot, PAGE_SIZE);
+	if (drm_mm_node_allocated(&ggtt->error_capture)) {
+		const u64 slot = ggtt->error_capture.start;
+
+		ggtt->vm.clear_range(&ggtt->vm, slot, PAGE_SIZE);
+	}
 }
 
 #define DAY_AS_SECONDS(x) (24 * 60 * 60 * (x))
@@ -1712,8 +1766,6 @@ i915_capture_gpu_state(struct drm_i915_private *i915)
 	error->overlay = intel_overlay_capture_error_state(i915);
 	error->display = intel_display_capture_error_state(i915);
 
-	error->epoch = capture_find_epoch(error);
-
 	capture_finish(error);
 	compress_fini(&compress);
 
diff --git a/drivers/gpu/drm/i915/i915_gpu_error.h b/drivers/gpu/drm/i915/i915_gpu_error.h
index df9f57766626..5d2c3372ff99 100644
--- a/drivers/gpu/drm/i915/i915_gpu_error.h
+++ b/drivers/gpu/drm/i915/i915_gpu_error.h
@@ -34,7 +34,6 @@ struct i915_gpu_state {
 	ktime_t boottime;
 	ktime_t uptime;
 	unsigned long capture;
-	unsigned long epoch;
 
 	struct drm_i915_private *i915;
 
@@ -74,6 +73,10 @@ struct i915_gpu_state {
 	u32 gam_ecochk;
 	u32 gab_ctl;
 	u32 gfx_mode;
+	u32 gtt_cache;
+	u32 aux_err; /* gen12 */
+	u32 sfc_done[GEN12_SFC_DONE_MAX]; /* gen12 */
+	u32 gam_done; /* gen12 */
 
 	u32 nfence;
 	u64 fence[I915_MAX_NUM_FENCES];
@@ -85,7 +88,6 @@ struct i915_gpu_state {
 
 		/* Software tracked state */
 		bool idle;
-		unsigned long hangcheck_timestamp;
 		int num_requests;
 		u32 reset_count;
 
@@ -118,7 +120,6 @@ struct i915_gpu_state {
 		struct drm_i915_error_context {
 			char comm[TASK_COMM_LEN];
 			pid_t pid;
-			u32 hw_id;
 			int active;
 			int guilty;
 			struct i915_sched_attr sched_attr;
@@ -127,6 +128,7 @@ struct i915_gpu_state {
 		struct drm_i915_error_object {
 			u64 gtt_offset;
 			u64 gtt_size;
+			u32 gtt_page_sizes;
 			int num_pages;
 			int page_count;
 			int unused;
diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c
index 37e3dd3c1a9d..dae00f7dd7df 100644
--- a/drivers/gpu/drm/i915/i915_irq.c
+++ b/drivers/gpu/drm/i915/i915_irq.c
@@ -29,7 +29,6 @@
 #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
 
 #include <linux/circ_buf.h>
-#include <linux/cpuidle.h>
 #include <linux/slab.h>
 #include <linux/sysrq.h>
 
@@ -46,6 +45,7 @@
 #include "gt/intel_gt.h"
 #include "gt/intel_gt_irq.h"
 #include "gt/intel_gt_pm_irq.h"
+#include "gt/intel_rps.h"
 
 #include "i915_drv.h"
 #include "i915_irq.h"
@@ -149,30 +149,24 @@ static const u32 hpd_gen12[HPD_NUM_PINS] = {
 };
 
 static const u32 hpd_icp[HPD_NUM_PINS] = {
-	[HPD_PORT_A] = SDE_DDIA_HOTPLUG_ICP,
-	[HPD_PORT_B] = SDE_DDIB_HOTPLUG_ICP,
-	[HPD_PORT_C] = SDE_TC1_HOTPLUG_ICP,
-	[HPD_PORT_D] = SDE_TC2_HOTPLUG_ICP,
-	[HPD_PORT_E] = SDE_TC3_HOTPLUG_ICP,
-	[HPD_PORT_F] = SDE_TC4_HOTPLUG_ICP
-};
-
-static const u32 hpd_mcc[HPD_NUM_PINS] = {
-	[HPD_PORT_A] = SDE_DDIA_HOTPLUG_ICP,
-	[HPD_PORT_B] = SDE_DDIB_HOTPLUG_ICP,
-	[HPD_PORT_C] = SDE_TC1_HOTPLUG_ICP
+	[HPD_PORT_A] = SDE_DDI_HOTPLUG_ICP(PORT_A),
+	[HPD_PORT_B] = SDE_DDI_HOTPLUG_ICP(PORT_B),
+	[HPD_PORT_C] = SDE_TC_HOTPLUG_ICP(PORT_TC1),
+	[HPD_PORT_D] = SDE_TC_HOTPLUG_ICP(PORT_TC2),
+	[HPD_PORT_E] = SDE_TC_HOTPLUG_ICP(PORT_TC3),
+	[HPD_PORT_F] = SDE_TC_HOTPLUG_ICP(PORT_TC4),
 };
 
 static const u32 hpd_tgp[HPD_NUM_PINS] = {
-	[HPD_PORT_A] = SDE_DDIA_HOTPLUG_ICP,
-	[HPD_PORT_B] = SDE_DDIB_HOTPLUG_ICP,
-	[HPD_PORT_C] = SDE_DDIC_HOTPLUG_TGP,
-	[HPD_PORT_D] = SDE_TC1_HOTPLUG_ICP,
-	[HPD_PORT_E] = SDE_TC2_HOTPLUG_ICP,
-	[HPD_PORT_F] = SDE_TC3_HOTPLUG_ICP,
-	[HPD_PORT_G] = SDE_TC4_HOTPLUG_ICP,
-	[HPD_PORT_H] = SDE_TC5_HOTPLUG_TGP,
-	[HPD_PORT_I] = SDE_TC6_HOTPLUG_TGP,
+	[HPD_PORT_A] = SDE_DDI_HOTPLUG_ICP(PORT_A),
+	[HPD_PORT_B] = SDE_DDI_HOTPLUG_ICP(PORT_B),
+	[HPD_PORT_C] = SDE_DDI_HOTPLUG_ICP(PORT_C),
+	[HPD_PORT_D] = SDE_TC_HOTPLUG_ICP(PORT_TC1),
+	[HPD_PORT_E] = SDE_TC_HOTPLUG_ICP(PORT_TC2),
+	[HPD_PORT_F] = SDE_TC_HOTPLUG_ICP(PORT_TC3),
+	[HPD_PORT_G] = SDE_TC_HOTPLUG_ICP(PORT_TC4),
+	[HPD_PORT_H] = SDE_TC_HOTPLUG_ICP(PORT_TC5),
+	[HPD_PORT_I] = SDE_TC_HOTPLUG_ICP(PORT_TC6),
 };
 
 void gen3_irq_reset(struct intel_uncore *uncore, i915_reg_t imr,
@@ -327,180 +321,6 @@ void ilk_update_display_irq(struct drm_i915_private *dev_priv,
 	}
 }
 
-static i915_reg_t gen6_pm_iir(struct drm_i915_private *dev_priv)
-{
-	WARN_ON_ONCE(INTEL_GEN(dev_priv) >= 11);
-
-	return INTEL_GEN(dev_priv) >= 8 ? GEN8_GT_IIR(2) : GEN6_PMIIR;
-}
-
-void gen11_reset_rps_interrupts(struct drm_i915_private *dev_priv)
-{
-	struct intel_gt *gt = &dev_priv->gt;
-
-	spin_lock_irq(&gt->irq_lock);
-
-	while (gen11_gt_reset_one_iir(gt, 0, GEN11_GTPM))
-		;
-
-	dev_priv->gt_pm.rps.pm_iir = 0;
-
-	spin_unlock_irq(&gt->irq_lock);
-}
-
-void gen6_reset_rps_interrupts(struct drm_i915_private *dev_priv)
-{
-	struct intel_gt *gt = &dev_priv->gt;
-
-	spin_lock_irq(&gt->irq_lock);
-	gen6_gt_pm_reset_iir(gt, GEN6_PM_RPS_EVENTS);
-	dev_priv->gt_pm.rps.pm_iir = 0;
-	spin_unlock_irq(&gt->irq_lock);
-}
-
-void gen6_enable_rps_interrupts(struct drm_i915_private *dev_priv)
-{
-	struct intel_gt *gt = &dev_priv->gt;
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
-
-	if (READ_ONCE(rps->interrupts_enabled))
-		return;
-
-	spin_lock_irq(&gt->irq_lock);
-	WARN_ON_ONCE(rps->pm_iir);
-
-	if (INTEL_GEN(dev_priv) >= 11)
-		WARN_ON_ONCE(gen11_gt_reset_one_iir(gt, 0, GEN11_GTPM));
-	else
-		WARN_ON_ONCE(I915_READ(gen6_pm_iir(dev_priv)) & dev_priv->pm_rps_events);
-
-	rps->interrupts_enabled = true;
-	gen6_gt_pm_enable_irq(gt, dev_priv->pm_rps_events);
-
-	spin_unlock_irq(&gt->irq_lock);
-}
-
-u32 gen6_sanitize_rps_pm_mask(const struct drm_i915_private *i915, u32 mask)
-{
-	return mask & ~i915->gt_pm.rps.pm_intrmsk_mbz;
-}
-
-void gen6_disable_rps_interrupts(struct drm_i915_private *dev_priv)
-{
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
-	struct intel_gt *gt = &dev_priv->gt;
-
-	if (!READ_ONCE(rps->interrupts_enabled))
-		return;
-
-	spin_lock_irq(&gt->irq_lock);
-	rps->interrupts_enabled = false;
-
-	I915_WRITE(GEN6_PMINTRMSK, gen6_sanitize_rps_pm_mask(dev_priv, ~0u));
-
-	gen6_gt_pm_disable_irq(gt, GEN6_PM_RPS_EVENTS);
-
-	spin_unlock_irq(&gt->irq_lock);
-	intel_synchronize_irq(dev_priv);
-
-	/* Now that we will not be generating any more work, flush any
-	 * outstanding tasks. As we are called on the RPS idle path,
-	 * we will reset the GPU to minimum frequencies, so the current
-	 * state of the worker can be discarded.
-	 */
-	cancel_work_sync(&rps->work);
-	if (INTEL_GEN(dev_priv) >= 11)
-		gen11_reset_rps_interrupts(dev_priv);
-	else
-		gen6_reset_rps_interrupts(dev_priv);
-}
-
-void gen9_reset_guc_interrupts(struct intel_guc *guc)
-{
-	struct intel_gt *gt = guc_to_gt(guc);
-
-	assert_rpm_wakelock_held(&gt->i915->runtime_pm);
-
-	spin_lock_irq(&gt->irq_lock);
-	gen6_gt_pm_reset_iir(gt, gt->pm_guc_events);
-	spin_unlock_irq(&gt->irq_lock);
-}
-
-void gen9_enable_guc_interrupts(struct intel_guc *guc)
-{
-	struct intel_gt *gt = guc_to_gt(guc);
-
-	assert_rpm_wakelock_held(&gt->i915->runtime_pm);
-
-	spin_lock_irq(&gt->irq_lock);
-	if (!guc->interrupts.enabled) {
-		WARN_ON_ONCE(intel_uncore_read(gt->uncore,
-					       gen6_pm_iir(gt->i915)) &
-			     gt->pm_guc_events);
-		guc->interrupts.enabled = true;
-		gen6_gt_pm_enable_irq(gt, gt->pm_guc_events);
-	}
-	spin_unlock_irq(&gt->irq_lock);
-}
-
-void gen9_disable_guc_interrupts(struct intel_guc *guc)
-{
-	struct intel_gt *gt = guc_to_gt(guc);
-
-	assert_rpm_wakelock_held(&gt->i915->runtime_pm);
-
-	spin_lock_irq(&gt->irq_lock);
-	guc->interrupts.enabled = false;
-
-	gen6_gt_pm_disable_irq(gt, gt->pm_guc_events);
-
-	spin_unlock_irq(&gt->irq_lock);
-	intel_synchronize_irq(gt->i915);
-
-	gen9_reset_guc_interrupts(guc);
-}
-
-void gen11_reset_guc_interrupts(struct intel_guc *guc)
-{
-	struct intel_gt *gt = guc_to_gt(guc);
-
-	spin_lock_irq(&gt->irq_lock);
-	gen11_gt_reset_one_iir(gt, 0, GEN11_GUC);
-	spin_unlock_irq(&gt->irq_lock);
-}
-
-void gen11_enable_guc_interrupts(struct intel_guc *guc)
-{
-	struct intel_gt *gt = guc_to_gt(guc);
-
-	spin_lock_irq(&gt->irq_lock);
-	if (!guc->interrupts.enabled) {
-		u32 events = REG_FIELD_PREP(ENGINE1_MASK, GUC_INTR_GUC2HOST);
-
-		WARN_ON_ONCE(gen11_gt_reset_one_iir(gt, 0, GEN11_GUC));
-		intel_uncore_write(gt->uncore, GEN11_GUC_SG_INTR_ENABLE, events);
-		intel_uncore_write(gt->uncore, GEN11_GUC_SG_INTR_MASK, ~events);
-		guc->interrupts.enabled = true;
-	}
-	spin_unlock_irq(&gt->irq_lock);
-}
-
-void gen11_disable_guc_interrupts(struct intel_guc *guc)
-{
-	struct intel_gt *gt = guc_to_gt(guc);
-
-	spin_lock_irq(&gt->irq_lock);
-	guc->interrupts.enabled = false;
-
-	intel_uncore_write(gt->uncore, GEN11_GUC_SG_INTR_MASK, ~0);
-	intel_uncore_write(gt->uncore, GEN11_GUC_SG_INTR_ENABLE, 0);
-
-	spin_unlock_irq(&gt->irq_lock);
-	intel_synchronize_irq(gt->i915);
-
-	gen11_reset_guc_interrupts(guc);
-}
-
 /**
  * bdw_update_port_irq - update DE port interrupt
  * @dev_priv: driver private
@@ -942,14 +762,14 @@ static int __intel_get_crtc_scanline(struct intel_crtc *crtc)
 	return (position + crtc->scanline_offset) % vtotal;
 }
 
-bool i915_get_crtc_scanoutpos(struct drm_device *dev, unsigned int pipe,
+bool i915_get_crtc_scanoutpos(struct drm_device *dev, unsigned int index,
 			      bool in_vblank_irq, int *vpos, int *hpos,
 			      ktime_t *stime, ktime_t *etime,
 			      const struct drm_display_mode *mode)
 {
 	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct intel_crtc *intel_crtc = intel_get_crtc_for_pipe(dev_priv,
-								pipe);
+	struct intel_crtc *crtc = to_intel_crtc(drm_crtc_from_index(dev, index));
+	enum pipe pipe = crtc->pipe;
 	int position;
 	int vbl_start, vbl_end, hsync_start, htotal, vtotal;
 	unsigned long irqflags;
@@ -992,7 +812,7 @@ bool i915_get_crtc_scanoutpos(struct drm_device *dev, unsigned int pipe,
 		/* No obvious pixelcount register. Only query vertical
 		 * scanout position from Display scan line register.
 		 */
-		position = __intel_get_crtc_scanline(intel_crtc);
+		position = __intel_get_crtc_scanline(crtc);
 	} else {
 		/* Have access to pixelcount since start of frame.
 		 * We can split this into vertical and horizontal
@@ -1072,199 +892,6 @@ int intel_get_crtc_scanline(struct intel_crtc *crtc)
 	return position;
 }
 
-static void ironlake_rps_change_irq_handler(struct drm_i915_private *dev_priv)
-{
-	struct intel_uncore *uncore = &dev_priv->uncore;
-	u32 busy_up, busy_down, max_avg, min_avg;
-	u8 new_delay;
-
-	spin_lock(&mchdev_lock);
-
-	intel_uncore_write16(uncore,
-			     MEMINTRSTS,
-			     intel_uncore_read(uncore, MEMINTRSTS));
-
-	new_delay = dev_priv->ips.cur_delay;
-
-	intel_uncore_write16(uncore, MEMINTRSTS, MEMINT_EVAL_CHG);
-	busy_up = intel_uncore_read(uncore, RCPREVBSYTUPAVG);
-	busy_down = intel_uncore_read(uncore, RCPREVBSYTDNAVG);
-	max_avg = intel_uncore_read(uncore, RCBMAXAVG);
-	min_avg = intel_uncore_read(uncore, RCBMINAVG);
-
-	/* Handle RCS change request from hw */
-	if (busy_up > max_avg) {
-		if (dev_priv->ips.cur_delay != dev_priv->ips.max_delay)
-			new_delay = dev_priv->ips.cur_delay - 1;
-		if (new_delay < dev_priv->ips.max_delay)
-			new_delay = dev_priv->ips.max_delay;
-	} else if (busy_down < min_avg) {
-		if (dev_priv->ips.cur_delay != dev_priv->ips.min_delay)
-			new_delay = dev_priv->ips.cur_delay + 1;
-		if (new_delay > dev_priv->ips.min_delay)
-			new_delay = dev_priv->ips.min_delay;
-	}
-
-	if (ironlake_set_drps(dev_priv, new_delay))
-		dev_priv->ips.cur_delay = new_delay;
-
-	spin_unlock(&mchdev_lock);
-
-	return;
-}
-
-static void vlv_c0_read(struct drm_i915_private *dev_priv,
-			struct intel_rps_ei *ei)
-{
-	ei->ktime = ktime_get_raw();
-	ei->render_c0 = I915_READ(VLV_RENDER_C0_COUNT);
-	ei->media_c0 = I915_READ(VLV_MEDIA_C0_COUNT);
-}
-
-void gen6_rps_reset_ei(struct drm_i915_private *dev_priv)
-{
-	memset(&dev_priv->gt_pm.rps.ei, 0, sizeof(dev_priv->gt_pm.rps.ei));
-}
-
-static u32 vlv_wa_c0_ei(struct drm_i915_private *dev_priv, u32 pm_iir)
-{
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
-	const struct intel_rps_ei *prev = &rps->ei;
-	struct intel_rps_ei now;
-	u32 events = 0;
-
-	if ((pm_iir & GEN6_PM_RP_UP_EI_EXPIRED) == 0)
-		return 0;
-
-	vlv_c0_read(dev_priv, &now);
-
-	if (prev->ktime) {
-		u64 time, c0;
-		u32 render, media;
-
-		time = ktime_us_delta(now.ktime, prev->ktime);
-
-		time *= dev_priv->czclk_freq;
-
-		/* Workload can be split between render + media,
-		 * e.g. SwapBuffers being blitted in X after being rendered in
-		 * mesa. To account for this we need to combine both engines
-		 * into our activity counter.
-		 */
-		render = now.render_c0 - prev->render_c0;
-		media = now.media_c0 - prev->media_c0;
-		c0 = max(render, media);
-		c0 *= 1000 * 100 << 8; /* to usecs and scale to threshold% */
-
-		if (c0 > time * rps->power.up_threshold)
-			events = GEN6_PM_RP_UP_THRESHOLD;
-		else if (c0 < time * rps->power.down_threshold)
-			events = GEN6_PM_RP_DOWN_THRESHOLD;
-	}
-
-	rps->ei = now;
-	return events;
-}
-
-static void gen6_pm_rps_work(struct work_struct *work)
-{
-	struct drm_i915_private *dev_priv =
-		container_of(work, struct drm_i915_private, gt_pm.rps.work);
-	struct intel_gt *gt = &dev_priv->gt;
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
-	bool client_boost = false;
-	int new_delay, adj, min, max;
-	u32 pm_iir = 0;
-
-	spin_lock_irq(&gt->irq_lock);
-	if (rps->interrupts_enabled) {
-		pm_iir = fetch_and_zero(&rps->pm_iir);
-		client_boost = atomic_read(&rps->num_waiters);
-	}
-	spin_unlock_irq(&gt->irq_lock);
-
-	/* Make sure we didn't queue anything we're not going to process. */
-	WARN_ON(pm_iir & ~dev_priv->pm_rps_events);
-	if ((pm_iir & dev_priv->pm_rps_events) == 0 && !client_boost)
-		goto out;
-
-	mutex_lock(&rps->lock);
-
-	pm_iir |= vlv_wa_c0_ei(dev_priv, pm_iir);
-
-	adj = rps->last_adj;
-	new_delay = rps->cur_freq;
-	min = rps->min_freq_softlimit;
-	max = rps->max_freq_softlimit;
-	if (client_boost)
-		max = rps->max_freq;
-	if (client_boost && new_delay < rps->boost_freq) {
-		new_delay = rps->boost_freq;
-		adj = 0;
-	} else if (pm_iir & GEN6_PM_RP_UP_THRESHOLD) {
-		if (adj > 0)
-			adj *= 2;
-		else /* CHV needs even encode values */
-			adj = IS_CHERRYVIEW(dev_priv) ? 2 : 1;
-
-		if (new_delay >= rps->max_freq_softlimit)
-			adj = 0;
-	} else if (client_boost) {
-		adj = 0;
-	} else if (pm_iir & GEN6_PM_RP_DOWN_TIMEOUT) {
-		if (rps->cur_freq > rps->efficient_freq)
-			new_delay = rps->efficient_freq;
-		else if (rps->cur_freq > rps->min_freq_softlimit)
-			new_delay = rps->min_freq_softlimit;
-		adj = 0;
-	} else if (pm_iir & GEN6_PM_RP_DOWN_THRESHOLD) {
-		if (adj < 0)
-			adj *= 2;
-		else /* CHV needs even encode values */
-			adj = IS_CHERRYVIEW(dev_priv) ? -2 : -1;
-
-		if (new_delay <= rps->min_freq_softlimit)
-			adj = 0;
-	} else { /* unknown event */
-		adj = 0;
-	}
-
-	rps->last_adj = adj;
-
-	/*
-	 * Limit deboosting and boosting to keep ourselves at the extremes
-	 * when in the respective power modes (i.e. slowly decrease frequencies
-	 * while in the HIGH_POWER zone and slowly increase frequencies while
-	 * in the LOW_POWER zone). On idle, we will hit the timeout and drop
-	 * to the next level quickly, and conversely if busy we expect to
-	 * hit a waitboost and rapidly switch into max power.
-	 */
-	if ((adj < 0 && rps->power.mode == HIGH_POWER) ||
-	    (adj > 0 && rps->power.mode == LOW_POWER))
-		rps->last_adj = 0;
-
-	/* sysfs frequency interfaces may have snuck in while servicing the
-	 * interrupt
-	 */
-	new_delay += adj;
-	new_delay = clamp_t(int, new_delay, min, max);
-
-	if (intel_set_rps(dev_priv, new_delay)) {
-		DRM_DEBUG_DRIVER("Failed to set new GPU frequency\n");
-		rps->last_adj = 0;
-	}
-
-	mutex_unlock(&rps->lock);
-
-out:
-	/* Make sure not to corrupt PMIMR state used by ringbuffer on GEN6 */
-	spin_lock_irq(&gt->irq_lock);
-	if (rps->interrupts_enabled)
-		gen6_gt_pm_unmask_irq(gt, dev_priv->pm_rps_events);
-	spin_unlock_irq(&gt->irq_lock);
-}
-
-
 /**
  * ivybridge_parity_work - Workqueue called when a parity error interrupt
  * occurred.
@@ -1401,11 +1028,11 @@ static bool icp_ddi_port_hotplug_long_detect(enum hpd_pin pin, u32 val)
 {
 	switch (pin) {
 	case HPD_PORT_A:
-		return val & ICP_DDIA_HPD_LONG_DETECT;
+		return val & SHOTPLUG_CTL_DDI_HPD_LONG_DETECT(PORT_A);
 	case HPD_PORT_B:
-		return val & ICP_DDIB_HPD_LONG_DETECT;
+		return val & SHOTPLUG_CTL_DDI_HPD_LONG_DETECT(PORT_B);
 	case HPD_PORT_C:
-		return val & TGP_DDIC_HPD_LONG_DETECT;
+		return val & SHOTPLUG_CTL_DDI_HPD_LONG_DETECT(PORT_C);
 	default:
 		return false;
 	}
@@ -1427,20 +1054,6 @@ static bool icp_tc_port_hotplug_long_detect(enum hpd_pin pin, u32 val)
 	}
 }
 
-static bool tgp_ddi_port_hotplug_long_detect(enum hpd_pin pin, u32 val)
-{
-	switch (pin) {
-	case HPD_PORT_A:
-		return val & ICP_DDIA_HPD_LONG_DETECT;
-	case HPD_PORT_B:
-		return val & ICP_DDIB_HPD_LONG_DETECT;
-	case HPD_PORT_C:
-		return val & TGP_DDIC_HPD_LONG_DETECT;
-	default:
-		return false;
-	}
-}
-
 static bool tgp_tc_port_hotplug_long_detect(enum hpd_pin pin, u32 val)
 {
 	switch (pin) {
@@ -1652,54 +1265,6 @@ static void i9xx_pipe_crc_irq_handler(struct drm_i915_private *dev_priv,
 				     res1, res2);
 }
 
-/* The RPS events need forcewake, so we add them to a work queue and mask their
- * IMR bits until the work is done. Other interrupts can be processed without
- * the work queue. */
-void gen11_rps_irq_handler(struct intel_gt *gt, u32 pm_iir)
-{
-	struct drm_i915_private *i915 = gt->i915;
-	struct intel_rps *rps = &i915->gt_pm.rps;
-	const u32 events = i915->pm_rps_events & pm_iir;
-
-	lockdep_assert_held(&gt->irq_lock);
-
-	if (unlikely(!events))
-		return;
-
-	gen6_gt_pm_mask_irq(gt, events);
-
-	if (!rps->interrupts_enabled)
-		return;
-
-	rps->pm_iir |= events;
-	schedule_work(&rps->work);
-}
-
-void gen6_rps_irq_handler(struct drm_i915_private *dev_priv, u32 pm_iir)
-{
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
-	struct intel_gt *gt = &dev_priv->gt;
-
-	if (pm_iir & dev_priv->pm_rps_events) {
-		spin_lock(&gt->irq_lock);
-		gen6_gt_pm_mask_irq(gt, pm_iir & dev_priv->pm_rps_events);
-		if (rps->interrupts_enabled) {
-			rps->pm_iir |= pm_iir & dev_priv->pm_rps_events;
-			schedule_work(&rps->work);
-		}
-		spin_unlock(&gt->irq_lock);
-	}
-
-	if (INTEL_GEN(dev_priv) >= 8)
-		return;
-
-	if (pm_iir & PM_VEBOX_USER_INTERRUPT)
-		intel_engine_breadcrumbs_irq(dev_priv->engine[VECS0]);
-
-	if (pm_iir & PM_VEBOX_CS_ERROR_INTERRUPT)
-		DRM_DEBUG("Command parser error, pm_iir 0x%08x\n", pm_iir);
-}
-
 static void i9xx_pipestat_irq_reset(struct drm_i915_private *dev_priv)
 {
 	enum pipe pipe;
@@ -1716,7 +1281,7 @@ static void i9xx_pipestat_irq_reset(struct drm_i915_private *dev_priv)
 static void i9xx_pipestat_irq_ack(struct drm_i915_private *dev_priv,
 				  u32 iir, u32 pipe_stats[I915_MAX_PIPES])
 {
-	int pipe;
+	enum pipe pipe;
 
 	spin_lock(&dev_priv->irq_lock);
 
@@ -1741,6 +1306,7 @@ static void i9xx_pipestat_irq_ack(struct drm_i915_private *dev_priv,
 		status_mask = PIPE_FIFO_UNDERRUN_STATUS;
 
 		switch (pipe) {
+		default:
 		case PIPE_A:
 			iir_bit = I915_DISPLAY_PIPE_A_EVENT_INTERRUPT;
 			break;
@@ -2009,7 +1575,7 @@ static irqreturn_t valleyview_irq_handler(int irq, void *arg)
 		if (gt_iir)
 			gen6_gt_irq_handler(&dev_priv->gt, gt_iir);
 		if (pm_iir)
-			gen6_rps_irq_handler(dev_priv, pm_iir);
+			gen6_rps_irq_handler(&dev_priv->gt.rps, pm_iir);
 
 		if (hotplug_status)
 			i9xx_hpd_irq_handler(dev_priv, hotplug_status);
@@ -2136,7 +1702,7 @@ static void ibx_hpd_irq_handler(struct drm_i915_private *dev_priv,
 
 static void ibx_irq_handler(struct drm_i915_private *dev_priv, u32 pch_iir)
 {
-	int pipe;
+	enum pipe pipe;
 	u32 hotplug_trigger = pch_iir & SDE_HOTPLUG_MASK;
 
 	ibx_hpd_irq_handler(dev_priv, hotplug_trigger, hpd_ibx);
@@ -2222,7 +1788,7 @@ static void cpt_serr_int_handler(struct drm_i915_private *dev_priv)
 
 static void cpt_irq_handler(struct drm_i915_private *dev_priv, u32 pch_iir)
 {
-	int pipe;
+	enum pipe pipe;
 	u32 hotplug_trigger = pch_iir & SDE_HOTPLUG_MASK_CPT;
 
 	ibx_hpd_irq_handler(dev_priv, hotplug_trigger, hpd_cpt);
@@ -2256,19 +1822,35 @@ static void cpt_irq_handler(struct drm_i915_private *dev_priv, u32 pch_iir)
 		cpt_serr_int_handler(dev_priv);
 }
 
-static void icp_irq_handler(struct drm_i915_private *dev_priv, u32 pch_iir,
-			    const u32 *pins)
+static void icp_irq_handler(struct drm_i915_private *dev_priv, u32 pch_iir)
 {
-	u32 ddi_hotplug_trigger;
-	u32 tc_hotplug_trigger;
+	u32 ddi_hotplug_trigger, tc_hotplug_trigger;
 	u32 pin_mask = 0, long_mask = 0;
+	bool (*tc_port_hotplug_long_detect)(enum hpd_pin pin, u32 val);
+	const u32 *pins;
 
-	if (HAS_PCH_MCC(dev_priv)) {
+	if (HAS_PCH_TGP(dev_priv)) {
+		ddi_hotplug_trigger = pch_iir & SDE_DDI_MASK_TGP;
+		tc_hotplug_trigger = pch_iir & SDE_TC_MASK_TGP;
+		tc_port_hotplug_long_detect = tgp_tc_port_hotplug_long_detect;
+		pins = hpd_tgp;
+	} else if (HAS_PCH_JSP(dev_priv)) {
 		ddi_hotplug_trigger = pch_iir & SDE_DDI_MASK_TGP;
 		tc_hotplug_trigger = 0;
+		pins = hpd_tgp;
+	} else if (HAS_PCH_MCC(dev_priv)) {
+		ddi_hotplug_trigger = pch_iir & SDE_DDI_MASK_ICP;
+		tc_hotplug_trigger = pch_iir & SDE_TC_HOTPLUG_ICP(PORT_TC1);
+		tc_port_hotplug_long_detect = icp_tc_port_hotplug_long_detect;
+		pins = hpd_icp;
 	} else {
+		WARN(!HAS_PCH_ICP(dev_priv),
+		     "Unrecognized PCH type 0x%x\n", INTEL_PCH_TYPE(dev_priv));
+
 		ddi_hotplug_trigger = pch_iir & SDE_DDI_MASK_ICP;
 		tc_hotplug_trigger = pch_iir & SDE_TC_MASK_ICP;
+		tc_port_hotplug_long_detect = icp_tc_port_hotplug_long_detect;
+		pins = hpd_icp;
 	}
 
 	if (ddi_hotplug_trigger) {
@@ -2292,44 +1874,7 @@ static void icp_irq_handler(struct drm_i915_private *dev_priv, u32 pch_iir,
 		intel_get_hpd_pins(dev_priv, &pin_mask, &long_mask,
 				   tc_hotplug_trigger,
 				   dig_hotplug_reg, pins,
-				   icp_tc_port_hotplug_long_detect);
-	}
-
-	if (pin_mask)
-		intel_hpd_irq_handler(dev_priv, pin_mask, long_mask);
-
-	if (pch_iir & SDE_GMBUS_ICP)
-		gmbus_irq_handler(dev_priv);
-}
-
-static void tgp_irq_handler(struct drm_i915_private *dev_priv, u32 pch_iir)
-{
-	u32 ddi_hotplug_trigger = pch_iir & SDE_DDI_MASK_TGP;
-	u32 tc_hotplug_trigger = pch_iir & SDE_TC_MASK_TGP;
-	u32 pin_mask = 0, long_mask = 0;
-
-	if (ddi_hotplug_trigger) {
-		u32 dig_hotplug_reg;
-
-		dig_hotplug_reg = I915_READ(SHOTPLUG_CTL_DDI);
-		I915_WRITE(SHOTPLUG_CTL_DDI, dig_hotplug_reg);
-
-		intel_get_hpd_pins(dev_priv, &pin_mask, &long_mask,
-				   ddi_hotplug_trigger,
-				   dig_hotplug_reg, hpd_tgp,
-				   tgp_ddi_port_hotplug_long_detect);
-	}
-
-	if (tc_hotplug_trigger) {
-		u32 dig_hotplug_reg;
-
-		dig_hotplug_reg = I915_READ(SHOTPLUG_CTL_TC);
-		I915_WRITE(SHOTPLUG_CTL_TC, dig_hotplug_reg);
-
-		intel_get_hpd_pins(dev_priv, &pin_mask, &long_mask,
-				   tc_hotplug_trigger,
-				   dig_hotplug_reg, hpd_tgp,
-				   tgp_tc_port_hotplug_long_detect);
+				   tc_port_hotplug_long_detect);
 	}
 
 	if (pin_mask)
@@ -2434,7 +1979,7 @@ static void ilk_display_irq_handler(struct drm_i915_private *dev_priv,
 	}
 
 	if (IS_GEN(dev_priv, 5) && de_iir & DE_PCU_EVENT)
-		ironlake_rps_change_irq_handler(dev_priv);
+		gen5_rps_irq_handler(&dev_priv->gt.rps);
 }
 
 static void ivb_display_irq_handler(struct drm_i915_private *dev_priv,
@@ -2539,7 +2084,7 @@ static irqreturn_t ironlake_irq_handler(int irq, void *arg)
 		if (pm_iir) {
 			I915_WRITE(GEN6_PMIIR, pm_iir);
 			ret = IRQ_HANDLED;
-			gen6_rps_irq_handler(dev_priv, pm_iir);
+			gen6_rps_irq_handler(&dev_priv->gt.rps, pm_iir);
 		}
 	}
 
@@ -2616,10 +2161,16 @@ static u32 gen8_de_port_aux_mask(struct drm_i915_private *dev_priv)
 	u32 mask;
 
 	if (INTEL_GEN(dev_priv) >= 12)
-		/* TODO: Add AUX entries for USBC */
 		return TGL_DE_PORT_AUX_DDIA |
 			TGL_DE_PORT_AUX_DDIB |
-			TGL_DE_PORT_AUX_DDIC;
+			TGL_DE_PORT_AUX_DDIC |
+			TGL_DE_PORT_AUX_USBC1 |
+			TGL_DE_PORT_AUX_USBC2 |
+			TGL_DE_PORT_AUX_USBC3 |
+			TGL_DE_PORT_AUX_USBC4 |
+			TGL_DE_PORT_AUX_USBC5 |
+			TGL_DE_PORT_AUX_USBC6;
+
 
 	mask = GEN8_AUX_CHANNEL_A;
 	if (INTEL_GEN(dev_priv) >= 9)
@@ -2638,7 +2189,9 @@ static u32 gen8_de_port_aux_mask(struct drm_i915_private *dev_priv)
 
 static u32 gen8_de_pipe_fault_mask(struct drm_i915_private *dev_priv)
 {
-	if (INTEL_GEN(dev_priv) >= 9)
+	if (INTEL_GEN(dev_priv) >= 11)
+		return GEN11_DE_PIPE_IRQ_FAULT_ERRORS;
+	else if (INTEL_GEN(dev_priv) >= 9)
 		return GEN9_DE_PIPE_IRQ_FAULT_ERRORS;
 	else
 		return GEN8_DE_PIPE_IRQ_FAULT_ERRORS;
@@ -2655,11 +2208,21 @@ gen8_de_misc_irq_handler(struct drm_i915_private *dev_priv, u32 iir)
 	}
 
 	if (iir & GEN8_DE_EDP_PSR) {
-		u32 psr_iir = I915_READ(EDP_PSR_IIR);
+		u32 psr_iir;
+		i915_reg_t iir_reg;
+
+		if (INTEL_GEN(dev_priv) >= 12)
+			iir_reg = TRANS_PSR_IIR(dev_priv->psr.transcoder);
+		else
+			iir_reg = EDP_PSR_IIR;
+
+		psr_iir = I915_READ(iir_reg);
+		I915_WRITE(iir_reg, psr_iir);
+
+		if (psr_iir)
+			found = true;
 
 		intel_psr_irq_handler(dev_priv, psr_iir);
-		I915_WRITE(EDP_PSR_IIR, psr_iir);
-		found = true;
 	}
 
 	if (!found)
@@ -2780,12 +2343,8 @@ gen8_de_irq_handler(struct drm_i915_private *dev_priv, u32 master_ctl)
 			I915_WRITE(SDEIIR, iir);
 			ret = IRQ_HANDLED;
 
-			if (INTEL_PCH_TYPE(dev_priv) >= PCH_TGP)
-				tgp_irq_handler(dev_priv, iir);
-			else if (INTEL_PCH_TYPE(dev_priv) >= PCH_MCC)
-				icp_irq_handler(dev_priv, iir, hpd_mcc);
-			else if (INTEL_PCH_TYPE(dev_priv) >= PCH_ICP)
-				icp_irq_handler(dev_priv, iir, hpd_icp);
+			if (INTEL_PCH_TYPE(dev_priv) >= PCH_ICP)
+				icp_irq_handler(dev_priv, iir);
 			else if (INTEL_PCH_TYPE(dev_priv) >= PCH_SPT)
 				spt_irq_handler(dev_priv, iir);
 			else
@@ -2894,9 +2453,11 @@ static inline void gen11_master_intr_enable(void __iomem * const regs)
 	raw_reg_write(regs, GEN11_GFX_MSTR_IRQ, GEN11_MASTER_IRQ);
 }
 
-static irqreturn_t gen11_irq_handler(int irq, void *arg)
+static __always_inline irqreturn_t
+__gen11_irq_handler(struct drm_i915_private * const i915,
+		    u32 (*intr_disable)(void __iomem * const regs),
+		    void (*intr_enable)(void __iomem * const regs))
 {
-	struct drm_i915_private * const i915 = arg;
 	void __iomem * const regs = i915->uncore.regs;
 	struct intel_gt *gt = &i915->gt;
 	u32 master_ctl;
@@ -2905,9 +2466,9 @@ static irqreturn_t gen11_irq_handler(int irq, void *arg)
 	if (!intel_irqs_enabled(i915))
 		return IRQ_NONE;
 
-	master_ctl = gen11_master_intr_disable(regs);
+	master_ctl = intr_disable(regs);
 	if (!master_ctl) {
-		gen11_master_intr_enable(regs);
+		intr_enable(regs);
 		return IRQ_NONE;
 	}
 
@@ -2929,13 +2490,20 @@ static irqreturn_t gen11_irq_handler(int irq, void *arg)
 
 	gu_misc_iir = gen11_gu_misc_irq_ack(gt, master_ctl);
 
-	gen11_master_intr_enable(regs);
+	intr_enable(regs);
 
 	gen11_gu_misc_irq_handler(gt, gu_misc_iir);
 
 	return IRQ_HANDLED;
 }
 
+static irqreturn_t gen11_irq_handler(int irq, void *arg)
+{
+	return __gen11_irq_handler(arg,
+				   gen11_master_intr_disable,
+				   gen11_master_intr_enable);
+}
+
 /* Called from drm generic code, passed 'crtc' which
  * we use as a pipe index
  */
@@ -2952,12 +2520,18 @@ int i8xx_enable_vblank(struct drm_crtc *crtc)
 	return 0;
 }
 
-int i945gm_enable_vblank(struct drm_crtc *crtc)
+int i915gm_enable_vblank(struct drm_crtc *crtc)
 {
 	struct drm_i915_private *dev_priv = to_i915(crtc->dev);
 
-	if (dev_priv->i945gm_vblank.enabled++ == 0)
-		schedule_work(&dev_priv->i945gm_vblank.work);
+	/*
+	 * Vblank interrupts fail to wake the device up from C2+.
+	 * Disabling render clock gating during C-states avoids
+	 * the problem. There is a small power cost so we do this
+	 * only when vblank interrupts are actually enabled.
+	 */
+	if (dev_priv->vblank_enabled++ == 0)
+		I915_WRITE(SCPD0, _MASKED_BIT_ENABLE(CSTATE_RENDER_CLOCK_GATE_DISABLE));
 
 	return i8xx_enable_vblank(crtc);
 }
@@ -3030,14 +2604,14 @@ void i8xx_disable_vblank(struct drm_crtc *crtc)
 	spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
 }
 
-void i945gm_disable_vblank(struct drm_crtc *crtc)
+void i915gm_disable_vblank(struct drm_crtc *crtc)
 {
 	struct drm_i915_private *dev_priv = to_i915(crtc->dev);
 
 	i8xx_disable_vblank(crtc);
 
-	if (--dev_priv->i945gm_vblank.enabled == 0)
-		schedule_work(&dev_priv->i945gm_vblank.work);
+	if (--dev_priv->vblank_enabled == 0)
+		I915_WRITE(SCPD0, _MASKED_BIT_DISABLE(CSTATE_RENDER_CLOCK_GATE_DISABLE));
 }
 
 void i965_disable_vblank(struct drm_crtc *crtc)
@@ -3076,60 +2650,6 @@ void bdw_disable_vblank(struct drm_crtc *crtc)
 	spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
 }
 
-static void i945gm_vblank_work_func(struct work_struct *work)
-{
-	struct drm_i915_private *dev_priv =
-		container_of(work, struct drm_i915_private, i945gm_vblank.work);
-
-	/*
-	 * Vblank interrupts fail to wake up the device from C3,
-	 * hence we want to prevent C3 usage while vblank interrupts
-	 * are enabled.
-	 */
-	pm_qos_update_request(&dev_priv->i945gm_vblank.pm_qos,
-			      READ_ONCE(dev_priv->i945gm_vblank.enabled) ?
-			      dev_priv->i945gm_vblank.c3_disable_latency :
-			      PM_QOS_DEFAULT_VALUE);
-}
-
-static int cstate_disable_latency(const char *name)
-{
-	const struct cpuidle_driver *drv;
-	int i;
-
-	drv = cpuidle_get_driver();
-	if (!drv)
-		return 0;
-
-	for (i = 0; i < drv->state_count; i++) {
-		const struct cpuidle_state *state = &drv->states[i];
-
-		if (!strcmp(state->name, name))
-			return state->exit_latency ?
-				state->exit_latency - 1 : 0;
-	}
-
-	return 0;
-}
-
-static void i945gm_vblank_work_init(struct drm_i915_private *dev_priv)
-{
-	INIT_WORK(&dev_priv->i945gm_vblank.work,
-		  i945gm_vblank_work_func);
-
-	dev_priv->i945gm_vblank.c3_disable_latency =
-		cstate_disable_latency("C3");
-	pm_qos_add_request(&dev_priv->i945gm_vblank.pm_qos,
-			   PM_QOS_CPU_DMA_LATENCY,
-			   PM_QOS_DEFAULT_VALUE);
-}
-
-static void i945gm_vblank_work_fini(struct drm_i915_private *dev_priv)
-{
-	cancel_work_sync(&dev_priv->i945gm_vblank.work);
-	pm_qos_remove_request(&dev_priv->i945gm_vblank.pm_qos);
-}
-
 static void ibx_irq_reset(struct drm_i915_private *dev_priv)
 {
 	struct intel_uncore *uncore = &dev_priv->uncore;
@@ -3246,7 +2766,7 @@ static void valleyview_irq_reset(struct drm_i915_private *dev_priv)
 static void gen8_irq_reset(struct drm_i915_private *dev_priv)
 {
 	struct intel_uncore *uncore = &dev_priv->uncore;
-	int pipe;
+	enum pipe pipe;
 
 	gen8_master_intr_disable(dev_priv->uncore.regs);
 
@@ -3271,7 +2791,7 @@ static void gen8_irq_reset(struct drm_i915_private *dev_priv)
 static void gen11_irq_reset(struct drm_i915_private *dev_priv)
 {
 	struct intel_uncore *uncore = &dev_priv->uncore;
-	int pipe;
+	enum pipe pipe;
 
 	gen11_master_intr_disable(dev_priv->uncore.regs);
 
@@ -3279,8 +2799,23 @@ static void gen11_irq_reset(struct drm_i915_private *dev_priv)
 
 	intel_uncore_write(uncore, GEN11_DISPLAY_INT_CTL, 0);
 
-	intel_uncore_write(uncore, EDP_PSR_IMR, 0xffffffff);
-	intel_uncore_write(uncore, EDP_PSR_IIR, 0xffffffff);
+	if (INTEL_GEN(dev_priv) >= 12) {
+		enum transcoder trans;
+
+		for (trans = TRANSCODER_A; trans <= TRANSCODER_D; trans++) {
+			enum intel_display_power_domain domain;
+
+			domain = POWER_DOMAIN_TRANSCODER(trans);
+			if (!intel_display_power_is_enabled(dev_priv, domain))
+				continue;
+
+			intel_uncore_write(uncore, TRANS_PSR_IMR(trans), 0xffffffff);
+			intel_uncore_write(uncore, TRANS_PSR_IIR(trans), 0xffffffff);
+		}
+	} else {
+		intel_uncore_write(uncore, EDP_PSR_IMR, 0xffffffff);
+		intel_uncore_write(uncore, EDP_PSR_IIR, 0xffffffff);
+	}
 
 	for_each_pipe(dev_priv, pipe)
 		if (intel_display_power_is_enabled(dev_priv,
@@ -3431,42 +2966,44 @@ static void icp_hpd_detection_setup(struct drm_i915_private *dev_priv,
 	}
 }
 
-static void icp_hpd_irq_setup(struct drm_i915_private *dev_priv)
+static void icp_hpd_irq_setup(struct drm_i915_private *dev_priv,
+			      u32 sde_ddi_mask, u32 sde_tc_mask,
+			      u32 ddi_enable_mask, u32 tc_enable_mask,
+			      const u32 *pins)
 {
 	u32 hotplug_irqs, enabled_irqs;
 
-	hotplug_irqs = SDE_DDI_MASK_ICP | SDE_TC_MASK_ICP;
-	enabled_irqs = intel_hpd_enabled_irqs(dev_priv, hpd_icp);
+	hotplug_irqs = sde_ddi_mask | sde_tc_mask;
+	enabled_irqs = intel_hpd_enabled_irqs(dev_priv, pins);
 
 	ibx_display_interrupt_update(dev_priv, hotplug_irqs, enabled_irqs);
 
-	icp_hpd_detection_setup(dev_priv, ICP_DDI_HPD_ENABLE_MASK,
-				ICP_TC_HPD_ENABLE_MASK);
+	icp_hpd_detection_setup(dev_priv, ddi_enable_mask, tc_enable_mask);
 }
 
+/*
+ * EHL doesn't need most of gen11_hpd_irq_setup, it's handling only the
+ * equivalent of SDE.
+ */
 static void mcc_hpd_irq_setup(struct drm_i915_private *dev_priv)
 {
-	u32 hotplug_irqs, enabled_irqs;
-
-	hotplug_irqs = SDE_DDI_MASK_TGP;
-	enabled_irqs = intel_hpd_enabled_irqs(dev_priv, hpd_mcc);
-
-	ibx_display_interrupt_update(dev_priv, hotplug_irqs, enabled_irqs);
-
-	icp_hpd_detection_setup(dev_priv, TGP_DDI_HPD_ENABLE_MASK, 0);
+	icp_hpd_irq_setup(dev_priv,
+			  SDE_DDI_MASK_ICP, SDE_TC_HOTPLUG_ICP(PORT_TC1),
+			  ICP_DDI_HPD_ENABLE_MASK, ICP_TC_HPD_ENABLE(PORT_TC1),
+			  hpd_icp);
 }
 
-static void tgp_hpd_irq_setup(struct drm_i915_private *dev_priv)
+/*
+ * JSP behaves exactly the same as MCC above except that port C is mapped to
+ * the DDI-C pins instead of the TC1 pins.  This means we should follow TGP's
+ * masks & tables rather than ICP's masks & tables.
+ */
+static void jsp_hpd_irq_setup(struct drm_i915_private *dev_priv)
 {
-	u32 hotplug_irqs, enabled_irqs;
-
-	hotplug_irqs = SDE_DDI_MASK_TGP | SDE_TC_MASK_TGP;
-	enabled_irqs = intel_hpd_enabled_irqs(dev_priv, hpd_tgp);
-
-	ibx_display_interrupt_update(dev_priv, hotplug_irqs, enabled_irqs);
-
-	icp_hpd_detection_setup(dev_priv, TGP_DDI_HPD_ENABLE_MASK,
-				TGP_TC_HPD_ENABLE_MASK);
+	icp_hpd_irq_setup(dev_priv,
+			  SDE_DDI_MASK_TGP, 0,
+			  TGP_DDI_HPD_ENABLE_MASK, 0,
+			  hpd_tgp);
 }
 
 static void gen11_hpd_detection_setup(struct drm_i915_private *dev_priv)
@@ -3506,9 +3043,13 @@ static void gen11_hpd_irq_setup(struct drm_i915_private *dev_priv)
 	gen11_hpd_detection_setup(dev_priv);
 
 	if (INTEL_PCH_TYPE(dev_priv) >= PCH_TGP)
-		tgp_hpd_irq_setup(dev_priv);
+		icp_hpd_irq_setup(dev_priv, SDE_DDI_MASK_TGP, SDE_TC_MASK_TGP,
+				  TGP_DDI_HPD_ENABLE_MASK,
+				  TGP_TC_HPD_ENABLE_MASK, hpd_tgp);
 	else if (INTEL_PCH_TYPE(dev_priv) >= PCH_ICP)
-		icp_hpd_irq_setup(dev_priv);
+		icp_hpd_irq_setup(dev_priv, SDE_DDI_MASK_ICP, SDE_TC_MASK_ICP,
+				  ICP_DDI_HPD_ENABLE_MASK,
+				  ICP_TC_HPD_ENABLE_MASK, hpd_icp);
 }
 
 static void spt_hpd_detection_setup(struct drm_i915_private *dev_priv)
@@ -3684,7 +3225,6 @@ static void ironlake_irq_postinstall(struct drm_i915_private *dev_priv)
 
 	if (IS_HASWELL(dev_priv)) {
 		gen3_assert_iir_is_zero(uncore, EDP_PSR_IIR);
-		intel_psr_irq_control(dev_priv, dev_priv->psr.debug);
 		display_mask |= DE_EDP_PSR_INT_HSW;
 	}
 
@@ -3794,8 +3334,21 @@ static void gen8_de_irq_postinstall(struct drm_i915_private *dev_priv)
 	else if (IS_BROADWELL(dev_priv))
 		de_port_enables |= GEN8_PORT_DP_A_HOTPLUG;
 
-	gen3_assert_iir_is_zero(uncore, EDP_PSR_IIR);
-	intel_psr_irq_control(dev_priv, dev_priv->psr.debug);
+	if (INTEL_GEN(dev_priv) >= 12) {
+		enum transcoder trans;
+
+		for (trans = TRANSCODER_A; trans <= TRANSCODER_D; trans++) {
+			enum intel_display_power_domain domain;
+
+			domain = POWER_DOMAIN_TRANSCODER(trans);
+			if (!intel_display_power_is_enabled(dev_priv, domain))
+				continue;
+
+			gen3_assert_iir_is_zero(uncore, TRANS_PSR_IIR(trans));
+		}
+	} else {
+		gen3_assert_iir_is_zero(uncore, EDP_PSR_IIR);
+	}
 
 	for_each_pipe(dev_priv, pipe) {
 		dev_priv->de_irq_mask[pipe] = ~de_pipe_masked;
@@ -3853,8 +3406,11 @@ static void icp_irq_postinstall(struct drm_i915_private *dev_priv)
 	if (HAS_PCH_TGP(dev_priv))
 		icp_hpd_detection_setup(dev_priv, TGP_DDI_HPD_ENABLE_MASK,
 					TGP_TC_HPD_ENABLE_MASK);
-	else if (HAS_PCH_MCC(dev_priv))
+	else if (HAS_PCH_JSP(dev_priv))
 		icp_hpd_detection_setup(dev_priv, TGP_DDI_HPD_ENABLE_MASK, 0);
+	else if (HAS_PCH_MCC(dev_priv))
+		icp_hpd_detection_setup(dev_priv, ICP_DDI_HPD_ENABLE_MASK,
+					ICP_TC_HPD_ENABLE(PORT_TC1));
 	else
 		icp_hpd_detection_setup(dev_priv, ICP_DDI_HPD_ENABLE_MASK,
 					ICP_TC_HPD_ENABLE_MASK);
@@ -4317,16 +3873,10 @@ static irqreturn_t i965_irq_handler(int irq, void *arg)
 void intel_irq_init(struct drm_i915_private *dev_priv)
 {
 	struct drm_device *dev = &dev_priv->drm;
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
 	int i;
 
-	if (IS_I945GM(dev_priv))
-		i945gm_vblank_work_init(dev_priv);
-
 	intel_hpd_init_work(dev_priv);
 
-	INIT_WORK(&rps->work, gen6_pm_rps_work);
-
 	INIT_WORK(&dev_priv->l3_parity.error_work, ivybridge_parity_work);
 	for (i = 0; i < MAX_L3_SLICES; ++i)
 		dev_priv->l3_parity.remap_info[i] = NULL;
@@ -4335,33 +3885,6 @@ void intel_irq_init(struct drm_i915_private *dev_priv)
 	if (HAS_GT_UC(dev_priv) && INTEL_GEN(dev_priv) < 11)
 		dev_priv->gt.pm_guc_events = GUC_INTR_GUC2HOST << 16;
 
-	/* Let's track the enabled rps events */
-	if (IS_VALLEYVIEW(dev_priv))
-		/* WaGsvRC0ResidencyMethod:vlv */
-		dev_priv->pm_rps_events = GEN6_PM_RP_UP_EI_EXPIRED;
-	else
-		dev_priv->pm_rps_events = (GEN6_PM_RP_UP_THRESHOLD |
-					   GEN6_PM_RP_DOWN_THRESHOLD |
-					   GEN6_PM_RP_DOWN_TIMEOUT);
-
-	/* We share the register with other engine */
-	if (INTEL_GEN(dev_priv) > 9)
-		GEM_WARN_ON(dev_priv->pm_rps_events & 0xffff0000);
-
-	rps->pm_intrmsk_mbz = 0;
-
-	/*
-	 * SNB,IVB,HSW can while VLV,CHV may hard hang on looping batchbuffer
-	 * if GEN6_PM_UP_EI_EXPIRED is masked.
-	 *
-	 * TODO: verify if this can be reproduced on VLV,CHV.
-	 */
-	if (INTEL_GEN(dev_priv) <= 7)
-		rps->pm_intrmsk_mbz |= GEN6_PM_RP_UP_EI_EXPIRED;
-
-	if (INTEL_GEN(dev_priv) >= 8)
-		rps->pm_intrmsk_mbz |= GEN8_PMINTR_DISABLE_REDIRECT_TO_GUC;
-
 	dev->vblank_disable_immediate = true;
 
 	/* Most platforms treat the display irq block as an always-on
@@ -4387,8 +3910,9 @@ void intel_irq_init(struct drm_i915_private *dev_priv)
 		if (I915_HAS_HOTPLUG(dev_priv))
 			dev_priv->display.hpd_irq_setup = i915_hpd_irq_setup;
 	} else {
-		if (HAS_PCH_MCC(dev_priv))
-			/* EHL doesn't need most of gen11_hpd_irq_setup */
+		if (HAS_PCH_JSP(dev_priv))
+			dev_priv->display.hpd_irq_setup = jsp_hpd_irq_setup;
+		else if (HAS_PCH_MCC(dev_priv))
 			dev_priv->display.hpd_irq_setup = mcc_hpd_irq_setup;
 		else if (INTEL_GEN(dev_priv) >= 11)
 			dev_priv->display.hpd_irq_setup = gen11_hpd_irq_setup;
@@ -4411,9 +3935,6 @@ void intel_irq_fini(struct drm_i915_private *i915)
 {
 	int i;
 
-	if (IS_I945GM(i915))
-		i945gm_vblank_work_fini(i915);
-
 	for (i = 0; i < MAX_L3_SLICES; ++i)
 		kfree(i915->l3_parity.remap_info[i]);
 }
@@ -4538,10 +4059,10 @@ void intel_irq_uninstall(struct drm_i915_private *dev_priv)
 	int irq = dev_priv->drm.pdev->irq;
 
 	/*
-	 * FIXME we can get called twice during driver load
-	 * error handling due to intel_modeset_cleanup()
-	 * calling us out of sequence. Would be nice if
-	 * it didn't do that...
+	 * FIXME we can get called twice during driver probe
+	 * error handling as well as during driver remove due to
+	 * intel_modeset_driver_remove() calling us out of sequence.
+	 * Would be nice if it didn't do that...
 	 */
 	if (!dev_priv->drm.irq_enabled)
 		return;
diff --git a/drivers/gpu/drm/i915/i915_irq.h b/drivers/gpu/drm/i915/i915_irq.h
index 8e7e6071777e..812c47a9c2d6 100644
--- a/drivers/gpu/drm/i915/i915_irq.h
+++ b/drivers/gpu/drm/i915/i915_irq.h
@@ -17,14 +17,8 @@ struct drm_device;
 struct drm_display_mode;
 struct drm_i915_private;
 struct intel_crtc;
-struct intel_crtc;
-struct intel_gt;
-struct intel_guc;
 struct intel_uncore;
 
-void gen11_rps_irq_handler(struct intel_gt *gt, u32 pm_iir);
-void gen6_rps_irq_handler(struct drm_i915_private *dev_priv, u32 pm_iir);
-
 void intel_irq_init(struct drm_i915_private *dev_priv);
 void intel_irq_fini(struct drm_i915_private *dev_priv);
 int intel_irq_install(struct drm_i915_private *dev_priv);
@@ -106,12 +100,6 @@ void gen8_irq_power_well_post_enable(struct drm_i915_private *dev_priv,
 				     u8 pipe_mask);
 void gen8_irq_power_well_pre_disable(struct drm_i915_private *dev_priv,
 				     u8 pipe_mask);
-void gen9_reset_guc_interrupts(struct intel_guc *guc);
-void gen9_enable_guc_interrupts(struct intel_guc *guc);
-void gen9_disable_guc_interrupts(struct intel_guc *guc);
-void gen11_reset_guc_interrupts(struct intel_guc *guc);
-void gen11_enable_guc_interrupts(struct intel_guc *guc);
-void gen11_disable_guc_interrupts(struct intel_guc *guc);
 
 bool i915_get_crtc_scanoutpos(struct drm_device *dev, unsigned int pipe,
 			      bool in_vblank_irq, int *vpos, int *hpos,
@@ -122,12 +110,12 @@ u32 i915_get_vblank_counter(struct drm_crtc *crtc);
 u32 g4x_get_vblank_counter(struct drm_crtc *crtc);
 
 int i8xx_enable_vblank(struct drm_crtc *crtc);
-int i945gm_enable_vblank(struct drm_crtc *crtc);
+int i915gm_enable_vblank(struct drm_crtc *crtc);
 int i965_enable_vblank(struct drm_crtc *crtc);
 int ilk_enable_vblank(struct drm_crtc *crtc);
 int bdw_enable_vblank(struct drm_crtc *crtc);
 void i8xx_disable_vblank(struct drm_crtc *crtc);
-void i945gm_disable_vblank(struct drm_crtc *crtc);
+void i915gm_disable_vblank(struct drm_crtc *crtc);
 void i965_disable_vblank(struct drm_crtc *crtc);
 void ilk_disable_vblank(struct drm_crtc *crtc);
 void bdw_disable_vblank(struct drm_crtc *crtc);
diff --git a/drivers/gpu/drm/i915/i915_params.c b/drivers/gpu/drm/i915/i915_params.c
index 296452f9efe4..1dd1f3652795 100644
--- a/drivers/gpu/drm/i915/i915_params.c
+++ b/drivers/gpu/drm/i915/i915_params.c
@@ -46,7 +46,8 @@ i915_param_named(modeset, int, 0400,
 
 i915_param_named_unsafe(enable_dc, int, 0400,
 	"Enable power-saving display C-states. "
-	"(-1=auto [default]; 0=disable; 1=up to DC5; 2=up to DC6)");
+	"(-1=auto [default]; 0=disable; 1=up to DC5; 2=up to DC6; "
+	"3=up to DC5 with DC3CO; 4=up to DC6 with DC3CO)");
 
 i915_param_named_unsafe(enable_fbc, int, 0600,
 	"Enable frame buffer compression for power savings "
@@ -165,7 +166,7 @@ i915_param_named_unsafe(enable_dp_mst, bool, 0600,
 	"Enable multi-stream transport (MST) for new DisplayPort sinks. (default: true)");
 
 #if IS_ENABLED(CONFIG_DRM_I915_DEBUG)
-i915_param_named_unsafe(inject_load_failure, uint, 0400,
+i915_param_named_unsafe(inject_probe_failure, uint, 0400,
 	"Force an error after a number of failure check points (0:disabled (default), N:force failure at the Nth failure check point)");
 #endif
 
@@ -178,6 +179,11 @@ i915_param_named(enable_gvt, bool, 0400,
 	"Enable support for Intel GVT-g graphics virtualization host support(default:false)");
 #endif
 
+#if IS_ENABLED(CONFIG_DRM_I915_UNSTABLE_FAKE_LMEM)
+i915_param_named_unsafe(fake_lmem_start, ulong, 0600,
+	"Fake LMEM start offset (default: 0)");
+#endif
+
 static __always_inline void _print_param(struct drm_printer *p,
 					 const char *name,
 					 const char *type,
@@ -189,6 +195,8 @@ static __always_inline void _print_param(struct drm_printer *p,
 		drm_printf(p, "i915.%s=%d\n", name, *(const int *)x);
 	else if (!__builtin_strcmp(type, "unsigned int"))
 		drm_printf(p, "i915.%s=%u\n", name, *(const unsigned int *)x);
+	else if (!__builtin_strcmp(type, "unsigned long"))
+		drm_printf(p, "i915.%s=%lu\n", name, *(const unsigned long *)x);
 	else if (!__builtin_strcmp(type, "char *"))
 		drm_printf(p, "i915.%s=%s\n", name, *(const char **)x);
 	else
diff --git a/drivers/gpu/drm/i915/i915_params.h b/drivers/gpu/drm/i915/i915_params.h
index d29ade3b7de6..31b88f297fbc 100644
--- a/drivers/gpu/drm/i915/i915_params.h
+++ b/drivers/gpu/drm/i915/i915_params.h
@@ -61,11 +61,12 @@ struct drm_printer;
 	param(char *, dmc_firmware_path, NULL) \
 	param(int, mmio_debug, -IS_ENABLED(CONFIG_DRM_I915_DEBUG_MMIO)) \
 	param(int, edp_vswing, 0) \
-	param(int, reset, 2) \
-	param(unsigned int, inject_load_failure, 0) \
+	param(int, reset, 3) \
+	param(unsigned int, inject_probe_failure, 0) \
 	param(int, fastboot, -1) \
 	param(int, enable_dpcd_backlight, 0) \
 	param(char *, force_probe, CONFIG_DRM_I915_FORCE_PROBE) \
+	param(unsigned long, fake_lmem_start, 0) \
 	/* leave bools at the end to not create holes */ \
 	param(bool, alpha_support, IS_ENABLED(CONFIG_DRM_I915_ALPHA_SUPPORT)) \
 	param(bool, enable_hangcheck, true) \
diff --git a/drivers/gpu/drm/i915/i915_pci.c b/drivers/gpu/drm/i915/i915_pci.c
index 1974e4c78a43..1bb701d32a5d 100644
--- a/drivers/gpu/drm/i915/i915_pci.c
+++ b/drivers/gpu/drm/i915/i915_pci.c
@@ -23,7 +23,6 @@
  */
 
 #include <linux/console.h>
-#include <linux/vgaarb.h>
 #include <linux/vga_switcheroo.h>
 
 #include <drm/drm_drv.h>
@@ -118,6 +117,14 @@
 		[PIPE_C] = IVB_CURSOR_C_OFFSET, \
 	}
 
+#define TGL_CURSOR_OFFSETS \
+	.cursor_offsets = { \
+		[PIPE_A] = CURSOR_A_OFFSET, \
+		[PIPE_B] = IVB_CURSOR_B_OFFSET, \
+		[PIPE_C] = IVB_CURSOR_C_OFFSET, \
+		[PIPE_D] = TGL_CURSOR_D_OFFSET, \
+	}
+
 #define I9XX_COLORS \
 	.color = { .gamma_lut_size = 256 }
 #define I965_COLORS \
@@ -144,10 +151,13 @@
 #define GEN_DEFAULT_PAGE_SIZES \
 	.page_sizes = I915_GTT_PAGE_SIZE_4K
 
+#define GEN_DEFAULT_REGIONS \
+	.memory_regions = REGION_SMEM | REGION_STOLEN
+
 #define I830_FEATURES \
 	GEN(2), \
 	.is_mobile = 1, \
-	.num_pipes = 2, \
+	.pipe_mask = BIT(PIPE_A) | BIT(PIPE_B), \
 	.display.has_overlay = 1, \
 	.display.cursor_needs_physical = 1, \
 	.display.overlay_needs_physical = 1, \
@@ -161,11 +171,12 @@
 	I9XX_PIPE_OFFSETS, \
 	I9XX_CURSOR_OFFSETS, \
 	I9XX_COLORS, \
-	GEN_DEFAULT_PAGE_SIZES
+	GEN_DEFAULT_PAGE_SIZES, \
+	GEN_DEFAULT_REGIONS
 
 #define I845_FEATURES \
 	GEN(2), \
-	.num_pipes = 1, \
+	.pipe_mask = BIT(PIPE_A), \
 	.display.has_overlay = 1, \
 	.display.overlay_needs_physical = 1, \
 	.display.has_gmch = 1, \
@@ -178,7 +189,8 @@
 	I845_PIPE_OFFSETS, \
 	I845_CURSOR_OFFSETS, \
 	I9XX_COLORS, \
-	GEN_DEFAULT_PAGE_SIZES
+	GEN_DEFAULT_PAGE_SIZES, \
+	GEN_DEFAULT_REGIONS
 
 static const struct intel_device_info intel_i830_info = {
 	I830_FEATURES,
@@ -203,7 +215,7 @@ static const struct intel_device_info intel_i865g_info = {
 
 #define GEN3_FEATURES \
 	GEN(3), \
-	.num_pipes = 2, \
+	.pipe_mask = BIT(PIPE_A) | BIT(PIPE_B), \
 	.display.has_gmch = 1, \
 	.gpu_reset_clobbers_display = true, \
 	.engine_mask = BIT(RCS0), \
@@ -212,7 +224,8 @@ static const struct intel_device_info intel_i865g_info = {
 	I9XX_PIPE_OFFSETS, \
 	I9XX_CURSOR_OFFSETS, \
 	I9XX_COLORS, \
-	GEN_DEFAULT_PAGE_SIZES
+	GEN_DEFAULT_PAGE_SIZES, \
+	GEN_DEFAULT_REGIONS
 
 static const struct intel_device_info intel_i915g_info = {
 	GEN3_FEATURES,
@@ -287,7 +300,7 @@ static const struct intel_device_info intel_pineview_m_info = {
 
 #define GEN4_FEATURES \
 	GEN(4), \
-	.num_pipes = 2, \
+	.pipe_mask = BIT(PIPE_A) | BIT(PIPE_B), \
 	.display.has_hotplug = 1, \
 	.display.has_gmch = 1, \
 	.gpu_reset_clobbers_display = true, \
@@ -297,7 +310,8 @@ static const struct intel_device_info intel_pineview_m_info = {
 	I9XX_PIPE_OFFSETS, \
 	I9XX_CURSOR_OFFSETS, \
 	I965_COLORS, \
-	GEN_DEFAULT_PAGE_SIZES
+	GEN_DEFAULT_PAGE_SIZES, \
+	GEN_DEFAULT_REGIONS
 
 static const struct intel_device_info intel_i965g_info = {
 	GEN4_FEATURES,
@@ -337,7 +351,7 @@ static const struct intel_device_info intel_gm45_info = {
 
 #define GEN5_FEATURES \
 	GEN(5), \
-	.num_pipes = 2, \
+	.pipe_mask = BIT(PIPE_A) | BIT(PIPE_B), \
 	.display.has_hotplug = 1, \
 	.engine_mask = BIT(RCS0) | BIT(VCS0), \
 	.has_snoop = true, \
@@ -347,7 +361,8 @@ static const struct intel_device_info intel_gm45_info = {
 	I9XX_PIPE_OFFSETS, \
 	I9XX_CURSOR_OFFSETS, \
 	ILK_COLORS, \
-	GEN_DEFAULT_PAGE_SIZES
+	GEN_DEFAULT_PAGE_SIZES, \
+	GEN_DEFAULT_REGIONS
 
 static const struct intel_device_info intel_ironlake_d_info = {
 	GEN5_FEATURES,
@@ -363,7 +378,7 @@ static const struct intel_device_info intel_ironlake_m_info = {
 
 #define GEN6_FEATURES \
 	GEN(6), \
-	.num_pipes = 2, \
+	.pipe_mask = BIT(PIPE_A) | BIT(PIPE_B), \
 	.display.has_hotplug = 1, \
 	.display.has_fbc = 1, \
 	.engine_mask = BIT(RCS0) | BIT(VCS0) | BIT(BCS0), \
@@ -377,7 +392,8 @@ static const struct intel_device_info intel_ironlake_m_info = {
 	I9XX_PIPE_OFFSETS, \
 	I9XX_CURSOR_OFFSETS, \
 	ILK_COLORS, \
-	GEN_DEFAULT_PAGE_SIZES
+	GEN_DEFAULT_PAGE_SIZES, \
+	GEN_DEFAULT_REGIONS
 
 #define SNB_D_PLATFORM \
 	GEN6_FEATURES, \
@@ -411,7 +427,7 @@ static const struct intel_device_info intel_sandybridge_m_gt2_info = {
 
 #define GEN7_FEATURES  \
 	GEN(7), \
-	.num_pipes = 3, \
+	.pipe_mask = BIT(PIPE_A) | BIT(PIPE_B) | BIT(PIPE_C), \
 	.display.has_hotplug = 1, \
 	.display.has_fbc = 1, \
 	.engine_mask = BIT(RCS0) | BIT(VCS0) | BIT(BCS0), \
@@ -420,12 +436,13 @@ static const struct intel_device_info intel_sandybridge_m_gt2_info = {
 	.has_rc6 = 1, \
 	.has_rc6p = 1, \
 	.has_rps = true, \
-	.ppgtt_type = INTEL_PPGTT_FULL, \
+	.ppgtt_type = INTEL_PPGTT_ALIASING, \
 	.ppgtt_size = 31, \
 	IVB_PIPE_OFFSETS, \
 	IVB_CURSOR_OFFSETS, \
 	IVB_COLORS, \
-	GEN_DEFAULT_PAGE_SIZES
+	GEN_DEFAULT_PAGE_SIZES, \
+	GEN_DEFAULT_REGIONS
 
 #define IVB_D_PLATFORM \
 	GEN7_FEATURES, \
@@ -462,7 +479,7 @@ static const struct intel_device_info intel_ivybridge_q_info = {
 	GEN7_FEATURES,
 	PLATFORM(INTEL_IVYBRIDGE),
 	.gt = 2,
-	.num_pipes = 0, /* legal, last one wins */
+	.pipe_mask = 0, /* legal, last one wins */
 	.has_l3_dpf = 1,
 };
 
@@ -470,13 +487,13 @@ static const struct intel_device_info intel_valleyview_info = {
 	PLATFORM(INTEL_VALLEYVIEW),
 	GEN(7),
 	.is_lp = 1,
-	.num_pipes = 2,
+	.pipe_mask = BIT(PIPE_A) | BIT(PIPE_B),
 	.has_runtime_pm = 1,
 	.has_rc6 = 1,
 	.has_rps = true,
 	.display.has_gmch = 1,
 	.display.has_hotplug = 1,
-	.ppgtt_type = INTEL_PPGTT_FULL,
+	.ppgtt_type = INTEL_PPGTT_ALIASING,
 	.ppgtt_size = 31,
 	.has_snoop = true,
 	.has_coherent_ggtt = false,
@@ -486,6 +503,7 @@ static const struct intel_device_info intel_valleyview_info = {
 	I9XX_CURSOR_OFFSETS,
 	I965_COLORS,
 	GEN_DEFAULT_PAGE_SIZES,
+	GEN_DEFAULT_REGIONS,
 };
 
 #define G75_FEATURES  \
@@ -560,7 +578,7 @@ static const struct intel_device_info intel_broadwell_gt3_info = {
 static const struct intel_device_info intel_cherryview_info = {
 	PLATFORM(INTEL_CHERRYVIEW),
 	GEN(8),
-	.num_pipes = 3,
+	.pipe_mask = BIT(PIPE_A) | BIT(PIPE_B) | BIT(PIPE_C),
 	.display.has_hotplug = 1,
 	.is_lp = 1,
 	.engine_mask = BIT(RCS0) | BIT(VCS0) | BIT(BCS0) | BIT(VECS0),
@@ -570,7 +588,7 @@ static const struct intel_device_info intel_cherryview_info = {
 	.has_rps = true,
 	.has_logical_ring_contexts = 1,
 	.display.has_gmch = 1,
-	.ppgtt_type = INTEL_PPGTT_FULL,
+	.ppgtt_type = INTEL_PPGTT_ALIASING,
 	.ppgtt_size = 32,
 	.has_reset_engine = 1,
 	.has_snoop = true,
@@ -580,6 +598,7 @@ static const struct intel_device_info intel_cherryview_info = {
 	CHV_CURSOR_OFFSETS,
 	CHV_COLORS,
 	GEN_DEFAULT_PAGE_SIZES,
+	GEN_DEFAULT_REGIONS,
 };
 
 #define GEN9_DEFAULT_PAGE_SIZES \
@@ -593,6 +612,7 @@ static const struct intel_device_info intel_cherryview_info = {
 	.has_logical_ring_preemption = 1, \
 	.display.has_csr = 1, \
 	.has_gt_uc = 1, \
+	.display.has_hdcp = 1, \
 	.display.has_ipc = 1, \
 	.ddb_size = 896
 
@@ -631,11 +651,12 @@ static const struct intel_device_info intel_skylake_gt4_info = {
 	.is_lp = 1, \
 	.display.has_hotplug = 1, \
 	.engine_mask = BIT(RCS0) | BIT(VCS0) | BIT(BCS0) | BIT(VECS0), \
-	.num_pipes = 3, \
+	.pipe_mask = BIT(PIPE_A) | BIT(PIPE_B) | BIT(PIPE_C), \
 	.has_64bit_reloc = 1, \
 	.display.has_ddi = 1, \
 	.has_fpga_dbg = 1, \
 	.display.has_fbc = 1, \
+	.display.has_hdcp = 1, \
 	.display.has_psr = 1, \
 	.has_runtime_pm = 1, \
 	.display.has_csr = 1, \
@@ -654,7 +675,8 @@ static const struct intel_device_info intel_skylake_gt4_info = {
 	HSW_PIPE_OFFSETS, \
 	IVB_CURSOR_OFFSETS, \
 	IVB_COLORS, \
-	GEN9_DEFAULT_PAGE_SIZES
+	GEN9_DEFAULT_PAGE_SIZES, \
+	GEN_DEFAULT_REGIONS
 
 static const struct intel_device_info intel_broxton_info = {
 	GEN9_LP_FEATURES,
@@ -715,6 +737,7 @@ static const struct intel_device_info intel_coffeelake_gt3_info = {
 	GEN9_FEATURES, \
 	GEN(10), \
 	.ddb_size = 1024, \
+	.display.has_dsc = 1, \
 	.has_coherent_ggtt = false, \
 	GLK_COLORS
 
@@ -787,18 +810,25 @@ static const struct intel_device_info intel_elkhartlake_info = {
 		[TRANSCODER_DSI_0] = TRANSCODER_DSI0_OFFSET, \
 		[TRANSCODER_DSI_1] = TRANSCODER_DSI1_OFFSET, \
 	}, \
-	.has_global_mocs = 1
+	TGL_CURSOR_OFFSETS, \
+	.has_global_mocs = 1, \
+	.display.has_dsb = 1
 
 static const struct intel_device_info intel_tigerlake_12_info = {
 	GEN12_FEATURES,
 	PLATFORM(INTEL_TIGERLAKE),
-	.num_pipes = 4,
+	.pipe_mask = BIT(PIPE_A) | BIT(PIPE_B) | BIT(PIPE_C) | BIT(PIPE_D),
 	.require_force_probe = 1,
 	.display.has_modular_fia = 1,
 	.engine_mask =
 		BIT(RCS0) | BIT(BCS0) | BIT(VECS0) | BIT(VCS0) | BIT(VCS2),
+	.has_rps = false, /* XXX disabled for debugging */
 };
 
+#define GEN12_DGFX_FEATURES \
+	GEN12_FEATURES, \
+	.is_dgfx = 1
+
 #undef GEN
 #undef PLATFORM
 
diff --git a/drivers/gpu/drm/i915/i915_perf.c b/drivers/gpu/drm/i915/i915_perf.c
index e42b86827d6b..65d7c2e599de 100644
--- a/drivers/gpu/drm/i915/i915_perf.c
+++ b/drivers/gpu/drm/i915/i915_perf.c
@@ -196,8 +196,11 @@
 #include <linux/uuid.h>
 
 #include "gem/i915_gem_context.h"
-#include "gem/i915_gem_pm.h"
+#include "gt/intel_engine_pm.h"
+#include "gt/intel_engine_user.h"
+#include "gt/intel_gt.h"
 #include "gt/intel_lrc_reg.h"
+#include "gt/intel_ring.h"
 
 #include "i915_drv.h"
 #include "i915_perf.h"
@@ -215,6 +218,7 @@
 #include "oa/i915_oa_cflgt3.h"
 #include "oa/i915_oa_cnl.h"
 #include "oa/i915_oa_icl.h"
+#include "oa/i915_oa_tgl.h"
 
 /* HW requires this to be a power of two, between 128k and 16M, though driver
  * is currently generally designed assuming the largest 16M size is used such
@@ -291,6 +295,7 @@ static u32 i915_perf_stream_paranoid = true;
 
 /* On Gen8+ automatically triggered OA reports include a 'reason' field... */
 #define OAREPORT_REASON_MASK           0x3f
+#define OAREPORT_REASON_MASK_EXTENDED  0x7f
 #define OAREPORT_REASON_SHIFT          19
 #define OAREPORT_REASON_TIMER          (1<<0)
 #define OAREPORT_REASON_CTX_SWITCH     (1<<3)
@@ -336,17 +341,24 @@ static const struct i915_oa_format gen8_plus_oa_formats[I915_OA_FORMAT_MAX] = {
 	[I915_OA_FORMAT_C4_B8]		    = { 7, 64 },
 };
 
+static const struct i915_oa_format gen12_oa_formats[I915_OA_FORMAT_MAX] = {
+	[I915_OA_FORMAT_A32u40_A4u32_B8_C8] = { 5, 256 },
+};
+
 #define SAMPLE_OA_REPORT      (1<<0)
 
 /**
  * struct perf_open_properties - for validated properties given to open a stream
  * @sample_flags: `DRM_I915_PERF_PROP_SAMPLE_*` properties are tracked as flags
  * @single_context: Whether a single or all gpu contexts should be monitored
+ * @hold_preemption: Whether the preemption is disabled for the filtered
+ *                   context
  * @ctx_handle: A gem ctx handle for use with @single_context
  * @metrics_set: An ID for an OA unit metric set advertised via sysfs
  * @oa_format: An OA unit HW report format
  * @oa_periodic: Whether to enable periodic OA unit sampling
  * @oa_period_exponent: The OA unit sampling period is derived from this
+ * @engine: The engine (typically rcs0) being monitored by the OA unit
  *
  * As read_properties_unlocked() enumerates and validates the properties given
  * to open a stream of metrics the configuration is built up in the structure
@@ -356,6 +368,7 @@ struct perf_open_properties {
 	u32 sample_flags;
 
 	u64 single_context:1;
+	u64 hold_preemption:1;
 	u64 ctx_handle;
 
 	/* OA sampling state */
@@ -363,69 +376,74 @@ struct perf_open_properties {
 	int oa_format;
 	bool oa_periodic;
 	int oa_period_exponent;
+
+	struct intel_engine_cs *engine;
+};
+
+struct i915_oa_config_bo {
+	struct llist_node node;
+
+	struct i915_oa_config *oa_config;
+	struct i915_vma *vma;
 };
 
 static enum hrtimer_restart oa_poll_check_timer_cb(struct hrtimer *hrtimer);
 
-static void free_oa_config(struct drm_i915_private *dev_priv,
-			   struct i915_oa_config *oa_config)
+void i915_oa_config_release(struct kref *ref)
 {
-	if (!PTR_ERR(oa_config->flex_regs))
-		kfree(oa_config->flex_regs);
-	if (!PTR_ERR(oa_config->b_counter_regs))
-		kfree(oa_config->b_counter_regs);
-	if (!PTR_ERR(oa_config->mux_regs))
-		kfree(oa_config->mux_regs);
-	kfree(oa_config);
-}
+	struct i915_oa_config *oa_config =
+		container_of(ref, typeof(*oa_config), ref);
 
-static void put_oa_config(struct drm_i915_private *dev_priv,
-			  struct i915_oa_config *oa_config)
-{
-	if (!atomic_dec_and_test(&oa_config->ref_count))
-		return;
+	kfree(oa_config->flex_regs);
+	kfree(oa_config->b_counter_regs);
+	kfree(oa_config->mux_regs);
 
-	free_oa_config(dev_priv, oa_config);
+	kfree_rcu(oa_config, rcu);
 }
 
-static int get_oa_config(struct drm_i915_private *dev_priv,
-			 int metrics_set,
-			 struct i915_oa_config **out_config)
+struct i915_oa_config *
+i915_perf_get_oa_config(struct i915_perf *perf, int metrics_set)
 {
-	int ret;
+	struct i915_oa_config *oa_config;
 
-	if (metrics_set == 1) {
-		*out_config = &dev_priv->perf.test_config;
-		atomic_inc(&dev_priv->perf.test_config.ref_count);
-		return 0;
-	}
+	rcu_read_lock();
+	if (metrics_set == 1)
+		oa_config = &perf->test_config;
+	else
+		oa_config = idr_find(&perf->metrics_idr, metrics_set);
+	if (oa_config)
+		oa_config = i915_oa_config_get(oa_config);
+	rcu_read_unlock();
 
-	ret = mutex_lock_interruptible(&dev_priv->perf.metrics_lock);
-	if (ret)
-		return ret;
+	return oa_config;
+}
 
-	*out_config = idr_find(&dev_priv->perf.metrics_idr, metrics_set);
-	if (!*out_config)
-		ret = -EINVAL;
-	else
-		atomic_inc(&(*out_config)->ref_count);
+static void free_oa_config_bo(struct i915_oa_config_bo *oa_bo)
+{
+	i915_oa_config_put(oa_bo->oa_config);
+	i915_vma_put(oa_bo->vma);
+	kfree(oa_bo);
+}
 
-	mutex_unlock(&dev_priv->perf.metrics_lock);
+static u32 gen12_oa_hw_tail_read(struct i915_perf_stream *stream)
+{
+	struct intel_uncore *uncore = stream->uncore;
 
-	return ret;
+	return intel_uncore_read(uncore, GEN12_OAG_OATAILPTR) &
+	       GEN12_OAG_OATAILPTR_MASK;
 }
 
 static u32 gen8_oa_hw_tail_read(struct i915_perf_stream *stream)
 {
-	struct drm_i915_private *dev_priv = stream->dev_priv;
+	struct intel_uncore *uncore = stream->uncore;
 
-	return I915_READ(GEN8_OATAILPTR) & GEN8_OATAILPTR_MASK;
+	return intel_uncore_read(uncore, GEN8_OATAILPTR) & GEN8_OATAILPTR_MASK;
 }
 
 static u32 gen7_oa_hw_tail_read(struct i915_perf_stream *stream)
 {
-	struct drm_i915_private *dev_priv = stream->dev_priv;
-	u32 oastatus1 = I915_READ(GEN7_OASTATUS1);
+	struct intel_uncore *uncore = stream->uncore;
+	u32 oastatus1 = intel_uncore_read(uncore, GEN7_OASTATUS1);
 
 	return oastatus1 & GEN7_OASTATUS1_TAIL_MASK;
 }
@@ -456,7 +474,6 @@ static u32 gen7_oa_hw_tail_read(struct i915_perf_stream *stream)
  */
 static bool oa_buffer_check_unlocked(struct i915_perf_stream *stream)
 {
-	struct drm_i915_private *dev_priv = stream->dev_priv;
 	int report_size = stream->oa_buffer.format_size;
 	unsigned long flags;
 	unsigned int aged_idx;
@@ -479,7 +496,7 @@ static bool oa_buffer_check_unlocked(struct i915_perf_stream *stream)
 	aged_tail = stream->oa_buffer.tails[aged_idx].offset;
 	aging_tail = stream->oa_buffer.tails[!aged_idx].offset;
 
-	hw_tail = dev_priv->perf.ops.oa_hw_tail_read(stream);
+	hw_tail = stream->perf->ops.oa_hw_tail_read(stream);
 
 	/* The tail pointer increases in 64 byte increments,
 	 * not in report_size steps...
@@ -536,7 +553,7 @@ static bool oa_buffer_check_unlocked(struct i915_perf_stream *stream)
 				aging_tail = hw_tail;
 			stream->oa_buffer.aging_timestamp = now;
 		} else {
-			DRM_ERROR("Ignoring spurious out of range OA buffer tail pointer = %u\n",
+			DRM_ERROR("Ignoring spurious out of range OA buffer tail pointer = %x\n",
 				  hw_tail);
 		}
 	}
@@ -655,7 +672,7 @@ static int gen8_append_oa_reports(struct i915_perf_stream *stream,
 				  size_t count,
 				  size_t *offset)
 {
-	struct drm_i915_private *dev_priv = stream->dev_priv;
+	struct intel_uncore *uncore = stream->uncore;
 	int report_size = stream->oa_buffer.format_size;
 	u8 *oa_buf_base = stream->oa_buffer.vaddr;
 	u32 gtt_offset = i915_ggtt_offset(stream->oa_buffer.vma);
@@ -738,9 +755,11 @@ static int gen8_append_oa_reports(struct i915_perf_stream *stream,
 		 * it to userspace...
 		 */
 		reason = ((report32[0] >> OAREPORT_REASON_SHIFT) &
-			  OAREPORT_REASON_MASK);
+			  (IS_GEN(stream->perf->i915, 12) ?
+			   OAREPORT_REASON_MASK_EXTENDED :
+			   OAREPORT_REASON_MASK));
 		if (reason == 0) {
-			if (__ratelimit(&dev_priv->perf.spurious_report_rs))
+			if (__ratelimit(&stream->perf->spurious_report_rs))
 				DRM_NOTE("Skipping spurious, invalid OA report\n");
 			continue;
 		}
@@ -755,7 +774,8 @@ static int gen8_append_oa_reports(struct i915_perf_stream *stream,
 		 * Note: that we don't clear the valid_ctx_bit so userspace can
 		 * understand that the ID has been squashed by the kernel.
 		 */
-		if (!(report32[0] & dev_priv->perf.gen8_valid_ctx_bit))
+		if (!(report32[0] & stream->perf->gen8_valid_ctx_bit) &&
+		    INTEL_GEN(stream->perf->i915) <= 11)
 			ctx_id = report32[2] = INVALID_CTX_ID;
 
 		/*
@@ -789,7 +809,7 @@ static int gen8_append_oa_reports(struct i915_perf_stream *stream,
 		 * switches since it's not-uncommon for periodic samples to
 		 * identify a switch before any 'context switch' report.
 		 */
-		if (!dev_priv->perf.exclusive_stream->ctx ||
+		if (!stream->perf->exclusive_stream->ctx ||
 		    stream->specific_ctx_id == ctx_id ||
 		    stream->oa_buffer.last_ctx_id == stream->specific_ctx_id ||
 		    reason & OAREPORT_REASON_CTX_SWITCH) {
@@ -798,7 +818,7 @@ static int gen8_append_oa_reports(struct i915_perf_stream *stream,
 			 * While filtering for a single context we avoid
 			 * leaking the IDs of other contexts.
 			 */
-			if (dev_priv->perf.exclusive_stream->ctx &&
+			if (stream->perf->exclusive_stream->ctx &&
 			    stream->specific_ctx_id != ctx_id) {
 				report32[2] = INVALID_CTX_ID;
 			}
@@ -822,6 +842,11 @@ static int gen8_append_oa_reports(struct i915_perf_stream *stream,
 	}
 
 	if (start_offset != *offset) {
+		i915_reg_t oaheadptr;
+
+		oaheadptr = IS_GEN(stream->perf->i915, 12) ?
+			    GEN12_OAG_OAHEADPTR : GEN8_OAHEADPTR;
+
 		spin_lock_irqsave(&stream->oa_buffer.ptr_lock, flags);
 
 		/*
@@ -829,8 +854,8 @@ static int gen8_append_oa_reports(struct i915_perf_stream *stream,
 		 * relative to oa_buf_base so put back here...
 		 */
 		head += gtt_offset;
-
-		I915_WRITE(GEN8_OAHEADPTR, head & GEN8_OAHEADPTR_MASK);
+		intel_uncore_write(uncore, oaheadptr,
+				   head & GEN12_OAG_OAHEADPTR_MASK);
 		stream->oa_buffer.head = head;
 
 		spin_unlock_irqrestore(&stream->oa_buffer.ptr_lock, flags);
@@ -864,14 +889,18 @@ static int gen8_oa_read(struct i915_perf_stream *stream,
 			size_t count,
 			size_t *offset)
 {
-	struct drm_i915_private *dev_priv = stream->dev_priv;
+	struct intel_uncore *uncore = stream->uncore;
 	u32 oastatus;
+	i915_reg_t oastatus_reg;
 	int ret;
 
 	if (WARN_ON(!stream->oa_buffer.vaddr))
 		return -EIO;
 
-	oastatus = I915_READ(GEN8_OASTATUS);
+	oastatus_reg = IS_GEN(stream->perf->i915, 12) ?
+		       GEN12_OAG_OASTATUS : GEN8_OASTATUS;
+
+	oastatus = intel_uncore_read(uncore, oastatus_reg);
 
 	/*
 	 * We treat OABUFFER_OVERFLOW as a significant error:
@@ -896,14 +925,14 @@ static int gen8_oa_read(struct i915_perf_stream *stream,
 		DRM_DEBUG("OA buffer overflow (exponent = %d): force restart\n",
 			  stream->period_exponent);
 
-		dev_priv->perf.ops.oa_disable(stream);
-		dev_priv->perf.ops.oa_enable(stream);
+		stream->perf->ops.oa_disable(stream);
+		stream->perf->ops.oa_enable(stream);
 
 		/*
 		 * Note: .oa_enable() is expected to re-init the oabuffer and
 		 * reset GEN8_OASTATUS for us
 		 */
-		oastatus = I915_READ(GEN8_OASTATUS);
+		oastatus = intel_uncore_read(uncore, oastatus_reg);
 	}
 
 	if (oastatus & GEN8_OASTATUS_REPORT_LOST) {
@@ -911,8 +940,8 @@ static int gen8_oa_read(struct i915_perf_stream *stream,
 				       DRM_I915_PERF_RECORD_OA_REPORT_LOST);
 		if (ret)
 			return ret;
-		I915_WRITE(GEN8_OASTATUS,
-			   oastatus & ~GEN8_OASTATUS_REPORT_LOST);
+		intel_uncore_write(uncore, oastatus_reg,
+				   oastatus & ~GEN8_OASTATUS_REPORT_LOST);
 	}
 
 	return gen8_append_oa_reports(stream, buf, count, offset);
@@ -943,7 +972,7 @@ static int gen7_append_oa_reports(struct i915_perf_stream *stream,
 				  size_t count,
 				  size_t *offset)
 {
-	struct drm_i915_private *dev_priv = stream->dev_priv;
+	struct intel_uncore *uncore = stream->uncore;
 	int report_size = stream->oa_buffer.format_size;
 	u8 *oa_buf_base = stream->oa_buffer.vaddr;
 	u32 gtt_offset = i915_ggtt_offset(stream->oa_buffer.vma);
@@ -1017,7 +1046,7 @@ static int gen7_append_oa_reports(struct i915_perf_stream *stream,
 		 * copying it to userspace...
 		 */
 		if (report32[0] == 0) {
-			if (__ratelimit(&dev_priv->perf.spurious_report_rs))
+			if (__ratelimit(&stream->perf->spurious_report_rs))
 				DRM_NOTE("Skipping spurious, invalid OA report\n");
 			continue;
 		}
@@ -1043,9 +1072,9 @@ static int gen7_append_oa_reports(struct i915_perf_stream *stream,
 		 */
 		head += gtt_offset;
 
-		I915_WRITE(GEN7_OASTATUS2,
-			   ((head & GEN7_OASTATUS2_HEAD_MASK) |
-			    GEN7_OASTATUS2_MEM_SELECT_GGTT));
+		intel_uncore_write(uncore, GEN7_OASTATUS2,
+				   (head & GEN7_OASTATUS2_HEAD_MASK) |
+				   GEN7_OASTATUS2_MEM_SELECT_GGTT);
 		stream->oa_buffer.head = head;
 
 		spin_unlock_irqrestore(&stream->oa_buffer.ptr_lock, flags);
@@ -1075,21 +1104,21 @@ static int gen7_oa_read(struct i915_perf_stream *stream,
 			size_t count,
 			size_t *offset)
 {
-	struct drm_i915_private *dev_priv = stream->dev_priv;
+	struct intel_uncore *uncore = stream->uncore;
 	u32 oastatus1;
 	int ret;
 
 	if (WARN_ON(!stream->oa_buffer.vaddr))
 		return -EIO;
 
-	oastatus1 = I915_READ(GEN7_OASTATUS1);
+	oastatus1 = intel_uncore_read(uncore, GEN7_OASTATUS1);
 
 	/* XXX: On Haswell we don't have a safe way to clear oastatus1
 	 * bits while the OA unit is enabled (while the tail pointer
 	 * may be updated asynchronously) so we ignore status bits
 	 * that have already been reported to userspace.
 	 */
-	oastatus1 &= ~dev_priv->perf.gen7_latched_oastatus1;
+	oastatus1 &= ~stream->perf->gen7_latched_oastatus1;
 
 	/* We treat OABUFFER_OVERFLOW as a significant error:
 	 *
@@ -1120,10 +1149,10 @@ static int gen7_oa_read(struct i915_perf_stream *stream,
 		DRM_DEBUG("OA buffer overflow (exponent = %d): force restart\n",
 			  stream->period_exponent);
 
-		dev_priv->perf.ops.oa_disable(stream);
-		dev_priv->perf.ops.oa_enable(stream);
+		stream->perf->ops.oa_disable(stream);
+		stream->perf->ops.oa_enable(stream);
 
-		oastatus1 = I915_READ(GEN7_OASTATUS1);
+		oastatus1 = intel_uncore_read(uncore, GEN7_OASTATUS1);
 	}
 
 	if (unlikely(oastatus1 & GEN7_OASTATUS1_REPORT_LOST)) {
@@ -1131,7 +1160,7 @@ static int gen7_oa_read(struct i915_perf_stream *stream,
 				       DRM_I915_PERF_RECORD_OA_REPORT_LOST);
 		if (ret)
 			return ret;
-		dev_priv->perf.gen7_latched_oastatus1 |=
+		stream->perf->gen7_latched_oastatus1 |=
 			GEN7_OASTATUS1_REPORT_LOST;
 	}
 
@@ -1196,25 +1225,18 @@ static int i915_oa_read(struct i915_perf_stream *stream,
 			size_t count,
 			size_t *offset)
 {
-	struct drm_i915_private *dev_priv = stream->dev_priv;
-
-	return dev_priv->perf.ops.read(stream, buf, count, offset);
+	return stream->perf->ops.read(stream, buf, count, offset);
 }
 
 static struct intel_context *oa_pin_context(struct i915_perf_stream *stream)
 {
 	struct i915_gem_engines_iter it;
-	struct drm_i915_private *i915 = stream->dev_priv;
 	struct i915_gem_context *ctx = stream->ctx;
 	struct intel_context *ce;
 	int err;
 
-	err = i915_mutex_lock_interruptible(&i915->drm);
-	if (err)
-		return ERR_PTR(err);
-
 	for_each_gem_engine(ce, i915_gem_context_lock_engines(ctx), it) {
-		if (ce->engine->class != RENDER_CLASS)
+		if (ce->engine != stream->engine) /* first match! */
 			continue;
 
 		/*
@@ -1229,10 +1251,6 @@ static struct intel_context *oa_pin_context(struct i915_perf_stream *stream)
 	}
 	i915_gem_context_unlock_engines(ctx);
 
-	mutex_unlock(&i915->drm.struct_mutex);
-	if (err)
-		return ERR_PTR(err);
-
 	return stream->pinned_ctx;
 }
 
@@ -1248,14 +1266,13 @@ static struct intel_context *oa_pin_context(struct i915_perf_stream *stream)
  */
 static int oa_get_render_ctx_id(struct i915_perf_stream *stream)
 {
-	struct drm_i915_private *i915 = stream->dev_priv;
 	struct intel_context *ce;
 
 	ce = oa_pin_context(stream);
 	if (IS_ERR(ce))
 		return PTR_ERR(ce);
 
-	switch (INTEL_GEN(i915)) {
+	switch (INTEL_GEN(ce->engine->i915)) {
 	case 7: {
 		/*
 		 * On Haswell we don't do any post processing of the reports
@@ -1269,7 +1286,11 @@ static int oa_get_render_ctx_id(struct i915_perf_stream *stream)
 	case 8:
 	case 9:
 	case 10:
-		if (USES_GUC_SUBMISSION(i915)) {
+		if (intel_engine_in_execlists_submission_mode(ce->engine)) {
+			stream->specific_ctx_id_mask =
+				(1U << GEN8_CTX_ID_WIDTH) - 1;
+			stream->specific_ctx_id = stream->specific_ctx_id_mask;
+		} else {
 			/*
 			 * When using GuC, the context descriptor we write in
 			 * i915 is read by GuC and rewritten before it's
@@ -1289,31 +1310,23 @@ static int oa_get_render_ctx_id(struct i915_perf_stream *stream)
 			 */
 			stream->specific_ctx_id_mask =
 				(1U << (GEN8_CTX_ID_WIDTH - 1)) - 1;
-		} else {
-			stream->specific_ctx_id_mask =
-				(1U << GEN8_CTX_ID_WIDTH) - 1;
-			stream->specific_ctx_id =
-				upper_32_bits(ce->lrc_desc);
-			stream->specific_ctx_id &=
-				stream->specific_ctx_id_mask;
 		}
 		break;
 
-	case 11: {
+	case 11:
+	case 12: {
 		stream->specific_ctx_id_mask =
-			((1U << GEN11_SW_CTX_ID_WIDTH) - 1) << (GEN11_SW_CTX_ID_SHIFT - 32) |
-			((1U << GEN11_ENGINE_INSTANCE_WIDTH) - 1) << (GEN11_ENGINE_INSTANCE_SHIFT - 32) |
-			((1 << GEN11_ENGINE_CLASS_WIDTH) - 1) << (GEN11_ENGINE_CLASS_SHIFT - 32);
-		stream->specific_ctx_id = upper_32_bits(ce->lrc_desc);
-		stream->specific_ctx_id &=
-			stream->specific_ctx_id_mask;
+			((1U << GEN11_SW_CTX_ID_WIDTH) - 1) << (GEN11_SW_CTX_ID_SHIFT - 32);
+		stream->specific_ctx_id = stream->specific_ctx_id_mask;
 		break;
 	}
 
 	default:
-		MISSING_CASE(INTEL_GEN(i915));
+		MISSING_CASE(INTEL_GEN(ce->engine->i915));
 	}
 
+	ce->tag = stream->specific_ctx_id_mask;
+
 	DRM_DEBUG_DRIVER("filtering on ctx_id=0x%x ctx_id_mask=0x%x\n",
 			 stream->specific_ctx_id,
 			 stream->specific_ctx_id_mask);
@@ -1330,69 +1343,76 @@ static int oa_get_render_ctx_id(struct i915_perf_stream *stream)
  */
 static void oa_put_render_ctx_id(struct i915_perf_stream *stream)
 {
-	struct drm_i915_private *dev_priv = stream->dev_priv;
 	struct intel_context *ce;
 
-	stream->specific_ctx_id = INVALID_CTX_ID;
-	stream->specific_ctx_id_mask = 0;
-
 	ce = fetch_and_zero(&stream->pinned_ctx);
 	if (ce) {
-		mutex_lock(&dev_priv->drm.struct_mutex);
+		ce->tag = 0; /* recomputed on next submission after parking */
 		intel_context_unpin(ce);
-		mutex_unlock(&dev_priv->drm.struct_mutex);
 	}
+
+	stream->specific_ctx_id = INVALID_CTX_ID;
+	stream->specific_ctx_id_mask = 0;
 }
 
 static void
 free_oa_buffer(struct i915_perf_stream *stream)
 {
-	struct drm_i915_private *i915 = stream->dev_priv;
-
-	mutex_lock(&i915->drm.struct_mutex);
-
 	i915_vma_unpin_and_release(&stream->oa_buffer.vma,
 				   I915_VMA_RELEASE_MAP);
 
-	mutex_unlock(&i915->drm.struct_mutex);
-
 	stream->oa_buffer.vaddr = NULL;
 }
 
+static void
+free_oa_configs(struct i915_perf_stream *stream)
+{
+	struct i915_oa_config_bo *oa_bo, *tmp;
+
+	i915_oa_config_put(stream->oa_config);
+	llist_for_each_entry_safe(oa_bo, tmp, stream->oa_config_bos.first, node)
+		free_oa_config_bo(oa_bo);
+}
+
+static void
+free_noa_wait(struct i915_perf_stream *stream)
+{
+	i915_vma_unpin_and_release(&stream->noa_wait, 0);
+}
+
 static void i915_oa_stream_destroy(struct i915_perf_stream *stream)
 {
-	struct drm_i915_private *dev_priv = stream->dev_priv;
+	struct i915_perf *perf = stream->perf;
 
-	BUG_ON(stream != dev_priv->perf.exclusive_stream);
+	BUG_ON(stream != perf->exclusive_stream);
 
 	/*
 	 * Unset exclusive_stream first, it will be checked while disabling
 	 * the metric set on gen8+.
 	 */
-	mutex_lock(&dev_priv->drm.struct_mutex);
-	dev_priv->perf.exclusive_stream = NULL;
-	dev_priv->perf.ops.disable_metric_set(stream);
-	mutex_unlock(&dev_priv->drm.struct_mutex);
+	perf->exclusive_stream = NULL;
+	perf->ops.disable_metric_set(stream);
 
 	free_oa_buffer(stream);
 
-	intel_uncore_forcewake_put(&dev_priv->uncore, FORCEWAKE_ALL);
-	intel_runtime_pm_put(&dev_priv->runtime_pm, stream->wakeref);
+	intel_uncore_forcewake_put(stream->uncore, FORCEWAKE_ALL);
+	intel_engine_pm_put(stream->engine);
 
 	if (stream->ctx)
 		oa_put_render_ctx_id(stream);
 
-	put_oa_config(dev_priv, stream->oa_config);
+	free_oa_configs(stream);
+	free_noa_wait(stream);
 
-	if (dev_priv->perf.spurious_report_rs.missed) {
+	if (perf->spurious_report_rs.missed) {
 		DRM_NOTE("%d spurious OA report notices suppressed due to ratelimiting\n",
-			 dev_priv->perf.spurious_report_rs.missed);
+			 perf->spurious_report_rs.missed);
 	}
 }
 
 static void gen7_init_oa_buffer(struct i915_perf_stream *stream)
 {
-	struct drm_i915_private *dev_priv = stream->dev_priv;
+	struct intel_uncore *uncore = stream->uncore;
 	u32 gtt_offset = i915_ggtt_offset(stream->oa_buffer.vma);
 	unsigned long flags;
 
@@ -1401,13 +1421,14 @@ static void gen7_init_oa_buffer(struct i915_perf_stream *stream)
 	/* Pre-DevBDW: OABUFFER must be set with counters off,
 	 * before OASTATUS1, but after OASTATUS2
 	 */
-	I915_WRITE(GEN7_OASTATUS2,
-		   gtt_offset | GEN7_OASTATUS2_MEM_SELECT_GGTT); /* head */
+	intel_uncore_write(uncore, GEN7_OASTATUS2, /* head */
+			   gtt_offset | GEN7_OASTATUS2_MEM_SELECT_GGTT);
 	stream->oa_buffer.head = gtt_offset;
 
-	I915_WRITE(GEN7_OABUFFER, gtt_offset);
+	intel_uncore_write(uncore, GEN7_OABUFFER, gtt_offset);
 
-	I915_WRITE(GEN7_OASTATUS1, gtt_offset | OABUFFER_SIZE_16M); /* tail */
+	intel_uncore_write(uncore, GEN7_OASTATUS1, /* tail */
+			   gtt_offset | OABUFFER_SIZE_16M);
 
 	/* Mark that we need updated tail pointers to read from... */
 	stream->oa_buffer.tails[0].offset = INVALID_TAIL_PTR;
@@ -1419,7 +1440,7 @@ static void gen7_init_oa_buffer(struct i915_perf_stream *stream)
 	 * already seen since they can't be cleared while periodic
 	 * sampling is enabled.
 	 */
-	dev_priv->perf.gen7_latched_oastatus1 = 0;
+	stream->perf->gen7_latched_oastatus1 = 0;
 
 	/* NB: although the OA buffer will initially be allocated
 	 * zeroed via shmfs (and so this memset is redundant when
@@ -1434,25 +1455,22 @@ static void gen7_init_oa_buffer(struct i915_perf_stream *stream)
 	 */
 	memset(stream->oa_buffer.vaddr, 0, OA_BUFFER_SIZE);
 
-	/* Maybe make ->pollin per-stream state if we support multiple
-	 * concurrent streams in the future.
-	 */
 	stream->pollin = false;
 }
 
 static void gen8_init_oa_buffer(struct i915_perf_stream *stream)
 {
-	struct drm_i915_private *dev_priv = stream->dev_priv;
+	struct intel_uncore *uncore = stream->uncore;
 	u32 gtt_offset = i915_ggtt_offset(stream->oa_buffer.vma);
 	unsigned long flags;
 
 	spin_lock_irqsave(&stream->oa_buffer.ptr_lock, flags);
 
-	I915_WRITE(GEN8_OASTATUS, 0);
-	I915_WRITE(GEN8_OAHEADPTR, gtt_offset);
+	intel_uncore_write(uncore, GEN8_OASTATUS, 0);
+	intel_uncore_write(uncore, GEN8_OAHEADPTR, gtt_offset);
 	stream->oa_buffer.head = gtt_offset;
 
-	I915_WRITE(GEN8_OABUFFER_UDW, 0);
+	intel_uncore_write(uncore, GEN8_OABUFFER_UDW, 0);
 
 	/*
 	 * PRM says:
@@ -1462,9 +1480,9 @@ static void gen8_init_oa_buffer(struct i915_perf_stream *stream)
 	 *  to enable proper functionality of the overflow
 	 *  bit."
 	 */
-	I915_WRITE(GEN8_OABUFFER, gtt_offset |
+	intel_uncore_write(uncore, GEN8_OABUFFER, gtt_offset |
 		   OABUFFER_SIZE_16M | GEN8_OABUFFER_MEM_SELECT_GGTT);
-	I915_WRITE(GEN8_OATAILPTR, gtt_offset & GEN8_OATAILPTR_MASK);
+	intel_uncore_write(uncore, GEN8_OATAILPTR, gtt_offset & GEN8_OATAILPTR_MASK);
 
 	/* Mark that we need updated tail pointers to read from... */
 	stream->oa_buffer.tails[0].offset = INVALID_TAIL_PTR;
@@ -1493,35 +1511,82 @@ static void gen8_init_oa_buffer(struct i915_perf_stream *stream)
 	 */
 	memset(stream->oa_buffer.vaddr, 0, OA_BUFFER_SIZE);
 
+	stream->pollin = false;
+}
+
+static void gen12_init_oa_buffer(struct i915_perf_stream *stream)
+{
+	struct intel_uncore *uncore = stream->uncore;
+	u32 gtt_offset = i915_ggtt_offset(stream->oa_buffer.vma);
+	unsigned long flags;
+
+	spin_lock_irqsave(&stream->oa_buffer.ptr_lock, flags);
+
+	intel_uncore_write(uncore, GEN12_OAG_OASTATUS, 0);
+	intel_uncore_write(uncore, GEN12_OAG_OAHEADPTR,
+			   gtt_offset & GEN12_OAG_OAHEADPTR_MASK);
+	stream->oa_buffer.head = gtt_offset;
+
+	/*
+	 * PRM says:
+	 *
+	 *  "This MMIO must be set before the OATAILPTR
+	 *  register and after the OAHEADPTR register. This is
+	 *  to enable proper functionality of the overflow
+	 *  bit."
+	 */
+	intel_uncore_write(uncore, GEN12_OAG_OABUFFER, gtt_offset |
+			   OABUFFER_SIZE_16M | GEN8_OABUFFER_MEM_SELECT_GGTT);
+	intel_uncore_write(uncore, GEN12_OAG_OATAILPTR,
+			   gtt_offset & GEN12_OAG_OATAILPTR_MASK);
+
+	/* Mark that we need updated tail pointers to read from... */
+	stream->oa_buffer.tails[0].offset = INVALID_TAIL_PTR;
+	stream->oa_buffer.tails[1].offset = INVALID_TAIL_PTR;
+
+	/*
+	 * Reset state used to recognise context switches, affecting which
+	 * reports we will forward to userspace while filtering for a single
+	 * context.
+	 */
+	stream->oa_buffer.last_ctx_id = INVALID_CTX_ID;
+
+	spin_unlock_irqrestore(&stream->oa_buffer.ptr_lock, flags);
+
 	/*
-	 * Maybe make ->pollin per-stream state if we support multiple
-	 * concurrent streams in the future.
+	 * NB: although the OA buffer will initially be allocated
+	 * zeroed via shmfs (and so this memset is redundant when
+	 * first allocating), we may re-init the OA buffer, either
+	 * when re-enabling a stream or in error/reset paths.
+	 *
+	 * The reason we clear the buffer for each re-init is for the
+	 * sanity check in gen8_append_oa_reports() that looks at the
+	 * reason field to make sure it's non-zero which relies on
+	 * the assumption that new reports are being written to zeroed
+	 * memory...
 	 */
+	memset(stream->oa_buffer.vaddr, 0,
+	       stream->oa_buffer.vma->size);
+
 	stream->pollin = false;
 }
 
 static int alloc_oa_buffer(struct i915_perf_stream *stream)
 {
 	struct drm_i915_gem_object *bo;
-	struct drm_i915_private *dev_priv = stream->dev_priv;
 	struct i915_vma *vma;
 	int ret;
 
 	if (WARN_ON(stream->oa_buffer.vma))
 		return -ENODEV;
 
-	ret = i915_mutex_lock_interruptible(&dev_priv->drm);
-	if (ret)
-		return ret;
-
 	BUILD_BUG_ON_NOT_POWER_OF_2(OA_BUFFER_SIZE);
 	BUILD_BUG_ON(OA_BUFFER_SIZE < SZ_128K || OA_BUFFER_SIZE > SZ_16M);
 
-	bo = i915_gem_object_create_shmem(dev_priv, OA_BUFFER_SIZE);
+	bo = i915_gem_object_create_shmem(stream->perf->i915, OA_BUFFER_SIZE);
 	if (IS_ERR(bo)) {
 		DRM_ERROR("Failed to allocate OA buffer\n");
-		ret = PTR_ERR(bo);
-		goto unlock;
+		return PTR_ERR(bo);
 	}
 
 	i915_gem_object_set_cache_coherency(bo, I915_CACHE_LLC);
@@ -1541,11 +1606,7 @@ static int alloc_oa_buffer(struct i915_perf_stream *stream)
 		goto err_unpin;
 	}
 
-	DRM_DEBUG_DRIVER("OA Buffer initialized, gtt offset = 0x%x, vaddr = %p\n",
-			 i915_ggtt_offset(stream->oa_buffer.vma),
-			 stream->oa_buffer.vaddr);
-
-	goto unlock;
+	return 0;
 
 err_unpin:
 	__i915_vma_unpin(vma);
@@ -1556,55 +1617,389 @@ err_unref:
 	stream->oa_buffer.vaddr = NULL;
 	stream->oa_buffer.vma = NULL;
 
-unlock:
-	mutex_unlock(&dev_priv->drm.struct_mutex);
 	return ret;
 }
 
-static void config_oa_regs(struct drm_i915_private *dev_priv,
-			   const struct i915_oa_reg *regs,
-			   u32 n_regs)
+static u32 *save_restore_register(struct i915_perf_stream *stream, u32 *cs,
+				  bool save, i915_reg_t reg, u32 offset,
+				  u32 dword_count)
+{
+	u32 cmd;
+	u32 d;
+
+	cmd = save ? MI_STORE_REGISTER_MEM : MI_LOAD_REGISTER_MEM;
+	if (INTEL_GEN(stream->perf->i915) >= 8)
+		cmd++;
+
+	for (d = 0; d < dword_count; d++) {
+		*cs++ = cmd;
+		*cs++ = i915_mmio_reg_offset(reg) + 4 * d;
+		*cs++ = intel_gt_scratch_offset(stream->engine->gt,
+						offset) + 4 * d;
+		*cs++ = 0;
+	}
+
+	return cs;
+}
+
+static int alloc_noa_wait(struct i915_perf_stream *stream)
+{
+	struct drm_i915_private *i915 = stream->perf->i915;
+	struct drm_i915_gem_object *bo;
+	struct i915_vma *vma;
+	const u64 delay_ticks = 0xffffffffffffffff -
+		DIV64_U64_ROUND_UP(
+			atomic64_read(&stream->perf->noa_programming_delay) *
+			RUNTIME_INFO(i915)->cs_timestamp_frequency_khz,
+			1000000ull);
+	const u32 base = stream->engine->mmio_base;
+#define CS_GPR(x) GEN8_RING_CS_GPR(base, x)
+	u32 *batch, *ts0, *cs, *jump;
+	int ret, i;
+	enum {
+		START_TS,
+		NOW_TS,
+		DELTA_TS,
+		JUMP_PREDICATE,
+		DELTA_TARGET,
+		N_CS_GPR
+	};
+
+	bo = i915_gem_object_create_internal(i915, 4096);
+	if (IS_ERR(bo)) {
+		DRM_ERROR("Failed to allocate NOA wait batchbuffer\n");
+		return PTR_ERR(bo);
+	}
+
+	/*
+	 * We pin in GGTT because we jump into this buffer now because
+	 * multiple OA config BOs will have a jump to this address and it
+	 * needs to be fixed during the lifetime of the i915/perf stream.
+	 */
+	vma = i915_gem_object_ggtt_pin(bo, NULL, 0, 0, PIN_HIGH);
+	if (IS_ERR(vma)) {
+		ret = PTR_ERR(vma);
+		goto err_unref;
+	}
+
+	batch = cs = i915_gem_object_pin_map(bo, I915_MAP_WB);
+	if (IS_ERR(batch)) {
+		ret = PTR_ERR(batch);
+		goto err_unpin;
+	}
+
+	/* Save registers. */
+	for (i = 0; i < N_CS_GPR; i++)
+		cs = save_restore_register(
+			stream, cs, true /* save */, CS_GPR(i),
+			INTEL_GT_SCRATCH_FIELD_PERF_CS_GPR + 8 * i, 2);
+	cs = save_restore_register(
+		stream, cs, true /* save */, MI_PREDICATE_RESULT_1,
+		INTEL_GT_SCRATCH_FIELD_PERF_PREDICATE_RESULT_1, 1);
+
+	/* First timestamp snapshot location. */
+	ts0 = cs;
+
+	/*
+	 * Initial snapshot of the timestamp register to implement the wait.
+	 * We work with 32b values, so clear out the top 32b bits of the
+	 * register because the ALU works 64bits.
+	 */
+	*cs++ = MI_LOAD_REGISTER_IMM(1);
+	*cs++ = i915_mmio_reg_offset(CS_GPR(START_TS)) + 4;
+	*cs++ = 0;
+	*cs++ = MI_LOAD_REGISTER_REG | (3 - 2);
+	*cs++ = i915_mmio_reg_offset(RING_TIMESTAMP(base));
+	*cs++ = i915_mmio_reg_offset(CS_GPR(START_TS));
+
+	/*
+	 * This is the location we're going to jump back into until the
+	 * required amount of time has passed.
+	 */
+	jump = cs;
+
+	/*
+	 * Take another snapshot of the timestamp register. Take care to clear
+	 * up the top 32bits of CS_GPR(1) as we're using it for other
+	 * operations below.
+	 */
+	*cs++ = MI_LOAD_REGISTER_IMM(1);
+	*cs++ = i915_mmio_reg_offset(CS_GPR(NOW_TS)) + 4;
+	*cs++ = 0;
+	*cs++ = MI_LOAD_REGISTER_REG | (3 - 2);
+	*cs++ = i915_mmio_reg_offset(RING_TIMESTAMP(base));
+	*cs++ = i915_mmio_reg_offset(CS_GPR(NOW_TS));
+
+	/*
+	 * Do a diff between the 2 timestamps and store the result back into
+	 * CS_GPR(1).
+	 */
+	*cs++ = MI_MATH(5);
+	*cs++ = MI_MATH_LOAD(MI_MATH_REG_SRCA, MI_MATH_REG(NOW_TS));
+	*cs++ = MI_MATH_LOAD(MI_MATH_REG_SRCB, MI_MATH_REG(START_TS));
+	*cs++ = MI_MATH_SUB;
+	*cs++ = MI_MATH_STORE(MI_MATH_REG(DELTA_TS), MI_MATH_REG_ACCU);
+	*cs++ = MI_MATH_STORE(MI_MATH_REG(JUMP_PREDICATE), MI_MATH_REG_CF);
+
+	/*
+	 * Transfer the carry flag (set to 1 if ts1 < ts0, meaning the
+	 * timestamp have rolled over the 32bits) into the predicate register
+	 * to be used for the predicated jump.
+	 */
+	*cs++ = MI_LOAD_REGISTER_REG | (3 - 2);
+	*cs++ = i915_mmio_reg_offset(CS_GPR(JUMP_PREDICATE));
+	*cs++ = i915_mmio_reg_offset(MI_PREDICATE_RESULT_1);
+
+	/* Restart from the beginning if we had timestamps roll over. */
+	*cs++ = (INTEL_GEN(i915) < 8 ?
+		 MI_BATCH_BUFFER_START :
+		 MI_BATCH_BUFFER_START_GEN8) |
+		MI_BATCH_PREDICATE;
+	*cs++ = i915_ggtt_offset(vma) + (ts0 - batch) * 4;
+	*cs++ = 0;
+
+	/*
+	 * Now add the diff between to previous timestamps and add it to :
+	 *      (((1 * << 64) - 1) - delay_ns)
+	 *
+	 * When the Carry Flag contains 1 this means the elapsed time is
+	 * longer than the expected delay, and we can exit the wait loop.
+	 */
+	*cs++ = MI_LOAD_REGISTER_IMM(2);
+	*cs++ = i915_mmio_reg_offset(CS_GPR(DELTA_TARGET));
+	*cs++ = lower_32_bits(delay_ticks);
+	*cs++ = i915_mmio_reg_offset(CS_GPR(DELTA_TARGET)) + 4;
+	*cs++ = upper_32_bits(delay_ticks);
+
+	*cs++ = MI_MATH(4);
+	*cs++ = MI_MATH_LOAD(MI_MATH_REG_SRCA, MI_MATH_REG(DELTA_TS));
+	*cs++ = MI_MATH_LOAD(MI_MATH_REG_SRCB, MI_MATH_REG(DELTA_TARGET));
+	*cs++ = MI_MATH_ADD;
+	*cs++ = MI_MATH_STOREINV(MI_MATH_REG(JUMP_PREDICATE), MI_MATH_REG_CF);
+
+	/*
+	 * Transfer the result into the predicate register to be used for the
+	 * predicated jump.
+	 */
+	*cs++ = MI_LOAD_REGISTER_REG | (3 - 2);
+	*cs++ = i915_mmio_reg_offset(CS_GPR(JUMP_PREDICATE));
+	*cs++ = i915_mmio_reg_offset(MI_PREDICATE_RESULT_1);
+
+	/* Predicate the jump.  */
+	*cs++ = (INTEL_GEN(i915) < 8 ?
+		 MI_BATCH_BUFFER_START :
+		 MI_BATCH_BUFFER_START_GEN8) |
+		MI_BATCH_PREDICATE;
+	*cs++ = i915_ggtt_offset(vma) + (jump - batch) * 4;
+	*cs++ = 0;
+
+	/* Restore registers. */
+	for (i = 0; i < N_CS_GPR; i++)
+		cs = save_restore_register(
+			stream, cs, false /* restore */, CS_GPR(i),
+			INTEL_GT_SCRATCH_FIELD_PERF_CS_GPR + 8 * i, 2);
+	cs = save_restore_register(
+		stream, cs, false /* restore */, MI_PREDICATE_RESULT_1,
+		INTEL_GT_SCRATCH_FIELD_PERF_PREDICATE_RESULT_1, 1);
+
+	/* And return to the ring. */
+	*cs++ = MI_BATCH_BUFFER_END;
+
+	GEM_BUG_ON(cs - batch > PAGE_SIZE / sizeof(*batch));
+
+	i915_gem_object_flush_map(bo);
+	i915_gem_object_unpin_map(bo);
+
+	stream->noa_wait = vma;
+	return 0;
+
+err_unpin:
+	i915_vma_unpin_and_release(&vma, 0);
+err_unref:
+	i915_gem_object_put(bo);
+	return ret;
+}
+
+static u32 *write_cs_mi_lri(u32 *cs,
+			    const struct i915_oa_reg *reg_data,
+			    u32 n_regs)
 {
 	u32 i;
 
 	for (i = 0; i < n_regs; i++) {
-		const struct i915_oa_reg *reg = regs + i;
+		if ((i % MI_LOAD_REGISTER_IMM_MAX_REGS) == 0) {
+			u32 n_lri = min_t(u32,
+					  n_regs - i,
+					  MI_LOAD_REGISTER_IMM_MAX_REGS);
 
-		I915_WRITE(reg->addr, reg->value);
+			*cs++ = MI_LOAD_REGISTER_IMM(n_lri);
+		}
+		*cs++ = i915_mmio_reg_offset(reg_data[i].addr);
+		*cs++ = reg_data[i].value;
 	}
+
+	return cs;
 }
 
-static void delay_after_mux(void)
+static int num_lri_dwords(int num_regs)
 {
+	int count = 0;
+
+	if (num_regs > 0) {
+		count += DIV_ROUND_UP(num_regs, MI_LOAD_REGISTER_IMM_MAX_REGS);
+		count += num_regs * 2;
+	}
+
+	return count;
+}
+
+static struct i915_oa_config_bo *
+alloc_oa_config_buffer(struct i915_perf_stream *stream,
+		       struct i915_oa_config *oa_config)
+{
+	struct drm_i915_gem_object *obj;
+	struct i915_oa_config_bo *oa_bo;
+	size_t config_length = 0;
+	u32 *cs;
+	int err;
+
+	oa_bo = kzalloc(sizeof(*oa_bo), GFP_KERNEL);
+	if (!oa_bo)
+		return ERR_PTR(-ENOMEM);
+
+	config_length += num_lri_dwords(oa_config->mux_regs_len);
+	config_length += num_lri_dwords(oa_config->b_counter_regs_len);
+	config_length += num_lri_dwords(oa_config->flex_regs_len);
+	config_length += 3; /* MI_BATCH_BUFFER_START */
+	config_length = ALIGN(sizeof(u32) * config_length, I915_GTT_PAGE_SIZE);
+
+	obj = i915_gem_object_create_shmem(stream->perf->i915, config_length);
+	if (IS_ERR(obj)) {
+		err = PTR_ERR(obj);
+		goto err_free;
+	}
+
+	cs = i915_gem_object_pin_map(obj, I915_MAP_WB);
+	if (IS_ERR(cs)) {
+		err = PTR_ERR(cs);
+		goto err_oa_bo;
+	}
+
+	cs = write_cs_mi_lri(cs,
+			     oa_config->mux_regs,
+			     oa_config->mux_regs_len);
+	cs = write_cs_mi_lri(cs,
+			     oa_config->b_counter_regs,
+			     oa_config->b_counter_regs_len);
+	cs = write_cs_mi_lri(cs,
+			     oa_config->flex_regs,
+			     oa_config->flex_regs_len);
+
+	/* Jump into the active wait. */
+	*cs++ = (INTEL_GEN(stream->perf->i915) < 8 ?
+		 MI_BATCH_BUFFER_START :
+		 MI_BATCH_BUFFER_START_GEN8);
+	*cs++ = i915_ggtt_offset(stream->noa_wait);
+	*cs++ = 0;
+
+	i915_gem_object_flush_map(obj);
+	i915_gem_object_unpin_map(obj);
+
+	oa_bo->vma = i915_vma_instance(obj,
+				       &stream->engine->gt->ggtt->vm,
+				       NULL);
+	if (IS_ERR(oa_bo->vma)) {
+		err = PTR_ERR(oa_bo->vma);
+		goto err_oa_bo;
+	}
+
+	oa_bo->oa_config = i915_oa_config_get(oa_config);
+	llist_add(&oa_bo->node, &stream->oa_config_bos);
+
+	return oa_bo;
+
+err_oa_bo:
+	i915_gem_object_put(obj);
+err_free:
+	kfree(oa_bo);
+	return ERR_PTR(err);
+}
+
+static struct i915_vma *
+get_oa_vma(struct i915_perf_stream *stream, struct i915_oa_config *oa_config)
+{
+	struct i915_oa_config_bo *oa_bo;
+
 	/*
-	 * It apparently takes a fairly long time for a new MUX
-	 * configuration to be be applied after these register writes.
-	 * This delay duration was derived empirically based on the
-	 * render_basic config but hopefully it covers the maximum
-	 * configuration latency.
-	 *
-	 * As a fallback, the checks in _append_oa_reports() to skip
-	 * invalid OA reports do also seem to work to discard reports
-	 * generated before this config has completed - albeit not
-	 * silently.
-	 *
-	 * Unfortunately this is essentially a magic number, since we
-	 * don't currently know of a reliable mechanism for predicting
-	 * how long the MUX config will take to apply and besides
-	 * seeing invalid reports we don't know of a reliable way to
-	 * explicitly check that the MUX config has landed.
-	 *
-	 * It's even possible we've miss characterized the underlying
-	 * problem - it just seems like the simplest explanation why
-	 * a delay at this location would mitigate any invalid reports.
+	 * Look for the buffer in the already allocated BOs attached
+	 * to the stream.
 	 */
-	usleep_range(15000, 20000);
+	llist_for_each_entry(oa_bo, stream->oa_config_bos.first, node) {
+		if (oa_bo->oa_config == oa_config &&
+		    memcmp(oa_bo->oa_config->uuid,
+			   oa_config->uuid,
+			   sizeof(oa_config->uuid)) == 0)
+			goto out;
+	}
+
+	oa_bo = alloc_oa_config_buffer(stream, oa_config);
+	if (IS_ERR(oa_bo))
+		return ERR_CAST(oa_bo);
+
+out:
+	return i915_vma_get(oa_bo->vma);
+}
+
+static int emit_oa_config(struct i915_perf_stream *stream,
+			  struct i915_oa_config *oa_config,
+			  struct intel_context *ce)
+{
+	struct i915_request *rq;
+	struct i915_vma *vma;
+	int err;
+
+	vma = get_oa_vma(stream, oa_config);
+	if (IS_ERR(vma))
+		return PTR_ERR(vma);
+
+	err = i915_vma_pin(vma, 0, 0, PIN_GLOBAL | PIN_HIGH);
+	if (err)
+		goto err_vma_put;
+
+	rq = i915_request_create(ce);
+	if (IS_ERR(rq)) {
+		err = PTR_ERR(rq);
+		goto err_vma_unpin;
+	}
+
+	i915_vma_lock(vma);
+	err = i915_request_await_object(rq, vma->obj, 0);
+	if (!err)
+		err = i915_vma_move_to_active(vma, rq, 0);
+	i915_vma_unlock(vma);
+	if (err)
+		goto err_add_request;
+
+	err = rq->engine->emit_bb_start(rq,
+					vma->node.start, 0,
+					I915_DISPATCH_SECURE);
+err_add_request:
+	i915_request_add(rq);
+err_vma_unpin:
+	i915_vma_unpin(vma);
+err_vma_put:
+	i915_vma_put(vma);
+	return err;
+}
+
+static struct intel_context *oa_context(struct i915_perf_stream *stream)
+{
+	return stream->pinned_ctx ?: stream->engine->kernel_context;
 }
 
 static int hsw_enable_metric_set(struct i915_perf_stream *stream)
 {
-	struct drm_i915_private *dev_priv = stream->dev_priv;
-	const struct i915_oa_config *oa_config = stream->oa_config;
+	struct intel_uncore *uncore = stream->uncore;
 
 	/*
 	 * PRM:
@@ -1616,31 +2011,24 @@ static int hsw_enable_metric_set(struct i915_perf_stream *stream)
 	 * count the events from non-render domain. Unit level clock
 	 * gating for RCS should also be disabled.
 	 */
-	I915_WRITE(GEN7_MISCCPCTL, (I915_READ(GEN7_MISCCPCTL) &
-				    ~GEN7_DOP_CLOCK_GATE_ENABLE));
-	I915_WRITE(GEN6_UCGCTL1, (I915_READ(GEN6_UCGCTL1) |
-				  GEN6_CSUNIT_CLOCK_GATE_DISABLE));
-
-	config_oa_regs(dev_priv, oa_config->mux_regs, oa_config->mux_regs_len);
-	delay_after_mux();
-
-	config_oa_regs(dev_priv, oa_config->b_counter_regs,
-		       oa_config->b_counter_regs_len);
+	intel_uncore_rmw(uncore, GEN7_MISCCPCTL,
+			 GEN7_DOP_CLOCK_GATE_ENABLE, 0);
+	intel_uncore_rmw(uncore, GEN6_UCGCTL1,
+			 0, GEN6_CSUNIT_CLOCK_GATE_DISABLE);
 
-	return 0;
+	return emit_oa_config(stream, stream->oa_config, oa_context(stream));
 }
 
 static void hsw_disable_metric_set(struct i915_perf_stream *stream)
 {
-	struct drm_i915_private *dev_priv = stream->dev_priv;
+	struct intel_uncore *uncore = stream->uncore;
 
-	I915_WRITE(GEN6_UCGCTL1, (I915_READ(GEN6_UCGCTL1) &
-				  ~GEN6_CSUNIT_CLOCK_GATE_DISABLE));
-	I915_WRITE(GEN7_MISCCPCTL, (I915_READ(GEN7_MISCCPCTL) |
-				    GEN7_DOP_CLOCK_GATE_ENABLE));
+	intel_uncore_rmw(uncore, GEN6_UCGCTL1,
+			 GEN6_CSUNIT_CLOCK_GATE_DISABLE, 0);
+	intel_uncore_rmw(uncore, GEN7_MISCCPCTL,
+			 0, GEN7_DOP_CLOCK_GATE_ENABLE);
 
-	I915_WRITE(GDT_CHICKEN_BITS, (I915_READ(GDT_CHICKEN_BITS) &
-				      ~GT_NOA_ENABLE));
+	intel_uncore_rmw(uncore, GDT_CHICKEN_BITS, GT_NOA_ENABLE, 0);
 }
 
 static u32 oa_config_flex_reg(const struct i915_oa_config *oa_config,
@@ -1672,14 +2060,11 @@ static u32 oa_config_flex_reg(const struct i915_oa_config *oa_config,
  * in the case that the OA unit has been disabled.
  */
 static void
-gen8_update_reg_state_unlocked(struct i915_perf_stream *stream,
-			       struct intel_context *ce,
-			       u32 *reg_state,
-			       const struct i915_oa_config *oa_config)
-{
-	struct drm_i915_private *i915 = ce->engine->i915;
-	u32 ctx_oactxctrl = i915->perf.ctx_oactxctrl_offset;
-	u32 ctx_flexeu0 = i915->perf.ctx_flexeu0_offset;
+gen8_update_reg_state_unlocked(const struct intel_context *ce,
+			       const struct i915_perf_stream *stream)
+{
+	u32 ctx_oactxctrl = stream->perf->ctx_oactxctrl_offset;
+	u32 ctx_flexeu0 = stream->perf->ctx_flexeu0_offset;
 	/* The MMIO offsets for Flex EU registers aren't contiguous */
 	i915_reg_t flex_regs[] = {
 		EU_PERF_CNTL0,
@@ -1690,21 +2075,28 @@ gen8_update_reg_state_unlocked(struct i915_perf_stream *stream,
 		EU_PERF_CNTL5,
 		EU_PERF_CNTL6,
 	};
+	u32 *reg_state = ce->lrc_reg_state;
 	int i;
 
-	CTX_REG(reg_state, ctx_oactxctrl, GEN8_OACTXCONTROL,
-		(stream->period_exponent << GEN8_OA_TIMER_PERIOD_SHIFT) |
-		(stream->periodic ? GEN8_OA_TIMER_ENABLE : 0) |
-		GEN8_OA_COUNTER_RESUME);
+	if (IS_GEN(stream->perf->i915, 12)) {
+		u32 format = stream->oa_buffer.format;
 
-	for (i = 0; i < ARRAY_SIZE(flex_regs); i++) {
-		CTX_REG(reg_state, ctx_flexeu0 + i * 2, flex_regs[i],
-			oa_config_flex_reg(oa_config, flex_regs[i]));
+		reg_state[ctx_oactxctrl + 1] =
+			(format << GEN12_OAR_OACONTROL_COUNTER_FORMAT_SHIFT) |
+			(stream->oa_config ? GEN12_OAR_OACONTROL_COUNTER_ENABLE : 0);
+	} else {
+		reg_state[ctx_oactxctrl + 1] =
+			(stream->period_exponent << GEN8_OA_TIMER_PERIOD_SHIFT) |
+			(stream->periodic ? GEN8_OA_TIMER_ENABLE : 0) |
+			GEN8_OA_COUNTER_RESUME;
 	}
 
-	CTX_REG(reg_state,
-		CTX_R_PWR_CLK_STATE, GEN8_R_PWR_CLK_STATE,
-		intel_sseu_make_rpcs(i915, &ce->sseu));
+	for (i = 0; !!ctx_flexeu0 && i < ARRAY_SIZE(flex_regs); i++)
+		reg_state[ctx_flexeu0 + i * 2 + 1] =
+			oa_config_flex_reg(stream->oa_config, flex_regs[i]);
+
+	reg_state[CTX_R_PWR_CLK_STATE] =
+		intel_sseu_make_rpcs(ce->engine->i915, &ce->sseu);
 }
 
 struct flex {
@@ -1728,7 +2120,7 @@ gen8_store_flex(struct i915_request *rq,
 	offset = i915_ggtt_offset(ce->state) + LRC_STATE_PN * PAGE_SIZE;
 	do {
 		*cs++ = MI_STORE_DWORD_IMM_GEN4 | MI_USE_GGTT;
-		*cs++ = offset + (flex->offset + 1) * sizeof(u32);
+		*cs++ = offset + flex->offset * sizeof(u32);
 		*cs++ = 0;
 		*cs++ = flex->value;
 	} while (flex++, --count);
@@ -1832,6 +2224,36 @@ static int gen8_configure_context(struct i915_gem_context *ctx,
 	return err;
 }
 
+static int gen12_emit_oar_config(struct intel_context *ce, bool enable)
+{
+	struct i915_request *rq;
+	u32 *cs;
+	int err = 0;
+
+	rq = i915_request_create(ce);
+	if (IS_ERR(rq))
+		return PTR_ERR(rq);
+
+	cs = intel_ring_begin(rq, 4);
+	if (IS_ERR(cs)) {
+		err = PTR_ERR(cs);
+		goto out;
+	}
+
+	*cs++ = MI_LOAD_REGISTER_IMM(1);
+	*cs++ = i915_mmio_reg_offset(RING_CONTEXT_CONTROL(ce->engine->mmio_base));
+	*cs++ = _MASKED_FIELD(GEN12_CTX_CTRL_OAR_CONTEXT_ENABLE,
+			      enable ? GEN12_CTX_CTRL_OAR_CONTEXT_ENABLE : 0);
+	*cs++ = MI_NOOP;
+
+	intel_ring_advance(rq, cs);
+
+out:
+	i915_request_add(rq);
+
+	return err;
+}
+
 /*
  * Manages updating the per-context aspects of the OA stream
  * configuration across all contexts.
@@ -1856,24 +2278,22 @@ static int gen8_configure_context(struct i915_gem_context *ctx,
  *
  * Note: it's only the RCS/Render context that has any OA state.
  */
-static int gen8_configure_all_contexts(struct i915_perf_stream *stream,
-				       const struct i915_oa_config *oa_config)
+static int lrc_configure_all_contexts(struct i915_perf_stream *stream,
+				      const struct i915_oa_config *oa_config)
 {
-	struct drm_i915_private *i915 = stream->dev_priv;
+	struct drm_i915_private *i915 = stream->perf->i915;
 	/* The MMIO offsets for Flex EU registers aren't contiguous */
-	const u32 ctx_flexeu0 = i915->perf.ctx_flexeu0_offset;
-#define ctx_flexeuN(N) (ctx_flexeu0 + 2 * (N))
+	const u32 ctx_flexeu0 = stream->perf->ctx_flexeu0_offset;
+#define ctx_flexeuN(N) (ctx_flexeu0 + 2 * (N) + 1)
 	struct flex regs[] = {
 		{
 			GEN8_R_PWR_CLK_STATE,
 			CTX_R_PWR_CLK_STATE,
 		},
 		{
-			GEN8_OACTXCONTROL,
-			i915->perf.ctx_oactxctrl_offset,
-			((stream->period_exponent << GEN8_OA_TIMER_PERIOD_SHIFT) |
-			 (stream->periodic ? GEN8_OA_TIMER_ENABLE : 0) |
-			 GEN8_OA_COUNTER_RESUME)
+			IS_GEN(i915, 12) ?
+			GEN12_OAR_OACONTROL : GEN8_OACTXCONTROL,
+			stream->perf->ctx_oactxctrl_offset + 1,
 		},
 		{ EU_PERF_CNTL0, ctx_flexeuN(0) },
 		{ EU_PERF_CNTL1, ctx_flexeuN(1) },
@@ -1885,13 +2305,27 @@ static int gen8_configure_all_contexts(struct i915_perf_stream *stream,
 	};
 #undef ctx_flexeuN
 	struct intel_engine_cs *engine;
-	struct i915_gem_context *ctx;
-	int i;
+	struct i915_gem_context *ctx, *cn;
+	size_t array_size = IS_GEN(i915, 12) ? 2 : ARRAY_SIZE(regs);
+	int i, err;
 
-	for (i = 2; i < ARRAY_SIZE(regs); i++)
+	if (IS_GEN(i915, 12)) {
+		u32 format = stream->oa_buffer.format;
+
+		regs[1].value =
+			(format << GEN12_OAR_OACONTROL_COUNTER_FORMAT_SHIFT) |
+			(oa_config ? GEN12_OAR_OACONTROL_COUNTER_ENABLE : 0);
+	} else {
+		regs[1].value =
+			(stream->period_exponent << GEN8_OA_TIMER_PERIOD_SHIFT) |
+			(stream->periodic ? GEN8_OA_TIMER_ENABLE : 0) |
+			GEN8_OA_COUNTER_RESUME;
+	}
+
+	for (i = 2; !!ctx_flexeu0 && i < array_size; i++)
 		regs[i].value = oa_config_flex_reg(oa_config, regs[i].reg);
 
-	lockdep_assert_held(&i915->drm.struct_mutex);
+	lockdep_assert_held(&stream->perf->lock);
 
 	/*
 	 * The OA register config is setup through the context image. This image
@@ -1909,16 +2343,27 @@ static int gen8_configure_all_contexts(struct i915_perf_stream *stream,
 	 * context. Contexts idle at the time of reconfiguration are not
 	 * trapped behind the barrier.
 	 */
-	list_for_each_entry(ctx, &i915->contexts.list, link) {
-		int err;
-
+	spin_lock(&i915->gem.contexts.lock);
+	list_for_each_entry_safe(ctx, cn, &i915->gem.contexts.list, link) {
 		if (ctx == i915->kernel_context)
 			continue;
 
-		err = gen8_configure_context(ctx, regs, ARRAY_SIZE(regs));
-		if (err)
+		if (!kref_get_unless_zero(&ctx->ref))
+			continue;
+
+		spin_unlock(&i915->gem.contexts.lock);
+
+		err = gen8_configure_context(ctx, regs, array_size);
+		if (err) {
+			i915_gem_context_put(ctx);
 			return err;
+		}
+
+		spin_lock(&i915->gem.contexts.lock);
+		list_safe_reset_next(ctx, cn, link);
+		i915_gem_context_put(ctx);
 	}
+	spin_unlock(&i915->gem.contexts.lock);
 
 	/*
 	 * After updating all other contexts, we need to modify ourselves.
@@ -1927,14 +2372,13 @@ static int gen8_configure_all_contexts(struct i915_perf_stream *stream,
 	 */
 	for_each_uabi_engine(engine, i915) {
 		struct intel_context *ce = engine->kernel_context;
-		int err;
 
 		if (engine->class != RENDER_CLASS)
 			continue;
 
 		regs[0].value = intel_sseu_make_rpcs(i915, &ce->sseu);
 
-		err = gen8_modify_self(ce, regs, ARRAY_SIZE(regs));
+		err = gen8_modify_self(ce, regs, array_size);
 		if (err)
 			return err;
 	}
@@ -1944,8 +2388,8 @@ static int gen8_configure_all_contexts(struct i915_perf_stream *stream,
 
 static int gen8_enable_metric_set(struct i915_perf_stream *stream)
 {
-	struct drm_i915_private *dev_priv = stream->dev_priv;
-	const struct i915_oa_config *oa_config = stream->oa_config;
+	struct intel_uncore *uncore = stream->uncore;
+	struct i915_oa_config *oa_config = stream->oa_config;
 	int ret;
 
 	/*
@@ -1971,10 +2415,10 @@ static int gen8_enable_metric_set(struct i915_perf_stream *stream)
 	 * be read back from automatically triggered reports, as part of the
 	 * RPT_ID field.
 	 */
-	if (IS_GEN_RANGE(dev_priv, 9, 11)) {
-		I915_WRITE(GEN8_OA_DEBUG,
-			   _MASKED_BIT_ENABLE(GEN9_OA_DEBUG_DISABLE_CLK_RATIO_REPORTS |
-					      GEN9_OA_DEBUG_INCLUDE_CLK_RATIO));
+	if (IS_GEN_RANGE(stream->perf->i915, 9, 11)) {
+		intel_uncore_write(uncore, GEN8_OA_DEBUG,
+				   _MASKED_BIT_ENABLE(GEN9_OA_DEBUG_DISABLE_CLK_RATIO_REPORTS |
+						      GEN9_OA_DEBUG_INCLUDE_CLK_RATIO));
 	}
 
 	/*
@@ -1982,45 +2426,102 @@ static int gen8_enable_metric_set(struct i915_perf_stream *stream)
 	 * to make sure all slices/subslices are ON before writing to NOA
 	 * registers.
 	 */
-	ret = gen8_configure_all_contexts(stream, oa_config);
+	ret = lrc_configure_all_contexts(stream, oa_config);
 	if (ret)
 		return ret;
 
-	config_oa_regs(dev_priv, oa_config->mux_regs, oa_config->mux_regs_len);
-	delay_after_mux();
+	return emit_oa_config(stream, oa_config, oa_context(stream));
+}
+
+static int gen12_enable_metric_set(struct i915_perf_stream *stream)
+{
+	struct intel_uncore *uncore = stream->uncore;
+	struct i915_oa_config *oa_config = stream->oa_config;
+	bool periodic = stream->periodic;
+	u32 period_exponent = stream->period_exponent;
+	int ret;
 
-	config_oa_regs(dev_priv, oa_config->b_counter_regs,
-		       oa_config->b_counter_regs_len);
+	intel_uncore_write(uncore, GEN12_OAG_OA_DEBUG,
+			   /* Disable clk ratio reports, like previous Gens. */
+			   _MASKED_BIT_ENABLE(GEN12_OAG_OA_DEBUG_DISABLE_CLK_RATIO_REPORTS |
+					      GEN12_OAG_OA_DEBUG_INCLUDE_CLK_RATIO) |
+			   /*
+			    * If the user didn't require OA reports, instruct the
+			    * hardware not to emit ctx switch reports.
+			    */
+			   !(stream->sample_flags & SAMPLE_OA_REPORT) ?
+			   _MASKED_BIT_ENABLE(GEN12_OAG_OA_DEBUG_DISABLE_CTX_SWITCH_REPORTS) :
+			   _MASKED_BIT_DISABLE(GEN12_OAG_OA_DEBUG_DISABLE_CTX_SWITCH_REPORTS));
+
+	intel_uncore_write(uncore, GEN12_OAG_OAGLBCTXCTRL, periodic ?
+			   (GEN12_OAG_OAGLBCTXCTRL_COUNTER_RESUME |
+			    GEN12_OAG_OAGLBCTXCTRL_TIMER_ENABLE |
+			    (period_exponent << GEN12_OAG_OAGLBCTXCTRL_TIMER_PERIOD_SHIFT))
+			    : 0);
 
-	return 0;
+	/*
+	 * Update all contexts prior writing the mux configurations as we need
+	 * to make sure all slices/subslices are ON before writing to NOA
+	 * registers.
+	 */
+	ret = lrc_configure_all_contexts(stream, oa_config);
+	if (ret)
+		return ret;
+
+	/*
+	 * For Gen12, performance counters are context
+	 * saved/restored. Only enable it for the context that
+	 * requested this.
+	 */
+	if (stream->ctx) {
+		ret = gen12_emit_oar_config(stream->pinned_ctx,
+					    oa_config != NULL);
+		if (ret)
+			return ret;
+	}
+
+	return emit_oa_config(stream, oa_config, oa_context(stream));
 }
 
 static void gen8_disable_metric_set(struct i915_perf_stream *stream)
 {
-	struct drm_i915_private *dev_priv = stream->dev_priv;
+	struct intel_uncore *uncore = stream->uncore;
 
 	/* Reset all contexts' slices/subslices configurations. */
-	gen8_configure_all_contexts(stream, NULL);
+	lrc_configure_all_contexts(stream, NULL);
 
-	I915_WRITE(GDT_CHICKEN_BITS, (I915_READ(GDT_CHICKEN_BITS) &
-				      ~GT_NOA_ENABLE));
+	intel_uncore_rmw(uncore, GDT_CHICKEN_BITS, GT_NOA_ENABLE, 0);
 }
 
 static void gen10_disable_metric_set(struct i915_perf_stream *stream)
 {
-	struct drm_i915_private *dev_priv = stream->dev_priv;
+	struct intel_uncore *uncore = stream->uncore;
+
+	/* Reset all contexts' slices/subslices configurations. */
+	lrc_configure_all_contexts(stream, NULL);
+
+	/* Make sure we disable noa to save power. */
+	intel_uncore_rmw(uncore, RPM_CONFIG1, GEN10_GT_NOA_ENABLE, 0);
+}
+
+static void gen12_disable_metric_set(struct i915_perf_stream *stream)
+{
+	struct intel_uncore *uncore = stream->uncore;
 
 	/* Reset all contexts' slices/subslices configurations. */
-	gen8_configure_all_contexts(stream, NULL);
+	lrc_configure_all_contexts(stream, NULL);
+
+	/* disable the context save/restore or OAR counters */
+	if (stream->ctx)
+		gen12_emit_oar_config(stream->pinned_ctx, false);
 
 	/* Make sure we disable noa to save power. */
-	I915_WRITE(RPM_CONFIG1,
-		   I915_READ(RPM_CONFIG1) & ~GEN10_GT_NOA_ENABLE);
+	intel_uncore_rmw(uncore, RPM_CONFIG1, GEN10_GT_NOA_ENABLE, 0);
 }
 
 static void gen7_oa_enable(struct i915_perf_stream *stream)
 {
-	struct drm_i915_private *dev_priv = stream->dev_priv;
+	struct intel_uncore *uncore = stream->uncore;
 	struct i915_gem_context *ctx = stream->ctx;
 	u32 ctx_id = stream->specific_ctx_id;
 	bool periodic = stream->periodic;
@@ -2038,19 +2539,19 @@ static void gen7_oa_enable(struct i915_perf_stream *stream)
 	 */
 	gen7_init_oa_buffer(stream);
 
-	I915_WRITE(GEN7_OACONTROL,
-		   (ctx_id & GEN7_OACONTROL_CTX_MASK) |
-		   (period_exponent <<
-		    GEN7_OACONTROL_TIMER_PERIOD_SHIFT) |
-		   (periodic ? GEN7_OACONTROL_TIMER_ENABLE : 0) |
-		   (report_format << GEN7_OACONTROL_FORMAT_SHIFT) |
-		   (ctx ? GEN7_OACONTROL_PER_CTX_ENABLE : 0) |
-		   GEN7_OACONTROL_ENABLE);
+	intel_uncore_write(uncore, GEN7_OACONTROL,
+			   (ctx_id & GEN7_OACONTROL_CTX_MASK) |
+			   (period_exponent <<
+			    GEN7_OACONTROL_TIMER_PERIOD_SHIFT) |
+			   (periodic ? GEN7_OACONTROL_TIMER_ENABLE : 0) |
+			   (report_format << GEN7_OACONTROL_FORMAT_SHIFT) |
+			   (ctx ? GEN7_OACONTROL_PER_CTX_ENABLE : 0) |
+			   GEN7_OACONTROL_ENABLE);
 }
 
 static void gen8_oa_enable(struct i915_perf_stream *stream)
 {
-	struct drm_i915_private *dev_priv = stream->dev_priv;
+	struct intel_uncore *uncore = stream->uncore;
 	u32 report_format = stream->oa_buffer.format;
 
 	/*
@@ -2069,9 +2570,28 @@ static void gen8_oa_enable(struct i915_perf_stream *stream)
 	 * filtering and instead filter on the cpu based on the context-id
 	 * field of reports
 	 */
-	I915_WRITE(GEN8_OACONTROL, (report_format <<
-				    GEN8_OA_REPORT_FORMAT_SHIFT) |
-				   GEN8_OA_COUNTER_ENABLE);
+	intel_uncore_write(uncore, GEN8_OACONTROL,
+			   (report_format << GEN8_OA_REPORT_FORMAT_SHIFT) |
+			   GEN8_OA_COUNTER_ENABLE);
+}
+
+static void gen12_oa_enable(struct i915_perf_stream *stream)
+{
+	struct intel_uncore *uncore = stream->uncore;
+	u32 report_format = stream->oa_buffer.format;
+
+	/*
+	 * If we don't want OA reports from the OA buffer, then we don't even
+	 * need to program the OAG unit.
+	 */
+	if (!(stream->sample_flags & SAMPLE_OA_REPORT))
+		return;
+
+	gen12_init_oa_buffer(stream);
+
+	intel_uncore_write(uncore, GEN12_OAG_OACONTROL,
+			   (report_format << GEN12_OAG_OACONTROL_OA_COUNTER_FORMAT_SHIFT) |
+			   GEN12_OAG_OACONTROL_OA_COUNTER_ENABLE);
 }
 
 /**
@@ -2085,9 +2605,7 @@ static void gen8_oa_enable(struct i915_perf_stream *stream)
  */
 static void i915_oa_stream_enable(struct i915_perf_stream *stream)
 {
-	struct drm_i915_private *dev_priv = stream->dev_priv;
-
-	dev_priv->perf.ops.oa_enable(stream);
+	stream->perf->ops.oa_enable(stream);
 
 	if (stream->periodic)
 		hrtimer_start(&stream->poll_check_timer,
@@ -2097,7 +2615,7 @@ static void i915_oa_stream_enable(struct i915_perf_stream *stream)
 
 static void gen7_oa_disable(struct i915_perf_stream *stream)
 {
-	struct intel_uncore *uncore = &stream->dev_priv->uncore;
+	struct intel_uncore *uncore = stream->uncore;
 
 	intel_uncore_write(uncore, GEN7_OACONTROL, 0);
 	if (intel_wait_for_register(uncore,
@@ -2108,7 +2626,7 @@ static void gen7_oa_disable(struct i915_perf_stream *stream)
 
 static void gen8_oa_disable(struct i915_perf_stream *stream)
 {
-	struct intel_uncore *uncore = &stream->dev_priv->uncore;
+	struct intel_uncore *uncore = stream->uncore;
 
 	intel_uncore_write(uncore, GEN8_OACONTROL, 0);
 	if (intel_wait_for_register(uncore,
@@ -2117,6 +2635,18 @@ static void gen8_oa_disable(struct i915_perf_stream *stream)
 		DRM_ERROR("wait for OA to be disabled timed out\n");
 }
 
+static void gen12_oa_disable(struct i915_perf_stream *stream)
+{
+	struct intel_uncore *uncore = stream->uncore;
+
+	intel_uncore_write(uncore, GEN12_OAG_OACONTROL, 0);
+	if (intel_wait_for_register(uncore,
+				    GEN12_OAG_OACONTROL,
+				    GEN12_OAG_OACONTROL_OA_COUNTER_ENABLE, 0,
+				    50))
+		DRM_ERROR("wait for OA to be disabled timed out\n");
+}
+
 /**
  * i915_oa_stream_disable - handle `I915_PERF_IOCTL_DISABLE` for OA stream
  * @stream: An i915 perf stream opened for OA metrics
@@ -2127,9 +2657,7 @@ static void gen8_oa_disable(struct i915_perf_stream *stream)
  */
 static void i915_oa_stream_disable(struct i915_perf_stream *stream)
 {
-	struct drm_i915_private *dev_priv = stream->dev_priv;
-
-	dev_priv->perf.ops.oa_disable(stream);
+	stream->perf->ops.oa_disable(stream);
 
 	if (stream->periodic)
 		hrtimer_cancel(&stream->poll_check_timer);
@@ -2166,15 +2694,21 @@ static int i915_oa_stream_init(struct i915_perf_stream *stream,
 			       struct drm_i915_perf_open_param *param,
 			       struct perf_open_properties *props)
 {
-	struct drm_i915_private *dev_priv = stream->dev_priv;
+	struct i915_perf *perf = stream->perf;
 	int format_size;
 	int ret;
 
-	/* If the sysfs metrics/ directory wasn't registered for some
+	if (!props->engine) {
+		DRM_DEBUG("OA engine not specified\n");
+		return -EINVAL;
+	}
+
+	/*
+	 * If the sysfs metrics/ directory wasn't registered for some
 	 * reason then don't let userspace try their luck with config
 	 * IDs
 	 */
-	if (!dev_priv->perf.metrics_kobj) {
+	if (!perf->metrics_kobj) {
 		DRM_DEBUG("OA metrics weren't advertised via sysfs\n");
 		return -EINVAL;
 	}
@@ -2184,16 +2718,17 @@ static int i915_oa_stream_init(struct i915_perf_stream *stream,
 		return -EINVAL;
 	}
 
-	if (!dev_priv->perf.ops.enable_metric_set) {
+	if (!perf->ops.enable_metric_set) {
 		DRM_DEBUG("OA unit not supported\n");
 		return -ENODEV;
 	}
 
-	/* To avoid the complexity of having to accurately filter
+	/*
+	 * To avoid the complexity of having to accurately filter
 	 * counter reports and marshal to the appropriate client
 	 * we currently only allow exclusive access
 	 */
-	if (dev_priv->perf.exclusive_stream) {
+	if (perf->exclusive_stream) {
 		DRM_DEBUG("OA unit already in use\n");
 		return -EBUSY;
 	}
@@ -2203,9 +2738,12 @@ static int i915_oa_stream_init(struct i915_perf_stream *stream,
 		return -EINVAL;
 	}
 
+	stream->engine = props->engine;
+	stream->uncore = stream->engine->gt->uncore;
+
 	stream->sample_size = sizeof(struct drm_i915_perf_record_header);
 
-	format_size = dev_priv->perf.oa_formats[props->oa_format].size;
+	format_size = perf->oa_formats[props->oa_format].size;
 
 	stream->sample_flags |= SAMPLE_OA_REPORT;
 	stream->sample_size += format_size;
@@ -2214,8 +2752,10 @@ static int i915_oa_stream_init(struct i915_perf_stream *stream,
 	if (WARN_ON(stream->oa_buffer.format_size == 0))
 		return -EINVAL;
 
+	stream->hold_preemption = props->hold_preemption;
+
 	stream->oa_buffer.format =
-		dev_priv->perf.oa_formats[props->oa_format].format;
+		perf->oa_formats[props->oa_format].format;
 
 	stream->periodic = props->oa_periodic;
 	if (stream->periodic)
@@ -2229,9 +2769,16 @@ static int i915_oa_stream_init(struct i915_perf_stream *stream,
 		}
 	}
 
-	ret = get_oa_config(dev_priv, props->metrics_set, &stream->oa_config);
+	ret = alloc_noa_wait(stream);
 	if (ret) {
+		DRM_DEBUG("Unable to allocate NOA wait batch buffer\n");
+		goto err_noa_wait_alloc;
+	}
+
+	stream->oa_config = i915_perf_get_oa_config(perf, props->metrics_set);
+	if (!stream->oa_config) {
 		DRM_DEBUG("Invalid OA config id=%i\n", props->metrics_set);
+		ret = -EINVAL;
 		goto err_config;
 	}
 
@@ -2247,27 +2794,24 @@ static int i915_oa_stream_init(struct i915_perf_stream *stream,
 	 *   In our case we are expecting that taking pm + FORCEWAKE
 	 *   references will effectively disable RC6.
 	 */
-	stream->wakeref = intel_runtime_pm_get(&dev_priv->runtime_pm);
-	intel_uncore_forcewake_get(&dev_priv->uncore, FORCEWAKE_ALL);
+	intel_engine_pm_get(stream->engine);
+	intel_uncore_forcewake_get(stream->uncore, FORCEWAKE_ALL);
 
 	ret = alloc_oa_buffer(stream);
 	if (ret)
 		goto err_oa_buf_alloc;
 
-	ret = i915_mutex_lock_interruptible(&dev_priv->drm);
-	if (ret)
-		goto err_lock;
-
 	stream->ops = &i915_oa_stream_ops;
-	dev_priv->perf.exclusive_stream = stream;
+	perf->exclusive_stream = stream;
 
-	ret = dev_priv->perf.ops.enable_metric_set(stream);
+	ret = perf->ops.enable_metric_set(stream);
 	if (ret) {
 		DRM_DEBUG("Unable to enable metric set\n");
 		goto err_enable;
 	}
 
-	mutex_unlock(&dev_priv->drm.struct_mutex);
+	DRM_DEBUG("opening stream oa config uuid=%s\n",
+		  stream->oa_config->uuid);
 
 	hrtimer_init(&stream->poll_check_timer,
 		     CLOCK_MONOTONIC, HRTIMER_MODE_REL);
@@ -2278,38 +2822,40 @@ static int i915_oa_stream_init(struct i915_perf_stream *stream,
 	return 0;
 
 err_enable:
-	dev_priv->perf.exclusive_stream = NULL;
-	dev_priv->perf.ops.disable_metric_set(stream);
-	mutex_unlock(&dev_priv->drm.struct_mutex);
+	perf->exclusive_stream = NULL;
+	perf->ops.disable_metric_set(stream);
 
-err_lock:
 	free_oa_buffer(stream);
 
 err_oa_buf_alloc:
-	put_oa_config(dev_priv, stream->oa_config);
+	free_oa_configs(stream);
 
-	intel_uncore_forcewake_put(&dev_priv->uncore, FORCEWAKE_ALL);
-	intel_runtime_pm_put(&dev_priv->runtime_pm, stream->wakeref);
+	intel_uncore_forcewake_put(stream->uncore, FORCEWAKE_ALL);
+	intel_engine_pm_put(stream->engine);
 
 err_config:
+	free_noa_wait(stream);
+
+err_noa_wait_alloc:
 	if (stream->ctx)
 		oa_put_render_ctx_id(stream);
 
 	return ret;
 }
 
-void i915_oa_init_reg_state(struct intel_engine_cs *engine,
-			    struct intel_context *ce,
-			    u32 *regs)
+void i915_oa_init_reg_state(const struct intel_context *ce,
+			    const struct intel_engine_cs *engine)
 {
 	struct i915_perf_stream *stream;
 
+	/* perf.exclusive_stream serialised by lrc_configure_all_contexts() */
+
 	if (engine->class != RENDER_CLASS)
 		return;
 
 	stream = engine->i915->perf.exclusive_stream;
 	if (stream)
-		gen8_update_reg_state_unlocked(stream, ce, regs, stream->oa_config);
+		gen8_update_reg_state_unlocked(ce, stream);
 }
 
 /**
@@ -2379,7 +2925,7 @@ static ssize_t i915_perf_read(struct file *file,
 			      loff_t *ppos)
 {
 	struct i915_perf_stream *stream = file->private_data;
-	struct drm_i915_private *dev_priv = stream->dev_priv;
+	struct i915_perf *perf = stream->perf;
 	ssize_t ret;
 
 	/* To ensure it's handled consistently we simply treat all reads of a
@@ -2402,15 +2948,15 @@ static ssize_t i915_perf_read(struct file *file,
 			if (ret)
 				return ret;
 
-			mutex_lock(&dev_priv->perf.lock);
+			mutex_lock(&perf->lock);
 			ret = i915_perf_read_locked(stream, file,
 						    buf, count, ppos);
-			mutex_unlock(&dev_priv->perf.lock);
+			mutex_unlock(&perf->lock);
 		} while (ret == -EAGAIN);
 	} else {
-		mutex_lock(&dev_priv->perf.lock);
+		mutex_lock(&perf->lock);
 		ret = i915_perf_read_locked(stream, file, buf, count, ppos);
-		mutex_unlock(&dev_priv->perf.lock);
+		mutex_unlock(&perf->lock);
 	}
 
 	/* We allow the poll checking to sometimes report false positive EPOLLIN
@@ -2448,7 +2994,6 @@ static enum hrtimer_restart oa_poll_check_timer_cb(struct hrtimer *hrtimer)
 
 /**
  * i915_perf_poll_locked - poll_wait() with a suitable wait queue for stream
- * @dev_priv: i915 device instance
  * @stream: An i915 perf stream
  * @file: An i915 perf stream file
  * @wait: poll() state table
@@ -2457,15 +3002,14 @@ static enum hrtimer_restart oa_poll_check_timer_cb(struct hrtimer *hrtimer)
  * &i915_perf_stream_ops->poll_wait to call poll_wait() with a wait queue that
  * will be woken for new stream data.
  *
- * Note: The &drm_i915_private->perf.lock mutex has been taken to serialize
+ * Note: The &perf->lock mutex has been taken to serialize
  * with any non-file-operation driver hooks.
  *
  * Returns: any poll events that are ready without sleeping
  */
-static __poll_t i915_perf_poll_locked(struct drm_i915_private *dev_priv,
-					  struct i915_perf_stream *stream,
-					  struct file *file,
-					  poll_table *wait)
+static __poll_t i915_perf_poll_locked(struct i915_perf_stream *stream,
+				      struct file *file,
+				      poll_table *wait)
 {
 	__poll_t events = 0;
 
@@ -2499,12 +3043,12 @@ static __poll_t i915_perf_poll_locked(struct drm_i915_private *dev_priv,
 static __poll_t i915_perf_poll(struct file *file, poll_table *wait)
 {
 	struct i915_perf_stream *stream = file->private_data;
-	struct drm_i915_private *dev_priv = stream->dev_priv;
+	struct i915_perf *perf = stream->perf;
 	__poll_t ret;
 
-	mutex_lock(&dev_priv->perf.lock);
-	ret = i915_perf_poll_locked(dev_priv, stream, file, wait);
-	mutex_unlock(&dev_priv->perf.lock);
+	mutex_lock(&perf->lock);
+	ret = i915_perf_poll_locked(stream, file, wait);
+	mutex_unlock(&perf->lock);
 
 	return ret;
 }
@@ -2529,6 +3073,9 @@ static void i915_perf_enable_locked(struct i915_perf_stream *stream)
 
 	if (stream->ops->enable)
 		stream->ops->enable(stream);
+
+	if (stream->hold_preemption)
+		i915_gem_context_set_nopreempt(stream->ctx);
 }
 
 /**
@@ -2553,17 +3100,54 @@ static void i915_perf_disable_locked(struct i915_perf_stream *stream)
 	/* Allow stream->ops->disable() to refer to this */
 	stream->enabled = false;
 
+	if (stream->hold_preemption)
+		i915_gem_context_clear_nopreempt(stream->ctx);
+
 	if (stream->ops->disable)
 		stream->ops->disable(stream);
 }
 
+static long i915_perf_config_locked(struct i915_perf_stream *stream,
+				    unsigned long metrics_set)
+{
+	struct i915_oa_config *config;
+	long ret = stream->oa_config->id;
+
+	config = i915_perf_get_oa_config(stream->perf, metrics_set);
+	if (!config)
+		return -EINVAL;
+
+	if (config != stream->oa_config) {
+		int err;
+
+		/*
+		 * If OA is bound to a specific context, emit the
+		 * reconfiguration inline from that context. The update
+		 * will then be ordered with respect to submission on that
+		 * context.
+		 *
+		 * When set globally, we use a low priority kernel context,
+		 * so it will effectively take effect when idle.
+		 */
+		err = emit_oa_config(stream, config, oa_context(stream));
+		if (err == 0)
+			config = xchg(&stream->oa_config, config);
+		else
+			ret = err;
+	}
+
+	i915_oa_config_put(config);
+
+	return ret;
+}
+
 /**
  * i915_perf_ioctl - support ioctl() usage with i915 perf stream FDs
  * @stream: An i915 perf stream
  * @cmd: the ioctl request
  * @arg: the ioctl data
  *
- * Note: The &drm_i915_private->perf.lock mutex has been taken to serialize
+ * Note: The &perf->lock mutex has been taken to serialize
  * with any non-file-operation driver hooks.
  *
  * Returns: zero on success or a negative error code. Returns -EINVAL for
@@ -2580,6 +3164,8 @@ static long i915_perf_ioctl_locked(struct i915_perf_stream *stream,
 	case I915_PERF_IOCTL_DISABLE:
 		i915_perf_disable_locked(stream);
 		return 0;
+	case I915_PERF_IOCTL_CONFIG:
+		return i915_perf_config_locked(stream, arg);
 	}
 
 	return -EINVAL;
@@ -2601,12 +3187,12 @@ static long i915_perf_ioctl(struct file *file,
 			    unsigned long arg)
 {
 	struct i915_perf_stream *stream = file->private_data;
-	struct drm_i915_private *dev_priv = stream->dev_priv;
+	struct i915_perf *perf = stream->perf;
 	long ret;
 
-	mutex_lock(&dev_priv->perf.lock);
+	mutex_lock(&perf->lock);
 	ret = i915_perf_ioctl_locked(stream, cmd, arg);
-	mutex_unlock(&dev_priv->perf.lock);
+	mutex_unlock(&perf->lock);
 
 	return ret;
 }
@@ -2618,7 +3204,7 @@ static long i915_perf_ioctl(struct file *file,
  * Frees all resources associated with the given i915 perf @stream, disabling
  * any associated data capture in the process.
  *
- * Note: The &drm_i915_private->perf.lock mutex has been taken to serialize
+ * Note: The &perf->lock mutex has been taken to serialize
  * with any non-file-operation driver hooks.
  */
 static void i915_perf_destroy_locked(struct i915_perf_stream *stream)
@@ -2629,8 +3215,6 @@ static void i915_perf_destroy_locked(struct i915_perf_stream *stream)
 	if (stream->ops->destroy)
 		stream->ops->destroy(stream);
 
-	list_del(&stream->link);
-
 	if (stream->ctx)
 		i915_gem_context_put(stream->ctx);
 
@@ -2651,14 +3235,14 @@ static void i915_perf_destroy_locked(struct i915_perf_stream *stream)
 static int i915_perf_release(struct inode *inode, struct file *file)
 {
 	struct i915_perf_stream *stream = file->private_data;
-	struct drm_i915_private *dev_priv = stream->dev_priv;
+	struct i915_perf *perf = stream->perf;
 
-	mutex_lock(&dev_priv->perf.lock);
+	mutex_lock(&perf->lock);
 	i915_perf_destroy_locked(stream);
-	mutex_unlock(&dev_priv->perf.lock);
+	mutex_unlock(&perf->lock);
 
 	/* Release the reference the perf stream kept on the driver. */
-	drm_dev_put(&dev_priv->drm);
+	drm_dev_put(&perf->i915->drm);
 
 	return 0;
 }
@@ -2680,7 +3264,7 @@ static const struct file_operations fops = {
 
 /**
  * i915_perf_open_ioctl_locked - DRM ioctl() for userspace to open a stream FD
- * @dev_priv: i915 device instance
+ * @perf: i915 perf instance
  * @param: The open parameters passed to 'DRM_I915_PERF_OPEN`
  * @props: individually validated u64 property value pairs
  * @file: drm file
@@ -2688,7 +3272,7 @@ static const struct file_operations fops = {
  * See i915_perf_ioctl_open() for interface details.
  *
  * Implements further stream config validation and stream initialization on
- * behalf of i915_perf_open_ioctl() with the &drm_i915_private->perf.lock mutex
+ * behalf of i915_perf_open_ioctl() with the &perf->lock mutex
  * taken to serialize with any non-file-operation driver hooks.
  *
  * Note: at this point the @props have only been validated in isolation and
@@ -2703,7 +3287,7 @@ static const struct file_operations fops = {
  * Returns: zero on success or a negative error code.
  */
 static int
-i915_perf_open_ioctl_locked(struct drm_i915_private *dev_priv,
+i915_perf_open_ioctl_locked(struct i915_perf *perf,
 			    struct drm_i915_perf_open_param *param,
 			    struct perf_open_properties *props,
 			    struct drm_file *file)
@@ -2734,17 +3318,34 @@ i915_perf_open_ioctl_locked(struct drm_i915_private *dev_priv,
 	 * rest of the system, which we consider acceptable for a
 	 * non-privileged client.
 	 *
-	 * For Gen8+ the OA unit no longer supports clock gating off for a
+	 * For Gen8->11 the OA unit no longer supports clock gating off for a
 	 * specific context and the kernel can't securely stop the counters
 	 * from updating as system-wide / global values. Even though we can
 	 * filter reports based on the included context ID we can't block
 	 * clients from seeing the raw / global counter values via
 	 * MI_REPORT_PERF_COUNT commands and so consider it a privileged op to
 	 * enable the OA unit by default.
+	 *
+	 * For Gen12+ we gain a new OAR unit that only monitors the RCS on a
+	 * per context basis. So we can relax requirements there if the user
+	 * doesn't request global stream access (i.e. query based sampling
+	 * using MI_RECORD_PERF_COUNT.
 	 */
-	if (IS_HASWELL(dev_priv) && specific_ctx)
+	if (IS_HASWELL(perf->i915) && specific_ctx)
+		privileged_op = false;
+	else if (IS_GEN(perf->i915, 12) && specific_ctx &&
+		 (props->sample_flags & SAMPLE_OA_REPORT) == 0)
 		privileged_op = false;
 
+	if (props->hold_preemption) {
+		if (!props->single_context) {
+			DRM_DEBUG("preemption disable with no context\n");
+			ret = -EINVAL;
+			goto err;
+		}
+		privileged_op = true;
+	}
+
 	/* Similar to perf's kernel.perf_paranoid_cpu sysctl option
 	 * we check a dev.i915.perf_stream_paranoid sysctl option
 	 * to determine if it's ok to access system wide OA counters
@@ -2752,7 +3353,7 @@ i915_perf_open_ioctl_locked(struct drm_i915_private *dev_priv,
 	 */
 	if (privileged_op &&
 	    i915_perf_stream_paranoid && !capable(CAP_SYS_ADMIN)) {
-		DRM_DEBUG("Insufficient privileges to open system-wide i915 perf stream\n");
+		DRM_DEBUG("Insufficient privileges to open i915 perf stream\n");
 		ret = -EACCES;
 		goto err_ctx;
 	}
@@ -2763,7 +3364,7 @@ i915_perf_open_ioctl_locked(struct drm_i915_private *dev_priv,
 		goto err_ctx;
 	}
 
-	stream->dev_priv = dev_priv;
+	stream->perf = perf;
 	stream->ctx = specific_ctx;
 
 	ret = i915_oa_stream_init(stream, param, props);
@@ -2779,8 +3380,6 @@ i915_perf_open_ioctl_locked(struct drm_i915_private *dev_priv,
 		goto err_flags;
 	}
 
-	list_add(&stream->link, &dev_priv->perf.streams);
-
 	if (param->flags & I915_PERF_FLAG_FD_CLOEXEC)
 		f_flags |= O_CLOEXEC;
 	if (param->flags & I915_PERF_FLAG_FD_NONBLOCK)
@@ -2789,7 +3388,7 @@ i915_perf_open_ioctl_locked(struct drm_i915_private *dev_priv,
 	stream_fd = anon_inode_getfd("[i915_perf]", &fops, stream, f_flags);
 	if (stream_fd < 0) {
 		ret = stream_fd;
-		goto err_open;
+		goto err_flags;
 	}
 
 	if (!(param->flags & I915_PERF_FLAG_DISABLED))
@@ -2798,12 +3397,10 @@ i915_perf_open_ioctl_locked(struct drm_i915_private *dev_priv,
 	/* Take a reference on the driver that will be kept with stream_fd
 	 * until its release.
 	 */
-	drm_dev_get(&dev_priv->drm);
+	drm_dev_get(&perf->i915->drm);
 
 	return stream_fd;
 
-err_open:
-	list_del(&stream->link);
 err_flags:
 	if (stream->ops->destroy)
 		stream->ops->destroy(stream);
@@ -2816,15 +3413,15 @@ err:
 	return ret;
 }
 
-static u64 oa_exponent_to_ns(struct drm_i915_private *dev_priv, int exponent)
+static u64 oa_exponent_to_ns(struct i915_perf *perf, int exponent)
 {
 	return div64_u64(1000000000ULL * (2ULL << exponent),
-			 1000ULL * RUNTIME_INFO(dev_priv)->cs_timestamp_frequency_khz);
+			 1000ULL * RUNTIME_INFO(perf->i915)->cs_timestamp_frequency_khz);
 }
 
 /**
  * read_properties_unlocked - validate + copy userspace stream open properties
- * @dev_priv: i915 device instance
+ * @perf: i915 perf instance
  * @uprops: The array of u64 key value pairs given by userspace
  * @n_props: The number of key value pairs expected in @uprops
  * @props: The stream configuration built up while validating properties
@@ -2837,7 +3434,7 @@ static u64 oa_exponent_to_ns(struct drm_i915_private *dev_priv, int exponent)
  * we shouldn't validate or assume anything about ordering here. This doesn't
  * rule out defining new properties with ordering requirements in the future.
  */
-static int read_properties_unlocked(struct drm_i915_private *dev_priv,
+static int read_properties_unlocked(struct i915_perf *perf,
 				    u64 __user *uprops,
 				    u32 n_props,
 				    struct perf_open_properties *props)
@@ -2852,6 +3449,15 @@ static int read_properties_unlocked(struct drm_i915_private *dev_priv,
 		return -EINVAL;
 	}
 
+	/* At the moment we only support using i915-perf on the RCS. */
+	props->engine = intel_engine_lookup_user(perf->i915,
+						 I915_ENGINE_CLASS_RENDER,
+						 0);
+	if (!props->engine) {
+		DRM_DEBUG("No RENDER-capable engines\n");
+		return -EINVAL;
+	}
+
 	/* Considering that ID = 0 is reserved and assuming that we don't
 	 * (currently) expect any configurations to ever specify duplicate
 	 * values for a particular property ID then the last _PROP_MAX value is
@@ -2903,7 +3509,7 @@ static int read_properties_unlocked(struct drm_i915_private *dev_priv,
 					  value);
 				return -EINVAL;
 			}
-			if (!dev_priv->perf.oa_formats[value].size) {
+			if (!perf->oa_formats[value].size) {
 				DRM_DEBUG("Unsupported OA report format %llu\n",
 					  value);
 				return -EINVAL;
@@ -2924,7 +3530,7 @@ static int read_properties_unlocked(struct drm_i915_private *dev_priv,
 			 */
 
 			BUILD_BUG_ON(sizeof(oa_period) != 8);
-			oa_period = oa_exponent_to_ns(dev_priv, value);
+			oa_period = oa_exponent_to_ns(perf, value);
 
 			/* This check is primarily to ensure that oa_period <=
 			 * UINT32_MAX (before passing to do_div which only
@@ -2949,6 +3555,9 @@ static int read_properties_unlocked(struct drm_i915_private *dev_priv,
 			props->oa_periodic = true;
 			props->oa_period_exponent = value;
 			break;
+		case DRM_I915_PERF_PROP_HOLD_PREEMPTION:
+			props->hold_preemption = !!value;
+			break;
 		case DRM_I915_PERF_PROP_MAX:
 			MISSING_CASE(id);
 			return -EINVAL;
@@ -2978,7 +3587,7 @@ static int read_properties_unlocked(struct drm_i915_private *dev_priv,
  * mutex to avoid an awkward lockdep with mmap_sem.
  *
  * Most of the implementation details are handled by
- * i915_perf_open_ioctl_locked() after taking the &drm_i915_private->perf.lock
+ * i915_perf_open_ioctl_locked() after taking the &perf->lock
  * mutex for serializing with any non-file-operation driver hooks.
  *
  * Return: A newly opened i915 Perf stream file descriptor or negative
@@ -2987,13 +3596,13 @@ static int read_properties_unlocked(struct drm_i915_private *dev_priv,
 int i915_perf_open_ioctl(struct drm_device *dev, void *data,
 			 struct drm_file *file)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct i915_perf *perf = &to_i915(dev)->perf;
 	struct drm_i915_perf_open_param *param = data;
 	struct perf_open_properties props;
 	u32 known_open_flags;
 	int ret;
 
-	if (!dev_priv->perf.initialized) {
+	if (!perf->i915) {
 		DRM_DEBUG("i915 perf interface not available for this system\n");
 		return -ENOTSUPP;
 	}
@@ -3006,124 +3615,130 @@ int i915_perf_open_ioctl(struct drm_device *dev, void *data,
 		return -EINVAL;
 	}
 
-	ret = read_properties_unlocked(dev_priv,
+	ret = read_properties_unlocked(perf,
 				       u64_to_user_ptr(param->properties_ptr),
 				       param->num_properties,
 				       &props);
 	if (ret)
 		return ret;
 
-	mutex_lock(&dev_priv->perf.lock);
-	ret = i915_perf_open_ioctl_locked(dev_priv, param, &props, file);
-	mutex_unlock(&dev_priv->perf.lock);
+	mutex_lock(&perf->lock);
+	ret = i915_perf_open_ioctl_locked(perf, param, &props, file);
+	mutex_unlock(&perf->lock);
 
 	return ret;
 }
 
 /**
  * i915_perf_register - exposes i915-perf to userspace
- * @dev_priv: i915 device instance
+ * @i915: i915 device instance
  *
  * In particular OA metric sets are advertised under a sysfs metrics/
  * directory allowing userspace to enumerate valid IDs that can be
  * used to open an i915-perf stream.
  */
-void i915_perf_register(struct drm_i915_private *dev_priv)
+void i915_perf_register(struct drm_i915_private *i915)
 {
+	struct i915_perf *perf = &i915->perf;
 	int ret;
 
-	if (!dev_priv->perf.initialized)
+	if (!perf->i915)
 		return;
 
 	/* To be sure we're synchronized with an attempted
 	 * i915_perf_open_ioctl(); considering that we register after
 	 * being exposed to userspace.
 	 */
-	mutex_lock(&dev_priv->perf.lock);
+	mutex_lock(&perf->lock);
 
-	dev_priv->perf.metrics_kobj =
+	perf->metrics_kobj =
 		kobject_create_and_add("metrics",
-				       &dev_priv->drm.primary->kdev->kobj);
-	if (!dev_priv->perf.metrics_kobj)
+				       &i915->drm.primary->kdev->kobj);
+	if (!perf->metrics_kobj)
 		goto exit;
 
-	sysfs_attr_init(&dev_priv->perf.test_config.sysfs_metric_id.attr);
-
-	if (INTEL_GEN(dev_priv) >= 11) {
-		i915_perf_load_test_config_icl(dev_priv);
-	} else if (IS_CANNONLAKE(dev_priv)) {
-		i915_perf_load_test_config_cnl(dev_priv);
-	} else if (IS_COFFEELAKE(dev_priv)) {
-		if (IS_CFL_GT2(dev_priv))
-			i915_perf_load_test_config_cflgt2(dev_priv);
-		if (IS_CFL_GT3(dev_priv))
-			i915_perf_load_test_config_cflgt3(dev_priv);
-	} else if (IS_GEMINILAKE(dev_priv)) {
-		i915_perf_load_test_config_glk(dev_priv);
-	} else if (IS_KABYLAKE(dev_priv)) {
-		if (IS_KBL_GT2(dev_priv))
-			i915_perf_load_test_config_kblgt2(dev_priv);
-		else if (IS_KBL_GT3(dev_priv))
-			i915_perf_load_test_config_kblgt3(dev_priv);
-	} else if (IS_BROXTON(dev_priv)) {
-		i915_perf_load_test_config_bxt(dev_priv);
-	} else if (IS_SKYLAKE(dev_priv)) {
-		if (IS_SKL_GT2(dev_priv))
-			i915_perf_load_test_config_sklgt2(dev_priv);
-		else if (IS_SKL_GT3(dev_priv))
-			i915_perf_load_test_config_sklgt3(dev_priv);
-		else if (IS_SKL_GT4(dev_priv))
-			i915_perf_load_test_config_sklgt4(dev_priv);
-	} else if (IS_CHERRYVIEW(dev_priv)) {
-		i915_perf_load_test_config_chv(dev_priv);
-	} else if (IS_BROADWELL(dev_priv)) {
-		i915_perf_load_test_config_bdw(dev_priv);
-	} else if (IS_HASWELL(dev_priv)) {
-		i915_perf_load_test_config_hsw(dev_priv);
-}
-
-	if (dev_priv->perf.test_config.id == 0)
+	sysfs_attr_init(&perf->test_config.sysfs_metric_id.attr);
+
+	if (IS_TIGERLAKE(i915)) {
+		i915_perf_load_test_config_tgl(i915);
+	} else if (INTEL_GEN(i915) >= 11) {
+		i915_perf_load_test_config_icl(i915);
+	} else if (IS_CANNONLAKE(i915)) {
+		i915_perf_load_test_config_cnl(i915);
+	} else if (IS_COFFEELAKE(i915)) {
+		if (IS_CFL_GT2(i915))
+			i915_perf_load_test_config_cflgt2(i915);
+		if (IS_CFL_GT3(i915))
+			i915_perf_load_test_config_cflgt3(i915);
+	} else if (IS_GEMINILAKE(i915)) {
+		i915_perf_load_test_config_glk(i915);
+	} else if (IS_KABYLAKE(i915)) {
+		if (IS_KBL_GT2(i915))
+			i915_perf_load_test_config_kblgt2(i915);
+		else if (IS_KBL_GT3(i915))
+			i915_perf_load_test_config_kblgt3(i915);
+	} else if (IS_BROXTON(i915)) {
+		i915_perf_load_test_config_bxt(i915);
+	} else if (IS_SKYLAKE(i915)) {
+		if (IS_SKL_GT2(i915))
+			i915_perf_load_test_config_sklgt2(i915);
+		else if (IS_SKL_GT3(i915))
+			i915_perf_load_test_config_sklgt3(i915);
+		else if (IS_SKL_GT4(i915))
+			i915_perf_load_test_config_sklgt4(i915);
+	} else if (IS_CHERRYVIEW(i915)) {
+		i915_perf_load_test_config_chv(i915);
+	} else if (IS_BROADWELL(i915)) {
+		i915_perf_load_test_config_bdw(i915);
+	} else if (IS_HASWELL(i915)) {
+		i915_perf_load_test_config_hsw(i915);
+	}
+
+	if (perf->test_config.id == 0)
 		goto sysfs_error;
 
-	ret = sysfs_create_group(dev_priv->perf.metrics_kobj,
-				 &dev_priv->perf.test_config.sysfs_metric);
+	ret = sysfs_create_group(perf->metrics_kobj,
+				 &perf->test_config.sysfs_metric);
 	if (ret)
 		goto sysfs_error;
 
-	atomic_set(&dev_priv->perf.test_config.ref_count, 1);
+	perf->test_config.perf = perf;
+	kref_init(&perf->test_config.ref);
 
 	goto exit;
 
 sysfs_error:
-	kobject_put(dev_priv->perf.metrics_kobj);
-	dev_priv->perf.metrics_kobj = NULL;
+	kobject_put(perf->metrics_kobj);
+	perf->metrics_kobj = NULL;
 
 exit:
-	mutex_unlock(&dev_priv->perf.lock);
+	mutex_unlock(&perf->lock);
 }
 
 /**
  * i915_perf_unregister - hide i915-perf from userspace
- * @dev_priv: i915 device instance
+ * @i915: i915 device instance
  *
  * i915-perf state cleanup is split up into an 'unregister' and
  * 'deinit' phase where the interface is first hidden from
  * userspace by i915_perf_unregister() before cleaning up
  * remaining state in i915_perf_fini().
  */
-void i915_perf_unregister(struct drm_i915_private *dev_priv)
+void i915_perf_unregister(struct drm_i915_private *i915)
 {
-	if (!dev_priv->perf.metrics_kobj)
+	struct i915_perf *perf = &i915->perf;
+
+	if (!perf->metrics_kobj)
 		return;
 
-	sysfs_remove_group(dev_priv->perf.metrics_kobj,
-			   &dev_priv->perf.test_config.sysfs_metric);
+	sysfs_remove_group(perf->metrics_kobj,
+			   &perf->test_config.sysfs_metric);
 
-	kobject_put(dev_priv->perf.metrics_kobj);
-	dev_priv->perf.metrics_kobj = NULL;
+	kobject_put(perf->metrics_kobj);
+	perf->metrics_kobj = NULL;
 }
 
-static bool gen8_is_valid_flex_addr(struct drm_i915_private *dev_priv, u32 addr)
+static bool gen8_is_valid_flex_addr(struct i915_perf *perf, u32 addr)
 {
 	static const i915_reg_t flex_eu_regs[] = {
 		EU_PERF_CNTL0,
@@ -3143,56 +3758,80 @@ static bool gen8_is_valid_flex_addr(struct drm_i915_private *dev_priv, u32 addr)
 	return false;
 }
 
-static bool gen7_is_valid_b_counter_addr(struct drm_i915_private *dev_priv, u32 addr)
+#define ADDR_IN_RANGE(addr, start, end) \
+	((addr) >= (start) && \
+	 (addr) <= (end))
+
+#define REG_IN_RANGE(addr, start, end) \
+	((addr) >= i915_mmio_reg_offset(start) && \
+	 (addr) <= i915_mmio_reg_offset(end))
+
+#define REG_EQUAL(addr, mmio) \
+	((addr) == i915_mmio_reg_offset(mmio))
+
+static bool gen7_is_valid_b_counter_addr(struct i915_perf *perf, u32 addr)
+{
+	return REG_IN_RANGE(addr, OASTARTTRIG1, OASTARTTRIG8) ||
+	       REG_IN_RANGE(addr, OAREPORTTRIG1, OAREPORTTRIG8) ||
+	       REG_IN_RANGE(addr, OACEC0_0, OACEC7_1);
+}
+
+static bool gen7_is_valid_mux_addr(struct i915_perf *perf, u32 addr)
+{
+	return REG_EQUAL(addr, HALF_SLICE_CHICKEN2) ||
+	       REG_IN_RANGE(addr, MICRO_BP0_0, NOA_WRITE) ||
+	       REG_IN_RANGE(addr, OA_PERFCNT1_LO, OA_PERFCNT2_HI) ||
+	       REG_IN_RANGE(addr, OA_PERFMATRIX_LO, OA_PERFMATRIX_HI);
+}
+
+static bool gen8_is_valid_mux_addr(struct i915_perf *perf, u32 addr)
 {
-	return (addr >= i915_mmio_reg_offset(OASTARTTRIG1) &&
-		addr <= i915_mmio_reg_offset(OASTARTTRIG8)) ||
-		(addr >= i915_mmio_reg_offset(OAREPORTTRIG1) &&
-		 addr <= i915_mmio_reg_offset(OAREPORTTRIG8)) ||
-		(addr >= i915_mmio_reg_offset(OACEC0_0) &&
-		 addr <= i915_mmio_reg_offset(OACEC7_1));
+	return gen7_is_valid_mux_addr(perf, addr) ||
+	       REG_EQUAL(addr, WAIT_FOR_RC6_EXIT) ||
+	       REG_IN_RANGE(addr, RPM_CONFIG0, NOA_CONFIG(8));
 }
 
-static bool gen7_is_valid_mux_addr(struct drm_i915_private *dev_priv, u32 addr)
+static bool gen10_is_valid_mux_addr(struct i915_perf *perf, u32 addr)
 {
-	return addr == i915_mmio_reg_offset(HALF_SLICE_CHICKEN2) ||
-		(addr >= i915_mmio_reg_offset(MICRO_BP0_0) &&
-		 addr <= i915_mmio_reg_offset(NOA_WRITE)) ||
-		(addr >= i915_mmio_reg_offset(OA_PERFCNT1_LO) &&
-		 addr <= i915_mmio_reg_offset(OA_PERFCNT2_HI)) ||
-		(addr >= i915_mmio_reg_offset(OA_PERFMATRIX_LO) &&
-		 addr <= i915_mmio_reg_offset(OA_PERFMATRIX_HI));
+	return gen8_is_valid_mux_addr(perf, addr) ||
+	       REG_EQUAL(addr, GEN10_NOA_WRITE_HIGH) ||
+	       REG_IN_RANGE(addr, OA_PERFCNT3_LO, OA_PERFCNT4_HI);
 }
 
-static bool gen8_is_valid_mux_addr(struct drm_i915_private *dev_priv, u32 addr)
+static bool hsw_is_valid_mux_addr(struct i915_perf *perf, u32 addr)
 {
-	return gen7_is_valid_mux_addr(dev_priv, addr) ||
-		addr == i915_mmio_reg_offset(WAIT_FOR_RC6_EXIT) ||
-		(addr >= i915_mmio_reg_offset(RPM_CONFIG0) &&
-		 addr <= i915_mmio_reg_offset(NOA_CONFIG(8)));
+	return gen7_is_valid_mux_addr(perf, addr) ||
+	       ADDR_IN_RANGE(addr, 0x25100, 0x2FF90) ||
+	       REG_IN_RANGE(addr, HSW_MBVID2_NOA0, HSW_MBVID2_NOA9) ||
+	       REG_EQUAL(addr, HSW_MBVID2_MISR0);
 }
 
-static bool gen10_is_valid_mux_addr(struct drm_i915_private *dev_priv, u32 addr)
+static bool chv_is_valid_mux_addr(struct i915_perf *perf, u32 addr)
 {
-	return gen8_is_valid_mux_addr(dev_priv, addr) ||
-		addr == i915_mmio_reg_offset(GEN10_NOA_WRITE_HIGH) ||
-		(addr >= i915_mmio_reg_offset(OA_PERFCNT3_LO) &&
-		 addr <= i915_mmio_reg_offset(OA_PERFCNT4_HI));
+	return gen7_is_valid_mux_addr(perf, addr) ||
+	       ADDR_IN_RANGE(addr, 0x182300, 0x1823A4);
 }
 
-static bool hsw_is_valid_mux_addr(struct drm_i915_private *dev_priv, u32 addr)
+static bool gen12_is_valid_b_counter_addr(struct i915_perf *perf, u32 addr)
 {
-	return gen7_is_valid_mux_addr(dev_priv, addr) ||
-		(addr >= 0x25100 && addr <= 0x2FF90) ||
-		(addr >= i915_mmio_reg_offset(HSW_MBVID2_NOA0) &&
-		 addr <= i915_mmio_reg_offset(HSW_MBVID2_NOA9)) ||
-		addr == i915_mmio_reg_offset(HSW_MBVID2_MISR0);
+	return REG_IN_RANGE(addr, GEN12_OAG_OASTARTTRIG1, GEN12_OAG_OASTARTTRIG8) ||
+	       REG_IN_RANGE(addr, GEN12_OAG_OAREPORTTRIG1, GEN12_OAG_OAREPORTTRIG8) ||
+	       REG_IN_RANGE(addr, GEN12_OAG_CEC0_0, GEN12_OAG_CEC7_1) ||
+	       REG_IN_RANGE(addr, GEN12_OAG_SCEC0_0, GEN12_OAG_SCEC7_1) ||
+	       REG_EQUAL(addr, GEN12_OAA_DBG_REG) ||
+	       REG_EQUAL(addr, GEN12_OAG_OA_PESS) ||
+	       REG_EQUAL(addr, GEN12_OAG_SPCTR_CNF);
 }
 
-static bool chv_is_valid_mux_addr(struct drm_i915_private *dev_priv, u32 addr)
+static bool gen12_is_valid_mux_addr(struct i915_perf *perf, u32 addr)
 {
-	return gen7_is_valid_mux_addr(dev_priv, addr) ||
-		(addr >= 0x182300 && addr <= 0x1823A4);
+	return REG_EQUAL(addr, NOA_WRITE) ||
+	       REG_EQUAL(addr, GEN10_NOA_WRITE_HIGH) ||
+	       REG_EQUAL(addr, GDT_CHICKEN_BITS) ||
+	       REG_EQUAL(addr, WAIT_FOR_RC6_EXIT) ||
+	       REG_EQUAL(addr, RPM_CONFIG0) ||
+	       REG_EQUAL(addr, RPM_CONFIG1) ||
+	       REG_IN_RANGE(addr, NOA_CONFIG(0), NOA_CONFIG(8));
 }
 
 static u32 mask_reg_value(u32 reg, u32 val)
@@ -3201,21 +3840,21 @@ static u32 mask_reg_value(u32 reg, u32 val)
 	 * WaDisableSTUnitPowerOptimization workaround. Make sure the value
 	 * programmed by userspace doesn't change this.
 	 */
-	if (i915_mmio_reg_offset(HALF_SLICE_CHICKEN2) == reg)
+	if (REG_EQUAL(reg, HALF_SLICE_CHICKEN2))
 		val = val & ~_MASKED_BIT_ENABLE(GEN8_ST_PO_DISABLE);
 
 	/* WAIT_FOR_RC6_EXIT has only one bit fullfilling the function
 	 * indicated by its name and a bunch of selection fields used by OA
 	 * configs.
 	 */
-	if (i915_mmio_reg_offset(WAIT_FOR_RC6_EXIT) == reg)
+	if (REG_EQUAL(reg, WAIT_FOR_RC6_EXIT))
 		val = val & ~_MASKED_BIT_ENABLE(HSW_WAIT_FOR_RC6_EXIT_ENABLE);
 
 	return val;
 }
 
-static struct i915_oa_reg *alloc_oa_regs(struct drm_i915_private *dev_priv,
-					 bool (*is_valid)(struct drm_i915_private *dev_priv, u32 addr),
+static struct i915_oa_reg *alloc_oa_regs(struct i915_perf *perf,
+					 bool (*is_valid)(struct i915_perf *perf, u32 addr),
 					 u32 __user *regs,
 					 u32 n_regs)
 {
@@ -3245,7 +3884,7 @@ static struct i915_oa_reg *alloc_oa_regs(struct drm_i915_private *dev_priv,
 		if (err)
 			goto addr_err;
 
-		if (!is_valid(dev_priv, addr)) {
+		if (!is_valid(perf, addr)) {
 			DRM_DEBUG("Invalid oa_reg address: %X\n", addr);
 			err = -EINVAL;
 			goto addr_err;
@@ -3278,7 +3917,7 @@ static ssize_t show_dynamic_id(struct device *dev,
 	return sprintf(buf, "%d\n", oa_config->id);
 }
 
-static int create_dynamic_oa_sysfs_entry(struct drm_i915_private *dev_priv,
+static int create_dynamic_oa_sysfs_entry(struct i915_perf *perf,
 					 struct i915_oa_config *oa_config)
 {
 	sysfs_attr_init(&oa_config->sysfs_metric_id.attr);
@@ -3293,7 +3932,7 @@ static int create_dynamic_oa_sysfs_entry(struct drm_i915_private *dev_priv,
 	oa_config->sysfs_metric.name = oa_config->uuid;
 	oa_config->sysfs_metric.attrs = oa_config->attrs;
 
-	return sysfs_create_group(dev_priv->perf.metrics_kobj,
+	return sysfs_create_group(perf->metrics_kobj,
 				  &oa_config->sysfs_metric);
 }
 
@@ -3313,17 +3952,18 @@ static int create_dynamic_oa_sysfs_entry(struct drm_i915_private *dev_priv,
 int i915_perf_add_config_ioctl(struct drm_device *dev, void *data,
 			       struct drm_file *file)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct i915_perf *perf = &to_i915(dev)->perf;
 	struct drm_i915_perf_oa_config *args = data;
 	struct i915_oa_config *oa_config, *tmp;
+	static struct i915_oa_reg *regs;
 	int err, id;
 
-	if (!dev_priv->perf.initialized) {
+	if (!perf->i915) {
 		DRM_DEBUG("i915 perf interface not available for this system\n");
 		return -ENOTSUPP;
 	}
 
-	if (!dev_priv->perf.metrics_kobj) {
+	if (!perf->metrics_kobj) {
 		DRM_DEBUG("OA metrics weren't advertised via sysfs\n");
 		return -EINVAL;
 	}
@@ -3346,7 +3986,8 @@ int i915_perf_add_config_ioctl(struct drm_device *dev, void *data,
 		return -ENOMEM;
 	}
 
-	atomic_set(&oa_config->ref_count, 1);
+	oa_config->perf = perf;
+	kref_init(&oa_config->ref);
 
 	if (!uuid_is_valid(args->uuid)) {
 		DRM_DEBUG("Invalid uuid format for OA config\n");
@@ -3360,59 +4001,59 @@ int i915_perf_add_config_ioctl(struct drm_device *dev, void *data,
 	memcpy(oa_config->uuid, args->uuid, sizeof(args->uuid));
 
 	oa_config->mux_regs_len = args->n_mux_regs;
-	oa_config->mux_regs =
-		alloc_oa_regs(dev_priv,
-			      dev_priv->perf.ops.is_valid_mux_reg,
-			      u64_to_user_ptr(args->mux_regs_ptr),
-			      args->n_mux_regs);
+	regs = alloc_oa_regs(perf,
+			     perf->ops.is_valid_mux_reg,
+			     u64_to_user_ptr(args->mux_regs_ptr),
+			     args->n_mux_regs);
 
-	if (IS_ERR(oa_config->mux_regs)) {
+	if (IS_ERR(regs)) {
 		DRM_DEBUG("Failed to create OA config for mux_regs\n");
-		err = PTR_ERR(oa_config->mux_regs);
+		err = PTR_ERR(regs);
 		goto reg_err;
 	}
+	oa_config->mux_regs = regs;
 
 	oa_config->b_counter_regs_len = args->n_boolean_regs;
-	oa_config->b_counter_regs =
-		alloc_oa_regs(dev_priv,
-			      dev_priv->perf.ops.is_valid_b_counter_reg,
-			      u64_to_user_ptr(args->boolean_regs_ptr),
-			      args->n_boolean_regs);
+	regs = alloc_oa_regs(perf,
+			     perf->ops.is_valid_b_counter_reg,
+			     u64_to_user_ptr(args->boolean_regs_ptr),
+			     args->n_boolean_regs);
 
-	if (IS_ERR(oa_config->b_counter_regs)) {
+	if (IS_ERR(regs)) {
 		DRM_DEBUG("Failed to create OA config for b_counter_regs\n");
-		err = PTR_ERR(oa_config->b_counter_regs);
+		err = PTR_ERR(regs);
 		goto reg_err;
 	}
+	oa_config->b_counter_regs = regs;
 
-	if (INTEL_GEN(dev_priv) < 8) {
+	if (INTEL_GEN(perf->i915) < 8) {
 		if (args->n_flex_regs != 0) {
 			err = -EINVAL;
 			goto reg_err;
 		}
 	} else {
 		oa_config->flex_regs_len = args->n_flex_regs;
-		oa_config->flex_regs =
-			alloc_oa_regs(dev_priv,
-				      dev_priv->perf.ops.is_valid_flex_reg,
-				      u64_to_user_ptr(args->flex_regs_ptr),
-				      args->n_flex_regs);
+		regs = alloc_oa_regs(perf,
+				     perf->ops.is_valid_flex_reg,
+				     u64_to_user_ptr(args->flex_regs_ptr),
+				     args->n_flex_regs);
 
-		if (IS_ERR(oa_config->flex_regs)) {
+		if (IS_ERR(regs)) {
 			DRM_DEBUG("Failed to create OA config for flex_regs\n");
-			err = PTR_ERR(oa_config->flex_regs);
+			err = PTR_ERR(regs);
 			goto reg_err;
 		}
+		oa_config->flex_regs = regs;
 	}
 
-	err = mutex_lock_interruptible(&dev_priv->perf.metrics_lock);
+	err = mutex_lock_interruptible(&perf->metrics_lock);
 	if (err)
 		goto reg_err;
 
 	/* We shouldn't have too many configs, so this iteration shouldn't be
 	 * too costly.
 	 */
-	idr_for_each_entry(&dev_priv->perf.metrics_idr, tmp, id) {
+	idr_for_each_entry(&perf->metrics_idr, tmp, id) {
 		if (!strcmp(tmp->uuid, oa_config->uuid)) {
 			DRM_DEBUG("OA config already exists with this uuid\n");
 			err = -EADDRINUSE;
@@ -3420,14 +4061,14 @@ int i915_perf_add_config_ioctl(struct drm_device *dev, void *data,
 		}
 	}
 
-	err = create_dynamic_oa_sysfs_entry(dev_priv, oa_config);
+	err = create_dynamic_oa_sysfs_entry(perf, oa_config);
 	if (err) {
 		DRM_DEBUG("Failed to create sysfs entry for OA config\n");
 		goto sysfs_err;
 	}
 
 	/* Config id 0 is invalid, id 1 for kernel stored test config. */
-	oa_config->id = idr_alloc(&dev_priv->perf.metrics_idr,
+	oa_config->id = idr_alloc(&perf->metrics_idr,
 				  oa_config, 2,
 				  0, GFP_KERNEL);
 	if (oa_config->id < 0) {
@@ -3436,16 +4077,16 @@ int i915_perf_add_config_ioctl(struct drm_device *dev, void *data,
 		goto sysfs_err;
 	}
 
-	mutex_unlock(&dev_priv->perf.metrics_lock);
+	mutex_unlock(&perf->metrics_lock);
 
 	DRM_DEBUG("Added config %s id=%i\n", oa_config->uuid, oa_config->id);
 
 	return oa_config->id;
 
 sysfs_err:
-	mutex_unlock(&dev_priv->perf.metrics_lock);
+	mutex_unlock(&perf->metrics_lock);
 reg_err:
-	put_oa_config(dev_priv, oa_config);
+	i915_oa_config_put(oa_config);
 	DRM_DEBUG("Failed to add new OA config\n");
 	return err;
 }
@@ -3464,12 +4105,12 @@ reg_err:
 int i915_perf_remove_config_ioctl(struct drm_device *dev, void *data,
 				  struct drm_file *file)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct i915_perf *perf = &to_i915(dev)->perf;
 	u64 *arg = data;
 	struct i915_oa_config *oa_config;
 	int ret;
 
-	if (!dev_priv->perf.initialized) {
+	if (!perf->i915) {
 		DRM_DEBUG("i915 perf interface not available for this system\n");
 		return -ENOTSUPP;
 	}
@@ -3479,31 +4120,33 @@ int i915_perf_remove_config_ioctl(struct drm_device *dev, void *data,
 		return -EACCES;
 	}
 
-	ret = mutex_lock_interruptible(&dev_priv->perf.metrics_lock);
+	ret = mutex_lock_interruptible(&perf->metrics_lock);
 	if (ret)
-		goto lock_err;
+		return ret;
 
-	oa_config = idr_find(&dev_priv->perf.metrics_idr, *arg);
+	oa_config = idr_find(&perf->metrics_idr, *arg);
 	if (!oa_config) {
 		DRM_DEBUG("Failed to remove unknown OA config\n");
 		ret = -ENOENT;
-		goto config_err;
+		goto err_unlock;
 	}
 
 	GEM_BUG_ON(*arg != oa_config->id);
 
-	sysfs_remove_group(dev_priv->perf.metrics_kobj,
-			   &oa_config->sysfs_metric);
+	sysfs_remove_group(perf->metrics_kobj, &oa_config->sysfs_metric);
+
+	idr_remove(&perf->metrics_idr, *arg);
 
-	idr_remove(&dev_priv->perf.metrics_idr, *arg);
+	mutex_unlock(&perf->metrics_lock);
 
 	DRM_DEBUG("Removed config %s id=%i\n", oa_config->uuid, oa_config->id);
 
-	put_oa_config(dev_priv, oa_config);
+	i915_oa_config_put(oa_config);
 
-config_err:
-	mutex_unlock(&dev_priv->perf.metrics_lock);
-lock_err:
+	return 0;
+
+err_unlock:
+	mutex_unlock(&perf->metrics_lock);
 	return ret;
 }
 
@@ -3551,103 +4194,126 @@ static struct ctl_table dev_root[] = {
 
 /**
  * i915_perf_init - initialize i915-perf state on module load
- * @dev_priv: i915 device instance
+ * @i915: i915 device instance
  *
  * Initializes i915-perf state without exposing anything to userspace.
  *
  * Note: i915-perf initialization is split into an 'init' and 'register'
  * phase with the i915_perf_register() exposing state to userspace.
  */
-void i915_perf_init(struct drm_i915_private *dev_priv)
-{
-	if (IS_HASWELL(dev_priv)) {
-		dev_priv->perf.ops.is_valid_b_counter_reg =
-			gen7_is_valid_b_counter_addr;
-		dev_priv->perf.ops.is_valid_mux_reg =
-			hsw_is_valid_mux_addr;
-		dev_priv->perf.ops.is_valid_flex_reg = NULL;
-		dev_priv->perf.ops.enable_metric_set = hsw_enable_metric_set;
-		dev_priv->perf.ops.disable_metric_set = hsw_disable_metric_set;
-		dev_priv->perf.ops.oa_enable = gen7_oa_enable;
-		dev_priv->perf.ops.oa_disable = gen7_oa_disable;
-		dev_priv->perf.ops.read = gen7_oa_read;
-		dev_priv->perf.ops.oa_hw_tail_read =
-			gen7_oa_hw_tail_read;
-
-		dev_priv->perf.oa_formats = hsw_oa_formats;
-	} else if (HAS_LOGICAL_RING_CONTEXTS(dev_priv)) {
+void i915_perf_init(struct drm_i915_private *i915)
+{
+	struct i915_perf *perf = &i915->perf;
+
+	/* XXX const struct i915_perf_ops! */
+
+	if (IS_HASWELL(i915)) {
+		perf->ops.is_valid_b_counter_reg = gen7_is_valid_b_counter_addr;
+		perf->ops.is_valid_mux_reg = hsw_is_valid_mux_addr;
+		perf->ops.is_valid_flex_reg = NULL;
+		perf->ops.enable_metric_set = hsw_enable_metric_set;
+		perf->ops.disable_metric_set = hsw_disable_metric_set;
+		perf->ops.oa_enable = gen7_oa_enable;
+		perf->ops.oa_disable = gen7_oa_disable;
+		perf->ops.read = gen7_oa_read;
+		perf->ops.oa_hw_tail_read = gen7_oa_hw_tail_read;
+
+		perf->oa_formats = hsw_oa_formats;
+	} else if (HAS_LOGICAL_RING_CONTEXTS(i915)) {
 		/* Note: that although we could theoretically also support the
 		 * legacy ringbuffer mode on BDW (and earlier iterations of
 		 * this driver, before upstreaming did this) it didn't seem
 		 * worth the complexity to maintain now that BDW+ enable
 		 * execlist mode by default.
 		 */
-		dev_priv->perf.oa_formats = gen8_plus_oa_formats;
+		perf->ops.read = gen8_oa_read;
 
-		dev_priv->perf.ops.oa_enable = gen8_oa_enable;
-		dev_priv->perf.ops.oa_disable = gen8_oa_disable;
-		dev_priv->perf.ops.read = gen8_oa_read;
-		dev_priv->perf.ops.oa_hw_tail_read = gen8_oa_hw_tail_read;
+		if (IS_GEN_RANGE(i915, 8, 9)) {
+			perf->oa_formats = gen8_plus_oa_formats;
 
-		if (IS_GEN_RANGE(dev_priv, 8, 9)) {
-			dev_priv->perf.ops.is_valid_b_counter_reg =
+			perf->ops.is_valid_b_counter_reg =
 				gen7_is_valid_b_counter_addr;
-			dev_priv->perf.ops.is_valid_mux_reg =
+			perf->ops.is_valid_mux_reg =
 				gen8_is_valid_mux_addr;
-			dev_priv->perf.ops.is_valid_flex_reg =
+			perf->ops.is_valid_flex_reg =
 				gen8_is_valid_flex_addr;
 
-			if (IS_CHERRYVIEW(dev_priv)) {
-				dev_priv->perf.ops.is_valid_mux_reg =
+			if (IS_CHERRYVIEW(i915)) {
+				perf->ops.is_valid_mux_reg =
 					chv_is_valid_mux_addr;
 			}
 
-			dev_priv->perf.ops.enable_metric_set = gen8_enable_metric_set;
-			dev_priv->perf.ops.disable_metric_set = gen8_disable_metric_set;
+			perf->ops.oa_enable = gen8_oa_enable;
+			perf->ops.oa_disable = gen8_oa_disable;
+			perf->ops.enable_metric_set = gen8_enable_metric_set;
+			perf->ops.disable_metric_set = gen8_disable_metric_set;
+			perf->ops.oa_hw_tail_read = gen8_oa_hw_tail_read;
 
-			if (IS_GEN(dev_priv, 8)) {
-				dev_priv->perf.ctx_oactxctrl_offset = 0x120;
-				dev_priv->perf.ctx_flexeu0_offset = 0x2ce;
+			if (IS_GEN(i915, 8)) {
+				perf->ctx_oactxctrl_offset = 0x120;
+				perf->ctx_flexeu0_offset = 0x2ce;
 
-				dev_priv->perf.gen8_valid_ctx_bit = BIT(25);
+				perf->gen8_valid_ctx_bit = BIT(25);
 			} else {
-				dev_priv->perf.ctx_oactxctrl_offset = 0x128;
-				dev_priv->perf.ctx_flexeu0_offset = 0x3de;
+				perf->ctx_oactxctrl_offset = 0x128;
+				perf->ctx_flexeu0_offset = 0x3de;
 
-				dev_priv->perf.gen8_valid_ctx_bit = BIT(16);
+				perf->gen8_valid_ctx_bit = BIT(16);
 			}
-		} else if (IS_GEN_RANGE(dev_priv, 10, 11)) {
-			dev_priv->perf.ops.is_valid_b_counter_reg =
+		} else if (IS_GEN_RANGE(i915, 10, 11)) {
+			perf->oa_formats = gen8_plus_oa_formats;
+
+			perf->ops.is_valid_b_counter_reg =
 				gen7_is_valid_b_counter_addr;
-			dev_priv->perf.ops.is_valid_mux_reg =
+			perf->ops.is_valid_mux_reg =
 				gen10_is_valid_mux_addr;
-			dev_priv->perf.ops.is_valid_flex_reg =
+			perf->ops.is_valid_flex_reg =
 				gen8_is_valid_flex_addr;
 
-			dev_priv->perf.ops.enable_metric_set = gen8_enable_metric_set;
-			dev_priv->perf.ops.disable_metric_set = gen10_disable_metric_set;
+			perf->ops.oa_enable = gen8_oa_enable;
+			perf->ops.oa_disable = gen8_oa_disable;
+			perf->ops.enable_metric_set = gen8_enable_metric_set;
+			perf->ops.disable_metric_set = gen10_disable_metric_set;
+			perf->ops.oa_hw_tail_read = gen8_oa_hw_tail_read;
 
-			if (IS_GEN(dev_priv, 10)) {
-				dev_priv->perf.ctx_oactxctrl_offset = 0x128;
-				dev_priv->perf.ctx_flexeu0_offset = 0x3de;
+			if (IS_GEN(i915, 10)) {
+				perf->ctx_oactxctrl_offset = 0x128;
+				perf->ctx_flexeu0_offset = 0x3de;
 			} else {
-				dev_priv->perf.ctx_oactxctrl_offset = 0x124;
-				dev_priv->perf.ctx_flexeu0_offset = 0x78e;
+				perf->ctx_oactxctrl_offset = 0x124;
+				perf->ctx_flexeu0_offset = 0x78e;
 			}
-			dev_priv->perf.gen8_valid_ctx_bit = BIT(16);
+			perf->gen8_valid_ctx_bit = BIT(16);
+		} else if (IS_GEN(i915, 12)) {
+			perf->oa_formats = gen12_oa_formats;
+
+			perf->ops.is_valid_b_counter_reg =
+				gen12_is_valid_b_counter_addr;
+			perf->ops.is_valid_mux_reg =
+				gen12_is_valid_mux_addr;
+			perf->ops.is_valid_flex_reg =
+				gen8_is_valid_flex_addr;
+
+			perf->ops.oa_enable = gen12_oa_enable;
+			perf->ops.oa_disable = gen12_oa_disable;
+			perf->ops.enable_metric_set = gen12_enable_metric_set;
+			perf->ops.disable_metric_set = gen12_disable_metric_set;
+			perf->ops.oa_hw_tail_read = gen12_oa_hw_tail_read;
+
+			perf->ctx_flexeu0_offset = 0;
+			perf->ctx_oactxctrl_offset = 0x144;
 		}
 	}
 
-	if (dev_priv->perf.ops.enable_metric_set) {
-		INIT_LIST_HEAD(&dev_priv->perf.streams);
-		mutex_init(&dev_priv->perf.lock);
+	if (perf->ops.enable_metric_set) {
+		mutex_init(&perf->lock);
 
 		oa_sample_rate_hard_limit = 1000 *
-			(RUNTIME_INFO(dev_priv)->cs_timestamp_frequency_khz / 2);
-		dev_priv->perf.sysctl_header = register_sysctl_table(dev_root);
+			(RUNTIME_INFO(i915)->cs_timestamp_frequency_khz / 2);
+		perf->sysctl_header = register_sysctl_table(dev_root);
 
-		mutex_init(&dev_priv->perf.metrics_lock);
-		idr_init(&dev_priv->perf.metrics_idr);
+		mutex_init(&perf->metrics_lock);
+		idr_init(&perf->metrics_idr);
 
 		/* We set up some ratelimit state to potentially throttle any
 		 * _NOTES about spurious, invalid OA reports which we don't
@@ -3659,44 +4325,70 @@ void i915_perf_init(struct drm_i915_private *dev_priv)
 		 *
 		 * Using the same limiting factors as printk_ratelimit()
 		 */
-		ratelimit_state_init(&dev_priv->perf.spurious_report_rs,
-				     5 * HZ, 10);
+		ratelimit_state_init(&perf->spurious_report_rs, 5 * HZ, 10);
 		/* Since we use a DRM_NOTE for spurious reports it would be
 		 * inconsistent to let __ratelimit() automatically print a
 		 * warning for throttling.
 		 */
-		ratelimit_set_flags(&dev_priv->perf.spurious_report_rs,
+		ratelimit_set_flags(&perf->spurious_report_rs,
 				    RATELIMIT_MSG_ON_RELEASE);
 
-		dev_priv->perf.initialized = true;
+		atomic64_set(&perf->noa_programming_delay,
+			     500 * 1000 /* 500us */);
+
+		perf->i915 = i915;
 	}
 }
 
 static int destroy_config(int id, void *p, void *data)
 {
-	struct drm_i915_private *dev_priv = data;
-	struct i915_oa_config *oa_config = p;
-
-	put_oa_config(dev_priv, oa_config);
-
+	i915_oa_config_put(p);
 	return 0;
 }
 
 /**
  * i915_perf_fini - Counter part to i915_perf_init()
- * @dev_priv: i915 device instance
+ * @i915: i915 device instance
  */
-void i915_perf_fini(struct drm_i915_private *dev_priv)
+void i915_perf_fini(struct drm_i915_private *i915)
 {
-	if (!dev_priv->perf.initialized)
+	struct i915_perf *perf = &i915->perf;
+
+	if (!perf->i915)
 		return;
 
-	idr_for_each(&dev_priv->perf.metrics_idr, destroy_config, dev_priv);
-	idr_destroy(&dev_priv->perf.metrics_idr);
+	idr_for_each(&perf->metrics_idr, destroy_config, perf);
+	idr_destroy(&perf->metrics_idr);
 
-	unregister_sysctl_table(dev_priv->perf.sysctl_header);
+	unregister_sysctl_table(perf->sysctl_header);
 
-	memset(&dev_priv->perf.ops, 0, sizeof(dev_priv->perf.ops));
+	memset(&perf->ops, 0, sizeof(perf->ops));
+	perf->i915 = NULL;
+}
 
-	dev_priv->perf.initialized = false;
+/**
+ * i915_perf_ioctl_version - Version of the i915-perf subsystem
+ *
+ * This version number is used by userspace to detect available features.
+ */
+int i915_perf_ioctl_version(void)
+{
+	/*
+	 * 1: Initial version
+	 *   I915_PERF_IOCTL_ENABLE
+	 *   I915_PERF_IOCTL_DISABLE
+	 *
+	 * 2: Added runtime modification of OA config.
+	 *   I915_PERF_IOCTL_CONFIG
+	 *
+	 * 3: Add DRM_I915_PERF_PROP_HOLD_PREEMPTION parameter to hold
+	 *    preemption on a particular context so that performance data is
+	 *    accessible from a delta of MI_RPC reports without looking at the
+	 *    OA buffer.
+	 */
+	return 3;
 }
+
+#if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
+#include "selftests/i915_perf.c"
+#endif
diff --git a/drivers/gpu/drm/i915/i915_perf.h b/drivers/gpu/drm/i915/i915_perf.h
index a412b16d9ffc..4ceebce72060 100644
--- a/drivers/gpu/drm/i915/i915_perf.h
+++ b/drivers/gpu/drm/i915/i915_perf.h
@@ -6,11 +6,15 @@
 #ifndef __I915_PERF_H__
 #define __I915_PERF_H__
 
+#include <linux/kref.h>
 #include <linux/types.h>
 
+#include "i915_perf_types.h"
+
 struct drm_device;
 struct drm_file;
 struct drm_i915_private;
+struct i915_oa_config;
 struct intel_context;
 struct intel_engine_cs;
 
@@ -18,6 +22,7 @@ void i915_perf_init(struct drm_i915_private *i915);
 void i915_perf_fini(struct drm_i915_private *i915);
 void i915_perf_register(struct drm_i915_private *i915);
 void i915_perf_unregister(struct drm_i915_private *i915);
+int i915_perf_ioctl_version(void);
 
 int i915_perf_open_ioctl(struct drm_device *dev, void *data,
 			 struct drm_file *file);
@@ -25,8 +30,29 @@ int i915_perf_add_config_ioctl(struct drm_device *dev, void *data,
 			       struct drm_file *file);
 int i915_perf_remove_config_ioctl(struct drm_device *dev, void *data,
 				  struct drm_file *file);
-void i915_oa_init_reg_state(struct intel_engine_cs *engine,
-			    struct intel_context *ce,
-			    u32 *reg_state);
+
+void i915_oa_init_reg_state(const struct intel_context *ce,
+			    const struct intel_engine_cs *engine);
+
+struct i915_oa_config *
+i915_perf_get_oa_config(struct i915_perf *perf, int metrics_set);
+
+static inline struct i915_oa_config *
+i915_oa_config_get(struct i915_oa_config *oa_config)
+{
+	if (kref_get_unless_zero(&oa_config->ref))
+		return oa_config;
+	else
+		return NULL;
+}
+
+void i915_oa_config_release(struct kref *ref);
+static inline void i915_oa_config_put(struct i915_oa_config *oa_config)
+{
+	if (!oa_config)
+		return;
+
+	kref_put(&oa_config->ref, i915_oa_config_release);
+}
 
 #endif /* __I915_PERF_H__ */
diff --git a/drivers/gpu/drm/i915/i915_perf_types.h b/drivers/gpu/drm/i915/i915_perf_types.h
new file mode 100644
index 000000000000..74ddc20a0d37
--- /dev/null
+++ b/drivers/gpu/drm/i915/i915_perf_types.h
@@ -0,0 +1,435 @@
+/* SPDX-License-Identifier: MIT */
+/*
+ * Copyright © 2019 Intel Corporation
+ */
+
+#ifndef _I915_PERF_TYPES_H_
+#define _I915_PERF_TYPES_H_
+
+#include <linux/atomic.h>
+#include <linux/device.h>
+#include <linux/hrtimer.h>
+#include <linux/llist.h>
+#include <linux/poll.h>
+#include <linux/sysfs.h>
+#include <linux/types.h>
+#include <linux/uuid.h>
+#include <linux/wait.h>
+
+#include "i915_reg.h"
+#include "intel_wakeref.h"
+
+struct drm_i915_private;
+struct file;
+struct i915_gem_context;
+struct i915_perf;
+struct i915_vma;
+struct intel_context;
+struct intel_engine_cs;
+
+struct i915_oa_format {
+	u32 format;
+	int size;
+};
+
+struct i915_oa_reg {
+	i915_reg_t addr;
+	u32 value;
+};
+
+struct i915_oa_config {
+	struct i915_perf *perf;
+
+	char uuid[UUID_STRING_LEN + 1];
+	int id;
+
+	const struct i915_oa_reg *mux_regs;
+	u32 mux_regs_len;
+	const struct i915_oa_reg *b_counter_regs;
+	u32 b_counter_regs_len;
+	const struct i915_oa_reg *flex_regs;
+	u32 flex_regs_len;
+
+	struct attribute_group sysfs_metric;
+	struct attribute *attrs[2];
+	struct device_attribute sysfs_metric_id;
+
+	struct kref ref;
+	struct rcu_head rcu;
+};
+
+struct i915_perf_stream;
+
+/**
+ * struct i915_perf_stream_ops - the OPs to support a specific stream type
+ */
+struct i915_perf_stream_ops {
+	/**
+	 * @enable: Enables the collection of HW samples, either in response to
+	 * `I915_PERF_IOCTL_ENABLE` or implicitly called when stream is opened
+	 * without `I915_PERF_FLAG_DISABLED`.
+	 */
+	void (*enable)(struct i915_perf_stream *stream);
+
+	/**
+	 * @disable: Disables the collection of HW samples, either in response
+	 * to `I915_PERF_IOCTL_DISABLE` or implicitly called before destroying
+	 * the stream.
+	 */
+	void (*disable)(struct i915_perf_stream *stream);
+
+	/**
+	 * @poll_wait: Call poll_wait, passing a wait queue that will be woken
+	 * once there is something ready to read() for the stream
+	 */
+	void (*poll_wait)(struct i915_perf_stream *stream,
+			  struct file *file,
+			  poll_table *wait);
+
+	/**
+	 * @wait_unlocked: For handling a blocking read, wait until there is
+	 * something to ready to read() for the stream. E.g. wait on the same
+	 * wait queue that would be passed to poll_wait().
+	 */
+	int (*wait_unlocked)(struct i915_perf_stream *stream);
+
+	/**
+	 * @read: Copy buffered metrics as records to userspace
+	 * **buf**: the userspace, destination buffer
+	 * **count**: the number of bytes to copy, requested by userspace
+	 * **offset**: zero at the start of the read, updated as the read
+	 * proceeds, it represents how many bytes have been copied so far and
+	 * the buffer offset for copying the next record.
+	 *
+	 * Copy as many buffered i915 perf samples and records for this stream
+	 * to userspace as will fit in the given buffer.
+	 *
+	 * Only write complete records; returning -%ENOSPC if there isn't room
+	 * for a complete record.
+	 *
+	 * Return any error condition that results in a short read such as
+	 * -%ENOSPC or -%EFAULT, even though these may be squashed before
+	 * returning to userspace.
+	 */
+	int (*read)(struct i915_perf_stream *stream,
+		    char __user *buf,
+		    size_t count,
+		    size_t *offset);
+
+	/**
+	 * @destroy: Cleanup any stream specific resources.
+	 *
+	 * The stream will always be disabled before this is called.
+	 */
+	void (*destroy)(struct i915_perf_stream *stream);
+};
+
+/**
+ * struct i915_perf_stream - state for a single open stream FD
+ */
+struct i915_perf_stream {
+	/**
+	 * @perf: i915_perf backpointer
+	 */
+	struct i915_perf *perf;
+
+	/**
+	 * @uncore: mmio access path
+	 */
+	struct intel_uncore *uncore;
+
+	/**
+	 * @engine: Engine associated with this performance stream.
+	 */
+	struct intel_engine_cs *engine;
+
+	/**
+	 * @sample_flags: Flags representing the `DRM_I915_PERF_PROP_SAMPLE_*`
+	 * properties given when opening a stream, representing the contents
+	 * of a single sample as read() by userspace.
+	 */
+	u32 sample_flags;
+
+	/**
+	 * @sample_size: Considering the configured contents of a sample
+	 * combined with the required header size, this is the total size
+	 * of a single sample record.
+	 */
+	int sample_size;
+
+	/**
+	 * @ctx: %NULL if measuring system-wide across all contexts or a
+	 * specific context that is being monitored.
+	 */
+	struct i915_gem_context *ctx;
+
+	/**
+	 * @enabled: Whether the stream is currently enabled, considering
+	 * whether the stream was opened in a disabled state and based
+	 * on `I915_PERF_IOCTL_ENABLE` and `I915_PERF_IOCTL_DISABLE` calls.
+	 */
+	bool enabled;
+
+	/**
+	 * @hold_preemption: Whether preemption is put on hold for command
+	 * submissions done on the @ctx. This is useful for some drivers that
+	 * cannot easily post process the OA buffer context to subtract delta
+	 * of performance counters not associated with @ctx.
+	 */
+	bool hold_preemption;
+
+	/**
+	 * @ops: The callbacks providing the implementation of this specific
+	 * type of configured stream.
+	 */
+	const struct i915_perf_stream_ops *ops;
+
+	/**
+	 * @oa_config: The OA configuration used by the stream.
+	 */
+	struct i915_oa_config *oa_config;
+
+	/**
+	 * @oa_config_bos: A list of struct i915_oa_config_bo allocated lazily
+	 * each time @oa_config changes.
+	 */
+	struct llist_head oa_config_bos;
+
+	/**
+	 * @pinned_ctx: The OA context specific information.
+	 */
+	struct intel_context *pinned_ctx;
+
+	/**
+	 * @specific_ctx_id: The id of the specific context.
+	 */
+	u32 specific_ctx_id;
+
+	/**
+	 * @specific_ctx_id_mask: The mask used to masking specific_ctx_id bits.
+	 */
+	u32 specific_ctx_id_mask;
+
+	/**
+	 * @poll_check_timer: High resolution timer that will periodically
+	 * check for data in the circular OA buffer for notifying userspace
+	 * (e.g. during a read() or poll()).
+	 */
+	struct hrtimer poll_check_timer;
+
+	/**
+	 * @poll_wq: The wait queue that hrtimer callback wakes when it
+	 * sees data ready to read in the circular OA buffer.
+	 */
+	wait_queue_head_t poll_wq;
+
+	/**
+	 * @pollin: Whether there is data available to read.
+	 */
+	bool pollin;
+
+	/**
+	 * @periodic: Whether periodic sampling is currently enabled.
+	 */
+	bool periodic;
+
+	/**
+	 * @period_exponent: The OA unit sampling frequency is derived from this.
+	 */
+	int period_exponent;
+
+	/**
+	 * @oa_buffer: State of the OA buffer.
+	 */
+	struct {
+		struct i915_vma *vma;
+		u8 *vaddr;
+		u32 last_ctx_id;
+		int format;
+		int format_size;
+		int size_exponent;
+
+		/**
+		 * @ptr_lock: Locks reads and writes to all head/tail state
+		 *
+		 * Consider: the head and tail pointer state needs to be read
+		 * consistently from a hrtimer callback (atomic context) and
+		 * read() fop (user context) with tail pointer updates happening
+		 * in atomic context and head updates in user context and the
+		 * (unlikely) possibility of read() errors needing to reset all
+		 * head/tail state.
+		 *
+		 * Note: Contention/performance aren't currently a significant
+		 * concern here considering the relatively low frequency of
+		 * hrtimer callbacks (5ms period) and that reads typically only
+		 * happen in response to a hrtimer event and likely complete
+		 * before the next callback.
+		 *
+		 * Note: This lock is not held *while* reading and copying data
+		 * to userspace so the value of head observed in htrimer
+		 * callbacks won't represent any partial consumption of data.
+		 */
+		spinlock_t ptr_lock;
+
+		/**
+		 * @tails: One 'aging' tail pointer and one 'aged' tail pointer ready to
+		 * used for reading.
+		 *
+		 * Initial values of 0xffffffff are invalid and imply that an
+		 * update is required (and should be ignored by an attempted
+		 * read)
+		 */
+		struct {
+			u32 offset;
+		} tails[2];
+
+		/**
+		 * @aged_tail_idx: Index for the aged tail ready to read() data up to.
+		 */
+		unsigned int aged_tail_idx;
+
+		/**
+		 * @aging_timestamp: A monotonic timestamp for when the current aging tail pointer
+		 * was read; used to determine when it is old enough to trust.
+		 */
+		u64 aging_timestamp;
+
+		/**
+		 * @head: Although we can always read back the head pointer register,
+		 * we prefer to avoid trusting the HW state, just to avoid any
+		 * risk that some hardware condition could * somehow bump the
+		 * head pointer unpredictably and cause us to forward the wrong
+		 * OA buffer data to userspace.
+		 */
+		u32 head;
+	} oa_buffer;
+
+	/**
+	 * @noa_wait: A batch buffer doing a wait on the GPU for the NOA logic to be
+	 * reprogrammed.
+	 */
+	struct i915_vma *noa_wait;
+};
+
+/**
+ * struct i915_oa_ops - Gen specific implementation of an OA unit stream
+ */
+struct i915_oa_ops {
+	/**
+	 * @is_valid_b_counter_reg: Validates register's address for
+	 * programming boolean counters for a particular platform.
+	 */
+	bool (*is_valid_b_counter_reg)(struct i915_perf *perf, u32 addr);
+
+	/**
+	 * @is_valid_mux_reg: Validates register's address for programming mux
+	 * for a particular platform.
+	 */
+	bool (*is_valid_mux_reg)(struct i915_perf *perf, u32 addr);
+
+	/**
+	 * @is_valid_flex_reg: Validates register's address for programming
+	 * flex EU filtering for a particular platform.
+	 */
+	bool (*is_valid_flex_reg)(struct i915_perf *perf, u32 addr);
+
+	/**
+	 * @enable_metric_set: Selects and applies any MUX configuration to set
+	 * up the Boolean and Custom (B/C) counters that are part of the
+	 * counter reports being sampled. May apply system constraints such as
+	 * disabling EU clock gating as required.
+	 */
+	int (*enable_metric_set)(struct i915_perf_stream *stream);
+
+	/**
+	 * @disable_metric_set: Remove system constraints associated with using
+	 * the OA unit.
+	 */
+	void (*disable_metric_set)(struct i915_perf_stream *stream);
+
+	/**
+	 * @oa_enable: Enable periodic sampling
+	 */
+	void (*oa_enable)(struct i915_perf_stream *stream);
+
+	/**
+	 * @oa_disable: Disable periodic sampling
+	 */
+	void (*oa_disable)(struct i915_perf_stream *stream);
+
+	/**
+	 * @read: Copy data from the circular OA buffer into a given userspace
+	 * buffer.
+	 */
+	int (*read)(struct i915_perf_stream *stream,
+		    char __user *buf,
+		    size_t count,
+		    size_t *offset);
+
+	/**
+	 * @oa_hw_tail_read: read the OA tail pointer register
+	 *
+	 * In particular this enables us to share all the fiddly code for
+	 * handling the OA unit tail pointer race that affects multiple
+	 * generations.
+	 */
+	u32 (*oa_hw_tail_read)(struct i915_perf_stream *stream);
+};
+
+struct i915_perf {
+	struct drm_i915_private *i915;
+
+	struct kobject *metrics_kobj;
+	struct ctl_table_header *sysctl_header;
+
+	/*
+	 * Lock associated with adding/modifying/removing OA configs
+	 * in perf->metrics_idr.
+	 */
+	struct mutex metrics_lock;
+
+	/*
+	 * List of dynamic configurations (struct i915_oa_config), you
+	 * need to hold perf->metrics_lock to access it.
+	 */
+	struct idr metrics_idr;
+
+	/*
+	 * Lock associated with anything below within this structure
+	 * except exclusive_stream.
+	 */
+	struct mutex lock;
+
+	/*
+	 * The stream currently using the OA unit. If accessed
+	 * outside a syscall associated to its file
+	 * descriptor.
+	 */
+	struct i915_perf_stream *exclusive_stream;
+
+	/**
+	 * For rate limiting any notifications of spurious
+	 * invalid OA reports
+	 */
+	struct ratelimit_state spurious_report_rs;
+
+	struct i915_oa_config test_config;
+
+	u32 gen7_latched_oastatus1;
+	u32 ctx_oactxctrl_offset;
+	u32 ctx_flexeu0_offset;
+
+	/**
+	 * The RPT_ID/reason field for Gen8+ includes a bit
+	 * to determine if the CTX ID in the report is valid
+	 * but the specific bit differs between Gen 8 and 9
+	 */
+	u32 gen8_valid_ctx_bit;
+
+	struct i915_oa_ops ops;
+	const struct i915_oa_format *oa_formats;
+
+	atomic64_t noa_programming_delay;
+};
+
+#endif /* _I915_PERF_TYPES_H_ */
diff --git a/drivers/gpu/drm/i915/i915_pmu.c b/drivers/gpu/drm/i915/i915_pmu.c
index 8e251e719390..0d40dccd1409 100644
--- a/drivers/gpu/drm/i915/i915_pmu.c
+++ b/drivers/gpu/drm/i915/i915_pmu.c
@@ -11,6 +11,8 @@
 #include "gt/intel_engine_pm.h"
 #include "gt/intel_engine_user.h"
 #include "gt/intel_gt_pm.h"
+#include "gt/intel_rc6.h"
+#include "gt/intel_rps.h"
 
 #include "i915_drv.h"
 #include "i915_pmu.h"
@@ -116,22 +118,124 @@ static bool pmu_needs_timer(struct i915_pmu *pmu, bool gpu_active)
 	return enable;
 }
 
-void i915_pmu_gt_parked(struct drm_i915_private *i915)
+static u64 __get_rc6(struct intel_gt *gt)
 {
-	struct i915_pmu *pmu = &i915->pmu;
+	struct drm_i915_private *i915 = gt->i915;
+	u64 val;
 
-	if (!pmu->base.event_init)
-		return;
+	val = intel_rc6_residency_ns(&gt->rc6,
+				     IS_VALLEYVIEW(i915) ?
+				     VLV_GT_RENDER_RC6 :
+				     GEN6_GT_GFX_RC6);
+
+	if (HAS_RC6p(i915))
+		val += intel_rc6_residency_ns(&gt->rc6, GEN6_GT_GFX_RC6p);
+
+	if (HAS_RC6pp(i915))
+		val += intel_rc6_residency_ns(&gt->rc6, GEN6_GT_GFX_RC6pp);
+
+	return val;
+}
+
+#if IS_ENABLED(CONFIG_PM)
+
+static inline s64 ktime_since(const ktime_t kt)
+{
+	return ktime_to_ns(ktime_sub(ktime_get(), kt));
+}
+
+static u64 __pmu_estimate_rc6(struct i915_pmu *pmu)
+{
+	u64 val;
 
-	spin_lock_irq(&pmu->lock);
 	/*
-	 * Signal sampling timer to stop if only engine events are enabled and
-	 * GPU went idle.
+	 * We think we are runtime suspended.
+	 *
+	 * Report the delta from when the device was suspended to now,
+	 * on top of the last known real value, as the approximated RC6
+	 * counter value.
 	 */
-	pmu->timer_enabled = pmu_needs_timer(pmu, false);
-	spin_unlock_irq(&pmu->lock);
+	val = ktime_since(pmu->sleep_last);
+	val += pmu->sample[__I915_SAMPLE_RC6].cur;
+
+	pmu->sample[__I915_SAMPLE_RC6_ESTIMATED].cur = val;
+
+	return val;
 }
 
+static u64 __pmu_update_rc6(struct i915_pmu *pmu, u64 val)
+{
+	/*
+	 * If we are coming back from being runtime suspended we must
+	 * be careful not to report a larger value than returned
+	 * previously.
+	 */
+	if (val >= pmu->sample[__I915_SAMPLE_RC6_ESTIMATED].cur) {
+		pmu->sample[__I915_SAMPLE_RC6_ESTIMATED].cur = 0;
+		pmu->sample[__I915_SAMPLE_RC6].cur = val;
+	} else {
+		val = pmu->sample[__I915_SAMPLE_RC6_ESTIMATED].cur;
+	}
+
+	return val;
+}
+
+static u64 get_rc6(struct intel_gt *gt)
+{
+	struct drm_i915_private *i915 = gt->i915;
+	struct i915_pmu *pmu = &i915->pmu;
+	unsigned long flags;
+	u64 val;
+
+	val = 0;
+	if (intel_gt_pm_get_if_awake(gt)) {
+		val = __get_rc6(gt);
+		intel_gt_pm_put(gt);
+	}
+
+	spin_lock_irqsave(&pmu->lock, flags);
+
+	if (val)
+		val = __pmu_update_rc6(pmu, val);
+	else
+		val = __pmu_estimate_rc6(pmu);
+
+	spin_unlock_irqrestore(&pmu->lock, flags);
+
+	return val;
+}
+
+static void park_rc6(struct drm_i915_private *i915)
+{
+	struct i915_pmu *pmu = &i915->pmu;
+
+	if (pmu->enable & config_enabled_mask(I915_PMU_RC6_RESIDENCY))
+		__pmu_update_rc6(pmu, __get_rc6(&i915->gt));
+
+	pmu->sleep_last = ktime_get();
+}
+
+static void unpark_rc6(struct drm_i915_private *i915)
+{
+	struct i915_pmu *pmu = &i915->pmu;
+
+	/* Estimate how long we slept and accumulate that into rc6 counters */
+	if (pmu->enable & config_enabled_mask(I915_PMU_RC6_RESIDENCY))
+		__pmu_estimate_rc6(pmu);
+}
+
+#else
+
+static u64 get_rc6(struct intel_gt *gt)
+{
+	return __get_rc6(gt);
+}
+
+static void park_rc6(struct drm_i915_private *i915) {}
+static void unpark_rc6(struct drm_i915_private *i915) {}
+
+#endif
+
 static void __i915_pmu_maybe_start_timer(struct i915_pmu *pmu)
 {
 	if (!pmu->timer_enabled && pmu_needs_timer(pmu, true)) {
@@ -143,6 +247,26 @@ static void __i915_pmu_maybe_start_timer(struct i915_pmu *pmu)
 	}
 }
 
+void i915_pmu_gt_parked(struct drm_i915_private *i915)
+{
+	struct i915_pmu *pmu = &i915->pmu;
+
+	if (!pmu->base.event_init)
+		return;
+
+	spin_lock_irq(&pmu->lock);
+
+	park_rc6(i915);
+
+	/*
+	 * Signal sampling timer to stop if only engine events are enabled and
+	 * GPU went idle.
+	 */
+	pmu->timer_enabled = pmu_needs_timer(pmu, false);
+
+	spin_unlock_irq(&pmu->lock);
+}
+
 void i915_pmu_gt_unparked(struct drm_i915_private *i915)
 {
 	struct i915_pmu *pmu = &i915->pmu;
@@ -151,10 +275,14 @@ void i915_pmu_gt_unparked(struct drm_i915_private *i915)
 		return;
 
 	spin_lock_irq(&pmu->lock);
+
 	/*
 	 * Re-enable sampling timer when GPU goes active.
 	 */
 	__i915_pmu_maybe_start_timer(pmu);
+
+	unpark_rc6(i915);
+
 	spin_unlock_irq(&pmu->lock);
 }
 
@@ -174,7 +302,7 @@ engines_sample(struct intel_gt *gt, unsigned int period_ns)
 	if ((i915->pmu.enable & ENGINE_SAMPLE_MASK) == 0)
 		return;
 
-	for_each_engine(engine, i915, id) {
+	for_each_engine(engine, gt, id) {
 		struct intel_engine_pmu *pmu = &engine->pmu;
 		unsigned long flags;
 		bool busy;
@@ -194,6 +322,10 @@ engines_sample(struct intel_gt *gt, unsigned int period_ns)
 		if (val & RING_WAIT_SEMAPHORE)
 			add_sample(&pmu->sample[I915_SAMPLE_SEMA], period_ns);
 
+		/* No need to sample when busy stats are supported. */
+		if (intel_engine_supports_stats(engine))
+			goto skip;
+
 		/*
 		 * While waiting on a semaphore or event, MI_MODE reports the
 		 * ring as idle. However, previously using the seqno, and with
@@ -227,25 +359,26 @@ frequency_sample(struct intel_gt *gt, unsigned int period_ns)
 	struct drm_i915_private *i915 = gt->i915;
 	struct intel_uncore *uncore = gt->uncore;
 	struct i915_pmu *pmu = &i915->pmu;
+	struct intel_rps *rps = &gt->rps;
 
 	if (pmu->enable & config_enabled_mask(I915_PMU_ACTUAL_FREQUENCY)) {
 		u32 val;
 
-		val = i915->gt_pm.rps.cur_freq;
+		val = rps->cur_freq;
 		if (intel_gt_pm_get_if_awake(gt)) {
 			val = intel_uncore_read_notrace(uncore, GEN6_RPSTAT1);
-			val = intel_get_cagf(i915, val);
+			val = intel_get_cagf(rps, val);
 			intel_gt_pm_put(gt);
 		}
 
 		add_sample_mult(&pmu->sample[__I915_SAMPLE_FREQ_ACT],
-				intel_gpu_freq(i915, val),
+				intel_gpu_freq(rps, val),
 				period_ns / 1000);
 	}
 
 	if (pmu->enable & config_enabled_mask(I915_PMU_REQUESTED_FREQUENCY)) {
 		add_sample_mult(&pmu->sample[__I915_SAMPLE_FREQ_REQ],
-				intel_gpu_freq(i915, i915->gt_pm.rps.cur_freq),
+				intel_gpu_freq(rps, rps->cur_freq),
 				period_ns / 1000);
 	}
 }
@@ -426,104 +559,6 @@ static int i915_pmu_event_init(struct perf_event *event)
 	return 0;
 }
 
-static u64 __get_rc6(struct intel_gt *gt)
-{
-	struct drm_i915_private *i915 = gt->i915;
-	u64 val;
-
-	val = intel_rc6_residency_ns(i915,
-				     IS_VALLEYVIEW(i915) ?
-				     VLV_GT_RENDER_RC6 :
-				     GEN6_GT_GFX_RC6);
-
-	if (HAS_RC6p(i915))
-		val += intel_rc6_residency_ns(i915, GEN6_GT_GFX_RC6p);
-
-	if (HAS_RC6pp(i915))
-		val += intel_rc6_residency_ns(i915, GEN6_GT_GFX_RC6pp);
-
-	return val;
-}
-
-static u64 get_rc6(struct intel_gt *gt)
-{
-#if IS_ENABLED(CONFIG_PM)
-	struct drm_i915_private *i915 = gt->i915;
-	struct intel_runtime_pm *rpm = &i915->runtime_pm;
-	struct i915_pmu *pmu = &i915->pmu;
-	intel_wakeref_t wakeref;
-	unsigned long flags;
-	u64 val;
-
-	wakeref = intel_runtime_pm_get_if_in_use(rpm);
-	if (wakeref) {
-		val = __get_rc6(gt);
-		intel_runtime_pm_put(rpm, wakeref);
-
-		/*
-		 * If we are coming back from being runtime suspended we must
-		 * be careful not to report a larger value than returned
-		 * previously.
-		 */
-
-		spin_lock_irqsave(&pmu->lock, flags);
-
-		if (val >= pmu->sample[__I915_SAMPLE_RC6_ESTIMATED].cur) {
-			pmu->sample[__I915_SAMPLE_RC6_ESTIMATED].cur = 0;
-			pmu->sample[__I915_SAMPLE_RC6].cur = val;
-		} else {
-			val = pmu->sample[__I915_SAMPLE_RC6_ESTIMATED].cur;
-		}
-
-		spin_unlock_irqrestore(&pmu->lock, flags);
-	} else {
-		struct device *kdev = rpm->kdev;
-
-		/*
-		 * We are runtime suspended.
-		 *
-		 * Report the delta from when the device was suspended to now,
-		 * on top of the last known real value, as the approximated RC6
-		 * counter value.
-		 */
-		spin_lock_irqsave(&pmu->lock, flags);
-
-		/*
-		 * After the above branch intel_runtime_pm_get_if_in_use failed
-		 * to get the runtime PM reference we cannot assume we are in
-		 * runtime suspend since we can either: a) race with coming out
-		 * of it before we took the power.lock, or b) there are other
-		 * states than suspended which can bring us here.
-		 *
-		 * We need to double-check that we are indeed currently runtime
-		 * suspended and if not we cannot do better than report the last
-		 * known RC6 value.
-		 */
-		if (pm_runtime_status_suspended(kdev)) {
-			val = pm_runtime_suspended_time(kdev);
-
-			if (!pmu->sample[__I915_SAMPLE_RC6_ESTIMATED].cur)
-				pmu->suspended_time_last = val;
-
-			val -= pmu->suspended_time_last;
-			val += pmu->sample[__I915_SAMPLE_RC6].cur;
-
-			pmu->sample[__I915_SAMPLE_RC6_ESTIMATED].cur = val;
-		} else if (pmu->sample[__I915_SAMPLE_RC6_ESTIMATED].cur) {
-			val = pmu->sample[__I915_SAMPLE_RC6_ESTIMATED].cur;
-		} else {
-			val = pmu->sample[__I915_SAMPLE_RC6].cur;
-		}
-
-		spin_unlock_irqrestore(&pmu->lock, flags);
-	}
-
-	return val;
-#else
-	return __get_rc6(gt);
-#endif
-}
-
 static u64 __i915_pmu_event_read(struct perf_event *event)
 {
 	struct drm_i915_private *i915 =
@@ -843,8 +878,8 @@ create_event_attributes(struct i915_pmu *pmu)
 		const char *name;
 		const char *unit;
 	} events[] = {
-		__event(I915_PMU_ACTUAL_FREQUENCY, "actual-frequency", "MHz"),
-		__event(I915_PMU_REQUESTED_FREQUENCY, "requested-frequency", "MHz"),
+		__event(I915_PMU_ACTUAL_FREQUENCY, "actual-frequency", "M"),
+		__event(I915_PMU_REQUESTED_FREQUENCY, "requested-frequency", "M"),
 		__event(I915_PMU_INTERRUPTS, "interrupts", NULL),
 		__event(I915_PMU_RC6_RESIDENCY, "rc6-residency", "ns"),
 	};
@@ -1047,21 +1082,43 @@ static void i915_pmu_unregister_cpuhp_state(struct i915_pmu *pmu)
 	cpuhp_remove_multi_state(cpuhp_slot);
 }
 
+static bool is_igp(struct drm_i915_private *i915)
+{
+	struct pci_dev *pdev = i915->drm.pdev;
+
+	/* IGP is 0000:00:02.0 */
+	return pci_domain_nr(pdev->bus) == 0 &&
+	       pdev->bus->number == 0 &&
+	       PCI_SLOT(pdev->devfn) == 2 &&
+	       PCI_FUNC(pdev->devfn) == 0;
+}
+
 void i915_pmu_register(struct drm_i915_private *i915)
 {
 	struct i915_pmu *pmu = &i915->pmu;
-	int ret;
+	int ret = -ENOMEM;
 
 	if (INTEL_GEN(i915) <= 2) {
 		dev_info(i915->drm.dev, "PMU not supported for this GPU.");
 		return;
 	}
 
-	i915_pmu_events_attr_group.attrs = create_event_attributes(pmu);
-	if (!i915_pmu_events_attr_group.attrs) {
-		ret = -ENOMEM;
+	spin_lock_init(&pmu->lock);
+	hrtimer_init(&pmu->timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL);
+	pmu->timer.function = i915_sample;
+
+	if (!is_igp(i915))
+		pmu->name = kasprintf(GFP_KERNEL,
+				      "i915-%s",
+				      dev_name(i915->drm.dev));
+	else
+		pmu->name = "i915";
+	if (!pmu->name)
 		goto err;
-	}
+
+	i915_pmu_events_attr_group.attrs = create_event_attributes(pmu);
+	if (!i915_pmu_events_attr_group.attrs)
+		goto err_name;
 
 	pmu->base.attr_groups	= i915_pmu_attr_groups;
 	pmu->base.task_ctx_nr	= perf_invalid_context;
@@ -1073,13 +1130,9 @@ void i915_pmu_register(struct drm_i915_private *i915)
 	pmu->base.read		= i915_pmu_event_read;
 	pmu->base.event_idx	= i915_pmu_event_event_idx;
 
-	spin_lock_init(&pmu->lock);
-	hrtimer_init(&pmu->timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL);
-	pmu->timer.function = i915_sample;
-
-	ret = perf_pmu_register(&pmu->base, "i915", -1);
+	ret = perf_pmu_register(&pmu->base, pmu->name, -1);
 	if (ret)
-		goto err;
+		goto err_attr;
 
 	ret = i915_pmu_register_cpuhp_state(pmu);
 	if (ret)
@@ -1089,10 +1142,14 @@ void i915_pmu_register(struct drm_i915_private *i915)
 
 err_unreg:
 	perf_pmu_unregister(&pmu->base);
-err:
+err_attr:
 	pmu->base.event_init = NULL;
 	free_event_attributes(pmu);
-	DRM_NOTE("Failed to register PMU! (err=%d)\n", ret);
+err_name:
+	if (!is_igp(i915))
+		kfree(pmu->name);
+err:
+	dev_notice(i915->drm.dev, "Failed to register PMU!\n");
 }
 
 void i915_pmu_unregister(struct drm_i915_private *i915)
@@ -1110,5 +1167,7 @@ void i915_pmu_unregister(struct drm_i915_private *i915)
 
 	perf_pmu_unregister(&pmu->base);
 	pmu->base.event_init = NULL;
+	if (!is_igp(i915))
+		kfree(pmu->name);
 	free_event_attributes(pmu);
 }
diff --git a/drivers/gpu/drm/i915/i915_pmu.h b/drivers/gpu/drm/i915/i915_pmu.h
index 4fc4f2478301..bf52e3983631 100644
--- a/drivers/gpu/drm/i915/i915_pmu.h
+++ b/drivers/gpu/drm/i915/i915_pmu.h
@@ -47,6 +47,10 @@ struct i915_pmu {
 	 */
 	struct pmu base;
 	/**
+	 * @name: Name as registered with perf core.
+	 */
+	const char *name;
+	/**
 	 * @lock: Lock protecting enable mask and ref count handling.
 	 */
 	spinlock_t lock;
@@ -97,9 +101,9 @@ struct i915_pmu {
 	 */
 	struct i915_pmu_sample sample[__I915_NUM_PMU_SAMPLERS];
 	/**
-	 * @suspended_time_last: Cached suspend time from PM core.
+	 * @sleep_last: Last time GT parked for RC6 estimation.
 	 */
-	u64 suspended_time_last;
+	ktime_t sleep_last;
 	/**
 	 * @i915_attr: Memory block holding device attributes.
 	 */
diff --git a/drivers/gpu/drm/i915/i915_priolist_types.h b/drivers/gpu/drm/i915/i915_priolist_types.h
index 21037a2e2038..732aad148881 100644
--- a/drivers/gpu/drm/i915/i915_priolist_types.h
+++ b/drivers/gpu/drm/i915/i915_priolist_types.h
@@ -16,6 +16,12 @@ enum {
 	I915_PRIORITY_MIN = I915_CONTEXT_MIN_USER_PRIORITY - 1,
 	I915_PRIORITY_NORMAL = I915_CONTEXT_DEFAULT_PRIORITY,
 	I915_PRIORITY_MAX = I915_CONTEXT_MAX_USER_PRIORITY + 1,
+
+	/* A preemptive pulse used to monitor the health of each engine */
+	I915_PRIORITY_HEARTBEAT,
+
+	/* Interactive workload, scheduled for immediate pageflipping */
+	I915_PRIORITY_DISPLAY,
 };
 
 #define I915_USER_PRIORITY_SHIFT 2
@@ -39,6 +45,7 @@ enum {
  * active request.
  */
 #define I915_PRIORITY_UNPREEMPTABLE INT_MAX
+#define I915_PRIORITY_BARRIER INT_MAX
 
 #define __NO_PREEMPTION (I915_PRIORITY_WAIT)
 
diff --git a/drivers/gpu/drm/i915/i915_query.c b/drivers/gpu/drm/i915/i915_query.c
index ad9240a0817a..c27cfef9281c 100644
--- a/drivers/gpu/drm/i915/i915_query.c
+++ b/drivers/gpu/drm/i915/i915_query.c
@@ -7,6 +7,7 @@
 #include <linux/nospec.h>
 
 #include "i915_drv.h"
+#include "i915_perf.h"
 #include "i915_query.h"
 #include <uapi/drm/i915_drm.h>
 
@@ -37,8 +38,6 @@ static int query_topology_info(struct drm_i915_private *dev_priv,
 	const struct sseu_dev_info *sseu = &RUNTIME_INFO(dev_priv)->sseu;
 	struct drm_i915_query_topology_info topo;
 	u32 slice_length, subslice_length, eu_length, total_length;
-	u8 subslice_stride = GEN_SSEU_STRIDE(sseu->max_subslices);
-	u8 eu_stride = GEN_SSEU_STRIDE(sseu->max_eus_per_subslice);
 	int ret;
 
 	if (query_item->flags != 0)
@@ -50,8 +49,8 @@ static int query_topology_info(struct drm_i915_private *dev_priv,
 	BUILD_BUG_ON(sizeof(u8) != sizeof(sseu->slice_mask));
 
 	slice_length = sizeof(sseu->slice_mask);
-	subslice_length = sseu->max_slices * subslice_stride;
-	eu_length = sseu->max_slices * sseu->max_subslices * eu_stride;
+	subslice_length = sseu->max_slices * sseu->ss_stride;
+	eu_length = sseu->max_slices * sseu->max_subslices * sseu->eu_stride;
 	total_length = sizeof(topo) + slice_length + subslice_length +
 		       eu_length;
 
@@ -69,9 +68,9 @@ static int query_topology_info(struct drm_i915_private *dev_priv,
 	topo.max_eus_per_subslice = sseu->max_eus_per_subslice;
 
 	topo.subslice_offset = slice_length;
-	topo.subslice_stride = subslice_stride;
+	topo.subslice_stride = sseu->ss_stride;
 	topo.eu_offset = slice_length + subslice_length;
-	topo.eu_stride = eu_stride;
+	topo.eu_stride = sseu->eu_stride;
 
 	if (__copy_to_user(u64_to_user_ptr(query_item->data_ptr),
 			   &topo, sizeof(topo)))
@@ -142,10 +141,305 @@ query_engine_info(struct drm_i915_private *i915,
 	return len;
 }
 
+static int can_copy_perf_config_registers_or_number(u32 user_n_regs,
+						    u64 user_regs_ptr,
+						    u32 kernel_n_regs)
+{
+	/*
+	 * We'll just put the number of registers, and won't copy the
+	 * register.
+	 */
+	if (user_n_regs == 0)
+		return 0;
+
+	if (user_n_regs < kernel_n_regs)
+		return -EINVAL;
+
+	if (!access_ok(u64_to_user_ptr(user_regs_ptr),
+		       2 * sizeof(u32) * kernel_n_regs))
+		return -EFAULT;
+
+	return 0;
+}
+
+static int copy_perf_config_registers_or_number(const struct i915_oa_reg *kernel_regs,
+						u32 kernel_n_regs,
+						u64 user_regs_ptr,
+						u32 *user_n_regs)
+{
+	u32 r;
+
+	if (*user_n_regs == 0) {
+		*user_n_regs = kernel_n_regs;
+		return 0;
+	}
+
+	*user_n_regs = kernel_n_regs;
+
+	for (r = 0; r < kernel_n_regs; r++) {
+		u32 __user *user_reg_ptr =
+			u64_to_user_ptr(user_regs_ptr + sizeof(u32) * r * 2);
+		u32 __user *user_val_ptr =
+			u64_to_user_ptr(user_regs_ptr + sizeof(u32) * r * 2 +
+					sizeof(u32));
+		int ret;
+
+		ret = __put_user(i915_mmio_reg_offset(kernel_regs[r].addr),
+				 user_reg_ptr);
+		if (ret)
+			return -EFAULT;
+
+		ret = __put_user(kernel_regs[r].value, user_val_ptr);
+		if (ret)
+			return -EFAULT;
+	}
+
+	return 0;
+}
+
+static int query_perf_config_data(struct drm_i915_private *i915,
+				  struct drm_i915_query_item *query_item,
+				  bool use_uuid)
+{
+	struct drm_i915_query_perf_config __user *user_query_config_ptr =
+		u64_to_user_ptr(query_item->data_ptr);
+	struct drm_i915_perf_oa_config __user *user_config_ptr =
+		u64_to_user_ptr(query_item->data_ptr +
+				sizeof(struct drm_i915_query_perf_config));
+	struct drm_i915_perf_oa_config user_config;
+	struct i915_perf *perf = &i915->perf;
+	struct i915_oa_config *oa_config;
+	char uuid[UUID_STRING_LEN + 1];
+	u64 config_id;
+	u32 flags, total_size;
+	int ret;
+
+	if (!perf->i915)
+		return -ENODEV;
+
+	total_size =
+		sizeof(struct drm_i915_query_perf_config) +
+		sizeof(struct drm_i915_perf_oa_config);
+
+	if (query_item->length == 0)
+		return total_size;
+
+	if (query_item->length < total_size) {
+		DRM_DEBUG("Invalid query config data item size=%u expected=%u\n",
+			  query_item->length, total_size);
+		return -EINVAL;
+	}
+
+	if (!access_ok(user_query_config_ptr, total_size))
+		return -EFAULT;
+
+	if (__get_user(flags, &user_query_config_ptr->flags))
+		return -EFAULT;
+
+	if (flags != 0)
+		return -EINVAL;
+
+	if (use_uuid) {
+		struct i915_oa_config *tmp;
+		int id;
+
+		BUILD_BUG_ON(sizeof(user_query_config_ptr->uuid) >= sizeof(uuid));
+
+		memset(&uuid, 0, sizeof(uuid));
+		if (__copy_from_user(uuid, user_query_config_ptr->uuid,
+				     sizeof(user_query_config_ptr->uuid)))
+			return -EFAULT;
+
+		oa_config = NULL;
+		rcu_read_lock();
+		idr_for_each_entry(&perf->metrics_idr, tmp, id) {
+			if (!strcmp(tmp->uuid, uuid)) {
+				oa_config = i915_oa_config_get(tmp);
+				break;
+			}
+		}
+		rcu_read_unlock();
+	} else {
+		if (__get_user(config_id, &user_query_config_ptr->config))
+			return -EFAULT;
+
+		oa_config = i915_perf_get_oa_config(perf, config_id);
+	}
+	if (!oa_config)
+		return -ENOENT;
+
+	if (__copy_from_user(&user_config, user_config_ptr,
+			     sizeof(user_config))) {
+		ret = -EFAULT;
+		goto out;
+	}
+
+	ret = can_copy_perf_config_registers_or_number(user_config.n_boolean_regs,
+						       user_config.boolean_regs_ptr,
+						       oa_config->b_counter_regs_len);
+	if (ret)
+		goto out;
+
+	ret = can_copy_perf_config_registers_or_number(user_config.n_flex_regs,
+						       user_config.flex_regs_ptr,
+						       oa_config->flex_regs_len);
+	if (ret)
+		goto out;
+
+	ret = can_copy_perf_config_registers_or_number(user_config.n_mux_regs,
+						       user_config.mux_regs_ptr,
+						       oa_config->mux_regs_len);
+	if (ret)
+		goto out;
+
+	ret = copy_perf_config_registers_or_number(oa_config->b_counter_regs,
+						   oa_config->b_counter_regs_len,
+						   user_config.boolean_regs_ptr,
+						   &user_config.n_boolean_regs);
+	if (ret)
+		goto out;
+
+	ret = copy_perf_config_registers_or_number(oa_config->flex_regs,
+						   oa_config->flex_regs_len,
+						   user_config.flex_regs_ptr,
+						   &user_config.n_flex_regs);
+	if (ret)
+		goto out;
+
+	ret = copy_perf_config_registers_or_number(oa_config->mux_regs,
+						   oa_config->mux_regs_len,
+						   user_config.mux_regs_ptr,
+						   &user_config.n_mux_regs);
+	if (ret)
+		goto out;
+
+	memcpy(user_config.uuid, oa_config->uuid, sizeof(user_config.uuid));
+
+	if (__copy_to_user(user_config_ptr, &user_config,
+			   sizeof(user_config))) {
+		ret = -EFAULT;
+		goto out;
+	}
+
+	ret = total_size;
+
+out:
+	i915_oa_config_put(oa_config);
+	return ret;
+}
+
+static size_t sizeof_perf_config_list(size_t count)
+{
+	return sizeof(struct drm_i915_query_perf_config) + sizeof(u64) * count;
+}
+
+static size_t sizeof_perf_metrics(struct i915_perf *perf)
+{
+	struct i915_oa_config *tmp;
+	size_t i;
+	int id;
+
+	i = 1;
+	rcu_read_lock();
+	idr_for_each_entry(&perf->metrics_idr, tmp, id)
+		i++;
+	rcu_read_unlock();
+
+	return sizeof_perf_config_list(i);
+}
+
+static int query_perf_config_list(struct drm_i915_private *i915,
+				  struct drm_i915_query_item *query_item)
+{
+	struct drm_i915_query_perf_config __user *user_query_config_ptr =
+		u64_to_user_ptr(query_item->data_ptr);
+	struct i915_perf *perf = &i915->perf;
+	u64 *oa_config_ids = NULL;
+	int alloc, n_configs;
+	u32 flags;
+	int ret;
+
+	if (!perf->i915)
+		return -ENODEV;
+
+	if (query_item->length == 0)
+		return sizeof_perf_metrics(perf);
+
+	if (get_user(flags, &user_query_config_ptr->flags))
+		return -EFAULT;
+
+	if (flags != 0)
+		return -EINVAL;
+
+	n_configs = 1;
+	do {
+		struct i915_oa_config *tmp;
+		u64 *ids;
+		int id;
+
+		ids = krealloc(oa_config_ids,
+			       n_configs * sizeof(*oa_config_ids),
+			       GFP_KERNEL);
+		if (!ids)
+			return -ENOMEM;
+
+		alloc = fetch_and_zero(&n_configs);
+
+		ids[n_configs++] = 1ull; /* reserved for test_config */
+		rcu_read_lock();
+		idr_for_each_entry(&perf->metrics_idr, tmp, id) {
+			if (n_configs < alloc)
+				ids[n_configs] = id;
+			n_configs++;
+		}
+		rcu_read_unlock();
+
+		oa_config_ids = ids;
+	} while (n_configs > alloc);
+
+	if (query_item->length < sizeof_perf_config_list(n_configs)) {
+		DRM_DEBUG("Invalid query config list item size=%u expected=%zu\n",
+			  query_item->length,
+			  sizeof_perf_config_list(n_configs));
+		kfree(oa_config_ids);
+		return -EINVAL;
+	}
+
+	if (put_user(n_configs, &user_query_config_ptr->config)) {
+		kfree(oa_config_ids);
+		return -EFAULT;
+	}
+
+	ret = copy_to_user(user_query_config_ptr + 1,
+			   oa_config_ids,
+			   n_configs * sizeof(*oa_config_ids));
+	kfree(oa_config_ids);
+	if (ret)
+		return -EFAULT;
+
+	return sizeof_perf_config_list(n_configs);
+}
+
+static int query_perf_config(struct drm_i915_private *i915,
+			     struct drm_i915_query_item *query_item)
+{
+	switch (query_item->flags) {
+	case DRM_I915_QUERY_PERF_CONFIG_LIST:
+		return query_perf_config_list(i915, query_item);
+	case DRM_I915_QUERY_PERF_CONFIG_DATA_FOR_UUID:
+		return query_perf_config_data(i915, query_item, true);
+	case DRM_I915_QUERY_PERF_CONFIG_DATA_FOR_ID:
+		return query_perf_config_data(i915, query_item, false);
+	default:
+		return -EINVAL;
+	}
+}
+
 static int (* const i915_query_funcs[])(struct drm_i915_private *dev_priv,
 					struct drm_i915_query_item *query_item) = {
 	query_topology_info,
 	query_engine_info,
+	query_perf_config,
 };
 
 int i915_query_ioctl(struct drm_device *dev, void *data, struct drm_file *file)
diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index 2abd199093c5..73079b503724 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -413,6 +413,9 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
 #define GEN11_VECS_SFC_USAGE(engine)		_MMIO((engine)->mmio_base + 0x2014)
 #define   GEN11_VECS_SFC_USAGE_BIT		(1 << 0)
 
+#define GEN12_SFC_DONE(n)		_MMIO(0x1cc00 + (n) * 0x100)
+#define GEN12_SFC_DONE_MAX		4
+
 #define RING_PP_DIR_BASE(base)		_MMIO((base) + 0x228)
 #define RING_PP_DIR_BASE_READ(base)	_MMIO((base) + 0x518)
 #define RING_PP_DIR_DCLV(base)		_MMIO((base) + 0x220)
@@ -471,6 +474,8 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
 #define   ECOCHK_PPGTT_WT_HSW		(0x2 << 3)
 #define   ECOCHK_PPGTT_WB_HSW		(0x3 << 3)
 
+#define GEN8_RC6_CTX_INFO		_MMIO(0x8504)
+
 #define GAC_ECO_BITS			_MMIO(0x14090)
 #define   ECOBITS_SNB_BIT		(1 << 13)
 #define   ECOBITS_PPGTT_CACHE64B	(3 << 8)
@@ -545,7 +550,9 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
 #define MI_PREDICATE_SRC0_UDW	_MMIO(0x2400 + 4)
 #define MI_PREDICATE_SRC1	_MMIO(0x2408)
 #define MI_PREDICATE_SRC1_UDW	_MMIO(0x2408 + 4)
-
+#define MI_PREDICATE_DATA       _MMIO(0x2410)
+#define MI_PREDICATE_RESULT     _MMIO(0x2418)
+#define MI_PREDICATE_RESULT_1   _MMIO(0x241c)
 #define MI_PREDICATE_RESULT_2	_MMIO(0x2214)
 #define  LOWER_SLICE_ENABLED	(1 << 0)
 #define  LOWER_SLICE_DISABLED	(0 << 0)
@@ -555,6 +562,10 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
  */
 #define BCS_SWCTRL _MMIO(0x22200)
 
+/* There are 16 GPR registers */
+#define BCS_GPR(n)	_MMIO(0x22600 + (n) * 8)
+#define BCS_GPR_UDW(n)	_MMIO(0x22600 + (n) * 8 + 4)
+
 #define GPGPU_THREADS_DISPATCHED        _MMIO(0x2290)
 #define GPGPU_THREADS_DISPATCHED_UDW	_MMIO(0x2290 + 4)
 #define HS_INVOCATION_COUNT             _MMIO(0x2300)
@@ -682,6 +693,45 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
 #define OABUFFER_SIZE_8M    (6 << 3)
 #define OABUFFER_SIZE_16M   (7 << 3)
 
+/* Gen12 OAR unit */
+#define GEN12_OAR_OACONTROL _MMIO(0x2960)
+#define  GEN12_OAR_OACONTROL_COUNTER_FORMAT_SHIFT 1
+#define  GEN12_OAR_OACONTROL_COUNTER_ENABLE       (1 << 0)
+
+#define GEN12_OACTXCONTROL _MMIO(0x2360)
+#define GEN12_OAR_OASTATUS _MMIO(0x2968)
+
+/* Gen12 OAG unit */
+#define GEN12_OAG_OAHEADPTR _MMIO(0xdb00)
+#define  GEN12_OAG_OAHEADPTR_MASK 0xffffffc0
+#define GEN12_OAG_OATAILPTR _MMIO(0xdb04)
+#define  GEN12_OAG_OATAILPTR_MASK 0xffffffc0
+
+#define GEN12_OAG_OABUFFER  _MMIO(0xdb08)
+#define  GEN12_OAG_OABUFFER_BUFFER_SIZE_MASK  (0x7)
+#define  GEN12_OAG_OABUFFER_BUFFER_SIZE_SHIFT (3)
+#define  GEN12_OAG_OABUFFER_MEMORY_SELECT     (1 << 0) /* 0: PPGTT, 1: GGTT */
+
+#define GEN12_OAG_OAGLBCTXCTRL _MMIO(0x2b28)
+#define  GEN12_OAG_OAGLBCTXCTRL_TIMER_PERIOD_SHIFT 2
+#define  GEN12_OAG_OAGLBCTXCTRL_TIMER_ENABLE       (1 << 1)
+#define  GEN12_OAG_OAGLBCTXCTRL_COUNTER_RESUME     (1 << 0)
+
+#define GEN12_OAG_OACONTROL _MMIO(0xdaf4)
+#define  GEN12_OAG_OACONTROL_OA_COUNTER_FORMAT_SHIFT 2
+#define  GEN12_OAG_OACONTROL_OA_COUNTER_ENABLE       (1 << 0)
+
+#define GEN12_OAG_OA_DEBUG _MMIO(0xdaf8)
+#define  GEN12_OAG_OA_DEBUG_INCLUDE_CLK_RATIO          (1 << 6)
+#define  GEN12_OAG_OA_DEBUG_DISABLE_CLK_RATIO_REPORTS  (1 << 5)
+#define  GEN12_OAG_OA_DEBUG_DISABLE_GO_1_0_REPORTS     (1 << 2)
+#define  GEN12_OAG_OA_DEBUG_DISABLE_CTX_SWITCH_REPORTS (1 << 1)
+
+#define GEN12_OAG_OASTATUS _MMIO(0xdafc)
+#define  GEN12_OAG_OASTATUS_COUNTER_OVERFLOW (1 << 2)
+#define  GEN12_OAG_OASTATUS_BUFFER_OVERFLOW  (1 << 1)
+#define  GEN12_OAG_OASTATUS_REPORT_LOST      (1 << 0)
+
 /*
  * Flexible, Aggregate EU Counter Registers.
  * Note: these aren't contiguous
@@ -918,6 +968,26 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
 #define OAREPORTTRIG8_NOA_SELECT_6_SHIFT    24
 #define OAREPORTTRIG8_NOA_SELECT_7_SHIFT    28
 
+/* Same layout as OASTARTTRIGX */
+#define GEN12_OAG_OASTARTTRIG1 _MMIO(0xd900)
+#define GEN12_OAG_OASTARTTRIG2 _MMIO(0xd904)
+#define GEN12_OAG_OASTARTTRIG3 _MMIO(0xd908)
+#define GEN12_OAG_OASTARTTRIG4 _MMIO(0xd90c)
+#define GEN12_OAG_OASTARTTRIG5 _MMIO(0xd910)
+#define GEN12_OAG_OASTARTTRIG6 _MMIO(0xd914)
+#define GEN12_OAG_OASTARTTRIG7 _MMIO(0xd918)
+#define GEN12_OAG_OASTARTTRIG8 _MMIO(0xd91c)
+
+/* Same layout as OAREPORTTRIGX */
+#define GEN12_OAG_OAREPORTTRIG1 _MMIO(0xd920)
+#define GEN12_OAG_OAREPORTTRIG2 _MMIO(0xd924)
+#define GEN12_OAG_OAREPORTTRIG3 _MMIO(0xd928)
+#define GEN12_OAG_OAREPORTTRIG4 _MMIO(0xd92c)
+#define GEN12_OAG_OAREPORTTRIG5 _MMIO(0xd930)
+#define GEN12_OAG_OAREPORTTRIG6 _MMIO(0xd934)
+#define GEN12_OAG_OAREPORTTRIG7 _MMIO(0xd938)
+#define GEN12_OAG_OAREPORTTRIG8 _MMIO(0xd93c)
+
 /* CECX_0 */
 #define OACEC_COMPARE_LESS_OR_EQUAL	6
 #define OACEC_COMPARE_NOT_EQUAL		5
@@ -934,6 +1004,10 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
 #define OACEC_SELECT_PREV	(1 << 19)
 #define OACEC_SELECT_BOOLEAN	(2 << 19)
 
+/* 11-bit array 0: pass-through, 1: negated */
+#define GEN12_OASCEC_NEGATE_MASK  0x7ff
+#define GEN12_OASCEC_NEGATE_SHIFT 21
+
 /* CECX_1 */
 #define OACEC_MASK_MASK		    0xffff
 #define OACEC_CONSIDERATIONS_MASK   0xffff
@@ -956,6 +1030,42 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
 #define OACEC7_0 _MMIO(0x27a8)
 #define OACEC7_1 _MMIO(0x27ac)
 
+/* Same layout as CECX_Y */
+#define GEN12_OAG_CEC0_0 _MMIO(0xd940)
+#define GEN12_OAG_CEC0_1 _MMIO(0xd944)
+#define GEN12_OAG_CEC1_0 _MMIO(0xd948)
+#define GEN12_OAG_CEC1_1 _MMIO(0xd94c)
+#define GEN12_OAG_CEC2_0 _MMIO(0xd950)
+#define GEN12_OAG_CEC2_1 _MMIO(0xd954)
+#define GEN12_OAG_CEC3_0 _MMIO(0xd958)
+#define GEN12_OAG_CEC3_1 _MMIO(0xd95c)
+#define GEN12_OAG_CEC4_0 _MMIO(0xd960)
+#define GEN12_OAG_CEC4_1 _MMIO(0xd964)
+#define GEN12_OAG_CEC5_0 _MMIO(0xd968)
+#define GEN12_OAG_CEC5_1 _MMIO(0xd96c)
+#define GEN12_OAG_CEC6_0 _MMIO(0xd970)
+#define GEN12_OAG_CEC6_1 _MMIO(0xd974)
+#define GEN12_OAG_CEC7_0 _MMIO(0xd978)
+#define GEN12_OAG_CEC7_1 _MMIO(0xd97c)
+
+/* Same layout as CECX_Y + negate 11-bit array */
+#define GEN12_OAG_SCEC0_0 _MMIO(0xdc00)
+#define GEN12_OAG_SCEC0_1 _MMIO(0xdc04)
+#define GEN12_OAG_SCEC1_0 _MMIO(0xdc08)
+#define GEN12_OAG_SCEC1_1 _MMIO(0xdc0c)
+#define GEN12_OAG_SCEC2_0 _MMIO(0xdc10)
+#define GEN12_OAG_SCEC2_1 _MMIO(0xdc14)
+#define GEN12_OAG_SCEC3_0 _MMIO(0xdc18)
+#define GEN12_OAG_SCEC3_1 _MMIO(0xdc1c)
+#define GEN12_OAG_SCEC4_0 _MMIO(0xdc20)
+#define GEN12_OAG_SCEC4_1 _MMIO(0xdc24)
+#define GEN12_OAG_SCEC5_0 _MMIO(0xdc28)
+#define GEN12_OAG_SCEC5_1 _MMIO(0xdc2c)
+#define GEN12_OAG_SCEC6_0 _MMIO(0xdc30)
+#define GEN12_OAG_SCEC6_1 _MMIO(0xdc34)
+#define GEN12_OAG_SCEC7_0 _MMIO(0xdc38)
+#define GEN12_OAG_SCEC7_1 _MMIO(0xdc3c)
+
 /* OA perf counters */
 #define OA_PERFCNT1_LO      _MMIO(0x91B8)
 #define OA_PERFCNT1_HI      _MMIO(0x91BC)
@@ -1036,6 +1146,10 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
 #define MICRO_BP3_COUNT_STATUS23	_MMIO(0x9838)
 #define MICRO_BP_FIRED_ARMED		_MMIO(0x983C)
 
+#define GEN12_OAA_DBG_REG _MMIO(0xdc44)
+#define GEN12_OAG_OA_PESS _MMIO(0x2b2c)
+#define GEN12_OAG_SPCTR_CNF _MMIO(0xdc40)
+
 #define GDT_CHICKEN_BITS    _MMIO(0x9840)
 #define   GT_NOA_ENABLE	    0x00000080
 
@@ -1956,8 +2070,8 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
 #define ICL_DPHY_CHKN(port)			_MMIO(_ICL_COMBOPHY(port) + _ICL_DPHY_CHKN_REG)
 #define   ICL_DPHY_CHKN_AFE_OVER_PPI_STRAP	REG_BIT(7)
 
-#define MG_PHY_PORT_LN(ln, port, ln0p1, ln0p2, ln1p1) \
-	_MMIO(_PORT((port) - PORT_C, ln0p1, ln0p2) + (ln) * ((ln1p1) - (ln0p1)))
+#define MG_PHY_PORT_LN(ln, tc_port, ln0p1, ln0p2, ln1p1) \
+	_MMIO(_PORT(tc_port, ln0p1, ln0p2) + (ln) * ((ln1p1) - (ln0p1)))
 
 #define MG_TX_LINK_PARAMS_TX1LN0_PORT1		0x16812C
 #define MG_TX_LINK_PARAMS_TX1LN1_PORT1		0x16852C
@@ -1967,10 +2081,10 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
 #define MG_TX_LINK_PARAMS_TX1LN1_PORT3		0x16A52C
 #define MG_TX_LINK_PARAMS_TX1LN0_PORT4		0x16B12C
 #define MG_TX_LINK_PARAMS_TX1LN1_PORT4		0x16B52C
-#define MG_TX1_LINK_PARAMS(ln, port) \
-	MG_PHY_PORT_LN(ln, port, MG_TX_LINK_PARAMS_TX1LN0_PORT1, \
-				 MG_TX_LINK_PARAMS_TX1LN0_PORT2, \
-				 MG_TX_LINK_PARAMS_TX1LN1_PORT1)
+#define MG_TX1_LINK_PARAMS(ln, tc_port) \
+	MG_PHY_PORT_LN(ln, tc_port, MG_TX_LINK_PARAMS_TX1LN0_PORT1, \
+				    MG_TX_LINK_PARAMS_TX1LN0_PORT2, \
+				    MG_TX_LINK_PARAMS_TX1LN1_PORT1)
 
 #define MG_TX_LINK_PARAMS_TX2LN0_PORT1		0x1680AC
 #define MG_TX_LINK_PARAMS_TX2LN1_PORT1		0x1684AC
@@ -1980,10 +2094,10 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
 #define MG_TX_LINK_PARAMS_TX2LN1_PORT3		0x16A4AC
 #define MG_TX_LINK_PARAMS_TX2LN0_PORT4		0x16B0AC
 #define MG_TX_LINK_PARAMS_TX2LN1_PORT4		0x16B4AC
-#define MG_TX2_LINK_PARAMS(ln, port) \
-	MG_PHY_PORT_LN(ln, port, MG_TX_LINK_PARAMS_TX2LN0_PORT1, \
-				 MG_TX_LINK_PARAMS_TX2LN0_PORT2, \
-				 MG_TX_LINK_PARAMS_TX2LN1_PORT1)
+#define MG_TX2_LINK_PARAMS(ln, tc_port) \
+	MG_PHY_PORT_LN(ln, tc_port, MG_TX_LINK_PARAMS_TX2LN0_PORT1, \
+				    MG_TX_LINK_PARAMS_TX2LN0_PORT2, \
+				    MG_TX_LINK_PARAMS_TX2LN1_PORT1)
 #define   CRI_USE_FS32			(1 << 5)
 
 #define MG_TX_PISO_READLOAD_TX1LN0_PORT1		0x16814C
@@ -1994,10 +2108,10 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
 #define MG_TX_PISO_READLOAD_TX1LN1_PORT3		0x16A54C
 #define MG_TX_PISO_READLOAD_TX1LN0_PORT4		0x16B14C
 #define MG_TX_PISO_READLOAD_TX1LN1_PORT4		0x16B54C
-#define MG_TX1_PISO_READLOAD(ln, port) \
-	MG_PHY_PORT_LN(ln, port, MG_TX_PISO_READLOAD_TX1LN0_PORT1, \
-				 MG_TX_PISO_READLOAD_TX1LN0_PORT2, \
-				 MG_TX_PISO_READLOAD_TX1LN1_PORT1)
+#define MG_TX1_PISO_READLOAD(ln, tc_port) \
+	MG_PHY_PORT_LN(ln, tc_port, MG_TX_PISO_READLOAD_TX1LN0_PORT1, \
+				    MG_TX_PISO_READLOAD_TX1LN0_PORT2, \
+				    MG_TX_PISO_READLOAD_TX1LN1_PORT1)
 
 #define MG_TX_PISO_READLOAD_TX2LN0_PORT1		0x1680CC
 #define MG_TX_PISO_READLOAD_TX2LN1_PORT1		0x1684CC
@@ -2007,10 +2121,10 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
 #define MG_TX_PISO_READLOAD_TX2LN1_PORT3		0x16A4CC
 #define MG_TX_PISO_READLOAD_TX2LN0_PORT4		0x16B0CC
 #define MG_TX_PISO_READLOAD_TX2LN1_PORT4		0x16B4CC
-#define MG_TX2_PISO_READLOAD(ln, port) \
-	MG_PHY_PORT_LN(ln, port, MG_TX_PISO_READLOAD_TX2LN0_PORT1, \
-				 MG_TX_PISO_READLOAD_TX2LN0_PORT2, \
-				 MG_TX_PISO_READLOAD_TX2LN1_PORT1)
+#define MG_TX2_PISO_READLOAD(ln, tc_port) \
+	MG_PHY_PORT_LN(ln, tc_port, MG_TX_PISO_READLOAD_TX2LN0_PORT1, \
+				    MG_TX_PISO_READLOAD_TX2LN0_PORT2, \
+				    MG_TX_PISO_READLOAD_TX2LN1_PORT1)
 #define   CRI_CALCINIT					(1 << 1)
 
 #define MG_TX_SWINGCTRL_TX1LN0_PORT1		0x168148
@@ -2021,10 +2135,10 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
 #define MG_TX_SWINGCTRL_TX1LN1_PORT3		0x16A548
 #define MG_TX_SWINGCTRL_TX1LN0_PORT4		0x16B148
 #define MG_TX_SWINGCTRL_TX1LN1_PORT4		0x16B548
-#define MG_TX1_SWINGCTRL(ln, port) \
-	MG_PHY_PORT_LN(ln, port, MG_TX_SWINGCTRL_TX1LN0_PORT1, \
-				 MG_TX_SWINGCTRL_TX1LN0_PORT2, \
-				 MG_TX_SWINGCTRL_TX1LN1_PORT1)
+#define MG_TX1_SWINGCTRL(ln, tc_port) \
+	MG_PHY_PORT_LN(ln, tc_port, MG_TX_SWINGCTRL_TX1LN0_PORT1, \
+				    MG_TX_SWINGCTRL_TX1LN0_PORT2, \
+				    MG_TX_SWINGCTRL_TX1LN1_PORT1)
 
 #define MG_TX_SWINGCTRL_TX2LN0_PORT1		0x1680C8
 #define MG_TX_SWINGCTRL_TX2LN1_PORT1		0x1684C8
@@ -2034,10 +2148,10 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
 #define MG_TX_SWINGCTRL_TX2LN1_PORT3		0x16A4C8
 #define MG_TX_SWINGCTRL_TX2LN0_PORT4		0x16B0C8
 #define MG_TX_SWINGCTRL_TX2LN1_PORT4		0x16B4C8
-#define MG_TX2_SWINGCTRL(ln, port) \
-	MG_PHY_PORT_LN(ln, port, MG_TX_SWINGCTRL_TX2LN0_PORT1, \
-				 MG_TX_SWINGCTRL_TX2LN0_PORT2, \
-				 MG_TX_SWINGCTRL_TX2LN1_PORT1)
+#define MG_TX2_SWINGCTRL(ln, tc_port) \
+	MG_PHY_PORT_LN(ln, tc_port, MG_TX_SWINGCTRL_TX2LN0_PORT1, \
+				    MG_TX_SWINGCTRL_TX2LN0_PORT2, \
+				    MG_TX_SWINGCTRL_TX2LN1_PORT1)
 #define   CRI_TXDEEMPH_OVERRIDE_17_12(x)		((x) << 0)
 #define   CRI_TXDEEMPH_OVERRIDE_17_12_MASK		(0x3F << 0)
 
@@ -2049,10 +2163,10 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
 #define MG_TX_DRVCTRL_TX1LN1_TXPORT3			0x16A544
 #define MG_TX_DRVCTRL_TX1LN0_TXPORT4			0x16B144
 #define MG_TX_DRVCTRL_TX1LN1_TXPORT4			0x16B544
-#define MG_TX1_DRVCTRL(ln, port) \
-	MG_PHY_PORT_LN(ln, port, MG_TX_DRVCTRL_TX1LN0_TXPORT1, \
-				 MG_TX_DRVCTRL_TX1LN0_TXPORT2, \
-				 MG_TX_DRVCTRL_TX1LN1_TXPORT1)
+#define MG_TX1_DRVCTRL(ln, tc_port) \
+	MG_PHY_PORT_LN(ln, tc_port, MG_TX_DRVCTRL_TX1LN0_TXPORT1, \
+				    MG_TX_DRVCTRL_TX1LN0_TXPORT2, \
+				    MG_TX_DRVCTRL_TX1LN1_TXPORT1)
 
 #define MG_TX_DRVCTRL_TX2LN0_PORT1			0x1680C4
 #define MG_TX_DRVCTRL_TX2LN1_PORT1			0x1684C4
@@ -2062,10 +2176,10 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
 #define MG_TX_DRVCTRL_TX2LN1_PORT3			0x16A4C4
 #define MG_TX_DRVCTRL_TX2LN0_PORT4			0x16B0C4
 #define MG_TX_DRVCTRL_TX2LN1_PORT4			0x16B4C4
-#define MG_TX2_DRVCTRL(ln, port) \
-	MG_PHY_PORT_LN(ln, port, MG_TX_DRVCTRL_TX2LN0_PORT1, \
-				 MG_TX_DRVCTRL_TX2LN0_PORT2, \
-				 MG_TX_DRVCTRL_TX2LN1_PORT1)
+#define MG_TX2_DRVCTRL(ln, tc_port) \
+	MG_PHY_PORT_LN(ln, tc_port, MG_TX_DRVCTRL_TX2LN0_PORT1, \
+				    MG_TX_DRVCTRL_TX2LN0_PORT2, \
+				    MG_TX_DRVCTRL_TX2LN1_PORT1)
 #define   CRI_TXDEEMPH_OVERRIDE_11_6(x)			((x) << 24)
 #define   CRI_TXDEEMPH_OVERRIDE_11_6_MASK		(0x3F << 24)
 #define   CRI_TXDEEMPH_OVERRIDE_EN			(1 << 22)
@@ -2082,10 +2196,10 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
 #define MG_CLKHUB_LN1_PORT3			0x16A79C
 #define MG_CLKHUB_LN0_PORT4			0x16B39C
 #define MG_CLKHUB_LN1_PORT4			0x16B79C
-#define MG_CLKHUB(ln, port) \
-	MG_PHY_PORT_LN(ln, port, MG_CLKHUB_LN0_PORT1, \
-				 MG_CLKHUB_LN0_PORT2, \
-				 MG_CLKHUB_LN1_PORT1)
+#define MG_CLKHUB(ln, tc_port) \
+	MG_PHY_PORT_LN(ln, tc_port, MG_CLKHUB_LN0_PORT1, \
+				    MG_CLKHUB_LN0_PORT2, \
+				    MG_CLKHUB_LN1_PORT1)
 #define   CFG_LOW_RATE_LKREN_EN				(1 << 11)
 
 #define MG_TX_DCC_TX1LN0_PORT1			0x168110
@@ -2096,10 +2210,10 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
 #define MG_TX_DCC_TX1LN1_PORT3			0x16A510
 #define MG_TX_DCC_TX1LN0_PORT4			0x16B110
 #define MG_TX_DCC_TX1LN1_PORT4			0x16B510
-#define MG_TX1_DCC(ln, port) \
-	MG_PHY_PORT_LN(ln, port, MG_TX_DCC_TX1LN0_PORT1, \
-				 MG_TX_DCC_TX1LN0_PORT2, \
-				 MG_TX_DCC_TX1LN1_PORT1)
+#define MG_TX1_DCC(ln, tc_port) \
+	MG_PHY_PORT_LN(ln, tc_port, MG_TX_DCC_TX1LN0_PORT1, \
+				    MG_TX_DCC_TX1LN0_PORT2, \
+				    MG_TX_DCC_TX1LN1_PORT1)
 #define MG_TX_DCC_TX2LN0_PORT1			0x168090
 #define MG_TX_DCC_TX2LN1_PORT1			0x168490
 #define MG_TX_DCC_TX2LN0_PORT2			0x169090
@@ -2108,10 +2222,10 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
 #define MG_TX_DCC_TX2LN1_PORT3			0x16A490
 #define MG_TX_DCC_TX2LN0_PORT4			0x16B090
 #define MG_TX_DCC_TX2LN1_PORT4			0x16B490
-#define MG_TX2_DCC(ln, port) \
-	MG_PHY_PORT_LN(ln, port, MG_TX_DCC_TX2LN0_PORT1, \
-				 MG_TX_DCC_TX2LN0_PORT2, \
-				 MG_TX_DCC_TX2LN1_PORT1)
+#define MG_TX2_DCC(ln, tc_port) \
+	MG_PHY_PORT_LN(ln, tc_port, MG_TX_DCC_TX2LN0_PORT1, \
+				    MG_TX_DCC_TX2LN0_PORT2, \
+				    MG_TX_DCC_TX2LN1_PORT1)
 #define   CFG_AMI_CK_DIV_OVERRIDE_VAL(x)	((x) << 25)
 #define   CFG_AMI_CK_DIV_OVERRIDE_VAL_MASK	(0x3 << 25)
 #define   CFG_AMI_CK_DIV_OVERRIDE_EN		(1 << 24)
@@ -2124,10 +2238,10 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
 #define MG_DP_MODE_LN1_ACU_PORT3			0x16A7A0
 #define MG_DP_MODE_LN0_ACU_PORT4			0x16B3A0
 #define MG_DP_MODE_LN1_ACU_PORT4			0x16B7A0
-#define MG_DP_MODE(ln, port)	\
-	MG_PHY_PORT_LN(ln, port, MG_DP_MODE_LN0_ACU_PORT1, \
-				 MG_DP_MODE_LN0_ACU_PORT2, \
-				 MG_DP_MODE_LN1_ACU_PORT1)
+#define MG_DP_MODE(ln, tc_port)	\
+	MG_PHY_PORT_LN(ln, tc_port, MG_DP_MODE_LN0_ACU_PORT1, \
+				    MG_DP_MODE_LN0_ACU_PORT2, \
+				    MG_DP_MODE_LN1_ACU_PORT1)
 #define   MG_DP_MODE_CFG_DP_X2_MODE			(1 << 7)
 #define   MG_DP_MODE_CFG_DP_X1_MODE			(1 << 6)
 #define   MG_DP_MODE_CFG_TR2PWR_GATING			(1 << 5)
@@ -2166,13 +2280,13 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
 #define _MMIO_FIA(fia, off)		_MMIO(_FIA(fia) + (off))
 
 /* ICL PHY DFLEX registers */
-#define PORT_TX_DFLEXDPMLE1(fia)	_MMIO_FIA((fia),  0x008C0)
-#define   DFLEXDPMLE1_DPMLETC_MASK(tc_port)	(0xf << (4 * (tc_port)))
-#define   DFLEXDPMLE1_DPMLETC_ML0(tc_port)	(1 << (4 * (tc_port)))
-#define   DFLEXDPMLE1_DPMLETC_ML1_0(tc_port)	(3 << (4 * (tc_port)))
-#define   DFLEXDPMLE1_DPMLETC_ML3(tc_port)	(8 << (4 * (tc_port)))
-#define   DFLEXDPMLE1_DPMLETC_ML3_2(tc_port)	(12 << (4 * (tc_port)))
-#define   DFLEXDPMLE1_DPMLETC_ML3_0(tc_port)	(15 << (4 * (tc_port)))
+#define PORT_TX_DFLEXDPMLE1(fia)		_MMIO_FIA((fia),  0x008C0)
+#define   DFLEXDPMLE1_DPMLETC_MASK(idx)		(0xf << (4 * (idx)))
+#define   DFLEXDPMLE1_DPMLETC_ML0(idx)		(1 << (4 * (idx)))
+#define   DFLEXDPMLE1_DPMLETC_ML1_0(idx)	(3 << (4 * (idx)))
+#define   DFLEXDPMLE1_DPMLETC_ML3(idx)		(8 << (4 * (idx)))
+#define   DFLEXDPMLE1_DPMLETC_ML3_2(idx)	(12 << (4 * (idx)))
+#define   DFLEXDPMLE1_DPMLETC_ML3_0(idx)	(15 << (4 * (idx)))
 
 /* BXT PHY Ref registers */
 #define _PORT_REF_DW3_A			0x16218C
@@ -2453,6 +2567,7 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
 #define   RING_FAULT_FAULT_TYPE(x) (((x) >> 1) & 0x3)
 #define   RING_FAULT_VALID	(1 << 0)
 #define DONE_REG		_MMIO(0x40b0)
+#define GEN12_GAM_DONE		_MMIO(0xcf68)
 #define GEN8_PRIVATE_PAT_LO	_MMIO(0x40e0)
 #define GEN8_PRIVATE_PAT_HI	_MMIO(0x40e0 + 4)
 #define GEN10_PAT_INDEX(index)	_MMIO(0x40e0 + (index) * 4)
@@ -2483,7 +2598,12 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
 #define   RING_WAIT		(1 << 11) /* gen3+, PRBx_CTL */
 #define   RING_WAIT_SEMAPHORE	(1 << 10) /* gen6+ */
 
+/* There are 16 64-bit CS General Purpose Registers per-engine on Gen8+ */
+#define GEN8_RING_CS_GPR(base, n)	_MMIO((base) + 0x600 + (n) * 8)
+#define GEN8_RING_CS_GPR_UDW(base, n)	_MMIO((base) + 0x600 + (n) * 8 + 4)
+
 #define RING_FORCE_TO_NONPRIV(base, i) _MMIO(((base) + 0x4D0) + (i) * 4)
+#define   RING_FORCE_TO_NONPRIV_ADDRESS_MASK	REG_GENMASK(25, 2)
 #define   RING_FORCE_TO_NONPRIV_ACCESS_RW	(0 << 28)    /* CFL+ & Gen11+ */
 #define   RING_FORCE_TO_NONPRIV_ACCESS_RD	(1 << 28)
 #define   RING_FORCE_TO_NONPRIV_ACCESS_WR	(2 << 28)
@@ -2596,6 +2716,8 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
 #define   FAULT_VA_HIGH_BITS		(0xf << 0)
 #define   FAULT_GTT_SEL			(1 << 4)
 
+#define GEN12_AUX_ERR_DBG		_MMIO(0x43f4)
+
 #define FPGA_DBG		_MMIO(0x42300)
 #define   FPGA_DBG_RM_NOCLAIM	(1 << 31)
 
@@ -2705,6 +2827,7 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
 #define VLV_GU_CTL0	_MMIO(VLV_DISPLAY_BASE + 0x2030)
 #define VLV_GU_CTL1	_MMIO(VLV_DISPLAY_BASE + 0x2034)
 #define SCPD0		_MMIO(0x209c) /* 915+ only */
+#define  CSTATE_RENDER_CLOCK_GATE_DISABLE	(1 << 5)
 #define GEN2_IER	_MMIO(0x20a0)
 #define GEN2_IIR	_MMIO(0x20a4)
 #define GEN2_IMR	_MMIO(0x20a8)
@@ -2878,6 +3001,7 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
 
 #define GEN6_RC_SLEEP_PSMI_CONTROL	_MMIO(0x2050)
 #define   GEN6_PSMI_SLEEP_MSG_DISABLE	(1 << 0)
+#define   GEN12_WAIT_FOR_EVENT_POWER_DOWN_DISABLE REG_BIT(7)
 #define   GEN8_RC_SEMA_IDLE_MSG_DISABLE	(1 << 12)
 #define   GEN8_FF_DOP_CLOCK_GATE_DISABLE	(1 << 10)
 
@@ -2956,6 +3080,8 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
 
 #define GEN11_GT_SUBSLICE_DISABLE _MMIO(0x913C)
 
+#define GEN12_GT_DSS_ENABLE _MMIO(0x913C)
+
 #define GEN6_BSD_SLEEP_PSMI_CONTROL	_MMIO(0x12050)
 #define   GEN6_BSD_SLEEP_MSG_DISABLE	(1 << 0)
 #define   GEN6_BSD_SLEEP_FLUSH_DISABLE	(1 << 2)
@@ -3558,6 +3684,9 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
 #define _PALETTE_A		0xa000
 #define _PALETTE_B		0xa800
 #define _CHV_PALETTE_C		0xc000
+#define PALETTE_RED_MASK        REG_GENMASK(23, 16)
+#define PALETTE_GREEN_MASK      REG_GENMASK(15, 8)
+#define PALETTE_BLUE_MASK       REG_GENMASK(7, 0)
 #define PALETTE(pipe, i)	_MMIO(DISPLAY_MMIO_BASE(dev_priv) + \
 				      _PICK((pipe), _PALETTE_A,		\
 					    _PALETTE_B, _CHV_PALETTE_C) + \
@@ -4038,10 +4167,15 @@ enum {
 #define  SARBUNIT_CLKGATE_DIS		(1 << 5)
 #define  RCCUNIT_CLKGATE_DIS		(1 << 7)
 #define  MSCUNIT_CLKGATE_DIS		(1 << 10)
+#define  L3_CLKGATE_DIS			REG_BIT(16)
+#define  L3_CR2X_CLKGATE_DIS		REG_BIT(17)
 
 #define SUBSLICE_UNIT_LEVEL_CLKGATE	_MMIO(0x9524)
 #define  GWUNIT_CLKGATE_DIS		(1 << 16)
 
+#define SUBSLICE_UNIT_LEVEL_CLKGATE2	_MMIO(0x9528)
+#define  CPSSUNIT_CLKGATE_DIS		REG_BIT(9)
+
 #define UNSLICE_UNIT_LEVEL_CLKGATE	_MMIO(0x9434)
 #define  VFUNIT_CLKGATE_DIS		(1 << 20)
 
@@ -4135,6 +4269,7 @@ enum {
 #define _VTOTAL_A	0x6000c
 #define _VBLANK_A	0x60010
 #define _VSYNC_A	0x60014
+#define _EXITLINE_A	0x60018
 #define _PIPEASRC	0x6001c
 #define _BCLRPAT_A	0x60020
 #define _VSYNCSHIFT_A	0x60028
@@ -4186,10 +4321,22 @@ enum {
 #define PIPESRC(trans)		_MMIO_TRANS2(trans, _PIPEASRC)
 #define PIPE_MULT(trans)	_MMIO_TRANS2(trans, _PIPE_MULT_A)
 
-/* HSW+ eDP PSR registers */
-#define HSW_EDP_PSR_BASE	0x64800
-#define BDW_EDP_PSR_BASE	0x6f800
-#define EDP_PSR_CTL				_MMIO(dev_priv->psr_mmio_base + 0)
+#define EXITLINE(trans)		_MMIO_TRANS2(trans, _EXITLINE_A)
+#define   EXITLINE_ENABLE	REG_BIT(31)
+#define   EXITLINE_MASK		REG_GENMASK(12, 0)
+#define   EXITLINE_SHIFT	0
+
+/*
+ * HSW+ eDP PSR registers
+ *
+ * HSW PSR registers are relative to DDIA(_DDI_BUF_CTL_A + 0x800) with just one
+ * instance of it
+ */
+#define _HSW_EDP_PSR_BASE			0x64800
+#define _SRD_CTL_A				0x60800
+#define _SRD_CTL_EDP				0x6f800
+#define _PSR_ADJ(tran, reg)			(_TRANS2(tran, reg) - dev_priv->hsw_psr_mmio_adjust)
+#define EDP_PSR_CTL(tran)			_MMIO(_PSR_ADJ(tran, _SRD_CTL_A))
 #define   EDP_PSR_ENABLE			(1 << 31)
 #define   BDW_PSR_SINGLE_FRAME			(1 << 30)
 #define   EDP_PSR_RESTORE_PSR_ACTIVE_CTX_MASK	(1 << 29) /* SW can't modify */
@@ -4215,27 +4362,40 @@ enum {
 #define   EDP_PSR_TP1_TIME_0us			(3 << 4)
 #define   EDP_PSR_IDLE_FRAME_SHIFT		0
 
-/* Bspec claims those aren't shifted but stay at 0x64800 */
+/*
+ * Until TGL, IMR/IIR are fixed at 0x648xx. On TGL+ those registers are relative
+ * to transcoder and bits defined for each one as if using no shift (i.e. as if
+ * it was for TRANSCODER_EDP)
+ */
 #define EDP_PSR_IMR				_MMIO(0x64834)
 #define EDP_PSR_IIR				_MMIO(0x64838)
-#define   EDP_PSR_ERROR(shift)			(1 << ((shift) + 2))
-#define   EDP_PSR_POST_EXIT(shift)		(1 << ((shift) + 1))
-#define   EDP_PSR_PRE_ENTRY(shift)		(1 << (shift))
-#define   EDP_PSR_TRANSCODER_C_SHIFT		24
-#define   EDP_PSR_TRANSCODER_B_SHIFT		16
-#define   EDP_PSR_TRANSCODER_A_SHIFT		8
-#define   EDP_PSR_TRANSCODER_EDP_SHIFT		0
-
-#define EDP_PSR_AUX_CTL				_MMIO(dev_priv->psr_mmio_base + 0x10)
+#define _PSR_IMR_A				0x60814
+#define _PSR_IIR_A				0x60818
+#define TRANS_PSR_IMR(tran)			_MMIO_TRANS2(tran, _PSR_IMR_A)
+#define TRANS_PSR_IIR(tran)			_MMIO_TRANS2(tran, _PSR_IIR_A)
+#define   _EDP_PSR_TRANS_SHIFT(trans)		((trans) == TRANSCODER_EDP ? \
+						 0 : ((trans) - TRANSCODER_A + 1) * 8)
+#define   EDP_PSR_TRANS_MASK(trans)		(0x7 << _EDP_PSR_TRANS_SHIFT(trans))
+#define   EDP_PSR_ERROR(trans)			(0x4 << _EDP_PSR_TRANS_SHIFT(trans))
+#define   EDP_PSR_POST_EXIT(trans)		(0x2 << _EDP_PSR_TRANS_SHIFT(trans))
+#define   EDP_PSR_PRE_ENTRY(trans)		(0x1 << _EDP_PSR_TRANS_SHIFT(trans))
+
+#define _SRD_AUX_CTL_A				0x60810
+#define _SRD_AUX_CTL_EDP			0x6f810
+#define EDP_PSR_AUX_CTL(tran)			_MMIO(_PSR_ADJ(tran, _SRD_AUX_CTL_A))
 #define   EDP_PSR_AUX_CTL_TIME_OUT_MASK		(3 << 26)
 #define   EDP_PSR_AUX_CTL_MESSAGE_SIZE_MASK	(0x1f << 20)
 #define   EDP_PSR_AUX_CTL_PRECHARGE_2US_MASK	(0xf << 16)
 #define   EDP_PSR_AUX_CTL_ERROR_INTERRUPT	(1 << 11)
 #define   EDP_PSR_AUX_CTL_BIT_CLOCK_2X_MASK	(0x7ff)
 
-#define EDP_PSR_AUX_DATA(i)			_MMIO(dev_priv->psr_mmio_base + 0x14 + (i) * 4) /* 5 registers */
+#define _SRD_AUX_DATA_A				0x60814
+#define _SRD_AUX_DATA_EDP			0x6f814
+#define EDP_PSR_AUX_DATA(tran, i)		_MMIO(_PSR_ADJ(tran, _SRD_AUX_DATA_A) + (i) + 4) /* 5 registers */
 
-#define EDP_PSR_STATUS				_MMIO(dev_priv->psr_mmio_base + 0x40)
+#define _SRD_STATUS_A				0x60840
+#define _SRD_STATUS_EDP				0x6f840
+#define EDP_PSR_STATUS(tran)			_MMIO(_PSR_ADJ(tran, _SRD_STATUS_A))
 #define   EDP_PSR_STATUS_STATE_MASK		(7 << 29)
 #define   EDP_PSR_STATUS_STATE_SHIFT		29
 #define   EDP_PSR_STATUS_STATE_IDLE		(0 << 29)
@@ -4260,10 +4420,15 @@ enum {
 #define   EDP_PSR_STATUS_SENDING_TP1		(1 << 4)
 #define   EDP_PSR_STATUS_IDLE_MASK		0xf
 
-#define EDP_PSR_PERF_CNT		_MMIO(dev_priv->psr_mmio_base + 0x44)
+#define _SRD_PERF_CNT_A			0x60844
+#define _SRD_PERF_CNT_EDP		0x6f844
+#define EDP_PSR_PERF_CNT(tran)		_MMIO(_PSR_ADJ(tran, _SRD_PERF_CNT_A))
 #define   EDP_PSR_PERF_CNT_MASK		0xffffff
 
-#define EDP_PSR_DEBUG				_MMIO(dev_priv->psr_mmio_base + 0x60) /* PSR_MASK on SKL+ */
+/* PSR_MASK on SKL+ */
+#define _SRD_DEBUG_A				0x60860
+#define _SRD_DEBUG_EDP				0x6f860
+#define EDP_PSR_DEBUG(tran)			_MMIO(_PSR_ADJ(tran, _SRD_DEBUG_A))
 #define   EDP_PSR_DEBUG_MASK_MAX_SLEEP         (1 << 28)
 #define   EDP_PSR_DEBUG_MASK_LPSP              (1 << 27)
 #define   EDP_PSR_DEBUG_MASK_MEMUP             (1 << 26)
@@ -4271,7 +4436,9 @@ enum {
 #define   EDP_PSR_DEBUG_MASK_DISP_REG_WRITE    (1 << 16) /* Reserved in ICL+ */
 #define   EDP_PSR_DEBUG_EXIT_ON_PIXEL_UNDERRUN (1 << 15) /* SKL+ */
 
-#define EDP_PSR2_CTL			_MMIO(0x6f900)
+#define _PSR2_CTL_A			0x60900
+#define _PSR2_CTL_EDP			0x6f900
+#define EDP_PSR2_CTL(tran)		_MMIO_TRANS2(tran, _PSR2_CTL_A)
 #define   EDP_PSR2_ENABLE		(1 << 31)
 #define   EDP_SU_TRACK_ENABLE		(1 << 30)
 #define   EDP_Y_COORDINATE_VALID	(1 << 26) /* GLK and CNL+ */
@@ -4293,8 +4460,8 @@ enum {
 #define _PSR_EVENT_TRANS_B			0x61848
 #define _PSR_EVENT_TRANS_C			0x62848
 #define _PSR_EVENT_TRANS_D			0x63848
-#define _PSR_EVENT_TRANS_EDP			0x6F848
-#define PSR_EVENT(trans)			_MMIO_TRANS2(trans, _PSR_EVENT_TRANS_A)
+#define _PSR_EVENT_TRANS_EDP			0x6f848
+#define PSR_EVENT(tran)				_MMIO_TRANS2(tran, _PSR_EVENT_TRANS_A)
 #define  PSR_EVENT_PSR2_WD_TIMER_EXPIRE		(1 << 17)
 #define  PSR_EVENT_PSR2_DISABLED		(1 << 16)
 #define  PSR_EVENT_SU_DIRTY_FIFO_UNDERRUN	(1 << 15)
@@ -4312,15 +4479,16 @@ enum {
 #define  PSR_EVENT_LPSP_MODE_EXIT		(1 << 1)
 #define  PSR_EVENT_PSR_DISABLE			(1 << 0)
 
-#define EDP_PSR2_STATUS			_MMIO(0x6f940)
+#define _PSR2_STATUS_A			0x60940
+#define _PSR2_STATUS_EDP		0x6f940
+#define EDP_PSR2_STATUS(tran)		_MMIO_TRANS2(tran, _PSR2_STATUS_A)
 #define EDP_PSR2_STATUS_STATE_MASK     (0xf << 28)
 #define EDP_PSR2_STATUS_STATE_SHIFT    28
 
-#define _PSR2_SU_STATUS_0		0x6F914
-#define _PSR2_SU_STATUS_1		0x6F918
-#define _PSR2_SU_STATUS_2		0x6F91C
-#define _PSR2_SU_STATUS(index)		_MMIO(_PICK_EVEN((index), _PSR2_SU_STATUS_0, _PSR2_SU_STATUS_1))
-#define PSR2_SU_STATUS(frame)		(_PSR2_SU_STATUS((frame) / 3))
+#define _PSR2_SU_STATUS_A		0x60914
+#define _PSR2_SU_STATUS_EDP		0x6f914
+#define _PSR2_SU_STATUS(tran, index)	_MMIO(_TRANS2(tran, _PSR2_SU_STATUS_A) + (index) * 4)
+#define PSR2_SU_STATUS(tran, frame)	(_PSR2_SU_STATUS(tran, (frame) / 3))
 #define PSR2_SU_STATUS_SHIFT(frame)	(((frame) % 3) * 10)
 #define PSR2_SU_STATUS_MASK(frame)	(0x3ff << PSR2_SU_STATUS_SHIFT(frame))
 #define PSR2_SU_STATUS_FRAMES		8
@@ -4646,6 +4814,7 @@ enum {
  * (Haswell and newer) to see which VIDEO_DIP_DATA byte corresponds to each byte
  * of the infoframe structure specified by CEA-861. */
 #define   VIDEO_DIP_DATA_SIZE	32
+#define   VIDEO_DIP_GMP_DATA_SIZE	36
 #define   VIDEO_DIP_VSC_DATA_SIZE	36
 #define   VIDEO_DIP_PPS_DATA_SIZE	132
 #define VIDEO_DIP_CTL		_MMIO(0x61170)
@@ -5482,45 +5651,9 @@ enum {
  */
 #define _DPA_AUX_CH_CTL		(DISPLAY_MMIO_BASE(dev_priv) + 0x64010)
 #define _DPA_AUX_CH_DATA1	(DISPLAY_MMIO_BASE(dev_priv) + 0x64014)
-#define _DPA_AUX_CH_DATA2	(DISPLAY_MMIO_BASE(dev_priv) + 0x64018)
-#define _DPA_AUX_CH_DATA3	(DISPLAY_MMIO_BASE(dev_priv) + 0x6401c)
-#define _DPA_AUX_CH_DATA4	(DISPLAY_MMIO_BASE(dev_priv) + 0x64020)
-#define _DPA_AUX_CH_DATA5	(DISPLAY_MMIO_BASE(dev_priv) + 0x64024)
 
 #define _DPB_AUX_CH_CTL		(DISPLAY_MMIO_BASE(dev_priv) + 0x64110)
 #define _DPB_AUX_CH_DATA1	(DISPLAY_MMIO_BASE(dev_priv) + 0x64114)
-#define _DPB_AUX_CH_DATA2	(DISPLAY_MMIO_BASE(dev_priv) + 0x64118)
-#define _DPB_AUX_CH_DATA3	(DISPLAY_MMIO_BASE(dev_priv) + 0x6411c)
-#define _DPB_AUX_CH_DATA4	(DISPLAY_MMIO_BASE(dev_priv) + 0x64120)
-#define _DPB_AUX_CH_DATA5	(DISPLAY_MMIO_BASE(dev_priv) + 0x64124)
-
-#define _DPC_AUX_CH_CTL		(DISPLAY_MMIO_BASE(dev_priv) + 0x64210)
-#define _DPC_AUX_CH_DATA1	(DISPLAY_MMIO_BASE(dev_priv) + 0x64214)
-#define _DPC_AUX_CH_DATA2	(DISPLAY_MMIO_BASE(dev_priv) + 0x64218)
-#define _DPC_AUX_CH_DATA3	(DISPLAY_MMIO_BASE(dev_priv) + 0x6421c)
-#define _DPC_AUX_CH_DATA4	(DISPLAY_MMIO_BASE(dev_priv) + 0x64220)
-#define _DPC_AUX_CH_DATA5	(DISPLAY_MMIO_BASE(dev_priv) + 0x64224)
-
-#define _DPD_AUX_CH_CTL		(DISPLAY_MMIO_BASE(dev_priv) + 0x64310)
-#define _DPD_AUX_CH_DATA1	(DISPLAY_MMIO_BASE(dev_priv) + 0x64314)
-#define _DPD_AUX_CH_DATA2	(DISPLAY_MMIO_BASE(dev_priv) + 0x64318)
-#define _DPD_AUX_CH_DATA3	(DISPLAY_MMIO_BASE(dev_priv) + 0x6431c)
-#define _DPD_AUX_CH_DATA4	(DISPLAY_MMIO_BASE(dev_priv) + 0x64320)
-#define _DPD_AUX_CH_DATA5	(DISPLAY_MMIO_BASE(dev_priv) + 0x64324)
-
-#define _DPE_AUX_CH_CTL		(DISPLAY_MMIO_BASE(dev_priv) + 0x64410)
-#define _DPE_AUX_CH_DATA1	(DISPLAY_MMIO_BASE(dev_priv) + 0x64414)
-#define _DPE_AUX_CH_DATA2	(DISPLAY_MMIO_BASE(dev_priv) + 0x64418)
-#define _DPE_AUX_CH_DATA3	(DISPLAY_MMIO_BASE(dev_priv) + 0x6441c)
-#define _DPE_AUX_CH_DATA4	(DISPLAY_MMIO_BASE(dev_priv) + 0x64420)
-#define _DPE_AUX_CH_DATA5	(DISPLAY_MMIO_BASE(dev_priv) + 0x64424)
-
-#define _DPF_AUX_CH_CTL		(DISPLAY_MMIO_BASE(dev_priv) + 0x64510)
-#define _DPF_AUX_CH_DATA1	(DISPLAY_MMIO_BASE(dev_priv) + 0x64514)
-#define _DPF_AUX_CH_DATA2	(DISPLAY_MMIO_BASE(dev_priv) + 0x64518)
-#define _DPF_AUX_CH_DATA3	(DISPLAY_MMIO_BASE(dev_priv) + 0x6451c)
-#define _DPF_AUX_CH_DATA4	(DISPLAY_MMIO_BASE(dev_priv) + 0x64520)
-#define _DPF_AUX_CH_DATA5	(DISPLAY_MMIO_BASE(dev_priv) + 0x64524)
 
 #define DP_AUX_CH_CTL(aux_ch)	_MMIO_PORT(aux_ch, _DPA_AUX_CH_CTL, _DPB_AUX_CH_CTL)
 #define DP_AUX_CH_DATA(aux_ch, i)	_MMIO(_PORT(aux_ch, _DPA_AUX_CH_DATA1, _DPB_AUX_CH_DATA1) + (i) * 4) /* 5 registers */
@@ -5652,6 +5785,11 @@ enum {
 #define   PIPECONF_CXSR_DOWNCLOCK	(1 << 16)
 #define   PIPECONF_EDP_RR_MODE_SWITCH_VLV	(1 << 14)
 #define   PIPECONF_COLOR_RANGE_SELECT	(1 << 13)
+#define   PIPECONF_OUTPUT_COLORSPACE_MASK	(3 << 11) /* ilk-ivb */
+#define   PIPECONF_OUTPUT_COLORSPACE_RGB	(0 << 11) /* ilk-ivb */
+#define   PIPECONF_OUTPUT_COLORSPACE_YUV601	(1 << 11) /* ilk-ivb */
+#define   PIPECONF_OUTPUT_COLORSPACE_YUV709	(2 << 11) /* ilk-ivb */
+#define   PIPECONF_OUTPUT_COLORSPACE_YUV_HSW	(1 << 11) /* hsw only */
 #define   PIPECONF_BPC_MASK	(0x7 << 5)
 #define   PIPECONF_8BPC		(0 << 5)
 #define   PIPECONF_10BPC	(1 << 5)
@@ -5739,12 +5877,13 @@ enum {
 
 #define  _PIPEAGCMAX           0x70010
 #define  _PIPEBGCMAX           0x71010
+#define PIPEGCMAX_RGB_MASK     REG_GENMASK(15, 0)
 #define PIPEGCMAX(pipe, i)     _MMIO_PIPE2(pipe, _PIPEAGCMAX + (i) * 4)
 
 #define _PIPE_MISC_A			0x70030
 #define _PIPE_MISC_B			0x71030
-#define   PIPEMISC_YUV420_ENABLE	(1 << 27)
-#define   PIPEMISC_YUV420_MODE_FULL_BLEND (1 << 26)
+#define   PIPEMISC_YUV420_ENABLE	(1 << 27) /* glk+ */
+#define   PIPEMISC_YUV420_MODE_FULL_BLEND (1 << 26) /* glk+ */
 #define   PIPEMISC_HDR_MODE_PRECISION	(1 << 23) /* icl+ */
 #define   PIPEMISC_OUTPUT_COLORSPACE_YUV  (1 << 11)
 #define   PIPEMISC_DITHER_BPC_MASK	(7 << 5)
@@ -6201,6 +6340,7 @@ enum {
 #define CHV_CURSOR_C_OFFSET 0x700e0
 #define IVB_CURSOR_B_OFFSET 0x71080
 #define IVB_CURSOR_C_OFFSET 0x72080
+#define TGL_CURSOR_D_OFFSET 0x73080
 
 /* Display A control */
 #define _DSPACNTR				0x70180
@@ -7171,11 +7311,17 @@ enum {
 /* legacy palette */
 #define _LGC_PALETTE_A           0x4a000
 #define _LGC_PALETTE_B           0x4a800
+#define LGC_PALETTE_RED_MASK     REG_GENMASK(23, 16)
+#define LGC_PALETTE_GREEN_MASK   REG_GENMASK(15, 8)
+#define LGC_PALETTE_BLUE_MASK    REG_GENMASK(7, 0)
 #define LGC_PALETTE(pipe, i) _MMIO(_PIPE(pipe, _LGC_PALETTE_A, _LGC_PALETTE_B) + (i) * 4)
 
 /* ilk/snb precision palette */
 #define _PREC_PALETTE_A           0x4b000
 #define _PREC_PALETTE_B           0x4c000
+#define   PREC_PALETTE_RED_MASK   REG_GENMASK(29, 20)
+#define   PREC_PALETTE_GREEN_MASK REG_GENMASK(19, 10)
+#define   PREC_PALETTE_BLUE_MASK  REG_GENMASK(9, 0)
 #define PREC_PALETTE(pipe, i) _MMIO(_PIPE(pipe, _PREC_PALETTE_A, _PREC_PALETTE_B) + (i) * 4)
 
 #define  _PREC_PIPEAGCMAX              0x4d000
@@ -7211,6 +7357,12 @@ enum {
 #define TGL_DMC_DEBUG_DC5_COUNT	_MMIO(0x101084)
 #define TGL_DMC_DEBUG_DC6_COUNT	_MMIO(0x101088)
 
+#define DMC_DEBUG3		_MMIO(0x101090)
+
+/* Display Internal Timeout Register */
+#define RM_TIMEOUT		_MMIO(0x42060)
+#define  MMIO_TIMEOUT_US(us)	((us) << 0)
+
 /* interrupts */
 #define DE_MASTER_IRQ_CONTROL   (1 << 31)
 #define DE_SPRITEB_FLIP_DONE    (1 << 29)
@@ -7322,6 +7474,9 @@ enum {
 #define  GEN8_PIPE_VSYNC		(1 << 1)
 #define  GEN8_PIPE_VBLANK		(1 << 0)
 #define  GEN9_PIPE_CURSOR_FAULT		(1 << 11)
+#define  GEN11_PIPE_PLANE7_FAULT	(1 << 22)
+#define  GEN11_PIPE_PLANE6_FAULT	(1 << 21)
+#define  GEN11_PIPE_PLANE5_FAULT	(1 << 20)
 #define  GEN9_PIPE_PLANE4_FAULT		(1 << 10)
 #define  GEN9_PIPE_PLANE3_FAULT		(1 << 9)
 #define  GEN9_PIPE_PLANE2_FAULT		(1 << 8)
@@ -7341,6 +7496,11 @@ enum {
 	 GEN9_PIPE_PLANE3_FAULT | \
 	 GEN9_PIPE_PLANE2_FAULT | \
 	 GEN9_PIPE_PLANE1_FAULT)
+#define GEN11_DE_PIPE_IRQ_FAULT_ERRORS \
+	(GEN9_DE_PIPE_IRQ_FAULT_ERRORS | \
+	 GEN11_PIPE_PLANE7_FAULT | \
+	 GEN11_PIPE_PLANE6_FAULT | \
+	 GEN11_PIPE_PLANE5_FAULT)
 
 #define GEN8_DE_PORT_ISR _MMIO(0x44440)
 #define GEN8_DE_PORT_IMR _MMIO(0x44444)
@@ -7360,6 +7520,12 @@ enum {
 #define  GEN8_PORT_DP_A_HOTPLUG		(1 << 3)
 #define  BXT_DE_PORT_GMBUS		(1 << 1)
 #define  GEN8_AUX_CHANNEL_A		(1 << 0)
+#define  TGL_DE_PORT_AUX_USBC6		(1 << 13)
+#define  TGL_DE_PORT_AUX_USBC5		(1 << 12)
+#define  TGL_DE_PORT_AUX_USBC4		(1 << 11)
+#define  TGL_DE_PORT_AUX_USBC3		(1 << 10)
+#define  TGL_DE_PORT_AUX_USBC2		(1 << 9)
+#define  TGL_DE_PORT_AUX_USBC1		(1 << 8)
 #define  TGL_DE_PORT_AUX_DDIC		(1 << 2)
 #define  TGL_DE_PORT_AUX_DDIB		(1 << 1)
 #define  TGL_DE_PORT_AUX_DDIA		(1 << 0)
@@ -7548,10 +7714,17 @@ enum {
 #define  BDW_DPRS_MASK_VBLANK_SRD	(1 << 0)
 #define CHICKEN_PIPESL_1(pipe) _MMIO_PIPE(pipe, _CHICKEN_PIPESL_1_A, _CHICKEN_PIPESL_1_B)
 
-#define CHICKEN_TRANS_A		_MMIO(0x420c0)
-#define CHICKEN_TRANS_B		_MMIO(0x420c4)
-#define CHICKEN_TRANS_C		_MMIO(0x420c8)
-#define CHICKEN_TRANS_EDP	_MMIO(0x420cc)
+#define _CHICKEN_TRANS_A	0x420c0
+#define _CHICKEN_TRANS_B	0x420c4
+#define _CHICKEN_TRANS_C	0x420c8
+#define _CHICKEN_TRANS_EDP	0x420cc
+#define _CHICKEN_TRANS_D	0x420d8
+#define CHICKEN_TRANS(trans)	_MMIO(_PICK((trans), \
+					    [TRANSCODER_EDP] = _CHICKEN_TRANS_EDP, \
+					    [TRANSCODER_A] = _CHICKEN_TRANS_A, \
+					    [TRANSCODER_B] = _CHICKEN_TRANS_B, \
+					    [TRANSCODER_C] = _CHICKEN_TRANS_C, \
+					    [TRANSCODER_D] = _CHICKEN_TRANS_D))
 #define  VSC_DATA_SEL_SOFTWARE_CONTROL	(1 << 25) /* GLK and CNL+ */
 #define  DDI_TRAINING_OVERRIDE_ENABLE	(1 << 19)
 #define  DDI_TRAINING_OVERRIDE_VALUE	(1 << 18)
@@ -7584,15 +7757,19 @@ enum {
 #define   CNL_DDI_CLOCK_REG_ACCESS_ON	(1 << 7)
 
 #define SKL_DFSM			_MMIO(0x51000)
-#define SKL_DFSM_CDCLK_LIMIT_MASK	(3 << 23)
-#define SKL_DFSM_CDCLK_LIMIT_675	(0 << 23)
-#define SKL_DFSM_CDCLK_LIMIT_540	(1 << 23)
-#define SKL_DFSM_CDCLK_LIMIT_450	(2 << 23)
-#define SKL_DFSM_CDCLK_LIMIT_337_5	(3 << 23)
-#define SKL_DFSM_PIPE_A_DISABLE		(1 << 30)
-#define SKL_DFSM_PIPE_B_DISABLE		(1 << 21)
-#define SKL_DFSM_PIPE_C_DISABLE		(1 << 28)
-#define TGL_DFSM_PIPE_D_DISABLE		(1 << 22)
+#define   SKL_DFSM_DISPLAY_PM_DISABLE	(1 << 27)
+#define   SKL_DFSM_DISPLAY_HDCP_DISABLE	(1 << 25)
+#define   SKL_DFSM_CDCLK_LIMIT_MASK	(3 << 23)
+#define   SKL_DFSM_CDCLK_LIMIT_675	(0 << 23)
+#define   SKL_DFSM_CDCLK_LIMIT_540	(1 << 23)
+#define   SKL_DFSM_CDCLK_LIMIT_450	(2 << 23)
+#define   SKL_DFSM_CDCLK_LIMIT_337_5	(3 << 23)
+#define   ICL_DFSM_DMC_DISABLE		(1 << 23)
+#define   SKL_DFSM_PIPE_A_DISABLE	(1 << 30)
+#define   SKL_DFSM_PIPE_B_DISABLE	(1 << 21)
+#define   SKL_DFSM_PIPE_C_DISABLE	(1 << 28)
+#define   TGL_DFSM_PIPE_D_DISABLE	(1 << 22)
+#define   CNL_DFSM_DISPLAY_DSC_DISABLE	(1 << 7)
 
 #define SKL_DSSM				_MMIO(0x51004)
 #define CNL_DSSM_CDCLK_PLL_REFCLK_24MHz		(1 << 31)
@@ -7609,7 +7786,10 @@ enum {
 #define  GEN9_POOLED_EU_LOAD_BALANCING_FIX_DISABLE  (1 << 10)
 
 #define GEN9_CS_DEBUG_MODE1		_MMIO(0x20ec)
+#define   FF_DOP_CLOCK_GATE_DISABLE	REG_BIT(1)
 #define GEN9_CTX_PREEMPT_REG		_MMIO(0x2248)
+#define   GEN12_DISABLE_POSH_BUSY_FF_DOP_CG REG_BIT(11)
+
 #define GEN8_CS_CHICKEN1		_MMIO(0x2580)
 #define GEN9_PREEMPT_3D_OBJECT_LEVEL		(1 << 0)
 #define GEN9_PREEMPT_GPGPU_LEVEL(hi, lo)	(((hi) << 2) | ((lo) << 1))
@@ -7634,6 +7814,7 @@ enum {
 
 #define GEN11_COMMON_SLICE_CHICKEN3		_MMIO(0x7304)
   #define GEN11_BLEND_EMB_FIX_DISABLE_IN_RCC	(1 << 11)
+  #define GEN12_DISABLE_CPS_AWARE_COLOR_PIPE	(1 << 9)
 
 #define HIZ_CHICKEN					_MMIO(0x7018)
 # define CHV_HZ_8X8_MODE_IN_1X				(1 << 15)
@@ -7818,29 +7999,24 @@ enum {
 				 SDE_FDI_RXA_CPT)
 
 /* south display engine interrupt: ICP/TGP */
-#define SDE_TC6_HOTPLUG_TGP		(1 << 29)
-#define SDE_TC5_HOTPLUG_TGP		(1 << 28)
-#define SDE_TC4_HOTPLUG_ICP		(1 << 27)
-#define SDE_TC3_HOTPLUG_ICP		(1 << 26)
-#define SDE_TC2_HOTPLUG_ICP		(1 << 25)
-#define SDE_TC1_HOTPLUG_ICP		(1 << 24)
 #define SDE_GMBUS_ICP			(1 << 23)
-#define SDE_DDIC_HOTPLUG_TGP		(1 << 18)
-#define SDE_DDIB_HOTPLUG_ICP		(1 << 17)
-#define SDE_DDIA_HOTPLUG_ICP		(1 << 16)
 #define SDE_TC_HOTPLUG_ICP(tc_port)	(1 << ((tc_port) + 24))
 #define SDE_DDI_HOTPLUG_ICP(port)	(1 << ((port) + 16))
-#define SDE_DDI_MASK_ICP		(SDE_DDIB_HOTPLUG_ICP |	\
-					 SDE_DDIA_HOTPLUG_ICP)
-#define SDE_TC_MASK_ICP			(SDE_TC4_HOTPLUG_ICP |	\
-					 SDE_TC3_HOTPLUG_ICP |	\
-					 SDE_TC2_HOTPLUG_ICP |	\
-					 SDE_TC1_HOTPLUG_ICP)
-#define SDE_DDI_MASK_TGP		(SDE_DDIC_HOTPLUG_TGP | \
-					 SDE_DDI_MASK_ICP)
-#define SDE_TC_MASK_TGP			(SDE_TC6_HOTPLUG_TGP |	\
-					 SDE_TC5_HOTPLUG_TGP |	\
-					 SDE_TC_MASK_ICP)
+#define SDE_DDI_MASK_ICP		(SDE_DDI_HOTPLUG_ICP(PORT_B) | \
+					 SDE_DDI_HOTPLUG_ICP(PORT_A))
+#define SDE_TC_MASK_ICP			(SDE_TC_HOTPLUG_ICP(PORT_TC4) | \
+					 SDE_TC_HOTPLUG_ICP(PORT_TC3) | \
+					 SDE_TC_HOTPLUG_ICP(PORT_TC2) | \
+					 SDE_TC_HOTPLUG_ICP(PORT_TC1))
+#define SDE_DDI_MASK_TGP		(SDE_DDI_HOTPLUG_ICP(PORT_C) | \
+					 SDE_DDI_HOTPLUG_ICP(PORT_B) | \
+					 SDE_DDI_HOTPLUG_ICP(PORT_A))
+#define SDE_TC_MASK_TGP			(SDE_TC_HOTPLUG_ICP(PORT_TC6) | \
+					 SDE_TC_HOTPLUG_ICP(PORT_TC5) | \
+					 SDE_TC_HOTPLUG_ICP(PORT_TC4) | \
+					 SDE_TC_HOTPLUG_ICP(PORT_TC3) | \
+					 SDE_TC_HOTPLUG_ICP(PORT_TC2) | \
+					 SDE_TC_HOTPLUG_ICP(PORT_TC1))
 
 #define SDEISR  _MMIO(0xc4000)
 #define SDEIMR  _MMIO(0xc4004)
@@ -7907,26 +8083,13 @@ enum {
  * SHOTPLUG_CTL_DDI and SHOTPLUG_CTL_TC.
  */
 
-#define SHOTPLUG_CTL_DDI			_MMIO(0xc4030)
-#define   TGP_DDIC_HPD_ENABLE			(1 << 11)
-#define   TGP_DDIC_HPD_STATUS_MASK		(3 << 8)
-#define   TGP_DDIC_HPD_NO_DETECT		(0 << 8)
-#define   TGP_DDIC_HPD_SHORT_DETECT		(1 << 8)
-#define   TGP_DDIC_HPD_LONG_DETECT		(2 << 8)
-#define   TGP_DDIC_HPD_SHORT_LONG_DETECT	(3 << 8)
-#define   ICP_DDIB_HPD_ENABLE			(1 << 7)
-#define   ICP_DDIB_HPD_STATUS_MASK		(3 << 4)
-#define   ICP_DDIB_HPD_NO_DETECT		(0 << 4)
-#define   ICP_DDIB_HPD_SHORT_DETECT		(1 << 4)
-#define   ICP_DDIB_HPD_LONG_DETECT		(2 << 4)
-#define   ICP_DDIB_HPD_SHORT_LONG_DETECT	(3 << 4)
-#define   ICP_DDIA_HPD_ENABLE			(1 << 3)
-#define   ICP_DDIA_HPD_OP_DRIVE_1		(1 << 2)
-#define   ICP_DDIA_HPD_STATUS_MASK		(3 << 0)
-#define   ICP_DDIA_HPD_NO_DETECT		(0 << 0)
-#define   ICP_DDIA_HPD_SHORT_DETECT		(1 << 0)
-#define   ICP_DDIA_HPD_LONG_DETECT		(2 << 0)
-#define   ICP_DDIA_HPD_SHORT_LONG_DETECT	(3 << 0)
+#define SHOTPLUG_CTL_DDI				_MMIO(0xc4030)
+#define   SHOTPLUG_CTL_DDI_HPD_ENABLE(port)		(0x8 << (4 * (port)))
+#define   SHOTPLUG_CTL_DDI_HPD_STATUS_MASK(port)	(0x3 << (4 * (port)))
+#define   SHOTPLUG_CTL_DDI_HPD_NO_DETECT(port)		(0x0 << (4 * (port)))
+#define   SHOTPLUG_CTL_DDI_HPD_SHORT_DETECT(port)	(0x1 << (4 * (port)))
+#define   SHOTPLUG_CTL_DDI_HPD_LONG_DETECT(port)	(0x2 << (4 * (port)))
+#define   SHOTPLUG_CTL_DDI_HPD_SHORT_LONG_DETECT(port)	(0x3 << (4 * (port)))
 
 #define SHOTPLUG_CTL_TC				_MMIO(0xc4034)
 #define   ICP_TC_HPD_ENABLE(tc_port)		(8 << (tc_port) * 4)
@@ -8037,14 +8200,15 @@ enum {
 #define   ICP_TC_HPD_LONG_DETECT(tc_port)	(2 << (tc_port) * 4)
 #define   ICP_TC_HPD_SHORT_DETECT(tc_port)	(1 << (tc_port) * 4)
 
-#define ICP_DDI_HPD_ENABLE_MASK		(ICP_DDIB_HPD_ENABLE |	\
-					 ICP_DDIA_HPD_ENABLE)
+#define ICP_DDI_HPD_ENABLE_MASK		(SHOTPLUG_CTL_DDI_HPD_ENABLE(PORT_B) | \
+					 SHOTPLUG_CTL_DDI_HPD_ENABLE(PORT_A))
 #define ICP_TC_HPD_ENABLE_MASK		(ICP_TC_HPD_ENABLE(PORT_TC4) | \
 					 ICP_TC_HPD_ENABLE(PORT_TC3) | \
 					 ICP_TC_HPD_ENABLE(PORT_TC2) | \
 					 ICP_TC_HPD_ENABLE(PORT_TC1))
-#define TGP_DDI_HPD_ENABLE_MASK		(TGP_DDIC_HPD_ENABLE |	\
-					 ICP_DDI_HPD_ENABLE_MASK)
+#define TGP_DDI_HPD_ENABLE_MASK		(SHOTPLUG_CTL_DDI_HPD_ENABLE(PORT_C) | \
+					 SHOTPLUG_CTL_DDI_HPD_ENABLE(PORT_B) | \
+					 SHOTPLUG_CTL_DDI_HPD_ENABLE(PORT_A))
 #define TGP_TC_HPD_ENABLE_MASK		(ICP_TC_HPD_ENABLE(PORT_TC6) | \
 					 ICP_TC_HPD_ENABLE(PORT_TC5) | \
 					 ICP_TC_HPD_ENABLE_MASK)
@@ -8594,6 +8758,10 @@ enum {
 #define   GEN9_PWRGT_MEDIA_STATUS_MASK		(1 << 0)
 #define   GEN9_PWRGT_RENDER_STATUS_MASK		(1 << 1)
 
+#define POWERGATE_ENABLE			_MMIO(0xa210)
+#define    VDN_HCP_POWERGATE_ENABLE(n)		BIT(((n) * 2) + 3)
+#define    VDN_MFX_POWERGATE_ENABLE(n)		BIT(((n) * 2) + 4)
+
 #define  GTFIFODBG				_MMIO(0x120000)
 #define    GT_FIFO_SBDEDICATE_FREE_ENTRY_CHV	(0x1f << 20)
 #define    GT_FIFO_FREE_ENTRIES_CHV		(0x7f << 13)
@@ -8831,6 +8999,7 @@ enum {
 #define     GEN9_SAGV_DISABLE			0x0
 #define     GEN9_SAGV_IS_DISABLED		0x1
 #define     GEN9_SAGV_ENABLE			0x3
+#define GEN12_PCODE_READ_SAGV_BLOCK_TIME_US	0x23
 #define GEN6_PCODE_DATA				_MMIO(0x138128)
 #define   GEN6_PCODE_FREQ_IA_RATIO_SHIFT	8
 #define   GEN6_PCODE_FREQ_RING_RATIO_SHIFT	16
@@ -9094,6 +9263,10 @@ enum {
 #define HSW_AUD_CHICKENBIT			_MMIO(0x65f10)
 #define   SKL_AUD_CODEC_WAKE_SIGNAL		(1 << 15)
 
+#define AUD_FREQ_CNTRL			_MMIO(0x65900)
+#define AUD_PIN_BUF_CTL		_MMIO(0x48414)
+#define   AUD_PIN_BUF_ENABLE		REG_BIT(31)
+
 /*
  * HSW - ICL power wells
  *
@@ -9256,12 +9429,20 @@ enum skl_power_gate {
 
 /* HDCP Repeater Registers */
 #define HDCP_REP_CTL			_MMIO(0x66d00)
+#define  HDCP_TRANSA_REP_PRESENT	BIT(31)
+#define  HDCP_TRANSB_REP_PRESENT	BIT(30)
+#define  HDCP_TRANSC_REP_PRESENT	BIT(29)
+#define  HDCP_TRANSD_REP_PRESENT	BIT(28)
 #define  HDCP_DDIB_REP_PRESENT		BIT(30)
 #define  HDCP_DDIA_REP_PRESENT		BIT(29)
 #define  HDCP_DDIC_REP_PRESENT		BIT(28)
 #define  HDCP_DDID_REP_PRESENT		BIT(27)
 #define  HDCP_DDIF_REP_PRESENT		BIT(26)
 #define  HDCP_DDIE_REP_PRESENT		BIT(25)
+#define  HDCP_TRANSA_SHA1_M0		(1 << 20)
+#define  HDCP_TRANSB_SHA1_M0		(2 << 20)
+#define  HDCP_TRANSC_SHA1_M0		(3 << 20)
+#define  HDCP_TRANSD_SHA1_M0		(4 << 20)
 #define  HDCP_DDIB_SHA1_M0		(1 << 20)
 #define  HDCP_DDIA_SHA1_M0		(2 << 20)
 #define  HDCP_DDIC_SHA1_M0		(3 << 20)
@@ -9301,15 +9482,92 @@ enum skl_power_gate {
 					  _PORTE_HDCP_AUTHENC, \
 					  _PORTF_HDCP_AUTHENC) + (x))
 #define PORT_HDCP_CONF(port)		_PORT_HDCP_AUTHENC(port, 0x0)
+#define _TRANSA_HDCP_CONF		0x66400
+#define _TRANSB_HDCP_CONF		0x66500
+#define TRANS_HDCP_CONF(trans)		_MMIO_TRANS(trans, _TRANSA_HDCP_CONF, \
+						    _TRANSB_HDCP_CONF)
+#define HDCP_CONF(dev_priv, trans, port) \
+					(INTEL_GEN(dev_priv) >= 12 ? \
+					 TRANS_HDCP_CONF(trans) : \
+					 PORT_HDCP_CONF(port))
+
 #define  HDCP_CONF_CAPTURE_AN		BIT(0)
 #define  HDCP_CONF_AUTH_AND_ENC		(BIT(1) | BIT(0))
 #define PORT_HDCP_ANINIT(port)		_PORT_HDCP_AUTHENC(port, 0x4)
+#define _TRANSA_HDCP_ANINIT		0x66404
+#define _TRANSB_HDCP_ANINIT		0x66504
+#define TRANS_HDCP_ANINIT(trans)	_MMIO_TRANS(trans, \
+						    _TRANSA_HDCP_ANINIT, \
+						    _TRANSB_HDCP_ANINIT)
+#define HDCP_ANINIT(dev_priv, trans, port) \
+					(INTEL_GEN(dev_priv) >= 12 ? \
+					 TRANS_HDCP_ANINIT(trans) : \
+					 PORT_HDCP_ANINIT(port))
+
 #define PORT_HDCP_ANLO(port)		_PORT_HDCP_AUTHENC(port, 0x8)
+#define _TRANSA_HDCP_ANLO		0x66408
+#define _TRANSB_HDCP_ANLO		0x66508
+#define TRANS_HDCP_ANLO(trans)		_MMIO_TRANS(trans, _TRANSA_HDCP_ANLO, \
+						    _TRANSB_HDCP_ANLO)
+#define HDCP_ANLO(dev_priv, trans, port) \
+					(INTEL_GEN(dev_priv) >= 12 ? \
+					 TRANS_HDCP_ANLO(trans) : \
+					 PORT_HDCP_ANLO(port))
+
 #define PORT_HDCP_ANHI(port)		_PORT_HDCP_AUTHENC(port, 0xC)
+#define _TRANSA_HDCP_ANHI		0x6640C
+#define _TRANSB_HDCP_ANHI		0x6650C
+#define TRANS_HDCP_ANHI(trans)		_MMIO_TRANS(trans, _TRANSA_HDCP_ANHI, \
+						    _TRANSB_HDCP_ANHI)
+#define HDCP_ANHI(dev_priv, trans, port) \
+					(INTEL_GEN(dev_priv) >= 12 ? \
+					 TRANS_HDCP_ANHI(trans) : \
+					 PORT_HDCP_ANHI(port))
+
 #define PORT_HDCP_BKSVLO(port)		_PORT_HDCP_AUTHENC(port, 0x10)
+#define _TRANSA_HDCP_BKSVLO		0x66410
+#define _TRANSB_HDCP_BKSVLO		0x66510
+#define TRANS_HDCP_BKSVLO(trans)	_MMIO_TRANS(trans, \
+						    _TRANSA_HDCP_BKSVLO, \
+						    _TRANSB_HDCP_BKSVLO)
+#define HDCP_BKSVLO(dev_priv, trans, port) \
+					(INTEL_GEN(dev_priv) >= 12 ? \
+					 TRANS_HDCP_BKSVLO(trans) : \
+					 PORT_HDCP_BKSVLO(port))
+
 #define PORT_HDCP_BKSVHI(port)		_PORT_HDCP_AUTHENC(port, 0x14)
+#define _TRANSA_HDCP_BKSVHI		0x66414
+#define _TRANSB_HDCP_BKSVHI		0x66514
+#define TRANS_HDCP_BKSVHI(trans)	_MMIO_TRANS(trans, \
+						    _TRANSA_HDCP_BKSVHI, \
+						    _TRANSB_HDCP_BKSVHI)
+#define HDCP_BKSVHI(dev_priv, trans, port) \
+					(INTEL_GEN(dev_priv) >= 12 ? \
+					 TRANS_HDCP_BKSVHI(trans) : \
+					 PORT_HDCP_BKSVHI(port))
+
 #define PORT_HDCP_RPRIME(port)		_PORT_HDCP_AUTHENC(port, 0x18)
+#define _TRANSA_HDCP_RPRIME		0x66418
+#define _TRANSB_HDCP_RPRIME		0x66518
+#define TRANS_HDCP_RPRIME(trans)	_MMIO_TRANS(trans, \
+						    _TRANSA_HDCP_RPRIME, \
+						    _TRANSB_HDCP_RPRIME)
+#define HDCP_RPRIME(dev_priv, trans, port) \
+					(INTEL_GEN(dev_priv) >= 12 ? \
+					 TRANS_HDCP_RPRIME(trans) : \
+					 PORT_HDCP_RPRIME(port))
+
 #define PORT_HDCP_STATUS(port)		_PORT_HDCP_AUTHENC(port, 0x1C)
+#define _TRANSA_HDCP_STATUS		0x6641C
+#define _TRANSB_HDCP_STATUS		0x6651C
+#define TRANS_HDCP_STATUS(trans)	_MMIO_TRANS(trans, \
+						    _TRANSA_HDCP_STATUS, \
+						    _TRANSB_HDCP_STATUS)
+#define HDCP_STATUS(dev_priv, trans, port) \
+					(INTEL_GEN(dev_priv) >= 12 ? \
+					 TRANS_HDCP_STATUS(trans) : \
+					 PORT_HDCP_STATUS(port))
+
 #define  HDCP_STATUS_STREAM_A_ENC	BIT(31)
 #define  HDCP_STATUS_STREAM_B_ENC	BIT(30)
 #define  HDCP_STATUS_STREAM_C_ENC	BIT(29)
@@ -9336,23 +9594,44 @@ enum skl_power_gate {
 					  _PORTD_HDCP2_BASE, \
 					  _PORTE_HDCP2_BASE, \
 					  _PORTF_HDCP2_BASE) + (x))
-
-#define HDCP2_AUTH_DDI(port)		_PORT_HDCP2_BASE(port, 0x98)
+#define PORT_HDCP2_AUTH(port)		_PORT_HDCP2_BASE(port, 0x98)
+#define _TRANSA_HDCP2_AUTH		0x66498
+#define _TRANSB_HDCP2_AUTH		0x66598
+#define TRANS_HDCP2_AUTH(trans)		_MMIO_TRANS(trans, _TRANSA_HDCP2_AUTH, \
+						    _TRANSB_HDCP2_AUTH)
 #define   AUTH_LINK_AUTHENTICATED	BIT(31)
 #define   AUTH_LINK_TYPE		BIT(30)
 #define   AUTH_FORCE_CLR_INPUTCTR	BIT(19)
 #define   AUTH_CLR_KEYS			BIT(18)
-
-#define HDCP2_CTL_DDI(port)		_PORT_HDCP2_BASE(port, 0xB0)
+#define HDCP2_AUTH(dev_priv, trans, port) \
+					(INTEL_GEN(dev_priv) >= 12 ? \
+					 TRANS_HDCP2_AUTH(trans) : \
+					 PORT_HDCP2_AUTH(port))
+
+#define PORT_HDCP2_CTL(port)		_PORT_HDCP2_BASE(port, 0xB0)
+#define _TRANSA_HDCP2_CTL		0x664B0
+#define _TRANSB_HDCP2_CTL		0x665B0
+#define TRANS_HDCP2_CTL(trans)		_MMIO_TRANS(trans, _TRANSA_HDCP2_CTL, \
+						    _TRANSB_HDCP2_CTL)
 #define   CTL_LINK_ENCRYPTION_REQ	BIT(31)
-
-#define HDCP2_STATUS_DDI(port)		_PORT_HDCP2_BASE(port, 0xB4)
-#define   STREAM_ENCRYPTION_STATUS_A	BIT(31)
-#define   STREAM_ENCRYPTION_STATUS_B	BIT(30)
-#define   STREAM_ENCRYPTION_STATUS_C	BIT(29)
+#define HDCP2_CTL(dev_priv, trans, port) \
+					(INTEL_GEN(dev_priv) >= 12 ? \
+					 TRANS_HDCP2_CTL(trans) : \
+					 PORT_HDCP2_CTL(port))
+
+#define PORT_HDCP2_STATUS(port)		_PORT_HDCP2_BASE(port, 0xB4)
+#define _TRANSA_HDCP2_STATUS		0x664B4
+#define _TRANSB_HDCP2_STATUS		0x665B4
+#define TRANS_HDCP2_STATUS(trans)	_MMIO_TRANS(trans, \
+						    _TRANSA_HDCP2_STATUS, \
+						    _TRANSB_HDCP2_STATUS)
 #define   LINK_TYPE_STATUS		BIT(22)
 #define   LINK_AUTH_STATUS		BIT(21)
 #define   LINK_ENCRYPTION_STATUS	BIT(20)
+#define HDCP2_STATUS(dev_priv, trans, port) \
+					(INTEL_GEN(dev_priv) >= 12 ? \
+					 TRANS_HDCP2_STATUS(trans) : \
+					 PORT_HDCP2_STATUS(port))
 
 /* Per-pipe DDI Function Control */
 #define _TRANS_DDI_FUNC_CTL_A		0x60400
@@ -9392,6 +9671,9 @@ enum skl_power_gate {
 #define  TRANS_DDI_EDP_INPUT_A_ONOFF	(4 << 12)
 #define  TRANS_DDI_EDP_INPUT_B_ONOFF	(5 << 12)
 #define  TRANS_DDI_EDP_INPUT_C_ONOFF	(6 << 12)
+#define  TRANS_DDI_MST_TRANSPORT_SELECT_MASK	REG_GENMASK(11, 10)
+#define  TRANS_DDI_MST_TRANSPORT_SELECT(trans)	\
+	REG_FIELD_PREP(TRANS_DDI_MST_TRANSPORT_SELECT_MASK, trans)
 #define  TRANS_DDI_HDCP_SIGNALLING	(1 << 9)
 #define  TRANS_DDI_DP_VC_PAYLOAD_ALLOC	(1 << 8)
 #define  TRANS_DDI_HDMI_SCRAMBLER_CTS_ENABLE (1 << 7)
@@ -9419,7 +9701,9 @@ enum skl_power_gate {
 /* DisplayPort Transport Control */
 #define _DP_TP_CTL_A			0x64040
 #define _DP_TP_CTL_B			0x64140
+#define _TGL_DP_TP_CTL_A		0x60540
 #define DP_TP_CTL(port) _MMIO_PORT(port, _DP_TP_CTL_A, _DP_TP_CTL_B)
+#define TGL_DP_TP_CTL(tran) _MMIO_TRANS2((tran), _TGL_DP_TP_CTL_A)
 #define  DP_TP_CTL_ENABLE			(1 << 31)
 #define  DP_TP_CTL_FEC_ENABLE			(1 << 30)
 #define  DP_TP_CTL_MODE_SST			(0 << 27)
@@ -9439,7 +9723,9 @@ enum skl_power_gate {
 /* DisplayPort Transport Status */
 #define _DP_TP_STATUS_A			0x64044
 #define _DP_TP_STATUS_B			0x64144
+#define _TGL_DP_TP_STATUS_A		0x60544
 #define DP_TP_STATUS(port) _MMIO_PORT(port, _DP_TP_STATUS_A, _DP_TP_STATUS_B)
+#define TGL_DP_TP_STATUS(tran) _MMIO_TRANS2((tran), _TGL_DP_TP_STATUS_A)
 #define  DP_TP_STATUS_FEC_ENABLE_LIVE		(1 << 28)
 #define  DP_TP_STATUS_IDLE_DONE			(1 << 25)
 #define  DP_TP_STATUS_ACT_SENT			(1 << 24)
@@ -9594,17 +9880,7 @@ enum skl_power_gate {
 #define _TRANSC_MSA_MISC		0x62410
 #define _TRANS_EDP_MSA_MISC		0x6f410
 #define TRANS_MSA_MISC(tran) _MMIO_TRANS2(tran, _TRANSA_MSA_MISC)
-
-#define  TRANS_MSA_SYNC_CLK		(1 << 0)
-#define  TRANS_MSA_SAMPLING_444		(2 << 1)
-#define  TRANS_MSA_CLRSP_YCBCR		(2 << 3)
-#define  TRANS_MSA_6_BPC		(0 << 5)
-#define  TRANS_MSA_8_BPC		(1 << 5)
-#define  TRANS_MSA_10_BPC		(2 << 5)
-#define  TRANS_MSA_12_BPC		(3 << 5)
-#define  TRANS_MSA_16_BPC		(4 << 5)
-#define  TRANS_MSA_CEA_RANGE		(1 << 3)
-#define  TRANS_MSA_USE_VSC_SDP		(1 << 14)
+/* See DP_MSA_MISC_* for the bit definitions */
 
 /* LCPLL Control */
 #define LCPLL_CTL			_MMIO(0x130040)
@@ -9645,7 +9921,10 @@ enum skl_power_gate {
 #define  BXT_CDCLK_CD2X_PIPE(pipe)	((pipe) << 20)
 #define  CDCLK_DIVMUX_CD_OVERRIDE	(1 << 19)
 #define  BXT_CDCLK_CD2X_PIPE_NONE	BXT_CDCLK_CD2X_PIPE(3)
+#define  ICL_CDCLK_CD2X_PIPE(pipe)	(_PICK(pipe, 0, 2, 6) << 19)
 #define  ICL_CDCLK_CD2X_PIPE_NONE	(7 << 19)
+#define  TGL_CDCLK_CD2X_PIPE(pipe)	BXT_CDCLK_CD2X_PIPE(pipe)
+#define  TGL_CDCLK_CD2X_PIPE_NONE	ICL_CDCLK_CD2X_PIPE_NONE
 #define  BXT_CDCLK_SSA_PRECHARGE_ENABLE	(1 << 16)
 #define  CDCLK_FREQ_DECIMAL_MASK	(0x7ff)
 
@@ -9966,6 +10245,166 @@ enum skl_power_gate {
 						   _TGL_DPLL1_CFGCR1, \
 						   _TGL_TBTPLL_CFGCR1)
 
+#define _DKL_PHY1_BASE			0x168000
+#define _DKL_PHY2_BASE			0x169000
+#define _DKL_PHY3_BASE			0x16A000
+#define _DKL_PHY4_BASE			0x16B000
+#define _DKL_PHY5_BASE			0x16C000
+#define _DKL_PHY6_BASE			0x16D000
+
+/* DEKEL PHY MMIO Address = Phy base + (internal address & ~index_mask) */
+#define _DKL_PLL_DIV0			0x200
+#define   DKL_PLL_DIV0_INTEG_COEFF(x)	((x) << 16)
+#define   DKL_PLL_DIV0_INTEG_COEFF_MASK	(0x1F << 16)
+#define   DKL_PLL_DIV0_PROP_COEFF(x)	((x) << 12)
+#define   DKL_PLL_DIV0_PROP_COEFF_MASK	(0xF << 12)
+#define   DKL_PLL_DIV0_FBPREDIV_SHIFT   (8)
+#define   DKL_PLL_DIV0_FBPREDIV(x)	((x) << DKL_PLL_DIV0_FBPREDIV_SHIFT)
+#define   DKL_PLL_DIV0_FBPREDIV_MASK	(0xF << DKL_PLL_DIV0_FBPREDIV_SHIFT)
+#define   DKL_PLL_DIV0_FBDIV_INT(x)	((x) << 0)
+#define   DKL_PLL_DIV0_FBDIV_INT_MASK	(0xFF << 0)
+#define DKL_PLL_DIV0(tc_port)		_MMIO(_PORT(tc_port, _DKL_PHY1_BASE, \
+						    _DKL_PHY2_BASE) + \
+						    _DKL_PLL_DIV0)
+
+#define _DKL_PLL_DIV1				0x204
+#define   DKL_PLL_DIV1_IREF_TRIM(x)		((x) << 16)
+#define   DKL_PLL_DIV1_IREF_TRIM_MASK		(0x1F << 16)
+#define   DKL_PLL_DIV1_TDC_TARGET_CNT(x)	((x) << 0)
+#define   DKL_PLL_DIV1_TDC_TARGET_CNT_MASK	(0xFF << 0)
+#define DKL_PLL_DIV1(tc_port)		_MMIO(_PORT(tc_port, _DKL_PHY1_BASE, \
+						    _DKL_PHY2_BASE) + \
+						    _DKL_PLL_DIV1)
+
+#define _DKL_PLL_SSC				0x210
+#define   DKL_PLL_SSC_IREF_NDIV_RATIO(x)	((x) << 29)
+#define   DKL_PLL_SSC_IREF_NDIV_RATIO_MASK	(0x7 << 29)
+#define   DKL_PLL_SSC_STEP_LEN(x)		((x) << 16)
+#define   DKL_PLL_SSC_STEP_LEN_MASK		(0xFF << 16)
+#define   DKL_PLL_SSC_STEP_NUM(x)		((x) << 11)
+#define   DKL_PLL_SSC_STEP_NUM_MASK		(0x7 << 11)
+#define   DKL_PLL_SSC_EN			(1 << 9)
+#define DKL_PLL_SSC(tc_port)		_MMIO(_PORT(tc_port, _DKL_PHY1_BASE, \
+						    _DKL_PHY2_BASE) + \
+						    _DKL_PLL_SSC)
+
+#define _DKL_PLL_BIAS			0x214
+#define   DKL_PLL_BIAS_FRAC_EN_H	(1 << 30)
+#define   DKL_PLL_BIAS_FBDIV_SHIFT	(8)
+#define   DKL_PLL_BIAS_FBDIV_FRAC(x)	((x) << DKL_PLL_BIAS_FBDIV_SHIFT)
+#define   DKL_PLL_BIAS_FBDIV_FRAC_MASK	(0x3FFFFF << DKL_PLL_BIAS_FBDIV_SHIFT)
+#define DKL_PLL_BIAS(tc_port)		_MMIO(_PORT(tc_port, _DKL_PHY1_BASE, \
+						    _DKL_PHY2_BASE) + \
+						    _DKL_PLL_BIAS)
+
+#define _DKL_PLL_TDC_COLDST_BIAS		0x218
+#define   DKL_PLL_TDC_SSC_STEP_SIZE(x)		((x) << 8)
+#define   DKL_PLL_TDC_SSC_STEP_SIZE_MASK	(0xFF << 8)
+#define   DKL_PLL_TDC_FEED_FWD_GAIN(x)		((x) << 0)
+#define   DKL_PLL_TDC_FEED_FWD_GAIN_MASK	(0xFF << 0)
+#define DKL_PLL_TDC_COLDST_BIAS(tc_port) _MMIO(_PORT(tc_port, \
+						     _DKL_PHY1_BASE, \
+						     _DKL_PHY2_BASE) + \
+						     _DKL_PLL_TDC_COLDST_BIAS)
+
+#define _DKL_REFCLKIN_CTL		0x12C
+/* Bits are the same as MG_REFCLKIN_CTL */
+#define DKL_REFCLKIN_CTL(tc_port)	_MMIO(_PORT(tc_port, \
+						    _DKL_PHY1_BASE, \
+						    _DKL_PHY2_BASE) + \
+					      _DKL_REFCLKIN_CTL)
+
+#define _DKL_CLKTOP2_HSCLKCTL		0xD4
+/* Bits are the same as MG_CLKTOP2_HSCLKCTL */
+#define DKL_CLKTOP2_HSCLKCTL(tc_port)	_MMIO(_PORT(tc_port, \
+						    _DKL_PHY1_BASE, \
+						    _DKL_PHY2_BASE) + \
+					      _DKL_CLKTOP2_HSCLKCTL)
+
+#define _DKL_CLKTOP2_CORECLKCTL1		0xD8
+/* Bits are the same as MG_CLKTOP2_CORECLKCTL1 */
+#define DKL_CLKTOP2_CORECLKCTL1(tc_port)	_MMIO(_PORT(tc_port, \
+							    _DKL_PHY1_BASE, \
+							    _DKL_PHY2_BASE) + \
+						      _DKL_CLKTOP2_CORECLKCTL1)
+
+#define _DKL_TX_DPCNTL0				0x2C0
+#define  DKL_TX_PRESHOOT_COEFF(x)			((x) << 13)
+#define  DKL_TX_PRESHOOT_COEFF_MASK			(0x1f << 13)
+#define  DKL_TX_DE_EMPHASIS_COEFF(x)		((x) << 8)
+#define  DKL_TX_DE_EMPAHSIS_COEFF_MASK		(0x1f << 8)
+#define  DKL_TX_VSWING_CONTROL(x)			((x) << 0)
+#define  DKL_TX_VSWING_CONTROL_MASK			(0x7 << 0)
+#define DKL_TX_DPCNTL0(tc_port) _MMIO(_PORT(tc_port, \
+						     _DKL_PHY1_BASE, \
+						     _DKL_PHY2_BASE) + \
+						     _DKL_TX_DPCNTL0)
+
+#define _DKL_TX_DPCNTL1				0x2C4
+/* Bits are the same as DKL_TX_DPCNTRL0 */
+#define DKL_TX_DPCNTL1(tc_port) _MMIO(_PORT(tc_port, \
+						     _DKL_PHY1_BASE, \
+						     _DKL_PHY2_BASE) + \
+						     _DKL_TX_DPCNTL1)
+
+#define _DKL_TX_DPCNTL2				0x2C8
+#define  DKL_TX_DP20BITMODE				(1 << 2)
+#define DKL_TX_DPCNTL2(tc_port) _MMIO(_PORT(tc_port, \
+						     _DKL_PHY1_BASE, \
+						     _DKL_PHY2_BASE) + \
+						     _DKL_TX_DPCNTL2)
+
+#define _DKL_TX_FW_CALIB				0x2F8
+#define  DKL_TX_CFG_DISABLE_WAIT_INIT			(1 << 7)
+#define DKL_TX_FW_CALIB(tc_port) _MMIO(_PORT(tc_port, \
+						     _DKL_PHY1_BASE, \
+						     _DKL_PHY2_BASE) + \
+						     _DKL_TX_FW_CALIB)
+
+#define _DKL_TX_PMD_LANE_SUS				0xD00
+#define DKL_TX_PMD_LANE_SUS(tc_port) _MMIO(_PORT(tc_port, \
+							  _DKL_PHY1_BASE, \
+							  _DKL_PHY2_BASE) + \
+							  _DKL_TX_PMD_LANE_SUS)
+
+#define _DKL_TX_DW17					0xDC4
+#define DKL_TX_DW17(tc_port) _MMIO(_PORT(tc_port, \
+						     _DKL_PHY1_BASE, \
+						     _DKL_PHY2_BASE) + \
+						     _DKL_TX_DW17)
+
+#define _DKL_TX_DW18					0xDC8
+#define DKL_TX_DW18(tc_port) _MMIO(_PORT(tc_port, \
+						     _DKL_PHY1_BASE, \
+						     _DKL_PHY2_BASE) + \
+						     _DKL_TX_DW18)
+
+#define _DKL_DP_MODE					0xA0
+#define DKL_DP_MODE(tc_port) _MMIO(_PORT(tc_port, \
+						     _DKL_PHY1_BASE, \
+						     _DKL_PHY2_BASE) + \
+						     _DKL_DP_MODE)
+
+#define _DKL_CMN_UC_DW27			0x36C
+#define  DKL_CMN_UC_DW27_UC_HEALTH		(0x1 << 15)
+#define DKL_CMN_UC_DW_27(tc_port)		_MMIO(_PORT(tc_port, \
+							    _DKL_PHY1_BASE, \
+							    _DKL_PHY2_BASE) + \
+							    _DKL_CMN_UC_DW27)
+
+/*
+ * Each Dekel PHY is addressed through a 4KB aperture. Each PHY has more than
+ * 4KB of register space, so a separate index is programmed in HIP_INDEX_REG0
+ * or HIP_INDEX_REG1, based on the port number, to set the upper 2 address
+ * bits that point the 4KB window into the full PHY register space.
+ */
+#define _HIP_INDEX_REG0			0x1010A0
+#define _HIP_INDEX_REG1			0x1010A4
+#define HIP_INDEX_REG(tc_port)		_MMIO((tc_port) < 4 ? _HIP_INDEX_REG0 \
+					      : _HIP_INDEX_REG1)
+#define _HIP_INDEX_SHIFT(tc_port)	(8 * ((tc_port) % 4))
+#define HIP_INDEX_VAL(tc_port, val)	((val) << _HIP_INDEX_SHIFT(tc_port))
+
 /* BXT display engine PLL */
 #define BXT_DE_PLL_CTL			_MMIO(0x6d000)
 #define   BXT_DE_PLL_RATIO(x)		(x)	/* {60,65,100} * 19.2MHz */
@@ -9980,6 +10419,8 @@ enum skl_power_gate {
 /* GEN9 DC */
 #define DC_STATE_EN			_MMIO(0x45504)
 #define  DC_STATE_DISABLE		0
+#define  DC_STATE_EN_DC3CO		REG_BIT(30)
+#define  DC_STATE_DC3CO_STATUS		REG_BIT(29)
 #define  DC_STATE_EN_UPTO_DC5		(1 << 0)
 #define  DC_STATE_EN_DC9		(1 << 3)
 #define  DC_STATE_EN_UPTO_DC6		(2 << 0)
@@ -10108,11 +10549,11 @@ enum skl_power_gate {
 #define _PIPE_A_CSC_COEFF_BV	0x49024
 
 #define _PIPE_A_CSC_MODE	0x49028
-#define  ICL_CSC_ENABLE			(1 << 31)
-#define  ICL_OUTPUT_CSC_ENABLE		(1 << 30)
-#define  CSC_BLACK_SCREEN_OFFSET	(1 << 2)
-#define  CSC_POSITION_BEFORE_GAMMA	(1 << 1)
-#define  CSC_MODE_YUV_TO_RGB		(1 << 0)
+#define  ICL_CSC_ENABLE			(1 << 31) /* icl+ */
+#define  ICL_OUTPUT_CSC_ENABLE		(1 << 30) /* icl+ */
+#define  CSC_BLACK_SCREEN_OFFSET	(1 << 2) /* ilk/snb */
+#define  CSC_POSITION_BEFORE_GAMMA	(1 << 1) /* pre-glk */
+#define  CSC_MODE_YUV_TO_RGB		(1 << 0) /* ilk/snb */
 
 #define _PIPE_A_CSC_PREOFF_HI	0x49030
 #define _PIPE_A_CSC_PREOFF_ME	0x49034
@@ -10228,6 +10669,9 @@ enum skl_power_gate {
 #define _PAL_PREC_GC_MAX_A	0x4A410
 #define _PAL_PREC_GC_MAX_B	0x4AC10
 #define _PAL_PREC_GC_MAX_C	0x4B410
+#define   PREC_PAL_DATA_RED_MASK	REG_GENMASK(29, 20)
+#define   PREC_PAL_DATA_GREEN_MASK	REG_GENMASK(19, 10)
+#define   PREC_PAL_DATA_BLUE_MASK	REG_GENMASK(9, 0)
 #define _PAL_PREC_EXT_GC_MAX_A	0x4A420
 #define _PAL_PREC_EXT_GC_MAX_B	0x4AC20
 #define _PAL_PREC_EXT_GC_MAX_C	0x4B420
@@ -10280,6 +10724,9 @@ enum skl_power_gate {
 #define   CGM_PIPE_MODE_GAMMA	(1 << 2)
 #define   CGM_PIPE_MODE_CSC	(1 << 1)
 #define   CGM_PIPE_MODE_DEGAMMA	(1 << 0)
+#define   CGM_PIPE_GAMMA_RED_MASK   REG_GENMASK(9, 0)
+#define   CGM_PIPE_GAMMA_GREEN_MASK REG_GENMASK(25, 16)
+#define   CGM_PIPE_GAMMA_BLUE_MASK  REG_GENMASK(9, 0)
 
 #define _CGM_PIPE_B_CSC_COEFF01	(VLV_DISPLAY_BASE + 0x69900)
 #define _CGM_PIPE_B_CSC_COEFF23	(VLV_DISPLAY_BASE + 0x69904)
@@ -11527,16 +11974,31 @@ enum skl_power_gate {
 
 #define PORT_TX_DFLEXDPSP(fia)			_MMIO_FIA((fia), 0x008A0)
 #define   MODULAR_FIA_MASK			(1 << 4)
-#define   TC_LIVE_STATE_TBT(tc_port)		(1 << ((tc_port) * 8 + 6))
-#define   TC_LIVE_STATE_TC(tc_port)		(1 << ((tc_port) * 8 + 5))
-#define   DP_LANE_ASSIGNMENT_SHIFT(tc_port)	((tc_port) * 8)
-#define   DP_LANE_ASSIGNMENT_MASK(tc_port)	(0xf << ((tc_port) * 8))
-#define   DP_LANE_ASSIGNMENT(tc_port, x)	((x) << ((tc_port) * 8))
+#define   TC_LIVE_STATE_TBT(idx)		(1 << ((idx) * 8 + 6))
+#define   TC_LIVE_STATE_TC(idx)			(1 << ((idx) * 8 + 5))
+#define   DP_LANE_ASSIGNMENT_SHIFT(idx)		((idx) * 8)
+#define   DP_LANE_ASSIGNMENT_MASK(idx)		(0xf << ((idx) * 8))
+#define   DP_LANE_ASSIGNMENT(idx, x)		((x) << ((idx) * 8))
 
 #define PORT_TX_DFLEXDPPMS(fia)			_MMIO_FIA((fia), 0x00890)
-#define   DP_PHY_MODE_STATUS_COMPLETED(tc_port)		(1 << (tc_port))
+#define   DP_PHY_MODE_STATUS_COMPLETED(idx)	(1 << (idx))
 
 #define PORT_TX_DFLEXDPCSSS(fia)		_MMIO_FIA((fia), 0x00894)
-#define   DP_PHY_MODE_STATUS_NOT_SAFE(tc_port)		(1 << (tc_port))
+#define   DP_PHY_MODE_STATUS_NOT_SAFE(idx)	(1 << (idx))
+
+#define PORT_TX_DFLEXPA1(fia)			_MMIO_FIA((fia), 0x00880)
+#define   DP_PIN_ASSIGNMENT_SHIFT(idx)		((idx) * 4)
+#define   DP_PIN_ASSIGNMENT_MASK(idx)		(0xf << ((idx) * 4))
+#define   DP_PIN_ASSIGNMENT(idx, x)		((x) << ((idx) * 4))
+
+/* This register controls the Display State Buffer (DSB) engines. */
+#define _DSBSL_INSTANCE_BASE		0x70B00
+#define DSBSL_INSTANCE(pipe, id)	(_DSBSL_INSTANCE_BASE + \
+					 (pipe) * 0x1000 + (id) * 100)
+#define DSB_HEAD(pipe, id)		_MMIO(DSBSL_INSTANCE(pipe, id) + 0x0)
+#define DSB_TAIL(pipe, id)		_MMIO(DSBSL_INSTANCE(pipe, id) + 0x4)
+#define DSB_CTRL(pipe, id)		_MMIO(DSBSL_INSTANCE(pipe, id) + 0x8)
+#define   DSB_ENABLE			(1 << 31)
+#define   DSB_STATUS			(1 << 0)
 
 #endif /* _I915_REG_H_ */
diff --git a/drivers/gpu/drm/i915/i915_request.c b/drivers/gpu/drm/i915/i915_request.c
index a53777dd371c..bbd71af00a91 100644
--- a/drivers/gpu/drm/i915/i915_request.c
+++ b/drivers/gpu/drm/i915/i915_request.c
@@ -31,6 +31,8 @@
 
 #include "gem/i915_gem_context.h"
 #include "gt/intel_context.h"
+#include "gt/intel_ring.h"
+#include "gt/intel_rps.h"
 
 #include "i915_active.h"
 #include "i915_drv.h"
@@ -169,16 +171,17 @@ remove_from_client(struct i915_request *request)
 {
 	struct drm_i915_file_private *file_priv;
 
-	file_priv = READ_ONCE(request->file_priv);
-	if (!file_priv)
+	if (!READ_ONCE(request->file_priv))
 		return;
 
-	spin_lock(&file_priv->mm.lock);
-	if (request->file_priv) {
+	rcu_read_lock();
+	file_priv = xchg(&request->file_priv, NULL);
+	if (file_priv) {
+		spin_lock(&file_priv->mm.lock);
 		list_del(&request->client_link);
-		request->file_priv = NULL;
+		spin_unlock(&file_priv->mm.lock);
 	}
-	spin_unlock(&file_priv->mm.lock);
+	rcu_read_unlock();
 }
 
 static void free_capture_list(struct i915_request *request)
@@ -194,11 +197,29 @@ static void free_capture_list(struct i915_request *request)
 	}
 }
 
-static bool i915_request_retire(struct i915_request *rq)
+static void remove_from_engine(struct i915_request *rq)
 {
-	struct i915_active_request *active, *next;
+	struct intel_engine_cs *engine, *locked;
 
-	lockdep_assert_held(&rq->timeline->mutex);
+	/*
+	 * Virtual engines complicate acquiring the engine timeline lock,
+	 * as their rq->engine pointer is not stable until under that
+	 * engine lock. The simple ploy we use is to take the lock then
+	 * check that the rq still belongs to the newly locked engine.
+	 */
+	locked = READ_ONCE(rq->engine);
+	spin_lock_irq(&locked->active.lock);
+	while (unlikely(locked != (engine = READ_ONCE(rq->engine)))) {
+		spin_unlock(&locked->active.lock);
+		spin_lock(&engine->active.lock);
+		locked = engine;
+	}
+	list_del(&rq->sched.link);
+	spin_unlock_irq(&locked->active.lock);
+}
+
+bool i915_request_retire(struct i915_request *rq)
+{
 	if (!i915_request_completed(rq))
 		return false;
 
@@ -219,68 +240,34 @@ static bool i915_request_retire(struct i915_request *rq)
 	 * Note this requires that we are always called in request
 	 * completion order.
 	 */
-	GEM_BUG_ON(!list_is_first(&rq->link, &rq->timeline->requests));
+	GEM_BUG_ON(!list_is_first(&rq->link,
+				  &i915_request_timeline(rq)->requests));
 	rq->ring->head = rq->postfix;
 
 	/*
-	 * Walk through the active list, calling retire on each. This allows
-	 * objects to track their GPU activity and mark themselves as idle
-	 * when their *last* active request is completed (updating state
-	 * tracking lists for eviction, active references for GEM, etc).
-	 *
-	 * As the ->retire() may free the node, we decouple it first and
-	 * pass along the auxiliary information (to avoid dereferencing
-	 * the node after the callback).
-	 */
-	list_for_each_entry_safe(active, next, &rq->active_list, link) {
-		/*
-		 * In microbenchmarks or focusing upon time inside the kernel,
-		 * we may spend an inordinate amount of time simply handling
-		 * the retirement of requests and processing their callbacks.
-		 * Of which, this loop itself is particularly hot due to the
-		 * cache misses when jumping around the list of
-		 * i915_active_request.  So we try to keep this loop as
-		 * streamlined as possible and also prefetch the next
-		 * i915_active_request to try and hide the likely cache miss.
-		 */
-		prefetchw(next);
-
-		INIT_LIST_HEAD(&active->link);
-		RCU_INIT_POINTER(active->request, NULL);
-
-		active->retire(active, rq);
-	}
-
-	local_irq_disable();
-
-	/*
 	 * We only loosely track inflight requests across preemption,
 	 * and so we may find ourselves attempting to retire a _completed_
 	 * request that we have removed from the HW and put back on a run
 	 * queue.
 	 */
-	spin_lock(&rq->engine->active.lock);
-	list_del(&rq->sched.link);
-	spin_unlock(&rq->engine->active.lock);
+	remove_from_engine(rq);
 
-	spin_lock(&rq->lock);
+	spin_lock_irq(&rq->lock);
 	i915_request_mark_complete(rq);
 	if (!i915_request_signaled(rq))
 		dma_fence_signal_locked(&rq->fence);
 	if (test_bit(DMA_FENCE_FLAG_ENABLE_SIGNAL_BIT, &rq->fence.flags))
 		i915_request_cancel_breadcrumb(rq);
 	if (i915_request_has_waitboost(rq)) {
-		GEM_BUG_ON(!atomic_read(&rq->i915->gt_pm.rps.num_waiters));
-		atomic_dec(&rq->i915->gt_pm.rps.num_waiters);
+		GEM_BUG_ON(!atomic_read(&rq->engine->gt->rps.num_waiters));
+		atomic_dec(&rq->engine->gt->rps.num_waiters);
 	}
 	if (!test_bit(I915_FENCE_FLAG_ACTIVE, &rq->fence.flags)) {
 		set_bit(I915_FENCE_FLAG_ACTIVE, &rq->fence.flags);
 		__notify_execute_cb(rq);
 	}
 	GEM_BUG_ON(!list_empty(&rq->execute_cb));
-	spin_unlock(&rq->lock);
-
-	local_irq_enable();
+	spin_unlock_irq(&rq->lock);
 
 	remove_from_client(rq);
 	list_del(&rq->link);
@@ -297,7 +284,7 @@ static bool i915_request_retire(struct i915_request *rq)
 
 void i915_request_retire_upto(struct i915_request *rq)
 {
-	struct intel_timeline * const tl = rq->timeline;
+	struct intel_timeline * const tl = i915_request_timeline(rq);
 	struct i915_request *tmp;
 
 	GEM_TRACE("%s fence %llx:%lld, current %d\n",
@@ -305,7 +292,6 @@ void i915_request_retire_upto(struct i915_request *rq)
 		  rq->fence.context, rq->fence.seqno,
 		  hwsp_seqno(rq));
 
-	lockdep_assert_held(&tl->mutex);
 	GEM_BUG_ON(!i915_request_completed(rq));
 
 	do {
@@ -358,9 +344,10 @@ __i915_request_await_execution(struct i915_request *rq,
 	return 0;
 }
 
-void __i915_request_submit(struct i915_request *request)
+bool __i915_request_submit(struct i915_request *request)
 {
 	struct intel_engine_cs *engine = request->engine;
+	bool result = false;
 
 	GEM_TRACE("%s fence %llx:%lld, current %d\n",
 		  engine->name,
@@ -370,6 +357,25 @@ void __i915_request_submit(struct i915_request *request)
 	GEM_BUG_ON(!irqs_disabled());
 	lockdep_assert_held(&engine->active.lock);
 
+	/*
+	 * With the advent of preempt-to-busy, we frequently encounter
+	 * requests that we have unsubmitted from HW, but left running
+	 * until the next ack and so have completed in the meantime. On
+	 * resubmission of that completed request, we can skip
+	 * updating the payload, and execlists can even skip submitting
+	 * the request.
+	 *
+	 * We must remove the request from the caller's priority queue,
+	 * and the caller must only call us when the request is in their
+	 * priority queue, under the active.lock. This ensures that the
+	 * request has *not* yet been retired and we can safely move
+	 * the request into the engine->active.list where it will be
+	 * dropped upon retiring. (Otherwise if resubmit a *retired*
+	 * request, this would be a horrible use-after-free.)
+	 */
+	if (i915_request_completed(request))
+		goto xfer;
+
 	if (i915_gem_context_is_banned(request->gem_context))
 		i915_request_skip(request, -EIO);
 
@@ -393,13 +399,18 @@ void __i915_request_submit(struct i915_request *request)
 	    i915_sw_fence_signaled(&request->semaphore))
 		engine->saturated |= request->sched.semaphores;
 
-	/* We may be recursing from the signal callback of another i915 fence */
-	spin_lock_nested(&request->lock, SINGLE_DEPTH_NESTING);
+	engine->emit_fini_breadcrumb(request,
+				     request->ring->vaddr + request->postfix);
 
-	list_move_tail(&request->sched.link, &engine->active.requests);
+	trace_i915_request_execute(request);
+	engine->serial++;
+	result = true;
 
-	GEM_BUG_ON(test_bit(I915_FENCE_FLAG_ACTIVE, &request->fence.flags));
-	set_bit(I915_FENCE_FLAG_ACTIVE, &request->fence.flags);
+xfer:	/* We may be recursing from the signal callback of another i915 fence */
+	spin_lock_nested(&request->lock, SINGLE_DEPTH_NESTING);
+
+	if (!test_and_set_bit(I915_FENCE_FLAG_ACTIVE, &request->fence.flags))
+		list_move_tail(&request->sched.link, &engine->active.requests);
 
 	if (test_bit(DMA_FENCE_FLAG_ENABLE_SIGNAL_BIT, &request->fence.flags) &&
 	    !test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &request->fence.flags) &&
@@ -410,12 +421,7 @@ void __i915_request_submit(struct i915_request *request)
 
 	spin_unlock(&request->lock);
 
-	engine->emit_fini_breadcrumb(request,
-				     request->ring->vaddr + request->postfix);
-
-	engine->serial++;
-
-	trace_i915_request_execute(request);
+	return result;
 }
 
 void i915_request_submit(struct i915_request *request)
@@ -641,9 +647,12 @@ __i915_request_create(struct intel_context *ce, gfp_t gfp)
 	rq->gem_context = ce->gem_context;
 	rq->engine = ce->engine;
 	rq->ring = ce->ring;
-	rq->timeline = tl;
+	rq->execution_mask = ce->engine->mask;
+
+	rcu_assign_pointer(rq->timeline, tl);
 	rq->hwsp_seqno = tl->hwsp_seqno;
 	rq->hwsp_cacheline = tl->hwsp_cacheline;
+
 	rq->rcustate = get_state_synchronize_rcu(); /* acts as smp_mb() */
 
 	spin_lock_init(&rq->lock);
@@ -661,9 +670,7 @@ __i915_request_create(struct intel_context *ce, gfp_t gfp)
 	rq->batch = NULL;
 	rq->capture_list = NULL;
 	rq->flags = 0;
-	rq->execution_mask = ALL_ENGINES;
 
-	INIT_LIST_HEAD(&rq->active_list);
 	INIT_LIST_HEAD(&rq->execute_cb);
 
 	/*
@@ -702,7 +709,6 @@ err_unwind:
 	ce->ring->emit = rq->head;
 
 	/* Make sure we didn't add ourselves to external state before freeing */
-	GEM_BUG_ON(!list_empty(&rq->active_list));
 	GEM_BUG_ON(!list_empty(&rq->sched.signalers_list));
 	GEM_BUG_ON(!list_empty(&rq->sched.waiters_list));
 
@@ -747,16 +753,43 @@ err_unlock:
 static int
 i915_request_await_start(struct i915_request *rq, struct i915_request *signal)
 {
-	if (list_is_first(&signal->link, &signal->timeline->requests))
-		return 0;
+	struct intel_timeline *tl;
+	struct dma_fence *fence;
+	int err;
+
+	GEM_BUG_ON(i915_request_timeline(rq) ==
+		   rcu_access_pointer(signal->timeline));
 
-	signal = list_prev_entry(signal, link);
-	if (intel_timeline_sync_is_later(rq->timeline, &signal->fence))
+	rcu_read_lock();
+	tl = rcu_dereference(signal->timeline);
+	if (i915_request_started(signal) || !kref_get_unless_zero(&tl->kref))
+		tl = NULL;
+	rcu_read_unlock();
+	if (!tl) /* already started or maybe even completed */
 		return 0;
 
-	return i915_sw_fence_await_dma_fence(&rq->submit,
-					     &signal->fence, 0,
-					     I915_FENCE_GFP);
+	fence = ERR_PTR(-EBUSY);
+	if (mutex_trylock(&tl->mutex)) {
+		fence = NULL;
+		if (!i915_request_started(signal) &&
+		    !list_is_first(&signal->link, &tl->requests)) {
+			signal = list_prev_entry(signal, link);
+			fence = dma_fence_get(&signal->fence);
+		}
+		mutex_unlock(&tl->mutex);
+	}
+	intel_timeline_put(tl);
+	if (IS_ERR_OR_NULL(fence))
+		return PTR_ERR_OR_ZERO(fence);
+
+	err = 0;
+	if (intel_timeline_sync_is_later(i915_request_timeline(rq), fence))
+		err = i915_sw_fence_await_dma_fence(&rq->submit,
+						    fence, 0,
+						    I915_FENCE_GFP);
+	dma_fence_put(fence);
+
+	return err;
 }
 
 static intel_engine_mask_t
@@ -782,34 +815,33 @@ emit_semaphore_wait(struct i915_request *to,
 		    struct i915_request *from,
 		    gfp_t gfp)
 {
+	const int has_token = INTEL_GEN(to->i915) >= 12;
 	u32 hwsp_offset;
+	int len;
 	u32 *cs;
-	int err;
 
-	GEM_BUG_ON(!from->timeline->has_initial_breadcrumb);
 	GEM_BUG_ON(INTEL_GEN(to->i915) < 8);
 
 	/* Just emit the first semaphore we see as request space is limited. */
 	if (already_busywaiting(to) & from->engine->mask)
-		return i915_sw_fence_await_dma_fence(&to->submit,
-						     &from->fence, 0,
-						     I915_FENCE_GFP);
+		goto await_fence;
 
-	err = i915_request_await_start(to, from);
-	if (err < 0)
-		return err;
+	if (i915_request_await_start(to, from) < 0)
+		goto await_fence;
 
 	/* Only submit our spinner after the signaler is running! */
-	err = __i915_request_await_execution(to, from, NULL, gfp);
-	if (err)
-		return err;
+	if (__i915_request_await_execution(to, from, NULL, gfp))
+		goto await_fence;
 
 	/* We need to pin the signaler's HWSP until we are finished reading. */
-	err = intel_timeline_read_hwsp(from, to, &hwsp_offset);
-	if (err)
-		return err;
+	if (intel_timeline_read_hwsp(from, to, &hwsp_offset))
+		goto await_fence;
 
-	cs = intel_ring_begin(to, 4);
+	len = 4;
+	if (has_token)
+		len += 2;
+
+	cs = intel_ring_begin(to, len);
 	if (IS_ERR(cs))
 		return PTR_ERR(cs);
 
@@ -821,18 +853,28 @@ emit_semaphore_wait(struct i915_request *to,
 	 * (post-wrap) values than they were expecting (and so wait
 	 * forever).
 	 */
-	*cs++ = MI_SEMAPHORE_WAIT |
-		MI_SEMAPHORE_GLOBAL_GTT |
-		MI_SEMAPHORE_POLL |
-		MI_SEMAPHORE_SAD_GTE_SDD;
+	*cs++ = (MI_SEMAPHORE_WAIT |
+		 MI_SEMAPHORE_GLOBAL_GTT |
+		 MI_SEMAPHORE_POLL |
+		 MI_SEMAPHORE_SAD_GTE_SDD) +
+		has_token;
 	*cs++ = from->fence.seqno;
 	*cs++ = hwsp_offset;
 	*cs++ = 0;
+	if (has_token) {
+		*cs++ = 0;
+		*cs++ = MI_NOOP;
+	}
 
 	intel_ring_advance(to, cs);
 	to->sched.semaphores |= from->engine->mask;
 	to->sched.flags |= I915_SCHED_HAS_SEMAPHORE_CHAIN;
 	return 0;
+
+await_fence:
+	return i915_sw_fence_await_dma_fence(&to->submit,
+					     &from->fence, 0,
+					     I915_FENCE_GFP);
 }
 
 static int
@@ -916,21 +958,23 @@ i915_request_await_dma_fence(struct i915_request *rq, struct dma_fence *fence)
 
 		/* Squash repeated waits to the same timelines */
 		if (fence->context &&
-		    intel_timeline_sync_is_later(rq->timeline, fence))
+		    intel_timeline_sync_is_later(i915_request_timeline(rq),
+						 fence))
 			continue;
 
 		if (dma_fence_is_i915(fence))
 			ret = i915_request_await_request(rq, to_request(fence));
 		else
 			ret = i915_sw_fence_await_dma_fence(&rq->submit, fence,
-							    I915_FENCE_TIMEOUT,
+							    fence->context ? I915_FENCE_TIMEOUT : 0,
 							    I915_FENCE_GFP);
 		if (ret < 0)
 			return ret;
 
 		/* Record the latest fence used against each timeline */
 		if (fence->context)
-			intel_timeline_sync_set(rq->timeline, fence);
+			intel_timeline_sync_set(i915_request_timeline(rq),
+						fence);
 	} while (--nchild);
 
 	return 0;
@@ -1072,7 +1116,7 @@ void i915_request_skip(struct i915_request *rq, int error)
 static struct i915_request *
 __i915_request_add_to_timeline(struct i915_request *rq)
 {
-	struct intel_timeline *timeline = rq->timeline;
+	struct intel_timeline *timeline = i915_request_timeline(rq);
 	struct i915_request *prev;
 
 	/*
@@ -1095,8 +1139,8 @@ __i915_request_add_to_timeline(struct i915_request *rq)
 	 * precludes optimising to use semaphores serialisation of a single
 	 * timeline across engines.
 	 */
-	prev = rcu_dereference_protected(timeline->last_request.request,
-					 lockdep_is_held(&timeline->mutex));
+	prev = to_request(__i915_active_fence_set(&timeline->last_request,
+						  &rq->fence));
 	if (prev && !i915_request_completed(prev)) {
 		if (is_power_of_2(prev->engine->mask | rq->engine->mask))
 			i915_sw_fence_await_sw_fence(&rq->submit,
@@ -1121,7 +1165,6 @@ __i915_request_add_to_timeline(struct i915_request *rq)
 	 * us, the timeline will hold its seqno which is later than ours.
 	 */
 	GEM_BUG_ON(timeline->seqno != rq->fence.seqno);
-	__i915_active_request_set(&timeline->last_request, rq);
 
 	return prev;
 }
@@ -1185,7 +1228,7 @@ void __i915_request_queue(struct i915_request *rq,
 void i915_request_add(struct i915_request *rq)
 {
 	struct i915_sched_attr attr = rq->gem_context->sched;
-	struct intel_timeline * const tl = rq->timeline;
+	struct intel_timeline * const tl = i915_request_timeline(rq);
 	struct i915_request *prev;
 
 	lockdep_assert_held(&tl->mutex);
@@ -1240,7 +1283,9 @@ void i915_request_add(struct i915_request *rq)
 	 * work on behalf of others -- but instead we should benefit from
 	 * improved resource management. (Well, that's the theory at least.)
 	 */
-	if (prev && i915_request_completed(prev) && prev->timeline == tl)
+	if (prev &&
+	    i915_request_completed(prev) &&
+	    rcu_access_pointer(prev->timeline) == tl)
 		i915_request_retire_upto(prev);
 
 	mutex_unlock(&tl->mutex);
@@ -1403,7 +1448,7 @@ long i915_request_wait(struct i915_request *rq,
 	 * completion. That requires having a good predictor for the request
 	 * duration, which we currently lack.
 	 */
-	if (CONFIG_DRM_I915_SPIN_REQUEST &&
+	if (IS_ACTIVE(CONFIG_DRM_I915_SPIN_REQUEST) &&
 	    __i915_spin_request(rq, state, CONFIG_DRM_I915_SPIN_REQUEST)) {
 		dma_fence_signal(&rq->fence);
 		goto out;
@@ -1423,7 +1468,7 @@ long i915_request_wait(struct i915_request *rq,
 	 */
 	if (flags & I915_WAIT_PRIORITY) {
 		if (!i915_request_started(rq) && INTEL_GEN(rq->i915) >= 6)
-			gen6_rps_boost(rq);
+			intel_rps_boost(rq);
 		i915_schedule_bump_priority(rq, I915_PRIORITY_WAIT);
 	}
 
@@ -1449,6 +1494,7 @@ long i915_request_wait(struct i915_request *rq,
 			break;
 		}
 
+		intel_engine_flush_submission(rq->engine);
 		timeout = io_schedule_timeout(timeout);
 	}
 	__set_current_state(TASK_RUNNING);
@@ -1456,53 +1502,11 @@ long i915_request_wait(struct i915_request *rq,
 	dma_fence_remove_callback(&rq->fence, &wait.cb);
 
 out:
-	mutex_release(&rq->engine->gt->reset.mutex.dep_map, 0, _THIS_IP_);
+	mutex_release(&rq->engine->gt->reset.mutex.dep_map, _THIS_IP_);
 	trace_i915_request_wait_end(rq);
 	return timeout;
 }
 
-bool i915_retire_requests(struct drm_i915_private *i915)
-{
-	struct intel_gt_timelines *timelines = &i915->gt.timelines;
-	struct intel_timeline *tl, *tn;
-	unsigned long flags;
-	LIST_HEAD(free);
-
-	spin_lock_irqsave(&timelines->lock, flags);
-	list_for_each_entry_safe(tl, tn, &timelines->active_list, link) {
-		if (!mutex_trylock(&tl->mutex))
-			continue;
-
-		intel_timeline_get(tl);
-		GEM_BUG_ON(!tl->active_count);
-		tl->active_count++; /* pin the list element */
-		spin_unlock_irqrestore(&timelines->lock, flags);
-
-		retire_requests(tl);
-
-		spin_lock_irqsave(&timelines->lock, flags);
-
-		/* Resume iteration after dropping lock */
-		list_safe_reset_next(tl, tn, link);
-		if (!--tl->active_count)
-			list_del(&tl->link);
-
-		mutex_unlock(&tl->mutex);
-
-		/* Defer the final release to after the spinlock */
-		if (refcount_dec_and_test(&tl->kref.refcount)) {
-			GEM_BUG_ON(tl->active_count);
-			list_add(&tl->link, &free);
-		}
-	}
-	spin_unlock_irqrestore(&timelines->lock, flags);
-
-	list_for_each_entry_safe(tl, tn, &free, link)
-		__intel_timeline_free(&tl->kref);
-
-	return !list_empty(&timelines->active_list);
-}
-
 #if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
 #include "selftests/mock_request.c"
 #include "selftests/i915_request.c"
diff --git a/drivers/gpu/drm/i915/i915_request.h b/drivers/gpu/drm/i915/i915_request.h
index 8ac6e1226a56..96991d64759c 100644
--- a/drivers/gpu/drm/i915/i915_request.h
+++ b/drivers/gpu/drm/i915/i915_request.h
@@ -113,7 +113,7 @@ struct i915_request {
 	struct intel_engine_cs *engine;
 	struct intel_context *hw_context;
 	struct intel_ring *ring;
-	struct intel_timeline *timeline;
+	struct intel_timeline __rcu *timeline;
 	struct list_head signal_link;
 
 	/*
@@ -211,14 +211,14 @@ struct i915_request {
 	 * on the active_list (of their final request).
 	 */
 	struct i915_capture_list *capture_list;
-	struct list_head active_list;
 
 	/** Time at which this request was emitted, in jiffies. */
 	unsigned long emitted_jiffies;
 
 	unsigned long flags;
-#define I915_REQUEST_WAITBOOST BIT(0)
-#define I915_REQUEST_NOPREEMPT BIT(1)
+#define I915_REQUEST_WAITBOOST	BIT(0)
+#define I915_REQUEST_NOPREEMPT	BIT(1)
+#define I915_REQUEST_SENTINEL	BIT(2)
 
 	/** timeline->request entry for this request */
 	struct list_head link;
@@ -251,6 +251,7 @@ struct i915_request *__i915_request_commit(struct i915_request *request);
 void __i915_request_queue(struct i915_request *rq,
 			  const struct i915_sched_attr *attr);
 
+bool i915_request_retire(struct i915_request *rq);
 void i915_request_retire_upto(struct i915_request *rq);
 
 static inline struct i915_request *
@@ -292,7 +293,7 @@ int i915_request_await_execution(struct i915_request *rq,
 
 void i915_request_add(struct i915_request *rq);
 
-void __i915_request_submit(struct i915_request *request);
+bool __i915_request_submit(struct i915_request *request);
 void i915_request_submit(struct i915_request *request);
 
 void i915_request_skip(struct i915_request *request, int error);
@@ -309,10 +310,8 @@ long i915_request_wait(struct i915_request *rq,
 		       long timeout)
 	__attribute__((nonnull(1)));
 #define I915_WAIT_INTERRUPTIBLE	BIT(0)
-#define I915_WAIT_LOCKED	BIT(1) /* struct_mutex held, handle GPU reset */
-#define I915_WAIT_PRIORITY	BIT(2) /* small priority bump for the request */
-#define I915_WAIT_ALL		BIT(3) /* used by i915_gem_object_wait() */
-#define I915_WAIT_FOR_IDLE_BOOST BIT(4)
+#define I915_WAIT_PRIORITY	BIT(1) /* small priority bump for the request */
+#define I915_WAIT_ALL		BIT(2) /* used by i915_gem_object_wait() */
 
 static inline bool i915_request_signaled(const struct i915_request *rq)
 {
@@ -442,6 +441,29 @@ static inline bool i915_request_has_nopreempt(const struct i915_request *rq)
 	return unlikely(rq->flags & I915_REQUEST_NOPREEMPT);
 }
 
-bool i915_retire_requests(struct drm_i915_private *i915);
+static inline bool i915_request_has_sentinel(const struct i915_request *rq)
+{
+	return unlikely(rq->flags & I915_REQUEST_SENTINEL);
+}
+
+static inline struct intel_timeline *
+i915_request_timeline(struct i915_request *rq)
+{
+	/* Valid only while the request is being constructed (or retired). */
+	return rcu_dereference_protected(rq->timeline,
+					 lockdep_is_held(&rcu_access_pointer(rq->timeline)->mutex));
+}
+
+static inline struct intel_timeline *
+i915_request_active_timeline(struct i915_request *rq)
+{
+	/*
+	 * When in use during submission, we are protected by a guarantee that
+	 * the context/timeline is pinned and must remain pinned until after
+	 * this submission.
+	 */
+	return rcu_dereference_protected(rq->timeline,
+					 lockdep_is_held(&rq->engine->active.lock));
+}
 
 #endif /* I915_REQUEST_H */
diff --git a/drivers/gpu/drm/i915/i915_scatterlist.h b/drivers/gpu/drm/i915/i915_scatterlist.h
index 6617963df9ed..b7b59328cb76 100644
--- a/drivers/gpu/drm/i915/i915_scatterlist.h
+++ b/drivers/gpu/drm/i915/i915_scatterlist.h
@@ -67,15 +67,15 @@ static inline struct scatterlist *__sg_next(struct scatterlist *sg)
 }
 
 /**
- * __for_each_sgt_dma - iterate over the DMA addresses of the given sg_table
- * @__dmap:	DMA address (output)
+ * __for_each_sgt_daddr - iterate over the device addresses of the given sg_table
+ * @__dp:	Device address (output)
  * @__iter:	'struct sgt_iter' (iterator state, internal)
  * @__sgt:	sg_table to iterate over (input)
  * @__step:	step size
  */
-#define __for_each_sgt_dma(__dmap, __iter, __sgt, __step)		\
+#define __for_each_sgt_daddr(__dp, __iter, __sgt, __step)		\
 	for ((__iter) = __sgt_iter((__sgt)->sgl, true);			\
-	     ((__dmap) = (__iter).dma + (__iter).curr);			\
+	     ((__dp) = (__iter).dma + (__iter).curr), (__iter).sgp;	\
 	     (((__iter).curr += (__step)) >= (__iter).max) ?		\
 	     (__iter) = __sgt_iter(__sg_next((__iter).sgp), true), 0 : 0)
 
diff --git a/drivers/gpu/drm/i915/i915_scheduler.c b/drivers/gpu/drm/i915/i915_scheduler.c
index 7b84ebca2901..010d67f48ad9 100644
--- a/drivers/gpu/drm/i915/i915_scheduler.c
+++ b/drivers/gpu/drm/i915/i915_scheduler.c
@@ -177,21 +177,51 @@ static inline int rq_prio(const struct i915_request *rq)
 	return rq->sched.attr.priority | __NO_PREEMPTION;
 }
 
-static void kick_submission(struct intel_engine_cs *engine, int prio)
+static inline bool need_preempt(int prio, int active)
 {
-	const struct i915_request *inflight = *engine->execlists.active;
+	/*
+	 * Allow preemption of low -> normal -> high, but we do
+	 * not allow low priority tasks to preempt other low priority
+	 * tasks under the impression that latency for low priority
+	 * tasks does not matter (as much as background throughput),
+	 * so kiss.
+	 */
+	return prio >= max(I915_PRIORITY_NORMAL, active);
+}
+
+static void kick_submission(struct intel_engine_cs *engine,
+			    const struct i915_request *rq,
+			    int prio)
+{
+	const struct i915_request *inflight;
 
 	/*
-	 * If we are already the currently executing context, don't
-	 * bother evaluating if we should preempt ourselves, or if
-	 * we expect nothing to change as a result of running the
-	 * tasklet, i.e. we have not change the priority queue
-	 * sufficiently to oust the running context.
+	 * We only need to kick the tasklet once for the high priority
+	 * new context we add into the queue.
 	 */
-	if (!inflight || !i915_scheduler_need_preempt(prio, rq_prio(inflight)))
+	if (prio <= engine->execlists.queue_priority_hint)
 		return;
 
-	tasklet_hi_schedule(&engine->execlists.tasklet);
+	rcu_read_lock();
+
+	/* Nothing currently active? We're overdue for a submission! */
+	inflight = execlists_active(&engine->execlists);
+	if (!inflight)
+		goto unlock;
+
+	/*
+	 * If we are already the currently executing context, don't
+	 * bother evaluating if we should preempt ourselves.
+	 */
+	if (inflight->hw_context == rq->hw_context)
+		goto unlock;
+
+	engine->execlists.queue_priority_hint = prio;
+	if (need_preempt(prio, rq_prio(inflight)))
+		tasklet_hi_schedule(&engine->execlists.tasklet);
+
+unlock:
+	rcu_read_unlock();
 }
 
 static void __i915_schedule(struct i915_sched_node *node,
@@ -317,13 +347,8 @@ static void __i915_schedule(struct i915_sched_node *node,
 			list_move_tail(&node->link, cache.priolist);
 		}
 
-		if (prio <= engine->execlists.queue_priority_hint)
-			continue;
-
-		engine->execlists.queue_priority_hint = prio;
-
 		/* Defer (tasklet) submission until after all of our updates. */
-		kick_submission(engine, prio);
+		kick_submission(engine, node_to_request(node), prio);
 	}
 
 	spin_unlock(&engine->active.lock);
diff --git a/drivers/gpu/drm/i915/i915_scheduler.h b/drivers/gpu/drm/i915/i915_scheduler.h
index 7eefccff39bf..07d243acf553 100644
--- a/drivers/gpu/drm/i915/i915_scheduler.h
+++ b/drivers/gpu/drm/i915/i915_scheduler.h
@@ -52,22 +52,4 @@ static inline void i915_priolist_free(struct i915_priolist *p)
 		__i915_priolist_free(p);
 }
 
-static inline bool i915_scheduler_need_preempt(int prio, int active)
-{
-	/*
-	 * Allow preemption of low -> normal -> high, but we do
-	 * not allow low priority tasks to preempt other low priority
-	 * tasks under the impression that latency for low priority
-	 * tasks does not matter (as much as background throughput),
-	 * so kiss.
-	 *
-	 * More naturally we would write
-	 *	prio >= max(0, last);
-	 * except that we wish to prevent triggering preemption at the same
-	 * priority level: the task that is running should remain running
-	 * to preserve FIFO ordering of dependencies.
-	 */
-	return prio > max(I915_PRIORITY_NORMAL - 1, active);
-}
-
 #endif /* _I915_SCHEDULER_H_ */
diff --git a/drivers/gpu/drm/i915/i915_scheduler_types.h b/drivers/gpu/drm/i915/i915_scheduler_types.h
index aad81acba9dc..d18e70550054 100644
--- a/drivers/gpu/drm/i915/i915_scheduler_types.h
+++ b/drivers/gpu/drm/i915/i915_scheduler_types.h
@@ -49,6 +49,15 @@ struct i915_sched_attr {
  * DAG of each request, we are able to insert it into a sorted queue when it
  * is ready, and are able to reorder its portion of the graph to accommodate
  * dynamic priority changes.
+ *
+ * Ok, there is now one active element to the "scheduler" in the backends.
+ * We let a new context run for a small amount of time before re-evaluating
+ * the run order. As we re-evaluate, we maintain the strict ordering of
+ * dependencies, but attempt to rotate the active contexts (the current context
+ * is put to the back of its priority queue, then reshuffling its dependents).
+ * This provides minimal timeslicing and prevents a userspace hog (e.g.
+ * something waiting on a user semaphore [VkEvent]) from denying service to
+ * others.
  */
 struct i915_sched_node {
 	struct list_head signalers_list; /* those before us, we depend upon */
diff --git a/drivers/gpu/drm/i915/i915_suspend.c b/drivers/gpu/drm/i915/i915_suspend.c
index 8508a01ad8b9..8812cdd9007f 100644
--- a/drivers/gpu/drm/i915/i915_suspend.c
+++ b/drivers/gpu/drm/i915/i915_suspend.c
@@ -28,6 +28,7 @@
 
 #include "display/intel_fbc.h"
 #include "display/intel_gmbus.h"
+#include "display/intel_vga.h"
 
 #include "i915_drv.h"
 #include "i915_reg.h"
@@ -57,7 +58,7 @@ static void i915_restore_display(struct drm_i915_private *dev_priv)
 	if (HAS_FBC(dev_priv) && INTEL_GEN(dev_priv) <= 4 && !IS_G4X(dev_priv))
 		I915_WRITE(FBC_CONTROL, dev_priv->regfile.saveFBC_CONTROL);
 
-	i915_redisable_vga(dev_priv);
+	intel_vga_redisable(dev_priv);
 }
 
 int i915_save_state(struct drm_i915_private *dev_priv)
@@ -65,8 +66,6 @@ int i915_save_state(struct drm_i915_private *dev_priv)
 	struct pci_dev *pdev = dev_priv->drm.pdev;
 	int i;
 
-	mutex_lock(&dev_priv->drm.struct_mutex);
-
 	i915_save_display(dev_priv);
 
 	if (IS_GEN(dev_priv, 4))
@@ -100,8 +99,6 @@ int i915_save_state(struct drm_i915_private *dev_priv)
 			dev_priv->regfile.saveSWF3[i] = I915_READ(SWF3(i));
 	}
 
-	mutex_unlock(&dev_priv->drm.struct_mutex);
-
 	return 0;
 }
 
@@ -110,8 +107,6 @@ int i915_restore_state(struct drm_i915_private *dev_priv)
 	struct pci_dev *pdev = dev_priv->drm.pdev;
 	int i;
 
-	mutex_lock(&dev_priv->drm.struct_mutex);
-
 	if (IS_GEN(dev_priv, 4))
 		pci_write_config_word(pdev, GCDGMBUS,
 				      dev_priv->regfile.saveGCDGMBUS);
@@ -145,8 +140,6 @@ int i915_restore_state(struct drm_i915_private *dev_priv)
 			I915_WRITE(SWF3(i), dev_priv->regfile.saveSWF3[i]);
 	}
 
-	mutex_unlock(&dev_priv->drm.struct_mutex);
-
 	intel_gmbus_reset(dev_priv);
 
 	return 0;
diff --git a/drivers/gpu/drm/i915/i915_switcheroo.c b/drivers/gpu/drm/i915/i915_switcheroo.c
new file mode 100644
index 000000000000..39c79e1c5b52
--- /dev/null
+++ b/drivers/gpu/drm/i915/i915_switcheroo.c
@@ -0,0 +1,67 @@
+// SPDX-License-Identifier: MIT
+/*
+ * Copyright © 2019 Intel Corporation
+ */
+
+#include <linux/vga_switcheroo.h>
+
+#include "i915_drv.h"
+#include "i915_switcheroo.h"
+
+static void i915_switcheroo_set_state(struct pci_dev *pdev,
+				      enum vga_switcheroo_state state)
+{
+	struct drm_i915_private *i915 = pdev_to_i915(pdev);
+	pm_message_t pmm = { .event = PM_EVENT_SUSPEND };
+
+	if (!i915) {
+		dev_err(&pdev->dev, "DRM not initialized, aborting switch.\n");
+		return;
+	}
+
+	if (state == VGA_SWITCHEROO_ON) {
+		pr_info("switched on\n");
+		i915->drm.switch_power_state = DRM_SWITCH_POWER_CHANGING;
+		/* i915 resume handler doesn't set to D0 */
+		pci_set_power_state(pdev, PCI_D0);
+		i915_resume_switcheroo(i915);
+		i915->drm.switch_power_state = DRM_SWITCH_POWER_ON;
+	} else {
+		pr_info("switched off\n");
+		i915->drm.switch_power_state = DRM_SWITCH_POWER_CHANGING;
+		i915_suspend_switcheroo(i915, pmm);
+		i915->drm.switch_power_state = DRM_SWITCH_POWER_OFF;
+	}
+}
+
+static bool i915_switcheroo_can_switch(struct pci_dev *pdev)
+{
+	struct drm_i915_private *i915 = pdev_to_i915(pdev);
+
+	/*
+	 * FIXME: open_count is protected by drm_global_mutex but that would lead to
+	 * locking inversion with the driver load path. And the access here is
+	 * completely racy anyway. So don't bother with locking for now.
+	 */
+	return i915 && i915->drm.open_count == 0;
+}
+
+static const struct vga_switcheroo_client_ops i915_switcheroo_ops = {
+	.set_gpu_state = i915_switcheroo_set_state,
+	.reprobe = NULL,
+	.can_switch = i915_switcheroo_can_switch,
+};
+
+int i915_switcheroo_register(struct drm_i915_private *i915)
+{
+	struct pci_dev *pdev = i915->drm.pdev;
+
+	return vga_switcheroo_register_client(pdev, &i915_switcheroo_ops, false);
+}
+
+void i915_switcheroo_unregister(struct drm_i915_private *i915)
+{
+	struct pci_dev *pdev = i915->drm.pdev;
+
+	vga_switcheroo_unregister_client(pdev);
+}
diff --git a/drivers/gpu/drm/i915/i915_switcheroo.h b/drivers/gpu/drm/i915/i915_switcheroo.h
new file mode 100644
index 000000000000..59b6c1e07d75
--- /dev/null
+++ b/drivers/gpu/drm/i915/i915_switcheroo.h
@@ -0,0 +1,14 @@
+/* SPDX-License-Identifier: MIT */
+/*
+ * Copyright © 2019 Intel Corporation
+ */
+
+#ifndef __I915_SWITCHEROO__
+#define __I915_SWITCHEROO__
+
+struct drm_i915_private;
+
+int i915_switcheroo_register(struct drm_i915_private *i915);
+void i915_switcheroo_unregister(struct drm_i915_private *i915);
+
+#endif /* __I915_SWITCHEROO__ */
diff --git a/drivers/gpu/drm/i915/i915_sysfs.c b/drivers/gpu/drm/i915/i915_sysfs.c
index d8a3b180c084..65476909d1bf 100644
--- a/drivers/gpu/drm/i915/i915_sysfs.c
+++ b/drivers/gpu/drm/i915/i915_sysfs.c
@@ -30,6 +30,9 @@
 #include <linux/stat.h>
 #include <linux/sysfs.h>
 
+#include "gt/intel_rc6.h"
+#include "gt/intel_rps.h"
+
 #include "i915_drv.h"
 #include "i915_sysfs.h"
 #include "intel_pm.h"
@@ -49,7 +52,7 @@ static u32 calc_residency(struct drm_i915_private *dev_priv,
 	u64 res = 0;
 
 	with_intel_runtime_pm(&dev_priv->runtime_pm, wakeref)
-		res = intel_rc6_residency_us(dev_priv, reg);
+		res = intel_rc6_residency_us(&dev_priv->gt.rc6, reg);
 
 	return DIV_ROUND_CLOSEST_ULL(res, 1000);
 }
@@ -142,12 +145,12 @@ static const struct attribute_group media_rc6_attr_group = {
 };
 #endif
 
-static int l3_access_valid(struct drm_i915_private *dev_priv, loff_t offset)
+static int l3_access_valid(struct drm_i915_private *i915, loff_t offset)
 {
-	if (!HAS_L3_DPF(dev_priv))
+	if (!HAS_L3_DPF(i915))
 		return -EPERM;
 
-	if (offset % 4 != 0)
+	if (!IS_ALIGNED(offset, sizeof(u32)))
 		return -EINVAL;
 
 	if (offset >= GEN7_L3LOG_SIZE)
@@ -162,31 +165,24 @@ i915_l3_read(struct file *filp, struct kobject *kobj,
 	     loff_t offset, size_t count)
 {
 	struct device *kdev = kobj_to_dev(kobj);
-	struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev);
-	struct drm_device *dev = &dev_priv->drm;
+	struct drm_i915_private *i915 = kdev_minor_to_i915(kdev);
 	int slice = (int)(uintptr_t)attr->private;
 	int ret;
 
-	count = round_down(count, 4);
-
-	ret = l3_access_valid(dev_priv, offset);
+	ret = l3_access_valid(i915, offset);
 	if (ret)
 		return ret;
 
+	count = round_down(count, sizeof(u32));
 	count = min_t(size_t, GEN7_L3LOG_SIZE - offset, count);
+	memset(buf, 0, count);
 
-	ret = i915_mutex_lock_interruptible(dev);
-	if (ret)
-		return ret;
-
-	if (dev_priv->l3_parity.remap_info[slice])
+	spin_lock(&i915->gem.contexts.lock);
+	if (i915->l3_parity.remap_info[slice])
 		memcpy(buf,
-		       dev_priv->l3_parity.remap_info[slice] + (offset/4),
+		       i915->l3_parity.remap_info[slice] + offset / sizeof(u32),
 		       count);
-	else
-		memset(buf, 0, count);
-
-	mutex_unlock(&dev->struct_mutex);
+	spin_unlock(&i915->gem.contexts.lock);
 
 	return count;
 }
@@ -197,46 +193,49 @@ i915_l3_write(struct file *filp, struct kobject *kobj,
 	      loff_t offset, size_t count)
 {
 	struct device *kdev = kobj_to_dev(kobj);
-	struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev);
-	struct drm_device *dev = &dev_priv->drm;
-	struct i915_gem_context *ctx;
+	struct drm_i915_private *i915 = kdev_minor_to_i915(kdev);
 	int slice = (int)(uintptr_t)attr->private;
-	u32 **remap_info;
+	u32 *remap_info, *freeme = NULL;
+	struct i915_gem_context *ctx;
 	int ret;
 
-	ret = l3_access_valid(dev_priv, offset);
+	ret = l3_access_valid(i915, offset);
 	if (ret)
 		return ret;
 
-	ret = i915_mutex_lock_interruptible(dev);
-	if (ret)
-		return ret;
+	if (count < sizeof(u32))
+		return -EINVAL;
 
-	remap_info = &dev_priv->l3_parity.remap_info[slice];
-	if (!*remap_info) {
-		*remap_info = kzalloc(GEN7_L3LOG_SIZE, GFP_KERNEL);
-		if (!*remap_info) {
-			ret = -ENOMEM;
-			goto out;
-		}
+	remap_info = kzalloc(GEN7_L3LOG_SIZE, GFP_KERNEL);
+	if (!remap_info)
+		return -ENOMEM;
+
+	spin_lock(&i915->gem.contexts.lock);
+
+	if (i915->l3_parity.remap_info[slice]) {
+		freeme = remap_info;
+		remap_info = i915->l3_parity.remap_info[slice];
+	} else {
+		i915->l3_parity.remap_info[slice] = remap_info;
 	}
 
-	/* TODO: Ideally we really want a GPU reset here to make sure errors
-	 * aren't propagated. Since I cannot find a stable way to reset the GPU
-	 * at this point it is left as a TODO.
-	*/
-	memcpy(*remap_info + (offset/4), buf, count);
+	count = round_down(count, sizeof(u32));
+	memcpy(remap_info + offset / sizeof(u32), buf, count);
 
 	/* NB: We defer the remapping until we switch to the context */
-	list_for_each_entry(ctx, &dev_priv->contexts.list, link)
-		ctx->remap_slice |= (1<<slice);
+	list_for_each_entry(ctx, &i915->gem.contexts.list, link)
+		ctx->remap_slice |= BIT(slice);
 
-	ret = count;
+	spin_unlock(&i915->gem.contexts.lock);
+	kfree(freeme);
 
-out:
-	mutex_unlock(&dev->struct_mutex);
+	/*
+	 * TODO: Ideally we really want a GPU reset here to make sure errors
+	 * aren't propagated. Since I cannot find a stable way to reset the GPU
+	 * at this point it is left as a TODO.
+	*/
 
-	return ret;
+	return count;
 }
 
 static const struct bin_attribute dpf_attrs = {
@@ -261,6 +260,7 @@ static ssize_t gt_act_freq_mhz_show(struct device *kdev,
 				    struct device_attribute *attr, char *buf)
 {
 	struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev);
+	struct intel_rps *rps = &dev_priv->gt.rps;
 	intel_wakeref_t wakeref;
 	u32 freq;
 
@@ -273,31 +273,31 @@ static ssize_t gt_act_freq_mhz_show(struct device *kdev,
 
 		freq = (freq >> 8) & 0xff;
 	} else {
-		freq = intel_get_cagf(dev_priv, I915_READ(GEN6_RPSTAT1));
+		freq = intel_get_cagf(rps, I915_READ(GEN6_RPSTAT1));
 	}
 
 	intel_runtime_pm_put(&dev_priv->runtime_pm, wakeref);
 
-	return snprintf(buf, PAGE_SIZE, "%d\n", intel_gpu_freq(dev_priv, freq));
+	return snprintf(buf, PAGE_SIZE, "%d\n", intel_gpu_freq(rps, freq));
 }
 
 static ssize_t gt_cur_freq_mhz_show(struct device *kdev,
 				    struct device_attribute *attr, char *buf)
 {
 	struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev);
+	struct intel_rps *rps = &dev_priv->gt.rps;
 
 	return snprintf(buf, PAGE_SIZE, "%d\n",
-			intel_gpu_freq(dev_priv,
-				       dev_priv->gt_pm.rps.cur_freq));
+			intel_gpu_freq(rps, rps->cur_freq));
 }
 
 static ssize_t gt_boost_freq_mhz_show(struct device *kdev, struct device_attribute *attr, char *buf)
 {
 	struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev);
+	struct intel_rps *rps = &dev_priv->gt.rps;
 
 	return snprintf(buf, PAGE_SIZE, "%d\n",
-			intel_gpu_freq(dev_priv,
-				       dev_priv->gt_pm.rps.boost_freq));
+			intel_gpu_freq(rps, rps->boost_freq));
 }
 
 static ssize_t gt_boost_freq_mhz_store(struct device *kdev,
@@ -305,7 +305,7 @@ static ssize_t gt_boost_freq_mhz_store(struct device *kdev,
 				       const char *buf, size_t count)
 {
 	struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev);
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
+	struct intel_rps *rps = &dev_priv->gt.rps;
 	bool boost = false;
 	ssize_t ret;
 	u32 val;
@@ -315,7 +315,7 @@ static ssize_t gt_boost_freq_mhz_store(struct device *kdev,
 		return ret;
 
 	/* Validate against (static) hardware limits */
-	val = intel_freq_opcode(dev_priv, val);
+	val = intel_freq_opcode(rps, val);
 	if (val < rps->min_freq || val > rps->max_freq)
 		return -EINVAL;
 
@@ -335,19 +335,19 @@ static ssize_t vlv_rpe_freq_mhz_show(struct device *kdev,
 				     struct device_attribute *attr, char *buf)
 {
 	struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev);
+	struct intel_rps *rps = &dev_priv->gt.rps;
 
 	return snprintf(buf, PAGE_SIZE, "%d\n",
-			intel_gpu_freq(dev_priv,
-				       dev_priv->gt_pm.rps.efficient_freq));
+			intel_gpu_freq(rps, rps->efficient_freq));
 }
 
 static ssize_t gt_max_freq_mhz_show(struct device *kdev, struct device_attribute *attr, char *buf)
 {
 	struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev);
+	struct intel_rps *rps = &dev_priv->gt.rps;
 
 	return snprintf(buf, PAGE_SIZE, "%d\n",
-			intel_gpu_freq(dev_priv,
-				       dev_priv->gt_pm.rps.max_freq_softlimit));
+			intel_gpu_freq(rps, rps->max_freq_softlimit));
 }
 
 static ssize_t gt_max_freq_mhz_store(struct device *kdev,
@@ -355,19 +355,17 @@ static ssize_t gt_max_freq_mhz_store(struct device *kdev,
 				     const char *buf, size_t count)
 {
 	struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev);
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
-	intel_wakeref_t wakeref;
-	u32 val;
+	struct intel_rps *rps = &dev_priv->gt.rps;
 	ssize_t ret;
+	u32 val;
 
 	ret = kstrtou32(buf, 0, &val);
 	if (ret)
 		return ret;
 
-	wakeref = intel_runtime_pm_get(&dev_priv->runtime_pm);
 	mutex_lock(&rps->lock);
 
-	val = intel_freq_opcode(dev_priv, val);
+	val = intel_freq_opcode(rps, val);
 	if (val < rps->min_freq ||
 	    val > rps->max_freq ||
 	    val < rps->min_freq_softlimit) {
@@ -377,7 +375,7 @@ static ssize_t gt_max_freq_mhz_store(struct device *kdev,
 
 	if (val > rps->rp0_freq)
 		DRM_DEBUG("User requested overclocking to %d\n",
-			  intel_gpu_freq(dev_priv, val));
+			  intel_gpu_freq(rps, val));
 
 	rps->max_freq_softlimit = val;
 
@@ -385,14 +383,15 @@ static ssize_t gt_max_freq_mhz_store(struct device *kdev,
 		      rps->min_freq_softlimit,
 		      rps->max_freq_softlimit);
 
-	/* We still need *_set_rps to process the new max_delay and
+	/*
+	 * We still need *_set_rps to process the new max_delay and
 	 * update the interrupt limits and PMINTRMSK even though
-	 * frequency request may be unchanged. */
-	ret = intel_set_rps(dev_priv, val);
+	 * frequency request may be unchanged.
+	 */
+	intel_rps_set(rps, val);
 
 unlock:
 	mutex_unlock(&rps->lock);
-	intel_runtime_pm_put(&dev_priv->runtime_pm, wakeref);
 
 	return ret ?: count;
 }
@@ -400,10 +399,10 @@ unlock:
 static ssize_t gt_min_freq_mhz_show(struct device *kdev, struct device_attribute *attr, char *buf)
 {
 	struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev);
+	struct intel_rps *rps = &dev_priv->gt.rps;
 
 	return snprintf(buf, PAGE_SIZE, "%d\n",
-			intel_gpu_freq(dev_priv,
-				       dev_priv->gt_pm.rps.min_freq_softlimit));
+			intel_gpu_freq(rps, rps->min_freq_softlimit));
 }
 
 static ssize_t gt_min_freq_mhz_store(struct device *kdev,
@@ -411,19 +410,17 @@ static ssize_t gt_min_freq_mhz_store(struct device *kdev,
 				     const char *buf, size_t count)
 {
 	struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev);
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
-	intel_wakeref_t wakeref;
-	u32 val;
+	struct intel_rps *rps = &dev_priv->gt.rps;
 	ssize_t ret;
+	u32 val;
 
 	ret = kstrtou32(buf, 0, &val);
 	if (ret)
 		return ret;
 
-	wakeref = intel_runtime_pm_get(&dev_priv->runtime_pm);
 	mutex_lock(&rps->lock);
 
-	val = intel_freq_opcode(dev_priv, val);
+	val = intel_freq_opcode(rps, val);
 	if (val < rps->min_freq ||
 	    val > rps->max_freq ||
 	    val > rps->max_freq_softlimit) {
@@ -437,14 +434,15 @@ static ssize_t gt_min_freq_mhz_store(struct device *kdev,
 		      rps->min_freq_softlimit,
 		      rps->max_freq_softlimit);
 
-	/* We still need *_set_rps to process the new min_delay and
+	/*
+	 * We still need *_set_rps to process the new min_delay and
 	 * update the interrupt limits and PMINTRMSK even though
-	 * frequency request may be unchanged. */
-	ret = intel_set_rps(dev_priv, val);
+	 * frequency request may be unchanged.
+	 */
+	intel_rps_set(rps, val);
 
 unlock:
 	mutex_unlock(&rps->lock);
-	intel_runtime_pm_put(&dev_priv->runtime_pm, wakeref);
 
 	return ret ?: count;
 }
@@ -466,15 +464,15 @@ static DEVICE_ATTR(gt_RPn_freq_mhz, S_IRUGO, gt_rp_mhz_show, NULL);
 static ssize_t gt_rp_mhz_show(struct device *kdev, struct device_attribute *attr, char *buf)
 {
 	struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev);
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
+	struct intel_rps *rps = &dev_priv->gt.rps;
 	u32 val;
 
 	if (attr == &dev_attr_gt_RP0_freq_mhz)
-		val = intel_gpu_freq(dev_priv, rps->rp0_freq);
+		val = intel_gpu_freq(rps, rps->rp0_freq);
 	else if (attr == &dev_attr_gt_RP1_freq_mhz)
-		val = intel_gpu_freq(dev_priv, rps->rp1_freq);
+		val = intel_gpu_freq(rps, rps->rp1_freq);
 	else if (attr == &dev_attr_gt_RPn_freq_mhz)
-		val = intel_gpu_freq(dev_priv, rps->min_freq);
+		val = intel_gpu_freq(rps, rps->min_freq);
 	else
 		BUG();
 
diff --git a/drivers/gpu/drm/i915/i915_trace.h b/drivers/gpu/drm/i915/i915_trace.h
index 24f2944da09d..7ef7a1e1664c 100644
--- a/drivers/gpu/drm/i915/i915_trace.h
+++ b/drivers/gpu/drm/i915/i915_trace.h
@@ -665,7 +665,6 @@ TRACE_EVENT(i915_request_queue,
 
 	    TP_STRUCT__entry(
 			     __field(u32, dev)
-			     __field(u32, hw_id)
 			     __field(u64, ctx)
 			     __field(u16, class)
 			     __field(u16, instance)
@@ -675,7 +674,6 @@ TRACE_EVENT(i915_request_queue,
 
 	    TP_fast_assign(
 			   __entry->dev = rq->i915->drm.primary->index;
-			   __entry->hw_id = rq->gem_context->hw_id;
 			   __entry->class = rq->engine->uabi_class;
 			   __entry->instance = rq->engine->uabi_instance;
 			   __entry->ctx = rq->fence.context;
@@ -683,10 +681,9 @@ TRACE_EVENT(i915_request_queue,
 			   __entry->flags = flags;
 			   ),
 
-	    TP_printk("dev=%u, engine=%u:%u, hw_id=%u, ctx=%llu, seqno=%u, flags=0x%x",
+	    TP_printk("dev=%u, engine=%u:%u, ctx=%llu, seqno=%u, flags=0x%x",
 		      __entry->dev, __entry->class, __entry->instance,
-		      __entry->hw_id, __entry->ctx, __entry->seqno,
-		      __entry->flags)
+		      __entry->ctx, __entry->seqno, __entry->flags)
 );
 
 DECLARE_EVENT_CLASS(i915_request,
@@ -695,7 +692,6 @@ DECLARE_EVENT_CLASS(i915_request,
 
 	    TP_STRUCT__entry(
 			     __field(u32, dev)
-			     __field(u32, hw_id)
 			     __field(u64, ctx)
 			     __field(u16, class)
 			     __field(u16, instance)
@@ -704,16 +700,15 @@ DECLARE_EVENT_CLASS(i915_request,
 
 	    TP_fast_assign(
 			   __entry->dev = rq->i915->drm.primary->index;
-			   __entry->hw_id = rq->gem_context->hw_id;
 			   __entry->class = rq->engine->uabi_class;
 			   __entry->instance = rq->engine->uabi_instance;
 			   __entry->ctx = rq->fence.context;
 			   __entry->seqno = rq->fence.seqno;
 			   ),
 
-	    TP_printk("dev=%u, engine=%u:%u, hw_id=%u, ctx=%llu, seqno=%u",
+	    TP_printk("dev=%u, engine=%u:%u, ctx=%llu, seqno=%u",
 		      __entry->dev, __entry->class, __entry->instance,
-		      __entry->hw_id, __entry->ctx, __entry->seqno)
+		      __entry->ctx, __entry->seqno)
 );
 
 DEFINE_EVENT(i915_request, i915_request_add,
@@ -738,7 +733,6 @@ TRACE_EVENT(i915_request_in,
 
 	    TP_STRUCT__entry(
 			     __field(u32, dev)
-			     __field(u32, hw_id)
 			     __field(u64, ctx)
 			     __field(u16, class)
 			     __field(u16, instance)
@@ -749,7 +743,6 @@ TRACE_EVENT(i915_request_in,
 
 	    TP_fast_assign(
 			   __entry->dev = rq->i915->drm.primary->index;
-			   __entry->hw_id = rq->gem_context->hw_id;
 			   __entry->class = rq->engine->uabi_class;
 			   __entry->instance = rq->engine->uabi_instance;
 			   __entry->ctx = rq->fence.context;
@@ -758,9 +751,9 @@ TRACE_EVENT(i915_request_in,
 			   __entry->port = port;
 			   ),
 
-	    TP_printk("dev=%u, engine=%u:%u, hw_id=%u, ctx=%llu, seqno=%u, prio=%u, port=%u",
+	    TP_printk("dev=%u, engine=%u:%u, ctx=%llu, seqno=%u, prio=%u, port=%u",
 		      __entry->dev, __entry->class, __entry->instance,
-		      __entry->hw_id, __entry->ctx, __entry->seqno,
+		      __entry->ctx, __entry->seqno,
 		      __entry->prio, __entry->port)
 );
 
@@ -770,7 +763,6 @@ TRACE_EVENT(i915_request_out,
 
 	    TP_STRUCT__entry(
 			     __field(u32, dev)
-			     __field(u32, hw_id)
 			     __field(u64, ctx)
 			     __field(u16, class)
 			     __field(u16, instance)
@@ -780,7 +772,6 @@ TRACE_EVENT(i915_request_out,
 
 	    TP_fast_assign(
 			   __entry->dev = rq->i915->drm.primary->index;
-			   __entry->hw_id = rq->gem_context->hw_id;
 			   __entry->class = rq->engine->uabi_class;
 			   __entry->instance = rq->engine->uabi_instance;
 			   __entry->ctx = rq->fence.context;
@@ -788,10 +779,9 @@ TRACE_EVENT(i915_request_out,
 			   __entry->completed = i915_request_completed(rq);
 			   ),
 
-		    TP_printk("dev=%u, engine=%u:%u, hw_id=%u, ctx=%llu, seqno=%u, completed?=%u",
+		    TP_printk("dev=%u, engine=%u:%u, ctx=%llu, seqno=%u, completed?=%u",
 			      __entry->dev, __entry->class, __entry->instance,
-			      __entry->hw_id, __entry->ctx, __entry->seqno,
-			      __entry->completed)
+			      __entry->ctx, __entry->seqno, __entry->completed)
 );
 
 #else
@@ -829,7 +819,6 @@ TRACE_EVENT(i915_request_wait_begin,
 
 	    TP_STRUCT__entry(
 			     __field(u32, dev)
-			     __field(u32, hw_id)
 			     __field(u64, ctx)
 			     __field(u16, class)
 			     __field(u16, instance)
@@ -845,7 +834,6 @@ TRACE_EVENT(i915_request_wait_begin,
 	     */
 	    TP_fast_assign(
 			   __entry->dev = rq->i915->drm.primary->index;
-			   __entry->hw_id = rq->gem_context->hw_id;
 			   __entry->class = rq->engine->uabi_class;
 			   __entry->instance = rq->engine->uabi_instance;
 			   __entry->ctx = rq->fence.context;
@@ -853,9 +841,9 @@ TRACE_EVENT(i915_request_wait_begin,
 			   __entry->flags = flags;
 			   ),
 
-	    TP_printk("dev=%u, engine=%u:%u, hw_id=%u, ctx=%llu, seqno=%u, flags=0x%x",
+	    TP_printk("dev=%u, engine=%u:%u, ctx=%llu, seqno=%u, flags=0x%x",
 		      __entry->dev, __entry->class, __entry->instance,
-		      __entry->hw_id, __entry->ctx, __entry->seqno,
+		      __entry->ctx, __entry->seqno,
 		      __entry->flags)
 );
 
@@ -958,19 +946,17 @@ DECLARE_EVENT_CLASS(i915_context,
 	TP_STRUCT__entry(
 			__field(u32, dev)
 			__field(struct i915_gem_context *, ctx)
-			__field(u32, hw_id)
 			__field(struct i915_address_space *, vm)
 	),
 
 	TP_fast_assign(
 			__entry->dev = ctx->i915->drm.primary->index;
 			__entry->ctx = ctx;
-			__entry->hw_id = ctx->hw_id;
-			__entry->vm = ctx->vm;
+			__entry->vm = rcu_access_pointer(ctx->vm);
 	),
 
-	TP_printk("dev=%u, ctx=%p, ctx_vm=%p, hw_id=%u",
-		  __entry->dev, __entry->ctx, __entry->vm, __entry->hw_id)
+	TP_printk("dev=%u, ctx=%p, ctx_vm=%p",
+		  __entry->dev, __entry->ctx, __entry->vm)
 )
 
 DEFINE_EVENT(i915_context, i915_context_create,
diff --git a/drivers/gpu/drm/i915/i915_utils.c b/drivers/gpu/drm/i915/i915_utils.c
index 16acdf7bdbe6..0348c6d0ef5f 100644
--- a/drivers/gpu/drm/i915/i915_utils.c
+++ b/drivers/gpu/drm/i915/i915_utils.c
@@ -54,25 +54,54 @@ __i915_printk(struct drm_i915_private *dev_priv, const char *level,
 #if IS_ENABLED(CONFIG_DRM_I915_DEBUG)
 static unsigned int i915_probe_fail_count;
 
-int __i915_inject_load_error(struct drm_i915_private *i915, int err,
-			     const char *func, int line)
+int __i915_inject_probe_error(struct drm_i915_private *i915, int err,
+			      const char *func, int line)
 {
-	if (i915_probe_fail_count >= i915_modparams.inject_load_failure)
+	if (i915_probe_fail_count >= i915_modparams.inject_probe_failure)
 		return 0;
 
-	if (++i915_probe_fail_count < i915_modparams.inject_load_failure)
+	if (++i915_probe_fail_count < i915_modparams.inject_probe_failure)
 		return 0;
 
 	__i915_printk(i915, KERN_INFO,
 		      "Injecting failure %d at checkpoint %u [%s:%d]\n",
-		      err, i915_modparams.inject_load_failure, func, line);
-	i915_modparams.inject_load_failure = 0;
+		      err, i915_modparams.inject_probe_failure, func, line);
+	i915_modparams.inject_probe_failure = 0;
 	return err;
 }
 
 bool i915_error_injected(void)
 {
-	return i915_probe_fail_count && !i915_modparams.inject_load_failure;
+	return i915_probe_fail_count && !i915_modparams.inject_probe_failure;
 }
 
 #endif
+
+void cancel_timer(struct timer_list *t)
+{
+	if (!READ_ONCE(t->expires))
+		return;
+
+	del_timer(t);
+	WRITE_ONCE(t->expires, 0);
+}
+
+void set_timer_ms(struct timer_list *t, unsigned long timeout)
+{
+	if (!timeout) {
+		cancel_timer(t);
+		return;
+	}
+
+	timeout = msecs_to_jiffies_timeout(timeout);
+
+	/*
+	 * Paranoia to make sure the compiler computes the timeout before
+	 * loading 'jiffies' as jiffies is volatile and may be updated in
+	 * the background by a timer tick. All to reduce the complexity
+	 * of the addition and reduce the risk of losing a jiffie.
+	 */
+	barrier();
+
+	mod_timer(t, jiffies + timeout);
+}
diff --git a/drivers/gpu/drm/i915/i915_utils.h b/drivers/gpu/drm/i915/i915_utils.h
index 562f756da421..04139ba1191e 100644
--- a/drivers/gpu/drm/i915/i915_utils.h
+++ b/drivers/gpu/drm/i915/i915_utils.h
@@ -32,6 +32,7 @@
 #include <linux/workqueue.h>
 
 struct drm_i915_private;
+struct timer_list;
 
 #undef WARN_ON
 /* Many gcc seem to no see through this and fall over :( */
@@ -60,20 +61,20 @@ __i915_printk(struct drm_i915_private *dev_priv, const char *level,
 
 #if IS_ENABLED(CONFIG_DRM_I915_DEBUG)
 
-int __i915_inject_load_error(struct drm_i915_private *i915, int err,
-			     const char *func, int line);
-#define i915_inject_load_error(_i915, _err) \
-	__i915_inject_load_error((_i915), (_err), __func__, __LINE__)
+int __i915_inject_probe_error(struct drm_i915_private *i915, int err,
+			      const char *func, int line);
+#define i915_inject_probe_error(_i915, _err) \
+	__i915_inject_probe_error((_i915), (_err), __func__, __LINE__)
 bool i915_error_injected(void);
 
 #else
 
-#define i915_inject_load_error(_i915, _err) 0
+#define i915_inject_probe_error(_i915, _err) 0
 #define i915_error_injected() false
 
 #endif
 
-#define i915_inject_probe_failure(i915) i915_inject_load_error((i915), -ENODEV)
+#define i915_inject_probe_failure(i915) i915_inject_probe_error((i915), -ENODEV)
 
 #define i915_probe_error(i915, fmt, ...)				   \
 	__i915_printk(i915, i915_error_injected() ? KERN_DEBUG : KERN_ERR, \
@@ -421,4 +422,25 @@ static inline void add_taint_for_CI(unsigned int taint)
 	add_taint(taint, LOCKDEP_STILL_OK);
 }
 
+void cancel_timer(struct timer_list *t);
+void set_timer_ms(struct timer_list *t, unsigned long timeout);
+
+static inline bool timer_expired(const struct timer_list *t)
+{
+	return READ_ONCE(t->expires) && !timer_pending(t);
+}
+
+/*
+ * This is a lookalike for IS_ENABLED() that takes a kconfig value,
+ * e.g. CONFIG_DRM_I915_SPIN_REQUEST, and evaluates whether it is non-zero
+ * i.e. whether the configuration is active. Wrapping up the config inside
+ * a boolean context prevents clang and smatch from complaining about potential
+ * issues in confusing logical-&& with bitwise-& for constants.
+ *
+ * Sadly IS_ENABLED() itself does not work with kconfig values.
+ *
+ * Returns 0 if @config is 0, 1 if set to any value.
+ */
+#define IS_ACTIVE(config) ((config) != 0)
+
 #endif /* !__I915_UTILS_H */
diff --git a/drivers/gpu/drm/i915/i915_vma.c b/drivers/gpu/drm/i915/i915_vma.c
index e0e677b2a3a9..e5512f26e20a 100644
--- a/drivers/gpu/drm/i915/i915_vma.c
+++ b/drivers/gpu/drm/i915/i915_vma.c
@@ -32,6 +32,7 @@
 
 #include "i915_drv.h"
 #include "i915_globals.h"
+#include "i915_sw_fence_work.h"
 #include "i915_trace.h"
 #include "i915_vma.h"
 
@@ -90,6 +91,7 @@ static int __i915_vma_active(struct i915_active *ref)
 	return i915_vma_tryget(active_to_vma(ref)) ? 0 : -ENOENT;
 }
 
+__i915_active_call
 static void __i915_vma_retire(struct i915_active *ref)
 {
 	i915_vma_put(active_to_vma(ref));
@@ -104,21 +106,21 @@ vma_create(struct drm_i915_gem_object *obj,
 	struct rb_node *rb, **p;
 
 	/* The aliasing_ppgtt should never be used directly! */
-	GEM_BUG_ON(vm == &vm->i915->ggtt.alias->vm);
+	GEM_BUG_ON(vm == &vm->gt->ggtt->alias->vm);
 
 	vma = i915_vma_alloc();
 	if (vma == NULL)
 		return ERR_PTR(-ENOMEM);
 
-	vma->vm = vm;
+	mutex_init(&vma->pages_mutex);
+	vma->vm = i915_vm_get(vm);
 	vma->ops = &vm->vma_ops;
 	vma->obj = obj;
 	vma->resv = obj->base.resv;
 	vma->size = obj->base.size;
 	vma->display_alignment = I915_GTT_MIN_ALIGNMENT;
 
-	i915_active_init(vm->i915, &vma->active,
-			 __i915_vma_active, __i915_vma_retire);
+	i915_active_init(&vma->active, __i915_vma_active, __i915_vma_retire);
 
 	/* Declare ourselves safe for use inside shrinkers */
 	if (IS_ENABLED(CONFIG_LOCKDEP)) {
@@ -171,7 +173,7 @@ vma_create(struct drm_i915_gem_object *obj,
 								i915_gem_object_get_stride(obj));
 		GEM_BUG_ON(!is_power_of_2(vma->fence_alignment));
 
-		vma->flags |= I915_VMA_GGTT;
+		__set_bit(I915_VMA_GGTT_BIT, __i915_vma_flags(vma));
 	}
 
 	spin_lock(&obj->vma.lock);
@@ -218,10 +220,6 @@ vma_create(struct drm_i915_gem_object *obj,
 
 	spin_unlock(&obj->vma.lock);
 
-	mutex_lock(&vm->mutex);
-	list_add(&vma->vm_link, &vm->unbound_list);
-	mutex_unlock(&vm->mutex);
-
 	return vma;
 
 err_vma:
@@ -265,8 +263,6 @@ vma_lookup(struct drm_i915_gem_object *obj,
  * Once created, the VMA is kept until either the object is freed, or the
  * address space is closed.
  *
- * Must be called with struct_mutex held.
- *
  * Returns the vma, or an error pointer.
  */
 struct i915_vma *
@@ -277,7 +273,7 @@ i915_vma_instance(struct drm_i915_gem_object *obj,
 	struct i915_vma *vma;
 
 	GEM_BUG_ON(view && !i915_is_ggtt(vm));
-	GEM_BUG_ON(vm->closed);
+	GEM_BUG_ON(!atomic_read(&vm->open));
 
 	spin_lock(&obj->vma.lock);
 	vma = vma_lookup(obj, vm, view);
@@ -291,18 +287,63 @@ i915_vma_instance(struct drm_i915_gem_object *obj,
 	return vma;
 }
 
+struct i915_vma_work {
+	struct dma_fence_work base;
+	struct i915_vma *vma;
+	enum i915_cache_level cache_level;
+	unsigned int flags;
+};
+
+static int __vma_bind(struct dma_fence_work *work)
+{
+	struct i915_vma_work *vw = container_of(work, typeof(*vw), base);
+	struct i915_vma *vma = vw->vma;
+	int err;
+
+	err = vma->ops->bind_vma(vma, vw->cache_level, vw->flags);
+	if (err)
+		atomic_or(I915_VMA_ERROR, &vma->flags);
+
+	if (vma->obj)
+		__i915_gem_object_unpin_pages(vma->obj);
+
+	return err;
+}
+
+static const struct dma_fence_work_ops bind_ops = {
+	.name = "bind",
+	.work = __vma_bind,
+};
+
+struct i915_vma_work *i915_vma_work(void)
+{
+	struct i915_vma_work *vw;
+
+	vw = kzalloc(sizeof(*vw), GFP_KERNEL);
+	if (!vw)
+		return NULL;
+
+	dma_fence_work_init(&vw->base, &bind_ops);
+	vw->base.dma.error = -EAGAIN; /* disable the worker by default */
+
+	return vw;
+}
+
 /**
  * i915_vma_bind - Sets up PTEs for an VMA in it's corresponding address space.
  * @vma: VMA to map
  * @cache_level: mapping cache level
  * @flags: flags like global or local mapping
+ * @work: preallocated worker for allocating and binding the PTE
  *
  * DMA addresses are taken from the scatter-gather table of this object (or of
  * this VMA in case of non-default GGTT views) and PTE entries set up.
  * Note that DMA addresses are also the only part of the SG table we care about.
  */
-int i915_vma_bind(struct i915_vma *vma, enum i915_cache_level cache_level,
-		  u32 flags)
+int i915_vma_bind(struct i915_vma *vma,
+		  enum i915_cache_level cache_level,
+		  u32 flags,
+		  struct i915_vma_work *work)
 {
 	u32 bind_flags;
 	u32 vma_flags;
@@ -319,13 +360,11 @@ int i915_vma_bind(struct i915_vma *vma, enum i915_cache_level cache_level,
 	if (GEM_DEBUG_WARN_ON(!flags))
 		return -EINVAL;
 
-	bind_flags = 0;
-	if (flags & PIN_GLOBAL)
-		bind_flags |= I915_VMA_GLOBAL_BIND;
-	if (flags & PIN_USER)
-		bind_flags |= I915_VMA_LOCAL_BIND;
+	bind_flags = flags;
+	bind_flags &= I915_VMA_GLOBAL_BIND | I915_VMA_LOCAL_BIND;
 
-	vma_flags = vma->flags & (I915_VMA_GLOBAL_BIND | I915_VMA_LOCAL_BIND);
+	vma_flags = atomic_read(&vma->flags);
+	vma_flags &= I915_VMA_GLOBAL_BIND | I915_VMA_LOCAL_BIND;
 	if (flags & PIN_UPDATE)
 		bind_flags |= vma_flags;
 	else
@@ -336,11 +375,34 @@ int i915_vma_bind(struct i915_vma *vma, enum i915_cache_level cache_level,
 	GEM_BUG_ON(!vma->pages);
 
 	trace_i915_vma_bind(vma, bind_flags);
-	ret = vma->ops->bind_vma(vma, cache_level, bind_flags);
-	if (ret)
-		return ret;
+	if (work && (bind_flags & ~vma_flags) & vma->vm->bind_async_flags) {
+		work->vma = vma;
+		work->cache_level = cache_level;
+		work->flags = bind_flags | I915_VMA_ALLOC;
 
-	vma->flags |= bind_flags;
+		/*
+		 * Note we only want to chain up to the migration fence on
+		 * the pages (not the object itself). As we don't track that,
+		 * yet, we have to use the exclusive fence instead.
+		 *
+		 * Also note that we do not want to track the async vma as
+		 * part of the obj->resv->excl_fence as it only affects
+		 * execution and not content or object's backing store lifetime.
+		 */
+		GEM_BUG_ON(i915_active_has_exclusive(&vma->active));
+		i915_active_set_exclusive(&vma->active, &work->base.dma);
+		work->base.dma.error = 0; /* enable the queue_work() */
+
+		if (vma->obj)
+			__i915_gem_object_pin_pages(vma->obj);
+	} else {
+		GEM_BUG_ON((bind_flags & ~vma_flags) & vma->vm->bind_async_flags);
+		ret = vma->ops->bind_vma(vma, cache_level, bind_flags);
+		if (ret)
+			return ret;
+	}
+
+	atomic_or(bind_flags, &vma->flags);
 	return 0;
 }
 
@@ -350,18 +412,16 @@ void __iomem *i915_vma_pin_iomap(struct i915_vma *vma)
 	int err;
 
 	/* Access through the GTT requires the device to be awake. */
-	assert_rpm_wakelock_held(&vma->vm->i915->runtime_pm);
-
-	lockdep_assert_held(&vma->vm->i915->drm.struct_mutex);
-	if (WARN_ON(!i915_vma_is_map_and_fenceable(vma))) {
+	assert_rpm_wakelock_held(vma->vm->gt->uncore->rpm);
+	if (GEM_WARN_ON(!i915_vma_is_map_and_fenceable(vma))) {
 		err = -ENODEV;
 		goto err;
 	}
 
 	GEM_BUG_ON(!i915_vma_is_ggtt(vma));
-	GEM_BUG_ON((vma->flags & I915_VMA_GLOBAL_BIND) == 0);
+	GEM_BUG_ON(!i915_vma_is_bound(vma, I915_VMA_GLOBAL_BIND));
 
-	ptr = vma->iomap;
+	ptr = READ_ONCE(vma->iomap);
 	if (ptr == NULL) {
 		ptr = io_mapping_map_wc(&i915_vm_to_ggtt(vma->vm)->iomap,
 					vma->node.start,
@@ -371,7 +431,10 @@ void __iomem *i915_vma_pin_iomap(struct i915_vma *vma)
 			goto err;
 		}
 
-		vma->iomap = ptr;
+		if (unlikely(cmpxchg(&vma->iomap, NULL, ptr))) {
+			io_mapping_unmap(ptr);
+			ptr = vma->iomap;
+		}
 	}
 
 	__i915_vma_pin(vma);
@@ -391,18 +454,12 @@ err:
 
 void i915_vma_flush_writes(struct i915_vma *vma)
 {
-	if (!i915_vma_has_ggtt_write(vma))
-		return;
-
-	intel_gt_flush_ggtt_writes(vma->vm->gt);
-
-	i915_vma_unset_ggtt_write(vma);
+	if (i915_vma_unset_ggtt_write(vma))
+		intel_gt_flush_ggtt_writes(vma->vm->gt);
 }
 
 void i915_vma_unpin_iomap(struct i915_vma *vma)
 {
-	lockdep_assert_held(&vma->vm->i915->drm.struct_mutex);
-
 	GEM_BUG_ON(vma->iomap == NULL);
 
 	i915_vma_flush_writes(vma);
@@ -438,6 +495,9 @@ bool i915_vma_misplaced(const struct i915_vma *vma,
 	if (!drm_mm_node_allocated(&vma->node))
 		return false;
 
+	if (test_bit(I915_VMA_ERROR_BIT, __i915_vma_flags(vma)))
+		return true;
+
 	if (vma->node.size < size)
 		return true;
 
@@ -472,17 +532,12 @@ void __i915_vma_set_map_and_fenceable(struct i915_vma *vma)
 	mappable = vma->node.start + vma->fence_size <= i915_vm_to_ggtt(vma->vm)->mappable_end;
 
 	if (mappable && fenceable)
-		vma->flags |= I915_VMA_CAN_FENCE;
+		set_bit(I915_VMA_CAN_FENCE_BIT, __i915_vma_flags(vma));
 	else
-		vma->flags &= ~I915_VMA_CAN_FENCE;
-}
-
-static bool color_differs(struct drm_mm_node *node, unsigned long color)
-{
-	return node->allocated && node->color != color;
+		clear_bit(I915_VMA_CAN_FENCE_BIT, __i915_vma_flags(vma));
 }
 
-bool i915_gem_valid_gtt_space(struct i915_vma *vma, unsigned long cache_level)
+bool i915_gem_valid_gtt_space(struct i915_vma *vma, unsigned long color)
 {
 	struct drm_mm_node *node = &vma->node;
 	struct drm_mm_node *other;
@@ -494,7 +549,7 @@ bool i915_gem_valid_gtt_space(struct i915_vma *vma, unsigned long cache_level)
 	 * these constraints apply and set the drm_mm.color_adjust
 	 * appropriately.
 	 */
-	if (vma->vm->mm.color_adjust == NULL)
+	if (!i915_vm_has_cache_coloring(vma->vm))
 		return true;
 
 	/* Only valid to be called on an already inserted vma */
@@ -502,11 +557,13 @@ bool i915_gem_valid_gtt_space(struct i915_vma *vma, unsigned long cache_level)
 	GEM_BUG_ON(list_empty(&node->node_list));
 
 	other = list_prev_entry(node, node_list);
-	if (color_differs(other, cache_level) && !drm_mm_hole_follows(other))
+	if (i915_node_color_differs(other, color) &&
+	    !drm_mm_hole_follows(other))
 		return false;
 
 	other = list_next_entry(node, node_list);
-	if (color_differs(other, cache_level) && !drm_mm_hole_follows(node))
+	if (i915_node_color_differs(other, color) &&
+	    !drm_mm_hole_follows(node))
 		return false;
 
 	return true;
@@ -541,13 +598,12 @@ static void assert_bind_count(const struct drm_i915_gem_object *obj)
 static int
 i915_vma_insert(struct i915_vma *vma, u64 size, u64 alignment, u64 flags)
 {
-	struct drm_i915_private *dev_priv = vma->vm->i915;
-	unsigned int cache_level;
+	unsigned long color;
 	u64 start, end;
 	int ret;
 
 	GEM_BUG_ON(i915_vma_is_closed(vma));
-	GEM_BUG_ON(vma->flags & (I915_VMA_GLOBAL_BIND | I915_VMA_LOCAL_BIND));
+	GEM_BUG_ON(i915_vma_is_bound(vma, I915_VMA_GLOBAL_BIND | I915_VMA_LOCAL_BIND));
 	GEM_BUG_ON(drm_mm_node_allocated(&vma->node));
 
 	size = max(size, vma->size);
@@ -567,7 +623,7 @@ i915_vma_insert(struct i915_vma *vma, u64 size, u64 alignment, u64 flags)
 
 	end = vma->vm->total;
 	if (flags & PIN_MAPPABLE)
-		end = min_t(u64, end, dev_priv->ggtt.mappable_end);
+		end = min_t(u64, end, i915_vm_to_ggtt(vma->vm)->mappable_end);
 	if (flags & PIN_ZONE_4G)
 		end = min_t(u64, end, (1ULL << 32) - I915_GTT_PAGE_SIZE);
 	GEM_BUG_ON(!IS_ALIGNED(end, I915_GTT_PAGE_SIZE));
@@ -583,35 +639,21 @@ i915_vma_insert(struct i915_vma *vma, u64 size, u64 alignment, u64 flags)
 		return -ENOSPC;
 	}
 
-	if (vma->obj) {
-		ret = i915_gem_object_pin_pages(vma->obj);
-		if (ret)
-			return ret;
-
-		cache_level = vma->obj->cache_level;
-	} else {
-		cache_level = 0;
-	}
-
-	GEM_BUG_ON(vma->pages);
-
-	ret = vma->ops->set_pages(vma);
-	if (ret)
-		goto err_unpin;
+	color = 0;
+	if (vma->obj && i915_vm_has_cache_coloring(vma->vm))
+		color = vma->obj->cache_level;
 
 	if (flags & PIN_OFFSET_FIXED) {
 		u64 offset = flags & PIN_OFFSET_MASK;
 		if (!IS_ALIGNED(offset, alignment) ||
-		    range_overflows(offset, size, end)) {
-			ret = -EINVAL;
-			goto err_clear;
-		}
+		    range_overflows(offset, size, end))
+			return -EINVAL;
 
 		ret = i915_gem_gtt_reserve(vma->vm, &vma->node,
-					   size, offset, cache_level,
+					   size, offset, color,
 					   flags);
 		if (ret)
-			goto err_clear;
+			return ret;
 	} else {
 		/*
 		 * We only support huge gtt pages through the 48b PPGTT,
@@ -647,116 +689,259 @@ i915_vma_insert(struct i915_vma *vma, u64 size, u64 alignment, u64 flags)
 		}
 
 		ret = i915_gem_gtt_insert(vma->vm, &vma->node,
-					  size, alignment, cache_level,
+					  size, alignment, color,
 					  start, end, flags);
 		if (ret)
-			goto err_clear;
+			return ret;
 
 		GEM_BUG_ON(vma->node.start < start);
 		GEM_BUG_ON(vma->node.start + vma->node.size > end);
 	}
 	GEM_BUG_ON(!drm_mm_node_allocated(&vma->node));
-	GEM_BUG_ON(!i915_gem_valid_gtt_space(vma, cache_level));
-
-	mutex_lock(&vma->vm->mutex);
-	list_move_tail(&vma->vm_link, &vma->vm->bound_list);
-	mutex_unlock(&vma->vm->mutex);
+	GEM_BUG_ON(!i915_gem_valid_gtt_space(vma, color));
 
 	if (vma->obj) {
-		atomic_inc(&vma->obj->bind_count);
-		assert_bind_count(vma->obj);
+		struct drm_i915_gem_object *obj = vma->obj;
+
+		atomic_inc(&obj->bind_count);
+		assert_bind_count(obj);
 	}
+	list_add_tail(&vma->vm_link, &vma->vm->bound_list);
 
 	return 0;
-
-err_clear:
-	vma->ops->clear_pages(vma);
-err_unpin:
-	if (vma->obj)
-		i915_gem_object_unpin_pages(vma->obj);
-	return ret;
 }
 
 static void
-i915_vma_remove(struct i915_vma *vma)
+i915_vma_detach(struct i915_vma *vma)
 {
 	GEM_BUG_ON(!drm_mm_node_allocated(&vma->node));
-	GEM_BUG_ON(vma->flags & (I915_VMA_GLOBAL_BIND | I915_VMA_LOCAL_BIND));
-
-	vma->ops->clear_pages(vma);
-
-	mutex_lock(&vma->vm->mutex);
-	drm_mm_remove_node(&vma->node);
-	list_move_tail(&vma->vm_link, &vma->vm->unbound_list);
-	mutex_unlock(&vma->vm->mutex);
+	GEM_BUG_ON(i915_vma_is_bound(vma, I915_VMA_GLOBAL_BIND | I915_VMA_LOCAL_BIND));
 
 	/*
-	 * Since the unbound list is global, only move to that list if
-	 * no more VMAs exist.
+	 * And finally now the object is completely decoupled from this
+	 * vma, we can drop its hold on the backing storage and allow
+	 * it to be reaped by the shrinker.
 	 */
+	list_del(&vma->vm_link);
 	if (vma->obj) {
 		struct drm_i915_gem_object *obj = vma->obj;
 
+		assert_bind_count(obj);
 		atomic_dec(&obj->bind_count);
+	}
+}
 
-		/*
-		 * And finally now the object is completely decoupled from this
-		 * vma, we can drop its hold on the backing storage and allow
-		 * it to be reaped by the shrinker.
-		 */
-		i915_gem_object_unpin_pages(obj);
-		assert_bind_count(obj);
+static bool try_qad_pin(struct i915_vma *vma, unsigned int flags)
+{
+	unsigned int bound;
+	bool pinned = true;
+
+	bound = atomic_read(&vma->flags);
+	do {
+		if (unlikely(flags & ~bound))
+			return false;
+
+		if (unlikely(bound & (I915_VMA_OVERFLOW | I915_VMA_ERROR)))
+			return false;
+
+		if (!(bound & I915_VMA_PIN_MASK))
+			goto unpinned;
+
+		GEM_BUG_ON(((bound + 1) & I915_VMA_PIN_MASK) == 0);
+	} while (!atomic_try_cmpxchg(&vma->flags, &bound, bound + 1));
+
+	return true;
+
+unpinned:
+	/*
+	 * If pin_count==0, but we are bound, check under the lock to avoid
+	 * racing with a concurrent i915_vma_unbind().
+	 */
+	mutex_lock(&vma->vm->mutex);
+	do {
+		if (unlikely(bound & (I915_VMA_OVERFLOW | I915_VMA_ERROR))) {
+			pinned = false;
+			break;
+		}
+
+		if (unlikely(flags & ~bound)) {
+			pinned = false;
+			break;
+		}
+	} while (!atomic_try_cmpxchg(&vma->flags, &bound, bound + 1));
+	mutex_unlock(&vma->vm->mutex);
+
+	return pinned;
+}
+
+static int vma_get_pages(struct i915_vma *vma)
+{
+	int err = 0;
+
+	if (atomic_add_unless(&vma->pages_count, 1, 0))
+		return 0;
+
+	/* Allocations ahoy! */
+	if (mutex_lock_interruptible(&vma->pages_mutex))
+		return -EINTR;
+
+	if (!atomic_read(&vma->pages_count)) {
+		if (vma->obj) {
+			err = i915_gem_object_pin_pages(vma->obj);
+			if (err)
+				goto unlock;
+		}
+
+		err = vma->ops->set_pages(vma);
+		if (err) {
+			if (vma->obj)
+				i915_gem_object_unpin_pages(vma->obj);
+			goto unlock;
+		}
 	}
+	atomic_inc(&vma->pages_count);
+
+unlock:
+	mutex_unlock(&vma->pages_mutex);
+
+	return err;
 }
 
-int __i915_vma_do_pin(struct i915_vma *vma,
-		      u64 size, u64 alignment, u64 flags)
+static void __vma_put_pages(struct i915_vma *vma, unsigned int count)
 {
-	const unsigned int bound = vma->flags;
-	int ret;
+	/* We allocate under vma_get_pages, so beware the shrinker */
+	mutex_lock_nested(&vma->pages_mutex, SINGLE_DEPTH_NESTING);
+	GEM_BUG_ON(atomic_read(&vma->pages_count) < count);
+	if (atomic_sub_return(count, &vma->pages_count) == 0) {
+		vma->ops->clear_pages(vma);
+		GEM_BUG_ON(vma->pages);
+		if (vma->obj)
+			i915_gem_object_unpin_pages(vma->obj);
+	}
+	mutex_unlock(&vma->pages_mutex);
+}
 
-	lockdep_assert_held(&vma->vm->i915->drm.struct_mutex);
-	GEM_BUG_ON((flags & (PIN_GLOBAL | PIN_USER)) == 0);
-	GEM_BUG_ON((flags & PIN_GLOBAL) && !i915_vma_is_ggtt(vma));
+static void vma_put_pages(struct i915_vma *vma)
+{
+	if (atomic_add_unless(&vma->pages_count, -1, 1))
+		return;
 
-	if (WARN_ON(bound & I915_VMA_PIN_OVERFLOW)) {
-		ret = -EBUSY;
-		goto err_unpin;
+	__vma_put_pages(vma, 1);
+}
+
+static void vma_unbind_pages(struct i915_vma *vma)
+{
+	unsigned int count;
+
+	lockdep_assert_held(&vma->vm->mutex);
+
+	/* The upper portion of pages_count is the number of bindings */
+	count = atomic_read(&vma->pages_count);
+	count >>= I915_VMA_PAGES_BIAS;
+	GEM_BUG_ON(!count);
+
+	__vma_put_pages(vma, count | count << I915_VMA_PAGES_BIAS);
+}
+
+int i915_vma_pin(struct i915_vma *vma, u64 size, u64 alignment, u64 flags)
+{
+	struct i915_vma_work *work = NULL;
+	unsigned int bound;
+	int err;
+
+	BUILD_BUG_ON(PIN_GLOBAL != I915_VMA_GLOBAL_BIND);
+	BUILD_BUG_ON(PIN_USER != I915_VMA_LOCAL_BIND);
+
+	GEM_BUG_ON(flags & PIN_UPDATE);
+	GEM_BUG_ON(!(flags & (PIN_USER | PIN_GLOBAL)));
+
+	/* First try and grab the pin without rebinding the vma */
+	if (try_qad_pin(vma, flags & I915_VMA_BIND_MASK))
+		return 0;
+
+	err = vma_get_pages(vma);
+	if (err)
+		return err;
+
+	if (flags & vma->vm->bind_async_flags) {
+		work = i915_vma_work();
+		if (!work) {
+			err = -ENOMEM;
+			goto err_pages;
+		}
 	}
 
-	if ((bound & I915_VMA_BIND_MASK) == 0) {
-		ret = i915_vma_insert(vma, size, alignment, flags);
-		if (ret)
-			goto err_unpin;
+	/* No more allocations allowed once we hold vm->mutex */
+	err = mutex_lock_interruptible(&vma->vm->mutex);
+	if (err)
+		goto err_fence;
+
+	bound = atomic_read(&vma->flags);
+	if (unlikely(bound & I915_VMA_ERROR)) {
+		err = -ENOMEM;
+		goto err_unlock;
 	}
-	GEM_BUG_ON(!drm_mm_node_allocated(&vma->node));
 
-	ret = i915_vma_bind(vma, vma->obj ? vma->obj->cache_level : 0, flags);
-	if (ret)
-		goto err_remove;
+	if (unlikely(!((bound + 1) & I915_VMA_PIN_MASK))) {
+		err = -EAGAIN; /* pins are meant to be fairly temporary */
+		goto err_unlock;
+	}
 
-	GEM_BUG_ON((vma->flags & I915_VMA_BIND_MASK) == 0);
+	if (unlikely(!(flags & ~bound & I915_VMA_BIND_MASK))) {
+		__i915_vma_pin(vma);
+		goto err_unlock;
+	}
 
-	if ((bound ^ vma->flags) & I915_VMA_GLOBAL_BIND)
-		__i915_vma_set_map_and_fenceable(vma);
+	err = i915_active_acquire(&vma->active);
+	if (err)
+		goto err_unlock;
+
+	if (!(bound & I915_VMA_BIND_MASK)) {
+		err = i915_vma_insert(vma, size, alignment, flags);
+		if (err)
+			goto err_active;
+
+		if (i915_is_ggtt(vma->vm))
+			__i915_vma_set_map_and_fenceable(vma);
+	}
+
+	GEM_BUG_ON(!vma->pages);
+	err = i915_vma_bind(vma,
+			    vma->obj ? vma->obj->cache_level : 0,
+			    flags, work);
+	if (err)
+		goto err_remove;
+
+	/* There should only be at most 2 active bindings (user, global) */
+	GEM_BUG_ON(bound + I915_VMA_PAGES_ACTIVE < bound);
+	atomic_add(I915_VMA_PAGES_ACTIVE, &vma->pages_count);
+	list_move_tail(&vma->vm_link, &vma->vm->bound_list);
 
+	__i915_vma_pin(vma);
+	GEM_BUG_ON(!i915_vma_is_pinned(vma));
+	GEM_BUG_ON(!i915_vma_is_bound(vma, flags));
 	GEM_BUG_ON(i915_vma_misplaced(vma, size, alignment, flags));
-	return 0;
 
 err_remove:
-	if ((bound & I915_VMA_BIND_MASK) == 0) {
-		i915_vma_remove(vma);
-		GEM_BUG_ON(vma->pages);
-		GEM_BUG_ON(vma->flags & I915_VMA_BIND_MASK);
+	if (!i915_vma_is_bound(vma, I915_VMA_BIND_MASK)) {
+		i915_vma_detach(vma);
+		drm_mm_remove_node(&vma->node);
 	}
-err_unpin:
-	__i915_vma_unpin(vma);
-	return ret;
+err_active:
+	i915_active_release(&vma->active);
+err_unlock:
+	mutex_unlock(&vma->vm->mutex);
+err_fence:
+	if (work)
+		dma_fence_work_commit(&work->base);
+err_pages:
+	vma_put_pages(vma);
+	return err;
 }
 
 void i915_vma_close(struct i915_vma *vma)
 {
-	struct drm_i915_private *i915 = vma->vm->i915;
+	struct intel_gt *gt = vma->vm->gt;
 	unsigned long flags;
 
 	GEM_BUG_ON(i915_vma_is_closed(vma));
@@ -773,79 +958,87 @@ void i915_vma_close(struct i915_vma *vma)
 	 * causing us to rebind the VMA once more. This ends up being a lot
 	 * of wasted work for the steady state.
 	 */
-	spin_lock_irqsave(&i915->gt.closed_lock, flags);
-	list_add(&vma->closed_link, &i915->gt.closed_vma);
-	spin_unlock_irqrestore(&i915->gt.closed_lock, flags);
+	spin_lock_irqsave(&gt->closed_lock, flags);
+	list_add(&vma->closed_link, &gt->closed_vma);
+	spin_unlock_irqrestore(&gt->closed_lock, flags);
 }
 
 static void __i915_vma_remove_closed(struct i915_vma *vma)
 {
-	struct drm_i915_private *i915 = vma->vm->i915;
-
-	if (!i915_vma_is_closed(vma))
-		return;
+	struct intel_gt *gt = vma->vm->gt;
 
-	spin_lock_irq(&i915->gt.closed_lock);
+	spin_lock_irq(&gt->closed_lock);
 	list_del_init(&vma->closed_link);
-	spin_unlock_irq(&i915->gt.closed_lock);
+	spin_unlock_irq(&gt->closed_lock);
 }
 
 void i915_vma_reopen(struct i915_vma *vma)
 {
-	__i915_vma_remove_closed(vma);
+	if (i915_vma_is_closed(vma))
+		__i915_vma_remove_closed(vma);
 }
 
-static void __i915_vma_destroy(struct i915_vma *vma)
+void i915_vma_destroy(struct i915_vma *vma)
 {
-	GEM_BUG_ON(vma->node.allocated);
-	GEM_BUG_ON(vma->fence);
-
-	mutex_lock(&vma->vm->mutex);
-	list_del(&vma->vm_link);
-	mutex_unlock(&vma->vm->mutex);
+	if (drm_mm_node_allocated(&vma->node)) {
+		mutex_lock(&vma->vm->mutex);
+		atomic_and(~I915_VMA_PIN_MASK, &vma->flags);
+		WARN_ON(__i915_vma_unbind(vma));
+		mutex_unlock(&vma->vm->mutex);
+		GEM_BUG_ON(drm_mm_node_allocated(&vma->node));
+	}
+	GEM_BUG_ON(i915_vma_is_active(vma));
 
 	if (vma->obj) {
 		struct drm_i915_gem_object *obj = vma->obj;
 
 		spin_lock(&obj->vma.lock);
 		list_del(&vma->obj_link);
-		rb_erase(&vma->obj_node, &vma->obj->vma.tree);
+		rb_erase(&vma->obj_node, &obj->vma.tree);
 		spin_unlock(&obj->vma.lock);
 	}
 
-	i915_active_fini(&vma->active);
+	__i915_vma_remove_closed(vma);
+	i915_vm_put(vma->vm);
 
+	i915_active_fini(&vma->active);
 	i915_vma_free(vma);
 }
 
-void i915_vma_destroy(struct i915_vma *vma)
+void i915_vma_parked(struct intel_gt *gt)
 {
-	lockdep_assert_held(&vma->vm->i915->drm.struct_mutex);
+	struct i915_vma *vma, *next;
 
-	GEM_BUG_ON(i915_vma_is_pinned(vma));
+	spin_lock_irq(&gt->closed_lock);
+	list_for_each_entry_safe(vma, next, &gt->closed_vma, closed_link) {
+		struct drm_i915_gem_object *obj = vma->obj;
+		struct i915_address_space *vm = vma->vm;
 
-	__i915_vma_remove_closed(vma);
+		/* XXX All to avoid keeping a reference on i915_vma itself */
 
-	WARN_ON(i915_vma_unbind(vma));
-	GEM_BUG_ON(i915_vma_is_active(vma));
+		if (!kref_get_unless_zero(&obj->base.refcount))
+			continue;
 
-	__i915_vma_destroy(vma);
-}
+		if (!i915_vm_tryopen(vm)) {
+			i915_gem_object_put(obj);
+			obj = NULL;
+		}
 
-void i915_vma_parked(struct drm_i915_private *i915)
-{
-	struct i915_vma *vma, *next;
+		spin_unlock_irq(&gt->closed_lock);
 
-	spin_lock_irq(&i915->gt.closed_lock);
-	list_for_each_entry_safe(vma, next, &i915->gt.closed_vma, closed_link) {
-		list_del_init(&vma->closed_link);
-		spin_unlock_irq(&i915->gt.closed_lock);
+		if (obj) {
+			i915_vma_destroy(vma);
+			i915_gem_object_put(obj);
+		}
 
-		i915_vma_destroy(vma);
+		i915_vm_close(vm);
 
-		spin_lock_irq(&i915->gt.closed_lock);
+		/* Restart after dropping lock */
+		spin_lock_irq(&gt->closed_lock);
+		next = list_first_entry(&gt->closed_vma,
+					typeof(*next), closed_link);
 	}
-	spin_unlock_irq(&i915->gt.closed_lock);
+	spin_unlock_irq(&gt->closed_lock);
 }
 
 static void __i915_vma_iounmap(struct i915_vma *vma)
@@ -883,6 +1076,20 @@ void i915_vma_revoke_mmap(struct i915_vma *vma)
 		list_del(&vma->obj->userfault_link);
 }
 
+int __i915_vma_move_to_active(struct i915_vma *vma, struct i915_request *rq)
+{
+	int err;
+
+	GEM_BUG_ON(!i915_vma_is_pinned(vma));
+
+	/* Wait for the vma to be bound before we start! */
+	err = i915_request_await_active(rq, &vma->active);
+	if (err)
+		return err;
+
+	return i915_active_add_request(&vma->active, rq);
+}
+
 int i915_vma_move_to_active(struct i915_vma *vma,
 			    struct i915_request *rq,
 			    unsigned int flags)
@@ -890,27 +1097,15 @@ int i915_vma_move_to_active(struct i915_vma *vma,
 	struct drm_i915_gem_object *obj = vma->obj;
 	int err;
 
-	assert_vma_held(vma);
 	assert_object_held(obj);
-	GEM_BUG_ON(!drm_mm_node_allocated(&vma->node));
 
-	/*
-	 * Add a reference if we're newly entering the active list.
-	 * The order in which we add operations to the retirement queue is
-	 * vital here: mark_active adds to the start of the callback list,
-	 * such that subsequent callbacks are called first. Therefore we
-	 * add the active reference first and queue for it to be dropped
-	 * *last*.
-	 */
-	err = i915_active_ref(&vma->active, rq->timeline, rq);
+	err = __i915_vma_move_to_active(vma, rq);
 	if (unlikely(err))
 		return err;
 
 	if (flags & EXEC_OBJECT_WRITE) {
 		if (intel_frontbuffer_invalidate(obj->frontbuffer, ORIGIN_CS))
-			i915_active_ref(&obj->frontbuffer->write,
-					rq->timeline,
-					rq);
+			i915_active_add_request(&obj->frontbuffer->write, rq);
 
 		dma_resv_add_excl_fence(vma->resv, &rq->fence);
 		obj->write_domain = I915_GEM_DOMAIN_RENDER;
@@ -930,44 +1125,31 @@ int i915_vma_move_to_active(struct i915_vma *vma,
 	return 0;
 }
 
-int i915_vma_unbind(struct i915_vma *vma)
+int __i915_vma_unbind(struct i915_vma *vma)
 {
 	int ret;
 
-	lockdep_assert_held(&vma->vm->i915->drm.struct_mutex);
+	lockdep_assert_held(&vma->vm->mutex);
 
 	/*
 	 * First wait upon any activity as retiring the request may
 	 * have side-effects such as unpinning or even unbinding this vma.
+	 *
+	 * XXX Actually waiting under the vm->mutex is a hinderance and
+	 * should be pipelined wherever possible. In cases where that is
+	 * unavoidable, we should lift the wait to before the mutex.
 	 */
-	might_sleep();
-	if (i915_vma_is_active(vma)) {
-		/*
-		 * When a closed VMA is retired, it is unbound - eek.
-		 * In order to prevent it from being recursively closed,
-		 * take a pin on the vma so that the second unbind is
-		 * aborted.
-		 *
-		 * Even more scary is that the retire callback may free
-		 * the object (last active vma). To prevent the explosion
-		 * we defer the actual object free to a worker that can
-		 * only proceed once it acquires the struct_mutex (which
-		 * we currently hold, therefore it cannot free this object
-		 * before we are finished).
-		 */
-		__i915_vma_pin(vma);
-		ret = i915_active_wait(&vma->active);
-		__i915_vma_unpin(vma);
-		if (ret)
-			return ret;
-	}
-	GEM_BUG_ON(i915_vma_is_active(vma));
+	ret = i915_vma_sync(vma);
+	if (ret)
+		return ret;
 
+	GEM_BUG_ON(i915_vma_is_active(vma));
 	if (i915_vma_is_pinned(vma)) {
 		vma_print_allocator(vma, "is pinned");
 		return -EBUSY;
 	}
 
+	GEM_BUG_ON(i915_vma_is_active(vma));
 	if (!drm_mm_node_allocated(&vma->node))
 		return 0;
 
@@ -982,34 +1164,47 @@ int i915_vma_unbind(struct i915_vma *vma)
 		GEM_BUG_ON(i915_vma_has_ggtt_write(vma));
 
 		/* release the fence reg _after_ flushing */
-		mutex_lock(&vma->vm->mutex);
 		ret = i915_vma_revoke_fence(vma);
-		mutex_unlock(&vma->vm->mutex);
 		if (ret)
 			return ret;
 
 		/* Force a pagefault for domain tracking on next user access */
-		mutex_lock(&vma->vm->mutex);
 		i915_vma_revoke_mmap(vma);
-		mutex_unlock(&vma->vm->mutex);
 
 		__i915_vma_iounmap(vma);
-		vma->flags &= ~I915_VMA_CAN_FENCE;
+		clear_bit(I915_VMA_CAN_FENCE_BIT, __i915_vma_flags(vma));
 	}
 	GEM_BUG_ON(vma->fence);
 	GEM_BUG_ON(i915_vma_has_userfault(vma));
 
-	if (likely(!vma->vm->closed)) {
+	if (likely(atomic_read(&vma->vm->open))) {
 		trace_i915_vma_unbind(vma);
 		vma->ops->unbind_vma(vma);
 	}
-	vma->flags &= ~(I915_VMA_GLOBAL_BIND | I915_VMA_LOCAL_BIND);
+	atomic_and(~(I915_VMA_BIND_MASK | I915_VMA_ERROR), &vma->flags);
 
-	i915_vma_remove(vma);
+	i915_vma_detach(vma);
+	vma_unbind_pages(vma);
 
+	drm_mm_remove_node(&vma->node); /* pairs with i915_vma_destroy() */
 	return 0;
 }
 
+int i915_vma_unbind(struct i915_vma *vma)
+{
+	struct i915_address_space *vm = vma->vm;
+	int err;
+
+	err = mutex_lock_interruptible(&vm->mutex);
+	if (err)
+		return err;
+
+	err = __i915_vma_unbind(vma);
+	mutex_unlock(&vm->mutex);
+
+	return err;
+}
+
 struct i915_vma *i915_vma_make_unshrinkable(struct i915_vma *vma)
 {
 	i915_gem_object_make_unshrinkable(vma->obj);
diff --git a/drivers/gpu/drm/i915/i915_vma.h b/drivers/gpu/drm/i915/i915_vma.h
index 889fc7cb910a..465932813bc5 100644
--- a/drivers/gpu/drm/i915/i915_vma.h
+++ b/drivers/gpu/drm/i915/i915_vma.h
@@ -72,7 +72,7 @@ struct i915_vma {
 	 * that exist in the ctx->handle_vmas LUT for this vma.
 	 */
 	atomic_t open_count;
-	unsigned long flags;
+	atomic_t flags;
 	/**
 	 * How many users have pinned this object in GTT space.
 	 *
@@ -96,22 +96,41 @@ struct i915_vma {
 	 * exclusive cachelines of a single page, so a maximum of 64 possible
 	 * users.
 	 */
-#define I915_VMA_PIN_MASK 0xff
-#define I915_VMA_PIN_OVERFLOW	BIT(8)
+#define I915_VMA_PIN_MASK 0x3ff
+#define I915_VMA_OVERFLOW 0x200
 
 	/** Flags and address space this VMA is bound to */
-#define I915_VMA_GLOBAL_BIND	BIT(9)
-#define I915_VMA_LOCAL_BIND	BIT(10)
-#define I915_VMA_BIND_MASK (I915_VMA_GLOBAL_BIND | I915_VMA_LOCAL_BIND | I915_VMA_PIN_OVERFLOW)
+#define I915_VMA_GLOBAL_BIND_BIT 10
+#define I915_VMA_LOCAL_BIND_BIT  11
 
-#define I915_VMA_GGTT		BIT(11)
-#define I915_VMA_CAN_FENCE	BIT(12)
-#define I915_VMA_USERFAULT_BIT	13
-#define I915_VMA_USERFAULT	BIT(I915_VMA_USERFAULT_BIT)
-#define I915_VMA_GGTT_WRITE	BIT(14)
+#define I915_VMA_GLOBAL_BIND	((int)BIT(I915_VMA_GLOBAL_BIND_BIT))
+#define I915_VMA_LOCAL_BIND	((int)BIT(I915_VMA_LOCAL_BIND_BIT))
+
+#define I915_VMA_BIND_MASK (I915_VMA_GLOBAL_BIND | I915_VMA_LOCAL_BIND)
+
+#define I915_VMA_ALLOC_BIT	12
+#define I915_VMA_ALLOC		((int)BIT(I915_VMA_ALLOC_BIT))
+
+#define I915_VMA_ERROR_BIT	13
+#define I915_VMA_ERROR		((int)BIT(I915_VMA_ERROR_BIT))
+
+#define I915_VMA_GGTT_BIT	14
+#define I915_VMA_CAN_FENCE_BIT	15
+#define I915_VMA_USERFAULT_BIT	16
+#define I915_VMA_GGTT_WRITE_BIT	17
+
+#define I915_VMA_GGTT		((int)BIT(I915_VMA_GGTT_BIT))
+#define I915_VMA_CAN_FENCE	((int)BIT(I915_VMA_CAN_FENCE_BIT))
+#define I915_VMA_USERFAULT	((int)BIT(I915_VMA_USERFAULT_BIT))
+#define I915_VMA_GGTT_WRITE	((int)BIT(I915_VMA_GGTT_WRITE_BIT))
 
 	struct i915_active active;
 
+#define I915_VMA_PAGES_BIAS 24
+#define I915_VMA_PAGES_ACTIVE (BIT(24) | 1)
+	atomic_t pages_count; /* number of active binds to the pages */
+	struct mutex pages_mutex; /* protect acquire/release of backing pages */
+
 	/**
 	 * Support different GGTT views into the same object.
 	 * This means there can be multiple VMA mappings per object and per VM.
@@ -158,52 +177,57 @@ static inline bool i915_vma_is_active(const struct i915_vma *vma)
 	return !i915_active_is_idle(&vma->active);
 }
 
+int __must_check __i915_vma_move_to_active(struct i915_vma *vma,
+					   struct i915_request *rq);
 int __must_check i915_vma_move_to_active(struct i915_vma *vma,
 					 struct i915_request *rq,
 					 unsigned int flags);
 
+#define __i915_vma_flags(v) ((unsigned long *)&(v)->flags.counter)
+
 static inline bool i915_vma_is_ggtt(const struct i915_vma *vma)
 {
-	return vma->flags & I915_VMA_GGTT;
+	return test_bit(I915_VMA_GGTT_BIT, __i915_vma_flags(vma));
 }
 
 static inline bool i915_vma_has_ggtt_write(const struct i915_vma *vma)
 {
-	return vma->flags & I915_VMA_GGTT_WRITE;
+	return test_bit(I915_VMA_GGTT_WRITE_BIT, __i915_vma_flags(vma));
 }
 
 static inline void i915_vma_set_ggtt_write(struct i915_vma *vma)
 {
 	GEM_BUG_ON(!i915_vma_is_ggtt(vma));
-	vma->flags |= I915_VMA_GGTT_WRITE;
+	set_bit(I915_VMA_GGTT_WRITE_BIT, __i915_vma_flags(vma));
 }
 
-static inline void i915_vma_unset_ggtt_write(struct i915_vma *vma)
+static inline bool i915_vma_unset_ggtt_write(struct i915_vma *vma)
 {
-	vma->flags &= ~I915_VMA_GGTT_WRITE;
+	return test_and_clear_bit(I915_VMA_GGTT_WRITE_BIT,
+				  __i915_vma_flags(vma));
 }
 
 void i915_vma_flush_writes(struct i915_vma *vma);
 
 static inline bool i915_vma_is_map_and_fenceable(const struct i915_vma *vma)
 {
-	return vma->flags & I915_VMA_CAN_FENCE;
+	return test_bit(I915_VMA_CAN_FENCE_BIT, __i915_vma_flags(vma));
 }
 
 static inline bool i915_vma_set_userfault(struct i915_vma *vma)
 {
 	GEM_BUG_ON(!i915_vma_is_map_and_fenceable(vma));
-	return __test_and_set_bit(I915_VMA_USERFAULT_BIT, &vma->flags);
+	return test_and_set_bit(I915_VMA_USERFAULT_BIT, __i915_vma_flags(vma));
 }
 
 static inline void i915_vma_unset_userfault(struct i915_vma *vma)
 {
-	return __clear_bit(I915_VMA_USERFAULT_BIT, &vma->flags);
+	return clear_bit(I915_VMA_USERFAULT_BIT, __i915_vma_flags(vma));
 }
 
 static inline bool i915_vma_has_userfault(const struct i915_vma *vma)
 {
-	return test_bit(I915_VMA_USERFAULT_BIT, &vma->flags);
+	return test_bit(I915_VMA_USERFAULT_BIT, __i915_vma_flags(vma));
 }
 
 static inline bool i915_vma_is_closed(const struct i915_vma *vma)
@@ -214,7 +238,7 @@ static inline bool i915_vma_is_closed(const struct i915_vma *vma)
 static inline u32 i915_ggtt_offset(const struct i915_vma *vma)
 {
 	GEM_BUG_ON(!i915_vma_is_ggtt(vma));
-	GEM_BUG_ON(!vma->node.allocated);
+	GEM_BUG_ON(!drm_mm_node_allocated(&vma->node));
 	GEM_BUG_ON(upper_32_bits(vma->node.start));
 	GEM_BUG_ON(upper_32_bits(vma->node.start + vma->node.size - 1));
 	return lower_32_bits(vma->node.start);
@@ -293,13 +317,18 @@ i915_vma_compare(struct i915_vma *vma,
 	return memcmp(&vma->ggtt_view.partial, &view->partial, view->type);
 }
 
-int i915_vma_bind(struct i915_vma *vma, enum i915_cache_level cache_level,
-		  u32 flags);
-bool i915_gem_valid_gtt_space(struct i915_vma *vma, unsigned long cache_level);
+struct i915_vma_work *i915_vma_work(void);
+int i915_vma_bind(struct i915_vma *vma,
+		  enum i915_cache_level cache_level,
+		  u32 flags,
+		  struct i915_vma_work *work);
+
+bool i915_gem_valid_gtt_space(struct i915_vma *vma, unsigned long color);
 bool i915_vma_misplaced(const struct i915_vma *vma,
 			u64 size, u64 alignment, u64 flags);
 void __i915_vma_set_map_and_fenceable(struct i915_vma *vma);
 void i915_vma_revoke_mmap(struct i915_vma *vma);
+int __i915_vma_unbind(struct i915_vma *vma);
 int __must_check i915_vma_unbind(struct i915_vma *vma);
 void i915_vma_unlink_ctx(struct i915_vma *vma);
 void i915_vma_close(struct i915_vma *vma);
@@ -318,30 +347,12 @@ static inline void i915_vma_unlock(struct i915_vma *vma)
 	dma_resv_unlock(vma->resv);
 }
 
-int __i915_vma_do_pin(struct i915_vma *vma,
-		      u64 size, u64 alignment, u64 flags);
-static inline int __must_check
-i915_vma_pin(struct i915_vma *vma, u64 size, u64 alignment, u64 flags)
-{
-	BUILD_BUG_ON(PIN_MBZ != I915_VMA_PIN_OVERFLOW);
-	BUILD_BUG_ON(PIN_GLOBAL != I915_VMA_GLOBAL_BIND);
-	BUILD_BUG_ON(PIN_USER != I915_VMA_LOCAL_BIND);
-
-	/* Pin early to prevent the shrinker/eviction logic from destroying
-	 * our vma as we insert and bind.
-	 */
-	if (likely(((++vma->flags ^ flags) & I915_VMA_BIND_MASK) == 0)) {
-		GEM_BUG_ON(!drm_mm_node_allocated(&vma->node));
-		GEM_BUG_ON(i915_vma_misplaced(vma, size, alignment, flags));
-		return 0;
-	}
-
-	return __i915_vma_do_pin(vma, size, alignment, flags);
-}
+int __must_check
+i915_vma_pin(struct i915_vma *vma, u64 size, u64 alignment, u64 flags);
 
 static inline int i915_vma_pin_count(const struct i915_vma *vma)
 {
-	return vma->flags & I915_VMA_PIN_MASK;
+	return atomic_read(&vma->flags) & I915_VMA_PIN_MASK;
 }
 
 static inline bool i915_vma_is_pinned(const struct i915_vma *vma)
@@ -351,18 +362,18 @@ static inline bool i915_vma_is_pinned(const struct i915_vma *vma)
 
 static inline void __i915_vma_pin(struct i915_vma *vma)
 {
-	vma->flags++;
-	GEM_BUG_ON(vma->flags & I915_VMA_PIN_OVERFLOW);
+	atomic_inc(&vma->flags);
+	GEM_BUG_ON(!i915_vma_is_pinned(vma));
 }
 
 static inline void __i915_vma_unpin(struct i915_vma *vma)
 {
-	vma->flags--;
+	GEM_BUG_ON(!i915_vma_is_pinned(vma));
+	atomic_dec(&vma->flags);
 }
 
 static inline void i915_vma_unpin(struct i915_vma *vma)
 {
-	GEM_BUG_ON(!i915_vma_is_pinned(vma));
 	GEM_BUG_ON(!drm_mm_node_allocated(&vma->node));
 	__i915_vma_unpin(vma);
 }
@@ -370,7 +381,13 @@ static inline void i915_vma_unpin(struct i915_vma *vma)
 static inline bool i915_vma_is_bound(const struct i915_vma *vma,
 				     unsigned int where)
 {
-	return vma->flags & where;
+	return atomic_read(&vma->flags) & where;
+}
+
+static inline bool i915_node_color_differs(const struct drm_mm_node *node,
+					   unsigned long color)
+{
+	return drm_mm_node_allocated(node) && node->color != color;
 }
 
 /**
@@ -382,8 +399,6 @@ static inline bool i915_vma_is_bound(const struct i915_vma *vma,
  * the caller must call i915_vma_unpin_iomap to relinquish the pinning
  * after the iomapping is no longer required.
  *
- * Callers must hold the struct_mutex.
- *
  * Returns a valid iomapped pointer or ERR_PTR.
  */
 void __iomem *i915_vma_pin_iomap(struct i915_vma *vma);
@@ -395,8 +410,8 @@ void __iomem *i915_vma_pin_iomap(struct i915_vma *vma);
  *
  * Unpins the previously iomapped VMA from i915_vma_pin_iomap().
  *
- * Callers must hold the struct_mutex. This function is only valid to be
- * called on a VMA previously iomapped by the caller with i915_vma_pin_iomap().
+ * This function is only valid to be called on a VMA previously
+ * iomapped by the caller with i915_vma_pin_iomap().
  */
 void i915_vma_unpin_iomap(struct i915_vma *vma);
 
@@ -424,6 +439,8 @@ static inline struct page *i915_vma_first_page(struct i915_vma *vma)
 int __must_check i915_vma_pin_fence(struct i915_vma *vma);
 int __must_check i915_vma_revoke_fence(struct i915_vma *vma);
 
+int __i915_vma_pin_fence(struct i915_vma *vma);
+
 static inline void __i915_vma_unpin_fence(struct i915_vma *vma)
 {
 	GEM_BUG_ON(atomic_read(&vma->fence->pin_count) <= 0);
@@ -441,12 +458,11 @@ static inline void __i915_vma_unpin_fence(struct i915_vma *vma)
 static inline void
 i915_vma_unpin_fence(struct i915_vma *vma)
 {
-	/* lockdep_assert_held(&vma->vm->i915->drm.struct_mutex); */
 	if (vma->fence)
 		__i915_vma_unpin_fence(vma);
 }
 
-void i915_vma_parked(struct drm_i915_private *i915);
+void i915_vma_parked(struct intel_gt *gt);
 
 #define for_each_until(cond) if (cond) break; else
 
@@ -470,4 +486,10 @@ struct i915_vma *i915_vma_make_unshrinkable(struct i915_vma *vma);
 void i915_vma_make_shrinkable(struct i915_vma *vma);
 void i915_vma_make_purgeable(struct i915_vma *vma);
 
+static inline int i915_vma_sync(struct i915_vma *vma)
+{
+	/* Wait for the asynchronous bindings and pending GPU reads */
+	return i915_active_wait(&vma->active);
+}
+
 #endif
diff --git a/drivers/gpu/drm/i915/intel_csr.c b/drivers/gpu/drm/i915/intel_csr.c
index 546577e39b4e..09870a31b4f0 100644
--- a/drivers/gpu/drm/i915/intel_csr.c
+++ b/drivers/gpu/drm/i915/intel_csr.c
@@ -44,8 +44,8 @@
 #define TGL_CSR_MAX_FW_SIZE		0x6000
 MODULE_FIRMWARE(TGL_CSR_PATH);
 
-#define ICL_CSR_PATH			"i915/icl_dmc_ver1_07.bin"
-#define ICL_CSR_VERSION_REQUIRED	CSR_VERSION(1, 7)
+#define ICL_CSR_PATH			"i915/icl_dmc_ver1_09.bin"
+#define ICL_CSR_VERSION_REQUIRED	CSR_VERSION(1, 9)
 #define ICL_CSR_MAX_FW_SIZE		0x6000
 MODULE_FIRMWARE(ICL_CSR_PATH);
 
diff --git a/drivers/gpu/drm/i915/intel_device_info.c b/drivers/gpu/drm/i915/intel_device_info.c
index d0ed44d33484..a5b571364cf6 100644
--- a/drivers/gpu/drm/i915/intel_device_info.c
+++ b/drivers/gpu/drm/i915/intel_device_info.c
@@ -93,9 +93,9 @@ static void sseu_dump(const struct sseu_dev_info *sseu, struct drm_printer *p)
 		   hweight8(sseu->slice_mask), sseu->slice_mask);
 	drm_printf(p, "subslice total: %u\n", intel_sseu_subslice_total(sseu));
 	for (s = 0; s < sseu->max_slices; s++) {
-		drm_printf(p, "slice%d: %u subslices, mask=%04x\n",
+		drm_printf(p, "slice%d: %u subslices, mask=%08x\n",
 			   s, intel_sseu_subslices_per_slice(sseu, s),
-			   sseu->subslice_mask[s]);
+			   intel_sseu_get_subslices(sseu, s));
 	}
 	drm_printf(p, "EU total: %u\n", sseu->eu_total);
 	drm_printf(p, "EU per subslice: %u\n", sseu->eu_per_subslice);
@@ -118,10 +118,9 @@ void intel_device_info_dump_runtime(const struct intel_runtime_info *info,
 static int sseu_eu_idx(const struct sseu_dev_info *sseu, int slice,
 		       int subslice)
 {
-	int subslice_stride = GEN_SSEU_STRIDE(sseu->max_eus_per_subslice);
-	int slice_stride = sseu->max_subslices * subslice_stride;
+	int slice_stride = sseu->max_subslices * sseu->eu_stride;
 
-	return slice * slice_stride + subslice * subslice_stride;
+	return slice * slice_stride + subslice * sseu->eu_stride;
 }
 
 static u16 sseu_get_eus(const struct sseu_dev_info *sseu, int slice,
@@ -130,7 +129,7 @@ static u16 sseu_get_eus(const struct sseu_dev_info *sseu, int slice,
 	int i, offset = sseu_eu_idx(sseu, slice, subslice);
 	u16 eu_mask = 0;
 
-	for (i = 0; i < GEN_SSEU_STRIDE(sseu->max_eus_per_subslice); i++) {
+	for (i = 0; i < sseu->eu_stride; i++) {
 		eu_mask |= ((u16)sseu->eu_mask[offset + i]) <<
 			(i * BITS_PER_BYTE);
 	}
@@ -143,7 +142,7 @@ static void sseu_set_eus(struct sseu_dev_info *sseu, int slice, int subslice,
 {
 	int i, offset = sseu_eu_idx(sseu, slice, subslice);
 
-	for (i = 0; i < GEN_SSEU_STRIDE(sseu->max_eus_per_subslice); i++) {
+	for (i = 0; i < sseu->eu_stride; i++) {
 		sseu->eu_mask[offset + i] =
 			(eu_mask >> (BITS_PER_BYTE * i)) & 0xff;
 	}
@@ -160,9 +159,9 @@ void intel_device_info_dump_topology(const struct sseu_dev_info *sseu,
 	}
 
 	for (s = 0; s < sseu->max_slices; s++) {
-		drm_printf(p, "slice%d: %u subslice(s) (0x%hhx):\n",
+		drm_printf(p, "slice%d: %u subslice(s) (0x%08x):\n",
 			   s, intel_sseu_subslices_per_slice(sseu, s),
-			   sseu->subslice_mask[s]);
+			   intel_sseu_get_subslices(sseu, s));
 
 		for (ss = 0; ss < sseu->max_subslices; ss++) {
 			u16 enabled_eus = sseu_get_eus(sseu, s, ss);
@@ -183,44 +182,80 @@ static u16 compute_eu_total(const struct sseu_dev_info *sseu)
 	return total;
 }
 
+static void gen11_compute_sseu_info(struct sseu_dev_info *sseu,
+				    u8 s_en, u32 ss_en, u16 eu_en)
+{
+	int s, ss;
+
+	/* ss_en represents entire subslice mask across all slices */
+	GEM_BUG_ON(sseu->max_slices * sseu->max_subslices >
+		   sizeof(ss_en) * BITS_PER_BYTE);
+
+	for (s = 0; s < sseu->max_slices; s++) {
+		if ((s_en & BIT(s)) == 0)
+			continue;
+
+		sseu->slice_mask |= BIT(s);
+
+		intel_sseu_set_subslices(sseu, s, ss_en);
+
+		for (ss = 0; ss < sseu->max_subslices; ss++)
+			if (intel_sseu_has_subslice(sseu, s, ss))
+				sseu_set_eus(sseu, s, ss, eu_en);
+	}
+	sseu->eu_per_subslice = hweight16(eu_en);
+	sseu->eu_total = compute_eu_total(sseu);
+}
+
+static void gen12_sseu_info_init(struct drm_i915_private *dev_priv)
+{
+	struct sseu_dev_info *sseu = &RUNTIME_INFO(dev_priv)->sseu;
+	u8 s_en;
+	u32 dss_en;
+	u16 eu_en = 0;
+	u8 eu_en_fuse;
+	int eu;
+
+	/*
+	 * Gen12 has Dual-Subslices, which behave similarly to 2 gen11 SS.
+	 * Instead of splitting these, provide userspace with an array
+	 * of DSS to more closely represent the hardware resource.
+	 */
+	intel_sseu_set_info(sseu, 1, 6, 16);
+
+	s_en = I915_READ(GEN11_GT_SLICE_ENABLE) & GEN11_GT_S_ENA_MASK;
+
+	dss_en = I915_READ(GEN12_GT_DSS_ENABLE);
+
+	/* one bit per pair of EUs */
+	eu_en_fuse = ~(I915_READ(GEN11_EU_DISABLE) & GEN11_EU_DIS_MASK);
+	for (eu = 0; eu < sseu->max_eus_per_subslice / 2; eu++)
+		if (eu_en_fuse & BIT(eu))
+			eu_en |= BIT(eu * 2) | BIT(eu * 2 + 1);
+
+	gen11_compute_sseu_info(sseu, s_en, dss_en, eu_en);
+
+	/* TGL only supports slice-level power gating */
+	sseu->has_slice_pg = 1;
+}
+
 static void gen11_sseu_info_init(struct drm_i915_private *dev_priv)
 {
 	struct sseu_dev_info *sseu = &RUNTIME_INFO(dev_priv)->sseu;
 	u8 s_en;
-	u32 ss_en, ss_en_mask;
+	u32 ss_en;
 	u8 eu_en;
-	int s;
 
-	if (IS_ELKHARTLAKE(dev_priv)) {
-		sseu->max_slices = 1;
-		sseu->max_subslices = 4;
-		sseu->max_eus_per_subslice = 8;
-	} else {
-		sseu->max_slices = 1;
-		sseu->max_subslices = 8;
-		sseu->max_eus_per_subslice = 8;
-	}
+	if (IS_ELKHARTLAKE(dev_priv))
+		intel_sseu_set_info(sseu, 1, 4, 8);
+	else
+		intel_sseu_set_info(sseu, 1, 8, 8);
 
 	s_en = I915_READ(GEN11_GT_SLICE_ENABLE) & GEN11_GT_S_ENA_MASK;
 	ss_en = ~I915_READ(GEN11_GT_SUBSLICE_DISABLE);
-	ss_en_mask = BIT(sseu->max_subslices) - 1;
 	eu_en = ~(I915_READ(GEN11_EU_DISABLE) & GEN11_EU_DIS_MASK);
 
-	for (s = 0; s < sseu->max_slices; s++) {
-		if (s_en & BIT(s)) {
-			int ss_idx = sseu->max_subslices * s;
-			int ss;
-
-			sseu->slice_mask |= BIT(s);
-			sseu->subslice_mask[s] = (ss_en >> ss_idx) & ss_en_mask;
-			for (ss = 0; ss < sseu->max_subslices; ss++) {
-				if (sseu->subslice_mask[s] & BIT(ss))
-					sseu_set_eus(sseu, s, ss, eu_en);
-			}
-		}
-	}
-	sseu->eu_per_subslice = hweight8(eu_en);
-	sseu->eu_total = compute_eu_total(sseu);
+	gen11_compute_sseu_info(sseu, s_en, ss_en, eu_en);
 
 	/* ICL has no power gating restrictions. */
 	sseu->has_slice_pg = 1;
@@ -236,23 +271,10 @@ static void gen10_sseu_info_init(struct drm_i915_private *dev_priv)
 	const int eu_mask = 0xff;
 	u32 subslice_mask, eu_en;
 
+	intel_sseu_set_info(sseu, 6, 4, 8);
+
 	sseu->slice_mask = (fuse2 & GEN10_F2_S_ENA_MASK) >>
 			    GEN10_F2_S_ENA_SHIFT;
-	sseu->max_slices = 6;
-	sseu->max_subslices = 4;
-	sseu->max_eus_per_subslice = 8;
-
-	subslice_mask = (1 << 4) - 1;
-	subslice_mask &= ~((fuse2 & GEN10_F2_SS_DIS_MASK) >>
-			   GEN10_F2_SS_DIS_SHIFT);
-
-	/*
-	 * Slice0 can have up to 3 subslices, but there are only 2 in
-	 * slice1/2.
-	 */
-	sseu->subslice_mask[0] = subslice_mask;
-	for (s = 1; s < sseu->max_slices; s++)
-		sseu->subslice_mask[s] = subslice_mask & 0x3;
 
 	/* Slice0 */
 	eu_en = ~I915_READ(GEN8_EU_DISABLE0);
@@ -277,14 +299,25 @@ static void gen10_sseu_info_init(struct drm_i915_private *dev_priv)
 	eu_en = ~I915_READ(GEN10_EU_DISABLE3);
 	sseu_set_eus(sseu, 5, 1, eu_en & eu_mask);
 
-	/* Do a second pass where we mark the subslices disabled if all their
-	 * eus are off.
-	 */
+	subslice_mask = (1 << 4) - 1;
+	subslice_mask &= ~((fuse2 & GEN10_F2_SS_DIS_MASK) >>
+			   GEN10_F2_SS_DIS_SHIFT);
+
 	for (s = 0; s < sseu->max_slices; s++) {
+		u32 subslice_mask_with_eus = subslice_mask;
+
 		for (ss = 0; ss < sseu->max_subslices; ss++) {
 			if (sseu_get_eus(sseu, s, ss) == 0)
-				sseu->subslice_mask[s] &= ~BIT(ss);
+				subslice_mask_with_eus &= ~BIT(ss);
 		}
+
+		/*
+		 * Slice0 can have up to 3 subslices, but there are only 2 in
+		 * slice1/2.
+		 */
+		intel_sseu_set_subslices(sseu, s, s == 0 ?
+						  subslice_mask_with_eus :
+						  subslice_mask_with_eus & 0x3);
 	}
 
 	sseu->eu_total = compute_eu_total(sseu);
@@ -310,13 +343,12 @@ static void cherryview_sseu_info_init(struct drm_i915_private *dev_priv)
 {
 	struct sseu_dev_info *sseu = &RUNTIME_INFO(dev_priv)->sseu;
 	u32 fuse;
+	u8 subslice_mask = 0;
 
 	fuse = I915_READ(CHV_FUSE_GT);
 
 	sseu->slice_mask = BIT(0);
-	sseu->max_slices = 1;
-	sseu->max_subslices = 2;
-	sseu->max_eus_per_subslice = 8;
+	intel_sseu_set_info(sseu, 1, 2, 8);
 
 	if (!(fuse & CHV_FGT_DISABLE_SS0)) {
 		u8 disabled_mask =
@@ -325,7 +357,7 @@ static void cherryview_sseu_info_init(struct drm_i915_private *dev_priv)
 			(((fuse & CHV_FGT_EU_DIS_SS0_R1_MASK) >>
 			  CHV_FGT_EU_DIS_SS0_R1_SHIFT) << 4);
 
-		sseu->subslice_mask[0] |= BIT(0);
+		subslice_mask |= BIT(0);
 		sseu_set_eus(sseu, 0, 0, ~disabled_mask);
 	}
 
@@ -336,10 +368,12 @@ static void cherryview_sseu_info_init(struct drm_i915_private *dev_priv)
 			(((fuse & CHV_FGT_EU_DIS_SS1_R1_MASK) >>
 			  CHV_FGT_EU_DIS_SS1_R1_SHIFT) << 4);
 
-		sseu->subslice_mask[0] |= BIT(1);
+		subslice_mask |= BIT(1);
 		sseu_set_eus(sseu, 0, 1, ~disabled_mask);
 	}
 
+	intel_sseu_set_subslices(sseu, 0, subslice_mask);
+
 	sseu->eu_total = compute_eu_total(sseu);
 
 	/*
@@ -372,9 +406,8 @@ static void gen9_sseu_info_init(struct drm_i915_private *dev_priv)
 	sseu->slice_mask = (fuse2 & GEN8_F2_S_ENA_MASK) >> GEN8_F2_S_ENA_SHIFT;
 
 	/* BXT has a single slice and at most 3 subslices. */
-	sseu->max_slices = IS_GEN9_LP(dev_priv) ? 1 : 3;
-	sseu->max_subslices = IS_GEN9_LP(dev_priv) ? 3 : 4;
-	sseu->max_eus_per_subslice = 8;
+	intel_sseu_set_info(sseu, IS_GEN9_LP(dev_priv) ? 1 : 3,
+			    IS_GEN9_LP(dev_priv) ? 3 : 4, 8);
 
 	/*
 	 * The subslice disable field is global, i.e. it applies
@@ -393,14 +426,14 @@ static void gen9_sseu_info_init(struct drm_i915_private *dev_priv)
 			/* skip disabled slice */
 			continue;
 
-		sseu->subslice_mask[s] = subslice_mask;
+		intel_sseu_set_subslices(sseu, s, subslice_mask);
 
 		eu_disable = I915_READ(GEN9_EU_DISABLE(s));
 		for (ss = 0; ss < sseu->max_subslices; ss++) {
 			int eu_per_ss;
 			u8 eu_disabled_mask;
 
-			if (!(sseu->subslice_mask[s] & BIT(ss)))
+			if (!intel_sseu_has_subslice(sseu, s, ss))
 				/* skip disabled subslice */
 				continue;
 
@@ -473,9 +506,7 @@ static void broadwell_sseu_info_init(struct drm_i915_private *dev_priv)
 
 	fuse2 = I915_READ(GEN8_FUSE2);
 	sseu->slice_mask = (fuse2 & GEN8_F2_S_ENA_MASK) >> GEN8_F2_S_ENA_SHIFT;
-	sseu->max_slices = 3;
-	sseu->max_subslices = 3;
-	sseu->max_eus_per_subslice = 8;
+	intel_sseu_set_info(sseu, 3, 3, 8);
 
 	/*
 	 * The subslice disable field is global, i.e. it applies
@@ -502,13 +533,13 @@ static void broadwell_sseu_info_init(struct drm_i915_private *dev_priv)
 			/* skip disabled slice */
 			continue;
 
-		sseu->subslice_mask[s] = subslice_mask;
+		intel_sseu_set_subslices(sseu, s, subslice_mask);
 
 		for (ss = 0; ss < sseu->max_subslices; ss++) {
 			u8 eu_disabled_mask;
 			u32 n_disabled;
 
-			if (!(sseu->subslice_mask[s] & BIT(ss)))
+			if (!intel_sseu_has_subslice(sseu, s, ss))
 				/* skip disabled subslice */
 				continue;
 
@@ -552,6 +583,7 @@ static void haswell_sseu_info_init(struct drm_i915_private *dev_priv)
 {
 	struct sseu_dev_info *sseu = &RUNTIME_INFO(dev_priv)->sseu;
 	u32 fuse1;
+	u8 subslice_mask = 0;
 	int s, ss;
 
 	/*
@@ -564,22 +596,18 @@ static void haswell_sseu_info_init(struct drm_i915_private *dev_priv)
 		/* fall through */
 	case 1:
 		sseu->slice_mask = BIT(0);
-		sseu->subslice_mask[0] = BIT(0);
+		subslice_mask = BIT(0);
 		break;
 	case 2:
 		sseu->slice_mask = BIT(0);
-		sseu->subslice_mask[0] = BIT(0) | BIT(1);
+		subslice_mask = BIT(0) | BIT(1);
 		break;
 	case 3:
 		sseu->slice_mask = BIT(0) | BIT(1);
-		sseu->subslice_mask[0] = BIT(0) | BIT(1);
-		sseu->subslice_mask[1] = BIT(0) | BIT(1);
+		subslice_mask = BIT(0) | BIT(1);
 		break;
 	}
 
-	sseu->max_slices = hweight8(sseu->slice_mask);
-	sseu->max_subslices = hweight8(sseu->subslice_mask[0]);
-
 	fuse1 = I915_READ(HSW_PAVP_FUSE1);
 	switch ((fuse1 & HSW_F1_EU_DIS_MASK) >> HSW_F1_EU_DIS_SHIFT) {
 	default:
@@ -596,9 +624,14 @@ static void haswell_sseu_info_init(struct drm_i915_private *dev_priv)
 		sseu->eu_per_subslice = 6;
 		break;
 	}
-	sseu->max_eus_per_subslice = sseu->eu_per_subslice;
+
+	intel_sseu_set_info(sseu, hweight8(sseu->slice_mask),
+			    hweight8(subslice_mask),
+			    sseu->eu_per_subslice);
 
 	for (s = 0; s < sseu->max_slices; s++) {
+		intel_sseu_set_subslices(sseu, s, subslice_mask);
+
 		for (ss = 0; ss < sseu->max_subslices; ss++) {
 			sseu_set_eus(sseu, s, ss,
 				     (1UL << sseu->eu_per_subslice) - 1);
@@ -900,12 +933,8 @@ void intel_device_info_runtime_init(struct drm_i915_private *dev_priv)
 			runtime->num_sprites[pipe] = 1;
 	}
 
-	if (i915_modparams.disable_display) {
-		DRM_INFO("Display disabled (module parameter)\n");
-		info->num_pipes = 0;
-	} else if (HAS_DISPLAY(dev_priv) &&
-		   (IS_GEN_RANGE(dev_priv, 7, 8)) &&
-		   HAS_PCH_SPLIT(dev_priv)) {
+	if (HAS_DISPLAY(dev_priv) && IS_GEN_RANGE(dev_priv, 7, 8) &&
+	    HAS_PCH_SPLIT(dev_priv)) {
 		u32 fuse_strap = I915_READ(FUSE_STRAP);
 		u32 sfuse_strap = I915_READ(SFUSE_STRAP);
 
@@ -923,14 +952,14 @@ void intel_device_info_runtime_init(struct drm_i915_private *dev_priv)
 		    (HAS_PCH_CPT(dev_priv) &&
 		     !(sfuse_strap & SFUSE_STRAP_FUSE_LOCK))) {
 			DRM_INFO("Display fused off, disabling\n");
-			info->num_pipes = 0;
+			info->pipe_mask = 0;
 		} else if (fuse_strap & IVB_PIPE_C_DISABLE) {
 			DRM_INFO("PipeC fused off\n");
-			info->num_pipes -= 1;
+			info->pipe_mask &= ~BIT(PIPE_C);
 		}
 	} else if (HAS_DISPLAY(dev_priv) && INTEL_GEN(dev_priv) >= 9) {
 		u32 dfsm = I915_READ(SKL_DFSM);
-		u8 enabled_mask = BIT(info->num_pipes) - 1;
+		u8 enabled_mask = info->pipe_mask;
 
 		if (dfsm & SKL_DFSM_PIPE_A_DISABLE)
 			enabled_mask &= ~BIT(PIPE_A);
@@ -951,7 +980,20 @@ void intel_device_info_runtime_init(struct drm_i915_private *dev_priv)
 			DRM_ERROR("invalid pipe fuse configuration: enabled_mask=0x%x\n",
 				  enabled_mask);
 		else
-			info->num_pipes = hweight8(enabled_mask);
+			info->pipe_mask = enabled_mask;
+
+		if (dfsm & SKL_DFSM_DISPLAY_HDCP_DISABLE)
+			info->display.has_hdcp = 0;
+
+		if (dfsm & SKL_DFSM_DISPLAY_PM_DISABLE)
+			info->display.has_fbc = 0;
+
+		if (INTEL_GEN(dev_priv) >= 11 && (dfsm & ICL_DFSM_DMC_DISABLE))
+			info->display.has_csr = 0;
+
+		if (INTEL_GEN(dev_priv) >= 10 &&
+		    (dfsm & CNL_DFSM_DISPLAY_DSC_DISABLE))
+			info->display.has_dsc = 0;
 	}
 
 	/* Initialize slice/subslice/EU info */
@@ -965,8 +1007,10 @@ void intel_device_info_runtime_init(struct drm_i915_private *dev_priv)
 		gen9_sseu_info_init(dev_priv);
 	else if (IS_GEN(dev_priv, 10))
 		gen10_sseu_info_init(dev_priv);
-	else if (INTEL_GEN(dev_priv) >= 11)
+	else if (IS_GEN(dev_priv, 11))
 		gen11_sseu_info_init(dev_priv);
+	else if (INTEL_GEN(dev_priv) >= 12)
+		gen12_sseu_info_init(dev_priv);
 
 	if (IS_GEN(dev_priv, 6) && intel_vtd_active()) {
 		DRM_INFO("Disabling ppGTT for VT-d support\n");
@@ -1010,8 +1054,10 @@ void intel_device_info_init_mmio(struct drm_i915_private *dev_priv)
 		      GEN11_GT_VEBOX_DISABLE_SHIFT;
 
 	for (i = 0; i < I915_MAX_VCS; i++) {
-		if (!HAS_ENGINE(dev_priv, _VCS(i)))
+		if (!HAS_ENGINE(dev_priv, _VCS(i))) {
+			vdbox_mask &= ~BIT(i);
 			continue;
+		}
 
 		if (!(BIT(i) & vdbox_mask)) {
 			info->engine_mask &= ~BIT(_VCS(i));
@@ -1032,8 +1078,10 @@ void intel_device_info_init_mmio(struct drm_i915_private *dev_priv)
 	GEM_BUG_ON(vdbox_mask != VDBOX_MASK(dev_priv));
 
 	for (i = 0; i < I915_MAX_VECS; i++) {
-		if (!HAS_ENGINE(dev_priv, _VECS(i)))
+		if (!HAS_ENGINE(dev_priv, _VECS(i))) {
+			vebox_mask &= ~BIT(i);
 			continue;
+		}
 
 		if (!(BIT(i) & vebox_mask)) {
 			info->engine_mask &= ~BIT(_VECS(i));
diff --git a/drivers/gpu/drm/i915/intel_device_info.h b/drivers/gpu/drm/i915/intel_device_info.h
index 92e0c2e0954c..4bdf8a6cfb47 100644
--- a/drivers/gpu/drm/i915/intel_device_info.h
+++ b/drivers/gpu/drm/i915/intel_device_info.h
@@ -107,6 +107,7 @@ enum intel_ppgtt_type {
 	func(is_mobile); \
 	func(is_lp); \
 	func(require_force_probe); \
+	func(is_dgfx); \
 	/* Keep has_* in alphabetical order */ \
 	func(has_64bit_reloc); \
 	func(gpu_reset_clobbers_display); \
@@ -135,8 +136,11 @@ enum intel_ppgtt_type {
 	func(has_csr); \
 	func(has_ddi); \
 	func(has_dp_mst); \
+	func(has_dsb); \
+	func(has_dsc); \
 	func(has_fbc); \
 	func(has_gmch); \
+	func(has_hdcp); \
 	func(has_hotplug); \
 	func(has_ipc); \
 	func(has_modular_fia); \
@@ -159,9 +163,11 @@ struct intel_device_info {
 
 	unsigned int page_sizes; /* page sizes supported by the HW */
 
+	u32 memory_regions; /* regions supported by the HW */
+
 	u32 display_mmio_offset;
 
-	u8 num_pipes;
+	u8 pipe_mask;
 
 #define DEFINE_FLAG(name) u8 name:1
 	DEV_INFO_FOR_EACH_FLAG(DEFINE_FLAG);
diff --git a/drivers/gpu/drm/i915/intel_memory_region.c b/drivers/gpu/drm/i915/intel_memory_region.c
new file mode 100644
index 000000000000..baaeaecc64af
--- /dev/null
+++ b/drivers/gpu/drm/i915/intel_memory_region.c
@@ -0,0 +1,272 @@
+// SPDX-License-Identifier: MIT
+/*
+ * Copyright © 2019 Intel Corporation
+ */
+
+#include "intel_memory_region.h"
+#include "i915_drv.h"
+
+/* XXX: Hysterical raisins. BIT(inst) needs to just be (inst) at some point. */
+#define REGION_MAP(type, inst) \
+	BIT((type) + INTEL_MEMORY_TYPE_SHIFT) | BIT(inst)
+
+const u32 intel_region_map[] = {
+	[INTEL_REGION_SMEM] = REGION_MAP(INTEL_MEMORY_SYSTEM, 0),
+	[INTEL_REGION_LMEM] = REGION_MAP(INTEL_MEMORY_LOCAL, 0),
+	[INTEL_REGION_STOLEN] = REGION_MAP(INTEL_MEMORY_STOLEN, 0),
+};
+
+static u64
+intel_memory_region_free_pages(struct intel_memory_region *mem,
+			       struct list_head *blocks)
+{
+	struct i915_buddy_block *block, *on;
+	u64 size = 0;
+
+	list_for_each_entry_safe(block, on, blocks, link) {
+		size += i915_buddy_block_size(&mem->mm, block);
+		i915_buddy_free(&mem->mm, block);
+	}
+	INIT_LIST_HEAD(blocks);
+
+	return size;
+}
+
+void
+__intel_memory_region_put_pages_buddy(struct intel_memory_region *mem,
+				      struct list_head *blocks)
+{
+	mutex_lock(&mem->mm_lock);
+	intel_memory_region_free_pages(mem, blocks);
+	mutex_unlock(&mem->mm_lock);
+}
+
+void
+__intel_memory_region_put_block_buddy(struct i915_buddy_block *block)
+{
+	struct list_head blocks;
+
+	INIT_LIST_HEAD(&blocks);
+	list_add(&block->link, &blocks);
+	__intel_memory_region_put_pages_buddy(block->private, &blocks);
+}
+
+int
+__intel_memory_region_get_pages_buddy(struct intel_memory_region *mem,
+				      resource_size_t size,
+				      unsigned int flags,
+				      struct list_head *blocks)
+{
+	unsigned int min_order = 0;
+	unsigned long n_pages;
+
+	GEM_BUG_ON(!IS_ALIGNED(size, mem->mm.chunk_size));
+	GEM_BUG_ON(!list_empty(blocks));
+
+	if (flags & I915_ALLOC_MIN_PAGE_SIZE) {
+		min_order = ilog2(mem->min_page_size) -
+			    ilog2(mem->mm.chunk_size);
+	}
+
+	if (flags & I915_ALLOC_CONTIGUOUS) {
+		size = roundup_pow_of_two(size);
+		min_order = ilog2(size) - ilog2(mem->mm.chunk_size);
+	}
+
+	n_pages = size >> ilog2(mem->mm.chunk_size);
+
+	mutex_lock(&mem->mm_lock);
+
+	do {
+		struct i915_buddy_block *block;
+		unsigned int order;
+
+		order = fls(n_pages) - 1;
+		GEM_BUG_ON(order > mem->mm.max_order);
+		GEM_BUG_ON(order < min_order);
+
+		do {
+			block = i915_buddy_alloc(&mem->mm, order);
+			if (!IS_ERR(block))
+				break;
+
+			if (order-- == min_order)
+				goto err_free_blocks;
+		} while (1);
+
+		n_pages -= BIT(order);
+
+		block->private = mem;
+		list_add(&block->link, blocks);
+
+		if (!n_pages)
+			break;
+	} while (1);
+
+	mutex_unlock(&mem->mm_lock);
+	return 0;
+
+err_free_blocks:
+	intel_memory_region_free_pages(mem, blocks);
+	mutex_unlock(&mem->mm_lock);
+	return -ENXIO;
+}
+
+struct i915_buddy_block *
+__intel_memory_region_get_block_buddy(struct intel_memory_region *mem,
+				      resource_size_t size,
+				      unsigned int flags)
+{
+	struct i915_buddy_block *block;
+	LIST_HEAD(blocks);
+	int ret;
+
+	ret = __intel_memory_region_get_pages_buddy(mem, size, flags, &blocks);
+	if (ret)
+		return ERR_PTR(ret);
+
+	block = list_first_entry(&blocks, typeof(*block), link);
+	list_del_init(&block->link);
+	return block;
+}
+
+int intel_memory_region_init_buddy(struct intel_memory_region *mem)
+{
+	return i915_buddy_init(&mem->mm, resource_size(&mem->region),
+			       PAGE_SIZE);
+}
+
+void intel_memory_region_release_buddy(struct intel_memory_region *mem)
+{
+	i915_buddy_fini(&mem->mm);
+}
+
+struct intel_memory_region *
+intel_memory_region_create(struct drm_i915_private *i915,
+			   resource_size_t start,
+			   resource_size_t size,
+			   resource_size_t min_page_size,
+			   resource_size_t io_start,
+			   const struct intel_memory_region_ops *ops)
+{
+	struct intel_memory_region *mem;
+	int err;
+
+	mem = kzalloc(sizeof(*mem), GFP_KERNEL);
+	if (!mem)
+		return ERR_PTR(-ENOMEM);
+
+	mem->i915 = i915;
+	mem->region = (struct resource)DEFINE_RES_MEM(start, size);
+	mem->io_start = io_start;
+	mem->min_page_size = min_page_size;
+	mem->ops = ops;
+
+	mutex_init(&mem->objects.lock);
+	INIT_LIST_HEAD(&mem->objects.list);
+	INIT_LIST_HEAD(&mem->objects.purgeable);
+
+	mutex_init(&mem->mm_lock);
+
+	if (ops->init) {
+		err = ops->init(mem);
+		if (err)
+			goto err_free;
+	}
+
+	kref_init(&mem->kref);
+	return mem;
+
+err_free:
+	kfree(mem);
+	return ERR_PTR(err);
+}
+
+static void __intel_memory_region_destroy(struct kref *kref)
+{
+	struct intel_memory_region *mem =
+		container_of(kref, typeof(*mem), kref);
+
+	if (mem->ops->release)
+		mem->ops->release(mem);
+
+	mutex_destroy(&mem->mm_lock);
+	mutex_destroy(&mem->objects.lock);
+	kfree(mem);
+}
+
+struct intel_memory_region *
+intel_memory_region_get(struct intel_memory_region *mem)
+{
+	kref_get(&mem->kref);
+	return mem;
+}
+
+void intel_memory_region_put(struct intel_memory_region *mem)
+{
+	kref_put(&mem->kref, __intel_memory_region_destroy);
+}
+
+/* Global memory region registration -- only slight layer inversions! */
+
+int intel_memory_regions_hw_probe(struct drm_i915_private *i915)
+{
+	int err, i;
+
+	for (i = 0; i < ARRAY_SIZE(i915->mm.regions); i++) {
+		struct intel_memory_region *mem = ERR_PTR(-ENODEV);
+		u32 type;
+
+		if (!HAS_REGION(i915, BIT(i)))
+			continue;
+
+		type = MEMORY_TYPE_FROM_REGION(intel_region_map[i]);
+		switch (type) {
+		case INTEL_MEMORY_SYSTEM:
+			mem = i915_gem_shmem_setup(i915);
+			break;
+		case INTEL_MEMORY_STOLEN:
+			mem = i915_gem_stolen_setup(i915);
+			break;
+		case INTEL_MEMORY_LOCAL:
+			mem = intel_setup_fake_lmem(i915);
+			break;
+		}
+
+		if (IS_ERR(mem)) {
+			err = PTR_ERR(mem);
+			DRM_ERROR("Failed to setup region(%d) type=%d\n", err, type);
+			goto out_cleanup;
+		}
+
+		mem->id = intel_region_map[i];
+		mem->type = type;
+		mem->instance = MEMORY_INSTANCE_FROM_REGION(intel_region_map[i]);
+
+		i915->mm.regions[i] = mem;
+	}
+
+	return 0;
+
+out_cleanup:
+	intel_memory_regions_driver_release(i915);
+	return err;
+}
+
+void intel_memory_regions_driver_release(struct drm_i915_private *i915)
+{
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(i915->mm.regions); i++) {
+		struct intel_memory_region *region =
+			fetch_and_zero(&i915->mm.regions[i]);
+
+		if (region)
+			intel_memory_region_put(region);
+	}
+}
+
+#if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
+#include "selftests/intel_memory_region.c"
+#include "selftests/mock_region.c"
+#endif
diff --git a/drivers/gpu/drm/i915/intel_memory_region.h b/drivers/gpu/drm/i915/intel_memory_region.h
new file mode 100644
index 000000000000..238722009677
--- /dev/null
+++ b/drivers/gpu/drm/i915/intel_memory_region.h
@@ -0,0 +1,129 @@
+/* SPDX-License-Identifier: MIT */
+/*
+ * Copyright © 2019 Intel Corporation
+ */
+
+#ifndef __INTEL_MEMORY_REGION_H__
+#define __INTEL_MEMORY_REGION_H__
+
+#include <linux/kref.h>
+#include <linux/ioport.h>
+#include <linux/mutex.h>
+#include <linux/io-mapping.h>
+#include <drm/drm_mm.h>
+
+#include "i915_buddy.h"
+
+struct drm_i915_private;
+struct drm_i915_gem_object;
+struct intel_memory_region;
+struct sg_table;
+
+/**
+ *  Base memory type
+ */
+enum intel_memory_type {
+	INTEL_MEMORY_SYSTEM = 0,
+	INTEL_MEMORY_LOCAL,
+	INTEL_MEMORY_STOLEN,
+};
+
+enum intel_region_id {
+	INTEL_REGION_SMEM = 0,
+	INTEL_REGION_LMEM,
+	INTEL_REGION_STOLEN,
+	INTEL_REGION_UNKNOWN, /* Should be last */
+};
+
+#define REGION_SMEM     BIT(INTEL_REGION_SMEM)
+#define REGION_LMEM     BIT(INTEL_REGION_LMEM)
+#define REGION_STOLEN   BIT(INTEL_REGION_STOLEN)
+
+#define INTEL_MEMORY_TYPE_SHIFT 16
+
+#define MEMORY_TYPE_FROM_REGION(r) (ilog2((r) >> INTEL_MEMORY_TYPE_SHIFT))
+#define MEMORY_INSTANCE_FROM_REGION(r) (ilog2((r) & 0xffff))
+
+#define I915_ALLOC_MIN_PAGE_SIZE  BIT(0)
+#define I915_ALLOC_CONTIGUOUS     BIT(1)
+
+/**
+ * Memory regions encoded as type | instance
+ */
+extern const u32 intel_region_map[];
+
+struct intel_memory_region_ops {
+	unsigned int flags;
+
+	int (*init)(struct intel_memory_region *mem);
+	void (*release)(struct intel_memory_region *mem);
+
+	struct drm_i915_gem_object *
+	(*create_object)(struct intel_memory_region *mem,
+			 resource_size_t size,
+			 unsigned int flags);
+};
+
+struct intel_memory_region {
+	struct drm_i915_private *i915;
+
+	const struct intel_memory_region_ops *ops;
+
+	struct io_mapping iomap;
+	struct resource region;
+
+	/* For fake LMEM */
+	struct drm_mm_node fake_mappable;
+
+	struct i915_buddy_mm mm;
+	struct mutex mm_lock;
+
+	struct kref kref;
+
+	resource_size_t io_start;
+	resource_size_t min_page_size;
+
+	unsigned int type;
+	unsigned int instance;
+	unsigned int id;
+
+	dma_addr_t remap_addr;
+
+	struct {
+		struct mutex lock; /* Protects access to objects */
+		struct list_head list;
+		struct list_head purgeable;
+	} objects;
+};
+
+int intel_memory_region_init_buddy(struct intel_memory_region *mem);
+void intel_memory_region_release_buddy(struct intel_memory_region *mem);
+
+int __intel_memory_region_get_pages_buddy(struct intel_memory_region *mem,
+					  resource_size_t size,
+					  unsigned int flags,
+					  struct list_head *blocks);
+struct i915_buddy_block *
+__intel_memory_region_get_block_buddy(struct intel_memory_region *mem,
+				      resource_size_t size,
+				      unsigned int flags);
+void __intel_memory_region_put_pages_buddy(struct intel_memory_region *mem,
+					   struct list_head *blocks);
+void __intel_memory_region_put_block_buddy(struct i915_buddy_block *block);
+
+struct intel_memory_region *
+intel_memory_region_create(struct drm_i915_private *i915,
+			   resource_size_t start,
+			   resource_size_t size,
+			   resource_size_t min_page_size,
+			   resource_size_t io_start,
+			   const struct intel_memory_region_ops *ops);
+
+struct intel_memory_region *
+intel_memory_region_get(struct intel_memory_region *mem);
+void intel_memory_region_put(struct intel_memory_region *mem);
+
+int intel_memory_regions_hw_probe(struct drm_i915_private *i915);
+void intel_memory_regions_driver_release(struct drm_i915_private *i915);
+
+#endif
diff --git a/drivers/gpu/drm/i915/intel_pch.c b/drivers/gpu/drm/i915/intel_pch.c
index fa864d8f2b73..8fd92b9130a7 100644
--- a/drivers/gpu/drm/i915/intel_pch.c
+++ b/drivers/gpu/drm/i915/intel_pch.c
@@ -52,7 +52,8 @@ intel_pch_type(const struct drm_i915_private *dev_priv, unsigned short id)
 		return PCH_SPT;
 	case INTEL_PCH_SPT_LP_DEVICE_ID_TYPE:
 		DRM_DEBUG_KMS("Found SunrisePoint LP PCH\n");
-		WARN_ON(!IS_SKYLAKE(dev_priv) && !IS_KABYLAKE(dev_priv));
+		WARN_ON(!IS_SKYLAKE(dev_priv) && !IS_KABYLAKE(dev_priv) &&
+			!IS_COFFEELAKE(dev_priv));
 		return PCH_SPT;
 	case INTEL_PCH_KBP_DEVICE_ID_TYPE:
 		DRM_DEBUG_KMS("Found Kaby Lake PCH (KBP)\n");
@@ -69,16 +70,21 @@ intel_pch_type(const struct drm_i915_private *dev_priv, unsigned short id)
 		WARN_ON(!IS_CANNONLAKE(dev_priv) && !IS_COFFEELAKE(dev_priv));
 		return PCH_CNP;
 	case INTEL_PCH_CMP_DEVICE_ID_TYPE:
+	case INTEL_PCH_CMP2_DEVICE_ID_TYPE:
 		DRM_DEBUG_KMS("Found Comet Lake PCH (CMP)\n");
 		WARN_ON(!IS_COFFEELAKE(dev_priv));
 		/* CometPoint is CNP Compatible */
 		return PCH_CNP;
+	case INTEL_PCH_CMP_V_DEVICE_ID_TYPE:
+		DRM_DEBUG_KMS("Found Comet Lake V PCH (CMP-V)\n");
+		WARN_ON(!IS_COFFEELAKE(dev_priv));
+		/* Comet Lake V PCH is based on KBP, which is SPT compatible */
+		return PCH_SPT;
 	case INTEL_PCH_ICP_DEVICE_ID_TYPE:
 		DRM_DEBUG_KMS("Found Ice Lake PCH\n");
 		WARN_ON(!IS_ICELAKE(dev_priv));
 		return PCH_ICP;
 	case INTEL_PCH_MCC_DEVICE_ID_TYPE:
-	case INTEL_PCH_MCC2_DEVICE_ID_TYPE:
 		DRM_DEBUG_KMS("Found Mule Creek Canyon PCH\n");
 		WARN_ON(!IS_ELKHARTLAKE(dev_priv));
 		return PCH_MCC;
@@ -86,6 +92,11 @@ intel_pch_type(const struct drm_i915_private *dev_priv, unsigned short id)
 		DRM_DEBUG_KMS("Found Tiger Lake LP PCH\n");
 		WARN_ON(!IS_TIGERLAKE(dev_priv));
 		return PCH_TGP;
+	case INTEL_PCH_JSP_DEVICE_ID_TYPE:
+	case INTEL_PCH_JSP2_DEVICE_ID_TYPE:
+		DRM_DEBUG_KMS("Found Jasper Lake PCH\n");
+		WARN_ON(!IS_ELKHARTLAKE(dev_priv));
+		return PCH_JSP;
 	default:
 		return PCH_NONE;
 	}
diff --git a/drivers/gpu/drm/i915/intel_pch.h b/drivers/gpu/drm/i915/intel_pch.h
index e6a2d65f19c6..d26c25dd8d54 100644
--- a/drivers/gpu/drm/i915/intel_pch.h
+++ b/drivers/gpu/drm/i915/intel_pch.h
@@ -23,6 +23,7 @@ enum intel_pch {
 	PCH_SPT,        /* Sunrisepoint/Kaby Lake PCH */
 	PCH_CNP,        /* Cannon/Comet Lake PCH */
 	PCH_ICP,	/* Ice Lake PCH */
+	PCH_JSP,	/* Jasper Lake PCH */
 	PCH_MCC,        /* Mule Creek Canyon PCH */
 	PCH_TGP,	/* Tiger Lake PCH */
 };
@@ -41,16 +42,20 @@ enum intel_pch {
 #define INTEL_PCH_CNP_DEVICE_ID_TYPE		0xA300
 #define INTEL_PCH_CNP_LP_DEVICE_ID_TYPE		0x9D80
 #define INTEL_PCH_CMP_DEVICE_ID_TYPE		0x0280
+#define INTEL_PCH_CMP2_DEVICE_ID_TYPE		0x0680
+#define INTEL_PCH_CMP_V_DEVICE_ID_TYPE		0xA380
 #define INTEL_PCH_ICP_DEVICE_ID_TYPE		0x3480
 #define INTEL_PCH_MCC_DEVICE_ID_TYPE		0x4B00
-#define INTEL_PCH_MCC2_DEVICE_ID_TYPE		0x3880
 #define INTEL_PCH_TGP_DEVICE_ID_TYPE		0xA080
+#define INTEL_PCH_JSP_DEVICE_ID_TYPE		0x4D80
+#define INTEL_PCH_JSP2_DEVICE_ID_TYPE		0x3880
 #define INTEL_PCH_P2X_DEVICE_ID_TYPE		0x7100
 #define INTEL_PCH_P3X_DEVICE_ID_TYPE		0x7000
 #define INTEL_PCH_QEMU_DEVICE_ID_TYPE		0x2900 /* qemu q35 has 2918 */
 
 #define INTEL_PCH_TYPE(dev_priv)		((dev_priv)->pch_type)
 #define INTEL_PCH_ID(dev_priv)			((dev_priv)->pch_id)
+#define HAS_PCH_JSP(dev_priv)			(INTEL_PCH_TYPE(dev_priv) == PCH_JSP)
 #define HAS_PCH_MCC(dev_priv)			(INTEL_PCH_TYPE(dev_priv) == PCH_MCC)
 #define HAS_PCH_TGP(dev_priv)			(INTEL_PCH_TYPE(dev_priv) == PCH_TGP)
 #define HAS_PCH_ICP(dev_priv)			(INTEL_PCH_TYPE(dev_priv) == PCH_ICP)
diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c
index 75ee027abb80..809bff955b5a 100644
--- a/drivers/gpu/drm/i915/intel_pm.c
+++ b/drivers/gpu/drm/i915/intel_pm.c
@@ -25,7 +25,6 @@
  *
  */
 
-#include <linux/cpufreq.h>
 #include <linux/module.h>
 #include <linux/pm_runtime.h>
 
@@ -38,6 +37,8 @@
 #include "display/intel_fbc.h"
 #include "display/intel_sprite.h"
 
+#include "gt/intel_llc.h"
+
 #include "i915_drv.h"
 #include "i915_irq.h"
 #include "i915_trace.h"
@@ -45,26 +46,6 @@
 #include "intel_sideband.h"
 #include "../../../platform/x86/intel_ips.h"
 
-/**
- * DOC: RC6
- *
- * RC6 is a special power stage which allows the GPU to enter an very
- * low-voltage mode when idle, using down to 0V while at this stage.  This
- * stage is entered automatically when the GPU is idle when RC6 support is
- * enabled, and as soon as new workload arises GPU wakes up automatically as well.
- *
- * There are different RC6 modes available in Intel GPU, which differentiate
- * among each other with the latency required to enter and leave RC6 and
- * voltage consumed by the GPU in different states.
- *
- * The combination of the following flags define which states GPU is allowed
- * to enter, while RC6 is the normal RC6 state, RC6p is the deep RC6, and
- * RC6pp is deepest RC6. Their support by hardware varies according to the
- * GPU, BIOS, chipset and platform. RC6 is usually the safest one and the one
- * which brings the most power savings; deeper states save more power, but
- * require higher latency to switch to and wake up.
- */
-
 static void gen9_init_clock_gating(struct drm_i915_private *dev_priv)
 {
 	if (HAS_LLC(dev_priv)) {
@@ -126,6 +107,14 @@ static void bxt_init_clock_gating(struct drm_i915_private *dev_priv)
 	 */
 	I915_WRITE(GEN9_CLKGATE_DIS_0, I915_READ(GEN9_CLKGATE_DIS_0) |
 		   PWM1_GATING_DIS | PWM2_GATING_DIS);
+
+	/*
+	 * Lower the display internal timeout.
+	 * This is needed to avoid any hard hangs when DSI port PLL
+	 * is off and a MMIO access is attempted by any privilege
+	 * application, using batch buffers or any other means.
+	 */
+	I915_WRITE(RM_TIMEOUT, MMIO_TIMEOUT_US(950));
 }
 
 static void glk_init_clock_gating(struct drm_i915_private *dev_priv)
@@ -216,8 +205,6 @@ static void i915_ironlake_get_mem_freq(struct drm_i915_private *dev_priv)
 		break;
 	}
 
-	dev_priv->ips.r_t = dev_priv->mem_freq;
-
 	switch (csipll & 0x3ff) {
 	case 0x00c:
 		dev_priv->fsb_freq = 3200;
@@ -246,14 +233,6 @@ static void i915_ironlake_get_mem_freq(struct drm_i915_private *dev_priv)
 		dev_priv->fsb_freq = 0;
 		break;
 	}
-
-	if (dev_priv->fsb_freq == 3200) {
-		dev_priv->ips.c_m = 0;
-	} else if (dev_priv->fsb_freq > 3200 && dev_priv->fsb_freq <= 4800) {
-		dev_priv->ips.c_m = 1;
-	} else {
-		dev_priv->ips.c_m = 2;
-	}
 }
 
 static const struct cxsr_latency cxsr_latency_table[] = {
@@ -1137,10 +1116,7 @@ static u16 g4x_compute_wm(const struct intel_crtc_state *crtc_state,
 	clock = adjusted_mode->crtc_clock;
 	htotal = adjusted_mode->crtc_htotal;
 
-	if (plane->id == PLANE_CURSOR)
-		width = plane_state->base.crtc_w;
-	else
-		width = drm_rect_width(&plane_state->base.dst);
+	width = drm_rect_width(&plane_state->base.dst);
 
 	if (plane->id == PLANE_CURSOR) {
 		wm = intel_wm_method2(clock, htotal, width, cpp, latency);
@@ -1327,8 +1303,8 @@ static int g4x_compute_pipe_wm(struct intel_crtc_state *crtc_state)
 	struct intel_atomic_state *state =
 		to_intel_atomic_state(crtc_state->base.state);
 	struct g4x_wm_state *wm_state = &crtc_state->wm.g4x.optimal;
-	int num_active_planes = hweight32(crtc_state->active_planes &
-					  ~BIT(PLANE_CURSOR));
+	int num_active_planes = hweight8(crtc_state->active_planes &
+					 ~BIT(PLANE_CURSOR));
 	const struct g4x_pipe_wm *raw;
 	const struct intel_plane_state *old_plane_state;
 	const struct intel_plane_state *new_plane_state;
@@ -1490,7 +1466,7 @@ static void g4x_merge_wm(struct drm_i915_private *dev_priv,
 			 struct g4x_wm_values *wm)
 {
 	struct intel_crtc *crtc;
-	int num_active_crtcs = 0;
+	int num_active_pipes = 0;
 
 	wm->cxsr = true;
 	wm->hpll_en = true;
@@ -1509,10 +1485,10 @@ static void g4x_merge_wm(struct drm_i915_private *dev_priv,
 		if (!wm_state->fbc_en)
 			wm->fbc_en = false;
 
-		num_active_crtcs++;
+		num_active_pipes++;
 	}
 
-	if (num_active_crtcs != 1) {
+	if (num_active_pipes != 1) {
 		wm->cxsr = false;
 		wm->hpll_en = false;
 		wm->fbc_en = false;
@@ -1659,7 +1635,7 @@ static int vlv_compute_fifo(struct intel_crtc_state *crtc_state)
 		&crtc_state->wm.vlv.raw[VLV_WM_LEVEL_PM2];
 	struct vlv_fifo_state *fifo_state = &crtc_state->wm.vlv.fifo_state;
 	unsigned int active_planes = crtc_state->active_planes & ~BIT(PLANE_CURSOR);
-	int num_active_planes = hweight32(active_planes);
+	int num_active_planes = hweight8(active_planes);
 	const int fifo_size = 511;
 	int fifo_extra, fifo_left = fifo_size;
 	int sprite0_fifo_extra = 0;
@@ -1848,8 +1824,8 @@ static int vlv_compute_pipe_wm(struct intel_crtc_state *crtc_state)
 	struct vlv_wm_state *wm_state = &crtc_state->wm.vlv.optimal;
 	const struct vlv_fifo_state *fifo_state =
 		&crtc_state->wm.vlv.fifo_state;
-	int num_active_planes = hweight32(crtc_state->active_planes &
-					  ~BIT(PLANE_CURSOR));
+	int num_active_planes = hweight8(crtc_state->active_planes &
+					 ~BIT(PLANE_CURSOR));
 	bool needs_modeset = drm_atomic_crtc_needs_modeset(&crtc_state->base);
 	const struct intel_plane_state *old_plane_state;
 	const struct intel_plane_state *new_plane_state;
@@ -1909,7 +1885,7 @@ static int vlv_compute_pipe_wm(struct intel_crtc_state *crtc_state)
 
 	for (level = 0; level < wm_state->num_levels; level++) {
 		const struct g4x_pipe_wm *raw = &crtc_state->wm.vlv.raw[level];
-		const int sr_fifo_size = INTEL_INFO(dev_priv)->num_pipes * 512 - 1;
+		const int sr_fifo_size = INTEL_NUM_PIPES(dev_priv) * 512 - 1;
 
 		if (!vlv_raw_crtc_wm_is_valid(crtc_state, level))
 			break;
@@ -2098,7 +2074,7 @@ static void vlv_merge_wm(struct drm_i915_private *dev_priv,
 			 struct vlv_wm_values *wm)
 {
 	struct intel_crtc *crtc;
-	int num_active_crtcs = 0;
+	int num_active_pipes = 0;
 
 	wm->level = dev_priv->wm.max_level;
 	wm->cxsr = true;
@@ -2112,14 +2088,14 @@ static void vlv_merge_wm(struct drm_i915_private *dev_priv,
 		if (!wm_state->cxsr)
 			wm->cxsr = false;
 
-		num_active_crtcs++;
+		num_active_pipes++;
 		wm->level = min_t(int, wm->level, wm_state->num_levels - 1);
 	}
 
-	if (num_active_crtcs != 1)
+	if (num_active_pipes != 1)
 		wm->cxsr = false;
 
-	if (num_active_crtcs > 1)
+	if (num_active_pipes > 1)
 		wm->level = VLV_WM_LEVEL_PM2;
 
 	for_each_intel_crtc(&dev_priv->drm, crtc) {
@@ -2569,7 +2545,8 @@ static u32 ilk_compute_cur_wm(const struct intel_crtc_state *crtc_state,
 
 	return ilk_wm_method2(crtc_state->pixel_rate,
 			      crtc_state->base.adjusted_mode.crtc_htotal,
-			      plane_state->base.crtc_w, cpp, mem_value);
+			      drm_rect_width(&plane_state->base.dst),
+			      cpp, mem_value);
 }
 
 /* Only for WM_LP. */
@@ -2648,7 +2625,7 @@ static unsigned int ilk_plane_wm_max(const struct drm_i915_private *dev_priv,
 
 	/* HSW allows LP1+ watermarks even with multiple pipes */
 	if (level == 0 || config->num_pipes_active > 1) {
-		fifo_size /= INTEL_INFO(dev_priv)->num_pipes;
+		fifo_size /= INTEL_NUM_PIPES(dev_priv);
 
 		/*
 		 * For some reason the non self refresh
@@ -3109,8 +3086,8 @@ static int ilk_compute_pipe_wm(struct intel_crtc_state *crtc_state)
 	struct intel_pipe_wm *pipe_wm;
 	struct drm_device *dev = state->dev;
 	const struct drm_i915_private *dev_priv = to_i915(dev);
-	struct drm_plane *plane;
-	const struct drm_plane_state *plane_state;
+	struct intel_plane *plane;
+	const struct intel_plane_state *plane_state;
 	const struct intel_plane_state *pristate = NULL;
 	const struct intel_plane_state *sprstate = NULL;
 	const struct intel_plane_state *curstate = NULL;
@@ -3119,15 +3096,13 @@ static int ilk_compute_pipe_wm(struct intel_crtc_state *crtc_state)
 
 	pipe_wm = &crtc_state->wm.ilk.optimal;
 
-	drm_atomic_crtc_state_for_each_plane_state(plane, plane_state, &crtc_state->base) {
-		const struct intel_plane_state *ps = to_intel_plane_state(plane_state);
-
-		if (plane->type == DRM_PLANE_TYPE_PRIMARY)
-			pristate = ps;
-		else if (plane->type == DRM_PLANE_TYPE_OVERLAY)
-			sprstate = ps;
-		else if (plane->type == DRM_PLANE_TYPE_CURSOR)
-			curstate = ps;
+	intel_atomic_crtc_state_for_each_plane_state(plane, plane_state, crtc_state) {
+		if (plane->base.type == DRM_PLANE_TYPE_PRIMARY)
+			pristate = plane_state;
+		else if (plane->base.type == DRM_PLANE_TYPE_OVERLAY)
+			sprstate = plane_state;
+		else if (plane->base.type == DRM_PLANE_TYPE_CURSOR)
+			curstate = plane_state;
 	}
 
 	pipe_wm->pipe_enabled = crtc_state->base.active;
@@ -3654,10 +3629,47 @@ static bool skl_needs_memory_bw_wa(struct drm_i915_private *dev_priv)
 static bool
 intel_has_sagv(struct drm_i915_private *dev_priv)
 {
+	/* HACK! */
+	if (IS_GEN(dev_priv, 12))
+		return false;
+
 	return (IS_GEN9_BC(dev_priv) || INTEL_GEN(dev_priv) >= 10) &&
 		dev_priv->sagv_status != I915_SAGV_NOT_CONTROLLED;
 }
 
+static void
+skl_setup_sagv_block_time(struct drm_i915_private *dev_priv)
+{
+	if (INTEL_GEN(dev_priv) >= 12) {
+		u32 val = 0;
+		int ret;
+
+		ret = sandybridge_pcode_read(dev_priv,
+					     GEN12_PCODE_READ_SAGV_BLOCK_TIME_US,
+					     &val, NULL);
+		if (!ret) {
+			dev_priv->sagv_block_time_us = val;
+			return;
+		}
+
+		DRM_DEBUG_DRIVER("Couldn't read SAGV block time!\n");
+	} else if (IS_GEN(dev_priv, 11)) {
+		dev_priv->sagv_block_time_us = 10;
+		return;
+	} else if (IS_GEN(dev_priv, 10)) {
+		dev_priv->sagv_block_time_us = 20;
+		return;
+	} else if (IS_GEN(dev_priv, 9)) {
+		dev_priv->sagv_block_time_us = 30;
+		return;
+	} else {
+		MISSING_CASE(INTEL_GEN(dev_priv));
+	}
+
+	/* Default to an unusable block time */
+	dev_priv->sagv_block_time_us = -1;
+}
+
 /*
  * SAGV dynamically adjusts the system agent voltage and clock frequencies
  * depending on power and performance requirements. The display engine access
@@ -3746,33 +3758,25 @@ bool intel_can_enable_sagv(struct intel_atomic_state *state)
 	struct intel_crtc_state *crtc_state;
 	enum pipe pipe;
 	int level, latency;
-	int sagv_block_time_us;
 
 	if (!intel_has_sagv(dev_priv))
 		return false;
 
-	if (IS_GEN(dev_priv, 9))
-		sagv_block_time_us = 30;
-	else if (IS_GEN(dev_priv, 10))
-		sagv_block_time_us = 20;
-	else
-		sagv_block_time_us = 10;
-
 	/*
 	 * If there are no active CRTCs, no additional checks need be performed
 	 */
-	if (hweight32(state->active_crtcs) == 0)
+	if (hweight8(state->active_pipes) == 0)
 		return true;
 
 	/*
 	 * SKL+ workaround: bspec recommends we disable SAGV when we have
 	 * more then one pipe enabled
 	 */
-	if (hweight32(state->active_crtcs) > 1)
+	if (hweight8(state->active_pipes) > 1)
 		return false;
 
 	/* Since we're now guaranteed to only have one active CRTC... */
-	pipe = ffs(state->active_crtcs) - 1;
+	pipe = ffs(state->active_pipes) - 1;
 	crtc = intel_get_crtc_for_pipe(dev_priv, pipe);
 	crtc_state = to_intel_crtc_state(crtc->base.state);
 
@@ -3804,7 +3808,7 @@ bool intel_can_enable_sagv(struct intel_atomic_state *state)
 		 * incur memory latencies higher than sagv_block_time_us we
 		 * can't enable SAGV.
 		 */
-		if (latency < sagv_block_time_us)
+		if (latency < dev_priv->sagv_block_time_us)
 			return false;
 	}
 
@@ -3867,14 +3871,14 @@ skl_ddb_get_pipe_allocation_limits(struct drm_i915_private *dev_priv,
 	if (WARN_ON(!state) || !crtc_state->base.active) {
 		alloc->start = 0;
 		alloc->end = 0;
-		*num_active = hweight32(dev_priv->active_crtcs);
+		*num_active = hweight8(dev_priv->active_pipes);
 		return;
 	}
 
 	if (intel_state->active_pipe_changes)
-		*num_active = hweight32(intel_state->active_crtcs);
+		*num_active = hweight8(intel_state->active_pipes);
 	else
-		*num_active = hweight32(dev_priv->active_crtcs);
+		*num_active = hweight8(dev_priv->active_pipes);
 
 	ddb_size = intel_get_ddb_size(dev_priv, crtc_state, total_data_rate,
 				      *num_active, ddb);
@@ -4005,7 +4009,8 @@ skl_ddb_get_hw_plane_state(struct drm_i915_private *dev_priv,
 		val = I915_READ(PLANE_BUF_CFG(pipe, plane_id));
 		val2 = I915_READ(PLANE_NV12_BUF_CFG(pipe, plane_id));
 
-		if (is_planar_yuv_format(fourcc))
+		if (fourcc &&
+		    drm_format_info_is_yuv_semiplanar(drm_format_info(fourcc)))
 			swap(val, val2);
 
 		skl_ddb_entry_init_from_hw(dev_priv, ddb_y, val);
@@ -4063,7 +4068,6 @@ static uint_fixed_16_16_t
 skl_plane_downscale_amount(const struct intel_crtc_state *crtc_state,
 			   const struct intel_plane_state *plane_state)
 {
-	struct intel_plane *plane = to_intel_plane(plane_state->base.plane);
 	u32 src_w, src_h, dst_w, dst_h;
 	uint_fixed_16_16_t fp_w_ratio, fp_h_ratio;
 	uint_fixed_16_16_t downscale_h, downscale_w;
@@ -4071,27 +4075,17 @@ skl_plane_downscale_amount(const struct intel_crtc_state *crtc_state,
 	if (WARN_ON(!intel_wm_plane_visible(crtc_state, plane_state)))
 		return u32_to_fixed16(0);
 
-	/* n.b., src is 16.16 fixed point, dst is whole integer */
-	if (plane->id == PLANE_CURSOR) {
-		/*
-		 * Cursors only support 0/180 degree rotation,
-		 * hence no need to account for rotation here.
-		 */
-		src_w = plane_state->base.src_w >> 16;
-		src_h = plane_state->base.src_h >> 16;
-		dst_w = plane_state->base.crtc_w;
-		dst_h = plane_state->base.crtc_h;
-	} else {
-		/*
-		 * Src coordinates are already rotated by 270 degrees for
-		 * the 90/270 degree plane rotation cases (to match the
-		 * GTT mapping), hence no need to account for rotation here.
-		 */
-		src_w = drm_rect_width(&plane_state->base.src) >> 16;
-		src_h = drm_rect_height(&plane_state->base.src) >> 16;
-		dst_w = drm_rect_width(&plane_state->base.dst);
-		dst_h = drm_rect_height(&plane_state->base.dst);
-	}
+	/*
+	 * Src coordinates are already rotated by 270 degrees for
+	 * the 90/270 degree plane rotation cases (to match the
+	 * GTT mapping), hence no need to account for rotation here.
+	 *
+	 * n.b., src is 16.16 fixed point, dst is whole integer.
+	 */
+	src_w = drm_rect_width(&plane_state->base.src) >> 16;
+	src_h = drm_rect_height(&plane_state->base.src) >> 16;
+	dst_w = drm_rect_width(&plane_state->base.dst);
+	dst_h = drm_rect_height(&plane_state->base.dst);
 
 	fp_w_ratio = div_fixed16(src_w, dst_w);
 	fp_h_ratio = div_fixed16(src_h, dst_h);
@@ -4101,117 +4095,26 @@ skl_plane_downscale_amount(const struct intel_crtc_state *crtc_state,
 	return mul_fixed16(downscale_w, downscale_h);
 }
 
-static uint_fixed_16_16_t
-skl_pipe_downscale_amount(const struct intel_crtc_state *crtc_state)
-{
-	uint_fixed_16_16_t pipe_downscale = u32_to_fixed16(1);
-
-	if (!crtc_state->base.enable)
-		return pipe_downscale;
-
-	if (crtc_state->pch_pfit.enabled) {
-		u32 src_w, src_h, dst_w, dst_h;
-		u32 pfit_size = crtc_state->pch_pfit.size;
-		uint_fixed_16_16_t fp_w_ratio, fp_h_ratio;
-		uint_fixed_16_16_t downscale_h, downscale_w;
-
-		src_w = crtc_state->pipe_src_w;
-		src_h = crtc_state->pipe_src_h;
-		dst_w = pfit_size >> 16;
-		dst_h = pfit_size & 0xffff;
-
-		if (!dst_w || !dst_h)
-			return pipe_downscale;
-
-		fp_w_ratio = div_fixed16(src_w, dst_w);
-		fp_h_ratio = div_fixed16(src_h, dst_h);
-		downscale_w = max_fixed16(fp_w_ratio, u32_to_fixed16(1));
-		downscale_h = max_fixed16(fp_h_ratio, u32_to_fixed16(1));
-
-		pipe_downscale = mul_fixed16(downscale_w, downscale_h);
-	}
-
-	return pipe_downscale;
-}
-
-int skl_check_pipe_max_pixel_rate(struct intel_crtc *intel_crtc,
-				  struct intel_crtc_state *crtc_state)
-{
-	struct drm_i915_private *dev_priv = to_i915(intel_crtc->base.dev);
-	struct drm_atomic_state *state = crtc_state->base.state;
-	struct drm_plane *plane;
-	const struct drm_plane_state *drm_plane_state;
-	int crtc_clock, dotclk;
-	u32 pipe_max_pixel_rate;
-	uint_fixed_16_16_t pipe_downscale;
-	uint_fixed_16_16_t max_downscale = u32_to_fixed16(1);
-
-	if (!crtc_state->base.enable)
-		return 0;
-
-	drm_atomic_crtc_state_for_each_plane_state(plane, drm_plane_state, &crtc_state->base) {
-		uint_fixed_16_16_t plane_downscale;
-		uint_fixed_16_16_t fp_9_div_8 = div_fixed16(9, 8);
-		int bpp;
-		const struct intel_plane_state *plane_state =
-			to_intel_plane_state(drm_plane_state);
-
-		if (!intel_wm_plane_visible(crtc_state, plane_state))
-			continue;
-
-		if (WARN_ON(!plane_state->base.fb))
-			return -EINVAL;
-
-		plane_downscale = skl_plane_downscale_amount(crtc_state, plane_state);
-		bpp = plane_state->base.fb->format->cpp[0] * 8;
-		if (bpp == 64)
-			plane_downscale = mul_fixed16(plane_downscale,
-						      fp_9_div_8);
-
-		max_downscale = max_fixed16(plane_downscale, max_downscale);
-	}
-	pipe_downscale = skl_pipe_downscale_amount(crtc_state);
-
-	pipe_downscale = mul_fixed16(pipe_downscale, max_downscale);
-
-	crtc_clock = crtc_state->base.adjusted_mode.crtc_clock;
-	dotclk = to_intel_atomic_state(state)->cdclk.logical.cdclk;
-
-	if (IS_GEMINILAKE(dev_priv) || INTEL_GEN(dev_priv) >= 10)
-		dotclk *= 2;
-
-	pipe_max_pixel_rate = div_round_up_u32_fixed16(dotclk, pipe_downscale);
-
-	if (pipe_max_pixel_rate < crtc_clock) {
-		DRM_DEBUG_KMS("Max supported pixel clock with scaling exceeded\n");
-		return -EINVAL;
-	}
-
-	return 0;
-}
-
 static u64
 skl_plane_relative_data_rate(const struct intel_crtc_state *crtc_state,
 			     const struct intel_plane_state *plane_state,
-			     const int plane)
+			     int color_plane)
 {
-	struct intel_plane *intel_plane = to_intel_plane(plane_state->base.plane);
+	struct intel_plane *plane = to_intel_plane(plane_state->base.plane);
+	const struct drm_framebuffer *fb = plane_state->base.fb;
 	u32 data_rate;
 	u32 width = 0, height = 0;
-	struct drm_framebuffer *fb;
-	u32 format;
 	uint_fixed_16_16_t down_scale_amount;
 	u64 rate;
 
 	if (!plane_state->base.visible)
 		return 0;
 
-	fb = plane_state->base.fb;
-	format = fb->format->format;
-
-	if (intel_plane->id == PLANE_CURSOR)
+	if (plane->id == PLANE_CURSOR)
 		return 0;
-	if (plane == 1 && !is_planar_yuv_format(format))
+
+	if (color_plane == 1 &&
+	    !drm_format_info_is_yuv_semiplanar(fb->format))
 		return 0;
 
 	/*
@@ -4223,7 +4126,7 @@ skl_plane_relative_data_rate(const struct intel_crtc_state *crtc_state,
 	height = drm_rect_height(&plane_state->base.src) >> 16;
 
 	/* UV plane does 1/2 pixel sub-sampling */
-	if (plane == 1 && is_planar_yuv_format(format)) {
+	if (color_plane == 1) {
 		width /= 2;
 		height /= 2;
 	}
@@ -4234,7 +4137,7 @@ skl_plane_relative_data_rate(const struct intel_crtc_state *crtc_state,
 
 	rate = mul_round_up_u32_fixed16(data_rate, down_scale_amount);
 
-	rate *= fb->format->cpp[plane];
+	rate *= fb->format->cpp[color_plane];
 	return rate;
 }
 
@@ -4244,18 +4147,16 @@ skl_get_total_relative_data_rate(struct intel_crtc_state *crtc_state,
 				 u64 *uv_plane_data_rate)
 {
 	struct drm_atomic_state *state = crtc_state->base.state;
-	struct drm_plane *plane;
-	const struct drm_plane_state *drm_plane_state;
+	struct intel_plane *plane;
+	const struct intel_plane_state *plane_state;
 	u64 total_data_rate = 0;
 
 	if (WARN_ON(!state))
 		return 0;
 
 	/* Calculate and cache data rate for each plane */
-	drm_atomic_crtc_state_for_each_plane_state(plane, drm_plane_state, &crtc_state->base) {
-		enum plane_id plane_id = to_intel_plane(plane)->id;
-		const struct intel_plane_state *plane_state =
-			to_intel_plane_state(drm_plane_state);
+	intel_atomic_crtc_state_for_each_plane_state(plane, plane_state, crtc_state) {
+		enum plane_id plane_id = plane->id;
 		u64 rate;
 
 		/* packed/y */
@@ -4276,21 +4177,19 @@ static u64
 icl_get_total_relative_data_rate(struct intel_crtc_state *crtc_state,
 				 u64 *plane_data_rate)
 {
-	struct drm_plane *plane;
-	const struct drm_plane_state *drm_plane_state;
+	struct intel_plane *plane;
+	const struct intel_plane_state *plane_state;
 	u64 total_data_rate = 0;
 
 	if (WARN_ON(!crtc_state->base.state))
 		return 0;
 
 	/* Calculate and cache data rate for each plane */
-	drm_atomic_crtc_state_for_each_plane_state(plane, drm_plane_state, &crtc_state->base) {
-		const struct intel_plane_state *plane_state =
-			to_intel_plane_state(drm_plane_state);
-		enum plane_id plane_id = to_intel_plane(plane)->id;
+	intel_atomic_crtc_state_for_each_plane_state(plane, plane_state, crtc_state) {
+		enum plane_id plane_id = plane->id;
 		u64 rate;
 
-		if (!plane_state->linked_plane) {
+		if (!plane_state->planar_linked_plane) {
 			rate = skl_plane_relative_data_rate(crtc_state, plane_state, 0);
 			plane_data_rate[plane_id] = rate;
 			total_data_rate += rate;
@@ -4299,17 +4198,17 @@ icl_get_total_relative_data_rate(struct intel_crtc_state *crtc_state,
 
 			/*
 			 * The slave plane might not iterate in
-			 * drm_atomic_crtc_state_for_each_plane_state(),
+			 * intel_atomic_crtc_state_for_each_plane_state(),
 			 * and needs the master plane state which may be
 			 * NULL if we try get_new_plane_state(), so we
 			 * always calculate from the master.
 			 */
-			if (plane_state->slave)
+			if (plane_state->planar_slave)
 				continue;
 
 			/* Y plane rate is calculated on the slave */
 			rate = skl_plane_relative_data_rate(crtc_state, plane_state, 0);
-			y_plane_id = plane_state->linked_plane->id;
+			y_plane_id = plane_state->planar_linked_plane->id;
 			plane_data_rate[y_plane_id] = rate;
 			total_data_rate += rate;
 
@@ -4639,7 +4538,7 @@ skl_compute_wm_params(const struct intel_crtc_state *crtc_state,
 	u32 interm_pbpl;
 
 	/* only planar format has two planes */
-	if (color_plane == 1 && !is_planar_yuv_format(format->format)) {
+	if (color_plane == 1 && !drm_format_info_is_yuv_semiplanar(format)) {
 		DRM_DEBUG_KMS("Non planar format have single plane\n");
 		return -EINVAL;
 	}
@@ -4651,7 +4550,7 @@ skl_compute_wm_params(const struct intel_crtc_state *crtc_state,
 	wp->x_tiled = modifier == I915_FORMAT_MOD_X_TILED;
 	wp->rc_surface = modifier == I915_FORMAT_MOD_Y_TILED_CCS ||
 			 modifier == I915_FORMAT_MOD_Yf_TILED_CCS;
-	wp->is_planar = is_planar_yuv_format(format->format);
+	wp->is_planar = drm_format_info_is_yuv_semiplanar(format);
 
 	wp->width = width;
 	if (color_plane == 1 && wp->is_planar)
@@ -4723,20 +4622,15 @@ skl_compute_plane_wm_params(const struct intel_crtc_state *crtc_state,
 			    const struct intel_plane_state *plane_state,
 			    struct skl_wm_params *wp, int color_plane)
 {
-	struct intel_plane *plane = to_intel_plane(plane_state->base.plane);
 	const struct drm_framebuffer *fb = plane_state->base.fb;
 	int width;
 
-	if (plane->id == PLANE_CURSOR) {
-		width = plane_state->base.crtc_w;
-	} else {
-		/*
-		 * Src coordinates are already rotated by 270 degrees for
-		 * the 90/270 degree plane rotation cases (to match the
-		 * GTT mapping), hence no need to account for rotation here.
-		 */
-		width = drm_rect_width(&plane_state->base.src) >> 16;
-	}
+	/*
+	 * Src coordinates are already rotated by 270 degrees for
+	 * the 90/270 degree plane rotation cases (to match the
+	 * GTT mapping), hence no need to account for rotation here.
+	 */
+	width = drm_rect_width(&plane_state->base.src) >> 16;
 
 	return skl_compute_wm_params(crtc_state, width,
 				     fb->format, fb->modifier,
@@ -5048,12 +4942,12 @@ static int icl_build_plane_wm(struct intel_crtc_state *crtc_state,
 	int ret;
 
 	/* Watermarks calculated in master */
-	if (plane_state->slave)
+	if (plane_state->planar_slave)
 		return 0;
 
-	if (plane_state->linked_plane) {
+	if (plane_state->planar_linked_plane) {
 		const struct drm_framebuffer *fb = plane_state->base.fb;
-		enum plane_id y_plane_id = plane_state->linked_plane->id;
+		enum plane_id y_plane_id = plane_state->planar_linked_plane->id;
 
 		WARN_ON(!intel_wm_plane_visible(crtc_state, plane_state));
 		WARN_ON(!fb->format->is_yuv ||
@@ -5082,8 +4976,8 @@ static int skl_build_pipe_wm(struct intel_crtc_state *crtc_state)
 {
 	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
 	struct skl_pipe_wm *pipe_wm = &crtc_state->wm.skl.optimal;
-	struct drm_plane *plane;
-	const struct drm_plane_state *drm_plane_state;
+	struct intel_plane *plane;
+	const struct intel_plane_state *plane_state;
 	int ret;
 
 	/*
@@ -5092,10 +4986,8 @@ static int skl_build_pipe_wm(struct intel_crtc_state *crtc_state)
 	 */
 	memset(pipe_wm->planes, 0, sizeof(pipe_wm->planes));
 
-	drm_atomic_crtc_state_for_each_plane_state(plane, drm_plane_state,
-						   &crtc_state->base) {
-		const struct intel_plane_state *plane_state =
-			to_intel_plane_state(drm_plane_state);
+	intel_atomic_crtc_state_for_each_plane_state(plane, plane_state,
+						     crtc_state) {
 
 		if (INTEL_GEN(dev_priv) >= 11)
 			ret = icl_build_plane_wm(crtc_state, plane_state);
@@ -5255,19 +5147,6 @@ bool skl_ddb_allocation_overlaps(const struct skl_ddb_entry *ddb,
 	return false;
 }
 
-static u32
-pipes_modified(struct intel_atomic_state *state)
-{
-	struct intel_crtc *crtc;
-	struct intel_crtc_state *crtc_state;
-	u32 i, ret = 0;
-
-	for_each_new_intel_crtc_in_state(state, crtc, crtc_state, i)
-		ret |= drm_crtc_mask(&crtc->base);
-
-	return ret;
-}
-
 static int
 skl_ddb_add_affected_planes(const struct intel_crtc_state *old_crtc_state,
 			    struct intel_crtc_state *new_crtc_state)
@@ -5443,36 +5322,27 @@ skl_print_wm_changes(struct intel_atomic_state *state)
 	}
 }
 
-static int
-skl_ddb_add_affected_pipes(struct intel_atomic_state *state, bool *changed)
+static int intel_add_all_pipes(struct intel_atomic_state *state)
 {
-	struct drm_device *dev = state->base.dev;
-	const struct drm_i915_private *dev_priv = to_i915(dev);
+	struct drm_i915_private *dev_priv = to_i915(state->base.dev);
 	struct intel_crtc *crtc;
-	struct intel_crtc_state *crtc_state;
-	u32 realloc_pipes = pipes_modified(state);
-	int ret, i;
 
-	/*
-	 * When we distrust bios wm we always need to recompute to set the
-	 * expected DDB allocations for each CRTC.
-	 */
-	if (dev_priv->wm.distrust_bios_wm)
-		(*changed) = true;
+	for_each_intel_crtc(&dev_priv->drm, crtc) {
+		struct intel_crtc_state *crtc_state;
 
-	/*
-	 * If this transaction isn't actually touching any CRTC's, don't
-	 * bother with watermark calculation.  Note that if we pass this
-	 * test, we're guaranteed to hold at least one CRTC state mutex,
-	 * which means we can safely use values like dev_priv->active_crtcs
-	 * since any racing commits that want to update them would need to
-	 * hold _all_ CRTC state mutexes.
-	 */
-	for_each_new_intel_crtc_in_state(state, crtc, crtc_state, i)
-		(*changed) = true;
+		crtc_state = intel_atomic_get_crtc_state(&state->base, crtc);
+		if (IS_ERR(crtc_state))
+			return PTR_ERR(crtc_state);
+	}
 
-	if (!*changed)
-		return 0;
+	return 0;
+}
+
+static int
+skl_ddb_add_affected_pipes(struct intel_atomic_state *state)
+{
+	struct drm_i915_private *dev_priv = to_i915(state->base.dev);
+	int ret;
 
 	/*
 	 * If this is our first atomic update following hardware readout,
@@ -5481,7 +5351,7 @@ skl_ddb_add_affected_pipes(struct intel_atomic_state *state, bool *changed)
 	 * ensure a full DDB recompute.
 	 */
 	if (dev_priv->wm.distrust_bios_wm) {
-		ret = drm_modeset_lock(&dev->mode_config.connection_mutex,
+		ret = drm_modeset_lock(&dev_priv->drm.mode_config.connection_mutex,
 				       state->base.acquire_ctx);
 		if (ret)
 			return ret;
@@ -5489,13 +5359,13 @@ skl_ddb_add_affected_pipes(struct intel_atomic_state *state, bool *changed)
 		state->active_pipe_changes = ~0;
 
 		/*
-		 * We usually only initialize state->active_crtcs if we
+		 * We usually only initialize state->active_pipes if we
 		 * we're doing a modeset; make sure this field is always
 		 * initialized during the sanitization process that happens
 		 * on the first commit too.
 		 */
 		if (!state->modeset)
-			state->active_crtcs = dev_priv->active_crtcs;
+			state->active_pipes = dev_priv->active_pipes;
 	}
 
 	/*
@@ -5512,18 +5382,11 @@ skl_ddb_add_affected_pipes(struct intel_atomic_state *state, bool *changed)
 	 * to grab the lock on *all* CRTC's.
 	 */
 	if (state->active_pipe_changes || state->modeset) {
-		realloc_pipes = ~0;
 		state->wm_results.dirty_pipes = ~0;
-	}
 
-	/*
-	 * We're not recomputing for the pipes not included in the commit, so
-	 * make sure we start with the current state.
-	 */
-	for_each_intel_crtc_mask(dev, crtc, realloc_pipes) {
-		crtc_state = intel_atomic_get_crtc_state(&state->base, crtc);
-		if (IS_ERR(crtc_state))
-			return PTR_ERR(crtc_state);
+		ret = intel_add_all_pipes(state);
+		if (ret)
+			return ret;
 	}
 
 	return 0;
@@ -5596,14 +5459,13 @@ skl_compute_wm(struct intel_atomic_state *state)
 	struct intel_crtc_state *new_crtc_state;
 	struct intel_crtc_state *old_crtc_state;
 	struct skl_ddb_values *results = &state->wm_results;
-	bool changed = false;
 	int ret, i;
 
 	/* Clear all dirty flags */
 	results->dirty_pipes = 0;
 
-	ret = skl_ddb_add_affected_pipes(state, &changed);
-	if (ret || !changed)
+	ret = skl_ddb_add_affected_pipes(state);
+	if (ret)
 		return ret;
 
 	/*
@@ -5625,7 +5487,7 @@ skl_compute_wm(struct intel_atomic_state *state)
 		if (!skl_pipe_wm_equals(crtc,
 					&old_crtc_state->wm.skl.optimal,
 					&new_crtc_state->wm.skl.optimal))
-			results->dirty_pipes |= drm_crtc_mask(&crtc->base);
+			results->dirty_pipes |= BIT(crtc->pipe);
 	}
 
 	ret = skl_compute_ddb(state);
@@ -5645,7 +5507,7 @@ static void skl_atomic_update_crtc_wm(struct intel_atomic_state *state,
 	struct skl_pipe_wm *pipe_wm = &crtc_state->wm.skl.optimal;
 	enum pipe pipe = crtc->pipe;
 
-	if (!(state->wm_results.dirty_pipes & drm_crtc_mask(&crtc->base)))
+	if ((state->wm_results.dirty_pipes & BIT(crtc->pipe)) == 0)
 		return;
 
 	I915_WRITE(PIPE_WM_LINETIME(pipe), pipe_wm->linetime);
@@ -5654,12 +5516,11 @@ static void skl_atomic_update_crtc_wm(struct intel_atomic_state *state,
 static void skl_initial_wm(struct intel_atomic_state *state,
 			   struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc_state->base.crtc);
-	struct drm_device *dev = intel_crtc->base.dev;
-	struct drm_i915_private *dev_priv = to_i915(dev);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	struct skl_ddb_values *results = &state->wm_results;
 
-	if ((results->dirty_pipes & drm_crtc_mask(&intel_crtc->base)) == 0)
+	if ((results->dirty_pipes & BIT(crtc->pipe)) == 0)
 		return;
 
 	mutex_lock(&dev_priv->wm.wm_mutex);
@@ -5808,10 +5669,10 @@ void skl_wm_get_hw_state(struct drm_i915_private *dev_priv)
 		skl_pipe_wm_get_hw_state(crtc, &crtc_state->wm.skl.optimal);
 
 		if (crtc->active)
-			hw->dirty_pipes |= drm_crtc_mask(&crtc->base);
+			hw->dirty_pipes |= BIT(crtc->pipe);
 	}
 
-	if (dev_priv->active_crtcs) {
+	if (dev_priv->active_pipes) {
 		/* Fully recompute DDB on first atomic commit */
 		dev_priv->wm.distrust_bios_wm = true;
 	}
@@ -6389,2378 +6250,6 @@ void intel_init_ipc(struct drm_i915_private *dev_priv)
 	intel_enable_ipc(dev_priv);
 }
 
-/*
- * Lock protecting IPS related data structures
- */
-DEFINE_SPINLOCK(mchdev_lock);
-
-bool ironlake_set_drps(struct drm_i915_private *i915, u8 val)
-{
-	struct intel_uncore *uncore = &i915->uncore;
-	u16 rgvswctl;
-
-	lockdep_assert_held(&mchdev_lock);
-
-	rgvswctl = intel_uncore_read16(uncore, MEMSWCTL);
-	if (rgvswctl & MEMCTL_CMD_STS) {
-		DRM_DEBUG("gpu busy, RCS change rejected\n");
-		return false; /* still busy with another command */
-	}
-
-	rgvswctl = (MEMCTL_CMD_CHFREQ << MEMCTL_CMD_SHIFT) |
-		(val << MEMCTL_FREQ_SHIFT) | MEMCTL_SFCAVM;
-	intel_uncore_write16(uncore, MEMSWCTL, rgvswctl);
-	intel_uncore_posting_read16(uncore, MEMSWCTL);
-
-	rgvswctl |= MEMCTL_CMD_STS;
-	intel_uncore_write16(uncore, MEMSWCTL, rgvswctl);
-
-	return true;
-}
-
-static void ironlake_enable_drps(struct drm_i915_private *dev_priv)
-{
-	struct intel_uncore *uncore = &dev_priv->uncore;
-	u32 rgvmodectl;
-	u8 fmax, fmin, fstart, vstart;
-
-	spin_lock_irq(&mchdev_lock);
-
-	rgvmodectl = intel_uncore_read(uncore, MEMMODECTL);
-
-	/* Enable temp reporting */
-	intel_uncore_write16(uncore, PMMISC, I915_READ(PMMISC) | MCPPCE_EN);
-	intel_uncore_write16(uncore, TSC1, I915_READ(TSC1) | TSE);
-
-	/* 100ms RC evaluation intervals */
-	intel_uncore_write(uncore, RCUPEI, 100000);
-	intel_uncore_write(uncore, RCDNEI, 100000);
-
-	/* Set max/min thresholds to 90ms and 80ms respectively */
-	intel_uncore_write(uncore, RCBMAXAVG, 90000);
-	intel_uncore_write(uncore, RCBMINAVG, 80000);
-
-	intel_uncore_write(uncore, MEMIHYST, 1);
-
-	/* Set up min, max, and cur for interrupt handling */
-	fmax = (rgvmodectl & MEMMODE_FMAX_MASK) >> MEMMODE_FMAX_SHIFT;
-	fmin = (rgvmodectl & MEMMODE_FMIN_MASK);
-	fstart = (rgvmodectl & MEMMODE_FSTART_MASK) >>
-		MEMMODE_FSTART_SHIFT;
-
-	vstart = (intel_uncore_read(uncore, PXVFREQ(fstart)) &
-		  PXVFREQ_PX_MASK) >> PXVFREQ_PX_SHIFT;
-
-	dev_priv->ips.fmax = fmax; /* IPS callback will increase this */
-	dev_priv->ips.fstart = fstart;
-
-	dev_priv->ips.max_delay = fstart;
-	dev_priv->ips.min_delay = fmin;
-	dev_priv->ips.cur_delay = fstart;
-
-	DRM_DEBUG_DRIVER("fmax: %d, fmin: %d, fstart: %d\n",
-			 fmax, fmin, fstart);
-
-	intel_uncore_write(uncore,
-			   MEMINTREN,
-			   MEMINT_CX_SUPR_EN | MEMINT_EVAL_CHG_EN);
-
-	/*
-	 * Interrupts will be enabled in ironlake_irq_postinstall
-	 */
-
-	intel_uncore_write(uncore, VIDSTART, vstart);
-	intel_uncore_posting_read(uncore, VIDSTART);
-
-	rgvmodectl |= MEMMODE_SWMODE_EN;
-	intel_uncore_write(uncore, MEMMODECTL, rgvmodectl);
-
-	if (wait_for_atomic((intel_uncore_read(uncore, MEMSWCTL) &
-			     MEMCTL_CMD_STS) == 0, 10))
-		DRM_ERROR("stuck trying to change perf mode\n");
-	mdelay(1);
-
-	ironlake_set_drps(dev_priv, fstart);
-
-	dev_priv->ips.last_count1 =
-		intel_uncore_read(uncore, DMIEC) +
-		intel_uncore_read(uncore, DDREC) +
-		intel_uncore_read(uncore, CSIEC);
-	dev_priv->ips.last_time1 = jiffies_to_msecs(jiffies);
-	dev_priv->ips.last_count2 = intel_uncore_read(uncore, GFXEC);
-	dev_priv->ips.last_time2 = ktime_get_raw_ns();
-
-	spin_unlock_irq(&mchdev_lock);
-}
-
-static void ironlake_disable_drps(struct drm_i915_private *i915)
-{
-	struct intel_uncore *uncore = &i915->uncore;
-	u16 rgvswctl;
-
-	spin_lock_irq(&mchdev_lock);
-
-	rgvswctl = intel_uncore_read16(uncore, MEMSWCTL);
-
-	/* Ack interrupts, disable EFC interrupt */
-	intel_uncore_write(uncore,
-			   MEMINTREN,
-			   intel_uncore_read(uncore, MEMINTREN) &
-			   ~MEMINT_EVAL_CHG_EN);
-	intel_uncore_write(uncore, MEMINTRSTS, MEMINT_EVAL_CHG);
-	intel_uncore_write(uncore,
-			   DEIER,
-			   intel_uncore_read(uncore, DEIER) & ~DE_PCU_EVENT);
-	intel_uncore_write(uncore, DEIIR, DE_PCU_EVENT);
-	intel_uncore_write(uncore,
-			   DEIMR,
-			   intel_uncore_read(uncore, DEIMR) | DE_PCU_EVENT);
-
-	/* Go back to the starting frequency */
-	ironlake_set_drps(i915, i915->ips.fstart);
-	mdelay(1);
-	rgvswctl |= MEMCTL_CMD_STS;
-	intel_uncore_write(uncore, MEMSWCTL, rgvswctl);
-	mdelay(1);
-
-	spin_unlock_irq(&mchdev_lock);
-}
-
-/* There's a funny hw issue where the hw returns all 0 when reading from
- * GEN6_RP_INTERRUPT_LIMITS. Hence we always need to compute the desired value
- * ourselves, instead of doing a rmw cycle (which might result in us clearing
- * all limits and the gpu stuck at whatever frequency it is at atm).
- */
-static u32 intel_rps_limits(struct drm_i915_private *dev_priv, u8 val)
-{
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
-	u32 limits;
-
-	/* Only set the down limit when we've reached the lowest level to avoid
-	 * getting more interrupts, otherwise leave this clear. This prevents a
-	 * race in the hw when coming out of rc6: There's a tiny window where
-	 * the hw runs at the minimal clock before selecting the desired
-	 * frequency, if the down threshold expires in that window we will not
-	 * receive a down interrupt. */
-	if (INTEL_GEN(dev_priv) >= 9) {
-		limits = (rps->max_freq_softlimit) << 23;
-		if (val <= rps->min_freq_softlimit)
-			limits |= (rps->min_freq_softlimit) << 14;
-	} else {
-		limits = rps->max_freq_softlimit << 24;
-		if (val <= rps->min_freq_softlimit)
-			limits |= rps->min_freq_softlimit << 16;
-	}
-
-	return limits;
-}
-
-static void rps_set_power(struct drm_i915_private *dev_priv, int new_power)
-{
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
-	u32 threshold_up = 0, threshold_down = 0; /* in % */
-	u32 ei_up = 0, ei_down = 0;
-
-	lockdep_assert_held(&rps->power.mutex);
-
-	if (new_power == rps->power.mode)
-		return;
-
-	/* Note the units here are not exactly 1us, but 1280ns. */
-	switch (new_power) {
-	case LOW_POWER:
-		/* Upclock if more than 95% busy over 16ms */
-		ei_up = 16000;
-		threshold_up = 95;
-
-		/* Downclock if less than 85% busy over 32ms */
-		ei_down = 32000;
-		threshold_down = 85;
-		break;
-
-	case BETWEEN:
-		/* Upclock if more than 90% busy over 13ms */
-		ei_up = 13000;
-		threshold_up = 90;
-
-		/* Downclock if less than 75% busy over 32ms */
-		ei_down = 32000;
-		threshold_down = 75;
-		break;
-
-	case HIGH_POWER:
-		/* Upclock if more than 85% busy over 10ms */
-		ei_up = 10000;
-		threshold_up = 85;
-
-		/* Downclock if less than 60% busy over 32ms */
-		ei_down = 32000;
-		threshold_down = 60;
-		break;
-	}
-
-	/* When byt can survive without system hang with dynamic
-	 * sw freq adjustments, this restriction can be lifted.
-	 */
-	if (IS_VALLEYVIEW(dev_priv))
-		goto skip_hw_write;
-
-	I915_WRITE(GEN6_RP_UP_EI,
-		   GT_INTERVAL_FROM_US(dev_priv, ei_up));
-	I915_WRITE(GEN6_RP_UP_THRESHOLD,
-		   GT_INTERVAL_FROM_US(dev_priv,
-				       ei_up * threshold_up / 100));
-
-	I915_WRITE(GEN6_RP_DOWN_EI,
-		   GT_INTERVAL_FROM_US(dev_priv, ei_down));
-	I915_WRITE(GEN6_RP_DOWN_THRESHOLD,
-		   GT_INTERVAL_FROM_US(dev_priv,
-				       ei_down * threshold_down / 100));
-
-	I915_WRITE(GEN6_RP_CONTROL,
-		   (INTEL_GEN(dev_priv) > 9 ? 0 : GEN6_RP_MEDIA_TURBO) |
-		   GEN6_RP_MEDIA_HW_NORMAL_MODE |
-		   GEN6_RP_MEDIA_IS_GFX |
-		   GEN6_RP_ENABLE |
-		   GEN6_RP_UP_BUSY_AVG |
-		   GEN6_RP_DOWN_IDLE_AVG);
-
-skip_hw_write:
-	rps->power.mode = new_power;
-	rps->power.up_threshold = threshold_up;
-	rps->power.down_threshold = threshold_down;
-}
-
-static void gen6_set_rps_thresholds(struct drm_i915_private *dev_priv, u8 val)
-{
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
-	int new_power;
-
-	new_power = rps->power.mode;
-	switch (rps->power.mode) {
-	case LOW_POWER:
-		if (val > rps->efficient_freq + 1 &&
-		    val > rps->cur_freq)
-			new_power = BETWEEN;
-		break;
-
-	case BETWEEN:
-		if (val <= rps->efficient_freq &&
-		    val < rps->cur_freq)
-			new_power = LOW_POWER;
-		else if (val >= rps->rp0_freq &&
-			 val > rps->cur_freq)
-			new_power = HIGH_POWER;
-		break;
-
-	case HIGH_POWER:
-		if (val < (rps->rp1_freq + rps->rp0_freq) >> 1 &&
-		    val < rps->cur_freq)
-			new_power = BETWEEN;
-		break;
-	}
-	/* Max/min bins are special */
-	if (val <= rps->min_freq_softlimit)
-		new_power = LOW_POWER;
-	if (val >= rps->max_freq_softlimit)
-		new_power = HIGH_POWER;
-
-	mutex_lock(&rps->power.mutex);
-	if (rps->power.interactive)
-		new_power = HIGH_POWER;
-	rps_set_power(dev_priv, new_power);
-	mutex_unlock(&rps->power.mutex);
-}
-
-void intel_rps_mark_interactive(struct drm_i915_private *i915, bool interactive)
-{
-	struct intel_rps *rps = &i915->gt_pm.rps;
-
-	if (INTEL_GEN(i915) < 6)
-		return;
-
-	mutex_lock(&rps->power.mutex);
-	if (interactive) {
-		if (!rps->power.interactive++ && READ_ONCE(i915->gt.awake))
-			rps_set_power(i915, HIGH_POWER);
-	} else {
-		GEM_BUG_ON(!rps->power.interactive);
-		rps->power.interactive--;
-	}
-	mutex_unlock(&rps->power.mutex);
-}
-
-static u32 gen6_rps_pm_mask(struct drm_i915_private *dev_priv, u8 val)
-{
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
-	u32 mask = 0;
-
-	/* We use UP_EI_EXPIRED interupts for both up/down in manual mode */
-	if (val > rps->min_freq_softlimit)
-		mask |= GEN6_PM_RP_UP_EI_EXPIRED | GEN6_PM_RP_DOWN_THRESHOLD | GEN6_PM_RP_DOWN_TIMEOUT;
-	if (val < rps->max_freq_softlimit)
-		mask |= GEN6_PM_RP_UP_EI_EXPIRED | GEN6_PM_RP_UP_THRESHOLD;
-
-	mask &= dev_priv->pm_rps_events;
-
-	return gen6_sanitize_rps_pm_mask(dev_priv, ~mask);
-}
-
-/* gen6_set_rps is called to update the frequency request, but should also be
- * called when the range (min_delay and max_delay) is modified so that we can
- * update the GEN6_RP_INTERRUPT_LIMITS register accordingly. */
-static int gen6_set_rps(struct drm_i915_private *dev_priv, u8 val)
-{
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
-
-	/* min/max delay may still have been modified so be sure to
-	 * write the limits value.
-	 */
-	if (val != rps->cur_freq) {
-		gen6_set_rps_thresholds(dev_priv, val);
-
-		if (INTEL_GEN(dev_priv) >= 9)
-			I915_WRITE(GEN6_RPNSWREQ,
-				   GEN9_FREQUENCY(val));
-		else if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv))
-			I915_WRITE(GEN6_RPNSWREQ,
-				   HSW_FREQUENCY(val));
-		else
-			I915_WRITE(GEN6_RPNSWREQ,
-				   GEN6_FREQUENCY(val) |
-				   GEN6_OFFSET(0) |
-				   GEN6_AGGRESSIVE_TURBO);
-	}
-
-	/* Make sure we continue to get interrupts
-	 * until we hit the minimum or maximum frequencies.
-	 */
-	I915_WRITE(GEN6_RP_INTERRUPT_LIMITS, intel_rps_limits(dev_priv, val));
-	I915_WRITE(GEN6_PMINTRMSK, gen6_rps_pm_mask(dev_priv, val));
-
-	rps->cur_freq = val;
-	trace_intel_gpu_freq_change(intel_gpu_freq(dev_priv, val));
-
-	return 0;
-}
-
-static int valleyview_set_rps(struct drm_i915_private *dev_priv, u8 val)
-{
-	int err;
-
-	if (WARN_ONCE(IS_CHERRYVIEW(dev_priv) && (val & 1),
-		      "Odd GPU freq value\n"))
-		val &= ~1;
-
-	I915_WRITE(GEN6_PMINTRMSK, gen6_rps_pm_mask(dev_priv, val));
-
-	if (val != dev_priv->gt_pm.rps.cur_freq) {
-		vlv_punit_get(dev_priv);
-		err = vlv_punit_write(dev_priv, PUNIT_REG_GPU_FREQ_REQ, val);
-		vlv_punit_put(dev_priv);
-		if (err)
-			return err;
-
-		gen6_set_rps_thresholds(dev_priv, val);
-	}
-
-	dev_priv->gt_pm.rps.cur_freq = val;
-	trace_intel_gpu_freq_change(intel_gpu_freq(dev_priv, val));
-
-	return 0;
-}
-
-/* vlv_set_rps_idle: Set the frequency to idle, if Gfx clocks are down
- *
- * * If Gfx is Idle, then
- * 1. Forcewake Media well.
- * 2. Request idle freq.
- * 3. Release Forcewake of Media well.
-*/
-static void vlv_set_rps_idle(struct drm_i915_private *dev_priv)
-{
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
-	u32 val = rps->idle_freq;
-	int err;
-
-	if (rps->cur_freq <= val)
-		return;
-
-	/* The punit delays the write of the frequency and voltage until it
-	 * determines the GPU is awake. During normal usage we don't want to
-	 * waste power changing the frequency if the GPU is sleeping (rc6).
-	 * However, the GPU and driver is now idle and we do not want to delay
-	 * switching to minimum voltage (reducing power whilst idle) as we do
-	 * not expect to be woken in the near future and so must flush the
-	 * change by waking the device.
-	 *
-	 * We choose to take the media powerwell (either would do to trick the
-	 * punit into committing the voltage change) as that takes a lot less
-	 * power than the render powerwell.
-	 */
-	intel_uncore_forcewake_get(&dev_priv->uncore, FORCEWAKE_MEDIA);
-	err = valleyview_set_rps(dev_priv, val);
-	intel_uncore_forcewake_put(&dev_priv->uncore, FORCEWAKE_MEDIA);
-
-	if (err)
-		DRM_ERROR("Failed to set RPS for idle\n");
-}
-
-void gen6_rps_busy(struct drm_i915_private *dev_priv)
-{
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
-
-	mutex_lock(&rps->lock);
-	if (rps->enabled) {
-		u8 freq;
-
-		if (dev_priv->pm_rps_events & GEN6_PM_RP_UP_EI_EXPIRED)
-			gen6_rps_reset_ei(dev_priv);
-		I915_WRITE(GEN6_PMINTRMSK,
-			   gen6_rps_pm_mask(dev_priv, rps->cur_freq));
-
-		gen6_enable_rps_interrupts(dev_priv);
-
-		/* Use the user's desired frequency as a guide, but for better
-		 * performance, jump directly to RPe as our starting frequency.
-		 */
-		freq = max(rps->cur_freq,
-			   rps->efficient_freq);
-
-		if (intel_set_rps(dev_priv,
-				  clamp(freq,
-					rps->min_freq_softlimit,
-					rps->max_freq_softlimit)))
-			DRM_DEBUG_DRIVER("Failed to set idle frequency\n");
-	}
-	mutex_unlock(&rps->lock);
-}
-
-void gen6_rps_idle(struct drm_i915_private *dev_priv)
-{
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
-
-	/* Flush our bottom-half so that it does not race with us
-	 * setting the idle frequency and so that it is bounded by
-	 * our rpm wakeref. And then disable the interrupts to stop any
-	 * futher RPS reclocking whilst we are asleep.
-	 */
-	gen6_disable_rps_interrupts(dev_priv);
-
-	mutex_lock(&rps->lock);
-	if (rps->enabled) {
-		if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv))
-			vlv_set_rps_idle(dev_priv);
-		else
-			gen6_set_rps(dev_priv, rps->idle_freq);
-		rps->last_adj = 0;
-		I915_WRITE(GEN6_PMINTRMSK,
-			   gen6_sanitize_rps_pm_mask(dev_priv, ~0));
-	}
-	mutex_unlock(&rps->lock);
-}
-
-void gen6_rps_boost(struct i915_request *rq)
-{
-	struct intel_rps *rps = &rq->i915->gt_pm.rps;
-	unsigned long flags;
-	bool boost;
-
-	/* This is intentionally racy! We peek at the state here, then
-	 * validate inside the RPS worker.
-	 */
-	if (!rps->enabled)
-		return;
-
-	if (i915_request_signaled(rq))
-		return;
-
-	/* Serializes with i915_request_retire() */
-	boost = false;
-	spin_lock_irqsave(&rq->lock, flags);
-	if (!i915_request_has_waitboost(rq) &&
-	    !dma_fence_is_signaled_locked(&rq->fence)) {
-		boost = !atomic_fetch_inc(&rps->num_waiters);
-		rq->flags |= I915_REQUEST_WAITBOOST;
-	}
-	spin_unlock_irqrestore(&rq->lock, flags);
-	if (!boost)
-		return;
-
-	if (READ_ONCE(rps->cur_freq) < rps->boost_freq)
-		schedule_work(&rps->work);
-
-	atomic_inc(&rps->boosts);
-}
-
-int intel_set_rps(struct drm_i915_private *dev_priv, u8 val)
-{
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
-	int err;
-
-	lockdep_assert_held(&rps->lock);
-	GEM_BUG_ON(val > rps->max_freq);
-	GEM_BUG_ON(val < rps->min_freq);
-
-	if (!rps->enabled) {
-		rps->cur_freq = val;
-		return 0;
-	}
-
-	if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv))
-		err = valleyview_set_rps(dev_priv, val);
-	else
-		err = gen6_set_rps(dev_priv, val);
-
-	return err;
-}
-
-static void gen9_disable_rc6(struct drm_i915_private *dev_priv)
-{
-	I915_WRITE(GEN6_RC_CONTROL, 0);
-	I915_WRITE(GEN9_PG_ENABLE, 0);
-}
-
-static void gen9_disable_rps(struct drm_i915_private *dev_priv)
-{
-	I915_WRITE(GEN6_RP_CONTROL, 0);
-}
-
-static void gen6_disable_rc6(struct drm_i915_private *dev_priv)
-{
-	I915_WRITE(GEN6_RC_CONTROL, 0);
-}
-
-static void gen6_disable_rps(struct drm_i915_private *dev_priv)
-{
-	I915_WRITE(GEN6_RPNSWREQ, 1 << 31);
-	I915_WRITE(GEN6_RP_CONTROL, 0);
-}
-
-static void cherryview_disable_rc6(struct drm_i915_private *dev_priv)
-{
-	I915_WRITE(GEN6_RC_CONTROL, 0);
-}
-
-static void cherryview_disable_rps(struct drm_i915_private *dev_priv)
-{
-	I915_WRITE(GEN6_RP_CONTROL, 0);
-}
-
-static void valleyview_disable_rc6(struct drm_i915_private *dev_priv)
-{
-	/* We're doing forcewake before Disabling RC6,
-	 * This what the BIOS expects when going into suspend */
-	intel_uncore_forcewake_get(&dev_priv->uncore, FORCEWAKE_ALL);
-
-	I915_WRITE(GEN6_RC_CONTROL, 0);
-
-	intel_uncore_forcewake_put(&dev_priv->uncore, FORCEWAKE_ALL);
-}
-
-static void valleyview_disable_rps(struct drm_i915_private *dev_priv)
-{
-	I915_WRITE(GEN6_RP_CONTROL, 0);
-}
-
-static bool bxt_check_bios_rc6_setup(struct drm_i915_private *dev_priv)
-{
-	bool enable_rc6 = true;
-	unsigned long rc6_ctx_base;
-	u32 rc_ctl;
-	int rc_sw_target;
-
-	rc_ctl = I915_READ(GEN6_RC_CONTROL);
-	rc_sw_target = (I915_READ(GEN6_RC_STATE) & RC_SW_TARGET_STATE_MASK) >>
-		       RC_SW_TARGET_STATE_SHIFT;
-	DRM_DEBUG_DRIVER("BIOS enabled RC states: "
-			 "HW_CTRL %s HW_RC6 %s SW_TARGET_STATE %x\n",
-			 onoff(rc_ctl & GEN6_RC_CTL_HW_ENABLE),
-			 onoff(rc_ctl & GEN6_RC_CTL_RC6_ENABLE),
-			 rc_sw_target);
-
-	if (!(I915_READ(RC6_LOCATION) & RC6_CTX_IN_DRAM)) {
-		DRM_DEBUG_DRIVER("RC6 Base location not set properly.\n");
-		enable_rc6 = false;
-	}
-
-	/*
-	 * The exact context size is not known for BXT, so assume a page size
-	 * for this check.
-	 */
-	rc6_ctx_base = I915_READ(RC6_CTX_BASE) & RC6_CTX_BASE_MASK;
-	if (!((rc6_ctx_base >= dev_priv->dsm_reserved.start) &&
-	      (rc6_ctx_base + PAGE_SIZE < dev_priv->dsm_reserved.end))) {
-		DRM_DEBUG_DRIVER("RC6 Base address not as expected.\n");
-		enable_rc6 = false;
-	}
-
-	if (!(((I915_READ(PWRCTX_MAXCNT_RCSUNIT) & IDLE_TIME_MASK) > 1) &&
-	      ((I915_READ(PWRCTX_MAXCNT_VCSUNIT0) & IDLE_TIME_MASK) > 1) &&
-	      ((I915_READ(PWRCTX_MAXCNT_BCSUNIT) & IDLE_TIME_MASK) > 1) &&
-	      ((I915_READ(PWRCTX_MAXCNT_VECSUNIT) & IDLE_TIME_MASK) > 1))) {
-		DRM_DEBUG_DRIVER("Engine Idle wait time not set properly.\n");
-		enable_rc6 = false;
-	}
-
-	if (!I915_READ(GEN8_PUSHBUS_CONTROL) ||
-	    !I915_READ(GEN8_PUSHBUS_ENABLE) ||
-	    !I915_READ(GEN8_PUSHBUS_SHIFT)) {
-		DRM_DEBUG_DRIVER("Pushbus not setup properly.\n");
-		enable_rc6 = false;
-	}
-
-	if (!I915_READ(GEN6_GFXPAUSE)) {
-		DRM_DEBUG_DRIVER("GFX pause not setup properly.\n");
-		enable_rc6 = false;
-	}
-
-	if (!I915_READ(GEN8_MISC_CTRL0)) {
-		DRM_DEBUG_DRIVER("GPM control not setup properly.\n");
-		enable_rc6 = false;
-	}
-
-	return enable_rc6;
-}
-
-static bool sanitize_rc6(struct drm_i915_private *i915)
-{
-	struct intel_device_info *info = mkwrite_device_info(i915);
-
-	/* Powersaving is controlled by the host when inside a VM */
-	if (intel_vgpu_active(i915)) {
-		info->has_rc6 = 0;
-		info->has_rps = false;
-	}
-
-	if (info->has_rc6 &&
-	    IS_GEN9_LP(i915) && !bxt_check_bios_rc6_setup(i915)) {
-		DRM_INFO("RC6 disabled by BIOS\n");
-		info->has_rc6 = 0;
-	}
-
-	/*
-	 * We assume that we do not have any deep rc6 levels if we don't have
-	 * have the previous rc6 level supported, i.e. we use HAS_RC6()
-	 * as the initial coarse check for rc6 in general, moving on to
-	 * progressively finer/deeper levels.
-	 */
-	if (!info->has_rc6 && info->has_rc6p)
-		info->has_rc6p = 0;
-
-	return info->has_rc6;
-}
-
-static void gen6_init_rps_frequencies(struct drm_i915_private *dev_priv)
-{
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
-
-	/* All of these values are in units of 50MHz */
-
-	/* static values from HW: RP0 > RP1 > RPn (min_freq) */
-	if (IS_GEN9_LP(dev_priv)) {
-		u32 rp_state_cap = I915_READ(BXT_RP_STATE_CAP);
-		rps->rp0_freq = (rp_state_cap >> 16) & 0xff;
-		rps->rp1_freq = (rp_state_cap >>  8) & 0xff;
-		rps->min_freq = (rp_state_cap >>  0) & 0xff;
-	} else {
-		u32 rp_state_cap = I915_READ(GEN6_RP_STATE_CAP);
-		rps->rp0_freq = (rp_state_cap >>  0) & 0xff;
-		rps->rp1_freq = (rp_state_cap >>  8) & 0xff;
-		rps->min_freq = (rp_state_cap >> 16) & 0xff;
-	}
-	/* hw_max = RP0 until we check for overclocking */
-	rps->max_freq = rps->rp0_freq;
-
-	rps->efficient_freq = rps->rp1_freq;
-	if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv) ||
-	    IS_GEN9_BC(dev_priv) || INTEL_GEN(dev_priv) >= 10) {
-		u32 ddcc_status = 0;
-
-		if (sandybridge_pcode_read(dev_priv,
-					   HSW_PCODE_DYNAMIC_DUTY_CYCLE_CONTROL,
-					   &ddcc_status, NULL) == 0)
-			rps->efficient_freq =
-				clamp_t(u8,
-					((ddcc_status >> 8) & 0xff),
-					rps->min_freq,
-					rps->max_freq);
-	}
-
-	if (IS_GEN9_BC(dev_priv) || INTEL_GEN(dev_priv) >= 10) {
-		/* Store the frequency values in 16.66 MHZ units, which is
-		 * the natural hardware unit for SKL
-		 */
-		rps->rp0_freq *= GEN9_FREQ_SCALER;
-		rps->rp1_freq *= GEN9_FREQ_SCALER;
-		rps->min_freq *= GEN9_FREQ_SCALER;
-		rps->max_freq *= GEN9_FREQ_SCALER;
-		rps->efficient_freq *= GEN9_FREQ_SCALER;
-	}
-}
-
-static void reset_rps(struct drm_i915_private *dev_priv,
-		      int (*set)(struct drm_i915_private *, u8))
-{
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
-	u8 freq = rps->cur_freq;
-
-	/* force a reset */
-	rps->power.mode = -1;
-	rps->cur_freq = -1;
-
-	if (set(dev_priv, freq))
-		DRM_ERROR("Failed to reset RPS to initial values\n");
-}
-
-/* See the Gen9_GT_PM_Programming_Guide doc for the below */
-static void gen9_enable_rps(struct drm_i915_private *dev_priv)
-{
-	intel_uncore_forcewake_get(&dev_priv->uncore, FORCEWAKE_ALL);
-
-	/* Program defaults and thresholds for RPS */
-	if (IS_GEN(dev_priv, 9))
-		I915_WRITE(GEN6_RC_VIDEO_FREQ,
-			GEN9_FREQUENCY(dev_priv->gt_pm.rps.rp1_freq));
-
-	/* 1 second timeout*/
-	I915_WRITE(GEN6_RP_DOWN_TIMEOUT,
-		GT_INTERVAL_FROM_US(dev_priv, 1000000));
-
-	I915_WRITE(GEN6_RP_IDLE_HYSTERSIS, 0xa);
-
-	/* Leaning on the below call to gen6_set_rps to program/setup the
-	 * Up/Down EI & threshold registers, as well as the RP_CONTROL,
-	 * RP_INTERRUPT_LIMITS & RPNSWREQ registers */
-	reset_rps(dev_priv, gen6_set_rps);
-
-	intel_uncore_forcewake_put(&dev_priv->uncore, FORCEWAKE_ALL);
-}
-
-static void gen11_enable_rc6(struct drm_i915_private *dev_priv)
-{
-	struct intel_engine_cs *engine;
-	enum intel_engine_id id;
-
-	/* 1a: Software RC state - RC0 */
-	I915_WRITE(GEN6_RC_STATE, 0);
-
-	/*
-	 * 1b: Get forcewake during program sequence. Although the driver
-	 * hasn't enabled a state yet where we need forcewake, BIOS may have.
-	 */
-	intel_uncore_forcewake_get(&dev_priv->uncore, FORCEWAKE_ALL);
-
-	/* 2a: Disable RC states. */
-	I915_WRITE(GEN6_RC_CONTROL, 0);
-
-	/* 2b: Program RC6 thresholds.*/
-	I915_WRITE(GEN6_RC6_WAKE_RATE_LIMIT, 54 << 16 | 85);
-	I915_WRITE(GEN10_MEDIA_WAKE_RATE_LIMIT, 150);
-
-	I915_WRITE(GEN6_RC_EVALUATION_INTERVAL, 125000); /* 12500 * 1280ns */
-	I915_WRITE(GEN6_RC_IDLE_HYSTERSIS, 25); /* 25 * 1280ns */
-	for_each_engine(engine, dev_priv, id)
-		I915_WRITE(RING_MAX_IDLE(engine->mmio_base), 10);
-
-	if (HAS_GT_UC(dev_priv))
-		I915_WRITE(GUC_MAX_IDLE_COUNT, 0xA);
-
-	I915_WRITE(GEN6_RC_SLEEP, 0);
-
-	I915_WRITE(GEN6_RC6_THRESHOLD, 50000); /* 50/125ms per EI */
-
-	/*
-	 * 2c: Program Coarse Power Gating Policies.
-	 *
-	 * Bspec's guidance is to use 25us (really 25 * 1280ns) here. What we
-	 * use instead is a more conservative estimate for the maximum time
-	 * it takes us to service a CS interrupt and submit a new ELSP - that
-	 * is the time which the GPU is idle waiting for the CPU to select the
-	 * next request to execute. If the idle hysteresis is less than that
-	 * interrupt service latency, the hardware will automatically gate
-	 * the power well and we will then incur the wake up cost on top of
-	 * the service latency. A similar guide from plane_state is that we
-	 * do not want the enable hysteresis to less than the wakeup latency.
-	 *
-	 * igt/gem_exec_nop/sequential provides a rough estimate for the
-	 * service latency, and puts it around 10us for Broadwell (and other
-	 * big core) and around 40us for Broxton (and other low power cores).
-	 * [Note that for legacy ringbuffer submission, this is less than 1us!]
-	 * However, the wakeup latency on Broxton is closer to 100us. To be
-	 * conservative, we have to factor in a context switch on top (due
-	 * to ksoftirqd).
-	 */
-	I915_WRITE(GEN9_MEDIA_PG_IDLE_HYSTERESIS, 250);
-	I915_WRITE(GEN9_RENDER_PG_IDLE_HYSTERESIS, 250);
-
-	/* 3a: Enable RC6 */
-	I915_WRITE(GEN6_RC_CONTROL,
-		   GEN6_RC_CTL_HW_ENABLE |
-		   GEN6_RC_CTL_RC6_ENABLE |
-		   GEN6_RC_CTL_EI_MODE(1));
-
-	/* 3b: Enable Coarse Power Gating only when RC6 is enabled. */
-	I915_WRITE(GEN9_PG_ENABLE,
-		   GEN9_RENDER_PG_ENABLE |
-		   GEN9_MEDIA_PG_ENABLE |
-		   GEN11_MEDIA_SAMPLER_PG_ENABLE);
-
-	intel_uncore_forcewake_put(&dev_priv->uncore, FORCEWAKE_ALL);
-}
-
-static void gen9_enable_rc6(struct drm_i915_private *dev_priv)
-{
-	struct intel_engine_cs *engine;
-	enum intel_engine_id id;
-	u32 rc6_mode;
-
-	/* 1a: Software RC state - RC0 */
-	I915_WRITE(GEN6_RC_STATE, 0);
-
-	/* 1b: Get forcewake during program sequence. Although the driver
-	 * hasn't enabled a state yet where we need forcewake, BIOS may have.*/
-	intel_uncore_forcewake_get(&dev_priv->uncore, FORCEWAKE_ALL);
-
-	/* 2a: Disable RC states. */
-	I915_WRITE(GEN6_RC_CONTROL, 0);
-
-	/* 2b: Program RC6 thresholds.*/
-	if (INTEL_GEN(dev_priv) >= 10) {
-		I915_WRITE(GEN6_RC6_WAKE_RATE_LIMIT, 54 << 16 | 85);
-		I915_WRITE(GEN10_MEDIA_WAKE_RATE_LIMIT, 150);
-	} else if (IS_SKYLAKE(dev_priv)) {
-		/*
-		 * WaRsDoubleRc6WrlWithCoarsePowerGating:skl Doubling WRL only
-		 * when CPG is enabled
-		 */
-		I915_WRITE(GEN6_RC6_WAKE_RATE_LIMIT, 108 << 16);
-	} else {
-		I915_WRITE(GEN6_RC6_WAKE_RATE_LIMIT, 54 << 16);
-	}
-
-	I915_WRITE(GEN6_RC_EVALUATION_INTERVAL, 125000); /* 12500 * 1280ns */
-	I915_WRITE(GEN6_RC_IDLE_HYSTERSIS, 25); /* 25 * 1280ns */
-	for_each_engine(engine, dev_priv, id)
-		I915_WRITE(RING_MAX_IDLE(engine->mmio_base), 10);
-
-	if (HAS_GT_UC(dev_priv))
-		I915_WRITE(GUC_MAX_IDLE_COUNT, 0xA);
-
-	I915_WRITE(GEN6_RC_SLEEP, 0);
-
-	/*
-	 * 2c: Program Coarse Power Gating Policies.
-	 *
-	 * Bspec's guidance is to use 25us (really 25 * 1280ns) here. What we
-	 * use instead is a more conservative estimate for the maximum time
-	 * it takes us to service a CS interrupt and submit a new ELSP - that
-	 * is the time which the GPU is idle waiting for the CPU to select the
-	 * next request to execute. If the idle hysteresis is less than that
-	 * interrupt service latency, the hardware will automatically gate
-	 * the power well and we will then incur the wake up cost on top of
-	 * the service latency. A similar guide from plane_state is that we
-	 * do not want the enable hysteresis to less than the wakeup latency.
-	 *
-	 * igt/gem_exec_nop/sequential provides a rough estimate for the
-	 * service latency, and puts it around 10us for Broadwell (and other
-	 * big core) and around 40us for Broxton (and other low power cores).
-	 * [Note that for legacy ringbuffer submission, this is less than 1us!]
-	 * However, the wakeup latency on Broxton is closer to 100us. To be
-	 * conservative, we have to factor in a context switch on top (due
-	 * to ksoftirqd).
-	 */
-	I915_WRITE(GEN9_MEDIA_PG_IDLE_HYSTERESIS, 250);
-	I915_WRITE(GEN9_RENDER_PG_IDLE_HYSTERESIS, 250);
-
-	/* 3a: Enable RC6 */
-	I915_WRITE(GEN6_RC6_THRESHOLD, 37500); /* 37.5/125ms per EI */
-
-	/* WaRsUseTimeoutMode:cnl (pre-prod) */
-	if (IS_CNL_REVID(dev_priv, CNL_REVID_A0, CNL_REVID_C0))
-		rc6_mode = GEN7_RC_CTL_TO_MODE;
-	else
-		rc6_mode = GEN6_RC_CTL_EI_MODE(1);
-
-	I915_WRITE(GEN6_RC_CONTROL,
-		   GEN6_RC_CTL_HW_ENABLE |
-		   GEN6_RC_CTL_RC6_ENABLE |
-		   rc6_mode);
-
-	/*
-	 * 3b: Enable Coarse Power Gating only when RC6 is enabled.
-	 * WaRsDisableCoarsePowerGating:skl,cnl - Render/Media PG need to be disabled with RC6.
-	 */
-	if (NEEDS_WaRsDisableCoarsePowerGating(dev_priv))
-		I915_WRITE(GEN9_PG_ENABLE, 0);
-	else
-		I915_WRITE(GEN9_PG_ENABLE,
-			   GEN9_RENDER_PG_ENABLE | GEN9_MEDIA_PG_ENABLE);
-
-	intel_uncore_forcewake_put(&dev_priv->uncore, FORCEWAKE_ALL);
-}
-
-static void gen8_enable_rc6(struct drm_i915_private *dev_priv)
-{
-	struct intel_engine_cs *engine;
-	enum intel_engine_id id;
-
-	/* 1a: Software RC state - RC0 */
-	I915_WRITE(GEN6_RC_STATE, 0);
-
-	/* 1b: Get forcewake during program sequence. Although the driver
-	 * hasn't enabled a state yet where we need forcewake, BIOS may have.*/
-	intel_uncore_forcewake_get(&dev_priv->uncore, FORCEWAKE_ALL);
-
-	/* 2a: Disable RC states. */
-	I915_WRITE(GEN6_RC_CONTROL, 0);
-
-	/* 2b: Program RC6 thresholds.*/
-	I915_WRITE(GEN6_RC6_WAKE_RATE_LIMIT, 40 << 16);
-	I915_WRITE(GEN6_RC_EVALUATION_INTERVAL, 125000); /* 12500 * 1280ns */
-	I915_WRITE(GEN6_RC_IDLE_HYSTERSIS, 25); /* 25 * 1280ns */
-	for_each_engine(engine, dev_priv, id)
-		I915_WRITE(RING_MAX_IDLE(engine->mmio_base), 10);
-	I915_WRITE(GEN6_RC_SLEEP, 0);
-	I915_WRITE(GEN6_RC6_THRESHOLD, 625); /* 800us/1.28 for TO */
-
-	/* 3: Enable RC6 */
-
-	I915_WRITE(GEN6_RC_CONTROL,
-		   GEN6_RC_CTL_HW_ENABLE |
-		   GEN7_RC_CTL_TO_MODE |
-		   GEN6_RC_CTL_RC6_ENABLE);
-
-	intel_uncore_forcewake_put(&dev_priv->uncore, FORCEWAKE_ALL);
-}
-
-static void gen8_enable_rps(struct drm_i915_private *dev_priv)
-{
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
-
-	intel_uncore_forcewake_get(&dev_priv->uncore, FORCEWAKE_ALL);
-
-	/* 1 Program defaults and thresholds for RPS*/
-	I915_WRITE(GEN6_RPNSWREQ,
-		   HSW_FREQUENCY(rps->rp1_freq));
-	I915_WRITE(GEN6_RC_VIDEO_FREQ,
-		   HSW_FREQUENCY(rps->rp1_freq));
-	/* NB: Docs say 1s, and 1000000 - which aren't equivalent */
-	I915_WRITE(GEN6_RP_DOWN_TIMEOUT, 100000000 / 128); /* 1 second timeout */
-
-	/* Docs recommend 900MHz, and 300 MHz respectively */
-	I915_WRITE(GEN6_RP_INTERRUPT_LIMITS,
-		   rps->max_freq_softlimit << 24 |
-		   rps->min_freq_softlimit << 16);
-
-	I915_WRITE(GEN6_RP_UP_THRESHOLD, 7600000 / 128); /* 76ms busyness per EI, 90% */
-	I915_WRITE(GEN6_RP_DOWN_THRESHOLD, 31300000 / 128); /* 313ms busyness per EI, 70%*/
-	I915_WRITE(GEN6_RP_UP_EI, 66000); /* 84.48ms, XXX: random? */
-	I915_WRITE(GEN6_RP_DOWN_EI, 350000); /* 448ms, XXX: random? */
-
-	I915_WRITE(GEN6_RP_IDLE_HYSTERSIS, 10);
-
-	/* 2: Enable RPS */
-	I915_WRITE(GEN6_RP_CONTROL,
-		   GEN6_RP_MEDIA_TURBO |
-		   GEN6_RP_MEDIA_HW_NORMAL_MODE |
-		   GEN6_RP_MEDIA_IS_GFX |
-		   GEN6_RP_ENABLE |
-		   GEN6_RP_UP_BUSY_AVG |
-		   GEN6_RP_DOWN_IDLE_AVG);
-
-	reset_rps(dev_priv, gen6_set_rps);
-
-	intel_uncore_forcewake_put(&dev_priv->uncore, FORCEWAKE_ALL);
-}
-
-static void gen6_enable_rc6(struct drm_i915_private *dev_priv)
-{
-	struct intel_engine_cs *engine;
-	enum intel_engine_id id;
-	u32 rc6vids, rc6_mask;
-	u32 gtfifodbg;
-	int ret;
-
-	I915_WRITE(GEN6_RC_STATE, 0);
-
-	/* Clear the DBG now so we don't confuse earlier errors */
-	gtfifodbg = I915_READ(GTFIFODBG);
-	if (gtfifodbg) {
-		DRM_ERROR("GT fifo had a previous error %x\n", gtfifodbg);
-		I915_WRITE(GTFIFODBG, gtfifodbg);
-	}
-
-	intel_uncore_forcewake_get(&dev_priv->uncore, FORCEWAKE_ALL);
-
-	/* disable the counters and set deterministic thresholds */
-	I915_WRITE(GEN6_RC_CONTROL, 0);
-
-	I915_WRITE(GEN6_RC1_WAKE_RATE_LIMIT, 1000 << 16);
-	I915_WRITE(GEN6_RC6_WAKE_RATE_LIMIT, 40 << 16 | 30);
-	I915_WRITE(GEN6_RC6pp_WAKE_RATE_LIMIT, 30);
-	I915_WRITE(GEN6_RC_EVALUATION_INTERVAL, 125000);
-	I915_WRITE(GEN6_RC_IDLE_HYSTERSIS, 25);
-
-	for_each_engine(engine, dev_priv, id)
-		I915_WRITE(RING_MAX_IDLE(engine->mmio_base), 10);
-
-	I915_WRITE(GEN6_RC_SLEEP, 0);
-	I915_WRITE(GEN6_RC1e_THRESHOLD, 1000);
-	if (IS_IVYBRIDGE(dev_priv))
-		I915_WRITE(GEN6_RC6_THRESHOLD, 125000);
-	else
-		I915_WRITE(GEN6_RC6_THRESHOLD, 50000);
-	I915_WRITE(GEN6_RC6p_THRESHOLD, 150000);
-	I915_WRITE(GEN6_RC6pp_THRESHOLD, 64000); /* unused */
-
-	/* We don't use those on Haswell */
-	rc6_mask = GEN6_RC_CTL_RC6_ENABLE;
-	if (HAS_RC6p(dev_priv))
-		rc6_mask |= GEN6_RC_CTL_RC6p_ENABLE;
-	if (HAS_RC6pp(dev_priv))
-		rc6_mask |= GEN6_RC_CTL_RC6pp_ENABLE;
-	I915_WRITE(GEN6_RC_CONTROL,
-		   rc6_mask |
-		   GEN6_RC_CTL_EI_MODE(1) |
-		   GEN6_RC_CTL_HW_ENABLE);
-
-	rc6vids = 0;
-	ret = sandybridge_pcode_read(dev_priv, GEN6_PCODE_READ_RC6VIDS,
-				     &rc6vids, NULL);
-	if (IS_GEN(dev_priv, 6) && ret) {
-		DRM_DEBUG_DRIVER("Couldn't check for BIOS workaround\n");
-	} else if (IS_GEN(dev_priv, 6) && (GEN6_DECODE_RC6_VID(rc6vids & 0xff) < 450)) {
-		DRM_DEBUG_DRIVER("You should update your BIOS. Correcting minimum rc6 voltage (%dmV->%dmV)\n",
-			  GEN6_DECODE_RC6_VID(rc6vids & 0xff), 450);
-		rc6vids &= 0xffff00;
-		rc6vids |= GEN6_ENCODE_RC6_VID(450);
-		ret = sandybridge_pcode_write(dev_priv, GEN6_PCODE_WRITE_RC6VIDS, rc6vids);
-		if (ret)
-			DRM_ERROR("Couldn't fix incorrect rc6 voltage\n");
-	}
-
-	intel_uncore_forcewake_put(&dev_priv->uncore, FORCEWAKE_ALL);
-}
-
-static void gen6_enable_rps(struct drm_i915_private *dev_priv)
-{
-	/* Here begins a magic sequence of register writes to enable
-	 * auto-downclocking.
-	 *
-	 * Perhaps there might be some value in exposing these to
-	 * userspace...
-	 */
-	intel_uncore_forcewake_get(&dev_priv->uncore, FORCEWAKE_ALL);
-
-	/* Power down if completely idle for over 50ms */
-	I915_WRITE(GEN6_RP_DOWN_TIMEOUT, 50000);
-	I915_WRITE(GEN6_RP_IDLE_HYSTERSIS, 10);
-
-	reset_rps(dev_priv, gen6_set_rps);
-
-	intel_uncore_forcewake_put(&dev_priv->uncore, FORCEWAKE_ALL);
-}
-
-static void gen6_update_ring_freq(struct drm_i915_private *dev_priv)
-{
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
-	const int min_freq = 15;
-	const int scaling_factor = 180;
-	unsigned int gpu_freq;
-	unsigned int max_ia_freq, min_ring_freq;
-	unsigned int max_gpu_freq, min_gpu_freq;
-	struct cpufreq_policy *policy;
-
-	lockdep_assert_held(&rps->lock);
-
-	if (rps->max_freq <= rps->min_freq)
-		return;
-
-	policy = cpufreq_cpu_get(0);
-	if (policy) {
-		max_ia_freq = policy->cpuinfo.max_freq;
-		cpufreq_cpu_put(policy);
-	} else {
-		/*
-		 * Default to measured freq if none found, PCU will ensure we
-		 * don't go over
-		 */
-		max_ia_freq = tsc_khz;
-	}
-
-	/* Convert from kHz to MHz */
-	max_ia_freq /= 1000;
-
-	min_ring_freq = I915_READ(DCLK) & 0xf;
-	/* convert DDR frequency from units of 266.6MHz to bandwidth */
-	min_ring_freq = mult_frac(min_ring_freq, 8, 3);
-
-	min_gpu_freq = rps->min_freq;
-	max_gpu_freq = rps->max_freq;
-	if (IS_GEN9_BC(dev_priv) || INTEL_GEN(dev_priv) >= 10) {
-		/* Convert GT frequency to 50 HZ units */
-		min_gpu_freq /= GEN9_FREQ_SCALER;
-		max_gpu_freq /= GEN9_FREQ_SCALER;
-	}
-
-	/*
-	 * For each potential GPU frequency, load a ring frequency we'd like
-	 * to use for memory access.  We do this by specifying the IA frequency
-	 * the PCU should use as a reference to determine the ring frequency.
-	 */
-	for (gpu_freq = max_gpu_freq; gpu_freq >= min_gpu_freq; gpu_freq--) {
-		const int diff = max_gpu_freq - gpu_freq;
-		unsigned int ia_freq = 0, ring_freq = 0;
-
-		if (IS_GEN9_BC(dev_priv) || INTEL_GEN(dev_priv) >= 10) {
-			/*
-			 * ring_freq = 2 * GT. ring_freq is in 100MHz units
-			 * No floor required for ring frequency on SKL.
-			 */
-			ring_freq = gpu_freq;
-		} else if (INTEL_GEN(dev_priv) >= 8) {
-			/* max(2 * GT, DDR). NB: GT is 50MHz units */
-			ring_freq = max(min_ring_freq, gpu_freq);
-		} else if (IS_HASWELL(dev_priv)) {
-			ring_freq = mult_frac(gpu_freq, 5, 4);
-			ring_freq = max(min_ring_freq, ring_freq);
-			/* leave ia_freq as the default, chosen by cpufreq */
-		} else {
-			/* On older processors, there is no separate ring
-			 * clock domain, so in order to boost the bandwidth
-			 * of the ring, we need to upclock the CPU (ia_freq).
-			 *
-			 * For GPU frequencies less than 750MHz,
-			 * just use the lowest ring freq.
-			 */
-			if (gpu_freq < min_freq)
-				ia_freq = 800;
-			else
-				ia_freq = max_ia_freq - ((diff * scaling_factor) / 2);
-			ia_freq = DIV_ROUND_CLOSEST(ia_freq, 100);
-		}
-
-		sandybridge_pcode_write(dev_priv,
-					GEN6_PCODE_WRITE_MIN_FREQ_TABLE,
-					ia_freq << GEN6_PCODE_FREQ_IA_RATIO_SHIFT |
-					ring_freq << GEN6_PCODE_FREQ_RING_RATIO_SHIFT |
-					gpu_freq);
-	}
-}
-
-static int cherryview_rps_max_freq(struct drm_i915_private *dev_priv)
-{
-	u32 val, rp0;
-
-	val = vlv_punit_read(dev_priv, FB_GFX_FMAX_AT_VMAX_FUSE);
-
-	switch (RUNTIME_INFO(dev_priv)->sseu.eu_total) {
-	case 8:
-		/* (2 * 4) config */
-		rp0 = (val >> FB_GFX_FMAX_AT_VMAX_2SS4EU_FUSE_SHIFT);
-		break;
-	case 12:
-		/* (2 * 6) config */
-		rp0 = (val >> FB_GFX_FMAX_AT_VMAX_2SS6EU_FUSE_SHIFT);
-		break;
-	case 16:
-		/* (2 * 8) config */
-	default:
-		/* Setting (2 * 8) Min RP0 for any other combination */
-		rp0 = (val >> FB_GFX_FMAX_AT_VMAX_2SS8EU_FUSE_SHIFT);
-		break;
-	}
-
-	rp0 = (rp0 & FB_GFX_FREQ_FUSE_MASK);
-
-	return rp0;
-}
-
-static int cherryview_rps_rpe_freq(struct drm_i915_private *dev_priv)
-{
-	u32 val, rpe;
-
-	val = vlv_punit_read(dev_priv, PUNIT_GPU_DUTYCYCLE_REG);
-	rpe = (val >> PUNIT_GPU_DUTYCYCLE_RPE_FREQ_SHIFT) & PUNIT_GPU_DUTYCYCLE_RPE_FREQ_MASK;
-
-	return rpe;
-}
-
-static int cherryview_rps_guar_freq(struct drm_i915_private *dev_priv)
-{
-	u32 val, rp1;
-
-	val = vlv_punit_read(dev_priv, FB_GFX_FMAX_AT_VMAX_FUSE);
-	rp1 = (val & FB_GFX_FREQ_FUSE_MASK);
-
-	return rp1;
-}
-
-static u32 cherryview_rps_min_freq(struct drm_i915_private *dev_priv)
-{
-	u32 val, rpn;
-
-	val = vlv_punit_read(dev_priv, FB_GFX_FMIN_AT_VMIN_FUSE);
-	rpn = ((val >> FB_GFX_FMIN_AT_VMIN_FUSE_SHIFT) &
-		       FB_GFX_FREQ_FUSE_MASK);
-
-	return rpn;
-}
-
-static int valleyview_rps_guar_freq(struct drm_i915_private *dev_priv)
-{
-	u32 val, rp1;
-
-	val = vlv_nc_read(dev_priv, IOSF_NC_FB_GFX_FREQ_FUSE);
-
-	rp1 = (val & FB_GFX_FGUARANTEED_FREQ_FUSE_MASK) >> FB_GFX_FGUARANTEED_FREQ_FUSE_SHIFT;
-
-	return rp1;
-}
-
-static int valleyview_rps_max_freq(struct drm_i915_private *dev_priv)
-{
-	u32 val, rp0;
-
-	val = vlv_nc_read(dev_priv, IOSF_NC_FB_GFX_FREQ_FUSE);
-
-	rp0 = (val & FB_GFX_MAX_FREQ_FUSE_MASK) >> FB_GFX_MAX_FREQ_FUSE_SHIFT;
-	/* Clamp to max */
-	rp0 = min_t(u32, rp0, 0xea);
-
-	return rp0;
-}
-
-static int valleyview_rps_rpe_freq(struct drm_i915_private *dev_priv)
-{
-	u32 val, rpe;
-
-	val = vlv_nc_read(dev_priv, IOSF_NC_FB_GFX_FMAX_FUSE_LO);
-	rpe = (val & FB_FMAX_VMIN_FREQ_LO_MASK) >> FB_FMAX_VMIN_FREQ_LO_SHIFT;
-	val = vlv_nc_read(dev_priv, IOSF_NC_FB_GFX_FMAX_FUSE_HI);
-	rpe |= (val & FB_FMAX_VMIN_FREQ_HI_MASK) << 5;
-
-	return rpe;
-}
-
-static int valleyview_rps_min_freq(struct drm_i915_private *dev_priv)
-{
-	u32 val;
-
-	val = vlv_punit_read(dev_priv, PUNIT_REG_GPU_LFM) & 0xff;
-	/*
-	 * According to the BYT Punit GPU turbo HAS 1.1.6.3 the minimum value
-	 * for the minimum frequency in GPLL mode is 0xc1. Contrary to this on
-	 * a BYT-M B0 the above register contains 0xbf. Moreover when setting
-	 * a frequency Punit will not allow values below 0xc0. Clamp it 0xc0
-	 * to make sure it matches what Punit accepts.
-	 */
-	return max_t(u32, val, 0xc0);
-}
-
-/* Check that the pctx buffer wasn't move under us. */
-static void valleyview_check_pctx(struct drm_i915_private *dev_priv)
-{
-	unsigned long pctx_addr = I915_READ(VLV_PCBR) & ~4095;
-
-	WARN_ON(pctx_addr != dev_priv->dsm.start +
-			     dev_priv->vlv_pctx->stolen->start);
-}
-
-
-/* Check that the pcbr address is not empty. */
-static void cherryview_check_pctx(struct drm_i915_private *dev_priv)
-{
-	unsigned long pctx_addr = I915_READ(VLV_PCBR) & ~4095;
-
-	WARN_ON((pctx_addr >> VLV_PCBR_ADDR_SHIFT) == 0);
-}
-
-static void cherryview_setup_pctx(struct drm_i915_private *dev_priv)
-{
-	resource_size_t pctx_paddr, paddr;
-	resource_size_t pctx_size = 32*1024;
-	u32 pcbr;
-
-	pcbr = I915_READ(VLV_PCBR);
-	if ((pcbr >> VLV_PCBR_ADDR_SHIFT) == 0) {
-		DRM_DEBUG_DRIVER("BIOS didn't set up PCBR, fixing up\n");
-		paddr = dev_priv->dsm.end + 1 - pctx_size;
-		GEM_BUG_ON(paddr > U32_MAX);
-
-		pctx_paddr = (paddr & (~4095));
-		I915_WRITE(VLV_PCBR, pctx_paddr);
-	}
-
-	DRM_DEBUG_DRIVER("PCBR: 0x%08x\n", I915_READ(VLV_PCBR));
-}
-
-static void valleyview_setup_pctx(struct drm_i915_private *dev_priv)
-{
-	struct drm_i915_gem_object *pctx;
-	resource_size_t pctx_paddr;
-	resource_size_t pctx_size = 24*1024;
-	u32 pcbr;
-
-	pcbr = I915_READ(VLV_PCBR);
-	if (pcbr) {
-		/* BIOS set it up already, grab the pre-alloc'd space */
-		resource_size_t pcbr_offset;
-
-		pcbr_offset = (pcbr & (~4095)) - dev_priv->dsm.start;
-		pctx = i915_gem_object_create_stolen_for_preallocated(dev_priv,
-								      pcbr_offset,
-								      I915_GTT_OFFSET_NONE,
-								      pctx_size);
-		goto out;
-	}
-
-	DRM_DEBUG_DRIVER("BIOS didn't set up PCBR, fixing up\n");
-
-	/*
-	 * From the Gunit register HAS:
-	 * The Gfx driver is expected to program this register and ensure
-	 * proper allocation within Gfx stolen memory.  For example, this
-	 * register should be programmed such than the PCBR range does not
-	 * overlap with other ranges, such as the frame buffer, protected
-	 * memory, or any other relevant ranges.
-	 */
-	pctx = i915_gem_object_create_stolen(dev_priv, pctx_size);
-	if (!pctx) {
-		DRM_DEBUG("not enough stolen space for PCTX, disabling\n");
-		goto out;
-	}
-
-	GEM_BUG_ON(range_overflows_t(u64,
-				     dev_priv->dsm.start,
-				     pctx->stolen->start,
-				     U32_MAX));
-	pctx_paddr = dev_priv->dsm.start + pctx->stolen->start;
-	I915_WRITE(VLV_PCBR, pctx_paddr);
-
-out:
-	DRM_DEBUG_DRIVER("PCBR: 0x%08x\n", I915_READ(VLV_PCBR));
-	dev_priv->vlv_pctx = pctx;
-}
-
-static void valleyview_cleanup_pctx(struct drm_i915_private *dev_priv)
-{
-	struct drm_i915_gem_object *pctx;
-
-	pctx = fetch_and_zero(&dev_priv->vlv_pctx);
-	if (pctx)
-		i915_gem_object_put(pctx);
-}
-
-static void vlv_init_gpll_ref_freq(struct drm_i915_private *dev_priv)
-{
-	dev_priv->gt_pm.rps.gpll_ref_freq =
-		vlv_get_cck_clock(dev_priv, "GPLL ref",
-				  CCK_GPLL_CLOCK_CONTROL,
-				  dev_priv->czclk_freq);
-
-	DRM_DEBUG_DRIVER("GPLL reference freq: %d kHz\n",
-			 dev_priv->gt_pm.rps.gpll_ref_freq);
-}
-
-static void valleyview_init_gt_powersave(struct drm_i915_private *dev_priv)
-{
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
-	u32 val;
-
-	valleyview_setup_pctx(dev_priv);
-
-	vlv_iosf_sb_get(dev_priv,
-			BIT(VLV_IOSF_SB_PUNIT) |
-			BIT(VLV_IOSF_SB_NC) |
-			BIT(VLV_IOSF_SB_CCK));
-
-	vlv_init_gpll_ref_freq(dev_priv);
-
-	val = vlv_punit_read(dev_priv, PUNIT_REG_GPU_FREQ_STS);
-	switch ((val >> 6) & 3) {
-	case 0:
-	case 1:
-		dev_priv->mem_freq = 800;
-		break;
-	case 2:
-		dev_priv->mem_freq = 1066;
-		break;
-	case 3:
-		dev_priv->mem_freq = 1333;
-		break;
-	}
-	DRM_DEBUG_DRIVER("DDR speed: %d MHz\n", dev_priv->mem_freq);
-
-	rps->max_freq = valleyview_rps_max_freq(dev_priv);
-	rps->rp0_freq = rps->max_freq;
-	DRM_DEBUG_DRIVER("max GPU freq: %d MHz (%u)\n",
-			 intel_gpu_freq(dev_priv, rps->max_freq),
-			 rps->max_freq);
-
-	rps->efficient_freq = valleyview_rps_rpe_freq(dev_priv);
-	DRM_DEBUG_DRIVER("RPe GPU freq: %d MHz (%u)\n",
-			 intel_gpu_freq(dev_priv, rps->efficient_freq),
-			 rps->efficient_freq);
-
-	rps->rp1_freq = valleyview_rps_guar_freq(dev_priv);
-	DRM_DEBUG_DRIVER("RP1(Guar Freq) GPU freq: %d MHz (%u)\n",
-			 intel_gpu_freq(dev_priv, rps->rp1_freq),
-			 rps->rp1_freq);
-
-	rps->min_freq = valleyview_rps_min_freq(dev_priv);
-	DRM_DEBUG_DRIVER("min GPU freq: %d MHz (%u)\n",
-			 intel_gpu_freq(dev_priv, rps->min_freq),
-			 rps->min_freq);
-
-	vlv_iosf_sb_put(dev_priv,
-			BIT(VLV_IOSF_SB_PUNIT) |
-			BIT(VLV_IOSF_SB_NC) |
-			BIT(VLV_IOSF_SB_CCK));
-}
-
-static void cherryview_init_gt_powersave(struct drm_i915_private *dev_priv)
-{
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
-	u32 val;
-
-	cherryview_setup_pctx(dev_priv);
-
-	vlv_iosf_sb_get(dev_priv,
-			BIT(VLV_IOSF_SB_PUNIT) |
-			BIT(VLV_IOSF_SB_NC) |
-			BIT(VLV_IOSF_SB_CCK));
-
-	vlv_init_gpll_ref_freq(dev_priv);
-
-	val = vlv_cck_read(dev_priv, CCK_FUSE_REG);
-
-	switch ((val >> 2) & 0x7) {
-	case 3:
-		dev_priv->mem_freq = 2000;
-		break;
-	default:
-		dev_priv->mem_freq = 1600;
-		break;
-	}
-	DRM_DEBUG_DRIVER("DDR speed: %d MHz\n", dev_priv->mem_freq);
-
-	rps->max_freq = cherryview_rps_max_freq(dev_priv);
-	rps->rp0_freq = rps->max_freq;
-	DRM_DEBUG_DRIVER("max GPU freq: %d MHz (%u)\n",
-			 intel_gpu_freq(dev_priv, rps->max_freq),
-			 rps->max_freq);
-
-	rps->efficient_freq = cherryview_rps_rpe_freq(dev_priv);
-	DRM_DEBUG_DRIVER("RPe GPU freq: %d MHz (%u)\n",
-			 intel_gpu_freq(dev_priv, rps->efficient_freq),
-			 rps->efficient_freq);
-
-	rps->rp1_freq = cherryview_rps_guar_freq(dev_priv);
-	DRM_DEBUG_DRIVER("RP1(Guar) GPU freq: %d MHz (%u)\n",
-			 intel_gpu_freq(dev_priv, rps->rp1_freq),
-			 rps->rp1_freq);
-
-	rps->min_freq = cherryview_rps_min_freq(dev_priv);
-	DRM_DEBUG_DRIVER("min GPU freq: %d MHz (%u)\n",
-			 intel_gpu_freq(dev_priv, rps->min_freq),
-			 rps->min_freq);
-
-	vlv_iosf_sb_put(dev_priv,
-			BIT(VLV_IOSF_SB_PUNIT) |
-			BIT(VLV_IOSF_SB_NC) |
-			BIT(VLV_IOSF_SB_CCK));
-
-	WARN_ONCE((rps->max_freq | rps->efficient_freq | rps->rp1_freq |
-		   rps->min_freq) & 1,
-		  "Odd GPU freq values\n");
-}
-
-static void valleyview_cleanup_gt_powersave(struct drm_i915_private *dev_priv)
-{
-	valleyview_cleanup_pctx(dev_priv);
-}
-
-static void cherryview_enable_rc6(struct drm_i915_private *dev_priv)
-{
-	struct intel_engine_cs *engine;
-	enum intel_engine_id id;
-	u32 gtfifodbg, rc6_mode, pcbr;
-
-	gtfifodbg = I915_READ(GTFIFODBG) & ~(GT_FIFO_SBDEDICATE_FREE_ENTRY_CHV |
-					     GT_FIFO_FREE_ENTRIES_CHV);
-	if (gtfifodbg) {
-		DRM_DEBUG_DRIVER("GT fifo had a previous error %x\n",
-				 gtfifodbg);
-		I915_WRITE(GTFIFODBG, gtfifodbg);
-	}
-
-	cherryview_check_pctx(dev_priv);
-
-	/* 1a & 1b: Get forcewake during program sequence. Although the driver
-	 * hasn't enabled a state yet where we need forcewake, BIOS may have.*/
-	intel_uncore_forcewake_get(&dev_priv->uncore, FORCEWAKE_ALL);
-
-	/*  Disable RC states. */
-	I915_WRITE(GEN6_RC_CONTROL, 0);
-
-	/* 2a: Program RC6 thresholds.*/
-	I915_WRITE(GEN6_RC6_WAKE_RATE_LIMIT, 40 << 16);
-	I915_WRITE(GEN6_RC_EVALUATION_INTERVAL, 125000); /* 12500 * 1280ns */
-	I915_WRITE(GEN6_RC_IDLE_HYSTERSIS, 25); /* 25 * 1280ns */
-
-	for_each_engine(engine, dev_priv, id)
-		I915_WRITE(RING_MAX_IDLE(engine->mmio_base), 10);
-	I915_WRITE(GEN6_RC_SLEEP, 0);
-
-	/* TO threshold set to 500 us ( 0x186 * 1.28 us) */
-	I915_WRITE(GEN6_RC6_THRESHOLD, 0x186);
-
-	/* Allows RC6 residency counter to work */
-	I915_WRITE(VLV_COUNTER_CONTROL,
-		   _MASKED_BIT_ENABLE(VLV_COUNT_RANGE_HIGH |
-				      VLV_MEDIA_RC6_COUNT_EN |
-				      VLV_RENDER_RC6_COUNT_EN));
-
-	/* For now we assume BIOS is allocating and populating the PCBR  */
-	pcbr = I915_READ(VLV_PCBR);
-
-	/* 3: Enable RC6 */
-	rc6_mode = 0;
-	if (pcbr >> VLV_PCBR_ADDR_SHIFT)
-		rc6_mode = GEN7_RC_CTL_TO_MODE;
-	I915_WRITE(GEN6_RC_CONTROL, rc6_mode);
-
-	intel_uncore_forcewake_put(&dev_priv->uncore, FORCEWAKE_ALL);
-}
-
-static void cherryview_enable_rps(struct drm_i915_private *dev_priv)
-{
-	u32 val;
-
-	intel_uncore_forcewake_get(&dev_priv->uncore, FORCEWAKE_ALL);
-
-	/* 1: Program defaults and thresholds for RPS*/
-	I915_WRITE(GEN6_RP_DOWN_TIMEOUT, 1000000);
-	I915_WRITE(GEN6_RP_UP_THRESHOLD, 59400);
-	I915_WRITE(GEN6_RP_DOWN_THRESHOLD, 245000);
-	I915_WRITE(GEN6_RP_UP_EI, 66000);
-	I915_WRITE(GEN6_RP_DOWN_EI, 350000);
-
-	I915_WRITE(GEN6_RP_IDLE_HYSTERSIS, 10);
-
-	/* 2: Enable RPS */
-	I915_WRITE(GEN6_RP_CONTROL,
-		   GEN6_RP_MEDIA_HW_NORMAL_MODE |
-		   GEN6_RP_MEDIA_IS_GFX |
-		   GEN6_RP_ENABLE |
-		   GEN6_RP_UP_BUSY_AVG |
-		   GEN6_RP_DOWN_IDLE_AVG);
-
-	/* Setting Fixed Bias */
-	vlv_punit_get(dev_priv);
-
-	val = VLV_OVERRIDE_EN | VLV_SOC_TDP_EN | CHV_BIAS_CPU_50_SOC_50;
-	vlv_punit_write(dev_priv, VLV_TURBO_SOC_OVERRIDE, val);
-
-	val = vlv_punit_read(dev_priv, PUNIT_REG_GPU_FREQ_STS);
-
-	vlv_punit_put(dev_priv);
-
-	/* RPS code assumes GPLL is used */
-	WARN_ONCE((val & GPLLENABLE) == 0, "GPLL not enabled\n");
-
-	DRM_DEBUG_DRIVER("GPLL enabled? %s\n", yesno(val & GPLLENABLE));
-	DRM_DEBUG_DRIVER("GPU status: 0x%08x\n", val);
-
-	reset_rps(dev_priv, valleyview_set_rps);
-
-	intel_uncore_forcewake_put(&dev_priv->uncore, FORCEWAKE_ALL);
-}
-
-static void valleyview_enable_rc6(struct drm_i915_private *dev_priv)
-{
-	struct intel_engine_cs *engine;
-	enum intel_engine_id id;
-	u32 gtfifodbg;
-
-	valleyview_check_pctx(dev_priv);
-
-	gtfifodbg = I915_READ(GTFIFODBG);
-	if (gtfifodbg) {
-		DRM_DEBUG_DRIVER("GT fifo had a previous error %x\n",
-				 gtfifodbg);
-		I915_WRITE(GTFIFODBG, gtfifodbg);
-	}
-
-	intel_uncore_forcewake_get(&dev_priv->uncore, FORCEWAKE_ALL);
-
-	/*  Disable RC states. */
-	I915_WRITE(GEN6_RC_CONTROL, 0);
-
-	I915_WRITE(GEN6_RC6_WAKE_RATE_LIMIT, 0x00280000);
-	I915_WRITE(GEN6_RC_EVALUATION_INTERVAL, 125000);
-	I915_WRITE(GEN6_RC_IDLE_HYSTERSIS, 25);
-
-	for_each_engine(engine, dev_priv, id)
-		I915_WRITE(RING_MAX_IDLE(engine->mmio_base), 10);
-
-	I915_WRITE(GEN6_RC6_THRESHOLD, 0x557);
-
-	/* Allows RC6 residency counter to work */
-	I915_WRITE(VLV_COUNTER_CONTROL,
-		   _MASKED_BIT_ENABLE(VLV_COUNT_RANGE_HIGH |
-				      VLV_MEDIA_RC0_COUNT_EN |
-				      VLV_RENDER_RC0_COUNT_EN |
-				      VLV_MEDIA_RC6_COUNT_EN |
-				      VLV_RENDER_RC6_COUNT_EN));
-
-	I915_WRITE(GEN6_RC_CONTROL,
-		   GEN7_RC_CTL_TO_MODE | VLV_RC_CTL_CTX_RST_PARALLEL);
-
-	intel_uncore_forcewake_put(&dev_priv->uncore, FORCEWAKE_ALL);
-}
-
-static void valleyview_enable_rps(struct drm_i915_private *dev_priv)
-{
-	u32 val;
-
-	intel_uncore_forcewake_get(&dev_priv->uncore, FORCEWAKE_ALL);
-
-	I915_WRITE(GEN6_RP_DOWN_TIMEOUT, 1000000);
-	I915_WRITE(GEN6_RP_UP_THRESHOLD, 59400);
-	I915_WRITE(GEN6_RP_DOWN_THRESHOLD, 245000);
-	I915_WRITE(GEN6_RP_UP_EI, 66000);
-	I915_WRITE(GEN6_RP_DOWN_EI, 350000);
-
-	I915_WRITE(GEN6_RP_IDLE_HYSTERSIS, 10);
-
-	I915_WRITE(GEN6_RP_CONTROL,
-		   GEN6_RP_MEDIA_TURBO |
-		   GEN6_RP_MEDIA_HW_NORMAL_MODE |
-		   GEN6_RP_MEDIA_IS_GFX |
-		   GEN6_RP_ENABLE |
-		   GEN6_RP_UP_BUSY_AVG |
-		   GEN6_RP_DOWN_IDLE_CONT);
-
-	vlv_punit_get(dev_priv);
-
-	/* Setting Fixed Bias */
-	val = VLV_OVERRIDE_EN | VLV_SOC_TDP_EN | VLV_BIAS_CPU_125_SOC_875;
-	vlv_punit_write(dev_priv, VLV_TURBO_SOC_OVERRIDE, val);
-
-	val = vlv_punit_read(dev_priv, PUNIT_REG_GPU_FREQ_STS);
-
-	vlv_punit_put(dev_priv);
-
-	/* RPS code assumes GPLL is used */
-	WARN_ONCE((val & GPLLENABLE) == 0, "GPLL not enabled\n");
-
-	DRM_DEBUG_DRIVER("GPLL enabled? %s\n", yesno(val & GPLLENABLE));
-	DRM_DEBUG_DRIVER("GPU status: 0x%08x\n", val);
-
-	reset_rps(dev_priv, valleyview_set_rps);
-
-	intel_uncore_forcewake_put(&dev_priv->uncore, FORCEWAKE_ALL);
-}
-
-static unsigned long intel_pxfreq(u32 vidfreq)
-{
-	unsigned long freq;
-	int div = (vidfreq & 0x3f0000) >> 16;
-	int post = (vidfreq & 0x3000) >> 12;
-	int pre = (vidfreq & 0x7);
-
-	if (!pre)
-		return 0;
-
-	freq = ((div * 133333) / ((1<<post) * pre));
-
-	return freq;
-}
-
-static const struct cparams {
-	u16 i;
-	u16 t;
-	u16 m;
-	u16 c;
-} cparams[] = {
-	{ 1, 1333, 301, 28664 },
-	{ 1, 1066, 294, 24460 },
-	{ 1, 800, 294, 25192 },
-	{ 0, 1333, 276, 27605 },
-	{ 0, 1066, 276, 27605 },
-	{ 0, 800, 231, 23784 },
-};
-
-static unsigned long __i915_chipset_val(struct drm_i915_private *dev_priv)
-{
-	u64 total_count, diff, ret;
-	u32 count1, count2, count3, m = 0, c = 0;
-	unsigned long now = jiffies_to_msecs(jiffies), diff1;
-	int i;
-
-	lockdep_assert_held(&mchdev_lock);
-
-	diff1 = now - dev_priv->ips.last_time1;
-
-	/* Prevent division-by-zero if we are asking too fast.
-	 * Also, we don't get interesting results if we are polling
-	 * faster than once in 10ms, so just return the saved value
-	 * in such cases.
-	 */
-	if (diff1 <= 10)
-		return dev_priv->ips.chipset_power;
-
-	count1 = I915_READ(DMIEC);
-	count2 = I915_READ(DDREC);
-	count3 = I915_READ(CSIEC);
-
-	total_count = count1 + count2 + count3;
-
-	/* FIXME: handle per-counter overflow */
-	if (total_count < dev_priv->ips.last_count1) {
-		diff = ~0UL - dev_priv->ips.last_count1;
-		diff += total_count;
-	} else {
-		diff = total_count - dev_priv->ips.last_count1;
-	}
-
-	for (i = 0; i < ARRAY_SIZE(cparams); i++) {
-		if (cparams[i].i == dev_priv->ips.c_m &&
-		    cparams[i].t == dev_priv->ips.r_t) {
-			m = cparams[i].m;
-			c = cparams[i].c;
-			break;
-		}
-	}
-
-	diff = div_u64(diff, diff1);
-	ret = ((m * diff) + c);
-	ret = div_u64(ret, 10);
-
-	dev_priv->ips.last_count1 = total_count;
-	dev_priv->ips.last_time1 = now;
-
-	dev_priv->ips.chipset_power = ret;
-
-	return ret;
-}
-
-unsigned long i915_chipset_val(struct drm_i915_private *dev_priv)
-{
-	intel_wakeref_t wakeref;
-	unsigned long val = 0;
-
-	if (!IS_GEN(dev_priv, 5))
-		return 0;
-
-	with_intel_runtime_pm(&dev_priv->runtime_pm, wakeref) {
-		spin_lock_irq(&mchdev_lock);
-		val = __i915_chipset_val(dev_priv);
-		spin_unlock_irq(&mchdev_lock);
-	}
-
-	return val;
-}
-
-unsigned long i915_mch_val(struct drm_i915_private *i915)
-{
-	unsigned long m, x, b;
-	u32 tsfs;
-
-	tsfs = intel_uncore_read(&i915->uncore, TSFS);
-
-	m = ((tsfs & TSFS_SLOPE_MASK) >> TSFS_SLOPE_SHIFT);
-	x = intel_uncore_read8(&i915->uncore, TR1);
-
-	b = tsfs & TSFS_INTR_MASK;
-
-	return ((m * x) / 127) - b;
-}
-
-static int _pxvid_to_vd(u8 pxvid)
-{
-	if (pxvid == 0)
-		return 0;
-
-	if (pxvid >= 8 && pxvid < 31)
-		pxvid = 31;
-
-	return (pxvid + 2) * 125;
-}
-
-static u32 pvid_to_extvid(struct drm_i915_private *dev_priv, u8 pxvid)
-{
-	const int vd = _pxvid_to_vd(pxvid);
-	const int vm = vd - 1125;
-
-	if (INTEL_INFO(dev_priv)->is_mobile)
-		return vm > 0 ? vm : 0;
-
-	return vd;
-}
-
-static void __i915_update_gfx_val(struct drm_i915_private *dev_priv)
-{
-	u64 now, diff, diffms;
-	u32 count;
-
-	lockdep_assert_held(&mchdev_lock);
-
-	now = ktime_get_raw_ns();
-	diffms = now - dev_priv->ips.last_time2;
-	do_div(diffms, NSEC_PER_MSEC);
-
-	/* Don't divide by 0 */
-	if (!diffms)
-		return;
-
-	count = I915_READ(GFXEC);
-
-	if (count < dev_priv->ips.last_count2) {
-		diff = ~0UL - dev_priv->ips.last_count2;
-		diff += count;
-	} else {
-		diff = count - dev_priv->ips.last_count2;
-	}
-
-	dev_priv->ips.last_count2 = count;
-	dev_priv->ips.last_time2 = now;
-
-	/* More magic constants... */
-	diff = diff * 1181;
-	diff = div_u64(diff, diffms * 10);
-	dev_priv->ips.gfx_power = diff;
-}
-
-void i915_update_gfx_val(struct drm_i915_private *dev_priv)
-{
-	intel_wakeref_t wakeref;
-
-	if (!IS_GEN(dev_priv, 5))
-		return;
-
-	with_intel_runtime_pm(&dev_priv->runtime_pm, wakeref) {
-		spin_lock_irq(&mchdev_lock);
-		__i915_update_gfx_val(dev_priv);
-		spin_unlock_irq(&mchdev_lock);
-	}
-}
-
-static unsigned long __i915_gfx_val(struct drm_i915_private *dev_priv)
-{
-	unsigned long t, corr, state1, corr2, state2;
-	u32 pxvid, ext_v;
-
-	lockdep_assert_held(&mchdev_lock);
-
-	pxvid = I915_READ(PXVFREQ(dev_priv->gt_pm.rps.cur_freq));
-	pxvid = (pxvid >> 24) & 0x7f;
-	ext_v = pvid_to_extvid(dev_priv, pxvid);
-
-	state1 = ext_v;
-
-	t = i915_mch_val(dev_priv);
-
-	/* Revel in the empirically derived constants */
-
-	/* Correction factor in 1/100000 units */
-	if (t > 80)
-		corr = ((t * 2349) + 135940);
-	else if (t >= 50)
-		corr = ((t * 964) + 29317);
-	else /* < 50 */
-		corr = ((t * 301) + 1004);
-
-	corr = corr * ((150142 * state1) / 10000 - 78642);
-	corr /= 100000;
-	corr2 = (corr * dev_priv->ips.corr);
-
-	state2 = (corr2 * state1) / 10000;
-	state2 /= 100; /* convert to mW */
-
-	__i915_update_gfx_val(dev_priv);
-
-	return dev_priv->ips.gfx_power + state2;
-}
-
-unsigned long i915_gfx_val(struct drm_i915_private *dev_priv)
-{
-	intel_wakeref_t wakeref;
-	unsigned long val = 0;
-
-	if (!IS_GEN(dev_priv, 5))
-		return 0;
-
-	with_intel_runtime_pm(&dev_priv->runtime_pm, wakeref) {
-		spin_lock_irq(&mchdev_lock);
-		val = __i915_gfx_val(dev_priv);
-		spin_unlock_irq(&mchdev_lock);
-	}
-
-	return val;
-}
-
-static struct drm_i915_private __rcu *i915_mch_dev;
-
-static struct drm_i915_private *mchdev_get(void)
-{
-	struct drm_i915_private *i915;
-
-	rcu_read_lock();
-	i915 = rcu_dereference(i915_mch_dev);
-	if (!kref_get_unless_zero(&i915->drm.ref))
-		i915 = NULL;
-	rcu_read_unlock();
-
-	return i915;
-}
-
-/**
- * i915_read_mch_val - return value for IPS use
- *
- * Calculate and return a value for the IPS driver to use when deciding whether
- * we have thermal and power headroom to increase CPU or GPU power budget.
- */
-unsigned long i915_read_mch_val(void)
-{
-	struct drm_i915_private *i915;
-	unsigned long chipset_val = 0;
-	unsigned long graphics_val = 0;
-	intel_wakeref_t wakeref;
-
-	i915 = mchdev_get();
-	if (!i915)
-		return 0;
-
-	with_intel_runtime_pm(&i915->runtime_pm, wakeref) {
-		spin_lock_irq(&mchdev_lock);
-		chipset_val = __i915_chipset_val(i915);
-		graphics_val = __i915_gfx_val(i915);
-		spin_unlock_irq(&mchdev_lock);
-	}
-
-	drm_dev_put(&i915->drm);
-	return chipset_val + graphics_val;
-}
-EXPORT_SYMBOL_GPL(i915_read_mch_val);
-
-/**
- * i915_gpu_raise - raise GPU frequency limit
- *
- * Raise the limit; IPS indicates we have thermal headroom.
- */
-bool i915_gpu_raise(void)
-{
-	struct drm_i915_private *i915;
-
-	i915 = mchdev_get();
-	if (!i915)
-		return false;
-
-	spin_lock_irq(&mchdev_lock);
-	if (i915->ips.max_delay > i915->ips.fmax)
-		i915->ips.max_delay--;
-	spin_unlock_irq(&mchdev_lock);
-
-	drm_dev_put(&i915->drm);
-	return true;
-}
-EXPORT_SYMBOL_GPL(i915_gpu_raise);
-
-/**
- * i915_gpu_lower - lower GPU frequency limit
- *
- * IPS indicates we're close to a thermal limit, so throttle back the GPU
- * frequency maximum.
- */
-bool i915_gpu_lower(void)
-{
-	struct drm_i915_private *i915;
-
-	i915 = mchdev_get();
-	if (!i915)
-		return false;
-
-	spin_lock_irq(&mchdev_lock);
-	if (i915->ips.max_delay < i915->ips.min_delay)
-		i915->ips.max_delay++;
-	spin_unlock_irq(&mchdev_lock);
-
-	drm_dev_put(&i915->drm);
-	return true;
-}
-EXPORT_SYMBOL_GPL(i915_gpu_lower);
-
-/**
- * i915_gpu_busy - indicate GPU business to IPS
- *
- * Tell the IPS driver whether or not the GPU is busy.
- */
-bool i915_gpu_busy(void)
-{
-	struct drm_i915_private *i915;
-	bool ret;
-
-	i915 = mchdev_get();
-	if (!i915)
-		return false;
-
-	ret = i915->gt.awake;
-
-	drm_dev_put(&i915->drm);
-	return ret;
-}
-EXPORT_SYMBOL_GPL(i915_gpu_busy);
-
-/**
- * i915_gpu_turbo_disable - disable graphics turbo
- *
- * Disable graphics turbo by resetting the max frequency and setting the
- * current frequency to the default.
- */
-bool i915_gpu_turbo_disable(void)
-{
-	struct drm_i915_private *i915;
-	bool ret;
-
-	i915 = mchdev_get();
-	if (!i915)
-		return false;
-
-	spin_lock_irq(&mchdev_lock);
-	i915->ips.max_delay = i915->ips.fstart;
-	ret = ironlake_set_drps(i915, i915->ips.fstart);
-	spin_unlock_irq(&mchdev_lock);
-
-	drm_dev_put(&i915->drm);
-	return ret;
-}
-EXPORT_SYMBOL_GPL(i915_gpu_turbo_disable);
-
-/**
- * Tells the intel_ips driver that the i915 driver is now loaded, if
- * IPS got loaded first.
- *
- * This awkward dance is so that neither module has to depend on the
- * other in order for IPS to do the appropriate communication of
- * GPU turbo limits to i915.
- */
-static void
-ips_ping_for_i915_load(void)
-{
-	void (*link)(void);
-
-	link = symbol_get(ips_link_to_i915_driver);
-	if (link) {
-		link();
-		symbol_put(ips_link_to_i915_driver);
-	}
-}
-
-void intel_gpu_ips_init(struct drm_i915_private *dev_priv)
-{
-	/* We only register the i915 ips part with intel-ips once everything is
-	 * set up, to avoid intel-ips sneaking in and reading bogus values. */
-	rcu_assign_pointer(i915_mch_dev, dev_priv);
-
-	ips_ping_for_i915_load();
-}
-
-void intel_gpu_ips_teardown(void)
-{
-	rcu_assign_pointer(i915_mch_dev, NULL);
-}
-
-static void intel_init_emon(struct drm_i915_private *dev_priv)
-{
-	u32 lcfuse;
-	u8 pxw[16];
-	int i;
-
-	/* Disable to program */
-	I915_WRITE(ECR, 0);
-	POSTING_READ(ECR);
-
-	/* Program energy weights for various events */
-	I915_WRITE(SDEW, 0x15040d00);
-	I915_WRITE(CSIEW0, 0x007f0000);
-	I915_WRITE(CSIEW1, 0x1e220004);
-	I915_WRITE(CSIEW2, 0x04000004);
-
-	for (i = 0; i < 5; i++)
-		I915_WRITE(PEW(i), 0);
-	for (i = 0; i < 3; i++)
-		I915_WRITE(DEW(i), 0);
-
-	/* Program P-state weights to account for frequency power adjustment */
-	for (i = 0; i < 16; i++) {
-		u32 pxvidfreq = I915_READ(PXVFREQ(i));
-		unsigned long freq = intel_pxfreq(pxvidfreq);
-		unsigned long vid = (pxvidfreq & PXVFREQ_PX_MASK) >>
-			PXVFREQ_PX_SHIFT;
-		unsigned long val;
-
-		val = vid * vid;
-		val *= (freq / 1000);
-		val *= 255;
-		val /= (127*127*900);
-		if (val > 0xff)
-			DRM_ERROR("bad pxval: %ld\n", val);
-		pxw[i] = val;
-	}
-	/* Render standby states get 0 weight */
-	pxw[14] = 0;
-	pxw[15] = 0;
-
-	for (i = 0; i < 4; i++) {
-		u32 val = (pxw[i*4] << 24) | (pxw[(i*4)+1] << 16) |
-			(pxw[(i*4)+2] << 8) | (pxw[(i*4)+3]);
-		I915_WRITE(PXW(i), val);
-	}
-
-	/* Adjust magic regs to magic values (more experimental results) */
-	I915_WRITE(OGW0, 0);
-	I915_WRITE(OGW1, 0);
-	I915_WRITE(EG0, 0x00007f00);
-	I915_WRITE(EG1, 0x0000000e);
-	I915_WRITE(EG2, 0x000e0000);
-	I915_WRITE(EG3, 0x68000300);
-	I915_WRITE(EG4, 0x42000000);
-	I915_WRITE(EG5, 0x00140031);
-	I915_WRITE(EG6, 0);
-	I915_WRITE(EG7, 0);
-
-	for (i = 0; i < 8; i++)
-		I915_WRITE(PXWL(i), 0);
-
-	/* Enable PMON + select events */
-	I915_WRITE(ECR, 0x80000019);
-
-	lcfuse = I915_READ(LCFUSE02);
-
-	dev_priv->ips.corr = (lcfuse & LCFUSE_HIV_MASK);
-}
-
-void intel_init_gt_powersave(struct drm_i915_private *dev_priv)
-{
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
-
-	/*
-	 * RPM depends on RC6 to save restore the GT HW context, so make RC6 a
-	 * requirement.
-	 */
-	if (!sanitize_rc6(dev_priv)) {
-		DRM_INFO("RC6 disabled, disabling runtime PM support\n");
-		pm_runtime_get(&dev_priv->drm.pdev->dev);
-	}
-
-	/* Initialize RPS limits (for userspace) */
-	if (IS_CHERRYVIEW(dev_priv))
-		cherryview_init_gt_powersave(dev_priv);
-	else if (IS_VALLEYVIEW(dev_priv))
-		valleyview_init_gt_powersave(dev_priv);
-	else if (INTEL_GEN(dev_priv) >= 6)
-		gen6_init_rps_frequencies(dev_priv);
-
-	/* Derive initial user preferences/limits from the hardware limits */
-	rps->max_freq_softlimit = rps->max_freq;
-	rps->min_freq_softlimit = rps->min_freq;
-
-	/* After setting max-softlimit, find the overclock max freq */
-	if (IS_GEN(dev_priv, 6) ||
-	    IS_IVYBRIDGE(dev_priv) || IS_HASWELL(dev_priv)) {
-		u32 params = 0;
-
-		sandybridge_pcode_read(dev_priv, GEN6_READ_OC_PARAMS,
-				       &params, NULL);
-		if (params & BIT(31)) { /* OC supported */
-			DRM_DEBUG_DRIVER("Overclocking supported, max: %dMHz, overclock: %dMHz\n",
-					 (rps->max_freq & 0xff) * 50,
-					 (params & 0xff) * 50);
-			rps->max_freq = params & 0xff;
-		}
-	}
-
-	/* Finally allow us to boost to max by default */
-	rps->boost_freq = rps->max_freq;
-	rps->idle_freq = rps->min_freq;
-	rps->cur_freq = rps->idle_freq;
-}
-
-void intel_cleanup_gt_powersave(struct drm_i915_private *dev_priv)
-{
-	if (IS_VALLEYVIEW(dev_priv))
-		valleyview_cleanup_gt_powersave(dev_priv);
-
-	if (!HAS_RC6(dev_priv))
-		pm_runtime_put(&dev_priv->drm.pdev->dev);
-}
-
-void intel_sanitize_gt_powersave(struct drm_i915_private *dev_priv)
-{
-	dev_priv->gt_pm.rps.enabled = true; /* force RPS disabling */
-	dev_priv->gt_pm.rc6.enabled = true; /* force RC6 disabling */
-	intel_disable_gt_powersave(dev_priv);
-
-	if (INTEL_GEN(dev_priv) >= 11)
-		gen11_reset_rps_interrupts(dev_priv);
-	else if (INTEL_GEN(dev_priv) >= 6)
-		gen6_reset_rps_interrupts(dev_priv);
-}
-
-static inline void intel_disable_llc_pstate(struct drm_i915_private *i915)
-{
-	lockdep_assert_held(&i915->gt_pm.rps.lock);
-
-	if (!i915->gt_pm.llc_pstate.enabled)
-		return;
-
-	/* Currently there is no HW configuration to be done to disable. */
-
-	i915->gt_pm.llc_pstate.enabled = false;
-}
-
-static void intel_disable_rc6(struct drm_i915_private *dev_priv)
-{
-	lockdep_assert_held(&dev_priv->gt_pm.rps.lock);
-
-	if (!dev_priv->gt_pm.rc6.enabled)
-		return;
-
-	if (INTEL_GEN(dev_priv) >= 9)
-		gen9_disable_rc6(dev_priv);
-	else if (IS_CHERRYVIEW(dev_priv))
-		cherryview_disable_rc6(dev_priv);
-	else if (IS_VALLEYVIEW(dev_priv))
-		valleyview_disable_rc6(dev_priv);
-	else if (INTEL_GEN(dev_priv) >= 6)
-		gen6_disable_rc6(dev_priv);
-
-	dev_priv->gt_pm.rc6.enabled = false;
-}
-
-static void intel_disable_rps(struct drm_i915_private *dev_priv)
-{
-	lockdep_assert_held(&dev_priv->gt_pm.rps.lock);
-
-	if (!dev_priv->gt_pm.rps.enabled)
-		return;
-
-	if (INTEL_GEN(dev_priv) >= 9)
-		gen9_disable_rps(dev_priv);
-	else if (IS_CHERRYVIEW(dev_priv))
-		cherryview_disable_rps(dev_priv);
-	else if (IS_VALLEYVIEW(dev_priv))
-		valleyview_disable_rps(dev_priv);
-	else if (INTEL_GEN(dev_priv) >= 6)
-		gen6_disable_rps(dev_priv);
-	else if (IS_IRONLAKE_M(dev_priv))
-		ironlake_disable_drps(dev_priv);
-
-	dev_priv->gt_pm.rps.enabled = false;
-}
-
-void intel_disable_gt_powersave(struct drm_i915_private *dev_priv)
-{
-	mutex_lock(&dev_priv->gt_pm.rps.lock);
-
-	intel_disable_rc6(dev_priv);
-	intel_disable_rps(dev_priv);
-	if (HAS_LLC(dev_priv))
-		intel_disable_llc_pstate(dev_priv);
-
-	mutex_unlock(&dev_priv->gt_pm.rps.lock);
-}
-
-static inline void intel_enable_llc_pstate(struct drm_i915_private *i915)
-{
-	lockdep_assert_held(&i915->gt_pm.rps.lock);
-
-	if (i915->gt_pm.llc_pstate.enabled)
-		return;
-
-	gen6_update_ring_freq(i915);
-
-	i915->gt_pm.llc_pstate.enabled = true;
-}
-
-static void intel_enable_rc6(struct drm_i915_private *dev_priv)
-{
-	lockdep_assert_held(&dev_priv->gt_pm.rps.lock);
-
-	if (dev_priv->gt_pm.rc6.enabled)
-		return;
-
-	if (IS_CHERRYVIEW(dev_priv))
-		cherryview_enable_rc6(dev_priv);
-	else if (IS_VALLEYVIEW(dev_priv))
-		valleyview_enable_rc6(dev_priv);
-	else if (INTEL_GEN(dev_priv) >= 11)
-		gen11_enable_rc6(dev_priv);
-	else if (INTEL_GEN(dev_priv) >= 9)
-		gen9_enable_rc6(dev_priv);
-	else if (IS_BROADWELL(dev_priv))
-		gen8_enable_rc6(dev_priv);
-	else if (INTEL_GEN(dev_priv) >= 6)
-		gen6_enable_rc6(dev_priv);
-
-	dev_priv->gt_pm.rc6.enabled = true;
-}
-
-static void intel_enable_rps(struct drm_i915_private *dev_priv)
-{
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
-
-	lockdep_assert_held(&rps->lock);
-
-	if (rps->enabled)
-		return;
-
-	if (IS_CHERRYVIEW(dev_priv)) {
-		cherryview_enable_rps(dev_priv);
-	} else if (IS_VALLEYVIEW(dev_priv)) {
-		valleyview_enable_rps(dev_priv);
-	} else if (INTEL_GEN(dev_priv) >= 9) {
-		gen9_enable_rps(dev_priv);
-	} else if (IS_BROADWELL(dev_priv)) {
-		gen8_enable_rps(dev_priv);
-	} else if (INTEL_GEN(dev_priv) >= 6) {
-		gen6_enable_rps(dev_priv);
-	} else if (IS_IRONLAKE_M(dev_priv)) {
-		ironlake_enable_drps(dev_priv);
-		intel_init_emon(dev_priv);
-	}
-
-	WARN_ON(rps->max_freq < rps->min_freq);
-	WARN_ON(rps->idle_freq > rps->max_freq);
-
-	WARN_ON(rps->efficient_freq < rps->min_freq);
-	WARN_ON(rps->efficient_freq > rps->max_freq);
-
-	rps->enabled = true;
-}
-
-void intel_enable_gt_powersave(struct drm_i915_private *dev_priv)
-{
-	/* Powersaving is controlled by the host when inside a VM */
-	if (intel_vgpu_active(dev_priv))
-		return;
-
-	mutex_lock(&dev_priv->gt_pm.rps.lock);
-
-	if (HAS_RC6(dev_priv))
-		intel_enable_rc6(dev_priv);
-	if (HAS_RPS(dev_priv))
-		intel_enable_rps(dev_priv);
-	if (HAS_LLC(dev_priv))
-		intel_enable_llc_pstate(dev_priv);
-
-	mutex_unlock(&dev_priv->gt_pm.rps.lock);
-}
-
 static void ibx_init_clock_gating(struct drm_i915_private *dev_priv)
 {
 	/*
@@ -8858,7 +6347,7 @@ static void ilk_init_clock_gating(struct drm_i915_private *dev_priv)
 
 static void cpt_init_clock_gating(struct drm_i915_private *dev_priv)
 {
-	int pipe;
+	enum pipe pipe;
 	u32 val;
 
 	/*
@@ -9078,6 +6567,22 @@ static void icl_init_clock_gating(struct drm_i915_private *dev_priv)
 		   _MASKED_BIT_ENABLE(GEN11_ENABLE_32_PLANE_MODE));
 }
 
+static void tgl_init_clock_gating(struct drm_i915_private *dev_priv)
+{
+	u32 vd_pg_enable = 0;
+	unsigned int i;
+
+	/* This is not a WA. Enable VD HCP & MFX_ENC powergate */
+	for (i = 0; i < I915_MAX_VCS; i++) {
+		if (HAS_ENGINE(dev_priv, _VCS(i)))
+			vd_pg_enable |= VDN_HCP_POWERGATE_ENABLE(i) |
+					VDN_MFX_POWERGATE_ENABLE(i);
+	}
+
+	I915_WRITE(POWERGATE_ENABLE,
+		   I915_READ(POWERGATE_ENABLE) | vd_pg_enable);
+}
+
 static void cnp_init_clock_gating(struct drm_i915_private *dev_priv)
 {
 	if (!HAS_PCH_CNP(dev_priv))
@@ -9598,7 +7103,7 @@ static void nop_init_clock_gating(struct drm_i915_private *dev_priv)
 void intel_init_clock_gating_hooks(struct drm_i915_private *dev_priv)
 {
 	if (IS_GEN(dev_priv, 12))
-		dev_priv->display.init_clock_gating = nop_init_clock_gating;
+		dev_priv->display.init_clock_gating = tgl_init_clock_gating;
 	else if (IS_GEN(dev_priv, 11))
 		dev_priv->display.init_clock_gating = icl_init_clock_gating;
 	else if (IS_CANNONLAKE(dev_priv))
@@ -9654,6 +7159,9 @@ void intel_init_pm(struct drm_i915_private *dev_priv)
 	else if (IS_GEN(dev_priv, 5))
 		i915_ironlake_get_mem_freq(dev_priv);
 
+	if (intel_has_sagv(dev_priv))
+		skl_setup_sagv_block_time(dev_priv);
+
 	/* For FIFO watermark updates */
 	if (INTEL_GEN(dev_priv) >= 9) {
 		skl_setup_wm_latency(dev_priv);
@@ -9712,7 +7220,7 @@ void intel_init_pm(struct drm_i915_private *dev_priv)
 		dev_priv->display.update_wm = i9xx_update_wm;
 		dev_priv->display.get_fifo_size = i9xx_get_fifo_size;
 	} else if (IS_GEN(dev_priv, 2)) {
-		if (INTEL_INFO(dev_priv)->num_pipes == 1) {
+		if (INTEL_NUM_PIPES(dev_priv) == 1) {
 			dev_priv->display.update_wm = i845_update_wm;
 			dev_priv->display.get_fifo_size = i845_get_fifo_size;
 		} else {
@@ -9724,217 +7232,8 @@ void intel_init_pm(struct drm_i915_private *dev_priv)
 	}
 }
 
-static int byt_gpu_freq(struct drm_i915_private *dev_priv, int val)
-{
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
-
-	/*
-	 * N = val - 0xb7
-	 * Slow = Fast = GPLL ref * N
-	 */
-	return DIV_ROUND_CLOSEST(rps->gpll_ref_freq * (val - 0xb7), 1000);
-}
-
-static int byt_freq_opcode(struct drm_i915_private *dev_priv, int val)
-{
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
-
-	return DIV_ROUND_CLOSEST(1000 * val, rps->gpll_ref_freq) + 0xb7;
-}
-
-static int chv_gpu_freq(struct drm_i915_private *dev_priv, int val)
-{
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
-
-	/*
-	 * N = val / 2
-	 * CU (slow) = CU2x (fast) / 2 = GPLL ref * N / 2
-	 */
-	return DIV_ROUND_CLOSEST(rps->gpll_ref_freq * val, 2 * 2 * 1000);
-}
-
-static int chv_freq_opcode(struct drm_i915_private *dev_priv, int val)
-{
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
-
-	/* CHV needs even values */
-	return DIV_ROUND_CLOSEST(2 * 1000 * val, rps->gpll_ref_freq) * 2;
-}
-
-int intel_gpu_freq(struct drm_i915_private *dev_priv, int val)
-{
-	if (INTEL_GEN(dev_priv) >= 9)
-		return DIV_ROUND_CLOSEST(val * GT_FREQUENCY_MULTIPLIER,
-					 GEN9_FREQ_SCALER);
-	else if (IS_CHERRYVIEW(dev_priv))
-		return chv_gpu_freq(dev_priv, val);
-	else if (IS_VALLEYVIEW(dev_priv))
-		return byt_gpu_freq(dev_priv, val);
-	else
-		return val * GT_FREQUENCY_MULTIPLIER;
-}
-
-int intel_freq_opcode(struct drm_i915_private *dev_priv, int val)
-{
-	if (INTEL_GEN(dev_priv) >= 9)
-		return DIV_ROUND_CLOSEST(val * GEN9_FREQ_SCALER,
-					 GT_FREQUENCY_MULTIPLIER);
-	else if (IS_CHERRYVIEW(dev_priv))
-		return chv_freq_opcode(dev_priv, val);
-	else if (IS_VALLEYVIEW(dev_priv))
-		return byt_freq_opcode(dev_priv, val);
-	else
-		return DIV_ROUND_CLOSEST(val, GT_FREQUENCY_MULTIPLIER);
-}
-
 void intel_pm_setup(struct drm_i915_private *dev_priv)
 {
-	mutex_init(&dev_priv->gt_pm.rps.lock);
-	mutex_init(&dev_priv->gt_pm.rps.power.mutex);
-
-	atomic_set(&dev_priv->gt_pm.rps.num_waiters, 0);
-
 	dev_priv->runtime_pm.suspended = false;
 	atomic_set(&dev_priv->runtime_pm.wakeref_count, 0);
 }
-
-static u64 vlv_residency_raw(struct drm_i915_private *dev_priv,
-			     const i915_reg_t reg)
-{
-	u32 lower, upper, tmp;
-	int loop = 2;
-
-	/*
-	 * The register accessed do not need forcewake. We borrow
-	 * uncore lock to prevent concurrent access to range reg.
-	 */
-	lockdep_assert_held(&dev_priv->uncore.lock);
-
-	/*
-	 * vlv and chv residency counters are 40 bits in width.
-	 * With a control bit, we can choose between upper or lower
-	 * 32bit window into this counter.
-	 *
-	 * Although we always use the counter in high-range mode elsewhere,
-	 * userspace may attempt to read the value before rc6 is initialised,
-	 * before we have set the default VLV_COUNTER_CONTROL value. So always
-	 * set the high bit to be safe.
-	 */
-	I915_WRITE_FW(VLV_COUNTER_CONTROL,
-		      _MASKED_BIT_ENABLE(VLV_COUNT_RANGE_HIGH));
-	upper = I915_READ_FW(reg);
-	do {
-		tmp = upper;
-
-		I915_WRITE_FW(VLV_COUNTER_CONTROL,
-			      _MASKED_BIT_DISABLE(VLV_COUNT_RANGE_HIGH));
-		lower = I915_READ_FW(reg);
-
-		I915_WRITE_FW(VLV_COUNTER_CONTROL,
-			      _MASKED_BIT_ENABLE(VLV_COUNT_RANGE_HIGH));
-		upper = I915_READ_FW(reg);
-	} while (upper != tmp && --loop);
-
-	/*
-	 * Everywhere else we always use VLV_COUNTER_CONTROL with the
-	 * VLV_COUNT_RANGE_HIGH bit set - so it is safe to leave it set
-	 * now.
-	 */
-
-	return lower | (u64)upper << 8;
-}
-
-u64 intel_rc6_residency_ns(struct drm_i915_private *dev_priv,
-			   const i915_reg_t reg)
-{
-	struct intel_uncore *uncore = &dev_priv->uncore;
-	u64 time_hw, prev_hw, overflow_hw;
-	unsigned int fw_domains;
-	unsigned long flags;
-	unsigned int i;
-	u32 mul, div;
-
-	if (!HAS_RC6(dev_priv))
-		return 0;
-
-	/*
-	 * Store previous hw counter values for counter wrap-around handling.
-	 *
-	 * There are only four interesting registers and they live next to each
-	 * other so we can use the relative address, compared to the smallest
-	 * one as the index into driver storage.
-	 */
-	i = (i915_mmio_reg_offset(reg) -
-	     i915_mmio_reg_offset(GEN6_GT_GFX_RC6_LOCKED)) / sizeof(u32);
-	if (WARN_ON_ONCE(i >= ARRAY_SIZE(dev_priv->gt_pm.rc6.cur_residency)))
-		return 0;
-
-	fw_domains = intel_uncore_forcewake_for_reg(uncore, reg, FW_REG_READ);
-
-	spin_lock_irqsave(&uncore->lock, flags);
-	intel_uncore_forcewake_get__locked(uncore, fw_domains);
-
-	/* On VLV and CHV, residency time is in CZ units rather than 1.28us */
-	if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)) {
-		mul = 1000000;
-		div = dev_priv->czclk_freq;
-		overflow_hw = BIT_ULL(40);
-		time_hw = vlv_residency_raw(dev_priv, reg);
-	} else {
-		/* 833.33ns units on Gen9LP, 1.28us elsewhere. */
-		if (IS_GEN9_LP(dev_priv)) {
-			mul = 10000;
-			div = 12;
-		} else {
-			mul = 1280;
-			div = 1;
-		}
-
-		overflow_hw = BIT_ULL(32);
-		time_hw = intel_uncore_read_fw(uncore, reg);
-	}
-
-	/*
-	 * Counter wrap handling.
-	 *
-	 * But relying on a sufficient frequency of queries otherwise counters
-	 * can still wrap.
-	 */
-	prev_hw = dev_priv->gt_pm.rc6.prev_hw_residency[i];
-	dev_priv->gt_pm.rc6.prev_hw_residency[i] = time_hw;
-
-	/* RC6 delta from last sample. */
-	if (time_hw >= prev_hw)
-		time_hw -= prev_hw;
-	else
-		time_hw += overflow_hw - prev_hw;
-
-	/* Add delta to RC6 extended raw driver copy. */
-	time_hw += dev_priv->gt_pm.rc6.cur_residency[i];
-	dev_priv->gt_pm.rc6.cur_residency[i] = time_hw;
-
-	intel_uncore_forcewake_put__locked(uncore, fw_domains);
-	spin_unlock_irqrestore(&uncore->lock, flags);
-
-	return mul_u64_u32_div(time_hw, mul, div);
-}
-
-u64 intel_rc6_residency_us(struct drm_i915_private *dev_priv,
-			   i915_reg_t reg)
-{
-	return DIV_ROUND_UP_ULL(intel_rc6_residency_ns(dev_priv, reg), 1000);
-}
-
-u32 intel_get_cagf(struct drm_i915_private *dev_priv, u32 rpstat)
-{
-	u32 cagf;
-
-	if (INTEL_GEN(dev_priv) >= 9)
-		cagf = (rpstat & GEN9_CAGF_MASK) >> GEN9_CAGF_SHIFT;
-	else if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv))
-		cagf = (rpstat & HSW_CAGF_MASK) >> HSW_CAGF_SHIFT;
-	else
-		cagf = (rpstat & GEN6_CAGF_MASK) >> GEN6_CAGF_SHIFT;
-
-	return  cagf;
-}
diff --git a/drivers/gpu/drm/i915/intel_pm.h b/drivers/gpu/drm/i915/intel_pm.h
index e3573e1e16e3..b579c724b915 100644
--- a/drivers/gpu/drm/i915/intel_pm.h
+++ b/drivers/gpu/drm/i915/intel_pm.h
@@ -29,16 +29,6 @@ void intel_update_watermarks(struct intel_crtc *crtc);
 void intel_init_pm(struct drm_i915_private *dev_priv);
 void intel_init_clock_gating_hooks(struct drm_i915_private *dev_priv);
 void intel_pm_setup(struct drm_i915_private *dev_priv);
-void intel_gpu_ips_init(struct drm_i915_private *dev_priv);
-void intel_gpu_ips_teardown(void);
-void intel_init_gt_powersave(struct drm_i915_private *dev_priv);
-void intel_cleanup_gt_powersave(struct drm_i915_private *dev_priv);
-void intel_sanitize_gt_powersave(struct drm_i915_private *dev_priv);
-void intel_enable_gt_powersave(struct drm_i915_private *dev_priv);
-void intel_disable_gt_powersave(struct drm_i915_private *dev_priv);
-void gen6_rps_busy(struct drm_i915_private *dev_priv);
-void gen6_rps_idle(struct drm_i915_private *dev_priv);
-void gen6_rps_boost(struct i915_request *rq);
 void g4x_wm_get_hw_state(struct drm_i915_private *dev_priv);
 void vlv_wm_get_hw_state(struct drm_i915_private *dev_priv);
 void ilk_wm_get_hw_state(struct drm_i915_private *dev_priv);
@@ -65,26 +55,9 @@ void skl_write_plane_wm(struct intel_plane *plane,
 void skl_write_cursor_wm(struct intel_plane *plane,
 			 const struct intel_crtc_state *crtc_state);
 bool ilk_disable_lp_wm(struct drm_device *dev);
-int skl_check_pipe_max_pixel_rate(struct intel_crtc *intel_crtc,
-				  struct intel_crtc_state *cstate);
 void intel_init_ipc(struct drm_i915_private *dev_priv);
 void intel_enable_ipc(struct drm_i915_private *dev_priv);
 
-int intel_gpu_freq(struct drm_i915_private *dev_priv, int val);
-int intel_freq_opcode(struct drm_i915_private *dev_priv, int val);
-u64 intel_rc6_residency_ns(struct drm_i915_private *dev_priv, i915_reg_t reg);
-u64 intel_rc6_residency_us(struct drm_i915_private *dev_priv, i915_reg_t reg);
-
-u32 intel_get_cagf(struct drm_i915_private *dev_priv, u32 rpstat1);
-
-unsigned long i915_chipset_val(struct drm_i915_private *dev_priv);
-unsigned long i915_mch_val(struct drm_i915_private *dev_priv);
-unsigned long i915_gfx_val(struct drm_i915_private *dev_priv);
-void i915_update_gfx_val(struct drm_i915_private *dev_priv);
-
-bool ironlake_set_drps(struct drm_i915_private *dev_priv, u8 val);
-int intel_set_rps(struct drm_i915_private *dev_priv, u8 val);
-void intel_rps_mark_interactive(struct drm_i915_private *i915, bool interactive);
 bool intel_set_memory_cxsr(struct drm_i915_private *dev_priv, bool enable);
 
 #endif /* __INTEL_PM_H__ */
diff --git a/drivers/gpu/drm/i915/intel_region_lmem.c b/drivers/gpu/drm/i915/intel_region_lmem.c
new file mode 100644
index 000000000000..583118095635
--- /dev/null
+++ b/drivers/gpu/drm/i915/intel_region_lmem.c
@@ -0,0 +1,132 @@
+// SPDX-License-Identifier: MIT
+/*
+ * Copyright © 2019 Intel Corporation
+ */
+
+#include "i915_drv.h"
+#include "intel_memory_region.h"
+#include "gem/i915_gem_lmem.h"
+#include "gem/i915_gem_region.h"
+#include "intel_region_lmem.h"
+
+static int init_fake_lmem_bar(struct intel_memory_region *mem)
+{
+	struct drm_i915_private *i915 = mem->i915;
+	struct i915_ggtt *ggtt = &i915->ggtt;
+	unsigned long n;
+	int ret;
+
+	/* We want to 1:1 map the mappable aperture to our reserved region */
+
+	mem->fake_mappable.start = 0;
+	mem->fake_mappable.size = resource_size(&mem->region);
+	mem->fake_mappable.color = I915_COLOR_UNEVICTABLE;
+
+	ret = drm_mm_reserve_node(&ggtt->vm.mm, &mem->fake_mappable);
+	if (ret)
+		return ret;
+
+	mem->remap_addr = dma_map_resource(&i915->drm.pdev->dev,
+					   mem->region.start,
+					   mem->fake_mappable.size,
+					   PCI_DMA_BIDIRECTIONAL,
+					   DMA_ATTR_FORCE_CONTIGUOUS);
+	if (dma_mapping_error(&i915->drm.pdev->dev, mem->remap_addr)) {
+		drm_mm_remove_node(&mem->fake_mappable);
+		return -EINVAL;
+	}
+
+	for (n = 0; n < mem->fake_mappable.size >> PAGE_SHIFT; ++n) {
+		ggtt->vm.insert_page(&ggtt->vm,
+				     mem->remap_addr + (n << PAGE_SHIFT),
+				     n << PAGE_SHIFT,
+				     I915_CACHE_NONE, 0);
+	}
+
+	mem->region = (struct resource)DEFINE_RES_MEM(mem->remap_addr,
+						      mem->fake_mappable.size);
+
+	return 0;
+}
+
+static void release_fake_lmem_bar(struct intel_memory_region *mem)
+{
+	if (drm_mm_node_allocated(&mem->fake_mappable))
+		drm_mm_remove_node(&mem->fake_mappable);
+
+	dma_unmap_resource(&mem->i915->drm.pdev->dev,
+			   mem->remap_addr,
+			   mem->fake_mappable.size,
+			   PCI_DMA_BIDIRECTIONAL,
+			   DMA_ATTR_FORCE_CONTIGUOUS);
+}
+
+static void
+region_lmem_release(struct intel_memory_region *mem)
+{
+	release_fake_lmem_bar(mem);
+	io_mapping_fini(&mem->iomap);
+	intel_memory_region_release_buddy(mem);
+}
+
+static int
+region_lmem_init(struct intel_memory_region *mem)
+{
+	int ret;
+
+	if (i915_modparams.fake_lmem_start) {
+		ret = init_fake_lmem_bar(mem);
+		GEM_BUG_ON(ret);
+	}
+
+	if (!io_mapping_init_wc(&mem->iomap,
+				mem->io_start,
+				resource_size(&mem->region)))
+		return -EIO;
+
+	ret = intel_memory_region_init_buddy(mem);
+	if (ret)
+		io_mapping_fini(&mem->iomap);
+
+	return ret;
+}
+
+const struct intel_memory_region_ops intel_region_lmem_ops = {
+	.init = region_lmem_init,
+	.release = region_lmem_release,
+	.create_object = __i915_gem_lmem_object_create,
+};
+
+struct intel_memory_region *
+intel_setup_fake_lmem(struct drm_i915_private *i915)
+{
+	struct pci_dev *pdev = i915->drm.pdev;
+	struct intel_memory_region *mem;
+	resource_size_t mappable_end;
+	resource_size_t io_start;
+	resource_size_t start;
+
+	GEM_BUG_ON(i915_ggtt_has_aperture(&i915->ggtt));
+	GEM_BUG_ON(!i915_modparams.fake_lmem_start);
+
+	/* Your mappable aperture belongs to me now! */
+	mappable_end = pci_resource_len(pdev, 2);
+	io_start = pci_resource_start(pdev, 2),
+	start = i915_modparams.fake_lmem_start;
+
+	mem = intel_memory_region_create(i915,
+					 start,
+					 mappable_end,
+					 PAGE_SIZE,
+					 io_start,
+					 &intel_region_lmem_ops);
+	if (!IS_ERR(mem)) {
+		DRM_INFO("Intel graphics fake LMEM: %pR\n", &mem->region);
+		DRM_INFO("Intel graphics fake LMEM IO start: %llx\n",
+			 (u64)mem->io_start);
+		DRM_INFO("Intel graphics fake LMEM size: %llx\n",
+			 (u64)resource_size(&mem->region));
+	}
+
+	return mem;
+}
diff --git a/drivers/gpu/drm/i915/intel_region_lmem.h b/drivers/gpu/drm/i915/intel_region_lmem.h
new file mode 100644
index 000000000000..213def7c7b8a
--- /dev/null
+++ b/drivers/gpu/drm/i915/intel_region_lmem.h
@@ -0,0 +1,16 @@
+/* SPDX-License-Identifier: MIT */
+/*
+ * Copyright © 2019 Intel Corporation
+ */
+
+#ifndef __INTEL_REGION_LMEM_H
+#define __INTEL_REGION_LMEM_H
+
+struct drm_i915_private;
+
+extern const struct intel_memory_region_ops intel_region_lmem_ops;
+
+struct intel_memory_region *
+intel_setup_fake_lmem(struct drm_i915_private *i915);
+
+#endif /* !__INTEL_REGION_LMEM_H */
diff --git a/drivers/gpu/drm/i915/intel_runtime_pm.c b/drivers/gpu/drm/i915/intel_runtime_pm.c
index 2fd3c097e1f5..ad719c9602af 100644
--- a/drivers/gpu/drm/i915/intel_runtime_pm.c
+++ b/drivers/gpu/drm/i915/intel_runtime_pm.c
@@ -27,7 +27,6 @@
  */
 
 #include <linux/pm_runtime.h>
-#include <linux/vgaarb.h>
 
 #include <drm/drm_print.h>
 
diff --git a/drivers/gpu/drm/i915/intel_uncore.c b/drivers/gpu/drm/i915/intel_uncore.c
index 9e583f13a9e4..94a97bf8c021 100644
--- a/drivers/gpu/drm/i915/intel_uncore.c
+++ b/drivers/gpu/drm/i915/intel_uncore.c
@@ -805,9 +805,6 @@ void assert_forcewakes_active(struct intel_uncore *uncore,
 /* We give fast paths for the really cool registers */
 #define NEEDS_FORCE_WAKE(reg) ((reg) < 0x40000)
 
-#define GEN11_NEEDS_FORCE_WAKE(reg) \
-	((reg) < 0x40000 || ((reg) >= 0x1c0000 && (reg) < 0x1dc000))
-
 #define __gen6_reg_read_fw_domains(uncore, offset) \
 ({ \
 	enum forcewake_domains __fwd; \
@@ -903,12 +900,10 @@ static const struct intel_forcewake_range __vlv_fw_ranges[] = {
 })
 
 #define __gen11_fwtable_reg_read_fw_domains(uncore, offset) \
-({ \
-	enum forcewake_domains __fwd = 0; \
-	if (GEN11_NEEDS_FORCE_WAKE((offset))) \
-		__fwd = find_fw_domain(uncore, offset); \
-	__fwd; \
-})
+	find_fw_domain(uncore, offset)
+
+#define __gen12_fwtable_reg_read_fw_domains(uncore, offset) \
+	find_fw_domain(uncore, offset)
 
 /* *Must* be sorted by offset! See intel_shadow_table_check(). */
 static const i915_reg_t gen8_shadowed_regs[] = {
@@ -935,6 +930,20 @@ static const i915_reg_t gen11_shadowed_regs[] = {
 	/* TODO: Other registers are not yet used */
 };
 
+static const i915_reg_t gen12_shadowed_regs[] = {
+	RING_TAIL(RENDER_RING_BASE),		/* 0x2000 (base) */
+	GEN6_RPNSWREQ,				/* 0xA008 */
+	GEN6_RC_VIDEO_FREQ,			/* 0xA00C */
+	RING_TAIL(BLT_RING_BASE),		/* 0x22000 (base) */
+	RING_TAIL(GEN11_BSD_RING_BASE),		/* 0x1C0000 (base) */
+	RING_TAIL(GEN11_BSD2_RING_BASE),	/* 0x1C4000 (base) */
+	RING_TAIL(GEN11_VEBOX_RING_BASE),	/* 0x1C8000 (base) */
+	RING_TAIL(GEN11_BSD3_RING_BASE),	/* 0x1D0000 (base) */
+	RING_TAIL(GEN11_BSD4_RING_BASE),	/* 0x1D4000 (base) */
+	RING_TAIL(GEN11_VEBOX2_RING_BASE),	/* 0x1D8000 (base) */
+	/* TODO: Other registers are not yet used */
+};
+
 static int mmio_reg_cmp(u32 key, const i915_reg_t *reg)
 {
 	u32 offset = i915_mmio_reg_offset(*reg);
@@ -957,6 +966,7 @@ static bool is_gen##x##_shadowed(u32 offset) \
 
 __is_genX_shadowed(8)
 __is_genX_shadowed(11)
+__is_genX_shadowed(12)
 
 static enum forcewake_domains
 gen6_reg_write_fw_domains(struct intel_uncore *uncore, i915_reg_t reg)
@@ -1005,8 +1015,18 @@ static const struct intel_forcewake_range __chv_fw_ranges[] = {
 #define __gen11_fwtable_reg_write_fw_domains(uncore, offset) \
 ({ \
 	enum forcewake_domains __fwd = 0; \
-	if (GEN11_NEEDS_FORCE_WAKE((offset)) && !is_gen11_shadowed(offset)) \
-		__fwd = find_fw_domain(uncore, offset); \
+	const u32 __offset = (offset); \
+	if (!is_gen11_shadowed(__offset)) \
+		__fwd = find_fw_domain(uncore, __offset); \
+	__fwd; \
+})
+
+#define __gen12_fwtable_reg_write_fw_domains(uncore, offset) \
+({ \
+	enum forcewake_domains __fwd = 0; \
+	const u32 __offset = (offset); \
+	if (!is_gen12_shadowed(__offset)) \
+		__fwd = find_fw_domain(uncore, __offset); \
 	__fwd; \
 })
 
@@ -1065,9 +1085,51 @@ static const struct intel_forcewake_range __gen11_fw_ranges[] = {
 	GEN_FW_RANGE(0x9400, 0x97ff, FORCEWAKE_ALL),
 	GEN_FW_RANGE(0x9800, 0xafff, FORCEWAKE_BLITTER),
 	GEN_FW_RANGE(0xb000, 0xb47f, FORCEWAKE_RENDER),
+	GEN_FW_RANGE(0xb480, 0xdeff, FORCEWAKE_BLITTER),
+	GEN_FW_RANGE(0xdf00, 0xe8ff, FORCEWAKE_RENDER),
+	GEN_FW_RANGE(0xe900, 0x16dff, FORCEWAKE_BLITTER),
+	GEN_FW_RANGE(0x16e00, 0x19fff, FORCEWAKE_RENDER),
+	GEN_FW_RANGE(0x1a000, 0x243ff, FORCEWAKE_BLITTER),
+	GEN_FW_RANGE(0x24400, 0x247ff, FORCEWAKE_RENDER),
+	GEN_FW_RANGE(0x24800, 0x3ffff, FORCEWAKE_BLITTER),
+	GEN_FW_RANGE(0x40000, 0x1bffff, 0),
+	GEN_FW_RANGE(0x1c0000, 0x1c3fff, FORCEWAKE_MEDIA_VDBOX0),
+	GEN_FW_RANGE(0x1c4000, 0x1c7fff, FORCEWAKE_MEDIA_VDBOX1),
+	GEN_FW_RANGE(0x1c8000, 0x1cbfff, FORCEWAKE_MEDIA_VEBOX0),
+	GEN_FW_RANGE(0x1cc000, 0x1cffff, FORCEWAKE_BLITTER),
+	GEN_FW_RANGE(0x1d0000, 0x1d3fff, FORCEWAKE_MEDIA_VDBOX2),
+	GEN_FW_RANGE(0x1d4000, 0x1d7fff, FORCEWAKE_MEDIA_VDBOX3),
+	GEN_FW_RANGE(0x1d8000, 0x1dbfff, FORCEWAKE_MEDIA_VEBOX1)
+};
+
+/* *Must* be sorted by offset ranges! See intel_fw_table_check(). */
+static const struct intel_forcewake_range __gen12_fw_ranges[] = {
+	GEN_FW_RANGE(0x0, 0xaff, FORCEWAKE_BLITTER),
+	GEN_FW_RANGE(0xb00, 0x1fff, 0), /* uncore range */
+	GEN_FW_RANGE(0x2000, 0x26ff, FORCEWAKE_RENDER),
+	GEN_FW_RANGE(0x2700, 0x2fff, FORCEWAKE_BLITTER),
+	GEN_FW_RANGE(0x3000, 0x3fff, FORCEWAKE_RENDER),
+	GEN_FW_RANGE(0x4000, 0x51ff, FORCEWAKE_BLITTER),
+	GEN_FW_RANGE(0x5200, 0x7fff, FORCEWAKE_RENDER),
+	GEN_FW_RANGE(0x8000, 0x813f, FORCEWAKE_BLITTER),
+	GEN_FW_RANGE(0x8140, 0x815f, FORCEWAKE_RENDER),
+	GEN_FW_RANGE(0x8160, 0x82ff, FORCEWAKE_BLITTER),
+	GEN_FW_RANGE(0x8300, 0x84ff, FORCEWAKE_RENDER),
+	GEN_FW_RANGE(0x8500, 0x8bff, FORCEWAKE_BLITTER),
+	GEN_FW_RANGE(0x8c00, 0x8cff, FORCEWAKE_RENDER),
+	GEN_FW_RANGE(0x8d00, 0x93ff, FORCEWAKE_BLITTER),
+	GEN_FW_RANGE(0x9400, 0x97ff, FORCEWAKE_ALL),
+	GEN_FW_RANGE(0x9800, 0xafff, FORCEWAKE_BLITTER),
+	GEN_FW_RANGE(0xb000, 0xb47f, FORCEWAKE_RENDER),
 	GEN_FW_RANGE(0xb480, 0xdfff, FORCEWAKE_BLITTER),
 	GEN_FW_RANGE(0xe000, 0xe8ff, FORCEWAKE_RENDER),
-	GEN_FW_RANGE(0xe900, 0x243ff, FORCEWAKE_BLITTER),
+	GEN_FW_RANGE(0xe900, 0x147ff, FORCEWAKE_BLITTER),
+	GEN_FW_RANGE(0x14800, 0x148ff, FORCEWAKE_RENDER),
+	GEN_FW_RANGE(0x14900, 0x19fff, FORCEWAKE_BLITTER),
+	GEN_FW_RANGE(0x1a000, 0x1a7ff, FORCEWAKE_RENDER),
+	GEN_FW_RANGE(0x1a800, 0x1afff, FORCEWAKE_BLITTER),
+	GEN_FW_RANGE(0x1b000, 0x1bfff, FORCEWAKE_RENDER),
+	GEN_FW_RANGE(0x1c000, 0x243ff, FORCEWAKE_BLITTER),
 	GEN_FW_RANGE(0x24400, 0x247ff, FORCEWAKE_RENDER),
 	GEN_FW_RANGE(0x24800, 0x3ffff, FORCEWAKE_BLITTER),
 	GEN_FW_RANGE(0x40000, 0x1bffff, 0),
@@ -1228,6 +1290,7 @@ __gen_read(func, 16) \
 __gen_read(func, 32) \
 __gen_read(func, 64)
 
+__gen_reg_read_funcs(gen12_fwtable);
 __gen_reg_read_funcs(gen11_fwtable);
 __gen_reg_read_funcs(fwtable);
 __gen_reg_read_funcs(gen6);
@@ -1319,6 +1382,7 @@ __gen_write(func, 8) \
 __gen_write(func, 16) \
 __gen_write(func, 32)
 
+__gen_reg_write_funcs(gen12_fwtable);
 __gen_reg_write_funcs(gen11_fwtable);
 __gen_reg_write_funcs(fwtable);
 __gen_reg_write_funcs(gen8);
@@ -1690,10 +1754,14 @@ static int uncore_forcewake_init(struct intel_uncore *uncore)
 		ASSIGN_FW_DOMAINS_TABLE(uncore, __gen9_fw_ranges);
 		ASSIGN_WRITE_MMIO_VFUNCS(uncore, fwtable);
 		ASSIGN_READ_MMIO_VFUNCS(uncore, fwtable);
-	} else {
+	} else if (IS_GEN(i915, 11)) {
 		ASSIGN_FW_DOMAINS_TABLE(uncore, __gen11_fw_ranges);
 		ASSIGN_WRITE_MMIO_VFUNCS(uncore, gen11_fwtable);
 		ASSIGN_READ_MMIO_VFUNCS(uncore, gen11_fwtable);
+	} else {
+		ASSIGN_FW_DOMAINS_TABLE(uncore, __gen12_fw_ranges);
+		ASSIGN_WRITE_MMIO_VFUNCS(uncore, gen12_fwtable);
+		ASSIGN_READ_MMIO_VFUNCS(uncore, gen12_fwtable);
 	}
 
 	uncore->pmic_bus_access_nb.notifier_call = i915_pmic_bus_access_notifier;
diff --git a/drivers/gpu/drm/i915/intel_uncore.h b/drivers/gpu/drm/i915/intel_uncore.h
index 414fc2cb0459..dcfa243892c6 100644
--- a/drivers/gpu/drm/i915/intel_uncore.h
+++ b/drivers/gpu/drm/i915/intel_uncore.h
@@ -378,23 +378,23 @@ intel_uncore_read64_2x32(struct intel_uncore *uncore,
 static inline void intel_uncore_rmw(struct intel_uncore *uncore,
 				    i915_reg_t reg, u32 clear, u32 set)
 {
-	u32 val;
+	u32 old, val;
 
-	val = intel_uncore_read(uncore, reg);
-	val &= ~clear;
-	val |= set;
-	intel_uncore_write(uncore, reg, val);
+	old = intel_uncore_read(uncore, reg);
+	val = (old & ~clear) | set;
+	if (val != old)
+		intel_uncore_write(uncore, reg, val);
 }
 
 static inline void intel_uncore_rmw_fw(struct intel_uncore *uncore,
 				       i915_reg_t reg, u32 clear, u32 set)
 {
-	u32 val;
+	u32 old, val;
 
-	val = intel_uncore_read_fw(uncore, reg);
-	val &= ~clear;
-	val |= set;
-	intel_uncore_write_fw(uncore, reg, val);
+	old = intel_uncore_read_fw(uncore, reg);
+	val = (old & ~clear) | set;
+	if (val != old)
+		intel_uncore_write_fw(uncore, reg, val);
 }
 
 static inline int intel_uncore_write_and_verify(struct intel_uncore *uncore,
diff --git a/drivers/gpu/drm/i915/oa/i915_oa_tgl.c b/drivers/gpu/drm/i915/oa/i915_oa_tgl.c
new file mode 100644
index 000000000000..a29d93707345
--- /dev/null
+++ b/drivers/gpu/drm/i915/oa/i915_oa_tgl.c
@@ -0,0 +1,121 @@
+// SPDX-License-Identifier: MIT
+/*
+ * Copyright © 2018 Intel Corporation
+ *
+ * Autogenerated file by GPU Top : https://github.com/rib/gputop
+ * DO NOT EDIT manually!
+ */
+
+#include <linux/sysfs.h>
+
+#include "i915_drv.h"
+#include "i915_oa_tgl.h"
+
+static const struct i915_oa_reg b_counter_config_test_oa[] = {
+	{ _MMIO(0xD920), 0x00000000 },
+	{ _MMIO(0xD900), 0x00000000 },
+	{ _MMIO(0xD904), 0xF0800000 },
+	{ _MMIO(0xD910), 0x00000000 },
+	{ _MMIO(0xD914), 0xF0800000 },
+	{ _MMIO(0xDC40), 0x00FF0000 },
+	{ _MMIO(0xD940), 0x00000004 },
+	{ _MMIO(0xD944), 0x0000FFFF },
+	{ _MMIO(0xDC00), 0x00000004 },
+	{ _MMIO(0xDC04), 0x0000FFFF },
+	{ _MMIO(0xD948), 0x00000003 },
+	{ _MMIO(0xD94C), 0x0000FFFF },
+	{ _MMIO(0xDC08), 0x00000003 },
+	{ _MMIO(0xDC0C), 0x0000FFFF },
+	{ _MMIO(0xD950), 0x00000007 },
+	{ _MMIO(0xD954), 0x0000FFFF },
+	{ _MMIO(0xDC10), 0x00000007 },
+	{ _MMIO(0xDC14), 0x0000FFFF },
+	{ _MMIO(0xD958), 0x00100002 },
+	{ _MMIO(0xD95C), 0x0000FFF7 },
+	{ _MMIO(0xDC18), 0x00100002 },
+	{ _MMIO(0xDC1C), 0x0000FFF7 },
+	{ _MMIO(0xD960), 0x00100002 },
+	{ _MMIO(0xD964), 0x0000FFCF },
+	{ _MMIO(0xDC20), 0x00100002 },
+	{ _MMIO(0xDC24), 0x0000FFCF },
+	{ _MMIO(0xD968), 0x00100082 },
+	{ _MMIO(0xD96C), 0x0000FFEF },
+	{ _MMIO(0xDC28), 0x00100082 },
+	{ _MMIO(0xDC2C), 0x0000FFEF },
+	{ _MMIO(0xD970), 0x001000C2 },
+	{ _MMIO(0xD974), 0x0000FFE7 },
+	{ _MMIO(0xDC30), 0x001000C2 },
+	{ _MMIO(0xDC34), 0x0000FFE7 },
+	{ _MMIO(0xD978), 0x00100001 },
+	{ _MMIO(0xD97C), 0x0000FFE7 },
+	{ _MMIO(0xDC38), 0x00100001 },
+	{ _MMIO(0xDC3C), 0x0000FFE7 },
+};
+
+static const struct i915_oa_reg flex_eu_config_test_oa[] = {
+};
+
+static const struct i915_oa_reg mux_config_test_oa[] = {
+	{ _MMIO(0x0D04), 0x00000200 },
+	{ _MMIO(0x9840), 0x00000000 },
+	{ _MMIO(0x9884), 0x00000000 },
+	{ _MMIO(0x9888), 0x280E0000 },
+	{ _MMIO(0x9888), 0x1E0E0147 },
+	{ _MMIO(0x9888), 0x180E0000 },
+	{ _MMIO(0x9888), 0x160E0000 },
+	{ _MMIO(0x9888), 0x1E0F1000 },
+	{ _MMIO(0x9888), 0x1E104000 },
+	{ _MMIO(0x9888), 0x2E020100 },
+	{ _MMIO(0x9888), 0x2C030004 },
+	{ _MMIO(0x9888), 0x38003000 },
+	{ _MMIO(0x9888), 0x1E0A8000 },
+	{ _MMIO(0x9884), 0x00000003 },
+	{ _MMIO(0x9888), 0x49110000 },
+	{ _MMIO(0x9888), 0x5D101400 },
+	{ _MMIO(0x9888), 0x1D140020 },
+	{ _MMIO(0x9888), 0x1D1103A3 },
+	{ _MMIO(0x9888), 0x01110000 },
+	{ _MMIO(0x9888), 0x61111000 },
+	{ _MMIO(0x9888), 0x1F128000 },
+	{ _MMIO(0x9888), 0x17100000 },
+	{ _MMIO(0x9888), 0x55100630 },
+	{ _MMIO(0x9888), 0x57100000 },
+	{ _MMIO(0x9888), 0x31100000 },
+	{ _MMIO(0x9884), 0x00000003 },
+	{ _MMIO(0x9888), 0x65100002 },
+	{ _MMIO(0x9884), 0x00000000 },
+	{ _MMIO(0x9888), 0x42000001 },
+};
+
+static ssize_t
+show_test_oa_id(struct device *kdev, struct device_attribute *attr, char *buf)
+{
+	return sprintf(buf, "1\n");
+}
+
+void
+i915_perf_load_test_config_tgl(struct drm_i915_private *dev_priv)
+{
+	strlcpy(dev_priv->perf.test_config.uuid,
+		"80a833f0-2504-4321-8894-e9277844ce7b",
+		sizeof(dev_priv->perf.test_config.uuid));
+	dev_priv->perf.test_config.id = 1;
+
+	dev_priv->perf.test_config.mux_regs = mux_config_test_oa;
+	dev_priv->perf.test_config.mux_regs_len = ARRAY_SIZE(mux_config_test_oa);
+
+	dev_priv->perf.test_config.b_counter_regs = b_counter_config_test_oa;
+	dev_priv->perf.test_config.b_counter_regs_len = ARRAY_SIZE(b_counter_config_test_oa);
+
+	dev_priv->perf.test_config.flex_regs = flex_eu_config_test_oa;
+	dev_priv->perf.test_config.flex_regs_len = ARRAY_SIZE(flex_eu_config_test_oa);
+
+	dev_priv->perf.test_config.sysfs_metric.name = "80a833f0-2504-4321-8894-e9277844ce7b";
+	dev_priv->perf.test_config.sysfs_metric.attrs = dev_priv->perf.test_config.attrs;
+
+	dev_priv->perf.test_config.attrs[0] = &dev_priv->perf.test_config.sysfs_metric_id.attr;
+
+	dev_priv->perf.test_config.sysfs_metric_id.attr.name = "id";
+	dev_priv->perf.test_config.sysfs_metric_id.attr.mode = 0444;
+	dev_priv->perf.test_config.sysfs_metric_id.show = show_test_oa_id;
+}
diff --git a/drivers/gpu/drm/i915/oa/i915_oa_tgl.h b/drivers/gpu/drm/i915/oa/i915_oa_tgl.h
new file mode 100644
index 000000000000..4c25f0be825c
--- /dev/null
+++ b/drivers/gpu/drm/i915/oa/i915_oa_tgl.h
@@ -0,0 +1,16 @@
+/* SPDX-License-Identifier: MIT */
+/*
+ * Copyright © 2018 Intel Corporation
+ *
+ * Autogenerated file by GPU Top : https://github.com/rib/gputop
+ * DO NOT EDIT manually!
+ */
+
+#ifndef __I915_OA_TGL_H__
+#define __I915_OA_TGL_H__
+
+struct drm_i915_private;
+
+void i915_perf_load_test_config_tgl(struct drm_i915_private *dev_priv);
+
+#endif
diff --git a/drivers/gpu/drm/i915/selftests/i915_active.c b/drivers/gpu/drm/i915/selftests/i915_active.c
index 77d844ac8b71..260b0ee5d1e3 100644
--- a/drivers/gpu/drm/i915/selftests/i915_active.c
+++ b/drivers/gpu/drm/i915/selftests/i915_active.c
@@ -68,7 +68,7 @@ static struct live_active *__live_alloc(struct drm_i915_private *i915)
 		return NULL;
 
 	kref_init(&active->ref);
-	i915_active_init(i915, &active->base, __live_active, __live_retire);
+	i915_active_init(&active->base, __live_active, __live_retire);
 
 	return active;
 }
@@ -79,7 +79,6 @@ __live_active_setup(struct drm_i915_private *i915)
 	struct intel_engine_cs *engine;
 	struct i915_sw_fence *submit;
 	struct live_active *active;
-	enum intel_engine_id id;
 	unsigned int count = 0;
 	int err = 0;
 
@@ -97,7 +96,7 @@ __live_active_setup(struct drm_i915_private *i915)
 	if (err)
 		goto out;
 
-	for_each_engine(engine, i915, id) {
+	for_each_uabi_engine(engine, i915) {
 		struct i915_request *rq;
 
 		rq = i915_request_create(engine->kernel_context);
@@ -110,7 +109,7 @@ __live_active_setup(struct drm_i915_private *i915)
 						       submit,
 						       GFP_KERNEL);
 		if (err >= 0)
-			err = i915_active_ref(&active->base, rq->timeline, rq);
+			err = i915_active_add_request(&active->base, rq);
 		i915_request_add(rq);
 		if (err) {
 			pr_err("Failed to track active ref!\n");
@@ -121,7 +120,7 @@ __live_active_setup(struct drm_i915_private *i915)
 	}
 
 	i915_active_release(&active->base);
-	if (active->retired && count) {
+	if (READ_ONCE(active->retired) && count) {
 		pr_err("i915_active retired before submission!\n");
 		err = -EINVAL;
 	}
@@ -146,35 +145,25 @@ static int live_active_wait(void *arg)
 {
 	struct drm_i915_private *i915 = arg;
 	struct live_active *active;
-	intel_wakeref_t wakeref;
 	int err = 0;
 
 	/* Check that we get a callback when requests retire upon waiting */
 
-	mutex_lock(&i915->drm.struct_mutex);
-	wakeref = intel_runtime_pm_get(&i915->runtime_pm);
-
 	active = __live_active_setup(i915);
-	if (IS_ERR(active)) {
-		err = PTR_ERR(active);
-		goto err;
-	}
+	if (IS_ERR(active))
+		return PTR_ERR(active);
 
 	i915_active_wait(&active->base);
-	if (!active->retired) {
+	if (!READ_ONCE(active->retired)) {
 		pr_err("i915_active not retired after waiting!\n");
 		err = -EINVAL;
 	}
 
 	__live_put(active);
 
-	if (igt_flush_test(i915, I915_WAIT_LOCKED))
+	if (igt_flush_test(i915))
 		err = -EIO;
 
-err:
-	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
-	mutex_unlock(&i915->drm.struct_mutex);
-
 	return err;
 }
 
@@ -182,35 +171,25 @@ static int live_active_retire(void *arg)
 {
 	struct drm_i915_private *i915 = arg;
 	struct live_active *active;
-	intel_wakeref_t wakeref;
 	int err = 0;
 
 	/* Check that we get a callback when requests are indirectly retired */
 
-	mutex_lock(&i915->drm.struct_mutex);
-	wakeref = intel_runtime_pm_get(&i915->runtime_pm);
-
 	active = __live_active_setup(i915);
-	if (IS_ERR(active)) {
-		err = PTR_ERR(active);
-		goto err;
-	}
+	if (IS_ERR(active))
+		return PTR_ERR(active);
 
 	/* waits for & retires all requests */
-	if (igt_flush_test(i915, I915_WAIT_LOCKED))
+	if (igt_flush_test(i915))
 		err = -EIO;
 
-	if (!active->retired) {
+	if (!READ_ONCE(active->retired)) {
 		pr_err("i915_active not retired after flushing!\n");
 		err = -EINVAL;
 	}
 
 	__live_put(active);
 
-err:
-	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
-	mutex_unlock(&i915->drm.struct_mutex);
-
 	return err;
 }
 
@@ -226,3 +205,48 @@ int i915_active_live_selftests(struct drm_i915_private *i915)
 
 	return i915_subtests(tests, i915);
 }
+
+static struct intel_engine_cs *node_to_barrier(struct active_node *it)
+{
+	struct intel_engine_cs *engine;
+
+	if (!is_barrier(&it->base))
+		return NULL;
+
+	engine = __barrier_to_engine(it);
+	smp_rmb(); /* serialise with add_active_barriers */
+	if (!is_barrier(&it->base))
+		return NULL;
+
+	return engine;
+}
+
+void i915_active_print(struct i915_active *ref, struct drm_printer *m)
+{
+	drm_printf(m, "active %pS:%pS\n", ref->active, ref->retire);
+	drm_printf(m, "\tcount: %d\n", atomic_read(&ref->count));
+	drm_printf(m, "\tpreallocated barriers? %s\n",
+		   yesno(!llist_empty(&ref->preallocated_barriers)));
+
+	if (i915_active_acquire_if_busy(ref)) {
+		struct active_node *it, *n;
+
+		rbtree_postorder_for_each_entry_safe(it, n, &ref->tree, node) {
+			struct intel_engine_cs *engine;
+
+			engine = node_to_barrier(it);
+			if (engine) {
+				drm_printf(m, "\tbarrier: %s\n", engine->name);
+				continue;
+			}
+
+			if (i915_active_fence_isset(&it->base)) {
+				drm_printf(m,
+					   "\ttimeline: %llx\n", it->timeline);
+				continue;
+			}
+		}
+
+		i915_active_release(ref);
+	}
+}
diff --git a/drivers/gpu/drm/i915/selftests/i915_buddy.c b/drivers/gpu/drm/i915/selftests/i915_buddy.c
index 23f784eae1e7..1b856bae67b5 100644
--- a/drivers/gpu/drm/i915/selftests/i915_buddy.c
+++ b/drivers/gpu/drm/i915/selftests/i915_buddy.c
@@ -375,6 +375,8 @@ retry:
 
 		if (err)
 			break;
+
+		cond_resched();
 	}
 
 	if (err == -ENOMEM)
@@ -687,6 +689,8 @@ static int igt_buddy_alloc_range(void *arg)
 		rem -= size;
 		if (!rem)
 			break;
+
+		cond_resched();
 	}
 
 	if (err == -ENOMEM)
diff --git a/drivers/gpu/drm/i915/selftests/i915_gem.c b/drivers/gpu/drm/i915/selftests/i915_gem.c
index bb6dd54a6ff3..d83f6bf6d9d4 100644
--- a/drivers/gpu/drm/i915/selftests/i915_gem.c
+++ b/drivers/gpu/drm/i915/selftests/i915_gem.c
@@ -15,23 +15,26 @@
 #include "igt_flush_test.h"
 #include "mock_drm.h"
 
-static int switch_to_context(struct drm_i915_private *i915,
-			     struct i915_gem_context *ctx)
+static int switch_to_context(struct i915_gem_context *ctx)
 {
-	struct intel_engine_cs *engine;
-	enum intel_engine_id id;
+	struct i915_gem_engines_iter it;
+	struct intel_context *ce;
+	int err = 0;
 
-	for_each_engine(engine, i915, id) {
+	for_each_gem_engine(ce, i915_gem_context_lock_engines(ctx), it) {
 		struct i915_request *rq;
 
-		rq = igt_request_alloc(ctx, engine);
-		if (IS_ERR(rq))
-			return PTR_ERR(rq);
+		rq = intel_context_create_request(ce);
+		if (IS_ERR(rq)) {
+			err = PTR_ERR(rq);
+			break;
+		}
 
 		i915_request_add(rq);
 	}
+	i915_gem_context_unlock_engines(ctx);
 
-	return 0;
+	return err;
 }
 
 static void trash_stolen(struct drm_i915_private *i915)
@@ -42,6 +45,10 @@ static void trash_stolen(struct drm_i915_private *i915)
 	unsigned long page;
 	u32 prng = 0x12345678;
 
+	/* XXX: fsck. needs some more thought... */
+	if (!i915_ggtt_has_aperture(ggtt))
+		return;
+
 	for (page = 0; page < size; page += PAGE_SIZE) {
 		const dma_addr_t dma = i915->dsm.start + page;
 		u32 __iomem *s;
@@ -117,7 +124,10 @@ static void pm_resume(struct drm_i915_private *i915)
 	 */
 	with_intel_runtime_pm(&i915->runtime_pm, wakeref) {
 		intel_gt_sanitize(&i915->gt, false);
-		i915_gem_sanitize(i915);
+
+		i915_gem_restore_gtt_mappings(i915);
+		i915_gem_restore_fences(&i915->ggtt);
+
 		i915_gem_resume(i915);
 	}
 }
@@ -134,11 +144,9 @@ static int igt_gem_suspend(void *arg)
 		return PTR_ERR(file);
 
 	err = -ENOMEM;
-	mutex_lock(&i915->drm.struct_mutex);
 	ctx = live_context(i915, file);
 	if (!IS_ERR(ctx))
-		err = switch_to_context(i915, ctx);
-	mutex_unlock(&i915->drm.struct_mutex);
+		err = switch_to_context(ctx);
 	if (err)
 		goto out;
 
@@ -153,9 +161,7 @@ static int igt_gem_suspend(void *arg)
 
 	pm_resume(i915);
 
-	mutex_lock(&i915->drm.struct_mutex);
-	err = switch_to_context(i915, ctx);
-	mutex_unlock(&i915->drm.struct_mutex);
+	err = switch_to_context(ctx);
 out:
 	mock_file_free(i915, file);
 	return err;
@@ -173,11 +179,9 @@ static int igt_gem_hibernate(void *arg)
 		return PTR_ERR(file);
 
 	err = -ENOMEM;
-	mutex_lock(&i915->drm.struct_mutex);
 	ctx = live_context(i915, file);
 	if (!IS_ERR(ctx))
-		err = switch_to_context(i915, ctx);
-	mutex_unlock(&i915->drm.struct_mutex);
+		err = switch_to_context(ctx);
 	if (err)
 		goto out;
 
@@ -192,9 +196,7 @@ static int igt_gem_hibernate(void *arg)
 
 	pm_resume(i915);
 
-	mutex_lock(&i915->drm.struct_mutex);
-	err = switch_to_context(i915, ctx);
-	mutex_unlock(&i915->drm.struct_mutex);
+	err = switch_to_context(ctx);
 out:
 	mock_file_free(i915, file);
 	return err;
diff --git a/drivers/gpu/drm/i915/selftests/i915_gem_evict.c b/drivers/gpu/drm/i915/selftests/i915_gem_evict.c
index cb30c669b1b7..42e948144f1b 100644
--- a/drivers/gpu/drm/i915/selftests/i915_gem_evict.c
+++ b/drivers/gpu/drm/i915/selftests/i915_gem_evict.c
@@ -43,8 +43,7 @@ static void quirk_add(struct drm_i915_gem_object *obj,
 	list_add(&obj->st_link, objects);
 }
 
-static int populate_ggtt(struct drm_i915_private *i915,
-			 struct list_head *objects)
+static int populate_ggtt(struct i915_ggtt *ggtt, struct list_head *objects)
 {
 	unsigned long unbound, bound, count;
 	struct drm_i915_gem_object *obj;
@@ -53,7 +52,8 @@ static int populate_ggtt(struct drm_i915_private *i915,
 	do {
 		struct i915_vma *vma;
 
-		obj = i915_gem_object_create_internal(i915, I915_GTT_PAGE_SIZE);
+		obj = i915_gem_object_create_internal(ggtt->vm.i915,
+						      I915_GTT_PAGE_SIZE);
 		if (IS_ERR(obj))
 			return PTR_ERR(obj);
 
@@ -70,7 +70,7 @@ static int populate_ggtt(struct drm_i915_private *i915,
 		count++;
 	} while (1);
 	pr_debug("Filled GGTT with %lu pages [%llu total]\n",
-		 count, i915->ggtt.vm.total / PAGE_SIZE);
+		 count, ggtt->vm.total / PAGE_SIZE);
 
 	bound = 0;
 	unbound = 0;
@@ -96,7 +96,7 @@ static int populate_ggtt(struct drm_i915_private *i915,
 		return -EINVAL;
 	}
 
-	if (list_empty(&i915->ggtt.vm.bound_list)) {
+	if (list_empty(&ggtt->vm.bound_list)) {
 		pr_err("No objects on the GGTT inactive list!\n");
 		return -EINVAL;
 	}
@@ -104,20 +104,16 @@ static int populate_ggtt(struct drm_i915_private *i915,
 	return 0;
 }
 
-static void unpin_ggtt(struct drm_i915_private *i915)
+static void unpin_ggtt(struct i915_ggtt *ggtt)
 {
-	struct i915_ggtt *ggtt = &i915->ggtt;
 	struct i915_vma *vma;
 
-	mutex_lock(&ggtt->vm.mutex);
-	list_for_each_entry(vma, &i915->ggtt.vm.bound_list, vm_link)
+	list_for_each_entry(vma, &ggtt->vm.bound_list, vm_link)
 		if (vma->obj->mm.quirked)
 			i915_vma_unpin(vma);
-	mutex_unlock(&ggtt->vm.mutex);
 }
 
-static void cleanup_objects(struct drm_i915_private *i915,
-			    struct list_head *list)
+static void cleanup_objects(struct i915_ggtt *ggtt, struct list_head *list)
 {
 	struct drm_i915_gem_object *obj, *on;
 
@@ -127,44 +123,44 @@ static void cleanup_objects(struct drm_i915_private *i915,
 		i915_gem_object_put(obj);
 	}
 
-	mutex_unlock(&i915->drm.struct_mutex);
-
-	i915_gem_drain_freed_objects(i915);
-
-	mutex_lock(&i915->drm.struct_mutex);
+	i915_gem_drain_freed_objects(ggtt->vm.i915);
 }
 
 static int igt_evict_something(void *arg)
 {
-	struct drm_i915_private *i915 = arg;
-	struct i915_ggtt *ggtt = &i915->ggtt;
+	struct intel_gt *gt = arg;
+	struct i915_ggtt *ggtt = gt->ggtt;
 	LIST_HEAD(objects);
 	int err;
 
 	/* Fill the GGTT with pinned objects and try to evict one. */
 
-	err = populate_ggtt(i915, &objects);
+	err = populate_ggtt(ggtt, &objects);
 	if (err)
 		goto cleanup;
 
 	/* Everything is pinned, nothing should happen */
+	mutex_lock(&ggtt->vm.mutex);
 	err = i915_gem_evict_something(&ggtt->vm,
 				       I915_GTT_PAGE_SIZE, 0, 0,
 				       0, U64_MAX,
 				       0);
+	mutex_unlock(&ggtt->vm.mutex);
 	if (err != -ENOSPC) {
 		pr_err("i915_gem_evict_something failed on a full GGTT with err=%d\n",
 		       err);
 		goto cleanup;
 	}
 
-	unpin_ggtt(i915);
+	unpin_ggtt(ggtt);
 
 	/* Everything is unpinned, we should be able to evict something */
+	mutex_lock(&ggtt->vm.mutex);
 	err = i915_gem_evict_something(&ggtt->vm,
 				       I915_GTT_PAGE_SIZE, 0, 0,
 				       0, U64_MAX,
 				       0);
+	mutex_unlock(&ggtt->vm.mutex);
 	if (err) {
 		pr_err("i915_gem_evict_something failed on a full GGTT with err=%d\n",
 		       err);
@@ -172,13 +168,14 @@ static int igt_evict_something(void *arg)
 	}
 
 cleanup:
-	cleanup_objects(i915, &objects);
+	cleanup_objects(ggtt, &objects);
 	return err;
 }
 
 static int igt_overcommit(void *arg)
 {
-	struct drm_i915_private *i915 = arg;
+	struct intel_gt *gt = arg;
+	struct i915_ggtt *ggtt = gt->ggtt;
 	struct drm_i915_gem_object *obj;
 	struct i915_vma *vma;
 	LIST_HEAD(objects);
@@ -188,11 +185,11 @@ static int igt_overcommit(void *arg)
 	 * We expect it to fail.
 	 */
 
-	err = populate_ggtt(i915, &objects);
+	err = populate_ggtt(ggtt, &objects);
 	if (err)
 		goto cleanup;
 
-	obj = i915_gem_object_create_internal(i915, I915_GTT_PAGE_SIZE);
+	obj = i915_gem_object_create_internal(gt->i915, I915_GTT_PAGE_SIZE);
 	if (IS_ERR(obj)) {
 		err = PTR_ERR(obj);
 		goto cleanup;
@@ -208,14 +205,14 @@ static int igt_overcommit(void *arg)
 	}
 
 cleanup:
-	cleanup_objects(i915, &objects);
+	cleanup_objects(ggtt, &objects);
 	return err;
 }
 
 static int igt_evict_for_vma(void *arg)
 {
-	struct drm_i915_private *i915 = arg;
-	struct i915_ggtt *ggtt = &i915->ggtt;
+	struct intel_gt *gt = arg;
+	struct i915_ggtt *ggtt = gt->ggtt;
 	struct drm_mm_node target = {
 		.start = 0,
 		.size = 4096,
@@ -225,22 +222,26 @@ static int igt_evict_for_vma(void *arg)
 
 	/* Fill the GGTT with pinned objects and try to evict a range. */
 
-	err = populate_ggtt(i915, &objects);
+	err = populate_ggtt(ggtt, &objects);
 	if (err)
 		goto cleanup;
 
 	/* Everything is pinned, nothing should happen */
+	mutex_lock(&ggtt->vm.mutex);
 	err = i915_gem_evict_for_node(&ggtt->vm, &target, 0);
+	mutex_unlock(&ggtt->vm.mutex);
 	if (err != -ENOSPC) {
 		pr_err("i915_gem_evict_for_node on a full GGTT returned err=%d\n",
 		       err);
 		goto cleanup;
 	}
 
-	unpin_ggtt(i915);
+	unpin_ggtt(ggtt);
 
 	/* Everything is unpinned, we should be able to evict the node */
+	mutex_lock(&ggtt->vm.mutex);
 	err = i915_gem_evict_for_node(&ggtt->vm, &target, 0);
+	mutex_unlock(&ggtt->vm.mutex);
 	if (err) {
 		pr_err("i915_gem_evict_for_node returned err=%d\n",
 		       err);
@@ -248,7 +249,7 @@ static int igt_evict_for_vma(void *arg)
 	}
 
 cleanup:
-	cleanup_objects(i915, &objects);
+	cleanup_objects(ggtt, &objects);
 	return err;
 }
 
@@ -261,8 +262,8 @@ static void mock_color_adjust(const struct drm_mm_node *node,
 
 static int igt_evict_for_cache_color(void *arg)
 {
-	struct drm_i915_private *i915 = arg;
-	struct i915_ggtt *ggtt = &i915->ggtt;
+	struct intel_gt *gt = arg;
+	struct i915_ggtt *ggtt = gt->ggtt;
 	const unsigned long flags = PIN_OFFSET_FIXED;
 	struct drm_mm_node target = {
 		.start = I915_GTT_PAGE_SIZE * 2,
@@ -274,14 +275,16 @@ static int igt_evict_for_cache_color(void *arg)
 	LIST_HEAD(objects);
 	int err;
 
-	/* Currently the use of color_adjust is limited to cache domains within
-	 * the ggtt, and so the presence of mm.color_adjust is assumed to be
-	 * i915_gtt_color_adjust throughout our driver, so using a mock color
-	 * adjust will work just fine for our purposes.
+	/*
+	 * Currently the use of color_adjust for the GGTT is limited to cache
+	 * coloring and guard pages, and so the presence of mm.color_adjust for
+	 * the GGTT is assumed to be i915_ggtt_color_adjust, hence using a mock
+	 * color adjust will work just fine for our purposes.
 	 */
 	ggtt->vm.mm.color_adjust = mock_color_adjust;
+	GEM_BUG_ON(!i915_vm_has_cache_coloring(&ggtt->vm));
 
-	obj = i915_gem_object_create_internal(i915, I915_GTT_PAGE_SIZE);
+	obj = i915_gem_object_create_internal(gt->i915, I915_GTT_PAGE_SIZE);
 	if (IS_ERR(obj)) {
 		err = PTR_ERR(obj);
 		goto cleanup;
@@ -297,7 +300,7 @@ static int igt_evict_for_cache_color(void *arg)
 		goto cleanup;
 	}
 
-	obj = i915_gem_object_create_internal(i915, I915_GTT_PAGE_SIZE);
+	obj = i915_gem_object_create_internal(gt->i915, I915_GTT_PAGE_SIZE);
 	if (IS_ERR(obj)) {
 		err = PTR_ERR(obj);
 		goto cleanup;
@@ -317,7 +320,9 @@ static int igt_evict_for_cache_color(void *arg)
 	i915_vma_unpin(vma);
 
 	/* Remove just the second vma */
+	mutex_lock(&ggtt->vm.mutex);
 	err = i915_gem_evict_for_node(&ggtt->vm, &target, 0);
+	mutex_unlock(&ggtt->vm.mutex);
 	if (err) {
 		pr_err("[0]i915_gem_evict_for_node returned err=%d\n", err);
 		goto cleanup;
@@ -328,7 +333,9 @@ static int igt_evict_for_cache_color(void *arg)
 	 */
 	target.color = I915_CACHE_L3_LLC;
 
+	mutex_lock(&ggtt->vm.mutex);
 	err = i915_gem_evict_for_node(&ggtt->vm, &target, 0);
+	mutex_unlock(&ggtt->vm.mutex);
 	if (!err) {
 		pr_err("[1]i915_gem_evict_for_node returned err=%d\n", err);
 		err = -EINVAL;
@@ -338,36 +345,40 @@ static int igt_evict_for_cache_color(void *arg)
 	err = 0;
 
 cleanup:
-	unpin_ggtt(i915);
-	cleanup_objects(i915, &objects);
+	unpin_ggtt(ggtt);
+	cleanup_objects(ggtt, &objects);
 	ggtt->vm.mm.color_adjust = NULL;
 	return err;
 }
 
 static int igt_evict_vm(void *arg)
 {
-	struct drm_i915_private *i915 = arg;
-	struct i915_ggtt *ggtt = &i915->ggtt;
+	struct intel_gt *gt = arg;
+	struct i915_ggtt *ggtt = gt->ggtt;
 	LIST_HEAD(objects);
 	int err;
 
 	/* Fill the GGTT with pinned objects and try to evict everything. */
 
-	err = populate_ggtt(i915, &objects);
+	err = populate_ggtt(ggtt, &objects);
 	if (err)
 		goto cleanup;
 
 	/* Everything is pinned, nothing should happen */
+	mutex_lock(&ggtt->vm.mutex);
 	err = i915_gem_evict_vm(&ggtt->vm);
+	mutex_unlock(&ggtt->vm.mutex);
 	if (err) {
 		pr_err("i915_gem_evict_vm on a full GGTT returned err=%d]\n",
 		       err);
 		goto cleanup;
 	}
 
-	unpin_ggtt(i915);
+	unpin_ggtt(ggtt);
 
+	mutex_lock(&ggtt->vm.mutex);
 	err = i915_gem_evict_vm(&ggtt->vm);
+	mutex_unlock(&ggtt->vm.mutex);
 	if (err) {
 		pr_err("i915_gem_evict_vm on a full GGTT returned err=%d]\n",
 		       err);
@@ -375,14 +386,16 @@ static int igt_evict_vm(void *arg)
 	}
 
 cleanup:
-	cleanup_objects(i915, &objects);
+	cleanup_objects(ggtt, &objects);
 	return err;
 }
 
 static int igt_evict_contexts(void *arg)
 {
 	const u64 PRETEND_GGTT_SIZE = 16ull << 20;
-	struct drm_i915_private *i915 = arg;
+	struct intel_gt *gt = arg;
+	struct i915_ggtt *ggtt = gt->ggtt;
+	struct drm_i915_private *i915 = gt->i915;
 	struct intel_engine_cs *engine;
 	enum intel_engine_id id;
 	struct reserved {
@@ -408,14 +421,14 @@ static int igt_evict_contexts(void *arg)
 	if (!HAS_FULL_PPGTT(i915))
 		return 0;
 
-	mutex_lock(&i915->drm.struct_mutex);
 	wakeref = intel_runtime_pm_get(&i915->runtime_pm);
 
 	/* Reserve a block so that we know we have enough to fit a few rq */
 	memset(&hole, 0, sizeof(hole));
-	err = i915_gem_gtt_insert(&i915->ggtt.vm, &hole,
+	mutex_lock(&ggtt->vm.mutex);
+	err = i915_gem_gtt_insert(&ggtt->vm, &hole,
 				  PRETEND_GGTT_SIZE, 0, I915_COLOR_UNEVICTABLE,
-				  0, i915->ggtt.vm.total,
+				  0, ggtt->vm.total,
 				  PIN_NOEVICT);
 	if (err)
 		goto out_locked;
@@ -425,15 +438,17 @@ static int igt_evict_contexts(void *arg)
 	do {
 		struct reserved *r;
 
+		mutex_unlock(&ggtt->vm.mutex);
 		r = kcalloc(1, sizeof(*r), GFP_KERNEL);
+		mutex_lock(&ggtt->vm.mutex);
 		if (!r) {
 			err = -ENOMEM;
 			goto out_locked;
 		}
 
-		if (i915_gem_gtt_insert(&i915->ggtt.vm, &r->node,
+		if (i915_gem_gtt_insert(&ggtt->vm, &r->node,
 					1ul << 20, 0, I915_COLOR_UNEVICTABLE,
-					0, i915->ggtt.vm.total,
+					0, ggtt->vm.total,
 					PIN_NOEVICT)) {
 			kfree(r);
 			break;
@@ -445,11 +460,11 @@ static int igt_evict_contexts(void *arg)
 		count++;
 	} while (1);
 	drm_mm_remove_node(&hole);
-	mutex_unlock(&i915->drm.struct_mutex);
+	mutex_unlock(&ggtt->vm.mutex);
 	pr_info("Filled GGTT with %lu 1MiB nodes\n", count);
 
 	/* Overfill the GGTT with context objects and so try to evict one. */
-	for_each_engine(engine, i915, id) {
+	for_each_engine(engine, gt, id) {
 		struct i915_sw_fence fence;
 		struct drm_file *file;
 
@@ -460,7 +475,6 @@ static int igt_evict_contexts(void *arg)
 		}
 
 		count = 0;
-		mutex_lock(&i915->drm.struct_mutex);
 		onstack_fence_init(&fence);
 		do {
 			struct i915_request *rq;
@@ -478,8 +492,8 @@ static int igt_evict_contexts(void *arg)
 			if (IS_ERR(rq)) {
 				/* When full, fail_if_busy will trigger EBUSY */
 				if (PTR_ERR(rq) != -EBUSY) {
-					pr_err("Unexpected error from request alloc (ctx hw id %u, on %s): %d\n",
-					       ctx->hw_id, engine->name,
+					pr_err("Unexpected error from request alloc (on %s): %d\n",
+					       engine->name,
 					       (int)PTR_ERR(rq));
 					err = PTR_ERR(rq);
 				}
@@ -497,8 +511,6 @@ static int igt_evict_contexts(void *arg)
 			count++;
 			err = 0;
 		} while(1);
-		mutex_unlock(&i915->drm.struct_mutex);
-
 		onstack_fence_fini(&fence);
 		pr_info("Submitted %lu contexts/requests on %s\n",
 			count, engine->name);
@@ -508,9 +520,9 @@ static int igt_evict_contexts(void *arg)
 			break;
 	}
 
-	mutex_lock(&i915->drm.struct_mutex);
+	mutex_lock(&ggtt->vm.mutex);
 out_locked:
-	if (igt_flush_test(i915, I915_WAIT_LOCKED))
+	if (igt_flush_test(i915))
 		err = -EIO;
 	while (reserved) {
 		struct reserved *next = reserved->next;
@@ -522,8 +534,8 @@ out_locked:
 	}
 	if (drm_mm_node_allocated(&hole))
 		drm_mm_remove_node(&hole);
+	mutex_unlock(&ggtt->vm.mutex);
 	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
-	mutex_unlock(&i915->drm.struct_mutex);
 
 	return err;
 }
@@ -545,11 +557,8 @@ int i915_gem_evict_mock_selftests(void)
 	if (!i915)
 		return -ENOMEM;
 
-	mutex_lock(&i915->drm.struct_mutex);
 	with_intel_runtime_pm(&i915->runtime_pm, wakeref)
-		err = i915_subtests(tests, i915);
-
-	mutex_unlock(&i915->drm.struct_mutex);
+		err = i915_subtests(tests, &i915->gt);
 
 	drm_dev_put(&i915->drm);
 	return err;
@@ -564,5 +573,5 @@ int i915_gem_evict_live_selftests(struct drm_i915_private *i915)
 	if (intel_gt_is_wedged(&i915->gt))
 		return 0;
 
-	return i915_subtests(tests, i915);
+	return intel_gt_live_subtests(tests, &i915->gt);
 }
diff --git a/drivers/gpu/drm/i915/selftests/i915_gem_gtt.c b/drivers/gpu/drm/i915/selftests/i915_gem_gtt.c
index 31a51ca1ddcb..3f7e80fb3bbd 100644
--- a/drivers/gpu/drm/i915/selftests/i915_gem_gtt.c
+++ b/drivers/gpu/drm/i915/selftests/i915_gem_gtt.c
@@ -25,26 +25,20 @@
 #include <linux/list_sort.h>
 #include <linux/prime_numbers.h>
 
+#include "gem/i915_gem_context.h"
 #include "gem/selftests/mock_context.h"
+#include "gt/intel_context.h"
 
 #include "i915_random.h"
 #include "i915_selftest.h"
 
 #include "mock_drm.h"
 #include "mock_gem_device.h"
+#include "igt_flush_test.h"
 
 static void cleanup_freed_objects(struct drm_i915_private *i915)
 {
-	/*
-	 * As we may hold onto the struct_mutex for inordinate lengths of
-	 * time, the NMI khungtaskd detector may fire for the free objects
-	 * worker.
-	 */
-	mutex_unlock(&i915->drm.struct_mutex);
-
 	i915_gem_drain_freed_objects(i915);
-
-	mutex_lock(&i915->drm.struct_mutex);
 }
 
 static void fake_free_pages(struct drm_i915_gem_object *obj,
@@ -88,8 +82,6 @@ static int fake_get_pages(struct drm_i915_gem_object *obj)
 	}
 	GEM_BUG_ON(rem);
 
-	obj->mm.madv = I915_MADV_DONTNEED;
-
 	__i915_gem_object_set_pages(obj, pages, sg_page_sizes);
 
 	return 0;
@@ -101,7 +93,6 @@ static void fake_put_pages(struct drm_i915_gem_object *obj,
 {
 	fake_free_pages(obj, pages);
 	obj->mm.dirty = false;
-	obj->mm.madv = I915_MADV_WILLNEED;
 }
 
 static const struct drm_i915_gem_object_ops fake_ops = {
@@ -113,6 +104,7 @@ static const struct drm_i915_gem_object_ops fake_ops = {
 static struct drm_i915_gem_object *
 fake_dma_object(struct drm_i915_private *i915, u64 size)
 {
+	static struct lock_class_key lock_class;
 	struct drm_i915_gem_object *obj;
 
 	GEM_BUG_ON(!size);
@@ -126,7 +118,9 @@ fake_dma_object(struct drm_i915_private *i915, u64 size)
 		goto err;
 
 	drm_gem_private_object_init(&i915->drm, &obj->base, size);
-	i915_gem_object_init(obj, &fake_ops);
+	i915_gem_object_init(obj, &fake_ops, &lock_class);
+
+	i915_gem_object_set_volatile(obj);
 
 	obj->write_domain = I915_GEM_DOMAIN_CPU;
 	obj->read_domains = I915_GEM_DOMAIN_CPU;
@@ -293,18 +287,20 @@ static int lowlevel_hole(struct drm_i915_private *i915,
 			mock_vma.node.size = BIT_ULL(size);
 			mock_vma.node.start = addr;
 
-			wakeref = intel_runtime_pm_get(&i915->runtime_pm);
-			vm->insert_entries(vm, &mock_vma, I915_CACHE_NONE, 0);
-			intel_runtime_pm_put(&i915->runtime_pm, wakeref);
+			with_intel_runtime_pm(&i915->runtime_pm, wakeref)
+				vm->insert_entries(vm, &mock_vma,
+						   I915_CACHE_NONE, 0);
 		}
 		count = n;
 
 		i915_random_reorder(order, count, &prng);
 		for (n = 0; n < count; n++) {
 			u64 addr = hole_start + order[n] * BIT_ULL(size);
+			intel_wakeref_t wakeref;
 
 			GEM_BUG_ON(addr + BIT_ULL(size) > vm->total);
-			vm->clear_range(vm, addr, BIT_ULL(size));
+			with_intel_runtime_pm(&i915->runtime_pm, wakeref)
+				vm->clear_range(vm, addr, BIT_ULL(size));
 		}
 
 		i915_gem_object_unpin_pages(obj);
@@ -875,6 +871,15 @@ static int __shrink_hole(struct drm_i915_private *i915,
 		i915_vma_unpin(vma);
 		addr += size;
 
+		/*
+		 * Since we are injecting allocation faults at random intervals,
+		 * wait for this allocation to complete before we change the
+		 * faultinjection.
+		 */
+		err = i915_vma_sync(vma);
+		if (err)
+			break;
+
 		if (igt_timeout(end_time,
 				"%s timed out at ofset %llx [%llx - %llx]\n",
 				__func__, addr, hole_start, hole_end)) {
@@ -1008,21 +1013,19 @@ static int exercise_ppgtt(struct drm_i915_private *dev_priv,
 	if (IS_ERR(file))
 		return PTR_ERR(file);
 
-	mutex_lock(&dev_priv->drm.struct_mutex);
 	ppgtt = i915_ppgtt_create(dev_priv);
 	if (IS_ERR(ppgtt)) {
 		err = PTR_ERR(ppgtt);
-		goto out_unlock;
+		goto out_free;
 	}
 	GEM_BUG_ON(offset_in_page(ppgtt->vm.total));
-	GEM_BUG_ON(ppgtt->vm.closed);
+	GEM_BUG_ON(!atomic_read(&ppgtt->vm.open));
 
 	err = func(dev_priv, &ppgtt->vm, 0, ppgtt->vm.total, end_time);
 
 	i915_vm_put(&ppgtt->vm);
-out_unlock:
-	mutex_unlock(&dev_priv->drm.struct_mutex);
 
+out_free:
 	mock_file_free(dev_priv, file);
 	return err;
 }
@@ -1085,7 +1088,6 @@ static int exercise_ggtt(struct drm_i915_private *i915,
 	IGT_TIMEOUT(end_time);
 	int err = 0;
 
-	mutex_lock(&i915->drm.struct_mutex);
 restart:
 	list_sort(NULL, &ggtt->vm.mm.hole_stack, sort_holes);
 	drm_mm_for_each_hole(node, &ggtt->vm.mm, hole_start, hole_end) {
@@ -1106,7 +1108,6 @@ restart:
 		last = hole_end;
 		goto restart;
 	}
-	mutex_unlock(&i915->drm.struct_mutex);
 
 	return err;
 }
@@ -1148,13 +1149,12 @@ static int igt_ggtt_page(void *arg)
 	unsigned int *order, n;
 	int err;
 
-	mutex_lock(&i915->drm.struct_mutex);
+	if (!i915_ggtt_has_aperture(ggtt))
+		return 0;
 
 	obj = i915_gem_object_create_internal(i915, PAGE_SIZE);
-	if (IS_ERR(obj)) {
-		err = PTR_ERR(obj);
-		goto out_unlock;
-	}
+	if (IS_ERR(obj))
+		return PTR_ERR(obj);
 
 	err = i915_gem_object_pin_pages(obj);
 	if (err)
@@ -1222,8 +1222,6 @@ out_unpin:
 	i915_gem_object_unpin_pages(obj);
 out_free:
 	i915_gem_object_put(obj);
-out_unlock:
-	mutex_unlock(&i915->drm.struct_mutex);
 	return err;
 }
 
@@ -1234,10 +1232,13 @@ static void track_vma_bind(struct i915_vma *vma)
 	atomic_inc(&obj->bind_count); /* track for eviction later */
 	__i915_gem_object_pin_pages(obj);
 
+	GEM_BUG_ON(vma->pages);
+	atomic_set(&vma->pages_count, I915_VMA_PAGES_ACTIVE);
+	__i915_gem_object_pin_pages(obj);
 	vma->pages = obj->mm.pages;
 
 	mutex_lock(&vma->vm->mutex);
-	list_move_tail(&vma->vm_link, &vma->vm->bound_list);
+	list_add_tail(&vma->vm_link, &vma->vm->bound_list);
 	mutex_unlock(&vma->vm->mutex);
 }
 
@@ -1248,6 +1249,7 @@ static int exercise_mock(struct drm_i915_private *i915,
 				     unsigned long end_time))
 {
 	const u64 limit = totalram_pages() << PAGE_SHIFT;
+	struct i915_address_space *vm;
 	struct i915_gem_context *ctx;
 	IGT_TIMEOUT(end_time);
 	int err;
@@ -1256,7 +1258,9 @@ static int exercise_mock(struct drm_i915_private *i915,
 	if (!ctx)
 		return -ENOMEM;
 
-	err = func(i915, ctx->vm, 0, min(ctx->vm->total, limit), end_time);
+	vm = i915_gem_context_get_vm_rcu(ctx);
+	err = func(i915, vm, 0, min(vm->total, limit), end_time);
+	i915_vm_put(vm);
 
 	mock_context_close(ctx);
 	return err;
@@ -1294,6 +1298,7 @@ static int igt_gtt_reserve(void *arg)
 {
 	struct i915_ggtt *ggtt = arg;
 	struct drm_i915_gem_object *obj, *on;
+	I915_RND_STATE(prng);
 	LIST_HEAD(objects);
 	u64 total;
 	int err = -ENODEV;
@@ -1330,11 +1335,13 @@ static int igt_gtt_reserve(void *arg)
 			goto out;
 		}
 
+		mutex_lock(&ggtt->vm.mutex);
 		err = i915_gem_gtt_reserve(&ggtt->vm, &vma->node,
 					   obj->base.size,
 					   total,
 					   obj->cache_level,
 					   0);
+		mutex_unlock(&ggtt->vm.mutex);
 		if (err) {
 			pr_err("i915_gem_gtt_reserve (pass 1) failed at %llu/%llu with err=%d\n",
 			       total, ggtt->vm.total, err);
@@ -1380,11 +1387,13 @@ static int igt_gtt_reserve(void *arg)
 			goto out;
 		}
 
+		mutex_lock(&ggtt->vm.mutex);
 		err = i915_gem_gtt_reserve(&ggtt->vm, &vma->node,
 					   obj->base.size,
 					   total,
 					   obj->cache_level,
 					   0);
+		mutex_unlock(&ggtt->vm.mutex);
 		if (err) {
 			pr_err("i915_gem_gtt_reserve (pass 2) failed at %llu/%llu with err=%d\n",
 			       total, ggtt->vm.total, err);
@@ -1420,15 +1429,18 @@ static int igt_gtt_reserve(void *arg)
 			goto out;
 		}
 
-		offset = random_offset(0, ggtt->vm.total,
-				       2*I915_GTT_PAGE_SIZE,
-				       I915_GTT_MIN_ALIGNMENT);
+		offset = igt_random_offset(&prng,
+					   0, ggtt->vm.total,
+					   2 * I915_GTT_PAGE_SIZE,
+					   I915_GTT_MIN_ALIGNMENT);
 
+		mutex_lock(&ggtt->vm.mutex);
 		err = i915_gem_gtt_reserve(&ggtt->vm, &vma->node,
 					   obj->base.size,
 					   offset,
 					   obj->cache_level,
 					   0);
+		mutex_unlock(&ggtt->vm.mutex);
 		if (err) {
 			pr_err("i915_gem_gtt_reserve (pass 3) failed at %llu/%llu with err=%d\n",
 			       total, ggtt->vm.total, err);
@@ -1497,11 +1509,13 @@ static int igt_gtt_insert(void *arg)
 
 	/* Check a couple of obviously invalid requests */
 	for (ii = invalid_insert; ii->size; ii++) {
+		mutex_lock(&ggtt->vm.mutex);
 		err = i915_gem_gtt_insert(&ggtt->vm, &tmp,
 					  ii->size, ii->alignment,
 					  I915_COLOR_UNEVICTABLE,
 					  ii->start, ii->end,
 					  0);
+		mutex_unlock(&ggtt->vm.mutex);
 		if (err != -ENOSPC) {
 			pr_err("Invalid i915_gem_gtt_insert(.size=%llx, .alignment=%llx, .start=%llx, .end=%llx) succeeded (err=%d)\n",
 			       ii->size, ii->alignment, ii->start, ii->end,
@@ -1537,10 +1551,12 @@ static int igt_gtt_insert(void *arg)
 			goto out;
 		}
 
+		mutex_lock(&ggtt->vm.mutex);
 		err = i915_gem_gtt_insert(&ggtt->vm, &vma->node,
 					  obj->base.size, 0, obj->cache_level,
 					  0, ggtt->vm.total,
 					  0);
+		mutex_unlock(&ggtt->vm.mutex);
 		if (err == -ENOSPC) {
 			/* maxed out the GGTT space */
 			i915_gem_object_put(obj);
@@ -1595,10 +1611,12 @@ static int igt_gtt_insert(void *arg)
 			goto out;
 		}
 
+		mutex_lock(&ggtt->vm.mutex);
 		err = i915_gem_gtt_insert(&ggtt->vm, &vma->node,
 					  obj->base.size, 0, obj->cache_level,
 					  0, ggtt->vm.total,
 					  0);
+		mutex_unlock(&ggtt->vm.mutex);
 		if (err) {
 			pr_err("i915_gem_gtt_insert (pass 2) failed at %llu/%llu with err=%d\n",
 			       total, ggtt->vm.total, err);
@@ -1642,10 +1660,12 @@ static int igt_gtt_insert(void *arg)
 			goto out;
 		}
 
+		mutex_lock(&ggtt->vm.mutex);
 		err = i915_gem_gtt_insert(&ggtt->vm, &vma->node,
 					  obj->base.size, 0, obj->cache_level,
 					  0, ggtt->vm.total,
 					  0);
+		mutex_unlock(&ggtt->vm.mutex);
 		if (err) {
 			pr_err("i915_gem_gtt_insert (pass 3) failed at %llu/%llu with err=%d\n",
 			       total, ggtt->vm.total, err);
@@ -1689,13 +1709,10 @@ int i915_gem_gtt_mock_selftests(void)
 	}
 	mock_init_ggtt(i915, ggtt);
 
-	mutex_lock(&i915->drm.struct_mutex);
 	err = i915_subtests(tests, ggtt);
-	mock_device_flush(i915);
-	mutex_unlock(&i915->drm.struct_mutex);
 
+	mock_device_flush(i915);
 	i915_gem_drain_freed_objects(i915);
-
 	mock_fini_ggtt(ggtt);
 	kfree(ggtt);
 out_put:
@@ -1703,6 +1720,312 @@ out_put:
 	return err;
 }
 
+static int context_sync(struct intel_context *ce)
+{
+	struct i915_request *rq;
+	long timeout;
+
+	rq = intel_context_create_request(ce);
+	if (IS_ERR(rq))
+		return PTR_ERR(rq);
+
+	i915_request_get(rq);
+	i915_request_add(rq);
+
+	timeout = i915_request_wait(rq, 0, HZ / 5);
+	i915_request_put(rq);
+
+	return timeout < 0 ? -EIO : 0;
+}
+
+static struct i915_request *
+submit_batch(struct intel_context *ce, u64 addr)
+{
+	struct i915_request *rq;
+	int err;
+
+	rq = intel_context_create_request(ce);
+	if (IS_ERR(rq))
+		return rq;
+
+	err = 0;
+	if (rq->engine->emit_init_breadcrumb) /* detect a hang */
+		err = rq->engine->emit_init_breadcrumb(rq);
+	if (err == 0)
+		err = rq->engine->emit_bb_start(rq, addr, 0, 0);
+
+	if (err == 0)
+		i915_request_get(rq);
+	i915_request_add(rq);
+
+	return err ? ERR_PTR(err) : rq;
+}
+
+static u32 *spinner(u32 *batch, int i)
+{
+	return batch + i * 64 / sizeof(*batch) + 4;
+}
+
+static void end_spin(u32 *batch, int i)
+{
+	*spinner(batch, i) = MI_BATCH_BUFFER_END;
+	wmb();
+}
+
+static int igt_cs_tlb(void *arg)
+{
+	const unsigned int count = PAGE_SIZE / 64;
+	const unsigned int chunk_size = count * PAGE_SIZE;
+	struct drm_i915_private *i915 = arg;
+	struct drm_i915_gem_object *bbe, *act, *out;
+	struct i915_gem_engines_iter it;
+	struct i915_address_space *vm;
+	struct i915_gem_context *ctx;
+	struct intel_context *ce;
+	struct drm_file *file;
+	struct i915_vma *vma;
+	I915_RND_STATE(prng);
+	unsigned int i;
+	u32 *result;
+	u32 *batch;
+	int err = 0;
+
+	/*
+	 * Our mission here is to fool the hardware to execute something
+	 * from scratch as it has not seen the batch move (due to missing
+	 * the TLB invalidate).
+	 */
+
+	file = mock_file(i915);
+	if (IS_ERR(file))
+		return PTR_ERR(file);
+
+	ctx = live_context(i915, file);
+	if (IS_ERR(ctx)) {
+		err = PTR_ERR(ctx);
+		goto out_unlock;
+	}
+
+	vm = i915_gem_context_get_vm_rcu(ctx);
+	if (i915_is_ggtt(vm))
+		goto out_vm;
+
+	/* Create two pages; dummy we prefill the TLB, and intended */
+	bbe = i915_gem_object_create_internal(i915, PAGE_SIZE);
+	if (IS_ERR(bbe)) {
+		err = PTR_ERR(bbe);
+		goto out_vm;
+	}
+
+	batch = i915_gem_object_pin_map(bbe, I915_MAP_WC);
+	if (IS_ERR(batch)) {
+		err = PTR_ERR(batch);
+		goto out_put_bbe;
+	}
+	memset32(batch, MI_BATCH_BUFFER_END, PAGE_SIZE / sizeof(u32));
+	i915_gem_object_flush_map(bbe);
+	i915_gem_object_unpin_map(bbe);
+
+	act = i915_gem_object_create_internal(i915, PAGE_SIZE);
+	if (IS_ERR(act)) {
+		err = PTR_ERR(act);
+		goto out_put_bbe;
+	}
+
+	/* Track the execution of each request by writing into different slot */
+	batch = i915_gem_object_pin_map(act, I915_MAP_WC);
+	if (IS_ERR(batch)) {
+		err = PTR_ERR(batch);
+		goto out_put_act;
+	}
+	for (i = 0; i < count; i++) {
+		u32 *cs = batch + i * 64 / sizeof(*cs);
+		u64 addr = (vm->total - PAGE_SIZE) + i * sizeof(u32);
+
+		GEM_BUG_ON(INTEL_GEN(i915) < 6);
+		cs[0] = MI_STORE_DWORD_IMM_GEN4;
+		if (INTEL_GEN(i915) >= 8) {
+			cs[1] = lower_32_bits(addr);
+			cs[2] = upper_32_bits(addr);
+			cs[3] = i;
+			cs[4] = MI_NOOP;
+			cs[5] = MI_BATCH_BUFFER_START_GEN8;
+		} else {
+			cs[1] = 0;
+			cs[2] = lower_32_bits(addr);
+			cs[3] = i;
+			cs[4] = MI_NOOP;
+			cs[5] = MI_BATCH_BUFFER_START;
+		}
+	}
+
+	out = i915_gem_object_create_internal(i915, PAGE_SIZE);
+	if (IS_ERR(out)) {
+		err = PTR_ERR(out);
+		goto out_put_batch;
+	}
+	i915_gem_object_set_cache_coherency(out, I915_CACHING_CACHED);
+
+	vma = i915_vma_instance(out, vm, NULL);
+	if (IS_ERR(vma)) {
+		err = PTR_ERR(vma);
+		goto out_put_batch;
+	}
+
+	err = i915_vma_pin(vma, 0, 0,
+			   PIN_USER |
+			   PIN_OFFSET_FIXED |
+			   (vm->total - PAGE_SIZE));
+	if (err)
+		goto out_put_out;
+	GEM_BUG_ON(vma->node.start != vm->total - PAGE_SIZE);
+
+	result = i915_gem_object_pin_map(out, I915_MAP_WB);
+	if (IS_ERR(result)) {
+		err = PTR_ERR(result);
+		goto out_put_out;
+	}
+
+	for_each_gem_engine(ce, i915_gem_context_lock_engines(ctx), it) {
+		IGT_TIMEOUT(end_time);
+		unsigned long pass = 0;
+
+		if (!intel_engine_can_store_dword(ce->engine))
+			continue;
+
+		while (!__igt_timeout(end_time, NULL)) {
+			struct i915_request *rq;
+			u64 offset;
+
+			offset = igt_random_offset(&prng,
+						   0, vm->total - PAGE_SIZE,
+						   chunk_size, PAGE_SIZE);
+
+			err = vm->allocate_va_range(vm, offset, chunk_size);
+			if (err)
+				goto end;
+
+			memset32(result, STACK_MAGIC, PAGE_SIZE / sizeof(u32));
+
+			vma = i915_vma_instance(bbe, vm, NULL);
+			if (IS_ERR(vma)) {
+				err = PTR_ERR(vma);
+				goto end;
+			}
+
+			err = vma->ops->set_pages(vma);
+			if (err)
+				goto end;
+
+			/* Prime the TLB with the dummy pages */
+			for (i = 0; i < count; i++) {
+				vma->node.start = offset + i * PAGE_SIZE;
+				vm->insert_entries(vm, vma, I915_CACHE_NONE, 0);
+
+				rq = submit_batch(ce, vma->node.start);
+				if (IS_ERR(rq)) {
+					err = PTR_ERR(rq);
+					goto end;
+				}
+				i915_request_put(rq);
+			}
+
+			vma->ops->clear_pages(vma);
+
+			err = context_sync(ce);
+			if (err) {
+				pr_err("%s: dummy setup timed out\n",
+				       ce->engine->name);
+				goto end;
+			}
+
+			vma = i915_vma_instance(act, vm, NULL);
+			if (IS_ERR(vma)) {
+				err = PTR_ERR(vma);
+				goto end;
+			}
+
+			err = vma->ops->set_pages(vma);
+			if (err)
+				goto end;
+
+			/* Replace the TLB with target batches */
+			for (i = 0; i < count; i++) {
+				struct i915_request *rq;
+				u32 *cs = batch + i * 64 / sizeof(*cs);
+				u64 addr;
+
+				vma->node.start = offset + i * PAGE_SIZE;
+				vm->insert_entries(vm, vma, I915_CACHE_NONE, 0);
+
+				addr = vma->node.start + i * 64;
+				cs[4] = MI_NOOP;
+				cs[6] = lower_32_bits(addr);
+				cs[7] = upper_32_bits(addr);
+				wmb();
+
+				rq = submit_batch(ce, addr);
+				if (IS_ERR(rq)) {
+					err = PTR_ERR(rq);
+					goto end;
+				}
+
+				/* Wait until the context chain has started */
+				if (i == 0) {
+					while (READ_ONCE(result[i]) &&
+					       !i915_request_completed(rq))
+						cond_resched();
+				} else {
+					end_spin(batch, i - 1);
+				}
+
+				i915_request_put(rq);
+			}
+			end_spin(batch, count - 1);
+
+			vma->ops->clear_pages(vma);
+
+			err = context_sync(ce);
+			if (err) {
+				pr_err("%s: writes timed out\n",
+				       ce->engine->name);
+				goto end;
+			}
+
+			for (i = 0; i < count; i++) {
+				if (result[i] != i) {
+					pr_err("%s: Write lost on pass %lu, at offset %llx, index %d, found %x, expected %x\n",
+					       ce->engine->name, pass,
+					       offset, i, result[i], i);
+					err = -EINVAL;
+					goto end;
+				}
+			}
+
+			vm->clear_range(vm, offset, chunk_size);
+			pass++;
+		}
+	}
+end:
+	if (igt_flush_test(i915))
+		err = -EIO;
+	i915_gem_context_unlock_engines(ctx);
+	i915_gem_object_unpin_map(out);
+out_put_out:
+	i915_gem_object_put(out);
+out_put_batch:
+	i915_gem_object_unpin_map(act);
+out_put_act:
+	i915_gem_object_put(act);
+out_put_bbe:
+	i915_gem_object_put(bbe);
+out_vm:
+	i915_vm_put(vm);
+out_unlock:
+	mock_file_free(i915, file);
+	return err;
+}
+
 int i915_gem_gtt_live_selftests(struct drm_i915_private *i915)
 {
 	static const struct i915_subtest tests[] = {
@@ -1720,6 +2043,7 @@ int i915_gem_gtt_live_selftests(struct drm_i915_private *i915)
 		SUBTEST(igt_ggtt_pot),
 		SUBTEST(igt_ggtt_fill),
 		SUBTEST(igt_ggtt_page),
+		SUBTEST(igt_cs_tlb),
 	};
 
 	GEM_BUG_ON(offset_in_page(i915->ggtt.vm.total));
diff --git a/drivers/gpu/drm/i915/selftests/i915_live_selftests.h b/drivers/gpu/drm/i915/selftests/i915_live_selftests.h
index 1ccf0f731ac0..4b3cac73e291 100644
--- a/drivers/gpu/drm/i915/selftests/i915_live_selftests.h
+++ b/drivers/gpu/drm/i915/selftests/i915_live_selftests.h
@@ -15,6 +15,9 @@ selftest(workarounds, intel_workarounds_live_selftests)
 selftest(gt_engines, intel_engine_live_selftests)
 selftest(gt_timelines, intel_timeline_live_selftests)
 selftest(gt_contexts, intel_context_live_selftests)
+selftest(gt_lrc, intel_lrc_live_selftests)
+selftest(gt_pm, intel_gt_pm_live_selftests)
+selftest(gt_heartbeat, intel_heartbeat_live_selftests)
 selftest(requests, i915_request_live_selftests)
 selftest(active, i915_active_live_selftests)
 selftest(objects, i915_gem_object_live_selftests)
@@ -30,6 +33,8 @@ selftest(gem_contexts, i915_gem_context_live_selftests)
 selftest(blt, i915_gem_object_blt_live_selftests)
 selftest(client, i915_gem_client_blt_live_selftests)
 selftest(reset, intel_reset_live_selftests)
+selftest(memory_region, intel_memory_region_live_selftests)
 selftest(hangcheck, intel_hangcheck_live_selftests)
 selftest(execlists, intel_execlists_live_selftests)
 selftest(guc, intel_guc_live_selftest)
+selftest(perf, i915_perf_live_selftests)
diff --git a/drivers/gpu/drm/i915/selftests/i915_mock_selftests.h b/drivers/gpu/drm/i915/selftests/i915_mock_selftests.h
index b88084fe3269..aa5a0e7f5d9e 100644
--- a/drivers/gpu/drm/i915/selftests/i915_mock_selftests.h
+++ b/drivers/gpu/drm/i915/selftests/i915_mock_selftests.h
@@ -26,3 +26,4 @@ selftest(gtt, i915_gem_gtt_mock_selftests)
 selftest(hugepages, i915_gem_huge_page_mock_selftests)
 selftest(contexts, i915_gem_context_mock_selftests)
 selftest(buddy, i915_buddy_mock_selftests)
+selftest(memory_region, intel_memory_region_mock_selftests)
diff --git a/drivers/gpu/drm/i915/selftests/i915_perf.c b/drivers/gpu/drm/i915/selftests/i915_perf.c
new file mode 100644
index 000000000000..aabd07f67e49
--- /dev/null
+++ b/drivers/gpu/drm/i915/selftests/i915_perf.c
@@ -0,0 +1,217 @@
+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2019 Intel Corporation
+ */
+
+#include <linux/kref.h>
+
+#include "gem/i915_gem_pm.h"
+#include "gt/intel_gt.h"
+
+#include "i915_selftest.h"
+
+#include "igt_flush_test.h"
+#include "lib_sw_fence.h"
+
+static struct i915_perf_stream *
+test_stream(struct i915_perf *perf)
+{
+	struct drm_i915_perf_open_param param = {};
+	struct perf_open_properties props = {
+		.engine = intel_engine_lookup_user(perf->i915,
+						   I915_ENGINE_CLASS_RENDER,
+						   0),
+		.sample_flags = SAMPLE_OA_REPORT,
+		.oa_format = IS_GEN(perf->i915, 12) ?
+		I915_OA_FORMAT_A32u40_A4u32_B8_C8 : I915_OA_FORMAT_C4_B8,
+		.metrics_set = 1,
+	};
+	struct i915_perf_stream *stream;
+
+	stream = kzalloc(sizeof(*stream), GFP_KERNEL);
+	if (!stream)
+		return NULL;
+
+	stream->perf = perf;
+
+	mutex_lock(&perf->lock);
+	if (i915_oa_stream_init(stream, &param, &props)) {
+		kfree(stream);
+		stream =  NULL;
+	}
+	mutex_unlock(&perf->lock);
+
+	return stream;
+}
+
+static void stream_destroy(struct i915_perf_stream *stream)
+{
+	struct i915_perf *perf = stream->perf;
+
+	mutex_lock(&perf->lock);
+	i915_perf_destroy_locked(stream);
+	mutex_unlock(&perf->lock);
+}
+
+static int live_sanitycheck(void *arg)
+{
+	struct drm_i915_private *i915 = arg;
+	struct i915_perf_stream *stream;
+
+	/* Quick check we can create a perf stream */
+
+	stream = test_stream(&i915->perf);
+	if (!stream)
+		return -EINVAL;
+
+	stream_destroy(stream);
+	return 0;
+}
+
+static int write_timestamp(struct i915_request *rq, int slot)
+{
+	u32 *cs;
+	int len;
+
+	cs = intel_ring_begin(rq, 6);
+	if (IS_ERR(cs))
+		return PTR_ERR(cs);
+
+	len = 5;
+	if (INTEL_GEN(rq->i915) >= 8)
+		len++;
+
+	*cs++ = GFX_OP_PIPE_CONTROL(len);
+	*cs++ = PIPE_CONTROL_GLOBAL_GTT_IVB |
+		PIPE_CONTROL_STORE_DATA_INDEX |
+		PIPE_CONTROL_WRITE_TIMESTAMP;
+	*cs++ = slot * sizeof(u32);
+	*cs++ = 0;
+	*cs++ = 0;
+	*cs++ = 0;
+
+	intel_ring_advance(rq, cs);
+
+	return 0;
+}
+
+static ktime_t poll_status(struct i915_request *rq, int slot)
+{
+	while (!intel_read_status_page(rq->engine, slot) &&
+	       !i915_request_completed(rq))
+		cpu_relax();
+
+	return ktime_get();
+}
+
+static int live_noa_delay(void *arg)
+{
+	struct drm_i915_private *i915 = arg;
+	struct i915_perf_stream *stream;
+	struct i915_request *rq;
+	ktime_t t0, t1;
+	u64 expected;
+	u32 delay;
+	int err;
+	int i;
+
+	/* Check that the GPU delays matches expectations */
+
+	stream = test_stream(&i915->perf);
+	if (!stream)
+		return -ENOMEM;
+
+	expected = atomic64_read(&stream->perf->noa_programming_delay);
+
+	if (stream->engine->class != RENDER_CLASS) {
+		err = -ENODEV;
+		goto out;
+	}
+
+	for (i = 0; i < 4; i++)
+		intel_write_status_page(stream->engine, 0x100 + i, 0);
+
+	rq = i915_request_create(stream->engine->kernel_context);
+	if (IS_ERR(rq)) {
+		err = PTR_ERR(rq);
+		goto out;
+	}
+
+	if (rq->engine->emit_init_breadcrumb &&
+	    i915_request_timeline(rq)->has_initial_breadcrumb) {
+		err = rq->engine->emit_init_breadcrumb(rq);
+		if (err) {
+			i915_request_add(rq);
+			goto out;
+		}
+	}
+
+	err = write_timestamp(rq, 0x100);
+	if (err) {
+		i915_request_add(rq);
+		goto out;
+	}
+
+	err = rq->engine->emit_bb_start(rq,
+					i915_ggtt_offset(stream->noa_wait), 0,
+					I915_DISPATCH_SECURE);
+	if (err) {
+		i915_request_add(rq);
+		goto out;
+	}
+
+	err = write_timestamp(rq, 0x102);
+	if (err) {
+		i915_request_add(rq);
+		goto out;
+	}
+
+	i915_request_get(rq);
+	i915_request_add(rq);
+
+	preempt_disable();
+	t0 = poll_status(rq, 0x100);
+	t1 = poll_status(rq, 0x102);
+	preempt_enable();
+
+	pr_info("CPU delay: %lluns, expected %lluns\n",
+		ktime_sub(t1, t0), expected);
+
+	delay = intel_read_status_page(stream->engine, 0x102);
+	delay -= intel_read_status_page(stream->engine, 0x100);
+	delay = div_u64(mul_u32_u32(delay, 1000 * 1000),
+			RUNTIME_INFO(i915)->cs_timestamp_frequency_khz);
+	pr_info("GPU delay: %uns, expected %lluns\n",
+		delay, expected);
+
+	if (4 * delay < 3 * expected || 2 * delay > 3 * expected) {
+		pr_err("GPU delay [%uus] outside of expected threshold! [%lluus, %lluus]\n",
+		       delay / 1000,
+		       div_u64(3 * expected, 4000),
+		       div_u64(3 * expected, 2000));
+		err = -EINVAL;
+	}
+
+	i915_request_put(rq);
+out:
+	stream_destroy(stream);
+	return err;
+}
+
+int i915_perf_live_selftests(struct drm_i915_private *i915)
+{
+	static const struct i915_subtest tests[] = {
+		SUBTEST(live_sanitycheck),
+		SUBTEST(live_noa_delay),
+	};
+	struct i915_perf *perf = &i915->perf;
+
+	if (!perf->metrics_kobj || !perf->ops.enable_metric_set)
+		return 0;
+
+	if (intel_gt_is_wedged(&i915->gt))
+		return 0;
+
+	return i915_subtests(tests, i915);
+}
diff --git a/drivers/gpu/drm/i915/selftests/i915_random.c b/drivers/gpu/drm/i915/selftests/i915_random.c
index 716a3f19f030..abdfadcf626b 100644
--- a/drivers/gpu/drm/i915/selftests/i915_random.c
+++ b/drivers/gpu/drm/i915/selftests/i915_random.c
@@ -29,6 +29,7 @@
 #include <linux/types.h>
 
 #include "i915_random.h"
+#include "i915_utils.h"
 
 u64 i915_prandom_u64_state(struct rnd_state *rnd)
 {
@@ -87,3 +88,22 @@ unsigned int *i915_random_order(unsigned int count, struct rnd_state *state)
 	i915_random_reorder(order, count, state);
 	return order;
 }
+
+u64 igt_random_offset(struct rnd_state *state,
+		      u64 start, u64 end,
+		      u64 len, u64 align)
+{
+	u64 range, addr;
+
+	BUG_ON(range_overflows(start, len, end));
+	BUG_ON(round_up(start, align) > round_down(end - len, align));
+
+	range = round_down(end - len, align) - round_up(start, align);
+	if (range) {
+		addr = i915_prandom_u64_state(state);
+		div64_u64_rem(addr, range, &addr);
+		start += addr;
+	}
+
+	return round_up(start, align);
+}
diff --git a/drivers/gpu/drm/i915/selftests/i915_random.h b/drivers/gpu/drm/i915/selftests/i915_random.h
index 8e1ff9c105b6..35cc69a3a1b9 100644
--- a/drivers/gpu/drm/i915/selftests/i915_random.h
+++ b/drivers/gpu/drm/i915/selftests/i915_random.h
@@ -57,4 +57,8 @@ void i915_random_reorder(unsigned int *order,
 void i915_prandom_shuffle(void *arr, size_t elsz, size_t count,
 			  struct rnd_state *state);
 
+u64 igt_random_offset(struct rnd_state *state,
+		      u64 start, u64 end,
+		      u64 len, u64 align);
+
 #endif /* !__I915_SELFTESTS_RANDOM_H__ */
diff --git a/drivers/gpu/drm/i915/selftests/i915_request.c b/drivers/gpu/drm/i915/selftests/i915_request.c
index b3688543ed7d..8618a4dc0701 100644
--- a/drivers/gpu/drm/i915/selftests/i915_request.c
+++ b/drivers/gpu/drm/i915/selftests/i915_request.c
@@ -37,25 +37,32 @@
 #include "mock_drm.h"
 #include "mock_gem_device.h"
 
+static unsigned int num_uabi_engines(struct drm_i915_private *i915)
+{
+	struct intel_engine_cs *engine;
+	unsigned int count;
+
+	count = 0;
+	for_each_uabi_engine(engine, i915)
+		count++;
+
+	return count;
+}
+
 static int igt_add_request(void *arg)
 {
 	struct drm_i915_private *i915 = arg;
 	struct i915_request *request;
-	int err = -ENOMEM;
 
 	/* Basic preliminary test to create a request and let it loose! */
 
-	mutex_lock(&i915->drm.struct_mutex);
 	request = mock_request(i915->engine[RCS0]->kernel_context, HZ / 10);
 	if (!request)
-		goto out_unlock;
+		return -ENOMEM;
 
 	i915_request_add(request);
 
-	err = 0;
-out_unlock:
-	mutex_unlock(&i915->drm.struct_mutex);
-	return err;
+	return 0;
 }
 
 static int igt_wait_request(void *arg)
@@ -67,12 +74,10 @@ static int igt_wait_request(void *arg)
 
 	/* Submit a request, then wait upon it */
 
-	mutex_lock(&i915->drm.struct_mutex);
 	request = mock_request(i915->engine[RCS0]->kernel_context, T);
-	if (!request) {
-		err = -ENOMEM;
-		goto out_unlock;
-	}
+	if (!request)
+		return -ENOMEM;
+
 	i915_request_get(request);
 
 	if (i915_request_wait(request, 0, 0) != -ETIME) {
@@ -125,9 +130,7 @@ static int igt_wait_request(void *arg)
 	err = 0;
 out_request:
 	i915_request_put(request);
-out_unlock:
 	mock_device_flush(i915);
-	mutex_unlock(&i915->drm.struct_mutex);
 	return err;
 }
 
@@ -140,52 +143,45 @@ static int igt_fence_wait(void *arg)
 
 	/* Submit a request, treat it as a fence and wait upon it */
 
-	mutex_lock(&i915->drm.struct_mutex);
 	request = mock_request(i915->engine[RCS0]->kernel_context, T);
-	if (!request) {
-		err = -ENOMEM;
-		goto out_locked;
-	}
+	if (!request)
+		return -ENOMEM;
 
 	if (dma_fence_wait_timeout(&request->fence, false, T) != -ETIME) {
 		pr_err("fence wait success before submit (expected timeout)!\n");
-		goto out_locked;
+		goto out;
 	}
 
 	i915_request_add(request);
-	mutex_unlock(&i915->drm.struct_mutex);
 
 	if (dma_fence_is_signaled(&request->fence)) {
 		pr_err("fence signaled immediately!\n");
-		goto out_device;
+		goto out;
 	}
 
 	if (dma_fence_wait_timeout(&request->fence, false, T / 2) != -ETIME) {
 		pr_err("fence wait success after submit (expected timeout)!\n");
-		goto out_device;
+		goto out;
 	}
 
 	if (dma_fence_wait_timeout(&request->fence, false, T) <= 0) {
 		pr_err("fence wait timed out (expected success)!\n");
-		goto out_device;
+		goto out;
 	}
 
 	if (!dma_fence_is_signaled(&request->fence)) {
 		pr_err("fence unsignaled after waiting!\n");
-		goto out_device;
+		goto out;
 	}
 
 	if (dma_fence_wait_timeout(&request->fence, false, T) <= 0) {
 		pr_err("fence wait timed out when complete (expected success)!\n");
-		goto out_device;
+		goto out;
 	}
 
 	err = 0;
-out_device:
-	mutex_lock(&i915->drm.struct_mutex);
-out_locked:
+out:
 	mock_device_flush(i915);
-	mutex_unlock(&i915->drm.struct_mutex);
 	return err;
 }
 
@@ -197,8 +193,8 @@ static int igt_request_rewind(void *arg)
 	struct intel_context *ce;
 	int err = -EINVAL;
 
-	mutex_lock(&i915->drm.struct_mutex);
 	ctx[0] = mock_context(i915, "A");
+
 	ce = i915_gem_context_get_engine(ctx[0], RCS0);
 	GEM_BUG_ON(IS_ERR(ce));
 	request = mock_request(ce, 2 * HZ);
@@ -212,6 +208,7 @@ static int igt_request_rewind(void *arg)
 	i915_request_add(request);
 
 	ctx[1] = mock_context(i915, "B");
+
 	ce = i915_gem_context_get_engine(ctx[1], RCS0);
 	GEM_BUG_ON(IS_ERR(ce));
 	vip = mock_request(ce, 0);
@@ -233,7 +230,6 @@ static int igt_request_rewind(void *arg)
 	request->engine->submit_request(request);
 	rcu_read_unlock();
 
-	mutex_unlock(&i915->drm.struct_mutex);
 
 	if (i915_request_wait(vip, 0, HZ) == -ETIME) {
 		pr_err("timed out waiting for high priority request\n");
@@ -248,14 +244,12 @@ static int igt_request_rewind(void *arg)
 	err = 0;
 err:
 	i915_request_put(vip);
-	mutex_lock(&i915->drm.struct_mutex);
 err_context_1:
 	mock_context_close(ctx[1]);
 	i915_request_put(request);
 err_context_0:
 	mock_context_close(ctx[0]);
 	mock_device_flush(i915);
-	mutex_unlock(&i915->drm.struct_mutex);
 	return err;
 }
 
@@ -282,7 +276,6 @@ __live_request_alloc(struct intel_context *ce)
 static int __igt_breadcrumbs_smoketest(void *arg)
 {
 	struct smoketest *t = arg;
-	struct mutex * const BKL = &t->engine->i915->drm.struct_mutex;
 	const unsigned int max_batch = min(t->ncontexts, t->max_batch) - 1;
 	const unsigned int total = 4 * t->ncontexts + 1;
 	unsigned int num_waits = 0, num_fences = 0;
@@ -300,7 +293,7 @@ static int __igt_breadcrumbs_smoketest(void *arg)
 	 * that the fences were marked as signaled.
 	 */
 
-	requests = kmalloc_array(total, sizeof(*requests), GFP_KERNEL);
+	requests = kcalloc(total, sizeof(*requests), GFP_KERNEL);
 	if (!requests)
 		return -ENOMEM;
 
@@ -337,14 +330,11 @@ static int __igt_breadcrumbs_smoketest(void *arg)
 			struct i915_request *rq;
 			struct intel_context *ce;
 
-			mutex_lock(BKL);
-
 			ce = i915_gem_context_get_engine(ctx, t->engine->legacy_idx);
 			GEM_BUG_ON(IS_ERR(ce));
 			rq = t->request_alloc(ce);
 			intel_context_put(ce);
 			if (IS_ERR(rq)) {
-				mutex_unlock(BKL);
 				err = PTR_ERR(rq);
 				count = n;
 				break;
@@ -357,8 +347,6 @@ static int __igt_breadcrumbs_smoketest(void *arg)
 			requests[n] = i915_request_get(rq);
 			i915_request_add(rq);
 
-			mutex_unlock(BKL);
-
 			if (err >= 0)
 				err = i915_sw_fence_await_dma_fence(wait,
 								    &rq->fence,
@@ -446,18 +434,16 @@ static int mock_breadcrumbs_smoketest(void *arg)
 	 * See __igt_breadcrumbs_smoketest();
 	 */
 
-	threads = kmalloc_array(ncpus, sizeof(*threads), GFP_KERNEL);
+	threads = kcalloc(ncpus, sizeof(*threads), GFP_KERNEL);
 	if (!threads)
 		return -ENOMEM;
 
-	t.contexts =
-		kmalloc_array(t.ncontexts, sizeof(*t.contexts), GFP_KERNEL);
+	t.contexts = kcalloc(t.ncontexts, sizeof(*t.contexts), GFP_KERNEL);
 	if (!t.contexts) {
 		ret = -ENOMEM;
 		goto out_threads;
 	}
 
-	mutex_lock(&t.engine->i915->drm.struct_mutex);
 	for (n = 0; n < t.ncontexts; n++) {
 		t.contexts[n] = mock_context(t.engine->i915, "mock");
 		if (!t.contexts[n]) {
@@ -465,7 +451,6 @@ static int mock_breadcrumbs_smoketest(void *arg)
 			goto out_contexts;
 		}
 	}
-	mutex_unlock(&t.engine->i915->drm.struct_mutex);
 
 	for (n = 0; n < ncpus; n++) {
 		threads[n] = kthread_run(__igt_breadcrumbs_smoketest,
@@ -479,6 +464,7 @@ static int mock_breadcrumbs_smoketest(void *arg)
 		get_task_struct(threads[n]);
 	}
 
+	yield(); /* start all threads before we begin */
 	msleep(jiffies_to_msecs(i915_selftest.timeout_jiffies));
 
 	for (n = 0; n < ncpus; n++) {
@@ -495,18 +481,15 @@ static int mock_breadcrumbs_smoketest(void *arg)
 		atomic_long_read(&t.num_fences),
 		ncpus);
 
-	mutex_lock(&t.engine->i915->drm.struct_mutex);
 out_contexts:
 	for (n = 0; n < t.ncontexts; n++) {
 		if (!t.contexts[n])
 			break;
 		mock_context_close(t.contexts[n]);
 	}
-	mutex_unlock(&t.engine->i915->drm.struct_mutex);
 	kfree(t.contexts);
 out_threads:
 	kfree(threads);
-
 	return ret;
 }
 
@@ -539,40 +522,37 @@ static int live_nop_request(void *arg)
 {
 	struct drm_i915_private *i915 = arg;
 	struct intel_engine_cs *engine;
-	intel_wakeref_t wakeref;
 	struct igt_live_test t;
-	unsigned int id;
 	int err = -ENODEV;
 
-	/* Submit various sized batches of empty requests, to each engine
+	/*
+	 * Submit various sized batches of empty requests, to each engine
 	 * (individually), and wait for the batch to complete. We can check
 	 * the overhead of submitting requests to the hardware.
 	 */
 
-	mutex_lock(&i915->drm.struct_mutex);
-	wakeref = intel_runtime_pm_get(&i915->runtime_pm);
-
-	for_each_engine(engine, i915, id) {
-		struct i915_request *request = NULL;
+	for_each_uabi_engine(engine, i915) {
 		unsigned long n, prime;
 		IGT_TIMEOUT(end_time);
 		ktime_t times[2] = {};
 
 		err = igt_live_test_begin(&t, i915, __func__, engine->name);
 		if (err)
-			goto out_unlock;
+			return err;
 
 		for_each_prime_number_from(prime, 1, 8192) {
+			struct i915_request *request = NULL;
+
 			times[1] = ktime_get_raw();
 
 			for (n = 0; n < prime; n++) {
+				i915_request_put(request);
 				request = i915_request_create(engine->kernel_context);
-				if (IS_ERR(request)) {
-					err = PTR_ERR(request);
-					goto out_unlock;
-				}
+				if (IS_ERR(request))
+					return PTR_ERR(request);
 
-				/* This space is left intentionally blank.
+				/*
+				 * This space is left intentionally blank.
 				 *
 				 * We do not actually want to perform any
 				 * action with this request, we just want
@@ -585,9 +565,11 @@ static int live_nop_request(void *arg)
 				 * for latency.
 				 */
 
+				i915_request_get(request);
 				i915_request_add(request);
 			}
 			i915_request_wait(request, 0, MAX_SCHEDULE_TIMEOUT);
+			i915_request_put(request);
 
 			times[1] = ktime_sub(ktime_get_raw(), times[1]);
 			if (prime == 1)
@@ -599,7 +581,7 @@ static int live_nop_request(void *arg)
 
 		err = igt_live_test_end(&t);
 		if (err)
-			goto out_unlock;
+			return err;
 
 		pr_info("Request latencies on %s: 1 = %lluns, %lu = %lluns\n",
 			engine->name,
@@ -607,9 +589,6 @@ static int live_nop_request(void *arg)
 			prime, div64_u64(ktime_to_ns(times[1]), prime));
 	}
 
-out_unlock:
-	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
-	mutex_unlock(&i915->drm.struct_mutex);
 	return err;
 }
 
@@ -647,8 +626,15 @@ static struct i915_vma *empty_batch(struct drm_i915_private *i915)
 	if (err)
 		goto err;
 
+	/* Force the wait wait now to avoid including it in the benchmark */
+	err = i915_vma_sync(vma);
+	if (err)
+		goto err_pin;
+
 	return vma;
 
+err_pin:
+	i915_vma_unpin(vma);
 err:
 	i915_gem_object_put(obj);
 	return ERR_PTR(err);
@@ -672,6 +658,7 @@ empty_request(struct intel_engine_cs *engine,
 	if (err)
 		goto out_request;
 
+	i915_request_get(request);
 out_request:
 	i915_request_add(request);
 	return err ? ERR_PTR(err) : request;
@@ -681,27 +668,21 @@ static int live_empty_request(void *arg)
 {
 	struct drm_i915_private *i915 = arg;
 	struct intel_engine_cs *engine;
-	intel_wakeref_t wakeref;
 	struct igt_live_test t;
 	struct i915_vma *batch;
-	unsigned int id;
 	int err = 0;
 
-	/* Submit various sized batches of empty requests, to each engine
+	/*
+	 * Submit various sized batches of empty requests, to each engine
 	 * (individually), and wait for the batch to complete. We can check
 	 * the overhead of submitting requests to the hardware.
 	 */
 
-	mutex_lock(&i915->drm.struct_mutex);
-	wakeref = intel_runtime_pm_get(&i915->runtime_pm);
-
 	batch = empty_batch(i915);
-	if (IS_ERR(batch)) {
-		err = PTR_ERR(batch);
-		goto out_unlock;
-	}
+	if (IS_ERR(batch))
+		return PTR_ERR(batch);
 
-	for_each_engine(engine, i915, id) {
+	for_each_uabi_engine(engine, i915) {
 		IGT_TIMEOUT(end_time);
 		struct i915_request *request;
 		unsigned long n, prime;
@@ -723,6 +704,7 @@ static int live_empty_request(void *arg)
 			times[1] = ktime_get_raw();
 
 			for (n = 0; n < prime; n++) {
+				i915_request_put(request);
 				request = empty_request(engine, batch);
 				if (IS_ERR(request)) {
 					err = PTR_ERR(request);
@@ -738,6 +720,7 @@ static int live_empty_request(void *arg)
 			if (__igt_timeout(end_time, NULL))
 				break;
 		}
+		i915_request_put(request);
 
 		err = igt_live_test_end(&t);
 		if (err)
@@ -752,18 +735,15 @@ static int live_empty_request(void *arg)
 out_batch:
 	i915_vma_unpin(batch);
 	i915_vma_put(batch);
-out_unlock:
-	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
-	mutex_unlock(&i915->drm.struct_mutex);
 	return err;
 }
 
 static struct i915_vma *recursive_batch(struct drm_i915_private *i915)
 {
 	struct i915_gem_context *ctx = i915->kernel_context;
-	struct i915_address_space *vm = ctx->vm ?: &i915->ggtt.vm;
 	struct drm_i915_gem_object *obj;
 	const int gen = INTEL_GEN(i915);
+	struct i915_address_space *vm;
 	struct i915_vma *vma;
 	u32 *cmd;
 	int err;
@@ -772,7 +752,9 @@ static struct i915_vma *recursive_batch(struct drm_i915_private *i915)
 	if (IS_ERR(obj))
 		return ERR_CAST(obj);
 
+	vm = i915_gem_context_get_vm_rcu(ctx);
 	vma = i915_vma_instance(obj, vm, NULL);
+	i915_vm_put(vm);
 	if (IS_ERR(vma)) {
 		err = PTR_ERR(vma);
 		goto err;
@@ -832,67 +814,73 @@ static int recursive_batch_resolve(struct i915_vma *batch)
 static int live_all_engines(void *arg)
 {
 	struct drm_i915_private *i915 = arg;
+	const unsigned int nengines = num_uabi_engines(i915);
 	struct intel_engine_cs *engine;
-	struct i915_request *request[I915_NUM_ENGINES];
-	intel_wakeref_t wakeref;
+	struct i915_request **request;
 	struct igt_live_test t;
 	struct i915_vma *batch;
-	unsigned int id;
+	unsigned int idx;
 	int err;
 
-	/* Check we can submit requests to all engines simultaneously. We
+	/*
+	 * Check we can submit requests to all engines simultaneously. We
 	 * send a recursive batch to each engine - checking that we don't
 	 * block doing so, and that they don't complete too soon.
 	 */
 
-	mutex_lock(&i915->drm.struct_mutex);
-	wakeref = intel_runtime_pm_get(&i915->runtime_pm);
+	request = kcalloc(nengines, sizeof(*request), GFP_KERNEL);
+	if (!request)
+		return -ENOMEM;
 
 	err = igt_live_test_begin(&t, i915, __func__, "");
 	if (err)
-		goto out_unlock;
+		goto out_free;
 
 	batch = recursive_batch(i915);
 	if (IS_ERR(batch)) {
 		err = PTR_ERR(batch);
 		pr_err("%s: Unable to create batch, err=%d\n", __func__, err);
-		goto out_unlock;
+		goto out_free;
 	}
 
-	for_each_engine(engine, i915, id) {
-		request[id] = i915_request_create(engine->kernel_context);
-		if (IS_ERR(request[id])) {
-			err = PTR_ERR(request[id]);
+	idx = 0;
+	for_each_uabi_engine(engine, i915) {
+		request[idx] = i915_request_create(engine->kernel_context);
+		if (IS_ERR(request[idx])) {
+			err = PTR_ERR(request[idx]);
 			pr_err("%s: Request allocation failed with err=%d\n",
 			       __func__, err);
 			goto out_request;
 		}
 
-		err = engine->emit_bb_start(request[id],
+		err = engine->emit_bb_start(request[idx],
 					    batch->node.start,
 					    batch->node.size,
 					    0);
 		GEM_BUG_ON(err);
-		request[id]->batch = batch;
+		request[idx]->batch = batch;
 
 		i915_vma_lock(batch);
-		err = i915_request_await_object(request[id], batch->obj, 0);
+		err = i915_request_await_object(request[idx], batch->obj, 0);
 		if (err == 0)
-			err = i915_vma_move_to_active(batch, request[id], 0);
+			err = i915_vma_move_to_active(batch, request[idx], 0);
 		i915_vma_unlock(batch);
 		GEM_BUG_ON(err);
 
-		i915_request_get(request[id]);
-		i915_request_add(request[id]);
+		i915_request_get(request[idx]);
+		i915_request_add(request[idx]);
+		idx++;
 	}
 
-	for_each_engine(engine, i915, id) {
-		if (i915_request_completed(request[id])) {
+	idx = 0;
+	for_each_uabi_engine(engine, i915) {
+		if (i915_request_completed(request[idx])) {
 			pr_err("%s(%s): request completed too early!\n",
 			       __func__, engine->name);
 			err = -EINVAL;
 			goto out_request;
 		}
+		idx++;
 	}
 
 	err = recursive_batch_resolve(batch);
@@ -901,10 +889,11 @@ static int live_all_engines(void *arg)
 		goto out_request;
 	}
 
-	for_each_engine(engine, i915, id) {
+	idx = 0;
+	for_each_uabi_engine(engine, i915) {
 		long timeout;
 
-		timeout = i915_request_wait(request[id], 0,
+		timeout = i915_request_wait(request[idx], 0,
 					    MAX_SCHEDULE_TIMEOUT);
 		if (timeout < 0) {
 			err = timeout;
@@ -913,50 +902,56 @@ static int live_all_engines(void *arg)
 			goto out_request;
 		}
 
-		GEM_BUG_ON(!i915_request_completed(request[id]));
-		i915_request_put(request[id]);
-		request[id] = NULL;
+		GEM_BUG_ON(!i915_request_completed(request[idx]));
+		i915_request_put(request[idx]);
+		request[idx] = NULL;
+		idx++;
 	}
 
 	err = igt_live_test_end(&t);
 
 out_request:
-	for_each_engine(engine, i915, id)
-		if (request[id])
-			i915_request_put(request[id]);
+	idx = 0;
+	for_each_uabi_engine(engine, i915) {
+		if (request[idx])
+			i915_request_put(request[idx]);
+		idx++;
+	}
 	i915_vma_unpin(batch);
 	i915_vma_put(batch);
-out_unlock:
-	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
-	mutex_unlock(&i915->drm.struct_mutex);
+out_free:
+	kfree(request);
 	return err;
 }
 
 static int live_sequential_engines(void *arg)
 {
 	struct drm_i915_private *i915 = arg;
-	struct i915_request *request[I915_NUM_ENGINES] = {};
+	const unsigned int nengines = num_uabi_engines(i915);
+	struct i915_request **request;
 	struct i915_request *prev = NULL;
 	struct intel_engine_cs *engine;
-	intel_wakeref_t wakeref;
 	struct igt_live_test t;
-	unsigned int id;
+	unsigned int idx;
 	int err;
 
-	/* Check we can submit requests to all engines sequentially, such
+	/*
+	 * Check we can submit requests to all engines sequentially, such
 	 * that each successive request waits for the earlier ones. This
 	 * tests that we don't execute requests out of order, even though
 	 * they are running on independent engines.
 	 */
 
-	mutex_lock(&i915->drm.struct_mutex);
-	wakeref = intel_runtime_pm_get(&i915->runtime_pm);
+	request = kcalloc(nengines, sizeof(*request), GFP_KERNEL);
+	if (!request)
+		return -ENOMEM;
 
 	err = igt_live_test_begin(&t, i915, __func__, "");
 	if (err)
-		goto out_unlock;
+		goto out_free;
 
-	for_each_engine(engine, i915, id) {
+	idx = 0;
+	for_each_uabi_engine(engine, i915) {
 		struct i915_vma *batch;
 
 		batch = recursive_batch(i915);
@@ -964,66 +959,69 @@ static int live_sequential_engines(void *arg)
 			err = PTR_ERR(batch);
 			pr_err("%s: Unable to create batch for %s, err=%d\n",
 			       __func__, engine->name, err);
-			goto out_unlock;
+			goto out_free;
 		}
 
-		request[id] = i915_request_create(engine->kernel_context);
-		if (IS_ERR(request[id])) {
-			err = PTR_ERR(request[id]);
+		request[idx] = i915_request_create(engine->kernel_context);
+		if (IS_ERR(request[idx])) {
+			err = PTR_ERR(request[idx]);
 			pr_err("%s: Request allocation failed for %s with err=%d\n",
 			       __func__, engine->name, err);
 			goto out_request;
 		}
 
 		if (prev) {
-			err = i915_request_await_dma_fence(request[id],
+			err = i915_request_await_dma_fence(request[idx],
 							   &prev->fence);
 			if (err) {
-				i915_request_add(request[id]);
+				i915_request_add(request[idx]);
 				pr_err("%s: Request await failed for %s with err=%d\n",
 				       __func__, engine->name, err);
 				goto out_request;
 			}
 		}
 
-		err = engine->emit_bb_start(request[id],
+		err = engine->emit_bb_start(request[idx],
 					    batch->node.start,
 					    batch->node.size,
 					    0);
 		GEM_BUG_ON(err);
-		request[id]->batch = batch;
+		request[idx]->batch = batch;
 
 		i915_vma_lock(batch);
-		err = i915_request_await_object(request[id], batch->obj, false);
+		err = i915_request_await_object(request[idx],
+						batch->obj, false);
 		if (err == 0)
-			err = i915_vma_move_to_active(batch, request[id], 0);
+			err = i915_vma_move_to_active(batch, request[idx], 0);
 		i915_vma_unlock(batch);
 		GEM_BUG_ON(err);
 
-		i915_request_get(request[id]);
-		i915_request_add(request[id]);
+		i915_request_get(request[idx]);
+		i915_request_add(request[idx]);
 
-		prev = request[id];
+		prev = request[idx];
+		idx++;
 	}
 
-	for_each_engine(engine, i915, id) {
+	idx = 0;
+	for_each_uabi_engine(engine, i915) {
 		long timeout;
 
-		if (i915_request_completed(request[id])) {
+		if (i915_request_completed(request[idx])) {
 			pr_err("%s(%s): request completed too early!\n",
 			       __func__, engine->name);
 			err = -EINVAL;
 			goto out_request;
 		}
 
-		err = recursive_batch_resolve(request[id]->batch);
+		err = recursive_batch_resolve(request[idx]->batch);
 		if (err) {
 			pr_err("%s: failed to resolve batch, err=%d\n",
 			       __func__, err);
 			goto out_request;
 		}
 
-		timeout = i915_request_wait(request[id], 0,
+		timeout = i915_request_wait(request[idx], 0,
 					    MAX_SCHEDULE_TIMEOUT);
 		if (timeout < 0) {
 			err = timeout;
@@ -1032,33 +1030,156 @@ static int live_sequential_engines(void *arg)
 			goto out_request;
 		}
 
-		GEM_BUG_ON(!i915_request_completed(request[id]));
+		GEM_BUG_ON(!i915_request_completed(request[idx]));
+		idx++;
 	}
 
 	err = igt_live_test_end(&t);
 
 out_request:
-	for_each_engine(engine, i915, id) {
+	idx = 0;
+	for_each_uabi_engine(engine, i915) {
 		u32 *cmd;
 
-		if (!request[id])
+		if (!request[idx])
 			break;
 
-		cmd = i915_gem_object_pin_map(request[id]->batch->obj,
+		cmd = i915_gem_object_pin_map(request[idx]->batch->obj,
 					      I915_MAP_WC);
 		if (!IS_ERR(cmd)) {
 			*cmd = MI_BATCH_BUFFER_END;
 			intel_gt_chipset_flush(engine->gt);
 
-			i915_gem_object_unpin_map(request[id]->batch->obj);
+			i915_gem_object_unpin_map(request[idx]->batch->obj);
 		}
 
-		i915_vma_put(request[id]->batch);
-		i915_request_put(request[id]);
+		i915_vma_put(request[idx]->batch);
+		i915_request_put(request[idx]);
+		idx++;
 	}
-out_unlock:
-	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
-	mutex_unlock(&i915->drm.struct_mutex);
+out_free:
+	kfree(request);
+	return err;
+}
+
+static int __live_parallel_engine1(void *arg)
+{
+	struct intel_engine_cs *engine = arg;
+	IGT_TIMEOUT(end_time);
+	unsigned long count;
+
+	count = 0;
+	do {
+		struct i915_request *rq;
+		int err;
+
+		rq = i915_request_create(engine->kernel_context);
+		if (IS_ERR(rq))
+			return PTR_ERR(rq);
+
+		i915_request_get(rq);
+		i915_request_add(rq);
+
+		err = 0;
+		if (i915_request_wait(rq, 0, HZ / 5) < 0)
+			err = -ETIME;
+		i915_request_put(rq);
+		if (err)
+			return err;
+
+		count++;
+	} while (!__igt_timeout(end_time, NULL));
+
+	pr_info("%s: %lu request + sync\n", engine->name, count);
+	return 0;
+}
+
+static int __live_parallel_engineN(void *arg)
+{
+	struct intel_engine_cs *engine = arg;
+	IGT_TIMEOUT(end_time);
+	unsigned long count;
+
+	count = 0;
+	do {
+		struct i915_request *rq;
+
+		rq = i915_request_create(engine->kernel_context);
+		if (IS_ERR(rq))
+			return PTR_ERR(rq);
+
+		i915_request_add(rq);
+		count++;
+	} while (!__igt_timeout(end_time, NULL));
+
+	pr_info("%s: %lu requests\n", engine->name, count);
+	return 0;
+}
+
+static int live_parallel_engines(void *arg)
+{
+	struct drm_i915_private *i915 = arg;
+	static int (* const func[])(void *arg) = {
+		__live_parallel_engine1,
+		__live_parallel_engineN,
+		NULL,
+	};
+	const unsigned int nengines = num_uabi_engines(i915);
+	struct intel_engine_cs *engine;
+	int (* const *fn)(void *arg);
+	struct task_struct **tsk;
+	int err = 0;
+
+	/*
+	 * Check we can submit requests to all engines concurrently. This
+	 * tests that we load up the system maximally.
+	 */
+
+	tsk = kcalloc(nengines, sizeof(*tsk), GFP_KERNEL);
+	if (!tsk)
+		return -ENOMEM;
+
+	for (fn = func; !err && *fn; fn++) {
+		struct igt_live_test t;
+		unsigned int idx;
+
+		err = igt_live_test_begin(&t, i915, __func__, "");
+		if (err)
+			break;
+
+		idx = 0;
+		for_each_uabi_engine(engine, i915) {
+			tsk[idx] = kthread_run(*fn, engine,
+					       "igt/parallel:%s",
+					       engine->name);
+			if (IS_ERR(tsk[idx])) {
+				err = PTR_ERR(tsk[idx]);
+				break;
+			}
+			get_task_struct(tsk[idx++]);
+		}
+
+		yield(); /* start all threads before we kthread_stop() */
+
+		idx = 0;
+		for_each_uabi_engine(engine, i915) {
+			int status;
+
+			if (IS_ERR(tsk[idx]))
+				break;
+
+			status = kthread_stop(tsk[idx]);
+			if (status && !err)
+				err = status;
+
+			put_task_struct(tsk[idx++]);
+		}
+
+		if (igt_live_test_end(&t))
+			err = -EIO;
+	}
+
+	kfree(tsk);
 	return err;
 }
 
@@ -1102,16 +1223,16 @@ max_batches(struct i915_gem_context *ctx, struct intel_engine_cs *engine)
 static int live_breadcrumbs_smoketest(void *arg)
 {
 	struct drm_i915_private *i915 = arg;
-	struct smoketest t[I915_NUM_ENGINES];
-	unsigned int ncpus = num_online_cpus();
+	const unsigned int nengines = num_uabi_engines(i915);
+	const unsigned int ncpus = num_online_cpus();
 	unsigned long num_waits, num_fences;
 	struct intel_engine_cs *engine;
 	struct task_struct **threads;
 	struct igt_live_test live;
-	enum intel_engine_id id;
 	intel_wakeref_t wakeref;
 	struct drm_file *file;
-	unsigned int n;
+	struct smoketest *smoke;
+	unsigned int n, idx;
 	int ret = 0;
 
 	/*
@@ -1130,29 +1251,31 @@ static int live_breadcrumbs_smoketest(void *arg)
 		goto out_rpm;
 	}
 
-	threads = kcalloc(ncpus * I915_NUM_ENGINES,
-			  sizeof(*threads),
-			  GFP_KERNEL);
-	if (!threads) {
+	smoke = kcalloc(nengines, sizeof(*smoke), GFP_KERNEL);
+	if (!smoke) {
 		ret = -ENOMEM;
 		goto out_file;
 	}
 
-	memset(&t[0], 0, sizeof(t[0]));
-	t[0].request_alloc = __live_request_alloc;
-	t[0].ncontexts = 64;
-	t[0].contexts = kmalloc_array(t[0].ncontexts,
-				      sizeof(*t[0].contexts),
-				      GFP_KERNEL);
-	if (!t[0].contexts) {
+	threads = kcalloc(ncpus * nengines, sizeof(*threads), GFP_KERNEL);
+	if (!threads) {
+		ret = -ENOMEM;
+		goto out_smoke;
+	}
+
+	smoke[0].request_alloc = __live_request_alloc;
+	smoke[0].ncontexts = 64;
+	smoke[0].contexts = kcalloc(smoke[0].ncontexts,
+				    sizeof(*smoke[0].contexts),
+				    GFP_KERNEL);
+	if (!smoke[0].contexts) {
 		ret = -ENOMEM;
 		goto out_threads;
 	}
 
-	mutex_lock(&i915->drm.struct_mutex);
-	for (n = 0; n < t[0].ncontexts; n++) {
-		t[0].contexts[n] = live_context(i915, file);
-		if (!t[0].contexts[n]) {
+	for (n = 0; n < smoke[0].ncontexts; n++) {
+		smoke[0].contexts[n] = live_context(i915, file);
+		if (!smoke[0].contexts[n]) {
 			ret = -ENOMEM;
 			goto out_contexts;
 		}
@@ -1162,45 +1285,48 @@ static int live_breadcrumbs_smoketest(void *arg)
 	if (ret)
 		goto out_contexts;
 
-	for_each_engine(engine, i915, id) {
-		t[id] = t[0];
-		t[id].engine = engine;
-		t[id].max_batch = max_batches(t[0].contexts[0], engine);
-		if (t[id].max_batch < 0) {
-			ret = t[id].max_batch;
-			mutex_unlock(&i915->drm.struct_mutex);
+	idx = 0;
+	for_each_uabi_engine(engine, i915) {
+		smoke[idx] = smoke[0];
+		smoke[idx].engine = engine;
+		smoke[idx].max_batch =
+			max_batches(smoke[0].contexts[0], engine);
+		if (smoke[idx].max_batch < 0) {
+			ret = smoke[idx].max_batch;
 			goto out_flush;
 		}
 		/* One ring interleaved between requests from all cpus */
-		t[id].max_batch /= num_online_cpus() + 1;
+		smoke[idx].max_batch /= num_online_cpus() + 1;
 		pr_debug("Limiting batches to %d requests on %s\n",
-			 t[id].max_batch, engine->name);
+			 smoke[idx].max_batch, engine->name);
 
 		for (n = 0; n < ncpus; n++) {
 			struct task_struct *tsk;
 
 			tsk = kthread_run(__igt_breadcrumbs_smoketest,
-					  &t[id], "igt/%d.%d", id, n);
+					  &smoke[idx], "igt/%d.%d", idx, n);
 			if (IS_ERR(tsk)) {
 				ret = PTR_ERR(tsk);
-				mutex_unlock(&i915->drm.struct_mutex);
 				goto out_flush;
 			}
 
 			get_task_struct(tsk);
-			threads[id * ncpus + n] = tsk;
+			threads[idx * ncpus + n] = tsk;
 		}
+
+		idx++;
 	}
-	mutex_unlock(&i915->drm.struct_mutex);
 
+	yield(); /* start all threads before we begin */
 	msleep(jiffies_to_msecs(i915_selftest.timeout_jiffies));
 
 out_flush:
+	idx = 0;
 	num_waits = 0;
 	num_fences = 0;
-	for_each_engine(engine, i915, id) {
+	for_each_uabi_engine(engine, i915) {
 		for (n = 0; n < ncpus; n++) {
-			struct task_struct *tsk = threads[id * ncpus + n];
+			struct task_struct *tsk = threads[idx * ncpus + n];
 			int err;
 
 			if (!tsk)
@@ -1213,19 +1339,20 @@ out_flush:
 			put_task_struct(tsk);
 		}
 
-		num_waits += atomic_long_read(&t[id].num_waits);
-		num_fences += atomic_long_read(&t[id].num_fences);
+		num_waits += atomic_long_read(&smoke[idx].num_waits);
+		num_fences += atomic_long_read(&smoke[idx].num_fences);
+		idx++;
 	}
 	pr_info("Completed %lu waits for %lu fences across %d engines and %d cpus\n",
 		num_waits, num_fences, RUNTIME_INFO(i915)->num_engines, ncpus);
 
-	mutex_lock(&i915->drm.struct_mutex);
 	ret = igt_live_test_end(&live) ?: ret;
 out_contexts:
-	mutex_unlock(&i915->drm.struct_mutex);
-	kfree(t[0].contexts);
+	kfree(smoke[0].contexts);
 out_threads:
 	kfree(threads);
+out_smoke:
+	kfree(smoke);
 out_file:
 	mock_file_free(i915, file);
 out_rpm:
@@ -1240,6 +1367,7 @@ int i915_request_live_selftests(struct drm_i915_private *i915)
 		SUBTEST(live_nop_request),
 		SUBTEST(live_all_engines),
 		SUBTEST(live_sequential_engines),
+		SUBTEST(live_parallel_engines),
 		SUBTEST(live_empty_request),
 		SUBTEST(live_breadcrumbs_smoketest),
 	};
diff --git a/drivers/gpu/drm/i915/selftests/i915_selftest.c b/drivers/gpu/drm/i915/selftests/i915_selftest.c
index 438ea0eaa416..a6cca4ad96f6 100644
--- a/drivers/gpu/drm/i915/selftests/i915_selftest.c
+++ b/drivers/gpu/drm/i915/selftests/i915_selftest.c
@@ -23,13 +23,14 @@
 
 #include <linux/random.h>
 
-#include "../i915_drv.h"
-#include "../i915_selftest.h"
+#include "gt/intel_gt_pm.h"
+#include "i915_drv.h"
+#include "i915_selftest.h"
 
 #include "igt_flush_test.h"
 
 struct i915_selftest i915_selftest __read_mostly = {
-	.timeout_ms = 1000,
+	.timeout_ms = 500,
 };
 
 int i915_mock_sanitycheck(void)
@@ -256,6 +257,10 @@ int __i915_live_setup(void *data)
 {
 	struct drm_i915_private *i915 = data;
 
+	/* The selftests expect an idle system */
+	if (intel_gt_pm_wait_for_idle(&i915->gt))
+		return -EIO;
+
 	return intel_gt_terminally_wedged(&i915->gt);
 }
 
@@ -263,10 +268,8 @@ int __i915_live_teardown(int err, void *data)
 {
 	struct drm_i915_private *i915 = data;
 
-	mutex_lock(&i915->drm.struct_mutex);
-	if (igt_flush_test(i915, I915_WAIT_LOCKED))
+	if (igt_flush_test(i915))
 		err = -EIO;
-	mutex_unlock(&i915->drm.struct_mutex);
 
 	i915_gem_drain_freed_objects(i915);
 
@@ -277,6 +280,10 @@ int __intel_gt_live_setup(void *data)
 {
 	struct intel_gt *gt = data;
 
+	/* The selftests expect an idle system */
+	if (intel_gt_pm_wait_for_idle(gt))
+		return -EIO;
+
 	return intel_gt_terminally_wedged(gt);
 }
 
@@ -284,10 +291,8 @@ int __intel_gt_live_teardown(int err, void *data)
 {
 	struct intel_gt *gt = data;
 
-	mutex_lock(&gt->i915->drm.struct_mutex);
-	if (igt_flush_test(gt->i915, I915_WAIT_LOCKED))
+	if (igt_flush_test(gt->i915))
 		err = -EIO;
-	mutex_unlock(&gt->i915->drm.struct_mutex);
 
 	i915_gem_drain_freed_objects(gt->i915);
 
diff --git a/drivers/gpu/drm/i915/selftests/i915_vma.c b/drivers/gpu/drm/i915/selftests/i915_vma.c
index a5bec0a4cdcc..58b5f40a07dd 100644
--- a/drivers/gpu/drm/i915/selftests/i915_vma.c
+++ b/drivers/gpu/drm/i915/selftests/i915_vma.c
@@ -24,6 +24,7 @@
 
 #include <linux/prime_numbers.h>
 
+#include "gem/i915_gem_context.h"
 #include "gem/selftests/mock_context.h"
 
 #include "i915_scatterlist.h"
@@ -38,7 +39,7 @@ static bool assert_vma(struct i915_vma *vma,
 {
 	bool ok = true;
 
-	if (vma->vm != ctx->vm) {
+	if (vma->vm != rcu_access_pointer(ctx->vm)) {
 		pr_err("VMA created with wrong VM\n");
 		ok = false;
 	}
@@ -113,11 +114,13 @@ static int create_vmas(struct drm_i915_private *i915,
 	list_for_each_entry(obj, objects, st_link) {
 		for (pinned = 0; pinned <= 1; pinned++) {
 			list_for_each_entry(ctx, contexts, link) {
-				struct i915_address_space *vm = ctx->vm;
+				struct i915_address_space *vm;
 				struct i915_vma *vma;
 				int err;
 
+				vm = i915_gem_context_get_vm_rcu(ctx);
 				vma = checked_vma_instance(obj, vm, NULL);
+				i915_vm_put(vm);
 				if (IS_ERR(vma))
 					return PTR_ERR(vma);
 
@@ -170,7 +173,7 @@ static int igt_vma_create(void *arg)
 		}
 
 		nc = 0;
-		for_each_prime_number(num_ctx, MAX_CONTEXT_HW_ID) {
+		for_each_prime_number(num_ctx, 2 * NUM_CONTEXT_TAG) {
 			for (; nc < num_ctx; nc++) {
 				ctx = mock_context(i915, "mock");
 				if (!ctx)
@@ -623,7 +626,7 @@ static bool assert_partial(struct drm_i915_gem_object *obj,
 	struct sgt_iter sgt;
 	dma_addr_t dma;
 
-	for_each_sgt_dma(dma, sgt, vma->pages) {
+	for_each_sgt_daddr(dma, sgt, vma->pages) {
 		dma_addr_t src;
 
 		if (!size) {
@@ -831,13 +834,10 @@ int i915_vma_mock_selftests(void)
 	}
 	mock_init_ggtt(i915, ggtt);
 
-	mutex_lock(&i915->drm.struct_mutex);
 	err = i915_subtests(tests, ggtt);
-	mock_device_flush(i915);
-	mutex_unlock(&i915->drm.struct_mutex);
 
+	mock_device_flush(i915);
 	i915_gem_drain_freed_objects(i915);
-
 	mock_fini_ggtt(ggtt);
 	kfree(ggtt);
 out_put:
@@ -879,8 +879,6 @@ static int igt_vma_remapped_gtt(void *arg)
 	if (IS_ERR(obj))
 		return PTR_ERR(obj);
 
-	mutex_lock(&i915->drm.struct_mutex);
-
 	wakeref = intel_runtime_pm_get(&i915->runtime_pm);
 
 	for (t = types; *t; t++) {
@@ -976,7 +974,6 @@ static int igt_vma_remapped_gtt(void *arg)
 
 out:
 	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
-	mutex_unlock(&i915->drm.struct_mutex);
 	i915_gem_object_put(obj);
 
 	return err;
diff --git a/drivers/gpu/drm/i915/selftests/igt_flush_test.c b/drivers/gpu/drm/i915/selftests/igt_flush_test.c
index d3b5eb402d33..7b0939e3f007 100644
--- a/drivers/gpu/drm/i915/selftests/igt_flush_test.c
+++ b/drivers/gpu/drm/i915/selftests/igt_flush_test.c
@@ -4,39 +4,32 @@
  * Copyright © 2018 Intel Corporation
  */
 
-#include "gem/i915_gem_context.h"
 #include "gt/intel_gt.h"
+#include "gt/intel_gt_requests.h"
 
 #include "i915_drv.h"
 #include "i915_selftest.h"
 
 #include "igt_flush_test.h"
 
-int igt_flush_test(struct drm_i915_private *i915, unsigned int flags)
+int igt_flush_test(struct drm_i915_private *i915)
 {
-	int ret = intel_gt_is_wedged(&i915->gt) ? -EIO : 0;
-	int repeat = !!(flags & I915_WAIT_LOCKED);
+	struct intel_gt *gt = &i915->gt;
+	int ret = intel_gt_is_wedged(gt) ? -EIO : 0;
 
 	cond_resched();
 
-	do {
-		if (i915_gem_wait_for_idle(i915, flags, HZ / 5) == -ETIME) {
-			pr_err("%pS timed out, cancelling all further testing.\n",
-			       __builtin_return_address(0));
+	if (intel_gt_wait_for_idle(gt, HZ / 5) == -ETIME) {
+		pr_err("%pS timed out, cancelling all further testing.\n",
+		       __builtin_return_address(0));
 
-			GEM_TRACE("%pS timed out.\n",
-				  __builtin_return_address(0));
-			GEM_TRACE_DUMP();
+		GEM_TRACE("%pS timed out.\n",
+			  __builtin_return_address(0));
+		GEM_TRACE_DUMP();
 
-			intel_gt_set_wedged(&i915->gt);
-			repeat = 0;
-			ret = -EIO;
-		}
-
-		/* Ensure we also flush after wedging. */
-		if (flags & I915_WAIT_LOCKED)
-			i915_retire_requests(i915);
-	} while (repeat--);
+		intel_gt_set_wedged(gt);
+		ret = -EIO;
+	}
 
 	return ret;
 }
diff --git a/drivers/gpu/drm/i915/selftests/igt_flush_test.h b/drivers/gpu/drm/i915/selftests/igt_flush_test.h
index 63e009927c43..7541fa74e641 100644
--- a/drivers/gpu/drm/i915/selftests/igt_flush_test.h
+++ b/drivers/gpu/drm/i915/selftests/igt_flush_test.h
@@ -9,6 +9,6 @@
 
 struct drm_i915_private;
 
-int igt_flush_test(struct drm_i915_private *i915, unsigned int flags);
+int igt_flush_test(struct drm_i915_private *i915);
 
 #endif /* IGT_FLUSH_TEST_H */
diff --git a/drivers/gpu/drm/i915/selftests/igt_live_test.c b/drivers/gpu/drm/i915/selftests/igt_live_test.c
index 3e902761cd16..c130010a7033 100644
--- a/drivers/gpu/drm/i915/selftests/igt_live_test.c
+++ b/drivers/gpu/drm/i915/selftests/igt_live_test.c
@@ -4,7 +4,8 @@
  * Copyright © 2018 Intel Corporation
  */
 
-#include "../i915_drv.h"
+#include "i915_drv.h"
+#include "gt/intel_gt_requests.h"
 
 #include "../i915_selftest.h"
 #include "igt_flush_test.h"
@@ -15,20 +16,16 @@ int igt_live_test_begin(struct igt_live_test *t,
 			const char *func,
 			const char *name)
 {
+	struct intel_gt *gt = &i915->gt;
 	struct intel_engine_cs *engine;
 	enum intel_engine_id id;
 	int err;
 
-	lockdep_assert_held(&i915->drm.struct_mutex);
-
 	t->i915 = i915;
 	t->func = func;
 	t->name = name;
 
-	err = i915_gem_wait_for_idle(i915,
-				     I915_WAIT_INTERRUPTIBLE |
-				     I915_WAIT_LOCKED,
-				     MAX_SCHEDULE_TIMEOUT);
+	err = intel_gt_wait_for_idle(gt, MAX_SCHEDULE_TIMEOUT);
 	if (err) {
 		pr_err("%s(%s): failed to idle before, with err=%d!",
 		       func, name, err);
@@ -37,7 +34,7 @@ int igt_live_test_begin(struct igt_live_test *t,
 
 	t->reset_global = i915_reset_count(&i915->gpu_error);
 
-	for_each_engine(engine, i915, id)
+	for_each_engine(engine, gt, id)
 		t->reset_engine[id] =
 			i915_reset_engine_count(&i915->gpu_error, engine);
 
@@ -50,9 +47,7 @@ int igt_live_test_end(struct igt_live_test *t)
 	struct intel_engine_cs *engine;
 	enum intel_engine_id id;
 
-	lockdep_assert_held(&i915->drm.struct_mutex);
-
-	if (igt_flush_test(i915, I915_WAIT_LOCKED))
+	if (igt_flush_test(i915))
 		return -EIO;
 
 	if (t->reset_global != i915_reset_count(&i915->gpu_error)) {
@@ -62,7 +57,7 @@ int igt_live_test_end(struct igt_live_test *t)
 		return -EIO;
 	}
 
-	for_each_engine(engine, i915, id) {
+	for_each_engine(engine, &i915->gt, id) {
 		if (t->reset_engine[id] ==
 		    i915_reset_engine_count(&i915->gpu_error, engine))
 			continue;
diff --git a/drivers/gpu/drm/i915/selftests/igt_reset.c b/drivers/gpu/drm/i915/selftests/igt_reset.c
index 7ec8f8b049c6..9f8590b868a9 100644
--- a/drivers/gpu/drm/i915/selftests/igt_reset.c
+++ b/drivers/gpu/drm/i915/selftests/igt_reset.c
@@ -22,7 +22,7 @@ void igt_global_reset_lock(struct intel_gt *gt)
 		wait_event(gt->reset.queue,
 			   !test_bit(I915_RESET_BACKOFF, &gt->reset.flags));
 
-	for_each_engine(engine, gt->i915, id) {
+	for_each_engine(engine, gt, id) {
 		while (test_and_set_bit(I915_RESET_ENGINE + id,
 					&gt->reset.flags))
 			wait_on_bit(&gt->reset.flags, I915_RESET_ENGINE + id,
@@ -35,7 +35,7 @@ void igt_global_reset_unlock(struct intel_gt *gt)
 	struct intel_engine_cs *engine;
 	enum intel_engine_id id;
 
-	for_each_engine(engine, gt->i915, id)
+	for_each_engine(engine, gt, id)
 		clear_bit(I915_RESET_ENGINE + id, &gt->reset.flags);
 
 	clear_bit(I915_RESET_BACKOFF, &gt->reset.flags);
diff --git a/drivers/gpu/drm/i915/selftests/igt_spinner.c b/drivers/gpu/drm/i915/selftests/igt_spinner.c
index 11f04ad48e68..ee8450b871da 100644
--- a/drivers/gpu/drm/i915/selftests/igt_spinner.c
+++ b/drivers/gpu/drm/i915/selftests/igt_spinner.c
@@ -147,7 +147,7 @@ igt_spinner_create_request(struct igt_spinner *spin,
 	intel_gt_chipset_flush(engine->gt);
 
 	if (engine->emit_init_breadcrumb &&
-	    rq->timeline->has_initial_breadcrumb) {
+	    i915_request_timeline(rq)->has_initial_breadcrumb) {
 		err = engine->emit_init_breadcrumb(rq);
 		if (err)
 			goto cancel_rq;
diff --git a/drivers/gpu/drm/i915/selftests/intel_memory_region.c b/drivers/gpu/drm/i915/selftests/intel_memory_region.c
new file mode 100644
index 000000000000..19e1cca8f143
--- /dev/null
+++ b/drivers/gpu/drm/i915/selftests/intel_memory_region.c
@@ -0,0 +1,624 @@
+// SPDX-License-Identifier: MIT
+/*
+ * Copyright © 2019 Intel Corporation
+ */
+
+#include <linux/prime_numbers.h>
+
+#include "../i915_selftest.h"
+
+#include "mock_drm.h"
+#include "mock_gem_device.h"
+#include "mock_region.h"
+
+#include "gem/i915_gem_context.h"
+#include "gem/i915_gem_lmem.h"
+#include "gem/i915_gem_region.h"
+#include "gem/i915_gem_object_blt.h"
+#include "gem/selftests/igt_gem_utils.h"
+#include "gem/selftests/mock_context.h"
+#include "gt/intel_engine_user.h"
+#include "gt/intel_gt.h"
+#include "selftests/igt_flush_test.h"
+#include "selftests/i915_random.h"
+
+static void close_objects(struct intel_memory_region *mem,
+			  struct list_head *objects)
+{
+	struct drm_i915_private *i915 = mem->i915;
+	struct drm_i915_gem_object *obj, *on;
+
+	list_for_each_entry_safe(obj, on, objects, st_link) {
+		if (i915_gem_object_has_pinned_pages(obj))
+			i915_gem_object_unpin_pages(obj);
+		/* No polluting the memory region between tests */
+		__i915_gem_object_put_pages(obj, I915_MM_NORMAL);
+		list_del(&obj->st_link);
+		i915_gem_object_put(obj);
+	}
+
+	cond_resched();
+
+	i915_gem_drain_freed_objects(i915);
+}
+
+static int igt_mock_fill(void *arg)
+{
+	struct intel_memory_region *mem = arg;
+	resource_size_t total = resource_size(&mem->region);
+	resource_size_t page_size;
+	resource_size_t rem;
+	unsigned long max_pages;
+	unsigned long page_num;
+	LIST_HEAD(objects);
+	int err = 0;
+
+	page_size = mem->mm.chunk_size;
+	max_pages = div64_u64(total, page_size);
+	rem = total;
+
+	for_each_prime_number_from(page_num, 1, max_pages) {
+		resource_size_t size = page_num * page_size;
+		struct drm_i915_gem_object *obj;
+
+		obj = i915_gem_object_create_region(mem, size, 0);
+		if (IS_ERR(obj)) {
+			err = PTR_ERR(obj);
+			break;
+		}
+
+		err = i915_gem_object_pin_pages(obj);
+		if (err) {
+			i915_gem_object_put(obj);
+			break;
+		}
+
+		list_add(&obj->st_link, &objects);
+		rem -= size;
+	}
+
+	if (err == -ENOMEM)
+		err = 0;
+	if (err == -ENXIO) {
+		if (page_num * page_size <= rem) {
+			pr_err("%s failed, space still left in region\n",
+			       __func__);
+			err = -EINVAL;
+		} else {
+			err = 0;
+		}
+	}
+
+	close_objects(mem, &objects);
+
+	return err;
+}
+
+static struct drm_i915_gem_object *
+igt_object_create(struct intel_memory_region *mem,
+		  struct list_head *objects,
+		  u64 size,
+		  unsigned int flags)
+{
+	struct drm_i915_gem_object *obj;
+	int err;
+
+	obj = i915_gem_object_create_region(mem, size, flags);
+	if (IS_ERR(obj))
+		return obj;
+
+	err = i915_gem_object_pin_pages(obj);
+	if (err)
+		goto put;
+
+	list_add(&obj->st_link, objects);
+	return obj;
+
+put:
+	i915_gem_object_put(obj);
+	return ERR_PTR(err);
+}
+
+static void igt_object_release(struct drm_i915_gem_object *obj)
+{
+	i915_gem_object_unpin_pages(obj);
+	__i915_gem_object_put_pages(obj, I915_MM_NORMAL);
+	list_del(&obj->st_link);
+	i915_gem_object_put(obj);
+}
+
+static int igt_mock_contiguous(void *arg)
+{
+	struct intel_memory_region *mem = arg;
+	struct drm_i915_gem_object *obj;
+	unsigned long n_objects;
+	LIST_HEAD(objects);
+	LIST_HEAD(holes);
+	I915_RND_STATE(prng);
+	resource_size_t total;
+	resource_size_t min;
+	u64 target;
+	int err = 0;
+
+	total = resource_size(&mem->region);
+
+	/* Min size */
+	obj = igt_object_create(mem, &objects, mem->mm.chunk_size,
+				I915_BO_ALLOC_CONTIGUOUS);
+	if (IS_ERR(obj))
+		return PTR_ERR(obj);
+
+	if (obj->mm.pages->nents != 1) {
+		pr_err("%s min object spans multiple sg entries\n", __func__);
+		err = -EINVAL;
+		goto err_close_objects;
+	}
+
+	igt_object_release(obj);
+
+	/* Max size */
+	obj = igt_object_create(mem, &objects, total, I915_BO_ALLOC_CONTIGUOUS);
+	if (IS_ERR(obj))
+		return PTR_ERR(obj);
+
+	if (obj->mm.pages->nents != 1) {
+		pr_err("%s max object spans multiple sg entries\n", __func__);
+		err = -EINVAL;
+		goto err_close_objects;
+	}
+
+	igt_object_release(obj);
+
+	/* Internal fragmentation should not bleed into the object size */
+	target = i915_prandom_u64_state(&prng);
+	div64_u64_rem(target, total, &target);
+	target = round_up(target, PAGE_SIZE);
+	target = max_t(u64, PAGE_SIZE, target);
+
+	obj = igt_object_create(mem, &objects, target,
+				I915_BO_ALLOC_CONTIGUOUS);
+	if (IS_ERR(obj))
+		return PTR_ERR(obj);
+
+	if (obj->base.size != target) {
+		pr_err("%s obj->base.size(%zx) != target(%llx)\n", __func__,
+		       obj->base.size, target);
+		err = -EINVAL;
+		goto err_close_objects;
+	}
+
+	if (obj->mm.pages->nents != 1) {
+		pr_err("%s object spans multiple sg entries\n", __func__);
+		err = -EINVAL;
+		goto err_close_objects;
+	}
+
+	igt_object_release(obj);
+
+	/*
+	 * Try to fragment the address space, such that half of it is free, but
+	 * the max contiguous block size is SZ_64K.
+	 */
+
+	target = SZ_64K;
+	n_objects = div64_u64(total, target);
+
+	while (n_objects--) {
+		struct list_head *list;
+
+		if (n_objects % 2)
+			list = &holes;
+		else
+			list = &objects;
+
+		obj = igt_object_create(mem, list, target,
+					I915_BO_ALLOC_CONTIGUOUS);
+		if (IS_ERR(obj)) {
+			err = PTR_ERR(obj);
+			goto err_close_objects;
+		}
+	}
+
+	close_objects(mem, &holes);
+
+	min = target;
+	target = total >> 1;
+
+	/* Make sure we can still allocate all the fragmented space */
+	obj = igt_object_create(mem, &objects, target, 0);
+	if (IS_ERR(obj)) {
+		err = PTR_ERR(obj);
+		goto err_close_objects;
+	}
+
+	igt_object_release(obj);
+
+	/*
+	 * Even though we have enough free space, we don't have a big enough
+	 * contiguous block. Make sure that holds true.
+	 */
+
+	do {
+		bool should_fail = target > min;
+
+		obj = igt_object_create(mem, &objects, target,
+					I915_BO_ALLOC_CONTIGUOUS);
+		if (should_fail != IS_ERR(obj)) {
+			pr_err("%s target allocation(%llx) mismatch\n",
+			       __func__, target);
+			err = -EINVAL;
+			goto err_close_objects;
+		}
+
+		target >>= 1;
+	} while (target >= mem->mm.chunk_size);
+
+err_close_objects:
+	list_splice_tail(&holes, &objects);
+	close_objects(mem, &objects);
+	return err;
+}
+
+static int igt_gpu_write_dw(struct intel_context *ce,
+			    struct i915_vma *vma,
+			    u32 dword,
+			    u32 value)
+{
+	return igt_gpu_fill_dw(ce, vma, dword * sizeof(u32),
+			       vma->size >> PAGE_SHIFT, value);
+}
+
+static int igt_cpu_check(struct drm_i915_gem_object *obj, u32 dword, u32 val)
+{
+	unsigned long n;
+	int err;
+
+	i915_gem_object_lock(obj);
+	err = i915_gem_object_set_to_wc_domain(obj, false);
+	i915_gem_object_unlock(obj);
+	if (err)
+		return err;
+
+	err = i915_gem_object_pin_pages(obj);
+	if (err)
+		return err;
+
+	for (n = 0; n < obj->base.size >> PAGE_SHIFT; ++n) {
+		u32 __iomem *base;
+		u32 read_val;
+
+		base = i915_gem_object_lmem_io_map_page_atomic(obj, n);
+
+		read_val = ioread32(base + dword);
+		io_mapping_unmap_atomic(base);
+		if (read_val != val) {
+			pr_err("n=%lu base[%u]=%u, val=%u\n",
+			       n, dword, read_val, val);
+			err = -EINVAL;
+			break;
+		}
+	}
+
+	i915_gem_object_unpin_pages(obj);
+	return err;
+}
+
+static int igt_gpu_write(struct i915_gem_context *ctx,
+			 struct drm_i915_gem_object *obj)
+{
+	struct i915_gem_engines *engines;
+	struct i915_gem_engines_iter it;
+	struct i915_address_space *vm;
+	struct intel_context *ce;
+	I915_RND_STATE(prng);
+	IGT_TIMEOUT(end_time);
+	unsigned int count;
+	struct i915_vma *vma;
+	int *order;
+	int i, n;
+	int err = 0;
+
+	GEM_BUG_ON(!i915_gem_object_has_pinned_pages(obj));
+
+	n = 0;
+	count = 0;
+	for_each_gem_engine(ce, i915_gem_context_lock_engines(ctx), it) {
+		count++;
+		if (!intel_engine_can_store_dword(ce->engine))
+			continue;
+
+		vm = ce->vm;
+		n++;
+	}
+	i915_gem_context_unlock_engines(ctx);
+	if (!n)
+		return 0;
+
+	order = i915_random_order(count * count, &prng);
+	if (!order)
+		return -ENOMEM;
+
+	vma = i915_vma_instance(obj, vm, NULL);
+	if (IS_ERR(vma)) {
+		err = PTR_ERR(vma);
+		goto out_free;
+	}
+
+	err = i915_vma_pin(vma, 0, 0, PIN_USER);
+	if (err)
+		goto out_free;
+
+	i = 0;
+	engines = i915_gem_context_lock_engines(ctx);
+	do {
+		u32 rng = prandom_u32_state(&prng);
+		u32 dword = offset_in_page(rng) / 4;
+
+		ce = engines->engines[order[i] % engines->num_engines];
+		i = (i + 1) % (count * count);
+		if (!ce || !intel_engine_can_store_dword(ce->engine))
+			continue;
+
+		err = igt_gpu_write_dw(ce, vma, dword, rng);
+		if (err)
+			break;
+
+		err = igt_cpu_check(obj, dword, rng);
+		if (err)
+			break;
+	} while (!__igt_timeout(end_time, NULL));
+	i915_gem_context_unlock_engines(ctx);
+
+out_free:
+	kfree(order);
+
+	if (err == -ENOMEM)
+		err = 0;
+
+	return err;
+}
+
+static int igt_lmem_create(void *arg)
+{
+	struct drm_i915_private *i915 = arg;
+	struct drm_i915_gem_object *obj;
+	int err = 0;
+
+	obj = i915_gem_object_create_lmem(i915, PAGE_SIZE, 0);
+	if (IS_ERR(obj))
+		return PTR_ERR(obj);
+
+	err = i915_gem_object_pin_pages(obj);
+	if (err)
+		goto out_put;
+
+	i915_gem_object_unpin_pages(obj);
+out_put:
+	i915_gem_object_put(obj);
+
+	return err;
+}
+
+static int igt_lmem_write_gpu(void *arg)
+{
+	struct drm_i915_private *i915 = arg;
+	struct drm_i915_gem_object *obj;
+	struct i915_gem_context *ctx;
+	struct drm_file *file;
+	I915_RND_STATE(prng);
+	u32 sz;
+	int err;
+
+	file = mock_file(i915);
+	if (IS_ERR(file))
+		return PTR_ERR(file);
+
+	ctx = live_context(i915, file);
+	if (IS_ERR(ctx)) {
+		err = PTR_ERR(ctx);
+		goto out_file;
+	}
+
+	sz = round_up(prandom_u32_state(&prng) % SZ_32M, PAGE_SIZE);
+
+	obj = i915_gem_object_create_lmem(i915, sz, 0);
+	if (IS_ERR(obj)) {
+		err = PTR_ERR(obj);
+		goto out_file;
+	}
+
+	err = i915_gem_object_pin_pages(obj);
+	if (err)
+		goto out_put;
+
+	err = igt_gpu_write(ctx, obj);
+	if (err)
+		pr_err("igt_gpu_write failed(%d)\n", err);
+
+	i915_gem_object_unpin_pages(obj);
+out_put:
+	i915_gem_object_put(obj);
+out_file:
+	mock_file_free(i915, file);
+	return err;
+}
+
+static struct intel_engine_cs *
+random_engine_class(struct drm_i915_private *i915,
+		    unsigned int class,
+		    struct rnd_state *prng)
+{
+	struct intel_engine_cs *engine;
+	unsigned int count;
+
+	count = 0;
+	for (engine = intel_engine_lookup_user(i915, class, 0);
+	     engine && engine->uabi_class == class;
+	     engine = rb_entry_safe(rb_next(&engine->uabi_node),
+				    typeof(*engine), uabi_node))
+		count++;
+
+	count = i915_prandom_u32_max_state(count, prng);
+	return intel_engine_lookup_user(i915, class, count);
+}
+
+static int igt_lmem_write_cpu(void *arg)
+{
+	struct drm_i915_private *i915 = arg;
+	struct drm_i915_gem_object *obj;
+	I915_RND_STATE(prng);
+	IGT_TIMEOUT(end_time);
+	u32 bytes[] = {
+		0, /* rng placeholder */
+		sizeof(u32),
+		sizeof(u64),
+		64, /* cl */
+		PAGE_SIZE,
+		PAGE_SIZE - sizeof(u32),
+		PAGE_SIZE - sizeof(u64),
+		PAGE_SIZE - 64,
+	};
+	struct intel_engine_cs *engine;
+	u32 *vaddr;
+	u32 sz;
+	u32 i;
+	int *order;
+	int count;
+	int err;
+
+	engine = random_engine_class(i915, I915_ENGINE_CLASS_COPY, &prng);
+	if (!engine)
+		return 0;
+
+	pr_info("%s: using %s\n", __func__, engine->name);
+
+	sz = round_up(prandom_u32_state(&prng) % SZ_32M, PAGE_SIZE);
+	sz = max_t(u32, 2 * PAGE_SIZE, sz);
+
+	obj = i915_gem_object_create_lmem(i915, sz, I915_BO_ALLOC_CONTIGUOUS);
+	if (IS_ERR(obj))
+		return PTR_ERR(obj);
+
+	vaddr = i915_gem_object_pin_map(obj, I915_MAP_WC);
+	if (IS_ERR(vaddr)) {
+		err = PTR_ERR(vaddr);
+		goto out_put;
+	}
+
+	/* Put the pages into a known state -- from the gpu for added fun */
+	err = i915_gem_object_fill_blt(obj, engine->kernel_context, 0xdeadbeaf);
+	if (err)
+		goto out_unpin;
+
+	i915_gem_object_lock(obj);
+	err = i915_gem_object_set_to_wc_domain(obj, true);
+	i915_gem_object_unlock(obj);
+	if (err)
+		goto out_unpin;
+
+	count = ARRAY_SIZE(bytes);
+	order = i915_random_order(count * count, &prng);
+	if (!order) {
+		err = -ENOMEM;
+		goto out_unpin;
+	}
+
+	/* We want to throw in a random width/align */
+	bytes[0] = igt_random_offset(&prng, 0, PAGE_SIZE, sizeof(u32),
+				     sizeof(u32));
+
+	i = 0;
+	do {
+		u32 offset;
+		u32 align;
+		u32 dword;
+		u32 size;
+		u32 val;
+
+		size = bytes[order[i] % count];
+		i = (i + 1) % (count * count);
+
+		align = bytes[order[i] % count];
+		i = (i + 1) % (count * count);
+
+		align = max_t(u32, sizeof(u32), rounddown_pow_of_two(align));
+
+		offset = igt_random_offset(&prng, 0, obj->base.size,
+					   size, align);
+
+		val = prandom_u32_state(&prng);
+		memset32(vaddr + offset / sizeof(u32), val ^ 0xdeadbeaf,
+			 size / sizeof(u32));
+
+		/*
+		 * Sample random dw -- don't waste precious time reading every
+		 * single dw.
+		 */
+		dword = igt_random_offset(&prng, offset,
+					  offset + size,
+					  sizeof(u32), sizeof(u32));
+		dword /= sizeof(u32);
+		if (vaddr[dword] != (val ^ 0xdeadbeaf)) {
+			pr_err("%s vaddr[%u]=%u, val=%u, size=%u, align=%u, offset=%u\n",
+			       __func__, dword, vaddr[dword], val ^ 0xdeadbeaf,
+			       size, align, offset);
+			err = -EINVAL;
+			break;
+		}
+	} while (!__igt_timeout(end_time, NULL));
+
+out_unpin:
+	i915_gem_object_unpin_map(obj);
+out_put:
+	i915_gem_object_put(obj);
+
+	return err;
+}
+
+int intel_memory_region_mock_selftests(void)
+{
+	static const struct i915_subtest tests[] = {
+		SUBTEST(igt_mock_fill),
+		SUBTEST(igt_mock_contiguous),
+	};
+	struct intel_memory_region *mem;
+	struct drm_i915_private *i915;
+	int err;
+
+	i915 = mock_gem_device();
+	if (!i915)
+		return -ENOMEM;
+
+	mem = mock_region_create(i915, 0, SZ_2G, I915_GTT_PAGE_SIZE_4K, 0);
+	if (IS_ERR(mem)) {
+		pr_err("failed to create memory region\n");
+		err = PTR_ERR(mem);
+		goto out_unref;
+	}
+
+	err = i915_subtests(tests, mem);
+
+	intel_memory_region_put(mem);
+out_unref:
+	drm_dev_put(&i915->drm);
+	return err;
+}
+
+int intel_memory_region_live_selftests(struct drm_i915_private *i915)
+{
+	static const struct i915_subtest tests[] = {
+		SUBTEST(igt_lmem_create),
+		SUBTEST(igt_lmem_write_cpu),
+		SUBTEST(igt_lmem_write_gpu),
+	};
+
+	if (!HAS_LMEM(i915)) {
+		pr_info("device lacks LMEM support, skipping\n");
+		return 0;
+	}
+
+	if (intel_gt_is_wedged(&i915->gt))
+		return 0;
+
+	return i915_live_subtests(tests, i915);
+}
diff --git a/drivers/gpu/drm/i915/selftests/intel_uncore.c b/drivers/gpu/drm/i915/selftests/intel_uncore.c
index 86815c6072a1..0e4e6be0101d 100644
--- a/drivers/gpu/drm/i915/selftests/intel_uncore.c
+++ b/drivers/gpu/drm/i915/selftests/intel_uncore.c
@@ -67,6 +67,7 @@ static int intel_shadow_table_check(void)
 	} reg_lists[] = {
 		{ gen8_shadowed_regs, ARRAY_SIZE(gen8_shadowed_regs) },
 		{ gen11_shadowed_regs, ARRAY_SIZE(gen11_shadowed_regs) },
+		{ gen12_shadowed_regs, ARRAY_SIZE(gen12_shadowed_regs) },
 	};
 	const i915_reg_t *reg;
 	unsigned int i, j;
@@ -101,6 +102,7 @@ int intel_uncore_mock_selftests(void)
 		{ __chv_fw_ranges, ARRAY_SIZE(__chv_fw_ranges), false },
 		{ __gen9_fw_ranges, ARRAY_SIZE(__gen9_fw_ranges), true },
 		{ __gen11_fw_ranges, ARRAY_SIZE(__gen11_fw_ranges), true },
+		{ __gen12_fw_ranges, ARRAY_SIZE(__gen12_fw_ranges), true },
 	};
 	int err, i;
 
@@ -138,19 +140,19 @@ static int live_forcewake_ops(void *arg)
 		}
 	};
 	const struct reg *r;
-	struct drm_i915_private *i915 = arg;
+	struct intel_gt *gt = arg;
 	struct intel_uncore_forcewake_domain *domain;
-	struct intel_uncore *uncore = &i915->uncore;
+	struct intel_uncore *uncore = gt->uncore;
 	struct intel_engine_cs *engine;
 	enum intel_engine_id id;
 	intel_wakeref_t wakeref;
 	unsigned int tmp;
 	int err = 0;
 
-	GEM_BUG_ON(i915->gt.awake);
+	GEM_BUG_ON(gt->awake);
 
 	/* vlv/chv with their pcu behave differently wrt reads */
-	if (IS_VALLEYVIEW(i915) || IS_CHERRYVIEW(i915)) {
+	if (IS_VALLEYVIEW(gt->i915) || IS_CHERRYVIEW(gt->i915)) {
 		pr_debug("PCU fakes forcewake badly; skipping\n");
 		return 0;
 	}
@@ -168,15 +170,15 @@ static int live_forcewake_ops(void *arg)
 
 	/* We have to pick carefully to get the exact behaviour we need */
 	for (r = registers; r->name; r++)
-		if (r->platforms & INTEL_INFO(i915)->gen_mask)
+		if (r->platforms & INTEL_INFO(gt->i915)->gen_mask)
 			break;
 	if (!r->name) {
 		pr_debug("Forcewaked register not known for %s; skipping\n",
-			 intel_platform_name(INTEL_INFO(i915)->platform));
+			 intel_platform_name(INTEL_INFO(gt->i915)->platform));
 		return 0;
 	}
 
-	wakeref = intel_runtime_pm_get(&i915->runtime_pm);
+	wakeref = intel_runtime_pm_get(uncore->rpm);
 
 	for_each_fw_domain(domain, uncore, tmp) {
 		smp_store_mb(domain->active, false);
@@ -186,7 +188,7 @@ static int live_forcewake_ops(void *arg)
 		intel_uncore_fw_release_timer(&domain->timer);
 	}
 
-	for_each_engine(engine, i915, id) {
+	for_each_engine(engine, gt, id) {
 		i915_reg_t mmio = _MMIO(engine->mmio_base + r->offset);
 		u32 __iomem *reg = uncore->regs + engine->mmio_base + r->offset;
 		enum forcewake_domains fw_domains;
@@ -247,22 +249,22 @@ static int live_forcewake_ops(void *arg)
 	}
 
 out_rpm:
-	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
+	intel_runtime_pm_put(uncore->rpm, wakeref);
 	return err;
 }
 
 static int live_forcewake_domains(void *arg)
 {
 #define FW_RANGE 0x40000
-	struct drm_i915_private *dev_priv = arg;
-	struct intel_uncore *uncore = &dev_priv->uncore;
+	struct intel_gt *gt = arg;
+	struct intel_uncore *uncore = gt->uncore;
 	unsigned long *valid;
 	u32 offset;
 	int err;
 
-	if (!HAS_FPGA_DBG_UNCLAIMED(dev_priv) &&
-	    !IS_VALLEYVIEW(dev_priv) &&
-	    !IS_CHERRYVIEW(dev_priv))
+	if (!HAS_FPGA_DBG_UNCLAIMED(gt->i915) &&
+	    !IS_VALLEYVIEW(gt->i915) &&
+	    !IS_CHERRYVIEW(gt->i915))
 		return 0;
 
 	/*
@@ -281,7 +283,7 @@ static int live_forcewake_domains(void *arg)
 	for (offset = 0; offset < FW_RANGE; offset += 4) {
 		i915_reg_t reg = { offset };
 
-		(void)I915_READ_FW(reg);
+		intel_uncore_posting_read_fw(uncore, reg);
 		if (!check_for_unclaimed_mmio(uncore))
 			set_bit(offset, valid);
 	}
@@ -298,7 +300,7 @@ static int live_forcewake_domains(void *arg)
 
 		check_for_unclaimed_mmio(uncore);
 
-		(void)I915_READ(reg);
+		intel_uncore_posting_read_fw(uncore, reg);
 		if (check_for_unclaimed_mmio(uncore)) {
 			pr_err("Unclaimed mmio read to register 0x%04x\n",
 			       offset);
@@ -310,21 +312,23 @@ static int live_forcewake_domains(void *arg)
 	return err;
 }
 
+static int live_fw_table(void *arg)
+{
+	struct intel_gt *gt = arg;
+
+	/* Confirm the table we load is still valid */
+	return intel_fw_table_check(gt->uncore->fw_domains_table,
+				    gt->uncore->fw_domains_table_entries,
+				    INTEL_GEN(gt->i915) >= 9);
+}
+
 int intel_uncore_live_selftests(struct drm_i915_private *i915)
 {
 	static const struct i915_subtest tests[] = {
+		SUBTEST(live_fw_table),
 		SUBTEST(live_forcewake_ops),
 		SUBTEST(live_forcewake_domains),
 	};
 
-	int err;
-
-	/* Confirm the table we load is still valid */
-	err = intel_fw_table_check(i915->uncore.fw_domains_table,
-				   i915->uncore.fw_domains_table_entries,
-				   INTEL_GEN(i915) >= 9);
-	if (err)
-		return err;
-
-	return i915_subtests(tests, i915);
+	return intel_gt_live_subtests(tests, &i915->gt);
 }
diff --git a/drivers/gpu/drm/i915/selftests/mock_gem_device.c b/drivers/gpu/drm/i915/selftests/mock_gem_device.c
index 01a89c071bf5..27ed3cee6a9b 100644
--- a/drivers/gpu/drm/i915/selftests/mock_gem_device.c
+++ b/drivers/gpu/drm/i915/selftests/mock_gem_device.c
@@ -26,27 +26,29 @@
 #include <linux/pm_runtime.h>
 
 #include "gt/intel_gt.h"
+#include "gt/intel_gt_requests.h"
 #include "gt/mock_engine.h"
+#include "intel_memory_region.h"
 
 #include "mock_request.h"
 #include "mock_gem_device.h"
 #include "mock_gtt.h"
 #include "mock_uncore.h"
+#include "mock_region.h"
 
 #include "gem/selftests/mock_context.h"
 #include "gem/selftests/mock_gem_object.h"
 
 void mock_device_flush(struct drm_i915_private *i915)
 {
+	struct intel_gt *gt = &i915->gt;
 	struct intel_engine_cs *engine;
 	enum intel_engine_id id;
 
-	lockdep_assert_held(&i915->drm.struct_mutex);
-
 	do {
-		for_each_engine(engine, i915, id)
+		for_each_engine(engine, gt, id)
 			mock_engine_flush(engine);
-	} while (i915_retire_requests(i915));
+	} while (intel_gt_retire_requests_timeout(gt, MAX_SCHEDULE_TIMEOUT));
 }
 
 static void mock_device_release(struct drm_device *dev)
@@ -55,31 +57,23 @@ static void mock_device_release(struct drm_device *dev)
 	struct intel_engine_cs *engine;
 	enum intel_engine_id id;
 
-	mutex_lock(&i915->drm.struct_mutex);
 	mock_device_flush(i915);
-	mutex_unlock(&i915->drm.struct_mutex);
 
-	flush_work(&i915->gem.idle_work);
 	i915_gem_drain_workqueue(i915);
 
-	mutex_lock(&i915->drm.struct_mutex);
-	for_each_engine(engine, i915, id)
+	for_each_engine(engine, &i915->gt, id)
 		mock_engine_free(engine);
-	i915_gem_contexts_fini(i915);
-	mutex_unlock(&i915->drm.struct_mutex);
+	i915_gem_driver_release__contexts(i915);
 
 	intel_timelines_fini(i915);
 
 	drain_workqueue(i915->wq);
 	i915_gem_drain_freed_objects(i915);
 
-	mutex_lock(&i915->drm.struct_mutex);
 	mock_fini_ggtt(&i915->ggtt);
-	mutex_unlock(&i915->drm.struct_mutex);
-
 	destroy_workqueue(i915->wq);
 
-	i915_gemfs_fini(i915);
+	intel_memory_regions_driver_release(i915);
 
 	drm_mode_config_cleanup(&i915->drm);
 
@@ -103,14 +97,6 @@ static void release_dev(struct device *dev)
 	kfree(pdev);
 }
 
-static void mock_retire_work_handler(struct work_struct *work)
-{
-}
-
-static void mock_idle_work_handler(struct work_struct *work)
-{
-}
-
 static int pm_domain_resume(struct device *dev)
 {
 	return pm_generic_runtime_resume(dev);
@@ -178,10 +164,15 @@ struct drm_i915_private *mock_gem_device(void)
 		I915_GTT_PAGE_SIZE_64K |
 		I915_GTT_PAGE_SIZE_2M;
 
-	mock_uncore_init(&i915->uncore);
+	mkwrite_device_info(i915)->memory_regions = REGION_SMEM;
+	intel_memory_regions_hw_probe(i915);
+
+	mock_uncore_init(&i915->uncore, i915);
+
 	i915_gem_init__mm(i915);
 	intel_gt_init_early(&i915->gt, i915);
 	atomic_inc(&i915->gt.wakeref.count); /* disable; no hw support */
+	i915->gt.awake = -ENODEV;
 
 	i915->wq = alloc_ordered_workqueue("mock", 0);
 	if (!i915->wq)
@@ -189,15 +180,8 @@ struct drm_i915_private *mock_gem_device(void)
 
 	mock_init_contexts(i915);
 
-	INIT_DELAYED_WORK(&i915->gem.retire_work, mock_retire_work_handler);
-	INIT_WORK(&i915->gem.idle_work, mock_idle_work_handler);
-
-	i915->gt.awake = true;
-
 	intel_timelines_init(i915);
 
-	mutex_lock(&i915->drm.struct_mutex);
-
 	mock_init_ggtt(i915, &i915->ggtt);
 
 	mkwrite_device_info(i915)->engine_mask = BIT(0);
@@ -214,21 +198,18 @@ struct drm_i915_private *mock_gem_device(void)
 		goto err_context;
 
 	intel_engines_driver_register(i915);
-	mutex_unlock(&i915->drm.struct_mutex);
-
-	WARN_ON(i915_gemfs_init(i915));
 
 	return i915;
 
 err_context:
-	i915_gem_contexts_fini(i915);
+	i915_gem_driver_release__contexts(i915);
 err_engine:
 	mock_engine_free(i915->engine[RCS0]);
 err_unlock:
-	mutex_unlock(&i915->drm.struct_mutex);
 	intel_timelines_fini(i915);
 	destroy_workqueue(i915->wq);
 err_drv:
+	intel_memory_regions_driver_release(i915);
 	drm_mode_config_cleanup(&i915->drm);
 	drm_dev_fini(&i915->drm);
 put_device:
diff --git a/drivers/gpu/drm/i915/selftests/mock_gtt.c b/drivers/gpu/drm/i915/selftests/mock_gtt.c
index e62a67e0f79c..20ac3844edec 100644
--- a/drivers/gpu/drm/i915/selftests/mock_gtt.c
+++ b/drivers/gpu/drm/i915/selftests/mock_gtt.c
@@ -43,7 +43,7 @@ static int mock_bind_ppgtt(struct i915_vma *vma,
 			   u32 flags)
 {
 	GEM_BUG_ON(flags & I915_VMA_GLOBAL_BIND);
-	vma->flags |= I915_VMA_LOCAL_BIND;
+	set_bit(I915_VMA_LOCAL_BIND_BIT, __i915_vma_flags(vma));
 	return 0;
 }
 
@@ -63,6 +63,7 @@ struct i915_ppgtt *mock_ppgtt(struct drm_i915_private *i915, const char *name)
 	if (!ppgtt)
 		return NULL;
 
+	ppgtt->vm.gt = &i915->gt;
 	ppgtt->vm.i915 = i915;
 	ppgtt->vm.total = round_down(U64_MAX, PAGE_SIZE);
 	ppgtt->vm.file = ERR_PTR(-ENODEV);
@@ -86,7 +87,7 @@ static int mock_bind_ggtt(struct i915_vma *vma,
 			  enum i915_cache_level cache_level,
 			  u32 flags)
 {
-	vma->flags |= I915_VMA_GLOBAL_BIND | I915_VMA_LOCAL_BIND;
+	atomic_or(I915_VMA_GLOBAL_BIND | I915_VMA_LOCAL_BIND, &vma->flags);
 	return 0;
 }
 
@@ -117,8 +118,7 @@ void mock_init_ggtt(struct drm_i915_private *i915, struct i915_ggtt *ggtt)
 	ggtt->vm.vma_ops.clear_pages = clear_pages;
 
 	i915_address_space_init(&ggtt->vm, VM_CLASS_GGTT);
-
-	intel_gt_init_hw(i915);
+	i915->gt.ggtt = ggtt;
 }
 
 void mock_fini_ggtt(struct i915_ggtt *ggtt)
diff --git a/drivers/gpu/drm/i915/selftests/mock_region.c b/drivers/gpu/drm/i915/selftests/mock_region.c
new file mode 100644
index 000000000000..b2ad41c27e67
--- /dev/null
+++ b/drivers/gpu/drm/i915/selftests/mock_region.c
@@ -0,0 +1,60 @@
+// SPDX-License-Identifier: MIT
+/*
+ * Copyright © 2019 Intel Corporation
+ */
+
+#include "gem/i915_gem_region.h"
+#include "intel_memory_region.h"
+
+#include "mock_region.h"
+
+static const struct drm_i915_gem_object_ops mock_region_obj_ops = {
+	.get_pages = i915_gem_object_get_pages_buddy,
+	.put_pages = i915_gem_object_put_pages_buddy,
+	.release = i915_gem_object_release_memory_region,
+};
+
+static struct drm_i915_gem_object *
+mock_object_create(struct intel_memory_region *mem,
+		   resource_size_t size,
+		   unsigned int flags)
+{
+	static struct lock_class_key lock_class;
+	struct drm_i915_private *i915 = mem->i915;
+	struct drm_i915_gem_object *obj;
+
+	if (size > BIT(mem->mm.max_order) * mem->mm.chunk_size)
+		return ERR_PTR(-E2BIG);
+
+	obj = i915_gem_object_alloc();
+	if (!obj)
+		return ERR_PTR(-ENOMEM);
+
+	drm_gem_private_object_init(&i915->drm, &obj->base, size);
+	i915_gem_object_init(obj, &mock_region_obj_ops, &lock_class);
+
+	obj->read_domains = I915_GEM_DOMAIN_CPU | I915_GEM_DOMAIN_GTT;
+
+	i915_gem_object_set_cache_coherency(obj, I915_CACHE_NONE);
+
+	i915_gem_object_init_memory_region(obj, mem, flags);
+
+	return obj;
+}
+
+static const struct intel_memory_region_ops mock_region_ops = {
+	.init = intel_memory_region_init_buddy,
+	.release = intel_memory_region_release_buddy,
+	.create_object = mock_object_create,
+};
+
+struct intel_memory_region *
+mock_region_create(struct drm_i915_private *i915,
+		   resource_size_t start,
+		   resource_size_t size,
+		   resource_size_t min_page_size,
+		   resource_size_t io_start)
+{
+	return intel_memory_region_create(i915, start, size, min_page_size,
+					  io_start, &mock_region_ops);
+}
diff --git a/drivers/gpu/drm/i915/selftests/mock_region.h b/drivers/gpu/drm/i915/selftests/mock_region.h
new file mode 100644
index 000000000000..24608089d833
--- /dev/null
+++ b/drivers/gpu/drm/i915/selftests/mock_region.h
@@ -0,0 +1,16 @@
+/* SPDX-License-Identifier: MIT */
+/*
+ * Copyright © 2019 Intel Corporation
+ */
+
+#ifndef __MOCK_REGION_H
+#define __MOCK_REGION_H
+
+struct intel_memory_region *
+mock_region_create(struct drm_i915_private *i915,
+		   resource_size_t start,
+		   resource_size_t size,
+		   resource_size_t min_page_size,
+		   resource_size_t io_start);
+
+#endif /* !__MOCK_REGION_H */
diff --git a/drivers/gpu/drm/i915/selftests/mock_uncore.c b/drivers/gpu/drm/i915/selftests/mock_uncore.c
index 49585f16d4a2..ca57e4008701 100644
--- a/drivers/gpu/drm/i915/selftests/mock_uncore.c
+++ b/drivers/gpu/drm/i915/selftests/mock_uncore.c
@@ -39,8 +39,11 @@ __nop_read(16)
 __nop_read(32)
 __nop_read(64)
 
-void mock_uncore_init(struct intel_uncore *uncore)
+void mock_uncore_init(struct intel_uncore *uncore,
+		      struct drm_i915_private *i915)
 {
+	intel_uncore_init_early(uncore, i915);
+
 	ASSIGN_RAW_WRITE_MMIO_VFUNCS(uncore, nop);
 	ASSIGN_RAW_READ_MMIO_VFUNCS(uncore, nop);
 }
diff --git a/drivers/gpu/drm/i915/selftests/mock_uncore.h b/drivers/gpu/drm/i915/selftests/mock_uncore.h
index dacb36b5ffcd..8a2cc553f466 100644
--- a/drivers/gpu/drm/i915/selftests/mock_uncore.h
+++ b/drivers/gpu/drm/i915/selftests/mock_uncore.h
@@ -25,6 +25,7 @@
 #ifndef __MOCK_UNCORE_H
 #define __MOCK_UNCORE_H
 
-void mock_uncore_init(struct intel_uncore *uncore);
+void mock_uncore_init(struct intel_uncore *uncore,
+		      struct drm_i915_private *i915);
 
 #endif /* !__MOCK_UNCORE_H */
diff --git a/drivers/gpu/drm/imx/imx-ldb.c b/drivers/gpu/drm/imx/imx-ldb.c
index 695f307f36b2..208069faf183 100644
--- a/drivers/gpu/drm/imx/imx-ldb.c
+++ b/drivers/gpu/drm/imx/imx-ldb.c
@@ -20,6 +20,7 @@
 
 #include <drm/drm_atomic.h>
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_fb_helper.h>
 #include <drm/drm_of.h>
 #include <drm/drm_panel.h>
diff --git a/drivers/gpu/drm/imx/parallel-display.c b/drivers/gpu/drm/imx/parallel-display.c
index e7ce17503ae1..35518e5de356 100644
--- a/drivers/gpu/drm/imx/parallel-display.c
+++ b/drivers/gpu/drm/imx/parallel-display.c
@@ -13,6 +13,7 @@
 #include <video/of_display_timing.h>
 
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_fb_helper.h>
 #include <drm/drm_of.h>
 #include <drm/drm_panel.h>
diff --git a/drivers/gpu/drm/ingenic/ingenic-drm.c b/drivers/gpu/drm/ingenic/ingenic-drm.c
index 2e2ed653e9c6..ec32e1c67335 100644
--- a/drivers/gpu/drm/ingenic/ingenic-drm.c
+++ b/drivers/gpu/drm/ingenic/ingenic-drm.c
@@ -13,6 +13,7 @@
 
 #include <drm/drm_atomic.h>
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_crtc.h>
 #include <drm/drm_crtc_helper.h>
 #include <drm/drm_drv.h>
@@ -676,8 +677,8 @@ static int ingenic_drm_probe(struct platform_device *pdev)
 	}
 
 	if (panel)
-		bridge = devm_drm_panel_bridge_add(dev, panel,
-						   DRM_MODE_CONNECTOR_DPI);
+		bridge = devm_drm_panel_bridge_add_typed(dev, panel,
+							 DRM_MODE_CONNECTOR_DPI);
 
 	priv->dma_hwdesc = dma_alloc_coherent(dev, sizeof(*priv->dma_hwdesc),
 					      &priv->dma_hwdesc_phys,
diff --git a/drivers/gpu/drm/lima/Kconfig b/drivers/gpu/drm/lima/Kconfig
index bb4ddc6bb0a6..571dc369a7e9 100644
--- a/drivers/gpu/drm/lima/Kconfig
+++ b/drivers/gpu/drm/lima/Kconfig
@@ -9,5 +9,6 @@ config DRM_LIMA
        depends on COMMON_CLK
        depends on OF
        select DRM_SCHED
+       select DRM_GEM_SHMEM_HELPER
        help
          DRM driver for ARM Mali 400/450 GPUs.
diff --git a/drivers/gpu/drm/lima/Makefile b/drivers/gpu/drm/lima/Makefile
index 38cc70281ba5..a85444b0a1d4 100644
--- a/drivers/gpu/drm/lima/Makefile
+++ b/drivers/gpu/drm/lima/Makefile
@@ -13,9 +13,7 @@ lima-y := \
 	lima_vm.o \
 	lima_sched.o \
 	lima_ctx.o \
-	lima_gem_prime.o \
 	lima_dlbu.o \
-	lima_bcast.o \
-	lima_object.o
+	lima_bcast.o
 
 obj-$(CONFIG_DRM_LIMA) += lima.o
diff --git a/drivers/gpu/drm/lima/lima_device.c b/drivers/gpu/drm/lima/lima_device.c
index d86b8d81a483..19829b543024 100644
--- a/drivers/gpu/drm/lima/lima_device.c
+++ b/drivers/gpu/drm/lima/lima_device.c
@@ -105,7 +105,8 @@ static int lima_clk_init(struct lima_device *dev)
 	if (err)
 		goto error_out0;
 
-	dev->reset = devm_reset_control_get_optional(dev->dev, NULL);
+	dev->reset = devm_reset_control_array_get_optional_shared(dev->dev);
+
 	if (IS_ERR(dev->reset)) {
 		err = PTR_ERR(dev->reset);
 		if (err != -EPROBE_DEFER)
@@ -313,7 +314,7 @@ int lima_device_init(struct lima_device *ldev)
 		ldev->va_end = LIMA_VA_RESERVE_START;
 		ldev->dlbu_cpu = dma_alloc_wc(
 			ldev->dev, LIMA_PAGE_SIZE,
-			&ldev->dlbu_dma, GFP_KERNEL);
+			&ldev->dlbu_dma, GFP_KERNEL | __GFP_NOWARN);
 		if (!ldev->dlbu_cpu) {
 			err = -ENOMEM;
 			goto err_out2;
diff --git a/drivers/gpu/drm/lima/lima_drv.c b/drivers/gpu/drm/lima/lima_drv.c
index 75ec703d22e0..124efe4fa97b 100644
--- a/drivers/gpu/drm/lima/lima_drv.c
+++ b/drivers/gpu/drm/lima/lima_drv.c
@@ -12,7 +12,6 @@
 
 #include "lima_drv.h"
 #include "lima_gem.h"
-#include "lima_gem_prime.h"
 #include "lima_vm.h"
 
 int lima_sched_timeout_ms;
@@ -240,16 +239,7 @@ static const struct drm_ioctl_desc lima_drm_driver_ioctls[] = {
 	DRM_IOCTL_DEF_DRV(LIMA_CTX_FREE, lima_ioctl_ctx_free, DRM_RENDER_ALLOW),
 };
 
-static const struct file_operations lima_drm_driver_fops = {
-	.owner              = THIS_MODULE,
-	.open               = drm_open,
-	.release            = drm_release,
-	.unlocked_ioctl     = drm_ioctl,
-#ifdef CONFIG_COMPAT
-	.compat_ioctl       = drm_compat_ioctl,
-#endif
-	.mmap               = lima_gem_mmap,
-};
+DEFINE_DRM_GEM_FOPS(lima_drm_driver_fops);
 
 static struct drm_driver lima_drm_driver = {
 	.driver_features    = DRIVER_RENDER | DRIVER_GEM | DRIVER_SYNCOBJ,
@@ -258,10 +248,6 @@ static struct drm_driver lima_drm_driver = {
 	.ioctls             = lima_drm_driver_ioctls,
 	.num_ioctls         = ARRAY_SIZE(lima_drm_driver_ioctls),
 	.fops               = &lima_drm_driver_fops,
-	.gem_free_object_unlocked = lima_gem_free_object,
-	.gem_open_object    = lima_gem_object_open,
-	.gem_close_object   = lima_gem_object_close,
-	.gem_vm_ops         = &lima_gem_vm_ops,
 	.name               = "lima",
 	.desc               = "lima DRM",
 	.date               = "20190217",
@@ -269,11 +255,11 @@ static struct drm_driver lima_drm_driver = {
 	.minor              = 0,
 	.patchlevel         = 0,
 
+	.gem_create_object  = lima_gem_create_object,
 	.prime_fd_to_handle = drm_gem_prime_fd_to_handle,
-	.gem_prime_import_sg_table = lima_gem_prime_import_sg_table,
+	.gem_prime_import_sg_table = drm_gem_shmem_prime_import_sg_table,
 	.prime_handle_to_fd = drm_gem_prime_handle_to_fd,
-	.gem_prime_get_sg_table = lima_gem_prime_get_sg_table,
-	.gem_prime_mmap = lima_gem_prime_mmap,
+	.gem_prime_mmap = drm_gem_prime_mmap,
 };
 
 static int lima_pdev_probe(struct platform_device *pdev)
diff --git a/drivers/gpu/drm/lima/lima_gem.c b/drivers/gpu/drm/lima/lima_gem.c
index 4da21353c3a2..d0059d8c97d8 100644
--- a/drivers/gpu/drm/lima/lima_gem.c
+++ b/drivers/gpu/drm/lima/lima_gem.c
@@ -3,7 +3,7 @@
 
 #include <linux/mm.h>
 #include <linux/sync_file.h>
-#include <linux/pfn_t.h>
+#include <linux/pagemap.h>
 
 #include <drm/drm_file.h>
 #include <drm/drm_syncobj.h>
@@ -13,40 +13,55 @@
 
 #include "lima_drv.h"
 #include "lima_gem.h"
-#include "lima_gem_prime.h"
 #include "lima_vm.h"
-#include "lima_object.h"
 
 int lima_gem_create_handle(struct drm_device *dev, struct drm_file *file,
 			   u32 size, u32 flags, u32 *handle)
 {
 	int err;
-	struct lima_bo *bo;
-	struct lima_device *ldev = to_lima_dev(dev);
+	gfp_t mask;
+	struct drm_gem_shmem_object *shmem;
+	struct drm_gem_object *obj;
+	struct sg_table *sgt;
+
+	shmem = drm_gem_shmem_create(dev, size);
+	if (IS_ERR(shmem))
+		return PTR_ERR(shmem);
+
+	obj = &shmem->base;
 
-	bo = lima_bo_create(ldev, size, flags, NULL);
-	if (IS_ERR(bo))
-		return PTR_ERR(bo);
+	/* Mali Utgard GPU can only support 32bit address space */
+	mask = mapping_gfp_mask(obj->filp->f_mapping);
+	mask &= ~__GFP_HIGHMEM;
+	mask |= __GFP_DMA32;
+	mapping_set_gfp_mask(obj->filp->f_mapping, mask);
 
-	err = drm_gem_handle_create(file, &bo->gem, handle);
+	sgt = drm_gem_shmem_get_pages_sgt(obj);
+	if (IS_ERR(sgt)) {
+		err = PTR_ERR(sgt);
+		goto out;
+	}
+
+	err = drm_gem_handle_create(file, obj, handle);
 
+out:
 	/* drop reference from allocate - handle holds it now */
-	drm_gem_object_put_unlocked(&bo->gem);
+	drm_gem_object_put_unlocked(obj);
 
 	return err;
 }
 
-void lima_gem_free_object(struct drm_gem_object *obj)
+static void lima_gem_free_object(struct drm_gem_object *obj)
 {
 	struct lima_bo *bo = to_lima_bo(obj);
 
 	if (!list_empty(&bo->va))
 		dev_err(obj->dev->dev, "lima gem free bo still has va\n");
 
-	lima_bo_destroy(bo);
+	drm_gem_shmem_free_object(obj);
 }
 
-int lima_gem_object_open(struct drm_gem_object *obj, struct drm_file *file)
+static int lima_gem_object_open(struct drm_gem_object *obj, struct drm_file *file)
 {
 	struct lima_bo *bo = to_lima_bo(obj);
 	struct lima_drm_priv *priv = to_lima_drm_priv(file);
@@ -55,7 +70,7 @@ int lima_gem_object_open(struct drm_gem_object *obj, struct drm_file *file)
 	return lima_vm_bo_add(vm, bo, true);
 }
 
-void lima_gem_object_close(struct drm_gem_object *obj, struct drm_file *file)
+static void lima_gem_object_close(struct drm_gem_object *obj, struct drm_file *file)
 {
 	struct lima_bo *bo = to_lima_bo(obj);
 	struct lima_drm_priv *priv = to_lima_drm_priv(file);
@@ -64,13 +79,41 @@ void lima_gem_object_close(struct drm_gem_object *obj, struct drm_file *file)
 	lima_vm_bo_del(vm, bo);
 }
 
+static const struct drm_gem_object_funcs lima_gem_funcs = {
+	.free = lima_gem_free_object,
+	.open = lima_gem_object_open,
+	.close = lima_gem_object_close,
+	.print_info = drm_gem_shmem_print_info,
+	.pin = drm_gem_shmem_pin,
+	.unpin = drm_gem_shmem_unpin,
+	.get_sg_table = drm_gem_shmem_get_sg_table,
+	.vmap = drm_gem_shmem_vmap,
+	.vunmap = drm_gem_shmem_vunmap,
+	.mmap = drm_gem_shmem_mmap,
+};
+
+struct drm_gem_object *lima_gem_create_object(struct drm_device *dev, size_t size)
+{
+	struct lima_bo *bo;
+
+	bo = kzalloc(sizeof(*bo), GFP_KERNEL);
+	if (!bo)
+		return NULL;
+
+	mutex_init(&bo->lock);
+	INIT_LIST_HEAD(&bo->va);
+
+	bo->base.base.funcs = &lima_gem_funcs;
+
+	return &bo->base.base;
+}
+
 int lima_gem_get_info(struct drm_file *file, u32 handle, u32 *va, u64 *offset)
 {
 	struct drm_gem_object *obj;
 	struct lima_bo *bo;
 	struct lima_drm_priv *priv = to_lima_drm_priv(file);
 	struct lima_vm *vm = priv->vm;
-	int err;
 
 	obj = drm_gem_object_lookup(file, handle);
 	if (!obj)
@@ -80,53 +123,9 @@ int lima_gem_get_info(struct drm_file *file, u32 handle, u32 *va, u64 *offset)
 
 	*va = lima_vm_get_va(vm, bo);
 
-	err = drm_gem_create_mmap_offset(obj);
-	if (!err)
-		*offset = drm_vma_node_offset_addr(&obj->vma_node);
+	*offset = drm_vma_node_offset_addr(&obj->vma_node);
 
 	drm_gem_object_put_unlocked(obj);
-	return err;
-}
-
-static vm_fault_t lima_gem_fault(struct vm_fault *vmf)
-{
-	struct vm_area_struct *vma = vmf->vma;
-	struct drm_gem_object *obj = vma->vm_private_data;
-	struct lima_bo *bo = to_lima_bo(obj);
-	pfn_t pfn;
-	pgoff_t pgoff;
-
-	/* We don't use vmf->pgoff since that has the fake offset: */
-	pgoff = (vmf->address - vma->vm_start) >> PAGE_SHIFT;
-	pfn = __pfn_to_pfn_t(page_to_pfn(bo->pages[pgoff]), PFN_DEV);
-
-	return vmf_insert_mixed(vma, vmf->address, pfn);
-}
-
-const struct vm_operations_struct lima_gem_vm_ops = {
-	.fault = lima_gem_fault,
-	.open = drm_gem_vm_open,
-	.close = drm_gem_vm_close,
-};
-
-void lima_set_vma_flags(struct vm_area_struct *vma)
-{
-	pgprot_t prot = vm_get_page_prot(vma->vm_flags);
-
-	vma->vm_flags |= VM_MIXEDMAP;
-	vma->vm_flags &= ~VM_PFNMAP;
-	vma->vm_page_prot = pgprot_writecombine(prot);
-}
-
-int lima_gem_mmap(struct file *filp, struct vm_area_struct *vma)
-{
-	int ret;
-
-	ret = drm_gem_mmap(filp, vma);
-	if (ret)
-		return ret;
-
-	lima_set_vma_flags(vma);
 	return 0;
 }
 
@@ -136,7 +135,7 @@ static int lima_gem_sync_bo(struct lima_sched_task *task, struct lima_bo *bo,
 	int err = 0;
 
 	if (!write) {
-		err = dma_resv_reserve_shared(bo->gem.resv, 1);
+		err = dma_resv_reserve_shared(lima_bo_resv(bo), 1);
 		if (err)
 			return err;
 	}
@@ -145,62 +144,7 @@ static int lima_gem_sync_bo(struct lima_sched_task *task, struct lima_bo *bo,
 	if (explicit)
 		return 0;
 
-	return drm_gem_fence_array_add_implicit(&task->deps, &bo->gem, write);
-}
-
-static int lima_gem_lock_bos(struct lima_bo **bos, u32 nr_bos,
-			     struct ww_acquire_ctx *ctx)
-{
-	int i, ret = 0, contended, slow_locked = -1;
-
-	ww_acquire_init(ctx, &reservation_ww_class);
-
-retry:
-	for (i = 0; i < nr_bos; i++) {
-		if (i == slow_locked) {
-			slow_locked = -1;
-			continue;
-		}
-
-		ret = ww_mutex_lock_interruptible(&bos[i]->gem.resv->lock, ctx);
-		if (ret < 0) {
-			contended = i;
-			goto err;
-		}
-	}
-
-	ww_acquire_done(ctx);
-	return 0;
-
-err:
-	for (i--; i >= 0; i--)
-		ww_mutex_unlock(&bos[i]->gem.resv->lock);
-
-	if (slow_locked >= 0)
-		ww_mutex_unlock(&bos[slow_locked]->gem.resv->lock);
-
-	if (ret == -EDEADLK) {
-		/* we lost out in a seqno race, lock and retry.. */
-		ret = ww_mutex_lock_slow_interruptible(
-			&bos[contended]->gem.resv->lock, ctx);
-		if (!ret) {
-			slow_locked = contended;
-			goto retry;
-		}
-	}
-	ww_acquire_fini(ctx);
-
-	return ret;
-}
-
-static void lima_gem_unlock_bos(struct lima_bo **bos, u32 nr_bos,
-				struct ww_acquire_ctx *ctx)
-{
-	int i;
-
-	for (i = 0; i < nr_bos; i++)
-		ww_mutex_unlock(&bos[i]->gem.resv->lock);
-	ww_acquire_fini(ctx);
+	return drm_gem_fence_array_add_implicit(&task->deps, &bo->base.base, write);
 }
 
 static int lima_gem_add_deps(struct drm_file *file, struct lima_submit *submit)
@@ -268,7 +212,8 @@ int lima_gem_submit(struct drm_file *file, struct lima_submit *submit)
 		bos[i] = bo;
 	}
 
-	err = lima_gem_lock_bos(bos, submit->nr_bos, &ctx);
+	err = drm_gem_lock_reservations((struct drm_gem_object **)bos,
+					submit->nr_bos, &ctx);
 	if (err)
 		goto err_out0;
 
@@ -296,15 +241,16 @@ int lima_gem_submit(struct drm_file *file, struct lima_submit *submit)
 
 	for (i = 0; i < submit->nr_bos; i++) {
 		if (submit->bos[i].flags & LIMA_SUBMIT_BO_WRITE)
-			dma_resv_add_excl_fence(bos[i]->gem.resv, fence);
+			dma_resv_add_excl_fence(lima_bo_resv(bos[i]), fence);
 		else
-			dma_resv_add_shared_fence(bos[i]->gem.resv, fence);
+			dma_resv_add_shared_fence(lima_bo_resv(bos[i]), fence);
 	}
 
-	lima_gem_unlock_bos(bos, submit->nr_bos, &ctx);
+	drm_gem_unlock_reservations((struct drm_gem_object **)bos,
+				    submit->nr_bos, &ctx);
 
 	for (i = 0; i < submit->nr_bos; i++)
-		drm_gem_object_put_unlocked(&bos[i]->gem);
+		drm_gem_object_put_unlocked(&bos[i]->base.base);
 
 	if (out_sync) {
 		drm_syncobj_replace_fence(out_sync, fence);
@@ -318,13 +264,14 @@ int lima_gem_submit(struct drm_file *file, struct lima_submit *submit)
 err_out2:
 	lima_sched_task_fini(submit->task);
 err_out1:
-	lima_gem_unlock_bos(bos, submit->nr_bos, &ctx);
+	drm_gem_unlock_reservations((struct drm_gem_object **)bos,
+				    submit->nr_bos, &ctx);
 err_out0:
 	for (i = 0; i < submit->nr_bos; i++) {
 		if (!bos[i])
 			break;
 		lima_vm_bo_del(vm, bos[i]);
-		drm_gem_object_put_unlocked(&bos[i]->gem);
+		drm_gem_object_put_unlocked(&bos[i]->base.base);
 	}
 	if (out_sync)
 		drm_syncobj_put(out_sync);
diff --git a/drivers/gpu/drm/lima/lima_gem.h b/drivers/gpu/drm/lima/lima_gem.h
index 556111a01135..1800feb3e47f 100644
--- a/drivers/gpu/drm/lima/lima_gem.h
+++ b/drivers/gpu/drm/lima/lima_gem.h
@@ -4,19 +4,37 @@
 #ifndef __LIMA_GEM_H__
 #define __LIMA_GEM_H__
 
-struct lima_bo;
+#include <drm/drm_gem_shmem_helper.h>
+
 struct lima_submit;
 
-extern const struct vm_operations_struct lima_gem_vm_ops;
+struct lima_bo {
+	struct drm_gem_shmem_object base;
+
+	struct mutex lock;
+	struct list_head va;
+};
+
+static inline struct lima_bo *
+to_lima_bo(struct drm_gem_object *obj)
+{
+	return container_of(to_drm_gem_shmem_obj(obj), struct lima_bo, base);
+}
+
+static inline size_t lima_bo_size(struct lima_bo *bo)
+{
+	return bo->base.base.size;
+}
+
+static inline struct dma_resv *lima_bo_resv(struct lima_bo *bo)
+{
+	return bo->base.base.resv;
+}
 
-struct lima_bo *lima_gem_create_bo(struct drm_device *dev, u32 size, u32 flags);
+struct drm_gem_object *lima_gem_create_object(struct drm_device *dev, size_t size);
 int lima_gem_create_handle(struct drm_device *dev, struct drm_file *file,
 			   u32 size, u32 flags, u32 *handle);
-void lima_gem_free_object(struct drm_gem_object *obj);
-int lima_gem_object_open(struct drm_gem_object *obj, struct drm_file *file);
-void lima_gem_object_close(struct drm_gem_object *obj, struct drm_file *file);
 int lima_gem_get_info(struct drm_file *file, u32 handle, u32 *va, u64 *offset);
-int lima_gem_mmap(struct file *filp, struct vm_area_struct *vma);
 int lima_gem_submit(struct drm_file *file, struct lima_submit *submit);
 int lima_gem_wait(struct drm_file *file, u32 handle, u32 op, s64 timeout_ns);
 
diff --git a/drivers/gpu/drm/lima/lima_gem_prime.c b/drivers/gpu/drm/lima/lima_gem_prime.c
deleted file mode 100644
index e3eb251e0a12..000000000000
--- a/drivers/gpu/drm/lima/lima_gem_prime.c
+++ /dev/null
@@ -1,46 +0,0 @@
-// SPDX-License-Identifier: GPL-2.0 OR MIT
-/* Copyright 2018-2019 Qiang Yu <yuq825@gmail.com> */
-
-#include <linux/dma-buf.h>
-#include <drm/drm_prime.h>
-#include <drm/drm_drv.h>
-#include <drm/drm_file.h>
-
-#include "lima_device.h"
-#include "lima_object.h"
-#include "lima_gem.h"
-#include "lima_gem_prime.h"
-
-struct drm_gem_object *lima_gem_prime_import_sg_table(
-	struct drm_device *dev, struct dma_buf_attachment *attach,
-	struct sg_table *sgt)
-{
-	struct lima_device *ldev = to_lima_dev(dev);
-	struct lima_bo *bo;
-
-	bo = lima_bo_create(ldev, attach->dmabuf->size, 0, sgt);
-	if (IS_ERR(bo))
-		return ERR_CAST(bo);
-
-	return &bo->gem;
-}
-
-struct sg_table *lima_gem_prime_get_sg_table(struct drm_gem_object *obj)
-{
-	struct lima_bo *bo = to_lima_bo(obj);
-	int npages = obj->size >> PAGE_SHIFT;
-
-	return drm_prime_pages_to_sg(bo->pages, npages);
-}
-
-int lima_gem_prime_mmap(struct drm_gem_object *obj, struct vm_area_struct *vma)
-{
-	int ret;
-
-	ret = drm_gem_mmap_obj(obj, obj->size, vma);
-	if (ret)
-		return ret;
-
-	lima_set_vma_flags(vma);
-	return 0;
-}
diff --git a/drivers/gpu/drm/lima/lima_gem_prime.h b/drivers/gpu/drm/lima/lima_gem_prime.h
deleted file mode 100644
index 34b4d35c21e3..000000000000
--- a/drivers/gpu/drm/lima/lima_gem_prime.h
+++ /dev/null
@@ -1,13 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0 OR MIT */
-/* Copyright 2018-2019 Qiang Yu <yuq825@gmail.com> */
-
-#ifndef __LIMA_GEM_PRIME_H__
-#define __LIMA_GEM_PRIME_H__
-
-struct drm_gem_object *lima_gem_prime_import_sg_table(
-	struct drm_device *dev, struct dma_buf_attachment *attach,
-	struct sg_table *sgt);
-struct sg_table *lima_gem_prime_get_sg_table(struct drm_gem_object *obj);
-int lima_gem_prime_mmap(struct drm_gem_object *obj, struct vm_area_struct *vma);
-
-#endif
diff --git a/drivers/gpu/drm/lima/lima_mmu.c b/drivers/gpu/drm/lima/lima_mmu.c
index 8e1651d6a61f..97ec09dee572 100644
--- a/drivers/gpu/drm/lima/lima_mmu.c
+++ b/drivers/gpu/drm/lima/lima_mmu.c
@@ -8,7 +8,6 @@
 #include "lima_device.h"
 #include "lima_mmu.h"
 #include "lima_vm.h"
-#include "lima_object.h"
 #include "lima_regs.h"
 
 #define mmu_write(reg, data) writel(data, ip->iomem + reg)
diff --git a/drivers/gpu/drm/lima/lima_object.c b/drivers/gpu/drm/lima/lima_object.c
deleted file mode 100644
index 87123b1d083c..000000000000
--- a/drivers/gpu/drm/lima/lima_object.c
+++ /dev/null
@@ -1,119 +0,0 @@
-// SPDX-License-Identifier: GPL-2.0 OR MIT
-/* Copyright 2018-2019 Qiang Yu <yuq825@gmail.com> */
-
-#include <drm/drm_prime.h>
-#include <linux/pagemap.h>
-#include <linux/dma-mapping.h>
-
-#include "lima_object.h"
-
-void lima_bo_destroy(struct lima_bo *bo)
-{
-	if (bo->sgt) {
-		kfree(bo->pages);
-		drm_prime_gem_destroy(&bo->gem, bo->sgt);
-	} else {
-		if (bo->pages_dma_addr) {
-			int i, npages = bo->gem.size >> PAGE_SHIFT;
-
-			for (i = 0; i < npages; i++) {
-				if (bo->pages_dma_addr[i])
-					dma_unmap_page(bo->gem.dev->dev,
-						       bo->pages_dma_addr[i],
-						       PAGE_SIZE, DMA_BIDIRECTIONAL);
-			}
-		}
-
-		if (bo->pages)
-			drm_gem_put_pages(&bo->gem, bo->pages, true, true);
-	}
-
-	kfree(bo->pages_dma_addr);
-	drm_gem_object_release(&bo->gem);
-	kfree(bo);
-}
-
-static struct lima_bo *lima_bo_create_struct(struct lima_device *dev, u32 size, u32 flags)
-{
-	struct lima_bo *bo;
-	int err;
-
-	size = PAGE_ALIGN(size);
-
-	bo = kzalloc(sizeof(*bo), GFP_KERNEL);
-	if (!bo)
-		return ERR_PTR(-ENOMEM);
-
-	mutex_init(&bo->lock);
-	INIT_LIST_HEAD(&bo->va);
-
-	err = drm_gem_object_init(dev->ddev, &bo->gem, size);
-	if (err) {
-		kfree(bo);
-		return ERR_PTR(err);
-	}
-
-	return bo;
-}
-
-struct lima_bo *lima_bo_create(struct lima_device *dev, u32 size,
-			       u32 flags, struct sg_table *sgt)
-{
-	int i, err;
-	size_t npages;
-	struct lima_bo *bo, *ret;
-
-	bo = lima_bo_create_struct(dev, size, flags);
-	if (IS_ERR(bo))
-		return bo;
-
-	npages = bo->gem.size >> PAGE_SHIFT;
-
-	bo->pages_dma_addr = kcalloc(npages, sizeof(dma_addr_t), GFP_KERNEL);
-	if (!bo->pages_dma_addr) {
-		ret = ERR_PTR(-ENOMEM);
-		goto err_out;
-	}
-
-	if (sgt) {
-		bo->sgt = sgt;
-
-		bo->pages = kcalloc(npages, sizeof(*bo->pages), GFP_KERNEL);
-		if (!bo->pages) {
-			ret = ERR_PTR(-ENOMEM);
-			goto err_out;
-		}
-
-		err = drm_prime_sg_to_page_addr_arrays(
-			sgt, bo->pages, bo->pages_dma_addr, npages);
-		if (err) {
-			ret = ERR_PTR(err);
-			goto err_out;
-		}
-	} else {
-		mapping_set_gfp_mask(bo->gem.filp->f_mapping, GFP_DMA32);
-		bo->pages = drm_gem_get_pages(&bo->gem);
-		if (IS_ERR(bo->pages)) {
-			ret = ERR_CAST(bo->pages);
-			bo->pages = NULL;
-			goto err_out;
-		}
-
-		for (i = 0; i < npages; i++) {
-			dma_addr_t addr = dma_map_page(dev->dev, bo->pages[i], 0,
-						       PAGE_SIZE, DMA_BIDIRECTIONAL);
-			if (dma_mapping_error(dev->dev, addr)) {
-				ret = ERR_PTR(-EFAULT);
-				goto err_out;
-			}
-			bo->pages_dma_addr[i] = addr;
-		}
-
-	}
-
-	return bo;
-
-err_out:
-	lima_bo_destroy(bo);
-	return ret;
-}
diff --git a/drivers/gpu/drm/lima/lima_object.h b/drivers/gpu/drm/lima/lima_object.h
deleted file mode 100644
index 31ca2d8dc0a1..000000000000
--- a/drivers/gpu/drm/lima/lima_object.h
+++ /dev/null
@@ -1,35 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0 OR MIT */
-/* Copyright 2018-2019 Qiang Yu <yuq825@gmail.com> */
-
-#ifndef __LIMA_OBJECT_H__
-#define __LIMA_OBJECT_H__
-
-#include <drm/drm_gem.h>
-
-#include "lima_device.h"
-
-struct lima_bo {
-	struct drm_gem_object gem;
-
-	struct page **pages;
-	dma_addr_t *pages_dma_addr;
-	struct sg_table *sgt;
-	void *vaddr;
-
-	struct mutex lock;
-	struct list_head va;
-};
-
-static inline struct lima_bo *
-to_lima_bo(struct drm_gem_object *obj)
-{
-	return container_of(obj, struct lima_bo, gem);
-}
-
-struct lima_bo *lima_bo_create(struct lima_device *dev, u32 size,
-			       u32 flags, struct sg_table *sgt);
-void lima_bo_destroy(struct lima_bo *bo);
-void *lima_bo_vmap(struct lima_bo *bo);
-void lima_bo_vunmap(struct lima_bo *bo);
-
-#endif
diff --git a/drivers/gpu/drm/lima/lima_sched.c b/drivers/gpu/drm/lima/lima_sched.c
index 4127cacac454..f522c5f99729 100644
--- a/drivers/gpu/drm/lima/lima_sched.c
+++ b/drivers/gpu/drm/lima/lima_sched.c
@@ -10,7 +10,7 @@
 #include "lima_vm.h"
 #include "lima_mmu.h"
 #include "lima_l2_cache.h"
-#include "lima_object.h"
+#include "lima_gem.h"
 
 struct lima_fence {
 	struct dma_fence base;
@@ -117,7 +117,7 @@ int lima_sched_task_init(struct lima_sched_task *task,
 		return -ENOMEM;
 
 	for (i = 0; i < num_bos; i++)
-		drm_gem_object_get(&bos[i]->gem);
+		drm_gem_object_get(&bos[i]->base.base);
 
 	err = drm_sched_job_init(&task->base, &context->base, vm);
 	if (err) {
@@ -148,7 +148,7 @@ void lima_sched_task_fini(struct lima_sched_task *task)
 
 	if (task->bos) {
 		for (i = 0; i < task->num_bos; i++)
-			drm_gem_object_put_unlocked(&task->bos[i]->gem);
+			drm_gem_object_put_unlocked(&task->bos[i]->base.base);
 		kfree(task->bos);
 	}
 
diff --git a/drivers/gpu/drm/lima/lima_vm.c b/drivers/gpu/drm/lima/lima_vm.c
index 19e88ca16527..840e2350d872 100644
--- a/drivers/gpu/drm/lima/lima_vm.c
+++ b/drivers/gpu/drm/lima/lima_vm.c
@@ -6,7 +6,7 @@
 
 #include "lima_device.h"
 #include "lima_vm.h"
-#include "lima_object.h"
+#include "lima_gem.h"
 #include "lima_regs.h"
 
 struct lima_bo_va {
@@ -32,7 +32,7 @@ struct lima_bo_va {
 #define LIMA_BTE(va) ((va & LIMA_VM_BT_MASK) >> LIMA_VM_BT_SHIFT)
 
 
-static void lima_vm_unmap_page_table(struct lima_vm *vm, u32 start, u32 end)
+static void lima_vm_unmap_range(struct lima_vm *vm, u32 start, u32 end)
 {
 	u32 addr;
 
@@ -44,41 +44,32 @@ static void lima_vm_unmap_page_table(struct lima_vm *vm, u32 start, u32 end)
 	}
 }
 
-static int lima_vm_map_page_table(struct lima_vm *vm, dma_addr_t *dma,
-				  u32 start, u32 end)
+static int lima_vm_map_page(struct lima_vm *vm, dma_addr_t pa, u32 va)
 {
-	u64 addr;
-	int i = 0;
-
-	for (addr = start; addr <= end; addr += LIMA_PAGE_SIZE) {
-		u32 pbe = LIMA_PBE(addr);
-		u32 bte = LIMA_BTE(addr);
-
-		if (!vm->bts[pbe].cpu) {
-			dma_addr_t pts;
-			u32 *pd;
-			int j;
-
-			vm->bts[pbe].cpu = dma_alloc_wc(
-				vm->dev->dev, LIMA_PAGE_SIZE << LIMA_VM_NUM_PT_PER_BT_SHIFT,
-				&vm->bts[pbe].dma, GFP_KERNEL | __GFP_ZERO);
-			if (!vm->bts[pbe].cpu) {
-				if (addr != start)
-					lima_vm_unmap_page_table(vm, start, addr - 1);
-				return -ENOMEM;
-			}
-
-			pts = vm->bts[pbe].dma;
-			pd = vm->pd.cpu + (pbe << LIMA_VM_NUM_PT_PER_BT_SHIFT);
-			for (j = 0; j < LIMA_VM_NUM_PT_PER_BT; j++) {
-				pd[j] = pts | LIMA_VM_FLAG_PRESENT;
-				pts += LIMA_PAGE_SIZE;
-			}
+	u32 pbe = LIMA_PBE(va);
+	u32 bte = LIMA_BTE(va);
+
+	if (!vm->bts[pbe].cpu) {
+		dma_addr_t pts;
+		u32 *pd;
+		int j;
+
+		vm->bts[pbe].cpu = dma_alloc_wc(
+			vm->dev->dev, LIMA_PAGE_SIZE << LIMA_VM_NUM_PT_PER_BT_SHIFT,
+			&vm->bts[pbe].dma, GFP_KERNEL | __GFP_NOWARN | __GFP_ZERO);
+		if (!vm->bts[pbe].cpu)
+			return -ENOMEM;
+
+		pts = vm->bts[pbe].dma;
+		pd = vm->pd.cpu + (pbe << LIMA_VM_NUM_PT_PER_BT_SHIFT);
+		for (j = 0; j < LIMA_VM_NUM_PT_PER_BT; j++) {
+			pd[j] = pts | LIMA_VM_FLAG_PRESENT;
+			pts += LIMA_PAGE_SIZE;
 		}
-
-		vm->bts[pbe].cpu[bte] = dma[i++] | LIMA_VM_FLAGS_CACHE;
 	}
 
+	vm->bts[pbe].cpu[bte] = pa | LIMA_VM_FLAGS_CACHE;
+
 	return 0;
 }
 
@@ -100,7 +91,8 @@ lima_vm_bo_find(struct lima_vm *vm, struct lima_bo *bo)
 int lima_vm_bo_add(struct lima_vm *vm, struct lima_bo *bo, bool create)
 {
 	struct lima_bo_va *bo_va;
-	int err;
+	struct sg_dma_page_iter sg_iter;
+	int offset = 0, err;
 
 	mutex_lock(&bo->lock);
 
@@ -128,14 +120,18 @@ int lima_vm_bo_add(struct lima_vm *vm, struct lima_bo *bo, bool create)
 
 	mutex_lock(&vm->lock);
 
-	err = drm_mm_insert_node(&vm->mm, &bo_va->node, bo->gem.size);
+	err = drm_mm_insert_node(&vm->mm, &bo_va->node, lima_bo_size(bo));
 	if (err)
 		goto err_out1;
 
-	err = lima_vm_map_page_table(vm, bo->pages_dma_addr, bo_va->node.start,
-				     bo_va->node.start + bo_va->node.size - 1);
-	if (err)
-		goto err_out2;
+	for_each_sg_dma_page(bo->base.sgt->sgl, &sg_iter, bo->base.sgt->nents, 0) {
+		err = lima_vm_map_page(vm, sg_page_iter_dma_address(&sg_iter),
+				       bo_va->node.start + offset);
+		if (err)
+			goto err_out2;
+
+		offset += PAGE_SIZE;
+	}
 
 	mutex_unlock(&vm->lock);
 
@@ -145,6 +141,8 @@ int lima_vm_bo_add(struct lima_vm *vm, struct lima_bo *bo, bool create)
 	return 0;
 
 err_out2:
+	if (offset)
+		lima_vm_unmap_range(vm, bo_va->node.start, bo_va->node.start + offset - 1);
 	drm_mm_remove_node(&bo_va->node);
 err_out1:
 	mutex_unlock(&vm->lock);
@@ -168,8 +166,8 @@ void lima_vm_bo_del(struct lima_vm *vm, struct lima_bo *bo)
 
 	mutex_lock(&vm->lock);
 
-	lima_vm_unmap_page_table(vm, bo_va->node.start,
-				 bo_va->node.start + bo_va->node.size - 1);
+	lima_vm_unmap_range(vm, bo_va->node.start,
+			    bo_va->node.start + bo_va->node.size - 1);
 
 	drm_mm_remove_node(&bo_va->node);
 
@@ -210,14 +208,13 @@ struct lima_vm *lima_vm_create(struct lima_device *dev)
 	kref_init(&vm->refcount);
 
 	vm->pd.cpu = dma_alloc_wc(dev->dev, LIMA_PAGE_SIZE, &vm->pd.dma,
-				  GFP_KERNEL | __GFP_ZERO);
+				  GFP_KERNEL | __GFP_NOWARN | __GFP_ZERO);
 	if (!vm->pd.cpu)
 		goto err_out0;
 
 	if (dev->dlbu_cpu) {
-		int err = lima_vm_map_page_table(
-			vm, &dev->dlbu_dma, LIMA_VA_RESERVE_DLBU,
-			LIMA_VA_RESERVE_DLBU + LIMA_PAGE_SIZE - 1);
+		int err = lima_vm_map_page(
+			vm, dev->dlbu_dma, LIMA_VA_RESERVE_DLBU);
 		if (err)
 			goto err_out1;
 	}
diff --git a/drivers/gpu/drm/mcde/mcde_drv.c b/drivers/gpu/drm/mcde/mcde_drv.c
index 9a09eba53182..5649887d2b90 100644
--- a/drivers/gpu/drm/mcde/mcde_drv.c
+++ b/drivers/gpu/drm/mcde/mcde_drv.c
@@ -484,7 +484,8 @@ static int mcde_probe(struct platform_device *pdev)
 	}
 	if (!match) {
 		dev_err(dev, "no matching components\n");
-		return -ENODEV;
+		ret = -ENODEV;
+		goto clk_disable;
 	}
 	if (IS_ERR(match)) {
 		dev_err(dev, "could not create component match\n");
diff --git a/drivers/gpu/drm/mcde/mcde_dsi.c b/drivers/gpu/drm/mcde/mcde_dsi.c
index f9c9e32b299c..d6214d3c8b33 100644
--- a/drivers/gpu/drm/mcde/mcde_dsi.c
+++ b/drivers/gpu/drm/mcde/mcde_dsi.c
@@ -946,8 +946,8 @@ static int mcde_dsi_bind(struct device *dev, struct device *master,
 		}
 	}
 	if (panel) {
-		bridge = drm_panel_bridge_add(panel,
-					      DRM_MODE_CONNECTOR_DSI);
+		bridge = drm_panel_bridge_add_typed(panel,
+						    DRM_MODE_CONNECTOR_DSI);
 		if (IS_ERR(bridge)) {
 			dev_err(dev, "error adding panel bridge\n");
 			return PTR_ERR(bridge);
diff --git a/drivers/gpu/drm/mediatek/Makefile b/drivers/gpu/drm/mediatek/Makefile
index 82ae49c64221..8067a4be8311 100644
--- a/drivers/gpu/drm/mediatek/Makefile
+++ b/drivers/gpu/drm/mediatek/Makefile
@@ -12,6 +12,8 @@ mediatek-drm-y := mtk_disp_color.o \
 		  mtk_drm_plane.o \
 		  mtk_dsi.o \
 		  mtk_mipi_tx.o \
+		  mtk_mt8173_mipi_tx.o \
+		  mtk_mt8183_mipi_tx.o \
 		  mtk_dpi.o
 
 obj-$(CONFIG_DRM_MEDIATEK) += mediatek-drm.o
diff --git a/drivers/gpu/drm/mediatek/mtk_disp_ovl.c b/drivers/gpu/drm/mediatek/mtk_disp_ovl.c
index 21851756c579..4a55bb6e2213 100644
--- a/drivers/gpu/drm/mediatek/mtk_disp_ovl.c
+++ b/drivers/gpu/drm/mediatek/mtk_disp_ovl.c
@@ -3,6 +3,8 @@
  * Copyright (c) 2015 MediaTek Inc.
  */
 
+#include <drm/drm_fourcc.h>
+
 #include <linux/clk.h>
 #include <linux/component.h>
 #include <linux/module.h>
@@ -19,6 +21,8 @@
 #define DISP_REG_OVL_EN				0x000c
 #define DISP_REG_OVL_RST			0x0014
 #define DISP_REG_OVL_ROI_SIZE			0x0020
+#define DISP_REG_OVL_DATAPATH_CON		0x0024
+#define OVL_BGCLR_SEL_IN				BIT(2)
 #define DISP_REG_OVL_ROI_BGCLR			0x0028
 #define DISP_REG_OVL_SRC_CON			0x002c
 #define DISP_REG_OVL_CON(n)			(0x0030 + 0x20 * (n))
@@ -31,7 +35,9 @@
 #define DISP_REG_OVL_ADDR_MT8173		0x0f40
 #define DISP_REG_OVL_ADDR(ovl, n)		((ovl)->data->addr + 0x20 * (n))
 
-#define	OVL_RDMA_MEM_GMC	0x40402020
+#define GMC_THRESHOLD_BITS	16
+#define GMC_THRESHOLD_HIGH	((1 << GMC_THRESHOLD_BITS) / 4)
+#define GMC_THRESHOLD_LOW	((1 << GMC_THRESHOLD_BITS) / 8)
 
 #define OVL_CON_BYTE_SWAP	BIT(24)
 #define OVL_CON_MTX_YUV_TO_RGB	(6 << 16)
@@ -46,9 +52,13 @@
 					OVL_CON_CLRFMT_RGB : 0)
 #define	OVL_CON_AEN		BIT(8)
 #define	OVL_CON_ALPHA		0xff
+#define	OVL_CON_VIRT_FLIP	BIT(9)
+#define	OVL_CON_HORZ_FLIP	BIT(10)
 
 struct mtk_disp_ovl_data {
 	unsigned int addr;
+	unsigned int gmc_bits;
+	unsigned int layer_nr;
 	bool fmt_rgb565_is_0;
 };
 
@@ -126,15 +136,65 @@ static void mtk_ovl_config(struct mtk_ddp_comp *comp, unsigned int w,
 
 static unsigned int mtk_ovl_layer_nr(struct mtk_ddp_comp *comp)
 {
-	return 4;
+	struct mtk_disp_ovl *ovl = comp_to_ovl(comp);
+
+	return ovl->data->layer_nr;
+}
+
+static unsigned int mtk_ovl_supported_rotations(struct mtk_ddp_comp *comp)
+{
+	return DRM_MODE_ROTATE_0 | DRM_MODE_ROTATE_180 |
+	       DRM_MODE_REFLECT_X | DRM_MODE_REFLECT_Y;
+}
+
+static int mtk_ovl_layer_check(struct mtk_ddp_comp *comp, unsigned int idx,
+			       struct mtk_plane_state *mtk_state)
+{
+	struct drm_plane_state *state = &mtk_state->base;
+	unsigned int rotation = 0;
+
+	rotation = drm_rotation_simplify(state->rotation,
+					 DRM_MODE_ROTATE_0 |
+					 DRM_MODE_REFLECT_X |
+					 DRM_MODE_REFLECT_Y);
+	rotation &= ~DRM_MODE_ROTATE_0;
+
+	/* We can only do reflection, not rotation */
+	if ((rotation & DRM_MODE_ROTATE_MASK) != 0)
+		return -EINVAL;
+
+	/*
+	 * TODO: Rotating/reflecting YUV buffers is not supported at this time.
+	 *	 Only RGB[AX] variants are supported.
+	 */
+	if (state->fb->format->is_yuv && rotation != 0)
+		return -EINVAL;
+
+	state->rotation = rotation;
+
+	return 0;
 }
 
 static void mtk_ovl_layer_on(struct mtk_ddp_comp *comp, unsigned int idx)
 {
 	unsigned int reg;
+	unsigned int gmc_thrshd_l;
+	unsigned int gmc_thrshd_h;
+	unsigned int gmc_value;
+	struct mtk_disp_ovl *ovl = comp_to_ovl(comp);
 
 	writel(0x1, comp->regs + DISP_REG_OVL_RDMA_CTRL(idx));
-	writel(OVL_RDMA_MEM_GMC, comp->regs + DISP_REG_OVL_RDMA_GMC(idx));
+
+	gmc_thrshd_l = GMC_THRESHOLD_LOW >>
+		      (GMC_THRESHOLD_BITS - ovl->data->gmc_bits);
+	gmc_thrshd_h = GMC_THRESHOLD_HIGH >>
+		      (GMC_THRESHOLD_BITS - ovl->data->gmc_bits);
+	if (ovl->data->gmc_bits == 10)
+		gmc_value = gmc_thrshd_h | gmc_thrshd_h << 16;
+	else
+		gmc_value = gmc_thrshd_l | gmc_thrshd_l << 8 |
+			    gmc_thrshd_h << 16 | gmc_thrshd_h << 24;
+	writel(gmc_value, comp->regs + DISP_REG_OVL_RDMA_GMC(idx));
 
 	reg = readl(comp->regs + DISP_REG_OVL_SRC_CON);
 	reg = reg | BIT(idx);
@@ -207,6 +267,16 @@ static void mtk_ovl_layer_config(struct mtk_ddp_comp *comp, unsigned int idx,
 	if (idx != 0)
 		con |= OVL_CON_AEN | OVL_CON_ALPHA;
 
+	if (pending->rotation & DRM_MODE_REFLECT_Y) {
+		con |= OVL_CON_VIRT_FLIP;
+		addr += (pending->height - 1) * pending->pitch;
+	}
+
+	if (pending->rotation & DRM_MODE_REFLECT_X) {
+		con |= OVL_CON_HORZ_FLIP;
+		addr += pending->pitch - 1;
+	}
+
 	writel_relaxed(con, comp->regs + DISP_REG_OVL_CON(idx));
 	writel_relaxed(pitch, comp->regs + DISP_REG_OVL_PITCH(idx));
 	writel_relaxed(src_size, comp->regs + DISP_REG_OVL_SRC_SIZE(idx));
@@ -217,16 +287,38 @@ static void mtk_ovl_layer_config(struct mtk_ddp_comp *comp, unsigned int idx,
 		mtk_ovl_layer_on(comp, idx);
 }
 
+static void mtk_ovl_bgclr_in_on(struct mtk_ddp_comp *comp)
+{
+	unsigned int reg;
+
+	reg = readl(comp->regs + DISP_REG_OVL_DATAPATH_CON);
+	reg = reg | OVL_BGCLR_SEL_IN;
+	writel(reg, comp->regs + DISP_REG_OVL_DATAPATH_CON);
+}
+
+static void mtk_ovl_bgclr_in_off(struct mtk_ddp_comp *comp)
+{
+	unsigned int reg;
+
+	reg = readl(comp->regs + DISP_REG_OVL_DATAPATH_CON);
+	reg = reg & ~OVL_BGCLR_SEL_IN;
+	writel(reg, comp->regs + DISP_REG_OVL_DATAPATH_CON);
+}
+
 static const struct mtk_ddp_comp_funcs mtk_disp_ovl_funcs = {
 	.config = mtk_ovl_config,
 	.start = mtk_ovl_start,
 	.stop = mtk_ovl_stop,
 	.enable_vblank = mtk_ovl_enable_vblank,
 	.disable_vblank = mtk_ovl_disable_vblank,
+	.supported_rotations = mtk_ovl_supported_rotations,
 	.layer_nr = mtk_ovl_layer_nr,
 	.layer_on = mtk_ovl_layer_on,
 	.layer_off = mtk_ovl_layer_off,
+	.layer_check = mtk_ovl_layer_check,
 	.layer_config = mtk_ovl_layer_config,
+	.bgclr_in_on = mtk_ovl_bgclr_in_on,
+	.bgclr_in_off = mtk_ovl_bgclr_in_off,
 };
 
 static int mtk_disp_ovl_bind(struct device *dev, struct device *master,
@@ -276,7 +368,12 @@ static int mtk_disp_ovl_probe(struct platform_device *pdev)
 	if (irq < 0)
 		return irq;
 
-	comp_id = mtk_ddp_comp_get_id(dev->of_node, MTK_DISP_OVL);
+	priv->data = of_device_get_match_data(dev);
+
+	comp_id = mtk_ddp_comp_get_id(dev->of_node,
+				      priv->data->layer_nr == 4 ?
+				      MTK_DISP_OVL :
+				      MTK_DISP_OVL_2L);
 	if (comp_id < 0) {
 		dev_err(dev, "Failed to identify by alias: %d\n", comp_id);
 		return comp_id;
@@ -289,8 +386,6 @@ static int mtk_disp_ovl_probe(struct platform_device *pdev)
 		return ret;
 	}
 
-	priv->data = of_device_get_match_data(dev);
-
 	platform_set_drvdata(pdev, priv);
 
 	ret = devm_request_irq(dev, irq, mtk_disp_ovl_irq_handler,
@@ -316,11 +411,15 @@ static int mtk_disp_ovl_remove(struct platform_device *pdev)
 
 static const struct mtk_disp_ovl_data mt2701_ovl_driver_data = {
 	.addr = DISP_REG_OVL_ADDR_MT2701,
+	.gmc_bits = 8,
+	.layer_nr = 4,
 	.fmt_rgb565_is_0 = false,
 };
 
 static const struct mtk_disp_ovl_data mt8173_ovl_driver_data = {
 	.addr = DISP_REG_OVL_ADDR_MT8173,
+	.gmc_bits = 8,
+	.layer_nr = 4,
 	.fmt_rgb565_is_0 = true,
 };
 
diff --git a/drivers/gpu/drm/mediatek/mtk_dpi.c b/drivers/gpu/drm/mediatek/mtk_dpi.c
index be6d95c5ff25..01fa8b8d763d 100644
--- a/drivers/gpu/drm/mediatek/mtk_dpi.c
+++ b/drivers/gpu/drm/mediatek/mtk_dpi.c
@@ -17,6 +17,7 @@
 #include <video/videomode.h>
 
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_crtc.h>
 #include <drm/drm_of.h>
 
diff --git a/drivers/gpu/drm/mediatek/mtk_drm_crtc.c b/drivers/gpu/drm/mediatek/mtk_drm_crtc.c
index 34a731755791..f80a8ba75977 100644
--- a/drivers/gpu/drm/mediatek/mtk_drm_crtc.c
+++ b/drivers/gpu/drm/mediatek/mtk_drm_crtc.c
@@ -207,6 +207,28 @@ static void mtk_crtc_ddp_clk_disable(struct mtk_drm_crtc *mtk_crtc)
 		clk_disable_unprepare(mtk_crtc->ddp_comp[i]->clk);
 }
 
+static
+struct mtk_ddp_comp *mtk_drm_ddp_comp_for_plane(struct drm_crtc *crtc,
+						struct drm_plane *plane,
+						unsigned int *local_layer)
+{
+	struct mtk_drm_crtc *mtk_crtc = to_mtk_crtc(crtc);
+	struct mtk_ddp_comp *comp;
+	int i, count = 0;
+
+	for (i = 0; i < mtk_crtc->ddp_comp_nr; i++) {
+		comp = mtk_crtc->ddp_comp[i];
+		if (plane->index < (count + mtk_ddp_comp_layer_nr(comp))) {
+			*local_layer = plane->index - count;
+			return comp;
+		}
+		count += mtk_ddp_comp_layer_nr(comp);
+	}
+
+	WARN(1, "Failed to find component for plane %d\n", plane->index);
+	return NULL;
+}
+
 static int mtk_crtc_ddp_hw_init(struct mtk_drm_crtc *mtk_crtc)
 {
 	struct drm_crtc *crtc = &mtk_crtc->base;
@@ -272,6 +294,9 @@ static int mtk_crtc_ddp_hw_init(struct mtk_drm_crtc *mtk_crtc)
 	for (i = 0; i < mtk_crtc->ddp_comp_nr; i++) {
 		struct mtk_ddp_comp *comp = mtk_crtc->ddp_comp[i];
 
+		if (i == 1)
+			mtk_ddp_comp_bgclr_in_on(comp);
+
 		mtk_ddp_comp_config(comp, width, height, vrefresh, bpc);
 		mtk_ddp_comp_start(comp);
 	}
@@ -280,10 +305,12 @@ static int mtk_crtc_ddp_hw_init(struct mtk_drm_crtc *mtk_crtc)
 	for (i = 0; i < mtk_crtc->layer_nr; i++) {
 		struct drm_plane *plane = &mtk_crtc->planes[i];
 		struct mtk_plane_state *plane_state;
+		struct mtk_ddp_comp *comp;
+		unsigned int local_layer;
 
 		plane_state = to_mtk_plane_state(plane->state);
-		mtk_ddp_comp_layer_config(mtk_crtc->ddp_comp[0], i,
-					  plane_state);
+		comp = mtk_drm_ddp_comp_for_plane(crtc, plane, &local_layer);
+		mtk_ddp_comp_layer_config(comp, local_layer, plane_state);
 	}
 
 	return 0;
@@ -301,8 +328,12 @@ static void mtk_crtc_ddp_hw_fini(struct mtk_drm_crtc *mtk_crtc)
 	int i;
 
 	DRM_DEBUG_DRIVER("%s\n", __func__);
-	for (i = 0; i < mtk_crtc->ddp_comp_nr; i++)
+	for (i = 0; i < mtk_crtc->ddp_comp_nr; i++) {
 		mtk_ddp_comp_stop(mtk_crtc->ddp_comp[i]);
+		if (i == 1)
+			mtk_ddp_comp_bgclr_in_off(mtk_crtc->ddp_comp[i]);
+	}
+
 	for (i = 0; i < mtk_crtc->ddp_comp_nr; i++)
 		mtk_disp_mutex_remove_comp(mtk_crtc->mutex,
 					   mtk_crtc->ddp_comp[i]->id);
@@ -327,6 +358,7 @@ static void mtk_crtc_ddp_config(struct drm_crtc *crtc)
 	struct mtk_crtc_state *state = to_mtk_crtc_state(mtk_crtc->base.state);
 	struct mtk_ddp_comp *comp = mtk_crtc->ddp_comp[0];
 	unsigned int i;
+	unsigned int local_layer;
 
 	/*
 	 * TODO: instead of updating the registers here, we should prepare
@@ -348,15 +380,30 @@ static void mtk_crtc_ddp_config(struct drm_crtc *crtc)
 
 			plane_state = to_mtk_plane_state(plane->state);
 
-			if (plane_state->pending.config) {
-				mtk_ddp_comp_layer_config(comp, i, plane_state);
-				plane_state->pending.config = false;
-			}
+			if (!plane_state->pending.config)
+				continue;
+
+			comp = mtk_drm_ddp_comp_for_plane(crtc, plane,
+							  &local_layer);
+
+			mtk_ddp_comp_layer_config(comp, local_layer,
+						  plane_state);
+			plane_state->pending.config = false;
 		}
 		mtk_crtc->pending_planes = false;
 	}
 }
 
+int mtk_drm_crtc_plane_check(struct drm_crtc *crtc, struct drm_plane *plane,
+			     struct mtk_plane_state *state)
+{
+	unsigned int local_layer;
+	struct mtk_ddp_comp *comp;
+
+	comp = mtk_drm_ddp_comp_for_plane(crtc, plane, &local_layer);
+	return mtk_ddp_comp_layer_check(comp, local_layer, state);
+}
+
 static void mtk_drm_crtc_atomic_enable(struct drm_crtc *crtc,
 				       struct drm_crtc_state *old_state)
 {
@@ -518,14 +565,65 @@ void mtk_crtc_ddp_irq(struct drm_crtc *crtc, struct mtk_ddp_comp *comp)
 	mtk_drm_finish_page_flip(mtk_crtc);
 }
 
+static int mtk_drm_crtc_num_comp_planes(struct mtk_drm_crtc *mtk_crtc,
+					int comp_idx)
+{
+	struct mtk_ddp_comp *comp;
+
+	if (comp_idx > 1)
+		return 0;
+
+	comp = mtk_crtc->ddp_comp[comp_idx];
+	if (!comp->funcs)
+		return 0;
+
+	if (comp_idx == 1 && !comp->funcs->bgclr_in_on)
+		return 0;
+
+	return mtk_ddp_comp_layer_nr(comp);
+}
+
+static inline
+enum drm_plane_type mtk_drm_crtc_plane_type(unsigned int plane_idx)
+{
+	if (plane_idx == 0)
+		return DRM_PLANE_TYPE_PRIMARY;
+	else if (plane_idx == 1)
+		return DRM_PLANE_TYPE_CURSOR;
+	else
+		return DRM_PLANE_TYPE_OVERLAY;
+
+}
+
+static int mtk_drm_crtc_init_comp_planes(struct drm_device *drm_dev,
+					 struct mtk_drm_crtc *mtk_crtc,
+					 int comp_idx, int pipe)
+{
+	int num_planes = mtk_drm_crtc_num_comp_planes(mtk_crtc, comp_idx);
+	struct mtk_ddp_comp *comp = mtk_crtc->ddp_comp[comp_idx];
+	int i, ret;
+
+	for (i = 0; i < num_planes; i++) {
+		ret = mtk_plane_init(drm_dev,
+				&mtk_crtc->planes[mtk_crtc->layer_nr],
+				BIT(pipe),
+				mtk_drm_crtc_plane_type(mtk_crtc->layer_nr),
+				mtk_ddp_comp_supported_rotations(comp));
+		if (ret)
+			return ret;
+
+		mtk_crtc->layer_nr++;
+	}
+	return 0;
+}
+
 int mtk_drm_crtc_create(struct drm_device *drm_dev,
 			const enum mtk_ddp_comp_id *path, unsigned int path_len)
 {
 	struct mtk_drm_private *priv = drm_dev->dev_private;
 	struct device *dev = drm_dev->dev;
 	struct mtk_drm_crtc *mtk_crtc;
-	enum drm_plane_type type;
-	unsigned int zpos;
+	unsigned int num_comp_planes = 0;
 	int pipe = priv->num_pipes;
 	int ret;
 	int i;
@@ -581,17 +679,15 @@ int mtk_drm_crtc_create(struct drm_device *drm_dev,
 		mtk_crtc->ddp_comp[i] = comp;
 	}
 
-	mtk_crtc->layer_nr = mtk_ddp_comp_layer_nr(mtk_crtc->ddp_comp[0]);
-	mtk_crtc->planes = devm_kcalloc(dev, mtk_crtc->layer_nr,
-					sizeof(struct drm_plane),
-					GFP_KERNEL);
-
-	for (zpos = 0; zpos < mtk_crtc->layer_nr; zpos++) {
-		type = (zpos == 0) ? DRM_PLANE_TYPE_PRIMARY :
-				(zpos == 1) ? DRM_PLANE_TYPE_CURSOR :
-						DRM_PLANE_TYPE_OVERLAY;
-		ret = mtk_plane_init(drm_dev, &mtk_crtc->planes[zpos],
-				     BIT(pipe), type);
+	for (i = 0; i < mtk_crtc->ddp_comp_nr; i++)
+		num_comp_planes += mtk_drm_crtc_num_comp_planes(mtk_crtc, i);
+
+	mtk_crtc->planes = devm_kcalloc(dev, num_comp_planes,
+					sizeof(struct drm_plane), GFP_KERNEL);
+
+	for (i = 0; i < mtk_crtc->ddp_comp_nr; i++) {
+		ret = mtk_drm_crtc_init_comp_planes(drm_dev, mtk_crtc, i,
+						    pipe);
 		if (ret)
 			return ret;
 	}
diff --git a/drivers/gpu/drm/mediatek/mtk_drm_crtc.h b/drivers/gpu/drm/mediatek/mtk_drm_crtc.h
index fcc134eb00c9..6afe1c19557a 100644
--- a/drivers/gpu/drm/mediatek/mtk_drm_crtc.h
+++ b/drivers/gpu/drm/mediatek/mtk_drm_crtc.h
@@ -19,5 +19,7 @@ void mtk_crtc_ddp_irq(struct drm_crtc *crtc, struct mtk_ddp_comp *comp);
 int mtk_drm_crtc_create(struct drm_device *drm_dev,
 			const enum mtk_ddp_comp_id *path,
 			unsigned int path_len);
+int mtk_drm_crtc_plane_check(struct drm_crtc *crtc, struct drm_plane *plane,
+			     struct mtk_plane_state *state);
 
 #endif /* MTK_DRM_CRTC_H */
diff --git a/drivers/gpu/drm/mediatek/mtk_drm_ddp.c b/drivers/gpu/drm/mediatek/mtk_drm_ddp.c
index 8106a71a7404..13035c906035 100644
--- a/drivers/gpu/drm/mediatek/mtk_drm_ddp.c
+++ b/drivers/gpu/drm/mediatek/mtk_drm_ddp.c
@@ -33,12 +33,15 @@
 #define DISP_REG_CONFIG_DSI_SEL			0x050
 #define DISP_REG_CONFIG_DPI_SEL			0x064
 
-#define DISP_REG_MUTEX_EN(n)	(0x20 + 0x20 * (n))
-#define DISP_REG_MUTEX(n)	(0x24 + 0x20 * (n))
-#define DISP_REG_MUTEX_RST(n)	(0x28 + 0x20 * (n))
-#define DISP_REG_MUTEX_MOD(n)	(0x2c + 0x20 * (n))
-#define DISP_REG_MUTEX_SOF(n)	(0x30 + 0x20 * (n))
-#define DISP_REG_MUTEX_MOD2(n)	(0x34 + 0x20 * (n))
+#define MT2701_DISP_MUTEX0_MOD0			0x2c
+#define MT2701_DISP_MUTEX0_SOF0			0x30
+
+#define DISP_REG_MUTEX_EN(n)			(0x20 + 0x20 * (n))
+#define DISP_REG_MUTEX(n)			(0x24 + 0x20 * (n))
+#define DISP_REG_MUTEX_RST(n)			(0x28 + 0x20 * (n))
+#define DISP_REG_MUTEX_MOD(mutex_mod_reg, n)	(mutex_mod_reg + 0x20 * (n))
+#define DISP_REG_MUTEX_SOF(mutex_sof_reg, n)	(mutex_sof_reg + 0x20 * (n))
+#define DISP_REG_MUTEX_MOD2(n)			(0x34 + 0x20 * (n))
 
 #define INT_MUTEX				BIT(1)
 
@@ -139,12 +142,30 @@ struct mtk_disp_mutex {
 	bool claimed;
 };
 
+enum mtk_ddp_mutex_sof_id {
+	DDP_MUTEX_SOF_SINGLE_MODE,
+	DDP_MUTEX_SOF_DSI0,
+	DDP_MUTEX_SOF_DSI1,
+	DDP_MUTEX_SOF_DPI0,
+	DDP_MUTEX_SOF_DPI1,
+	DDP_MUTEX_SOF_DSI2,
+	DDP_MUTEX_SOF_DSI3,
+};
+
+struct mtk_ddp_data {
+	const unsigned int *mutex_mod;
+	const unsigned int *mutex_sof;
+	const unsigned int mutex_mod_reg;
+	const unsigned int mutex_sof_reg;
+	const bool no_clk;
+};
+
 struct mtk_ddp {
 	struct device			*dev;
 	struct clk			*clk;
 	void __iomem			*regs;
 	struct mtk_disp_mutex		mutex[10];
-	const unsigned int		*mutex_mod;
+	const struct mtk_ddp_data	*data;
 };
 
 static const unsigned int mt2701_mutex_mod[DDP_COMPONENT_ID_MAX] = {
@@ -194,6 +215,37 @@ static const unsigned int mt8173_mutex_mod[DDP_COMPONENT_ID_MAX] = {
 	[DDP_COMPONENT_WDMA1] = MT8173_MUTEX_MOD_DISP_WDMA1,
 };
 
+static const unsigned int mt2712_mutex_sof[DDP_MUTEX_SOF_DSI3 + 1] = {
+	[DDP_MUTEX_SOF_SINGLE_MODE] = MUTEX_SOF_SINGLE_MODE,
+	[DDP_MUTEX_SOF_DSI0] = MUTEX_SOF_DSI0,
+	[DDP_MUTEX_SOF_DSI1] = MUTEX_SOF_DSI1,
+	[DDP_MUTEX_SOF_DPI0] = MUTEX_SOF_DPI0,
+	[DDP_MUTEX_SOF_DPI1] = MUTEX_SOF_DPI1,
+	[DDP_MUTEX_SOF_DSI2] = MUTEX_SOF_DSI2,
+	[DDP_MUTEX_SOF_DSI3] = MUTEX_SOF_DSI3,
+};
+
+static const struct mtk_ddp_data mt2701_ddp_driver_data = {
+	.mutex_mod = mt2701_mutex_mod,
+	.mutex_sof = mt2712_mutex_sof,
+	.mutex_mod_reg = MT2701_DISP_MUTEX0_MOD0,
+	.mutex_sof_reg = MT2701_DISP_MUTEX0_SOF0,
+};
+
+static const struct mtk_ddp_data mt2712_ddp_driver_data = {
+	.mutex_mod = mt2712_mutex_mod,
+	.mutex_sof = mt2712_mutex_sof,
+	.mutex_mod_reg = MT2701_DISP_MUTEX0_MOD0,
+	.mutex_sof_reg = MT2701_DISP_MUTEX0_SOF0,
+};
+
+static const struct mtk_ddp_data mt8173_ddp_driver_data = {
+	.mutex_mod = mt8173_mutex_mod,
+	.mutex_sof = mt2712_mutex_sof,
+	.mutex_mod_reg = MT2701_DISP_MUTEX0_MOD0,
+	.mutex_sof_reg = MT2701_DISP_MUTEX0_SOF0,
+};
+
 static unsigned int mtk_ddp_mout_en(enum mtk_ddp_comp_id cur,
 				    enum mtk_ddp_comp_id next,
 				    unsigned int *addr)
@@ -432,45 +484,49 @@ void mtk_disp_mutex_add_comp(struct mtk_disp_mutex *mutex,
 	struct mtk_ddp *ddp = container_of(mutex, struct mtk_ddp,
 					   mutex[mutex->id]);
 	unsigned int reg;
+	unsigned int sof_id;
 	unsigned int offset;
 
 	WARN_ON(&ddp->mutex[mutex->id] != mutex);
 
 	switch (id) {
 	case DDP_COMPONENT_DSI0:
-		reg = MUTEX_SOF_DSI0;
+		sof_id = DDP_MUTEX_SOF_DSI0;
 		break;
 	case DDP_COMPONENT_DSI1:
-		reg = MUTEX_SOF_DSI0;
+		sof_id = DDP_MUTEX_SOF_DSI0;
 		break;
 	case DDP_COMPONENT_DSI2:
-		reg = MUTEX_SOF_DSI2;
+		sof_id = DDP_MUTEX_SOF_DSI2;
 		break;
 	case DDP_COMPONENT_DSI3:
-		reg = MUTEX_SOF_DSI3;
+		sof_id = DDP_MUTEX_SOF_DSI3;
 		break;
 	case DDP_COMPONENT_DPI0:
-		reg = MUTEX_SOF_DPI0;
+		sof_id = DDP_MUTEX_SOF_DPI0;
 		break;
 	case DDP_COMPONENT_DPI1:
-		reg = MUTEX_SOF_DPI1;
+		sof_id = DDP_MUTEX_SOF_DPI1;
 		break;
 	default:
-		if (ddp->mutex_mod[id] < 32) {
-			offset = DISP_REG_MUTEX_MOD(mutex->id);
+		if (ddp->data->mutex_mod[id] < 32) {
+			offset = DISP_REG_MUTEX_MOD(ddp->data->mutex_mod_reg,
+						    mutex->id);
 			reg = readl_relaxed(ddp->regs + offset);
-			reg |= 1 << ddp->mutex_mod[id];
+			reg |= 1 << ddp->data->mutex_mod[id];
 			writel_relaxed(reg, ddp->regs + offset);
 		} else {
 			offset = DISP_REG_MUTEX_MOD2(mutex->id);
 			reg = readl_relaxed(ddp->regs + offset);
-			reg |= 1 << (ddp->mutex_mod[id] - 32);
+			reg |= 1 << (ddp->data->mutex_mod[id] - 32);
 			writel_relaxed(reg, ddp->regs + offset);
 		}
 		return;
 	}
 
-	writel_relaxed(reg, ddp->regs + DISP_REG_MUTEX_SOF(mutex->id));
+	writel_relaxed(ddp->data->mutex_sof[sof_id],
+		       ddp->regs +
+		       DISP_REG_MUTEX_SOF(ddp->data->mutex_sof_reg, mutex->id));
 }
 
 void mtk_disp_mutex_remove_comp(struct mtk_disp_mutex *mutex,
@@ -491,18 +547,21 @@ void mtk_disp_mutex_remove_comp(struct mtk_disp_mutex *mutex,
 	case DDP_COMPONENT_DPI0:
 	case DDP_COMPONENT_DPI1:
 		writel_relaxed(MUTEX_SOF_SINGLE_MODE,
-			       ddp->regs + DISP_REG_MUTEX_SOF(mutex->id));
+			       ddp->regs +
+			       DISP_REG_MUTEX_SOF(ddp->data->mutex_sof_reg,
+						  mutex->id));
 		break;
 	default:
-		if (ddp->mutex_mod[id] < 32) {
-			offset = DISP_REG_MUTEX_MOD(mutex->id);
+		if (ddp->data->mutex_mod[id] < 32) {
+			offset = DISP_REG_MUTEX_MOD(ddp->data->mutex_mod_reg,
+						    mutex->id);
 			reg = readl_relaxed(ddp->regs + offset);
-			reg &= ~(1 << ddp->mutex_mod[id]);
+			reg &= ~(1 << ddp->data->mutex_mod[id]);
 			writel_relaxed(reg, ddp->regs + offset);
 		} else {
 			offset = DISP_REG_MUTEX_MOD2(mutex->id);
 			reg = readl_relaxed(ddp->regs + offset);
-			reg &= ~(1 << (ddp->mutex_mod[id] - 32));
+			reg &= ~(1 << (ddp->data->mutex_mod[id] - 32));
 			writel_relaxed(reg, ddp->regs + offset);
 		}
 		break;
@@ -564,10 +623,14 @@ static int mtk_ddp_probe(struct platform_device *pdev)
 	for (i = 0; i < 10; i++)
 		ddp->mutex[i].id = i;
 
-	ddp->clk = devm_clk_get(dev, NULL);
-	if (IS_ERR(ddp->clk)) {
-		dev_err(dev, "Failed to get clock\n");
-		return PTR_ERR(ddp->clk);
+	ddp->data = of_device_get_match_data(dev);
+
+	if (!ddp->data->no_clk) {
+		ddp->clk = devm_clk_get(dev, NULL);
+		if (IS_ERR(ddp->clk)) {
+			dev_err(dev, "Failed to get clock\n");
+			return PTR_ERR(ddp->clk);
+		}
 	}
 
 	regs = platform_get_resource(pdev, IORESOURCE_MEM, 0);
@@ -577,8 +640,6 @@ static int mtk_ddp_probe(struct platform_device *pdev)
 		return PTR_ERR(ddp->regs);
 	}
 
-	ddp->mutex_mod = of_device_get_match_data(dev);
-
 	platform_set_drvdata(pdev, ddp);
 
 	return 0;
@@ -590,9 +651,12 @@ static int mtk_ddp_remove(struct platform_device *pdev)
 }
 
 static const struct of_device_id ddp_driver_dt_match[] = {
-	{ .compatible = "mediatek,mt2701-disp-mutex", .data = mt2701_mutex_mod},
-	{ .compatible = "mediatek,mt2712-disp-mutex", .data = mt2712_mutex_mod},
-	{ .compatible = "mediatek,mt8173-disp-mutex", .data = mt8173_mutex_mod},
+	{ .compatible = "mediatek,mt2701-disp-mutex",
+	  .data = &mt2701_ddp_driver_data},
+	{ .compatible = "mediatek,mt2712-disp-mutex",
+	  .data = &mt2712_ddp_driver_data},
+	{ .compatible = "mediatek,mt8173-disp-mutex",
+	  .data = &mt8173_ddp_driver_data},
 	{},
 };
 MODULE_DEVICE_TABLE(of, ddp_driver_dt_match);
diff --git a/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.c b/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.c
index efa85973e46b..7f21307cda75 100644
--- a/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.c
+++ b/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.c
@@ -33,6 +33,18 @@
 #define DISP_AAL_EN				0x0000
 #define DISP_AAL_SIZE				0x0030
 
+#define DISP_CCORR_EN				0x0000
+#define CCORR_EN				BIT(0)
+#define DISP_CCORR_CFG				0x0020
+#define CCORR_RELAY_MODE			BIT(0)
+#define DISP_CCORR_SIZE				0x0030
+
+#define DISP_DITHER_EN				0x0000
+#define DITHER_EN				BIT(0)
+#define DISP_DITHER_CFG				0x0020
+#define DITHER_RELAY_MODE			BIT(0)
+#define DISP_DITHER_SIZE			0x0030
+
 #define DISP_GAMMA_EN				0x0000
 #define DISP_GAMMA_CFG				0x0020
 #define DISP_GAMMA_SIZE				0x0030
@@ -123,6 +135,42 @@ static void mtk_aal_stop(struct mtk_ddp_comp *comp)
 	writel_relaxed(0x0, comp->regs + DISP_AAL_EN);
 }
 
+static void mtk_ccorr_config(struct mtk_ddp_comp *comp, unsigned int w,
+			     unsigned int h, unsigned int vrefresh,
+			     unsigned int bpc)
+{
+	writel(h << 16 | w, comp->regs + DISP_CCORR_SIZE);
+	writel(CCORR_RELAY_MODE, comp->regs + DISP_CCORR_CFG);
+}
+
+static void mtk_ccorr_start(struct mtk_ddp_comp *comp)
+{
+	writel(CCORR_EN, comp->regs + DISP_CCORR_EN);
+}
+
+static void mtk_ccorr_stop(struct mtk_ddp_comp *comp)
+{
+	writel_relaxed(0x0, comp->regs + DISP_CCORR_EN);
+}
+
+static void mtk_dither_config(struct mtk_ddp_comp *comp, unsigned int w,
+			      unsigned int h, unsigned int vrefresh,
+			      unsigned int bpc)
+{
+	writel(h << 16 | w, comp->regs + DISP_DITHER_SIZE);
+	writel(DITHER_RELAY_MODE, comp->regs + DISP_DITHER_CFG);
+}
+
+static void mtk_dither_start(struct mtk_ddp_comp *comp)
+{
+	writel(DITHER_EN, comp->regs + DISP_DITHER_EN);
+}
+
+static void mtk_dither_stop(struct mtk_ddp_comp *comp)
+{
+	writel_relaxed(0x0, comp->regs + DISP_DITHER_EN);
+}
+
 static void mtk_gamma_config(struct mtk_ddp_comp *comp, unsigned int w,
 			     unsigned int h, unsigned int vrefresh,
 			     unsigned int bpc)
@@ -171,6 +219,18 @@ static const struct mtk_ddp_comp_funcs ddp_aal = {
 	.stop = mtk_aal_stop,
 };
 
+static const struct mtk_ddp_comp_funcs ddp_ccorr = {
+	.config = mtk_ccorr_config,
+	.start = mtk_ccorr_start,
+	.stop = mtk_ccorr_stop,
+};
+
+static const struct mtk_ddp_comp_funcs ddp_dither = {
+	.config = mtk_dither_config,
+	.start = mtk_dither_start,
+	.stop = mtk_dither_stop,
+};
+
 static const struct mtk_ddp_comp_funcs ddp_gamma = {
 	.gamma_set = mtk_gamma_set,
 	.config = mtk_gamma_config,
@@ -189,11 +249,14 @@ static const struct mtk_ddp_comp_funcs ddp_ufoe = {
 
 static const char * const mtk_ddp_comp_stem[MTK_DDP_COMP_TYPE_MAX] = {
 	[MTK_DISP_OVL] = "ovl",
+	[MTK_DISP_OVL_2L] = "ovl_2l",
 	[MTK_DISP_RDMA] = "rdma",
 	[MTK_DISP_WDMA] = "wdma",
 	[MTK_DISP_COLOR] = "color",
+	[MTK_DISP_CCORR] = "ccorr",
 	[MTK_DISP_AAL] = "aal",
 	[MTK_DISP_GAMMA] = "gamma",
+	[MTK_DISP_DITHER] = "dither",
 	[MTK_DISP_UFOE] = "ufoe",
 	[MTK_DSI] = "dsi",
 	[MTK_DPI] = "dpi",
@@ -213,8 +276,10 @@ static const struct mtk_ddp_comp_match mtk_ddp_matches[DDP_COMPONENT_ID_MAX] = {
 	[DDP_COMPONENT_AAL0]	= { MTK_DISP_AAL,	0, &ddp_aal },
 	[DDP_COMPONENT_AAL1]	= { MTK_DISP_AAL,	1, &ddp_aal },
 	[DDP_COMPONENT_BLS]	= { MTK_DISP_BLS,	0, NULL },
+	[DDP_COMPONENT_CCORR]	= { MTK_DISP_CCORR,	0, &ddp_ccorr },
 	[DDP_COMPONENT_COLOR0]	= { MTK_DISP_COLOR,	0, NULL },
 	[DDP_COMPONENT_COLOR1]	= { MTK_DISP_COLOR,	1, NULL },
+	[DDP_COMPONENT_DITHER]	= { MTK_DISP_DITHER,	0, &ddp_dither },
 	[DDP_COMPONENT_DPI0]	= { MTK_DPI,		0, NULL },
 	[DDP_COMPONENT_DPI1]	= { MTK_DPI,		1, NULL },
 	[DDP_COMPONENT_DSI0]	= { MTK_DSI,		0, NULL },
@@ -226,6 +291,8 @@ static const struct mtk_ddp_comp_match mtk_ddp_matches[DDP_COMPONENT_ID_MAX] = {
 	[DDP_COMPONENT_OD1]	= { MTK_DISP_OD,	1, &ddp_od },
 	[DDP_COMPONENT_OVL0]	= { MTK_DISP_OVL,	0, NULL },
 	[DDP_COMPONENT_OVL1]	= { MTK_DISP_OVL,	1, NULL },
+	[DDP_COMPONENT_OVL_2L0]	= { MTK_DISP_OVL_2L,	0, NULL },
+	[DDP_COMPONENT_OVL_2L1]	= { MTK_DISP_OVL_2L,	1, NULL },
 	[DDP_COMPONENT_PWM0]	= { MTK_DISP_PWM,	0, NULL },
 	[DDP_COMPONENT_PWM1]	= { MTK_DISP_PWM,	1, NULL },
 	[DDP_COMPONENT_PWM2]	= { MTK_DISP_PWM,	2, NULL },
diff --git a/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.h b/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.h
index 0ad287f427cc..2f1e9e75b8da 100644
--- a/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.h
+++ b/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.h
@@ -17,9 +17,12 @@ struct drm_crtc_state;
 
 enum mtk_ddp_comp_type {
 	MTK_DISP_OVL,
+	MTK_DISP_OVL_2L,
 	MTK_DISP_RDMA,
 	MTK_DISP_WDMA,
 	MTK_DISP_COLOR,
+	MTK_DISP_CCORR,
+	MTK_DISP_DITHER,
 	MTK_DISP_AAL,
 	MTK_DISP_GAMMA,
 	MTK_DISP_UFOE,
@@ -36,8 +39,10 @@ enum mtk_ddp_comp_id {
 	DDP_COMPONENT_AAL0,
 	DDP_COMPONENT_AAL1,
 	DDP_COMPONENT_BLS,
+	DDP_COMPONENT_CCORR,
 	DDP_COMPONENT_COLOR0,
 	DDP_COMPONENT_COLOR1,
+	DDP_COMPONENT_DITHER,
 	DDP_COMPONENT_DPI0,
 	DDP_COMPONENT_DPI1,
 	DDP_COMPONENT_DSI0,
@@ -48,6 +53,8 @@ enum mtk_ddp_comp_id {
 	DDP_COMPONENT_OD0,
 	DDP_COMPONENT_OD1,
 	DDP_COMPONENT_OVL0,
+	DDP_COMPONENT_OVL_2L0,
+	DDP_COMPONENT_OVL_2L1,
 	DDP_COMPONENT_OVL1,
 	DDP_COMPONENT_PWM0,
 	DDP_COMPONENT_PWM1,
@@ -70,13 +77,19 @@ struct mtk_ddp_comp_funcs {
 	void (*stop)(struct mtk_ddp_comp *comp);
 	void (*enable_vblank)(struct mtk_ddp_comp *comp, struct drm_crtc *crtc);
 	void (*disable_vblank)(struct mtk_ddp_comp *comp);
+	unsigned int (*supported_rotations)(struct mtk_ddp_comp *comp);
 	unsigned int (*layer_nr)(struct mtk_ddp_comp *comp);
 	void (*layer_on)(struct mtk_ddp_comp *comp, unsigned int idx);
 	void (*layer_off)(struct mtk_ddp_comp *comp, unsigned int idx);
+	int (*layer_check)(struct mtk_ddp_comp *comp,
+			   unsigned int idx,
+			   struct mtk_plane_state *state);
 	void (*layer_config)(struct mtk_ddp_comp *comp, unsigned int idx,
 			     struct mtk_plane_state *state);
 	void (*gamma_set)(struct mtk_ddp_comp *comp,
 			  struct drm_crtc_state *state);
+	void (*bgclr_in_on)(struct mtk_ddp_comp *comp);
+	void (*bgclr_in_off)(struct mtk_ddp_comp *comp);
 };
 
 struct mtk_ddp_comp {
@@ -121,6 +134,15 @@ static inline void mtk_ddp_comp_disable_vblank(struct mtk_ddp_comp *comp)
 		comp->funcs->disable_vblank(comp);
 }
 
+static inline
+unsigned int mtk_ddp_comp_supported_rotations(struct mtk_ddp_comp *comp)
+{
+	if (comp->funcs && comp->funcs->supported_rotations)
+		return comp->funcs->supported_rotations(comp);
+
+	return 0;
+}
+
 static inline unsigned int mtk_ddp_comp_layer_nr(struct mtk_ddp_comp *comp)
 {
 	if (comp->funcs && comp->funcs->layer_nr)
@@ -143,6 +165,15 @@ static inline void mtk_ddp_comp_layer_off(struct mtk_ddp_comp *comp,
 		comp->funcs->layer_off(comp, idx);
 }
 
+static inline int mtk_ddp_comp_layer_check(struct mtk_ddp_comp *comp,
+					   unsigned int idx,
+					   struct mtk_plane_state *state)
+{
+	if (comp->funcs && comp->funcs->layer_check)
+		return comp->funcs->layer_check(comp, idx, state);
+	return 0;
+}
+
 static inline void mtk_ddp_comp_layer_config(struct mtk_ddp_comp *comp,
 					     unsigned int idx,
 					     struct mtk_plane_state *state)
@@ -158,6 +189,18 @@ static inline void mtk_ddp_gamma_set(struct mtk_ddp_comp *comp,
 		comp->funcs->gamma_set(comp, state);
 }
 
+static inline void mtk_ddp_comp_bgclr_in_on(struct mtk_ddp_comp *comp)
+{
+	if (comp->funcs && comp->funcs->bgclr_in_on)
+		comp->funcs->bgclr_in_on(comp);
+}
+
+static inline void mtk_ddp_comp_bgclr_in_off(struct mtk_ddp_comp *comp)
+{
+	if (comp->funcs && comp->funcs->bgclr_in_off)
+		comp->funcs->bgclr_in_off(comp);
+}
+
 int mtk_ddp_comp_get_id(struct device_node *node,
 			enum mtk_ddp_comp_type comp_type);
 int mtk_ddp_comp_init(struct device *dev, struct device_node *comp_node,
diff --git a/drivers/gpu/drm/mediatek/mtk_drm_drv.c b/drivers/gpu/drm/mediatek/mtk_drm_drv.c
index 352b81a7a670..84d14213d992 100644
--- a/drivers/gpu/drm/mediatek/mtk_drm_drv.c
+++ b/drivers/gpu/drm/mediatek/mtk_drm_drv.c
@@ -547,6 +547,7 @@ static int mtk_drm_probe(struct platform_device *pdev)
 		 */
 		if (comp_type == MTK_DISP_COLOR ||
 		    comp_type == MTK_DISP_OVL ||
+		    comp_type == MTK_DISP_OVL_2L ||
 		    comp_type == MTK_DISP_RDMA ||
 		    comp_type == MTK_DSI ||
 		    comp_type == MTK_DPI) {
@@ -669,8 +670,8 @@ static struct platform_driver * const mtk_drm_drivers[] = {
 	&mtk_disp_rdma_driver,
 	&mtk_dpi_driver,
 	&mtk_drm_platform_driver,
-	&mtk_dsi_driver,
 	&mtk_mipi_tx_driver,
+	&mtk_dsi_driver,
 };
 
 static int __init mtk_drm_init(void)
diff --git a/drivers/gpu/drm/mediatek/mtk_drm_gem.c b/drivers/gpu/drm/mediatek/mtk_drm_gem.c
index ca672f1d140d..b04a3c2b111e 100644
--- a/drivers/gpu/drm/mediatek/mtk_drm_gem.c
+++ b/drivers/gpu/drm/mediatek/mtk_drm_gem.c
@@ -271,7 +271,7 @@ void *mtk_drm_gem_prime_vmap(struct drm_gem_object *obj)
 			       pgprot_writecombine(PAGE_KERNEL));
 
 out:
-	kfree((void *)sgt);
+	kfree(sgt);
 
 	return mtk_gem->kvaddr;
 }
@@ -285,5 +285,5 @@ void mtk_drm_gem_prime_vunmap(struct drm_gem_object *obj, void *vaddr)
 
 	vunmap(vaddr);
 	mtk_gem->kvaddr = 0;
-	kfree((void *)mtk_gem->pages);
+	kfree(mtk_gem->pages);
 }
diff --git a/drivers/gpu/drm/mediatek/mtk_drm_plane.c b/drivers/gpu/drm/mediatek/mtk_drm_plane.c
index 584a9ecadce6..3b0cc91c7023 100644
--- a/drivers/gpu/drm/mediatek/mtk_drm_plane.c
+++ b/drivers/gpu/drm/mediatek/mtk_drm_plane.c
@@ -20,6 +20,12 @@
 static const u32 formats[] = {
 	DRM_FORMAT_XRGB8888,
 	DRM_FORMAT_ARGB8888,
+	DRM_FORMAT_BGRX8888,
+	DRM_FORMAT_BGRA8888,
+	DRM_FORMAT_ABGR8888,
+	DRM_FORMAT_XBGR8888,
+	DRM_FORMAT_RGB888,
+	DRM_FORMAT_BGR888,
 	DRM_FORMAT_RGB565,
 	DRM_FORMAT_UYVY,
 	DRM_FORMAT_YUYV,
@@ -84,6 +90,7 @@ static int mtk_plane_atomic_check(struct drm_plane *plane,
 {
 	struct drm_framebuffer *fb = state->fb;
 	struct drm_crtc_state *crtc_state;
+	int ret;
 
 	if (!fb)
 		return 0;
@@ -91,6 +98,11 @@ static int mtk_plane_atomic_check(struct drm_plane *plane,
 	if (!state->crtc)
 		return 0;
 
+	ret = mtk_drm_crtc_plane_check(state->crtc, plane,
+				       to_mtk_plane_state(state));
+	if (ret)
+		return ret;
+
 	crtc_state = drm_atomic_get_crtc_state(state->state, state->crtc);
 	if (IS_ERR(crtc_state))
 		return PTR_ERR(crtc_state);
@@ -132,6 +144,7 @@ static void mtk_plane_atomic_update(struct drm_plane *plane,
 	state->pending.y = plane->state->dst.y1;
 	state->pending.width = drm_rect_width(&plane->state->dst);
 	state->pending.height = drm_rect_height(&plane->state->dst);
+	state->pending.rotation = plane->state->rotation;
 	wmb(); /* Make sure the above parameters are set before update */
 	state->pending.dirty = true;
 }
@@ -154,7 +167,8 @@ static const struct drm_plane_helper_funcs mtk_plane_helper_funcs = {
 };
 
 int mtk_plane_init(struct drm_device *dev, struct drm_plane *plane,
-		   unsigned long possible_crtcs, enum drm_plane_type type)
+		   unsigned long possible_crtcs, enum drm_plane_type type,
+		   unsigned int supported_rotations)
 {
 	int err;
 
@@ -166,6 +180,14 @@ int mtk_plane_init(struct drm_device *dev, struct drm_plane *plane,
 		return err;
 	}
 
+	if (supported_rotations & ~DRM_MODE_ROTATE_0) {
+		err = drm_plane_create_rotation_property(plane,
+							 DRM_MODE_ROTATE_0,
+							 supported_rotations);
+		if (err)
+			DRM_INFO("Create rotation property failed\n");
+	}
+
 	drm_plane_helper_add(plane, &mtk_plane_helper_funcs);
 
 	return 0;
diff --git a/drivers/gpu/drm/mediatek/mtk_drm_plane.h b/drivers/gpu/drm/mediatek/mtk_drm_plane.h
index 6f842df722c7..760885e35b27 100644
--- a/drivers/gpu/drm/mediatek/mtk_drm_plane.h
+++ b/drivers/gpu/drm/mediatek/mtk_drm_plane.h
@@ -20,6 +20,7 @@ struct mtk_plane_pending_state {
 	unsigned int			y;
 	unsigned int			width;
 	unsigned int			height;
+	unsigned int			rotation;
 	bool				dirty;
 };
 
@@ -35,6 +36,7 @@ to_mtk_plane_state(struct drm_plane_state *state)
 }
 
 int mtk_plane_init(struct drm_device *dev, struct drm_plane *plane,
-		   unsigned long possible_crtcs, enum drm_plane_type type);
+		   unsigned long possible_crtcs, enum drm_plane_type type,
+		   unsigned int supported_rotations);
 
 #endif
diff --git a/drivers/gpu/drm/mediatek/mtk_dsi.c b/drivers/gpu/drm/mediatek/mtk_dsi.c
index 224afb666881..e9931bbbe846 100644
--- a/drivers/gpu/drm/mediatek/mtk_dsi.c
+++ b/drivers/gpu/drm/mediatek/mtk_dsi.c
@@ -16,6 +16,7 @@
 #include <video/videomode.h>
 
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_mipi_dsi.h>
 #include <drm/drm_of.h>
 #include <drm/drm_panel.h>
@@ -39,6 +40,7 @@
 #define DSI_CON_CTRL		0x10
 #define DSI_RESET			BIT(0)
 #define DSI_EN				BIT(1)
+#define DPHY_RESET			BIT(2)
 
 #define DSI_MODE_CTRL		0x14
 #define MODE				(3)
@@ -72,6 +74,7 @@
 #define DSI_VBP_NL		0x24
 #define DSI_VFP_NL		0x28
 #define DSI_VACT_NL		0x2C
+#define DSI_SIZE_CON		0x38
 #define DSI_HSA_WC		0x50
 #define DSI_HBP_WC		0x54
 #define DSI_HFP_WC		0x58
@@ -125,7 +128,10 @@
 #define VM_CMD_EN			BIT(0)
 #define TS_VFP_EN			BIT(5)
 
-#define DSI_CMDQ0		0x180
+#define DSI_SHADOW_DEBUG	0x190U
+#define FORCE_COMMIT			BIT(0)
+#define BYPASS_SHADOW			BIT(1)
+
 #define CONFIG				(0xff << 0)
 #define SHORT_PACKET			0
 #define LONG_PACKET			2
@@ -134,12 +140,6 @@
 #define DATA_0				(0xff << 16)
 #define DATA_1				(0xff << 24)
 
-#define T_LPX		5
-#define T_HS_PREP	6
-#define T_HS_TRAIL	8
-#define T_HS_EXIT	7
-#define T_HS_ZERO	10
-
 #define NS_TO_CYCLE(n, c)    ((n) / (c) + (((n) % (c)) ? 1 : 0))
 
 #define MTK_DSI_HOST_IS_READ(type) \
@@ -148,8 +148,33 @@
 	(type == MIPI_DSI_GENERIC_READ_REQUEST_2_PARAM) || \
 	(type == MIPI_DSI_DCS_READ))
 
+struct mtk_phy_timing {
+	u32 lpx;
+	u32 da_hs_prepare;
+	u32 da_hs_zero;
+	u32 da_hs_trail;
+
+	u32 ta_go;
+	u32 ta_sure;
+	u32 ta_get;
+	u32 da_hs_exit;
+
+	u32 clk_hs_zero;
+	u32 clk_hs_trail;
+
+	u32 clk_hs_prepare;
+	u32 clk_hs_post;
+	u32 clk_hs_exit;
+};
+
 struct phy;
 
+struct mtk_dsi_driver_data {
+	const u32 reg_cmdq_off;
+	bool has_shadow_ctl;
+	bool has_size_ctl;
+};
+
 struct mtk_dsi {
 	struct mtk_ddp_comp ddp_comp;
 	struct device *dev;
@@ -172,10 +197,12 @@ struct mtk_dsi {
 	enum mipi_dsi_pixel_format format;
 	unsigned int lanes;
 	struct videomode vm;
+	struct mtk_phy_timing phy_timing;
 	int refcount;
 	bool enabled;
 	u32 irq_data;
 	wait_queue_head_t irq_wait_queue;
+	const struct mtk_dsi_driver_data *driver_data;
 };
 
 static inline struct mtk_dsi *encoder_to_dsi(struct drm_encoder *e)
@@ -204,17 +231,36 @@ static void mtk_dsi_phy_timconfig(struct mtk_dsi *dsi)
 {
 	u32 timcon0, timcon1, timcon2, timcon3;
 	u32 ui, cycle_time;
+	struct mtk_phy_timing *timing = &dsi->phy_timing;
+
+	ui = DIV_ROUND_UP(1000000000, dsi->data_rate);
+	cycle_time = div_u64(8000000000ULL, dsi->data_rate);
+
+	timing->lpx = NS_TO_CYCLE(60, cycle_time);
+	timing->da_hs_prepare = NS_TO_CYCLE(50 + 5 * ui, cycle_time);
+	timing->da_hs_zero = NS_TO_CYCLE(110 + 6 * ui, cycle_time);
+	timing->da_hs_trail = NS_TO_CYCLE(77 + 4 * ui, cycle_time);
+
+	timing->ta_go = 4 * timing->lpx;
+	timing->ta_sure = 3 * timing->lpx / 2;
+	timing->ta_get = 5 * timing->lpx;
+	timing->da_hs_exit = 2 * timing->lpx;
 
-	ui = 1000 / dsi->data_rate + 0x01;
-	cycle_time = 8000 / dsi->data_rate + 0x01;
+	timing->clk_hs_zero = NS_TO_CYCLE(336, cycle_time);
+	timing->clk_hs_trail = NS_TO_CYCLE(100, cycle_time) + 10;
 
-	timcon0 = T_LPX | T_HS_PREP << 8 | T_HS_ZERO << 16 | T_HS_TRAIL << 24;
-	timcon1 = 4 * T_LPX | (3 * T_LPX / 2) << 8 | 5 * T_LPX << 16 |
-		  T_HS_EXIT << 24;
-	timcon2 = ((NS_TO_CYCLE(0x64, cycle_time) + 0xa) << 24) |
-		  (NS_TO_CYCLE(0x150, cycle_time) << 16);
-	timcon3 = NS_TO_CYCLE(0x40, cycle_time) | (2 * T_LPX) << 16 |
-		  NS_TO_CYCLE(80 + 52 * ui, cycle_time) << 8;
+	timing->clk_hs_prepare = NS_TO_CYCLE(64, cycle_time);
+	timing->clk_hs_post = NS_TO_CYCLE(80 + 52 * ui, cycle_time);
+	timing->clk_hs_exit = 2 * timing->lpx;
+
+	timcon0 = timing->lpx | timing->da_hs_prepare << 8 |
+		  timing->da_hs_zero << 16 | timing->da_hs_trail << 24;
+	timcon1 = timing->ta_go | timing->ta_sure << 8 |
+		  timing->ta_get << 16 | timing->da_hs_exit << 24;
+	timcon2 = 1 << 8 | timing->clk_hs_zero << 16 |
+		  timing->clk_hs_trail << 24;
+	timcon3 = timing->clk_hs_prepare | timing->clk_hs_post << 8 |
+		  timing->clk_hs_exit << 16;
 
 	writel(timcon0, dsi->regs + DSI_PHY_TIMECON0);
 	writel(timcon1, dsi->regs + DSI_PHY_TIMECON1);
@@ -238,6 +284,12 @@ static void mtk_dsi_reset_engine(struct mtk_dsi *dsi)
 	mtk_dsi_mask(dsi, DSI_CON_CTRL, DSI_RESET, 0);
 }
 
+static void mtk_dsi_reset_dphy(struct mtk_dsi *dsi)
+{
+	mtk_dsi_mask(dsi, DSI_CON_CTRL, DPHY_RESET, DPHY_RESET);
+	mtk_dsi_mask(dsi, DSI_CON_CTRL, DPHY_RESET, 0);
+}
+
 static void mtk_dsi_clk_ulp_mode_enter(struct mtk_dsi *dsi)
 {
 	mtk_dsi_mask(dsi, DSI_PHY_LCCON, LC_HS_TX_EN, 0);
@@ -401,7 +453,8 @@ static void mtk_dsi_config_vdo_timing(struct mtk_dsi *dsi)
 	u32 horizontal_sync_active_byte;
 	u32 horizontal_backporch_byte;
 	u32 horizontal_frontporch_byte;
-	u32 dsi_tmp_buf_bpp;
+	u32 dsi_tmp_buf_bpp, data_phy_cycles;
+	struct mtk_phy_timing *timing = &dsi->phy_timing;
 
 	struct videomode *vm = &dsi->vm;
 
@@ -415,6 +468,10 @@ static void mtk_dsi_config_vdo_timing(struct mtk_dsi *dsi)
 	writel(vm->vfront_porch, dsi->regs + DSI_VFP_NL);
 	writel(vm->vactive, dsi->regs + DSI_VACT_NL);
 
+	if (dsi->driver_data->has_size_ctl)
+		writel(vm->vactive << 16 | vm->hactive,
+		       dsi->regs + DSI_SIZE_CON);
+
 	horizontal_sync_active_byte = (vm->hsync_len * dsi_tmp_buf_bpp - 10);
 
 	if (dsi->mode_flags & MIPI_DSI_MODE_VIDEO_SYNC_PULSE)
@@ -424,7 +481,34 @@ static void mtk_dsi_config_vdo_timing(struct mtk_dsi *dsi)
 		horizontal_backporch_byte = ((vm->hback_porch + vm->hsync_len) *
 			dsi_tmp_buf_bpp - 10);
 
-	horizontal_frontporch_byte = (vm->hfront_porch * dsi_tmp_buf_bpp - 12);
+	data_phy_cycles = timing->lpx + timing->da_hs_prepare +
+				  timing->da_hs_zero + timing->da_hs_exit + 2;
+
+	if (dsi->mode_flags & MIPI_DSI_MODE_VIDEO_BURST) {
+		if (vm->hfront_porch * dsi_tmp_buf_bpp >
+		    data_phy_cycles * dsi->lanes + 18) {
+			horizontal_frontporch_byte = vm->hfront_porch *
+						     dsi_tmp_buf_bpp -
+						     data_phy_cycles *
+						     dsi->lanes - 18;
+		} else {
+			DRM_WARN("HFP less than d-phy, FPS will under 60Hz\n");
+			horizontal_frontporch_byte = vm->hfront_porch *
+						     dsi_tmp_buf_bpp;
+		}
+	} else {
+		if (vm->hfront_porch * dsi_tmp_buf_bpp >
+		    data_phy_cycles * dsi->lanes + 12) {
+			horizontal_frontporch_byte = vm->hfront_porch *
+						     dsi_tmp_buf_bpp -
+						     data_phy_cycles *
+						     dsi->lanes - 12;
+		} else {
+			DRM_WARN("HFP less than d-phy, FPS will under 60Hz\n");
+			horizontal_frontporch_byte = vm->hfront_porch *
+						     dsi_tmp_buf_bpp;
+		}
+	}
 
 	writel(horizontal_sync_active_byte, dsi->regs + DSI_HSA_WC);
 	writel(horizontal_backporch_byte, dsi->regs + DSI_HBP_WC);
@@ -522,10 +606,9 @@ static s32 mtk_dsi_switch_to_cmd_mode(struct mtk_dsi *dsi, u8 irq_flag, u32 t)
 
 static int mtk_dsi_poweron(struct mtk_dsi *dsi)
 {
-	struct device *dev = dsi->dev;
+	struct device *dev = dsi->host.dev;
 	int ret;
-	u64 pixel_clock, total_bits;
-	u32 htotal, htotal_bits, bit_per_pixel, overhead_cycles, overhead_bits;
+	u32 bit_per_pixel;
 
 	if (++dsi->refcount != 1)
 		return 0;
@@ -544,24 +627,8 @@ static int mtk_dsi_poweron(struct mtk_dsi *dsi)
 		break;
 	}
 
-	/**
-	 * htotal_time = htotal * byte_per_pixel / num_lanes
-	 * overhead_time = lpx + hs_prepare + hs_zero + hs_trail + hs_exit
-	 * mipi_ratio = (htotal_time + overhead_time) / htotal_time
-	 * data_rate = pixel_clock * bit_per_pixel * mipi_ratio / num_lanes;
-	 */
-	pixel_clock = dsi->vm.pixelclock;
-	htotal = dsi->vm.hactive + dsi->vm.hback_porch + dsi->vm.hfront_porch +
-			dsi->vm.hsync_len;
-	htotal_bits = htotal * bit_per_pixel;
-
-	overhead_cycles = T_LPX + T_HS_PREP + T_HS_ZERO + T_HS_TRAIL +
-			T_HS_EXIT;
-	overhead_bits = overhead_cycles * dsi->lanes * 8;
-	total_bits = htotal_bits + overhead_bits;
-
-	dsi->data_rate = DIV_ROUND_UP_ULL(pixel_clock * total_bits,
-					  htotal * dsi->lanes);
+	dsi->data_rate = DIV_ROUND_UP_ULL(dsi->vm.pixelclock * bit_per_pixel,
+					  dsi->lanes);
 
 	ret = clk_set_rate(dsi->hs_clk, dsi->data_rate);
 	if (ret < 0) {
@@ -584,10 +651,17 @@ static int mtk_dsi_poweron(struct mtk_dsi *dsi)
 	}
 
 	mtk_dsi_enable(dsi);
+
+	if (dsi->driver_data->has_shadow_ctl)
+		writel(FORCE_COMMIT | BYPASS_SHADOW,
+		       dsi->regs + DSI_SHADOW_DEBUG);
+
 	mtk_dsi_reset_engine(dsi);
 	mtk_dsi_phy_timconfig(dsi);
 
 	mtk_dsi_rxtx_control(dsi);
+	usleep_range(30, 100);
+	mtk_dsi_reset_dphy(dsi);
 	mtk_dsi_ps_control_vact(dsi);
 	mtk_dsi_set_vm_cmd(dsi);
 	mtk_dsi_config_vdo_timing(dsi);
@@ -938,6 +1012,7 @@ static void mtk_dsi_cmdq(struct mtk_dsi *dsi, const struct mipi_dsi_msg *msg)
 	const char *tx_buf = msg->tx_buf;
 	u8 config, cmdq_size, cmdq_off, type = msg->type;
 	u32 reg_val, cmdq_mask, i;
+	u32 reg_cmdq_off = dsi->driver_data->reg_cmdq_off;
 
 	if (MTK_DSI_HOST_IS_READ(type))
 		config = BTA;
@@ -957,9 +1032,11 @@ static void mtk_dsi_cmdq(struct mtk_dsi *dsi, const struct mipi_dsi_msg *msg)
 	}
 
 	for (i = 0; i < msg->tx_len; i++)
-		writeb(tx_buf[i], dsi->regs + DSI_CMDQ0 + cmdq_off + i);
+		mtk_dsi_mask(dsi, (reg_cmdq_off + cmdq_off + i) & (~0x3U),
+			     (0xffUL << (((i + cmdq_off) & 3U) * 8U)),
+			     tx_buf[i] << (((i + cmdq_off) & 3U) * 8U));
 
-	mtk_dsi_mask(dsi, DSI_CMDQ0, cmdq_mask, reg_val);
+	mtk_dsi_mask(dsi, reg_cmdq_off, cmdq_mask, reg_val);
 	mtk_dsi_mask(dsi, DSI_CMDQ_SIZE, CMDQ_SIZE, cmdq_size);
 }
 
@@ -1049,12 +1126,6 @@ static int mtk_dsi_bind(struct device *dev, struct device *master, void *data)
 		return ret;
 	}
 
-	ret = mipi_dsi_host_register(&dsi->host);
-	if (ret < 0) {
-		dev_err(dev, "failed to register DSI host: %d\n", ret);
-		goto err_ddp_comp_unregister;
-	}
-
 	ret = mtk_dsi_create_conn_enc(drm, dsi);
 	if (ret) {
 		DRM_ERROR("Encoder create failed with %d\n", ret);
@@ -1064,8 +1135,6 @@ static int mtk_dsi_bind(struct device *dev, struct device *master, void *data)
 	return 0;
 
 err_unregister:
-	mipi_dsi_host_unregister(&dsi->host);
-err_ddp_comp_unregister:
 	mtk_ddp_comp_unregister(drm, &dsi->ddp_comp);
 	return ret;
 }
@@ -1077,7 +1146,6 @@ static void mtk_dsi_unbind(struct device *dev, struct device *master,
 	struct mtk_dsi *dsi = dev_get_drvdata(dev);
 
 	mtk_dsi_destroy_conn_enc(dsi);
-	mipi_dsi_host_unregister(&dsi->host);
 	mtk_ddp_comp_unregister(drm, &dsi->ddp_comp);
 }
 
@@ -1101,31 +1169,38 @@ static int mtk_dsi_probe(struct platform_device *pdev)
 
 	dsi->host.ops = &mtk_dsi_ops;
 	dsi->host.dev = dev;
+	ret = mipi_dsi_host_register(&dsi->host);
+	if (ret < 0) {
+		dev_err(dev, "failed to register DSI host: %d\n", ret);
+		return ret;
+	}
 
 	ret = drm_of_find_panel_or_bridge(dev->of_node, 0, 0,
 					  &dsi->panel, &dsi->bridge);
 	if (ret)
-		return ret;
+		goto err_unregister_host;
+
+	dsi->driver_data = of_device_get_match_data(dev);
 
 	dsi->engine_clk = devm_clk_get(dev, "engine");
 	if (IS_ERR(dsi->engine_clk)) {
 		ret = PTR_ERR(dsi->engine_clk);
 		dev_err(dev, "Failed to get engine clock: %d\n", ret);
-		return ret;
+		goto err_unregister_host;
 	}
 
 	dsi->digital_clk = devm_clk_get(dev, "digital");
 	if (IS_ERR(dsi->digital_clk)) {
 		ret = PTR_ERR(dsi->digital_clk);
 		dev_err(dev, "Failed to get digital clock: %d\n", ret);
-		return ret;
+		goto err_unregister_host;
 	}
 
 	dsi->hs_clk = devm_clk_get(dev, "hs");
 	if (IS_ERR(dsi->hs_clk)) {
 		ret = PTR_ERR(dsi->hs_clk);
 		dev_err(dev, "Failed to get hs clock: %d\n", ret);
-		return ret;
+		goto err_unregister_host;
 	}
 
 	regs = platform_get_resource(pdev, IORESOURCE_MEM, 0);
@@ -1133,33 +1208,35 @@ static int mtk_dsi_probe(struct platform_device *pdev)
 	if (IS_ERR(dsi->regs)) {
 		ret = PTR_ERR(dsi->regs);
 		dev_err(dev, "Failed to ioremap memory: %d\n", ret);
-		return ret;
+		goto err_unregister_host;
 	}
 
 	dsi->phy = devm_phy_get(dev, "dphy");
 	if (IS_ERR(dsi->phy)) {
 		ret = PTR_ERR(dsi->phy);
 		dev_err(dev, "Failed to get MIPI-DPHY: %d\n", ret);
-		return ret;
+		goto err_unregister_host;
 	}
 
 	comp_id = mtk_ddp_comp_get_id(dev->of_node, MTK_DSI);
 	if (comp_id < 0) {
 		dev_err(dev, "Failed to identify by alias: %d\n", comp_id);
-		return comp_id;
+		ret = comp_id;
+		goto err_unregister_host;
 	}
 
 	ret = mtk_ddp_comp_init(dev, dev->of_node, &dsi->ddp_comp, comp_id,
 				&mtk_dsi_funcs);
 	if (ret) {
 		dev_err(dev, "Failed to initialize component: %d\n", ret);
-		return ret;
+		goto err_unregister_host;
 	}
 
 	irq_num = platform_get_irq(pdev, 0);
 	if (irq_num < 0) {
-		dev_err(&pdev->dev, "failed to request dsi irq resource\n");
-		return -EPROBE_DEFER;
+		dev_err(&pdev->dev, "failed to get dsi irq_num: %d\n", irq_num);
+		ret = irq_num;
+		goto err_unregister_host;
 	}
 
 	irq_set_status_flags(irq_num, IRQ_TYPE_LEVEL_LOW);
@@ -1167,14 +1244,24 @@ static int mtk_dsi_probe(struct platform_device *pdev)
 			       IRQF_TRIGGER_LOW, dev_name(&pdev->dev), dsi);
 	if (ret) {
 		dev_err(&pdev->dev, "failed to request mediatek dsi irq\n");
-		return -EPROBE_DEFER;
+		goto err_unregister_host;
 	}
 
 	init_waitqueue_head(&dsi->irq_wait_queue);
 
 	platform_set_drvdata(pdev, dsi);
 
-	return component_add(&pdev->dev, &mtk_dsi_component_ops);
+	ret = component_add(&pdev->dev, &mtk_dsi_component_ops);
+	if (ret) {
+		dev_err(&pdev->dev, "failed to add component: %d\n", ret);
+		goto err_unregister_host;
+	}
+
+	return 0;
+
+err_unregister_host:
+	mipi_dsi_host_unregister(&dsi->host);
+	return ret;
 }
 
 static int mtk_dsi_remove(struct platform_device *pdev)
@@ -1183,13 +1270,32 @@ static int mtk_dsi_remove(struct platform_device *pdev)
 
 	mtk_output_dsi_disable(dsi);
 	component_del(&pdev->dev, &mtk_dsi_component_ops);
+	mipi_dsi_host_unregister(&dsi->host);
 
 	return 0;
 }
 
+static const struct mtk_dsi_driver_data mt8173_dsi_driver_data = {
+	.reg_cmdq_off = 0x200,
+};
+
+static const struct mtk_dsi_driver_data mt2701_dsi_driver_data = {
+	.reg_cmdq_off = 0x180,
+};
+
+static const struct mtk_dsi_driver_data mt8183_dsi_driver_data = {
+	.reg_cmdq_off = 0x200,
+	.has_shadow_ctl = true,
+	.has_size_ctl = true,
+};
+
 static const struct of_device_id mtk_dsi_of_match[] = {
-	{ .compatible = "mediatek,mt2701-dsi" },
-	{ .compatible = "mediatek,mt8173-dsi" },
+	{ .compatible = "mediatek,mt2701-dsi",
+	  .data = &mt2701_dsi_driver_data },
+	{ .compatible = "mediatek,mt8173-dsi",
+	  .data = &mt8173_dsi_driver_data },
+	{ .compatible = "mediatek,mt8183-dsi",
+	  .data = &mt8183_dsi_driver_data },
 	{ },
 };
 
diff --git a/drivers/gpu/drm/mediatek/mtk_hdmi.c b/drivers/gpu/drm/mediatek/mtk_hdmi.c
index ce91b61364eb..c79b1f855d89 100644
--- a/drivers/gpu/drm/mediatek/mtk_hdmi.c
+++ b/drivers/gpu/drm/mediatek/mtk_hdmi.c
@@ -23,6 +23,7 @@
 #include <sound/hdmi-codec.h>
 
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_crtc.h>
 #include <drm/drm_edid.h>
 #include <drm/drm_print.h>
diff --git a/drivers/gpu/drm/mediatek/mtk_mipi_tx.c b/drivers/gpu/drm/mediatek/mtk_mipi_tx.c
index 1842dc2caae9..e4d34484ecc8 100644
--- a/drivers/gpu/drm/mediatek/mtk_mipi_tx.c
+++ b/drivers/gpu/drm/mediatek/mtk_mipi_tx.c
@@ -3,292 +3,39 @@
  * Copyright (c) 2015 MediaTek Inc.
  */
 
-#include <linux/clk.h>
-#include <linux/clk-provider.h>
-#include <linux/delay.h>
-#include <linux/io.h>
-#include <linux/module.h>
-#include <linux/of_device.h>
-#include <linux/platform_device.h>
-#include <linux/phy/phy.h>
-
-#define MIPITX_DSI_CON		0x00
-#define RG_DSI_LDOCORE_EN		BIT(0)
-#define RG_DSI_CKG_LDOOUT_EN		BIT(1)
-#define RG_DSI_BCLK_SEL			(3 << 2)
-#define RG_DSI_LD_IDX_SEL		(7 << 4)
-#define RG_DSI_PHYCLK_SEL		(2 << 8)
-#define RG_DSI_DSICLK_FREQ_SEL		BIT(10)
-#define RG_DSI_LPTX_CLMP_EN		BIT(11)
-
-#define MIPITX_DSI_CLOCK_LANE	0x04
-#define MIPITX_DSI_DATA_LANE0	0x08
-#define MIPITX_DSI_DATA_LANE1	0x0c
-#define MIPITX_DSI_DATA_LANE2	0x10
-#define MIPITX_DSI_DATA_LANE3	0x14
-#define RG_DSI_LNTx_LDOOUT_EN		BIT(0)
-#define RG_DSI_LNTx_CKLANE_EN		BIT(1)
-#define RG_DSI_LNTx_LPTX_IPLUS1		BIT(2)
-#define RG_DSI_LNTx_LPTX_IPLUS2		BIT(3)
-#define RG_DSI_LNTx_LPTX_IMINUS		BIT(4)
-#define RG_DSI_LNTx_LPCD_IPLUS		BIT(5)
-#define RG_DSI_LNTx_LPCD_IMINUS		BIT(6)
-#define RG_DSI_LNTx_RT_CODE		(0xf << 8)
-
-#define MIPITX_DSI_TOP_CON	0x40
-#define RG_DSI_LNT_INTR_EN		BIT(0)
-#define RG_DSI_LNT_HS_BIAS_EN		BIT(1)
-#define RG_DSI_LNT_IMP_CAL_EN		BIT(2)
-#define RG_DSI_LNT_TESTMODE_EN		BIT(3)
-#define RG_DSI_LNT_IMP_CAL_CODE		(0xf << 4)
-#define RG_DSI_LNT_AIO_SEL		(7 << 8)
-#define RG_DSI_PAD_TIE_LOW_EN		BIT(11)
-#define RG_DSI_DEBUG_INPUT_EN		BIT(12)
-#define RG_DSI_PRESERVE			(7 << 13)
-
-#define MIPITX_DSI_BG_CON	0x44
-#define RG_DSI_BG_CORE_EN		BIT(0)
-#define RG_DSI_BG_CKEN			BIT(1)
-#define RG_DSI_BG_DIV			(0x3 << 2)
-#define RG_DSI_BG_FAST_CHARGE		BIT(4)
-#define RG_DSI_VOUT_MSK			(0x3ffff << 5)
-#define RG_DSI_V12_SEL			(7 << 5)
-#define RG_DSI_V10_SEL			(7 << 8)
-#define RG_DSI_V072_SEL			(7 << 11)
-#define RG_DSI_V04_SEL			(7 << 14)
-#define RG_DSI_V032_SEL			(7 << 17)
-#define RG_DSI_V02_SEL			(7 << 20)
-#define RG_DSI_BG_R1_TRIM		(0xf << 24)
-#define RG_DSI_BG_R2_TRIM		(0xf << 28)
-
-#define MIPITX_DSI_PLL_CON0	0x50
-#define RG_DSI_MPPLL_PLL_EN		BIT(0)
-#define RG_DSI_MPPLL_DIV_MSK		(0x1ff << 1)
-#define RG_DSI_MPPLL_PREDIV		(3 << 1)
-#define RG_DSI_MPPLL_TXDIV0		(3 << 3)
-#define RG_DSI_MPPLL_TXDIV1		(3 << 5)
-#define RG_DSI_MPPLL_POSDIV		(7 << 7)
-#define RG_DSI_MPPLL_MONVC_EN		BIT(10)
-#define RG_DSI_MPPLL_MONREF_EN		BIT(11)
-#define RG_DSI_MPPLL_VOD_EN		BIT(12)
-
-#define MIPITX_DSI_PLL_CON1	0x54
-#define RG_DSI_MPPLL_SDM_FRA_EN		BIT(0)
-#define RG_DSI_MPPLL_SDM_SSC_PH_INIT	BIT(1)
-#define RG_DSI_MPPLL_SDM_SSC_EN		BIT(2)
-#define RG_DSI_MPPLL_SDM_SSC_PRD	(0xffff << 16)
-
-#define MIPITX_DSI_PLL_CON2	0x58
-
-#define MIPITX_DSI_PLL_TOP	0x64
-#define RG_DSI_MPPLL_PRESERVE		(0xff << 8)
-
-#define MIPITX_DSI_PLL_PWR	0x68
-#define RG_DSI_MPPLL_SDM_PWR_ON		BIT(0)
-#define RG_DSI_MPPLL_SDM_ISO_EN		BIT(1)
-#define RG_DSI_MPPLL_SDM_PWR_ACK	BIT(8)
-
-#define MIPITX_DSI_SW_CTRL	0x80
-#define SW_CTRL_EN			BIT(0)
-
-#define MIPITX_DSI_SW_CTRL_CON0	0x84
-#define SW_LNTC_LPTX_PRE_OE		BIT(0)
-#define SW_LNTC_LPTX_OE			BIT(1)
-#define SW_LNTC_LPTX_P			BIT(2)
-#define SW_LNTC_LPTX_N			BIT(3)
-#define SW_LNTC_HSTX_PRE_OE		BIT(4)
-#define SW_LNTC_HSTX_OE			BIT(5)
-#define SW_LNTC_HSTX_ZEROCLK		BIT(6)
-#define SW_LNT0_LPTX_PRE_OE		BIT(7)
-#define SW_LNT0_LPTX_OE			BIT(8)
-#define SW_LNT0_LPTX_P			BIT(9)
-#define SW_LNT0_LPTX_N			BIT(10)
-#define SW_LNT0_HSTX_PRE_OE		BIT(11)
-#define SW_LNT0_HSTX_OE			BIT(12)
-#define SW_LNT0_LPRX_EN			BIT(13)
-#define SW_LNT1_LPTX_PRE_OE		BIT(14)
-#define SW_LNT1_LPTX_OE			BIT(15)
-#define SW_LNT1_LPTX_P			BIT(16)
-#define SW_LNT1_LPTX_N			BIT(17)
-#define SW_LNT1_HSTX_PRE_OE		BIT(18)
-#define SW_LNT1_HSTX_OE			BIT(19)
-#define SW_LNT2_LPTX_PRE_OE		BIT(20)
-#define SW_LNT2_LPTX_OE			BIT(21)
-#define SW_LNT2_LPTX_P			BIT(22)
-#define SW_LNT2_LPTX_N			BIT(23)
-#define SW_LNT2_HSTX_PRE_OE		BIT(24)
-#define SW_LNT2_HSTX_OE			BIT(25)
-
-struct mtk_mipitx_data {
-	const u32 mppll_preserve;
-};
-
-struct mtk_mipi_tx {
-	struct device *dev;
-	void __iomem *regs;
-	u32 data_rate;
-	const struct mtk_mipitx_data *driver_data;
-	struct clk_hw pll_hw;
-	struct clk *pll;
-};
+#include "mtk_mipi_tx.h"
 
-static inline struct mtk_mipi_tx *mtk_mipi_tx_from_clk_hw(struct clk_hw *hw)
+inline struct mtk_mipi_tx *mtk_mipi_tx_from_clk_hw(struct clk_hw *hw)
 {
 	return container_of(hw, struct mtk_mipi_tx, pll_hw);
 }
 
-static void mtk_mipi_tx_clear_bits(struct mtk_mipi_tx *mipi_tx, u32 offset,
-				   u32 bits)
+void mtk_mipi_tx_clear_bits(struct mtk_mipi_tx *mipi_tx, u32 offset,
+			    u32 bits)
 {
 	u32 temp = readl(mipi_tx->regs + offset);
 
 	writel(temp & ~bits, mipi_tx->regs + offset);
 }
 
-static void mtk_mipi_tx_set_bits(struct mtk_mipi_tx *mipi_tx, u32 offset,
-				 u32 bits)
+void mtk_mipi_tx_set_bits(struct mtk_mipi_tx *mipi_tx, u32 offset,
+			  u32 bits)
 {
 	u32 temp = readl(mipi_tx->regs + offset);
 
 	writel(temp | bits, mipi_tx->regs + offset);
 }
 
-static void mtk_mipi_tx_update_bits(struct mtk_mipi_tx *mipi_tx, u32 offset,
-				    u32 mask, u32 data)
+void mtk_mipi_tx_update_bits(struct mtk_mipi_tx *mipi_tx, u32 offset,
+			     u32 mask, u32 data)
 {
 	u32 temp = readl(mipi_tx->regs + offset);
 
 	writel((temp & ~mask) | (data & mask), mipi_tx->regs + offset);
 }
 
-static int mtk_mipi_tx_pll_prepare(struct clk_hw *hw)
-{
-	struct mtk_mipi_tx *mipi_tx = mtk_mipi_tx_from_clk_hw(hw);
-	u8 txdiv, txdiv0, txdiv1;
-	u64 pcw;
-
-	dev_dbg(mipi_tx->dev, "prepare: %u Hz\n", mipi_tx->data_rate);
-
-	if (mipi_tx->data_rate >= 500000000) {
-		txdiv = 1;
-		txdiv0 = 0;
-		txdiv1 = 0;
-	} else if (mipi_tx->data_rate >= 250000000) {
-		txdiv = 2;
-		txdiv0 = 1;
-		txdiv1 = 0;
-	} else if (mipi_tx->data_rate >= 125000000) {
-		txdiv = 4;
-		txdiv0 = 2;
-		txdiv1 = 0;
-	} else if (mipi_tx->data_rate > 62000000) {
-		txdiv = 8;
-		txdiv0 = 2;
-		txdiv1 = 1;
-	} else if (mipi_tx->data_rate >= 50000000) {
-		txdiv = 16;
-		txdiv0 = 2;
-		txdiv1 = 2;
-	} else {
-		return -EINVAL;
-	}
-
-	mtk_mipi_tx_update_bits(mipi_tx, MIPITX_DSI_BG_CON,
-				RG_DSI_VOUT_MSK |
-				RG_DSI_BG_CKEN | RG_DSI_BG_CORE_EN,
-				(4 << 20) | (4 << 17) | (4 << 14) |
-				(4 << 11) | (4 << 8) | (4 << 5) |
-				RG_DSI_BG_CKEN | RG_DSI_BG_CORE_EN);
-
-	usleep_range(30, 100);
-
-	mtk_mipi_tx_update_bits(mipi_tx, MIPITX_DSI_TOP_CON,
-				RG_DSI_LNT_IMP_CAL_CODE | RG_DSI_LNT_HS_BIAS_EN,
-				(8 << 4) | RG_DSI_LNT_HS_BIAS_EN);
-
-	mtk_mipi_tx_set_bits(mipi_tx, MIPITX_DSI_CON,
-			     RG_DSI_CKG_LDOOUT_EN | RG_DSI_LDOCORE_EN);
-
-	mtk_mipi_tx_update_bits(mipi_tx, MIPITX_DSI_PLL_PWR,
-				RG_DSI_MPPLL_SDM_PWR_ON |
-				RG_DSI_MPPLL_SDM_ISO_EN,
-				RG_DSI_MPPLL_SDM_PWR_ON);
-
-	mtk_mipi_tx_clear_bits(mipi_tx, MIPITX_DSI_PLL_CON0,
-			       RG_DSI_MPPLL_PLL_EN);
-
-	mtk_mipi_tx_update_bits(mipi_tx, MIPITX_DSI_PLL_CON0,
-				RG_DSI_MPPLL_TXDIV0 | RG_DSI_MPPLL_TXDIV1 |
-				RG_DSI_MPPLL_PREDIV,
-				(txdiv0 << 3) | (txdiv1 << 5));
-
-	/*
-	 * PLL PCW config
-	 * PCW bit 24~30 = integer part of pcw
-	 * PCW bit 0~23 = fractional part of pcw
-	 * pcw = data_Rate*4*txdiv/(Ref_clk*2);
-	 * Post DIV =4, so need data_Rate*4
-	 * Ref_clk is 26MHz
-	 */
-	pcw = div_u64(((u64)mipi_tx->data_rate * 2 * txdiv) << 24,
-		      26000000);
-	writel(pcw, mipi_tx->regs + MIPITX_DSI_PLL_CON2);
-
-	mtk_mipi_tx_set_bits(mipi_tx, MIPITX_DSI_PLL_CON1,
-			     RG_DSI_MPPLL_SDM_FRA_EN);
-
-	mtk_mipi_tx_set_bits(mipi_tx, MIPITX_DSI_PLL_CON0, RG_DSI_MPPLL_PLL_EN);
-
-	usleep_range(20, 100);
-
-	mtk_mipi_tx_clear_bits(mipi_tx, MIPITX_DSI_PLL_CON1,
-			       RG_DSI_MPPLL_SDM_SSC_EN);
-
-	mtk_mipi_tx_update_bits(mipi_tx, MIPITX_DSI_PLL_TOP,
-				RG_DSI_MPPLL_PRESERVE,
-				mipi_tx->driver_data->mppll_preserve);
-
-	return 0;
-}
-
-static void mtk_mipi_tx_pll_unprepare(struct clk_hw *hw)
-{
-	struct mtk_mipi_tx *mipi_tx = mtk_mipi_tx_from_clk_hw(hw);
-
-	dev_dbg(mipi_tx->dev, "unprepare\n");
-
-	mtk_mipi_tx_clear_bits(mipi_tx, MIPITX_DSI_PLL_CON0,
-			       RG_DSI_MPPLL_PLL_EN);
-
-	mtk_mipi_tx_update_bits(mipi_tx, MIPITX_DSI_PLL_TOP,
-				RG_DSI_MPPLL_PRESERVE, 0);
-
-	mtk_mipi_tx_update_bits(mipi_tx, MIPITX_DSI_PLL_PWR,
-				RG_DSI_MPPLL_SDM_ISO_EN |
-				RG_DSI_MPPLL_SDM_PWR_ON,
-				RG_DSI_MPPLL_SDM_ISO_EN);
-
-	mtk_mipi_tx_clear_bits(mipi_tx, MIPITX_DSI_TOP_CON,
-			       RG_DSI_LNT_HS_BIAS_EN);
-
-	mtk_mipi_tx_clear_bits(mipi_tx, MIPITX_DSI_CON,
-			       RG_DSI_CKG_LDOOUT_EN | RG_DSI_LDOCORE_EN);
-
-	mtk_mipi_tx_clear_bits(mipi_tx, MIPITX_DSI_BG_CON,
-			       RG_DSI_BG_CKEN | RG_DSI_BG_CORE_EN);
-
-	mtk_mipi_tx_clear_bits(mipi_tx, MIPITX_DSI_PLL_CON0,
-			       RG_DSI_MPPLL_DIV_MSK);
-}
-
-static long mtk_mipi_tx_pll_round_rate(struct clk_hw *hw, unsigned long rate,
-				       unsigned long *prate)
-{
-	return clamp_val(rate, 50000000, 1250000000);
-}
-
-static int mtk_mipi_tx_pll_set_rate(struct clk_hw *hw, unsigned long rate,
-				    unsigned long parent_rate)
+int mtk_mipi_tx_pll_set_rate(struct clk_hw *hw, unsigned long rate,
+			     unsigned long parent_rate)
 {
 	struct mtk_mipi_tx *mipi_tx = mtk_mipi_tx_from_clk_hw(hw);
 
@@ -299,37 +46,14 @@ static int mtk_mipi_tx_pll_set_rate(struct clk_hw *hw, unsigned long rate,
 	return 0;
 }
 
-static unsigned long mtk_mipi_tx_pll_recalc_rate(struct clk_hw *hw,
-						 unsigned long parent_rate)
+unsigned long mtk_mipi_tx_pll_recalc_rate(struct clk_hw *hw,
+					  unsigned long parent_rate)
 {
 	struct mtk_mipi_tx *mipi_tx = mtk_mipi_tx_from_clk_hw(hw);
 
 	return mipi_tx->data_rate;
 }
 
-static const struct clk_ops mtk_mipi_tx_pll_ops = {
-	.prepare = mtk_mipi_tx_pll_prepare,
-	.unprepare = mtk_mipi_tx_pll_unprepare,
-	.round_rate = mtk_mipi_tx_pll_round_rate,
-	.set_rate = mtk_mipi_tx_pll_set_rate,
-	.recalc_rate = mtk_mipi_tx_pll_recalc_rate,
-};
-
-static int mtk_mipi_tx_power_on_signal(struct phy *phy)
-{
-	struct mtk_mipi_tx *mipi_tx = phy_get_drvdata(phy);
-	u32 reg;
-
-	for (reg = MIPITX_DSI_CLOCK_LANE;
-	     reg <= MIPITX_DSI_DATA_LANE3; reg += 4)
-		mtk_mipi_tx_set_bits(mipi_tx, reg, RG_DSI_LNTx_LDOOUT_EN);
-
-	mtk_mipi_tx_clear_bits(mipi_tx, MIPITX_DSI_TOP_CON,
-			       RG_DSI_PAD_TIE_LOW_EN);
-
-	return 0;
-}
-
 static int mtk_mipi_tx_power_on(struct phy *phy)
 {
 	struct mtk_mipi_tx *mipi_tx = phy_get_drvdata(phy);
@@ -341,30 +65,16 @@ static int mtk_mipi_tx_power_on(struct phy *phy)
 		return ret;
 
 	/* Enable DSI Lane LDO outputs, disable pad tie low */
-	mtk_mipi_tx_power_on_signal(phy);
-
+	mipi_tx->driver_data->mipi_tx_enable_signal(phy);
 	return 0;
 }
 
-static void mtk_mipi_tx_power_off_signal(struct phy *phy)
-{
-	struct mtk_mipi_tx *mipi_tx = phy_get_drvdata(phy);
-	u32 reg;
-
-	mtk_mipi_tx_set_bits(mipi_tx, MIPITX_DSI_TOP_CON,
-			     RG_DSI_PAD_TIE_LOW_EN);
-
-	for (reg = MIPITX_DSI_CLOCK_LANE;
-	     reg <= MIPITX_DSI_DATA_LANE3; reg += 4)
-		mtk_mipi_tx_clear_bits(mipi_tx, reg, RG_DSI_LNTx_LDOOUT_EN);
-}
-
 static int mtk_mipi_tx_power_off(struct phy *phy)
 {
 	struct mtk_mipi_tx *mipi_tx = phy_get_drvdata(phy);
 
 	/* Enable pad tie low, disable DSI Lane LDO outputs */
-	mtk_mipi_tx_power_off_signal(phy);
+	mipi_tx->driver_data->mipi_tx_disable_signal(phy);
 
 	/* Disable PLL and power down core */
 	clk_disable_unprepare(mipi_tx->pll);
@@ -383,10 +93,9 @@ static int mtk_mipi_tx_probe(struct platform_device *pdev)
 	struct device *dev = &pdev->dev;
 	struct mtk_mipi_tx *mipi_tx;
 	struct resource *mem;
-	struct clk *ref_clk;
 	const char *ref_clk_name;
+	struct clk *ref_clk;
 	struct clk_init_data clk_init = {
-		.ops = &mtk_mipi_tx_pll_ops,
 		.num_parents = 1,
 		.parent_names = (const char * const *)&ref_clk_name,
 		.flags = CLK_SET_RATE_GATE,
@@ -400,6 +109,7 @@ static int mtk_mipi_tx_probe(struct platform_device *pdev)
 		return -ENOMEM;
 
 	mipi_tx->driver_data = of_device_get_match_data(dev);
+
 	mem = platform_get_resource(pdev, IORESOURCE_MEM, 0);
 	mipi_tx->regs = devm_ioremap_resource(dev, mem);
 	if (IS_ERR(mipi_tx->regs)) {
@@ -414,6 +124,7 @@ static int mtk_mipi_tx_probe(struct platform_device *pdev)
 		dev_err(dev, "Failed to get reference clock: %d\n", ret);
 		return ret;
 	}
+
 	ref_clk_name = __clk_get_name(ref_clk);
 
 	ret = of_property_read_string(dev->of_node, "clock-output-names",
@@ -423,6 +134,8 @@ static int mtk_mipi_tx_probe(struct platform_device *pdev)
 		return ret;
 	}
 
+	clk_init.ops = mipi_tx->driver_data->mipi_tx_clk_ops;
+
 	mipi_tx->pll_hw.init = &clk_init;
 	mipi_tx->pll = devm_clk_register(dev, &mipi_tx->pll_hw);
 	if (IS_ERR(mipi_tx->pll)) {
@@ -457,20 +170,14 @@ static int mtk_mipi_tx_remove(struct platform_device *pdev)
 	return 0;
 }
 
-static const struct mtk_mipitx_data mt2701_mipitx_data = {
-	.mppll_preserve = (3 << 8)
-};
-
-static const struct mtk_mipitx_data mt8173_mipitx_data = {
-	.mppll_preserve = (0 << 8)
-};
-
 static const struct of_device_id mtk_mipi_tx_match[] = {
 	{ .compatible = "mediatek,mt2701-mipi-tx",
 	  .data = &mt2701_mipitx_data },
 	{ .compatible = "mediatek,mt8173-mipi-tx",
 	  .data = &mt8173_mipitx_data },
-	{},
+	{ .compatible = "mediatek,mt8183-mipi-tx",
+	  .data = &mt8183_mipitx_data },
+	{ },
 };
 
 struct platform_driver mtk_mipi_tx_driver = {
@@ -481,3 +188,4 @@ struct platform_driver mtk_mipi_tx_driver = {
 		.of_match_table = mtk_mipi_tx_match,
 	},
 };
+
diff --git a/drivers/gpu/drm/mediatek/mtk_mipi_tx.h b/drivers/gpu/drm/mediatek/mtk_mipi_tx.h
new file mode 100644
index 000000000000..413f35d86219
--- /dev/null
+++ b/drivers/gpu/drm/mediatek/mtk_mipi_tx.h
@@ -0,0 +1,49 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * Copyright (c) 2019 MediaTek Inc.
+ * Author: Jitao Shi <jitao.shi@mediatek.com>
+ */
+
+#ifndef _MTK_MIPI_TX_H
+#define _MTK_MIPI_TX_H
+
+#include <linux/clk.h>
+#include <linux/clk-provider.h>
+#include <linux/delay.h>
+#include <linux/io.h>
+#include <linux/module.h>
+#include <linux/of_device.h>
+#include <linux/platform_device.h>
+#include <linux/phy/phy.h>
+
+struct mtk_mipitx_data {
+	const u32 mppll_preserve;
+	const struct clk_ops *mipi_tx_clk_ops;
+	void (*mipi_tx_enable_signal)(struct phy *phy);
+	void (*mipi_tx_disable_signal)(struct phy *phy);
+};
+
+struct mtk_mipi_tx {
+	struct device *dev;
+	void __iomem *regs;
+	u32 data_rate;
+	const struct mtk_mipitx_data *driver_data;
+	struct clk_hw pll_hw;
+	struct clk *pll;
+};
+
+struct mtk_mipi_tx *mtk_mipi_tx_from_clk_hw(struct clk_hw *hw);
+void mtk_mipi_tx_clear_bits(struct mtk_mipi_tx *mipi_tx, u32 offset, u32 bits);
+void mtk_mipi_tx_set_bits(struct mtk_mipi_tx *mipi_tx, u32 offset, u32 bits);
+void mtk_mipi_tx_update_bits(struct mtk_mipi_tx *mipi_tx, u32 offset, u32 mask,
+			     u32 data);
+int mtk_mipi_tx_pll_set_rate(struct clk_hw *hw, unsigned long rate,
+			     unsigned long parent_rate);
+unsigned long mtk_mipi_tx_pll_recalc_rate(struct clk_hw *hw,
+					  unsigned long parent_rate);
+
+extern const struct mtk_mipitx_data mt2701_mipitx_data;
+extern const struct mtk_mipitx_data mt8173_mipitx_data;
+extern const struct mtk_mipitx_data mt8183_mipitx_data;
+
+#endif
diff --git a/drivers/gpu/drm/mediatek/mtk_mt8173_mipi_tx.c b/drivers/gpu/drm/mediatek/mtk_mt8173_mipi_tx.c
new file mode 100644
index 000000000000..f18db14d8b63
--- /dev/null
+++ b/drivers/gpu/drm/mediatek/mtk_mt8173_mipi_tx.c
@@ -0,0 +1,288 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (c) 2019 MediaTek Inc.
+ * Author: jitao.shi <jitao.shi@mediatek.com>
+ */
+
+#include "mtk_mipi_tx.h"
+
+#define MIPITX_DSI_CON		0x00
+#define RG_DSI_LDOCORE_EN		BIT(0)
+#define RG_DSI_CKG_LDOOUT_EN		BIT(1)
+#define RG_DSI_BCLK_SEL			(3 << 2)
+#define RG_DSI_LD_IDX_SEL		(7 << 4)
+#define RG_DSI_PHYCLK_SEL		(2 << 8)
+#define RG_DSI_DSICLK_FREQ_SEL		BIT(10)
+#define RG_DSI_LPTX_CLMP_EN		BIT(11)
+
+#define MIPITX_DSI_CLOCK_LANE	0x04
+#define MIPITX_DSI_DATA_LANE0	0x08
+#define MIPITX_DSI_DATA_LANE1	0x0c
+#define MIPITX_DSI_DATA_LANE2	0x10
+#define MIPITX_DSI_DATA_LANE3	0x14
+#define RG_DSI_LNTx_LDOOUT_EN		BIT(0)
+#define RG_DSI_LNTx_CKLANE_EN		BIT(1)
+#define RG_DSI_LNTx_LPTX_IPLUS1		BIT(2)
+#define RG_DSI_LNTx_LPTX_IPLUS2		BIT(3)
+#define RG_DSI_LNTx_LPTX_IMINUS		BIT(4)
+#define RG_DSI_LNTx_LPCD_IPLUS		BIT(5)
+#define RG_DSI_LNTx_LPCD_IMINUS		BIT(6)
+#define RG_DSI_LNTx_RT_CODE		(0xf << 8)
+
+#define MIPITX_DSI_TOP_CON	0x40
+#define RG_DSI_LNT_INTR_EN		BIT(0)
+#define RG_DSI_LNT_HS_BIAS_EN		BIT(1)
+#define RG_DSI_LNT_IMP_CAL_EN		BIT(2)
+#define RG_DSI_LNT_TESTMODE_EN		BIT(3)
+#define RG_DSI_LNT_IMP_CAL_CODE		(0xf << 4)
+#define RG_DSI_LNT_AIO_SEL		(7 << 8)
+#define RG_DSI_PAD_TIE_LOW_EN		BIT(11)
+#define RG_DSI_DEBUG_INPUT_EN		BIT(12)
+#define RG_DSI_PRESERVE			(7 << 13)
+
+#define MIPITX_DSI_BG_CON	0x44
+#define RG_DSI_BG_CORE_EN		BIT(0)
+#define RG_DSI_BG_CKEN			BIT(1)
+#define RG_DSI_BG_DIV			(0x3 << 2)
+#define RG_DSI_BG_FAST_CHARGE		BIT(4)
+#define RG_DSI_VOUT_MSK			(0x3ffff << 5)
+#define RG_DSI_V12_SEL			(7 << 5)
+#define RG_DSI_V10_SEL			(7 << 8)
+#define RG_DSI_V072_SEL			(7 << 11)
+#define RG_DSI_V04_SEL			(7 << 14)
+#define RG_DSI_V032_SEL			(7 << 17)
+#define RG_DSI_V02_SEL			(7 << 20)
+#define RG_DSI_BG_R1_TRIM		(0xf << 24)
+#define RG_DSI_BG_R2_TRIM		(0xf << 28)
+
+#define MIPITX_DSI_PLL_CON0	0x50
+#define RG_DSI_MPPLL_PLL_EN		BIT(0)
+#define RG_DSI_MPPLL_DIV_MSK		(0x1ff << 1)
+#define RG_DSI_MPPLL_PREDIV		(3 << 1)
+#define RG_DSI_MPPLL_TXDIV0		(3 << 3)
+#define RG_DSI_MPPLL_TXDIV1		(3 << 5)
+#define RG_DSI_MPPLL_POSDIV		(7 << 7)
+#define RG_DSI_MPPLL_MONVC_EN		BIT(10)
+#define RG_DSI_MPPLL_MONREF_EN		BIT(11)
+#define RG_DSI_MPPLL_VOD_EN		BIT(12)
+
+#define MIPITX_DSI_PLL_CON1	0x54
+#define RG_DSI_MPPLL_SDM_FRA_EN		BIT(0)
+#define RG_DSI_MPPLL_SDM_SSC_PH_INIT	BIT(1)
+#define RG_DSI_MPPLL_SDM_SSC_EN		BIT(2)
+#define RG_DSI_MPPLL_SDM_SSC_PRD	(0xffff << 16)
+
+#define MIPITX_DSI_PLL_CON2	0x58
+
+#define MIPITX_DSI_PLL_TOP	0x64
+#define RG_DSI_MPPLL_PRESERVE		(0xff << 8)
+
+#define MIPITX_DSI_PLL_PWR	0x68
+#define RG_DSI_MPPLL_SDM_PWR_ON		BIT(0)
+#define RG_DSI_MPPLL_SDM_ISO_EN		BIT(1)
+#define RG_DSI_MPPLL_SDM_PWR_ACK	BIT(8)
+
+#define MIPITX_DSI_SW_CTRL	0x80
+#define SW_CTRL_EN			BIT(0)
+
+#define MIPITX_DSI_SW_CTRL_CON0	0x84
+#define SW_LNTC_LPTX_PRE_OE		BIT(0)
+#define SW_LNTC_LPTX_OE			BIT(1)
+#define SW_LNTC_LPTX_P			BIT(2)
+#define SW_LNTC_LPTX_N			BIT(3)
+#define SW_LNTC_HSTX_PRE_OE		BIT(4)
+#define SW_LNTC_HSTX_OE			BIT(5)
+#define SW_LNTC_HSTX_ZEROCLK		BIT(6)
+#define SW_LNT0_LPTX_PRE_OE		BIT(7)
+#define SW_LNT0_LPTX_OE			BIT(8)
+#define SW_LNT0_LPTX_P			BIT(9)
+#define SW_LNT0_LPTX_N			BIT(10)
+#define SW_LNT0_HSTX_PRE_OE		BIT(11)
+#define SW_LNT0_HSTX_OE			BIT(12)
+#define SW_LNT0_LPRX_EN			BIT(13)
+#define SW_LNT1_LPTX_PRE_OE		BIT(14)
+#define SW_LNT1_LPTX_OE			BIT(15)
+#define SW_LNT1_LPTX_P			BIT(16)
+#define SW_LNT1_LPTX_N			BIT(17)
+#define SW_LNT1_HSTX_PRE_OE		BIT(18)
+#define SW_LNT1_HSTX_OE			BIT(19)
+#define SW_LNT2_LPTX_PRE_OE		BIT(20)
+#define SW_LNT2_LPTX_OE			BIT(21)
+#define SW_LNT2_LPTX_P			BIT(22)
+#define SW_LNT2_LPTX_N			BIT(23)
+#define SW_LNT2_HSTX_PRE_OE		BIT(24)
+#define SW_LNT2_HSTX_OE			BIT(25)
+
+static int mtk_mipi_tx_pll_prepare(struct clk_hw *hw)
+{
+	struct mtk_mipi_tx *mipi_tx = mtk_mipi_tx_from_clk_hw(hw);
+	u8 txdiv, txdiv0, txdiv1;
+	u64 pcw;
+
+	dev_dbg(mipi_tx->dev, "prepare: %u Hz\n", mipi_tx->data_rate);
+
+	if (mipi_tx->data_rate >= 500000000) {
+		txdiv = 1;
+		txdiv0 = 0;
+		txdiv1 = 0;
+	} else if (mipi_tx->data_rate >= 250000000) {
+		txdiv = 2;
+		txdiv0 = 1;
+		txdiv1 = 0;
+	} else if (mipi_tx->data_rate >= 125000000) {
+		txdiv = 4;
+		txdiv0 = 2;
+		txdiv1 = 0;
+	} else if (mipi_tx->data_rate > 62000000) {
+		txdiv = 8;
+		txdiv0 = 2;
+		txdiv1 = 1;
+	} else if (mipi_tx->data_rate >= 50000000) {
+		txdiv = 16;
+		txdiv0 = 2;
+		txdiv1 = 2;
+	} else {
+		return -EINVAL;
+	}
+
+	mtk_mipi_tx_update_bits(mipi_tx, MIPITX_DSI_BG_CON,
+				RG_DSI_VOUT_MSK |
+				RG_DSI_BG_CKEN | RG_DSI_BG_CORE_EN,
+				(4 << 20) | (4 << 17) | (4 << 14) |
+				(4 << 11) | (4 << 8) | (4 << 5) |
+				RG_DSI_BG_CKEN | RG_DSI_BG_CORE_EN);
+
+	usleep_range(30, 100);
+
+	mtk_mipi_tx_update_bits(mipi_tx, MIPITX_DSI_TOP_CON,
+				RG_DSI_LNT_IMP_CAL_CODE | RG_DSI_LNT_HS_BIAS_EN,
+				(8 << 4) | RG_DSI_LNT_HS_BIAS_EN);
+
+	mtk_mipi_tx_set_bits(mipi_tx, MIPITX_DSI_CON,
+			     RG_DSI_CKG_LDOOUT_EN | RG_DSI_LDOCORE_EN);
+
+	mtk_mipi_tx_update_bits(mipi_tx, MIPITX_DSI_PLL_PWR,
+				RG_DSI_MPPLL_SDM_PWR_ON |
+				RG_DSI_MPPLL_SDM_ISO_EN,
+				RG_DSI_MPPLL_SDM_PWR_ON);
+
+	mtk_mipi_tx_clear_bits(mipi_tx, MIPITX_DSI_PLL_CON0,
+			       RG_DSI_MPPLL_PLL_EN);
+
+	mtk_mipi_tx_update_bits(mipi_tx, MIPITX_DSI_PLL_CON0,
+				RG_DSI_MPPLL_TXDIV0 | RG_DSI_MPPLL_TXDIV1 |
+				RG_DSI_MPPLL_PREDIV,
+				(txdiv0 << 3) | (txdiv1 << 5));
+
+	/*
+	 * PLL PCW config
+	 * PCW bit 24~30 = integer part of pcw
+	 * PCW bit 0~23 = fractional part of pcw
+	 * pcw = data_Rate*4*txdiv/(Ref_clk*2);
+	 * Post DIV =4, so need data_Rate*4
+	 * Ref_clk is 26MHz
+	 */
+	pcw = div_u64(((u64)mipi_tx->data_rate * 2 * txdiv) << 24,
+		      26000000);
+	writel(pcw, mipi_tx->regs + MIPITX_DSI_PLL_CON2);
+
+	mtk_mipi_tx_set_bits(mipi_tx, MIPITX_DSI_PLL_CON1,
+			     RG_DSI_MPPLL_SDM_FRA_EN);
+
+	mtk_mipi_tx_set_bits(mipi_tx, MIPITX_DSI_PLL_CON0, RG_DSI_MPPLL_PLL_EN);
+
+	usleep_range(20, 100);
+
+	mtk_mipi_tx_clear_bits(mipi_tx, MIPITX_DSI_PLL_CON1,
+			       RG_DSI_MPPLL_SDM_SSC_EN);
+
+	mtk_mipi_tx_update_bits(mipi_tx, MIPITX_DSI_PLL_TOP,
+				RG_DSI_MPPLL_PRESERVE,
+				mipi_tx->driver_data->mppll_preserve);
+
+	return 0;
+}
+
+static void mtk_mipi_tx_pll_unprepare(struct clk_hw *hw)
+{
+	struct mtk_mipi_tx *mipi_tx = mtk_mipi_tx_from_clk_hw(hw);
+
+	dev_dbg(mipi_tx->dev, "unprepare\n");
+
+	mtk_mipi_tx_clear_bits(mipi_tx, MIPITX_DSI_PLL_CON0,
+			       RG_DSI_MPPLL_PLL_EN);
+
+	mtk_mipi_tx_update_bits(mipi_tx, MIPITX_DSI_PLL_TOP,
+				RG_DSI_MPPLL_PRESERVE, 0);
+
+	mtk_mipi_tx_update_bits(mipi_tx, MIPITX_DSI_PLL_PWR,
+				RG_DSI_MPPLL_SDM_ISO_EN |
+				RG_DSI_MPPLL_SDM_PWR_ON,
+				RG_DSI_MPPLL_SDM_ISO_EN);
+
+	mtk_mipi_tx_clear_bits(mipi_tx, MIPITX_DSI_TOP_CON,
+			       RG_DSI_LNT_HS_BIAS_EN);
+
+	mtk_mipi_tx_clear_bits(mipi_tx, MIPITX_DSI_CON,
+			       RG_DSI_CKG_LDOOUT_EN | RG_DSI_LDOCORE_EN);
+
+	mtk_mipi_tx_clear_bits(mipi_tx, MIPITX_DSI_BG_CON,
+			       RG_DSI_BG_CKEN | RG_DSI_BG_CORE_EN);
+
+	mtk_mipi_tx_clear_bits(mipi_tx, MIPITX_DSI_PLL_CON0,
+			       RG_DSI_MPPLL_DIV_MSK);
+}
+
+static long mtk_mipi_tx_pll_round_rate(struct clk_hw *hw, unsigned long rate,
+				       unsigned long *prate)
+{
+	return clamp_val(rate, 50000000, 1250000000);
+}
+
+static const struct clk_ops mtk_mipi_tx_pll_ops = {
+	.prepare = mtk_mipi_tx_pll_prepare,
+	.unprepare = mtk_mipi_tx_pll_unprepare,
+	.round_rate = mtk_mipi_tx_pll_round_rate,
+	.set_rate = mtk_mipi_tx_pll_set_rate,
+	.recalc_rate = mtk_mipi_tx_pll_recalc_rate,
+};
+
+static void mtk_mipi_tx_power_on_signal(struct phy *phy)
+{
+	struct mtk_mipi_tx *mipi_tx = phy_get_drvdata(phy);
+	u32 reg;
+
+	for (reg = MIPITX_DSI_CLOCK_LANE;
+	     reg <= MIPITX_DSI_DATA_LANE3; reg += 4)
+		mtk_mipi_tx_set_bits(mipi_tx, reg, RG_DSI_LNTx_LDOOUT_EN);
+
+	mtk_mipi_tx_clear_bits(mipi_tx, MIPITX_DSI_TOP_CON,
+			       RG_DSI_PAD_TIE_LOW_EN);
+}
+
+static void mtk_mipi_tx_power_off_signal(struct phy *phy)
+{
+	struct mtk_mipi_tx *mipi_tx = phy_get_drvdata(phy);
+	u32 reg;
+
+	mtk_mipi_tx_set_bits(mipi_tx, MIPITX_DSI_TOP_CON,
+			     RG_DSI_PAD_TIE_LOW_EN);
+
+	for (reg = MIPITX_DSI_CLOCK_LANE;
+	     reg <= MIPITX_DSI_DATA_LANE3; reg += 4)
+		mtk_mipi_tx_clear_bits(mipi_tx, reg, RG_DSI_LNTx_LDOOUT_EN);
+}
+
+const struct mtk_mipitx_data mt2701_mipitx_data = {
+	.mppll_preserve = (3 << 8),
+	.mipi_tx_clk_ops = &mtk_mipi_tx_pll_ops,
+	.mipi_tx_enable_signal = mtk_mipi_tx_power_on_signal,
+	.mipi_tx_disable_signal = mtk_mipi_tx_power_off_signal,
+};
+
+const struct mtk_mipitx_data mt8173_mipitx_data = {
+	.mppll_preserve = (0 << 8),
+	.mipi_tx_clk_ops = &mtk_mipi_tx_pll_ops,
+	.mipi_tx_enable_signal = mtk_mipi_tx_power_on_signal,
+	.mipi_tx_disable_signal = mtk_mipi_tx_power_off_signal,
+};
diff --git a/drivers/gpu/drm/mediatek/mtk_mt8183_mipi_tx.c b/drivers/gpu/drm/mediatek/mtk_mt8183_mipi_tx.c
new file mode 100644
index 000000000000..91f08a351fd0
--- /dev/null
+++ b/drivers/gpu/drm/mediatek/mtk_mt8183_mipi_tx.c
@@ -0,0 +1,149 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (c) 2019 MediaTek Inc.
+ * Author: jitao.shi <jitao.shi@mediatek.com>
+ */
+
+#include "mtk_mipi_tx.h"
+
+#define MIPITX_LANE_CON		0x000c
+#define RG_DSI_CPHY_T1DRV_EN		BIT(0)
+#define RG_DSI_ANA_CK_SEL		BIT(1)
+#define RG_DSI_PHY_CK_SEL		BIT(2)
+#define RG_DSI_CPHY_EN			BIT(3)
+#define RG_DSI_PHYCK_INV_EN		BIT(4)
+#define RG_DSI_PWR04_EN			BIT(5)
+#define RG_DSI_BG_LPF_EN		BIT(6)
+#define RG_DSI_BG_CORE_EN		BIT(7)
+#define RG_DSI_PAD_TIEL_SEL		BIT(8)
+
+#define MIPITX_PLL_PWR		0x0028
+#define MIPITX_PLL_CON0		0x002c
+#define MIPITX_PLL_CON1		0x0030
+#define MIPITX_PLL_CON2		0x0034
+#define MIPITX_PLL_CON3		0x0038
+#define MIPITX_PLL_CON4		0x003c
+#define RG_DSI_PLL_IBIAS		(3 << 10)
+
+#define MIPITX_D2_SW_CTL_EN	0x0144
+#define MIPITX_D0_SW_CTL_EN	0x0244
+#define MIPITX_CK_CKMODE_EN	0x0328
+#define DSI_CK_CKMODE_EN		BIT(0)
+#define MIPITX_CK_SW_CTL_EN	0x0344
+#define MIPITX_D1_SW_CTL_EN	0x0444
+#define MIPITX_D3_SW_CTL_EN	0x0544
+#define DSI_SW_CTL_EN			BIT(0)
+#define AD_DSI_PLL_SDM_PWR_ON		BIT(0)
+#define AD_DSI_PLL_SDM_ISO_EN		BIT(1)
+
+#define RG_DSI_PLL_EN			BIT(4)
+#define RG_DSI_PLL_POSDIV		(0x7 << 8)
+
+static int mtk_mipi_tx_pll_enable(struct clk_hw *hw)
+{
+	struct mtk_mipi_tx *mipi_tx = mtk_mipi_tx_from_clk_hw(hw);
+	unsigned int txdiv, txdiv0;
+	u64 pcw;
+
+	dev_dbg(mipi_tx->dev, "enable: %u bps\n", mipi_tx->data_rate);
+
+	if (mipi_tx->data_rate >= 2000000000) {
+		txdiv = 1;
+		txdiv0 = 0;
+	} else if (mipi_tx->data_rate >= 1000000000) {
+		txdiv = 2;
+		txdiv0 = 1;
+	} else if (mipi_tx->data_rate >= 500000000) {
+		txdiv = 4;
+		txdiv0 = 2;
+	} else if (mipi_tx->data_rate > 250000000) {
+		txdiv = 8;
+		txdiv0 = 3;
+	} else if (mipi_tx->data_rate >= 125000000) {
+		txdiv = 16;
+		txdiv0 = 4;
+	} else {
+		return -EINVAL;
+	}
+
+	mtk_mipi_tx_clear_bits(mipi_tx, MIPITX_PLL_CON4, RG_DSI_PLL_IBIAS);
+
+	mtk_mipi_tx_set_bits(mipi_tx, MIPITX_PLL_PWR, AD_DSI_PLL_SDM_PWR_ON);
+	mtk_mipi_tx_clear_bits(mipi_tx, MIPITX_PLL_CON1, RG_DSI_PLL_EN);
+	udelay(1);
+	mtk_mipi_tx_clear_bits(mipi_tx, MIPITX_PLL_PWR, AD_DSI_PLL_SDM_ISO_EN);
+	pcw = div_u64(((u64)mipi_tx->data_rate * txdiv) << 24, 26000000);
+	writel(pcw, mipi_tx->regs + MIPITX_PLL_CON0);
+	mtk_mipi_tx_update_bits(mipi_tx, MIPITX_PLL_CON1, RG_DSI_PLL_POSDIV,
+				txdiv0 << 8);
+	mtk_mipi_tx_set_bits(mipi_tx, MIPITX_PLL_CON1, RG_DSI_PLL_EN);
+
+	return 0;
+}
+
+static void mtk_mipi_tx_pll_disable(struct clk_hw *hw)
+{
+	struct mtk_mipi_tx *mipi_tx = mtk_mipi_tx_from_clk_hw(hw);
+
+	mtk_mipi_tx_clear_bits(mipi_tx, MIPITX_PLL_CON1, RG_DSI_PLL_EN);
+
+	mtk_mipi_tx_set_bits(mipi_tx, MIPITX_PLL_PWR, AD_DSI_PLL_SDM_ISO_EN);
+	mtk_mipi_tx_clear_bits(mipi_tx, MIPITX_PLL_PWR, AD_DSI_PLL_SDM_PWR_ON);
+}
+
+static long mtk_mipi_tx_pll_round_rate(struct clk_hw *hw, unsigned long rate,
+				       unsigned long *prate)
+{
+	return clamp_val(rate, 50000000, 1600000000);
+}
+
+static const struct clk_ops mtk_mipi_tx_pll_ops = {
+	.enable = mtk_mipi_tx_pll_enable,
+	.disable = mtk_mipi_tx_pll_disable,
+	.round_rate = mtk_mipi_tx_pll_round_rate,
+	.set_rate = mtk_mipi_tx_pll_set_rate,
+	.recalc_rate = mtk_mipi_tx_pll_recalc_rate,
+};
+
+static void mtk_mipi_tx_power_on_signal(struct phy *phy)
+{
+	struct mtk_mipi_tx *mipi_tx = phy_get_drvdata(phy);
+
+	/* BG_LPF_EN / BG_CORE_EN */
+	writel(RG_DSI_PAD_TIEL_SEL | RG_DSI_BG_CORE_EN,
+	       mipi_tx->regs + MIPITX_LANE_CON);
+	usleep_range(30, 100);
+	writel(RG_DSI_BG_CORE_EN | RG_DSI_BG_LPF_EN,
+	       mipi_tx->regs + MIPITX_LANE_CON);
+
+	/* Switch OFF each Lane */
+	mtk_mipi_tx_clear_bits(mipi_tx, MIPITX_D0_SW_CTL_EN, DSI_SW_CTL_EN);
+	mtk_mipi_tx_clear_bits(mipi_tx, MIPITX_D1_SW_CTL_EN, DSI_SW_CTL_EN);
+	mtk_mipi_tx_clear_bits(mipi_tx, MIPITX_D2_SW_CTL_EN, DSI_SW_CTL_EN);
+	mtk_mipi_tx_clear_bits(mipi_tx, MIPITX_D3_SW_CTL_EN, DSI_SW_CTL_EN);
+	mtk_mipi_tx_clear_bits(mipi_tx, MIPITX_CK_SW_CTL_EN, DSI_SW_CTL_EN);
+
+	mtk_mipi_tx_set_bits(mipi_tx, MIPITX_CK_CKMODE_EN, DSI_CK_CKMODE_EN);
+}
+
+static void mtk_mipi_tx_power_off_signal(struct phy *phy)
+{
+	struct mtk_mipi_tx *mipi_tx = phy_get_drvdata(phy);
+
+	/* Switch ON each Lane */
+	mtk_mipi_tx_set_bits(mipi_tx, MIPITX_D0_SW_CTL_EN, DSI_SW_CTL_EN);
+	mtk_mipi_tx_set_bits(mipi_tx, MIPITX_D1_SW_CTL_EN, DSI_SW_CTL_EN);
+	mtk_mipi_tx_set_bits(mipi_tx, MIPITX_D2_SW_CTL_EN, DSI_SW_CTL_EN);
+	mtk_mipi_tx_set_bits(mipi_tx, MIPITX_D3_SW_CTL_EN, DSI_SW_CTL_EN);
+	mtk_mipi_tx_set_bits(mipi_tx, MIPITX_CK_SW_CTL_EN, DSI_SW_CTL_EN);
+
+	writel(RG_DSI_PAD_TIEL_SEL | RG_DSI_BG_CORE_EN,
+	       mipi_tx->regs + MIPITX_LANE_CON);
+	writel(RG_DSI_PAD_TIEL_SEL, mipi_tx->regs + MIPITX_LANE_CON);
+}
+
+const struct mtk_mipitx_data mt8183_mipitx_data = {
+	.mipi_tx_clk_ops = &mtk_mipi_tx_pll_ops,
+	.mipi_tx_enable_signal = mtk_mipi_tx_power_on_signal,
+	.mipi_tx_disable_signal = mtk_mipi_tx_power_off_signal,
+};
diff --git a/drivers/gpu/drm/meson/meson_drv.c b/drivers/gpu/drm/meson/meson_drv.c
index a24f8dec5adc..397c33182f4f 100644
--- a/drivers/gpu/drm/meson/meson_drv.c
+++ b/drivers/gpu/drm/meson/meson_drv.c
@@ -372,6 +372,33 @@ static const struct component_master_ops meson_drv_master_ops = {
 	.unbind	= meson_drv_unbind,
 };
 
+static int __maybe_unused meson_drv_pm_suspend(struct device *dev)
+{
+	struct meson_drm *priv = dev_get_drvdata(dev);
+
+	if (!priv)
+		return 0;
+
+	return drm_mode_config_helper_suspend(priv->drm);
+}
+
+static int __maybe_unused meson_drv_pm_resume(struct device *dev)
+{
+	struct meson_drm *priv = dev_get_drvdata(dev);
+
+	if (!priv)
+		return 0;
+
+	meson_vpu_init(priv);
+	meson_venc_init(priv);
+	meson_vpp_init(priv);
+	meson_viu_init(priv);
+
+	drm_mode_config_helper_resume(priv->drm);
+
+	return 0;
+}
+
 static int compare_of(struct device *dev, void *data)
 {
 	DRM_DEBUG_DRIVER("Comparing of node %pOF with %pOF\n",
@@ -467,11 +494,16 @@ static const struct of_device_id dt_match[] = {
 };
 MODULE_DEVICE_TABLE(of, dt_match);
 
+static const struct dev_pm_ops meson_drv_pm_ops = {
+	SET_SYSTEM_SLEEP_PM_OPS(meson_drv_pm_suspend, meson_drv_pm_resume)
+};
+
 static struct platform_driver meson_drm_platform_driver = {
 	.probe      = meson_drv_probe,
 	.driver     = {
 		.name	= "meson-drm",
 		.of_match_table = dt_match,
+		.pm = &meson_drv_pm_ops,
 	},
 };
 
diff --git a/drivers/gpu/drm/meson/meson_dw_hdmi.c b/drivers/gpu/drm/meson/meson_dw_hdmi.c
index 68bbd987147b..3bb7ffe5fc39 100644
--- a/drivers/gpu/drm/meson/meson_dw_hdmi.c
+++ b/drivers/gpu/drm/meson/meson_dw_hdmi.c
@@ -802,6 +802,47 @@ static bool meson_hdmi_connector_is_available(struct device *dev)
 	return false;
 }
 
+static void meson_dw_hdmi_init(struct meson_dw_hdmi *meson_dw_hdmi)
+{
+	struct meson_drm *priv = meson_dw_hdmi->priv;
+
+	/* Enable clocks */
+	regmap_update_bits(priv->hhi, HHI_HDMI_CLK_CNTL, 0xffff, 0x100);
+
+	/* Bring HDMITX MEM output of power down */
+	regmap_update_bits(priv->hhi, HHI_MEM_PD_REG0, 0xff << 8, 0);
+
+	/* Reset HDMITX APB & TX & PHY */
+	reset_control_reset(meson_dw_hdmi->hdmitx_apb);
+	reset_control_reset(meson_dw_hdmi->hdmitx_ctrl);
+	reset_control_reset(meson_dw_hdmi->hdmitx_phy);
+
+	/* Enable APB3 fail on error */
+	if (!meson_vpu_is_compatible(priv, VPU_COMPATIBLE_G12A)) {
+		writel_bits_relaxed(BIT(15), BIT(15),
+				    meson_dw_hdmi->hdmitx + HDMITX_TOP_CTRL_REG);
+		writel_bits_relaxed(BIT(15), BIT(15),
+				    meson_dw_hdmi->hdmitx + HDMITX_DWC_CTRL_REG);
+	}
+
+	/* Bring out of reset */
+	meson_dw_hdmi->data->top_write(meson_dw_hdmi,
+				       HDMITX_TOP_SW_RESET,  0);
+
+	msleep(20);
+
+	meson_dw_hdmi->data->top_write(meson_dw_hdmi,
+				       HDMITX_TOP_CLK_CNTL, 0xff);
+
+	/* Enable HDMI-TX Interrupt */
+	meson_dw_hdmi->data->top_write(meson_dw_hdmi, HDMITX_TOP_INTR_STAT_CLR,
+				       HDMITX_TOP_INTR_CORE);
+
+	meson_dw_hdmi->data->top_write(meson_dw_hdmi, HDMITX_TOP_INTR_MASKN,
+				       HDMITX_TOP_INTR_CORE);
+
+}
+
 static int meson_dw_hdmi_bind(struct device *dev, struct device *master,
 				void *data)
 {
@@ -925,40 +966,7 @@ static int meson_dw_hdmi_bind(struct device *dev, struct device *master,
 
 	DRM_DEBUG_DRIVER("encoder initialized\n");
 
-	/* Enable clocks */
-	regmap_update_bits(priv->hhi, HHI_HDMI_CLK_CNTL, 0xffff, 0x100);
-
-	/* Bring HDMITX MEM output of power down */
-	regmap_update_bits(priv->hhi, HHI_MEM_PD_REG0, 0xff << 8, 0);
-
-	/* Reset HDMITX APB & TX & PHY */
-	reset_control_reset(meson_dw_hdmi->hdmitx_apb);
-	reset_control_reset(meson_dw_hdmi->hdmitx_ctrl);
-	reset_control_reset(meson_dw_hdmi->hdmitx_phy);
-
-	/* Enable APB3 fail on error */
-	if (!meson_vpu_is_compatible(priv, VPU_COMPATIBLE_G12A)) {
-		writel_bits_relaxed(BIT(15), BIT(15),
-				    meson_dw_hdmi->hdmitx + HDMITX_TOP_CTRL_REG);
-		writel_bits_relaxed(BIT(15), BIT(15),
-				    meson_dw_hdmi->hdmitx + HDMITX_DWC_CTRL_REG);
-	}
-
-	/* Bring out of reset */
-	meson_dw_hdmi->data->top_write(meson_dw_hdmi,
-				       HDMITX_TOP_SW_RESET,  0);
-
-	msleep(20);
-
-	meson_dw_hdmi->data->top_write(meson_dw_hdmi,
-				       HDMITX_TOP_CLK_CNTL, 0xff);
-
-	/* Enable HDMI-TX Interrupt */
-	meson_dw_hdmi->data->top_write(meson_dw_hdmi, HDMITX_TOP_INTR_STAT_CLR,
-				       HDMITX_TOP_INTR_CORE);
-
-	meson_dw_hdmi->data->top_write(meson_dw_hdmi, HDMITX_TOP_INTR_MASKN,
-				       HDMITX_TOP_INTR_CORE);
+	meson_dw_hdmi_init(meson_dw_hdmi);
 
 	/* Bridge / Connector */
 
@@ -969,6 +977,11 @@ static int meson_dw_hdmi_bind(struct device *dev, struct device *master,
 	dw_plat_data->input_bus_format = MEDIA_BUS_FMT_YUV8_1X24;
 	dw_plat_data->input_bus_encoding = V4L2_YCBCR_ENC_709;
 
+	if (dw_hdmi_is_compatible(meson_dw_hdmi, "amlogic,meson-gxl-dw-hdmi") ||
+	    dw_hdmi_is_compatible(meson_dw_hdmi, "amlogic,meson-gxm-dw-hdmi") ||
+	    dw_hdmi_is_compatible(meson_dw_hdmi, "amlogic,meson-g12a-dw-hdmi"))
+		dw_plat_data->use_drm_infoframe = true;
+
 	platform_set_drvdata(pdev, meson_dw_hdmi);
 
 	meson_dw_hdmi->hdmi = dw_hdmi_bind(pdev, encoder,
@@ -994,6 +1007,34 @@ static const struct component_ops meson_dw_hdmi_ops = {
 	.unbind	= meson_dw_hdmi_unbind,
 };
 
+static int __maybe_unused meson_dw_hdmi_pm_suspend(struct device *dev)
+{
+	struct meson_dw_hdmi *meson_dw_hdmi = dev_get_drvdata(dev);
+
+	if (!meson_dw_hdmi)
+		return 0;
+
+	/* Reset TOP */
+	meson_dw_hdmi->data->top_write(meson_dw_hdmi,
+				       HDMITX_TOP_SW_RESET, 0);
+
+	return 0;
+}
+
+static int __maybe_unused meson_dw_hdmi_pm_resume(struct device *dev)
+{
+	struct meson_dw_hdmi *meson_dw_hdmi = dev_get_drvdata(dev);
+
+	if (!meson_dw_hdmi)
+		return 0;
+
+	meson_dw_hdmi_init(meson_dw_hdmi);
+
+	dw_hdmi_resume(meson_dw_hdmi->hdmi);
+
+	return 0;
+}
+
 static int meson_dw_hdmi_probe(struct platform_device *pdev)
 {
 	return component_add(&pdev->dev, &meson_dw_hdmi_ops);
@@ -1006,6 +1047,11 @@ static int meson_dw_hdmi_remove(struct platform_device *pdev)
 	return 0;
 }
 
+static const struct dev_pm_ops meson_dw_hdmi_pm_ops = {
+	SET_SYSTEM_SLEEP_PM_OPS(meson_dw_hdmi_pm_suspend,
+				meson_dw_hdmi_pm_resume)
+};
+
 static const struct of_device_id meson_dw_hdmi_of_table[] = {
 	{ .compatible = "amlogic,meson-gxbb-dw-hdmi",
 	  .data = &meson_dw_hdmi_gx_data },
@@ -1025,6 +1071,7 @@ static struct platform_driver meson_dw_hdmi_platform_driver = {
 	.driver		= {
 		.name		= DRIVER_NAME,
 		.of_match_table	= meson_dw_hdmi_of_table,
+		.pm = &meson_dw_hdmi_pm_ops,
 	},
 };
 module_platform_driver(meson_dw_hdmi_platform_driver);
diff --git a/drivers/gpu/drm/meson/meson_vclk.c b/drivers/gpu/drm/meson/meson_vclk.c
index ac491a781952..f690793ae2d5 100644
--- a/drivers/gpu/drm/meson/meson_vclk.c
+++ b/drivers/gpu/drm/meson/meson_vclk.c
@@ -638,13 +638,18 @@ static bool meson_hdmi_pll_validate_params(struct meson_drm *priv,
 		if (frac >= HDMI_FRAC_MAX_GXBB)
 			return false;
 	} else if (meson_vpu_is_compatible(priv, VPU_COMPATIBLE_GXM) ||
-		   meson_vpu_is_compatible(priv, VPU_COMPATIBLE_GXL) ||
-		   meson_vpu_is_compatible(priv, VPU_COMPATIBLE_G12A)) {
+		   meson_vpu_is_compatible(priv, VPU_COMPATIBLE_GXL)) {
 		/* Empiric supported min/max dividers */
 		if (m < 106 || m > 247)
 			return false;
 		if (frac >= HDMI_FRAC_MAX_GXL)
 			return false;
+	} else if (meson_vpu_is_compatible(priv, VPU_COMPATIBLE_G12A)) {
+		/* Empiric supported min/max dividers */
+		if (m < 106 || m > 247)
+			return false;
+		if (frac >= HDMI_FRAC_MAX_G12A)
+			return false;
 	}
 
 	return true;
diff --git a/drivers/gpu/drm/mgag200/Kconfig b/drivers/gpu/drm/mgag200/Kconfig
index 76fee0fbdcae..aed11f4f4c55 100644
--- a/drivers/gpu/drm/mgag200/Kconfig
+++ b/drivers/gpu/drm/mgag200/Kconfig
@@ -4,6 +4,8 @@ config DRM_MGAG200
 	depends on DRM && PCI && MMU
 	select DRM_KMS_HELPER
 	select DRM_VRAM_HELPER
+	select DRM_TTM
+	select DRM_TTM_HELPER
 	help
 	 This is a KMS driver for the MGA G200 server chips, it
          does not support the original MGA G200 or any of the desktop
diff --git a/drivers/gpu/drm/mgag200/mgag200_cursor.c b/drivers/gpu/drm/mgag200/mgag200_cursor.c
index 289ce3e29032..79711dbb5b03 100644
--- a/drivers/gpu/drm/mgag200/mgag200_cursor.c
+++ b/drivers/gpu/drm/mgag200/mgag200_cursor.c
@@ -12,35 +12,10 @@
 static bool warn_transparent = true;
 static bool warn_palette = true;
 
-/*
-  Hide the cursor off screen. We can't disable the cursor hardware because it
-  takes too long to re-activate and causes momentary corruption
-*/
-static void mga_hide_cursor(struct mga_device *mdev)
-{
-	WREG8(MGA_CURPOSXL, 0);
-	WREG8(MGA_CURPOSXH, 0);
-	if (mdev->cursor.pixels_current)
-		drm_gem_vram_unpin(mdev->cursor.pixels_current);
-	mdev->cursor.pixels_current = NULL;
-}
-
-int mga_crtc_cursor_set(struct drm_crtc *crtc,
-			struct drm_file *file_priv,
-			uint32_t handle,
-			uint32_t width,
-			uint32_t height)
+static int mgag200_cursor_update(struct mga_device *mdev, void *dst, void *src,
+				 unsigned int width, unsigned int height)
 {
-	struct drm_device *dev = crtc->dev;
-	struct mga_device *mdev = (struct mga_device *)dev->dev_private;
-	struct drm_gem_vram_object *pixels_1 = mdev->cursor.pixels_1;
-	struct drm_gem_vram_object *pixels_2 = mdev->cursor.pixels_2;
-	struct drm_gem_vram_object *pixels_current = mdev->cursor.pixels_current;
-	struct drm_gem_vram_object *pixels_next;
-	struct drm_gem_object *obj;
-	struct drm_gem_vram_object *gbo = NULL;
-	int ret = 0;
-	u8 *src, *dst;
+	struct drm_device *dev = mdev->dev;
 	unsigned int i, row, col;
 	uint32_t colour_set[16];
 	uint32_t *next_space = &colour_set[0];
@@ -48,79 +23,9 @@ int mga_crtc_cursor_set(struct drm_crtc *crtc,
 	uint32_t this_colour;
 	bool found = false;
 	int colour_count = 0;
-	s64 gpu_addr;
-	u64 dst_gpu;
 	u8 reg_index;
 	u8 this_row[48];
 
-	if (!pixels_1 || !pixels_2) {
-		WREG8(MGA_CURPOSXL, 0);
-		WREG8(MGA_CURPOSXH, 0);
-		return -ENOTSUPP; /* Didn't allocate space for cursors */
-	}
-
-	if (WARN_ON(pixels_current &&
-		    pixels_1 != pixels_current &&
-		    pixels_2 != pixels_current)) {
-		return -ENOTSUPP; /* inconsistent state */
-	}
-
-	if (!handle || !file_priv) {
-		mga_hide_cursor(mdev);
-		return 0;
-	}
-
-	if (width != 64 || height != 64) {
-		WREG8(MGA_CURPOSXL, 0);
-		WREG8(MGA_CURPOSXH, 0);
-		return -EINVAL;
-	}
-
-	if (pixels_current == pixels_1)
-		pixels_next = pixels_2;
-	else
-		pixels_next = pixels_1;
-
-	obj = drm_gem_object_lookup(file_priv, handle);
-	if (!obj)
-		return -ENOENT;
-	gbo = drm_gem_vram_of_gem(obj);
-	ret = drm_gem_vram_pin(gbo, 0);
-	if (ret) {
-		dev_err(&dev->pdev->dev, "failed to lock user bo\n");
-		goto err_drm_gem_object_put_unlocked;
-	}
-	src = drm_gem_vram_kmap(gbo, true, NULL);
-	if (IS_ERR(src)) {
-		ret = PTR_ERR(src);
-		dev_err(&dev->pdev->dev,
-			"failed to kmap user buffer updates\n");
-		goto err_drm_gem_vram_unpin_src;
-	}
-
-	/* Pin and map up-coming buffer to write colour indices */
-	ret = drm_gem_vram_pin(pixels_next, DRM_GEM_VRAM_PL_FLAG_VRAM);
-	if (ret) {
-		dev_err(&dev->pdev->dev,
-			"failed to pin cursor buffer: %d\n", ret);
-		goto err_drm_gem_vram_kunmap_src;
-	}
-	dst = drm_gem_vram_kmap(pixels_next, true, NULL);
-	if (IS_ERR(dst)) {
-		ret = PTR_ERR(dst);
-		dev_err(&dev->pdev->dev,
-			"failed to kmap cursor updates: %d\n", ret);
-		goto err_drm_gem_vram_unpin_dst;
-	}
-	gpu_addr = drm_gem_vram_offset(pixels_next);
-	if (gpu_addr < 0) {
-		ret = (int)gpu_addr;
-		dev_err(&dev->pdev->dev,
-			"failed to get cursor scanout address: %d\n", ret);
-		goto err_drm_gem_vram_kunmap_dst;
-	}
-	dst_gpu = (u64)gpu_addr;
-
 	memset(&colour_set[0], 0, sizeof(uint32_t)*16);
 	/* width*height*4 = 16384 */
 	for (i = 0; i < 16384; i += 4) {
@@ -133,8 +38,7 @@ int mga_crtc_cursor_set(struct drm_crtc *crtc,
 				dev_info(&dev->pdev->dev, "Not enabling hardware cursor.\n");
 				warn_transparent = false; /* Only tell the user once. */
 			}
-			ret = -EINVAL;
-			goto err_drm_gem_vram_kunmap_dst;
+			return -EINVAL;
 		}
 		/* Don't need to store transparent pixels as colours */
 		if (this_colour>>24 == 0x0)
@@ -155,8 +59,7 @@ int mga_crtc_cursor_set(struct drm_crtc *crtc,
 				dev_info(&dev->pdev->dev, "Not enabling hardware cursor.\n");
 				warn_palette = false; /* Only tell the user once. */
 			}
-			ret = -EINVAL;
-			goto err_drm_gem_vram_kunmap_dst;
+			return -EINVAL;
 		}
 		*next_space = this_colour;
 		next_space++;
@@ -200,54 +103,218 @@ int mga_crtc_cursor_set(struct drm_crtc *crtc,
 		memcpy_toio(dst + row*48, &this_row[0], 48);
 	}
 
+	return 0;
+}
+
+static void mgag200_cursor_set_base(struct mga_device *mdev, u64 address)
+{
+	u8 addrl = (address >> 10) & 0xff;
+	u8 addrh = (address >> 18) & 0x3f;
+
 	/* Program gpu address of cursor buffer */
-	WREG_DAC(MGA1064_CURSOR_BASE_ADR_LOW, (u8)((dst_gpu>>10) & 0xff));
-	WREG_DAC(MGA1064_CURSOR_BASE_ADR_HI, (u8)((dst_gpu>>18) & 0x3f));
+	WREG_DAC(MGA1064_CURSOR_BASE_ADR_LOW, addrl);
+	WREG_DAC(MGA1064_CURSOR_BASE_ADR_HI, addrh);
+}
+
+static int mgag200_show_cursor(struct mga_device *mdev, void *src,
+			       unsigned int width, unsigned int height)
+{
+	struct drm_device *dev = mdev->dev;
+	struct drm_gem_vram_object *gbo;
+	void *dst;
+	s64 off;
+	int ret;
+
+	gbo = mdev->cursor.gbo[mdev->cursor.next_index];
+	if (!gbo) {
+		WREG8(MGA_CURPOSXL, 0);
+		WREG8(MGA_CURPOSXH, 0);
+		return -ENOTSUPP; /* Didn't allocate space for cursors */
+	}
+	dst = drm_gem_vram_vmap(gbo);
+	if (IS_ERR(dst)) {
+		ret = PTR_ERR(dst);
+		dev_err(&dev->pdev->dev,
+			"failed to map cursor updates: %d\n", ret);
+		return ret;
+	}
+	off = drm_gem_vram_offset(gbo);
+	if (off < 0) {
+		ret = (int)off;
+		dev_err(&dev->pdev->dev,
+			"failed to get cursor scanout address: %d\n", ret);
+		goto err_drm_gem_vram_vunmap;
+	}
+
+	ret = mgag200_cursor_update(mdev, dst, src, width, height);
+	if (ret)
+		goto err_drm_gem_vram_vunmap;
+	mgag200_cursor_set_base(mdev, off);
 
 	/* Adjust cursor control register to turn on the cursor */
 	WREG_DAC(MGA1064_CURSOR_CTL, 4); /* 16-colour palletized cursor mode */
 
-	/* Now update internal buffer pointers */
-	if (pixels_current)
-		drm_gem_vram_unpin(pixels_current);
-	mdev->cursor.pixels_current = pixels_next;
+	drm_gem_vram_vunmap(gbo, dst);
 
-	drm_gem_vram_kunmap(pixels_next);
-	drm_gem_vram_kunmap(gbo);
-	drm_gem_vram_unpin(gbo);
-	drm_gem_object_put_unlocked(obj);
+	++mdev->cursor.next_index;
+	mdev->cursor.next_index %= ARRAY_SIZE(mdev->cursor.gbo);
+
+	return 0;
+
+err_drm_gem_vram_vunmap:
+	drm_gem_vram_vunmap(gbo, dst);
+	return ret;
+}
+
+/*
+ * Hide the cursor off screen. We can't disable the cursor hardware because
+ * it takes too long to re-activate and causes momentary corruption.
+ */
+static void mgag200_hide_cursor(struct mga_device *mdev)
+{
+	WREG8(MGA_CURPOSXL, 0);
+	WREG8(MGA_CURPOSXH, 0);
+}
+
+static void mgag200_move_cursor(struct mga_device *mdev, int x, int y)
+{
+	if (WARN_ON(x <= 0))
+		return;
+	if (WARN_ON(y <= 0))
+		return;
+	if (WARN_ON(x & ~0xffff))
+		return;
+	if (WARN_ON(y & ~0xffff))
+		return;
+
+	WREG8(MGA_CURPOSXL, x & 0xff);
+	WREG8(MGA_CURPOSXH, (x>>8) & 0xff);
+
+	WREG8(MGA_CURPOSYL, y & 0xff);
+	WREG8(MGA_CURPOSYH, (y>>8) & 0xff);
+}
+
+int mgag200_cursor_init(struct mga_device *mdev)
+{
+	struct drm_device *dev = mdev->dev;
+	size_t ncursors = ARRAY_SIZE(mdev->cursor.gbo);
+	size_t size;
+	int ret;
+	size_t i;
+	struct drm_gem_vram_object *gbo;
+
+	size = roundup(64 * 48, PAGE_SIZE);
+	if (size * ncursors > mdev->vram_fb_available)
+		return -ENOMEM;
+
+	for (i = 0; i < ncursors; ++i) {
+		gbo = drm_gem_vram_create(dev, &dev->vram_mm->bdev,
+					  size, 0, false);
+		if (IS_ERR(gbo)) {
+			ret = PTR_ERR(gbo);
+			goto err_drm_gem_vram_put;
+		}
+		ret = drm_gem_vram_pin(gbo, DRM_GEM_VRAM_PL_FLAG_VRAM |
+					    DRM_GEM_VRAM_PL_FLAG_TOPDOWN);
+		if (ret) {
+			drm_gem_vram_put(gbo);
+			goto err_drm_gem_vram_put;
+		}
+
+		mdev->cursor.gbo[i] = gbo;
+	}
+
+	/*
+	 * At the high end of video memory, we reserve space for
+	 * buffer objects. The cursor plane uses this memory to store
+	 * a double-buffered image of the current cursor. Hence, it's
+	 * not available for framebuffers.
+	 */
+	mdev->vram_fb_available -= ncursors * size;
 
 	return 0;
 
-err_drm_gem_vram_kunmap_dst:
-	drm_gem_vram_kunmap(pixels_next);
-err_drm_gem_vram_unpin_dst:
-	drm_gem_vram_unpin(pixels_next);
-err_drm_gem_vram_kunmap_src:
-	drm_gem_vram_kunmap(gbo);
-err_drm_gem_vram_unpin_src:
-	drm_gem_vram_unpin(gbo);
+err_drm_gem_vram_put:
+	while (i) {
+		--i;
+		gbo = mdev->cursor.gbo[i];
+		drm_gem_vram_unpin(gbo);
+		drm_gem_vram_put(gbo);
+		mdev->cursor.gbo[i] = NULL;
+	}
+	return ret;
+}
+
+void mgag200_cursor_fini(struct mga_device *mdev)
+{
+	size_t i;
+	struct drm_gem_vram_object *gbo;
+
+	for (i = 0; i < ARRAY_SIZE(mdev->cursor.gbo); ++i) {
+		gbo = mdev->cursor.gbo[i];
+		drm_gem_vram_unpin(gbo);
+		drm_gem_vram_put(gbo);
+	}
+}
+
+int mgag200_crtc_cursor_set(struct drm_crtc *crtc, struct drm_file *file_priv,
+			    uint32_t handle, uint32_t width, uint32_t height)
+{
+	struct drm_device *dev = crtc->dev;
+	struct mga_device *mdev = (struct mga_device *)dev->dev_private;
+	struct drm_gem_object *obj;
+	struct drm_gem_vram_object *gbo = NULL;
+	int ret;
+	u8 *src;
+
+	if (!handle || !file_priv) {
+		mgag200_hide_cursor(mdev);
+		return 0;
+	}
+
+	if (width != 64 || height != 64) {
+		WREG8(MGA_CURPOSXL, 0);
+		WREG8(MGA_CURPOSXH, 0);
+		return -EINVAL;
+	}
+
+	obj = drm_gem_object_lookup(file_priv, handle);
+	if (!obj)
+		return -ENOENT;
+	gbo = drm_gem_vram_of_gem(obj);
+	src = drm_gem_vram_vmap(gbo);
+	if (IS_ERR(src)) {
+		ret = PTR_ERR(src);
+		dev_err(&dev->pdev->dev,
+			"failed to map user buffer updates\n");
+		goto err_drm_gem_object_put_unlocked;
+	}
+
+	ret = mgag200_show_cursor(mdev, src, width, height);
+	if (ret)
+		goto err_drm_gem_vram_vunmap;
+
+	/* Now update internal buffer pointers */
+	drm_gem_vram_vunmap(gbo, src);
+	drm_gem_object_put_unlocked(obj);
+
+	return 0;
+err_drm_gem_vram_vunmap:
+	drm_gem_vram_vunmap(gbo, src);
 err_drm_gem_object_put_unlocked:
 	drm_gem_object_put_unlocked(obj);
 	return ret;
 }
 
-int mga_crtc_cursor_move(struct drm_crtc *crtc, int x, int y)
+int mgag200_crtc_cursor_move(struct drm_crtc *crtc, int x, int y)
 {
 	struct mga_device *mdev = (struct mga_device *)crtc->dev->dev_private;
+
 	/* Our origin is at (64,64) */
 	x += 64;
 	y += 64;
 
-	BUG_ON(x <= 0);
-	BUG_ON(y <= 0);
-	BUG_ON(x & ~0xffff);
-	BUG_ON(y & ~0xffff);
+	mgag200_move_cursor(mdev, x, y);
 
-	WREG8(MGA_CURPOSXL, x & 0xff);
-	WREG8(MGA_CURPOSXH, (x>>8) & 0xff);
-
-	WREG8(MGA_CURPOSYL, y & 0xff);
-	WREG8(MGA_CURPOSYH, (y>>8) & 0xff);
 	return 0;
 }
diff --git a/drivers/gpu/drm/mgag200/mgag200_drv.c b/drivers/gpu/drm/mgag200/mgag200_drv.c
index afd9119b6cf1..397f8b0a9af8 100644
--- a/drivers/gpu/drm/mgag200/mgag200_drv.c
+++ b/drivers/gpu/drm/mgag200/mgag200_drv.c
@@ -46,7 +46,7 @@ MODULE_DEVICE_TABLE(pci, pciidlist);
 
 static int mga_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 {
-	drm_fb_helper_remove_conflicting_pci_framebuffers(pdev, 0, "mgag200drmfb");
+	drm_fb_helper_remove_conflicting_pci_framebuffers(pdev, "mgag200drmfb");
 
 	return drm_get_pci_dev(pdev, ent, &driver);
 }
@@ -58,10 +58,7 @@ static void mga_pci_remove(struct pci_dev *pdev)
 	drm_put_dev(dev);
 }
 
-static const struct file_operations mgag200_driver_fops = {
-	.owner = THIS_MODULE,
-	DRM_VRAM_MM_FILE_OPERATIONS
-};
+DEFINE_DRM_GEM_FOPS(mgag200_driver_fops);
 
 static struct drm_driver driver = {
 	.driver_features = DRIVER_GEM | DRIVER_MODESET,
diff --git a/drivers/gpu/drm/mgag200/mgag200_drv.h b/drivers/gpu/drm/mgag200/mgag200_drv.h
index 1c93f8dc08c7..0ea9a525e57d 100644
--- a/drivers/gpu/drm/mgag200/mgag200_drv.h
+++ b/drivers/gpu/drm/mgag200/mgag200_drv.h
@@ -19,7 +19,6 @@
 #include <drm/drm_fb_helper.h>
 #include <drm/drm_gem.h>
 #include <drm/drm_gem_vram_helper.h>
-#include <drm/drm_vram_mm_helper.h>
 
 #include "mgag200_reg.h"
 
@@ -130,16 +129,8 @@ struct mga_connector {
 };
 
 struct mga_cursor {
-	/*
-	   We have to have 2 buffers for the cursor to avoid occasional
-	   corruption while switching cursor icons.
-	   If either of these is NULL, then don't do hardware cursors, and
-	   fall back to software.
-	*/
-	struct drm_gem_vram_object *pixels_1;
-	struct drm_gem_vram_object *pixels_2;
-	/* The currently displayed icon, this points to one of pixels_1, or pixels_2 */
-	struct drm_gem_vram_object *pixels_current;
+	struct drm_gem_vram_object *gbo[2];
+	unsigned int next_index;
 };
 
 struct mga_mc {
@@ -174,6 +165,8 @@ struct mga_device {
 
 	struct mga_cursor cursor;
 
+	size_t vram_fb_available;
+
 	bool				suspended;
 	int				num_crtc;
 	enum mga_type			type;
@@ -204,8 +197,10 @@ int mgag200_mm_init(struct mga_device *mdev);
 void mgag200_mm_fini(struct mga_device *mdev);
 int mgag200_mmap(struct file *filp, struct vm_area_struct *vma);
 
-int mga_crtc_cursor_set(struct drm_crtc *crtc, struct drm_file *file_priv,
-						uint32_t handle, uint32_t width, uint32_t height);
-int mga_crtc_cursor_move(struct drm_crtc *crtc, int x, int y);
+int mgag200_cursor_init(struct mga_device *mdev);
+void mgag200_cursor_fini(struct mga_device *mdev);
+int mgag200_crtc_cursor_set(struct drm_crtc *crtc, struct drm_file *file_priv,
+			    uint32_t handle, uint32_t width, uint32_t height);
+int mgag200_crtc_cursor_move(struct drm_crtc *crtc, int x, int y);
 
 #endif				/* __MGAG200_DRV_H__ */
diff --git a/drivers/gpu/drm/mgag200/mgag200_main.c b/drivers/gpu/drm/mgag200/mgag200_main.c
index a9773334dedf..5f74aabcd3df 100644
--- a/drivers/gpu/drm/mgag200/mgag200_main.c
+++ b/drivers/gpu/drm/mgag200/mgag200_main.c
@@ -159,7 +159,7 @@ int mgag200_driver_load(struct drm_device *dev, unsigned long flags)
 
 	drm_mode_config_init(dev);
 	dev->mode_config.funcs = (void *)&mga_mode_funcs;
-	if (IS_G200_SE(mdev) && mdev->mc.vram_size < (2048*1024))
+	if (IS_G200_SE(mdev) && mdev->vram_fb_available < (2048*1024))
 		dev->mode_config.preferred_depth = 16;
 	else
 		dev->mode_config.preferred_depth = 32;
@@ -171,20 +171,10 @@ int mgag200_driver_load(struct drm_device *dev, unsigned long flags)
 		goto err_modeset;
 	}
 
-	/* Make small buffers to store a hardware cursor (double buffered icon updates) */
-	mdev->cursor.pixels_1 = drm_gem_vram_create(dev, &dev->vram_mm->bdev,
-						    roundup(48*64, PAGE_SIZE),
-						    0, 0);
-	mdev->cursor.pixels_2 = drm_gem_vram_create(dev, &dev->vram_mm->bdev,
-						    roundup(48*64, PAGE_SIZE),
-						    0, 0);
-	if (IS_ERR(mdev->cursor.pixels_2) || IS_ERR(mdev->cursor.pixels_1)) {
-		mdev->cursor.pixels_1 = NULL;
-		mdev->cursor.pixels_2 = NULL;
+	r = mgag200_cursor_init(mdev);
+	if (r)
 		dev_warn(&dev->pdev->dev,
-			"Could not allocate space for cursors. Not doing hardware cursors.\n");
-	}
-	mdev->cursor.pixels_current = NULL;
+			"Could not initialize cursors. Not doing hardware cursors.\n");
 
 	r = drm_fbdev_generic_setup(mdev->dev, 0);
 	if (r)
@@ -194,6 +184,7 @@ int mgag200_driver_load(struct drm_device *dev, unsigned long flags)
 
 err_modeset:
 	drm_mode_config_cleanup(dev);
+	mgag200_cursor_fini(mdev);
 	mgag200_mm_fini(mdev);
 err_mm:
 	dev->dev_private = NULL;
@@ -209,6 +200,7 @@ void mgag200_driver_unload(struct drm_device *dev)
 		return;
 	mgag200_modeset_fini(mdev);
 	drm_mode_config_cleanup(dev);
+	mgag200_cursor_fini(mdev);
 	mgag200_mm_fini(mdev);
 	dev->dev_private = NULL;
 }
diff --git a/drivers/gpu/drm/mgag200/mgag200_mode.c b/drivers/gpu/drm/mgag200/mgag200_mode.c
index 5e778b5f1a10..5ec697148fc1 100644
--- a/drivers/gpu/drm/mgag200/mgag200_mode.c
+++ b/drivers/gpu/drm/mgag200/mgag200_mode.c
@@ -1413,8 +1413,8 @@ static void mga_crtc_disable(struct drm_crtc *crtc)
 
 /* These provide the minimum set of functions required to handle a CRTC */
 static const struct drm_crtc_funcs mga_crtc_funcs = {
-	.cursor_set = mga_crtc_cursor_set,
-	.cursor_move = mga_crtc_cursor_move,
+	.cursor_set = mgag200_crtc_cursor_set,
+	.cursor_move = mgag200_crtc_cursor_move,
 	.gamma_set = mga_crtc_gamma_set,
 	.set_config = drm_crtc_helper_set_config,
 	.destroy = mga_crtc_destroy,
@@ -1629,7 +1629,7 @@ static enum drm_mode_status mga_vga_mode_valid(struct drm_connector *connector,
 			bpp = connector->cmdline_mode.bpp;
 	}
 
-	if ((mode->hdisplay * mode->vdisplay * (bpp/8)) > mdev->mc.vram_size) {
+	if ((mode->hdisplay * mode->vdisplay * (bpp/8)) > mdev->vram_fb_available) {
 		if (connector->cmdline_mode.specified)
 			connector->cmdline_mode.specified = false;
 		return MODE_BAD;
@@ -1638,16 +1638,6 @@ static enum drm_mode_status mga_vga_mode_valid(struct drm_connector *connector,
 	return MODE_OK;
 }
 
-static struct drm_encoder *mga_connector_best_encoder(struct drm_connector
-						  *connector)
-{
-	int enc_id = connector->encoder_ids[0];
-	/* pick the encoder ids */
-	if (enc_id)
-		return drm_encoder_find(connector->dev, NULL, enc_id);
-	return NULL;
-}
-
 static void mga_connector_destroy(struct drm_connector *connector)
 {
 	struct mga_connector *mga_connector = to_mga_connector(connector);
@@ -1659,7 +1649,6 @@ static void mga_connector_destroy(struct drm_connector *connector)
 static const struct drm_connector_helper_funcs mga_vga_connector_helper_funcs = {
 	.get_modes = mga_vga_get_modes,
 	.mode_valid = mga_vga_mode_valid,
-	.best_encoder = mga_connector_best_encoder,
 };
 
 static const struct drm_connector_funcs mga_vga_connector_funcs = {
diff --git a/drivers/gpu/drm/mgag200/mgag200_ttm.c b/drivers/gpu/drm/mgag200/mgag200_ttm.c
index 73a6b848601c..99997d737362 100644
--- a/drivers/gpu/drm/mgag200/mgag200_ttm.c
+++ b/drivers/gpu/drm/mgag200/mgag200_ttm.c
@@ -37,8 +37,7 @@ int mgag200_mm_init(struct mga_device *mdev)
 	struct drm_device *dev = mdev->dev;
 
 	vmm = drm_vram_helper_alloc_mm(dev, pci_resource_start(dev->pdev, 0),
-				       mdev->mc.vram_size,
-				       &drm_gem_vram_mm_funcs);
+				       mdev->mc.vram_size);
 	if (IS_ERR(vmm)) {
 		ret = PTR_ERR(vmm);
 		DRM_ERROR("Error initializing VRAM MM; %d\n", ret);
@@ -51,6 +50,8 @@ int mgag200_mm_init(struct mga_device *mdev)
 	mdev->fb_mtrr = arch_phys_wc_add(pci_resource_start(dev->pdev, 0),
 					 pci_resource_len(dev->pdev, 0));
 
+	mdev->vram_fb_available = mdev->mc.vram_size;
+
 	return 0;
 }
 
@@ -58,6 +59,8 @@ void mgag200_mm_fini(struct mga_device *mdev)
 {
 	struct drm_device *dev = mdev->dev;
 
+	mdev->vram_fb_available = 0;
+
 	drm_vram_helper_release_mm(dev);
 
 	arch_io_free_memtype_wc(pci_resource_start(dev->pdev, 0),
diff --git a/drivers/gpu/drm/msm/adreno/a6xx_gpu_state.c b/drivers/gpu/drm/msm/adreno/a6xx_gpu_state.c
index e686331fa089..691c1a277d91 100644
--- a/drivers/gpu/drm/msm/adreno/a6xx_gpu_state.c
+++ b/drivers/gpu/drm/msm/adreno/a6xx_gpu_state.c
@@ -352,26 +352,26 @@ static void a6xx_get_debugbus(struct msm_gpu *gpu,
 		cxdbg = ioremap(res->start, resource_size(res));
 
 	if (cxdbg) {
-		cxdbg_write(cxdbg, REG_A6XX_DBGC_CFG_DBGBUS_CNTLT,
+		cxdbg_write(cxdbg, REG_A6XX_CX_DBGC_CFG_DBGBUS_CNTLT,
 			A6XX_DBGC_CFG_DBGBUS_CNTLT_SEGT(0xf));
 
-		cxdbg_write(cxdbg, REG_A6XX_DBGC_CFG_DBGBUS_CNTLM,
+		cxdbg_write(cxdbg, REG_A6XX_CX_DBGC_CFG_DBGBUS_CNTLM,
 			A6XX_DBGC_CFG_DBGBUS_CNTLM_ENABLE(0xf));
 
-		cxdbg_write(cxdbg, REG_A6XX_DBGC_CFG_DBGBUS_IVTL_0, 0);
-		cxdbg_write(cxdbg, REG_A6XX_DBGC_CFG_DBGBUS_IVTL_1, 0);
-		cxdbg_write(cxdbg, REG_A6XX_DBGC_CFG_DBGBUS_IVTL_2, 0);
-		cxdbg_write(cxdbg, REG_A6XX_DBGC_CFG_DBGBUS_IVTL_3, 0);
+		cxdbg_write(cxdbg, REG_A6XX_CX_DBGC_CFG_DBGBUS_IVTL_0, 0);
+		cxdbg_write(cxdbg, REG_A6XX_CX_DBGC_CFG_DBGBUS_IVTL_1, 0);
+		cxdbg_write(cxdbg, REG_A6XX_CX_DBGC_CFG_DBGBUS_IVTL_2, 0);
+		cxdbg_write(cxdbg, REG_A6XX_CX_DBGC_CFG_DBGBUS_IVTL_3, 0);
 
-		cxdbg_write(cxdbg, REG_A6XX_DBGC_CFG_DBGBUS_BYTEL_0,
+		cxdbg_write(cxdbg, REG_A6XX_CX_DBGC_CFG_DBGBUS_BYTEL_0,
 			0x76543210);
-		cxdbg_write(cxdbg, REG_A6XX_DBGC_CFG_DBGBUS_BYTEL_1,
+		cxdbg_write(cxdbg, REG_A6XX_CX_DBGC_CFG_DBGBUS_BYTEL_1,
 			0xFEDCBA98);
 
-		cxdbg_write(cxdbg, REG_A6XX_DBGC_CFG_DBGBUS_MASKL_0, 0);
-		cxdbg_write(cxdbg, REG_A6XX_DBGC_CFG_DBGBUS_MASKL_1, 0);
-		cxdbg_write(cxdbg, REG_A6XX_DBGC_CFG_DBGBUS_MASKL_2, 0);
-		cxdbg_write(cxdbg, REG_A6XX_DBGC_CFG_DBGBUS_MASKL_3, 0);
+		cxdbg_write(cxdbg, REG_A6XX_CX_DBGC_CFG_DBGBUS_MASKL_0, 0);
+		cxdbg_write(cxdbg, REG_A6XX_CX_DBGC_CFG_DBGBUS_MASKL_1, 0);
+		cxdbg_write(cxdbg, REG_A6XX_CX_DBGC_CFG_DBGBUS_MASKL_2, 0);
+		cxdbg_write(cxdbg, REG_A6XX_CX_DBGC_CFG_DBGBUS_MASKL_3, 0);
 	}
 
 	a6xx_state->debugbus = state_kcalloc(a6xx_state,
diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_kms.h b/drivers/gpu/drm/msm/disp/dpu1/dpu_kms.h
index 4c889aabdaf9..959d03e007fa 100644
--- a/drivers/gpu/drm/msm/disp/dpu1/dpu_kms.h
+++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_kms.h
@@ -31,7 +31,7 @@
  */
 #define DPU_DEBUG(fmt, ...)                                                \
 	do {                                                               \
-		if (unlikely(drm_debug & DRM_UT_KMS))                      \
+		if (drm_debug_enabled(DRM_UT_KMS))                         \
 			DRM_DEBUG(fmt, ##__VA_ARGS__); \
 		else                                                       \
 			pr_debug(fmt, ##__VA_ARGS__);                      \
@@ -43,7 +43,7 @@
  */
 #define DPU_DEBUG_DRIVER(fmt, ...)                                         \
 	do {                                                               \
-		if (unlikely(drm_debug & DRM_UT_DRIVER))                   \
+		if (drm_debug_enabled(DRM_UT_DRIVER))                      \
 			DRM_ERROR(fmt, ##__VA_ARGS__); \
 		else                                                       \
 			pr_debug(fmt, ##__VA_ARGS__);                      \
diff --git a/drivers/gpu/drm/msm/dsi/dsi.h b/drivers/gpu/drm/msm/dsi/dsi.h
index 0da8a4e428ad..eff1a4c61258 100644
--- a/drivers/gpu/drm/msm/dsi/dsi.h
+++ b/drivers/gpu/drm/msm/dsi/dsi.h
@@ -9,6 +9,7 @@
 #include <linux/of_platform.h>
 #include <linux/platform_device.h>
 
+#include <drm/drm_bridge.h>
 #include <drm/drm_crtc.h>
 #include <drm/drm_mipi_dsi.h>
 #include <drm/drm_panel.h>
diff --git a/drivers/gpu/drm/msm/dsi/dsi_host.c b/drivers/gpu/drm/msm/dsi/dsi_host.c
index 663ff9f4fac9..1e7b1be25bb0 100644
--- a/drivers/gpu/drm/msm/dsi/dsi_host.c
+++ b/drivers/gpu/drm/msm/dsi/dsi_host.c
@@ -26,6 +26,8 @@
 #include "dsi_cfg.h"
 #include "msm_kms.h"
 
+#define DSI_RESET_TOGGLE_DELAY_MS 20
+
 static int dsi_get_version(const void __iomem *base, u32 *major, u32 *minor)
 {
 	u32 ver;
@@ -986,7 +988,7 @@ static void dsi_sw_reset(struct msm_dsi_host *msm_host)
 	wmb(); /* clocks need to be enabled before reset */
 
 	dsi_write(msm_host, REG_DSI_RESET, 1);
-	wmb(); /* make sure reset happen */
+	msleep(DSI_RESET_TOGGLE_DELAY_MS); /* make sure reset happen */
 	dsi_write(msm_host, REG_DSI_RESET, 0);
 }
 
@@ -1396,7 +1398,7 @@ static void dsi_sw_reset_restore(struct msm_dsi_host *msm_host)
 
 	/* dsi controller can only be reset while clocks are running */
 	dsi_write(msm_host, REG_DSI_RESET, 1);
-	wmb();	/* make sure reset happen */
+	msleep(DSI_RESET_TOGGLE_DELAY_MS); /* make sure reset happen */
 	dsi_write(msm_host, REG_DSI_RESET, 0);
 	wmb();	/* controller out of reset */
 	dsi_write(msm_host, REG_DSI_CTRL, data0);
diff --git a/drivers/gpu/drm/msm/edp/edp.c b/drivers/gpu/drm/msm/edp/edp.c
index 0f312ac5b624..ad4e963ccd9b 100644
--- a/drivers/gpu/drm/msm/edp/edp.c
+++ b/drivers/gpu/drm/msm/edp/edp.c
@@ -178,7 +178,9 @@ int msm_edp_modeset_init(struct msm_edp *edp, struct drm_device *dev,
 		goto fail;
 	}
 
-	encoder->bridge = edp->bridge;
+	ret = drm_bridge_attach(encoder, edp->bridge, NULL);
+	if (ret)
+		goto fail;
 
 	priv->bridges[priv->num_bridges++]       = edp->bridge;
 	priv->connectors[priv->num_connectors++] = edp->connector;
diff --git a/drivers/gpu/drm/msm/edp/edp.h b/drivers/gpu/drm/msm/edp/edp.h
index f2c17858a703..eb34243dad53 100644
--- a/drivers/gpu/drm/msm/edp/edp.h
+++ b/drivers/gpu/drm/msm/edp/edp.h
@@ -10,6 +10,7 @@
 #include <linux/interrupt.h>
 #include <linux/kernel.h>
 #include <linux/platform_device.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_crtc.h>
 #include <drm/drm_dp_helper.h>
 
diff --git a/drivers/gpu/drm/msm/edp/edp_ctrl.c b/drivers/gpu/drm/msm/edp/edp_ctrl.c
index 7f3dd3ffe2c9..0d9657cc70db 100644
--- a/drivers/gpu/drm/msm/edp/edp_ctrl.c
+++ b/drivers/gpu/drm/msm/edp/edp_ctrl.c
@@ -89,7 +89,6 @@ struct edp_ctrl {
 	/* edid raw data */
 	struct edid *edid;
 
-	struct drm_dp_link dp_link;
 	struct drm_dp_aux *drm_aux;
 
 	/* dpcd raw data */
@@ -403,7 +402,7 @@ static void edp_fill_link_cfg(struct edp_ctrl *ctrl)
 	u32 prate;
 	u32 lrate;
 	u32 bpp;
-	u8 max_lane = ctrl->dp_link.num_lanes;
+	u8 max_lane = drm_dp_max_lane_count(ctrl->dpcd);
 	u8 lane;
 
 	prate = ctrl->pixel_rate;
@@ -413,7 +412,7 @@ static void edp_fill_link_cfg(struct edp_ctrl *ctrl)
 	 * By default, use the maximum link rate and minimum lane count,
 	 * so that we can do rate down shift during link training.
 	 */
-	ctrl->link_rate = drm_dp_link_rate_to_bw_code(ctrl->dp_link.rate);
+	ctrl->link_rate = ctrl->dpcd[DP_MAX_LINK_RATE];
 
 	prate *= bpp;
 	prate /= 8; /* in kByte */
@@ -439,7 +438,7 @@ static void edp_config_ctrl(struct edp_ctrl *ctrl)
 
 	data = EDP_CONFIGURATION_CTRL_LANES(ctrl->lane_cnt - 1);
 
-	if (ctrl->dp_link.capabilities & DP_LINK_CAP_ENHANCED_FRAMING)
+	if (drm_dp_enhanced_frame_cap(ctrl->dpcd))
 		data |= EDP_CONFIGURATION_CTRL_ENHANCED_FRAMING;
 
 	depth = EDP_6BIT;
@@ -701,7 +700,7 @@ static int edp_link_rate_down_shift(struct edp_ctrl *ctrl)
 
 	rate = ctrl->link_rate;
 	lane = ctrl->lane_cnt;
-	max_lane = ctrl->dp_link.num_lanes;
+	max_lane = drm_dp_max_lane_count(ctrl->dpcd);
 
 	bpp = ctrl->color_depth * 3;
 	prate = ctrl->pixel_rate;
@@ -751,18 +750,22 @@ static int edp_clear_training_pattern(struct edp_ctrl *ctrl)
 
 static int edp_do_link_train(struct edp_ctrl *ctrl)
 {
+	u8 values[2];
 	int ret;
-	struct drm_dp_link dp_link;
 
 	DBG("");
 	/*
 	 * Set the current link rate and lane cnt to panel. They may have been
 	 * adjusted and the values are different from them in DPCD CAP
 	 */
-	dp_link.num_lanes = ctrl->lane_cnt;
-	dp_link.rate = drm_dp_bw_code_to_link_rate(ctrl->link_rate);
-	dp_link.capabilities = ctrl->dp_link.capabilities;
-	if (drm_dp_link_configure(ctrl->drm_aux, &dp_link) < 0)
+	values[0] = ctrl->lane_cnt;
+	values[1] = ctrl->link_rate;
+
+	if (drm_dp_enhanced_frame_cap(ctrl->dpcd))
+		values[1] |= DP_LANE_COUNT_ENHANCED_FRAME_EN;
+
+	if (drm_dp_dpcd_write(ctrl->drm_aux, DP_LINK_BW_SET, values,
+			      sizeof(values)) < 0)
 		return EDP_TRAIN_FAIL;
 
 	ctrl->v_level = 0; /* start from default level */
@@ -952,6 +955,7 @@ static void edp_ctrl_on_worker(struct work_struct *work)
 {
 	struct edp_ctrl *ctrl = container_of(
 				work, struct edp_ctrl, on_work);
+	u8 value;
 	int ret;
 
 	mutex_lock(&ctrl->dev_mutex);
@@ -965,9 +969,27 @@ static void edp_ctrl_on_worker(struct work_struct *work)
 	edp_ctrl_link_enable(ctrl, 1);
 
 	edp_ctrl_irq_enable(ctrl, 1);
-	ret = drm_dp_link_power_up(ctrl->drm_aux, &ctrl->dp_link);
-	if (ret)
-		goto fail;
+
+	/* DP_SET_POWER register is only available on DPCD v1.1 and later */
+	if (ctrl->dpcd[DP_DPCD_REV] >= 0x11) {
+		ret = drm_dp_dpcd_readb(ctrl->drm_aux, DP_SET_POWER, &value);
+		if (ret < 0)
+			goto fail;
+
+		value &= ~DP_SET_POWER_MASK;
+		value |= DP_SET_POWER_D0;
+
+		ret = drm_dp_dpcd_writeb(ctrl->drm_aux, DP_SET_POWER, value);
+		if (ret < 0)
+			goto fail;
+
+		/*
+		 * According to the DP 1.1 specification, a "Sink Device must
+		 * exit the power saving state within 1 ms" (Section 2.5.3.1,
+		 * Table 5-52, "Sink Control Field" (register 0x600).
+		 */
+		usleep_range(1000, 2000);
+	}
 
 	ctrl->power_on = true;
 
@@ -1011,7 +1033,19 @@ static void edp_ctrl_off_worker(struct work_struct *work)
 
 	edp_state_ctrl(ctrl, 0);
 
-	drm_dp_link_power_down(ctrl->drm_aux, &ctrl->dp_link);
+	/* DP_SET_POWER register is only available on DPCD v1.1 and later */
+	if (ctrl->dpcd[DP_DPCD_REV] >= 0x11) {
+		u8 value;
+		int ret;
+
+		ret = drm_dp_dpcd_readb(ctrl->drm_aux, DP_SET_POWER, &value);
+		if (ret > 0) {
+			value &= ~DP_SET_POWER_MASK;
+			value |= DP_SET_POWER_D3;
+
+			drm_dp_dpcd_writeb(ctrl->drm_aux, DP_SET_POWER, value);
+		}
+	}
 
 	edp_ctrl_irq_enable(ctrl, 0);
 
@@ -1225,14 +1259,8 @@ int msm_edp_ctrl_get_panel_info(struct edp_ctrl *ctrl,
 		edp_ctrl_irq_enable(ctrl, 1);
 	}
 
-	ret = drm_dp_link_probe(ctrl->drm_aux, &ctrl->dp_link);
-	if (ret) {
-		pr_err("%s: read dpcd cap failed, %d\n", __func__, ret);
-		goto disable_ret;
-	}
-
 	/* Initialize link rate as panel max link rate */
-	ctrl->link_rate = drm_dp_link_rate_to_bw_code(ctrl->dp_link.rate);
+	ctrl->link_rate = ctrl->dpcd[DP_MAX_LINK_RATE];
 
 	ctrl->edid = drm_get_edid(connector, &ctrl->drm_aux->ddc);
 	if (!ctrl->edid) {
diff --git a/drivers/gpu/drm/msm/hdmi/hdmi.c b/drivers/gpu/drm/msm/hdmi/hdmi.c
index 355afb936401..1a9b6289637d 100644
--- a/drivers/gpu/drm/msm/hdmi/hdmi.c
+++ b/drivers/gpu/drm/msm/hdmi/hdmi.c
@@ -327,7 +327,9 @@ int msm_hdmi_modeset_init(struct hdmi *hdmi,
 		goto fail;
 	}
 
-	encoder->bridge = hdmi->bridge;
+	ret = drm_bridge_attach(encoder, hdmi->bridge, NULL);
+	if (ret)
+		goto fail;
 
 	priv->bridges[priv->num_bridges++]       = hdmi->bridge;
 	priv->connectors[priv->num_connectors++] = hdmi->connector;
diff --git a/drivers/gpu/drm/msm/hdmi/hdmi.h b/drivers/gpu/drm/msm/hdmi/hdmi.h
index bdac452b00fb..d0b84f0abee1 100644
--- a/drivers/gpu/drm/msm/hdmi/hdmi.h
+++ b/drivers/gpu/drm/msm/hdmi/hdmi.h
@@ -14,6 +14,8 @@
 #include <linux/gpio/consumer.h>
 #include <linux/hdmi.h>
 
+#include <drm/drm_bridge.h>
+
 #include "msm_drv.h"
 #include "hdmi.xml.h"
 
diff --git a/drivers/gpu/drm/msm/msm_debugfs.c b/drivers/gpu/drm/msm/msm_debugfs.c
index 6be879578140..1c74381a4fc9 100644
--- a/drivers/gpu/drm/msm/msm_debugfs.c
+++ b/drivers/gpu/drm/msm/msm_debugfs.c
@@ -47,12 +47,8 @@ static int msm_gpu_release(struct inode *inode, struct file *file)
 	struct msm_gpu_show_priv *show_priv = m->private;
 	struct msm_drm_private *priv = show_priv->dev->dev_private;
 	struct msm_gpu *gpu = priv->gpu;
-	int ret;
-
-	ret = mutex_lock_interruptible(&show_priv->dev->struct_mutex);
-	if (ret)
-		return ret;
 
+	mutex_lock(&show_priv->dev->struct_mutex);
 	gpu->funcs->gpu_state_put(show_priv->state);
 	mutex_unlock(&show_priv->dev->struct_mutex);
 
diff --git a/drivers/gpu/drm/mxsfb/mxsfb_crtc.c b/drivers/gpu/drm/mxsfb/mxsfb_crtc.c
index 12421567af89..b69ace8bf526 100644
--- a/drivers/gpu/drm/mxsfb/mxsfb_crtc.c
+++ b/drivers/gpu/drm/mxsfb/mxsfb_crtc.c
@@ -95,8 +95,11 @@ static void mxsfb_set_bus_fmt(struct mxsfb_drm_private *mxsfb)
 
 	reg = readl(mxsfb->base + LCDC_CTRL);
 
-	if (mxsfb->connector.display_info.num_bus_formats)
-		bus_format = mxsfb->connector.display_info.bus_formats[0];
+	if (mxsfb->connector->display_info.num_bus_formats)
+		bus_format = mxsfb->connector->display_info.bus_formats[0];
+
+	DRM_DEV_DEBUG_DRIVER(drm->dev, "Using bus_format: 0x%08X\n",
+			     bus_format);
 
 	reg &= ~CTRL_BUS_WIDTH_MASK;
 	switch (bus_format) {
@@ -204,8 +207,9 @@ static dma_addr_t mxsfb_get_fb_paddr(struct mxsfb_drm_private *mxsfb)
 
 static void mxsfb_crtc_mode_set_nofb(struct mxsfb_drm_private *mxsfb)
 {
+	struct drm_device *drm = mxsfb->pipe.crtc.dev;
 	struct drm_display_mode *m = &mxsfb->pipe.crtc.state->adjusted_mode;
-	const u32 bus_flags = mxsfb->connector.display_info.bus_flags;
+	u32 bus_flags = mxsfb->connector->display_info.bus_flags;
 	u32 vdctrl0, vsync_pulse_len, hsync_pulse_len;
 	int err;
 
@@ -229,6 +233,16 @@ static void mxsfb_crtc_mode_set_nofb(struct mxsfb_drm_private *mxsfb)
 
 	clk_set_rate(mxsfb->clk, m->crtc_clock * 1000);
 
+	if (mxsfb->bridge && mxsfb->bridge->timings)
+		bus_flags = mxsfb->bridge->timings->input_bus_flags;
+
+	DRM_DEV_DEBUG_DRIVER(drm->dev, "Pixel clock: %dkHz (actual: %dkHz)\n",
+			     m->crtc_clock,
+			     (int)(clk_get_rate(mxsfb->clk) / 1000));
+	DRM_DEV_DEBUG_DRIVER(drm->dev, "Connector bus_flags: 0x%08X\n",
+			     bus_flags);
+	DRM_DEV_DEBUG_DRIVER(drm->dev, "Mode flags: 0x%08X\n", m->flags);
+
 	writel(TRANSFER_COUNT_SET_VCOUNT(m->crtc_vdisplay) |
 	       TRANSFER_COUNT_SET_HCOUNT(m->crtc_hdisplay),
 	       mxsfb->base + mxsfb->devdata->transfer_count);
diff --git a/drivers/gpu/drm/mxsfb/mxsfb_drv.c b/drivers/gpu/drm/mxsfb/mxsfb_drv.c
index e8506335cd15..497cf443a9af 100644
--- a/drivers/gpu/drm/mxsfb/mxsfb_drv.c
+++ b/drivers/gpu/drm/mxsfb/mxsfb_drv.c
@@ -101,9 +101,25 @@ static void mxsfb_pipe_enable(struct drm_simple_display_pipe *pipe,
 			      struct drm_crtc_state *crtc_state,
 			      struct drm_plane_state *plane_state)
 {
+	struct drm_connector *connector;
 	struct mxsfb_drm_private *mxsfb = drm_pipe_to_mxsfb_drm_private(pipe);
 	struct drm_device *drm = pipe->plane.dev;
 
+	if (!mxsfb->connector) {
+		list_for_each_entry(connector,
+				    &drm->mode_config.connector_list,
+				    head)
+			if (connector->encoder == &mxsfb->pipe.encoder) {
+				mxsfb->connector = connector;
+				break;
+			}
+	}
+
+	if (!mxsfb->connector) {
+		dev_warn(drm->dev, "No connector attached, using default\n");
+		mxsfb->connector = &mxsfb->panel_connector;
+	}
+
 	pm_runtime_get_sync(drm->dev);
 	drm_panel_prepare(mxsfb->panel);
 	mxsfb_crtc_enable(mxsfb);
@@ -129,6 +145,9 @@ static void mxsfb_pipe_disable(struct drm_simple_display_pipe *pipe)
 		drm_crtc_send_vblank_event(crtc, event);
 	}
 	spin_unlock_irq(&drm->event_lock);
+
+	if (mxsfb->connector != &mxsfb->panel_connector)
+		mxsfb->connector = NULL;
 }
 
 static void mxsfb_pipe_update(struct drm_simple_display_pipe *pipe,
@@ -226,16 +245,33 @@ static int mxsfb_load(struct drm_device *drm, unsigned long flags)
 
 	ret = drm_simple_display_pipe_init(drm, &mxsfb->pipe, &mxsfb_funcs,
 			mxsfb_formats, ARRAY_SIZE(mxsfb_formats), NULL,
-			&mxsfb->connector);
+			mxsfb->connector);
 	if (ret < 0) {
 		dev_err(drm->dev, "Cannot setup simple display pipe\n");
 		goto err_vblank;
 	}
 
-	ret = drm_panel_attach(mxsfb->panel, &mxsfb->connector);
-	if (ret) {
-		dev_err(drm->dev, "Cannot connect panel\n");
-		goto err_vblank;
+	/*
+	 * Attach panel only if there is one.
+	 * If there is no panel attach, it must be a bridge. In this case, we
+	 * need a reference to its connector for a proper initialization.
+	 * We will do this check in pipe->enable(), since the connector won't
+	 * be attached to an encoder until then.
+	 */
+
+	if (mxsfb->panel) {
+		ret = drm_panel_attach(mxsfb->panel, mxsfb->connector);
+		if (ret) {
+			dev_err(drm->dev, "Cannot connect panel: %d\n", ret);
+			goto err_vblank;
+		}
+	} else if (mxsfb->bridge) {
+		ret = drm_simple_display_pipe_attach_bridge(&mxsfb->pipe,
+							    mxsfb->bridge);
+		if (ret) {
+			dev_err(drm->dev, "Cannot connect bridge: %d\n", ret);
+			goto err_vblank;
+		}
 	}
 
 	drm->mode_config.min_width	= MXSFB_MIN_XRES;
diff --git a/drivers/gpu/drm/mxsfb/mxsfb_drv.h b/drivers/gpu/drm/mxsfb/mxsfb_drv.h
index d975300dca05..0b65b5194a9c 100644
--- a/drivers/gpu/drm/mxsfb/mxsfb_drv.h
+++ b/drivers/gpu/drm/mxsfb/mxsfb_drv.h
@@ -27,8 +27,10 @@ struct mxsfb_drm_private {
 	struct clk			*clk_disp_axi;
 
 	struct drm_simple_display_pipe	pipe;
-	struct drm_connector		connector;
+	struct drm_connector		panel_connector;
+	struct drm_connector		*connector;
 	struct drm_panel		*panel;
+	struct drm_bridge		*bridge;
 };
 
 int mxsfb_setup_crtc(struct drm_device *dev);
diff --git a/drivers/gpu/drm/mxsfb/mxsfb_out.c b/drivers/gpu/drm/mxsfb/mxsfb_out.c
index be36f4d6cc96..4eb94744c526 100644
--- a/drivers/gpu/drm/mxsfb/mxsfb_out.c
+++ b/drivers/gpu/drm/mxsfb/mxsfb_out.c
@@ -21,7 +21,8 @@
 static struct mxsfb_drm_private *
 drm_connector_to_mxsfb_drm_private(struct drm_connector *connector)
 {
-	return container_of(connector, struct mxsfb_drm_private, connector);
+	return container_of(connector, struct mxsfb_drm_private,
+			    panel_connector);
 }
 
 static int mxsfb_panel_get_modes(struct drm_connector *connector)
@@ -76,22 +77,23 @@ static const struct drm_connector_funcs mxsfb_panel_connector_funcs = {
 int mxsfb_create_output(struct drm_device *drm)
 {
 	struct mxsfb_drm_private *mxsfb = drm->dev_private;
-	struct drm_panel *panel;
 	int ret;
 
-	ret = drm_of_find_panel_or_bridge(drm->dev->of_node, 0, 0, &panel, NULL);
+	ret = drm_of_find_panel_or_bridge(drm->dev->of_node, 0, 0,
+					  &mxsfb->panel, &mxsfb->bridge);
 	if (ret)
 		return ret;
 
-	mxsfb->connector.dpms = DRM_MODE_DPMS_OFF;
-	mxsfb->connector.polled = 0;
-	drm_connector_helper_add(&mxsfb->connector,
-			&mxsfb_panel_connector_helper_funcs);
-	ret = drm_connector_init(drm, &mxsfb->connector,
-				 &mxsfb_panel_connector_funcs,
-				 DRM_MODE_CONNECTOR_Unknown);
-	if (!ret)
-		mxsfb->panel = panel;
+	if (mxsfb->panel) {
+		mxsfb->connector = &mxsfb->panel_connector;
+		mxsfb->connector->dpms = DRM_MODE_DPMS_OFF;
+		mxsfb->connector->polled = 0;
+		drm_connector_helper_add(mxsfb->connector,
+					 &mxsfb_panel_connector_helper_funcs);
+		ret = drm_connector_init(drm, mxsfb->connector,
+					 &mxsfb_panel_connector_funcs,
+					 DRM_MODE_CONNECTOR_Unknown);
+	}
 
 	return ret;
 }
diff --git a/drivers/gpu/drm/nouveau/dispnv04/disp.c b/drivers/gpu/drm/nouveau/dispnv04/disp.c
index dc64863b5fd8..44ee82d0c9b6 100644
--- a/drivers/gpu/drm/nouveau/dispnv04/disp.c
+++ b/drivers/gpu/drm/nouveau/dispnv04/disp.c
@@ -256,7 +256,7 @@ nv04_display_create(struct drm_device *dev)
 
 	list_for_each_entry_safe(connector, ct,
 				 &dev->mode_config.connector_list, head) {
-		if (!connector->encoder_ids[0]) {
+		if (!connector->possible_encoders) {
 			NV_WARN(drm, "%s has no encoders, removing\n",
 				connector->name);
 			connector->funcs->destroy(connector);
diff --git a/drivers/gpu/drm/nouveau/dispnv50/disp.c b/drivers/gpu/drm/nouveau/dispnv50/disp.c
index b46be8a091e9..549486f1d937 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/disp.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/disp.c
@@ -986,20 +986,11 @@ nv50_mstc_atomic_check(struct drm_connector *connector,
 	return drm_dp_atomic_release_vcpi_slots(state, mgr, mstc->port);
 }
 
-static const struct drm_connector_helper_funcs
-nv50_mstc_help = {
-	.get_modes = nv50_mstc_get_modes,
-	.mode_valid = nv50_mstc_mode_valid,
-	.best_encoder = nv50_mstc_best_encoder,
-	.atomic_best_encoder = nv50_mstc_atomic_best_encoder,
-	.atomic_check = nv50_mstc_atomic_check,
-};
-
-static enum drm_connector_status
-nv50_mstc_detect(struct drm_connector *connector, bool force)
+static int
+nv50_mstc_detect(struct drm_connector *connector,
+		 struct drm_modeset_acquire_ctx *ctx, bool force)
 {
 	struct nv50_mstc *mstc = nv50_mstc(connector);
-	enum drm_connector_status conn_status;
 	int ret;
 
 	if (drm_connector_is_unregistered(connector))
@@ -1009,14 +1000,24 @@ nv50_mstc_detect(struct drm_connector *connector, bool force)
 	if (ret < 0 && ret != -EACCES)
 		return connector_status_disconnected;
 
-	conn_status = drm_dp_mst_detect_port(connector, mstc->port->mgr,
-					     mstc->port);
+	ret = drm_dp_mst_detect_port(connector, ctx, mstc->port->mgr,
+				     mstc->port);
 
 	pm_runtime_mark_last_busy(connector->dev->dev);
 	pm_runtime_put_autosuspend(connector->dev->dev);
-	return conn_status;
+	return ret;
 }
 
+static const struct drm_connector_helper_funcs
+nv50_mstc_help = {
+	.get_modes = nv50_mstc_get_modes,
+	.mode_valid = nv50_mstc_mode_valid,
+	.best_encoder = nv50_mstc_best_encoder,
+	.atomic_best_encoder = nv50_mstc_atomic_best_encoder,
+	.atomic_check = nv50_mstc_atomic_check,
+	.detect_ctx = nv50_mstc_detect,
+};
+
 static void
 nv50_mstc_destroy(struct drm_connector *connector)
 {
@@ -1031,7 +1032,6 @@ nv50_mstc_destroy(struct drm_connector *connector)
 static const struct drm_connector_funcs
 nv50_mstc = {
 	.reset = nouveau_conn_reset,
-	.detect = nv50_mstc_detect,
 	.fill_modes = drm_helper_probe_single_connector_modes,
 	.destroy = nv50_mstc_destroy,
 	.atomic_duplicate_state = nouveau_conn_atomic_duplicate_state,
@@ -1309,14 +1309,14 @@ nv50_mstm_fini(struct nv50_mstm *mstm)
 }
 
 static void
-nv50_mstm_init(struct nv50_mstm *mstm)
+nv50_mstm_init(struct nv50_mstm *mstm, bool runtime)
 {
 	int ret;
 
 	if (!mstm || !mstm->mgr.mst_state)
 		return;
 
-	ret = drm_dp_mst_topology_mgr_resume(&mstm->mgr);
+	ret = drm_dp_mst_topology_mgr_resume(&mstm->mgr, !runtime);
 	if (ret == -1) {
 		drm_dp_mst_topology_mgr_set_mst(&mstm->mgr, false);
 		drm_kms_helper_hotplug_event(mstm->mgr.dev);
@@ -2263,7 +2263,7 @@ nv50_display_init(struct drm_device *dev, bool resume, bool runtime)
 		if (encoder->encoder_type != DRM_MODE_ENCODER_DPMST) {
 			struct nouveau_encoder *nv_encoder =
 				nouveau_encoder(encoder);
-			nv50_mstm_init(nv_encoder->dp.mstm);
+			nv50_mstm_init(nv_encoder->dp.mstm, runtime);
 		}
 	}
 
@@ -2392,7 +2392,7 @@ nv50_display_create(struct drm_device *dev)
 
 	/* cull any connectors we created that don't have an encoder */
 	list_for_each_entry_safe(connector, tmp, &dev->mode_config.connector_list, head) {
-		if (connector->encoder_ids[0])
+		if (connector->possible_encoders)
 			continue;
 
 		NV_WARN(drm, "%s has no encoders, removing\n",
diff --git a/drivers/gpu/drm/nouveau/nouveau_connector.c b/drivers/gpu/drm/nouveau/nouveau_connector.c
index 94dfa2e5a9ab..5b413588b823 100644
--- a/drivers/gpu/drm/nouveau/nouveau_connector.c
+++ b/drivers/gpu/drm/nouveau/nouveau_connector.c
@@ -365,9 +365,8 @@ find_encoder(struct drm_connector *connector, int type)
 {
 	struct nouveau_encoder *nv_encoder;
 	struct drm_encoder *enc;
-	int i;
 
-	drm_connector_for_each_possible_encoder(connector, enc, i) {
+	drm_connector_for_each_possible_encoder(connector, enc) {
 		nv_encoder = nouveau_encoder(enc);
 
 		if (type == DCB_OUTPUT_ANY ||
@@ -414,10 +413,10 @@ nouveau_connector_ddc_detect(struct drm_connector *connector)
 	struct drm_device *dev = connector->dev;
 	struct nouveau_encoder *nv_encoder = NULL, *found = NULL;
 	struct drm_encoder *encoder;
-	int i, ret;
+	int ret;
 	bool switcheroo_ddc = false;
 
-	drm_connector_for_each_possible_encoder(connector, encoder, i) {
+	drm_connector_for_each_possible_encoder(connector, encoder) {
 		nv_encoder = nouveau_encoder(encoder);
 
 		switch (nv_encoder->dcb->type) {
@@ -1131,6 +1130,16 @@ nouveau_connector_hotplug(struct nvif_notify *notify)
 	const char *name = connector->name;
 	struct nouveau_encoder *nv_encoder;
 	int ret;
+	bool plugged = (rep->mask != NVIF_NOTIFY_CONN_V0_UNPLUG);
+
+	if (rep->mask & NVIF_NOTIFY_CONN_V0_IRQ) {
+		NV_DEBUG(drm, "service %s\n", name);
+		drm_dp_cec_irq(&nv_connector->aux);
+		if ((nv_encoder = find_encoder(connector, DCB_OUTPUT_DP)))
+			nv50_mstm_service(nv_encoder->dp.mstm);
+
+		return NVIF_NOTIFY_KEEP;
+	}
 
 	ret = pm_runtime_get(drm->dev->dev);
 	if (ret == 0) {
@@ -1151,25 +1160,16 @@ nouveau_connector_hotplug(struct nvif_notify *notify)
 		return NVIF_NOTIFY_DROP;
 	}
 
-	if (rep->mask & NVIF_NOTIFY_CONN_V0_IRQ) {
-		NV_DEBUG(drm, "service %s\n", name);
-		drm_dp_cec_irq(&nv_connector->aux);
-		if ((nv_encoder = find_encoder(connector, DCB_OUTPUT_DP)))
-			nv50_mstm_service(nv_encoder->dp.mstm);
-	} else {
-		bool plugged = (rep->mask != NVIF_NOTIFY_CONN_V0_UNPLUG);
-
+	if (!plugged)
+		drm_dp_cec_unset_edid(&nv_connector->aux);
+	NV_DEBUG(drm, "%splugged %s\n", plugged ? "" : "un", name);
+	if ((nv_encoder = find_encoder(connector, DCB_OUTPUT_DP))) {
 		if (!plugged)
-			drm_dp_cec_unset_edid(&nv_connector->aux);
-		NV_DEBUG(drm, "%splugged %s\n", plugged ? "" : "un", name);
-		if ((nv_encoder = find_encoder(connector, DCB_OUTPUT_DP))) {
-			if (!plugged)
-				nv50_mstm_remove(nv_encoder->dp.mstm);
-		}
-
-		drm_helper_hpd_irq_event(connector->dev);
+			nv50_mstm_remove(nv_encoder->dp.mstm);
 	}
 
+	drm_helper_hpd_irq_event(connector->dev);
+
 	pm_runtime_mark_last_busy(drm->dev->dev);
 	pm_runtime_put_autosuspend(drm->dev->dev);
 	return NVIF_NOTIFY_KEEP;
@@ -1415,8 +1415,7 @@ nouveau_connector_create(struct drm_device *dev,
 	switch (type) {
 	case DRM_MODE_CONNECTOR_DisplayPort:
 	case DRM_MODE_CONNECTOR_eDP:
-		drm_dp_cec_register_connector(&nv_connector->aux,
-					      connector->name, dev->dev);
+		drm_dp_cec_register_connector(&nv_connector->aux, connector);
 		break;
 	}
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_display.c b/drivers/gpu/drm/nouveau/nouveau_display.c
index 6f038511a03a..53f9bceaf17a 100644
--- a/drivers/gpu/drm/nouveau/nouveau_display.c
+++ b/drivers/gpu/drm/nouveau/nouveau_display.c
@@ -407,6 +407,17 @@ nouveau_display_init(struct drm_device *dev, bool resume, bool runtime)
 	struct drm_connector_list_iter conn_iter;
 	int ret;
 
+	/*
+	 * Enable hotplug interrupts (done as early as possible, since we need
+	 * them for MST)
+	 */
+	drm_connector_list_iter_begin(dev, &conn_iter);
+	nouveau_for_each_non_mst_connector_iter(connector, &conn_iter) {
+		struct nouveau_connector *conn = nouveau_connector(connector);
+		nvif_notify_get(&conn->hpd);
+	}
+	drm_connector_list_iter_end(&conn_iter);
+
 	ret = disp->init(dev, resume, runtime);
 	if (ret)
 		return ret;
@@ -416,14 +427,6 @@ nouveau_display_init(struct drm_device *dev, bool resume, bool runtime)
 	 */
 	drm_kms_helper_poll_enable(dev);
 
-	/* enable hotplug interrupts */
-	drm_connector_list_iter_begin(dev, &conn_iter);
-	nouveau_for_each_non_mst_connector_iter(connector, &conn_iter) {
-		struct nouveau_connector *conn = nouveau_connector(connector);
-		nvif_notify_get(&conn->hpd);
-	}
-	drm_connector_list_iter_end(&conn_iter);
-
 	return ret;
 }
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_svm.c b/drivers/gpu/drm/nouveau/nouveau_svm.c
index 668d4bd0c118..df9bf1fd1bc0 100644
--- a/drivers/gpu/drm/nouveau/nouveau_svm.c
+++ b/drivers/gpu/drm/nouveau/nouveau_svm.c
@@ -88,6 +88,7 @@ nouveau_ivmm_find(struct nouveau_svm *svm, u64 inst)
 }
 
 struct nouveau_svmm {
+	struct mmu_notifier notifier;
 	struct nouveau_vmm *vmm;
 	struct {
 		unsigned long start;
@@ -95,9 +96,6 @@ struct nouveau_svmm {
 	} unmanaged;
 
 	struct mutex mutex;
-
-	struct mm_struct *mm;
-	struct hmm_mirror mirror;
 };
 
 #define SVMM_DBG(s,f,a...)                                                     \
@@ -251,10 +249,11 @@ nouveau_svmm_invalidate(struct nouveau_svmm *svmm, u64 start, u64 limit)
 }
 
 static int
-nouveau_svmm_sync_cpu_device_pagetables(struct hmm_mirror *mirror,
-					const struct mmu_notifier_range *update)
+nouveau_svmm_invalidate_range_start(struct mmu_notifier *mn,
+				    const struct mmu_notifier_range *update)
 {
-	struct nouveau_svmm *svmm = container_of(mirror, typeof(*svmm), mirror);
+	struct nouveau_svmm *svmm =
+		container_of(mn, struct nouveau_svmm, notifier);
 	unsigned long start = update->start;
 	unsigned long limit = update->end;
 
@@ -264,6 +263,9 @@ nouveau_svmm_sync_cpu_device_pagetables(struct hmm_mirror *mirror,
 	SVMM_DBG(svmm, "invalidate %016lx-%016lx", start, limit);
 
 	mutex_lock(&svmm->mutex);
+	if (unlikely(!svmm->vmm))
+		goto out;
+
 	if (limit > svmm->unmanaged.start && start < svmm->unmanaged.limit) {
 		if (start < svmm->unmanaged.start) {
 			nouveau_svmm_invalidate(svmm, start,
@@ -273,19 +275,20 @@ nouveau_svmm_sync_cpu_device_pagetables(struct hmm_mirror *mirror,
 	}
 
 	nouveau_svmm_invalidate(svmm, start, limit);
+
+out:
 	mutex_unlock(&svmm->mutex);
 	return 0;
 }
 
-static void
-nouveau_svmm_release(struct hmm_mirror *mirror)
+static void nouveau_svmm_free_notifier(struct mmu_notifier *mn)
 {
+	kfree(container_of(mn, struct nouveau_svmm, notifier));
 }
 
-static const struct hmm_mirror_ops
-nouveau_svmm = {
-	.sync_cpu_device_pagetables = nouveau_svmm_sync_cpu_device_pagetables,
-	.release = nouveau_svmm_release,
+static const struct mmu_notifier_ops nouveau_mn_ops = {
+	.invalidate_range_start = nouveau_svmm_invalidate_range_start,
+	.free_notifier = nouveau_svmm_free_notifier,
 };
 
 void
@@ -293,8 +296,10 @@ nouveau_svmm_fini(struct nouveau_svmm **psvmm)
 {
 	struct nouveau_svmm *svmm = *psvmm;
 	if (svmm) {
-		hmm_mirror_unregister(&svmm->mirror);
-		kfree(*psvmm);
+		mutex_lock(&svmm->mutex);
+		svmm->vmm = NULL;
+		mutex_unlock(&svmm->mutex);
+		mmu_notifier_put(&svmm->notifier);
 		*psvmm = NULL;
 	}
 }
@@ -320,7 +325,7 @@ nouveau_svmm_init(struct drm_device *dev, void *data,
 	mutex_lock(&cli->mutex);
 	if (cli->svm.cli) {
 		ret = -EBUSY;
-		goto done;
+		goto out_free;
 	}
 
 	/* Allocate a new GPU VMM that can support SVM (managed by the
@@ -335,24 +340,26 @@ nouveau_svmm_init(struct drm_device *dev, void *data,
 				.fault_replay = true,
 			    }, sizeof(struct gp100_vmm_v0), &cli->svm.vmm);
 	if (ret)
-		goto done;
-
-	/* Enable HMM mirroring of CPU address-space to VMM. */
-	svmm->mm = get_task_mm(current);
-	down_write(&svmm->mm->mmap_sem);
-	svmm->mirror.ops = &nouveau_svmm;
-	ret = hmm_mirror_register(&svmm->mirror, svmm->mm);
-	if (ret == 0) {
-		cli->svm.svmm = svmm;
-		cli->svm.cli = cli;
-	}
-	up_write(&svmm->mm->mmap_sem);
-	mmput(svmm->mm);
+		goto out_free;
 
-done:
+	down_write(&current->mm->mmap_sem);
+	svmm->notifier.ops = &nouveau_mn_ops;
+	ret = __mmu_notifier_register(&svmm->notifier, current->mm);
 	if (ret)
-		nouveau_svmm_fini(&svmm);
+		goto out_mm_unlock;
+	/* Note, ownership of svmm transfers to mmu_notifier */
+
+	cli->svm.svmm = svmm;
+	cli->svm.cli = cli;
+	up_write(&current->mm->mmap_sem);
 	mutex_unlock(&cli->mutex);
+	return 0;
+
+out_mm_unlock:
+	up_write(&current->mm->mmap_sem);
+out_free:
+	mutex_unlock(&cli->mutex);
+	kfree(svmm);
 	return ret;
 }
 
@@ -475,43 +482,90 @@ nouveau_svm_fault_cache(struct nouveau_svm *svm,
 		fault->inst, fault->addr, fault->access);
 }
 
-static inline bool
-nouveau_range_done(struct hmm_range *range)
+struct svm_notifier {
+	struct mmu_interval_notifier notifier;
+	struct nouveau_svmm *svmm;
+};
+
+static bool nouveau_svm_range_invalidate(struct mmu_interval_notifier *mni,
+					 const struct mmu_notifier_range *range,
+					 unsigned long cur_seq)
 {
-	bool ret = hmm_range_valid(range);
+	struct svm_notifier *sn =
+		container_of(mni, struct svm_notifier, notifier);
 
-	hmm_range_unregister(range);
-	return ret;
+	/*
+	 * serializes the update to mni->invalidate_seq done by caller and
+	 * prevents invalidation of the PTE from progressing while HW is being
+	 * programmed. This is very hacky and only works because the normal
+	 * notifier that does invalidation is always called after the range
+	 * notifier.
+	 */
+	if (mmu_notifier_range_blockable(range))
+		mutex_lock(&sn->svmm->mutex);
+	else if (!mutex_trylock(&sn->svmm->mutex))
+		return false;
+	mmu_interval_set_seq(mni, cur_seq);
+	mutex_unlock(&sn->svmm->mutex);
+	return true;
 }
 
-static int
-nouveau_range_fault(struct nouveau_svmm *svmm, struct hmm_range *range)
+static const struct mmu_interval_notifier_ops nouveau_svm_mni_ops = {
+	.invalidate = nouveau_svm_range_invalidate,
+};
+
+static int nouveau_range_fault(struct nouveau_svmm *svmm,
+			       struct nouveau_drm *drm, void *data, u32 size,
+			       u64 *pfns, struct svm_notifier *notifier)
 {
+	unsigned long timeout =
+		jiffies + msecs_to_jiffies(HMM_RANGE_DEFAULT_TIMEOUT);
+	/* Have HMM fault pages within the fault window to the GPU. */
+	struct hmm_range range = {
+		.notifier = &notifier->notifier,
+		.start = notifier->notifier.interval_tree.start,
+		.end = notifier->notifier.interval_tree.last + 1,
+		.pfns = pfns,
+		.flags = nouveau_svm_pfn_flags,
+		.values = nouveau_svm_pfn_values,
+		.pfn_shift = NVIF_VMM_PFNMAP_V0_ADDR_SHIFT,
+	};
+	struct mm_struct *mm = notifier->notifier.mm;
 	long ret;
 
-	range->default_flags = 0;
-	range->pfn_flags_mask = -1UL;
+	while (true) {
+		if (time_after(jiffies, timeout))
+			return -EBUSY;
+
+		range.notifier_seq = mmu_interval_read_begin(range.notifier);
+		range.default_flags = 0;
+		range.pfn_flags_mask = -1UL;
+		down_read(&mm->mmap_sem);
+		ret = hmm_range_fault(&range, 0);
+		up_read(&mm->mmap_sem);
+		if (ret <= 0) {
+			if (ret == 0 || ret == -EBUSY)
+				continue;
+			return ret;
+		}
 
-	ret = hmm_range_register(range, &svmm->mirror);
-	if (ret) {
-		up_read(&svmm->mm->mmap_sem);
-		return (int)ret;
+		mutex_lock(&svmm->mutex);
+		if (mmu_interval_read_retry(range.notifier,
+					    range.notifier_seq)) {
+			mutex_unlock(&svmm->mutex);
+			continue;
+		}
+		break;
 	}
 
-	if (!hmm_range_wait_until_valid(range, HMM_RANGE_DEFAULT_TIMEOUT)) {
-		up_read(&svmm->mm->mmap_sem);
-		return -EBUSY;
-	}
+	nouveau_dmem_convert_pfn(drm, &range);
 
-	ret = hmm_range_fault(range, 0);
-	if (ret <= 0) {
-		if (ret == 0)
-			ret = -EBUSY;
-		up_read(&svmm->mm->mmap_sem);
-		hmm_range_unregister(range);
-		return ret;
-	}
-	return 0;
+	svmm->vmm->vmm.object.client->super = true;
+	ret = nvif_object_ioctl(&svmm->vmm->vmm.object, data, size, NULL);
+	svmm->vmm->vmm.object.client->super = false;
+	mutex_unlock(&svmm->mutex);
+
+	return ret;
 }
 
 static int
@@ -531,7 +585,6 @@ nouveau_svm_fault(struct nvif_notify *notify)
 		} i;
 		u64 phys[16];
 	} args;
-	struct hmm_range range;
 	struct vm_area_struct *vma;
 	u64 inst, start, limit;
 	int fi, fn, pi, fill;
@@ -587,6 +640,9 @@ nouveau_svm_fault(struct nvif_notify *notify)
 	args.i.p.version = 0;
 
 	for (fi = 0; fn = fi + 1, fi < buffer->fault_nr; fi = fn) {
+		struct svm_notifier notifier;
+		struct mm_struct *mm;
+
 		/* Cancel any faults from non-SVM channels. */
 		if (!(svmm = buffer->fault[fi]->svmm)) {
 			nouveau_svm_fault_cancel_fault(svm, buffer->fault[fi]);
@@ -606,24 +662,32 @@ nouveau_svm_fault(struct nvif_notify *notify)
 			start = max_t(u64, start, svmm->unmanaged.limit);
 		SVMM_DBG(svmm, "wndw %016llx-%016llx", start, limit);
 
+		mm = svmm->notifier.mm;
+		if (!mmget_not_zero(mm)) {
+			nouveau_svm_fault_cancel_fault(svm, buffer->fault[fi]);
+			continue;
+		}
+
 		/* Intersect fault window with the CPU VMA, cancelling
 		 * the fault if the address is invalid.
 		 */
-		down_read(&svmm->mm->mmap_sem);
-		vma = find_vma_intersection(svmm->mm, start, limit);
+		down_read(&mm->mmap_sem);
+		vma = find_vma_intersection(mm, start, limit);
 		if (!vma) {
 			SVMM_ERR(svmm, "wndw %016llx-%016llx", start, limit);
-			up_read(&svmm->mm->mmap_sem);
+			up_read(&mm->mmap_sem);
+			mmput(mm);
 			nouveau_svm_fault_cancel_fault(svm, buffer->fault[fi]);
 			continue;
 		}
 		start = max_t(u64, start, vma->vm_start);
 		limit = min_t(u64, limit, vma->vm_end);
+		up_read(&mm->mmap_sem);
 		SVMM_DBG(svmm, "wndw %016llx-%016llx", start, limit);
 
 		if (buffer->fault[fi]->addr != start) {
 			SVMM_ERR(svmm, "addr %016llx", buffer->fault[fi]->addr);
-			up_read(&svmm->mm->mmap_sem);
+			mmput(mm);
 			nouveau_svm_fault_cancel_fault(svm, buffer->fault[fi]);
 			continue;
 		}
@@ -679,33 +743,19 @@ nouveau_svm_fault(struct nvif_notify *notify)
 			 args.i.p.addr,
 			 args.i.p.addr + args.i.p.size, fn - fi);
 
-		/* Have HMM fault pages within the fault window to the GPU. */
-		range.start = args.i.p.addr;
-		range.end = args.i.p.addr + args.i.p.size;
-		range.pfns = args.phys;
-		range.flags = nouveau_svm_pfn_flags;
-		range.values = nouveau_svm_pfn_values;
-		range.pfn_shift = NVIF_VMM_PFNMAP_V0_ADDR_SHIFT;
-again:
-		ret = nouveau_range_fault(svmm, &range);
-		if (ret == 0) {
-			mutex_lock(&svmm->mutex);
-			if (!nouveau_range_done(&range)) {
-				mutex_unlock(&svmm->mutex);
-				goto again;
-			}
-
-			nouveau_dmem_convert_pfn(svm->drm, &range);
-
-			svmm->vmm->vmm.object.client->super = true;
-			ret = nvif_object_ioctl(&svmm->vmm->vmm.object,
-						&args, sizeof(args.i) +
-						pi * sizeof(args.phys[0]),
-						NULL);
-			svmm->vmm->vmm.object.client->super = false;
-			mutex_unlock(&svmm->mutex);
-			up_read(&svmm->mm->mmap_sem);
+		notifier.svmm = svmm;
+		ret = mmu_interval_notifier_insert(&notifier.notifier,
+						   svmm->notifier.mm,
+						   args.i.p.addr, args.i.p.size,
+						   &nouveau_svm_mni_ops);
+		if (!ret) {
+			ret = nouveau_range_fault(
+				svmm, svm->drm, &args,
+				sizeof(args.i) + pi * sizeof(args.phys[0]),
+				args.phys, &notifier);
+			mmu_interval_notifier_remove(&notifier.notifier);
 		}
+		mmput(mm);
 
 		/* Cancel any faults in the window whose pages didn't manage
 		 * to keep their valid bit, or stay writeable when required.
@@ -714,10 +764,10 @@ again:
 		 */
 		while (fi < fn) {
 			struct nouveau_svm_fault *fault = buffer->fault[fi++];
-			pi = (fault->addr - range.start) >> PAGE_SHIFT;
+			pi = (fault->addr - args.i.p.addr) >> PAGE_SHIFT;
 			if (ret ||
-			     !(range.pfns[pi] & NVIF_VMM_PFNMAP_V0_V) ||
-			    (!(range.pfns[pi] & NVIF_VMM_PFNMAP_V0_W) &&
+			     !(args.phys[pi] & NVIF_VMM_PFNMAP_V0_V) ||
+			    (!(args.phys[pi] & NVIF_VMM_PFNMAP_V0_W) &&
 			     fault->access != 0 && fault->access != 3)) {
 				nouveau_svm_fault_cancel_fault(svm, fault);
 				continue;
diff --git a/drivers/gpu/drm/nouveau/nouveau_ttm.c b/drivers/gpu/drm/nouveau/nouveau_ttm.c
index f0daf958e03a..77a0c6ad3cef 100644
--- a/drivers/gpu/drm/nouveau/nouveau_ttm.c
+++ b/drivers/gpu/drm/nouveau/nouveau_ttm.c
@@ -236,6 +236,7 @@ nouveau_ttm_init(struct nouveau_drm *drm)
 	ret = ttm_bo_device_init(&drm->ttm.bdev,
 				  &nouveau_bo_driver,
 				  dev->anon_inode->i_mapping,
+				  dev->vma_offset_manager,
 				  drm->client.mmu.dmabits <= 32 ? true : false);
 	if (ret) {
 		NV_ERROR(drm, "error initialising bo driver, %d\n", ret);
diff --git a/drivers/gpu/drm/omapdrm/dss/Makefile b/drivers/gpu/drm/omapdrm/dss/Makefile
index 904101c5e79d..5950c3f52c2e 100644
--- a/drivers/gpu/drm/omapdrm/dss/Makefile
+++ b/drivers/gpu/drm/omapdrm/dss/Makefile
@@ -6,7 +6,7 @@ omapdss-base-y := base.o display.o dss-of.o output.o
 
 obj-$(CONFIG_OMAP2_DSS) += omapdss.o
 # Core DSS files
-omapdss-y := core.o dss.o dispc.o dispc_coefs.o \
+omapdss-y := dss.o dispc.o dispc_coefs.o \
 	pll.o video-pll.o
 omapdss-$(CONFIG_OMAP2_DSS_DPI) += dpi.o
 omapdss-$(CONFIG_OMAP2_DSS_VENC) += venc.o
diff --git a/drivers/gpu/drm/omapdrm/dss/core.c b/drivers/gpu/drm/omapdrm/dss/core.c
deleted file mode 100644
index 6ac497b63711..000000000000
--- a/drivers/gpu/drm/omapdrm/dss/core.c
+++ /dev/null
@@ -1,55 +0,0 @@
-// SPDX-License-Identifier: GPL-2.0-only
-/*
- * Copyright (C) 2009 Nokia Corporation
- * Author: Tomi Valkeinen <tomi.valkeinen@ti.com>
- *
- * Some code and ideas taken from drivers/video/omap/ driver
- * by Imre Deak.
- */
-
-#define DSS_SUBSYS_NAME "CORE"
-
-#include <linux/kernel.h>
-#include <linux/module.h>
-#include <linux/platform_device.h>
-
-#include "omapdss.h"
-#include "dss.h"
-
-/* INIT */
-static struct platform_driver * const omap_dss_drivers[] = {
-	&omap_dsshw_driver,
-	&omap_dispchw_driver,
-#ifdef CONFIG_OMAP2_DSS_DSI
-	&omap_dsihw_driver,
-#endif
-#ifdef CONFIG_OMAP2_DSS_VENC
-	&omap_venchw_driver,
-#endif
-#ifdef CONFIG_OMAP4_DSS_HDMI
-	&omapdss_hdmi4hw_driver,
-#endif
-#ifdef CONFIG_OMAP5_DSS_HDMI
-	&omapdss_hdmi5hw_driver,
-#endif
-};
-
-static int __init omap_dss_init(void)
-{
-	return platform_register_drivers(omap_dss_drivers,
-					 ARRAY_SIZE(omap_dss_drivers));
-}
-
-static void __exit omap_dss_exit(void)
-{
-	platform_unregister_drivers(omap_dss_drivers,
-				    ARRAY_SIZE(omap_dss_drivers));
-}
-
-module_init(omap_dss_init);
-module_exit(omap_dss_exit);
-
-MODULE_AUTHOR("Tomi Valkeinen <tomi.valkeinen@ti.com>");
-MODULE_DESCRIPTION("OMAP2/3 Display Subsystem");
-MODULE_LICENSE("GPL v2");
-
diff --git a/drivers/gpu/drm/omapdrm/dss/dispc.c b/drivers/gpu/drm/omapdrm/dss/dispc.c
index ed0ccbeed70f..413dbdd1771e 100644
--- a/drivers/gpu/drm/omapdrm/dss/dispc.c
+++ b/drivers/gpu/drm/omapdrm/dss/dispc.c
@@ -114,6 +114,7 @@ struct dispc_features {
 	const unsigned int num_reg_fields;
 	const enum omap_overlay_caps *overlay_caps;
 	const u32 **supported_color_modes;
+	const u32 *supported_scaler_color_modes;
 	unsigned int num_mgrs;
 	unsigned int num_ovls;
 	unsigned int buffer_size_unit;
@@ -184,9 +185,6 @@ struct dispc_device {
 
 	struct regmap *syscon_pol;
 	u32 syscon_pol_offset;
-
-	/* DISPC_CONTROL & DISPC_CONFIG lock*/
-	spinlock_t control_lock;
 };
 
 enum omap_color_component {
@@ -368,25 +366,17 @@ static inline u32 dispc_read_reg(struct dispc_device *dispc, u16 idx)
 static u32 mgr_fld_read(struct dispc_device *dispc, enum omap_channel channel,
 			enum mgr_reg_fields regfld)
 {
-	const struct dispc_reg_field rfld = mgr_desc[channel].reg_desc[regfld];
+	const struct dispc_reg_field *rfld = &mgr_desc[channel].reg_desc[regfld];
 
-	return REG_GET(dispc, rfld.reg, rfld.high, rfld.low);
+	return REG_GET(dispc, rfld->reg, rfld->high, rfld->low);
 }
 
 static void mgr_fld_write(struct dispc_device *dispc, enum omap_channel channel,
 			  enum mgr_reg_fields regfld, int val)
 {
-	const struct dispc_reg_field rfld = mgr_desc[channel].reg_desc[regfld];
-	const bool need_lock = rfld.reg == DISPC_CONTROL || rfld.reg == DISPC_CONFIG;
-	unsigned long flags;
+	const struct dispc_reg_field *rfld = &mgr_desc[channel].reg_desc[regfld];
 
-	if (need_lock) {
-		spin_lock_irqsave(&dispc->control_lock, flags);
-		REG_FLD_MOD(dispc, rfld.reg, val, rfld.high, rfld.low);
-		spin_unlock_irqrestore(&dispc->control_lock, flags);
-	} else {
-		REG_FLD_MOD(dispc, rfld.reg, val, rfld.high, rfld.low);
-	}
+	REG_FLD_MOD(dispc, rfld->reg, val, rfld->high, rfld->low);
 }
 
 static int dispc_get_num_ovls(struct dispc_device *dispc)
@@ -2510,6 +2500,19 @@ static int dispc_ovl_calc_scaling(struct dispc_device *dispc,
 	if (width == out_width && height == out_height)
 		return 0;
 
+	if (dispc->feat->supported_scaler_color_modes) {
+		const u32 *modes = dispc->feat->supported_scaler_color_modes;
+		unsigned int i;
+
+		for (i = 0; modes[i]; ++i) {
+			if (modes[i] == fourcc)
+				break;
+		}
+
+		if (modes[i] == 0)
+			return -EINVAL;
+	}
+
 	if (plane == OMAP_DSS_WB) {
 		switch (fourcc) {
 		case DRM_FORMAT_NV12:
@@ -4225,6 +4228,12 @@ static const u32 *omap4_dispc_supported_color_modes[] = {
 	DRM_FORMAT_RGBX8888),
 };
 
+static const u32 omap3_dispc_supported_scaler_color_modes[] = {
+	DRM_FORMAT_XRGB8888, DRM_FORMAT_RGB565, DRM_FORMAT_YUYV,
+	DRM_FORMAT_UYVY,
+	0,
+};
+
 static const struct dispc_features omap24xx_dispc_feats = {
 	.sw_start		=	5,
 	.fp_start		=	15,
@@ -4253,6 +4262,7 @@ static const struct dispc_features omap24xx_dispc_feats = {
 	.num_reg_fields		=	ARRAY_SIZE(omap2_dispc_reg_fields),
 	.overlay_caps		=	omap2_dispc_overlay_caps,
 	.supported_color_modes	=	omap2_dispc_supported_color_modes,
+	.supported_scaler_color_modes = COLOR_ARRAY(DRM_FORMAT_XRGB8888),
 	.num_mgrs		=	2,
 	.num_ovls		=	3,
 	.buffer_size_unit	=	1,
@@ -4287,6 +4297,7 @@ static const struct dispc_features omap34xx_rev1_0_dispc_feats = {
 	.num_reg_fields		=	ARRAY_SIZE(omap3_dispc_reg_fields),
 	.overlay_caps		=	omap3430_dispc_overlay_caps,
 	.supported_color_modes	=	omap3_dispc_supported_color_modes,
+	.supported_scaler_color_modes = omap3_dispc_supported_scaler_color_modes,
 	.num_mgrs		=	2,
 	.num_ovls		=	3,
 	.buffer_size_unit	=	1,
@@ -4321,6 +4332,7 @@ static const struct dispc_features omap34xx_rev3_0_dispc_feats = {
 	.num_reg_fields		=	ARRAY_SIZE(omap3_dispc_reg_fields),
 	.overlay_caps		=	omap3430_dispc_overlay_caps,
 	.supported_color_modes	=	omap3_dispc_supported_color_modes,
+	.supported_scaler_color_modes = omap3_dispc_supported_scaler_color_modes,
 	.num_mgrs		=	2,
 	.num_ovls		=	3,
 	.buffer_size_unit	=	1,
@@ -4355,6 +4367,7 @@ static const struct dispc_features omap36xx_dispc_feats = {
 	.num_reg_fields		=	ARRAY_SIZE(omap3_dispc_reg_fields),
 	.overlay_caps		=	omap3630_dispc_overlay_caps,
 	.supported_color_modes	=	omap3_dispc_supported_color_modes,
+	.supported_scaler_color_modes = omap3_dispc_supported_scaler_color_modes,
 	.num_mgrs		=	2,
 	.num_ovls		=	3,
 	.buffer_size_unit	=	1,
@@ -4389,6 +4402,7 @@ static const struct dispc_features am43xx_dispc_feats = {
 	.num_reg_fields		=	ARRAY_SIZE(omap3_dispc_reg_fields),
 	.overlay_caps		=	omap3430_dispc_overlay_caps,
 	.supported_color_modes	=	omap3_dispc_supported_color_modes,
+	.supported_scaler_color_modes = omap3_dispc_supported_scaler_color_modes,
 	.num_mgrs		=	1,
 	.num_ovls		=	3,
 	.buffer_size_unit	=	1,
@@ -4768,8 +4782,6 @@ static int dispc_bind(struct device *dev, struct device *master, void *data)
 	platform_set_drvdata(pdev, dispc);
 	dispc->dss = dss;
 
-	spin_lock_init(&dispc->control_lock);
-
 	/*
 	 * The OMAP3-based models can't be told apart using the compatible
 	 * string, use SoC device matching.
diff --git a/drivers/gpu/drm/omapdrm/dss/dsi.c b/drivers/gpu/drm/omapdrm/dss/dsi.c
index b30fcaa2d0f5..da16ea095f13 100644
--- a/drivers/gpu/drm/omapdrm/dss/dsi.c
+++ b/drivers/gpu/drm/omapdrm/dss/dsi.c
@@ -3548,7 +3548,7 @@ static int dsi_proto_config(struct dsi_data *dsi)
 
 static void dsi_proto_timings(struct dsi_data *dsi)
 {
-	unsigned int tlpx, tclk_zero, tclk_prepare, tclk_trail;
+	unsigned int tlpx, tclk_zero, tclk_prepare;
 	unsigned int tclk_pre, tclk_post;
 	unsigned int ths_prepare, ths_prepare_ths_zero, ths_zero;
 	unsigned int ths_trail, ths_exit;
@@ -3567,7 +3567,6 @@ static void dsi_proto_timings(struct dsi_data *dsi)
 
 	r = dsi_read_reg(dsi, DSI_DSIPHY_CFG1);
 	tlpx = FLD_GET(r, 20, 16) * 2;
-	tclk_trail = FLD_GET(r, 15, 8);
 	tclk_zero = FLD_GET(r, 7, 0);
 
 	r = dsi_read_reg(dsi, DSI_DSIPHY_CFG2);
diff --git a/drivers/gpu/drm/omapdrm/dss/dss.c b/drivers/gpu/drm/omapdrm/dss/dss.c
index e226324adb69..225ec808b01a 100644
--- a/drivers/gpu/drm/omapdrm/dss/dss.c
+++ b/drivers/gpu/drm/omapdrm/dss/dss.c
@@ -1083,7 +1083,7 @@ static const struct dss_features omap34xx_dss_feats = {
 
 static const struct dss_features omap3630_dss_feats = {
 	.model			=	DSS_MODEL_OMAP3,
-	.fck_div_max		=	32,
+	.fck_div_max		=	31,
 	.fck_freq_max		=	173000000,
 	.dss_fck_multiplier	=	1,
 	.parent_clk_name	=	"dpll4_ck",
@@ -1598,3 +1598,40 @@ struct platform_driver omap_dsshw_driver = {
 		.suppress_bind_attrs = true,
 	},
 };
+
+/* INIT */
+static struct platform_driver * const omap_dss_drivers[] = {
+	&omap_dsshw_driver,
+	&omap_dispchw_driver,
+#ifdef CONFIG_OMAP2_DSS_DSI
+	&omap_dsihw_driver,
+#endif
+#ifdef CONFIG_OMAP2_DSS_VENC
+	&omap_venchw_driver,
+#endif
+#ifdef CONFIG_OMAP4_DSS_HDMI
+	&omapdss_hdmi4hw_driver,
+#endif
+#ifdef CONFIG_OMAP5_DSS_HDMI
+	&omapdss_hdmi5hw_driver,
+#endif
+};
+
+static int __init omap_dss_init(void)
+{
+	return platform_register_drivers(omap_dss_drivers,
+					 ARRAY_SIZE(omap_dss_drivers));
+}
+
+static void __exit omap_dss_exit(void)
+{
+	platform_unregister_drivers(omap_dss_drivers,
+				    ARRAY_SIZE(omap_dss_drivers));
+}
+
+module_init(omap_dss_init);
+module_exit(omap_dss_exit);
+
+MODULE_AUTHOR("Tomi Valkeinen <tomi.valkeinen@ti.com>");
+MODULE_DESCRIPTION("OMAP2/3/4/5 Display Subsystem");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/gpu/drm/omapdrm/dss/hdmi4_core.c b/drivers/gpu/drm/omapdrm/dss/hdmi4_core.c
index 5d5d5588ebc1..ea5d5c228534 100644
--- a/drivers/gpu/drm/omapdrm/dss/hdmi4_core.c
+++ b/drivers/gpu/drm/omapdrm/dss/hdmi4_core.c
@@ -542,8 +542,9 @@ static void hdmi_core_audio_config(struct hdmi_core_data *core,
 	}
 
 	/* Set ACR clock divisor */
-	REG_FLD_MOD(av_base,
-			HDMI_CORE_AV_FREQ_SVAL, cfg->mclk_mode, 2, 0);
+	if (cfg->use_mclk)
+		REG_FLD_MOD(av_base, HDMI_CORE_AV_FREQ_SVAL,
+			    cfg->mclk_mode, 2, 0);
 
 	r = hdmi_read_reg(av_base, HDMI_CORE_AV_ACR_CTRL);
 	/*
@@ -675,7 +676,7 @@ int hdmi4_audio_config(struct hdmi_core_data *core, struct hdmi_wp_data *wp,
 	struct hdmi_audio_format audio_format;
 	struct hdmi_audio_dma audio_dma;
 	struct hdmi_core_audio_config acore;
-	int err, n, cts, channel_count;
+	int n, cts, channel_count;
 	unsigned int fs_nr;
 	bool word_length_16b = false;
 
@@ -737,7 +738,7 @@ int hdmi4_audio_config(struct hdmi_core_data *core, struct hdmi_wp_data *wp,
 		return -EINVAL;
 	}
 
-	err = hdmi_compute_acr(pclk, fs_nr, &n, &cts);
+	hdmi_compute_acr(pclk, fs_nr, &n, &cts);
 
 	/* Audio clock regeneration settings */
 	acore.n = n;
diff --git a/drivers/gpu/drm/omapdrm/dss/hdmi5_core.c b/drivers/gpu/drm/omapdrm/dss/hdmi5_core.c
index 7400fb99d453..ff4d35c8771f 100644
--- a/drivers/gpu/drm/omapdrm/dss/hdmi5_core.c
+++ b/drivers/gpu/drm/omapdrm/dss/hdmi5_core.c
@@ -23,24 +23,12 @@
 
 #include "hdmi5_core.h"
 
-/* only 24 bit color depth used for now */
-static const struct csc_table csc_table_deepcolor[] = {
-	/* HDMI_DEEP_COLOR_24BIT */
-	[0] = { 7036, 0, 0, 32, 0, 7036, 0, 32, 0, 0, 7036, 32, },
-	/* HDMI_DEEP_COLOR_30BIT */
-	[1] = { 7015, 0, 0, 128, 0, 7015, 0, 128, 0, 0, 7015, 128, },
-	/* HDMI_DEEP_COLOR_36BIT */
-	[2] = { 7010, 0, 0, 512, 0, 7010, 0, 512, 0, 0, 7010, 512, },
-	/* FULL RANGE */
-	[3] = { 8192, 0, 0, 0, 0, 8192, 0, 0, 0, 0, 8192, 0, },
-};
-
 static void hdmi_core_ddc_init(struct hdmi_core_data *core)
 {
 	void __iomem *base = core->base;
 	const unsigned long long iclk = 266000000;	/* DSS L3 ICLK */
-	const unsigned int ss_scl_high = 4600;		/* ns */
-	const unsigned int ss_scl_low = 5400;		/* ns */
+	const unsigned int ss_scl_high = 4700;		/* ns */
+	const unsigned int ss_scl_low = 5500;		/* ns */
 	const unsigned int fs_scl_high = 600;		/* ns */
 	const unsigned int fs_scl_low = 1300;		/* ns */
 	const unsigned int sda_hold = 1000;		/* ns */
@@ -397,14 +385,6 @@ static void hdmi_core_config_video_packetizer(struct hdmi_core_data *core)
 	REG_FLD_MOD(base, HDMI_CORE_VP_CONF, clr_depth ? 0 : 2, 1, 0);
 }
 
-static void hdmi_core_config_csc(struct hdmi_core_data *core)
-{
-	int clr_depth = 0;	/* 24 bit color depth */
-
-	/* CSC_COLORDEPTH */
-	REG_FLD_MOD(core->base, HDMI_CORE_CSC_SCALE, clr_depth, 7, 4);
-}
-
 static void hdmi_core_config_video_sampler(struct hdmi_core_data *core)
 {
 	int video_mapping = 1;	/* for 24 bit color depth */
@@ -469,47 +449,67 @@ static void hdmi_core_write_avi_infoframe(struct hdmi_core_data *core,
 	REG_FLD_MOD(base, HDMI_CORE_FC_PRCONF, pr, 3, 0);
 }
 
-static void hdmi_core_csc_config(struct hdmi_core_data *core,
-		struct csc_table csc_coeff)
+static void hdmi_core_write_csc(struct hdmi_core_data *core,
+		const struct csc_table *csc_coeff)
 {
 	void __iomem *base = core->base;
 
-	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_A1_MSB, csc_coeff.a1 >> 8 , 6, 0);
-	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_A1_LSB, csc_coeff.a1, 7, 0);
-	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_A2_MSB, csc_coeff.a2 >> 8, 6, 0);
-	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_A2_LSB, csc_coeff.a2, 7, 0);
-	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_A3_MSB, csc_coeff.a3 >> 8, 6, 0);
-	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_A3_LSB, csc_coeff.a3, 7, 0);
-	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_A4_MSB, csc_coeff.a4 >> 8, 6, 0);
-	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_A4_LSB, csc_coeff.a4, 7, 0);
-	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_B1_MSB, csc_coeff.b1 >> 8, 6, 0);
-	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_B1_LSB, csc_coeff.b1, 7, 0);
-	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_B2_MSB, csc_coeff.b2 >> 8, 6, 0);
-	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_B2_LSB, csc_coeff.b2, 7, 0);
-	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_B3_MSB, csc_coeff.b3 >> 8, 6, 0);
-	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_B3_LSB, csc_coeff.b3, 7, 0);
-	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_B4_MSB, csc_coeff.b4 >> 8, 6, 0);
-	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_B4_LSB, csc_coeff.b4, 7, 0);
-	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_C1_MSB, csc_coeff.c1 >> 8, 6, 0);
-	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_C1_LSB, csc_coeff.c1, 7, 0);
-	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_C2_MSB, csc_coeff.c2 >> 8, 6, 0);
-	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_C2_LSB, csc_coeff.c2, 7, 0);
-	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_C3_MSB, csc_coeff.c3 >> 8, 6, 0);
-	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_C3_LSB, csc_coeff.c3, 7, 0);
-	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_C4_MSB, csc_coeff.c4 >> 8, 6, 0);
-	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_C4_LSB, csc_coeff.c4, 7, 0);
-
+	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_A1_MSB, csc_coeff->a1 >> 8, 6, 0);
+	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_A1_LSB, csc_coeff->a1, 7, 0);
+	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_A2_MSB, csc_coeff->a2 >> 8, 6, 0);
+	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_A2_LSB, csc_coeff->a2, 7, 0);
+	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_A3_MSB, csc_coeff->a3 >> 8, 6, 0);
+	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_A3_LSB, csc_coeff->a3, 7, 0);
+	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_A4_MSB, csc_coeff->a4 >> 8, 6, 0);
+	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_A4_LSB, csc_coeff->a4, 7, 0);
+	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_B1_MSB, csc_coeff->b1 >> 8, 6, 0);
+	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_B1_LSB, csc_coeff->b1, 7, 0);
+	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_B2_MSB, csc_coeff->b2 >> 8, 6, 0);
+	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_B2_LSB, csc_coeff->b2, 7, 0);
+	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_B3_MSB, csc_coeff->b3 >> 8, 6, 0);
+	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_B3_LSB, csc_coeff->b3, 7, 0);
+	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_B4_MSB, csc_coeff->b4 >> 8, 6, 0);
+	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_B4_LSB, csc_coeff->b4, 7, 0);
+	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_C1_MSB, csc_coeff->c1 >> 8, 6, 0);
+	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_C1_LSB, csc_coeff->c1, 7, 0);
+	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_C2_MSB, csc_coeff->c2 >> 8, 6, 0);
+	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_C2_LSB, csc_coeff->c2, 7, 0);
+	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_C3_MSB, csc_coeff->c3 >> 8, 6, 0);
+	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_C3_LSB, csc_coeff->c3, 7, 0);
+	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_C4_MSB, csc_coeff->c4 >> 8, 6, 0);
+	REG_FLD_MOD(base, HDMI_CORE_CSC_COEF_C4_LSB, csc_coeff->c4, 7, 0);
+
+	/* enable CSC */
 	REG_FLD_MOD(base, HDMI_CORE_MC_FLOWCTRL, 0x1, 0, 0);
 }
 
-static void hdmi_core_configure_range(struct hdmi_core_data *core)
+static void hdmi_core_configure_range(struct hdmi_core_data *core,
+				      enum hdmi_quantization_range range)
 {
-	struct csc_table csc_coeff = { 0 };
+	static const struct csc_table csc_limited_range = {
+		7036, 0, 0, 32, 0, 7036, 0, 32, 0, 0, 7036, 32
+	};
+	static const struct csc_table csc_full_range = {
+		8192, 0, 0, 0, 0, 8192, 0, 0, 0, 0, 8192, 0
+	};
+	const struct csc_table *csc_coeff;
+
+	/* CSC_COLORDEPTH  = 24 bits*/
+	REG_FLD_MOD(core->base, HDMI_CORE_CSC_SCALE, 0, 7, 4);
+
+	switch (range) {
+	case HDMI_QUANTIZATION_RANGE_FULL:
+		csc_coeff = &csc_full_range;
+		break;
 
-	/* support limited range with 24 bit color depth for now */
-	csc_coeff = csc_table_deepcolor[0];
+	case HDMI_QUANTIZATION_RANGE_DEFAULT:
+	case HDMI_QUANTIZATION_RANGE_LIMITED:
+	default:
+		csc_coeff = &csc_limited_range;
+		break;
+	}
 
-	hdmi_core_csc_config(core, csc_coeff);
+	hdmi_core_write_csc(core, csc_coeff);
 }
 
 static void hdmi_core_enable_video_path(struct hdmi_core_data *core)
@@ -600,9 +600,20 @@ void hdmi5_configure(struct hdmi_core_data *core, struct hdmi_wp_data *wp,
 	struct videomode vm;
 	struct hdmi_video_format video_format;
 	struct hdmi_core_vid_config v_core_cfg;
+	enum hdmi_quantization_range range;
 
 	hdmi_core_mask_interrupts(core);
 
+	if (cfg->hdmi_dvi_mode == HDMI_HDMI) {
+		char vic = cfg->infoframe.video_code;
+
+		/* All CEA modes other than VIC 1 use limited quantization range. */
+		range = vic > 1 ? HDMI_QUANTIZATION_RANGE_LIMITED :
+			HDMI_QUANTIZATION_RANGE_FULL;
+	} else {
+		range = HDMI_QUANTIZATION_RANGE_FULL;
+	}
+
 	hdmi_core_init(&v_core_cfg, cfg);
 
 	hdmi_wp_init_vid_fmt_timings(&video_format, &vm, cfg);
@@ -616,9 +627,8 @@ void hdmi5_configure(struct hdmi_core_data *core, struct hdmi_wp_data *wp,
 
 	hdmi_wp_video_config_interface(wp, &vm);
 
-	/* support limited range with 24 bit color depth for now */
-	hdmi_core_configure_range(core);
-	cfg->infoframe.quantization_range = HDMI_QUANTIZATION_RANGE_LIMITED;
+	hdmi_core_configure_range(core, range);
+	cfg->infoframe.quantization_range = range;
 
 	/*
 	 * configure core video part, set software reset in the core
@@ -628,7 +638,6 @@ void hdmi5_configure(struct hdmi_core_data *core, struct hdmi_wp_data *wp,
 	hdmi_core_video_config(core, &v_core_cfg);
 
 	hdmi_core_config_video_packetizer(core);
-	hdmi_core_config_csc(core);
 	hdmi_core_config_video_sampler(core);
 
 	if (cfg->hdmi_dvi_mode == HDMI_HDMI)
@@ -798,7 +807,7 @@ int hdmi5_audio_config(struct hdmi_core_data *core, struct hdmi_wp_data *wp,
 	struct hdmi_audio_format audio_format;
 	struct hdmi_audio_dma audio_dma;
 	struct hdmi_core_audio_config core_cfg;
-	int err, n, cts, channel_count;
+	int n, cts, channel_count;
 	unsigned int fs_nr;
 	bool word_length_16b = false;
 
@@ -841,7 +850,7 @@ int hdmi5_audio_config(struct hdmi_core_data *core, struct hdmi_wp_data *wp,
 		return -EINVAL;
 	}
 
-	err = hdmi_compute_acr(pclk, fs_nr, &n, &cts);
+	hdmi_compute_acr(pclk, fs_nr, &n, &cts);
 	core_cfg.n = n;
 	core_cfg.cts = cts;
 
diff --git a/drivers/gpu/drm/omapdrm/dss/output.c b/drivers/gpu/drm/omapdrm/dss/output.c
index 14b41de44ebc..0693d34fca1b 100644
--- a/drivers/gpu/drm/omapdrm/dss/output.c
+++ b/drivers/gpu/drm/omapdrm/dss/output.c
@@ -12,6 +12,7 @@
 #include <linux/of.h>
 #include <linux/of_graph.h>
 
+#include <drm/drm_bridge.h>
 #include <drm/drm_panel.h>
 
 #include "dss.h"
diff --git a/drivers/gpu/drm/omapdrm/omap_dmm_tiler.h b/drivers/gpu/drm/omapdrm/omap_dmm_tiler.h
index 835e6654fa82..43c1d096b021 100644
--- a/drivers/gpu/drm/omapdrm/omap_dmm_tiler.h
+++ b/drivers/gpu/drm/omapdrm/omap_dmm_tiler.h
@@ -113,7 +113,7 @@ extern struct platform_driver omap_dmm_driver;
 /* GEM bo flags -> tiler fmt */
 static inline enum tiler_fmt gem2fmt(u32 flags)
 {
-	switch (flags & OMAP_BO_TILED) {
+	switch (flags & OMAP_BO_TILED_MASK) {
 	case OMAP_BO_TILED_8:
 		return TILFMT_8BIT;
 	case OMAP_BO_TILED_16:
diff --git a/drivers/gpu/drm/omapdrm/omap_drv.c b/drivers/gpu/drm/omapdrm/omap_drv.c
index 2983c003698e..b3e22c890c51 100644
--- a/drivers/gpu/drm/omapdrm/omap_drv.c
+++ b/drivers/gpu/drm/omapdrm/omap_drv.c
@@ -11,6 +11,7 @@
 
 #include <drm/drm_atomic.h>
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_drv.h>
 #include <drm/drm_fb_helper.h>
 #include <drm/drm_file.h>
diff --git a/drivers/gpu/drm/omapdrm/omap_encoder.c b/drivers/gpu/drm/omapdrm/omap_encoder.c
index 6fe14111cd95..24bbe9f2a32e 100644
--- a/drivers/gpu/drm/omapdrm/omap_encoder.c
+++ b/drivers/gpu/drm/omapdrm/omap_encoder.c
@@ -6,6 +6,7 @@
 
 #include <linux/list.h>
 
+#include <drm/drm_bridge.h>
 #include <drm/drm_crtc.h>
 #include <drm/drm_modeset_helper_vtables.h>
 #include <drm/drm_edid.h>
diff --git a/drivers/gpu/drm/omapdrm/omap_fb.c b/drivers/gpu/drm/omapdrm/omap_fb.c
index 1b8b5108caf8..9aeab81dfb90 100644
--- a/drivers/gpu/drm/omapdrm/omap_fb.c
+++ b/drivers/gpu/drm/omapdrm/omap_fb.c
@@ -95,7 +95,7 @@ static u32 get_linear_addr(struct drm_framebuffer *fb,
 
 bool omap_framebuffer_supports_rotation(struct drm_framebuffer *fb)
 {
-	return omap_gem_flags(fb->obj[0]) & OMAP_BO_TILED;
+	return omap_gem_flags(fb->obj[0]) & OMAP_BO_TILED_MASK;
 }
 
 /* Note: DRM rotates counter-clockwise, TILER & DSS rotates clockwise */
@@ -135,7 +135,6 @@ void omap_framebuffer_update_scanout(struct drm_framebuffer *fb,
 {
 	struct omap_framebuffer *omap_fb = to_omap_framebuffer(fb);
 	const struct drm_format_info *format = omap_fb->format;
-	struct plane *plane = &omap_fb->planes[0];
 	u32 x, y, orient = 0;
 
 	info->fourcc = fb->format->format;
@@ -154,7 +153,7 @@ void omap_framebuffer_update_scanout(struct drm_framebuffer *fb,
 	x = state->src_x >> 16;
 	y = state->src_y >> 16;
 
-	if (omap_gem_flags(fb->obj[0]) & OMAP_BO_TILED) {
+	if (omap_gem_flags(fb->obj[0]) & OMAP_BO_TILED_MASK) {
 		u32 w = state->src_w >> 16;
 		u32 h = state->src_h >> 16;
 
@@ -209,10 +208,8 @@ void omap_framebuffer_update_scanout(struct drm_framebuffer *fb,
 	info->screen_width /= format->cpp[0];
 
 	if (fb->format->format == DRM_FORMAT_NV12) {
-		plane = &omap_fb->planes[1];
-
 		if (info->rotation_type == OMAP_DSS_ROT_TILER) {
-			WARN_ON(!(omap_gem_flags(fb->obj[1]) & OMAP_BO_TILED));
+			WARN_ON(!(omap_gem_flags(fb->obj[1]) & OMAP_BO_TILED_MASK));
 			omap_gem_rotated_dma_addr(fb->obj[1], orient, x/2, y/2,
 						  &info->p_uv_addr);
 		} else {
diff --git a/drivers/gpu/drm/omapdrm/omap_gem.c b/drivers/gpu/drm/omapdrm/omap_gem.c
index 08f539efddfb..e518d93ca6df 100644
--- a/drivers/gpu/drm/omapdrm/omap_gem.c
+++ b/drivers/gpu/drm/omapdrm/omap_gem.c
@@ -67,7 +67,7 @@ struct omap_gem_object {
 	/**
 	 * # of users of dma_addr
 	 */
-	u32 dma_addr_cnt;
+	refcount_t dma_addr_cnt;
 
 	/**
 	 * If the buffer has been imported from a dmabuf the OMAP_DB_DMABUF flag
@@ -196,7 +196,7 @@ static void omap_gem_evict(struct drm_gem_object *obj)
 	struct omap_gem_object *omap_obj = to_omap_bo(obj);
 	struct omap_drm_private *priv = obj->dev->dev_private;
 
-	if (omap_obj->flags & OMAP_BO_TILED) {
+	if (omap_obj->flags & OMAP_BO_TILED_MASK) {
 		enum tiler_fmt fmt = gem2fmt(omap_obj->flags);
 		int i;
 
@@ -324,7 +324,7 @@ size_t omap_gem_mmap_size(struct drm_gem_object *obj)
 	struct omap_gem_object *omap_obj = to_omap_bo(obj);
 	size_t size = obj->size;
 
-	if (omap_obj->flags & OMAP_BO_TILED) {
+	if (omap_obj->flags & OMAP_BO_TILED_MASK) {
 		/* for tiled buffers, the virtual size has stride rounded up
 		 * to 4kb.. (to hide the fact that row n+1 might start 16kb or
 		 * 32kb later!).  But we don't back the entire buffer with
@@ -513,7 +513,7 @@ vm_fault_t omap_gem_fault(struct vm_fault *vmf)
 	 * probably trigger put_pages()?
 	 */
 
-	if (omap_obj->flags & OMAP_BO_TILED)
+	if (omap_obj->flags & OMAP_BO_TILED_MASK)
 		ret = omap_gem_fault_2d(obj, vma, vmf);
 	else
 		ret = omap_gem_fault_1d(obj, vma, vmf);
@@ -773,18 +773,20 @@ int omap_gem_pin(struct drm_gem_object *obj, dma_addr_t *dma_addr)
 	mutex_lock(&omap_obj->lock);
 
 	if (!omap_gem_is_contiguous(omap_obj) && priv->has_dmm) {
-		if (omap_obj->dma_addr_cnt == 0) {
+		if (refcount_read(&omap_obj->dma_addr_cnt) == 0) {
 			u32 npages = obj->size >> PAGE_SHIFT;
 			enum tiler_fmt fmt = gem2fmt(omap_obj->flags);
 			struct tiler_block *block;
 
 			BUG_ON(omap_obj->block);
 
+			refcount_set(&omap_obj->dma_addr_cnt, 1);
+
 			ret = omap_gem_attach_pages(obj);
 			if (ret)
 				goto fail;
 
-			if (omap_obj->flags & OMAP_BO_TILED) {
+			if (omap_obj->flags & OMAP_BO_TILED_MASK) {
 				block = tiler_reserve_2d(fmt,
 						omap_obj->width,
 						omap_obj->height, 0);
@@ -813,13 +815,15 @@ int omap_gem_pin(struct drm_gem_object *obj, dma_addr_t *dma_addr)
 			omap_obj->block = block;
 
 			DBG("got dma address: %pad", &omap_obj->dma_addr);
+		} else {
+			refcount_inc(&omap_obj->dma_addr_cnt);
 		}
 
-		omap_obj->dma_addr_cnt++;
-
-		*dma_addr = omap_obj->dma_addr;
+		if (dma_addr)
+			*dma_addr = omap_obj->dma_addr;
 	} else if (omap_gem_is_contiguous(omap_obj)) {
-		*dma_addr = omap_obj->dma_addr;
+		if (dma_addr)
+			*dma_addr = omap_obj->dma_addr;
 	} else {
 		ret = -EINVAL;
 		goto fail;
@@ -832,38 +836,46 @@ fail:
 }
 
 /**
+ * omap_gem_unpin_locked() - Unpin a GEM object from memory
+ * @obj: the GEM object
+ *
+ * omap_gem_unpin() without locking.
+ */
+static void omap_gem_unpin_locked(struct drm_gem_object *obj)
+{
+	struct omap_gem_object *omap_obj = to_omap_bo(obj);
+	int ret;
+
+	if (refcount_dec_and_test(&omap_obj->dma_addr_cnt)) {
+		ret = tiler_unpin(omap_obj->block);
+		if (ret) {
+			dev_err(obj->dev->dev,
+				"could not unpin pages: %d\n", ret);
+		}
+		ret = tiler_release(omap_obj->block);
+		if (ret) {
+			dev_err(obj->dev->dev,
+				"could not release unmap: %d\n", ret);
+		}
+		omap_obj->dma_addr = 0;
+		omap_obj->block = NULL;
+	}
+}
+
+/**
  * omap_gem_unpin() - Unpin a GEM object from memory
  * @obj: the GEM object
  *
  * Unpin the given GEM object previously pinned with omap_gem_pin(). Pins are
- * reference-counted, the actualy unpin will only be performed when the number
+ * reference-counted, the actual unpin will only be performed when the number
  * of calls to this function matches the number of calls to omap_gem_pin().
  */
 void omap_gem_unpin(struct drm_gem_object *obj)
 {
 	struct omap_gem_object *omap_obj = to_omap_bo(obj);
-	int ret;
 
 	mutex_lock(&omap_obj->lock);
-
-	if (omap_obj->dma_addr_cnt > 0) {
-		omap_obj->dma_addr_cnt--;
-		if (omap_obj->dma_addr_cnt == 0) {
-			ret = tiler_unpin(omap_obj->block);
-			if (ret) {
-				dev_err(obj->dev->dev,
-					"could not unpin pages: %d\n", ret);
-			}
-			ret = tiler_release(omap_obj->block);
-			if (ret) {
-				dev_err(obj->dev->dev,
-					"could not release unmap: %d\n", ret);
-			}
-			omap_obj->dma_addr = 0;
-			omap_obj->block = NULL;
-		}
-	}
-
+	omap_gem_unpin_locked(obj);
 	mutex_unlock(&omap_obj->lock);
 }
 
@@ -879,8 +891,8 @@ int omap_gem_rotated_dma_addr(struct drm_gem_object *obj, u32 orient,
 
 	mutex_lock(&omap_obj->lock);
 
-	if ((omap_obj->dma_addr_cnt > 0) && omap_obj->block &&
-			(omap_obj->flags & OMAP_BO_TILED)) {
+	if ((refcount_read(&omap_obj->dma_addr_cnt) > 0) && omap_obj->block &&
+			(omap_obj->flags & OMAP_BO_TILED_MASK)) {
 		*dma_addr = tiler_tsptr(omap_obj->block, orient, x, y);
 		ret = 0;
 	}
@@ -895,7 +907,7 @@ int omap_gem_tiled_stride(struct drm_gem_object *obj, u32 orient)
 {
 	struct omap_gem_object *omap_obj = to_omap_bo(obj);
 	int ret = -EINVAL;
-	if (omap_obj->flags & OMAP_BO_TILED)
+	if (omap_obj->flags & OMAP_BO_TILED_MASK)
 		ret = tiler_stride(gem2fmt(omap_obj->flags), orient);
 	return ret;
 }
@@ -1030,10 +1042,11 @@ void omap_gem_describe(struct drm_gem_object *obj, struct seq_file *m)
 
 	seq_printf(m, "%08x: %2d (%2d) %08llx %pad (%2d) %p %4d",
 			omap_obj->flags, obj->name, kref_read(&obj->refcount),
-			off, &omap_obj->dma_addr, omap_obj->dma_addr_cnt,
+			off, &omap_obj->dma_addr,
+			refcount_read(&omap_obj->dma_addr_cnt),
 			omap_obj->vaddr, omap_obj->roll);
 
-	if (omap_obj->flags & OMAP_BO_TILED) {
+	if (omap_obj->flags & OMAP_BO_TILED_MASK) {
 		seq_printf(m, " %dx%d", omap_obj->width, omap_obj->height);
 		if (omap_obj->block) {
 			struct tcm_area *area = &omap_obj->block->area;
@@ -1093,7 +1106,7 @@ void omap_gem_free_object(struct drm_gem_object *obj)
 	mutex_lock(&omap_obj->lock);
 
 	/* The object should not be pinned. */
-	WARN_ON(omap_obj->dma_addr_cnt > 0);
+	WARN_ON(refcount_read(&omap_obj->dma_addr_cnt) > 0);
 
 	if (omap_obj->pages) {
 		if (omap_obj->flags & OMAP_BO_MEM_DMABUF)
@@ -1120,6 +1133,38 @@ void omap_gem_free_object(struct drm_gem_object *obj)
 	kfree(omap_obj);
 }
 
+static bool omap_gem_validate_flags(struct drm_device *dev, u32 flags)
+{
+	struct omap_drm_private *priv = dev->dev_private;
+
+	switch (flags & OMAP_BO_CACHE_MASK) {
+	case OMAP_BO_CACHED:
+	case OMAP_BO_WC:
+	case OMAP_BO_CACHE_MASK:
+		break;
+
+	default:
+		return false;
+	}
+
+	if (flags & OMAP_BO_TILED_MASK) {
+		if (!priv->usergart)
+			return false;
+
+		switch (flags & OMAP_BO_TILED_MASK) {
+		case OMAP_BO_TILED_8:
+		case OMAP_BO_TILED_16:
+		case OMAP_BO_TILED_32:
+			break;
+
+		default:
+			return false;
+		}
+	}
+
+	return true;
+}
+
 /* GEM buffer object constructor */
 struct drm_gem_object *omap_gem_new(struct drm_device *dev,
 		union omap_gem_size gsize, u32 flags)
@@ -1131,18 +1176,15 @@ struct drm_gem_object *omap_gem_new(struct drm_device *dev,
 	size_t size;
 	int ret;
 
-	/* Validate the flags and compute the memory and cache flags. */
-	if (flags & OMAP_BO_TILED) {
-		if (!priv->usergart) {
-			dev_err(dev->dev, "Tiled buffers require DMM\n");
-			return NULL;
-		}
+	if (!omap_gem_validate_flags(dev, flags))
+		return NULL;
 
+	/* Validate the flags and compute the memory and cache flags. */
+	if (flags & OMAP_BO_TILED_MASK) {
 		/*
 		 * Tiled buffers are always shmem paged backed. When they are
 		 * scanned out, they are remapped into DMM/TILER.
 		 */
-		flags &= ~OMAP_BO_SCANOUT;
 		flags |= OMAP_BO_MEM_SHMEM;
 
 		/*
@@ -1153,9 +1195,8 @@ struct drm_gem_object *omap_gem_new(struct drm_device *dev,
 		flags |= tiler_get_cpu_cache_flags();
 	} else if ((flags & OMAP_BO_SCANOUT) && !priv->has_dmm) {
 		/*
-		 * OMAP_BO_SCANOUT hints that the buffer doesn't need to be
-		 * tiled. However, to lower the pressure on memory allocation,
-		 * use contiguous memory only if no TILER is available.
+		 * If we don't have DMM, we must allocate scanout buffers
+		 * from contiguous DMA memory.
 		 */
 		flags |= OMAP_BO_MEM_DMA_API;
 	} else if (!(flags & OMAP_BO_MEM_DMABUF)) {
@@ -1174,7 +1215,7 @@ struct drm_gem_object *omap_gem_new(struct drm_device *dev,
 	omap_obj->flags = flags;
 	mutex_init(&omap_obj->lock);
 
-	if (flags & OMAP_BO_TILED) {
+	if (flags & OMAP_BO_TILED_MASK) {
 		/*
 		 * For tiled buffers align dimensions to slot boundaries and
 		 * calculate size based on aligned dimensions.
diff --git a/drivers/gpu/drm/omapdrm/omap_gem_dmabuf.c b/drivers/gpu/drm/omapdrm/omap_gem_dmabuf.c
index e8c3ae7ac77e..7344bb61936c 100644
--- a/drivers/gpu/drm/omapdrm/omap_gem_dmabuf.c
+++ b/drivers/gpu/drm/omapdrm/omap_gem_dmabuf.c
@@ -67,7 +67,7 @@ static int omap_gem_dmabuf_begin_cpu_access(struct dma_buf *buffer,
 {
 	struct drm_gem_object *obj = buffer->priv;
 	struct page **pages;
-	if (omap_gem_flags(obj) & OMAP_BO_TILED) {
+	if (omap_gem_flags(obj) & OMAP_BO_TILED_MASK) {
 		/* TODO we would need to pin at least part of the buffer to
 		 * get de-tiled view.  For now just reject it.
 		 */
diff --git a/drivers/gpu/drm/panel/panel-arm-versatile.c b/drivers/gpu/drm/panel/panel-arm-versatile.c
index 5f72c922a04b..a0574dc03e16 100644
--- a/drivers/gpu/drm/panel/panel-arm-versatile.c
+++ b/drivers/gpu/drm/panel/panel-arm-versatile.c
@@ -350,9 +350,8 @@ static int versatile_panel_probe(struct platform_device *pdev)
 			dev_info(dev, "panel mounted on IB2 daughterboard\n");
 	}
 
-	drm_panel_init(&vpanel->panel);
-	vpanel->panel.dev = dev;
-	vpanel->panel.funcs = &versatile_panel_drm_funcs;
+	drm_panel_init(&vpanel->panel, dev, &versatile_panel_drm_funcs,
+		       DRM_MODE_CONNECTOR_DPI);
 
 	return drm_panel_add(&vpanel->panel);
 }
diff --git a/drivers/gpu/drm/panel/panel-feiyang-fy07024di26a30d.c b/drivers/gpu/drm/panel/panel-feiyang-fy07024di26a30d.c
index dabf59e0f56f..98f184b81187 100644
--- a/drivers/gpu/drm/panel/panel-feiyang-fy07024di26a30d.c
+++ b/drivers/gpu/drm/panel/panel-feiyang-fy07024di26a30d.c
@@ -204,9 +204,8 @@ static int feiyang_dsi_probe(struct mipi_dsi_device *dsi)
 	mipi_dsi_set_drvdata(dsi, ctx);
 	ctx->dsi = dsi;
 
-	drm_panel_init(&ctx->panel);
-	ctx->panel.dev = &dsi->dev;
-	ctx->panel.funcs = &feiyang_funcs;
+	drm_panel_init(&ctx->panel, &dsi->dev, &feiyang_funcs,
+		       DRM_MODE_CONNECTOR_DSI);
 
 	ctx->dvdd = devm_regulator_get(&dsi->dev, "dvdd");
 	if (IS_ERR(ctx->dvdd)) {
diff --git a/drivers/gpu/drm/panel/panel-ilitek-ili9322.c b/drivers/gpu/drm/panel/panel-ilitek-ili9322.c
index 3c58f63adbf7..24955bec1958 100644
--- a/drivers/gpu/drm/panel/panel-ilitek-ili9322.c
+++ b/drivers/gpu/drm/panel/panel-ilitek-ili9322.c
@@ -895,9 +895,8 @@ static int ili9322_probe(struct spi_device *spi)
 		ili->input = ili->conf->input;
 	}
 
-	drm_panel_init(&ili->panel);
-	ili->panel.dev = dev;
-	ili->panel.funcs = &ili9322_drm_funcs;
+	drm_panel_init(&ili->panel, dev, &ili9322_drm_funcs,
+		       DRM_MODE_CONNECTOR_DPI);
 
 	return drm_panel_add(&ili->panel);
 }
diff --git a/drivers/gpu/drm/panel/panel-ilitek-ili9881c.c b/drivers/gpu/drm/panel/panel-ilitek-ili9881c.c
index 3ad4a46c4e94..e8789e460a16 100644
--- a/drivers/gpu/drm/panel/panel-ilitek-ili9881c.c
+++ b/drivers/gpu/drm/panel/panel-ilitek-ili9881c.c
@@ -433,9 +433,8 @@ static int ili9881c_dsi_probe(struct mipi_dsi_device *dsi)
 	mipi_dsi_set_drvdata(dsi, ctx);
 	ctx->dsi = dsi;
 
-	drm_panel_init(&ctx->panel);
-	ctx->panel.dev = &dsi->dev;
-	ctx->panel.funcs = &ili9881c_funcs;
+	drm_panel_init(&ctx->panel, &dsi->dev, &ili9881c_funcs,
+		       DRM_MODE_CONNECTOR_DSI);
 
 	ctx->power = devm_regulator_get(&dsi->dev, "power");
 	if (IS_ERR(ctx->power)) {
diff --git a/drivers/gpu/drm/panel/panel-innolux-p079zca.c b/drivers/gpu/drm/panel/panel-innolux-p079zca.c
index d92d1c98878c..83df1ac4211f 100644
--- a/drivers/gpu/drm/panel/panel-innolux-p079zca.c
+++ b/drivers/gpu/drm/panel/panel-innolux-p079zca.c
@@ -487,9 +487,8 @@ static int innolux_panel_add(struct mipi_dsi_device *dsi,
 	if (IS_ERR(innolux->backlight))
 		return PTR_ERR(innolux->backlight);
 
-	drm_panel_init(&innolux->base);
-	innolux->base.funcs = &innolux_panel_funcs;
-	innolux->base.dev = dev;
+	drm_panel_init(&innolux->base, dev, &innolux_panel_funcs,
+		       DRM_MODE_CONNECTOR_DSI);
 
 	err = drm_panel_add(&innolux->base);
 	if (err < 0)
diff --git a/drivers/gpu/drm/panel/panel-jdi-lt070me05000.c b/drivers/gpu/drm/panel/panel-jdi-lt070me05000.c
index ff3e89e61e3f..56364a93f0b8 100644
--- a/drivers/gpu/drm/panel/panel-jdi-lt070me05000.c
+++ b/drivers/gpu/drm/panel/panel-jdi-lt070me05000.c
@@ -437,9 +437,8 @@ static int jdi_panel_add(struct jdi_panel *jdi)
 		return ret;
 	}
 
-	drm_panel_init(&jdi->base);
-	jdi->base.funcs = &jdi_panel_funcs;
-	jdi->base.dev = &jdi->dsi->dev;
+	drm_panel_init(&jdi->base, &jdi->dsi->dev, &jdi_panel_funcs,
+		       DRM_MODE_CONNECTOR_DSI);
 
 	ret = drm_panel_add(&jdi->base);
 
diff --git a/drivers/gpu/drm/panel/panel-kingdisplay-kd097d04.c b/drivers/gpu/drm/panel/panel-kingdisplay-kd097d04.c
index 3ac04eb8d0fe..45f96556ec8c 100644
--- a/drivers/gpu/drm/panel/panel-kingdisplay-kd097d04.c
+++ b/drivers/gpu/drm/panel/panel-kingdisplay-kd097d04.c
@@ -391,9 +391,8 @@ static int kingdisplay_panel_add(struct kingdisplay_panel *kingdisplay)
 	if (IS_ERR(kingdisplay->backlight))
 		return PTR_ERR(kingdisplay->backlight);
 
-	drm_panel_init(&kingdisplay->base);
-	kingdisplay->base.funcs = &kingdisplay_panel_funcs;
-	kingdisplay->base.dev = &kingdisplay->link->dev;
+	drm_panel_init(&kingdisplay->base, &kingdisplay->link->dev,
+		       &kingdisplay_panel_funcs, DRM_MODE_CONNECTOR_DSI);
 
 	return drm_panel_add(&kingdisplay->base);
 }
diff --git a/drivers/gpu/drm/panel/panel-lg-lb035q02.c b/drivers/gpu/drm/panel/panel-lg-lb035q02.c
index fc82a525b071..7a1385e834f0 100644
--- a/drivers/gpu/drm/panel/panel-lg-lb035q02.c
+++ b/drivers/gpu/drm/panel/panel-lg-lb035q02.c
@@ -196,9 +196,8 @@ static int lb035q02_probe(struct spi_device *spi)
 	if (ret < 0)
 		return ret;
 
-	drm_panel_init(&lcd->panel);
-	lcd->panel.dev = &lcd->spi->dev;
-	lcd->panel.funcs = &lb035q02_funcs;
+	drm_panel_init(&lcd->panel, &lcd->spi->dev, &lb035q02_funcs,
+		       DRM_MODE_CONNECTOR_DPI);
 
 	return drm_panel_add(&lcd->panel);
 }
@@ -220,9 +219,17 @@ static const struct of_device_id lb035q02_of_match[] = {
 
 MODULE_DEVICE_TABLE(of, lb035q02_of_match);
 
+static const struct spi_device_id lb035q02_ids[] = {
+	{ "lb035q02", 0 },
+	{ /* sentinel */ }
+};
+
+MODULE_DEVICE_TABLE(spi, lb035q02_ids);
+
 static struct spi_driver lb035q02_driver = {
 	.probe		= lb035q02_probe,
 	.remove		= lb035q02_remove,
+	.id_table	= lb035q02_ids,
 	.driver		= {
 		.name	= "panel-lg-lb035q02",
 		.of_match_table = lb035q02_of_match,
@@ -231,7 +238,6 @@ static struct spi_driver lb035q02_driver = {
 
 module_spi_driver(lb035q02_driver);
 
-MODULE_ALIAS("spi:lgphilips,lb035q02");
 MODULE_AUTHOR("Tomi Valkeinen <tomi.valkeinen@ti.com>");
 MODULE_DESCRIPTION("LG.Philips LB035Q02 LCD Panel driver");
 MODULE_LICENSE("GPL");
diff --git a/drivers/gpu/drm/panel/panel-lg-lg4573.c b/drivers/gpu/drm/panel/panel-lg-lg4573.c
index 41bf02d122a1..db4865a4c2b9 100644
--- a/drivers/gpu/drm/panel/panel-lg-lg4573.c
+++ b/drivers/gpu/drm/panel/panel-lg-lg4573.c
@@ -259,9 +259,8 @@ static int lg4573_probe(struct spi_device *spi)
 		return ret;
 	}
 
-	drm_panel_init(&ctx->panel);
-	ctx->panel.dev = &spi->dev;
-	ctx->panel.funcs = &lg4573_drm_funcs;
+	drm_panel_init(&ctx->panel, &spi->dev, &lg4573_drm_funcs,
+		       DRM_MODE_CONNECTOR_DPI);
 
 	return drm_panel_add(&ctx->panel);
 }
diff --git a/drivers/gpu/drm/panel/panel-lvds.c b/drivers/gpu/drm/panel/panel-lvds.c
index ad47cc95459e..2405f26e5d31 100644
--- a/drivers/gpu/drm/panel/panel-lvds.c
+++ b/drivers/gpu/drm/panel/panel-lvds.c
@@ -197,7 +197,6 @@ static int panel_lvds_parse_dt(struct panel_lvds *lvds)
 static int panel_lvds_probe(struct platform_device *pdev)
 {
 	struct panel_lvds *lvds;
-	struct device_node *np;
 	int ret;
 
 	lvds = devm_kzalloc(&pdev->dev, sizeof(*lvds), GFP_KERNEL);
@@ -243,14 +242,9 @@ static int panel_lvds_probe(struct platform_device *pdev)
 		return ret;
 	}
 
-	np = of_parse_phandle(lvds->dev->of_node, "backlight", 0);
-	if (np) {
-		lvds->backlight = of_find_backlight_by_node(np);
-		of_node_put(np);
-
-		if (!lvds->backlight)
-			return -EPROBE_DEFER;
-	}
+	lvds->backlight = devm_of_find_backlight(lvds->dev);
+	if (IS_ERR(lvds->backlight))
+		return PTR_ERR(lvds->backlight);
 
 	/*
 	 * TODO: Handle all power supplies specified in the DT node in a generic
@@ -260,20 +254,15 @@ static int panel_lvds_probe(struct platform_device *pdev)
 	 */
 
 	/* Register the panel. */
-	drm_panel_init(&lvds->panel);
-	lvds->panel.dev = lvds->dev;
-	lvds->panel.funcs = &panel_lvds_funcs;
+	drm_panel_init(&lvds->panel, lvds->dev, &panel_lvds_funcs,
+		       DRM_MODE_CONNECTOR_LVDS);
 
 	ret = drm_panel_add(&lvds->panel);
 	if (ret < 0)
-		goto error;
+		return ret;
 
 	dev_set_drvdata(lvds->dev, lvds);
 	return 0;
-
-error:
-	put_device(&lvds->backlight->dev);
-	return ret;
 }
 
 static int panel_lvds_remove(struct platform_device *pdev)
@@ -284,9 +273,6 @@ static int panel_lvds_remove(struct platform_device *pdev)
 
 	panel_lvds_disable(&lvds->panel);
 
-	if (lvds->backlight)
-		put_device(&lvds->backlight->dev);
-
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/panel/panel-nec-nl8048hl11.c b/drivers/gpu/drm/panel/panel-nec-nl8048hl11.c
index 299b217c83e1..fd593532ab23 100644
--- a/drivers/gpu/drm/panel/panel-nec-nl8048hl11.c
+++ b/drivers/gpu/drm/panel/panel-nec-nl8048hl11.c
@@ -205,9 +205,8 @@ static int nl8048_probe(struct spi_device *spi)
 	if (ret < 0)
 		return ret;
 
-	drm_panel_init(&lcd->panel);
-	lcd->panel.dev = &lcd->spi->dev;
-	lcd->panel.funcs = &nl8048_funcs;
+	drm_panel_init(&lcd->panel, &lcd->spi->dev, &nl8048_funcs,
+		       DRM_MODE_CONNECTOR_DPI);
 
 	return drm_panel_add(&lcd->panel);
 }
@@ -230,9 +229,17 @@ static const struct of_device_id nl8048_of_match[] = {
 
 MODULE_DEVICE_TABLE(of, nl8048_of_match);
 
+static const struct spi_device_id nl8048_ids[] = {
+	{ "nl8048hl11", 0 },
+	{ /* sentinel */ }
+};
+
+MODULE_DEVICE_TABLE(spi, nl8048_ids);
+
 static struct spi_driver nl8048_driver = {
 	.probe		= nl8048_probe,
 	.remove		= nl8048_remove,
+	.id_table	= nl8048_ids,
 	.driver		= {
 		.name	= "panel-nec-nl8048hl11",
 		.pm	= &nl8048_pm_ops,
@@ -242,7 +249,6 @@ static struct spi_driver nl8048_driver = {
 
 module_spi_driver(nl8048_driver);
 
-MODULE_ALIAS("spi:nec,nl8048hl11");
 MODULE_AUTHOR("Erik Gilling <konkers@android.com>");
 MODULE_DESCRIPTION("NEC-NL8048HL11 Driver");
 MODULE_LICENSE("GPL");
diff --git a/drivers/gpu/drm/panel/panel-novatek-nt39016.c b/drivers/gpu/drm/panel/panel-novatek-nt39016.c
index 2ad1063b068d..60ccedce530c 100644
--- a/drivers/gpu/drm/panel/panel-novatek-nt39016.c
+++ b/drivers/gpu/drm/panel/panel-novatek-nt39016.c
@@ -292,9 +292,8 @@ static int nt39016_probe(struct spi_device *spi)
 		return err;
 	}
 
-	drm_panel_init(&panel->drm_panel);
-	panel->drm_panel.dev = dev;
-	panel->drm_panel.funcs = &nt39016_funcs;
+	drm_panel_init(&panel->drm_panel, dev, &nt39016_funcs,
+		       DRM_MODE_CONNECTOR_DPI);
 
 	err = drm_panel_add(&panel->drm_panel);
 	if (err < 0) {
diff --git a/drivers/gpu/drm/panel/panel-olimex-lcd-olinuxino.c b/drivers/gpu/drm/panel/panel-olimex-lcd-olinuxino.c
index 2bae1db3ff34..f2a72ee6ee07 100644
--- a/drivers/gpu/drm/panel/panel-olimex-lcd-olinuxino.c
+++ b/drivers/gpu/drm/panel/panel-olimex-lcd-olinuxino.c
@@ -288,9 +288,8 @@ static int lcd_olinuxino_probe(struct i2c_client *client,
 	if (IS_ERR(lcd->backlight))
 		return PTR_ERR(lcd->backlight);
 
-	drm_panel_init(&lcd->panel);
-	lcd->panel.dev = dev;
-	lcd->panel.funcs = &lcd_olinuxino_funcs;
+	drm_panel_init(&lcd->panel, dev, &lcd_olinuxino_funcs,
+		       DRM_MODE_CONNECTOR_DPI);
 
 	return drm_panel_add(&lcd->panel);
 }
diff --git a/drivers/gpu/drm/panel/panel-orisetech-otm8009a.c b/drivers/gpu/drm/panel/panel-orisetech-otm8009a.c
index c7b48df8869a..bf1f928b215f 100644
--- a/drivers/gpu/drm/panel/panel-orisetech-otm8009a.c
+++ b/drivers/gpu/drm/panel/panel-orisetech-otm8009a.c
@@ -455,9 +455,8 @@ static int otm8009a_probe(struct mipi_dsi_device *dsi)
 	dsi->mode_flags = MIPI_DSI_MODE_VIDEO | MIPI_DSI_MODE_VIDEO_BURST |
 			  MIPI_DSI_MODE_LPM;
 
-	drm_panel_init(&ctx->panel);
-	ctx->panel.dev = dev;
-	ctx->panel.funcs = &otm8009a_drm_funcs;
+	drm_panel_init(&ctx->panel, dev, &otm8009a_drm_funcs,
+		       DRM_MODE_CONNECTOR_DSI);
 
 	ctx->bl_dev = devm_backlight_device_register(dev, dev_name(dev),
 						     dsi->host->dev, ctx,
diff --git a/drivers/gpu/drm/panel/panel-osd-osd101t2587-53ts.c b/drivers/gpu/drm/panel/panel-osd-osd101t2587-53ts.c
index e0e20ecff916..2b40913899d8 100644
--- a/drivers/gpu/drm/panel/panel-osd-osd101t2587-53ts.c
+++ b/drivers/gpu/drm/panel/panel-osd-osd101t2587-53ts.c
@@ -166,9 +166,8 @@ static int osd101t2587_panel_add(struct osd101t2587_panel *osd101t2587)
 	if (IS_ERR(osd101t2587->backlight))
 		return PTR_ERR(osd101t2587->backlight);
 
-	drm_panel_init(&osd101t2587->base);
-	osd101t2587->base.funcs = &osd101t2587_panel_funcs;
-	osd101t2587->base.dev = &osd101t2587->dsi->dev;
+	drm_panel_init(&osd101t2587->base, &osd101t2587->dsi->dev,
+		       &osd101t2587_panel_funcs, DRM_MODE_CONNECTOR_DSI);
 
 	return drm_panel_add(&osd101t2587->base);
 }
diff --git a/drivers/gpu/drm/panel/panel-panasonic-vvx10f034n00.c b/drivers/gpu/drm/panel/panel-panasonic-vvx10f034n00.c
index 3dff0b3f73c2..664605071d34 100644
--- a/drivers/gpu/drm/panel/panel-panasonic-vvx10f034n00.c
+++ b/drivers/gpu/drm/panel/panel-panasonic-vvx10f034n00.c
@@ -223,9 +223,8 @@ static int wuxga_nt_panel_add(struct wuxga_nt_panel *wuxga_nt)
 			return -EPROBE_DEFER;
 	}
 
-	drm_panel_init(&wuxga_nt->base);
-	wuxga_nt->base.funcs = &wuxga_nt_panel_funcs;
-	wuxga_nt->base.dev = &wuxga_nt->dsi->dev;
+	drm_panel_init(&wuxga_nt->base, &wuxga_nt->dsi->dev,
+		       &wuxga_nt_panel_funcs, DRM_MODE_CONNECTOR_DSI);
 
 	ret = drm_panel_add(&wuxga_nt->base);
 	if (ret < 0)
diff --git a/drivers/gpu/drm/panel/panel-raspberrypi-touchscreen.c b/drivers/gpu/drm/panel/panel-raspberrypi-touchscreen.c
index b5b14aa059ea..09824e92fc78 100644
--- a/drivers/gpu/drm/panel/panel-raspberrypi-touchscreen.c
+++ b/drivers/gpu/drm/panel/panel-raspberrypi-touchscreen.c
@@ -426,8 +426,8 @@ static int rpi_touchscreen_probe(struct i2c_client *i2c,
 		return PTR_ERR(ts->dsi);
 	}
 
-	ts->base.dev = dev;
-	ts->base.funcs = &rpi_touchscreen_funcs;
+	drm_panel_init(&ts->base, dev, &rpi_touchscreen_funcs,
+		       DRM_MODE_CONNECTOR_DSI);
 
 	/* This appears last, as it's what will unblock the DSI host
 	 * driver's component bind function.
diff --git a/drivers/gpu/drm/panel/panel-raydium-rm67191.c b/drivers/gpu/drm/panel/panel-raydium-rm67191.c
index 6a5d37006103..fd67fc6185c4 100644
--- a/drivers/gpu/drm/panel/panel-raydium-rm67191.c
+++ b/drivers/gpu/drm/panel/panel-raydium-rm67191.c
@@ -606,9 +606,8 @@ static int rad_panel_probe(struct mipi_dsi_device *dsi)
 	if (ret)
 		return ret;
 
-	drm_panel_init(&panel->panel);
-	panel->panel.funcs = &rad_panel_funcs;
-	panel->panel.dev = dev;
+	drm_panel_init(&panel->panel, dev, &rad_panel_funcs,
+		       DRM_MODE_CONNECTOR_DSI);
 	dev_set_drvdata(dev, panel);
 
 	ret = drm_panel_add(&panel->panel);
diff --git a/drivers/gpu/drm/panel/panel-raydium-rm68200.c b/drivers/gpu/drm/panel/panel-raydium-rm68200.c
index ba889625ad43..994e855721f4 100644
--- a/drivers/gpu/drm/panel/panel-raydium-rm68200.c
+++ b/drivers/gpu/drm/panel/panel-raydium-rm68200.c
@@ -404,9 +404,8 @@ static int rm68200_probe(struct mipi_dsi_device *dsi)
 	dsi->mode_flags = MIPI_DSI_MODE_VIDEO | MIPI_DSI_MODE_VIDEO_BURST |
 			  MIPI_DSI_MODE_LPM;
 
-	drm_panel_init(&ctx->panel);
-	ctx->panel.dev = dev;
-	ctx->panel.funcs = &rm68200_drm_funcs;
+	drm_panel_init(&ctx->panel, dev, &rm68200_drm_funcs,
+		       DRM_MODE_CONNECTOR_DSI);
 
 	drm_panel_add(&ctx->panel);
 
diff --git a/drivers/gpu/drm/panel/panel-rocktech-jh057n00900.c b/drivers/gpu/drm/panel/panel-rocktech-jh057n00900.c
index b9109922397f..31234b79d3b1 100644
--- a/drivers/gpu/drm/panel/panel-rocktech-jh057n00900.c
+++ b/drivers/gpu/drm/panel/panel-rocktech-jh057n00900.c
@@ -343,9 +343,8 @@ static int jh057n_probe(struct mipi_dsi_device *dsi)
 		return ret;
 	}
 
-	drm_panel_init(&ctx->panel);
-	ctx->panel.dev = dev;
-	ctx->panel.funcs = &jh057n_drm_funcs;
+	drm_panel_init(&ctx->panel, dev, &jh057n_drm_funcs,
+		       DRM_MODE_CONNECTOR_DSI);
 
 	drm_panel_add(&ctx->panel);
 
diff --git a/drivers/gpu/drm/panel/panel-ronbo-rb070d30.c b/drivers/gpu/drm/panel/panel-ronbo-rb070d30.c
index 3c15764f0c03..170a5cda21b9 100644
--- a/drivers/gpu/drm/panel/panel-ronbo-rb070d30.c
+++ b/drivers/gpu/drm/panel/panel-ronbo-rb070d30.c
@@ -173,9 +173,8 @@ static int rb070d30_panel_dsi_probe(struct mipi_dsi_device *dsi)
 	mipi_dsi_set_drvdata(dsi, ctx);
 	ctx->dsi = dsi;
 
-	drm_panel_init(&ctx->panel);
-	ctx->panel.dev = &dsi->dev;
-	ctx->panel.funcs = &rb070d30_panel_funcs;
+	drm_panel_init(&ctx->panel, &dsi->dev, &rb070d30_panel_funcs,
+		       DRM_MODE_CONNECTOR_DSI);
 
 	ctx->gpios.reset = devm_gpiod_get(&dsi->dev, "reset", GPIOD_OUT_LOW);
 	if (IS_ERR(ctx->gpios.reset)) {
diff --git a/drivers/gpu/drm/panel/panel-samsung-ld9040.c b/drivers/gpu/drm/panel/panel-samsung-ld9040.c
index 3be902dcedc0..250809ba37c7 100644
--- a/drivers/gpu/drm/panel/panel-samsung-ld9040.c
+++ b/drivers/gpu/drm/panel/panel-samsung-ld9040.c
@@ -351,9 +351,8 @@ static int ld9040_probe(struct spi_device *spi)
 		return ret;
 	}
 
-	drm_panel_init(&ctx->panel);
-	ctx->panel.dev = dev;
-	ctx->panel.funcs = &ld9040_drm_funcs;
+	drm_panel_init(&ctx->panel, dev, &ld9040_drm_funcs,
+		       DRM_MODE_CONNECTOR_DPI);
 
 	return drm_panel_add(&ctx->panel);
 }
diff --git a/drivers/gpu/drm/panel/panel-samsung-s6d16d0.c b/drivers/gpu/drm/panel/panel-samsung-s6d16d0.c
index f75bef24e050..e3a0397e953e 100644
--- a/drivers/gpu/drm/panel/panel-samsung-s6d16d0.c
+++ b/drivers/gpu/drm/panel/panel-samsung-s6d16d0.c
@@ -215,9 +215,8 @@ static int s6d16d0_probe(struct mipi_dsi_device *dsi)
 		return ret;
 	}
 
-	drm_panel_init(&s6->panel);
-	s6->panel.dev = dev;
-	s6->panel.funcs = &s6d16d0_drm_funcs;
+	drm_panel_init(&s6->panel, dev, &s6d16d0_drm_funcs,
+		       DRM_MODE_CONNECTOR_DSI);
 
 	ret = drm_panel_add(&s6->panel);
 	if (ret < 0)
diff --git a/drivers/gpu/drm/panel/panel-samsung-s6e3ha2.c b/drivers/gpu/drm/panel/panel-samsung-s6e3ha2.c
index b923de23ed65..938ab72c5540 100644
--- a/drivers/gpu/drm/panel/panel-samsung-s6e3ha2.c
+++ b/drivers/gpu/drm/panel/panel-samsung-s6e3ha2.c
@@ -732,9 +732,8 @@ static int s6e3ha2_probe(struct mipi_dsi_device *dsi)
 	ctx->bl_dev->props.brightness = S6E3HA2_DEFAULT_BRIGHTNESS;
 	ctx->bl_dev->props.power = FB_BLANK_POWERDOWN;
 
-	drm_panel_init(&ctx->panel);
-	ctx->panel.dev = dev;
-	ctx->panel.funcs = &s6e3ha2_drm_funcs;
+	drm_panel_init(&ctx->panel, dev, &s6e3ha2_drm_funcs,
+		       DRM_MODE_CONNECTOR_DSI);
 
 	ret = drm_panel_add(&ctx->panel);
 	if (ret < 0)
diff --git a/drivers/gpu/drm/panel/panel-samsung-s6e63j0x03.c b/drivers/gpu/drm/panel/panel-samsung-s6e63j0x03.c
index cd90fa700c49..a60635e9226d 100644
--- a/drivers/gpu/drm/panel/panel-samsung-s6e63j0x03.c
+++ b/drivers/gpu/drm/panel/panel-samsung-s6e63j0x03.c
@@ -466,9 +466,8 @@ static int s6e63j0x03_probe(struct mipi_dsi_device *dsi)
 		return PTR_ERR(ctx->reset_gpio);
 	}
 
-	drm_panel_init(&ctx->panel);
-	ctx->panel.dev = dev;
-	ctx->panel.funcs = &s6e63j0x03_funcs;
+	drm_panel_init(&ctx->panel, dev, &s6e63j0x03_funcs,
+		       DRM_MODE_CONNECTOR_DSI);
 
 	ctx->bl_dev = backlight_device_register("s6e63j0x03", dev, ctx,
 						&s6e63j0x03_bl_ops, NULL);
diff --git a/drivers/gpu/drm/panel/panel-samsung-s6e63m0.c b/drivers/gpu/drm/panel/panel-samsung-s6e63m0.c
index 142d395ea512..ba01af0b14fd 100644
--- a/drivers/gpu/drm/panel/panel-samsung-s6e63m0.c
+++ b/drivers/gpu/drm/panel/panel-samsung-s6e63m0.c
@@ -473,9 +473,8 @@ static int s6e63m0_probe(struct spi_device *spi)
 		return ret;
 	}
 
-	drm_panel_init(&ctx->panel);
-	ctx->panel.dev = dev;
-	ctx->panel.funcs = &s6e63m0_drm_funcs;
+	drm_panel_init(&ctx->panel, dev, &s6e63m0_drm_funcs,
+		       DRM_MODE_CONNECTOR_DPI);
 
 	ret = s6e63m0_backlight_register(ctx);
 	if (ret < 0)
diff --git a/drivers/gpu/drm/panel/panel-samsung-s6e8aa0.c b/drivers/gpu/drm/panel/panel-samsung-s6e8aa0.c
index 81858267723a..dbced6501204 100644
--- a/drivers/gpu/drm/panel/panel-samsung-s6e8aa0.c
+++ b/drivers/gpu/drm/panel/panel-samsung-s6e8aa0.c
@@ -1017,9 +1017,8 @@ static int s6e8aa0_probe(struct mipi_dsi_device *dsi)
 
 	ctx->brightness = GAMMA_LEVEL_NUM - 1;
 
-	drm_panel_init(&ctx->panel);
-	ctx->panel.dev = dev;
-	ctx->panel.funcs = &s6e8aa0_drm_funcs;
+	drm_panel_init(&ctx->panel, dev, &s6e8aa0_drm_funcs,
+		       DRM_MODE_CONNECTOR_DSI);
 
 	ret = drm_panel_add(&ctx->panel);
 	if (ret < 0)
diff --git a/drivers/gpu/drm/panel/panel-seiko-43wvf1g.c b/drivers/gpu/drm/panel/panel-seiko-43wvf1g.c
index 18b22b1294fb..b3619ba443bd 100644
--- a/drivers/gpu/drm/panel/panel-seiko-43wvf1g.c
+++ b/drivers/gpu/drm/panel/panel-seiko-43wvf1g.c
@@ -274,9 +274,8 @@ static int seiko_panel_probe(struct device *dev,
 			return -EPROBE_DEFER;
 	}
 
-	drm_panel_init(&panel->base);
-	panel->base.dev = dev;
-	panel->base.funcs = &seiko_panel_funcs;
+	drm_panel_init(&panel->base, dev, &seiko_panel_funcs,
+		       DRM_MODE_CONNECTOR_DPI);
 
 	err = drm_panel_add(&panel->base);
 	if (err < 0)
diff --git a/drivers/gpu/drm/panel/panel-sharp-lq101r1sx01.c b/drivers/gpu/drm/panel/panel-sharp-lq101r1sx01.c
index e910b4ad1310..5e136c3ba185 100644
--- a/drivers/gpu/drm/panel/panel-sharp-lq101r1sx01.c
+++ b/drivers/gpu/drm/panel/panel-sharp-lq101r1sx01.c
@@ -329,9 +329,8 @@ static int sharp_panel_add(struct sharp_panel *sharp)
 	if (IS_ERR(sharp->backlight))
 		return PTR_ERR(sharp->backlight);
 
-	drm_panel_init(&sharp->base);
-	sharp->base.funcs = &sharp_panel_funcs;
-	sharp->base.dev = &sharp->link1->dev;
+	drm_panel_init(&sharp->base, &sharp->link1->dev, &sharp_panel_funcs,
+		       DRM_MODE_CONNECTOR_DSI);
 
 	return drm_panel_add(&sharp->base);
 }
diff --git a/drivers/gpu/drm/panel/panel-sharp-ls037v7dw01.c b/drivers/gpu/drm/panel/panel-sharp-ls037v7dw01.c
index 46cd9a250129..eeab7998c7de 100644
--- a/drivers/gpu/drm/panel/panel-sharp-ls037v7dw01.c
+++ b/drivers/gpu/drm/panel/panel-sharp-ls037v7dw01.c
@@ -185,9 +185,8 @@ static int ls037v7dw01_probe(struct platform_device *pdev)
 		return PTR_ERR(lcd->ud_gpio);
 	}
 
-	drm_panel_init(&lcd->panel);
-	lcd->panel.dev = &pdev->dev;
-	lcd->panel.funcs = &ls037v7dw01_funcs;
+	drm_panel_init(&lcd->panel, &pdev->dev, &ls037v7dw01_funcs,
+		       DRM_MODE_CONNECTOR_DPI);
 
 	return drm_panel_add(&lcd->panel);
 }
diff --git a/drivers/gpu/drm/panel/panel-sharp-ls043t1le01.c b/drivers/gpu/drm/panel/panel-sharp-ls043t1le01.c
index c39abde9f9f1..b963ba4ab589 100644
--- a/drivers/gpu/drm/panel/panel-sharp-ls043t1le01.c
+++ b/drivers/gpu/drm/panel/panel-sharp-ls043t1le01.c
@@ -264,9 +264,8 @@ static int sharp_nt_panel_add(struct sharp_nt_panel *sharp_nt)
 	if (IS_ERR(sharp_nt->backlight))
 		return PTR_ERR(sharp_nt->backlight);
 
-	drm_panel_init(&sharp_nt->base);
-	sharp_nt->base.funcs = &sharp_nt_panel_funcs;
-	sharp_nt->base.dev = &sharp_nt->dsi->dev;
+	drm_panel_init(&sharp_nt->base, &sharp_nt->dsi->dev,
+		       &sharp_nt_panel_funcs, DRM_MODE_CONNECTOR_DSI);
 
 	return drm_panel_add(&sharp_nt->base);
 }
diff --git a/drivers/gpu/drm/panel/panel-simple.c b/drivers/gpu/drm/panel/panel-simple.c
index 28fa6ba7b767..5d487686d25c 100644
--- a/drivers/gpu/drm/panel/panel-simple.c
+++ b/drivers/gpu/drm/panel/panel-simple.c
@@ -94,6 +94,7 @@ struct panel_desc {
 
 	u32 bus_format;
 	u32 bus_flags;
+	int connector_type;
 };
 
 struct panel_simple {
@@ -464,9 +465,8 @@ static int panel_simple_probe(struct device *dev, const struct panel_desc *desc)
 	if (!of_get_display_timing(dev->of_node, "panel-timing", &dt))
 		panel_simple_parse_panel_timing_node(dev, panel, &dt);
 
-	drm_panel_init(&panel->base);
-	panel->base.dev = dev;
-	panel->base.funcs = &panel_simple_funcs;
+	drm_panel_init(&panel->base, dev, &panel_simple_funcs,
+		       desc->connector_type);
 
 	err = drm_panel_add(&panel->base);
 	if (err < 0)
@@ -833,6 +833,7 @@ static const struct panel_desc auo_g133han01 = {
 		.unprepare = 1000,
 	},
 	.bus_format = MEDIA_BUS_FMT_RGB888_1X7X4_JEIDA,
+	.connector_type = DRM_MODE_CONNECTOR_LVDS,
 };
 
 static const struct display_timing auo_g185han01_timings = {
@@ -862,6 +863,7 @@ static const struct panel_desc auo_g185han01 = {
 		.unprepare = 1000,
 	},
 	.bus_format = MEDIA_BUS_FMT_RGB888_1X7X4_SPWG,
+	.connector_type = DRM_MODE_CONNECTOR_LVDS,
 };
 
 static const struct display_timing auo_p320hvn03_timings = {
@@ -890,6 +892,7 @@ static const struct panel_desc auo_p320hvn03 = {
 		.unprepare = 500,
 	},
 	.bus_format = MEDIA_BUS_FMT_RGB888_1X7X4_SPWG,
+	.connector_type = DRM_MODE_CONNECTOR_LVDS,
 };
 
 static const struct drm_display_mode auo_t215hvn01_mode = {
@@ -1205,6 +1208,7 @@ static const struct panel_desc dlc_dlc0700yzg_1 = {
 		.disable = 200,
 	},
 	.bus_format = MEDIA_BUS_FMT_RGB666_1X7X3_SPWG,
+	.connector_type = DRM_MODE_CONNECTOR_LVDS,
 };
 
 static const struct display_timing dlc_dlc1010gig_timing = {
@@ -1235,6 +1239,7 @@ static const struct panel_desc dlc_dlc1010gig = {
 		.unprepare = 60,
 	},
 	.bus_format = MEDIA_BUS_FMT_RGB888_1X7X4_SPWG,
+	.connector_type = DRM_MODE_CONNECTOR_LVDS,
 };
 
 static const struct drm_display_mode edt_et035012dm6_mode = {
@@ -1501,6 +1506,7 @@ static const struct panel_desc hannstar_hsd070pww1 = {
 		.height = 94,
 	},
 	.bus_format = MEDIA_BUS_FMT_RGB666_1X7X3_SPWG,
+	.connector_type = DRM_MODE_CONNECTOR_LVDS,
 };
 
 static const struct display_timing hannstar_hsd100pxn1_timing = {
@@ -1525,6 +1531,7 @@ static const struct panel_desc hannstar_hsd100pxn1 = {
 		.height = 152,
 	},
 	.bus_format = MEDIA_BUS_FMT_RGB666_1X7X3_SPWG,
+	.connector_type = DRM_MODE_CONNECTOR_LVDS,
 };
 
 static const struct drm_display_mode hitachi_tx23d38vm0caa_mode = {
@@ -1631,6 +1638,7 @@ static const struct panel_desc innolux_g070y2_l01 = {
 		.unprepare = 800,
 	},
 	.bus_format = MEDIA_BUS_FMT_RGB888_1X7X4_SPWG,
+	.connector_type = DRM_MODE_CONNECTOR_LVDS,
 };
 
 static const struct display_timing innolux_g101ice_l01_timing = {
@@ -1659,6 +1667,7 @@ static const struct panel_desc innolux_g101ice_l01 = {
 		.disable = 200,
 	},
 	.bus_format = MEDIA_BUS_FMT_RGB888_1X7X4_SPWG,
+	.connector_type = DRM_MODE_CONNECTOR_LVDS,
 };
 
 static const struct display_timing innolux_g121i1_l01_timing = {
@@ -1686,6 +1695,7 @@ static const struct panel_desc innolux_g121i1_l01 = {
 		.disable = 20,
 	},
 	.bus_format = MEDIA_BUS_FMT_RGB888_1X7X4_SPWG,
+	.connector_type = DRM_MODE_CONNECTOR_LVDS,
 };
 
 static const struct drm_display_mode innolux_g121x1_l03_mode = {
@@ -1869,6 +1879,7 @@ static const struct panel_desc koe_tx31d200vm0baa = {
 		.height = 109,
 	},
 	.bus_format = MEDIA_BUS_FMT_RGB666_1X7X3_SPWG,
+	.connector_type = DRM_MODE_CONNECTOR_LVDS,
 };
 
 static const struct display_timing kyo_tcg121xglp_timing = {
@@ -1893,6 +1904,7 @@ static const struct panel_desc kyo_tcg121xglp = {
 		.height = 184,
 	},
 	.bus_format = MEDIA_BUS_FMT_RGB888_1X7X4_SPWG,
+	.connector_type = DRM_MODE_CONNECTOR_LVDS,
 };
 
 static const struct drm_display_mode lemaker_bl035_rgb_002_mode = {
@@ -1941,6 +1953,7 @@ static const struct panel_desc lg_lb070wv8 = {
 		.height = 91,
 	},
 	.bus_format = MEDIA_BUS_FMT_RGB888_1X7X4_SPWG,
+	.connector_type = DRM_MODE_CONNECTOR_LVDS,
 };
 
 static const struct drm_display_mode lg_lp079qx1_sp0v_mode = {
@@ -2063,6 +2076,7 @@ static const struct panel_desc mitsubishi_aa070mc01 = {
 		.disable = 400,
 	},
 	.bus_format = MEDIA_BUS_FMT_RGB888_1X7X4_SPWG,
+	.connector_type = DRM_MODE_CONNECTOR_LVDS,
 	.bus_flags = DRM_BUS_FLAG_DE_HIGH,
 };
 
@@ -2091,6 +2105,7 @@ static const struct panel_desc nec_nl12880bc20_05 = {
 		.disable = 50,
 	},
 	.bus_format = MEDIA_BUS_FMT_RGB888_1X7X4_SPWG,
+	.connector_type = DRM_MODE_CONNECTOR_LVDS,
 };
 
 static const struct drm_display_mode nec_nl4827hc19_05b_mode = {
@@ -2193,6 +2208,7 @@ static const struct panel_desc nlt_nl192108ac18_02d = {
 		.unprepare = 500,
 	},
 	.bus_format = MEDIA_BUS_FMT_RGB888_1X7X4_SPWG,
+	.connector_type = DRM_MODE_CONNECTOR_LVDS,
 };
 
 static const struct drm_display_mode nvd_9128_mode = {
@@ -2216,6 +2232,7 @@ static const struct panel_desc nvd_9128 = {
 		.height = 88,
 	},
 	.bus_format = MEDIA_BUS_FMT_RGB888_1X7X4_SPWG,
+	.connector_type = DRM_MODE_CONNECTOR_LVDS,
 };
 
 static const struct display_timing okaya_rs800480t_7x0gp_timing = {
@@ -2381,6 +2398,7 @@ static const struct panel_desc osddisplays_osd070t1718_19ts = {
 	},
 	.bus_format = MEDIA_BUS_FMT_RGB888_1X24,
 	.bus_flags = DRM_BUS_FLAG_DE_HIGH | DRM_BUS_FLAG_PIXDATA_DRIVE_POSEDGE,
+	.connector_type = DRM_MODE_CONNECTOR_DPI,
 };
 
 static const struct drm_display_mode pda_91_00156_a0_mode = {
@@ -2628,6 +2646,7 @@ static const struct panel_desc sharp_lq101k1ly04 = {
 		.height = 136,
 	},
 	.bus_format = MEDIA_BUS_FMT_RGB888_1X7X4_JEIDA,
+	.connector_type = DRM_MODE_CONNECTOR_LVDS,
 };
 
 static const struct display_timing sharp_lq123p1jx31_timing = {
@@ -2807,6 +2826,7 @@ static const struct panel_desc tianma_tm070jdhg30 = {
 		.height = 95,
 	},
 	.bus_format = MEDIA_BUS_FMT_RGB888_1X7X4_SPWG,
+	.connector_type = DRM_MODE_CONNECTOR_LVDS,
 };
 
 static const struct display_timing tianma_tm070rvhg71_timing = {
@@ -2831,6 +2851,7 @@ static const struct panel_desc tianma_tm070rvhg71 = {
 		.height = 86,
 	},
 	.bus_format = MEDIA_BUS_FMT_RGB888_1X7X4_SPWG,
+	.connector_type = DRM_MODE_CONNECTOR_LVDS,
 };
 
 static const struct drm_display_mode ti_nspire_cx_lcd_mode[] = {
@@ -2913,6 +2934,7 @@ static const struct panel_desc toshiba_lt089ac29000 = {
 	},
 	.bus_format = MEDIA_BUS_FMT_RGB888_1X24,
 	.bus_flags = DRM_BUS_FLAG_DE_HIGH | DRM_BUS_FLAG_PIXDATA_DRIVE_POSEDGE,
+	.connector_type = DRM_MODE_CONNECTOR_LVDS,
 };
 
 static const struct drm_display_mode tpk_f07a_0102_mode = {
@@ -2983,6 +3005,7 @@ static const struct panel_desc urt_umsh_8596md_lvds = {
 		.height = 91,
 	},
 	.bus_format = MEDIA_BUS_FMT_RGB666_1X7X3_SPWG,
+	.connector_type = DRM_MODE_CONNECTOR_LVDS,
 };
 
 static const struct panel_desc urt_umsh_8596md_parallel = {
diff --git a/drivers/gpu/drm/panel/panel-sitronix-st7701.c b/drivers/gpu/drm/panel/panel-sitronix-st7701.c
index 09c5d9a6f9fa..ee3f23f45755 100644
--- a/drivers/gpu/drm/panel/panel-sitronix-st7701.c
+++ b/drivers/gpu/drm/panel/panel-sitronix-st7701.c
@@ -369,7 +369,8 @@ static int st7701_dsi_probe(struct mipi_dsi_device *dsi)
 	if (IS_ERR(st7701->backlight))
 		return PTR_ERR(st7701->backlight);
 
-	drm_panel_init(&st7701->panel);
+	drm_panel_init(&st7701->panel, &dsi->dev, &st7701_funcs,
+		       DRM_MODE_CONNECTOR_DSI);
 
 	/**
 	 * Once sleep out has been issued, ST7701 IC required to wait 120ms
@@ -381,8 +382,6 @@ static int st7701_dsi_probe(struct mipi_dsi_device *dsi)
 	 * ts8550b and there is no valid documentation for that.
 	 */
 	st7701->sleep_delay = 120 + desc->panel_sleep_delay;
-	st7701->panel.funcs = &st7701_funcs;
-	st7701->panel.dev = &dsi->dev;
 
 	ret = drm_panel_add(&st7701->panel);
 	if (ret < 0)
diff --git a/drivers/gpu/drm/panel/panel-sitronix-st7789v.c b/drivers/gpu/drm/panel/panel-sitronix-st7789v.c
index 5e3e92ea9ea6..108a85bb6667 100644
--- a/drivers/gpu/drm/panel/panel-sitronix-st7789v.c
+++ b/drivers/gpu/drm/panel/panel-sitronix-st7789v.c
@@ -381,8 +381,8 @@ static int st7789v_probe(struct spi_device *spi)
 	spi_set_drvdata(spi, ctx);
 	ctx->spi = spi;
 
-	ctx->panel.dev = &spi->dev;
-	ctx->panel.funcs = &st7789v_drm_funcs;
+	drm_panel_init(&ctx->panel, &spi->dev, &st7789v_drm_funcs,
+		       DRM_MODE_CONNECTOR_DPI);
 
 	ctx->power = devm_regulator_get(&spi->dev, "power");
 	if (IS_ERR(ctx->power))
diff --git a/drivers/gpu/drm/panel/panel-sony-acx565akm.c b/drivers/gpu/drm/panel/panel-sony-acx565akm.c
index 305259b58767..d6387d8f88a3 100644
--- a/drivers/gpu/drm/panel/panel-sony-acx565akm.c
+++ b/drivers/gpu/drm/panel/panel-sony-acx565akm.c
@@ -648,9 +648,8 @@ static int acx565akm_probe(struct spi_device *spi)
 			return ret;
 	}
 
-	drm_panel_init(&lcd->panel);
-	lcd->panel.dev = &lcd->spi->dev;
-	lcd->panel.funcs = &acx565akm_funcs;
+	drm_panel_init(&lcd->panel, &lcd->spi->dev, &acx565akm_funcs,
+		       DRM_MODE_CONNECTOR_DPI);
 
 	ret = drm_panel_add(&lcd->panel);
 	if (ret < 0) {
@@ -684,9 +683,17 @@ static const struct of_device_id acx565akm_of_match[] = {
 
 MODULE_DEVICE_TABLE(of, acx565akm_of_match);
 
+static const struct spi_device_id acx565akm_ids[] = {
+	{ "acx565akm", 0 },
+	{ /* sentinel */ }
+};
+
+MODULE_DEVICE_TABLE(spi, acx565akm_ids);
+
 static struct spi_driver acx565akm_driver = {
 	.probe		= acx565akm_probe,
 	.remove		= acx565akm_remove,
+	.id_table	= acx565akm_ids,
 	.driver		= {
 		.name	= "panel-sony-acx565akm",
 		.of_match_table = acx565akm_of_match,
@@ -695,7 +702,6 @@ static struct spi_driver acx565akm_driver = {
 
 module_spi_driver(acx565akm_driver);
 
-MODULE_ALIAS("spi:sony,acx565akm");
 MODULE_AUTHOR("Nokia Corporation");
 MODULE_DESCRIPTION("Sony ACX565AKM LCD Panel Driver");
 MODULE_LICENSE("GPL");
diff --git a/drivers/gpu/drm/panel/panel-tpo-td028ttec1.c b/drivers/gpu/drm/panel/panel-tpo-td028ttec1.c
index d7b2e34626ef..c44d6a65c0aa 100644
--- a/drivers/gpu/drm/panel/panel-tpo-td028ttec1.c
+++ b/drivers/gpu/drm/panel/panel-tpo-td028ttec1.c
@@ -347,9 +347,8 @@ static int td028ttec1_probe(struct spi_device *spi)
 		return ret;
 	}
 
-	drm_panel_init(&lcd->panel);
-	lcd->panel.dev = &lcd->spi->dev;
-	lcd->panel.funcs = &td028ttec1_funcs;
+	drm_panel_init(&lcd->panel, &lcd->spi->dev, &td028ttec1_funcs,
+		       DRM_MODE_CONNECTOR_DPI);
 
 	return drm_panel_add(&lcd->panel);
 }
@@ -375,8 +374,7 @@ static const struct of_device_id td028ttec1_of_match[] = {
 MODULE_DEVICE_TABLE(of, td028ttec1_of_match);
 
 static const struct spi_device_id td028ttec1_ids[] = {
-	{ "tpo,td028ttec1", 0},
-	{ "toppoly,td028ttec1", 0 },
+	{ "td028ttec1", 0 },
 	{ /* sentinel */ }
 };
 
diff --git a/drivers/gpu/drm/panel/panel-tpo-td043mtea1.c b/drivers/gpu/drm/panel/panel-tpo-td043mtea1.c
index 84370562910f..621b65feec07 100644
--- a/drivers/gpu/drm/panel/panel-tpo-td043mtea1.c
+++ b/drivers/gpu/drm/panel/panel-tpo-td043mtea1.c
@@ -458,9 +458,8 @@ static int td043mtea1_probe(struct spi_device *spi)
 		return ret;
 	}
 
-	drm_panel_init(&lcd->panel);
-	lcd->panel.dev = &lcd->spi->dev;
-	lcd->panel.funcs = &td043mtea1_funcs;
+	drm_panel_init(&lcd->panel, &lcd->spi->dev, &td043mtea1_funcs,
+		       DRM_MODE_CONNECTOR_DPI);
 
 	ret = drm_panel_add(&lcd->panel);
 	if (ret < 0) {
@@ -491,9 +490,17 @@ static const struct of_device_id td043mtea1_of_match[] = {
 
 MODULE_DEVICE_TABLE(of, td043mtea1_of_match);
 
+static const struct spi_device_id td043mtea1_ids[] = {
+	{ "td043mtea1", 0 },
+	{ /* sentinel */ }
+};
+
+MODULE_DEVICE_TABLE(spi, td043mtea1_ids);
+
 static struct spi_driver td043mtea1_driver = {
 	.probe		= td043mtea1_probe,
 	.remove		= td043mtea1_remove,
+	.id_table	= td043mtea1_ids,
 	.driver		= {
 		.name	= "panel-tpo-td043mtea1",
 		.pm	= &td043mtea1_pm_ops,
@@ -503,7 +510,6 @@ static struct spi_driver td043mtea1_driver = {
 
 module_spi_driver(td043mtea1_driver);
 
-MODULE_ALIAS("spi:tpo,td043mtea1");
 MODULE_AUTHOR("Gražvydas Ignotas <notasas@gmail.com>");
 MODULE_DESCRIPTION("TPO TD043MTEA1 Panel Driver");
 MODULE_LICENSE("GPL");
diff --git a/drivers/gpu/drm/panel/panel-tpo-tpg110.c b/drivers/gpu/drm/panel/panel-tpo-tpg110.c
index 71591e5f5938..1a5418ae2ccf 100644
--- a/drivers/gpu/drm/panel/panel-tpo-tpg110.c
+++ b/drivers/gpu/drm/panel/panel-tpo-tpg110.c
@@ -457,9 +457,8 @@ static int tpg110_probe(struct spi_device *spi)
 	if (ret)
 		return ret;
 
-	drm_panel_init(&tpg->panel);
-	tpg->panel.dev = dev;
-	tpg->panel.funcs = &tpg110_drm_funcs;
+	drm_panel_init(&tpg->panel, dev, &tpg110_drm_funcs,
+		       DRM_MODE_CONNECTOR_DPI);
 	spi_set_drvdata(spi, tpg);
 
 	return drm_panel_add(&tpg->panel);
diff --git a/drivers/gpu/drm/panel/panel-truly-nt35597.c b/drivers/gpu/drm/panel/panel-truly-nt35597.c
index 77e1311b7c69..0feea2456e14 100644
--- a/drivers/gpu/drm/panel/panel-truly-nt35597.c
+++ b/drivers/gpu/drm/panel/panel-truly-nt35597.c
@@ -518,9 +518,8 @@ static int truly_nt35597_panel_add(struct truly_nt35597 *ctx)
 	/* dual port */
 	gpiod_set_value(ctx->mode_gpio, 0);
 
-	drm_panel_init(&ctx->panel);
-	ctx->panel.dev = dev;
-	ctx->panel.funcs = &truly_nt35597_drm_funcs;
+	drm_panel_init(&ctx->panel, dev, &truly_nt35597_drm_funcs,
+		       DRM_MODE_CONNECTOR_DSI);
 	drm_panel_add(&ctx->panel);
 
 	return 0;
diff --git a/drivers/gpu/drm/panfrost/TODO b/drivers/gpu/drm/panfrost/TODO
index 536a0d4f8d29..8c811a9e683b 100644
--- a/drivers/gpu/drm/panfrost/TODO
+++ b/drivers/gpu/drm/panfrost/TODO
@@ -10,3 +10,5 @@
 
 - Compute job support. So called 'compute only' jobs need to be plumbed up to
   userspace.
+
+- Support core dump on job failure
diff --git a/drivers/gpu/drm/panfrost/panfrost_devfreq.c b/drivers/gpu/drm/panfrost/panfrost_devfreq.c
index 12ff77dacc95..4c4e8a30a1ac 100644
--- a/drivers/gpu/drm/panfrost/panfrost_devfreq.c
+++ b/drivers/gpu/drm/panfrost/panfrost_devfreq.c
@@ -13,97 +13,42 @@
 #include "panfrost_gpu.h"
 #include "panfrost_regs.h"
 
-static void panfrost_devfreq_update_utilization(struct panfrost_device *pfdev, int slot);
+static void panfrost_devfreq_update_utilization(struct panfrost_device *pfdev);
 
 static int panfrost_devfreq_target(struct device *dev, unsigned long *freq,
 				   u32 flags)
 {
-	struct panfrost_device *pfdev = platform_get_drvdata(to_platform_device(dev));
-	struct dev_pm_opp *opp;
-	unsigned long old_clk_rate = pfdev->devfreq.cur_freq;
-	unsigned long target_volt, target_rate;
+	struct panfrost_device *pfdev = dev_get_drvdata(dev);
 	int err;
 
-	opp = devfreq_recommended_opp(dev, freq, flags);
-	if (IS_ERR(opp))
-		return PTR_ERR(opp);
-
-	target_rate = dev_pm_opp_get_freq(opp);
-	target_volt = dev_pm_opp_get_voltage(opp);
-	dev_pm_opp_put(opp);
-
-	if (old_clk_rate == target_rate)
-		return 0;
-
-	/*
-	 * If frequency scaling from low to high, adjust voltage first.
-	 * If frequency scaling from high to low, adjust frequency first.
-	 */
-	if (old_clk_rate < target_rate) {
-		err = regulator_set_voltage(pfdev->regulator, target_volt,
-					    target_volt);
-		if (err) {
-			dev_err(dev, "Cannot set voltage %lu uV\n",
-				target_volt);
-			return err;
-		}
-	}
-
-	err = clk_set_rate(pfdev->clock, target_rate);
-	if (err) {
-		dev_err(dev, "Cannot set frequency %lu (%d)\n", target_rate,
-			err);
-		regulator_set_voltage(pfdev->regulator, pfdev->devfreq.cur_volt,
-				      pfdev->devfreq.cur_volt);
+	err = dev_pm_opp_set_rate(dev, *freq);
+	if (err)
 		return err;
-	}
 
-	if (old_clk_rate > target_rate) {
-		err = regulator_set_voltage(pfdev->regulator, target_volt,
-					    target_volt);
-		if (err)
-			dev_err(dev, "Cannot set voltage %lu uV\n", target_volt);
-	}
-
-	pfdev->devfreq.cur_freq = target_rate;
-	pfdev->devfreq.cur_volt = target_volt;
+	*freq = clk_get_rate(pfdev->clock);
 
 	return 0;
 }
 
 static void panfrost_devfreq_reset(struct panfrost_device *pfdev)
 {
-	ktime_t now = ktime_get();
-	int i;
-
-	for (i = 0; i < NUM_JOB_SLOTS; i++) {
-		pfdev->devfreq.slot[i].busy_time = 0;
-		pfdev->devfreq.slot[i].idle_time = 0;
-		pfdev->devfreq.slot[i].time_last_update = now;
-	}
+	pfdev->devfreq.busy_time = 0;
+	pfdev->devfreq.idle_time = 0;
+	pfdev->devfreq.time_last_update = ktime_get();
 }
 
 static int panfrost_devfreq_get_dev_status(struct device *dev,
 					   struct devfreq_dev_status *status)
 {
-	struct panfrost_device *pfdev = platform_get_drvdata(to_platform_device(dev));
-	int i;
+	struct panfrost_device *pfdev = dev_get_drvdata(dev);
 
-	for (i = 0; i < NUM_JOB_SLOTS; i++) {
-		panfrost_devfreq_update_utilization(pfdev, i);
-	}
+	panfrost_devfreq_update_utilization(pfdev);
 
 	status->current_frequency = clk_get_rate(pfdev->clock);
-	status->total_time = ktime_to_ns(ktime_add(pfdev->devfreq.slot[0].busy_time,
-						   pfdev->devfreq.slot[0].idle_time));
+	status->total_time = ktime_to_ns(ktime_add(pfdev->devfreq.busy_time,
+						   pfdev->devfreq.idle_time));
 
-	status->busy_time = 0;
-	for (i = 0; i < NUM_JOB_SLOTS; i++) {
-		status->busy_time += ktime_to_ns(pfdev->devfreq.slot[i].busy_time);
-	}
-
-	/* We're scheduling only to one core atm, so don't divide for now */
-	/* status->busy_time /= NUM_JOB_SLOTS; */
+	status->busy_time = ktime_to_ns(pfdev->devfreq.busy_time);
 
 	panfrost_devfreq_reset(pfdev);
 
@@ -119,7 +64,7 @@ static int panfrost_devfreq_get_cur_freq(struct device *dev, unsigned long *freq
 {
 	struct panfrost_device *pfdev = platform_get_drvdata(to_platform_device(dev));
 
-	*freq = pfdev->devfreq.cur_freq;
+	*freq = clk_get_rate(pfdev->clock);
 
 	return 0;
 }
@@ -135,6 +80,7 @@ int panfrost_devfreq_init(struct panfrost_device *pfdev)
 {
 	int ret;
 	struct dev_pm_opp *opp;
+	unsigned long cur_freq;
 
 	ret = dev_pm_opp_of_add_table(&pfdev->pdev->dev);
 	if (ret == -ENODEV) /* Optional, continue without devfreq */
@@ -144,13 +90,13 @@ int panfrost_devfreq_init(struct panfrost_device *pfdev)
 
 	panfrost_devfreq_reset(pfdev);
 
-	pfdev->devfreq.cur_freq = clk_get_rate(pfdev->clock);
+	cur_freq = clk_get_rate(pfdev->clock);
 
-	opp = devfreq_recommended_opp(&pfdev->pdev->dev, &pfdev->devfreq.cur_freq, 0);
+	opp = devfreq_recommended_opp(&pfdev->pdev->dev, &cur_freq, 0);
 	if (IS_ERR(opp))
 		return PTR_ERR(opp);
 
-	panfrost_devfreq_profile.initial_freq = pfdev->devfreq.cur_freq;
+	panfrost_devfreq_profile.initial_freq = cur_freq;
 	dev_pm_opp_put(opp);
 
 	pfdev->devfreq.devfreq = devm_devfreq_add_device(&pfdev->pdev->dev,
@@ -174,14 +120,10 @@ void panfrost_devfreq_fini(struct panfrost_device *pfdev)
 
 void panfrost_devfreq_resume(struct panfrost_device *pfdev)
 {
-	int i;
-
 	if (!pfdev->devfreq.devfreq)
 		return;
 
 	panfrost_devfreq_reset(pfdev);
-	for (i = 0; i < NUM_JOB_SLOTS; i++)
-		pfdev->devfreq.slot[i].busy = false;
 
 	devfreq_resume_device(pfdev->devfreq.devfreq);
 }
@@ -194,9 +136,8 @@ void panfrost_devfreq_suspend(struct panfrost_device *pfdev)
 	devfreq_suspend_device(pfdev->devfreq.devfreq);
 }
 
-static void panfrost_devfreq_update_utilization(struct panfrost_device *pfdev, int slot)
+static void panfrost_devfreq_update_utilization(struct panfrost_device *pfdev)
 {
-	struct panfrost_devfreq_slot *devfreq_slot = &pfdev->devfreq.slot[slot];
 	ktime_t now;
 	ktime_t last;
 
@@ -204,22 +145,27 @@ static void panfrost_devfreq_update_utilization(struct panfrost_device *pfdev, i
 		return;
 
 	now = ktime_get();
-	last = pfdev->devfreq.slot[slot].time_last_update;
+	last = pfdev->devfreq.time_last_update;
 
-	/* If we last recorded a transition to busy, we have been idle since */
-	if (devfreq_slot->busy)
-		pfdev->devfreq.slot[slot].busy_time += ktime_sub(now, last);
+	if (atomic_read(&pfdev->devfreq.busy_count) > 0)
+		pfdev->devfreq.busy_time += ktime_sub(now, last);
 	else
-		pfdev->devfreq.slot[slot].idle_time += ktime_sub(now, last);
+		pfdev->devfreq.idle_time += ktime_sub(now, last);
 
-	pfdev->devfreq.slot[slot].time_last_update = now;
+	pfdev->devfreq.time_last_update = now;
+}
+
+void panfrost_devfreq_record_busy(struct panfrost_device *pfdev)
+{
+	panfrost_devfreq_update_utilization(pfdev);
+	atomic_inc(&pfdev->devfreq.busy_count);
 }
 
-/* The job scheduler is expected to call this at every transition busy <-> idle */
-void panfrost_devfreq_record_transition(struct panfrost_device *pfdev, int slot)
+void panfrost_devfreq_record_idle(struct panfrost_device *pfdev)
 {
-	struct panfrost_devfreq_slot *devfreq_slot = &pfdev->devfreq.slot[slot];
+	int count;
 
-	panfrost_devfreq_update_utilization(pfdev, slot);
-	devfreq_slot->busy = !devfreq_slot->busy;
+	panfrost_devfreq_update_utilization(pfdev);
+	count = atomic_dec_if_positive(&pfdev->devfreq.busy_count);
+	WARN_ON(count < 0);
 }
diff --git a/drivers/gpu/drm/panfrost/panfrost_devfreq.h b/drivers/gpu/drm/panfrost/panfrost_devfreq.h
index e3bc63e82843..0611beffc8d0 100644
--- a/drivers/gpu/drm/panfrost/panfrost_devfreq.h
+++ b/drivers/gpu/drm/panfrost/panfrost_devfreq.h
@@ -10,6 +10,7 @@ void panfrost_devfreq_fini(struct panfrost_device *pfdev);
 void panfrost_devfreq_resume(struct panfrost_device *pfdev);
 void panfrost_devfreq_suspend(struct panfrost_device *pfdev);
 
-void panfrost_devfreq_record_transition(struct panfrost_device *pfdev, int slot);
+void panfrost_devfreq_record_busy(struct panfrost_device *pfdev);
+void panfrost_devfreq_record_idle(struct panfrost_device *pfdev);
 
 #endif /* __PANFROST_DEVFREQ_H__ */
diff --git a/drivers/gpu/drm/panfrost/panfrost_device.h b/drivers/gpu/drm/panfrost/panfrost_device.h
index 9c39b9794811..06713811b92c 100644
--- a/drivers/gpu/drm/panfrost/panfrost_device.h
+++ b/drivers/gpu/drm/panfrost/panfrost_device.h
@@ -51,13 +51,6 @@ struct panfrost_features {
 	unsigned long hw_issues[64 / BITS_PER_LONG];
 };
 
-struct panfrost_devfreq_slot {
-	ktime_t busy_time;
-	ktime_t idle_time;
-	ktime_t time_last_update;
-	bool busy;
-};
-
 struct panfrost_device {
 	struct device *dev;
 	struct drm_device *ddev;
@@ -93,9 +86,10 @@ struct panfrost_device {
 	struct {
 		struct devfreq *devfreq;
 		struct thermal_cooling_device *cooling;
-		unsigned long cur_freq;
-		unsigned long cur_volt;
-		struct panfrost_devfreq_slot slot[NUM_JOB_SLOTS];
+		ktime_t busy_time;
+		ktime_t idle_time;
+		ktime_t time_last_update;
+		atomic_t busy_count;
 	} devfreq;
 };
 
diff --git a/drivers/gpu/drm/panfrost/panfrost_drv.c b/drivers/gpu/drm/panfrost/panfrost_drv.c
index bc2ddeb55f5d..9458dc6c750c 100644
--- a/drivers/gpu/drm/panfrost/panfrost_drv.c
+++ b/drivers/gpu/drm/panfrost/panfrost_drv.c
@@ -470,7 +470,7 @@ static const struct drm_ioctl_desc panfrost_drm_driver_ioctls[] = {
 	PANFROST_IOCTL(MADVISE,		madvise,	DRM_RENDER_ALLOW),
 };
 
-DEFINE_DRM_GEM_SHMEM_FOPS(panfrost_drm_driver_fops);
+DEFINE_DRM_GEM_FOPS(panfrost_drm_driver_fops);
 
 /*
  * Panfrost driver version:
@@ -556,11 +556,11 @@ static int panfrost_probe(struct platform_device *pdev)
 	return 0;
 
 err_out2:
+	pm_runtime_disable(pfdev->dev);
 	panfrost_devfreq_fini(pfdev);
 err_out1:
 	panfrost_device_fini(pfdev);
 err_out0:
-	pm_runtime_disable(pfdev->dev);
 	drm_dev_put(ddev);
 	return err;
 }
diff --git a/drivers/gpu/drm/panfrost/panfrost_gem.c b/drivers/gpu/drm/panfrost/panfrost_gem.c
index acb07fe06580..deca0c30bbd4 100644
--- a/drivers/gpu/drm/panfrost/panfrost_gem.c
+++ b/drivers/gpu/drm/panfrost/panfrost_gem.c
@@ -112,7 +112,7 @@ static const struct drm_gem_object_funcs panfrost_gem_funcs = {
 	.get_sg_table = drm_gem_shmem_get_sg_table,
 	.vmap = drm_gem_shmem_vmap,
 	.vunmap = drm_gem_shmem_vunmap,
-	.vm_ops = &drm_gem_shmem_vm_ops,
+	.mmap = drm_gem_shmem_mmap,
 };
 
 /**
diff --git a/drivers/gpu/drm/panfrost/panfrost_gpu.c b/drivers/gpu/drm/panfrost/panfrost_gpu.c
index f67ed925c0ef..8822ec13a0d6 100644
--- a/drivers/gpu/drm/panfrost/panfrost_gpu.c
+++ b/drivers/gpu/drm/panfrost/panfrost_gpu.c
@@ -208,6 +208,9 @@ static void panfrost_gpu_init_features(struct panfrost_device *pfdev)
 	pfdev->features.mem_features = gpu_read(pfdev, GPU_MEM_FEATURES);
 	pfdev->features.mmu_features = gpu_read(pfdev, GPU_MMU_FEATURES);
 	pfdev->features.thread_features = gpu_read(pfdev, GPU_THREAD_FEATURES);
+	pfdev->features.max_threads = gpu_read(pfdev, GPU_THREAD_MAX_THREADS);
+	pfdev->features.thread_max_workgroup_sz = gpu_read(pfdev, GPU_THREAD_MAX_WORKGROUP_SIZE);
+	pfdev->features.thread_max_barrier_sz = gpu_read(pfdev, GPU_THREAD_MAX_BARRIER_SIZE);
 	pfdev->features.coherency_features = gpu_read(pfdev, GPU_COHERENCY_FEATURES);
 	for (i = 0; i < 4; i++)
 		pfdev->features.texture_features[i] = gpu_read(pfdev, GPU_TEXTURE_FEATURES(i));
diff --git a/drivers/gpu/drm/panfrost/panfrost_issues.h b/drivers/gpu/drm/panfrost/panfrost_issues.h
index cec6dcdadb5c..8e59d765bf19 100644
--- a/drivers/gpu/drm/panfrost/panfrost_issues.h
+++ b/drivers/gpu/drm/panfrost/panfrost_issues.h
@@ -13,37 +13,118 @@
  * to care about.
  */
 enum panfrost_hw_issue {
+	/* Need way to guarantee that all previously-translated memory accesses
+	 * are commited */
 	HW_ISSUE_6367,
+
+	/* On job complete with non-done the cache is not flushed */
 	HW_ISSUE_6787,
+
+	/* Write of PRFCNT_CONFIG_MODE_MANUAL to PRFCNT_CONFIG causes a
+	 * instrumentation dump if PRFCNT_TILER_EN is enabled */
 	HW_ISSUE_8186,
+
+	/* TIB: Reports faults from a vtile which has not yet been allocated */
 	HW_ISSUE_8245,
+
+	/* uTLB deadlock could occur when writing to an invalid page at the
+	 * same time as access to a valid page in the same uTLB cache line ( ==
+	 * 4 PTEs == 16K block of mapping) */
 	HW_ISSUE_8316,
+
+	/* HT: TERMINATE for RUN command ignored if previous LOAD_DESCRIPTOR is
+	 * still executing */
 	HW_ISSUE_8394,
+
+	/* CSE: Sends a TERMINATED response for a task that should not be
+	 * terminated */
 	HW_ISSUE_8401,
+
+	/* Repeatedly Soft-stopping a job chain consisting of (Vertex Shader,
+	 * Cache Flush, Tiler) jobs causes DATA_INVALID_FAULT on tiler job. */
 	HW_ISSUE_8408,
+
+	/* Disable the Pause Buffer in the LS pipe. */
 	HW_ISSUE_8443,
+
+	/* Change in RMUs in use causes problems related with the core's SDC */
 	HW_ISSUE_8987,
+
+	/* Compute endpoint has a 4-deep queue of tasks, meaning a soft stop
+	 * won't complete until all 4 tasks have completed */
 	HW_ISSUE_9435,
+
+	/* HT: Tiler returns TERMINATED for non-terminated command */
 	HW_ISSUE_9510,
+
+	/* Occasionally the GPU will issue multiple page faults for the same
+	 * address before the MMU page table has been read by the GPU */
 	HW_ISSUE_9630,
+
+	/* RA DCD load request to SDC returns invalid load ignore causing
+	 * colour buffer mismatch */
 	HW_ISSUE_10327,
+
+	/* MMU TLB invalidation hazards */
 	HW_ISSUE_10649,
+
+	/* Missing cache flush in multi core-group configuration */
 	HW_ISSUE_10676,
+
+	/* Chicken bit on T72X for a hardware workaround in compiler */
 	HW_ISSUE_10797,
+
+	/* Soft-stopping fragment jobs might fail with TILE_RANGE_FAULT */
 	HW_ISSUE_10817,
+
+	/* Intermittent missing interrupt on job completion */
 	HW_ISSUE_10883,
+
+	/* Soft-stopping fragment jobs might fail with TILE_RANGE_ERROR
+	 * (similar to issue 10817) and can use #10817 workaround */
 	HW_ISSUE_10959,
+
+	/* Soft-stopped fragment shader job can restart with out-of-bound
+	 * restart index */
 	HW_ISSUE_10969,
+
+	/* Race condition can cause tile list corruption */
 	HW_ISSUE_11020,
+
+	/* Write buffer can cause tile list corruption */
 	HW_ISSUE_11024,
+
+	/* Pause buffer can cause a fragment job hang */
 	HW_ISSUE_11035,
+
+	/* Dynamic Core Scaling not supported due to errata */
 	HW_ISSUE_11056,
+
+	/* Clear encoder state for a hard stopped fragment job which is AFBC
+	 * encoded by soft resetting the GPU. Only for T76X r0p0, r0p1 and
+	 * r0p1_50rel0 */
 	HW_ISSUE_T76X_3542,
+
+	/* Keep tiler module clock on to prevent GPU stall */
 	HW_ISSUE_T76X_3953,
+
+	/* Must ensure L2 is not transitioning when we reset. Workaround with a
+	 * busy wait until L2 completes transition; ensure there is a maximum
+	 * loop count as she may never complete her transition. (On chips
+	 * without this errata, it's totally okay if L2 transitions.) */
 	HW_ISSUE_TMIX_8463,
+
+	/* Don't set SC_LS_ATTR_CHECK_DISABLE/SC_LS_ALLOW_ATTR_TYPES */
 	GPUCORE_1619,
+
+	/* When a hard-stop follows close after a soft-stop, the completion
+	 * code for the terminated job may be incorrectly set to STOPPED */
 	HW_ISSUE_TMIX_8438,
+
+	/* "Protected mode" is buggy on Mali-G31 some Bifrost chips, so the
+	 * kernel must fiddle with L2 caches to prevent data leakage */
 	HW_ISSUE_TGOX_R1_1234,
+
 	HW_ISSUE_END
 };
 
diff --git a/drivers/gpu/drm/panfrost/panfrost_job.c b/drivers/gpu/drm/panfrost/panfrost_job.c
index a58551668d9a..d411eb6c8eb9 100644
--- a/drivers/gpu/drm/panfrost/panfrost_job.c
+++ b/drivers/gpu/drm/panfrost/panfrost_job.c
@@ -155,8 +155,7 @@ static void panfrost_job_hw_submit(struct panfrost_job *job, int js)
 	}
 
 	cfg = panfrost_mmu_as_get(pfdev, &job->file_priv->mmu);
-
-	panfrost_devfreq_record_transition(pfdev, js);
+	panfrost_devfreq_record_busy(pfdev);
 
 	job_write(pfdev, JS_HEAD_NEXT_LO(js), jc_head & 0xFFFFFFFF);
 	job_write(pfdev, JS_HEAD_NEXT_HI(js), jc_head >> 32);
@@ -381,13 +380,19 @@ static void panfrost_job_timedout(struct drm_sched_job *sched_job)
 		job_read(pfdev, JS_TAIL_LO(js)),
 		sched_job);
 
-	mutex_lock(&pfdev->reset_lock);
+	if (!mutex_trylock(&pfdev->reset_lock))
+		return;
 
-	for (i = 0; i < NUM_JOB_SLOTS; i++)
-		drm_sched_stop(&pfdev->js->queue[i].sched, sched_job);
+	for (i = 0; i < NUM_JOB_SLOTS; i++) {
+		struct drm_gpu_scheduler *sched = &pfdev->js->queue[i].sched;
+
+		drm_sched_stop(sched, sched_job);
+		if (js != i)
+			/* Ensure any timeouts on other slots have finished */
+			cancel_delayed_work_sync(&sched->work_tdr);
+	}
 
-	if (sched_job)
-		drm_sched_increase_karma(sched_job);
+	drm_sched_increase_karma(sched_job);
 
 	spin_lock_irqsave(&pfdev->js->job_lock, flags);
 	for (i = 0; i < NUM_JOB_SLOTS; i++) {
@@ -398,9 +403,7 @@ static void panfrost_job_timedout(struct drm_sched_job *sched_job)
 	}
 	spin_unlock_irqrestore(&pfdev->js->job_lock, flags);
 
-	/* panfrost_core_dump(pfdev); */
-
-	panfrost_devfreq_record_transition(pfdev, js);
+	panfrost_devfreq_record_idle(pfdev);
 	panfrost_device_reset(pfdev);
 
 	for (i = 0; i < NUM_JOB_SLOTS; i++)
@@ -463,7 +466,7 @@ static irqreturn_t panfrost_job_irq_handler(int irq, void *data)
 				pfdev->jobs[j] = NULL;
 
 				panfrost_mmu_as_put(pfdev, &job->file_priv->mmu);
-				panfrost_devfreq_record_transition(pfdev, j);
+				panfrost_devfreq_record_idle(pfdev);
 
 				dma_fence_signal_locked(job->done_fence);
 				pm_runtime_put_autosuspend(pfdev->dev);
@@ -564,14 +567,14 @@ int panfrost_job_is_idle(struct panfrost_device *pfdev)
 	struct panfrost_job_slot *js = pfdev->js;
 	int i;
 
+	/* Check whether the hardware is idle */
+	if (atomic_read(&pfdev->devfreq.busy_count))
+		return false;
+
 	for (i = 0; i < NUM_JOB_SLOTS; i++) {
 		/* If there are any jobs in the HW queue, we're not idle */
 		if (atomic_read(&js->queue[i].sched.hw_rq_count))
 			return false;
-
-		/* Check whether the hardware is idle */
-		if (pfdev->devfreq.slot[i].busy)
-			return false;
 	}
 
 	return true;
diff --git a/drivers/gpu/drm/panfrost/panfrost_mmu.c b/drivers/gpu/drm/panfrost/panfrost_mmu.c
index bdd990568476..a3ed64a1f15e 100644
--- a/drivers/gpu/drm/panfrost/panfrost_mmu.c
+++ b/drivers/gpu/drm/panfrost/panfrost_mmu.c
@@ -224,9 +224,9 @@ static size_t get_pgsize(u64 addr, size_t size)
 	return SZ_2M;
 }
 
-void panfrost_mmu_flush_range(struct panfrost_device *pfdev,
-			      struct panfrost_mmu *mmu,
-			      u64 iova, size_t size)
+static void panfrost_mmu_flush_range(struct panfrost_device *pfdev,
+				     struct panfrost_mmu *mmu,
+				     u64 iova, size_t size)
 {
 	if (mmu->as < 0)
 		return;
@@ -406,11 +406,11 @@ addr_to_drm_mm_node(struct panfrost_device *pfdev, int as, u64 addr)
 	spin_lock(&pfdev->as_lock);
 	list_for_each_entry(mmu, &pfdev->as_lru_list, list) {
 		if (as == mmu->as)
-			break;
+			goto found_mmu;
 	}
-	if (as != mmu->as)
-		goto out;
+	goto out;
 
+found_mmu:
 	priv = container_of(mmu, struct panfrost_file_priv, mmu);
 
 	spin_lock(&priv->mm_lock);
@@ -432,7 +432,8 @@ out:
 
 #define NUM_FAULT_PAGES (SZ_2M / PAGE_SIZE)
 
-int panfrost_mmu_map_fault_addr(struct panfrost_device *pfdev, int as, u64 addr)
+static int panfrost_mmu_map_fault_addr(struct panfrost_device *pfdev, int as,
+				       u64 addr)
 {
 	int ret, i;
 	struct panfrost_gem_object *bo;
diff --git a/drivers/gpu/drm/panfrost/panfrost_perfcnt.c b/drivers/gpu/drm/panfrost/panfrost_perfcnt.c
index 83c57d325ca8..2dba192bf198 100644
--- a/drivers/gpu/drm/panfrost/panfrost_perfcnt.c
+++ b/drivers/gpu/drm/panfrost/panfrost_perfcnt.c
@@ -16,6 +16,7 @@
 #include "panfrost_issues.h"
 #include "panfrost_job.h"
 #include "panfrost_mmu.h"
+#include "panfrost_perfcnt.h"
 #include "panfrost_regs.h"
 
 #define COUNTERS_PER_BLOCK		64
diff --git a/drivers/gpu/drm/pl111/pl111_display.c b/drivers/gpu/drm/pl111/pl111_display.c
index 024771a4083e..703ddc803c55 100644
--- a/drivers/gpu/drm/pl111/pl111_display.c
+++ b/drivers/gpu/drm/pl111/pl111_display.c
@@ -48,10 +48,10 @@ irqreturn_t pl111_irq(int irq, void *data)
 }
 
 static enum drm_mode_status
-pl111_mode_valid(struct drm_crtc *crtc,
+pl111_mode_valid(struct drm_simple_display_pipe *pipe,
 		 const struct drm_display_mode *mode)
 {
-	struct drm_device *drm = crtc->dev;
+	struct drm_device *drm = pipe->crtc.dev;
 	struct pl111_drm_dev_private *priv = drm->dev_private;
 	u32 cpp = priv->variant->fb_bpp / 8;
 	u64 bw;
diff --git a/drivers/gpu/drm/pl111/pl111_drv.c b/drivers/gpu/drm/pl111/pl111_drv.c
index 276b53473a84..63dfcda04147 100644
--- a/drivers/gpu/drm/pl111/pl111_drv.c
+++ b/drivers/gpu/drm/pl111/pl111_drv.c
@@ -150,8 +150,8 @@ static int pl111_modeset_init(struct drm_device *dev)
 		return -EPROBE_DEFER;
 
 	if (panel) {
-		bridge = drm_panel_bridge_add(panel,
-					      DRM_MODE_CONNECTOR_Unknown);
+		bridge = drm_panel_bridge_add_typed(panel,
+						    DRM_MODE_CONNECTOR_Unknown);
 		if (IS_ERR(bridge)) {
 			ret = PTR_ERR(bridge);
 			goto out_config;
diff --git a/drivers/gpu/drm/qxl/Kconfig b/drivers/gpu/drm/qxl/Kconfig
index d0d691b31f4a..ca3f51c2a8fe 100644
--- a/drivers/gpu/drm/qxl/Kconfig
+++ b/drivers/gpu/drm/qxl/Kconfig
@@ -4,6 +4,7 @@ config DRM_QXL
 	depends on DRM && PCI && MMU
 	select DRM_KMS_HELPER
 	select DRM_TTM
+	select DRM_TTM_HELPER
 	select CRC32
 	help
 	  QXL virtual GPU for Spice virtualization desktop integration.
diff --git a/drivers/gpu/drm/qxl/qxl_drv.c b/drivers/gpu/drm/qxl/qxl_drv.c
index 265bfe9f8016..1d601f57a6ba 100644
--- a/drivers/gpu/drm/qxl/qxl_drv.c
+++ b/drivers/gpu/drm/qxl/qxl_drv.c
@@ -88,7 +88,7 @@ qxl_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 	if (ret)
 		goto free_dev;
 
-	ret = drm_fb_helper_remove_conflicting_pci_framebuffers(pdev, 0, "qxl");
+	ret = drm_fb_helper_remove_conflicting_pci_framebuffers(pdev, "qxl");
 	if (ret)
 		goto disable_pci;
 
@@ -150,15 +150,7 @@ qxl_pci_remove(struct pci_dev *pdev)
 	drm_dev_put(dev);
 }
 
-static const struct file_operations qxl_fops = {
-	.owner = THIS_MODULE,
-	.open = drm_open,
-	.release = drm_release,
-	.unlocked_ioctl = drm_ioctl,
-	.poll = drm_poll,
-	.read = drm_read,
-	.mmap = qxl_mmap,
-};
+DEFINE_DRM_GEM_FOPS(qxl_fops);
 
 static int qxl_drm_freeze(struct drm_device *dev)
 {
@@ -276,16 +268,8 @@ static struct drm_driver qxl_driver = {
 #endif
 	.prime_handle_to_fd = drm_gem_prime_handle_to_fd,
 	.prime_fd_to_handle = drm_gem_prime_fd_to_handle,
-	.gem_prime_pin = qxl_gem_prime_pin,
-	.gem_prime_unpin = qxl_gem_prime_unpin,
-	.gem_prime_get_sg_table = qxl_gem_prime_get_sg_table,
 	.gem_prime_import_sg_table = qxl_gem_prime_import_sg_table,
-	.gem_prime_vmap = qxl_gem_prime_vmap,
-	.gem_prime_vunmap = qxl_gem_prime_vunmap,
 	.gem_prime_mmap = qxl_gem_prime_mmap,
-	.gem_free_object_unlocked = qxl_gem_object_free,
-	.gem_open_object = qxl_gem_object_open,
-	.gem_close_object = qxl_gem_object_close,
 	.fops = &qxl_fops,
 	.ioctls = qxl_ioctls,
 	.irq_handler = qxl_irq_handler,
diff --git a/drivers/gpu/drm/qxl/qxl_drv.h b/drivers/gpu/drm/qxl/qxl_drv.h
index 9e034c5fa87d..27e45a2d6b52 100644
--- a/drivers/gpu/drm/qxl/qxl_drv.h
+++ b/drivers/gpu/drm/qxl/qxl_drv.h
@@ -38,6 +38,7 @@
 #include <drm/drm_crtc.h>
 #include <drm/drm_encoder.h>
 #include <drm/drm_fb_helper.h>
+#include <drm/drm_gem_ttm_helper.h>
 #include <drm/drm_ioctl.h>
 #include <drm/drm_gem.h>
 #include <drm/qxl_drm.h>
@@ -354,7 +355,8 @@ int qxl_mode_dumb_mmap(struct drm_file *filp,
 /* qxl ttm */
 int qxl_ttm_init(struct qxl_device *qdev);
 void qxl_ttm_fini(struct qxl_device *qdev);
-int qxl_mmap(struct file *filp, struct vm_area_struct *vma);
+int qxl_ttm_io_mem_reserve(struct ttm_bo_device *bdev,
+			   struct ttm_mem_reg *mem);
 
 /* qxl image */
 
diff --git a/drivers/gpu/drm/qxl/qxl_object.c b/drivers/gpu/drm/qxl/qxl_object.c
index 548dfe6f3b26..ab72dc3476e9 100644
--- a/drivers/gpu/drm/qxl/qxl_object.c
+++ b/drivers/gpu/drm/qxl/qxl_object.c
@@ -54,9 +54,14 @@ bool qxl_ttm_bo_is_qxl_bo(struct ttm_buffer_object *bo)
 void qxl_ttm_placement_from_domain(struct qxl_bo *qbo, u32 domain, bool pinned)
 {
 	u32 c = 0;
-	u32 pflag = pinned ? TTM_PL_FLAG_NO_EVICT : 0;
+	u32 pflag = 0;
 	unsigned int i;
 
+	if (pinned)
+		pflag |= TTM_PL_FLAG_NO_EVICT;
+	if (qbo->tbo.base.size <= PAGE_SIZE)
+		pflag |= TTM_PL_FLAG_TOPDOWN;
+
 	qbo->placement.placement = qbo->placements;
 	qbo->placement.busy_placement = qbo->placements;
 	if (domain == QXL_GEM_DOMAIN_VRAM)
@@ -77,6 +82,19 @@ void qxl_ttm_placement_from_domain(struct qxl_bo *qbo, u32 domain, bool pinned)
 	}
 }
 
+static const struct drm_gem_object_funcs qxl_object_funcs = {
+	.free = qxl_gem_object_free,
+	.open = qxl_gem_object_open,
+	.close = qxl_gem_object_close,
+	.pin = qxl_gem_prime_pin,
+	.unpin = qxl_gem_prime_unpin,
+	.get_sg_table = qxl_gem_prime_get_sg_table,
+	.vmap = qxl_gem_prime_vmap,
+	.vunmap = qxl_gem_prime_vunmap,
+	.mmap = drm_gem_ttm_mmap,
+	.print_info = drm_gem_ttm_print_info,
+};
+
 int qxl_bo_create(struct qxl_device *qdev,
 		  unsigned long size, bool kernel, bool pinned, u32 domain,
 		  struct qxl_surface *surf,
@@ -100,6 +118,7 @@ int qxl_bo_create(struct qxl_device *qdev,
 		kfree(bo);
 		return r;
 	}
+	bo->tbo.base.funcs = &qxl_object_funcs;
 	bo->type = domain;
 	bo->pin_count = pinned ? 1 : 0;
 	bo->surface_id = 0;
@@ -148,7 +167,6 @@ int qxl_bo_kmap(struct qxl_bo *bo, void **ptr)
 void *qxl_bo_kmap_atomic_page(struct qxl_device *qdev,
 			      struct qxl_bo *bo, int page_offset)
 {
-	struct ttm_mem_type_manager *man = &bo->tbo.bdev->man[bo->tbo.mem.mem_type];
 	void *rptr;
 	int ret;
 	struct io_mapping *map;
@@ -160,9 +178,7 @@ void *qxl_bo_kmap_atomic_page(struct qxl_device *qdev,
 	else
 		goto fallback;
 
-	(void) ttm_mem_io_lock(man, false);
-	ret = ttm_mem_io_reserve(bo->tbo.bdev, &bo->tbo.mem);
-	ttm_mem_io_unlock(man);
+	ret = qxl_ttm_io_mem_reserve(bo->tbo.bdev, &bo->tbo.mem);
 
 	return io_mapping_map_atomic_wc(map, bo->tbo.mem.bus.offset + page_offset);
 fallback:
@@ -193,17 +209,11 @@ void qxl_bo_kunmap(struct qxl_bo *bo)
 void qxl_bo_kunmap_atomic_page(struct qxl_device *qdev,
 			       struct qxl_bo *bo, void *pmap)
 {
-	struct ttm_mem_type_manager *man = &bo->tbo.bdev->man[bo->tbo.mem.mem_type];
-
 	if ((bo->tbo.mem.mem_type != TTM_PL_VRAM) &&
 	    (bo->tbo.mem.mem_type != TTM_PL_PRIV))
 		goto fallback;
 
 	io_mapping_unmap_atomic(pmap);
-
-	(void) ttm_mem_io_lock(man, false);
-	ttm_mem_io_free(bo->tbo.bdev, &bo->tbo.mem);
-	ttm_mem_io_unlock(man);
 	return;
  fallback:
 	qxl_bo_kunmap(bo);
diff --git a/drivers/gpu/drm/qxl/qxl_release.c b/drivers/gpu/drm/qxl/qxl_release.c
index 312216caeea2..2feca734c7b1 100644
--- a/drivers/gpu/drm/qxl/qxl_release.c
+++ b/drivers/gpu/drm/qxl/qxl_release.c
@@ -260,7 +260,7 @@ int qxl_release_reserve_list(struct qxl_release *release, bool no_intr)
 		return 0;
 
 	ret = ttm_eu_reserve_buffers(&release->ticket, &release->bos,
-				     !no_intr, NULL, true);
+				     !no_intr, NULL);
 	if (ret)
 		return ret;
 
@@ -429,7 +429,6 @@ void qxl_release_unmap(struct qxl_device *qdev,
 void qxl_release_fence_buffer_objects(struct qxl_release *release)
 {
 	struct ttm_buffer_object *bo;
-	struct ttm_bo_global *glob;
 	struct ttm_bo_device *bdev;
 	struct ttm_validate_buffer *entry;
 	struct qxl_device *qdev;
@@ -451,18 +450,16 @@ void qxl_release_fence_buffer_objects(struct qxl_release *release)
 		       release->id | 0xf0000000, release->base.seqno);
 	trace_dma_fence_emit(&release->base);
 
-	glob = bdev->glob;
-
-	spin_lock(&glob->lru_lock);
+	spin_lock(&ttm_bo_glob.lru_lock);
 
 	list_for_each_entry(entry, &release->bos, head) {
 		bo = entry->bo;
 
 		dma_resv_add_shared_fence(bo->base.resv, &release->base);
-		ttm_bo_add_to_lru(bo);
+		ttm_bo_move_to_lru_tail(bo, NULL);
 		dma_resv_unlock(bo->base.resv);
 	}
-	spin_unlock(&glob->lru_lock);
+	spin_unlock(&ttm_bo_glob.lru_lock);
 	ww_acquire_fini(&release->ticket);
 }
 
diff --git a/drivers/gpu/drm/qxl/qxl_ttm.c b/drivers/gpu/drm/qxl/qxl_ttm.c
index 9b24514c75aa..16a5e903533d 100644
--- a/drivers/gpu/drm/qxl/qxl_ttm.c
+++ b/drivers/gpu/drm/qxl/qxl_ttm.c
@@ -48,47 +48,6 @@ static struct qxl_device *qxl_get_qdev(struct ttm_bo_device *bdev)
 	return qdev;
 }
 
-static struct vm_operations_struct qxl_ttm_vm_ops;
-static const struct vm_operations_struct *ttm_vm_ops;
-
-static vm_fault_t qxl_ttm_fault(struct vm_fault *vmf)
-{
-	struct ttm_buffer_object *bo;
-	vm_fault_t ret;
-
-	bo = (struct ttm_buffer_object *)vmf->vma->vm_private_data;
-	if (bo == NULL)
-		return VM_FAULT_NOPAGE;
-	ret = ttm_vm_ops->fault(vmf);
-	return ret;
-}
-
-int qxl_mmap(struct file *filp, struct vm_area_struct *vma)
-{
-	int r;
-	struct drm_file *file_priv = filp->private_data;
-	struct qxl_device *qdev = file_priv->minor->dev->dev_private;
-
-	if (qdev == NULL) {
-		DRM_ERROR(
-		 "filp->private_data->minor->dev->dev_private == NULL\n");
-		return -EINVAL;
-	}
-	DRM_DEBUG_DRIVER("filp->private_data = 0x%p, vma->vm_pgoff = %lx\n",
-		  filp->private_data, vma->vm_pgoff);
-
-	r = ttm_bo_mmap(filp, vma, &qdev->mman.bdev);
-	if (unlikely(r != 0))
-		return r;
-	if (unlikely(ttm_vm_ops == NULL)) {
-		ttm_vm_ops = vma->vm_ops;
-		qxl_ttm_vm_ops = *ttm_vm_ops;
-		qxl_ttm_vm_ops.fault = &qxl_ttm_fault;
-	}
-	vma->vm_ops = &qxl_ttm_vm_ops;
-	return 0;
-}
-
 static int qxl_invalidate_caches(struct ttm_bo_device *bdev, uint32_t flags)
 {
 	return 0;
@@ -151,16 +110,8 @@ static void qxl_evict_flags(struct ttm_buffer_object *bo,
 	*placement = qbo->placement;
 }
 
-static int qxl_verify_access(struct ttm_buffer_object *bo, struct file *filp)
-{
-	struct qxl_bo *qbo = to_qxl_bo(bo);
-
-	return drm_vma_node_verify_access(&qbo->tbo.base.vma_node,
-					  filp->private_data);
-}
-
-static int qxl_ttm_io_mem_reserve(struct ttm_bo_device *bdev,
-				  struct ttm_mem_reg *mem)
+int qxl_ttm_io_mem_reserve(struct ttm_bo_device *bdev,
+			   struct ttm_mem_reg *mem)
 {
 	struct ttm_mem_type_manager *man = &bdev->man[mem->mem_type];
 	struct qxl_device *qdev = qxl_get_qdev(bdev);
@@ -310,7 +261,6 @@ static struct ttm_bo_driver qxl_bo_driver = {
 	.eviction_valuable = ttm_bo_eviction_valuable,
 	.evict_flags = &qxl_evict_flags,
 	.move = &qxl_bo_move,
-	.verify_access = &qxl_verify_access,
 	.io_mem_reserve = &qxl_ttm_io_mem_reserve,
 	.io_mem_free = &qxl_ttm_io_mem_free,
 	.move_notify = &qxl_bo_move_notify,
@@ -325,6 +275,7 @@ int qxl_ttm_init(struct qxl_device *qdev)
 	r = ttm_bo_device_init(&qdev->mman.bdev,
 			       &qxl_bo_driver,
 			       qdev->ddev.anon_inode->i_mapping,
+			       qdev->ddev.vma_offset_manager,
 			       false);
 	if (r) {
 		DRM_ERROR("failed initializing buffer object driver(%d).\n", r);
@@ -368,14 +319,11 @@ static int qxl_mm_dump_table(struct seq_file *m, void *data)
 {
 	struct drm_info_node *node = (struct drm_info_node *)m->private;
 	struct drm_mm *mm = (struct drm_mm *)node->info_ent->data;
-	struct drm_device *dev = node->minor->dev;
-	struct qxl_device *rdev = dev->dev_private;
-	struct ttm_bo_global *glob = rdev->mman.bdev.glob;
 	struct drm_printer p = drm_seq_file_printer(m);
 
-	spin_lock(&glob->lru_lock);
+	spin_lock(&ttm_bo_glob.lru_lock);
 	drm_mm_print(mm, &p);
-	spin_unlock(&glob->lru_lock);
+	spin_unlock(&ttm_bo_glob.lru_lock);
 	return 0;
 }
 #endif
diff --git a/drivers/gpu/drm/radeon/cik.c b/drivers/gpu/drm/radeon/cik.c
index 62eab82a64f9..40a7e702c2a9 100644
--- a/drivers/gpu/drm/radeon/cik.c
+++ b/drivers/gpu/drm/radeon/cik.c
@@ -221,9 +221,7 @@ int ci_get_temp(struct radeon_device *rdev)
 	else
 		actual_temp = temp & 0x1ff;
 
-	actual_temp = actual_temp * 1000;
-
-	return actual_temp;
+	return actual_temp * 1000;
 }
 
 /* get temperature in millidegrees */
@@ -239,9 +237,7 @@ int kv_get_temp(struct radeon_device *rdev)
 	else
 		actual_temp = 0;
 
-	actual_temp = actual_temp * 1000;
-
-	return actual_temp;
+	return actual_temp * 1000;
 }
 
 /*
@@ -6969,8 +6965,8 @@ static int cik_irq_init(struct radeon_device *rdev)
 	}
 
 	/* setup interrupt control */
-	/* XXX this should actually be a bus address, not an MC address. same on older asics */
-	WREG32(INTERRUPT_CNTL2, rdev->ih.gpu_addr >> 8);
+	/* set dummy read address to dummy page address */
+	WREG32(INTERRUPT_CNTL2, rdev->dummy_page.addr >> 8);
 	interrupt_cntl = RREG32(INTERRUPT_CNTL);
 	/* IH_DUMMY_RD_OVERRIDE=0 - dummy read disabled with msi, enabled without msi
 	 * IH_DUMMY_RD_OVERRIDE=1 - dummy read controlled by IH_DUMMY_RD_EN
@@ -9504,7 +9500,6 @@ static void cik_pcie_gen3_enable(struct radeon_device *rdev)
 {
 	struct pci_dev *root = rdev->pdev->bus->self;
 	enum pci_bus_speed speed_cap;
-	int bridge_pos, gpu_pos;
 	u32 speed_cntl, current_data_rate;
 	int i;
 	u16 tmp16;
@@ -9546,12 +9541,7 @@ static void cik_pcie_gen3_enable(struct radeon_device *rdev)
 		DRM_INFO("enabling PCIE gen 2 link speeds, disable with radeon.pcie_gen2=0\n");
 	}
 
-	bridge_pos = pci_pcie_cap(root);
-	if (!bridge_pos)
-		return;
-
-	gpu_pos = pci_pcie_cap(rdev->pdev);
-	if (!gpu_pos)
+	if (!pci_is_pcie(root) || !pci_is_pcie(rdev->pdev))
 		return;
 
 	if (speed_cap == PCIE_SPEED_8_0GT) {
@@ -9561,14 +9551,17 @@ static void cik_pcie_gen3_enable(struct radeon_device *rdev)
 			u16 bridge_cfg2, gpu_cfg2;
 			u32 max_lw, current_lw, tmp;
 
-			pci_read_config_word(root, bridge_pos + PCI_EXP_LNKCTL, &bridge_cfg);
-			pci_read_config_word(rdev->pdev, gpu_pos + PCI_EXP_LNKCTL, &gpu_cfg);
+			pcie_capability_read_word(root, PCI_EXP_LNKCTL,
+						  &bridge_cfg);
+			pcie_capability_read_word(rdev->pdev, PCI_EXP_LNKCTL,
+						  &gpu_cfg);
 
 			tmp16 = bridge_cfg | PCI_EXP_LNKCTL_HAWD;
-			pci_write_config_word(root, bridge_pos + PCI_EXP_LNKCTL, tmp16);
+			pcie_capability_write_word(root, PCI_EXP_LNKCTL, tmp16);
 
 			tmp16 = gpu_cfg | PCI_EXP_LNKCTL_HAWD;
-			pci_write_config_word(rdev->pdev, gpu_pos + PCI_EXP_LNKCTL, tmp16);
+			pcie_capability_write_word(rdev->pdev, PCI_EXP_LNKCTL,
+						   tmp16);
 
 			tmp = RREG32_PCIE_PORT(PCIE_LC_STATUS1);
 			max_lw = (tmp & LC_DETECTED_LINK_WIDTH_MASK) >> LC_DETECTED_LINK_WIDTH_SHIFT;
@@ -9586,15 +9579,23 @@ static void cik_pcie_gen3_enable(struct radeon_device *rdev)
 
 			for (i = 0; i < 10; i++) {
 				/* check status */
-				pci_read_config_word(rdev->pdev, gpu_pos + PCI_EXP_DEVSTA, &tmp16);
+				pcie_capability_read_word(rdev->pdev,
+							  PCI_EXP_DEVSTA,
+							  &tmp16);
 				if (tmp16 & PCI_EXP_DEVSTA_TRPND)
 					break;
 
-				pci_read_config_word(root, bridge_pos + PCI_EXP_LNKCTL, &bridge_cfg);
-				pci_read_config_word(rdev->pdev, gpu_pos + PCI_EXP_LNKCTL, &gpu_cfg);
+				pcie_capability_read_word(root, PCI_EXP_LNKCTL,
+							  &bridge_cfg);
+				pcie_capability_read_word(rdev->pdev,
+							  PCI_EXP_LNKCTL,
+							  &gpu_cfg);
 
-				pci_read_config_word(root, bridge_pos + PCI_EXP_LNKCTL2, &bridge_cfg2);
-				pci_read_config_word(rdev->pdev, gpu_pos + PCI_EXP_LNKCTL2, &gpu_cfg2);
+				pcie_capability_read_word(root, PCI_EXP_LNKCTL2,
+							  &bridge_cfg2);
+				pcie_capability_read_word(rdev->pdev,
+							  PCI_EXP_LNKCTL2,
+							  &gpu_cfg2);
 
 				tmp = RREG32_PCIE_PORT(PCIE_LC_CNTL4);
 				tmp |= LC_SET_QUIESCE;
@@ -9607,26 +9608,45 @@ static void cik_pcie_gen3_enable(struct radeon_device *rdev)
 				msleep(100);
 
 				/* linkctl */
-				pci_read_config_word(root, bridge_pos + PCI_EXP_LNKCTL, &tmp16);
+				pcie_capability_read_word(root, PCI_EXP_LNKCTL,
+							  &tmp16);
 				tmp16 &= ~PCI_EXP_LNKCTL_HAWD;
 				tmp16 |= (bridge_cfg & PCI_EXP_LNKCTL_HAWD);
-				pci_write_config_word(root, bridge_pos + PCI_EXP_LNKCTL, tmp16);
+				pcie_capability_write_word(root, PCI_EXP_LNKCTL,
+							   tmp16);
 
-				pci_read_config_word(rdev->pdev, gpu_pos + PCI_EXP_LNKCTL, &tmp16);
+				pcie_capability_read_word(rdev->pdev,
+							  PCI_EXP_LNKCTL,
+							  &tmp16);
 				tmp16 &= ~PCI_EXP_LNKCTL_HAWD;
 				tmp16 |= (gpu_cfg & PCI_EXP_LNKCTL_HAWD);
-				pci_write_config_word(rdev->pdev, gpu_pos + PCI_EXP_LNKCTL, tmp16);
+				pcie_capability_write_word(rdev->pdev,
+							   PCI_EXP_LNKCTL,
+							   tmp16);
 
 				/* linkctl2 */
-				pci_read_config_word(root, bridge_pos + PCI_EXP_LNKCTL2, &tmp16);
-				tmp16 &= ~((1 << 4) | (7 << 9));
-				tmp16 |= (bridge_cfg2 & ((1 << 4) | (7 << 9)));
-				pci_write_config_word(root, bridge_pos + PCI_EXP_LNKCTL2, tmp16);
-
-				pci_read_config_word(rdev->pdev, gpu_pos + PCI_EXP_LNKCTL2, &tmp16);
-				tmp16 &= ~((1 << 4) | (7 << 9));
-				tmp16 |= (gpu_cfg2 & ((1 << 4) | (7 << 9)));
-				pci_write_config_word(rdev->pdev, gpu_pos + PCI_EXP_LNKCTL2, tmp16);
+				pcie_capability_read_word(root, PCI_EXP_LNKCTL2,
+							  &tmp16);
+				tmp16 &= ~(PCI_EXP_LNKCTL2_ENTER_COMP |
+					   PCI_EXP_LNKCTL2_TX_MARGIN);
+				tmp16 |= (bridge_cfg2 &
+					  (PCI_EXP_LNKCTL2_ENTER_COMP |
+					   PCI_EXP_LNKCTL2_TX_MARGIN));
+				pcie_capability_write_word(root,
+							   PCI_EXP_LNKCTL2,
+							   tmp16);
+
+				pcie_capability_read_word(rdev->pdev,
+							  PCI_EXP_LNKCTL2,
+							  &tmp16);
+				tmp16 &= ~(PCI_EXP_LNKCTL2_ENTER_COMP |
+					   PCI_EXP_LNKCTL2_TX_MARGIN);
+				tmp16 |= (gpu_cfg2 &
+					  (PCI_EXP_LNKCTL2_ENTER_COMP |
+					   PCI_EXP_LNKCTL2_TX_MARGIN));
+				pcie_capability_write_word(rdev->pdev,
+							   PCI_EXP_LNKCTL2,
+							   tmp16);
 
 				tmp = RREG32_PCIE_PORT(PCIE_LC_CNTL4);
 				tmp &= ~LC_SET_QUIESCE;
@@ -9640,15 +9660,15 @@ static void cik_pcie_gen3_enable(struct radeon_device *rdev)
 	speed_cntl &= ~LC_FORCE_DIS_SW_SPEED_CHANGE;
 	WREG32_PCIE_PORT(PCIE_LC_SPEED_CNTL, speed_cntl);
 
-	pci_read_config_word(rdev->pdev, gpu_pos + PCI_EXP_LNKCTL2, &tmp16);
-	tmp16 &= ~0xf;
+	pcie_capability_read_word(rdev->pdev, PCI_EXP_LNKCTL2, &tmp16);
+	tmp16 &= ~PCI_EXP_LNKCTL2_TLS;
 	if (speed_cap == PCIE_SPEED_8_0GT)
-		tmp16 |= 3; /* gen3 */
+		tmp16 |= PCI_EXP_LNKCTL2_TLS_8_0GT; /* gen3 */
 	else if (speed_cap == PCIE_SPEED_5_0GT)
-		tmp16 |= 2; /* gen2 */
+		tmp16 |= PCI_EXP_LNKCTL2_TLS_5_0GT; /* gen2 */
 	else
-		tmp16 |= 1; /* gen1 */
-	pci_write_config_word(rdev->pdev, gpu_pos + PCI_EXP_LNKCTL2, tmp16);
+		tmp16 |= PCI_EXP_LNKCTL2_TLS_2_5GT; /* gen1 */
+	pcie_capability_write_word(rdev->pdev, PCI_EXP_LNKCTL2, tmp16);
 
 	speed_cntl = RREG32_PCIE_PORT(PCIE_LC_SPEED_CNTL);
 	speed_cntl |= LC_INITIATE_LINK_SPEED_CHANGE;
diff --git a/drivers/gpu/drm/radeon/r600.c b/drivers/gpu/drm/radeon/r600.c
index e937cc01910d..033bc466a862 100644
--- a/drivers/gpu/drm/radeon/r600.c
+++ b/drivers/gpu/drm/radeon/r600.c
@@ -3696,8 +3696,8 @@ int r600_irq_init(struct radeon_device *rdev)
 	}
 
 	/* setup interrupt control */
-	/* set dummy read address to ring address */
-	WREG32(INTERRUPT_CNTL2, rdev->ih.gpu_addr >> 8);
+	/* set dummy read address to dummy page address */
+	WREG32(INTERRUPT_CNTL2, rdev->dummy_page.addr >> 8);
 	interrupt_cntl = RREG32(INTERRUPT_CNTL);
 	/* IH_DUMMY_RD_OVERRIDE=0 - dummy read disabled with msi, enabled without msi
 	 * IH_DUMMY_RD_OVERRIDE=1 - dummy read controlled by IH_DUMMY_RD_EN
diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h
index d59b004f6695..30e32adc1fc6 100644
--- a/drivers/gpu/drm/radeon/radeon.h
+++ b/drivers/gpu/drm/radeon/radeon.h
@@ -68,6 +68,10 @@
 #include <linux/hashtable.h>
 #include <linux/dma-fence.h>
 
+#ifdef CONFIG_MMU_NOTIFIER
+#include <linux/mmu_notifier.h>
+#endif
+
 #include <drm/ttm/ttm_bo_api.h>
 #include <drm/ttm/ttm_bo_driver.h>
 #include <drm/ttm/ttm_placement.h>
@@ -509,8 +513,9 @@ struct radeon_bo {
 	struct ttm_bo_kmap_obj		dma_buf_vmap;
 	pid_t				pid;
 
-	struct radeon_mn		*mn;
-	struct list_head		mn_list;
+#ifdef CONFIG_MMU_NOTIFIER
+	struct mmu_interval_notifier	notifier;
+#endif
 };
 #define gem_to_radeon_bo(gobj) container_of((gobj), struct radeon_bo, tbo.base)
 
diff --git a/drivers/gpu/drm/radeon/radeon_audio.c b/drivers/gpu/drm/radeon/radeon_audio.c
index b9aea5776d3d..72db2b41e96d 100644
--- a/drivers/gpu/drm/radeon/radeon_audio.c
+++ b/drivers/gpu/drm/radeon/radeon_audio.c
@@ -367,10 +367,10 @@ static void radeon_audio_write_sad_regs(struct drm_encoder *encoder)
 		return;
 
 	sad_count = drm_edid_to_sad(radeon_connector_edid(connector), &sads);
-	if (sad_count <= 0) {
+	if (sad_count < 0)
 		DRM_ERROR("Couldn't read SADs: %d\n", sad_count);
+	if (sad_count <= 0)
 		return;
-	}
 	BUG_ON(!sads);
 
 	if (radeon_encoder->audio && radeon_encoder->audio->write_sad_regs)
diff --git a/drivers/gpu/drm/radeon/radeon_connectors.c b/drivers/gpu/drm/radeon/radeon_connectors.c
index b684cd719612..c07427d3c199 100644
--- a/drivers/gpu/drm/radeon/radeon_connectors.c
+++ b/drivers/gpu/drm/radeon/radeon_connectors.c
@@ -249,11 +249,10 @@ radeon_connector_update_scratch_regs(struct drm_connector *connector, enum drm_c
 	struct drm_encoder *encoder;
 	const struct drm_connector_helper_funcs *connector_funcs = connector->helper_private;
 	bool connected;
-	int i;
 
 	best_encoder = connector_funcs->best_encoder(connector);
 
-	drm_connector_for_each_possible_encoder(connector, encoder, i) {
+	drm_connector_for_each_possible_encoder(connector, encoder) {
 		if ((encoder == best_encoder) && (status == connector_status_connected))
 			connected = true;
 		else
@@ -269,9 +268,8 @@ radeon_connector_update_scratch_regs(struct drm_connector *connector, enum drm_c
 static struct drm_encoder *radeon_find_encoder(struct drm_connector *connector, int encoder_type)
 {
 	struct drm_encoder *encoder;
-	int i;
 
-	drm_connector_for_each_possible_encoder(connector, encoder, i) {
+	drm_connector_for_each_possible_encoder(connector, encoder) {
 		if (encoder->encoder_type == encoder_type)
 			return encoder;
 	}
@@ -380,10 +378,9 @@ static int radeon_ddc_get_modes(struct drm_connector *connector)
 static struct drm_encoder *radeon_best_single_encoder(struct drm_connector *connector)
 {
 	struct drm_encoder *encoder;
-	int i;
 
 	/* pick the first one */
-	drm_connector_for_each_possible_encoder(connector, encoder, i)
+	drm_connector_for_each_possible_encoder(connector, encoder)
 		return encoder;
 
 	return NULL;
@@ -428,14 +425,13 @@ radeon_connector_analog_encoder_conflict_solve(struct drm_connector *connector,
 
 	list_for_each_entry(conflict, &dev->mode_config.connector_list, head) {
 		struct drm_encoder *enc;
-		int i;
 
 		if (conflict == connector)
 			continue;
 
 		radeon_conflict = to_radeon_connector(conflict);
 
-		drm_connector_for_each_possible_encoder(conflict, enc, i) {
+		drm_connector_for_each_possible_encoder(conflict, enc) {
 			/* if the IDs match */
 			if (enc == encoder) {
 				if (conflict->status != connector_status_connected)
@@ -1363,9 +1359,7 @@ radeon_dvi_detect(struct drm_connector *connector, bool force)
 
 	/* find analog encoder */
 	if (radeon_connector->dac_load_detect) {
-		int i;
-
-		drm_connector_for_each_possible_encoder(connector, encoder, i) {
+		drm_connector_for_each_possible_encoder(connector, encoder) {
 			if (encoder->encoder_type != DRM_MODE_ENCODER_DAC &&
 			    encoder->encoder_type != DRM_MODE_ENCODER_TVDAC)
 				continue;
@@ -1443,9 +1437,8 @@ static struct drm_encoder *radeon_dvi_encoder(struct drm_connector *connector)
 {
 	struct radeon_connector *radeon_connector = to_radeon_connector(connector);
 	struct drm_encoder *encoder;
-	int i;
 
-	drm_connector_for_each_possible_encoder(connector, encoder, i) {
+	drm_connector_for_each_possible_encoder(connector, encoder) {
 		if (radeon_connector->use_digital == true) {
 			if (encoder->encoder_type == DRM_MODE_ENCODER_TMDS)
 				return encoder;
@@ -1460,7 +1453,7 @@ static struct drm_encoder *radeon_dvi_encoder(struct drm_connector *connector)
 
 	/* then check use digitial */
 	/* pick the first one */
-	drm_connector_for_each_possible_encoder(connector, encoder, i)
+	drm_connector_for_each_possible_encoder(connector, encoder)
 		return encoder;
 
 	return NULL;
@@ -1603,9 +1596,8 @@ u16 radeon_connector_encoder_get_dp_bridge_encoder_id(struct drm_connector *conn
 {
 	struct drm_encoder *encoder;
 	struct radeon_encoder *radeon_encoder;
-	int i;
 
-	drm_connector_for_each_possible_encoder(connector, encoder, i) {
+	drm_connector_for_each_possible_encoder(connector, encoder) {
 		radeon_encoder = to_radeon_encoder(encoder);
 
 		switch (radeon_encoder->encoder_id) {
@@ -1624,10 +1616,9 @@ static bool radeon_connector_encoder_is_hbr2(struct drm_connector *connector)
 {
 	struct drm_encoder *encoder;
 	struct radeon_encoder *radeon_encoder;
-	int i;
 	bool found = false;
 
-	drm_connector_for_each_possible_encoder(connector, encoder, i) {
+	drm_connector_for_each_possible_encoder(connector, encoder) {
 		radeon_encoder = to_radeon_encoder(encoder);
 		if (radeon_encoder->caps & ATOM_ENCODER_CAP_RECORD_HBR2)
 			found = true;
diff --git a/drivers/gpu/drm/radeon/radeon_dp_mst.c b/drivers/gpu/drm/radeon/radeon_dp_mst.c
index 2994f07fbad9..ee28f5b3785e 100644
--- a/drivers/gpu/drm/radeon/radeon_dp_mst.c
+++ b/drivers/gpu/drm/radeon/radeon_dp_mst.c
@@ -233,21 +233,26 @@ drm_encoder *radeon_mst_best_encoder(struct drm_connector *connector)
 	return &radeon_connector->mst_encoder->base;
 }
 
+static int
+radeon_dp_mst_detect(struct drm_connector *connector,
+		     struct drm_modeset_acquire_ctx *ctx,
+		     bool force)
+{
+	struct radeon_connector *radeon_connector =
+		to_radeon_connector(connector);
+	struct radeon_connector *master = radeon_connector->mst_port;
+
+	return drm_dp_mst_detect_port(connector, ctx, &master->mst_mgr,
+				      radeon_connector->port);
+}
+
 static const struct drm_connector_helper_funcs radeon_dp_mst_connector_helper_funcs = {
 	.get_modes = radeon_dp_mst_get_modes,
 	.mode_valid = radeon_dp_mst_mode_valid,
 	.best_encoder = radeon_mst_best_encoder,
+	.detect_ctx = radeon_dp_mst_detect,
 };
 
-static enum drm_connector_status
-radeon_dp_mst_detect(struct drm_connector *connector, bool force)
-{
-	struct radeon_connector *radeon_connector = to_radeon_connector(connector);
-	struct radeon_connector *master = radeon_connector->mst_port;
-
-	return drm_dp_mst_detect_port(connector, &master->mst_mgr, radeon_connector->port);
-}
-
 static void
 radeon_dp_mst_connector_destroy(struct drm_connector *connector)
 {
@@ -262,7 +267,6 @@ radeon_dp_mst_connector_destroy(struct drm_connector *connector)
 
 static const struct drm_connector_funcs radeon_dp_mst_connector_funcs = {
 	.dpms = drm_helper_connector_dpms,
-	.detect = radeon_dp_mst_detect,
 	.fill_modes = drm_helper_probe_single_connector_modes,
 	.destroy = radeon_dp_mst_connector_destroy,
 };
diff --git a/drivers/gpu/drm/radeon/radeon_drv.c b/drivers/gpu/drm/radeon/radeon_drv.c
index d0bc91ed7c90..fd74e2611185 100644
--- a/drivers/gpu/drm/radeon/radeon_drv.c
+++ b/drivers/gpu/drm/radeon/radeon_drv.c
@@ -361,7 +361,7 @@ static int radeon_pci_probe(struct pci_dev *pdev,
 		return -EPROBE_DEFER;
 
 	/* Get rid of things like offb */
-	ret = drm_fb_helper_remove_conflicting_pci_framebuffers(pdev, 0, "radeondrmfb");
+	ret = drm_fb_helper_remove_conflicting_pci_framebuffers(pdev, "radeondrmfb");
 	if (ret)
 		return ret;
 
@@ -379,19 +379,22 @@ radeon_pci_remove(struct pci_dev *pdev)
 static void
 radeon_pci_shutdown(struct pci_dev *pdev)
 {
-	struct drm_device *ddev = pci_get_drvdata(pdev);
-
 	/* if we are running in a VM, make sure the device
 	 * torn down properly on reboot/shutdown
 	 */
 	if (radeon_device_is_virtual())
 		radeon_pci_remove(pdev);
 
-	/* Some adapters need to be suspended before a
-	* shutdown occurs in order to prevent an error
-	* during kexec.
-	*/
-	radeon_suspend_kms(ddev, true, true, false);
+#ifdef CONFIG_PPC64
+	/*
+	 * Some adapters need to be suspended before a
+	 * shutdown occurs in order to prevent an error
+	 * during kexec.
+	 * Make this power specific becauase it breaks
+	 * some non-power boards.
+	 */
+	radeon_suspend_kms(pci_get_drvdata(pdev), true, true, false);
+#endif
 }
 
 static int radeon_pmops_suspend(struct device *dev)
diff --git a/drivers/gpu/drm/radeon/radeon_gem.c b/drivers/gpu/drm/radeon/radeon_gem.c
index b2b076606f54..67298a0739cb 100644
--- a/drivers/gpu/drm/radeon/radeon_gem.c
+++ b/drivers/gpu/drm/radeon/radeon_gem.c
@@ -566,7 +566,7 @@ static void radeon_gem_va_update_vm(struct radeon_device *rdev,
 	if (!vm_bos)
 		return;
 
-	r = ttm_eu_reserve_buffers(&ticket, &list, true, NULL, true);
+	r = ttm_eu_reserve_buffers(&ticket, &list, true, NULL);
 	if (r)
 		goto error_free;
 
diff --git a/drivers/gpu/drm/radeon/radeon_mn.c b/drivers/gpu/drm/radeon/radeon_mn.c
index dbab9a3a969b..f93829f08a4d 100644
--- a/drivers/gpu/drm/radeon/radeon_mn.c
+++ b/drivers/gpu/drm/radeon/radeon_mn.c
@@ -36,131 +36,51 @@
 
 #include "radeon.h"
 
-struct radeon_mn {
-	struct mmu_notifier	mn;
-
-	/* objects protected by lock */
-	struct mutex		lock;
-	struct rb_root_cached	objects;
-};
-
-struct radeon_mn_node {
-	struct interval_tree_node	it;
-	struct list_head		bos;
-};
-
 /**
- * radeon_mn_invalidate_range_start - callback to notify about mm change
+ * radeon_mn_invalidate - callback to notify about mm change
  *
  * @mn: our notifier
- * @mn: the mm this callback is about
- * @start: start of updated range
- * @end: end of updated range
+ * @range: the VMA under invalidation
  *
  * We block for all BOs between start and end to be idle and
  * unmap them by move them into system domain again.
  */
-static int radeon_mn_invalidate_range_start(struct mmu_notifier *mn,
-				const struct mmu_notifier_range *range)
+static bool radeon_mn_invalidate(struct mmu_interval_notifier *mn,
+				 const struct mmu_notifier_range *range,
+				 unsigned long cur_seq)
 {
-	struct radeon_mn *rmn = container_of(mn, struct radeon_mn, mn);
+	struct radeon_bo *bo = container_of(mn, struct radeon_bo, notifier);
 	struct ttm_operation_ctx ctx = { false, false };
-	struct interval_tree_node *it;
-	unsigned long end;
-	int ret = 0;
-
-	/* notification is exclusive, but interval is inclusive */
-	end = range->end - 1;
-
-	/* TODO we should be able to split locking for interval tree and
-	 * the tear down.
-	 */
-	if (mmu_notifier_range_blockable(range))
-		mutex_lock(&rmn->lock);
-	else if (!mutex_trylock(&rmn->lock))
-		return -EAGAIN;
-
-	it = interval_tree_iter_first(&rmn->objects, range->start, end);
-	while (it) {
-		struct radeon_mn_node *node;
-		struct radeon_bo *bo;
-		long r;
-
-		if (!mmu_notifier_range_blockable(range)) {
-			ret = -EAGAIN;
-			goto out_unlock;
-		}
-
-		node = container_of(it, struct radeon_mn_node, it);
-		it = interval_tree_iter_next(it, range->start, end);
+	long r;
 
-		list_for_each_entry(bo, &node->bos, mn_list) {
+	if (!bo->tbo.ttm || bo->tbo.ttm->state != tt_bound)
+		return true;
 
-			if (!bo->tbo.ttm || bo->tbo.ttm->state != tt_bound)
-				continue;
+	if (!mmu_notifier_range_blockable(range))
+		return false;
 
-			r = radeon_bo_reserve(bo, true);
-			if (r) {
-				DRM_ERROR("(%ld) failed to reserve user bo\n", r);
-				continue;
-			}
-
-			r = dma_resv_wait_timeout_rcu(bo->tbo.base.resv,
-				true, false, MAX_SCHEDULE_TIMEOUT);
-			if (r <= 0)
-				DRM_ERROR("(%ld) failed to wait for user bo\n", r);
-
-			radeon_ttm_placement_from_domain(bo, RADEON_GEM_DOMAIN_CPU);
-			r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx);
-			if (r)
-				DRM_ERROR("(%ld) failed to validate user bo\n", r);
-
-			radeon_bo_unreserve(bo);
-		}
+	r = radeon_bo_reserve(bo, true);
+	if (r) {
+		DRM_ERROR("(%ld) failed to reserve user bo\n", r);
+		return true;
 	}
-	
-out_unlock:
-	mutex_unlock(&rmn->lock);
-
-	return ret;
-}
-
-static void radeon_mn_release(struct mmu_notifier *mn, struct mm_struct *mm)
-{
-	struct mmu_notifier_range range = {
-		.mm = mm,
-		.start = 0,
-		.end = ULONG_MAX,
-		.flags = 0,
-		.event = MMU_NOTIFY_UNMAP,
-	};
-
-	radeon_mn_invalidate_range_start(mn, &range);
-}
-
-static struct mmu_notifier *radeon_mn_alloc_notifier(struct mm_struct *mm)
-{
-	struct radeon_mn *rmn;
 
-	rmn = kzalloc(sizeof(*rmn), GFP_KERNEL);
-	if (!rmn)
-		return ERR_PTR(-ENOMEM);
+	r = dma_resv_wait_timeout_rcu(bo->tbo.base.resv, true, false,
+				      MAX_SCHEDULE_TIMEOUT);
+	if (r <= 0)
+		DRM_ERROR("(%ld) failed to wait for user bo\n", r);
 
-	mutex_init(&rmn->lock);
-	rmn->objects = RB_ROOT_CACHED;
-	return &rmn->mn;
-}
+	radeon_ttm_placement_from_domain(bo, RADEON_GEM_DOMAIN_CPU);
+	r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx);
+	if (r)
+		DRM_ERROR("(%ld) failed to validate user bo\n", r);
 
-static void radeon_mn_free_notifier(struct mmu_notifier *mn)
-{
-	kfree(container_of(mn, struct radeon_mn, mn));
+	radeon_bo_unreserve(bo);
+	return true;
 }
 
-static const struct mmu_notifier_ops radeon_mn_ops = {
-	.release = radeon_mn_release,
-	.invalidate_range_start = radeon_mn_invalidate_range_start,
-	.alloc_notifier = radeon_mn_alloc_notifier,
-	.free_notifier = radeon_mn_free_notifier,
+static const struct mmu_interval_notifier_ops radeon_mn_ops = {
+	.invalidate = radeon_mn_invalidate,
 };
 
 /**
@@ -174,51 +94,20 @@ static const struct mmu_notifier_ops radeon_mn_ops = {
  */
 int radeon_mn_register(struct radeon_bo *bo, unsigned long addr)
 {
-	unsigned long end = addr + radeon_bo_size(bo) - 1;
-	struct mmu_notifier *mn;
-	struct radeon_mn *rmn;
-	struct radeon_mn_node *node = NULL;
-	struct list_head bos;
-	struct interval_tree_node *it;
-
-	mn = mmu_notifier_get(&radeon_mn_ops, current->mm);
-	if (IS_ERR(mn))
-		return PTR_ERR(mn);
-	rmn = container_of(mn, struct radeon_mn, mn);
-
-	INIT_LIST_HEAD(&bos);
-
-	mutex_lock(&rmn->lock);
-
-	while ((it = interval_tree_iter_first(&rmn->objects, addr, end))) {
-		kfree(node);
-		node = container_of(it, struct radeon_mn_node, it);
-		interval_tree_remove(&node->it, &rmn->objects);
-		addr = min(it->start, addr);
-		end = max(it->last, end);
-		list_splice(&node->bos, &bos);
-	}
-
-	if (!node) {
-		node = kmalloc(sizeof(struct radeon_mn_node), GFP_KERNEL);
-		if (!node) {
-			mutex_unlock(&rmn->lock);
-			return -ENOMEM;
-		}
-	}
-
-	bo->mn = rmn;
-
-	node->it.start = addr;
-	node->it.last = end;
-	INIT_LIST_HEAD(&node->bos);
-	list_splice(&bos, &node->bos);
-	list_add(&bo->mn_list, &node->bos);
-
-	interval_tree_insert(&node->it, &rmn->objects);
-
-	mutex_unlock(&rmn->lock);
-
+	int ret;
+
+	ret = mmu_interval_notifier_insert(&bo->notifier, current->mm, addr,
+					   radeon_bo_size(bo), &radeon_mn_ops);
+	if (ret)
+		return ret;
+
+	/*
+	 * FIXME: radeon appears to allow get_user_pages to run during
+	 * invalidate_range_start/end, which is not a safe way to read the
+	 * PTEs. It should use the mmu_interval_read_begin() scheme around the
+	 * get_user_pages to ensure that the PTEs are read properly
+	 */
+	mmu_interval_read_begin(&bo->notifier);
 	return 0;
 }
 
@@ -231,27 +120,8 @@ int radeon_mn_register(struct radeon_bo *bo, unsigned long addr)
  */
 void radeon_mn_unregister(struct radeon_bo *bo)
 {
-	struct radeon_mn *rmn = bo->mn;
-	struct list_head *head;
-
-	if (!rmn)
+	if (!bo->notifier.mm)
 		return;
-
-	mutex_lock(&rmn->lock);
-	/* save the next list entry for later */
-	head = bo->mn_list.next;
-
-	list_del(&bo->mn_list);
-
-	if (list_empty(head)) {
-		struct radeon_mn_node *node;
-		node = container_of(head, struct radeon_mn_node, bos);
-		interval_tree_remove(&node->it, &rmn->objects);
-		kfree(node);
-	}
-
-	mutex_unlock(&rmn->lock);
-
-	mmu_notifier_put(&rmn->mn);
-	bo->mn = NULL;
+	mmu_interval_notifier_remove(&bo->notifier);
+	bo->notifier.mm = NULL;
 }
diff --git a/drivers/gpu/drm/radeon/radeon_object.c b/drivers/gpu/drm/radeon/radeon_object.c
index 2abe1eab471f..140d94cc080d 100644
--- a/drivers/gpu/drm/radeon/radeon_object.c
+++ b/drivers/gpu/drm/radeon/radeon_object.c
@@ -542,7 +542,7 @@ int radeon_bo_list_validate(struct radeon_device *rdev,
 	u64 bytes_moved_threshold = radeon_bo_get_threshold_for_moves(rdev);
 
 	INIT_LIST_HEAD(&duplicates);
-	r = ttm_eu_reserve_buffers(ticket, head, true, &duplicates, true);
+	r = ttm_eu_reserve_buffers(ticket, head, true, &duplicates);
 	if (unlikely(r != 0)) {
 		return r;
 	}
diff --git a/drivers/gpu/drm/radeon/radeon_ttm.c b/drivers/gpu/drm/radeon/radeon_ttm.c
index a05e10724d46..098bc9f40b98 100644
--- a/drivers/gpu/drm/radeon/radeon_ttm.c
+++ b/drivers/gpu/drm/radeon/radeon_ttm.c
@@ -794,6 +794,7 @@ int radeon_ttm_init(struct radeon_device *rdev)
 	r = ttm_bo_device_init(&rdev->mman.bdev,
 			       &radeon_bo_driver,
 			       rdev->ddev->anon_inode->i_mapping,
+			       rdev->ddev->vma_offset_manager,
 			       dma_addressing_limited(&rdev->pdev->dev));
 	if (r) {
 		DRM_ERROR("failed initializing buffer object driver(%d).\n", r);
diff --git a/drivers/gpu/drm/radeon/si.c b/drivers/gpu/drm/radeon/si.c
index 05894d198a79..d7eea75b2c27 100644
--- a/drivers/gpu/drm/radeon/si.c
+++ b/drivers/gpu/drm/radeon/si.c
@@ -3257,7 +3257,7 @@ static void si_gpu_init(struct radeon_device *rdev)
 		/* XXX what about 12? */
 		rdev->config.si.tile_config |= (3 << 0);
 		break;
-	}	
+	}
 	switch ((mc_arb_ramcfg & NOOFBANK_MASK) >> NOOFBANK_SHIFT) {
 	case 0: /* four banks */
 		rdev->config.si.tile_config |= 0 << 4;
@@ -5997,8 +5997,8 @@ static int si_irq_init(struct radeon_device *rdev)
 	}
 
 	/* setup interrupt control */
-	/* set dummy read address to ring address */
-	WREG32(INTERRUPT_CNTL2, rdev->ih.gpu_addr >> 8);
+	/* set dummy read address to dummy page address */
+	WREG32(INTERRUPT_CNTL2, rdev->dummy_page.addr >> 8);
 	interrupt_cntl = RREG32(INTERRUPT_CNTL);
 	/* IH_DUMMY_RD_OVERRIDE=0 - dummy read disabled with msi, enabled without msi
 	 * IH_DUMMY_RD_OVERRIDE=1 - dummy read controlled by IH_DUMMY_RD_EN
@@ -7087,7 +7087,6 @@ static void si_pcie_gen3_enable(struct radeon_device *rdev)
 {
 	struct pci_dev *root = rdev->pdev->bus->self;
 	enum pci_bus_speed speed_cap;
-	int bridge_pos, gpu_pos;
 	u32 speed_cntl, current_data_rate;
 	int i;
 	u16 tmp16;
@@ -7129,12 +7128,7 @@ static void si_pcie_gen3_enable(struct radeon_device *rdev)
 		DRM_INFO("enabling PCIE gen 2 link speeds, disable with radeon.pcie_gen2=0\n");
 	}
 
-	bridge_pos = pci_pcie_cap(root);
-	if (!bridge_pos)
-		return;
-
-	gpu_pos = pci_pcie_cap(rdev->pdev);
-	if (!gpu_pos)
+	if (!pci_is_pcie(root) || !pci_is_pcie(rdev->pdev))
 		return;
 
 	if (speed_cap == PCIE_SPEED_8_0GT) {
@@ -7144,14 +7138,17 @@ static void si_pcie_gen3_enable(struct radeon_device *rdev)
 			u16 bridge_cfg2, gpu_cfg2;
 			u32 max_lw, current_lw, tmp;
 
-			pci_read_config_word(root, bridge_pos + PCI_EXP_LNKCTL, &bridge_cfg);
-			pci_read_config_word(rdev->pdev, gpu_pos + PCI_EXP_LNKCTL, &gpu_cfg);
+			pcie_capability_read_word(root, PCI_EXP_LNKCTL,
+						  &bridge_cfg);
+			pcie_capability_read_word(rdev->pdev, PCI_EXP_LNKCTL,
+						  &gpu_cfg);
 
 			tmp16 = bridge_cfg | PCI_EXP_LNKCTL_HAWD;
-			pci_write_config_word(root, bridge_pos + PCI_EXP_LNKCTL, tmp16);
+			pcie_capability_write_word(root, PCI_EXP_LNKCTL, tmp16);
 
 			tmp16 = gpu_cfg | PCI_EXP_LNKCTL_HAWD;
-			pci_write_config_word(rdev->pdev, gpu_pos + PCI_EXP_LNKCTL, tmp16);
+			pcie_capability_write_word(rdev->pdev, PCI_EXP_LNKCTL,
+						   tmp16);
 
 			tmp = RREG32_PCIE(PCIE_LC_STATUS1);
 			max_lw = (tmp & LC_DETECTED_LINK_WIDTH_MASK) >> LC_DETECTED_LINK_WIDTH_SHIFT;
@@ -7169,15 +7166,23 @@ static void si_pcie_gen3_enable(struct radeon_device *rdev)
 
 			for (i = 0; i < 10; i++) {
 				/* check status */
-				pci_read_config_word(rdev->pdev, gpu_pos + PCI_EXP_DEVSTA, &tmp16);
+				pcie_capability_read_word(rdev->pdev,
+							  PCI_EXP_DEVSTA,
+							  &tmp16);
 				if (tmp16 & PCI_EXP_DEVSTA_TRPND)
 					break;
 
-				pci_read_config_word(root, bridge_pos + PCI_EXP_LNKCTL, &bridge_cfg);
-				pci_read_config_word(rdev->pdev, gpu_pos + PCI_EXP_LNKCTL, &gpu_cfg);
+				pcie_capability_read_word(root, PCI_EXP_LNKCTL,
+							  &bridge_cfg);
+				pcie_capability_read_word(rdev->pdev,
+							  PCI_EXP_LNKCTL,
+							  &gpu_cfg);
 
-				pci_read_config_word(root, bridge_pos + PCI_EXP_LNKCTL2, &bridge_cfg2);
-				pci_read_config_word(rdev->pdev, gpu_pos + PCI_EXP_LNKCTL2, &gpu_cfg2);
+				pcie_capability_read_word(root, PCI_EXP_LNKCTL2,
+							  &bridge_cfg2);
+				pcie_capability_read_word(rdev->pdev,
+							  PCI_EXP_LNKCTL2,
+							  &gpu_cfg2);
 
 				tmp = RREG32_PCIE_PORT(PCIE_LC_CNTL4);
 				tmp |= LC_SET_QUIESCE;
@@ -7190,26 +7195,46 @@ static void si_pcie_gen3_enable(struct radeon_device *rdev)
 				msleep(100);
 
 				/* linkctl */
-				pci_read_config_word(root, bridge_pos + PCI_EXP_LNKCTL, &tmp16);
+				pcie_capability_read_word(root, PCI_EXP_LNKCTL,
+							  &tmp16);
 				tmp16 &= ~PCI_EXP_LNKCTL_HAWD;
 				tmp16 |= (bridge_cfg & PCI_EXP_LNKCTL_HAWD);
-				pci_write_config_word(root, bridge_pos + PCI_EXP_LNKCTL, tmp16);
+				pcie_capability_write_word(root,
+							   PCI_EXP_LNKCTL,
+							   tmp16);
 
-				pci_read_config_word(rdev->pdev, gpu_pos + PCI_EXP_LNKCTL, &tmp16);
+				pcie_capability_read_word(rdev->pdev,
+							  PCI_EXP_LNKCTL,
+							  &tmp16);
 				tmp16 &= ~PCI_EXP_LNKCTL_HAWD;
 				tmp16 |= (gpu_cfg & PCI_EXP_LNKCTL_HAWD);
-				pci_write_config_word(rdev->pdev, gpu_pos + PCI_EXP_LNKCTL, tmp16);
+				pcie_capability_write_word(rdev->pdev,
+							   PCI_EXP_LNKCTL,
+							   tmp16);
 
 				/* linkctl2 */
-				pci_read_config_word(root, bridge_pos + PCI_EXP_LNKCTL2, &tmp16);
-				tmp16 &= ~((1 << 4) | (7 << 9));
-				tmp16 |= (bridge_cfg2 & ((1 << 4) | (7 << 9)));
-				pci_write_config_word(root, bridge_pos + PCI_EXP_LNKCTL2, tmp16);
-
-				pci_read_config_word(rdev->pdev, gpu_pos + PCI_EXP_LNKCTL2, &tmp16);
-				tmp16 &= ~((1 << 4) | (7 << 9));
-				tmp16 |= (gpu_cfg2 & ((1 << 4) | (7 << 9)));
-				pci_write_config_word(rdev->pdev, gpu_pos + PCI_EXP_LNKCTL2, tmp16);
+				pcie_capability_read_word(root, PCI_EXP_LNKCTL2,
+							  &tmp16);
+				tmp16 &= ~(PCI_EXP_LNKCTL2_ENTER_COMP |
+					   PCI_EXP_LNKCTL2_TX_MARGIN);
+				tmp16 |= (bridge_cfg2 &
+					  (PCI_EXP_LNKCTL2_ENTER_COMP |
+					   PCI_EXP_LNKCTL2_TX_MARGIN));
+				pcie_capability_write_word(root,
+							   PCI_EXP_LNKCTL2,
+							   tmp16);
+
+				pcie_capability_read_word(rdev->pdev,
+							  PCI_EXP_LNKCTL2,
+							  &tmp16);
+				tmp16 &= ~(PCI_EXP_LNKCTL2_ENTER_COMP |
+					   PCI_EXP_LNKCTL2_TX_MARGIN);
+				tmp16 |= (gpu_cfg2 &
+					  (PCI_EXP_LNKCTL2_ENTER_COMP |
+					   PCI_EXP_LNKCTL2_TX_MARGIN));
+				pcie_capability_write_word(rdev->pdev,
+							   PCI_EXP_LNKCTL2,
+							   tmp16);
 
 				tmp = RREG32_PCIE_PORT(PCIE_LC_CNTL4);
 				tmp &= ~LC_SET_QUIESCE;
@@ -7223,15 +7248,15 @@ static void si_pcie_gen3_enable(struct radeon_device *rdev)
 	speed_cntl &= ~LC_FORCE_DIS_SW_SPEED_CHANGE;
 	WREG32_PCIE_PORT(PCIE_LC_SPEED_CNTL, speed_cntl);
 
-	pci_read_config_word(rdev->pdev, gpu_pos + PCI_EXP_LNKCTL2, &tmp16);
-	tmp16 &= ~0xf;
+	pcie_capability_read_word(rdev->pdev, PCI_EXP_LNKCTL2, &tmp16);
+	tmp16 &= ~PCI_EXP_LNKCTL2_TLS;
 	if (speed_cap == PCIE_SPEED_8_0GT)
-		tmp16 |= 3; /* gen3 */
+		tmp16 |= PCI_EXP_LNKCTL2_TLS_8_0GT; /* gen3 */
 	else if (speed_cap == PCIE_SPEED_5_0GT)
-		tmp16 |= 2; /* gen2 */
+		tmp16 |= PCI_EXP_LNKCTL2_TLS_5_0GT; /* gen2 */
 	else
-		tmp16 |= 1; /* gen1 */
-	pci_write_config_word(rdev->pdev, gpu_pos + PCI_EXP_LNKCTL2, tmp16);
+		tmp16 |= PCI_EXP_LNKCTL2_TLS_2_5GT; /* gen1 */
+	pcie_capability_write_word(rdev->pdev, PCI_EXP_LNKCTL2, tmp16);
 
 	speed_cntl = RREG32_PCIE_PORT(PCIE_LC_SPEED_CNTL);
 	speed_cntl |= LC_INITIATE_LINK_SPEED_CHANGE;
diff --git a/drivers/gpu/drm/radeon/si_dpm.c b/drivers/gpu/drm/radeon/si_dpm.c
index 460fd98e40a7..a0b382a637a6 100644
--- a/drivers/gpu/drm/radeon/si_dpm.c
+++ b/drivers/gpu/drm/radeon/si_dpm.c
@@ -1958,6 +1958,7 @@ static void si_initialize_powertune_defaults(struct radeon_device *rdev)
 		case 0x682C:
 			si_pi->cac_weights = cac_weights_cape_verde_pro;
 			si_pi->dte_data = dte_data_sun_xt;
+			update_dte_from_pl2 = true;
 			break;
 		case 0x6825:
 		case 0x6827:
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_drv.c b/drivers/gpu/drm/rcar-du/rcar_du_drv.c
index 9c93eb4fad8b..f266c17b907a 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_drv.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_drv.c
@@ -131,6 +131,35 @@ static const struct rcar_du_device_info rcar_du_r8a774a1_info = {
 	.dpll_mask =  BIT(1),
 };
 
+static const struct rcar_du_device_info rcar_du_r8a774b1_info = {
+	.gen = 3,
+	.features = RCAR_DU_FEATURE_CRTC_IRQ_CLOCK
+		  | RCAR_DU_FEATURE_VSP1_SOURCE
+		  | RCAR_DU_FEATURE_INTERLACED
+		  | RCAR_DU_FEATURE_TVM_SYNC,
+	.channels_mask = BIT(3) | BIT(1) | BIT(0),
+	.routes = {
+		/*
+		 * R8A774B1 has one RGB output, one LVDS output and one HDMI
+		 * output.
+		 */
+		[RCAR_DU_OUTPUT_DPAD0] = {
+			.possible_crtcs = BIT(2),
+			.port = 0,
+		},
+		[RCAR_DU_OUTPUT_HDMI0] = {
+			.possible_crtcs = BIT(1),
+			.port = 1,
+		},
+		[RCAR_DU_OUTPUT_LVDS0] = {
+			.possible_crtcs = BIT(0),
+			.port = 2,
+		},
+	},
+	.num_lvds = 1,
+	.dpll_mask =  BIT(1),
+};
+
 static const struct rcar_du_device_info rcar_du_r8a774c0_info = {
 	.gen = 3,
 	.features = RCAR_DU_FEATURE_CRTC_IRQ_CLOCK
@@ -416,6 +445,7 @@ static const struct of_device_id rcar_du_of_table[] = {
 	{ .compatible = "renesas,du-r8a7745", .data = &rzg1_du_r8a7745_info },
 	{ .compatible = "renesas,du-r8a77470", .data = &rzg1_du_r8a77470_info },
 	{ .compatible = "renesas,du-r8a774a1", .data = &rcar_du_r8a774a1_info },
+	{ .compatible = "renesas,du-r8a774b1", .data = &rcar_du_r8a774b1_info },
 	{ .compatible = "renesas,du-r8a774c0", .data = &rcar_du_r8a774c0_info },
 	{ .compatible = "renesas,du-r8a7779", .data = &rcar_du_r8a7779_info },
 	{ .compatible = "renesas,du-r8a7790", .data = &rcar_du_r8a7790_info },
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_encoder.c b/drivers/gpu/drm/rcar-du/rcar_du_encoder.c
index 0f00bdfe2366..3cd83a030a04 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_encoder.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_encoder.c
@@ -9,6 +9,7 @@
 
 #include <linux/export.h>
 
+#include <drm/drm_bridge.h>
 #include <drm/drm_crtc.h>
 #include <drm/drm_modeset_helper_vtables.h>
 #include <drm/drm_panel.h>
@@ -84,8 +85,8 @@ int rcar_du_encoder_init(struct rcar_du_device *rcdu,
 			goto done;
 		}
 
-		bridge = devm_drm_panel_bridge_add(rcdu->dev, panel,
-						   DRM_MODE_CONNECTOR_DPI);
+		bridge = devm_drm_panel_bridge_add_typed(rcdu->dev, panel,
+							 DRM_MODE_CONNECTOR_DPI);
 		if (IS_ERR(bridge)) {
 			ret = PTR_ERR(bridge);
 			goto done;
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_kms.c b/drivers/gpu/drm/rcar-du/rcar_du_kms.c
index 2dc9caee8767..0d59f390de19 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_kms.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_kms.c
@@ -585,7 +585,11 @@ static int rcar_du_vsps_init(struct rcar_du_device *rcdu)
 
 		vsps[j].crtcs_mask |= BIT(i);
 
-		/* Store the VSP pointer and pipe index in the CRTC. */
+		/*
+		 * Store the VSP pointer and pipe index in the CRTC. If the
+		 * second cell of the 'vsps' specifier isn't present, default
+		 * to 0 to remain compatible with older DT bindings.
+		 */
 		rcdu->crtcs[i].vsp = &rcdu->vsps[j];
 		rcdu->crtcs[i].vsp_pipe = cells >= 1 ? args.args[0] : 0;
 	}
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_writeback.c b/drivers/gpu/drm/rcar-du/rcar_du_writeback.c
index ae07290bba6a..04efa78d70b6 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_writeback.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_writeback.c
@@ -147,7 +147,7 @@ static int rcar_du_wb_enc_atomic_check(struct drm_encoder *encoder,
 	struct drm_device *dev = encoder->dev;
 	struct drm_framebuffer *fb;
 
-	if (!conn_state->writeback_job || !conn_state->writeback_job->fb)
+	if (!conn_state->writeback_job)
 		return 0;
 
 	fb = conn_state->writeback_job->fb;
@@ -221,7 +221,7 @@ void rcar_du_writeback_setup(struct rcar_du_crtc *rcrtc,
 	unsigned int i;
 
 	state = rcrtc->writeback.base.state;
-	if (!state || !state->writeback_job || !state->writeback_job->fb)
+	if (!state || !state->writeback_job)
 		return;
 
 	fb = state->writeback_job->fb;
diff --git a/drivers/gpu/drm/rcar-du/rcar_lvds.c b/drivers/gpu/drm/rcar-du/rcar_lvds.c
index 3fc7e6899cab..8c6c172bbf2e 100644
--- a/drivers/gpu/drm/rcar-du/rcar_lvds.c
+++ b/drivers/gpu/drm/rcar-du/rcar_lvds.c
@@ -16,6 +16,7 @@
 #include <linux/of_graph.h>
 #include <linux/platform_device.h>
 #include <linux/slab.h>
+#include <linux/sys_soc.h>
 
 #include <drm/drm_atomic.h>
 #include <drm/drm_atomic_helper.h>
@@ -842,8 +843,23 @@ static int rcar_lvds_get_clocks(struct rcar_lvds *lvds)
 	return 0;
 }
 
+static const struct rcar_lvds_device_info rcar_lvds_r8a7790es1_info = {
+	.gen = 2,
+	.quirks = RCAR_LVDS_QUIRK_LANES,
+	.pll_setup = rcar_lvds_pll_setup_gen2,
+};
+
+static const struct soc_device_attribute lvds_quirk_matches[] = {
+	{
+		.soc_id = "r8a7790", .revision = "ES1.*",
+		.data = &rcar_lvds_r8a7790es1_info,
+	},
+	{ /* sentinel */ }
+};
+
 static int rcar_lvds_probe(struct platform_device *pdev)
 {
+	const struct soc_device_attribute *attr;
 	struct rcar_lvds *lvds;
 	struct resource *mem;
 	int ret;
@@ -857,6 +873,10 @@ static int rcar_lvds_probe(struct platform_device *pdev)
 	lvds->dev = &pdev->dev;
 	lvds->info = of_device_get_match_data(&pdev->dev);
 
+	attr = soc_device_match(lvds_quirk_matches);
+	if (attr)
+		lvds->info = attr->data;
+
 	ret = rcar_lvds_parse_dt(lvds);
 	if (ret < 0)
 		return ret;
@@ -893,12 +913,6 @@ static const struct rcar_lvds_device_info rcar_lvds_gen2_info = {
 	.pll_setup = rcar_lvds_pll_setup_gen2,
 };
 
-static const struct rcar_lvds_device_info rcar_lvds_r8a7790_info = {
-	.gen = 2,
-	.quirks = RCAR_LVDS_QUIRK_LANES,
-	.pll_setup = rcar_lvds_pll_setup_gen2,
-};
-
 static const struct rcar_lvds_device_info rcar_lvds_gen3_info = {
 	.gen = 3,
 	.quirks = RCAR_LVDS_QUIRK_PWD,
@@ -929,8 +943,9 @@ static const struct of_device_id rcar_lvds_of_table[] = {
 	{ .compatible = "renesas,r8a7743-lvds", .data = &rcar_lvds_gen2_info },
 	{ .compatible = "renesas,r8a7744-lvds", .data = &rcar_lvds_gen2_info },
 	{ .compatible = "renesas,r8a774a1-lvds", .data = &rcar_lvds_gen3_info },
+	{ .compatible = "renesas,r8a774b1-lvds", .data = &rcar_lvds_gen3_info },
 	{ .compatible = "renesas,r8a774c0-lvds", .data = &rcar_lvds_r8a77990_info },
-	{ .compatible = "renesas,r8a7790-lvds", .data = &rcar_lvds_r8a7790_info },
+	{ .compatible = "renesas,r8a7790-lvds", .data = &rcar_lvds_gen2_info },
 	{ .compatible = "renesas,r8a7791-lvds", .data = &rcar_lvds_gen2_info },
 	{ .compatible = "renesas,r8a7793-lvds", .data = &rcar_lvds_gen2_info },
 	{ .compatible = "renesas,r8a7795-lvds", .data = &rcar_lvds_gen3_info },
diff --git a/drivers/gpu/drm/rockchip/cdn-dp-core.c b/drivers/gpu/drm/rockchip/cdn-dp-core.c
index d505ea7d5384..eed594bd38d3 100644
--- a/drivers/gpu/drm/rockchip/cdn-dp-core.c
+++ b/drivers/gpu/drm/rockchip/cdn-dp-core.c
@@ -477,8 +477,8 @@ static int cdn_dp_disable(struct cdn_dp_device *dp)
 	cdn_dp_set_firmware_active(dp, false);
 	cdn_dp_clk_disable(dp);
 	dp->active = false;
-	dp->link.rate = 0;
-	dp->link.num_lanes = 0;
+	dp->max_lanes = 0;
+	dp->max_rate = 0;
 	if (!dp->connected) {
 		kfree(dp->edid);
 		dp->edid = NULL;
@@ -570,7 +570,7 @@ static bool cdn_dp_check_link_status(struct cdn_dp_device *dp)
 	struct cdn_dp_port *port = cdn_dp_connected_port(dp);
 	u8 sink_lanes = drm_dp_max_lane_count(dp->dpcd);
 
-	if (!port || !dp->link.rate || !dp->link.num_lanes)
+	if (!port || !dp->max_rate || !dp->max_lanes)
 		return false;
 
 	if (cdn_dp_dpcd_read(dp, DP_LANE0_1_STATUS, link_status,
@@ -952,8 +952,8 @@ static void cdn_dp_pd_event_work(struct work_struct *work)
 
 	/* Enabled and connected with a sink, re-train if requested */
 	} else if (!cdn_dp_check_link_status(dp)) {
-		unsigned int rate = dp->link.rate;
-		unsigned int lanes = dp->link.num_lanes;
+		unsigned int rate = dp->max_rate;
+		unsigned int lanes = dp->max_lanes;
 		struct drm_display_mode *mode = &dp->mode;
 
 		DRM_DEV_INFO(dp->dev, "Connected with sink. Re-train link\n");
@@ -966,7 +966,7 @@ static void cdn_dp_pd_event_work(struct work_struct *work)
 
 		/* If training result is changed, update the video config */
 		if (mode->clock &&
-		    (rate != dp->link.rate || lanes != dp->link.num_lanes)) {
+		    (rate != dp->max_rate || lanes != dp->max_lanes)) {
 			ret = cdn_dp_config_video(dp);
 			if (ret) {
 				dp->connected = false;
diff --git a/drivers/gpu/drm/rockchip/cdn-dp-core.h b/drivers/gpu/drm/rockchip/cdn-dp-core.h
index b85ea89eb60b..83c4586665b4 100644
--- a/drivers/gpu/drm/rockchip/cdn-dp-core.h
+++ b/drivers/gpu/drm/rockchip/cdn-dp-core.h
@@ -92,9 +92,10 @@ struct cdn_dp_device {
 	struct reset_control *core_rst;
 	struct audio_info audio_info;
 	struct video_info video_info;
-	struct drm_dp_link link;
 	struct cdn_dp_port *port[MAX_PHY];
 	u8 ports;
+	u8 max_lanes;
+	u8 max_rate;
 	u8 lanes;
 	int active_port;
 
diff --git a/drivers/gpu/drm/rockchip/cdn-dp-reg.c b/drivers/gpu/drm/rockchip/cdn-dp-reg.c
index 077c87021908..7361c07cb4a7 100644
--- a/drivers/gpu/drm/rockchip/cdn-dp-reg.c
+++ b/drivers/gpu/drm/rockchip/cdn-dp-reg.c
@@ -535,8 +535,8 @@ static int cdn_dp_get_training_status(struct cdn_dp_device *dp)
 	if (ret)
 		goto err_get_training_status;
 
-	dp->link.rate = drm_dp_bw_code_to_link_rate(status[0]);
-	dp->link.num_lanes = status[1];
+	dp->max_rate = drm_dp_bw_code_to_link_rate(status[0]);
+	dp->max_lanes = status[1];
 
 err_get_training_status:
 	if (ret)
@@ -560,8 +560,8 @@ int cdn_dp_train_link(struct cdn_dp_device *dp)
 		return ret;
 	}
 
-	DRM_DEV_DEBUG_KMS(dp->dev, "rate:0x%x, lanes:%d\n", dp->link.rate,
-			  dp->link.num_lanes);
+	DRM_DEV_DEBUG_KMS(dp->dev, "rate:0x%x, lanes:%d\n", dp->max_rate,
+			  dp->max_lanes);
 	return ret;
 }
 
@@ -639,7 +639,7 @@ int cdn_dp_config_video(struct cdn_dp_device *dp)
 	bit_per_pix = (video->color_fmt == YCBCR_4_2_2) ?
 		      (video->color_depth * 2) : (video->color_depth * 3);
 
-	link_rate = dp->link.rate / 1000;
+	link_rate = dp->max_rate / 1000;
 
 	ret = cdn_dp_reg_write(dp, BND_HSYNC2VSYNC, VIF_BYPASS_INTERLACE);
 	if (ret)
@@ -659,14 +659,13 @@ int cdn_dp_config_video(struct cdn_dp_device *dp)
 	do {
 		tu_size_reg += 2;
 		symbol = tu_size_reg * mode->clock * bit_per_pix;
-		do_div(symbol, dp->link.num_lanes * link_rate * 8);
+		do_div(symbol, dp->max_lanes * link_rate * 8);
 		rem = do_div(symbol, 1000);
 		if (tu_size_reg > 64) {
 			ret = -EINVAL;
 			DRM_DEV_ERROR(dp->dev,
 				      "tu error, clk:%d, lanes:%d, rate:%d\n",
-				      mode->clock, dp->link.num_lanes,
-				      link_rate);
+				      mode->clock, dp->max_lanes, link_rate);
 			goto err_config_video;
 		}
 	} while ((symbol <= 1) || (tu_size_reg - symbol < 4) ||
@@ -680,7 +679,7 @@ int cdn_dp_config_video(struct cdn_dp_device *dp)
 
 	/* set the FIFO Buffer size */
 	val = div_u64(mode->clock * (symbol + 1), 1000) + link_rate;
-	val /= (dp->link.num_lanes * link_rate);
+	val /= (dp->max_lanes * link_rate);
 	val = div_u64(8 * (symbol + 1), bit_per_pix) - val;
 	val += 2;
 	ret = cdn_dp_reg_write(dp, DP_VC_TABLE(15), val);
@@ -833,7 +832,7 @@ static void cdn_dp_audio_config_i2s(struct cdn_dp_device *dp,
 	u32 val;
 
 	if (audio->channels == 2) {
-		if (dp->link.num_lanes == 1)
+		if (dp->max_lanes == 1)
 			sub_pckt_num = 2;
 		else
 			sub_pckt_num = 4;
diff --git a/drivers/gpu/drm/rockchip/dw_hdmi-rockchip.c b/drivers/gpu/drm/rockchip/dw_hdmi-rockchip.c
index 906891b03a38..7f56d8c3491d 100644
--- a/drivers/gpu/drm/rockchip/dw_hdmi-rockchip.c
+++ b/drivers/gpu/drm/rockchip/dw_hdmi-rockchip.c
@@ -450,6 +450,7 @@ static const struct dw_hdmi_plat_data rk3328_hdmi_drv_data = {
 	.phy_ops = &rk3328_hdmi_phy_ops,
 	.phy_name = "inno_dw_hdmi_phy2",
 	.phy_force_vendor = true,
+	.use_drm_infoframe = true,
 };
 
 static struct rockchip_hdmi_chip_data rk3399_chip_data = {
@@ -464,6 +465,7 @@ static const struct dw_hdmi_plat_data rk3399_hdmi_drv_data = {
 	.cur_ctr    = rockchip_cur_ctr,
 	.phy_config = rockchip_phy_config,
 	.phy_data = &rk3399_chip_data,
+	.use_drm_infoframe = true,
 };
 
 static const struct of_device_id dw_hdmi_rockchip_dt_ids[] = {
diff --git a/drivers/gpu/drm/rockchip/rk3066_hdmi.c b/drivers/gpu/drm/rockchip/rk3066_hdmi.c
index 85fc5f01f761..cdb401f4283d 100644
--- a/drivers/gpu/drm/rockchip/rk3066_hdmi.c
+++ b/drivers/gpu/drm/rockchip/rk3066_hdmi.c
@@ -743,7 +743,6 @@ static int rk3066_hdmi_bind(struct device *dev, struct device *master,
 	struct platform_device *pdev = to_platform_device(dev);
 	struct drm_device *drm = data;
 	struct rk3066_hdmi *hdmi;
-	struct resource *iores;
 	int irq;
 	int ret;
 
@@ -753,12 +752,7 @@ static int rk3066_hdmi_bind(struct device *dev, struct device *master,
 
 	hdmi->dev = dev;
 	hdmi->drm_dev = drm;
-
-	iores = platform_get_resource(pdev, IORESOURCE_MEM, 0);
-	if (!iores)
-		return -ENXIO;
-
-	hdmi->regs = devm_ioremap_resource(dev, iores);
+	hdmi->regs = devm_platform_ioremap_resource(pdev, 0);
 	if (IS_ERR(hdmi->regs))
 		return PTR_ERR(hdmi->regs);
 
diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_gem.c b/drivers/gpu/drm/rockchip/rockchip_drm_gem.c
index 291e89b4045f..7582d0e6a60a 100644
--- a/drivers/gpu/drm/rockchip/rockchip_drm_gem.c
+++ b/drivers/gpu/drm/rockchip/rockchip_drm_gem.c
@@ -294,7 +294,7 @@ static void rockchip_gem_release_object(struct rockchip_gem_object *rk_obj)
 	kfree(rk_obj);
 }
 
-struct rockchip_gem_object *
+static struct rockchip_gem_object *
 	rockchip_gem_alloc_object(struct drm_device *drm, unsigned int size)
 {
 	struct rockchip_gem_object *rk_obj;
diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_vop.c b/drivers/gpu/drm/rockchip/rockchip_drm_vop.c
index 613404f86668..d04b3492bdac 100644
--- a/drivers/gpu/drm/rockchip/rockchip_drm_vop.c
+++ b/drivers/gpu/drm/rockchip/rockchip_drm_vop.c
@@ -139,6 +139,7 @@ struct vop {
 
 	uint32_t *regsbak;
 	void __iomem *regs;
+	void __iomem *lut_regs;
 
 	/* physical map length of vop register */
 	uint32_t len;
@@ -1040,14 +1041,118 @@ static bool vop_crtc_mode_fixup(struct drm_crtc *crtc,
 				struct drm_display_mode *adjusted_mode)
 {
 	struct vop *vop = to_vop(crtc);
+	unsigned long rate;
 
-	adjusted_mode->clock =
-		DIV_ROUND_UP(clk_round_rate(vop->dclk,
-					    adjusted_mode->clock * 1000), 1000);
+	/*
+	 * Clock craziness.
+	 *
+	 * Key points:
+	 *
+	 * - DRM works in in kHz.
+	 * - Clock framework works in Hz.
+	 * - Rockchip's clock driver picks the clock rate that is the
+	 *   same _OR LOWER_ than the one requested.
+	 *
+	 * Action plan:
+	 *
+	 * 1. When DRM gives us a mode, we should add 999 Hz to it.  That way
+	 *    if the clock we need is 60000001 Hz (~60 MHz) and DRM tells us to
+	 *    make 60000 kHz then the clock framework will actually give us
+	 *    the right clock.
+	 *
+	 *    NOTE: if the PLL (maybe through a divider) could actually make
+	 *    a clock rate 999 Hz higher instead of the one we want then this
+	 *    could be a problem.  Unfortunately there's not much we can do
+	 *    since it's baked into DRM to use kHz.  It shouldn't matter in
+	 *    practice since Rockchip PLLs are controlled by tables and
+	 *    even if there is a divider in the middle I wouldn't expect PLL
+	 *    rates in the table that are just a few kHz different.
+	 *
+	 * 2. Get the clock framework to round the rate for us to tell us
+	 *    what it will actually make.
+	 *
+	 * 3. Store the rounded up rate so that we don't need to worry about
+	 *    this in the actual clk_set_rate().
+	 */
+	rate = clk_round_rate(vop->dclk, adjusted_mode->clock * 1000 + 999);
+	adjusted_mode->clock = DIV_ROUND_UP(rate, 1000);
 
 	return true;
 }
 
+static bool vop_dsp_lut_is_enabled(struct vop *vop)
+{
+	return vop_read_reg(vop, 0, &vop->data->common->dsp_lut_en);
+}
+
+static void vop_crtc_write_gamma_lut(struct vop *vop, struct drm_crtc *crtc)
+{
+	struct drm_color_lut *lut = crtc->state->gamma_lut->data;
+	unsigned int i;
+
+	for (i = 0; i < crtc->gamma_size; i++) {
+		u32 word;
+
+		word = (drm_color_lut_extract(lut[i].red, 10) << 20) |
+		       (drm_color_lut_extract(lut[i].green, 10) << 10) |
+			drm_color_lut_extract(lut[i].blue, 10);
+		writel(word, vop->lut_regs + i * 4);
+	}
+}
+
+static void vop_crtc_gamma_set(struct vop *vop, struct drm_crtc *crtc,
+			       struct drm_crtc_state *old_state)
+{
+	struct drm_crtc_state *state = crtc->state;
+	unsigned int idle;
+	int ret;
+
+	if (!vop->lut_regs)
+		return;
+	/*
+	 * To disable gamma (gamma_lut is null) or to write
+	 * an update to the LUT, clear dsp_lut_en.
+	 */
+	spin_lock(&vop->reg_lock);
+	VOP_REG_SET(vop, common, dsp_lut_en, 0);
+	vop_cfg_done(vop);
+	spin_unlock(&vop->reg_lock);
+
+	/*
+	 * In order to write the LUT to the internal memory,
+	 * we need to first make sure the dsp_lut_en bit is cleared.
+	 */
+	ret = readx_poll_timeout(vop_dsp_lut_is_enabled, vop,
+				 idle, !idle, 5, 30 * 1000);
+	if (ret) {
+		DRM_DEV_ERROR(vop->dev, "display LUT RAM enable timeout!\n");
+		return;
+	}
+
+	if (!state->gamma_lut)
+		return;
+
+	spin_lock(&vop->reg_lock);
+	vop_crtc_write_gamma_lut(vop, crtc);
+	VOP_REG_SET(vop, common, dsp_lut_en, 1);
+	vop_cfg_done(vop);
+	spin_unlock(&vop->reg_lock);
+}
+
+static void vop_crtc_atomic_begin(struct drm_crtc *crtc,
+				  struct drm_crtc_state *old_crtc_state)
+{
+	struct vop *vop = to_vop(crtc);
+
+	/*
+	 * Only update GAMMA if the 'active' flag is not changed,
+	 * otherwise it's updated by .atomic_enable.
+	 */
+	if (crtc->state->color_mgmt_changed &&
+	    !crtc->state->active_changed)
+		vop_crtc_gamma_set(vop, crtc, old_crtc_state);
+}
+
 static void vop_crtc_atomic_enable(struct drm_crtc *crtc,
 				   struct drm_crtc_state *old_state)
 {
@@ -1075,6 +1180,14 @@ static void vop_crtc_atomic_enable(struct drm_crtc *crtc,
 		return;
 	}
 
+	/*
+	 * If we have a GAMMA LUT in the state, then let's make sure
+	 * it's updated. We might be coming out of suspend,
+	 * which means the LUT internal memory needs to be re-written.
+	 */
+	if (crtc->state->gamma_lut)
+		vop_crtc_gamma_set(vop, crtc, old_state);
+
 	mutex_lock(&vop->vop_lock);
 
 	WARN_ON(vop->event);
@@ -1085,9 +1198,7 @@ static void vop_crtc_atomic_enable(struct drm_crtc *crtc,
 		DRM_DEV_ERROR(vop->dev, "Failed to enable vop (%d)\n", ret);
 		return;
 	}
-
-	pin_pol = BIT(DCLK_INVERT);
-	pin_pol |= (adjusted_mode->flags & DRM_MODE_FLAG_PHSYNC) ?
+	pin_pol = (adjusted_mode->flags & DRM_MODE_FLAG_PHSYNC) ?
 		   BIT(HSYNC_POSITIVE) : 0;
 	pin_pol |= (adjusted_mode->flags & DRM_MODE_FLAG_PVSYNC) ?
 		   BIT(VSYNC_POSITIVE) : 0;
@@ -1096,25 +1207,29 @@ static void vop_crtc_atomic_enable(struct drm_crtc *crtc,
 
 	switch (s->output_type) {
 	case DRM_MODE_CONNECTOR_LVDS:
-		VOP_REG_SET(vop, output, rgb_en, 1);
+		VOP_REG_SET(vop, output, rgb_dclk_pol, 1);
 		VOP_REG_SET(vop, output, rgb_pin_pol, pin_pol);
+		VOP_REG_SET(vop, output, rgb_en, 1);
 		break;
 	case DRM_MODE_CONNECTOR_eDP:
+		VOP_REG_SET(vop, output, edp_dclk_pol, 1);
 		VOP_REG_SET(vop, output, edp_pin_pol, pin_pol);
 		VOP_REG_SET(vop, output, edp_en, 1);
 		break;
 	case DRM_MODE_CONNECTOR_HDMIA:
+		VOP_REG_SET(vop, output, hdmi_dclk_pol, 1);
 		VOP_REG_SET(vop, output, hdmi_pin_pol, pin_pol);
 		VOP_REG_SET(vop, output, hdmi_en, 1);
 		break;
 	case DRM_MODE_CONNECTOR_DSI:
+		VOP_REG_SET(vop, output, mipi_dclk_pol, 1);
 		VOP_REG_SET(vop, output, mipi_pin_pol, pin_pol);
 		VOP_REG_SET(vop, output, mipi_en, 1);
 		VOP_REG_SET(vop, output, mipi_dual_channel_en,
 			    !!(s->output_flags & ROCKCHIP_OUTPUT_DSI_DUAL));
 		break;
 	case DRM_MODE_CONNECTOR_DisplayPort:
-		pin_pol &= ~BIT(DCLK_INVERT);
+		VOP_REG_SET(vop, output, dp_dclk_pol, 0);
 		VOP_REG_SET(vop, output, dp_pin_pol, pin_pol);
 		VOP_REG_SET(vop, output, dp_en, 1);
 		break;
@@ -1191,6 +1306,26 @@ static void vop_wait_for_irq_handler(struct vop *vop)
 	synchronize_irq(vop->irq);
 }
 
+static int vop_crtc_atomic_check(struct drm_crtc *crtc,
+				 struct drm_crtc_state *crtc_state)
+{
+	struct vop *vop = to_vop(crtc);
+
+	if (vop->lut_regs && crtc_state->color_mgmt_changed &&
+	    crtc_state->gamma_lut) {
+		unsigned int len;
+
+		len = drm_color_lut_size(crtc_state->gamma_lut);
+		if (len != crtc->gamma_size) {
+			DRM_DEBUG_KMS("Invalid LUT size; got %d, expected %d\n",
+				      len, crtc->gamma_size);
+			return -EINVAL;
+		}
+	}
+
+	return 0;
+}
+
 static void vop_crtc_atomic_flush(struct drm_crtc *crtc,
 				  struct drm_crtc_state *old_crtc_state)
 {
@@ -1243,6 +1378,8 @@ static void vop_crtc_atomic_flush(struct drm_crtc *crtc,
 
 static const struct drm_crtc_helper_funcs vop_crtc_helper_funcs = {
 	.mode_fixup = vop_crtc_mode_fixup,
+	.atomic_check = vop_crtc_atomic_check,
+	.atomic_begin = vop_crtc_atomic_begin,
 	.atomic_flush = vop_crtc_atomic_flush,
 	.atomic_enable = vop_crtc_atomic_enable,
 	.atomic_disable = vop_crtc_atomic_disable,
@@ -1361,6 +1498,7 @@ static const struct drm_crtc_funcs vop_crtc_funcs = {
 	.disable_vblank = vop_crtc_disable_vblank,
 	.set_crc_source = vop_crtc_set_crc_source,
 	.verify_crc_source = vop_crtc_verify_crc_source,
+	.gamma_set = drm_atomic_helper_legacy_gamma_set,
 };
 
 static void vop_fb_unref_worker(struct drm_flip_work *work, void *val)
@@ -1518,6 +1656,10 @@ static int vop_create_crtc(struct vop *vop)
 		goto err_cleanup_planes;
 
 	drm_crtc_helper_add(crtc, &vop_crtc_helper_funcs);
+	if (vop->lut_regs) {
+		drm_mode_crtc_set_gamma_size(crtc, vop_data->lut_size);
+		drm_crtc_enable_color_mgmt(crtc, 0, false, vop_data->lut_size);
+	}
 
 	/*
 	 * Create drm_planes for overlay windows with possible_crtcs restricted
@@ -1822,6 +1964,17 @@ static int vop_bind(struct device *dev, struct device *master, void *data)
 	if (IS_ERR(vop->regs))
 		return PTR_ERR(vop->regs);
 
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 1);
+	if (res) {
+		if (!vop_data->lut_size) {
+			DRM_DEV_ERROR(dev, "no gamma LUT size defined\n");
+			return -EINVAL;
+		}
+		vop->lut_regs = devm_ioremap_resource(dev, res);
+		if (IS_ERR(vop->lut_regs))
+			return PTR_ERR(vop->lut_regs);
+	}
+
 	vop->regsbak = devm_kzalloc(dev, vop->len, GFP_KERNEL);
 	if (!vop->regsbak)
 		return -ENOMEM;
diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_vop.h b/drivers/gpu/drm/rockchip/rockchip_drm_vop.h
index 2149a889c29d..0b3d18c457b2 100644
--- a/drivers/gpu/drm/rockchip/rockchip_drm_vop.h
+++ b/drivers/gpu/drm/rockchip/rockchip_drm_vop.h
@@ -46,10 +46,15 @@ struct vop_modeset {
 struct vop_output {
 	struct vop_reg pin_pol;
 	struct vop_reg dp_pin_pol;
+	struct vop_reg dp_dclk_pol;
 	struct vop_reg edp_pin_pol;
+	struct vop_reg edp_dclk_pol;
 	struct vop_reg hdmi_pin_pol;
+	struct vop_reg hdmi_dclk_pol;
 	struct vop_reg mipi_pin_pol;
+	struct vop_reg mipi_dclk_pol;
 	struct vop_reg rgb_pin_pol;
+	struct vop_reg rgb_dclk_pol;
 	struct vop_reg dp_en;
 	struct vop_reg edp_en;
 	struct vop_reg hdmi_en;
@@ -67,6 +72,7 @@ struct vop_common {
 	struct vop_reg dither_down_mode;
 	struct vop_reg dither_down_en;
 	struct vop_reg dither_up;
+	struct vop_reg dsp_lut_en;
 	struct vop_reg gate_en;
 	struct vop_reg mmu_en;
 	struct vop_reg out_mode;
@@ -170,6 +176,7 @@ struct vop_data {
 	const struct vop_win_yuv2yuv_data *win_yuv2yuv;
 	const struct vop_win_data *win;
 	unsigned int win_size;
+	unsigned int lut_size;
 
 #define VOP_FEATURE_OUTPUT_RGB10	BIT(0)
 #define VOP_FEATURE_INTERNAL_RGB	BIT(1)
@@ -294,8 +301,7 @@ enum dither_down_mode_sel {
 enum vop_pol {
 	HSYNC_POSITIVE = 0,
 	VSYNC_POSITIVE = 1,
-	DEN_NEGATIVE   = 2,
-	DCLK_INVERT    = 3
+	DEN_NEGATIVE   = 2
 };
 
 #define FRAC_16_16(mult, div)    (((mult) << 16) / (div))
diff --git a/drivers/gpu/drm/rockchip/rockchip_lvds.c b/drivers/gpu/drm/rockchip/rockchip_lvds.c
index 64aefa856896..8a4c9af0ba73 100644
--- a/drivers/gpu/drm/rockchip/rockchip_lvds.c
+++ b/drivers/gpu/drm/rockchip/rockchip_lvds.c
@@ -16,6 +16,7 @@
 #include <linux/regmap.h>
 #include <linux/reset.h>
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_bridge.h>
 
 #include <drm/drm_dp_helper.h>
 #include <drm/drm_of.h>
diff --git a/drivers/gpu/drm/rockchip/rockchip_rgb.c b/drivers/gpu/drm/rockchip/rockchip_rgb.c
index 89e0bb0fe0ab..ae730275a34f 100644
--- a/drivers/gpu/drm/rockchip/rockchip_rgb.c
+++ b/drivers/gpu/drm/rockchip/rockchip_rgb.c
@@ -9,6 +9,7 @@
 #include <linux/of_graph.h>
 
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_dp_helper.h>
 #include <drm/drm_of.h>
 #include <drm/drm_panel.h>
@@ -135,7 +136,8 @@ struct rockchip_rgb *rockchip_rgb_init(struct device *dev,
 	drm_encoder_helper_add(encoder, &rockchip_rgb_encoder_helper_funcs);
 
 	if (panel) {
-		bridge = drm_panel_bridge_add(panel, DRM_MODE_CONNECTOR_LVDS);
+		bridge = drm_panel_bridge_add_typed(panel,
+						    DRM_MODE_CONNECTOR_LVDS);
 		if (IS_ERR(bridge))
 			return ERR_CAST(bridge);
 	}
diff --git a/drivers/gpu/drm/rockchip/rockchip_vop_reg.c b/drivers/gpu/drm/rockchip/rockchip_vop_reg.c
index d1494be14471..7a9d979c8d5d 100644
--- a/drivers/gpu/drm/rockchip/rockchip_vop_reg.c
+++ b/drivers/gpu/drm/rockchip/rockchip_vop_reg.c
@@ -16,6 +16,7 @@
 
 #include "rockchip_drm_vop.h"
 #include "rockchip_vop_reg.h"
+#include "rockchip_drm_drv.h"
 
 #define _VOP_REG(off, _mask, _shift, _write_mask, _relaxed) \
 		{ \
@@ -214,9 +215,11 @@ static const struct vop_modeset px30_modeset = {
 };
 
 static const struct vop_output px30_output = {
-	.rgb_pin_pol = VOP_REG(PX30_DSP_CTRL0, 0xf, 1),
-	.mipi_pin_pol = VOP_REG(PX30_DSP_CTRL0, 0xf, 25),
+	.rgb_dclk_pol = VOP_REG(PX30_DSP_CTRL0, 0x1, 1),
+	.rgb_pin_pol = VOP_REG(PX30_DSP_CTRL0, 0x7, 2),
 	.rgb_en = VOP_REG(PX30_DSP_CTRL0, 0x1, 0),
+	.mipi_dclk_pol = VOP_REG(PX30_DSP_CTRL0, 0x1, 25),
+	.mipi_pin_pol = VOP_REG(PX30_DSP_CTRL0, 0x7, 26),
 	.mipi_en = VOP_REG(PX30_DSP_CTRL0, 0x1, 24),
 };
 
@@ -598,6 +601,7 @@ static const struct vop_common rk3288_common = {
 	.dither_down_en = VOP_REG(RK3288_DSP_CTRL1, 0x1, 2),
 	.pre_dither_down = VOP_REG(RK3288_DSP_CTRL1, 0x1, 1),
 	.dither_up = VOP_REG(RK3288_DSP_CTRL1, 0x1, 6),
+	.dsp_lut_en = VOP_REG(RK3288_DSP_CTRL1, 0x1, 0),
 	.data_blank = VOP_REG(RK3288_DSP_CTRL0, 0x1, 19),
 	.dsp_blank = VOP_REG(RK3288_DSP_CTRL0, 0x3, 18),
 	.out_mode = VOP_REG(RK3288_DSP_CTRL0, 0xf, 0),
@@ -646,6 +650,7 @@ static const struct vop_data rk3288_vop = {
 	.output = &rk3288_output,
 	.win = rk3288_vop_win_data,
 	.win_size = ARRAY_SIZE(rk3288_vop_win_data),
+	.lut_size = 1024,
 };
 
 static const int rk3368_vop_intrs[] = {
@@ -717,10 +722,14 @@ static const struct vop_win_data rk3368_vop_win_data[] = {
 };
 
 static const struct vop_output rk3368_output = {
-	.rgb_pin_pol = VOP_REG(RK3368_DSP_CTRL1, 0xf, 16),
-	.hdmi_pin_pol = VOP_REG(RK3368_DSP_CTRL1, 0xf, 20),
-	.edp_pin_pol = VOP_REG(RK3368_DSP_CTRL1, 0xf, 24),
-	.mipi_pin_pol = VOP_REG(RK3368_DSP_CTRL1, 0xf, 28),
+	.rgb_dclk_pol = VOP_REG(RK3368_DSP_CTRL1, 0x1, 19),
+	.hdmi_dclk_pol = VOP_REG(RK3368_DSP_CTRL1, 0x1, 23),
+	.edp_dclk_pol = VOP_REG(RK3368_DSP_CTRL1, 0x1, 27),
+	.mipi_dclk_pol = VOP_REG(RK3368_DSP_CTRL1, 0x1, 31),
+	.rgb_pin_pol = VOP_REG(RK3368_DSP_CTRL1, 0x7, 16),
+	.hdmi_pin_pol = VOP_REG(RK3368_DSP_CTRL1, 0x7, 20),
+	.edp_pin_pol = VOP_REG(RK3368_DSP_CTRL1, 0x7, 24),
+	.mipi_pin_pol = VOP_REG(RK3368_DSP_CTRL1, 0x7, 28),
 	.rgb_en = VOP_REG(RK3288_SYS_CTRL, 0x1, 12),
 	.hdmi_en = VOP_REG(RK3288_SYS_CTRL, 0x1, 13),
 	.edp_en = VOP_REG(RK3288_SYS_CTRL, 0x1, 14),
@@ -764,11 +773,16 @@ static const struct vop_data rk3366_vop = {
 };
 
 static const struct vop_output rk3399_output = {
-	.dp_pin_pol = VOP_REG(RK3399_DSP_CTRL1, 0xf, 16),
-	.rgb_pin_pol = VOP_REG(RK3368_DSP_CTRL1, 0xf, 16),
-	.hdmi_pin_pol = VOP_REG(RK3368_DSP_CTRL1, 0xf, 20),
-	.edp_pin_pol = VOP_REG(RK3368_DSP_CTRL1, 0xf, 24),
-	.mipi_pin_pol = VOP_REG(RK3368_DSP_CTRL1, 0xf, 28),
+	.dp_dclk_pol = VOP_REG(RK3399_DSP_CTRL1, 0x1, 19),
+	.rgb_dclk_pol = VOP_REG(RK3368_DSP_CTRL1, 0x1, 19),
+	.hdmi_dclk_pol = VOP_REG(RK3368_DSP_CTRL1, 0x1, 23),
+	.edp_dclk_pol = VOP_REG(RK3368_DSP_CTRL1, 0x1, 27),
+	.mipi_dclk_pol = VOP_REG(RK3368_DSP_CTRL1, 0x1, 31),
+	.dp_pin_pol = VOP_REG(RK3399_DSP_CTRL1, 0x7, 16),
+	.rgb_pin_pol = VOP_REG(RK3368_DSP_CTRL1, 0x7, 16),
+	.hdmi_pin_pol = VOP_REG(RK3368_DSP_CTRL1, 0x7, 20),
+	.edp_pin_pol = VOP_REG(RK3368_DSP_CTRL1, 0x7, 24),
+	.mipi_pin_pol = VOP_REG(RK3368_DSP_CTRL1, 0x7, 28),
 	.dp_en = VOP_REG(RK3399_SYS_CTRL, 0x1, 11),
 	.rgb_en = VOP_REG(RK3288_SYS_CTRL, 0x1, 12),
 	.hdmi_en = VOP_REG(RK3288_SYS_CTRL, 0x1, 13),
@@ -872,14 +886,18 @@ static const struct vop_modeset rk3328_modeset = {
 };
 
 static const struct vop_output rk3328_output = {
+	.rgb_dclk_pol = VOP_REG(RK3328_DSP_CTRL1, 0x1, 19),
+	.hdmi_dclk_pol = VOP_REG(RK3328_DSP_CTRL1, 0x1, 23),
+	.edp_dclk_pol = VOP_REG(RK3328_DSP_CTRL1, 0x1, 27),
+	.mipi_dclk_pol = VOP_REG(RK3328_DSP_CTRL1, 0x1, 31),
 	.rgb_en = VOP_REG(RK3328_SYS_CTRL, 0x1, 12),
 	.hdmi_en = VOP_REG(RK3328_SYS_CTRL, 0x1, 13),
 	.edp_en = VOP_REG(RK3328_SYS_CTRL, 0x1, 14),
 	.mipi_en = VOP_REG(RK3328_SYS_CTRL, 0x1, 15),
-	.rgb_pin_pol = VOP_REG(RK3328_DSP_CTRL1, 0xf, 16),
-	.hdmi_pin_pol = VOP_REG(RK3328_DSP_CTRL1, 0xf, 20),
-	.edp_pin_pol = VOP_REG(RK3328_DSP_CTRL1, 0xf, 24),
-	.mipi_pin_pol = VOP_REG(RK3328_DSP_CTRL1, 0xf, 28),
+	.rgb_pin_pol = VOP_REG(RK3328_DSP_CTRL1, 0x7, 16),
+	.hdmi_pin_pol = VOP_REG(RK3328_DSP_CTRL1, 0x7, 20),
+	.edp_pin_pol = VOP_REG(RK3328_DSP_CTRL1, 0x7, 24),
+	.mipi_pin_pol = VOP_REG(RK3328_DSP_CTRL1, 0x7, 28),
 };
 
 static const struct vop_misc rk3328_misc = {
diff --git a/drivers/gpu/drm/scheduler/sched_entity.c b/drivers/gpu/drm/scheduler/sched_entity.c
index 1a5153197fe9..461a7a8129f4 100644
--- a/drivers/gpu/drm/scheduler/sched_entity.c
+++ b/drivers/gpu/drm/scheduler/sched_entity.c
@@ -23,6 +23,7 @@
 
 #include <linux/kthread.h>
 #include <linux/slab.h>
+#include <linux/completion.h>
 
 #include <drm/drm_print.h>
 #include <drm/gpu_scheduler.h>
@@ -68,6 +69,8 @@ int drm_sched_entity_init(struct drm_sched_entity *entity,
 	if (!entity->rq_list)
 		return -ENOMEM;
 
+	init_completion(&entity->entity_idle);
+
 	for (i = 0; i < num_rq_list; ++i)
 		entity->rq_list[i] = rq_list[i];
 
@@ -286,11 +289,12 @@ void drm_sched_entity_fini(struct drm_sched_entity *entity)
 	 */
 	if (spsc_queue_count(&entity->job_queue)) {
 		if (sched) {
-			/* Park the kernel for a moment to make sure it isn't processing
-			 * our enity.
+			/*
+			 * Wait for thread to idle to make sure it isn't processing
+			 * this entity.
 			 */
-			kthread_park(sched->thread);
-			kthread_unpark(sched->thread);
+			wait_for_completion(&entity->entity_idle);
+
 		}
 		if (entity->dependency) {
 			dma_fence_remove_callback(entity->dependency,
diff --git a/drivers/gpu/drm/scheduler/sched_fence.c b/drivers/gpu/drm/scheduler/sched_fence.c
index 54977408f574..8b45c3a1b84e 100644
--- a/drivers/gpu/drm/scheduler/sched_fence.c
+++ b/drivers/gpu/drm/scheduler/sched_fence.c
@@ -128,13 +128,13 @@ static void drm_sched_fence_release_finished(struct dma_fence *f)
 	dma_fence_put(&fence->scheduled);
 }
 
-const struct dma_fence_ops drm_sched_fence_ops_scheduled = {
+static const struct dma_fence_ops drm_sched_fence_ops_scheduled = {
 	.get_driver_name = drm_sched_fence_get_driver_name,
 	.get_timeline_name = drm_sched_fence_get_timeline_name,
 	.release = drm_sched_fence_release_scheduled,
 };
 
-const struct dma_fence_ops drm_sched_fence_ops_finished = {
+static const struct dma_fence_ops drm_sched_fence_ops_finished = {
 	.get_driver_name = drm_sched_fence_get_driver_name,
 	.get_timeline_name = drm_sched_fence_get_timeline_name,
 	.release = drm_sched_fence_release_finished,
diff --git a/drivers/gpu/drm/scheduler/sched_main.c b/drivers/gpu/drm/scheduler/sched_main.c
index 9a0ee74d82dc..3c57e84222ca 100644
--- a/drivers/gpu/drm/scheduler/sched_main.c
+++ b/drivers/gpu/drm/scheduler/sched_main.c
@@ -47,6 +47,7 @@
 #include <linux/kthread.h>
 #include <linux/wait.h>
 #include <linux/sched.h>
+#include <linux/completion.h>
 #include <uapi/linux/sched/types.h>
 
 #include <drm/drm_print.h>
@@ -134,6 +135,7 @@ drm_sched_rq_select_entity(struct drm_sched_rq *rq)
 		list_for_each_entry_continue(entity, &rq->entities, list) {
 			if (drm_sched_entity_is_ready(entity)) {
 				rq->current_entity = entity;
+				reinit_completion(&entity->entity_idle);
 				spin_unlock(&rq->lock);
 				return entity;
 			}
@@ -144,6 +146,7 @@ drm_sched_rq_select_entity(struct drm_sched_rq *rq)
 
 		if (drm_sched_entity_is_ready(entity)) {
 			rq->current_entity = entity;
+			reinit_completion(&entity->entity_idle);
 			spin_unlock(&rq->lock);
 			return entity;
 		}
@@ -479,6 +482,7 @@ void drm_sched_resubmit_jobs(struct drm_gpu_scheduler *sched)
 	struct drm_sched_job *s_job, *tmp;
 	uint64_t guilty_context;
 	bool found_guilty = false;
+	struct dma_fence *fence;
 
 	list_for_each_entry_safe(s_job, tmp, &sched->ring_mirror_list, node) {
 		struct drm_sched_fence *s_fence = s_job->s_fence;
@@ -492,7 +496,18 @@ void drm_sched_resubmit_jobs(struct drm_gpu_scheduler *sched)
 			dma_fence_set_error(&s_fence->finished, -ECANCELED);
 
 		dma_fence_put(s_job->s_fence->parent);
-		s_job->s_fence->parent = sched->ops->run_job(s_job);
+		fence = sched->ops->run_job(s_job);
+
+		if (IS_ERR_OR_NULL(fence)) {
+			if (IS_ERR(fence))
+				dma_fence_set_error(&s_fence->finished, PTR_ERR(fence));
+
+			s_job->s_fence->parent = NULL;
+		} else {
+			s_job->s_fence->parent = fence;
+		}
+
+
 	}
 }
 EXPORT_SYMBOL(drm_sched_resubmit_jobs);
@@ -622,43 +637,45 @@ static void drm_sched_process_job(struct dma_fence *f, struct dma_fence_cb *cb)
 }
 
 /**
- * drm_sched_cleanup_jobs - destroy finished jobs
+ * drm_sched_get_cleanup_job - fetch the next finished job to be destroyed
  *
  * @sched: scheduler instance
  *
- * Remove all finished jobs from the mirror list and destroy them.
+ * Returns the next finished job from the mirror list (if there is one)
+ * ready for it to be destroyed.
  */
-static void drm_sched_cleanup_jobs(struct drm_gpu_scheduler *sched)
+static struct drm_sched_job *
+drm_sched_get_cleanup_job(struct drm_gpu_scheduler *sched)
 {
+	struct drm_sched_job *job;
 	unsigned long flags;
 
-	/* Don't destroy jobs while the timeout worker is running */
-	if (sched->timeout != MAX_SCHEDULE_TIMEOUT &&
-	    !cancel_delayed_work(&sched->work_tdr))
-		return;
-
+	/*
+	 * Don't destroy jobs while the timeout worker is running  OR thread
+	 * is being parked and hence assumed to not touch ring_mirror_list
+	 */
+	if ((sched->timeout != MAX_SCHEDULE_TIMEOUT &&
+	    !cancel_delayed_work(&sched->work_tdr)) ||
+	    __kthread_should_park(sched->thread))
+		return NULL;
 
-	while (!list_empty(&sched->ring_mirror_list)) {
-		struct drm_sched_job *job;
+	spin_lock_irqsave(&sched->job_list_lock, flags);
 
-		job = list_first_entry(&sched->ring_mirror_list,
+	job = list_first_entry_or_null(&sched->ring_mirror_list,
 				       struct drm_sched_job, node);
-		if (!dma_fence_is_signaled(&job->s_fence->finished))
-			break;
 
-		spin_lock_irqsave(&sched->job_list_lock, flags);
+	if (job && dma_fence_is_signaled(&job->s_fence->finished)) {
 		/* remove job from ring_mirror_list */
 		list_del_init(&job->node);
-		spin_unlock_irqrestore(&sched->job_list_lock, flags);
-
-		sched->ops->free_job(job);
+	} else {
+		job = NULL;
+		/* queue timeout for next job */
+		drm_sched_start_timeout(sched);
 	}
 
-	/* queue timeout for next job */
-	spin_lock_irqsave(&sched->job_list_lock, flags);
-	drm_sched_start_timeout(sched);
 	spin_unlock_irqrestore(&sched->job_list_lock, flags);
 
+	return job;
 }
 
 /**
@@ -698,17 +715,27 @@ static int drm_sched_main(void *param)
 		struct drm_sched_fence *s_fence;
 		struct drm_sched_job *sched_job;
 		struct dma_fence *fence;
+		struct drm_sched_job *cleanup_job = NULL;
 
 		wait_event_interruptible(sched->wake_up_worker,
-					 (drm_sched_cleanup_jobs(sched),
+					 (cleanup_job = drm_sched_get_cleanup_job(sched)) ||
 					 (!drm_sched_blocked(sched) &&
 					  (entity = drm_sched_select_entity(sched))) ||
-					 kthread_should_stop()));
+					 kthread_should_stop());
+
+		if (cleanup_job) {
+			sched->ops->free_job(cleanup_job);
+			/* queue timeout for next job */
+			drm_sched_start_timeout(sched);
+		}
 
 		if (!entity)
 			continue;
 
 		sched_job = drm_sched_entity_pop_job(entity);
+
+		complete(&entity->entity_idle);
+
 		if (!sched_job)
 			continue;
 
@@ -720,7 +747,7 @@ static int drm_sched_main(void *param)
 		fence = sched->ops->run_job(sched_job);
 		drm_sched_fence_scheduled(s_fence);
 
-		if (fence) {
+		if (!IS_ERR_OR_NULL(fence)) {
 			s_fence->parent = dma_fence_get(fence);
 			r = dma_fence_add_callback(fence, &sched_job->cb,
 						   drm_sched_process_job);
@@ -730,8 +757,12 @@ static int drm_sched_main(void *param)
 				DRM_ERROR("fence add callback failed (%d)\n",
 					  r);
 			dma_fence_put(fence);
-		} else
+		} else {
+			if (IS_ERR(fence))
+				dma_fence_set_error(&s_fence->finished, PTR_ERR(fence));
+
 			drm_sched_process_job(NULL, &sched_job->cb);
+		}
 
 		wake_up(&sched->job_scheduled);
 	}
diff --git a/drivers/gpu/drm/selftests/Makefile b/drivers/gpu/drm/selftests/Makefile
index aae88f8a016c..d2137342b371 100644
--- a/drivers/gpu/drm/selftests/Makefile
+++ b/drivers/gpu/drm/selftests/Makefile
@@ -1,6 +1,6 @@
 # SPDX-License-Identifier: GPL-2.0-only
 test-drm_modeset-y := test-drm_modeset_common.o test-drm_plane_helper.o \
                       test-drm_format.o test-drm_framebuffer.o \
-		      test-drm_damage_helper.o
+		      test-drm_damage_helper.o test-drm_dp_mst_helper.o
 
 obj-$(CONFIG_DRM_DEBUG_SELFTEST) += test-drm_mm.o test-drm_modeset.o test-drm_cmdline_parser.o
diff --git a/drivers/gpu/drm/selftests/drm_modeset_selftests.h b/drivers/gpu/drm/selftests/drm_modeset_selftests.h
index 464753746013..1898de0b4a4d 100644
--- a/drivers/gpu/drm/selftests/drm_modeset_selftests.h
+++ b/drivers/gpu/drm/selftests/drm_modeset_selftests.h
@@ -32,3 +32,5 @@ selftest(damage_iter_damage_one_intersect, igt_damage_iter_damage_one_intersect)
 selftest(damage_iter_damage_one_outside, igt_damage_iter_damage_one_outside)
 selftest(damage_iter_damage_src_moved, igt_damage_iter_damage_src_moved)
 selftest(damage_iter_damage_not_visible, igt_damage_iter_damage_not_visible)
+selftest(dp_mst_calc_pbn_mode, igt_dp_mst_calc_pbn_mode)
+selftest(dp_mst_sideband_msg_req_decode, igt_dp_mst_sideband_msg_req_decode)
diff --git a/drivers/gpu/drm/selftests/test-drm_dp_mst_helper.c b/drivers/gpu/drm/selftests/test-drm_dp_mst_helper.c
new file mode 100644
index 000000000000..af2b2de65316
--- /dev/null
+++ b/drivers/gpu/drm/selftests/test-drm_dp_mst_helper.c
@@ -0,0 +1,238 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Test cases for for the DRM DP MST helpers
+ */
+
+#define PREFIX_STR "[drm_dp_mst_helper]"
+
+#include <drm/drm_dp_mst_helper.h>
+#include <drm/drm_print.h>
+
+#include "../drm_dp_mst_topology_internal.h"
+#include "test-drm_modeset_common.h"
+
+int igt_dp_mst_calc_pbn_mode(void *ignored)
+{
+	int pbn, i;
+	const struct {
+		int rate;
+		int bpp;
+		int expected;
+	} test_params[] = {
+		{ 154000, 30, 689 },
+		{ 234000, 30, 1047 },
+		{ 297000, 24, 1063 },
+	};
+
+	for (i = 0; i < ARRAY_SIZE(test_params); i++) {
+		pbn = drm_dp_calc_pbn_mode(test_params[i].rate,
+					   test_params[i].bpp);
+		FAIL(pbn != test_params[i].expected,
+		     "Expected PBN %d for clock %d bpp %d, got %d\n",
+		     test_params[i].expected, test_params[i].rate,
+		     test_params[i].bpp, pbn);
+	}
+
+	return 0;
+}
+
+static bool
+sideband_msg_req_equal(const struct drm_dp_sideband_msg_req_body *in,
+		       const struct drm_dp_sideband_msg_req_body *out)
+{
+	const struct drm_dp_remote_i2c_read_tx *txin, *txout;
+	int i;
+
+	if (in->req_type != out->req_type)
+		return false;
+
+	switch (in->req_type) {
+	/*
+	 * Compare struct members manually for request types which can't be
+	 * compared simply using memcmp(). This is because said request types
+	 * contain pointers to other allocated structs
+	 */
+	case DP_REMOTE_I2C_READ:
+#define IN in->u.i2c_read
+#define OUT out->u.i2c_read
+		if (IN.num_bytes_read != OUT.num_bytes_read ||
+		    IN.num_transactions != OUT.num_transactions ||
+		    IN.port_number != OUT.port_number ||
+		    IN.read_i2c_device_id != OUT.read_i2c_device_id)
+			return false;
+
+		for (i = 0; i < IN.num_transactions; i++) {
+			txin = &IN.transactions[i];
+			txout = &OUT.transactions[i];
+
+			if (txin->i2c_dev_id != txout->i2c_dev_id ||
+			    txin->no_stop_bit != txout->no_stop_bit ||
+			    txin->num_bytes != txout->num_bytes ||
+			    txin->i2c_transaction_delay !=
+			    txout->i2c_transaction_delay)
+				return false;
+
+			if (memcmp(txin->bytes, txout->bytes,
+				   txin->num_bytes) != 0)
+				return false;
+		}
+		break;
+#undef IN
+#undef OUT
+
+	case DP_REMOTE_DPCD_WRITE:
+#define IN in->u.dpcd_write
+#define OUT out->u.dpcd_write
+		if (IN.dpcd_address != OUT.dpcd_address ||
+		    IN.num_bytes != OUT.num_bytes ||
+		    IN.port_number != OUT.port_number)
+			return false;
+
+		return memcmp(IN.bytes, OUT.bytes, IN.num_bytes) == 0;
+#undef IN
+#undef OUT
+
+	case DP_REMOTE_I2C_WRITE:
+#define IN in->u.i2c_write
+#define OUT out->u.i2c_write
+		if (IN.port_number != OUT.port_number ||
+		    IN.write_i2c_device_id != OUT.write_i2c_device_id ||
+		    IN.num_bytes != OUT.num_bytes)
+			return false;
+
+		return memcmp(IN.bytes, OUT.bytes, IN.num_bytes) == 0;
+#undef IN
+#undef OUT
+
+	default:
+		return memcmp(in, out, sizeof(*in)) == 0;
+	}
+
+	return true;
+}
+
+static bool
+sideband_msg_req_encode_decode(struct drm_dp_sideband_msg_req_body *in)
+{
+	struct drm_dp_sideband_msg_req_body out = {0};
+	struct drm_printer p = drm_err_printer(PREFIX_STR);
+	struct drm_dp_sideband_msg_tx txmsg;
+	int i, ret;
+
+	drm_dp_encode_sideband_req(in, &txmsg);
+	ret = drm_dp_decode_sideband_req(&txmsg, &out);
+	if (ret < 0) {
+		drm_printf(&p, "Failed to decode sideband request: %d\n",
+			   ret);
+		return false;
+	}
+
+	if (!sideband_msg_req_equal(in, &out)) {
+		drm_printf(&p, "Encode/decode failed, expected:\n");
+		drm_dp_dump_sideband_msg_req_body(in, 1, &p);
+		drm_printf(&p, "Got:\n");
+		drm_dp_dump_sideband_msg_req_body(&out, 1, &p);
+		return false;
+	}
+
+	switch (in->req_type) {
+	case DP_REMOTE_DPCD_WRITE:
+		kfree(out.u.dpcd_write.bytes);
+		break;
+	case DP_REMOTE_I2C_READ:
+		for (i = 0; i < out.u.i2c_read.num_transactions; i++)
+			kfree(out.u.i2c_read.transactions[i].bytes);
+		break;
+	case DP_REMOTE_I2C_WRITE:
+		kfree(out.u.i2c_write.bytes);
+		break;
+	}
+
+	/* Clear everything but the req_type for the input */
+	memset(&in->u, 0, sizeof(in->u));
+
+	return true;
+}
+
+int igt_dp_mst_sideband_msg_req_decode(void *unused)
+{
+	struct drm_dp_sideband_msg_req_body in = { 0 };
+	u8 data[] = { 0xff, 0x0, 0xdd };
+	int i;
+
+#define DO_TEST() FAIL_ON(!sideband_msg_req_encode_decode(&in))
+
+	in.req_type = DP_ENUM_PATH_RESOURCES;
+	in.u.port_num.port_number = 5;
+	DO_TEST();
+
+	in.req_type = DP_POWER_UP_PHY;
+	in.u.port_num.port_number = 5;
+	DO_TEST();
+
+	in.req_type = DP_POWER_DOWN_PHY;
+	in.u.port_num.port_number = 5;
+	DO_TEST();
+
+	in.req_type = DP_ALLOCATE_PAYLOAD;
+	in.u.allocate_payload.number_sdp_streams = 3;
+	for (i = 0; i < in.u.allocate_payload.number_sdp_streams; i++)
+		in.u.allocate_payload.sdp_stream_sink[i] = i + 1;
+	DO_TEST();
+	in.u.allocate_payload.port_number = 0xf;
+	DO_TEST();
+	in.u.allocate_payload.vcpi = 0x7f;
+	DO_TEST();
+	in.u.allocate_payload.pbn = U16_MAX;
+	DO_TEST();
+
+	in.req_type = DP_QUERY_PAYLOAD;
+	in.u.query_payload.port_number = 0xf;
+	DO_TEST();
+	in.u.query_payload.vcpi = 0x7f;
+	DO_TEST();
+
+	in.req_type = DP_REMOTE_DPCD_READ;
+	in.u.dpcd_read.port_number = 0xf;
+	DO_TEST();
+	in.u.dpcd_read.dpcd_address = 0xfedcb;
+	DO_TEST();
+	in.u.dpcd_read.num_bytes = U8_MAX;
+	DO_TEST();
+
+	in.req_type = DP_REMOTE_DPCD_WRITE;
+	in.u.dpcd_write.port_number = 0xf;
+	DO_TEST();
+	in.u.dpcd_write.dpcd_address = 0xfedcb;
+	DO_TEST();
+	in.u.dpcd_write.num_bytes = ARRAY_SIZE(data);
+	in.u.dpcd_write.bytes = data;
+	DO_TEST();
+
+	in.req_type = DP_REMOTE_I2C_READ;
+	in.u.i2c_read.port_number = 0xf;
+	DO_TEST();
+	in.u.i2c_read.read_i2c_device_id = 0x7f;
+	DO_TEST();
+	in.u.i2c_read.num_transactions = 3;
+	in.u.i2c_read.num_bytes_read = ARRAY_SIZE(data) * 3;
+	for (i = 0; i < in.u.i2c_read.num_transactions; i++) {
+		in.u.i2c_read.transactions[i].bytes = data;
+		in.u.i2c_read.transactions[i].num_bytes = ARRAY_SIZE(data);
+		in.u.i2c_read.transactions[i].i2c_dev_id = 0x7f & ~i;
+		in.u.i2c_read.transactions[i].i2c_transaction_delay = 0xf & ~i;
+	}
+	DO_TEST();
+
+	in.req_type = DP_REMOTE_I2C_WRITE;
+	in.u.i2c_write.port_number = 0xf;
+	DO_TEST();
+	in.u.i2c_write.write_i2c_device_id = 0x7f;
+	DO_TEST();
+	in.u.i2c_write.num_bytes = ARRAY_SIZE(data);
+	in.u.i2c_write.bytes = data;
+	DO_TEST();
+
+#undef DO_TEST
+	return 0;
+}
diff --git a/drivers/gpu/drm/selftests/test-drm_framebuffer.c b/drivers/gpu/drm/selftests/test-drm_framebuffer.c
index 74d5561a862b..2d29ea6f92e2 100644
--- a/drivers/gpu/drm/selftests/test-drm_framebuffer.c
+++ b/drivers/gpu/drm/selftests/test-drm_framebuffer.c
@@ -126,7 +126,7 @@ static struct drm_framebuffer_test createbuffer_tests[] = {
 		 .handles = { 1, 1, 0 }, .pitches = { MAX_WIDTH, MAX_WIDTH - 1, 0 },
 	}
 },
-{ .buffer_created = 0, .name = "NV12 Invalid modifier/misssing DRM_MODE_FB_MODIFIERS flag",
+{ .buffer_created = 0, .name = "NV12 Invalid modifier/missing DRM_MODE_FB_MODIFIERS flag",
 	.cmd = { .width = MAX_WIDTH, .height = MAX_HEIGHT, .pixel_format = DRM_FORMAT_NV12,
 		 .handles = { 1, 1, 0 }, .modifier = { DRM_FORMAT_MOD_SAMSUNG_64_32_TILE, 0, 0 },
 		 .pitches = { MAX_WIDTH, MAX_WIDTH, 0 },
diff --git a/drivers/gpu/drm/selftests/test-drm_mm.c b/drivers/gpu/drm/selftests/test-drm_mm.c
index 388f9844f4ba..9aabe82dcd3a 100644
--- a/drivers/gpu/drm/selftests/test-drm_mm.c
+++ b/drivers/gpu/drm/selftests/test-drm_mm.c
@@ -854,7 +854,7 @@ static bool assert_contiguous_in_range(struct drm_mm *mm,
 
 	if (start > 0) {
 		node = __drm_mm_interval_first(mm, 0, start - 1);
-		if (node->allocated) {
+		if (drm_mm_node_allocated(node)) {
 			pr_err("node before start: node=%llx+%llu, start=%llx\n",
 			       node->start, node->size, start);
 			return false;
@@ -863,7 +863,7 @@ static bool assert_contiguous_in_range(struct drm_mm *mm,
 
 	if (end < U64_MAX) {
 		node = __drm_mm_interval_first(mm, end, U64_MAX);
-		if (node->allocated) {
+		if (drm_mm_node_allocated(node)) {
 			pr_err("node after end: node=%llx+%llu, end=%llx\n",
 			       node->start, node->size, end);
 			return false;
@@ -1156,12 +1156,12 @@ static void show_holes(const struct drm_mm *mm, int count)
 		struct drm_mm_node *next = list_next_entry(hole, node_list);
 		const char *node1 = NULL, *node2 = NULL;
 
-		if (hole->allocated)
+		if (drm_mm_node_allocated(hole))
 			node1 = kasprintf(GFP_KERNEL,
 					  "[%llx + %lld, color=%ld], ",
 					  hole->start, hole->size, hole->color);
 
-		if (next->allocated)
+		if (drm_mm_node_allocated(next))
 			node2 = kasprintf(GFP_KERNEL,
 					  ", [%llx + %lld, color=%ld]",
 					  next->start, next->size, next->color);
@@ -1900,18 +1900,18 @@ static void separate_adjacent_colors(const struct drm_mm_node *node,
 				     u64 *start,
 				     u64 *end)
 {
-	if (node->allocated && node->color != color)
+	if (drm_mm_node_allocated(node) && node->color != color)
 		++*start;
 
 	node = list_next_entry(node, node_list);
-	if (node->allocated && node->color != color)
+	if (drm_mm_node_allocated(node) && node->color != color)
 		--*end;
 }
 
 static bool colors_abutt(const struct drm_mm_node *node)
 {
 	if (!drm_mm_hole_follows(node) &&
-	    list_next_entry(node, node_list)->allocated) {
+	    drm_mm_node_allocated(list_next_entry(node, node_list))) {
 		pr_err("colors abutt; %ld [%llx + %llx] is next to %ld [%llx + %llx]!\n",
 		       node->color, node->start, node->size,
 		       list_next_entry(node, node_list)->color,
diff --git a/drivers/gpu/drm/selftests/test-drm_modeset_common.h b/drivers/gpu/drm/selftests/test-drm_modeset_common.h
index 8c76f09c12d1..0fcb8bbc6a1b 100644
--- a/drivers/gpu/drm/selftests/test-drm_modeset_common.h
+++ b/drivers/gpu/drm/selftests/test-drm_modeset_common.h
@@ -39,5 +39,7 @@ int igt_damage_iter_damage_one_intersect(void *ignored);
 int igt_damage_iter_damage_one_outside(void *ignored);
 int igt_damage_iter_damage_src_moved(void *ignored);
 int igt_damage_iter_damage_not_visible(void *ignored);
+int igt_dp_mst_calc_pbn_mode(void *ignored);
+int igt_dp_mst_sideband_msg_req_decode(void *ignored);
 
 #endif
diff --git a/drivers/gpu/drm/sti/sti_cursor.c b/drivers/gpu/drm/sti/sti_cursor.c
index 0bf7c332cf0b..ea64c1dcaf63 100644
--- a/drivers/gpu/drm/sti/sti_cursor.c
+++ b/drivers/gpu/drm/sti/sti_cursor.c
@@ -47,7 +47,7 @@ struct dma_pixmap {
 	void *base;
 };
 
-/**
+/*
  * STI Cursor structure
  *
  * @sti_plane:    sti_plane structure
diff --git a/drivers/gpu/drm/sti/sti_dvo.c b/drivers/gpu/drm/sti/sti_dvo.c
index e55870190bf5..68289b0b063a 100644
--- a/drivers/gpu/drm/sti/sti_dvo.c
+++ b/drivers/gpu/drm/sti/sti_dvo.c
@@ -12,6 +12,7 @@
 #include <linux/platform_device.h>
 
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_device.h>
 #include <drm/drm_panel.h>
 #include <drm/drm_print.h>
@@ -65,7 +66,7 @@ static struct dvo_config rgb_24bit_de_cfg = {
 	.awg_fwgen_fct = sti_awg_generate_code_data_enable_mode,
 };
 
-/**
+/*
  * STI digital video output structure
  *
  * @dev: driver device
diff --git a/drivers/gpu/drm/sti/sti_gdp.c b/drivers/gpu/drm/sti/sti_gdp.c
index 8e926cd6a1c8..11595c748844 100644
--- a/drivers/gpu/drm/sti/sti_gdp.c
+++ b/drivers/gpu/drm/sti/sti_gdp.c
@@ -103,7 +103,7 @@ struct sti_gdp_node_list {
 	dma_addr_t btm_field_paddr;
 };
 
-/**
+/*
  * STI GDP structure
  *
  * @sti_plane:          sti_plane structure
diff --git a/drivers/gpu/drm/sti/sti_hda.c b/drivers/gpu/drm/sti/sti_hda.c
index 94e404f13234..8f7bf33815fd 100644
--- a/drivers/gpu/drm/sti/sti_hda.c
+++ b/drivers/gpu/drm/sti/sti_hda.c
@@ -12,6 +12,7 @@
 #include <linux/seq_file.h>
 
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_debugfs.h>
 #include <drm/drm_device.h>
 #include <drm/drm_file.h>
@@ -230,7 +231,7 @@ static const struct sti_hda_video_config hda_supported_modes[] = {
 	 AWGi_720x480p_60, NN_720x480p_60, VID_ED}
 };
 
-/**
+/*
  * STI hd analog structure
  *
  * @dev: driver device
diff --git a/drivers/gpu/drm/sti/sti_hdmi.c b/drivers/gpu/drm/sti/sti_hdmi.c
index 9862c322f0c4..814560ead4e1 100644
--- a/drivers/gpu/drm/sti/sti_hdmi.c
+++ b/drivers/gpu/drm/sti/sti_hdmi.c
@@ -9,11 +9,12 @@
 #include <linux/debugfs.h>
 #include <linux/hdmi.h>
 #include <linux/module.h>
-#include <linux/of_gpio.h>
+#include <linux/io.h>
 #include <linux/platform_device.h>
 #include <linux/reset.h>
 
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_debugfs.h>
 #include <drm/drm_drv.h>
 #include <drm/drm_edid.h>
@@ -333,7 +334,6 @@ static void hdmi_infoframe_reset(struct sti_hdmi *hdmi,
  * Helper to concatenate infoframe in 32 bits word
  *
  * @ptr: pointer on the hdmi internal structure
- * @data: infoframe to write
  * @size: size to write
  */
 static inline unsigned int hdmi_infoframe_subpack(const u8 *ptr, size_t size)
@@ -543,13 +543,14 @@ static int hdmi_vendor_infoframe_config(struct sti_hdmi *hdmi)
 	return 0;
 }
 
+#define HDMI_TIMEOUT_SWRESET  100   /*milliseconds */
+
 /**
  * Software reset of the hdmi subsystem
  *
  * @hdmi: pointer on the hdmi internal structure
  *
  */
-#define HDMI_TIMEOUT_SWRESET  100   /*milliseconds */
 static void hdmi_swreset(struct sti_hdmi *hdmi)
 {
 	u32 val;
@@ -1256,6 +1257,7 @@ static int sti_hdmi_bind(struct device *dev, struct device *master, void *data)
 	struct drm_device *drm_dev = data;
 	struct drm_encoder *encoder;
 	struct sti_hdmi_connector *connector;
+	struct cec_connector_info conn_info;
 	struct drm_connector *drm_connector;
 	struct drm_bridge *bridge;
 	int err;
@@ -1318,6 +1320,14 @@ static int sti_hdmi_bind(struct device *dev, struct device *master, void *data)
 		goto err_sysfs;
 	}
 
+	cec_fill_conn_info_from_drm(&conn_info, drm_connector);
+	hdmi->notifier = cec_notifier_conn_register(&hdmi->dev, NULL,
+						    &conn_info);
+	if (!hdmi->notifier) {
+		hdmi->drm_connector = NULL;
+		return -ENOMEM;
+	}
+
 	/* Enable default interrupts */
 	hdmi_write(hdmi, HDMI_DEFAULT_INT, HDMI_INT_EN);
 
@@ -1331,6 +1341,9 @@ err_sysfs:
 static void sti_hdmi_unbind(struct device *dev,
 		struct device *master, void *data)
 {
+	struct sti_hdmi *hdmi = dev_get_drvdata(dev);
+
+	cec_notifier_conn_unregister(hdmi->notifier);
 }
 
 static const struct component_ops sti_hdmi_ops = {
@@ -1436,10 +1449,6 @@ static int sti_hdmi_probe(struct platform_device *pdev)
 		goto release_adapter;
 	}
 
-	hdmi->notifier = cec_notifier_get(&pdev->dev);
-	if (!hdmi->notifier)
-		goto release_adapter;
-
 	hdmi->reset = devm_reset_control_get(dev, "hdmi");
 	/* Take hdmi out of reset */
 	if (!IS_ERR(hdmi->reset))
@@ -1459,14 +1468,11 @@ static int sti_hdmi_remove(struct platform_device *pdev)
 {
 	struct sti_hdmi *hdmi = dev_get_drvdata(&pdev->dev);
 
-	cec_notifier_set_phys_addr(hdmi->notifier, CEC_PHYS_ADDR_INVALID);
-
 	i2c_put_adapter(hdmi->ddc_adapt);
 	if (hdmi->audio_pdev)
 		platform_device_unregister(hdmi->audio_pdev);
 	component_del(&pdev->dev, &sti_hdmi_ops);
 
-	cec_notifier_put(hdmi->notifier);
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/sti/sti_tvout.c b/drivers/gpu/drm/sti/sti_tvout.c
index aba79c172512..5767e93dd1cd 100644
--- a/drivers/gpu/drm/sti/sti_tvout.c
+++ b/drivers/gpu/drm/sti/sti_tvout.c
@@ -157,9 +157,9 @@ static void tvout_write(struct sti_tvout *tvout, u32 val, int offset)
  *
  * @tvout: tvout structure
  * @reg: register to set
- * @cr_r:
- * @y_g:
- * @cb_b:
+ * @cr_r: red chroma or red order
+ * @y_g: y or green order
+ * @cb_b: blue chroma or blue order
  */
 static void tvout_vip_set_color_order(struct sti_tvout *tvout, int reg,
 				      u32 cr_r, u32 y_g, u32 cb_b)
@@ -214,7 +214,7 @@ static void tvout_vip_set_rnd(struct sti_tvout *tvout, int reg, u32 rnd)
  * @tvout: tvout structure
  * @reg: register to set
  * @main_path: main or auxiliary path
- * @sel_input: selected_input (main/aux + conv)
+ * @video_out: selected_input (main/aux + conv)
  */
 static void tvout_vip_set_sel_input(struct sti_tvout *tvout,
 				    int reg,
@@ -251,7 +251,7 @@ static void tvout_vip_set_sel_input(struct sti_tvout *tvout,
  *
  * @tvout: tvout structure
  * @reg: register to set
- * @in_vid_signed: used video input format
+ * @in_vid_fmt: used video input format
  */
 static void tvout_vip_set_in_vid_fmt(struct sti_tvout *tvout,
 		int reg, u32 in_vid_fmt)
diff --git a/drivers/gpu/drm/sti/sti_vtg.c b/drivers/gpu/drm/sti/sti_vtg.c
index ef4009f11396..0b17ac8a3faa 100644
--- a/drivers/gpu/drm/sti/sti_vtg.c
+++ b/drivers/gpu/drm/sti/sti_vtg.c
@@ -121,7 +121,7 @@ struct sti_vtg_sync_params {
 	u32 vsync_off_bot;
 };
 
-/**
+/*
  * STI VTG structure
  *
  * @regs: register mapping
diff --git a/drivers/gpu/drm/stm/dw_mipi_dsi-stm.c b/drivers/gpu/drm/stm/dw_mipi_dsi-stm.c
index a03a642c147c..514efefb0016 100644
--- a/drivers/gpu/drm/stm/dw_mipi_dsi-stm.c
+++ b/drivers/gpu/drm/stm/dw_mipi_dsi-stm.c
@@ -260,8 +260,11 @@ dw_mipi_dsi_get_lane_mbps(void *priv_data, const struct drm_display_mode *mode,
 	/* Compute requested pll out */
 	bpp = mipi_dsi_pixel_format_to_bpp(format);
 	pll_out_khz = mode->clock * bpp / lanes;
+
 	/* Add 20% to pll out to be higher than pixel bw (burst mode only) */
-	pll_out_khz = (pll_out_khz * 12) / 10;
+	if (mode_flags & MIPI_DSI_MODE_VIDEO_BURST)
+		pll_out_khz = (pll_out_khz * 12) / 10;
+
 	if (pll_out_khz > dsi->lane_max_kbps) {
 		pll_out_khz = dsi->lane_max_kbps;
 		DRM_WARN("Warning max phy mbps is used\n");
diff --git a/drivers/gpu/drm/stm/ltdc.c b/drivers/gpu/drm/stm/ltdc.c
index 3ab4fbf8eb0d..5b51298921cf 100644
--- a/drivers/gpu/drm/stm/ltdc.c
+++ b/drivers/gpu/drm/stm/ltdc.c
@@ -15,6 +15,7 @@
 #include <linux/module.h>
 #include <linux/of_address.h>
 #include <linux/of_graph.h>
+#include <linux/pinctrl/consumer.h>
 #include <linux/platform_device.h>
 #include <linux/pm_runtime.h>
 #include <linux/reset.h>
@@ -1040,6 +1041,36 @@ static const struct drm_encoder_funcs ltdc_encoder_funcs = {
 	.destroy = drm_encoder_cleanup,
 };
 
+static void ltdc_encoder_disable(struct drm_encoder *encoder)
+{
+	struct drm_device *ddev = encoder->dev;
+
+	DRM_DEBUG_DRIVER("\n");
+
+	/* Set to sleep state the pinctrl whatever type of encoder */
+	pinctrl_pm_select_sleep_state(ddev->dev);
+}
+
+static void ltdc_encoder_enable(struct drm_encoder *encoder)
+{
+	struct drm_device *ddev = encoder->dev;
+
+	DRM_DEBUG_DRIVER("\n");
+
+	/*
+	 * Set to default state the pinctrl only with DPI type.
+	 * Others types like DSI, don't need pinctrl due to
+	 * internal bridge (the signals do not come out of the chipset).
+	 */
+	if (encoder->encoder_type == DRM_MODE_ENCODER_DPI)
+		pinctrl_pm_select_default_state(ddev->dev);
+}
+
+static const struct drm_encoder_helper_funcs ltdc_encoder_helper_funcs = {
+	.disable = ltdc_encoder_disable,
+	.enable = ltdc_encoder_enable,
+};
+
 static int ltdc_encoder_init(struct drm_device *ddev, struct drm_bridge *bridge)
 {
 	struct drm_encoder *encoder;
@@ -1055,6 +1086,8 @@ static int ltdc_encoder_init(struct drm_device *ddev, struct drm_bridge *bridge)
 	drm_encoder_init(ddev, encoder, &ltdc_encoder_funcs,
 			 DRM_MODE_ENCODER_DPI, NULL);
 
+	drm_encoder_helper_add(encoder, &ltdc_encoder_helper_funcs);
+
 	ret = drm_bridge_attach(encoder, bridge, NULL);
 	if (ret) {
 		drm_encoder_cleanup(encoder);
@@ -1236,8 +1269,8 @@ int ltdc_load(struct drm_device *ddev)
 	/* Add endpoints panels or bridges if any */
 	for (i = 0; i < MAX_ENDPOINTS; i++) {
 		if (panel[i]) {
-			bridge[i] = drm_panel_bridge_add(panel[i],
-							DRM_MODE_CONNECTOR_DPI);
+			bridge[i] = drm_panel_bridge_add_typed(panel[i],
+							       DRM_MODE_CONNECTOR_DPI);
 			if (IS_ERR(bridge[i])) {
 				DRM_ERROR("panel-bridge endpoint %d\n", i);
 				ret = PTR_ERR(bridge[i]);
@@ -1280,6 +1313,8 @@ int ltdc_load(struct drm_device *ddev)
 
 	clk_disable_unprepare(ldev->pixel_clk);
 
+	pinctrl_pm_select_sleep_state(ddev->dev);
+
 	pm_runtime_enable(ddev->dev);
 
 	return 0;
diff --git a/drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c b/drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c
index eb8071a4d6d0..a7c4654445c7 100644
--- a/drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c
+++ b/drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c
@@ -490,6 +490,7 @@ static int sun4i_hdmi_bind(struct device *dev, struct device *master,
 {
 	struct platform_device *pdev = to_platform_device(dev);
 	struct drm_device *drm = data;
+	struct cec_connector_info conn_info;
 	struct sun4i_drv *drv = drm->dev_private;
 	struct sun4i_hdmi *hdmi;
 	struct resource *res;
@@ -629,8 +630,7 @@ static int sun4i_hdmi_bind(struct device *dev, struct device *master,
 
 #ifdef CONFIG_DRM_SUN4I_HDMI_CEC
 	hdmi->cec_adap = cec_pin_allocate_adapter(&sun4i_hdmi_cec_pin_ops,
-		hdmi, "sun4i", CEC_CAP_TRANSMIT | CEC_CAP_LOG_ADDRS |
-		CEC_CAP_PASSTHROUGH | CEC_CAP_RC);
+		hdmi, "sun4i", CEC_CAP_DEFAULTS | CEC_CAP_CONNECTOR_INFO);
 	ret = PTR_ERR_OR_ZERO(hdmi->cec_adap);
 	if (ret < 0)
 		goto err_cleanup_connector;
@@ -649,6 +649,8 @@ static int sun4i_hdmi_bind(struct device *dev, struct device *master,
 			"Couldn't initialise the HDMI connector\n");
 		goto err_cleanup_connector;
 	}
+	cec_fill_conn_info_from_drm(&conn_info, &hdmi->connector);
+	cec_s_conn_info(hdmi->cec_adap, &conn_info);
 
 	/* There is no HPD interrupt, so we need to poll the controller */
 	hdmi->connector.polled = DRM_CONNECTOR_POLL_CONNECT |
diff --git a/drivers/gpu/drm/sun4i/sun4i_lvds.c b/drivers/gpu/drm/sun4i/sun4i_lvds.c
index 7fbf425acb55..25ab2ef6d545 100644
--- a/drivers/gpu/drm/sun4i/sun4i_lvds.c
+++ b/drivers/gpu/drm/sun4i/sun4i_lvds.c
@@ -7,6 +7,7 @@
 #include <linux/clk.h>
 
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_of.h>
 #include <drm/drm_panel.h>
 #include <drm/drm_print.h>
diff --git a/drivers/gpu/drm/sun4i/sun4i_rgb.c b/drivers/gpu/drm/sun4i/sun4i_rgb.c
index aac56983f208..e74b9eddca01 100644
--- a/drivers/gpu/drm/sun4i/sun4i_rgb.c
+++ b/drivers/gpu/drm/sun4i/sun4i_rgb.c
@@ -9,6 +9,7 @@
 #include <linux/clk.h>
 
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_of.h>
 #include <drm/drm_panel.h>
 #include <drm/drm_print.h>
diff --git a/drivers/gpu/drm/sun4i/sun4i_tcon.c b/drivers/gpu/drm/sun4i/sun4i_tcon.c
index 04c721d0d3b9..42651d737c55 100644
--- a/drivers/gpu/drm/sun4i/sun4i_tcon.c
+++ b/drivers/gpu/drm/sun4i/sun4i_tcon.c
@@ -16,6 +16,7 @@
 #include <linux/reset.h>
 
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_connector.h>
 #include <drm/drm_crtc.h>
 #include <drm/drm_encoder.h>
@@ -488,7 +489,7 @@ static void sun4i_tcon0_mode_set_rgb(struct sun4i_tcon *tcon,
 
 	WARN_ON(!tcon->quirks->has_channel_0);
 
-	tcon->dclk_min_div = 6;
+	tcon->dclk_min_div = 1;
 	tcon->dclk_max_div = 127;
 	sun4i_tcon0_mode_set_common(tcon, mode);
 
diff --git a/drivers/gpu/drm/sun4i/sun6i_mipi_dsi.c b/drivers/gpu/drm/sun4i/sun6i_mipi_dsi.c
index 1636344ba9ec..c958ca9bae63 100644
--- a/drivers/gpu/drm/sun4i/sun6i_mipi_dsi.c
+++ b/drivers/gpu/drm/sun4i/sun6i_mipi_dsi.c
@@ -16,6 +16,7 @@
 #include <linux/platform_device.h>
 #include <linux/pm_runtime.h>
 #include <linux/regmap.h>
+#include <linux/regulator/consumer.h>
 #include <linux/reset.h>
 #include <linux/slab.h>
 
@@ -365,8 +366,7 @@ static void sun6i_dsi_inst_init(struct sun6i_dsi *dsi,
 static u16 sun6i_dsi_get_video_start_delay(struct sun6i_dsi *dsi,
 					   struct drm_display_mode *mode)
 {
-	u16 start = clamp(mode->vtotal - mode->vdisplay - 10, 8, 100);
-	u16 delay = mode->vtotal - (mode->vsync_end - mode->vdisplay) + start;
+	u16 delay = mode->vtotal - (mode->vsync_start - mode->vdisplay) + 1;
 
 	if (delay > mode->vtotal)
 		delay = delay % mode->vtotal;
@@ -437,9 +437,9 @@ static void sun6i_dsi_setup_burst(struct sun6i_dsi *dsi,
 			     SUN6I_DSI_BURST_LINE_SYNC_POINT(SUN6I_DSI_SYNC_POINT));
 
 		val = SUN6I_DSI_TCON_DRQ_ENABLE_MODE;
-	} else if ((mode->hsync_end - mode->hdisplay) > 20) {
+	} else if ((mode->hsync_start - mode->hdisplay) > 20) {
 		/* Maaaaaagic */
-		u16 drq = (mode->hsync_end - mode->hdisplay) - 20;
+		u16 drq = (mode->hsync_start - mode->hdisplay) - 20;
 
 		drq *= mipi_dsi_pixel_format_to_bpp(device->format);
 		drq /= 32;
@@ -569,11 +569,12 @@ static void sun6i_dsi_setup_timings(struct sun6i_dsi *dsi,
 			  (mode->htotal - mode->hsync_end) * Bpp - HBP_PACKET_OVERHEAD);
 
 		/*
-		 * The frontporch is set using a blanking packet (4
-		 * bytes + payload + 2 bytes). Its minimal size is
-		 * therefore 6 bytes
+		 * The frontporch is set using a sync event (4 bytes)
+		 * and two blanking packets (each one is 4 bytes +
+		 * payload + 2 bytes). Its minimal size is therefore
+		 * 16 bytes
 		 */
-#define HFP_PACKET_OVERHEAD	6
+#define HFP_PACKET_OVERHEAD	16
 		hfp = max((unsigned int)HFP_PACKET_OVERHEAD,
 			  (mode->hsync_start - mode->hdisplay) * Bpp - HFP_PACKET_OVERHEAD);
 
@@ -831,8 +832,8 @@ static u32 sun6i_dsi_dcs_build_pkt_hdr(struct sun6i_dsi *dsi,
 	u32 pkt = msg->type;
 
 	if (msg->type == MIPI_DSI_DCS_LONG_WRITE) {
-		pkt |= ((msg->tx_len + 1) & 0xffff) << 8;
-		pkt |= (((msg->tx_len + 1) >> 8) & 0xffff) << 16;
+		pkt |= ((msg->tx_len) & 0xffff) << 8;
+		pkt |= (((msg->tx_len) >> 8) & 0xffff) << 16;
 	} else {
 		pkt |= (((u8 *)msg->tx_buf)[0] << 8);
 		if (msg->tx_len > 1)
@@ -1100,6 +1101,12 @@ static int sun6i_dsi_probe(struct platform_device *pdev)
 		return PTR_ERR(base);
 	}
 
+	dsi->regulator = devm_regulator_get(dev, "vcc-dsi");
+	if (IS_ERR(dsi->regulator)) {
+		dev_err(dev, "Couldn't get VCC-DSI supply\n");
+		return PTR_ERR(dsi->regulator);
+	}
+
 	dsi->regs = devm_regmap_init_mmio_clk(dev, "bus", base,
 					      &sun6i_dsi_regmap_config);
 	if (IS_ERR(dsi->regs)) {
@@ -1173,6 +1180,13 @@ static int sun6i_dsi_remove(struct platform_device *pdev)
 static int __maybe_unused sun6i_dsi_runtime_resume(struct device *dev)
 {
 	struct sun6i_dsi *dsi = dev_get_drvdata(dev);
+	int err;
+
+	err = regulator_enable(dsi->regulator);
+	if (err) {
+		dev_err(dsi->dev, "failed to enable VCC-DSI supply: %d\n", err);
+		return err;
+	}
 
 	reset_control_deassert(dsi->reset);
 	clk_prepare_enable(dsi->mod_clk);
@@ -1205,6 +1219,7 @@ static int __maybe_unused sun6i_dsi_runtime_suspend(struct device *dev)
 
 	clk_disable_unprepare(dsi->mod_clk);
 	reset_control_assert(dsi->reset);
+	regulator_disable(dsi->regulator);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/sun4i/sun6i_mipi_dsi.h b/drivers/gpu/drm/sun4i/sun6i_mipi_dsi.h
index 5c3ad5be0690..3f4846f581ef 100644
--- a/drivers/gpu/drm/sun4i/sun6i_mipi_dsi.h
+++ b/drivers/gpu/drm/sun4i/sun6i_mipi_dsi.h
@@ -23,6 +23,7 @@ struct sun6i_dsi {
 	struct clk		*bus_clk;
 	struct clk		*mod_clk;
 	struct regmap		*regs;
+	struct regulator	*regulator;
 	struct reset_control	*reset;
 	struct phy		*dphy;
 
diff --git a/drivers/gpu/drm/sun4i/sun8i_dw_hdmi.c b/drivers/gpu/drm/sun4i/sun8i_dw_hdmi.c
index a44dca4b0219..e8a317d5ba19 100644
--- a/drivers/gpu/drm/sun4i/sun8i_dw_hdmi.c
+++ b/drivers/gpu/drm/sun4i/sun8i_dw_hdmi.c
@@ -226,6 +226,7 @@ static int sun8i_dw_hdmi_bind(struct device *dev, struct device *master,
 	sun8i_hdmi_phy_init(hdmi->phy);
 
 	plat_data->mode_valid = hdmi->quirks->mode_valid;
+	plat_data->use_drm_infoframe = hdmi->quirks->use_drm_infoframe;
 	sun8i_hdmi_phy_set_ops(hdmi->phy, plat_data);
 
 	platform_set_drvdata(pdev, hdmi);
@@ -300,6 +301,7 @@ static const struct sun8i_dw_hdmi_quirks sun8i_a83t_quirks = {
 
 static const struct sun8i_dw_hdmi_quirks sun50i_h6_quirks = {
 	.mode_valid = sun8i_dw_hdmi_mode_valid_h6,
+	.use_drm_infoframe = true,
 };
 
 static const struct of_device_id sun8i_dw_hdmi_dt_ids[] = {
diff --git a/drivers/gpu/drm/sun4i/sun8i_dw_hdmi.h b/drivers/gpu/drm/sun4i/sun8i_dw_hdmi.h
index d707c9171824..8e64945167e9 100644
--- a/drivers/gpu/drm/sun4i/sun8i_dw_hdmi.h
+++ b/drivers/gpu/drm/sun4i/sun8i_dw_hdmi.h
@@ -179,6 +179,7 @@ struct sun8i_dw_hdmi_quirks {
 	enum drm_mode_status (*mode_valid)(struct drm_connector *connector,
 					   const struct drm_display_mode *mode);
 	unsigned int set_rate : 1;
+	unsigned int use_drm_infoframe : 1;
 };
 
 struct sun8i_dw_hdmi {
diff --git a/drivers/gpu/drm/tegra/Kconfig b/drivers/gpu/drm/tegra/Kconfig
index 1d1269fde3c1..5043dcaf1cf9 100644
--- a/drivers/gpu/drm/tegra/Kconfig
+++ b/drivers/gpu/drm/tegra/Kconfig
@@ -9,7 +9,7 @@ config DRM_TEGRA
 	select DRM_MIPI_DSI
 	select DRM_PANEL
 	select TEGRA_HOST1X
-	select IOMMU_IOVA if IOMMU_SUPPORT
+	select IOMMU_IOVA
 	select CEC_CORE if CEC_NOTIFIER
 	help
 	  Choose this option if you have an NVIDIA Tegra SoC.
diff --git a/drivers/gpu/drm/tegra/Makefile b/drivers/gpu/drm/tegra/Makefile
index 33c463e8d49f..d6cf202414f0 100644
--- a/drivers/gpu/drm/tegra/Makefile
+++ b/drivers/gpu/drm/tegra/Makefile
@@ -5,6 +5,7 @@ tegra-drm-y := \
 	drm.o \
 	gem.o \
 	fb.o \
+	dp.o \
 	hub.o \
 	plane.o \
 	dc.o \
diff --git a/drivers/gpu/drm/tegra/dc.c b/drivers/gpu/drm/tegra/dc.c
index fbf57bc3cdab..5b1f9ff97576 100644
--- a/drivers/gpu/drm/tegra/dc.c
+++ b/drivers/gpu/drm/tegra/dc.c
@@ -715,9 +715,7 @@ static void tegra_plane_atomic_update(struct drm_plane *plane,
 	window.swap = state->swap;
 
 	for (i = 0; i < fb->format->num_planes; i++) {
-		struct tegra_bo *bo = tegra_fb_get_plane(fb, i);
-
-		window.base[i] = bo->paddr + fb->offsets[i];
+		window.base[i] = state->iova[i] + fb->offsets[i];
 
 		/*
 		 * Tegra uses a shared stride for UV planes. Framebuffers are
@@ -732,6 +730,8 @@ static void tegra_plane_atomic_update(struct drm_plane *plane,
 }
 
 static const struct drm_plane_helper_funcs tegra_plane_helper_funcs = {
+	.prepare_fb = tegra_plane_prepare_fb,
+	.cleanup_fb = tegra_plane_cleanup_fb,
 	.atomic_check = tegra_plane_atomic_check,
 	.atomic_disable = tegra_plane_atomic_disable,
 	.atomic_update = tegra_plane_atomic_update,
@@ -869,11 +869,11 @@ static void tegra_cursor_atomic_update(struct drm_plane *plane,
 		return;
 	}
 
-	value |= (bo->paddr >> 10) & 0x3fffff;
+	value |= (bo->iova >> 10) & 0x3fffff;
 	tegra_dc_writel(dc, value, DC_DISP_CURSOR_START_ADDR);
 
 #ifdef CONFIG_ARCH_DMA_ADDR_T_64BIT
-	value = (bo->paddr >> 32) & 0x3;
+	value = (bo->iova >> 32) & 0x3;
 	tegra_dc_writel(dc, value, DC_DISP_CURSOR_START_ADDR_HI);
 #endif
 
@@ -914,6 +914,8 @@ static void tegra_cursor_atomic_disable(struct drm_plane *plane,
 }
 
 static const struct drm_plane_helper_funcs tegra_cursor_plane_helper_funcs = {
+	.prepare_fb = tegra_plane_prepare_fb,
+	.cleanup_fb = tegra_plane_cleanup_fb,
 	.atomic_check = tegra_cursor_atomic_check,
 	.atomic_update = tegra_cursor_atomic_update,
 	.atomic_disable = tegra_cursor_atomic_disable,
@@ -2014,9 +2016,8 @@ static int tegra_dc_init(struct host1x_client *client)
 	if (!dc->syncpt)
 		dev_warn(dc->dev, "failed to allocate syncpoint\n");
 
-	dc->group = host1x_client_iommu_attach(client, true);
-	if (IS_ERR(dc->group)) {
-		err = PTR_ERR(dc->group);
+	err = host1x_client_iommu_attach(client);
+	if (err < 0) {
 		dev_err(client->dev, "failed to attach to domain: %d\n", err);
 		return err;
 	}
@@ -2074,6 +2075,12 @@ static int tegra_dc_init(struct host1x_client *client)
 		goto cleanup;
 	}
 
+	/*
+	 * Inherit the DMA parameters (such as maximum segment size) from the
+	 * parent device.
+	 */
+	client->dev->dma_parms = client->parent->dma_parms;
+
 	return 0;
 
 cleanup:
@@ -2083,7 +2090,7 @@ cleanup:
 	if (!IS_ERR(primary))
 		drm_plane_cleanup(primary);
 
-	host1x_client_iommu_detach(client, dc->group);
+	host1x_client_iommu_detach(client);
 	host1x_syncpt_free(dc->syncpt);
 
 	return err;
@@ -2097,6 +2104,9 @@ static int tegra_dc_exit(struct host1x_client *client)
 	if (!tegra_dc_has_window_groups(dc))
 		return 0;
 
+	/* avoid a dangling pointer just in case this disappears */
+	client->dev->dma_parms = NULL;
+
 	devm_free_irq(dc->dev, dc->irq, dc);
 
 	err = tegra_dc_rgb_exit(dc);
@@ -2105,7 +2115,7 @@ static int tegra_dc_exit(struct host1x_client *client)
 		return err;
 	}
 
-	host1x_client_iommu_detach(client, dc->group);
+	host1x_client_iommu_detach(client);
 	host1x_syncpt_free(dc->syncpt);
 
 	return 0;
diff --git a/drivers/gpu/drm/tegra/dc.h b/drivers/gpu/drm/tegra/dc.h
index 0c4d17851f47..3d8ddccd758f 100644
--- a/drivers/gpu/drm/tegra/dc.h
+++ b/drivers/gpu/drm/tegra/dc.h
@@ -90,8 +90,6 @@ struct tegra_dc {
 	struct drm_info_list *debugfs_files;
 
 	const struct tegra_dc_soc_info *soc;
-
-	struct iommu_group *group;
 };
 
 static inline struct tegra_dc *
diff --git a/drivers/gpu/drm/tegra/dp.c b/drivers/gpu/drm/tegra/dp.c
new file mode 100644
index 000000000000..70dfb7d1dec5
--- /dev/null
+++ b/drivers/gpu/drm/tegra/dp.c
@@ -0,0 +1,876 @@
+// SPDX-License-Identifier: MIT
+/*
+ * Copyright (C) 2013-2019 NVIDIA Corporation
+ * Copyright (C) 2015 Rob Clark
+ */
+
+#include <drm/drm_crtc.h>
+#include <drm/drm_dp_helper.h>
+#include <drm/drm_print.h>
+
+#include "dp.h"
+
+static const u8 drm_dp_edp_revisions[] = { 0x11, 0x12, 0x13, 0x14 };
+
+static void drm_dp_link_caps_reset(struct drm_dp_link_caps *caps)
+{
+	caps->enhanced_framing = false;
+	caps->tps3_supported = false;
+	caps->fast_training = false;
+	caps->channel_coding = false;
+	caps->alternate_scrambler_reset = false;
+}
+
+void drm_dp_link_caps_copy(struct drm_dp_link_caps *dest,
+			   const struct drm_dp_link_caps *src)
+{
+	dest->enhanced_framing = src->enhanced_framing;
+	dest->tps3_supported = src->tps3_supported;
+	dest->fast_training = src->fast_training;
+	dest->channel_coding = src->channel_coding;
+	dest->alternate_scrambler_reset = src->alternate_scrambler_reset;
+}
+
+static void drm_dp_link_reset(struct drm_dp_link *link)
+{
+	unsigned int i;
+
+	if (!link)
+		return;
+
+	link->revision = 0;
+	link->max_rate = 0;
+	link->max_lanes = 0;
+
+	drm_dp_link_caps_reset(&link->caps);
+	link->aux_rd_interval.cr = 0;
+	link->aux_rd_interval.ce = 0;
+	link->edp = 0;
+
+	link->rate = 0;
+	link->lanes = 0;
+
+	for (i = 0; i < DP_MAX_SUPPORTED_RATES; i++)
+		link->rates[i] = 0;
+
+	link->num_rates = 0;
+}
+
+/**
+ * drm_dp_link_add_rate() - add a rate to the list of supported rates
+ * @link: the link to add the rate to
+ * @rate: the rate to add
+ *
+ * Add a link rate to the list of supported link rates.
+ *
+ * Returns:
+ * 0 on success or one of the following negative error codes on failure:
+ * - ENOSPC if the maximum number of supported rates has been reached
+ * - EEXISTS if the link already supports this rate
+ *
+ * See also:
+ * drm_dp_link_remove_rate()
+ */
+int drm_dp_link_add_rate(struct drm_dp_link *link, unsigned long rate)
+{
+	unsigned int i, pivot;
+
+	if (link->num_rates == DP_MAX_SUPPORTED_RATES)
+		return -ENOSPC;
+
+	for (pivot = 0; pivot < link->num_rates; pivot++)
+		if (rate <= link->rates[pivot])
+			break;
+
+	if (pivot != link->num_rates && rate == link->rates[pivot])
+		return -EEXIST;
+
+	for (i = link->num_rates; i > pivot; i--)
+		link->rates[i] = link->rates[i - 1];
+
+	link->rates[pivot] = rate;
+	link->num_rates++;
+
+	return 0;
+}
+
+/**
+ * drm_dp_link_remove_rate() - remove a rate from the list of supported rates
+ * @link: the link from which to remove the rate
+ * @rate: the rate to remove
+ *
+ * Removes a link rate from the list of supported link rates.
+ *
+ * Returns:
+ * 0 on success or one of the following negative error codes on failure:
+ * - EINVAL if the specified rate is not among the supported rates
+ *
+ * See also:
+ * drm_dp_link_add_rate()
+ */
+int drm_dp_link_remove_rate(struct drm_dp_link *link, unsigned long rate)
+{
+	unsigned int i;
+
+	for (i = 0; i < link->num_rates; i++)
+		if (rate == link->rates[i])
+			break;
+
+	if (i == link->num_rates)
+		return -EINVAL;
+
+	link->num_rates--;
+
+	while (i < link->num_rates) {
+		link->rates[i] = link->rates[i + 1];
+		i++;
+	}
+
+	return 0;
+}
+
+/**
+ * drm_dp_link_update_rates() - normalize the supported link rates array
+ * @link: the link for which to normalize the supported link rates
+ *
+ * Users should call this function after they've manually modified the array
+ * of supported link rates. This function removes any stale entries, compacts
+ * the array and updates the supported link rate count. Note that calling the
+ * drm_dp_link_remove_rate() function already does this janitorial work.
+ *
+ * See also:
+ * drm_dp_link_add_rate(), drm_dp_link_remove_rate()
+ */
+void drm_dp_link_update_rates(struct drm_dp_link *link)
+{
+	unsigned int i, count = 0;
+
+	for (i = 0; i < link->num_rates; i++) {
+		if (link->rates[i] != 0)
+			link->rates[count++] = link->rates[i];
+	}
+
+	for (i = count; i < link->num_rates; i++)
+		link->rates[i] = 0;
+
+	link->num_rates = count;
+}
+
+/**
+ * drm_dp_link_probe() - probe a DisplayPort link for capabilities
+ * @aux: DisplayPort AUX channel
+ * @link: pointer to structure in which to return link capabilities
+ *
+ * The structure filled in by this function can usually be passed directly
+ * into drm_dp_link_power_up() and drm_dp_link_configure() to power up and
+ * configure the link based on the link's capabilities.
+ *
+ * Returns 0 on success or a negative error code on failure.
+ */
+int drm_dp_link_probe(struct drm_dp_aux *aux, struct drm_dp_link *link)
+{
+	u8 dpcd[DP_RECEIVER_CAP_SIZE], value;
+	unsigned int rd_interval;
+	int err;
+
+	drm_dp_link_reset(link);
+
+	err = drm_dp_dpcd_read(aux, DP_DPCD_REV, dpcd, sizeof(dpcd));
+	if (err < 0)
+		return err;
+
+	link->revision = dpcd[DP_DPCD_REV];
+	link->max_rate = drm_dp_max_link_rate(dpcd);
+	link->max_lanes = drm_dp_max_lane_count(dpcd);
+
+	link->caps.enhanced_framing = drm_dp_enhanced_frame_cap(dpcd);
+	link->caps.tps3_supported = drm_dp_tps3_supported(dpcd);
+	link->caps.fast_training = drm_dp_fast_training_cap(dpcd);
+	link->caps.channel_coding = drm_dp_channel_coding_supported(dpcd);
+
+	if (drm_dp_alternate_scrambler_reset_cap(dpcd)) {
+		link->caps.alternate_scrambler_reset = true;
+
+		err = drm_dp_dpcd_readb(aux, DP_EDP_DPCD_REV, &value);
+		if (err < 0)
+			return err;
+
+		if (value >= ARRAY_SIZE(drm_dp_edp_revisions))
+			DRM_ERROR("unsupported eDP version: %02x\n", value);
+		else
+			link->edp = drm_dp_edp_revisions[value];
+	}
+
+	/*
+	 * The DPCD stores the AUX read interval in units of 4 ms. There are
+	 * two special cases:
+	 *
+	 *   1) if the TRAINING_AUX_RD_INTERVAL field is 0, the clock recovery
+	 *      and channel equalization should use 100 us or 400 us AUX read
+	 *      intervals, respectively
+	 *
+	 *   2) for DP v1.4 and above, clock recovery should always use 100 us
+	 *      AUX read intervals
+	 */
+	rd_interval = dpcd[DP_TRAINING_AUX_RD_INTERVAL] &
+			   DP_TRAINING_AUX_RD_MASK;
+
+	if (rd_interval > 4) {
+		DRM_DEBUG_KMS("AUX interval %u out of range (max. 4)\n",
+			      rd_interval);
+		rd_interval = 4;
+	}
+
+	rd_interval *= 4 * USEC_PER_MSEC;
+
+	if (rd_interval == 0 || link->revision >= DP_DPCD_REV_14)
+		link->aux_rd_interval.cr = 100;
+
+	if (rd_interval == 0)
+		link->aux_rd_interval.ce = 400;
+
+	link->rate = link->max_rate;
+	link->lanes = link->max_lanes;
+
+	/* Parse SUPPORTED_LINK_RATES from eDP 1.4 */
+	if (link->edp >= 0x14) {
+		u8 supported_rates[DP_MAX_SUPPORTED_RATES * 2];
+		unsigned int i;
+		u16 rate;
+
+		err = drm_dp_dpcd_read(aux, DP_SUPPORTED_LINK_RATES,
+				       supported_rates,
+				       sizeof(supported_rates));
+		if (err < 0)
+			return err;
+
+		for (i = 0; i < DP_MAX_SUPPORTED_RATES; i++) {
+			rate = supported_rates[i * 2 + 1] << 8 |
+			       supported_rates[i * 2 + 0];
+
+			drm_dp_link_add_rate(link, rate * 200);
+		}
+	}
+
+	return 0;
+}
+
+/**
+ * drm_dp_link_power_up() - power up a DisplayPort link
+ * @aux: DisplayPort AUX channel
+ * @link: pointer to a structure containing the link configuration
+ *
+ * Returns 0 on success or a negative error code on failure.
+ */
+int drm_dp_link_power_up(struct drm_dp_aux *aux, struct drm_dp_link *link)
+{
+	u8 value;
+	int err;
+
+	/* DP_SET_POWER register is only available on DPCD v1.1 and later */
+	if (link->revision < 0x11)
+		return 0;
+
+	err = drm_dp_dpcd_readb(aux, DP_SET_POWER, &value);
+	if (err < 0)
+		return err;
+
+	value &= ~DP_SET_POWER_MASK;
+	value |= DP_SET_POWER_D0;
+
+	err = drm_dp_dpcd_writeb(aux, DP_SET_POWER, value);
+	if (err < 0)
+		return err;
+
+	/*
+	 * According to the DP 1.1 specification, a "Sink Device must exit the
+	 * power saving state within 1 ms" (Section 2.5.3.1, Table 5-52, "Sink
+	 * Control Field" (register 0x600).
+	 */
+	usleep_range(1000, 2000);
+
+	return 0;
+}
+
+/**
+ * drm_dp_link_power_down() - power down a DisplayPort link
+ * @aux: DisplayPort AUX channel
+ * @link: pointer to a structure containing the link configuration
+ *
+ * Returns 0 on success or a negative error code on failure.
+ */
+int drm_dp_link_power_down(struct drm_dp_aux *aux, struct drm_dp_link *link)
+{
+	u8 value;
+	int err;
+
+	/* DP_SET_POWER register is only available on DPCD v1.1 and later */
+	if (link->revision < 0x11)
+		return 0;
+
+	err = drm_dp_dpcd_readb(aux, DP_SET_POWER, &value);
+	if (err < 0)
+		return err;
+
+	value &= ~DP_SET_POWER_MASK;
+	value |= DP_SET_POWER_D3;
+
+	err = drm_dp_dpcd_writeb(aux, DP_SET_POWER, value);
+	if (err < 0)
+		return err;
+
+	return 0;
+}
+
+/**
+ * drm_dp_link_configure() - configure a DisplayPort link
+ * @aux: DisplayPort AUX channel
+ * @link: pointer to a structure containing the link configuration
+ *
+ * Returns 0 on success or a negative error code on failure.
+ */
+int drm_dp_link_configure(struct drm_dp_aux *aux, struct drm_dp_link *link)
+{
+	u8 values[2], value;
+	int err;
+
+	if (link->ops && link->ops->configure) {
+		err = link->ops->configure(link);
+		if (err < 0) {
+			DRM_ERROR("failed to configure DP link: %d\n", err);
+			return err;
+		}
+	}
+
+	values[0] = drm_dp_link_rate_to_bw_code(link->rate);
+	values[1] = link->lanes;
+
+	if (link->caps.enhanced_framing)
+		values[1] |= DP_LANE_COUNT_ENHANCED_FRAME_EN;
+
+	err = drm_dp_dpcd_write(aux, DP_LINK_BW_SET, values, sizeof(values));
+	if (err < 0)
+		return err;
+
+	if (link->caps.channel_coding)
+		value = DP_SET_ANSI_8B10B;
+	else
+		value = 0;
+
+	err = drm_dp_dpcd_writeb(aux, DP_MAIN_LINK_CHANNEL_CODING_SET, value);
+	if (err < 0)
+		return err;
+
+	if (link->caps.alternate_scrambler_reset) {
+		err = drm_dp_dpcd_writeb(aux, DP_EDP_CONFIGURATION_SET,
+					 DP_ALTERNATE_SCRAMBLER_RESET_ENABLE);
+		if (err < 0)
+			return err;
+	}
+
+	return 0;
+}
+
+/**
+ * drm_dp_link_choose() - choose the lowest possible configuration for a mode
+ * @link: DRM DP link object
+ * @mode: DRM display mode
+ * @info: DRM display information
+ *
+ * According to the eDP specification, a source should select a configuration
+ * with the lowest number of lanes and the lowest possible link rate that can
+ * match the bitrate requirements of a video mode. However it must ensure not
+ * to exceed the capabilities of the sink.
+ *
+ * Returns: 0 on success or a negative error code on failure.
+ */
+int drm_dp_link_choose(struct drm_dp_link *link,
+		       const struct drm_display_mode *mode,
+		       const struct drm_display_info *info)
+{
+	/* available link symbol clock rates */
+	static const unsigned int rates[3] = { 162000, 270000, 540000 };
+	/* available number of lanes */
+	static const unsigned int lanes[3] = { 1, 2, 4 };
+	unsigned long requirement, capacity;
+	unsigned int rate = link->max_rate;
+	unsigned int i, j;
+
+	/* bandwidth requirement */
+	requirement = mode->clock * info->bpc * 3;
+
+	for (i = 0; i < ARRAY_SIZE(lanes) && lanes[i] <= link->max_lanes; i++) {
+		for (j = 0; j < ARRAY_SIZE(rates) && rates[j] <= rate; j++) {
+			/*
+			 * Capacity for this combination of lanes and rate,
+			 * factoring in the ANSI 8B/10B encoding.
+			 *
+			 * Link rates in the DRM DP helpers are really link
+			 * symbol frequencies, so a tenth of the actual rate
+			 * of the link.
+			 */
+			capacity = lanes[i] * (rates[j] * 10) * 8 / 10;
+
+			if (capacity >= requirement) {
+				DRM_DEBUG_KMS("using %u lanes at %u kHz (%lu/%lu kbps)\n",
+					      lanes[i], rates[j], requirement,
+					      capacity);
+				link->lanes = lanes[i];
+				link->rate = rates[j];
+				return 0;
+			}
+		}
+	}
+
+	return -ERANGE;
+}
+
+/**
+ * DOC: Link training
+ *
+ * These functions contain common logic and helpers to implement DisplayPort
+ * link training.
+ */
+
+/**
+ * drm_dp_link_train_init() - initialize DisplayPort link training state
+ * @train: DisplayPort link training state
+ */
+void drm_dp_link_train_init(struct drm_dp_link_train *train)
+{
+	struct drm_dp_link_train_set *request = &train->request;
+	struct drm_dp_link_train_set *adjust = &train->adjust;
+	unsigned int i;
+
+	for (i = 0; i < 4; i++) {
+		request->voltage_swing[i] = 0;
+		adjust->voltage_swing[i] = 0;
+
+		request->pre_emphasis[i] = 0;
+		adjust->pre_emphasis[i] = 0;
+
+		request->post_cursor[i] = 0;
+		adjust->post_cursor[i] = 0;
+	}
+
+	train->pattern = DP_TRAINING_PATTERN_DISABLE;
+	train->clock_recovered = false;
+	train->channel_equalized = false;
+}
+
+static bool drm_dp_link_train_valid(const struct drm_dp_link_train *train)
+{
+	return train->clock_recovered && train->channel_equalized;
+}
+
+static int drm_dp_link_apply_training(struct drm_dp_link *link)
+{
+	struct drm_dp_link_train_set *request = &link->train.request;
+	unsigned int lanes = link->lanes, *vs, *pe, *pc, i;
+	struct drm_dp_aux *aux = link->aux;
+	u8 values[4], pattern = 0;
+	int err;
+
+	err = link->ops->apply_training(link);
+	if (err < 0) {
+		DRM_ERROR("failed to apply link training: %d\n", err);
+		return err;
+	}
+
+	vs = request->voltage_swing;
+	pe = request->pre_emphasis;
+	pc = request->post_cursor;
+
+	/* write currently selected voltage-swing and pre-emphasis levels */
+	for (i = 0; i < lanes; i++)
+		values[i] = DP_TRAIN_VOLTAGE_SWING_LEVEL(vs[i]) |
+			    DP_TRAIN_PRE_EMPHASIS_LEVEL(pe[i]);
+
+	err = drm_dp_dpcd_write(aux, DP_TRAINING_LANE0_SET, values, lanes);
+	if (err < 0) {
+		DRM_ERROR("failed to set training parameters: %d\n", err);
+		return err;
+	}
+
+	/* write currently selected post-cursor level (if supported) */
+	if (link->revision >= 0x12 && link->rate == 540000) {
+		values[0] = values[1] = 0;
+
+		for (i = 0; i < lanes; i++)
+			values[i / 2] |= DP_LANE_POST_CURSOR(i, pc[i]);
+
+		err = drm_dp_dpcd_write(aux, DP_TRAINING_LANE0_1_SET2, values,
+					DIV_ROUND_UP(lanes, 2));
+		if (err < 0) {
+			DRM_ERROR("failed to set post-cursor: %d\n", err);
+			return err;
+		}
+	}
+
+	/* write link pattern */
+	if (link->train.pattern != DP_TRAINING_PATTERN_DISABLE)
+		pattern |= DP_LINK_SCRAMBLING_DISABLE;
+
+	pattern |= link->train.pattern;
+
+	err = drm_dp_dpcd_writeb(aux, DP_TRAINING_PATTERN_SET, pattern);
+	if (err < 0) {
+		DRM_ERROR("failed to set training pattern: %d\n", err);
+		return err;
+	}
+
+	return 0;
+}
+
+static void drm_dp_link_train_wait(struct drm_dp_link *link)
+{
+	unsigned long min = 0;
+
+	switch (link->train.pattern) {
+	case DP_TRAINING_PATTERN_1:
+		min = link->aux_rd_interval.cr;
+		break;
+
+	case DP_TRAINING_PATTERN_2:
+	case DP_TRAINING_PATTERN_3:
+		min = link->aux_rd_interval.ce;
+		break;
+
+	default:
+		break;
+	}
+
+	if (min > 0)
+		usleep_range(min, 2 * min);
+}
+
+static void drm_dp_link_get_adjustments(struct drm_dp_link *link,
+					u8 status[DP_LINK_STATUS_SIZE])
+{
+	struct drm_dp_link_train_set *adjust = &link->train.adjust;
+	unsigned int i;
+
+	for (i = 0; i < link->lanes; i++) {
+		adjust->voltage_swing[i] =
+			drm_dp_get_adjust_request_voltage(status, i) >>
+				DP_TRAIN_VOLTAGE_SWING_SHIFT;
+
+		adjust->pre_emphasis[i] =
+			drm_dp_get_adjust_request_pre_emphasis(status, i) >>
+				DP_TRAIN_PRE_EMPHASIS_SHIFT;
+
+		adjust->post_cursor[i] =
+			drm_dp_get_adjust_request_post_cursor(status, i);
+	}
+}
+
+static void drm_dp_link_train_adjust(struct drm_dp_link_train *train)
+{
+	struct drm_dp_link_train_set *request = &train->request;
+	struct drm_dp_link_train_set *adjust = &train->adjust;
+	unsigned int i;
+
+	for (i = 0; i < 4; i++)
+		if (request->voltage_swing[i] != adjust->voltage_swing[i])
+			request->voltage_swing[i] = adjust->voltage_swing[i];
+
+	for (i = 0; i < 4; i++)
+		if (request->pre_emphasis[i] != adjust->pre_emphasis[i])
+			request->pre_emphasis[i] = adjust->pre_emphasis[i];
+
+	for (i = 0; i < 4; i++)
+		if (request->post_cursor[i] != adjust->post_cursor[i])
+			request->post_cursor[i] = adjust->post_cursor[i];
+}
+
+static int drm_dp_link_recover_clock(struct drm_dp_link *link)
+{
+	u8 status[DP_LINK_STATUS_SIZE];
+	int err;
+
+	err = drm_dp_link_apply_training(link);
+	if (err < 0)
+		return err;
+
+	drm_dp_link_train_wait(link);
+
+	err = drm_dp_dpcd_read_link_status(link->aux, status);
+	if (err < 0) {
+		DRM_ERROR("failed to read link status: %d\n", err);
+		return err;
+	}
+
+	if (!drm_dp_clock_recovery_ok(status, link->lanes))
+		drm_dp_link_get_adjustments(link, status);
+	else
+		link->train.clock_recovered = true;
+
+	return 0;
+}
+
+static int drm_dp_link_clock_recovery(struct drm_dp_link *link)
+{
+	unsigned int repeat;
+	int err;
+
+	/* start clock recovery using training pattern 1 */
+	link->train.pattern = DP_TRAINING_PATTERN_1;
+
+	for (repeat = 1; repeat < 5; repeat++) {
+		err = drm_dp_link_recover_clock(link);
+		if (err < 0) {
+			DRM_ERROR("failed to recover clock: %d\n", err);
+			return err;
+		}
+
+		if (link->train.clock_recovered)
+			break;
+
+		drm_dp_link_train_adjust(&link->train);
+	}
+
+	return 0;
+}
+
+static int drm_dp_link_equalize_channel(struct drm_dp_link *link)
+{
+	struct drm_dp_aux *aux = link->aux;
+	u8 status[DP_LINK_STATUS_SIZE];
+	int err;
+
+	err = drm_dp_link_apply_training(link);
+	if (err < 0)
+		return err;
+
+	drm_dp_link_train_wait(link);
+
+	err = drm_dp_dpcd_read_link_status(aux, status);
+	if (err < 0) {
+		DRM_ERROR("failed to read link status: %d\n", err);
+		return err;
+	}
+
+	if (!drm_dp_clock_recovery_ok(status, link->lanes)) {
+		DRM_ERROR("clock recovery lost while equalizing channel\n");
+		link->train.clock_recovered = false;
+		return 0;
+	}
+
+	if (!drm_dp_channel_eq_ok(status, link->lanes))
+		drm_dp_link_get_adjustments(link, status);
+	else
+		link->train.channel_equalized = true;
+
+	return 0;
+}
+
+static int drm_dp_link_channel_equalization(struct drm_dp_link *link)
+{
+	unsigned int repeat;
+	int err;
+
+	/* start channel equalization using pattern 2 or 3 */
+	if (link->caps.tps3_supported)
+		link->train.pattern = DP_TRAINING_PATTERN_3;
+	else
+		link->train.pattern = DP_TRAINING_PATTERN_2;
+
+	for (repeat = 1; repeat < 5; repeat++) {
+		err = drm_dp_link_equalize_channel(link);
+		if (err < 0) {
+			DRM_ERROR("failed to equalize channel: %d\n", err);
+			return err;
+		}
+
+		if (link->train.channel_equalized)
+			break;
+
+		drm_dp_link_train_adjust(&link->train);
+	}
+
+	return 0;
+}
+
+static int drm_dp_link_downgrade(struct drm_dp_link *link)
+{
+	switch (link->rate) {
+	case 162000:
+		return -EINVAL;
+
+	case 270000:
+		link->rate = 162000;
+		break;
+
+	case 540000:
+		link->rate = 270000;
+		return 0;
+	}
+
+	return 0;
+}
+
+static void drm_dp_link_train_disable(struct drm_dp_link *link)
+{
+	int err;
+
+	link->train.pattern = DP_TRAINING_PATTERN_DISABLE;
+
+	err = drm_dp_link_apply_training(link);
+	if (err < 0)
+		DRM_ERROR("failed to disable link training: %d\n", err);
+}
+
+static int drm_dp_link_train_full(struct drm_dp_link *link)
+{
+	int err;
+
+retry:
+	DRM_DEBUG_KMS("full-training link: %u lane%s at %u MHz\n",
+		      link->lanes, (link->lanes > 1) ? "s" : "",
+		      link->rate / 100);
+
+	err = drm_dp_link_configure(link->aux, link);
+	if (err < 0) {
+		DRM_ERROR("failed to configure DP link: %d\n", err);
+		return err;
+	}
+
+	err = drm_dp_link_clock_recovery(link);
+	if (err < 0) {
+		DRM_ERROR("clock recovery failed: %d\n", err);
+		goto out;
+	}
+
+	if (!link->train.clock_recovered) {
+		DRM_ERROR("clock recovery failed, downgrading link\n");
+
+		err = drm_dp_link_downgrade(link);
+		if (err < 0)
+			goto out;
+
+		goto retry;
+	}
+
+	DRM_DEBUG_KMS("clock recovery succeeded\n");
+
+	err = drm_dp_link_channel_equalization(link);
+	if (err < 0) {
+		DRM_ERROR("channel equalization failed: %d\n", err);
+		goto out;
+	}
+
+	if (!link->train.channel_equalized) {
+		DRM_ERROR("channel equalization failed, downgrading link\n");
+
+		err = drm_dp_link_downgrade(link);
+		if (err < 0)
+			goto out;
+
+		goto retry;
+	}
+
+	DRM_DEBUG_KMS("channel equalization succeeded\n");
+
+out:
+	drm_dp_link_train_disable(link);
+	return err;
+}
+
+static int drm_dp_link_train_fast(struct drm_dp_link *link)
+{
+	u8 status[DP_LINK_STATUS_SIZE];
+	int err;
+
+	DRM_DEBUG_KMS("fast-training link: %u lane%s at %u MHz\n",
+		      link->lanes, (link->lanes > 1) ? "s" : "",
+		      link->rate / 100);
+
+	err = drm_dp_link_configure(link->aux, link);
+	if (err < 0) {
+		DRM_ERROR("failed to configure DP link: %d\n", err);
+		return err;
+	}
+
+	/* transmit training pattern 1 for 500 microseconds */
+	link->train.pattern = DP_TRAINING_PATTERN_1;
+
+	err = drm_dp_link_apply_training(link);
+	if (err < 0)
+		goto out;
+
+	usleep_range(500, 1000);
+
+	/* transmit training pattern 2 or 3 for 500 microseconds */
+	if (link->caps.tps3_supported)
+		link->train.pattern = DP_TRAINING_PATTERN_3;
+	else
+		link->train.pattern = DP_TRAINING_PATTERN_2;
+
+	err = drm_dp_link_apply_training(link);
+	if (err < 0)
+		goto out;
+
+	usleep_range(500, 1000);
+
+	err = drm_dp_dpcd_read_link_status(link->aux, status);
+	if (err < 0) {
+		DRM_ERROR("failed to read link status: %d\n", err);
+		goto out;
+	}
+
+	if (!drm_dp_clock_recovery_ok(status, link->lanes)) {
+		DRM_ERROR("clock recovery failed\n");
+		err = -EIO;
+	}
+
+	if (!drm_dp_channel_eq_ok(status, link->lanes)) {
+		DRM_ERROR("channel equalization failed\n");
+		err = -EIO;
+	}
+
+out:
+	drm_dp_link_train_disable(link);
+	return err;
+}
+
+/**
+ * drm_dp_link_train() - perform DisplayPort link training
+ * @link: a DP link object
+ *
+ * Uses the context stored in the DP link object to perform link training. It
+ * is expected that drivers will call drm_dp_link_probe() to obtain the link
+ * capabilities before performing link training.
+ *
+ * If the sink supports fast link training (no AUX CH handshake) and valid
+ * training settings are available, this function will try to perform fast
+ * link training and fall back to full link training on failure.
+ *
+ * Returns: 0 on success or a negative error code on failure.
+ */
+int drm_dp_link_train(struct drm_dp_link *link)
+{
+	int err;
+
+	drm_dp_link_train_init(&link->train);
+
+	if (link->caps.fast_training) {
+		if (drm_dp_link_train_valid(&link->train)) {
+			err = drm_dp_link_train_fast(link);
+			if (err < 0)
+				DRM_ERROR("fast link training failed: %d\n",
+					  err);
+			else
+				return 0;
+		} else {
+			DRM_DEBUG_KMS("training parameters not available\n");
+		}
+	} else {
+		DRM_DEBUG_KMS("fast link training not supported\n");
+	}
+
+	err = drm_dp_link_train_full(link);
+	if (err < 0)
+		DRM_ERROR("full link training failed: %d\n", err);
+
+	return err;
+}
diff --git a/drivers/gpu/drm/tegra/dp.h b/drivers/gpu/drm/tegra/dp.h
new file mode 100644
index 000000000000..cb12ed0c54e7
--- /dev/null
+++ b/drivers/gpu/drm/tegra/dp.h
@@ -0,0 +1,177 @@
+/* SPDX-License-Identifier: MIT */
+/*
+ * Copyright (C) 2013-2019 NVIDIA Corporation.
+ * Copyright (C) 2015 Rob Clark
+ */
+
+#ifndef DRM_TEGRA_DP_H
+#define DRM_TEGRA_DP_H 1
+
+#include <linux/types.h>
+
+struct drm_display_info;
+struct drm_display_mode;
+struct drm_dp_aux;
+struct drm_dp_link;
+
+/**
+ * struct drm_dp_link_caps - DP link capabilities
+ */
+struct drm_dp_link_caps {
+	/**
+	 * @enhanced_framing:
+	 *
+	 * enhanced framing capability (mandatory as of DP 1.2)
+	 */
+	bool enhanced_framing;
+
+	/**
+	 * tps3_supported:
+	 *
+	 * training pattern sequence 3 supported for equalization
+	 */
+	bool tps3_supported;
+
+	/**
+	 * @fast_training:
+	 *
+	 * AUX CH handshake not required for link training
+	 */
+	bool fast_training;
+
+	/**
+	 * @channel_coding:
+	 *
+	 * ANSI 8B/10B channel coding capability
+	 */
+	bool channel_coding;
+
+	/**
+	 * @alternate_scrambler_reset:
+	 *
+	 * eDP alternate scrambler reset capability
+	 */
+	bool alternate_scrambler_reset;
+};
+
+void drm_dp_link_caps_copy(struct drm_dp_link_caps *dest,
+			   const struct drm_dp_link_caps *src);
+
+/**
+ * struct drm_dp_link_ops - DP link operations
+ */
+struct drm_dp_link_ops {
+	/**
+	 * @apply_training:
+	 */
+	int (*apply_training)(struct drm_dp_link *link);
+
+	/**
+	 * @configure:
+	 */
+	int (*configure)(struct drm_dp_link *link);
+};
+
+#define DP_TRAIN_VOLTAGE_SWING_LEVEL(x) ((x) << 0)
+#define DP_TRAIN_PRE_EMPHASIS_LEVEL(x) ((x) << 3)
+#define DP_LANE_POST_CURSOR(i, x) (((x) & 0x3) << (((i) & 1) << 2))
+
+/**
+ * struct drm_dp_link_train_set - link training settings
+ * @voltage_swing: per-lane voltage swing
+ * @pre_emphasis: per-lane pre-emphasis
+ * @post_cursor: per-lane post-cursor
+ */
+struct drm_dp_link_train_set {
+	unsigned int voltage_swing[4];
+	unsigned int pre_emphasis[4];
+	unsigned int post_cursor[4];
+};
+
+/**
+ * struct drm_dp_link_train - link training state information
+ * @request: currently requested settings
+ * @adjust: adjustments requested by sink
+ * @pattern: currently requested training pattern
+ * @clock_recovered: flag to track if clock recovery has completed
+ * @channel_equalized: flag to track if channel equalization has completed
+ */
+struct drm_dp_link_train {
+	struct drm_dp_link_train_set request;
+	struct drm_dp_link_train_set adjust;
+
+	unsigned int pattern;
+
+	bool clock_recovered;
+	bool channel_equalized;
+};
+
+/**
+ * struct drm_dp_link - DP link capabilities and configuration
+ * @revision: DP specification revision supported on the link
+ * @max_rate: maximum clock rate supported on the link
+ * @max_lanes: maximum number of lanes supported on the link
+ * @caps: capabilities supported on the link (see &drm_dp_link_caps)
+ * @aux_rd_interval: AUX read interval to use for training (in microseconds)
+ * @edp: eDP revision (0x11: eDP 1.1, 0x12: eDP 1.2, ...)
+ * @rate: currently configured link rate
+ * @lanes: currently configured number of lanes
+ * @rates: additional supported link rates in kHz (eDP 1.4)
+ * @num_rates: number of additional supported link rates (eDP 1.4)
+ */
+struct drm_dp_link {
+	unsigned char revision;
+	unsigned int max_rate;
+	unsigned int max_lanes;
+
+	struct drm_dp_link_caps caps;
+
+	/**
+	 * @cr: clock recovery read interval
+	 * @ce: channel equalization read interval
+	 */
+	struct {
+		unsigned int cr;
+		unsigned int ce;
+	} aux_rd_interval;
+
+	unsigned char edp;
+
+	unsigned int rate;
+	unsigned int lanes;
+
+	unsigned long rates[DP_MAX_SUPPORTED_RATES];
+	unsigned int num_rates;
+
+	/**
+	 * @ops: DP link operations
+	 */
+	const struct drm_dp_link_ops *ops;
+
+	/**
+	 * @aux: DP AUX channel
+	 */
+	struct drm_dp_aux *aux;
+
+	/**
+	 * @train: DP link training state
+	 */
+	struct drm_dp_link_train train;
+};
+
+int drm_dp_link_add_rate(struct drm_dp_link *link, unsigned long rate);
+int drm_dp_link_remove_rate(struct drm_dp_link *link, unsigned long rate);
+void drm_dp_link_update_rates(struct drm_dp_link *link);
+
+int drm_dp_link_probe(struct drm_dp_aux *aux, struct drm_dp_link *link);
+int drm_dp_link_power_up(struct drm_dp_aux *aux, struct drm_dp_link *link);
+int drm_dp_link_power_down(struct drm_dp_aux *aux, struct drm_dp_link *link);
+int drm_dp_link_configure(struct drm_dp_aux *aux, struct drm_dp_link *link);
+int drm_dp_link_choose(struct drm_dp_link *link,
+		       const struct drm_display_mode *mode,
+		       const struct drm_display_info *info);
+
+void drm_dp_link_train_init(struct drm_dp_link_train *train);
+int drm_dp_link_train(struct drm_dp_link *link);
+
+#endif
diff --git a/drivers/gpu/drm/tegra/dpaux.c b/drivers/gpu/drm/tegra/dpaux.c
index a0f6f9b0d258..622cdf1ad246 100644
--- a/drivers/gpu/drm/tegra/dpaux.c
+++ b/drivers/gpu/drm/tegra/dpaux.c
@@ -9,6 +9,7 @@
 #include <linux/interrupt.h>
 #include <linux/io.h>
 #include <linux/module.h>
+#include <linux/of_device.h>
 #include <linux/of_gpio.h>
 #include <linux/pinctrl/pinconf-generic.h>
 #include <linux/pinctrl/pinctrl.h>
@@ -22,6 +23,7 @@
 #include <drm/drm_dp_helper.h>
 #include <drm/drm_panel.h>
 
+#include "dp.h"
 #include "dpaux.h"
 #include "drm.h"
 #include "trace.h"
@@ -29,10 +31,18 @@
 static DEFINE_MUTEX(dpaux_lock);
 static LIST_HEAD(dpaux_list);
 
+struct tegra_dpaux_soc {
+	unsigned int cmh;
+	unsigned int drvz;
+	unsigned int drvi;
+};
+
 struct tegra_dpaux {
 	struct drm_dp_aux aux;
 	struct device *dev;
 
+	const struct tegra_dpaux_soc *soc;
+
 	void __iomem *regs;
 	int irq;
 
@@ -120,6 +130,7 @@ static ssize_t tegra_dpaux_transfer(struct drm_dp_aux *aux,
 	struct tegra_dpaux *dpaux = to_dpaux(aux);
 	unsigned long status;
 	ssize_t ret = 0;
+	u8 reply = 0;
 	u32 value;
 
 	/* Tegra has 4x4 byte DP AUX transmit and receive FIFOs. */
@@ -214,23 +225,23 @@ static ssize_t tegra_dpaux_transfer(struct drm_dp_aux *aux,
 
 	switch ((value & DPAUX_DP_AUXSTAT_REPLY_TYPE_MASK) >> 16) {
 	case 0x00:
-		msg->reply = DP_AUX_NATIVE_REPLY_ACK;
+		reply = DP_AUX_NATIVE_REPLY_ACK;
 		break;
 
 	case 0x01:
-		msg->reply = DP_AUX_NATIVE_REPLY_NACK;
+		reply = DP_AUX_NATIVE_REPLY_NACK;
 		break;
 
 	case 0x02:
-		msg->reply = DP_AUX_NATIVE_REPLY_DEFER;
+		reply = DP_AUX_NATIVE_REPLY_DEFER;
 		break;
 
 	case 0x04:
-		msg->reply = DP_AUX_I2C_REPLY_NACK;
+		reply = DP_AUX_I2C_REPLY_NACK;
 		break;
 
 	case 0x08:
-		msg->reply = DP_AUX_I2C_REPLY_DEFER;
+		reply = DP_AUX_I2C_REPLY_DEFER;
 		break;
 	}
 
@@ -238,14 +249,24 @@ static ssize_t tegra_dpaux_transfer(struct drm_dp_aux *aux,
 		if (msg->request & DP_AUX_I2C_READ) {
 			size_t count = value & DPAUX_DP_AUXSTAT_REPLY_MASK;
 
-			if (WARN_ON(count != msg->size))
-				count = min_t(size_t, count, msg->size);
+			/*
+			 * There might be a smarter way to do this, but since
+			 * the DP helpers will already retry transactions for
+			 * an -EBUSY return value, simply reuse that instead.
+			 */
+			if (count != msg->size) {
+				ret = -EBUSY;
+				goto out;
+			}
 
 			tegra_dpaux_read_fifo(dpaux, msg->buffer, count);
 			ret = count;
 		}
 	}
 
+	msg->reply = reply;
+
+out:
 	return ret;
 }
 
@@ -310,9 +331,9 @@ static int tegra_dpaux_pad_config(struct tegra_dpaux *dpaux, unsigned function)
 
 	switch (function) {
 	case DPAUX_PADCTL_FUNC_AUX:
-		value = DPAUX_HYBRID_PADCTL_AUX_CMH(2) |
-			DPAUX_HYBRID_PADCTL_AUX_DRVZ(4) |
-			DPAUX_HYBRID_PADCTL_AUX_DRVI(0x18) |
+		value = DPAUX_HYBRID_PADCTL_AUX_CMH(dpaux->soc->cmh) |
+			DPAUX_HYBRID_PADCTL_AUX_DRVZ(dpaux->soc->drvz) |
+			DPAUX_HYBRID_PADCTL_AUX_DRVI(dpaux->soc->drvi) |
 			DPAUX_HYBRID_PADCTL_AUX_INPUT_RCV |
 			DPAUX_HYBRID_PADCTL_MODE_AUX;
 		break;
@@ -320,9 +341,9 @@ static int tegra_dpaux_pad_config(struct tegra_dpaux *dpaux, unsigned function)
 	case DPAUX_PADCTL_FUNC_I2C:
 		value = DPAUX_HYBRID_PADCTL_I2C_SDA_INPUT_RCV |
 			DPAUX_HYBRID_PADCTL_I2C_SCL_INPUT_RCV |
-			DPAUX_HYBRID_PADCTL_AUX_CMH(2) |
-			DPAUX_HYBRID_PADCTL_AUX_DRVZ(4) |
-			DPAUX_HYBRID_PADCTL_AUX_DRVI(0x18) |
+			DPAUX_HYBRID_PADCTL_AUX_CMH(dpaux->soc->cmh) |
+			DPAUX_HYBRID_PADCTL_AUX_DRVZ(dpaux->soc->drvz) |
+			DPAUX_HYBRID_PADCTL_AUX_DRVI(dpaux->soc->drvi) |
 			DPAUX_HYBRID_PADCTL_MODE_I2C;
 		break;
 
@@ -436,6 +457,7 @@ static int tegra_dpaux_probe(struct platform_device *pdev)
 	if (!dpaux)
 		return -ENOMEM;
 
+	dpaux->soc = of_device_get_match_data(&pdev->dev);
 	INIT_WORK(&dpaux->work, tegra_dpaux_hotplug);
 	init_completion(&dpaux->complete);
 	INIT_LIST_HEAD(&dpaux->list);
@@ -493,6 +515,8 @@ static int tegra_dpaux_probe(struct platform_device *pdev)
 
 			return PTR_ERR(dpaux->vdd);
 		}
+
+		dpaux->vdd = NULL;
 	}
 
 	platform_set_drvdata(pdev, dpaux);
@@ -641,11 +665,29 @@ static const struct dev_pm_ops tegra_dpaux_pm_ops = {
 	SET_RUNTIME_PM_OPS(tegra_dpaux_suspend, tegra_dpaux_resume, NULL)
 };
 
+static const struct tegra_dpaux_soc tegra124_dpaux_soc = {
+	.cmh = 0x02,
+	.drvz = 0x04,
+	.drvi = 0x18,
+};
+
+static const struct tegra_dpaux_soc tegra210_dpaux_soc = {
+	.cmh = 0x02,
+	.drvz = 0x04,
+	.drvi = 0x30,
+};
+
+static const struct tegra_dpaux_soc tegra194_dpaux_soc = {
+	.cmh = 0x02,
+	.drvz = 0x04,
+	.drvi = 0x2c,
+};
+
 static const struct of_device_id tegra_dpaux_of_match[] = {
-	{ .compatible = "nvidia,tegra194-dpaux", },
-	{ .compatible = "nvidia,tegra186-dpaux", },
-	{ .compatible = "nvidia,tegra210-dpaux", },
-	{ .compatible = "nvidia,tegra124-dpaux", },
+	{ .compatible = "nvidia,tegra194-dpaux", .data = &tegra194_dpaux_soc },
+	{ .compatible = "nvidia,tegra186-dpaux", .data = &tegra210_dpaux_soc },
+	{ .compatible = "nvidia,tegra210-dpaux", .data = &tegra210_dpaux_soc },
+	{ .compatible = "nvidia,tegra124-dpaux", .data = &tegra124_dpaux_soc },
 	{ },
 };
 MODULE_DEVICE_TABLE(of, tegra_dpaux_of_match);
@@ -686,25 +728,32 @@ int drm_dp_aux_attach(struct drm_dp_aux *aux, struct tegra_output *output)
 	output->connector.polled = DRM_CONNECTOR_POLL_HPD;
 	dpaux->output = output;
 
-	err = regulator_enable(dpaux->vdd);
-	if (err < 0)
-		return err;
+	if (output->panel) {
+		enum drm_connector_status status;
 
-	timeout = jiffies + msecs_to_jiffies(250);
+		if (dpaux->vdd) {
+			err = regulator_enable(dpaux->vdd);
+			if (err < 0)
+				return err;
+		}
 
-	while (time_before(jiffies, timeout)) {
-		enum drm_connector_status status;
+		timeout = jiffies + msecs_to_jiffies(250);
+
+		while (time_before(jiffies, timeout)) {
+			status = drm_dp_aux_detect(aux);
+
+			if (status == connector_status_connected)
+				break;
 
-		status = drm_dp_aux_detect(aux);
-		if (status == connector_status_connected) {
-			enable_irq(dpaux->irq);
-			return 0;
+			usleep_range(1000, 2000);
 		}
 
-		usleep_range(1000, 2000);
+		if (status != connector_status_connected)
+			return -ETIMEDOUT;
 	}
 
-	return -ETIMEDOUT;
+	enable_irq(dpaux->irq);
+	return 0;
 }
 
 int drm_dp_aux_detach(struct drm_dp_aux *aux)
@@ -715,25 +764,33 @@ int drm_dp_aux_detach(struct drm_dp_aux *aux)
 
 	disable_irq(dpaux->irq);
 
-	err = regulator_disable(dpaux->vdd);
-	if (err < 0)
-		return err;
+	if (dpaux->output->panel) {
+		enum drm_connector_status status;
 
-	timeout = jiffies + msecs_to_jiffies(250);
+		if (dpaux->vdd) {
+			err = regulator_disable(dpaux->vdd);
+			if (err < 0)
+				return err;
+		}
 
-	while (time_before(jiffies, timeout)) {
-		enum drm_connector_status status;
+		timeout = jiffies + msecs_to_jiffies(250);
+
+		while (time_before(jiffies, timeout)) {
+			status = drm_dp_aux_detect(aux);
+
+			if (status == connector_status_disconnected)
+				break;
 
-		status = drm_dp_aux_detect(aux);
-		if (status == connector_status_disconnected) {
-			dpaux->output = NULL;
-			return 0;
+			usleep_range(1000, 2000);
 		}
 
-		usleep_range(1000, 2000);
+		if (status != connector_status_disconnected)
+			return -ETIMEDOUT;
+
+		dpaux->output = NULL;
 	}
 
-	return -ETIMEDOUT;
+	return 0;
 }
 
 enum drm_connector_status drm_dp_aux_detect(struct drm_dp_aux *aux)
@@ -764,72 +821,3 @@ int drm_dp_aux_disable(struct drm_dp_aux *aux)
 
 	return 0;
 }
-
-int drm_dp_aux_prepare(struct drm_dp_aux *aux, u8 encoding)
-{
-	int err;
-
-	err = drm_dp_dpcd_writeb(aux, DP_MAIN_LINK_CHANNEL_CODING_SET,
-				 encoding);
-	if (err < 0)
-		return err;
-
-	return 0;
-}
-
-int drm_dp_aux_train(struct drm_dp_aux *aux, struct drm_dp_link *link,
-		     u8 pattern)
-{
-	u8 tp = pattern & DP_TRAINING_PATTERN_MASK;
-	u8 status[DP_LINK_STATUS_SIZE], values[4];
-	unsigned int i;
-	int err;
-
-	err = drm_dp_dpcd_writeb(aux, DP_TRAINING_PATTERN_SET, pattern);
-	if (err < 0)
-		return err;
-
-	if (tp == DP_TRAINING_PATTERN_DISABLE)
-		return 0;
-
-	for (i = 0; i < link->num_lanes; i++)
-		values[i] = DP_TRAIN_MAX_PRE_EMPHASIS_REACHED |
-			    DP_TRAIN_PRE_EMPH_LEVEL_0 |
-			    DP_TRAIN_MAX_SWING_REACHED |
-			    DP_TRAIN_VOLTAGE_SWING_LEVEL_0;
-
-	err = drm_dp_dpcd_write(aux, DP_TRAINING_LANE0_SET, values,
-				link->num_lanes);
-	if (err < 0)
-		return err;
-
-	usleep_range(500, 1000);
-
-	err = drm_dp_dpcd_read_link_status(aux, status);
-	if (err < 0)
-		return err;
-
-	switch (tp) {
-	case DP_TRAINING_PATTERN_1:
-		if (!drm_dp_clock_recovery_ok(status, link->num_lanes))
-			return -EAGAIN;
-
-		break;
-
-	case DP_TRAINING_PATTERN_2:
-		if (!drm_dp_channel_eq_ok(status, link->num_lanes))
-			return -EAGAIN;
-
-		break;
-
-	default:
-		dev_err(aux->dev, "unsupported training pattern %u\n", tp);
-		return -EINVAL;
-	}
-
-	err = drm_dp_dpcd_writeb(aux, DP_EDP_CONFIGURATION_SET, 0);
-	if (err < 0)
-		return err;
-
-	return 0;
-}
diff --git a/drivers/gpu/drm/tegra/drm.c b/drivers/gpu/drm/tegra/drm.c
index 6fb7d74ff553..56e5e7a5c108 100644
--- a/drivers/gpu/drm/tegra/drm.c
+++ b/drivers/gpu/drm/tegra/drm.c
@@ -20,10 +20,6 @@
 #include <drm/drm_prime.h>
 #include <drm/drm_vblank.h>
 
-#if IS_ENABLED(CONFIG_ARM_DMA_USE_IOMMU)
-#include <asm/dma-iommu.h>
-#endif
-
 #include "drm.h"
 #include "gem.h"
 
@@ -86,168 +82,6 @@ tegra_drm_mode_config_helpers = {
 	.atomic_commit_tail = tegra_atomic_commit_tail,
 };
 
-static int tegra_drm_load(struct drm_device *drm, unsigned long flags)
-{
-	struct host1x_device *device = to_host1x_device(drm->dev);
-	struct tegra_drm *tegra;
-	int err;
-
-	tegra = kzalloc(sizeof(*tegra), GFP_KERNEL);
-	if (!tegra)
-		return -ENOMEM;
-
-	if (iommu_present(&platform_bus_type)) {
-		tegra->domain = iommu_domain_alloc(&platform_bus_type);
-		if (!tegra->domain) {
-			err = -ENOMEM;
-			goto free;
-		}
-
-		err = iova_cache_get();
-		if (err < 0)
-			goto domain;
-	}
-
-	mutex_init(&tegra->clients_lock);
-	INIT_LIST_HEAD(&tegra->clients);
-
-	drm->dev_private = tegra;
-	tegra->drm = drm;
-
-	drm_mode_config_init(drm);
-
-	drm->mode_config.min_width = 0;
-	drm->mode_config.min_height = 0;
-
-	drm->mode_config.max_width = 4096;
-	drm->mode_config.max_height = 4096;
-
-	drm->mode_config.allow_fb_modifiers = true;
-
-	drm->mode_config.normalize_zpos = true;
-
-	drm->mode_config.funcs = &tegra_drm_mode_config_funcs;
-	drm->mode_config.helper_private = &tegra_drm_mode_config_helpers;
-
-	err = tegra_drm_fb_prepare(drm);
-	if (err < 0)
-		goto config;
-
-	drm_kms_helper_poll_init(drm);
-
-	err = host1x_device_init(device);
-	if (err < 0)
-		goto fbdev;
-
-	if (tegra->domain) {
-		u64 carveout_start, carveout_end, gem_start, gem_end;
-		u64 dma_mask = dma_get_mask(&device->dev);
-		dma_addr_t start, end;
-		unsigned long order;
-
-		start = tegra->domain->geometry.aperture_start & dma_mask;
-		end = tegra->domain->geometry.aperture_end & dma_mask;
-
-		gem_start = start;
-		gem_end = end - CARVEOUT_SZ;
-		carveout_start = gem_end + 1;
-		carveout_end = end;
-
-		order = __ffs(tegra->domain->pgsize_bitmap);
-		init_iova_domain(&tegra->carveout.domain, 1UL << order,
-				 carveout_start >> order);
-
-		tegra->carveout.shift = iova_shift(&tegra->carveout.domain);
-		tegra->carveout.limit = carveout_end >> tegra->carveout.shift;
-
-		drm_mm_init(&tegra->mm, gem_start, gem_end - gem_start + 1);
-		mutex_init(&tegra->mm_lock);
-
-		DRM_DEBUG("IOMMU apertures:\n");
-		DRM_DEBUG("  GEM: %#llx-%#llx\n", gem_start, gem_end);
-		DRM_DEBUG("  Carveout: %#llx-%#llx\n", carveout_start,
-			  carveout_end);
-	}
-
-	if (tegra->hub) {
-		err = tegra_display_hub_prepare(tegra->hub);
-		if (err < 0)
-			goto device;
-	}
-
-	/*
-	 * We don't use the drm_irq_install() helpers provided by the DRM
-	 * core, so we need to set this manually in order to allow the
-	 * DRM_IOCTL_WAIT_VBLANK to operate correctly.
-	 */
-	drm->irq_enabled = true;
-
-	/* syncpoints are used for full 32-bit hardware VBLANK counters */
-	drm->max_vblank_count = 0xffffffff;
-
-	err = drm_vblank_init(drm, drm->mode_config.num_crtc);
-	if (err < 0)
-		goto hub;
-
-	drm_mode_config_reset(drm);
-
-	err = tegra_drm_fb_init(drm);
-	if (err < 0)
-		goto hub;
-
-	return 0;
-
-hub:
-	if (tegra->hub)
-		tegra_display_hub_cleanup(tegra->hub);
-device:
-	host1x_device_exit(device);
-fbdev:
-	drm_kms_helper_poll_fini(drm);
-	tegra_drm_fb_free(drm);
-config:
-	drm_mode_config_cleanup(drm);
-
-	if (tegra->domain) {
-		mutex_destroy(&tegra->mm_lock);
-		drm_mm_takedown(&tegra->mm);
-		put_iova_domain(&tegra->carveout.domain);
-		iova_cache_put();
-	}
-domain:
-	if (tegra->domain)
-		iommu_domain_free(tegra->domain);
-free:
-	kfree(tegra);
-	return err;
-}
-
-static void tegra_drm_unload(struct drm_device *drm)
-{
-	struct host1x_device *device = to_host1x_device(drm->dev);
-	struct tegra_drm *tegra = drm->dev_private;
-	int err;
-
-	drm_kms_helper_poll_fini(drm);
-	tegra_drm_fb_exit(drm);
-	drm_atomic_helper_shutdown(drm);
-	drm_mode_config_cleanup(drm);
-
-	err = host1x_device_exit(device);
-	if (err < 0)
-		return;
-
-	if (tegra->domain) {
-		mutex_destroy(&tegra->mm_lock);
-		drm_mm_takedown(&tegra->mm);
-		put_iova_domain(&tegra->carveout.domain);
-		iova_cache_put();
-		iommu_domain_free(tegra->domain);
-	}
-
-	kfree(tegra);
-}
-
 static int tegra_drm_open(struct drm_device *drm, struct drm_file *filp)
 {
 	struct tegra_drm_file *fpriv;
@@ -311,6 +145,8 @@ static int host1x_reloc_copy_from_user(struct host1x_reloc *dest,
 	if (err < 0)
 		return err;
 
+	dest->flags = HOST1X_RELOC_READ | HOST1X_RELOC_WRITE;
+
 	dest->cmdbuf.bo = host1x_bo_lookup(file, cmdbuf);
 	if (!dest->cmdbuf.bo)
 		return -ENOENT;
@@ -1014,8 +850,6 @@ static int tegra_debugfs_init(struct drm_minor *minor)
 static struct drm_driver tegra_drm_driver = {
 	.driver_features = DRIVER_MODESET | DRIVER_GEM |
 			   DRIVER_ATOMIC | DRIVER_RENDER,
-	.load = tegra_drm_load,
-	.unload = tegra_drm_unload,
 	.open = tegra_drm_open,
 	.postclose = tegra_drm_postclose,
 	.lastclose = drm_fb_helper_lastclose,
@@ -1068,57 +902,63 @@ int tegra_drm_unregister_client(struct tegra_drm *tegra,
 	return 0;
 }
 
-struct iommu_group *host1x_client_iommu_attach(struct host1x_client *client,
-					       bool shared)
+int host1x_client_iommu_attach(struct host1x_client *client)
 {
+	struct iommu_domain *domain = iommu_get_domain_for_dev(client->dev);
 	struct drm_device *drm = dev_get_drvdata(client->parent);
 	struct tegra_drm *tegra = drm->dev_private;
 	struct iommu_group *group = NULL;
 	int err;
 
+	/*
+	 * If the host1x client is already attached to an IOMMU domain that is
+	 * not the shared IOMMU domain, don't try to attach it to a different
+	 * domain. This allows using the IOMMU-backed DMA API.
+	 */
+	if (domain && domain != tegra->domain)
+		return 0;
+
 	if (tegra->domain) {
 		group = iommu_group_get(client->dev);
 		if (!group) {
 			dev_err(client->dev, "failed to get IOMMU group\n");
-			return ERR_PTR(-ENODEV);
+			return -ENODEV;
 		}
 
-		if (!shared || (shared && (group != tegra->group))) {
-#if IS_ENABLED(CONFIG_ARM_DMA_USE_IOMMU)
-			if (client->dev->archdata.mapping) {
-				struct dma_iommu_mapping *mapping =
-					to_dma_iommu_mapping(client->dev);
-				arm_iommu_detach_device(client->dev);
-				arm_iommu_release_mapping(mapping);
-			}
-#endif
+		if (domain != tegra->domain) {
 			err = iommu_attach_group(tegra->domain, group);
 			if (err < 0) {
 				iommu_group_put(group);
-				return ERR_PTR(err);
+				return err;
 			}
-
-			if (shared && !tegra->group)
-				tegra->group = group;
 		}
+
+		tegra->use_explicit_iommu = true;
 	}
 
-	return group;
+	client->group = group;
+
+	return 0;
 }
 
-void host1x_client_iommu_detach(struct host1x_client *client,
-				struct iommu_group *group)
+void host1x_client_iommu_detach(struct host1x_client *client)
 {
 	struct drm_device *drm = dev_get_drvdata(client->parent);
 	struct tegra_drm *tegra = drm->dev_private;
+	struct iommu_domain *domain;
 
-	if (group) {
-		if (group == tegra->group) {
-			iommu_detach_group(tegra->domain, group);
-			tegra->group = NULL;
-		}
+	if (client->group) {
+		/*
+		 * Devices that are part of the same group may no longer be
+		 * attached to a domain at this point because their group may
+		 * have been detached by an earlier client.
+		 */
+		domain = iommu_get_domain_for_dev(client->dev);
+		if (domain)
+			iommu_detach_group(tegra->domain, client->group);
 
-		iommu_group_put(group);
+		iommu_group_put(client->group);
+		client->group = NULL;
 	}
 }
 
@@ -1202,6 +1042,8 @@ void tegra_drm_free(struct tegra_drm *tegra, size_t size, void *virt,
 static int host1x_drm_probe(struct host1x_device *dev)
 {
 	struct drm_driver *driver = &tegra_drm_driver;
+	struct iommu_domain *domain;
+	struct tegra_drm *tegra;
 	struct drm_device *drm;
 	int err;
 
@@ -1209,18 +1051,180 @@ static int host1x_drm_probe(struct host1x_device *dev)
 	if (IS_ERR(drm))
 		return PTR_ERR(drm);
 
+	tegra = kzalloc(sizeof(*tegra), GFP_KERNEL);
+	if (!tegra) {
+		err = -ENOMEM;
+		goto put;
+	}
+
+	/*
+	 * If the Tegra DRM clients are backed by an IOMMU, push buffers are
+	 * likely to be allocated beyond the 32-bit boundary if sufficient
+	 * system memory is available. This is problematic on earlier Tegra
+	 * generations where host1x supports a maximum of 32 address bits in
+	 * the GATHER opcode. In this case, unless host1x is behind an IOMMU
+	 * as well it won't be able to process buffers allocated beyond the
+	 * 32-bit boundary.
+	 *
+	 * The DMA API will use bounce buffers in this case, so that could
+	 * perhaps still be made to work, even if less efficient, but there
+	 * is another catch: in order to perform cache maintenance on pages
+	 * allocated for discontiguous buffers we need to map and unmap the
+	 * SG table representing these buffers. This is fine for something
+	 * small like a push buffer, but it exhausts the bounce buffer pool
+	 * (typically on the order of a few MiB) for framebuffers (many MiB
+	 * for any modern resolution).
+	 *
+	 * Work around this by making sure that Tegra DRM clients only use
+	 * an IOMMU if the parent host1x also uses an IOMMU.
+	 *
+	 * Note that there's still a small gap here that we don't cover: if
+	 * the DMA API is backed by an IOMMU there's no way to control which
+	 * device is attached to an IOMMU and which isn't, except via wiring
+	 * up the device tree appropriately. This is considered an problem
+	 * of integration, so care must be taken for the DT to be consistent.
+	 */
+	domain = iommu_get_domain_for_dev(drm->dev->parent);
+
+	if (domain && iommu_present(&platform_bus_type)) {
+		tegra->domain = iommu_domain_alloc(&platform_bus_type);
+		if (!tegra->domain) {
+			err = -ENOMEM;
+			goto free;
+		}
+
+		err = iova_cache_get();
+		if (err < 0)
+			goto domain;
+	}
+
+	mutex_init(&tegra->clients_lock);
+	INIT_LIST_HEAD(&tegra->clients);
+
 	dev_set_drvdata(&dev->dev, drm);
+	drm->dev_private = tegra;
+	tegra->drm = drm;
+
+	drm_mode_config_init(drm);
 
-	err = drm_fb_helper_remove_conflicting_framebuffers(NULL, "tegradrmfb", false);
+	drm->mode_config.min_width = 0;
+	drm->mode_config.min_height = 0;
+
+	drm->mode_config.max_width = 4096;
+	drm->mode_config.max_height = 4096;
+
+	drm->mode_config.allow_fb_modifiers = true;
+
+	drm->mode_config.normalize_zpos = true;
+
+	drm->mode_config.funcs = &tegra_drm_mode_config_funcs;
+	drm->mode_config.helper_private = &tegra_drm_mode_config_helpers;
+
+	err = tegra_drm_fb_prepare(drm);
 	if (err < 0)
-		goto put;
+		goto config;
+
+	drm_kms_helper_poll_init(drm);
+
+	err = host1x_device_init(dev);
+	if (err < 0)
+		goto fbdev;
+
+	if (tegra->use_explicit_iommu) {
+		u64 carveout_start, carveout_end, gem_start, gem_end;
+		u64 dma_mask = dma_get_mask(&dev->dev);
+		dma_addr_t start, end;
+		unsigned long order;
+
+		start = tegra->domain->geometry.aperture_start & dma_mask;
+		end = tegra->domain->geometry.aperture_end & dma_mask;
+
+		gem_start = start;
+		gem_end = end - CARVEOUT_SZ;
+		carveout_start = gem_end + 1;
+		carveout_end = end;
+
+		order = __ffs(tegra->domain->pgsize_bitmap);
+		init_iova_domain(&tegra->carveout.domain, 1UL << order,
+				 carveout_start >> order);
+
+		tegra->carveout.shift = iova_shift(&tegra->carveout.domain);
+		tegra->carveout.limit = carveout_end >> tegra->carveout.shift;
+
+		drm_mm_init(&tegra->mm, gem_start, gem_end - gem_start + 1);
+		mutex_init(&tegra->mm_lock);
+
+		DRM_DEBUG_DRIVER("IOMMU apertures:\n");
+		DRM_DEBUG_DRIVER("  GEM: %#llx-%#llx\n", gem_start, gem_end);
+		DRM_DEBUG_DRIVER("  Carveout: %#llx-%#llx\n", carveout_start,
+				 carveout_end);
+	} else if (tegra->domain) {
+		iommu_domain_free(tegra->domain);
+		tegra->domain = NULL;
+		iova_cache_put();
+	}
+
+	if (tegra->hub) {
+		err = tegra_display_hub_prepare(tegra->hub);
+		if (err < 0)
+			goto device;
+	}
+
+	/*
+	 * We don't use the drm_irq_install() helpers provided by the DRM
+	 * core, so we need to set this manually in order to allow the
+	 * DRM_IOCTL_WAIT_VBLANK to operate correctly.
+	 */
+	drm->irq_enabled = true;
+
+	/* syncpoints are used for full 32-bit hardware VBLANK counters */
+	drm->max_vblank_count = 0xffffffff;
+
+	err = drm_vblank_init(drm, drm->mode_config.num_crtc);
+	if (err < 0)
+		goto hub;
+
+	drm_mode_config_reset(drm);
+
+	err = drm_fb_helper_remove_conflicting_framebuffers(NULL, "tegradrmfb",
+							    false);
+	if (err < 0)
+		goto hub;
+
+	err = tegra_drm_fb_init(drm);
+	if (err < 0)
+		goto hub;
 
 	err = drm_dev_register(drm, 0);
 	if (err < 0)
-		goto put;
+		goto fb;
 
 	return 0;
 
+fb:
+	tegra_drm_fb_exit(drm);
+hub:
+	if (tegra->hub)
+		tegra_display_hub_cleanup(tegra->hub);
+device:
+	if (tegra->domain) {
+		mutex_destroy(&tegra->mm_lock);
+		drm_mm_takedown(&tegra->mm);
+		put_iova_domain(&tegra->carveout.domain);
+		iova_cache_put();
+	}
+
+	host1x_device_exit(dev);
+fbdev:
+	drm_kms_helper_poll_fini(drm);
+	tegra_drm_fb_free(drm);
+config:
+	drm_mode_config_cleanup(drm);
+domain:
+	if (tegra->domain)
+		iommu_domain_free(tegra->domain);
+free:
+	kfree(tegra);
 put:
 	drm_dev_put(drm);
 	return err;
@@ -1229,8 +1233,29 @@ put:
 static int host1x_drm_remove(struct host1x_device *dev)
 {
 	struct drm_device *drm = dev_get_drvdata(&dev->dev);
+	struct tegra_drm *tegra = drm->dev_private;
+	int err;
 
 	drm_dev_unregister(drm);
+
+	drm_kms_helper_poll_fini(drm);
+	tegra_drm_fb_exit(drm);
+	drm_atomic_helper_shutdown(drm);
+	drm_mode_config_cleanup(drm);
+
+	err = host1x_device_exit(dev);
+	if (err < 0)
+		dev_err(&dev->dev, "host1x device cleanup failed: %d\n", err);
+
+	if (tegra->domain) {
+		mutex_destroy(&tegra->mm_lock);
+		drm_mm_takedown(&tegra->mm);
+		put_iova_domain(&tegra->carveout.domain);
+		iova_cache_put();
+		iommu_domain_free(tegra->domain);
+	}
+
+	kfree(tegra);
 	drm_dev_put(drm);
 
 	return 0;
diff --git a/drivers/gpu/drm/tegra/drm.h b/drivers/gpu/drm/tegra/drm.h
index 29911eff9ceb..d941553f7a3d 100644
--- a/drivers/gpu/drm/tegra/drm.h
+++ b/drivers/gpu/drm/tegra/drm.h
@@ -36,7 +36,7 @@ struct tegra_drm {
 	struct drm_device *drm;
 
 	struct iommu_domain *domain;
-	struct iommu_group *group;
+	bool use_explicit_iommu;
 	struct mutex mm_lock;
 	struct drm_mm mm;
 
@@ -100,10 +100,8 @@ int tegra_drm_register_client(struct tegra_drm *tegra,
 			      struct tegra_drm_client *client);
 int tegra_drm_unregister_client(struct tegra_drm *tegra,
 				struct tegra_drm_client *client);
-struct iommu_group *host1x_client_iommu_attach(struct host1x_client *client,
-					       bool shared);
-void host1x_client_iommu_detach(struct host1x_client *client,
-				struct iommu_group *group);
+int host1x_client_iommu_attach(struct host1x_client *client);
+void host1x_client_iommu_detach(struct host1x_client *client);
 
 int tegra_drm_init(struct tegra_drm *tegra, struct drm_device *drm);
 int tegra_drm_exit(struct tegra_drm *tegra);
@@ -155,17 +153,12 @@ void tegra_output_connector_destroy(struct drm_connector *connector);
 void tegra_output_encoder_destroy(struct drm_encoder *encoder);
 
 /* from dpaux.c */
-struct drm_dp_link;
-
 struct drm_dp_aux *drm_dp_aux_find_by_of_node(struct device_node *np);
 enum drm_connector_status drm_dp_aux_detect(struct drm_dp_aux *aux);
 int drm_dp_aux_attach(struct drm_dp_aux *aux, struct tegra_output *output);
 int drm_dp_aux_detach(struct drm_dp_aux *aux);
 int drm_dp_aux_enable(struct drm_dp_aux *aux);
 int drm_dp_aux_disable(struct drm_dp_aux *aux);
-int drm_dp_aux_prepare(struct drm_dp_aux *aux, u8 encoding);
-int drm_dp_aux_train(struct drm_dp_aux *aux, struct drm_dp_link *link,
-		     u8 pattern);
 
 /* from fb.c */
 struct tegra_bo *tegra_fb_get_plane(struct drm_framebuffer *framebuffer,
diff --git a/drivers/gpu/drm/tegra/falcon.c b/drivers/gpu/drm/tegra/falcon.c
index f49ad36e24db..56edef06c48e 100644
--- a/drivers/gpu/drm/tegra/falcon.c
+++ b/drivers/gpu/drm/tegra/falcon.c
@@ -58,32 +58,17 @@ static int falcon_copy_chunk(struct falcon *falcon,
 static void falcon_copy_firmware_image(struct falcon *falcon,
 				       const struct firmware *firmware)
 {
-	u32 *firmware_vaddr = falcon->firmware.vaddr;
-	dma_addr_t daddr;
+	u32 *virt = falcon->firmware.virt;
 	size_t i;
-	int err;
 
 	/* copy the whole thing taking into account endianness */
 	for (i = 0; i < firmware->size / sizeof(u32); i++)
-		firmware_vaddr[i] = le32_to_cpu(((u32 *)firmware->data)[i]);
-
-	/* ensure that caches are flushed and falcon can see the firmware */
-	daddr = dma_map_single(falcon->dev, firmware_vaddr,
-			       falcon->firmware.size, DMA_TO_DEVICE);
-	err = dma_mapping_error(falcon->dev, daddr);
-	if (err) {
-		dev_err(falcon->dev, "failed to map firmware: %d\n", err);
-		return;
-	}
-	dma_sync_single_for_device(falcon->dev, daddr,
-				   falcon->firmware.size, DMA_TO_DEVICE);
-	dma_unmap_single(falcon->dev, daddr, falcon->firmware.size,
-			 DMA_TO_DEVICE);
+		virt[i] = le32_to_cpu(((u32 *)firmware->data)[i]);
 }
 
 static int falcon_parse_firmware_image(struct falcon *falcon)
 {
-	struct falcon_fw_bin_header_v1 *bin = (void *)falcon->firmware.vaddr;
+	struct falcon_fw_bin_header_v1 *bin = (void *)falcon->firmware.virt;
 	struct falcon_fw_os_header_v1 *os;
 
 	/* endian problems would show up right here */
@@ -104,7 +89,7 @@ static int falcon_parse_firmware_image(struct falcon *falcon)
 		return -EINVAL;
 	}
 
-	os = falcon->firmware.vaddr + bin->os_header_offset;
+	os = falcon->firmware.virt + bin->os_header_offset;
 
 	falcon->firmware.bin_data.size = bin->os_size;
 	falcon->firmware.bin_data.offset = bin->os_data_offset;
@@ -125,6 +110,8 @@ int falcon_read_firmware(struct falcon *falcon, const char *name)
 	if (err < 0)
 		return err;
 
+	falcon->firmware.size = falcon->firmware.firmware->size;
+
 	return 0;
 }
 
@@ -133,16 +120,6 @@ int falcon_load_firmware(struct falcon *falcon)
 	const struct firmware *firmware = falcon->firmware.firmware;
 	int err;
 
-	falcon->firmware.size = firmware->size;
-
-	/* allocate iova space for the firmware */
-	falcon->firmware.vaddr = falcon->ops->alloc(falcon, firmware->size,
-						    &falcon->firmware.paddr);
-	if (IS_ERR(falcon->firmware.vaddr)) {
-		dev_err(falcon->dev, "DMA memory mapping failed\n");
-		return PTR_ERR(falcon->firmware.vaddr);
-	}
-
 	/* copy firmware image into local area. this also ensures endianness */
 	falcon_copy_firmware_image(falcon, firmware);
 
@@ -150,45 +127,26 @@ int falcon_load_firmware(struct falcon *falcon)
 	err = falcon_parse_firmware_image(falcon);
 	if (err < 0) {
 		dev_err(falcon->dev, "failed to parse firmware image\n");
-		goto err_setup_firmware_image;
+		return err;
 	}
 
 	release_firmware(firmware);
 	falcon->firmware.firmware = NULL;
 
 	return 0;
-
-err_setup_firmware_image:
-	falcon->ops->free(falcon, falcon->firmware.size,
-			  falcon->firmware.paddr, falcon->firmware.vaddr);
-
-	return err;
 }
 
 int falcon_init(struct falcon *falcon)
 {
-	/* check mandatory ops */
-	if (!falcon->ops || !falcon->ops->alloc || !falcon->ops->free)
-		return -EINVAL;
-
-	falcon->firmware.vaddr = NULL;
+	falcon->firmware.virt = NULL;
 
 	return 0;
 }
 
 void falcon_exit(struct falcon *falcon)
 {
-	if (falcon->firmware.firmware) {
+	if (falcon->firmware.firmware)
 		release_firmware(falcon->firmware.firmware);
-		falcon->firmware.firmware = NULL;
-	}
-
-	if (falcon->firmware.vaddr) {
-		falcon->ops->free(falcon, falcon->firmware.size,
-				  falcon->firmware.paddr,
-				  falcon->firmware.vaddr);
-		falcon->firmware.vaddr = NULL;
-	}
 }
 
 int falcon_boot(struct falcon *falcon)
@@ -197,7 +155,7 @@ int falcon_boot(struct falcon *falcon)
 	u32 value;
 	int err;
 
-	if (!falcon->firmware.vaddr)
+	if (!falcon->firmware.virt)
 		return -EINVAL;
 
 	err = readl_poll_timeout(falcon->regs + FALCON_DMACTL, value,
@@ -210,7 +168,7 @@ int falcon_boot(struct falcon *falcon)
 	falcon_writel(falcon, 0, FALCON_DMACTL);
 
 	/* setup the address of the binary data so Falcon can access it later */
-	falcon_writel(falcon, (falcon->firmware.paddr +
+	falcon_writel(falcon, (falcon->firmware.iova +
 			       falcon->firmware.bin_data.offset) >> 8,
 		      FALCON_DMATRFBASE);
 
diff --git a/drivers/gpu/drm/tegra/falcon.h b/drivers/gpu/drm/tegra/falcon.h
index 3d1243217410..c56ee32d92ee 100644
--- a/drivers/gpu/drm/tegra/falcon.h
+++ b/drivers/gpu/drm/tegra/falcon.h
@@ -74,15 +74,6 @@ struct falcon_fw_os_header_v1 {
 	u32 data_size;
 };
 
-struct falcon;
-
-struct falcon_ops {
-	void *(*alloc)(struct falcon *falcon, size_t size,
-		       dma_addr_t *paddr);
-	void (*free)(struct falcon *falcon, size_t size,
-		     dma_addr_t paddr, void *vaddr);
-};
-
 struct falcon_firmware_section {
 	unsigned long offset;
 	size_t size;
@@ -93,8 +84,9 @@ struct falcon_firmware {
 	const struct firmware *firmware;
 
 	/* Raw firmware data */
-	dma_addr_t paddr;
-	void *vaddr;
+	dma_addr_t iova;
+	dma_addr_t phys;
+	void *virt;
 	size_t size;
 
 	/* Parsed firmware information */
@@ -107,8 +99,6 @@ struct falcon {
 	/* Set by falcon client */
 	struct device *dev;
 	void __iomem *regs;
-	const struct falcon_ops *ops;
-	void *data;
 
 	struct falcon_firmware firmware;
 };
diff --git a/drivers/gpu/drm/tegra/fb.c b/drivers/gpu/drm/tegra/fb.c
index e34325c83d28..7cea89f29a5c 100644
--- a/drivers/gpu/drm/tegra/fb.c
+++ b/drivers/gpu/drm/tegra/fb.c
@@ -269,10 +269,10 @@ static int tegra_fbdev_probe(struct drm_fb_helper *helper,
 		}
 	}
 
-	drm->mode_config.fb_base = (resource_size_t)bo->paddr;
+	drm->mode_config.fb_base = (resource_size_t)bo->iova;
 	info->screen_base = (void __iomem *)bo->vaddr + offset;
 	info->screen_size = size;
-	info->fix.smem_start = (unsigned long)(bo->paddr + offset);
+	info->fix.smem_start = (unsigned long)(bo->iova + offset);
 	info->fix.smem_len = size;
 
 	return 0;
diff --git a/drivers/gpu/drm/tegra/gem.c b/drivers/gpu/drm/tegra/gem.c
index fb7667c8dd4c..746dae32c484 100644
--- a/drivers/gpu/drm/tegra/gem.c
+++ b/drivers/gpu/drm/tegra/gem.c
@@ -27,17 +27,55 @@ static void tegra_bo_put(struct host1x_bo *bo)
 	drm_gem_object_put_unlocked(&obj->gem);
 }
 
-static dma_addr_t tegra_bo_pin(struct host1x_bo *bo, struct sg_table **sgt)
+static struct sg_table *tegra_bo_pin(struct device *dev, struct host1x_bo *bo,
+				     dma_addr_t *phys)
 {
 	struct tegra_bo *obj = host1x_to_tegra_bo(bo);
+	struct sg_table *sgt;
+	int err;
+
+	/*
+	 * If we've manually mapped the buffer object through the IOMMU, make
+	 * sure to return the IOVA address of our mapping.
+	 */
+	if (phys && obj->mm) {
+		*phys = obj->iova;
+		return NULL;
+	}
+
+	/*
+	 * If we don't have a mapping for this buffer yet, return an SG table
+	 * so that host1x can do the mapping for us via the DMA API.
+	 */
+	sgt = kzalloc(sizeof(*sgt), GFP_KERNEL);
+	if (!sgt)
+		return ERR_PTR(-ENOMEM);
 
-	*sgt = obj->sgt;
+	if (obj->pages) {
+		err = sg_alloc_table_from_pages(sgt, obj->pages, obj->num_pages,
+						0, obj->gem.size, GFP_KERNEL);
+		if (err < 0)
+			goto free;
+	} else {
+		err = dma_get_sgtable(dev, sgt, obj->vaddr, obj->iova,
+				      obj->gem.size);
+		if (err < 0)
+			goto free;
+	}
 
-	return obj->paddr;
+	return sgt;
+
+free:
+	kfree(sgt);
+	return ERR_PTR(err);
 }
 
-static void tegra_bo_unpin(struct host1x_bo *bo, struct sg_table *sgt)
+static void tegra_bo_unpin(struct device *dev, struct sg_table *sgt)
 {
+	if (sgt) {
+		sg_free_table(sgt);
+		kfree(sgt);
+	}
 }
 
 static void *tegra_bo_mmap(struct host1x_bo *bo)
@@ -133,9 +171,9 @@ static int tegra_bo_iommu_map(struct tegra_drm *tegra, struct tegra_bo *bo)
 		goto unlock;
 	}
 
-	bo->paddr = bo->mm->start;
+	bo->iova = bo->mm->start;
 
-	bo->size = iommu_map_sg(tegra->domain, bo->paddr, bo->sgt->sgl,
+	bo->size = iommu_map_sg(tegra->domain, bo->iova, bo->sgt->sgl,
 				bo->sgt->nents, prot);
 	if (!bo->size) {
 		dev_err(tegra->drm->dev, "failed to map buffer\n");
@@ -161,7 +199,7 @@ static int tegra_bo_iommu_unmap(struct tegra_drm *tegra, struct tegra_bo *bo)
 		return 0;
 
 	mutex_lock(&tegra->mm_lock);
-	iommu_unmap(tegra->domain, bo->paddr, bo->size);
+	iommu_unmap(tegra->domain, bo->iova, bo->size);
 	drm_mm_remove_node(bo->mm);
 	mutex_unlock(&tegra->mm_lock);
 
@@ -209,7 +247,7 @@ static void tegra_bo_free(struct drm_device *drm, struct tegra_bo *bo)
 		sg_free_table(bo->sgt);
 		kfree(bo->sgt);
 	} else if (bo->vaddr) {
-		dma_free_wc(drm->dev, bo->gem.size, bo->vaddr, bo->paddr);
+		dma_free_wc(drm->dev, bo->gem.size, bo->vaddr, bo->iova);
 	}
 }
 
@@ -264,7 +302,7 @@ static int tegra_bo_alloc(struct drm_device *drm, struct tegra_bo *bo)
 	} else {
 		size_t size = bo->gem.size;
 
-		bo->vaddr = dma_alloc_wc(drm->dev, size, &bo->paddr,
+		bo->vaddr = dma_alloc_wc(drm->dev, size, &bo->iova,
 					 GFP_KERNEL | __GFP_NOWARN);
 		if (!bo->vaddr) {
 			dev_err(drm->dev,
@@ -365,7 +403,7 @@ static struct tegra_bo *tegra_bo_import(struct drm_device *drm,
 			goto detach;
 		}
 
-		bo->paddr = sg_dma_address(bo->sgt->sgl);
+		bo->iova = sg_dma_address(bo->sgt->sgl);
 	}
 
 	bo->gem.import_attach = attach;
@@ -461,7 +499,7 @@ int __tegra_gem_mmap(struct drm_gem_object *gem, struct vm_area_struct *vma)
 		vma->vm_flags &= ~VM_PFNMAP;
 		vma->vm_pgoff = 0;
 
-		err = dma_mmap_wc(gem->dev->dev, vma, bo->vaddr, bo->paddr,
+		err = dma_mmap_wc(gem->dev->dev, vma, bo->vaddr, bo->iova,
 				  gem->size);
 		if (err < 0) {
 			drm_gem_vm_close(vma);
@@ -508,25 +546,18 @@ tegra_gem_prime_map_dma_buf(struct dma_buf_attachment *attach,
 		return NULL;
 
 	if (bo->pages) {
-		struct scatterlist *sg;
-		unsigned int i;
-
-		if (sg_alloc_table(sgt, bo->num_pages, GFP_KERNEL))
-			goto free;
-
-		for_each_sg(sgt->sgl, sg, bo->num_pages, i)
-			sg_set_page(sg, bo->pages[i], PAGE_SIZE, 0);
-
-		if (dma_map_sg(attach->dev, sgt->sgl, sgt->nents, dir) == 0)
+		if (sg_alloc_table_from_pages(sgt, bo->pages, bo->num_pages,
+					      0, gem->size, GFP_KERNEL) < 0)
 			goto free;
 	} else {
-		if (sg_alloc_table(sgt, 1, GFP_KERNEL))
+		if (dma_get_sgtable(attach->dev, sgt, bo->vaddr, bo->iova,
+				    gem->size) < 0)
 			goto free;
-
-		sg_dma_address(sgt->sgl) = bo->paddr;
-		sg_dma_len(sgt->sgl) = gem->size;
 	}
 
+	if (dma_map_sg(attach->dev, sgt->sgl, sgt->nents, dir) == 0)
+		goto free;
+
 	return sgt;
 
 free:
diff --git a/drivers/gpu/drm/tegra/gem.h b/drivers/gpu/drm/tegra/gem.h
index 83ffb1e14ca3..fafb5724499b 100644
--- a/drivers/gpu/drm/tegra/gem.h
+++ b/drivers/gpu/drm/tegra/gem.h
@@ -31,7 +31,7 @@ struct tegra_bo {
 	struct host1x_bo base;
 	unsigned long flags;
 	struct sg_table *sgt;
-	dma_addr_t paddr;
+	dma_addr_t iova;
 	void *vaddr;
 
 	struct drm_mm_node *mm;
diff --git a/drivers/gpu/drm/tegra/gr2d.c b/drivers/gpu/drm/tegra/gr2d.c
index 641299cc85b8..1fc4e56c7cc5 100644
--- a/drivers/gpu/drm/tegra/gr2d.c
+++ b/drivers/gpu/drm/tegra/gr2d.c
@@ -17,7 +17,6 @@ struct gr2d_soc {
 };
 
 struct gr2d {
-	struct iommu_group *group;
 	struct tegra_drm_client client;
 	struct host1x_channel *channel;
 	struct clk *clk;
@@ -40,7 +39,7 @@ static int gr2d_init(struct host1x_client *client)
 	struct gr2d *gr2d = to_gr2d(drm);
 	int err;
 
-	gr2d->channel = host1x_channel_request(client->dev);
+	gr2d->channel = host1x_channel_request(client);
 	if (!gr2d->channel)
 		return -ENOMEM;
 
@@ -51,9 +50,8 @@ static int gr2d_init(struct host1x_client *client)
 		goto put;
 	}
 
-	gr2d->group = host1x_client_iommu_attach(client, false);
-	if (IS_ERR(gr2d->group)) {
-		err = PTR_ERR(gr2d->group);
+	err = host1x_client_iommu_attach(client);
+	if (err < 0) {
 		dev_err(client->dev, "failed to attach to domain: %d\n", err);
 		goto free;
 	}
@@ -67,7 +65,7 @@ static int gr2d_init(struct host1x_client *client)
 	return 0;
 
 detach:
-	host1x_client_iommu_detach(client, gr2d->group);
+	host1x_client_iommu_detach(client);
 free:
 	host1x_syncpt_free(client->syncpts[0]);
 put:
@@ -87,7 +85,7 @@ static int gr2d_exit(struct host1x_client *client)
 	if (err < 0)
 		return err;
 
-	host1x_client_iommu_detach(client, gr2d->group);
+	host1x_client_iommu_detach(client);
 	host1x_syncpt_free(client->syncpts[0]);
 	host1x_channel_put(gr2d->channel);
 
diff --git a/drivers/gpu/drm/tegra/gr3d.c b/drivers/gpu/drm/tegra/gr3d.c
index 8b9a35b1cbb3..24fae0f64032 100644
--- a/drivers/gpu/drm/tegra/gr3d.c
+++ b/drivers/gpu/drm/tegra/gr3d.c
@@ -23,7 +23,6 @@ struct gr3d_soc {
 };
 
 struct gr3d {
-	struct iommu_group *group;
 	struct tegra_drm_client client;
 	struct host1x_channel *channel;
 	struct clk *clk_secondary;
@@ -49,7 +48,7 @@ static int gr3d_init(struct host1x_client *client)
 	struct gr3d *gr3d = to_gr3d(drm);
 	int err;
 
-	gr3d->channel = host1x_channel_request(client->dev);
+	gr3d->channel = host1x_channel_request(client);
 	if (!gr3d->channel)
 		return -ENOMEM;
 
@@ -60,9 +59,8 @@ static int gr3d_init(struct host1x_client *client)
 		goto put;
 	}
 
-	gr3d->group = host1x_client_iommu_attach(client, false);
-	if (IS_ERR(gr3d->group)) {
-		err = PTR_ERR(gr3d->group);
+	err = host1x_client_iommu_attach(client);
+	if (err < 0) {
 		dev_err(client->dev, "failed to attach to domain: %d\n", err);
 		goto free;
 	}
@@ -76,7 +74,7 @@ static int gr3d_init(struct host1x_client *client)
 	return 0;
 
 detach:
-	host1x_client_iommu_detach(client, gr3d->group);
+	host1x_client_iommu_detach(client);
 free:
 	host1x_syncpt_free(client->syncpts[0]);
 put:
@@ -95,7 +93,7 @@ static int gr3d_exit(struct host1x_client *client)
 	if (err < 0)
 		return err;
 
-	host1x_client_iommu_detach(client, gr3d->group);
+	host1x_client_iommu_detach(client);
 	host1x_syncpt_free(client->syncpts[0]);
 	host1x_channel_put(gr3d->channel);
 
diff --git a/drivers/gpu/drm/tegra/hub.c b/drivers/gpu/drm/tegra/hub.c
index 839b49c40e51..2b4082d0bc9e 100644
--- a/drivers/gpu/drm/tegra/hub.c
+++ b/drivers/gpu/drm/tegra/hub.c
@@ -413,7 +413,6 @@ static void tegra_shared_plane_atomic_update(struct drm_plane *plane,
 	unsigned int zpos = plane->state->normalized_zpos;
 	struct drm_framebuffer *fb = plane->state->fb;
 	struct tegra_plane *p = to_tegra_plane(plane);
-	struct tegra_bo *bo;
 	dma_addr_t base;
 	u32 value;
 
@@ -456,8 +455,7 @@ static void tegra_shared_plane_atomic_update(struct drm_plane *plane,
 	/* disable compression */
 	tegra_plane_writel(p, 0, DC_WINBUF_CDE_CONTROL);
 
-	bo = tegra_fb_get_plane(fb, 0);
-	base = bo->paddr;
+	base = state->iova[0] + fb->offsets[0];
 
 	tegra_plane_writel(p, state->format, DC_WIN_COLOR_DEPTH);
 	tegra_plane_writel(p, 0, DC_WIN_PRECOMP_WGRP_PARAMS);
@@ -521,6 +519,8 @@ static void tegra_shared_plane_atomic_update(struct drm_plane *plane,
 }
 
 static const struct drm_plane_helper_funcs tegra_shared_plane_helper_funcs = {
+	.prepare_fb = tegra_plane_prepare_fb,
+	.cleanup_fb = tegra_plane_cleanup_fb,
 	.atomic_check = tegra_shared_plane_atomic_check,
 	.atomic_update = tegra_shared_plane_atomic_update,
 	.atomic_disable = tegra_shared_plane_atomic_disable,
diff --git a/drivers/gpu/drm/tegra/output.c b/drivers/gpu/drm/tegra/output.c
index bdcaa4c7168c..34373734ff68 100644
--- a/drivers/gpu/drm/tegra/output.c
+++ b/drivers/gpu/drm/tegra/output.c
@@ -70,6 +70,11 @@ tegra_output_connector_detect(struct drm_connector *connector, bool force)
 
 void tegra_output_connector_destroy(struct drm_connector *connector)
 {
+	struct tegra_output *output = connector_to_output(connector);
+
+	if (output->cec)
+		cec_notifier_conn_unregister(output->cec);
+
 	drm_connector_unregister(connector);
 	drm_connector_cleanup(connector);
 }
@@ -163,18 +168,11 @@ int tegra_output_probe(struct tegra_output *output)
 		disable_irq(output->hpd_irq);
 	}
 
-	output->cec = cec_notifier_get(output->dev);
-	if (!output->cec)
-		return -ENOMEM;
-
 	return 0;
 }
 
 void tegra_output_remove(struct tegra_output *output)
 {
-	if (output->cec)
-		cec_notifier_put(output->cec);
-
 	if (output->hpd_gpio)
 		free_irq(output->hpd_irq, output);
 
@@ -184,6 +182,7 @@ void tegra_output_remove(struct tegra_output *output)
 
 int tegra_output_init(struct drm_device *drm, struct tegra_output *output)
 {
+	int connector_type;
 	int err;
 
 	if (output->panel) {
@@ -199,6 +198,21 @@ int tegra_output_init(struct drm_device *drm, struct tegra_output *output)
 	if (output->hpd_gpio)
 		enable_irq(output->hpd_irq);
 
+	connector_type = output->connector.connector_type;
+	/*
+	 * Create a CEC notifier for HDMI connector.
+	 */
+	if (connector_type == DRM_MODE_CONNECTOR_HDMIA ||
+	    connector_type == DRM_MODE_CONNECTOR_HDMIB) {
+		struct cec_connector_info conn_info;
+
+		cec_fill_conn_info_from_drm(&conn_info, &output->connector);
+		output->cec = cec_notifier_conn_register(output->dev, NULL,
+							 &conn_info);
+		if (!output->cec)
+			return -ENOMEM;
+	}
+
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/tegra/plane.c b/drivers/gpu/drm/tegra/plane.c
index 6bab71d6e81d..163b590be224 100644
--- a/drivers/gpu/drm/tegra/plane.c
+++ b/drivers/gpu/drm/tegra/plane.c
@@ -6,6 +6,7 @@
 #include <drm/drm_atomic.h>
 #include <drm/drm_atomic_helper.h>
 #include <drm/drm_fourcc.h>
+#include <drm/drm_gem_framebuffer_helper.h>
 #include <drm/drm_plane_helper.h>
 
 #include "dc.h"
@@ -23,6 +24,7 @@ static void tegra_plane_reset(struct drm_plane *plane)
 {
 	struct tegra_plane *p = to_tegra_plane(plane);
 	struct tegra_plane_state *state;
+	unsigned int i;
 
 	if (plane->state)
 		__drm_atomic_helper_plane_destroy_state(plane->state);
@@ -36,6 +38,9 @@ static void tegra_plane_reset(struct drm_plane *plane)
 		plane->state->plane = plane;
 		plane->state->zpos = p->index;
 		plane->state->normalized_zpos = p->index;
+
+		for (i = 0; i < 3; i++)
+			state->iova[i] = DMA_MAPPING_ERROR;
 	}
 }
 
@@ -60,6 +65,11 @@ tegra_plane_atomic_duplicate_state(struct drm_plane *plane)
 	for (i = 0; i < 2; i++)
 		copy->blending[i] = state->blending[i];
 
+	for (i = 0; i < 3; i++) {
+		copy->iova[i] = DMA_MAPPING_ERROR;
+		copy->sgt[i] = NULL;
+	}
+
 	return &copy->base;
 }
 
@@ -95,6 +105,100 @@ const struct drm_plane_funcs tegra_plane_funcs = {
 	.format_mod_supported = tegra_plane_format_mod_supported,
 };
 
+static int tegra_dc_pin(struct tegra_dc *dc, struct tegra_plane_state *state)
+{
+	unsigned int i;
+	int err;
+
+	for (i = 0; i < state->base.fb->format->num_planes; i++) {
+		struct tegra_bo *bo = tegra_fb_get_plane(state->base.fb, i);
+
+		if (!dc->client.group) {
+			struct sg_table *sgt;
+
+			sgt = host1x_bo_pin(dc->dev, &bo->base, NULL);
+			if (IS_ERR(sgt)) {
+				err = PTR_ERR(sgt);
+				goto unpin;
+			}
+
+			err = dma_map_sg(dc->dev, sgt->sgl, sgt->nents,
+					 DMA_TO_DEVICE);
+			if (err == 0) {
+				err = -ENOMEM;
+				goto unpin;
+			}
+
+			state->iova[i] = sg_dma_address(sgt->sgl);
+			state->sgt[i] = sgt;
+		} else {
+			state->iova[i] = bo->iova;
+		}
+	}
+
+	return 0;
+
+unpin:
+	dev_err(dc->dev, "failed to map plane %u: %d\n", i, err);
+
+	while (i--) {
+		struct tegra_bo *bo = tegra_fb_get_plane(state->base.fb, i);
+		struct sg_table *sgt = state->sgt[i];
+
+		dma_unmap_sg(dc->dev, sgt->sgl, sgt->nents, DMA_TO_DEVICE);
+		host1x_bo_unpin(dc->dev, &bo->base, sgt);
+
+		state->iova[i] = DMA_MAPPING_ERROR;
+		state->sgt[i] = NULL;
+	}
+
+	return err;
+}
+
+static void tegra_dc_unpin(struct tegra_dc *dc, struct tegra_plane_state *state)
+{
+	unsigned int i;
+
+	for (i = 0; i < state->base.fb->format->num_planes; i++) {
+		struct tegra_bo *bo = tegra_fb_get_plane(state->base.fb, i);
+
+		if (!dc->client.group) {
+			struct sg_table *sgt = state->sgt[i];
+
+			if (sgt) {
+				dma_unmap_sg(dc->dev, sgt->sgl, sgt->nents,
+					     DMA_TO_DEVICE);
+				host1x_bo_unpin(dc->dev, &bo->base, sgt);
+			}
+		}
+
+		state->iova[i] = DMA_MAPPING_ERROR;
+		state->sgt[i] = NULL;
+	}
+}
+
+int tegra_plane_prepare_fb(struct drm_plane *plane,
+			   struct drm_plane_state *state)
+{
+	struct tegra_dc *dc = to_tegra_dc(state->crtc);
+
+	if (!state->fb)
+		return 0;
+
+	drm_gem_fb_prepare_fb(plane, state);
+
+	return tegra_dc_pin(dc, to_tegra_plane_state(state));
+}
+
+void tegra_plane_cleanup_fb(struct drm_plane *plane,
+			    struct drm_plane_state *state)
+{
+	struct tegra_dc *dc = to_tegra_dc(state->crtc);
+
+	if (dc)
+		tegra_dc_unpin(dc, to_tegra_plane_state(state));
+}
+
 int tegra_plane_state_add(struct tegra_plane *plane,
 			  struct drm_plane_state *state)
 {
diff --git a/drivers/gpu/drm/tegra/plane.h b/drivers/gpu/drm/tegra/plane.h
index 510c394e6d9a..a158a915109a 100644
--- a/drivers/gpu/drm/tegra/plane.h
+++ b/drivers/gpu/drm/tegra/plane.h
@@ -39,6 +39,9 @@ struct tegra_plane_legacy_blending_state {
 struct tegra_plane_state {
 	struct drm_plane_state base;
 
+	struct sg_table *sgt[3];
+	dma_addr_t iova[3];
+
 	struct tegra_bo_tiling tiling;
 	u32 format;
 	u32 swap;
@@ -61,6 +64,11 @@ to_tegra_plane_state(struct drm_plane_state *state)
 
 extern const struct drm_plane_funcs tegra_plane_funcs;
 
+int tegra_plane_prepare_fb(struct drm_plane *plane,
+			   struct drm_plane_state *state);
+void tegra_plane_cleanup_fb(struct drm_plane *plane,
+			    struct drm_plane_state *state);
+
 int tegra_plane_state_add(struct tegra_plane *plane,
 			  struct drm_plane_state *state);
 
diff --git a/drivers/gpu/drm/tegra/sor.c b/drivers/gpu/drm/tegra/sor.c
index e1669ada0a40..615cb319fa8b 100644
--- a/drivers/gpu/drm/tegra/sor.c
+++ b/drivers/gpu/drm/tegra/sor.c
@@ -25,6 +25,7 @@
 #include <drm/drm_scdc_helper.h>
 
 #include "dc.h"
+#include "dp.h"
 #include "drm.h"
 #include "hda.h"
 #include "sor.h"
@@ -370,10 +371,11 @@ struct tegra_sor_regs {
 };
 
 struct tegra_sor_soc {
-	bool supports_edp;
 	bool supports_lvds;
 	bool supports_hdmi;
 	bool supports_dp;
+	bool supports_audio;
+	bool supports_hdcp;
 
 	const struct tegra_sor_regs *regs;
 	bool has_nvdisplay;
@@ -382,6 +384,12 @@ struct tegra_sor_soc {
 	unsigned int num_settings;
 
 	const u8 *xbar_cfg;
+	const u8 *lane_map;
+
+	const u8 (*voltage_swing)[4][4];
+	const u8 (*pre_emphasis)[4][4];
+	const u8 (*post_cursor)[4][4];
+	const u8 (*tx_pu)[4][4];
 };
 
 struct tegra_sor;
@@ -390,6 +398,8 @@ struct tegra_sor_ops {
 	const char *name;
 	int (*probe)(struct tegra_sor *sor);
 	int (*remove)(struct tegra_sor *sor);
+	void (*audio_enable)(struct tegra_sor *sor);
+	void (*audio_disable)(struct tegra_sor *sor);
 };
 
 struct tegra_sor {
@@ -412,6 +422,7 @@ struct tegra_sor {
 
 	u8 xbar_cfg[5];
 
+	struct drm_dp_link link;
 	struct drm_dp_aux *aux;
 
 	struct drm_info_list *debugfs_files;
@@ -514,10 +525,19 @@ static inline struct tegra_clk_sor_pad *to_pad(struct clk_hw *hw)
 	return container_of(hw, struct tegra_clk_sor_pad, hw);
 }
 
-static const char * const tegra_clk_sor_pad_parents[] = {
-	"pll_d2_out0", "pll_dp"
+static const char * const tegra_clk_sor_pad_parents[2][2] = {
+	{ "pll_d_out0", "pll_dp" },
+	{ "pll_d2_out0", "pll_dp" },
 };
 
+/*
+ * Implementing ->set_parent() here isn't really required because the parent
+ * will be explicitly selected in the driver code via the DP_CLK_SEL mux in
+ * the SOR_CLK_CNTRL register. This is primarily for compatibility with the
+ * Tegra186 and later SoC generations where the BPMP implements this clock
+ * and doesn't expose the mux via the common clock framework.
+ */
+
 static int tegra_clk_sor_pad_set_parent(struct clk_hw *hw, u8 index)
 {
 	struct tegra_clk_sor_pad *pad = to_pad(hw);
@@ -586,8 +606,8 @@ static struct clk *tegra_clk_sor_pad_register(struct tegra_sor *sor,
 
 	init.name = name;
 	init.flags = 0;
-	init.parent_names = tegra_clk_sor_pad_parents;
-	init.num_parents = ARRAY_SIZE(tegra_clk_sor_pad_parents);
+	init.parent_names = tegra_clk_sor_pad_parents[sor->index];
+	init.num_parents = ARRAY_SIZE(tegra_clk_sor_pad_parents[sor->index]);
 	init.ops = &tegra_clk_sor_pad_ops;
 
 	pad->hw.init = &init;
@@ -597,112 +617,340 @@ static struct clk *tegra_clk_sor_pad_register(struct tegra_sor *sor,
 	return clk;
 }
 
-static int tegra_sor_dp_train_fast(struct tegra_sor *sor,
-				   struct drm_dp_link *link)
+static void tegra_sor_filter_rates(struct tegra_sor *sor)
 {
+	struct drm_dp_link *link = &sor->link;
 	unsigned int i;
-	u8 pattern;
+
+	/* Tegra only supports RBR, HBR and HBR2 */
+	for (i = 0; i < link->num_rates; i++) {
+		switch (link->rates[i]) {
+		case 1620000:
+		case 2700000:
+		case 5400000:
+			break;
+
+		default:
+			DRM_DEBUG_KMS("link rate %lu kHz not supported\n",
+				      link->rates[i]);
+			link->rates[i] = 0;
+			break;
+		}
+	}
+
+	drm_dp_link_update_rates(link);
+}
+
+static int tegra_sor_power_up_lanes(struct tegra_sor *sor, unsigned int lanes)
+{
+	unsigned long timeout;
 	u32 value;
-	int err;
 
-	/* setup lane parameters */
-	value = SOR_LANE_DRIVE_CURRENT_LANE3(0x40) |
-		SOR_LANE_DRIVE_CURRENT_LANE2(0x40) |
-		SOR_LANE_DRIVE_CURRENT_LANE1(0x40) |
-		SOR_LANE_DRIVE_CURRENT_LANE0(0x40);
-	tegra_sor_writel(sor, value, SOR_LANE_DRIVE_CURRENT0);
+	/*
+	 * Clear or set the PD_TXD bit corresponding to each lane, depending
+	 * on whether it is used or not.
+	 */
+	value = tegra_sor_readl(sor, sor->soc->regs->dp_padctl0);
 
-	value = SOR_LANE_PREEMPHASIS_LANE3(0x0f) |
-		SOR_LANE_PREEMPHASIS_LANE2(0x0f) |
-		SOR_LANE_PREEMPHASIS_LANE1(0x0f) |
-		SOR_LANE_PREEMPHASIS_LANE0(0x0f);
-	tegra_sor_writel(sor, value, SOR_LANE_PREEMPHASIS0);
+	if (lanes <= 2)
+		value &= ~(SOR_DP_PADCTL_PD_TXD(sor->soc->lane_map[3]) |
+			   SOR_DP_PADCTL_PD_TXD(sor->soc->lane_map[2]));
+	else
+		value |= SOR_DP_PADCTL_PD_TXD(sor->soc->lane_map[3]) |
+			 SOR_DP_PADCTL_PD_TXD(sor->soc->lane_map[2]);
 
-	value = SOR_LANE_POSTCURSOR_LANE3(0x00) |
-		SOR_LANE_POSTCURSOR_LANE2(0x00) |
-		SOR_LANE_POSTCURSOR_LANE1(0x00) |
-		SOR_LANE_POSTCURSOR_LANE0(0x00);
-	tegra_sor_writel(sor, value, SOR_LANE_POSTCURSOR0);
+	if (lanes <= 1)
+		value &= ~SOR_DP_PADCTL_PD_TXD(sor->soc->lane_map[1]);
+	else
+		value |= SOR_DP_PADCTL_PD_TXD(sor->soc->lane_map[1]);
 
-	/* disable LVDS mode */
-	tegra_sor_writel(sor, 0, SOR_LVDS);
+	if (lanes == 0)
+		value &= ~SOR_DP_PADCTL_PD_TXD(sor->soc->lane_map[0]);
+	else
+		value |= SOR_DP_PADCTL_PD_TXD(sor->soc->lane_map[0]);
+
+	tegra_sor_writel(sor, value, sor->soc->regs->dp_padctl0);
+
+	/* start lane sequencer */
+	value = SOR_LANE_SEQ_CTL_TRIGGER | SOR_LANE_SEQ_CTL_SEQUENCE_DOWN |
+		SOR_LANE_SEQ_CTL_POWER_STATE_UP;
+	tegra_sor_writel(sor, value, SOR_LANE_SEQ_CTL);
+
+	timeout = jiffies + msecs_to_jiffies(250);
 
+	while (time_before(jiffies, timeout)) {
+		value = tegra_sor_readl(sor, SOR_LANE_SEQ_CTL);
+		if ((value & SOR_LANE_SEQ_CTL_TRIGGER) == 0)
+			break;
+
+		usleep_range(250, 1000);
+	}
+
+	if ((value & SOR_LANE_SEQ_CTL_TRIGGER) != 0)
+		return -ETIMEDOUT;
+
+	return 0;
+}
+
+static int tegra_sor_power_down_lanes(struct tegra_sor *sor)
+{
+	unsigned long timeout;
+	u32 value;
+
+	/* power down all lanes */
 	value = tegra_sor_readl(sor, sor->soc->regs->dp_padctl0);
-	value |= SOR_DP_PADCTL_TX_PU_ENABLE;
-	value &= ~SOR_DP_PADCTL_TX_PU_MASK;
-	value |= SOR_DP_PADCTL_TX_PU(2); /* XXX: don't hardcode? */
+	value &= ~(SOR_DP_PADCTL_PD_TXD_3 | SOR_DP_PADCTL_PD_TXD_0 |
+		   SOR_DP_PADCTL_PD_TXD_1 | SOR_DP_PADCTL_PD_TXD_2);
 	tegra_sor_writel(sor, value, sor->soc->regs->dp_padctl0);
 
+	/* start lane sequencer */
+	value = SOR_LANE_SEQ_CTL_TRIGGER | SOR_LANE_SEQ_CTL_SEQUENCE_UP |
+		SOR_LANE_SEQ_CTL_POWER_STATE_DOWN;
+	tegra_sor_writel(sor, value, SOR_LANE_SEQ_CTL);
+
+	timeout = jiffies + msecs_to_jiffies(250);
+
+	while (time_before(jiffies, timeout)) {
+		value = tegra_sor_readl(sor, SOR_LANE_SEQ_CTL);
+		if ((value & SOR_LANE_SEQ_CTL_TRIGGER) == 0)
+			break;
+
+		usleep_range(25, 100);
+	}
+
+	if ((value & SOR_LANE_SEQ_CTL_TRIGGER) != 0)
+		return -ETIMEDOUT;
+
+	return 0;
+}
+
+static void tegra_sor_dp_precharge(struct tegra_sor *sor, unsigned int lanes)
+{
+	u32 value;
+
+	/* pre-charge all used lanes */
 	value = tegra_sor_readl(sor, sor->soc->regs->dp_padctl0);
-	value |= SOR_DP_PADCTL_CM_TXD_3 | SOR_DP_PADCTL_CM_TXD_2 |
-		 SOR_DP_PADCTL_CM_TXD_1 | SOR_DP_PADCTL_CM_TXD_0;
+
+	if (lanes <= 2)
+		value &= ~(SOR_DP_PADCTL_CM_TXD(sor->soc->lane_map[3]) |
+			   SOR_DP_PADCTL_CM_TXD(sor->soc->lane_map[2]));
+	else
+		value |= SOR_DP_PADCTL_CM_TXD(sor->soc->lane_map[3]) |
+			 SOR_DP_PADCTL_CM_TXD(sor->soc->lane_map[2]);
+
+	if (lanes <= 1)
+		value &= ~SOR_DP_PADCTL_CM_TXD(sor->soc->lane_map[1]);
+	else
+		value |= SOR_DP_PADCTL_CM_TXD(sor->soc->lane_map[1]);
+
+	if (lanes == 0)
+		value &= ~SOR_DP_PADCTL_CM_TXD(sor->soc->lane_map[0]);
+	else
+		value |= SOR_DP_PADCTL_CM_TXD(sor->soc->lane_map[0]);
+
 	tegra_sor_writel(sor, value, sor->soc->regs->dp_padctl0);
 
-	usleep_range(10, 100);
+	usleep_range(15, 100);
 
 	value = tegra_sor_readl(sor, sor->soc->regs->dp_padctl0);
 	value &= ~(SOR_DP_PADCTL_CM_TXD_3 | SOR_DP_PADCTL_CM_TXD_2 |
 		   SOR_DP_PADCTL_CM_TXD_1 | SOR_DP_PADCTL_CM_TXD_0);
 	tegra_sor_writel(sor, value, sor->soc->regs->dp_padctl0);
+}
 
-	err = drm_dp_aux_prepare(sor->aux, DP_SET_ANSI_8B10B);
-	if (err < 0)
-		return err;
+static void tegra_sor_dp_term_calibrate(struct tegra_sor *sor)
+{
+	u32 mask = 0x08, adj = 0, value;
+
+	/* enable pad calibration logic */
+	value = tegra_sor_readl(sor, sor->soc->regs->dp_padctl0);
+	value &= ~SOR_DP_PADCTL_PAD_CAL_PD;
+	tegra_sor_writel(sor, value, sor->soc->regs->dp_padctl0);
+
+	value = tegra_sor_readl(sor, sor->soc->regs->pll1);
+	value |= SOR_PLL1_TMDS_TERM;
+	tegra_sor_writel(sor, value, sor->soc->regs->pll1);
+
+	while (mask) {
+		adj |= mask;
+
+		value = tegra_sor_readl(sor, sor->soc->regs->pll1);
+		value &= ~SOR_PLL1_TMDS_TERMADJ_MASK;
+		value |= SOR_PLL1_TMDS_TERMADJ(adj);
+		tegra_sor_writel(sor, value, sor->soc->regs->pll1);
 
-	for (i = 0, value = 0; i < link->num_lanes; i++) {
-		unsigned long lane = SOR_DP_TPG_CHANNEL_CODING |
-				     SOR_DP_TPG_SCRAMBLER_NONE |
-				     SOR_DP_TPG_PATTERN_TRAIN1;
-		value = (value << 8) | lane;
+		usleep_range(100, 200);
+
+		value = tegra_sor_readl(sor, sor->soc->regs->pll1);
+		if (value & SOR_PLL1_TERM_COMPOUT)
+			adj &= ~mask;
+
+		mask >>= 1;
 	}
 
-	tegra_sor_writel(sor, value, SOR_DP_TPG);
+	value = tegra_sor_readl(sor, sor->soc->regs->pll1);
+	value &= ~SOR_PLL1_TMDS_TERMADJ_MASK;
+	value |= SOR_PLL1_TMDS_TERMADJ(adj);
+	tegra_sor_writel(sor, value, sor->soc->regs->pll1);
 
-	pattern = DP_TRAINING_PATTERN_1;
+	/* disable pad calibration logic */
+	value = tegra_sor_readl(sor, sor->soc->regs->dp_padctl0);
+	value |= SOR_DP_PADCTL_PAD_CAL_PD;
+	tegra_sor_writel(sor, value, sor->soc->regs->dp_padctl0);
+}
 
-	err = drm_dp_aux_train(sor->aux, link, pattern);
-	if (err < 0)
-		return err;
+static int tegra_sor_dp_link_apply_training(struct drm_dp_link *link)
+{
+	struct tegra_sor *sor = container_of(link, struct tegra_sor, link);
+	u32 voltage_swing = 0, pre_emphasis = 0, post_cursor = 0;
+	const struct tegra_sor_soc *soc = sor->soc;
+	u32 pattern = 0, tx_pu = 0, value;
+	unsigned int i;
 
-	value = tegra_sor_readl(sor, SOR_DP_SPARE0);
-	value |= SOR_DP_SPARE_SEQ_ENABLE;
-	value &= ~SOR_DP_SPARE_PANEL_INTERNAL;
-	value |= SOR_DP_SPARE_MACRO_SOR_CLK;
-	tegra_sor_writel(sor, value, SOR_DP_SPARE0);
+	for (value = 0, i = 0; i < link->lanes; i++) {
+		u8 vs = link->train.request.voltage_swing[i];
+		u8 pe = link->train.request.pre_emphasis[i];
+		u8 pc = link->train.request.post_cursor[i];
+		u8 shift = sor->soc->lane_map[i] << 3;
+
+		voltage_swing |= soc->voltage_swing[pc][vs][pe] << shift;
+		pre_emphasis |= soc->pre_emphasis[pc][vs][pe] << shift;
+		post_cursor |= soc->post_cursor[pc][vs][pe] << shift;
+
+		if (sor->soc->tx_pu[pc][vs][pe] > tx_pu)
+			tx_pu = sor->soc->tx_pu[pc][vs][pe];
+
+		switch (link->train.pattern) {
+		case DP_TRAINING_PATTERN_DISABLE:
+			value = SOR_DP_TPG_SCRAMBLER_GALIOS |
+				SOR_DP_TPG_PATTERN_NONE;
+			break;
+
+		case DP_TRAINING_PATTERN_1:
+			value = SOR_DP_TPG_SCRAMBLER_NONE |
+				SOR_DP_TPG_PATTERN_TRAIN1;
+			break;
+
+		case DP_TRAINING_PATTERN_2:
+			value = SOR_DP_TPG_SCRAMBLER_NONE |
+				SOR_DP_TPG_PATTERN_TRAIN2;
+			break;
+
+		case DP_TRAINING_PATTERN_3:
+			value = SOR_DP_TPG_SCRAMBLER_NONE |
+				SOR_DP_TPG_PATTERN_TRAIN3;
+			break;
+
+		default:
+			return -EINVAL;
+		}
+
+		if (link->caps.channel_coding)
+			value |= SOR_DP_TPG_CHANNEL_CODING;
 
-	for (i = 0, value = 0; i < link->num_lanes; i++) {
-		unsigned long lane = SOR_DP_TPG_CHANNEL_CODING |
-				     SOR_DP_TPG_SCRAMBLER_NONE |
-				     SOR_DP_TPG_PATTERN_TRAIN2;
-		value = (value << 8) | lane;
+		pattern = pattern << 8 | value;
 	}
 
-	tegra_sor_writel(sor, value, SOR_DP_TPG);
+	tegra_sor_writel(sor, voltage_swing, SOR_LANE_DRIVE_CURRENT0);
+	tegra_sor_writel(sor, pre_emphasis, SOR_LANE_PREEMPHASIS0);
 
-	pattern = DP_LINK_SCRAMBLING_DISABLE | DP_TRAINING_PATTERN_2;
+	if (link->caps.tps3_supported)
+		tegra_sor_writel(sor, post_cursor, SOR_LANE_POSTCURSOR0);
 
-	err = drm_dp_aux_train(sor->aux, link, pattern);
-	if (err < 0)
-		return err;
+	tegra_sor_writel(sor, pattern, SOR_DP_TPG);
 
-	for (i = 0, value = 0; i < link->num_lanes; i++) {
-		unsigned long lane = SOR_DP_TPG_CHANNEL_CODING |
-				     SOR_DP_TPG_SCRAMBLER_GALIOS |
-				     SOR_DP_TPG_PATTERN_NONE;
-		value = (value << 8) | lane;
+	value = tegra_sor_readl(sor, sor->soc->regs->dp_padctl0);
+	value &= ~SOR_DP_PADCTL_TX_PU_MASK;
+	value |= SOR_DP_PADCTL_TX_PU_ENABLE;
+	value |= SOR_DP_PADCTL_TX_PU(tx_pu);
+	tegra_sor_writel(sor, value, sor->soc->regs->dp_padctl0);
+
+	usleep_range(20, 100);
+
+	return 0;
+}
+
+static int tegra_sor_dp_link_configure(struct drm_dp_link *link)
+{
+	struct tegra_sor *sor = container_of(link, struct tegra_sor, link);
+	unsigned int rate, lanes;
+	u32 value;
+	int err;
+
+	rate = drm_dp_link_rate_to_bw_code(link->rate);
+	lanes = link->lanes;
+
+	/* configure link speed and lane count */
+	value = tegra_sor_readl(sor, SOR_CLK_CNTRL);
+	value &= ~SOR_CLK_CNTRL_DP_LINK_SPEED_MASK;
+	value |= SOR_CLK_CNTRL_DP_LINK_SPEED(rate);
+	tegra_sor_writel(sor, value, SOR_CLK_CNTRL);
+
+	value = tegra_sor_readl(sor, SOR_DP_LINKCTL0);
+	value &= ~SOR_DP_LINKCTL_LANE_COUNT_MASK;
+	value |= SOR_DP_LINKCTL_LANE_COUNT(lanes);
+
+	if (link->caps.enhanced_framing)
+		value |= SOR_DP_LINKCTL_ENHANCED_FRAME;
+
+	tegra_sor_writel(sor, value, SOR_DP_LINKCTL0);
+
+	usleep_range(400, 1000);
+
+	/* configure load pulse position adjustment */
+	value = tegra_sor_readl(sor, sor->soc->regs->pll1);
+	value &= ~SOR_PLL1_LOADADJ_MASK;
+
+	switch (rate) {
+	case DP_LINK_BW_1_62:
+		value |= SOR_PLL1_LOADADJ(0x3);
+		break;
+
+	case DP_LINK_BW_2_7:
+		value |= SOR_PLL1_LOADADJ(0x4);
+		break;
+
+	case DP_LINK_BW_5_4:
+		value |= SOR_PLL1_LOADADJ(0x6);
+		break;
 	}
 
-	tegra_sor_writel(sor, value, SOR_DP_TPG);
+	tegra_sor_writel(sor, value, sor->soc->regs->pll1);
 
-	pattern = DP_TRAINING_PATTERN_DISABLE;
+	/* use alternate scrambler reset for eDP */
+	value = tegra_sor_readl(sor, SOR_DP_SPARE0);
 
-	err = drm_dp_aux_train(sor->aux, link, pattern);
-	if (err < 0)
+	if (link->edp == 0)
+		value &= ~SOR_DP_SPARE_PANEL_INTERNAL;
+	else
+		value |= SOR_DP_SPARE_PANEL_INTERNAL;
+
+	tegra_sor_writel(sor, value, SOR_DP_SPARE0);
+
+	err = tegra_sor_power_down_lanes(sor);
+	if (err < 0) {
+		dev_err(sor->dev, "failed to power down lanes: %d\n", err);
+		return err;
+	}
+
+	/* power up and pre-charge lanes */
+	err = tegra_sor_power_up_lanes(sor, lanes);
+	if (err < 0) {
+		dev_err(sor->dev, "failed to power up %u lane%s: %d\n",
+			lanes, (lanes != 1) ? "s" : "", err);
 		return err;
+	}
+
+	tegra_sor_dp_precharge(sor, lanes);
 
 	return 0;
 }
 
+static const struct drm_dp_link_ops tegra_sor_dp_link_ops = {
+	.apply_training = tegra_sor_dp_link_apply_training,
+	.configure = tegra_sor_dp_link_configure,
+};
+
 static void tegra_sor_super_update(struct tegra_sor *sor)
 {
 	tegra_sor_writel(sor, 0, SOR_SUPER_STATE0);
@@ -912,11 +1160,11 @@ static int tegra_sor_compute_config(struct tegra_sor *sor,
 	u32 num_syms_per_line;
 	unsigned int i;
 
-	if (!link_rate || !link->num_lanes || !pclk || !config->bits_per_pixel)
+	if (!link_rate || !link->lanes || !pclk || !config->bits_per_pixel)
 		return -EINVAL;
 
-	output = link_rate * 8 * link->num_lanes;
 	input = pclk * config->bits_per_pixel;
+	output = link_rate * 8 * link->lanes;
 
 	if (input >= output)
 		return -ERANGE;
@@ -959,7 +1207,7 @@ static int tegra_sor_compute_config(struct tegra_sor *sor,
 	watermark = div_u64(watermark + params.error, f);
 	config->watermark = watermark + (config->bits_per_pixel / 8) + 2;
 	num_syms_per_line = (mode->hdisplay * config->bits_per_pixel) *
-			    (link->num_lanes * 8);
+			    (link->lanes * 8);
 
 	if (config->watermark > 30) {
 		config->watermark = 30;
@@ -976,15 +1224,15 @@ static int tegra_sor_compute_config(struct tegra_sor *sor,
 	num = ((mode->htotal - mode->hdisplay) - 7) * link_rate;
 	config->hblank_symbols = div_u64(num, pclk);
 
-	if (link->capabilities & DP_LINK_CAP_ENHANCED_FRAMING)
+	if (link->caps.enhanced_framing)
 		config->hblank_symbols -= 3;
 
-	config->hblank_symbols -= 12 / link->num_lanes;
+	config->hblank_symbols -= 12 / link->lanes;
 
 	/* compute the number of symbols per vertical blanking interval */
 	num = (mode->hdisplay - 25) * link_rate;
 	config->vblank_symbols = div_u64(num, pclk);
-	config->vblank_symbols -= 36 / link->num_lanes + 4;
+	config->vblank_symbols -= 36 / link->lanes + 4;
 
 	dev_dbg(sor->dev, "blank symbols: H:%u V:%u\n", config->hblank_symbols,
 		config->vblank_symbols);
@@ -1200,29 +1448,6 @@ static int tegra_sor_power_down(struct tegra_sor *sor)
 		return err;
 	}
 
-	value = tegra_sor_readl(sor, sor->soc->regs->dp_padctl0);
-	value &= ~(SOR_DP_PADCTL_PD_TXD_3 | SOR_DP_PADCTL_PD_TXD_0 |
-		   SOR_DP_PADCTL_PD_TXD_1 | SOR_DP_PADCTL_PD_TXD_2);
-	tegra_sor_writel(sor, value, sor->soc->regs->dp_padctl0);
-
-	/* stop lane sequencer */
-	value = SOR_LANE_SEQ_CTL_TRIGGER | SOR_LANE_SEQ_CTL_SEQUENCE_UP |
-		SOR_LANE_SEQ_CTL_POWER_STATE_DOWN;
-	tegra_sor_writel(sor, value, SOR_LANE_SEQ_CTL);
-
-	timeout = jiffies + msecs_to_jiffies(250);
-
-	while (time_before(jiffies, timeout)) {
-		value = tegra_sor_readl(sor, SOR_LANE_SEQ_CTL);
-		if ((value & SOR_LANE_SEQ_CTL_TRIGGER) == 0)
-			break;
-
-		usleep_range(25, 100);
-	}
-
-	if ((value & SOR_LANE_SEQ_CTL_TRIGGER) != 0)
-		return -ETIMEDOUT;
-
 	value = tegra_sor_readl(sor, sor->soc->regs->pll2);
 	value |= SOR_PLL2_PORT_POWERDOWN;
 	tegra_sor_writel(sor, value, sor->soc->regs->pll2);
@@ -1584,403 +1809,6 @@ static const struct drm_encoder_funcs tegra_sor_encoder_funcs = {
 	.destroy = tegra_output_encoder_destroy,
 };
 
-static void tegra_sor_edp_disable(struct drm_encoder *encoder)
-{
-	struct tegra_output *output = encoder_to_output(encoder);
-	struct tegra_dc *dc = to_tegra_dc(encoder->crtc);
-	struct tegra_sor *sor = to_sor(output);
-	u32 value;
-	int err;
-
-	if (output->panel)
-		drm_panel_disable(output->panel);
-
-	err = tegra_sor_detach(sor);
-	if (err < 0)
-		dev_err(sor->dev, "failed to detach SOR: %d\n", err);
-
-	tegra_sor_writel(sor, 0, SOR_STATE1);
-	tegra_sor_update(sor);
-
-	/*
-	 * The following accesses registers of the display controller, so make
-	 * sure it's only executed when the output is attached to one.
-	 */
-	if (dc) {
-		value = tegra_dc_readl(dc, DC_DISP_DISP_WIN_OPTIONS);
-		value &= ~SOR_ENABLE(0);
-		tegra_dc_writel(dc, value, DC_DISP_DISP_WIN_OPTIONS);
-
-		tegra_dc_commit(dc);
-	}
-
-	err = tegra_sor_power_down(sor);
-	if (err < 0)
-		dev_err(sor->dev, "failed to power down SOR: %d\n", err);
-
-	if (sor->aux) {
-		err = drm_dp_aux_disable(sor->aux);
-		if (err < 0)
-			dev_err(sor->dev, "failed to disable DP: %d\n", err);
-	}
-
-	err = tegra_io_pad_power_disable(sor->pad);
-	if (err < 0)
-		dev_err(sor->dev, "failed to power off I/O pad: %d\n", err);
-
-	if (output->panel)
-		drm_panel_unprepare(output->panel);
-
-	pm_runtime_put(sor->dev);
-}
-
-#if 0
-static int calc_h_ref_to_sync(const struct drm_display_mode *mode,
-			      unsigned int *value)
-{
-	unsigned int hfp, hsw, hbp, a = 0, b;
-
-	hfp = mode->hsync_start - mode->hdisplay;
-	hsw = mode->hsync_end - mode->hsync_start;
-	hbp = mode->htotal - mode->hsync_end;
-
-	pr_info("hfp: %u, hsw: %u, hbp: %u\n", hfp, hsw, hbp);
-
-	b = hfp - 1;
-
-	pr_info("a: %u, b: %u\n", a, b);
-	pr_info("a + hsw + hbp = %u\n", a + hsw + hbp);
-
-	if (a + hsw + hbp <= 11) {
-		a = 1 + 11 - hsw - hbp;
-		pr_info("a: %u\n", a);
-	}
-
-	if (a > b)
-		return -EINVAL;
-
-	if (hsw < 1)
-		return -EINVAL;
-
-	if (mode->hdisplay < 16)
-		return -EINVAL;
-
-	if (value) {
-		if (b > a && a % 2)
-			*value = a + 1;
-		else
-			*value = a;
-	}
-
-	return 0;
-}
-#endif
-
-static void tegra_sor_edp_enable(struct drm_encoder *encoder)
-{
-	struct drm_display_mode *mode = &encoder->crtc->state->adjusted_mode;
-	struct tegra_output *output = encoder_to_output(encoder);
-	struct tegra_dc *dc = to_tegra_dc(encoder->crtc);
-	struct tegra_sor *sor = to_sor(output);
-	struct tegra_sor_config config;
-	struct tegra_sor_state *state;
-	struct drm_dp_link link;
-	u8 rate, lanes;
-	unsigned int i;
-	int err = 0;
-	u32 value;
-
-	state = to_sor_state(output->connector.state);
-
-	pm_runtime_get_sync(sor->dev);
-
-	if (output->panel)
-		drm_panel_prepare(output->panel);
-
-	err = drm_dp_aux_enable(sor->aux);
-	if (err < 0)
-		dev_err(sor->dev, "failed to enable DP: %d\n", err);
-
-	err = drm_dp_link_probe(sor->aux, &link);
-	if (err < 0) {
-		dev_err(sor->dev, "failed to probe eDP link: %d\n", err);
-		return;
-	}
-
-	/* switch to safe parent clock */
-	err = tegra_sor_set_parent_clock(sor, sor->clk_safe);
-	if (err < 0)
-		dev_err(sor->dev, "failed to set safe parent clock: %d\n", err);
-
-	memset(&config, 0, sizeof(config));
-	config.bits_per_pixel = state->bpc * 3;
-
-	err = tegra_sor_compute_config(sor, mode, &config, &link);
-	if (err < 0)
-		dev_err(sor->dev, "failed to compute configuration: %d\n", err);
-
-	value = tegra_sor_readl(sor, SOR_CLK_CNTRL);
-	value &= ~SOR_CLK_CNTRL_DP_CLK_SEL_MASK;
-	value |= SOR_CLK_CNTRL_DP_CLK_SEL_SINGLE_DPCLK;
-	tegra_sor_writel(sor, value, SOR_CLK_CNTRL);
-
-	value = tegra_sor_readl(sor, sor->soc->regs->pll2);
-	value &= ~SOR_PLL2_BANDGAP_POWERDOWN;
-	tegra_sor_writel(sor, value, sor->soc->regs->pll2);
-	usleep_range(20, 100);
-
-	value = tegra_sor_readl(sor, sor->soc->regs->pll3);
-	value |= SOR_PLL3_PLL_VDD_MODE_3V3;
-	tegra_sor_writel(sor, value, sor->soc->regs->pll3);
-
-	value = SOR_PLL0_ICHPMP(0xf) | SOR_PLL0_VCOCAP_RST |
-		SOR_PLL0_PLLREG_LEVEL_V45 | SOR_PLL0_RESISTOR_EXT;
-	tegra_sor_writel(sor, value, sor->soc->regs->pll0);
-
-	value = tegra_sor_readl(sor, sor->soc->regs->pll2);
-	value |= SOR_PLL2_SEQ_PLLCAPPD;
-	value &= ~SOR_PLL2_SEQ_PLLCAPPD_ENFORCE;
-	value |= SOR_PLL2_LVDS_ENABLE;
-	tegra_sor_writel(sor, value, sor->soc->regs->pll2);
-
-	value = SOR_PLL1_TERM_COMPOUT | SOR_PLL1_TMDS_TERM;
-	tegra_sor_writel(sor, value, sor->soc->regs->pll1);
-
-	while (true) {
-		value = tegra_sor_readl(sor, sor->soc->regs->pll2);
-		if ((value & SOR_PLL2_SEQ_PLLCAPPD_ENFORCE) == 0)
-			break;
-
-		usleep_range(250, 1000);
-	}
-
-	value = tegra_sor_readl(sor, sor->soc->regs->pll2);
-	value &= ~SOR_PLL2_POWERDOWN_OVERRIDE;
-	value &= ~SOR_PLL2_PORT_POWERDOWN;
-	tegra_sor_writel(sor, value, sor->soc->regs->pll2);
-
-	/*
-	 * power up
-	 */
-
-	/* set safe link bandwidth (1.62 Gbps) */
-	value = tegra_sor_readl(sor, SOR_CLK_CNTRL);
-	value &= ~SOR_CLK_CNTRL_DP_LINK_SPEED_MASK;
-	value |= SOR_CLK_CNTRL_DP_LINK_SPEED_G1_62;
-	tegra_sor_writel(sor, value, SOR_CLK_CNTRL);
-
-	/* step 1 */
-	value = tegra_sor_readl(sor, sor->soc->regs->pll2);
-	value |= SOR_PLL2_SEQ_PLLCAPPD_ENFORCE | SOR_PLL2_PORT_POWERDOWN |
-		 SOR_PLL2_BANDGAP_POWERDOWN;
-	tegra_sor_writel(sor, value, sor->soc->regs->pll2);
-
-	value = tegra_sor_readl(sor, sor->soc->regs->pll0);
-	value |= SOR_PLL0_VCOPD | SOR_PLL0_PWR;
-	tegra_sor_writel(sor, value, sor->soc->regs->pll0);
-
-	value = tegra_sor_readl(sor, sor->soc->regs->dp_padctl0);
-	value &= ~SOR_DP_PADCTL_PAD_CAL_PD;
-	tegra_sor_writel(sor, value, sor->soc->regs->dp_padctl0);
-
-	/* step 2 */
-	err = tegra_io_pad_power_enable(sor->pad);
-	if (err < 0)
-		dev_err(sor->dev, "failed to power on I/O pad: %d\n", err);
-
-	usleep_range(5, 100);
-
-	/* step 3 */
-	value = tegra_sor_readl(sor, sor->soc->regs->pll2);
-	value &= ~SOR_PLL2_BANDGAP_POWERDOWN;
-	tegra_sor_writel(sor, value, sor->soc->regs->pll2);
-
-	usleep_range(20, 100);
-
-	/* step 4 */
-	value = tegra_sor_readl(sor, sor->soc->regs->pll0);
-	value &= ~SOR_PLL0_VCOPD;
-	value &= ~SOR_PLL0_PWR;
-	tegra_sor_writel(sor, value, sor->soc->regs->pll0);
-
-	value = tegra_sor_readl(sor, sor->soc->regs->pll2);
-	value &= ~SOR_PLL2_SEQ_PLLCAPPD_ENFORCE;
-	tegra_sor_writel(sor, value, sor->soc->regs->pll2);
-
-	usleep_range(200, 1000);
-
-	/* step 5 */
-	value = tegra_sor_readl(sor, sor->soc->regs->pll2);
-	value &= ~SOR_PLL2_PORT_POWERDOWN;
-	tegra_sor_writel(sor, value, sor->soc->regs->pll2);
-
-	/* XXX not in TRM */
-	for (value = 0, i = 0; i < 5; i++)
-		value |= SOR_XBAR_CTRL_LINK0_XSEL(i, sor->xbar_cfg[i]) |
-			 SOR_XBAR_CTRL_LINK1_XSEL(i, i);
-
-	tegra_sor_writel(sor, 0x00000000, SOR_XBAR_POL);
-	tegra_sor_writel(sor, value, SOR_XBAR_CTRL);
-
-	/* switch to DP parent clock */
-	err = tegra_sor_set_parent_clock(sor, sor->clk_dp);
-	if (err < 0)
-		dev_err(sor->dev, "failed to set parent clock: %d\n", err);
-
-	/* power DP lanes */
-	value = tegra_sor_readl(sor, sor->soc->regs->dp_padctl0);
-
-	if (link.num_lanes <= 2)
-		value &= ~(SOR_DP_PADCTL_PD_TXD_3 | SOR_DP_PADCTL_PD_TXD_2);
-	else
-		value |= SOR_DP_PADCTL_PD_TXD_3 | SOR_DP_PADCTL_PD_TXD_2;
-
-	if (link.num_lanes <= 1)
-		value &= ~SOR_DP_PADCTL_PD_TXD_1;
-	else
-		value |= SOR_DP_PADCTL_PD_TXD_1;
-
-	if (link.num_lanes == 0)
-		value &= ~SOR_DP_PADCTL_PD_TXD_0;
-	else
-		value |= SOR_DP_PADCTL_PD_TXD_0;
-
-	tegra_sor_writel(sor, value, sor->soc->regs->dp_padctl0);
-
-	value = tegra_sor_readl(sor, SOR_DP_LINKCTL0);
-	value &= ~SOR_DP_LINKCTL_LANE_COUNT_MASK;
-	value |= SOR_DP_LINKCTL_LANE_COUNT(link.num_lanes);
-	tegra_sor_writel(sor, value, SOR_DP_LINKCTL0);
-
-	/* start lane sequencer */
-	value = SOR_LANE_SEQ_CTL_TRIGGER | SOR_LANE_SEQ_CTL_SEQUENCE_DOWN |
-		SOR_LANE_SEQ_CTL_POWER_STATE_UP;
-	tegra_sor_writel(sor, value, SOR_LANE_SEQ_CTL);
-
-	while (true) {
-		value = tegra_sor_readl(sor, SOR_LANE_SEQ_CTL);
-		if ((value & SOR_LANE_SEQ_CTL_TRIGGER) == 0)
-			break;
-
-		usleep_range(250, 1000);
-	}
-
-	/* set link bandwidth */
-	value = tegra_sor_readl(sor, SOR_CLK_CNTRL);
-	value &= ~SOR_CLK_CNTRL_DP_LINK_SPEED_MASK;
-	value |= drm_dp_link_rate_to_bw_code(link.rate) << 2;
-	tegra_sor_writel(sor, value, SOR_CLK_CNTRL);
-
-	tegra_sor_apply_config(sor, &config);
-
-	/* enable link */
-	value = tegra_sor_readl(sor, SOR_DP_LINKCTL0);
-	value |= SOR_DP_LINKCTL_ENABLE;
-	value |= SOR_DP_LINKCTL_ENHANCED_FRAME;
-	tegra_sor_writel(sor, value, SOR_DP_LINKCTL0);
-
-	for (i = 0, value = 0; i < 4; i++) {
-		unsigned long lane = SOR_DP_TPG_CHANNEL_CODING |
-				     SOR_DP_TPG_SCRAMBLER_GALIOS |
-				     SOR_DP_TPG_PATTERN_NONE;
-		value = (value << 8) | lane;
-	}
-
-	tegra_sor_writel(sor, value, SOR_DP_TPG);
-
-	/* enable pad calibration logic */
-	value = tegra_sor_readl(sor, sor->soc->regs->dp_padctl0);
-	value |= SOR_DP_PADCTL_PAD_CAL_PD;
-	tegra_sor_writel(sor, value, sor->soc->regs->dp_padctl0);
-
-	err = drm_dp_link_probe(sor->aux, &link);
-	if (err < 0)
-		dev_err(sor->dev, "failed to probe eDP link: %d\n", err);
-
-	err = drm_dp_link_power_up(sor->aux, &link);
-	if (err < 0)
-		dev_err(sor->dev, "failed to power up eDP link: %d\n", err);
-
-	err = drm_dp_link_configure(sor->aux, &link);
-	if (err < 0)
-		dev_err(sor->dev, "failed to configure eDP link: %d\n", err);
-
-	rate = drm_dp_link_rate_to_bw_code(link.rate);
-	lanes = link.num_lanes;
-
-	value = tegra_sor_readl(sor, SOR_CLK_CNTRL);
-	value &= ~SOR_CLK_CNTRL_DP_LINK_SPEED_MASK;
-	value |= SOR_CLK_CNTRL_DP_LINK_SPEED(rate);
-	tegra_sor_writel(sor, value, SOR_CLK_CNTRL);
-
-	value = tegra_sor_readl(sor, SOR_DP_LINKCTL0);
-	value &= ~SOR_DP_LINKCTL_LANE_COUNT_MASK;
-	value |= SOR_DP_LINKCTL_LANE_COUNT(lanes);
-
-	if (link.capabilities & DP_LINK_CAP_ENHANCED_FRAMING)
-		value |= SOR_DP_LINKCTL_ENHANCED_FRAME;
-
-	tegra_sor_writel(sor, value, SOR_DP_LINKCTL0);
-
-	/* disable training pattern generator */
-
-	for (i = 0; i < link.num_lanes; i++) {
-		unsigned long lane = SOR_DP_TPG_CHANNEL_CODING |
-				     SOR_DP_TPG_SCRAMBLER_GALIOS |
-				     SOR_DP_TPG_PATTERN_NONE;
-		value = (value << 8) | lane;
-	}
-
-	tegra_sor_writel(sor, value, SOR_DP_TPG);
-
-	err = tegra_sor_dp_train_fast(sor, &link);
-	if (err < 0)
-		dev_err(sor->dev, "DP fast link training failed: %d\n", err);
-
-	dev_dbg(sor->dev, "fast link training succeeded\n");
-
-	err = tegra_sor_power_up(sor, 250);
-	if (err < 0)
-		dev_err(sor->dev, "failed to power up SOR: %d\n", err);
-
-	/* CSTM (LVDS, link A/B, upper) */
-	value = SOR_CSTM_LVDS | SOR_CSTM_LINK_ACT_A | SOR_CSTM_LINK_ACT_B |
-		SOR_CSTM_UPPER;
-	tegra_sor_writel(sor, value, SOR_CSTM);
-
-	/* use DP-A protocol */
-	value = tegra_sor_readl(sor, SOR_STATE1);
-	value &= ~SOR_STATE_ASY_PROTOCOL_MASK;
-	value |= SOR_STATE_ASY_PROTOCOL_DP_A;
-	tegra_sor_writel(sor, value, SOR_STATE1);
-
-	tegra_sor_mode_set(sor, mode, state);
-
-	/* PWM setup */
-	err = tegra_sor_setup_pwm(sor, 250);
-	if (err < 0)
-		dev_err(sor->dev, "failed to setup PWM: %d\n", err);
-
-	tegra_sor_update(sor);
-
-	value = tegra_dc_readl(dc, DC_DISP_DISP_WIN_OPTIONS);
-	value |= SOR_ENABLE(0);
-	tegra_dc_writel(dc, value, DC_DISP_DISP_WIN_OPTIONS);
-
-	tegra_dc_commit(dc);
-
-	err = tegra_sor_attach(sor);
-	if (err < 0)
-		dev_err(sor->dev, "failed to attach SOR: %d\n", err);
-
-	err = tegra_sor_wakeup(sor);
-	if (err < 0)
-		dev_err(sor->dev, "failed to enable DC: %d\n", err);
-
-	if (output->panel)
-		drm_panel_enable(output->panel);
-}
-
 static int
 tegra_sor_encoder_atomic_check(struct drm_encoder *encoder,
 			       struct drm_crtc_state *crtc_state,
@@ -2030,12 +1858,6 @@ tegra_sor_encoder_atomic_check(struct drm_encoder *encoder,
 	return 0;
 }
 
-static const struct drm_encoder_helper_funcs tegra_sor_edp_helpers = {
-	.disable = tegra_sor_edp_disable,
-	.enable = tegra_sor_edp_enable,
-	.atomic_check = tegra_sor_encoder_atomic_check,
-};
-
 static inline u32 tegra_sor_hdmi_subpack(const u8 *ptr, size_t size)
 {
 	u32 value = 0;
@@ -2160,6 +1982,15 @@ static void tegra_sor_audio_prepare(struct tegra_sor *sor)
 {
 	u32 value;
 
+	/*
+	 * Enable and unmask the HDA codec SCRATCH0 register interrupt. This
+	 * is used for interoperability between the HDA codec driver and the
+	 * HDMI/DP driver.
+	 */
+	value = SOR_INT_CODEC_SCRATCH1 | SOR_INT_CODEC_SCRATCH0;
+	tegra_sor_writel(sor, value, SOR_INT_ENABLE);
+	tegra_sor_writel(sor, value, SOR_INT_MASK);
+
 	tegra_sor_write_eld(sor);
 
 	value = SOR_AUDIO_HDA_PRESENSE_ELDV | SOR_AUDIO_HDA_PRESENSE_PD;
@@ -2169,6 +2000,32 @@ static void tegra_sor_audio_prepare(struct tegra_sor *sor)
 static void tegra_sor_audio_unprepare(struct tegra_sor *sor)
 {
 	tegra_sor_writel(sor, 0, SOR_AUDIO_HDA_PRESENSE);
+	tegra_sor_writel(sor, 0, SOR_INT_MASK);
+	tegra_sor_writel(sor, 0, SOR_INT_ENABLE);
+}
+
+static void tegra_sor_audio_enable(struct tegra_sor *sor)
+{
+	u32 value;
+
+	value = tegra_sor_readl(sor, SOR_AUDIO_CNTRL);
+
+	/* select HDA audio input */
+	value &= ~SOR_AUDIO_CNTRL_SOURCE_SELECT(SOURCE_SELECT_MASK);
+	value |= SOR_AUDIO_CNTRL_SOURCE_SELECT(SOURCE_SELECT_HDA);
+
+	/* inject null samples */
+	if (sor->format.channels != 2)
+		value &= ~SOR_AUDIO_CNTRL_INJECT_NULLSMPL;
+	else
+		value |= SOR_AUDIO_CNTRL_INJECT_NULLSMPL;
+
+	value |= SOR_AUDIO_CNTRL_AFIFO_FLUSH;
+
+	tegra_sor_writel(sor, value, SOR_AUDIO_CNTRL);
+
+	/* enable advertising HBR capability */
+	tegra_sor_writel(sor, SOR_AUDIO_SPARE_HBR_ENABLE, SOR_AUDIO_SPARE);
 }
 
 static int tegra_sor_hdmi_enable_audio_infoframe(struct tegra_sor *sor)
@@ -2206,24 +2063,7 @@ static void tegra_sor_hdmi_audio_enable(struct tegra_sor *sor)
 {
 	u32 value;
 
-	value = tegra_sor_readl(sor, SOR_AUDIO_CNTRL);
-
-	/* select HDA audio input */
-	value &= ~SOR_AUDIO_CNTRL_SOURCE_SELECT(SOURCE_SELECT_MASK);
-	value |= SOR_AUDIO_CNTRL_SOURCE_SELECT(SOURCE_SELECT_HDA);
-
-	/* inject null samples */
-	if (sor->format.channels != 2)
-		value &= ~SOR_AUDIO_CNTRL_INJECT_NULLSMPL;
-	else
-		value |= SOR_AUDIO_CNTRL_INJECT_NULLSMPL;
-
-	value |= SOR_AUDIO_CNTRL_AFIFO_FLUSH;
-
-	tegra_sor_writel(sor, value, SOR_AUDIO_CNTRL);
-
-	/* enable advertising HBR capability */
-	tegra_sor_writel(sor, SOR_AUDIO_SPARE_HBR_ENABLE, SOR_AUDIO_SPARE);
+	tegra_sor_audio_enable(sor);
 
 	tegra_sor_writel(sor, 0, SOR_HDMI_ACR_CTRL);
 
@@ -2399,9 +2239,9 @@ static void tegra_sor_hdmi_disable(struct drm_encoder *encoder)
 	value = tegra_dc_readl(dc, DC_DISP_DISP_WIN_OPTIONS);
 
 	if (!sor->soc->has_nvdisplay)
-		value &= ~(SOR1_TIMING_CYA | SOR_ENABLE(1));
-	else
-		value &= ~SOR_ENABLE(sor->index);
+		value &= ~SOR1_TIMING_CYA;
+
+	value &= ~SOR_ENABLE(sor->index);
 
 	tegra_dc_writel(dc, value, DC_DISP_DISP_WIN_OPTIONS);
 
@@ -2559,16 +2399,34 @@ static void tegra_sor_hdmi_enable(struct drm_encoder *encoder)
 	tegra_sor_writel(sor, 0x00000000, SOR_XBAR_POL);
 	tegra_sor_writel(sor, value, SOR_XBAR_CTRL);
 
-	/* switch to parent clock */
-	err = clk_set_parent(sor->clk, sor->clk_parent);
+	/*
+	 * Switch the pad clock to the DP clock. Note that we cannot actually
+	 * do this because Tegra186 and later don't support clk_set_parent()
+	 * on the sorX_pad_clkout clocks. We already do the equivalent above
+	 * using the DP_CLK_SEL mux of the SOR_CLK_CNTRL register.
+	 */
+#if 0
+	err = clk_set_parent(sor->clk_pad, sor->clk_dp);
 	if (err < 0) {
-		dev_err(sor->dev, "failed to set parent clock: %d\n", err);
+		dev_err(sor->dev, "failed to select pad parent clock: %d\n",
+			err);
 		return;
 	}
+#endif
 
+	/* switch the SOR clock to the pad clock */
 	err = tegra_sor_set_parent_clock(sor, sor->clk_pad);
 	if (err < 0) {
-		dev_err(sor->dev, "failed to set pad clock: %d\n", err);
+		dev_err(sor->dev, "failed to select SOR parent clock: %d\n",
+			err);
+		return;
+	}
+
+	/* switch the output clock to the parent pixel clock */
+	err = clk_set_parent(sor->clk, sor->clk_parent);
+	if (err < 0) {
+		dev_err(sor->dev, "failed to select output parent clock: %d\n",
+			err);
 		return;
 	}
 
@@ -2774,9 +2632,9 @@ static void tegra_sor_hdmi_enable(struct drm_encoder *encoder)
 	value = tegra_dc_readl(dc, DC_DISP_DISP_WIN_OPTIONS);
 
 	if (!sor->soc->has_nvdisplay)
-		value |= SOR_ENABLE(1) | SOR1_TIMING_CYA;
-	else
-		value |= SOR_ENABLE(sor->index);
+		value |= SOR1_TIMING_CYA;
+
+	value |= SOR_ENABLE(sor->index);
 
 	tegra_dc_writel(dc, value, DC_DISP_DISP_WIN_OPTIONS);
 
@@ -2803,6 +2661,396 @@ static const struct drm_encoder_helper_funcs tegra_sor_hdmi_helpers = {
 	.atomic_check = tegra_sor_encoder_atomic_check,
 };
 
+static void tegra_sor_dp_disable(struct drm_encoder *encoder)
+{
+	struct tegra_output *output = encoder_to_output(encoder);
+	struct tegra_dc *dc = to_tegra_dc(encoder->crtc);
+	struct tegra_sor *sor = to_sor(output);
+	u32 value;
+	int err;
+
+	if (output->panel)
+		drm_panel_disable(output->panel);
+
+	/*
+	 * Do not attempt to power down a DP link if we're not connected since
+	 * the AUX transactions would just be timing out.
+	 */
+	if (output->connector.status != connector_status_disconnected) {
+		err = drm_dp_link_power_down(sor->aux, &sor->link);
+		if (err < 0)
+			dev_err(sor->dev, "failed to power down link: %d\n",
+				err);
+	}
+
+	err = tegra_sor_detach(sor);
+	if (err < 0)
+		dev_err(sor->dev, "failed to detach SOR: %d\n", err);
+
+	tegra_sor_writel(sor, 0, SOR_STATE1);
+	tegra_sor_update(sor);
+
+	value = tegra_dc_readl(dc, DC_DISP_DISP_WIN_OPTIONS);
+	value &= ~SOR_ENABLE(sor->index);
+	tegra_dc_writel(dc, value, DC_DISP_DISP_WIN_OPTIONS);
+	tegra_dc_commit(dc);
+
+	value = tegra_sor_readl(sor, SOR_STATE1);
+	value &= ~SOR_STATE_ASY_PROTOCOL_MASK;
+	value &= ~SOR_STATE_ASY_SUBOWNER_MASK;
+	value &= ~SOR_STATE_ASY_OWNER_MASK;
+	tegra_sor_writel(sor, value, SOR_STATE1);
+	tegra_sor_update(sor);
+
+	/* switch to safe parent clock */
+	err = tegra_sor_set_parent_clock(sor, sor->clk_safe);
+	if (err < 0)
+		dev_err(sor->dev, "failed to set safe clock: %d\n", err);
+
+	err = tegra_sor_power_down(sor);
+	if (err < 0)
+		dev_err(sor->dev, "failed to power down SOR: %d\n", err);
+
+	err = tegra_io_pad_power_disable(sor->pad);
+	if (err < 0)
+		dev_err(sor->dev, "failed to power off I/O pad: %d\n", err);
+
+	err = drm_dp_aux_disable(sor->aux);
+	if (err < 0)
+		dev_err(sor->dev, "failed disable DPAUX: %d\n", err);
+
+	if (output->panel)
+		drm_panel_unprepare(output->panel);
+
+	pm_runtime_put(sor->dev);
+}
+
+static void tegra_sor_dp_enable(struct drm_encoder *encoder)
+{
+	struct tegra_output *output = encoder_to_output(encoder);
+	struct tegra_dc *dc = to_tegra_dc(encoder->crtc);
+	struct tegra_sor *sor = to_sor(output);
+	struct tegra_sor_config config;
+	struct tegra_sor_state *state;
+	struct drm_display_mode *mode;
+	struct drm_display_info *info;
+	unsigned int i;
+	u32 value;
+	int err;
+
+	state = to_sor_state(output->connector.state);
+	mode = &encoder->crtc->state->adjusted_mode;
+	info = &output->connector.display_info;
+
+	pm_runtime_get_sync(sor->dev);
+
+	/* switch to safe parent clock */
+	err = tegra_sor_set_parent_clock(sor, sor->clk_safe);
+	if (err < 0)
+		dev_err(sor->dev, "failed to set safe parent clock: %d\n", err);
+
+	err = tegra_io_pad_power_enable(sor->pad);
+	if (err < 0)
+		dev_err(sor->dev, "failed to power on LVDS rail: %d\n", err);
+
+	usleep_range(20, 100);
+
+	err = drm_dp_aux_enable(sor->aux);
+	if (err < 0)
+		dev_err(sor->dev, "failed to enable DPAUX: %d\n", err);
+
+	err = drm_dp_link_probe(sor->aux, &sor->link);
+	if (err < 0)
+		dev_err(sor->dev, "failed to probe DP link: %d\n", err);
+
+	tegra_sor_filter_rates(sor);
+
+	err = drm_dp_link_choose(&sor->link, mode, info);
+	if (err < 0)
+		dev_err(sor->dev, "failed to choose link: %d\n", err);
+
+	if (output->panel)
+		drm_panel_prepare(output->panel);
+
+	value = tegra_sor_readl(sor, sor->soc->regs->pll2);
+	value &= ~SOR_PLL2_BANDGAP_POWERDOWN;
+	tegra_sor_writel(sor, value, sor->soc->regs->pll2);
+
+	usleep_range(20, 40);
+
+	value = tegra_sor_readl(sor, sor->soc->regs->pll3);
+	value |= SOR_PLL3_PLL_VDD_MODE_3V3;
+	tegra_sor_writel(sor, value, sor->soc->regs->pll3);
+
+	value = tegra_sor_readl(sor, sor->soc->regs->pll0);
+	value &= ~(SOR_PLL0_VCOPD | SOR_PLL0_PWR);
+	tegra_sor_writel(sor, value, sor->soc->regs->pll0);
+
+	value = tegra_sor_readl(sor, sor->soc->regs->pll2);
+	value &= ~SOR_PLL2_SEQ_PLLCAPPD_ENFORCE;
+	value |= SOR_PLL2_SEQ_PLLCAPPD;
+	tegra_sor_writel(sor, value, sor->soc->regs->pll2);
+
+	usleep_range(200, 400);
+
+	value = tegra_sor_readl(sor, sor->soc->regs->pll2);
+	value &= ~SOR_PLL2_POWERDOWN_OVERRIDE;
+	value &= ~SOR_PLL2_PORT_POWERDOWN;
+	tegra_sor_writel(sor, value, sor->soc->regs->pll2);
+
+	value = tegra_sor_readl(sor, SOR_CLK_CNTRL);
+	value &= ~SOR_CLK_CNTRL_DP_CLK_SEL_MASK;
+
+	if (output->panel)
+		value |= SOR_CLK_CNTRL_DP_CLK_SEL_SINGLE_DPCLK;
+	else
+		value |= SOR_CLK_CNTRL_DP_CLK_SEL_DIFF_DPCLK;
+
+	tegra_sor_writel(sor, value, SOR_CLK_CNTRL);
+
+	usleep_range(200, 400);
+
+	value = tegra_sor_readl(sor, SOR_DP_SPARE0);
+	/* XXX not in TRM */
+	if (output->panel)
+		value |= SOR_DP_SPARE_PANEL_INTERNAL;
+	else
+		value &= ~SOR_DP_SPARE_PANEL_INTERNAL;
+
+	value |= SOR_DP_SPARE_SEQ_ENABLE;
+	tegra_sor_writel(sor, value, SOR_DP_SPARE0);
+
+	/* XXX not in TRM */
+	tegra_sor_writel(sor, 0, SOR_LVDS);
+
+	value = tegra_sor_readl(sor, sor->soc->regs->pll0);
+	value &= ~SOR_PLL0_ICHPMP_MASK;
+	value &= ~SOR_PLL0_VCOCAP_MASK;
+	value |= SOR_PLL0_ICHPMP(0x1);
+	value |= SOR_PLL0_VCOCAP(0x3);
+	value |= SOR_PLL0_RESISTOR_EXT;
+	tegra_sor_writel(sor, value, sor->soc->regs->pll0);
+
+	/* XXX not in TRM */
+	for (value = 0, i = 0; i < 5; i++)
+		value |= SOR_XBAR_CTRL_LINK0_XSEL(i, sor->soc->xbar_cfg[i]) |
+			 SOR_XBAR_CTRL_LINK1_XSEL(i, i);
+
+	tegra_sor_writel(sor, 0x00000000, SOR_XBAR_POL);
+	tegra_sor_writel(sor, value, SOR_XBAR_CTRL);
+
+	/*
+	 * Switch the pad clock to the DP clock. Note that we cannot actually
+	 * do this because Tegra186 and later don't support clk_set_parent()
+	 * on the sorX_pad_clkout clocks. We already do the equivalent above
+	 * using the DP_CLK_SEL mux of the SOR_CLK_CNTRL register.
+	 */
+#if 0
+	err = clk_set_parent(sor->clk_pad, sor->clk_parent);
+	if (err < 0) {
+		dev_err(sor->dev, "failed to select pad parent clock: %d\n",
+			err);
+		return;
+	}
+#endif
+
+	/* switch the SOR clock to the pad clock */
+	err = tegra_sor_set_parent_clock(sor, sor->clk_pad);
+	if (err < 0) {
+		dev_err(sor->dev, "failed to select SOR parent clock: %d\n",
+			err);
+		return;
+	}
+
+	/* switch the output clock to the parent pixel clock */
+	err = clk_set_parent(sor->clk, sor->clk_parent);
+	if (err < 0) {
+		dev_err(sor->dev, "failed to select output parent clock: %d\n",
+			err);
+		return;
+	}
+
+	/* use DP-A protocol */
+	value = tegra_sor_readl(sor, SOR_STATE1);
+	value &= ~SOR_STATE_ASY_PROTOCOL_MASK;
+	value |= SOR_STATE_ASY_PROTOCOL_DP_A;
+	tegra_sor_writel(sor, value, SOR_STATE1);
+
+	/* enable port */
+	value = tegra_sor_readl(sor, SOR_DP_LINKCTL0);
+	value |= SOR_DP_LINKCTL_ENABLE;
+	tegra_sor_writel(sor, value, SOR_DP_LINKCTL0);
+
+	tegra_sor_dp_term_calibrate(sor);
+
+	err = drm_dp_link_train(&sor->link);
+	if (err < 0)
+		dev_err(sor->dev, "link training failed: %d\n", err);
+	else
+		dev_dbg(sor->dev, "link training succeeded\n");
+
+	err = drm_dp_link_power_up(sor->aux, &sor->link);
+	if (err < 0)
+		dev_err(sor->dev, "failed to power up DP link: %d\n", err);
+
+	/* compute configuration */
+	memset(&config, 0, sizeof(config));
+	config.bits_per_pixel = state->bpc * 3;
+
+	err = tegra_sor_compute_config(sor, mode, &config, &sor->link);
+	if (err < 0)
+		dev_err(sor->dev, "failed to compute configuration: %d\n", err);
+
+	tegra_sor_apply_config(sor, &config);
+	tegra_sor_mode_set(sor, mode, state);
+
+	if (output->panel) {
+		/* CSTM (LVDS, link A/B, upper) */
+		value = SOR_CSTM_LVDS | SOR_CSTM_LINK_ACT_A | SOR_CSTM_LINK_ACT_B |
+			SOR_CSTM_UPPER;
+		tegra_sor_writel(sor, value, SOR_CSTM);
+
+		/* PWM setup */
+		err = tegra_sor_setup_pwm(sor, 250);
+		if (err < 0)
+			dev_err(sor->dev, "failed to setup PWM: %d\n", err);
+	}
+
+	tegra_sor_update(sor);
+
+	err = tegra_sor_power_up(sor, 250);
+	if (err < 0)
+		dev_err(sor->dev, "failed to power up SOR: %d\n", err);
+
+	/* attach and wake up */
+	err = tegra_sor_attach(sor);
+	if (err < 0)
+		dev_err(sor->dev, "failed to attach SOR: %d\n", err);
+
+	value = tegra_dc_readl(dc, DC_DISP_DISP_WIN_OPTIONS);
+	value |= SOR_ENABLE(sor->index);
+	tegra_dc_writel(dc, value, DC_DISP_DISP_WIN_OPTIONS);
+
+	tegra_dc_commit(dc);
+
+	err = tegra_sor_wakeup(sor);
+	if (err < 0)
+		dev_err(sor->dev, "failed to wakeup SOR: %d\n", err);
+
+	if (output->panel)
+		drm_panel_enable(output->panel);
+}
+
+static const struct drm_encoder_helper_funcs tegra_sor_dp_helpers = {
+	.disable = tegra_sor_dp_disable,
+	.enable = tegra_sor_dp_enable,
+	.atomic_check = tegra_sor_encoder_atomic_check,
+};
+
+static int tegra_sor_hdmi_probe(struct tegra_sor *sor)
+{
+	int err;
+
+	sor->avdd_io_supply = devm_regulator_get(sor->dev, "avdd-io");
+	if (IS_ERR(sor->avdd_io_supply)) {
+		dev_err(sor->dev, "cannot get AVDD I/O supply: %ld\n",
+			PTR_ERR(sor->avdd_io_supply));
+		return PTR_ERR(sor->avdd_io_supply);
+	}
+
+	err = regulator_enable(sor->avdd_io_supply);
+	if (err < 0) {
+		dev_err(sor->dev, "failed to enable AVDD I/O supply: %d\n",
+			err);
+		return err;
+	}
+
+	sor->vdd_pll_supply = devm_regulator_get(sor->dev, "vdd-pll");
+	if (IS_ERR(sor->vdd_pll_supply)) {
+		dev_err(sor->dev, "cannot get VDD PLL supply: %ld\n",
+			PTR_ERR(sor->vdd_pll_supply));
+		return PTR_ERR(sor->vdd_pll_supply);
+	}
+
+	err = regulator_enable(sor->vdd_pll_supply);
+	if (err < 0) {
+		dev_err(sor->dev, "failed to enable VDD PLL supply: %d\n",
+			err);
+		return err;
+	}
+
+	sor->hdmi_supply = devm_regulator_get(sor->dev, "hdmi");
+	if (IS_ERR(sor->hdmi_supply)) {
+		dev_err(sor->dev, "cannot get HDMI supply: %ld\n",
+			PTR_ERR(sor->hdmi_supply));
+		return PTR_ERR(sor->hdmi_supply);
+	}
+
+	err = regulator_enable(sor->hdmi_supply);
+	if (err < 0) {
+		dev_err(sor->dev, "failed to enable HDMI supply: %d\n", err);
+		return err;
+	}
+
+	INIT_DELAYED_WORK(&sor->scdc, tegra_sor_hdmi_scdc_work);
+
+	return 0;
+}
+
+static int tegra_sor_hdmi_remove(struct tegra_sor *sor)
+{
+	regulator_disable(sor->hdmi_supply);
+	regulator_disable(sor->vdd_pll_supply);
+	regulator_disable(sor->avdd_io_supply);
+
+	return 0;
+}
+
+static const struct tegra_sor_ops tegra_sor_hdmi_ops = {
+	.name = "HDMI",
+	.probe = tegra_sor_hdmi_probe,
+	.remove = tegra_sor_hdmi_remove,
+	.audio_enable = tegra_sor_hdmi_audio_enable,
+	.audio_disable = tegra_sor_hdmi_audio_disable,
+};
+
+static int tegra_sor_dp_probe(struct tegra_sor *sor)
+{
+	int err;
+
+	sor->avdd_io_supply = devm_regulator_get(sor->dev, "avdd-io-hdmi-dp");
+	if (IS_ERR(sor->avdd_io_supply))
+		return PTR_ERR(sor->avdd_io_supply);
+
+	err = regulator_enable(sor->avdd_io_supply);
+	if (err < 0)
+		return err;
+
+	sor->vdd_pll_supply = devm_regulator_get(sor->dev, "vdd-hdmi-dp-pll");
+	if (IS_ERR(sor->vdd_pll_supply))
+		return PTR_ERR(sor->vdd_pll_supply);
+
+	err = regulator_enable(sor->vdd_pll_supply);
+	if (err < 0)
+		return err;
+
+	return 0;
+}
+
+static int tegra_sor_dp_remove(struct tegra_sor *sor)
+{
+	regulator_disable(sor->vdd_pll_supply);
+	regulator_disable(sor->avdd_io_supply);
+
+	return 0;
+}
+
+static const struct tegra_sor_ops tegra_sor_dp_ops = {
+	.name = "DP",
+	.probe = tegra_sor_dp_probe,
+	.remove = tegra_sor_dp_remove,
+};
+
 static int tegra_sor_init(struct host1x_client *client)
 {
 	struct drm_device *drm = dev_get_drvdata(client->parent);
@@ -2810,11 +3058,10 @@ static int tegra_sor_init(struct host1x_client *client)
 	struct tegra_sor *sor = host1x_client_to_sor(client);
 	int connector = DRM_MODE_CONNECTOR_Unknown;
 	int encoder = DRM_MODE_ENCODER_NONE;
-	u32 value;
 	int err;
 
 	if (!sor->aux) {
-		if (sor->soc->supports_hdmi) {
+		if (sor->ops == &tegra_sor_hdmi_ops) {
 			connector = DRM_MODE_CONNECTOR_HDMIA;
 			encoder = DRM_MODE_ENCODER_TMDS;
 			helpers = &tegra_sor_hdmi_helpers;
@@ -2823,14 +3070,18 @@ static int tegra_sor_init(struct host1x_client *client)
 			encoder = DRM_MODE_ENCODER_LVDS;
 		}
 	} else {
-		if (sor->soc->supports_edp) {
+		if (sor->output.panel) {
 			connector = DRM_MODE_CONNECTOR_eDP;
 			encoder = DRM_MODE_ENCODER_TMDS;
-			helpers = &tegra_sor_edp_helpers;
-		} else if (sor->soc->supports_dp) {
+			helpers = &tegra_sor_dp_helpers;
+		} else {
 			connector = DRM_MODE_CONNECTOR_DisplayPort;
 			encoder = DRM_MODE_ENCODER_TMDS;
+			helpers = &tegra_sor_dp_helpers;
 		}
+
+		sor->link.ops = &tegra_sor_dp_link_ops;
+		sor->link.aux = sor->aux;
 	}
 
 	sor->output.dev = sor->dev;
@@ -2913,15 +3164,6 @@ static int tegra_sor_init(struct host1x_client *client)
 	if (err < 0)
 		return err;
 
-	/*
-	 * Enable and unmask the HDA codec SCRATCH0 register interrupt. This
-	 * is used for interoperability between the HDA codec driver and the
-	 * HDMI/DP driver.
-	 */
-	value = SOR_INT_CODEC_SCRATCH1 | SOR_INT_CODEC_SCRATCH0;
-	tegra_sor_writel(sor, value, SOR_INT_ENABLE);
-	tegra_sor_writel(sor, value, SOR_INT_MASK);
-
 	return 0;
 }
 
@@ -2930,9 +3172,6 @@ static int tegra_sor_exit(struct host1x_client *client)
 	struct tegra_sor *sor = host1x_client_to_sor(client);
 	int err;
 
-	tegra_sor_writel(sor, 0, SOR_INT_MASK);
-	tegra_sor_writel(sor, 0, SOR_INT_ENABLE);
-
 	tegra_output_exit(&sor->output);
 
 	if (sor->aux) {
@@ -2955,75 +3194,6 @@ static const struct host1x_client_ops sor_client_ops = {
 	.exit = tegra_sor_exit,
 };
 
-static const struct tegra_sor_ops tegra_sor_edp_ops = {
-	.name = "eDP",
-};
-
-static int tegra_sor_hdmi_probe(struct tegra_sor *sor)
-{
-	int err;
-
-	sor->avdd_io_supply = devm_regulator_get(sor->dev, "avdd-io");
-	if (IS_ERR(sor->avdd_io_supply)) {
-		dev_err(sor->dev, "cannot get AVDD I/O supply: %ld\n",
-			PTR_ERR(sor->avdd_io_supply));
-		return PTR_ERR(sor->avdd_io_supply);
-	}
-
-	err = regulator_enable(sor->avdd_io_supply);
-	if (err < 0) {
-		dev_err(sor->dev, "failed to enable AVDD I/O supply: %d\n",
-			err);
-		return err;
-	}
-
-	sor->vdd_pll_supply = devm_regulator_get(sor->dev, "vdd-pll");
-	if (IS_ERR(sor->vdd_pll_supply)) {
-		dev_err(sor->dev, "cannot get VDD PLL supply: %ld\n",
-			PTR_ERR(sor->vdd_pll_supply));
-		return PTR_ERR(sor->vdd_pll_supply);
-	}
-
-	err = regulator_enable(sor->vdd_pll_supply);
-	if (err < 0) {
-		dev_err(sor->dev, "failed to enable VDD PLL supply: %d\n",
-			err);
-		return err;
-	}
-
-	sor->hdmi_supply = devm_regulator_get(sor->dev, "hdmi");
-	if (IS_ERR(sor->hdmi_supply)) {
-		dev_err(sor->dev, "cannot get HDMI supply: %ld\n",
-			PTR_ERR(sor->hdmi_supply));
-		return PTR_ERR(sor->hdmi_supply);
-	}
-
-	err = regulator_enable(sor->hdmi_supply);
-	if (err < 0) {
-		dev_err(sor->dev, "failed to enable HDMI supply: %d\n", err);
-		return err;
-	}
-
-	INIT_DELAYED_WORK(&sor->scdc, tegra_sor_hdmi_scdc_work);
-
-	return 0;
-}
-
-static int tegra_sor_hdmi_remove(struct tegra_sor *sor)
-{
-	regulator_disable(sor->hdmi_supply);
-	regulator_disable(sor->vdd_pll_supply);
-	regulator_disable(sor->avdd_io_supply);
-
-	return 0;
-}
-
-static const struct tegra_sor_ops tegra_sor_hdmi_ops = {
-	.name = "HDMI",
-	.probe = tegra_sor_hdmi_probe,
-	.remove = tegra_sor_hdmi_remove,
-};
-
 static const u8 tegra124_sor_xbar_cfg[5] = {
 	0, 1, 2, 3, 4
 };
@@ -3043,14 +3213,161 @@ static const struct tegra_sor_regs tegra124_sor_regs = {
 	.dp_padctl2 = 0x73,
 };
 
+/* Tegra124 and Tegra132 have lanes 0 and 2 swapped. */
+static const u8 tegra124_sor_lane_map[4] = {
+	2, 1, 0, 3,
+};
+
+static const u8 tegra124_sor_voltage_swing[4][4][4] = {
+	{
+		{ 0x13, 0x19, 0x1e, 0x28 },
+		{ 0x1e, 0x25, 0x2d, },
+		{ 0x28, 0x32, },
+		{ 0x3c, },
+	}, {
+		{ 0x12, 0x17, 0x1b, 0x25 },
+		{ 0x1c, 0x23, 0x2a, },
+		{ 0x25, 0x2f, },
+		{ 0x39, }
+	}, {
+		{ 0x12, 0x16, 0x1a, 0x22 },
+		{ 0x1b, 0x20, 0x27, },
+		{ 0x24, 0x2d, },
+		{ 0x36, },
+	}, {
+		{ 0x11, 0x14, 0x17, 0x1f },
+		{ 0x19, 0x1e, 0x24, },
+		{ 0x22, 0x2a, },
+		{ 0x32, },
+	},
+};
+
+static const u8 tegra124_sor_pre_emphasis[4][4][4] = {
+	{
+		{ 0x00, 0x09, 0x13, 0x25 },
+		{ 0x00, 0x0f, 0x1e, },
+		{ 0x00, 0x14, },
+		{ 0x00, },
+	}, {
+		{ 0x00, 0x0a, 0x14, 0x28 },
+		{ 0x00, 0x0f, 0x1e, },
+		{ 0x00, 0x14, },
+		{ 0x00 },
+	}, {
+		{ 0x00, 0x0a, 0x14, 0x28 },
+		{ 0x00, 0x0f, 0x1e, },
+		{ 0x00, 0x14, },
+		{ 0x00, },
+	}, {
+		{ 0x00, 0x0a, 0x14, 0x28 },
+		{ 0x00, 0x0f, 0x1e, },
+		{ 0x00, 0x14, },
+		{ 0x00, },
+	},
+};
+
+static const u8 tegra124_sor_post_cursor[4][4][4] = {
+	{
+		{ 0x00, 0x00, 0x00, 0x00 },
+		{ 0x00, 0x00, 0x00, },
+		{ 0x00, 0x00, },
+		{ 0x00, },
+	}, {
+		{ 0x02, 0x02, 0x04, 0x05 },
+		{ 0x02, 0x04, 0x05, },
+		{ 0x04, 0x05, },
+		{ 0x05, },
+	}, {
+		{ 0x04, 0x05, 0x08, 0x0b },
+		{ 0x05, 0x09, 0x0b, },
+		{ 0x08, 0x0a, },
+		{ 0x0b, },
+	}, {
+		{ 0x05, 0x09, 0x0b, 0x12 },
+		{ 0x09, 0x0d, 0x12, },
+		{ 0x0b, 0x0f, },
+		{ 0x12, },
+	},
+};
+
+static const u8 tegra124_sor_tx_pu[4][4][4] = {
+	{
+		{ 0x20, 0x30, 0x40, 0x60 },
+		{ 0x30, 0x40, 0x60, },
+		{ 0x40, 0x60, },
+		{ 0x60, },
+	}, {
+		{ 0x20, 0x20, 0x30, 0x50 },
+		{ 0x30, 0x40, 0x50, },
+		{ 0x40, 0x50, },
+		{ 0x60, },
+	}, {
+		{ 0x20, 0x20, 0x30, 0x40, },
+		{ 0x30, 0x30, 0x40, },
+		{ 0x40, 0x50, },
+		{ 0x60, },
+	}, {
+		{ 0x20, 0x20, 0x20, 0x40, },
+		{ 0x30, 0x30, 0x40, },
+		{ 0x40, 0x40, },
+		{ 0x60, },
+	},
+};
+
 static const struct tegra_sor_soc tegra124_sor = {
-	.supports_edp = true,
 	.supports_lvds = true,
 	.supports_hdmi = false,
-	.supports_dp = false,
+	.supports_dp = true,
+	.supports_audio = false,
+	.supports_hdcp = false,
 	.regs = &tegra124_sor_regs,
 	.has_nvdisplay = false,
 	.xbar_cfg = tegra124_sor_xbar_cfg,
+	.lane_map = tegra124_sor_lane_map,
+	.voltage_swing = tegra124_sor_voltage_swing,
+	.pre_emphasis = tegra124_sor_pre_emphasis,
+	.post_cursor = tegra124_sor_post_cursor,
+	.tx_pu = tegra124_sor_tx_pu,
+};
+
+static const u8 tegra132_sor_pre_emphasis[4][4][4] = {
+	{
+		{ 0x00, 0x08, 0x12, 0x24 },
+		{ 0x01, 0x0e, 0x1d, },
+		{ 0x01, 0x13, },
+		{ 0x00, },
+	}, {
+		{ 0x00, 0x08, 0x12, 0x24 },
+		{ 0x00, 0x0e, 0x1d, },
+		{ 0x00, 0x13, },
+		{ 0x00 },
+	}, {
+		{ 0x00, 0x08, 0x12, 0x24 },
+		{ 0x00, 0x0e, 0x1d, },
+		{ 0x00, 0x13, },
+		{ 0x00, },
+	}, {
+		{ 0x00, 0x08, 0x12, 0x24 },
+		{ 0x00, 0x0e, 0x1d, },
+		{ 0x00, 0x13, },
+		{ 0x00, },
+	},
+};
+
+static const struct tegra_sor_soc tegra132_sor = {
+	.supports_lvds = true,
+	.supports_hdmi = false,
+	.supports_dp = true,
+	.supports_audio = false,
+	.supports_hdcp = false,
+	.regs = &tegra124_sor_regs,
+	.has_nvdisplay = false,
+	.xbar_cfg = tegra124_sor_xbar_cfg,
+	.lane_map = tegra124_sor_lane_map,
+	.voltage_swing = tegra124_sor_voltage_swing,
+	.pre_emphasis = tegra132_sor_pre_emphasis,
+	.post_cursor = tegra124_sor_post_cursor,
+	.tx_pu = tegra124_sor_tx_pu,
 };
 
 static const struct tegra_sor_regs tegra210_sor_regs = {
@@ -3068,33 +3385,50 @@ static const struct tegra_sor_regs tegra210_sor_regs = {
 	.dp_padctl2 = 0x73,
 };
 
+static const u8 tegra210_sor_xbar_cfg[5] = {
+	2, 1, 0, 3, 4
+};
+
+static const u8 tegra210_sor_lane_map[4] = {
+	0, 1, 2, 3,
+};
+
 static const struct tegra_sor_soc tegra210_sor = {
-	.supports_edp = true,
 	.supports_lvds = false,
 	.supports_hdmi = false,
-	.supports_dp = false,
+	.supports_dp = true,
+	.supports_audio = false,
+	.supports_hdcp = false,
+
 	.regs = &tegra210_sor_regs,
 	.has_nvdisplay = false,
-	.xbar_cfg = tegra124_sor_xbar_cfg,
-};
 
-static const u8 tegra210_sor_xbar_cfg[5] = {
-	2, 1, 0, 3, 4
+	.xbar_cfg = tegra210_sor_xbar_cfg,
+	.lane_map = tegra210_sor_lane_map,
+	.voltage_swing = tegra124_sor_voltage_swing,
+	.pre_emphasis = tegra124_sor_pre_emphasis,
+	.post_cursor = tegra124_sor_post_cursor,
+	.tx_pu = tegra124_sor_tx_pu,
 };
 
 static const struct tegra_sor_soc tegra210_sor1 = {
-	.supports_edp = false,
 	.supports_lvds = false,
 	.supports_hdmi = true,
 	.supports_dp = true,
+	.supports_audio = true,
+	.supports_hdcp = true,
 
 	.regs = &tegra210_sor_regs,
 	.has_nvdisplay = false,
 
 	.num_settings = ARRAY_SIZE(tegra210_sor_hdmi_defaults),
 	.settings = tegra210_sor_hdmi_defaults,
-
 	.xbar_cfg = tegra210_sor_xbar_cfg,
+	.lane_map = tegra210_sor_lane_map,
+	.voltage_swing = tegra124_sor_voltage_swing,
+	.pre_emphasis = tegra124_sor_pre_emphasis,
+	.post_cursor = tegra124_sor_post_cursor,
+	.tx_pu = tegra124_sor_tx_pu,
 };
 
 static const struct tegra_sor_regs tegra186_sor_regs = {
@@ -3112,31 +3446,72 @@ static const struct tegra_sor_regs tegra186_sor_regs = {
 	.dp_padctl2 = 0x16a,
 };
 
-static const struct tegra_sor_soc tegra186_sor = {
-	.supports_edp = false,
-	.supports_lvds = false,
-	.supports_hdmi = false,
-	.supports_dp = true,
-
-	.regs = &tegra186_sor_regs,
-	.has_nvdisplay = true,
+static const u8 tegra186_sor_voltage_swing[4][4][4] = {
+	{
+		{ 0x13, 0x19, 0x1e, 0x28 },
+		{ 0x1e, 0x25, 0x2d, },
+		{ 0x28, 0x32, },
+		{ 0x39, },
+	}, {
+		{ 0x12, 0x16, 0x1b, 0x25 },
+		{ 0x1c, 0x23, 0x2a, },
+		{ 0x25, 0x2f, },
+		{ 0x37, }
+	}, {
+		{ 0x12, 0x16, 0x1a, 0x22 },
+		{ 0x1b, 0x20, 0x27, },
+		{ 0x24, 0x2d, },
+		{ 0x35, },
+	}, {
+		{ 0x11, 0x14, 0x17, 0x1f },
+		{ 0x19, 0x1e, 0x24, },
+		{ 0x22, 0x2a, },
+		{ 0x32, },
+	},
+};
 
-	.xbar_cfg = tegra124_sor_xbar_cfg,
+static const u8 tegra186_sor_pre_emphasis[4][4][4] = {
+	{
+		{ 0x00, 0x08, 0x12, 0x24 },
+		{ 0x01, 0x0e, 0x1d, },
+		{ 0x01, 0x13, },
+		{ 0x00, },
+	}, {
+		{ 0x00, 0x08, 0x12, 0x24 },
+		{ 0x00, 0x0e, 0x1d, },
+		{ 0x00, 0x13, },
+		{ 0x00 },
+	}, {
+		{ 0x00, 0x08, 0x14, 0x24 },
+		{ 0x00, 0x0e, 0x1d, },
+		{ 0x00, 0x13, },
+		{ 0x00, },
+	}, {
+		{ 0x00, 0x08, 0x12, 0x24 },
+		{ 0x00, 0x0e, 0x1d, },
+		{ 0x00, 0x13, },
+		{ 0x00, },
+	},
 };
 
-static const struct tegra_sor_soc tegra186_sor1 = {
-	.supports_edp = false,
+static const struct tegra_sor_soc tegra186_sor = {
 	.supports_lvds = false,
 	.supports_hdmi = true,
 	.supports_dp = true,
+	.supports_audio = true,
+	.supports_hdcp = true,
 
 	.regs = &tegra186_sor_regs,
 	.has_nvdisplay = true,
 
 	.num_settings = ARRAY_SIZE(tegra186_sor_hdmi_defaults),
 	.settings = tegra186_sor_hdmi_defaults,
-
 	.xbar_cfg = tegra124_sor_xbar_cfg,
+	.lane_map = tegra124_sor_lane_map,
+	.voltage_swing = tegra186_sor_voltage_swing,
+	.pre_emphasis = tegra186_sor_pre_emphasis,
+	.post_cursor = tegra124_sor_post_cursor,
+	.tx_pu = tegra124_sor_tx_pu,
 };
 
 static const struct tegra_sor_regs tegra194_sor_regs = {
@@ -3155,10 +3530,11 @@ static const struct tegra_sor_regs tegra194_sor_regs = {
 };
 
 static const struct tegra_sor_soc tegra194_sor = {
-	.supports_edp = true,
 	.supports_lvds = false,
 	.supports_hdmi = true,
 	.supports_dp = true,
+	.supports_audio = true,
+	.supports_hdcp = true,
 
 	.regs = &tegra194_sor_regs,
 	.has_nvdisplay = true,
@@ -3167,14 +3543,19 @@ static const struct tegra_sor_soc tegra194_sor = {
 	.settings = tegra194_sor_hdmi_defaults,
 
 	.xbar_cfg = tegra210_sor_xbar_cfg,
+	.lane_map = tegra124_sor_lane_map,
+	.voltage_swing = tegra186_sor_voltage_swing,
+	.pre_emphasis = tegra186_sor_pre_emphasis,
+	.post_cursor = tegra124_sor_post_cursor,
+	.tx_pu = tegra124_sor_tx_pu,
 };
 
 static const struct of_device_id tegra_sor_of_match[] = {
 	{ .compatible = "nvidia,tegra194-sor", .data = &tegra194_sor },
-	{ .compatible = "nvidia,tegra186-sor1", .data = &tegra186_sor1 },
 	{ .compatible = "nvidia,tegra186-sor", .data = &tegra186_sor },
 	{ .compatible = "nvidia,tegra210-sor1", .data = &tegra210_sor1 },
 	{ .compatible = "nvidia,tegra210-sor", .data = &tegra210_sor },
+	{ .compatible = "nvidia,tegra132-sor", .data = &tegra132_sor },
 	{ .compatible = "nvidia,tegra124-sor", .data = &tegra124_sor },
 	{ },
 };
@@ -3200,6 +3581,11 @@ static int tegra_sor_parse_dt(struct tegra_sor *sor)
 		 * earlier
 		 */
 		sor->pad = TEGRA_IO_PAD_HDMI_DP0 + sor->index;
+	} else {
+		if (!sor->soc->supports_audio)
+			sor->index = 0;
+		else
+			sor->index = 1;
 	}
 
 	err = of_property_read_u32_array(np, "nvidia,xbar-cfg", xbar_cfg, 5);
@@ -3234,9 +3620,11 @@ static irqreturn_t tegra_sor_irq(int irq, void *data)
 
 			tegra_hda_parse_format(format, &sor->format);
 
-			tegra_sor_hdmi_audio_enable(sor);
+			if (sor->ops->audio_enable)
+				sor->ops->audio_enable(sor);
 		} else {
-			tegra_sor_hdmi_audio_disable(sor);
+			if (sor->ops->audio_disable)
+				sor->ops->audio_disable(sor);
 		}
 	}
 
@@ -3273,6 +3661,8 @@ static int tegra_sor_probe(struct platform_device *pdev)
 
 		if (!sor->aux)
 			return -EPROBE_DEFER;
+
+		sor->output.ddc = &sor->aux->ddc;
 	}
 
 	if (!sor->aux) {
@@ -3287,16 +3677,15 @@ static int tegra_sor_probe(struct platform_device *pdev)
 			return -ENODEV;
 		}
 	} else {
-		if (sor->soc->supports_edp) {
-			sor->ops = &tegra_sor_edp_ops;
-			sor->pad = TEGRA_IO_PAD_LVDS;
-		} else if (sor->soc->supports_dp) {
-			dev_err(&pdev->dev, "DisplayPort not supported yet\n");
-			return -ENODEV;
-		} else {
-			dev_err(&pdev->dev, "unknown (DP) support\n");
-			return -ENODEV;
-		}
+		np = of_parse_phandle(pdev->dev.of_node, "nvidia,panel", 0);
+		/*
+		 * No need to keep this around since we only use it as a check
+		 * to see if a panel is connected (eDP) or not (DP).
+		 */
+		of_node_put(np);
+
+		sor->ops = &tegra_sor_dp_ops;
+		sor->pad = TEGRA_IO_PAD_LVDS;
 	}
 
 	err = tegra_sor_parse_dt(sor);
@@ -3451,6 +3840,8 @@ static int tegra_sor_probe(struct platform_device *pdev)
 	 * pad output clock.
 	 */
 	if (!sor->clk_pad) {
+		char *name;
+
 		err = pm_runtime_get_sync(&pdev->dev);
 		if (err < 0) {
 			dev_err(&pdev->dev, "failed to get runtime PM: %d\n",
@@ -3458,8 +3849,13 @@ static int tegra_sor_probe(struct platform_device *pdev)
 			goto remove;
 		}
 
-		sor->clk_pad = tegra_clk_sor_pad_register(sor,
-							  "sor1_pad_clkout");
+		name = devm_kasprintf(&pdev->dev, GFP_KERNEL, "sor%u_pad_clkout", sor->index);
+		if (!name) {
+			err = -ENOMEM;
+			goto remove;
+		}
+
+		sor->clk_pad = tegra_clk_sor_pad_register(sor, name);
 		pm_runtime_put(&pdev->dev);
 	}
 
diff --git a/drivers/gpu/drm/tegra/sor.h b/drivers/gpu/drm/tegra/sor.h
index f8efd8be4b7c..00e09d5dca30 100644
--- a/drivers/gpu/drm/tegra/sor.h
+++ b/drivers/gpu/drm/tegra/sor.h
@@ -39,6 +39,7 @@
 #define  SOR_STATE_ASY_CRC_MODE_NON_ACTIVE	(0x2 << 6)
 #define  SOR_STATE_ASY_CRC_MODE_COMPLETE	(0x1 << 6)
 #define  SOR_STATE_ASY_CRC_MODE_ACTIVE		(0x0 << 6)
+#define  SOR_STATE_ASY_SUBOWNER_MASK		(0x3 << 4)
 #define  SOR_STATE_ASY_OWNER_MASK		0xf
 #define  SOR_STATE_ASY_OWNER(x)			(((x) & 0xf) << 0)
 
@@ -283,10 +284,12 @@
 #define  SOR_DP_PADCTL_CM_TXD_2		(1 << 6)
 #define  SOR_DP_PADCTL_CM_TXD_1		(1 << 5)
 #define  SOR_DP_PADCTL_CM_TXD_0		(1 << 4)
+#define  SOR_DP_PADCTL_CM_TXD(x)	(1 << (4 + (x)))
 #define  SOR_DP_PADCTL_PD_TXD_3		(1 << 3)
 #define  SOR_DP_PADCTL_PD_TXD_0		(1 << 2)
 #define  SOR_DP_PADCTL_PD_TXD_1		(1 << 1)
 #define  SOR_DP_PADCTL_PD_TXD_2		(1 << 0)
+#define  SOR_DP_PADCTL_PD_TXD(x)	(1 << (0 + (x)))
 
 #define SOR_DP_PADCTL1 0x5d
 
diff --git a/drivers/gpu/drm/tegra/vic.c b/drivers/gpu/drm/tegra/vic.c
index cd0399fd8c63..9444ba183990 100644
--- a/drivers/gpu/drm/tegra/vic.c
+++ b/drivers/gpu/drm/tegra/vic.c
@@ -34,7 +34,6 @@ struct vic {
 	void __iomem *regs;
 	struct tegra_drm_client client;
 	struct host1x_channel *channel;
-	struct iommu_domain *domain;
 	struct device *dev;
 	struct clk *clk;
 	struct reset_control *rst;
@@ -97,6 +96,9 @@ static int vic_runtime_suspend(struct device *dev)
 
 static int vic_boot(struct vic *vic)
 {
+#ifdef CONFIG_IOMMU_API
+	struct iommu_fwspec *spec = dev_iommu_fwspec_get(vic->dev);
+#endif
 	u32 fce_ucode_size, fce_bin_data_offset;
 	void *hdr;
 	int err = 0;
@@ -105,15 +107,14 @@ static int vic_boot(struct vic *vic)
 		return 0;
 
 #ifdef CONFIG_IOMMU_API
-	if (vic->config->supports_sid) {
-		struct iommu_fwspec *spec = dev_iommu_fwspec_get(vic->dev);
+	if (vic->config->supports_sid && spec) {
 		u32 value;
 
 		value = TRANSCFG_ATT(1, TRANSCFG_SID_FALCON) |
 			TRANSCFG_ATT(0, TRANSCFG_SID_HW);
 		vic_writel(vic, value, VIC_TFBIF_TRANSCFG);
 
-		if (spec && spec->num_ids > 0) {
+		if (spec->num_ids > 0) {
 			value = spec->ids[0] & 0xffff;
 
 			vic_writel(vic, value, VIC_THI_STREAMID0);
@@ -132,9 +133,9 @@ static int vic_boot(struct vic *vic)
 	if (err < 0)
 		return err;
 
-	hdr = vic->falcon.firmware.vaddr;
+	hdr = vic->falcon.firmware.virt;
 	fce_bin_data_offset = *(u32 *)(hdr + VIC_UCODE_FCE_DATA_OFFSET);
-	hdr = vic->falcon.firmware.vaddr +
+	hdr = vic->falcon.firmware.virt +
 		*(u32 *)(hdr + VIC_UCODE_FCE_HEADER_OFFSET);
 	fce_ucode_size = *(u32 *)(hdr + FCE_UCODE_SIZE_OFFSET);
 
@@ -142,7 +143,7 @@ static int vic_boot(struct vic *vic)
 	falcon_execute_method(&vic->falcon, VIC_SET_FCE_UCODE_SIZE,
 			      fce_ucode_size);
 	falcon_execute_method(&vic->falcon, VIC_SET_FCE_UCODE_OFFSET,
-			      (vic->falcon.firmware.paddr + fce_bin_data_offset)
+			      (vic->falcon.firmware.iova + fce_bin_data_offset)
 				>> 8);
 
 	err = falcon_wait_idle(&vic->falcon);
@@ -157,48 +158,21 @@ static int vic_boot(struct vic *vic)
 	return 0;
 }
 
-static void *vic_falcon_alloc(struct falcon *falcon, size_t size,
-			      dma_addr_t *iova)
-{
-	struct tegra_drm *tegra = falcon->data;
-
-	return tegra_drm_alloc(tegra, size, iova);
-}
-
-static void vic_falcon_free(struct falcon *falcon, size_t size,
-			    dma_addr_t iova, void *va)
-{
-	struct tegra_drm *tegra = falcon->data;
-
-	return tegra_drm_free(tegra, size, va, iova);
-}
-
-static const struct falcon_ops vic_falcon_ops = {
-	.alloc = vic_falcon_alloc,
-	.free = vic_falcon_free
-};
-
 static int vic_init(struct host1x_client *client)
 {
 	struct tegra_drm_client *drm = host1x_to_drm_client(client);
-	struct iommu_group *group = iommu_group_get(client->dev);
 	struct drm_device *dev = dev_get_drvdata(client->parent);
 	struct tegra_drm *tegra = dev->dev_private;
 	struct vic *vic = to_vic(drm);
 	int err;
 
-	if (group && tegra->domain) {
-		err = iommu_attach_group(tegra->domain, group);
-		if (err < 0) {
-			dev_err(vic->dev, "failed to attach to domain: %d\n",
-				err);
-			return err;
-		}
-
-		vic->domain = tegra->domain;
+	err = host1x_client_iommu_attach(client);
+	if (err < 0) {
+		dev_err(vic->dev, "failed to attach to domain: %d\n", err);
+		return err;
 	}
 
-	vic->channel = host1x_channel_request(client->dev);
+	vic->channel = host1x_channel_request(client);
 	if (!vic->channel) {
 		err = -ENOMEM;
 		goto detach;
@@ -214,6 +188,12 @@ static int vic_init(struct host1x_client *client)
 	if (err < 0)
 		goto free_syncpt;
 
+	/*
+	 * Inherit the DMA parameters (such as maximum segment size) from the
+	 * parent device.
+	 */
+	client->dev->dma_parms = client->parent->dma_parms;
+
 	return 0;
 
 free_syncpt:
@@ -221,8 +201,7 @@ free_syncpt:
 free_channel:
 	host1x_channel_put(vic->channel);
 detach:
-	if (group && tegra->domain)
-		iommu_detach_group(tegra->domain, group);
+	host1x_client_iommu_detach(client);
 
 	return err;
 }
@@ -230,22 +209,32 @@ detach:
 static int vic_exit(struct host1x_client *client)
 {
 	struct tegra_drm_client *drm = host1x_to_drm_client(client);
-	struct iommu_group *group = iommu_group_get(client->dev);
 	struct drm_device *dev = dev_get_drvdata(client->parent);
 	struct tegra_drm *tegra = dev->dev_private;
 	struct vic *vic = to_vic(drm);
 	int err;
 
+	/* avoid a dangling pointer just in case this disappears */
+	client->dev->dma_parms = NULL;
+
 	err = tegra_drm_unregister_client(tegra, drm);
 	if (err < 0)
 		return err;
 
 	host1x_syncpt_free(client->syncpts[0]);
 	host1x_channel_put(vic->channel);
-
-	if (vic->domain) {
-		iommu_detach_group(vic->domain, group);
-		vic->domain = NULL;
+	host1x_client_iommu_detach(client);
+
+	if (client->group) {
+		dma_unmap_single(vic->dev, vic->falcon.firmware.phys,
+				 vic->falcon.firmware.size, DMA_TO_DEVICE);
+		tegra_drm_free(tegra, vic->falcon.firmware.size,
+			       vic->falcon.firmware.virt,
+			       vic->falcon.firmware.iova);
+	} else {
+		dma_free_coherent(vic->dev, vic->falcon.firmware.size,
+				  vic->falcon.firmware.virt,
+				  vic->falcon.firmware.iova);
 	}
 
 	return 0;
@@ -258,25 +247,64 @@ static const struct host1x_client_ops vic_client_ops = {
 
 static int vic_load_firmware(struct vic *vic)
 {
+	struct host1x_client *client = &vic->client.base;
+	struct tegra_drm *tegra = vic->client.drm;
+	dma_addr_t iova;
+	size_t size;
+	void *virt;
 	int err;
 
-	if (vic->falcon.data)
+	if (vic->falcon.firmware.virt)
 		return 0;
 
-	vic->falcon.data = vic->client.drm;
-
 	err = falcon_read_firmware(&vic->falcon, vic->config->firmware);
 	if (err < 0)
-		goto cleanup;
+		return err;
+
+	size = vic->falcon.firmware.size;
+
+	if (!client->group) {
+		virt = dma_alloc_coherent(vic->dev, size, &iova, GFP_KERNEL);
+
+		err = dma_mapping_error(vic->dev, iova);
+		if (err < 0)
+			return err;
+	} else {
+		virt = tegra_drm_alloc(tegra, size, &iova);
+	}
+
+	vic->falcon.firmware.virt = virt;
+	vic->falcon.firmware.iova = iova;
 
 	err = falcon_load_firmware(&vic->falcon);
 	if (err < 0)
 		goto cleanup;
 
+	/*
+	 * In this case we have received an IOVA from the shared domain, so we
+	 * need to make sure to get the physical address so that the DMA API
+	 * knows what memory pages to flush the cache for.
+	 */
+	if (client->group) {
+		dma_addr_t phys;
+
+		phys = dma_map_single(vic->dev, virt, size, DMA_TO_DEVICE);
+
+		err = dma_mapping_error(vic->dev, phys);
+		if (err < 0)
+			goto cleanup;
+
+		vic->falcon.firmware.phys = phys;
+	}
+
 	return 0;
 
 cleanup:
-	vic->falcon.data = NULL;
+	if (!client->group)
+		dma_free_coherent(vic->dev, size, virt, iova);
+	else
+		tegra_drm_free(tegra, size, virt, iova);
+
 	return err;
 }
 
@@ -374,6 +402,13 @@ static int vic_probe(struct platform_device *pdev)
 	struct vic *vic;
 	int err;
 
+	/* inherit DMA mask from host1x parent */
+	err = dma_coerce_mask_and_coherent(dev, *dev->parent->dma_mask);
+	if (err < 0) {
+		dev_err(&pdev->dev, "failed to set DMA mask: %d\n", err);
+		return err;
+	}
+
 	vic = devm_kzalloc(dev, sizeof(*vic), GFP_KERNEL);
 	if (!vic)
 		return -ENOMEM;
@@ -410,7 +445,6 @@ static int vic_probe(struct platform_device *pdev)
 
 	vic->falcon.dev = dev;
 	vic->falcon.regs = vic->regs;
-	vic->falcon.ops = &vic_falcon_ops;
 
 	err = falcon_init(&vic->falcon);
 	if (err < 0)
diff --git a/drivers/gpu/drm/tilcdc/tilcdc_external.c b/drivers/gpu/drm/tilcdc/tilcdc_external.c
index 43d756b7810e..51d034e095f4 100644
--- a/drivers/gpu/drm/tilcdc/tilcdc_external.c
+++ b/drivers/gpu/drm/tilcdc/tilcdc_external.c
@@ -8,6 +8,7 @@
 #include <linux/of_graph.h>
 
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_of.h>
 
 #include "tilcdc_drv.h"
@@ -139,8 +140,8 @@ int tilcdc_attach_external_device(struct drm_device *ddev)
 	}
 
 	if (panel) {
-		bridge = devm_drm_panel_bridge_add(ddev->dev, panel,
-						   DRM_MODE_CONNECTOR_DPI);
+		bridge = devm_drm_panel_bridge_add_typed(ddev->dev, panel,
+							 DRM_MODE_CONNECTOR_DPI);
 		if (IS_ERR(bridge)) {
 			ret = PTR_ERR(bridge);
 			goto err_encoder_cleanup;
diff --git a/drivers/gpu/drm/tilcdc/tilcdc_plane.c b/drivers/gpu/drm/tilcdc/tilcdc_plane.c
index 3abb9641f212..e2090020b3a0 100644
--- a/drivers/gpu/drm/tilcdc/tilcdc_plane.c
+++ b/drivers/gpu/drm/tilcdc/tilcdc_plane.c
@@ -11,7 +11,7 @@
 
 #include "tilcdc_drv.h"
 
-static struct drm_plane_funcs tilcdc_plane_funcs = {
+static const struct drm_plane_funcs tilcdc_plane_funcs = {
 	.update_plane	= drm_atomic_helper_update_plane,
 	.disable_plane	= drm_atomic_helper_disable_plane,
 	.destroy	= drm_plane_cleanup,
diff --git a/drivers/gpu/drm/tilcdc/tilcdc_tfp410.c b/drivers/gpu/drm/tilcdc/tilcdc_tfp410.c
index 525dc1c0f1c1..530edb3b51cc 100644
--- a/drivers/gpu/drm/tilcdc/tilcdc_tfp410.c
+++ b/drivers/gpu/drm/tilcdc/tilcdc_tfp410.c
@@ -7,6 +7,7 @@
 #include <linux/gpio.h>
 #include <linux/mod_devicetable.h>
 #include <linux/of_gpio.h>
+#include <linux/pinctrl/consumer.h>
 #include <linux/platform_device.h>
 
 #include <drm/drm_atomic_helper.h>
diff --git a/drivers/gpu/drm/tiny/Kconfig b/drivers/gpu/drm/tiny/Kconfig
index 504763423d46..a46ac284dd5e 100644
--- a/drivers/gpu/drm/tiny/Kconfig
+++ b/drivers/gpu/drm/tiny/Kconfig
@@ -63,7 +63,6 @@ config TINYDRM_REPAPER
 	depends on DRM && SPI
 	select DRM_KMS_HELPER
 	select DRM_KMS_CMA_HELPER
-	depends on THERMAL || !THERMAL
 	help
 	  DRM driver for the following Pervasive Displays panels:
 	  1.44" TFT EPD Panel (E1144CS021)
diff --git a/drivers/gpu/drm/tiny/gm12u320.c b/drivers/gpu/drm/tiny/gm12u320.c
index 03d0e2df6774..94fb1f593564 100644
--- a/drivers/gpu/drm/tiny/gm12u320.c
+++ b/drivers/gpu/drm/tiny/gm12u320.c
@@ -649,7 +649,7 @@ static void gm12u320_driver_release(struct drm_device *dev)
 	kfree(gm12u320);
 }
 
-DEFINE_DRM_GEM_SHMEM_FOPS(gm12u320_fops);
+DEFINE_DRM_GEM_FOPS(gm12u320_fops);
 
 static struct drm_driver gm12u320_drm_driver = {
 	.driver_features = DRIVER_MODESET | DRIVER_GEM | DRIVER_ATOMIC,
diff --git a/drivers/gpu/drm/ttm/Makefile b/drivers/gpu/drm/ttm/Makefile
index 01fc670ce7a2..caea2a099496 100644
--- a/drivers/gpu/drm/ttm/Makefile
+++ b/drivers/gpu/drm/ttm/Makefile
@@ -4,8 +4,8 @@
 
 ttm-y := ttm_memory.o ttm_tt.o ttm_bo.o \
 	ttm_bo_util.o ttm_bo_vm.o ttm_module.o \
-	ttm_execbuf_util.o ttm_page_alloc.o ttm_bo_manager.o \
-	ttm_page_alloc_dma.o
+	ttm_execbuf_util.o ttm_page_alloc.o ttm_bo_manager.o
 ttm-$(CONFIG_AGP) += ttm_agp_backend.o
+ttm-$(CONFIG_DRM_TTM_DMA_PAGE_POOL) += ttm_page_alloc_dma.o
 
 obj-$(CONFIG_DRM_TTM) += ttm.o
diff --git a/drivers/gpu/drm/ttm/ttm_agp_backend.c b/drivers/gpu/drm/ttm/ttm_agp_backend.c
index ea4d59eb8966..6050dc846894 100644
--- a/drivers/gpu/drm/ttm/ttm_agp_backend.c
+++ b/drivers/gpu/drm/ttm/ttm_agp_backend.c
@@ -51,7 +51,7 @@ struct ttm_agp_backend {
 static int ttm_agp_bind(struct ttm_tt *ttm, struct ttm_mem_reg *bo_mem)
 {
 	struct ttm_agp_backend *agp_be = container_of(ttm, struct ttm_agp_backend, ttm);
-	struct page *dummy_read_page = ttm->bdev->glob->dummy_read_page;
+	struct page *dummy_read_page = ttm_bo_glob.dummy_read_page;
 	struct drm_mm_node *node = bo_mem->mm_node;
 	struct agp_memory *mem;
 	int ret, cached = (bo_mem->placement & TTM_PL_FLAG_CACHED);
diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c
index 20ff56f27aa4..8d91b0428af1 100644
--- a/drivers/gpu/drm/ttm/ttm_bo.c
+++ b/drivers/gpu/drm/ttm/ttm_bo.c
@@ -51,6 +51,7 @@ static void ttm_bo_global_kobj_release(struct kobject *kobj);
 DEFINE_MUTEX(ttm_global_mutex);
 unsigned ttm_bo_glob_use_count;
 struct ttm_bo_global ttm_bo_glob;
+EXPORT_SYMBOL(ttm_bo_glob);
 
 static struct attribute ttm_bo_count = {
 	.name = "bo_count",
@@ -148,23 +149,21 @@ static void ttm_bo_release_list(struct kref *list_kref)
 {
 	struct ttm_buffer_object *bo =
 	    container_of(list_kref, struct ttm_buffer_object, list_kref);
-	struct ttm_bo_device *bdev = bo->bdev;
 	size_t acc_size = bo->acc_size;
 
 	BUG_ON(kref_read(&bo->list_kref));
 	BUG_ON(kref_read(&bo->kref));
-	BUG_ON(atomic_read(&bo->cpu_writers));
 	BUG_ON(bo->mem.mm_node != NULL);
 	BUG_ON(!list_empty(&bo->lru));
 	BUG_ON(!list_empty(&bo->ddestroy));
 	ttm_tt_destroy(bo->ttm);
-	atomic_dec(&bo->bdev->glob->bo_count);
+	atomic_dec(&ttm_bo_glob.bo_count);
 	dma_fence_put(bo->moving);
 	if (!ttm_bo_uses_embedded_gem_object(bo))
 		dma_resv_fini(&bo->base._resv);
 	mutex_destroy(&bo->wu_mutex);
 	bo->destroy(bo);
-	ttm_mem_global_free(bdev->glob->mem_glob, acc_size);
+	ttm_mem_global_free(&ttm_mem_glob, acc_size);
 }
 
 static void ttm_bo_add_mem_to_lru(struct ttm_buffer_object *bo,
@@ -185,25 +184,20 @@ static void ttm_bo_add_mem_to_lru(struct ttm_buffer_object *bo,
 	list_add_tail(&bo->lru, &man->lru[bo->priority]);
 	kref_get(&bo->list_kref);
 
-	if (bo->ttm && !(bo->ttm->page_flags &
-			 (TTM_PAGE_FLAG_SG | TTM_PAGE_FLAG_SWAPPED))) {
-		list_add_tail(&bo->swap, &bdev->glob->swap_lru[bo->priority]);
+	if (!(man->flags & TTM_MEMTYPE_FLAG_FIXED) && bo->ttm &&
+	    !(bo->ttm->page_flags & (TTM_PAGE_FLAG_SG |
+				     TTM_PAGE_FLAG_SWAPPED))) {
+		list_add_tail(&bo->swap, &ttm_bo_glob.swap_lru[bo->priority]);
 		kref_get(&bo->list_kref);
 	}
 }
 
-void ttm_bo_add_to_lru(struct ttm_buffer_object *bo)
-{
-	ttm_bo_add_mem_to_lru(bo, &bo->mem);
-}
-EXPORT_SYMBOL(ttm_bo_add_to_lru);
-
 static void ttm_bo_ref_bug(struct kref *list_kref)
 {
 	BUG();
 }
 
-void ttm_bo_del_from_lru(struct ttm_buffer_object *bo)
+static void ttm_bo_del_from_lru(struct ttm_buffer_object *bo)
 {
 	struct ttm_bo_device *bdev = bo->bdev;
 	bool notify = false;
@@ -223,16 +217,6 @@ void ttm_bo_del_from_lru(struct ttm_buffer_object *bo)
 		bdev->driver->del_from_lru_notify(bo);
 }
 
-void ttm_bo_del_sub_from_lru(struct ttm_buffer_object *bo)
-{
-	struct ttm_bo_global *glob = bo->bdev->glob;
-
-	spin_lock(&glob->lru_lock);
-	ttm_bo_del_from_lru(bo);
-	spin_unlock(&glob->lru_lock);
-}
-EXPORT_SYMBOL(ttm_bo_del_sub_from_lru);
-
 static void ttm_bo_bulk_move_set_pos(struct ttm_lru_bulk_move_pos *pos,
 				     struct ttm_buffer_object *bo)
 {
@@ -247,7 +231,7 @@ void ttm_bo_move_to_lru_tail(struct ttm_buffer_object *bo,
 	dma_resv_assert_held(bo->base.resv);
 
 	ttm_bo_del_from_lru(bo);
-	ttm_bo_add_to_lru(bo);
+	ttm_bo_add_mem_to_lru(bo, &bo->mem);
 
 	if (bulk && !(bo->mem.placement & TTM_PL_FLAG_NO_EVICT)) {
 		switch (bo->mem.mem_type) {
@@ -310,7 +294,7 @@ void ttm_bo_bulk_move_lru_tail(struct ttm_lru_bulk_move *bulk)
 		dma_resv_assert_held(pos->first->base.resv);
 		dma_resv_assert_held(pos->last->base.resv);
 
-		lru = &pos->first->bdev->glob->swap_lru[i];
+		lru = &ttm_bo_glob.swap_lru[i];
 		list_bulk_move_tail(lru, &pos->first->swap, &pos->last->swap);
 	}
 }
@@ -474,7 +458,6 @@ static void ttm_bo_flush_all_fences(struct ttm_buffer_object *bo)
 static void ttm_bo_cleanup_refs_or_queue(struct ttm_buffer_object *bo)
 {
 	struct ttm_bo_device *bdev = bo->bdev;
-	struct ttm_bo_global *glob = bdev->glob;
 	int ret;
 
 	ret = ttm_bo_individualize_resv(bo);
@@ -484,16 +467,16 @@ static void ttm_bo_cleanup_refs_or_queue(struct ttm_buffer_object *bo)
 		 */
 		dma_resv_wait_timeout_rcu(bo->base.resv, true, false,
 						    30 * HZ);
-		spin_lock(&glob->lru_lock);
+		spin_lock(&ttm_bo_glob.lru_lock);
 		goto error;
 	}
 
-	spin_lock(&glob->lru_lock);
+	spin_lock(&ttm_bo_glob.lru_lock);
 	ret = dma_resv_trylock(bo->base.resv) ? 0 : -EBUSY;
 	if (!ret) {
 		if (dma_resv_test_signaled_rcu(&bo->base._resv, true)) {
 			ttm_bo_del_from_lru(bo);
-			spin_unlock(&glob->lru_lock);
+			spin_unlock(&ttm_bo_glob.lru_lock);
 			if (bo->base.resv != &bo->base._resv)
 				dma_resv_unlock(&bo->base._resv);
 
@@ -511,7 +494,7 @@ static void ttm_bo_cleanup_refs_or_queue(struct ttm_buffer_object *bo)
 		 */
 		if (bo->mem.placement & TTM_PL_FLAG_NO_EVICT) {
 			bo->mem.placement &= ~TTM_PL_FLAG_NO_EVICT;
-			ttm_bo_add_to_lru(bo);
+			ttm_bo_move_to_lru_tail(bo, NULL);
 		}
 
 		dma_resv_unlock(bo->base.resv);
@@ -522,7 +505,7 @@ static void ttm_bo_cleanup_refs_or_queue(struct ttm_buffer_object *bo)
 error:
 	kref_get(&bo->list_kref);
 	list_add_tail(&bo->ddestroy, &bdev->ddestroy);
-	spin_unlock(&glob->lru_lock);
+	spin_unlock(&ttm_bo_glob.lru_lock);
 
 	schedule_delayed_work(&bdev->wq,
 			      ((HZ / 100) < 1) ? 1 : HZ / 100);
@@ -545,7 +528,6 @@ static int ttm_bo_cleanup_refs(struct ttm_buffer_object *bo,
 			       bool interruptible, bool no_wait_gpu,
 			       bool unlock_resv)
 {
-	struct ttm_bo_global *glob = bo->bdev->glob;
 	struct dma_resv *resv;
 	int ret;
 
@@ -564,7 +546,7 @@ static int ttm_bo_cleanup_refs(struct ttm_buffer_object *bo,
 
 		if (unlock_resv)
 			dma_resv_unlock(bo->base.resv);
-		spin_unlock(&glob->lru_lock);
+		spin_unlock(&ttm_bo_glob.lru_lock);
 
 		lret = dma_resv_wait_timeout_rcu(resv, true,
 							   interruptible,
@@ -575,7 +557,7 @@ static int ttm_bo_cleanup_refs(struct ttm_buffer_object *bo,
 		else if (lret == 0)
 			return -EBUSY;
 
-		spin_lock(&glob->lru_lock);
+		spin_lock(&ttm_bo_glob.lru_lock);
 		if (unlock_resv && !dma_resv_trylock(bo->base.resv)) {
 			/*
 			 * We raced, and lost, someone else holds the reservation now,
@@ -585,7 +567,7 @@ static int ttm_bo_cleanup_refs(struct ttm_buffer_object *bo,
 			 * delayed destruction would succeed, so just return success
 			 * here.
 			 */
-			spin_unlock(&glob->lru_lock);
+			spin_unlock(&ttm_bo_glob.lru_lock);
 			return 0;
 		}
 		ret = 0;
@@ -594,7 +576,7 @@ static int ttm_bo_cleanup_refs(struct ttm_buffer_object *bo,
 	if (ret || unlikely(list_empty(&bo->ddestroy))) {
 		if (unlock_resv)
 			dma_resv_unlock(bo->base.resv);
-		spin_unlock(&glob->lru_lock);
+		spin_unlock(&ttm_bo_glob.lru_lock);
 		return ret;
 	}
 
@@ -602,7 +584,7 @@ static int ttm_bo_cleanup_refs(struct ttm_buffer_object *bo,
 	list_del_init(&bo->ddestroy);
 	kref_put(&bo->list_kref, ttm_bo_ref_bug);
 
-	spin_unlock(&glob->lru_lock);
+	spin_unlock(&ttm_bo_glob.lru_lock);
 	ttm_bo_cleanup_memtype_use(bo);
 
 	if (unlock_resv)
@@ -617,7 +599,7 @@ static int ttm_bo_cleanup_refs(struct ttm_buffer_object *bo,
  */
 static bool ttm_bo_delayed_delete(struct ttm_bo_device *bdev, bool remove_all)
 {
-	struct ttm_bo_global *glob = bdev->glob;
+	struct ttm_bo_global *glob = &ttm_bo_glob;
 	struct list_head removed;
 	bool empty;
 
@@ -675,7 +657,7 @@ static void ttm_bo_release(struct kref *kref)
 	if (bo->bdev->driver->release_notify)
 		bo->bdev->driver->release_notify(bo);
 
-	drm_vma_offset_remove(&bdev->vma_manager, &bo->base.vma_node);
+	drm_vma_offset_remove(bdev->vma_manager, &bo->base.vma_node);
 	ttm_mem_io_lock(man, false);
 	ttm_mem_io_free_vm(bo);
 	ttm_mem_io_unlock(man);
@@ -841,13 +823,12 @@ static int ttm_mem_evict_first(struct ttm_bo_device *bdev,
 			       struct ww_acquire_ctx *ticket)
 {
 	struct ttm_buffer_object *bo = NULL, *busy_bo = NULL;
-	struct ttm_bo_global *glob = bdev->glob;
 	struct ttm_mem_type_manager *man = &bdev->man[mem_type];
 	bool locked = false;
 	unsigned i;
 	int ret;
 
-	spin_lock(&glob->lru_lock);
+	spin_lock(&ttm_bo_glob.lru_lock);
 	for (i = 0; i < TTM_MAX_BO_PRIORITY; ++i) {
 		list_for_each_entry(bo, &man->lru[i], lru) {
 			bool busy;
@@ -878,11 +859,11 @@ static int ttm_mem_evict_first(struct ttm_bo_device *bdev,
 
 	if (!bo) {
 		if (busy_bo)
-			ttm_bo_get(busy_bo);
-		spin_unlock(&glob->lru_lock);
+			kref_get(&busy_bo->list_kref);
+		spin_unlock(&ttm_bo_glob.lru_lock);
 		ret = ttm_mem_evict_wait_busy(busy_bo, ctx, ticket);
 		if (busy_bo)
-			ttm_bo_put(busy_bo);
+			kref_put(&busy_bo->list_kref, ttm_bo_release_list);
 		return ret;
 	}
 
@@ -895,17 +876,11 @@ static int ttm_mem_evict_first(struct ttm_bo_device *bdev,
 		return ret;
 	}
 
-	ttm_bo_del_from_lru(bo);
-	spin_unlock(&glob->lru_lock);
+	spin_unlock(&ttm_bo_glob.lru_lock);
 
 	ret = ttm_bo_evict(bo, ctx);
-	if (locked) {
+	if (locked)
 		ttm_bo_unreserve(bo);
-	} else {
-		spin_lock(&glob->lru_lock);
-		ttm_bo_add_to_lru(bo);
-		spin_unlock(&glob->lru_lock);
-	}
 
 	kref_put(&bo->list_kref, ttm_bo_release_list);
 	return ret;
@@ -925,7 +900,8 @@ EXPORT_SYMBOL(ttm_bo_mem_put);
  */
 static int ttm_bo_add_move_fence(struct ttm_buffer_object *bo,
 				 struct ttm_mem_type_manager *man,
-				 struct ttm_mem_reg *mem)
+				 struct ttm_mem_reg *mem,
+				 bool no_wait_gpu)
 {
 	struct dma_fence *fence;
 	int ret;
@@ -934,19 +910,22 @@ static int ttm_bo_add_move_fence(struct ttm_buffer_object *bo,
 	fence = dma_fence_get(man->move);
 	spin_unlock(&man->move_lock);
 
-	if (fence) {
-		dma_resv_add_shared_fence(bo->base.resv, fence);
+	if (!fence)
+		return 0;
 
-		ret = dma_resv_reserve_shared(bo->base.resv, 1);
-		if (unlikely(ret)) {
-			dma_fence_put(fence);
-			return ret;
-		}
+	if (no_wait_gpu)
+		return -EBUSY;
 
-		dma_fence_put(bo->moving);
-		bo->moving = fence;
+	dma_resv_add_shared_fence(bo->base.resv, fence);
+
+	ret = dma_resv_reserve_shared(bo->base.resv, 1);
+	if (unlikely(ret)) {
+		dma_fence_put(fence);
+		return ret;
 	}
 
+	dma_fence_put(bo->moving);
+	bo->moving = fence;
 	return 0;
 }
 
@@ -977,7 +956,7 @@ static int ttm_bo_mem_force_space(struct ttm_buffer_object *bo,
 			return ret;
 	} while (1);
 
-	return ttm_bo_add_move_fence(bo, man, mem);
+	return ttm_bo_add_move_fence(bo, man, mem, ctx->no_wait_gpu);
 }
 
 static uint32_t ttm_bo_select_caching(struct ttm_mem_type_manager *man,
@@ -1067,12 +1046,10 @@ static int ttm_bo_mem_placement(struct ttm_buffer_object *bo,
 	mem->mem_type = mem_type;
 	mem->placement = cur_flags;
 
-	if (bo->mem.mem_type < mem_type && !list_empty(&bo->lru)) {
-		spin_lock(&bo->bdev->glob->lru_lock);
-		ttm_bo_del_from_lru(bo);
-		ttm_bo_add_mem_to_lru(bo, mem);
-		spin_unlock(&bo->bdev->glob->lru_lock);
-	}
+	spin_lock(&ttm_bo_glob.lru_lock);
+	ttm_bo_del_from_lru(bo);
+	ttm_bo_add_mem_to_lru(bo, mem);
+	spin_unlock(&ttm_bo_glob.lru_lock);
 
 	return 0;
 }
@@ -1119,14 +1096,18 @@ int ttm_bo_mem_space(struct ttm_buffer_object *bo,
 		if (unlikely(ret))
 			goto error;
 
-		if (mem->mm_node) {
-			ret = ttm_bo_add_move_fence(bo, man, mem);
-			if (unlikely(ret)) {
-				(*man->func->put_node)(man, mem);
-				goto error;
-			}
-			return 0;
+		if (!mem->mm_node)
+			continue;
+
+		ret = ttm_bo_add_move_fence(bo, man, mem, ctx->no_wait_gpu);
+		if (unlikely(ret)) {
+			(*man->func->put_node)(man, mem);
+			if (ret == -EBUSY)
+				continue;
+
+			goto error;
 		}
+		return 0;
 	}
 
 	for (i = 0; i < placement->num_busy_placement; ++i) {
@@ -1159,9 +1140,9 @@ int ttm_bo_mem_space(struct ttm_buffer_object *bo,
 
 error:
 	if (bo->mem.mem_type == TTM_PL_SYSTEM && !list_empty(&bo->lru)) {
-		spin_lock(&bo->bdev->glob->lru_lock);
+		spin_lock(&ttm_bo_glob.lru_lock);
 		ttm_bo_move_to_lru_tail(bo, NULL);
-		spin_unlock(&bo->bdev->glob->lru_lock);
+		spin_unlock(&ttm_bo_glob.lru_lock);
 	}
 
 	return ret;
@@ -1285,9 +1266,9 @@ int ttm_bo_init_reserved(struct ttm_bo_device *bdev,
 			 struct dma_resv *resv,
 			 void (*destroy) (struct ttm_buffer_object *))
 {
+	struct ttm_mem_global *mem_glob = &ttm_mem_glob;
 	int ret = 0;
 	unsigned long num_pages;
-	struct ttm_mem_global *mem_glob = bdev->glob->mem_glob;
 	bool locked;
 
 	ret = ttm_mem_global_alloc(mem_glob, acc_size, ctx);
@@ -1314,7 +1295,6 @@ int ttm_bo_init_reserved(struct ttm_bo_device *bdev,
 
 	kref_init(&bo->kref);
 	kref_init(&bo->list_kref);
-	atomic_set(&bo->cpu_writers, 0);
 	INIT_LIST_HEAD(&bo->lru);
 	INIT_LIST_HEAD(&bo->ddestroy);
 	INIT_LIST_HEAD(&bo->swap);
@@ -1348,7 +1328,7 @@ int ttm_bo_init_reserved(struct ttm_bo_device *bdev,
 		dma_resv_init(&bo->base._resv);
 		drm_vma_node_reset(&bo->base.vma_node);
 	}
-	atomic_inc(&bo->bdev->glob->bo_count);
+	atomic_inc(&ttm_bo_glob.bo_count);
 
 	/*
 	 * For ttm_bo_type_device buffers, allocate
@@ -1356,7 +1336,7 @@ int ttm_bo_init_reserved(struct ttm_bo_device *bdev,
 	 */
 	if (bo->type == ttm_bo_type_device ||
 	    bo->type == ttm_bo_type_sg)
-		ret = drm_vma_offset_add(&bdev->vma_manager, &bo->base.vma_node,
+		ret = drm_vma_offset_add(bdev->vma_manager, &bo->base.vma_node,
 					 bo->mem.num_pages);
 
 	/* passed reservation objects should already be locked,
@@ -1378,11 +1358,9 @@ int ttm_bo_init_reserved(struct ttm_bo_device *bdev,
 		return ret;
 	}
 
-	if (resv && !(bo->mem.placement & TTM_PL_FLAG_NO_EVICT)) {
-		spin_lock(&bdev->glob->lru_lock);
-		ttm_bo_add_to_lru(bo);
-		spin_unlock(&bdev->glob->lru_lock);
-	}
+	spin_lock(&ttm_bo_glob.lru_lock);
+	ttm_bo_move_to_lru_tail(bo, NULL);
+	spin_unlock(&ttm_bo_glob.lru_lock);
 
 	return ret;
 }
@@ -1480,7 +1458,7 @@ static int ttm_bo_force_list_clean(struct ttm_bo_device *bdev,
 		.flags = TTM_OPT_FLAG_FORCE_ALLOC
 	};
 	struct ttm_mem_type_manager *man = &bdev->man[mem_type];
-	struct ttm_bo_global *glob = bdev->glob;
+	struct ttm_bo_global *glob = &ttm_bo_glob;
 	struct dma_fence *fence;
 	int ret;
 	unsigned i;
@@ -1649,8 +1627,6 @@ static int ttm_bo_global_init(void)
 		goto out;
 
 	spin_lock_init(&glob->lru_lock);
-	glob->mem_glob = &ttm_mem_glob;
-	glob->mem_glob->bo_glob = glob;
 	glob->dummy_read_page = alloc_page(__GFP_ZERO | GFP_DMA32);
 
 	if (unlikely(glob->dummy_read_page == NULL)) {
@@ -1674,10 +1650,10 @@ out:
 
 int ttm_bo_device_release(struct ttm_bo_device *bdev)
 {
+	struct ttm_bo_global *glob = &ttm_bo_glob;
 	int ret = 0;
 	unsigned i = TTM_NUM_MEM_TYPES;
 	struct ttm_mem_type_manager *man;
-	struct ttm_bo_global *glob = bdev->glob;
 
 	while (i--) {
 		man = &bdev->man[i];
@@ -1707,8 +1683,6 @@ int ttm_bo_device_release(struct ttm_bo_device *bdev)
 			pr_debug("Swap list %d was clean\n", i);
 	spin_unlock(&glob->lru_lock);
 
-	drm_vma_offset_manager_destroy(&bdev->vma_manager);
-
 	if (!ret)
 		ttm_bo_global_release();
 
@@ -1719,11 +1693,15 @@ EXPORT_SYMBOL(ttm_bo_device_release);
 int ttm_bo_device_init(struct ttm_bo_device *bdev,
 		       struct ttm_bo_driver *driver,
 		       struct address_space *mapping,
+		       struct drm_vma_offset_manager *vma_manager,
 		       bool need_dma32)
 {
 	struct ttm_bo_global *glob = &ttm_bo_glob;
 	int ret;
 
+	if (WARN_ON(vma_manager == NULL))
+		return -EINVAL;
+
 	ret = ttm_bo_global_init();
 	if (ret)
 		return ret;
@@ -1740,13 +1718,10 @@ int ttm_bo_device_init(struct ttm_bo_device *bdev,
 	if (unlikely(ret != 0))
 		goto out_no_sys;
 
-	drm_vma_offset_manager_init(&bdev->vma_manager,
-				    DRM_FILE_PAGE_OFFSET_START,
-				    DRM_FILE_PAGE_OFFSET_SIZE);
+	bdev->vma_manager = vma_manager;
 	INIT_DELAYED_WORK(&bdev->wq, ttm_bo_delayed_workqueue);
 	INIT_LIST_HEAD(&bdev->ddestroy);
 	bdev->dev_mapping = mapping;
-	bdev->glob = glob;
 	bdev->need_dma32 = need_dma32;
 	mutex_lock(&ttm_global_mutex);
 	list_add_tail(&bdev->device_list, &glob->device_list);
@@ -1826,31 +1801,6 @@ int ttm_bo_wait(struct ttm_buffer_object *bo,
 }
 EXPORT_SYMBOL(ttm_bo_wait);
 
-int ttm_bo_synccpu_write_grab(struct ttm_buffer_object *bo, bool no_wait)
-{
-	int ret = 0;
-
-	/*
-	 * Using ttm_bo_reserve makes sure the lru lists are updated.
-	 */
-
-	ret = ttm_bo_reserve(bo, true, no_wait, NULL);
-	if (unlikely(ret != 0))
-		return ret;
-	ret = ttm_bo_wait(bo, true, no_wait);
-	if (likely(ret == 0))
-		atomic_inc(&bo->cpu_writers);
-	ttm_bo_unreserve(bo);
-	return ret;
-}
-EXPORT_SYMBOL(ttm_bo_synccpu_write_grab);
-
-void ttm_bo_synccpu_write_release(struct ttm_buffer_object *bo)
-{
-	atomic_dec(&bo->cpu_writers);
-}
-EXPORT_SYMBOL(ttm_bo_synccpu_write_release);
-
 /**
  * A buffer object shrink method that tries to swap out the first
  * buffer object on the bo_global::swap_lru list.
@@ -1950,8 +1900,7 @@ void ttm_bo_swapout_all(struct ttm_bo_device *bdev)
 		.no_wait_gpu = false
 	};
 
-	while (ttm_bo_swapout(bdev->glob, &ctx) == 0)
-		;
+	while (ttm_bo_swapout(&ttm_bo_glob, &ctx) == 0);
 }
 EXPORT_SYMBOL(ttm_bo_swapout_all);
 
diff --git a/drivers/gpu/drm/ttm/ttm_bo_util.c b/drivers/gpu/drm/ttm/ttm_bo_util.c
index fe81c565e7ef..6b0883a1776e 100644
--- a/drivers/gpu/drm/ttm/ttm_bo_util.c
+++ b/drivers/gpu/drm/ttm/ttm_bo_util.c
@@ -102,7 +102,6 @@ int ttm_mem_io_lock(struct ttm_mem_type_manager *man, bool interruptible)
 	mutex_lock(&man->io_reserve_mutex);
 	return 0;
 }
-EXPORT_SYMBOL(ttm_mem_io_lock);
 
 void ttm_mem_io_unlock(struct ttm_mem_type_manager *man)
 {
@@ -111,7 +110,6 @@ void ttm_mem_io_unlock(struct ttm_mem_type_manager *man)
 
 	mutex_unlock(&man->io_reserve_mutex);
 }
-EXPORT_SYMBOL(ttm_mem_io_unlock);
 
 static int ttm_mem_io_evict(struct ttm_mem_type_manager *man)
 {
@@ -153,7 +151,6 @@ retry:
 	}
 	return ret;
 }
-EXPORT_SYMBOL(ttm_mem_io_reserve);
 
 void ttm_mem_io_free(struct ttm_bo_device *bdev,
 		     struct ttm_mem_reg *mem)
@@ -169,7 +166,6 @@ void ttm_mem_io_free(struct ttm_bo_device *bdev,
 		bdev->driver->io_mem_free(bdev, mem);
 
 }
-EXPORT_SYMBOL(ttm_mem_io_free);
 
 int ttm_mem_io_reserve_vm(struct ttm_buffer_object *bo)
 {
@@ -503,7 +499,7 @@ static int ttm_buffer_object_transfer(struct ttm_buffer_object *bo,
 	 * TODO: Explicit member copy would probably be better here.
 	 */
 
-	atomic_inc(&bo->bdev->glob->bo_count);
+	atomic_inc(&ttm_bo_glob.bo_count);
 	INIT_LIST_HEAD(&fbo->base.ddestroy);
 	INIT_LIST_HEAD(&fbo->base.lru);
 	INIT_LIST_HEAD(&fbo->base.swap);
@@ -511,15 +507,16 @@ static int ttm_buffer_object_transfer(struct ttm_buffer_object *bo,
 	mutex_init(&fbo->base.wu_mutex);
 	fbo->base.moving = NULL;
 	drm_vma_node_reset(&fbo->base.base.vma_node);
-	atomic_set(&fbo->base.cpu_writers, 0);
 
 	kref_init(&fbo->base.list_kref);
 	kref_init(&fbo->base.kref);
 	fbo->base.destroy = &ttm_transfered_destroy;
 	fbo->base.acc_size = 0;
-	fbo->base.base.resv = &fbo->base.base._resv;
-	dma_resv_init(fbo->base.base.resv);
-	ret = dma_resv_trylock(fbo->base.base.resv);
+	if (bo->base.resv == &bo->base._resv)
+		fbo->base.base.resv = &fbo->base.base._resv;
+
+	dma_resv_init(&fbo->base.base._resv);
+	ret = dma_resv_trylock(&fbo->base.base._resv);
 	WARN_ON(!ret);
 
 	*new_obj = &fbo->base;
@@ -716,7 +713,7 @@ int ttm_bo_move_accel_cleanup(struct ttm_buffer_object *bo,
 		if (ret)
 			return ret;
 
-		dma_resv_add_excl_fence(ghost_obj->base.resv, fence);
+		dma_resv_add_excl_fence(&ghost_obj->base._resv, fence);
 
 		/**
 		 * If we're not moving to fixed memory, the TTM object
@@ -729,7 +726,7 @@ int ttm_bo_move_accel_cleanup(struct ttm_buffer_object *bo,
 		else
 			bo->ttm = NULL;
 
-		ttm_bo_unreserve(ghost_obj);
+		dma_resv_unlock(&ghost_obj->base._resv);
 		ttm_bo_put(ghost_obj);
 	}
 
@@ -772,7 +769,7 @@ int ttm_bo_pipeline_move(struct ttm_buffer_object *bo,
 		if (ret)
 			return ret;
 
-		dma_resv_add_excl_fence(ghost_obj->base.resv, fence);
+		dma_resv_add_excl_fence(&ghost_obj->base._resv, fence);
 
 		/**
 		 * If we're not moving to fixed memory, the TTM object
@@ -785,7 +782,7 @@ int ttm_bo_pipeline_move(struct ttm_buffer_object *bo,
 		else
 			bo->ttm = NULL;
 
-		ttm_bo_unreserve(ghost_obj);
+		dma_resv_unlock(&ghost_obj->base._resv);
 		ttm_bo_put(ghost_obj);
 
 	} else if (from->flags & TTM_MEMTYPE_FLAG_FIXED) {
@@ -841,7 +838,7 @@ int ttm_bo_pipeline_gutting(struct ttm_buffer_object *bo)
 	if (ret)
 		return ret;
 
-	ret = dma_resv_copy_fences(ghost->base.resv, bo->base.resv);
+	ret = dma_resv_copy_fences(&ghost->base._resv, bo->base.resv);
 	/* Last resort, wait for the BO to be idle when we are OOM */
 	if (ret)
 		ttm_bo_wait(bo, false, false);
@@ -850,7 +847,7 @@ int ttm_bo_pipeline_gutting(struct ttm_buffer_object *bo)
 	bo->mem.mem_type = TTM_PL_SYSTEM;
 	bo->ttm = NULL;
 
-	ttm_bo_unreserve(ghost);
+	dma_resv_unlock(&ghost->base._resv);
 	ttm_bo_put(ghost);
 
 	return 0;
diff --git a/drivers/gpu/drm/ttm/ttm_bo_vm.c b/drivers/gpu/drm/ttm/ttm_bo_vm.c
index 76eedb963693..11863fbdd5d6 100644
--- a/drivers/gpu/drm/ttm/ttm_bo_vm.c
+++ b/drivers/gpu/drm/ttm/ttm_bo_vm.c
@@ -42,8 +42,6 @@
 #include <linux/uaccess.h>
 #include <linux/mem_encrypt.h>
 
-#define TTM_BO_VM_NUM_PREFAULT 16
-
 static vm_fault_t ttm_bo_vm_fault_idle(struct ttm_buffer_object *bo,
 				struct vm_fault *vmf)
 {
@@ -106,25 +104,30 @@ static unsigned long ttm_bo_io_mem_pfn(struct ttm_buffer_object *bo,
 		+ page_offset;
 }
 
-static vm_fault_t ttm_bo_vm_fault(struct vm_fault *vmf)
+/**
+ * ttm_bo_vm_reserve - Reserve a buffer object in a retryable vm callback
+ * @bo: The buffer object
+ * @vmf: The fault structure handed to the callback
+ *
+ * vm callbacks like fault() and *_mkwrite() allow for the mm_sem to be dropped
+ * during long waits, and after the wait the callback will be restarted. This
+ * is to allow other threads using the same virtual memory space concurrent
+ * access to map(), unmap() completely unrelated buffer objects. TTM buffer
+ * object reservations sometimes wait for GPU and should therefore be
+ * considered long waits. This function reserves the buffer object interruptibly
+ * taking this into account. Starvation is avoided by the vm system not
+ * allowing too many repeated restarts.
+ * This function is intended to be used in customized fault() and _mkwrite()
+ * handlers.
+ *
+ * Return:
+ *    0 on success and the bo was reserved.
+ *    VM_FAULT_RETRY if blocking wait.
+ *    VM_FAULT_NOPAGE if blocking wait and retrying was not allowed.
+ */
+vm_fault_t ttm_bo_vm_reserve(struct ttm_buffer_object *bo,
+			     struct vm_fault *vmf)
 {
-	struct vm_area_struct *vma = vmf->vma;
-	struct ttm_buffer_object *bo = (struct ttm_buffer_object *)
-	    vma->vm_private_data;
-	struct ttm_bo_device *bdev = bo->bdev;
-	unsigned long page_offset;
-	unsigned long page_last;
-	unsigned long pfn;
-	struct ttm_tt *ttm = NULL;
-	struct page *page;
-	int err;
-	int i;
-	vm_fault_t ret = VM_FAULT_NOPAGE;
-	unsigned long address = vmf->address;
-	struct ttm_mem_type_manager *man =
-		&bdev->man[bo->mem.mem_type];
-	struct vm_area_struct cvma;
-
 	/*
 	 * Work around locking order reversal in fault / nopfn
 	 * between mmap_sem and bo_reserve: Perform a trylock operation
@@ -151,14 +154,54 @@ static vm_fault_t ttm_bo_vm_fault(struct vm_fault *vmf)
 		return VM_FAULT_NOPAGE;
 	}
 
+	return 0;
+}
+EXPORT_SYMBOL(ttm_bo_vm_reserve);
+
+/**
+ * ttm_bo_vm_fault_reserved - TTM fault helper
+ * @vmf: The struct vm_fault given as argument to the fault callback
+ * @prot: The page protection to be used for this memory area.
+ * @num_prefault: Maximum number of prefault pages. The caller may want to
+ * specify this based on madvice settings and the size of the GPU object
+ * backed by the memory.
+ *
+ * This function inserts one or more page table entries pointing to the
+ * memory backing the buffer object, and then returns a return code
+ * instructing the caller to retry the page access.
+ *
+ * Return:
+ *   VM_FAULT_NOPAGE on success or pending signal
+ *   VM_FAULT_SIGBUS on unspecified error
+ *   VM_FAULT_OOM on out-of-memory
+ *   VM_FAULT_RETRY if retryable wait
+ */
+vm_fault_t ttm_bo_vm_fault_reserved(struct vm_fault *vmf,
+				    pgprot_t prot,
+				    pgoff_t num_prefault)
+{
+	struct vm_area_struct *vma = vmf->vma;
+	struct vm_area_struct cvma = *vma;
+	struct ttm_buffer_object *bo = vma->vm_private_data;
+	struct ttm_bo_device *bdev = bo->bdev;
+	unsigned long page_offset;
+	unsigned long page_last;
+	unsigned long pfn;
+	struct ttm_tt *ttm = NULL;
+	struct page *page;
+	int err;
+	pgoff_t i;
+	vm_fault_t ret = VM_FAULT_NOPAGE;
+	unsigned long address = vmf->address;
+	struct ttm_mem_type_manager *man =
+		&bdev->man[bo->mem.mem_type];
+
 	/*
 	 * Refuse to fault imported pages. This should be handled
 	 * (if at all) by redirecting mmap to the exporter.
 	 */
-	if (bo->ttm && (bo->ttm->page_flags & TTM_PAGE_FLAG_SG)) {
-		ret = VM_FAULT_SIGBUS;
-		goto out_unlock;
-	}
+	if (bo->ttm && (bo->ttm->page_flags & TTM_PAGE_FLAG_SG))
+		return VM_FAULT_SIGBUS;
 
 	if (bdev->driver->fault_reserve_notify) {
 		struct dma_fence *moving = dma_fence_get(bo->moving);
@@ -169,17 +212,15 @@ static vm_fault_t ttm_bo_vm_fault(struct vm_fault *vmf)
 			break;
 		case -EBUSY:
 		case -ERESTARTSYS:
-			ret = VM_FAULT_NOPAGE;
-			goto out_unlock;
+			return VM_FAULT_NOPAGE;
 		default:
-			ret = VM_FAULT_SIGBUS;
-			goto out_unlock;
+			return VM_FAULT_SIGBUS;
 		}
 
 		if (bo->moving != moving) {
-			spin_lock(&bdev->glob->lru_lock);
+			spin_lock(&ttm_bo_glob.lru_lock);
 			ttm_bo_move_to_lru_tail(bo, NULL);
-			spin_unlock(&bdev->glob->lru_lock);
+			spin_unlock(&ttm_bo_glob.lru_lock);
 		}
 		dma_fence_put(moving);
 	}
@@ -189,21 +230,12 @@ static vm_fault_t ttm_bo_vm_fault(struct vm_fault *vmf)
 	 * move.
 	 */
 	ret = ttm_bo_vm_fault_idle(bo, vmf);
-	if (unlikely(ret != 0)) {
-		if (ret == VM_FAULT_RETRY &&
-		    !(vmf->flags & FAULT_FLAG_RETRY_NOWAIT)) {
-			/* The BO has already been unreserved. */
-			return ret;
-		}
-
-		goto out_unlock;
-	}
+	if (unlikely(ret != 0))
+		return ret;
 
 	err = ttm_mem_io_lock(man, true);
-	if (unlikely(err != 0)) {
-		ret = VM_FAULT_NOPAGE;
-		goto out_unlock;
-	}
+	if (unlikely(err != 0))
+		return VM_FAULT_NOPAGE;
 	err = ttm_mem_io_reserve_vm(bo);
 	if (unlikely(err != 0)) {
 		ret = VM_FAULT_SIGBUS;
@@ -220,18 +252,8 @@ static vm_fault_t ttm_bo_vm_fault(struct vm_fault *vmf)
 		goto out_io_unlock;
 	}
 
-	/*
-	 * Make a local vma copy to modify the page_prot member
-	 * and vm_flags if necessary. The vma parameter is protected
-	 * by mmap_sem in write mode.
-	 */
-	cvma = *vma;
-	cvma.vm_page_prot = vm_get_page_prot(cvma.vm_flags);
-
-	if (bo->mem.bus.is_iomem) {
-		cvma.vm_page_prot = ttm_io_prot(bo->mem.placement,
-						cvma.vm_page_prot);
-	} else {
+	cvma.vm_page_prot = ttm_io_prot(bo->mem.placement, prot);
+	if (!bo->mem.bus.is_iomem) {
 		struct ttm_operation_ctx ctx = {
 			.interruptible = false,
 			.no_wait_gpu = false,
@@ -240,24 +262,21 @@ static vm_fault_t ttm_bo_vm_fault(struct vm_fault *vmf)
 		};
 
 		ttm = bo->ttm;
-		cvma.vm_page_prot = ttm_io_prot(bo->mem.placement,
-						cvma.vm_page_prot);
-
-		/* Allocate all page at once, most common usage */
-		if (ttm_tt_populate(ttm, &ctx)) {
+		if (ttm_tt_populate(bo->ttm, &ctx)) {
 			ret = VM_FAULT_OOM;
 			goto out_io_unlock;
 		}
+	} else {
+		/* Iomem should not be marked encrypted */
+		cvma.vm_page_prot = pgprot_decrypted(cvma.vm_page_prot);
 	}
 
 	/*
 	 * Speculatively prefault a number of pages. Only error on
 	 * first page.
 	 */
-	for (i = 0; i < TTM_BO_VM_NUM_PREFAULT; ++i) {
+	for (i = 0; i < num_prefault; ++i) {
 		if (bo->mem.bus.is_iomem) {
-			/* Iomem should not be marked encrypted */
-			cvma.vm_page_prot = pgprot_decrypted(cvma.vm_page_prot);
 			pfn = ttm_bo_io_mem_pfn(bo, page_offset);
 		} else {
 			page = ttm->pages[page_offset];
@@ -278,15 +297,13 @@ static vm_fault_t ttm_bo_vm_fault(struct vm_fault *vmf)
 		else
 			ret = vmf_insert_pfn(&cvma, address, pfn);
 
-		/*
-		 * Somebody beat us to this PTE or prefaulting to
-		 * an already populated PTE, or prefaulting error.
-		 */
-
-		if (unlikely((ret == VM_FAULT_NOPAGE && i > 0)))
-			break;
-		else if (unlikely(ret & VM_FAULT_ERROR))
-			goto out_io_unlock;
+		/* Never error on prefaulted PTEs */
+		if (unlikely((ret & VM_FAULT_ERROR))) {
+			if (i == 0)
+				goto out_io_unlock;
+			else
+				break;
+		}
 
 		address += PAGE_SIZE;
 		if (unlikely(++page_offset >= page_last))
@@ -295,28 +312,49 @@ static vm_fault_t ttm_bo_vm_fault(struct vm_fault *vmf)
 	ret = VM_FAULT_NOPAGE;
 out_io_unlock:
 	ttm_mem_io_unlock(man);
-out_unlock:
+	return ret;
+}
+EXPORT_SYMBOL(ttm_bo_vm_fault_reserved);
+
+static vm_fault_t ttm_bo_vm_fault(struct vm_fault *vmf)
+{
+	struct vm_area_struct *vma = vmf->vma;
+	pgprot_t prot;
+	struct ttm_buffer_object *bo = vma->vm_private_data;
+	vm_fault_t ret;
+
+	ret = ttm_bo_vm_reserve(bo, vmf);
+	if (ret)
+		return ret;
+
+	prot = vm_get_page_prot(vma->vm_flags);
+	ret = ttm_bo_vm_fault_reserved(vmf, prot, TTM_BO_VM_NUM_PREFAULT);
+	if (ret == VM_FAULT_RETRY && !(vmf->flags & FAULT_FLAG_RETRY_NOWAIT))
+		return ret;
+
 	dma_resv_unlock(bo->base.resv);
+
 	return ret;
 }
 
-static void ttm_bo_vm_open(struct vm_area_struct *vma)
+void ttm_bo_vm_open(struct vm_area_struct *vma)
 {
-	struct ttm_buffer_object *bo =
-	    (struct ttm_buffer_object *)vma->vm_private_data;
+	struct ttm_buffer_object *bo = vma->vm_private_data;
 
 	WARN_ON(bo->bdev->dev_mapping != vma->vm_file->f_mapping);
 
 	ttm_bo_get(bo);
 }
+EXPORT_SYMBOL(ttm_bo_vm_open);
 
-static void ttm_bo_vm_close(struct vm_area_struct *vma)
+void ttm_bo_vm_close(struct vm_area_struct *vma)
 {
-	struct ttm_buffer_object *bo = (struct ttm_buffer_object *)vma->vm_private_data;
+	struct ttm_buffer_object *bo = vma->vm_private_data;
 
 	ttm_bo_put(bo);
 	vma->vm_private_data = NULL;
 }
+EXPORT_SYMBOL(ttm_bo_vm_close);
 
 static int ttm_bo_vm_access_kmap(struct ttm_buffer_object *bo,
 				 unsigned long offset,
@@ -409,16 +447,16 @@ static struct ttm_buffer_object *ttm_bo_vm_lookup(struct ttm_bo_device *bdev,
 	struct drm_vma_offset_node *node;
 	struct ttm_buffer_object *bo = NULL;
 
-	drm_vma_offset_lock_lookup(&bdev->vma_manager);
+	drm_vma_offset_lock_lookup(bdev->vma_manager);
 
-	node = drm_vma_offset_lookup_locked(&bdev->vma_manager, offset, pages);
+	node = drm_vma_offset_lookup_locked(bdev->vma_manager, offset, pages);
 	if (likely(node)) {
 		bo = container_of(node, struct ttm_buffer_object,
 				  base.vma_node);
 		bo = ttm_bo_get_unless_zero(bo);
 	}
 
-	drm_vma_offset_unlock_lookup(&bdev->vma_manager);
+	drm_vma_offset_unlock_lookup(bdev->vma_manager);
 
 	if (!bo)
 		pr_err("Could not find buffer object to map\n");
@@ -426,6 +464,28 @@ static struct ttm_buffer_object *ttm_bo_vm_lookup(struct ttm_bo_device *bdev,
 	return bo;
 }
 
+static void ttm_bo_mmap_vma_setup(struct ttm_buffer_object *bo, struct vm_area_struct *vma)
+{
+	vma->vm_ops = &ttm_bo_vm_ops;
+
+	/*
+	 * Note: We're transferring the bo reference to
+	 * vma->vm_private_data here.
+	 */
+
+	vma->vm_private_data = bo;
+
+	/*
+	 * We'd like to use VM_PFNMAP on shared mappings, where
+	 * (vma->vm_flags & VM_SHARED) != 0, for performance reasons,
+	 * but for some reason VM_PFNMAP + x86 PAT + write-combine is very
+	 * bad for performance. Until that has been sorted out, use
+	 * VM_MIXEDMAP on all mappings. See freedesktop.org bug #75719
+	 */
+	vma->vm_flags |= VM_MIXEDMAP;
+	vma->vm_flags |= VM_IO | VM_DONTEXPAND | VM_DONTDUMP;
+}
+
 int ttm_bo_mmap(struct file *filp, struct vm_area_struct *vma,
 		struct ttm_bo_device *bdev)
 {
@@ -449,24 +509,7 @@ int ttm_bo_mmap(struct file *filp, struct vm_area_struct *vma,
 	if (unlikely(ret != 0))
 		goto out_unref;
 
-	vma->vm_ops = &ttm_bo_vm_ops;
-
-	/*
-	 * Note: We're transferring the bo reference to
-	 * vma->vm_private_data here.
-	 */
-
-	vma->vm_private_data = bo;
-
-	/*
-	 * We'd like to use VM_PFNMAP on shared mappings, where
-	 * (vma->vm_flags & VM_SHARED) != 0, for performance reasons,
-	 * but for some reason VM_PFNMAP + x86 PAT + write-combine is very
-	 * bad for performance. Until that has been sorted out, use
-	 * VM_MIXEDMAP on all mappings. See freedesktop.org bug #75719
-	 */
-	vma->vm_flags |= VM_MIXEDMAP;
-	vma->vm_flags |= VM_IO | VM_DONTEXPAND | VM_DONTDUMP;
+	ttm_bo_mmap_vma_setup(bo, vma);
 	return 0;
 out_unref:
 	ttm_bo_put(bo);
@@ -474,17 +517,17 @@ out_unref:
 }
 EXPORT_SYMBOL(ttm_bo_mmap);
 
-int ttm_fbdev_mmap(struct vm_area_struct *vma, struct ttm_buffer_object *bo)
+int ttm_bo_mmap_obj(struct vm_area_struct *vma, struct ttm_buffer_object *bo)
 {
-	if (vma->vm_pgoff != 0)
-		return -EACCES;
-
 	ttm_bo_get(bo);
 
-	vma->vm_ops = &ttm_bo_vm_ops;
-	vma->vm_private_data = bo;
-	vma->vm_flags |= VM_MIXEDMAP;
-	vma->vm_flags |= VM_IO | VM_DONTEXPAND;
+	/*
+	 * FIXME: &drm_gem_object_funcs.mmap is called with the fake offset
+	 * removed. Add it back here until the rest of TTM works without it.
+	 */
+	vma->vm_pgoff += drm_vma_node_start(&bo->base.vma_node);
+
+	ttm_bo_mmap_vma_setup(bo, vma);
 	return 0;
 }
-EXPORT_SYMBOL(ttm_fbdev_mmap);
+EXPORT_SYMBOL(ttm_bo_mmap_obj);
diff --git a/drivers/gpu/drm/ttm/ttm_execbuf_util.c b/drivers/gpu/drm/ttm/ttm_execbuf_util.c
index 131dae8f4170..1797f04c0534 100644
--- a/drivers/gpu/drm/ttm/ttm_execbuf_util.c
+++ b/drivers/gpu/drm/ttm/ttm_execbuf_util.c
@@ -43,37 +43,22 @@ static void ttm_eu_backoff_reservation_reverse(struct list_head *list,
 	}
 }
 
-static void ttm_eu_del_from_lru_locked(struct list_head *list)
-{
-	struct ttm_validate_buffer *entry;
-
-	list_for_each_entry(entry, list, head) {
-		struct ttm_buffer_object *bo = entry->bo;
-		ttm_bo_del_from_lru(bo);
-	}
-}
-
 void ttm_eu_backoff_reservation(struct ww_acquire_ctx *ticket,
 				struct list_head *list)
 {
 	struct ttm_validate_buffer *entry;
-	struct ttm_bo_global *glob;
 
 	if (list_empty(list))
 		return;
 
-	entry = list_first_entry(list, struct ttm_validate_buffer, head);
-	glob = entry->bo->bdev->glob;
-
-	spin_lock(&glob->lru_lock);
+	spin_lock(&ttm_bo_glob.lru_lock);
 	list_for_each_entry(entry, list, head) {
 		struct ttm_buffer_object *bo = entry->bo;
 
-		if (list_empty(&bo->lru))
-			ttm_bo_add_to_lru(bo);
+		ttm_bo_move_to_lru_tail(bo, NULL);
 		dma_resv_unlock(bo->base.resv);
 	}
-	spin_unlock(&glob->lru_lock);
+	spin_unlock(&ttm_bo_glob.lru_lock);
 
 	if (ticket)
 		ww_acquire_fini(ticket);
@@ -94,18 +79,14 @@ EXPORT_SYMBOL(ttm_eu_backoff_reservation);
 
 int ttm_eu_reserve_buffers(struct ww_acquire_ctx *ticket,
 			   struct list_head *list, bool intr,
-			   struct list_head *dups, bool del_lru)
+			   struct list_head *dups)
 {
-	struct ttm_bo_global *glob;
 	struct ttm_validate_buffer *entry;
 	int ret;
 
 	if (list_empty(list))
 		return 0;
 
-	entry = list_first_entry(list, struct ttm_validate_buffer, head);
-	glob = entry->bo->bdev->glob;
-
 	if (ticket)
 		ww_acquire_init(ticket, &reservation_ww_class);
 
@@ -113,12 +94,7 @@ int ttm_eu_reserve_buffers(struct ww_acquire_ctx *ticket,
 		struct ttm_buffer_object *bo = entry->bo;
 
 		ret = __ttm_bo_reserve(bo, intr, (ticket == NULL), ticket);
-		if (!ret && unlikely(atomic_read(&bo->cpu_writers) > 0)) {
-			dma_resv_unlock(bo->base.resv);
-
-			ret = -EBUSY;
-
-		} else if (ret == -EALREADY && dups) {
+		if (ret == -EALREADY && dups) {
 			struct ttm_validate_buffer *safe = entry;
 			entry = list_prev_entry(entry, head);
 			list_del(&safe->head);
@@ -173,11 +149,6 @@ int ttm_eu_reserve_buffers(struct ww_acquire_ctx *ticket,
 		list_add(&entry->head, list);
 	}
 
-	if (del_lru) {
-		spin_lock(&glob->lru_lock);
-		ttm_eu_del_from_lru_locked(list);
-		spin_unlock(&glob->lru_lock);
-	}
 	return 0;
 }
 EXPORT_SYMBOL(ttm_eu_reserve_buffers);
@@ -187,30 +158,22 @@ void ttm_eu_fence_buffer_objects(struct ww_acquire_ctx *ticket,
 				 struct dma_fence *fence)
 {
 	struct ttm_validate_buffer *entry;
-	struct ttm_buffer_object *bo;
-	struct ttm_bo_global *glob;
 
 	if (list_empty(list))
 		return;
 
-	bo = list_first_entry(list, struct ttm_validate_buffer, head)->bo;
-	glob = bo->bdev->glob;
-
-	spin_lock(&glob->lru_lock);
-
+	spin_lock(&ttm_bo_glob.lru_lock);
 	list_for_each_entry(entry, list, head) {
-		bo = entry->bo;
+		struct ttm_buffer_object *bo = entry->bo;
+
 		if (entry->num_shared)
 			dma_resv_add_shared_fence(bo->base.resv, fence);
 		else
 			dma_resv_add_excl_fence(bo->base.resv, fence);
-		if (list_empty(&bo->lru))
-			ttm_bo_add_to_lru(bo);
-		else
-			ttm_bo_move_to_lru_tail(bo, NULL);
+		ttm_bo_move_to_lru_tail(bo, NULL);
 		dma_resv_unlock(bo->base.resv);
 	}
-	spin_unlock(&glob->lru_lock);
+	spin_unlock(&ttm_bo_glob.lru_lock);
 	if (ticket)
 		ww_acquire_fini(ticket);
 }
diff --git a/drivers/gpu/drm/ttm/ttm_memory.c b/drivers/gpu/drm/ttm/ttm_memory.c
index 8617958b7ae6..acd63b70d814 100644
--- a/drivers/gpu/drm/ttm/ttm_memory.c
+++ b/drivers/gpu/drm/ttm/ttm_memory.c
@@ -275,7 +275,7 @@ static void ttm_shrink(struct ttm_mem_global *glob, bool from_wq,
 
 	while (ttm_zones_above_swap_target(glob, from_wq, extra)) {
 		spin_unlock(&glob->lock);
-		ret = ttm_bo_swapout(glob->bo_glob, ctx);
+		ret = ttm_bo_swapout(&ttm_bo_glob, ctx);
 		spin_lock(&glob->lock);
 		if (unlikely(ret != 0))
 			break;
diff --git a/drivers/gpu/drm/ttm/ttm_page_alloc.c b/drivers/gpu/drm/ttm/ttm_page_alloc.c
index 627f8dc91d0e..b40a4678c296 100644
--- a/drivers/gpu/drm/ttm/ttm_page_alloc.c
+++ b/drivers/gpu/drm/ttm/ttm_page_alloc.c
@@ -1028,7 +1028,7 @@ void ttm_page_alloc_fini(void)
 static void
 ttm_pool_unpopulate_helper(struct ttm_tt *ttm, unsigned mem_count_update)
 {
-	struct ttm_mem_global *mem_glob = ttm->bdev->glob->mem_glob;
+	struct ttm_mem_global *mem_glob = &ttm_mem_glob;
 	unsigned i;
 
 	if (mem_count_update == 0)
@@ -1049,7 +1049,7 @@ put_pages:
 
 int ttm_pool_populate(struct ttm_tt *ttm, struct ttm_operation_ctx *ctx)
 {
-	struct ttm_mem_global *mem_glob = ttm->bdev->glob->mem_glob;
+	struct ttm_mem_global *mem_glob = &ttm_mem_glob;
 	unsigned i;
 	int ret;
 
diff --git a/drivers/gpu/drm/ttm/ttm_page_alloc_dma.c b/drivers/gpu/drm/ttm/ttm_page_alloc_dma.c
index 7d78e6deac89..bf876faea592 100644
--- a/drivers/gpu/drm/ttm/ttm_page_alloc_dma.c
+++ b/drivers/gpu/drm/ttm/ttm_page_alloc_dma.c
@@ -33,7 +33,6 @@
  *   when freed).
  */
 
-#if defined(CONFIG_SWIOTLB) || defined(CONFIG_INTEL_IOMMU)
 #define pr_fmt(fmt) "[TTM] " fmt
 
 #include <linux/dma-mapping.h>
@@ -886,8 +885,8 @@ static gfp_t ttm_dma_pool_gfp_flags(struct ttm_dma_tt *ttm_dma, bool huge)
 int ttm_dma_populate(struct ttm_dma_tt *ttm_dma, struct device *dev,
 			struct ttm_operation_ctx *ctx)
 {
+	struct ttm_mem_global *mem_glob = &ttm_mem_glob;
 	struct ttm_tt *ttm = &ttm_dma->ttm;
-	struct ttm_mem_global *mem_glob = ttm->bdev->glob->mem_glob;
 	unsigned long num_pages = ttm->num_pages;
 	struct dma_pool *pool;
 	struct dma_page *d_page;
@@ -991,8 +990,8 @@ EXPORT_SYMBOL_GPL(ttm_dma_populate);
 /* Put all pages in pages list to correct pool to wait for reuse */
 void ttm_dma_unpopulate(struct ttm_dma_tt *ttm_dma, struct device *dev)
 {
+	struct ttm_mem_global *mem_glob = &ttm_mem_glob;
 	struct ttm_tt *ttm = &ttm_dma->ttm;
-	struct ttm_mem_global *mem_glob = ttm->bdev->glob->mem_glob;
 	struct dma_pool *pool;
 	struct dma_page *d_page, *next;
 	enum pool_type type;
@@ -1238,5 +1237,3 @@ int ttm_dma_page_alloc_debugfs(struct seq_file *m, void *data)
 	return 0;
 }
 EXPORT_SYMBOL_GPL(ttm_dma_page_alloc_debugfs);
-
-#endif
diff --git a/drivers/gpu/drm/tve200/tve200_drv.c b/drivers/gpu/drm/tve200/tve200_drv.c
index 416f24823c0a..954b09c948eb 100644
--- a/drivers/gpu/drm/tve200/tve200_drv.c
+++ b/drivers/gpu/drm/tve200/tve200_drv.c
@@ -80,8 +80,8 @@ static int tve200_modeset_init(struct drm_device *dev)
 	if (ret && ret != -ENODEV)
 		return ret;
 	if (panel) {
-		bridge = drm_panel_bridge_add(panel,
-					      DRM_MODE_CONNECTOR_Unknown);
+		bridge = drm_panel_bridge_add_typed(panel,
+						    DRM_MODE_CONNECTOR_Unknown);
 		if (IS_ERR(bridge)) {
 			ret = PTR_ERR(bridge);
 			goto out_bridge;
diff --git a/drivers/gpu/drm/udl/udl_connector.c b/drivers/gpu/drm/udl/udl_connector.c
index ddb61a60c610..b4ae3e89a7b4 100644
--- a/drivers/gpu/drm/udl/udl_connector.c
+++ b/drivers/gpu/drm/udl/udl_connector.c
@@ -90,13 +90,6 @@ udl_detect(struct drm_connector *connector, bool force)
 	return connector_status_connected;
 }
 
-static struct drm_encoder*
-udl_best_single_encoder(struct drm_connector *connector)
-{
-	int enc_id = connector->encoder_ids[0];
-	return drm_encoder_find(connector->dev, NULL, enc_id);
-}
-
 static int udl_connector_set_property(struct drm_connector *connector,
 				      struct drm_property *property,
 				      uint64_t val)
@@ -120,7 +113,6 @@ static void udl_connector_destroy(struct drm_connector *connector)
 static const struct drm_connector_helper_funcs udl_connector_helper_funcs = {
 	.get_modes = udl_get_modes,
 	.mode_valid = udl_mode_valid,
-	.best_encoder = udl_best_single_encoder,
 };
 
 static const struct drm_connector_funcs udl_connector_funcs = {
diff --git a/drivers/gpu/drm/v3d/v3d_bo.c b/drivers/gpu/drm/v3d/v3d_bo.c
index a22b75a3a533..edd299ab53d8 100644
--- a/drivers/gpu/drm/v3d/v3d_bo.c
+++ b/drivers/gpu/drm/v3d/v3d_bo.c
@@ -58,7 +58,7 @@ static const struct drm_gem_object_funcs v3d_gem_funcs = {
 	.get_sg_table = drm_gem_shmem_get_sg_table,
 	.vmap = drm_gem_shmem_vmap,
 	.vunmap = drm_gem_shmem_vunmap,
-	.vm_ops = &drm_gem_shmem_vm_ops,
+	.mmap = drm_gem_shmem_mmap,
 };
 
 /* gem_create_object function for allocating a BO struct and doing
diff --git a/drivers/gpu/drm/v3d/v3d_drv.c b/drivers/gpu/drm/v3d/v3d_drv.c
index 3506ae2723ae..1a07462b4528 100644
--- a/drivers/gpu/drm/v3d/v3d_drv.c
+++ b/drivers/gpu/drm/v3d/v3d_drv.c
@@ -126,6 +126,9 @@ static int v3d_get_param_ioctl(struct drm_device *dev, void *data,
 	case DRM_V3D_PARAM_SUPPORTS_CSD:
 		args->value = v3d_has_csd(v3d);
 		return 0;
+	case DRM_V3D_PARAM_SUPPORTS_CACHE_FLUSH:
+		args->value = 1;
+		return 0;
 	default:
 		DRM_DEBUG("Unknown parameter %d\n", args->param);
 		return -EINVAL;
@@ -169,7 +172,7 @@ v3d_postclose(struct drm_device *dev, struct drm_file *file)
 	kfree(v3d_priv);
 }
 
-DEFINE_DRM_GEM_SHMEM_FOPS(v3d_drm_fops);
+DEFINE_DRM_GEM_FOPS(v3d_drm_fops);
 
 /* DRM_AUTH is required on SUBMIT_CL for now, while we don't have GMP
  * protection between clients.  Note that render nodes would be be
diff --git a/drivers/gpu/drm/v3d/v3d_gem.c b/drivers/gpu/drm/v3d/v3d_gem.c
index 5d80507b539b..549dde83408b 100644
--- a/drivers/gpu/drm/v3d/v3d_gem.c
+++ b/drivers/gpu/drm/v3d/v3d_gem.c
@@ -530,13 +530,16 @@ v3d_submit_cl_ioctl(struct drm_device *dev, void *data,
 	struct drm_v3d_submit_cl *args = data;
 	struct v3d_bin_job *bin = NULL;
 	struct v3d_render_job *render;
+	struct v3d_job *clean_job = NULL;
+	struct v3d_job *last_job;
 	struct ww_acquire_ctx acquire_ctx;
 	int ret = 0;
 
 	trace_v3d_submit_cl_ioctl(&v3d->drm, args->rcl_start, args->rcl_end);
 
-	if (args->pad != 0) {
-		DRM_INFO("pad must be zero: %d\n", args->pad);
+	if (args->flags != 0 &&
+	    args->flags != DRM_V3D_SUBMIT_CL_FLUSH_CACHE) {
+		DRM_INFO("invalid flags: %d\n", args->flags);
 		return -EINVAL;
 	}
 
@@ -557,13 +560,17 @@ v3d_submit_cl_ioctl(struct drm_device *dev, void *data,
 
 	if (args->bcl_start != args->bcl_end) {
 		bin = kcalloc(1, sizeof(*bin), GFP_KERNEL);
-		if (!bin)
+		if (!bin) {
+			v3d_job_put(&render->base);
 			return -ENOMEM;
+		}
 
 		ret = v3d_job_init(v3d, file_priv, &bin->base,
 				   v3d_job_free, args->in_sync_bcl);
 		if (ret) {
+			kfree(bin);
 			v3d_job_put(&render->base);
+			kfree(bin);
 			return ret;
 		}
 
@@ -575,12 +582,31 @@ v3d_submit_cl_ioctl(struct drm_device *dev, void *data,
 		bin->render = render;
 	}
 
-	ret = v3d_lookup_bos(dev, file_priv, &render->base,
+	if (args->flags & DRM_V3D_SUBMIT_CL_FLUSH_CACHE) {
+		clean_job = kcalloc(1, sizeof(*clean_job), GFP_KERNEL);
+		if (!clean_job) {
+			ret = -ENOMEM;
+			goto fail;
+		}
+
+		ret = v3d_job_init(v3d, file_priv, clean_job, v3d_job_free, 0);
+		if (ret) {
+			kfree(clean_job);
+			clean_job = NULL;
+			goto fail;
+		}
+
+		last_job = clean_job;
+	} else {
+		last_job = &render->base;
+	}
+
+	ret = v3d_lookup_bos(dev, file_priv, last_job,
 			     args->bo_handles, args->bo_handle_count);
 	if (ret)
 		goto fail;
 
-	ret = v3d_lock_bo_reservations(&render->base, &acquire_ctx);
+	ret = v3d_lock_bo_reservations(last_job, &acquire_ctx);
 	if (ret)
 		goto fail;
 
@@ -599,28 +625,44 @@ v3d_submit_cl_ioctl(struct drm_device *dev, void *data,
 	ret = v3d_push_job(v3d_priv, &render->base, V3D_RENDER);
 	if (ret)
 		goto fail_unreserve;
+
+	if (clean_job) {
+		struct dma_fence *render_fence =
+			dma_fence_get(render->base.done_fence);
+		ret = drm_gem_fence_array_add(&clean_job->deps, render_fence);
+		if (ret)
+			goto fail_unreserve;
+		ret = v3d_push_job(v3d_priv, clean_job, V3D_CACHE_CLEAN);
+		if (ret)
+			goto fail_unreserve;
+	}
+
 	mutex_unlock(&v3d->sched_lock);
 
 	v3d_attach_fences_and_unlock_reservation(file_priv,
-						 &render->base,
+						 last_job,
 						 &acquire_ctx,
 						 args->out_sync,
-						 render->base.done_fence);
+						 last_job->done_fence);
 
 	if (bin)
 		v3d_job_put(&bin->base);
 	v3d_job_put(&render->base);
+	if (clean_job)
+		v3d_job_put(clean_job);
 
 	return 0;
 
 fail_unreserve:
 	mutex_unlock(&v3d->sched_lock);
-	drm_gem_unlock_reservations(render->base.bo,
-				    render->base.bo_count, &acquire_ctx);
+	drm_gem_unlock_reservations(last_job->bo,
+				    last_job->bo_count, &acquire_ctx);
 fail:
 	if (bin)
 		v3d_job_put(&bin->base);
 	v3d_job_put(&render->base);
+	if (clean_job)
+		v3d_job_put(clean_job);
 
 	return ret;
 }
diff --git a/drivers/gpu/drm/vboxvideo/Kconfig b/drivers/gpu/drm/vboxvideo/Kconfig
index 56ba510f21a2..45fe135d6e43 100644
--- a/drivers/gpu/drm/vboxvideo/Kconfig
+++ b/drivers/gpu/drm/vboxvideo/Kconfig
@@ -4,6 +4,8 @@ config DRM_VBOXVIDEO
 	depends on DRM && X86 && PCI
 	select DRM_KMS_HELPER
 	select DRM_VRAM_HELPER
+	select DRM_TTM
+	select DRM_TTM_HELPER
 	select GENERIC_ALLOCATOR
 	help
 	  This is a KMS driver for the virtual Graphics Card used in
diff --git a/drivers/gpu/drm/vboxvideo/Makefile b/drivers/gpu/drm/vboxvideo/Makefile
index 55d798c76b21..f2e968b5ffa6 100644
--- a/drivers/gpu/drm/vboxvideo/Makefile
+++ b/drivers/gpu/drm/vboxvideo/Makefile
@@ -1,6 +1,6 @@
 # SPDX-License-Identifier: GPL-2.0
 vboxvideo-y :=  hgsmi_base.o modesetting.o vbva_base.o \
-		vbox_drv.o vbox_fb.o vbox_hgsmi.o vbox_irq.o vbox_main.o \
+		vbox_drv.o vbox_hgsmi.o vbox_irq.o vbox_main.o \
 		vbox_mode.o vbox_ttm.o
 
 obj-$(CONFIG_DRM_VBOXVIDEO) += vboxvideo.o
diff --git a/drivers/gpu/drm/vboxvideo/vbox_drv.c b/drivers/gpu/drm/vboxvideo/vbox_drv.c
index 862db495d111..8512d970a09f 100644
--- a/drivers/gpu/drm/vboxvideo/vbox_drv.c
+++ b/drivers/gpu/drm/vboxvideo/vbox_drv.c
@@ -14,6 +14,7 @@
 
 #include <drm/drm_crtc_helper.h>
 #include <drm/drm_drv.h>
+#include <drm/drm_fb_helper.h>
 #include <drm/drm_file.h>
 #include <drm/drm_ioctl.h>
 
@@ -32,10 +33,6 @@ static const struct pci_device_id pciidlist[] = {
 };
 MODULE_DEVICE_TABLE(pci, pciidlist);
 
-static const struct drm_fb_helper_funcs vbox_fb_helper_funcs = {
-	.fb_probe = vboxfb_create,
-};
-
 static int vbox_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 {
 	struct vbox_private *vbox;
@@ -79,20 +76,16 @@ static int vbox_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 	if (ret)
 		goto err_mode_fini;
 
-	ret = drm_fb_helper_fbdev_setup(&vbox->ddev, &vbox->fb_helper,
-					&vbox_fb_helper_funcs, 32,
-					vbox->num_crtcs);
+	ret = drm_fbdev_generic_setup(&vbox->ddev, 32);
 	if (ret)
 		goto err_irq_fini;
 
 	ret = drm_dev_register(&vbox->ddev, 0);
 	if (ret)
-		goto err_fbdev_fini;
+		goto err_irq_fini;
 
 	return 0;
 
-err_fbdev_fini:
-	vbox_fbdev_fini(vbox);
 err_irq_fini:
 	vbox_irq_fini(vbox);
 err_mode_fini:
@@ -113,7 +106,6 @@ static void vbox_pci_remove(struct pci_dev *pdev)
 	struct vbox_private *vbox = pci_get_drvdata(pdev);
 
 	drm_dev_unregister(&vbox->ddev);
-	vbox_fbdev_fini(vbox);
 	vbox_irq_fini(vbox);
 	vbox_mode_fini(vbox);
 	vbox_mm_fini(vbox);
@@ -189,10 +181,7 @@ static struct pci_driver vbox_pci_driver = {
 #endif
 };
 
-static const struct file_operations vbox_fops = {
-	.owner = THIS_MODULE,
-	DRM_VRAM_MM_FILE_OPERATIONS
-};
+DEFINE_DRM_GEM_FOPS(vbox_fops);
 
 static struct drm_driver driver = {
 	.driver_features =
diff --git a/drivers/gpu/drm/vboxvideo/vbox_drv.h b/drivers/gpu/drm/vboxvideo/vbox_drv.h
index e8cb9efc6088..87421903816c 100644
--- a/drivers/gpu/drm/vboxvideo/vbox_drv.h
+++ b/drivers/gpu/drm/vboxvideo/vbox_drv.h
@@ -16,12 +16,9 @@
 #include <linux/string.h>
 
 #include <drm/drm_encoder.h>
-#include <drm/drm_fb_helper.h>
 #include <drm/drm_gem.h>
 #include <drm/drm_gem_vram_helper.h>
 
-#include <drm/drm_vram_mm_helper.h>
-
 #include "vboxvideo_guest.h"
 #include "vboxvideo_vbe.h"
 #include "hgsmi_ch_setup.h"
@@ -48,16 +45,9 @@
 				sizeof(struct hgsmi_host_flags))
 #define HOST_FLAGS_OFFSET GUEST_HEAP_USABLE_SIZE
 
-struct vbox_framebuffer {
-	struct drm_framebuffer base;
-	struct drm_gem_object *obj;
-};
-
 struct vbox_private {
 	/* Must be first; or we must define our own release callback */
 	struct drm_device ddev;
-	struct drm_fb_helper fb_helper;
-	struct vbox_framebuffer afb;
 
 	u8 __iomem *guest_heap;
 	u8 __iomem *vbva_buffers;
@@ -137,7 +127,6 @@ struct vbox_encoder {
 #define to_vbox_crtc(x) container_of(x, struct vbox_crtc, base)
 #define to_vbox_connector(x) container_of(x, struct vbox_connector, base)
 #define to_vbox_encoder(x) container_of(x, struct vbox_encoder, base)
-#define to_vbox_framebuffer(x) container_of(x, struct vbox_framebuffer, base)
 
 bool vbox_check_supported(u16 id);
 int vbox_hw_init(struct vbox_private *vbox);
@@ -148,25 +137,9 @@ void vbox_mode_fini(struct vbox_private *vbox);
 
 void vbox_report_caps(struct vbox_private *vbox);
 
-void vbox_framebuffer_dirty_rectangles(struct drm_framebuffer *fb,
-				       struct drm_clip_rect *rects,
-				       unsigned int num_rects);
-
-int vbox_framebuffer_init(struct vbox_private *vbox,
-			  struct vbox_framebuffer *vbox_fb,
-			  const struct drm_mode_fb_cmd2 *mode_cmd,
-			  struct drm_gem_object *obj);
-
-int vboxfb_create(struct drm_fb_helper *helper,
-		  struct drm_fb_helper_surface_size *sizes);
-void vbox_fbdev_fini(struct vbox_private *vbox);
-
 int vbox_mm_init(struct vbox_private *vbox);
 void vbox_mm_fini(struct vbox_private *vbox);
 
-int vbox_gem_create(struct vbox_private *vbox,
-		    u32 size, bool iskernel, struct drm_gem_object **obj);
-
 /* vbox_irq.c */
 int vbox_irq_init(struct vbox_private *vbox);
 void vbox_irq_fini(struct vbox_private *vbox);
diff --git a/drivers/gpu/drm/vboxvideo/vbox_fb.c b/drivers/gpu/drm/vboxvideo/vbox_fb.c
deleted file mode 100644
index 8f74bcffc034..000000000000
--- a/drivers/gpu/drm/vboxvideo/vbox_fb.c
+++ /dev/null
@@ -1,149 +0,0 @@
-// SPDX-License-Identifier: MIT
-/*
- * Copyright (C) 2013-2017 Oracle Corporation
- * This file is based on ast_fb.c
- * Copyright 2012 Red Hat Inc.
- * Authors: Dave Airlie <airlied@redhat.com>
- *          Michael Thayer <michael.thayer@oracle.com,
- */
-#include <linux/delay.h>
-#include <linux/errno.h>
-#include <linux/fb.h>
-#include <linux/init.h>
-#include <linux/kernel.h>
-#include <linux/mm.h>
-#include <linux/module.h>
-#include <linux/pci.h>
-#include <linux/string.h>
-#include <linux/sysrq.h>
-#include <linux/tty.h>
-
-#include <drm/drm_crtc.h>
-#include <drm/drm_crtc_helper.h>
-#include <drm/drm_fb_helper.h>
-#include <drm/drm_fourcc.h>
-
-#include "vbox_drv.h"
-#include "vboxvideo.h"
-
-#ifdef CONFIG_DRM_KMS_FB_HELPER
-static struct fb_deferred_io vbox_defio = {
-	.delay = HZ / 30,
-	.deferred_io = drm_fb_helper_deferred_io,
-};
-#endif
-
-static struct fb_ops vboxfb_ops = {
-	.owner = THIS_MODULE,
-	DRM_FB_HELPER_DEFAULT_OPS,
-	.fb_fillrect = drm_fb_helper_sys_fillrect,
-	.fb_copyarea = drm_fb_helper_sys_copyarea,
-	.fb_imageblit = drm_fb_helper_sys_imageblit,
-};
-
-int vboxfb_create(struct drm_fb_helper *helper,
-		  struct drm_fb_helper_surface_size *sizes)
-{
-	struct vbox_private *vbox =
-		container_of(helper, struct vbox_private, fb_helper);
-	struct pci_dev *pdev = vbox->ddev.pdev;
-	struct drm_mode_fb_cmd2 mode_cmd;
-	struct drm_framebuffer *fb;
-	struct fb_info *info;
-	struct drm_gem_object *gobj;
-	struct drm_gem_vram_object *gbo;
-	int size, ret;
-	s64 gpu_addr;
-	u32 pitch;
-
-	mode_cmd.width = sizes->surface_width;
-	mode_cmd.height = sizes->surface_height;
-	pitch = mode_cmd.width * ((sizes->surface_bpp + 7) / 8);
-	mode_cmd.pixel_format = drm_mode_legacy_fb_format(sizes->surface_bpp,
-							  sizes->surface_depth);
-	mode_cmd.pitches[0] = pitch;
-
-	size = pitch * mode_cmd.height;
-
-	ret = vbox_gem_create(vbox, size, true, &gobj);
-	if (ret) {
-		DRM_ERROR("failed to create fbcon backing object %d\n", ret);
-		return ret;
-	}
-
-	ret = vbox_framebuffer_init(vbox, &vbox->afb, &mode_cmd, gobj);
-	if (ret)
-		return ret;
-
-	gbo = drm_gem_vram_of_gem(gobj);
-
-	ret = drm_gem_vram_pin(gbo, DRM_GEM_VRAM_PL_FLAG_VRAM);
-	if (ret)
-		return ret;
-
-	info = drm_fb_helper_alloc_fbi(helper);
-	if (IS_ERR(info))
-		return PTR_ERR(info);
-
-	info->screen_size = size;
-	info->screen_base = (char __iomem *)drm_gem_vram_kmap(gbo, true, NULL);
-	if (IS_ERR(info->screen_base))
-		return PTR_ERR(info->screen_base);
-
-	fb = &vbox->afb.base;
-	helper->fb = fb;
-
-	info->fbops = &vboxfb_ops;
-
-	/*
-	 * This seems to be done for safety checking that the framebuffer
-	 * is not registered twice by different drivers.
-	 */
-	info->apertures->ranges[0].base = pci_resource_start(pdev, 0);
-	info->apertures->ranges[0].size = pci_resource_len(pdev, 0);
-
-	drm_fb_helper_fill_info(info, helper, sizes);
-
-	gpu_addr = drm_gem_vram_offset(gbo);
-	if (gpu_addr < 0)
-		return (int)gpu_addr;
-	info->fix.smem_start = info->apertures->ranges[0].base + gpu_addr;
-	info->fix.smem_len = vbox->available_vram_size - gpu_addr;
-
-#ifdef CONFIG_DRM_KMS_FB_HELPER
-	info->fbdefio = &vbox_defio;
-	fb_deferred_io_init(info);
-#endif
-
-	info->pixmap.flags = FB_PIXMAP_SYSTEM;
-
-	DRM_DEBUG_KMS("allocated %dx%d\n", fb->width, fb->height);
-
-	return 0;
-}
-
-void vbox_fbdev_fini(struct vbox_private *vbox)
-{
-	struct vbox_framebuffer *afb = &vbox->afb;
-
-#ifdef CONFIG_DRM_KMS_FB_HELPER
-	if (vbox->fb_helper.fbdev && vbox->fb_helper.fbdev->fbdefio)
-		fb_deferred_io_cleanup(vbox->fb_helper.fbdev);
-#endif
-
-	drm_fb_helper_unregister_fbi(&vbox->fb_helper);
-
-	if (afb->obj) {
-		struct drm_gem_vram_object *gbo = drm_gem_vram_of_gem(afb->obj);
-
-		drm_gem_vram_kunmap(gbo);
-		drm_gem_vram_unpin(gbo);
-
-		drm_gem_object_put_unlocked(afb->obj);
-		afb->obj = NULL;
-	}
-	drm_fb_helper_fini(&vbox->fb_helper);
-
-	drm_framebuffer_unregister_private(&afb->base);
-	drm_framebuffer_cleanup(&afb->base);
-}
diff --git a/drivers/gpu/drm/vboxvideo/vbox_main.c b/drivers/gpu/drm/vboxvideo/vbox_main.c
index 02fa8277ff1e..9dcab115a261 100644
--- a/drivers/gpu/drm/vboxvideo/vbox_main.c
+++ b/drivers/gpu/drm/vboxvideo/vbox_main.c
@@ -11,22 +11,12 @@
 #include <linux/vbox_err.h>
 #include <drm/drm_fb_helper.h>
 #include <drm/drm_crtc_helper.h>
+#include <drm/drm_damage_helper.h>
 
 #include "vbox_drv.h"
 #include "vboxvideo_guest.h"
 #include "vboxvideo_vbe.h"
 
-static void vbox_user_framebuffer_destroy(struct drm_framebuffer *fb)
-{
-	struct vbox_framebuffer *vbox_fb = to_vbox_framebuffer(fb);
-
-	if (vbox_fb->obj)
-		drm_gem_object_put_unlocked(vbox_fb->obj);
-
-	drm_framebuffer_cleanup(fb);
-	kfree(fb);
-}
-
 void vbox_report_caps(struct vbox_private *vbox)
 {
 	u32 caps = VBVACAPS_DISABLE_CURSOR_INTEGRATION |
@@ -38,87 +28,6 @@ void vbox_report_caps(struct vbox_private *vbox)
 	hgsmi_send_caps_info(vbox->guest_pool, caps);
 }
 
-/* Send information about dirty rectangles to VBVA. */
-void vbox_framebuffer_dirty_rectangles(struct drm_framebuffer *fb,
-				       struct drm_clip_rect *rects,
-				       unsigned int num_rects)
-{
-	struct vbox_private *vbox = fb->dev->dev_private;
-	struct drm_display_mode *mode;
-	struct drm_crtc *crtc;
-	int crtc_x, crtc_y;
-	unsigned int i;
-
-	mutex_lock(&vbox->hw_mutex);
-	list_for_each_entry(crtc, &fb->dev->mode_config.crtc_list, head) {
-		if (crtc->primary->state->fb != fb)
-			continue;
-
-		mode = &crtc->state->mode;
-		crtc_x = crtc->primary->state->src_x >> 16;
-		crtc_y = crtc->primary->state->src_y >> 16;
-
-		for (i = 0; i < num_rects; ++i) {
-			struct vbva_cmd_hdr cmd_hdr;
-			unsigned int crtc_id = to_vbox_crtc(crtc)->crtc_id;
-
-			if (rects[i].x1 > crtc_x + mode->hdisplay ||
-			    rects[i].y1 > crtc_y + mode->vdisplay ||
-			    rects[i].x2 < crtc_x ||
-			    rects[i].y2 < crtc_y)
-				continue;
-
-			cmd_hdr.x = (s16)rects[i].x1;
-			cmd_hdr.y = (s16)rects[i].y1;
-			cmd_hdr.w = (u16)rects[i].x2 - rects[i].x1;
-			cmd_hdr.h = (u16)rects[i].y2 - rects[i].y1;
-
-			if (!vbva_buffer_begin_update(&vbox->vbva_info[crtc_id],
-						      vbox->guest_pool))
-				continue;
-
-			vbva_write(&vbox->vbva_info[crtc_id], vbox->guest_pool,
-				   &cmd_hdr, sizeof(cmd_hdr));
-			vbva_buffer_end_update(&vbox->vbva_info[crtc_id]);
-		}
-	}
-	mutex_unlock(&vbox->hw_mutex);
-}
-
-static int vbox_user_framebuffer_dirty(struct drm_framebuffer *fb,
-				       struct drm_file *file_priv,
-				       unsigned int flags, unsigned int color,
-				       struct drm_clip_rect *rects,
-				       unsigned int num_rects)
-{
-	vbox_framebuffer_dirty_rectangles(fb, rects, num_rects);
-
-	return 0;
-}
-
-static const struct drm_framebuffer_funcs vbox_fb_funcs = {
-	.destroy = vbox_user_framebuffer_destroy,
-	.dirty = vbox_user_framebuffer_dirty,
-};
-
-int vbox_framebuffer_init(struct vbox_private *vbox,
-			  struct vbox_framebuffer *vbox_fb,
-			  const struct drm_mode_fb_cmd2 *mode_cmd,
-			  struct drm_gem_object *obj)
-{
-	int ret;
-
-	drm_helper_mode_fill_fb_struct(&vbox->ddev, &vbox_fb->base, mode_cmd);
-	vbox_fb->obj = obj;
-	ret = drm_framebuffer_init(&vbox->ddev, &vbox_fb->base, &vbox_fb_funcs);
-	if (ret) {
-		DRM_ERROR("framebuffer init failed %d\n", ret);
-		return ret;
-	}
-
-	return 0;
-}
-
 static int vbox_accel_init(struct vbox_private *vbox)
 {
 	struct vbva_buffer *vbva;
@@ -270,29 +179,3 @@ void vbox_hw_fini(struct vbox_private *vbox)
 	gen_pool_destroy(vbox->guest_pool);
 	pci_iounmap(vbox->ddev.pdev, vbox->guest_heap);
 }
-
-int vbox_gem_create(struct vbox_private *vbox,
-		    u32 size, bool iskernel, struct drm_gem_object **obj)
-{
-	struct drm_gem_vram_object *gbo;
-	int ret;
-
-	*obj = NULL;
-
-	size = roundup(size, PAGE_SIZE);
-	if (size == 0)
-		return -EINVAL;
-
-	gbo = drm_gem_vram_create(&vbox->ddev, &vbox->ddev.vram_mm->bdev,
-				  size, 0, false);
-	if (IS_ERR(gbo)) {
-		ret = PTR_ERR(gbo);
-		if (ret != -ERESTARTSYS)
-			DRM_ERROR("failed to allocate GEM object\n");
-		return ret;
-	}
-
-	*obj = &gbo->bo.base;
-
-	return 0;
-}
diff --git a/drivers/gpu/drm/vboxvideo/vbox_mode.c b/drivers/gpu/drm/vboxvideo/vbox_mode.c
index e1e48ba919eb..19612132c8a3 100644
--- a/drivers/gpu/drm/vboxvideo/vbox_mode.c
+++ b/drivers/gpu/drm/vboxvideo/vbox_mode.c
@@ -13,7 +13,9 @@
 
 #include <drm/drm_atomic.h>
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_fb_helper.h>
 #include <drm/drm_fourcc.h>
+#include <drm/drm_gem_framebuffer_helper.h>
 #include <drm/drm_plane_helper.h>
 #include <drm/drm_probe_helper.h>
 #include <drm/drm_vblank.h>
@@ -133,7 +135,7 @@ static bool vbox_set_up_input_mapping(struct vbox_private *vbox)
 
 		if (!fb1) {
 			fb1 = fb;
-			if (to_vbox_framebuffer(fb1) == &vbox->afb)
+			if (fb1 == vbox->ddev.fb_helper->fb)
 				break;
 		} else if (fb != fb1) {
 			single_framebuffer = false;
@@ -172,8 +174,7 @@ static void vbox_crtc_set_base_and_mode(struct drm_crtc *crtc,
 					struct drm_framebuffer *fb,
 					int x, int y)
 {
-	struct drm_gem_vram_object *gbo =
-		drm_gem_vram_of_gem(to_vbox_framebuffer(fb)->obj);
+	struct drm_gem_vram_object *gbo = drm_gem_vram_of_gem(fb->obj[0]);
 	struct vbox_private *vbox = crtc->dev->dev_private;
 	struct vbox_crtc *vbox_crtc = to_vbox_crtc(crtc);
 	bool needs_modeset = drm_atomic_crtc_needs_modeset(crtc->state);
@@ -283,10 +284,43 @@ static void vbox_primary_atomic_update(struct drm_plane *plane,
 {
 	struct drm_crtc *crtc = plane->state->crtc;
 	struct drm_framebuffer *fb = plane->state->fb;
+	struct vbox_private *vbox = fb->dev->dev_private;
+	struct drm_mode_rect *clips;
+	uint32_t num_clips, i;
 
 	vbox_crtc_set_base_and_mode(crtc, fb,
 				    plane->state->src_x >> 16,
 				    plane->state->src_y >> 16);
+
+	/* Send information about dirty rectangles to VBVA. */
+
+	clips = drm_plane_get_damage_clips(plane->state);
+	num_clips = drm_plane_get_damage_clips_count(plane->state);
+
+	if (!num_clips)
+		return;
+
+	mutex_lock(&vbox->hw_mutex);
+
+	for (i = 0; i < num_clips; ++i, ++clips) {
+		struct vbva_cmd_hdr cmd_hdr;
+		unsigned int crtc_id = to_vbox_crtc(crtc)->crtc_id;
+
+		cmd_hdr.x = (s16)clips->x1;
+		cmd_hdr.y = (s16)clips->y1;
+		cmd_hdr.w = (u16)clips->x2 - clips->x1;
+		cmd_hdr.h = (u16)clips->y2 - clips->y1;
+
+		if (!vbva_buffer_begin_update(&vbox->vbva_info[crtc_id],
+					      vbox->guest_pool))
+			continue;
+
+		vbva_write(&vbox->vbva_info[crtc_id], vbox->guest_pool,
+			   &cmd_hdr, sizeof(cmd_hdr));
+		vbva_buffer_end_update(&vbox->vbva_info[crtc_id]);
+	}
+
+	mutex_unlock(&vbox->hw_mutex);
 }
 
 static void vbox_primary_atomic_disable(struct drm_plane *plane,
@@ -300,35 +334,6 @@ static void vbox_primary_atomic_disable(struct drm_plane *plane,
 				    old_state->src_y >> 16);
 }
 
-static int vbox_primary_prepare_fb(struct drm_plane *plane,
-				   struct drm_plane_state *new_state)
-{
-	struct drm_gem_vram_object *gbo;
-	int ret;
-
-	if (!new_state->fb)
-		return 0;
-
-	gbo = drm_gem_vram_of_gem(to_vbox_framebuffer(new_state->fb)->obj);
-	ret = drm_gem_vram_pin(gbo, DRM_GEM_VRAM_PL_FLAG_VRAM);
-	if (ret)
-		DRM_WARN("Error %d pinning new fb, out of video mem?\n", ret);
-
-	return ret;
-}
-
-static void vbox_primary_cleanup_fb(struct drm_plane *plane,
-				    struct drm_plane_state *old_state)
-{
-	struct drm_gem_vram_object *gbo;
-
-	if (!old_state->fb)
-		return;
-
-	gbo = drm_gem_vram_of_gem(to_vbox_framebuffer(old_state->fb)->obj);
-	drm_gem_vram_unpin(gbo);
-}
-
 static int vbox_cursor_atomic_check(struct drm_plane *plane,
 				    struct drm_plane_state *new_state)
 {
@@ -386,8 +391,7 @@ static void vbox_cursor_atomic_update(struct drm_plane *plane,
 		container_of(plane->dev, struct vbox_private, ddev);
 	struct vbox_crtc *vbox_crtc = to_vbox_crtc(plane->state->crtc);
 	struct drm_framebuffer *fb = plane->state->fb;
-	struct drm_gem_vram_object *gbo =
-		drm_gem_vram_of_gem(to_vbox_framebuffer(fb)->obj);
+	struct drm_gem_vram_object *gbo = drm_gem_vram_of_gem(fb->obj[0]);
 	u32 width = plane->state->crtc_w;
 	u32 height = plane->state->crtc_h;
 	size_t data_size, mask_size;
@@ -459,30 +463,6 @@ static void vbox_cursor_atomic_disable(struct drm_plane *plane,
 	mutex_unlock(&vbox->hw_mutex);
 }
 
-static int vbox_cursor_prepare_fb(struct drm_plane *plane,
-				  struct drm_plane_state *new_state)
-{
-	struct drm_gem_vram_object *gbo;
-
-	if (!new_state->fb)
-		return 0;
-
-	gbo = drm_gem_vram_of_gem(to_vbox_framebuffer(new_state->fb)->obj);
-	return drm_gem_vram_pin(gbo, DRM_GEM_VRAM_PL_FLAG_SYSTEM);
-}
-
-static void vbox_cursor_cleanup_fb(struct drm_plane *plane,
-				   struct drm_plane_state *old_state)
-{
-	struct drm_gem_vram_object *gbo;
-
-	if (!plane->state->fb)
-		return;
-
-	gbo = drm_gem_vram_of_gem(to_vbox_framebuffer(plane->state->fb)->obj);
-	drm_gem_vram_unpin(gbo);
-}
-
 static const u32 vbox_cursor_plane_formats[] = {
 	DRM_FORMAT_ARGB8888,
 };
@@ -491,8 +471,8 @@ static const struct drm_plane_helper_funcs vbox_cursor_helper_funcs = {
 	.atomic_check	= vbox_cursor_atomic_check,
 	.atomic_update	= vbox_cursor_atomic_update,
 	.atomic_disable	= vbox_cursor_atomic_disable,
-	.prepare_fb	= vbox_cursor_prepare_fb,
-	.cleanup_fb	= vbox_cursor_cleanup_fb,
+	.prepare_fb	= drm_gem_vram_plane_helper_prepare_fb,
+	.cleanup_fb	= drm_gem_vram_plane_helper_cleanup_fb,
 };
 
 static const struct drm_plane_funcs vbox_cursor_plane_funcs = {
@@ -513,8 +493,8 @@ static const struct drm_plane_helper_funcs vbox_primary_helper_funcs = {
 	.atomic_check = vbox_primary_atomic_check,
 	.atomic_update = vbox_primary_atomic_update,
 	.atomic_disable = vbox_primary_atomic_disable,
-	.prepare_fb = vbox_primary_prepare_fb,
-	.cleanup_fb = vbox_primary_cleanup_fb,
+	.prepare_fb	= drm_gem_vram_plane_helper_prepare_fb,
+	.cleanup_fb	= drm_gem_vram_plane_helper_cleanup_fb,
 };
 
 static const struct drm_plane_funcs vbox_primary_plane_funcs = {
@@ -856,40 +836,8 @@ static int vbox_connector_init(struct drm_device *dev,
 	return 0;
 }
 
-static struct drm_framebuffer *vbox_user_framebuffer_create(
-		struct drm_device *dev,
-		struct drm_file *filp,
-		const struct drm_mode_fb_cmd2 *mode_cmd)
-{
-	struct vbox_private *vbox =
-		container_of(dev, struct vbox_private, ddev);
-	struct drm_gem_object *obj;
-	struct vbox_framebuffer *vbox_fb;
-	int ret = -ENOMEM;
-
-	obj = drm_gem_object_lookup(filp, mode_cmd->handles[0]);
-	if (!obj)
-		return ERR_PTR(-ENOENT);
-
-	vbox_fb = kzalloc(sizeof(*vbox_fb), GFP_KERNEL);
-	if (!vbox_fb)
-		goto err_unref_obj;
-
-	ret = vbox_framebuffer_init(vbox, vbox_fb, mode_cmd, obj);
-	if (ret)
-		goto err_free_vbox_fb;
-
-	return &vbox_fb->base;
-
-err_free_vbox_fb:
-	kfree(vbox_fb);
-err_unref_obj:
-	drm_gem_object_put_unlocked(obj);
-	return ERR_PTR(ret);
-}
-
 static const struct drm_mode_config_funcs vbox_mode_funcs = {
-	.fb_create = vbox_user_framebuffer_create,
+	.fb_create = drm_gem_fb_create_with_dirty,
 	.atomic_check = drm_atomic_helper_check,
 	.atomic_commit = drm_atomic_helper_commit,
 };
diff --git a/drivers/gpu/drm/vboxvideo/vbox_ttm.c b/drivers/gpu/drm/vboxvideo/vbox_ttm.c
index b82595a9ed0f..976423d0c3cc 100644
--- a/drivers/gpu/drm/vboxvideo/vbox_ttm.c
+++ b/drivers/gpu/drm/vboxvideo/vbox_ttm.c
@@ -17,8 +17,7 @@ int vbox_mm_init(struct vbox_private *vbox)
 	struct drm_device *dev = &vbox->ddev;
 
 	vmm = drm_vram_helper_alloc_mm(dev, pci_resource_start(dev->pdev, 0),
-				       vbox->available_vram_size,
-				       &drm_gem_vram_mm_funcs);
+				       vbox->available_vram_size);
 	if (IS_ERR(vmm)) {
 		ret = PTR_ERR(vmm);
 		DRM_ERROR("Error initializing VRAM MM; %d\n", ret);
diff --git a/drivers/gpu/drm/vc4/vc4_crtc.c b/drivers/gpu/drm/vc4/vc4_crtc.c
index f1f0a7c87771..b00e20f5ce05 100644
--- a/drivers/gpu/drm/vc4/vc4_crtc.c
+++ b/drivers/gpu/drm/vc4/vc4_crtc.c
@@ -994,7 +994,7 @@ static void vc4_crtc_destroy_state(struct drm_crtc *crtc,
 	struct vc4_dev *vc4 = to_vc4_dev(crtc->dev);
 	struct vc4_crtc_state *vc4_state = to_vc4_crtc_state(state);
 
-	if (vc4_state->mm.allocated) {
+	if (drm_mm_node_allocated(&vc4_state->mm)) {
 		unsigned long flags;
 
 		spin_lock_irqsave(&vc4->hvs->mm_lock, flags);
diff --git a/drivers/gpu/drm/vc4/vc4_dpi.c b/drivers/gpu/drm/vc4/vc4_dpi.c
index 8a27a6acee61..c586325de2a5 100644
--- a/drivers/gpu/drm/vc4/vc4_dpi.c
+++ b/drivers/gpu/drm/vc4/vc4_dpi.c
@@ -249,7 +249,8 @@ static int vc4_dpi_init_bridge(struct vc4_dpi *dpi)
 	}
 
 	if (panel)
-		bridge = drm_panel_bridge_add(panel, DRM_MODE_CONNECTOR_DPI);
+		bridge = drm_panel_bridge_add_typed(panel,
+						    DRM_MODE_CONNECTOR_DPI);
 
 	return drm_bridge_attach(dpi->encoder, bridge, NULL);
 }
diff --git a/drivers/gpu/drm/vc4/vc4_dsi.c b/drivers/gpu/drm/vc4/vc4_dsi.c
index c78fa8144776..c9ba83ed49b9 100644
--- a/drivers/gpu/drm/vc4/vc4_dsi.c
+++ b/drivers/gpu/drm/vc4/vc4_dsi.c
@@ -31,6 +31,7 @@
 #include <linux/pm_runtime.h>
 
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_bridge.h>
 #include <drm/drm_edid.h>
 #include <drm/drm_mipi_dsi.h>
 #include <drm/drm_of.h>
@@ -1575,8 +1576,8 @@ static int vc4_dsi_bind(struct device *dev, struct device *master, void *data)
 	}
 
 	if (panel) {
-		dsi->bridge = devm_drm_panel_bridge_add(dev, panel,
-							DRM_MODE_CONNECTOR_DSI);
+		dsi->bridge = devm_drm_panel_bridge_add_typed(dev, panel,
+							      DRM_MODE_CONNECTOR_DSI);
 		if (IS_ERR(dsi->bridge))
 			return PTR_ERR(dsi->bridge);
 	}
diff --git a/drivers/gpu/drm/vc4/vc4_hdmi.c b/drivers/gpu/drm/vc4/vc4_hdmi.c
index ee7d4e7b0ee3..1c62c6c9244b 100644
--- a/drivers/gpu/drm/vc4/vc4_hdmi.c
+++ b/drivers/gpu/drm/vc4/vc4_hdmi.c
@@ -398,10 +398,7 @@ static void vc4_hdmi_set_avi_infoframe(struct drm_encoder *encoder)
 					   HDMI_QUANTIZATION_RANGE_LIMITED :
 					   HDMI_QUANTIZATION_RANGE_FULL);
 
-	frame.avi.right_bar = cstate->tv.margins.right;
-	frame.avi.left_bar = cstate->tv.margins.left;
-	frame.avi.top_bar = cstate->tv.margins.top;
-	frame.avi.bottom_bar = cstate->tv.margins.bottom;
+	drm_hdmi_avi_infoframe_bars(&frame.avi, cstate);
 
 	vc4_hdmi_write_infoframe(encoder, &frame);
 }
@@ -1285,6 +1282,9 @@ static const struct cec_adap_ops vc4_hdmi_cec_adap_ops = {
 
 static int vc4_hdmi_bind(struct device *dev, struct device *master, void *data)
 {
+#ifdef CONFIG_DRM_VC4_HDMI_CEC
+	struct cec_connector_info conn_info;
+#endif
 	struct platform_device *pdev = to_platform_device(dev);
 	struct drm_device *drm = dev_get_drvdata(master);
 	struct vc4_dev *vc4 = drm->dev_private;
@@ -1403,13 +1403,15 @@ static int vc4_hdmi_bind(struct device *dev, struct device *master, void *data)
 #ifdef CONFIG_DRM_VC4_HDMI_CEC
 	hdmi->cec_adap = cec_allocate_adapter(&vc4_hdmi_cec_adap_ops,
 					      vc4, "vc4",
-					      CEC_CAP_TRANSMIT |
-					      CEC_CAP_LOG_ADDRS |
-					      CEC_CAP_PASSTHROUGH |
-					      CEC_CAP_RC, 1);
+					      CEC_CAP_DEFAULTS |
+					      CEC_CAP_CONNECTOR_INFO, 1);
 	ret = PTR_ERR_OR_ZERO(hdmi->cec_adap);
 	if (ret < 0)
 		goto err_destroy_conn;
+
+	cec_fill_conn_info_from_drm(&conn_info, hdmi->connector);
+	cec_s_conn_info(hdmi->cec_adap, &conn_info);
+
 	HDMI_WRITE(VC4_HDMI_CPU_MASK_SET, 0xffffffff);
 	value = HDMI_READ(VC4_HDMI_CEC_CNTRL_1);
 	value &= ~VC4_HDMI_CEC_DIV_CLK_CNT_MASK;
diff --git a/drivers/gpu/drm/vc4/vc4_hvs.c b/drivers/gpu/drm/vc4/vc4_hvs.c
index 9936b15d0bf1..5a43659da319 100644
--- a/drivers/gpu/drm/vc4/vc4_hvs.c
+++ b/drivers/gpu/drm/vc4/vc4_hvs.c
@@ -315,7 +315,7 @@ static void vc4_hvs_unbind(struct device *dev, struct device *master,
 	struct drm_device *drm = dev_get_drvdata(master);
 	struct vc4_dev *vc4 = drm->dev_private;
 
-	if (vc4->hvs->mitchell_netravali_filter.allocated)
+	if (drm_mm_node_allocated(&vc4->hvs->mitchell_netravali_filter))
 		drm_mm_remove_node(&vc4->hvs->mitchell_netravali_filter);
 
 	drm_mm_takedown(&vc4->hvs->dlist_mm);
diff --git a/drivers/gpu/drm/vc4/vc4_plane.c b/drivers/gpu/drm/vc4/vc4_plane.c
index 5e5f90810aca..4934127f0d76 100644
--- a/drivers/gpu/drm/vc4/vc4_plane.c
+++ b/drivers/gpu/drm/vc4/vc4_plane.c
@@ -178,7 +178,7 @@ static void vc4_plane_destroy_state(struct drm_plane *plane,
 	struct vc4_dev *vc4 = to_vc4_dev(plane->dev);
 	struct vc4_plane_state *vc4_state = to_vc4_plane_state(state);
 
-	if (vc4_state->lbm.allocated) {
+	if (drm_mm_node_allocated(&vc4_state->lbm)) {
 		unsigned long irqflags;
 
 		spin_lock_irqsave(&vc4->hvs->mm_lock, irqflags);
@@ -557,7 +557,7 @@ static int vc4_plane_allocate_lbm(struct drm_plane_state *state)
 	/* Allocate the LBM memory that the HVS will use for temporary
 	 * storage due to our scaling/format conversion.
 	 */
-	if (!vc4_state->lbm.allocated) {
+	if (!drm_mm_node_allocated(&vc4_state->lbm)) {
 		int ret;
 
 		spin_lock_irqsave(&vc4->hvs->mm_lock, irqflags);
diff --git a/drivers/gpu/drm/vc4/vc4_txp.c b/drivers/gpu/drm/vc4/vc4_txp.c
index 1ce4d7142b6e..bf720206727f 100644
--- a/drivers/gpu/drm/vc4/vc4_txp.c
+++ b/drivers/gpu/drm/vc4/vc4_txp.c
@@ -231,7 +231,7 @@ static int vc4_txp_connector_atomic_check(struct drm_connector *conn,
 	int i;
 
 	conn_state = drm_atomic_get_new_connector_state(state, conn);
-	if (!conn_state->writeback_job || !conn_state->writeback_job->fb)
+	if (!conn_state->writeback_job)
 		return 0;
 
 	crtc_state = drm_atomic_get_new_crtc_state(state, conn_state->crtc);
@@ -271,8 +271,7 @@ static void vc4_txp_connector_atomic_commit(struct drm_connector *conn,
 	u32 ctrl;
 	int i;
 
-	if (WARN_ON(!conn_state->writeback_job ||
-		    !conn_state->writeback_job->fb))
+	if (WARN_ON(!conn_state->writeback_job))
 		return;
 
 	mode = &conn_state->crtc->state->adjusted_mode;
diff --git a/drivers/gpu/drm/virtio/Kconfig b/drivers/gpu/drm/virtio/Kconfig
index ba36e933bb49..eff3047052d4 100644
--- a/drivers/gpu/drm/virtio/Kconfig
+++ b/drivers/gpu/drm/virtio/Kconfig
@@ -3,7 +3,7 @@ config DRM_VIRTIO_GPU
 	tristate "Virtio GPU driver"
 	depends on DRM && VIRTIO && MMU
 	select DRM_KMS_HELPER
-	select DRM_TTM
+	select DRM_GEM_SHMEM_HELPER
 	help
 	   This is the virtual GPU driver for virtio.  It can be used with
 	   QEMU based VMMs (like KVM or Xen).
diff --git a/drivers/gpu/drm/virtio/Makefile b/drivers/gpu/drm/virtio/Makefile
index 458e606a936f..92aa2b3d349d 100644
--- a/drivers/gpu/drm/virtio/Makefile
+++ b/drivers/gpu/drm/virtio/Makefile
@@ -4,7 +4,7 @@
 # Direct Rendering Infrastructure (DRI) in XFree86 4.1.0 and higher.
 
 virtio-gpu-y := virtgpu_drv.o virtgpu_kms.o virtgpu_gem.o \
-	virtgpu_display.o virtgpu_vq.o virtgpu_ttm.o \
+	virtgpu_display.o virtgpu_vq.o \
 	virtgpu_fence.o virtgpu_object.o virtgpu_debugfs.o virtgpu_plane.o \
 	virtgpu_ioctl.o virtgpu_prime.o virtgpu_trace_points.o
 
diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.c b/drivers/gpu/drm/virtio/virtgpu_drv.c
index 0fc32fa0b3c0..8dee698c90ff 100644
--- a/drivers/gpu/drm/virtio/virtgpu_drv.c
+++ b/drivers/gpu/drm/virtio/virtgpu_drv.c
@@ -56,7 +56,6 @@ static int virtio_gpu_pci_quirk(struct drm_device *dev, struct virtio_device *vd
 	dev->pdev = pdev;
 	if (vga)
 		drm_fb_helper_remove_conflicting_pci_framebuffers(pdev,
-								  0,
 								  "virtiodrmfb");
 
 	/*
@@ -185,17 +184,7 @@ MODULE_AUTHOR("Dave Airlie <airlied@redhat.com>");
 MODULE_AUTHOR("Gerd Hoffmann <kraxel@redhat.com>");
 MODULE_AUTHOR("Alon Levy");
 
-static const struct file_operations virtio_gpu_driver_fops = {
-	.owner = THIS_MODULE,
-	.open = drm_open,
-	.mmap = virtio_gpu_mmap,
-	.poll = drm_poll,
-	.read = drm_read,
-	.unlocked_ioctl	= drm_ioctl,
-	.release = drm_release,
-	.compat_ioctl = drm_compat_ioctl,
-	.llseek = noop_llseek,
-};
+DEFINE_DRM_GEM_FOPS(virtio_gpu_driver_fops);
 
 static struct drm_driver driver = {
 	.driver_features = DRIVER_MODESET | DRIVER_GEM | DRIVER_RENDER | DRIVER_ATOMIC,
@@ -210,15 +199,10 @@ static struct drm_driver driver = {
 #endif
 	.prime_handle_to_fd = drm_gem_prime_handle_to_fd,
 	.prime_fd_to_handle = drm_gem_prime_fd_to_handle,
-	.gem_prime_get_sg_table = virtgpu_gem_prime_get_sg_table,
+	.gem_prime_mmap = drm_gem_prime_mmap,
 	.gem_prime_import_sg_table = virtgpu_gem_prime_import_sg_table,
-	.gem_prime_vmap = virtgpu_gem_prime_vmap,
-	.gem_prime_vunmap = virtgpu_gem_prime_vunmap,
-	.gem_prime_mmap = virtgpu_gem_prime_mmap,
 
-	.gem_free_object_unlocked = virtio_gpu_gem_free_object,
-	.gem_open_object = virtio_gpu_gem_object_open,
-	.gem_close_object = virtio_gpu_gem_object_close,
+	.gem_create_object = virtio_gpu_create_object,
 	.fops = &virtio_gpu_driver_fops,
 
 	.ioctls = virtio_gpu_ioctls,
diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h b/drivers/gpu/drm/virtio/virtgpu_drv.h
index e28829661724..0b56ba005e25 100644
--- a/drivers/gpu/drm/virtio/virtgpu_drv.h
+++ b/drivers/gpu/drm/virtio/virtgpu_drv.h
@@ -35,12 +35,9 @@
 #include <drm/drm_encoder.h>
 #include <drm/drm_fb_helper.h>
 #include <drm/drm_gem.h>
+#include <drm/drm_gem_shmem_helper.h>
 #include <drm/drm_ioctl.h>
 #include <drm/drm_probe_helper.h>
-#include <drm/ttm/ttm_bo_api.h>
-#include <drm/ttm/ttm_bo_driver.h>
-#include <drm/ttm/ttm_module.h>
-#include <drm/ttm/ttm_placement.h>
 
 #define DRIVER_NAME "virtio_gpu"
 #define DRIVER_DESC "virtio GPU"
@@ -68,21 +65,23 @@ struct virtio_gpu_object_params {
 };
 
 struct virtio_gpu_object {
-	struct drm_gem_object gem_base;
+	struct drm_gem_shmem_object base;
 	uint32_t hw_res_handle;
 
 	struct sg_table *pages;
 	uint32_t mapped;
-	void *vmap;
 	bool dumb;
-	struct ttm_place                placement_code;
-	struct ttm_placement		placement;
-	struct ttm_buffer_object	tbo;
-	struct ttm_bo_kmap_obj		kmap;
 	bool created;
 };
 #define gem_to_virtio_gpu_obj(gobj) \
-	container_of((gobj), struct virtio_gpu_object, gem_base)
+	container_of((gobj), struct virtio_gpu_object, base.base)
+
+struct virtio_gpu_object_array {
+	struct ww_acquire_ctx ticket;
+	struct list_head next;
+	u32 nents, total;
+	struct drm_gem_object *objs[];
+};
 
 struct virtio_gpu_vbuffer;
 struct virtio_gpu_device;
@@ -115,9 +114,9 @@ struct virtio_gpu_vbuffer {
 
 	char *resp_buf;
 	int resp_size;
-
 	virtio_gpu_resp_cb resp_cb;
 
+	struct virtio_gpu_object_array *objs;
 	struct list_head list;
 };
 
@@ -147,10 +146,6 @@ struct virtio_gpu_framebuffer {
 #define to_virtio_gpu_framebuffer(x) \
 	container_of(x, struct virtio_gpu_framebuffer, base)
 
-struct virtio_gpu_mman {
-	struct ttm_bo_device		bdev;
-};
-
 struct virtio_gpu_queue {
 	struct virtqueue *vq;
 	spinlock_t qlock;
@@ -179,8 +174,6 @@ struct virtio_gpu_device {
 
 	struct virtio_device *vdev;
 
-	struct virtio_gpu_mman mman;
-
 	struct virtio_gpu_output outputs[VIRTIO_GPU_MAX_SCANOUTS];
 	uint32_t num_scanouts;
 
@@ -205,6 +198,10 @@ struct virtio_gpu_device {
 
 	struct work_struct config_changed_work;
 
+	struct work_struct obj_free_work;
+	spinlock_t obj_free_lock;
+	struct list_head obj_free_list;
+
 	struct virtio_gpu_drv_capset *capsets;
 	uint32_t num_capsets;
 	struct list_head cap_cache;
@@ -217,9 +214,6 @@ struct virtio_gpu_fpriv {
 /* virtio_ioctl.c */
 #define DRM_VIRTIO_NUM_IOCTLS 10
 extern struct drm_ioctl_desc virtio_gpu_ioctls[DRM_VIRTIO_NUM_IOCTLS];
-int virtio_gpu_object_list_validate(struct ww_acquire_ctx *ticket,
-				    struct list_head *head);
-void virtio_gpu_unref_list(struct list_head *head);
 
 /* virtio_kms.c */
 int virtio_gpu_init(struct drm_device *dev);
@@ -240,10 +234,6 @@ int virtio_gpu_gem_object_open(struct drm_gem_object *obj,
 			       struct drm_file *file);
 void virtio_gpu_gem_object_close(struct drm_gem_object *obj,
 				 struct drm_file *file);
-struct virtio_gpu_object*
-virtio_gpu_alloc_object(struct drm_device *dev,
-			struct virtio_gpu_object_params *params,
-			struct virtio_gpu_fence *fence);
 int virtio_gpu_mode_dumb_create(struct drm_file *file_priv,
 				struct drm_device *dev,
 				struct drm_mode_create_dumb *args);
@@ -251,20 +241,35 @@ int virtio_gpu_mode_dumb_mmap(struct drm_file *file_priv,
 			      struct drm_device *dev,
 			      uint32_t handle, uint64_t *offset_p);
 
+struct virtio_gpu_object_array *virtio_gpu_array_alloc(u32 nents);
+struct virtio_gpu_object_array*
+virtio_gpu_array_from_handles(struct drm_file *drm_file, u32 *handles, u32 nents);
+void virtio_gpu_array_add_obj(struct virtio_gpu_object_array *objs,
+			      struct drm_gem_object *obj);
+int virtio_gpu_array_lock_resv(struct virtio_gpu_object_array *objs);
+void virtio_gpu_array_unlock_resv(struct virtio_gpu_object_array *objs);
+void virtio_gpu_array_add_fence(struct virtio_gpu_object_array *objs,
+				struct dma_fence *fence);
+void virtio_gpu_array_put_free(struct virtio_gpu_object_array *objs);
+void virtio_gpu_array_put_free_delayed(struct virtio_gpu_device *vgdev,
+				       struct virtio_gpu_object_array *objs);
+void virtio_gpu_array_put_free_work(struct work_struct *work);
+
 /* virtio vg */
 int virtio_gpu_alloc_vbufs(struct virtio_gpu_device *vgdev);
 void virtio_gpu_free_vbufs(struct virtio_gpu_device *vgdev);
 void virtio_gpu_cmd_create_resource(struct virtio_gpu_device *vgdev,
 				    struct virtio_gpu_object *bo,
 				    struct virtio_gpu_object_params *params,
+				    struct virtio_gpu_object_array *objs,
 				    struct virtio_gpu_fence *fence);
 void virtio_gpu_cmd_unref_resource(struct virtio_gpu_device *vgdev,
 				   uint32_t resource_id);
 void virtio_gpu_cmd_transfer_to_host_2d(struct virtio_gpu_device *vgdev,
-					struct virtio_gpu_object *bo,
 					uint64_t offset,
-					__le32 width, __le32 height,
-					__le32 x, __le32 y,
+					uint32_t width, uint32_t height,
+					uint32_t x, uint32_t y,
+					struct virtio_gpu_object_array *objs,
 					struct virtio_gpu_fence *fence);
 void virtio_gpu_cmd_resource_flush(struct virtio_gpu_device *vgdev,
 				   uint32_t resource_id,
@@ -295,28 +300,32 @@ void virtio_gpu_cmd_context_destroy(struct virtio_gpu_device *vgdev,
 				    uint32_t id);
 void virtio_gpu_cmd_context_attach_resource(struct virtio_gpu_device *vgdev,
 					    uint32_t ctx_id,
-					    uint32_t resource_id);
+					    struct virtio_gpu_object_array *objs);
 void virtio_gpu_cmd_context_detach_resource(struct virtio_gpu_device *vgdev,
 					    uint32_t ctx_id,
-					    uint32_t resource_id);
+					    struct virtio_gpu_object_array *objs);
 void virtio_gpu_cmd_submit(struct virtio_gpu_device *vgdev,
 			   void *data, uint32_t data_size,
-			   uint32_t ctx_id, struct virtio_gpu_fence *fence);
+			   uint32_t ctx_id,
+			   struct virtio_gpu_object_array *objs,
+			   struct virtio_gpu_fence *fence);
 void virtio_gpu_cmd_transfer_from_host_3d(struct virtio_gpu_device *vgdev,
-					  uint32_t resource_id, uint32_t ctx_id,
+					  uint32_t ctx_id,
 					  uint64_t offset, uint32_t level,
 					  struct virtio_gpu_box *box,
+					  struct virtio_gpu_object_array *objs,
 					  struct virtio_gpu_fence *fence);
 void virtio_gpu_cmd_transfer_to_host_3d(struct virtio_gpu_device *vgdev,
-					struct virtio_gpu_object *bo,
 					uint32_t ctx_id,
 					uint64_t offset, uint32_t level,
 					struct virtio_gpu_box *box,
+					struct virtio_gpu_object_array *objs,
 					struct virtio_gpu_fence *fence);
 void
 virtio_gpu_cmd_resource_create_3d(struct virtio_gpu_device *vgdev,
 				  struct virtio_gpu_object *bo,
 				  struct virtio_gpu_object_params *params,
+				  struct virtio_gpu_object_array *objs,
 				  struct virtio_gpu_fence *fence);
 void virtio_gpu_ctrl_ack(struct virtqueue *vq);
 void virtio_gpu_cursor_ack(struct virtqueue *vq);
@@ -339,11 +348,6 @@ struct drm_plane *virtio_gpu_plane_init(struct virtio_gpu_device *vgdev,
 					enum drm_plane_type type,
 					int index);
 
-/* virtio_gpu_ttm.c */
-int virtio_gpu_ttm_init(struct virtio_gpu_device *vgdev);
-void virtio_gpu_ttm_fini(struct virtio_gpu_device *vgdev);
-int virtio_gpu_mmap(struct file *filp, struct vm_area_struct *vma);
-
 /* virtio_gpu_fence.c */
 bool virtio_fence_signaled(struct dma_fence *f);
 struct virtio_gpu_fence *virtio_gpu_fence_alloc(
@@ -355,70 +359,21 @@ void virtio_gpu_fence_event_process(struct virtio_gpu_device *vdev,
 				    u64 last_seq);
 
 /* virtio_gpu_object */
+struct drm_gem_object *virtio_gpu_create_object(struct drm_device *dev,
+						size_t size);
 int virtio_gpu_object_create(struct virtio_gpu_device *vgdev,
 			     struct virtio_gpu_object_params *params,
 			     struct virtio_gpu_object **bo_ptr,
 			     struct virtio_gpu_fence *fence);
-void virtio_gpu_object_kunmap(struct virtio_gpu_object *bo);
-int virtio_gpu_object_kmap(struct virtio_gpu_object *bo);
-int virtio_gpu_object_get_sg_table(struct virtio_gpu_device *qdev,
-				   struct virtio_gpu_object *bo);
-void virtio_gpu_object_free_sg_table(struct virtio_gpu_object *bo);
-int virtio_gpu_object_wait(struct virtio_gpu_object *bo, bool no_wait);
 
 /* virtgpu_prime.c */
-struct sg_table *virtgpu_gem_prime_get_sg_table(struct drm_gem_object *obj);
 struct drm_gem_object *virtgpu_gem_prime_import_sg_table(
 	struct drm_device *dev, struct dma_buf_attachment *attach,
 	struct sg_table *sgt);
-void *virtgpu_gem_prime_vmap(struct drm_gem_object *obj);
-void virtgpu_gem_prime_vunmap(struct drm_gem_object *obj, void *vaddr);
-int virtgpu_gem_prime_mmap(struct drm_gem_object *obj,
-			   struct vm_area_struct *vma);
-
-static inline struct virtio_gpu_object*
-virtio_gpu_object_ref(struct virtio_gpu_object *bo)
-{
-	ttm_bo_get(&bo->tbo);
-	return bo;
-}
-
-static inline void virtio_gpu_object_unref(struct virtio_gpu_object **bo)
-{
-	struct ttm_buffer_object *tbo;
-
-	if ((*bo) == NULL)
-		return;
-	tbo = &((*bo)->tbo);
-	ttm_bo_put(tbo);
-	*bo = NULL;
-}
 
 static inline u64 virtio_gpu_object_mmap_offset(struct virtio_gpu_object *bo)
 {
-	return drm_vma_node_offset_addr(&bo->tbo.base.vma_node);
-}
-
-static inline int virtio_gpu_object_reserve(struct virtio_gpu_object *bo,
-					 bool no_wait)
-{
-	int r;
-
-	r = ttm_bo_reserve(&bo->tbo, true, no_wait, NULL);
-	if (unlikely(r != 0)) {
-		if (r != -ERESTARTSYS) {
-			struct virtio_gpu_device *qdev =
-				bo->gem_base.dev->dev_private;
-			dev_err(qdev->dev, "%p reserve failed\n", bo);
-		}
-		return r;
-	}
-	return 0;
-}
-
-static inline void virtio_gpu_object_unreserve(struct virtio_gpu_object *bo)
-{
-	ttm_bo_unreserve(&bo->tbo);
+	return drm_vma_node_offset_addr(&bo->base.base.vma_node);
 }
 
 /* virgl debufs */
diff --git a/drivers/gpu/drm/virtio/virtgpu_fence.c b/drivers/gpu/drm/virtio/virtgpu_fence.c
index a0514f5bd006..a4b9881ca1d3 100644
--- a/drivers/gpu/drm/virtio/virtgpu_fence.c
+++ b/drivers/gpu/drm/virtio/virtgpu_fence.c
@@ -41,6 +41,10 @@ bool virtio_fence_signaled(struct dma_fence *f)
 {
 	struct virtio_gpu_fence *fence = to_virtio_fence(f);
 
+	if (WARN_ON_ONCE(fence->f.seqno == 0))
+		/* leaked fence outside driver before completing
+		 * initialization with virtio_gpu_fence_emit */
+		return false;
 	if (atomic64_read(&fence->drv->last_seq) >= fence->f.seqno)
 		return true;
 	return false;
diff --git a/drivers/gpu/drm/virtio/virtgpu_gem.c b/drivers/gpu/drm/virtio/virtgpu_gem.c
index 292566146814..4c1f579edfb3 100644
--- a/drivers/gpu/drm/virtio/virtgpu_gem.c
+++ b/drivers/gpu/drm/virtio/virtgpu_gem.c
@@ -28,54 +28,31 @@
 
 #include "virtgpu_drv.h"
 
-void virtio_gpu_gem_free_object(struct drm_gem_object *gem_obj)
-{
-	struct virtio_gpu_object *obj = gem_to_virtio_gpu_obj(gem_obj);
-
-	if (obj)
-		virtio_gpu_object_unref(&obj);
-}
-
-struct virtio_gpu_object*
-virtio_gpu_alloc_object(struct drm_device *dev,
-			struct virtio_gpu_object_params *params,
-			struct virtio_gpu_fence *fence)
-{
-	struct virtio_gpu_device *vgdev = dev->dev_private;
-	struct virtio_gpu_object *obj;
-	int ret;
-
-	ret = virtio_gpu_object_create(vgdev, params, &obj, fence);
-	if (ret)
-		return ERR_PTR(ret);
-
-	return obj;
-}
-
 int virtio_gpu_gem_create(struct drm_file *file,
 			  struct drm_device *dev,
 			  struct virtio_gpu_object_params *params,
 			  struct drm_gem_object **obj_p,
 			  uint32_t *handle_p)
 {
+	struct virtio_gpu_device *vgdev = dev->dev_private;
 	struct virtio_gpu_object *obj;
 	int ret;
 	u32 handle;
 
-	obj = virtio_gpu_alloc_object(dev, params, NULL);
-	if (IS_ERR(obj))
-		return PTR_ERR(obj);
+	ret = virtio_gpu_object_create(vgdev, params, &obj, NULL);
+	if (ret < 0)
+		return ret;
 
-	ret = drm_gem_handle_create(file, &obj->gem_base, &handle);
+	ret = drm_gem_handle_create(file, &obj->base.base, &handle);
 	if (ret) {
-		drm_gem_object_release(&obj->gem_base);
+		drm_gem_object_release(&obj->base.base);
 		return ret;
 	}
 
-	*obj_p = &obj->gem_base;
+	*obj_p = &obj->base.base;
 
 	/* drop reference from allocate - handle holds it now */
-	drm_gem_object_put_unlocked(&obj->gem_base);
+	drm_gem_object_put_unlocked(&obj->base.base);
 
 	*handle_p = handle;
 	return 0;
@@ -136,19 +113,18 @@ int virtio_gpu_gem_object_open(struct drm_gem_object *obj,
 {
 	struct virtio_gpu_device *vgdev = obj->dev->dev_private;
 	struct virtio_gpu_fpriv *vfpriv = file->driver_priv;
-	struct virtio_gpu_object *qobj = gem_to_virtio_gpu_obj(obj);
-	int r;
+	struct virtio_gpu_object_array *objs;
 
 	if (!vgdev->has_virgl_3d)
 		return 0;
 
-	r = virtio_gpu_object_reserve(qobj, false);
-	if (r)
-		return r;
+	objs = virtio_gpu_array_alloc(1);
+	if (!objs)
+		return -ENOMEM;
+	virtio_gpu_array_add_obj(objs, obj);
 
 	virtio_gpu_cmd_context_attach_resource(vgdev, vfpriv->ctx_id,
-					       qobj->hw_res_handle);
-	virtio_gpu_object_unreserve(qobj);
+					       objs);
 	return 0;
 }
 
@@ -157,17 +133,136 @@ void virtio_gpu_gem_object_close(struct drm_gem_object *obj,
 {
 	struct virtio_gpu_device *vgdev = obj->dev->dev_private;
 	struct virtio_gpu_fpriv *vfpriv = file->driver_priv;
-	struct virtio_gpu_object *qobj = gem_to_virtio_gpu_obj(obj);
-	int r;
+	struct virtio_gpu_object_array *objs;
 
 	if (!vgdev->has_virgl_3d)
 		return;
 
-	r = virtio_gpu_object_reserve(qobj, false);
-	if (r)
+	objs = virtio_gpu_array_alloc(1);
+	if (!objs)
 		return;
+	virtio_gpu_array_add_obj(objs, obj);
 
 	virtio_gpu_cmd_context_detach_resource(vgdev, vfpriv->ctx_id,
-						qobj->hw_res_handle);
-	virtio_gpu_object_unreserve(qobj);
+					       objs);
+}
+
+struct virtio_gpu_object_array *virtio_gpu_array_alloc(u32 nents)
+{
+	struct virtio_gpu_object_array *objs;
+	size_t size = sizeof(*objs) + sizeof(objs->objs[0]) * nents;
+
+	objs = kmalloc(size, GFP_KERNEL);
+	if (!objs)
+		return NULL;
+
+	objs->nents = 0;
+	objs->total = nents;
+	return objs;
+}
+
+static void virtio_gpu_array_free(struct virtio_gpu_object_array *objs)
+{
+	kfree(objs);
+}
+
+struct virtio_gpu_object_array*
+virtio_gpu_array_from_handles(struct drm_file *drm_file, u32 *handles, u32 nents)
+{
+	struct virtio_gpu_object_array *objs;
+	u32 i;
+
+	objs = virtio_gpu_array_alloc(nents);
+	if (!objs)
+		return NULL;
+
+	for (i = 0; i < nents; i++) {
+		objs->objs[i] = drm_gem_object_lookup(drm_file, handles[i]);
+		if (!objs->objs[i]) {
+			objs->nents = i;
+			virtio_gpu_array_put_free(objs);
+			return NULL;
+		}
+	}
+	objs->nents = i;
+	return objs;
+}
+
+void virtio_gpu_array_add_obj(struct virtio_gpu_object_array *objs,
+			      struct drm_gem_object *obj)
+{
+	if (WARN_ON_ONCE(objs->nents == objs->total))
+		return;
+
+	drm_gem_object_get(obj);
+	objs->objs[objs->nents] = obj;
+	objs->nents++;
+}
+
+int virtio_gpu_array_lock_resv(struct virtio_gpu_object_array *objs)
+{
+	int ret;
+
+	if (objs->nents == 1) {
+		ret = dma_resv_lock_interruptible(objs->objs[0]->resv, NULL);
+	} else {
+		ret = drm_gem_lock_reservations(objs->objs, objs->nents,
+						&objs->ticket);
+	}
+	return ret;
+}
+
+void virtio_gpu_array_unlock_resv(struct virtio_gpu_object_array *objs)
+{
+	if (objs->nents == 1) {
+		dma_resv_unlock(objs->objs[0]->resv);
+	} else {
+		drm_gem_unlock_reservations(objs->objs, objs->nents,
+					    &objs->ticket);
+	}
+}
+
+void virtio_gpu_array_add_fence(struct virtio_gpu_object_array *objs,
+				struct dma_fence *fence)
+{
+	int i;
+
+	for (i = 0; i < objs->nents; i++)
+		dma_resv_add_excl_fence(objs->objs[i]->resv, fence);
+}
+
+void virtio_gpu_array_put_free(struct virtio_gpu_object_array *objs)
+{
+	u32 i;
+
+	for (i = 0; i < objs->nents; i++)
+		drm_gem_object_put_unlocked(objs->objs[i]);
+	virtio_gpu_array_free(objs);
+}
+
+void virtio_gpu_array_put_free_delayed(struct virtio_gpu_device *vgdev,
+				       struct virtio_gpu_object_array *objs)
+{
+	spin_lock(&vgdev->obj_free_lock);
+	list_add_tail(&objs->next, &vgdev->obj_free_list);
+	spin_unlock(&vgdev->obj_free_lock);
+	schedule_work(&vgdev->obj_free_work);
+}
+
+void virtio_gpu_array_put_free_work(struct work_struct *work)
+{
+	struct virtio_gpu_device *vgdev =
+		container_of(work, struct virtio_gpu_device, obj_free_work);
+	struct virtio_gpu_object_array *objs;
+
+	spin_lock(&vgdev->obj_free_lock);
+	while (!list_empty(&vgdev->obj_free_list)) {
+		objs = list_first_entry(&vgdev->obj_free_list,
+					struct virtio_gpu_object_array, next);
+		list_del(&objs->next);
+		spin_unlock(&vgdev->obj_free_lock);
+		virtio_gpu_array_put_free(objs);
+		spin_lock(&vgdev->obj_free_lock);
+	}
+	spin_unlock(&vgdev->obj_free_lock);
 }
diff --git a/drivers/gpu/drm/virtio/virtgpu_ioctl.c b/drivers/gpu/drm/virtio/virtgpu_ioctl.c
index 0a88ef11b9d3..9af1ec62434f 100644
--- a/drivers/gpu/drm/virtio/virtgpu_ioctl.c
+++ b/drivers/gpu/drm/virtio/virtgpu_ioctl.c
@@ -29,7 +29,6 @@
 #include <linux/sync_file.h>
 
 #include <drm/drm_file.h>
-#include <drm/ttm/ttm_execbuf_util.h>
 #include <drm/virtgpu_drm.h>
 
 #include "virtgpu_drv.h"
@@ -56,45 +55,6 @@ static int virtio_gpu_map_ioctl(struct drm_device *dev, void *data,
 					 &virtio_gpu_map->offset);
 }
 
-int virtio_gpu_object_list_validate(struct ww_acquire_ctx *ticket,
-				    struct list_head *head)
-{
-	struct ttm_operation_ctx ctx = { false, false };
-	struct ttm_validate_buffer *buf;
-	struct ttm_buffer_object *bo;
-	struct virtio_gpu_object *qobj;
-	int ret;
-
-	ret = ttm_eu_reserve_buffers(ticket, head, true, NULL, true);
-	if (ret != 0)
-		return ret;
-
-	list_for_each_entry(buf, head, head) {
-		bo = buf->bo;
-		qobj = container_of(bo, struct virtio_gpu_object, tbo);
-		ret = ttm_bo_validate(bo, &qobj->placement, &ctx);
-		if (ret) {
-			ttm_eu_backoff_reservation(ticket, head);
-			return ret;
-		}
-	}
-	return 0;
-}
-
-void virtio_gpu_unref_list(struct list_head *head)
-{
-	struct ttm_validate_buffer *buf;
-	struct ttm_buffer_object *bo;
-	struct virtio_gpu_object *qobj;
-
-	list_for_each_entry(buf, head, head) {
-		bo = buf->bo;
-		qobj = container_of(bo, struct virtio_gpu_object, tbo);
-
-		drm_gem_object_put_unlocked(&qobj->gem_base);
-	}
-}
-
 /*
  * Usage of execbuffer:
  * Relocations need to take into account the full VIRTIO_GPUDrawable size.
@@ -107,16 +67,11 @@ static int virtio_gpu_execbuffer_ioctl(struct drm_device *dev, void *data,
 	struct drm_virtgpu_execbuffer *exbuf = data;
 	struct virtio_gpu_device *vgdev = dev->dev_private;
 	struct virtio_gpu_fpriv *vfpriv = drm_file->driver_priv;
-	struct drm_gem_object *gobj;
 	struct virtio_gpu_fence *out_fence;
-	struct virtio_gpu_object *qobj;
 	int ret;
 	uint32_t *bo_handles = NULL;
 	void __user *user_bo_handles = NULL;
-	struct list_head validate_list;
-	struct ttm_validate_buffer *buflist = NULL;
-	int i;
-	struct ww_acquire_ctx ticket;
+	struct virtio_gpu_object_array *buflist = NULL;
 	struct sync_file *sync_file;
 	int in_fence_fd = exbuf->fence_fd;
 	int out_fence_fd = -1;
@@ -157,15 +112,10 @@ static int virtio_gpu_execbuffer_ioctl(struct drm_device *dev, void *data,
 			return out_fence_fd;
 	}
 
-	INIT_LIST_HEAD(&validate_list);
 	if (exbuf->num_bo_handles) {
-
 		bo_handles = kvmalloc_array(exbuf->num_bo_handles,
-					   sizeof(uint32_t), GFP_KERNEL);
-		buflist = kvmalloc_array(exbuf->num_bo_handles,
-					   sizeof(struct ttm_validate_buffer),
-					   GFP_KERNEL | __GFP_ZERO);
-		if (!bo_handles || !buflist) {
+					    sizeof(uint32_t), GFP_KERNEL);
+		if (!bo_handles) {
 			ret = -ENOMEM;
 			goto out_unused_fd;
 		}
@@ -177,27 +127,23 @@ static int virtio_gpu_execbuffer_ioctl(struct drm_device *dev, void *data,
 			goto out_unused_fd;
 		}
 
-		for (i = 0; i < exbuf->num_bo_handles; i++) {
-			gobj = drm_gem_object_lookup(drm_file, bo_handles[i]);
-			if (!gobj) {
-				ret = -ENOENT;
-				goto out_unused_fd;
-			}
-
-			qobj = gem_to_virtio_gpu_obj(gobj);
-			buflist[i].bo = &qobj->tbo;
-
-			list_add(&buflist[i].head, &validate_list);
+		buflist = virtio_gpu_array_from_handles(drm_file, bo_handles,
+							exbuf->num_bo_handles);
+		if (!buflist) {
+			ret = -ENOENT;
+			goto out_unused_fd;
 		}
 		kvfree(bo_handles);
 		bo_handles = NULL;
 	}
 
-	ret = virtio_gpu_object_list_validate(&ticket, &validate_list);
-	if (ret)
-		goto out_free;
+	if (buflist) {
+		ret = virtio_gpu_array_lock_resv(buflist);
+		if (ret)
+			goto out_unused_fd;
+	}
 
-	buf = memdup_user(u64_to_user_ptr(exbuf->command), exbuf->size);
+	buf = vmemdup_user(u64_to_user_ptr(exbuf->command), exbuf->size);
 	if (IS_ERR(buf)) {
 		ret = PTR_ERR(buf);
 		goto out_unresv;
@@ -222,24 +168,18 @@ static int virtio_gpu_execbuffer_ioctl(struct drm_device *dev, void *data,
 	}
 
 	virtio_gpu_cmd_submit(vgdev, buf, exbuf->size,
-			      vfpriv->ctx_id, out_fence);
-
-	ttm_eu_fence_buffer_objects(&ticket, &validate_list, &out_fence->f);
-
-	/* fence the command bo */
-	virtio_gpu_unref_list(&validate_list);
-	kvfree(buflist);
+			      vfpriv->ctx_id, buflist, out_fence);
 	return 0;
 
 out_memdup:
-	kfree(buf);
+	kvfree(buf);
 out_unresv:
-	ttm_eu_backoff_reservation(&ticket, &validate_list);
-out_free:
-	virtio_gpu_unref_list(&validate_list);
+	if (buflist)
+		virtio_gpu_array_unlock_resv(buflist);
 out_unused_fd:
 	kvfree(bo_handles);
-	kvfree(buflist);
+	if (buflist)
+		virtio_gpu_array_put_free(buflist);
 
 	if (out_fence_fd >= 0)
 		put_unused_fd(out_fence_fd);
@@ -316,11 +256,11 @@ static int virtio_gpu_resource_create_ioctl(struct drm_device *dev, void *data,
 	fence = virtio_gpu_fence_alloc(vgdev);
 	if (!fence)
 		return -ENOMEM;
-	qobj = virtio_gpu_alloc_object(dev, &params, fence);
+	ret = virtio_gpu_object_create(vgdev, &params, &qobj, fence);
 	dma_fence_put(&fence->f);
-	if (IS_ERR(qobj))
-		return PTR_ERR(qobj);
-	obj = &qobj->gem_base;
+	if (ret < 0)
+		return ret;
+	obj = &qobj->base.base;
 
 	ret = drm_gem_handle_create(file_priv, obj, &handle);
 	if (ret) {
@@ -347,7 +287,7 @@ static int virtio_gpu_resource_info_ioctl(struct drm_device *dev, void *data,
 
 	qobj = gem_to_virtio_gpu_obj(gobj);
 
-	ri->size = qobj->gem_base.size;
+	ri->size = qobj->base.base.size;
 	ri->res_handle = qobj->hw_res_handle;
 	drm_gem_object_put_unlocked(gobj);
 	return 0;
@@ -360,9 +300,7 @@ static int virtio_gpu_transfer_from_host_ioctl(struct drm_device *dev,
 	struct virtio_gpu_device *vgdev = dev->dev_private;
 	struct virtio_gpu_fpriv *vfpriv = file->driver_priv;
 	struct drm_virtgpu_3d_transfer_from_host *args = data;
-	struct ttm_operation_ctx ctx = { true, false };
-	struct drm_gem_object *gobj = NULL;
-	struct virtio_gpu_object *qobj = NULL;
+	struct virtio_gpu_object_array *objs;
 	struct virtio_gpu_fence *fence;
 	int ret;
 	u32 offset = args->offset;
@@ -371,39 +309,31 @@ static int virtio_gpu_transfer_from_host_ioctl(struct drm_device *dev,
 	if (vgdev->has_virgl_3d == false)
 		return -ENOSYS;
 
-	gobj = drm_gem_object_lookup(file, args->bo_handle);
-	if (gobj == NULL)
+	objs = virtio_gpu_array_from_handles(file, &args->bo_handle, 1);
+	if (objs == NULL)
 		return -ENOENT;
 
-	qobj = gem_to_virtio_gpu_obj(gobj);
-
-	ret = virtio_gpu_object_reserve(qobj, false);
-	if (ret)
-		goto out;
-
-	ret = ttm_bo_validate(&qobj->tbo, &qobj->placement, &ctx);
-	if (unlikely(ret))
-		goto out_unres;
+	ret = virtio_gpu_array_lock_resv(objs);
+	if (ret != 0)
+		goto err_put_free;
 
 	convert_to_hw_box(&box, &args->box);
 
 	fence = virtio_gpu_fence_alloc(vgdev);
 	if (!fence) {
 		ret = -ENOMEM;
-		goto out_unres;
+		goto err_unlock;
 	}
 	virtio_gpu_cmd_transfer_from_host_3d
-		(vgdev, qobj->hw_res_handle,
-		 vfpriv->ctx_id, offset, args->level,
-		 &box, fence);
-	dma_resv_add_excl_fence(qobj->tbo.base.resv,
-					  &fence->f);
-
+		(vgdev, vfpriv->ctx_id, offset, args->level,
+		 &box, objs, fence);
 	dma_fence_put(&fence->f);
-out_unres:
-	virtio_gpu_object_unreserve(qobj);
-out:
-	drm_gem_object_put_unlocked(gobj);
+	return 0;
+
+err_unlock:
+	virtio_gpu_array_unlock_resv(objs);
+err_put_free:
+	virtio_gpu_array_put_free(objs);
 	return ret;
 }
 
@@ -413,75 +343,71 @@ static int virtio_gpu_transfer_to_host_ioctl(struct drm_device *dev, void *data,
 	struct virtio_gpu_device *vgdev = dev->dev_private;
 	struct virtio_gpu_fpriv *vfpriv = file->driver_priv;
 	struct drm_virtgpu_3d_transfer_to_host *args = data;
-	struct ttm_operation_ctx ctx = { true, false };
-	struct drm_gem_object *gobj = NULL;
-	struct virtio_gpu_object *qobj = NULL;
+	struct virtio_gpu_object_array *objs;
 	struct virtio_gpu_fence *fence;
 	struct virtio_gpu_box box;
 	int ret;
 	u32 offset = args->offset;
 
-	gobj = drm_gem_object_lookup(file, args->bo_handle);
-	if (gobj == NULL)
+	objs = virtio_gpu_array_from_handles(file, &args->bo_handle, 1);
+	if (objs == NULL)
 		return -ENOENT;
 
-	qobj = gem_to_virtio_gpu_obj(gobj);
-
-	ret = virtio_gpu_object_reserve(qobj, false);
-	if (ret)
-		goto out;
-
-	ret = ttm_bo_validate(&qobj->tbo, &qobj->placement, &ctx);
-	if (unlikely(ret))
-		goto out_unres;
-
 	convert_to_hw_box(&box, &args->box);
 	if (!vgdev->has_virgl_3d) {
 		virtio_gpu_cmd_transfer_to_host_2d
-			(vgdev, qobj, offset,
-			 box.w, box.h, box.x, box.y, NULL);
+			(vgdev, offset,
+			 box.w, box.h, box.x, box.y,
+			 objs, NULL);
 	} else {
+		ret = virtio_gpu_array_lock_resv(objs);
+		if (ret != 0)
+			goto err_put_free;
+
+		ret = -ENOMEM;
 		fence = virtio_gpu_fence_alloc(vgdev);
-		if (!fence) {
-			ret = -ENOMEM;
-			goto out_unres;
-		}
+		if (!fence)
+			goto err_unlock;
+
 		virtio_gpu_cmd_transfer_to_host_3d
-			(vgdev, qobj,
+			(vgdev,
 			 vfpriv ? vfpriv->ctx_id : 0, offset,
-			 args->level, &box, fence);
-		dma_resv_add_excl_fence(qobj->tbo.base.resv,
-						  &fence->f);
+			 args->level, &box, objs, fence);
 		dma_fence_put(&fence->f);
 	}
+	return 0;
 
-out_unres:
-	virtio_gpu_object_unreserve(qobj);
-out:
-	drm_gem_object_put_unlocked(gobj);
+err_unlock:
+	virtio_gpu_array_unlock_resv(objs);
+err_put_free:
+	virtio_gpu_array_put_free(objs);
 	return ret;
 }
 
 static int virtio_gpu_wait_ioctl(struct drm_device *dev, void *data,
-			    struct drm_file *file)
+				 struct drm_file *file)
 {
 	struct drm_virtgpu_3d_wait *args = data;
-	struct drm_gem_object *gobj = NULL;
-	struct virtio_gpu_object *qobj = NULL;
+	struct drm_gem_object *obj;
+	long timeout = 15 * HZ;
 	int ret;
-	bool nowait = false;
 
-	gobj = drm_gem_object_lookup(file, args->handle);
-	if (gobj == NULL)
+	obj = drm_gem_object_lookup(file, args->handle);
+	if (obj == NULL)
 		return -ENOENT;
 
-	qobj = gem_to_virtio_gpu_obj(gobj);
-
-	if (args->flags & VIRTGPU_WAIT_NOWAIT)
-		nowait = true;
-	ret = virtio_gpu_object_wait(qobj, nowait);
+	if (args->flags & VIRTGPU_WAIT_NOWAIT) {
+		ret = dma_resv_test_signaled_rcu(obj->resv, true);
+	} else {
+		ret = dma_resv_wait_timeout_rcu(obj->resv, true, true,
+						timeout);
+	}
+	if (ret == 0)
+		ret = -EBUSY;
+	else if (ret > 0)
+		ret = 0;
 
-	drm_gem_object_put_unlocked(gobj);
+	drm_gem_object_put_unlocked(obj);
 	return ret;
 }
 
diff --git a/drivers/gpu/drm/virtio/virtgpu_kms.c b/drivers/gpu/drm/virtio/virtgpu_kms.c
index c190702fab72..2f5773e43557 100644
--- a/drivers/gpu/drm/virtio/virtgpu_kms.c
+++ b/drivers/gpu/drm/virtio/virtgpu_kms.c
@@ -147,19 +147,23 @@ int virtio_gpu_init(struct drm_device *dev)
 	INIT_WORK(&vgdev->config_changed_work,
 		  virtio_gpu_config_changed_work_func);
 
+	INIT_WORK(&vgdev->obj_free_work,
+		  virtio_gpu_array_put_free_work);
+	INIT_LIST_HEAD(&vgdev->obj_free_list);
+	spin_lock_init(&vgdev->obj_free_lock);
+
 #ifdef __LITTLE_ENDIAN
 	if (virtio_has_feature(vgdev->vdev, VIRTIO_GPU_F_VIRGL))
 		vgdev->has_virgl_3d = true;
-	DRM_INFO("virgl 3d acceleration %s\n",
-		 vgdev->has_virgl_3d ? "enabled" : "not supported by host");
-#else
-	DRM_INFO("virgl 3d acceleration not supported by guest\n");
 #endif
 	if (virtio_has_feature(vgdev->vdev, VIRTIO_GPU_F_EDID)) {
 		vgdev->has_edid = true;
-		DRM_INFO("EDID support available.\n");
 	}
 
+	DRM_INFO("features: %cvirgl %cedid\n",
+		 vgdev->has_virgl_3d ? '+' : '-',
+		 vgdev->has_edid     ? '+' : '-');
+
 	ret = virtio_find_vqs(vgdev->vdev, 2, vqs, callbacks, names, NULL);
 	if (ret) {
 		DRM_ERROR("failed to find virt queues\n");
@@ -173,12 +177,6 @@ int virtio_gpu_init(struct drm_device *dev)
 		goto err_vbufs;
 	}
 
-	ret = virtio_gpu_ttm_init(vgdev);
-	if (ret) {
-		DRM_ERROR("failed to init ttm %d\n", ret);
-		goto err_ttm;
-	}
-
 	/* get display info */
 	virtio_cread(vgdev->vdev, struct virtio_gpu_config,
 		     num_scanouts, &num_scanouts);
@@ -210,8 +208,6 @@ int virtio_gpu_init(struct drm_device *dev)
 	return 0;
 
 err_scanouts:
-	virtio_gpu_ttm_fini(vgdev);
-err_ttm:
 	virtio_gpu_free_vbufs(vgdev);
 err_vbufs:
 	vgdev->vdev->config->del_vqs(vgdev->vdev);
@@ -234,6 +230,7 @@ void virtio_gpu_deinit(struct drm_device *dev)
 {
 	struct virtio_gpu_device *vgdev = dev->dev_private;
 
+	flush_work(&vgdev->obj_free_work);
 	vgdev->vqs_ready = false;
 	flush_work(&vgdev->ctrlq.dequeue_work);
 	flush_work(&vgdev->cursorq.dequeue_work);
@@ -242,7 +239,6 @@ void virtio_gpu_deinit(struct drm_device *dev)
 	vgdev->vdev->config->del_vqs(vgdev->vdev);
 
 	virtio_gpu_modeset_fini(vgdev);
-	virtio_gpu_ttm_fini(vgdev);
 	virtio_gpu_free_vbufs(vgdev);
 	virtio_gpu_cleanup_cap_cache(vgdev);
 	kfree(vgdev->capsets);
diff --git a/drivers/gpu/drm/virtio/virtgpu_object.c b/drivers/gpu/drm/virtio/virtgpu_object.c
index 09b526518f5a..017a9e0fc3bb 100644
--- a/drivers/gpu/drm/virtio/virtgpu_object.c
+++ b/drivers/gpu/drm/virtio/virtgpu_object.c
@@ -23,73 +23,83 @@
  * WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
  */
 
-#include <drm/ttm/ttm_execbuf_util.h>
+#include <linux/moduleparam.h>
 
 #include "virtgpu_drv.h"
 
+static int virtio_gpu_virglrenderer_workaround = 1;
+module_param_named(virglhack, virtio_gpu_virglrenderer_workaround, int, 0400);
+
 static int virtio_gpu_resource_id_get(struct virtio_gpu_device *vgdev,
 				       uint32_t *resid)
 {
-#if 0
-	int handle = ida_alloc(&vgdev->resource_ida, GFP_KERNEL);
-
-	if (handle < 0)
-		return handle;
-#else
-	static int handle;
-
-	/*
-	 * FIXME: dirty hack to avoid re-using IDs, virglrenderer
-	 * can't deal with that.  Needs fixing in virglrenderer, also
-	 * should figure a better way to handle that in the guest.
-	 */
-	handle++;
-#endif
-
-	*resid = handle + 1;
+	if (virtio_gpu_virglrenderer_workaround) {
+		/*
+		 * Hack to avoid re-using resource IDs.
+		 *
+		 * virglrenderer versions up to (and including) 0.7.0
+		 * can't deal with that.  virglrenderer commit
+		 * "f91a9dd35715 Fix unlinking resources from hash
+		 * table." (Feb 2019) fixes the bug.
+		 */
+		static int handle;
+		handle++;
+		*resid = handle + 1;
+	} else {
+		int handle = ida_alloc(&vgdev->resource_ida, GFP_KERNEL);
+		if (handle < 0)
+			return handle;
+		*resid = handle + 1;
+	}
 	return 0;
 }
 
 static void virtio_gpu_resource_id_put(struct virtio_gpu_device *vgdev, uint32_t id)
 {
-#if 0
-	ida_free(&vgdev->resource_ida, id - 1);
-#endif
+	if (!virtio_gpu_virglrenderer_workaround) {
+		ida_free(&vgdev->resource_ida, id - 1);
+	}
 }
 
-static void virtio_gpu_ttm_bo_destroy(struct ttm_buffer_object *tbo)
+static void virtio_gpu_free_object(struct drm_gem_object *obj)
 {
-	struct virtio_gpu_object *bo;
-	struct virtio_gpu_device *vgdev;
-
-	bo = container_of(tbo, struct virtio_gpu_object, tbo);
-	vgdev = (struct virtio_gpu_device *)bo->gem_base.dev->dev_private;
+	struct virtio_gpu_object *bo = gem_to_virtio_gpu_obj(obj);
+	struct virtio_gpu_device *vgdev = bo->base.base.dev->dev_private;
 
+	if (bo->pages)
+		virtio_gpu_object_detach(vgdev, bo);
 	if (bo->created)
 		virtio_gpu_cmd_unref_resource(vgdev, bo->hw_res_handle);
-	if (bo->pages)
-		virtio_gpu_object_free_sg_table(bo);
-	if (bo->vmap)
-		virtio_gpu_object_kunmap(bo);
-	drm_gem_object_release(&bo->gem_base);
 	virtio_gpu_resource_id_put(vgdev, bo->hw_res_handle);
-	kfree(bo);
+
+	drm_gem_shmem_free_object(obj);
 }
 
-static void virtio_gpu_init_ttm_placement(struct virtio_gpu_object *vgbo)
+static const struct drm_gem_object_funcs virtio_gpu_gem_funcs = {
+	.free = virtio_gpu_free_object,
+	.open = virtio_gpu_gem_object_open,
+	.close = virtio_gpu_gem_object_close,
+
+	.print_info = drm_gem_shmem_print_info,
+	.pin = drm_gem_shmem_pin,
+	.unpin = drm_gem_shmem_unpin,
+	.get_sg_table = drm_gem_shmem_get_sg_table,
+	.vmap = drm_gem_shmem_vmap,
+	.vunmap = drm_gem_shmem_vunmap,
+	.mmap = &drm_gem_shmem_mmap,
+};
+
+struct drm_gem_object *virtio_gpu_create_object(struct drm_device *dev,
+						size_t size)
 {
-	u32 c = 1;
-
-	vgbo->placement.placement = &vgbo->placement_code;
-	vgbo->placement.busy_placement = &vgbo->placement_code;
-	vgbo->placement_code.fpfn = 0;
-	vgbo->placement_code.lpfn = 0;
-	vgbo->placement_code.flags =
-		TTM_PL_MASK_CACHING | TTM_PL_FLAG_TT |
-		TTM_PL_FLAG_NO_EVICT;
-	vgbo->placement.num_placement = c;
-	vgbo->placement.num_busy_placement = c;
+	struct virtio_gpu_object *bo;
+
+	bo = kzalloc(sizeof(*bo), GFP_KERNEL);
+	if (!bo)
+		return NULL;
 
+	bo->base.base.funcs = &virtio_gpu_gem_funcs;
+	return &bo->base.base;
 }
 
 int virtio_gpu_object_create(struct virtio_gpu_device *vgdev,
@@ -97,157 +107,59 @@ int virtio_gpu_object_create(struct virtio_gpu_device *vgdev,
 			     struct virtio_gpu_object **bo_ptr,
 			     struct virtio_gpu_fence *fence)
 {
+	struct virtio_gpu_object_array *objs = NULL;
+	struct drm_gem_shmem_object *shmem_obj;
 	struct virtio_gpu_object *bo;
-	size_t acc_size;
 	int ret;
 
 	*bo_ptr = NULL;
 
-	acc_size = ttm_bo_dma_acc_size(&vgdev->mman.bdev, params->size,
-				       sizeof(struct virtio_gpu_object));
+	params->size = roundup(params->size, PAGE_SIZE);
+	shmem_obj = drm_gem_shmem_create(vgdev->ddev, params->size);
+	if (IS_ERR(shmem_obj))
+		return PTR_ERR(shmem_obj);
+	bo = gem_to_virtio_gpu_obj(&shmem_obj->base);
 
-	bo = kzalloc(sizeof(struct virtio_gpu_object), GFP_KERNEL);
-	if (bo == NULL)
-		return -ENOMEM;
 	ret = virtio_gpu_resource_id_get(vgdev, &bo->hw_res_handle);
-	if (ret < 0) {
-		kfree(bo);
-		return ret;
-	}
-	params->size = roundup(params->size, PAGE_SIZE);
-	ret = drm_gem_object_init(vgdev->ddev, &bo->gem_base, params->size);
-	if (ret != 0) {
-		virtio_gpu_resource_id_put(vgdev, bo->hw_res_handle);
-		kfree(bo);
-		return ret;
-	}
+	if (ret < 0)
+		goto err_free_gem;
+
 	bo->dumb = params->dumb;
 
+	if (fence) {
+		ret = -ENOMEM;
+		objs = virtio_gpu_array_alloc(1);
+		if (!objs)
+			goto err_put_id;
+		virtio_gpu_array_add_obj(objs, &bo->base.base);
+
+		ret = virtio_gpu_array_lock_resv(objs);
+		if (ret != 0)
+			goto err_put_objs;
+	}
+
 	if (params->virgl) {
-		virtio_gpu_cmd_resource_create_3d(vgdev, bo, params, fence);
+		virtio_gpu_cmd_resource_create_3d(vgdev, bo, params,
+						  objs, fence);
 	} else {
-		virtio_gpu_cmd_create_resource(vgdev, bo, params, fence);
+		virtio_gpu_cmd_create_resource(vgdev, bo, params,
+					       objs, fence);
 	}
 
-	virtio_gpu_init_ttm_placement(bo);
-	ret = ttm_bo_init(&vgdev->mman.bdev, &bo->tbo, params->size,
-			  ttm_bo_type_device, &bo->placement, 0,
-			  true, acc_size, NULL, NULL,
-			  &virtio_gpu_ttm_bo_destroy);
-	/* ttm_bo_init failure will call the destroy */
-	if (ret != 0)
+	ret = virtio_gpu_object_attach(vgdev, bo, NULL);
+	if (ret != 0) {
+		virtio_gpu_free_object(&shmem_obj->base);
 		return ret;
-
-	if (fence) {
-		struct virtio_gpu_fence_driver *drv = &vgdev->fence_drv;
-		struct list_head validate_list;
-		struct ttm_validate_buffer mainbuf;
-		struct ww_acquire_ctx ticket;
-		unsigned long irq_flags;
-		bool signaled;
-
-		INIT_LIST_HEAD(&validate_list);
-		memset(&mainbuf, 0, sizeof(struct ttm_validate_buffer));
-
-		/* use a gem reference since unref list undoes them */
-		drm_gem_object_get(&bo->gem_base);
-		mainbuf.bo = &bo->tbo;
-		list_add(&mainbuf.head, &validate_list);
-
-		ret = virtio_gpu_object_list_validate(&ticket, &validate_list);
-		if (ret == 0) {
-			spin_lock_irqsave(&drv->lock, irq_flags);
-			signaled = virtio_fence_signaled(&fence->f);
-			if (!signaled)
-				/* virtio create command still in flight */
-				ttm_eu_fence_buffer_objects(&ticket, &validate_list,
-							    &fence->f);
-			spin_unlock_irqrestore(&drv->lock, irq_flags);
-			if (signaled)
-				/* virtio create command finished */
-				ttm_eu_backoff_reservation(&ticket, &validate_list);
-		}
-		virtio_gpu_unref_list(&validate_list);
 	}
 
 	*bo_ptr = bo;
 	return 0;
-}
-
-void virtio_gpu_object_kunmap(struct virtio_gpu_object *bo)
-{
-	bo->vmap = NULL;
-	ttm_bo_kunmap(&bo->kmap);
-}
-
-int virtio_gpu_object_kmap(struct virtio_gpu_object *bo)
-{
-	bool is_iomem;
-	int r;
-
-	WARN_ON(bo->vmap);
-
-	r = ttm_bo_kmap(&bo->tbo, 0, bo->tbo.num_pages, &bo->kmap);
-	if (r)
-		return r;
-	bo->vmap = ttm_kmap_obj_virtual(&bo->kmap, &is_iomem);
-	return 0;
-}
 
-int virtio_gpu_object_get_sg_table(struct virtio_gpu_device *qdev,
-				   struct virtio_gpu_object *bo)
-{
-	int ret;
-	struct page **pages = bo->tbo.ttm->pages;
-	int nr_pages = bo->tbo.num_pages;
-	struct ttm_operation_ctx ctx = {
-		.interruptible = false,
-		.no_wait_gpu = false
-	};
-	size_t max_segment;
-
-	/* wtf swapping */
-	if (bo->pages)
-		return 0;
-
-	if (bo->tbo.ttm->state == tt_unpopulated)
-		bo->tbo.ttm->bdev->driver->ttm_tt_populate(bo->tbo.ttm, &ctx);
-	bo->pages = kmalloc(sizeof(struct sg_table), GFP_KERNEL);
-	if (!bo->pages)
-		goto out;
-
-	max_segment = virtio_max_dma_size(qdev->vdev);
-	max_segment &= PAGE_MASK;
-	if (max_segment > SCATTERLIST_MAX_SEGMENT)
-		max_segment = SCATTERLIST_MAX_SEGMENT;
-	ret = __sg_alloc_table_from_pages(bo->pages, pages, nr_pages, 0,
-					  nr_pages << PAGE_SHIFT,
-					  max_segment, GFP_KERNEL);
-	if (ret)
-		goto out;
-	return 0;
-out:
-	kfree(bo->pages);
-	bo->pages = NULL;
-	return -ENOMEM;
-}
-
-void virtio_gpu_object_free_sg_table(struct virtio_gpu_object *bo)
-{
-	sg_free_table(bo->pages);
-	kfree(bo->pages);
-	bo->pages = NULL;
-}
-
-int virtio_gpu_object_wait(struct virtio_gpu_object *bo, bool no_wait)
-{
-	int r;
-
-	r = ttm_bo_reserve(&bo->tbo, true, no_wait, NULL);
-	if (unlikely(r != 0))
-		return r;
-	r = ttm_bo_wait(&bo->tbo, true, no_wait);
-	ttm_bo_unreserve(&bo->tbo);
-	return r;
+err_put_objs:
+	virtio_gpu_array_put_free(objs);
+err_put_id:
+	virtio_gpu_resource_id_put(vgdev, bo->hw_res_handle);
+err_free_gem:
+	drm_gem_shmem_free_object(&shmem_obj->base);
+	return ret;
 }
-
diff --git a/drivers/gpu/drm/virtio/virtgpu_plane.c b/drivers/gpu/drm/virtio/virtgpu_plane.c
index a492ac3f4a7e..390524143139 100644
--- a/drivers/gpu/drm/virtio/virtgpu_plane.c
+++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
@@ -84,7 +84,22 @@ static const struct drm_plane_funcs virtio_gpu_plane_funcs = {
 static int virtio_gpu_plane_atomic_check(struct drm_plane *plane,
 					 struct drm_plane_state *state)
 {
-	return 0;
+	bool is_cursor = plane->type == DRM_PLANE_TYPE_CURSOR;
+	struct drm_crtc_state *crtc_state;
+	int ret;
+
+	if (!state->fb || !state->crtc)
+		return 0;
+
+	crtc_state = drm_atomic_get_crtc_state(state->state, state->crtc);
+	if (IS_ERR(crtc_state))
+                return PTR_ERR(crtc_state);
+
+	ret = drm_atomic_helper_check_plane_state(state, crtc_state,
+						  DRM_PLANE_HELPER_NO_SCALING,
+						  DRM_PLANE_HELPER_NO_SCALING,
+						  is_cursor, true);
+	return ret;
 }
 
 static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
@@ -109,12 +124,19 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
 		bo = gem_to_virtio_gpu_obj(vgfb->base.obj[0]);
 		handle = bo->hw_res_handle;
 		if (bo->dumb) {
+			struct virtio_gpu_object_array *objs;
+
+			objs = virtio_gpu_array_alloc(1);
+			if (!objs)
+				return;
+			virtio_gpu_array_add_obj(objs, vgfb->base.obj[0]);
 			virtio_gpu_cmd_transfer_to_host_2d
-				(vgdev, bo, 0,
-				 cpu_to_le32(plane->state->src_w >> 16),
-				 cpu_to_le32(plane->state->src_h >> 16),
-				 cpu_to_le32(plane->state->src_x >> 16),
-				 cpu_to_le32(plane->state->src_y >> 16), NULL);
+				(vgdev, 0,
+				 plane->state->src_w >> 16,
+				 plane->state->src_h >> 16,
+				 plane->state->src_x >> 16,
+				 plane->state->src_y >> 16,
+				 objs, NULL);
 		}
 	} else {
 		handle = 0;
@@ -186,7 +208,6 @@ static void virtio_gpu_cursor_plane_update(struct drm_plane *plane,
 	struct virtio_gpu_framebuffer *vgfb;
 	struct virtio_gpu_object *bo = NULL;
 	uint32_t handle;
-	int ret = 0;
 
 	if (plane->state->crtc)
 		output = drm_crtc_to_virtio_gpu_output(plane->state->crtc);
@@ -205,20 +226,20 @@ static void virtio_gpu_cursor_plane_update(struct drm_plane *plane,
 
 	if (bo && bo->dumb && (plane->state->fb != old_state->fb)) {
 		/* new cursor -- update & wait */
+		struct virtio_gpu_object_array *objs;
+
+		objs = virtio_gpu_array_alloc(1);
+		if (!objs)
+			return;
+		virtio_gpu_array_add_obj(objs, vgfb->base.obj[0]);
 		virtio_gpu_cmd_transfer_to_host_2d
-			(vgdev, bo, 0,
-			 cpu_to_le32(plane->state->crtc_w),
-			 cpu_to_le32(plane->state->crtc_h),
-			 0, 0, vgfb->fence);
-		ret = virtio_gpu_object_reserve(bo, false);
-		if (!ret) {
-			dma_resv_add_excl_fence(bo->tbo.base.resv,
-							  &vgfb->fence->f);
-			dma_fence_put(&vgfb->fence->f);
-			vgfb->fence = NULL;
-			virtio_gpu_object_unreserve(bo);
-			virtio_gpu_object_wait(bo, false);
-		}
+			(vgdev, 0,
+			 plane->state->crtc_w,
+			 plane->state->crtc_h,
+			 0, 0, objs, vgfb->fence);
+		dma_fence_wait(&vgfb->fence->f, true);
+		dma_fence_put(&vgfb->fence->f);
+		vgfb->fence = NULL;
 	}
 
 	if (plane->state->fb != old_state->fb) {
diff --git a/drivers/gpu/drm/virtio/virtgpu_prime.c b/drivers/gpu/drm/virtio/virtgpu_prime.c
index dc642a884b88..050d24c39a8f 100644
--- a/drivers/gpu/drm/virtio/virtgpu_prime.c
+++ b/drivers/gpu/drm/virtio/virtgpu_prime.c
@@ -30,43 +30,9 @@
  * device that might share buffers with virtgpu
  */
 
-struct sg_table *virtgpu_gem_prime_get_sg_table(struct drm_gem_object *obj)
-{
-	struct virtio_gpu_object *bo = gem_to_virtio_gpu_obj(obj);
-
-	if (!bo->tbo.ttm->pages || !bo->tbo.ttm->num_pages)
-		/* should not happen */
-		return ERR_PTR(-EINVAL);
-
-	return drm_prime_pages_to_sg(bo->tbo.ttm->pages,
-				     bo->tbo.ttm->num_pages);
-}
-
 struct drm_gem_object *virtgpu_gem_prime_import_sg_table(
 	struct drm_device *dev, struct dma_buf_attachment *attach,
 	struct sg_table *table)
 {
 	return ERR_PTR(-ENODEV);
 }
-
-void *virtgpu_gem_prime_vmap(struct drm_gem_object *obj)
-{
-	struct virtio_gpu_object *bo = gem_to_virtio_gpu_obj(obj);
-	int ret;
-
-	ret = virtio_gpu_object_kmap(bo);
-	if (ret)
-		return NULL;
-	return bo->vmap;
-}
-
-void virtgpu_gem_prime_vunmap(struct drm_gem_object *obj, void *vaddr)
-{
-	virtio_gpu_object_kunmap(gem_to_virtio_gpu_obj(obj));
-}
-
-int virtgpu_gem_prime_mmap(struct drm_gem_object *obj,
-			   struct vm_area_struct *vma)
-{
-	return drm_gem_prime_mmap(obj, vma);
-}
diff --git a/drivers/gpu/drm/virtio/virtgpu_ttm.c b/drivers/gpu/drm/virtio/virtgpu_ttm.c
deleted file mode 100644
index f87903641847..000000000000
--- a/drivers/gpu/drm/virtio/virtgpu_ttm.c
+++ /dev/null
@@ -1,305 +0,0 @@
-/*
- * Copyright (C) 2015 Red Hat, Inc.
- * All Rights Reserved.
- *
- * Authors:
- *    Dave Airlie
- *    Alon Levy
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- */
-
-#include <linux/delay.h>
-
-#include <drm/drm.h>
-#include <drm/drm_file.h>
-#include <drm/ttm/ttm_bo_api.h>
-#include <drm/ttm/ttm_bo_driver.h>
-#include <drm/ttm/ttm_module.h>
-#include <drm/ttm/ttm_page_alloc.h>
-#include <drm/ttm/ttm_placement.h>
-#include <drm/virtgpu_drm.h>
-
-#include "virtgpu_drv.h"
-
-static struct
-virtio_gpu_device *virtio_gpu_get_vgdev(struct ttm_bo_device *bdev)
-{
-	struct virtio_gpu_mman *mman;
-	struct virtio_gpu_device *vgdev;
-
-	mman = container_of(bdev, struct virtio_gpu_mman, bdev);
-	vgdev = container_of(mman, struct virtio_gpu_device, mman);
-	return vgdev;
-}
-
-int virtio_gpu_mmap(struct file *filp, struct vm_area_struct *vma)
-{
-	struct drm_file *file_priv;
-	struct virtio_gpu_device *vgdev;
-	int r;
-
-	file_priv = filp->private_data;
-	vgdev = file_priv->minor->dev->dev_private;
-	if (vgdev == NULL) {
-		DRM_ERROR(
-		 "filp->private_data->minor->dev->dev_private == NULL\n");
-		return -EINVAL;
-	}
-	r = ttm_bo_mmap(filp, vma, &vgdev->mman.bdev);
-
-	return r;
-}
-
-static int virtio_gpu_invalidate_caches(struct ttm_bo_device *bdev,
-					uint32_t flags)
-{
-	return 0;
-}
-
-static int ttm_bo_man_get_node(struct ttm_mem_type_manager *man,
-			       struct ttm_buffer_object *bo,
-			       const struct ttm_place *place,
-			       struct ttm_mem_reg *mem)
-{
-	mem->mm_node = (void *)1;
-	return 0;
-}
-
-static void ttm_bo_man_put_node(struct ttm_mem_type_manager *man,
-				struct ttm_mem_reg *mem)
-{
-	mem->mm_node = (void *)NULL;
-}
-
-static int ttm_bo_man_init(struct ttm_mem_type_manager *man,
-			   unsigned long p_size)
-{
-	return 0;
-}
-
-static int ttm_bo_man_takedown(struct ttm_mem_type_manager *man)
-{
-	return 0;
-}
-
-static void ttm_bo_man_debug(struct ttm_mem_type_manager *man,
-			     struct drm_printer *printer)
-{
-}
-
-static const struct ttm_mem_type_manager_func virtio_gpu_bo_manager_func = {
-	.init = ttm_bo_man_init,
-	.takedown = ttm_bo_man_takedown,
-	.get_node = ttm_bo_man_get_node,
-	.put_node = ttm_bo_man_put_node,
-	.debug = ttm_bo_man_debug
-};
-
-static int virtio_gpu_init_mem_type(struct ttm_bo_device *bdev, uint32_t type,
-				    struct ttm_mem_type_manager *man)
-{
-	switch (type) {
-	case TTM_PL_SYSTEM:
-		/* System memory */
-		man->flags = TTM_MEMTYPE_FLAG_MAPPABLE;
-		man->available_caching = TTM_PL_MASK_CACHING;
-		man->default_caching = TTM_PL_FLAG_CACHED;
-		break;
-	case TTM_PL_TT:
-		man->func = &virtio_gpu_bo_manager_func;
-		man->flags = TTM_MEMTYPE_FLAG_MAPPABLE;
-		man->available_caching = TTM_PL_MASK_CACHING;
-		man->default_caching = TTM_PL_FLAG_CACHED;
-		break;
-	default:
-		DRM_ERROR("Unsupported memory type %u\n", (unsigned int)type);
-		return -EINVAL;
-	}
-	return 0;
-}
-
-static void virtio_gpu_evict_flags(struct ttm_buffer_object *bo,
-				struct ttm_placement *placement)
-{
-	static const struct ttm_place placements = {
-		.fpfn  = 0,
-		.lpfn  = 0,
-		.flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM,
-	};
-
-	placement->placement = &placements;
-	placement->busy_placement = &placements;
-	placement->num_placement = 1;
-	placement->num_busy_placement = 1;
-}
-
-static int virtio_gpu_verify_access(struct ttm_buffer_object *bo,
-				    struct file *filp)
-{
-	return 0;
-}
-
-static int virtio_gpu_ttm_io_mem_reserve(struct ttm_bo_device *bdev,
-					 struct ttm_mem_reg *mem)
-{
-	struct ttm_mem_type_manager *man = &bdev->man[mem->mem_type];
-
-	mem->bus.addr = NULL;
-	mem->bus.offset = 0;
-	mem->bus.size = mem->num_pages << PAGE_SHIFT;
-	mem->bus.base = 0;
-	mem->bus.is_iomem = false;
-	if (!(man->flags & TTM_MEMTYPE_FLAG_MAPPABLE))
-		return -EINVAL;
-	switch (mem->mem_type) {
-	case TTM_PL_SYSTEM:
-	case TTM_PL_TT:
-		/* system memory */
-		return 0;
-	default:
-		return -EINVAL;
-	}
-	return 0;
-}
-
-static void virtio_gpu_ttm_io_mem_free(struct ttm_bo_device *bdev,
-				       struct ttm_mem_reg *mem)
-{
-}
-
-/*
- * TTM backend functions.
- */
-struct virtio_gpu_ttm_tt {
-	struct ttm_dma_tt		ttm;
-	struct virtio_gpu_object        *obj;
-};
-
-static int virtio_gpu_ttm_tt_bind(struct ttm_tt *ttm,
-				  struct ttm_mem_reg *bo_mem)
-{
-	struct virtio_gpu_ttm_tt *gtt =
-		container_of(ttm, struct virtio_gpu_ttm_tt, ttm.ttm);
-	struct virtio_gpu_device *vgdev =
-		virtio_gpu_get_vgdev(gtt->obj->tbo.bdev);
-
-	virtio_gpu_object_attach(vgdev, gtt->obj, NULL);
-	return 0;
-}
-
-static int virtio_gpu_ttm_tt_unbind(struct ttm_tt *ttm)
-{
-	struct virtio_gpu_ttm_tt *gtt =
-		container_of(ttm, struct virtio_gpu_ttm_tt, ttm.ttm);
-	struct virtio_gpu_device *vgdev =
-		virtio_gpu_get_vgdev(gtt->obj->tbo.bdev);
-
-	virtio_gpu_object_detach(vgdev, gtt->obj);
-	return 0;
-}
-
-static void virtio_gpu_ttm_tt_destroy(struct ttm_tt *ttm)
-{
-	struct virtio_gpu_ttm_tt *gtt =
-		container_of(ttm, struct virtio_gpu_ttm_tt, ttm.ttm);
-
-	ttm_dma_tt_fini(&gtt->ttm);
-	kfree(gtt);
-}
-
-static struct ttm_backend_func virtio_gpu_tt_func = {
-	.bind = &virtio_gpu_ttm_tt_bind,
-	.unbind = &virtio_gpu_ttm_tt_unbind,
-	.destroy = &virtio_gpu_ttm_tt_destroy,
-};
-
-static struct ttm_tt *virtio_gpu_ttm_tt_create(struct ttm_buffer_object *bo,
-					       uint32_t page_flags)
-{
-	struct virtio_gpu_device *vgdev;
-	struct virtio_gpu_ttm_tt *gtt;
-
-	vgdev = virtio_gpu_get_vgdev(bo->bdev);
-	gtt = kzalloc(sizeof(struct virtio_gpu_ttm_tt), GFP_KERNEL);
-	if (gtt == NULL)
-		return NULL;
-	gtt->ttm.ttm.func = &virtio_gpu_tt_func;
-	gtt->obj = container_of(bo, struct virtio_gpu_object, tbo);
-	if (ttm_dma_tt_init(&gtt->ttm, bo, page_flags)) {
-		kfree(gtt);
-		return NULL;
-	}
-	return &gtt->ttm.ttm;
-}
-
-static void virtio_gpu_bo_swap_notify(struct ttm_buffer_object *tbo)
-{
-	struct virtio_gpu_object *bo;
-
-	bo = container_of(tbo, struct virtio_gpu_object, tbo);
-
-	if (bo->pages)
-		virtio_gpu_object_free_sg_table(bo);
-}
-
-static struct ttm_bo_driver virtio_gpu_bo_driver = {
-	.ttm_tt_create = &virtio_gpu_ttm_tt_create,
-	.invalidate_caches = &virtio_gpu_invalidate_caches,
-	.init_mem_type = &virtio_gpu_init_mem_type,
-	.eviction_valuable = ttm_bo_eviction_valuable,
-	.evict_flags = &virtio_gpu_evict_flags,
-	.verify_access = &virtio_gpu_verify_access,
-	.io_mem_reserve = &virtio_gpu_ttm_io_mem_reserve,
-	.io_mem_free = &virtio_gpu_ttm_io_mem_free,
-	.swap_notify = &virtio_gpu_bo_swap_notify,
-};
-
-int virtio_gpu_ttm_init(struct virtio_gpu_device *vgdev)
-{
-	int r;
-
-	/* No others user of address space so set it to 0 */
-	r = ttm_bo_device_init(&vgdev->mman.bdev,
-			       &virtio_gpu_bo_driver,
-			       vgdev->ddev->anon_inode->i_mapping,
-			       false);
-	if (r) {
-		DRM_ERROR("failed initializing buffer object driver(%d).\n", r);
-		goto err_dev_init;
-	}
-
-	r = ttm_bo_init_mm(&vgdev->mman.bdev, TTM_PL_TT, 0);
-	if (r) {
-		DRM_ERROR("Failed initializing GTT heap.\n");
-		goto err_mm_init;
-	}
-	return 0;
-
-err_mm_init:
-	ttm_bo_device_release(&vgdev->mman.bdev);
-err_dev_init:
-	return r;
-}
-
-void virtio_gpu_ttm_fini(struct virtio_gpu_device *vgdev)
-{
-	ttm_bo_device_release(&vgdev->mman.bdev);
-	DRM_INFO("virtio_gpu: ttm finalized\n");
-}
diff --git a/drivers/gpu/drm/virtio/virtgpu_vq.c b/drivers/gpu/drm/virtio/virtgpu_vq.c
index 7ac20490e1b4..74ad3bc3ebe8 100644
--- a/drivers/gpu/drm/virtio/virtgpu_vq.c
+++ b/drivers/gpu/drm/virtio/virtgpu_vq.c
@@ -155,7 +155,7 @@ static void free_vbuf(struct virtio_gpu_device *vgdev,
 {
 	if (vbuf->resp_size > MAX_INLINE_RESP_SIZE)
 		kfree(vbuf->resp_buf);
-	kfree(vbuf->data_buf);
+	kvfree(vbuf->data_buf);
 	kmem_cache_free(vgdev->vbufs, vbuf);
 }
 
@@ -192,7 +192,7 @@ void virtio_gpu_dequeue_ctrl_func(struct work_struct *work)
 	} while (!virtqueue_enable_cb(vgdev->ctrlq.vq));
 	spin_unlock(&vgdev->ctrlq.qlock);
 
-	list_for_each_entry_safe(entry, tmp, &reclaim_list, list) {
+	list_for_each_entry(entry, &reclaim_list, list) {
 		resp = (struct virtio_gpu_ctrl_hdr *)entry->resp_buf;
 
 		trace_virtio_gpu_cmd_response(vgdev->ctrlq.vq, resp);
@@ -219,14 +219,18 @@ void virtio_gpu_dequeue_ctrl_func(struct work_struct *work)
 		}
 		if (entry->resp_cb)
 			entry->resp_cb(vgdev, entry);
-
-		list_del(&entry->list);
-		free_vbuf(vgdev, entry);
 	}
 	wake_up(&vgdev->ctrlq.ack_queue);
 
 	if (fence_id)
 		virtio_gpu_fence_event_process(vgdev, fence_id);
+
+	list_for_each_entry_safe(entry, tmp, &reclaim_list, list) {
+		if (entry->objs)
+			virtio_gpu_array_put_free_delayed(vgdev, entry->objs);
+		list_del(&entry->list);
+		free_vbuf(vgdev, entry);
+	}
 }
 
 void virtio_gpu_dequeue_cursor_func(struct work_struct *work)
@@ -252,26 +256,67 @@ void virtio_gpu_dequeue_cursor_func(struct work_struct *work)
 	wake_up(&vgdev->cursorq.ack_queue);
 }
 
-static int virtio_gpu_queue_ctrl_buffer_locked(struct virtio_gpu_device *vgdev,
-					       struct virtio_gpu_vbuffer *vbuf)
+/* Create sg_table from a vmalloc'd buffer. */
+static struct sg_table *vmalloc_to_sgt(char *data, uint32_t size, int *sg_ents)
+{
+	int ret, s, i;
+	struct sg_table *sgt;
+	struct scatterlist *sg;
+	struct page *pg;
+
+	if (WARN_ON(!PAGE_ALIGNED(data)))
+		return NULL;
+
+	sgt = kmalloc(sizeof(*sgt), GFP_KERNEL);
+	if (!sgt)
+		return NULL;
+
+	*sg_ents = DIV_ROUND_UP(size, PAGE_SIZE);
+	ret = sg_alloc_table(sgt, *sg_ents, GFP_KERNEL);
+	if (ret) {
+		kfree(sgt);
+		return NULL;
+	}
+
+	for_each_sg(sgt->sgl, sg, *sg_ents, i) {
+		pg = vmalloc_to_page(data);
+		if (!pg) {
+			sg_free_table(sgt);
+			kfree(sgt);
+			return NULL;
+		}
+
+		s = min_t(int, PAGE_SIZE, size);
+		sg_set_page(sg, pg, s, 0);
+
+		size -= s;
+		data += s;
+	}
+
+	return sgt;
+}
+
+static bool virtio_gpu_queue_ctrl_buffer_locked(struct virtio_gpu_device *vgdev,
+						struct virtio_gpu_vbuffer *vbuf,
+						struct scatterlist *vout)
 		__releases(&vgdev->ctrlq.qlock)
 		__acquires(&vgdev->ctrlq.qlock)
 {
 	struct virtqueue *vq = vgdev->ctrlq.vq;
-	struct scatterlist *sgs[3], vcmd, vout, vresp;
+	struct scatterlist *sgs[3], vcmd, vresp;
 	int outcnt = 0, incnt = 0;
+	bool notify = false;
 	int ret;
 
 	if (!vgdev->vqs_ready)
-		return -ENODEV;
+		return notify;
 
 	sg_init_one(&vcmd, vbuf->buf, vbuf->size);
 	sgs[outcnt + incnt] = &vcmd;
 	outcnt++;
 
-	if (vbuf->data_size) {
-		sg_init_one(&vout, vbuf->data_buf, vbuf->data_size);
-		sgs[outcnt + incnt] = &vout;
+	if (vout) {
+		sgs[outcnt + incnt] = vout;
 		outcnt++;
 	}
 
@@ -292,32 +337,35 @@ retry:
 		trace_virtio_gpu_cmd_queue(vq,
 			(struct virtio_gpu_ctrl_hdr *)vbuf->buf);
 
-		virtqueue_kick(vq);
+		notify = virtqueue_kick_prepare(vq);
 	}
-
-	if (!ret)
-		ret = vq->num_free;
-	return ret;
-}
-
-static int virtio_gpu_queue_ctrl_buffer(struct virtio_gpu_device *vgdev,
-					struct virtio_gpu_vbuffer *vbuf)
-{
-	int rc;
-
-	spin_lock(&vgdev->ctrlq.qlock);
-	rc = virtio_gpu_queue_ctrl_buffer_locked(vgdev, vbuf);
-	spin_unlock(&vgdev->ctrlq.qlock);
-	return rc;
+	return notify;
 }
 
-static int virtio_gpu_queue_fenced_ctrl_buffer(struct virtio_gpu_device *vgdev,
-					       struct virtio_gpu_vbuffer *vbuf,
-					       struct virtio_gpu_ctrl_hdr *hdr,
-					       struct virtio_gpu_fence *fence)
+static void virtio_gpu_queue_fenced_ctrl_buffer(struct virtio_gpu_device *vgdev,
+						struct virtio_gpu_vbuffer *vbuf,
+						struct virtio_gpu_ctrl_hdr *hdr,
+						struct virtio_gpu_fence *fence)
 {
 	struct virtqueue *vq = vgdev->ctrlq.vq;
-	int rc;
+	struct scatterlist *vout = NULL, sg;
+	struct sg_table *sgt = NULL;
+	bool notify;
+	int outcnt = 0;
+
+	if (vbuf->data_size) {
+		if (is_vmalloc_addr(vbuf->data_buf)) {
+			sgt = vmalloc_to_sgt(vbuf->data_buf, vbuf->data_size,
+					     &outcnt);
+			if (!sgt)
+				return;
+			vout = sgt->sgl;
+		} else {
+			sg_init_one(&sg, vbuf->data_buf, vbuf->data_size);
+			vout = &sg;
+			outcnt = 1;
+		}
+	}
 
 again:
 	spin_lock(&vgdev->ctrlq.qlock);
@@ -330,29 +378,47 @@ again:
 	 * to wait for free space, which can result in fence ids being
 	 * submitted out-of-order.
 	 */
-	if (vq->num_free < 3) {
+	if (vq->num_free < 2 + outcnt) {
 		spin_unlock(&vgdev->ctrlq.qlock);
 		wait_event(vgdev->ctrlq.ack_queue, vq->num_free >= 3);
 		goto again;
 	}
 
-	if (fence)
+	if (hdr && fence) {
 		virtio_gpu_fence_emit(vgdev, hdr, fence);
-	rc = virtio_gpu_queue_ctrl_buffer_locked(vgdev, vbuf);
+		if (vbuf->objs) {
+			virtio_gpu_array_add_fence(vbuf->objs, &fence->f);
+			virtio_gpu_array_unlock_resv(vbuf->objs);
+		}
+	}
+	notify = virtio_gpu_queue_ctrl_buffer_locked(vgdev, vbuf, vout);
 	spin_unlock(&vgdev->ctrlq.qlock);
-	return rc;
+	if (notify)
+		virtqueue_notify(vgdev->ctrlq.vq);
+
+	if (sgt) {
+		sg_free_table(sgt);
+		kfree(sgt);
+	}
+}
+
+static void virtio_gpu_queue_ctrl_buffer(struct virtio_gpu_device *vgdev,
+					 struct virtio_gpu_vbuffer *vbuf)
+{
+	virtio_gpu_queue_fenced_ctrl_buffer(vgdev, vbuf, NULL, NULL);
 }
 
-static int virtio_gpu_queue_cursor(struct virtio_gpu_device *vgdev,
-				   struct virtio_gpu_vbuffer *vbuf)
+static void virtio_gpu_queue_cursor(struct virtio_gpu_device *vgdev,
+				    struct virtio_gpu_vbuffer *vbuf)
 {
 	struct virtqueue *vq = vgdev->cursorq.vq;
 	struct scatterlist *sgs[1], ccmd;
+	bool notify;
 	int ret;
 	int outcnt;
 
 	if (!vgdev->vqs_ready)
-		return -ENODEV;
+		return;
 
 	sg_init_one(&ccmd, vbuf->buf, vbuf->size);
 	sgs[0] = &ccmd;
@@ -370,14 +436,13 @@ retry:
 		trace_virtio_gpu_cmd_queue(vq,
 			(struct virtio_gpu_ctrl_hdr *)vbuf->buf);
 
-		virtqueue_kick(vq);
+		notify = virtqueue_kick_prepare(vq);
 	}
 
 	spin_unlock(&vgdev->cursorq.qlock);
 
-	if (!ret)
-		ret = vq->num_free;
-	return ret;
+	if (notify)
+		virtqueue_notify(vq);
 }
 
 /* just create gem objects for userspace and long lived objects,
@@ -388,6 +453,7 @@ retry:
 void virtio_gpu_cmd_create_resource(struct virtio_gpu_device *vgdev,
 				    struct virtio_gpu_object *bo,
 				    struct virtio_gpu_object_params *params,
+				    struct virtio_gpu_object_array *objs,
 				    struct virtio_gpu_fence *fence)
 {
 	struct virtio_gpu_resource_create_2d *cmd_p;
@@ -395,6 +461,7 @@ void virtio_gpu_cmd_create_resource(struct virtio_gpu_device *vgdev,
 
 	cmd_p = virtio_gpu_alloc_cmd(vgdev, &vbuf, sizeof(*cmd_p));
 	memset(cmd_p, 0, sizeof(*cmd_p));
+	vbuf->objs = objs;
 
 	cmd_p->hdr.type = cpu_to_le32(VIRTIO_GPU_CMD_RESOURCE_CREATE_2D);
 	cmd_p->resource_id = cpu_to_le32(bo->hw_res_handle);
@@ -481,12 +548,13 @@ void virtio_gpu_cmd_resource_flush(struct virtio_gpu_device *vgdev,
 }
 
 void virtio_gpu_cmd_transfer_to_host_2d(struct virtio_gpu_device *vgdev,
-					struct virtio_gpu_object *bo,
 					uint64_t offset,
-					__le32 width, __le32 height,
-					__le32 x, __le32 y,
+					uint32_t width, uint32_t height,
+					uint32_t x, uint32_t y,
+					struct virtio_gpu_object_array *objs,
 					struct virtio_gpu_fence *fence)
 {
+	struct virtio_gpu_object *bo = gem_to_virtio_gpu_obj(objs->objs[0]);
 	struct virtio_gpu_transfer_to_host_2d *cmd_p;
 	struct virtio_gpu_vbuffer *vbuf;
 	bool use_dma_api = !virtio_has_iommu_quirk(vgdev->vdev);
@@ -498,14 +566,15 @@ void virtio_gpu_cmd_transfer_to_host_2d(struct virtio_gpu_device *vgdev,
 
 	cmd_p = virtio_gpu_alloc_cmd(vgdev, &vbuf, sizeof(*cmd_p));
 	memset(cmd_p, 0, sizeof(*cmd_p));
+	vbuf->objs = objs;
 
 	cmd_p->hdr.type = cpu_to_le32(VIRTIO_GPU_CMD_TRANSFER_TO_HOST_2D);
 	cmd_p->resource_id = cpu_to_le32(bo->hw_res_handle);
 	cmd_p->offset = cpu_to_le64(offset);
-	cmd_p->r.width = width;
-	cmd_p->r.height = height;
-	cmd_p->r.x = x;
-	cmd_p->r.y = y;
+	cmd_p->r.width = cpu_to_le32(width);
+	cmd_p->r.height = cpu_to_le32(height);
+	cmd_p->r.x = cpu_to_le32(x);
+	cmd_p->r.y = cpu_to_le32(y);
 
 	virtio_gpu_queue_fenced_ctrl_buffer(vgdev, vbuf, &cmd_p->hdr, fence);
 }
@@ -826,34 +895,38 @@ void virtio_gpu_cmd_context_destroy(struct virtio_gpu_device *vgdev,
 
 void virtio_gpu_cmd_context_attach_resource(struct virtio_gpu_device *vgdev,
 					    uint32_t ctx_id,
-					    uint32_t resource_id)
+					    struct virtio_gpu_object_array *objs)
 {
+	struct virtio_gpu_object *bo = gem_to_virtio_gpu_obj(objs->objs[0]);
 	struct virtio_gpu_ctx_resource *cmd_p;
 	struct virtio_gpu_vbuffer *vbuf;
 
 	cmd_p = virtio_gpu_alloc_cmd(vgdev, &vbuf, sizeof(*cmd_p));
 	memset(cmd_p, 0, sizeof(*cmd_p));
+	vbuf->objs = objs;
 
 	cmd_p->hdr.type = cpu_to_le32(VIRTIO_GPU_CMD_CTX_ATTACH_RESOURCE);
 	cmd_p->hdr.ctx_id = cpu_to_le32(ctx_id);
-	cmd_p->resource_id = cpu_to_le32(resource_id);
+	cmd_p->resource_id = cpu_to_le32(bo->hw_res_handle);
 	virtio_gpu_queue_ctrl_buffer(vgdev, vbuf);
 
 }
 
 void virtio_gpu_cmd_context_detach_resource(struct virtio_gpu_device *vgdev,
 					    uint32_t ctx_id,
-					    uint32_t resource_id)
+					    struct virtio_gpu_object_array *objs)
 {
+	struct virtio_gpu_object *bo = gem_to_virtio_gpu_obj(objs->objs[0]);
 	struct virtio_gpu_ctx_resource *cmd_p;
 	struct virtio_gpu_vbuffer *vbuf;
 
 	cmd_p = virtio_gpu_alloc_cmd(vgdev, &vbuf, sizeof(*cmd_p));
 	memset(cmd_p, 0, sizeof(*cmd_p));
+	vbuf->objs = objs;
 
 	cmd_p->hdr.type = cpu_to_le32(VIRTIO_GPU_CMD_CTX_DETACH_RESOURCE);
 	cmd_p->hdr.ctx_id = cpu_to_le32(ctx_id);
-	cmd_p->resource_id = cpu_to_le32(resource_id);
+	cmd_p->resource_id = cpu_to_le32(bo->hw_res_handle);
 	virtio_gpu_queue_ctrl_buffer(vgdev, vbuf);
 }
 
@@ -861,6 +934,7 @@ void
 virtio_gpu_cmd_resource_create_3d(struct virtio_gpu_device *vgdev,
 				  struct virtio_gpu_object *bo,
 				  struct virtio_gpu_object_params *params,
+				  struct virtio_gpu_object_array *objs,
 				  struct virtio_gpu_fence *fence)
 {
 	struct virtio_gpu_resource_create_3d *cmd_p;
@@ -868,6 +942,7 @@ virtio_gpu_cmd_resource_create_3d(struct virtio_gpu_device *vgdev,
 
 	cmd_p = virtio_gpu_alloc_cmd(vgdev, &vbuf, sizeof(*cmd_p));
 	memset(cmd_p, 0, sizeof(*cmd_p));
+	vbuf->objs = objs;
 
 	cmd_p->hdr.type = cpu_to_le32(VIRTIO_GPU_CMD_RESOURCE_CREATE_3D);
 	cmd_p->resource_id = cpu_to_le32(bo->hw_res_handle);
@@ -888,12 +963,13 @@ virtio_gpu_cmd_resource_create_3d(struct virtio_gpu_device *vgdev,
 }
 
 void virtio_gpu_cmd_transfer_to_host_3d(struct virtio_gpu_device *vgdev,
-					struct virtio_gpu_object *bo,
 					uint32_t ctx_id,
 					uint64_t offset, uint32_t level,
 					struct virtio_gpu_box *box,
+					struct virtio_gpu_object_array *objs,
 					struct virtio_gpu_fence *fence)
 {
+	struct virtio_gpu_object *bo = gem_to_virtio_gpu_obj(objs->objs[0]);
 	struct virtio_gpu_transfer_host_3d *cmd_p;
 	struct virtio_gpu_vbuffer *vbuf;
 	bool use_dma_api = !virtio_has_iommu_quirk(vgdev->vdev);
@@ -906,6 +982,8 @@ void virtio_gpu_cmd_transfer_to_host_3d(struct virtio_gpu_device *vgdev,
 	cmd_p = virtio_gpu_alloc_cmd(vgdev, &vbuf, sizeof(*cmd_p));
 	memset(cmd_p, 0, sizeof(*cmd_p));
 
+	vbuf->objs = objs;
+
 	cmd_p->hdr.type = cpu_to_le32(VIRTIO_GPU_CMD_TRANSFER_TO_HOST_3D);
 	cmd_p->hdr.ctx_id = cpu_to_le32(ctx_id);
 	cmd_p->resource_id = cpu_to_le32(bo->hw_res_handle);
@@ -917,20 +995,24 @@ void virtio_gpu_cmd_transfer_to_host_3d(struct virtio_gpu_device *vgdev,
 }
 
 void virtio_gpu_cmd_transfer_from_host_3d(struct virtio_gpu_device *vgdev,
-					  uint32_t resource_id, uint32_t ctx_id,
+					  uint32_t ctx_id,
 					  uint64_t offset, uint32_t level,
 					  struct virtio_gpu_box *box,
+					  struct virtio_gpu_object_array *objs,
 					  struct virtio_gpu_fence *fence)
 {
+	struct virtio_gpu_object *bo = gem_to_virtio_gpu_obj(objs->objs[0]);
 	struct virtio_gpu_transfer_host_3d *cmd_p;
 	struct virtio_gpu_vbuffer *vbuf;
 
 	cmd_p = virtio_gpu_alloc_cmd(vgdev, &vbuf, sizeof(*cmd_p));
 	memset(cmd_p, 0, sizeof(*cmd_p));
 
+	vbuf->objs = objs;
+
 	cmd_p->hdr.type = cpu_to_le32(VIRTIO_GPU_CMD_TRANSFER_FROM_HOST_3D);
 	cmd_p->hdr.ctx_id = cpu_to_le32(ctx_id);
-	cmd_p->resource_id = cpu_to_le32(resource_id);
+	cmd_p->resource_id = cpu_to_le32(bo->hw_res_handle);
 	cmd_p->box = *box;
 	cmd_p->offset = cpu_to_le64(offset);
 	cmd_p->level = cpu_to_le32(level);
@@ -940,7 +1022,9 @@ void virtio_gpu_cmd_transfer_from_host_3d(struct virtio_gpu_device *vgdev,
 
 void virtio_gpu_cmd_submit(struct virtio_gpu_device *vgdev,
 			   void *data, uint32_t data_size,
-			   uint32_t ctx_id, struct virtio_gpu_fence *fence)
+			   uint32_t ctx_id,
+			   struct virtio_gpu_object_array *objs,
+			   struct virtio_gpu_fence *fence)
 {
 	struct virtio_gpu_cmd_submit *cmd_p;
 	struct virtio_gpu_vbuffer *vbuf;
@@ -950,6 +1034,7 @@ void virtio_gpu_cmd_submit(struct virtio_gpu_device *vgdev,
 
 	vbuf->data_buf = data;
 	vbuf->data_size = data_size;
+	vbuf->objs = objs;
 
 	cmd_p->hdr.type = cpu_to_le32(VIRTIO_GPU_CMD_SUBMIT_3D);
 	cmd_p->hdr.ctx_id = cpu_to_le32(ctx_id);
@@ -965,17 +1050,21 @@ int virtio_gpu_object_attach(struct virtio_gpu_device *vgdev,
 	bool use_dma_api = !virtio_has_iommu_quirk(vgdev->vdev);
 	struct virtio_gpu_mem_entry *ents;
 	struct scatterlist *sg;
-	int si, nents;
+	int si, nents, ret;
 
 	if (WARN_ON_ONCE(!obj->created))
 		return -EINVAL;
+	if (WARN_ON_ONCE(obj->pages))
+		return -EINVAL;
 
-	if (!obj->pages) {
-		int ret;
+	ret = drm_gem_shmem_pin(&obj->base.base);
+	if (ret < 0)
+		return -EINVAL;
 
-		ret = virtio_gpu_object_get_sg_table(vgdev, obj);
-		if (ret)
-			return ret;
+	obj->pages = drm_gem_shmem_get_sg_table(&obj->base.base);
+	if (obj->pages == NULL) {
+		drm_gem_shmem_unpin(&obj->base.base);
+		return -EINVAL;
 	}
 
 	if (use_dma_api) {
@@ -1014,6 +1103,9 @@ void virtio_gpu_object_detach(struct virtio_gpu_device *vgdev,
 {
 	bool use_dma_api = !virtio_has_iommu_quirk(vgdev->vdev);
 
+	if (WARN_ON_ONCE(!obj->pages))
+		return;
+
 	if (use_dma_api && obj->mapped) {
 		struct virtio_gpu_fence *fence = virtio_gpu_fence_alloc(vgdev);
 		/* detach backing and wait for the host process it ... */
@@ -1029,6 +1121,11 @@ void virtio_gpu_object_detach(struct virtio_gpu_device *vgdev,
 	} else {
 		virtio_gpu_cmd_resource_inval_backing(vgdev, obj->hw_res_handle, NULL);
 	}
+
+	sg_free_table(obj->pages);
+	obj->pages = NULL;
+
+	drm_gem_shmem_unpin(&obj->base.base);
 }
 
 void virtio_gpu_cursor_ping(struct virtio_gpu_device *vgdev,
diff --git a/drivers/gpu/drm/vkms/vkms_crtc.c b/drivers/gpu/drm/vkms/vkms_crtc.c
index 927dafaebc76..74f703b8d22a 100644
--- a/drivers/gpu/drm/vkms/vkms_crtc.c
+++ b/drivers/gpu/drm/vkms/vkms_crtc.c
@@ -16,17 +16,18 @@ static enum hrtimer_restart vkms_vblank_simulate(struct hrtimer *timer)
 	u64 ret_overrun;
 	bool ret;
 
-	spin_lock(&output->lock);
-
 	ret_overrun = hrtimer_forward_now(&output->vblank_hrtimer,
 					  output->period_ns);
 	WARN_ON(ret_overrun != 1);
 
+	spin_lock(&output->lock);
 	ret = drm_crtc_handle_vblank(crtc);
 	if (!ret)
 		DRM_ERROR("vkms failure on handling vblank");
 
 	state = output->composer_state;
+	spin_unlock(&output->lock);
+
 	if (state && output->composer_enabled) {
 		u64 frame = drm_crtc_accurate_vblank_count(crtc);
 
@@ -48,8 +49,6 @@ static enum hrtimer_restart vkms_vblank_simulate(struct hrtimer *timer)
 			DRM_DEBUG_DRIVER("Composer worker already queued\n");
 	}
 
-	spin_unlock(&output->lock);
-
 	return HRTIMER_RESTART;
 }
 
@@ -85,7 +84,7 @@ bool vkms_get_vblank_timestamp(struct drm_device *dev, unsigned int pipe,
 	struct vkms_output *output = &vkmsdev->output;
 	struct drm_vblank_crtc *vblank = &dev->vblank[pipe];
 
-	*vblank_time = output->vblank_hrtimer.node.expires;
+	*vblank_time = READ_ONCE(output->vblank_hrtimer.node.expires);
 
 	if (WARN_ON(*vblank_time == vblank->time))
 		return true;
diff --git a/drivers/gpu/drm/vkms/vkms_drv.c b/drivers/gpu/drm/vkms/vkms_drv.c
index 44ab9f8ef8be..d1fe144aa289 100644
--- a/drivers/gpu/drm/vkms/vkms_drv.c
+++ b/drivers/gpu/drm/vkms/vkms_drv.c
@@ -11,13 +11,14 @@
 
 #include <linux/module.h>
 #include <linux/platform_device.h>
+#include <linux/dma-mapping.h>
 
+#include <drm/drm_gem.h>
 #include <drm/drm_atomic.h>
 #include <drm/drm_atomic_helper.h>
 #include <drm/drm_drv.h>
 #include <drm/drm_fb_helper.h>
 #include <drm/drm_file.h>
-#include <drm/drm_gem.h>
 #include <drm/drm_gem_framebuffer_helper.h>
 #include <drm/drm_ioctl.h>
 #include <drm/drm_probe_helper.h>
@@ -83,7 +84,7 @@ static void vkms_atomic_commit_tail(struct drm_atomic_state *old_state)
 
 	drm_atomic_helper_commit_hw_done(old_state);
 
-	drm_atomic_helper_wait_for_vblanks(dev, old_state);
+	drm_atomic_helper_wait_for_flip_done(dev, old_state);
 
 	for_each_old_crtc_in_state(old_state, crtc, old_crtc_state, i) {
 		struct vkms_crtc_state *vkms_state =
@@ -103,6 +104,8 @@ static struct drm_driver vkms_driver = {
 	.gem_vm_ops		= &vkms_gem_vm_ops,
 	.gem_free_object_unlocked = vkms_gem_free_object,
 	.get_vblank_timestamp	= vkms_get_vblank_timestamp,
+	.prime_fd_to_handle	= drm_gem_prime_fd_to_handle,
+	.gem_prime_import_sg_table = vkms_prime_import_sg_table,
 
 	.name			= DRIVER_NAME,
 	.desc			= DRIVER_DESC,
@@ -157,6 +160,14 @@ static int __init vkms_init(void)
 	if (ret)
 		goto out_unregister;
 
+	ret = dma_coerce_mask_and_coherent(vkms_device->drm.dev,
+					   DMA_BIT_MASK(64));
+
+	if (ret) {
+		DRM_ERROR("Could not initialize DMA support\n");
+		goto out_fini;
+	}
+
 	vkms_device->drm.irq_enabled = true;
 
 	ret = drm_vblank_init(&vkms_device->drm, 1);
diff --git a/drivers/gpu/drm/vkms/vkms_drv.h b/drivers/gpu/drm/vkms/vkms_drv.h
index 5a95100fa18b..7d52e24564db 100644
--- a/drivers/gpu/drm/vkms/vkms_drv.h
+++ b/drivers/gpu/drm/vkms/vkms_drv.h
@@ -137,6 +137,12 @@ int vkms_gem_vmap(struct drm_gem_object *obj);
 
 void vkms_gem_vunmap(struct drm_gem_object *obj);
 
+/* Prime */
+struct drm_gem_object *
+vkms_prime_import_sg_table(struct drm_device *dev,
+			   struct dma_buf_attachment *attach,
+			   struct sg_table *sg);
+
 /* CRC Support */
 const char *const *vkms_get_crc_sources(struct drm_crtc *crtc,
 					size_t *count);
diff --git a/drivers/gpu/drm/vkms/vkms_gem.c b/drivers/gpu/drm/vkms/vkms_gem.c
index 6489bfe0a149..2e01186fb943 100644
--- a/drivers/gpu/drm/vkms/vkms_gem.c
+++ b/drivers/gpu/drm/vkms/vkms_gem.c
@@ -1,7 +1,9 @@
 // SPDX-License-Identifier: GPL-2.0+
 
+#include <linux/dma-buf.h>
 #include <linux/shmem_fs.h>
 #include <linux/vmalloc.h>
+#include <drm/drm_prime.h>
 
 #include "vkms_drv.h"
 
@@ -218,3 +220,28 @@ out:
 	mutex_unlock(&vkms_obj->pages_lock);
 	return ret;
 }
+
+struct drm_gem_object *
+vkms_prime_import_sg_table(struct drm_device *dev,
+			   struct dma_buf_attachment *attach,
+			   struct sg_table *sg)
+{
+	struct vkms_gem_object *obj;
+	int npages;
+
+	obj = __vkms_gem_create(dev, attach->dmabuf->size);
+	if (IS_ERR(obj))
+		return ERR_CAST(obj);
+
+	npages = PAGE_ALIGN(attach->dmabuf->size) / PAGE_SIZE;
+	DRM_DEBUG_PRIME("Importing %d pages\n", npages);
+
+	obj->pages = kvmalloc_array(npages, sizeof(struct page *), GFP_KERNEL);
+	if (!obj->pages) {
+		vkms_gem_free_object(&obj->gem);
+		return ERR_PTR(-ENOMEM);
+	}
+
+	drm_prime_sg_to_page_addr_arrays(sg, obj->pages, NULL, npages);
+	return &obj->gem;
+}
diff --git a/drivers/gpu/drm/vmwgfx/Kconfig b/drivers/gpu/drm/vmwgfx/Kconfig
index 6b28a326f8bb..15acdf2a7c0f 100644
--- a/drivers/gpu/drm/vmwgfx/Kconfig
+++ b/drivers/gpu/drm/vmwgfx/Kconfig
@@ -8,6 +8,7 @@ config DRM_VMWGFX
 	select FB_CFB_IMAGEBLIT
 	select DRM_TTM
 	select FB
+	select MAPPING_DIRTY_HELPERS
 	# Only needed for the transitional use of drm_crtc_init - can be removed
 	# again once vmwgfx sets up the primary plane itself.
 	select DRM_KMS_HELPER
diff --git a/drivers/gpu/drm/vmwgfx/Makefile b/drivers/gpu/drm/vmwgfx/Makefile
index 8841bd30e1e5..c877a21a0739 100644
--- a/drivers/gpu/drm/vmwgfx/Makefile
+++ b/drivers/gpu/drm/vmwgfx/Makefile
@@ -8,7 +8,7 @@ vmwgfx-y := vmwgfx_execbuf.o vmwgfx_gmr.o vmwgfx_kms.o vmwgfx_drv.o \
 	    vmwgfx_cmdbuf_res.o vmwgfx_cmdbuf.o vmwgfx_stdu.o \
 	    vmwgfx_cotable.o vmwgfx_so.o vmwgfx_binding.o vmwgfx_msg.o \
 	    vmwgfx_simple_resource.o vmwgfx_va.o vmwgfx_blit.o \
-	    vmwgfx_validation.o \
+	    vmwgfx_validation.o vmwgfx_page_dirty.o \
 	    ttm_object.o ttm_lock.o
 
 obj-$(CONFIG_DRM_VMWGFX) := vmwgfx.o
diff --git a/drivers/gpu/drm/vmwgfx/device_include/svga3d_surfacedefs.h b/drivers/gpu/drm/vmwgfx/device_include/svga3d_surfacedefs.h
index f2bfd3d80598..61414f105c67 100644
--- a/drivers/gpu/drm/vmwgfx/device_include/svga3d_surfacedefs.h
+++ b/drivers/gpu/drm/vmwgfx/device_include/svga3d_surfacedefs.h
@@ -1280,7 +1280,6 @@ svga3dsurface_get_pixel_offset(SVGA3dSurfaceFormat format,
 	return offset;
 }
 
-
 static inline u32
 svga3dsurface_get_image_offset(SVGA3dSurfaceFormat format,
 			       surf_size_struct baseLevelSize,
@@ -1375,4 +1374,236 @@ svga3dsurface_is_screen_target_format(SVGA3dSurfaceFormat format)
 	return svga3dsurface_is_dx_screen_target_format(format);
 }
 
+/**
+ * struct svga3dsurface_mip - Mimpmap level information
+ * @bytes: Bytes required in the backing store of this mipmap level.
+ * @img_stride: Byte stride per image.
+ * @row_stride: Byte stride per block row.
+ * @size: The size of the mipmap.
+ */
+struct svga3dsurface_mip {
+	size_t bytes;
+	size_t img_stride;
+	size_t row_stride;
+	struct drm_vmw_size size;
+
+};
+
+/**
+ * struct svga3dsurface_cache - Cached surface information
+ * @desc: Pointer to the surface descriptor
+ * @mip: Array of mipmap level information. Valid size is @num_mip_levels.
+ * @mip_chain_bytes: Bytes required in the backing store for the whole chain
+ * of mip levels.
+ * @sheet_bytes: Bytes required in the backing store for a sheet
+ * representing a single sample.
+ * @num_mip_levels: Valid size of the @mip array. Number of mipmap levels in
+ * a chain.
+ * @num_layers: Number of slices in an array texture or number of faces in
+ * a cubemap texture.
+ */
+struct svga3dsurface_cache {
+	const struct svga3d_surface_desc *desc;
+	struct svga3dsurface_mip mip[DRM_VMW_MAX_MIP_LEVELS];
+	size_t mip_chain_bytes;
+	size_t sheet_bytes;
+	u32 num_mip_levels;
+	u32 num_layers;
+};
+
+/**
+ * struct svga3dsurface_loc - Surface location
+ * @sub_resource: Surface subresource. Defined as layer * num_mip_levels +
+ * mip_level.
+ * @x: X coordinate.
+ * @y: Y coordinate.
+ * @z: Z coordinate.
+ */
+struct svga3dsurface_loc {
+	u32 sub_resource;
+	u32 x, y, z;
+};
+
+/**
+ * svga3dsurface_subres - Compute the subresource from layer and mipmap.
+ * @cache: Surface layout data.
+ * @mip_level: The mipmap level.
+ * @layer: The surface layer (face or array slice).
+ *
+ * Return: The subresource.
+ */
+static inline u32 svga3dsurface_subres(const struct svga3dsurface_cache *cache,
+				       u32 mip_level, u32 layer)
+{
+	return cache->num_mip_levels * layer + mip_level;
+}
+
+/**
+ * svga3dsurface_setup_cache - Build a surface cache entry
+ * @size: The surface base level dimensions.
+ * @format: The surface format.
+ * @num_mip_levels: Number of mipmap levels.
+ * @num_layers: Number of layers.
+ * @cache: Pointer to a struct svga3dsurface_cach object to be filled in.
+ *
+ * Return: Zero on success, -EINVAL on invalid surface layout.
+ */
+static inline int svga3dsurface_setup_cache(const struct drm_vmw_size *size,
+					    SVGA3dSurfaceFormat format,
+					    u32 num_mip_levels,
+					    u32 num_layers,
+					    u32 num_samples,
+					    struct svga3dsurface_cache *cache)
+{
+	const struct svga3d_surface_desc *desc;
+	u32 i;
+
+	memset(cache, 0, sizeof(*cache));
+	cache->desc = desc = svga3dsurface_get_desc(format);
+	cache->num_mip_levels = num_mip_levels;
+	cache->num_layers = num_layers;
+	for (i = 0; i < cache->num_mip_levels; i++) {
+		struct svga3dsurface_mip *mip = &cache->mip[i];
+
+		mip->size = svga3dsurface_get_mip_size(*size, i);
+		mip->bytes = svga3dsurface_get_image_buffer_size
+			(desc, &mip->size, 0);
+		mip->row_stride =
+			__KERNEL_DIV_ROUND_UP(mip->size.width,
+					      desc->block_size.width) *
+			desc->bytes_per_block * num_samples;
+		if (!mip->row_stride)
+			goto invalid_dim;
+
+		mip->img_stride =
+			__KERNEL_DIV_ROUND_UP(mip->size.height,
+					      desc->block_size.height) *
+			mip->row_stride;
+		if (!mip->img_stride)
+			goto invalid_dim;
+
+		cache->mip_chain_bytes += mip->bytes;
+	}
+	cache->sheet_bytes = cache->mip_chain_bytes * num_layers;
+	if (!cache->sheet_bytes)
+		goto invalid_dim;
+
+	return 0;
+
+invalid_dim:
+	VMW_DEBUG_USER("Invalid surface layout for dirty tracking.\n");
+	return -EINVAL;
+}
+
+/**
+ * svga3dsurface_get_loc - Get a surface location from an offset into the
+ * backing store
+ * @cache: Surface layout data.
+ * @loc: Pointer to a struct svga3dsurface_loc to be filled in.
+ * @offset: Offset into the surface backing store.
+ */
+static inline void
+svga3dsurface_get_loc(const struct svga3dsurface_cache *cache,
+		      struct svga3dsurface_loc *loc,
+		      size_t offset)
+{
+	const struct svga3dsurface_mip *mip = &cache->mip[0];
+	const struct svga3d_surface_desc *desc = cache->desc;
+	u32 layer;
+	int i;
+
+	if (offset >= cache->sheet_bytes)
+		offset %= cache->sheet_bytes;
+
+	layer = offset / cache->mip_chain_bytes;
+	offset -= layer * cache->mip_chain_bytes;
+	for (i = 0; i < cache->num_mip_levels; ++i, ++mip) {
+		if (mip->bytes > offset)
+			break;
+		offset -= mip->bytes;
+	}
+
+	loc->sub_resource = svga3dsurface_subres(cache, i, layer);
+	loc->z = offset / mip->img_stride;
+	offset -= loc->z * mip->img_stride;
+	loc->z *= desc->block_size.depth;
+	loc->y = offset / mip->row_stride;
+	offset -= loc->y * mip->row_stride;
+	loc->y *= desc->block_size.height;
+	loc->x = offset / desc->bytes_per_block;
+	loc->x *= desc->block_size.width;
+}
+
+/**
+ * svga3dsurface_inc_loc - Clamp increment a surface location with one block
+ * size
+ * in each dimension.
+ * @loc: Pointer to a struct svga3dsurface_loc to be incremented.
+ *
+ * When computing the size of a range as size = end - start, the range does not
+ * include the end element. However a location representing the last byte
+ * of a touched region in the backing store *is* included in the range.
+ * This function modifies such a location to match the end definition
+ * given as start + size which is the one used in a SVGA3dBox.
+ */
+static inline void
+svga3dsurface_inc_loc(const struct svga3dsurface_cache *cache,
+		      struct svga3dsurface_loc *loc)
+{
+	const struct svga3d_surface_desc *desc = cache->desc;
+	u32 mip = loc->sub_resource % cache->num_mip_levels;
+	const struct drm_vmw_size *size = &cache->mip[mip].size;
+
+	loc->sub_resource++;
+	loc->x += desc->block_size.width;
+	if (loc->x > size->width)
+		loc->x = size->width;
+	loc->y += desc->block_size.height;
+	if (loc->y > size->height)
+		loc->y = size->height;
+	loc->z += desc->block_size.depth;
+	if (loc->z > size->depth)
+		loc->z = size->depth;
+}
+
+/**
+ * svga3dsurface_min_loc - The start location in a subresource
+ * @cache: Surface layout data.
+ * @sub_resource: The subresource.
+ * @loc: Pointer to a struct svga3dsurface_loc to be filled in.
+ */
+static inline void
+svga3dsurface_min_loc(const struct svga3dsurface_cache *cache,
+		      u32 sub_resource,
+		      struct svga3dsurface_loc *loc)
+{
+	loc->sub_resource = sub_resource;
+	loc->x = loc->y = loc->z = 0;
+}
+
+/**
+ * svga3dsurface_min_loc - The end location in a subresource
+ * @cache: Surface layout data.
+ * @sub_resource: The subresource.
+ * @loc: Pointer to a struct svga3dsurface_loc to be filled in.
+ *
+ * Following the end definition given in svga3dsurface_inc_loc(),
+ * Compute the end location of a surface subresource.
+ */
+static inline void
+svga3dsurface_max_loc(const struct svga3dsurface_cache *cache,
+		      u32 sub_resource,
+		      struct svga3dsurface_loc *loc)
+{
+	const struct drm_vmw_size *size;
+	u32 mip;
+
+	loc->sub_resource = sub_resource + 1;
+	mip = sub_resource % cache->num_mip_levels;
+	size = &cache->mip[mip].size;
+	loc->x = size->width;
+	loc->y = size->height;
+	loc->z = size->depth;
+}
+
 #endif /* _SVGA3D_SURFACEDEFS_H_ */
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_bo.c b/drivers/gpu/drm/vmwgfx/vmwgfx_bo.c
index aad8d8140259..8b71bf6b58ef 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_bo.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_bo.c
@@ -462,6 +462,8 @@ void vmw_bo_bo_free(struct ttm_buffer_object *bo)
 {
 	struct vmw_buffer_object *vmw_bo = vmw_buffer_object(bo);
 
+	WARN_ON(vmw_bo->dirty);
+	WARN_ON(!RB_EMPTY_ROOT(&vmw_bo->res_tree));
 	vmw_bo_unmap(vmw_bo);
 	kfree(vmw_bo);
 }
@@ -475,8 +477,11 @@ void vmw_bo_bo_free(struct ttm_buffer_object *bo)
 static void vmw_user_bo_destroy(struct ttm_buffer_object *bo)
 {
 	struct vmw_user_buffer_object *vmw_user_bo = vmw_user_buffer_object(bo);
+	struct vmw_buffer_object *vbo = &vmw_user_bo->vbo;
 
-	vmw_bo_unmap(&vmw_user_bo->vbo);
+	WARN_ON(vbo->dirty);
+	WARN_ON(!RB_EMPTY_ROOT(&vbo->res_tree));
+	vmw_bo_unmap(vbo);
 	ttm_prime_object_kfree(vmw_user_bo, prime);
 }
 
@@ -511,8 +516,7 @@ int vmw_bo_init(struct vmw_private *dev_priv,
 	memset(vmw_bo, 0, sizeof(*vmw_bo));
 	BUILD_BUG_ON(TTM_MAX_BO_PRIORITY <= 3);
 	vmw_bo->base.priority = 3;
-
-	INIT_LIST_HEAD(&vmw_bo->res_list);
+	vmw_bo->res_tree = RB_ROOT;
 
 	ret = ttm_bo_init(bdev, &vmw_bo->base, size,
 			  ttm_bo_type_device, placement,
@@ -566,7 +570,7 @@ static void vmw_user_bo_ref_obj_release(struct ttm_base_object *base,
 
 	switch (ref_type) {
 	case TTM_REF_SYNCCPU_WRITE:
-		ttm_bo_synccpu_write_release(&user_bo->vbo.base);
+		atomic_dec(&user_bo->vbo.cpu_writers);
 		break;
 	default:
 		WARN_ONCE(true, "Undefined buffer object reference release.\n");
@@ -682,12 +686,12 @@ static int vmw_user_bo_synccpu_grab(struct vmw_user_buffer_object *user_bo,
 				    struct ttm_object_file *tfile,
 				    uint32_t flags)
 {
+	bool nonblock = !!(flags & drm_vmw_synccpu_dontblock);
 	struct ttm_buffer_object *bo = &user_bo->vbo.base;
 	bool existed;
 	int ret;
 
 	if (flags & drm_vmw_synccpu_allow_cs) {
-		bool nonblock = !!(flags & drm_vmw_synccpu_dontblock);
 		long lret;
 
 		lret = dma_resv_wait_timeout_rcu
@@ -700,15 +704,22 @@ static int vmw_user_bo_synccpu_grab(struct vmw_user_buffer_object *user_bo,
 		return 0;
 	}
 
-	ret = ttm_bo_synccpu_write_grab
-		(bo, !!(flags & drm_vmw_synccpu_dontblock));
+	ret = ttm_bo_reserve(bo, true, nonblock, NULL);
+	if (unlikely(ret != 0))
+		return ret;
+
+	ret = ttm_bo_wait(bo, true, nonblock);
+	if (likely(ret == 0))
+		atomic_inc(&user_bo->vbo.cpu_writers);
+
+	ttm_bo_unreserve(bo);
 	if (unlikely(ret != 0))
 		return ret;
 
 	ret = ttm_ref_object_add(tfile, &user_bo->prime.base,
 				 TTM_REF_SYNCCPU_WRITE, &existed, false);
 	if (ret != 0 || existed)
-		ttm_bo_synccpu_write_release(&user_bo->vbo.base);
+		atomic_dec(&user_bo->vbo.cpu_writers);
 
 	return ret;
 }
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c
index b38bcb032c99..e962048f65d2 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c
@@ -576,8 +576,7 @@ static int vmw_dma_select_mode(struct vmw_private *dev_priv)
 	else
 		dev_priv->map_mode = vmw_dma_map_populate;
 
-	/* No TTM coherent page pool? FIXME: Ask TTM instead! */
-        if (!(IS_ENABLED(CONFIG_SWIOTLB) || IS_ENABLED(CONFIG_INTEL_IOMMU)) &&
+        if (!IS_ENABLED(CONFIG_DRM_TTM_DMA_PAGE_POOL) &&
 	    (dev_priv->map_mode == vmw_dma_alloc_coherent))
 		return -EINVAL;
 
@@ -827,9 +826,13 @@ static int vmw_driver_load(struct drm_device *dev, unsigned long chipset)
 		goto out_no_fman;
 	}
 
+	drm_vma_offset_manager_init(&dev_priv->vma_manager,
+				    DRM_FILE_PAGE_OFFSET_START,
+				    DRM_FILE_PAGE_OFFSET_SIZE);
 	ret = ttm_bo_device_init(&dev_priv->bdev,
 				 &vmw_bo_driver,
 				 dev->anon_inode->i_mapping,
+				 &dev_priv->vma_manager,
 				 false);
 	if (unlikely(ret != 0)) {
 		DRM_ERROR("Failed initializing TTM buffer object driver.\n");
@@ -986,6 +989,7 @@ static void vmw_driver_unload(struct drm_device *dev)
 	if (dev_priv->has_mob)
 		(void) ttm_bo_clean_mm(&dev_priv->bdev, VMW_PL_MOB);
 	(void) ttm_bo_device_release(&dev_priv->bdev);
+	drm_vma_offset_manager_destroy(&dev_priv->vma_manager);
 	vmw_release_device_late(dev_priv);
 	vmw_fence_manager_takedown(dev_priv->fman);
 	if (dev_priv->capabilities & SVGA_CAP_IRQMASK)
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h
index 5eb73ded8e07..a31e726d6d71 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h
@@ -56,9 +56,9 @@
 
 
 #define VMWGFX_DRIVER_NAME "vmwgfx"
-#define VMWGFX_DRIVER_DATE "20180704"
+#define VMWGFX_DRIVER_DATE "20190328"
 #define VMWGFX_DRIVER_MAJOR 2
-#define VMWGFX_DRIVER_MINOR 15
+#define VMWGFX_DRIVER_MINOR 16
 #define VMWGFX_DRIVER_PATCHLEVEL 0
 #define VMWGFX_FIFO_STATIC_SIZE (1024*1024)
 #define VMWGFX_MAX_RELOCATIONS 2048
@@ -100,21 +100,26 @@ struct vmw_fpriv {
 /**
  * struct vmw_buffer_object - TTM buffer object with vmwgfx additions
  * @base: The TTM buffer object
- * @res_list: List of resources using this buffer object as a backing MOB
+ * @res_tree: RB tree of resources using this buffer object as a backing MOB
  * @pin_count: pin depth
+ * @cpu_writers: Number of synccpu write grabs. Protected by reservation when
+ * increased. May be decreased without reservation.
  * @dx_query_ctx: DX context if this buffer object is used as a DX query MOB
  * @map: Kmap object for semi-persistent mappings
  * @res_prios: Eviction priority counts for attached resources
+ * @dirty: structure for user-space dirty-tracking
  */
 struct vmw_buffer_object {
 	struct ttm_buffer_object base;
-	struct list_head res_list;
+	struct rb_root res_tree;
 	s32 pin_count;
+	atomic_t cpu_writers;
 	/* Not ref-counted.  Protected by binding_mutex */
 	struct vmw_resource *dx_query_ctx;
 	/* Protected by reservation */
 	struct ttm_bo_kmap_obj map;
 	u32 res_prios[TTM_MAX_BO_PRIORITY];
+	struct vmw_bo_dirty *dirty;
 };
 
 /**
@@ -145,7 +150,8 @@ struct vmw_res_func;
  * @res_dirty: Resource contains data not yet in the backup buffer. Protected
  * by resource reserved.
  * @backup_dirty: Backup buffer contains data not yet in the HW resource.
- * Protecte by resource reserved.
+ * Protected by resource reserved.
+ * @coherent: Emulate coherency by tracking vm accesses.
  * @backup: The backup buffer if any. Protected by resource reserved.
  * @backup_offset: Offset into the backup buffer if any. Protected by resource
  * reserved. Note that only a few resource types can have a @backup_offset
@@ -154,29 +160,32 @@ struct vmw_res_func;
  * pin-count greater than zero. It is not on the resource LRU lists and its
  * backup buffer is pinned. Hence it can't be evicted.
  * @func: Method vtable for this resource. Immutable.
+ * @mob_node; Node for the MOB backup rbtree. Protected by @backup reserved.
  * @lru_head: List head for the LRU list. Protected by @dev_priv::resource_lock.
- * @mob_head: List head for the MOB backup list. Protected by @backup reserved.
  * @binding_head: List head for the context binding list. Protected by
  * the @dev_priv::binding_mutex
  * @res_free: The resource destructor.
  * @hw_destroy: Callback to destroy the resource on the device, as part of
  * resource destruction.
  */
+struct vmw_resource_dirty;
 struct vmw_resource {
 	struct kref kref;
 	struct vmw_private *dev_priv;
 	int id;
 	u32 used_prio;
 	unsigned long backup_size;
-	bool res_dirty;
-	bool backup_dirty;
+	u32 res_dirty : 1;
+	u32 backup_dirty : 1;
+	u32 coherent : 1;
 	struct vmw_buffer_object *backup;
 	unsigned long backup_offset;
 	unsigned long pin_count;
 	const struct vmw_res_func *func;
+	struct rb_node mob_node;
 	struct list_head lru_head;
-	struct list_head mob_head;
 	struct list_head binding_head;
+	struct vmw_resource_dirty *dirty;
 	void (*res_free) (struct vmw_resource *res);
 	void (*hw_destroy) (struct vmw_resource *res);
 };
@@ -438,6 +447,7 @@ struct vmw_private {
 	struct vmw_fifo_state fifo;
 
 	struct drm_device *dev;
+	struct drm_vma_offset_manager vma_manager;
 	unsigned long vmw_chipset;
 	unsigned int io_start;
 	uint32_t vram_start;
@@ -674,7 +684,8 @@ extern void vmw_resource_unreference(struct vmw_resource **p_res);
 extern struct vmw_resource *vmw_resource_reference(struct vmw_resource *res);
 extern struct vmw_resource *
 vmw_resource_reference_unless_doomed(struct vmw_resource *res);
-extern int vmw_resource_validate(struct vmw_resource *res, bool intr);
+extern int vmw_resource_validate(struct vmw_resource *res, bool intr,
+				 bool dirtying);
 extern int vmw_resource_reserve(struct vmw_resource *res, bool interruptible,
 				bool no_backup);
 extern bool vmw_resource_needs_backup(const struct vmw_resource *res);
@@ -716,6 +727,10 @@ extern void vmw_resource_evict_all(struct vmw_private *dev_priv);
 extern void vmw_resource_unbind_list(struct vmw_buffer_object *vbo);
 void vmw_resource_mob_attach(struct vmw_resource *res);
 void vmw_resource_mob_detach(struct vmw_resource *res);
+void vmw_resource_dirty_update(struct vmw_resource *res, pgoff_t start,
+			       pgoff_t end);
+int vmw_resources_clean(struct vmw_buffer_object *vbo, pgoff_t start,
+			pgoff_t end, pgoff_t *num_prefault);
 
 /**
  * vmw_resource_mob_attached - Whether a resource currently has a mob attached
@@ -725,7 +740,7 @@ void vmw_resource_mob_detach(struct vmw_resource *res);
  */
 static inline bool vmw_resource_mob_attached(const struct vmw_resource *res)
 {
-	return !list_empty(&res->mob_head);
+	return !RB_EMPTY_NODE(&res->mob_node);
 }
 
 /**
@@ -1403,6 +1418,17 @@ int vmw_host_log(const char *log);
 #define VMW_DEBUG_USER(fmt, ...)                                              \
 	DRM_DEBUG_DRIVER(fmt, ##__VA_ARGS__)
 
+/* Resource dirtying - vmwgfx_page_dirty.c */
+void vmw_bo_dirty_scan(struct vmw_buffer_object *vbo);
+int vmw_bo_dirty_add(struct vmw_buffer_object *vbo);
+void vmw_bo_dirty_transfer_to_res(struct vmw_resource *res);
+void vmw_bo_dirty_clear_res(struct vmw_resource *res);
+void vmw_bo_dirty_release(struct vmw_buffer_object *vbo);
+void vmw_bo_dirty_unmap(struct vmw_buffer_object *vbo,
+			pgoff_t start, pgoff_t end);
+vm_fault_t vmw_bo_vm_fault(struct vm_fault *vmf);
+vm_fault_t vmw_bo_vm_mkwrite(struct vm_fault *vmf);
+
 /**
  * VMW_DEBUG_KMS - Debug output for kernel mode-setting
  *
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_execbuf.c b/drivers/gpu/drm/vmwgfx/vmwgfx_execbuf.c
index ff86d49dc5e8..934ad7c0c342 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_execbuf.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_execbuf.c
@@ -2560,7 +2560,6 @@ static int vmw_cmd_dx_check_subresource(struct vmw_private *dev_priv,
 		     offsetof(typeof(*cmd), sid));
 
 	cmd = container_of(header, typeof(*cmd), header);
-
 	return vmw_cmd_res_check(dev_priv, sw_context, vmw_res_surface,
 				 VMW_RES_DIRTY_NONE, user_surface_converter,
 				 &cmd->sid, NULL);
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_page_dirty.c b/drivers/gpu/drm/vmwgfx/vmwgfx_page_dirty.c
new file mode 100644
index 000000000000..f07aa857587c
--- /dev/null
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_page_dirty.c
@@ -0,0 +1,488 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
+/**************************************************************************
+ *
+ * Copyright 2019 VMware, Inc., Palo Alto, CA., USA
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the
+ * "Software"), to deal in the Software without restriction, including
+ * without limitation the rights to use, copy, modify, merge, publish,
+ * distribute, sub license, and/or sell copies of the Software, and to
+ * permit persons to whom the Software is furnished to do so, subject to
+ * the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the
+ * next paragraph) shall be included in all copies or substantial portions
+ * of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDERS, AUTHORS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM,
+ * DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
+ * OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE
+ * USE OR OTHER DEALINGS IN THE SOFTWARE.
+ *
+ **************************************************************************/
+#include "vmwgfx_drv.h"
+
+/*
+ * Different methods for tracking dirty:
+ * VMW_BO_DIRTY_PAGETABLE - Scan the pagetable for hardware dirty bits
+ * VMW_BO_DIRTY_MKWRITE - Write-protect page table entries and record write-
+ * accesses in the VM mkwrite() callback
+ */
+enum vmw_bo_dirty_method {
+	VMW_BO_DIRTY_PAGETABLE,
+	VMW_BO_DIRTY_MKWRITE,
+};
+
+/*
+ * No dirtied pages at scan trigger a transition to the _MKWRITE method,
+ * similarly a certain percentage of dirty pages trigger a transition to
+ * the _PAGETABLE method. How many triggers should we wait for before
+ * changing method?
+ */
+#define VMW_DIRTY_NUM_CHANGE_TRIGGERS 2
+
+/* Percentage to trigger a transition to the _PAGETABLE method */
+#define VMW_DIRTY_PERCENTAGE 10
+
+/**
+ * struct vmw_bo_dirty - Dirty information for buffer objects
+ * @start: First currently dirty bit
+ * @end: Last currently dirty bit + 1
+ * @method: The currently used dirty method
+ * @change_count: Number of consecutive method change triggers
+ * @ref_count: Reference count for this structure
+ * @bitmap_size: The size of the bitmap in bits. Typically equal to the
+ * nuber of pages in the bo.
+ * @size: The accounting size for this struct.
+ * @bitmap: A bitmap where each bit represents a page. A set bit means a
+ * dirty page.
+ */
+struct vmw_bo_dirty {
+	unsigned long start;
+	unsigned long end;
+	enum vmw_bo_dirty_method method;
+	unsigned int change_count;
+	unsigned int ref_count;
+	unsigned long bitmap_size;
+	size_t size;
+	unsigned long bitmap[0];
+};
+
+/**
+ * vmw_bo_dirty_scan_pagetable - Perform a pagetable scan for dirty bits
+ * @vbo: The buffer object to scan
+ *
+ * Scans the pagetable for dirty bits. Clear those bits and modify the
+ * dirty structure with the results. This function may change the
+ * dirty-tracking method.
+ */
+static void vmw_bo_dirty_scan_pagetable(struct vmw_buffer_object *vbo)
+{
+	struct vmw_bo_dirty *dirty = vbo->dirty;
+	pgoff_t offset = drm_vma_node_start(&vbo->base.base.vma_node);
+	struct address_space *mapping = vbo->base.bdev->dev_mapping;
+	pgoff_t num_marked;
+
+	num_marked = clean_record_shared_mapping_range
+		(mapping,
+		 offset, dirty->bitmap_size,
+		 offset, &dirty->bitmap[0],
+		 &dirty->start, &dirty->end);
+	if (num_marked == 0)
+		dirty->change_count++;
+	else
+		dirty->change_count = 0;
+
+	if (dirty->change_count > VMW_DIRTY_NUM_CHANGE_TRIGGERS) {
+		dirty->change_count = 0;
+		dirty->method = VMW_BO_DIRTY_MKWRITE;
+		wp_shared_mapping_range(mapping,
+					offset, dirty->bitmap_size);
+		clean_record_shared_mapping_range(mapping,
+						  offset, dirty->bitmap_size,
+						  offset, &dirty->bitmap[0],
+						  &dirty->start, &dirty->end);
+	}
+}
+
+/**
+ * vmw_bo_dirty_scan_mkwrite - Reset the mkwrite dirty-tracking method
+ * @vbo: The buffer object to scan
+ *
+ * Write-protect pages written to so that consecutive write accesses will
+ * trigger a call to mkwrite.
+ *
+ * This function may change the dirty-tracking method.
+ */
+static void vmw_bo_dirty_scan_mkwrite(struct vmw_buffer_object *vbo)
+{
+	struct vmw_bo_dirty *dirty = vbo->dirty;
+	unsigned long offset = drm_vma_node_start(&vbo->base.base.vma_node);
+	struct address_space *mapping = vbo->base.bdev->dev_mapping;
+	pgoff_t num_marked;
+
+	if (dirty->end <= dirty->start)
+		return;
+
+	num_marked = wp_shared_mapping_range(vbo->base.bdev->dev_mapping,
+					dirty->start + offset,
+					dirty->end - dirty->start);
+
+	if (100UL * num_marked / dirty->bitmap_size >
+	    VMW_DIRTY_PERCENTAGE) {
+		dirty->change_count++;
+	} else {
+		dirty->change_count = 0;
+	}
+
+	if (dirty->change_count > VMW_DIRTY_NUM_CHANGE_TRIGGERS) {
+		pgoff_t start = 0;
+		pgoff_t end = dirty->bitmap_size;
+
+		dirty->method = VMW_BO_DIRTY_PAGETABLE;
+		clean_record_shared_mapping_range(mapping, offset, end, offset,
+						  &dirty->bitmap[0],
+						  &start, &end);
+		bitmap_clear(&dirty->bitmap[0], 0, dirty->bitmap_size);
+		if (dirty->start < dirty->end)
+			bitmap_set(&dirty->bitmap[0], dirty->start,
+				   dirty->end - dirty->start);
+		dirty->change_count = 0;
+	}
+}
+
+/**
+ * vmw_bo_dirty_scan - Scan for dirty pages and add them to the dirty
+ * tracking structure
+ * @vbo: The buffer object to scan
+ *
+ * This function may change the dirty tracking method.
+ */
+void vmw_bo_dirty_scan(struct vmw_buffer_object *vbo)
+{
+	struct vmw_bo_dirty *dirty = vbo->dirty;
+
+	if (dirty->method == VMW_BO_DIRTY_PAGETABLE)
+		vmw_bo_dirty_scan_pagetable(vbo);
+	else
+		vmw_bo_dirty_scan_mkwrite(vbo);
+}
+
+/**
+ * vmw_bo_dirty_pre_unmap - write-protect and pick up dirty pages before
+ * an unmap_mapping_range operation.
+ * @vbo: The buffer object,
+ * @start: First page of the range within the buffer object.
+ * @end: Last page of the range within the buffer object + 1.
+ *
+ * If we're using the _PAGETABLE scan method, we may leak dirty pages
+ * when calling unmap_mapping_range(). This function makes sure we pick
+ * up all dirty pages.
+ */
+static void vmw_bo_dirty_pre_unmap(struct vmw_buffer_object *vbo,
+				   pgoff_t start, pgoff_t end)
+{
+	struct vmw_bo_dirty *dirty = vbo->dirty;
+	unsigned long offset = drm_vma_node_start(&vbo->base.base.vma_node);
+	struct address_space *mapping = vbo->base.bdev->dev_mapping;
+
+	if (dirty->method != VMW_BO_DIRTY_PAGETABLE || start >= end)
+		return;
+
+	wp_shared_mapping_range(mapping, start + offset, end - start);
+	clean_record_shared_mapping_range(mapping, start + offset,
+					  end - start, offset,
+					  &dirty->bitmap[0], &dirty->start,
+					  &dirty->end);
+}
+
+/**
+ * vmw_bo_dirty_unmap - Clear all ptes pointing to a range within a bo
+ * @vbo: The buffer object,
+ * @start: First page of the range within the buffer object.
+ * @end: Last page of the range within the buffer object + 1.
+ *
+ * This is similar to ttm_bo_unmap_virtual_locked() except it takes a subrange.
+ */
+void vmw_bo_dirty_unmap(struct vmw_buffer_object *vbo,
+			pgoff_t start, pgoff_t end)
+{
+	unsigned long offset = drm_vma_node_start(&vbo->base.base.vma_node);
+	struct address_space *mapping = vbo->base.bdev->dev_mapping;
+
+	vmw_bo_dirty_pre_unmap(vbo, start, end);
+	unmap_shared_mapping_range(mapping, (offset + start) << PAGE_SHIFT,
+				   (loff_t) (end - start) << PAGE_SHIFT);
+}
+
+/**
+ * vmw_bo_dirty_add - Add a dirty-tracking user to a buffer object
+ * @vbo: The buffer object
+ *
+ * This function registers a dirty-tracking user to a buffer object.
+ * A user can be for example a resource or a vma in a special user-space
+ * mapping.
+ *
+ * Return: Zero on success, -ENOMEM on memory allocation failure.
+ */
+int vmw_bo_dirty_add(struct vmw_buffer_object *vbo)
+{
+	struct vmw_bo_dirty *dirty = vbo->dirty;
+	pgoff_t num_pages = vbo->base.num_pages;
+	size_t size, acc_size;
+	int ret;
+	static struct ttm_operation_ctx ctx = {
+		.interruptible = false,
+		.no_wait_gpu = false
+	};
+
+	if (dirty) {
+		dirty->ref_count++;
+		return 0;
+	}
+
+	size = sizeof(*dirty) + BITS_TO_LONGS(num_pages) * sizeof(long);
+	acc_size = ttm_round_pot(size);
+	ret = ttm_mem_global_alloc(&ttm_mem_glob, acc_size, &ctx);
+	if (ret) {
+		VMW_DEBUG_USER("Out of graphics memory for buffer object "
+			       "dirty tracker.\n");
+		return ret;
+	}
+	dirty = kvzalloc(size, GFP_KERNEL);
+	if (!dirty) {
+		ret = -ENOMEM;
+		goto out_no_dirty;
+	}
+
+	dirty->size = acc_size;
+	dirty->bitmap_size = num_pages;
+	dirty->start = dirty->bitmap_size;
+	dirty->end = 0;
+	dirty->ref_count = 1;
+	if (num_pages < PAGE_SIZE / sizeof(pte_t)) {
+		dirty->method = VMW_BO_DIRTY_PAGETABLE;
+	} else {
+		struct address_space *mapping = vbo->base.bdev->dev_mapping;
+		pgoff_t offset = drm_vma_node_start(&vbo->base.base.vma_node);
+
+		dirty->method = VMW_BO_DIRTY_MKWRITE;
+
+		/* Write-protect and then pick up already dirty bits */
+		wp_shared_mapping_range(mapping, offset, num_pages);
+		clean_record_shared_mapping_range(mapping, offset, num_pages,
+						  offset,
+						  &dirty->bitmap[0],
+						  &dirty->start, &dirty->end);
+	}
+
+	vbo->dirty = dirty;
+
+	return 0;
+
+out_no_dirty:
+	ttm_mem_global_free(&ttm_mem_glob, acc_size);
+	return ret;
+}
+
+/**
+ * vmw_bo_dirty_release - Release a dirty-tracking user from a buffer object
+ * @vbo: The buffer object
+ *
+ * This function releases a dirty-tracking user from a buffer object.
+ * If the reference count reaches zero, then the dirty-tracking object is
+ * freed and the pointer to it cleared.
+ *
+ * Return: Zero on success, -ENOMEM on memory allocation failure.
+ */
+void vmw_bo_dirty_release(struct vmw_buffer_object *vbo)
+{
+	struct vmw_bo_dirty *dirty = vbo->dirty;
+
+	if (dirty && --dirty->ref_count == 0) {
+		size_t acc_size = dirty->size;
+
+		kvfree(dirty);
+		ttm_mem_global_free(&ttm_mem_glob, acc_size);
+		vbo->dirty = NULL;
+	}
+}
+
+/**
+ * vmw_bo_dirty_transfer_to_res - Pick up a resource's dirty region from
+ * its backing mob.
+ * @res: The resource
+ *
+ * This function will pick up all dirty ranges affecting the resource from
+ * it's backup mob, and call vmw_resource_dirty_update() once for each
+ * range. The transferred ranges will be cleared from the backing mob's
+ * dirty tracking.
+ */
+void vmw_bo_dirty_transfer_to_res(struct vmw_resource *res)
+{
+	struct vmw_buffer_object *vbo = res->backup;
+	struct vmw_bo_dirty *dirty = vbo->dirty;
+	pgoff_t start, cur, end;
+	unsigned long res_start = res->backup_offset;
+	unsigned long res_end = res->backup_offset + res->backup_size;
+
+	WARN_ON_ONCE(res_start & ~PAGE_MASK);
+	res_start >>= PAGE_SHIFT;
+	res_end = DIV_ROUND_UP(res_end, PAGE_SIZE);
+
+	if (res_start >= dirty->end || res_end <= dirty->start)
+		return;
+
+	cur = max(res_start, dirty->start);
+	res_end = max(res_end, dirty->end);
+	while (cur < res_end) {
+		unsigned long num;
+
+		start = find_next_bit(&dirty->bitmap[0], res_end, cur);
+		if (start >= res_end)
+			break;
+
+		end = find_next_zero_bit(&dirty->bitmap[0], res_end, start + 1);
+		cur = end + 1;
+		num = end - start;
+		bitmap_clear(&dirty->bitmap[0], start, num);
+		vmw_resource_dirty_update(res, start, end);
+	}
+
+	if (res_start <= dirty->start && res_end > dirty->start)
+		dirty->start = res_end;
+	if (res_start < dirty->end && res_end >= dirty->end)
+		dirty->end = res_start;
+}
+
+/**
+ * vmw_bo_dirty_clear_res - Clear a resource's dirty region from
+ * its backing mob.
+ * @res: The resource
+ *
+ * This function will clear all dirty ranges affecting the resource from
+ * it's backup mob's dirty tracking.
+ */
+void vmw_bo_dirty_clear_res(struct vmw_resource *res)
+{
+	unsigned long res_start = res->backup_offset;
+	unsigned long res_end = res->backup_offset + res->backup_size;
+	struct vmw_buffer_object *vbo = res->backup;
+	struct vmw_bo_dirty *dirty = vbo->dirty;
+
+	res_start >>= PAGE_SHIFT;
+	res_end = DIV_ROUND_UP(res_end, PAGE_SIZE);
+
+	if (res_start >= dirty->end || res_end <= dirty->start)
+		return;
+
+	res_start = max(res_start, dirty->start);
+	res_end = min(res_end, dirty->end);
+	bitmap_clear(&dirty->bitmap[0], res_start, res_end - res_start);
+
+	if (res_start <= dirty->start && res_end > dirty->start)
+		dirty->start = res_end;
+	if (res_start < dirty->end && res_end >= dirty->end)
+		dirty->end = res_start;
+}
+
+vm_fault_t vmw_bo_vm_mkwrite(struct vm_fault *vmf)
+{
+	struct vm_area_struct *vma = vmf->vma;
+	struct ttm_buffer_object *bo = (struct ttm_buffer_object *)
+	    vma->vm_private_data;
+	vm_fault_t ret;
+	unsigned long page_offset;
+	unsigned int save_flags;
+	struct vmw_buffer_object *vbo =
+		container_of(bo, typeof(*vbo), base);
+
+	/*
+	 * mkwrite() doesn't handle the VM_FAULT_RETRY return value correctly.
+	 * So make sure the TTM helpers are aware.
+	 */
+	save_flags = vmf->flags;
+	vmf->flags &= ~FAULT_FLAG_ALLOW_RETRY;
+	ret = ttm_bo_vm_reserve(bo, vmf);
+	vmf->flags = save_flags;
+	if (ret)
+		return ret;
+
+	page_offset = vmf->pgoff - drm_vma_node_start(&bo->base.vma_node);
+	if (unlikely(page_offset >= bo->num_pages)) {
+		ret = VM_FAULT_SIGBUS;
+		goto out_unlock;
+	}
+
+	if (vbo->dirty && vbo->dirty->method == VMW_BO_DIRTY_MKWRITE &&
+	    !test_bit(page_offset, &vbo->dirty->bitmap[0])) {
+		struct vmw_bo_dirty *dirty = vbo->dirty;
+
+		__set_bit(page_offset, &dirty->bitmap[0]);
+		dirty->start = min(dirty->start, page_offset);
+		dirty->end = max(dirty->end, page_offset + 1);
+	}
+
+out_unlock:
+	dma_resv_unlock(bo->base.resv);
+	return ret;
+}
+
+vm_fault_t vmw_bo_vm_fault(struct vm_fault *vmf)
+{
+	struct vm_area_struct *vma = vmf->vma;
+	struct ttm_buffer_object *bo = (struct ttm_buffer_object *)
+	    vma->vm_private_data;
+	struct vmw_buffer_object *vbo =
+		container_of(bo, struct vmw_buffer_object, base);
+	pgoff_t num_prefault;
+	pgprot_t prot;
+	vm_fault_t ret;
+
+	ret = ttm_bo_vm_reserve(bo, vmf);
+	if (ret)
+		return ret;
+
+	num_prefault = (vma->vm_flags & VM_RAND_READ) ? 1 :
+		TTM_BO_VM_NUM_PREFAULT;
+
+	if (vbo->dirty) {
+		pgoff_t allowed_prefault;
+		unsigned long page_offset;
+
+		page_offset = vmf->pgoff -
+			drm_vma_node_start(&bo->base.vma_node);
+		if (page_offset >= bo->num_pages ||
+		    vmw_resources_clean(vbo, page_offset,
+					page_offset + PAGE_SIZE,
+					&allowed_prefault)) {
+			ret = VM_FAULT_SIGBUS;
+			goto out_unlock;
+		}
+
+		num_prefault = min(num_prefault, allowed_prefault);
+	}
+
+	/*
+	 * If we don't track dirty using the MKWRITE method, make sure
+	 * sure the page protection is write-enabled so we don't get
+	 * a lot of unnecessary write faults.
+	 */
+	if (vbo->dirty && vbo->dirty->method == VMW_BO_DIRTY_MKWRITE)
+		prot = vma->vm_page_prot;
+	else
+		prot = vm_get_page_prot(vma->vm_flags);
+
+	ret = ttm_bo_vm_fault_reserved(vmf, prot, num_prefault);
+	if (ret == VM_FAULT_RETRY && !(vmf->flags & FAULT_FLAG_RETRY_NOWAIT))
+		return ret;
+
+out_unlock:
+	dma_resv_unlock(bo->base.resv);
+
+	return ret;
+}
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c b/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c
index 5581a7826b4c..c8441030637a 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c
@@ -40,11 +40,24 @@
 void vmw_resource_mob_attach(struct vmw_resource *res)
 {
 	struct vmw_buffer_object *backup = res->backup;
+	struct rb_node **new = &backup->res_tree.rb_node, *parent = NULL;
 
 	dma_resv_assert_held(res->backup->base.base.resv);
 	res->used_prio = (res->res_dirty) ? res->func->dirty_prio :
 		res->func->prio;
-	list_add_tail(&res->mob_head, &backup->res_list);
+
+	while (*new) {
+		struct vmw_resource *this =
+			container_of(*new, struct vmw_resource, mob_node);
+
+		parent = *new;
+		new = (res->backup_offset < this->backup_offset) ?
+			&((*new)->rb_left) : &((*new)->rb_right);
+	}
+
+	rb_link_node(&res->mob_node, parent, new);
+	rb_insert_color(&res->mob_node, &backup->res_tree);
+
 	vmw_bo_prio_add(backup, res->used_prio);
 }
 
@@ -58,7 +71,8 @@ void vmw_resource_mob_detach(struct vmw_resource *res)
 
 	dma_resv_assert_held(backup->base.base.resv);
 	if (vmw_resource_mob_attached(res)) {
-		list_del_init(&res->mob_head);
+		rb_erase(&res->mob_node, &backup->res_tree);
+		RB_CLEAR_NODE(&res->mob_node);
 		vmw_bo_prio_del(backup, res->used_prio);
 	}
 }
@@ -119,6 +133,10 @@ static void vmw_resource_release(struct kref *kref)
 		}
 		res->backup_dirty = false;
 		vmw_resource_mob_detach(res);
+		if (res->dirty)
+			res->func->dirty_free(res);
+		if (res->coherent)
+			vmw_bo_dirty_release(res->backup);
 		ttm_bo_unreserve(bo);
 		vmw_bo_unreference(&res->backup);
 	}
@@ -200,15 +218,17 @@ int vmw_resource_init(struct vmw_private *dev_priv, struct vmw_resource *res,
 	res->res_free = res_free;
 	res->dev_priv = dev_priv;
 	res->func = func;
+	RB_CLEAR_NODE(&res->mob_node);
 	INIT_LIST_HEAD(&res->lru_head);
-	INIT_LIST_HEAD(&res->mob_head);
 	INIT_LIST_HEAD(&res->binding_head);
 	res->id = -1;
 	res->backup = NULL;
 	res->backup_offset = 0;
 	res->backup_dirty = false;
 	res->res_dirty = false;
+	res->coherent = false;
 	res->used_prio = 3;
+	res->dirty = NULL;
 	if (delay_id)
 		return 0;
 	else
@@ -373,7 +393,8 @@ out_no_bo:
  * should be retried once resources have been freed up.
  */
 static int vmw_resource_do_validate(struct vmw_resource *res,
-				    struct ttm_validate_buffer *val_buf)
+				    struct ttm_validate_buffer *val_buf,
+				    bool dirtying)
 {
 	int ret = 0;
 	const struct vmw_res_func *func = res->func;
@@ -395,6 +416,39 @@ static int vmw_resource_do_validate(struct vmw_resource *res,
 			vmw_resource_mob_attach(res);
 	}
 
+	/*
+	 * Handle the case where the backup mob is marked coherent but
+	 * the resource isn't.
+	 */
+	if (func->dirty_alloc && vmw_resource_mob_attached(res) &&
+	    !res->coherent) {
+		if (res->backup->dirty && !res->dirty) {
+			ret = func->dirty_alloc(res);
+			if (ret)
+				return ret;
+		} else if (!res->backup->dirty && res->dirty) {
+			func->dirty_free(res);
+		}
+	}
+
+	/*
+	 * Transfer the dirty regions to the resource and update
+	 * the resource.
+	 */
+	if (res->dirty) {
+		if (dirtying && !res->res_dirty) {
+			pgoff_t start = res->backup_offset >> PAGE_SHIFT;
+			pgoff_t end = __KERNEL_DIV_ROUND_UP
+				(res->backup_offset + res->backup_size,
+				 PAGE_SIZE);
+
+			vmw_bo_dirty_unmap(res->backup, start, end);
+		}
+
+		vmw_bo_dirty_transfer_to_res(res);
+		return func->dirty_sync(res);
+	}
+
 	return 0;
 
 out_bind_failed:
@@ -433,16 +487,28 @@ void vmw_resource_unreserve(struct vmw_resource *res,
 	if (switch_backup && new_backup != res->backup) {
 		if (res->backup) {
 			vmw_resource_mob_detach(res);
+			if (res->coherent)
+				vmw_bo_dirty_release(res->backup);
 			vmw_bo_unreference(&res->backup);
 		}
 
 		if (new_backup) {
 			res->backup = vmw_bo_reference(new_backup);
+
+			/*
+			 * The validation code should already have added a
+			 * dirty tracker here.
+			 */
+			WARN_ON(res->coherent && !new_backup->dirty);
+
 			vmw_resource_mob_attach(res);
 		} else {
 			res->backup = NULL;
 		}
+	} else if (switch_backup && res->coherent) {
+		vmw_bo_dirty_release(res->backup);
 	}
+
 	if (switch_backup)
 		res->backup_offset = new_backup_offset;
 
@@ -492,8 +558,7 @@ vmw_resource_check_buffer(struct ww_acquire_ctx *ticket,
 	val_buf->bo = &res->backup->base;
 	val_buf->num_shared = 0;
 	list_add_tail(&val_buf->head, &val_list);
-	ret = ttm_eu_reserve_buffers(ticket, &val_list, interruptible, NULL,
-				     true);
+	ret = ttm_eu_reserve_buffers(ticket, &val_list, interruptible, NULL);
 	if (unlikely(ret != 0))
 		goto out_no_reserve;
 
@@ -623,6 +688,7 @@ out_no_unbind:
  *                         to the device.
  * @res: The resource to make visible to the device.
  * @intr: Perform waits interruptible if possible.
+ * @dirtying: Pending GPU operation will dirty the resource
  *
  * On succesful return, any backup DMA buffer pointed to by @res->backup will
  * be reserved and validated.
@@ -632,7 +698,8 @@ out_no_unbind:
  * Return: Zero on success, -ERESTARTSYS if interrupted, negative error code
  * on failure.
  */
-int vmw_resource_validate(struct vmw_resource *res, bool intr)
+int vmw_resource_validate(struct vmw_resource *res, bool intr,
+			  bool dirtying)
 {
 	int ret;
 	struct vmw_resource *evict_res;
@@ -649,7 +716,7 @@ int vmw_resource_validate(struct vmw_resource *res, bool intr)
 	if (res->backup)
 		val_buf.bo = &res->backup->base;
 	do {
-		ret = vmw_resource_do_validate(res, &val_buf);
+		ret = vmw_resource_do_validate(res, &val_buf, dirtying);
 		if (likely(ret != -EBUSY))
 			break;
 
@@ -712,19 +779,20 @@ out_no_validate:
  */
 void vmw_resource_unbind_list(struct vmw_buffer_object *vbo)
 {
-
-	struct vmw_resource *res, *next;
 	struct ttm_validate_buffer val_buf = {
 		.bo = &vbo->base,
 		.num_shared = 0
 	};
 
 	dma_resv_assert_held(vbo->base.base.resv);
-	list_for_each_entry_safe(res, next, &vbo->res_list, mob_head) {
-		if (!res->func->unbind)
-			continue;
+	while (!RB_EMPTY_ROOT(&vbo->res_tree)) {
+		struct rb_node *node = vbo->res_tree.rb_node;
+		struct vmw_resource *res =
+			container_of(node, struct vmw_resource, mob_node);
+
+		if (!WARN_ON_ONCE(!res->func->unbind))
+			(void) res->func->unbind(res, res->res_dirty, &val_buf);
 
-		(void) res->func->unbind(res, res->res_dirty, &val_buf);
 		res->backup_dirty = true;
 		res->res_dirty = false;
 		vmw_resource_mob_detach(res);
@@ -948,7 +1016,7 @@ int vmw_resource_pin(struct vmw_resource *res, bool interruptible)
 			/* Do we really need to pin the MOB as well? */
 			vmw_bo_pin_reserved(vbo, true);
 		}
-		ret = vmw_resource_validate(res, interruptible);
+		ret = vmw_resource_validate(res, interruptible, true);
 		if (vbo)
 			ttm_bo_unreserve(&vbo->base);
 		if (ret)
@@ -1008,3 +1076,101 @@ enum vmw_res_type vmw_res_type(const struct vmw_resource *res)
 {
 	return res->func->res_type;
 }
+
+/**
+ * vmw_resource_update_dirty - Update a resource's dirty tracker with a
+ * sequential range of touched backing store memory.
+ * @res: The resource.
+ * @start: The first page touched.
+ * @end: The last page touched + 1.
+ */
+void vmw_resource_dirty_update(struct vmw_resource *res, pgoff_t start,
+			       pgoff_t end)
+{
+	if (res->dirty)
+		res->func->dirty_range_add(res, start << PAGE_SHIFT,
+					   end << PAGE_SHIFT);
+}
+
+/**
+ * vmw_resources_clean - Clean resources intersecting a mob range
+ * @vbo: The mob buffer object
+ * @start: The mob page offset starting the range
+ * @end: The mob page offset ending the range
+ * @num_prefault: Returns how many pages including the first have been
+ * cleaned and are ok to prefault
+ */
+int vmw_resources_clean(struct vmw_buffer_object *vbo, pgoff_t start,
+			pgoff_t end, pgoff_t *num_prefault)
+{
+	struct rb_node *cur = vbo->res_tree.rb_node;
+	struct vmw_resource *found = NULL;
+	unsigned long res_start = start << PAGE_SHIFT;
+	unsigned long res_end = end << PAGE_SHIFT;
+	unsigned long last_cleaned = 0;
+
+	/*
+	 * Find the resource with lowest backup_offset that intersects the
+	 * range.
+	 */
+	while (cur) {
+		struct vmw_resource *cur_res =
+			container_of(cur, struct vmw_resource, mob_node);
+
+		if (cur_res->backup_offset >= res_end) {
+			cur = cur->rb_left;
+		} else if (cur_res->backup_offset + cur_res->backup_size <=
+			   res_start) {
+			cur = cur->rb_right;
+		} else {
+			found = cur_res;
+			cur = cur->rb_left;
+			/* Continue to look for resources with lower offsets */
+		}
+	}
+
+	/*
+	 * In order of increasing backup_offset, clean dirty resorces
+	 * intersecting the range.
+	 */
+	while (found) {
+		if (found->res_dirty) {
+			int ret;
+
+			if (!found->func->clean)
+				return -EINVAL;
+
+			ret = found->func->clean(found);
+			if (ret)
+				return ret;
+
+			found->res_dirty = false;
+		}
+		last_cleaned = found->backup_offset + found->backup_size;
+		cur = rb_next(&found->mob_node);
+		if (!cur)
+			break;
+
+		found = container_of(cur, struct vmw_resource, mob_node);
+		if (found->backup_offset >= res_end)
+			break;
+	}
+
+	/*
+	 * Set number of pages allowed prefaulting and fence the buffer object
+	 */
+	*num_prefault = 1;
+	if (last_cleaned > res_start) {
+		struct ttm_buffer_object *bo = &vbo->base;
+
+		*num_prefault = __KERNEL_DIV_ROUND_UP(last_cleaned - res_start,
+						      PAGE_SIZE);
+		vmw_bo_fence_single(bo, NULL);
+		if (bo->moving)
+			dma_fence_put(bo->moving);
+		bo->moving = dma_fence_get
+			(dma_resv_get_excl(bo->base.resv));
+	}
+
+	return 0;
+}
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_resource_priv.h b/drivers/gpu/drm/vmwgfx/vmwgfx_resource_priv.h
index 984e588c62ca..3b7438b2d289 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_resource_priv.h
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_resource_priv.h
@@ -71,6 +71,13 @@ struct vmw_user_resource_conv {
  * @commit_notify:     If the resource is a command buffer managed resource,
  *                     callback to notify that a define or remove command
  *                     has been committed to the device.
+ * @dirty_alloc:       Allocate a dirty tracker. NULL if dirty-tracking is not
+ *                     supported.
+ * @dirty_free:        Free the dirty tracker.
+ * @dirty_sync:        Upload the dirty mob contents to the resource.
+ * @dirty_add_range:   Add a sequential dirty range to the resource
+ *                     dirty tracker.
+ * @clean:             Clean the resource.
  */
 struct vmw_res_func {
 	enum vmw_res_type res_type;
@@ -90,6 +97,12 @@ struct vmw_res_func {
 		       struct ttm_validate_buffer *val_buf);
 	void (*commit_notify)(struct vmw_resource *res,
 			      enum vmw_cmdbuf_res_state state);
+	int (*dirty_alloc)(struct vmw_resource *res);
+	void (*dirty_free)(struct vmw_resource *res);
+	int (*dirty_sync)(struct vmw_resource *res);
+	void (*dirty_range_add)(struct vmw_resource *res, size_t start,
+				 size_t end);
+	int (*clean)(struct vmw_resource *res);
 };
 
 /**
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_surface.c b/drivers/gpu/drm/vmwgfx/vmwgfx_surface.c
index 29d8794f0421..32b9131b2bae 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_surface.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_surface.c
@@ -68,6 +68,20 @@ struct vmw_surface_offset {
 	uint32_t bo_offset;
 };
 
+/**
+ * vmw_surface_dirty - Surface dirty-tracker
+ * @cache: Cached layout information of the surface.
+ * @size: Accounting size for the struct vmw_surface_dirty.
+ * @num_subres: Number of subresources.
+ * @boxes: Array of SVGA3dBoxes indicating dirty regions. One per subresource.
+ */
+struct vmw_surface_dirty {
+	struct svga3dsurface_cache cache;
+	size_t size;
+	u32 num_subres;
+	SVGA3dBox boxes[0];
+};
+
 static void vmw_user_surface_free(struct vmw_resource *res);
 static struct vmw_resource *
 vmw_user_surface_base_to_res(struct ttm_base_object *base);
@@ -96,6 +110,13 @@ vmw_gb_surface_reference_internal(struct drm_device *dev,
 				  struct drm_vmw_gb_surface_ref_ext_rep *rep,
 				  struct drm_file *file_priv);
 
+static void vmw_surface_dirty_free(struct vmw_resource *res);
+static int vmw_surface_dirty_alloc(struct vmw_resource *res);
+static int vmw_surface_dirty_sync(struct vmw_resource *res);
+static void vmw_surface_dirty_range_add(struct vmw_resource *res, size_t start,
+					size_t end);
+static int vmw_surface_clean(struct vmw_resource *res);
+
 static const struct vmw_user_resource_conv user_surface_conv = {
 	.object_type = VMW_RES_SURFACE,
 	.base_obj_to_res = vmw_user_surface_base_to_res,
@@ -133,7 +154,12 @@ static const struct vmw_res_func vmw_gb_surface_func = {
 	.create = vmw_gb_surface_create,
 	.destroy = vmw_gb_surface_destroy,
 	.bind = vmw_gb_surface_bind,
-	.unbind = vmw_gb_surface_unbind
+	.unbind = vmw_gb_surface_unbind,
+	.dirty_alloc = vmw_surface_dirty_alloc,
+	.dirty_free = vmw_surface_dirty_free,
+	.dirty_sync = vmw_surface_dirty_sync,
+	.dirty_range_add = vmw_surface_dirty_range_add,
+	.clean = vmw_surface_clean,
 };
 
 /**
@@ -336,7 +362,6 @@ static void vmw_hw_surface_destroy(struct vmw_resource *res)
 {
 
 	struct vmw_private *dev_priv = res->dev_priv;
-	struct vmw_surface *srf;
 	void *cmd;
 
 	if (res->func->destroy == vmw_gb_surface_destroy) {
@@ -360,7 +385,6 @@ static void vmw_hw_surface_destroy(struct vmw_resource *res)
 		 */
 
 		mutex_lock(&dev_priv->cmdbuf_mutex);
-		srf = vmw_res_to_srf(res);
 		dev_priv->used_memory_size -= res->backup_size;
 		mutex_unlock(&dev_priv->cmdbuf_mutex);
 	}
@@ -641,6 +665,7 @@ static void vmw_user_surface_free(struct vmw_resource *res)
 	struct vmw_private *dev_priv = srf->res.dev_priv;
 	uint32_t size = user_srf->size;
 
+	WARN_ON_ONCE(res->dirty);
 	if (user_srf->master)
 		drm_master_put(&user_srf->master);
 	kfree(srf->offsets);
@@ -1168,10 +1193,16 @@ static int vmw_gb_surface_bind(struct vmw_resource *res,
 		cmd2->header.id = SVGA_3D_CMD_UPDATE_GB_SURFACE;
 		cmd2->header.size = sizeof(cmd2->body);
 		cmd2->body.sid = res->id;
-		res->backup_dirty = false;
 	}
 	vmw_fifo_commit(dev_priv, submit_size);
 
+	if (res->backup->dirty && res->backup_dirty) {
+		/* We've just made a full upload. Cear dirty regions. */
+		vmw_bo_dirty_clear_res(res);
+	}
+
+	res->backup_dirty = false;
+
 	return 0;
 }
 
@@ -1636,7 +1667,8 @@ vmw_gb_surface_define_internal(struct drm_device *dev,
 			}
 		}
 	} else if (req->base.drm_surface_flags &
-		   drm_vmw_surface_flag_create_buffer)
+		   (drm_vmw_surface_flag_create_buffer |
+		    drm_vmw_surface_flag_coherent))
 		ret = vmw_user_bo_alloc(dev_priv, tfile,
 					res->backup_size,
 					req->base.drm_surface_flags &
@@ -1650,6 +1682,26 @@ vmw_gb_surface_define_internal(struct drm_device *dev,
 		goto out_unlock;
 	}
 
+	if (req->base.drm_surface_flags & drm_vmw_surface_flag_coherent) {
+		struct vmw_buffer_object *backup = res->backup;
+
+		ttm_bo_reserve(&backup->base, false, false, NULL);
+		if (!res->func->dirty_alloc)
+			ret = -EINVAL;
+		if (!ret)
+			ret = vmw_bo_dirty_add(backup);
+		if (!ret) {
+			res->coherent = true;
+			ret = res->func->dirty_alloc(res);
+		}
+		ttm_bo_unreserve(&backup->base);
+		if (ret) {
+			vmw_resource_unreference(&res);
+			goto out_unlock;
+		}
+
+	}
+
 	tmp = vmw_resource_reference(res);
 	ret = ttm_prime_object_init(tfile, res->backup_size, &user_srf->prime,
 				    req->base.drm_surface_flags &
@@ -1758,3 +1810,338 @@ out_bad_resource:
 
 	return ret;
 }
+
+/**
+ * vmw_subres_dirty_add - Add a dirty region to a subresource
+ * @dirty: The surfaces's dirty tracker.
+ * @loc_start: The location corresponding to the start of the region.
+ * @loc_end: The location corresponding to the end of the region.
+ *
+ * As we are assuming that @loc_start and @loc_end represent a sequential
+ * range of backing store memory, if the region spans multiple lines then
+ * regardless of the x coordinate, the full lines are dirtied.
+ * Correspondingly if the region spans multiple z slices, then full rather
+ * than partial z slices are dirtied.
+ */
+static void vmw_subres_dirty_add(struct vmw_surface_dirty *dirty,
+				 const struct svga3dsurface_loc *loc_start,
+				 const struct svga3dsurface_loc *loc_end)
+{
+	const struct svga3dsurface_cache *cache = &dirty->cache;
+	SVGA3dBox *box = &dirty->boxes[loc_start->sub_resource];
+	u32 mip = loc_start->sub_resource % cache->num_mip_levels;
+	const struct drm_vmw_size *size = &cache->mip[mip].size;
+	u32 box_c2 = box->z + box->d;
+
+	if (WARN_ON(loc_start->sub_resource >= dirty->num_subres))
+		return;
+
+	if (box->d == 0 || box->z > loc_start->z)
+		box->z = loc_start->z;
+	if (box_c2 < loc_end->z)
+		box->d = loc_end->z - box->z;
+
+	if (loc_start->z + 1 == loc_end->z) {
+		box_c2 = box->y + box->h;
+		if (box->h == 0 || box->y > loc_start->y)
+			box->y = loc_start->y;
+		if (box_c2 < loc_end->y)
+			box->h = loc_end->y - box->y;
+
+		if (loc_start->y + 1 == loc_end->y) {
+			box_c2 = box->x + box->w;
+			if (box->w == 0 || box->x > loc_start->x)
+				box->x = loc_start->x;
+			if (box_c2 < loc_end->x)
+				box->w = loc_end->x - box->x;
+		} else {
+			box->x = 0;
+			box->w = size->width;
+		}
+	} else {
+		box->y = 0;
+		box->h = size->height;
+		box->x = 0;
+		box->w = size->width;
+	}
+}
+
+/**
+ * vmw_subres_dirty_full - Mark a full subresource as dirty
+ * @dirty: The surface's dirty tracker.
+ * @subres: The subresource
+ */
+static void vmw_subres_dirty_full(struct vmw_surface_dirty *dirty, u32 subres)
+{
+	const struct svga3dsurface_cache *cache = &dirty->cache;
+	u32 mip = subres % cache->num_mip_levels;
+	const struct drm_vmw_size *size = &cache->mip[mip].size;
+	SVGA3dBox *box = &dirty->boxes[subres];
+
+	box->x = 0;
+	box->y = 0;
+	box->z = 0;
+	box->w = size->width;
+	box->h = size->height;
+	box->d = size->depth;
+}
+
+/*
+ * vmw_surface_tex_dirty_add_range - The dirty_add_range callback for texture
+ * surfaces.
+ */
+static void vmw_surface_tex_dirty_range_add(struct vmw_resource *res,
+					    size_t start, size_t end)
+{
+	struct vmw_surface_dirty *dirty =
+		(struct vmw_surface_dirty *) res->dirty;
+	size_t backup_end = res->backup_offset + res->backup_size;
+	struct svga3dsurface_loc loc1, loc2;
+	const struct svga3dsurface_cache *cache;
+
+	start = max_t(size_t, start, res->backup_offset) - res->backup_offset;
+	end = min(end, backup_end) - res->backup_offset;
+	cache = &dirty->cache;
+	svga3dsurface_get_loc(cache, &loc1, start);
+	svga3dsurface_get_loc(cache, &loc2, end - 1);
+	svga3dsurface_inc_loc(cache, &loc2);
+
+	if (loc1.sub_resource + 1 == loc2.sub_resource) {
+		/* Dirty range covers a single sub-resource */
+		vmw_subres_dirty_add(dirty, &loc1, &loc2);
+	} else {
+		/* Dirty range covers multiple sub-resources */
+		struct svga3dsurface_loc loc_min, loc_max;
+		u32 sub_res;
+
+		svga3dsurface_max_loc(cache, loc1.sub_resource, &loc_max);
+		vmw_subres_dirty_add(dirty, &loc1, &loc_max);
+		svga3dsurface_min_loc(cache, loc2.sub_resource - 1, &loc_min);
+		vmw_subres_dirty_add(dirty, &loc_min, &loc2);
+		for (sub_res = loc1.sub_resource + 1;
+		     sub_res < loc2.sub_resource - 1; ++sub_res)
+			vmw_subres_dirty_full(dirty, sub_res);
+	}
+}
+
+/*
+ * vmw_surface_tex_dirty_add_range - The dirty_add_range callback for buffer
+ * surfaces.
+ */
+static void vmw_surface_buf_dirty_range_add(struct vmw_resource *res,
+					    size_t start, size_t end)
+{
+	struct vmw_surface_dirty *dirty =
+		(struct vmw_surface_dirty *) res->dirty;
+	const struct svga3dsurface_cache *cache = &dirty->cache;
+	size_t backup_end = res->backup_offset + cache->mip_chain_bytes;
+	SVGA3dBox *box = &dirty->boxes[0];
+	u32 box_c2;
+
+	box->h = box->d = 1;
+	start = max_t(size_t, start, res->backup_offset) - res->backup_offset;
+	end = min(end, backup_end) - res->backup_offset;
+	box_c2 = box->x + box->w;
+	if (box->w == 0 || box->x > start)
+		box->x = start;
+	if (box_c2 < end)
+		box->w = end - box->x;
+}
+
+/*
+ * vmw_surface_tex_dirty_add_range - The dirty_add_range callback for surfaces
+ */
+static void vmw_surface_dirty_range_add(struct vmw_resource *res, size_t start,
+					size_t end)
+{
+	struct vmw_surface *srf = vmw_res_to_srf(res);
+
+	if (WARN_ON(end <= res->backup_offset ||
+		    start >= res->backup_offset + res->backup_size))
+		return;
+
+	if (srf->format == SVGA3D_BUFFER)
+		vmw_surface_buf_dirty_range_add(res, start, end);
+	else
+		vmw_surface_tex_dirty_range_add(res, start, end);
+}
+
+/*
+ * vmw_surface_dirty_sync - The surface's dirty_sync callback.
+ */
+static int vmw_surface_dirty_sync(struct vmw_resource *res)
+{
+	struct vmw_private *dev_priv = res->dev_priv;
+	bool has_dx = 0;
+	u32 i, num_dirty;
+	struct vmw_surface_dirty *dirty =
+		(struct vmw_surface_dirty *) res->dirty;
+	size_t alloc_size;
+	const struct svga3dsurface_cache *cache = &dirty->cache;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdDXUpdateSubResource body;
+	} *cmd1;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdUpdateGBImage body;
+	} *cmd2;
+	void *cmd;
+
+	num_dirty = 0;
+	for (i = 0; i < dirty->num_subres; ++i) {
+		const SVGA3dBox *box = &dirty->boxes[i];
+
+		if (box->d)
+			num_dirty++;
+	}
+
+	if (!num_dirty)
+		goto out;
+
+	alloc_size = num_dirty * ((has_dx) ? sizeof(*cmd1) : sizeof(*cmd2));
+	cmd = VMW_FIFO_RESERVE(dev_priv, alloc_size);
+	if (!cmd)
+		return -ENOMEM;
+
+	cmd1 = cmd;
+	cmd2 = cmd;
+
+	for (i = 0; i < dirty->num_subres; ++i) {
+		const SVGA3dBox *box = &dirty->boxes[i];
+
+		if (!box->d)
+			continue;
+
+		/*
+		 * DX_UPDATE_SUBRESOURCE is aware of array surfaces.
+		 * UPDATE_GB_IMAGE is not.
+		 */
+		if (has_dx) {
+			cmd1->header.id = SVGA_3D_CMD_DX_UPDATE_SUBRESOURCE;
+			cmd1->header.size = sizeof(cmd1->body);
+			cmd1->body.sid = res->id;
+			cmd1->body.subResource = i;
+			cmd1->body.box = *box;
+			cmd1++;
+		} else {
+			cmd2->header.id = SVGA_3D_CMD_UPDATE_GB_IMAGE;
+			cmd2->header.size = sizeof(cmd2->body);
+			cmd2->body.image.sid = res->id;
+			cmd2->body.image.face = i / cache->num_mip_levels;
+			cmd2->body.image.mipmap = i -
+				(cache->num_mip_levels * cmd2->body.image.face);
+			cmd2->body.box = *box;
+			cmd2++;
+		}
+
+	}
+	vmw_fifo_commit(dev_priv, alloc_size);
+ out:
+	memset(&dirty->boxes[0], 0, sizeof(dirty->boxes[0]) *
+	       dirty->num_subres);
+
+	return 0;
+}
+
+/*
+ * vmw_surface_dirty_alloc - The surface's dirty_alloc callback.
+ */
+static int vmw_surface_dirty_alloc(struct vmw_resource *res)
+{
+	struct vmw_surface *srf = vmw_res_to_srf(res);
+	struct vmw_surface_dirty *dirty;
+	u32 num_layers = 1;
+	u32 num_mip;
+	u32 num_subres;
+	u32 num_samples;
+	size_t dirty_size, acc_size;
+	static struct ttm_operation_ctx ctx = {
+		.interruptible = false,
+		.no_wait_gpu = false
+	};
+	int ret;
+
+	if (srf->array_size)
+		num_layers = srf->array_size;
+	else if (srf->flags & SVGA3D_SURFACE_CUBEMAP)
+		num_layers *= SVGA3D_MAX_SURFACE_FACES;
+
+	num_mip = srf->mip_levels[0];
+	if (!num_mip)
+		num_mip = 1;
+
+	num_subres = num_layers * num_mip;
+	dirty_size = sizeof(*dirty) + num_subres * sizeof(dirty->boxes[0]);
+	acc_size = ttm_round_pot(dirty_size);
+	ret = ttm_mem_global_alloc(vmw_mem_glob(res->dev_priv),
+				   acc_size, &ctx);
+	if (ret) {
+		VMW_DEBUG_USER("Out of graphics memory for surface "
+			       "dirty tracker.\n");
+		return ret;
+	}
+
+	dirty = kvzalloc(dirty_size, GFP_KERNEL);
+	if (!dirty) {
+		ret = -ENOMEM;
+		goto out_no_dirty;
+	}
+
+	num_samples = max_t(u32, 1, srf->multisample_count);
+	ret = svga3dsurface_setup_cache(&srf->base_size, srf->format, num_mip,
+					num_layers, num_samples, &dirty->cache);
+	if (ret)
+		goto out_no_cache;
+
+	dirty->num_subres = num_subres;
+	dirty->size = acc_size;
+	res->dirty = (struct vmw_resource_dirty *) dirty;
+
+	return 0;
+
+out_no_cache:
+	kvfree(dirty);
+out_no_dirty:
+	ttm_mem_global_free(vmw_mem_glob(res->dev_priv), acc_size);
+	return ret;
+}
+
+/*
+ * vmw_surface_dirty_free - The surface's dirty_free callback
+ */
+static void vmw_surface_dirty_free(struct vmw_resource *res)
+{
+	struct vmw_surface_dirty *dirty =
+		(struct vmw_surface_dirty *) res->dirty;
+	size_t acc_size = dirty->size;
+
+	kvfree(dirty);
+	ttm_mem_global_free(vmw_mem_glob(res->dev_priv), acc_size);
+	res->dirty = NULL;
+}
+
+/*
+ * vmw_surface_clean - The surface's clean callback
+ */
+static int vmw_surface_clean(struct vmw_resource *res)
+{
+	struct vmw_private *dev_priv = res->dev_priv;
+	size_t alloc_size;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdReadbackGBSurface body;
+	} *cmd;
+
+	alloc_size = sizeof(*cmd);
+	cmd = VMW_FIFO_RESERVE(dev_priv, alloc_size);
+	if (!cmd)
+		return -ENOMEM;
+
+	cmd->header.id = SVGA_3D_CMD_READBACK_GB_SURFACE;
+	cmd->header.size = sizeof(cmd->body);
+	cmd->body.sid = res->id;
+	vmw_fifo_commit(dev_priv, alloc_size);
+
+	return 0;
+}
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_ttm_glue.c b/drivers/gpu/drm/vmwgfx/vmwgfx_ttm_glue.c
index 5a7b8bb420de..ce288756531b 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_ttm_glue.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_ttm_glue.c
@@ -29,10 +29,23 @@
 
 int vmw_mmap(struct file *filp, struct vm_area_struct *vma)
 {
+	static const struct vm_operations_struct vmw_vm_ops = {
+		.pfn_mkwrite = vmw_bo_vm_mkwrite,
+		.page_mkwrite = vmw_bo_vm_mkwrite,
+		.fault = vmw_bo_vm_fault,
+		.open = ttm_bo_vm_open,
+		.close = ttm_bo_vm_close
+	};
 	struct drm_file *file_priv = filp->private_data;
 	struct vmw_private *dev_priv = vmw_priv(file_priv->minor->dev);
+	int ret = ttm_bo_mmap(filp, vma, &dev_priv->bdev);
 
-	return ttm_bo_mmap(filp, vma, &dev_priv->bdev);
+	if (ret)
+		return ret;
+
+	vma->vm_ops = &vmw_vm_ops;
+
+	return 0;
 }
 
 /* struct vmw_validation_mem callback */
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_validation.c b/drivers/gpu/drm/vmwgfx/vmwgfx_validation.c
index f611b2290a1b..e69bc373ae2e 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_validation.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_validation.c
@@ -33,6 +33,8 @@
  * struct vmw_validation_bo_node - Buffer object validation metadata.
  * @base: Metadata used for TTM reservation- and validation.
  * @hash: A hash entry used for the duplicate detection hash table.
+ * @coherent_count: If switching backup buffers, number of new coherent
+ * resources that will have this buffer as a backup buffer.
  * @as_mob: Validate as mob.
  * @cpu_blit: Validate for cpu blit access.
  *
@@ -42,6 +44,7 @@
 struct vmw_validation_bo_node {
 	struct ttm_validate_buffer base;
 	struct drm_hash_item hash;
+	unsigned int coherent_count;
 	u32 as_mob : 1;
 	u32 cpu_blit : 1;
 };
@@ -459,6 +462,19 @@ int vmw_validation_res_reserve(struct vmw_validation_context *ctx,
 			if (ret)
 				goto out_unreserve;
 		}
+
+		if (val->switching_backup && val->new_backup &&
+		    res->coherent) {
+			struct vmw_validation_bo_node *bo_node =
+				vmw_validation_find_bo_dup(ctx,
+							   val->new_backup);
+
+			if (WARN_ON(!bo_node)) {
+				ret = -EINVAL;
+				goto out_unreserve;
+			}
+			bo_node->coherent_count++;
+		}
 	}
 
 	return 0;
@@ -521,6 +537,9 @@ int vmw_validation_bo_validate_single(struct ttm_buffer_object *bo,
 	};
 	int ret;
 
+	if (atomic_read(&vbo->cpu_writers))
+		return -EBUSY;
+
 	if (vbo->pin_count > 0)
 		return 0;
 
@@ -562,6 +581,9 @@ int vmw_validation_bo_validate(struct vmw_validation_context *ctx, bool intr)
 	int ret;
 
 	list_for_each_entry(entry, &ctx->bo_list, base.head) {
+		struct vmw_buffer_object *vbo =
+			container_of(entry->base.bo, typeof(*vbo), base);
+
 		if (entry->cpu_blit) {
 			struct ttm_operation_ctx ctx = {
 				.interruptible = intr,
@@ -576,6 +598,27 @@ int vmw_validation_bo_validate(struct vmw_validation_context *ctx, bool intr)
 		}
 		if (ret)
 			return ret;
+
+		/*
+		 * Rather than having the resource code allocating the bo
+		 * dirty tracker in resource_unreserve() where we can't fail,
+		 * Do it here when validating the buffer object.
+		 */
+		if (entry->coherent_count) {
+			unsigned int coherent_count = entry->coherent_count;
+
+			while (coherent_count) {
+				ret = vmw_bo_dirty_add(vbo);
+				if (ret)
+					return ret;
+
+				coherent_count--;
+			}
+			entry->coherent_count -= coherent_count;
+		}
+
+		if (vbo->dirty)
+			vmw_bo_dirty_scan(vbo);
 	}
 	return 0;
 }
@@ -601,7 +644,8 @@ int vmw_validation_res_validate(struct vmw_validation_context *ctx, bool intr)
 		struct vmw_resource *res = val->res;
 		struct vmw_buffer_object *backup = res->backup;
 
-		ret = vmw_resource_validate(res, intr);
+		ret = vmw_resource_validate(res, intr, val->dirty_set &&
+					    val->dirty);
 		if (ret) {
 			if (ret != -ERESTARTSYS)
 				DRM_ERROR("Failed to validate resource.\n");
@@ -828,3 +872,34 @@ int vmw_validation_preload_res(struct vmw_validation_context *ctx,
 	ctx->mem_size_left += size;
 	return 0;
 }
+
+/**
+ * vmw_validation_bo_backoff - Unreserve buffer objects registered with a
+ * validation context
+ * @ctx: The validation context
+ *
+ * This function unreserves the buffer objects previously reserved using
+ * vmw_validation_bo_reserve. It's typically used as part of an error path
+ */
+void vmw_validation_bo_backoff(struct vmw_validation_context *ctx)
+{
+	struct vmw_validation_bo_node *entry;
+
+	/*
+	 * Switching coherent resource backup buffers failed.
+	 * Release corresponding buffer object dirty trackers.
+	 */
+	list_for_each_entry(entry, &ctx->bo_list, base.head) {
+		if (entry->coherent_count) {
+			unsigned int coherent_count = entry->coherent_count;
+			struct vmw_buffer_object *vbo =
+				container_of(entry->base.bo, typeof(*vbo),
+					     base);
+
+			while (coherent_count--)
+				vmw_bo_dirty_release(vbo);
+		}
+	}
+
+	ttm_eu_backoff_reservation(&ctx->ticket, &ctx->bo_list);
+}
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_validation.h b/drivers/gpu/drm/vmwgfx/vmwgfx_validation.h
index 0e063743dd86..739906d1b3eb 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_validation.h
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_validation.h
@@ -170,21 +170,7 @@ vmw_validation_bo_reserve(struct vmw_validation_context *ctx,
 			  bool intr)
 {
 	return ttm_eu_reserve_buffers(&ctx->ticket, &ctx->bo_list, intr,
-				      NULL, true);
-}
-
-/**
- * vmw_validation_bo_backoff - Unreserve buffer objects registered with a
- * validation context
- * @ctx: The validation context
- *
- * This function unreserves the buffer objects previously reserved using
- * vmw_validation_bo_reserve. It's typically used as part of an error path
- */
-static inline void
-vmw_validation_bo_backoff(struct vmw_validation_context *ctx)
-{
-	ttm_eu_backoff_reservation(&ctx->ticket, &ctx->bo_list);
+				      NULL);
 }
 
 /**
@@ -269,4 +255,6 @@ int vmw_validation_preload_res(struct vmw_validation_context *ctx,
 			       unsigned int size);
 void vmw_validation_res_set_dirty(struct vmw_validation_context *ctx,
 				  void *val_private, u32 dirty);
+void vmw_validation_bo_backoff(struct vmw_validation_context *ctx);
+
 #endif
diff --git a/drivers/gpu/drm/xen/xen_drm_front.c b/drivers/gpu/drm/xen/xen_drm_front.c
index ba1828acd8c9..4be49c1aef51 100644
--- a/drivers/gpu/drm/xen/xen_drm_front.c
+++ b/drivers/gpu/drm/xen/xen_drm_front.c
@@ -718,17 +718,9 @@ static int xen_drv_probe(struct xenbus_device *xb_dev,
 	struct device *dev = &xb_dev->dev;
 	int ret;
 
-	/*
-	 * The device is not spawn from a device tree, so arch_setup_dma_ops
-	 * is not called, thus leaving the device with dummy DMA ops.
-	 * This makes the device return error on PRIME buffer import, which
-	 * is not correct: to fix this call of_dma_configure() with a NULL
-	 * node to set default DMA ops.
-	 */
-	dev->coherent_dma_mask = DMA_BIT_MASK(32);
-	ret = of_dma_configure(dev, NULL, true);
+	ret = dma_coerce_mask_and_coherent(dev, DMA_BIT_MASK(64));
 	if (ret < 0) {
-		DRM_ERROR("Cannot setup DMA ops, ret %d", ret);
+		DRM_ERROR("Cannot setup DMA mask, ret %d", ret);
 		return ret;
 	}
 
diff --git a/drivers/gpu/drm/xen/xen_drm_front_kms.c b/drivers/gpu/drm/xen/xen_drm_front_kms.c
index 21ad1c359b61..ff506bc99414 100644
--- a/drivers/gpu/drm/xen/xen_drm_front_kms.c
+++ b/drivers/gpu/drm/xen/xen_drm_front_kms.c
@@ -270,11 +270,12 @@ static void display_update(struct drm_simple_display_pipe *pipe,
 }
 
 static enum drm_mode_status
-display_mode_valid(struct drm_crtc *crtc, const struct drm_display_mode *mode)
+display_mode_valid(struct drm_simple_display_pipe *pipe,
+		   const struct drm_display_mode *mode)
 {
 	struct xen_drm_front_drm_pipeline *pipeline =
-			container_of(crtc, struct xen_drm_front_drm_pipeline,
-				     pipe.crtc);
+			container_of(pipe, struct xen_drm_front_drm_pipeline,
+				     pipe);
 
 	if (mode->hdisplay != pipeline->width)
 		return MODE_ERROR;
diff --git a/drivers/gpu/host1x/Kconfig b/drivers/gpu/host1x/Kconfig
index cf987a317a55..6dab94adf25e 100644
--- a/drivers/gpu/host1x/Kconfig
+++ b/drivers/gpu/host1x/Kconfig
@@ -2,7 +2,7 @@
 config TEGRA_HOST1X
 	tristate "NVIDIA Tegra host1x driver"
 	depends on ARCH_TEGRA || (ARM && COMPILE_TEST)
-	select IOMMU_IOVA if IOMMU_SUPPORT
+	select IOMMU_IOVA
 	help
 	  Driver for the NVIDIA Tegra host1x hardware.
 
diff --git a/drivers/gpu/host1x/bus.c b/drivers/gpu/host1x/bus.c
index 742aa9ff21b8..2c8559ff3481 100644
--- a/drivers/gpu/host1x/bus.c
+++ b/drivers/gpu/host1x/bus.c
@@ -445,7 +445,7 @@ static int host1x_device_add(struct host1x *host1x,
 	of_dma_configure(&device->dev, host1x->dev->of_node, true);
 
 	device->dev.dma_parms = &device->dma_parms;
-	dma_set_max_seg_size(&device->dev, SZ_4M);
+	dma_set_max_seg_size(&device->dev, UINT_MAX);
 
 	err = host1x_device_parse_dt(device, driver);
 	if (err < 0) {
diff --git a/drivers/gpu/host1x/cdma.c b/drivers/gpu/host1x/cdma.c
index 48c84c48299c..e8d3fda91d8a 100644
--- a/drivers/gpu/host1x/cdma.c
+++ b/drivers/gpu/host1x/cdma.c
@@ -232,9 +232,9 @@ unsigned int host1x_cdma_wait_locked(struct host1x_cdma *cdma,
  *
  * Must be called with the cdma lock held.
  */
-int host1x_cdma_wait_pushbuffer_space(struct host1x *host1x,
-				      struct host1x_cdma *cdma,
-				      unsigned int needed)
+static int host1x_cdma_wait_pushbuffer_space(struct host1x *host1x,
+					     struct host1x_cdma *cdma,
+					     unsigned int needed)
 {
 	while (true) {
 		struct push_buffer *pb = &cdma->push_buffer;
diff --git a/drivers/gpu/host1x/channel.c b/drivers/gpu/host1x/channel.c
index 1436295aa450..4cd212bb570d 100644
--- a/drivers/gpu/host1x/channel.c
+++ b/drivers/gpu/host1x/channel.c
@@ -115,14 +115,14 @@ static struct host1x_channel *acquire_unused_channel(struct host1x *host)
 
 /**
  * host1x_channel_request() - Allocate a channel
- * @device: Host1x unit this channel will be used to send commands to
+ * @client: Host1x client this channel will be used to send commands to
  *
- * Allocates a new host1x channel for @device. May return NULL if CDMA
+ * Allocates a new host1x channel for @client. May return NULL if CDMA
  * initialization fails.
  */
-struct host1x_channel *host1x_channel_request(struct device *dev)
+struct host1x_channel *host1x_channel_request(struct host1x_client *client)
 {
-	struct host1x *host = dev_get_drvdata(dev->parent);
+	struct host1x *host = dev_get_drvdata(client->dev->parent);
 	struct host1x_channel_list *chlist = &host->channel_list;
 	struct host1x_channel *channel;
 	int err;
@@ -133,7 +133,8 @@ struct host1x_channel *host1x_channel_request(struct device *dev)
 
 	kref_init(&channel->refcount);
 	mutex_init(&channel->submitlock);
-	channel->dev = dev;
+	channel->client = client;
+	channel->dev = client->dev;
 
 	err = host1x_hw_channel_init(host, channel, channel->id);
 	if (err < 0)
@@ -148,7 +149,7 @@ struct host1x_channel *host1x_channel_request(struct device *dev)
 fail:
 	clear_bit(channel->id, chlist->allocated_channels);
 
-	dev_err(dev, "failed to initialize channel\n");
+	dev_err(client->dev, "failed to initialize channel\n");
 
 	return NULL;
 }
diff --git a/drivers/gpu/host1x/channel.h b/drivers/gpu/host1x/channel.h
index 4fd694834f74..39044ff6c3aa 100644
--- a/drivers/gpu/host1x/channel.h
+++ b/drivers/gpu/host1x/channel.h
@@ -26,6 +26,7 @@ struct host1x_channel {
 	unsigned int id;
 	struct mutex submitlock;
 	void __iomem *regs;
+	struct host1x_client *client;
 	struct device *dev;
 	struct host1x_cdma cdma;
 };
diff --git a/drivers/gpu/host1x/dev.c b/drivers/gpu/host1x/dev.c
index 5a3f797240d4..a738ea55e407 100644
--- a/drivers/gpu/host1x/dev.c
+++ b/drivers/gpu/host1x/dev.c
@@ -18,10 +18,6 @@
 #include <trace/events/host1x.h>
 #undef CREATE_TRACE_POINTS
 
-#if IS_ENABLED(CONFIG_ARM_DMA_USE_IOMMU)
-#include <asm/dma-iommu.h>
-#endif
-
 #include "bus.h"
 #include "channel.h"
 #include "debug.h"
@@ -77,6 +73,10 @@ static const struct host1x_info host1x01_info = {
 	.init = host1x01_init,
 	.sync_offset = 0x3000,
 	.dma_mask = DMA_BIT_MASK(32),
+	.has_wide_gather = false,
+	.has_hypervisor = false,
+	.num_sid_entries = 0,
+	.sid_table = NULL,
 };
 
 static const struct host1x_info host1x02_info = {
@@ -87,6 +87,10 @@ static const struct host1x_info host1x02_info = {
 	.init = host1x02_init,
 	.sync_offset = 0x3000,
 	.dma_mask = DMA_BIT_MASK(32),
+	.has_wide_gather = false,
+	.has_hypervisor = false,
+	.num_sid_entries = 0,
+	.sid_table = NULL,
 };
 
 static const struct host1x_info host1x04_info = {
@@ -97,6 +101,10 @@ static const struct host1x_info host1x04_info = {
 	.init = host1x04_init,
 	.sync_offset = 0x2100,
 	.dma_mask = DMA_BIT_MASK(34),
+	.has_wide_gather = false,
+	.has_hypervisor = false,
+	.num_sid_entries = 0,
+	.sid_table = NULL,
 };
 
 static const struct host1x_info host1x05_info = {
@@ -107,6 +115,10 @@ static const struct host1x_info host1x05_info = {
 	.init = host1x05_init,
 	.sync_offset = 0x2100,
 	.dma_mask = DMA_BIT_MASK(34),
+	.has_wide_gather = false,
+	.has_hypervisor = false,
+	.num_sid_entries = 0,
+	.sid_table = NULL,
 };
 
 static const struct host1x_sid_entry tegra186_sid_table[] = {
@@ -126,6 +138,7 @@ static const struct host1x_info host1x06_info = {
 	.init = host1x06_init,
 	.sync_offset = 0x0,
 	.dma_mask = DMA_BIT_MASK(40),
+	.has_wide_gather = true,
 	.has_hypervisor = true,
 	.num_sid_entries = ARRAY_SIZE(tegra186_sid_table),
 	.sid_table = tegra186_sid_table,
@@ -148,6 +161,7 @@ static const struct host1x_info host1x07_info = {
 	.init = host1x07_init,
 	.sync_offset = 0x0,
 	.dma_mask = DMA_BIT_MASK(40),
+	.has_wide_gather = true,
 	.has_hypervisor = true,
 	.num_sid_entries = ARRAY_SIZE(tegra194_sid_table),
 	.sid_table = tegra194_sid_table,
@@ -178,6 +192,117 @@ static void host1x_setup_sid_table(struct host1x *host)
 	}
 }
 
+static struct iommu_domain *host1x_iommu_attach(struct host1x *host)
+{
+	struct iommu_domain *domain = iommu_get_domain_for_dev(host->dev);
+	int err;
+
+	/*
+	 * If the host1x firewall is enabled, there's no need to enable IOMMU
+	 * support. Similarly, if host1x is already attached to an IOMMU (via
+	 * the DMA API), don't try to attach again.
+	 */
+	if (IS_ENABLED(CONFIG_TEGRA_HOST1X_FIREWALL) || domain)
+		return domain;
+
+	host->group = iommu_group_get(host->dev);
+	if (host->group) {
+		struct iommu_domain_geometry *geometry;
+		dma_addr_t start, end;
+		unsigned long order;
+
+		err = iova_cache_get();
+		if (err < 0)
+			goto put_group;
+
+		host->domain = iommu_domain_alloc(&platform_bus_type);
+		if (!host->domain) {
+			err = -ENOMEM;
+			goto put_cache;
+		}
+
+		err = iommu_attach_group(host->domain, host->group);
+		if (err) {
+			if (err == -ENODEV)
+				err = 0;
+
+			goto free_domain;
+		}
+
+		geometry = &host->domain->geometry;
+		start = geometry->aperture_start & host->info->dma_mask;
+		end = geometry->aperture_end & host->info->dma_mask;
+
+		order = __ffs(host->domain->pgsize_bitmap);
+		init_iova_domain(&host->iova, 1UL << order, start >> order);
+		host->iova_end = end;
+
+		domain = host->domain;
+	}
+
+	return domain;
+
+free_domain:
+	iommu_domain_free(host->domain);
+	host->domain = NULL;
+put_cache:
+	iova_cache_put();
+put_group:
+	iommu_group_put(host->group);
+	host->group = NULL;
+
+	return ERR_PTR(err);
+}
+
+static int host1x_iommu_init(struct host1x *host)
+{
+	u64 mask = host->info->dma_mask;
+	struct iommu_domain *domain;
+	int err;
+
+	domain = host1x_iommu_attach(host);
+	if (IS_ERR(domain)) {
+		err = PTR_ERR(domain);
+		dev_err(host->dev, "failed to attach to IOMMU: %d\n", err);
+		return err;
+	}
+
+	/*
+	 * If we're not behind an IOMMU make sure we don't get push buffers
+	 * that are allocated outside of the range addressable by the GATHER
+	 * opcode.
+	 *
+	 * Newer generations of Tegra (Tegra186 and later) support a wide
+	 * variant of the GATHER opcode that allows addressing more bits.
+	 */
+	if (!domain && !host->info->has_wide_gather)
+		mask = DMA_BIT_MASK(32);
+
+	err = dma_coerce_mask_and_coherent(host->dev, mask);
+	if (err < 0) {
+		dev_err(host->dev, "failed to set DMA mask: %d\n", err);
+		return err;
+	}
+
+	return 0;
+}
+
+static void host1x_iommu_exit(struct host1x *host)
+{
+	if (host->domain) {
+		put_iova_domain(&host->iova);
+		iommu_detach_group(host->domain, host->group);
+
+		iommu_domain_free(host->domain);
+		host->domain = NULL;
+
+		iova_cache_put();
+
+		iommu_group_put(host->group);
+		host->group = NULL;
+	}
+}
+
 static int host1x_probe(struct platform_device *pdev)
 {
 	struct host1x *host;
@@ -237,7 +362,8 @@ static int host1x_probe(struct platform_device *pdev)
 			return PTR_ERR(host->hv_regs);
 	}
 
-	dma_set_mask_and_coherent(host->dev, host->info->dma_mask);
+	host->dev->dma_parms = &host->dma_parms;
+	dma_set_max_seg_size(host->dev, UINT_MAX);
 
 	if (host->info->init) {
 		err = host->info->init(host);
@@ -261,87 +387,42 @@ static int host1x_probe(struct platform_device *pdev)
 		dev_err(&pdev->dev, "failed to get reset: %d\n", err);
 		return err;
 	}
-#if IS_ENABLED(CONFIG_ARM_DMA_USE_IOMMU)
-	if (host->dev->archdata.mapping) {
-		struct dma_iommu_mapping *mapping =
-				to_dma_iommu_mapping(host->dev);
-		arm_iommu_detach_device(host->dev);
-		arm_iommu_release_mapping(mapping);
-	}
-#endif
-	if (IS_ENABLED(CONFIG_TEGRA_HOST1X_FIREWALL))
-		goto skip_iommu;
-
-	host->group = iommu_group_get(&pdev->dev);
-	if (host->group) {
-		struct iommu_domain_geometry *geometry;
-		u64 mask = dma_get_mask(host->dev);
-		dma_addr_t start, end;
-		unsigned long order;
-
-		err = iova_cache_get();
-		if (err < 0)
-			goto put_group;
-
-		host->domain = iommu_domain_alloc(&platform_bus_type);
-		if (!host->domain) {
-			err = -ENOMEM;
-			goto put_cache;
-		}
 
-		err = iommu_attach_group(host->domain, host->group);
-		if (err) {
-			if (err == -ENODEV) {
-				iommu_domain_free(host->domain);
-				host->domain = NULL;
-				iova_cache_put();
-				iommu_group_put(host->group);
-				host->group = NULL;
-				goto skip_iommu;
-			}
-
-			goto fail_free_domain;
-		}
-
-		geometry = &host->domain->geometry;
-		start = geometry->aperture_start & mask;
-		end = geometry->aperture_end & mask;
-
-		order = __ffs(host->domain->pgsize_bitmap);
-		init_iova_domain(&host->iova, 1UL << order, start >> order);
-		host->iova_end = end;
+	err = host1x_iommu_init(host);
+	if (err < 0) {
+		dev_err(&pdev->dev, "failed to setup IOMMU: %d\n", err);
+		return err;
 	}
 
-skip_iommu:
 	err = host1x_channel_list_init(&host->channel_list,
 				       host->info->nb_channels);
 	if (err) {
 		dev_err(&pdev->dev, "failed to initialize channel list\n");
-		goto fail_detach_device;
+		goto iommu_exit;
 	}
 
 	err = clk_prepare_enable(host->clk);
 	if (err < 0) {
 		dev_err(&pdev->dev, "failed to enable clock\n");
-		goto fail_free_channels;
+		goto free_channels;
 	}
 
 	err = reset_control_deassert(host->rst);
 	if (err < 0) {
 		dev_err(&pdev->dev, "failed to deassert reset: %d\n", err);
-		goto fail_unprepare_disable;
+		goto unprepare_disable;
 	}
 
 	err = host1x_syncpt_init(host);
 	if (err) {
 		dev_err(&pdev->dev, "failed to initialize syncpts\n");
-		goto fail_reset_assert;
+		goto reset_assert;
 	}
 
 	err = host1x_intr_init(host, syncpt_irq);
 	if (err) {
 		dev_err(&pdev->dev, "failed to initialize interrupts\n");
-		goto fail_deinit_syncpt;
+		goto deinit_syncpt;
 	}
 
 	host1x_debug_init(host);
@@ -351,33 +432,22 @@ skip_iommu:
 
 	err = host1x_register(host);
 	if (err < 0)
-		goto fail_deinit_intr;
+		goto deinit_intr;
 
 	return 0;
 
-fail_deinit_intr:
+deinit_intr:
 	host1x_intr_deinit(host);
-fail_deinit_syncpt:
+deinit_syncpt:
 	host1x_syncpt_deinit(host);
-fail_reset_assert:
+reset_assert:
 	reset_control_assert(host->rst);
-fail_unprepare_disable:
+unprepare_disable:
 	clk_disable_unprepare(host->clk);
-fail_free_channels:
+free_channels:
 	host1x_channel_list_free(&host->channel_list);
-fail_detach_device:
-	if (host->group && host->domain) {
-		put_iova_domain(&host->iova);
-		iommu_detach_group(host->domain, host->group);
-	}
-fail_free_domain:
-	if (host->domain)
-		iommu_domain_free(host->domain);
-put_cache:
-	if (host->group)
-		iova_cache_put();
-put_group:
-	iommu_group_put(host->group);
+iommu_exit:
+	host1x_iommu_exit(host);
 
 	return err;
 }
@@ -387,18 +457,12 @@ static int host1x_remove(struct platform_device *pdev)
 	struct host1x *host = platform_get_drvdata(pdev);
 
 	host1x_unregister(host);
+	host1x_debug_deinit(host);
 	host1x_intr_deinit(host);
 	host1x_syncpt_deinit(host);
 	reset_control_assert(host->rst);
 	clk_disable_unprepare(host->clk);
-
-	if (host->domain) {
-		put_iova_domain(&host->iova);
-		iommu_detach_group(host->domain, host->group);
-		iommu_domain_free(host->domain);
-		iova_cache_put();
-		iommu_group_put(host->group);
-	}
+	host1x_iommu_exit(host);
 
 	return 0;
 }
diff --git a/drivers/gpu/host1x/dev.h b/drivers/gpu/host1x/dev.h
index ff56f5e23a02..f781a9b0f39d 100644
--- a/drivers/gpu/host1x/dev.h
+++ b/drivers/gpu/host1x/dev.h
@@ -97,6 +97,7 @@ struct host1x_info {
 	int (*init)(struct host1x *host1x); /* initialize per SoC ops */
 	unsigned int sync_offset; /* offset of syncpoint registers */
 	u64 dma_mask; /* mask of addressable memory */
+	bool has_wide_gather; /* supports GATHER_W opcode */
 	bool has_hypervisor; /* has hypervisor registers */
 	unsigned int num_sid_entries;
 	const struct host1x_sid_entry *sid_table;
@@ -140,6 +141,8 @@ struct host1x {
 	struct list_head devices;
 
 	struct list_head list;
+
+	struct device_dma_parameters dma_parms;
 };
 
 void host1x_hypervisor_writel(struct host1x *host1x, u32 r, u32 v);
diff --git a/drivers/gpu/host1x/intr.c b/drivers/gpu/host1x/intr.c
index 26f3c741d085..9245add23b5d 100644
--- a/drivers/gpu/host1x/intr.c
+++ b/drivers/gpu/host1x/intr.c
@@ -105,7 +105,6 @@ static void action_submit_complete(struct host1x_waitlist *waiter)
 	/*  Add nr_completed to trace */
 	trace_host1x_channel_submit_complete(dev_name(channel->dev),
 					     waiter->count, waiter->thresh);
-
 }
 
 static void action_wakeup(struct host1x_waitlist *waiter)
diff --git a/drivers/gpu/host1x/job.c b/drivers/gpu/host1x/job.c
index eaa5c3352c13..25ca54de8fc5 100644
--- a/drivers/gpu/host1x/job.c
+++ b/drivers/gpu/host1x/job.c
@@ -99,6 +99,8 @@ EXPORT_SYMBOL(host1x_job_add_gather);
 
 static unsigned int pin_job(struct host1x *host, struct host1x_job *job)
 {
+	struct host1x_client *client = job->client;
+	struct device *dev = client->dev;
 	unsigned int i;
 	int err;
 
@@ -106,8 +108,8 @@ static unsigned int pin_job(struct host1x *host, struct host1x_job *job)
 
 	for (i = 0; i < job->num_relocs; i++) {
 		struct host1x_reloc *reloc = &job->relocs[i];
+		dma_addr_t phys_addr, *phys;
 		struct sg_table *sgt;
-		dma_addr_t phys_addr;
 
 		reloc->target.bo = host1x_bo_get(reloc->target.bo);
 		if (!reloc->target.bo) {
@@ -115,7 +117,50 @@ static unsigned int pin_job(struct host1x *host, struct host1x_job *job)
 			goto unpin;
 		}
 
-		phys_addr = host1x_bo_pin(reloc->target.bo, &sgt);
+		if (client->group)
+			phys = &phys_addr;
+		else
+			phys = NULL;
+
+		sgt = host1x_bo_pin(dev, reloc->target.bo, phys);
+		if (IS_ERR(sgt)) {
+			err = PTR_ERR(sgt);
+			goto unpin;
+		}
+
+		if (sgt) {
+			unsigned long mask = HOST1X_RELOC_READ |
+					     HOST1X_RELOC_WRITE;
+			enum dma_data_direction dir;
+
+			switch (reloc->flags & mask) {
+			case HOST1X_RELOC_READ:
+				dir = DMA_TO_DEVICE;
+				break;
+
+			case HOST1X_RELOC_WRITE:
+				dir = DMA_FROM_DEVICE;
+				break;
+
+			case HOST1X_RELOC_READ | HOST1X_RELOC_WRITE:
+				dir = DMA_BIDIRECTIONAL;
+				break;
+
+			default:
+				err = -EINVAL;
+				goto unpin;
+			}
+
+			err = dma_map_sg(dev, sgt->sgl, sgt->nents, dir);
+			if (!err) {
+				err = -ENOMEM;
+				goto unpin;
+			}
+
+			job->unpins[job->num_unpins].dev = dev;
+			job->unpins[job->num_unpins].dir = dir;
+			phys_addr = sg_dma_address(sgt->sgl);
+		}
 
 		job->addr_phys[job->num_unpins] = phys_addr;
 		job->unpins[job->num_unpins].bo = reloc->target.bo;
@@ -139,7 +184,11 @@ static unsigned int pin_job(struct host1x *host, struct host1x_job *job)
 			goto unpin;
 		}
 
-		phys_addr = host1x_bo_pin(g->bo, &sgt);
+		sgt = host1x_bo_pin(host->dev, g->bo, NULL);
+		if (IS_ERR(sgt)) {
+			err = PTR_ERR(sgt);
+			goto unpin;
+		}
 
 		if (!IS_ENABLED(CONFIG_TEGRA_HOST1X_FIREWALL) && host->domain) {
 			for_each_sg(sgt->sgl, sg, sgt->nents, j)
@@ -163,15 +212,24 @@ static unsigned int pin_job(struct host1x *host, struct host1x_job *job)
 				goto unpin;
 			}
 
-			job->addr_phys[job->num_unpins] =
-				iova_dma_addr(&host->iova, alloc);
 			job->unpins[job->num_unpins].size = gather_size;
+			phys_addr = iova_dma_addr(&host->iova, alloc);
 		} else {
-			job->addr_phys[job->num_unpins] = phys_addr;
+			err = dma_map_sg(host->dev, sgt->sgl, sgt->nents,
+					 DMA_TO_DEVICE);
+			if (!err) {
+				err = -ENOMEM;
+				goto unpin;
+			}
+
+			job->unpins[job->num_unpins].dev = host->dev;
+			phys_addr = sg_dma_address(sgt->sgl);
 		}
 
-		job->gather_addr_phys[i] = job->addr_phys[job->num_unpins];
+		job->addr_phys[job->num_unpins] = phys_addr;
+		job->gather_addr_phys[i] = phys_addr;
 
+		job->unpins[job->num_unpins].dir = DMA_TO_DEVICE;
 		job->unpins[job->num_unpins].bo = g->bo;
 		job->unpins[job->num_unpins].sgt = sgt;
 		job->num_unpins++;
@@ -436,7 +494,8 @@ out:
 	return err;
 }
 
-static inline int copy_gathers(struct host1x_job *job, struct device *dev)
+static inline int copy_gathers(struct device *host, struct host1x_job *job,
+			       struct device *dev)
 {
 	struct host1x_firewall fw;
 	size_t size = 0;
@@ -459,12 +518,12 @@ static inline int copy_gathers(struct host1x_job *job, struct device *dev)
 	 * Try a non-blocking allocation from a higher priority pools first,
 	 * as awaiting for the allocation here is a major performance hit.
 	 */
-	job->gather_copy_mapped = dma_alloc_wc(dev, size, &job->gather_copy,
+	job->gather_copy_mapped = dma_alloc_wc(host, size, &job->gather_copy,
 					       GFP_NOWAIT);
 
 	/* the higher priority allocation failed, try the generic-blocking */
 	if (!job->gather_copy_mapped)
-		job->gather_copy_mapped = dma_alloc_wc(dev, size,
+		job->gather_copy_mapped = dma_alloc_wc(host, size,
 						       &job->gather_copy,
 						       GFP_KERNEL);
 	if (!job->gather_copy_mapped)
@@ -512,7 +571,7 @@ int host1x_job_pin(struct host1x_job *job, struct device *dev)
 		goto out;
 
 	if (IS_ENABLED(CONFIG_TEGRA_HOST1X_FIREWALL)) {
-		err = copy_gathers(job, dev);
+		err = copy_gathers(host->dev, job, dev);
 		if (err)
 			goto out;
 	}
@@ -557,6 +616,8 @@ void host1x_job_unpin(struct host1x_job *job)
 
 	for (i = 0; i < job->num_unpins; i++) {
 		struct host1x_job_unpin_data *unpin = &job->unpins[i];
+		struct device *dev = unpin->dev ?: host->dev;
+		struct sg_table *sgt = unpin->sgt;
 
 		if (!IS_ENABLED(CONFIG_TEGRA_HOST1X_FIREWALL) &&
 		    unpin->size && host->domain) {
@@ -566,14 +627,18 @@ void host1x_job_unpin(struct host1x_job *job)
 				iova_pfn(&host->iova, job->addr_phys[i]));
 		}
 
-		host1x_bo_unpin(unpin->bo, unpin->sgt);
+		if (unpin->dev && sgt)
+			dma_unmap_sg(unpin->dev, sgt->sgl, sgt->nents,
+				     unpin->dir);
+
+		host1x_bo_unpin(dev, unpin->bo, sgt);
 		host1x_bo_put(unpin->bo);
 	}
 
 	job->num_unpins = 0;
 
 	if (job->gather_copy_size)
-		dma_free_wc(job->channel->dev, job->gather_copy_size,
+		dma_free_wc(host->dev, job->gather_copy_size,
 			    job->gather_copy_mapped, job->gather_copy);
 }
 EXPORT_SYMBOL(host1x_job_unpin);
diff --git a/drivers/gpu/host1x/job.h b/drivers/gpu/host1x/job.h
index 62b8805e6b35..94bc2e4ae241 100644
--- a/drivers/gpu/host1x/job.h
+++ b/drivers/gpu/host1x/job.h
@@ -8,6 +8,8 @@
 #ifndef __HOST1X_JOB_H
 #define __HOST1X_JOB_H
 
+#include <linux/dma-direction.h>
+
 struct host1x_job_gather {
 	unsigned int words;
 	dma_addr_t base;
@@ -19,7 +21,9 @@ struct host1x_job_gather {
 struct host1x_job_unpin_data {
 	struct host1x_bo *bo;
 	struct sg_table *sgt;
+	struct device *dev;
 	size_t size;
+	enum dma_data_direction dir;
 };
 
 /*