Merge tag 'drm-fixes-2020-04-18' of git://anongit.freedesktop.org/drm/drm

Pull drm fixes from Dave Airlie:
 "Quiet enough for rc2, mostly amdgpu fixes, a couple of i915 fixes, and
  one nouveau module firmware fix:

  i915:
   - Fix guest page access by using the brand new VFIO dma r/w interface (Yan)
   - Fix for i915 perf read buffers (Ashutosh)

  amdgpu:
   - gfx10 fix
   - SMU7 overclocking fix
   - RAS fix
   - GPU reset fix
   - Fix a regression in a previous suspend/resume fix
   - Add a gfxoff quirk

  nouveau:
   - fix missing MODULE_FIRMWARE"

* tag 'drm-fixes-2020-04-18' of git://anongit.freedesktop.org/drm/drm:
  drm/nouveau/sec2/gv100-: add missing MODULE_FIRMWARE()
  drm/amdgpu/gfx9: add gfxoff quirk
  drm/amdgpu: fix the hw hang during perform system reboot and reset
  drm/i915/gvt: switch to user vfio_group_pin/upin_pages
  drm/i915/gvt: subsitute kvm_read/write_guest with vfio_dma_rw
  drm/i915/gvt: hold reference of VFIO group during opening of vgpu
  drm/i915/perf: Do not clear pollin for small user read buffers
  drm/amdgpu: fix wrong vram lost counter increment V2
  drm/amd/powerplay: unload mp1 for Arcturus RAS baco reset
  drm/amd/powerplay: force the trim of the mclk dpm_levels if OD is enabled
  Revert "drm/amdgpu: change SH MEM alignment mode for gfx10"

7 files changed, 23 insertions, 15 deletions

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index 559dc24ef436..f84f9e35a73b 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -2008,8 +2008,24 @@ static void amdgpu_device_fill_reset_magic(struct amdgpu_device *adev)
  */
 static bool amdgpu_device_check_vram_lost(struct amdgpu_device *adev)
 {
-	return !!memcmp(adev->gart.ptr, adev->reset_magic,
-			AMDGPU_RESET_MAGIC_NUM);
+	if (memcmp(adev->gart.ptr, adev->reset_magic,
+			AMDGPU_RESET_MAGIC_NUM))
+		return true;
+
+	if (!adev->in_gpu_reset)
+		return false;
+
+	/*
+	 * For all ASICs with baco/mode1 reset, the VRAM is
+	 * always assumed to be lost.
+	 */
+	switch (amdgpu_asic_reset_method(adev)) {
+	case AMD_RESET_METHOD_BACO:
+	case AMD_RESET_METHOD_MODE1:
+		return true;
+	default:
+		return false;
+	}
 }
 
 /**
@@ -2340,6 +2356,8 @@ static int amdgpu_device_ip_suspend_phase1(struct amdgpu_device *adev)
 {
 	int i, r;
 
+	amdgpu_device_set_pg_state(adev, AMD_PG_STATE_UNGATE);
+	amdgpu_device_set_cg_state(adev, AMD_CG_STATE_UNGATE);
 
 	for (i = adev->num_ip_blocks - 1; i >= 0; i--) {
 		if (!adev->ip_blocks[i].status.valid)
diff --git a/drivers/gpu/drm/amd/amdgpu/cik.c b/drivers/gpu/drm/amd/amdgpu/cik.c
index 006f21ef7ddf..62635e58e45e 100644
--- a/drivers/gpu/drm/amd/amdgpu/cik.c
+++ b/drivers/gpu/drm/amd/amdgpu/cik.c
@@ -1358,8 +1358,6 @@ static int cik_asic_reset(struct amdgpu_device *adev)
 	int r;
 
 	if (cik_asic_reset_method(adev) == AMD_RESET_METHOD_BACO) {
-		if (!adev->in_suspend)
-			amdgpu_inc_vram_lost(adev);
 		r = amdgpu_dpm_baco_reset(adev);
 	} else {
 		r = cik_asic_pci_config_reset(adev);
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c
index d78059fd2c72..f92c158d89a1 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c
@@ -279,7 +279,7 @@ static const struct soc15_reg_golden golden_settings_gc_10_1_2_nv12[] =
 
 #define DEFAULT_SH_MEM_CONFIG \
 	((SH_MEM_ADDRESS_MODE_64 << SH_MEM_CONFIG__ADDRESS_MODE__SHIFT) | \
-	 (SH_MEM_ALIGNMENT_MODE_DWORD << SH_MEM_CONFIG__ALIGNMENT_MODE__SHIFT) | \
+	 (SH_MEM_ALIGNMENT_MODE_UNALIGNED << SH_MEM_CONFIG__ALIGNMENT_MODE__SHIFT) | \
 	 (SH_MEM_RETRY_MODE_ALL << SH_MEM_CONFIG__RETRY_MODE__SHIFT) | \
 	 (3 << SH_MEM_CONFIG__INITIAL_INST_PREFETCH__SHIFT))
 
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
index e6b113ed2f40..0c390485bc10 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
@@ -1234,6 +1234,8 @@ struct amdgpu_gfxoff_quirk {
 static const struct amdgpu_gfxoff_quirk amdgpu_gfxoff_quirk_list[] = {
 	/* https://bugzilla.kernel.org/show_bug.cgi?id=204689 */
 	{ 0x1002, 0x15dd, 0x1002, 0x15dd, 0xc8 },
+	/* https://bugzilla.kernel.org/show_bug.cgi?id=207171 */
+	{ 0x1002, 0x15dd, 0x103c, 0x83e7, 0xd3 },
 	{ 0, 0, 0, 0, 0 },
 };
 
diff --git a/drivers/gpu/drm/amd/amdgpu/nv.c b/drivers/gpu/drm/amd/amdgpu/nv.c
index 033cbbca2072..52318b03c424 100644
--- a/drivers/gpu/drm/amd/amdgpu/nv.c
+++ b/drivers/gpu/drm/amd/amdgpu/nv.c
@@ -351,8 +351,6 @@ 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) {
-		if (!adev->in_suspend)
-			amdgpu_inc_vram_lost(adev);
 		ret = smu_baco_enter(smu);
 		if (ret)
 			return ret;
@@ -360,8 +358,6 @@ static int nv_asic_reset(struct amdgpu_device *adev)
 		if (ret)
 			return ret;
 	} else {
-		if (!adev->in_suspend)
-			amdgpu_inc_vram_lost(adev);
 		ret = nv_asic_mode1_reset(adev);
 	}
 
diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c
index a40499d51c93..d42a8d8a0dea 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc15.c
+++ b/drivers/gpu/drm/amd/amdgpu/soc15.c
@@ -569,14 +569,10 @@ static int soc15_asic_reset(struct amdgpu_device *adev)
 
 	switch (soc15_asic_reset_method(adev)) {
 		case AMD_RESET_METHOD_BACO:
-			if (!adev->in_suspend)
-				amdgpu_inc_vram_lost(adev);
 			return soc15_asic_baco_reset(adev);
 		case AMD_RESET_METHOD_MODE2:
 			return amdgpu_dpm_mode2_reset(adev);
 		default:
-			if (!adev->in_suspend)
-				amdgpu_inc_vram_lost(adev);
 			return soc15_asic_mode1_reset(adev);
 	}
 }
diff --git a/drivers/gpu/drm/amd/amdgpu/vi.c b/drivers/gpu/drm/amd/amdgpu/vi.c
index 78b35901643b..3ce10e05d0d6 100644
--- a/drivers/gpu/drm/amd/amdgpu/vi.c
+++ b/drivers/gpu/drm/amd/amdgpu/vi.c
@@ -765,8 +765,6 @@ static int vi_asic_reset(struct amdgpu_device *adev)
 	int r;
 
 	if (vi_asic_reset_method(adev) == AMD_RESET_METHOD_BACO) {
-		if (!adev->in_suspend)
-			amdgpu_inc_vram_lost(adev);
 		r = amdgpu_dpm_baco_reset(adev);
 	} else {
 		r = vi_asic_pci_config_reset(adev);