Merge branch 'drm-next' of git://people.freedesktop.org/~airlied/linux

Pull drm updates from Dave Airlie: "This is the main git pull for the drm, I pretty much froze major pulls at -rc5/6 time, and haven't had much fallout, so will probably continue doing that. Lots of changes all over, big internal header cleanup to make it clear drm features are legacy things and what are things that modern KMS drivers should be using. Also big move to use the new generic fences in all the TTM drivers. core: atomic prep work, vblank rework changes, allows immediate vblank disables major header reworking and cleanups to better delinate legacy interfaces from what KMS drivers should be using. cursor planes locking fixes ttm: move to generic fences (affects all TTM drivers) ppc64 caching fixes radeon: userptr support, uvd for old asics, reset rework for fence changes better buffer placement changes, dpm feature enablement hdmi audio support fixes intel: Cherryview work, 180 degree rotation, skylake prep work, execlist command submission full ppgtt prep work cursor improvements edid caching, vdd handling improvements nouveau: fence reworking kepler memory clock work gt21x clock work fan control improvements hdmi infoframe fixes DP audio ast: ppc64 fixes caching fix rcar: rcar-du DT support ipuv3: prep work for capture support msm: LVDS support for mdp4, new panel, gpu refactoring exynos: exynos3250 SoC support, drop bad mmap interface, mipi dsi changes, and component match support" * 'drm-next' of git://people.freedesktop.org/~airlied/linux: (640 commits) drm/mst: rework payload table allocation to conform better. drm/ast: Fix HW cursor image drm/radeon/kv: add uvd/vce info to dpm debugfs output drm/radeon/ci: add uvd/vce info to dpm debugfs output drm/radeon: export reservation_object from dmabuf to ttm drm/radeon: cope with foreign fences inside the reservation object drm/radeon: cope with foreign fences inside display drm/core: use helper to check driver features drm/radeon/cik: write gfx ucode version to ucode addr reg drm/radeon/si: print full CS when we hit a packet 0 drm/radeon: remove unecessary includes drm/radeon/combios: declare legacy_connector_convert as static drm/radeon/atombios: declare connector convert tables as static drm/radeon: drop btc_get_max_clock_from_voltage_dependency_table drm/radeon/dpm: drop clk/voltage dependency filters for BTC drm/radeon/dpm: drop clk/voltage dependency filters for CI drm/radeon/dpm: drop clk/voltage dependency filters for SI drm/radeon/dpm: drop clk/voltage dependency filters for NI drm/radeon: disable audio when we disable hdmi (v2) drm/radeon: split audio enable between eg and r600 (v2) ...
author: Linus Torvalds <torvalds@linux-foundation.org> 2014-10-14 09:39:08 +0200
committer: Linus Torvalds <torvalds@linux-foundation.org> 2014-10-14 09:39:08 +0200
commit: 2d65a9f48fcdf7866aab6457bc707ca233e0c791 (patch)
tree: f93e5838d6ac2e59434367f4ff905f7d9c45fc2b /drivers/gpu
parent: Merge branch 'misc' of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild (diff)
parent: drm/mst: rework payload table allocation to conform better. (diff)
download: linux-dev-2d65a9f48fcdf7866aab6457bc707ca233e0c791.tar.xz
linux-dev-2d65a9f48fcdf7866aab6457bc707ca233e0c791.zip
479 files changed, 23511 insertions, 10099 deletions
diff --git a/drivers/gpu/drm/Kconfig b/drivers/gpu/drm/Kconfig
index b066bb3ca01a..e3b4b0f02b3d 100644
--- a/drivers/gpu/drm/Kconfig
+++ b/drivers/gpu/drm/Kconfig
@@ -8,6 +8,7 @@ menuconfig DRM
 	tristate "Direct Rendering Manager (XFree86 4.1.0 and higher DRI support)"
 	depends on (AGP || AGP=n) && !EMULATED_CMPXCHG && MMU && HAS_DMA
 	select HDMI
+	select FB_CMDLINE
 	select I2C
 	select I2C_ALGOBIT
 	select DMA_SHARED_BUFFER
@@ -24,12 +25,6 @@ config DRM_MIPI_DSI
 	bool
 	depends on DRM
 
-config DRM_USB
-	tristate
-	depends on DRM
-	depends on USB_SUPPORT && USB_ARCH_HAS_HCD
-	select USB
-
 config DRM_KMS_HELPER
 	tristate
 	depends on DRM
@@ -115,6 +110,7 @@ config DRM_RADEON
 	select HWMON
 	select BACKLIGHT_CLASS_DEVICE
 	select INTERVAL_TREE
+	select MMU_NOTIFIER
 	help
 	  Choose this option if you have an ATI Radeon graphics card.  There
 	  are both PCI and AGP versions.  You don't need to choose this to
diff --git a/drivers/gpu/drm/Makefile b/drivers/gpu/drm/Makefile
index 4a55d59ccd22..9292a761ea6d 100644
--- a/drivers/gpu/drm/Makefile
+++ b/drivers/gpu/drm/Makefile
@@ -4,7 +4,7 @@
 
 ccflags-y := -Iinclude/drm
 
-drm-y       :=	drm_auth.o drm_buffer.o drm_bufs.o drm_cache.o \
+drm-y       :=	drm_auth.o drm_bufs.o drm_cache.o \
 		drm_context.o drm_dma.o \
 		drm_fops.o drm_gem.o drm_ioctl.o drm_irq.o \
 		drm_lock.o drm_memory.o drm_drv.o drm_vm.o \
@@ -22,8 +22,6 @@ drm-$(CONFIG_PCI) += ati_pcigart.o
 drm-$(CONFIG_DRM_PANEL) += drm_panel.o
 drm-$(CONFIG_OF) += drm_of.o
 
-drm-usb-y   := drm_usb.o
-
 drm_kms_helper-y := drm_crtc_helper.o drm_dp_helper.o drm_probe_helper.o \
 		drm_plane_helper.o drm_dp_mst_topology.o
 drm_kms_helper-$(CONFIG_DRM_LOAD_EDID_FIRMWARE) += drm_edid_load.o
@@ -36,7 +34,6 @@ CFLAGS_drm_trace_points.o := -I$(src)
 
 obj-$(CONFIG_DRM)	+= drm.o
 obj-$(CONFIG_DRM_MIPI_DSI) += drm_mipi_dsi.o
-obj-$(CONFIG_DRM_USB)   += drm_usb.o
 obj-$(CONFIG_DRM_TTM)	+= ttm/
 obj-$(CONFIG_DRM_TDFX)	+= tdfx/
 obj-$(CONFIG_DRM_R128)	+= r128/
diff --git a/drivers/gpu/drm/armada/armada_drv.c b/drivers/gpu/drm/armada/armada_drv.c
index e2d5792b140f..f672e6ad8afa 100644
--- a/drivers/gpu/drm/armada/armada_drv.c
+++ b/drivers/gpu/drm/armada/armada_drv.c
@@ -308,6 +308,7 @@ static struct drm_driver armada_drm_driver = {
 	.postclose		= NULL,
 	.lastclose		= armada_drm_lastclose,
 	.unload			= armada_drm_unload,
+	.set_busid		= drm_platform_set_busid,
 	.get_vblank_counter	= drm_vblank_count,
 	.enable_vblank		= armada_drm_enable_vblank,
 	.disable_vblank		= armada_drm_disable_vblank,
diff --git a/drivers/gpu/drm/armada/armada_gem.h b/drivers/gpu/drm/armada/armada_gem.h
index 00b6cd461a03..b000ea3a829a 100644
--- a/drivers/gpu/drm/armada/armada_gem.h
+++ b/drivers/gpu/drm/armada/armada_gem.h
@@ -8,6 +8,8 @@
 #ifndef ARMADA_GEM_H
 #define ARMADA_GEM_H
 
+#include <drm/drm_gem.h>
+
 /* GEM */
 struct armada_gem_object {
 	struct drm_gem_object	obj;
diff --git a/drivers/gpu/drm/ast/ast_dp501.c b/drivers/gpu/drm/ast/ast_dp501.c
index 5da4b62285fa..76f07f38b941 100644
--- a/drivers/gpu/drm/ast/ast_dp501.c
+++ b/drivers/gpu/drm/ast/ast_dp501.c
@@ -379,11 +379,39 @@ static bool ast_init_dvo(struct drm_device *dev)
 	return true;
 }
 
+
+static void ast_init_analog(struct drm_device *dev)
+{
+	struct ast_private *ast = dev->dev_private;
+	u32 data;
+
+	/*
+	 * Set DAC source to VGA mode in SCU2C via the P2A
+	 * bridge. First configure the P2U to target the SCU
+	 * in case it isn't at this stage.
+	 */
+	ast_write32(ast, 0xf004, 0x1e6e0000);
+	ast_write32(ast, 0xf000, 0x1);
+
+	/* Then unlock the SCU with the magic password */
+	ast_write32(ast, 0x12000, 0x1688a8a8);
+	ast_write32(ast, 0x12000, 0x1688a8a8);
+	ast_write32(ast, 0x12000, 0x1688a8a8);
+
+	/* Finally, clear bits [17:16] of SCU2c */
+	data = ast_read32(ast, 0x1202c);
+	data &= 0xfffcffff;
+	ast_write32(ast, 0, data);
+
+	/* Disable DVO */
+	ast_set_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xa3, 0xcf, 0x00);
+}
+
 void ast_init_3rdtx(struct drm_device *dev)
 {
 	struct ast_private *ast = dev->dev_private;
 	u8 jreg;
-	u32 data;
+
 	if (ast->chip == AST2300 || ast->chip == AST2400) {
 		jreg = ast_get_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xd1, 0xff);
 		switch (jreg & 0x0e) {
@@ -399,12 +427,8 @@ void ast_init_3rdtx(struct drm_device *dev)
 		default:
 			if (ast->tx_chip_type == AST_TX_SIL164)
 				ast_init_dvo(dev);
-			else {
-				ast_write32(ast, 0x12000, 0x1688a8a8);
-				data = ast_read32(ast, 0x1202c);
-				data &= 0xfffcffff;
-				ast_write32(ast, 0, data);
-			}
+			else
+				ast_init_analog(dev);
 		}
 	}
 }
diff --git a/drivers/gpu/drm/ast/ast_drv.c b/drivers/gpu/drm/ast/ast_drv.c
index f19682a93c24..9a32d9dfdd26 100644
--- a/drivers/gpu/drm/ast/ast_drv.c
+++ b/drivers/gpu/drm/ast/ast_drv.c
@@ -199,6 +199,7 @@ static struct drm_driver driver = {
 
 	.load = ast_driver_load,
 	.unload = ast_driver_unload,
+	.set_busid = drm_pci_set_busid,
 
 	.fops = &ast_fops,
 	.name = DRIVER_NAME,
diff --git a/drivers/gpu/drm/ast/ast_drv.h b/drivers/gpu/drm/ast/ast_drv.h
index 957d4fabf1e1..86205a28e56b 100644
--- a/drivers/gpu/drm/ast/ast_drv.h
+++ b/drivers/gpu/drm/ast/ast_drv.h
@@ -36,6 +36,8 @@
 #include <drm/ttm/ttm_memory.h>
 #include <drm/ttm/ttm_module.h>
 
+#include <drm/drm_gem.h>
+
 #include <linux/i2c.h>
 #include <linux/i2c-algo-bit.h>
 
@@ -125,8 +127,9 @@ struct ast_gem_object;
 
 #define AST_IO_AR_PORT_WRITE		(0x40)
 #define AST_IO_MISC_PORT_WRITE		(0x42)
+#define AST_IO_VGA_ENABLE_PORT		(0x43)
 #define AST_IO_SEQ_PORT			(0x44)
-#define AST_DAC_INDEX_READ		(0x3c7)
+#define AST_IO_DAC_INDEX_READ		(0x47)
 #define AST_IO_DAC_INDEX_WRITE		(0x48)
 #define AST_IO_DAC_DATA		        (0x49)
 #define AST_IO_GR_PORT			(0x4E)
@@ -134,6 +137,8 @@ struct ast_gem_object;
 #define AST_IO_INPUT_STATUS1_READ	(0x5A)
 #define AST_IO_MISC_PORT_READ		(0x4C)
 
+#define AST_IO_MM_OFFSET		(0x380)
+
 #define __ast_read(x) \
 static inline u##x ast_read##x(struct ast_private *ast, u32 reg) { \
 u##x val = 0;\
@@ -316,7 +321,7 @@ struct ast_bo {
 	struct ttm_placement placement;
 	struct ttm_bo_kmap_obj kmap;
 	struct drm_gem_object gem;
-	u32 placements[3];
+	struct ttm_place placements[3];
 	int pin_count;
 };
 #define gem_to_ast_bo(gobj) container_of((gobj), struct ast_bo, gem)
@@ -381,6 +386,9 @@ int ast_bo_push_sysram(struct ast_bo *bo);
 int ast_mmap(struct file *filp, struct vm_area_struct *vma);
 
 /* ast post */
+void ast_enable_vga(struct drm_device *dev);
+void ast_enable_mmio(struct drm_device *dev);
+bool ast_is_vga_enabled(struct drm_device *dev);
 void ast_post_gpu(struct drm_device *dev);
 u32 ast_mindwm(struct ast_private *ast, u32 r);
 void ast_moutdwm(struct ast_private *ast, u32 r, u32 v);
diff --git a/drivers/gpu/drm/ast/ast_fb.c b/drivers/gpu/drm/ast/ast_fb.c
index cba45c774552..5c60ae524c45 100644
--- a/drivers/gpu/drm/ast/ast_fb.c
+++ b/drivers/gpu/drm/ast/ast_fb.c
@@ -186,7 +186,8 @@ static int astfb_create_object(struct ast_fbdev *afbdev,
 static int astfb_create(struct drm_fb_helper *helper,
 			struct drm_fb_helper_surface_size *sizes)
 {
-	struct ast_fbdev *afbdev = (struct ast_fbdev *)helper;
+	struct ast_fbdev *afbdev =
+		container_of(helper, struct ast_fbdev, helper);
 	struct drm_device *dev = afbdev->helper.dev;
 	struct drm_mode_fb_cmd2 mode_cmd;
 	struct drm_framebuffer *fb;
diff --git a/drivers/gpu/drm/ast/ast_main.c b/drivers/gpu/drm/ast/ast_main.c
index b792194e0d9c..035dacc93382 100644
--- a/drivers/gpu/drm/ast/ast_main.c
+++ b/drivers/gpu/drm/ast/ast_main.c
@@ -63,7 +63,7 @@ uint8_t ast_get_index_reg_mask(struct ast_private *ast,
 }
 
 
-static int ast_detect_chip(struct drm_device *dev)
+static int ast_detect_chip(struct drm_device *dev, bool *need_post)
 {
 	struct ast_private *ast = dev->dev_private;
 	uint32_t data, jreg;
@@ -110,6 +110,21 @@ static int ast_detect_chip(struct drm_device *dev)
 		}
 	}
 
+	/*
+	 * If VGA isn't enabled, we need to enable now or subsequent
+	 * access to the scratch registers will fail. We also inform
+	 * our caller that it needs to POST the chip
+	 * (Assumption: VGA not enabled -> need to POST)
+	 */
+	if (!ast_is_vga_enabled(dev)) {
+		ast_enable_vga(dev);
+		ast_enable_mmio(dev);
+		DRM_INFO("VGA not enabled on entry, requesting chip POST\n");
+		*need_post = true;
+	} else
+		*need_post = false;
+
+	/* Check if we support wide screen */
 	switch (ast->chip) {
 	case AST1180:
 		ast->support_wide_screen = true;
@@ -125,6 +140,7 @@ static int ast_detect_chip(struct drm_device *dev)
 			ast->support_wide_screen = true;
 		else {
 			ast->support_wide_screen = false;
+			/* Read SCU7c (silicon revision register) */
 			ast_write32(ast, 0xf004, 0x1e6e0000);
 			ast_write32(ast, 0xf000, 0x1);
 			data = ast_read32(ast, 0x1207c);
@@ -137,11 +153,29 @@ static int ast_detect_chip(struct drm_device *dev)
 		break;
 	}
 
+	/* Check 3rd Tx option (digital output afaik) */
 	ast->tx_chip_type = AST_TX_NONE;
-	jreg = ast_get_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xa3, 0xff);
-	if (jreg & 0x80)
-		ast->tx_chip_type = AST_TX_SIL164;
+
+	/*
+	 * VGACRA3 Enhanced Color Mode Register, check if DVO is already
+	 * enabled, in that case, assume we have a SIL164 TMDS transmitter
+	 *
+	 * Don't make that assumption if we the chip wasn't enabled and
+	 * is at power-on reset, otherwise we'll incorrectly "detect" a
+	 * SIL164 when there is none.
+	 */
+	if (!*need_post) {
+		jreg = ast_get_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xa3, 0xff);
+		if (jreg & 0x80)
+			ast->tx_chip_type = AST_TX_SIL164;
+	}
+
 	if ((ast->chip == AST2300) || (ast->chip == AST2400)) {
+		/*
+		 * On AST2300 and 2400, look the configuration set by the SoC in
+		 * the SOC scratch register #1 bits 11:8 (interestingly marked
+		 * as "reserved" in the spec)
+		 */
 		jreg = ast_get_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xd1, 0xff);
 		switch (jreg) {
 		case 0x04:
@@ -162,6 +196,17 @@ static int ast_detect_chip(struct drm_device *dev)
 		}
 	}
 
+	/* Print stuff for diagnostic purposes */
+	switch(ast->tx_chip_type) {
+	case AST_TX_SIL164:
+		DRM_INFO("Using Sil164 TMDS transmitter\n");
+		break;
+	case AST_TX_DP501:
+		DRM_INFO("Using DP501 DisplayPort transmitter\n");
+		break;
+	default:
+		DRM_INFO("Analog VGA only\n");
+	}
 	return 0;
 }
 
@@ -346,6 +391,7 @@ static u32 ast_get_vram_info(struct drm_device *dev)
 int ast_driver_load(struct drm_device *dev, unsigned long flags)
 {
 	struct ast_private *ast;
+	bool need_post;
 	int ret = 0;
 
 	ast = kzalloc(sizeof(struct ast_private), GFP_KERNEL);
@@ -360,13 +406,27 @@ int ast_driver_load(struct drm_device *dev, unsigned long flags)
 		ret = -EIO;
 		goto out_free;
 	}
-	ast->ioregs = pci_iomap(dev->pdev, 2, 0);
+
+	/*
+	 * If we don't have IO space at all, use MMIO now and
+	 * assume the chip has MMIO enabled by default (rev 0x20
+	 * and higher).
+	 */
+	if (!(pci_resource_flags(dev->pdev, 2) & IORESOURCE_IO)) {
+		DRM_INFO("platform has no IO space, trying MMIO\n");
+		ast->ioregs = ast->regs + AST_IO_MM_OFFSET;
+	}
+
+	/* "map" IO regs if the above hasn't done so already */
 	if (!ast->ioregs) {
-		ret = -EIO;
-		goto out_free;
+		ast->ioregs = pci_iomap(dev->pdev, 2, 0);
+		if (!ast->ioregs) {
+			ret = -EIO;
+			goto out_free;
+		}
 	}
 
-	ast_detect_chip(dev);
+	ast_detect_chip(dev, &need_post);
 
 	if (ast->chip != AST1180) {
 		ast_get_dram_info(dev);
@@ -374,6 +434,9 @@ int ast_driver_load(struct drm_device *dev, unsigned long flags)
 		DRM_INFO("dram %d %d %d %08x\n", ast->mclk, ast->dram_type, ast->dram_bus_width, ast->vram_size);
 	}
 
+	if (need_post)
+		ast_post_gpu(dev);
+
 	ret = ast_mm_init(ast);
 	if (ret)
 		goto out_free;
diff --git a/drivers/gpu/drm/ast/ast_mode.c b/drivers/gpu/drm/ast/ast_mode.c
index 5389350244f2..9dc0fd5c1ea4 100644
--- a/drivers/gpu/drm/ast/ast_mode.c
+++ b/drivers/gpu/drm/ast/ast_mode.c
@@ -80,6 +80,8 @@ static bool ast_get_vbios_mode_info(struct drm_crtc *crtc, struct drm_display_mo
 	struct ast_private *ast = crtc->dev->dev_private;
 	u32 refresh_rate_index = 0, mode_id, color_index, refresh_rate;
 	u32 hborder, vborder;
+	bool check_sync;
+	struct ast_vbios_enhtable *best = NULL;
 
 	switch (crtc->primary->fb->bits_per_pixel) {
 	case 8:
@@ -141,14 +143,34 @@ static bool ast_get_vbios_mode_info(struct drm_crtc *crtc, struct drm_display_mo
 	}
 
 	refresh_rate = drm_mode_vrefresh(mode);
-	while (vbios_mode->enh_table->refresh_rate < refresh_rate) {
-		vbios_mode->enh_table++;
-		if ((vbios_mode->enh_table->refresh_rate > refresh_rate) ||
-		    (vbios_mode->enh_table->refresh_rate == 0xff)) {
-			vbios_mode->enh_table--;
-			break;
+	check_sync = vbios_mode->enh_table->flags & WideScreenMode;
+	do {
+		struct ast_vbios_enhtable *loop = vbios_mode->enh_table;
+
+		while (loop->refresh_rate != 0xff) {
+			if ((check_sync) &&
+			    (((mode->flags & DRM_MODE_FLAG_NVSYNC)  &&
+			      (loop->flags & PVSync))  ||
+			     ((mode->flags & DRM_MODE_FLAG_PVSYNC)  &&
+			      (loop->flags & NVSync))  ||
+			     ((mode->flags & DRM_MODE_FLAG_NHSYNC)  &&
+			      (loop->flags & PHSync))  ||
+			     ((mode->flags & DRM_MODE_FLAG_PHSYNC)  &&
+			      (loop->flags & NHSync)))) {
+				loop++;
+				continue;
+			}
+			if (loop->refresh_rate <= refresh_rate
+			    && (!best || loop->refresh_rate > best->refresh_rate))
+				best = loop;
+			loop++;
 		}
-	}
+		if (best || !check_sync)
+			break;
+		check_sync = 0;
+	} while (1);
+	if (best)
+		vbios_mode->enh_table = best;
 
 	hborder = (vbios_mode->enh_table->flags & HBorder) ? 8 : 0;
 	vborder = (vbios_mode->enh_table->flags & VBorder) ? 8 : 0;
@@ -419,8 +441,10 @@ static void ast_set_sync_reg(struct drm_device *dev, struct drm_display_mode *mo
 	struct ast_private *ast = dev->dev_private;
 	u8 jreg;
 
-	jreg = ast_io_read8(ast, AST_IO_MISC_PORT_READ);
-	jreg |= (vbios_mode->enh_table->flags & SyncNN);
+	jreg  = ast_io_read8(ast, AST_IO_MISC_PORT_READ);
+	jreg &= ~0xC0;
+	if (vbios_mode->enh_table->flags & NVSync) jreg |= 0x80;
+	if (vbios_mode->enh_table->flags & NHSync) jreg |= 0x40;
 	ast_io_write8(ast, AST_IO_MISC_PORT_WRITE, jreg);
 }
 
@@ -1080,8 +1104,8 @@ static u32 copy_cursor_image(u8 *src, u8 *dst, int width, int height)
 			srcdata32[1].ul = *((u32 *)(srcxor + 4)) & 0xf0f0f0f0;
 			data32.b[0] = srcdata32[0].b[1] | (srcdata32[0].b[0] >> 4);
 			data32.b[1] = srcdata32[0].b[3] | (srcdata32[0].b[2] >> 4);
-			data32.b[2] = srcdata32[0].b[1] | (srcdata32[1].b[0] >> 4);
-			data32.b[3] = srcdata32[0].b[3] | (srcdata32[1].b[2] >> 4);
+			data32.b[2] = srcdata32[1].b[1] | (srcdata32[1].b[0] >> 4);
+			data32.b[3] = srcdata32[1].b[3] | (srcdata32[1].b[2] >> 4);
 
 			writel(data32.ul, dstxor);
 			csum += data32.ul;
diff --git a/drivers/gpu/drm/ast/ast_post.c b/drivers/gpu/drm/ast/ast_post.c
index 38d437f3a267..810c51d92b99 100644
--- a/drivers/gpu/drm/ast/ast_post.c
+++ b/drivers/gpu/drm/ast/ast_post.c
@@ -33,18 +33,23 @@
 
 static void ast_init_dram_2300(struct drm_device *dev);
 
-static void
-ast_enable_vga(struct drm_device *dev)
+void ast_enable_vga(struct drm_device *dev)
+{
+	struct ast_private *ast = dev->dev_private;
+
+	ast_io_write8(ast, AST_IO_VGA_ENABLE_PORT, 0x01);
+	ast_io_write8(ast, AST_IO_MISC_PORT_WRITE, 0x01);
+}
+
+void ast_enable_mmio(struct drm_device *dev)
 {
 	struct ast_private *ast = dev->dev_private;
 
-	ast_io_write8(ast, 0x43, 0x01);
-	ast_io_write8(ast, 0x42, 0x01);
+	ast_set_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xa1, 0xff, 0x04);
 }
 
-#if 0 /* will use later */
-static bool
-ast_is_vga_enabled(struct drm_device *dev)
+
+bool ast_is_vga_enabled(struct drm_device *dev)
 {
 	struct ast_private *ast = dev->dev_private;
 	u8 ch;
@@ -52,7 +57,7 @@ ast_is_vga_enabled(struct drm_device *dev)
 	if (ast->chip == AST1180) {
 		/* TODO 1180 */
 	} else {
-		ch = ast_io_read8(ast, 0x43);
+		ch = ast_io_read8(ast, AST_IO_VGA_ENABLE_PORT);
 		if (ch) {
 			ast_open_key(ast);
 			ch = ast_get_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xb6, 0xff);
@@ -61,7 +66,6 @@ ast_is_vga_enabled(struct drm_device *dev)
 	}
 	return 0;
 }
-#endif
 
 static const u8 extreginfo[] = { 0x0f, 0x04, 0x1c, 0xff };
 static const u8 extreginfo_ast2300a0[] = { 0x0f, 0x04, 0x1c, 0xff };
@@ -371,6 +375,7 @@ void ast_post_gpu(struct drm_device *dev)
 	pci_write_config_dword(ast->dev->pdev, 0x04, reg);
 
 	ast_enable_vga(dev);
+	ast_enable_mmio(dev);
 	ast_open_key(ast);
 	ast_set_def_ext_reg(dev);
 
diff --git a/drivers/gpu/drm/ast/ast_tables.h b/drivers/gpu/drm/ast/ast_tables.h
index 05c01ea85294..3608d5aa7451 100644
--- a/drivers/gpu/drm/ast/ast_tables.h
+++ b/drivers/gpu/drm/ast/ast_tables.h
@@ -35,14 +35,18 @@
 #define HalfDCLK                0x00000002
 #define DoubleScanMode          0x00000004
 #define LineCompareOff          0x00000008
-#define SyncPP                  0x00000000
-#define SyncPN                  0x00000040
-#define SyncNP                  0x00000080
-#define SyncNN                  0x000000C0
 #define HBorder                 0x00000020
 #define VBorder                 0x00000010
 #define WideScreenMode		0x00000100
 #define NewModeInfo		0x00000200
+#define NHSync			0x00000400
+#define PHSync			0x00000800
+#define NVSync			0x00001000
+#define PVSync			0x00002000
+#define SyncPP			(PVSync | PHSync)
+#define SyncPN			(PVSync | NHSync)
+#define SyncNP			(NVSync | PHSync)
+#define SyncNN			(NVSync | NHSync)
 
 /* DCLK Index */
 #define VCLK25_175     		0x00
@@ -72,6 +76,7 @@
 #define VCLK119     		0x17
 #define VCLK85_5     		0x18
 #define VCLK97_75     		0x19
+#define VCLK118_25			0x1A
 
 static struct ast_vbios_dclk_info dclk_table[] = {
 	{0x2C, 0xE7, 0x03},					/* 00: VCLK25_175	*/
@@ -100,6 +105,7 @@ static struct ast_vbios_dclk_info dclk_table[] = {
 	{0x77, 0x58, 0x80},					/* 17: VCLK119      */
 	{0x32, 0x67, 0x80},				    /* 18: VCLK85_5     */
 	{0x6a, 0x6d, 0x80},					/* 19: VCLK97_75	*/
+	{0x3b, 0x2c, 0x81},					/* 1A: VCLK118_25	*/
 };
 
 static struct ast_vbios_stdtable vbios_stdtable[] = {
@@ -246,8 +252,10 @@ static struct ast_vbios_enhtable res_1360x768[] = {
 static struct ast_vbios_enhtable res_1600x900[] = {
 	{1760, 1600, 48, 32, 926,  900, 3, 5, VCLK97_75,	/* 60Hz CVT RB */
 	 (SyncNP | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 60, 1, 0x3A },
-	{1760, 1600, 48, 32, 926,  900, 3, 5, VCLK97_75,	/* end */
-	 (SyncNP | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 0xFF, 1, 0x3A }
+	{2112, 1600, 88,168, 934,  900, 3, 5, VCLK118_25,	/* 60Hz CVT */
+	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 60, 2, 0x3A },
+	{2112, 1600, 88,168, 934,  900, 3, 5, VCLK118_25,	/* 60Hz CVT */
+	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 0xFF, 2, 0x3A },
 };
 
 static struct ast_vbios_enhtable res_1920x1080[] = {
@@ -261,11 +269,11 @@ static struct ast_vbios_enhtable res_1920x1080[] = {
 /* 16:10 */
 static struct ast_vbios_enhtable res_1280x800[] = {
 	{1440, 1280, 48, 32,  823,  800, 3, 6, VCLK71,	/* 60Hz RB */
-	 (SyncNP | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 60, 1, 35 },
+	 (SyncNP | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 60, 1, 0x35 },
 	{1680, 1280, 72,128,  831,  800, 3, 6, VCLK83_5,	/* 60Hz */
-	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 60, 1, 0x35 },
+	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 60, 2, 0x35 },
 	{1680, 1280, 72,128,  831,  800, 3, 6, VCLK83_5,	/* 60Hz */
-	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 0xFF, 1, 0x35 },
+	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 0xFF, 2, 0x35 },
 
 };
 
@@ -273,24 +281,24 @@ static struct ast_vbios_enhtable res_1440x900[] = {
 	{1600, 1440, 48, 32,  926,  900, 3, 6, VCLK88_75,	/* 60Hz RB */
 	 (SyncNP | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 60, 1, 0x36 },
 	{1904, 1440, 80,152,  934,  900, 3, 6, VCLK106_5,	/* 60Hz */
-	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 60, 1, 0x36 },
+	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 60, 2, 0x36 },
 	{1904, 1440, 80,152,  934,  900, 3, 6, VCLK106_5,	/* 60Hz */
-	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 0xFF, 1, 0x36 },
+	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 0xFF, 2, 0x36 },
 };
 
 static struct ast_vbios_enhtable res_1680x1050[] = {
 	{1840, 1680, 48, 32, 1080, 1050, 3, 6, VCLK119,	/* 60Hz RB */
 	 (SyncNP | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 60, 1, 0x37 },
 	{2240, 1680,104,176, 1089, 1050, 3, 6, VCLK146_25,	/* 60Hz */
-	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 60, 1, 0x37 },
+	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 60, 2, 0x37 },
 	{2240, 1680,104,176, 1089, 1050, 3, 6, VCLK146_25,	/* 60Hz */
-	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 0xFF, 1, 0x37 },
+	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 0xFF, 2, 0x37 },
 };
 
 static struct ast_vbios_enhtable res_1920x1200[] = {
-	{2080, 1920, 48, 32, 1235, 1200, 3, 6, VCLK154,	/* 60Hz */
+	{2080, 1920, 48, 32, 1235, 1200, 3, 6, VCLK154,	/* 60Hz RB*/
 	 (SyncNP | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 60, 1, 0x34 },
-	{2080, 1920, 48, 32, 1235, 1200, 3, 6, VCLK154,	/* 60Hz */
+	{2080, 1920, 48, 32, 1235, 1200, 3, 6, VCLK154,	/* 60Hz RB */
 	 (SyncNP | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 0xFF, 1, 0x34 },
 };
 
diff --git a/drivers/gpu/drm/ast/ast_ttm.c b/drivers/gpu/drm/ast/ast_ttm.c
index b8246227bab0..08f82eae6939 100644
--- a/drivers/gpu/drm/ast/ast_ttm.c
+++ b/drivers/gpu/drm/ast/ast_ttm.c
@@ -293,18 +293,22 @@ void ast_mm_fini(struct ast_private *ast)
 void ast_ttm_placement(struct ast_bo *bo, int domain)
 {
 	u32 c = 0;
-	bo->placement.fpfn = 0;
-	bo->placement.lpfn = 0;
+	unsigned i;
+
 	bo->placement.placement = bo->placements;
 	bo->placement.busy_placement = bo->placements;
 	if (domain & TTM_PL_FLAG_VRAM)
-		bo->placements[c++] = TTM_PL_FLAG_WC | TTM_PL_FLAG_UNCACHED | TTM_PL_FLAG_VRAM;
+		bo->placements[c++].flags = TTM_PL_FLAG_WC | TTM_PL_FLAG_UNCACHED | TTM_PL_FLAG_VRAM;
 	if (domain & TTM_PL_FLAG_SYSTEM)
-		bo->placements[c++] = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
+		bo->placements[c++].flags = TTM_PL_FLAG_CACHED | TTM_PL_FLAG_SYSTEM;
 	if (!c)
-		bo->placements[c++] = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
+		bo->placements[c++].flags = TTM_PL_FLAG_CACHED | TTM_PL_FLAG_SYSTEM;
 	bo->placement.num_placement = c;
 	bo->placement.num_busy_placement = c;
+	for (i = 0; i < c; ++i) {
+		bo->placements[i].fpfn = 0;
+		bo->placements[i].lpfn = 0;
+	}
 }
 
 int ast_bo_create(struct drm_device *dev, int size, int align,
@@ -335,7 +339,7 @@ int ast_bo_create(struct drm_device *dev, int size, int align,
 	ret = ttm_bo_init(&ast->ttm.bdev, &astbo->bo, size,
 			  ttm_bo_type_device, &astbo->placement,
 			  align >> PAGE_SHIFT, false, NULL, acc_size,
-			  NULL, ast_bo_ttm_destroy);
+			  NULL, NULL, ast_bo_ttm_destroy);
 	if (ret)
 		return ret;
 
@@ -360,7 +364,7 @@ int ast_bo_pin(struct ast_bo *bo, u32 pl_flag, u64 *gpu_addr)
 
 	ast_ttm_placement(bo, pl_flag);
 	for (i = 0; i < bo->placement.num_placement; i++)
-		bo->placements[i] |= TTM_PL_FLAG_NO_EVICT;
+		bo->placements[i].flags |= TTM_PL_FLAG_NO_EVICT;
 	ret = ttm_bo_validate(&bo->bo, &bo->placement, false, false);
 	if (ret)
 		return ret;
@@ -383,7 +387,7 @@ int ast_bo_unpin(struct ast_bo *bo)
 		return 0;
 
 	for (i = 0; i < bo->placement.num_placement ; i++)
-		bo->placements[i] &= ~TTM_PL_FLAG_NO_EVICT;
+		bo->placements[i].flags &= ~TTM_PL_FLAG_NO_EVICT;
 	ret = ttm_bo_validate(&bo->bo, &bo->placement, false, false);
 	if (ret)
 		return ret;
@@ -407,7 +411,7 @@ int ast_bo_push_sysram(struct ast_bo *bo)
 
 	ast_ttm_placement(bo, TTM_PL_FLAG_SYSTEM);
 	for (i = 0; i < bo->placement.num_placement ; i++)
-		bo->placements[i] |= TTM_PL_FLAG_NO_EVICT;
+		bo->placements[i].flags |= TTM_PL_FLAG_NO_EVICT;
 
 	ret = ttm_bo_validate(&bo->bo, &bo->placement, false, false);
 	if (ret) {
@@ -423,7 +427,7 @@ int ast_mmap(struct file *filp, struct vm_area_struct *vma)
 	struct ast_private *ast;
 
 	if (unlikely(vma->vm_pgoff < DRM_FILE_PAGE_OFFSET))
-		return drm_mmap(filp, vma);
+		return -EINVAL;
 
 	file_priv = filp->private_data;
 	ast = file_priv->minor->dev->dev_private;
diff --git a/drivers/gpu/drm/ati_pcigart.c b/drivers/gpu/drm/ati_pcigart.c
index c399dea27a3b..6c4d4b6eba80 100644
--- a/drivers/gpu/drm/ati_pcigart.c
+++ b/drivers/gpu/drm/ati_pcigart.c
@@ -34,6 +34,8 @@
 #include <linux/export.h>
 #include <drm/drmP.h>
 
+#include <drm/ati_pcigart.h>
+
 # define ATI_PCIGART_PAGE_SIZE		4096	/**< PCI GART page size */
 
 static int drm_ati_alloc_pcigart_table(struct drm_device *dev,
diff --git a/drivers/gpu/drm/bochs/bochs.h b/drivers/gpu/drm/bochs/bochs.h
index 7eb52dd44b01..71f2687fc3cc 100644
--- a/drivers/gpu/drm/bochs/bochs.h
+++ b/drivers/gpu/drm/bochs/bochs.h
@@ -7,6 +7,8 @@
 #include <drm/drm_crtc_helper.h>
 #include <drm/drm_fb_helper.h>
 
+#include <drm/drm_gem.h>
+
 #include <ttm/ttm_bo_driver.h>
 #include <ttm/ttm_page_alloc.h>
 
@@ -99,7 +101,7 @@ struct bochs_bo {
 	struct ttm_placement placement;
 	struct ttm_bo_kmap_obj kmap;
 	struct drm_gem_object gem;
-	u32 placements[3];
+	struct ttm_place placements[3];
 	int pin_count;
 };
 
diff --git a/drivers/gpu/drm/bochs/bochs_drv.c b/drivers/gpu/drm/bochs/bochs_drv.c
index 9738e9b14708..98837bde2d25 100644
--- a/drivers/gpu/drm/bochs/bochs_drv.c
+++ b/drivers/gpu/drm/bochs/bochs_drv.c
@@ -82,6 +82,7 @@ static struct drm_driver bochs_driver = {
 	.driver_features	= DRIVER_GEM | DRIVER_MODESET,
 	.load			= bochs_load,
 	.unload			= bochs_unload,
+	.set_busid		= drm_pci_set_busid,
 	.fops			= &bochs_fops,
 	.name			= "bochs-drm",
 	.desc			= "bochs dispi vga interface (qemu stdvga)",
diff --git a/drivers/gpu/drm/bochs/bochs_mm.c b/drivers/gpu/drm/bochs/bochs_mm.c
index 1728a1b0b813..66286ff518d4 100644
--- a/drivers/gpu/drm/bochs/bochs_mm.c
+++ b/drivers/gpu/drm/bochs/bochs_mm.c
@@ -257,20 +257,26 @@ void bochs_mm_fini(struct bochs_device *bochs)
 
 static void bochs_ttm_placement(struct bochs_bo *bo, int domain)
 {
+	unsigned i;
 	u32 c = 0;
-	bo->placement.fpfn = 0;
-	bo->placement.lpfn = 0;
 	bo->placement.placement = bo->placements;
 	bo->placement.busy_placement = bo->placements;
 	if (domain & TTM_PL_FLAG_VRAM) {
-		bo->placements[c++] = TTM_PL_FLAG_WC | TTM_PL_FLAG_UNCACHED
+		bo->placements[c++].flags = TTM_PL_FLAG_WC
+			| TTM_PL_FLAG_UNCACHED
 			| TTM_PL_FLAG_VRAM;
 	}
 	if (domain & TTM_PL_FLAG_SYSTEM) {
-		bo->placements[c++] = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
+		bo->placements[c++].flags = TTM_PL_MASK_CACHING
+			| TTM_PL_FLAG_SYSTEM;
 	}
 	if (!c) {
-		bo->placements[c++] = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
+		bo->placements[c++].flags = TTM_PL_MASK_CACHING
+			| TTM_PL_FLAG_SYSTEM;
+	}
+	for (i = 0; i < c; ++i) {
+		bo->placements[i].fpfn = 0;
+		bo->placements[i].lpfn = 0;
 	}
 	bo->placement.num_placement = c;
 	bo->placement.num_busy_placement = c;
@@ -294,7 +300,7 @@ int bochs_bo_pin(struct bochs_bo *bo, u32 pl_flag, u64 *gpu_addr)
 
 	bochs_ttm_placement(bo, pl_flag);
 	for (i = 0; i < bo->placement.num_placement; i++)
-		bo->placements[i] |= TTM_PL_FLAG_NO_EVICT;
+		bo->placements[i].flags |= TTM_PL_FLAG_NO_EVICT;
 	ret = ttm_bo_validate(&bo->bo, &bo->placement, false, false);
 	if (ret)
 		return ret;
@@ -319,7 +325,7 @@ int bochs_bo_unpin(struct bochs_bo *bo)
 		return 0;
 
 	for (i = 0; i < bo->placement.num_placement; i++)
-		bo->placements[i] &= ~TTM_PL_FLAG_NO_EVICT;
+		bo->placements[i].flags &= ~TTM_PL_FLAG_NO_EVICT;
 	ret = ttm_bo_validate(&bo->bo, &bo->placement, false, false);
 	if (ret)
 		return ret;
@@ -333,7 +339,7 @@ int bochs_mmap(struct file *filp, struct vm_area_struct *vma)
 	struct bochs_device *bochs;
 
 	if (unlikely(vma->vm_pgoff < DRM_FILE_PAGE_OFFSET))
-		return drm_mmap(filp, vma);
+		return -EINVAL;
 
 	file_priv = filp->private_data;
 	bochs = file_priv->minor->dev->dev_private;
@@ -371,7 +377,7 @@ static int bochs_bo_create(struct drm_device *dev, int size, int align,
 	ret = ttm_bo_init(&bochs->ttm.bdev, &bochsbo->bo, size,
 			  ttm_bo_type_device, &bochsbo->placement,
 			  align >> PAGE_SHIFT, false, NULL, acc_size,
-			  NULL, bochs_bo_ttm_destroy);
+			  NULL, NULL, bochs_bo_ttm_destroy);
 	if (ret)
 		return ret;
 
diff --git a/drivers/gpu/drm/cirrus/cirrus_drv.c b/drivers/gpu/drm/cirrus/cirrus_drv.c
index 919c73b94447..e705335101a5 100644
--- a/drivers/gpu/drm/cirrus/cirrus_drv.c
+++ b/drivers/gpu/drm/cirrus/cirrus_drv.c
@@ -128,6 +128,7 @@ static struct drm_driver driver = {
 	.driver_features = DRIVER_MODESET | DRIVER_GEM,
 	.load = cirrus_driver_load,
 	.unload = cirrus_driver_unload,
+	.set_busid = drm_pci_set_busid,
 	.fops = &cirrus_driver_fops,
 	.name = DRIVER_NAME,
 	.desc = DRIVER_DESC,
diff --git a/drivers/gpu/drm/cirrus/cirrus_drv.h b/drivers/gpu/drm/cirrus/cirrus_drv.h
index 401c890b6c6a..d44e69daa239 100644
--- a/drivers/gpu/drm/cirrus/cirrus_drv.h
+++ b/drivers/gpu/drm/cirrus/cirrus_drv.h
@@ -21,6 +21,8 @@
 #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"
@@ -163,7 +165,7 @@ struct cirrus_bo {
 	struct ttm_placement placement;
 	struct ttm_bo_kmap_obj kmap;
 	struct drm_gem_object gem;
-	u32 placements[3];
+	struct ttm_place placements[3];
 	int pin_count;
 };
 #define gem_to_cirrus_bo(gobj) container_of((gobj), struct cirrus_bo, gem)
diff --git a/drivers/gpu/drm/cirrus/cirrus_fbdev.c b/drivers/gpu/drm/cirrus/cirrus_fbdev.c
index 2a135f253e29..d231b1c317af 100644
--- a/drivers/gpu/drm/cirrus/cirrus_fbdev.c
+++ b/drivers/gpu/drm/cirrus/cirrus_fbdev.c
@@ -160,7 +160,8 @@ static int cirrusfb_create_object(struct cirrus_fbdev *afbdev,
 static int cirrusfb_create(struct drm_fb_helper *helper,
 			   struct drm_fb_helper_surface_size *sizes)
 {
-	struct cirrus_fbdev *gfbdev = (struct cirrus_fbdev *)helper;
+	struct cirrus_fbdev *gfbdev =
+		container_of(helper, struct cirrus_fbdev, helper);
 	struct drm_device *dev = gfbdev->helper.dev;
 	struct cirrus_device *cdev = gfbdev->helper.dev->dev_private;
 	struct fb_info *info;
diff --git a/drivers/gpu/drm/cirrus/cirrus_ttm.c b/drivers/gpu/drm/cirrus/cirrus_ttm.c
index 92e6b7786097..dfffd528517a 100644
--- a/drivers/gpu/drm/cirrus/cirrus_ttm.c
+++ b/drivers/gpu/drm/cirrus/cirrus_ttm.c
@@ -298,18 +298,21 @@ void cirrus_mm_fini(struct cirrus_device *cirrus)
 void cirrus_ttm_placement(struct cirrus_bo *bo, int domain)
 {
 	u32 c = 0;
-	bo->placement.fpfn = 0;
-	bo->placement.lpfn = 0;
+	unsigned i;
 	bo->placement.placement = bo->placements;
 	bo->placement.busy_placement = bo->placements;
 	if (domain & TTM_PL_FLAG_VRAM)
-		bo->placements[c++] = TTM_PL_FLAG_WC | TTM_PL_FLAG_UNCACHED | TTM_PL_FLAG_VRAM;
+		bo->placements[c++].flags = TTM_PL_FLAG_WC | TTM_PL_FLAG_UNCACHED | TTM_PL_FLAG_VRAM;
 	if (domain & TTM_PL_FLAG_SYSTEM)
-		bo->placements[c++] = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
+		bo->placements[c++].flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
 	if (!c)
-		bo->placements[c++] = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
+		bo->placements[c++].flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
 	bo->placement.num_placement = c;
 	bo->placement.num_busy_placement = c;
+	for (i = 0; i < c; ++i) {
+		bo->placements[i].fpfn = 0;
+		bo->placements[i].lpfn = 0;
+	}
 }
 
 int cirrus_bo_create(struct drm_device *dev, int size, int align,
@@ -340,7 +343,7 @@ int cirrus_bo_create(struct drm_device *dev, int size, int align,
 	ret = ttm_bo_init(&cirrus->ttm.bdev, &cirrusbo->bo, size,
 			  ttm_bo_type_device, &cirrusbo->placement,
 			  align >> PAGE_SHIFT, false, NULL, acc_size,
-			  NULL, cirrus_bo_ttm_destroy);
+			  NULL, NULL, cirrus_bo_ttm_destroy);
 	if (ret)
 		return ret;
 
@@ -365,7 +368,7 @@ int cirrus_bo_pin(struct cirrus_bo *bo, u32 pl_flag, u64 *gpu_addr)
 
 	cirrus_ttm_placement(bo, pl_flag);
 	for (i = 0; i < bo->placement.num_placement; i++)
-		bo->placements[i] |= TTM_PL_FLAG_NO_EVICT;
+		bo->placements[i].flags |= TTM_PL_FLAG_NO_EVICT;
 	ret = ttm_bo_validate(&bo->bo, &bo->placement, false, false);
 	if (ret)
 		return ret;
@@ -392,7 +395,7 @@ int cirrus_bo_push_sysram(struct cirrus_bo *bo)
 
 	cirrus_ttm_placement(bo, TTM_PL_FLAG_SYSTEM);
 	for (i = 0; i < bo->placement.num_placement ; i++)
-		bo->placements[i] |= TTM_PL_FLAG_NO_EVICT;
+		bo->placements[i].flags |= TTM_PL_FLAG_NO_EVICT;
 
 	ret = ttm_bo_validate(&bo->bo, &bo->placement, false, false);
 	if (ret) {
@@ -408,7 +411,7 @@ int cirrus_mmap(struct file *filp, struct vm_area_struct *vma)
 	struct cirrus_device *cirrus;
 
 	if (unlikely(vma->vm_pgoff < DRM_FILE_PAGE_OFFSET))
-		return drm_mmap(filp, vma);
+		return -EINVAL;
 
 	file_priv = filp->private_data;
 	cirrus = file_priv->minor->dev->dev_private;
diff --git a/drivers/gpu/drm/drm_agpsupport.c b/drivers/gpu/drm/drm_agpsupport.c
index dde205cef384..4b2b4aa5033b 100644
--- a/drivers/gpu/drm/drm_agpsupport.c
+++ b/drivers/gpu/drm/drm_agpsupport.c
@@ -34,6 +34,7 @@
 #include <drm/drmP.h>
 #include <linux/module.h>
 #include <linux/slab.h>
+#include "drm_legacy.h"
 
 #if __OS_HAS_AGP
 
diff --git a/drivers/gpu/drm/drm_auth.c b/drivers/gpu/drm/drm_auth.c
index 3cedae12b3c1..fc8e8aaa34fb 100644
--- a/drivers/gpu/drm/drm_auth.c
+++ b/drivers/gpu/drm/drm_auth.c
@@ -34,6 +34,13 @@
  */
 
 #include <drm/drmP.h>
+#include "drm_internal.h"
+
+struct drm_magic_entry {
+	struct list_head head;
+	struct drm_hash_item hash_item;
+	struct drm_file *priv;
+};
 
 /**
  * Find the file with the given magic number.
diff --git a/drivers/gpu/drm/drm_bufs.c b/drivers/gpu/drm/drm_bufs.c
index 61acb8f6756d..569064a00693 100644
--- a/drivers/gpu/drm/drm_bufs.c
+++ b/drivers/gpu/drm/drm_bufs.c
@@ -1,18 +1,13 @@
-/**
- * \file drm_bufs.c
- * Generic buffer template
- *
- * \author Rickard E. (Rik) Faith <faith@valinux.com>
- * \author Gareth Hughes <gareth@valinux.com>
- */
-
 /*
- * Created: Thu Nov 23 03:10:50 2000 by gareth@valinux.com
+ * Legacy: Generic DRM Buffer Management
  *
  * Copyright 1999, 2000 Precision Insight, Inc., Cedar Park, Texas.
  * Copyright 2000 VA Linux Systems, Inc., Sunnyvale, California.
  * All Rights Reserved.
  *
+ * Author: Rickard E. (Rik) Faith <faith@valinux.com>
+ * Author: Gareth Hughes <gareth@valinux.com>
+ *
  * 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
@@ -39,6 +34,7 @@
 #include <linux/export.h>
 #include <asm/shmparam.h>
 #include <drm/drmP.h>
+#include "drm_legacy.h"
 
 static struct drm_map_list *drm_find_matching_map(struct drm_device *dev,
 						  struct drm_local_map *map)
@@ -365,9 +361,9 @@ static int drm_addmap_core(struct drm_device * dev, resource_size_t offset,
 	return 0;
 }
 
-int drm_addmap(struct drm_device * dev, resource_size_t offset,
-	       unsigned int size, enum drm_map_type type,
-	       enum drm_map_flags flags, struct drm_local_map ** map_ptr)
+int drm_legacy_addmap(struct drm_device * dev, resource_size_t offset,
+		      unsigned int size, enum drm_map_type type,
+		      enum drm_map_flags flags, struct drm_local_map **map_ptr)
 {
 	struct drm_map_list *list;
 	int rc;
@@ -377,8 +373,7 @@ int drm_addmap(struct drm_device * dev, resource_size_t offset,
 		*map_ptr = list->map;
 	return rc;
 }
-
-EXPORT_SYMBOL(drm_addmap);
+EXPORT_SYMBOL(drm_legacy_addmap);
 
 /**
  * Ioctl to specify a range of memory that is available for mapping by a
@@ -391,8 +386,8 @@ EXPORT_SYMBOL(drm_addmap);
  * \return zero on success or a negative value on error.
  *
  */
-int drm_addmap_ioctl(struct drm_device *dev, void *data,
-		     struct drm_file *file_priv)
+int drm_legacy_addmap_ioctl(struct drm_device *dev, void *data,
+			    struct drm_file *file_priv)
 {
 	struct drm_map *map = data;
 	struct drm_map_list *maplist;
@@ -429,9 +424,9 @@ int drm_addmap_ioctl(struct drm_device *dev, void *data,
  * its being used, and free any associate resource (such as MTRR's) if it's not
  * being on use.
  *
- * \sa drm_addmap
+ * \sa drm_legacy_addmap
  */
-int drm_rmmap_locked(struct drm_device *dev, struct drm_local_map *map)
+int drm_legacy_rmmap_locked(struct drm_device *dev, struct drm_local_map *map)
 {
 	struct drm_map_list *r_list = NULL, *list_t;
 	drm_dma_handle_t dmah;
@@ -478,26 +473,26 @@ int drm_rmmap_locked(struct drm_device *dev, struct drm_local_map *map)
 		dmah.vaddr = map->handle;
 		dmah.busaddr = map->offset;
 		dmah.size = map->size;
-		__drm_pci_free(dev, &dmah);
+		__drm_legacy_pci_free(dev, &dmah);
 		break;
 	}
 	kfree(map);
 
 	return 0;
 }
-EXPORT_SYMBOL(drm_rmmap_locked);
+EXPORT_SYMBOL(drm_legacy_rmmap_locked);
 
-int drm_rmmap(struct drm_device *dev, struct drm_local_map *map)
+int drm_legacy_rmmap(struct drm_device *dev, struct drm_local_map *map)
 {
 	int ret;
 
 	mutex_lock(&dev->struct_mutex);
-	ret = drm_rmmap_locked(dev, map);
+	ret = drm_legacy_rmmap_locked(dev, map);
 	mutex_unlock(&dev->struct_mutex);
 
 	return ret;
 }
-EXPORT_SYMBOL(drm_rmmap);
+EXPORT_SYMBOL(drm_legacy_rmmap);
 
 /* The rmmap ioctl appears to be unnecessary.  All mappings are torn down on
  * the last close of the device, and this is necessary for cleanup when things
@@ -514,8 +509,8 @@ EXPORT_SYMBOL(drm_rmmap);
  * \param arg pointer to a struct drm_map structure.
  * \return zero on success or a negative value on error.
  */
-int drm_rmmap_ioctl(struct drm_device *dev, void *data,
-		    struct drm_file *file_priv)
+int drm_legacy_rmmap_ioctl(struct drm_device *dev, void *data,
+			   struct drm_file *file_priv)
 {
 	struct drm_map *request = data;
 	struct drm_local_map *map = NULL;
@@ -546,7 +541,7 @@ int drm_rmmap_ioctl(struct drm_device *dev, void *data,
 		return 0;
 	}
 
-	ret = drm_rmmap_locked(dev, map);
+	ret = drm_legacy_rmmap_locked(dev, map);
 
 	mutex_unlock(&dev->struct_mutex);
 
@@ -599,7 +594,8 @@ static void drm_cleanup_buf_error(struct drm_device * dev,
  * reallocates the buffer list of the same size order to accommodate the new
  * buffers.
  */
-int drm_addbufs_agp(struct drm_device * dev, struct drm_buf_desc * request)
+int drm_legacy_addbufs_agp(struct drm_device *dev,
+			   struct drm_buf_desc *request)
 {
 	struct drm_device_dma *dma = dev->dma;
 	struct drm_buf_entry *entry;
@@ -759,10 +755,11 @@ int drm_addbufs_agp(struct drm_device * dev, struct drm_buf_desc * request)
 	atomic_dec(&dev->buf_alloc);
 	return 0;
 }
-EXPORT_SYMBOL(drm_addbufs_agp);
+EXPORT_SYMBOL(drm_legacy_addbufs_agp);
 #endif				/* __OS_HAS_AGP */
 
-int drm_addbufs_pci(struct drm_device * dev, struct drm_buf_desc * request)
+int drm_legacy_addbufs_pci(struct drm_device *dev,
+			   struct drm_buf_desc *request)
 {
 	struct drm_device_dma *dma = dev->dma;
 	int count;
@@ -964,9 +961,10 @@ int drm_addbufs_pci(struct drm_device * dev, struct drm_buf_desc * request)
 	return 0;
 
 }
-EXPORT_SYMBOL(drm_addbufs_pci);
+EXPORT_SYMBOL(drm_legacy_addbufs_pci);
 
-static int drm_addbufs_sg(struct drm_device * dev, struct drm_buf_desc * request)
+static int drm_legacy_addbufs_sg(struct drm_device *dev,
+				 struct drm_buf_desc *request)
 {
 	struct drm_device_dma *dma = dev->dma;
 	struct drm_buf_entry *entry;
@@ -1135,8 +1133,8 @@ static int drm_addbufs_sg(struct drm_device * dev, struct drm_buf_desc * request
  * addbufs_sg() or addbufs_pci() for AGP, scatter-gather or consistent
  * PCI memory respectively.
  */
-int drm_addbufs(struct drm_device *dev, void *data,
-		struct drm_file *file_priv)
+int drm_legacy_addbufs(struct drm_device *dev, void *data,
+		       struct drm_file *file_priv)
 {
 	struct drm_buf_desc *request = data;
 	int ret;
@@ -1149,15 +1147,15 @@ int drm_addbufs(struct drm_device *dev, void *data,
 
 #if __OS_HAS_AGP
 	if (request->flags & _DRM_AGP_BUFFER)
-		ret = drm_addbufs_agp(dev, request);
+		ret = drm_legacy_addbufs_agp(dev, request);
 	else
 #endif
 	if (request->flags & _DRM_SG_BUFFER)
-		ret = drm_addbufs_sg(dev, request);
+		ret = drm_legacy_addbufs_sg(dev, request);
 	else if (request->flags & _DRM_FB_BUFFER)
 		ret = -EINVAL;
 	else
-		ret = drm_addbufs_pci(dev, request);
+		ret = drm_legacy_addbufs_pci(dev, request);
 
 	return ret;
 }
@@ -1179,8 +1177,8 @@ int drm_addbufs(struct drm_device *dev, void *data,
  * lock, preventing of allocating more buffers after this call. Information
  * about each requested buffer is then copied into user space.
  */
-int drm_infobufs(struct drm_device *dev, void *data,
-		 struct drm_file *file_priv)
+int drm_legacy_infobufs(struct drm_device *dev, void *data,
+			struct drm_file *file_priv)
 {
 	struct drm_device_dma *dma = dev->dma;
 	struct drm_buf_info *request = data;
@@ -1260,8 +1258,8 @@ int drm_infobufs(struct drm_device *dev, void *data,
  *
  * \note This ioctl is deprecated and mostly never used.
  */
-int drm_markbufs(struct drm_device *dev, void *data,
-		 struct drm_file *file_priv)
+int drm_legacy_markbufs(struct drm_device *dev, void *data,
+			struct drm_file *file_priv)
 {
 	struct drm_device_dma *dma = dev->dma;
 	struct drm_buf_desc *request = data;
@@ -1307,8 +1305,8 @@ int drm_markbufs(struct drm_device *dev, void *data,
  * Calls free_buffer() for each used buffer.
  * This function is primarily used for debugging.
  */
-int drm_freebufs(struct drm_device *dev, void *data,
-		 struct drm_file *file_priv)
+int drm_legacy_freebufs(struct drm_device *dev, void *data,
+			struct drm_file *file_priv)
 {
 	struct drm_device_dma *dma = dev->dma;
 	struct drm_buf_free *request = data;
@@ -1340,7 +1338,7 @@ int drm_freebufs(struct drm_device *dev, void *data,
 				  task_pid_nr(current));
 			return -EINVAL;
 		}
-		drm_free_buffer(dev, buf);
+		drm_legacy_free_buffer(dev, buf);
 	}
 
 	return 0;
@@ -1360,8 +1358,8 @@ int drm_freebufs(struct drm_device *dev, void *data,
  * offset equal to 0, which drm_mmap() interpretes as PCI buffers and calls
  * drm_mmap_dma().
  */
-int drm_mapbufs(struct drm_device *dev, void *data,
-	        struct drm_file *file_priv)
+int drm_legacy_mapbufs(struct drm_device *dev, void *data,
+		       struct drm_file *file_priv)
 {
 	struct drm_device_dma *dma = dev->dma;
 	int retcode = 0;
@@ -1448,7 +1446,7 @@ int drm_mapbufs(struct drm_device *dev, void *data,
 	return retcode;
 }
 
-int drm_dma_ioctl(struct drm_device *dev, void *data,
+int drm_legacy_dma_ioctl(struct drm_device *dev, void *data,
 		  struct drm_file *file_priv)
 {
 	if (drm_core_check_feature(dev, DRIVER_MODESET))
@@ -1460,7 +1458,7 @@ int drm_dma_ioctl(struct drm_device *dev, void *data,
 		return -EINVAL;
 }
 
-struct drm_local_map *drm_getsarea(struct drm_device *dev)
+struct drm_local_map *drm_legacy_getsarea(struct drm_device *dev)
 {
 	struct drm_map_list *entry;
 
@@ -1472,4 +1470,4 @@ struct drm_local_map *drm_getsarea(struct drm_device *dev)
 	}
 	return NULL;
 }
-EXPORT_SYMBOL(drm_getsarea);
+EXPORT_SYMBOL(drm_legacy_getsarea);
diff --git a/drivers/gpu/drm/drm_crtc.c b/drivers/gpu/drm/drm_crtc.c
index 90e773019eac..e79c8d3700d8 100644
--- a/drivers/gpu/drm/drm_crtc.c
+++ b/drivers/gpu/drm/drm_crtc.c
@@ -40,106 +40,12 @@
 #include <drm/drm_modeset_lock.h>
 
 #include "drm_crtc_internal.h"
+#include "drm_internal.h"
 
 static struct drm_framebuffer *add_framebuffer_internal(struct drm_device *dev,
 							struct drm_mode_fb_cmd2 *r,
 							struct drm_file *file_priv);
 
-/**
- * drm_modeset_lock_all - take all modeset locks
- * @dev: drm device
- *
- * This function takes all modeset locks, suitable where a more fine-grained
- * scheme isn't (yet) implemented. Locks must be dropped with
- * drm_modeset_unlock_all.
- */
-void drm_modeset_lock_all(struct drm_device *dev)
-{
-	struct drm_mode_config *config = &dev->mode_config;
-	struct drm_modeset_acquire_ctx *ctx;
-	int ret;
-
-	ctx = kzalloc(sizeof(*ctx), GFP_KERNEL);
-	if (WARN_ON(!ctx))
-		return;
-
-	mutex_lock(&config->mutex);
-
-	drm_modeset_acquire_init(ctx, 0);
-
-retry:
-	ret = drm_modeset_lock(&config->connection_mutex, ctx);
-	if (ret)
-		goto fail;
-	ret = drm_modeset_lock_all_crtcs(dev, ctx);
-	if (ret)
-		goto fail;
-
-	WARN_ON(config->acquire_ctx);
-
-	/* now we hold the locks, so now that it is safe, stash the
-	 * ctx for drm_modeset_unlock_all():
-	 */
-	config->acquire_ctx = ctx;
-
-	drm_warn_on_modeset_not_all_locked(dev);
-
-	return;
-
-fail:
-	if (ret == -EDEADLK) {
-		drm_modeset_backoff(ctx);
-		goto retry;
-	}
-}
-EXPORT_SYMBOL(drm_modeset_lock_all);
-
-/**
- * drm_modeset_unlock_all - drop all modeset locks
- * @dev: device
- *
- * This function drop all modeset locks taken by drm_modeset_lock_all.
- */
-void drm_modeset_unlock_all(struct drm_device *dev)
-{
-	struct drm_mode_config *config = &dev->mode_config;
-	struct drm_modeset_acquire_ctx *ctx = config->acquire_ctx;
-
-	if (WARN_ON(!ctx))
-		return;
-
-	config->acquire_ctx = NULL;
-	drm_modeset_drop_locks(ctx);
-	drm_modeset_acquire_fini(ctx);
-
-	kfree(ctx);
-
-	mutex_unlock(&dev->mode_config.mutex);
-}
-EXPORT_SYMBOL(drm_modeset_unlock_all);
-
-/**
- * drm_warn_on_modeset_not_all_locked - check that all modeset locks are locked
- * @dev: device
- *
- * Useful as a debug assert.
- */
-void drm_warn_on_modeset_not_all_locked(struct drm_device *dev)
-{
-	struct drm_crtc *crtc;
-
-	/* Locking is currently fubar in the panic handler. */
-	if (oops_in_progress)
-		return;
-
-	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head)
-		WARN_ON(!drm_modeset_is_locked(&crtc->mutex));
-
-	WARN_ON(!drm_modeset_is_locked(&dev->mode_config.connection_mutex));
-	WARN_ON(!mutex_is_locked(&dev->mode_config.mutex));
-}
-EXPORT_SYMBOL(drm_warn_on_modeset_not_all_locked);
-
 /* Avoid boilerplate.  I'm tired of typing. */
 #define DRM_ENUM_NAME_FN(fnname, list)				\
 	const char *fnname(int val)				\
@@ -515,9 +421,6 @@ int drm_framebuffer_init(struct drm_device *dev, struct drm_framebuffer *fb,
 	if (ret)
 		goto out;
 
-	/* Grab the idr reference. */
-	drm_framebuffer_reference(fb);
-
 	dev->mode_config.num_fb++;
 	list_add(&fb->head, &dev->mode_config.fb_list);
 out:
@@ -527,10 +430,34 @@ out:
 }
 EXPORT_SYMBOL(drm_framebuffer_init);
 
+/* dev->mode_config.fb_lock must be held! */
+static void __drm_framebuffer_unregister(struct drm_device *dev,
+					 struct drm_framebuffer *fb)
+{
+	mutex_lock(&dev->mode_config.idr_mutex);
+	idr_remove(&dev->mode_config.crtc_idr, fb->base.id);
+	mutex_unlock(&dev->mode_config.idr_mutex);
+
+	fb->base.id = 0;
+}
+
 static void drm_framebuffer_free(struct kref *kref)
 {
 	struct drm_framebuffer *fb =
 			container_of(kref, struct drm_framebuffer, refcount);
+	struct drm_device *dev = fb->dev;
+
+	/*
+	 * The lookup idr holds a weak reference, which has not necessarily been
+	 * removed at this point. Check for that.
+	 */
+	mutex_lock(&dev->mode_config.fb_lock);
+	if (fb->base.id) {
+		/* Mark fb as reaped and drop idr ref. */
+		__drm_framebuffer_unregister(dev, fb);
+	}
+	mutex_unlock(&dev->mode_config.fb_lock);
+
 	fb->funcs->destroy(fb);
 }
 
@@ -567,8 +494,10 @@ struct drm_framebuffer *drm_framebuffer_lookup(struct drm_device *dev,
 
 	mutex_lock(&dev->mode_config.fb_lock);
 	fb = __drm_framebuffer_lookup(dev, id);
-	if (fb)
-		drm_framebuffer_reference(fb);
+	if (fb) {
+		if (!kref_get_unless_zero(&fb->refcount))
+			fb = NULL;
+	}
 	mutex_unlock(&dev->mode_config.fb_lock);
 
 	return fb;
@@ -612,19 +541,6 @@ static void __drm_framebuffer_unreference(struct drm_framebuffer *fb)
 	kref_put(&fb->refcount, drm_framebuffer_free_bug);
 }
 
-/* dev->mode_config.fb_lock must be held! */
-static void __drm_framebuffer_unregister(struct drm_device *dev,
-					 struct drm_framebuffer *fb)
-{
-	mutex_lock(&dev->mode_config.idr_mutex);
-	idr_remove(&dev->mode_config.crtc_idr, fb->base.id);
-	mutex_unlock(&dev->mode_config.idr_mutex);
-
-	fb->base.id = 0;
-
-	__drm_framebuffer_unreference(fb);
-}
-
 /**
  * drm_framebuffer_unregister_private - unregister a private fb from the lookup idr
  * @fb: fb to unregister
@@ -764,11 +680,7 @@ int drm_crtc_init_with_planes(struct drm_device *dev, struct drm_crtc *crtc,
 	crtc->funcs = funcs;
 	crtc->invert_dimensions = false;
 
-	drm_modeset_lock_all(dev);
 	drm_modeset_lock_init(&crtc->mutex);
-	/* dropped by _unlock_all(): */
-	drm_modeset_lock(&crtc->mutex, config->acquire_ctx);
-
 	ret = drm_mode_object_get(dev, &crtc->base, DRM_MODE_OBJECT_CRTC);
 	if (ret)
 		goto out;
@@ -786,7 +698,6 @@ int drm_crtc_init_with_planes(struct drm_device *dev, struct drm_crtc *crtc,
 		cursor->possible_crtcs = 1 << drm_crtc_index(crtc);
 
  out:
-	drm_modeset_unlock_all(dev);
 
 	return ret;
 }
@@ -853,6 +764,59 @@ static void drm_mode_remove(struct drm_connector *connector,
 }
 
 /**
+ * drm_connector_get_cmdline_mode - reads the user's cmdline mode
+ * @connector: connector to quwery
+ * @mode: returned mode
+ *
+ * The kernel supports per-connector configration of its consoles through
+ * use of the video= parameter. This function parses that option and
+ * extracts the user's specified mode (or enable/disable status) for a
+ * particular connector. This is typically only used during the early fbdev
+ * setup.
+ */
+static void drm_connector_get_cmdline_mode(struct drm_connector *connector)
+{
+	struct drm_cmdline_mode *mode = &connector->cmdline_mode;
+	char *option = NULL;
+
+	if (fb_get_options(connector->name, &option))
+		return;
+
+	if (!drm_mode_parse_command_line_for_connector(option,
+						       connector,
+						       mode))
+		return;
+
+	if (mode->force) {
+		const char *s;
+
+		switch (mode->force) {
+		case DRM_FORCE_OFF:
+			s = "OFF";
+			break;
+		case DRM_FORCE_ON_DIGITAL:
+			s = "ON - dig";
+			break;
+		default:
+		case DRM_FORCE_ON:
+			s = "ON";
+			break;
+		}
+
+		DRM_INFO("forcing %s connector %s\n", connector->name, s);
+		connector->force = mode->force;
+	}
+
+	DRM_DEBUG_KMS("cmdline mode for connector %s %dx%d@%dHz%s%s%s\n",
+		      connector->name,
+		      mode->xres, mode->yres,
+		      mode->refresh_specified ? mode->refresh : 60,
+		      mode->rb ? " reduced blanking" : "",
+		      mode->margins ? " with margins" : "",
+		      mode->interlace ?  " interlaced" : "");
+}
+
+/**
  * drm_connector_init - Init a preallocated connector
  * @dev: DRM device
  * @connector: the connector to init
@@ -904,6 +868,8 @@ int drm_connector_init(struct drm_device *dev,
 	connector->edid_blob_ptr = NULL;
 	connector->status = connector_status_unknown;
 
+	drm_connector_get_cmdline_mode(connector);
+
 	list_add_tail(&connector->head, &dev->mode_config.connector_list);
 	dev->mode_config.num_connector++;
 
@@ -957,6 +923,29 @@ void drm_connector_cleanup(struct drm_connector *connector)
 EXPORT_SYMBOL(drm_connector_cleanup);
 
 /**
+ * drm_connector_index - find the index of a registered connector
+ * @connector: connector to find index for
+ *
+ * Given a registered connector, return the index of that connector within a DRM
+ * device's list of connectors.
+ */
+unsigned int drm_connector_index(struct drm_connector *connector)
+{
+	unsigned int index = 0;
+	struct drm_connector *tmp;
+
+	list_for_each_entry(tmp, &connector->dev->mode_config.connector_list, head) {
+		if (tmp == connector)
+			return index;
+
+		index++;
+	}
+
+	BUG();
+}
+EXPORT_SYMBOL(drm_connector_index);
+
+/**
  * drm_connector_register - register a connector
  * @connector: the connector to register
  *
@@ -1261,6 +1250,29 @@ void drm_plane_cleanup(struct drm_plane *plane)
 EXPORT_SYMBOL(drm_plane_cleanup);
 
 /**
+ * drm_plane_index - find the index of a registered plane
+ * @plane: plane to find index for
+ *
+ * Given a registered plane, return the index of that CRTC within a DRM
+ * device's list of planes.
+ */
+unsigned int drm_plane_index(struct drm_plane *plane)
+{
+	unsigned int index = 0;
+	struct drm_plane *tmp;
+
+	list_for_each_entry(tmp, &plane->dev->mode_config.plane_list, head) {
+		if (tmp == plane)
+			return index;
+
+		index++;
+	}
+
+	BUG();
+}
+EXPORT_SYMBOL(drm_plane_index);
+
+/**
  * drm_plane_force_disable - Forcibly disable a plane
  * @plane: plane to disable
  *
@@ -1271,19 +1283,21 @@ EXPORT_SYMBOL(drm_plane_cleanup);
  */
 void drm_plane_force_disable(struct drm_plane *plane)
 {
-	struct drm_framebuffer *old_fb = plane->fb;
 	int ret;
 
-	if (!old_fb)
+	if (!plane->fb)
 		return;
 
+	plane->old_fb = plane->fb;
 	ret = plane->funcs->disable_plane(plane);
 	if (ret) {
 		DRM_ERROR("failed to disable plane with busy fb\n");
+		plane->old_fb = NULL;
 		return;
 	}
 	/* disconnect the plane from the fb and crtc: */
-	__drm_framebuffer_unreference(old_fb);
+	__drm_framebuffer_unreference(plane->old_fb);
+	plane->old_fb = NULL;
 	plane->fb = NULL;
 	plane->crtc = NULL;
 }
@@ -2249,33 +2263,29 @@ out:
  *
  * src_{x,y,w,h} are provided in 16.16 fixed point format
  */
-static int setplane_internal(struct drm_plane *plane,
-			     struct drm_crtc *crtc,
-			     struct drm_framebuffer *fb,
-			     int32_t crtc_x, int32_t crtc_y,
-			     uint32_t crtc_w, uint32_t crtc_h,
-			     /* src_{x,y,w,h} values are 16.16 fixed point */
-			     uint32_t src_x, uint32_t src_y,
-			     uint32_t src_w, uint32_t src_h)
+static int __setplane_internal(struct drm_plane *plane,
+			       struct drm_crtc *crtc,
+			       struct drm_framebuffer *fb,
+			       int32_t crtc_x, int32_t crtc_y,
+			       uint32_t crtc_w, uint32_t crtc_h,
+			       /* src_{x,y,w,h} values are 16.16 fixed point */
+			       uint32_t src_x, uint32_t src_y,
+			       uint32_t src_w, uint32_t src_h)
 {
-	struct drm_device *dev = plane->dev;
-	struct drm_framebuffer *old_fb = NULL;
 	int ret = 0;
 	unsigned int fb_width, fb_height;
 	int i;
 
 	/* No fb means shut it down */
 	if (!fb) {
-		drm_modeset_lock_all(dev);
-		old_fb = plane->fb;
+		plane->old_fb = plane->fb;
 		ret = plane->funcs->disable_plane(plane);
 		if (!ret) {
 			plane->crtc = NULL;
 			plane->fb = NULL;
 		} else {
-			old_fb = NULL;
+			plane->old_fb = NULL;
 		}
-		drm_modeset_unlock_all(dev);
 		goto out;
 	}
 
@@ -2315,8 +2325,7 @@ static int setplane_internal(struct drm_plane *plane,
 		goto out;
 	}
 
-	drm_modeset_lock_all(dev);
-	old_fb = plane->fb;
+	plane->old_fb = plane->fb;
 	ret = plane->funcs->update_plane(plane, crtc, fb,
 					 crtc_x, crtc_y, crtc_w, crtc_h,
 					 src_x, src_y, src_w, src_h);
@@ -2325,18 +2334,37 @@ static int setplane_internal(struct drm_plane *plane,
 		plane->fb = fb;
 		fb = NULL;
 	} else {
-		old_fb = NULL;
+		plane->old_fb = NULL;
 	}
-	drm_modeset_unlock_all(dev);
 
 out:
 	if (fb)
 		drm_framebuffer_unreference(fb);
-	if (old_fb)
-		drm_framebuffer_unreference(old_fb);
+	if (plane->old_fb)
+		drm_framebuffer_unreference(plane->old_fb);
+	plane->old_fb = NULL;
 
 	return ret;
+}
+
+static int setplane_internal(struct drm_plane *plane,
+			     struct drm_crtc *crtc,
+			     struct drm_framebuffer *fb,
+			     int32_t crtc_x, int32_t crtc_y,
+			     uint32_t crtc_w, uint32_t crtc_h,
+			     /* src_{x,y,w,h} values are 16.16 fixed point */
+			     uint32_t src_x, uint32_t src_y,
+			     uint32_t src_w, uint32_t src_h)
+{
+	int ret;
+
+	drm_modeset_lock_all(plane->dev);
+	ret = __setplane_internal(plane, crtc, fb,
+				  crtc_x, crtc_y, crtc_w, crtc_h,
+				  src_x, src_y, src_w, src_h);
+	drm_modeset_unlock_all(plane->dev);
 
+	return ret;
 }
 
 /**
@@ -2440,7 +2468,7 @@ int drm_mode_set_config_internal(struct drm_mode_set *set)
 	 * crtcs. Atomic modeset will have saner semantics ...
 	 */
 	list_for_each_entry(tmp, &crtc->dev->mode_config.crtc_list, head)
-		tmp->old_fb = tmp->primary->fb;
+		tmp->primary->old_fb = tmp->primary->fb;
 
 	fb = set->fb;
 
@@ -2453,8 +2481,9 @@ int drm_mode_set_config_internal(struct drm_mode_set *set)
 	list_for_each_entry(tmp, &crtc->dev->mode_config.crtc_list, head) {
 		if (tmp->primary->fb)
 			drm_framebuffer_reference(tmp->primary->fb);
-		if (tmp->old_fb)
-			drm_framebuffer_unreference(tmp->old_fb);
+		if (tmp->primary->old_fb)
+			drm_framebuffer_unreference(tmp->primary->old_fb);
+		tmp->primary->old_fb = NULL;
 	}
 
 	return ret;
@@ -2701,6 +2730,7 @@ static int drm_mode_cursor_universal(struct drm_crtc *crtc,
 	int ret = 0;
 
 	BUG_ON(!crtc->cursor);
+	WARN_ON(crtc->cursor->crtc != crtc && crtc->cursor->crtc != NULL);
 
 	/*
 	 * Obtain fb we'll be using (either new or existing) and take an extra
@@ -2720,11 +2750,9 @@ static int drm_mode_cursor_universal(struct drm_crtc *crtc,
 			fb = NULL;
 		}
 	} else {
-		mutex_lock(&dev->mode_config.mutex);
 		fb = crtc->cursor->fb;
 		if (fb)
 			drm_framebuffer_reference(fb);
-		mutex_unlock(&dev->mode_config.mutex);
 	}
 
 	if (req->flags & DRM_MODE_CURSOR_MOVE) {
@@ -2746,7 +2774,7 @@ static int drm_mode_cursor_universal(struct drm_crtc *crtc,
 	 * setplane_internal will take care of deref'ing either the old or new
 	 * framebuffer depending on success.
 	 */
-	ret = setplane_internal(crtc->cursor, crtc, fb,
+	ret = __setplane_internal(crtc->cursor, crtc, fb,
 				crtc_x, crtc_y, crtc_w, crtc_h,
 				0, 0, src_w, src_h);
 
@@ -2782,10 +2810,12 @@ static int drm_mode_cursor_common(struct drm_device *dev,
 	 * If this crtc has a universal cursor plane, call that plane's update
 	 * handler rather than using legacy cursor handlers.
 	 */
-	if (crtc->cursor)
-		return drm_mode_cursor_universal(crtc, req, file_priv);
+	drm_modeset_lock_crtc(crtc);
+	if (crtc->cursor) {
+		ret = drm_mode_cursor_universal(crtc, req, file_priv);
+		goto out;
+	}
 
-	drm_modeset_lock(&crtc->mutex, NULL);
 	if (req->flags & DRM_MODE_CURSOR_BO) {
 		if (!crtc->funcs->cursor_set && !crtc->funcs->cursor_set2) {
 			ret = -ENXIO;
@@ -2809,7 +2839,7 @@ static int drm_mode_cursor_common(struct drm_device *dev,
 		}
 	}
 out:
-	drm_modeset_unlock(&crtc->mutex);
+	drm_modeset_unlock_crtc(crtc);
 
 	return ret;
 
@@ -3370,7 +3400,16 @@ void drm_fb_release(struct drm_file *priv)
 	struct drm_device *dev = priv->minor->dev;
 	struct drm_framebuffer *fb, *tfb;
 
-	mutex_lock(&priv->fbs_lock);
+	/*
+	 * When the file gets released that means no one else can access the fb
+	 * list any more, so no need to grab fpriv->fbs_lock. And we need to to
+	 * avoid upsetting lockdep since the universal cursor code adds a
+	 * framebuffer while holding mutex locks.
+	 *
+	 * Note that a real deadlock between fpriv->fbs_lock and the modeset
+	 * locks is impossible here since no one else but this function can get
+	 * at it any more.
+	 */
 	list_for_each_entry_safe(fb, tfb, &priv->fbs, filp_head) {
 
 		mutex_lock(&dev->mode_config.fb_lock);
@@ -3383,7 +3422,6 @@ void drm_fb_release(struct drm_file *priv)
 		/* This will also drop the fpriv->fbs reference. */
 		drm_framebuffer_remove(fb);
 	}
-	mutex_unlock(&priv->fbs_lock);
 }
 
 /**
@@ -3495,9 +3533,10 @@ EXPORT_SYMBOL(drm_property_create_enum);
  * @flags: flags specifying the property type
  * @name: name of the property
  * @props: enumeration lists with property bitflags
- * @num_values: number of pre-defined values
+ * @num_props: size of the @props array
+ * @supported_bits: bitmask of all supported enumeration values
  *
- * This creates a new generic drm property which can then be attached to a drm
+ * This creates a new bitmask drm property which can then be attached to a drm
  * object with drm_object_attach_property. The returned property object must be
  * freed with drm_property_destroy.
  *
@@ -4157,12 +4196,25 @@ static int drm_mode_crtc_set_obj_prop(struct drm_mode_object *obj,
 	return ret;
 }
 
-static int drm_mode_plane_set_obj_prop(struct drm_mode_object *obj,
-				      struct drm_property *property,
-				      uint64_t value)
+/**
+ * drm_mode_plane_set_obj_prop - set the value of a property
+ * @plane: drm plane object to set property value for
+ * @property: property to set
+ * @value: value the property should be set to
+ *
+ * This functions sets a given property on a given plane object. This function
+ * calls the driver's ->set_property callback and changes the software state of
+ * the property if the callback succeeds.
+ *
+ * Returns:
+ * Zero on success, error code on failure.
+ */
+int drm_mode_plane_set_obj_prop(struct drm_plane *plane,
+				struct drm_property *property,
+				uint64_t value)
 {
 	int ret = -EINVAL;
-	struct drm_plane *plane = obj_to_plane(obj);
+	struct drm_mode_object *obj = &plane->base;
 
 	if (plane->funcs->set_property)
 		ret = plane->funcs->set_property(plane, property, value);
@@ -4171,6 +4223,7 @@ static int drm_mode_plane_set_obj_prop(struct drm_mode_object *obj,
 
 	return ret;
 }
+EXPORT_SYMBOL(drm_mode_plane_set_obj_prop);
 
 /**
  * drm_mode_getproperty_ioctl - get the current value of a object's property
@@ -4309,7 +4362,8 @@ int drm_mode_obj_set_property_ioctl(struct drm_device *dev, void *data,
 		ret = drm_mode_crtc_set_obj_prop(arg_obj, property, arg->value);
 		break;
 	case DRM_MODE_OBJECT_PLANE:
-		ret = drm_mode_plane_set_obj_prop(arg_obj, property, arg->value);
+		ret = drm_mode_plane_set_obj_prop(obj_to_plane(arg_obj),
+						  property, arg->value);
 		break;
 	}
 
@@ -4529,7 +4583,7 @@ int drm_mode_page_flip_ioctl(struct drm_device *dev,
 {
 	struct drm_mode_crtc_page_flip *page_flip = data;
 	struct drm_crtc *crtc;
-	struct drm_framebuffer *fb = NULL, *old_fb = NULL;
+	struct drm_framebuffer *fb = NULL;
 	struct drm_pending_vblank_event *e = NULL;
 	unsigned long flags;
 	int ret = -EINVAL;
@@ -4545,7 +4599,7 @@ int drm_mode_page_flip_ioctl(struct drm_device *dev,
 	if (!crtc)
 		return -ENOENT;
 
-	drm_modeset_lock(&crtc->mutex, NULL);
+	drm_modeset_lock_crtc(crtc);
 	if (crtc->primary->fb == NULL) {
 		/* The framebuffer is currently unbound, presumably
 		 * due to a hotplug event, that userspace has not
@@ -4601,7 +4655,7 @@ int drm_mode_page_flip_ioctl(struct drm_device *dev,
 			(void (*) (struct drm_pending_event *)) kfree;
 	}
 
-	old_fb = crtc->primary->fb;
+	crtc->primary->old_fb = crtc->primary->fb;
 	ret = crtc->funcs->page_flip(crtc, fb, e, page_flip->flags);
 	if (ret) {
 		if (page_flip->flags & DRM_MODE_PAGE_FLIP_EVENT) {
@@ -4611,7 +4665,7 @@ int drm_mode_page_flip_ioctl(struct drm_device *dev,
 			kfree(e);
 		}
 		/* Keep the old fb, don't unref it. */
-		old_fb = NULL;
+		crtc->primary->old_fb = NULL;
 	} else {
 		/*
 		 * Warn if the driver hasn't properly updated the crtc->fb
@@ -4627,9 +4681,10 @@ int drm_mode_page_flip_ioctl(struct drm_device *dev,
 out:
 	if (fb)
 		drm_framebuffer_unreference(fb);
-	if (old_fb)
-		drm_framebuffer_unreference(old_fb);
-	drm_modeset_unlock(&crtc->mutex);
+	if (crtc->primary->old_fb)
+		drm_framebuffer_unreference(crtc->primary->old_fb);
+	crtc->primary->old_fb = NULL;
+	drm_modeset_unlock_crtc(crtc);
 
 	return ret;
 }
@@ -4645,9 +4700,14 @@ out:
 void drm_mode_config_reset(struct drm_device *dev)
 {
 	struct drm_crtc *crtc;
+	struct drm_plane *plane;
 	struct drm_encoder *encoder;
 	struct drm_connector *connector;
 
+	list_for_each_entry(plane, &dev->mode_config.plane_list, head)
+		if (plane->funcs->reset)
+			plane->funcs->reset(plane);
+
 	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head)
 		if (crtc->funcs->reset)
 			crtc->funcs->reset(crtc);
diff --git a/drivers/gpu/drm/drm_debugfs.c b/drivers/gpu/drm/drm_debugfs.c
index 13bd42923dd4..3bcf8e6a85b3 100644
--- a/drivers/gpu/drm/drm_debugfs.c
+++ b/drivers/gpu/drm/drm_debugfs.c
@@ -36,6 +36,7 @@
 #include <linux/export.h>
 #include <drm/drmP.h>
 #include <drm/drm_edid.h>
+#include "drm_internal.h"
 
 #if defined(CONFIG_DEBUG_FS)
 
@@ -49,9 +50,7 @@ static const struct drm_info_list drm_debugfs_list[] = {
 	{"clients", drm_clients_info, 0},
 	{"bufs", drm_bufs_info, 0},
 	{"gem_names", drm_gem_name_info, DRIVER_GEM},
-#if DRM_DEBUG_CODE
 	{"vma", drm_vma_info, 0},
-#endif
 };
 #define DRM_DEBUGFS_ENTRIES ARRAY_SIZE(drm_debugfs_list)
 
diff --git a/drivers/gpu/drm/drm_dma.c b/drivers/gpu/drm/drm_dma.c
index 8a140a953754..ea481800ef56 100644
--- a/drivers/gpu/drm/drm_dma.c
+++ b/drivers/gpu/drm/drm_dma.c
@@ -35,6 +35,7 @@
 
 #include <linux/export.h>
 #include <drm/drmP.h>
+#include "drm_legacy.h"
 
 /**
  * Initialize the DMA data.
@@ -124,7 +125,7 @@ void drm_legacy_dma_takedown(struct drm_device *dev)
  *
  * Resets the fields of \p buf.
  */
-void drm_free_buffer(struct drm_device *dev, struct drm_buf * buf)
+void drm_legacy_free_buffer(struct drm_device *dev, struct drm_buf * buf)
 {
 	if (!buf)
 		return;
@@ -142,8 +143,8 @@ void drm_free_buffer(struct drm_device *dev, struct drm_buf * buf)
  *
  * Frees each buffer associated with \p file_priv not already on the hardware.
  */
-void drm_core_reclaim_buffers(struct drm_device *dev,
-			      struct drm_file *file_priv)
+void drm_legacy_reclaim_buffers(struct drm_device *dev,
+				struct drm_file *file_priv)
 {
 	struct drm_device_dma *dma = dev->dma;
 	int i;
@@ -154,7 +155,7 @@ void drm_core_reclaim_buffers(struct drm_device *dev,
 		if (dma->buflist[i]->file_priv == file_priv) {
 			switch (dma->buflist[i]->list) {
 			case DRM_LIST_NONE:
-				drm_free_buffer(dev, dma->buflist[i]);
+				drm_legacy_free_buffer(dev, dma->buflist[i]);
 				break;
 			case DRM_LIST_WAIT:
 				dma->buflist[i]->list = DRM_LIST_RECLAIM;
@@ -166,5 +167,3 @@ void drm_core_reclaim_buffers(struct drm_device *dev,
 		}
 	}
 }
-
-EXPORT_SYMBOL(drm_core_reclaim_buffers);
diff --git a/drivers/gpu/drm/drm_dp_mst_topology.c b/drivers/gpu/drm/drm_dp_mst_topology.c
index ac3c2738db94..070f913d2dba 100644
--- a/drivers/gpu/drm/drm_dp_mst_topology.c
+++ b/drivers/gpu/drm/drm_dp_mst_topology.c
@@ -682,7 +682,7 @@ static int build_allocate_payload(struct drm_dp_sideband_msg_tx *msg, int port_n
 static int drm_dp_mst_assign_payload_id(struct drm_dp_mst_topology_mgr *mgr,
 					struct drm_dp_vcpi *vcpi)
 {
-	int ret;
+	int ret, vcpi_ret;
 
 	mutex_lock(&mgr->payload_lock);
 	ret = find_first_zero_bit(&mgr->payload_mask, mgr->max_payloads + 1);
@@ -692,8 +692,16 @@ static int drm_dp_mst_assign_payload_id(struct drm_dp_mst_topology_mgr *mgr,
 		goto out_unlock;
 	}
 
+	vcpi_ret = find_first_zero_bit(&mgr->vcpi_mask, mgr->max_payloads + 1);
+	if (vcpi_ret > mgr->max_payloads) {
+		ret = -EINVAL;
+		DRM_DEBUG_KMS("out of vcpi ids %d\n", ret);
+		goto out_unlock;
+	}
+
 	set_bit(ret, &mgr->payload_mask);
-	vcpi->vcpi = ret;
+	set_bit(vcpi_ret, &mgr->vcpi_mask);
+	vcpi->vcpi = vcpi_ret + 1;
 	mgr->proposed_vcpis[ret - 1] = vcpi;
 out_unlock:
 	mutex_unlock(&mgr->payload_lock);
@@ -701,15 +709,23 @@ out_unlock:
 }
 
 static void drm_dp_mst_put_payload_id(struct drm_dp_mst_topology_mgr *mgr,
-				      int id)
+				      int vcpi)
 {
-	if (id == 0)
+	int i;
+	if (vcpi == 0)
 		return;
 
 	mutex_lock(&mgr->payload_lock);
-	DRM_DEBUG_KMS("putting payload %d\n", id);
-	clear_bit(id, &mgr->payload_mask);
-	mgr->proposed_vcpis[id - 1] = NULL;
+	DRM_DEBUG_KMS("putting payload %d\n", vcpi);
+	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);
+			}
+	}
 	mutex_unlock(&mgr->payload_lock);
 }
 
@@ -1563,7 +1579,7 @@ static int drm_dp_destroy_payload_step1(struct drm_dp_mst_topology_mgr *mgr,
 	}
 
 	drm_dp_dpcd_write_payload(mgr, id, payload);
-	payload->payload_state = 0;
+	payload->payload_state = DP_PAYLOAD_DELETE_LOCAL;
 	return 0;
 }
 
@@ -1590,7 +1606,7 @@ static int drm_dp_destroy_payload_step2(struct drm_dp_mst_topology_mgr *mgr,
  */
 int drm_dp_update_payload_part1(struct drm_dp_mst_topology_mgr *mgr)
 {
-	int i;
+	int i, j;
 	int cur_slots = 1;
 	struct drm_dp_payload req_payload;
 	struct drm_dp_mst_port *port;
@@ -1607,26 +1623,46 @@ int drm_dp_update_payload_part1(struct drm_dp_mst_topology_mgr *mgr)
 			port = NULL;
 			req_payload.num_slots = 0;
 		}
+
+		if (mgr->payloads[i].start_slot != req_payload.start_slot) {
+			mgr->payloads[i].start_slot = req_payload.start_slot;
+		}
 		/* work out what is required to happen with this payload */
-		if (mgr->payloads[i].start_slot != req_payload.start_slot ||
-		    mgr->payloads[i].num_slots != req_payload.num_slots) {
+		if (mgr->payloads[i].num_slots != req_payload.num_slots) {
 
 			/* need to push an update for this payload */
 			if (req_payload.num_slots) {
-				drm_dp_create_payload_step1(mgr, i + 1, &req_payload);
+				drm_dp_create_payload_step1(mgr, mgr->proposed_vcpis[i]->vcpi, &req_payload);
 				mgr->payloads[i].num_slots = req_payload.num_slots;
 			} else if (mgr->payloads[i].num_slots) {
 				mgr->payloads[i].num_slots = 0;
-				drm_dp_destroy_payload_step1(mgr, port, i + 1, &mgr->payloads[i]);
+				drm_dp_destroy_payload_step1(mgr, port, port->vcpi.vcpi, &mgr->payloads[i]);
 				req_payload.payload_state = mgr->payloads[i].payload_state;
-			} else
-				req_payload.payload_state = 0;
-
-			mgr->payloads[i].start_slot = req_payload.start_slot;
+				mgr->payloads[i].start_slot = 0;
+			}
 			mgr->payloads[i].payload_state = req_payload.payload_state;
 		}
 		cur_slots += req_payload.num_slots;
 	}
+
+	for (i = 0; i < mgr->max_payloads; i++) {
+		if (mgr->payloads[i].payload_state == DP_PAYLOAD_DELETE_LOCAL) {
+			DRM_DEBUG_KMS("removing payload %d\n", i);
+			for (j = i; j < mgr->max_payloads - 1; j++) {
+				memcpy(&mgr->payloads[j], &mgr->payloads[j + 1], sizeof(struct drm_dp_payload));
+				mgr->proposed_vcpis[j] = mgr->proposed_vcpis[j + 1];
+				if (mgr->proposed_vcpis[j] && mgr->proposed_vcpis[j]->num_slots) {
+					set_bit(j + 1, &mgr->payload_mask);
+				} else {
+					clear_bit(j + 1, &mgr->payload_mask);
+				}
+			}
+			memset(&mgr->payloads[mgr->max_payloads - 1], 0, sizeof(struct drm_dp_payload));
+			mgr->proposed_vcpis[mgr->max_payloads - 1] = NULL;
+			clear_bit(mgr->max_payloads, &mgr->payload_mask);
+
+		}
+	}
 	mutex_unlock(&mgr->payload_lock);
 
 	return 0;
@@ -1657,9 +1693,9 @@ int drm_dp_update_payload_part2(struct drm_dp_mst_topology_mgr *mgr)
 
 		DRM_DEBUG_KMS("payload %d %d\n", i, mgr->payloads[i].payload_state);
 		if (mgr->payloads[i].payload_state == DP_PAYLOAD_LOCAL) {
-			ret = drm_dp_create_payload_step2(mgr, port, i + 1, &mgr->payloads[i]);
+			ret = drm_dp_create_payload_step2(mgr, port, mgr->proposed_vcpis[i]->vcpi, &mgr->payloads[i]);
 		} else if (mgr->payloads[i].payload_state == DP_PAYLOAD_DELETE_LOCAL) {
-			ret = drm_dp_destroy_payload_step2(mgr, i + 1, &mgr->payloads[i]);
+			ret = drm_dp_destroy_payload_step2(mgr, mgr->proposed_vcpis[i]->vcpi, &mgr->payloads[i]);
 		}
 		if (ret) {
 			mutex_unlock(&mgr->payload_lock);
@@ -1772,7 +1808,7 @@ static int drm_dp_get_vc_payload_bw(int dp_link_bw, int dp_link_count)
 	case DP_LINK_BW_5_4:
 		return 10 * dp_link_count;
 	}
-	return 0;
+	BUG();
 }
 
 /**
@@ -1861,6 +1897,7 @@ int drm_dp_mst_topology_mgr_set_mst(struct drm_dp_mst_topology_mgr *mgr, bool ms
 		memset(mgr->payloads, 0, mgr->max_payloads * sizeof(struct drm_dp_payload));
 		mgr->payload_mask = 0;
 		set_bit(0, &mgr->payload_mask);
+		mgr->vcpi_mask = 0;
 	}
 
 out_unlock:
@@ -2071,6 +2108,7 @@ static int drm_dp_mst_handle_up_req(struct drm_dp_mst_topology_mgr *mgr)
  * drm_dp_mst_hpd_irq() - MST hotplug IRQ notify
  * @mgr: manager to notify irq for.
  * @esi: 4 bytes from SINK_COUNT_ESI
+ * @handled: whether the hpd interrupt was consumed or not
  *
  * This should be called from the driver when it detects a short IRQ,
  * along with the value of the DEVICE_SERVICE_IRQ_VECTOR_ESI0. The
@@ -2474,7 +2512,7 @@ void drm_dp_mst_dump_topology(struct seq_file *m,
 	mutex_unlock(&mgr->lock);
 
 	mutex_lock(&mgr->payload_lock);
-	seq_printf(m, "vcpi: %lx\n", mgr->payload_mask);
+	seq_printf(m, "vcpi: %lx %lx\n", mgr->payload_mask, mgr->vcpi_mask);
 
 	for (i = 0; i < mgr->max_payloads; i++) {
 		if (mgr->proposed_vcpis[i]) {
diff --git a/drivers/gpu/drm/drm_drv.c b/drivers/gpu/drm/drm_drv.c
index 3242e208c0d0..bc3da32d4585 100644
--- a/drivers/gpu/drm/drm_drv.c
+++ b/drivers/gpu/drm/drm_drv.c
@@ -35,32 +35,20 @@
 #include <drm/drmP.h>
 #include <drm/drm_core.h>
 #include "drm_legacy.h"
+#include "drm_internal.h"
 
 unsigned int drm_debug = 0;	/* 1 to enable debug output */
 EXPORT_SYMBOL(drm_debug);
 
-unsigned int drm_vblank_offdelay = 5000;    /* Default to 5000 msecs. */
-
-unsigned int drm_timestamp_precision = 20;  /* Default to 20 usecs. */
-
-/*
- * Default to use monotonic timestamps for wait-for-vblank and page-flip
- * complete events.
- */
-unsigned int drm_timestamp_monotonic = 1;
-
 MODULE_AUTHOR(CORE_AUTHOR);
 MODULE_DESCRIPTION(CORE_DESC);
 MODULE_LICENSE("GPL and additional rights");
 MODULE_PARM_DESC(debug, "Enable debug output");
-MODULE_PARM_DESC(vblankoffdelay, "Delay until vblank irq auto-disable [msecs]");
+MODULE_PARM_DESC(vblankoffdelay, "Delay until vblank irq auto-disable [msecs] (0: never disable, <0: disable immediately)");
 MODULE_PARM_DESC(timestamp_precision_usec, "Max. error on timestamps [usecs]");
 MODULE_PARM_DESC(timestamp_monotonic, "Use monotonic timestamps");
 
 module_param_named(debug, drm_debug, int, 0600);
-module_param_named(vblankoffdelay, drm_vblank_offdelay, int, 0600);
-module_param_named(timestamp_precision_usec, drm_timestamp_precision, int, 0600);
-module_param_named(timestamp_monotonic, drm_timestamp_monotonic, int, 0600);
 
 static DEFINE_SPINLOCK(drm_minor_lock);
 static struct idr drm_minors_idr;
@@ -68,22 +56,19 @@ static struct idr drm_minors_idr;
 struct class *drm_class;
 static struct dentry *drm_debugfs_root;
 
-int drm_err(const char *func, const char *format, ...)
+void drm_err(const char *func, const char *format, ...)
 {
 	struct va_format vaf;
 	va_list args;
-	int r;
 
 	va_start(args, format);
 
 	vaf.fmt = format;
 	vaf.va = &args;
 
-	r = printk(KERN_ERR "[" DRM_NAME ":%s] *ERROR* %pV", func, &vaf);
+	printk(KERN_ERR "[" DRM_NAME ":%s] *ERROR* %pV", func, &vaf);
 
 	va_end(args);
-
-	return r;
 }
 EXPORT_SYMBOL(drm_err);
 
@@ -102,6 +87,8 @@ void drm_ut_debug_printk(const char *function_name, const char *format, ...)
 }
 EXPORT_SYMBOL(drm_ut_debug_printk);
 
+#define DRM_MAGIC_HASH_ORDER  4  /**< Size of key hash table. Must be power of 2. */
+
 struct drm_master *drm_master_create(struct drm_minor *minor)
 {
 	struct drm_master *master;
@@ -133,7 +120,6 @@ EXPORT_SYMBOL(drm_master_get);
 static void drm_master_destroy(struct kref *kref)
 {
 	struct drm_master *master = container_of(kref, struct drm_master, refcount);
-	struct drm_magic_entry *pt, *next;
 	struct drm_device *dev = master->minor->dev;
 	struct drm_map_list *r_list, *list_temp;
 
@@ -143,7 +129,7 @@ static void drm_master_destroy(struct kref *kref)
 
 	list_for_each_entry_safe(r_list, list_temp, &dev->maplist, head) {
 		if (r_list->master == master) {
-			drm_rmmap_locked(dev, r_list->map);
+			drm_legacy_rmmap_locked(dev, r_list->map);
 			r_list = NULL;
 		}
 	}
@@ -154,12 +140,6 @@ static void drm_master_destroy(struct kref *kref)
 		master->unique_len = 0;
 	}
 
-	list_for_each_entry_safe(pt, next, &master->magicfree, head) {
-		list_del(&pt->head);
-		drm_ht_remove_item(&master->magiclist, &pt->hash_item);
-		kfree(pt);
-	}
-
 	drm_ht_remove(&master->magiclist);
 
 	mutex_unlock(&dev->struct_mutex);
@@ -615,7 +595,7 @@ struct drm_device *drm_dev_alloc(struct drm_driver *driver,
 		goto err_ht;
 	}
 
-	if (driver->driver_features & DRIVER_GEM) {
+	if (drm_core_check_feature(dev, DRIVER_GEM)) {
 		ret = drm_gem_init(dev);
 		if (ret) {
 			DRM_ERROR("Cannot initialize graphics execution manager (GEM)\n");
@@ -645,7 +625,7 @@ static void drm_dev_release(struct kref *ref)
 {
 	struct drm_device *dev = container_of(ref, struct drm_device, ref);
 
-	if (dev->driver->driver_features & DRIVER_GEM)
+	if (drm_core_check_feature(dev, DRIVER_GEM))
 		drm_gem_destroy(dev);
 
 	drm_legacy_ctxbitmap_cleanup(dev);
@@ -779,7 +759,7 @@ void drm_dev_unregister(struct drm_device *dev)
 	drm_vblank_cleanup(dev);
 
 	list_for_each_entry_safe(r_list, list_temp, &dev->maplist, head)
-		drm_rmmap(dev, r_list->map);
+		drm_legacy_rmmap(dev, r_list->map);
 
 	drm_minor_unregister(dev, DRM_MINOR_LEGACY);
 	drm_minor_unregister(dev, DRM_MINOR_RENDER);
diff --git a/drivers/gpu/drm/drm_edid.c b/drivers/gpu/drm/drm_edid.c
index 1dbf3bc4c6a3..3bf999134bcc 100644
--- a/drivers/gpu/drm/drm_edid.c
+++ b/drivers/gpu/drm/drm_edid.c
@@ -632,27 +632,27 @@ static const struct drm_display_mode edid_cea_modes[] = {
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC |
 			DRM_MODE_FLAG_INTERLACE),
 	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 6 - 1440x480i@60Hz */
-	{ DRM_MODE("1440x480i", DRM_MODE_TYPE_DRIVER, 27000, 1440, 1478,
-		   1602, 1716, 0, 480, 488, 494, 525, 0,
+	/* 6 - 720(1440)x480i@60Hz */
+	{ DRM_MODE("720x480i", DRM_MODE_TYPE_DRIVER, 13500, 720, 739,
+		   801, 858, 0, 480, 488, 494, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 7 - 1440x480i@60Hz */
-	{ DRM_MODE("1440x480i", DRM_MODE_TYPE_DRIVER, 27000, 1440, 1478,
-		   1602, 1716, 0, 480, 488, 494, 525, 0,
+	/* 7 - 720(1440)x480i@60Hz */
+	{ DRM_MODE("720x480i", DRM_MODE_TYPE_DRIVER, 13500, 720, 739,
+		   801, 858, 0, 480, 488, 494, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 8 - 1440x240@60Hz */
-	{ DRM_MODE("1440x240", DRM_MODE_TYPE_DRIVER, 27000, 1440, 1478,
-		   1602, 1716, 0, 240, 244, 247, 262, 0,
+	/* 8 - 720(1440)x240@60Hz */
+	{ DRM_MODE("720x240", DRM_MODE_TYPE_DRIVER, 13500, 720, 739,
+		   801, 858, 0, 240, 244, 247, 262, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 9 - 1440x240@60Hz */
-	{ DRM_MODE("1440x240", DRM_MODE_TYPE_DRIVER, 27000, 1440, 1478,
-		   1602, 1716, 0, 240, 244, 247, 262, 0,
+	/* 9 - 720(1440)x240@60Hz */
+	{ DRM_MODE("720x240", DRM_MODE_TYPE_DRIVER, 13500, 720, 739,
+		   801, 858, 0, 240, 244, 247, 262, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
@@ -714,27 +714,27 @@ static const struct drm_display_mode edid_cea_modes[] = {
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC |
 			DRM_MODE_FLAG_INTERLACE),
 	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 21 - 1440x576i@50Hz */
-	{ DRM_MODE("1440x576i", DRM_MODE_TYPE_DRIVER, 27000, 1440, 1464,
-		   1590, 1728, 0, 576, 580, 586, 625, 0,
+	/* 21 - 720(1440)x576i@50Hz */
+	{ DRM_MODE("720x576i", DRM_MODE_TYPE_DRIVER, 13500, 720, 732,
+		   795, 864, 0, 576, 580, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 22 - 1440x576i@50Hz */
-	{ DRM_MODE("1440x576i", DRM_MODE_TYPE_DRIVER, 27000, 1440, 1464,
-		   1590, 1728, 0, 576, 580, 586, 625, 0,
+	/* 22 - 720(1440)x576i@50Hz */
+	{ DRM_MODE("720x576i", DRM_MODE_TYPE_DRIVER, 13500, 720, 732,
+		   795, 864, 0, 576, 580, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 23 - 1440x288@50Hz */
-	{ DRM_MODE("1440x288", DRM_MODE_TYPE_DRIVER, 27000, 1440, 1464,
-		   1590, 1728, 0, 288, 290, 293, 312, 0,
+	/* 23 - 720(1440)x288@50Hz */
+	{ DRM_MODE("720x288", DRM_MODE_TYPE_DRIVER, 13500, 720, 732,
+		   795, 864, 0, 288, 290, 293, 312, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 24 - 1440x288@50Hz */
-	{ DRM_MODE("1440x288", DRM_MODE_TYPE_DRIVER, 27000, 1440, 1464,
-		   1590, 1728, 0, 288, 290, 293, 312, 0,
+	/* 24 - 720(1440)x288@50Hz */
+	{ DRM_MODE("720x288", DRM_MODE_TYPE_DRIVER, 13500, 720, 732,
+		   795, 864, 0, 288, 290, 293, 312, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
@@ -837,17 +837,17 @@ static const struct drm_display_mode edid_cea_modes[] = {
 		   796, 864, 0, 576, 581, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
 	  .vrefresh = 100, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 44 - 1440x576i@100Hz */
-	{ DRM_MODE("1440x576", DRM_MODE_TYPE_DRIVER, 54000, 1440, 1464,
-		   1590, 1728, 0, 576, 580, 586, 625, 0,
+	/* 44 - 720(1440)x576i@100Hz */
+	{ DRM_MODE("720x576i", DRM_MODE_TYPE_DRIVER, 27000, 720, 732,
+		   795, 864, 0, 576, 580, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
-			DRM_MODE_FLAG_DBLCLK),
+			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 100, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 45 - 1440x576i@100Hz */
-	{ DRM_MODE("1440x576", DRM_MODE_TYPE_DRIVER, 54000, 1440, 1464,
-		   1590, 1728, 0, 576, 580, 586, 625, 0,
+	/* 45 - 720(1440)x576i@100Hz */
+	{ DRM_MODE("720x576i", DRM_MODE_TYPE_DRIVER, 27000, 720, 732,
+		   795, 864, 0, 576, 580, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
-			DRM_MODE_FLAG_DBLCLK),
+			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 100, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 46 - 1920x1080i@120Hz */
 	{ DRM_MODE("1920x1080i", DRM_MODE_TYPE_DRIVER, 148500, 1920, 2008,
@@ -870,15 +870,15 @@ static const struct drm_display_mode edid_cea_modes[] = {
 		   798, 858, 0, 480, 489, 495, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
 	  .vrefresh = 120, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 50 - 1440x480i@120Hz */
-	{ DRM_MODE("1440x480i", DRM_MODE_TYPE_DRIVER, 54000, 1440, 1478,
-		   1602, 1716, 0, 480, 488, 494, 525, 0,
+	/* 50 - 720(1440)x480i@120Hz */
+	{ DRM_MODE("720x480i", DRM_MODE_TYPE_DRIVER, 27000, 720, 739,
+		   801, 858, 0, 480, 488, 494, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 120, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 51 - 1440x480i@120Hz */
-	{ DRM_MODE("1440x480i", DRM_MODE_TYPE_DRIVER, 54000, 1440, 1478,
-		   1602, 1716, 0, 480, 488, 494, 525, 0,
+	/* 51 - 720(1440)x480i@120Hz */
+	{ DRM_MODE("720x480i", DRM_MODE_TYPE_DRIVER, 27000, 720, 739,
+		   801, 858, 0, 480, 488, 494, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 120, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
@@ -892,15 +892,15 @@ static const struct drm_display_mode edid_cea_modes[] = {
 		   796, 864, 0, 576, 581, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
 	  .vrefresh = 200, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 54 - 1440x576i@200Hz */
-	{ DRM_MODE("1440x576i", DRM_MODE_TYPE_DRIVER, 108000, 1440, 1464,
-		   1590, 1728, 0, 576, 580, 586, 625, 0,
+	/* 54 - 720(1440)x576i@200Hz */
+	{ DRM_MODE("720x576i", DRM_MODE_TYPE_DRIVER, 54000, 720, 732,
+		   795, 864, 0, 576, 580, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 200, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 55 - 1440x576i@200Hz */
-	{ DRM_MODE("1440x576i", DRM_MODE_TYPE_DRIVER, 108000, 1440, 1464,
-		   1590, 1728, 0, 576, 580, 586, 625, 0,
+	/* 55 - 720(1440)x576i@200Hz */
+	{ DRM_MODE("720x576i", DRM_MODE_TYPE_DRIVER, 54000, 720, 732,
+		   795, 864, 0, 576, 580, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 200, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
@@ -914,15 +914,15 @@ static const struct drm_display_mode edid_cea_modes[] = {
 		   798, 858, 0, 480, 489, 495, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
 	  .vrefresh = 240, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 58 - 1440x480i@240 */
-	{ DRM_MODE("1440x480i", DRM_MODE_TYPE_DRIVER, 108000, 1440, 1478,
-		   1602, 1716, 0, 480, 488, 494, 525, 0,
+	/* 58 - 720(1440)x480i@240 */
+	{ DRM_MODE("720x480i", DRM_MODE_TYPE_DRIVER, 54000, 720, 739,
+		   801, 858, 0, 480, 488, 494, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 240, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 59 - 1440x480i@240 */
-	{ DRM_MODE("1440x480i", DRM_MODE_TYPE_DRIVER, 108000, 1440, 1478,
-		   1602, 1716, 0, 480, 488, 494, 525, 0,
+	/* 59 - 720(1440)x480i@240 */
+	{ DRM_MODE("720x480i", DRM_MODE_TYPE_DRIVER, 54000, 720, 739,
+		   801, 858, 0, 480, 488, 494, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 240, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
@@ -2103,7 +2103,8 @@ static int
 add_inferred_modes(struct drm_connector *connector, struct edid *edid)
 {
 	struct detailed_mode_closure closure = {
-		connector, edid, 0, 0, 0
+		.connector = connector,
+		.edid = edid,
 	};
 
 	if (version_greater(edid, 1, 0))
@@ -2169,7 +2170,8 @@ add_established_modes(struct drm_connector *connector, struct edid *edid)
 		((edid->established_timings.mfg_rsvd & 0x80) << 9);
 	int i, modes = 0;
 	struct detailed_mode_closure closure = {
-		connector, edid, 0, 0, 0
+		.connector = connector,
+		.edid = edid,
 	};
 
 	for (i = 0; i <= EDID_EST_TIMINGS; i++) {
@@ -2227,7 +2229,8 @@ add_standard_modes(struct drm_connector *connector, struct edid *edid)
 {
 	int i, modes = 0;
 	struct detailed_mode_closure closure = {
-		connector, edid, 0, 0, 0
+		.connector = connector,
+		.edid = edid,
 	};
 
 	for (i = 0; i < EDID_STD_TIMINGS; i++) {
@@ -2313,7 +2316,8 @@ static int
 add_cvt_modes(struct drm_connector *connector, struct edid *edid)
 {	
 	struct detailed_mode_closure closure = {
-		connector, edid, 0, 0, 0
+		.connector = connector,
+		.edid = edid,
 	};
 
 	if (version_greater(edid, 1, 2))
@@ -2357,11 +2361,10 @@ add_detailed_modes(struct drm_connector *connector, struct edid *edid,
 		   u32 quirks)
 {
 	struct detailed_mode_closure closure = {
-		connector,
-		edid,
-		1,
-		quirks,
-		0
+		.connector = connector,
+		.edid = edid,
+		.preferred = 1,
+		.quirks = quirks,
 	};
 
 	if (closure.preferred && !version_greater(edid, 1, 3))
@@ -3433,10 +3436,10 @@ EXPORT_SYMBOL(drm_rgb_quant_range_selectable);
 /**
  * drm_assign_hdmi_deep_color_info - detect whether monitor supports
  * hdmi deep color modes and update drm_display_info if so.
- *
  * @edid: monitor EDID information
  * @info: Updated with maximum supported deep color bpc and color format
  *        if deep color supported.
+ * @connector: DRM connector, used only for debug output
  *
  * Parse the CEA extension according to CEA-861-B.
  * Return true if HDMI deep color supported, false if not or unknown.
diff --git a/drivers/gpu/drm/drm_fb_helper.c b/drivers/gpu/drm/drm_fb_helper.c
index 3144db9dc0f1..0c0c39bac23d 100644
--- a/drivers/gpu/drm/drm_fb_helper.c
+++ b/drivers/gpu/drm/drm_fb_helper.c
@@ -126,7 +126,7 @@ int drm_fb_helper_add_one_connector(struct drm_fb_helper *fb_helper, struct drm_
 
 	WARN_ON(!mutex_is_locked(&fb_helper->dev->mode_config.mutex));
 	if (fb_helper->connector_count + 1 > fb_helper->connector_info_alloc_count) {
-		temp = krealloc(fb_helper->connector_info, sizeof(struct drm_fb_helper_connector) * (fb_helper->connector_count + 1), GFP_KERNEL);
+		temp = krealloc(fb_helper->connector_info, sizeof(struct drm_fb_helper_connector *) * (fb_helper->connector_count + 1), GFP_KERNEL);
 		if (!temp)
 			return -ENOMEM;
 
@@ -171,60 +171,6 @@ int drm_fb_helper_remove_one_connector(struct drm_fb_helper *fb_helper,
 }
 EXPORT_SYMBOL(drm_fb_helper_remove_one_connector);
 
-static int drm_fb_helper_parse_command_line(struct drm_fb_helper *fb_helper)
-{
-	struct drm_fb_helper_connector *fb_helper_conn;
-	int i;
-
-	for (i = 0; i < fb_helper->connector_count; i++) {
-		struct drm_cmdline_mode *mode;
-		struct drm_connector *connector;
-		char *option = NULL;
-
-		fb_helper_conn = fb_helper->connector_info[i];
-		connector = fb_helper_conn->connector;
-		mode = &fb_helper_conn->cmdline_mode;
-
-		/* do something on return - turn off connector maybe */
-		if (fb_get_options(connector->name, &option))
-			continue;
-
-		if (drm_mode_parse_command_line_for_connector(option,
-							      connector,
-							      mode)) {
-			if (mode->force) {
-				const char *s;
-				switch (mode->force) {
-				case DRM_FORCE_OFF:
-					s = "OFF";
-					break;
-				case DRM_FORCE_ON_DIGITAL:
-					s = "ON - dig";
-					break;
-				default:
-				case DRM_FORCE_ON:
-					s = "ON";
-					break;
-				}
-
-				DRM_INFO("forcing %s connector %s\n",
-					 connector->name, s);
-				connector->force = mode->force;
-			}
-
-			DRM_DEBUG_KMS("cmdline mode for connector %s %dx%d@%dHz%s%s%s\n",
-				      connector->name,
-				      mode->xres, mode->yres,
-				      mode->refresh_specified ? mode->refresh : 60,
-				      mode->rb ? " reduced blanking" : "",
-				      mode->margins ? " with margins" : "",
-				      mode->interlace ?  " interlaced" : "");
-		}
-
-	}
-	return 0;
-}
-
 static void drm_fb_helper_save_lut_atomic(struct drm_crtc *crtc, struct drm_fb_helper *helper)
 {
 	uint16_t *r_base, *g_base, *b_base;
@@ -345,10 +291,17 @@ static bool restore_fbdev_mode(struct drm_fb_helper *fb_helper)
 
 	drm_warn_on_modeset_not_all_locked(dev);
 
-	list_for_each_entry(plane, &dev->mode_config.plane_list, head)
+	list_for_each_entry(plane, &dev->mode_config.plane_list, head) {
 		if (plane->type != DRM_PLANE_TYPE_PRIMARY)
 			drm_plane_force_disable(plane);
 
+		if (dev->mode_config.rotation_property) {
+			drm_mode_plane_set_obj_prop(plane,
+						    dev->mode_config.rotation_property,
+						    BIT(DRM_ROTATE_0));
+		}
+	}
+
 	for (i = 0; i < fb_helper->crtc_count; i++) {
 		struct drm_mode_set *mode_set = &fb_helper->crtc_info[i].mode_set;
 		struct drm_crtc *crtc = mode_set->crtc;
@@ -419,11 +372,11 @@ static bool drm_fb_helper_force_kernel_mode(void)
 		if (dev->switch_power_state == DRM_SWITCH_POWER_OFF)
 			continue;
 
-		/* NOTE: we use lockless flag below to avoid grabbing other
-		 * modeset locks.  So just trylock the underlying mutex
-		 * directly:
+		/*
+		 * NOTE: Use trylock mode to avoid deadlocks and sleeping in
+		 * panic context.
 		 */
-		if (!mutex_trylock(&dev->mode_config.mutex)) {
+		if (__drm_modeset_lock_all(dev, true) != 0) {
 			error = true;
 			continue;
 		}
@@ -432,7 +385,7 @@ static bool drm_fb_helper_force_kernel_mode(void)
 		if (ret)
 			error = true;
 
-		mutex_unlock(&dev->mode_config.mutex);
+		drm_modeset_unlock_all(dev);
 	}
 	return error;
 }
@@ -1013,7 +966,7 @@ static int drm_fb_helper_single_fb_probe(struct drm_fb_helper *fb_helper,
 		struct drm_fb_helper_connector *fb_helper_conn = fb_helper->connector_info[i];
 		struct drm_cmdline_mode *cmdline_mode;
 
-		cmdline_mode = &fb_helper_conn->cmdline_mode;
+		cmdline_mode = &fb_helper_conn->connector->cmdline_mode;
 
 		if (cmdline_mode->bpp_specified) {
 			switch (cmdline_mode->bpp) {
@@ -1260,9 +1213,7 @@ EXPORT_SYMBOL(drm_has_preferred_mode);
 
 static bool drm_has_cmdline_mode(struct drm_fb_helper_connector *fb_connector)
 {
-	struct drm_cmdline_mode *cmdline_mode;
-	cmdline_mode = &fb_connector->cmdline_mode;
-	return cmdline_mode->specified;
+	return fb_connector->connector->cmdline_mode.specified;
 }
 
 struct drm_display_mode *drm_pick_cmdline_mode(struct drm_fb_helper_connector *fb_helper_conn,
@@ -1272,7 +1223,7 @@ struct drm_display_mode *drm_pick_cmdline_mode(struct drm_fb_helper_connector *f
 	struct drm_display_mode *mode = NULL;
 	bool prefer_non_interlace;
 
-	cmdline_mode = &fb_helper_conn->cmdline_mode;
+	cmdline_mode = &fb_helper_conn->connector->cmdline_mode;
 	if (cmdline_mode->specified == false)
 		return mode;
 
@@ -1657,8 +1608,6 @@ bool drm_fb_helper_initial_config(struct drm_fb_helper *fb_helper, int bpp_sel)
 	struct drm_device *dev = fb_helper->dev;
 	int count = 0;
 
-	drm_fb_helper_parse_command_line(fb_helper);
-
 	mutex_lock(&dev->mode_config.mutex);
 	count = drm_fb_helper_probe_connector_modes(fb_helper,
 						    dev->mode_config.max_width,
diff --git a/drivers/gpu/drm/drm_fops.c b/drivers/gpu/drm/drm_fops.c
index 79d5221c6e41..ed7bc68f7e87 100644
--- a/drivers/gpu/drm/drm_fops.c
+++ b/drivers/gpu/drm/drm_fops.c
@@ -39,10 +39,10 @@
 #include <linux/slab.h>
 #include <linux/module.h>
 #include "drm_legacy.h"
+#include "drm_internal.h"
 
 /* from BKL pushdown */
 DEFINE_MUTEX(drm_global_mutex);
-EXPORT_SYMBOL(drm_global_mutex);
 
 static int drm_open_helper(struct file *filp, struct drm_minor *minor);
 
@@ -171,7 +171,7 @@ static int drm_open_helper(struct file *filp, struct drm_minor *minor)
 	init_waitqueue_head(&priv->event_wait);
 	priv->event_space = 4096; /* set aside 4k for event buffer */
 
-	if (dev->driver->driver_features & DRIVER_GEM)
+	if (drm_core_check_feature(dev, DRIVER_GEM))
 		drm_gem_open(dev, priv);
 
 	if (drm_core_check_feature(dev, DRIVER_PRIME))
@@ -256,7 +256,7 @@ out_close:
 out_prime_destroy:
 	if (drm_core_check_feature(dev, DRIVER_PRIME))
 		drm_prime_destroy_file_private(&priv->prime);
-	if (dev->driver->driver_features & DRIVER_GEM)
+	if (drm_core_check_feature(dev, DRIVER_GEM))
 		drm_gem_release(dev, priv);
 	put_pid(priv->pid);
 	kfree(priv);
@@ -268,11 +268,11 @@ static void drm_master_release(struct drm_device *dev, struct file *filp)
 {
 	struct drm_file *file_priv = filp->private_data;
 
-	if (drm_i_have_hw_lock(dev, file_priv)) {
+	if (drm_legacy_i_have_hw_lock(dev, file_priv)) {
 		DRM_DEBUG("File %p released, freeing lock for context %d\n",
 			  filp, _DRM_LOCKING_CONTEXT(file_priv->master->lock.hw_lock->lock));
-		drm_lock_free(&file_priv->master->lock,
-			      _DRM_LOCKING_CONTEXT(file_priv->master->lock.hw_lock->lock));
+		drm_legacy_lock_free(&file_priv->master->lock,
+				     _DRM_LOCKING_CONTEXT(file_priv->master->lock.hw_lock->lock));
 	}
 }
 
@@ -330,8 +330,6 @@ static void drm_legacy_dev_reinit(struct drm_device *dev)
  */
 int drm_lastclose(struct drm_device * dev)
 {
-	struct drm_vma_entry *vma, *vma_temp;
-
 	DRM_DEBUG("\n");
 
 	if (dev->driver->lastclose)
@@ -346,13 +344,7 @@ int drm_lastclose(struct drm_device * dev)
 	drm_agp_clear(dev);
 
 	drm_legacy_sg_cleanup(dev);
-
-	/* Clear vma list (only built for debugging) */
-	list_for_each_entry_safe(vma, vma_temp, &dev->vmalist, head) {
-		list_del(&vma->head);
-		kfree(vma);
-	}
-
+	drm_legacy_vma_flush(dev);
 	drm_legacy_dma_takedown(dev);
 
 	mutex_unlock(&dev->struct_mutex);
@@ -412,14 +404,14 @@ int drm_release(struct inode *inode, struct file *filp)
 		drm_master_release(dev, filp);
 
 	if (drm_core_check_feature(dev, DRIVER_HAVE_DMA))
-		drm_core_reclaim_buffers(dev, file_priv);
+		drm_legacy_reclaim_buffers(dev, file_priv);
 
 	drm_events_release(file_priv);
 
-	if (dev->driver->driver_features & DRIVER_MODESET)
+	if (drm_core_check_feature(dev, DRIVER_MODESET))
 		drm_fb_release(file_priv);
 
-	if (dev->driver->driver_features & DRIVER_GEM)
+	if (drm_core_check_feature(dev, DRIVER_GEM))
 		drm_gem_release(dev, file_priv);
 
 	drm_legacy_ctxbitmap_flush(dev, file_priv);
@@ -464,6 +456,8 @@ int drm_release(struct inode *inode, struct file *filp)
 	if (drm_core_check_feature(dev, DRIVER_PRIME))
 		drm_prime_destroy_file_private(&file_priv->prime);
 
+	WARN_ON(!list_empty(&file_priv->event_list));
+
 	put_pid(file_priv->pid);
 	kfree(file_priv);
 
diff --git a/drivers/gpu/drm/drm_gem.c b/drivers/gpu/drm/drm_gem.c
index 6adee4c2afc0..f6ca51259fa3 100644
--- a/drivers/gpu/drm/drm_gem.c
+++ b/drivers/gpu/drm/drm_gem.c
@@ -38,6 +38,8 @@
 #include <linux/dma-buf.h>
 #include <drm/drmP.h>
 #include <drm/drm_vma_manager.h>
+#include <drm/drm_gem.h>
+#include "drm_internal.h"
 
 /** @file drm_gem.c
  *
@@ -146,7 +148,7 @@ int drm_gem_object_init(struct drm_device *dev,
 EXPORT_SYMBOL(drm_gem_object_init);
 
 /**
- * drm_gem_object_init - initialize an allocated private GEM object
+ * drm_gem_private_object_init - initialize an allocated private GEM object
  * @dev: drm_device the object should be initialized for
  * @obj: drm_gem_object to initialize
  * @size: object size
@@ -579,7 +581,7 @@ drm_gem_close_ioctl(struct drm_device *dev, void *data,
 	struct drm_gem_close *args = data;
 	int ret;
 
-	if (!(dev->driver->driver_features & DRIVER_GEM))
+	if (!drm_core_check_feature(dev, DRIVER_GEM))
 		return -ENODEV;
 
 	ret = drm_gem_handle_delete(file_priv, args->handle);
@@ -606,7 +608,7 @@ drm_gem_flink_ioctl(struct drm_device *dev, void *data,
 	struct drm_gem_object *obj;
 	int ret;
 
-	if (!(dev->driver->driver_features & DRIVER_GEM))
+	if (!drm_core_check_feature(dev, DRIVER_GEM))
 		return -ENODEV;
 
 	obj = drm_gem_object_lookup(dev, file_priv, args->handle);
@@ -659,7 +661,7 @@ drm_gem_open_ioctl(struct drm_device *dev, void *data,
 	int ret;
 	u32 handle;
 
-	if (!(dev->driver->driver_features & DRIVER_GEM))
+	if (!drm_core_check_feature(dev, DRIVER_GEM))
 		return -ENODEV;
 
 	mutex_lock(&dev->object_name_lock);
@@ -887,7 +889,7 @@ int drm_gem_mmap(struct file *filp, struct vm_area_struct *vma)
 					   vma_pages(vma));
 	if (!node) {
 		mutex_unlock(&dev->struct_mutex);
-		return drm_mmap(filp, vma);
+		return -EINVAL;
 	} else if (!drm_vma_node_is_allowed(node, filp)) {
 		mutex_unlock(&dev->struct_mutex);
 		return -EACCES;
diff --git a/drivers/gpu/drm/drm_gem_cma_helper.c b/drivers/gpu/drm/drm_gem_cma_helper.c
index e467e67af6e7..0316310e2cc4 100644
--- a/drivers/gpu/drm/drm_gem_cma_helper.c
+++ b/drivers/gpu/drm/drm_gem_cma_helper.c
@@ -316,7 +316,8 @@ out:
 EXPORT_SYMBOL_GPL(drm_gem_cma_prime_get_sg_table);
 
 struct drm_gem_object *
-drm_gem_cma_prime_import_sg_table(struct drm_device *dev, size_t size,
+drm_gem_cma_prime_import_sg_table(struct drm_device *dev,
+				  struct dma_buf_attachment *attach,
 				  struct sg_table *sgt)
 {
 	struct drm_gem_cma_object *cma_obj;
@@ -325,14 +326,14 @@ drm_gem_cma_prime_import_sg_table(struct drm_device *dev, size_t size,
 		return ERR_PTR(-EINVAL);
 
 	/* Create a CMA GEM buffer. */
-	cma_obj = __drm_gem_cma_create(dev, size);
+	cma_obj = __drm_gem_cma_create(dev, attach->dmabuf->size);
 	if (IS_ERR(cma_obj))
 		return ERR_CAST(cma_obj);
 
 	cma_obj->paddr = sg_dma_address(sgt->sgl);
 	cma_obj->sgt = sgt;
 
-	DRM_DEBUG_PRIME("dma_addr = %pad, size = %zu\n", &cma_obj->paddr, size);
+	DRM_DEBUG_PRIME("dma_addr = %pad, size = %zu\n", &cma_obj->paddr, attach->dmabuf->size);
 
 	return &cma_obj->base;
 }
diff --git a/drivers/gpu/drm/drm_info.c b/drivers/gpu/drm/drm_info.c
index ecaf0fa2eec8..51efebd434f3 100644
--- a/drivers/gpu/drm/drm_info.c
+++ b/drivers/gpu/drm/drm_info.c
@@ -35,6 +35,9 @@
 
 #include <linux/seq_file.h>
 #include <drm/drmP.h>
+#include <drm/drm_gem.h>
+
+#include "drm_legacy.h"
 
 /**
  * Called when "/proc/dri/.../name" is read.
@@ -183,15 +186,32 @@ int drm_clients_info(struct seq_file *m, void *data)
 	struct drm_device *dev = node->minor->dev;
 	struct drm_file *priv;
 
+	seq_printf(m,
+		   "%20s %5s %3s master a %5s %10s\n",
+		   "command",
+		   "pid",
+		   "dev",
+		   "uid",
+		   "magic");
+
+	/* dev->filelist is sorted youngest first, but we want to present
+	 * oldest first (i.e. kernel, servers, clients), so walk backwardss.
+	 */
 	mutex_lock(&dev->struct_mutex);
-	seq_printf(m, "a dev	pid    uid	magic\n\n");
-	list_for_each_entry(priv, &dev->filelist, lhead) {
-		seq_printf(m, "%c %3d %5d %5d %10u\n",
-			   priv->authenticated ? 'y' : 'n',
-			   priv->minor->index,
+	list_for_each_entry_reverse(priv, &dev->filelist, lhead) {
+		struct task_struct *task;
+
+		rcu_read_lock(); /* locks pid_task()->comm */
+		task = pid_task(priv->pid, PIDTYPE_PID);
+		seq_printf(m, "%20s %5d %3d   %c    %c %5d %10u\n",
+			   task ? task->comm : "<unknown>",
 			   pid_vnr(priv->pid),
+			   priv->minor->index,
+			   priv->is_master ? 'y' : 'n',
+			   priv->authenticated ? 'y' : 'n',
 			   from_kuid_munged(seq_user_ns(m), priv->uid),
 			   priv->magic);
+		rcu_read_unlock();
 	}
 	mutex_unlock(&dev->struct_mutex);
 	return 0;
@@ -223,62 +243,3 @@ int drm_gem_name_info(struct seq_file *m, void *data)
 
 	return 0;
 }
-
-#if DRM_DEBUG_CODE
-
-int drm_vma_info(struct seq_file *m, void *data)
-{
-	struct drm_info_node *node = (struct drm_info_node *) m->private;
-	struct drm_device *dev = node->minor->dev;
-	struct drm_vma_entry *pt;
-	struct vm_area_struct *vma;
-	unsigned long vma_count = 0;
-#if defined(__i386__)
-	unsigned int pgprot;
-#endif
-
-	mutex_lock(&dev->struct_mutex);
-	list_for_each_entry(pt, &dev->vmalist, head)
-		vma_count++;
-
-	seq_printf(m, "vma use count: %lu, high_memory = %pK, 0x%pK\n",
-		   vma_count, high_memory,
-		   (void *)(unsigned long)virt_to_phys(high_memory));
-
-	list_for_each_entry(pt, &dev->vmalist, head) {
-		vma = pt->vma;
-		if (!vma)
-			continue;
-		seq_printf(m,
-			   "\n%5d 0x%pK-0x%pK %c%c%c%c%c%c 0x%08lx000",
-			   pt->pid,
-			   (void *)vma->vm_start, (void *)vma->vm_end,
-			   vma->vm_flags & VM_READ ? 'r' : '-',
-			   vma->vm_flags & VM_WRITE ? 'w' : '-',
-			   vma->vm_flags & VM_EXEC ? 'x' : '-',
-			   vma->vm_flags & VM_MAYSHARE ? 's' : 'p',
-			   vma->vm_flags & VM_LOCKED ? 'l' : '-',
-			   vma->vm_flags & VM_IO ? 'i' : '-',
-			   vma->vm_pgoff);
-
-#if defined(__i386__)
-		pgprot = pgprot_val(vma->vm_page_prot);
-		seq_printf(m, " %c%c%c%c%c%c%c%c%c",
-			   pgprot & _PAGE_PRESENT ? 'p' : '-',
-			   pgprot & _PAGE_RW ? 'w' : 'r',
-			   pgprot & _PAGE_USER ? 'u' : 's',
-			   pgprot & _PAGE_PWT ? 't' : 'b',
-			   pgprot & _PAGE_PCD ? 'u' : 'c',
-			   pgprot & _PAGE_ACCESSED ? 'a' : '-',
-			   pgprot & _PAGE_DIRTY ? 'd' : '-',
-			   pgprot & _PAGE_PSE ? 'm' : 'k',
-			   pgprot & _PAGE_GLOBAL ? 'g' : 'l');
-#endif
-		seq_printf(m, "\n");
-	}
-	mutex_unlock(&dev->struct_mutex);
-	return 0;
-}
-
-#endif
-
diff --git a/drivers/gpu/drm/drm_internal.h b/drivers/gpu/drm/drm_internal.h
new file mode 100644
index 000000000000..7cc0a3516871
--- /dev/null
+++ b/drivers/gpu/drm/drm_internal.h
@@ -0,0 +1,132 @@
+/*
+ * Copyright © 2014 Intel Corporation
+ *   Daniel Vetter <daniel.vetter@ffwll.ch>
+ *
+ * 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.
+ */
+
+/* drm_irq.c */
+extern unsigned int drm_timestamp_monotonic;
+
+/* drm_fops.c */
+extern struct mutex drm_global_mutex;
+int drm_lastclose(struct drm_device *dev);
+
+/* drm_pci.c */
+int drm_pci_set_unique(struct drm_device *dev,
+		       struct drm_master *master,
+		       struct drm_unique *u);
+int drm_irq_by_busid(struct drm_device *dev, void *data,
+		     struct drm_file *file_priv);
+
+/* drm_vm.c */
+int drm_vma_info(struct seq_file *m, void *data);
+void drm_vm_open_locked(struct drm_device *dev, struct vm_area_struct *vma);
+void drm_vm_close_locked(struct drm_device *dev, struct vm_area_struct *vma);
+
+/* drm_prime.c */
+int drm_prime_handle_to_fd_ioctl(struct drm_device *dev, void *data,
+				 struct drm_file *file_priv);
+int drm_prime_fd_to_handle_ioctl(struct drm_device *dev, void *data,
+				 struct drm_file *file_priv);
+
+void drm_prime_init_file_private(struct drm_prime_file_private *prime_fpriv);
+void drm_prime_destroy_file_private(struct drm_prime_file_private *prime_fpriv);
+void drm_prime_remove_buf_handle_locked(struct drm_prime_file_private *prime_fpriv,
+					struct dma_buf *dma_buf);
+
+/* drm_info.c */
+int drm_name_info(struct seq_file *m, void *data);
+int drm_vm_info(struct seq_file *m, void *data);
+int drm_bufs_info(struct seq_file *m, void *data);
+int drm_vblank_info(struct seq_file *m, void *data);
+int drm_clients_info(struct seq_file *m, void* data);
+int drm_gem_name_info(struct seq_file *m, void *data);
+
+/* drm_irq.c */
+int drm_control(struct drm_device *dev, void *data,
+		struct drm_file *file_priv);
+int drm_modeset_ctl(struct drm_device *dev, void *data,
+		    struct drm_file *file_priv);
+
+/* drm_auth.c */
+int drm_getmagic(struct drm_device *dev, void *data,
+		 struct drm_file *file_priv);
+int drm_authmagic(struct drm_device *dev, void *data,
+		  struct drm_file *file_priv);
+int drm_remove_magic(struct drm_master *master, drm_magic_t magic);
+
+/* drm_sysfs.c */
+extern struct class *drm_class;
+
+struct class *drm_sysfs_create(struct module *owner, char *name);
+void drm_sysfs_destroy(void);
+struct device *drm_sysfs_minor_alloc(struct drm_minor *minor);
+int drm_sysfs_connector_add(struct drm_connector *connector);
+void drm_sysfs_connector_remove(struct drm_connector *connector);
+
+/* drm_gem.c */
+int drm_gem_init(struct drm_device *dev);
+void drm_gem_destroy(struct drm_device *dev);
+int drm_gem_handle_create_tail(struct drm_file *file_priv,
+			       struct drm_gem_object *obj,
+			       u32 *handlep);
+int drm_gem_close_ioctl(struct drm_device *dev, void *data,
+			struct drm_file *file_priv);
+int drm_gem_flink_ioctl(struct drm_device *dev, void *data,
+			struct drm_file *file_priv);
+int drm_gem_open_ioctl(struct drm_device *dev, void *data,
+		       struct drm_file *file_priv);
+void drm_gem_open(struct drm_device *dev, struct drm_file *file_private);
+void drm_gem_release(struct drm_device *dev, struct drm_file *file_private);
+
+/* drm_drv.c */
+int drm_setmaster_ioctl(struct drm_device *dev, void *data,
+			struct drm_file *file_priv);
+int drm_dropmaster_ioctl(struct drm_device *dev, void *data,
+			 struct drm_file *file_priv);
+struct drm_master *drm_master_create(struct drm_minor *minor);
+
+/* drm_debugfs.c */
+#if defined(CONFIG_DEBUG_FS)
+int drm_debugfs_init(struct drm_minor *minor, int minor_id,
+		     struct dentry *root);
+int drm_debugfs_cleanup(struct drm_minor *minor);
+int drm_debugfs_connector_add(struct drm_connector *connector);
+void drm_debugfs_connector_remove(struct drm_connector *connector);
+#else
+static inline int drm_debugfs_init(struct drm_minor *minor, int minor_id,
+				   struct dentry *root)
+{
+	return 0;
+}
+
+static inline int drm_debugfs_cleanup(struct drm_minor *minor)
+{
+	return 0;
+}
+
+static inline int drm_debugfs_connector_add(struct drm_connector *connector)
+{
+	return 0;
+}
+static inline void drm_debugfs_connector_remove(struct drm_connector *connector)
+{
+}
+#endif
diff --git a/drivers/gpu/drm/drm_ioctl.c b/drivers/gpu/drm/drm_ioctl.c
index 40be746b7e68..00587a1e3c83 100644
--- a/drivers/gpu/drm/drm_ioctl.c
+++ b/drivers/gpu/drm/drm_ioctl.c
@@ -31,6 +31,7 @@
 #include <drm/drmP.h>
 #include <drm/drm_core.h>
 #include "drm_legacy.h"
+#include "drm_internal.h"
 
 #include <linux/pci.h>
 #include <linux/export.h>
@@ -41,121 +42,6 @@
 static int drm_version(struct drm_device *dev, void *data,
 		       struct drm_file *file_priv);
 
-#define DRM_IOCTL_DEF(ioctl, _func, _flags) \
-	[DRM_IOCTL_NR(ioctl)] = {.cmd = ioctl, .func = _func, .flags = _flags, .cmd_drv = 0, .name = #ioctl}
-
-/** Ioctl table */
-static const struct drm_ioctl_desc drm_ioctls[] = {
-	DRM_IOCTL_DEF(DRM_IOCTL_VERSION, drm_version, DRM_UNLOCKED|DRM_RENDER_ALLOW),
-	DRM_IOCTL_DEF(DRM_IOCTL_GET_UNIQUE, drm_getunique, 0),
-	DRM_IOCTL_DEF(DRM_IOCTL_GET_MAGIC, drm_getmagic, 0),
-	DRM_IOCTL_DEF(DRM_IOCTL_IRQ_BUSID, drm_irq_by_busid, DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_GET_MAP, drm_getmap, DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_GET_CLIENT, drm_getclient, DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_GET_STATS, drm_getstats, DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_GET_CAP, drm_getcap, DRM_UNLOCKED|DRM_RENDER_ALLOW),
-	DRM_IOCTL_DEF(DRM_IOCTL_SET_CLIENT_CAP, drm_setclientcap, 0),
-	DRM_IOCTL_DEF(DRM_IOCTL_SET_VERSION, drm_setversion, DRM_MASTER),
-
-	DRM_IOCTL_DEF(DRM_IOCTL_SET_UNIQUE, drm_setunique, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_BLOCK, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_UNBLOCK, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_AUTH_MAGIC, drm_authmagic, DRM_AUTH|DRM_MASTER),
-
-	DRM_IOCTL_DEF(DRM_IOCTL_ADD_MAP, drm_addmap_ioctl, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_RM_MAP, drm_rmmap_ioctl, DRM_AUTH),
-
-	DRM_IOCTL_DEF(DRM_IOCTL_SET_SAREA_CTX, drm_legacy_setsareactx, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_GET_SAREA_CTX, drm_legacy_getsareactx, DRM_AUTH),
-
-	DRM_IOCTL_DEF(DRM_IOCTL_SET_MASTER, drm_setmaster_ioctl, DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_DROP_MASTER, drm_dropmaster_ioctl, DRM_ROOT_ONLY),
-
-	DRM_IOCTL_DEF(DRM_IOCTL_ADD_CTX, drm_legacy_addctx, DRM_AUTH|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_RM_CTX, drm_legacy_rmctx, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_MOD_CTX, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_GET_CTX, drm_legacy_getctx, DRM_AUTH),
-	DRM_IOCTL_DEF(DRM_IOCTL_SWITCH_CTX, drm_legacy_switchctx, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_NEW_CTX, drm_legacy_newctx, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_RES_CTX, drm_legacy_resctx, DRM_AUTH),
-
-	DRM_IOCTL_DEF(DRM_IOCTL_ADD_DRAW, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_RM_DRAW, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-
-	DRM_IOCTL_DEF(DRM_IOCTL_LOCK, drm_lock, DRM_AUTH),
-	DRM_IOCTL_DEF(DRM_IOCTL_UNLOCK, drm_unlock, DRM_AUTH),
-
-	DRM_IOCTL_DEF(DRM_IOCTL_FINISH, drm_noop, DRM_AUTH),
-
-	DRM_IOCTL_DEF(DRM_IOCTL_ADD_BUFS, drm_addbufs, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_MARK_BUFS, drm_markbufs, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_INFO_BUFS, drm_infobufs, DRM_AUTH),
-	DRM_IOCTL_DEF(DRM_IOCTL_MAP_BUFS, drm_mapbufs, DRM_AUTH),
-	DRM_IOCTL_DEF(DRM_IOCTL_FREE_BUFS, drm_freebufs, DRM_AUTH),
-	DRM_IOCTL_DEF(DRM_IOCTL_DMA, drm_dma_ioctl, DRM_AUTH),
-
-	DRM_IOCTL_DEF(DRM_IOCTL_CONTROL, drm_control, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-
-#if __OS_HAS_AGP
-	DRM_IOCTL_DEF(DRM_IOCTL_AGP_ACQUIRE, drm_agp_acquire_ioctl, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_AGP_RELEASE, drm_agp_release_ioctl, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_AGP_ENABLE, drm_agp_enable_ioctl, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_AGP_INFO, drm_agp_info_ioctl, DRM_AUTH),
-	DRM_IOCTL_DEF(DRM_IOCTL_AGP_ALLOC, drm_agp_alloc_ioctl, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_AGP_FREE, drm_agp_free_ioctl, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_AGP_BIND, drm_agp_bind_ioctl, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_AGP_UNBIND, drm_agp_unbind_ioctl, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-#endif
-
-	DRM_IOCTL_DEF(DRM_IOCTL_SG_ALLOC, drm_sg_alloc, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-	DRM_IOCTL_DEF(DRM_IOCTL_SG_FREE, drm_sg_free, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-
-	DRM_IOCTL_DEF(DRM_IOCTL_WAIT_VBLANK, drm_wait_vblank, DRM_UNLOCKED),
-
-	DRM_IOCTL_DEF(DRM_IOCTL_MODESET_CTL, drm_modeset_ctl, 0),
-
-	DRM_IOCTL_DEF(DRM_IOCTL_UPDATE_DRAW, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
-
-	DRM_IOCTL_DEF(DRM_IOCTL_GEM_CLOSE, drm_gem_close_ioctl, DRM_UNLOCKED|DRM_RENDER_ALLOW),
-	DRM_IOCTL_DEF(DRM_IOCTL_GEM_FLINK, drm_gem_flink_ioctl, DRM_AUTH|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_GEM_OPEN, drm_gem_open_ioctl, DRM_AUTH|DRM_UNLOCKED),
-
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETRESOURCES, drm_mode_getresources, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-
-	DRM_IOCTL_DEF(DRM_IOCTL_PRIME_HANDLE_TO_FD, drm_prime_handle_to_fd_ioctl, DRM_AUTH|DRM_UNLOCKED|DRM_RENDER_ALLOW),
-	DRM_IOCTL_DEF(DRM_IOCTL_PRIME_FD_TO_HANDLE, drm_prime_fd_to_handle_ioctl, DRM_AUTH|DRM_UNLOCKED|DRM_RENDER_ALLOW),
-
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETPLANERESOURCES, drm_mode_getplane_res, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETCRTC, drm_mode_getcrtc, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_SETCRTC, drm_mode_setcrtc, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETPLANE, drm_mode_getplane, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_SETPLANE, drm_mode_setplane, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_CURSOR, drm_mode_cursor_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETGAMMA, drm_mode_gamma_get_ioctl, DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_SETGAMMA, drm_mode_gamma_set_ioctl, DRM_MASTER|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETENCODER, drm_mode_getencoder, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETCONNECTOR, drm_mode_getconnector, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_ATTACHMODE, drm_noop, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_DETACHMODE, drm_noop, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETPROPERTY, drm_mode_getproperty_ioctl, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_SETPROPERTY, drm_mode_connector_property_set_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETPROPBLOB, drm_mode_getblob_ioctl, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETFB, drm_mode_getfb, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_ADDFB, drm_mode_addfb, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_ADDFB2, drm_mode_addfb2, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_RMFB, drm_mode_rmfb, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_PAGE_FLIP, drm_mode_page_flip_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_DIRTYFB, drm_mode_dirtyfb_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_CREATE_DUMB, drm_mode_create_dumb_ioctl, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_MAP_DUMB, drm_mode_mmap_dumb_ioctl, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_DESTROY_DUMB, drm_mode_destroy_dumb_ioctl, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_OBJ_GETPROPERTIES, drm_mode_obj_get_properties_ioctl, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_OBJ_SETPROPERTY, drm_mode_obj_set_property_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_CURSOR2, drm_mode_cursor2_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-};
-
-#define DRM_CORE_IOCTL_COUNT	ARRAY_SIZE( drm_ioctls )
-
 /**
  * Get the bus id.
  *
@@ -167,7 +53,7 @@ static const struct drm_ioctl_desc drm_ioctls[] = {
  *
  * Copies the bus id from drm_device::unique into user space.
  */
-int drm_getunique(struct drm_device *dev, void *data,
+static int drm_getunique(struct drm_device *dev, void *data,
 		  struct drm_file *file_priv)
 {
 	struct drm_unique *u = data;
@@ -189,7 +75,6 @@ drm_unset_busid(struct drm_device *dev,
 	kfree(master->unique);
 	master->unique = NULL;
 	master->unique_len = 0;
-	master->unique_size = 0;
 }
 
 /**
@@ -207,7 +92,7 @@ drm_unset_busid(struct drm_device *dev,
  * version 1.1 or greater. Also note that KMS is all version 1.1 and later and
  * UMS was only ever supported on pci devices.
  */
-int drm_setunique(struct drm_device *dev, void *data,
+static int drm_setunique(struct drm_device *dev, void *data,
 		  struct drm_file *file_priv)
 {
 	struct drm_unique *u = data;
@@ -245,15 +130,15 @@ static int drm_set_busid(struct drm_device *dev, struct drm_file *file_priv)
 	if (master->unique != NULL)
 		drm_unset_busid(dev, master);
 
-	if (dev->driver->bus && dev->driver->bus->set_busid) {
-		ret = dev->driver->bus->set_busid(dev, master);
+	if (dev->driver->set_busid) {
+		ret = dev->driver->set_busid(dev, master);
 		if (ret) {
 			drm_unset_busid(dev, master);
 			return ret;
 		}
 	} else {
 		if (WARN(dev->unique == NULL,
-			 "No drm_bus.set_busid() implementation provided by "
+			 "No drm_driver.set_busid() implementation provided by "
 			 "%ps. Use drm_dev_set_unique() to set the unique "
 			 "name explicitly.", dev->driver))
 			return -EINVAL;
@@ -279,7 +164,7 @@ static int drm_set_busid(struct drm_device *dev, struct drm_file *file_priv)
  * Searches for the mapping with the specified offset and copies its information
  * into userspace
  */
-int drm_getmap(struct drm_device *dev, void *data,
+static int drm_getmap(struct drm_device *dev, void *data,
 	       struct drm_file *file_priv)
 {
 	struct drm_map *map = data;
@@ -340,7 +225,7 @@ int drm_getmap(struct drm_device *dev, void *data,
  * Searches for the client with the specified index and copies its information
  * into userspace
  */
-int drm_getclient(struct drm_device *dev, void *data,
+static int drm_getclient(struct drm_device *dev, void *data,
 		  struct drm_file *file_priv)
 {
 	struct drm_client *client = data;
@@ -380,7 +265,7 @@ int drm_getclient(struct drm_device *dev, void *data,
  *
  * \return zero on success or a negative number on failure.
  */
-int drm_getstats(struct drm_device *dev, void *data,
+static int drm_getstats(struct drm_device *dev, void *data,
 		 struct drm_file *file_priv)
 {
 	struct drm_stats *stats = data;
@@ -394,7 +279,7 @@ int drm_getstats(struct drm_device *dev, void *data,
 /**
  * Get device/driver capabilities
  */
-int drm_getcap(struct drm_device *dev, void *data, struct drm_file *file_priv)
+static int drm_getcap(struct drm_device *dev, void *data, struct drm_file *file_priv)
 {
 	struct drm_get_cap *req = data;
 
@@ -444,7 +329,7 @@ int drm_getcap(struct drm_device *dev, void *data, struct drm_file *file_priv)
 /**
  * Set device/driver capabilities
  */
-int
+static int
 drm_setclientcap(struct drm_device *dev, void *data, struct drm_file *file_priv)
 {
 	struct drm_set_client_cap *req = data;
@@ -478,7 +363,7 @@ drm_setclientcap(struct drm_device *dev, void *data, struct drm_file *file_priv)
  *
  * Sets the requested interface version
  */
-int drm_setversion(struct drm_device *dev, void *data, struct drm_file *file_priv)
+static int drm_setversion(struct drm_device *dev, void *data, struct drm_file *file_priv)
 {
 	struct drm_set_version *sv = data;
 	int if_version, retcode = 0;
@@ -624,6 +509,121 @@ static int drm_ioctl_permit(u32 flags, struct drm_file *file_priv)
 	return 0;
 }
 
+#define DRM_IOCTL_DEF(ioctl, _func, _flags) \
+	[DRM_IOCTL_NR(ioctl)] = {.cmd = ioctl, .func = _func, .flags = _flags, .cmd_drv = 0, .name = #ioctl}
+
+/** Ioctl table */
+static const struct drm_ioctl_desc drm_ioctls[] = {
+	DRM_IOCTL_DEF(DRM_IOCTL_VERSION, drm_version, DRM_UNLOCKED|DRM_RENDER_ALLOW),
+	DRM_IOCTL_DEF(DRM_IOCTL_GET_UNIQUE, drm_getunique, 0),
+	DRM_IOCTL_DEF(DRM_IOCTL_GET_MAGIC, drm_getmagic, 0),
+	DRM_IOCTL_DEF(DRM_IOCTL_IRQ_BUSID, drm_irq_by_busid, DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_GET_MAP, drm_getmap, DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_GET_CLIENT, drm_getclient, DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_GET_STATS, drm_getstats, DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_GET_CAP, drm_getcap, DRM_UNLOCKED|DRM_RENDER_ALLOW),
+	DRM_IOCTL_DEF(DRM_IOCTL_SET_CLIENT_CAP, drm_setclientcap, 0),
+	DRM_IOCTL_DEF(DRM_IOCTL_SET_VERSION, drm_setversion, DRM_MASTER),
+
+	DRM_IOCTL_DEF(DRM_IOCTL_SET_UNIQUE, drm_setunique, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_BLOCK, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_UNBLOCK, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_AUTH_MAGIC, drm_authmagic, DRM_AUTH|DRM_MASTER),
+
+	DRM_IOCTL_DEF(DRM_IOCTL_ADD_MAP, drm_legacy_addmap_ioctl, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_RM_MAP, drm_legacy_rmmap_ioctl, DRM_AUTH),
+
+	DRM_IOCTL_DEF(DRM_IOCTL_SET_SAREA_CTX, drm_legacy_setsareactx, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_GET_SAREA_CTX, drm_legacy_getsareactx, DRM_AUTH),
+
+	DRM_IOCTL_DEF(DRM_IOCTL_SET_MASTER, drm_setmaster_ioctl, DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_DROP_MASTER, drm_dropmaster_ioctl, DRM_ROOT_ONLY),
+
+	DRM_IOCTL_DEF(DRM_IOCTL_ADD_CTX, drm_legacy_addctx, DRM_AUTH|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_RM_CTX, drm_legacy_rmctx, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_MOD_CTX, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_GET_CTX, drm_legacy_getctx, DRM_AUTH),
+	DRM_IOCTL_DEF(DRM_IOCTL_SWITCH_CTX, drm_legacy_switchctx, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_NEW_CTX, drm_legacy_newctx, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_RES_CTX, drm_legacy_resctx, DRM_AUTH),
+
+	DRM_IOCTL_DEF(DRM_IOCTL_ADD_DRAW, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_RM_DRAW, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+
+	DRM_IOCTL_DEF(DRM_IOCTL_LOCK, drm_legacy_lock, DRM_AUTH),
+	DRM_IOCTL_DEF(DRM_IOCTL_UNLOCK, drm_legacy_unlock, DRM_AUTH),
+
+	DRM_IOCTL_DEF(DRM_IOCTL_FINISH, drm_noop, DRM_AUTH),
+
+	DRM_IOCTL_DEF(DRM_IOCTL_ADD_BUFS, drm_legacy_addbufs, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_MARK_BUFS, drm_legacy_markbufs, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_INFO_BUFS, drm_legacy_infobufs, DRM_AUTH),
+	DRM_IOCTL_DEF(DRM_IOCTL_MAP_BUFS, drm_legacy_mapbufs, DRM_AUTH),
+	DRM_IOCTL_DEF(DRM_IOCTL_FREE_BUFS, drm_legacy_freebufs, DRM_AUTH),
+	DRM_IOCTL_DEF(DRM_IOCTL_DMA, drm_legacy_dma_ioctl, DRM_AUTH),
+
+	DRM_IOCTL_DEF(DRM_IOCTL_CONTROL, drm_control, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+
+#if __OS_HAS_AGP
+	DRM_IOCTL_DEF(DRM_IOCTL_AGP_ACQUIRE, drm_agp_acquire_ioctl, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_AGP_RELEASE, drm_agp_release_ioctl, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_AGP_ENABLE, drm_agp_enable_ioctl, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_AGP_INFO, drm_agp_info_ioctl, DRM_AUTH),
+	DRM_IOCTL_DEF(DRM_IOCTL_AGP_ALLOC, drm_agp_alloc_ioctl, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_AGP_FREE, drm_agp_free_ioctl, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_AGP_BIND, drm_agp_bind_ioctl, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_AGP_UNBIND, drm_agp_unbind_ioctl, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+#endif
+
+	DRM_IOCTL_DEF(DRM_IOCTL_SG_ALLOC, drm_legacy_sg_alloc, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF(DRM_IOCTL_SG_FREE, drm_legacy_sg_free, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+
+	DRM_IOCTL_DEF(DRM_IOCTL_WAIT_VBLANK, drm_wait_vblank, DRM_UNLOCKED),
+
+	DRM_IOCTL_DEF(DRM_IOCTL_MODESET_CTL, drm_modeset_ctl, 0),
+
+	DRM_IOCTL_DEF(DRM_IOCTL_UPDATE_DRAW, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+
+	DRM_IOCTL_DEF(DRM_IOCTL_GEM_CLOSE, drm_gem_close_ioctl, DRM_UNLOCKED|DRM_RENDER_ALLOW),
+	DRM_IOCTL_DEF(DRM_IOCTL_GEM_FLINK, drm_gem_flink_ioctl, DRM_AUTH|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_GEM_OPEN, drm_gem_open_ioctl, DRM_AUTH|DRM_UNLOCKED),
+
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETRESOURCES, drm_mode_getresources, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+
+	DRM_IOCTL_DEF(DRM_IOCTL_PRIME_HANDLE_TO_FD, drm_prime_handle_to_fd_ioctl, DRM_AUTH|DRM_UNLOCKED|DRM_RENDER_ALLOW),
+	DRM_IOCTL_DEF(DRM_IOCTL_PRIME_FD_TO_HANDLE, drm_prime_fd_to_handle_ioctl, DRM_AUTH|DRM_UNLOCKED|DRM_RENDER_ALLOW),
+
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETPLANERESOURCES, drm_mode_getplane_res, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETCRTC, drm_mode_getcrtc, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_SETCRTC, drm_mode_setcrtc, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETPLANE, drm_mode_getplane, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_SETPLANE, drm_mode_setplane, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_CURSOR, drm_mode_cursor_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETGAMMA, drm_mode_gamma_get_ioctl, DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_SETGAMMA, drm_mode_gamma_set_ioctl, DRM_MASTER|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETENCODER, drm_mode_getencoder, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETCONNECTOR, drm_mode_getconnector, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_ATTACHMODE, drm_noop, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_DETACHMODE, drm_noop, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETPROPERTY, drm_mode_getproperty_ioctl, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_SETPROPERTY, drm_mode_connector_property_set_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETPROPBLOB, drm_mode_getblob_ioctl, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETFB, drm_mode_getfb, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_ADDFB, drm_mode_addfb, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_ADDFB2, drm_mode_addfb2, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_RMFB, drm_mode_rmfb, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_PAGE_FLIP, drm_mode_page_flip_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_DIRTYFB, drm_mode_dirtyfb_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_CREATE_DUMB, drm_mode_create_dumb_ioctl, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_MAP_DUMB, drm_mode_mmap_dumb_ioctl, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_DESTROY_DUMB, drm_mode_destroy_dumb_ioctl, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_OBJ_GETPROPERTIES, drm_mode_obj_get_properties_ioctl, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_OBJ_SETPROPERTY, drm_mode_obj_set_property_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_CURSOR2, drm_mode_cursor2_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+};
+
+#define DRM_CORE_IOCTL_COUNT	ARRAY_SIZE( drm_ioctls )
+
 /**
  * Called whenever a process performs an ioctl on /dev/drm.
  *
diff --git a/drivers/gpu/drm/drm_irq.c b/drivers/gpu/drm/drm_irq.c
index 08ba1209228e..5ef03c216a27 100644
--- a/drivers/gpu/drm/drm_irq.c
+++ b/drivers/gpu/drm/drm_irq.c
@@ -34,6 +34,7 @@
 
 #include <drm/drmP.h>
 #include "drm_trace.h"
+#include "drm_internal.h"
 
 #include <linux/interrupt.h>	/* For task queue support */
 #include <linux/slab.h>
@@ -55,12 +56,91 @@
  */
 #define DRM_REDUNDANT_VBLIRQ_THRESH_NS 1000000
 
+static bool
+drm_get_last_vbltimestamp(struct drm_device *dev, int crtc,
+			  struct timeval *tvblank, unsigned flags);
+
+static unsigned int drm_timestamp_precision = 20;  /* Default to 20 usecs. */
+
 /*
- * Clear vblank timestamp buffer for a crtc.
+ * Default to use monotonic timestamps for wait-for-vblank and page-flip
+ * complete events.
+ */
+unsigned int drm_timestamp_monotonic = 1;
+
+static int drm_vblank_offdelay = 5000;    /* Default to 5000 msecs. */
+
+module_param_named(vblankoffdelay, drm_vblank_offdelay, int, 0600);
+module_param_named(timestamp_precision_usec, drm_timestamp_precision, int, 0600);
+module_param_named(timestamp_monotonic, drm_timestamp_monotonic, int, 0600);
+
+/**
+ * drm_update_vblank_count - update the master vblank counter
+ * @dev: DRM device
+ * @crtc: counter to update
+ *
+ * Call back into the driver to update the appropriate vblank counter
+ * (specified by @crtc).  Deal with wraparound, if it occurred, and
+ * update the last read value so we can deal with wraparound on the next
+ * call if necessary.
+ *
+ * Only necessary when going from off->on, to account for frames we
+ * didn't get an interrupt for.
+ *
+ * Note: caller must hold dev->vbl_lock since this reads & writes
+ * device vblank fields.
  */
-static void clear_vblank_timestamps(struct drm_device *dev, int crtc)
+static void drm_update_vblank_count(struct drm_device *dev, int crtc)
 {
-	memset(dev->vblank[crtc].time, 0, sizeof(dev->vblank[crtc].time));
+	struct drm_vblank_crtc *vblank = &dev->vblank[crtc];
+	u32 cur_vblank, diff, tslot;
+	bool rc;
+	struct timeval t_vblank;
+
+	/*
+	 * Interrupts were disabled prior to this call, so deal with counter
+	 * wrap if needed.
+	 * NOTE!  It's possible we lost a full dev->max_vblank_count events
+	 * here if the register is small or we had vblank interrupts off for
+	 * a long time.
+	 *
+	 * We repeat the hardware vblank counter & timestamp query until
+	 * we get consistent results. This to prevent races between gpu
+	 * updating its hardware counter while we are retrieving the
+	 * corresponding vblank timestamp.
+	 */
+	do {
+		cur_vblank = dev->driver->get_vblank_counter(dev, crtc);
+		rc = drm_get_last_vbltimestamp(dev, crtc, &t_vblank, 0);
+	} while (cur_vblank != dev->driver->get_vblank_counter(dev, crtc));
+
+	/* Deal with counter wrap */
+	diff = cur_vblank - vblank->last;
+	if (cur_vblank < vblank->last) {
+		diff += dev->max_vblank_count;
+
+		DRM_DEBUG("last_vblank[%d]=0x%x, cur_vblank=0x%x => diff=0x%x\n",
+			  crtc, vblank->last, cur_vblank, diff);
+	}
+
+	DRM_DEBUG("updating vblank count on crtc %d, missed %d\n",
+		  crtc, diff);
+
+	if (diff == 0)
+		return;
+
+	/* Reinitialize corresponding vblank timestamp if high-precision query
+	 * available. Skip this step if query unsupported or failed. Will
+	 * reinitialize delayed at next vblank interrupt in that case.
+	 */
+	if (rc) {
+		tslot = atomic_read(&vblank->count) + diff;
+		vblanktimestamp(dev, crtc, tslot) = t_vblank;
+	}
+
+	smp_mb__before_atomic();
+	atomic_add(diff, &vblank->count);
+	smp_mb__after_atomic();
 }
 
 /*
@@ -71,10 +151,11 @@ static void clear_vblank_timestamps(struct drm_device *dev, int crtc)
  */
 static void vblank_disable_and_save(struct drm_device *dev, int crtc)
 {
+	struct drm_vblank_crtc *vblank = &dev->vblank[crtc];
 	unsigned long irqflags;
 	u32 vblcount;
 	s64 diff_ns;
-	int vblrc;
+	bool vblrc;
 	struct timeval tvblank;
 	int count = DRM_TIMESTAMP_MAXRETRIES;
 
@@ -84,8 +165,28 @@ static void vblank_disable_and_save(struct drm_device *dev, int crtc)
 	 */
 	spin_lock_irqsave(&dev->vblank_time_lock, irqflags);
 
+	/*
+	 * If the vblank interrupt was already disbled update the count
+	 * and timestamp to maintain the appearance that the counter
+	 * has been ticking all along until this time. This makes the
+	 * count account for the entire time between drm_vblank_on() and
+	 * drm_vblank_off().
+	 *
+	 * But only do this if precise vblank timestamps are available.
+	 * Otherwise we might read a totally bogus timestamp since drivers
+	 * lacking precise timestamp support rely upon sampling the system clock
+	 * at vblank interrupt time. Which obviously won't work out well if the
+	 * vblank interrupt is disabled.
+	 */
+	if (!vblank->enabled &&
+	    drm_get_last_vbltimestamp(dev, crtc, &tvblank, 0)) {
+		drm_update_vblank_count(dev, crtc);
+		spin_unlock_irqrestore(&dev->vblank_time_lock, irqflags);
+		return;
+	}
+
 	dev->driver->disable_vblank(dev, crtc);
-	dev->vblank[crtc].enabled = false;
+	vblank->enabled = false;
 
 	/* No further vblank irq's will be processed after
 	 * this point. Get current hardware vblank count and
@@ -100,9 +201,9 @@ static void vblank_disable_and_save(struct drm_device *dev, int crtc)
 	 * delayed gpu counter increment.
 	 */
 	do {
-		dev->vblank[crtc].last = dev->driver->get_vblank_counter(dev, crtc);
+		vblank->last = dev->driver->get_vblank_counter(dev, crtc);
 		vblrc = drm_get_last_vbltimestamp(dev, crtc, &tvblank, 0);
-	} while (dev->vblank[crtc].last != dev->driver->get_vblank_counter(dev, crtc) && (--count) && vblrc);
+	} while (vblank->last != dev->driver->get_vblank_counter(dev, crtc) && (--count) && vblrc);
 
 	if (!count)
 		vblrc = 0;
@@ -110,7 +211,7 @@ static void vblank_disable_and_save(struct drm_device *dev, int crtc)
 	/* Compute time difference to stored timestamp of last vblank
 	 * as updated by last invocation of drm_handle_vblank() in vblank irq.
 	 */
-	vblcount = atomic_read(&dev->vblank[crtc].count);
+	vblcount = atomic_read(&vblank->count);
 	diff_ns = timeval_to_ns(&tvblank) -
 		  timeval_to_ns(&vblanktimestamp(dev, crtc, vblcount));
 
@@ -126,14 +227,18 @@ static void vblank_disable_and_save(struct drm_device *dev, int crtc)
 	 * available. In that case we can't account for this and just
 	 * hope for the best.
 	 */
-	if ((vblrc > 0) && (abs64(diff_ns) > 1000000)) {
-		atomic_inc(&dev->vblank[crtc].count);
+	if (vblrc && (abs64(diff_ns) > 1000000)) {
+		/* Store new timestamp in ringbuffer. */
+		vblanktimestamp(dev, crtc, vblcount + 1) = tvblank;
+
+		/* Increment cooked vblank count. This also atomically commits
+		 * the timestamp computed above.
+		 */
+		smp_mb__before_atomic();
+		atomic_inc(&vblank->count);
 		smp_mb__after_atomic();
 	}
 
-	/* Invalidate all timestamps while vblank irq's are off. */
-	clear_vblank_timestamps(dev, crtc);
-
 	spin_unlock_irqrestore(&dev->vblank_time_lock, irqflags);
 }
 
@@ -164,14 +269,20 @@ static void vblank_disable_fn(unsigned long arg)
 void drm_vblank_cleanup(struct drm_device *dev)
 {
 	int crtc;
+	unsigned long irqflags;
 
 	/* Bail if the driver didn't call drm_vblank_init() */
 	if (dev->num_crtcs == 0)
 		return;
 
 	for (crtc = 0; crtc < dev->num_crtcs; crtc++) {
-		del_timer_sync(&dev->vblank[crtc].disable_timer);
-		vblank_disable_fn((unsigned long)&dev->vblank[crtc]);
+		struct drm_vblank_crtc *vblank = &dev->vblank[crtc];
+
+		del_timer_sync(&vblank->disable_timer);
+
+		spin_lock_irqsave(&dev->vbl_lock, irqflags);
+		vblank_disable_and_save(dev, crtc);
+		spin_unlock_irqrestore(&dev->vbl_lock, irqflags);
 	}
 
 	kfree(dev->vblank);
@@ -204,11 +315,13 @@ int drm_vblank_init(struct drm_device *dev, int num_crtcs)
 		goto err;
 
 	for (i = 0; i < num_crtcs; i++) {
-		dev->vblank[i].dev = dev;
-		dev->vblank[i].crtc = i;
-		init_waitqueue_head(&dev->vblank[i].queue);
-		setup_timer(&dev->vblank[i].disable_timer, vblank_disable_fn,
-			    (unsigned long)&dev->vblank[i]);
+		struct drm_vblank_crtc *vblank = &dev->vblank[i];
+
+		vblank->dev = dev;
+		vblank->crtc = i;
+		init_waitqueue_head(&vblank->queue);
+		setup_timer(&vblank->disable_timer, vblank_disable_fn,
+			    (unsigned long)vblank);
 	}
 
 	DRM_INFO("Supports vblank timestamp caching Rev 2 (21.10.2013).\n");
@@ -224,7 +337,7 @@ int drm_vblank_init(struct drm_device *dev, int num_crtcs)
 	return 0;
 
 err:
-	drm_vblank_cleanup(dev);
+	dev->num_crtcs = 0;
 	return ret;
 }
 EXPORT_SYMBOL(drm_vblank_init);
@@ -360,9 +473,11 @@ int drm_irq_uninstall(struct drm_device *dev)
 	if (dev->num_crtcs) {
 		spin_lock_irqsave(&dev->vbl_lock, irqflags);
 		for (i = 0; i < dev->num_crtcs; i++) {
-			wake_up(&dev->vblank[i].queue);
-			dev->vblank[i].enabled = false;
-			dev->vblank[i].last =
+			struct drm_vblank_crtc *vblank = &dev->vblank[i];
+
+			wake_up(&vblank->queue);
+			vblank->enabled = false;
+			vblank->last =
 				dev->driver->get_vblank_counter(dev, i);
 		}
 		spin_unlock_irqrestore(&dev->vbl_lock, irqflags);
@@ -617,7 +732,7 @@ int drm_calc_vbltimestamp_from_scanoutpos(struct drm_device *dev, int crtc,
 	 * within vblank area, counting down the number of lines until
 	 * start of scanout.
 	 */
-	invbl = vbl_status & DRM_SCANOUTPOS_INVBL;
+	invbl = vbl_status & DRM_SCANOUTPOS_IN_VBLANK;
 
 	/* Convert scanout position into elapsed time at raw_time query
 	 * since start of scanout at first display scanline. delta_ns
@@ -647,7 +762,7 @@ int drm_calc_vbltimestamp_from_scanoutpos(struct drm_device *dev, int crtc,
 
 	vbl_status = DRM_VBLANKTIME_SCANOUTPOS_METHOD;
 	if (invbl)
-		vbl_status |= DRM_VBLANKTIME_INVBL;
+		vbl_status |= DRM_VBLANKTIME_IN_VBLANK;
 
 	return vbl_status;
 }
@@ -679,10 +794,11 @@ static struct timeval get_drm_timestamp(void)
  * call, i.e., it isn't very precisely locked to the true vblank.
  *
  * Returns:
- * Non-zero if timestamp is considered to be very precise, zero otherwise.
+ * True if timestamp is considered to be very precise, false otherwise.
  */
-u32 drm_get_last_vbltimestamp(struct drm_device *dev, int crtc,
-			      struct timeval *tvblank, unsigned flags)
+static bool
+drm_get_last_vbltimestamp(struct drm_device *dev, int crtc,
+			  struct timeval *tvblank, unsigned flags)
 {
 	int ret;
 
@@ -694,7 +810,7 @@ u32 drm_get_last_vbltimestamp(struct drm_device *dev, int crtc,
 		ret = dev->driver->get_vblank_timestamp(dev, crtc, &max_error,
 							tvblank, flags);
 		if (ret > 0)
-			return (u32) ret;
+			return true;
 	}
 
 	/* GPU high precision timestamp query unsupported or failed.
@@ -702,9 +818,8 @@ u32 drm_get_last_vbltimestamp(struct drm_device *dev, int crtc,
 	 */
 	*tvblank = get_drm_timestamp();
 
-	return 0;
+	return false;
 }
-EXPORT_SYMBOL(drm_get_last_vbltimestamp);
 
 /**
  * drm_vblank_count - retrieve "cooked" vblank counter value
@@ -720,7 +835,11 @@ EXPORT_SYMBOL(drm_get_last_vbltimestamp);
  */
 u32 drm_vblank_count(struct drm_device *dev, int crtc)
 {
-	return atomic_read(&dev->vblank[crtc].count);
+	struct drm_vblank_crtc *vblank = &dev->vblank[crtc];
+
+	if (WARN_ON(crtc >= dev->num_crtcs))
+		return 0;
+	return atomic_read(&vblank->count);
 }
 EXPORT_SYMBOL(drm_vblank_count);
 
@@ -740,18 +859,22 @@ EXPORT_SYMBOL(drm_vblank_count);
 u32 drm_vblank_count_and_time(struct drm_device *dev, int crtc,
 			      struct timeval *vblanktime)
 {
+	struct drm_vblank_crtc *vblank = &dev->vblank[crtc];
 	u32 cur_vblank;
 
+	if (WARN_ON(crtc >= dev->num_crtcs))
+		return 0;
+
 	/* Read timestamp from slot of _vblank_time ringbuffer
 	 * that corresponds to current vblank count. Retry if
 	 * count has incremented during readout. This works like
 	 * a seqlock.
 	 */
 	do {
-		cur_vblank = atomic_read(&dev->vblank[crtc].count);
+		cur_vblank = atomic_read(&vblank->count);
 		*vblanktime = vblanktimestamp(dev, crtc, cur_vblank);
 		smp_rmb();
-	} while (cur_vblank != atomic_read(&dev->vblank[crtc].count));
+	} while (cur_vblank != atomic_read(&vblank->count));
 
 	return cur_vblank;
 }
@@ -800,83 +923,20 @@ void drm_send_vblank_event(struct drm_device *dev, int crtc,
 EXPORT_SYMBOL(drm_send_vblank_event);
 
 /**
- * drm_update_vblank_count - update the master vblank counter
- * @dev: DRM device
- * @crtc: counter to update
- *
- * Call back into the driver to update the appropriate vblank counter
- * (specified by @crtc).  Deal with wraparound, if it occurred, and
- * update the last read value so we can deal with wraparound on the next
- * call if necessary.
- *
- * Only necessary when going from off->on, to account for frames we
- * didn't get an interrupt for.
- *
- * Note: caller must hold dev->vbl_lock since this reads & writes
- * device vblank fields.
- */
-static void drm_update_vblank_count(struct drm_device *dev, int crtc)
-{
-	u32 cur_vblank, diff, tslot, rc;
-	struct timeval t_vblank;
-
-	/*
-	 * Interrupts were disabled prior to this call, so deal with counter
-	 * wrap if needed.
-	 * NOTE!  It's possible we lost a full dev->max_vblank_count events
-	 * here if the register is small or we had vblank interrupts off for
-	 * a long time.
-	 *
-	 * We repeat the hardware vblank counter & timestamp query until
-	 * we get consistent results. This to prevent races between gpu
-	 * updating its hardware counter while we are retrieving the
-	 * corresponding vblank timestamp.
-	 */
-	do {
-		cur_vblank = dev->driver->get_vblank_counter(dev, crtc);
-		rc = drm_get_last_vbltimestamp(dev, crtc, &t_vblank, 0);
-	} while (cur_vblank != dev->driver->get_vblank_counter(dev, crtc));
-
-	/* Deal with counter wrap */
-	diff = cur_vblank - dev->vblank[crtc].last;
-	if (cur_vblank < dev->vblank[crtc].last) {
-		diff += dev->max_vblank_count;
-
-		DRM_DEBUG("last_vblank[%d]=0x%x, cur_vblank=0x%x => diff=0x%x\n",
-			  crtc, dev->vblank[crtc].last, cur_vblank, diff);
-	}
-
-	DRM_DEBUG("enabling vblank interrupts on crtc %d, missed %d\n",
-		  crtc, diff);
-
-	/* Reinitialize corresponding vblank timestamp if high-precision query
-	 * available. Skip this step if query unsupported or failed. Will
-	 * reinitialize delayed at next vblank interrupt in that case.
-	 */
-	if (rc) {
-		tslot = atomic_read(&dev->vblank[crtc].count) + diff;
-		vblanktimestamp(dev, crtc, tslot) = t_vblank;
-	}
-
-	smp_mb__before_atomic();
-	atomic_add(diff, &dev->vblank[crtc].count);
-	smp_mb__after_atomic();
-}
-
-/**
  * drm_vblank_enable - enable the vblank interrupt on a CRTC
  * @dev: DRM device
  * @crtc: CRTC in question
  */
 static int drm_vblank_enable(struct drm_device *dev, int crtc)
 {
+	struct drm_vblank_crtc *vblank = &dev->vblank[crtc];
 	int ret = 0;
 
 	assert_spin_locked(&dev->vbl_lock);
 
 	spin_lock(&dev->vblank_time_lock);
 
-	if (!dev->vblank[crtc].enabled) {
+	if (!vblank->enabled) {
 		/*
 		 * Enable vblank irqs under vblank_time_lock protection.
 		 * All vblank count & timestamp updates are held off
@@ -887,9 +947,9 @@ static int drm_vblank_enable(struct drm_device *dev, int crtc)
 		ret = dev->driver->enable_vblank(dev, crtc);
 		DRM_DEBUG("enabling vblank on crtc %d, ret: %d\n", crtc, ret);
 		if (ret)
-			atomic_dec(&dev->vblank[crtc].refcount);
+			atomic_dec(&vblank->refcount);
 		else {
-			dev->vblank[crtc].enabled = true;
+			vblank->enabled = true;
 			drm_update_vblank_count(dev, crtc);
 		}
 	}
@@ -914,16 +974,20 @@ static int drm_vblank_enable(struct drm_device *dev, int crtc)
  */
 int drm_vblank_get(struct drm_device *dev, int crtc)
 {
+	struct drm_vblank_crtc *vblank = &dev->vblank[crtc];
 	unsigned long irqflags;
 	int ret = 0;
 
+	if (WARN_ON(crtc >= dev->num_crtcs))
+		return -EINVAL;
+
 	spin_lock_irqsave(&dev->vbl_lock, irqflags);
 	/* Going from 0->1 means we have to enable interrupts again */
-	if (atomic_add_return(1, &dev->vblank[crtc].refcount) == 1) {
+	if (atomic_add_return(1, &vblank->refcount) == 1) {
 		ret = drm_vblank_enable(dev, crtc);
 	} else {
-		if (!dev->vblank[crtc].enabled) {
-			atomic_dec(&dev->vblank[crtc].refcount);
+		if (!vblank->enabled) {
+			atomic_dec(&vblank->refcount);
 			ret = -EINVAL;
 		}
 	}
@@ -963,13 +1027,23 @@ EXPORT_SYMBOL(drm_crtc_vblank_get);
  */
 void drm_vblank_put(struct drm_device *dev, int crtc)
 {
-	BUG_ON(atomic_read(&dev->vblank[crtc].refcount) == 0);
+	struct drm_vblank_crtc *vblank = &dev->vblank[crtc];
+
+	BUG_ON(atomic_read(&vblank->refcount) == 0);
+
+	if (WARN_ON(crtc >= dev->num_crtcs))
+		return;
 
 	/* Last user schedules interrupt disable */
-	if (atomic_dec_and_test(&dev->vblank[crtc].refcount) &&
-	    (drm_vblank_offdelay > 0))
-		mod_timer(&dev->vblank[crtc].disable_timer,
-			  jiffies + ((drm_vblank_offdelay * HZ)/1000));
+	if (atomic_dec_and_test(&vblank->refcount)) {
+		if (drm_vblank_offdelay == 0)
+			return;
+		else if (dev->vblank_disable_immediate || drm_vblank_offdelay < 0)
+			vblank_disable_fn((unsigned long)vblank);
+		else
+			mod_timer(&vblank->disable_timer,
+				  jiffies + ((drm_vblank_offdelay * HZ)/1000));
+	}
 }
 EXPORT_SYMBOL(drm_vblank_put);
 
@@ -989,6 +1063,50 @@ void drm_crtc_vblank_put(struct drm_crtc *crtc)
 EXPORT_SYMBOL(drm_crtc_vblank_put);
 
 /**
+ * drm_wait_one_vblank - wait for one vblank
+ * @dev: DRM device
+ * @crtc: crtc index
+ *
+ * This waits for one vblank to pass on @crtc, using the irq driver interfaces.
+ * It is a failure to call this when the vblank irq for @crtc is disabled, e.g.
+ * due to lack of driver support or because the crtc is off.
+ */
+void drm_wait_one_vblank(struct drm_device *dev, int crtc)
+{
+	int ret;
+	u32 last;
+
+	ret = drm_vblank_get(dev, crtc);
+	if (WARN(ret, "vblank not available on crtc %i, ret=%i\n", crtc, ret))
+		return;
+
+	last = drm_vblank_count(dev, crtc);
+
+	ret = wait_event_timeout(dev->vblank[crtc].queue,
+				 last != drm_vblank_count(dev, crtc),
+				 msecs_to_jiffies(100));
+
+	WARN(ret == 0, "vblank wait timed out on crtc %i\n", crtc);
+
+	drm_vblank_put(dev, crtc);
+}
+EXPORT_SYMBOL(drm_wait_one_vblank);
+
+/**
+ * drm_crtc_wait_one_vblank - wait for one vblank
+ * @crtc: DRM crtc
+ *
+ * This waits for one vblank to pass on @crtc, using the irq driver interfaces.
+ * It is a failure to call this when the vblank irq for @crtc is disabled, e.g.
+ * due to lack of driver support or because the crtc is off.
+ */
+void drm_crtc_wait_one_vblank(struct drm_crtc *crtc)
+{
+	drm_wait_one_vblank(crtc->dev, drm_crtc_index(crtc));
+}
+EXPORT_SYMBOL(drm_crtc_wait_one_vblank);
+
+/**
  * drm_vblank_off - disable vblank events on a CRTC
  * @dev: DRM device
  * @crtc: CRTC in question
@@ -1004,19 +1122,34 @@ EXPORT_SYMBOL(drm_crtc_vblank_put);
  */
 void drm_vblank_off(struct drm_device *dev, int crtc)
 {
+	struct drm_vblank_crtc *vblank = &dev->vblank[crtc];
 	struct drm_pending_vblank_event *e, *t;
 	struct timeval now;
 	unsigned long irqflags;
 	unsigned int seq;
 
-	spin_lock_irqsave(&dev->vbl_lock, irqflags);
+	if (WARN_ON(crtc >= dev->num_crtcs))
+		return;
+
+	spin_lock_irqsave(&dev->event_lock, irqflags);
+
+	spin_lock(&dev->vbl_lock);
 	vblank_disable_and_save(dev, crtc);
-	wake_up(&dev->vblank[crtc].queue);
+	wake_up(&vblank->queue);
+
+	/*
+	 * Prevent subsequent drm_vblank_get() from re-enabling
+	 * the vblank interrupt by bumping the refcount.
+	 */
+	if (!vblank->inmodeset) {
+		atomic_inc(&vblank->refcount);
+		vblank->inmodeset = 1;
+	}
+	spin_unlock(&dev->vbl_lock);
 
 	/* Send any queued vblank events, lest the natives grow disquiet */
 	seq = drm_vblank_count_and_time(dev, crtc, &now);
 
-	spin_lock(&dev->event_lock);
 	list_for_each_entry_safe(e, t, &dev->vblank_event_list, base.link) {
 		if (e->pipe != crtc)
 			continue;
@@ -1027,9 +1160,7 @@ void drm_vblank_off(struct drm_device *dev, int crtc)
 		drm_vblank_put(dev, e->pipe);
 		send_vblank_event(dev, e, seq, &now);
 	}
-	spin_unlock(&dev->event_lock);
-
-	spin_unlock_irqrestore(&dev->vbl_lock, irqflags);
+	spin_unlock_irqrestore(&dev->event_lock, irqflags);
 }
 EXPORT_SYMBOL(drm_vblank_off);
 
@@ -1066,11 +1197,35 @@ EXPORT_SYMBOL(drm_crtc_vblank_off);
  */
 void drm_vblank_on(struct drm_device *dev, int crtc)
 {
+	struct drm_vblank_crtc *vblank = &dev->vblank[crtc];
 	unsigned long irqflags;
 
+	if (WARN_ON(crtc >= dev->num_crtcs))
+		return;
+
 	spin_lock_irqsave(&dev->vbl_lock, irqflags);
-	/* re-enable interrupts if there's are users left */
-	if (atomic_read(&dev->vblank[crtc].refcount) != 0)
+	/* Drop our private "prevent drm_vblank_get" refcount */
+	if (vblank->inmodeset) {
+		atomic_dec(&vblank->refcount);
+		vblank->inmodeset = 0;
+	}
+
+	/*
+	 * sample the current counter to avoid random jumps
+	 * when drm_vblank_enable() applies the diff
+	 *
+	 * -1 to make sure user will never see the same
+	 * vblank counter value before and after a modeset
+	 */
+	vblank->last =
+		(dev->driver->get_vblank_counter(dev, crtc) - 1) &
+		dev->max_vblank_count;
+	/*
+	 * re-enable interrupts if there are users left, or the
+	 * user wishes vblank interrupts to be enabled all the time.
+	 */
+	if (atomic_read(&vblank->refcount) != 0 ||
+	    (!dev->vblank_disable_immediate && drm_vblank_offdelay == 0))
 		WARN_ON(drm_vblank_enable(dev, crtc));
 	spin_unlock_irqrestore(&dev->vbl_lock, irqflags);
 }
@@ -1118,9 +1273,15 @@ EXPORT_SYMBOL(drm_crtc_vblank_on);
  */
 void drm_vblank_pre_modeset(struct drm_device *dev, int crtc)
 {
+	struct drm_vblank_crtc *vblank = &dev->vblank[crtc];
+
 	/* vblank is not initialized (IRQ not installed ?), or has been freed */
 	if (!dev->num_crtcs)
 		return;
+
+	if (WARN_ON(crtc >= dev->num_crtcs))
+		return;
+
 	/*
 	 * To avoid all the problems that might happen if interrupts
 	 * were enabled/disabled around or between these calls, we just
@@ -1128,10 +1289,10 @@ void drm_vblank_pre_modeset(struct drm_device *dev, int crtc)
 	 * to avoid corrupting the count if multiple, mismatch calls occur),
 	 * so that interrupts remain enabled in the interim.
 	 */
-	if (!dev->vblank[crtc].inmodeset) {
-		dev->vblank[crtc].inmodeset = 0x1;
+	if (!vblank->inmodeset) {
+		vblank->inmodeset = 0x1;
 		if (drm_vblank_get(dev, crtc) == 0)
-			dev->vblank[crtc].inmodeset |= 0x2;
+			vblank->inmodeset |= 0x2;
 	}
 }
 EXPORT_SYMBOL(drm_vblank_pre_modeset);
@@ -1146,21 +1307,22 @@ EXPORT_SYMBOL(drm_vblank_pre_modeset);
  */
 void drm_vblank_post_modeset(struct drm_device *dev, int crtc)
 {
+	struct drm_vblank_crtc *vblank = &dev->vblank[crtc];
 	unsigned long irqflags;
 
 	/* vblank is not initialized (IRQ not installed ?), or has been freed */
 	if (!dev->num_crtcs)
 		return;
 
-	if (dev->vblank[crtc].inmodeset) {
+	if (vblank->inmodeset) {
 		spin_lock_irqsave(&dev->vbl_lock, irqflags);
 		dev->vblank_disable_allowed = true;
 		spin_unlock_irqrestore(&dev->vbl_lock, irqflags);
 
-		if (dev->vblank[crtc].inmodeset & 0x2)
+		if (vblank->inmodeset & 0x2)
 			drm_vblank_put(dev, crtc);
 
-		dev->vblank[crtc].inmodeset = 0;
+		vblank->inmodeset = 0;
 	}
 }
 EXPORT_SYMBOL(drm_vblank_post_modeset);
@@ -1212,6 +1374,7 @@ static int drm_queue_vblank_event(struct drm_device *dev, int pipe,
 				  union drm_wait_vblank *vblwait,
 				  struct drm_file *file_priv)
 {
+	struct drm_vblank_crtc *vblank = &dev->vblank[pipe];
 	struct drm_pending_vblank_event *e;
 	struct timeval now;
 	unsigned long flags;
@@ -1235,6 +1398,18 @@ static int drm_queue_vblank_event(struct drm_device *dev, int pipe,
 
 	spin_lock_irqsave(&dev->event_lock, flags);
 
+	/*
+	 * drm_vblank_off() might have been called after we called
+	 * drm_vblank_get(). drm_vblank_off() holds event_lock
+	 * around the vblank disable, so no need for further locking.
+	 * The reference from drm_vblank_get() protects against
+	 * vblank disable from another source.
+	 */
+	if (!vblank->enabled) {
+		ret = -EINVAL;
+		goto err_unlock;
+	}
+
 	if (file_priv->event_space < sizeof e->event) {
 		ret = -EBUSY;
 		goto err_unlock;
@@ -1295,6 +1470,7 @@ err_put:
 int drm_wait_vblank(struct drm_device *dev, void *data,
 		    struct drm_file *file_priv)
 {
+	struct drm_vblank_crtc *vblank;
 	union drm_wait_vblank *vblwait = data;
 	int ret;
 	unsigned int flags, seq, crtc, high_crtc;
@@ -1324,6 +1500,8 @@ int drm_wait_vblank(struct drm_device *dev, void *data,
 	if (crtc >= dev->num_crtcs)
 		return -EINVAL;
 
+	vblank = &dev->vblank[crtc];
+
 	ret = drm_vblank_get(dev, crtc);
 	if (ret) {
 		DRM_DEBUG("failed to acquire vblank counter, %d\n", ret);
@@ -1356,11 +1534,11 @@ int drm_wait_vblank(struct drm_device *dev, void *data,
 
 	DRM_DEBUG("waiting on vblank count %d, crtc %d\n",
 		  vblwait->request.sequence, crtc);
-	dev->vblank[crtc].last_wait = vblwait->request.sequence;
-	DRM_WAIT_ON(ret, dev->vblank[crtc].queue, 3 * HZ,
+	vblank->last_wait = vblwait->request.sequence;
+	DRM_WAIT_ON(ret, vblank->queue, 3 * HZ,
 		    (((drm_vblank_count(dev, crtc) -
 		       vblwait->request.sequence) <= (1 << 23)) ||
-		     !dev->vblank[crtc].enabled ||
+		     !vblank->enabled ||
 		     !dev->irq_enabled));
 
 	if (ret != -EINTR) {
@@ -1385,12 +1563,11 @@ static void drm_handle_vblank_events(struct drm_device *dev, int crtc)
 {
 	struct drm_pending_vblank_event *e, *t;
 	struct timeval now;
-	unsigned long flags;
 	unsigned int seq;
 
-	seq = drm_vblank_count_and_time(dev, crtc, &now);
+	assert_spin_locked(&dev->event_lock);
 
-	spin_lock_irqsave(&dev->event_lock, flags);
+	seq = drm_vblank_count_and_time(dev, crtc, &now);
 
 	list_for_each_entry_safe(e, t, &dev->vblank_event_list, base.link) {
 		if (e->pipe != crtc)
@@ -1406,8 +1583,6 @@ static void drm_handle_vblank_events(struct drm_device *dev, int crtc)
 		send_vblank_event(dev, e, seq, &now);
 	}
 
-	spin_unlock_irqrestore(&dev->event_lock, flags);
-
 	trace_drm_vblank_event(crtc, seq);
 }
 
@@ -1421,6 +1596,7 @@ static void drm_handle_vblank_events(struct drm_device *dev, int crtc)
  */
 bool drm_handle_vblank(struct drm_device *dev, int crtc)
 {
+	struct drm_vblank_crtc *vblank = &dev->vblank[crtc];
 	u32 vblcount;
 	s64 diff_ns;
 	struct timeval tvblank;
@@ -1429,15 +1605,21 @@ bool drm_handle_vblank(struct drm_device *dev, int crtc)
 	if (!dev->num_crtcs)
 		return false;
 
+	if (WARN_ON(crtc >= dev->num_crtcs))
+		return false;
+
+	spin_lock_irqsave(&dev->event_lock, irqflags);
+
 	/* Need timestamp lock to prevent concurrent execution with
 	 * vblank enable/disable, as this would cause inconsistent
 	 * or corrupted timestamps and vblank counts.
 	 */
-	spin_lock_irqsave(&dev->vblank_time_lock, irqflags);
+	spin_lock(&dev->vblank_time_lock);
 
 	/* Vblank irq handling disabled. Nothing to do. */
-	if (!dev->vblank[crtc].enabled) {
-		spin_unlock_irqrestore(&dev->vblank_time_lock, irqflags);
+	if (!vblank->enabled) {
+		spin_unlock(&dev->vblank_time_lock);
+		spin_unlock_irqrestore(&dev->event_lock, irqflags);
 		return false;
 	}
 
@@ -1446,7 +1628,7 @@ bool drm_handle_vblank(struct drm_device *dev, int crtc)
 	 */
 
 	/* Get current timestamp and count. */
-	vblcount = atomic_read(&dev->vblank[crtc].count);
+	vblcount = atomic_read(&vblank->count);
 	drm_get_last_vbltimestamp(dev, crtc, &tvblank, DRM_CALLED_FROM_VBLIRQ);
 
 	/* Compute time difference to timestamp of last vblank */
@@ -1470,17 +1652,20 @@ bool drm_handle_vblank(struct drm_device *dev, int crtc)
 		 * the timestamp computed above.
 		 */
 		smp_mb__before_atomic();
-		atomic_inc(&dev->vblank[crtc].count);
+		atomic_inc(&vblank->count);
 		smp_mb__after_atomic();
 	} else {
 		DRM_DEBUG("crtc %d: Redundant vblirq ignored. diff_ns = %d\n",
 			  crtc, (int) diff_ns);
 	}
 
-	wake_up(&dev->vblank[crtc].queue);
+	spin_unlock(&dev->vblank_time_lock);
+
+	wake_up(&vblank->queue);
 	drm_handle_vblank_events(dev, crtc);
 
-	spin_unlock_irqrestore(&dev->vblank_time_lock, irqflags);
+	spin_unlock_irqrestore(&dev->event_lock, irqflags);
+
 	return true;
 }
 EXPORT_SYMBOL(drm_handle_vblank);
diff --git a/drivers/gpu/drm/drm_legacy.h b/drivers/gpu/drm/drm_legacy.h
index d34f20a79b7c..c1dc61473db5 100644
--- a/drivers/gpu/drm/drm_legacy.h
+++ b/drivers/gpu/drm/drm_legacy.h
@@ -23,6 +23,15 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  */
 
+/*
+ * This file contains legacy interfaces that modern drm drivers
+ * should no longer be using. They cannot be removed as legacy
+ * drivers use them, and removing them are API breaks.
+ */
+#include <linux/list.h>
+#include <drm/drm_legacy.h>
+
+struct agp_memory;
 struct drm_device;
 struct drm_file;
 
@@ -48,4 +57,57 @@ int drm_legacy_rmctx(struct drm_device *d, void *v, struct drm_file *f);
 int drm_legacy_setsareactx(struct drm_device *d, void *v, struct drm_file *f);
 int drm_legacy_getsareactx(struct drm_device *d, void *v, struct drm_file *f);
 
+/*
+ * Generic Buffer Management
+ */
+
+#define DRM_MAP_HASH_OFFSET 0x10000000
+
+int drm_legacy_addmap_ioctl(struct drm_device *d, void *v, struct drm_file *f);
+int drm_legacy_rmmap_ioctl(struct drm_device *d, void *v, struct drm_file *f);
+int drm_legacy_addbufs(struct drm_device *d, void *v, struct drm_file *f);
+int drm_legacy_infobufs(struct drm_device *d, void *v, struct drm_file *f);
+int drm_legacy_markbufs(struct drm_device *d, void *v, struct drm_file *f);
+int drm_legacy_freebufs(struct drm_device *d, void *v, struct drm_file *f);
+int drm_legacy_mapbufs(struct drm_device *d, void *v, struct drm_file *f);
+int drm_legacy_dma_ioctl(struct drm_device *d, void *v, struct drm_file *f);
+
+void drm_legacy_vma_flush(struct drm_device *d);
+
+/*
+ * AGP Support
+ */
+
+struct drm_agp_mem {
+	unsigned long handle;
+	struct agp_memory *memory;
+	unsigned long bound;
+	int pages;
+	struct list_head head;
+};
+
+/*
+ * Generic Userspace Locking-API
+ */
+
+int drm_legacy_i_have_hw_lock(struct drm_device *d, struct drm_file *f);
+int drm_legacy_lock(struct drm_device *d, void *v, struct drm_file *f);
+int drm_legacy_unlock(struct drm_device *d, void *v, struct drm_file *f);
+int drm_legacy_lock_free(struct drm_lock_data *lock, unsigned int ctx);
+
+/* DMA support */
+int drm_legacy_dma_setup(struct drm_device *dev);
+void drm_legacy_dma_takedown(struct drm_device *dev);
+void drm_legacy_free_buffer(struct drm_device *dev,
+			    struct drm_buf * buf);
+void drm_legacy_reclaim_buffers(struct drm_device *dev,
+				struct drm_file *filp);
+
+/* Scatter Gather Support */
+void drm_legacy_sg_cleanup(struct drm_device *dev);
+int drm_legacy_sg_alloc(struct drm_device *dev, void *data,
+			struct drm_file *file_priv);
+int drm_legacy_sg_free(struct drm_device *dev, void *data,
+		       struct drm_file *file_priv);
+
 #endif /* __DRM_LEGACY_H__ */
diff --git a/drivers/gpu/drm/drm_lock.c b/drivers/gpu/drm/drm_lock.c
index e26b59e385ff..f861361a635e 100644
--- a/drivers/gpu/drm/drm_lock.c
+++ b/drivers/gpu/drm/drm_lock.c
@@ -36,6 +36,7 @@
 #include <linux/export.h>
 #include <drm/drmP.h>
 #include "drm_legacy.h"
+#include "drm_internal.h"
 
 static int drm_notifier(void *priv);
 
@@ -52,7 +53,8 @@ static int drm_lock_take(struct drm_lock_data *lock_data, unsigned int context);
  *
  * Add the current task to the lock wait queue, and attempt to take to lock.
  */
-int drm_lock(struct drm_device *dev, void *data, struct drm_file *file_priv)
+int drm_legacy_lock(struct drm_device *dev, void *data,
+		    struct drm_file *file_priv)
 {
 	DECLARE_WAITQUEUE(entry, current);
 	struct drm_lock *lock = data;
@@ -120,7 +122,7 @@ int drm_lock(struct drm_device *dev, void *data, struct drm_file *file_priv)
 		sigaddset(&dev->sigmask, SIGTTOU);
 		dev->sigdata.context = lock->context;
 		dev->sigdata.lock = master->lock.hw_lock;
-		block_all_signals(drm_notifier, &dev->sigdata, &dev->sigmask);
+		block_all_signals(drm_notifier, dev, &dev->sigmask);
 	}
 
 	if (dev->driver->dma_quiescent && (lock->flags & _DRM_LOCK_QUIESCENT))
@@ -146,7 +148,7 @@ int drm_lock(struct drm_device *dev, void *data, struct drm_file *file_priv)
  *
  * Transfer and free the lock.
  */
-int drm_unlock(struct drm_device *dev, void *data, struct drm_file *file_priv)
+int drm_legacy_unlock(struct drm_device *dev, void *data, struct drm_file *file_priv)
 {
 	struct drm_lock *lock = data;
 	struct drm_master *master = file_priv->master;
@@ -157,7 +159,7 @@ int drm_unlock(struct drm_device *dev, void *data, struct drm_file *file_priv)
 		return -EINVAL;
 	}
 
-	if (drm_lock_free(&master->lock, lock->context)) {
+	if (drm_legacy_lock_free(&master->lock, lock->context)) {
 		/* FIXME: Should really bail out here. */
 	}
 
@@ -250,7 +252,7 @@ static int drm_lock_transfer(struct drm_lock_data *lock_data,
  * Marks the lock as not held, via the \p cmpxchg instruction. Wakes any task
  * waiting on the lock queue.
  */
-int drm_lock_free(struct drm_lock_data *lock_data, unsigned int context)
+int drm_legacy_lock_free(struct drm_lock_data *lock_data, unsigned int context)
 {
 	unsigned int old, new, prev;
 	volatile unsigned int *lock = &lock_data->hw_lock->lock;
@@ -286,26 +288,27 @@ int drm_lock_free(struct drm_lock_data *lock_data, unsigned int context)
  * If the lock is not held, then let the signal proceed as usual.  If the lock
  * is held, then set the contended flag and keep the signal blocked.
  *
- * \param priv pointer to a drm_sigdata structure.
+ * \param priv pointer to a drm_device structure.
  * \return one if the signal should be delivered normally, or zero if the
  * signal should be blocked.
  */
 static int drm_notifier(void *priv)
 {
-	struct drm_sigdata *s = (struct drm_sigdata *) priv;
+	struct drm_device *dev = priv;
+	struct drm_hw_lock *lock = dev->sigdata.lock;
 	unsigned int old, new, prev;
 
 	/* Allow signal delivery if lock isn't held */
-	if (!s->lock || !_DRM_LOCK_IS_HELD(s->lock->lock)
-	    || _DRM_LOCKING_CONTEXT(s->lock->lock) != s->context)
+	if (!lock || !_DRM_LOCK_IS_HELD(lock->lock)
+	    || _DRM_LOCKING_CONTEXT(lock->lock) != dev->sigdata.context)
 		return 1;
 
 	/* Otherwise, set flag to force call to
 	   drmUnlock */
 	do {
-		old = s->lock->lock;
+		old = lock->lock;
 		new = old | _DRM_LOCK_CONT;
-		prev = cmpxchg(&s->lock->lock, old, new);
+		prev = cmpxchg(&lock->lock, old, new);
 	} while (prev != old);
 	return 0;
 }
@@ -323,7 +326,7 @@ static int drm_notifier(void *priv)
  * having to worry about starvation.
  */
 
-void drm_idlelock_take(struct drm_lock_data *lock_data)
+void drm_legacy_idlelock_take(struct drm_lock_data *lock_data)
 {
 	int ret;
 
@@ -340,9 +343,9 @@ void drm_idlelock_take(struct drm_lock_data *lock_data)
 	}
 	spin_unlock_bh(&lock_data->spinlock);
 }
-EXPORT_SYMBOL(drm_idlelock_take);
+EXPORT_SYMBOL(drm_legacy_idlelock_take);
 
-void drm_idlelock_release(struct drm_lock_data *lock_data)
+void drm_legacy_idlelock_release(struct drm_lock_data *lock_data)
 {
 	unsigned int old, prev;
 	volatile unsigned int *lock = &lock_data->hw_lock->lock;
@@ -360,9 +363,10 @@ void drm_idlelock_release(struct drm_lock_data *lock_data)
 	}
 	spin_unlock_bh(&lock_data->spinlock);
 }
-EXPORT_SYMBOL(drm_idlelock_release);
+EXPORT_SYMBOL(drm_legacy_idlelock_release);
 
-int drm_i_have_hw_lock(struct drm_device *dev, struct drm_file *file_priv)
+int drm_legacy_i_have_hw_lock(struct drm_device *dev,
+			      struct drm_file *file_priv)
 {
 	struct drm_master *master = file_priv->master;
 	return (file_priv->lock_count && master->lock.hw_lock &&
diff --git a/drivers/gpu/drm/drm_memory.c b/drivers/gpu/drm/drm_memory.c
index 00c67c0f2381..a521ef6ff807 100644
--- a/drivers/gpu/drm/drm_memory.c
+++ b/drivers/gpu/drm/drm_memory.c
@@ -36,8 +36,20 @@
 #include <linux/highmem.h>
 #include <linux/export.h>
 #include <drm/drmP.h>
+#include "drm_legacy.h"
 
 #if __OS_HAS_AGP
+
+#ifdef HAVE_PAGE_AGP
+# include <asm/agp.h>
+#else
+# ifdef __powerpc__
+#  define PAGE_AGP	__pgprot(_PAGE_KERNEL | _PAGE_NO_CACHE)
+# else
+#  define PAGE_AGP	PAGE_KERNEL
+# endif
+#endif
+
 static void *agp_remap(unsigned long offset, unsigned long size,
 		       struct drm_device * dev)
 {
@@ -108,25 +120,25 @@ static inline void *agp_remap(unsigned long offset, unsigned long size,
 
 #endif				/* agp */
 
-void drm_core_ioremap(struct drm_local_map *map, struct drm_device *dev)
+void drm_legacy_ioremap(struct drm_local_map *map, struct drm_device *dev)
 {
 	if (dev->agp && dev->agp->cant_use_aperture && map->type == _DRM_AGP)
 		map->handle = agp_remap(map->offset, map->size, dev);
 	else
 		map->handle = ioremap(map->offset, map->size);
 }
-EXPORT_SYMBOL(drm_core_ioremap);
+EXPORT_SYMBOL(drm_legacy_ioremap);
 
-void drm_core_ioremap_wc(struct drm_local_map *map, struct drm_device *dev)
+void drm_legacy_ioremap_wc(struct drm_local_map *map, struct drm_device *dev)
 {
 	if (dev->agp && dev->agp->cant_use_aperture && map->type == _DRM_AGP)
 		map->handle = agp_remap(map->offset, map->size, dev);
 	else
 		map->handle = ioremap_wc(map->offset, map->size);
 }
-EXPORT_SYMBOL(drm_core_ioremap_wc);
+EXPORT_SYMBOL(drm_legacy_ioremap_wc);
 
-void drm_core_ioremapfree(struct drm_local_map *map, struct drm_device *dev)
+void drm_legacy_ioremapfree(struct drm_local_map *map, struct drm_device *dev)
 {
 	if (!map->handle || !map->size)
 		return;
@@ -136,4 +148,4 @@ void drm_core_ioremapfree(struct drm_local_map *map, struct drm_device *dev)
 	else
 		iounmap(map->handle);
 }
-EXPORT_SYMBOL(drm_core_ioremapfree);
+EXPORT_SYMBOL(drm_legacy_ioremapfree);
diff --git a/drivers/gpu/drm/drm_mipi_dsi.c b/drivers/gpu/drm/drm_mipi_dsi.c
index 6aa6a9e95570..eb6dfe52cab2 100644
--- a/drivers/gpu/drm/drm_mipi_dsi.c
+++ b/drivers/gpu/drm/drm_mipi_dsi.c
@@ -231,6 +231,9 @@ ssize_t mipi_dsi_dcs_write(struct mipi_dsi_device *dsi, const void *data,
 		break;
 	}
 
+	if (dsi->mode_flags & MIPI_DSI_MODE_LPM)
+		msg.flags = MIPI_DSI_MSG_USE_LPM;
+
 	return ops->transfer(dsi->host, &msg);
 }
 EXPORT_SYMBOL(mipi_dsi_dcs_write);
@@ -260,6 +263,9 @@ ssize_t mipi_dsi_dcs_read(struct mipi_dsi_device *dsi, u8 cmd, void *data,
 	if (!ops || !ops->transfer)
 		return -ENOSYS;
 
+	if (dsi->mode_flags & MIPI_DSI_MODE_LPM)
+		msg.flags = MIPI_DSI_MSG_USE_LPM;
+
 	return ops->transfer(dsi->host, &msg);
 }
 EXPORT_SYMBOL(mipi_dsi_dcs_read);
diff --git a/drivers/gpu/drm/drm_modes.c b/drivers/gpu/drm/drm_modes.c
index bedf1894e17e..d1b7d2006529 100644
--- a/drivers/gpu/drm/drm_modes.c
+++ b/drivers/gpu/drm/drm_modes.c
@@ -1259,6 +1259,7 @@ drm_mode_create_from_cmdline_mode(struct drm_device *dev,
 	if (!mode)
 		return NULL;
 
+	mode->type |= DRM_MODE_TYPE_USERDEF;
 	drm_mode_set_crtcinfo(mode, CRTC_INTERLACE_HALVE_V);
 	return mode;
 }
diff --git a/drivers/gpu/drm/drm_modeset_lock.c b/drivers/gpu/drm/drm_modeset_lock.c
index 3a02e5e3e9f3..474e4d12a2d8 100644
--- a/drivers/gpu/drm/drm_modeset_lock.c
+++ b/drivers/gpu/drm/drm_modeset_lock.c
@@ -57,6 +57,212 @@
 
 
 /**
+ * __drm_modeset_lock_all - internal helper to grab all modeset locks
+ * @dev: DRM device
+ * @trylock: trylock mode for atomic contexts
+ *
+ * This is a special version of drm_modeset_lock_all() which can also be used in
+ * atomic contexts. Then @trylock must be set to true.
+ *
+ * Returns:
+ * 0 on success or negative error code on failure.
+ */
+int __drm_modeset_lock_all(struct drm_device *dev,
+			   bool trylock)
+{
+	struct drm_mode_config *config = &dev->mode_config;
+	struct drm_modeset_acquire_ctx *ctx;
+	int ret;
+
+	ctx = kzalloc(sizeof(*ctx),
+		      trylock ? GFP_ATOMIC : GFP_KERNEL);
+	if (!ctx)
+		return -ENOMEM;
+
+	if (trylock) {
+		if (!mutex_trylock(&config->mutex))
+			return -EBUSY;
+	} else {
+		mutex_lock(&config->mutex);
+	}
+
+	drm_modeset_acquire_init(ctx, 0);
+	ctx->trylock_only = trylock;
+
+retry:
+	ret = drm_modeset_lock(&config->connection_mutex, ctx);
+	if (ret)
+		goto fail;
+	ret = drm_modeset_lock_all_crtcs(dev, ctx);
+	if (ret)
+		goto fail;
+
+	WARN_ON(config->acquire_ctx);
+
+	/* now we hold the locks, so now that it is safe, stash the
+	 * ctx for drm_modeset_unlock_all():
+	 */
+	config->acquire_ctx = ctx;
+
+	drm_warn_on_modeset_not_all_locked(dev);
+
+	return 0;
+
+fail:
+	if (ret == -EDEADLK) {
+		drm_modeset_backoff(ctx);
+		goto retry;
+	}
+
+	return ret;
+}
+EXPORT_SYMBOL(__drm_modeset_lock_all);
+
+/**
+ * drm_modeset_lock_all - take all modeset locks
+ * @dev: drm device
+ *
+ * This function takes all modeset locks, suitable where a more fine-grained
+ * scheme isn't (yet) implemented. Locks must be dropped with
+ * drm_modeset_unlock_all.
+ */
+void drm_modeset_lock_all(struct drm_device *dev)
+{
+	WARN_ON(__drm_modeset_lock_all(dev, false) != 0);
+}
+EXPORT_SYMBOL(drm_modeset_lock_all);
+
+/**
+ * drm_modeset_unlock_all - drop all modeset locks
+ * @dev: device
+ *
+ * This function drop all modeset locks taken by drm_modeset_lock_all.
+ */
+void drm_modeset_unlock_all(struct drm_device *dev)
+{
+	struct drm_mode_config *config = &dev->mode_config;
+	struct drm_modeset_acquire_ctx *ctx = config->acquire_ctx;
+
+	if (WARN_ON(!ctx))
+		return;
+
+	config->acquire_ctx = NULL;
+	drm_modeset_drop_locks(ctx);
+	drm_modeset_acquire_fini(ctx);
+
+	kfree(ctx);
+
+	mutex_unlock(&dev->mode_config.mutex);
+}
+EXPORT_SYMBOL(drm_modeset_unlock_all);
+
+/**
+ * drm_modeset_lock_crtc - lock crtc with hidden acquire ctx
+ * @crtc: drm crtc
+ *
+ * This function locks the given crtc using a hidden acquire context. This is
+ * necessary so that drivers internally using the atomic interfaces can grab
+ * further locks with the lock acquire context.
+ */
+void drm_modeset_lock_crtc(struct drm_crtc *crtc)
+{
+	struct drm_modeset_acquire_ctx *ctx;
+	int ret;
+
+	ctx = kzalloc(sizeof(*ctx), GFP_KERNEL);
+	if (WARN_ON(!ctx))
+		return;
+
+	drm_modeset_acquire_init(ctx, 0);
+
+retry:
+	ret = drm_modeset_lock(&crtc->mutex, ctx);
+	if (ret)
+		goto fail;
+
+	WARN_ON(crtc->acquire_ctx);
+
+	/* now we hold the locks, so now that it is safe, stash the
+	 * ctx for drm_modeset_unlock_crtc():
+	 */
+	crtc->acquire_ctx = ctx;
+
+	return;
+
+fail:
+	if (ret == -EDEADLK) {
+		drm_modeset_backoff(ctx);
+		goto retry;
+	}
+}
+EXPORT_SYMBOL(drm_modeset_lock_crtc);
+
+/**
+ * drm_modeset_legacy_acquire_ctx - find acquire ctx for legacy ioctls
+ * @crtc: drm crtc
+ *
+ * Legacy ioctl operations like cursor updates or page flips only have per-crtc
+ * locking, and store the acquire ctx in the corresponding crtc. All other
+ * legacy operations take all locks and use a global acquire context. This
+ * function grabs the right one.
+ */
+struct drm_modeset_acquire_ctx *
+drm_modeset_legacy_acquire_ctx(struct drm_crtc *crtc)
+{
+	if (crtc->acquire_ctx)
+		return crtc->acquire_ctx;
+
+	WARN_ON(!crtc->dev->mode_config.acquire_ctx);
+
+	return crtc->dev->mode_config.acquire_ctx;
+}
+EXPORT_SYMBOL(drm_modeset_legacy_acquire_ctx);
+
+/**
+ * drm_modeset_unlock_crtc - drop crtc lock
+ * @crtc: drm crtc
+ *
+ * This drops the crtc lock acquire with drm_modeset_lock_crtc() and all other
+ * locks acquired through the hidden context.
+ */
+void drm_modeset_unlock_crtc(struct drm_crtc *crtc)
+{
+	struct drm_modeset_acquire_ctx *ctx = crtc->acquire_ctx;
+
+	if (WARN_ON(!ctx))
+		return;
+
+	crtc->acquire_ctx = NULL;
+	drm_modeset_drop_locks(ctx);
+	drm_modeset_acquire_fini(ctx);
+
+	kfree(ctx);
+}
+EXPORT_SYMBOL(drm_modeset_unlock_crtc);
+
+/**
+ * drm_warn_on_modeset_not_all_locked - check that all modeset locks are locked
+ * @dev: device
+ *
+ * Useful as a debug assert.
+ */
+void drm_warn_on_modeset_not_all_locked(struct drm_device *dev)
+{
+	struct drm_crtc *crtc;
+
+	/* Locking is currently fubar in the panic handler. */
+	if (oops_in_progress)
+		return;
+
+	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head)
+		WARN_ON(!drm_modeset_is_locked(&crtc->mutex));
+
+	WARN_ON(!drm_modeset_is_locked(&dev->mode_config.connection_mutex));
+	WARN_ON(!mutex_is_locked(&dev->mode_config.mutex));
+}
+EXPORT_SYMBOL(drm_warn_on_modeset_not_all_locked);
+
+/**
  * drm_modeset_acquire_init - initialize acquire context
  * @ctx: the acquire context
  * @flags: for future
@@ -108,7 +314,12 @@ static inline int modeset_lock(struct drm_modeset_lock *lock,
 
 	WARN_ON(ctx->contended);
 
-	if (interruptible && slow) {
+	if (ctx->trylock_only) {
+		if (!ww_mutex_trylock(&lock->mutex))
+			return -EBUSY;
+		else
+			return 0;
+	} else if (interruptible && slow) {
 		ret = ww_mutex_lock_slow_interruptible(&lock->mutex, &ctx->ww_ctx);
 	} else if (interruptible) {
 		ret = ww_mutex_lock_interruptible(&lock->mutex, &ctx->ww_ctx);
diff --git a/drivers/gpu/drm/drm_pci.c b/drivers/gpu/drm/drm_pci.c
index 020cfd934854..fd29f03645b8 100644
--- a/drivers/gpu/drm/drm_pci.c
+++ b/drivers/gpu/drm/drm_pci.c
@@ -27,6 +27,7 @@
 #include <linux/dma-mapping.h>
 #include <linux/export.h>
 #include <drm/drmP.h>
+#include "drm_legacy.h"
 
 /**
  * drm_pci_alloc - Allocate a PCI consistent memory block, for DMA.
@@ -81,7 +82,7 @@ EXPORT_SYMBOL(drm_pci_alloc);
  *
  * This function is for internal use in the Linux-specific DRM core code.
  */
-void __drm_pci_free(struct drm_device * dev, drm_dma_handle_t * dmah)
+void __drm_legacy_pci_free(struct drm_device * dev, drm_dma_handle_t * dmah)
 {
 	unsigned long addr;
 	size_t sz;
@@ -105,7 +106,7 @@ void __drm_pci_free(struct drm_device * dev, drm_dma_handle_t * dmah)
  */
 void drm_pci_free(struct drm_device * dev, drm_dma_handle_t * dmah)
 {
-	__drm_pci_free(dev, dmah);
+	__drm_legacy_pci_free(dev, dmah);
 	kfree(dmah);
 }
 
@@ -127,34 +128,20 @@ static int drm_get_pci_domain(struct drm_device *dev)
 	return pci_domain_nr(dev->pdev->bus);
 }
 
-static int drm_pci_set_busid(struct drm_device *dev, struct drm_master *master)
+int drm_pci_set_busid(struct drm_device *dev, struct drm_master *master)
 {
-	int len, ret;
-	master->unique_len = 40;
-	master->unique_size = master->unique_len;
-	master->unique = kmalloc(master->unique_size, GFP_KERNEL);
-	if (master->unique == NULL)
+	master->unique = kasprintf(GFP_KERNEL, "pci:%04x:%02x:%02x.%d",
+					drm_get_pci_domain(dev),
+					dev->pdev->bus->number,
+					PCI_SLOT(dev->pdev->devfn),
+					PCI_FUNC(dev->pdev->devfn));
+	if (!master->unique)
 		return -ENOMEM;
 
-
-	len = snprintf(master->unique, master->unique_len,
-		       "pci:%04x:%02x:%02x.%d",
-		       drm_get_pci_domain(dev),
-		       dev->pdev->bus->number,
-		       PCI_SLOT(dev->pdev->devfn),
-		       PCI_FUNC(dev->pdev->devfn));
-
-	if (len >= master->unique_len) {
-		DRM_ERROR("buffer overflow");
-		ret = -EINVAL;
-		goto err;
-	} else
-		master->unique_len = len;
-
+	master->unique_len = strlen(master->unique);
 	return 0;
-err:
-	return ret;
 }
+EXPORT_SYMBOL(drm_pci_set_busid);
 
 int drm_pci_set_unique(struct drm_device *dev,
 		       struct drm_master *master,
@@ -163,8 +150,7 @@ int drm_pci_set_unique(struct drm_device *dev,
 	int domain, bus, slot, func, ret;
 
 	master->unique_len = u->unique_len;
-	master->unique_size = u->unique_len + 1;
-	master->unique = kmalloc(master->unique_size, GFP_KERNEL);
+	master->unique = kmalloc(master->unique_len + 1, GFP_KERNEL);
 	if (!master->unique) {
 		ret = -ENOMEM;
 		goto err;
@@ -269,10 +255,6 @@ void drm_pci_agp_destroy(struct drm_device *dev)
 	}
 }
 
-static struct drm_bus drm_pci_bus = {
-	.set_busid = drm_pci_set_busid,
-};
-
 /**
  * drm_get_pci_dev - Register a PCI device with the DRM subsystem
  * @pdev: PCI device
@@ -353,8 +335,6 @@ int drm_pci_init(struct drm_driver *driver, struct pci_driver *pdriver)
 
 	DRM_DEBUG("\n");
 
-	driver->bus = &drm_pci_bus;
-
 	if (driver->driver_features & DRIVER_MODESET)
 		return pci_register_driver(pdriver);
 
diff --git a/drivers/gpu/drm/drm_platform.c b/drivers/gpu/drm/drm_platform.c
index d5b76f148c12..5314c9d5fef4 100644
--- a/drivers/gpu/drm/drm_platform.c
+++ b/drivers/gpu/drm/drm_platform.c
@@ -68,42 +68,23 @@ err_free:
 	return ret;
 }
 
-static int drm_platform_set_busid(struct drm_device *dev, struct drm_master *master)
+int drm_platform_set_busid(struct drm_device *dev, struct drm_master *master)
 {
-	int len, ret, id;
-
-	master->unique_len = 13 + strlen(dev->platformdev->name);
-	master->unique_size = master->unique_len;
-	master->unique = kmalloc(master->unique_len + 1, GFP_KERNEL);
-
-	if (master->unique == NULL)
-		return -ENOMEM;
+	int id;
 
 	id = dev->platformdev->id;
-
-	/* if only a single instance of the platform device, id will be
-	 * set to -1.. use 0 instead to avoid a funny looking bus-id:
-	 */
-	if (id == -1)
+	if (id < 0)
 		id = 0;
 
-	len = snprintf(master->unique, master->unique_len,
-			"platform:%s:%02d", dev->platformdev->name, id);
-
-	if (len > master->unique_len) {
-		DRM_ERROR("Unique buffer overflowed\n");
-		ret = -EINVAL;
-		goto err;
-	}
+	master->unique = kasprintf(GFP_KERNEL, "platform:%s:%02d",
+						dev->platformdev->name, id);
+	if (!master->unique)
+		return -ENOMEM;
 
+	master->unique_len = strlen(master->unique);
 	return 0;
-err:
-	return ret;
 }
-
-static struct drm_bus drm_platform_bus = {
-	.set_busid = drm_platform_set_busid,
-};
+EXPORT_SYMBOL(drm_platform_set_busid);
 
 /**
  * drm_platform_init - Register a platform device with the DRM subsystem
@@ -120,7 +101,6 @@ int drm_platform_init(struct drm_driver *driver, struct platform_device *platfor
 {
 	DRM_DEBUG("\n");
 
-	driver->bus = &drm_platform_bus;
 	return drm_get_platform_dev(platform_device, driver);
 }
 EXPORT_SYMBOL(drm_platform_init);
diff --git a/drivers/gpu/drm/drm_prime.c b/drivers/gpu/drm/drm_prime.c
index 99d578bad17e..78ca30808422 100644
--- a/drivers/gpu/drm/drm_prime.c
+++ b/drivers/gpu/drm/drm_prime.c
@@ -29,6 +29,9 @@
 #include <linux/export.h>
 #include <linux/dma-buf.h>
 #include <drm/drmP.h>
+#include <drm/drm_gem.h>
+
+#include "drm_internal.h"
 
 /*
  * DMA-BUF/GEM Object references and lifetime overview:
@@ -522,7 +525,7 @@ struct drm_gem_object *drm_gem_prime_import(struct drm_device *dev,
 		goto fail_detach;
 	}
 
-	obj = dev->driver->gem_prime_import_sg_table(dev, dma_buf->size, sgt);
+	obj = dev->driver->gem_prime_import_sg_table(dev, attach, sgt);
 	if (IS_ERR(obj)) {
 		ret = PTR_ERR(obj);
 		goto fail_unmap;
diff --git a/drivers/gpu/drm/drm_probe_helper.c b/drivers/gpu/drm/drm_probe_helper.c
index db7d250f7ac7..6857e9ad6339 100644
--- a/drivers/gpu/drm/drm_probe_helper.c
+++ b/drivers/gpu/drm/drm_probe_helper.c
@@ -82,6 +82,22 @@ static void drm_mode_validate_flag(struct drm_connector *connector,
 	return;
 }
 
+static int drm_helper_probe_add_cmdline_mode(struct drm_connector *connector)
+{
+	struct drm_display_mode *mode;
+
+	if (!connector->cmdline_mode.specified)
+		return 0;
+
+	mode = drm_mode_create_from_cmdline_mode(connector->dev,
+						 &connector->cmdline_mode);
+	if (mode == NULL)
+		return 0;
+
+	drm_mode_probed_add(connector, mode);
+	return 1;
+}
+
 static int drm_helper_probe_single_connector_modes_merge_bits(struct drm_connector *connector,
 							      uint32_t maxX, uint32_t maxY, bool merge_type_bits)
 {
@@ -141,6 +157,7 @@ static int drm_helper_probe_single_connector_modes_merge_bits(struct drm_connect
 
 	if (count == 0 && connector->status == connector_status_connected)
 		count = drm_add_modes_noedid(connector, 1024, 768);
+	count += drm_helper_probe_add_cmdline_mode(connector);
 	if (count == 0)
 		goto prune;
 
diff --git a/drivers/gpu/drm/drm_scatter.c b/drivers/gpu/drm/drm_scatter.c
index 1c78406f6e71..4f0f3b36d537 100644
--- a/drivers/gpu/drm/drm_scatter.c
+++ b/drivers/gpu/drm/drm_scatter.c
@@ -34,6 +34,7 @@
 #include <linux/vmalloc.h>
 #include <linux/slab.h>
 #include <drm/drmP.h>
+#include "drm_legacy.h"
 
 #define DEBUG_SCATTER 0
 
@@ -78,8 +79,8 @@ void drm_legacy_sg_cleanup(struct drm_device *dev)
 # define ScatterHandle(x) (unsigned int)(x)
 #endif
 
-int drm_sg_alloc(struct drm_device *dev, void *data,
-		 struct drm_file *file_priv)
+int drm_legacy_sg_alloc(struct drm_device *dev, void *data,
+			struct drm_file *file_priv)
 {
 	struct drm_scatter_gather *request = data;
 	struct drm_sg_mem *entry;
@@ -194,8 +195,8 @@ int drm_sg_alloc(struct drm_device *dev, void *data,
 	return -ENOMEM;
 }
 
-int drm_sg_free(struct drm_device *dev, void *data,
-		struct drm_file *file_priv)
+int drm_legacy_sg_free(struct drm_device *dev, void *data,
+		       struct drm_file *file_priv)
 {
 	struct drm_scatter_gather *request = data;
 	struct drm_sg_mem *entry;
diff --git a/drivers/gpu/drm/drm_sysfs.c b/drivers/gpu/drm/drm_sysfs.c
index ab1a5f6dde8a..cc3d6d6d67e0 100644
--- a/drivers/gpu/drm/drm_sysfs.c
+++ b/drivers/gpu/drm/drm_sysfs.c
@@ -21,6 +21,7 @@
 #include <drm/drm_sysfs.h>
 #include <drm/drm_core.h>
 #include <drm/drmP.h>
+#include "drm_internal.h"
 
 #define to_drm_minor(d) dev_get_drvdata(d)
 #define to_drm_connector(d) dev_get_drvdata(d)
diff --git a/drivers/gpu/drm/drm_usb.c b/drivers/gpu/drm/drm_usb.c
deleted file mode 100644
index f2fe94aab901..000000000000
--- a/drivers/gpu/drm/drm_usb.c
+++ /dev/null
@@ -1,88 +0,0 @@
-#include <drm/drmP.h>
-#include <drm/drm_usb.h>
-#include <linux/usb.h>
-#include <linux/module.h>
-
-int drm_get_usb_dev(struct usb_interface *interface,
-		    const struct usb_device_id *id,
-		    struct drm_driver *driver)
-{
-	struct drm_device *dev;
-	int ret;
-
-	DRM_DEBUG("\n");
-
-	dev = drm_dev_alloc(driver, &interface->dev);
-	if (!dev)
-		return -ENOMEM;
-
-	dev->usbdev = interface_to_usbdev(interface);
-	usb_set_intfdata(interface, dev);
-
-	ret = drm_dev_register(dev, 0);
-	if (ret)
-		goto err_free;
-
-	DRM_INFO("Initialized %s %d.%d.%d %s on minor %d\n",
-		 driver->name, driver->major, driver->minor, driver->patchlevel,
-		 driver->date, dev->primary->index);
-
-	return 0;
-
-err_free:
-	drm_dev_unref(dev);
-	return ret;
-
-}
-EXPORT_SYMBOL(drm_get_usb_dev);
-
-static int drm_usb_set_busid(struct drm_device *dev,
-			       struct drm_master *master)
-{
-	return 0;
-}
-
-static struct drm_bus drm_usb_bus = {
-	.set_busid = drm_usb_set_busid,
-};
-
-/**
- * drm_usb_init - Register matching USB devices with the DRM subsystem
- * @driver: DRM device driver
- * @udriver: USB device driver
- *
- * Registers one or more devices matched by a USB driver with the DRM
- * subsystem.
- *
- * Return: 0 on success or a negative error code on failure.
- */
-int drm_usb_init(struct drm_driver *driver, struct usb_driver *udriver)
-{
-	int res;
-	DRM_DEBUG("\n");
-
-	driver->bus = &drm_usb_bus;
-
-	res = usb_register(udriver);
-	return res;
-}
-EXPORT_SYMBOL(drm_usb_init);
-
-/**
- * drm_usb_exit - Unregister matching USB devices from the DRM subsystem
- * @driver: DRM device driver
- * @udriver: USB device driver
- *
- * Unregisters one or more devices matched by a USB driver from the DRM
- * subsystem.
- */
-void drm_usb_exit(struct drm_driver *driver,
-		  struct usb_driver *udriver)
-{
-	usb_deregister(udriver);
-}
-EXPORT_SYMBOL(drm_usb_exit);
-
-MODULE_AUTHOR("David Airlie");
-MODULE_DESCRIPTION("USB DRM support");
-MODULE_LICENSE("GPL and additional rights");
diff --git a/drivers/gpu/drm/drm_vm.c b/drivers/gpu/drm/drm_vm.c
index 24e045c4f531..4a2c328959e5 100644
--- a/drivers/gpu/drm/drm_vm.c
+++ b/drivers/gpu/drm/drm_vm.c
@@ -35,10 +35,19 @@
 
 #include <drm/drmP.h>
 #include <linux/export.h>
+#include <linux/seq_file.h>
 #if defined(__ia64__)
 #include <linux/efi.h>
 #include <linux/slab.h>
 #endif
+#include <asm/pgtable.h>
+#include "drm_legacy.h"
+
+struct drm_vma_entry {
+	struct list_head head;
+	struct vm_area_struct *vma;
+	pid_t pid;
+};
 
 static void drm_vm_open(struct vm_area_struct *vma);
 static void drm_vm_close(struct vm_area_struct *vma);
@@ -48,15 +57,11 @@ static pgprot_t drm_io_prot(struct drm_local_map *map,
 {
 	pgprot_t tmp = vm_get_page_prot(vma->vm_flags);
 
-#if defined(__i386__) || defined(__x86_64__)
+#if defined(__i386__) || defined(__x86_64__) || defined(__powerpc__)
 	if (map->type == _DRM_REGISTERS && !(map->flags & _DRM_WRITE_COMBINING))
 		tmp = pgprot_noncached(tmp);
 	else
 		tmp = pgprot_writecombine(tmp);
-#elif defined(__powerpc__)
-	pgprot_val(tmp) |= _PAGE_NO_CACHE;
-	if (map->type == _DRM_REGISTERS)
-		pgprot_val(tmp) |= _PAGE_GUARDED;
 #elif defined(__ia64__)
 	if (efi_range_is_wc(vma->vm_start, vma->vm_end -
 				    vma->vm_start))
@@ -263,7 +268,7 @@ static void drm_vm_shm_close(struct vm_area_struct *vma)
 				dmah.vaddr = map->handle;
 				dmah.busaddr = map->offset;
 				dmah.size = map->size;
-				__drm_pci_free(dev, &dmah);
+				__drm_legacy_pci_free(dev, &dmah);
 				break;
 			}
 			kfree(map);
@@ -412,7 +417,6 @@ void drm_vm_open_locked(struct drm_device *dev,
 		list_add(&vma_entry->head, &dev->vmalist);
 	}
 }
-EXPORT_SYMBOL_GPL(drm_vm_open_locked);
 
 static void drm_vm_open(struct vm_area_struct *vma)
 {
@@ -532,7 +536,7 @@ static resource_size_t drm_core_get_reg_ofs(struct drm_device *dev)
  * according to the mapping type and remaps the pages. Finally sets the file
  * pointer and calls vm_open().
  */
-int drm_mmap_locked(struct file *filp, struct vm_area_struct *vma)
+static int drm_mmap_locked(struct file *filp, struct vm_area_struct *vma)
 {
 	struct drm_file *priv = filp->private_data;
 	struct drm_device *dev = priv->minor->dev;
@@ -646,7 +650,7 @@ int drm_mmap_locked(struct file *filp, struct vm_area_struct *vma)
 	return 0;
 }
 
-int drm_mmap(struct file *filp, struct vm_area_struct *vma)
+int drm_legacy_mmap(struct file *filp, struct vm_area_struct *vma)
 {
 	struct drm_file *priv = filp->private_data;
 	struct drm_device *dev = priv->minor->dev;
@@ -661,4 +665,69 @@ int drm_mmap(struct file *filp, struct vm_area_struct *vma)
 
 	return ret;
 }
-EXPORT_SYMBOL(drm_mmap);
+EXPORT_SYMBOL(drm_legacy_mmap);
+
+void drm_legacy_vma_flush(struct drm_device *dev)
+{
+	struct drm_vma_entry *vma, *vma_temp;
+
+	/* Clear vma list (only needed for legacy drivers) */
+	list_for_each_entry_safe(vma, vma_temp, &dev->vmalist, head) {
+		list_del(&vma->head);
+		kfree(vma);
+	}
+}
+
+int drm_vma_info(struct seq_file *m, void *data)
+{
+	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_device *dev = node->minor->dev;
+	struct drm_vma_entry *pt;
+	struct vm_area_struct *vma;
+	unsigned long vma_count = 0;
+#if defined(__i386__)
+	unsigned int pgprot;
+#endif
+
+	mutex_lock(&dev->struct_mutex);
+	list_for_each_entry(pt, &dev->vmalist, head)
+		vma_count++;
+
+	seq_printf(m, "vma use count: %lu, high_memory = %pK, 0x%pK\n",
+		   vma_count, high_memory,
+		   (void *)(unsigned long)virt_to_phys(high_memory));
+
+	list_for_each_entry(pt, &dev->vmalist, head) {
+		vma = pt->vma;
+		if (!vma)
+			continue;
+		seq_printf(m,
+			   "\n%5d 0x%pK-0x%pK %c%c%c%c%c%c 0x%08lx000",
+			   pt->pid,
+			   (void *)vma->vm_start, (void *)vma->vm_end,
+			   vma->vm_flags & VM_READ ? 'r' : '-',
+			   vma->vm_flags & VM_WRITE ? 'w' : '-',
+			   vma->vm_flags & VM_EXEC ? 'x' : '-',
+			   vma->vm_flags & VM_MAYSHARE ? 's' : 'p',
+			   vma->vm_flags & VM_LOCKED ? 'l' : '-',
+			   vma->vm_flags & VM_IO ? 'i' : '-',
+			   vma->vm_pgoff);
+
+#if defined(__i386__)
+		pgprot = pgprot_val(vma->vm_page_prot);
+		seq_printf(m, " %c%c%c%c%c%c%c%c%c",
+			   pgprot & _PAGE_PRESENT ? 'p' : '-',
+			   pgprot & _PAGE_RW ? 'w' : 'r',
+			   pgprot & _PAGE_USER ? 'u' : 's',
+			   pgprot & _PAGE_PWT ? 't' : 'b',
+			   pgprot & _PAGE_PCD ? 'u' : 'c',
+			   pgprot & _PAGE_ACCESSED ? 'a' : '-',
+			   pgprot & _PAGE_DIRTY ? 'd' : '-',
+			   pgprot & _PAGE_PSE ? 'm' : 'k',
+			   pgprot & _PAGE_GLOBAL ? 'g' : 'l');
+#endif
+		seq_printf(m, "\n");
+	}
+	mutex_unlock(&dev->struct_mutex);
+	return 0;
+}
diff --git a/drivers/gpu/drm/exynos/exynos_dp_core.c b/drivers/gpu/drm/exynos/exynos_dp_core.c
index 4f3c7eb2d37d..cd50ece31601 100644
--- a/drivers/gpu/drm/exynos/exynos_dp_core.c
+++ b/drivers/gpu/drm/exynos/exynos_dp_core.c
@@ -329,8 +329,8 @@ static int exynos_dp_link_start(struct exynos_dp_device *dp)
 		return retval;
 
 	for (lane = 0; lane < lane_count; lane++)
-		buf[lane] = DP_TRAIN_PRE_EMPHASIS_0 |
-			    DP_TRAIN_VOLTAGE_SWING_400;
+		buf[lane] = DP_TRAIN_PRE_EMPH_LEVEL_0 |
+			    DP_TRAIN_VOLTAGE_SWING_LEVEL_0;
 
 	retval = exynos_dp_write_bytes_to_dpcd(dp, DP_TRAINING_LANE0_SET,
 			lane_count, buf);
@@ -937,6 +937,8 @@ static enum drm_connector_status exynos_dp_detect(
 
 static void exynos_dp_connector_destroy(struct drm_connector *connector)
 {
+	drm_connector_unregister(connector);
+	drm_connector_cleanup(connector);
 }
 
 static struct drm_connector_funcs exynos_dp_connector_funcs = {
@@ -1358,8 +1360,8 @@ static void exynos_dp_unbind(struct device *dev, struct device *master,
 
 	exynos_dp_dpms(display, DRM_MODE_DPMS_OFF);
 
+	exynos_dp_connector_destroy(&dp->connector);
 	encoder->funcs->destroy(encoder);
-	drm_connector_cleanup(&dp->connector);
 }
 
 static const struct component_ops exynos_dp_ops = {
diff --git a/drivers/gpu/drm/exynos/exynos_drm_crtc.c b/drivers/gpu/drm/exynos/exynos_drm_crtc.c
index b68e58f78cd1..8e38e9f8e542 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_crtc.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_crtc.c
@@ -32,7 +32,6 @@ enum exynos_crtc_mode {
  * Exynos specific crtc structure.
  *
  * @drm_crtc: crtc object.
- * @drm_plane: pointer of private plane object for this crtc
  * @manager: the manager associated with this crtc
  * @pipe: a crtc index created at load() with a new crtc object creation
  *	and the crtc object would be set to private->crtc array
@@ -46,7 +45,6 @@ enum exynos_crtc_mode {
  */
 struct exynos_drm_crtc {
 	struct drm_crtc			drm_crtc;
-	struct drm_plane		*plane;
 	struct exynos_drm_manager	*manager;
 	unsigned int			pipe;
 	unsigned int			dpms;
@@ -94,12 +92,12 @@ static void exynos_drm_crtc_commit(struct drm_crtc *crtc)
 
 	exynos_drm_crtc_dpms(crtc, DRM_MODE_DPMS_ON);
 
-	exynos_plane_commit(exynos_crtc->plane);
+	exynos_plane_commit(crtc->primary);
 
 	if (manager->ops->commit)
 		manager->ops->commit(manager);
 
-	exynos_plane_dpms(exynos_crtc->plane, DRM_MODE_DPMS_ON);
+	exynos_plane_dpms(crtc->primary, DRM_MODE_DPMS_ON);
 }
 
 static bool
@@ -123,10 +121,9 @@ exynos_drm_crtc_mode_set(struct drm_crtc *crtc, struct drm_display_mode *mode,
 {
 	struct exynos_drm_crtc *exynos_crtc = to_exynos_crtc(crtc);
 	struct exynos_drm_manager *manager = exynos_crtc->manager;
-	struct drm_plane *plane = exynos_crtc->plane;
+	struct drm_framebuffer *fb = crtc->primary->fb;
 	unsigned int crtc_w;
 	unsigned int crtc_h;
-	int ret;
 
 	/*
 	 * copy the mode data adjusted by mode_fixup() into crtc->mode
@@ -134,29 +131,21 @@ exynos_drm_crtc_mode_set(struct drm_crtc *crtc, struct drm_display_mode *mode,
 	 */
 	memcpy(&crtc->mode, adjusted_mode, sizeof(*adjusted_mode));
 
-	crtc_w = crtc->primary->fb->width - x;
-	crtc_h = crtc->primary->fb->height - y;
+	crtc_w = fb->width - x;
+	crtc_h = fb->height - y;
 
 	if (manager->ops->mode_set)
 		manager->ops->mode_set(manager, &crtc->mode);
 
-	ret = exynos_plane_mode_set(plane, crtc, crtc->primary->fb, 0, 0, crtc_w, crtc_h,
-				    x, y, crtc_w, crtc_h);
-	if (ret)
-		return ret;
-
-	plane->crtc = crtc;
-	plane->fb = crtc->primary->fb;
-	drm_framebuffer_reference(plane->fb);
-
-	return 0;
+	return exynos_plane_mode_set(crtc->primary, crtc, fb, 0, 0,
+				     crtc_w, crtc_h, x, y, crtc_w, crtc_h);
 }
 
 static int exynos_drm_crtc_mode_set_commit(struct drm_crtc *crtc, int x, int y,
 					  struct drm_framebuffer *old_fb)
 {
 	struct exynos_drm_crtc *exynos_crtc = to_exynos_crtc(crtc);
-	struct drm_plane *plane = exynos_crtc->plane;
+	struct drm_framebuffer *fb = crtc->primary->fb;
 	unsigned int crtc_w;
 	unsigned int crtc_h;
 	int ret;
@@ -167,11 +156,11 @@ static int exynos_drm_crtc_mode_set_commit(struct drm_crtc *crtc, int x, int y,
 		return -EPERM;
 	}
 
-	crtc_w = crtc->primary->fb->width - x;
-	crtc_h = crtc->primary->fb->height - y;
+	crtc_w = fb->width - x;
+	crtc_h = fb->height - y;
 
-	ret = exynos_plane_mode_set(plane, crtc, crtc->primary->fb, 0, 0, crtc_w, crtc_h,
-				    x, y, crtc_w, crtc_h);
+	ret = exynos_plane_mode_set(crtc->primary, crtc, fb, 0, 0,
+				    crtc_w, crtc_h, x, y, crtc_w, crtc_h);
 	if (ret)
 		return ret;
 
@@ -304,8 +293,7 @@ static int exynos_drm_crtc_set_property(struct drm_crtc *crtc,
 			exynos_drm_crtc_commit(crtc);
 			break;
 		case CRTC_MODE_BLANK:
-			exynos_plane_dpms(exynos_crtc->plane,
-					  DRM_MODE_DPMS_OFF);
+			exynos_plane_dpms(crtc->primary, DRM_MODE_DPMS_OFF);
 			break;
 		default:
 			break;
@@ -351,8 +339,10 @@ static void exynos_drm_crtc_attach_mode_property(struct drm_crtc *crtc)
 int exynos_drm_crtc_create(struct exynos_drm_manager *manager)
 {
 	struct exynos_drm_crtc *exynos_crtc;
+	struct drm_plane *plane;
 	struct exynos_drm_private *private = manager->drm_dev->dev_private;
 	struct drm_crtc *crtc;
+	int ret;
 
 	exynos_crtc = kzalloc(sizeof(*exynos_crtc), GFP_KERNEL);
 	if (!exynos_crtc)
@@ -364,11 +354,11 @@ int exynos_drm_crtc_create(struct exynos_drm_manager *manager)
 	exynos_crtc->dpms = DRM_MODE_DPMS_OFF;
 	exynos_crtc->manager = manager;
 	exynos_crtc->pipe = manager->pipe;
-	exynos_crtc->plane = exynos_plane_init(manager->drm_dev,
-				1 << manager->pipe, true);
-	if (!exynos_crtc->plane) {
-		kfree(exynos_crtc);
-		return -ENOMEM;
+	plane = exynos_plane_init(manager->drm_dev, 1 << manager->pipe,
+				  DRM_PLANE_TYPE_PRIMARY);
+	if (IS_ERR(plane)) {
+		ret = PTR_ERR(plane);
+		goto err_plane;
 	}
 
 	manager->crtc = &exynos_crtc->drm_crtc;
@@ -376,12 +366,22 @@ int exynos_drm_crtc_create(struct exynos_drm_manager *manager)
 
 	private->crtc[manager->pipe] = crtc;
 
-	drm_crtc_init(manager->drm_dev, crtc, &exynos_crtc_funcs);
+	ret = drm_crtc_init_with_planes(manager->drm_dev, crtc, plane, NULL,
+					&exynos_crtc_funcs);
+	if (ret < 0)
+		goto err_crtc;
+
 	drm_crtc_helper_add(crtc, &exynos_crtc_helper_funcs);
 
 	exynos_drm_crtc_attach_mode_property(crtc);
 
 	return 0;
+
+err_crtc:
+	plane->funcs->destroy(plane);
+err_plane:
+	kfree(exynos_crtc);
+	return ret;
 }
 
 int exynos_drm_crtc_enable_vblank(struct drm_device *dev, int pipe)
diff --git a/drivers/gpu/drm/exynos/exynos_drm_dpi.c b/drivers/gpu/drm/exynos/exynos_drm_dpi.c
index fa08f05e3e34..96c87db388fb 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_dpi.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_dpi.c
@@ -342,8 +342,12 @@ int exynos_dpi_remove(struct device *dev)
 	struct exynos_dpi *ctx = exynos_dpi_display.ctx;
 
 	exynos_dpi_dpms(&exynos_dpi_display, DRM_MODE_DPMS_OFF);
+
+	exynos_dpi_connector_destroy(&ctx->connector);
 	encoder->funcs->destroy(encoder);
-	drm_connector_cleanup(&ctx->connector);
+
+	if (ctx->panel)
+		drm_panel_detach(ctx->panel);
 
 	exynos_drm_component_del(dev, EXYNOS_DEVICE_TYPE_CONNECTOR);
 
diff --git a/drivers/gpu/drm/exynos/exynos_drm_drv.c b/drivers/gpu/drm/exynos/exynos_drm_drv.c
index 0d74e9b99c4e..443a2069858a 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_drv.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_drv.c
@@ -15,7 +15,6 @@
 #include <drm/drmP.h>
 #include <drm/drm_crtc_helper.h>
 
-#include <linux/anon_inodes.h>
 #include <linux/component.h>
 
 #include <drm/exynos_drm.h>
@@ -86,8 +85,9 @@ static int exynos_drm_load(struct drm_device *dev, unsigned long flags)
 		struct drm_plane *plane;
 		unsigned long possible_crtcs = (1 << MAX_CRTC) - 1;
 
-		plane = exynos_plane_init(dev, possible_crtcs, false);
-		if (!plane)
+		plane = exynos_plane_init(dev, possible_crtcs,
+					  DRM_PLANE_TYPE_OVERLAY);
+		if (IS_ERR(plane))
 			goto err_mode_config_cleanup;
 	}
 
@@ -116,6 +116,23 @@ static int exynos_drm_load(struct drm_device *dev, unsigned long flags)
 	/* force connectors detection */
 	drm_helper_hpd_irq_event(dev);
 
+	/*
+	 * enable drm irq mode.
+	 * - with irq_enabled = true, we can use the vblank feature.
+	 *
+	 * P.S. note that we wouldn't use drm irq handler but
+	 *	just specific driver own one instead because
+	 *	drm framework supports only one irq handler.
+	 */
+	dev->irq_enabled = true;
+
+	/*
+	 * with vblank_disable_allowed = true, vblank interrupt will be disabled
+	 * by drm timer once a current process gives up ownership of
+	 * vblank event.(after drm_vblank_put function is called)
+	 */
+	dev->vblank_disable_allowed = true;
+
 	return 0;
 
 err_unbind_all:
@@ -136,23 +153,19 @@ static int exynos_drm_unload(struct drm_device *dev)
 	exynos_drm_device_subdrv_remove(dev);
 
 	exynos_drm_fbdev_fini(dev);
-	drm_vblank_cleanup(dev);
 	drm_kms_helper_poll_fini(dev);
-	drm_mode_config_cleanup(dev);
 
+	component_unbind_all(dev->dev, dev);
+	drm_vblank_cleanup(dev);
+	drm_mode_config_cleanup(dev);
 	drm_release_iommu_mapping(dev);
-	kfree(dev->dev_private);
 
-	component_unbind_all(dev->dev, dev);
+	kfree(dev->dev_private);
 	dev->dev_private = NULL;
 
 	return 0;
 }
 
-static const struct file_operations exynos_drm_gem_fops = {
-	.mmap = exynos_drm_gem_mmap_buffer,
-};
-
 static int exynos_drm_suspend(struct drm_device *dev, pm_message_t state)
 {
 	struct drm_connector *connector;
@@ -191,7 +204,6 @@ static int exynos_drm_resume(struct drm_device *dev)
 static int exynos_drm_open(struct drm_device *dev, struct drm_file *file)
 {
 	struct drm_exynos_file_private *file_priv;
-	struct file *anon_filp;
 	int ret;
 
 	file_priv = kzalloc(sizeof(*file_priv), GFP_KERNEL);
@@ -204,21 +216,8 @@ static int exynos_drm_open(struct drm_device *dev, struct drm_file *file)
 	if (ret)
 		goto err_file_priv_free;
 
-	anon_filp = anon_inode_getfile("exynos_gem", &exynos_drm_gem_fops,
-					NULL, 0);
-	if (IS_ERR(anon_filp)) {
-		ret = PTR_ERR(anon_filp);
-		goto err_subdrv_close;
-	}
-
-	anon_filp->f_mode = FMODE_READ | FMODE_WRITE;
-	file_priv->anon_filp = anon_filp;
-
 	return ret;
 
-err_subdrv_close:
-	exynos_drm_subdrv_close(dev, file);
-
 err_file_priv_free:
 	kfree(file_priv);
 	file->driver_priv = NULL;
@@ -234,7 +233,6 @@ static void exynos_drm_preclose(struct drm_device *dev,
 static void exynos_drm_postclose(struct drm_device *dev, struct drm_file *file)
 {
 	struct exynos_drm_private *private = dev->dev_private;
-	struct drm_exynos_file_private *file_priv;
 	struct drm_pending_vblank_event *v, *vt;
 	struct drm_pending_event *e, *et;
 	unsigned long flags;
@@ -260,10 +258,6 @@ static void exynos_drm_postclose(struct drm_device *dev, struct drm_file *file)
 	}
 	spin_unlock_irqrestore(&dev->event_lock, flags);
 
-	file_priv = file->driver_priv;
-	if (file_priv->anon_filp)
-		fput(file_priv->anon_filp);
-
 	kfree(file->driver_priv);
 	file->driver_priv = NULL;
 }
@@ -282,11 +276,6 @@ static const struct vm_operations_struct exynos_drm_gem_vm_ops = {
 static const struct drm_ioctl_desc exynos_ioctls[] = {
 	DRM_IOCTL_DEF_DRV(EXYNOS_GEM_CREATE, exynos_drm_gem_create_ioctl,
 			DRM_UNLOCKED | DRM_AUTH),
-	DRM_IOCTL_DEF_DRV(EXYNOS_GEM_MAP_OFFSET,
-			exynos_drm_gem_map_offset_ioctl, DRM_UNLOCKED |
-			DRM_AUTH),
-	DRM_IOCTL_DEF_DRV(EXYNOS_GEM_MMAP,
-			exynos_drm_gem_mmap_ioctl, DRM_UNLOCKED | DRM_AUTH),
 	DRM_IOCTL_DEF_DRV(EXYNOS_GEM_GET,
 			exynos_drm_gem_get_ioctl, DRM_UNLOCKED),
 	DRM_IOCTL_DEF_DRV(EXYNOS_VIDI_CONNECTION,
@@ -330,6 +319,7 @@ static struct drm_driver exynos_drm_driver = {
 	.preclose		= exynos_drm_preclose,
 	.lastclose		= exynos_drm_lastclose,
 	.postclose		= exynos_drm_postclose,
+	.set_busid		= drm_platform_set_busid,
 	.get_vblank_counter	= drm_vblank_count,
 	.enable_vblank		= exynos_drm_crtc_enable_vblank,
 	.disable_vblank		= exynos_drm_crtc_disable_vblank,
@@ -485,21 +475,20 @@ void exynos_drm_component_del(struct device *dev,
 	mutex_unlock(&drm_component_lock);
 }
 
-static int compare_of(struct device *dev, void *data)
+static int compare_dev(struct device *dev, void *data)
 {
 	return dev == (struct device *)data;
 }
 
-static int exynos_drm_add_components(struct device *dev, struct master *m)
+static struct component_match *exynos_drm_match_add(struct device *dev)
 {
+	struct component_match *match = NULL;
 	struct component_dev *cdev;
 	unsigned int attach_cnt = 0;
 
 	mutex_lock(&drm_component_lock);
 
 	list_for_each_entry(cdev, &drm_component_list, list) {
-		int ret;
-
 		/*
 		 * Add components to master only in case that crtc and
 		 * encoder/connector device objects exist.
@@ -514,16 +503,10 @@ static int exynos_drm_add_components(struct device *dev, struct master *m)
 		/*
 		 * fimd and dpi modules have same device object so add
 		 * only crtc device object in this case.
-		 *
-		 * TODO. if dpi module follows driver-model driver then
-		 * below codes can be removed.
 		 */
 		if (cdev->crtc_dev == cdev->conn_dev) {
-			ret = component_master_add_child(m, compare_of,
-					cdev->crtc_dev);
-			if (ret < 0)
-				return ret;
-
+			component_match_add(dev, &match, compare_dev,
+						cdev->crtc_dev);
 			goto out_lock;
 		}
 
@@ -533,11 +516,8 @@ static int exynos_drm_add_components(struct device *dev, struct master *m)
 		 * connector/encoder need pipe number of crtc when they
 		 * are created.
 		 */
-		ret = component_master_add_child(m, compare_of, cdev->crtc_dev);
-		ret |= component_master_add_child(m, compare_of,
-							cdev->conn_dev);
-		if (ret < 0)
-			return ret;
+		component_match_add(dev, &match, compare_dev, cdev->crtc_dev);
+		component_match_add(dev, &match, compare_dev, cdev->conn_dev);
 
 out_lock:
 		mutex_lock(&drm_component_lock);
@@ -545,7 +525,7 @@ out_lock:
 
 	mutex_unlock(&drm_component_lock);
 
-	return attach_cnt ? 0 : -ENODEV;
+	return attach_cnt ? match : ERR_PTR(-EPROBE_DEFER);
 }
 
 static int exynos_drm_bind(struct device *dev)
@@ -559,13 +539,13 @@ static void exynos_drm_unbind(struct device *dev)
 }
 
 static const struct component_master_ops exynos_drm_ops = {
-	.add_components = exynos_drm_add_components,
 	.bind		= exynos_drm_bind,
 	.unbind		= exynos_drm_unbind,
 };
 
 static int exynos_drm_platform_probe(struct platform_device *pdev)
 {
+	struct component_match *match;
 	int ret;
 
 	pdev->dev.coherent_dma_mask = DMA_BIT_MASK(32);
@@ -632,13 +612,23 @@ static int exynos_drm_platform_probe(struct platform_device *pdev)
 		goto err_unregister_ipp_drv;
 #endif
 
-	ret = component_master_add(&pdev->dev, &exynos_drm_ops);
+	match = exynos_drm_match_add(&pdev->dev);
+	if (IS_ERR(match)) {
+		ret = PTR_ERR(match);
+		goto err_unregister_resources;
+	}
+
+	ret = component_master_add_with_match(&pdev->dev, &exynos_drm_ops,
+						match);
 	if (ret < 0)
-		DRM_DEBUG_KMS("re-tried by last sub driver probed later.\n");
+		goto err_unregister_resources;
 
-	return 0;
+	return ret;
+
+err_unregister_resources:
 
 #ifdef CONFIG_DRM_EXYNOS_IPP
+	exynos_platform_device_ipp_unregister();
 err_unregister_ipp_drv:
 	platform_driver_unregister(&ipp_driver);
 err_unregister_gsc_drv:
diff --git a/drivers/gpu/drm/exynos/exynos_drm_drv.h b/drivers/gpu/drm/exynos/exynos_drm_drv.h
index 69a6fa397d75..d22e640f59a0 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_drv.h
+++ b/drivers/gpu/drm/exynos/exynos_drm_drv.h
@@ -240,7 +240,6 @@ struct exynos_drm_g2d_private {
 struct drm_exynos_file_private {
 	struct exynos_drm_g2d_private	*g2d_priv;
 	struct device			*ipp_dev;
-	struct file			*anon_filp;
 };
 
 /*
diff --git a/drivers/gpu/drm/exynos/exynos_drm_dsi.c b/drivers/gpu/drm/exynos/exynos_drm_dsi.c
index 442aa2d00132..24741d8758e8 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_dsi.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_dsi.c
@@ -114,6 +114,8 @@
 #define DSIM_SYNC_INFORM		(1 << 27)
 #define DSIM_EOT_DISABLE		(1 << 28)
 #define DSIM_MFLUSH_VS			(1 << 29)
+/* This flag is valid only for exynos3250/3472/4415/5260/5430 */
+#define DSIM_CLKLANE_STOP		(1 << 30)
 
 /* DSIM_ESCMODE */
 #define DSIM_TX_TRIGGER_RST		(1 << 4)
@@ -262,6 +264,7 @@ struct exynos_dsi_driver_data {
 	unsigned int plltmr_reg;
 
 	unsigned int has_freqband:1;
+	unsigned int has_clklane_stop:1;
 };
 
 struct exynos_dsi {
@@ -301,9 +304,16 @@ struct exynos_dsi {
 #define host_to_dsi(host) container_of(host, struct exynos_dsi, dsi_host)
 #define connector_to_dsi(c) container_of(c, struct exynos_dsi, connector)
 
+static struct exynos_dsi_driver_data exynos3_dsi_driver_data = {
+	.plltmr_reg = 0x50,
+	.has_freqband = 1,
+	.has_clklane_stop = 1,
+};
+
 static struct exynos_dsi_driver_data exynos4_dsi_driver_data = {
 	.plltmr_reg = 0x50,
 	.has_freqband = 1,
+	.has_clklane_stop = 1,
 };
 
 static struct exynos_dsi_driver_data exynos5_dsi_driver_data = {
@@ -311,6 +321,8 @@ static struct exynos_dsi_driver_data exynos5_dsi_driver_data = {
 };
 
 static struct of_device_id exynos_dsi_of_match[] = {
+	{ .compatible = "samsung,exynos3250-mipi-dsi",
+	  .data = &exynos3_dsi_driver_data },
 	{ .compatible = "samsung,exynos4210-mipi-dsi",
 	  .data = &exynos4_dsi_driver_data },
 	{ .compatible = "samsung,exynos5410-mipi-dsi",
@@ -421,7 +433,7 @@ static unsigned long exynos_dsi_set_pll(struct exynos_dsi *dsi,
 	if (!fout) {
 		dev_err(dsi->dev,
 			"failed to find PLL PMS for requested frequency\n");
-		return -EFAULT;
+		return 0;
 	}
 	dev_dbg(dsi->dev, "PLL freq %lu, (p %d, m %d, s %d)\n", fout, p, m, s);
 
@@ -453,7 +465,7 @@ static unsigned long exynos_dsi_set_pll(struct exynos_dsi *dsi,
 	do {
 		if (timeout-- == 0) {
 			dev_err(dsi->dev, "PLL failed to stabilize\n");
-			return -EFAULT;
+			return 0;
 		}
 		reg = readl(dsi->reg_base + DSIM_STATUS_REG);
 	} while ((reg & DSIM_PLL_STABLE) == 0);
@@ -569,6 +581,7 @@ static void exynos_dsi_disable_clock(struct exynos_dsi *dsi)
 
 static int exynos_dsi_init_link(struct exynos_dsi *dsi)
 {
+	struct exynos_dsi_driver_data *driver_data = dsi->driver_data;
 	int timeout;
 	u32 reg;
 	u32 lanes_mask;
@@ -650,6 +663,20 @@ static int exynos_dsi_init_link(struct exynos_dsi *dsi)
 	reg |= DSIM_LANE_EN(lanes_mask);
 	writel(reg, dsi->reg_base + DSIM_CONFIG_REG);
 
+	/*
+	 * Use non-continuous clock mode if the periparal wants and
+	 * host controller supports
+	 *
+	 * In non-continous clock mode, host controller will turn off
+	 * the HS clock between high-speed transmissions to reduce
+	 * power consumption.
+	 */
+	if (driver_data->has_clklane_stop &&
+			dsi->mode_flags & MIPI_DSI_CLOCK_NON_CONTINUOUS) {
+		reg |= DSIM_CLKLANE_STOP;
+		writel(reg, dsi->reg_base + DSIM_CONFIG_REG);
+	}
+
 	/* Check clock and data lane state are stop state */
 	timeout = 100;
 	do {
@@ -1414,6 +1441,9 @@ exynos_dsi_detect(struct drm_connector *connector, bool force)
 
 static void exynos_dsi_connector_destroy(struct drm_connector *connector)
 {
+	drm_connector_unregister(connector);
+	drm_connector_cleanup(connector);
+	connector->dev = NULL;
 }
 
 static struct drm_connector_funcs exynos_dsi_connector_funcs = {
@@ -1634,10 +1664,10 @@ static void exynos_dsi_unbind(struct device *dev, struct device *master,
 
 	exynos_dsi_dpms(&exynos_dsi_display, DRM_MODE_DPMS_OFF);
 
-	mipi_dsi_host_unregister(&dsi->dsi_host);
-
+	exynos_dsi_connector_destroy(&dsi->connector);
 	encoder->funcs->destroy(encoder);
-	drm_connector_cleanup(&dsi->connector);
+
+	mipi_dsi_host_unregister(&dsi->dsi_host);
 }
 
 static const struct component_ops exynos_dsi_component_ops = {
diff --git a/drivers/gpu/drm/exynos/exynos_drm_fb.c b/drivers/gpu/drm/exynos/exynos_drm_fb.c
index 65a22cad7b36..d346d1e6eda0 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_fb.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_fb.c
@@ -165,6 +165,7 @@ exynos_drm_framebuffer_init(struct drm_device *dev,
 
 	ret = drm_framebuffer_init(dev, &exynos_fb->fb, &exynos_drm_fb_funcs);
 	if (ret) {
+		kfree(exynos_fb);
 		DRM_ERROR("failed to initialize framebuffer\n");
 		return ERR_PTR(ret);
 	}
diff --git a/drivers/gpu/drm/exynos/exynos_drm_fbdev.c b/drivers/gpu/drm/exynos/exynos_drm_fbdev.c
index 32e63f60e1d1..e12ea90c6237 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_fbdev.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_fbdev.c
@@ -123,6 +123,7 @@ static int exynos_drm_fbdev_update(struct drm_fb_helper *helper,
 
 	fbi->screen_base = buffer->kvaddr + offset;
 	fbi->screen_size = size;
+	fbi->fix.smem_len = size;
 
 	return 0;
 }
@@ -353,9 +354,6 @@ void exynos_drm_fbdev_fini(struct drm_device *dev)
 
 	fbdev = to_exynos_fbdev(private->fb_helper);
 
-	if (fbdev->exynos_gem_obj)
-		exynos_drm_gem_destroy(fbdev->exynos_gem_obj);
-
 	exynos_drm_fbdev_destroy(dev, private->fb_helper);
 	kfree(fbdev);
 	private->fb_helper = NULL;
diff --git a/drivers/gpu/drm/exynos/exynos_drm_fimc.c b/drivers/gpu/drm/exynos/exynos_drm_fimc.c
index ec7cc9ea50df..68d38eb6774d 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_fimc.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_fimc.c
@@ -336,9 +336,6 @@ static bool fimc_check_ovf(struct fimc_context *ctx)
 		fimc_set_bits(ctx, EXYNOS_CIWDOFST,
 			EXYNOS_CIWDOFST_CLROVFIY | EXYNOS_CIWDOFST_CLROVFICB |
 			EXYNOS_CIWDOFST_CLROVFICR);
-		fimc_clear_bits(ctx, EXYNOS_CIWDOFST,
-			EXYNOS_CIWDOFST_CLROVFIY | EXYNOS_CIWDOFST_CLROVFICB |
-			EXYNOS_CIWDOFST_CLROVFICR);
 
 		dev_err(ippdrv->dev, "occurred overflow at %d, status 0x%x.\n",
 			ctx->id, status);
@@ -718,24 +715,24 @@ static int fimc_src_set_addr(struct device *dev,
 	case IPP_BUF_ENQUEUE:
 		config = &property->config[EXYNOS_DRM_OPS_SRC];
 		fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_Y],
-			EXYNOS_CIIYSA(buf_id));
+			EXYNOS_CIIYSA0);
 
 		if (config->fmt == DRM_FORMAT_YVU420) {
 			fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_CR],
-				EXYNOS_CIICBSA(buf_id));
+				EXYNOS_CIICBSA0);
 			fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_CB],
-				EXYNOS_CIICRSA(buf_id));
+				EXYNOS_CIICRSA0);
 		} else {
 			fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_CB],
-				EXYNOS_CIICBSA(buf_id));
+				EXYNOS_CIICBSA0);
 			fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_CR],
-				EXYNOS_CIICRSA(buf_id));
+				EXYNOS_CIICRSA0);
 		}
 		break;
 	case IPP_BUF_DEQUEUE:
-		fimc_write(ctx, 0x0, EXYNOS_CIIYSA(buf_id));
-		fimc_write(ctx, 0x0, EXYNOS_CIICBSA(buf_id));
-		fimc_write(ctx, 0x0, EXYNOS_CIICRSA(buf_id));
+		fimc_write(ctx, 0x0, EXYNOS_CIIYSA0);
+		fimc_write(ctx, 0x0, EXYNOS_CIICBSA0);
+		fimc_write(ctx, 0x0, EXYNOS_CIICRSA0);
 		break;
 	default:
 		/* bypass */
@@ -1122,67 +1119,34 @@ static int fimc_dst_set_size(struct device *dev, int swap,
 	return 0;
 }
 
-static int fimc_dst_get_buf_count(struct fimc_context *ctx)
-{
-	u32 cfg, buf_num;
-
-	cfg = fimc_read(ctx, EXYNOS_CIFCNTSEQ);
-
-	buf_num = hweight32(cfg);
-
-	DRM_DEBUG_KMS("buf_num[%d]\n", buf_num);
-
-	return buf_num;
-}
-
-static int fimc_dst_set_buf_seq(struct fimc_context *ctx, u32 buf_id,
+static void fimc_dst_set_buf_seq(struct fimc_context *ctx, u32 buf_id,
 		enum drm_exynos_ipp_buf_type buf_type)
 {
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
-	bool enable;
-	u32 cfg;
-	u32 mask = 0x00000001 << buf_id;
-	int ret = 0;
 	unsigned long flags;
+	u32 buf_num;
+	u32 cfg;
 
 	DRM_DEBUG_KMS("buf_id[%d]buf_type[%d]\n", buf_id, buf_type);
 
 	spin_lock_irqsave(&ctx->lock, flags);
 
-	/* mask register set */
 	cfg = fimc_read(ctx, EXYNOS_CIFCNTSEQ);
 
-	switch (buf_type) {
-	case IPP_BUF_ENQUEUE:
-		enable = true;
-		break;
-	case IPP_BUF_DEQUEUE:
-		enable = false;
-		break;
-	default:
-		dev_err(ippdrv->dev, "invalid buf ctrl parameter.\n");
-		ret =  -EINVAL;
-		goto err_unlock;
-	}
+	if (buf_type == IPP_BUF_ENQUEUE)
+		cfg |= (1 << buf_id);
+	else
+		cfg &= ~(1 << buf_id);
 
-	/* sequence id */
-	cfg &= ~mask;
-	cfg |= (enable << buf_id);
 	fimc_write(ctx, cfg, EXYNOS_CIFCNTSEQ);
 
-	/* interrupt enable */
-	if (buf_type == IPP_BUF_ENQUEUE &&
-	    fimc_dst_get_buf_count(ctx) >= FIMC_BUF_START)
-		fimc_mask_irq(ctx, true);
+	buf_num = hweight32(cfg);
 
-	/* interrupt disable */
-	if (buf_type == IPP_BUF_DEQUEUE &&
-	    fimc_dst_get_buf_count(ctx) <= FIMC_BUF_STOP)
+	if (buf_type == IPP_BUF_ENQUEUE && buf_num >= FIMC_BUF_START)
+		fimc_mask_irq(ctx, true);
+	else if (buf_type == IPP_BUF_DEQUEUE && buf_num <= FIMC_BUF_STOP)
 		fimc_mask_irq(ctx, false);
 
-err_unlock:
 	spin_unlock_irqrestore(&ctx->lock, flags);
-	return ret;
 }
 
 static int fimc_dst_set_addr(struct device *dev,
@@ -1240,7 +1204,9 @@ static int fimc_dst_set_addr(struct device *dev,
 		break;
 	}
 
-	return fimc_dst_set_buf_seq(ctx, buf_id, buf_type);
+	fimc_dst_set_buf_seq(ctx, buf_id, buf_type);
+
+	return 0;
 }
 
 static struct exynos_drm_ipp_ops fimc_dst_ops = {
@@ -1291,14 +1257,11 @@ static irqreturn_t fimc_irq_handler(int irq, void *dev_id)
 
 	DRM_DEBUG_KMS("buf_id[%d]\n", buf_id);
 
-	if (fimc_dst_set_buf_seq(ctx, buf_id, IPP_BUF_DEQUEUE) < 0) {
-		DRM_ERROR("failed to dequeue.\n");
-		return IRQ_HANDLED;
-	}
+	fimc_dst_set_buf_seq(ctx, buf_id, IPP_BUF_DEQUEUE);
 
 	event_work->ippdrv = ippdrv;
 	event_work->buf_id[EXYNOS_DRM_OPS_DST] = buf_id;
-	queue_work(ippdrv->event_workq, (struct work_struct *)event_work);
+	queue_work(ippdrv->event_workq, &event_work->work);
 
 	return IRQ_HANDLED;
 }
@@ -1590,11 +1553,8 @@ static int fimc_ippdrv_start(struct device *dev, enum drm_exynos_ipp_cmd cmd)
 
 	fimc_clear_bits(ctx, EXYNOS_CIOCTRL, EXYNOS_CIOCTRL_WEAVE_MASK);
 
-	if (cmd == IPP_CMD_M2M) {
-		fimc_set_bits(ctx, EXYNOS_MSCTRL, EXYNOS_MSCTRL_ENVID);
-
+	if (cmd == IPP_CMD_M2M)
 		fimc_set_bits(ctx, EXYNOS_MSCTRL, EXYNOS_MSCTRL_ENVID);
-	}
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/exynos/exynos_drm_fimd.c b/drivers/gpu/drm/exynos/exynos_drm_fimd.c
index 5d09e33fef87..085b066a9993 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_fimd.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_fimd.c
@@ -104,6 +104,14 @@ static struct fimd_driver_data s3c64xx_fimd_driver_data = {
 	.has_limited_fmt = 1,
 };
 
+static struct fimd_driver_data exynos3_fimd_driver_data = {
+	.timing_base = 0x20000,
+	.lcdblk_offset = 0x210,
+	.lcdblk_bypass_shift = 1,
+	.has_shadowcon = 1,
+	.has_vidoutcon = 1,
+};
+
 static struct fimd_driver_data exynos4_fimd_driver_data = {
 	.timing_base = 0x0,
 	.lcdblk_offset = 0x210,
@@ -168,6 +176,8 @@ struct fimd_context {
 static const struct of_device_id fimd_driver_dt_match[] = {
 	{ .compatible = "samsung,s3c6400-fimd",
 	  .data = &s3c64xx_fimd_driver_data },
+	{ .compatible = "samsung,exynos3250-fimd",
+	  .data = &exynos3_fimd_driver_data },
 	{ .compatible = "samsung,exynos4210-fimd",
 	  .data = &exynos4_fimd_driver_data },
 	{ .compatible = "samsung,exynos5250-fimd",
@@ -204,7 +214,6 @@ static void fimd_wait_for_vblank(struct exynos_drm_manager *mgr)
 		DRM_DEBUG_KMS("vblank wait timed out.\n");
 }
 
-
 static void fimd_clear_channel(struct exynos_drm_manager *mgr)
 {
 	struct fimd_context *ctx = mgr->ctx;
@@ -214,17 +223,31 @@ static void fimd_clear_channel(struct exynos_drm_manager *mgr)
 
 	/* Check if any channel is enabled. */
 	for (win = 0; win < WINDOWS_NR; win++) {
-		u32 val = readl(ctx->regs + SHADOWCON);
-		if (val & SHADOWCON_CHx_ENABLE(win)) {
-			val &= ~SHADOWCON_CHx_ENABLE(win);
-			writel(val, ctx->regs + SHADOWCON);
+		u32 val = readl(ctx->regs + WINCON(win));
+
+		if (val & WINCONx_ENWIN) {
+			/* wincon */
+			val &= ~WINCONx_ENWIN;
+			writel(val, ctx->regs + WINCON(win));
+
+			/* unprotect windows */
+			if (ctx->driver_data->has_shadowcon) {
+				val = readl(ctx->regs + SHADOWCON);
+				val &= ~SHADOWCON_CHx_ENABLE(win);
+				writel(val, ctx->regs + SHADOWCON);
+			}
 			ch_enabled = 1;
 		}
 	}
 
 	/* Wait for vsync, as disable channel takes effect at next vsync */
-	if (ch_enabled)
+	if (ch_enabled) {
+		unsigned int state = ctx->suspended;
+
+		ctx->suspended = 0;
 		fimd_wait_for_vblank(mgr);
+		ctx->suspended = state;
+	}
 }
 
 static int fimd_mgr_initialize(struct exynos_drm_manager *mgr,
@@ -237,23 +260,6 @@ static int fimd_mgr_initialize(struct exynos_drm_manager *mgr,
 	mgr->drm_dev = ctx->drm_dev = drm_dev;
 	mgr->pipe = ctx->pipe = priv->pipe++;
 
-	/*
-	 * enable drm irq mode.
-	 * - with irq_enabled = true, we can use the vblank feature.
-	 *
-	 * P.S. note that we wouldn't use drm irq handler but
-	 *	just specific driver own one instead because
-	 *	drm framework supports only one irq handler.
-	 */
-	drm_dev->irq_enabled = true;
-
-	/*
-	 * with vblank_disable_allowed = true, vblank interrupt will be disabled
-	 * by drm timer once a current process gives up ownership of
-	 * vblank event.(after drm_vblank_put function is called)
-	 */
-	drm_dev->vblank_disable_allowed = true;
-
 	/* attach this sub driver to iommu mapping if supported. */
 	if (is_drm_iommu_supported(ctx->drm_dev)) {
 		/*
@@ -1051,7 +1057,6 @@ static void fimd_unbind(struct device *dev, struct device *master,
 {
 	struct exynos_drm_manager *mgr = dev_get_drvdata(dev);
 	struct fimd_context *ctx = fimd_manager.ctx;
-	struct drm_crtc *crtc = mgr->crtc;
 
 	fimd_dpms(mgr, DRM_MODE_DPMS_OFF);
 
@@ -1059,8 +1064,6 @@ static void fimd_unbind(struct device *dev, struct device *master,
 		exynos_dpi_remove(dev);
 
 	fimd_mgr_remove(mgr);
-
-	crtc->funcs->destroy(crtc);
 }
 
 static const struct component_ops fimd_component_ops = {
diff --git a/drivers/gpu/drm/exynos/exynos_drm_gem.c b/drivers/gpu/drm/exynos/exynos_drm_gem.c
index 15db80138382..0d5b9698d384 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_gem.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_gem.c
@@ -318,40 +318,16 @@ void exynos_drm_gem_put_dma_addr(struct drm_device *dev,
 	drm_gem_object_unreference_unlocked(obj);
 }
 
-int exynos_drm_gem_map_offset_ioctl(struct drm_device *dev, void *data,
-				    struct drm_file *file_priv)
-{
-	struct drm_exynos_gem_map_off *args = data;
-
-	DRM_DEBUG_KMS("handle = 0x%x, offset = 0x%lx\n",
-			args->handle, (unsigned long)args->offset);
-
-	if (!(dev->driver->driver_features & DRIVER_GEM)) {
-		DRM_ERROR("does not support GEM.\n");
-		return -ENODEV;
-	}
-
-	return exynos_drm_gem_dumb_map_offset(file_priv, dev, args->handle,
-			&args->offset);
-}
-
-int exynos_drm_gem_mmap_buffer(struct file *filp,
+int exynos_drm_gem_mmap_buffer(struct exynos_drm_gem_obj *exynos_gem_obj,
 				      struct vm_area_struct *vma)
 {
-	struct drm_gem_object *obj = filp->private_data;
-	struct exynos_drm_gem_obj *exynos_gem_obj = to_exynos_gem_obj(obj);
-	struct drm_device *drm_dev = obj->dev;
+	struct drm_device *drm_dev = exynos_gem_obj->base.dev;
 	struct exynos_drm_gem_buf *buffer;
 	unsigned long vm_size;
 	int ret;
 
-	WARN_ON(!mutex_is_locked(&obj->dev->struct_mutex));
-
-	vma->vm_flags |= VM_IO | VM_DONTEXPAND | VM_DONTDUMP;
-	vma->vm_private_data = obj;
-	vma->vm_ops = drm_dev->driver->gem_vm_ops;
-
-	update_vm_cache_attr(exynos_gem_obj, vma);
+	vma->vm_flags &= ~VM_PFNMAP;
+	vma->vm_pgoff = 0;
 
 	vm_size = vma->vm_end - vma->vm_start;
 
@@ -373,60 +349,6 @@ int exynos_drm_gem_mmap_buffer(struct file *filp,
 		return ret;
 	}
 
-	/*
-	 * take a reference to this mapping of the object. And this reference
-	 * is unreferenced by the corresponding vm_close call.
-	 */
-	drm_gem_object_reference(obj);
-
-	drm_vm_open_locked(drm_dev, vma);
-
-	return 0;
-}
-
-int exynos_drm_gem_mmap_ioctl(struct drm_device *dev, void *data,
-			      struct drm_file *file_priv)
-{
-	struct drm_exynos_file_private *exynos_file_priv;
-	struct drm_exynos_gem_mmap *args = data;
-	struct drm_gem_object *obj;
-	struct file *anon_filp;
-	unsigned long addr;
-
-	if (!(dev->driver->driver_features & DRIVER_GEM)) {
-		DRM_ERROR("does not support GEM.\n");
-		return -ENODEV;
-	}
-
-	mutex_lock(&dev->struct_mutex);
-
-	obj = drm_gem_object_lookup(dev, file_priv, args->handle);
-	if (!obj) {
-		DRM_ERROR("failed to lookup gem object.\n");
-		mutex_unlock(&dev->struct_mutex);
-		return -EINVAL;
-	}
-
-	exynos_file_priv = file_priv->driver_priv;
-	anon_filp = exynos_file_priv->anon_filp;
-	anon_filp->private_data = obj;
-
-	addr = vm_mmap(anon_filp, 0, args->size, PROT_READ | PROT_WRITE,
-			MAP_SHARED, 0);
-
-	drm_gem_object_unreference(obj);
-
-	if (IS_ERR_VALUE(addr)) {
-		mutex_unlock(&dev->struct_mutex);
-		return (int)addr;
-	}
-
-	mutex_unlock(&dev->struct_mutex);
-
-	args->mapped = addr;
-
-	DRM_DEBUG_KMS("mapped = 0x%lx\n", (unsigned long)args->mapped);
-
 	return 0;
 }
 
@@ -710,16 +632,20 @@ int exynos_drm_gem_mmap(struct file *filp, struct vm_area_struct *vma)
 	exynos_gem_obj = to_exynos_gem_obj(obj);
 
 	ret = check_gem_flags(exynos_gem_obj->flags);
-	if (ret) {
-		drm_gem_vm_close(vma);
-		drm_gem_free_mmap_offset(obj);
-		return ret;
-	}
-
-	vma->vm_flags &= ~VM_PFNMAP;
-	vma->vm_flags |= VM_MIXEDMAP;
+	if (ret)
+		goto err_close_vm;
 
 	update_vm_cache_attr(exynos_gem_obj, vma);
 
+	ret = exynos_drm_gem_mmap_buffer(exynos_gem_obj, vma);
+	if (ret)
+		goto err_close_vm;
+
+	return ret;
+
+err_close_vm:
+	drm_gem_vm_close(vma);
+	drm_gem_free_mmap_offset(obj);
+
 	return ret;
 }
diff --git a/drivers/gpu/drm/exynos/exynos_drm_gem.h b/drivers/gpu/drm/exynos/exynos_drm_gem.h
index 1592c0ba7de8..ec58fe9c40df 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_gem.h
+++ b/drivers/gpu/drm/exynos/exynos_drm_gem.h
@@ -12,6 +12,8 @@
 #ifndef _EXYNOS_DRM_GEM_H_
 #define _EXYNOS_DRM_GEM_H_
 
+#include <drm/drm_gem.h>
+
 #define to_exynos_gem_obj(x)	container_of(x,\
 			struct exynos_drm_gem_obj, base)
 
@@ -111,20 +113,6 @@ void exynos_drm_gem_put_dma_addr(struct drm_device *dev,
 					unsigned int gem_handle,
 					struct drm_file *filp);
 
-/* get buffer offset to map to user space. */
-int exynos_drm_gem_map_offset_ioctl(struct drm_device *dev, void *data,
-				    struct drm_file *file_priv);
-
-/*
- * mmap the physically continuous memory that a gem object contains
- * to user space.
- */
-int exynos_drm_gem_mmap_ioctl(struct drm_device *dev, void *data,
-			      struct drm_file *file_priv);
-
-int exynos_drm_gem_mmap_buffer(struct file *filp,
-				      struct vm_area_struct *vma);
-
 /* map user space allocated by malloc to pages. */
 int exynos_drm_gem_userptr_ioctl(struct drm_device *dev, void *data,
 				      struct drm_file *file_priv);
diff --git a/drivers/gpu/drm/exynos/exynos_drm_gsc.c b/drivers/gpu/drm/exynos/exynos_drm_gsc.c
index 9e3ff1672965..c6a013fc321c 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_gsc.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_gsc.c
@@ -1326,8 +1326,7 @@ static irqreturn_t gsc_irq_handler(int irq, void *dev_id)
 			buf_id[EXYNOS_DRM_OPS_SRC];
 		event_work->buf_id[EXYNOS_DRM_OPS_DST] =
 			buf_id[EXYNOS_DRM_OPS_DST];
-		queue_work(ippdrv->event_workq,
-			(struct work_struct *)event_work);
+		queue_work(ippdrv->event_workq, &event_work->work);
 	}
 
 	return IRQ_HANDLED;
diff --git a/drivers/gpu/drm/exynos/exynos_drm_ipp.c b/drivers/gpu/drm/exynos/exynos_drm_ipp.c
index c411399070d6..00d74b18f7cb 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_ipp.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_ipp.c
@@ -75,7 +75,6 @@ struct drm_exynos_ipp_mem_node {
 	u32	prop_id;
 	u32	buf_id;
 	struct drm_exynos_ipp_buf_info	buf_info;
-	struct drm_file		*filp;
 };
 
 /*
@@ -319,44 +318,6 @@ static void ipp_print_property(struct drm_exynos_ipp_property *property,
 		sz->hsize, sz->vsize, config->flip, config->degree);
 }
 
-static int ipp_find_and_set_property(struct drm_exynos_ipp_property *property)
-{
-	struct exynos_drm_ippdrv *ippdrv;
-	struct drm_exynos_ipp_cmd_node *c_node;
-	u32 prop_id = property->prop_id;
-
-	DRM_DEBUG_KMS("prop_id[%d]\n", prop_id);
-
-	ippdrv = ipp_find_drv_by_handle(prop_id);
-	if (IS_ERR(ippdrv)) {
-		DRM_ERROR("failed to get ipp driver.\n");
-		return -EINVAL;
-	}
-
-	/*
-	 * Find command node using command list in ippdrv.
-	 * when we find this command no using prop_id.
-	 * return property information set in this command node.
-	 */
-	mutex_lock(&ippdrv->cmd_lock);
-	list_for_each_entry(c_node, &ippdrv->cmd_list, list) {
-		if ((c_node->property.prop_id == prop_id) &&
-		    (c_node->state == IPP_STATE_STOP)) {
-			mutex_unlock(&ippdrv->cmd_lock);
-			DRM_DEBUG_KMS("found cmd[%d]ippdrv[0x%x]\n",
-				property->cmd, (int)ippdrv);
-
-			c_node->property = *property;
-			return 0;
-		}
-	}
-	mutex_unlock(&ippdrv->cmd_lock);
-
-	DRM_ERROR("failed to search property.\n");
-
-	return -EINVAL;
-}
-
 static struct drm_exynos_ipp_cmd_work *ipp_create_cmd_work(void)
 {
 	struct drm_exynos_ipp_cmd_work *cmd_work;
@@ -392,6 +353,7 @@ int exynos_drm_ipp_set_property(struct drm_device *drm_dev, void *data,
 	struct drm_exynos_ipp_property *property = data;
 	struct exynos_drm_ippdrv *ippdrv;
 	struct drm_exynos_ipp_cmd_node *c_node;
+	u32 prop_id;
 	int ret, i;
 
 	if (!ctx) {
@@ -404,6 +366,8 @@ int exynos_drm_ipp_set_property(struct drm_device *drm_dev, void *data,
 		return -EINVAL;
 	}
 
+	prop_id = property->prop_id;
+
 	/*
 	 * This is log print for user application property.
 	 * user application set various property.
@@ -412,14 +376,24 @@ int exynos_drm_ipp_set_property(struct drm_device *drm_dev, void *data,
 		ipp_print_property(property, i);
 
 	/*
-	 * set property ioctl generated new prop_id.
-	 * but in this case already asigned prop_id using old set property.
-	 * e.g PAUSE state. this case supports find current prop_id and use it
-	 * instead of allocation.
+	 * In case prop_id is not zero try to set existing property.
 	 */
-	if (property->prop_id) {
-		DRM_DEBUG_KMS("prop_id[%d]\n", property->prop_id);
-		return ipp_find_and_set_property(property);
+	if (prop_id) {
+		c_node = ipp_find_obj(&ctx->prop_idr, &ctx->prop_lock, prop_id);
+
+		if (!c_node || c_node->filp != file) {
+			DRM_DEBUG_KMS("prop_id[%d] not found\n", prop_id);
+			return -EINVAL;
+		}
+
+		if (c_node->state != IPP_STATE_STOP) {
+			DRM_DEBUG_KMS("prop_id[%d] not stopped\n", prop_id);
+			return -EINVAL;
+		}
+
+		c_node->property = *property;
+
+		return 0;
 	}
 
 	/* find ipp driver using ipp id */
@@ -445,9 +419,9 @@ int exynos_drm_ipp_set_property(struct drm_device *drm_dev, void *data,
 		property->prop_id, property->cmd, (int)ippdrv);
 
 	/* stored property information and ippdrv in private data */
-	c_node->dev = dev;
 	c_node->property = *property;
 	c_node->state = IPP_STATE_IDLE;
+	c_node->filp = file;
 
 	c_node->start_work = ipp_create_cmd_work();
 	if (IS_ERR(c_node->start_work)) {
@@ -499,105 +473,37 @@ err_clear:
 	return ret;
 }
 
-static void ipp_clean_cmd_node(struct ipp_context *ctx,
-				struct drm_exynos_ipp_cmd_node *c_node)
-{
-	/* delete list */
-	list_del(&c_node->list);
-
-	ipp_remove_id(&ctx->prop_idr, &ctx->prop_lock,
-			c_node->property.prop_id);
-
-	/* destroy mutex */
-	mutex_destroy(&c_node->lock);
-	mutex_destroy(&c_node->mem_lock);
-	mutex_destroy(&c_node->event_lock);
-
-	/* free command node */
-	kfree(c_node->start_work);
-	kfree(c_node->stop_work);
-	kfree(c_node->event_work);
-	kfree(c_node);
-}
-
-static bool ipp_check_mem_list(struct drm_exynos_ipp_cmd_node *c_node)
-{
-	switch (c_node->property.cmd) {
-	case IPP_CMD_WB:
-		return !list_empty(&c_node->mem_list[EXYNOS_DRM_OPS_DST]);
-	case IPP_CMD_OUTPUT:
-		return !list_empty(&c_node->mem_list[EXYNOS_DRM_OPS_SRC]);
-	case IPP_CMD_M2M:
-	default:
-		return !list_empty(&c_node->mem_list[EXYNOS_DRM_OPS_SRC]) &&
-		       !list_empty(&c_node->mem_list[EXYNOS_DRM_OPS_DST]);
-	}
-}
-
-static struct drm_exynos_ipp_mem_node
-		*ipp_find_mem_node(struct drm_exynos_ipp_cmd_node *c_node,
-		struct drm_exynos_ipp_queue_buf *qbuf)
-{
-	struct drm_exynos_ipp_mem_node *m_node;
-	struct list_head *head;
-	int count = 0;
-
-	DRM_DEBUG_KMS("buf_id[%d]\n", qbuf->buf_id);
-
-	/* source/destination memory list */
-	head = &c_node->mem_list[qbuf->ops_id];
-
-	/* find memory node from memory list */
-	list_for_each_entry(m_node, head, list) {
-		DRM_DEBUG_KMS("count[%d]m_node[0x%x]\n", count++, (int)m_node);
-
-		/* compare buffer id */
-		if (m_node->buf_id == qbuf->buf_id)
-			return m_node;
-	}
-
-	return NULL;
-}
-
-static int ipp_set_mem_node(struct exynos_drm_ippdrv *ippdrv,
+static int ipp_put_mem_node(struct drm_device *drm_dev,
 		struct drm_exynos_ipp_cmd_node *c_node,
 		struct drm_exynos_ipp_mem_node *m_node)
 {
-	struct exynos_drm_ipp_ops *ops = NULL;
-	int ret = 0;
+	int i;
 
 	DRM_DEBUG_KMS("node[0x%x]\n", (int)m_node);
 
 	if (!m_node) {
-		DRM_ERROR("invalid queue node.\n");
+		DRM_ERROR("invalid dequeue node.\n");
 		return -EFAULT;
 	}
 
 	DRM_DEBUG_KMS("ops_id[%d]\n", m_node->ops_id);
 
-	/* get operations callback */
-	ops = ippdrv->ops[m_node->ops_id];
-	if (!ops) {
-		DRM_ERROR("not support ops.\n");
-		return -EFAULT;
+	/* put gem buffer */
+	for_each_ipp_planar(i) {
+		unsigned long handle = m_node->buf_info.handles[i];
+		if (handle)
+			exynos_drm_gem_put_dma_addr(drm_dev, handle,
+							c_node->filp);
 	}
 
-	/* set address and enable irq */
-	if (ops->set_addr) {
-		ret = ops->set_addr(ippdrv->dev, &m_node->buf_info,
-			m_node->buf_id, IPP_BUF_ENQUEUE);
-		if (ret) {
-			DRM_ERROR("failed to set addr.\n");
-			return ret;
-		}
-	}
+	list_del(&m_node->list);
+	kfree(m_node);
 
-	return ret;
+	return 0;
 }
 
 static struct drm_exynos_ipp_mem_node
 		*ipp_get_mem_node(struct drm_device *drm_dev,
-		struct drm_file *file,
 		struct drm_exynos_ipp_cmd_node *c_node,
 		struct drm_exynos_ipp_queue_buf *qbuf)
 {
@@ -615,6 +521,7 @@ static struct drm_exynos_ipp_mem_node
 	m_node->ops_id = qbuf->ops_id;
 	m_node->prop_id = qbuf->prop_id;
 	m_node->buf_id = qbuf->buf_id;
+	INIT_LIST_HEAD(&m_node->list);
 
 	DRM_DEBUG_KMS("m_node[0x%x]ops_id[%d]\n", (int)m_node, qbuf->ops_id);
 	DRM_DEBUG_KMS("prop_id[%d]buf_id[%d]\n", qbuf->prop_id, m_node->buf_id);
@@ -627,10 +534,11 @@ static struct drm_exynos_ipp_mem_node
 			dma_addr_t *addr;
 
 			addr = exynos_drm_gem_get_dma_addr(drm_dev,
-					qbuf->handle[i], file);
+					qbuf->handle[i], c_node->filp);
 			if (IS_ERR(addr)) {
 				DRM_ERROR("failed to get addr.\n");
-				goto err_clear;
+				ipp_put_mem_node(drm_dev, c_node, m_node);
+				return ERR_PTR(-EFAULT);
 			}
 
 			buf_info->handles[i] = qbuf->handle[i];
@@ -640,46 +548,30 @@ static struct drm_exynos_ipp_mem_node
 		}
 	}
 
-	m_node->filp = file;
 	mutex_lock(&c_node->mem_lock);
 	list_add_tail(&m_node->list, &c_node->mem_list[qbuf->ops_id]);
 	mutex_unlock(&c_node->mem_lock);
 
 	return m_node;
-
-err_clear:
-	kfree(m_node);
-	return ERR_PTR(-EFAULT);
 }
 
-static int ipp_put_mem_node(struct drm_device *drm_dev,
-		struct drm_exynos_ipp_cmd_node *c_node,
-		struct drm_exynos_ipp_mem_node *m_node)
+static void ipp_clean_mem_nodes(struct drm_device *drm_dev,
+			       struct drm_exynos_ipp_cmd_node *c_node, int ops)
 {
-	int i;
-
-	DRM_DEBUG_KMS("node[0x%x]\n", (int)m_node);
+	struct drm_exynos_ipp_mem_node *m_node, *tm_node;
+	struct list_head *head = &c_node->mem_list[ops];
 
-	if (!m_node) {
-		DRM_ERROR("invalid dequeue node.\n");
-		return -EFAULT;
-	}
+	mutex_lock(&c_node->mem_lock);
 
-	DRM_DEBUG_KMS("ops_id[%d]\n", m_node->ops_id);
+	list_for_each_entry_safe(m_node, tm_node, head, list) {
+		int ret;
 
-	/* put gem buffer */
-	for_each_ipp_planar(i) {
-		unsigned long handle = m_node->buf_info.handles[i];
-		if (handle)
-			exynos_drm_gem_put_dma_addr(drm_dev, handle,
-							m_node->filp);
+		ret = ipp_put_mem_node(drm_dev, c_node, m_node);
+		if (ret)
+			DRM_ERROR("failed to put m_node.\n");
 	}
 
-	/* delete list in queue */
-	list_del(&m_node->list);
-	kfree(m_node);
-
-	return 0;
+	mutex_unlock(&c_node->mem_lock);
 }
 
 static void ipp_free_event(struct drm_pending_event *event)
@@ -688,7 +580,6 @@ static void ipp_free_event(struct drm_pending_event *event)
 }
 
 static int ipp_get_event(struct drm_device *drm_dev,
-		struct drm_file *file,
 		struct drm_exynos_ipp_cmd_node *c_node,
 		struct drm_exynos_ipp_queue_buf *qbuf)
 {
@@ -700,7 +591,7 @@ static int ipp_get_event(struct drm_device *drm_dev,
 	e = kzalloc(sizeof(*e), GFP_KERNEL);
 	if (!e) {
 		spin_lock_irqsave(&drm_dev->event_lock, flags);
-		file->event_space += sizeof(e->event);
+		c_node->filp->event_space += sizeof(e->event);
 		spin_unlock_irqrestore(&drm_dev->event_lock, flags);
 		return -ENOMEM;
 	}
@@ -712,7 +603,7 @@ static int ipp_get_event(struct drm_device *drm_dev,
 	e->event.prop_id = qbuf->prop_id;
 	e->event.buf_id[EXYNOS_DRM_OPS_DST] = qbuf->buf_id;
 	e->base.event = &e->event.base;
-	e->base.file_priv = file;
+	e->base.file_priv = c_node->filp;
 	e->base.destroy = ipp_free_event;
 	mutex_lock(&c_node->event_lock);
 	list_add_tail(&e->base.link, &c_node->event_list);
@@ -757,6 +648,115 @@ out_unlock:
 	return;
 }
 
+static void ipp_clean_cmd_node(struct ipp_context *ctx,
+				struct drm_exynos_ipp_cmd_node *c_node)
+{
+	int i;
+
+	/* cancel works */
+	cancel_work_sync(&c_node->start_work->work);
+	cancel_work_sync(&c_node->stop_work->work);
+	cancel_work_sync(&c_node->event_work->work);
+
+	/* put event */
+	ipp_put_event(c_node, NULL);
+
+	for_each_ipp_ops(i)
+		ipp_clean_mem_nodes(ctx->subdrv.drm_dev, c_node, i);
+
+	/* delete list */
+	list_del(&c_node->list);
+
+	ipp_remove_id(&ctx->prop_idr, &ctx->prop_lock,
+			c_node->property.prop_id);
+
+	/* destroy mutex */
+	mutex_destroy(&c_node->lock);
+	mutex_destroy(&c_node->mem_lock);
+	mutex_destroy(&c_node->event_lock);
+
+	/* free command node */
+	kfree(c_node->start_work);
+	kfree(c_node->stop_work);
+	kfree(c_node->event_work);
+	kfree(c_node);
+}
+
+static bool ipp_check_mem_list(struct drm_exynos_ipp_cmd_node *c_node)
+{
+	switch (c_node->property.cmd) {
+	case IPP_CMD_WB:
+		return !list_empty(&c_node->mem_list[EXYNOS_DRM_OPS_DST]);
+	case IPP_CMD_OUTPUT:
+		return !list_empty(&c_node->mem_list[EXYNOS_DRM_OPS_SRC]);
+	case IPP_CMD_M2M:
+	default:
+		return !list_empty(&c_node->mem_list[EXYNOS_DRM_OPS_SRC]) &&
+		       !list_empty(&c_node->mem_list[EXYNOS_DRM_OPS_DST]);
+	}
+}
+
+static struct drm_exynos_ipp_mem_node
+		*ipp_find_mem_node(struct drm_exynos_ipp_cmd_node *c_node,
+		struct drm_exynos_ipp_queue_buf *qbuf)
+{
+	struct drm_exynos_ipp_mem_node *m_node;
+	struct list_head *head;
+	int count = 0;
+
+	DRM_DEBUG_KMS("buf_id[%d]\n", qbuf->buf_id);
+
+	/* source/destination memory list */
+	head = &c_node->mem_list[qbuf->ops_id];
+
+	/* find memory node from memory list */
+	list_for_each_entry(m_node, head, list) {
+		DRM_DEBUG_KMS("count[%d]m_node[0x%x]\n", count++, (int)m_node);
+
+		/* compare buffer id */
+		if (m_node->buf_id == qbuf->buf_id)
+			return m_node;
+	}
+
+	return NULL;
+}
+
+static int ipp_set_mem_node(struct exynos_drm_ippdrv *ippdrv,
+		struct drm_exynos_ipp_cmd_node *c_node,
+		struct drm_exynos_ipp_mem_node *m_node)
+{
+	struct exynos_drm_ipp_ops *ops = NULL;
+	int ret = 0;
+
+	DRM_DEBUG_KMS("node[0x%x]\n", (int)m_node);
+
+	if (!m_node) {
+		DRM_ERROR("invalid queue node.\n");
+		return -EFAULT;
+	}
+
+	DRM_DEBUG_KMS("ops_id[%d]\n", m_node->ops_id);
+
+	/* get operations callback */
+	ops = ippdrv->ops[m_node->ops_id];
+	if (!ops) {
+		DRM_ERROR("not support ops.\n");
+		return -EFAULT;
+	}
+
+	/* set address and enable irq */
+	if (ops->set_addr) {
+		ret = ops->set_addr(ippdrv->dev, &m_node->buf_info,
+			m_node->buf_id, IPP_BUF_ENQUEUE);
+		if (ret) {
+			DRM_ERROR("failed to set addr.\n");
+			return ret;
+		}
+	}
+
+	return ret;
+}
+
 static void ipp_handle_cmd_work(struct device *dev,
 		struct exynos_drm_ippdrv *ippdrv,
 		struct drm_exynos_ipp_cmd_work *cmd_work,
@@ -766,7 +766,7 @@ static void ipp_handle_cmd_work(struct device *dev,
 
 	cmd_work->ippdrv = ippdrv;
 	cmd_work->c_node = c_node;
-	queue_work(ctx->cmd_workq, (struct work_struct *)cmd_work);
+	queue_work(ctx->cmd_workq, &cmd_work->work);
 }
 
 static int ipp_queue_buf_with_run(struct device *dev,
@@ -872,7 +872,7 @@ int exynos_drm_ipp_queue_buf(struct drm_device *drm_dev, void *data,
 	/* find command node */
 	c_node = ipp_find_obj(&ctx->prop_idr, &ctx->prop_lock,
 		qbuf->prop_id);
-	if (!c_node) {
+	if (!c_node || c_node->filp != file) {
 		DRM_ERROR("failed to get command node.\n");
 		return -ENODEV;
 	}
@@ -881,7 +881,7 @@ int exynos_drm_ipp_queue_buf(struct drm_device *drm_dev, void *data,
 	switch (qbuf->buf_type) {
 	case IPP_BUF_ENQUEUE:
 		/* get memory node */
-		m_node = ipp_get_mem_node(drm_dev, file, c_node, qbuf);
+		m_node = ipp_get_mem_node(drm_dev, c_node, qbuf);
 		if (IS_ERR(m_node)) {
 			DRM_ERROR("failed to get m_node.\n");
 			return PTR_ERR(m_node);
@@ -894,7 +894,7 @@ int exynos_drm_ipp_queue_buf(struct drm_device *drm_dev, void *data,
 		 */
 		if (qbuf->ops_id == EXYNOS_DRM_OPS_DST) {
 			/* get event for destination buffer */
-			ret = ipp_get_event(drm_dev, file, c_node, qbuf);
+			ret = ipp_get_event(drm_dev, c_node, qbuf);
 			if (ret) {
 				DRM_ERROR("failed to get event.\n");
 				goto err_clean_node;
@@ -1007,7 +1007,7 @@ int exynos_drm_ipp_cmd_ctrl(struct drm_device *drm_dev, void *data,
 
 	c_node = ipp_find_obj(&ctx->prop_idr, &ctx->prop_lock,
 		cmd_ctrl->prop_id);
-	if (!c_node) {
+	if (!c_node || c_node->filp != file) {
 		DRM_ERROR("invalid command node list.\n");
 		return -ENODEV;
 	}
@@ -1257,80 +1257,39 @@ static int ipp_stop_property(struct drm_device *drm_dev,
 		struct exynos_drm_ippdrv *ippdrv,
 		struct drm_exynos_ipp_cmd_node *c_node)
 {
-	struct drm_exynos_ipp_mem_node *m_node, *tm_node;
 	struct drm_exynos_ipp_property *property = &c_node->property;
-	struct list_head *head;
-	int ret = 0, i;
+	int i;
 
 	DRM_DEBUG_KMS("prop_id[%d]\n", property->prop_id);
 
-	/* put event */
-	ipp_put_event(c_node, NULL);
-
-	mutex_lock(&c_node->mem_lock);
+	/* stop operations */
+	if (ippdrv->stop)
+		ippdrv->stop(ippdrv->dev, property->cmd);
 
 	/* check command */
 	switch (property->cmd) {
 	case IPP_CMD_M2M:
-		for_each_ipp_ops(i) {
-			/* source/destination memory list */
-			head = &c_node->mem_list[i];
-
-			list_for_each_entry_safe(m_node, tm_node,
-				head, list) {
-				ret = ipp_put_mem_node(drm_dev, c_node,
-					m_node);
-				if (ret) {
-					DRM_ERROR("failed to put m_node.\n");
-					goto err_clear;
-				}
-			}
-		}
+		for_each_ipp_ops(i)
+			ipp_clean_mem_nodes(drm_dev, c_node, i);
 		break;
 	case IPP_CMD_WB:
-		/* destination memory list */
-		head = &c_node->mem_list[EXYNOS_DRM_OPS_DST];
-
-		list_for_each_entry_safe(m_node, tm_node, head, list) {
-			ret = ipp_put_mem_node(drm_dev, c_node, m_node);
-			if (ret) {
-				DRM_ERROR("failed to put m_node.\n");
-				goto err_clear;
-			}
-		}
+		ipp_clean_mem_nodes(drm_dev, c_node, EXYNOS_DRM_OPS_DST);
 		break;
 	case IPP_CMD_OUTPUT:
-		/* source memory list */
-		head = &c_node->mem_list[EXYNOS_DRM_OPS_SRC];
-
-		list_for_each_entry_safe(m_node, tm_node, head, list) {
-			ret = ipp_put_mem_node(drm_dev, c_node, m_node);
-			if (ret) {
-				DRM_ERROR("failed to put m_node.\n");
-				goto err_clear;
-			}
-		}
+		ipp_clean_mem_nodes(drm_dev, c_node, EXYNOS_DRM_OPS_SRC);
 		break;
 	default:
 		DRM_ERROR("invalid operations.\n");
-		ret = -EINVAL;
-		goto err_clear;
+		return -EINVAL;
 	}
 
-err_clear:
-	mutex_unlock(&c_node->mem_lock);
-
-	/* stop operations */
-	if (ippdrv->stop)
-		ippdrv->stop(ippdrv->dev, property->cmd);
-
-	return ret;
+	return 0;
 }
 
 void ipp_sched_cmd(struct work_struct *work)
 {
 	struct drm_exynos_ipp_cmd_work *cmd_work =
-		(struct drm_exynos_ipp_cmd_work *)work;
+		container_of(work, struct drm_exynos_ipp_cmd_work, work);
 	struct exynos_drm_ippdrv *ippdrv;
 	struct drm_exynos_ipp_cmd_node *c_node;
 	struct drm_exynos_ipp_property *property;
@@ -1543,7 +1502,7 @@ err_event_unlock:
 void ipp_sched_event(struct work_struct *work)
 {
 	struct drm_exynos_ipp_event_work *event_work =
-		(struct drm_exynos_ipp_event_work *)work;
+		container_of(work, struct drm_exynos_ipp_event_work, work);
 	struct exynos_drm_ippdrv *ippdrv;
 	struct drm_exynos_ipp_cmd_node *c_node;
 	int ret;
@@ -1646,11 +1605,11 @@ err:
 
 static void ipp_subdrv_remove(struct drm_device *drm_dev, struct device *dev)
 {
-	struct exynos_drm_ippdrv *ippdrv;
+	struct exynos_drm_ippdrv *ippdrv, *t;
 	struct ipp_context *ctx = get_ipp_context(dev);
 
 	/* get ipp driver entry */
-	list_for_each_entry(ippdrv, &exynos_drm_ippdrv_list, drv_list) {
+	list_for_each_entry_safe(ippdrv, t, &exynos_drm_ippdrv_list, drv_list) {
 		if (is_drm_iommu_supported(drm_dev))
 			drm_iommu_detach_device(drm_dev, ippdrv->dev);
 
@@ -1677,14 +1636,11 @@ static int ipp_subdrv_open(struct drm_device *drm_dev, struct device *dev,
 static void ipp_subdrv_close(struct drm_device *drm_dev, struct device *dev,
 		struct drm_file *file)
 {
-	struct drm_exynos_file_private *file_priv = file->driver_priv;
 	struct exynos_drm_ippdrv *ippdrv = NULL;
 	struct ipp_context *ctx = get_ipp_context(dev);
 	struct drm_exynos_ipp_cmd_node *c_node, *tc_node;
 	int count = 0;
 
-	DRM_DEBUG_KMS("for priv[0x%x]\n", (int)file_priv->ipp_dev);
-
 	list_for_each_entry(ippdrv, &exynos_drm_ippdrv_list, drv_list) {
 		mutex_lock(&ippdrv->cmd_lock);
 		list_for_each_entry_safe(c_node, tc_node,
@@ -1692,7 +1648,7 @@ static void ipp_subdrv_close(struct drm_device *drm_dev, struct device *dev,
 			DRM_DEBUG_KMS("count[%d]ippdrv[0x%x]\n",
 				count++, (int)ippdrv);
 
-			if (c_node->dev == file_priv->ipp_dev) {
+			if (c_node->filp == file) {
 				/*
 				 * userland goto unnormal state. process killed.
 				 * and close the file.
@@ -1808,63 +1764,12 @@ static int ipp_remove(struct platform_device *pdev)
 	return 0;
 }
 
-static int ipp_power_ctrl(struct ipp_context *ctx, bool enable)
-{
-	DRM_DEBUG_KMS("enable[%d]\n", enable);
-
-	return 0;
-}
-
-#ifdef CONFIG_PM_SLEEP
-static int ipp_suspend(struct device *dev)
-{
-	struct ipp_context *ctx = get_ipp_context(dev);
-
-	if (pm_runtime_suspended(dev))
-		return 0;
-
-	return ipp_power_ctrl(ctx, false);
-}
-
-static int ipp_resume(struct device *dev)
-{
-	struct ipp_context *ctx = get_ipp_context(dev);
-
-	if (!pm_runtime_suspended(dev))
-		return ipp_power_ctrl(ctx, true);
-
-	return 0;
-}
-#endif
-
-#ifdef CONFIG_PM_RUNTIME
-static int ipp_runtime_suspend(struct device *dev)
-{
-	struct ipp_context *ctx = get_ipp_context(dev);
-
-	return ipp_power_ctrl(ctx, false);
-}
-
-static int ipp_runtime_resume(struct device *dev)
-{
-	struct ipp_context *ctx = get_ipp_context(dev);
-
-	return ipp_power_ctrl(ctx, true);
-}
-#endif
-
-static const struct dev_pm_ops ipp_pm_ops = {
-	SET_SYSTEM_SLEEP_PM_OPS(ipp_suspend, ipp_resume)
-	SET_RUNTIME_PM_OPS(ipp_runtime_suspend, ipp_runtime_resume, NULL)
-};
-
 struct platform_driver ipp_driver = {
 	.probe		= ipp_probe,
 	.remove		= ipp_remove,
 	.driver		= {
 		.name	= "exynos-drm-ipp",
 		.owner	= THIS_MODULE,
-		.pm	= &ipp_pm_ops,
 	},
 };
 
diff --git a/drivers/gpu/drm/exynos/exynos_drm_ipp.h b/drivers/gpu/drm/exynos/exynos_drm_ipp.h
index 6f48d62aeb30..2a61547a39d0 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_ipp.h
+++ b/drivers/gpu/drm/exynos/exynos_drm_ipp.h
@@ -48,7 +48,6 @@ struct drm_exynos_ipp_cmd_work {
 /*
  * A structure of command node.
  *
- * @dev: IPP device.
  * @list: list head to command queue information.
  * @event_list: list head of event.
  * @mem_list: list head to source,destination memory queue information.
@@ -62,9 +61,9 @@ struct drm_exynos_ipp_cmd_work {
  * @stop_work: stop command work structure.
  * @event_work: event work structure.
  * @state: state of command node.
+ * @filp: associated file pointer.
  */
 struct drm_exynos_ipp_cmd_node {
-	struct device		*dev;
 	struct list_head	list;
 	struct list_head	event_list;
 	struct list_head	mem_list[EXYNOS_DRM_OPS_MAX];
@@ -78,6 +77,7 @@ struct drm_exynos_ipp_cmd_node {
 	struct drm_exynos_ipp_cmd_work *stop_work;
 	struct drm_exynos_ipp_event_work *event_work;
 	enum drm_exynos_ipp_state	state;
+	struct drm_file	*filp;
 };
 
 /*
diff --git a/drivers/gpu/drm/exynos/exynos_drm_plane.c b/drivers/gpu/drm/exynos/exynos_drm_plane.c
index 8371cbd7631d..c7045a663763 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_plane.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_plane.c
@@ -139,6 +139,8 @@ int exynos_plane_mode_set(struct drm_plane *plane, struct drm_crtc *crtc,
 			overlay->crtc_x, overlay->crtc_y,
 			overlay->crtc_width, overlay->crtc_height);
 
+	plane->crtc = crtc;
+
 	exynos_drm_crtc_plane_mode_set(crtc, overlay);
 
 	return 0;
@@ -187,8 +189,6 @@ exynos_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 	if (ret < 0)
 		return ret;
 
-	plane->crtc = crtc;
-
 	exynos_plane_commit(plane);
 	exynos_plane_dpms(plane, DRM_MODE_DPMS_ON);
 
@@ -254,25 +254,26 @@ static void exynos_plane_attach_zpos_property(struct drm_plane *plane)
 }
 
 struct drm_plane *exynos_plane_init(struct drm_device *dev,
-				    unsigned long possible_crtcs, bool priv)
+				    unsigned long possible_crtcs,
+				    enum drm_plane_type type)
 {
 	struct exynos_plane *exynos_plane;
 	int err;
 
 	exynos_plane = kzalloc(sizeof(struct exynos_plane), GFP_KERNEL);
 	if (!exynos_plane)
-		return NULL;
+		return ERR_PTR(-ENOMEM);
 
-	err = drm_plane_init(dev, &exynos_plane->base, possible_crtcs,
-			      &exynos_plane_funcs, formats, ARRAY_SIZE(formats),
-			      priv);
+	err = drm_universal_plane_init(dev, &exynos_plane->base, possible_crtcs,
+				       &exynos_plane_funcs, formats,
+				       ARRAY_SIZE(formats), type);
 	if (err) {
 		DRM_ERROR("failed to initialize plane\n");
 		kfree(exynos_plane);
-		return NULL;
+		return ERR_PTR(err);
 	}
 
-	if (priv)
+	if (type == DRM_PLANE_TYPE_PRIMARY)
 		exynos_plane->overlay.zpos = DEFAULT_ZPOS;
 	else
 		exynos_plane_attach_zpos_property(&exynos_plane->base);
diff --git a/drivers/gpu/drm/exynos/exynos_drm_plane.h b/drivers/gpu/drm/exynos/exynos_drm_plane.h
index 84d464c90d3d..0d1986b115f8 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_plane.h
+++ b/drivers/gpu/drm/exynos/exynos_drm_plane.h
@@ -17,4 +17,5 @@ int exynos_plane_mode_set(struct drm_plane *plane, struct drm_crtc *crtc,
 void exynos_plane_commit(struct drm_plane *plane);
 void exynos_plane_dpms(struct drm_plane *plane, int mode);
 struct drm_plane *exynos_plane_init(struct drm_device *dev,
-				    unsigned long possible_crtcs, bool priv);
+				    unsigned long possible_crtcs,
+				    enum drm_plane_type type);
diff --git a/drivers/gpu/drm/exynos/exynos_drm_rotator.c b/drivers/gpu/drm/exynos/exynos_drm_rotator.c
index 55af6b41c1df..b6a37d4f5b13 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_rotator.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_rotator.c
@@ -156,8 +156,7 @@ static irqreturn_t rotator_irq_handler(int irq, void *arg)
 		event_work->ippdrv = ippdrv;
 		event_work->buf_id[EXYNOS_DRM_OPS_DST] =
 			rot->cur_buf_id[EXYNOS_DRM_OPS_DST];
-		queue_work(ippdrv->event_workq,
-			(struct work_struct *)event_work);
+		queue_work(ippdrv->event_workq, &event_work->work);
 	} else {
 		DRM_ERROR("the SFR is set illegally\n");
 	}
diff --git a/drivers/gpu/drm/exynos/exynos_drm_vidi.c b/drivers/gpu/drm/exynos/exynos_drm_vidi.c
index 9528d81d8004..d565207040a2 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_vidi.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_vidi.c
@@ -303,23 +303,6 @@ static int vidi_mgr_initialize(struct exynos_drm_manager *mgr,
 	mgr->drm_dev = ctx->drm_dev = drm_dev;
 	mgr->pipe = ctx->pipe = priv->pipe++;
 
-	/*
-	 * enable drm irq mode.
-	 * - with irq_enabled = 1, we can use the vblank feature.
-	 *
-	 * P.S. note that we wouldn't use drm irq handler but
-	 *	just specific driver own one instead because
-	 *	drm framework supports only one irq handler.
-	 */
-	drm_dev->irq_enabled = 1;
-
-	/*
-	 * with vblank_disable_allowed = 1, vblank interrupt will be disabled
-	 * by drm timer once a current process gives up ownership of
-	 * vblank event.(after drm_vblank_put function is called)
-	 */
-	drm_dev->vblank_disable_allowed = 1;
-
 	return 0;
 }
 
@@ -648,7 +631,6 @@ static int vidi_remove(struct platform_device *pdev)
 	struct exynos_drm_manager *mgr = platform_get_drvdata(pdev);
 	struct vidi_context *ctx = mgr->ctx;
 	struct drm_encoder *encoder = ctx->encoder;
-	struct drm_crtc *crtc = mgr->crtc;
 
 	if (ctx->raw_edid != (struct edid *)fake_edid_info) {
 		kfree(ctx->raw_edid);
@@ -657,7 +639,6 @@ static int vidi_remove(struct platform_device *pdev)
 		return -EINVAL;
 	}
 
-	crtc->funcs->destroy(crtc);
 	encoder->funcs->destroy(encoder);
 	drm_connector_cleanup(&ctx->connector);
 
diff --git a/drivers/gpu/drm/exynos/exynos_hdmi.c b/drivers/gpu/drm/exynos/exynos_hdmi.c
index 562966db2aa1..7910fb37d9bb 100644
--- a/drivers/gpu/drm/exynos/exynos_hdmi.c
+++ b/drivers/gpu/drm/exynos/exynos_hdmi.c
@@ -1040,6 +1040,8 @@ static enum drm_connector_status hdmi_detect(struct drm_connector *connector,
 
 static void hdmi_connector_destroy(struct drm_connector *connector)
 {
+	drm_connector_unregister(connector);
+	drm_connector_cleanup(connector);
 }
 
 static struct drm_connector_funcs hdmi_connector_funcs = {
@@ -2314,8 +2316,8 @@ static void hdmi_unbind(struct device *dev, struct device *master, void *data)
 	struct drm_encoder *encoder = display->encoder;
 	struct hdmi_context *hdata = display->ctx;
 
+	hdmi_connector_destroy(&hdata->connector);
 	encoder->funcs->destroy(encoder);
-	drm_connector_cleanup(&hdata->connector);
 }
 
 static const struct component_ops hdmi_component_ops = {
diff --git a/drivers/gpu/drm/exynos/exynos_mixer.c b/drivers/gpu/drm/exynos/exynos_mixer.c
index e8b4ec84b312..a41c84ee3a2d 100644
--- a/drivers/gpu/drm/exynos/exynos_mixer.c
+++ b/drivers/gpu/drm/exynos/exynos_mixer.c
@@ -1302,15 +1302,12 @@ static int mixer_bind(struct device *dev, struct device *manager, void *data)
 static void mixer_unbind(struct device *dev, struct device *master, void *data)
 {
 	struct exynos_drm_manager *mgr = dev_get_drvdata(dev);
-	struct drm_crtc *crtc = mgr->crtc;
 
 	dev_info(dev, "remove successful\n");
 
 	mixer_mgr_remove(mgr);
 
 	pm_runtime_disable(dev);
-
-	crtc->funcs->destroy(crtc);
 }
 
 static const struct component_ops mixer_component_ops = {
diff --git a/drivers/gpu/drm/gma500/cdv_intel_dp.c b/drivers/gpu/drm/gma500/cdv_intel_dp.c
index a4cc0e60a1be..9f158eab517a 100644
--- a/drivers/gpu/drm/gma500/cdv_intel_dp.c
+++ b/drivers/gpu/drm/gma500/cdv_intel_dp.c
@@ -1089,7 +1089,7 @@ static char	*link_train_names[] = {
 };
 #endif
 
-#define CDV_DP_VOLTAGE_MAX	    DP_TRAIN_VOLTAGE_SWING_1200
+#define CDV_DP_VOLTAGE_MAX	    DP_TRAIN_VOLTAGE_SWING_LEVEL_3
 /*
 static uint8_t
 cdv_intel_dp_pre_emphasis_max(uint8_t voltage_swing)
@@ -1276,7 +1276,7 @@ cdv_intel_dp_set_vswing_premph(struct gma_encoder *encoder, uint8_t signal_level
 		cdv_sb_write(dev, ddi_reg->VSwing2, dp_vswing_premph_table[index]);
 
 	/* ;gfx_dpio_set_reg(0x814c, 0x40802040) */
-	if ((vswing + premph) == DP_TRAIN_VOLTAGE_SWING_1200)
+	if ((vswing + premph) == DP_TRAIN_VOLTAGE_SWING_LEVEL_3)
 		cdv_sb_write(dev, ddi_reg->VSwing3, 0x70802040);
 	else
 		cdv_sb_write(dev, ddi_reg->VSwing3, 0x40802040);
diff --git a/drivers/gpu/drm/gma500/framebuffer.c b/drivers/gpu/drm/gma500/framebuffer.c
index d0dd3bea8aa5..ddd90ddbc200 100644
--- a/drivers/gpu/drm/gma500/framebuffer.c
+++ b/drivers/gpu/drm/gma500/framebuffer.c
@@ -540,7 +540,8 @@ static void psbfb_gamma_get(struct drm_crtc *crtc, u16 *red,
 static int psbfb_probe(struct drm_fb_helper *helper,
 				struct drm_fb_helper_surface_size *sizes)
 {
-	struct psb_fbdev *psb_fbdev = (struct psb_fbdev *)helper;
+	struct psb_fbdev *psb_fbdev =
+		container_of(helper, struct psb_fbdev, psb_fb_helper);
 	struct drm_device *dev = psb_fbdev->psb_fb_helper.dev;
 	struct drm_psb_private *dev_priv = dev->dev_private;
 	int bytespp;
diff --git a/drivers/gpu/drm/gma500/gtt.h b/drivers/gpu/drm/gma500/gtt.h
index f5860a739bd8..cdbb350c9d5d 100644
--- a/drivers/gpu/drm/gma500/gtt.h
+++ b/drivers/gpu/drm/gma500/gtt.h
@@ -21,6 +21,7 @@
 #define _PSB_GTT_H_
 
 #include <drm/drmP.h>
+#include <drm/drm_gem.h>
 
 /* This wants cleaning up with respect to the psb_dev and un-needed stuff */
 struct psb_gtt {
diff --git a/drivers/gpu/drm/gma500/intel_bios.c b/drivers/gpu/drm/gma500/intel_bios.c
index d3497348c4d5..63bde4e86c6a 100644
--- a/drivers/gpu/drm/gma500/intel_bios.c
+++ b/drivers/gpu/drm/gma500/intel_bios.c
@@ -116,30 +116,30 @@ parse_edp(struct drm_psb_private *dev_priv, struct bdb_header *bdb)
 
 	switch (edp_link_params->preemphasis) {
 	case 0:
-		dev_priv->edp.preemphasis = DP_TRAIN_PRE_EMPHASIS_0;
+		dev_priv->edp.preemphasis = DP_TRAIN_PRE_EMPH_LEVEL_0;
 		break;
 	case 1:
-		dev_priv->edp.preemphasis = DP_TRAIN_PRE_EMPHASIS_3_5;
+		dev_priv->edp.preemphasis = DP_TRAIN_PRE_EMPH_LEVEL_1;
 		break;
 	case 2:
-		dev_priv->edp.preemphasis = DP_TRAIN_PRE_EMPHASIS_6;
+		dev_priv->edp.preemphasis = DP_TRAIN_PRE_EMPH_LEVEL_2;
 		break;
 	case 3:
-		dev_priv->edp.preemphasis = DP_TRAIN_PRE_EMPHASIS_9_5;
+		dev_priv->edp.preemphasis = DP_TRAIN_PRE_EMPH_LEVEL_3;
 		break;
 	}
 	switch (edp_link_params->vswing) {
 	case 0:
-		dev_priv->edp.vswing = DP_TRAIN_VOLTAGE_SWING_400;
+		dev_priv->edp.vswing = DP_TRAIN_VOLTAGE_SWING_LEVEL_0;
 		break;
 	case 1:
-		dev_priv->edp.vswing = DP_TRAIN_VOLTAGE_SWING_600;
+		dev_priv->edp.vswing = DP_TRAIN_VOLTAGE_SWING_LEVEL_1;
 		break;
 	case 2:
-		dev_priv->edp.vswing = DP_TRAIN_VOLTAGE_SWING_800;
+		dev_priv->edp.vswing = DP_TRAIN_VOLTAGE_SWING_LEVEL_2;
 		break;
 	case 3:
-		dev_priv->edp.vswing = DP_TRAIN_VOLTAGE_SWING_1200;
+		dev_priv->edp.vswing = DP_TRAIN_VOLTAGE_SWING_LEVEL_3;
 		break;
 	}
 	DRM_DEBUG_KMS("VBT reports EDP: VSwing  %d, Preemph %d\n",
diff --git a/drivers/gpu/drm/gma500/psb_drv.c b/drivers/gpu/drm/gma500/psb_drv.c
index eec993f93b1a..6ec3a905fdd2 100644
--- a/drivers/gpu/drm/gma500/psb_drv.c
+++ b/drivers/gpu/drm/gma500/psb_drv.c
@@ -476,6 +476,7 @@ static struct drm_driver driver = {
 	.unload = psb_driver_unload,
 	.lastclose = psb_driver_lastclose,
 	.preclose = psb_driver_preclose,
+	.set_busid = drm_pci_set_busid,
 
 	.num_ioctls = ARRAY_SIZE(psb_ioctls),
 	.device_is_agp = psb_driver_device_is_agp,
diff --git a/drivers/gpu/drm/i810/i810_dma.c b/drivers/gpu/drm/i810/i810_dma.c
index bae897de9468..d91856779beb 100644
--- a/drivers/gpu/drm/i810/i810_dma.c
+++ b/drivers/gpu/drm/i810/i810_dma.c
@@ -213,7 +213,7 @@ static int i810_dma_cleanup(struct drm_device *dev)
 		    (drm_i810_private_t *) dev->dev_private;
 
 		if (dev_priv->ring.virtual_start)
-			drm_core_ioremapfree(&dev_priv->ring.map, dev);
+			drm_legacy_ioremapfree(&dev_priv->ring.map, dev);
 		if (dev_priv->hw_status_page) {
 			pci_free_consistent(dev->pdev, PAGE_SIZE,
 					    dev_priv->hw_status_page,
@@ -227,7 +227,7 @@ static int i810_dma_cleanup(struct drm_device *dev)
 			drm_i810_buf_priv_t *buf_priv = buf->dev_private;
 
 			if (buf_priv->kernel_virtual && buf->total)
-				drm_core_ioremapfree(&buf_priv->map, dev);
+				drm_legacy_ioremapfree(&buf_priv->map, dev);
 		}
 	}
 	return 0;
@@ -306,7 +306,7 @@ static int i810_freelist_init(struct drm_device *dev, drm_i810_private_t *dev_pr
 		buf_priv->map.flags = 0;
 		buf_priv->map.mtrr = 0;
 
-		drm_core_ioremap(&buf_priv->map, dev);
+		drm_legacy_ioremap(&buf_priv->map, dev);
 		buf_priv->kernel_virtual = buf_priv->map.handle;
 
 	}
@@ -334,7 +334,7 @@ static int i810_dma_initialize(struct drm_device *dev,
 		DRM_ERROR("can not find sarea!\n");
 		return -EINVAL;
 	}
-	dev_priv->mmio_map = drm_core_findmap(dev, init->mmio_offset);
+	dev_priv->mmio_map = drm_legacy_findmap(dev, init->mmio_offset);
 	if (!dev_priv->mmio_map) {
 		dev->dev_private = (void *)dev_priv;
 		i810_dma_cleanup(dev);
@@ -342,7 +342,7 @@ static int i810_dma_initialize(struct drm_device *dev,
 		return -EINVAL;
 	}
 	dev->agp_buffer_token = init->buffers_offset;
-	dev->agp_buffer_map = drm_core_findmap(dev, init->buffers_offset);
+	dev->agp_buffer_map = drm_legacy_findmap(dev, init->buffers_offset);
 	if (!dev->agp_buffer_map) {
 		dev->dev_private = (void *)dev_priv;
 		i810_dma_cleanup(dev);
@@ -363,7 +363,7 @@ static int i810_dma_initialize(struct drm_device *dev,
 	dev_priv->ring.map.flags = 0;
 	dev_priv->ring.map.mtrr = 0;
 
-	drm_core_ioremap(&dev_priv->ring.map, dev);
+	drm_legacy_ioremap(&dev_priv->ring.map, dev);
 
 	if (dev_priv->ring.map.handle == NULL) {
 		dev->dev_private = (void *)dev_priv;
@@ -1215,9 +1215,9 @@ void i810_driver_preclose(struct drm_device *dev, struct drm_file *file_priv)
 	}
 
 	if (file_priv->master && file_priv->master->lock.hw_lock) {
-		drm_idlelock_take(&file_priv->master->lock);
+		drm_legacy_idlelock_take(&file_priv->master->lock);
 		i810_driver_reclaim_buffers(dev, file_priv);
-		drm_idlelock_release(&file_priv->master->lock);
+		drm_legacy_idlelock_release(&file_priv->master->lock);
 	} else {
 		/* master disappeared, clean up stuff anyway and hope nothing
 		 * goes wrong */
diff --git a/drivers/gpu/drm/i810/i810_drv.c b/drivers/gpu/drm/i810/i810_drv.c
index 441ccf8f5bdc..44f4a131c8dd 100644
--- a/drivers/gpu/drm/i810/i810_drv.c
+++ b/drivers/gpu/drm/i810/i810_drv.c
@@ -47,7 +47,7 @@ static const struct file_operations i810_driver_fops = {
 	.open = drm_open,
 	.release = drm_release,
 	.unlocked_ioctl = drm_ioctl,
-	.mmap = drm_mmap,
+	.mmap = drm_legacy_mmap,
 	.poll = drm_poll,
 #ifdef CONFIG_COMPAT
 	.compat_ioctl = drm_compat_ioctl,
@@ -63,6 +63,7 @@ static struct drm_driver driver = {
 	.load = i810_driver_load,
 	.lastclose = i810_driver_lastclose,
 	.preclose = i810_driver_preclose,
+	.set_busid = drm_pci_set_busid,
 	.device_is_agp = i810_driver_device_is_agp,
 	.dma_quiescent = i810_driver_dma_quiescent,
 	.ioctls = i810_ioctls,
diff --git a/drivers/gpu/drm/i810/i810_drv.h b/drivers/gpu/drm/i810/i810_drv.h
index d4d16eddd651..93ec5dc4e7d3 100644
--- a/drivers/gpu/drm/i810/i810_drv.h
+++ b/drivers/gpu/drm/i810/i810_drv.h
@@ -32,6 +32,8 @@
 #ifndef _I810_DRV_H_
 #define _I810_DRV_H_
 
+#include <drm/drm_legacy.h>
+
 /* General customization:
  */
 
diff --git a/drivers/gpu/drm/i915/Makefile b/drivers/gpu/drm/i915/Makefile
index 91bd167e1cb7..c1dd485aeb6c 100644
--- a/drivers/gpu/drm/i915/Makefile
+++ b/drivers/gpu/drm/i915/Makefile
@@ -31,6 +31,7 @@ i915-y += i915_cmd_parser.o \
 	  i915_gpu_error.o \
 	  i915_irq.o \
 	  i915_trace_points.o \
+	  intel_lrc.o \
 	  intel_ringbuffer.o \
 	  intel_uncore.o
 
diff --git a/drivers/gpu/drm/i915/dvo_ns2501.c b/drivers/gpu/drm/i915/dvo_ns2501.c
index 74f2af7c2d3e..441630434d34 100644
--- a/drivers/gpu/drm/i915/dvo_ns2501.c
+++ b/drivers/gpu/drm/i915/dvo_ns2501.c
@@ -60,16 +60,297 @@
 
 #define NS2501_REGC 0x0c
 
+enum {
+	MODE_640x480,
+	MODE_800x600,
+	MODE_1024x768,
+};
+
+struct ns2501_reg {
+	 uint8_t offset;
+	 uint8_t value;
+};
+
+/*
+ * Magic values based on what the BIOS on
+ * Fujitsu-Siemens Lifebook S6010 programs (1024x768 panel).
+ */
+static const struct ns2501_reg regs_1024x768[][86] = {
+	[MODE_640x480] = {
+		[0] = { .offset = 0x0a, .value = 0x81, },
+		[1] = { .offset = 0x18, .value = 0x07, },
+		[2] = { .offset = 0x19, .value = 0x00, },
+		[3] = { .offset = 0x1a, .value = 0x00, },
+		[4] = { .offset = 0x1b, .value = 0x11, },
+		[5] = { .offset = 0x1c, .value = 0x54, },
+		[6] = { .offset = 0x1d, .value = 0x03, },
+		[7] = { .offset = 0x1e, .value = 0x02, },
+		[8] = { .offset = 0xf3, .value = 0x90, },
+		[9] = { .offset = 0xf9, .value = 0x00, },
+		[10] = { .offset = 0xc1, .value = 0x90, },
+		[11] = { .offset = 0xc2, .value = 0x00, },
+		[12] = { .offset = 0xc3, .value = 0x0f, },
+		[13] = { .offset = 0xc4, .value = 0x03, },
+		[14] = { .offset = 0xc5, .value = 0x16, },
+		[15] = { .offset = 0xc6, .value = 0x00, },
+		[16] = { .offset = 0xc7, .value = 0x02, },
+		[17] = { .offset = 0xc8, .value = 0x02, },
+		[18] = { .offset = 0xf4, .value = 0x00, },
+		[19] = { .offset = 0x80, .value = 0xff, },
+		[20] = { .offset = 0x81, .value = 0x07, },
+		[21] = { .offset = 0x82, .value = 0x3d, },
+		[22] = { .offset = 0x83, .value = 0x05, },
+		[23] = { .offset = 0x94, .value = 0x00, },
+		[24] = { .offset = 0x95, .value = 0x00, },
+		[25] = { .offset = 0x96, .value = 0x05, },
+		[26] = { .offset = 0x97, .value = 0x00, },
+		[27] = { .offset = 0x9a, .value = 0x88, },
+		[28] = { .offset = 0x9b, .value = 0x00, },
+		[29] = { .offset = 0x98, .value = 0x00, },
+		[30] = { .offset = 0x99, .value = 0x00, },
+		[31] = { .offset = 0xf7, .value = 0x88, },
+		[32] = { .offset = 0xf8, .value = 0x0a, },
+		[33] = { .offset = 0x9c, .value = 0x24, },
+		[34] = { .offset = 0x9d, .value = 0x00, },
+		[35] = { .offset = 0x9e, .value = 0x25, },
+		[36] = { .offset = 0x9f, .value = 0x03, },
+		[37] = { .offset = 0xa0, .value = 0x28, },
+		[38] = { .offset = 0xa1, .value = 0x01, },
+		[39] = { .offset = 0xa2, .value = 0x28, },
+		[40] = { .offset = 0xa3, .value = 0x05, },
+		[41] = { .offset = 0xb6, .value = 0x09, },
+		[42] = { .offset = 0xb8, .value = 0x00, },
+		[43] = { .offset = 0xb9, .value = 0xa0, },
+		[44] = { .offset = 0xba, .value = 0x00, },
+		[45] = { .offset = 0xbb, .value = 0x20, },
+		[46] = { .offset = 0x10, .value = 0x00, },
+		[47] = { .offset = 0x11, .value = 0xa0, },
+		[48] = { .offset = 0x12, .value = 0x02, },
+		[49] = { .offset = 0x20, .value = 0x00, },
+		[50] = { .offset = 0x22, .value = 0x00, },
+		[51] = { .offset = 0x23, .value = 0x00, },
+		[52] = { .offset = 0x24, .value = 0x00, },
+		[53] = { .offset = 0x25, .value = 0x00, },
+		[54] = { .offset = 0x8c, .value = 0x10, },
+		[55] = { .offset = 0x8d, .value = 0x02, },
+		[56] = { .offset = 0x8e, .value = 0x10, },
+		[57] = { .offset = 0x8f, .value = 0x00, },
+		[58] = { .offset = 0x90, .value = 0xff, },
+		[59] = { .offset = 0x91, .value = 0x07, },
+		[60] = { .offset = 0x92, .value = 0xa0, },
+		[61] = { .offset = 0x93, .value = 0x02, },
+		[62] = { .offset = 0xa5, .value = 0x00, },
+		[63] = { .offset = 0xa6, .value = 0x00, },
+		[64] = { .offset = 0xa7, .value = 0x00, },
+		[65] = { .offset = 0xa8, .value = 0x00, },
+		[66] = { .offset = 0xa9, .value = 0x04, },
+		[67] = { .offset = 0xaa, .value = 0x70, },
+		[68] = { .offset = 0xab, .value = 0x4f, },
+		[69] = { .offset = 0xac, .value = 0x00, },
+		[70] = { .offset = 0xa4, .value = 0x84, },
+		[71] = { .offset = 0x7e, .value = 0x18, },
+		[72] = { .offset = 0x84, .value = 0x00, },
+		[73] = { .offset = 0x85, .value = 0x00, },
+		[74] = { .offset = 0x86, .value = 0x00, },
+		[75] = { .offset = 0x87, .value = 0x00, },
+		[76] = { .offset = 0x88, .value = 0x00, },
+		[77] = { .offset = 0x89, .value = 0x00, },
+		[78] = { .offset = 0x8a, .value = 0x00, },
+		[79] = { .offset = 0x8b, .value = 0x00, },
+		[80] = { .offset = 0x26, .value = 0x00, },
+		[81] = { .offset = 0x27, .value = 0x00, },
+		[82] = { .offset = 0xad, .value = 0x00, },
+		[83] = { .offset = 0x08, .value = 0x30, }, /* 0x31 */
+		[84] = { .offset = 0x41, .value = 0x00, },
+		[85] = { .offset = 0xc0, .value = 0x05, },
+	},
+	[MODE_800x600] = {
+		[0] = { .offset = 0x0a, .value = 0x81, },
+		[1] = { .offset = 0x18, .value = 0x07, },
+		[2] = { .offset = 0x19, .value = 0x00, },
+		[3] = { .offset = 0x1a, .value = 0x00, },
+		[4] = { .offset = 0x1b, .value = 0x19, },
+		[5] = { .offset = 0x1c, .value = 0x64, },
+		[6] = { .offset = 0x1d, .value = 0x02, },
+		[7] = { .offset = 0x1e, .value = 0x02, },
+		[8] = { .offset = 0xf3, .value = 0x90, },
+		[9] = { .offset = 0xf9, .value = 0x00, },
+		[10] = { .offset = 0xc1, .value = 0xd7, },
+		[11] = { .offset = 0xc2, .value = 0x00, },
+		[12] = { .offset = 0xc3, .value = 0xf8, },
+		[13] = { .offset = 0xc4, .value = 0x03, },
+		[14] = { .offset = 0xc5, .value = 0x1a, },
+		[15] = { .offset = 0xc6, .value = 0x00, },
+		[16] = { .offset = 0xc7, .value = 0x73, },
+		[17] = { .offset = 0xc8, .value = 0x02, },
+		[18] = { .offset = 0xf4, .value = 0x00, },
+		[19] = { .offset = 0x80, .value = 0x27, },
+		[20] = { .offset = 0x81, .value = 0x03, },
+		[21] = { .offset = 0x82, .value = 0x41, },
+		[22] = { .offset = 0x83, .value = 0x05, },
+		[23] = { .offset = 0x94, .value = 0x00, },
+		[24] = { .offset = 0x95, .value = 0x00, },
+		[25] = { .offset = 0x96, .value = 0x05, },
+		[26] = { .offset = 0x97, .value = 0x00, },
+		[27] = { .offset = 0x9a, .value = 0x88, },
+		[28] = { .offset = 0x9b, .value = 0x00, },
+		[29] = { .offset = 0x98, .value = 0x00, },
+		[30] = { .offset = 0x99, .value = 0x00, },
+		[31] = { .offset = 0xf7, .value = 0x88, },
+		[32] = { .offset = 0xf8, .value = 0x06, },
+		[33] = { .offset = 0x9c, .value = 0x23, },
+		[34] = { .offset = 0x9d, .value = 0x00, },
+		[35] = { .offset = 0x9e, .value = 0x25, },
+		[36] = { .offset = 0x9f, .value = 0x03, },
+		[37] = { .offset = 0xa0, .value = 0x28, },
+		[38] = { .offset = 0xa1, .value = 0x01, },
+		[39] = { .offset = 0xa2, .value = 0x28, },
+		[40] = { .offset = 0xa3, .value = 0x05, },
+		[41] = { .offset = 0xb6, .value = 0x09, },
+		[42] = { .offset = 0xb8, .value = 0x30, },
+		[43] = { .offset = 0xb9, .value = 0xc8, },
+		[44] = { .offset = 0xba, .value = 0x00, },
+		[45] = { .offset = 0xbb, .value = 0x20, },
+		[46] = { .offset = 0x10, .value = 0x20, },
+		[47] = { .offset = 0x11, .value = 0xc8, },
+		[48] = { .offset = 0x12, .value = 0x02, },
+		[49] = { .offset = 0x20, .value = 0x00, },
+		[50] = { .offset = 0x22, .value = 0x00, },
+		[51] = { .offset = 0x23, .value = 0x00, },
+		[52] = { .offset = 0x24, .value = 0x00, },
+		[53] = { .offset = 0x25, .value = 0x00, },
+		[54] = { .offset = 0x8c, .value = 0x10, },
+		[55] = { .offset = 0x8d, .value = 0x02, },
+		[56] = { .offset = 0x8e, .value = 0x04, },
+		[57] = { .offset = 0x8f, .value = 0x00, },
+		[58] = { .offset = 0x90, .value = 0xff, },
+		[59] = { .offset = 0x91, .value = 0x07, },
+		[60] = { .offset = 0x92, .value = 0xa0, },
+		[61] = { .offset = 0x93, .value = 0x02, },
+		[62] = { .offset = 0xa5, .value = 0x00, },
+		[63] = { .offset = 0xa6, .value = 0x00, },
+		[64] = { .offset = 0xa7, .value = 0x00, },
+		[65] = { .offset = 0xa8, .value = 0x00, },
+		[66] = { .offset = 0xa9, .value = 0x83, },
+		[67] = { .offset = 0xaa, .value = 0x40, },
+		[68] = { .offset = 0xab, .value = 0x32, },
+		[69] = { .offset = 0xac, .value = 0x00, },
+		[70] = { .offset = 0xa4, .value = 0x80, },
+		[71] = { .offset = 0x7e, .value = 0x18, },
+		[72] = { .offset = 0x84, .value = 0x00, },
+		[73] = { .offset = 0x85, .value = 0x00, },
+		[74] = { .offset = 0x86, .value = 0x00, },
+		[75] = { .offset = 0x87, .value = 0x00, },
+		[76] = { .offset = 0x88, .value = 0x00, },
+		[77] = { .offset = 0x89, .value = 0x00, },
+		[78] = { .offset = 0x8a, .value = 0x00, },
+		[79] = { .offset = 0x8b, .value = 0x00, },
+		[80] = { .offset = 0x26, .value = 0x00, },
+		[81] = { .offset = 0x27, .value = 0x00, },
+		[82] = { .offset = 0xad, .value = 0x00, },
+		[83] = { .offset = 0x08, .value = 0x30, }, /* 0x31 */
+		[84] = { .offset = 0x41, .value = 0x00, },
+		[85] = { .offset = 0xc0, .value = 0x07, },
+	},
+	[MODE_1024x768] = {
+		[0] = { .offset = 0x0a, .value = 0x81, },
+		[1] = { .offset = 0x18, .value = 0x07, },
+		[2] = { .offset = 0x19, .value = 0x00, },
+		[3] = { .offset = 0x1a, .value = 0x00, },
+		[4] = { .offset = 0x1b, .value = 0x11, },
+		[5] = { .offset = 0x1c, .value = 0x54, },
+		[6] = { .offset = 0x1d, .value = 0x03, },
+		[7] = { .offset = 0x1e, .value = 0x02, },
+		[8] = { .offset = 0xf3, .value = 0x90, },
+		[9] = { .offset = 0xf9, .value = 0x00, },
+		[10] = { .offset = 0xc1, .value = 0x90, },
+		[11] = { .offset = 0xc2, .value = 0x00, },
+		[12] = { .offset = 0xc3, .value = 0x0f, },
+		[13] = { .offset = 0xc4, .value = 0x03, },
+		[14] = { .offset = 0xc5, .value = 0x16, },
+		[15] = { .offset = 0xc6, .value = 0x00, },
+		[16] = { .offset = 0xc7, .value = 0x02, },
+		[17] = { .offset = 0xc8, .value = 0x02, },
+		[18] = { .offset = 0xf4, .value = 0x00, },
+		[19] = { .offset = 0x80, .value = 0xff, },
+		[20] = { .offset = 0x81, .value = 0x07, },
+		[21] = { .offset = 0x82, .value = 0x3d, },
+		[22] = { .offset = 0x83, .value = 0x05, },
+		[23] = { .offset = 0x94, .value = 0x00, },
+		[24] = { .offset = 0x95, .value = 0x00, },
+		[25] = { .offset = 0x96, .value = 0x05, },
+		[26] = { .offset = 0x97, .value = 0x00, },
+		[27] = { .offset = 0x9a, .value = 0x88, },
+		[28] = { .offset = 0x9b, .value = 0x00, },
+		[29] = { .offset = 0x98, .value = 0x00, },
+		[30] = { .offset = 0x99, .value = 0x00, },
+		[31] = { .offset = 0xf7, .value = 0x88, },
+		[32] = { .offset = 0xf8, .value = 0x0a, },
+		[33] = { .offset = 0x9c, .value = 0x24, },
+		[34] = { .offset = 0x9d, .value = 0x00, },
+		[35] = { .offset = 0x9e, .value = 0x25, },
+		[36] = { .offset = 0x9f, .value = 0x03, },
+		[37] = { .offset = 0xa0, .value = 0x28, },
+		[38] = { .offset = 0xa1, .value = 0x01, },
+		[39] = { .offset = 0xa2, .value = 0x28, },
+		[40] = { .offset = 0xa3, .value = 0x05, },
+		[41] = { .offset = 0xb6, .value = 0x09, },
+		[42] = { .offset = 0xb8, .value = 0x00, },
+		[43] = { .offset = 0xb9, .value = 0xa0, },
+		[44] = { .offset = 0xba, .value = 0x00, },
+		[45] = { .offset = 0xbb, .value = 0x20, },
+		[46] = { .offset = 0x10, .value = 0x00, },
+		[47] = { .offset = 0x11, .value = 0xa0, },
+		[48] = { .offset = 0x12, .value = 0x02, },
+		[49] = { .offset = 0x20, .value = 0x00, },
+		[50] = { .offset = 0x22, .value = 0x00, },
+		[51] = { .offset = 0x23, .value = 0x00, },
+		[52] = { .offset = 0x24, .value = 0x00, },
+		[53] = { .offset = 0x25, .value = 0x00, },
+		[54] = { .offset = 0x8c, .value = 0x10, },
+		[55] = { .offset = 0x8d, .value = 0x02, },
+		[56] = { .offset = 0x8e, .value = 0x10, },
+		[57] = { .offset = 0x8f, .value = 0x00, },
+		[58] = { .offset = 0x90, .value = 0xff, },
+		[59] = { .offset = 0x91, .value = 0x07, },
+		[60] = { .offset = 0x92, .value = 0xa0, },
+		[61] = { .offset = 0x93, .value = 0x02, },
+		[62] = { .offset = 0xa5, .value = 0x00, },
+		[63] = { .offset = 0xa6, .value = 0x00, },
+		[64] = { .offset = 0xa7, .value = 0x00, },
+		[65] = { .offset = 0xa8, .value = 0x00, },
+		[66] = { .offset = 0xa9, .value = 0x04, },
+		[67] = { .offset = 0xaa, .value = 0x70, },
+		[68] = { .offset = 0xab, .value = 0x4f, },
+		[69] = { .offset = 0xac, .value = 0x00, },
+		[70] = { .offset = 0xa4, .value = 0x84, },
+		[71] = { .offset = 0x7e, .value = 0x18, },
+		[72] = { .offset = 0x84, .value = 0x00, },
+		[73] = { .offset = 0x85, .value = 0x00, },
+		[74] = { .offset = 0x86, .value = 0x00, },
+		[75] = { .offset = 0x87, .value = 0x00, },
+		[76] = { .offset = 0x88, .value = 0x00, },
+		[77] = { .offset = 0x89, .value = 0x00, },
+		[78] = { .offset = 0x8a, .value = 0x00, },
+		[79] = { .offset = 0x8b, .value = 0x00, },
+		[80] = { .offset = 0x26, .value = 0x00, },
+		[81] = { .offset = 0x27, .value = 0x00, },
+		[82] = { .offset = 0xad, .value = 0x00, },
+		[83] = { .offset = 0x08, .value = 0x34, }, /* 0x35 */
+		[84] = { .offset = 0x41, .value = 0x00, },
+		[85] = { .offset = 0xc0, .value = 0x01, },
+	},
+};
+
+static const struct ns2501_reg regs_init[] = {
+	[0] = { .offset = 0x35, .value = 0xff, },
+	[1] = { .offset = 0x34, .value = 0x00, },
+	[2] = { .offset = 0x08, .value = 0x30, },
+};
+
 struct ns2501_priv {
-	//I2CDevRec d;
 	bool quiet;
-	int reg_8_shadow;
-	int reg_8_set;
-	// Shadow registers for i915
-	int dvoc;
-	int pll_a;
-	int srcdim;
-	int fw_blc;
+	const struct ns2501_reg *regs;
 };
 
 #define NSPTR(d) ((NS2501Ptr)(d->DriverPrivate.ptr))
@@ -205,11 +486,9 @@ static bool ns2501_init(struct intel_dvo_device *dvo,
 		goto out;
 	}
 	ns->quiet = false;
-	ns->reg_8_set = 0;
-	ns->reg_8_shadow =
-	    NS2501_8_PD | NS2501_8_BPAS | NS2501_8_VEN | NS2501_8_HEN;
 
 	DRM_DEBUG_KMS("init ns2501 dvo controller successfully!\n");
+
 	return true;
 
 out:
@@ -242,9 +521,9 @@ static enum drm_mode_status ns2501_mode_valid(struct intel_dvo_device *dvo,
 	 * of the panel in here so we could always accept it
 	 * by disabling the scaler.
 	 */
-	if ((mode->hdisplay == 800 && mode->vdisplay == 600) ||
-	    (mode->hdisplay == 640 && mode->vdisplay == 480) ||
-	    (mode->hdisplay == 1024 && mode->vdisplay == 768)) {
+	if ((mode->hdisplay == 640 && mode->vdisplay == 480 && mode->clock == 25175) ||
+	    (mode->hdisplay == 800 && mode->vdisplay == 600 && mode->clock == 40000) ||
+	    (mode->hdisplay == 1024 && mode->vdisplay == 768 && mode->clock == 65000)) {
 		return MODE_OK;
 	} else {
 		return MODE_ONE_SIZE;	/* Is this a reasonable error? */
@@ -255,180 +534,30 @@ static void ns2501_mode_set(struct intel_dvo_device *dvo,
 			    struct drm_display_mode *mode,
 			    struct drm_display_mode *adjusted_mode)
 {
-	bool ok;
-	int retries = 10;
 	struct ns2501_priv *ns = (struct ns2501_priv *)(dvo->dev_priv);
+	int mode_idx, i;
 
 	DRM_DEBUG_KMS
 	    ("set mode (hdisplay=%d,htotal=%d,vdisplay=%d,vtotal=%d).\n",
 	     mode->hdisplay, mode->htotal, mode->vdisplay, mode->vtotal);
 
-	/*
-	 * Where do I find the native resolution for which scaling is not required???
-	 *
-	 * First trigger the DVO on as otherwise the chip does not appear on the i2c
-	 * bus.
-	 */
-	do {
-		ok = true;
-
-		if (mode->hdisplay == 800 && mode->vdisplay == 600) {
-			/* mode 277 */
-			ns->reg_8_shadow &= ~NS2501_8_BPAS;
-			DRM_DEBUG_KMS("switching to 800x600\n");
-
-			/*
-			 * No, I do not know where this data comes from.
-			 * It is just what the video bios left in the DVO, so
-			 * I'm just copying it here over.
-			 * This also means that I cannot support any other modes
-			 * except the ones supported by the bios.
-			 */
-			ok &= ns2501_writeb(dvo, 0x11, 0xc8);	// 0xc7 also works.
-			ok &= ns2501_writeb(dvo, 0x1b, 0x19);
-			ok &= ns2501_writeb(dvo, 0x1c, 0x62);	// VBIOS left 0x64 here, but 0x62 works nicer
-			ok &= ns2501_writeb(dvo, 0x1d, 0x02);
-
-			ok &= ns2501_writeb(dvo, 0x34, 0x03);
-			ok &= ns2501_writeb(dvo, 0x35, 0xff);
+	if (mode->hdisplay == 640 && mode->vdisplay == 480)
+		mode_idx = MODE_640x480;
+	else if (mode->hdisplay == 800 && mode->vdisplay == 600)
+		mode_idx = MODE_800x600;
+	else if (mode->hdisplay == 1024 && mode->vdisplay == 768)
+		mode_idx = MODE_1024x768;
+	else
+		return;
 
-			ok &= ns2501_writeb(dvo, 0x80, 0x27);
-			ok &= ns2501_writeb(dvo, 0x81, 0x03);
-			ok &= ns2501_writeb(dvo, 0x82, 0x41);
-			ok &= ns2501_writeb(dvo, 0x83, 0x05);
+	/* Hopefully doing it every time won't hurt... */
+	for (i = 0; i < ARRAY_SIZE(regs_init); i++)
+		ns2501_writeb(dvo, regs_init[i].offset, regs_init[i].value);
 
-			ok &= ns2501_writeb(dvo, 0x8d, 0x02);
-			ok &= ns2501_writeb(dvo, 0x8e, 0x04);
-			ok &= ns2501_writeb(dvo, 0x8f, 0x00);
+	ns->regs = regs_1024x768[mode_idx];
 
-			ok &= ns2501_writeb(dvo, 0x90, 0xfe);	/* vertical. VBIOS left 0xff here, but 0xfe works better */
-			ok &= ns2501_writeb(dvo, 0x91, 0x07);
-			ok &= ns2501_writeb(dvo, 0x94, 0x00);
-			ok &= ns2501_writeb(dvo, 0x95, 0x00);
-
-			ok &= ns2501_writeb(dvo, 0x96, 0x00);
-
-			ok &= ns2501_writeb(dvo, 0x99, 0x00);
-			ok &= ns2501_writeb(dvo, 0x9a, 0x88);
-
-			ok &= ns2501_writeb(dvo, 0x9c, 0x23);	/* Looks like first and last line of the image. */
-			ok &= ns2501_writeb(dvo, 0x9d, 0x00);
-			ok &= ns2501_writeb(dvo, 0x9e, 0x25);
-			ok &= ns2501_writeb(dvo, 0x9f, 0x03);
-
-			ok &= ns2501_writeb(dvo, 0xa4, 0x80);
-
-			ok &= ns2501_writeb(dvo, 0xb6, 0x00);
-
-			ok &= ns2501_writeb(dvo, 0xb9, 0xc8);	/* horizontal? */
-			ok &= ns2501_writeb(dvo, 0xba, 0x00);	/* horizontal? */
-
-			ok &= ns2501_writeb(dvo, 0xc0, 0x05);	/* horizontal? */
-			ok &= ns2501_writeb(dvo, 0xc1, 0xd7);
-
-			ok &= ns2501_writeb(dvo, 0xc2, 0x00);
-			ok &= ns2501_writeb(dvo, 0xc3, 0xf8);
-
-			ok &= ns2501_writeb(dvo, 0xc4, 0x03);
-			ok &= ns2501_writeb(dvo, 0xc5, 0x1a);
-
-			ok &= ns2501_writeb(dvo, 0xc6, 0x00);
-			ok &= ns2501_writeb(dvo, 0xc7, 0x73);
-			ok &= ns2501_writeb(dvo, 0xc8, 0x02);
-
-		} else if (mode->hdisplay == 640 && mode->vdisplay == 480) {
-			/* mode 274 */
-			DRM_DEBUG_KMS("switching to 640x480\n");
-			/*
-			 * No, I do not know where this data comes from.
-			 * It is just what the video bios left in the DVO, so
-			 * I'm just copying it here over.
-			 * This also means that I cannot support any other modes
-			 * except the ones supported by the bios.
-			 */
-			ns->reg_8_shadow &= ~NS2501_8_BPAS;
-
-			ok &= ns2501_writeb(dvo, 0x11, 0xa0);
-			ok &= ns2501_writeb(dvo, 0x1b, 0x11);
-			ok &= ns2501_writeb(dvo, 0x1c, 0x54);
-			ok &= ns2501_writeb(dvo, 0x1d, 0x03);
-
-			ok &= ns2501_writeb(dvo, 0x34, 0x03);
-			ok &= ns2501_writeb(dvo, 0x35, 0xff);
-
-			ok &= ns2501_writeb(dvo, 0x80, 0xff);
-			ok &= ns2501_writeb(dvo, 0x81, 0x07);
-			ok &= ns2501_writeb(dvo, 0x82, 0x3d);
-			ok &= ns2501_writeb(dvo, 0x83, 0x05);
-
-			ok &= ns2501_writeb(dvo, 0x8d, 0x02);
-			ok &= ns2501_writeb(dvo, 0x8e, 0x10);
-			ok &= ns2501_writeb(dvo, 0x8f, 0x00);
-
-			ok &= ns2501_writeb(dvo, 0x90, 0xff);	/* vertical */
-			ok &= ns2501_writeb(dvo, 0x91, 0x07);
-			ok &= ns2501_writeb(dvo, 0x94, 0x00);
-			ok &= ns2501_writeb(dvo, 0x95, 0x00);
-
-			ok &= ns2501_writeb(dvo, 0x96, 0x05);
-
-			ok &= ns2501_writeb(dvo, 0x99, 0x00);
-			ok &= ns2501_writeb(dvo, 0x9a, 0x88);
-
-			ok &= ns2501_writeb(dvo, 0x9c, 0x24);
-			ok &= ns2501_writeb(dvo, 0x9d, 0x00);
-			ok &= ns2501_writeb(dvo, 0x9e, 0x25);
-			ok &= ns2501_writeb(dvo, 0x9f, 0x03);
-
-			ok &= ns2501_writeb(dvo, 0xa4, 0x84);
-
-			ok &= ns2501_writeb(dvo, 0xb6, 0x09);
-
-			ok &= ns2501_writeb(dvo, 0xb9, 0xa0);	/* horizontal? */
-			ok &= ns2501_writeb(dvo, 0xba, 0x00);	/* horizontal? */
-
-			ok &= ns2501_writeb(dvo, 0xc0, 0x05);	/* horizontal? */
-			ok &= ns2501_writeb(dvo, 0xc1, 0x90);
-
-			ok &= ns2501_writeb(dvo, 0xc2, 0x00);
-			ok &= ns2501_writeb(dvo, 0xc3, 0x0f);
-
-			ok &= ns2501_writeb(dvo, 0xc4, 0x03);
-			ok &= ns2501_writeb(dvo, 0xc5, 0x16);
-
-			ok &= ns2501_writeb(dvo, 0xc6, 0x00);
-			ok &= ns2501_writeb(dvo, 0xc7, 0x02);
-			ok &= ns2501_writeb(dvo, 0xc8, 0x02);
-
-		} else if (mode->hdisplay == 1024 && mode->vdisplay == 768) {
-			/* mode 280 */
-			DRM_DEBUG_KMS("switching to 1024x768\n");
-			/*
-			 * This might or might not work, actually. I'm silently
-			 * assuming here that the native panel resolution is
-			 * 1024x768. If not, then this leaves the scaler disabled
-			 * generating a picture that is likely not the expected.
-			 *
-			 * Problem is that I do not know where to take the panel
-			 * dimensions from.
-			 *
-			 * Enable the bypass, scaling not required.
-			 *
-			 * The scaler registers are irrelevant here....
-			 *
-			 */
-			ns->reg_8_shadow |= NS2501_8_BPAS;
-			ok &= ns2501_writeb(dvo, 0x37, 0x44);
-		} else {
-			/*
-			 * Data not known. Bummer!
-			 * Hopefully, the code should not go here
-			 * as mode_OK delivered no other modes.
-			 */
-			ns->reg_8_shadow |= NS2501_8_BPAS;
-		}
-		ok &= ns2501_writeb(dvo, NS2501_REG8, ns->reg_8_shadow);
-	} while (!ok && retries--);
+	for (i = 0; i < 84; i++)
+		ns2501_writeb(dvo, ns->regs[i].offset, ns->regs[i].value);
 }
 
 /* set the NS2501 power state */
@@ -439,60 +568,46 @@ static bool ns2501_get_hw_state(struct intel_dvo_device *dvo)
 	if (!ns2501_readb(dvo, NS2501_REG8, &ch))
 		return false;
 
-	if (ch & NS2501_8_PD)
-		return true;
-	else
-		return false;
+	return ch & NS2501_8_PD;
 }
 
 /* set the NS2501 power state */
 static void ns2501_dpms(struct intel_dvo_device *dvo, bool enable)
 {
-	bool ok;
-	int retries = 10;
 	struct ns2501_priv *ns = (struct ns2501_priv *)(dvo->dev_priv);
-	unsigned char ch;
 
 	DRM_DEBUG_KMS("Trying set the dpms of the DVO to %i\n", enable);
 
-	ch = ns->reg_8_shadow;
+	if (enable) {
+		if (WARN_ON(ns->regs[83].offset != 0x08 ||
+			    ns->regs[84].offset != 0x41 ||
+			    ns->regs[85].offset != 0xc0))
+			return;
 
-	if (enable)
-		ch |= NS2501_8_PD;
-	else
-		ch &= ~NS2501_8_PD;
-
-	if (ns->reg_8_set == 0 || ns->reg_8_shadow != ch) {
-		ns->reg_8_set = 1;
-		ns->reg_8_shadow = ch;
-
-		do {
-			ok = true;
-			ok &= ns2501_writeb(dvo, NS2501_REG8, ch);
-			ok &=
-			    ns2501_writeb(dvo, 0x34,
-					  enable ? 0x03 : 0x00);
-			ok &=
-			    ns2501_writeb(dvo, 0x35,
-					  enable ? 0xff : 0x00);
-		} while (!ok && retries--);
-	}
-}
+		ns2501_writeb(dvo, 0xc0, ns->regs[85].value | 0x08);
 
-static void ns2501_dump_regs(struct intel_dvo_device *dvo)
-{
-	uint8_t val;
-
-	ns2501_readb(dvo, NS2501_FREQ_LO, &val);
-	DRM_DEBUG_KMS("NS2501_FREQ_LO: 0x%02x\n", val);
-	ns2501_readb(dvo, NS2501_FREQ_HI, &val);
-	DRM_DEBUG_KMS("NS2501_FREQ_HI: 0x%02x\n", val);
-	ns2501_readb(dvo, NS2501_REG8, &val);
-	DRM_DEBUG_KMS("NS2501_REG8: 0x%02x\n", val);
-	ns2501_readb(dvo, NS2501_REG9, &val);
-	DRM_DEBUG_KMS("NS2501_REG9: 0x%02x\n", val);
-	ns2501_readb(dvo, NS2501_REGC, &val);
-	DRM_DEBUG_KMS("NS2501_REGC: 0x%02x\n", val);
+		ns2501_writeb(dvo, 0x41, ns->regs[84].value);
+
+		ns2501_writeb(dvo, 0x34, 0x01);
+		msleep(15);
+
+		ns2501_writeb(dvo, 0x08, 0x35);
+		if (!(ns->regs[83].value & NS2501_8_BPAS))
+			ns2501_writeb(dvo, 0x08, 0x31);
+		msleep(200);
+
+		ns2501_writeb(dvo, 0x34, 0x03);
+
+		ns2501_writeb(dvo, 0xc0, ns->regs[85].value);
+	} else {
+		ns2501_writeb(dvo, 0x34, 0x01);
+		msleep(200);
+
+		ns2501_writeb(dvo, 0x08, 0x34);
+		msleep(15);
+
+		ns2501_writeb(dvo, 0x34, 0x00);
+	}
 }
 
 static void ns2501_destroy(struct intel_dvo_device *dvo)
@@ -512,6 +627,5 @@ struct intel_dvo_dev_ops ns2501_ops = {
 	.mode_set = ns2501_mode_set,
 	.dpms = ns2501_dpms,
 	.get_hw_state = ns2501_get_hw_state,
-	.dump_regs = ns2501_dump_regs,
 	.destroy = ns2501_destroy,
 };
diff --git a/drivers/gpu/drm/i915/i915_cmd_parser.c b/drivers/gpu/drm/i915/i915_cmd_parser.c
index 4b7ed5289217..593b657d3e59 100644
--- a/drivers/gpu/drm/i915/i915_cmd_parser.c
+++ b/drivers/gpu/drm/i915/i915_cmd_parser.c
@@ -844,8 +844,6 @@ finish:
  */
 bool i915_needs_cmd_parser(struct intel_engine_cs *ring)
 {
-	struct drm_i915_private *dev_priv = ring->dev->dev_private;
-
 	if (!ring->needs_cmd_parser)
 		return false;
 
@@ -854,7 +852,7 @@ bool i915_needs_cmd_parser(struct intel_engine_cs *ring)
 	 * disabled. That will cause all of the parser's PPGTT checks to
 	 * fail. For now, disable parsing when PPGTT is off.
 	 */
-	if (!dev_priv->mm.aliasing_ppgtt)
+	if (USES_PPGTT(ring->dev))
 		return false;
 
 	return (i915.enable_cmd_parser == 1);
diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c
index 9e737b771c40..063b44817e08 100644
--- a/drivers/gpu/drm/i915/i915_debugfs.c
+++ b/drivers/gpu/drm/i915/i915_debugfs.c
@@ -136,7 +136,7 @@ describe_obj(struct seq_file *m, struct drm_i915_gem_object *obj)
 		   obj->last_read_seqno,
 		   obj->last_write_seqno,
 		   obj->last_fenced_seqno,
-		   i915_cache_level_str(obj->cache_level),
+		   i915_cache_level_str(to_i915(obj->base.dev), obj->cache_level),
 		   obj->dirty ? " dirty" : "",
 		   obj->madv == I915_MADV_DONTNEED ? " purgeable" : "");
 	if (obj->base.name)
@@ -333,7 +333,7 @@ static int per_file_stats(int id, void *ptr, void *data)
 			}
 
 			ppgtt = container_of(vma->vm, struct i915_hw_ppgtt, base);
-			if (ppgtt->ctx && ppgtt->ctx->file_priv != stats->file_priv)
+			if (ppgtt->file_priv != stats->file_priv)
 				continue;
 
 			if (obj->ring) /* XXX per-vma statistic */
@@ -515,6 +515,7 @@ static int i915_gem_pageflip_info(struct seq_file *m, void *data)
 {
 	struct drm_info_node *node = m->private;
 	struct drm_device *dev = node->minor->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	unsigned long flags;
 	struct intel_crtc *crtc;
 	int ret;
@@ -534,6 +535,8 @@ static int i915_gem_pageflip_info(struct seq_file *m, void *data)
 			seq_printf(m, "No flip due on pipe %c (plane %c)\n",
 				   pipe, plane);
 		} else {
+			u32 addr;
+
 			if (atomic_read(&work->pending) < INTEL_FLIP_COMPLETE) {
 				seq_printf(m, "Flip queued on pipe %c (plane %c)\n",
 					   pipe, plane);
@@ -541,23 +544,35 @@ static int i915_gem_pageflip_info(struct seq_file *m, void *data)
 				seq_printf(m, "Flip pending (waiting for vsync) on pipe %c (plane %c)\n",
 					   pipe, plane);
 			}
+			if (work->flip_queued_ring) {
+				seq_printf(m, "Flip queued on %s at seqno %u, next seqno %u [current breadcrumb %u], completed? %d\n",
+					   work->flip_queued_ring->name,
+					   work->flip_queued_seqno,
+					   dev_priv->next_seqno,
+					   work->flip_queued_ring->get_seqno(work->flip_queued_ring, true),
+					   i915_seqno_passed(work->flip_queued_ring->get_seqno(work->flip_queued_ring, true),
+							     work->flip_queued_seqno));
+			} else
+				seq_printf(m, "Flip not associated with any ring\n");
+			seq_printf(m, "Flip queued on frame %d, (was ready on frame %d), now %d\n",
+				   work->flip_queued_vblank,
+				   work->flip_ready_vblank,
+				   drm_vblank_count(dev, crtc->pipe));
 			if (work->enable_stall_check)
 				seq_puts(m, "Stall check enabled, ");
 			else
 				seq_puts(m, "Stall check waiting for page flip ioctl, ");
 			seq_printf(m, "%d prepares\n", atomic_read(&work->pending));
 
-			if (work->old_fb_obj) {
-				struct drm_i915_gem_object *obj = work->old_fb_obj;
-				if (obj)
-					seq_printf(m, "Old framebuffer gtt_offset 0x%08lx\n",
-						   i915_gem_obj_ggtt_offset(obj));
-			}
+			if (INTEL_INFO(dev)->gen >= 4)
+				addr = I915_HI_DISPBASE(I915_READ(DSPSURF(crtc->plane)));
+			else
+				addr = I915_READ(DSPADDR(crtc->plane));
+			seq_printf(m, "Current scanout address 0x%08x\n", addr);
+
 			if (work->pending_flip_obj) {
-				struct drm_i915_gem_object *obj = work->pending_flip_obj;
-				if (obj)
-					seq_printf(m, "New framebuffer gtt_offset 0x%08lx\n",
-						   i915_gem_obj_ggtt_offset(obj));
+				seq_printf(m, "New framebuffer address 0x%08lx\n", (long)work->gtt_offset);
+				seq_printf(m, "MMIO update completed? %d\n",  addr == work->gtt_offset);
 			}
 		}
 		spin_unlock_irqrestore(&dev->event_lock, flags);
@@ -650,7 +665,6 @@ static int i915_interrupt_info(struct seq_file *m, void *data)
 	intel_runtime_pm_get(dev_priv);
 
 	if (IS_CHERRYVIEW(dev)) {
-		int i;
 		seq_printf(m, "Master Interrupt Control:\t%08x\n",
 			   I915_READ(GEN8_MASTER_IRQ));
 
@@ -662,7 +676,7 @@ static int i915_interrupt_info(struct seq_file *m, void *data)
 			   I915_READ(VLV_IIR_RW));
 		seq_printf(m, "Display IMR:\t%08x\n",
 			   I915_READ(VLV_IMR));
-		for_each_pipe(pipe)
+		for_each_pipe(dev_priv, pipe)
 			seq_printf(m, "Pipe %c stat:\t%08x\n",
 				   pipe_name(pipe),
 				   I915_READ(PIPESTAT(pipe)));
@@ -702,7 +716,13 @@ static int i915_interrupt_info(struct seq_file *m, void *data)
 				   i, I915_READ(GEN8_GT_IER(i)));
 		}
 
-		for_each_pipe(pipe) {
+		for_each_pipe(dev_priv, pipe) {
+			if (!intel_display_power_enabled(dev_priv,
+						POWER_DOMAIN_PIPE(pipe))) {
+				seq_printf(m, "Pipe %c power disabled\n",
+					   pipe_name(pipe));
+				continue;
+			}
 			seq_printf(m, "Pipe %c IMR:\t%08x\n",
 				   pipe_name(pipe),
 				   I915_READ(GEN8_DE_PIPE_IMR(pipe)));
@@ -743,7 +763,7 @@ static int i915_interrupt_info(struct seq_file *m, void *data)
 			   I915_READ(VLV_IIR_RW));
 		seq_printf(m, "Display IMR:\t%08x\n",
 			   I915_READ(VLV_IMR));
-		for_each_pipe(pipe)
+		for_each_pipe(dev_priv, pipe)
 			seq_printf(m, "Pipe %c stat:\t%08x\n",
 				   pipe_name(pipe),
 				   I915_READ(PIPESTAT(pipe)));
@@ -779,7 +799,7 @@ static int i915_interrupt_info(struct seq_file *m, void *data)
 			   I915_READ(IIR));
 		seq_printf(m, "Interrupt mask:      %08x\n",
 			   I915_READ(IMR));
-		for_each_pipe(pipe)
+		for_each_pipe(dev_priv, pipe)
 			seq_printf(m, "Pipe %c stat:         %08x\n",
 				   pipe_name(pipe),
 				   I915_READ(PIPESTAT(pipe)));
@@ -927,7 +947,7 @@ static ssize_t i915_error_state_read(struct file *file, char __user *userbuf,
 	ssize_t ret_count = 0;
 	int ret;
 
-	ret = i915_error_state_buf_init(&error_str, count, *pos);
+	ret = i915_error_state_buf_init(&error_str, to_i915(error_priv->dev), count, *pos);
 	if (ret)
 		return ret;
 
@@ -1024,6 +1044,7 @@ static int i915_frequency_info(struct seq_file *m, void *unused)
 		u32 rpstat, cagf, reqf;
 		u32 rpupei, rpcurup, rpprevup;
 		u32 rpdownei, rpcurdown, rpprevdown;
+		u32 pm_ier, pm_imr, pm_isr, pm_iir, pm_mask;
 		int max_freq;
 
 		/* RPSTAT1 is in the GT power well */
@@ -1061,12 +1082,21 @@ static int i915_frequency_info(struct seq_file *m, void *unused)
 		gen6_gt_force_wake_put(dev_priv, FORCEWAKE_ALL);
 		mutex_unlock(&dev->struct_mutex);
 
+		if (IS_GEN6(dev) || IS_GEN7(dev)) {
+			pm_ier = I915_READ(GEN6_PMIER);
+			pm_imr = I915_READ(GEN6_PMIMR);
+			pm_isr = I915_READ(GEN6_PMISR);
+			pm_iir = I915_READ(GEN6_PMIIR);
+			pm_mask = I915_READ(GEN6_PMINTRMSK);
+		} else {
+			pm_ier = I915_READ(GEN8_GT_IER(2));
+			pm_imr = I915_READ(GEN8_GT_IMR(2));
+			pm_isr = I915_READ(GEN8_GT_ISR(2));
+			pm_iir = I915_READ(GEN8_GT_IIR(2));
+			pm_mask = I915_READ(GEN6_PMINTRMSK);
+		}
 		seq_printf(m, "PM IER=0x%08x IMR=0x%08x ISR=0x%08x IIR=0x%08x, MASK=0x%08x\n",
-			   I915_READ(GEN6_PMIER),
-			   I915_READ(GEN6_PMIMR),
-			   I915_READ(GEN6_PMISR),
-			   I915_READ(GEN6_PMIIR),
-			   I915_READ(GEN6_PMINTRMSK));
+			   pm_ier, pm_imr, pm_isr, pm_iir, pm_mask);
 		seq_printf(m, "GT_PERF_STATUS: 0x%08x\n", gt_perf_status);
 		seq_printf(m, "Render p-state ratio: %d\n",
 			   (gt_perf_status & 0xff00) >> 8);
@@ -1365,7 +1395,7 @@ static int i915_drpc_info(struct seq_file *m, void *unused)
 
 	if (IS_VALLEYVIEW(dev))
 		return vlv_drpc_info(m);
-	else if (IS_GEN6(dev) || IS_GEN7(dev))
+	else if (INTEL_INFO(dev)->gen >= 6)
 		return gen6_drpc_info(m);
 	else
 		return ironlake_drpc_info(m);
@@ -1433,6 +1463,47 @@ static int i915_fbc_status(struct seq_file *m, void *unused)
 	return 0;
 }
 
+static int i915_fbc_fc_get(void *data, u64 *val)
+{
+	struct drm_device *dev = data;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	if (INTEL_INFO(dev)->gen < 7 || !HAS_FBC(dev))
+		return -ENODEV;
+
+	drm_modeset_lock_all(dev);
+	*val = dev_priv->fbc.false_color;
+	drm_modeset_unlock_all(dev);
+
+	return 0;
+}
+
+static int i915_fbc_fc_set(void *data, u64 val)
+{
+	struct drm_device *dev = data;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 reg;
+
+	if (INTEL_INFO(dev)->gen < 7 || !HAS_FBC(dev))
+		return -ENODEV;
+
+	drm_modeset_lock_all(dev);
+
+	reg = I915_READ(ILK_DPFC_CONTROL);
+	dev_priv->fbc.false_color = val;
+
+	I915_WRITE(ILK_DPFC_CONTROL, val ?
+		   (reg | FBC_CTL_FALSE_COLOR) :
+		   (reg & ~FBC_CTL_FALSE_COLOR));
+
+	drm_modeset_unlock_all(dev);
+	return 0;
+}
+
+DEFINE_SIMPLE_ATTRIBUTE(i915_fbc_fc_fops,
+			i915_fbc_fc_get, i915_fbc_fc_set,
+			"%llu\n");
+
 static int i915_ips_status(struct seq_file *m, void *unused)
 {
 	struct drm_info_node *node = m->private;
@@ -1630,6 +1701,14 @@ static int i915_gem_framebuffer_info(struct seq_file *m, void *data)
 	return 0;
 }
 
+static void describe_ctx_ringbuf(struct seq_file *m,
+				 struct intel_ringbuffer *ringbuf)
+{
+	seq_printf(m, " (ringbuffer, space: %d, head: %u, tail: %u, last head: %d)",
+		   ringbuf->space, ringbuf->head, ringbuf->tail,
+		   ringbuf->last_retired_head);
+}
+
 static int i915_context_status(struct seq_file *m, void *unused)
 {
 	struct drm_info_node *node = m->private;
@@ -1656,16 +1735,168 @@ static int i915_context_status(struct seq_file *m, void *unused)
 	}
 
 	list_for_each_entry(ctx, &dev_priv->context_list, link) {
-		if (ctx->legacy_hw_ctx.rcs_state == NULL)
+		if (!i915.enable_execlists &&
+		    ctx->legacy_hw_ctx.rcs_state == NULL)
 			continue;
 
 		seq_puts(m, "HW context ");
 		describe_ctx(m, ctx);
-		for_each_ring(ring, dev_priv, i)
+		for_each_ring(ring, dev_priv, i) {
+			if (ring->default_context == ctx)
+				seq_printf(m, "(default context %s) ",
+					   ring->name);
+		}
+
+		if (i915.enable_execlists) {
+			seq_putc(m, '\n');
+			for_each_ring(ring, dev_priv, i) {
+				struct drm_i915_gem_object *ctx_obj =
+					ctx->engine[i].state;
+				struct intel_ringbuffer *ringbuf =
+					ctx->engine[i].ringbuf;
+
+				seq_printf(m, "%s: ", ring->name);
+				if (ctx_obj)
+					describe_obj(m, ctx_obj);
+				if (ringbuf)
+					describe_ctx_ringbuf(m, ringbuf);
+				seq_putc(m, '\n');
+			}
+		} else {
+			describe_obj(m, ctx->legacy_hw_ctx.rcs_state);
+		}
+
+		seq_putc(m, '\n');
+	}
+
+	mutex_unlock(&dev->struct_mutex);
+
+	return 0;
+}
+
+static int i915_dump_lrc(struct seq_file *m, void *unused)
+{
+	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_device *dev = node->minor->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring;
+	struct intel_context *ctx;
+	int ret, i;
+
+	if (!i915.enable_execlists) {
+		seq_printf(m, "Logical Ring Contexts are disabled\n");
+		return 0;
+	}
+
+	ret = mutex_lock_interruptible(&dev->struct_mutex);
+	if (ret)
+		return ret;
+
+	list_for_each_entry(ctx, &dev_priv->context_list, link) {
+		for_each_ring(ring, dev_priv, i) {
+			struct drm_i915_gem_object *ctx_obj = ctx->engine[i].state;
+
 			if (ring->default_context == ctx)
-				seq_printf(m, "(default context %s) ", ring->name);
+				continue;
+
+			if (ctx_obj) {
+				struct page *page = i915_gem_object_get_page(ctx_obj, 1);
+				uint32_t *reg_state = kmap_atomic(page);
+				int j;
+
+				seq_printf(m, "CONTEXT: %s %u\n", ring->name,
+						intel_execlists_ctx_id(ctx_obj));
+
+				for (j = 0; j < 0x600 / sizeof(u32) / 4; j += 4) {
+					seq_printf(m, "\t[0x%08lx] 0x%08x 0x%08x 0x%08x 0x%08x\n",
+					i915_gem_obj_ggtt_offset(ctx_obj) + 4096 + (j * 4),
+					reg_state[j], reg_state[j + 1],
+					reg_state[j + 2], reg_state[j + 3]);
+				}
+				kunmap_atomic(reg_state);
+
+				seq_putc(m, '\n');
+			}
+		}
+	}
+
+	mutex_unlock(&dev->struct_mutex);
+
+	return 0;
+}
+
+static int i915_execlists(struct seq_file *m, void *data)
+{
+	struct drm_info_node *node = (struct drm_info_node *)m->private;
+	struct drm_device *dev = node->minor->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring;
+	u32 status_pointer;
+	u8 read_pointer;
+	u8 write_pointer;
+	u32 status;
+	u32 ctx_id;
+	struct list_head *cursor;
+	int ring_id, i;
+	int ret;
+
+	if (!i915.enable_execlists) {
+		seq_puts(m, "Logical Ring Contexts are disabled\n");
+		return 0;
+	}
+
+	ret = mutex_lock_interruptible(&dev->struct_mutex);
+	if (ret)
+		return ret;
+
+	for_each_ring(ring, dev_priv, ring_id) {
+		struct intel_ctx_submit_request *head_req = NULL;
+		int count = 0;
+		unsigned long flags;
+
+		seq_printf(m, "%s\n", ring->name);
+
+		status = I915_READ(RING_EXECLIST_STATUS(ring));
+		ctx_id = I915_READ(RING_EXECLIST_STATUS(ring) + 4);
+		seq_printf(m, "\tExeclist status: 0x%08X, context: %u\n",
+			   status, ctx_id);
+
+		status_pointer = I915_READ(RING_CONTEXT_STATUS_PTR(ring));
+		seq_printf(m, "\tStatus pointer: 0x%08X\n", status_pointer);
+
+		read_pointer = ring->next_context_status_buffer;
+		write_pointer = status_pointer & 0x07;
+		if (read_pointer > write_pointer)
+			write_pointer += 6;
+		seq_printf(m, "\tRead pointer: 0x%08X, write pointer 0x%08X\n",
+			   read_pointer, write_pointer);
+
+		for (i = 0; i < 6; i++) {
+			status = I915_READ(RING_CONTEXT_STATUS_BUF(ring) + 8*i);
+			ctx_id = I915_READ(RING_CONTEXT_STATUS_BUF(ring) + 8*i + 4);
+
+			seq_printf(m, "\tStatus buffer %d: 0x%08X, context: %u\n",
+				   i, status, ctx_id);
+		}
+
+		spin_lock_irqsave(&ring->execlist_lock, flags);
+		list_for_each(cursor, &ring->execlist_queue)
+			count++;
+		head_req = list_first_entry_or_null(&ring->execlist_queue,
+				struct intel_ctx_submit_request, execlist_link);
+		spin_unlock_irqrestore(&ring->execlist_lock, flags);
+
+		seq_printf(m, "\t%d requests in queue\n", count);
+		if (head_req) {
+			struct drm_i915_gem_object *ctx_obj;
+
+			ctx_obj = head_req->ctx->engine[ring_id].state;
+			seq_printf(m, "\tHead request id: %u\n",
+				   intel_execlists_ctx_id(ctx_obj));
+			seq_printf(m, "\tHead request tail: %u\n",
+				   head_req->tail);
+		}
 
-		describe_obj(m, ctx->legacy_hw_ctx.rcs_state);
 		seq_putc(m, '\n');
 	}
 
@@ -1774,7 +2005,13 @@ static int per_file_ctx(int id, void *ptr, void *data)
 {
 	struct intel_context *ctx = ptr;
 	struct seq_file *m = data;
-	struct i915_hw_ppgtt *ppgtt = ctx_to_ppgtt(ctx);
+	struct i915_hw_ppgtt *ppgtt = ctx->ppgtt;
+
+	if (!ppgtt) {
+		seq_printf(m, "  no ppgtt for context %d\n",
+			   ctx->user_handle);
+		return 0;
+	}
 
 	if (i915_gem_context_is_default(ctx))
 		seq_puts(m, "  default context:\n");
@@ -1834,8 +2071,7 @@ static void gen6_ppgtt_info(struct seq_file *m, struct drm_device *dev)
 		seq_printf(m, "pd gtt offset: 0x%08x\n", ppgtt->pd_offset);
 
 		ppgtt->debug_dump(ppgtt, m);
-	} else
-		return;
+	}
 
 	list_for_each_entry_reverse(file, &dev->filelist, lhead) {
 		struct drm_i915_file_private *file_priv = file->driver_priv;
@@ -2406,6 +2642,40 @@ static int i915_shared_dplls_info(struct seq_file *m, void *unused)
 	return 0;
 }
 
+static int i915_wa_registers(struct seq_file *m, void *unused)
+{
+	int i;
+	int ret;
+	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_device *dev = node->minor->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	ret = mutex_lock_interruptible(&dev->struct_mutex);
+	if (ret)
+		return ret;
+
+	intel_runtime_pm_get(dev_priv);
+
+	seq_printf(m, "Workarounds applied: %d\n", dev_priv->num_wa_regs);
+	for (i = 0; i < dev_priv->num_wa_regs; ++i) {
+		u32 addr, mask;
+
+		addr = dev_priv->intel_wa_regs[i].addr;
+		mask = dev_priv->intel_wa_regs[i].mask;
+		dev_priv->intel_wa_regs[i].value = I915_READ(addr) | mask;
+		if (dev_priv->intel_wa_regs[i].addr)
+			seq_printf(m, "0x%X: 0x%08X, mask: 0x%08X\n",
+				   dev_priv->intel_wa_regs[i].addr,
+				   dev_priv->intel_wa_regs[i].value,
+				   dev_priv->intel_wa_regs[i].mask);
+	}
+
+	intel_runtime_pm_put(dev_priv);
+	mutex_unlock(&dev->struct_mutex);
+
+	return 0;
+}
+
 struct pipe_crc_info {
 	const char *name;
 	struct drm_device *dev;
@@ -2667,8 +2937,7 @@ static int i9xx_pipe_crc_auto_source(struct drm_device *dev, enum pipe pipe,
 	*source = INTEL_PIPE_CRC_SOURCE_PIPE;
 
 	drm_modeset_lock_all(dev);
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list,
-			    base.head) {
+	for_each_intel_encoder(dev, encoder) {
 		if (!encoder->base.crtc)
 			continue;
 
@@ -3557,9 +3826,6 @@ i915_drop_caches_set(void *data, u64 val)
 {
 	struct drm_device *dev = data;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_i915_gem_object *obj, *next;
-	struct i915_address_space *vm;
-	struct i915_vma *vma, *x;
 	int ret;
 
 	DRM_DEBUG("Dropping caches: 0x%08llx\n", val);
@@ -3579,29 +3845,11 @@ i915_drop_caches_set(void *data, u64 val)
 	if (val & (DROP_RETIRE | DROP_ACTIVE))
 		i915_gem_retire_requests(dev);
 
-	if (val & DROP_BOUND) {
-		list_for_each_entry(vm, &dev_priv->vm_list, global_link) {
-			list_for_each_entry_safe(vma, x, &vm->inactive_list,
-						 mm_list) {
-				if (vma->pin_count)
-					continue;
+	if (val & DROP_BOUND)
+		i915_gem_shrink(dev_priv, LONG_MAX, I915_SHRINK_BOUND);
 
-				ret = i915_vma_unbind(vma);
-				if (ret)
-					goto unlock;
-			}
-		}
-	}
-
-	if (val & DROP_UNBOUND) {
-		list_for_each_entry_safe(obj, next, &dev_priv->mm.unbound_list,
-					 global_list)
-			if (obj->pages_pin_count == 0) {
-				ret = i915_gem_object_put_pages(obj);
-				if (ret)
-					goto unlock;
-			}
-	}
+	if (val & DROP_UNBOUND)
+		i915_gem_shrink(dev_priv, LONG_MAX, I915_SHRINK_UNBOUND);
 
 unlock:
 	mutex_unlock(&dev->struct_mutex);
@@ -3923,6 +4171,8 @@ static const struct drm_info_list i915_debugfs_list[] = {
 	{"i915_opregion", i915_opregion, 0},
 	{"i915_gem_framebuffer", i915_gem_framebuffer_info, 0},
 	{"i915_context_status", i915_context_status, 0},
+	{"i915_dump_lrc", i915_dump_lrc, 0},
+	{"i915_execlists", i915_execlists, 0},
 	{"i915_gen6_forcewake_count", i915_gen6_forcewake_count_info, 0},
 	{"i915_swizzle_info", i915_swizzle_info, 0},
 	{"i915_ppgtt_info", i915_ppgtt_info, 0},
@@ -3936,6 +4186,7 @@ static const struct drm_info_list i915_debugfs_list[] = {
 	{"i915_semaphore_status", i915_semaphore_status, 0},
 	{"i915_shared_dplls_info", i915_shared_dplls_info, 0},
 	{"i915_dp_mst_info", i915_dp_mst_info, 0},
+	{"i915_wa_registers", i915_wa_registers, 0},
 };
 #define I915_DEBUGFS_ENTRIES ARRAY_SIZE(i915_debugfs_list)
 
@@ -3957,6 +4208,7 @@ static const struct i915_debugfs_files {
 	{"i915_pri_wm_latency", &i915_pri_wm_latency_fops},
 	{"i915_spr_wm_latency", &i915_spr_wm_latency_fops},
 	{"i915_cur_wm_latency", &i915_cur_wm_latency_fops},
+	{"i915_fbc_false_color", &i915_fbc_fc_fops},
 };
 
 void intel_display_crc_init(struct drm_device *dev)
@@ -3964,7 +4216,7 @@ void intel_display_crc_init(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	enum pipe pipe;
 
-	for_each_pipe(pipe) {
+	for_each_pipe(dev_priv, pipe) {
 		struct intel_pipe_crc *pipe_crc = &dev_priv->pipe_crc[pipe];
 
 		pipe_crc->opened = false;
diff --git a/drivers/gpu/drm/i915/i915_dma.c b/drivers/gpu/drm/i915/i915_dma.c
index 9933c26017ed..1403b01e8216 100644
--- a/drivers/gpu/drm/i915/i915_dma.c
+++ b/drivers/gpu/drm/i915/i915_dma.c
@@ -28,9 +28,11 @@
 
 #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
 
+#include <linux/async.h>
 #include <drm/drmP.h>
 #include <drm/drm_crtc_helper.h>
 #include <drm/drm_fb_helper.h>
+#include <drm/drm_legacy.h>
 #include "intel_drv.h"
 #include <drm/i915_drm.h>
 #include "i915_drv.h"
@@ -196,7 +198,7 @@ static int i915_initialize(struct drm_device *dev, drm_i915_init_t *init)
 	struct drm_i915_master_private *master_priv = dev->primary->master->driver_priv;
 	int ret;
 
-	master_priv->sarea = drm_getsarea(dev);
+	master_priv->sarea = drm_legacy_getsarea(dev);
 	if (master_priv->sarea) {
 		master_priv->sarea_priv = (drm_i915_sarea_t *)
 			((u8 *)master_priv->sarea->handle + init->sarea_priv_offset);
@@ -999,7 +1001,7 @@ static int i915_getparam(struct drm_device *dev, void *data,
 		value = HAS_WT(dev);
 		break;
 	case I915_PARAM_HAS_ALIASING_PPGTT:
-		value = dev_priv->mm.aliasing_ppgtt || USES_FULL_PPGTT(dev);
+		value = USES_PPGTT(dev);
 		break;
 	case I915_PARAM_HAS_WAIT_TIMEOUT:
 		value = 1;
@@ -1355,8 +1357,6 @@ static int i915_load_modeset_init(struct drm_device *dev)
 	if (ret)
 		goto cleanup_irq;
 
-	INIT_WORK(&dev_priv->console_resume_work, intel_console_resume);
-
 	intel_modeset_gem_init(dev);
 
 	/* Always safe in the mode setting case. */
@@ -1382,7 +1382,7 @@ static int i915_load_modeset_init(struct drm_device *dev)
 	 * scanning against hotplug events. Hence do this first and ignore the
 	 * tiny window where we will loose hotplug notifactions.
 	 */
-	intel_fbdev_initial_config(dev);
+	async_schedule(intel_fbdev_initial_config, dev_priv);
 
 	drm_kms_helper_poll_init(dev);
 
@@ -1393,7 +1393,6 @@ cleanup_gem:
 	i915_gem_cleanup_ringbuffer(dev);
 	i915_gem_context_fini(dev);
 	mutex_unlock(&dev->struct_mutex);
-	WARN_ON(dev_priv->mm.aliasing_ppgtt);
 cleanup_irq:
 	drm_irq_uninstall(dev);
 cleanup_gem_stolen:
@@ -1536,10 +1535,10 @@ static void intel_device_info_runtime_init(struct drm_device *dev)
 	info = (struct intel_device_info *)&dev_priv->info;
 
 	if (IS_VALLEYVIEW(dev))
-		for_each_pipe(pipe)
+		for_each_pipe(dev_priv, pipe)
 			info->num_sprites[pipe] = 2;
 	else
-		for_each_pipe(pipe)
+		for_each_pipe(dev_priv, pipe)
 			info->num_sprites[pipe] = 1;
 
 	if (i915.disable_display) {
@@ -1608,9 +1607,10 @@ int i915_driver_load(struct drm_device *dev, unsigned long flags)
 	dev->dev_private = dev_priv;
 	dev_priv->dev = dev;
 
-	/* copy initial configuration to dev_priv->info */
+	/* Setup the write-once "constant" device info */
 	device_info = (struct intel_device_info *)&dev_priv->info;
-	*device_info = *info;
+	memcpy(device_info, info, sizeof(dev_priv->info));
+	device_info->device_id = dev->pdev->device;
 
 	spin_lock_init(&dev_priv->irq_lock);
 	spin_lock_init(&dev_priv->gpu_error.lock);
@@ -1822,7 +1822,7 @@ out_mtrrfree:
 	arch_phys_wc_del(dev_priv->gtt.mtrr);
 	io_mapping_free(dev_priv->gtt.mappable);
 out_gtt:
-	dev_priv->gtt.base.cleanup(&dev_priv->gtt.base);
+	i915_global_gtt_cleanup(dev);
 out_regs:
 	intel_uncore_fini(dev);
 	pci_iounmap(dev->pdev, dev_priv->regs);
@@ -1869,7 +1869,6 @@ int i915_driver_unload(struct drm_device *dev)
 	if (drm_core_check_feature(dev, DRIVER_MODESET)) {
 		intel_fbdev_fini(dev);
 		intel_modeset_cleanup(dev);
-		cancel_work_sync(&dev_priv->console_resume_work);
 
 		/*
 		 * free the memory space allocated for the child device
@@ -1902,7 +1901,6 @@ int i915_driver_unload(struct drm_device *dev)
 		mutex_lock(&dev->struct_mutex);
 		i915_gem_cleanup_ringbuffer(dev);
 		i915_gem_context_fini(dev);
-		WARN_ON(dev_priv->mm.aliasing_ppgtt);
 		mutex_unlock(&dev->struct_mutex);
 		i915_gem_cleanup_stolen(dev);
 
@@ -1910,8 +1908,6 @@ int i915_driver_unload(struct drm_device *dev)
 			i915_free_hws(dev);
 	}
 
-	WARN_ON(!list_empty(&dev_priv->vm_list));
-
 	drm_vblank_cleanup(dev);
 
 	intel_teardown_gmbus(dev);
@@ -1921,7 +1917,7 @@ int i915_driver_unload(struct drm_device *dev)
 	destroy_workqueue(dev_priv->wq);
 	pm_qos_remove_request(&dev_priv->pm_qos);
 
-	dev_priv->gtt.base.cleanup(&dev_priv->gtt.base);
+	i915_global_gtt_cleanup(dev);
 
 	intel_uncore_fini(dev);
 	if (dev_priv->regs != NULL)
@@ -1986,6 +1982,9 @@ void i915_driver_preclose(struct drm_device *dev, struct drm_file *file)
 	i915_gem_context_close(dev, file);
 	i915_gem_release(dev, file);
 	mutex_unlock(&dev->struct_mutex);
+
+	if (drm_core_check_feature(dev, DRIVER_MODESET))
+		intel_modeset_preclose(dev, file);
 }
 
 void i915_driver_postclose(struct drm_device *dev, struct drm_file *file)
diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c
index e27cdbe9d524..055d5e7fbf12 100644
--- a/drivers/gpu/drm/i915/i915_drv.c
+++ b/drivers/gpu/drm/i915/i915_drv.c
@@ -481,6 +481,10 @@ bool i915_semaphore_is_enabled(struct drm_device *dev)
 	if (i915.semaphores >= 0)
 		return i915.semaphores;
 
+	/* TODO: make semaphores and Execlists play nicely together */
+	if (i915.enable_execlists)
+		return false;
+
 	/* Until we get further testing... */
 	if (IS_GEN8(dev))
 		return false;
@@ -524,6 +528,10 @@ static void intel_suspend_encoders(struct drm_i915_private *dev_priv)
 	drm_modeset_unlock_all(dev);
 }
 
+static int intel_suspend_complete(struct drm_i915_private *dev_priv);
+static int intel_resume_prepare(struct drm_i915_private *dev_priv,
+				bool rpm_resume);
+
 static int i915_drm_freeze(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -591,9 +599,7 @@ static int i915_drm_freeze(struct drm_device *dev)
 	intel_uncore_forcewake_reset(dev, false);
 	intel_opregion_fini(dev);
 
-	console_lock();
-	intel_fbdev_set_suspend(dev, FBINFO_STATE_SUSPENDED);
-	console_unlock();
+	intel_fbdev_set_suspend(dev, FBINFO_STATE_SUSPENDED, true);
 
 	dev_priv->suspend_count++;
 
@@ -632,30 +638,20 @@ int i915_suspend(struct drm_device *dev, pm_message_t state)
 	return 0;
 }
 
-void intel_console_resume(struct work_struct *work)
-{
-	struct drm_i915_private *dev_priv =
-		container_of(work, struct drm_i915_private,
-			     console_resume_work);
-	struct drm_device *dev = dev_priv->dev;
-
-	console_lock();
-	intel_fbdev_set_suspend(dev, FBINFO_STATE_RUNNING);
-	console_unlock();
-}
-
 static int i915_drm_thaw_early(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	int ret;
 
-	if (IS_HASWELL(dev) || IS_BROADWELL(dev))
-		hsw_disable_pc8(dev_priv);
+	ret = intel_resume_prepare(dev_priv, false);
+	if (ret)
+		DRM_ERROR("Resume prepare failed: %d,Continuing resume\n", ret);
 
 	intel_uncore_early_sanitize(dev, true);
 	intel_uncore_sanitize(dev);
 	intel_power_domains_init_hw(dev_priv);
 
-	return 0;
+	return ret;
 }
 
 static int __i915_drm_thaw(struct drm_device *dev, bool restore_gtt_mappings)
@@ -714,17 +710,7 @@ static int __i915_drm_thaw(struct drm_device *dev, bool restore_gtt_mappings)
 
 	intel_opregion_init(dev);
 
-	/*
-	 * The console lock can be pretty contented on resume due
-	 * to all the printk activity.  Try to keep it out of the hot
-	 * path of resume if possible.
-	 */
-	if (console_trylock()) {
-		intel_fbdev_set_suspend(dev, FBINFO_STATE_RUNNING);
-		console_unlock();
-	} else {
-		schedule_work(&dev_priv->console_resume_work);
-	}
+	intel_fbdev_set_suspend(dev, FBINFO_STATE_RUNNING, false);
 
 	mutex_lock(&dev_priv->modeset_restore_lock);
 	dev_priv->modeset_restore = MODESET_DONE;
@@ -858,7 +844,13 @@ int i915_reset(struct drm_device *dev)
 			!dev_priv->ums.mm_suspended) {
 		dev_priv->ums.mm_suspended = 0;
 
+		/* Used to prevent gem_check_wedged returning -EAGAIN during gpu reset */
+		dev_priv->gpu_error.reload_in_reset = true;
+
 		ret = i915_gem_init_hw(dev);
+
+		dev_priv->gpu_error.reload_in_reset = false;
+
 		mutex_unlock(&dev->struct_mutex);
 		if (ret) {
 			DRM_ERROR("Failed hw init on reset %d\n", ret);
@@ -879,8 +871,6 @@ int i915_reset(struct drm_device *dev)
 		 */
 		if (INTEL_INFO(dev)->gen > 5)
 			intel_reset_gt_powersave(dev);
-
-		intel_hpd_init(dev);
 	} else {
 		mutex_unlock(&dev->struct_mutex);
 	}
@@ -941,6 +931,7 @@ static int i915_pm_suspend_late(struct device *dev)
 	struct pci_dev *pdev = to_pci_dev(dev);
 	struct drm_device *drm_dev = pci_get_drvdata(pdev);
 	struct drm_i915_private *dev_priv = drm_dev->dev_private;
+	int ret;
 
 	/*
 	 * We have a suspedn ordering issue with the snd-hda driver also
@@ -954,13 +945,16 @@ static int i915_pm_suspend_late(struct device *dev)
 	if (drm_dev->switch_power_state == DRM_SWITCH_POWER_OFF)
 		return 0;
 
-	if (IS_HASWELL(drm_dev) || IS_BROADWELL(drm_dev))
-		hsw_enable_pc8(dev_priv);
+	ret = intel_suspend_complete(dev_priv);
 
-	pci_disable_device(pdev);
-	pci_set_power_state(pdev, PCI_D3hot);
+	if (ret)
+		DRM_ERROR("Suspend complete failed: %d\n", ret);
+	else {
+		pci_disable_device(pdev);
+		pci_set_power_state(pdev, PCI_D3hot);
+	}
 
-	return 0;
+	return ret;
 }
 
 static int i915_pm_resume_early(struct device *dev)
@@ -1016,23 +1010,26 @@ static int i915_pm_poweroff(struct device *dev)
 	return i915_drm_freeze(drm_dev);
 }
 
-static int hsw_runtime_suspend(struct drm_i915_private *dev_priv)
+static int hsw_suspend_complete(struct drm_i915_private *dev_priv)
 {
 	hsw_enable_pc8(dev_priv);
 
 	return 0;
 }
 
-static int snb_runtime_resume(struct drm_i915_private *dev_priv)
+static int snb_resume_prepare(struct drm_i915_private *dev_priv,
+				bool rpm_resume)
 {
 	struct drm_device *dev = dev_priv->dev;
 
-	intel_init_pch_refclk(dev);
+	if (rpm_resume)
+		intel_init_pch_refclk(dev);
 
 	return 0;
 }
 
-static int hsw_runtime_resume(struct drm_i915_private *dev_priv)
+static int hsw_resume_prepare(struct drm_i915_private *dev_priv,
+				bool rpm_resume)
 {
 	hsw_disable_pc8(dev_priv);
 
@@ -1328,7 +1325,7 @@ static void vlv_check_no_gt_access(struct drm_i915_private *dev_priv)
 	I915_WRITE(VLV_GTLC_PW_STATUS, VLV_GTLC_ALLOWWAKEERR);
 }
 
-static int vlv_runtime_suspend(struct drm_i915_private *dev_priv)
+static int vlv_suspend_complete(struct drm_i915_private *dev_priv)
 {
 	u32 mask;
 	int err;
@@ -1368,7 +1365,8 @@ err1:
 	return err;
 }
 
-static int vlv_runtime_resume(struct drm_i915_private *dev_priv)
+static int vlv_resume_prepare(struct drm_i915_private *dev_priv,
+				bool rpm_resume)
 {
 	struct drm_device *dev = dev_priv->dev;
 	int err;
@@ -1393,8 +1391,10 @@ static int vlv_runtime_resume(struct drm_i915_private *dev_priv)
 
 	vlv_check_no_gt_access(dev_priv);
 
-	intel_init_clock_gating(dev);
-	i915_gem_restore_fences(dev);
+	if (rpm_resume) {
+		intel_init_clock_gating(dev);
+		i915_gem_restore_fences(dev);
+	}
 
 	return ret;
 }
@@ -1409,7 +1409,9 @@ static int intel_runtime_suspend(struct device *device)
 	if (WARN_ON_ONCE(!(dev_priv->rps.enabled && intel_enable_rc6(dev))))
 		return -ENODEV;
 
-	WARN_ON(!HAS_RUNTIME_PM(dev));
+	if (WARN_ON_ONCE(!HAS_RUNTIME_PM(dev)))
+		return -ENODEV;
+
 	assert_force_wake_inactive(dev_priv);
 
 	DRM_DEBUG_KMS("Suspending device\n");
@@ -1446,17 +1448,7 @@ static int intel_runtime_suspend(struct device *device)
 	cancel_work_sync(&dev_priv->rps.work);
 	intel_runtime_pm_disable_interrupts(dev);
 
-	if (IS_GEN6(dev)) {
-		ret = 0;
-	} else if (IS_HASWELL(dev) || IS_BROADWELL(dev)) {
-		ret = hsw_runtime_suspend(dev_priv);
-	} else if (IS_VALLEYVIEW(dev)) {
-		ret = vlv_runtime_suspend(dev_priv);
-	} else {
-		ret = -ENODEV;
-		WARN_ON(1);
-	}
-
+	ret = intel_suspend_complete(dev_priv);
 	if (ret) {
 		DRM_ERROR("Runtime suspend failed, disabling it (%d)\n", ret);
 		intel_runtime_pm_restore_interrupts(dev);
@@ -1468,13 +1460,29 @@ static int intel_runtime_suspend(struct device *device)
 	dev_priv->pm.suspended = true;
 
 	/*
-	 * current versions of firmware which depend on this opregion
-	 * notification have repurposed the D1 definition to mean
-	 * "runtime suspended" vs. what you would normally expect (D3)
-	 * to distinguish it from notifications that might be sent
-	 * via the suspend path.
+	 * FIXME: We really should find a document that references the arguments
+	 * used below!
 	 */
-	intel_opregion_notify_adapter(dev, PCI_D1);
+	if (IS_HASWELL(dev)) {
+		/*
+		 * current versions of firmware which depend on this opregion
+		 * notification have repurposed the D1 definition to mean
+		 * "runtime suspended" vs. what you would normally expect (D3)
+		 * to distinguish it from notifications that might be sent via
+		 * the suspend path.
+		 */
+		intel_opregion_notify_adapter(dev, PCI_D1);
+	} else {
+		/*
+		 * On Broadwell, if we use PCI_D1 the PCH DDI ports will stop
+		 * being detected, and the call we do at intel_runtime_resume()
+		 * won't be able to restore them. Since PCI_D3hot matches the
+		 * actual specification and appears to be working, use it. Let's
+		 * assume the other non-Haswell platforms will stay the same as
+		 * Broadwell.
+		 */
+		intel_opregion_notify_adapter(dev, PCI_D3hot);
+	}
 
 	DRM_DEBUG_KMS("Device suspended\n");
 	return 0;
@@ -1487,24 +1495,15 @@ static int intel_runtime_resume(struct device *device)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	int ret;
 
-	WARN_ON(!HAS_RUNTIME_PM(dev));
+	if (WARN_ON_ONCE(!HAS_RUNTIME_PM(dev)))
+		return -ENODEV;
 
 	DRM_DEBUG_KMS("Resuming device\n");
 
 	intel_opregion_notify_adapter(dev, PCI_D0);
 	dev_priv->pm.suspended = false;
 
-	if (IS_GEN6(dev)) {
-		ret = snb_runtime_resume(dev_priv);
-	} else if (IS_HASWELL(dev) || IS_BROADWELL(dev)) {
-		ret = hsw_runtime_resume(dev_priv);
-	} else if (IS_VALLEYVIEW(dev)) {
-		ret = vlv_runtime_resume(dev_priv);
-	} else {
-		WARN_ON(1);
-		ret = -ENODEV;
-	}
-
+	ret = intel_resume_prepare(dev_priv, true);
 	/*
 	 * No point of rolling back things in case of an error, as the best
 	 * we can do is to hope that things will still work (and disable RPM).
@@ -1523,6 +1522,48 @@ static int intel_runtime_resume(struct device *device)
 	return ret;
 }
 
+/*
+ * This function implements common functionality of runtime and system
+ * suspend sequence.
+ */
+static int intel_suspend_complete(struct drm_i915_private *dev_priv)
+{
+	struct drm_device *dev = dev_priv->dev;
+	int ret;
+
+	if (IS_HASWELL(dev) || IS_BROADWELL(dev))
+		ret = hsw_suspend_complete(dev_priv);
+	else if (IS_VALLEYVIEW(dev))
+		ret = vlv_suspend_complete(dev_priv);
+	else
+		ret = 0;
+
+	return ret;
+}
+
+/*
+ * This function implements common functionality of runtime and system
+ * resume sequence. Variable rpm_resume used for implementing different
+ * code paths.
+ */
+static int intel_resume_prepare(struct drm_i915_private *dev_priv,
+				bool rpm_resume)
+{
+	struct drm_device *dev = dev_priv->dev;
+	int ret;
+
+	if (IS_GEN6(dev))
+		ret = snb_resume_prepare(dev_priv, rpm_resume);
+	else if (IS_HASWELL(dev) || IS_BROADWELL(dev))
+		ret = hsw_resume_prepare(dev_priv, rpm_resume);
+	else if (IS_VALLEYVIEW(dev))
+		ret = vlv_resume_prepare(dev_priv, rpm_resume);
+	else
+		ret = 0;
+
+	return ret;
+}
+
 static const struct dev_pm_ops i915_pm_ops = {
 	.suspend = i915_pm_suspend,
 	.suspend_late = i915_pm_suspend_late,
@@ -1572,6 +1613,7 @@ static struct drm_driver driver = {
 	.lastclose = i915_driver_lastclose,
 	.preclose = i915_driver_preclose,
 	.postclose = i915_driver_postclose,
+	.set_busid = drm_pci_set_busid,
 
 	/* Used in place of i915_pm_ops for non-DRIVER_MODESET */
 	.suspend = i915_suspend,
@@ -1663,6 +1705,8 @@ static void __exit i915_exit(void)
 module_init(i915_init);
 module_exit(i915_exit);
 
-MODULE_AUTHOR(DRIVER_AUTHOR);
+MODULE_AUTHOR("Tungsten Graphics, Inc.");
+MODULE_AUTHOR("Intel Corporation");
+
 MODULE_DESCRIPTION(DRIVER_DESC);
 MODULE_LICENSE("GPL and additional rights");
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 3524306d8cfb..16a6f6d187a1 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -35,11 +35,15 @@
 #include "i915_reg.h"
 #include "intel_bios.h"
 #include "intel_ringbuffer.h"
+#include "intel_lrc.h"
 #include "i915_gem_gtt.h"
+#include "i915_gem_render_state.h"
 #include <linux/io-mapping.h>
 #include <linux/i2c.h>
 #include <linux/i2c-algo-bit.h>
 #include <drm/intel-gtt.h>
+#include <drm/drm_legacy.h> /* for struct drm_dma_handle */
+#include <drm/drm_gem.h>
 #include <linux/backlight.h>
 #include <linux/hashtable.h>
 #include <linux/intel-iommu.h>
@@ -49,11 +53,9 @@
 /* General customization:
  */
 
-#define DRIVER_AUTHOR		"Tungsten Graphics, Inc."
-
 #define DRIVER_NAME		"i915"
 #define DRIVER_DESC		"Intel Graphics"
-#define DRIVER_DATE		"20140725"
+#define DRIVER_DATE		"20140905"
 
 enum pipe {
 	INVALID_PIPE = -1,
@@ -162,7 +164,10 @@ enum hpd_pin {
 	 I915_GEM_DOMAIN_INSTRUCTION | \
 	 I915_GEM_DOMAIN_VERTEX)
 
-#define for_each_pipe(p) for ((p) = 0; (p) < INTEL_INFO(dev)->num_pipes; (p)++)
+#define for_each_pipe(__dev_priv, __p) \
+	for ((__p) = 0; (__p) < INTEL_INFO(__dev_priv)->num_pipes; (__p)++)
+#define for_each_plane(pipe, p) \
+	for ((p) = 0; (p) < INTEL_INFO(dev)->num_sprites[(pipe)] + 1; (p)++)
 #define for_each_sprite(p, s) for ((s) = 0; (s) < INTEL_INFO(dev)->num_sprites[(p)]; (s)++)
 
 #define for_each_crtc(dev, crtc) \
@@ -171,6 +176,11 @@ enum hpd_pin {
 #define for_each_intel_crtc(dev, intel_crtc) \
 	list_for_each_entry(intel_crtc, &dev->mode_config.crtc_list, base.head)
 
+#define for_each_intel_encoder(dev, intel_encoder)		\
+	list_for_each_entry(intel_encoder,			\
+			    &(dev)->mode_config.encoder_list,	\
+			    base.head)
+
 #define for_each_encoder_on_crtc(dev, __crtc, intel_encoder) \
 	list_for_each_entry((intel_encoder), &(dev)->mode_config.encoder_list, base.head) \
 		if ((intel_encoder)->base.crtc == (__crtc))
@@ -198,10 +208,13 @@ enum intel_dpll_id {
 #define I915_NUM_PLLS 2
 
 struct intel_dpll_hw_state {
+	/* i9xx, pch plls */
 	uint32_t dpll;
 	uint32_t dpll_md;
 	uint32_t fp0;
 	uint32_t fp1;
+
+	/* hsw, bdw */
 	uint32_t wrpll;
 };
 
@@ -277,8 +290,10 @@ struct intel_opregion {
 struct intel_overlay;
 struct intel_overlay_error_state;
 
+struct drm_local_map;
+
 struct drm_i915_master_private {
-	drm_local_map_t *sarea;
+	struct drm_local_map *sarea;
 	struct _drm_i915_sarea *sarea_priv;
 };
 #define I915_FENCE_REG_NONE -1
@@ -388,6 +403,7 @@ struct drm_i915_error_state {
 		pid_t pid;
 		char comm[TASK_COMM_LEN];
 	} ring[I915_NUM_RINGS];
+
 	struct drm_i915_error_buffer {
 		u32 size;
 		u32 name;
@@ -406,6 +422,7 @@ struct drm_i915_error_state {
 	} **active_bo, **pinned_bo;
 
 	u32 *active_bo_count, *pinned_bo_count;
+	u32 vm_count;
 };
 
 struct intel_connector;
@@ -551,6 +568,7 @@ struct intel_uncore {
 
 struct intel_device_info {
 	u32 display_mmio_offset;
+	u16 device_id;
 	u8 num_pipes:3;
 	u8 num_sprites[I915_MAX_PIPES];
 	u8 gen;
@@ -615,13 +633,21 @@ struct intel_context {
 	uint8_t remap_slice;
 	struct drm_i915_file_private *file_priv;
 	struct i915_ctx_hang_stats hang_stats;
-	struct i915_address_space *vm;
+	struct i915_hw_ppgtt *ppgtt;
 
+	/* Legacy ring buffer submission */
 	struct {
 		struct drm_i915_gem_object *rcs_state;
 		bool initialized;
 	} legacy_hw_ctx;
 
+	/* Execlists */
+	bool rcs_initialized;
+	struct {
+		struct drm_i915_gem_object *state;
+		struct intel_ringbuffer *ringbuf;
+	} engine[I915_NUM_RINGS];
+
 	struct list_head link;
 };
 
@@ -635,6 +661,8 @@ struct i915_fbc {
 	struct drm_mm_node compressed_fb;
 	struct drm_mm_node *compressed_llb;
 
+	bool false_color;
+
 	struct intel_fbc_work {
 		struct delayed_work work;
 		struct drm_crtc *crtc;
@@ -688,6 +716,7 @@ enum intel_sbi_destination {
 #define QUIRK_LVDS_SSC_DISABLE (1<<1)
 #define QUIRK_INVERT_BRIGHTNESS (1<<2)
 #define QUIRK_BACKLIGHT_PRESENT (1<<3)
+#define QUIRK_PIPEB_FORCE (1<<4)
 
 struct intel_fbdev;
 struct intel_fbc_work;
@@ -1147,6 +1176,7 @@ struct i915_gem_mm {
 };
 
 struct drm_i915_error_state_buf {
+	struct drm_i915_private *i915;
 	unsigned bytes;
 	unsigned size;
 	int err;
@@ -1219,6 +1249,9 @@ struct i915_gpu_error {
 
 	/* For missed irq/seqno simulation. */
 	unsigned int test_irq_rings;
+
+	/* Used to prevent gem_check_wedged returning -EAGAIN during gpu reset   */
+	bool reload_in_reset;
 };
 
 enum modeset_restore {
@@ -1228,6 +1261,12 @@ enum modeset_restore {
 };
 
 struct ddi_vbt_port_info {
+	/*
+	 * This is an index in the HDMI/DVI DDI buffer translation table.
+	 * The special value HDMI_LEVEL_SHIFT_UNKNOWN means the VBT didn't
+	 * populate this field.
+	 */
+#define HDMI_LEVEL_SHIFT_UNKNOWN	0xff
 	uint8_t hdmi_level_shift;
 
 	uint8_t supports_dvi:1;
@@ -1421,7 +1460,7 @@ struct drm_i915_private {
 	struct drm_i915_gem_object *semaphore_obj;
 	uint32_t last_seqno, next_seqno;
 
-	drm_dma_handle_t *status_page_dmah;
+	struct drm_dma_handle *status_page_dmah;
 	struct resource mch_res;
 
 	/* protects the irq masks */
@@ -1475,6 +1514,9 @@ struct drm_i915_private {
 	/* LVDS info */
 	bool no_aux_handshake;
 
+	/* protects panel power sequencer state */
+	struct mutex pps_mutex;
+
 	struct drm_i915_fence_reg fence_regs[I915_MAX_NUM_FENCES]; /* assume 965 */
 	int fence_reg_start; /* 4 if userland hasn't ioctl'd us yet */
 	int num_fence_regs; /* 8 on pre-965, 16 otherwise */
@@ -1526,6 +1568,20 @@ struct drm_i915_private {
 	struct intel_shared_dpll shared_dplls[I915_NUM_PLLS];
 	int dpio_phy_iosf_port[I915_NUM_PHYS_VLV];
 
+	/*
+	 * workarounds are currently applied at different places and
+	 * changes are being done to consolidate them so exact count is
+	 * not clear at this point, use a max value for now.
+	 */
+#define I915_MAX_WA_REGS  16
+	struct {
+		u32 addr;
+		u32 value;
+		/* bitmask representing WA bits */
+		u32 mask;
+	} intel_wa_regs[I915_MAX_WA_REGS];
+	u32 num_wa_regs;
+
 	/* Reclocking support */
 	bool render_reclock_avail;
 	bool lvds_downclock_avail;
@@ -1561,14 +1617,9 @@ struct drm_i915_private {
 #ifdef CONFIG_DRM_I915_FBDEV
 	/* list of fbdev register on this device */
 	struct intel_fbdev *fbdev;
+	struct work_struct fbdev_suspend_work;
 #endif
 
-	/*
-	 * The console may be contended at resume, but we don't
-	 * want it to block on it.
-	 */
-	struct work_struct console_resume_work;
-
 	struct drm_property *broadcast_rgb_property;
 	struct drm_property *force_audio_property;
 
@@ -1614,12 +1665,28 @@ struct drm_i915_private {
 	 */
 	struct workqueue_struct *dp_wq;
 
+	uint32_t bios_vgacntr;
+
 	/* Old dri1 support infrastructure, beware the dragons ya fools entering
 	 * here! */
 	struct i915_dri1_state dri1;
 	/* Old ums support infrastructure, same warning applies. */
 	struct i915_ums_state ums;
 
+	/* Abstract the submission mechanism (legacy ringbuffer or execlists) away */
+	struct {
+		int (*do_execbuf)(struct drm_device *dev, struct drm_file *file,
+				  struct intel_engine_cs *ring,
+				  struct intel_context *ctx,
+				  struct drm_i915_gem_execbuffer2 *args,
+				  struct list_head *vmas,
+				  struct drm_i915_gem_object *batch_obj,
+				  u64 exec_start, u32 flags);
+		int (*init_rings)(struct drm_device *dev);
+		void (*cleanup_ring)(struct intel_engine_cs *ring);
+		void (*stop_ring)(struct intel_engine_cs *ring);
+	} gt;
+
 	/*
 	 * NOTE: This is the dri1/ums dungeon, don't add stuff here. Your patch
 	 * will be rejected. Instead look for a better place.
@@ -1761,13 +1828,6 @@ struct drm_i915_gem_object {
 	 * Only honoured if hardware has relevant pte bit
 	 */
 	unsigned long gt_ro:1;
-
-	/*
-	 * Is the GPU currently using a fence to access this buffer,
-	 */
-	unsigned int pending_fenced_gpu_access:1;
-	unsigned int fenced_gpu_access:1;
-
 	unsigned int cache_level:3;
 
 	unsigned int has_aliasing_ppgtt_mapping:1;
@@ -1805,7 +1865,7 @@ struct drm_i915_gem_object {
 	struct drm_file *pin_filp;
 
 	/** for phy allocated objects */
-	drm_dma_handle_t *phys_handle;
+	struct drm_dma_handle *phys_handle;
 
 	union {
 		struct i915_gem_userptr {
@@ -1971,51 +2031,63 @@ struct drm_i915_cmd_table {
 	int count;
 };
 
-#define INTEL_INFO(dev)	(&to_i915(dev)->info)
-
-#define IS_I830(dev)		((dev)->pdev->device == 0x3577)
-#define IS_845G(dev)		((dev)->pdev->device == 0x2562)
+/* Note that the (struct drm_i915_private *) cast is just to shut up gcc. */
+#define __I915__(p) ({ \
+	struct drm_i915_private *__p; \
+	if (__builtin_types_compatible_p(typeof(*p), struct drm_i915_private)) \
+		__p = (struct drm_i915_private *)p; \
+	else if (__builtin_types_compatible_p(typeof(*p), struct drm_device)) \
+		__p = to_i915((struct drm_device *)p); \
+	else \
+		BUILD_BUG(); \
+	__p; \
+})
+#define INTEL_INFO(p) 	(&__I915__(p)->info)
+#define INTEL_DEVID(p)	(INTEL_INFO(p)->device_id)
+
+#define IS_I830(dev)		(INTEL_DEVID(dev) == 0x3577)
+#define IS_845G(dev)		(INTEL_DEVID(dev) == 0x2562)
 #define IS_I85X(dev)		(INTEL_INFO(dev)->is_i85x)
-#define IS_I865G(dev)		((dev)->pdev->device == 0x2572)
+#define IS_I865G(dev)		(INTEL_DEVID(dev) == 0x2572)
 #define IS_I915G(dev)		(INTEL_INFO(dev)->is_i915g)
-#define IS_I915GM(dev)		((dev)->pdev->device == 0x2592)
-#define IS_I945G(dev)		((dev)->pdev->device == 0x2772)
+#define IS_I915GM(dev)		(INTEL_DEVID(dev) == 0x2592)
+#define IS_I945G(dev)		(INTEL_DEVID(dev) == 0x2772)
 #define IS_I945GM(dev)		(INTEL_INFO(dev)->is_i945gm)
 #define IS_BROADWATER(dev)	(INTEL_INFO(dev)->is_broadwater)
 #define IS_CRESTLINE(dev)	(INTEL_INFO(dev)->is_crestline)
-#define IS_GM45(dev)		((dev)->pdev->device == 0x2A42)
+#define IS_GM45(dev)		(INTEL_DEVID(dev) == 0x2A42)
 #define IS_G4X(dev)		(INTEL_INFO(dev)->is_g4x)
-#define IS_PINEVIEW_G(dev)	((dev)->pdev->device == 0xa001)
-#define IS_PINEVIEW_M(dev)	((dev)->pdev->device == 0xa011)
+#define IS_PINEVIEW_G(dev)	(INTEL_DEVID(dev) == 0xa001)
+#define IS_PINEVIEW_M(dev)	(INTEL_DEVID(dev) == 0xa011)
 #define IS_PINEVIEW(dev)	(INTEL_INFO(dev)->is_pineview)
 #define IS_G33(dev)		(INTEL_INFO(dev)->is_g33)
-#define IS_IRONLAKE_M(dev)	((dev)->pdev->device == 0x0046)
+#define IS_IRONLAKE_M(dev)	(INTEL_DEVID(dev) == 0x0046)
 #define IS_IVYBRIDGE(dev)	(INTEL_INFO(dev)->is_ivybridge)
-#define IS_IVB_GT1(dev)		((dev)->pdev->device == 0x0156 || \
-				 (dev)->pdev->device == 0x0152 || \
-				 (dev)->pdev->device == 0x015a)
-#define IS_SNB_GT1(dev)		((dev)->pdev->device == 0x0102 || \
-				 (dev)->pdev->device == 0x0106 || \
-				 (dev)->pdev->device == 0x010A)
+#define IS_IVB_GT1(dev)		(INTEL_DEVID(dev) == 0x0156 || \
+				 INTEL_DEVID(dev) == 0x0152 || \
+				 INTEL_DEVID(dev) == 0x015a)
+#define IS_SNB_GT1(dev)		(INTEL_DEVID(dev) == 0x0102 || \
+				 INTEL_DEVID(dev) == 0x0106 || \
+				 INTEL_DEVID(dev) == 0x010A)
 #define IS_VALLEYVIEW(dev)	(INTEL_INFO(dev)->is_valleyview)
 #define IS_CHERRYVIEW(dev)	(INTEL_INFO(dev)->is_valleyview && IS_GEN8(dev))
 #define IS_HASWELL(dev)	(INTEL_INFO(dev)->is_haswell)
 #define IS_BROADWELL(dev)	(!INTEL_INFO(dev)->is_valleyview && IS_GEN8(dev))
 #define IS_MOBILE(dev)		(INTEL_INFO(dev)->is_mobile)
 #define IS_HSW_EARLY_SDV(dev)	(IS_HASWELL(dev) && \
-				 ((dev)->pdev->device & 0xFF00) == 0x0C00)
+				 (INTEL_DEVID(dev) & 0xFF00) == 0x0C00)
 #define IS_BDW_ULT(dev)		(IS_BROADWELL(dev) && \
-				 (((dev)->pdev->device & 0xf) == 0x2  || \
-				 ((dev)->pdev->device & 0xf) == 0x6 || \
-				 ((dev)->pdev->device & 0xf) == 0xe))
+				 ((INTEL_DEVID(dev) & 0xf) == 0x2  || \
+				 (INTEL_DEVID(dev) & 0xf) == 0x6 || \
+				 (INTEL_DEVID(dev) & 0xf) == 0xe))
 #define IS_HSW_ULT(dev)		(IS_HASWELL(dev) && \
-				 ((dev)->pdev->device & 0xFF00) == 0x0A00)
+				 (INTEL_DEVID(dev) & 0xFF00) == 0x0A00)
 #define IS_ULT(dev)		(IS_HSW_ULT(dev) || IS_BDW_ULT(dev))
 #define IS_HSW_GT3(dev)		(IS_HASWELL(dev) && \
-				 ((dev)->pdev->device & 0x00F0) == 0x0020)
+				 (INTEL_DEVID(dev) & 0x00F0) == 0x0020)
 /* ULX machines are also considered ULT. */
-#define IS_HSW_ULX(dev)		((dev)->pdev->device == 0x0A0E || \
-				 (dev)->pdev->device == 0x0A1E)
+#define IS_HSW_ULX(dev)		(INTEL_DEVID(dev) == 0x0A0E || \
+				 INTEL_DEVID(dev) == 0x0A1E)
 #define IS_PRELIMINARY_HW(intel_info) ((intel_info)->is_preliminary)
 
 /*
@@ -2047,10 +2119,11 @@ struct drm_i915_cmd_table {
 #define I915_NEED_GFX_HWS(dev)	(INTEL_INFO(dev)->need_gfx_hws)
 
 #define HAS_HW_CONTEXTS(dev)	(INTEL_INFO(dev)->gen >= 6)
+#define HAS_LOGICAL_RING_CONTEXTS(dev)	(INTEL_INFO(dev)->gen >= 8)
 #define HAS_ALIASING_PPGTT(dev)	(INTEL_INFO(dev)->gen >= 6)
 #define HAS_PPGTT(dev)		(INTEL_INFO(dev)->gen >= 7 && !IS_GEN8(dev))
-#define USES_PPGTT(dev)		intel_enable_ppgtt(dev, false)
-#define USES_FULL_PPGTT(dev)	intel_enable_ppgtt(dev, true)
+#define USES_PPGTT(dev)		(i915.enable_ppgtt)
+#define USES_FULL_PPGTT(dev)	(i915.enable_ppgtt == 2)
 
 #define HAS_OVERLAY(dev)		(INTEL_INFO(dev)->has_overlay)
 #define OVERLAY_NEEDS_PHYSICAL(dev)	(INTEL_INFO(dev)->overlay_needs_physical)
@@ -2134,6 +2207,7 @@ struct i915_params {
 	int enable_rc6;
 	int enable_fbc;
 	int enable_ppgtt;
+	int enable_execlists;
 	int enable_psr;
 	unsigned int preliminary_hw_support;
 	int disable_power_well;
@@ -2180,8 +2254,6 @@ extern void i915_update_gfx_val(struct drm_i915_private *dev_priv);
 int vlv_force_gfx_clock(struct drm_i915_private *dev_priv, bool on);
 void intel_hpd_cancel_work(struct drm_i915_private *dev_priv);
 
-extern void intel_console_resume(struct work_struct *work);
-
 /* i915_irq.c */
 void i915_queue_hangcheck(struct drm_device *dev);
 __printf(3, 4)
@@ -2229,6 +2301,20 @@ int i915_gem_set_domain_ioctl(struct drm_device *dev, void *data,
 			      struct drm_file *file_priv);
 int i915_gem_sw_finish_ioctl(struct drm_device *dev, void *data,
 			     struct drm_file *file_priv);
+void i915_gem_execbuffer_move_to_active(struct list_head *vmas,
+					struct intel_engine_cs *ring);
+void i915_gem_execbuffer_retire_commands(struct drm_device *dev,
+					 struct drm_file *file,
+					 struct intel_engine_cs *ring,
+					 struct drm_i915_gem_object *obj);
+int i915_gem_ringbuffer_submission(struct drm_device *dev,
+				   struct drm_file *file,
+				   struct intel_engine_cs *ring,
+				   struct intel_context *ctx,
+				   struct drm_i915_gem_execbuffer2 *args,
+				   struct list_head *vmas,
+				   struct drm_i915_gem_object *batch_obj,
+				   u64 exec_start, u32 flags);
 int i915_gem_execbuffer(struct drm_device *dev, void *data,
 			struct drm_file *file_priv);
 int i915_gem_execbuffer2(struct drm_device *dev, void *data,
@@ -2263,6 +2349,12 @@ int i915_gem_get_aperture_ioctl(struct drm_device *dev, void *data,
 int i915_gem_wait_ioctl(struct drm_device *dev, void *data,
 			struct drm_file *file_priv);
 void i915_gem_load(struct drm_device *dev);
+unsigned long i915_gem_shrink(struct drm_i915_private *dev_priv,
+			      long target,
+			      unsigned flags);
+#define I915_SHRINK_PURGEABLE 0x1
+#define I915_SHRINK_UNBOUND 0x2
+#define I915_SHRINK_BOUND 0x4
 void *i915_gem_object_alloc(struct drm_device *dev);
 void i915_gem_object_free(struct drm_i915_gem_object *obj);
 void i915_gem_object_init(struct drm_i915_gem_object *obj,
@@ -2381,6 +2473,7 @@ void i915_gem_reset(struct drm_device *dev);
 bool i915_gem_clflush_object(struct drm_i915_gem_object *obj, bool force);
 int __must_check i915_gem_object_finish_gpu(struct drm_i915_gem_object *obj);
 int __must_check i915_gem_init(struct drm_device *dev);
+int i915_gem_init_rings(struct drm_device *dev);
 int __must_check i915_gem_init_hw(struct drm_device *dev);
 int i915_gem_l3_remap(struct intel_engine_cs *ring, int slice);
 void i915_gem_init_swizzling(struct drm_device *dev);
@@ -2451,7 +2544,7 @@ static inline bool i915_gem_obj_is_pinned(struct drm_i915_gem_object *obj) {
 }
 
 /* Some GGTT VM helpers */
-#define obj_to_ggtt(obj) \
+#define i915_obj_to_ggtt(obj) \
 	(&((struct drm_i915_private *)(obj)->base.dev->dev_private)->gtt.base)
 static inline bool i915_is_ggtt(struct i915_address_space *vm)
 {
@@ -2460,21 +2553,30 @@ static inline bool i915_is_ggtt(struct i915_address_space *vm)
 	return vm == ggtt;
 }
 
+static inline struct i915_hw_ppgtt *
+i915_vm_to_ppgtt(struct i915_address_space *vm)
+{
+	WARN_ON(i915_is_ggtt(vm));
+
+	return container_of(vm, struct i915_hw_ppgtt, base);
+}
+
+
 static inline bool i915_gem_obj_ggtt_bound(struct drm_i915_gem_object *obj)
 {
-	return i915_gem_obj_bound(obj, obj_to_ggtt(obj));
+	return i915_gem_obj_bound(obj, i915_obj_to_ggtt(obj));
 }
 
 static inline unsigned long
 i915_gem_obj_ggtt_offset(struct drm_i915_gem_object *obj)
 {
-	return i915_gem_obj_offset(obj, obj_to_ggtt(obj));
+	return i915_gem_obj_offset(obj, i915_obj_to_ggtt(obj));
 }
 
 static inline unsigned long
 i915_gem_obj_ggtt_size(struct drm_i915_gem_object *obj)
 {
-	return i915_gem_obj_size(obj, obj_to_ggtt(obj));
+	return i915_gem_obj_size(obj, i915_obj_to_ggtt(obj));
 }
 
 static inline int __must_check
@@ -2482,7 +2584,8 @@ i915_gem_obj_ggtt_pin(struct drm_i915_gem_object *obj,
 		      uint32_t alignment,
 		      unsigned flags)
 {
-	return i915_gem_object_pin(obj, obj_to_ggtt(obj), alignment, flags | PIN_GLOBAL);
+	return i915_gem_object_pin(obj, i915_obj_to_ggtt(obj),
+				   alignment, flags | PIN_GLOBAL);
 }
 
 static inline int
@@ -2494,7 +2597,6 @@ i915_gem_object_ggtt_unbind(struct drm_i915_gem_object *obj)
 void i915_gem_object_ggtt_unpin(struct drm_i915_gem_object *obj);
 
 /* i915_gem_context.c */
-#define ctx_to_ppgtt(ctx) container_of((ctx)->vm, struct i915_hw_ppgtt, base)
 int __must_check i915_gem_context_init(struct drm_device *dev);
 void i915_gem_context_fini(struct drm_device *dev);
 void i915_gem_context_reset(struct drm_device *dev);
@@ -2506,6 +2608,8 @@ int i915_switch_context(struct intel_engine_cs *ring,
 struct intel_context *
 i915_gem_context_get(struct drm_i915_file_private *file_priv, u32 id);
 void i915_gem_context_free(struct kref *ctx_ref);
+struct drm_i915_gem_object *
+i915_gem_alloc_context_obj(struct drm_device *dev, size_t size);
 static inline void i915_gem_context_reference(struct intel_context *ctx)
 {
 	kref_get(&ctx->ref);
@@ -2526,8 +2630,6 @@ int i915_gem_context_create_ioctl(struct drm_device *dev, void *data,
 int i915_gem_context_destroy_ioctl(struct drm_device *dev, void *data,
 				   struct drm_file *file);
 
-/* i915_gem_render_state.c */
-int i915_gem_render_state_init(struct intel_engine_cs *ring);
 /* i915_gem_evict.c */
 int __must_check i915_gem_evict_something(struct drm_device *dev,
 					  struct i915_address_space *vm,
@@ -2595,6 +2697,7 @@ void i915_error_printf(struct drm_i915_error_state_buf *e, const char *f, ...);
 int i915_error_state_to_str(struct drm_i915_error_state_buf *estr,
 			    const struct i915_error_state_file_priv *error);
 int i915_error_state_buf_init(struct drm_i915_error_state_buf *eb,
+			      struct drm_i915_private *i915,
 			      size_t count, loff_t pos);
 static inline void i915_error_state_buf_release(
 	struct drm_i915_error_state_buf *eb)
@@ -2609,7 +2712,7 @@ void i915_error_state_put(struct i915_error_state_file_priv *error_priv);
 void i915_destroy_error_state(struct drm_device *dev);
 
 void i915_get_extra_instdone(struct drm_device *dev, uint32_t *instdone);
-const char *i915_cache_level_str(int type);
+const char *i915_cache_level_str(struct drm_i915_private *i915, int type);
 
 /* i915_cmd_parser.c */
 int i915_cmd_parser_get_version(void);
@@ -2701,6 +2804,7 @@ extern void intel_modeset_setup_hw_state(struct drm_device *dev,
 extern void i915_redisable_vga(struct drm_device *dev);
 extern void i915_redisable_vga_power_on(struct drm_device *dev);
 extern bool intel_fbc_enabled(struct drm_device *dev);
+extern void gen8_fbc_sw_flush(struct drm_device *dev, u32 value);
 extern void intel_disable_fbc(struct drm_device *dev);
 extern bool ironlake_set_drps(struct drm_device *dev, u8 val);
 extern void intel_init_pch_refclk(struct drm_device *dev);
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index ad55b06a3cb1..28f91df2604d 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -60,7 +60,6 @@ static unsigned long i915_gem_shrinker_scan(struct shrinker *shrinker,
 static int i915_gem_shrinker_oom(struct notifier_block *nb,
 				 unsigned long event,
 				 void *ptr);
-static unsigned long i915_gem_purge(struct drm_i915_private *dev_priv, long target);
 static unsigned long i915_gem_shrink_all(struct drm_i915_private *dev_priv);
 
 static bool cpu_cache_is_coherent(struct drm_device *dev,
@@ -1085,7 +1084,13 @@ i915_gem_check_wedge(struct i915_gpu_error *error,
 		if (i915_terminally_wedged(error))
 			return -EIO;
 
-		return -EAGAIN;
+		/*
+		 * Check if GPU Reset is in progress - we need intel_ring_begin
+		 * to work properly to reinit the hw state while the gpu is
+		 * still marked as reset-in-progress. Handle this with a flag.
+		 */
+		if (!error->reload_in_reset)
+			return -EAGAIN;
 	}
 
 	return 0;
@@ -1735,7 +1740,11 @@ static int i915_gem_object_create_mmap_offset(struct drm_i915_gem_object *obj)
 	 * offsets on purgeable objects by truncating it and marking it purged,
 	 * which prevents userspace from ever using that object again.
 	 */
-	i915_gem_purge(dev_priv, obj->base.size >> PAGE_SHIFT);
+	i915_gem_shrink(dev_priv,
+			obj->base.size >> PAGE_SHIFT,
+			I915_SHRINK_BOUND |
+			I915_SHRINK_UNBOUND |
+			I915_SHRINK_PURGEABLE);
 	ret = drm_gem_create_mmap_offset(&obj->base);
 	if (ret != -ENOSPC)
 		goto out;
@@ -1932,12 +1941,11 @@ i915_gem_object_put_pages(struct drm_i915_gem_object *obj)
 	return 0;
 }
 
-static unsigned long
-__i915_gem_shrink(struct drm_i915_private *dev_priv, long target,
-		  bool purgeable_only)
+unsigned long
+i915_gem_shrink(struct drm_i915_private *dev_priv,
+		long target, unsigned flags)
 {
-	struct list_head still_in_list;
-	struct drm_i915_gem_object *obj;
+	const bool purgeable_only = flags & I915_SHRINK_PURGEABLE;
 	unsigned long count = 0;
 
 	/*
@@ -1959,62 +1967,68 @@ __i915_gem_shrink(struct drm_i915_private *dev_priv, long target,
 	 * dev->struct_mutex and so we won't ever be able to observe an
 	 * object on the bound_list with a reference count equals 0.
 	 */
-	INIT_LIST_HEAD(&still_in_list);
-	while (count < target && !list_empty(&dev_priv->mm.unbound_list)) {
-		obj = list_first_entry(&dev_priv->mm.unbound_list,
-				       typeof(*obj), global_list);
-		list_move_tail(&obj->global_list, &still_in_list);
+	if (flags & I915_SHRINK_UNBOUND) {
+		struct list_head still_in_list;
 
-		if (!i915_gem_object_is_purgeable(obj) && purgeable_only)
-			continue;
+		INIT_LIST_HEAD(&still_in_list);
+		while (count < target && !list_empty(&dev_priv->mm.unbound_list)) {
+			struct drm_i915_gem_object *obj;
 
-		drm_gem_object_reference(&obj->base);
+			obj = list_first_entry(&dev_priv->mm.unbound_list,
+					       typeof(*obj), global_list);
+			list_move_tail(&obj->global_list, &still_in_list);
 
-		if (i915_gem_object_put_pages(obj) == 0)
-			count += obj->base.size >> PAGE_SHIFT;
+			if (!i915_gem_object_is_purgeable(obj) && purgeable_only)
+				continue;
 
-		drm_gem_object_unreference(&obj->base);
+			drm_gem_object_reference(&obj->base);
+
+			if (i915_gem_object_put_pages(obj) == 0)
+				count += obj->base.size >> PAGE_SHIFT;
+
+			drm_gem_object_unreference(&obj->base);
+		}
+		list_splice(&still_in_list, &dev_priv->mm.unbound_list);
 	}
-	list_splice(&still_in_list, &dev_priv->mm.unbound_list);
 
-	INIT_LIST_HEAD(&still_in_list);
-	while (count < target && !list_empty(&dev_priv->mm.bound_list)) {
-		struct i915_vma *vma, *v;
+	if (flags & I915_SHRINK_BOUND) {
+		struct list_head still_in_list;
 
-		obj = list_first_entry(&dev_priv->mm.bound_list,
-				       typeof(*obj), global_list);
-		list_move_tail(&obj->global_list, &still_in_list);
+		INIT_LIST_HEAD(&still_in_list);
+		while (count < target && !list_empty(&dev_priv->mm.bound_list)) {
+			struct drm_i915_gem_object *obj;
+			struct i915_vma *vma, *v;
 
-		if (!i915_gem_object_is_purgeable(obj) && purgeable_only)
-			continue;
+			obj = list_first_entry(&dev_priv->mm.bound_list,
+					       typeof(*obj), global_list);
+			list_move_tail(&obj->global_list, &still_in_list);
 
-		drm_gem_object_reference(&obj->base);
+			if (!i915_gem_object_is_purgeable(obj) && purgeable_only)
+				continue;
 
-		list_for_each_entry_safe(vma, v, &obj->vma_list, vma_link)
-			if (i915_vma_unbind(vma))
-				break;
+			drm_gem_object_reference(&obj->base);
 
-		if (i915_gem_object_put_pages(obj) == 0)
-			count += obj->base.size >> PAGE_SHIFT;
+			list_for_each_entry_safe(vma, v, &obj->vma_list, vma_link)
+				if (i915_vma_unbind(vma))
+					break;
 
-		drm_gem_object_unreference(&obj->base);
+			if (i915_gem_object_put_pages(obj) == 0)
+				count += obj->base.size >> PAGE_SHIFT;
+
+			drm_gem_object_unreference(&obj->base);
+		}
+		list_splice(&still_in_list, &dev_priv->mm.bound_list);
 	}
-	list_splice(&still_in_list, &dev_priv->mm.bound_list);
 
 	return count;
 }
 
 static unsigned long
-i915_gem_purge(struct drm_i915_private *dev_priv, long target)
-{
-	return __i915_gem_shrink(dev_priv, target, true);
-}
-
-static unsigned long
 i915_gem_shrink_all(struct drm_i915_private *dev_priv)
 {
 	i915_gem_evict_everything(dev_priv->dev);
-	return __i915_gem_shrink(dev_priv, LONG_MAX, false);
+	return i915_gem_shrink(dev_priv, LONG_MAX,
+			       I915_SHRINK_BOUND | I915_SHRINK_UNBOUND);
 }
 
 static int
@@ -2061,7 +2075,11 @@ i915_gem_object_get_pages_gtt(struct drm_i915_gem_object *obj)
 	for (i = 0; i < page_count; i++) {
 		page = shmem_read_mapping_page_gfp(mapping, i, gfp);
 		if (IS_ERR(page)) {
-			i915_gem_purge(dev_priv, page_count);
+			i915_gem_shrink(dev_priv,
+					page_count,
+					I915_SHRINK_BOUND |
+					I915_SHRINK_UNBOUND |
+					I915_SHRINK_PURGEABLE);
 			page = shmem_read_mapping_page_gfp(mapping, i, gfp);
 		}
 		if (IS_ERR(page)) {
@@ -2163,8 +2181,6 @@ static void
 i915_gem_object_move_to_active(struct drm_i915_gem_object *obj,
 			       struct intel_engine_cs *ring)
 {
-	struct drm_device *dev = obj->base.dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 seqno = intel_ring_get_seqno(ring);
 
 	BUG_ON(ring == NULL);
@@ -2183,19 +2199,6 @@ i915_gem_object_move_to_active(struct drm_i915_gem_object *obj,
 	list_move_tail(&obj->ring_list, &ring->active_list);
 
 	obj->last_read_seqno = seqno;
-
-	if (obj->fenced_gpu_access) {
-		obj->last_fenced_seqno = seqno;
-
-		/* Bump MRU to take account of the delayed flush */
-		if (obj->fence_reg != I915_FENCE_REG_NONE) {
-			struct drm_i915_fence_reg *reg;
-
-			reg = &dev_priv->fence_regs[obj->fence_reg];
-			list_move_tail(&reg->lru_list,
-				       &dev_priv->mm.fence_list);
-		}
-	}
 }
 
 void i915_vma_move_to_active(struct i915_vma *vma,
@@ -2231,7 +2234,6 @@ i915_gem_object_move_to_inactive(struct drm_i915_gem_object *obj)
 	obj->base.write_domain = 0;
 
 	obj->last_fenced_seqno = 0;
-	obj->fenced_gpu_access = false;
 
 	obj->active = 0;
 	drm_gem_object_unreference(&obj->base);
@@ -2329,10 +2331,21 @@ int __i915_add_request(struct intel_engine_cs *ring,
 {
 	struct drm_i915_private *dev_priv = ring->dev->dev_private;
 	struct drm_i915_gem_request *request;
+	struct intel_ringbuffer *ringbuf;
 	u32 request_ring_position, request_start;
 	int ret;
 
-	request_start = intel_ring_get_tail(ring->buffer);
+	request = ring->preallocated_lazy_request;
+	if (WARN_ON(request == NULL))
+		return -ENOMEM;
+
+	if (i915.enable_execlists) {
+		struct intel_context *ctx = request->ctx;
+		ringbuf = ctx->engine[ring->id].ringbuf;
+	} else
+		ringbuf = ring->buffer;
+
+	request_start = intel_ring_get_tail(ringbuf);
 	/*
 	 * Emit any outstanding flushes - execbuf can fail to emit the flush
 	 * after having emitted the batchbuffer command. Hence we need to fix
@@ -2340,24 +2353,32 @@ int __i915_add_request(struct intel_engine_cs *ring,
 	 * is that the flush _must_ happen before the next request, no matter
 	 * what.
 	 */
-	ret = intel_ring_flush_all_caches(ring);
-	if (ret)
-		return ret;
-
-	request = ring->preallocated_lazy_request;
-	if (WARN_ON(request == NULL))
-		return -ENOMEM;
+	if (i915.enable_execlists) {
+		ret = logical_ring_flush_all_caches(ringbuf);
+		if (ret)
+			return ret;
+	} else {
+		ret = intel_ring_flush_all_caches(ring);
+		if (ret)
+			return ret;
+	}
 
 	/* Record the position of the start of the request so that
 	 * should we detect the updated seqno part-way through the
 	 * GPU processing the request, we never over-estimate the
 	 * position of the head.
 	 */
-	request_ring_position = intel_ring_get_tail(ring->buffer);
+	request_ring_position = intel_ring_get_tail(ringbuf);
 
-	ret = ring->add_request(ring);
-	if (ret)
-		return ret;
+	if (i915.enable_execlists) {
+		ret = ring->emit_request(ringbuf);
+		if (ret)
+			return ret;
+	} else {
+		ret = ring->add_request(ring);
+		if (ret)
+			return ret;
+	}
 
 	request->seqno = intel_ring_get_seqno(ring);
 	request->ring = ring;
@@ -2372,12 +2393,14 @@ int __i915_add_request(struct intel_engine_cs *ring,
 	 */
 	request->batch_obj = obj;
 
-	/* Hold a reference to the current context so that we can inspect
-	 * it later in case a hangcheck error event fires.
-	 */
-	request->ctx = ring->last_context;
-	if (request->ctx)
-		i915_gem_context_reference(request->ctx);
+	if (!i915.enable_execlists) {
+		/* Hold a reference to the current context so that we can inspect
+		 * it later in case a hangcheck error event fires.
+		 */
+		request->ctx = ring->last_context;
+		if (request->ctx)
+			i915_gem_context_reference(request->ctx);
+	}
 
 	request->emitted_jiffies = jiffies;
 	list_add_tail(&request->list, &ring->request_list);
@@ -2548,6 +2571,18 @@ static void i915_gem_reset_ring_cleanup(struct drm_i915_private *dev_priv,
 		i915_gem_free_request(request);
 	}
 
+	while (!list_empty(&ring->execlist_queue)) {
+		struct intel_ctx_submit_request *submit_req;
+
+		submit_req = list_first_entry(&ring->execlist_queue,
+				struct intel_ctx_submit_request,
+				execlist_link);
+		list_del(&submit_req->execlist_link);
+		intel_runtime_pm_put(dev_priv);
+		i915_gem_context_unreference(submit_req->ctx);
+		kfree(submit_req);
+	}
+
 	/* These may not have been flush before the reset, do so now */
 	kfree(ring->preallocated_lazy_request);
 	ring->preallocated_lazy_request = NULL;
@@ -2632,6 +2667,7 @@ i915_gem_retire_requests_ring(struct intel_engine_cs *ring)
 
 	while (!list_empty(&ring->request_list)) {
 		struct drm_i915_gem_request *request;
+		struct intel_ringbuffer *ringbuf;
 
 		request = list_first_entry(&ring->request_list,
 					   struct drm_i915_gem_request,
@@ -2641,12 +2677,24 @@ i915_gem_retire_requests_ring(struct intel_engine_cs *ring)
 			break;
 
 		trace_i915_gem_request_retire(ring, request->seqno);
+
+		/* This is one of the few common intersection points
+		 * between legacy ringbuffer submission and execlists:
+		 * we need to tell them apart in order to find the correct
+		 * ringbuffer to which the request belongs to.
+		 */
+		if (i915.enable_execlists) {
+			struct intel_context *ctx = request->ctx;
+			ringbuf = ctx->engine[ring->id].ringbuf;
+		} else
+			ringbuf = ring->buffer;
+
 		/* We know the GPU must have read the request to have
 		 * sent us the seqno + interrupt, so use the position
 		 * of tail of the request to update the last known position
 		 * of the GPU head.
 		 */
-		ring->buffer->last_retired_head = request->tail;
+		ringbuf->last_retired_head = request->tail;
 
 		i915_gem_free_request(request);
 	}
@@ -2908,6 +2956,9 @@ int i915_vma_unbind(struct i915_vma *vma)
 	 * cause memory corruption through use-after-free.
 	 */
 
+	/* Throw away the active reference before moving to the unbound list */
+	i915_gem_object_retire(obj);
+
 	if (i915_is_ggtt(vma->vm)) {
 		i915_gem_object_finish_gtt(obj);
 
@@ -2922,9 +2973,8 @@ int i915_vma_unbind(struct i915_vma *vma)
 	vma->unbind_vma(vma);
 
 	list_del_init(&vma->mm_list);
-	/* Avoid an unnecessary call to unbind on rebind. */
 	if (i915_is_ggtt(vma->vm))
-		obj->map_and_fenceable = true;
+		obj->map_and_fenceable = false;
 
 	drm_mm_remove_node(&vma->node);
 	i915_gem_vma_destroy(vma);
@@ -2953,9 +3003,11 @@ int i915_gpu_idle(struct drm_device *dev)
 
 	/* Flush everything onto the inactive list. */
 	for_each_ring(ring, dev_priv, i) {
-		ret = i915_switch_context(ring, ring->default_context);
-		if (ret)
-			return ret;
+		if (!i915.enable_execlists) {
+			ret = i915_switch_context(ring, ring->default_context);
+			if (ret)
+				return ret;
+		}
 
 		ret = intel_ring_idle(ring);
 		if (ret)
@@ -3169,7 +3221,6 @@ i915_gem_object_wait_fence(struct drm_i915_gem_object *obj)
 		obj->last_fenced_seqno = 0;
 	}
 
-	obj->fenced_gpu_access = false;
 	return 0;
 }
 
@@ -3276,6 +3327,9 @@ i915_gem_object_get_fence(struct drm_i915_gem_object *obj)
 			return 0;
 		}
 	} else if (enable) {
+		if (WARN_ON(!obj->map_and_fenceable))
+			return -EINVAL;
+
 		reg = i915_find_fence_reg(dev);
 		if (IS_ERR(reg))
 			return PTR_ERR(reg);
@@ -3297,17 +3351,20 @@ i915_gem_object_get_fence(struct drm_i915_gem_object *obj)
 	return 0;
 }
 
-static bool i915_gem_valid_gtt_space(struct drm_device *dev,
-				     struct drm_mm_node *gtt_space,
+static bool i915_gem_valid_gtt_space(struct i915_vma *vma,
 				     unsigned long cache_level)
 {
+	struct drm_mm_node *gtt_space = &vma->node;
 	struct drm_mm_node *other;
 
-	/* On non-LLC machines we have to be careful when putting differing
-	 * types of snoopable memory together to avoid the prefetcher
-	 * crossing memory domains and dying.
+	/*
+	 * On some machines we have to be careful when putting differing types
+	 * of snoopable memory together to avoid the prefetcher crossing memory
+	 * domains and dying. During vm initialisation, we decide whether or not
+	 * these constraints apply and set the drm_mm.color_adjust
+	 * appropriately.
 	 */
-	if (HAS_LLC(dev))
+	if (vma->vm->mm.color_adjust == NULL)
 		return true;
 
 	if (!drm_mm_node_allocated(gtt_space))
@@ -3445,8 +3502,7 @@ search_free:
 
 		goto err_free_vma;
 	}
-	if (WARN_ON(!i915_gem_valid_gtt_space(dev, &vma->node,
-					      obj->cache_level))) {
+	if (WARN_ON(!i915_gem_valid_gtt_space(vma, obj->cache_level))) {
 		ret = -EINVAL;
 		goto err_remove_node;
 	}
@@ -3586,11 +3642,12 @@ int
 i915_gem_object_set_to_gtt_domain(struct drm_i915_gem_object *obj, bool write)
 {
 	struct drm_i915_private *dev_priv = obj->base.dev->dev_private;
+	struct i915_vma *vma = i915_gem_obj_to_ggtt(obj);
 	uint32_t old_write_domain, old_read_domains;
 	int ret;
 
 	/* Not valid to be called on unbound objects. */
-	if (!i915_gem_obj_bound_any(obj))
+	if (vma == NULL)
 		return -EINVAL;
 
 	if (obj->base.write_domain == I915_GEM_DOMAIN_GTT)
@@ -3632,13 +3689,9 @@ i915_gem_object_set_to_gtt_domain(struct drm_i915_gem_object *obj, bool write)
 					    old_write_domain);
 
 	/* And bump the LRU for this access */
-	if (i915_gem_object_is_inactive(obj)) {
-		struct i915_vma *vma = i915_gem_obj_to_ggtt(obj);
-		if (vma)
-			list_move_tail(&vma->mm_list,
-				       &dev_priv->gtt.base.inactive_list);
-
-	}
+	if (i915_gem_object_is_inactive(obj))
+		list_move_tail(&vma->mm_list,
+			       &dev_priv->gtt.base.inactive_list);
 
 	return 0;
 }
@@ -3659,7 +3712,7 @@ int i915_gem_object_set_cache_level(struct drm_i915_gem_object *obj,
 	}
 
 	list_for_each_entry_safe(vma, next, &obj->vma_list, vma_link) {
-		if (!i915_gem_valid_gtt_space(dev, &vma->node, cache_level)) {
+		if (!i915_gem_valid_gtt_space(vma, cache_level)) {
 			ret = i915_vma_unbind(vma);
 			if (ret)
 				return ret;
@@ -3802,9 +3855,6 @@ static bool is_pin_display(struct drm_i915_gem_object *obj)
 {
 	struct i915_vma *vma;
 
-	if (list_empty(&obj->vma_list))
-		return false;
-
 	vma = i915_gem_obj_to_ggtt(obj);
 	if (!vma)
 		return false;
@@ -4331,8 +4381,6 @@ void i915_gem_object_init(struct drm_i915_gem_object *obj,
 
 	obj->fence_reg = I915_FENCE_REG_NONE;
 	obj->madv = I915_MADV_WILLNEED;
-	/* Avoid an unnecessary call to unbind on the first bind. */
-	obj->map_and_fenceable = true;
 
 	i915_gem_info_add_obj(obj->base.dev->dev_private, obj->base.size);
 }
@@ -4493,12 +4541,18 @@ struct i915_vma *i915_gem_obj_to_vma(struct drm_i915_gem_object *obj,
 
 void i915_gem_vma_destroy(struct i915_vma *vma)
 {
+	struct i915_address_space *vm = NULL;
 	WARN_ON(vma->node.allocated);
 
 	/* Keep the vma as a placeholder in the execbuffer reservation lists */
 	if (!list_empty(&vma->exec_list))
 		return;
 
+	vm = vma->vm;
+
+	if (!i915_is_ggtt(vm))
+		i915_ppgtt_put(i915_vm_to_ppgtt(vm));
+
 	list_del(&vma->vma_link);
 
 	kfree(vma);
@@ -4512,7 +4566,7 @@ i915_gem_stop_ringbuffers(struct drm_device *dev)
 	int i;
 
 	for_each_ring(ring, dev_priv, i)
-		intel_stop_ring_buffer(ring);
+		dev_priv->gt.stop_ring(ring);
 }
 
 int
@@ -4629,11 +4683,46 @@ intel_enable_blt(struct drm_device *dev)
 	return true;
 }
 
-static int i915_gem_init_rings(struct drm_device *dev)
+static void init_unused_ring(struct drm_device *dev, u32 base)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	I915_WRITE(RING_CTL(base), 0);
+	I915_WRITE(RING_HEAD(base), 0);
+	I915_WRITE(RING_TAIL(base), 0);
+	I915_WRITE(RING_START(base), 0);
+}
+
+static void init_unused_rings(struct drm_device *dev)
+{
+	if (IS_I830(dev)) {
+		init_unused_ring(dev, PRB1_BASE);
+		init_unused_ring(dev, SRB0_BASE);
+		init_unused_ring(dev, SRB1_BASE);
+		init_unused_ring(dev, SRB2_BASE);
+		init_unused_ring(dev, SRB3_BASE);
+	} else if (IS_GEN2(dev)) {
+		init_unused_ring(dev, SRB0_BASE);
+		init_unused_ring(dev, SRB1_BASE);
+	} else if (IS_GEN3(dev)) {
+		init_unused_ring(dev, PRB1_BASE);
+		init_unused_ring(dev, PRB2_BASE);
+	}
+}
+
+int i915_gem_init_rings(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	int ret;
 
+	/*
+	 * 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(dev);
+
 	ret = intel_init_render_ring_buffer(dev);
 	if (ret)
 		return ret;
@@ -4712,7 +4801,7 @@ i915_gem_init_hw(struct drm_device *dev)
 
 	i915_gem_init_swizzling(dev);
 
-	ret = i915_gem_init_rings(dev);
+	ret = dev_priv->gt.init_rings(dev);
 	if (ret)
 		return ret;
 
@@ -4730,6 +4819,14 @@ i915_gem_init_hw(struct drm_device *dev)
 	if (ret && ret != -EIO) {
 		DRM_ERROR("Context enable failed %d\n", ret);
 		i915_gem_cleanup_ringbuffer(dev);
+
+		return ret;
+	}
+
+	ret = i915_ppgtt_init_hw(dev);
+	if (ret && ret != -EIO) {
+		DRM_ERROR("PPGTT enable failed %d\n", ret);
+		i915_gem_cleanup_ringbuffer(dev);
 	}
 
 	return ret;
@@ -4740,6 +4837,9 @@ int i915_gem_init(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	int ret;
 
+	i915.enable_execlists = intel_sanitize_enable_execlists(dev,
+			i915.enable_execlists);
+
 	mutex_lock(&dev->struct_mutex);
 
 	if (IS_VALLEYVIEW(dev)) {
@@ -4750,7 +4850,24 @@ int i915_gem_init(struct drm_device *dev)
 			DRM_DEBUG_DRIVER("allow wake ack timed out\n");
 	}
 
-	i915_gem_init_userptr(dev);
+	if (!i915.enable_execlists) {
+		dev_priv->gt.do_execbuf = i915_gem_ringbuffer_submission;
+		dev_priv->gt.init_rings = i915_gem_init_rings;
+		dev_priv->gt.cleanup_ring = intel_cleanup_ring_buffer;
+		dev_priv->gt.stop_ring = intel_stop_ring_buffer;
+	} else {
+		dev_priv->gt.do_execbuf = intel_execlists_submission;
+		dev_priv->gt.init_rings = intel_logical_rings_init;
+		dev_priv->gt.cleanup_ring = intel_logical_ring_cleanup;
+		dev_priv->gt.stop_ring = intel_logical_ring_stop;
+	}
+
+	ret = i915_gem_init_userptr(dev);
+	if (ret) {
+		mutex_unlock(&dev->struct_mutex);
+		return ret;
+	}
+
 	i915_gem_init_global_gtt(dev);
 
 	ret = i915_gem_context_init(dev);
@@ -4785,7 +4902,7 @@ i915_gem_cleanup_ringbuffer(struct drm_device *dev)
 	int i;
 
 	for_each_ring(ring, dev_priv, i)
-		intel_cleanup_ring_buffer(ring);
+		dev_priv->gt.cleanup_ring(ring);
 }
 
 int
@@ -5097,9 +5214,7 @@ unsigned long i915_gem_obj_offset(struct drm_i915_gem_object *o,
 	struct drm_i915_private *dev_priv = o->base.dev->dev_private;
 	struct i915_vma *vma;
 
-	if (!dev_priv->mm.aliasing_ppgtt ||
-	    vm == &dev_priv->mm.aliasing_ppgtt->base)
-		vm = &dev_priv->gtt.base;
+	WARN_ON(vm == &dev_priv->mm.aliasing_ppgtt->base);
 
 	list_for_each_entry(vma, &o->vma_list, vma_link) {
 		if (vma->vm == vm)
@@ -5140,9 +5255,7 @@ unsigned long i915_gem_obj_size(struct drm_i915_gem_object *o,
 	struct drm_i915_private *dev_priv = o->base.dev->dev_private;
 	struct i915_vma *vma;
 
-	if (!dev_priv->mm.aliasing_ppgtt ||
-	    vm == &dev_priv->mm.aliasing_ppgtt->base)
-		vm = &dev_priv->gtt.base;
+	WARN_ON(vm == &dev_priv->mm.aliasing_ppgtt->base);
 
 	BUG_ON(list_empty(&o->vma_list));
 
@@ -5165,11 +5278,16 @@ i915_gem_shrinker_scan(struct shrinker *shrinker, struct shrink_control *sc)
 	if (!i915_gem_shrinker_lock(dev, &unlock))
 		return SHRINK_STOP;
 
-	freed = i915_gem_purge(dev_priv, sc->nr_to_scan);
+	freed = i915_gem_shrink(dev_priv,
+				sc->nr_to_scan,
+				I915_SHRINK_BOUND |
+				I915_SHRINK_UNBOUND |
+				I915_SHRINK_PURGEABLE);
 	if (freed < sc->nr_to_scan)
-		freed += __i915_gem_shrink(dev_priv,
-					   sc->nr_to_scan - freed,
-					   false);
+		freed += i915_gem_shrink(dev_priv,
+					 sc->nr_to_scan - freed,
+					 I915_SHRINK_BOUND |
+					 I915_SHRINK_UNBOUND);
 	if (unlock)
 		mutex_unlock(&dev->struct_mutex);
 
@@ -5247,14 +5365,8 @@ struct i915_vma *i915_gem_obj_to_ggtt(struct drm_i915_gem_object *obj)
 {
 	struct i915_vma *vma;
 
-	/* This WARN has probably outlived its usefulness (callers already
-	 * WARN if they don't find the GGTT vma they expect). When removing,
-	 * remember to remove the pre-check in is_pin_display() as well */
-	if (WARN_ON(list_empty(&obj->vma_list)))
-		return NULL;
-
 	vma = list_first_entry(&obj->vma_list, typeof(*vma), vma_link);
-	if (vma->vm != obj_to_ggtt(obj))
+	if (vma->vm != i915_obj_to_ggtt(obj))
 		return NULL;
 
 	return vma;
diff --git a/drivers/gpu/drm/i915/i915_gem_context.c b/drivers/gpu/drm/i915/i915_gem_context.c
index 3b99390e467a..a5221d8f1580 100644
--- a/drivers/gpu/drm/i915/i915_gem_context.c
+++ b/drivers/gpu/drm/i915/i915_gem_context.c
@@ -96,50 +96,6 @@
 #define GEN6_CONTEXT_ALIGN (64<<10)
 #define GEN7_CONTEXT_ALIGN 4096
 
-static void do_ppgtt_cleanup(struct i915_hw_ppgtt *ppgtt)
-{
-	struct drm_device *dev = ppgtt->base.dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct i915_address_space *vm = &ppgtt->base;
-
-	if (ppgtt == dev_priv->mm.aliasing_ppgtt ||
-	    (list_empty(&vm->active_list) && list_empty(&vm->inactive_list))) {
-		ppgtt->base.cleanup(&ppgtt->base);
-		return;
-	}
-
-	/*
-	 * Make sure vmas are unbound before we take down the drm_mm
-	 *
-	 * FIXME: Proper refcounting should take care of this, this shouldn't be
-	 * needed at all.
-	 */
-	if (!list_empty(&vm->active_list)) {
-		struct i915_vma *vma;
-
-		list_for_each_entry(vma, &vm->active_list, mm_list)
-			if (WARN_ON(list_empty(&vma->vma_link) ||
-				    list_is_singular(&vma->vma_link)))
-				break;
-
-		i915_gem_evict_vm(&ppgtt->base, true);
-	} else {
-		i915_gem_retire_requests(dev);
-		i915_gem_evict_vm(&ppgtt->base, false);
-	}
-
-	ppgtt->base.cleanup(&ppgtt->base);
-}
-
-static void ppgtt_release(struct kref *kref)
-{
-	struct i915_hw_ppgtt *ppgtt =
-		container_of(kref, struct i915_hw_ppgtt, ref);
-
-	do_ppgtt_cleanup(ppgtt);
-	kfree(ppgtt);
-}
-
 static size_t get_context_alignment(struct drm_device *dev)
 {
 	if (IS_GEN6(dev))
@@ -179,24 +135,20 @@ static int get_context_size(struct drm_device *dev)
 void i915_gem_context_free(struct kref *ctx_ref)
 {
 	struct intel_context *ctx = container_of(ctx_ref,
-						   typeof(*ctx), ref);
-	struct i915_hw_ppgtt *ppgtt = NULL;
+						 typeof(*ctx), ref);
 
-	if (ctx->legacy_hw_ctx.rcs_state) {
-		/* We refcount even the aliasing PPGTT to keep the code symmetric */
-		if (USES_PPGTT(ctx->legacy_hw_ctx.rcs_state->base.dev))
-			ppgtt = ctx_to_ppgtt(ctx);
-	}
+	if (i915.enable_execlists)
+		intel_lr_context_free(ctx);
+
+	i915_ppgtt_put(ctx->ppgtt);
 
-	if (ppgtt)
-		kref_put(&ppgtt->ref, ppgtt_release);
 	if (ctx->legacy_hw_ctx.rcs_state)
 		drm_gem_object_unreference(&ctx->legacy_hw_ctx.rcs_state->base);
 	list_del(&ctx->link);
 	kfree(ctx);
 }
 
-static struct drm_i915_gem_object *
+struct drm_i915_gem_object *
 i915_gem_alloc_context_obj(struct drm_device *dev, size_t size)
 {
 	struct drm_i915_gem_object *obj;
@@ -226,29 +178,9 @@ i915_gem_alloc_context_obj(struct drm_device *dev, size_t size)
 	return obj;
 }
 
-static struct i915_hw_ppgtt *
-create_vm_for_ctx(struct drm_device *dev, struct intel_context *ctx)
-{
-	struct i915_hw_ppgtt *ppgtt;
-	int ret;
-
-	ppgtt = kzalloc(sizeof(*ppgtt), GFP_KERNEL);
-	if (!ppgtt)
-		return ERR_PTR(-ENOMEM);
-
-	ret = i915_gem_init_ppgtt(dev, ppgtt);
-	if (ret) {
-		kfree(ppgtt);
-		return ERR_PTR(ret);
-	}
-
-	ppgtt->ctx = ctx;
-	return ppgtt;
-}
-
 static struct intel_context *
 __create_hw_context(struct drm_device *dev,
-		  struct drm_i915_file_private *file_priv)
+		    struct drm_i915_file_private *file_priv)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_context *ctx;
@@ -301,11 +233,9 @@ err_out:
  */
 static struct intel_context *
 i915_gem_create_context(struct drm_device *dev,
-			struct drm_i915_file_private *file_priv,
-			bool create_vm)
+			struct drm_i915_file_private *file_priv)
 {
 	const bool is_global_default_ctx = file_priv == NULL;
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_context *ctx;
 	int ret = 0;
 
@@ -331,34 +261,18 @@ i915_gem_create_context(struct drm_device *dev,
 		}
 	}
 
-	if (create_vm) {
-		struct i915_hw_ppgtt *ppgtt = create_vm_for_ctx(dev, ctx);
+	if (USES_FULL_PPGTT(dev)) {
+		struct i915_hw_ppgtt *ppgtt = i915_ppgtt_create(dev, file_priv);
 
 		if (IS_ERR_OR_NULL(ppgtt)) {
 			DRM_DEBUG_DRIVER("PPGTT setup failed (%ld)\n",
 					 PTR_ERR(ppgtt));
 			ret = PTR_ERR(ppgtt);
 			goto err_unpin;
-		} else
-			ctx->vm = &ppgtt->base;
-
-		/* This case is reserved for the global default context and
-		 * should only happen once. */
-		if (is_global_default_ctx) {
-			if (WARN_ON(dev_priv->mm.aliasing_ppgtt)) {
-				ret = -EEXIST;
-				goto err_unpin;
-			}
-
-			dev_priv->mm.aliasing_ppgtt = ppgtt;
 		}
-	} else if (USES_PPGTT(dev)) {
-		/* For platforms which only have aliasing PPGTT, we fake the
-		 * address space and refcounting. */
-		ctx->vm = &dev_priv->mm.aliasing_ppgtt->base;
-		kref_get(&dev_priv->mm.aliasing_ppgtt->ref);
-	} else
-		ctx->vm = &dev_priv->gtt.base;
+
+		ctx->ppgtt = ppgtt;
+	}
 
 	return ctx;
 
@@ -375,34 +289,23 @@ void i915_gem_context_reset(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	int i;
 
-	/* Prevent the hardware from restoring the last context (which hung) on
-	 * the next switch */
+	/* In execlists mode we will unreference the context when the execlist
+	 * queue is cleared and the requests destroyed.
+	 */
+	if (i915.enable_execlists)
+		return;
+
 	for (i = 0; i < I915_NUM_RINGS; i++) {
 		struct intel_engine_cs *ring = &dev_priv->ring[i];
-		struct intel_context *dctx = ring->default_context;
 		struct intel_context *lctx = ring->last_context;
 
-		/* Do a fake switch to the default context */
-		if (lctx == dctx)
-			continue;
-
-		if (!lctx)
-			continue;
+		if (lctx) {
+			if (lctx->legacy_hw_ctx.rcs_state && i == RCS)
+				i915_gem_object_ggtt_unpin(lctx->legacy_hw_ctx.rcs_state);
 
-		if (dctx->legacy_hw_ctx.rcs_state && i == RCS) {
-			WARN_ON(i915_gem_obj_ggtt_pin(dctx->legacy_hw_ctx.rcs_state,
-						      get_context_alignment(dev), 0));
-			/* Fake a finish/inactive */
-			dctx->legacy_hw_ctx.rcs_state->base.write_domain = 0;
-			dctx->legacy_hw_ctx.rcs_state->active = 0;
+			i915_gem_context_unreference(lctx);
+			ring->last_context = NULL;
 		}
-
-		if (lctx->legacy_hw_ctx.rcs_state && i == RCS)
-			i915_gem_object_ggtt_unpin(lctx->legacy_hw_ctx.rcs_state);
-
-		i915_gem_context_unreference(lctx);
-		i915_gem_context_reference(dctx);
-		ring->last_context = dctx;
 	}
 }
 
@@ -417,7 +320,11 @@ int i915_gem_context_init(struct drm_device *dev)
 	if (WARN_ON(dev_priv->ring[RCS].default_context))
 		return 0;
 
-	if (HAS_HW_CONTEXTS(dev)) {
+	if (i915.enable_execlists) {
+		/* NB: intentionally left blank. We will allocate our own
+		 * backing objects as we need them, thank you very much */
+		dev_priv->hw_context_size = 0;
+	} else if (HAS_HW_CONTEXTS(dev)) {
 		dev_priv->hw_context_size = round_up(get_context_size(dev), 4096);
 		if (dev_priv->hw_context_size > (1<<20)) {
 			DRM_DEBUG_DRIVER("Disabling HW Contexts; invalid size %d\n",
@@ -426,18 +333,23 @@ int i915_gem_context_init(struct drm_device *dev)
 		}
 	}
 
-	ctx = i915_gem_create_context(dev, NULL, USES_PPGTT(dev));
+	ctx = i915_gem_create_context(dev, NULL);
 	if (IS_ERR(ctx)) {
 		DRM_ERROR("Failed to create default global context (error %ld)\n",
 			  PTR_ERR(ctx));
 		return PTR_ERR(ctx);
 	}
 
-	/* NB: RCS will hold a ref for all rings */
-	for (i = 0; i < I915_NUM_RINGS; i++)
-		dev_priv->ring[i].default_context = ctx;
+	for (i = 0; i < I915_NUM_RINGS; i++) {
+		struct intel_engine_cs *ring = &dev_priv->ring[i];
 
-	DRM_DEBUG_DRIVER("%s context support initialized\n", dev_priv->hw_context_size ? "HW" : "fake");
+		/* NB: RCS will hold a ref for all rings */
+		ring->default_context = ctx;
+	}
+
+	DRM_DEBUG_DRIVER("%s context support initialized\n",
+			i915.enable_execlists ? "LR" :
+			dev_priv->hw_context_size ? "HW" : "fake");
 	return 0;
 }
 
@@ -489,19 +401,11 @@ int i915_gem_context_enable(struct drm_i915_private *dev_priv)
 	struct intel_engine_cs *ring;
 	int ret, i;
 
-	/* This is the only place the aliasing PPGTT gets enabled, which means
-	 * it has to happen before we bail on reset */
-	if (dev_priv->mm.aliasing_ppgtt) {
-		struct i915_hw_ppgtt *ppgtt = dev_priv->mm.aliasing_ppgtt;
-		ppgtt->enable(ppgtt);
-	}
+	BUG_ON(!dev_priv->ring[RCS].default_context);
 
-	/* FIXME: We should make this work, even in reset */
-	if (i915_reset_in_progress(&dev_priv->gpu_error))
+	if (i915.enable_execlists)
 		return 0;
 
-	BUG_ON(!dev_priv->ring[RCS].default_context);
-
 	for_each_ring(ring, dev_priv, i) {
 		ret = i915_switch_context(ring, ring->default_context);
 		if (ret)
@@ -527,7 +431,7 @@ int i915_gem_context_open(struct drm_device *dev, struct drm_file *file)
 	idr_init(&file_priv->context_idr);
 
 	mutex_lock(&dev->struct_mutex);
-	ctx = i915_gem_create_context(dev, file_priv, USES_FULL_PPGTT(dev));
+	ctx = i915_gem_create_context(dev, file_priv);
 	mutex_unlock(&dev->struct_mutex);
 
 	if (IS_ERR(ctx)) {
@@ -563,6 +467,7 @@ mi_set_context(struct intel_engine_cs *ring,
 	       struct intel_context *new_context,
 	       u32 hw_flags)
 {
+	u32 flags = hw_flags | MI_MM_SPACE_GTT;
 	int ret;
 
 	/* w/a: If Flush TLB Invalidation Mode is enabled, driver must do a TLB
@@ -576,6 +481,10 @@ mi_set_context(struct intel_engine_cs *ring,
 			return ret;
 	}
 
+	/* These flags are for resource streamer on HSW+ */
+	if (!IS_HASWELL(ring->dev) && INTEL_INFO(ring->dev)->gen < 8)
+		flags |= (MI_SAVE_EXT_STATE_EN | MI_RESTORE_EXT_STATE_EN);
+
 	ret = intel_ring_begin(ring, 6);
 	if (ret)
 		return ret;
@@ -589,10 +498,7 @@ mi_set_context(struct intel_engine_cs *ring,
 	intel_ring_emit(ring, MI_NOOP);
 	intel_ring_emit(ring, MI_SET_CONTEXT);
 	intel_ring_emit(ring, i915_gem_obj_ggtt_offset(new_context->legacy_hw_ctx.rcs_state) |
-			MI_MM_SPACE_GTT |
-			MI_SAVE_EXT_STATE_EN |
-			MI_RESTORE_EXT_STATE_EN |
-			hw_flags);
+			flags);
 	/*
 	 * w/a: MI_SET_CONTEXT must always be followed by MI_NOOP
 	 * WaMiSetContext_Hang:snb,ivb,vlv
@@ -614,7 +520,6 @@ static int do_switch(struct intel_engine_cs *ring,
 {
 	struct drm_i915_private *dev_priv = ring->dev->dev_private;
 	struct intel_context *from = ring->last_context;
-	struct i915_hw_ppgtt *ppgtt = ctx_to_ppgtt(to);
 	u32 hw_flags = 0;
 	bool uninitialized = false;
 	int ret, i;
@@ -642,8 +547,8 @@ static int do_switch(struct intel_engine_cs *ring,
 	 */
 	from = ring->last_context;
 
-	if (USES_FULL_PPGTT(ring->dev)) {
-		ret = ppgtt->switch_mm(ppgtt, ring, false);
+	if (to->ppgtt) {
+		ret = to->ppgtt->switch_mm(to->ppgtt, ring);
 		if (ret)
 			goto unpin_out;
 	}
@@ -723,6 +628,12 @@ done:
 	ring->last_context = to;
 
 	if (uninitialized) {
+		if (ring->init_context) {
+			ret = ring->init_context(ring);
+			if (ret)
+				DRM_ERROR("ring init context: %d\n", ret);
+		}
+
 		ret = i915_gem_render_state_init(ring);
 		if (ret)
 			DRM_ERROR("init render state: %d\n", ret);
@@ -743,14 +654,19 @@ unpin_out:
  *
  * The context life cycle is simple. The context refcount is incremented and
  * decremented by 1 and create and destroy. If the context is in use by the GPU,
- * it will have a refoucnt > 1. This allows us to destroy the context abstract
+ * it will have a refcount > 1. This allows us to destroy the context abstract
  * object while letting the normal object tracking destroy the backing BO.
+ *
+ * This function should not be used in execlists mode.  Instead the context is
+ * switched by writing to the ELSP and requests keep a reference to their
+ * context.
  */
 int i915_switch_context(struct intel_engine_cs *ring,
 			struct intel_context *to)
 {
 	struct drm_i915_private *dev_priv = ring->dev->dev_private;
 
+	WARN_ON(i915.enable_execlists);
 	WARN_ON(!mutex_is_locked(&dev_priv->dev->struct_mutex));
 
 	if (to->legacy_hw_ctx.rcs_state == NULL) { /* We have the fake context */
@@ -766,9 +682,9 @@ int i915_switch_context(struct intel_engine_cs *ring,
 	return do_switch(ring, to);
 }
 
-static bool hw_context_enabled(struct drm_device *dev)
+static bool contexts_enabled(struct drm_device *dev)
 {
-	return to_i915(dev)->hw_context_size;
+	return i915.enable_execlists || to_i915(dev)->hw_context_size;
 }
 
 int i915_gem_context_create_ioctl(struct drm_device *dev, void *data,
@@ -779,14 +695,14 @@ int i915_gem_context_create_ioctl(struct drm_device *dev, void *data,
 	struct intel_context *ctx;
 	int ret;
 
-	if (!hw_context_enabled(dev))
+	if (!contexts_enabled(dev))
 		return -ENODEV;
 
 	ret = i915_mutex_lock_interruptible(dev);
 	if (ret)
 		return ret;
 
-	ctx = i915_gem_create_context(dev, file_priv, USES_FULL_PPGTT(dev));
+	ctx = i915_gem_create_context(dev, file_priv);
 	mutex_unlock(&dev->struct_mutex);
 	if (IS_ERR(ctx))
 		return PTR_ERR(ctx);
diff --git a/drivers/gpu/drm/i915/i915_gem_evict.c b/drivers/gpu/drm/i915/i915_gem_evict.c
index bbf4b12d842e..886ff2ee7a28 100644
--- a/drivers/gpu/drm/i915/i915_gem_evict.c
+++ b/drivers/gpu/drm/i915/i915_gem_evict.c
@@ -243,7 +243,7 @@ int
 i915_gem_evict_everything(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct i915_address_space *vm;
+	struct i915_address_space *vm, *v;
 	bool lists_empty = true;
 	int ret;
 
@@ -270,7 +270,7 @@ i915_gem_evict_everything(struct drm_device *dev)
 	i915_gem_retire_requests(dev);
 
 	/* Having flushed everything, unbind() should never raise an error */
-	list_for_each_entry(vm, &dev_priv->vm_list, global_link)
+	list_for_each_entry_safe(vm, v, &dev_priv->vm_list, global_link)
 		WARN_ON(i915_gem_evict_vm(vm, false));
 
 	return 0;
diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
index 60998fc4e5b2..1a0611bb576b 100644
--- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
+++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
@@ -35,6 +35,7 @@
 
 #define  __EXEC_OBJECT_HAS_PIN (1<<31)
 #define  __EXEC_OBJECT_HAS_FENCE (1<<30)
+#define  __EXEC_OBJECT_NEEDS_MAP (1<<29)
 #define  __EXEC_OBJECT_NEEDS_BIAS (1<<28)
 
 #define BATCH_OFFSET_BIAS (256*1024)
@@ -94,7 +95,6 @@ eb_lookup_vmas(struct eb_vmas *eb,
 	       struct i915_address_space *vm,
 	       struct drm_file *file)
 {
-	struct drm_i915_private *dev_priv = vm->dev->dev_private;
 	struct drm_i915_gem_object *obj;
 	struct list_head objects;
 	int i, ret;
@@ -129,20 +129,6 @@ eb_lookup_vmas(struct eb_vmas *eb,
 	i = 0;
 	while (!list_empty(&objects)) {
 		struct i915_vma *vma;
-		struct i915_address_space *bind_vm = vm;
-
-		if (exec[i].flags & EXEC_OBJECT_NEEDS_GTT &&
-		    USES_FULL_PPGTT(vm->dev)) {
-			ret = -EINVAL;
-			goto err;
-		}
-
-		/* If we have secure dispatch, or the userspace assures us that
-		 * they know what they're doing, use the GGTT VM.
-		 */
-		if (((args->flags & I915_EXEC_SECURE) &&
-		    (i == (args->buffer_count - 1))))
-			bind_vm = &dev_priv->gtt.base;
 
 		obj = list_first_entry(&objects,
 				       struct drm_i915_gem_object,
@@ -156,7 +142,7 @@ eb_lookup_vmas(struct eb_vmas *eb,
 		 * from the (obj, vm) we don't run the risk of creating
 		 * duplicated vmas for the same vm.
 		 */
-		vma = i915_gem_obj_lookup_or_create_vma(obj, bind_vm);
+		vma = i915_gem_obj_lookup_or_create_vma(obj, vm);
 		if (IS_ERR(vma)) {
 			DRM_DEBUG("Failed to lookup VMA\n");
 			ret = PTR_ERR(vma);
@@ -307,7 +293,7 @@ relocate_entry_gtt(struct drm_i915_gem_object *obj,
 	struct drm_device *dev = obj->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	uint64_t delta = reloc->delta + target_offset;
-	uint32_t __iomem *reloc_entry;
+	uint64_t offset;
 	void __iomem *reloc_page;
 	int ret;
 
@@ -320,25 +306,24 @@ relocate_entry_gtt(struct drm_i915_gem_object *obj,
 		return ret;
 
 	/* Map the page containing the relocation we're going to perform.  */
-	reloc->offset += i915_gem_obj_ggtt_offset(obj);
+	offset = i915_gem_obj_ggtt_offset(obj);
+	offset += reloc->offset;
 	reloc_page = io_mapping_map_atomic_wc(dev_priv->gtt.mappable,
-			reloc->offset & PAGE_MASK);
-	reloc_entry = (uint32_t __iomem *)
-		(reloc_page + offset_in_page(reloc->offset));
-	iowrite32(lower_32_bits(delta), reloc_entry);
+					      offset & PAGE_MASK);
+	iowrite32(lower_32_bits(delta), reloc_page + offset_in_page(offset));
 
 	if (INTEL_INFO(dev)->gen >= 8) {
-		reloc_entry += 1;
+		offset += sizeof(uint32_t);
 
-		if (offset_in_page(reloc->offset + sizeof(uint32_t)) == 0) {
+		if (offset_in_page(offset) == 0) {
 			io_mapping_unmap_atomic(reloc_page);
-			reloc_page = io_mapping_map_atomic_wc(
-					dev_priv->gtt.mappable,
-					reloc->offset + sizeof(uint32_t));
-			reloc_entry = reloc_page;
+			reloc_page =
+				io_mapping_map_atomic_wc(dev_priv->gtt.mappable,
+							 offset);
 		}
 
-		iowrite32(upper_32_bits(delta), reloc_entry);
+		iowrite32(upper_32_bits(delta),
+			  reloc_page + offset_in_page(offset));
 	}
 
 	io_mapping_unmap_atomic(reloc_page);
@@ -535,34 +520,18 @@ i915_gem_execbuffer_relocate(struct eb_vmas *eb)
 }
 
 static int
-need_reloc_mappable(struct i915_vma *vma)
-{
-	struct drm_i915_gem_exec_object2 *entry = vma->exec_entry;
-	return entry->relocation_count && !use_cpu_reloc(vma->obj) &&
-		i915_is_ggtt(vma->vm);
-}
-
-static int
 i915_gem_execbuffer_reserve_vma(struct i915_vma *vma,
 				struct intel_engine_cs *ring,
 				bool *need_reloc)
 {
 	struct drm_i915_gem_object *obj = vma->obj;
 	struct drm_i915_gem_exec_object2 *entry = vma->exec_entry;
-	bool has_fenced_gpu_access = INTEL_INFO(ring->dev)->gen < 4;
-	bool need_fence;
 	uint64_t flags;
 	int ret;
 
 	flags = 0;
-
-	need_fence =
-		has_fenced_gpu_access &&
-		entry->flags & EXEC_OBJECT_NEEDS_FENCE &&
-		obj->tiling_mode != I915_TILING_NONE;
-	if (need_fence || need_reloc_mappable(vma))
+	if (entry->flags & __EXEC_OBJECT_NEEDS_MAP)
 		flags |= PIN_MAPPABLE;
-
 	if (entry->flags & EXEC_OBJECT_NEEDS_GTT)
 		flags |= PIN_GLOBAL;
 	if (entry->flags & __EXEC_OBJECT_NEEDS_BIAS)
@@ -574,17 +543,13 @@ i915_gem_execbuffer_reserve_vma(struct i915_vma *vma,
 
 	entry->flags |= __EXEC_OBJECT_HAS_PIN;
 
-	if (has_fenced_gpu_access) {
-		if (entry->flags & EXEC_OBJECT_NEEDS_FENCE) {
-			ret = i915_gem_object_get_fence(obj);
-			if (ret)
-				return ret;
-
-			if (i915_gem_object_pin_fence(obj))
-				entry->flags |= __EXEC_OBJECT_HAS_FENCE;
+	if (entry->flags & EXEC_OBJECT_NEEDS_FENCE) {
+		ret = i915_gem_object_get_fence(obj);
+		if (ret)
+			return ret;
 
-			obj->pending_fenced_gpu_access = true;
-		}
+		if (i915_gem_object_pin_fence(obj))
+			entry->flags |= __EXEC_OBJECT_HAS_FENCE;
 	}
 
 	if (entry->offset != vma->node.start) {
@@ -601,26 +566,40 @@ i915_gem_execbuffer_reserve_vma(struct i915_vma *vma,
 }
 
 static bool
-eb_vma_misplaced(struct i915_vma *vma, bool has_fenced_gpu_access)
+need_reloc_mappable(struct i915_vma *vma)
 {
 	struct drm_i915_gem_exec_object2 *entry = vma->exec_entry;
-	struct drm_i915_gem_object *obj = vma->obj;
-	bool need_fence, need_mappable;
 
-	need_fence =
-		has_fenced_gpu_access &&
-		entry->flags & EXEC_OBJECT_NEEDS_FENCE &&
-		obj->tiling_mode != I915_TILING_NONE;
-	need_mappable = need_fence || need_reloc_mappable(vma);
+	if (entry->relocation_count == 0)
+		return false;
+
+	if (!i915_is_ggtt(vma->vm))
+		return false;
+
+	/* See also use_cpu_reloc() */
+	if (HAS_LLC(vma->obj->base.dev))
+		return false;
 
-	WARN_ON((need_mappable || need_fence) &&
+	if (vma->obj->base.write_domain == I915_GEM_DOMAIN_CPU)
+		return false;
+
+	return true;
+}
+
+static bool
+eb_vma_misplaced(struct i915_vma *vma)
+{
+	struct drm_i915_gem_exec_object2 *entry = vma->exec_entry;
+	struct drm_i915_gem_object *obj = vma->obj;
+
+	WARN_ON(entry->flags & __EXEC_OBJECT_NEEDS_MAP &&
 	       !i915_is_ggtt(vma->vm));
 
 	if (entry->alignment &&
 	    vma->node.start & (entry->alignment - 1))
 		return true;
 
-	if (need_mappable && !obj->map_and_fenceable)
+	if (entry->flags & __EXEC_OBJECT_NEEDS_MAP && !obj->map_and_fenceable)
 		return true;
 
 	if (entry->flags & __EXEC_OBJECT_NEEDS_BIAS &&
@@ -642,9 +621,6 @@ i915_gem_execbuffer_reserve(struct intel_engine_cs *ring,
 	bool has_fenced_gpu_access = INTEL_INFO(ring->dev)->gen < 4;
 	int retry;
 
-	if (list_empty(vmas))
-		return 0;
-
 	i915_gem_retire_requests_ring(ring);
 
 	vm = list_first_entry(vmas, struct i915_vma, exec_list)->vm;
@@ -658,20 +634,21 @@ i915_gem_execbuffer_reserve(struct intel_engine_cs *ring,
 		obj = vma->obj;
 		entry = vma->exec_entry;
 
+		if (!has_fenced_gpu_access)
+			entry->flags &= ~EXEC_OBJECT_NEEDS_FENCE;
 		need_fence =
-			has_fenced_gpu_access &&
 			entry->flags & EXEC_OBJECT_NEEDS_FENCE &&
 			obj->tiling_mode != I915_TILING_NONE;
 		need_mappable = need_fence || need_reloc_mappable(vma);
 
-		if (need_mappable)
+		if (need_mappable) {
+			entry->flags |= __EXEC_OBJECT_NEEDS_MAP;
 			list_move(&vma->exec_list, &ordered_vmas);
-		else
+		} else
 			list_move_tail(&vma->exec_list, &ordered_vmas);
 
 		obj->base.pending_read_domains = I915_GEM_GPU_DOMAINS & ~I915_GEM_DOMAIN_COMMAND;
 		obj->base.pending_write_domain = 0;
-		obj->pending_fenced_gpu_access = false;
 	}
 	list_splice(&ordered_vmas, vmas);
 
@@ -696,7 +673,7 @@ i915_gem_execbuffer_reserve(struct intel_engine_cs *ring,
 			if (!drm_mm_node_allocated(&vma->node))
 				continue;
 
-			if (eb_vma_misplaced(vma, has_fenced_gpu_access))
+			if (eb_vma_misplaced(vma))
 				ret = i915_vma_unbind(vma);
 			else
 				ret = i915_gem_execbuffer_reserve_vma(vma, ring, need_relocs);
@@ -744,9 +721,6 @@ i915_gem_execbuffer_relocate_slow(struct drm_device *dev,
 	int i, total, ret;
 	unsigned count = args->buffer_count;
 
-	if (WARN_ON(list_empty(&eb->vmas)))
-		return 0;
-
 	vm = list_first_entry(&eb->vmas, struct i915_vma, exec_list)->vm;
 
 	/* We may process another execbuffer during the unlock... */
@@ -890,18 +864,24 @@ i915_gem_check_execbuffer(struct drm_i915_gem_execbuffer2 *exec)
 }
 
 static int
-validate_exec_list(struct drm_i915_gem_exec_object2 *exec,
+validate_exec_list(struct drm_device *dev,
+		   struct drm_i915_gem_exec_object2 *exec,
 		   int count)
 {
-	int i;
 	unsigned relocs_total = 0;
 	unsigned relocs_max = UINT_MAX / sizeof(struct drm_i915_gem_relocation_entry);
+	unsigned invalid_flags;
+	int i;
+
+	invalid_flags = __EXEC_OBJECT_UNKNOWN_FLAGS;
+	if (USES_FULL_PPGTT(dev))
+		invalid_flags |= EXEC_OBJECT_NEEDS_GTT;
 
 	for (i = 0; i < count; i++) {
 		char __user *ptr = to_user_ptr(exec[i].relocs_ptr);
 		int length; /* limited by fault_in_pages_readable() */
 
-		if (exec[i].flags & __EXEC_OBJECT_UNKNOWN_FLAGS)
+		if (exec[i].flags & invalid_flags)
 			return -EINVAL;
 
 		/* First check for malicious input causing overflow in
@@ -951,16 +931,26 @@ i915_gem_validate_context(struct drm_device *dev, struct drm_file *file,
 		return ERR_PTR(-EIO);
 	}
 
+	if (i915.enable_execlists && !ctx->engine[ring->id].state) {
+		int ret = intel_lr_context_deferred_create(ctx, ring);
+		if (ret) {
+			DRM_DEBUG("Could not create LRC %u: %d\n", ctx_id, ret);
+			return ERR_PTR(ret);
+		}
+	}
+
 	return ctx;
 }
 
-static void
+void
 i915_gem_execbuffer_move_to_active(struct list_head *vmas,
 				   struct intel_engine_cs *ring)
 {
+	u32 seqno = intel_ring_get_seqno(ring);
 	struct i915_vma *vma;
 
 	list_for_each_entry(vma, vmas, exec_list) {
+		struct drm_i915_gem_exec_object2 *entry = vma->exec_entry;
 		struct drm_i915_gem_object *obj = vma->obj;
 		u32 old_read = obj->base.read_domains;
 		u32 old_write = obj->base.write_domain;
@@ -969,24 +959,31 @@ i915_gem_execbuffer_move_to_active(struct list_head *vmas,
 		if (obj->base.write_domain == 0)
 			obj->base.pending_read_domains |= obj->base.read_domains;
 		obj->base.read_domains = obj->base.pending_read_domains;
-		obj->fenced_gpu_access = obj->pending_fenced_gpu_access;
 
 		i915_vma_move_to_active(vma, ring);
 		if (obj->base.write_domain) {
 			obj->dirty = 1;
-			obj->last_write_seqno = intel_ring_get_seqno(ring);
+			obj->last_write_seqno = seqno;
 
 			intel_fb_obj_invalidate(obj, ring);
 
 			/* update for the implicit flush after a batch */
 			obj->base.write_domain &= ~I915_GEM_GPU_DOMAINS;
 		}
+		if (entry->flags & EXEC_OBJECT_NEEDS_FENCE) {
+			obj->last_fenced_seqno = seqno;
+			if (entry->flags & __EXEC_OBJECT_HAS_FENCE) {
+				struct drm_i915_private *dev_priv = to_i915(ring->dev);
+				list_move_tail(&dev_priv->fence_regs[obj->fence_reg].lru_list,
+					       &dev_priv->mm.fence_list);
+			}
+		}
 
 		trace_i915_gem_object_change_domain(obj, old_read, old_write);
 	}
 }
 
-static void
+void
 i915_gem_execbuffer_retire_commands(struct drm_device *dev,
 				    struct drm_file *file,
 				    struct intel_engine_cs *ring,
@@ -1026,14 +1023,14 @@ i915_reset_gen7_sol_offsets(struct drm_device *dev,
 	return 0;
 }
 
-static int
-legacy_ringbuffer_submission(struct drm_device *dev, struct drm_file *file,
-			     struct intel_engine_cs *ring,
-			     struct intel_context *ctx,
-			     struct drm_i915_gem_execbuffer2 *args,
-			     struct list_head *vmas,
-			     struct drm_i915_gem_object *batch_obj,
-			     u64 exec_start, u32 flags)
+int
+i915_gem_ringbuffer_submission(struct drm_device *dev, struct drm_file *file,
+			       struct intel_engine_cs *ring,
+			       struct intel_context *ctx,
+			       struct drm_i915_gem_execbuffer2 *args,
+			       struct list_head *vmas,
+			       struct drm_i915_gem_object *batch_obj,
+			       u64 exec_start, u32 flags)
 {
 	struct drm_clip_rect *cliprects = NULL;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -1254,7 +1251,7 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
 	if (!i915_gem_check_execbuffer(args))
 		return -EINVAL;
 
-	ret = validate_exec_list(exec, args->buffer_count);
+	ret = validate_exec_list(dev, exec, args->buffer_count);
 	if (ret)
 		return ret;
 
@@ -1318,8 +1315,9 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
 
 	i915_gem_context_reference(ctx);
 
-	vm = ctx->vm;
-	if (!USES_FULL_PPGTT(dev))
+	if (ctx->ppgtt)
+		vm = &ctx->ppgtt->base;
+	else
 		vm = &dev_priv->gtt.base;
 
 	eb = eb_create(args);
@@ -1386,25 +1384,36 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
 	/* 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.
 	 * hsw should have this fixed, but bdw mucks it up again. */
-	if (flags & I915_DISPATCH_SECURE &&
-	    !batch_obj->has_global_gtt_mapping) {
-		/* When we have multiple VMs, we'll need to make sure that we
-		 * allocate space first */
-		struct i915_vma *vma = i915_gem_obj_to_ggtt(batch_obj);
-		BUG_ON(!vma);
-		vma->bind_vma(vma, batch_obj->cache_level, GLOBAL_BIND);
-	}
+	if (flags & I915_DISPATCH_SECURE) {
+		/*
+		 * So on first glance it looks freaky that we pin the batch here
+		 * outside of the reservation loop. But:
+		 * - The batch is already pinned into the relevant ppgtt, so we
+		 *   already have the backing storage fully allocated.
+		 * - No other BO uses the global gtt (well contexts, but meh),
+		 *   so we don't really have issues with mutliple objects not
+		 *   fitting due to fragmentation.
+		 * So this is actually safe.
+		 */
+		ret = i915_gem_obj_ggtt_pin(batch_obj, 0, 0);
+		if (ret)
+			goto err;
 
-	if (flags & I915_DISPATCH_SECURE)
 		exec_start += i915_gem_obj_ggtt_offset(batch_obj);
-	else
+	} else
 		exec_start += i915_gem_obj_offset(batch_obj, vm);
 
-	ret = legacy_ringbuffer_submission(dev, file, ring, ctx,
-			args, &eb->vmas, batch_obj, exec_start, flags);
-	if (ret)
-		goto err;
+	ret = dev_priv->gt.do_execbuf(dev, file, ring, ctx, args,
+				      &eb->vmas, batch_obj, exec_start, flags);
 
+	/*
+	 * FIXME: We crucially rely upon the active tracking for the (ppgtt)
+	 * batch vma for correctness. For less ugly and less fragility this
+	 * needs to be adjusted to also track the ggtt batch vma properly as
+	 * active.
+	 */
+	if (flags & I915_DISPATCH_SECURE)
+		i915_gem_object_ggtt_unpin(batch_obj);
 err:
 	/* the request owns the ref now */
 	i915_gem_context_unreference(ctx);
diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c
index e42925f76b4b..b672b843fd5e 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.c
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.c
@@ -33,17 +33,6 @@
 static void bdw_setup_private_ppat(struct drm_i915_private *dev_priv);
 static void chv_setup_private_ppat(struct drm_i915_private *dev_priv);
 
-bool intel_enable_ppgtt(struct drm_device *dev, bool full)
-{
-	if (i915.enable_ppgtt == 0)
-		return false;
-
-	if (i915.enable_ppgtt == 1 && full)
-		return false;
-
-	return true;
-}
-
 static int sanitize_enable_ppgtt(struct drm_device *dev, int enable_ppgtt)
 {
 	if (enable_ppgtt == 0 || !HAS_ALIASING_PPGTT(dev))
@@ -78,7 +67,6 @@ static void ppgtt_bind_vma(struct i915_vma *vma,
 			   enum i915_cache_level cache_level,
 			   u32 flags);
 static void ppgtt_unbind_vma(struct i915_vma *vma);
-static int gen8_ppgtt_enable(struct i915_hw_ppgtt *ppgtt);
 
 static inline gen8_gtt_pte_t gen8_pte_encode(dma_addr_t addr,
 					     enum i915_cache_level level,
@@ -216,19 +204,12 @@ static gen6_gtt_pte_t iris_pte_encode(dma_addr_t addr,
 
 /* Broadwell Page Directory Pointer Descriptors */
 static int gen8_write_pdp(struct intel_engine_cs *ring, unsigned entry,
-			   uint64_t val, bool synchronous)
+			   uint64_t val)
 {
-	struct drm_i915_private *dev_priv = ring->dev->dev_private;
 	int ret;
 
 	BUG_ON(entry >= 4);
 
-	if (synchronous) {
-		I915_WRITE(GEN8_RING_PDP_UDW(ring, entry), val >> 32);
-		I915_WRITE(GEN8_RING_PDP_LDW(ring, entry), (u32)val);
-		return 0;
-	}
-
 	ret = intel_ring_begin(ring, 6);
 	if (ret)
 		return ret;
@@ -245,8 +226,7 @@ static int gen8_write_pdp(struct intel_engine_cs *ring, unsigned entry,
 }
 
 static int gen8_mm_switch(struct i915_hw_ppgtt *ppgtt,
-			  struct intel_engine_cs *ring,
-			  bool synchronous)
+			  struct intel_engine_cs *ring)
 {
 	int i, ret;
 
@@ -255,7 +235,7 @@ static int gen8_mm_switch(struct i915_hw_ppgtt *ppgtt,
 
 	for (i = used_pd - 1; i >= 0; i--) {
 		dma_addr_t addr = ppgtt->pd_dma_addr[i];
-		ret = gen8_write_pdp(ring, i, addr, synchronous);
+		ret = gen8_write_pdp(ring, i, addr);
 		if (ret)
 			return ret;
 	}
@@ -403,9 +383,6 @@ static void gen8_ppgtt_cleanup(struct i915_address_space *vm)
 	struct i915_hw_ppgtt *ppgtt =
 		container_of(vm, struct i915_hw_ppgtt, base);
 
-	list_del(&vm->global_link);
-	drm_mm_takedown(&vm->mm);
-
 	gen8_ppgtt_unmap_pages(ppgtt);
 	gen8_ppgtt_free(ppgtt);
 }
@@ -615,7 +592,6 @@ static int gen8_ppgtt_init(struct i915_hw_ppgtt *ppgtt, uint64_t size)
 		kunmap_atomic(pd_vaddr);
 	}
 
-	ppgtt->enable = gen8_ppgtt_enable;
 	ppgtt->switch_mm = gen8_mm_switch;
 	ppgtt->base.clear_range = gen8_ppgtt_clear_range;
 	ppgtt->base.insert_entries = gen8_ppgtt_insert_entries;
@@ -724,29 +700,10 @@ static uint32_t get_pd_offset(struct i915_hw_ppgtt *ppgtt)
 }
 
 static int hsw_mm_switch(struct i915_hw_ppgtt *ppgtt,
-			 struct intel_engine_cs *ring,
-			 bool synchronous)
+			 struct intel_engine_cs *ring)
 {
-	struct drm_device *dev = ppgtt->base.dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	int ret;
 
-	/* If we're in reset, we can assume the GPU is sufficiently idle to
-	 * manually frob these bits. Ideally we could use the ring functions,
-	 * except our error handling makes it quite difficult (can't use
-	 * intel_ring_begin, ring->flush, or intel_ring_advance)
-	 *
-	 * FIXME: We should try not to special case reset
-	 */
-	if (synchronous ||
-	    i915_reset_in_progress(&dev_priv->gpu_error)) {
-		WARN_ON(ppgtt != dev_priv->mm.aliasing_ppgtt);
-		I915_WRITE(RING_PP_DIR_DCLV(ring), PP_DIR_DCLV_2G);
-		I915_WRITE(RING_PP_DIR_BASE(ring), get_pd_offset(ppgtt));
-		POSTING_READ(RING_PP_DIR_BASE(ring));
-		return 0;
-	}
-
 	/* NB: TLBs must be flushed and invalidated before a switch */
 	ret = ring->flush(ring, I915_GEM_GPU_DOMAINS, I915_GEM_GPU_DOMAINS);
 	if (ret)
@@ -768,29 +725,10 @@ static int hsw_mm_switch(struct i915_hw_ppgtt *ppgtt,
 }
 
 static int gen7_mm_switch(struct i915_hw_ppgtt *ppgtt,
-			  struct intel_engine_cs *ring,
-			  bool synchronous)
+			  struct intel_engine_cs *ring)
 {
-	struct drm_device *dev = ppgtt->base.dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	int ret;
 
-	/* If we're in reset, we can assume the GPU is sufficiently idle to
-	 * manually frob these bits. Ideally we could use the ring functions,
-	 * except our error handling makes it quite difficult (can't use
-	 * intel_ring_begin, ring->flush, or intel_ring_advance)
-	 *
-	 * FIXME: We should try not to special case reset
-	 */
-	if (synchronous ||
-	    i915_reset_in_progress(&dev_priv->gpu_error)) {
-		WARN_ON(ppgtt != dev_priv->mm.aliasing_ppgtt);
-		I915_WRITE(RING_PP_DIR_DCLV(ring), PP_DIR_DCLV_2G);
-		I915_WRITE(RING_PP_DIR_BASE(ring), get_pd_offset(ppgtt));
-		POSTING_READ(RING_PP_DIR_BASE(ring));
-		return 0;
-	}
-
 	/* NB: TLBs must be flushed and invalidated before a switch */
 	ret = ring->flush(ring, I915_GEM_GPU_DOMAINS, I915_GEM_GPU_DOMAINS);
 	if (ret)
@@ -819,14 +757,11 @@ static int gen7_mm_switch(struct i915_hw_ppgtt *ppgtt,
 }
 
 static int gen6_mm_switch(struct i915_hw_ppgtt *ppgtt,
-			  struct intel_engine_cs *ring,
-			  bool synchronous)
+			  struct intel_engine_cs *ring)
 {
 	struct drm_device *dev = ppgtt->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	if (!synchronous)
-		return 0;
 
 	I915_WRITE(RING_PP_DIR_DCLV(ring), PP_DIR_DCLV_2G);
 	I915_WRITE(RING_PP_DIR_BASE(ring), get_pd_offset(ppgtt));
@@ -836,39 +771,20 @@ static int gen6_mm_switch(struct i915_hw_ppgtt *ppgtt,
 	return 0;
 }
 
-static int gen8_ppgtt_enable(struct i915_hw_ppgtt *ppgtt)
+static void gen8_ppgtt_enable(struct drm_device *dev)
 {
-	struct drm_device *dev = ppgtt->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_engine_cs *ring;
-	int j, ret;
+	int j;
 
 	for_each_ring(ring, dev_priv, j) {
 		I915_WRITE(RING_MODE_GEN7(ring),
 			   _MASKED_BIT_ENABLE(GFX_PPGTT_ENABLE));
-
-		/* We promise to do a switch later with FULL PPGTT. If this is
-		 * aliasing, this is the one and only switch we'll do */
-		if (USES_FULL_PPGTT(dev))
-			continue;
-
-		ret = ppgtt->switch_mm(ppgtt, ring, true);
-		if (ret)
-			goto err_out;
 	}
-
-	return 0;
-
-err_out:
-	for_each_ring(ring, dev_priv, j)
-		I915_WRITE(RING_MODE_GEN7(ring),
-			   _MASKED_BIT_DISABLE(GFX_PPGTT_ENABLE));
-	return ret;
 }
 
-static int gen7_ppgtt_enable(struct i915_hw_ppgtt *ppgtt)
+static void gen7_ppgtt_enable(struct drm_device *dev)
 {
-	struct drm_device *dev = ppgtt->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_engine_cs *ring;
 	uint32_t ecochk, ecobits;
@@ -887,31 +803,16 @@ static int gen7_ppgtt_enable(struct i915_hw_ppgtt *ppgtt)
 	I915_WRITE(GAM_ECOCHK, ecochk);
 
 	for_each_ring(ring, dev_priv, i) {
-		int ret;
 		/* GFX_MODE is per-ring on gen7+ */
 		I915_WRITE(RING_MODE_GEN7(ring),
 			   _MASKED_BIT_ENABLE(GFX_PPGTT_ENABLE));
-
-		/* We promise to do a switch later with FULL PPGTT. If this is
-		 * aliasing, this is the one and only switch we'll do */
-		if (USES_FULL_PPGTT(dev))
-			continue;
-
-		ret = ppgtt->switch_mm(ppgtt, ring, true);
-		if (ret)
-			return ret;
 	}
-
-	return 0;
 }
 
-static int gen6_ppgtt_enable(struct i915_hw_ppgtt *ppgtt)
+static void gen6_ppgtt_enable(struct drm_device *dev)
 {
-	struct drm_device *dev = ppgtt->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_engine_cs *ring;
 	uint32_t ecochk, gab_ctl, ecobits;
-	int i;
 
 	ecobits = I915_READ(GAC_ECO_BITS);
 	I915_WRITE(GAC_ECO_BITS, ecobits | ECOBITS_SNB_BIT |
@@ -924,14 +825,6 @@ static int gen6_ppgtt_enable(struct i915_hw_ppgtt *ppgtt)
 	I915_WRITE(GAM_ECOCHK, ecochk | ECOCHK_SNB_BIT | ECOCHK_PPGTT_CACHE64B);
 
 	I915_WRITE(GFX_MODE, _MASKED_BIT_ENABLE(GFX_PPGTT_ENABLE));
-
-	for_each_ring(ring, dev_priv, i) {
-		int ret = ppgtt->switch_mm(ppgtt, ring, true);
-		if (ret)
-			return ret;
-	}
-
-	return 0;
 }
 
 /* PPGTT support for Sandybdrige/Gen6 and later */
@@ -1029,8 +922,6 @@ static void gen6_ppgtt_cleanup(struct i915_address_space *vm)
 	struct i915_hw_ppgtt *ppgtt =
 		container_of(vm, struct i915_hw_ppgtt, base);
 
-	list_del(&vm->global_link);
-	drm_mm_takedown(&ppgtt->base.mm);
 	drm_mm_remove_node(&ppgtt->node);
 
 	gen6_ppgtt_unmap_pages(ppgtt);
@@ -1151,13 +1042,10 @@ static int gen6_ppgtt_init(struct i915_hw_ppgtt *ppgtt)
 
 	ppgtt->base.pte_encode = dev_priv->gtt.base.pte_encode;
 	if (IS_GEN6(dev)) {
-		ppgtt->enable = gen6_ppgtt_enable;
 		ppgtt->switch_mm = gen6_mm_switch;
 	} else if (IS_HASWELL(dev)) {
-		ppgtt->enable = gen7_ppgtt_enable;
 		ppgtt->switch_mm = hsw_mm_switch;
 	} else if (IS_GEN7(dev)) {
-		ppgtt->enable = gen7_ppgtt_enable;
 		ppgtt->switch_mm = gen7_mm_switch;
 	} else
 		BUG();
@@ -1188,39 +1076,114 @@ static int gen6_ppgtt_init(struct i915_hw_ppgtt *ppgtt)
 			 ppgtt->node.size >> 20,
 			 ppgtt->node.start / PAGE_SIZE);
 
+	gen6_write_pdes(ppgtt);
+	DRM_DEBUG("Adding PPGTT at offset %x\n",
+		  ppgtt->pd_offset << 10);
+
 	return 0;
 }
 
-int i915_gem_init_ppgtt(struct drm_device *dev, struct i915_hw_ppgtt *ppgtt)
+static int __hw_ppgtt_init(struct drm_device *dev, struct i915_hw_ppgtt *ppgtt)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	int ret = 0;
 
 	ppgtt->base.dev = dev;
 	ppgtt->base.scratch = dev_priv->gtt.base.scratch;
 
 	if (INTEL_INFO(dev)->gen < 8)
-		ret = gen6_ppgtt_init(ppgtt);
+		return gen6_ppgtt_init(ppgtt);
 	else if (IS_GEN8(dev))
-		ret = gen8_ppgtt_init(ppgtt, dev_priv->gtt.base.total);
+		return gen8_ppgtt_init(ppgtt, dev_priv->gtt.base.total);
 	else
 		BUG();
+}
+int i915_ppgtt_init(struct drm_device *dev, struct i915_hw_ppgtt *ppgtt)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	int ret = 0;
 
-	if (!ret) {
-		struct drm_i915_private *dev_priv = dev->dev_private;
+	ret = __hw_ppgtt_init(dev, ppgtt);
+	if (ret == 0) {
 		kref_init(&ppgtt->ref);
 		drm_mm_init(&ppgtt->base.mm, ppgtt->base.start,
 			    ppgtt->base.total);
 		i915_init_vm(dev_priv, &ppgtt->base);
-		if (INTEL_INFO(dev)->gen < 8) {
-			gen6_write_pdes(ppgtt);
-			DRM_DEBUG("Adding PPGTT at offset %x\n",
-				  ppgtt->pd_offset << 10);
+	}
+
+	return ret;
+}
+
+int i915_ppgtt_init_hw(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring;
+	struct i915_hw_ppgtt *ppgtt = dev_priv->mm.aliasing_ppgtt;
+	int i, ret = 0;
+
+	/* In the case of execlists, PPGTT is enabled by the context descriptor
+	 * and the PDPs are contained within the context itself.  We don't
+	 * need to do anything here. */
+	if (i915.enable_execlists)
+		return 0;
+
+	if (!USES_PPGTT(dev))
+		return 0;
+
+	if (IS_GEN6(dev))
+		gen6_ppgtt_enable(dev);
+	else if (IS_GEN7(dev))
+		gen7_ppgtt_enable(dev);
+	else if (INTEL_INFO(dev)->gen >= 8)
+		gen8_ppgtt_enable(dev);
+	else
+		WARN_ON(1);
+
+	if (ppgtt) {
+		for_each_ring(ring, dev_priv, i) {
+			ret = ppgtt->switch_mm(ppgtt, ring);
+			if (ret != 0)
+				return ret;
 		}
 	}
 
 	return ret;
 }
+struct i915_hw_ppgtt *
+i915_ppgtt_create(struct drm_device *dev, struct drm_i915_file_private *fpriv)
+{
+	struct i915_hw_ppgtt *ppgtt;
+	int ret;
+
+	ppgtt = kzalloc(sizeof(*ppgtt), GFP_KERNEL);
+	if (!ppgtt)
+		return ERR_PTR(-ENOMEM);
+
+	ret = i915_ppgtt_init(dev, ppgtt);
+	if (ret) {
+		kfree(ppgtt);
+		return ERR_PTR(ret);
+	}
+
+	ppgtt->file_priv = fpriv;
+
+	return ppgtt;
+}
+
+void  i915_ppgtt_release(struct kref *kref)
+{
+	struct i915_hw_ppgtt *ppgtt =
+		container_of(kref, struct i915_hw_ppgtt, ref);
+
+	/* vmas should already be unbound */
+	WARN_ON(!list_empty(&ppgtt->base.active_list));
+	WARN_ON(!list_empty(&ppgtt->base.inactive_list));
+
+	list_del(&ppgtt->base.global_link);
+	drm_mm_takedown(&ppgtt->base.mm);
+
+	ppgtt->base.cleanup(&ppgtt->base);
+	kfree(ppgtt);
+}
 
 static void
 ppgtt_bind_vma(struct i915_vma *vma,
@@ -1687,10 +1650,10 @@ static void i915_gtt_color_adjust(struct drm_mm_node *node,
 	}
 }
 
-void i915_gem_setup_global_gtt(struct drm_device *dev,
-			       unsigned long start,
-			       unsigned long mappable_end,
-			       unsigned long end)
+int i915_gem_setup_global_gtt(struct drm_device *dev,
+			      unsigned long start,
+			      unsigned long mappable_end,
+			      unsigned long end)
 {
 	/* Let GEM Manage all of the aperture.
 	 *
@@ -1706,6 +1669,7 @@ void i915_gem_setup_global_gtt(struct drm_device *dev,
 	struct drm_mm_node *entry;
 	struct drm_i915_gem_object *obj;
 	unsigned long hole_start, hole_end;
+	int ret;
 
 	BUG_ON(mappable_end > end);
 
@@ -1717,14 +1681,16 @@ void i915_gem_setup_global_gtt(struct drm_device *dev,
 	/* Mark any preallocated objects as occupied */
 	list_for_each_entry(obj, &dev_priv->mm.bound_list, global_list) {
 		struct i915_vma *vma = i915_gem_obj_to_vma(obj, ggtt_vm);
-		int ret;
+
 		DRM_DEBUG_KMS("reserving preallocated space: %lx + %zx\n",
 			      i915_gem_obj_ggtt_offset(obj), obj->base.size);
 
 		WARN_ON(i915_gem_obj_ggtt_bound(obj));
 		ret = drm_mm_reserve_node(&ggtt_vm->mm, &vma->node);
-		if (ret)
-			DRM_DEBUG_KMS("Reservation failed\n");
+		if (ret) {
+			DRM_DEBUG_KMS("Reservation failed: %i\n", ret);
+			return ret;
+		}
 		obj->has_global_gtt_mapping = 1;
 	}
 
@@ -1741,6 +1707,22 @@ void i915_gem_setup_global_gtt(struct drm_device *dev,
 
 	/* And finally clear the reserved guard page */
 	ggtt_vm->clear_range(ggtt_vm, end - PAGE_SIZE, PAGE_SIZE, true);
+
+	if (USES_PPGTT(dev) && !USES_FULL_PPGTT(dev)) {
+		struct i915_hw_ppgtt *ppgtt;
+
+		ppgtt = kzalloc(sizeof(*ppgtt), GFP_KERNEL);
+		if (!ppgtt)
+			return -ENOMEM;
+
+		ret = __hw_ppgtt_init(dev, ppgtt);
+		if (ret != 0)
+			return ret;
+
+		dev_priv->mm.aliasing_ppgtt = ppgtt;
+	}
+
+	return 0;
 }
 
 void i915_gem_init_global_gtt(struct drm_device *dev)
@@ -1754,6 +1736,25 @@ void i915_gem_init_global_gtt(struct drm_device *dev)
 	i915_gem_setup_global_gtt(dev, 0, mappable_size, gtt_size);
 }
 
+void i915_global_gtt_cleanup(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct i915_address_space *vm = &dev_priv->gtt.base;
+
+	if (dev_priv->mm.aliasing_ppgtt) {
+		struct i915_hw_ppgtt *ppgtt = dev_priv->mm.aliasing_ppgtt;
+
+		ppgtt->base.cleanup(&ppgtt->base);
+	}
+
+	if (drm_mm_initialized(&vm->mm)) {
+		drm_mm_takedown(&vm->mm);
+		list_del(&vm->global_link);
+	}
+
+	vm->cleanup(vm);
+}
+
 static int setup_scratch_page(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -2022,10 +2023,6 @@ static void gen6_gmch_remove(struct i915_address_space *vm)
 
 	struct i915_gtt *gtt = container_of(vm, struct i915_gtt, base);
 
-	if (drm_mm_initialized(&vm->mm)) {
-		drm_mm_takedown(&vm->mm);
-		list_del(&vm->global_link);
-	}
 	iounmap(gtt->gsm);
 	teardown_scratch_page(vm->dev);
 }
@@ -2058,10 +2055,6 @@ static int i915_gmch_probe(struct drm_device *dev,
 
 static void i915_gmch_remove(struct i915_address_space *vm)
 {
-	if (drm_mm_initialized(&vm->mm)) {
-		drm_mm_takedown(&vm->mm);
-		list_del(&vm->global_link);
-	}
 	intel_gmch_remove();
 }
 
@@ -2160,8 +2153,10 @@ static struct i915_vma *__i915_gem_vma_create(struct drm_i915_gem_object *obj,
 	/* Keep GGTT vmas first to make debug easier */
 	if (i915_is_ggtt(vm))
 		list_add(&vma->vma_link, &obj->vma_list);
-	else
+	else {
 		list_add_tail(&vma->vma_link, &obj->vma_list);
+		i915_ppgtt_get(i915_vm_to_ppgtt(vm));
+	}
 
 	return vma;
 }
diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.h b/drivers/gpu/drm/i915/i915_gem_gtt.h
index 8d6f7c18c404..d5c14af51e99 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.h
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.h
@@ -34,6 +34,8 @@
 #ifndef __I915_GEM_GTT_H__
 #define __I915_GEM_GTT_H__
 
+struct drm_i915_file_private;
+
 typedef uint32_t gen6_gtt_pte_t;
 typedef uint64_t gen8_gtt_pte_t;
 typedef gen8_gtt_pte_t gen8_ppgtt_pde_t;
@@ -258,22 +260,36 @@ struct i915_hw_ppgtt {
 		dma_addr_t *gen8_pt_dma_addr[4];
 	};
 
-	struct intel_context *ctx;
+	struct drm_i915_file_private *file_priv;
 
 	int (*enable)(struct i915_hw_ppgtt *ppgtt);
 	int (*switch_mm)(struct i915_hw_ppgtt *ppgtt,
-			 struct intel_engine_cs *ring,
-			 bool synchronous);
+			 struct intel_engine_cs *ring);
 	void (*debug_dump)(struct i915_hw_ppgtt *ppgtt, struct seq_file *m);
 };
 
 int i915_gem_gtt_init(struct drm_device *dev);
 void i915_gem_init_global_gtt(struct drm_device *dev);
-void i915_gem_setup_global_gtt(struct drm_device *dev, unsigned long start,
-			       unsigned long mappable_end, unsigned long end);
-
-bool intel_enable_ppgtt(struct drm_device *dev, bool full);
-int i915_gem_init_ppgtt(struct drm_device *dev, struct i915_hw_ppgtt *ppgtt);
+int i915_gem_setup_global_gtt(struct drm_device *dev, unsigned long start,
+			      unsigned long mappable_end, unsigned long end);
+void i915_global_gtt_cleanup(struct drm_device *dev);
+
+
+int i915_ppgtt_init(struct drm_device *dev, struct i915_hw_ppgtt *ppgtt);
+int i915_ppgtt_init_hw(struct drm_device *dev);
+void i915_ppgtt_release(struct kref *kref);
+struct i915_hw_ppgtt *i915_ppgtt_create(struct drm_device *dev,
+					struct drm_i915_file_private *fpriv);
+static inline void i915_ppgtt_get(struct i915_hw_ppgtt *ppgtt)
+{
+	if (ppgtt)
+		kref_get(&ppgtt->ref);
+}
+static inline void i915_ppgtt_put(struct i915_hw_ppgtt *ppgtt)
+{
+	if (ppgtt)
+		kref_put(&ppgtt->ref, i915_ppgtt_release);
+}
 
 void i915_check_and_clear_faults(struct drm_device *dev);
 void i915_gem_suspend_gtt_mappings(struct drm_device *dev);
diff --git a/drivers/gpu/drm/i915/i915_gem_render_state.c b/drivers/gpu/drm/i915/i915_gem_render_state.c
index e60be3f552a6..a9a62d75aa57 100644
--- a/drivers/gpu/drm/i915/i915_gem_render_state.c
+++ b/drivers/gpu/drm/i915/i915_gem_render_state.c
@@ -28,13 +28,6 @@
 #include "i915_drv.h"
 #include "intel_renderstate.h"
 
-struct render_state {
-	const struct intel_renderstate_rodata *rodata;
-	struct drm_i915_gem_object *obj;
-	u64 ggtt_offset;
-	int gen;
-};
-
 static const struct intel_renderstate_rodata *
 render_state_get_rodata(struct drm_device *dev, const int gen)
 {
@@ -127,30 +120,47 @@ static int render_state_setup(struct render_state *so)
 	return 0;
 }
 
-static void render_state_fini(struct render_state *so)
+void i915_gem_render_state_fini(struct render_state *so)
 {
 	i915_gem_object_ggtt_unpin(so->obj);
 	drm_gem_object_unreference(&so->obj->base);
 }
 
-int i915_gem_render_state_init(struct intel_engine_cs *ring)
+int i915_gem_render_state_prepare(struct intel_engine_cs *ring,
+				  struct render_state *so)
 {
-	struct render_state so;
 	int ret;
 
 	if (WARN_ON(ring->id != RCS))
 		return -ENOENT;
 
-	ret = render_state_init(&so, ring->dev);
+	ret = render_state_init(so, ring->dev);
 	if (ret)
 		return ret;
 
-	if (so.rodata == NULL)
+	if (so->rodata == NULL)
 		return 0;
 
-	ret = render_state_setup(&so);
+	ret = render_state_setup(so);
+	if (ret) {
+		i915_gem_render_state_fini(so);
+		return ret;
+	}
+
+	return 0;
+}
+
+int i915_gem_render_state_init(struct intel_engine_cs *ring)
+{
+	struct render_state so;
+	int ret;
+
+	ret = i915_gem_render_state_prepare(ring, &so);
 	if (ret)
-		goto out;
+		return ret;
+
+	if (so.rodata == NULL)
+		return 0;
 
 	ret = ring->dispatch_execbuffer(ring,
 					so.ggtt_offset,
@@ -164,6 +174,6 @@ int i915_gem_render_state_init(struct intel_engine_cs *ring)
 	ret = __i915_add_request(ring, NULL, so.obj, NULL);
 	/* __i915_add_request moves object to inactive if it fails */
 out:
-	render_state_fini(&so);
+	i915_gem_render_state_fini(&so);
 	return ret;
 }
diff --git a/drivers/gpu/drm/i915/i915_gem_render_state.h b/drivers/gpu/drm/i915/i915_gem_render_state.h
new file mode 100644
index 000000000000..c44961ed3fad
--- /dev/null
+++ b/drivers/gpu/drm/i915/i915_gem_render_state.h
@@ -0,0 +1,47 @@
+/*
+ * Copyright © 2014 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.
+ */
+
+#ifndef _I915_GEM_RENDER_STATE_H_
+#define _I915_GEM_RENDER_STATE_H_
+
+#include <linux/types.h>
+
+struct intel_renderstate_rodata {
+	const u32 *reloc;
+	const u32 *batch;
+	const u32 batch_items;
+};
+
+struct render_state {
+	const struct intel_renderstate_rodata *rodata;
+	struct drm_i915_gem_object *obj;
+	u64 ggtt_offset;
+	int gen;
+};
+
+int i915_gem_render_state_init(struct intel_engine_cs *ring);
+void i915_gem_render_state_fini(struct render_state *so);
+int i915_gem_render_state_prepare(struct intel_engine_cs *ring,
+				  struct render_state *so);
+
+#endif /* _I915_GEM_RENDER_STATE_H_ */
diff --git a/drivers/gpu/drm/i915/i915_gem_stolen.c b/drivers/gpu/drm/i915/i915_gem_stolen.c
index 21c025a209c0..85fda6b803e4 100644
--- a/drivers/gpu/drm/i915/i915_gem_stolen.c
+++ b/drivers/gpu/drm/i915/i915_gem_stolen.c
@@ -289,6 +289,7 @@ void i915_gem_cleanup_stolen(struct drm_device *dev)
 int i915_gem_init_stolen(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 tmp;
 	int bios_reserved = 0;
 
 #ifdef CONFIG_INTEL_IOMMU
@@ -308,8 +309,16 @@ int i915_gem_init_stolen(struct drm_device *dev)
 	DRM_DEBUG_KMS("found %zd bytes of stolen memory at %08lx\n",
 		      dev_priv->gtt.stolen_size, dev_priv->mm.stolen_base);
 
-	if (IS_VALLEYVIEW(dev))
-		bios_reserved = 1024*1024; /* top 1M on VLV/BYT */
+	if (INTEL_INFO(dev)->gen >= 8) {
+		tmp = I915_READ(GEN7_BIOS_RESERVED);
+		tmp >>= GEN8_BIOS_RESERVED_SHIFT;
+		tmp &= GEN8_BIOS_RESERVED_MASK;
+		bios_reserved = (1024*1024) << tmp;
+	} else if (IS_GEN7(dev)) {
+		tmp = I915_READ(GEN7_BIOS_RESERVED);
+		bios_reserved = tmp & GEN7_BIOS_RESERVED_256K ?
+			256*1024 : 1024*1024;
+	}
 
 	if (WARN_ON(bios_reserved > dev_priv->gtt.stolen_size))
 		return 0;
diff --git a/drivers/gpu/drm/i915/i915_gem_tiling.c b/drivers/gpu/drm/i915/i915_gem_tiling.c
index cb150e8b4336..2cefb597df6d 100644
--- a/drivers/gpu/drm/i915/i915_gem_tiling.c
+++ b/drivers/gpu/drm/i915/i915_gem_tiling.c
@@ -91,7 +91,14 @@ i915_gem_detect_bit_6_swizzle(struct drm_device *dev)
 	uint32_t swizzle_x = I915_BIT_6_SWIZZLE_UNKNOWN;
 	uint32_t swizzle_y = I915_BIT_6_SWIZZLE_UNKNOWN;
 
-	if (IS_VALLEYVIEW(dev)) {
+	if (INTEL_INFO(dev)->gen >= 8 || IS_VALLEYVIEW(dev)) {
+		/*
+		 * On BDW+, swizzling is not used. We leave the CPU memory
+		 * controller in charge of optimizing memory accesses without
+		 * the extra address manipulation GPU side.
+		 *
+		 * VLV and CHV don't have GPU swizzling.
+		 */
 		swizzle_x = I915_BIT_6_SWIZZLE_NONE;
 		swizzle_y = I915_BIT_6_SWIZZLE_NONE;
 	} else if (INTEL_INFO(dev)->gen >= 6) {
@@ -376,7 +383,7 @@ i915_gem_set_tiling(struct drm_device *dev, void *data,
 
 		if (ret == 0) {
 			obj->fence_dirty =
-				obj->fenced_gpu_access ||
+				obj->last_fenced_seqno ||
 				obj->fence_reg != I915_FENCE_REG_NONE;
 
 			obj->tiling_mode = args->tiling_mode;
diff --git a/drivers/gpu/drm/i915/i915_gem_userptr.c b/drivers/gpu/drm/i915/i915_gem_userptr.c
index d38413997379..d182058383a9 100644
--- a/drivers/gpu/drm/i915/i915_gem_userptr.c
+++ b/drivers/gpu/drm/i915/i915_gem_userptr.c
@@ -293,15 +293,23 @@ i915_gem_userptr_release__mmu_notifier(struct drm_i915_gem_object *obj)
 static struct i915_mmu_notifier *
 i915_mmu_notifier_find(struct i915_mm_struct *mm)
 {
-	if (mm->mn == NULL) {
-		down_write(&mm->mm->mmap_sem);
-		mutex_lock(&to_i915(mm->dev)->mm_lock);
-		if (mm->mn == NULL)
-			mm->mn = i915_mmu_notifier_create(mm->mm);
-		mutex_unlock(&to_i915(mm->dev)->mm_lock);
-		up_write(&mm->mm->mmap_sem);
+	struct i915_mmu_notifier *mn = mm->mn;
+
+	mn = mm->mn;
+	if (mn)
+		return mn;
+
+	down_write(&mm->mm->mmap_sem);
+	mutex_lock(&to_i915(mm->dev)->mm_lock);
+	if ((mn = mm->mn) == NULL) {
+		mn = i915_mmu_notifier_create(mm->mm);
+		if (!IS_ERR(mn))
+			mm->mn = mn;
 	}
-	return mm->mn;
+	mutex_unlock(&to_i915(mm->dev)->mm_lock);
+	up_write(&mm->mm->mmap_sem);
+
+	return mn;
 }
 
 static int
@@ -681,16 +689,15 @@ i915_gem_userptr_get_pages(struct drm_i915_gem_object *obj)
 static void
 i915_gem_userptr_put_pages(struct drm_i915_gem_object *obj)
 {
-	struct scatterlist *sg;
-	int i;
+	struct sg_page_iter sg_iter;
 
 	BUG_ON(obj->userptr.work != NULL);
 
 	if (obj->madv != I915_MADV_WILLNEED)
 		obj->dirty = 0;
 
-	for_each_sg(obj->pages->sgl, sg, obj->pages->nents, i) {
-		struct page *page = sg_page(sg);
+	for_each_sg_page(obj->pages->sgl, &sg_iter, obj->pages->nents, 0) {
+		struct page *page = sg_page_iter_page(&sg_iter);
 
 		if (obj->dirty)
 			set_page_dirty(page);
diff --git a/drivers/gpu/drm/i915/i915_gpu_error.c b/drivers/gpu/drm/i915/i915_gpu_error.c
index eab41f9390f8..2c87a797213f 100644
--- a/drivers/gpu/drm/i915/i915_gpu_error.c
+++ b/drivers/gpu/drm/i915/i915_gpu_error.c
@@ -192,10 +192,10 @@ static void print_error_buffers(struct drm_i915_error_state_buf *m,
 				struct drm_i915_error_buffer *err,
 				int count)
 {
-	err_printf(m, "%s [%d]:\n", name, count);
+	err_printf(m, "  %s [%d]:\n", name, count);
 
 	while (count--) {
-		err_printf(m, "  %08x %8u %02x %02x %x %x",
+		err_printf(m, "    %08x %8u %02x %02x %x %x",
 			   err->gtt_offset,
 			   err->size,
 			   err->read_domains,
@@ -208,7 +208,7 @@ static void print_error_buffers(struct drm_i915_error_state_buf *m,
 		err_puts(m, err->userptr ? " userptr" : "");
 		err_puts(m, err->ring != -1 ? " " : "");
 		err_puts(m, ring_str(err->ring));
-		err_puts(m, i915_cache_level_str(err->cache_level));
+		err_puts(m, i915_cache_level_str(m->i915, err->cache_level));
 
 		if (err->name)
 			err_printf(m, " (name: %d)", err->name);
@@ -393,15 +393,17 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m,
 		i915_ring_error_state(m, dev, &error->ring[i]);
 	}
 
-	if (error->active_bo)
+	for (i = 0; i < error->vm_count; i++) {
+		err_printf(m, "vm[%d]\n", i);
+
 		print_error_buffers(m, "Active",
-				    error->active_bo[0],
-				    error->active_bo_count[0]);
+				    error->active_bo[i],
+				    error->active_bo_count[i]);
 
-	if (error->pinned_bo)
 		print_error_buffers(m, "Pinned",
-				    error->pinned_bo[0],
-				    error->pinned_bo_count[0]);
+				    error->pinned_bo[i],
+				    error->pinned_bo_count[i]);
+	}
 
 	for (i = 0; i < ARRAY_SIZE(error->ring); i++) {
 		obj = error->ring[i].batchbuffer;
@@ -492,9 +494,11 @@ out:
 }
 
 int i915_error_state_buf_init(struct drm_i915_error_state_buf *ebuf,
+			      struct drm_i915_private *i915,
 			      size_t count, loff_t pos)
 {
 	memset(ebuf, 0, sizeof(*ebuf));
+	ebuf->i915 = i915;
 
 	/* We need to have enough room to store any i915_error_state printf
 	 * so that we can move it to start position.
@@ -556,24 +560,54 @@ static void i915_error_state_free(struct kref *error_ref)
 }
 
 static struct drm_i915_error_object *
-i915_error_object_create_sized(struct drm_i915_private *dev_priv,
-			       struct drm_i915_gem_object *src,
-			       struct i915_address_space *vm,
-			       const int num_pages)
+i915_error_object_create(struct drm_i915_private *dev_priv,
+			 struct drm_i915_gem_object *src,
+			 struct i915_address_space *vm)
 {
 	struct drm_i915_error_object *dst;
-	int i;
+	int num_pages;
+	bool use_ggtt;
+	int i = 0;
 	u32 reloc_offset;
 
 	if (src == NULL || src->pages == NULL)
 		return NULL;
 
+	num_pages = src->base.size >> PAGE_SHIFT;
+
 	dst = kmalloc(sizeof(*dst) + num_pages * sizeof(u32 *), GFP_ATOMIC);
 	if (dst == NULL)
 		return NULL;
 
-	reloc_offset = dst->gtt_offset = i915_gem_obj_offset(src, vm);
-	for (i = 0; i < num_pages; i++) {
+	if (i915_gem_obj_bound(src, vm))
+		dst->gtt_offset = i915_gem_obj_offset(src, vm);
+	else
+		dst->gtt_offset = -1;
+
+	reloc_offset = dst->gtt_offset;
+	use_ggtt = (src->cache_level == I915_CACHE_NONE &&
+		    i915_is_ggtt(vm) &&
+		    src->has_global_gtt_mapping &&
+		    reloc_offset + num_pages * PAGE_SIZE <= dev_priv->gtt.mappable_end);
+
+	/* Cannot access stolen address directly, try to use the aperture */
+	if (src->stolen) {
+		use_ggtt = true;
+
+		if (!src->has_global_gtt_mapping)
+			goto unwind;
+
+		reloc_offset = i915_gem_obj_ggtt_offset(src);
+		if (reloc_offset + num_pages * PAGE_SIZE > dev_priv->gtt.mappable_end)
+			goto unwind;
+	}
+
+	/* Cannot access snooped pages through the aperture */
+	if (use_ggtt && src->cache_level != I915_CACHE_NONE && !HAS_LLC(dev_priv->dev))
+		goto unwind;
+
+	dst->page_count = num_pages;
+	while (num_pages--) {
 		unsigned long flags;
 		void *d;
 
@@ -582,10 +616,7 @@ i915_error_object_create_sized(struct drm_i915_private *dev_priv,
 			goto unwind;
 
 		local_irq_save(flags);
-		if (src->cache_level == I915_CACHE_NONE &&
-		    reloc_offset < dev_priv->gtt.mappable_end &&
-		    src->has_global_gtt_mapping &&
-		    i915_is_ggtt(vm)) {
+		if (use_ggtt) {
 			void __iomem *s;
 
 			/* Simply ignore tiling or any overlapping fence.
@@ -597,14 +628,6 @@ i915_error_object_create_sized(struct drm_i915_private *dev_priv,
 						     reloc_offset);
 			memcpy_fromio(d, s, PAGE_SIZE);
 			io_mapping_unmap_atomic(s);
-		} else if (src->stolen) {
-			unsigned long offset;
-
-			offset = dev_priv->mm.stolen_base;
-			offset += src->stolen->start;
-			offset += i << PAGE_SHIFT;
-
-			memcpy_fromio(d, (void __iomem *) offset, PAGE_SIZE);
 		} else {
 			struct page *page;
 			void *s;
@@ -621,11 +644,9 @@ i915_error_object_create_sized(struct drm_i915_private *dev_priv,
 		}
 		local_irq_restore(flags);
 
-		dst->pages[i] = d;
-
+		dst->pages[i++] = d;
 		reloc_offset += PAGE_SIZE;
 	}
-	dst->page_count = num_pages;
 
 	return dst;
 
@@ -635,22 +656,19 @@ unwind:
 	kfree(dst);
 	return NULL;
 }
-#define i915_error_object_create(dev_priv, src, vm) \
-	i915_error_object_create_sized((dev_priv), (src), (vm), \
-				       (src)->base.size>>PAGE_SHIFT)
-
 #define i915_error_ggtt_object_create(dev_priv, src) \
-	i915_error_object_create_sized((dev_priv), (src), &(dev_priv)->gtt.base, \
-				       (src)->base.size>>PAGE_SHIFT)
+	i915_error_object_create((dev_priv), (src), &(dev_priv)->gtt.base)
 
 static void capture_bo(struct drm_i915_error_buffer *err,
-		       struct drm_i915_gem_object *obj)
+		       struct i915_vma *vma)
 {
+	struct drm_i915_gem_object *obj = vma->obj;
+
 	err->size = obj->base.size;
 	err->name = obj->base.name;
 	err->rseqno = obj->last_read_seqno;
 	err->wseqno = obj->last_write_seqno;
-	err->gtt_offset = i915_gem_obj_ggtt_offset(obj);
+	err->gtt_offset = vma->node.start;
 	err->read_domains = obj->base.read_domains;
 	err->write_domain = obj->base.write_domain;
 	err->fence_reg = obj->fence_reg;
@@ -674,7 +692,7 @@ static u32 capture_active_bo(struct drm_i915_error_buffer *err,
 	int i = 0;
 
 	list_for_each_entry(vma, head, mm_list) {
-		capture_bo(err++, vma->obj);
+		capture_bo(err++, vma);
 		if (++i == count)
 			break;
 	}
@@ -683,21 +701,27 @@ static u32 capture_active_bo(struct drm_i915_error_buffer *err,
 }
 
 static u32 capture_pinned_bo(struct drm_i915_error_buffer *err,
-			     int count, struct list_head *head)
+			     int count, struct list_head *head,
+			     struct i915_address_space *vm)
 {
 	struct drm_i915_gem_object *obj;
-	int i = 0;
+	struct drm_i915_error_buffer * const first = err;
+	struct drm_i915_error_buffer * const last = err + count;
 
 	list_for_each_entry(obj, head, global_list) {
-		if (!i915_gem_obj_is_pinned(obj))
-			continue;
+		struct i915_vma *vma;
 
-		capture_bo(err++, obj);
-		if (++i == count)
+		if (err == last)
 			break;
+
+		list_for_each_entry(vma, &obj->vma_list, vma_link)
+			if (vma->vm == vm && vma->pin_count > 0) {
+				capture_bo(err++, vma);
+				break;
+			}
 	}
 
-	return i;
+	return err - first;
 }
 
 /* Generate a semi-unique error code. The code is not meant to have meaning, The
@@ -890,9 +914,6 @@ static void i915_record_ring_state(struct drm_device *dev,
 		ering->hws = I915_READ(mmio);
 	}
 
-	ering->cpu_ring_head = ring->buffer->head;
-	ering->cpu_ring_tail = ring->buffer->tail;
-
 	ering->hangcheck_score = ring->hangcheck.score;
 	ering->hangcheck_action = ring->hangcheck.action;
 
@@ -955,6 +976,7 @@ static void i915_gem_record_rings(struct drm_device *dev,
 
 	for (i = 0; i < I915_NUM_RINGS; i++) {
 		struct intel_engine_cs *ring = &dev_priv->ring[i];
+		struct intel_ringbuffer *rbuf;
 
 		error->ring[i].pid = -1;
 
@@ -967,6 +989,12 @@ static void i915_gem_record_rings(struct drm_device *dev,
 
 		request = i915_gem_find_active_request(ring);
 		if (request) {
+			struct i915_address_space *vm;
+
+			vm = request->ctx && request->ctx->ppgtt ?
+				&request->ctx->ppgtt->base :
+				&dev_priv->gtt.base;
+
 			/* We need to copy these to an anonymous buffer
 			 * as the simplest method to avoid being overwritten
 			 * by userspace.
@@ -974,12 +1002,9 @@ static void i915_gem_record_rings(struct drm_device *dev,
 			error->ring[i].batchbuffer =
 				i915_error_object_create(dev_priv,
 							 request->batch_obj,
-							 request->ctx ?
-							 request->ctx->vm :
-							 &dev_priv->gtt.base);
+							 vm);
 
-			if (HAS_BROKEN_CS_TLB(dev_priv->dev) &&
-			    ring->scratch.obj)
+			if (HAS_BROKEN_CS_TLB(dev_priv->dev))
 				error->ring[i].wa_batchbuffer =
 					i915_error_ggtt_object_create(dev_priv,
 							     ring->scratch.obj);
@@ -998,12 +1023,27 @@ static void i915_gem_record_rings(struct drm_device *dev,
 			}
 		}
 
+		if (i915.enable_execlists) {
+			/* TODO: This is only a small fix to keep basic error
+			 * capture working, but we need to add more information
+			 * for it to be useful (e.g. dump the context being
+			 * executed).
+			 */
+			if (request)
+				rbuf = request->ctx->engine[ring->id].ringbuf;
+			else
+				rbuf = ring->default_context->engine[ring->id].ringbuf;
+		} else
+			rbuf = ring->buffer;
+
+		error->ring[i].cpu_ring_head = rbuf->head;
+		error->ring[i].cpu_ring_tail = rbuf->tail;
+
 		error->ring[i].ringbuffer =
-			i915_error_ggtt_object_create(dev_priv, ring->buffer->obj);
+			i915_error_ggtt_object_create(dev_priv, rbuf->obj);
 
-		if (ring->status_page.obj)
-			error->ring[i].hws_page =
-				i915_error_ggtt_object_create(dev_priv, ring->status_page.obj);
+		error->ring[i].hws_page =
+			i915_error_ggtt_object_create(dev_priv, ring->status_page.obj);
 
 		i915_gem_record_active_context(ring, error, &error->ring[i]);
 
@@ -1049,9 +1089,14 @@ static void i915_gem_capture_vm(struct drm_i915_private *dev_priv,
 	list_for_each_entry(vma, &vm->active_list, mm_list)
 		i++;
 	error->active_bo_count[ndx] = i;
-	list_for_each_entry(obj, &dev_priv->mm.bound_list, global_list)
-		if (i915_gem_obj_is_pinned(obj))
-			i++;
+
+	list_for_each_entry(obj, &dev_priv->mm.bound_list, global_list) {
+		list_for_each_entry(vma, &obj->vma_list, vma_link)
+			if (vma->vm == vm && vma->pin_count > 0) {
+				i++;
+				break;
+			}
+	}
 	error->pinned_bo_count[ndx] = i - error->active_bo_count[ndx];
 
 	if (i) {
@@ -1070,7 +1115,7 @@ static void i915_gem_capture_vm(struct drm_i915_private *dev_priv,
 		error->pinned_bo_count[ndx] =
 			capture_pinned_bo(pinned_bo,
 					  error->pinned_bo_count[ndx],
-					  &dev_priv->mm.bound_list);
+					  &dev_priv->mm.bound_list, vm);
 	error->active_bo[ndx] = active_bo;
 	error->pinned_bo[ndx] = pinned_bo;
 }
@@ -1091,8 +1136,25 @@ static void i915_gem_capture_buffers(struct drm_i915_private *dev_priv,
 	error->pinned_bo_count = kcalloc(cnt, sizeof(*error->pinned_bo_count),
 					 GFP_ATOMIC);
 
-	list_for_each_entry(vm, &dev_priv->vm_list, global_link)
-		i915_gem_capture_vm(dev_priv, error, vm, i++);
+	if (error->active_bo == NULL ||
+	    error->pinned_bo == NULL ||
+	    error->active_bo_count == NULL ||
+	    error->pinned_bo_count == NULL) {
+		kfree(error->active_bo);
+		kfree(error->active_bo_count);
+		kfree(error->pinned_bo);
+		kfree(error->pinned_bo_count);
+
+		error->active_bo = NULL;
+		error->active_bo_count = NULL;
+		error->pinned_bo = NULL;
+		error->pinned_bo_count = NULL;
+	} else {
+		list_for_each_entry(vm, &dev_priv->vm_list, global_link)
+			i915_gem_capture_vm(dev_priv, error, vm, i++);
+
+		error->vm_count = cnt;
+	}
 }
 
 /* Capture all registers which don't fit into another category. */
@@ -1295,11 +1357,11 @@ void i915_destroy_error_state(struct drm_device *dev)
 		kref_put(&error->ref, i915_error_state_free);
 }
 
-const char *i915_cache_level_str(int type)
+const char *i915_cache_level_str(struct drm_i915_private *i915, int type)
 {
 	switch (type) {
 	case I915_CACHE_NONE: return " uncached";
-	case I915_CACHE_LLC: return " snooped or LLC";
+	case I915_CACHE_LLC: return HAS_LLC(i915) ? " LLC" : " snooped";
 	case I915_CACHE_L3_LLC: return " L3+LLC";
 	case I915_CACHE_WT: return " WT";
 	default: return "";
diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c
index 0050ee9470f1..3201986bf25e 100644
--- a/drivers/gpu/drm/i915/i915_irq.c
+++ b/drivers/gpu/drm/i915/i915_irq.c
@@ -151,7 +151,7 @@ ironlake_disable_display_irq(struct drm_i915_private *dev_priv, u32 mask)
 {
 	assert_spin_locked(&dev_priv->irq_lock);
 
-	if (!intel_irqs_enabled(dev_priv))
+	if (WARN_ON(!intel_irqs_enabled(dev_priv)))
 		return;
 
 	if ((dev_priv->irq_mask & mask) != mask) {
@@ -238,7 +238,7 @@ static bool ivb_can_enable_err_int(struct drm_device *dev)
 
 	assert_spin_locked(&dev_priv->irq_lock);
 
-	for_each_pipe(pipe) {
+	for_each_pipe(dev_priv, pipe) {
 		crtc = to_intel_crtc(dev_priv->pipe_to_crtc_mapping[pipe]);
 
 		if (crtc->cpu_fifo_underrun_disabled)
@@ -296,7 +296,7 @@ static bool cpt_can_enable_serr_int(struct drm_device *dev)
 
 	assert_spin_locked(&dev_priv->irq_lock);
 
-	for_each_pipe(pipe) {
+	for_each_pipe(dev_priv, pipe) {
 		crtc = to_intel_crtc(dev_priv->pipe_to_crtc_mapping[pipe]);
 
 		if (crtc->pch_fifo_underrun_disabled)
@@ -497,7 +497,7 @@ static bool __intel_set_cpu_fifo_underrun_reporting(struct drm_device *dev,
 	old = !intel_crtc->cpu_fifo_underrun_disabled;
 	intel_crtc->cpu_fifo_underrun_disabled = !enable;
 
-	if (INTEL_INFO(dev)->gen < 5 || IS_VALLEYVIEW(dev))
+	if (HAS_GMCH_DISPLAY(dev))
 		i9xx_set_fifo_underrun_reporting(dev, pipe, enable, old);
 	else if (IS_GEN5(dev) || IS_GEN6(dev))
 		ironlake_set_fifo_underrun_reporting(dev, pipe, enable);
@@ -1020,7 +1020,7 @@ static int i915_get_crtc_scanoutpos(struct drm_device *dev, int pipe,
 
 	/* In vblank? */
 	if (in_vbl)
-		ret |= DRM_SCANOUTPOS_INVBL;
+		ret |= DRM_SCANOUTPOS_IN_VBLANK;
 
 	return ret;
 }
@@ -1322,10 +1322,10 @@ static u32 vlv_c0_residency(struct drm_i915_private *dev_priv,
  * @dev_priv: DRM device private
  *
  */
-static u32 vlv_calc_delay_from_C0_counters(struct drm_i915_private *dev_priv)
+static int vlv_calc_delay_from_C0_counters(struct drm_i915_private *dev_priv)
 {
 	u32 residency_C0_up = 0, residency_C0_down = 0;
-	u8 new_delay, adj;
+	int new_delay, adj;
 
 	dev_priv->rps.ei_interrupt_count++;
 
@@ -1627,6 +1627,7 @@ static irqreturn_t gen8_gt_irq_handler(struct drm_device *dev,
 				       struct drm_i915_private *dev_priv,
 				       u32 master_ctl)
 {
+	struct intel_engine_cs *ring;
 	u32 rcs, bcs, vcs;
 	uint32_t tmp = 0;
 	irqreturn_t ret = IRQ_NONE;
@@ -1636,12 +1637,20 @@ static irqreturn_t gen8_gt_irq_handler(struct drm_device *dev,
 		if (tmp) {
 			I915_WRITE(GEN8_GT_IIR(0), tmp);
 			ret = IRQ_HANDLED;
+
 			rcs = tmp >> GEN8_RCS_IRQ_SHIFT;
-			bcs = tmp >> GEN8_BCS_IRQ_SHIFT;
+			ring = &dev_priv->ring[RCS];
 			if (rcs & GT_RENDER_USER_INTERRUPT)
-				notify_ring(dev, &dev_priv->ring[RCS]);
+				notify_ring(dev, ring);
+			if (rcs & GT_CONTEXT_SWITCH_INTERRUPT)
+				intel_execlists_handle_ctx_events(ring);
+
+			bcs = tmp >> GEN8_BCS_IRQ_SHIFT;
+			ring = &dev_priv->ring[BCS];
 			if (bcs & GT_RENDER_USER_INTERRUPT)
-				notify_ring(dev, &dev_priv->ring[BCS]);
+				notify_ring(dev, ring);
+			if (bcs & GT_CONTEXT_SWITCH_INTERRUPT)
+				intel_execlists_handle_ctx_events(ring);
 		} else
 			DRM_ERROR("The master control interrupt lied (GT0)!\n");
 	}
@@ -1651,12 +1660,20 @@ static irqreturn_t gen8_gt_irq_handler(struct drm_device *dev,
 		if (tmp) {
 			I915_WRITE(GEN8_GT_IIR(1), tmp);
 			ret = IRQ_HANDLED;
+
 			vcs = tmp >> GEN8_VCS1_IRQ_SHIFT;
+			ring = &dev_priv->ring[VCS];
 			if (vcs & GT_RENDER_USER_INTERRUPT)
-				notify_ring(dev, &dev_priv->ring[VCS]);
+				notify_ring(dev, ring);
+			if (vcs & GT_CONTEXT_SWITCH_INTERRUPT)
+				intel_execlists_handle_ctx_events(ring);
+
 			vcs = tmp >> GEN8_VCS2_IRQ_SHIFT;
+			ring = &dev_priv->ring[VCS2];
 			if (vcs & GT_RENDER_USER_INTERRUPT)
-				notify_ring(dev, &dev_priv->ring[VCS2]);
+				notify_ring(dev, ring);
+			if (vcs & GT_CONTEXT_SWITCH_INTERRUPT)
+				intel_execlists_handle_ctx_events(ring);
 		} else
 			DRM_ERROR("The master control interrupt lied (GT1)!\n");
 	}
@@ -1677,9 +1694,13 @@ static irqreturn_t gen8_gt_irq_handler(struct drm_device *dev,
 		if (tmp) {
 			I915_WRITE(GEN8_GT_IIR(3), tmp);
 			ret = IRQ_HANDLED;
+
 			vcs = tmp >> GEN8_VECS_IRQ_SHIFT;
+			ring = &dev_priv->ring[VECS];
 			if (vcs & GT_RENDER_USER_INTERRUPT)
-				notify_ring(dev, &dev_priv->ring[VECS]);
+				notify_ring(dev, ring);
+			if (vcs & GT_CONTEXT_SWITCH_INTERRUPT)
+				intel_execlists_handle_ctx_events(ring);
 		} else
 			DRM_ERROR("The master control interrupt lied (GT3)!\n");
 	}
@@ -1772,7 +1793,9 @@ static inline void intel_hpd_irq_handler(struct drm_device *dev,
 				long_hpd = (dig_hotplug_reg >> dig_shift) & PORTB_HOTPLUG_LONG_DETECT;
 			}
 
-			DRM_DEBUG_DRIVER("digital hpd port %d %d\n", port, long_hpd);
+			DRM_DEBUG_DRIVER("digital hpd port %c - %s\n",
+					 port_name(port),
+					 long_hpd ? "long" : "short");
 			/* for long HPD pulses we want to have the digital queue happen,
 			   but we still want HPD storm detection to function. */
 			if (long_hpd) {
@@ -1984,14 +2007,9 @@ static void gen6_rps_irq_handler(struct drm_i915_private *dev_priv, u32 pm_iir)
 
 static bool intel_pipe_handle_vblank(struct drm_device *dev, enum pipe pipe)
 {
-	struct intel_crtc *crtc;
-
 	if (!drm_handle_vblank(dev, pipe))
 		return false;
 
-	crtc = to_intel_crtc(intel_get_crtc_for_pipe(dev, pipe));
-	wake_up(&crtc->vbl_wait);
-
 	return true;
 }
 
@@ -2002,7 +2020,7 @@ static void valleyview_pipestat_irq_handler(struct drm_device *dev, u32 iir)
 	int pipe;
 
 	spin_lock(&dev_priv->irq_lock);
-	for_each_pipe(pipe) {
+	for_each_pipe(dev_priv, pipe) {
 		int reg;
 		u32 mask, iir_bit = 0;
 
@@ -2047,9 +2065,10 @@ static void valleyview_pipestat_irq_handler(struct drm_device *dev, u32 iir)
 	}
 	spin_unlock(&dev_priv->irq_lock);
 
-	for_each_pipe(pipe) {
-		if (pipe_stats[pipe] & PIPE_START_VBLANK_INTERRUPT_STATUS)
-			intel_pipe_handle_vblank(dev, pipe);
+	for_each_pipe(dev_priv, pipe) {
+		if (pipe_stats[pipe] & PIPE_START_VBLANK_INTERRUPT_STATUS &&
+		    intel_pipe_handle_vblank(dev, pipe))
+			intel_check_page_flip(dev, pipe);
 
 		if (pipe_stats[pipe] & PLANE_FLIP_DONE_INT_STATUS_VLV) {
 			intel_prepare_page_flip(dev, pipe);
@@ -2216,7 +2235,7 @@ static void ibx_irq_handler(struct drm_device *dev, u32 pch_iir)
 		DRM_ERROR("PCH poison interrupt\n");
 
 	if (pch_iir & SDE_FDI_MASK)
-		for_each_pipe(pipe)
+		for_each_pipe(dev_priv, pipe)
 			DRM_DEBUG_DRIVER("  pipe %c FDI IIR: 0x%08x\n",
 					 pipe_name(pipe),
 					 I915_READ(FDI_RX_IIR(pipe)));
@@ -2247,7 +2266,7 @@ static void ivb_err_int_handler(struct drm_device *dev)
 	if (err_int & ERR_INT_POISON)
 		DRM_ERROR("Poison interrupt\n");
 
-	for_each_pipe(pipe) {
+	for_each_pipe(dev_priv, pipe) {
 		if (err_int & ERR_INT_FIFO_UNDERRUN(pipe)) {
 			if (intel_set_cpu_fifo_underrun_reporting(dev, pipe,
 								  false))
@@ -2324,7 +2343,7 @@ static void cpt_irq_handler(struct drm_device *dev, u32 pch_iir)
 		DRM_DEBUG_DRIVER("Audio CP change interrupt\n");
 
 	if (pch_iir & SDE_FDI_MASK_CPT)
-		for_each_pipe(pipe)
+		for_each_pipe(dev_priv, pipe)
 			DRM_DEBUG_DRIVER("  pipe %c FDI IIR: 0x%08x\n",
 					 pipe_name(pipe),
 					 I915_READ(FDI_RX_IIR(pipe)));
@@ -2347,9 +2366,10 @@ static void ilk_display_irq_handler(struct drm_device *dev, u32 de_iir)
 	if (de_iir & DE_POISON)
 		DRM_ERROR("Poison interrupt\n");
 
-	for_each_pipe(pipe) {
-		if (de_iir & DE_PIPE_VBLANK(pipe))
-			intel_pipe_handle_vblank(dev, pipe);
+	for_each_pipe(dev_priv, pipe) {
+		if (de_iir & DE_PIPE_VBLANK(pipe) &&
+		    intel_pipe_handle_vblank(dev, pipe))
+			intel_check_page_flip(dev, pipe);
 
 		if (de_iir & DE_PIPE_FIFO_UNDERRUN(pipe))
 			if (intel_set_cpu_fifo_underrun_reporting(dev, pipe, false))
@@ -2397,9 +2417,10 @@ static void ivb_display_irq_handler(struct drm_device *dev, u32 de_iir)
 	if (de_iir & DE_GSE_IVB)
 		intel_opregion_asle_intr(dev);
 
-	for_each_pipe(pipe) {
-		if (de_iir & (DE_PIPE_VBLANK_IVB(pipe)))
-			intel_pipe_handle_vblank(dev, pipe);
+	for_each_pipe(dev_priv, pipe) {
+		if (de_iir & (DE_PIPE_VBLANK_IVB(pipe)) &&
+		    intel_pipe_handle_vblank(dev, pipe))
+			intel_check_page_flip(dev, pipe);
 
 		/* plane/pipes map 1:1 on ilk+ */
 		if (de_iir & DE_PLANE_FLIP_DONE_IVB(pipe)) {
@@ -2544,7 +2565,7 @@ static irqreturn_t gen8_irq_handler(int irq, void *arg)
 			DRM_ERROR("The master control interrupt lied (DE PORT)!\n");
 	}
 
-	for_each_pipe(pipe) {
+	for_each_pipe(dev_priv, pipe) {
 		uint32_t pipe_iir;
 
 		if (!(master_ctl & GEN8_DE_PIPE_IRQ(pipe)))
@@ -2554,8 +2575,9 @@ static irqreturn_t gen8_irq_handler(int irq, void *arg)
 		if (pipe_iir) {
 			ret = IRQ_HANDLED;
 			I915_WRITE(GEN8_DE_PIPE_IIR(pipe), pipe_iir);
-			if (pipe_iir & GEN8_PIPE_VBLANK)
-				intel_pipe_handle_vblank(dev, pipe);
+			if (pipe_iir & GEN8_PIPE_VBLANK &&
+			    intel_pipe_handle_vblank(dev, pipe))
+				intel_check_page_flip(dev, pipe);
 
 			if (pipe_iir & GEN8_PIPE_PRIMARY_FLIP_DONE) {
 				intel_prepare_page_flip(dev, pipe);
@@ -2763,7 +2785,7 @@ static void i915_report_and_clear_eir(struct drm_device *dev)
 
 	if (eir & I915_ERROR_MEMORY_REFRESH) {
 		pr_err("memory refresh error:\n");
-		for_each_pipe(pipe)
+		for_each_pipe(dev_priv, pipe)
 			pr_err("pipe %c stat: 0x%08x\n",
 			       pipe_name(pipe), I915_READ(PIPESTAT(pipe)));
 		/* pipestat has already been acked */
@@ -2860,52 +2882,6 @@ void i915_handle_error(struct drm_device *dev, bool wedged,
 	schedule_work(&dev_priv->gpu_error.work);
 }
 
-static void __always_unused i915_pageflip_stall_check(struct drm_device *dev, int pipe)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_crtc *crtc = dev_priv->pipe_to_crtc_mapping[pipe];
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	struct drm_i915_gem_object *obj;
-	struct intel_unpin_work *work;
-	unsigned long flags;
-	bool stall_detected;
-
-	/* Ignore early vblank irqs */
-	if (intel_crtc == NULL)
-		return;
-
-	spin_lock_irqsave(&dev->event_lock, flags);
-	work = intel_crtc->unpin_work;
-
-	if (work == NULL ||
-	    atomic_read(&work->pending) >= INTEL_FLIP_COMPLETE ||
-	    !work->enable_stall_check) {
-		/* Either the pending flip IRQ arrived, or we're too early. Don't check */
-		spin_unlock_irqrestore(&dev->event_lock, flags);
-		return;
-	}
-
-	/* Potential stall - if we see that the flip has happened, assume a missed interrupt */
-	obj = work->pending_flip_obj;
-	if (INTEL_INFO(dev)->gen >= 4) {
-		int dspsurf = DSPSURF(intel_crtc->plane);
-		stall_detected = I915_HI_DISPBASE(I915_READ(dspsurf)) ==
-					i915_gem_obj_ggtt_offset(obj);
-	} else {
-		int dspaddr = DSPADDR(intel_crtc->plane);
-		stall_detected = I915_READ(dspaddr) == (i915_gem_obj_ggtt_offset(obj) +
-							crtc->y * crtc->primary->fb->pitches[0] +
-							crtc->x * crtc->primary->fb->bits_per_pixel/8);
-	}
-
-	spin_unlock_irqrestore(&dev->event_lock, flags);
-
-	if (stall_detected) {
-		DRM_DEBUG_DRIVER("Pageflip stall detected\n");
-		intel_prepare_page_flip(dev, intel_crtc->plane);
-	}
-}
-
 /* Called from drm generic code, passed 'crtc' which
  * we use as a pipe index
  */
@@ -3441,7 +3417,7 @@ static void valleyview_irq_preinstall(struct drm_device *dev)
 
 	I915_WRITE(PORT_HOTPLUG_EN, 0);
 	I915_WRITE(PORT_HOTPLUG_STAT, I915_READ(PORT_HOTPLUG_STAT));
-	for_each_pipe(pipe)
+	for_each_pipe(dev_priv, pipe)
 		I915_WRITE(PIPESTAT(pipe), 0xffff);
 	I915_WRITE(VLV_IIR, 0xffffffff);
 	I915_WRITE(VLV_IMR, 0xffffffff);
@@ -3467,7 +3443,7 @@ static void gen8_irq_reset(struct drm_device *dev)
 
 	gen8_gt_irq_reset(dev_priv);
 
-	for_each_pipe(pipe)
+	for_each_pipe(dev_priv, pipe)
 		if (intel_display_power_enabled(dev_priv,
 						POWER_DOMAIN_PIPE(pipe)))
 			GEN8_IRQ_RESET_NDX(DE_PIPE, pipe);
@@ -3510,7 +3486,7 @@ static void cherryview_irq_preinstall(struct drm_device *dev)
 	I915_WRITE(PORT_HOTPLUG_EN, 0);
 	I915_WRITE(PORT_HOTPLUG_STAT, I915_READ(PORT_HOTPLUG_STAT));
 
-	for_each_pipe(pipe)
+	for_each_pipe(dev_priv, pipe)
 		I915_WRITE(PIPESTAT(pipe), 0xffff);
 
 	I915_WRITE(VLV_IMR, 0xffffffff);
@@ -3522,18 +3498,17 @@ static void cherryview_irq_preinstall(struct drm_device *dev)
 static void ibx_hpd_irq_setup(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_mode_config *mode_config = &dev->mode_config;
 	struct intel_encoder *intel_encoder;
 	u32 hotplug_irqs, hotplug, enabled_irqs = 0;
 
 	if (HAS_PCH_IBX(dev)) {
 		hotplug_irqs = SDE_HOTPLUG_MASK;
-		list_for_each_entry(intel_encoder, &mode_config->encoder_list, base.head)
+		for_each_intel_encoder(dev, intel_encoder)
 			if (dev_priv->hpd_stats[intel_encoder->hpd_pin].hpd_mark == HPD_ENABLED)
 				enabled_irqs |= hpd_ibx[intel_encoder->hpd_pin];
 	} else {
 		hotplug_irqs = SDE_HOTPLUG_MASK_CPT;
-		list_for_each_entry(intel_encoder, &mode_config->encoder_list, base.head)
+		for_each_intel_encoder(dev, intel_encoder)
 			if (dev_priv->hpd_stats[intel_encoder->hpd_pin].hpd_mark == HPD_ENABLED)
 				enabled_irqs |= hpd_cpt[intel_encoder->hpd_pin];
 	}
@@ -3782,28 +3757,31 @@ static int valleyview_irq_postinstall(struct drm_device *dev)
 
 static void gen8_gt_irq_postinstall(struct drm_i915_private *dev_priv)
 {
-	int i;
-
 	/* These are interrupts we'll toggle with the ring mask register */
 	uint32_t gt_interrupts[] = {
 		GT_RENDER_USER_INTERRUPT << GEN8_RCS_IRQ_SHIFT |
+			GT_CONTEXT_SWITCH_INTERRUPT << GEN8_RCS_IRQ_SHIFT |
 			GT_RENDER_L3_PARITY_ERROR_INTERRUPT |
-			GT_RENDER_USER_INTERRUPT << GEN8_BCS_IRQ_SHIFT,
+			GT_RENDER_USER_INTERRUPT << GEN8_BCS_IRQ_SHIFT |
+			GT_CONTEXT_SWITCH_INTERRUPT << GEN8_BCS_IRQ_SHIFT,
 		GT_RENDER_USER_INTERRUPT << GEN8_VCS1_IRQ_SHIFT |
-			GT_RENDER_USER_INTERRUPT << GEN8_VCS2_IRQ_SHIFT,
+			GT_CONTEXT_SWITCH_INTERRUPT << GEN8_VCS1_IRQ_SHIFT |
+			GT_RENDER_USER_INTERRUPT << GEN8_VCS2_IRQ_SHIFT |
+			GT_CONTEXT_SWITCH_INTERRUPT << GEN8_VCS2_IRQ_SHIFT,
 		0,
-		GT_RENDER_USER_INTERRUPT << GEN8_VECS_IRQ_SHIFT
+		GT_RENDER_USER_INTERRUPT << GEN8_VECS_IRQ_SHIFT |
+			GT_CONTEXT_SWITCH_INTERRUPT << GEN8_VECS_IRQ_SHIFT
 		};
 
-	for (i = 0; i < ARRAY_SIZE(gt_interrupts); i++)
-		GEN8_IRQ_INIT_NDX(GT, i, ~gt_interrupts[i], gt_interrupts[i]);
-
 	dev_priv->pm_irq_mask = 0xffffffff;
+	GEN8_IRQ_INIT_NDX(GT, 0, ~gt_interrupts[0], gt_interrupts[0]);
+	GEN8_IRQ_INIT_NDX(GT, 1, ~gt_interrupts[1], gt_interrupts[1]);
+	GEN8_IRQ_INIT_NDX(GT, 2, dev_priv->pm_irq_mask, dev_priv->pm_rps_events);
+	GEN8_IRQ_INIT_NDX(GT, 3, ~gt_interrupts[3], gt_interrupts[3]);
 }
 
 static void gen8_de_irq_postinstall(struct drm_i915_private *dev_priv)
 {
-	struct drm_device *dev = dev_priv->dev;
 	uint32_t de_pipe_masked = GEN8_PIPE_PRIMARY_FLIP_DONE |
 		GEN8_PIPE_CDCLK_CRC_DONE |
 		GEN8_DE_PIPE_IRQ_FAULT_ERRORS;
@@ -3814,7 +3792,7 @@ static void gen8_de_irq_postinstall(struct drm_i915_private *dev_priv)
 	dev_priv->de_irq_mask[PIPE_B] = ~de_pipe_masked;
 	dev_priv->de_irq_mask[PIPE_C] = ~de_pipe_masked;
 
-	for_each_pipe(pipe)
+	for_each_pipe(dev_priv, pipe)
 		if (intel_display_power_enabled(dev_priv,
 				POWER_DOMAIN_PIPE(pipe)))
 			GEN8_IRQ_INIT_NDX(DE_PIPE, pipe,
@@ -3859,12 +3837,12 @@ static int cherryview_irq_postinstall(struct drm_device *dev)
 	 */
 	dev_priv->irq_mask = ~enable_mask;
 
-	for_each_pipe(pipe)
+	for_each_pipe(dev_priv, pipe)
 		I915_WRITE(PIPESTAT(pipe), 0xffff);
 
 	spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
 	i915_enable_pipestat(dev_priv, PIPE_A, PIPE_GMBUS_INTERRUPT_STATUS);
-	for_each_pipe(pipe)
+	for_each_pipe(dev_priv, pipe)
 		i915_enable_pipestat(dev_priv, pipe, pipestat_enable);
 	spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
 
@@ -3901,7 +3879,7 @@ static void valleyview_irq_uninstall(struct drm_device *dev)
 
 	I915_WRITE(VLV_MASTER_IER, 0);
 
-	for_each_pipe(pipe)
+	for_each_pipe(dev_priv, pipe)
 		I915_WRITE(PIPESTAT(pipe), 0xffff);
 
 	I915_WRITE(HWSTAM, 0xffffffff);
@@ -3963,7 +3941,7 @@ do {							\
 	I915_WRITE(PORT_HOTPLUG_EN, 0);
 	I915_WRITE(PORT_HOTPLUG_STAT, I915_READ(PORT_HOTPLUG_STAT));
 
-	for_each_pipe(pipe)
+	for_each_pipe(dev_priv, pipe)
 		I915_WRITE(PIPESTAT(pipe), 0xffff);
 
 	I915_WRITE(VLV_IMR, 0xffffffff);
@@ -3987,7 +3965,7 @@ static void i8xx_irq_preinstall(struct drm_device * dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	int pipe;
 
-	for_each_pipe(pipe)
+	for_each_pipe(dev_priv, pipe)
 		I915_WRITE(PIPESTAT(pipe), 0);
 	I915_WRITE16(IMR, 0xffff);
 	I915_WRITE16(IER, 0x0);
@@ -4041,7 +4019,7 @@ static bool i8xx_handle_vblank(struct drm_device *dev,
 		return false;
 
 	if ((iir & flip_pending) == 0)
-		return false;
+		goto check_page_flip;
 
 	intel_prepare_page_flip(dev, plane);
 
@@ -4052,11 +4030,14 @@ static bool i8xx_handle_vblank(struct drm_device *dev,
 	 * an interrupt per se, we watch for the change at vblank.
 	 */
 	if (I915_READ16(ISR) & flip_pending)
-		return false;
+		goto check_page_flip;
 
 	intel_finish_page_flip(dev, pipe);
-
 	return true;
+
+check_page_flip:
+	intel_check_page_flip(dev, pipe);
+	return false;
 }
 
 static irqreturn_t i8xx_irq_handler(int irq, void *arg)
@@ -4087,7 +4068,7 @@ static irqreturn_t i8xx_irq_handler(int irq, void *arg)
 					  "Command parser error, iir 0x%08x",
 					  iir);
 
-		for_each_pipe(pipe) {
+		for_each_pipe(dev_priv, pipe) {
 			int reg = PIPESTAT(pipe);
 			pipe_stats[pipe] = I915_READ(reg);
 
@@ -4107,7 +4088,7 @@ static irqreturn_t i8xx_irq_handler(int irq, void *arg)
 		if (iir & I915_USER_INTERRUPT)
 			notify_ring(dev, &dev_priv->ring[RCS]);
 
-		for_each_pipe(pipe) {
+		for_each_pipe(dev_priv, pipe) {
 			int plane = pipe;
 			if (HAS_FBC(dev))
 				plane = !plane;
@@ -4135,7 +4116,7 @@ static void i8xx_irq_uninstall(struct drm_device * dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	int pipe;
 
-	for_each_pipe(pipe) {
+	for_each_pipe(dev_priv, pipe) {
 		/* Clear enable bits; then clear status bits */
 		I915_WRITE(PIPESTAT(pipe), 0);
 		I915_WRITE(PIPESTAT(pipe), I915_READ(PIPESTAT(pipe)));
@@ -4156,7 +4137,7 @@ static void i915_irq_preinstall(struct drm_device * dev)
 	}
 
 	I915_WRITE16(HWSTAM, 0xeffe);
-	for_each_pipe(pipe)
+	for_each_pipe(dev_priv, pipe)
 		I915_WRITE(PIPESTAT(pipe), 0);
 	I915_WRITE(IMR, 0xffffffff);
 	I915_WRITE(IER, 0x0);
@@ -4226,7 +4207,7 @@ static bool i915_handle_vblank(struct drm_device *dev,
 		return false;
 
 	if ((iir & flip_pending) == 0)
-		return false;
+		goto check_page_flip;
 
 	intel_prepare_page_flip(dev, plane);
 
@@ -4237,11 +4218,14 @@ static bool i915_handle_vblank(struct drm_device *dev,
 	 * an interrupt per se, we watch for the change at vblank.
 	 */
 	if (I915_READ(ISR) & flip_pending)
-		return false;
+		goto check_page_flip;
 
 	intel_finish_page_flip(dev, pipe);
-
 	return true;
+
+check_page_flip:
+	intel_check_page_flip(dev, pipe);
+	return false;
 }
 
 static irqreturn_t i915_irq_handler(int irq, void *arg)
@@ -4271,7 +4255,7 @@ static irqreturn_t i915_irq_handler(int irq, void *arg)
 					  "Command parser error, iir 0x%08x",
 					  iir);
 
-		for_each_pipe(pipe) {
+		for_each_pipe(dev_priv, pipe) {
 			int reg = PIPESTAT(pipe);
 			pipe_stats[pipe] = I915_READ(reg);
 
@@ -4297,7 +4281,7 @@ static irqreturn_t i915_irq_handler(int irq, void *arg)
 		if (iir & I915_USER_INTERRUPT)
 			notify_ring(dev, &dev_priv->ring[RCS]);
 
-		for_each_pipe(pipe) {
+		for_each_pipe(dev_priv, pipe) {
 			int plane = pipe;
 			if (HAS_FBC(dev))
 				plane = !plane;
@@ -4355,7 +4339,7 @@ static void i915_irq_uninstall(struct drm_device * dev)
 	}
 
 	I915_WRITE16(HWSTAM, 0xffff);
-	for_each_pipe(pipe) {
+	for_each_pipe(dev_priv, pipe) {
 		/* Clear enable bits; then clear status bits */
 		I915_WRITE(PIPESTAT(pipe), 0);
 		I915_WRITE(PIPESTAT(pipe), I915_READ(PIPESTAT(pipe)));
@@ -4375,7 +4359,7 @@ static void i965_irq_preinstall(struct drm_device * dev)
 	I915_WRITE(PORT_HOTPLUG_STAT, I915_READ(PORT_HOTPLUG_STAT));
 
 	I915_WRITE(HWSTAM, 0xeffe);
-	for_each_pipe(pipe)
+	for_each_pipe(dev_priv, pipe)
 		I915_WRITE(PIPESTAT(pipe), 0);
 	I915_WRITE(IMR, 0xffffffff);
 	I915_WRITE(IER, 0x0);
@@ -4444,7 +4428,6 @@ static int i965_irq_postinstall(struct drm_device *dev)
 static void i915_hpd_irq_setup(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_mode_config *mode_config = &dev->mode_config;
 	struct intel_encoder *intel_encoder;
 	u32 hotplug_en;
 
@@ -4455,7 +4438,7 @@ static void i915_hpd_irq_setup(struct drm_device *dev)
 		hotplug_en &= ~HOTPLUG_INT_EN_MASK;
 		/* Note HDMI and DP share hotplug bits */
 		/* enable bits are the same for all generations */
-		list_for_each_entry(intel_encoder, &mode_config->encoder_list, base.head)
+		for_each_intel_encoder(dev, intel_encoder)
 			if (dev_priv->hpd_stats[intel_encoder->hpd_pin].hpd_mark == HPD_ENABLED)
 				hotplug_en |= hpd_mask_i915[intel_encoder->hpd_pin];
 		/* Programming the CRT detection parameters tends
@@ -4501,7 +4484,7 @@ static irqreturn_t i965_irq_handler(int irq, void *arg)
 					  "Command parser error, iir 0x%08x",
 					  iir);
 
-		for_each_pipe(pipe) {
+		for_each_pipe(dev_priv, pipe) {
 			int reg = PIPESTAT(pipe);
 			pipe_stats[pipe] = I915_READ(reg);
 
@@ -4532,7 +4515,7 @@ static irqreturn_t i965_irq_handler(int irq, void *arg)
 		if (iir & I915_BSD_USER_INTERRUPT)
 			notify_ring(dev, &dev_priv->ring[VCS]);
 
-		for_each_pipe(pipe) {
+		for_each_pipe(dev_priv, pipe) {
 			if (pipe_stats[pipe] & PIPE_START_VBLANK_INTERRUPT_STATUS &&
 			    i915_handle_vblank(dev, pipe, pipe, iir))
 				flip_mask &= ~DISPLAY_PLANE_FLIP_PENDING(pipe);
@@ -4589,12 +4572,12 @@ static void i965_irq_uninstall(struct drm_device * dev)
 	I915_WRITE(PORT_HOTPLUG_STAT, I915_READ(PORT_HOTPLUG_STAT));
 
 	I915_WRITE(HWSTAM, 0xffffffff);
-	for_each_pipe(pipe)
+	for_each_pipe(dev_priv, pipe)
 		I915_WRITE(PIPESTAT(pipe), 0);
 	I915_WRITE(IMR, 0xffffffff);
 	I915_WRITE(IER, 0x0);
 
-	for_each_pipe(pipe)
+	for_each_pipe(dev_priv, pipe)
 		I915_WRITE(PIPESTAT(pipe),
 			   I915_READ(PIPESTAT(pipe)) & 0x8000ffff);
 	I915_WRITE(IIR, I915_READ(IIR));
@@ -4652,8 +4635,8 @@ void intel_irq_init(struct drm_device *dev)
 	INIT_WORK(&dev_priv->l3_parity.error_work, ivybridge_parity_work);
 
 	/* Let's track the enabled rps events */
-	if (IS_VALLEYVIEW(dev))
-		/* WaGsvRC0ResidenncyMethod:VLV */
+	if (IS_VALLEYVIEW(dev) && !IS_CHERRYVIEW(dev))
+		/* WaGsvRC0ResidencyMethod:vlv */
 		dev_priv->pm_rps_events = GEN6_PM_RP_UP_EI_EXPIRED;
 	else
 		dev_priv->pm_rps_events = GEN6_PM_RPS_EVENTS;
@@ -4680,6 +4663,14 @@ void intel_irq_init(struct drm_device *dev)
 		dev->max_vblank_count = 0xffffff; /* only 24 bits of frame count */
 	}
 
+	/*
+	 * Opt out of the vblank disable timer on everything except gen2.
+	 * Gen2 doesn't have a hardware frame counter and so depends on
+	 * vblank interrupts to produce sane vblank seuquence numbers.
+	 */
+	if (!IS_GEN2(dev))
+		dev->vblank_disable_immediate = true;
+
 	if (drm_core_check_feature(dev, DRIVER_MODESET)) {
 		dev->driver->get_vblank_timestamp = i915_get_vblank_timestamp;
 		dev->driver->get_scanout_position = i915_get_crtc_scanoutpos;
diff --git a/drivers/gpu/drm/i915/i915_params.c b/drivers/gpu/drm/i915/i915_params.c
index 9842fd2e742a..c91cb2033cc5 100644
--- a/drivers/gpu/drm/i915/i915_params.c
+++ b/drivers/gpu/drm/i915/i915_params.c
@@ -35,6 +35,7 @@ struct i915_params i915 __read_mostly = {
 	.vbt_sdvo_panel_type = -1,
 	.enable_rc6 = -1,
 	.enable_fbc = -1,
+	.enable_execlists = 0,
 	.enable_hangcheck = true,
 	.enable_ppgtt = -1,
 	.enable_psr = 0,
@@ -118,6 +119,11 @@ MODULE_PARM_DESC(enable_ppgtt,
 	"Override PPGTT usage. "
 	"(-1=auto [default], 0=disabled, 1=aliasing, 2=full)");
 
+module_param_named(enable_execlists, i915.enable_execlists, int, 0400);
+MODULE_PARM_DESC(enable_execlists,
+	"Override execlists usage. "
+	"(-1=auto, 0=disabled [default], 1=enabled)");
+
 module_param_named(enable_psr, i915.enable_psr, int, 0600);
 MODULE_PARM_DESC(enable_psr, "Enable PSR (default: false)");
 
diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index f29b44c86a2f..c01e5f31430e 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -143,6 +143,14 @@
 #define GAB_CTL				0x24000
 #define   GAB_CTL_CONT_AFTER_PAGEFAULT	(1<<8)
 
+#define GEN7_BIOS_RESERVED		0x1082C0
+#define GEN7_BIOS_RESERVED_1M		(0 << 5)
+#define GEN7_BIOS_RESERVED_256K		(1 << 5)
+#define GEN8_BIOS_RESERVED_SHIFT       7
+#define GEN7_BIOS_RESERVED_MASK        0x1
+#define GEN8_BIOS_RESERVED_MASK        0x3
+
+
 /* VGA stuff */
 
 #define VGA_ST01_MDA 0x3ba
@@ -272,6 +280,7 @@
 #define   MI_SEMAPHORE_POLL		(1<<15)
 #define   MI_SEMAPHORE_SAD_GTE_SDD	(1<<12)
 #define MI_STORE_DWORD_IMM	MI_INSTR(0x20, 1)
+#define MI_STORE_DWORD_IMM_GEN8	MI_INSTR(0x20, 2)
 #define   MI_MEM_VIRTUAL	(1 << 22) /* 965+ only */
 #define MI_STORE_DWORD_INDEX	MI_INSTR(0x21, 1)
 #define   MI_STORE_DWORD_INDEX_SHIFT 2
@@ -282,6 +291,7 @@
  *   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)
+#define   MI_LRI_FORCE_POSTED		(1<<12)
 #define MI_STORE_REGISTER_MEM(x) MI_INSTR(0x24, 2*(x)-1)
 #define MI_STORE_REGISTER_MEM_GEN8(x) MI_INSTR(0x24, 3*(x)-1)
 #define   MI_SRM_LRM_GLOBAL_GTT		(1<<22)
@@ -501,10 +511,26 @@
 #define BUNIT_REG_BISOC				0x11
 
 #define PUNIT_REG_DSPFREQ			0x36
+#define   DSPFREQSTAT_SHIFT_CHV			24
+#define   DSPFREQSTAT_MASK_CHV			(0x1f << DSPFREQSTAT_SHIFT_CHV)
+#define   DSPFREQGUAR_SHIFT_CHV			8
+#define   DSPFREQGUAR_MASK_CHV			(0x1f << DSPFREQGUAR_SHIFT_CHV)
 #define   DSPFREQSTAT_SHIFT			30
 #define   DSPFREQSTAT_MASK			(0x3 << DSPFREQSTAT_SHIFT)
 #define   DSPFREQGUAR_SHIFT			14
 #define   DSPFREQGUAR_MASK			(0x3 << DSPFREQGUAR_SHIFT)
+#define   _DP_SSC(val, pipe)			((val) << (2 * (pipe)))
+#define   DP_SSC_MASK(pipe)			_DP_SSC(0x3, (pipe))
+#define   DP_SSC_PWR_ON(pipe)			_DP_SSC(0x0, (pipe))
+#define   DP_SSC_CLK_GATE(pipe)			_DP_SSC(0x1, (pipe))
+#define   DP_SSC_RESET(pipe)			_DP_SSC(0x2, (pipe))
+#define   DP_SSC_PWR_GATE(pipe)			_DP_SSC(0x3, (pipe))
+#define   _DP_SSS(val, pipe)			((val) << (2 * (pipe) + 16))
+#define   DP_SSS_MASK(pipe)			_DP_SSS(0x3, (pipe))
+#define   DP_SSS_PWR_ON(pipe)			_DP_SSS(0x0, (pipe))
+#define   DP_SSS_CLK_GATE(pipe)			_DP_SSS(0x1, (pipe))
+#define   DP_SSS_RESET(pipe)			_DP_SSS(0x2, (pipe))
+#define   DP_SSS_PWR_GATE(pipe)			_DP_SSS(0x3, (pipe))
 
 /* See the PUNIT HAS v0.8 for the below bits */
 enum punit_power_well {
@@ -518,6 +544,11 @@ enum punit_power_well {
 	PUNIT_POWER_WELL_DPIO_TX_C_LANES_23	= 9,
 	PUNIT_POWER_WELL_DPIO_RX0		= 10,
 	PUNIT_POWER_WELL_DPIO_RX1		= 11,
+	PUNIT_POWER_WELL_DPIO_CMN_D		= 12,
+	/* FIXME: guesswork below */
+	PUNIT_POWER_WELL_DPIO_TX_D_LANES_01	= 13,
+	PUNIT_POWER_WELL_DPIO_TX_D_LANES_23	= 14,
+	PUNIT_POWER_WELL_DPIO_RX2		= 15,
 
 	PUNIT_POWER_WELL_NUM,
 };
@@ -838,8 +869,8 @@ enum punit_power_well {
 
 #define _VLV_TX_DW2_CH0			0x8288
 #define _VLV_TX_DW2_CH1			0x8488
-#define   DPIO_SWING_MARGIN_SHIFT	16
-#define   DPIO_SWING_MARGIN_MASK	(0xff << DPIO_SWING_MARGIN_SHIFT)
+#define   DPIO_SWING_MARGIN000_SHIFT	16
+#define   DPIO_SWING_MARGIN000_MASK	(0xff << DPIO_SWING_MARGIN000_SHIFT)
 #define   DPIO_UNIQ_TRANS_SCALE_SHIFT	8
 #define VLV_TX_DW2(ch) _PORT(ch, _VLV_TX_DW2_CH0, _VLV_TX_DW2_CH1)
 
@@ -847,12 +878,16 @@ enum punit_power_well {
 #define _VLV_TX_DW3_CH1			0x848c
 /* The following bit for CHV phy */
 #define   DPIO_TX_UNIQ_TRANS_SCALE_EN	(1<<27)
+#define   DPIO_SWING_MARGIN101_SHIFT	16
+#define   DPIO_SWING_MARGIN101_MASK	(0xff << DPIO_SWING_MARGIN101_SHIFT)
 #define VLV_TX_DW3(ch) _PORT(ch, _VLV_TX_DW3_CH0, _VLV_TX_DW3_CH1)
 
 #define _VLV_TX_DW4_CH0			0x8290
 #define _VLV_TX_DW4_CH1			0x8490
 #define   DPIO_SWING_DEEMPH9P5_SHIFT	24
 #define   DPIO_SWING_DEEMPH9P5_MASK	(0xff << DPIO_SWING_DEEMPH9P5_SHIFT)
+#define   DPIO_SWING_DEEMPH6P0_SHIFT	16
+#define   DPIO_SWING_DEEMPH6P0_MASK	(0xff << DPIO_SWING_DEEMPH6P0_SHIFT)
 #define VLV_TX_DW4(ch) _PORT(ch, _VLV_TX_DW4_CH0, _VLV_TX_DW4_CH1)
 
 #define _VLV_TX3_DW4_CH0		0x690
@@ -1003,6 +1038,13 @@ enum punit_power_well {
 #define   PGTBL_ADDRESS_LO_MASK	0xfffff000 /* bits [31:12] */
 #define   PGTBL_ADDRESS_HI_MASK	0x000000f0 /* bits [35:32] (gen4) */
 #define PGTBL_ER	0x02024
+#define PRB0_BASE (0x2030-0x30)
+#define PRB1_BASE (0x2040-0x30) /* 830,gen3 */
+#define PRB2_BASE (0x2050-0x30) /* gen3 */
+#define SRB0_BASE (0x2100-0x30) /* gen2 */
+#define SRB1_BASE (0x2110-0x30) /* gen2 */
+#define SRB2_BASE (0x2120-0x30) /* 830 */
+#define SRB3_BASE (0x2130-0x30) /* 830 */
 #define RENDER_RING_BASE	0x02000
 #define BSD_RING_BASE		0x04000
 #define GEN6_BSD_RING_BASE	0x12000
@@ -1064,6 +1106,7 @@ enum punit_power_well {
 #define RING_ACTHD_UDW(base)	((base)+0x5c)
 #define RING_NOPID(base)	((base)+0x94)
 #define RING_IMR(base)		((base)+0xa8)
+#define RING_HWSTAM(base)	((base)+0x98)
 #define RING_TIMESTAMP(base)	((base)+0x358)
 #define   TAIL_ADDR		0x001FFFF8
 #define   HEAD_WRAP_COUNT	0xFFE00000
@@ -1248,6 +1291,10 @@ enum punit_power_well {
 #define   INSTPM_TLB_INVALIDATE	(1<<9)
 #define   INSTPM_SYNC_FLUSH	(1<<5)
 #define ACTHD	        0x020c8
+#define MEM_MODE	0x020cc
+#define   MEM_DISPLAY_B_TRICKLE_FEED_DISABLE (1<<3) /* 830 only */
+#define   MEM_DISPLAY_A_TRICKLE_FEED_DISABLE (1<<2) /* 830/845 only */
+#define   MEM_DISPLAY_TRICKLE_FEED_DISABLE (1<<2) /* 85x only */
 #define FW_BLC		0x020d8
 #define FW_BLC2		0x020dc
 #define FW_BLC_SELF	0x020e0 /* 915+ only */
@@ -1380,6 +1427,7 @@ enum punit_power_well {
 #define GT_BSD_CS_ERROR_INTERRUPT		(1 << 15)
 #define GT_BSD_USER_INTERRUPT			(1 << 12)
 #define GT_RENDER_L3_PARITY_ERROR_INTERRUPT_S1	(1 << 11) /* hsw+; rsvd on snb, ivb, vlv */
+#define GT_CONTEXT_SWITCH_INTERRUPT		(1 <<  8)
 #define GT_RENDER_L3_PARITY_ERROR_INTERRUPT	(1 <<  5) /* !snb */
 #define GT_RENDER_PIPECTL_NOTIFY_INTERRUPT	(1 <<  4)
 #define GT_RENDER_CS_MASTER_ERROR_INTERRUPT	(1 <<  3)
@@ -1519,6 +1567,7 @@ enum punit_power_well {
 /* Framebuffer compression for Ironlake */
 #define ILK_DPFC_CB_BASE	0x43200
 #define ILK_DPFC_CONTROL	0x43208
+#define   FBC_CTL_FALSE_COLOR	(1<<10)
 /* The bit 28-8 is reserved */
 #define   DPFC_RESERVED		(0x1FFFFF00)
 #define ILK_DPFC_RECOMP_CTL	0x4320c
@@ -1675,12 +1724,9 @@ enum punit_power_well {
 #define DPIO_PHY_STATUS			(VLV_DISPLAY_BASE + 0x6240)
 #define   DPLL_PORTD_READY_MASK		(0xf)
 #define DISPLAY_PHY_CONTROL (VLV_DISPLAY_BASE + 0x60100)
-#define   PHY_COM_LANE_RESET_DEASSERT(phy, val) \
-				((phy == DPIO_PHY0) ? (val | 1) : (val | 2))
-#define   PHY_COM_LANE_RESET_ASSERT(phy, val) \
-				((phy == DPIO_PHY0) ? (val & ~1) : (val & ~2))
+#define   PHY_COM_LANE_RESET_DEASSERT(phy) (1 << (phy))
 #define DISPLAY_PHY_STATUS (VLV_DISPLAY_BASE + 0x60104)
-#define   PHY_POWERGOOD(phy)	((phy == DPIO_PHY0) ? (1<<31) : (1<<30))
+#define   PHY_POWERGOOD(phy)	(((phy) == DPIO_PHY0) ? (1<<31) : (1<<30))
 
 /*
  * The i830 generation, in LVDS mode, defines P1 as the bit number set within
@@ -2397,6 +2443,7 @@ enum punit_power_well {
 #define _PIPEASRC	0x6001c
 #define _BCLRPAT_A	0x60020
 #define _VSYNCSHIFT_A	0x60028
+#define _PIPE_MULT_A	0x6002c
 
 /* Pipe B timing regs */
 #define _HTOTAL_B	0x61000
@@ -2408,6 +2455,7 @@ enum punit_power_well {
 #define _PIPEBSRC	0x6101c
 #define _BCLRPAT_B	0x61020
 #define _VSYNCSHIFT_B	0x61028
+#define _PIPE_MULT_B	0x6102c
 
 #define TRANSCODER_A_OFFSET 0x60000
 #define TRANSCODER_B_OFFSET 0x61000
@@ -2428,6 +2476,7 @@ enum punit_power_well {
 #define BCLRPAT(trans) _TRANSCODER2(trans, _BCLRPAT_A)
 #define VSYNCSHIFT(trans) _TRANSCODER2(trans, _VSYNCSHIFT_A)
 #define PIPESRC(trans) _TRANSCODER2(trans, _PIPEASRC)
+#define PIPE_MULT(trans) _TRANSCODER2(trans, _PIPE_MULT_A)
 
 /* HSW+ eDP PSR registers */
 #define EDP_PSR_BASE(dev)                       (IS_HASWELL(dev) ? 0x64800 : 0x6f800)
@@ -3476,6 +3525,8 @@ enum punit_power_well {
 #define   DP_LINK_TRAIN_OFF		(3 << 28)
 #define   DP_LINK_TRAIN_MASK		(3 << 28)
 #define   DP_LINK_TRAIN_SHIFT		28
+#define   DP_LINK_TRAIN_PAT_3_CHV	(1 << 14)
+#define   DP_LINK_TRAIN_MASK_CHV	((3 << 28)|(1<<14))
 
 /* CPT Link training mode */
 #define   DP_LINK_TRAIN_PAT_1_CPT	(0 << 8)
@@ -3732,7 +3783,6 @@ enum punit_power_well {
 #define   PIPE_VSYNC_INTERRUPT_STATUS		(1UL<<9)
 #define   PIPE_DISPLAY_LINE_COMPARE_STATUS	(1UL<<8)
 #define   PIPE_DPST_EVENT_STATUS		(1UL<<7)
-#define   PIPE_LEGACY_BLC_EVENT_STATUS		(1UL<<6)
 #define   PIPE_A_PSR_STATUS_VLV			(1UL<<6)
 #define   PIPE_LEGACY_BLC_EVENT_STATUS		(1UL<<6)
 #define   PIPE_ODD_FIELD_INTERRUPT_STATUS	(1UL<<5)
@@ -3842,73 +3892,151 @@ enum punit_power_well {
 #define   DSPARB_BEND_SHIFT	9 /* on 855 */
 #define   DSPARB_AEND_SHIFT	0
 
+/* pnv/gen4/g4x/vlv/chv */
 #define DSPFW1			(dev_priv->info.display_mmio_offset + 0x70034)
-#define   DSPFW_SR_SHIFT	23
-#define   DSPFW_SR_MASK		(0x1ff<<23)
-#define   DSPFW_CURSORB_SHIFT	16
-#define   DSPFW_CURSORB_MASK	(0x3f<<16)
-#define   DSPFW_PLANEB_SHIFT	8
-#define   DSPFW_PLANEB_MASK	(0x7f<<8)
-#define   DSPFW_PLANEA_MASK	(0x7f)
+#define   DSPFW_SR_SHIFT		23
+#define   DSPFW_SR_MASK			(0x1ff<<23)
+#define   DSPFW_CURSORB_SHIFT		16
+#define   DSPFW_CURSORB_MASK		(0x3f<<16)
+#define   DSPFW_PLANEB_SHIFT		8
+#define   DSPFW_PLANEB_MASK		(0x7f<<8)
+#define   DSPFW_PLANEB_MASK_VLV		(0xff<<8) /* vlv/chv */
+#define   DSPFW_PLANEA_SHIFT		0
+#define   DSPFW_PLANEA_MASK		(0x7f<<0)
+#define   DSPFW_PLANEA_MASK_VLV		(0xff<<0) /* vlv/chv */
 #define DSPFW2			(dev_priv->info.display_mmio_offset + 0x70038)
-#define   DSPFW_CURSORA_MASK	0x00003f00
-#define   DSPFW_CURSORA_SHIFT	8
-#define   DSPFW_PLANEC_MASK	(0x7f)
+#define   DSPFW_FBC_SR_EN		(1<<31)	  /* g4x */
+#define   DSPFW_FBC_SR_SHIFT		28
+#define   DSPFW_FBC_SR_MASK		(0x7<<28) /* g4x */
+#define   DSPFW_FBC_HPLL_SR_SHIFT	24
+#define   DSPFW_FBC_HPLL_SR_MASK	(0xf<<24) /* g4x */
+#define   DSPFW_SPRITEB_SHIFT		(16)
+#define   DSPFW_SPRITEB_MASK		(0x7f<<16) /* g4x */
+#define   DSPFW_SPRITEB_MASK_VLV	(0xff<<16) /* vlv/chv */
+#define   DSPFW_CURSORA_SHIFT		8
+#define   DSPFW_CURSORA_MASK		(0x3f<<8)
+#define   DSPFW_PLANEC_SHIFT_OLD	0
+#define   DSPFW_PLANEC_MASK_OLD		(0x7f<<0) /* pre-gen4 sprite C */
+#define   DSPFW_SPRITEA_SHIFT		0
+#define   DSPFW_SPRITEA_MASK		(0x7f<<0) /* g4x */
+#define   DSPFW_SPRITEA_MASK_VLV	(0xff<<0) /* vlv/chv */
 #define DSPFW3			(dev_priv->info.display_mmio_offset + 0x7003c)
-#define   DSPFW_HPLL_SR_EN	(1<<31)
-#define   DSPFW_CURSOR_SR_SHIFT	24
+#define   DSPFW_HPLL_SR_EN		(1<<31)
 #define   PINEVIEW_SELF_REFRESH_EN	(1<<30)
+#define   DSPFW_CURSOR_SR_SHIFT		24
 #define   DSPFW_CURSOR_SR_MASK		(0x3f<<24)
 #define   DSPFW_HPLL_CURSOR_SHIFT	16
 #define   DSPFW_HPLL_CURSOR_MASK	(0x3f<<16)
-#define   DSPFW_HPLL_SR_MASK		(0x1ff)
-#define DSPFW4			(dev_priv->info.display_mmio_offset + 0x70070)
-#define DSPFW7			(dev_priv->info.display_mmio_offset + 0x7007c)
+#define   DSPFW_HPLL_SR_SHIFT		0
+#define   DSPFW_HPLL_SR_MASK		(0x1ff<<0)
+
+/* vlv/chv */
+#define DSPFW4			(VLV_DISPLAY_BASE + 0x70070)
+#define   DSPFW_SPRITEB_WM1_SHIFT	16
+#define   DSPFW_SPRITEB_WM1_MASK	(0xff<<16)
+#define   DSPFW_CURSORA_WM1_SHIFT	8
+#define   DSPFW_CURSORA_WM1_MASK	(0x3f<<8)
+#define   DSPFW_SPRITEA_WM1_SHIFT	0
+#define   DSPFW_SPRITEA_WM1_MASK	(0xff<<0)
+#define DSPFW5			(VLV_DISPLAY_BASE + 0x70074)
+#define   DSPFW_PLANEB_WM1_SHIFT	24
+#define   DSPFW_PLANEB_WM1_MASK		(0xff<<24)
+#define   DSPFW_PLANEA_WM1_SHIFT	16
+#define   DSPFW_PLANEA_WM1_MASK		(0xff<<16)
+#define   DSPFW_CURSORB_WM1_SHIFT	8
+#define   DSPFW_CURSORB_WM1_MASK	(0x3f<<8)
+#define   DSPFW_CURSOR_SR_WM1_SHIFT	0
+#define   DSPFW_CURSOR_SR_WM1_MASK	(0x3f<<0)
+#define DSPFW6			(VLV_DISPLAY_BASE + 0x70078)
+#define   DSPFW_SR_WM1_SHIFT		0
+#define   DSPFW_SR_WM1_MASK		(0x1ff<<0)
+#define DSPFW7			(VLV_DISPLAY_BASE + 0x7007c)
+#define DSPFW7_CHV		(VLV_DISPLAY_BASE + 0x700b4) /* wtf #1? */
+#define   DSPFW_SPRITED_WM1_SHIFT	24
+#define   DSPFW_SPRITED_WM1_MASK	(0xff<<24)
+#define   DSPFW_SPRITED_SHIFT		16
+#define   DSPFW_SPRITED_MASK		(0xff<<16)
+#define   DSPFW_SPRITEC_WM1_SHIFT	8
+#define   DSPFW_SPRITEC_WM1_MASK	(0xff<<8)
+#define   DSPFW_SPRITEC_SHIFT		0
+#define   DSPFW_SPRITEC_MASK		(0xff<<0)
+#define DSPFW8_CHV		(VLV_DISPLAY_BASE + 0x700b8)
+#define   DSPFW_SPRITEF_WM1_SHIFT	24
+#define   DSPFW_SPRITEF_WM1_MASK	(0xff<<24)
+#define   DSPFW_SPRITEF_SHIFT		16
+#define   DSPFW_SPRITEF_MASK		(0xff<<16)
+#define   DSPFW_SPRITEE_WM1_SHIFT	8
+#define   DSPFW_SPRITEE_WM1_MASK	(0xff<<8)
+#define   DSPFW_SPRITEE_SHIFT		0
+#define   DSPFW_SPRITEE_MASK		(0xff<<0)
+#define DSPFW9_CHV		(VLV_DISPLAY_BASE + 0x7007c) /* wtf #2? */
+#define   DSPFW_PLANEC_WM1_SHIFT	24
+#define   DSPFW_PLANEC_WM1_MASK		(0xff<<24)
+#define   DSPFW_PLANEC_SHIFT		16
+#define   DSPFW_PLANEC_MASK		(0xff<<16)
+#define   DSPFW_CURSORC_WM1_SHIFT	8
+#define   DSPFW_CURSORC_WM1_MASK	(0x3f<<16)
+#define   DSPFW_CURSORC_SHIFT		0
+#define   DSPFW_CURSORC_MASK		(0x3f<<0)
+
+/* vlv/chv high order bits */
+#define DSPHOWM			(VLV_DISPLAY_BASE + 0x70064)
+#define   DSPFW_SR_HI_SHIFT		24
+#define   DSPFW_SR_HI_MASK		(1<<24)
+#define   DSPFW_SPRITEF_HI_SHIFT	23
+#define   DSPFW_SPRITEF_HI_MASK		(1<<23)
+#define   DSPFW_SPRITEE_HI_SHIFT	22
+#define   DSPFW_SPRITEE_HI_MASK		(1<<22)
+#define   DSPFW_PLANEC_HI_SHIFT		21
+#define   DSPFW_PLANEC_HI_MASK		(1<<21)
+#define   DSPFW_SPRITED_HI_SHIFT	20
+#define   DSPFW_SPRITED_HI_MASK		(1<<20)
+#define   DSPFW_SPRITEC_HI_SHIFT	16
+#define   DSPFW_SPRITEC_HI_MASK		(1<<16)
+#define   DSPFW_PLANEB_HI_SHIFT		12
+#define   DSPFW_PLANEB_HI_MASK		(1<<12)
+#define   DSPFW_SPRITEB_HI_SHIFT	8
+#define   DSPFW_SPRITEB_HI_MASK		(1<<8)
+#define   DSPFW_SPRITEA_HI_SHIFT	4
+#define   DSPFW_SPRITEA_HI_MASK		(1<<4)
+#define   DSPFW_PLANEA_HI_SHIFT		0
+#define   DSPFW_PLANEA_HI_MASK		(1<<0)
+#define DSPHOWM1		(VLV_DISPLAY_BASE + 0x70068)
+#define   DSPFW_SR_WM1_HI_SHIFT		24
+#define   DSPFW_SR_WM1_HI_MASK		(1<<24)
+#define   DSPFW_SPRITEF_WM1_HI_SHIFT	23
+#define   DSPFW_SPRITEF_WM1_HI_MASK	(1<<23)
+#define   DSPFW_SPRITEE_WM1_HI_SHIFT	22
+#define   DSPFW_SPRITEE_WM1_HI_MASK	(1<<22)
+#define   DSPFW_PLANEC_WM1_HI_SHIFT	21
+#define   DSPFW_PLANEC_WM1_HI_MASK	(1<<21)
+#define   DSPFW_SPRITED_WM1_HI_SHIFT	20
+#define   DSPFW_SPRITED_WM1_HI_MASK	(1<<20)
+#define   DSPFW_SPRITEC_WM1_HI_SHIFT	16
+#define   DSPFW_SPRITEC_WM1_HI_MASK	(1<<16)
+#define   DSPFW_PLANEB_WM1_HI_SHIFT	12
+#define   DSPFW_PLANEB_WM1_HI_MASK	(1<<12)
+#define   DSPFW_SPRITEB_WM1_HI_SHIFT	8
+#define   DSPFW_SPRITEB_WM1_HI_MASK	(1<<8)
+#define   DSPFW_SPRITEA_WM1_HI_SHIFT	4
+#define   DSPFW_SPRITEA_WM1_HI_MASK	(1<<4)
+#define   DSPFW_PLANEA_WM1_HI_SHIFT	0
+#define   DSPFW_PLANEA_WM1_HI_MASK	(1<<0)
 
 /* drain latency register values*/
 #define DRAIN_LATENCY_PRECISION_32	32
 #define DRAIN_LATENCY_PRECISION_64	64
-#define VLV_DDL1			(VLV_DISPLAY_BASE + 0x70050)
-#define DDL_CURSORA_PRECISION_64	(1<<31)
-#define DDL_CURSORA_PRECISION_32	(0<<31)
-#define DDL_CURSORA_SHIFT		24
-#define DDL_SPRITEB_PRECISION_64	(1<<23)
-#define DDL_SPRITEB_PRECISION_32	(0<<23)
-#define DDL_SPRITEB_SHIFT		16
-#define DDL_SPRITEA_PRECISION_64	(1<<15)
-#define DDL_SPRITEA_PRECISION_32	(0<<15)
-#define DDL_SPRITEA_SHIFT		8
-#define DDL_PLANEA_PRECISION_64		(1<<7)
-#define DDL_PLANEA_PRECISION_32		(0<<7)
-#define DDL_PLANEA_SHIFT		0
-
-#define VLV_DDL2			(VLV_DISPLAY_BASE + 0x70054)
-#define DDL_CURSORB_PRECISION_64	(1<<31)
-#define DDL_CURSORB_PRECISION_32	(0<<31)
-#define DDL_CURSORB_SHIFT		24
-#define DDL_SPRITED_PRECISION_64	(1<<23)
-#define DDL_SPRITED_PRECISION_32	(0<<23)
-#define DDL_SPRITED_SHIFT		16
-#define DDL_SPRITEC_PRECISION_64	(1<<15)
-#define DDL_SPRITEC_PRECISION_32	(0<<15)
-#define DDL_SPRITEC_SHIFT		8
-#define DDL_PLANEB_PRECISION_64		(1<<7)
-#define DDL_PLANEB_PRECISION_32		(0<<7)
-#define DDL_PLANEB_SHIFT		0
-
-#define VLV_DDL3			(VLV_DISPLAY_BASE + 0x70058)
-#define DDL_CURSORC_PRECISION_64	(1<<31)
-#define DDL_CURSORC_PRECISION_32	(0<<31)
-#define DDL_CURSORC_SHIFT		24
-#define DDL_SPRITEF_PRECISION_64	(1<<23)
-#define DDL_SPRITEF_PRECISION_32	(0<<23)
-#define DDL_SPRITEF_SHIFT		16
-#define DDL_SPRITEE_PRECISION_64	(1<<15)
-#define DDL_SPRITEE_PRECISION_32	(0<<15)
-#define DDL_SPRITEE_SHIFT		8
-#define DDL_PLANEC_PRECISION_64		(1<<7)
-#define DDL_PLANEC_PRECISION_32		(0<<7)
-#define DDL_PLANEC_SHIFT		0
+#define VLV_DDL(pipe)			(VLV_DISPLAY_BASE + 0x70050 + 4 * (pipe))
+#define DDL_CURSOR_PRECISION_64		(1<<31)
+#define DDL_CURSOR_PRECISION_32		(0<<31)
+#define DDL_CURSOR_SHIFT		24
+#define DDL_SPRITE_PRECISION_64(sprite)	(1<<(15+8*(sprite)))
+#define DDL_SPRITE_PRECISION_32(sprite)	(0<<(15+8*(sprite)))
+#define DDL_SPRITE_SHIFT(sprite)	(8+8*(sprite))
+#define DDL_PLANE_PRECISION_64		(1<<7)
+#define DDL_PLANE_PRECISION_32		(0<<7)
+#define DDL_PLANE_SHIFT			0
+#define DRAIN_LATENCY_MASK		0x7f
 
 /* FIFO watermark sizes etc */
 #define G4X_FIFO_LINE_SIZE	64
@@ -4026,7 +4154,8 @@ enum punit_power_well {
 /* Old style CUR*CNTR flags (desktop 8xx) */
 #define   CURSOR_ENABLE		0x80000000
 #define   CURSOR_GAMMA_ENABLE	0x40000000
-#define   CURSOR_STRIDE_MASK	0x30000000
+#define   CURSOR_STRIDE_SHIFT	28
+#define   CURSOR_STRIDE(x)	((ffs(x)-9) << CURSOR_STRIDE_SHIFT) /* 256,512,1k,2k */
 #define   CURSOR_PIPE_CSC_ENABLE (1<<24)
 #define   CURSOR_FORMAT_SHIFT	24
 #define   CURSOR_FORMAT_MASK	(0x07 << CURSOR_FORMAT_SHIFT)
@@ -4111,6 +4240,7 @@ enum punit_power_well {
 #define   DISPPLANE_NO_LINE_DOUBLE		0
 #define   DISPPLANE_STEREO_POLARITY_FIRST	0
 #define   DISPPLANE_STEREO_POLARITY_SECOND	(1<<18)
+#define   DISPPLANE_ROTATE_180         (1<<15)
 #define   DISPPLANE_TRICKLE_FEED_DISABLE	(1<<14) /* Ironlake */
 #define   DISPPLANE_TILED			(1<<10)
 #define _DSPAADDR				0x70184
@@ -4195,6 +4325,7 @@ enum punit_power_well {
 #define   DVS_YUV_ORDER_UYVY	(1<<16)
 #define   DVS_YUV_ORDER_YVYU	(2<<16)
 #define   DVS_YUV_ORDER_VYUY	(3<<16)
+#define   DVS_ROTATE_180	(1<<15)
 #define   DVS_DEST_KEY		(1<<2)
 #define   DVS_TRICKLE_FEED_DISABLE (1<<14)
 #define   DVS_TILED		(1<<10)
@@ -4265,6 +4396,7 @@ enum punit_power_well {
 #define   SPRITE_YUV_ORDER_UYVY		(1<<16)
 #define   SPRITE_YUV_ORDER_YVYU		(2<<16)
 #define   SPRITE_YUV_ORDER_VYUY		(3<<16)
+#define   SPRITE_ROTATE_180		(1<<15)
 #define   SPRITE_TRICKLE_FEED_DISABLE	(1<<14)
 #define   SPRITE_INT_GAMMA_ENABLE	(1<<13)
 #define   SPRITE_TILED			(1<<10)
@@ -4338,6 +4470,7 @@ enum punit_power_well {
 #define   SP_YUV_ORDER_UYVY		(1<<16)
 #define   SP_YUV_ORDER_YVYU		(2<<16)
 #define   SP_YUV_ORDER_VYUY		(3<<16)
+#define   SP_ROTATE_180			(1<<15)
 #define   SP_TILED			(1<<10)
 #define _SPALINOFF		(VLV_DISPLAY_BASE + 0x72184)
 #define _SPASTRIDE		(VLV_DISPLAY_BASE + 0x72188)
@@ -5246,8 +5379,7 @@ enum punit_power_well {
 #define PIPEA_PP_STATUS         (VLV_DISPLAY_BASE + 0x61200)
 #define PIPEA_PP_CONTROL        (VLV_DISPLAY_BASE + 0x61204)
 #define PIPEA_PP_ON_DELAYS      (VLV_DISPLAY_BASE + 0x61208)
-#define  PANEL_PORT_SELECT_DPB_VLV	(1 << 30)
-#define  PANEL_PORT_SELECT_DPC_VLV	(2 << 30)
+#define  PANEL_PORT_SELECT_VLV(port)	((port) << 30)
 #define PIPEA_PP_OFF_DELAYS     (VLV_DISPLAY_BASE + 0x6120c)
 #define PIPEA_PP_DIVISOR        (VLV_DISPLAY_BASE + 0x61210)
 
@@ -5407,7 +5539,6 @@ enum punit_power_well {
 #define   VLV_GTLC_ALLOWWAKEERR			(1 << 1)
 #define   VLV_GTLC_PW_MEDIA_STATUS_MASK		(1 << 5)
 #define   VLV_GTLC_PW_RENDER_STATUS_MASK	(1 << 7)
-#define VLV_GTLC_SURVIVABILITY_REG              0x130098
 #define  FORCEWAKE_MT				0xa188 /* multi-threaded */
 #define   FORCEWAKE_KERNEL			0x1
 #define   FORCEWAKE_USER			0x2
@@ -5545,12 +5676,6 @@ enum punit_power_well {
 						 GEN6_PM_RP_DOWN_THRESHOLD | \
 						 GEN6_PM_RP_DOWN_TIMEOUT)
 
-#define CHV_CZ_CLOCK_FREQ_MODE_200			200
-#define CHV_CZ_CLOCK_FREQ_MODE_267			267
-#define CHV_CZ_CLOCK_FREQ_MODE_320			320
-#define CHV_CZ_CLOCK_FREQ_MODE_333			333
-#define CHV_CZ_CLOCK_FREQ_MODE_400			400
-
 #define GEN7_GT_SCRATCH_BASE			0x4F100
 #define GEN7_GT_SCRATCH_REG_NUM			8
 
@@ -5866,15 +5991,7 @@ enum punit_power_well {
 #define DDI_BUF_CTL_B				0x64100
 #define DDI_BUF_CTL(port) _PORT(port, DDI_BUF_CTL_A, DDI_BUF_CTL_B)
 #define  DDI_BUF_CTL_ENABLE			(1<<31)
-#define  DDI_BUF_EMP_400MV_0DB_HSW		(0<<24)   /* Sel0 */
-#define  DDI_BUF_EMP_400MV_3_5DB_HSW		(1<<24)   /* Sel1 */
-#define  DDI_BUF_EMP_400MV_6DB_HSW		(2<<24)   /* Sel2 */
-#define  DDI_BUF_EMP_400MV_9_5DB_HSW		(3<<24)   /* Sel3 */
-#define  DDI_BUF_EMP_600MV_0DB_HSW		(4<<24)   /* Sel4 */
-#define  DDI_BUF_EMP_600MV_3_5DB_HSW		(5<<24)   /* Sel5 */
-#define  DDI_BUF_EMP_600MV_6DB_HSW		(6<<24)   /* Sel6 */
-#define  DDI_BUF_EMP_800MV_0DB_HSW		(7<<24)   /* Sel7 */
-#define  DDI_BUF_EMP_800MV_3_5DB_HSW		(8<<24)   /* Sel8 */
+#define  DDI_BUF_TRANS_SELECT(n)	((n) << 24)
 #define  DDI_BUF_EMP_MASK			(0xf<<24)
 #define  DDI_BUF_PORT_REVERSAL			(1<<16)
 #define  DDI_BUF_IS_IDLE			(1<<7)
diff --git a/drivers/gpu/drm/i915/i915_sysfs.c b/drivers/gpu/drm/i915/i915_sysfs.c
index ae7fd8fc27f0..503847f18fdd 100644
--- a/drivers/gpu/drm/i915/i915_sysfs.c
+++ b/drivers/gpu/drm/i915/i915_sysfs.c
@@ -540,7 +540,7 @@ static ssize_t error_state_read(struct file *filp, struct kobject *kobj,
 
 	memset(&error_priv, 0, sizeof(error_priv));
 
-	ret = i915_error_state_buf_init(&error_str, count, off);
+	ret = i915_error_state_buf_init(&error_str, to_i915(dev), count, off);
 	if (ret)
 		return ret;
 
diff --git a/drivers/gpu/drm/i915/intel_bios.c b/drivers/gpu/drm/i915/intel_bios.c
index afcc8dd40bdd..a4bd90f36a03 100644
--- a/drivers/gpu/drm/i915/intel_bios.c
+++ b/drivers/gpu/drm/i915/intel_bios.c
@@ -627,16 +627,16 @@ parse_edp(struct drm_i915_private *dev_priv, struct bdb_header *bdb)
 
 	switch (edp_link_params->preemphasis) {
 	case EDP_PREEMPHASIS_NONE:
-		dev_priv->vbt.edp_preemphasis = DP_TRAIN_PRE_EMPHASIS_0;
+		dev_priv->vbt.edp_preemphasis = DP_TRAIN_PRE_EMPH_LEVEL_0;
 		break;
 	case EDP_PREEMPHASIS_3_5dB:
-		dev_priv->vbt.edp_preemphasis = DP_TRAIN_PRE_EMPHASIS_3_5;
+		dev_priv->vbt.edp_preemphasis = DP_TRAIN_PRE_EMPH_LEVEL_1;
 		break;
 	case EDP_PREEMPHASIS_6dB:
-		dev_priv->vbt.edp_preemphasis = DP_TRAIN_PRE_EMPHASIS_6;
+		dev_priv->vbt.edp_preemphasis = DP_TRAIN_PRE_EMPH_LEVEL_2;
 		break;
 	case EDP_PREEMPHASIS_9_5dB:
-		dev_priv->vbt.edp_preemphasis = DP_TRAIN_PRE_EMPHASIS_9_5;
+		dev_priv->vbt.edp_preemphasis = DP_TRAIN_PRE_EMPH_LEVEL_3;
 		break;
 	default:
 		DRM_DEBUG_KMS("VBT has unknown eDP pre-emphasis value %u\n",
@@ -646,16 +646,16 @@ parse_edp(struct drm_i915_private *dev_priv, struct bdb_header *bdb)
 
 	switch (edp_link_params->vswing) {
 	case EDP_VSWING_0_4V:
-		dev_priv->vbt.edp_vswing = DP_TRAIN_VOLTAGE_SWING_400;
+		dev_priv->vbt.edp_vswing = DP_TRAIN_VOLTAGE_SWING_LEVEL_0;
 		break;
 	case EDP_VSWING_0_6V:
-		dev_priv->vbt.edp_vswing = DP_TRAIN_VOLTAGE_SWING_600;
+		dev_priv->vbt.edp_vswing = DP_TRAIN_VOLTAGE_SWING_LEVEL_1;
 		break;
 	case EDP_VSWING_0_8V:
-		dev_priv->vbt.edp_vswing = DP_TRAIN_VOLTAGE_SWING_800;
+		dev_priv->vbt.edp_vswing = DP_TRAIN_VOLTAGE_SWING_LEVEL_2;
 		break;
 	case EDP_VSWING_1_2V:
-		dev_priv->vbt.edp_vswing = DP_TRAIN_VOLTAGE_SWING_1200;
+		dev_priv->vbt.edp_vswing = DP_TRAIN_VOLTAGE_SWING_LEVEL_3;
 		break;
 	default:
 		DRM_DEBUG_KMS("VBT has unknown eDP voltage swing value %u\n",
@@ -976,12 +976,10 @@ static void parse_ddi_port(struct drm_i915_private *dev_priv, enum port port,
 	if (bdb->version >= 158) {
 		/* The VBT HDMI level shift values match the table we have. */
 		hdmi_level_shift = child->raw[7] & 0xF;
-		if (hdmi_level_shift < 0xC) {
-			DRM_DEBUG_KMS("VBT HDMI level shift for port %c: %d\n",
-				      port_name(port),
-				      hdmi_level_shift);
-			info->hdmi_level_shift = hdmi_level_shift;
-		}
+		DRM_DEBUG_KMS("VBT HDMI level shift for port %c: %d\n",
+			      port_name(port),
+			      hdmi_level_shift);
+		info->hdmi_level_shift = hdmi_level_shift;
 	}
 }
 
@@ -1114,8 +1112,7 @@ init_vbt_defaults(struct drm_i915_private *dev_priv)
 		struct ddi_vbt_port_info *info =
 			&dev_priv->vbt.ddi_port_info[port];
 
-		/* Recommended BSpec default: 800mV 0dB. */
-		info->hdmi_level_shift = 6;
+		info->hdmi_level_shift = HDMI_LEVEL_SHIFT_UNKNOWN;
 
 		info->supports_dvi = (port != PORT_A && port != PORT_E);
 		info->supports_hdmi = info->supports_dvi;
diff --git a/drivers/gpu/drm/i915/intel_bios.h b/drivers/gpu/drm/i915/intel_bios.h
index b98667796337..905999bee2ac 100644
--- a/drivers/gpu/drm/i915/intel_bios.h
+++ b/drivers/gpu/drm/i915/intel_bios.h
@@ -802,7 +802,8 @@ struct mipi_config {
 
 	u16 rsvd4;
 
-	u8 rsvd5[5];
+	u8 rsvd5;
+	u32 target_burst_mode_freq;
 	u32 dsi_ddr_clk;
 	u32 bridge_ref_clk;
 
diff --git a/drivers/gpu/drm/i915/intel_ddi.c b/drivers/gpu/drm/i915/intel_ddi.c
index 5db0b5552e39..b63d4fa204a3 100644
--- a/drivers/gpu/drm/i915/intel_ddi.c
+++ b/drivers/gpu/drm/i915/intel_ddi.c
@@ -28,87 +28,103 @@
 #include "i915_drv.h"
 #include "intel_drv.h"
 
+struct ddi_buf_trans {
+	u32 trans1;	/* balance leg enable, de-emph level */
+	u32 trans2;	/* vref sel, vswing */
+};
+
 /* HDMI/DVI modes ignore everything but the last 2 items. So we share
  * them for both DP and FDI transports, allowing those ports to
  * automatically adapt to HDMI connections as well
  */
-static const u32 hsw_ddi_translations_dp[] = {
-	0x00FFFFFF, 0x0006000E,		/* DP parameters */
-	0x00D75FFF, 0x0005000A,
-	0x00C30FFF, 0x00040006,
-	0x80AAAFFF, 0x000B0000,
-	0x00FFFFFF, 0x0005000A,
-	0x00D75FFF, 0x000C0004,
-	0x80C30FFF, 0x000B0000,
-	0x00FFFFFF, 0x00040006,
-	0x80D75FFF, 0x000B0000,
+static const struct ddi_buf_trans hsw_ddi_translations_dp[] = {
+	{ 0x00FFFFFF, 0x0006000E },
+	{ 0x00D75FFF, 0x0005000A },
+	{ 0x00C30FFF, 0x00040006 },
+	{ 0x80AAAFFF, 0x000B0000 },
+	{ 0x00FFFFFF, 0x0005000A },
+	{ 0x00D75FFF, 0x000C0004 },
+	{ 0x80C30FFF, 0x000B0000 },
+	{ 0x00FFFFFF, 0x00040006 },
+	{ 0x80D75FFF, 0x000B0000 },
 };
 
-static const u32 hsw_ddi_translations_fdi[] = {
-	0x00FFFFFF, 0x0007000E,		/* FDI parameters */
-	0x00D75FFF, 0x000F000A,
-	0x00C30FFF, 0x00060006,
-	0x00AAAFFF, 0x001E0000,
-	0x00FFFFFF, 0x000F000A,
-	0x00D75FFF, 0x00160004,
-	0x00C30FFF, 0x001E0000,
-	0x00FFFFFF, 0x00060006,
-	0x00D75FFF, 0x001E0000,
+static const struct ddi_buf_trans hsw_ddi_translations_fdi[] = {
+	{ 0x00FFFFFF, 0x0007000E },
+	{ 0x00D75FFF, 0x000F000A },
+	{ 0x00C30FFF, 0x00060006 },
+	{ 0x00AAAFFF, 0x001E0000 },
+	{ 0x00FFFFFF, 0x000F000A },
+	{ 0x00D75FFF, 0x00160004 },
+	{ 0x00C30FFF, 0x001E0000 },
+	{ 0x00FFFFFF, 0x00060006 },
+	{ 0x00D75FFF, 0x001E0000 },
 };
 
-static const u32 hsw_ddi_translations_hdmi[] = {
-				/* Idx	NT mV diff	T mV diff	db  */
-	0x00FFFFFF, 0x0006000E, /* 0:	400		400		0   */
-	0x00E79FFF, 0x000E000C, /* 1:	400		500		2   */
-	0x00D75FFF, 0x0005000A, /* 2:	400		600		3.5 */
-	0x00FFFFFF, 0x0005000A, /* 3:	600		600		0   */
-	0x00E79FFF, 0x001D0007, /* 4:	600		750		2   */
-	0x00D75FFF, 0x000C0004, /* 5:	600		900		3.5 */
-	0x00FFFFFF, 0x00040006, /* 6:	800		800		0   */
-	0x80E79FFF, 0x00030002, /* 7:	800		1000		2   */
-	0x00FFFFFF, 0x00140005, /* 8:	850		850		0   */
-	0x00FFFFFF, 0x000C0004, /* 9:	900		900		0   */
-	0x00FFFFFF, 0x001C0003, /* 10:	950		950		0   */
-	0x80FFFFFF, 0x00030002, /* 11:	1000		1000		0   */
+static const struct ddi_buf_trans hsw_ddi_translations_hdmi[] = {
+					/* Idx	NT mV d	T mV d	db	*/
+	{ 0x00FFFFFF, 0x0006000E },	/* 0:	400	400	0	*/
+	{ 0x00E79FFF, 0x000E000C },	/* 1:	400	500	2	*/
+	{ 0x00D75FFF, 0x0005000A },	/* 2:	400	600	3.5	*/
+	{ 0x00FFFFFF, 0x0005000A },	/* 3:	600	600	0	*/
+	{ 0x00E79FFF, 0x001D0007 },	/* 4:	600	750	2	*/
+	{ 0x00D75FFF, 0x000C0004 },	/* 5:	600	900	3.5	*/
+	{ 0x00FFFFFF, 0x00040006 },	/* 6:	800	800	0	*/
+	{ 0x80E79FFF, 0x00030002 },	/* 7:	800	1000	2	*/
+	{ 0x00FFFFFF, 0x00140005 },	/* 8:	850	850	0	*/
+	{ 0x00FFFFFF, 0x000C0004 },	/* 9:	900	900	0	*/
+	{ 0x00FFFFFF, 0x001C0003 },	/* 10:	950	950	0	*/
+	{ 0x80FFFFFF, 0x00030002 },	/* 11:	1000	1000	0	*/
 };
 
-static const u32 bdw_ddi_translations_edp[] = {
-	0x00FFFFFF, 0x00000012,		/* eDP parameters */
-	0x00EBAFFF, 0x00020011,
-	0x00C71FFF, 0x0006000F,
-	0x00AAAFFF, 0x000E000A,
-	0x00FFFFFF, 0x00020011,
-	0x00DB6FFF, 0x0005000F,
-	0x00BEEFFF, 0x000A000C,
-	0x00FFFFFF, 0x0005000F,
-	0x00DB6FFF, 0x000A000C,
-	0x00FFFFFF, 0x00140006		/* HDMI parameters 800mV 0dB*/
+static const struct ddi_buf_trans bdw_ddi_translations_edp[] = {
+	{ 0x00FFFFFF, 0x00000012 },
+	{ 0x00EBAFFF, 0x00020011 },
+	{ 0x00C71FFF, 0x0006000F },
+	{ 0x00AAAFFF, 0x000E000A },
+	{ 0x00FFFFFF, 0x00020011 },
+	{ 0x00DB6FFF, 0x0005000F },
+	{ 0x00BEEFFF, 0x000A000C },
+	{ 0x00FFFFFF, 0x0005000F },
+	{ 0x00DB6FFF, 0x000A000C },
 };
 
-static const u32 bdw_ddi_translations_dp[] = {
-	0x00FFFFFF, 0x0007000E,		/* DP parameters */
-	0x00D75FFF, 0x000E000A,
-	0x00BEFFFF, 0x00140006,
-	0x80B2CFFF, 0x001B0002,
-	0x00FFFFFF, 0x000E000A,
-	0x00D75FFF, 0x00180004,
-	0x80CB2FFF, 0x001B0002,
-	0x00F7DFFF, 0x00180004,
-	0x80D75FFF, 0x001B0002,
-	0x00FFFFFF, 0x00140006		/* HDMI parameters 800mV 0dB*/
+static const struct ddi_buf_trans bdw_ddi_translations_dp[] = {
+	{ 0x00FFFFFF, 0x0007000E },
+	{ 0x00D75FFF, 0x000E000A },
+	{ 0x00BEFFFF, 0x00140006 },
+	{ 0x80B2CFFF, 0x001B0002 },
+	{ 0x00FFFFFF, 0x000E000A },
+	{ 0x00D75FFF, 0x00180004 },
+	{ 0x80CB2FFF, 0x001B0002 },
+	{ 0x00F7DFFF, 0x00180004 },
+	{ 0x80D75FFF, 0x001B0002 },
 };
 
-static const u32 bdw_ddi_translations_fdi[] = {
-	0x00FFFFFF, 0x0001000E,		/* FDI parameters */
-	0x00D75FFF, 0x0004000A,
-	0x00C30FFF, 0x00070006,
-	0x00AAAFFF, 0x000C0000,
-	0x00FFFFFF, 0x0004000A,
-	0x00D75FFF, 0x00090004,
-	0x00C30FFF, 0x000C0000,
-	0x00FFFFFF, 0x00070006,
-	0x00D75FFF, 0x000C0000,
-	0x00FFFFFF, 0x00140006		/* HDMI parameters 800mV 0dB*/
+static const struct ddi_buf_trans bdw_ddi_translations_fdi[] = {
+	{ 0x00FFFFFF, 0x0001000E },
+	{ 0x00D75FFF, 0x0004000A },
+	{ 0x00C30FFF, 0x00070006 },
+	{ 0x00AAAFFF, 0x000C0000 },
+	{ 0x00FFFFFF, 0x0004000A },
+	{ 0x00D75FFF, 0x00090004 },
+	{ 0x00C30FFF, 0x000C0000 },
+	{ 0x00FFFFFF, 0x00070006 },
+	{ 0x00D75FFF, 0x000C0000 },
+};
+
+static const struct ddi_buf_trans bdw_ddi_translations_hdmi[] = {
+					/* Idx	NT mV d	T mV df	db	*/
+	{ 0x00FFFFFF, 0x0007000E },	/* 0:	400	400	0	*/
+	{ 0x00D75FFF, 0x000E000A },	/* 1:	400	600	3.5	*/
+	{ 0x00BEFFFF, 0x00140006 },	/* 2:	400	800	6	*/
+	{ 0x00FFFFFF, 0x0009000D },	/* 3:	450	450	0	*/
+	{ 0x00FFFFFF, 0x000E000A },	/* 4:	600	600	0	*/
+	{ 0x00D7FFFF, 0x00140006 },	/* 5:	600	800	2.5	*/
+	{ 0x80CB2FFF, 0x001B0002 },	/* 6:	600	1000	4.5	*/
+	{ 0x00FFFFFF, 0x00140006 },	/* 7:	800	800	0	*/
+	{ 0x80E79FFF, 0x001B0002 },	/* 8:	800	1000	2	*/
+	{ 0x80FFFFFF, 0x001B0002 },	/* 9:	1000	1000	0	*/
 };
 
 enum port intel_ddi_get_encoder_port(struct intel_encoder *intel_encoder)
@@ -145,26 +161,36 @@ static void intel_prepare_ddi_buffers(struct drm_device *dev, enum port port)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 reg;
-	int i;
+	int i, n_hdmi_entries, hdmi_800mV_0dB;
 	int hdmi_level = dev_priv->vbt.ddi_port_info[port].hdmi_level_shift;
-	const u32 *ddi_translations_fdi;
-	const u32 *ddi_translations_dp;
-	const u32 *ddi_translations_edp;
-	const u32 *ddi_translations;
+	const struct ddi_buf_trans *ddi_translations_fdi;
+	const struct ddi_buf_trans *ddi_translations_dp;
+	const struct ddi_buf_trans *ddi_translations_edp;
+	const struct ddi_buf_trans *ddi_translations_hdmi;
+	const struct ddi_buf_trans *ddi_translations;
 
 	if (IS_BROADWELL(dev)) {
 		ddi_translations_fdi = bdw_ddi_translations_fdi;
 		ddi_translations_dp = bdw_ddi_translations_dp;
 		ddi_translations_edp = bdw_ddi_translations_edp;
+		ddi_translations_hdmi = bdw_ddi_translations_hdmi;
+		n_hdmi_entries = ARRAY_SIZE(bdw_ddi_translations_hdmi);
+		hdmi_800mV_0dB = 7;
 	} else if (IS_HASWELL(dev)) {
 		ddi_translations_fdi = hsw_ddi_translations_fdi;
 		ddi_translations_dp = hsw_ddi_translations_dp;
 		ddi_translations_edp = hsw_ddi_translations_dp;
+		ddi_translations_hdmi = hsw_ddi_translations_hdmi;
+		n_hdmi_entries = ARRAY_SIZE(hsw_ddi_translations_hdmi);
+		hdmi_800mV_0dB = 6;
 	} else {
 		WARN(1, "ddi translation table missing\n");
 		ddi_translations_edp = bdw_ddi_translations_dp;
 		ddi_translations_fdi = bdw_ddi_translations_fdi;
 		ddi_translations_dp = bdw_ddi_translations_dp;
+		ddi_translations_hdmi = bdw_ddi_translations_hdmi;
+		n_hdmi_entries = ARRAY_SIZE(bdw_ddi_translations_hdmi);
+		hdmi_800mV_0dB = 7;
 	}
 
 	switch (port) {
@@ -190,14 +216,22 @@ static void intel_prepare_ddi_buffers(struct drm_device *dev, enum port port)
 
 	for (i = 0, reg = DDI_BUF_TRANS(port);
 	     i < ARRAY_SIZE(hsw_ddi_translations_fdi); i++) {
-		I915_WRITE(reg, ddi_translations[i]);
+		I915_WRITE(reg, ddi_translations[i].trans1);
 		reg += 4;
-	}
-	/* Entry 9 is for HDMI: */
-	for (i = 0; i < 2; i++) {
-		I915_WRITE(reg, hsw_ddi_translations_hdmi[hdmi_level * 2 + i]);
+		I915_WRITE(reg, ddi_translations[i].trans2);
 		reg += 4;
 	}
+
+	/* Choose a good default if VBT is badly populated */
+	if (hdmi_level == HDMI_LEVEL_SHIFT_UNKNOWN ||
+	    hdmi_level >= n_hdmi_entries)
+		hdmi_level = hdmi_800mV_0dB;
+
+	/* Entry 9 is for HDMI: */
+	I915_WRITE(reg, ddi_translations_hdmi[hdmi_level].trans1);
+	reg += 4;
+	I915_WRITE(reg, ddi_translations_hdmi[hdmi_level].trans2);
+	reg += 4;
 }
 
 /* Program DDI buffers translations for DP. By default, program ports A-D in DP
@@ -214,18 +248,6 @@ void intel_prepare_ddi(struct drm_device *dev)
 		intel_prepare_ddi_buffers(dev, port);
 }
 
-static const long hsw_ddi_buf_ctl_values[] = {
-	DDI_BUF_EMP_400MV_0DB_HSW,
-	DDI_BUF_EMP_400MV_3_5DB_HSW,
-	DDI_BUF_EMP_400MV_6DB_HSW,
-	DDI_BUF_EMP_400MV_9_5DB_HSW,
-	DDI_BUF_EMP_600MV_0DB_HSW,
-	DDI_BUF_EMP_600MV_3_5DB_HSW,
-	DDI_BUF_EMP_600MV_6DB_HSW,
-	DDI_BUF_EMP_800MV_0DB_HSW,
-	DDI_BUF_EMP_800MV_3_5DB_HSW
-};
-
 static void intel_wait_ddi_buf_idle(struct drm_i915_private *dev_priv,
 				    enum port port)
 {
@@ -285,7 +307,7 @@ void hsw_fdi_link_train(struct drm_crtc *crtc)
 
 	/* Start the training iterating through available voltages and emphasis,
 	 * testing each value twice. */
-	for (i = 0; i < ARRAY_SIZE(hsw_ddi_buf_ctl_values) * 2; i++) {
+	for (i = 0; i < ARRAY_SIZE(hsw_ddi_translations_fdi) * 2; i++) {
 		/* Configure DP_TP_CTL with auto-training */
 		I915_WRITE(DP_TP_CTL(PORT_E),
 					DP_TP_CTL_FDI_AUTOTRAIN |
@@ -300,7 +322,7 @@ void hsw_fdi_link_train(struct drm_crtc *crtc)
 		I915_WRITE(DDI_BUF_CTL(PORT_E),
 			   DDI_BUF_CTL_ENABLE |
 			   ((intel_crtc->config.fdi_lanes - 1) << 1) |
-			   hsw_ddi_buf_ctl_values[i / 2]);
+			   DDI_BUF_TRANS_SELECT(i / 2));
 		POSTING_READ(DDI_BUF_CTL(PORT_E));
 
 		udelay(600);
@@ -375,7 +397,7 @@ void intel_ddi_init_dp_buf_reg(struct intel_encoder *encoder)
 		enc_to_dig_port(&encoder->base);
 
 	intel_dp->DP = intel_dig_port->saved_port_bits |
-		DDI_BUF_CTL_ENABLE | DDI_BUF_EMP_400MV_0DB_HSW;
+		DDI_BUF_CTL_ENABLE | DDI_BUF_TRANS_SELECT(0);
 	intel_dp->DP |= DDI_PORT_WIDTH(intel_dp->lane_count);
 
 }
@@ -402,7 +424,7 @@ intel_ddi_get_crtc_encoder(struct drm_crtc *crtc)
 }
 
 #define LC_FREQ 2700
-#define LC_FREQ_2K (LC_FREQ * 2000)
+#define LC_FREQ_2K U64_C(LC_FREQ * 2000)
 
 #define P_MIN 2
 #define P_MAX 64
@@ -414,7 +436,11 @@ intel_ddi_get_crtc_encoder(struct drm_crtc *crtc)
 #define VCO_MIN 2400
 #define VCO_MAX 4800
 
-#define ABS_DIFF(a, b) ((a > b) ? (a - b) : (b - a))
+#define abs_diff(a, b) ({			\
+	typeof(a) __a = (a);			\
+	typeof(b) __b = (b);			\
+	(void) (&__a == &__b);			\
+	__a > __b ? (__a - __b) : (__b - __a); })
 
 struct wrpll_rnp {
 	unsigned p, n2, r2;
@@ -524,9 +550,9 @@ static void wrpll_update_rnp(uint64_t freq2k, unsigned budget,
 	 */
 	a = freq2k * budget * p * r2;
 	b = freq2k * budget * best->p * best->r2;
-	diff = ABS_DIFF((freq2k * p * r2), (LC_FREQ_2K * n2));
-	diff_best = ABS_DIFF((freq2k * best->p * best->r2),
-			     (LC_FREQ_2K * best->n2));
+	diff = abs_diff(freq2k * p * r2, LC_FREQ_2K * n2);
+	diff_best = abs_diff(freq2k * best->p * best->r2,
+			     LC_FREQ_2K * best->n2);
 	c = 1000000 * diff;
 	d = 1000000 * diff_best;
 
@@ -587,8 +613,8 @@ static int intel_ddi_calc_wrpll_link(struct drm_i915_private *dev_priv,
 	return (refclk * n * 100) / (p * r);
 }
 
-void intel_ddi_clock_get(struct intel_encoder *encoder,
-			 struct intel_crtc_config *pipe_config)
+static void hsw_ddi_clock_get(struct intel_encoder *encoder,
+			      struct intel_crtc_config *pipe_config)
 {
 	struct drm_i915_private *dev_priv = encoder->base.dev->dev_private;
 	int link_clock = 0;
@@ -643,9 +669,15 @@ void intel_ddi_clock_get(struct intel_encoder *encoder,
 		pipe_config->adjusted_mode.crtc_clock = pipe_config->port_clock;
 }
 
+void intel_ddi_clock_get(struct intel_encoder *encoder,
+			 struct intel_crtc_config *pipe_config)
+{
+	hsw_ddi_clock_get(encoder, pipe_config);
+}
+
 static void
-intel_ddi_calculate_wrpll(int clock /* in Hz */,
-			  unsigned *r2_out, unsigned *n2_out, unsigned *p_out)
+hsw_ddi_calculate_wrpll(int clock /* in Hz */,
+			unsigned *r2_out, unsigned *n2_out, unsigned *p_out)
 {
 	uint64_t freq2k;
 	unsigned p, n2, r2;
@@ -708,27 +740,17 @@ intel_ddi_calculate_wrpll(int clock /* in Hz */,
 	*r2_out = best.r2;
 }
 
-/*
- * Tries to find a PLL for the CRTC. If it finds, it increases the refcount and
- * stores it in intel_crtc->ddi_pll_sel, so other mode sets won't be able to
- * steal the selected PLL. You need to call intel_ddi_pll_enable to actually
- * enable the PLL.
- */
-bool intel_ddi_pll_select(struct intel_crtc *intel_crtc)
+static bool
+hsw_ddi_pll_select(struct intel_crtc *intel_crtc,
+		   struct intel_encoder *intel_encoder,
+		   int clock)
 {
-	struct drm_crtc *crtc = &intel_crtc->base;
-	struct intel_encoder *intel_encoder = intel_ddi_get_crtc_encoder(crtc);
-	int type = intel_encoder->type;
-	int clock = intel_crtc->config.port_clock;
-
-	intel_put_shared_dpll(intel_crtc);
-
-	if (type == INTEL_OUTPUT_HDMI) {
+	if (intel_encoder->type == INTEL_OUTPUT_HDMI) {
 		struct intel_shared_dpll *pll;
 		uint32_t val;
 		unsigned p, n2, r2;
 
-		intel_ddi_calculate_wrpll(clock * 1000, &r2, &n2, &p);
+		hsw_ddi_calculate_wrpll(clock * 1000, &r2, &n2, &p);
 
 		val = WRPLL_PLL_ENABLE | WRPLL_PLL_LCPLL |
 		      WRPLL_DIVIDER_REFERENCE(r2) | WRPLL_DIVIDER_FEEDBACK(n2) |
@@ -749,6 +771,25 @@ bool intel_ddi_pll_select(struct intel_crtc *intel_crtc)
 	return true;
 }
 
+
+/*
+ * Tries to find a *shared* PLL for the CRTC and store it in
+ * intel_crtc->ddi_pll_sel.
+ *
+ * For private DPLLs, compute_config() should do the selection for us. This
+ * function should be folded into compute_config() eventually.
+ */
+bool intel_ddi_pll_select(struct intel_crtc *intel_crtc)
+{
+	struct drm_crtc *crtc = &intel_crtc->base;
+	struct intel_encoder *intel_encoder = intel_ddi_get_crtc_encoder(crtc);
+	int clock = intel_crtc->config.port_clock;
+
+	intel_put_shared_dpll(intel_crtc);
+
+	return hsw_ddi_pll_select(intel_crtc, intel_encoder, clock);
+}
+
 void intel_ddi_set_pipe_settings(struct drm_crtc *crtc)
 {
 	struct drm_i915_private *dev_priv = crtc->dev->dev_private;
@@ -1183,31 +1224,52 @@ static void intel_disable_ddi(struct intel_encoder *intel_encoder)
 	}
 }
 
-int intel_ddi_get_cdclk_freq(struct drm_i915_private *dev_priv)
+static int bdw_get_cdclk_freq(struct drm_i915_private *dev_priv)
+{
+	uint32_t lcpll = I915_READ(LCPLL_CTL);
+	uint32_t freq = lcpll & LCPLL_CLK_FREQ_MASK;
+
+	if (lcpll & LCPLL_CD_SOURCE_FCLK)
+		return 800000;
+	else if (I915_READ(FUSE_STRAP) & HSW_CDCLK_LIMIT)
+		return 450000;
+	else if (freq == LCPLL_CLK_FREQ_450)
+		return 450000;
+	else if (freq == LCPLL_CLK_FREQ_54O_BDW)
+		return 540000;
+	else if (freq == LCPLL_CLK_FREQ_337_5_BDW)
+		return 337500;
+	else
+		return 675000;
+}
+
+static int hsw_get_cdclk_freq(struct drm_i915_private *dev_priv)
 {
 	struct drm_device *dev = dev_priv->dev;
 	uint32_t lcpll = I915_READ(LCPLL_CTL);
 	uint32_t freq = lcpll & LCPLL_CLK_FREQ_MASK;
 
-	if (lcpll & LCPLL_CD_SOURCE_FCLK) {
+	if (lcpll & LCPLL_CD_SOURCE_FCLK)
 		return 800000;
-	} else if (I915_READ(FUSE_STRAP) & HSW_CDCLK_LIMIT) {
+	else if (I915_READ(FUSE_STRAP) & HSW_CDCLK_LIMIT)
 		return 450000;
-	} else if (freq == LCPLL_CLK_FREQ_450) {
+	else if (freq == LCPLL_CLK_FREQ_450)
 		return 450000;
-	} else if (IS_HASWELL(dev)) {
-		if (IS_ULT(dev))
-			return 337500;
-		else
-			return 540000;
-	} else {
-		if (freq == LCPLL_CLK_FREQ_54O_BDW)
-			return 540000;
-		else if (freq == LCPLL_CLK_FREQ_337_5_BDW)
-			return 337500;
-		else
-			return 675000;
-	}
+	else if (IS_ULT(dev))
+		return 337500;
+	else
+		return 540000;
+}
+
+int intel_ddi_get_cdclk_freq(struct drm_i915_private *dev_priv)
+{
+	struct drm_device *dev = dev_priv->dev;
+
+	if (IS_BROADWELL(dev))
+		return bdw_get_cdclk_freq(dev_priv);
+
+	/* Haswell */
+	return hsw_get_cdclk_freq(dev_priv);
 }
 
 static void hsw_ddi_pll_enable(struct drm_i915_private *dev_priv,
@@ -1248,10 +1310,8 @@ static const char * const hsw_ddi_pll_names[] = {
 	"WRPLL 2",
 };
 
-void intel_ddi_pll_init(struct drm_device *dev)
+static void hsw_shared_dplls_init(struct drm_i915_private *dev_priv)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	uint32_t val = I915_READ(LCPLL_CTL);
 	int i;
 
 	dev_priv->num_shared_dpll = 2;
@@ -1264,6 +1324,14 @@ void intel_ddi_pll_init(struct drm_device *dev)
 		dev_priv->shared_dplls[i].get_hw_state =
 			hsw_ddi_pll_get_hw_state;
 	}
+}
+
+void intel_ddi_pll_init(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	uint32_t val = I915_READ(LCPLL_CTL);
+
+	hsw_shared_dplls_init(dev_priv);
 
 	/* The LCPLL register should be turned on by the BIOS. For now let's
 	 * just check its state and print errors in case something is wrong.
@@ -1444,7 +1512,7 @@ void intel_ddi_get_config(struct intel_encoder *encoder,
 		dev_priv->vbt.edp_bpp = pipe_config->pipe_bpp;
 	}
 
-	intel_ddi_clock_get(encoder, pipe_config);
+	hsw_ddi_clock_get(encoder, pipe_config);
 }
 
 static void intel_ddi_destroy(struct drm_encoder *encoder)
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index d8324c69fa86..507370513f3d 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -91,15 +91,16 @@ static int intel_framebuffer_init(struct drm_device *dev,
 				  struct intel_framebuffer *ifb,
 				  struct drm_mode_fb_cmd2 *mode_cmd,
 				  struct drm_i915_gem_object *obj);
-static void intel_dp_set_m_n(struct intel_crtc *crtc);
 static void i9xx_set_pipeconf(struct intel_crtc *intel_crtc);
 static void intel_set_pipe_timings(struct intel_crtc *intel_crtc);
 static void intel_cpu_transcoder_set_m_n(struct intel_crtc *crtc,
-					 struct intel_link_m_n *m_n);
+					 struct intel_link_m_n *m_n,
+					 struct intel_link_m_n *m2_n2);
 static void ironlake_set_pipeconf(struct drm_crtc *crtc);
 static void haswell_set_pipeconf(struct drm_crtc *crtc);
 static void intel_set_pipe_csc(struct drm_crtc *crtc);
 static void vlv_prepare_pll(struct intel_crtc *crtc);
+static void chv_prepare_pll(struct intel_crtc *crtc);
 
 static struct intel_encoder *intel_find_encoder(struct intel_connector *connector, int pipe)
 {
@@ -899,7 +900,8 @@ static void g4x_wait_for_vblank(struct drm_device *dev, int pipe)
 	frame = I915_READ(frame_reg);
 
 	if (wait_for(I915_READ_NOTRACE(frame_reg) != frame, 50))
-		WARN(1, "vblank wait timed out\n");
+		WARN(1, "vblank wait on pipe %c timed out\n",
+		     pipe_name(pipe));
 }
 
 /**
@@ -940,7 +942,8 @@ void intel_wait_for_vblank(struct drm_device *dev, int pipe)
 	if (wait_for(I915_READ(pipestat_reg) &
 		     PIPE_VBLANK_INTERRUPT_STATUS,
 		     50))
-		DRM_DEBUG_KMS("vblank wait timed out\n");
+		DRM_DEBUG_KMS("vblank wait on pipe %c timed out\n",
+			      pipe_name(pipe));
 }
 
 static bool pipe_dsl_stopped(struct drm_device *dev, enum pipe pipe)
@@ -964,8 +967,7 @@ static bool pipe_dsl_stopped(struct drm_device *dev, enum pipe pipe)
 
 /*
  * intel_wait_for_pipe_off - wait for pipe to turn off
- * @dev: drm device
- * @pipe: pipe to wait for
+ * @crtc: crtc whose pipe to wait for
  *
  * After disabling a pipe, we can't wait for vblank in the usual way,
  * spinning on the vblank interrupt status bit, since we won't actually
@@ -979,11 +981,12 @@ static bool pipe_dsl_stopped(struct drm_device *dev, enum pipe pipe)
  *   ends up stopping at the start of the next frame).
  *
  */
-void intel_wait_for_pipe_off(struct drm_device *dev, int pipe)
+static void intel_wait_for_pipe_off(struct intel_crtc *crtc)
 {
+	struct drm_device *dev = crtc->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	enum transcoder cpu_transcoder = intel_pipe_to_cpu_transcoder(dev_priv,
-								      pipe);
+	enum transcoder cpu_transcoder = crtc->config.cpu_transcoder;
+	enum pipe pipe = crtc->pipe;
 
 	if (INTEL_INFO(dev)->gen >= 4) {
 		int reg = PIPECONF(cpu_transcoder);
@@ -1192,27 +1195,40 @@ void assert_fdi_rx_pll(struct drm_i915_private *dev_priv,
 static void assert_panel_unlocked(struct drm_i915_private *dev_priv,
 				  enum pipe pipe)
 {
-	int pp_reg, lvds_reg;
+	struct drm_device *dev = dev_priv->dev;
+	int pp_reg;
 	u32 val;
 	enum pipe panel_pipe = PIPE_A;
 	bool locked = true;
 
-	if (HAS_PCH_SPLIT(dev_priv->dev)) {
+	if (WARN_ON(HAS_DDI(dev)))
+		return;
+
+	if (HAS_PCH_SPLIT(dev)) {
+		u32 port_sel;
+
 		pp_reg = PCH_PP_CONTROL;
-		lvds_reg = PCH_LVDS;
+		port_sel = I915_READ(PCH_PP_ON_DELAYS) & PANEL_PORT_SELECT_MASK;
+
+		if (port_sel == PANEL_PORT_SELECT_LVDS &&
+		    I915_READ(PCH_LVDS) & LVDS_PIPEB_SELECT)
+			panel_pipe = PIPE_B;
+		/* XXX: else fix for eDP */
+	} else if (IS_VALLEYVIEW(dev)) {
+		/* presumably write lock depends on pipe, not port select */
+		pp_reg = VLV_PIPE_PP_CONTROL(pipe);
+		panel_pipe = pipe;
 	} else {
 		pp_reg = PP_CONTROL;
-		lvds_reg = LVDS;
+		if (I915_READ(LVDS) & LVDS_PIPEB_SELECT)
+			panel_pipe = PIPE_B;
 	}
 
 	val = I915_READ(pp_reg);
 	if (!(val & PANEL_POWER_ON) ||
-	    ((val & PANEL_UNLOCK_REGS) == PANEL_UNLOCK_REGS))
+	    ((val & PANEL_UNLOCK_MASK) == PANEL_UNLOCK_REGS))
 		locked = false;
 
-	if (I915_READ(lvds_reg) & LVDS_PIPEB_SELECT)
-		panel_pipe = PIPE_B;
-
 	WARN(panel_pipe == pipe && locked,
 	     "panel assertion failure, pipe %c regs locked\n",
 	     pipe_name(pipe));
@@ -1245,8 +1261,9 @@ void assert_pipe(struct drm_i915_private *dev_priv,
 	enum transcoder cpu_transcoder = intel_pipe_to_cpu_transcoder(dev_priv,
 								      pipe);
 
-	/* if we need the pipe A quirk it must be always on */
-	if (pipe == PIPE_A && dev_priv->quirks & QUIRK_PIPEA_FORCE)
+	/* if we need the pipe quirk it must be always on */
+	if ((pipe == PIPE_A && dev_priv->quirks & QUIRK_PIPEA_FORCE) ||
+	    (pipe == PIPE_B && dev_priv->quirks & QUIRK_PIPEB_FORCE))
 		state = true;
 
 	if (!intel_display_power_enabled(dev_priv,
@@ -1300,7 +1317,7 @@ static void assert_planes_disabled(struct drm_i915_private *dev_priv,
 	}
 
 	/* Need to check both planes against the pipe */
-	for_each_pipe(i) {
+	for_each_pipe(dev_priv, i) {
 		reg = DSPCNTR(i);
 		val = I915_READ(reg);
 		cur_pipe = (val & DISPPLANE_SEL_PIPE_MASK) >>
@@ -1341,6 +1358,12 @@ static void assert_sprites_disabled(struct drm_i915_private *dev_priv,
 	}
 }
 
+static void assert_vblank_disabled(struct drm_crtc *crtc)
+{
+	if (WARN_ON(drm_crtc_vblank_get(crtc) == 0))
+		drm_crtc_vblank_put(crtc);
+}
+
 static void ibx_assert_pch_refclk_enabled(struct drm_i915_private *dev_priv)
 {
 	u32 val;
@@ -1513,34 +1536,6 @@ static void intel_init_dpio(struct drm_device *dev)
 	}
 }
 
-static void intel_reset_dpio(struct drm_device *dev)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-
-	if (IS_CHERRYVIEW(dev)) {
-		enum dpio_phy phy;
-		u32 val;
-
-		for (phy = DPIO_PHY0; phy < I915_NUM_PHYS_VLV; phy++) {
-			/* Poll for phypwrgood signal */
-			if (wait_for(I915_READ(DISPLAY_PHY_STATUS) &
-						PHY_POWERGOOD(phy), 1))
-				DRM_ERROR("Display PHY %d is not power up\n", phy);
-
-			/*
-			 * Deassert common lane reset for PHY.
-			 *
-			 * This should only be done on init and resume from S3
-			 * with both PLLs disabled, or we risk losing DPIO and
-			 * PLL synchronization.
-			 */
-			val = I915_READ(DISPLAY_PHY_CONTROL);
-			I915_WRITE(DISPLAY_PHY_CONTROL,
-				PHY_COM_LANE_RESET_DEASSERT(phy, val));
-		}
-	}
-}
-
 static void vlv_enable_pll(struct intel_crtc *crtc)
 {
 	struct drm_device *dev = crtc->base.dev;
@@ -1554,7 +1549,7 @@ static void vlv_enable_pll(struct intel_crtc *crtc)
 	BUG_ON(!IS_VALLEYVIEW(dev_priv->dev));
 
 	/* PLL is protected by panel, make sure we can write it */
-	if (IS_MOBILE(dev_priv->dev) && !IS_I830(dev_priv->dev))
+	if (IS_MOBILE(dev_priv->dev))
 		assert_panel_unlocked(dev_priv, crtc->pipe);
 
 	I915_WRITE(reg, dpll);
@@ -1617,6 +1612,18 @@ static void chv_enable_pll(struct intel_crtc *crtc)
 	mutex_unlock(&dev_priv->dpio_lock);
 }
 
+static int intel_num_dvo_pipes(struct drm_device *dev)
+{
+	struct intel_crtc *crtc;
+	int count = 0;
+
+	for_each_intel_crtc(dev, crtc)
+		count += crtc->active &&
+			intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_DVO);
+
+	return count;
+}
+
 static void i9xx_enable_pll(struct intel_crtc *crtc)
 {
 	struct drm_device *dev = crtc->base.dev;
@@ -1633,7 +1640,18 @@ static void i9xx_enable_pll(struct intel_crtc *crtc)
 	if (IS_MOBILE(dev) && !IS_I830(dev))
 		assert_panel_unlocked(dev_priv, crtc->pipe);
 
-	I915_WRITE(reg, dpll);
+	/* Enable DVO 2x clock on both PLLs if necessary */
+	if (IS_I830(dev) && intel_num_dvo_pipes(dev) > 0) {
+		/*
+		 * It appears to be important that we don't enable this
+		 * for the current pipe before otherwise configuring the
+		 * PLL. No idea how this should be handled if multiple
+		 * DVO outputs are enabled simultaneosly.
+		 */
+		dpll |= DPLL_DVO_2X_MODE;
+		I915_WRITE(DPLL(!crtc->pipe),
+			   I915_READ(DPLL(!crtc->pipe)) | DPLL_DVO_2X_MODE);
+	}
 
 	/* Wait for the clocks to stabilize. */
 	POSTING_READ(reg);
@@ -1672,10 +1690,25 @@ static void i9xx_enable_pll(struct intel_crtc *crtc)
  *
  * Note!  This is for pre-ILK only.
  */
-static void i9xx_disable_pll(struct drm_i915_private *dev_priv, enum pipe pipe)
+static void i9xx_disable_pll(struct intel_crtc *crtc)
 {
-	/* Don't disable pipe A or pipe A PLLs if needed */
-	if (pipe == PIPE_A && (dev_priv->quirks & QUIRK_PIPEA_FORCE))
+	struct drm_device *dev = crtc->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	enum pipe pipe = crtc->pipe;
+
+	/* Disable DVO 2x clock on both PLLs if necessary */
+	if (IS_I830(dev) &&
+	    intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_DVO) &&
+	    intel_num_dvo_pipes(dev) == 1) {
+		I915_WRITE(DPLL(PIPE_B),
+			   I915_READ(DPLL(PIPE_B)) & ~DPLL_DVO_2X_MODE);
+		I915_WRITE(DPLL(PIPE_A),
+			   I915_READ(DPLL(PIPE_A)) & ~DPLL_DVO_2X_MODE);
+	}
+
+	/* Don't disable pipe or pipe PLLs if needed */
+	if ((pipe == PIPE_A && dev_priv->quirks & QUIRK_PIPEA_FORCE) ||
+	    (pipe == PIPE_B && dev_priv->quirks & QUIRK_PIPEB_FORCE))
 		return;
 
 	/* Make sure the pipe isn't still relying on us */
@@ -1712,7 +1745,7 @@ static void chv_disable_pll(struct drm_i915_private *dev_priv, enum pipe pipe)
 	assert_pipe_disabled(dev_priv, pipe);
 
 	/* Set PLL en = 0 */
-	val = DPLL_SSC_REF_CLOCK_CHV;
+	val = DPLL_SSC_REF_CLOCK_CHV | DPLL_REFA_CLK_ENABLE_VLV;
 	if (pipe != PIPE_A)
 		val |= DPLL_INTEGRATED_CRI_CLK_VLV;
 	I915_WRITE(DPLL(pipe), val);
@@ -1806,7 +1839,7 @@ static void intel_enable_shared_dpll(struct intel_crtc *crtc)
 	if (WARN_ON(pll->refcount == 0))
 		return;
 
-	DRM_DEBUG_KMS("enable %s (active %d, on? %d)for crtc %d\n",
+	DRM_DEBUG_KMS("enable %s (active %d, on? %d) for crtc %d\n",
 		      pll->name, pll->active, pll->on,
 		      crtc->base.base.id);
 
@@ -1824,7 +1857,7 @@ static void intel_enable_shared_dpll(struct intel_crtc *crtc)
 	pll->on = true;
 }
 
-void intel_disable_shared_dpll(struct intel_crtc *crtc)
+static void intel_disable_shared_dpll(struct intel_crtc *crtc)
 {
 	struct drm_device *dev = crtc->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -1868,7 +1901,7 @@ static void ironlake_enable_pch_transcoder(struct drm_i915_private *dev_priv,
 	uint32_t reg, val, pipeconf_val;
 
 	/* PCH only available on ILK+ */
-	BUG_ON(INTEL_INFO(dev)->gen < 5);
+	BUG_ON(!HAS_PCH_SPLIT(dev));
 
 	/* Make sure PCH DPLL is enabled */
 	assert_shared_dpll_enabled(dev_priv,
@@ -1921,7 +1954,7 @@ static void lpt_enable_pch_transcoder(struct drm_i915_private *dev_priv,
 	u32 val, pipeconf_val;
 
 	/* PCH only available on ILK+ */
-	BUG_ON(INTEL_INFO(dev_priv->dev)->gen < 5);
+	BUG_ON(!HAS_PCH_SPLIT(dev_priv->dev));
 
 	/* FDI must be feeding us bits for PCH ports */
 	assert_fdi_tx_enabled(dev_priv, (enum pipe) cpu_transcoder);
@@ -2043,8 +2076,8 @@ static void intel_enable_pipe(struct intel_crtc *crtc)
 	reg = PIPECONF(cpu_transcoder);
 	val = I915_READ(reg);
 	if (val & PIPECONF_ENABLE) {
-		WARN_ON(!(pipe == PIPE_A &&
-			  dev_priv->quirks & QUIRK_PIPEA_FORCE));
+		WARN_ON(!((pipe == PIPE_A && dev_priv->quirks & QUIRK_PIPEA_FORCE) ||
+			  (pipe == PIPE_B && dev_priv->quirks & QUIRK_PIPEB_FORCE)));
 		return;
 	}
 
@@ -2054,21 +2087,19 @@ static void intel_enable_pipe(struct intel_crtc *crtc)
 
 /**
  * intel_disable_pipe - disable a pipe, asserting requirements
- * @dev_priv: i915 private structure
- * @pipe: pipe to disable
+ * @crtc: crtc whose pipes is to be disabled
  *
- * Disable @pipe, making sure that various hardware specific requirements
- * are met, if applicable, e.g. plane disabled, panel fitter off, etc.
- *
- * @pipe should be %PIPE_A or %PIPE_B.
+ * Disable the pipe of @crtc, making sure that various hardware
+ * specific requirements are met, if applicable, e.g. plane
+ * disabled, panel fitter off, etc.
  *
  * Will wait until the pipe has shut down before returning.
  */
-static void intel_disable_pipe(struct drm_i915_private *dev_priv,
-			       enum pipe pipe)
+static void intel_disable_pipe(struct intel_crtc *crtc)
 {
-	enum transcoder cpu_transcoder = intel_pipe_to_cpu_transcoder(dev_priv,
-								      pipe);
+	struct drm_i915_private *dev_priv = crtc->base.dev->dev_private;
+	enum transcoder cpu_transcoder = crtc->config.cpu_transcoder;
+	enum pipe pipe = crtc->pipe;
 	int reg;
 	u32 val;
 
@@ -2080,17 +2111,26 @@ static void intel_disable_pipe(struct drm_i915_private *dev_priv,
 	assert_cursor_disabled(dev_priv, pipe);
 	assert_sprites_disabled(dev_priv, pipe);
 
-	/* Don't disable pipe A or pipe A PLLs if needed */
-	if (pipe == PIPE_A && (dev_priv->quirks & QUIRK_PIPEA_FORCE))
-		return;
-
 	reg = PIPECONF(cpu_transcoder);
 	val = I915_READ(reg);
 	if ((val & PIPECONF_ENABLE) == 0)
 		return;
 
-	I915_WRITE(reg, val & ~PIPECONF_ENABLE);
-	intel_wait_for_pipe_off(dev_priv->dev, pipe);
+	/*
+	 * Double wide has implications for planes
+	 * so best keep it disabled when not needed.
+	 */
+	if (crtc->config.double_wide)
+		val &= ~PIPECONF_DOUBLE_WIDE;
+
+	/* Don't disable pipe or pipe PLLs if needed */
+	if (!(pipe == PIPE_A && dev_priv->quirks & QUIRK_PIPEA_FORCE) &&
+	    !(pipe == PIPE_B && dev_priv->quirks & QUIRK_PIPEB_FORCE))
+		val &= ~PIPECONF_ENABLE;
+
+	I915_WRITE(reg, val);
+	if ((val & PIPECONF_ENABLE) == 0)
+		intel_wait_for_pipe_off(crtc);
 }
 
 /*
@@ -2109,35 +2149,28 @@ void intel_flush_primary_plane(struct drm_i915_private *dev_priv,
 
 /**
  * intel_enable_primary_hw_plane - enable the primary plane on a given pipe
- * @dev_priv: i915 private structure
- * @plane: plane to enable
- * @pipe: pipe being fed
+ * @plane:  plane to be enabled
+ * @crtc: crtc for the plane
  *
- * Enable @plane on @pipe, making sure that @pipe is running first.
+ * Enable @plane on @crtc, making sure that the pipe is running first.
  */
-static void intel_enable_primary_hw_plane(struct drm_i915_private *dev_priv,
-					  enum plane plane, enum pipe pipe)
+static void intel_enable_primary_hw_plane(struct drm_plane *plane,
+					  struct drm_crtc *crtc)
 {
-	struct drm_device *dev = dev_priv->dev;
-	struct intel_crtc *intel_crtc =
-		to_intel_crtc(dev_priv->pipe_to_crtc_mapping[pipe]);
-	int reg;
-	u32 val;
+	struct drm_device *dev = plane->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 
 	/* If the pipe isn't enabled, we can't pump pixels and may hang */
-	assert_pipe_enabled(dev_priv, pipe);
+	assert_pipe_enabled(dev_priv, intel_crtc->pipe);
 
 	if (intel_crtc->primary_enabled)
 		return;
 
 	intel_crtc->primary_enabled = true;
 
-	reg = DSPCNTR(plane);
-	val = I915_READ(reg);
-	WARN_ON(val & DISPLAY_PLANE_ENABLE);
-
-	I915_WRITE(reg, val | DISPLAY_PLANE_ENABLE);
-	intel_flush_primary_plane(dev_priv, plane);
+	dev_priv->display.update_primary_plane(crtc, plane->fb,
+					       crtc->x, crtc->y);
 
 	/*
 	 * BDW signals flip done immediately if the plane
@@ -2150,31 +2183,27 @@ static void intel_enable_primary_hw_plane(struct drm_i915_private *dev_priv,
 
 /**
  * intel_disable_primary_hw_plane - disable the primary hardware plane
- * @dev_priv: i915 private structure
- * @plane: plane to disable
- * @pipe: pipe consuming the data
+ * @plane: plane to be disabled
+ * @crtc: crtc for the plane
  *
- * Disable @plane; should be an independent operation.
+ * Disable @plane on @crtc, making sure that the pipe is running first.
  */
-static void intel_disable_primary_hw_plane(struct drm_i915_private *dev_priv,
-					   enum plane plane, enum pipe pipe)
+static void intel_disable_primary_hw_plane(struct drm_plane *plane,
+					   struct drm_crtc *crtc)
 {
-	struct intel_crtc *intel_crtc =
-		to_intel_crtc(dev_priv->pipe_to_crtc_mapping[pipe]);
-	int reg;
-	u32 val;
+	struct drm_device *dev = plane->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+
+	assert_pipe_enabled(dev_priv, intel_crtc->pipe);
 
 	if (!intel_crtc->primary_enabled)
 		return;
 
 	intel_crtc->primary_enabled = false;
 
-	reg = DSPCNTR(plane);
-	val = I915_READ(reg);
-	WARN_ON((val & DISPLAY_PLANE_ENABLE) == 0);
-
-	I915_WRITE(reg, val & ~DISPLAY_PLANE_ENABLE);
-	intel_flush_primary_plane(dev_priv, plane);
+	dev_priv->display.update_primary_plane(crtc, plane->fb,
+					       crtc->x, crtc->y);
 }
 
 static bool need_vtd_wa(struct drm_device *dev)
@@ -2422,16 +2451,46 @@ static void i9xx_update_primary_plane(struct drm_crtc *crtc,
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	struct drm_i915_gem_object *obj = intel_fb_obj(fb);
+	struct drm_i915_gem_object *obj;
 	int plane = intel_crtc->plane;
 	unsigned long linear_offset;
 	u32 dspcntr;
-	u32 reg;
+	u32 reg = DSPCNTR(plane);
+	int pixel_size;
+
+	if (!intel_crtc->primary_enabled) {
+		I915_WRITE(reg, 0);
+		if (INTEL_INFO(dev)->gen >= 4)
+			I915_WRITE(DSPSURF(plane), 0);
+		else
+			I915_WRITE(DSPADDR(plane), 0);
+		POSTING_READ(reg);
+		return;
+	}
+
+	obj = intel_fb_obj(fb);
+	if (WARN_ON(obj == NULL))
+		return;
+
+	pixel_size = drm_format_plane_cpp(fb->pixel_format, 0);
+
+	dspcntr = DISPPLANE_GAMMA_ENABLE;
+
+	dspcntr |= DISPLAY_PLANE_ENABLE;
+
+	if (INTEL_INFO(dev)->gen < 4) {
+		if (intel_crtc->pipe == PIPE_B)
+			dspcntr |= DISPPLANE_SEL_PIPE_B;
+
+		/* pipesrc and dspsize control the size that is scaled from,
+		 * which should always be the user's requested size.
+		 */
+		I915_WRITE(DSPSIZE(plane),
+			   ((intel_crtc->config.pipe_src_h - 1) << 16) |
+			   (intel_crtc->config.pipe_src_w - 1));
+		I915_WRITE(DSPPOS(plane), 0);
+	}
 
-	reg = DSPCNTR(plane);
-	dspcntr = I915_READ(reg);
-	/* Mask out pixel format bits in case we change it */
-	dspcntr &= ~DISPPLANE_PIXFORMAT_MASK;
 	switch (fb->pixel_format) {
 	case DRM_FORMAT_C8:
 		dspcntr |= DISPPLANE_8BPP;
@@ -2463,30 +2522,40 @@ static void i9xx_update_primary_plane(struct drm_crtc *crtc,
 		BUG();
 	}
 
-	if (INTEL_INFO(dev)->gen >= 4) {
-		if (obj->tiling_mode != I915_TILING_NONE)
-			dspcntr |= DISPPLANE_TILED;
-		else
-			dspcntr &= ~DISPPLANE_TILED;
-	}
+	if (INTEL_INFO(dev)->gen >= 4 &&
+	    obj->tiling_mode != I915_TILING_NONE)
+		dspcntr |= DISPPLANE_TILED;
 
 	if (IS_G4X(dev))
 		dspcntr |= DISPPLANE_TRICKLE_FEED_DISABLE;
 
-	I915_WRITE(reg, dspcntr);
-
-	linear_offset = y * fb->pitches[0] + x * (fb->bits_per_pixel / 8);
+	linear_offset = y * fb->pitches[0] + x * pixel_size;
 
 	if (INTEL_INFO(dev)->gen >= 4) {
 		intel_crtc->dspaddr_offset =
 			intel_gen4_compute_page_offset(&x, &y, obj->tiling_mode,
-						       fb->bits_per_pixel / 8,
+						       pixel_size,
 						       fb->pitches[0]);
 		linear_offset -= intel_crtc->dspaddr_offset;
 	} else {
 		intel_crtc->dspaddr_offset = linear_offset;
 	}
 
+	if (to_intel_plane(crtc->primary)->rotation == BIT(DRM_ROTATE_180)) {
+		dspcntr |= DISPPLANE_ROTATE_180;
+
+		x += (intel_crtc->config.pipe_src_w - 1);
+		y += (intel_crtc->config.pipe_src_h - 1);
+
+		/* Finding the last pixel of the last line of the display
+		data and adding to linear_offset*/
+		linear_offset +=
+			(intel_crtc->config.pipe_src_h - 1) * fb->pitches[0] +
+			(intel_crtc->config.pipe_src_w - 1) * pixel_size;
+	}
+
+	I915_WRITE(reg, dspcntr);
+
 	DRM_DEBUG_KMS("Writing base %08lX %08lX %d %d %d\n",
 		      i915_gem_obj_ggtt_offset(obj), linear_offset, x, y,
 		      fb->pitches[0]);
@@ -2508,16 +2577,33 @@ static void ironlake_update_primary_plane(struct drm_crtc *crtc,
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	struct drm_i915_gem_object *obj = intel_fb_obj(fb);
+	struct drm_i915_gem_object *obj;
 	int plane = intel_crtc->plane;
 	unsigned long linear_offset;
 	u32 dspcntr;
-	u32 reg;
+	u32 reg = DSPCNTR(plane);
+	int pixel_size;
+
+	if (!intel_crtc->primary_enabled) {
+		I915_WRITE(reg, 0);
+		I915_WRITE(DSPSURF(plane), 0);
+		POSTING_READ(reg);
+		return;
+	}
+
+	obj = intel_fb_obj(fb);
+	if (WARN_ON(obj == NULL))
+		return;
+
+	pixel_size = drm_format_plane_cpp(fb->pixel_format, 0);
+
+	dspcntr = DISPPLANE_GAMMA_ENABLE;
+
+	dspcntr |= DISPLAY_PLANE_ENABLE;
+
+	if (IS_HASWELL(dev) || IS_BROADWELL(dev))
+		dspcntr |= DISPPLANE_PIPE_CSC_ENABLE;
 
-	reg = DSPCNTR(plane);
-	dspcntr = I915_READ(reg);
-	/* Mask out pixel format bits in case we change it */
-	dspcntr &= ~DISPPLANE_PIXFORMAT_MASK;
 	switch (fb->pixel_format) {
 	case DRM_FORMAT_C8:
 		dspcntr |= DISPPLANE_8BPP;
@@ -2547,22 +2633,32 @@ static void ironlake_update_primary_plane(struct drm_crtc *crtc,
 
 	if (obj->tiling_mode != I915_TILING_NONE)
 		dspcntr |= DISPPLANE_TILED;
-	else
-		dspcntr &= ~DISPPLANE_TILED;
 
-	if (IS_HASWELL(dev) || IS_BROADWELL(dev))
-		dspcntr &= ~DISPPLANE_TRICKLE_FEED_DISABLE;
-	else
+	if (!IS_HASWELL(dev) && !IS_BROADWELL(dev))
 		dspcntr |= DISPPLANE_TRICKLE_FEED_DISABLE;
 
-	I915_WRITE(reg, dspcntr);
-
-	linear_offset = y * fb->pitches[0] + x * (fb->bits_per_pixel / 8);
+	linear_offset = y * fb->pitches[0] + x * pixel_size;
 	intel_crtc->dspaddr_offset =
 		intel_gen4_compute_page_offset(&x, &y, obj->tiling_mode,
-					       fb->bits_per_pixel / 8,
+					       pixel_size,
 					       fb->pitches[0]);
 	linear_offset -= intel_crtc->dspaddr_offset;
+	if (to_intel_plane(crtc->primary)->rotation == BIT(DRM_ROTATE_180)) {
+		dspcntr |= DISPPLANE_ROTATE_180;
+
+		if (!IS_HASWELL(dev) && !IS_BROADWELL(dev)) {
+			x += (intel_crtc->config.pipe_src_w - 1);
+			y += (intel_crtc->config.pipe_src_h - 1);
+
+			/* Finding the last pixel of the last line of the display
+			data and adding to linear_offset*/
+			linear_offset +=
+				(intel_crtc->config.pipe_src_h - 1) * fb->pitches[0] +
+				(intel_crtc->config.pipe_src_w - 1) * pixel_size;
+		}
+	}
+
+	I915_WRITE(reg, dspcntr);
 
 	DRM_DEBUG_KMS("Writing base %08lX %08lX %d %d %d\n",
 		      i915_gem_obj_ggtt_offset(obj), linear_offset, x, y,
@@ -3346,23 +3442,54 @@ bool intel_has_pending_fb_unpin(struct drm_device *dev)
 	return false;
 }
 
+static void page_flip_completed(struct intel_crtc *intel_crtc)
+{
+	struct drm_i915_private *dev_priv = to_i915(intel_crtc->base.dev);
+	struct intel_unpin_work *work = intel_crtc->unpin_work;
+
+	/* ensure that the unpin work is consistent wrt ->pending. */
+	smp_rmb();
+	intel_crtc->unpin_work = NULL;
+
+	if (work->event)
+		drm_send_vblank_event(intel_crtc->base.dev,
+				      intel_crtc->pipe,
+				      work->event);
+
+	drm_crtc_vblank_put(&intel_crtc->base);
+
+	wake_up_all(&dev_priv->pending_flip_queue);
+	queue_work(dev_priv->wq, &work->work);
+
+	trace_i915_flip_complete(intel_crtc->plane,
+				 work->pending_flip_obj);
+}
+
 void intel_crtc_wait_for_pending_flips(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	if (crtc->primary->fb == NULL)
-		return;
-
 	WARN_ON(waitqueue_active(&dev_priv->pending_flip_queue));
+	if (WARN_ON(wait_event_timeout(dev_priv->pending_flip_queue,
+				       !intel_crtc_has_pending_flip(crtc),
+				       60*HZ) == 0)) {
+		struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+		unsigned long flags;
 
-	WARN_ON(wait_event_timeout(dev_priv->pending_flip_queue,
-				   !intel_crtc_has_pending_flip(crtc),
-				   60*HZ) == 0);
+		spin_lock_irqsave(&dev->event_lock, flags);
+		if (intel_crtc->unpin_work) {
+			WARN_ONCE(1, "Removing stuck page flip\n");
+			page_flip_completed(intel_crtc);
+		}
+		spin_unlock_irqrestore(&dev->event_lock, flags);
+	}
 
-	mutex_lock(&dev->struct_mutex);
-	intel_finish_fb(crtc->primary->fb);
-	mutex_unlock(&dev->struct_mutex);
+	if (crtc->primary->fb) {
+		mutex_lock(&dev->struct_mutex);
+		intel_finish_fb(crtc->primary->fb);
+		mutex_unlock(&dev->struct_mutex);
+	}
 }
 
 /* Program iCLKIP clock to the desired frequency */
@@ -3911,14 +4038,14 @@ static void intel_crtc_dpms_overlay(struct intel_crtc *intel_crtc, bool enable)
 static void intel_crtc_enable_planes(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	int pipe = intel_crtc->pipe;
-	int plane = intel_crtc->plane;
+
+	assert_vblank_disabled(crtc);
 
 	drm_vblank_on(dev, pipe);
 
-	intel_enable_primary_hw_plane(dev_priv, plane, pipe);
+	intel_enable_primary_hw_plane(crtc->primary, crtc);
 	intel_enable_planes(crtc);
 	intel_crtc_update_cursor(crtc, true);
 	intel_crtc_dpms_overlay(intel_crtc, true);
@@ -3955,7 +4082,7 @@ static void intel_crtc_disable_planes(struct drm_crtc *crtc)
 	intel_crtc_dpms_overlay(intel_crtc, false);
 	intel_crtc_update_cursor(crtc, false);
 	intel_disable_planes(crtc);
-	intel_disable_primary_hw_plane(dev_priv, plane, pipe);
+	intel_disable_primary_hw_plane(crtc->primary, crtc);
 
 	/*
 	 * FIXME: Once we grow proper nuclear flip support out of this we need
@@ -3965,6 +4092,8 @@ static void intel_crtc_disable_planes(struct drm_crtc *crtc)
 	intel_frontbuffer_flip(dev, INTEL_FRONTBUFFER_ALL_MASK(pipe));
 
 	drm_vblank_off(dev, pipe);
+
+	assert_vblank_disabled(crtc);
 }
 
 static void ironlake_crtc_enable(struct drm_crtc *crtc)
@@ -3974,7 +4103,6 @@ static void ironlake_crtc_enable(struct drm_crtc *crtc)
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	struct intel_encoder *encoder;
 	int pipe = intel_crtc->pipe;
-	enum plane plane = intel_crtc->plane;
 
 	WARN_ON(!crtc->enabled);
 
@@ -3991,18 +4119,11 @@ static void ironlake_crtc_enable(struct drm_crtc *crtc)
 
 	if (intel_crtc->config.has_pch_encoder) {
 		intel_cpu_transcoder_set_m_n(intel_crtc,
-					     &intel_crtc->config.fdi_m_n);
+				     &intel_crtc->config.fdi_m_n, NULL);
 	}
 
 	ironlake_set_pipeconf(crtc);
 
-	/* Set up the display plane register */
-	I915_WRITE(DSPCNTR(plane), DISPPLANE_GAMMA_ENABLE);
-	POSTING_READ(DSPCNTR(plane));
-
-	dev_priv->display.update_primary_plane(crtc, crtc->primary->fb,
-					       crtc->x, crtc->y);
-
 	intel_crtc->active = true;
 
 	intel_set_cpu_fifo_underrun_reporting(dev, pipe, true);
@@ -4087,7 +4208,6 @@ static void haswell_crtc_enable(struct drm_crtc *crtc)
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	struct intel_encoder *encoder;
 	int pipe = intel_crtc->pipe;
-	enum plane plane = intel_crtc->plane;
 
 	WARN_ON(!crtc->enabled);
 
@@ -4102,22 +4222,20 @@ static void haswell_crtc_enable(struct drm_crtc *crtc)
 
 	intel_set_pipe_timings(intel_crtc);
 
+	if (intel_crtc->config.cpu_transcoder != TRANSCODER_EDP) {
+		I915_WRITE(PIPE_MULT(intel_crtc->config.cpu_transcoder),
+			   intel_crtc->config.pixel_multiplier - 1);
+	}
+
 	if (intel_crtc->config.has_pch_encoder) {
 		intel_cpu_transcoder_set_m_n(intel_crtc,
-					     &intel_crtc->config.fdi_m_n);
+				     &intel_crtc->config.fdi_m_n, NULL);
 	}
 
 	haswell_set_pipeconf(crtc);
 
 	intel_set_pipe_csc(crtc);
 
-	/* Set up the display plane register */
-	I915_WRITE(DSPCNTR(plane), DISPPLANE_GAMMA_ENABLE | DISPPLANE_PIPE_CSC_ENABLE);
-	POSTING_READ(DSPCNTR(plane));
-
-	dev_priv->display.update_primary_plane(crtc, crtc->primary->fb,
-					       crtc->x, crtc->y);
-
 	intel_crtc->active = true;
 
 	intel_set_cpu_fifo_underrun_reporting(dev, pipe, true);
@@ -4198,7 +4316,8 @@ static void ironlake_crtc_disable(struct drm_crtc *crtc)
 	if (intel_crtc->config.has_pch_encoder)
 		intel_set_pch_fifo_underrun_reporting(dev, pipe, false);
 
-	intel_disable_pipe(dev_priv, pipe);
+	intel_disable_pipe(intel_crtc);
+
 	ironlake_pfit_disable(intel_crtc);
 
 	for_each_encoder_on_crtc(dev, crtc, encoder)
@@ -4246,7 +4365,6 @@ static void haswell_crtc_disable(struct drm_crtc *crtc)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	struct intel_encoder *encoder;
-	int pipe = intel_crtc->pipe;
 	enum transcoder cpu_transcoder = intel_crtc->config.cpu_transcoder;
 
 	if (!intel_crtc->active)
@@ -4261,7 +4379,7 @@ static void haswell_crtc_disable(struct drm_crtc *crtc)
 
 	if (intel_crtc->config.has_pch_encoder)
 		intel_set_pch_fifo_underrun_reporting(dev, TRANSCODER_A, false);
-	intel_disable_pipe(dev_priv, pipe);
+	intel_disable_pipe(intel_crtc);
 
 	if (intel_crtc->config.dp_encoder_is_mst)
 		intel_ddi_set_vc_payload_alloc(crtc, false);
@@ -4539,12 +4657,57 @@ static void valleyview_set_cdclk(struct drm_device *dev, int cdclk)
 	vlv_update_cdclk(dev);
 }
 
+static void cherryview_set_cdclk(struct drm_device *dev, int cdclk)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 val, cmd;
+
+	WARN_ON(dev_priv->display.get_display_clock_speed(dev) != dev_priv->vlv_cdclk_freq);
+
+	switch (cdclk) {
+	case 400000:
+		cmd = 3;
+		break;
+	case 333333:
+	case 320000:
+		cmd = 2;
+		break;
+	case 266667:
+		cmd = 1;
+		break;
+	case 200000:
+		cmd = 0;
+		break;
+	default:
+		WARN_ON(1);
+		return;
+	}
+
+	mutex_lock(&dev_priv->rps.hw_lock);
+	val = vlv_punit_read(dev_priv, PUNIT_REG_DSPFREQ);
+	val &= ~DSPFREQGUAR_MASK_CHV;
+	val |= (cmd << DSPFREQGUAR_SHIFT_CHV);
+	vlv_punit_write(dev_priv, PUNIT_REG_DSPFREQ, val);
+	if (wait_for((vlv_punit_read(dev_priv, PUNIT_REG_DSPFREQ) &
+		      DSPFREQSTAT_MASK_CHV) == (cmd << DSPFREQSTAT_SHIFT_CHV),
+		     50)) {
+		DRM_ERROR("timed out waiting for CDclk change\n");
+	}
+	mutex_unlock(&dev_priv->rps.hw_lock);
+
+	vlv_update_cdclk(dev);
+}
+
 static int valleyview_calc_cdclk(struct drm_i915_private *dev_priv,
 				 int max_pixclk)
 {
 	int vco = valleyview_get_vco(dev_priv);
 	int freq_320 = (vco <<  1) % 320000 != 0 ? 333333 : 320000;
 
+	/* FIXME: Punit isn't quite ready yet */
+	if (IS_CHERRYVIEW(dev_priv->dev))
+		return 400000;
+
 	/*
 	 * Really only a few cases to deal with, as only 4 CDclks are supported:
 	 *   200MHz
@@ -4607,21 +4770,23 @@ static void valleyview_modeset_global_resources(struct drm_device *dev)
 	int max_pixclk = intel_mode_max_pixclk(dev_priv);
 	int req_cdclk = valleyview_calc_cdclk(dev_priv, max_pixclk);
 
-	if (req_cdclk != dev_priv->vlv_cdclk_freq)
-		valleyview_set_cdclk(dev, req_cdclk);
+	if (req_cdclk != dev_priv->vlv_cdclk_freq) {
+		if (IS_CHERRYVIEW(dev))
+			cherryview_set_cdclk(dev, req_cdclk);
+		else
+			valleyview_set_cdclk(dev, req_cdclk);
+	}
+
 	modeset_update_crtc_power_domains(dev);
 }
 
 static void valleyview_crtc_enable(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	struct intel_encoder *encoder;
 	int pipe = intel_crtc->pipe;
-	int plane = intel_crtc->plane;
 	bool is_dsi;
-	u32 dspcntr;
 
 	WARN_ON(!crtc->enabled);
 
@@ -4630,33 +4795,20 @@ static void valleyview_crtc_enable(struct drm_crtc *crtc)
 
 	is_dsi = intel_pipe_has_type(crtc, INTEL_OUTPUT_DSI);
 
-	if (!is_dsi && !IS_CHERRYVIEW(dev))
-		vlv_prepare_pll(intel_crtc);
-
-	/* Set up the display plane register */
-	dspcntr = DISPPLANE_GAMMA_ENABLE;
+	if (!is_dsi) {
+		if (IS_CHERRYVIEW(dev))
+			chv_prepare_pll(intel_crtc);
+		else
+			vlv_prepare_pll(intel_crtc);
+	}
 
 	if (intel_crtc->config.has_dp_encoder)
 		intel_dp_set_m_n(intel_crtc);
 
 	intel_set_pipe_timings(intel_crtc);
 
-	/* pipesrc and dspsize control the size that is scaled from,
-	 * which should always be the user's requested size.
-	 */
-	I915_WRITE(DSPSIZE(plane),
-		   ((intel_crtc->config.pipe_src_h - 1) << 16) |
-		   (intel_crtc->config.pipe_src_w - 1));
-	I915_WRITE(DSPPOS(plane), 0);
-
 	i9xx_set_pipeconf(intel_crtc);
 
-	I915_WRITE(DSPCNTR(plane), dspcntr);
-	POSTING_READ(DSPCNTR(plane));
-
-	dev_priv->display.update_primary_plane(crtc, crtc->primary->fb,
-					       crtc->x, crtc->y);
-
 	intel_crtc->active = true;
 
 	intel_set_cpu_fifo_underrun_reporting(dev, pipe, true);
@@ -4704,12 +4856,9 @@ static void i9xx_set_pll_dividers(struct intel_crtc *crtc)
 static void i9xx_crtc_enable(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	struct intel_encoder *encoder;
 	int pipe = intel_crtc->pipe;
-	int plane = intel_crtc->plane;
-	u32 dspcntr;
 
 	WARN_ON(!crtc->enabled);
 
@@ -4718,35 +4867,13 @@ static void i9xx_crtc_enable(struct drm_crtc *crtc)
 
 	i9xx_set_pll_dividers(intel_crtc);
 
-	/* Set up the display plane register */
-	dspcntr = DISPPLANE_GAMMA_ENABLE;
-
-	if (pipe == 0)
-		dspcntr &= ~DISPPLANE_SEL_PIPE_MASK;
-	else
-		dspcntr |= DISPPLANE_SEL_PIPE_B;
-
 	if (intel_crtc->config.has_dp_encoder)
 		intel_dp_set_m_n(intel_crtc);
 
 	intel_set_pipe_timings(intel_crtc);
 
-	/* pipesrc and dspsize control the size that is scaled from,
-	 * which should always be the user's requested size.
-	 */
-	I915_WRITE(DSPSIZE(plane),
-		   ((intel_crtc->config.pipe_src_h - 1) << 16) |
-		   (intel_crtc->config.pipe_src_w - 1));
-	I915_WRITE(DSPPOS(plane), 0);
-
 	i9xx_set_pipeconf(intel_crtc);
 
-	I915_WRITE(DSPCNTR(plane), dspcntr);
-	POSTING_READ(DSPCNTR(plane));
-
-	dev_priv->display.update_primary_plane(crtc, crtc->primary->fb,
-					       crtc->x, crtc->y);
-
 	intel_crtc->active = true;
 
 	if (!IS_GEN2(dev))
@@ -4842,7 +4969,7 @@ static void i9xx_crtc_disable(struct drm_crtc *crtc)
 	 */
 	intel_wait_for_vblank(dev, pipe);
 
-	intel_disable_pipe(dev_priv, pipe);
+	intel_disable_pipe(intel_crtc);
 
 	i9xx_pfit_disable(intel_crtc);
 
@@ -4856,7 +4983,7 @@ static void i9xx_crtc_disable(struct drm_crtc *crtc)
 		else if (IS_VALLEYVIEW(dev))
 			vlv_disable_pll(dev_priv, pipe);
 		else
-			i9xx_disable_pll(dev_priv, pipe);
+			i9xx_disable_pll(intel_crtc);
 	}
 
 	if (!IS_GEN2(dev))
@@ -5275,6 +5402,10 @@ static int valleyview_get_display_clock_speed(struct drm_device *dev)
 	u32 val;
 	int divider;
 
+	/* FIXME: Punit isn't quite ready yet */
+	if (IS_CHERRYVIEW(dev))
+		return 400000;
+
 	mutex_lock(&dev_priv->dpio_lock);
 	val = vlv_cck_read(dev_priv, CCK_DISPLAY_CLOCK_CONTROL);
 	mutex_unlock(&dev_priv->dpio_lock);
@@ -5519,7 +5650,8 @@ static void intel_pch_transcoder_set_m_n(struct intel_crtc *crtc,
 }
 
 static void intel_cpu_transcoder_set_m_n(struct intel_crtc *crtc,
-					 struct intel_link_m_n *m_n)
+					 struct intel_link_m_n *m_n,
+					 struct intel_link_m_n *m2_n2)
 {
 	struct drm_device *dev = crtc->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -5531,6 +5663,18 @@ static void intel_cpu_transcoder_set_m_n(struct intel_crtc *crtc,
 		I915_WRITE(PIPE_DATA_N1(transcoder), m_n->gmch_n);
 		I915_WRITE(PIPE_LINK_M1(transcoder), m_n->link_m);
 		I915_WRITE(PIPE_LINK_N1(transcoder), m_n->link_n);
+		/* M2_N2 registers to be set only for gen < 8 (M2_N2 available
+		 * for gen < 8) and if DRRS is supported (to make sure the
+		 * registers are not unnecessarily accessed).
+		 */
+		if (m2_n2 && INTEL_INFO(dev)->gen < 8 &&
+			crtc->config.has_drrs) {
+			I915_WRITE(PIPE_DATA_M2(transcoder),
+					TU_SIZE(m2_n2->tu) | m2_n2->gmch_m);
+			I915_WRITE(PIPE_DATA_N2(transcoder), m2_n2->gmch_n);
+			I915_WRITE(PIPE_LINK_M2(transcoder), m2_n2->link_m);
+			I915_WRITE(PIPE_LINK_N2(transcoder), m2_n2->link_n);
+		}
 	} else {
 		I915_WRITE(PIPE_DATA_M_G4X(pipe), TU_SIZE(m_n->tu) | m_n->gmch_m);
 		I915_WRITE(PIPE_DATA_N_G4X(pipe), m_n->gmch_n);
@@ -5539,12 +5683,13 @@ static void intel_cpu_transcoder_set_m_n(struct intel_crtc *crtc,
 	}
 }
 
-static void intel_dp_set_m_n(struct intel_crtc *crtc)
+void intel_dp_set_m_n(struct intel_crtc *crtc)
 {
 	if (crtc->config.has_pch_encoder)
 		intel_pch_transcoder_set_m_n(crtc, &crtc->config.dp_m_n);
 	else
-		intel_cpu_transcoder_set_m_n(crtc, &crtc->config.dp_m_n);
+		intel_cpu_transcoder_set_m_n(crtc, &crtc->config.dp_m_n,
+						   &crtc->config.dp_m2_n2);
 }
 
 static void vlv_update_pll(struct intel_crtc *crtc)
@@ -5662,6 +5807,18 @@ static void vlv_prepare_pll(struct intel_crtc *crtc)
 
 static void chv_update_pll(struct intel_crtc *crtc)
 {
+	crtc->config.dpll_hw_state.dpll = DPLL_SSC_REF_CLOCK_CHV |
+		DPLL_REFA_CLK_ENABLE_VLV | DPLL_VGA_MODE_DIS |
+		DPLL_VCO_ENABLE;
+	if (crtc->pipe != PIPE_A)
+		crtc->config.dpll_hw_state.dpll |= DPLL_INTEGRATED_CRI_CLK_VLV;
+
+	crtc->config.dpll_hw_state.dpll_md =
+		(crtc->config.pixel_multiplier - 1) << DPLL_MD_UDI_MULTIPLIER_SHIFT;
+}
+
+static void chv_prepare_pll(struct intel_crtc *crtc)
+{
 	struct drm_device *dev = crtc->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	int pipe = crtc->pipe;
@@ -5671,15 +5828,6 @@ static void chv_update_pll(struct intel_crtc *crtc)
 	u32 bestn, bestm1, bestm2, bestp1, bestp2, bestm2_frac;
 	int refclk;
 
-	crtc->config.dpll_hw_state.dpll = DPLL_SSC_REF_CLOCK_CHV |
-		DPLL_REFA_CLK_ENABLE_VLV | DPLL_VGA_MODE_DIS |
-		DPLL_VCO_ENABLE;
-	if (pipe != PIPE_A)
-		crtc->config.dpll_hw_state.dpll |= DPLL_INTEGRATED_CRI_CLK_VLV;
-
-	crtc->config.dpll_hw_state.dpll_md =
-		(crtc->config.pixel_multiplier - 1) << DPLL_MD_UDI_MULTIPLIER_SHIFT;
-
 	bestn = crtc->config.dpll.n;
 	bestm2_frac = crtc->config.dpll.m2 & 0x3fffff;
 	bestm1 = crtc->config.dpll.m1;
@@ -5839,7 +5987,7 @@ static void i8xx_update_pll(struct intel_crtc *crtc,
 			dpll |= PLL_P2_DIVIDE_BY_4;
 	}
 
-	if (intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_DVO))
+	if (!IS_I830(dev) && intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_DVO))
 		dpll |= DPLL_DVO_2X_MODE;
 
 	if (intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_LVDS) &&
@@ -5990,9 +6138,9 @@ static void i9xx_set_pipeconf(struct intel_crtc *intel_crtc)
 
 	pipeconf = 0;
 
-	if (dev_priv->quirks & QUIRK_PIPEA_FORCE &&
-	    I915_READ(PIPECONF(intel_crtc->pipe)) & PIPECONF_ENABLE)
-		pipeconf |= PIPECONF_ENABLE;
+	if ((intel_crtc->pipe == PIPE_A && dev_priv->quirks & QUIRK_PIPEA_FORCE) ||
+	    (intel_crtc->pipe == PIPE_B && dev_priv->quirks & QUIRK_PIPEB_FORCE))
+		pipeconf |= I915_READ(PIPECONF(intel_crtc->pipe)) & PIPECONF_ENABLE;
 
 	if (intel_crtc->config.double_wide)
 		pipeconf |= PIPECONF_DOUBLE_WIDE;
@@ -6235,7 +6383,7 @@ static void i9xx_get_plane_config(struct intel_crtc *crtc,
 	crtc->base.primary->fb->height = ((val >> 0) & 0xfff) + 1;
 
 	val = I915_READ(DSPSTRIDE(pipe));
-	crtc->base.primary->fb->pitches[0] = val & 0xffffff80;
+	crtc->base.primary->fb->pitches[0] = val & 0xffffffc0;
 
 	aligned_height = intel_align_height(dev, crtc->base.primary->fb->height,
 					    plane_config->tiled);
@@ -6345,6 +6493,14 @@ static bool i9xx_get_pipe_config(struct intel_crtc *crtc,
 	}
 	pipe_config->dpll_hw_state.dpll = I915_READ(DPLL(crtc->pipe));
 	if (!IS_VALLEYVIEW(dev)) {
+		/*
+		 * DPLL_DVO_2X_MODE must be enabled for both DPLLs
+		 * on 830. Filter it out here so that we don't
+		 * report errors due to that.
+		 */
+		if (IS_I830(dev))
+			pipe_config->dpll_hw_state.dpll &= ~DPLL_DVO_2X_MODE;
+
 		pipe_config->dpll_hw_state.fp0 = I915_READ(FP0(crtc->pipe));
 		pipe_config->dpll_hw_state.fp1 = I915_READ(FP1(crtc->pipe));
 	} else {
@@ -6367,7 +6523,6 @@ static bool i9xx_get_pipe_config(struct intel_crtc *crtc,
 static void ironlake_init_pch_refclk(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_mode_config *mode_config = &dev->mode_config;
 	struct intel_encoder *encoder;
 	u32 val, final;
 	bool has_lvds = false;
@@ -6377,8 +6532,7 @@ static void ironlake_init_pch_refclk(struct drm_device *dev)
 	bool can_ssc = false;
 
 	/* We need to take the global config into account */
-	list_for_each_entry(encoder, &mode_config->encoder_list,
-			    base.head) {
+	for_each_intel_encoder(dev, encoder) {
 		switch (encoder->type) {
 		case INTEL_OUTPUT_LVDS:
 			has_panel = true;
@@ -6685,11 +6839,10 @@ static void lpt_disable_clkout_dp(struct drm_device *dev)
 
 static void lpt_init_pch_refclk(struct drm_device *dev)
 {
-	struct drm_mode_config *mode_config = &dev->mode_config;
 	struct intel_encoder *encoder;
 	bool has_vga = false;
 
-	list_for_each_entry(encoder, &mode_config->encoder_list, base.head) {
+	for_each_intel_encoder(dev, encoder) {
 		switch (encoder->type) {
 		case INTEL_OUTPUT_ANALOG:
 			has_vga = true;
@@ -7145,7 +7298,8 @@ static void intel_pch_transcoder_get_m_n(struct intel_crtc *crtc,
 
 static void intel_cpu_transcoder_get_m_n(struct intel_crtc *crtc,
 					 enum transcoder transcoder,
-					 struct intel_link_m_n *m_n)
+					 struct intel_link_m_n *m_n,
+					 struct intel_link_m_n *m2_n2)
 {
 	struct drm_device *dev = crtc->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -7159,6 +7313,20 @@ static void intel_cpu_transcoder_get_m_n(struct intel_crtc *crtc,
 		m_n->gmch_n = I915_READ(PIPE_DATA_N1(transcoder));
 		m_n->tu = ((I915_READ(PIPE_DATA_M1(transcoder))
 			    & TU_SIZE_MASK) >> TU_SIZE_SHIFT) + 1;
+		/* Read M2_N2 registers only for gen < 8 (M2_N2 available for
+		 * gen < 8) and if DRRS is supported (to make sure the
+		 * registers are not unnecessarily read).
+		 */
+		if (m2_n2 && INTEL_INFO(dev)->gen < 8 &&
+			crtc->config.has_drrs) {
+			m2_n2->link_m = I915_READ(PIPE_LINK_M2(transcoder));
+			m2_n2->link_n =	I915_READ(PIPE_LINK_N2(transcoder));
+			m2_n2->gmch_m =	I915_READ(PIPE_DATA_M2(transcoder))
+					& ~TU_SIZE_MASK;
+			m2_n2->gmch_n =	I915_READ(PIPE_DATA_N2(transcoder));
+			m2_n2->tu = ((I915_READ(PIPE_DATA_M2(transcoder))
+					& TU_SIZE_MASK) >> TU_SIZE_SHIFT) + 1;
+		}
 	} else {
 		m_n->link_m = I915_READ(PIPE_LINK_M_G4X(pipe));
 		m_n->link_n = I915_READ(PIPE_LINK_N_G4X(pipe));
@@ -7177,14 +7345,15 @@ void intel_dp_get_m_n(struct intel_crtc *crtc,
 		intel_pch_transcoder_get_m_n(crtc, &pipe_config->dp_m_n);
 	else
 		intel_cpu_transcoder_get_m_n(crtc, pipe_config->cpu_transcoder,
-					     &pipe_config->dp_m_n);
+					     &pipe_config->dp_m_n,
+					     &pipe_config->dp_m2_n2);
 }
 
 static void ironlake_get_fdi_m_n_config(struct intel_crtc *crtc,
 					struct intel_crtc_config *pipe_config)
 {
 	intel_cpu_transcoder_get_m_n(crtc, pipe_config->cpu_transcoder,
-				     &pipe_config->fdi_m_n);
+				     &pipe_config->fdi_m_n, NULL);
 }
 
 static void ironlake_get_pfit_config(struct intel_crtc *crtc,
@@ -7255,7 +7424,7 @@ static void ironlake_get_plane_config(struct intel_crtc *crtc,
 	crtc->base.primary->fb->height = ((val >> 0) & 0xfff) + 1;
 
 	val = I915_READ(DSPSTRIDE(pipe));
-	crtc->base.primary->fb->pitches[0] = val & 0xffffff80;
+	crtc->base.primary->fb->pitches[0] = val & 0xffffffc0;
 
 	aligned_height = intel_align_height(dev, crtc->base.primary->fb->height,
 					    plane_config->tiled);
@@ -7615,6 +7784,22 @@ static int haswell_crtc_mode_set(struct drm_crtc *crtc,
 	return 0;
 }
 
+static void haswell_get_ddi_pll(struct drm_i915_private *dev_priv,
+				enum port port,
+				struct intel_crtc_config *pipe_config)
+{
+	pipe_config->ddi_pll_sel = I915_READ(PORT_CLK_SEL(port));
+
+	switch (pipe_config->ddi_pll_sel) {
+	case PORT_CLK_SEL_WRPLL1:
+		pipe_config->shared_dpll = DPLL_ID_WRPLL1;
+		break;
+	case PORT_CLK_SEL_WRPLL2:
+		pipe_config->shared_dpll = DPLL_ID_WRPLL2;
+		break;
+	}
+}
+
 static void haswell_get_ddi_port_state(struct intel_crtc *crtc,
 				       struct intel_crtc_config *pipe_config)
 {
@@ -7628,16 +7813,7 @@ static void haswell_get_ddi_port_state(struct intel_crtc *crtc,
 
 	port = (tmp & TRANS_DDI_PORT_MASK) >> TRANS_DDI_PORT_SHIFT;
 
-	pipe_config->ddi_pll_sel = I915_READ(PORT_CLK_SEL(port));
-
-	switch (pipe_config->ddi_pll_sel) {
-	case PORT_CLK_SEL_WRPLL1:
-		pipe_config->shared_dpll = DPLL_ID_WRPLL1;
-		break;
-	case PORT_CLK_SEL_WRPLL2:
-		pipe_config->shared_dpll = DPLL_ID_WRPLL2;
-		break;
-	}
+	haswell_get_ddi_pll(dev_priv, port, pipe_config);
 
 	if (pipe_config->shared_dpll >= 0) {
 		pll = &dev_priv->shared_dplls[pipe_config->shared_dpll];
@@ -7719,7 +7895,12 @@ static bool haswell_get_pipe_config(struct intel_crtc *crtc,
 		pipe_config->ips_enabled = hsw_crtc_supports_ips(crtc) &&
 			(I915_READ(IPS_CTL) & IPS_ENABLE);
 
-	pipe_config->pixel_multiplier = 1;
+	if (pipe_config->cpu_transcoder != TRANSCODER_EDP) {
+		pipe_config->pixel_multiplier =
+			I915_READ(PIPE_MULT(pipe_config->cpu_transcoder)) + 1;
+	} else {
+		pipe_config->pixel_multiplier = 1;
+	}
 
 	return true;
 }
@@ -8037,74 +8218,62 @@ static void i845_update_cursor(struct drm_crtc *crtc, u32 base)
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	uint32_t cntl;
+	uint32_t cntl = 0, size = 0;
 
-	if (base != intel_crtc->cursor_base) {
-		/* On these chipsets we can only modify the base whilst
-		 * the cursor is disabled.
-		 */
-		if (intel_crtc->cursor_cntl) {
-			I915_WRITE(_CURACNTR, 0);
-			POSTING_READ(_CURACNTR);
-			intel_crtc->cursor_cntl = 0;
+	if (base) {
+		unsigned int width = intel_crtc->cursor_width;
+		unsigned int height = intel_crtc->cursor_height;
+		unsigned int stride = roundup_pow_of_two(width) * 4;
+
+		switch (stride) {
+		default:
+			WARN_ONCE(1, "Invalid cursor width/stride, width=%u, stride=%u\n",
+				  width, stride);
+			stride = 256;
+			/* fallthrough */
+		case 256:
+		case 512:
+		case 1024:
+		case 2048:
+			break;
 		}
 
-		I915_WRITE(_CURABASE, base);
-		POSTING_READ(_CURABASE);
+		cntl |= CURSOR_ENABLE |
+			CURSOR_GAMMA_ENABLE |
+			CURSOR_FORMAT_ARGB |
+			CURSOR_STRIDE(stride);
+
+		size = (height << 12) | width;
 	}
 
-	/* XXX width must be 64, stride 256 => 0x00 << 28 */
-	cntl = 0;
-	if (base)
-		cntl = (CURSOR_ENABLE |
-			CURSOR_GAMMA_ENABLE |
-			CURSOR_FORMAT_ARGB);
-	if (intel_crtc->cursor_cntl != cntl) {
-		I915_WRITE(_CURACNTR, cntl);
+	if (intel_crtc->cursor_cntl != 0 &&
+	    (intel_crtc->cursor_base != base ||
+	     intel_crtc->cursor_size != size ||
+	     intel_crtc->cursor_cntl != cntl)) {
+		/* On these chipsets we can only modify the base/size/stride
+		 * whilst the cursor is disabled.
+		 */
+		I915_WRITE(_CURACNTR, 0);
 		POSTING_READ(_CURACNTR);
-		intel_crtc->cursor_cntl = cntl;
+		intel_crtc->cursor_cntl = 0;
 	}
-}
 
-static void i9xx_update_cursor(struct drm_crtc *crtc, u32 base)
-{
-	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	int pipe = intel_crtc->pipe;
-	uint32_t cntl;
+	if (intel_crtc->cursor_base != base)
+		I915_WRITE(_CURABASE, base);
 
-	cntl = 0;
-	if (base) {
-		cntl = MCURSOR_GAMMA_ENABLE;
-		switch (intel_crtc->cursor_width) {
-			case 64:
-				cntl |= CURSOR_MODE_64_ARGB_AX;
-				break;
-			case 128:
-				cntl |= CURSOR_MODE_128_ARGB_AX;
-				break;
-			case 256:
-				cntl |= CURSOR_MODE_256_ARGB_AX;
-				break;
-			default:
-				WARN_ON(1);
-				return;
-		}
-		cntl |= pipe << 28; /* Connect to correct pipe */
+	if (intel_crtc->cursor_size != size) {
+		I915_WRITE(CURSIZE, size);
+		intel_crtc->cursor_size = size;
 	}
+
 	if (intel_crtc->cursor_cntl != cntl) {
-		I915_WRITE(CURCNTR(pipe), cntl);
-		POSTING_READ(CURCNTR(pipe));
+		I915_WRITE(_CURACNTR, cntl);
+		POSTING_READ(_CURACNTR);
 		intel_crtc->cursor_cntl = cntl;
 	}
-
-	/* and commit changes on next vblank */
-	I915_WRITE(CURBASE(pipe), base);
-	POSTING_READ(CURBASE(pipe));
 }
 
-static void ivb_update_cursor(struct drm_crtc *crtc, u32 base)
+static void i9xx_update_cursor(struct drm_crtc *crtc, u32 base)
 {
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -8129,6 +8298,7 @@ static void ivb_update_cursor(struct drm_crtc *crtc, u32 base)
 				WARN_ON(1);
 				return;
 		}
+		cntl |= pipe << 28; /* Connect to correct pipe */
 	}
 	if (IS_HASWELL(dev) || IS_BROADWELL(dev))
 		cntl |= CURSOR_PIPE_CSC_ENABLE;
@@ -8188,15 +8358,50 @@ static void intel_crtc_update_cursor(struct drm_crtc *crtc,
 
 	I915_WRITE(CURPOS(pipe), pos);
 
-	if (IS_IVYBRIDGE(dev) || IS_HASWELL(dev) || IS_BROADWELL(dev))
-		ivb_update_cursor(crtc, base);
-	else if (IS_845G(dev) || IS_I865G(dev))
+	if (IS_845G(dev) || IS_I865G(dev))
 		i845_update_cursor(crtc, base);
 	else
 		i9xx_update_cursor(crtc, base);
 	intel_crtc->cursor_base = base;
 }
 
+static bool cursor_size_ok(struct drm_device *dev,
+			   uint32_t width, uint32_t height)
+{
+	if (width == 0 || height == 0)
+		return false;
+
+	/*
+	 * 845g/865g are special in that they are only limited by
+	 * the width of their cursors, the height is arbitrary up to
+	 * the precision of the register. Everything else requires
+	 * square cursors, limited to a few power-of-two sizes.
+	 */
+	if (IS_845G(dev) || IS_I865G(dev)) {
+		if ((width & 63) != 0)
+			return false;
+
+		if (width > (IS_845G(dev) ? 64 : 512))
+			return false;
+
+		if (height > 1023)
+			return false;
+	} else {
+		switch (width | height) {
+		case 256:
+		case 128:
+			if (IS_GEN2(dev))
+				return false;
+		case 64:
+			break;
+		default:
+			return false;
+		}
+	}
+
+	return true;
+}
+
 /*
  * intel_crtc_cursor_set_obj - Set cursor to specified GEM object
  *
@@ -8212,7 +8417,7 @@ static int intel_crtc_cursor_set_obj(struct drm_crtc *crtc,
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	enum pipe pipe = intel_crtc->pipe;
-	unsigned old_width;
+	unsigned old_width, stride;
 	uint32_t addr;
 	int ret;
 
@@ -8220,20 +8425,18 @@ static int intel_crtc_cursor_set_obj(struct drm_crtc *crtc,
 	if (!obj) {
 		DRM_DEBUG_KMS("cursor off\n");
 		addr = 0;
-		obj = NULL;
 		mutex_lock(&dev->struct_mutex);
 		goto finish;
 	}
 
 	/* Check for which cursor types we support */
-	if (!((width == 64 && height == 64) ||
-			(width == 128 && height == 128 && !IS_GEN2(dev)) ||
-			(width == 256 && height == 256 && !IS_GEN2(dev)))) {
+	if (!cursor_size_ok(dev, width, height)) {
 		DRM_DEBUG("Cursor dimension not supported\n");
 		return -EINVAL;
 	}
 
-	if (obj->base.size < width * height * 4) {
+	stride = roundup_pow_of_two(width) * 4;
+	if (obj->base.size < stride * height) {
 		DRM_DEBUG_KMS("buffer is too small\n");
 		ret = -ENOMEM;
 		goto fail;
@@ -8295,9 +8498,6 @@ static int intel_crtc_cursor_set_obj(struct drm_crtc *crtc,
 		addr = obj->phys_handle->busaddr;
 	}
 
-	if (IS_GEN2(dev))
-		I915_WRITE(CURSIZE, (height << 12) | width);
-
  finish:
 	if (intel_crtc->cursor_bo) {
 		if (!INTEL_INFO(dev)->cursor_needs_physical)
@@ -8944,12 +9144,13 @@ static void intel_mark_fb_busy(struct drm_device *dev,
 			       unsigned frontbuffer_bits,
 			       struct intel_engine_cs *ring)
 {
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	enum pipe pipe;
 
 	if (!i915.powersave)
 		return;
 
-	for_each_pipe(pipe) {
+	for_each_pipe(dev_priv, pipe) {
 		if (!(frontbuffer_bits & INTEL_FRONTBUFFER_ALL_MASK(pipe)))
 			continue;
 
@@ -9019,6 +9220,14 @@ void intel_frontbuffer_flush(struct drm_device *dev,
 	intel_mark_fb_busy(dev, frontbuffer_bits, NULL);
 
 	intel_edp_psr_flush(dev, frontbuffer_bits);
+
+	/*
+	 * FIXME: Unconditional fbc flushing here is a rather gross hack and
+	 * needs to be reworked into a proper frontbuffer tracking scheme like
+	 * psr employs.
+	 */
+	if (IS_BROADWELL(dev))
+		gen8_fbc_sw_flush(dev, FBC_REND_CACHE_CLEAN);
 }
 
 /**
@@ -9151,7 +9360,6 @@ static void intel_unpin_work_fn(struct work_struct *__work)
 static void do_intel_finish_page_flip(struct drm_device *dev,
 				      struct drm_crtc *crtc)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	struct intel_unpin_work *work;
 	unsigned long flags;
@@ -9171,23 +9379,9 @@ static void do_intel_finish_page_flip(struct drm_device *dev,
 		return;
 	}
 
-	/* and that the unpin work is consistent wrt ->pending. */
-	smp_rmb();
-
-	intel_crtc->unpin_work = NULL;
-
-	if (work->event)
-		drm_send_vblank_event(dev, intel_crtc->pipe, work->event);
-
-	drm_crtc_vblank_put(crtc);
+	page_flip_completed(intel_crtc);
 
 	spin_unlock_irqrestore(&dev->event_lock, flags);
-
-	wake_up_all(&dev_priv->pending_flip_queue);
-
-	queue_work(dev_priv->wq, &work->work);
-
-	trace_i915_flip_complete(intel_crtc->plane, work->pending_flip_obj);
 }
 
 void intel_finish_page_flip(struct drm_device *dev, int pipe)
@@ -9532,6 +9726,8 @@ static bool use_mmio_flip(struct intel_engine_cs *ring,
 		return false;
 	else if (i915.use_mmio_flip > 0)
 		return true;
+	else if (i915.enable_execlists)
+		return true;
 	else
 		return ring != obj->ring;
 }
@@ -9665,6 +9861,65 @@ static int intel_default_queue_flip(struct drm_device *dev,
 	return -ENODEV;
 }
 
+static bool __intel_pageflip_stall_check(struct drm_device *dev,
+					 struct drm_crtc *crtc)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+	struct intel_unpin_work *work = intel_crtc->unpin_work;
+	u32 addr;
+
+	if (atomic_read(&work->pending) >= INTEL_FLIP_COMPLETE)
+		return true;
+
+	if (!work->enable_stall_check)
+		return false;
+
+	if (work->flip_ready_vblank == 0) {
+		if (work->flip_queued_ring &&
+		    !i915_seqno_passed(work->flip_queued_ring->get_seqno(work->flip_queued_ring, true),
+				       work->flip_queued_seqno))
+			return false;
+
+		work->flip_ready_vblank = drm_vblank_count(dev, intel_crtc->pipe);
+	}
+
+	if (drm_vblank_count(dev, intel_crtc->pipe) - work->flip_ready_vblank < 3)
+		return false;
+
+	/* Potential stall - if we see that the flip has happened,
+	 * assume a missed interrupt. */
+	if (INTEL_INFO(dev)->gen >= 4)
+		addr = I915_HI_DISPBASE(I915_READ(DSPSURF(intel_crtc->plane)));
+	else
+		addr = I915_READ(DSPADDR(intel_crtc->plane));
+
+	/* There is a potential issue here with a false positive after a flip
+	 * to the same address. We could address this by checking for a
+	 * non-incrementing frame counter.
+	 */
+	return addr == work->gtt_offset;
+}
+
+void intel_check_page_flip(struct drm_device *dev, int pipe)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct drm_crtc *crtc = dev_priv->pipe_to_crtc_mapping[pipe];
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+	unsigned long flags;
+
+	if (crtc == NULL)
+		return;
+
+	spin_lock_irqsave(&dev->event_lock, flags);
+	if (intel_crtc->unpin_work && __intel_pageflip_stall_check(dev, crtc)) {
+		WARN_ONCE(1, "Kicking stuck page flip: queued at %d, now %d\n",
+			 intel_crtc->unpin_work->flip_queued_vblank, drm_vblank_count(dev, pipe));
+		page_flip_completed(intel_crtc);
+	}
+	spin_unlock_irqrestore(&dev->event_lock, flags);
+}
+
 static int intel_crtc_page_flip(struct drm_crtc *crtc,
 				struct drm_framebuffer *fb,
 				struct drm_pending_vblank_event *event,
@@ -9721,12 +9976,20 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc,
 	/* We borrow the event spin lock for protecting unpin_work */
 	spin_lock_irqsave(&dev->event_lock, flags);
 	if (intel_crtc->unpin_work) {
-		spin_unlock_irqrestore(&dev->event_lock, flags);
-		kfree(work);
-		drm_crtc_vblank_put(crtc);
+		/* Before declaring the flip queue wedged, check if
+		 * the hardware completed the operation behind our backs.
+		 */
+		if (__intel_pageflip_stall_check(dev, crtc)) {
+			DRM_DEBUG_DRIVER("flip queue: previous flip completed, continuing\n");
+			page_flip_completed(intel_crtc);
+		} else {
+			DRM_DEBUG_DRIVER("flip queue: crtc already busy\n");
+			spin_unlock_irqrestore(&dev->event_lock, flags);
 
-		DRM_DEBUG_DRIVER("flip queue: crtc already busy\n");
-		return -EBUSY;
+			drm_crtc_vblank_put(crtc);
+			kfree(work);
+			return -EBUSY;
+		}
 	}
 	intel_crtc->unpin_work = work;
 	spin_unlock_irqrestore(&dev->event_lock, flags);
@@ -9746,8 +10009,6 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc,
 
 	work->pending_flip_obj = obj;
 
-	work->enable_stall_check = true;
-
 	atomic_inc(&intel_crtc->unpin_work_count);
 	intel_crtc->reset_counter = atomic_read(&dev_priv->gpu_error.reset_counter);
 
@@ -9776,14 +10037,26 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc,
 	work->gtt_offset =
 		i915_gem_obj_ggtt_offset(obj) + intel_crtc->dspaddr_offset;
 
-	if (use_mmio_flip(ring, obj))
+	if (use_mmio_flip(ring, obj)) {
 		ret = intel_queue_mmio_flip(dev, crtc, fb, obj, ring,
 					    page_flip_flags);
-	else
+		if (ret)
+			goto cleanup_unpin;
+
+		work->flip_queued_seqno = obj->last_write_seqno;
+		work->flip_queued_ring = obj->ring;
+	} else {
 		ret = dev_priv->display.queue_flip(dev, crtc, fb, obj, ring,
-				page_flip_flags);
-	if (ret)
-		goto cleanup_unpin;
+						   page_flip_flags);
+		if (ret)
+			goto cleanup_unpin;
+
+		work->flip_queued_seqno = intel_ring_get_seqno(ring);
+		work->flip_queued_ring = ring;
+	}
+
+	work->flip_queued_vblank = drm_vblank_count(dev, intel_crtc->pipe);
+	work->enable_stall_check = true;
 
 	i915_gem_track_fb(work->old_fb_obj, obj,
 			  INTEL_FRONTBUFFER_PRIMARY(pipe));
@@ -9818,8 +10091,11 @@ free_work:
 out_hang:
 		intel_crtc_wait_for_pending_flips(crtc);
 		ret = intel_pipe_set_base(crtc, crtc->x, crtc->y, fb);
-		if (ret == 0 && event)
+		if (ret == 0 && event) {
+			spin_lock_irqsave(&dev->event_lock, flags);
 			drm_send_vblank_event(dev, pipe, event);
+			spin_unlock_irqrestore(&dev->event_lock, flags);
+		}
 	}
 	return ret;
 }
@@ -9847,8 +10123,7 @@ static void intel_modeset_update_staged_output_state(struct drm_device *dev)
 			to_intel_encoder(connector->base.encoder);
 	}
 
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list,
-			    base.head) {
+	for_each_intel_encoder(dev, encoder) {
 		encoder->new_crtc =
 			to_intel_crtc(encoder->base.crtc);
 	}
@@ -9879,8 +10154,7 @@ static void intel_modeset_commit_output_state(struct drm_device *dev)
 		connector->base.encoder = &connector->new_encoder->base;
 	}
 
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list,
-			    base.head) {
+	for_each_intel_encoder(dev, encoder) {
 		encoder->base.crtc = &encoder->new_crtc->base;
 	}
 
@@ -10007,6 +10281,15 @@ static void intel_dump_pipe_config(struct intel_crtc *crtc,
 		      pipe_config->dp_m_n.gmch_m, pipe_config->dp_m_n.gmch_n,
 		      pipe_config->dp_m_n.link_m, pipe_config->dp_m_n.link_n,
 		      pipe_config->dp_m_n.tu);
+
+	DRM_DEBUG_KMS("dp: %i, gmch_m2: %u, gmch_n2: %u, link_m2: %u, link_n2: %u, tu2: %u\n",
+		      pipe_config->has_dp_encoder,
+		      pipe_config->dp_m2_n2.gmch_m,
+		      pipe_config->dp_m2_n2.gmch_n,
+		      pipe_config->dp_m2_n2.link_m,
+		      pipe_config->dp_m2_n2.link_n,
+		      pipe_config->dp_m2_n2.tu);
+
 	DRM_DEBUG_KMS("requested mode:\n");
 	drm_mode_debug_printmodeline(&pipe_config->requested_mode);
 	DRM_DEBUG_KMS("adjusted mode:\n");
@@ -10041,8 +10324,7 @@ static bool check_single_encoder_cloning(struct intel_crtc *crtc,
 	struct drm_device *dev = crtc->base.dev;
 	struct intel_encoder *source_encoder;
 
-	list_for_each_entry(source_encoder,
-			    &dev->mode_config.encoder_list, base.head) {
+	for_each_intel_encoder(dev, source_encoder) {
 		if (source_encoder->new_crtc != crtc)
 			continue;
 
@@ -10058,8 +10340,7 @@ static bool check_encoder_cloning(struct intel_crtc *crtc)
 	struct drm_device *dev = crtc->base.dev;
 	struct intel_encoder *encoder;
 
-	list_for_each_entry(encoder,
-			    &dev->mode_config.encoder_list, base.head) {
+	for_each_intel_encoder(dev, encoder) {
 		if (encoder->new_crtc != crtc)
 			continue;
 
@@ -10143,8 +10424,7 @@ encoder_retry:
 	 * adjust it according to limitations or connector properties, and also
 	 * a chance to reject the mode entirely.
 	 */
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list,
-			    base.head) {
+	for_each_intel_encoder(dev, encoder) {
 
 		if (&encoder->new_crtc->base != crtc)
 			continue;
@@ -10222,8 +10502,7 @@ intel_modeset_affected_pipes(struct drm_crtc *crtc, unsigned *modeset_pipes,
 				1 << connector->new_encoder->new_crtc->pipe;
 	}
 
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list,
-			    base.head) {
+	for_each_intel_encoder(dev, encoder) {
 		if (encoder->base.crtc == &encoder->new_crtc->base)
 			continue;
 
@@ -10297,8 +10576,7 @@ intel_modeset_update_state(struct drm_device *dev, unsigned prepare_pipes)
 	struct intel_crtc *intel_crtc;
 	struct drm_connector *connector;
 
-	list_for_each_entry(intel_encoder, &dev->mode_config.encoder_list,
-			    base.head) {
+	for_each_intel_encoder(dev, intel_encoder) {
 		if (!intel_encoder->base.crtc)
 			continue;
 
@@ -10387,6 +10665,22 @@ intel_pipe_config_compare(struct drm_device *dev,
 		return false; \
 	}
 
+/* This is required for BDW+ where there is only one set of registers for
+ * switching between high and low RR.
+ * This macro can be used whenever a comparison has to be made between one
+ * hw state and multiple sw state variables.
+ */
+#define PIPE_CONF_CHECK_I_ALT(name, alt_name) \
+	if ((current_config->name != pipe_config->name) && \
+		(current_config->alt_name != pipe_config->name)) { \
+			DRM_ERROR("mismatch in " #name " " \
+				  "(expected %i or %i, found %i)\n", \
+				  current_config->name, \
+				  current_config->alt_name, \
+				  pipe_config->name); \
+			return false; \
+	}
+
 #define PIPE_CONF_CHECK_FLAGS(name, mask)	\
 	if ((current_config->name ^ pipe_config->name) & (mask)) { \
 		DRM_ERROR("mismatch in " #name "(" #mask ") "	   \
@@ -10419,11 +10713,28 @@ intel_pipe_config_compare(struct drm_device *dev,
 	PIPE_CONF_CHECK_I(fdi_m_n.tu);
 
 	PIPE_CONF_CHECK_I(has_dp_encoder);
-	PIPE_CONF_CHECK_I(dp_m_n.gmch_m);
-	PIPE_CONF_CHECK_I(dp_m_n.gmch_n);
-	PIPE_CONF_CHECK_I(dp_m_n.link_m);
-	PIPE_CONF_CHECK_I(dp_m_n.link_n);
-	PIPE_CONF_CHECK_I(dp_m_n.tu);
+
+	if (INTEL_INFO(dev)->gen < 8) {
+		PIPE_CONF_CHECK_I(dp_m_n.gmch_m);
+		PIPE_CONF_CHECK_I(dp_m_n.gmch_n);
+		PIPE_CONF_CHECK_I(dp_m_n.link_m);
+		PIPE_CONF_CHECK_I(dp_m_n.link_n);
+		PIPE_CONF_CHECK_I(dp_m_n.tu);
+
+		if (current_config->has_drrs) {
+			PIPE_CONF_CHECK_I(dp_m2_n2.gmch_m);
+			PIPE_CONF_CHECK_I(dp_m2_n2.gmch_n);
+			PIPE_CONF_CHECK_I(dp_m2_n2.link_m);
+			PIPE_CONF_CHECK_I(dp_m2_n2.link_n);
+			PIPE_CONF_CHECK_I(dp_m2_n2.tu);
+		}
+	} else {
+		PIPE_CONF_CHECK_I_ALT(dp_m_n.gmch_m, dp_m2_n2.gmch_m);
+		PIPE_CONF_CHECK_I_ALT(dp_m_n.gmch_n, dp_m2_n2.gmch_n);
+		PIPE_CONF_CHECK_I_ALT(dp_m_n.link_m, dp_m2_n2.link_m);
+		PIPE_CONF_CHECK_I_ALT(dp_m_n.link_n, dp_m2_n2.link_n);
+		PIPE_CONF_CHECK_I_ALT(dp_m_n.tu, dp_m2_n2.tu);
+	}
 
 	PIPE_CONF_CHECK_I(adjusted_mode.crtc_hdisplay);
 	PIPE_CONF_CHECK_I(adjusted_mode.crtc_htotal);
@@ -10509,6 +10820,7 @@ intel_pipe_config_compare(struct drm_device *dev,
 
 #undef PIPE_CONF_CHECK_X
 #undef PIPE_CONF_CHECK_I
+#undef PIPE_CONF_CHECK_I_ALT
 #undef PIPE_CONF_CHECK_FLAGS
 #undef PIPE_CONF_CHECK_CLOCK_FUZZY
 #undef PIPE_CONF_QUIRK
@@ -10538,8 +10850,7 @@ check_encoder_state(struct drm_device *dev)
 	struct intel_encoder *encoder;
 	struct intel_connector *connector;
 
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list,
-			    base.head) {
+	for_each_intel_encoder(dev, encoder) {
 		bool enabled = false;
 		bool active = false;
 		enum pipe pipe, tracked_pipe;
@@ -10618,8 +10929,7 @@ check_crtc_state(struct drm_device *dev)
 		WARN(crtc->active && !crtc->base.enabled,
 		     "active crtc, but not enabled in sw tracking\n");
 
-		list_for_each_entry(encoder, &dev->mode_config.encoder_list,
-				    base.head) {
+		for_each_intel_encoder(dev, encoder) {
 			if (encoder->base.crtc != &crtc->base)
 				continue;
 			enabled = true;
@@ -10637,12 +10947,12 @@ check_crtc_state(struct drm_device *dev)
 		active = dev_priv->display.get_pipe_config(crtc,
 							   &pipe_config);
 
-		/* hw state is inconsistent with the pipe A quirk */
-		if (crtc->pipe == PIPE_A && dev_priv->quirks & QUIRK_PIPEA_FORCE)
+		/* hw state is inconsistent with the pipe quirk */
+		if ((crtc->pipe == PIPE_A && dev_priv->quirks & QUIRK_PIPEA_FORCE) ||
+		    (crtc->pipe == PIPE_B && dev_priv->quirks & QUIRK_PIPEB_FORCE))
 			active = crtc->active;
 
-		list_for_each_entry(encoder, &dev->mode_config.encoder_list,
-				    base.head) {
+		for_each_intel_encoder(dev, encoder) {
 			enum pipe pipe;
 			if (encoder->base.crtc != &crtc->base)
 				continue;
@@ -11010,7 +11320,7 @@ static void intel_set_config_restore_state(struct drm_device *dev,
 	}
 
 	count = 0;
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list, base.head) {
+	for_each_intel_encoder(dev, encoder) {
 		encoder->new_crtc =
 			to_intel_crtc(config->save_encoder_crtcs[count++]);
 	}
@@ -11169,8 +11479,7 @@ intel_modeset_stage_output_state(struct drm_device *dev,
 	}
 
 	/* Check for any encoders that needs to be disabled. */
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list,
-			    base.head) {
+	for_each_intel_encoder(dev, encoder) {
 		int num_connectors = 0;
 		list_for_each_entry(connector,
 				    &dev->mode_config.connector_list,
@@ -11203,9 +11512,7 @@ intel_modeset_stage_output_state(struct drm_device *dev,
 	for_each_intel_crtc(dev, crtc) {
 		crtc->new_enabled = false;
 
-		list_for_each_entry(encoder,
-				    &dev->mode_config.encoder_list,
-				    base.head) {
+		for_each_intel_encoder(dev, encoder) {
 			if (encoder->new_crtc == crtc) {
 				crtc->new_enabled = true;
 				break;
@@ -11242,7 +11549,7 @@ static void disable_crtc_nofb(struct intel_crtc *crtc)
 			connector->new_encoder = NULL;
 	}
 
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list, base.head) {
+	for_each_intel_encoder(dev, encoder) {
 		if (encoder->new_crtc == crtc)
 			encoder->new_crtc = NULL;
 	}
@@ -11305,7 +11612,6 @@ static int intel_crtc_set_config(struct drm_mode_set *set)
 		ret = intel_set_mode(set->crtc, set->mode,
 				     set->x, set->y, set->fb);
 	} else if (config->fb_changed) {
-		struct drm_i915_private *dev_priv = dev->dev_private;
 		struct intel_crtc *intel_crtc = to_intel_crtc(set->crtc);
 
 		intel_crtc_wait_for_pending_flips(set->crtc);
@@ -11319,8 +11625,7 @@ static int intel_crtc_set_config(struct drm_mode_set *set)
 		 */
 		if (!intel_crtc->primary_enabled && ret == 0) {
 			WARN_ON(!intel_crtc->active);
-			intel_enable_primary_hw_plane(dev_priv, intel_crtc->plane,
-						      intel_crtc->pipe);
+			intel_enable_primary_hw_plane(set->crtc->primary, set->crtc);
 		}
 
 		/*
@@ -11473,8 +11778,6 @@ static int
 intel_primary_plane_disable(struct drm_plane *plane)
 {
 	struct drm_device *dev = plane->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_plane *intel_plane = to_intel_plane(plane);
 	struct intel_crtc *intel_crtc;
 
 	if (!plane->fb)
@@ -11497,8 +11800,8 @@ intel_primary_plane_disable(struct drm_plane *plane)
 		goto disable_unpin;
 
 	intel_crtc_wait_for_pending_flips(plane->crtc);
-	intel_disable_primary_hw_plane(dev_priv, intel_plane->plane,
-				       intel_plane->pipe);
+	intel_disable_primary_hw_plane(plane, plane->crtc);
+
 disable_unpin:
 	mutex_lock(&dev->struct_mutex);
 	i915_gem_track_fb(intel_fb_obj(plane->fb), NULL,
@@ -11520,7 +11823,6 @@ intel_primary_plane_setplane(struct drm_plane *plane, struct drm_crtc *crtc,
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	struct intel_plane *intel_plane = to_intel_plane(plane);
 	struct drm_i915_gem_object *obj = intel_fb_obj(fb);
 	struct drm_i915_gem_object *old_obj = intel_fb_obj(plane->fb);
 	struct drm_rect dest = {
@@ -11542,6 +11844,21 @@ intel_primary_plane_setplane(struct drm_plane *plane, struct drm_crtc *crtc,
 		.x2 = intel_crtc->active ? intel_crtc->config.pipe_src_w : 0,
 		.y2 = intel_crtc->active ? intel_crtc->config.pipe_src_h : 0,
 	};
+	const struct {
+		int crtc_x, crtc_y;
+		unsigned int crtc_w, crtc_h;
+		uint32_t src_x, src_y, src_w, src_h;
+	} orig = {
+		.crtc_x = crtc_x,
+		.crtc_y = crtc_y,
+		.crtc_w = crtc_w,
+		.crtc_h = crtc_h,
+		.src_x = src_x,
+		.src_y = src_y,
+		.src_w = src_w,
+		.src_h = src_h,
+	};
+	struct intel_plane *intel_plane = to_intel_plane(plane);
 	bool visible;
 	int ret;
 
@@ -11607,9 +11924,7 @@ intel_primary_plane_setplane(struct drm_plane *plane, struct drm_crtc *crtc,
 				  INTEL_FRONTBUFFER_PRIMARY(intel_crtc->pipe));
 
 		if (intel_crtc->primary_enabled)
-			intel_disable_primary_hw_plane(dev_priv,
-						       intel_plane->plane,
-						       intel_plane->pipe);
+			intel_disable_primary_hw_plane(plane, crtc);
 
 
 		if (plane->fb != fb)
@@ -11618,16 +11933,42 @@ intel_primary_plane_setplane(struct drm_plane *plane, struct drm_crtc *crtc,
 
 		mutex_unlock(&dev->struct_mutex);
 
-		return 0;
-	}
+	} else {
+		if (intel_crtc && intel_crtc->active &&
+		    intel_crtc->primary_enabled) {
+			/*
+			 * FBC does not work on some platforms for rotated
+			 * planes, so disable it when rotation is not 0 and
+			 * update it when rotation is set back to 0.
+			 *
+			 * FIXME: This is redundant with the fbc update done in
+			 * the primary plane enable function except that that
+			 * one is done too late. We eventually need to unify
+			 * this.
+			 */
+			if (INTEL_INFO(dev)->gen <= 4 && !IS_G4X(dev) &&
+			    dev_priv->fbc.plane == intel_crtc->plane &&
+			    intel_plane->rotation != BIT(DRM_ROTATE_0)) {
+				intel_disable_fbc(dev);
+			}
+		}
+		ret = intel_pipe_set_base(crtc, src.x1, src.y1, fb);
+		if (ret)
+			return ret;
 
-	ret = intel_pipe_set_base(crtc, src.x1, src.y1, fb);
-	if (ret)
-		return ret;
+		if (!intel_crtc->primary_enabled)
+			intel_enable_primary_hw_plane(plane, crtc);
+	}
 
-	if (!intel_crtc->primary_enabled)
-		intel_enable_primary_hw_plane(dev_priv, intel_crtc->plane,
-					      intel_crtc->pipe);
+	intel_plane->crtc_x = orig.crtc_x;
+	intel_plane->crtc_y = orig.crtc_y;
+	intel_plane->crtc_w = orig.crtc_w;
+	intel_plane->crtc_h = orig.crtc_h;
+	intel_plane->src_x = orig.src_x;
+	intel_plane->src_y = orig.src_y;
+	intel_plane->src_w = orig.src_w;
+	intel_plane->src_h = orig.src_h;
+	intel_plane->obj = obj;
 
 	return 0;
 }
@@ -11644,6 +11985,7 @@ static const struct drm_plane_funcs intel_primary_plane_funcs = {
 	.update_plane = intel_primary_plane_setplane,
 	.disable_plane = intel_primary_plane_disable,
 	.destroy = intel_plane_destroy,
+	.set_property = intel_plane_set_property
 };
 
 static struct drm_plane *intel_primary_plane_create(struct drm_device *dev,
@@ -11661,6 +12003,7 @@ static struct drm_plane *intel_primary_plane_create(struct drm_device *dev,
 	primary->max_downscale = 1;
 	primary->pipe = pipe;
 	primary->plane = pipe;
+	primary->rotation = BIT(DRM_ROTATE_0);
 	if (HAS_FBC(dev) && INTEL_INFO(dev)->gen < 4)
 		primary->plane = !pipe;
 
@@ -11676,6 +12019,19 @@ static struct drm_plane *intel_primary_plane_create(struct drm_device *dev,
 				 &intel_primary_plane_funcs,
 				 intel_primary_formats, num_formats,
 				 DRM_PLANE_TYPE_PRIMARY);
+
+	if (INTEL_INFO(dev)->gen >= 4) {
+		if (!dev->mode_config.rotation_property)
+			dev->mode_config.rotation_property =
+				drm_mode_create_rotation_property(dev,
+							BIT(DRM_ROTATE_0) |
+							BIT(DRM_ROTATE_180));
+		if (dev->mode_config.rotation_property)
+			drm_object_attach_property(&primary->base.base,
+				dev->mode_config.rotation_property,
+				primary->rotation);
+	}
+
 	return &primary->base;
 }
 
@@ -11736,6 +12092,10 @@ intel_cursor_plane_update(struct drm_plane *plane, struct drm_crtc *crtc,
 		return intel_crtc_cursor_set_obj(crtc, obj, crtc_w, crtc_h);
 	} else {
 		intel_crtc_update_cursor(crtc, visible);
+
+		intel_frontbuffer_flip(crtc->dev,
+				       INTEL_FRONTBUFFER_CURSOR(intel_crtc->pipe));
+
 		return 0;
 	}
 }
@@ -11812,8 +12172,7 @@ static void intel_crtc_init(struct drm_device *dev, int pipe)
 
 	intel_crtc->cursor_base = ~0;
 	intel_crtc->cursor_cntl = ~0;
-
-	init_waitqueue_head(&intel_crtc->vbl_wait);
+	intel_crtc->cursor_size = ~0;
 
 	BUG_ON(pipe >= ARRAY_SIZE(dev_priv->plane_to_crtc_mapping) ||
 	       dev_priv->plane_to_crtc_mapping[intel_crtc->plane] != NULL);
@@ -11876,8 +12235,7 @@ static int intel_encoder_clones(struct intel_encoder *encoder)
 	int index_mask = 0;
 	int entry = 0;
 
-	list_for_each_entry(source_encoder,
-			    &dev->mode_config.encoder_list, base.head) {
+	for_each_intel_encoder(dev, source_encoder) {
 		if (encoders_cloneable(encoder, source_encoder))
 			index_mask |= (1 << entry);
 
@@ -12066,7 +12424,7 @@ static void intel_setup_outputs(struct drm_device *dev)
 
 	intel_edp_psr_init(dev);
 
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list, base.head) {
+	for_each_intel_encoder(dev, encoder) {
 		encoder->base.possible_crtcs = encoder->crtc_mask;
 		encoder->base.possible_clones =
 			intel_encoder_clones(encoder);
@@ -12332,29 +12690,27 @@ static void intel_init_display(struct drm_device *dev)
 		dev_priv->display.get_display_clock_speed =
 			i830_get_display_clock_speed;
 
-	if (HAS_PCH_SPLIT(dev)) {
-		if (IS_GEN5(dev)) {
-			dev_priv->display.fdi_link_train = ironlake_fdi_link_train;
-			dev_priv->display.write_eld = ironlake_write_eld;
-		} else if (IS_GEN6(dev)) {
-			dev_priv->display.fdi_link_train = gen6_fdi_link_train;
-			dev_priv->display.write_eld = ironlake_write_eld;
-			dev_priv->display.modeset_global_resources =
-				snb_modeset_global_resources;
-		} else if (IS_IVYBRIDGE(dev)) {
-			/* FIXME: detect B0+ stepping and use auto training */
-			dev_priv->display.fdi_link_train = ivb_manual_fdi_link_train;
-			dev_priv->display.write_eld = ironlake_write_eld;
-			dev_priv->display.modeset_global_resources =
-				ivb_modeset_global_resources;
-		} else if (IS_HASWELL(dev) || IS_GEN8(dev)) {
-			dev_priv->display.fdi_link_train = hsw_fdi_link_train;
-			dev_priv->display.write_eld = haswell_write_eld;
-			dev_priv->display.modeset_global_resources =
-				haswell_modeset_global_resources;
-		}
-	} else if (IS_G4X(dev)) {
+	if (IS_G4X(dev)) {
 		dev_priv->display.write_eld = g4x_write_eld;
+	} else if (IS_GEN5(dev)) {
+		dev_priv->display.fdi_link_train = ironlake_fdi_link_train;
+		dev_priv->display.write_eld = ironlake_write_eld;
+	} else if (IS_GEN6(dev)) {
+		dev_priv->display.fdi_link_train = gen6_fdi_link_train;
+		dev_priv->display.write_eld = ironlake_write_eld;
+		dev_priv->display.modeset_global_resources =
+			snb_modeset_global_resources;
+	} else if (IS_IVYBRIDGE(dev)) {
+		/* FIXME: detect B0+ stepping and use auto training */
+		dev_priv->display.fdi_link_train = ivb_manual_fdi_link_train;
+		dev_priv->display.write_eld = ironlake_write_eld;
+		dev_priv->display.modeset_global_resources =
+			ivb_modeset_global_resources;
+	} else if (IS_HASWELL(dev) || IS_BROADWELL(dev)) {
+		dev_priv->display.fdi_link_train = hsw_fdi_link_train;
+		dev_priv->display.write_eld = haswell_write_eld;
+		dev_priv->display.modeset_global_resources =
+			haswell_modeset_global_resources;
 	} else if (IS_VALLEYVIEW(dev)) {
 		dev_priv->display.modeset_global_resources =
 			valleyview_modeset_global_resources;
@@ -12388,6 +12744,8 @@ static void intel_init_display(struct drm_device *dev)
 	}
 
 	intel_panel_init_backlight_funcs(dev);
+
+	mutex_init(&dev_priv->pps_mutex);
 }
 
 /*
@@ -12403,6 +12761,14 @@ static void quirk_pipea_force(struct drm_device *dev)
 	DRM_INFO("applying pipe a force quirk\n");
 }
 
+static void quirk_pipeb_force(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	dev_priv->quirks |= QUIRK_PIPEB_FORCE;
+	DRM_INFO("applying pipe b force quirk\n");
+}
+
 /*
  * Some machines (Lenovo U160) do not work with SSC on LVDS for some reason
  */
@@ -12477,6 +12843,12 @@ static struct intel_quirk intel_quirks[] = {
 	/* ThinkPad T60 needs pipe A force quirk (bug #16494) */
 	{ 0x2782, 0x17aa, 0x201a, quirk_pipea_force },
 
+	/* 830 needs to leave pipe A & dpll A up */
+	{ 0x3577, PCI_ANY_ID, PCI_ANY_ID, quirk_pipea_force },
+
+	/* 830 needs to leave pipe B & dpll B up */
+	{ 0x3577, PCI_ANY_ID, PCI_ANY_ID, quirk_pipeb_force },
+
 	/* Lenovo U160 cannot use SSC on LVDS */
 	{ 0x0046, 0x17aa, 0x3920, quirk_ssc_force_disable },
 
@@ -12550,7 +12922,11 @@ static void i915_disable_vga(struct drm_device *dev)
 	vga_put(dev->pdev, VGA_RSRC_LEGACY_IO);
 	udelay(300);
 
-	I915_WRITE(vga_reg, VGA_DISP_DISABLE);
+	/*
+	 * Fujitsu-Siemens Lifebook S6010 (830) has problems resuming
+	 * from S3 without preserving (some of?) the other bits.
+	 */
+	I915_WRITE(vga_reg, dev_priv->bios_vgacntr | VGA_DISP_DISABLE);
 	POSTING_READ(vga_reg);
 }
 
@@ -12563,8 +12939,6 @@ void intel_modeset_init_hw(struct drm_device *dev)
 
 	intel_init_clock_gating(dev);
 
-	intel_reset_dpio(dev);
-
 	intel_enable_gt_powersave(dev);
 }
 
@@ -12610,7 +12984,10 @@ void intel_modeset_init(struct drm_device *dev)
 		dev->mode_config.max_height = 8192;
 	}
 
-	if (IS_GEN2(dev)) {
+	if (IS_845G(dev) || IS_I865G(dev)) {
+		dev->mode_config.cursor_width = IS_845G(dev) ? 64 : 512;
+		dev->mode_config.cursor_height = 1023;
+	} else if (IS_GEN2(dev)) {
 		dev->mode_config.cursor_width = GEN2_CURSOR_WIDTH;
 		dev->mode_config.cursor_height = GEN2_CURSOR_HEIGHT;
 	} else {
@@ -12624,7 +13001,7 @@ void intel_modeset_init(struct drm_device *dev)
 		      INTEL_INFO(dev)->num_pipes,
 		      INTEL_INFO(dev)->num_pipes > 1 ? "s" : "");
 
-	for_each_pipe(pipe) {
+	for_each_pipe(dev_priv, pipe) {
 		intel_crtc_init(dev, pipe);
 		for_each_sprite(pipe, sprite) {
 			ret = intel_plane_init(dev, pipe, sprite);
@@ -12635,10 +13012,11 @@ void intel_modeset_init(struct drm_device *dev)
 	}
 
 	intel_init_dpio(dev);
-	intel_reset_dpio(dev);
 
 	intel_shared_dpll_init(dev);
 
+	/* save the BIOS value before clobbering it */
+	dev_priv->bios_vgacntr = I915_READ(i915_vgacntrl_reg(dev));
 	/* Just disable it once at startup */
 	i915_disable_vga(dev);
 	intel_setup_outputs(dev);
@@ -12730,9 +13108,10 @@ static void intel_sanitize_crtc(struct intel_crtc *crtc)
 	I915_WRITE(reg, I915_READ(reg) & ~PIPECONF_FRAME_START_DELAY_MASK);
 
 	/* restore vblank interrupts to correct state */
-	if (crtc->active)
+	if (crtc->active) {
+		update_scanline_offset(crtc);
 		drm_vblank_on(dev, crtc->pipe);
-	else
+	} else
 		drm_vblank_off(dev, crtc->pipe);
 
 	/* We need to sanitize the plane -> pipe mapping first because this will
@@ -12815,7 +13194,7 @@ static void intel_sanitize_crtc(struct intel_crtc *crtc)
 		}
 	}
 
-	if (crtc->active || IS_VALLEYVIEW(dev) || INTEL_INFO(dev)->gen < 5) {
+	if (crtc->active || HAS_GMCH_DISPLAY(dev)) {
 		/*
 		 * We start out with underrun reporting disabled to avoid races.
 		 * For correct bookkeeping mark this on active crtcs.
@@ -12831,8 +13210,6 @@ static void intel_sanitize_crtc(struct intel_crtc *crtc)
 		 */
 		crtc->cpu_fifo_underrun_disabled = true;
 		crtc->pch_fifo_underrun_disabled = true;
-
-		update_scanline_offset(crtc);
 	}
 }
 
@@ -12964,8 +13341,7 @@ static void intel_modeset_readout_hw_state(struct drm_device *dev)
 			intel_display_power_get(dev_priv, POWER_DOMAIN_PLLS);
 	}
 
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list,
-			    base.head) {
+	for_each_intel_encoder(dev, encoder) {
 		pipe = 0;
 
 		if (encoder->get_hw_state(encoder, &pipe)) {
@@ -13029,12 +13405,11 @@ void intel_modeset_setup_hw_state(struct drm_device *dev,
 	}
 
 	/* HW state is read out, now we need to sanitize this mess. */
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list,
-			    base.head) {
+	for_each_intel_encoder(dev, encoder) {
 		intel_sanitize_encoder(encoder);
 	}
 
-	for_each_pipe(pipe) {
+	for_each_pipe(dev_priv, pipe) {
 		crtc = to_intel_crtc(dev_priv->pipe_to_crtc_mapping[pipe]);
 		intel_sanitize_crtc(crtc);
 		intel_dump_pipe_config(crtc, &crtc->config, "[setup_hw_state]");
@@ -13062,7 +13437,7 @@ void intel_modeset_setup_hw_state(struct drm_device *dev,
 		 * We need to use raw interfaces for restoring state to avoid
 		 * checking (bogus) intermediate states.
 		 */
-		for_each_pipe(pipe) {
+		for_each_pipe(dev_priv, pipe) {
 			struct drm_crtc *crtc =
 				dev_priv->pipe_to_crtc_mapping[pipe];
 
@@ -13283,7 +13658,7 @@ intel_display_capture_error_state(struct drm_device *dev)
 	if (IS_HASWELL(dev) || IS_BROADWELL(dev))
 		error->power_well_driver = I915_READ(HSW_PWR_WELL_DRIVER);
 
-	for_each_pipe(i) {
+	for_each_pipe(dev_priv, i) {
 		error->pipe[i].power_domain_on =
 			intel_display_power_enabled_unlocked(dev_priv,
 							   POWER_DOMAIN_PIPE(i));
@@ -13347,6 +13722,7 @@ intel_display_print_error_state(struct drm_i915_error_state_buf *m,
 				struct drm_device *dev,
 				struct intel_display_error_state *error)
 {
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	int i;
 
 	if (!error)
@@ -13356,7 +13732,7 @@ intel_display_print_error_state(struct drm_i915_error_state_buf *m,
 	if (IS_HASWELL(dev) || IS_BROADWELL(dev))
 		err_printf(m, "PWR_WELL_CTL2: %08x\n",
 			   error->power_well_driver);
-	for_each_pipe(i) {
+	for_each_pipe(dev_priv, i) {
 		err_printf(m, "Pipe [%d]:\n", i);
 		err_printf(m, "  Power: %s\n",
 			   error->pipe[i].power_domain_on ? "on" : "off");
@@ -13397,3 +13773,25 @@ intel_display_print_error_state(struct drm_i915_error_state_buf *m,
 		err_printf(m, "  VSYNC: %08x\n", error->transcoder[i].vsync);
 	}
 }
+
+void intel_modeset_preclose(struct drm_device *dev, struct drm_file *file)
+{
+	struct intel_crtc *crtc;
+
+	for_each_intel_crtc(dev, crtc) {
+		struct intel_unpin_work *work;
+		unsigned long irqflags;
+
+		spin_lock_irqsave(&dev->event_lock, irqflags);
+
+		work = crtc->unpin_work;
+
+		if (work && work->event &&
+		    work->event->base.file_priv == file) {
+			kfree(work->event);
+			work->event = NULL;
+		}
+
+		spin_unlock_irqrestore(&dev->event_lock, irqflags);
+	}
+}
diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c
index fdff1d420c14..f6a3fdd5589e 100644
--- a/drivers/gpu/drm/i915/intel_dp.c
+++ b/drivers/gpu/drm/i915/intel_dp.c
@@ -111,7 +111,7 @@ static struct intel_dp *intel_attached_dp(struct drm_connector *connector)
 }
 
 static void intel_dp_link_down(struct intel_dp *intel_dp);
-static bool _edp_panel_vdd_on(struct intel_dp *intel_dp);
+static bool edp_panel_vdd_on(struct intel_dp *intel_dp);
 static void edp_panel_vdd_off(struct intel_dp *intel_dp, bool sync);
 
 int
@@ -290,32 +290,201 @@ intel_dp_init_panel_power_sequencer_registers(struct drm_device *dev,
 					      struct intel_dp *intel_dp,
 					      struct edp_power_seq *out);
 
+static void pps_lock(struct intel_dp *intel_dp)
+{
+	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
+	struct intel_encoder *encoder = &intel_dig_port->base;
+	struct drm_device *dev = encoder->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	enum intel_display_power_domain power_domain;
+
+	/*
+	 * See vlv_power_sequencer_reset() why we need
+	 * a power domain reference here.
+	 */
+	power_domain = intel_display_port_power_domain(encoder);
+	intel_display_power_get(dev_priv, power_domain);
+
+	mutex_lock(&dev_priv->pps_mutex);
+}
+
+static void pps_unlock(struct intel_dp *intel_dp)
+{
+	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
+	struct intel_encoder *encoder = &intel_dig_port->base;
+	struct drm_device *dev = encoder->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	enum intel_display_power_domain power_domain;
+
+	mutex_unlock(&dev_priv->pps_mutex);
+
+	power_domain = intel_display_port_power_domain(encoder);
+	intel_display_power_put(dev_priv, power_domain);
+}
+
 static enum pipe
 vlv_power_sequencer_pipe(struct intel_dp *intel_dp)
 {
 	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
-	struct drm_crtc *crtc = intel_dig_port->base.base.crtc;
 	struct drm_device *dev = intel_dig_port->base.base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	enum port port = intel_dig_port->port;
-	enum pipe pipe;
+	struct intel_encoder *encoder;
+	unsigned int pipes = (1 << PIPE_A) | (1 << PIPE_B);
+	struct edp_power_seq power_seq;
+
+	lockdep_assert_held(&dev_priv->pps_mutex);
+
+	if (intel_dp->pps_pipe != INVALID_PIPE)
+		return intel_dp->pps_pipe;
+
+	/*
+	 * We don't have power sequencer currently.
+	 * Pick one that's not used by other ports.
+	 */
+	list_for_each_entry(encoder, &dev->mode_config.encoder_list,
+			    base.head) {
+		struct intel_dp *tmp;
+
+		if (encoder->type != INTEL_OUTPUT_EDP)
+			continue;
+
+		tmp = enc_to_intel_dp(&encoder->base);
+
+		if (tmp->pps_pipe != INVALID_PIPE)
+			pipes &= ~(1 << tmp->pps_pipe);
+	}
+
+	/*
+	 * Didn't find one. This should not happen since there
+	 * are two power sequencers and up to two eDP ports.
+	 */
+	if (WARN_ON(pipes == 0))
+		return PIPE_A;
+
+	intel_dp->pps_pipe = ffs(pipes) - 1;
+
+	DRM_DEBUG_KMS("picked pipe %c power sequencer for port %c\n",
+		      pipe_name(intel_dp->pps_pipe),
+		      port_name(intel_dig_port->port));
 
-	/* modeset should have pipe */
-	if (crtc)
-		return to_intel_crtc(crtc)->pipe;
+	/* init power sequencer on this pipe and port */
+	intel_dp_init_panel_power_sequencer(dev, intel_dp, &power_seq);
+	intel_dp_init_panel_power_sequencer_registers(dev, intel_dp,
+						      &power_seq);
+
+	return intel_dp->pps_pipe;
+}
+
+typedef bool (*vlv_pipe_check)(struct drm_i915_private *dev_priv,
+			       enum pipe pipe);
+
+static bool vlv_pipe_has_pp_on(struct drm_i915_private *dev_priv,
+			       enum pipe pipe)
+{
+	return I915_READ(VLV_PIPE_PP_STATUS(pipe)) & PP_ON;
+}
+
+static bool vlv_pipe_has_vdd_on(struct drm_i915_private *dev_priv,
+				enum pipe pipe)
+{
+	return I915_READ(VLV_PIPE_PP_CONTROL(pipe)) & EDP_FORCE_VDD;
+}
+
+static bool vlv_pipe_any(struct drm_i915_private *dev_priv,
+			 enum pipe pipe)
+{
+	return true;
+}
+
+static enum pipe
+vlv_initial_pps_pipe(struct drm_i915_private *dev_priv,
+		     enum port port,
+		     vlv_pipe_check pipe_check)
+{
+	enum pipe pipe;
 
-	/* init time, try to find a pipe with this port selected */
 	for (pipe = PIPE_A; pipe <= PIPE_B; pipe++) {
 		u32 port_sel = I915_READ(VLV_PIPE_PP_ON_DELAYS(pipe)) &
 			PANEL_PORT_SELECT_MASK;
-		if (port_sel == PANEL_PORT_SELECT_DPB_VLV && port == PORT_B)
-			return pipe;
-		if (port_sel == PANEL_PORT_SELECT_DPC_VLV && port == PORT_C)
-			return pipe;
+
+		if (port_sel != PANEL_PORT_SELECT_VLV(port))
+			continue;
+
+		if (!pipe_check(dev_priv, pipe))
+			continue;
+
+		return pipe;
+	}
+
+	return INVALID_PIPE;
+}
+
+static void
+vlv_initial_power_sequencer_setup(struct intel_dp *intel_dp)
+{
+	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
+	struct drm_device *dev = intel_dig_port->base.base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct edp_power_seq power_seq;
+	enum port port = intel_dig_port->port;
+
+	lockdep_assert_held(&dev_priv->pps_mutex);
+
+	/* try to find a pipe with this port selected */
+	/* first pick one where the panel is on */
+	intel_dp->pps_pipe = vlv_initial_pps_pipe(dev_priv, port,
+						  vlv_pipe_has_pp_on);
+	/* didn't find one? pick one where vdd is on */
+	if (intel_dp->pps_pipe == INVALID_PIPE)
+		intel_dp->pps_pipe = vlv_initial_pps_pipe(dev_priv, port,
+							  vlv_pipe_has_vdd_on);
+	/* didn't find one? pick one with just the correct port */
+	if (intel_dp->pps_pipe == INVALID_PIPE)
+		intel_dp->pps_pipe = vlv_initial_pps_pipe(dev_priv, port,
+							  vlv_pipe_any);
+
+	/* 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));
+		return;
 	}
 
-	/* shrug */
-	return PIPE_A;
+	DRM_DEBUG_KMS("initial power sequencer for port %c: pipe %c\n",
+		      port_name(port), pipe_name(intel_dp->pps_pipe));
+
+	intel_dp_init_panel_power_sequencer(dev, intel_dp, &power_seq);
+	intel_dp_init_panel_power_sequencer_registers(dev, intel_dp,
+						      &power_seq);
+}
+
+void vlv_power_sequencer_reset(struct drm_i915_private *dev_priv)
+{
+	struct drm_device *dev = dev_priv->dev;
+	struct intel_encoder *encoder;
+
+	if (WARN_ON(!IS_VALLEYVIEW(dev)))
+		return;
+
+	/*
+	 * We can't grab pps_mutex here due to deadlock with power_domain
+	 * mutex when power_domain functions are called while holding pps_mutex.
+	 * That also means that in order to use pps_pipe the code needs to
+	 * hold both a power domain reference and pps_mutex, and the power domain
+	 * reference get/put must be done while _not_ holding pps_mutex.
+	 * pps_{lock,unlock}() do these steps in the correct order, so one
+	 * should use them always.
+	 */
+
+	list_for_each_entry(encoder, &dev->mode_config.encoder_list, base.head) {
+		struct intel_dp *intel_dp;
+
+		if (encoder->type != INTEL_OUTPUT_EDP)
+			continue;
+
+		intel_dp = enc_to_intel_dp(&encoder->base);
+		intel_dp->pps_pipe = INVALID_PIPE;
+	}
 }
 
 static u32 _pp_ctrl_reg(struct intel_dp *intel_dp)
@@ -349,12 +518,15 @@ static int edp_notify_handler(struct notifier_block *this, unsigned long code,
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 pp_div;
 	u32 pp_ctrl_reg, pp_div_reg;
-	enum pipe pipe = vlv_power_sequencer_pipe(intel_dp);
 
 	if (!is_edp(intel_dp) || code != SYS_RESTART)
 		return 0;
 
+	pps_lock(intel_dp);
+
 	if (IS_VALLEYVIEW(dev)) {
+		enum pipe pipe = vlv_power_sequencer_pipe(intel_dp);
+
 		pp_ctrl_reg = VLV_PIPE_PP_CONTROL(pipe);
 		pp_div_reg  = VLV_PIPE_PP_DIVISOR(pipe);
 		pp_div = I915_READ(pp_div_reg);
@@ -366,6 +538,8 @@ static int edp_notify_handler(struct notifier_block *this, unsigned long code,
 		msleep(intel_dp->panel_power_cycle_delay);
 	}
 
+	pps_unlock(intel_dp);
+
 	return 0;
 }
 
@@ -374,6 +548,8 @@ static bool edp_have_panel_power(struct intel_dp *intel_dp)
 	struct drm_device *dev = intel_dp_to_dev(intel_dp);
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
+	lockdep_assert_held(&dev_priv->pps_mutex);
+
 	return (I915_READ(_pp_stat_reg(intel_dp)) & PP_ON) != 0;
 }
 
@@ -381,13 +557,10 @@ static bool edp_have_panel_vdd(struct intel_dp *intel_dp)
 {
 	struct drm_device *dev = intel_dp_to_dev(intel_dp);
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
-	struct intel_encoder *intel_encoder = &intel_dig_port->base;
-	enum intel_display_power_domain power_domain;
 
-	power_domain = intel_display_port_power_domain(intel_encoder);
-	return intel_display_power_enabled(dev_priv, power_domain) &&
-	       (I915_READ(_pp_ctrl_reg(intel_dp)) & EDP_FORCE_VDD) != 0;
+	lockdep_assert_held(&dev_priv->pps_mutex);
+
+	return I915_READ(_pp_ctrl_reg(intel_dp)) & EDP_FORCE_VDD;
 }
 
 static void
@@ -535,7 +708,15 @@ intel_dp_aux_ch(struct intel_dp *intel_dp,
 	bool has_aux_irq = HAS_AUX_IRQ(dev);
 	bool vdd;
 
-	vdd = _edp_panel_vdd_on(intel_dp);
+	pps_lock(intel_dp);
+
+	/*
+	 * We will be called with VDD already enabled for dpcd/edid/oui reads.
+	 * In such cases we want to leave VDD enabled and it's up to upper layers
+	 * to turn it off. But for eg. i2c-dev access we need to turn it on/off
+	 * ourselves.
+	 */
+	vdd = edp_panel_vdd_on(intel_dp);
 
 	/* dp aux is extremely sensitive to irq latency, hence request the
 	 * lowest possible wakeup latency and so prevent the cpu from going into
@@ -644,6 +825,8 @@ out:
 	if (vdd)
 		edp_panel_vdd_off(intel_dp, false);
 
+	pps_unlock(intel_dp);
+
 	return ret;
 }
 
@@ -828,20 +1011,6 @@ intel_dp_set_clock(struct intel_encoder *encoder,
 	}
 }
 
-static void
-intel_dp_set_m2_n2(struct intel_crtc *crtc, struct intel_link_m_n *m_n)
-{
-	struct drm_device *dev = crtc->base.dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	enum transcoder transcoder = crtc->config.cpu_transcoder;
-
-	I915_WRITE(PIPE_DATA_M2(transcoder),
-		TU_SIZE(m_n->tu) | m_n->gmch_m);
-	I915_WRITE(PIPE_DATA_N2(transcoder), m_n->gmch_n);
-	I915_WRITE(PIPE_LINK_M2(transcoder), m_n->link_m);
-	I915_WRITE(PIPE_LINK_N2(transcoder), m_n->link_n);
-}
-
 bool
 intel_dp_compute_config(struct intel_encoder *encoder,
 			struct intel_crtc_config *pipe_config)
@@ -867,6 +1036,7 @@ intel_dp_compute_config(struct intel_encoder *encoder,
 		pipe_config->has_pch_encoder = true;
 
 	pipe_config->has_dp_encoder = true;
+	pipe_config->has_drrs = false;
 	pipe_config->has_audio = intel_dp->has_audio;
 
 	if (is_edp(intel_dp) && intel_connector->panel.fixed_mode) {
@@ -898,23 +1068,15 @@ intel_dp_compute_config(struct intel_encoder *encoder,
 			bpp = dev_priv->vbt.edp_bpp;
 		}
 
-		if (IS_BROADWELL(dev)) {
-			/* Yes, it's an ugly hack. */
-			min_lane_count = max_lane_count;
-			DRM_DEBUG_KMS("forcing lane count to max (%u) on BDW\n",
-				      min_lane_count);
-		} else if (dev_priv->vbt.edp_lanes) {
-			min_lane_count = min(dev_priv->vbt.edp_lanes,
-					     max_lane_count);
-			DRM_DEBUG_KMS("using min %u lanes per VBT\n",
-				      min_lane_count);
-		}
-
-		if (dev_priv->vbt.edp_rate) {
-			min_clock = min(dev_priv->vbt.edp_rate >> 3, max_clock);
-			DRM_DEBUG_KMS("using min %02x link bw per VBT\n",
-				      bws[min_clock]);
-		}
+		/*
+		 * Use the maximum clock and number of lanes the eDP panel
+		 * advertizes being capable of. The panels are generally
+		 * designed to support only a single clock and lane
+		 * configuration, and typically these values correspond to the
+		 * native resolution of the panel.
+		 */
+		min_lane_count = max_lane_count;
+		min_clock = max_clock;
 	}
 
 	for (; bpp >= 6*3; bpp -= 2*3) {
@@ -970,13 +1132,14 @@ found:
 
 	if (intel_connector->panel.downclock_mode != NULL &&
 		intel_dp->drrs_state.type == SEAMLESS_DRRS_SUPPORT) {
+			pipe_config->has_drrs = true;
 			intel_link_compute_m_n(bpp, lane_count,
 				intel_connector->panel.downclock_mode->clock,
 				pipe_config->port_clock,
 				&pipe_config->dp_m2_n2);
 	}
 
-	if (HAS_DDI(dev))
+	if (IS_HASWELL(dev) || IS_BROADWELL(dev))
 		hsw_dp_set_ddi_pll_sel(pipe_config, intel_dp->link_bw);
 	else
 		intel_dp_set_clock(encoder, pipe_config, intel_dp->link_bw);
@@ -1110,6 +1273,8 @@ static void wait_panel_status(struct intel_dp *intel_dp,
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 pp_stat_reg, pp_ctrl_reg;
 
+	lockdep_assert_held(&dev_priv->pps_mutex);
+
 	pp_stat_reg = _pp_stat_reg(intel_dp);
 	pp_ctrl_reg = _pp_ctrl_reg(intel_dp);
 
@@ -1173,13 +1338,20 @@ static  u32 ironlake_get_pp_control(struct intel_dp *intel_dp)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 control;
 
+	lockdep_assert_held(&dev_priv->pps_mutex);
+
 	control = I915_READ(_pp_ctrl_reg(intel_dp));
 	control &= ~PANEL_UNLOCK_MASK;
 	control |= PANEL_UNLOCK_REGS;
 	return control;
 }
 
-static bool _edp_panel_vdd_on(struct intel_dp *intel_dp)
+/*
+ * Must be paired with edp_panel_vdd_off().
+ * Must hold pps_mutex around the whole on/off sequence.
+ * Can be nested with intel_edp_panel_vdd_{on,off}() calls.
+ */
+static bool edp_panel_vdd_on(struct intel_dp *intel_dp)
 {
 	struct drm_device *dev = intel_dp_to_dev(intel_dp);
 	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
@@ -1190,6 +1362,8 @@ static bool _edp_panel_vdd_on(struct intel_dp *intel_dp)
 	u32 pp_stat_reg, pp_ctrl_reg;
 	bool need_to_disable = !intel_dp->want_panel_vdd;
 
+	lockdep_assert_held(&dev_priv->pps_mutex);
+
 	if (!is_edp(intel_dp))
 		return false;
 
@@ -1227,62 +1401,76 @@ static bool _edp_panel_vdd_on(struct intel_dp *intel_dp)
 	return need_to_disable;
 }
 
+/*
+ * Must be paired with intel_edp_panel_vdd_off() or
+ * intel_edp_panel_off().
+ * Nested calls to these functions are not allowed since
+ * we drop the lock. Caller must use some higher level
+ * locking to prevent nested calls from other threads.
+ */
 void intel_edp_panel_vdd_on(struct intel_dp *intel_dp)
 {
-	if (is_edp(intel_dp)) {
-		bool vdd = _edp_panel_vdd_on(intel_dp);
+	bool vdd;
 
-		WARN(!vdd, "eDP VDD already requested on\n");
-	}
+	if (!is_edp(intel_dp))
+		return;
+
+	pps_lock(intel_dp);
+	vdd = edp_panel_vdd_on(intel_dp);
+	pps_unlock(intel_dp);
+
+	WARN(!vdd, "eDP VDD already requested on\n");
 }
 
 static void edp_panel_vdd_off_sync(struct intel_dp *intel_dp)
 {
 	struct drm_device *dev = intel_dp_to_dev(intel_dp);
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_digital_port *intel_dig_port =
+		dp_to_dig_port(intel_dp);
+	struct intel_encoder *intel_encoder = &intel_dig_port->base;
+	enum intel_display_power_domain power_domain;
 	u32 pp;
 	u32 pp_stat_reg, pp_ctrl_reg;
 
-	WARN_ON(!drm_modeset_is_locked(&dev->mode_config.connection_mutex));
+	lockdep_assert_held(&dev_priv->pps_mutex);
 
-	if (!intel_dp->want_panel_vdd && edp_have_panel_vdd(intel_dp)) {
-		struct intel_digital_port *intel_dig_port =
-						dp_to_dig_port(intel_dp);
-		struct intel_encoder *intel_encoder = &intel_dig_port->base;
-		enum intel_display_power_domain power_domain;
+	WARN_ON(intel_dp->want_panel_vdd);
+
+	if (!edp_have_panel_vdd(intel_dp))
+		return;
 
-		DRM_DEBUG_KMS("Turning eDP VDD off\n");
+	DRM_DEBUG_KMS("Turning eDP VDD off\n");
 
-		pp = ironlake_get_pp_control(intel_dp);
-		pp &= ~EDP_FORCE_VDD;
+	pp = ironlake_get_pp_control(intel_dp);
+	pp &= ~EDP_FORCE_VDD;
 
-		pp_ctrl_reg = _pp_ctrl_reg(intel_dp);
-		pp_stat_reg = _pp_stat_reg(intel_dp);
+	pp_ctrl_reg = _pp_ctrl_reg(intel_dp);
+	pp_stat_reg = _pp_stat_reg(intel_dp);
 
-		I915_WRITE(pp_ctrl_reg, pp);
-		POSTING_READ(pp_ctrl_reg);
+	I915_WRITE(pp_ctrl_reg, pp);
+	POSTING_READ(pp_ctrl_reg);
 
-		/* Make sure sequencer is idle before allowing subsequent activity */
-		DRM_DEBUG_KMS("PP_STATUS: 0x%08x PP_CONTROL: 0x%08x\n",
-		I915_READ(pp_stat_reg), I915_READ(pp_ctrl_reg));
+	/* Make sure sequencer is idle before allowing subsequent activity */
+	DRM_DEBUG_KMS("PP_STATUS: 0x%08x PP_CONTROL: 0x%08x\n",
+	I915_READ(pp_stat_reg), I915_READ(pp_ctrl_reg));
 
-		if ((pp & POWER_TARGET_ON) == 0)
-			intel_dp->last_power_cycle = jiffies;
+	if ((pp & POWER_TARGET_ON) == 0)
+		intel_dp->last_power_cycle = jiffies;
 
-		power_domain = intel_display_port_power_domain(intel_encoder);
-		intel_display_power_put(dev_priv, power_domain);
-	}
+	power_domain = intel_display_port_power_domain(intel_encoder);
+	intel_display_power_put(dev_priv, power_domain);
 }
 
 static void edp_panel_vdd_work(struct work_struct *__work)
 {
 	struct intel_dp *intel_dp = container_of(to_delayed_work(__work),
 						 struct intel_dp, panel_vdd_work);
-	struct drm_device *dev = intel_dp_to_dev(intel_dp);
 
-	drm_modeset_lock(&dev->mode_config.connection_mutex, NULL);
-	edp_panel_vdd_off_sync(intel_dp);
-	drm_modeset_unlock(&dev->mode_config.connection_mutex);
+	pps_lock(intel_dp);
+	if (!intel_dp->want_panel_vdd)
+		edp_panel_vdd_off_sync(intel_dp);
+	pps_unlock(intel_dp);
 }
 
 static void edp_panel_vdd_schedule_off(struct intel_dp *intel_dp)
@@ -1298,8 +1486,18 @@ static void edp_panel_vdd_schedule_off(struct intel_dp *intel_dp)
 	schedule_delayed_work(&intel_dp->panel_vdd_work, delay);
 }
 
+/*
+ * Must be paired with edp_panel_vdd_on().
+ * Must hold pps_mutex around the whole on/off sequence.
+ * Can be nested with intel_edp_panel_vdd_{on,off}() calls.
+ */
 static void edp_panel_vdd_off(struct intel_dp *intel_dp, bool sync)
 {
+	struct drm_i915_private *dev_priv =
+		intel_dp_to_dev(intel_dp)->dev_private;
+
+	lockdep_assert_held(&dev_priv->pps_mutex);
+
 	if (!is_edp(intel_dp))
 		return;
 
@@ -1313,6 +1511,22 @@ static void edp_panel_vdd_off(struct intel_dp *intel_dp, bool sync)
 		edp_panel_vdd_schedule_off(intel_dp);
 }
 
+/*
+ * Must be paired with intel_edp_panel_vdd_on().
+ * Nested calls to these functions are not allowed since
+ * we drop the lock. Caller must use some higher level
+ * locking to prevent nested calls from other threads.
+ */
+static void intel_edp_panel_vdd_off(struct intel_dp *intel_dp, bool sync)
+{
+	if (!is_edp(intel_dp))
+		return;
+
+	pps_lock(intel_dp);
+	edp_panel_vdd_off(intel_dp, sync);
+	pps_unlock(intel_dp);
+}
+
 void intel_edp_panel_on(struct intel_dp *intel_dp)
 {
 	struct drm_device *dev = intel_dp_to_dev(intel_dp);
@@ -1325,9 +1539,11 @@ void intel_edp_panel_on(struct intel_dp *intel_dp)
 
 	DRM_DEBUG_KMS("Turn eDP power on\n");
 
+	pps_lock(intel_dp);
+
 	if (edp_have_panel_power(intel_dp)) {
 		DRM_DEBUG_KMS("eDP power already on\n");
-		return;
+		goto out;
 	}
 
 	wait_panel_power_cycle(intel_dp);
@@ -1356,6 +1572,9 @@ void intel_edp_panel_on(struct intel_dp *intel_dp)
 		I915_WRITE(pp_ctrl_reg, pp);
 		POSTING_READ(pp_ctrl_reg);
 	}
+
+ out:
+	pps_unlock(intel_dp);
 }
 
 void intel_edp_panel_off(struct intel_dp *intel_dp)
@@ -1373,6 +1592,8 @@ void intel_edp_panel_off(struct intel_dp *intel_dp)
 
 	DRM_DEBUG_KMS("Turn eDP power off\n");
 
+	pps_lock(intel_dp);
+
 	WARN(!intel_dp->want_panel_vdd, "Need VDD to turn off panel\n");
 
 	pp = ironlake_get_pp_control(intel_dp);
@@ -1394,9 +1615,12 @@ void intel_edp_panel_off(struct intel_dp *intel_dp)
 	/* We got a reference when we enabled the VDD. */
 	power_domain = intel_display_port_power_domain(intel_encoder);
 	intel_display_power_put(dev_priv, power_domain);
+
+	pps_unlock(intel_dp);
 }
 
-void intel_edp_backlight_on(struct intel_dp *intel_dp)
+/* Enable backlight in the panel power control. */
+static void _intel_edp_backlight_on(struct intel_dp *intel_dp)
 {
 	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
 	struct drm_device *dev = intel_dig_port->base.base.dev;
@@ -1404,13 +1628,6 @@ void intel_edp_backlight_on(struct intel_dp *intel_dp)
 	u32 pp;
 	u32 pp_ctrl_reg;
 
-	if (!is_edp(intel_dp))
-		return;
-
-	DRM_DEBUG_KMS("\n");
-
-	intel_panel_enable_backlight(intel_dp->attached_connector);
-
 	/*
 	 * If we enable the backlight right away following a panel power
 	 * on, we may see slight flicker as the panel syncs with the eDP
@@ -1418,6 +1635,9 @@ void intel_edp_backlight_on(struct intel_dp *intel_dp)
 	 * allowing it to appear.
 	 */
 	wait_backlight_on(intel_dp);
+
+	pps_lock(intel_dp);
+
 	pp = ironlake_get_pp_control(intel_dp);
 	pp |= EDP_BLC_ENABLE;
 
@@ -1425,9 +1645,24 @@ void intel_edp_backlight_on(struct intel_dp *intel_dp)
 
 	I915_WRITE(pp_ctrl_reg, pp);
 	POSTING_READ(pp_ctrl_reg);
+
+	pps_unlock(intel_dp);
 }
 
-void intel_edp_backlight_off(struct intel_dp *intel_dp)
+/* Enable backlight PWM and backlight PP control. */
+void intel_edp_backlight_on(struct intel_dp *intel_dp)
+{
+	if (!is_edp(intel_dp))
+		return;
+
+	DRM_DEBUG_KMS("\n");
+
+	intel_panel_enable_backlight(intel_dp->attached_connector);
+	_intel_edp_backlight_on(intel_dp);
+}
+
+/* Disable backlight in the panel power control. */
+static void _intel_edp_backlight_off(struct intel_dp *intel_dp)
 {
 	struct drm_device *dev = intel_dp_to_dev(intel_dp);
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -1437,7 +1672,8 @@ void intel_edp_backlight_off(struct intel_dp *intel_dp)
 	if (!is_edp(intel_dp))
 		return;
 
-	DRM_DEBUG_KMS("\n");
+	pps_lock(intel_dp);
+
 	pp = ironlake_get_pp_control(intel_dp);
 	pp &= ~EDP_BLC_ENABLE;
 
@@ -1445,13 +1681,51 @@ void intel_edp_backlight_off(struct intel_dp *intel_dp)
 
 	I915_WRITE(pp_ctrl_reg, pp);
 	POSTING_READ(pp_ctrl_reg);
-	intel_dp->last_backlight_off = jiffies;
 
+	pps_unlock(intel_dp);
+
+	intel_dp->last_backlight_off = jiffies;
 	edp_wait_backlight_off(intel_dp);
+}
+
+/* Disable backlight PP control and backlight PWM. */
+void intel_edp_backlight_off(struct intel_dp *intel_dp)
+{
+	if (!is_edp(intel_dp))
+		return;
+
+	DRM_DEBUG_KMS("\n");
 
+	_intel_edp_backlight_off(intel_dp);
 	intel_panel_disable_backlight(intel_dp->attached_connector);
 }
 
+/*
+ * Hook for controlling the panel power control backlight through the bl_power
+ * sysfs attribute. Take care to handle multiple calls.
+ */
+static void intel_edp_backlight_power(struct intel_connector *connector,
+				      bool enable)
+{
+	struct intel_dp *intel_dp = intel_attached_dp(&connector->base);
+	bool is_enabled;
+
+	pps_lock(intel_dp);
+	is_enabled = ironlake_get_pp_control(intel_dp) & EDP_BLC_ENABLE;
+	pps_unlock(intel_dp);
+
+	if (is_enabled == enable)
+		return;
+
+	DRM_DEBUG_KMS("panel power control backlight %s\n",
+		      enable ? "enable" : "disable");
+
+	if (enable)
+		_intel_edp_backlight_on(intel_dp);
+	else
+		_intel_edp_backlight_off(intel_dp);
+}
+
 static void ironlake_edp_pll_on(struct intel_dp *intel_dp)
 {
 	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
@@ -1515,8 +1789,6 @@ void intel_dp_sink_dpms(struct intel_dp *intel_dp, int mode)
 	if (mode != DRM_MODE_DPMS_ON) {
 		ret = drm_dp_dpcd_writeb(&intel_dp->aux, DP_SET_POWER,
 					 DP_SET_POWER_D3);
-		if (ret != 1)
-			DRM_DEBUG_DRIVER("failed to write sink power state\n");
 	} else {
 		/*
 		 * When turning on, we need to retry for 1ms to give the sink
@@ -1530,6 +1802,10 @@ void intel_dp_sink_dpms(struct intel_dp *intel_dp, int mode)
 			msleep(1);
 		}
 	}
+
+	if (ret != 1)
+		DRM_DEBUG_KMS("failed to %s sink power state\n",
+			      mode == DRM_MODE_DPMS_ON ? "enable" : "disable");
 }
 
 static bool intel_dp_get_hw_state(struct intel_encoder *encoder,
@@ -1576,7 +1852,7 @@ static bool intel_dp_get_hw_state(struct intel_encoder *encoder,
 			return true;
 		}
 
-		for_each_pipe(i) {
+		for_each_pipe(dev_priv, i) {
 			trans_dp = I915_READ(TRANS_DP_CTL(i));
 			if ((trans_dp & TRANS_DP_PORT_SEL_MASK) == trans_sel) {
 				*pipe = i;
@@ -2036,7 +2312,6 @@ void intel_edp_psr_init(struct drm_device *dev)
 static void intel_disable_dp(struct intel_encoder *encoder)
 {
 	struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
-	enum port port = dp_to_dig_port(intel_dp)->port;
 	struct drm_device *dev = encoder->base.dev;
 
 	/* Make sure the panel is off before trying to change the mode. But also
@@ -2046,21 +2321,19 @@ static void intel_disable_dp(struct intel_encoder *encoder)
 	intel_dp_sink_dpms(intel_dp, DRM_MODE_DPMS_OFF);
 	intel_edp_panel_off(intel_dp);
 
-	/* cpu edp my only be disable _after_ the cpu pipe/plane is disabled. */
-	if (!(port == PORT_A || IS_VALLEYVIEW(dev)))
+	/* disable the port before the pipe on g4x */
+	if (INTEL_INFO(dev)->gen < 5)
 		intel_dp_link_down(intel_dp);
 }
 
-static void g4x_post_disable_dp(struct intel_encoder *encoder)
+static void ilk_post_disable_dp(struct intel_encoder *encoder)
 {
 	struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
 	enum port port = dp_to_dig_port(intel_dp)->port;
 
-	if (port != PORT_A)
-		return;
-
 	intel_dp_link_down(intel_dp);
-	ironlake_edp_pll_off(intel_dp);
+	if (port == PORT_A)
+		ironlake_edp_pll_off(intel_dp);
 }
 
 static void vlv_post_disable_dp(struct intel_encoder *encoder)
@@ -2106,6 +2379,104 @@ static void chv_post_disable_dp(struct intel_encoder *encoder)
 	mutex_unlock(&dev_priv->dpio_lock);
 }
 
+static void
+_intel_dp_set_link_train(struct intel_dp *intel_dp,
+			 uint32_t *DP,
+			 uint8_t dp_train_pat)
+{
+	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
+	struct drm_device *dev = intel_dig_port->base.base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	enum port port = intel_dig_port->port;
+
+	if (HAS_DDI(dev)) {
+		uint32_t temp = I915_READ(DP_TP_CTL(port));
+
+		if (dp_train_pat & DP_LINK_SCRAMBLING_DISABLE)
+			temp |= DP_TP_CTL_SCRAMBLE_DISABLE;
+		else
+			temp &= ~DP_TP_CTL_SCRAMBLE_DISABLE;
+
+		temp &= ~DP_TP_CTL_LINK_TRAIN_MASK;
+		switch (dp_train_pat & DP_TRAINING_PATTERN_MASK) {
+		case DP_TRAINING_PATTERN_DISABLE:
+			temp |= DP_TP_CTL_LINK_TRAIN_NORMAL;
+
+			break;
+		case DP_TRAINING_PATTERN_1:
+			temp |= DP_TP_CTL_LINK_TRAIN_PAT1;
+			break;
+		case DP_TRAINING_PATTERN_2:
+			temp |= DP_TP_CTL_LINK_TRAIN_PAT2;
+			break;
+		case DP_TRAINING_PATTERN_3:
+			temp |= DP_TP_CTL_LINK_TRAIN_PAT3;
+			break;
+		}
+		I915_WRITE(DP_TP_CTL(port), temp);
+
+	} else if (HAS_PCH_CPT(dev) && (IS_GEN7(dev) || port != PORT_A)) {
+		*DP &= ~DP_LINK_TRAIN_MASK_CPT;
+
+		switch (dp_train_pat & DP_TRAINING_PATTERN_MASK) {
+		case DP_TRAINING_PATTERN_DISABLE:
+			*DP |= DP_LINK_TRAIN_OFF_CPT;
+			break;
+		case DP_TRAINING_PATTERN_1:
+			*DP |= DP_LINK_TRAIN_PAT_1_CPT;
+			break;
+		case DP_TRAINING_PATTERN_2:
+			*DP |= DP_LINK_TRAIN_PAT_2_CPT;
+			break;
+		case DP_TRAINING_PATTERN_3:
+			DRM_ERROR("DP training pattern 3 not supported\n");
+			*DP |= DP_LINK_TRAIN_PAT_2_CPT;
+			break;
+		}
+
+	} else {
+		if (IS_CHERRYVIEW(dev))
+			*DP &= ~DP_LINK_TRAIN_MASK_CHV;
+		else
+			*DP &= ~DP_LINK_TRAIN_MASK;
+
+		switch (dp_train_pat & DP_TRAINING_PATTERN_MASK) {
+		case DP_TRAINING_PATTERN_DISABLE:
+			*DP |= DP_LINK_TRAIN_OFF;
+			break;
+		case DP_TRAINING_PATTERN_1:
+			*DP |= DP_LINK_TRAIN_PAT_1;
+			break;
+		case DP_TRAINING_PATTERN_2:
+			*DP |= DP_LINK_TRAIN_PAT_2;
+			break;
+		case DP_TRAINING_PATTERN_3:
+			if (IS_CHERRYVIEW(dev)) {
+				*DP |= DP_LINK_TRAIN_PAT_3_CHV;
+			} else {
+				DRM_ERROR("DP training pattern 3 not supported\n");
+				*DP |= DP_LINK_TRAIN_PAT_2;
+			}
+			break;
+		}
+	}
+}
+
+static void intel_dp_enable_port(struct intel_dp *intel_dp)
+{
+	struct drm_device *dev = intel_dp_to_dev(intel_dp);
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	intel_dp->DP |= DP_PORT_EN;
+
+	/* enable with pattern 1 (as per spec) */
+	_intel_dp_set_link_train(intel_dp, &intel_dp->DP,
+				 DP_TRAINING_PATTERN_1);
+
+	I915_WRITE(intel_dp->output_reg, intel_dp->DP);
+	POSTING_READ(intel_dp->output_reg);
+}
+
 static void intel_enable_dp(struct intel_encoder *encoder)
 {
 	struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
@@ -2116,11 +2487,12 @@ static void intel_enable_dp(struct intel_encoder *encoder)
 	if (WARN_ON(dp_reg & DP_PORT_EN))
 		return;
 
+	intel_dp_enable_port(intel_dp);
 	intel_edp_panel_vdd_on(intel_dp);
+	intel_edp_panel_on(intel_dp);
+	intel_edp_panel_vdd_off(intel_dp, true);
 	intel_dp_sink_dpms(intel_dp, DRM_MODE_DPMS_ON);
 	intel_dp_start_link_train(intel_dp);
-	intel_edp_panel_on(intel_dp);
-	edp_panel_vdd_off(intel_dp, true);
 	intel_dp_complete_link_train(intel_dp);
 	intel_dp_stop_link_train(intel_dp);
 }
@@ -2154,6 +2526,78 @@ static void g4x_pre_enable_dp(struct intel_encoder *encoder)
 	}
 }
 
+static void vlv_steal_power_sequencer(struct drm_device *dev,
+				      enum pipe pipe)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_encoder *encoder;
+
+	lockdep_assert_held(&dev_priv->pps_mutex);
+
+	list_for_each_entry(encoder, &dev->mode_config.encoder_list,
+			    base.head) {
+		struct intel_dp *intel_dp;
+		enum port port;
+
+		if (encoder->type != INTEL_OUTPUT_EDP)
+			continue;
+
+		intel_dp = enc_to_intel_dp(&encoder->base);
+		port = dp_to_dig_port(intel_dp)->port;
+
+		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));
+
+		/* make sure vdd is off before we steal it */
+		edp_panel_vdd_off_sync(intel_dp);
+
+		intel_dp->pps_pipe = INVALID_PIPE;
+	}
+}
+
+static void vlv_init_panel_power_sequencer(struct intel_dp *intel_dp)
+{
+	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
+	struct intel_encoder *encoder = &intel_dig_port->base;
+	struct drm_device *dev = encoder->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_crtc *crtc = to_intel_crtc(encoder->base.crtc);
+	struct edp_power_seq power_seq;
+
+	lockdep_assert_held(&dev_priv->pps_mutex);
+
+	if (intel_dp->pps_pipe == crtc->pipe)
+		return;
+
+	/*
+	 * If another power sequencer was being used on this
+	 * port previously make sure to turn off vdd there while
+	 * we still have control of it.
+	 */
+	if (intel_dp->pps_pipe != INVALID_PIPE)
+		edp_panel_vdd_off_sync(intel_dp);
+
+	/*
+	 * We may be stealing the power
+	 * sequencer from another port.
+	 */
+	vlv_steal_power_sequencer(dev, crtc->pipe);
+
+	/* 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(intel_dig_port->port));
+
+	/* init power sequencer on this pipe and port */
+	intel_dp_init_panel_power_sequencer(dev, intel_dp, &power_seq);
+	intel_dp_init_panel_power_sequencer_registers(dev, intel_dp,
+						      &power_seq);
+}
+
 static void vlv_pre_enable_dp(struct intel_encoder *encoder)
 {
 	struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
@@ -2163,7 +2607,6 @@ static void vlv_pre_enable_dp(struct intel_encoder *encoder)
 	struct intel_crtc *intel_crtc = to_intel_crtc(encoder->base.crtc);
 	enum dpio_channel port = vlv_dport_to_channel(dport);
 	int pipe = intel_crtc->pipe;
-	struct edp_power_seq power_seq;
 	u32 val;
 
 	mutex_lock(&dev_priv->dpio_lock);
@@ -2182,10 +2625,9 @@ static void vlv_pre_enable_dp(struct intel_encoder *encoder)
 	mutex_unlock(&dev_priv->dpio_lock);
 
 	if (is_edp(intel_dp)) {
-		/* init power sequencer on this pipe and port */
-		intel_dp_init_panel_power_sequencer(dev, intel_dp, &power_seq);
-		intel_dp_init_panel_power_sequencer_registers(dev, intel_dp,
-							      &power_seq);
+		pps_lock(intel_dp);
+		vlv_init_panel_power_sequencer(intel_dp);
+		pps_unlock(intel_dp);
 	}
 
 	intel_enable_dp(encoder);
@@ -2229,7 +2671,6 @@ static void chv_pre_enable_dp(struct intel_encoder *encoder)
 	struct intel_digital_port *dport = dp_to_dig_port(intel_dp);
 	struct drm_device *dev = encoder->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct edp_power_seq power_seq;
 	struct intel_crtc *intel_crtc =
 		to_intel_crtc(encoder->base.crtc);
 	enum dpio_channel ch = vlv_dport_to_channel(dport);
@@ -2275,10 +2716,9 @@ static void chv_pre_enable_dp(struct intel_encoder *encoder)
 	mutex_unlock(&dev_priv->dpio_lock);
 
 	if (is_edp(intel_dp)) {
-		/* init power sequencer on this pipe and port */
-		intel_dp_init_panel_power_sequencer(dev, intel_dp, &power_seq);
-		intel_dp_init_panel_power_sequencer_registers(dev, intel_dp,
-							      &power_seq);
+		pps_lock(intel_dp);
+		vlv_init_panel_power_sequencer(intel_dp);
+		pps_unlock(intel_dp);
 	}
 
 	intel_enable_dp(encoder);
@@ -2297,6 +2737,8 @@ static void chv_dp_pre_pll_enable(struct intel_encoder *encoder)
 	enum pipe pipe = intel_crtc->pipe;
 	u32 val;
 
+	intel_dp_prepare(encoder);
+
 	mutex_lock(&dev_priv->dpio_lock);
 
 	/* program left/right clock distribution */
@@ -2395,13 +2837,13 @@ intel_dp_voltage_max(struct intel_dp *intel_dp)
 	enum port port = dp_to_dig_port(intel_dp)->port;
 
 	if (IS_VALLEYVIEW(dev))
-		return DP_TRAIN_VOLTAGE_SWING_1200;
+		return DP_TRAIN_VOLTAGE_SWING_LEVEL_3;
 	else if (IS_GEN7(dev) && port == PORT_A)
-		return DP_TRAIN_VOLTAGE_SWING_800;
+		return DP_TRAIN_VOLTAGE_SWING_LEVEL_2;
 	else if (HAS_PCH_CPT(dev) && port != PORT_A)
-		return DP_TRAIN_VOLTAGE_SWING_1200;
+		return DP_TRAIN_VOLTAGE_SWING_LEVEL_3;
 	else
-		return DP_TRAIN_VOLTAGE_SWING_800;
+		return DP_TRAIN_VOLTAGE_SWING_LEVEL_2;
 }
 
 static uint8_t
@@ -2412,49 +2854,49 @@ intel_dp_pre_emphasis_max(struct intel_dp *intel_dp, uint8_t voltage_swing)
 
 	if (IS_HASWELL(dev) || IS_BROADWELL(dev)) {
 		switch (voltage_swing & DP_TRAIN_VOLTAGE_SWING_MASK) {
-		case DP_TRAIN_VOLTAGE_SWING_400:
-			return DP_TRAIN_PRE_EMPHASIS_9_5;
-		case DP_TRAIN_VOLTAGE_SWING_600:
-			return DP_TRAIN_PRE_EMPHASIS_6;
-		case DP_TRAIN_VOLTAGE_SWING_800:
-			return DP_TRAIN_PRE_EMPHASIS_3_5;
-		case DP_TRAIN_VOLTAGE_SWING_1200:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_0:
+			return DP_TRAIN_PRE_EMPH_LEVEL_3;
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_1:
+			return DP_TRAIN_PRE_EMPH_LEVEL_2;
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_2:
+			return DP_TRAIN_PRE_EMPH_LEVEL_1;
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_3:
 		default:
-			return DP_TRAIN_PRE_EMPHASIS_0;
+			return DP_TRAIN_PRE_EMPH_LEVEL_0;
 		}
 	} else if (IS_VALLEYVIEW(dev)) {
 		switch (voltage_swing & DP_TRAIN_VOLTAGE_SWING_MASK) {
-		case DP_TRAIN_VOLTAGE_SWING_400:
-			return DP_TRAIN_PRE_EMPHASIS_9_5;
-		case DP_TRAIN_VOLTAGE_SWING_600:
-			return DP_TRAIN_PRE_EMPHASIS_6;
-		case DP_TRAIN_VOLTAGE_SWING_800:
-			return DP_TRAIN_PRE_EMPHASIS_3_5;
-		case DP_TRAIN_VOLTAGE_SWING_1200:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_0:
+			return DP_TRAIN_PRE_EMPH_LEVEL_3;
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_1:
+			return DP_TRAIN_PRE_EMPH_LEVEL_2;
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_2:
+			return DP_TRAIN_PRE_EMPH_LEVEL_1;
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_3:
 		default:
-			return DP_TRAIN_PRE_EMPHASIS_0;
+			return DP_TRAIN_PRE_EMPH_LEVEL_0;
 		}
 	} else if (IS_GEN7(dev) && port == PORT_A) {
 		switch (voltage_swing & DP_TRAIN_VOLTAGE_SWING_MASK) {
-		case DP_TRAIN_VOLTAGE_SWING_400:
-			return DP_TRAIN_PRE_EMPHASIS_6;
-		case DP_TRAIN_VOLTAGE_SWING_600:
-		case DP_TRAIN_VOLTAGE_SWING_800:
-			return DP_TRAIN_PRE_EMPHASIS_3_5;
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_0:
+			return DP_TRAIN_PRE_EMPH_LEVEL_2;
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_1:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_2:
+			return DP_TRAIN_PRE_EMPH_LEVEL_1;
 		default:
-			return DP_TRAIN_PRE_EMPHASIS_0;
+			return DP_TRAIN_PRE_EMPH_LEVEL_0;
 		}
 	} else {
 		switch (voltage_swing & DP_TRAIN_VOLTAGE_SWING_MASK) {
-		case DP_TRAIN_VOLTAGE_SWING_400:
-			return DP_TRAIN_PRE_EMPHASIS_6;
-		case DP_TRAIN_VOLTAGE_SWING_600:
-			return DP_TRAIN_PRE_EMPHASIS_6;
-		case DP_TRAIN_VOLTAGE_SWING_800:
-			return DP_TRAIN_PRE_EMPHASIS_3_5;
-		case DP_TRAIN_VOLTAGE_SWING_1200:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_0:
+			return DP_TRAIN_PRE_EMPH_LEVEL_2;
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_1:
+			return DP_TRAIN_PRE_EMPH_LEVEL_2;
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_2:
+			return DP_TRAIN_PRE_EMPH_LEVEL_1;
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_3:
 		default:
-			return DP_TRAIN_PRE_EMPHASIS_0;
+			return DP_TRAIN_PRE_EMPH_LEVEL_0;
 		}
 	}
 }
@@ -2473,22 +2915,22 @@ static uint32_t intel_vlv_signal_levels(struct intel_dp *intel_dp)
 	int pipe = intel_crtc->pipe;
 
 	switch (train_set & DP_TRAIN_PRE_EMPHASIS_MASK) {
-	case DP_TRAIN_PRE_EMPHASIS_0:
+	case DP_TRAIN_PRE_EMPH_LEVEL_0:
 		preemph_reg_value = 0x0004000;
 		switch (train_set & DP_TRAIN_VOLTAGE_SWING_MASK) {
-		case DP_TRAIN_VOLTAGE_SWING_400:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_0:
 			demph_reg_value = 0x2B405555;
 			uniqtranscale_reg_value = 0x552AB83A;
 			break;
-		case DP_TRAIN_VOLTAGE_SWING_600:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_1:
 			demph_reg_value = 0x2B404040;
 			uniqtranscale_reg_value = 0x5548B83A;
 			break;
-		case DP_TRAIN_VOLTAGE_SWING_800:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_2:
 			demph_reg_value = 0x2B245555;
 			uniqtranscale_reg_value = 0x5560B83A;
 			break;
-		case DP_TRAIN_VOLTAGE_SWING_1200:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_3:
 			demph_reg_value = 0x2B405555;
 			uniqtranscale_reg_value = 0x5598DA3A;
 			break;
@@ -2496,18 +2938,18 @@ static uint32_t intel_vlv_signal_levels(struct intel_dp *intel_dp)
 			return 0;
 		}
 		break;
-	case DP_TRAIN_PRE_EMPHASIS_3_5:
+	case DP_TRAIN_PRE_EMPH_LEVEL_1:
 		preemph_reg_value = 0x0002000;
 		switch (train_set & DP_TRAIN_VOLTAGE_SWING_MASK) {
-		case DP_TRAIN_VOLTAGE_SWING_400:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_0:
 			demph_reg_value = 0x2B404040;
 			uniqtranscale_reg_value = 0x5552B83A;
 			break;
-		case DP_TRAIN_VOLTAGE_SWING_600:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_1:
 			demph_reg_value = 0x2B404848;
 			uniqtranscale_reg_value = 0x5580B83A;
 			break;
-		case DP_TRAIN_VOLTAGE_SWING_800:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_2:
 			demph_reg_value = 0x2B404040;
 			uniqtranscale_reg_value = 0x55ADDA3A;
 			break;
@@ -2515,14 +2957,14 @@ static uint32_t intel_vlv_signal_levels(struct intel_dp *intel_dp)
 			return 0;
 		}
 		break;
-	case DP_TRAIN_PRE_EMPHASIS_6:
+	case DP_TRAIN_PRE_EMPH_LEVEL_2:
 		preemph_reg_value = 0x0000000;
 		switch (train_set & DP_TRAIN_VOLTAGE_SWING_MASK) {
-		case DP_TRAIN_VOLTAGE_SWING_400:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_0:
 			demph_reg_value = 0x2B305555;
 			uniqtranscale_reg_value = 0x5570B83A;
 			break;
-		case DP_TRAIN_VOLTAGE_SWING_600:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_1:
 			demph_reg_value = 0x2B2B4040;
 			uniqtranscale_reg_value = 0x55ADDA3A;
 			break;
@@ -2530,10 +2972,10 @@ static uint32_t intel_vlv_signal_levels(struct intel_dp *intel_dp)
 			return 0;
 		}
 		break;
-	case DP_TRAIN_PRE_EMPHASIS_9_5:
+	case DP_TRAIN_PRE_EMPH_LEVEL_3:
 		preemph_reg_value = 0x0006000;
 		switch (train_set & DP_TRAIN_VOLTAGE_SWING_MASK) {
-		case DP_TRAIN_VOLTAGE_SWING_400:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_0:
 			demph_reg_value = 0x1B405555;
 			uniqtranscale_reg_value = 0x55ADDA3A;
 			break;
@@ -2572,21 +3014,21 @@ static uint32_t intel_chv_signal_levels(struct intel_dp *intel_dp)
 	int i;
 
 	switch (train_set & DP_TRAIN_PRE_EMPHASIS_MASK) {
-	case DP_TRAIN_PRE_EMPHASIS_0:
+	case DP_TRAIN_PRE_EMPH_LEVEL_0:
 		switch (train_set & DP_TRAIN_VOLTAGE_SWING_MASK) {
-		case DP_TRAIN_VOLTAGE_SWING_400:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_0:
 			deemph_reg_value = 128;
 			margin_reg_value = 52;
 			break;
-		case DP_TRAIN_VOLTAGE_SWING_600:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_1:
 			deemph_reg_value = 128;
 			margin_reg_value = 77;
 			break;
-		case DP_TRAIN_VOLTAGE_SWING_800:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_2:
 			deemph_reg_value = 128;
 			margin_reg_value = 102;
 			break;
-		case DP_TRAIN_VOLTAGE_SWING_1200:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_3:
 			deemph_reg_value = 128;
 			margin_reg_value = 154;
 			/* FIXME extra to set for 1200 */
@@ -2595,17 +3037,17 @@ static uint32_t intel_chv_signal_levels(struct intel_dp *intel_dp)
 			return 0;
 		}
 		break;
-	case DP_TRAIN_PRE_EMPHASIS_3_5:
+	case DP_TRAIN_PRE_EMPH_LEVEL_1:
 		switch (train_set & DP_TRAIN_VOLTAGE_SWING_MASK) {
-		case DP_TRAIN_VOLTAGE_SWING_400:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_0:
 			deemph_reg_value = 85;
 			margin_reg_value = 78;
 			break;
-		case DP_TRAIN_VOLTAGE_SWING_600:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_1:
 			deemph_reg_value = 85;
 			margin_reg_value = 116;
 			break;
-		case DP_TRAIN_VOLTAGE_SWING_800:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_2:
 			deemph_reg_value = 85;
 			margin_reg_value = 154;
 			break;
@@ -2613,13 +3055,13 @@ static uint32_t intel_chv_signal_levels(struct intel_dp *intel_dp)
 			return 0;
 		}
 		break;
-	case DP_TRAIN_PRE_EMPHASIS_6:
+	case DP_TRAIN_PRE_EMPH_LEVEL_2:
 		switch (train_set & DP_TRAIN_VOLTAGE_SWING_MASK) {
-		case DP_TRAIN_VOLTAGE_SWING_400:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_0:
 			deemph_reg_value = 64;
 			margin_reg_value = 104;
 			break;
-		case DP_TRAIN_VOLTAGE_SWING_600:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_1:
 			deemph_reg_value = 64;
 			margin_reg_value = 154;
 			break;
@@ -2627,9 +3069,9 @@ static uint32_t intel_chv_signal_levels(struct intel_dp *intel_dp)
 			return 0;
 		}
 		break;
-	case DP_TRAIN_PRE_EMPHASIS_9_5:
+	case DP_TRAIN_PRE_EMPH_LEVEL_3:
 		switch (train_set & DP_TRAIN_VOLTAGE_SWING_MASK) {
-		case DP_TRAIN_VOLTAGE_SWING_400:
+		case DP_TRAIN_VOLTAGE_SWING_LEVEL_0:
 			deemph_reg_value = 43;
 			margin_reg_value = 154;
 			break;
@@ -2663,8 +3105,8 @@ static uint32_t intel_chv_signal_levels(struct intel_dp *intel_dp)
 	/* Program swing margin */
 	for (i = 0; i < 4; i++) {
 		val = vlv_dpio_read(dev_priv, pipe, CHV_TX_DW2(ch, i));
-		val &= ~DPIO_SWING_MARGIN_MASK;
-		val |= margin_reg_value << DPIO_SWING_MARGIN_SHIFT;
+		val &= ~DPIO_SWING_MARGIN000_MASK;
+		val |= margin_reg_value << DPIO_SWING_MARGIN000_SHIFT;
 		vlv_dpio_write(dev_priv, pipe, CHV_TX_DW2(ch, i), val);
 	}
 
@@ -2676,9 +3118,9 @@ static uint32_t intel_chv_signal_levels(struct intel_dp *intel_dp)
 	}
 
 	if (((train_set & DP_TRAIN_PRE_EMPHASIS_MASK)
-			== DP_TRAIN_PRE_EMPHASIS_0) &&
+			== DP_TRAIN_PRE_EMPH_LEVEL_0) &&
 		((train_set & DP_TRAIN_VOLTAGE_SWING_MASK)
-			== DP_TRAIN_VOLTAGE_SWING_1200)) {
+			== DP_TRAIN_VOLTAGE_SWING_LEVEL_3)) {
 
 		/*
 		 * The document said it needs to set bit 27 for ch0 and bit 26
@@ -2757,32 +3199,32 @@ intel_gen4_signal_levels(uint8_t train_set)
 	uint32_t	signal_levels = 0;
 
 	switch (train_set & DP_TRAIN_VOLTAGE_SWING_MASK) {
-	case DP_TRAIN_VOLTAGE_SWING_400:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_0:
 	default:
 		signal_levels |= DP_VOLTAGE_0_4;
 		break;
-	case DP_TRAIN_VOLTAGE_SWING_600:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_1:
 		signal_levels |= DP_VOLTAGE_0_6;
 		break;
-	case DP_TRAIN_VOLTAGE_SWING_800:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_2:
 		signal_levels |= DP_VOLTAGE_0_8;
 		break;
-	case DP_TRAIN_VOLTAGE_SWING_1200:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_3:
 		signal_levels |= DP_VOLTAGE_1_2;
 		break;
 	}
 	switch (train_set & DP_TRAIN_PRE_EMPHASIS_MASK) {
-	case DP_TRAIN_PRE_EMPHASIS_0:
+	case DP_TRAIN_PRE_EMPH_LEVEL_0:
 	default:
 		signal_levels |= DP_PRE_EMPHASIS_0;
 		break;
-	case DP_TRAIN_PRE_EMPHASIS_3_5:
+	case DP_TRAIN_PRE_EMPH_LEVEL_1:
 		signal_levels |= DP_PRE_EMPHASIS_3_5;
 		break;
-	case DP_TRAIN_PRE_EMPHASIS_6:
+	case DP_TRAIN_PRE_EMPH_LEVEL_2:
 		signal_levels |= DP_PRE_EMPHASIS_6;
 		break;
-	case DP_TRAIN_PRE_EMPHASIS_9_5:
+	case DP_TRAIN_PRE_EMPH_LEVEL_3:
 		signal_levels |= DP_PRE_EMPHASIS_9_5;
 		break;
 	}
@@ -2796,19 +3238,19 @@ intel_gen6_edp_signal_levels(uint8_t train_set)
 	int signal_levels = train_set & (DP_TRAIN_VOLTAGE_SWING_MASK |
 					 DP_TRAIN_PRE_EMPHASIS_MASK);
 	switch (signal_levels) {
-	case DP_TRAIN_VOLTAGE_SWING_400 | DP_TRAIN_PRE_EMPHASIS_0:
-	case DP_TRAIN_VOLTAGE_SWING_600 | DP_TRAIN_PRE_EMPHASIS_0:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_0 | DP_TRAIN_PRE_EMPH_LEVEL_0:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_1 | DP_TRAIN_PRE_EMPH_LEVEL_0:
 		return EDP_LINK_TRAIN_400_600MV_0DB_SNB_B;
-	case DP_TRAIN_VOLTAGE_SWING_400 | DP_TRAIN_PRE_EMPHASIS_3_5:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_0 | DP_TRAIN_PRE_EMPH_LEVEL_1:
 		return EDP_LINK_TRAIN_400MV_3_5DB_SNB_B;
-	case DP_TRAIN_VOLTAGE_SWING_400 | DP_TRAIN_PRE_EMPHASIS_6:
-	case DP_TRAIN_VOLTAGE_SWING_600 | DP_TRAIN_PRE_EMPHASIS_6:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_0 | DP_TRAIN_PRE_EMPH_LEVEL_2:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_1 | DP_TRAIN_PRE_EMPH_LEVEL_2:
 		return EDP_LINK_TRAIN_400_600MV_6DB_SNB_B;
-	case DP_TRAIN_VOLTAGE_SWING_600 | DP_TRAIN_PRE_EMPHASIS_3_5:
-	case DP_TRAIN_VOLTAGE_SWING_800 | DP_TRAIN_PRE_EMPHASIS_3_5:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_1 | DP_TRAIN_PRE_EMPH_LEVEL_1:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_2 | DP_TRAIN_PRE_EMPH_LEVEL_1:
 		return EDP_LINK_TRAIN_600_800MV_3_5DB_SNB_B;
-	case DP_TRAIN_VOLTAGE_SWING_800 | DP_TRAIN_PRE_EMPHASIS_0:
-	case DP_TRAIN_VOLTAGE_SWING_1200 | DP_TRAIN_PRE_EMPHASIS_0:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_2 | DP_TRAIN_PRE_EMPH_LEVEL_0:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_3 | DP_TRAIN_PRE_EMPH_LEVEL_0:
 		return EDP_LINK_TRAIN_800_1200MV_0DB_SNB_B;
 	default:
 		DRM_DEBUG_KMS("Unsupported voltage swing/pre-emphasis level:"
@@ -2824,21 +3266,21 @@ intel_gen7_edp_signal_levels(uint8_t train_set)
 	int signal_levels = train_set & (DP_TRAIN_VOLTAGE_SWING_MASK |
 					 DP_TRAIN_PRE_EMPHASIS_MASK);
 	switch (signal_levels) {
-	case DP_TRAIN_VOLTAGE_SWING_400 | DP_TRAIN_PRE_EMPHASIS_0:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_0 | DP_TRAIN_PRE_EMPH_LEVEL_0:
 		return EDP_LINK_TRAIN_400MV_0DB_IVB;
-	case DP_TRAIN_VOLTAGE_SWING_400 | DP_TRAIN_PRE_EMPHASIS_3_5:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_0 | DP_TRAIN_PRE_EMPH_LEVEL_1:
 		return EDP_LINK_TRAIN_400MV_3_5DB_IVB;
-	case DP_TRAIN_VOLTAGE_SWING_400 | DP_TRAIN_PRE_EMPHASIS_6:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_0 | DP_TRAIN_PRE_EMPH_LEVEL_2:
 		return EDP_LINK_TRAIN_400MV_6DB_IVB;
 
-	case DP_TRAIN_VOLTAGE_SWING_600 | DP_TRAIN_PRE_EMPHASIS_0:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_1 | DP_TRAIN_PRE_EMPH_LEVEL_0:
 		return EDP_LINK_TRAIN_600MV_0DB_IVB;
-	case DP_TRAIN_VOLTAGE_SWING_600 | DP_TRAIN_PRE_EMPHASIS_3_5:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_1 | DP_TRAIN_PRE_EMPH_LEVEL_1:
 		return EDP_LINK_TRAIN_600MV_3_5DB_IVB;
 
-	case DP_TRAIN_VOLTAGE_SWING_800 | DP_TRAIN_PRE_EMPHASIS_0:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_2 | DP_TRAIN_PRE_EMPH_LEVEL_0:
 		return EDP_LINK_TRAIN_800MV_0DB_IVB;
-	case DP_TRAIN_VOLTAGE_SWING_800 | DP_TRAIN_PRE_EMPHASIS_3_5:
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_2 | DP_TRAIN_PRE_EMPH_LEVEL_1:
 		return EDP_LINK_TRAIN_800MV_3_5DB_IVB;
 
 	default:
@@ -2855,30 +3297,30 @@ intel_hsw_signal_levels(uint8_t train_set)
 	int signal_levels = train_set & (DP_TRAIN_VOLTAGE_SWING_MASK |
 					 DP_TRAIN_PRE_EMPHASIS_MASK);
 	switch (signal_levels) {
-	case DP_TRAIN_VOLTAGE_SWING_400 | DP_TRAIN_PRE_EMPHASIS_0:
-		return DDI_BUF_EMP_400MV_0DB_HSW;
-	case DP_TRAIN_VOLTAGE_SWING_400 | DP_TRAIN_PRE_EMPHASIS_3_5:
-		return DDI_BUF_EMP_400MV_3_5DB_HSW;
-	case DP_TRAIN_VOLTAGE_SWING_400 | DP_TRAIN_PRE_EMPHASIS_6:
-		return DDI_BUF_EMP_400MV_6DB_HSW;
-	case DP_TRAIN_VOLTAGE_SWING_400 | DP_TRAIN_PRE_EMPHASIS_9_5:
-		return DDI_BUF_EMP_400MV_9_5DB_HSW;
-
-	case DP_TRAIN_VOLTAGE_SWING_600 | DP_TRAIN_PRE_EMPHASIS_0:
-		return DDI_BUF_EMP_600MV_0DB_HSW;
-	case DP_TRAIN_VOLTAGE_SWING_600 | DP_TRAIN_PRE_EMPHASIS_3_5:
-		return DDI_BUF_EMP_600MV_3_5DB_HSW;
-	case DP_TRAIN_VOLTAGE_SWING_600 | DP_TRAIN_PRE_EMPHASIS_6:
-		return DDI_BUF_EMP_600MV_6DB_HSW;
-
-	case DP_TRAIN_VOLTAGE_SWING_800 | DP_TRAIN_PRE_EMPHASIS_0:
-		return DDI_BUF_EMP_800MV_0DB_HSW;
-	case DP_TRAIN_VOLTAGE_SWING_800 | DP_TRAIN_PRE_EMPHASIS_3_5:
-		return DDI_BUF_EMP_800MV_3_5DB_HSW;
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_0 | DP_TRAIN_PRE_EMPH_LEVEL_0:
+		return DDI_BUF_TRANS_SELECT(0);
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_0 | DP_TRAIN_PRE_EMPH_LEVEL_1:
+		return DDI_BUF_TRANS_SELECT(1);
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_0 | DP_TRAIN_PRE_EMPH_LEVEL_2:
+		return DDI_BUF_TRANS_SELECT(2);
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_0 | DP_TRAIN_PRE_EMPH_LEVEL_3:
+		return DDI_BUF_TRANS_SELECT(3);
+
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_1 | DP_TRAIN_PRE_EMPH_LEVEL_0:
+		return DDI_BUF_TRANS_SELECT(4);
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_1 | DP_TRAIN_PRE_EMPH_LEVEL_1:
+		return DDI_BUF_TRANS_SELECT(5);
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_1 | DP_TRAIN_PRE_EMPH_LEVEL_2:
+		return DDI_BUF_TRANS_SELECT(6);
+
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_2 | DP_TRAIN_PRE_EMPH_LEVEL_0:
+		return DDI_BUF_TRANS_SELECT(7);
+	case DP_TRAIN_VOLTAGE_SWING_LEVEL_2 | DP_TRAIN_PRE_EMPH_LEVEL_1:
+		return DDI_BUF_TRANS_SELECT(8);
 	default:
 		DRM_DEBUG_KMS("Unsupported voltage swing/pre-emphasis level:"
 			      "0x%x\n", signal_levels);
-		return DDI_BUF_EMP_400MV_0DB_HSW;
+		return DDI_BUF_TRANS_SELECT(0);
 	}
 }
 
@@ -2925,74 +3367,10 @@ intel_dp_set_link_train(struct intel_dp *intel_dp,
 	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
 	struct drm_device *dev = intel_dig_port->base.base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	enum port port = intel_dig_port->port;
 	uint8_t buf[sizeof(intel_dp->train_set) + 1];
 	int ret, len;
 
-	if (HAS_DDI(dev)) {
-		uint32_t temp = I915_READ(DP_TP_CTL(port));
-
-		if (dp_train_pat & DP_LINK_SCRAMBLING_DISABLE)
-			temp |= DP_TP_CTL_SCRAMBLE_DISABLE;
-		else
-			temp &= ~DP_TP_CTL_SCRAMBLE_DISABLE;
-
-		temp &= ~DP_TP_CTL_LINK_TRAIN_MASK;
-		switch (dp_train_pat & DP_TRAINING_PATTERN_MASK) {
-		case DP_TRAINING_PATTERN_DISABLE:
-			temp |= DP_TP_CTL_LINK_TRAIN_NORMAL;
-
-			break;
-		case DP_TRAINING_PATTERN_1:
-			temp |= DP_TP_CTL_LINK_TRAIN_PAT1;
-			break;
-		case DP_TRAINING_PATTERN_2:
-			temp |= DP_TP_CTL_LINK_TRAIN_PAT2;
-			break;
-		case DP_TRAINING_PATTERN_3:
-			temp |= DP_TP_CTL_LINK_TRAIN_PAT3;
-			break;
-		}
-		I915_WRITE(DP_TP_CTL(port), temp);
-
-	} else if (HAS_PCH_CPT(dev) && (IS_GEN7(dev) || port != PORT_A)) {
-		*DP &= ~DP_LINK_TRAIN_MASK_CPT;
-
-		switch (dp_train_pat & DP_TRAINING_PATTERN_MASK) {
-		case DP_TRAINING_PATTERN_DISABLE:
-			*DP |= DP_LINK_TRAIN_OFF_CPT;
-			break;
-		case DP_TRAINING_PATTERN_1:
-			*DP |= DP_LINK_TRAIN_PAT_1_CPT;
-			break;
-		case DP_TRAINING_PATTERN_2:
-			*DP |= DP_LINK_TRAIN_PAT_2_CPT;
-			break;
-		case DP_TRAINING_PATTERN_3:
-			DRM_ERROR("DP training pattern 3 not supported\n");
-			*DP |= DP_LINK_TRAIN_PAT_2_CPT;
-			break;
-		}
-
-	} else {
-		*DP &= ~DP_LINK_TRAIN_MASK;
-
-		switch (dp_train_pat & DP_TRAINING_PATTERN_MASK) {
-		case DP_TRAINING_PATTERN_DISABLE:
-			*DP |= DP_LINK_TRAIN_OFF;
-			break;
-		case DP_TRAINING_PATTERN_1:
-			*DP |= DP_LINK_TRAIN_PAT_1;
-			break;
-		case DP_TRAINING_PATTERN_2:
-			*DP |= DP_LINK_TRAIN_PAT_2;
-			break;
-		case DP_TRAINING_PATTERN_3:
-			DRM_ERROR("DP training pattern 3 not supported\n");
-			*DP |= DP_LINK_TRAIN_PAT_2;
-			break;
-		}
-	}
+	_intel_dp_set_link_train(intel_dp, DP, dp_train_pat);
 
 	I915_WRITE(intel_dp->output_reg, *DP);
 	POSTING_READ(intel_dp->output_reg);
@@ -3276,7 +3654,10 @@ intel_dp_link_down(struct intel_dp *intel_dp)
 		DP &= ~DP_LINK_TRAIN_MASK_CPT;
 		I915_WRITE(intel_dp->output_reg, DP | DP_LINK_TRAIN_PAT_IDLE_CPT);
 	} else {
-		DP &= ~DP_LINK_TRAIN_MASK;
+		if (IS_CHERRYVIEW(dev))
+			DP &= ~DP_LINK_TRAIN_MASK_CHV;
+		else
+			DP &= ~DP_LINK_TRAIN_MASK;
 		I915_WRITE(intel_dp->output_reg, DP | DP_LINK_TRAIN_PAT_IDLE);
 	}
 	POSTING_READ(intel_dp->output_reg);
@@ -3322,15 +3703,11 @@ intel_dp_get_dpcd(struct intel_dp *intel_dp)
 	struct drm_device *dev = dig_port->base.base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	char dpcd_hex_dump[sizeof(intel_dp->dpcd) * 3];
-
 	if (intel_dp_dpcd_read_wake(&intel_dp->aux, 0x000, intel_dp->dpcd,
 				    sizeof(intel_dp->dpcd)) < 0)
 		return false; /* aux transfer failed */
 
-	hex_dump_to_buffer(intel_dp->dpcd, sizeof(intel_dp->dpcd),
-			   32, 1, dpcd_hex_dump, sizeof(dpcd_hex_dump), false);
-	DRM_DEBUG_KMS("DPCD: %s\n", dpcd_hex_dump);
+	DRM_DEBUG_KMS("DPCD: %*ph\n", (int) sizeof(intel_dp->dpcd), intel_dp->dpcd);
 
 	if (intel_dp->dpcd[DP_DPCD_REV] == 0)
 		return false; /* DPCD not present */
@@ -3351,7 +3728,7 @@ intel_dp_get_dpcd(struct intel_dp *intel_dp)
 	if (intel_dp->dpcd[DP_DPCD_REV] >= 0x12 &&
 	    intel_dp->dpcd[DP_MAX_LANE_COUNT] & DP_TPS3_SUPPORTED) {
 		intel_dp->use_tps3 = true;
-		DRM_DEBUG_KMS("Displayport TPS3 supported");
+		DRM_DEBUG_KMS("Displayport TPS3 supported\n");
 	} else
 		intel_dp->use_tps3 = false;
 
@@ -3388,7 +3765,7 @@ intel_dp_probe_oui(struct intel_dp *intel_dp)
 		DRM_DEBUG_KMS("Branch OUI: %02hx%02hx%02hx\n",
 			      buf[0], buf[1], buf[2]);
 
-	edp_panel_vdd_off(intel_dp, false);
+	intel_edp_panel_vdd_off(intel_dp, false);
 }
 
 static bool
@@ -3402,7 +3779,7 @@ intel_dp_probe_mst(struct intel_dp *intel_dp)
 	if (intel_dp->dpcd[DP_DPCD_REV] < 0x12)
 		return false;
 
-	_edp_panel_vdd_on(intel_dp);
+	intel_edp_panel_vdd_on(intel_dp);
 	if (intel_dp_dpcd_read_wake(&intel_dp->aux, DP_MSTM_CAP, buf, 1)) {
 		if (buf[0] & DP_MST_CAP) {
 			DRM_DEBUG_KMS("Sink is MST capable\n");
@@ -3412,7 +3789,7 @@ intel_dp_probe_mst(struct intel_dp *intel_dp)
 			intel_dp->is_mst = false;
 		}
 	}
-	edp_panel_vdd_off(intel_dp, false);
+	intel_edp_panel_vdd_off(intel_dp, false);
 
 	drm_dp_mst_topology_mgr_set_mst(&intel_dp->mst_mgr, intel_dp->is_mst);
 	return intel_dp->is_mst;
@@ -3427,21 +3804,21 @@ int intel_dp_sink_crc(struct intel_dp *intel_dp, u8 *crc)
 	u8 buf[1];
 
 	if (drm_dp_dpcd_readb(&intel_dp->aux, DP_TEST_SINK_MISC, buf) < 0)
-		return -EAGAIN;
+		return -EIO;
 
 	if (!(buf[0] & DP_TEST_CRC_SUPPORTED))
 		return -ENOTTY;
 
 	if (drm_dp_dpcd_writeb(&intel_dp->aux, DP_TEST_SINK,
 			       DP_TEST_SINK_START) < 0)
-		return -EAGAIN;
+		return -EIO;
 
 	/* Wait 2 vblanks to be sure we will have the correct CRC value */
 	intel_wait_for_vblank(dev, intel_crtc->pipe);
 	intel_wait_for_vblank(dev, intel_crtc->pipe);
 
 	if (drm_dp_dpcd_read(&intel_dp->aux, DP_TEST_CRC_R_CR, crc, 6) < 0)
-		return -EAGAIN;
+		return -EIO;
 
 	drm_dp_dpcd_writeb(&intel_dp->aux, DP_TEST_SINK, 0);
 	return 0;
@@ -3644,20 +4021,24 @@ intel_dp_detect_dpcd(struct intel_dp *intel_dp)
 }
 
 static enum drm_connector_status
+edp_detect(struct intel_dp *intel_dp)
+{
+	struct drm_device *dev = intel_dp_to_dev(intel_dp);
+	enum drm_connector_status status;
+
+	status = intel_panel_detect(dev);
+	if (status == connector_status_unknown)
+		status = connector_status_connected;
+
+	return status;
+}
+
+static enum drm_connector_status
 ironlake_dp_detect(struct intel_dp *intel_dp)
 {
 	struct drm_device *dev = intel_dp_to_dev(intel_dp);
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
-	enum drm_connector_status status;
-
-	/* Can't disconnect eDP, but you can close the lid... */
-	if (is_edp(intel_dp)) {
-		status = intel_panel_detect(dev);
-		if (status == connector_status_unknown)
-			status = connector_status_connected;
-		return status;
-	}
 
 	if (!ibx_digital_port_connected(dev_priv, intel_dig_port))
 		return connector_status_disconnected;
@@ -3733,9 +4114,9 @@ g4x_dp_detect(struct intel_dp *intel_dp)
 }
 
 static struct edid *
-intel_dp_get_edid(struct drm_connector *connector, struct i2c_adapter *adapter)
+intel_dp_get_edid(struct intel_dp *intel_dp)
 {
-	struct intel_connector *intel_connector = to_intel_connector(connector);
+	struct intel_connector *intel_connector = intel_dp->attached_connector;
 
 	/* use cached edid if we have one */
 	if (intel_connector->edid) {
@@ -3744,27 +4125,55 @@ intel_dp_get_edid(struct drm_connector *connector, struct i2c_adapter *adapter)
 			return NULL;
 
 		return drm_edid_duplicate(intel_connector->edid);
-	}
+	} else
+		return drm_get_edid(&intel_connector->base,
+				    &intel_dp->aux.ddc);
+}
+
+static void
+intel_dp_set_edid(struct intel_dp *intel_dp)
+{
+	struct intel_connector *intel_connector = intel_dp->attached_connector;
+	struct edid *edid;
+
+	edid = intel_dp_get_edid(intel_dp);
+	intel_connector->detect_edid = edid;
 
-	return drm_get_edid(connector, adapter);
+	if (intel_dp->force_audio != HDMI_AUDIO_AUTO)
+		intel_dp->has_audio = intel_dp->force_audio == HDMI_AUDIO_ON;
+	else
+		intel_dp->has_audio = drm_detect_monitor_audio(edid);
 }
 
-static int
-intel_dp_get_edid_modes(struct drm_connector *connector, struct i2c_adapter *adapter)
+static void
+intel_dp_unset_edid(struct intel_dp *intel_dp)
 {
-	struct intel_connector *intel_connector = to_intel_connector(connector);
+	struct intel_connector *intel_connector = intel_dp->attached_connector;
 
-	/* use cached edid if we have one */
-	if (intel_connector->edid) {
-		/* invalid edid */
-		if (IS_ERR(intel_connector->edid))
-			return 0;
+	kfree(intel_connector->detect_edid);
+	intel_connector->detect_edid = NULL;
 
-		return intel_connector_update_modes(connector,
-						    intel_connector->edid);
-	}
+	intel_dp->has_audio = false;
+}
+
+static enum intel_display_power_domain
+intel_dp_power_get(struct intel_dp *dp)
+{
+	struct intel_encoder *encoder = &dp_to_dig_port(dp)->base;
+	enum intel_display_power_domain power_domain;
+
+	power_domain = intel_display_port_power_domain(encoder);
+	intel_display_power_get(to_i915(encoder->base.dev), power_domain);
+
+	return power_domain;
+}
 
-	return intel_ddc_get_modes(connector, adapter);
+static void
+intel_dp_power_put(struct intel_dp *dp,
+		   enum intel_display_power_domain power_domain)
+{
+	struct intel_encoder *encoder = &dp_to_dig_port(dp)->base;
+	intel_display_power_put(to_i915(encoder->base.dev), power_domain);
 }
 
 static enum drm_connector_status
@@ -3774,33 +4183,30 @@ intel_dp_detect(struct drm_connector *connector, bool force)
 	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
 	struct intel_encoder *intel_encoder = &intel_dig_port->base;
 	struct drm_device *dev = connector->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	enum drm_connector_status status;
 	enum intel_display_power_domain power_domain;
-	struct edid *edid = NULL;
 	bool ret;
 
-	power_domain = intel_display_port_power_domain(intel_encoder);
-	intel_display_power_get(dev_priv, power_domain);
-
 	DRM_DEBUG_KMS("[CONNECTOR:%d:%s]\n",
 		      connector->base.id, connector->name);
+	intel_dp_unset_edid(intel_dp);
 
 	if (intel_dp->is_mst) {
 		/* MST devices are disconnected from a monitor POV */
 		if (intel_encoder->type != INTEL_OUTPUT_EDP)
 			intel_encoder->type = INTEL_OUTPUT_DISPLAYPORT;
-		status = connector_status_disconnected;
-		goto out;
+		return connector_status_disconnected;
 	}
 
-	intel_dp->has_audio = false;
+	power_domain = intel_dp_power_get(intel_dp);
 
-	if (HAS_PCH_SPLIT(dev))
+	/* Can't disconnect eDP, but you can close the lid... */
+	if (is_edp(intel_dp))
+		status = edp_detect(intel_dp);
+	else if (HAS_PCH_SPLIT(dev))
 		status = ironlake_dp_detect(intel_dp);
 	else
 		status = g4x_dp_detect(intel_dp);
-
 	if (status != connector_status_connected)
 		goto out;
 
@@ -3816,82 +4222,78 @@ intel_dp_detect(struct drm_connector *connector, bool force)
 		goto out;
 	}
 
-	if (intel_dp->force_audio != HDMI_AUDIO_AUTO) {
-		intel_dp->has_audio = (intel_dp->force_audio == HDMI_AUDIO_ON);
-	} else {
-		edid = intel_dp_get_edid(connector, &intel_dp->aux.ddc);
-		if (edid) {
-			intel_dp->has_audio = drm_detect_monitor_audio(edid);
-			kfree(edid);
-		}
-	}
+	intel_dp_set_edid(intel_dp);
 
 	if (intel_encoder->type != INTEL_OUTPUT_EDP)
 		intel_encoder->type = INTEL_OUTPUT_DISPLAYPORT;
 	status = connector_status_connected;
 
 out:
-	intel_display_power_put(dev_priv, power_domain);
+	intel_dp_power_put(intel_dp, power_domain);
 	return status;
 }
 
-static int intel_dp_get_modes(struct drm_connector *connector)
+static void
+intel_dp_force(struct drm_connector *connector)
 {
 	struct intel_dp *intel_dp = intel_attached_dp(connector);
-	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
-	struct intel_encoder *intel_encoder = &intel_dig_port->base;
-	struct intel_connector *intel_connector = to_intel_connector(connector);
-	struct drm_device *dev = connector->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_encoder *intel_encoder = &dp_to_dig_port(intel_dp)->base;
 	enum intel_display_power_domain power_domain;
-	int ret;
 
-	/* We should parse the EDID data and find out if it has an audio sink
-	 */
+	DRM_DEBUG_KMS("[CONNECTOR:%d:%s]\n",
+		      connector->base.id, connector->name);
+	intel_dp_unset_edid(intel_dp);
 
-	power_domain = intel_display_port_power_domain(intel_encoder);
-	intel_display_power_get(dev_priv, power_domain);
+	if (connector->status != connector_status_connected)
+		return;
 
-	ret = intel_dp_get_edid_modes(connector, &intel_dp->aux.ddc);
-	intel_display_power_put(dev_priv, power_domain);
-	if (ret)
-		return ret;
+	power_domain = intel_dp_power_get(intel_dp);
+
+	intel_dp_set_edid(intel_dp);
+
+	intel_dp_power_put(intel_dp, power_domain);
+
+	if (intel_encoder->type != INTEL_OUTPUT_EDP)
+		intel_encoder->type = INTEL_OUTPUT_DISPLAYPORT;
+}
+
+static int intel_dp_get_modes(struct drm_connector *connector)
+{
+	struct intel_connector *intel_connector = to_intel_connector(connector);
+	struct edid *edid;
+
+	edid = intel_connector->detect_edid;
+	if (edid) {
+		int ret = intel_connector_update_modes(connector, edid);
+		if (ret)
+			return ret;
+	}
 
 	/* if eDP has no EDID, fall back to fixed mode */
-	if (is_edp(intel_dp) && intel_connector->panel.fixed_mode) {
+	if (is_edp(intel_attached_dp(connector)) &&
+	    intel_connector->panel.fixed_mode) {
 		struct drm_display_mode *mode;
-		mode = drm_mode_duplicate(dev,
+
+		mode = drm_mode_duplicate(connector->dev,
 					  intel_connector->panel.fixed_mode);
 		if (mode) {
 			drm_mode_probed_add(connector, mode);
 			return 1;
 		}
 	}
+
 	return 0;
 }
 
 static bool
 intel_dp_detect_audio(struct drm_connector *connector)
 {
-	struct intel_dp *intel_dp = intel_attached_dp(connector);
-	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
-	struct intel_encoder *intel_encoder = &intel_dig_port->base;
-	struct drm_device *dev = connector->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	enum intel_display_power_domain power_domain;
-	struct edid *edid;
 	bool has_audio = false;
+	struct edid *edid;
 
-	power_domain = intel_display_port_power_domain(intel_encoder);
-	intel_display_power_get(dev_priv, power_domain);
-
-	edid = intel_dp_get_edid(connector, &intel_dp->aux.ddc);
-	if (edid) {
+	edid = to_intel_connector(connector)->detect_edid;
+	if (edid)
 		has_audio = drm_detect_monitor_audio(edid);
-		kfree(edid);
-	}
-
-	intel_display_power_put(dev_priv, power_domain);
 
 	return has_audio;
 }
@@ -3989,6 +4391,8 @@ intel_dp_connector_destroy(struct drm_connector *connector)
 {
 	struct intel_connector *intel_connector = to_intel_connector(connector);
 
+	kfree(intel_connector->detect_edid);
+
 	if (!IS_ERR_OR_NULL(intel_connector->edid))
 		kfree(intel_connector->edid);
 
@@ -4005,16 +4409,20 @@ void intel_dp_encoder_destroy(struct drm_encoder *encoder)
 {
 	struct intel_digital_port *intel_dig_port = enc_to_dig_port(encoder);
 	struct intel_dp *intel_dp = &intel_dig_port->dp;
-	struct drm_device *dev = intel_dp_to_dev(intel_dp);
 
 	drm_dp_aux_unregister(&intel_dp->aux);
 	intel_dp_mst_encoder_cleanup(intel_dig_port);
 	drm_encoder_cleanup(encoder);
 	if (is_edp(intel_dp)) {
 		cancel_delayed_work_sync(&intel_dp->panel_vdd_work);
-		drm_modeset_lock(&dev->mode_config.connection_mutex, NULL);
+		/*
+		 * vdd might still be enabled do to the delayed vdd off.
+		 * Make sure vdd is actually turned off here.
+		 */
+		pps_lock(intel_dp);
 		edp_panel_vdd_off_sync(intel_dp);
-		drm_modeset_unlock(&dev->mode_config.connection_mutex);
+		pps_unlock(intel_dp);
+
 		if (intel_dp->edp_notifier.notifier_call) {
 			unregister_reboot_notifier(&intel_dp->edp_notifier);
 			intel_dp->edp_notifier.notifier_call = NULL;
@@ -4030,7 +4438,13 @@ static void intel_dp_encoder_suspend(struct intel_encoder *intel_encoder)
 	if (!is_edp(intel_dp))
 		return;
 
+	/*
+	 * vdd might still be enabled do to the delayed vdd off.
+	 * Make sure vdd is actually turned off here.
+	 */
+	pps_lock(intel_dp);
 	edp_panel_vdd_off_sync(intel_dp);
+	pps_unlock(intel_dp);
 }
 
 static void intel_dp_encoder_reset(struct drm_encoder *encoder)
@@ -4041,6 +4455,7 @@ static void intel_dp_encoder_reset(struct drm_encoder *encoder)
 static const struct drm_connector_funcs intel_dp_connector_funcs = {
 	.dpms = intel_connector_dpms,
 	.detect = intel_dp_detect,
+	.force = intel_dp_force,
 	.fill_modes = drm_helper_probe_single_connector_modes,
 	.set_property = intel_dp_set_property,
 	.destroy = intel_dp_connector_destroy,
@@ -4076,7 +4491,8 @@ intel_dp_hpd_pulse(struct intel_digital_port *intel_dig_port, bool long_hpd)
 	if (intel_dig_port->base.type != INTEL_OUTPUT_EDP)
 		intel_dig_port->base.type = INTEL_OUTPUT_DISPLAYPORT;
 
-	DRM_DEBUG_KMS("got hpd irq on port %d - %s\n", intel_dig_port->port,
+	DRM_DEBUG_KMS("got hpd irq on port %c - %s\n",
+		      port_name(intel_dig_port->port),
 		      long_hpd ? "long" : "short");
 
 	power_domain = intel_display_port_power_domain(intel_encoder);
@@ -4216,6 +4632,8 @@ intel_dp_init_panel_power_sequencer(struct drm_device *dev,
 	u32 pp_on, pp_off, pp_div, pp;
 	int pp_ctrl_reg, pp_on_reg, pp_off_reg, pp_div_reg;
 
+	lockdep_assert_held(&dev_priv->pps_mutex);
+
 	if (HAS_PCH_SPLIT(dev)) {
 		pp_ctrl_reg = PCH_PP_CONTROL;
 		pp_on_reg = PCH_PP_ON_DELAYS;
@@ -4315,6 +4733,9 @@ intel_dp_init_panel_power_sequencer_registers(struct drm_device *dev,
 	u32 pp_on, pp_off, pp_div, port_sel = 0;
 	int div = HAS_PCH_SPLIT(dev) ? intel_pch_rawclk(dev) : intel_hrawclk(dev);
 	int pp_on_reg, pp_off_reg, pp_div_reg;
+	enum port port = dp_to_dig_port(intel_dp)->port;
+
+	lockdep_assert_held(&dev_priv->pps_mutex);
 
 	if (HAS_PCH_SPLIT(dev)) {
 		pp_on_reg = PCH_PP_ON_DELAYS;
@@ -4349,12 +4770,9 @@ intel_dp_init_panel_power_sequencer_registers(struct drm_device *dev,
 	/* Haswell doesn't have any port selection bits for the panel
 	 * power sequencer any more. */
 	if (IS_VALLEYVIEW(dev)) {
-		if (dp_to_dig_port(intel_dp)->port == PORT_B)
-			port_sel = PANEL_PORT_SELECT_DPB_VLV;
-		else
-			port_sel = PANEL_PORT_SELECT_DPC_VLV;
+		port_sel = PANEL_PORT_SELECT_VLV(port);
 	} else if (HAS_PCH_IBX(dev) || HAS_PCH_CPT(dev)) {
-		if (dp_to_dig_port(intel_dp)->port == PORT_A)
+		if (port == PORT_A)
 			port_sel = PANEL_PORT_SELECT_DPA;
 		else
 			port_sel = PANEL_PORT_SELECT_DPD;
@@ -4438,7 +4856,7 @@ void intel_dp_set_drrs_state(struct drm_device *dev, int refresh_rate)
 		val = I915_READ(reg);
 		if (index > DRRS_HIGH_RR) {
 			val |= PIPECONF_EDP_RR_MODE_SWITCH;
-			intel_dp_set_m2_n2(intel_crtc, &config->dp_m2_n2);
+			intel_dp_set_m_n(intel_crtc);
 		} else {
 			val &= ~PIPECONF_EDP_RR_MODE_SWITCH;
 		}
@@ -4478,7 +4896,7 @@ intel_dp_drrs_init(struct intel_digital_port *intel_dig_port,
 	}
 
 	if (dev_priv->vbt.drrs_type != SEAMLESS_DRRS_SUPPORT) {
-		DRM_INFO("VBT doesn't support DRRS\n");
+		DRM_DEBUG_KMS("VBT doesn't support DRRS\n");
 		return NULL;
 	}
 
@@ -4486,7 +4904,7 @@ intel_dp_drrs_init(struct intel_digital_port *intel_dig_port,
 					(dev, fixed_mode, connector);
 
 	if (!downclock_mode) {
-		DRM_INFO("DRRS not supported\n");
+		DRM_DEBUG_KMS("DRRS not supported\n");
 		return NULL;
 	}
 
@@ -4497,7 +4915,7 @@ intel_dp_drrs_init(struct intel_digital_port *intel_dig_port,
 	intel_dp->drrs_state.type = dev_priv->vbt.drrs_type;
 
 	intel_dp->drrs_state.refresh_rate_type = DRRS_HIGH_RR;
-	DRM_INFO("seamless DRRS supported for eDP panel.\n");
+	DRM_DEBUG_KMS("seamless DRRS supported for eDP panel.\n");
 	return downclock_mode;
 }
 
@@ -4512,8 +4930,11 @@ void intel_edp_panel_vdd_sanitize(struct intel_encoder *intel_encoder)
 		return;
 
 	intel_dp = enc_to_intel_dp(&intel_encoder->base);
+
+	pps_lock(intel_dp);
+
 	if (!edp_have_panel_vdd(intel_dp))
-		return;
+		goto out;
 	/*
 	 * The VDD bit needs a power domain reference, so if the bit is
 	 * already enabled when we boot or resume, grab this reference and
@@ -4525,6 +4946,8 @@ void intel_edp_panel_vdd_sanitize(struct intel_encoder *intel_encoder)
 	intel_display_power_get(dev_priv, power_domain);
 
 	edp_panel_vdd_schedule_off(intel_dp);
+ out:
+	pps_unlock(intel_dp);
 }
 
 static bool intel_edp_init_connector(struct intel_dp *intel_dp,
@@ -4552,7 +4975,7 @@ static bool intel_edp_init_connector(struct intel_dp *intel_dp,
 	/* Cache DPCD and EDID for edp. */
 	intel_edp_panel_vdd_on(intel_dp);
 	has_dpcd = intel_dp_get_dpcd(intel_dp);
-	edp_panel_vdd_off(intel_dp, false);
+	intel_edp_panel_vdd_off(intel_dp, false);
 
 	if (has_dpcd) {
 		if (intel_dp->dpcd[DP_DPCD_REV] >= 0x11)
@@ -4566,7 +4989,9 @@ static bool intel_edp_init_connector(struct intel_dp *intel_dp,
 	}
 
 	/* We now know it's not a ghost, init power sequence regs. */
+	pps_lock(intel_dp);
 	intel_dp_init_panel_power_sequencer_registers(dev, intel_dp, power_seq);
+	pps_unlock(intel_dp);
 
 	mutex_lock(&dev->mode_config.mutex);
 	edid = drm_get_edid(connector, &intel_dp->aux.ddc);
@@ -4610,6 +5035,7 @@ static bool intel_edp_init_connector(struct intel_dp *intel_dp,
 	}
 
 	intel_panel_init(&intel_connector->panel, fixed_mode, downclock_mode);
+	intel_connector->panel.backlight_power = intel_edp_backlight_power;
 	intel_panel_setup_backlight(connector);
 
 	return true;
@@ -4628,6 +5054,8 @@ intel_dp_init_connector(struct intel_digital_port *intel_dig_port,
 	struct edp_power_seq power_seq = { 0 };
 	int type;
 
+	intel_dp->pps_pipe = INVALID_PIPE;
+
 	/* intel_dp vfuncs */
 	if (IS_VALLEYVIEW(dev))
 		intel_dp->get_aux_clock_divider = vlv_get_aux_clock_divider;
@@ -4698,8 +5126,15 @@ intel_dp_init_connector(struct intel_digital_port *intel_dig_port,
 	}
 
 	if (is_edp(intel_dp)) {
-		intel_dp_init_panel_power_timestamps(intel_dp);
-		intel_dp_init_panel_power_sequencer(dev, intel_dp, &power_seq);
+		pps_lock(intel_dp);
+		if (IS_VALLEYVIEW(dev)) {
+			vlv_initial_power_sequencer_setup(intel_dp);
+		} else {
+			intel_dp_init_panel_power_timestamps(intel_dp);
+			intel_dp_init_panel_power_sequencer(dev, intel_dp,
+							    &power_seq);
+		}
+		pps_unlock(intel_dp);
 	}
 
 	intel_dp_aux_init(intel_dp, intel_connector);
@@ -4707,7 +5142,8 @@ intel_dp_init_connector(struct intel_digital_port *intel_dig_port,
 	/* init MST on ports that can support it */
 	if (IS_HASWELL(dev) || IS_BROADWELL(dev)) {
 		if (port == PORT_B || port == PORT_C || port == PORT_D) {
-			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);
 		}
 	}
 
@@ -4715,9 +5151,13 @@ intel_dp_init_connector(struct intel_digital_port *intel_dig_port,
 		drm_dp_aux_unregister(&intel_dp->aux);
 		if (is_edp(intel_dp)) {
 			cancel_delayed_work_sync(&intel_dp->panel_vdd_work);
-			drm_modeset_lock(&dev->mode_config.connection_mutex, NULL);
+			/*
+			 * vdd might still be enabled do to the delayed vdd off.
+			 * Make sure vdd is actually turned off here.
+			 */
+			pps_lock(intel_dp);
 			edp_panel_vdd_off_sync(intel_dp);
-			drm_modeset_unlock(&dev->mode_config.connection_mutex);
+			pps_unlock(intel_dp);
 		}
 		drm_connector_unregister(connector);
 		drm_connector_cleanup(connector);
@@ -4781,7 +5221,8 @@ intel_dp_init(struct drm_device *dev, int output_reg, enum port port)
 	} else {
 		intel_encoder->pre_enable = g4x_pre_enable_dp;
 		intel_encoder->enable = g4x_enable_dp;
-		intel_encoder->post_disable = g4x_post_disable_dp;
+		if (INTEL_INFO(dev)->gen >= 5)
+			intel_encoder->post_disable = ilk_post_disable_dp;
 	}
 
 	intel_dig_port->port = port;
diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h
index b8c8bbd8e5f9..07ce04683c30 100644
--- a/drivers/gpu/drm/i915/intel_drv.h
+++ b/drivers/gpu/drm/i915/intel_drv.h
@@ -25,6 +25,7 @@
 #ifndef __INTEL_DRV_H__
 #define __INTEL_DRV_H__
 
+#include <linux/async.h>
 #include <linux/i2c.h>
 #include <linux/hdmi.h>
 #include <drm/i915_drm.h>
@@ -179,6 +180,8 @@ struct intel_panel {
 		bool active_low_pwm;
 		struct backlight_device *device;
 	} backlight;
+
+	void (*backlight_power)(struct intel_connector *, bool enable);
 };
 
 struct intel_connector {
@@ -211,6 +214,7 @@ struct intel_connector {
 
 	/* Cached EDID for eDP and LVDS. May hold ERR_PTR for invalid EDID. */
 	struct edid *edid;
+	struct edid *detect_edid;
 
 	/* since POLL and HPD connectors may use the same HPD line keep the native
 	   state of connector->polled in case hotplug storm detection changes it */
@@ -330,6 +334,7 @@ struct intel_crtc_config {
 
 	/* m2_n2 for eDP downclock */
 	struct intel_link_m_n dp_m2_n2;
+	bool has_drrs;
 
 	/*
 	 * Frequence the dpll for the port should run at. Differs from the
@@ -410,6 +415,7 @@ struct intel_crtc {
 	uint32_t cursor_addr;
 	int16_t cursor_width, cursor_height;
 	uint32_t cursor_cntl;
+	uint32_t cursor_size;
 	uint32_t cursor_base;
 
 	struct intel_plane_config plane_config;
@@ -430,8 +436,6 @@ struct intel_crtc {
 		struct intel_pipe_wm active;
 	} wm;
 
-	wait_queue_head_t vbl_wait;
-
 	int scanline_offset;
 	struct intel_mmio_flip mmio_flip;
 };
@@ -455,6 +459,7 @@ struct intel_plane {
 	unsigned int crtc_w, crtc_h;
 	uint32_t src_x, src_y;
 	uint32_t src_w, src_h;
+	unsigned int rotation;
 
 	/* Since we need to change the watermarks before/after
 	 * enabling/disabling the planes, we need to store the parameters here
@@ -565,6 +570,12 @@ struct intel_dp {
 
 	struct notifier_block edp_notifier;
 
+	/*
+	 * Pipe whose power sequencer is currently locked into
+	 * this port. Only relevant on VLV/CHV.
+	 */
+	enum pipe pps_pipe;
+
 	bool use_tps3;
 	bool can_mst; /* this port supports mst */
 	bool is_mst;
@@ -663,6 +674,10 @@ struct intel_unpin_work {
 #define INTEL_FLIP_COMPLETE	2
 	u32 flip_count;
 	u32 gtt_offset;
+	struct intel_engine_cs *flip_queued_ring;
+	u32 flip_queued_seqno;
+	int flip_queued_vblank;
+	int flip_ready_vblank;
 	bool enable_stall_check;
 };
 
@@ -827,7 +842,6 @@ int intel_get_pipe_from_crtc_id(struct drm_device *dev, void *data,
 enum transcoder intel_pipe_to_cpu_transcoder(struct drm_i915_private *dev_priv,
 					     enum pipe pipe);
 void intel_wait_for_vblank(struct drm_device *dev, int pipe);
-void intel_wait_for_pipe_off(struct drm_device *dev, int pipe);
 int ironlake_get_lanes_required(int target_clock, int link_bw, int bpp);
 void vlv_wait_port_ready(struct drm_i915_private *dev_priv,
 			 struct intel_digital_port *dport);
@@ -848,6 +862,7 @@ __intel_framebuffer_create(struct drm_device *dev,
 void intel_prepare_page_flip(struct drm_device *dev, int plane);
 void intel_finish_page_flip(struct drm_device *dev, int pipe);
 void intel_finish_page_flip_plane(struct drm_device *dev, int plane);
+void intel_check_page_flip(struct drm_device *dev, int pipe);
 
 /* shared dpll functions */
 struct intel_shared_dpll *intel_crtc_to_shared_dpll(struct intel_crtc *crtc);
@@ -882,6 +897,7 @@ void hsw_enable_pc8(struct drm_i915_private *dev_priv);
 void hsw_disable_pc8(struct drm_i915_private *dev_priv);
 void intel_dp_get_m_n(struct intel_crtc *crtc,
 		      struct intel_crtc_config *pipe_config);
+void intel_dp_set_m_n(struct intel_crtc *crtc);
 int intel_dotclock_calculate(int link_freq, const struct intel_link_m_n *m_n);
 void
 ironlake_check_encoder_dotclock(const struct intel_crtc_config *pipe_config,
@@ -896,7 +912,7 @@ void intel_mode_from_pipe_config(struct drm_display_mode *mode,
 				 struct intel_crtc_config *pipe_config);
 int intel_format_to_fourcc(int format);
 void intel_crtc_wait_for_pending_flips(struct drm_crtc *crtc);
-
+void intel_modeset_preclose(struct drm_device *dev, struct drm_file *file);
 
 /* intel_dp.c */
 void intel_dp_init(struct drm_device *dev, int output_reg, enum port port);
@@ -935,6 +951,7 @@ void intel_dp_mst_suspend(struct drm_device *dev);
 void intel_dp_mst_resume(struct drm_device *dev);
 int intel_dp_max_link_bw(struct intel_dp *intel_dp);
 void intel_dp_hot_plug(struct intel_encoder *intel_encoder);
+void vlv_power_sequencer_reset(struct drm_i915_private *dev_priv);
 /* intel_dp_mst.c */
 int intel_dp_mst_encoder_init(struct intel_digital_port *intel_dig_port, int conn_id);
 void intel_dp_mst_encoder_cleanup(struct intel_digital_port *intel_dig_port);
@@ -949,9 +966,9 @@ void intel_dvo_init(struct drm_device *dev);
 /* legacy fbdev emulation in intel_fbdev.c */
 #ifdef CONFIG_DRM_I915_FBDEV
 extern int intel_fbdev_init(struct drm_device *dev);
-extern void intel_fbdev_initial_config(struct drm_device *dev);
+extern void intel_fbdev_initial_config(void *data, async_cookie_t cookie);
 extern void intel_fbdev_fini(struct drm_device *dev);
-extern void intel_fbdev_set_suspend(struct drm_device *dev, int state);
+extern void intel_fbdev_set_suspend(struct drm_device *dev, int state, bool synchronous);
 extern void intel_fbdev_output_poll_changed(struct drm_device *dev);
 extern void intel_fbdev_restore_mode(struct drm_device *dev);
 #else
@@ -960,7 +977,7 @@ static inline int intel_fbdev_init(struct drm_device *dev)
 	return 0;
 }
 
-static inline void intel_fbdev_initial_config(struct drm_device *dev)
+static inline void intel_fbdev_initial_config(void *data, async_cookie_t cookie)
 {
 }
 
@@ -968,7 +985,7 @@ static inline void intel_fbdev_fini(struct drm_device *dev)
 {
 }
 
-static inline void intel_fbdev_set_suspend(struct drm_device *dev, int state)
+static inline void intel_fbdev_set_suspend(struct drm_device *dev, int state, bool synchronous)
 {
 }
 
@@ -1091,7 +1108,10 @@ bool intel_sdvo_init(struct drm_device *dev, uint32_t sdvo_reg, bool is_sdvob);
 int intel_plane_init(struct drm_device *dev, enum pipe pipe, int plane);
 void intel_flush_primary_plane(struct drm_i915_private *dev_priv,
 			       enum plane plane);
-void intel_plane_restore(struct drm_plane *plane);
+int intel_plane_set_property(struct drm_plane *plane,
+			     struct drm_property *prop,
+			     uint64_t val);
+int intel_plane_restore(struct drm_plane *plane);
 void intel_plane_disable(struct drm_plane *plane);
 int intel_sprite_set_colorkey(struct drm_device *dev, void *data,
 			      struct drm_file *file_priv);
diff --git a/drivers/gpu/drm/i915/intel_dsi.c b/drivers/gpu/drm/i915/intel_dsi.c
index 670c29a7b5dd..5bd9e09ad3c5 100644
--- a/drivers/gpu/drm/i915/intel_dsi.c
+++ b/drivers/gpu/drm/i915/intel_dsi.c
@@ -184,7 +184,7 @@ static void intel_dsi_pre_enable(struct intel_encoder *encoder)
 
 	/* update the hw state for DPLL */
 	intel_crtc->config.dpll_hw_state.dpll = DPLL_INTEGRATED_CLOCK_VLV |
-						DPLL_REFA_CLK_ENABLE_VLV;
+		DPLL_REFA_CLK_ENABLE_VLV;
 
 	tmp = I915_READ(DSPCLK_GATE_D);
 	tmp |= DPOUNIT_CLOCK_GATE_DISABLE;
@@ -259,8 +259,8 @@ static void intel_dsi_disable(struct intel_encoder *encoder)
 	temp = I915_READ(MIPI_CTRL(pipe));
 	temp &= ~ESCAPE_CLOCK_DIVIDER_MASK;
 	I915_WRITE(MIPI_CTRL(pipe), temp |
-			intel_dsi->escape_clk_div <<
-			ESCAPE_CLOCK_DIVIDER_SHIFT);
+		   intel_dsi->escape_clk_div <<
+		   ESCAPE_CLOCK_DIVIDER_SHIFT);
 
 	I915_WRITE(MIPI_EOT_DISABLE(pipe), CLOCKSTOP);
 
@@ -297,7 +297,7 @@ static void intel_dsi_clear_device_ready(struct intel_encoder *encoder)
 	usleep_range(2000, 2500);
 
 	if (wait_for(((I915_READ(MIPI_PORT_CTRL(pipe)) & AFE_LATCHOUT)
-					== 0x00000), 30))
+		      == 0x00000), 30))
 		DRM_ERROR("DSI LP not going Low\n");
 
 	val = I915_READ(MIPI_PORT_CTRL(pipe));
@@ -423,9 +423,11 @@ static u16 txclkesc(u32 divider, unsigned int us)
 }
 
 /* return pixels in terms of txbyteclkhs */
-static u16 txbyteclkhs(u16 pixels, int bpp, int lane_count)
+static u16 txbyteclkhs(u16 pixels, int bpp, int lane_count,
+		       u16 burst_mode_ratio)
 {
-	return DIV_ROUND_UP(DIV_ROUND_UP(pixels * bpp, 8), lane_count);
+	return DIV_ROUND_UP(DIV_ROUND_UP(pixels * bpp * burst_mode_ratio,
+					 8 * 100), lane_count);
 }
 
 static void set_dsi_timings(struct drm_encoder *encoder,
@@ -451,10 +453,12 @@ static void set_dsi_timings(struct drm_encoder *encoder,
 	vbp = mode->vtotal - mode->vsync_end;
 
 	/* horizontal values are in terms of high speed byte clock */
-	hactive = txbyteclkhs(hactive, bpp, lane_count);
-	hfp = txbyteclkhs(hfp, bpp, lane_count);
-	hsync = txbyteclkhs(hsync, bpp, lane_count);
-	hbp = txbyteclkhs(hbp, bpp, lane_count);
+	hactive = txbyteclkhs(hactive, bpp, lane_count,
+			      intel_dsi->burst_mode_ratio);
+	hfp = txbyteclkhs(hfp, bpp, lane_count, intel_dsi->burst_mode_ratio);
+	hsync = txbyteclkhs(hsync, bpp, lane_count,
+			    intel_dsi->burst_mode_ratio);
+	hbp = txbyteclkhs(hbp, bpp, lane_count, intel_dsi->burst_mode_ratio);
 
 	I915_WRITE(MIPI_HACTIVE_AREA_COUNT(pipe), hactive);
 	I915_WRITE(MIPI_HFP_COUNT(pipe), hfp);
@@ -541,12 +545,14 @@ static void intel_dsi_prepare(struct intel_encoder *intel_encoder)
 	    intel_dsi->video_mode_format == VIDEO_MODE_BURST) {
 		I915_WRITE(MIPI_HS_TX_TIMEOUT(pipe),
 			   txbyteclkhs(adjusted_mode->htotal, bpp,
-				       intel_dsi->lane_count) + 1);
+				       intel_dsi->lane_count,
+				       intel_dsi->burst_mode_ratio) + 1);
 	} else {
 		I915_WRITE(MIPI_HS_TX_TIMEOUT(pipe),
 			   txbyteclkhs(adjusted_mode->vtotal *
 				       adjusted_mode->htotal,
-				       bpp, intel_dsi->lane_count) + 1);
+				       bpp, intel_dsi->lane_count,
+				       intel_dsi->burst_mode_ratio) + 1);
 	}
 	I915_WRITE(MIPI_LP_RX_TIMEOUT(pipe), intel_dsi->lp_rx_timeout);
 	I915_WRITE(MIPI_TURN_AROUND_TIMEOUT(pipe), intel_dsi->turn_arnd_val);
@@ -576,7 +582,7 @@ static void intel_dsi_prepare(struct intel_encoder *intel_encoder)
 	 * XXX: write MIPI_STOP_STATE_STALL?
 	 */
 	I915_WRITE(MIPI_HIGH_LOW_SWITCH_COUNT(pipe),
-						intel_dsi->hs_to_lp_count);
+		   intel_dsi->hs_to_lp_count);
 
 	/* XXX: low power clock equivalence in terms of byte clock. the number
 	 * of byte clocks occupied in one low power clock. based on txbyteclkhs
@@ -601,10 +607,10 @@ static void intel_dsi_prepare(struct intel_encoder *intel_encoder)
 		 * 64 like 1366 x 768. Enable RANDOM resolution support for such
 		 * panels by default */
 		I915_WRITE(MIPI_VIDEO_MODE_FORMAT(pipe),
-				intel_dsi->video_frmt_cfg_bits |
-				intel_dsi->video_mode_format |
-				IP_TG_CONFIG |
-				RANDOM_DPI_DISPLAY_RESOLUTION);
+			   intel_dsi->video_frmt_cfg_bits |
+			   intel_dsi->video_mode_format |
+			   IP_TG_CONFIG |
+			   RANDOM_DPI_DISPLAY_RESOLUTION);
 }
 
 static void intel_dsi_pre_pll_enable(struct intel_encoder *encoder)
diff --git a/drivers/gpu/drm/i915/intel_dsi.h b/drivers/gpu/drm/i915/intel_dsi.h
index fd51867fd0d3..657eb5c1b9d8 100644
--- a/drivers/gpu/drm/i915/intel_dsi.h
+++ b/drivers/gpu/drm/i915/intel_dsi.h
@@ -116,6 +116,8 @@ struct intel_dsi {
 	u16 clk_hs_to_lp_count;
 
 	u16 init_count;
+	u32 pclk;
+	u16 burst_mode_ratio;
 
 	/* all delays in ms */
 	u16 backlight_off_delay;
diff --git a/drivers/gpu/drm/i915/intel_dsi_cmd.c b/drivers/gpu/drm/i915/intel_dsi_cmd.c
index 7f1430ac8543..f4767fd2ebeb 100644
--- a/drivers/gpu/drm/i915/intel_dsi_cmd.c
+++ b/drivers/gpu/drm/i915/intel_dsi_cmd.c
@@ -430,7 +430,7 @@ void wait_for_dsi_fifo_empty(struct intel_dsi *intel_dsi)
 	u32 mask;
 
 	mask = LP_CTRL_FIFO_EMPTY | HS_CTRL_FIFO_EMPTY |
-					LP_DATA_FIFO_EMPTY | HS_DATA_FIFO_EMPTY;
+		LP_DATA_FIFO_EMPTY | HS_DATA_FIFO_EMPTY;
 
 	if (wait_for((I915_READ(MIPI_GEN_FIFO_STAT(pipe)) & mask) == mask, 100))
 		DRM_ERROR("DPI FIFOs are not empty\n");
diff --git a/drivers/gpu/drm/i915/intel_dsi_panel_vbt.c b/drivers/gpu/drm/i915/intel_dsi_panel_vbt.c
index 47c7584a4aa0..f6bdd44069ce 100644
--- a/drivers/gpu/drm/i915/intel_dsi_panel_vbt.c
+++ b/drivers/gpu/drm/i915/intel_dsi_panel_vbt.c
@@ -271,6 +271,8 @@ static bool generic_init(struct intel_dsi_device *dsi)
 	u32 ths_prepare_ns, tclk_trail_ns;
 	u32 tclk_prepare_clkzero, ths_prepare_hszero;
 	u32 lp_to_hs_switch, hs_to_lp_switch;
+	u32 pclk, computed_ddr;
+	u16 burst_mode_ratio;
 
 	DRM_DEBUG_KMS("\n");
 
@@ -284,8 +286,6 @@ static bool generic_init(struct intel_dsi_device *dsi)
 	else if (intel_dsi->pixel_format == VID_MODE_FORMAT_RGB565)
 		bits_per_pixel = 16;
 
-	bitrate = (mode->clock * bits_per_pixel) / intel_dsi->lane_count;
-
 	intel_dsi->operation_mode = mipi_config->is_cmd_mode;
 	intel_dsi->video_mode_format = mipi_config->video_transfer_mode;
 	intel_dsi->escape_clk_div = mipi_config->byte_clk_sel;
@@ -297,6 +297,40 @@ static bool generic_init(struct intel_dsi_device *dsi)
 	intel_dsi->video_frmt_cfg_bits =
 		mipi_config->bta_enabled ? DISABLE_VIDEO_BTA : 0;
 
+	pclk = mode->clock;
+
+	/* Burst Mode Ratio
+	 * Target ddr frequency from VBT / non burst ddr freq
+	 * multiply by 100 to preserve remainder
+	 */
+	if (intel_dsi->video_mode_format == VIDEO_MODE_BURST) {
+		if (mipi_config->target_burst_mode_freq) {
+			computed_ddr =
+				(pclk * bits_per_pixel) / intel_dsi->lane_count;
+
+			if (mipi_config->target_burst_mode_freq <
+								computed_ddr) {
+				DRM_ERROR("Burst mode freq is less than computed\n");
+				return false;
+			}
+
+			burst_mode_ratio = DIV_ROUND_UP(
+				mipi_config->target_burst_mode_freq * 100,
+				computed_ddr);
+
+			pclk = DIV_ROUND_UP(pclk * burst_mode_ratio, 100);
+		} else {
+			DRM_ERROR("Burst mode target is not set\n");
+			return false;
+		}
+	} else
+		burst_mode_ratio = 100;
+
+	intel_dsi->burst_mode_ratio = burst_mode_ratio;
+	intel_dsi->pclk = pclk;
+
+	bitrate = (pclk * bits_per_pixel) / intel_dsi->lane_count;
+
 	switch (intel_dsi->escape_clk_div) {
 	case 0:
 		tlpx_ns = 50;
diff --git a/drivers/gpu/drm/i915/intel_dsi_pll.c b/drivers/gpu/drm/i915/intel_dsi_pll.c
index d8bb1ea2f0da..fa7a6ca34cd6 100644
--- a/drivers/gpu/drm/i915/intel_dsi_pll.c
+++ b/drivers/gpu/drm/i915/intel_dsi_pll.c
@@ -134,8 +134,7 @@ static u32 dsi_rr_formula(const struct drm_display_mode *mode,
 #else
 
 /* Get DSI clock from pixel clock */
-static u32 dsi_clk_from_pclk(const struct drm_display_mode *mode,
-			  int pixel_format, int lane_count)
+static u32 dsi_clk_from_pclk(u32 pclk, int pixel_format, int lane_count)
 {
 	u32 dsi_clk_khz;
 	u32 bpp;
@@ -156,7 +155,7 @@ static u32 dsi_clk_from_pclk(const struct drm_display_mode *mode,
 
 	/* DSI data rate = pixel clock * bits per pixel / lane count
 	   pixel clock is converted from KHz to Hz */
-	dsi_clk_khz = DIV_ROUND_CLOSEST(mode->clock * bpp, lane_count);
+	dsi_clk_khz = DIV_ROUND_CLOSEST(pclk * bpp, lane_count);
 
 	return dsi_clk_khz;
 }
@@ -191,7 +190,7 @@ static int dsi_calc_mnp(u32 dsi_clk, struct dsi_mnp *dsi_mnp)
 	for (m = 62; m <= 92; m++) {
 		for (p = 2; p <= 6; p++) {
 			/* Find the optimal m and p divisors
-			with minimal error +/- the required clock */
+			   with minimal error +/- the required clock */
 			calc_dsi_clk = (m * ref_clk) / p;
 			if (calc_dsi_clk == target_dsi_clk) {
 				calc_m = m;
@@ -228,15 +227,13 @@ static int dsi_calc_mnp(u32 dsi_clk, struct dsi_mnp *dsi_mnp)
 static void vlv_configure_dsi_pll(struct intel_encoder *encoder)
 {
 	struct drm_i915_private *dev_priv = encoder->base.dev->dev_private;
-	struct intel_crtc *intel_crtc = to_intel_crtc(encoder->base.crtc);
-	const struct drm_display_mode *mode = &intel_crtc->config.adjusted_mode;
 	struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base);
 	int ret;
 	struct dsi_mnp dsi_mnp;
 	u32 dsi_clk;
 
-	dsi_clk = dsi_clk_from_pclk(mode, intel_dsi->pixel_format,
-						intel_dsi->lane_count);
+	dsi_clk = dsi_clk_from_pclk(intel_dsi->pclk, intel_dsi->pixel_format,
+				    intel_dsi->lane_count);
 
 	ret = dsi_calc_mnp(dsi_clk, &dsi_mnp);
 	if (ret) {
@@ -318,8 +315,8 @@ static void assert_bpp_mismatch(int pixel_format, int pipe_bpp)
 	}
 
 	WARN(bpp != pipe_bpp,
-		"bpp match assertion failure (expected %d, current %d)\n",
-		bpp, pipe_bpp);
+	     "bpp match assertion failure (expected %d, current %d)\n",
+	     bpp, pipe_bpp);
 }
 
 u32 vlv_get_dsi_pclk(struct intel_encoder *encoder, int pipe_bpp)
diff --git a/drivers/gpu/drm/i915/intel_dvo.c b/drivers/gpu/drm/i915/intel_dvo.c
index 56b47d2ffaf7..e40e3df33517 100644
--- a/drivers/gpu/drm/i915/intel_dvo.c
+++ b/drivers/gpu/drm/i915/intel_dvo.c
@@ -85,7 +85,7 @@ static const struct intel_dvo_device intel_dvo_devices[] = {
 	{
 	        .type = INTEL_DVO_CHIP_TMDS,
 		.name = "ns2501",
-		.dvo_reg = DVOC,
+		.dvo_reg = DVOB,
 		.slave_addr = NS2501_ADDR,
 		.dev_ops = &ns2501_ops,
        }
@@ -185,12 +185,13 @@ static void intel_enable_dvo(struct intel_encoder *encoder)
 	u32 dvo_reg = intel_dvo->dev.dvo_reg;
 	u32 temp = I915_READ(dvo_reg);
 
-	I915_WRITE(dvo_reg, temp | DVO_ENABLE);
-	I915_READ(dvo_reg);
 	intel_dvo->dev.dev_ops->mode_set(&intel_dvo->dev,
 					 &crtc->config.requested_mode,
 					 &crtc->config.adjusted_mode);
 
+	I915_WRITE(dvo_reg, temp | DVO_ENABLE);
+	I915_READ(dvo_reg);
+
 	intel_dvo->dev.dev_ops->dpms(&intel_dvo->dev, true);
 }
 
@@ -226,10 +227,6 @@ static void intel_dvo_dpms(struct drm_connector *connector, int mode)
 
 		intel_crtc_update_dpms(crtc);
 
-		intel_dvo->dev.dev_ops->mode_set(&intel_dvo->dev,
-						 &config->requested_mode,
-						 &config->adjusted_mode);
-
 		intel_dvo->dev.dev_ops->dpms(&intel_dvo->dev, true);
 	} else {
 		intel_dvo->dev.dev_ops->dpms(&intel_dvo->dev, false);
diff --git a/drivers/gpu/drm/i915/intel_fbdev.c b/drivers/gpu/drm/i915/intel_fbdev.c
index f475414671d8..9b584f3fbb99 100644
--- a/drivers/gpu/drm/i915/intel_fbdev.c
+++ b/drivers/gpu/drm/i915/intel_fbdev.c
@@ -24,8 +24,10 @@
  *     David Airlie
  */
 
+#include <linux/async.h>
 #include <linux/module.h>
 #include <linux/kernel.h>
+#include <linux/console.h>
 #include <linux/errno.h>
 #include <linux/string.h>
 #include <linux/mm.h>
@@ -331,24 +333,6 @@ static bool intel_fb_initial_config(struct drm_fb_helper *fb_helper,
 	int num_connectors_enabled = 0;
 	int num_connectors_detected = 0;
 
-	/*
-	 * If the user specified any force options, just bail here
-	 * and use that config.
-	 */
-	for (i = 0; i < fb_helper->connector_count; i++) {
-		struct drm_fb_helper_connector *fb_conn;
-		struct drm_connector *connector;
-
-		fb_conn = fb_helper->connector_info[i];
-		connector = fb_conn->connector;
-
-		if (!enabled[i])
-			continue;
-
-		if (connector->force != DRM_FORCE_UNSPECIFIED)
-			return false;
-	}
-
 	save_enabled = kcalloc(dev->mode_config.num_connector, sizeof(bool),
 			       GFP_KERNEL);
 	if (!save_enabled)
@@ -374,8 +358,18 @@ static bool intel_fb_initial_config(struct drm_fb_helper *fb_helper,
 			continue;
 		}
 
+		if (connector->force == DRM_FORCE_OFF) {
+			DRM_DEBUG_KMS("connector %s is disabled by user, skipping\n",
+				      connector->name);
+			enabled[i] = false;
+			continue;
+		}
+
 		encoder = connector->encoder;
 		if (!encoder || WARN_ON(!encoder->crtc)) {
+			if (connector->force > DRM_FORCE_OFF)
+				goto bail;
+
 			DRM_DEBUG_KMS("connector %s has no encoder or crtc, skipping\n",
 				      connector->name);
 			enabled[i] = false;
@@ -394,8 +388,7 @@ static bool intel_fb_initial_config(struct drm_fb_helper *fb_helper,
 		for (j = 0; j < fb_helper->connector_count; j++) {
 			if (crtcs[j] == new_crtc) {
 				DRM_DEBUG_KMS("fallback: cloned configuration\n");
-				fallback = true;
-				goto out;
+				goto bail;
 			}
 		}
 
@@ -466,8 +459,8 @@ static bool intel_fb_initial_config(struct drm_fb_helper *fb_helper,
 		fallback = true;
 	}
 
-out:
 	if (fallback) {
+bail:
 		DRM_DEBUG_KMS("Not using firmware configuration\n");
 		memcpy(enabled, save_enabled, dev->mode_config.num_connector);
 		kfree(save_enabled);
@@ -636,6 +629,15 @@ out:
 	return false;
 }
 
+static void intel_fbdev_suspend_worker(struct work_struct *work)
+{
+	intel_fbdev_set_suspend(container_of(work,
+					     struct drm_i915_private,
+					     fbdev_suspend_work)->dev,
+				FBINFO_STATE_RUNNING,
+				true);
+}
+
 int intel_fbdev_init(struct drm_device *dev)
 {
 	struct intel_fbdev *ifbdev;
@@ -662,14 +664,16 @@ int intel_fbdev_init(struct drm_device *dev)
 	}
 
 	dev_priv->fbdev = ifbdev;
+	INIT_WORK(&dev_priv->fbdev_suspend_work, intel_fbdev_suspend_worker);
+
 	drm_fb_helper_single_add_all_connectors(&ifbdev->helper);
 
 	return 0;
 }
 
-void intel_fbdev_initial_config(struct drm_device *dev)
+void intel_fbdev_initial_config(void *data, async_cookie_t cookie)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = data;
 	struct intel_fbdev *ifbdev = dev_priv->fbdev;
 
 	/* Due to peculiar init order wrt to hpd handling this is separate. */
@@ -682,12 +686,15 @@ void intel_fbdev_fini(struct drm_device *dev)
 	if (!dev_priv->fbdev)
 		return;
 
+	flush_work(&dev_priv->fbdev_suspend_work);
+
+	async_synchronize_full();
 	intel_fbdev_destroy(dev, dev_priv->fbdev);
 	kfree(dev_priv->fbdev);
 	dev_priv->fbdev = NULL;
 }
 
-void intel_fbdev_set_suspend(struct drm_device *dev, int state)
+void intel_fbdev_set_suspend(struct drm_device *dev, int state, bool synchronous)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_fbdev *ifbdev = dev_priv->fbdev;
@@ -698,6 +705,33 @@ void intel_fbdev_set_suspend(struct drm_device *dev, int state)
 
 	info = ifbdev->helper.fbdev;
 
+	if (synchronous) {
+		/* Flush any pending work to turn the console on, and then
+		 * wait to turn it off. It must be synchronous as we are
+		 * about to suspend or unload the driver.
+		 *
+		 * Note that from within the work-handler, we cannot flush
+		 * ourselves, so only flush outstanding work upon suspend!
+		 */
+		if (state != FBINFO_STATE_RUNNING)
+			flush_work(&dev_priv->fbdev_suspend_work);
+		console_lock();
+	} else {
+		/*
+		 * The console lock can be pretty contented on resume due
+		 * to all the printk activity.  Try to keep it out of the hot
+		 * path of resume if possible.
+		 */
+		WARN_ON(state != FBINFO_STATE_RUNNING);
+		if (!console_trylock()) {
+			/* Don't block our own workqueue as this can
+			 * be run in parallel with other i915.ko tasks.
+			 */
+			schedule_work(&dev_priv->fbdev_suspend_work);
+			return;
+		}
+	}
+
 	/* On resume from hibernation: If the object is shmemfs backed, it has
 	 * been restored from swap. If the object is stolen however, it will be
 	 * full of whatever garbage was left in there.
@@ -706,6 +740,7 @@ void intel_fbdev_set_suspend(struct drm_device *dev, int state)
 		memset_io(info->screen_base, 0, info->screen_size);
 
 	fb_set_suspend(info, state);
+	console_unlock();
 }
 
 void intel_fbdev_output_poll_changed(struct drm_device *dev)
diff --git a/drivers/gpu/drm/i915/intel_hdmi.c b/drivers/gpu/drm/i915/intel_hdmi.c
index 5a9de21637b7..29ec1535992d 100644
--- a/drivers/gpu/drm/i915/intel_hdmi.c
+++ b/drivers/gpu/drm/i915/intel_hdmi.c
@@ -869,10 +869,15 @@ static enum drm_mode_status
 intel_hdmi_mode_valid(struct drm_connector *connector,
 		      struct drm_display_mode *mode)
 {
-	if (mode->clock > hdmi_portclock_limit(intel_attached_hdmi(connector),
-					       true))
+	int clock = mode->clock;
+
+	if (mode->flags & DRM_MODE_FLAG_DBLCLK)
+		clock *= 2;
+
+	if (clock > hdmi_portclock_limit(intel_attached_hdmi(connector),
+					 true))
 		return MODE_CLOCK_HIGH;
-	if (mode->clock < 20000)
+	if (clock < 20000)
 		return MODE_CLOCK_LOW;
 
 	if (mode->flags & DRM_MODE_FLAG_DBLSCAN)
@@ -890,7 +895,7 @@ static bool hdmi_12bpc_possible(struct intel_crtc *crtc)
 	if (HAS_GMCH_DISPLAY(dev))
 		return false;
 
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list, base.head) {
+	for_each_intel_encoder(dev, encoder) {
 		if (encoder->new_crtc != crtc)
 			continue;
 
@@ -926,6 +931,10 @@ bool intel_hdmi_compute_config(struct intel_encoder *encoder,
 			intel_hdmi->color_range = 0;
 	}
 
+	if (adjusted_mode->flags & DRM_MODE_FLAG_DBLCLK) {
+		pipe_config->pixel_multiplier = 2;
+	}
+
 	if (intel_hdmi->color_range)
 		pipe_config->limited_color_range = true;
 
@@ -967,104 +976,117 @@ bool intel_hdmi_compute_config(struct intel_encoder *encoder,
 	return true;
 }
 
-static enum drm_connector_status
-intel_hdmi_detect(struct drm_connector *connector, bool force)
+static void
+intel_hdmi_unset_edid(struct drm_connector *connector)
 {
-	struct drm_device *dev = connector->dev;
 	struct intel_hdmi *intel_hdmi = intel_attached_hdmi(connector);
-	struct intel_digital_port *intel_dig_port =
-		hdmi_to_dig_port(intel_hdmi);
-	struct intel_encoder *intel_encoder = &intel_dig_port->base;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct edid *edid;
-	enum intel_display_power_domain power_domain;
-	enum drm_connector_status status = connector_status_disconnected;
 
-	DRM_DEBUG_KMS("[CONNECTOR:%d:%s]\n",
-		      connector->base.id, connector->name);
+	intel_hdmi->has_hdmi_sink = false;
+	intel_hdmi->has_audio = false;
+	intel_hdmi->rgb_quant_range_selectable = false;
+
+	kfree(to_intel_connector(connector)->detect_edid);
+	to_intel_connector(connector)->detect_edid = NULL;
+}
+
+static bool
+intel_hdmi_set_edid(struct drm_connector *connector)
+{
+	struct drm_i915_private *dev_priv = to_i915(connector->dev);
+	struct intel_hdmi *intel_hdmi = intel_attached_hdmi(connector);
+	struct intel_encoder *intel_encoder =
+		&hdmi_to_dig_port(intel_hdmi)->base;
+	enum intel_display_power_domain power_domain;
+	struct edid *edid;
+	bool connected = false;
 
 	power_domain = intel_display_port_power_domain(intel_encoder);
 	intel_display_power_get(dev_priv, power_domain);
 
-	intel_hdmi->has_hdmi_sink = false;
-	intel_hdmi->has_audio = false;
-	intel_hdmi->rgb_quant_range_selectable = false;
 	edid = drm_get_edid(connector,
 			    intel_gmbus_get_adapter(dev_priv,
 						    intel_hdmi->ddc_bus));
 
-	if (edid) {
-		if (edid->input & DRM_EDID_INPUT_DIGITAL) {
-			status = connector_status_connected;
-			if (intel_hdmi->force_audio != HDMI_AUDIO_OFF_DVI)
-				intel_hdmi->has_hdmi_sink =
-						drm_detect_hdmi_monitor(edid);
-			intel_hdmi->has_audio = drm_detect_monitor_audio(edid);
-			intel_hdmi->rgb_quant_range_selectable =
-				drm_rgb_quant_range_selectable(edid);
-		}
-		kfree(edid);
-	}
+	intel_display_power_put(dev_priv, power_domain);
+
+	to_intel_connector(connector)->detect_edid = edid;
+	if (edid && edid->input & DRM_EDID_INPUT_DIGITAL) {
+		intel_hdmi->rgb_quant_range_selectable =
+			drm_rgb_quant_range_selectable(edid);
 
-	if (status == connector_status_connected) {
+		intel_hdmi->has_audio = drm_detect_monitor_audio(edid);
 		if (intel_hdmi->force_audio != HDMI_AUDIO_AUTO)
 			intel_hdmi->has_audio =
-				(intel_hdmi->force_audio == HDMI_AUDIO_ON);
-		intel_encoder->type = INTEL_OUTPUT_HDMI;
+				intel_hdmi->force_audio == HDMI_AUDIO_ON;
+
+		if (intel_hdmi->force_audio != HDMI_AUDIO_OFF_DVI)
+			intel_hdmi->has_hdmi_sink =
+				drm_detect_hdmi_monitor(edid);
+
+		connected = true;
 	}
 
-	intel_display_power_put(dev_priv, power_domain);
+	return connected;
+}
+
+static enum drm_connector_status
+intel_hdmi_detect(struct drm_connector *connector, bool force)
+{
+	enum drm_connector_status status;
+
+	DRM_DEBUG_KMS("[CONNECTOR:%d:%s]\n",
+		      connector->base.id, connector->name);
+
+	intel_hdmi_unset_edid(connector);
+
+	if (intel_hdmi_set_edid(connector)) {
+		struct intel_hdmi *intel_hdmi = intel_attached_hdmi(connector);
+
+		hdmi_to_dig_port(intel_hdmi)->base.type = INTEL_OUTPUT_HDMI;
+		status = connector_status_connected;
+	} else
+		status = connector_status_disconnected;
 
 	return status;
 }
 
-static int intel_hdmi_get_modes(struct drm_connector *connector)
+static void
+intel_hdmi_force(struct drm_connector *connector)
 {
-	struct intel_encoder *intel_encoder = intel_attached_encoder(connector);
-	struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(&intel_encoder->base);
-	struct drm_i915_private *dev_priv = connector->dev->dev_private;
-	enum intel_display_power_domain power_domain;
-	int ret;
+	struct intel_hdmi *intel_hdmi = intel_attached_hdmi(connector);
 
-	/* We should parse the EDID data and find out if it's an HDMI sink so
-	 * we can send audio to it.
-	 */
+	DRM_DEBUG_KMS("[CONNECTOR:%d:%s]\n",
+		      connector->base.id, connector->name);
 
-	power_domain = intel_display_port_power_domain(intel_encoder);
-	intel_display_power_get(dev_priv, power_domain);
+	intel_hdmi_unset_edid(connector);
 
-	ret = intel_ddc_get_modes(connector,
-				   intel_gmbus_get_adapter(dev_priv,
-							   intel_hdmi->ddc_bus));
+	if (connector->status != connector_status_connected)
+		return;
 
-	intel_display_power_put(dev_priv, power_domain);
+	intel_hdmi_set_edid(connector);
+	hdmi_to_dig_port(intel_hdmi)->base.type = INTEL_OUTPUT_HDMI;
+}
 
-	return ret;
+static int intel_hdmi_get_modes(struct drm_connector *connector)
+{
+	struct edid *edid;
+
+	edid = to_intel_connector(connector)->detect_edid;
+	if (edid == NULL)
+		return 0;
+
+	return intel_connector_update_modes(connector, edid);
 }
 
 static bool
 intel_hdmi_detect_audio(struct drm_connector *connector)
 {
-	struct intel_encoder *intel_encoder = intel_attached_encoder(connector);
-	struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(&intel_encoder->base);
-	struct drm_i915_private *dev_priv = connector->dev->dev_private;
-	enum intel_display_power_domain power_domain;
-	struct edid *edid;
 	bool has_audio = false;
+	struct edid *edid;
 
-	power_domain = intel_display_port_power_domain(intel_encoder);
-	intel_display_power_get(dev_priv, power_domain);
-
-	edid = drm_get_edid(connector,
-			    intel_gmbus_get_adapter(dev_priv,
-						    intel_hdmi->ddc_bus));
-	if (edid) {
-		if (edid->input & DRM_EDID_INPUT_DIGITAL)
-			has_audio = drm_detect_monitor_audio(edid);
-		kfree(edid);
-	}
-
-	intel_display_power_put(dev_priv, power_domain);
+	edid = to_intel_connector(connector)->detect_edid;
+	if (edid && edid->input & DRM_EDID_INPUT_DIGITAL)
+		has_audio = drm_detect_monitor_audio(edid);
 
 	return has_audio;
 }
@@ -1265,6 +1287,8 @@ static void chv_hdmi_pre_pll_enable(struct intel_encoder *encoder)
 	enum pipe pipe = intel_crtc->pipe;
 	u32 val;
 
+	intel_hdmi_prepare(encoder);
+
 	mutex_lock(&dev_priv->dpio_lock);
 
 	/* program left/right clock distribution */
@@ -1434,8 +1458,8 @@ static void chv_hdmi_pre_enable(struct intel_encoder *encoder)
 
 	for (i = 0; i < 4; i++) {
 		val = vlv_dpio_read(dev_priv, pipe, CHV_TX_DW2(ch, i));
-		val &= ~DPIO_SWING_MARGIN_MASK;
-		val |= 102 << DPIO_SWING_MARGIN_SHIFT;
+		val &= ~DPIO_SWING_MARGIN000_MASK;
+		val |= 102 << DPIO_SWING_MARGIN000_SHIFT;
 		vlv_dpio_write(dev_priv, pipe, CHV_TX_DW2(ch, i), val);
 	}
 
@@ -1482,6 +1506,7 @@ static void chv_hdmi_pre_enable(struct intel_encoder *encoder)
 
 static void intel_hdmi_destroy(struct drm_connector *connector)
 {
+	kfree(to_intel_connector(connector)->detect_edid);
 	drm_connector_cleanup(connector);
 	kfree(connector);
 }
@@ -1489,6 +1514,7 @@ static void intel_hdmi_destroy(struct drm_connector *connector)
 static const struct drm_connector_funcs intel_hdmi_connector_funcs = {
 	.dpms = intel_connector_dpms,
 	.detect = intel_hdmi_detect,
+	.force = intel_hdmi_force,
 	.fill_modes = drm_helper_probe_single_connector_modes,
 	.set_property = intel_hdmi_set_property,
 	.destroy = intel_hdmi_destroy,
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
new file mode 100644
index 000000000000..bafd38b5703e
--- /dev/null
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -0,0 +1,1766 @@
+/*
+ * Copyright © 2014 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.
+ *
+ * Authors:
+ *    Ben Widawsky <ben@bwidawsk.net>
+ *    Michel Thierry <michel.thierry@intel.com>
+ *    Thomas Daniel <thomas.daniel@intel.com>
+ *    Oscar Mateo <oscar.mateo@intel.com>
+ *
+ */
+
+/**
+ * DOC: Logical Rings, Logical Ring Contexts and Execlists
+ *
+ * Motivation:
+ * GEN8 brings an expansion of the HW contexts: "Logical Ring Contexts".
+ * These expanded contexts enable a number of new abilities, especially
+ * "Execlists" (also implemented in this file).
+ *
+ * One of the main differences with the legacy HW contexts is that logical
+ * ring contexts incorporate many more things to the context's state, like
+ * PDPs or ringbuffer control registers:
+ *
+ * The reason why PDPs are included in the context is straightforward: as
+ * PPGTTs (per-process GTTs) are actually per-context, having the PDPs
+ * contained there mean you don't need to do a ppgtt->switch_mm yourself,
+ * instead, the GPU will do it for you on the context switch.
+ *
+ * But, what about the ringbuffer control registers (head, tail, etc..)?
+ * shouldn't we just need a set of those per engine command streamer? This is
+ * where the name "Logical Rings" starts to make sense: by virtualizing the
+ * rings, the engine cs shifts to a new "ring buffer" with every context
+ * switch. When you want to submit a workload to the GPU you: A) choose your
+ * context, B) find its appropriate virtualized ring, C) write commands to it
+ * and then, finally, D) tell the GPU to switch to that context.
+ *
+ * Instead of the legacy MI_SET_CONTEXT, the way you tell the GPU to switch
+ * to a contexts is via a context execution list, ergo "Execlists".
+ *
+ * LRC implementation:
+ * Regarding the creation of contexts, we have:
+ *
+ * - One global default context.
+ * - One local default context for each opened fd.
+ * - One local extra context for each context create ioctl call.
+ *
+ * Now that ringbuffers belong per-context (and not per-engine, like before)
+ * and that contexts are uniquely tied to a given engine (and not reusable,
+ * like before) we need:
+ *
+ * - One ringbuffer per-engine inside each context.
+ * - One backing object per-engine inside each context.
+ *
+ * The global default context starts its life with these new objects fully
+ * allocated and populated. The local default context for each opened fd is
+ * more complex, because we don't know at creation time which engine is going
+ * to use them. To handle this, we have implemented a deferred creation of LR
+ * contexts:
+ *
+ * The local context starts its life as a hollow or blank holder, that only
+ * gets populated for a given engine once we receive an execbuffer. If later
+ * on we receive another execbuffer ioctl for the same context but a different
+ * engine, we allocate/populate a new ringbuffer and context backing object and
+ * so on.
+ *
+ * Finally, regarding local contexts created using the ioctl call: as they are
+ * only allowed with the render ring, we can allocate & populate them right
+ * away (no need to defer anything, at least for now).
+ *
+ * Execlists implementation:
+ * Execlists are the new method by which, on gen8+ hardware, workloads are
+ * submitted for execution (as opposed to the legacy, ringbuffer-based, method).
+ * This method works as follows:
+ *
+ * When a request is committed, its commands (the BB start and any leading or
+ * trailing commands, like the seqno breadcrumbs) are placed in the ringbuffer
+ * for the appropriate context. The tail pointer in the hardware context is not
+ * updated at this time, but instead, kept by the driver in the ringbuffer
+ * structure. A structure representing this request is added to a request queue
+ * for the appropriate engine: this structure contains a copy of the context's
+ * tail after the request was written to the ring buffer and a pointer to the
+ * context itself.
+ *
+ * If the engine's request queue was empty before the request was added, the
+ * queue is processed immediately. Otherwise the queue will be processed during
+ * a context switch interrupt. In any case, elements on the queue will get sent
+ * (in pairs) to the GPU's ExecLists Submit Port (ELSP, for short) with a
+ * globally unique 20-bits submission ID.
+ *
+ * When execution of a request completes, the GPU updates the context status
+ * buffer with a context complete event and generates a context switch interrupt.
+ * During the interrupt handling, the driver examines the events in the buffer:
+ * for each context complete event, if the announced ID matches that on the head
+ * of the request queue, then that request is retired and removed from the queue.
+ *
+ * After processing, if any requests were retired and the queue is not empty
+ * then a new execution list can be submitted. The two requests at the front of
+ * the queue are next to be submitted but since a context may not occur twice in
+ * an execution list, if subsequent requests have the same ID as the first then
+ * the two requests must be combined. This is done simply by discarding requests
+ * at the head of the queue until either only one requests is left (in which case
+ * we use a NULL second context) or the first two requests have unique IDs.
+ *
+ * By always executing the first two requests in the queue the driver ensures
+ * that the GPU is kept as busy as possible. In the case where a single context
+ * completes but a second context is still executing, the request for this second
+ * context will be at the head of the queue when we remove the first one. This
+ * request will then be resubmitted along with a new request for a different context,
+ * which will cause the hardware to continue executing the second request and queue
+ * the new request (the GPU detects the condition of a context getting preempted
+ * with the same context and optimizes the context switch flow by not doing
+ * preemption, but just sampling the new tail pointer).
+ *
+ */
+
+#include <drm/drmP.h>
+#include <drm/i915_drm.h>
+#include "i915_drv.h"
+
+#define GEN8_LR_CONTEXT_RENDER_SIZE (20 * PAGE_SIZE)
+#define GEN8_LR_CONTEXT_OTHER_SIZE (2 * PAGE_SIZE)
+
+#define GEN8_LR_CONTEXT_ALIGN 4096
+
+#define RING_EXECLIST_QFULL		(1 << 0x2)
+#define RING_EXECLIST1_VALID		(1 << 0x3)
+#define RING_EXECLIST0_VALID		(1 << 0x4)
+#define RING_EXECLIST_ACTIVE_STATUS	(3 << 0xE)
+#define RING_EXECLIST1_ACTIVE		(1 << 0x11)
+#define RING_EXECLIST0_ACTIVE		(1 << 0x12)
+
+#define GEN8_CTX_STATUS_IDLE_ACTIVE	(1 << 0)
+#define GEN8_CTX_STATUS_PREEMPTED	(1 << 1)
+#define GEN8_CTX_STATUS_ELEMENT_SWITCH	(1 << 2)
+#define GEN8_CTX_STATUS_ACTIVE_IDLE	(1 << 3)
+#define GEN8_CTX_STATUS_COMPLETE	(1 << 4)
+#define GEN8_CTX_STATUS_LITE_RESTORE	(1 << 15)
+
+#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_GPGPU_CSR_BASE_ADDRESS	0x44
+
+#define GEN8_CTX_VALID (1<<0)
+#define GEN8_CTX_FORCE_PD_RESTORE (1<<1)
+#define GEN8_CTX_FORCE_RESTORE (1<<2)
+#define GEN8_CTX_L3LLC_COHERENT (1<<5)
+#define GEN8_CTX_PRIVILEGE (1<<8)
+enum {
+	ADVANCED_CONTEXT = 0,
+	LEGACY_CONTEXT,
+	ADVANCED_AD_CONTEXT,
+	LEGACY_64B_CONTEXT
+};
+#define GEN8_CTX_MODE_SHIFT 3
+enum {
+	FAULT_AND_HANG = 0,
+	FAULT_AND_HALT, /* Debug only */
+	FAULT_AND_STREAM,
+	FAULT_AND_CONTINUE /* Unsupported */
+};
+#define GEN8_CTX_ID_SHIFT 32
+
+/**
+ * intel_sanitize_enable_execlists() - sanitize i915.enable_execlists
+ * @dev: DRM device.
+ * @enable_execlists: value of i915.enable_execlists module parameter.
+ *
+ * Only certain platforms support Execlists (the prerequisites being
+ * support for Logical Ring Contexts and Aliasing PPGTT or better),
+ * and only when enabled via module parameter.
+ *
+ * Return: 1 if Execlists is supported and has to be enabled.
+ */
+int intel_sanitize_enable_execlists(struct drm_device *dev, int enable_execlists)
+{
+	WARN_ON(i915.enable_ppgtt == -1);
+
+	if (enable_execlists == 0)
+		return 0;
+
+	if (HAS_LOGICAL_RING_CONTEXTS(dev) && USES_PPGTT(dev) &&
+	    i915.use_mmio_flip >= 0)
+		return 1;
+
+	return 0;
+}
+
+/**
+ * intel_execlists_ctx_id() - get the Execlists Context ID
+ * @ctx_obj: Logical Ring Context backing object.
+ *
+ * Do not confuse with ctx->id! Unfortunately we have a name overload
+ * here: the old context ID we pass to userspace as a handler so that
+ * they can refer to a context, and the new context ID we pass to the
+ * ELSP so that the GPU can inform us of the context status via
+ * interrupts.
+ *
+ * Return: 20-bits globally unique context ID.
+ */
+u32 intel_execlists_ctx_id(struct drm_i915_gem_object *ctx_obj)
+{
+	u32 lrca = i915_gem_obj_ggtt_offset(ctx_obj);
+
+	/* LRCA is required to be 4K aligned so the more significant 20 bits
+	 * are globally unique */
+	return lrca >> 12;
+}
+
+static uint64_t execlists_ctx_descriptor(struct drm_i915_gem_object *ctx_obj)
+{
+	uint64_t desc;
+	uint64_t lrca = i915_gem_obj_ggtt_offset(ctx_obj);
+
+	WARN_ON(lrca & 0xFFFFFFFF00000FFFULL);
+
+	desc = GEN8_CTX_VALID;
+	desc |= LEGACY_CONTEXT << GEN8_CTX_MODE_SHIFT;
+	desc |= GEN8_CTX_L3LLC_COHERENT;
+	desc |= GEN8_CTX_PRIVILEGE;
+	desc |= lrca;
+	desc |= (u64)intel_execlists_ctx_id(ctx_obj) << GEN8_CTX_ID_SHIFT;
+
+	/* TODO: WaDisableLiteRestore when we start using semaphore
+	 * signalling between Command Streamers */
+	/* desc |= GEN8_CTX_FORCE_RESTORE; */
+
+	return desc;
+}
+
+static void execlists_elsp_write(struct intel_engine_cs *ring,
+				 struct drm_i915_gem_object *ctx_obj0,
+				 struct drm_i915_gem_object *ctx_obj1)
+{
+	struct drm_i915_private *dev_priv = ring->dev->dev_private;
+	uint64_t temp = 0;
+	uint32_t desc[4];
+	unsigned long flags;
+
+	/* XXX: You must always write both descriptors in the order below. */
+	if (ctx_obj1)
+		temp = execlists_ctx_descriptor(ctx_obj1);
+	else
+		temp = 0;
+	desc[1] = (u32)(temp >> 32);
+	desc[0] = (u32)temp;
+
+	temp = execlists_ctx_descriptor(ctx_obj0);
+	desc[3] = (u32)(temp >> 32);
+	desc[2] = (u32)temp;
+
+	/* Set Force Wakeup bit to prevent GT from entering C6 while ELSP writes
+	 * are in progress.
+	 *
+	 * The other problem is that we can't just call gen6_gt_force_wake_get()
+	 * because that function calls intel_runtime_pm_get(), which might sleep.
+	 * Instead, we do the runtime_pm_get/put when creating/destroying requests.
+	 */
+	spin_lock_irqsave(&dev_priv->uncore.lock, flags);
+	if (IS_CHERRYVIEW(dev_priv->dev)) {
+		if (dev_priv->uncore.fw_rendercount++ == 0)
+			dev_priv->uncore.funcs.force_wake_get(dev_priv,
+							      FORCEWAKE_RENDER);
+		if (dev_priv->uncore.fw_mediacount++ == 0)
+			dev_priv->uncore.funcs.force_wake_get(dev_priv,
+							      FORCEWAKE_MEDIA);
+	} else {
+		if (dev_priv->uncore.forcewake_count++ == 0)
+			dev_priv->uncore.funcs.force_wake_get(dev_priv,
+							      FORCEWAKE_ALL);
+	}
+	spin_unlock_irqrestore(&dev_priv->uncore.lock, flags);
+
+	I915_WRITE(RING_ELSP(ring), desc[1]);
+	I915_WRITE(RING_ELSP(ring), desc[0]);
+	I915_WRITE(RING_ELSP(ring), desc[3]);
+	/* The context is automatically loaded after the following */
+	I915_WRITE(RING_ELSP(ring), desc[2]);
+
+	/* ELSP is a wo register, so use another nearby reg for posting instead */
+	POSTING_READ(RING_EXECLIST_STATUS(ring));
+
+	/* Release Force Wakeup (see the big comment above). */
+	spin_lock_irqsave(&dev_priv->uncore.lock, flags);
+	if (IS_CHERRYVIEW(dev_priv->dev)) {
+		if (--dev_priv->uncore.fw_rendercount == 0)
+			dev_priv->uncore.funcs.force_wake_put(dev_priv,
+							      FORCEWAKE_RENDER);
+		if (--dev_priv->uncore.fw_mediacount == 0)
+			dev_priv->uncore.funcs.force_wake_put(dev_priv,
+							      FORCEWAKE_MEDIA);
+	} else {
+		if (--dev_priv->uncore.forcewake_count == 0)
+			dev_priv->uncore.funcs.force_wake_put(dev_priv,
+							      FORCEWAKE_ALL);
+	}
+
+	spin_unlock_irqrestore(&dev_priv->uncore.lock, flags);
+}
+
+static int execlists_ctx_write_tail(struct drm_i915_gem_object *ctx_obj, u32 tail)
+{
+	struct page *page;
+	uint32_t *reg_state;
+
+	page = i915_gem_object_get_page(ctx_obj, 1);
+	reg_state = kmap_atomic(page);
+
+	reg_state[CTX_RING_TAIL+1] = tail;
+
+	kunmap_atomic(reg_state);
+
+	return 0;
+}
+
+static int execlists_submit_context(struct intel_engine_cs *ring,
+				    struct intel_context *to0, u32 tail0,
+				    struct intel_context *to1, u32 tail1)
+{
+	struct drm_i915_gem_object *ctx_obj0;
+	struct drm_i915_gem_object *ctx_obj1 = NULL;
+
+	ctx_obj0 = to0->engine[ring->id].state;
+	BUG_ON(!ctx_obj0);
+	WARN_ON(!i915_gem_obj_is_pinned(ctx_obj0));
+
+	execlists_ctx_write_tail(ctx_obj0, tail0);
+
+	if (to1) {
+		ctx_obj1 = to1->engine[ring->id].state;
+		BUG_ON(!ctx_obj1);
+		WARN_ON(!i915_gem_obj_is_pinned(ctx_obj1));
+
+		execlists_ctx_write_tail(ctx_obj1, tail1);
+	}
+
+	execlists_elsp_write(ring, ctx_obj0, ctx_obj1);
+
+	return 0;
+}
+
+static void execlists_context_unqueue(struct intel_engine_cs *ring)
+{
+	struct intel_ctx_submit_request *req0 = NULL, *req1 = NULL;
+	struct intel_ctx_submit_request *cursor = NULL, *tmp = NULL;
+	struct drm_i915_private *dev_priv = ring->dev->dev_private;
+
+	assert_spin_locked(&ring->execlist_lock);
+
+	if (list_empty(&ring->execlist_queue))
+		return;
+
+	/* Try to read in pairs */
+	list_for_each_entry_safe(cursor, tmp, &ring->execlist_queue,
+				 execlist_link) {
+		if (!req0) {
+			req0 = cursor;
+		} else if (req0->ctx == cursor->ctx) {
+			/* Same ctx: ignore first request, as second request
+			 * will update tail past first request's workload */
+			cursor->elsp_submitted = req0->elsp_submitted;
+			list_del(&req0->execlist_link);
+			queue_work(dev_priv->wq, &req0->work);
+			req0 = cursor;
+		} else {
+			req1 = cursor;
+			break;
+		}
+	}
+
+	WARN_ON(req1 && req1->elsp_submitted);
+
+	WARN_ON(execlists_submit_context(ring, req0->ctx, req0->tail,
+					 req1 ? req1->ctx : NULL,
+					 req1 ? req1->tail : 0));
+
+	req0->elsp_submitted++;
+	if (req1)
+		req1->elsp_submitted++;
+}
+
+static bool execlists_check_remove_request(struct intel_engine_cs *ring,
+					   u32 request_id)
+{
+	struct drm_i915_private *dev_priv = ring->dev->dev_private;
+	struct intel_ctx_submit_request *head_req;
+
+	assert_spin_locked(&ring->execlist_lock);
+
+	head_req = list_first_entry_or_null(&ring->execlist_queue,
+					    struct intel_ctx_submit_request,
+					    execlist_link);
+
+	if (head_req != NULL) {
+		struct drm_i915_gem_object *ctx_obj =
+				head_req->ctx->engine[ring->id].state;
+		if (intel_execlists_ctx_id(ctx_obj) == request_id) {
+			WARN(head_req->elsp_submitted == 0,
+			     "Never submitted head request\n");
+
+			if (--head_req->elsp_submitted <= 0) {
+				list_del(&head_req->execlist_link);
+				queue_work(dev_priv->wq, &head_req->work);
+				return true;
+			}
+		}
+	}
+
+	return false;
+}
+
+/**
+ * intel_execlists_handle_ctx_events() - handle Context Switch interrupts
+ * @ring: Engine Command Streamer to handle.
+ *
+ * Check the unread Context Status Buffers and manage the submission of new
+ * contexts to the ELSP accordingly.
+ */
+void intel_execlists_handle_ctx_events(struct intel_engine_cs *ring)
+{
+	struct drm_i915_private *dev_priv = ring->dev->dev_private;
+	u32 status_pointer;
+	u8 read_pointer;
+	u8 write_pointer;
+	u32 status;
+	u32 status_id;
+	u32 submit_contexts = 0;
+
+	status_pointer = I915_READ(RING_CONTEXT_STATUS_PTR(ring));
+
+	read_pointer = ring->next_context_status_buffer;
+	write_pointer = status_pointer & 0x07;
+	if (read_pointer > write_pointer)
+		write_pointer += 6;
+
+	spin_lock(&ring->execlist_lock);
+
+	while (read_pointer < write_pointer) {
+		read_pointer++;
+		status = I915_READ(RING_CONTEXT_STATUS_BUF(ring) +
+				(read_pointer % 6) * 8);
+		status_id = I915_READ(RING_CONTEXT_STATUS_BUF(ring) +
+				(read_pointer % 6) * 8 + 4);
+
+		if (status & GEN8_CTX_STATUS_PREEMPTED) {
+			if (status & GEN8_CTX_STATUS_LITE_RESTORE) {
+				if (execlists_check_remove_request(ring, status_id))
+					WARN(1, "Lite Restored request removed from queue\n");
+			} else
+				WARN(1, "Preemption without Lite Restore\n");
+		}
+
+		 if ((status & GEN8_CTX_STATUS_ACTIVE_IDLE) ||
+		     (status & GEN8_CTX_STATUS_ELEMENT_SWITCH)) {
+			if (execlists_check_remove_request(ring, status_id))
+				submit_contexts++;
+		}
+	}
+
+	if (submit_contexts != 0)
+		execlists_context_unqueue(ring);
+
+	spin_unlock(&ring->execlist_lock);
+
+	WARN(submit_contexts > 2, "More than two context complete events?\n");
+	ring->next_context_status_buffer = write_pointer % 6;
+
+	I915_WRITE(RING_CONTEXT_STATUS_PTR(ring),
+		   ((u32)ring->next_context_status_buffer & 0x07) << 8);
+}
+
+static void execlists_free_request_task(struct work_struct *work)
+{
+	struct intel_ctx_submit_request *req =
+		container_of(work, struct intel_ctx_submit_request, work);
+	struct drm_device *dev = req->ring->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	intel_runtime_pm_put(dev_priv);
+
+	mutex_lock(&dev->struct_mutex);
+	i915_gem_context_unreference(req->ctx);
+	mutex_unlock(&dev->struct_mutex);
+
+	kfree(req);
+}
+
+static int execlists_context_queue(struct intel_engine_cs *ring,
+				   struct intel_context *to,
+				   u32 tail)
+{
+	struct intel_ctx_submit_request *req = NULL, *cursor;
+	struct drm_i915_private *dev_priv = ring->dev->dev_private;
+	unsigned long flags;
+	int num_elements = 0;
+
+	req = kzalloc(sizeof(*req), GFP_KERNEL);
+	if (req == NULL)
+		return -ENOMEM;
+	req->ctx = to;
+	i915_gem_context_reference(req->ctx);
+	req->ring = ring;
+	req->tail = tail;
+	INIT_WORK(&req->work, execlists_free_request_task);
+
+	intel_runtime_pm_get(dev_priv);
+
+	spin_lock_irqsave(&ring->execlist_lock, flags);
+
+	list_for_each_entry(cursor, &ring->execlist_queue, execlist_link)
+		if (++num_elements > 2)
+			break;
+
+	if (num_elements > 2) {
+		struct intel_ctx_submit_request *tail_req;
+
+		tail_req = list_last_entry(&ring->execlist_queue,
+					   struct intel_ctx_submit_request,
+					   execlist_link);
+
+		if (to == tail_req->ctx) {
+			WARN(tail_req->elsp_submitted != 0,
+			     "More than 2 already-submitted reqs queued\n");
+			list_del(&tail_req->execlist_link);
+			queue_work(dev_priv->wq, &tail_req->work);
+		}
+	}
+
+	list_add_tail(&req->execlist_link, &ring->execlist_queue);
+	if (num_elements == 0)
+		execlists_context_unqueue(ring);
+
+	spin_unlock_irqrestore(&ring->execlist_lock, flags);
+
+	return 0;
+}
+
+static int logical_ring_invalidate_all_caches(struct intel_ringbuffer *ringbuf)
+{
+	struct intel_engine_cs *ring = ringbuf->ring;
+	uint32_t flush_domains;
+	int ret;
+
+	flush_domains = 0;
+	if (ring->gpu_caches_dirty)
+		flush_domains = I915_GEM_GPU_DOMAINS;
+
+	ret = ring->emit_flush(ringbuf, I915_GEM_GPU_DOMAINS, flush_domains);
+	if (ret)
+		return ret;
+
+	ring->gpu_caches_dirty = false;
+	return 0;
+}
+
+static int execlists_move_to_gpu(struct intel_ringbuffer *ringbuf,
+				 struct list_head *vmas)
+{
+	struct intel_engine_cs *ring = ringbuf->ring;
+	struct i915_vma *vma;
+	uint32_t flush_domains = 0;
+	bool flush_chipset = false;
+	int ret;
+
+	list_for_each_entry(vma, vmas, exec_list) {
+		struct drm_i915_gem_object *obj = vma->obj;
+
+		ret = i915_gem_object_sync(obj, ring);
+		if (ret)
+			return ret;
+
+		if (obj->base.write_domain & I915_GEM_DOMAIN_CPU)
+			flush_chipset |= i915_gem_clflush_object(obj, false);
+
+		flush_domains |= obj->base.write_domain;
+	}
+
+	if (flush_domains & I915_GEM_DOMAIN_GTT)
+		wmb();
+
+	/* Unconditionally invalidate gpu caches and ensure that we do flush
+	 * any residual writes from the previous batch.
+	 */
+	return logical_ring_invalidate_all_caches(ringbuf);
+}
+
+/**
+ * execlists_submission() - submit a batchbuffer for execution, Execlists style
+ * @dev: DRM device.
+ * @file: DRM file.
+ * @ring: Engine Command Streamer to submit to.
+ * @ctx: Context to employ for this submission.
+ * @args: execbuffer call arguments.
+ * @vmas: list of vmas.
+ * @batch_obj: the batchbuffer to submit.
+ * @exec_start: batchbuffer start virtual address pointer.
+ * @flags: translated execbuffer call flags.
+ *
+ * This is the evil twin version of i915_gem_ringbuffer_submission. It abstracts
+ * away the submission details of the execbuffer ioctl call.
+ *
+ * Return: non-zero if the submission fails.
+ */
+int intel_execlists_submission(struct drm_device *dev, struct drm_file *file,
+			       struct intel_engine_cs *ring,
+			       struct intel_context *ctx,
+			       struct drm_i915_gem_execbuffer2 *args,
+			       struct list_head *vmas,
+			       struct drm_i915_gem_object *batch_obj,
+			       u64 exec_start, u32 flags)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_ringbuffer *ringbuf = ctx->engine[ring->id].ringbuf;
+	int instp_mode;
+	u32 instp_mask;
+	int ret;
+
+	instp_mode = args->flags & I915_EXEC_CONSTANTS_MASK;
+	instp_mask = I915_EXEC_CONSTANTS_MASK;
+	switch (instp_mode) {
+	case I915_EXEC_CONSTANTS_REL_GENERAL:
+	case I915_EXEC_CONSTANTS_ABSOLUTE:
+	case I915_EXEC_CONSTANTS_REL_SURFACE:
+		if (instp_mode != 0 && ring != &dev_priv->ring[RCS]) {
+			DRM_DEBUG("non-0 rel constants mode on non-RCS\n");
+			return -EINVAL;
+		}
+
+		if (instp_mode != dev_priv->relative_constants_mode) {
+			if (instp_mode == I915_EXEC_CONSTANTS_REL_SURFACE) {
+				DRM_DEBUG("rel surface constants mode invalid on gen5+\n");
+				return -EINVAL;
+			}
+
+			/* The HW changed the meaning on this bit on gen6 */
+			instp_mask &= ~I915_EXEC_CONSTANTS_REL_SURFACE;
+		}
+		break;
+	default:
+		DRM_DEBUG("execbuf with unknown constants: %d\n", instp_mode);
+		return -EINVAL;
+	}
+
+	if (args->num_cliprects != 0) {
+		DRM_DEBUG("clip rectangles are only valid on pre-gen5\n");
+		return -EINVAL;
+	} else {
+		if (args->DR4 == 0xffffffff) {
+			DRM_DEBUG("UXA submitting garbage DR4, fixing up\n");
+			args->DR4 = 0;
+		}
+
+		if (args->DR1 || args->DR4 || args->cliprects_ptr) {
+			DRM_DEBUG("0 cliprects but dirt in cliprects fields\n");
+			return -EINVAL;
+		}
+	}
+
+	if (args->flags & I915_EXEC_GEN7_SOL_RESET) {
+		DRM_DEBUG("sol reset is gen7 only\n");
+		return -EINVAL;
+	}
+
+	ret = execlists_move_to_gpu(ringbuf, vmas);
+	if (ret)
+		return ret;
+
+	if (ring == &dev_priv->ring[RCS] &&
+	    instp_mode != dev_priv->relative_constants_mode) {
+		ret = intel_logical_ring_begin(ringbuf, 4);
+		if (ret)
+			return ret;
+
+		intel_logical_ring_emit(ringbuf, MI_NOOP);
+		intel_logical_ring_emit(ringbuf, MI_LOAD_REGISTER_IMM(1));
+		intel_logical_ring_emit(ringbuf, INSTPM);
+		intel_logical_ring_emit(ringbuf, instp_mask << 16 | instp_mode);
+		intel_logical_ring_advance(ringbuf);
+
+		dev_priv->relative_constants_mode = instp_mode;
+	}
+
+	ret = ring->emit_bb_start(ringbuf, exec_start, flags);
+	if (ret)
+		return ret;
+
+	i915_gem_execbuffer_move_to_active(vmas, ring);
+	i915_gem_execbuffer_retire_commands(dev, file, ring, batch_obj);
+
+	return 0;
+}
+
+void intel_logical_ring_stop(struct intel_engine_cs *ring)
+{
+	struct drm_i915_private *dev_priv = ring->dev->dev_private;
+	int ret;
+
+	if (!intel_ring_initialized(ring))
+		return;
+
+	ret = intel_ring_idle(ring);
+	if (ret && !i915_reset_in_progress(&to_i915(ring->dev)->gpu_error))
+		DRM_ERROR("failed to quiesce %s whilst cleaning up: %d\n",
+			  ring->name, ret);
+
+	/* TODO: Is this correct with Execlists enabled? */
+	I915_WRITE_MODE(ring, _MASKED_BIT_ENABLE(STOP_RING));
+	if (wait_for_atomic((I915_READ_MODE(ring) & MODE_IDLE) != 0, 1000)) {
+		DRM_ERROR("%s :timed out trying to stop ring\n", ring->name);
+		return;
+	}
+	I915_WRITE_MODE(ring, _MASKED_BIT_DISABLE(STOP_RING));
+}
+
+int logical_ring_flush_all_caches(struct intel_ringbuffer *ringbuf)
+{
+	struct intel_engine_cs *ring = ringbuf->ring;
+	int ret;
+
+	if (!ring->gpu_caches_dirty)
+		return 0;
+
+	ret = ring->emit_flush(ringbuf, 0, I915_GEM_GPU_DOMAINS);
+	if (ret)
+		return ret;
+
+	ring->gpu_caches_dirty = false;
+	return 0;
+}
+
+/**
+ * intel_logical_ring_advance_and_submit() - advance the tail and submit the workload
+ * @ringbuf: Logical Ringbuffer to advance.
+ *
+ * The tail is updated in our logical ringbuffer struct, not in the actual context. What
+ * really happens during submission is that the context and current tail will be placed
+ * on a queue waiting for the ELSP to be ready to accept a new context submission. At that
+ * point, the tail *inside* the context is updated and the ELSP written to.
+ */
+void intel_logical_ring_advance_and_submit(struct intel_ringbuffer *ringbuf)
+{
+	struct intel_engine_cs *ring = ringbuf->ring;
+	struct intel_context *ctx = ringbuf->FIXME_lrc_ctx;
+
+	intel_logical_ring_advance(ringbuf);
+
+	if (intel_ring_stopped(ring))
+		return;
+
+	execlists_context_queue(ring, ctx, ringbuf->tail);
+}
+
+static int logical_ring_alloc_seqno(struct intel_engine_cs *ring,
+				    struct intel_context *ctx)
+{
+	if (ring->outstanding_lazy_seqno)
+		return 0;
+
+	if (ring->preallocated_lazy_request == NULL) {
+		struct drm_i915_gem_request *request;
+
+		request = kmalloc(sizeof(*request), GFP_KERNEL);
+		if (request == NULL)
+			return -ENOMEM;
+
+		/* Hold a reference to the context this request belongs to
+		 * (we will need it when the time comes to emit/retire the
+		 * request).
+		 */
+		request->ctx = ctx;
+		i915_gem_context_reference(request->ctx);
+
+		ring->preallocated_lazy_request = request;
+	}
+
+	return i915_gem_get_seqno(ring->dev, &ring->outstanding_lazy_seqno);
+}
+
+static int logical_ring_wait_request(struct intel_ringbuffer *ringbuf,
+				     int bytes)
+{
+	struct intel_engine_cs *ring = ringbuf->ring;
+	struct drm_i915_gem_request *request;
+	u32 seqno = 0;
+	int ret;
+
+	if (ringbuf->last_retired_head != -1) {
+		ringbuf->head = ringbuf->last_retired_head;
+		ringbuf->last_retired_head = -1;
+
+		ringbuf->space = intel_ring_space(ringbuf);
+		if (ringbuf->space >= bytes)
+			return 0;
+	}
+
+	list_for_each_entry(request, &ring->request_list, list) {
+		if (__intel_ring_space(request->tail, ringbuf->tail,
+				       ringbuf->size) >= bytes) {
+			seqno = request->seqno;
+			break;
+		}
+	}
+
+	if (seqno == 0)
+		return -ENOSPC;
+
+	ret = i915_wait_seqno(ring, seqno);
+	if (ret)
+		return ret;
+
+	i915_gem_retire_requests_ring(ring);
+	ringbuf->head = ringbuf->last_retired_head;
+	ringbuf->last_retired_head = -1;
+
+	ringbuf->space = intel_ring_space(ringbuf);
+	return 0;
+}
+
+static int logical_ring_wait_for_space(struct intel_ringbuffer *ringbuf,
+				       int bytes)
+{
+	struct intel_engine_cs *ring = ringbuf->ring;
+	struct drm_device *dev = ring->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	unsigned long end;
+	int ret;
+
+	ret = logical_ring_wait_request(ringbuf, bytes);
+	if (ret != -ENOSPC)
+		return ret;
+
+	/* Force the context submission in case we have been skipping it */
+	intel_logical_ring_advance_and_submit(ringbuf);
+
+	/* With GEM the hangcheck timer should kick us out of the loop,
+	 * leaving it early runs the risk of corrupting GEM state (due
+	 * to running on almost untested codepaths). But on resume
+	 * timers don't work yet, so prevent a complete hang in that
+	 * case by choosing an insanely large timeout. */
+	end = jiffies + 60 * HZ;
+
+	do {
+		ringbuf->head = I915_READ_HEAD(ring);
+		ringbuf->space = intel_ring_space(ringbuf);
+		if (ringbuf->space >= bytes) {
+			ret = 0;
+			break;
+		}
+
+		msleep(1);
+
+		if (dev_priv->mm.interruptible && signal_pending(current)) {
+			ret = -ERESTARTSYS;
+			break;
+		}
+
+		ret = i915_gem_check_wedge(&dev_priv->gpu_error,
+					   dev_priv->mm.interruptible);
+		if (ret)
+			break;
+
+		if (time_after(jiffies, end)) {
+			ret = -EBUSY;
+			break;
+		}
+	} while (1);
+
+	return ret;
+}
+
+static int logical_ring_wrap_buffer(struct intel_ringbuffer *ringbuf)
+{
+	uint32_t __iomem *virt;
+	int rem = ringbuf->size - ringbuf->tail;
+
+	if (ringbuf->space < rem) {
+		int ret = logical_ring_wait_for_space(ringbuf, rem);
+
+		if (ret)
+			return ret;
+	}
+
+	virt = ringbuf->virtual_start + ringbuf->tail;
+	rem /= 4;
+	while (rem--)
+		iowrite32(MI_NOOP, virt++);
+
+	ringbuf->tail = 0;
+	ringbuf->space = intel_ring_space(ringbuf);
+
+	return 0;
+}
+
+static int logical_ring_prepare(struct intel_ringbuffer *ringbuf, int bytes)
+{
+	int ret;
+
+	if (unlikely(ringbuf->tail + bytes > ringbuf->effective_size)) {
+		ret = logical_ring_wrap_buffer(ringbuf);
+		if (unlikely(ret))
+			return ret;
+	}
+
+	if (unlikely(ringbuf->space < bytes)) {
+		ret = logical_ring_wait_for_space(ringbuf, bytes);
+		if (unlikely(ret))
+			return ret;
+	}
+
+	return 0;
+}
+
+/**
+ * intel_logical_ring_begin() - prepare the logical ringbuffer to accept some commands
+ *
+ * @ringbuf: Logical ringbuffer.
+ * @num_dwords: number of DWORDs that we plan to write to the ringbuffer.
+ *
+ * The ringbuffer might not be ready to accept the commands right away (maybe it needs to
+ * be wrapped, or wait a bit for the tail to be updated). This function takes care of that
+ * and also preallocates a request (every workload submission is still mediated through
+ * requests, same as it did with legacy ringbuffer submission).
+ *
+ * Return: non-zero if the ringbuffer is not ready to be written to.
+ */
+int intel_logical_ring_begin(struct intel_ringbuffer *ringbuf, int num_dwords)
+{
+	struct intel_engine_cs *ring = ringbuf->ring;
+	struct drm_device *dev = ring->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	int ret;
+
+	ret = i915_gem_check_wedge(&dev_priv->gpu_error,
+				   dev_priv->mm.interruptible);
+	if (ret)
+		return ret;
+
+	ret = logical_ring_prepare(ringbuf, num_dwords * sizeof(uint32_t));
+	if (ret)
+		return ret;
+
+	/* Preallocate the olr before touching the ring */
+	ret = logical_ring_alloc_seqno(ring, ringbuf->FIXME_lrc_ctx);
+	if (ret)
+		return ret;
+
+	ringbuf->space -= num_dwords * sizeof(uint32_t);
+	return 0;
+}
+
+static int gen8_init_common_ring(struct intel_engine_cs *ring)
+{
+	struct drm_device *dev = ring->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	I915_WRITE_IMR(ring, ~(ring->irq_enable_mask | ring->irq_keep_mask));
+	I915_WRITE(RING_HWSTAM(ring->mmio_base), 0xffffffff);
+
+	I915_WRITE(RING_MODE_GEN7(ring),
+		   _MASKED_BIT_DISABLE(GFX_REPLAY_MODE) |
+		   _MASKED_BIT_ENABLE(GFX_RUN_LIST_ENABLE));
+	POSTING_READ(RING_MODE_GEN7(ring));
+	DRM_DEBUG_DRIVER("Execlists enabled for %s\n", ring->name);
+
+	memset(&ring->hangcheck, 0, sizeof(ring->hangcheck));
+
+	return 0;
+}
+
+static int gen8_init_render_ring(struct intel_engine_cs *ring)
+{
+	struct drm_device *dev = ring->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	int ret;
+
+	ret = gen8_init_common_ring(ring);
+	if (ret)
+		return ret;
+
+	/* We need to disable the AsyncFlip performance optimisations in order
+	 * to use MI_WAIT_FOR_EVENT within the CS. It should already be
+	 * programmed to '1' on all products.
+	 *
+	 * WaDisableAsyncFlipPerfMode:snb,ivb,hsw,vlv,bdw,chv
+	 */
+	I915_WRITE(MI_MODE, _MASKED_BIT_ENABLE(ASYNC_FLIP_PERF_DISABLE));
+
+	ret = intel_init_pipe_control(ring);
+	if (ret)
+		return ret;
+
+	I915_WRITE(INSTPM, _MASKED_BIT_ENABLE(INSTPM_FORCE_ORDERING));
+
+	return ret;
+}
+
+static int gen8_emit_bb_start(struct intel_ringbuffer *ringbuf,
+			      u64 offset, unsigned flags)
+{
+	bool ppgtt = !(flags & I915_DISPATCH_SECURE);
+	int ret;
+
+	ret = intel_logical_ring_begin(ringbuf, 4);
+	if (ret)
+		return ret;
+
+	/* FIXME(BDW): Address space and security selectors. */
+	intel_logical_ring_emit(ringbuf, MI_BATCH_BUFFER_START_GEN8 | (ppgtt<<8));
+	intel_logical_ring_emit(ringbuf, lower_32_bits(offset));
+	intel_logical_ring_emit(ringbuf, upper_32_bits(offset));
+	intel_logical_ring_emit(ringbuf, MI_NOOP);
+	intel_logical_ring_advance(ringbuf);
+
+	return 0;
+}
+
+static bool gen8_logical_ring_get_irq(struct intel_engine_cs *ring)
+{
+	struct drm_device *dev = ring->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	unsigned long flags;
+
+	if (!dev->irq_enabled)
+		return false;
+
+	spin_lock_irqsave(&dev_priv->irq_lock, flags);
+	if (ring->irq_refcount++ == 0) {
+		I915_WRITE_IMR(ring, ~(ring->irq_enable_mask | ring->irq_keep_mask));
+		POSTING_READ(RING_IMR(ring->mmio_base));
+	}
+	spin_unlock_irqrestore(&dev_priv->irq_lock, flags);
+
+	return true;
+}
+
+static void gen8_logical_ring_put_irq(struct intel_engine_cs *ring)
+{
+	struct drm_device *dev = ring->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	unsigned long flags;
+
+	spin_lock_irqsave(&dev_priv->irq_lock, flags);
+	if (--ring->irq_refcount == 0) {
+		I915_WRITE_IMR(ring, ~ring->irq_keep_mask);
+		POSTING_READ(RING_IMR(ring->mmio_base));
+	}
+	spin_unlock_irqrestore(&dev_priv->irq_lock, flags);
+}
+
+static int gen8_emit_flush(struct intel_ringbuffer *ringbuf,
+			   u32 invalidate_domains,
+			   u32 unused)
+{
+	struct intel_engine_cs *ring = ringbuf->ring;
+	struct drm_device *dev = ring->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	uint32_t cmd;
+	int ret;
+
+	ret = intel_logical_ring_begin(ringbuf, 4);
+	if (ret)
+		return ret;
+
+	cmd = MI_FLUSH_DW + 1;
+
+	if (ring == &dev_priv->ring[VCS]) {
+		if (invalidate_domains & I915_GEM_GPU_DOMAINS)
+			cmd |= MI_INVALIDATE_TLB | MI_INVALIDATE_BSD |
+				MI_FLUSH_DW_STORE_INDEX |
+				MI_FLUSH_DW_OP_STOREDW;
+	} else {
+		if (invalidate_domains & I915_GEM_DOMAIN_RENDER)
+			cmd |= MI_INVALIDATE_TLB | MI_FLUSH_DW_STORE_INDEX |
+				MI_FLUSH_DW_OP_STOREDW;
+	}
+
+	intel_logical_ring_emit(ringbuf, cmd);
+	intel_logical_ring_emit(ringbuf,
+				I915_GEM_HWS_SCRATCH_ADDR |
+				MI_FLUSH_DW_USE_GTT);
+	intel_logical_ring_emit(ringbuf, 0); /* upper addr */
+	intel_logical_ring_emit(ringbuf, 0); /* value */
+	intel_logical_ring_advance(ringbuf);
+
+	return 0;
+}
+
+static int gen8_emit_flush_render(struct intel_ringbuffer *ringbuf,
+				  u32 invalidate_domains,
+				  u32 flush_domains)
+{
+	struct intel_engine_cs *ring = ringbuf->ring;
+	u32 scratch_addr = ring->scratch.gtt_offset + 2 * CACHELINE_BYTES;
+	u32 flags = 0;
+	int ret;
+
+	flags |= PIPE_CONTROL_CS_STALL;
+
+	if (flush_domains) {
+		flags |= PIPE_CONTROL_RENDER_TARGET_CACHE_FLUSH;
+		flags |= PIPE_CONTROL_DEPTH_CACHE_FLUSH;
+	}
+
+	if (invalidate_domains) {
+		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_QW_WRITE;
+		flags |= PIPE_CONTROL_GLOBAL_GTT_IVB;
+	}
+
+	ret = intel_logical_ring_begin(ringbuf, 6);
+	if (ret)
+		return ret;
+
+	intel_logical_ring_emit(ringbuf, GFX_OP_PIPE_CONTROL(6));
+	intel_logical_ring_emit(ringbuf, flags);
+	intel_logical_ring_emit(ringbuf, scratch_addr);
+	intel_logical_ring_emit(ringbuf, 0);
+	intel_logical_ring_emit(ringbuf, 0);
+	intel_logical_ring_emit(ringbuf, 0);
+	intel_logical_ring_advance(ringbuf);
+
+	return 0;
+}
+
+static u32 gen8_get_seqno(struct intel_engine_cs *ring, bool lazy_coherency)
+{
+	return intel_read_status_page(ring, I915_GEM_HWS_INDEX);
+}
+
+static void gen8_set_seqno(struct intel_engine_cs *ring, u32 seqno)
+{
+	intel_write_status_page(ring, I915_GEM_HWS_INDEX, seqno);
+}
+
+static int gen8_emit_request(struct intel_ringbuffer *ringbuf)
+{
+	struct intel_engine_cs *ring = ringbuf->ring;
+	u32 cmd;
+	int ret;
+
+	ret = intel_logical_ring_begin(ringbuf, 6);
+	if (ret)
+		return ret;
+
+	cmd = MI_STORE_DWORD_IMM_GEN8;
+	cmd |= MI_GLOBAL_GTT;
+
+	intel_logical_ring_emit(ringbuf, cmd);
+	intel_logical_ring_emit(ringbuf,
+				(ring->status_page.gfx_addr +
+				(I915_GEM_HWS_INDEX << MI_STORE_DWORD_INDEX_SHIFT)));
+	intel_logical_ring_emit(ringbuf, 0);
+	intel_logical_ring_emit(ringbuf, ring->outstanding_lazy_seqno);
+	intel_logical_ring_emit(ringbuf, MI_USER_INTERRUPT);
+	intel_logical_ring_emit(ringbuf, MI_NOOP);
+	intel_logical_ring_advance_and_submit(ringbuf);
+
+	return 0;
+}
+
+/**
+ * intel_logical_ring_cleanup() - deallocate the Engine Command Streamer
+ *
+ * @ring: Engine Command Streamer.
+ *
+ */
+void intel_logical_ring_cleanup(struct intel_engine_cs *ring)
+{
+	struct drm_i915_private *dev_priv = ring->dev->dev_private;
+
+	if (!intel_ring_initialized(ring))
+		return;
+
+	intel_logical_ring_stop(ring);
+	WARN_ON((I915_READ_MODE(ring) & MODE_IDLE) == 0);
+	ring->preallocated_lazy_request = NULL;
+	ring->outstanding_lazy_seqno = 0;
+
+	if (ring->cleanup)
+		ring->cleanup(ring);
+
+	i915_cmd_parser_fini_ring(ring);
+
+	if (ring->status_page.obj) {
+		kunmap(sg_page(ring->status_page.obj->pages->sgl));
+		ring->status_page.obj = NULL;
+	}
+}
+
+static int logical_ring_init(struct drm_device *dev, struct intel_engine_cs *ring)
+{
+	int ret;
+
+	/* Intentionally left blank. */
+	ring->buffer = NULL;
+
+	ring->dev = dev;
+	INIT_LIST_HEAD(&ring->active_list);
+	INIT_LIST_HEAD(&ring->request_list);
+	init_waitqueue_head(&ring->irq_queue);
+
+	INIT_LIST_HEAD(&ring->execlist_queue);
+	spin_lock_init(&ring->execlist_lock);
+	ring->next_context_status_buffer = 0;
+
+	ret = i915_cmd_parser_init_ring(ring);
+	if (ret)
+		return ret;
+
+	if (ring->init) {
+		ret = ring->init(ring);
+		if (ret)
+			return ret;
+	}
+
+	ret = intel_lr_context_deferred_create(ring->default_context, ring);
+
+	return ret;
+}
+
+static int logical_render_ring_init(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring = &dev_priv->ring[RCS];
+
+	ring->name = "render ring";
+	ring->id = RCS;
+	ring->mmio_base = RENDER_RING_BASE;
+	ring->irq_enable_mask =
+		GT_RENDER_USER_INTERRUPT << GEN8_RCS_IRQ_SHIFT;
+	ring->irq_keep_mask =
+		GT_CONTEXT_SWITCH_INTERRUPT << GEN8_RCS_IRQ_SHIFT;
+	if (HAS_L3_DPF(dev))
+		ring->irq_keep_mask |= GT_RENDER_L3_PARITY_ERROR_INTERRUPT;
+
+	ring->init = gen8_init_render_ring;
+	ring->cleanup = intel_fini_pipe_control;
+	ring->get_seqno = gen8_get_seqno;
+	ring->set_seqno = gen8_set_seqno;
+	ring->emit_request = gen8_emit_request;
+	ring->emit_flush = gen8_emit_flush_render;
+	ring->irq_get = gen8_logical_ring_get_irq;
+	ring->irq_put = gen8_logical_ring_put_irq;
+	ring->emit_bb_start = gen8_emit_bb_start;
+
+	return logical_ring_init(dev, ring);
+}
+
+static int logical_bsd_ring_init(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring = &dev_priv->ring[VCS];
+
+	ring->name = "bsd ring";
+	ring->id = VCS;
+	ring->mmio_base = GEN6_BSD_RING_BASE;
+	ring->irq_enable_mask =
+		GT_RENDER_USER_INTERRUPT << GEN8_VCS1_IRQ_SHIFT;
+	ring->irq_keep_mask =
+		GT_CONTEXT_SWITCH_INTERRUPT << GEN8_VCS1_IRQ_SHIFT;
+
+	ring->init = gen8_init_common_ring;
+	ring->get_seqno = gen8_get_seqno;
+	ring->set_seqno = gen8_set_seqno;
+	ring->emit_request = gen8_emit_request;
+	ring->emit_flush = gen8_emit_flush;
+	ring->irq_get = gen8_logical_ring_get_irq;
+	ring->irq_put = gen8_logical_ring_put_irq;
+	ring->emit_bb_start = gen8_emit_bb_start;
+
+	return logical_ring_init(dev, ring);
+}
+
+static int logical_bsd2_ring_init(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring = &dev_priv->ring[VCS2];
+
+	ring->name = "bds2 ring";
+	ring->id = VCS2;
+	ring->mmio_base = GEN8_BSD2_RING_BASE;
+	ring->irq_enable_mask =
+		GT_RENDER_USER_INTERRUPT << GEN8_VCS2_IRQ_SHIFT;
+	ring->irq_keep_mask =
+		GT_CONTEXT_SWITCH_INTERRUPT << GEN8_VCS2_IRQ_SHIFT;
+
+	ring->init = gen8_init_common_ring;
+	ring->get_seqno = gen8_get_seqno;
+	ring->set_seqno = gen8_set_seqno;
+	ring->emit_request = gen8_emit_request;
+	ring->emit_flush = gen8_emit_flush;
+	ring->irq_get = gen8_logical_ring_get_irq;
+	ring->irq_put = gen8_logical_ring_put_irq;
+	ring->emit_bb_start = gen8_emit_bb_start;
+
+	return logical_ring_init(dev, ring);
+}
+
+static int logical_blt_ring_init(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring = &dev_priv->ring[BCS];
+
+	ring->name = "blitter ring";
+	ring->id = BCS;
+	ring->mmio_base = BLT_RING_BASE;
+	ring->irq_enable_mask =
+		GT_RENDER_USER_INTERRUPT << GEN8_BCS_IRQ_SHIFT;
+	ring->irq_keep_mask =
+		GT_CONTEXT_SWITCH_INTERRUPT << GEN8_BCS_IRQ_SHIFT;
+
+	ring->init = gen8_init_common_ring;
+	ring->get_seqno = gen8_get_seqno;
+	ring->set_seqno = gen8_set_seqno;
+	ring->emit_request = gen8_emit_request;
+	ring->emit_flush = gen8_emit_flush;
+	ring->irq_get = gen8_logical_ring_get_irq;
+	ring->irq_put = gen8_logical_ring_put_irq;
+	ring->emit_bb_start = gen8_emit_bb_start;
+
+	return logical_ring_init(dev, ring);
+}
+
+static int logical_vebox_ring_init(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring = &dev_priv->ring[VECS];
+
+	ring->name = "video enhancement ring";
+	ring->id = VECS;
+	ring->mmio_base = VEBOX_RING_BASE;
+	ring->irq_enable_mask =
+		GT_RENDER_USER_INTERRUPT << GEN8_VECS_IRQ_SHIFT;
+	ring->irq_keep_mask =
+		GT_CONTEXT_SWITCH_INTERRUPT << GEN8_VECS_IRQ_SHIFT;
+
+	ring->init = gen8_init_common_ring;
+	ring->get_seqno = gen8_get_seqno;
+	ring->set_seqno = gen8_set_seqno;
+	ring->emit_request = gen8_emit_request;
+	ring->emit_flush = gen8_emit_flush;
+	ring->irq_get = gen8_logical_ring_get_irq;
+	ring->irq_put = gen8_logical_ring_put_irq;
+	ring->emit_bb_start = gen8_emit_bb_start;
+
+	return logical_ring_init(dev, ring);
+}
+
+/**
+ * intel_logical_rings_init() - allocate, populate and init the Engine Command Streamers
+ * @dev: DRM device.
+ *
+ * This function inits the engines for an Execlists submission style (the equivalent in the
+ * legacy ringbuffer submission world would be i915_gem_init_rings). It does it only for
+ * those engines that are present in the hardware.
+ *
+ * Return: non-zero if the initialization failed.
+ */
+int intel_logical_rings_init(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	int ret;
+
+	ret = logical_render_ring_init(dev);
+	if (ret)
+		return ret;
+
+	if (HAS_BSD(dev)) {
+		ret = logical_bsd_ring_init(dev);
+		if (ret)
+			goto cleanup_render_ring;
+	}
+
+	if (HAS_BLT(dev)) {
+		ret = logical_blt_ring_init(dev);
+		if (ret)
+			goto cleanup_bsd_ring;
+	}
+
+	if (HAS_VEBOX(dev)) {
+		ret = logical_vebox_ring_init(dev);
+		if (ret)
+			goto cleanup_blt_ring;
+	}
+
+	if (HAS_BSD2(dev)) {
+		ret = logical_bsd2_ring_init(dev);
+		if (ret)
+			goto cleanup_vebox_ring;
+	}
+
+	ret = i915_gem_set_seqno(dev, ((u32)~0 - 0x1000));
+	if (ret)
+		goto cleanup_bsd2_ring;
+
+	return 0;
+
+cleanup_bsd2_ring:
+	intel_logical_ring_cleanup(&dev_priv->ring[VCS2]);
+cleanup_vebox_ring:
+	intel_logical_ring_cleanup(&dev_priv->ring[VECS]);
+cleanup_blt_ring:
+	intel_logical_ring_cleanup(&dev_priv->ring[BCS]);
+cleanup_bsd_ring:
+	intel_logical_ring_cleanup(&dev_priv->ring[VCS]);
+cleanup_render_ring:
+	intel_logical_ring_cleanup(&dev_priv->ring[RCS]);
+
+	return ret;
+}
+
+int intel_lr_context_render_state_init(struct intel_engine_cs *ring,
+				       struct intel_context *ctx)
+{
+	struct intel_ringbuffer *ringbuf = ctx->engine[ring->id].ringbuf;
+	struct render_state so;
+	struct drm_i915_file_private *file_priv = ctx->file_priv;
+	struct drm_file *file = file_priv ? file_priv->file : NULL;
+	int ret;
+
+	ret = i915_gem_render_state_prepare(ring, &so);
+	if (ret)
+		return ret;
+
+	if (so.rodata == NULL)
+		return 0;
+
+	ret = ring->emit_bb_start(ringbuf,
+			so.ggtt_offset,
+			I915_DISPATCH_SECURE);
+	if (ret)
+		goto out;
+
+	i915_vma_move_to_active(i915_gem_obj_to_ggtt(so.obj), ring);
+
+	ret = __i915_add_request(ring, file, so.obj, NULL);
+	/* intel_logical_ring_add_request moves object to inactive if it
+	 * fails */
+out:
+	i915_gem_render_state_fini(&so);
+	return ret;
+}
+
+static int
+populate_lr_context(struct intel_context *ctx, struct drm_i915_gem_object *ctx_obj,
+		    struct intel_engine_cs *ring, struct intel_ringbuffer *ringbuf)
+{
+	struct drm_device *dev = ring->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct drm_i915_gem_object *ring_obj = ringbuf->obj;
+	struct i915_hw_ppgtt *ppgtt = ctx->ppgtt;
+	struct page *page;
+	uint32_t *reg_state;
+	int ret;
+
+	if (!ppgtt)
+		ppgtt = dev_priv->mm.aliasing_ppgtt;
+
+	ret = i915_gem_object_set_to_cpu_domain(ctx_obj, true);
+	if (ret) {
+		DRM_DEBUG_DRIVER("Could not set to CPU domain\n");
+		return ret;
+	}
+
+	ret = i915_gem_object_get_pages(ctx_obj);
+	if (ret) {
+		DRM_DEBUG_DRIVER("Could not get object pages\n");
+		return ret;
+	}
+
+	i915_gem_object_pin_pages(ctx_obj);
+
+	/* The second page of the context object contains some fields which must
+	 * be set up prior to the first execution. */
+	page = i915_gem_object_get_page(ctx_obj, 1);
+	reg_state = kmap_atomic(page);
+
+	/* 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). */
+	if (ring->id == RCS)
+		reg_state[CTX_LRI_HEADER_0] = MI_LOAD_REGISTER_IMM(14);
+	else
+		reg_state[CTX_LRI_HEADER_0] = MI_LOAD_REGISTER_IMM(11);
+	reg_state[CTX_LRI_HEADER_0] |= MI_LRI_FORCE_POSTED;
+	reg_state[CTX_CONTEXT_CONTROL] = RING_CONTEXT_CONTROL(ring);
+	reg_state[CTX_CONTEXT_CONTROL+1] =
+			_MASKED_BIT_ENABLE((1<<3) | MI_RESTORE_INHIBIT);
+	reg_state[CTX_RING_HEAD] = RING_HEAD(ring->mmio_base);
+	reg_state[CTX_RING_HEAD+1] = 0;
+	reg_state[CTX_RING_TAIL] = RING_TAIL(ring->mmio_base);
+	reg_state[CTX_RING_TAIL+1] = 0;
+	reg_state[CTX_RING_BUFFER_START] = RING_START(ring->mmio_base);
+	reg_state[CTX_RING_BUFFER_START+1] = i915_gem_obj_ggtt_offset(ring_obj);
+	reg_state[CTX_RING_BUFFER_CONTROL] = RING_CTL(ring->mmio_base);
+	reg_state[CTX_RING_BUFFER_CONTROL+1] =
+			((ringbuf->size - PAGE_SIZE) & RING_NR_PAGES) | RING_VALID;
+	reg_state[CTX_BB_HEAD_U] = ring->mmio_base + 0x168;
+	reg_state[CTX_BB_HEAD_U+1] = 0;
+	reg_state[CTX_BB_HEAD_L] = ring->mmio_base + 0x140;
+	reg_state[CTX_BB_HEAD_L+1] = 0;
+	reg_state[CTX_BB_STATE] = ring->mmio_base + 0x110;
+	reg_state[CTX_BB_STATE+1] = (1<<5);
+	reg_state[CTX_SECOND_BB_HEAD_U] = ring->mmio_base + 0x11c;
+	reg_state[CTX_SECOND_BB_HEAD_U+1] = 0;
+	reg_state[CTX_SECOND_BB_HEAD_L] = ring->mmio_base + 0x114;
+	reg_state[CTX_SECOND_BB_HEAD_L+1] = 0;
+	reg_state[CTX_SECOND_BB_STATE] = ring->mmio_base + 0x118;
+	reg_state[CTX_SECOND_BB_STATE+1] = 0;
+	if (ring->id == RCS) {
+		/* TODO: according to BSpec, the register state context
+		 * for CHV does not have these. OTOH, these registers do
+		 * exist in CHV. I'm waiting for a clarification */
+		reg_state[CTX_BB_PER_CTX_PTR] = ring->mmio_base + 0x1c0;
+		reg_state[CTX_BB_PER_CTX_PTR+1] = 0;
+		reg_state[CTX_RCS_INDIRECT_CTX] = ring->mmio_base + 0x1c4;
+		reg_state[CTX_RCS_INDIRECT_CTX+1] = 0;
+		reg_state[CTX_RCS_INDIRECT_CTX_OFFSET] = ring->mmio_base + 0x1c8;
+		reg_state[CTX_RCS_INDIRECT_CTX_OFFSET+1] = 0;
+	}
+	reg_state[CTX_LRI_HEADER_1] = MI_LOAD_REGISTER_IMM(9);
+	reg_state[CTX_LRI_HEADER_1] |= MI_LRI_FORCE_POSTED;
+	reg_state[CTX_CTX_TIMESTAMP] = ring->mmio_base + 0x3a8;
+	reg_state[CTX_CTX_TIMESTAMP+1] = 0;
+	reg_state[CTX_PDP3_UDW] = GEN8_RING_PDP_UDW(ring, 3);
+	reg_state[CTX_PDP3_LDW] = GEN8_RING_PDP_LDW(ring, 3);
+	reg_state[CTX_PDP2_UDW] = GEN8_RING_PDP_UDW(ring, 2);
+	reg_state[CTX_PDP2_LDW] = GEN8_RING_PDP_LDW(ring, 2);
+	reg_state[CTX_PDP1_UDW] = GEN8_RING_PDP_UDW(ring, 1);
+	reg_state[CTX_PDP1_LDW] = GEN8_RING_PDP_LDW(ring, 1);
+	reg_state[CTX_PDP0_UDW] = GEN8_RING_PDP_UDW(ring, 0);
+	reg_state[CTX_PDP0_LDW] = GEN8_RING_PDP_LDW(ring, 0);
+	reg_state[CTX_PDP3_UDW+1] = upper_32_bits(ppgtt->pd_dma_addr[3]);
+	reg_state[CTX_PDP3_LDW+1] = lower_32_bits(ppgtt->pd_dma_addr[3]);
+	reg_state[CTX_PDP2_UDW+1] = upper_32_bits(ppgtt->pd_dma_addr[2]);
+	reg_state[CTX_PDP2_LDW+1] = lower_32_bits(ppgtt->pd_dma_addr[2]);
+	reg_state[CTX_PDP1_UDW+1] = upper_32_bits(ppgtt->pd_dma_addr[1]);
+	reg_state[CTX_PDP1_LDW+1] = lower_32_bits(ppgtt->pd_dma_addr[1]);
+	reg_state[CTX_PDP0_UDW+1] = upper_32_bits(ppgtt->pd_dma_addr[0]);
+	reg_state[CTX_PDP0_LDW+1] = lower_32_bits(ppgtt->pd_dma_addr[0]);
+	if (ring->id == RCS) {
+		reg_state[CTX_LRI_HEADER_2] = MI_LOAD_REGISTER_IMM(1);
+		reg_state[CTX_R_PWR_CLK_STATE] = 0x20c8;
+		reg_state[CTX_R_PWR_CLK_STATE+1] = 0;
+	}
+
+	kunmap_atomic(reg_state);
+
+	ctx_obj->dirty = 1;
+	set_page_dirty(page);
+	i915_gem_object_unpin_pages(ctx_obj);
+
+	return 0;
+}
+
+/**
+ * intel_lr_context_free() - free the LRC specific bits of a context
+ * @ctx: the LR context to free.
+ *
+ * The real context freeing is done in i915_gem_context_free: this only
+ * takes care of the bits that are LRC related: the per-engine backing
+ * objects and the logical ringbuffer.
+ */
+void intel_lr_context_free(struct intel_context *ctx)
+{
+	int i;
+
+	for (i = 0; i < I915_NUM_RINGS; i++) {
+		struct drm_i915_gem_object *ctx_obj = ctx->engine[i].state;
+		struct intel_ringbuffer *ringbuf = ctx->engine[i].ringbuf;
+
+		if (ctx_obj) {
+			intel_destroy_ringbuffer_obj(ringbuf);
+			kfree(ringbuf);
+			i915_gem_object_ggtt_unpin(ctx_obj);
+			drm_gem_object_unreference(&ctx_obj->base);
+		}
+	}
+}
+
+static uint32_t get_lr_context_size(struct intel_engine_cs *ring)
+{
+	int ret = 0;
+
+	WARN_ON(INTEL_INFO(ring->dev)->gen != 8);
+
+	switch (ring->id) {
+	case RCS:
+		ret = GEN8_LR_CONTEXT_RENDER_SIZE;
+		break;
+	case VCS:
+	case BCS:
+	case VECS:
+	case VCS2:
+		ret = GEN8_LR_CONTEXT_OTHER_SIZE;
+		break;
+	}
+
+	return ret;
+}
+
+/**
+ * intel_lr_context_deferred_create() - create the LRC specific bits of a context
+ * @ctx: LR context to create.
+ * @ring: engine to be used with the context.
+ *
+ * This function can be called more than once, with different engines, if we plan
+ * to use the context with them. The context backing objects and the ringbuffers
+ * (specially the ringbuffer backing objects) suck a lot of memory up, and that's why
+ * the creation is a deferred call: it's better to make sure first that we need to use
+ * a given ring with the context.
+ *
+ * Return: non-zero on eror.
+ */
+int intel_lr_context_deferred_create(struct intel_context *ctx,
+				     struct intel_engine_cs *ring)
+{
+	struct drm_device *dev = ring->dev;
+	struct drm_i915_gem_object *ctx_obj;
+	uint32_t context_size;
+	struct intel_ringbuffer *ringbuf;
+	int ret;
+
+	WARN_ON(ctx->legacy_hw_ctx.rcs_state != NULL);
+	if (ctx->engine[ring->id].state)
+		return 0;
+
+	context_size = round_up(get_lr_context_size(ring), 4096);
+
+	ctx_obj = i915_gem_alloc_context_obj(dev, context_size);
+	if (IS_ERR(ctx_obj)) {
+		ret = PTR_ERR(ctx_obj);
+		DRM_DEBUG_DRIVER("Alloc LRC backing obj failed: %d\n", ret);
+		return ret;
+	}
+
+	ret = i915_gem_obj_ggtt_pin(ctx_obj, GEN8_LR_CONTEXT_ALIGN, 0);
+	if (ret) {
+		DRM_DEBUG_DRIVER("Pin LRC backing obj failed: %d\n", ret);
+		drm_gem_object_unreference(&ctx_obj->base);
+		return ret;
+	}
+
+	ringbuf = kzalloc(sizeof(*ringbuf), GFP_KERNEL);
+	if (!ringbuf) {
+		DRM_DEBUG_DRIVER("Failed to allocate ringbuffer %s\n",
+				ring->name);
+		i915_gem_object_ggtt_unpin(ctx_obj);
+		drm_gem_object_unreference(&ctx_obj->base);
+		ret = -ENOMEM;
+		return ret;
+	}
+
+	ringbuf->ring = ring;
+	ringbuf->FIXME_lrc_ctx = ctx;
+
+	ringbuf->size = 32 * PAGE_SIZE;
+	ringbuf->effective_size = ringbuf->size;
+	ringbuf->head = 0;
+	ringbuf->tail = 0;
+	ringbuf->space = ringbuf->size;
+	ringbuf->last_retired_head = -1;
+
+	/* TODO: For now we put this in the mappable region so that we can reuse
+	 * the existing ringbuffer code which ioremaps it. When we start
+	 * creating many contexts, this will no longer work and we must switch
+	 * to a kmapish interface.
+	 */
+	ret = intel_alloc_ringbuffer_obj(dev, ringbuf);
+	if (ret) {
+		DRM_DEBUG_DRIVER("Failed to allocate ringbuffer obj %s: %d\n",
+				ring->name, ret);
+		goto error;
+	}
+
+	ret = populate_lr_context(ctx, ctx_obj, ring, ringbuf);
+	if (ret) {
+		DRM_DEBUG_DRIVER("Failed to populate LRC: %d\n", ret);
+		intel_destroy_ringbuffer_obj(ringbuf);
+		goto error;
+	}
+
+	ctx->engine[ring->id].ringbuf = ringbuf;
+	ctx->engine[ring->id].state = ctx_obj;
+
+	if (ctx == ring->default_context) {
+		/* The status page is offset 0 from the default context object
+		 * in LRC mode. */
+		ring->status_page.gfx_addr = i915_gem_obj_ggtt_offset(ctx_obj);
+		ring->status_page.page_addr =
+				kmap(sg_page(ctx_obj->pages->sgl));
+		if (ring->status_page.page_addr == NULL)
+			return -ENOMEM;
+		ring->status_page.obj = ctx_obj;
+	}
+
+	if (ring->id == RCS && !ctx->rcs_initialized) {
+		ret = intel_lr_context_render_state_init(ring, ctx);
+		if (ret) {
+			DRM_ERROR("Init render state failed: %d\n", ret);
+			ctx->engine[ring->id].ringbuf = NULL;
+			ctx->engine[ring->id].state = NULL;
+			intel_destroy_ringbuffer_obj(ringbuf);
+			goto error;
+		}
+		ctx->rcs_initialized = true;
+	}
+
+	return 0;
+
+error:
+	kfree(ringbuf);
+	i915_gem_object_ggtt_unpin(ctx_obj);
+	drm_gem_object_unreference(&ctx_obj->base);
+	return ret;
+}
diff --git a/drivers/gpu/drm/i915/intel_lrc.h b/drivers/gpu/drm/i915/intel_lrc.h
new file mode 100644
index 000000000000..33c3b4bf28c5
--- /dev/null
+++ b/drivers/gpu/drm/i915/intel_lrc.h
@@ -0,0 +1,114 @@
+/*
+ * Copyright © 2014 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.
+ */
+
+#ifndef _INTEL_LRC_H_
+#define _INTEL_LRC_H_
+
+/* Execlists regs */
+#define RING_ELSP(ring)			((ring)->mmio_base+0x230)
+#define RING_EXECLIST_STATUS(ring)	((ring)->mmio_base+0x234)
+#define RING_CONTEXT_CONTROL(ring)	((ring)->mmio_base+0x244)
+#define RING_CONTEXT_STATUS_BUF(ring)	((ring)->mmio_base+0x370)
+#define RING_CONTEXT_STATUS_PTR(ring)	((ring)->mmio_base+0x3a0)
+
+/* Logical Rings */
+void intel_logical_ring_stop(struct intel_engine_cs *ring);
+void intel_logical_ring_cleanup(struct intel_engine_cs *ring);
+int intel_logical_rings_init(struct drm_device *dev);
+
+int logical_ring_flush_all_caches(struct intel_ringbuffer *ringbuf);
+void intel_logical_ring_advance_and_submit(struct intel_ringbuffer *ringbuf);
+/**
+ * intel_logical_ring_advance() - advance the ringbuffer tail
+ * @ringbuf: Ringbuffer to advance.
+ *
+ * The tail is only updated in our logical ringbuffer struct.
+ */
+static inline void intel_logical_ring_advance(struct intel_ringbuffer *ringbuf)
+{
+	ringbuf->tail &= ringbuf->size - 1;
+}
+/**
+ * intel_logical_ring_emit() - write a DWORD to the ringbuffer.
+ * @ringbuf: Ringbuffer to write to.
+ * @data: DWORD to write.
+ */
+static inline void intel_logical_ring_emit(struct intel_ringbuffer *ringbuf,
+					   u32 data)
+{
+	iowrite32(data, ringbuf->virtual_start + ringbuf->tail);
+	ringbuf->tail += 4;
+}
+int intel_logical_ring_begin(struct intel_ringbuffer *ringbuf, int num_dwords);
+
+/* Logical Ring Contexts */
+int intel_lr_context_render_state_init(struct intel_engine_cs *ring,
+				       struct intel_context *ctx);
+void intel_lr_context_free(struct intel_context *ctx);
+int intel_lr_context_deferred_create(struct intel_context *ctx,
+				     struct intel_engine_cs *ring);
+
+/* Execlists */
+int intel_sanitize_enable_execlists(struct drm_device *dev, int enable_execlists);
+int intel_execlists_submission(struct drm_device *dev, struct drm_file *file,
+			       struct intel_engine_cs *ring,
+			       struct intel_context *ctx,
+			       struct drm_i915_gem_execbuffer2 *args,
+			       struct list_head *vmas,
+			       struct drm_i915_gem_object *batch_obj,
+			       u64 exec_start, u32 flags);
+u32 intel_execlists_ctx_id(struct drm_i915_gem_object *ctx_obj);
+
+/**
+ * struct intel_ctx_submit_request - queued context submission request
+ * @ctx: Context to submit to the ELSP.
+ * @ring: Engine to submit it to.
+ * @tail: how far in the context's ringbuffer this request goes to.
+ * @execlist_link: link in the submission queue.
+ * @work: workqueue for processing this request in a bottom half.
+ * @elsp_submitted: no. of times this request has been sent to the ELSP.
+ *
+ * The ELSP only accepts two elements at a time, so we queue context/tail
+ * pairs on a given queue (ring->execlist_queue) until the hardware is
+ * available. The queue serves a double purpose: we also use it to keep track
+ * of the up to 2 contexts currently in the hardware (usually one in execution
+ * and the other queued up by the GPU): We only remove elements from the head
+ * of the queue when the hardware informs us that an element has been
+ * completed.
+ *
+ * All accesses to the queue are mediated by a spinlock (ring->execlist_lock).
+ */
+struct intel_ctx_submit_request {
+	struct intel_context *ctx;
+	struct intel_engine_cs *ring;
+	u32 tail;
+
+	struct list_head execlist_link;
+	struct work_struct work;
+
+	int elsp_submitted;
+};
+
+void intel_execlists_handle_ctx_events(struct intel_engine_cs *ring);
+
+#endif /* _INTEL_LRC_H_ */
diff --git a/drivers/gpu/drm/i915/intel_lvds.c b/drivers/gpu/drm/i915/intel_lvds.c
index fdf40267249c..a6bd1422e38f 100644
--- a/drivers/gpu/drm/i915/intel_lvds.c
+++ b/drivers/gpu/drm/i915/intel_lvds.c
@@ -823,8 +823,7 @@ bool intel_is_dual_link_lvds(struct drm_device *dev)
 	struct intel_encoder *encoder;
 	struct intel_lvds_encoder *lvds_encoder;
 
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list,
-			    base.head) {
+	for_each_intel_encoder(dev, encoder) {
 		if (encoder->type == INTEL_OUTPUT_LVDS) {
 			lvds_encoder = to_lvds_encoder(&encoder->base);
 
diff --git a/drivers/gpu/drm/i915/intel_panel.c b/drivers/gpu/drm/i915/intel_panel.c
index 8e374449c6b5..18784470a760 100644
--- a/drivers/gpu/drm/i915/intel_panel.c
+++ b/drivers/gpu/drm/i915/intel_panel.c
@@ -751,6 +751,8 @@ void intel_panel_disable_backlight(struct intel_connector *connector)
 
 	spin_lock_irqsave(&dev_priv->backlight_lock, flags);
 
+	if (panel->backlight.device)
+		panel->backlight.device->props.power = FB_BLANK_POWERDOWN;
 	panel->backlight.enabled = false;
 	dev_priv->display.disable_backlight(connector);
 
@@ -957,6 +959,8 @@ void intel_panel_enable_backlight(struct intel_connector *connector)
 
 	dev_priv->display.enable_backlight(connector);
 	panel->backlight.enabled = true;
+	if (panel->backlight.device)
+		panel->backlight.device->props.power = FB_BLANK_UNBLANK;
 
 	spin_unlock_irqrestore(&dev_priv->backlight_lock, flags);
 }
@@ -965,6 +969,7 @@ void intel_panel_enable_backlight(struct intel_connector *connector)
 static int intel_backlight_device_update_status(struct backlight_device *bd)
 {
 	struct intel_connector *connector = bl_get_data(bd);
+	struct intel_panel *panel = &connector->panel;
 	struct drm_device *dev = connector->base.dev;
 
 	drm_modeset_lock(&dev->mode_config.connection_mutex, NULL);
@@ -972,6 +977,23 @@ static int intel_backlight_device_update_status(struct backlight_device *bd)
 		      bd->props.brightness, bd->props.max_brightness);
 	intel_panel_set_backlight(connector, bd->props.brightness,
 				  bd->props.max_brightness);
+
+	/*
+	 * Allow flipping bl_power as a sub-state of enabled. Sadly the
+	 * backlight class device does not make it easy to to differentiate
+	 * between callbacks for brightness and bl_power, so our backlight_power
+	 * callback needs to take this into account.
+	 */
+	if (panel->backlight.enabled) {
+		if (panel->backlight_power) {
+			bool enable = bd->props.power == FB_BLANK_UNBLANK &&
+				bd->props.brightness != 0;
+			panel->backlight_power(connector, enable);
+		}
+	} else {
+		bd->props.power = FB_BLANK_POWERDOWN;
+	}
+
 	drm_modeset_unlock(&dev->mode_config.connection_mutex);
 	return 0;
 }
@@ -1023,6 +1045,11 @@ static int intel_backlight_device_register(struct intel_connector *connector)
 					    panel->backlight.level,
 					    props.max_brightness);
 
+	if (panel->backlight.enabled)
+		props.power = FB_BLANK_UNBLANK;
+	else
+		props.power = FB_BLANK_POWERDOWN;
+
 	/*
 	 * Note: using the same name independent of the connector prevents
 	 * registration of multiple backlight devices in the driver.
@@ -1203,7 +1230,7 @@ static int vlv_setup_backlight(struct intel_connector *connector)
 	enum pipe pipe;
 	u32 ctl, ctl2, val;
 
-	for_each_pipe(pipe) {
+	for_each_pipe(dev_priv, pipe) {
 		u32 cur_val = I915_READ(VLV_BLC_PWM_CTL(pipe));
 
 		/* Skip if the modulation freq is already set */
diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c
index 40c12295c0bd..c27b6140bfd1 100644
--- a/drivers/gpu/drm/i915/intel_pm.c
+++ b/drivers/gpu/drm/i915/intel_pm.c
@@ -309,6 +309,9 @@ static void gen7_enable_fbc(struct drm_crtc *crtc)
 
 	dpfc_ctl |= IVB_DPFC_CTL_FENCE_EN;
 
+	if (dev_priv->fbc.false_color)
+		dpfc_ctl |= FBC_CTL_FALSE_COLOR;
+
 	I915_WRITE(ILK_DPFC_CONTROL, dpfc_ctl | DPFC_CTL_EN);
 
 	if (IS_IVYBRIDGE(dev)) {
@@ -342,6 +345,16 @@ bool intel_fbc_enabled(struct drm_device *dev)
 	return dev_priv->display.fbc_enabled(dev);
 }
 
+void gen8_fbc_sw_flush(struct drm_device *dev, u32 value)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	if (!IS_GEN8(dev))
+		return;
+
+	I915_WRITE(MSG_FBC_REND_STATE, value);
+}
+
 static void intel_fbc_work_fn(struct work_struct *__work)
 {
 	struct intel_fbc_work *work =
@@ -578,6 +591,12 @@ void intel_update_fbc(struct drm_device *dev)
 			DRM_DEBUG_KMS("framebuffer not tiled or fenced, disabling compression\n");
 		goto out_disable;
 	}
+	if (INTEL_INFO(dev)->gen <= 4 && !IS_G4X(dev) &&
+	    to_intel_plane(crtc->primary)->rotation != BIT(DRM_ROTATE_0)) {
+		if (set_no_fbc_reason(dev_priv, FBC_UNSUPPORTED_MODE))
+			DRM_DEBUG_KMS("Rotation unsupported, disabling\n");
+		goto out_disable;
+	}
 
 	/* If the kernel debugger is active, always disable compression */
 	if (in_dbg_master())
@@ -853,7 +872,7 @@ void intel_set_memory_cxsr(struct drm_i915_private *dev_priv, bool enable)
  * A value of 5us seems to be a good balance; safe for very low end
  * platforms but not overly aggressive on lower latency configs.
  */
-static const int latency_ns = 5000;
+static const int pessimal_latency_ns = 5000;
 
 static int i9xx_get_fifo_size(struct drm_device *dev, int plane)
 {
@@ -982,13 +1001,20 @@ static const struct intel_watermark_params i915_wm_info = {
 	.guard_size = 2,
 	.cacheline_size = I915_FIFO_LINE_SIZE,
 };
-static const struct intel_watermark_params i830_wm_info = {
+static const struct intel_watermark_params i830_a_wm_info = {
 	.fifo_size = I855GM_FIFO_SIZE,
 	.max_wm = I915_MAX_WM,
 	.default_wm = 1,
 	.guard_size = 2,
 	.cacheline_size = I830_FIFO_LINE_SIZE,
 };
+static const struct intel_watermark_params i830_bc_wm_info = {
+	.fifo_size = I855GM_FIFO_SIZE,
+	.max_wm = I915_MAX_WM/2,
+	.default_wm = 1,
+	.guard_size = 2,
+	.cacheline_size = I830_FIFO_LINE_SIZE,
+};
 static const struct intel_watermark_params i845_wm_info = {
 	.fifo_size = I830_FIFO_SIZE,
 	.max_wm = I915_MAX_WM,
@@ -1044,6 +1070,17 @@ static unsigned long intel_calculate_wm(unsigned long clock_in_khz,
 		wm_size = wm->max_wm;
 	if (wm_size <= 0)
 		wm_size = wm->default_wm;
+
+	/*
+	 * Bspec seems to indicate that the value shouldn't be lower than
+	 * 'burst size + 1'. Certainly 830 is quite unhappy with low values.
+	 * Lets go for 8 which is the burst size since certain platforms
+	 * already use a hardcoded 8 (which is what the spec says should be
+	 * done).
+	 */
+	if (wm_size <= 8)
+		wm_size = 8;
+
 	return wm_size;
 }
 
@@ -1268,33 +1305,27 @@ static bool g4x_compute_srwm(struct drm_device *dev,
 			      display, cursor);
 }
 
-static bool vlv_compute_drain_latency(struct drm_device *dev,
-				     int plane,
-				     int *plane_prec_mult,
-				     int *plane_dl,
-				     int *cursor_prec_mult,
-				     int *cursor_dl)
+static bool vlv_compute_drain_latency(struct drm_crtc *crtc,
+				      int pixel_size,
+				      int *prec_mult,
+				      int *drain_latency)
 {
-	struct drm_crtc *crtc;
-	int clock, pixel_size;
 	int entries;
+	int clock = to_intel_crtc(crtc)->config.adjusted_mode.crtc_clock;
 
-	crtc = intel_get_crtc_for_plane(dev, plane);
-	if (!intel_crtc_active(crtc))
+	if (WARN(clock == 0, "Pixel clock is zero!\n"))
 		return false;
 
-	clock = to_intel_crtc(crtc)->config.adjusted_mode.crtc_clock;
-	pixel_size = crtc->primary->fb->bits_per_pixel / 8;	/* BPP */
+	if (WARN(pixel_size == 0, "Pixel size is zero!\n"))
+		return false;
 
-	entries = (clock / 1000) * pixel_size;
-	*plane_prec_mult = (entries > 128) ?
-		DRAIN_LATENCY_PRECISION_64 : DRAIN_LATENCY_PRECISION_32;
-	*plane_dl = (64 * (*plane_prec_mult) * 4) / entries;
+	entries = DIV_ROUND_UP(clock, 1000) * pixel_size;
+	*prec_mult = (entries > 128) ? DRAIN_LATENCY_PRECISION_64 :
+				       DRAIN_LATENCY_PRECISION_32;
+	*drain_latency = (64 * (*prec_mult) * 4) / entries;
 
-	entries = (clock / 1000) * 4;	/* BPP is always 4 for cursor */
-	*cursor_prec_mult = (entries > 128) ?
-		DRAIN_LATENCY_PRECISION_64 : DRAIN_LATENCY_PRECISION_32;
-	*cursor_dl = (64 * (*cursor_prec_mult) * 4) / entries;
+	if (*drain_latency > DRAIN_LATENCY_MASK)
+		*drain_latency = DRAIN_LATENCY_MASK;
 
 	return true;
 }
@@ -1307,39 +1338,48 @@ static bool vlv_compute_drain_latency(struct drm_device *dev,
  * latency value.
  */
 
-static void vlv_update_drain_latency(struct drm_device *dev)
+static void vlv_update_drain_latency(struct drm_crtc *crtc)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	int planea_prec, planea_dl, planeb_prec, planeb_dl;
-	int cursora_prec, cursora_dl, cursorb_prec, cursorb_dl;
-	int plane_prec_mult, cursor_prec_mult; /* Precision multiplier is
-							either 16 or 32 */
+	struct drm_i915_private *dev_priv = crtc->dev->dev_private;
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+	int pixel_size;
+	int drain_latency;
+	enum pipe pipe = intel_crtc->pipe;
+	int plane_prec, prec_mult, plane_dl;
+
+	plane_dl = I915_READ(VLV_DDL(pipe)) & ~(DDL_PLANE_PRECISION_64 |
+		   DRAIN_LATENCY_MASK | DDL_CURSOR_PRECISION_64 |
+		   (DRAIN_LATENCY_MASK << DDL_CURSOR_SHIFT));
 
-	/* For plane A, Cursor A */
-	if (vlv_compute_drain_latency(dev, 0, &plane_prec_mult, &planea_dl,
-				      &cursor_prec_mult, &cursora_dl)) {
-		cursora_prec = (cursor_prec_mult == DRAIN_LATENCY_PRECISION_32) ?
-			DDL_CURSORA_PRECISION_32 : DDL_CURSORA_PRECISION_64;
-		planea_prec = (plane_prec_mult == DRAIN_LATENCY_PRECISION_32) ?
-			DDL_PLANEA_PRECISION_32 : DDL_PLANEA_PRECISION_64;
+	if (!intel_crtc_active(crtc)) {
+		I915_WRITE(VLV_DDL(pipe), plane_dl);
+		return;
+	}
 
-		I915_WRITE(VLV_DDL1, cursora_prec |
-				(cursora_dl << DDL_CURSORA_SHIFT) |
-				planea_prec | planea_dl);
+	/* Primary plane Drain Latency */
+	pixel_size = crtc->primary->fb->bits_per_pixel / 8;	/* BPP */
+	if (vlv_compute_drain_latency(crtc, pixel_size, &prec_mult, &drain_latency)) {
+		plane_prec = (prec_mult == DRAIN_LATENCY_PRECISION_64) ?
+					   DDL_PLANE_PRECISION_64 :
+					   DDL_PLANE_PRECISION_32;
+		plane_dl |= plane_prec | drain_latency;
 	}
 
-	/* For plane B, Cursor B */
-	if (vlv_compute_drain_latency(dev, 1, &plane_prec_mult, &planeb_dl,
-				      &cursor_prec_mult, &cursorb_dl)) {
-		cursorb_prec = (cursor_prec_mult == DRAIN_LATENCY_PRECISION_32) ?
-			DDL_CURSORB_PRECISION_32 : DDL_CURSORB_PRECISION_64;
-		planeb_prec = (plane_prec_mult == DRAIN_LATENCY_PRECISION_32) ?
-			DDL_PLANEB_PRECISION_32 : DDL_PLANEB_PRECISION_64;
+	/* Cursor Drain Latency
+	 * BPP is always 4 for cursor
+	 */
+	pixel_size = 4;
 
-		I915_WRITE(VLV_DDL2, cursorb_prec |
-				(cursorb_dl << DDL_CURSORB_SHIFT) |
-				planeb_prec | planeb_dl);
+	/* Program cursor DL only if it is enabled */
+	if (intel_crtc->cursor_base &&
+	    vlv_compute_drain_latency(crtc, pixel_size, &prec_mult, &drain_latency)) {
+		plane_prec = (prec_mult == DRAIN_LATENCY_PRECISION_64) ?
+					   DDL_CURSOR_PRECISION_64 :
+					   DDL_CURSOR_PRECISION_32;
+		plane_dl |= plane_prec | (drain_latency << DDL_CURSOR_SHIFT);
 	}
+
+	I915_WRITE(VLV_DDL(pipe), plane_dl);
 }
 
 #define single_plane_enabled(mask) is_power_of_2(mask)
@@ -1355,20 +1395,92 @@ static void valleyview_update_wm(struct drm_crtc *crtc)
 	unsigned int enabled = 0;
 	bool cxsr_enabled;
 
-	vlv_update_drain_latency(dev);
+	vlv_update_drain_latency(crtc);
+
+	if (g4x_compute_wm0(dev, PIPE_A,
+			    &valleyview_wm_info, pessimal_latency_ns,
+			    &valleyview_cursor_wm_info, pessimal_latency_ns,
+			    &planea_wm, &cursora_wm))
+		enabled |= 1 << PIPE_A;
+
+	if (g4x_compute_wm0(dev, PIPE_B,
+			    &valleyview_wm_info, pessimal_latency_ns,
+			    &valleyview_cursor_wm_info, pessimal_latency_ns,
+			    &planeb_wm, &cursorb_wm))
+		enabled |= 1 << PIPE_B;
+
+	if (single_plane_enabled(enabled) &&
+	    g4x_compute_srwm(dev, ffs(enabled) - 1,
+			     sr_latency_ns,
+			     &valleyview_wm_info,
+			     &valleyview_cursor_wm_info,
+			     &plane_sr, &ignore_cursor_sr) &&
+	    g4x_compute_srwm(dev, ffs(enabled) - 1,
+			     2*sr_latency_ns,
+			     &valleyview_wm_info,
+			     &valleyview_cursor_wm_info,
+			     &ignore_plane_sr, &cursor_sr)) {
+		cxsr_enabled = true;
+	} else {
+		cxsr_enabled = false;
+		intel_set_memory_cxsr(dev_priv, false);
+		plane_sr = cursor_sr = 0;
+	}
+
+	DRM_DEBUG_KMS("Setting FIFO watermarks - A: plane=%d, cursor=%d, "
+		      "B: plane=%d, cursor=%d, SR: plane=%d, cursor=%d\n",
+		      planea_wm, cursora_wm,
+		      planeb_wm, cursorb_wm,
+		      plane_sr, cursor_sr);
+
+	I915_WRITE(DSPFW1,
+		   (plane_sr << DSPFW_SR_SHIFT) |
+		   (cursorb_wm << DSPFW_CURSORB_SHIFT) |
+		   (planeb_wm << DSPFW_PLANEB_SHIFT) |
+		   (planea_wm << DSPFW_PLANEA_SHIFT));
+	I915_WRITE(DSPFW2,
+		   (I915_READ(DSPFW2) & ~DSPFW_CURSORA_MASK) |
+		   (cursora_wm << DSPFW_CURSORA_SHIFT));
+	I915_WRITE(DSPFW3,
+		   (I915_READ(DSPFW3) & ~DSPFW_CURSOR_SR_MASK) |
+		   (cursor_sr << DSPFW_CURSOR_SR_SHIFT));
+
+	if (cxsr_enabled)
+		intel_set_memory_cxsr(dev_priv, true);
+}
+
+static void cherryview_update_wm(struct drm_crtc *crtc)
+{
+	struct drm_device *dev = crtc->dev;
+	static const int sr_latency_ns = 12000;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	int planea_wm, planeb_wm, planec_wm;
+	int cursora_wm, cursorb_wm, cursorc_wm;
+	int plane_sr, cursor_sr;
+	int ignore_plane_sr, ignore_cursor_sr;
+	unsigned int enabled = 0;
+	bool cxsr_enabled;
+
+	vlv_update_drain_latency(crtc);
 
 	if (g4x_compute_wm0(dev, PIPE_A,
-			    &valleyview_wm_info, latency_ns,
-			    &valleyview_cursor_wm_info, latency_ns,
+			    &valleyview_wm_info, pessimal_latency_ns,
+			    &valleyview_cursor_wm_info, pessimal_latency_ns,
 			    &planea_wm, &cursora_wm))
 		enabled |= 1 << PIPE_A;
 
 	if (g4x_compute_wm0(dev, PIPE_B,
-			    &valleyview_wm_info, latency_ns,
-			    &valleyview_cursor_wm_info, latency_ns,
+			    &valleyview_wm_info, pessimal_latency_ns,
+			    &valleyview_cursor_wm_info, pessimal_latency_ns,
 			    &planeb_wm, &cursorb_wm))
 		enabled |= 1 << PIPE_B;
 
+	if (g4x_compute_wm0(dev, PIPE_C,
+			    &valleyview_wm_info, pessimal_latency_ns,
+			    &valleyview_cursor_wm_info, pessimal_latency_ns,
+			    &planec_wm, &cursorc_wm))
+		enabled |= 1 << PIPE_C;
+
 	if (single_plane_enabled(enabled) &&
 	    g4x_compute_srwm(dev, ffs(enabled) - 1,
 			     sr_latency_ns,
@@ -1387,27 +1499,66 @@ static void valleyview_update_wm(struct drm_crtc *crtc)
 		plane_sr = cursor_sr = 0;
 	}
 
-	DRM_DEBUG_KMS("Setting FIFO watermarks - A: plane=%d, cursor=%d, B: plane=%d, cursor=%d, SR: plane=%d, cursor=%d\n",
+	DRM_DEBUG_KMS("Setting FIFO watermarks - A: plane=%d, cursor=%d, "
+		      "B: plane=%d, cursor=%d, C: plane=%d, cursor=%d, "
+		      "SR: plane=%d, cursor=%d\n",
 		      planea_wm, cursora_wm,
 		      planeb_wm, cursorb_wm,
+		      planec_wm, cursorc_wm,
 		      plane_sr, cursor_sr);
 
 	I915_WRITE(DSPFW1,
 		   (plane_sr << DSPFW_SR_SHIFT) |
 		   (cursorb_wm << DSPFW_CURSORB_SHIFT) |
 		   (planeb_wm << DSPFW_PLANEB_SHIFT) |
-		   planea_wm);
+		   (planea_wm << DSPFW_PLANEA_SHIFT));
 	I915_WRITE(DSPFW2,
 		   (I915_READ(DSPFW2) & ~DSPFW_CURSORA_MASK) |
 		   (cursora_wm << DSPFW_CURSORA_SHIFT));
 	I915_WRITE(DSPFW3,
 		   (I915_READ(DSPFW3) & ~DSPFW_CURSOR_SR_MASK) |
 		   (cursor_sr << DSPFW_CURSOR_SR_SHIFT));
+	I915_WRITE(DSPFW9_CHV,
+		   (I915_READ(DSPFW9_CHV) & ~(DSPFW_PLANEC_MASK |
+					      DSPFW_CURSORC_MASK)) |
+		   (planec_wm << DSPFW_PLANEC_SHIFT) |
+		   (cursorc_wm << DSPFW_CURSORC_SHIFT));
 
 	if (cxsr_enabled)
 		intel_set_memory_cxsr(dev_priv, true);
 }
 
+static void valleyview_update_sprite_wm(struct drm_plane *plane,
+					struct drm_crtc *crtc,
+					uint32_t sprite_width,
+					uint32_t sprite_height,
+					int pixel_size,
+					bool enabled, bool scaled)
+{
+	struct drm_device *dev = crtc->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	int pipe = to_intel_plane(plane)->pipe;
+	int sprite = to_intel_plane(plane)->plane;
+	int drain_latency;
+	int plane_prec;
+	int sprite_dl;
+	int prec_mult;
+
+	sprite_dl = I915_READ(VLV_DDL(pipe)) & ~(DDL_SPRITE_PRECISION_64(sprite) |
+		    (DRAIN_LATENCY_MASK << DDL_SPRITE_SHIFT(sprite)));
+
+	if (enabled && vlv_compute_drain_latency(crtc, pixel_size, &prec_mult,
+						 &drain_latency)) {
+		plane_prec = (prec_mult == DRAIN_LATENCY_PRECISION_64) ?
+					   DDL_SPRITE_PRECISION_64(sprite) :
+					   DDL_SPRITE_PRECISION_32(sprite);
+		sprite_dl |= plane_prec |
+			     (drain_latency << DDL_SPRITE_SHIFT(sprite));
+	}
+
+	I915_WRITE(VLV_DDL(pipe), sprite_dl);
+}
+
 static void g4x_update_wm(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
@@ -1419,14 +1570,14 @@ static void g4x_update_wm(struct drm_crtc *crtc)
 	bool cxsr_enabled;
 
 	if (g4x_compute_wm0(dev, PIPE_A,
-			    &g4x_wm_info, latency_ns,
-			    &g4x_cursor_wm_info, latency_ns,
+			    &g4x_wm_info, pessimal_latency_ns,
+			    &g4x_cursor_wm_info, pessimal_latency_ns,
 			    &planea_wm, &cursora_wm))
 		enabled |= 1 << PIPE_A;
 
 	if (g4x_compute_wm0(dev, PIPE_B,
-			    &g4x_wm_info, latency_ns,
-			    &g4x_cursor_wm_info, latency_ns,
+			    &g4x_wm_info, pessimal_latency_ns,
+			    &g4x_cursor_wm_info, pessimal_latency_ns,
 			    &planeb_wm, &cursorb_wm))
 		enabled |= 1 << PIPE_B;
 
@@ -1443,7 +1594,8 @@ static void g4x_update_wm(struct drm_crtc *crtc)
 		plane_sr = cursor_sr = 0;
 	}
 
-	DRM_DEBUG_KMS("Setting FIFO watermarks - A: plane=%d, cursor=%d, B: plane=%d, cursor=%d, SR: plane=%d, cursor=%d\n",
+	DRM_DEBUG_KMS("Setting FIFO watermarks - A: plane=%d, cursor=%d, "
+		      "B: plane=%d, cursor=%d, SR: plane=%d, cursor=%d\n",
 		      planea_wm, cursora_wm,
 		      planeb_wm, cursorb_wm,
 		      plane_sr, cursor_sr);
@@ -1452,7 +1604,7 @@ static void g4x_update_wm(struct drm_crtc *crtc)
 		   (plane_sr << DSPFW_SR_SHIFT) |
 		   (cursorb_wm << DSPFW_CURSORB_SHIFT) |
 		   (planeb_wm << DSPFW_PLANEB_SHIFT) |
-		   planea_wm);
+		   (planea_wm << DSPFW_PLANEA_SHIFT));
 	I915_WRITE(DSPFW2,
 		   (I915_READ(DSPFW2) & ~DSPFW_CURSORA_MASK) |
 		   (cursora_wm << DSPFW_CURSORA_SHIFT));
@@ -1526,8 +1678,11 @@ static void i965_update_wm(struct drm_crtc *unused_crtc)
 
 	/* 965 has limitations... */
 	I915_WRITE(DSPFW1, (srwm << DSPFW_SR_SHIFT) |
-		   (8 << 16) | (8 << 8) | (8 << 0));
-	I915_WRITE(DSPFW2, (8 << 8) | (8 << 0));
+		   (8 << DSPFW_CURSORB_SHIFT) |
+		   (8 << DSPFW_PLANEB_SHIFT) |
+		   (8 << DSPFW_PLANEA_SHIFT));
+	I915_WRITE(DSPFW2, (8 << DSPFW_CURSORA_SHIFT) |
+		   (8 << DSPFW_PLANEC_SHIFT_OLD));
 	/* update cursor SR watermark */
 	I915_WRITE(DSPFW3, (cursor_sr << DSPFW_CURSOR_SR_SHIFT));
 
@@ -1552,7 +1707,7 @@ static void i9xx_update_wm(struct drm_crtc *unused_crtc)
 	else if (!IS_GEN2(dev))
 		wm_info = &i915_wm_info;
 	else
-		wm_info = &i830_wm_info;
+		wm_info = &i830_a_wm_info;
 
 	fifo_size = dev_priv->display.get_fifo_size(dev, 0);
 	crtc = intel_get_crtc_for_plane(dev, 0);
@@ -1565,10 +1720,16 @@ static void i9xx_update_wm(struct drm_crtc *unused_crtc)
 		adjusted_mode = &to_intel_crtc(crtc)->config.adjusted_mode;
 		planea_wm = intel_calculate_wm(adjusted_mode->crtc_clock,
 					       wm_info, fifo_size, cpp,
-					       latency_ns);
+					       pessimal_latency_ns);
 		enabled = crtc;
-	} else
+	} else {
 		planea_wm = fifo_size - wm_info->guard_size;
+		if (planea_wm > (long)wm_info->max_wm)
+			planea_wm = wm_info->max_wm;
+	}
+
+	if (IS_GEN2(dev))
+		wm_info = &i830_bc_wm_info;
 
 	fifo_size = dev_priv->display.get_fifo_size(dev, 1);
 	crtc = intel_get_crtc_for_plane(dev, 1);
@@ -1581,13 +1742,16 @@ static void i9xx_update_wm(struct drm_crtc *unused_crtc)
 		adjusted_mode = &to_intel_crtc(crtc)->config.adjusted_mode;
 		planeb_wm = intel_calculate_wm(adjusted_mode->crtc_clock,
 					       wm_info, fifo_size, cpp,
-					       latency_ns);
+					       pessimal_latency_ns);
 		if (enabled == NULL)
 			enabled = crtc;
 		else
 			enabled = NULL;
-	} else
+	} else {
 		planeb_wm = fifo_size - wm_info->guard_size;
+		if (planeb_wm > (long)wm_info->max_wm)
+			planeb_wm = wm_info->max_wm;
+	}
 
 	DRM_DEBUG_KMS("FIFO watermarks - A: %d, B: %d\n", planea_wm, planeb_wm);
 
@@ -1674,7 +1838,7 @@ static void i845_update_wm(struct drm_crtc *unused_crtc)
 	planea_wm = intel_calculate_wm(adjusted_mode->crtc_clock,
 				       &i845_wm_info,
 				       dev_priv->display.get_fifo_size(dev, 0),
-				       4, latency_ns);
+				       4, pessimal_latency_ns);
 	fwater_lo = I915_READ(FW_BLC) & ~0xfff;
 	fwater_lo |= (3<<8) | planea_wm;
 
@@ -2527,7 +2691,7 @@ static struct intel_pipe_wm *ilk_find_best_result(struct drm_device *dev,
 #define WM_DIRTY_FBC (1 << 24)
 #define WM_DIRTY_DDB (1 << 25)
 
-static unsigned int ilk_compute_wm_dirty(struct drm_device *dev,
+static unsigned int ilk_compute_wm_dirty(struct drm_i915_private *dev_priv,
 					 const struct ilk_wm_values *old,
 					 const struct ilk_wm_values *new)
 {
@@ -2535,7 +2699,7 @@ static unsigned int ilk_compute_wm_dirty(struct drm_device *dev,
 	enum pipe pipe;
 	int wm_lp;
 
-	for_each_pipe(pipe) {
+	for_each_pipe(dev_priv, pipe) {
 		if (old->wm_linetime[pipe] != new->wm_linetime[pipe]) {
 			dirty |= WM_DIRTY_LINETIME(pipe);
 			/* Must disable LP1+ watermarks too */
@@ -2621,7 +2785,7 @@ static void ilk_write_wm_values(struct drm_i915_private *dev_priv,
 	unsigned int dirty;
 	uint32_t val;
 
-	dirty = ilk_compute_wm_dirty(dev, previous, results);
+	dirty = ilk_compute_wm_dirty(dev_priv, previous, results);
 	if (!dirty)
 		return;
 
@@ -3327,13 +3491,18 @@ void valleyview_set_rps(struct drm_device *dev, u8 val)
 	WARN_ON(val > dev_priv->rps.max_freq_softlimit);
 	WARN_ON(val < dev_priv->rps.min_freq_softlimit);
 
-	DRM_DEBUG_DRIVER("GPU freq request from %d MHz (%u) to %d MHz (%u)\n",
-			 vlv_gpu_freq(dev_priv, dev_priv->rps.cur_freq),
-			 dev_priv->rps.cur_freq,
-			 vlv_gpu_freq(dev_priv, val), val);
+	if (WARN_ONCE(IS_CHERRYVIEW(dev) && (val & 1),
+		      "Odd GPU freq value\n"))
+		val &= ~1;
+
+	if (val != dev_priv->rps.cur_freq) {
+		DRM_DEBUG_DRIVER("GPU freq request from %d MHz (%u) to %d MHz (%u)\n",
+				 vlv_gpu_freq(dev_priv, dev_priv->rps.cur_freq),
+				 dev_priv->rps.cur_freq,
+				 vlv_gpu_freq(dev_priv, val), val);
 
-	if (val != dev_priv->rps.cur_freq)
 		vlv_punit_write(dev_priv, PUNIT_REG_GPU_FREQ_REQ, val);
+	}
 
 	I915_WRITE(GEN6_PMINTRMSK, gen6_rps_pm_mask(dev_priv, val));
 
@@ -3406,8 +3575,14 @@ static void valleyview_disable_rps(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
+	/* we're doing forcewake before Disabling RC6,
+	 * This what the BIOS expects when going into suspend */
+	gen6_gt_force_wake_get(dev_priv, FORCEWAKE_ALL);
+
 	I915_WRITE(GEN6_RC_CONTROL, 0);
 
+	gen6_gt_force_wake_put(dev_priv, FORCEWAKE_ALL);
+
 	gen6_disable_rps_interrupts(dev);
 }
 
@@ -3598,7 +3773,6 @@ static void gen6_enable_rps(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_engine_cs *ring;
 	u32 rp_state_cap;
-	u32 gt_perf_status;
 	u32 rc6vids, pcu_mbox = 0, rc6_mask = 0;
 	u32 gtfifodbg;
 	int rc6_mode;
@@ -3623,7 +3797,6 @@ static void gen6_enable_rps(struct drm_device *dev)
 	gen6_gt_force_wake_get(dev_priv, FORCEWAKE_ALL);
 
 	rp_state_cap = I915_READ(GEN6_RP_STATE_CAP);
-	gt_perf_status = I915_READ(GEN6_GT_PERF_STATUS);
 
 	parse_rp_state_cap(dev_priv, rp_state_cap);
 
@@ -3965,11 +4138,27 @@ static void valleyview_cleanup_pctx(struct drm_device *dev)
 static void valleyview_init_gt_powersave(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 val;
 
 	valleyview_setup_pctx(dev);
 
 	mutex_lock(&dev_priv->rps.hw_lock);
 
+	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", dev_priv->mem_freq);
+
 	dev_priv->rps.max_freq = valleyview_rps_max_freq(dev_priv);
 	dev_priv->rps.rp0_freq = dev_priv->rps.max_freq;
 	DRM_DEBUG_DRIVER("max GPU freq: %d MHz (%u)\n",
@@ -4004,11 +4193,38 @@ static void valleyview_init_gt_powersave(struct drm_device *dev)
 static void cherryview_init_gt_powersave(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 val;
 
 	cherryview_setup_pctx(dev);
 
 	mutex_lock(&dev_priv->rps.hw_lock);
 
+	val = vlv_punit_read(dev_priv, CCK_FUSE_REG);
+	switch ((val >> 2) & 0x7) {
+	case 0:
+	case 1:
+		dev_priv->rps.cz_freq = 200;
+		dev_priv->mem_freq = 1600;
+		break;
+	case 2:
+		dev_priv->rps.cz_freq = 267;
+		dev_priv->mem_freq = 1600;
+		break;
+	case 3:
+		dev_priv->rps.cz_freq = 333;
+		dev_priv->mem_freq = 2000;
+		break;
+	case 4:
+		dev_priv->rps.cz_freq = 320;
+		dev_priv->mem_freq = 1600;
+		break;
+	case 5:
+		dev_priv->rps.cz_freq = 400;
+		dev_priv->mem_freq = 1600;
+		break;
+	}
+	DRM_DEBUG_DRIVER("DDR speed: %d MHz", dev_priv->mem_freq);
+
 	dev_priv->rps.max_freq = cherryview_rps_max_freq(dev_priv);
 	dev_priv->rps.rp0_freq = dev_priv->rps.max_freq;
 	DRM_DEBUG_DRIVER("max GPU freq: %d MHz (%u)\n",
@@ -4030,6 +4246,12 @@ static void cherryview_init_gt_powersave(struct drm_device *dev)
 			 vlv_gpu_freq(dev_priv, dev_priv->rps.min_freq),
 			 dev_priv->rps.min_freq);
 
+	WARN_ONCE((dev_priv->rps.max_freq |
+		   dev_priv->rps.efficient_freq |
+		   dev_priv->rps.rp1_freq |
+		   dev_priv->rps.min_freq) & 1,
+		  "Odd GPU freq values\n");
+
 	/* Preserve min/max settings in case of re-init */
 	if (dev_priv->rps.max_freq_softlimit == 0)
 		dev_priv->rps.max_freq_softlimit = dev_priv->rps.max_freq;
@@ -5088,7 +5310,7 @@ static void g4x_disable_trickle_feed(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	int pipe;
 
-	for_each_pipe(pipe) {
+	for_each_pipe(dev_priv, pipe) {
 		I915_WRITE(DSPCNTR(pipe),
 			   I915_READ(DSPCNTR(pipe)) |
 			   DISPPLANE_TRICKLE_FEED_DISABLE);
@@ -5203,7 +5425,7 @@ static void cpt_init_clock_gating(struct drm_device *dev)
 	/* The below fixes the weird display corruption, a few pixels shifted
 	 * downward, on (only) LVDS of some HP laptops with IVY.
 	 */
-	for_each_pipe(pipe) {
+	for_each_pipe(dev_priv, pipe) {
 		val = I915_READ(TRANS_CHICKEN2(pipe));
 		val |= TRANS_CHICKEN2_TIMING_OVERRIDE;
 		val &= ~TRANS_CHICKEN2_FDI_POLARITY_REVERSED;
@@ -5215,7 +5437,7 @@ static void cpt_init_clock_gating(struct drm_device *dev)
 		I915_WRITE(TRANS_CHICKEN2(pipe), val);
 	}
 	/* WADP0ClockGatingDisable */
-	for_each_pipe(pipe) {
+	for_each_pipe(dev_priv, pipe) {
 		I915_WRITE(TRANS_CHICKEN1(pipe),
 			   TRANS_CHICKEN1_DP0UNIT_GC_DISABLE);
 	}
@@ -5383,7 +5605,7 @@ static void lpt_suspend_hw(struct drm_device *dev)
 	}
 }
 
-static void gen8_init_clock_gating(struct drm_device *dev)
+static void broadwell_init_clock_gating(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	enum pipe pipe;
@@ -5395,37 +5617,12 @@ static void gen8_init_clock_gating(struct drm_device *dev)
 	/* FIXME(BDW): Check all the w/a, some might only apply to
 	 * pre-production hw. */
 
-	/* WaDisablePartialInstShootdown:bdw */
-	I915_WRITE(GEN8_ROW_CHICKEN,
-		   _MASKED_BIT_ENABLE(PARTIAL_INSTRUCTION_SHOOTDOWN_DISABLE));
 
-	/* WaDisableThreadStallDopClockGating:bdw */
-	/* FIXME: Unclear whether we really need this on production bdw. */
-	I915_WRITE(GEN8_ROW_CHICKEN,
-		   _MASKED_BIT_ENABLE(STALL_DOP_GATING_DISABLE));
-
-	/*
-	 * This GEN8_CENTROID_PIXEL_OPT_DIS W/A is only needed for
-	 * pre-production hardware
-	 */
-	I915_WRITE(HALF_SLICE_CHICKEN3,
-		   _MASKED_BIT_ENABLE(GEN8_CENTROID_PIXEL_OPT_DIS));
-	I915_WRITE(HALF_SLICE_CHICKEN3,
-		   _MASKED_BIT_ENABLE(GEN8_SAMPLER_POWER_BYPASS_DIS));
 	I915_WRITE(GAMTARBMODE, _MASKED_BIT_ENABLE(ARB_MODE_BWGTLB_DISABLE));
 
 	I915_WRITE(_3D_CHICKEN3,
 		   _MASKED_BIT_ENABLE(_3D_CHICKEN_SDE_LIMIT_FIFO_POLY_DEPTH(2)));
 
-	I915_WRITE(COMMON_SLICE_CHICKEN2,
-		   _MASKED_BIT_ENABLE(GEN8_CSC2_SBE_VUE_CACHE_CONSERVATIVE));
-
-	I915_WRITE(GEN7_HALF_SLICE_CHICKEN1,
-		   _MASKED_BIT_ENABLE(GEN7_SINGLE_SUBSCAN_DISPATCH_ENABLE));
-
-	/* WaDisableDopClockGating:bdw May not be needed for production */
-	I915_WRITE(GEN7_ROW_CHICKEN2,
-		   _MASKED_BIT_ENABLE(DOP_CLOCK_GATING_DISABLE));
 
 	/* WaSwitchSolVfFArbitrationPriority:bdw */
 	I915_WRITE(GAM_ECOCHK, I915_READ(GAM_ECOCHK) | HSW_ECOCHK_ARB_PRIO_SOL);
@@ -5435,37 +5632,18 @@ static void gen8_init_clock_gating(struct drm_device *dev)
 		   I915_READ(CHICKEN_PAR1_1) | DPA_MASK_VBLANK_SRD);
 
 	/* WaPsrDPRSUnmaskVBlankInSRD:bdw */
-	for_each_pipe(pipe) {
+	for_each_pipe(dev_priv, pipe) {
 		I915_WRITE(CHICKEN_PIPESL_1(pipe),
 			   I915_READ(CHICKEN_PIPESL_1(pipe)) |
 			   BDW_DPRS_MASK_VBLANK_SRD);
 	}
 
-	/* Use Force Non-Coherent whenever executing a 3D context. This is a
-	 * workaround for for a possible hang in the unlikely event a TLB
-	 * invalidation occurs during a PSD flush.
-	 */
-	I915_WRITE(HDC_CHICKEN0,
-		   I915_READ(HDC_CHICKEN0) |
-		   _MASKED_BIT_ENABLE(HDC_FORCE_NON_COHERENT));
-
 	/* WaVSRefCountFullforceMissDisable:bdw */
 	/* WaDSRefCountFullforceMissDisable:bdw */
 	I915_WRITE(GEN7_FF_THREAD_MODE,
 		   I915_READ(GEN7_FF_THREAD_MODE) &
 		   ~(GEN8_FF_DS_REF_CNT_FFME | GEN7_FF_VS_REF_CNT_FFME));
 
-	/*
-	 * BSpec recommends 8x4 when MSAA is used,
-	 * however in practice 16x4 seems fastest.
-	 *
-	 * Note that PS/WM thread counts depend on the WIZ hashing
-	 * disable bit, which we don't touch here, but it's good
-	 * to keep in mind (see 3DSTATE_PS and 3DSTATE_WM).
-	 */
-	I915_WRITE(GEN7_GT_MODE,
-		   GEN6_WIZ_HASHING_MASK | GEN6_WIZ_HASHING_16x4);
-
 	I915_WRITE(GEN6_RC_SLEEP_PSMI_CONTROL,
 		   _MASKED_BIT_ENABLE(GEN8_RC_SEMA_IDLE_MSG_DISABLE));
 
@@ -5473,9 +5651,7 @@ static void gen8_init_clock_gating(struct drm_device *dev)
 	I915_WRITE(GEN8_UCGCTL6, I915_READ(GEN8_UCGCTL6) |
 		   GEN8_SDEUNIT_CLOCK_GATE_DISABLE);
 
-	/* Wa4x4STCOptimizationDisable:bdw */
-	I915_WRITE(CACHE_MODE_1,
-		   _MASKED_BIT_ENABLE(GEN8_4x4_STC_OPTIMIZATION_DISABLE));
+	lpt_init_clock_gating(dev);
 }
 
 static void haswell_init_clock_gating(struct drm_device *dev)
@@ -5631,24 +5807,6 @@ static void ivybridge_init_clock_gating(struct drm_device *dev)
 static void valleyview_init_clock_gating(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	u32 val;
-
-	mutex_lock(&dev_priv->rps.hw_lock);
-	val = vlv_punit_read(dev_priv, PUNIT_REG_GPU_FREQ_STS);
-	mutex_unlock(&dev_priv->rps.hw_lock);
-	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", dev_priv->mem_freq);
 
 	I915_WRITE(DSPCLK_GATE_D, VRHUNIT_CLOCK_GATE_DISABLE);
 
@@ -5724,48 +5882,11 @@ static void valleyview_init_clock_gating(struct drm_device *dev)
 static void cherryview_init_clock_gating(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	u32 val;
-
-	mutex_lock(&dev_priv->rps.hw_lock);
-	val = vlv_punit_read(dev_priv, CCK_FUSE_REG);
-	mutex_unlock(&dev_priv->rps.hw_lock);
-	switch ((val >> 2) & 0x7) {
-	case 0:
-	case 1:
-			dev_priv->rps.cz_freq = CHV_CZ_CLOCK_FREQ_MODE_200;
-			dev_priv->mem_freq = 1600;
-			break;
-	case 2:
-			dev_priv->rps.cz_freq = CHV_CZ_CLOCK_FREQ_MODE_267;
-			dev_priv->mem_freq = 1600;
-			break;
-	case 3:
-			dev_priv->rps.cz_freq = CHV_CZ_CLOCK_FREQ_MODE_333;
-			dev_priv->mem_freq = 2000;
-			break;
-	case 4:
-			dev_priv->rps.cz_freq = CHV_CZ_CLOCK_FREQ_MODE_320;
-			dev_priv->mem_freq = 1600;
-			break;
-	case 5:
-			dev_priv->rps.cz_freq = CHV_CZ_CLOCK_FREQ_MODE_400;
-			dev_priv->mem_freq = 1600;
-			break;
-	}
-	DRM_DEBUG_DRIVER("DDR speed: %d MHz", dev_priv->mem_freq);
 
 	I915_WRITE(DSPCLK_GATE_D, VRHUNIT_CLOCK_GATE_DISABLE);
 
 	I915_WRITE(MI_ARB_VLV, MI_ARB_DISPLAY_TRICKLE_FEED_DISABLE);
 
-	/* WaDisablePartialInstShootdown:chv */
-	I915_WRITE(GEN8_ROW_CHICKEN,
-		   _MASKED_BIT_ENABLE(PARTIAL_INSTRUCTION_SHOOTDOWN_DISABLE));
-
-	/* WaDisableThreadStallDopClockGating:chv */
-	I915_WRITE(GEN8_ROW_CHICKEN,
-		   _MASKED_BIT_ENABLE(STALL_DOP_GATING_DISABLE));
-
 	/* WaVSRefCountFullforceMissDisable:chv */
 	/* WaDSRefCountFullforceMissDisable:chv */
 	I915_WRITE(GEN7_FF_THREAD_MODE,
@@ -5784,10 +5905,6 @@ static void cherryview_init_clock_gating(struct drm_device *dev)
 	I915_WRITE(GEN8_UCGCTL6, I915_READ(GEN8_UCGCTL6) |
 		   GEN8_SDEUNIT_CLOCK_GATE_DISABLE);
 
-	/* WaDisableSamplerPowerBypass:chv (pre-production hw) */
-	I915_WRITE(HALF_SLICE_CHICKEN3,
-		   _MASKED_BIT_ENABLE(GEN8_SAMPLER_POWER_BYPASS_DIS));
-
 	/* WaDisableGunitClockGating:chv (pre-production hw) */
 	I915_WRITE(VLV_GUNIT_CLOCK_GATE, I915_READ(VLV_GUNIT_CLOCK_GATE) |
 		   GINT_DIS);
@@ -5797,8 +5914,6 @@ static void cherryview_init_clock_gating(struct drm_device *dev)
 		   _MASKED_BIT_ENABLE(GEN8_FF_DOP_CLOCK_GATE_DISABLE));
 
 	/* WaDisableDopClockGating:chv (pre-production hw) */
-	I915_WRITE(GEN7_ROW_CHICKEN2,
-		   _MASKED_BIT_ENABLE(DOP_CLOCK_GATING_DISABLE));
 	I915_WRITE(GEN6_UCGCTL1, I915_READ(GEN6_UCGCTL1) |
 		   GEN6_EU_TCUNIT_CLOCK_GATE_DISABLE);
 }
@@ -5883,6 +5998,9 @@ static void gen3_init_clock_gating(struct drm_device *dev)
 
 	/* On GEN3 we really need to make sure the ARB C3 LP bit is set */
 	I915_WRITE(MI_ARB_STATE, _MASKED_BIT_ENABLE(MI_ARB_C3_LP_WRITE_ENABLE));
+
+	I915_WRITE(MI_ARB_STATE,
+		   _MASKED_BIT_ENABLE(MI_ARB_DISPLAY_TRICKLE_FEED_DISABLE));
 }
 
 static void i85x_init_clock_gating(struct drm_device *dev)
@@ -5894,6 +6012,9 @@ static void i85x_init_clock_gating(struct drm_device *dev)
 	/* interrupts should cause a wake up from C3 */
 	I915_WRITE(MI_STATE, _MASKED_BIT_ENABLE(MI_AGPBUSY_INT_EN) |
 		   _MASKED_BIT_DISABLE(MI_AGPBUSY_830_MODE));
+
+	I915_WRITE(MEM_MODE,
+		   _MASKED_BIT_ENABLE(MEM_DISPLAY_TRICKLE_FEED_DISABLE));
 }
 
 static void i830_init_clock_gating(struct drm_device *dev)
@@ -5901,6 +6022,10 @@ static void i830_init_clock_gating(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
 	I915_WRITE(DSPCLK_GATE_D, OVRUNIT_CLOCK_GATE_DISABLE);
+
+	I915_WRITE(MEM_MODE,
+		   _MASKED_BIT_ENABLE(MEM_DISPLAY_A_TRICKLE_FEED_DISABLE) |
+		   _MASKED_BIT_ENABLE(MEM_DISPLAY_B_TRICKLE_FEED_DISABLE));
 }
 
 void intel_init_clock_gating(struct drm_device *dev)
@@ -6203,6 +6328,8 @@ static void vlv_display_power_well_disable(struct drm_i915_private *dev_priv,
 	spin_unlock_irq(&dev_priv->irq_lock);
 
 	vlv_set_power_well(dev_priv, power_well, false);
+
+	vlv_power_sequencer_reset(dev_priv);
 }
 
 static void vlv_dpio_cmn_power_well_enable(struct drm_i915_private *dev_priv,
@@ -6238,12 +6365,11 @@ static void vlv_dpio_cmn_power_well_enable(struct drm_i915_private *dev_priv,
 static void vlv_dpio_cmn_power_well_disable(struct drm_i915_private *dev_priv,
 					    struct i915_power_well *power_well)
 {
-	struct drm_device *dev = dev_priv->dev;
 	enum pipe pipe;
 
 	WARN_ON_ONCE(power_well->data != PUNIT_POWER_WELL_DPIO_CMN_BC);
 
-	for_each_pipe(pipe)
+	for_each_pipe(dev_priv, pipe)
 		assert_pll_disabled(dev_priv, pipe);
 
 	/* Assert common reset */
@@ -6252,6 +6378,153 @@ static void vlv_dpio_cmn_power_well_disable(struct drm_i915_private *dev_priv,
 	vlv_set_power_well(dev_priv, power_well, false);
 }
 
+static void chv_dpio_cmn_power_well_enable(struct drm_i915_private *dev_priv,
+					   struct i915_power_well *power_well)
+{
+	enum dpio_phy phy;
+
+	WARN_ON_ONCE(power_well->data != PUNIT_POWER_WELL_DPIO_CMN_BC &&
+		     power_well->data != PUNIT_POWER_WELL_DPIO_CMN_D);
+
+	/*
+	 * Enable the CRI clock source so we can get at the
+	 * display and the reference clock for VGA
+	 * hotplug / manual detection.
+	 */
+	if (power_well->data == PUNIT_POWER_WELL_DPIO_CMN_BC) {
+		phy = DPIO_PHY0;
+		I915_WRITE(DPLL(PIPE_B), I915_READ(DPLL(PIPE_B)) |
+			   DPLL_REFA_CLK_ENABLE_VLV);
+		I915_WRITE(DPLL(PIPE_B), I915_READ(DPLL(PIPE_B)) |
+			   DPLL_REFA_CLK_ENABLE_VLV | DPLL_INTEGRATED_CRI_CLK_VLV);
+	} else {
+		phy = DPIO_PHY1;
+		I915_WRITE(DPLL(PIPE_C), I915_READ(DPLL(PIPE_C)) |
+			   DPLL_REFA_CLK_ENABLE_VLV | DPLL_INTEGRATED_CRI_CLK_VLV);
+	}
+	udelay(1); /* >10ns for cmnreset, >0ns for sidereset */
+	vlv_set_power_well(dev_priv, power_well, true);
+
+	/* Poll for phypwrgood signal */
+	if (wait_for(I915_READ(DISPLAY_PHY_STATUS) & PHY_POWERGOOD(phy), 1))
+		DRM_ERROR("Display PHY %d is not power up\n", phy);
+
+	I915_WRITE(DISPLAY_PHY_CONTROL, I915_READ(DISPLAY_PHY_CONTROL) |
+		   PHY_COM_LANE_RESET_DEASSERT(phy));
+}
+
+static void chv_dpio_cmn_power_well_disable(struct drm_i915_private *dev_priv,
+					    struct i915_power_well *power_well)
+{
+	enum dpio_phy phy;
+
+	WARN_ON_ONCE(power_well->data != PUNIT_POWER_WELL_DPIO_CMN_BC &&
+		     power_well->data != PUNIT_POWER_WELL_DPIO_CMN_D);
+
+	if (power_well->data == PUNIT_POWER_WELL_DPIO_CMN_BC) {
+		phy = DPIO_PHY0;
+		assert_pll_disabled(dev_priv, PIPE_A);
+		assert_pll_disabled(dev_priv, PIPE_B);
+	} else {
+		phy = DPIO_PHY1;
+		assert_pll_disabled(dev_priv, PIPE_C);
+	}
+
+	I915_WRITE(DISPLAY_PHY_CONTROL, I915_READ(DISPLAY_PHY_CONTROL) &
+		   ~PHY_COM_LANE_RESET_DEASSERT(phy));
+
+	vlv_set_power_well(dev_priv, power_well, false);
+}
+
+static bool chv_pipe_power_well_enabled(struct drm_i915_private *dev_priv,
+					struct i915_power_well *power_well)
+{
+	enum pipe pipe = power_well->data;
+	bool enabled;
+	u32 state, ctrl;
+
+	mutex_lock(&dev_priv->rps.hw_lock);
+
+	state = vlv_punit_read(dev_priv, PUNIT_REG_DSPFREQ) & DP_SSS_MASK(pipe);
+	/*
+	 * We only ever set the power-on and power-gate states, anything
+	 * else is unexpected.
+	 */
+	WARN_ON(state != DP_SSS_PWR_ON(pipe) && state != DP_SSS_PWR_GATE(pipe));
+	enabled = state == DP_SSS_PWR_ON(pipe);
+
+	/*
+	 * A transient state at this point would mean some unexpected party
+	 * is poking at the power controls too.
+	 */
+	ctrl = vlv_punit_read(dev_priv, PUNIT_REG_DSPFREQ) & DP_SSC_MASK(pipe);
+	WARN_ON(ctrl << 16 != state);
+
+	mutex_unlock(&dev_priv->rps.hw_lock);
+
+	return enabled;
+}
+
+static void chv_set_pipe_power_well(struct drm_i915_private *dev_priv,
+				    struct i915_power_well *power_well,
+				    bool enable)
+{
+	enum pipe pipe = power_well->data;
+	u32 state;
+	u32 ctrl;
+
+	state = enable ? DP_SSS_PWR_ON(pipe) : DP_SSS_PWR_GATE(pipe);
+
+	mutex_lock(&dev_priv->rps.hw_lock);
+
+#define COND \
+	((vlv_punit_read(dev_priv, PUNIT_REG_DSPFREQ) & DP_SSS_MASK(pipe)) == state)
+
+	if (COND)
+		goto out;
+
+	ctrl = vlv_punit_read(dev_priv, PUNIT_REG_DSPFREQ);
+	ctrl &= ~DP_SSC_MASK(pipe);
+	ctrl |= enable ? DP_SSC_PWR_ON(pipe) : DP_SSC_PWR_GATE(pipe);
+	vlv_punit_write(dev_priv, PUNIT_REG_DSPFREQ, ctrl);
+
+	if (wait_for(COND, 100))
+		DRM_ERROR("timout setting power well state %08x (%08x)\n",
+			  state,
+			  vlv_punit_read(dev_priv, PUNIT_REG_DSPFREQ));
+
+#undef COND
+
+out:
+	mutex_unlock(&dev_priv->rps.hw_lock);
+}
+
+static void chv_pipe_power_well_sync_hw(struct drm_i915_private *dev_priv,
+					struct i915_power_well *power_well)
+{
+	chv_set_pipe_power_well(dev_priv, power_well, power_well->count > 0);
+}
+
+static void chv_pipe_power_well_enable(struct drm_i915_private *dev_priv,
+				       struct i915_power_well *power_well)
+{
+	WARN_ON_ONCE(power_well->data != PIPE_A &&
+		     power_well->data != PIPE_B &&
+		     power_well->data != PIPE_C);
+
+	chv_set_pipe_power_well(dev_priv, power_well, true);
+}
+
+static void chv_pipe_power_well_disable(struct drm_i915_private *dev_priv,
+					struct i915_power_well *power_well)
+{
+	WARN_ON_ONCE(power_well->data != PIPE_A &&
+		     power_well->data != PIPE_B &&
+		     power_well->data != PIPE_C);
+
+	chv_set_pipe_power_well(dev_priv, power_well, false);
+}
+
 static void check_power_well_state(struct drm_i915_private *dev_priv,
 				   struct i915_power_well *power_well)
 {
@@ -6443,6 +6716,39 @@ EXPORT_SYMBOL_GPL(i915_get_cdclk_freq);
 	BIT(POWER_DOMAIN_PORT_DDI_C_4_LANES) |	\
 	BIT(POWER_DOMAIN_INIT))
 
+#define CHV_PIPE_A_POWER_DOMAINS (	\
+	BIT(POWER_DOMAIN_PIPE_A) |	\
+	BIT(POWER_DOMAIN_INIT))
+
+#define CHV_PIPE_B_POWER_DOMAINS (	\
+	BIT(POWER_DOMAIN_PIPE_B) |	\
+	BIT(POWER_DOMAIN_INIT))
+
+#define CHV_PIPE_C_POWER_DOMAINS (	\
+	BIT(POWER_DOMAIN_PIPE_C) |	\
+	BIT(POWER_DOMAIN_INIT))
+
+#define CHV_DPIO_CMN_BC_POWER_DOMAINS (		\
+	BIT(POWER_DOMAIN_PORT_DDI_B_2_LANES) |	\
+	BIT(POWER_DOMAIN_PORT_DDI_B_4_LANES) |	\
+	BIT(POWER_DOMAIN_PORT_DDI_C_2_LANES) |	\
+	BIT(POWER_DOMAIN_PORT_DDI_C_4_LANES) |	\
+	BIT(POWER_DOMAIN_INIT))
+
+#define CHV_DPIO_CMN_D_POWER_DOMAINS (		\
+	BIT(POWER_DOMAIN_PORT_DDI_D_2_LANES) |	\
+	BIT(POWER_DOMAIN_PORT_DDI_D_4_LANES) |	\
+	BIT(POWER_DOMAIN_INIT))
+
+#define CHV_DPIO_TX_D_LANES_01_POWER_DOMAINS (	\
+	BIT(POWER_DOMAIN_PORT_DDI_D_2_LANES) |	\
+	BIT(POWER_DOMAIN_PORT_DDI_D_4_LANES) |	\
+	BIT(POWER_DOMAIN_INIT))
+
+#define CHV_DPIO_TX_D_LANES_23_POWER_DOMAINS (	\
+	BIT(POWER_DOMAIN_PORT_DDI_D_4_LANES) |	\
+	BIT(POWER_DOMAIN_INIT))
+
 static const struct i915_power_well_ops i9xx_always_on_power_well_ops = {
 	.sync_hw = i9xx_always_on_power_well_noop,
 	.enable = i9xx_always_on_power_well_noop,
@@ -6450,6 +6756,20 @@ static const struct i915_power_well_ops i9xx_always_on_power_well_ops = {
 	.is_enabled = i9xx_always_on_power_well_enabled,
 };
 
+static const struct i915_power_well_ops chv_pipe_power_well_ops = {
+	.sync_hw = chv_pipe_power_well_sync_hw,
+	.enable = chv_pipe_power_well_enable,
+	.disable = chv_pipe_power_well_disable,
+	.is_enabled = chv_pipe_power_well_enabled,
+};
+
+static const struct i915_power_well_ops chv_dpio_cmn_power_well_ops = {
+	.sync_hw = vlv_power_well_sync_hw,
+	.enable = chv_dpio_cmn_power_well_enable,
+	.disable = chv_dpio_cmn_power_well_disable,
+	.is_enabled = vlv_power_well_enabled,
+};
+
 static struct i915_power_well i9xx_always_on_power_well[] = {
 	{
 		.name = "always-on",
@@ -6572,6 +6892,107 @@ static struct i915_power_well vlv_power_wells[] = {
 	},
 };
 
+static struct i915_power_well chv_power_wells[] = {
+	{
+		.name = "always-on",
+		.always_on = 1,
+		.domains = VLV_ALWAYS_ON_POWER_DOMAINS,
+		.ops = &i9xx_always_on_power_well_ops,
+	},
+#if 0
+	{
+		.name = "display",
+		.domains = VLV_DISPLAY_POWER_DOMAINS,
+		.data = PUNIT_POWER_WELL_DISP2D,
+		.ops = &vlv_display_power_well_ops,
+	},
+	{
+		.name = "pipe-a",
+		.domains = CHV_PIPE_A_POWER_DOMAINS,
+		.data = PIPE_A,
+		.ops = &chv_pipe_power_well_ops,
+	},
+	{
+		.name = "pipe-b",
+		.domains = CHV_PIPE_B_POWER_DOMAINS,
+		.data = PIPE_B,
+		.ops = &chv_pipe_power_well_ops,
+	},
+	{
+		.name = "pipe-c",
+		.domains = CHV_PIPE_C_POWER_DOMAINS,
+		.data = PIPE_C,
+		.ops = &chv_pipe_power_well_ops,
+	},
+#endif
+	{
+		.name = "dpio-common-bc",
+		/*
+		 * XXX: cmnreset for one PHY seems to disturb the other.
+		 * As a workaround keep both powered on at the same
+		 * time for now.
+		 */
+		.domains = CHV_DPIO_CMN_BC_POWER_DOMAINS | CHV_DPIO_CMN_D_POWER_DOMAINS,
+		.data = PUNIT_POWER_WELL_DPIO_CMN_BC,
+		.ops = &chv_dpio_cmn_power_well_ops,
+	},
+	{
+		.name = "dpio-common-d",
+		/*
+		 * XXX: cmnreset for one PHY seems to disturb the other.
+		 * As a workaround keep both powered on at the same
+		 * time for now.
+		 */
+		.domains = CHV_DPIO_CMN_BC_POWER_DOMAINS | CHV_DPIO_CMN_D_POWER_DOMAINS,
+		.data = PUNIT_POWER_WELL_DPIO_CMN_D,
+		.ops = &chv_dpio_cmn_power_well_ops,
+	},
+#if 0
+	{
+		.name = "dpio-tx-b-01",
+		.domains = VLV_DPIO_TX_B_LANES_01_POWER_DOMAINS |
+			   VLV_DPIO_TX_B_LANES_23_POWER_DOMAINS,
+		.ops = &vlv_dpio_power_well_ops,
+		.data = PUNIT_POWER_WELL_DPIO_TX_B_LANES_01,
+	},
+	{
+		.name = "dpio-tx-b-23",
+		.domains = VLV_DPIO_TX_B_LANES_01_POWER_DOMAINS |
+			   VLV_DPIO_TX_B_LANES_23_POWER_DOMAINS,
+		.ops = &vlv_dpio_power_well_ops,
+		.data = PUNIT_POWER_WELL_DPIO_TX_B_LANES_23,
+	},
+	{
+		.name = "dpio-tx-c-01",
+		.domains = VLV_DPIO_TX_C_LANES_01_POWER_DOMAINS |
+			   VLV_DPIO_TX_C_LANES_23_POWER_DOMAINS,
+		.ops = &vlv_dpio_power_well_ops,
+		.data = PUNIT_POWER_WELL_DPIO_TX_C_LANES_01,
+	},
+	{
+		.name = "dpio-tx-c-23",
+		.domains = VLV_DPIO_TX_C_LANES_01_POWER_DOMAINS |
+			   VLV_DPIO_TX_C_LANES_23_POWER_DOMAINS,
+		.ops = &vlv_dpio_power_well_ops,
+		.data = PUNIT_POWER_WELL_DPIO_TX_C_LANES_23,
+	},
+	{
+		.name = "dpio-tx-d-01",
+		.domains = CHV_DPIO_TX_D_LANES_01_POWER_DOMAINS |
+			   CHV_DPIO_TX_D_LANES_23_POWER_DOMAINS,
+		.ops = &vlv_dpio_power_well_ops,
+		.data = PUNIT_POWER_WELL_DPIO_TX_D_LANES_01,
+	},
+	{
+		.name = "dpio-tx-d-23",
+		.domains = CHV_DPIO_TX_D_LANES_01_POWER_DOMAINS |
+			   CHV_DPIO_TX_D_LANES_23_POWER_DOMAINS,
+		.ops = &vlv_dpio_power_well_ops,
+		.data = PUNIT_POWER_WELL_DPIO_TX_D_LANES_23,
+	},
+#endif
+};
+
 static struct i915_power_well *lookup_power_well(struct drm_i915_private *dev_priv,
 						 enum punit_power_well power_well_id)
 {
@@ -6608,6 +7029,8 @@ int intel_power_domains_init(struct drm_i915_private *dev_priv)
 	} else if (IS_BROADWELL(dev_priv->dev)) {
 		set_power_wells(power_domains, bdw_power_wells);
 		hsw_pwr = power_domains;
+	} else if (IS_CHERRYVIEW(dev_priv->dev)) {
+		set_power_wells(power_domains, chv_power_wells);
 	} else if (IS_VALLEYVIEW(dev_priv->dev)) {
 		set_power_wells(power_domains, vlv_power_wells);
 	} else {
@@ -6833,13 +7256,15 @@ void intel_init_pm(struct drm_device *dev)
 		else if (IS_HASWELL(dev))
 			dev_priv->display.init_clock_gating = haswell_init_clock_gating;
 		else if (INTEL_INFO(dev)->gen == 8)
-			dev_priv->display.init_clock_gating = gen8_init_clock_gating;
+			dev_priv->display.init_clock_gating = broadwell_init_clock_gating;
 	} else if (IS_CHERRYVIEW(dev)) {
-		dev_priv->display.update_wm = valleyview_update_wm;
+		dev_priv->display.update_wm = cherryview_update_wm;
+		dev_priv->display.update_sprite_wm = valleyview_update_sprite_wm;
 		dev_priv->display.init_clock_gating =
 			cherryview_init_clock_gating;
 	} else if (IS_VALLEYVIEW(dev)) {
 		dev_priv->display.update_wm = valleyview_update_wm;
+		dev_priv->display.update_sprite_wm = valleyview_update_sprite_wm;
 		dev_priv->display.init_clock_gating =
 			valleyview_init_clock_gating;
 	} else if (IS_PINEVIEW(dev)) {
@@ -7025,6 +7450,7 @@ static int chv_freq_opcode(struct drm_i915_private *dev_priv, int val)
 		return -1;
 	}
 
+	/* CHV needs even values */
 	opcode = (DIV_ROUND_CLOSEST((val * 2 * mul), dev_priv->rps.cz_freq) * 2);
 
 	return opcode;
diff --git a/drivers/gpu/drm/i915/intel_renderstate.h b/drivers/gpu/drm/i915/intel_renderstate.h
index fd4f66231d30..6c792d3a9c9c 100644
--- a/drivers/gpu/drm/i915/intel_renderstate.h
+++ b/drivers/gpu/drm/i915/intel_renderstate.h
@@ -24,13 +24,7 @@
 #ifndef _INTEL_RENDERSTATE_H
 #define _INTEL_RENDERSTATE_H
 
-#include <linux/types.h>
-
-struct intel_renderstate_rodata {
-	const u32 *reloc;
-	const u32 *batch;
-	const u32 batch_items;
-};
+#include "i915_drv.h"
 
 extern const struct intel_renderstate_rodata gen6_null_state;
 extern const struct intel_renderstate_rodata gen7_null_state;
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index 47a126a0493f..0a80e419b589 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -33,14 +33,24 @@
 #include "i915_trace.h"
 #include "intel_drv.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
+bool
+intel_ring_initialized(struct intel_engine_cs *ring)
+{
+	struct drm_device *dev = ring->dev;
+
+	if (!dev)
+		return false;
+
+	if (i915.enable_execlists) {
+		struct intel_context *dctx = ring->default_context;
+		struct intel_ringbuffer *ringbuf = dctx->engine[ring->id].ringbuf;
+
+		return ringbuf->obj;
+	} else
+		return ring->buffer && ring->buffer->obj;
+}
 
-static inline int __ring_space(int head, int tail, int size)
+int __intel_ring_space(int head, int tail, int size)
 {
 	int space = head - (tail + I915_RING_FREE_SPACE);
 	if (space < 0)
@@ -48,12 +58,13 @@ static inline int __ring_space(int head, int tail, int size)
 	return space;
 }
 
-static inline int ring_space(struct intel_ringbuffer *ringbuf)
+int intel_ring_space(struct intel_ringbuffer *ringbuf)
 {
-	return __ring_space(ringbuf->head & HEAD_ADDR, ringbuf->tail, ringbuf->size);
+	return __intel_ring_space(ringbuf->head & HEAD_ADDR,
+				  ringbuf->tail, ringbuf->size);
 }
 
-static bool intel_ring_stopped(struct intel_engine_cs *ring)
+bool intel_ring_stopped(struct intel_engine_cs *ring)
 {
 	struct drm_i915_private *dev_priv = ring->dev->dev_private;
 	return dev_priv->gpu_error.stop_rings & intel_ring_flag(ring);
@@ -433,7 +444,14 @@ gen8_render_ring_flush(struct intel_engine_cs *ring,
 			return ret;
 	}
 
-	return gen8_emit_pipe_control(ring, flags, scratch_addr);
+	ret = gen8_emit_pipe_control(ring, flags, scratch_addr);
+	if (ret)
+		return ret;
+
+	if (!invalidate_domains && flush_domains)
+		return gen7_ring_fbc_flush(ring, FBC_REND_NUKE);
+
+	return 0;
 }
 
 static void ring_write_tail(struct intel_engine_cs *ring,
@@ -476,9 +494,14 @@ static bool stop_ring(struct intel_engine_cs *ring)
 
 	if (!IS_GEN2(ring->dev)) {
 		I915_WRITE_MODE(ring, _MASKED_BIT_ENABLE(STOP_RING));
-		if (wait_for_atomic((I915_READ_MODE(ring) & MODE_IDLE) != 0, 1000)) {
-			DRM_ERROR("%s :timed out trying to stop ring\n", ring->name);
-			return false;
+		if (wait_for((I915_READ_MODE(ring) & MODE_IDLE) != 0, 1000)) {
+			DRM_ERROR("%s : timed out trying to stop ring\n", ring->name);
+			/* Sometimes we observe that the idle flag is not
+			 * set even though the ring is empty. So double
+			 * check before giving up.
+			 */
+			if (I915_READ_HEAD(ring) != I915_READ_TAIL(ring))
+				return false;
 		}
 	}
 
@@ -540,6 +563,14 @@ static int init_ring_common(struct intel_engine_cs *ring)
 	 * also enforces ordering), otherwise the hw might lose the new ring
 	 * register values. */
 	I915_WRITE_START(ring, i915_gem_obj_ggtt_offset(obj));
+
+	/* WaClearRingBufHeadRegAtInit:ctg,elk */
+	if (I915_READ_HEAD(ring))
+		DRM_DEBUG("%s initialization failed [head=%08x], fudging\n",
+			  ring->name, I915_READ_HEAD(ring));
+	I915_WRITE_HEAD(ring, 0);
+	(void)I915_READ_HEAD(ring);
+
 	I915_WRITE_CTL(ring,
 			((ringbuf->size - PAGE_SIZE) & RING_NR_PAGES)
 			| RING_VALID);
@@ -563,7 +594,7 @@ static int init_ring_common(struct intel_engine_cs *ring)
 	else {
 		ringbuf->head = I915_READ_HEAD(ring);
 		ringbuf->tail = I915_READ_TAIL(ring) & TAIL_ADDR;
-		ringbuf->space = ring_space(ringbuf);
+		ringbuf->space = intel_ring_space(ringbuf);
 		ringbuf->last_retired_head = -1;
 	}
 
@@ -575,8 +606,25 @@ out:
 	return ret;
 }
 
-static int
-init_pipe_control(struct intel_engine_cs *ring)
+void
+intel_fini_pipe_control(struct intel_engine_cs *ring)
+{
+	struct drm_device *dev = ring->dev;
+
+	if (ring->scratch.obj == NULL)
+		return;
+
+	if (INTEL_INFO(dev)->gen >= 5) {
+		kunmap(sg_page(ring->scratch.obj->pages->sgl));
+		i915_gem_object_ggtt_unpin(ring->scratch.obj);
+	}
+
+	drm_gem_object_unreference(&ring->scratch.obj->base);
+	ring->scratch.obj = NULL;
+}
+
+int
+intel_init_pipe_control(struct intel_engine_cs *ring)
 {
 	int ret;
 
@@ -617,6 +665,135 @@ err:
 	return ret;
 }
 
+static inline void intel_ring_emit_wa(struct intel_engine_cs *ring,
+				       u32 addr, u32 value)
+{
+	struct drm_device *dev = ring->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	if (WARN_ON(dev_priv->num_wa_regs >= I915_MAX_WA_REGS))
+		return;
+
+	intel_ring_emit(ring, MI_LOAD_REGISTER_IMM(1));
+	intel_ring_emit(ring, addr);
+	intel_ring_emit(ring, value);
+
+	dev_priv->intel_wa_regs[dev_priv->num_wa_regs].addr = addr;
+	dev_priv->intel_wa_regs[dev_priv->num_wa_regs].mask = value & 0xFFFF;
+	/* value is updated with the status of remaining bits of this
+	 * register when it is read from debugfs file
+	 */
+	dev_priv->intel_wa_regs[dev_priv->num_wa_regs].value = value;
+	dev_priv->num_wa_regs++;
+
+	return;
+}
+
+static int bdw_init_workarounds(struct intel_engine_cs *ring)
+{
+	int ret;
+	struct drm_device *dev = ring->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	/*
+	 * workarounds applied in this fn are part of register state context,
+	 * they need to be re-initialized followed by gpu reset, suspend/resume,
+	 * module reload.
+	 */
+	dev_priv->num_wa_regs = 0;
+	memset(dev_priv->intel_wa_regs, 0, sizeof(dev_priv->intel_wa_regs));
+
+	/*
+	 * update the number of dwords required based on the
+	 * actual number of workarounds applied
+	 */
+	ret = intel_ring_begin(ring, 18);
+	if (ret)
+		return ret;
+
+	/* WaDisablePartialInstShootdown:bdw */
+	/* WaDisableThreadStallDopClockGating:bdw */
+	/* FIXME: Unclear whether we really need this on production bdw. */
+	intel_ring_emit_wa(ring, GEN8_ROW_CHICKEN,
+			   _MASKED_BIT_ENABLE(PARTIAL_INSTRUCTION_SHOOTDOWN_DISABLE
+					     | STALL_DOP_GATING_DISABLE));
+
+	/* WaDisableDopClockGating:bdw May not be needed for production */
+	intel_ring_emit_wa(ring, GEN7_ROW_CHICKEN2,
+			   _MASKED_BIT_ENABLE(DOP_CLOCK_GATING_DISABLE));
+
+	intel_ring_emit_wa(ring, HALF_SLICE_CHICKEN3,
+			   _MASKED_BIT_ENABLE(GEN8_SAMPLER_POWER_BYPASS_DIS));
+
+	/* Use Force Non-Coherent whenever executing a 3D context. This is a
+	 * workaround for for a possible hang in the unlikely event a TLB
+	 * invalidation occurs during a PSD flush.
+	 */
+	intel_ring_emit_wa(ring, HDC_CHICKEN0,
+			   _MASKED_BIT_ENABLE(HDC_FORCE_NON_COHERENT));
+
+	/* Wa4x4STCOptimizationDisable:bdw */
+	intel_ring_emit_wa(ring, CACHE_MODE_1,
+			   _MASKED_BIT_ENABLE(GEN8_4x4_STC_OPTIMIZATION_DISABLE));
+
+	/*
+	 * BSpec recommends 8x4 when MSAA is used,
+	 * however in practice 16x4 seems fastest.
+	 *
+	 * Note that PS/WM thread counts depend on the WIZ hashing
+	 * disable bit, which we don't touch here, but it's good
+	 * to keep in mind (see 3DSTATE_PS and 3DSTATE_WM).
+	 */
+	intel_ring_emit_wa(ring, GEN7_GT_MODE,
+			   GEN6_WIZ_HASHING_MASK | GEN6_WIZ_HASHING_16x4);
+
+	intel_ring_advance(ring);
+
+	DRM_DEBUG_DRIVER("Number of Workarounds applied: %d\n",
+			 dev_priv->num_wa_regs);
+
+	return 0;
+}
+
+static int chv_init_workarounds(struct intel_engine_cs *ring)
+{
+	int ret;
+	struct drm_device *dev = ring->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	/*
+	 * workarounds applied in this fn are part of register state context,
+	 * they need to be re-initialized followed by gpu reset, suspend/resume,
+	 * module reload.
+	 */
+	dev_priv->num_wa_regs = 0;
+	memset(dev_priv->intel_wa_regs, 0, sizeof(dev_priv->intel_wa_regs));
+
+	ret = intel_ring_begin(ring, 12);
+	if (ret)
+		return ret;
+
+	/* WaDisablePartialInstShootdown:chv */
+	intel_ring_emit_wa(ring, GEN8_ROW_CHICKEN,
+			   _MASKED_BIT_ENABLE(PARTIAL_INSTRUCTION_SHOOTDOWN_DISABLE));
+
+	/* WaDisableThreadStallDopClockGating:chv */
+	intel_ring_emit_wa(ring, GEN8_ROW_CHICKEN,
+			   _MASKED_BIT_ENABLE(STALL_DOP_GATING_DISABLE));
+
+	/* WaDisableDopClockGating:chv (pre-production hw) */
+	intel_ring_emit_wa(ring, GEN7_ROW_CHICKEN2,
+			   _MASKED_BIT_ENABLE(DOP_CLOCK_GATING_DISABLE));
+
+	/* WaDisableSamplerPowerBypass:chv (pre-production hw) */
+	intel_ring_emit_wa(ring, HALF_SLICE_CHICKEN3,
+			   _MASKED_BIT_ENABLE(GEN8_SAMPLER_POWER_BYPASS_DIS));
+
+	intel_ring_advance(ring);
+
+	return 0;
+}
+
 static int init_render_ring(struct intel_engine_cs *ring)
 {
 	struct drm_device *dev = ring->dev;
@@ -651,7 +828,7 @@ static int init_render_ring(struct intel_engine_cs *ring)
 			   _MASKED_BIT_ENABLE(GFX_REPLAY_MODE));
 
 	if (INTEL_INFO(dev)->gen >= 5) {
-		ret = init_pipe_control(ring);
+		ret = intel_init_pipe_control(ring);
 		if (ret)
 			return ret;
 	}
@@ -686,16 +863,7 @@ static void render_ring_cleanup(struct intel_engine_cs *ring)
 		dev_priv->semaphore_obj = NULL;
 	}
 
-	if (ring->scratch.obj == NULL)
-		return;
-
-	if (INTEL_INFO(dev)->gen >= 5) {
-		kunmap(sg_page(ring->scratch.obj->pages->sgl));
-		i915_gem_object_ggtt_unpin(ring->scratch.obj);
-	}
-
-	drm_gem_object_unreference(&ring->scratch.obj->base);
-	ring->scratch.obj = NULL;
+	intel_fini_pipe_control(ring);
 }
 
 static int gen8_rcs_signal(struct intel_engine_cs *signaller,
@@ -1526,7 +1694,7 @@ static int init_phys_status_page(struct intel_engine_cs *ring)
 	return 0;
 }
 
-static void intel_destroy_ringbuffer_obj(struct intel_ringbuffer *ringbuf)
+void intel_destroy_ringbuffer_obj(struct intel_ringbuffer *ringbuf)
 {
 	if (!ringbuf->obj)
 		return;
@@ -1537,8 +1705,8 @@ static void intel_destroy_ringbuffer_obj(struct intel_ringbuffer *ringbuf)
 	ringbuf->obj = NULL;
 }
 
-static int intel_alloc_ringbuffer_obj(struct drm_device *dev,
-				      struct intel_ringbuffer *ringbuf)
+int intel_alloc_ringbuffer_obj(struct drm_device *dev,
+			       struct intel_ringbuffer *ringbuf)
 {
 	struct drm_i915_private *dev_priv = to_i915(dev);
 	struct drm_i915_gem_object *obj;
@@ -1600,7 +1768,9 @@ static int intel_init_ring_buffer(struct drm_device *dev,
 	ring->dev = dev;
 	INIT_LIST_HEAD(&ring->active_list);
 	INIT_LIST_HEAD(&ring->request_list);
+	INIT_LIST_HEAD(&ring->execlist_queue);
 	ringbuf->size = 32 * PAGE_SIZE;
+	ringbuf->ring = ring;
 	memset(ring->semaphore.sync_seqno, 0, sizeof(ring->semaphore.sync_seqno));
 
 	init_waitqueue_head(&ring->irq_queue);
@@ -1683,13 +1853,14 @@ static int intel_ring_wait_request(struct intel_engine_cs *ring, int n)
 		ringbuf->head = ringbuf->last_retired_head;
 		ringbuf->last_retired_head = -1;
 
-		ringbuf->space = ring_space(ringbuf);
+		ringbuf->space = intel_ring_space(ringbuf);
 		if (ringbuf->space >= n)
 			return 0;
 	}
 
 	list_for_each_entry(request, &ring->request_list, list) {
-		if (__ring_space(request->tail, ringbuf->tail, ringbuf->size) >= n) {
+		if (__intel_ring_space(request->tail, ringbuf->tail,
+				       ringbuf->size) >= n) {
 			seqno = request->seqno;
 			break;
 		}
@@ -1706,7 +1877,7 @@ static int intel_ring_wait_request(struct intel_engine_cs *ring, int n)
 	ringbuf->head = ringbuf->last_retired_head;
 	ringbuf->last_retired_head = -1;
 
-	ringbuf->space = ring_space(ringbuf);
+	ringbuf->space = intel_ring_space(ringbuf);
 	return 0;
 }
 
@@ -1735,7 +1906,7 @@ static int ring_wait_for_space(struct intel_engine_cs *ring, int n)
 	trace_i915_ring_wait_begin(ring);
 	do {
 		ringbuf->head = I915_READ_HEAD(ring);
-		ringbuf->space = ring_space(ringbuf);
+		ringbuf->space = intel_ring_space(ringbuf);
 		if (ringbuf->space >= n) {
 			ret = 0;
 			break;
@@ -1787,7 +1958,7 @@ static int intel_wrap_ring_buffer(struct intel_engine_cs *ring)
 		iowrite32(MI_NOOP, virt++);
 
 	ringbuf->tail = 0;
-	ringbuf->space = ring_space(ringbuf);
+	ringbuf->space = intel_ring_space(ringbuf);
 
 	return 0;
 }
@@ -1992,9 +2163,7 @@ gen8_ring_dispatch_execbuffer(struct intel_engine_cs *ring,
 			      u64 offset, u32 len,
 			      unsigned flags)
 {
-	struct drm_i915_private *dev_priv = ring->dev->dev_private;
-	bool ppgtt = dev_priv->mm.aliasing_ppgtt != NULL &&
-		!(flags & I915_DISPATCH_SECURE);
+	bool ppgtt = USES_PPGTT(ring->dev) && !(flags & I915_DISPATCH_SECURE);
 	int ret;
 
 	ret = intel_ring_begin(ring, 4);
@@ -2023,8 +2192,9 @@ hsw_ring_dispatch_execbuffer(struct intel_engine_cs *ring,
 		return ret;
 
 	intel_ring_emit(ring,
-			MI_BATCH_BUFFER_START | MI_BATCH_PPGTT_HSW |
-			(flags & I915_DISPATCH_SECURE ? 0 : MI_BATCH_NON_SECURE_HSW));
+			MI_BATCH_BUFFER_START |
+			(flags & I915_DISPATCH_SECURE ?
+			 0 : MI_BATCH_PPGTT_HSW | MI_BATCH_NON_SECURE_HSW));
 	/* bit0-7 is the length on GEN6+ */
 	intel_ring_emit(ring, offset);
 	intel_ring_advance(ring);
@@ -2123,6 +2293,10 @@ int intel_init_render_ring_buffer(struct drm_device *dev)
 					dev_priv->semaphore_obj = obj;
 			}
 		}
+		if (IS_CHERRYVIEW(dev))
+			ring->init_context = chv_init_workarounds;
+		else
+			ring->init_context = bdw_init_workarounds;
 		ring->add_request = gen6_add_request;
 		ring->flush = gen8_render_ring_flush;
 		ring->irq_get = gen8_ring_get_irq;
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
index 70525d0c2c74..96479c89f4bd 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.h
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
@@ -5,6 +5,13 @@
 
 #define I915_CMD_HASH_ORDER 9
 
+/* 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
+
 /*
  * Gen2 BSpec "1. Programming Environment" / 1.4.4.6 "Ring Buffer Use"
  * Gen3 BSpec "vol1c Memory Interface Functions" / 2.3.4.5 "Ring Buffer Use"
@@ -90,6 +97,15 @@ struct intel_ringbuffer {
 	struct drm_i915_gem_object *obj;
 	void __iomem *virtual_start;
 
+	struct intel_engine_cs *ring;
+
+	/*
+	 * FIXME: This backpointer is an artifact of the history of how the
+	 * execlist patches came into being. It will get removed once the basic
+	 * code has landed.
+	 */
+	struct intel_context *FIXME_lrc_ctx;
+
 	u32 head;
 	u32 tail;
 	int space;
@@ -132,6 +148,8 @@ struct  intel_engine_cs {
 
 	int		(*init)(struct intel_engine_cs *ring);
 
+	int		(*init_context)(struct intel_engine_cs *ring);
+
 	void		(*write_tail)(struct intel_engine_cs *ring,
 				      u32 value);
 	int __must_check (*flush)(struct intel_engine_cs *ring,
@@ -214,6 +232,18 @@ struct  intel_engine_cs {
 				  unsigned int num_dwords);
 	} semaphore;
 
+	/* Execlists */
+	spinlock_t execlist_lock;
+	struct list_head execlist_queue;
+	u8 next_context_status_buffer;
+	u32             irq_keep_mask; /* bitmask for interrupts that should not be masked */
+	int		(*emit_request)(struct intel_ringbuffer *ringbuf);
+	int		(*emit_flush)(struct intel_ringbuffer *ringbuf,
+				      u32 invalidate_domains,
+				      u32 flush_domains);
+	int		(*emit_bb_start)(struct intel_ringbuffer *ringbuf,
+					 u64 offset, unsigned flags);
+
 	/**
 	 * List of objects currently involved in rendering from the
 	 * ringbuffer.
@@ -287,11 +317,7 @@ struct  intel_engine_cs {
 	u32 (*get_cmd_length_mask)(u32 cmd_header);
 };
 
-static inline bool
-intel_ring_initialized(struct intel_engine_cs *ring)
-{
-	return ring->buffer && ring->buffer->obj;
-}
+bool intel_ring_initialized(struct intel_engine_cs *ring);
 
 static inline unsigned
 intel_ring_flag(struct intel_engine_cs *ring)
@@ -355,6 +381,10 @@ intel_write_status_page(struct intel_engine_cs *ring,
 #define I915_GEM_HWS_SCRATCH_INDEX	0x30
 #define I915_GEM_HWS_SCRATCH_ADDR (I915_GEM_HWS_SCRATCH_INDEX << MI_STORE_DWORD_INDEX_SHIFT)
 
+void intel_destroy_ringbuffer_obj(struct intel_ringbuffer *ringbuf);
+int intel_alloc_ringbuffer_obj(struct drm_device *dev,
+			       struct intel_ringbuffer *ringbuf);
+
 void intel_stop_ring_buffer(struct intel_engine_cs *ring);
 void intel_cleanup_ring_buffer(struct intel_engine_cs *ring);
 
@@ -372,6 +402,9 @@ static inline void intel_ring_advance(struct intel_engine_cs *ring)
 	struct intel_ringbuffer *ringbuf = ring->buffer;
 	ringbuf->tail &= ringbuf->size - 1;
 }
+int __intel_ring_space(int head, int tail, int size);
+int intel_ring_space(struct intel_ringbuffer *ringbuf);
+bool intel_ring_stopped(struct intel_engine_cs *ring);
 void __intel_ring_advance(struct intel_engine_cs *ring);
 
 int __must_check intel_ring_idle(struct intel_engine_cs *ring);
@@ -379,6 +412,9 @@ void intel_ring_init_seqno(struct intel_engine_cs *ring, u32 seqno);
 int intel_ring_flush_all_caches(struct intel_engine_cs *ring);
 int intel_ring_invalidate_all_caches(struct intel_engine_cs *ring);
 
+void intel_fini_pipe_control(struct intel_engine_cs *ring);
+int intel_init_pipe_control(struct intel_engine_cs *ring);
+
 int intel_init_render_ring_buffer(struct drm_device *dev);
 int intel_init_bsd_ring_buffer(struct drm_device *dev);
 int intel_init_bsd2_ring_buffer(struct drm_device *dev);
diff --git a/drivers/gpu/drm/i915/intel_sprite.c b/drivers/gpu/drm/i915/intel_sprite.c
index 168c6652cda1..07a74ef589bd 100644
--- a/drivers/gpu/drm/i915/intel_sprite.c
+++ b/drivers/gpu/drm/i915/intel_sprite.c
@@ -53,6 +53,7 @@ static bool intel_pipe_update_start(struct intel_crtc *crtc, uint32_t *start_vbl
 	enum pipe pipe = crtc->pipe;
 	long timeout = msecs_to_jiffies_timeout(1);
 	int scanline, min, max, vblank_start;
+	wait_queue_head_t *wq = drm_crtc_vblank_waitqueue(&crtc->base);
 	DEFINE_WAIT(wait);
 
 	WARN_ON(!drm_modeset_is_locked(&crtc->base.mutex));
@@ -81,7 +82,7 @@ static bool intel_pipe_update_start(struct intel_crtc *crtc, uint32_t *start_vbl
 		 * other CPUs can see the task state update by the time we
 		 * read the scanline.
 		 */
-		prepare_to_wait(&crtc->vbl_wait, &wait, TASK_UNINTERRUPTIBLE);
+		prepare_to_wait(wq, &wait, TASK_UNINTERRUPTIBLE);
 
 		scanline = intel_get_crtc_scanline(crtc);
 		if (scanline < min || scanline > max)
@@ -100,7 +101,7 @@ static bool intel_pipe_update_start(struct intel_crtc *crtc, uint32_t *start_vbl
 		local_irq_disable();
 	}
 
-	finish_wait(&crtc->vbl_wait, &wait);
+	finish_wait(wq, &wait);
 
 	drm_vblank_put(dev, pipe);
 
@@ -163,6 +164,7 @@ vlv_update_plane(struct drm_plane *dplane, struct drm_crtc *crtc,
 	sprctl &= ~SP_PIXFORMAT_MASK;
 	sprctl &= ~SP_YUV_BYTE_ORDER_MASK;
 	sprctl &= ~SP_TILED;
+	sprctl &= ~SP_ROTATE_180;
 
 	switch (fb->pixel_format) {
 	case DRM_FORMAT_YUYV:
@@ -235,6 +237,14 @@ vlv_update_plane(struct drm_plane *dplane, struct drm_crtc *crtc,
 							fb->pitches[0]);
 	linear_offset -= sprsurf_offset;
 
+	if (intel_plane->rotation == BIT(DRM_ROTATE_180)) {
+		sprctl |= SP_ROTATE_180;
+
+		x += src_w;
+		y += src_h;
+		linear_offset += src_h * fb->pitches[0] + src_w * pixel_size;
+	}
+
 	atomic_update = intel_pipe_update_start(intel_crtc, &start_vbl_count);
 
 	intel_update_primary_plane(intel_crtc);
@@ -364,6 +374,7 @@ ivb_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 	sprctl &= ~SPRITE_RGB_ORDER_RGBX;
 	sprctl &= ~SPRITE_YUV_BYTE_ORDER_MASK;
 	sprctl &= ~SPRITE_TILED;
+	sprctl &= ~SPRITE_ROTATE_180;
 
 	switch (fb->pixel_format) {
 	case DRM_FORMAT_XBGR8888:
@@ -426,6 +437,18 @@ ivb_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 					       pixel_size, fb->pitches[0]);
 	linear_offset -= sprsurf_offset;
 
+	if (intel_plane->rotation == BIT(DRM_ROTATE_180)) {
+		sprctl |= SPRITE_ROTATE_180;
+
+		/* HSW and BDW does this automagically in hardware */
+		if (!IS_HASWELL(dev) && !IS_BROADWELL(dev)) {
+			x += src_w;
+			y += src_h;
+			linear_offset += src_h * fb->pitches[0] +
+				src_w * pixel_size;
+		}
+	}
+
 	atomic_update = intel_pipe_update_start(intel_crtc, &start_vbl_count);
 
 	intel_update_primary_plane(intel_crtc);
@@ -571,6 +594,7 @@ ilk_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 	dvscntr &= ~DVS_RGB_ORDER_XBGR;
 	dvscntr &= ~DVS_YUV_BYTE_ORDER_MASK;
 	dvscntr &= ~DVS_TILED;
+	dvscntr &= ~DVS_ROTATE_180;
 
 	switch (fb->pixel_format) {
 	case DRM_FORMAT_XBGR8888:
@@ -628,6 +652,14 @@ ilk_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 					       pixel_size, fb->pitches[0]);
 	linear_offset -= dvssurf_offset;
 
+	if (intel_plane->rotation == BIT(DRM_ROTATE_180)) {
+		dvscntr |= DVS_ROTATE_180;
+
+		x += src_w;
+		y += src_h;
+		linear_offset += src_h * fb->pitches[0] + src_w * pixel_size;
+	}
+
 	atomic_update = intel_pipe_update_start(intel_crtc, &start_vbl_count);
 
 	intel_update_primary_plane(intel_crtc);
@@ -895,6 +927,9 @@ intel_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 	max_scale = intel_plane->max_downscale << 16;
 	min_scale = intel_plane->can_scale ? 1 : (1 << 16);
 
+	drm_rect_rotate(&src, fb->width << 16, fb->height << 16,
+			intel_plane->rotation);
+
 	hscale = drm_rect_calc_hscale_relaxed(&src, &dst, min_scale, max_scale);
 	BUG_ON(hscale < 0);
 
@@ -933,6 +968,9 @@ intel_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 				     drm_rect_width(&dst) * hscale - drm_rect_width(&src),
 				     drm_rect_height(&dst) * vscale - drm_rect_height(&src));
 
+		drm_rect_rotate_inv(&src, fb->width << 16, fb->height << 16,
+				    intel_plane->rotation);
+
 		/* sanity check to make sure the src viewport wasn't enlarged */
 		WARN_ON(src.x1 < (int) src_x ||
 			src.y1 < (int) src_y ||
@@ -1180,18 +1218,45 @@ out_unlock:
 	return ret;
 }
 
-void intel_plane_restore(struct drm_plane *plane)
+int intel_plane_set_property(struct drm_plane *plane,
+			     struct drm_property *prop,
+			     uint64_t val)
+{
+	struct drm_device *dev = plane->dev;
+	struct intel_plane *intel_plane = to_intel_plane(plane);
+	uint64_t old_val;
+	int ret = -ENOENT;
+
+	if (prop == dev->mode_config.rotation_property) {
+		/* exactly one rotation angle please */
+		if (hweight32(val & 0xf) != 1)
+			return -EINVAL;
+
+		if (intel_plane->rotation == val)
+			return 0;
+
+		old_val = intel_plane->rotation;
+		intel_plane->rotation = val;
+		ret = intel_plane_restore(plane);
+		if (ret)
+			intel_plane->rotation = old_val;
+	}
+
+	return ret;
+}
+
+int intel_plane_restore(struct drm_plane *plane)
 {
 	struct intel_plane *intel_plane = to_intel_plane(plane);
 
 	if (!plane->crtc || !plane->fb)
-		return;
+		return 0;
 
-	intel_update_plane(plane, plane->crtc, plane->fb,
-			   intel_plane->crtc_x, intel_plane->crtc_y,
-			   intel_plane->crtc_w, intel_plane->crtc_h,
-			   intel_plane->src_x, intel_plane->src_y,
-			   intel_plane->src_w, intel_plane->src_h);
+	return plane->funcs->update_plane(plane, plane->crtc, plane->fb,
+				  intel_plane->crtc_x, intel_plane->crtc_y,
+				  intel_plane->crtc_w, intel_plane->crtc_h,
+				  intel_plane->src_x, intel_plane->src_y,
+				  intel_plane->src_w, intel_plane->src_h);
 }
 
 void intel_plane_disable(struct drm_plane *plane)
@@ -1206,6 +1271,7 @@ static const struct drm_plane_funcs intel_plane_funcs = {
 	.update_plane = intel_update_plane,
 	.disable_plane = intel_disable_plane,
 	.destroy = intel_destroy_plane,
+	.set_property = intel_plane_set_property,
 };
 
 static uint32_t ilk_plane_formats[] = {
@@ -1310,13 +1376,28 @@ intel_plane_init(struct drm_device *dev, enum pipe pipe, int plane)
 
 	intel_plane->pipe = pipe;
 	intel_plane->plane = plane;
+	intel_plane->rotation = BIT(DRM_ROTATE_0);
 	possible_crtcs = (1 << pipe);
-	ret = drm_plane_init(dev, &intel_plane->base, possible_crtcs,
-			     &intel_plane_funcs,
-			     plane_formats, num_plane_formats,
-			     false);
-	if (ret)
+	ret = drm_universal_plane_init(dev, &intel_plane->base, possible_crtcs,
+				       &intel_plane_funcs,
+				       plane_formats, num_plane_formats,
+				       DRM_PLANE_TYPE_OVERLAY);
+	if (ret) {
 		kfree(intel_plane);
+		goto out;
+	}
+
+	if (!dev->mode_config.rotation_property)
+		dev->mode_config.rotation_property =
+			drm_mode_create_rotation_property(dev,
+							  BIT(DRM_ROTATE_0) |
+							  BIT(DRM_ROTATE_180));
+
+	if (dev->mode_config.rotation_property)
+		drm_object_attach_property(&intel_plane->base.base,
+					   dev->mode_config.rotation_property,
+					   intel_plane->rotation);
 
+ out:
 	return ret;
 }
diff --git a/drivers/gpu/drm/i915/intel_uncore.c b/drivers/gpu/drm/i915/intel_uncore.c
index e81bc3bdc533..918b76163965 100644
--- a/drivers/gpu/drm/i915/intel_uncore.c
+++ b/drivers/gpu/drm/i915/intel_uncore.c
@@ -101,7 +101,7 @@ static void __gen7_gt_force_wake_mt_get(struct drm_i915_private *dev_priv,
 {
 	u32 forcewake_ack;
 
-	if (IS_HASWELL(dev_priv->dev) || IS_GEN8(dev_priv->dev))
+	if (IS_HASWELL(dev_priv->dev) || IS_BROADWELL(dev_priv->dev))
 		forcewake_ack = FORCEWAKE_ACK_HSW;
 	else
 		forcewake_ack = FORCEWAKE_MT_ACK;
@@ -334,7 +334,7 @@ void intel_uncore_forcewake_reset(struct drm_device *dev, bool restore)
 	else if (IS_GEN6(dev) || IS_GEN7(dev))
 		__gen6_gt_force_wake_reset(dev_priv);
 
-	if (IS_IVYBRIDGE(dev) || IS_HASWELL(dev) || IS_GEN8(dev))
+	if (IS_IVYBRIDGE(dev) || IS_HASWELL(dev) || IS_BROADWELL(dev))
 		__gen7_gt_force_wake_mt_reset(dev_priv);
 
 	if (restore) { /* If reset with a user forcewake, try to restore */
@@ -838,7 +838,7 @@ void intel_uncore_init(struct drm_device *dev)
 	if (IS_VALLEYVIEW(dev)) {
 		dev_priv->uncore.funcs.force_wake_get = __vlv_force_wake_get;
 		dev_priv->uncore.funcs.force_wake_put = __vlv_force_wake_put;
-	} else if (IS_HASWELL(dev) || IS_GEN8(dev)) {
+	} else if (IS_HASWELL(dev) || IS_BROADWELL(dev)) {
 		dev_priv->uncore.funcs.force_wake_get = __gen7_gt_force_wake_mt_get;
 		dev_priv->uncore.funcs.force_wake_put = __gen7_gt_force_wake_mt_put;
 	} else if (IS_IVYBRIDGE(dev)) {
diff --git a/drivers/gpu/drm/mga/mga_dma.c b/drivers/gpu/drm/mga/mga_dma.c
index c3bf059ba720..8cfa9cb74c86 100644
--- a/drivers/gpu/drm/mga/mga_dma.c
+++ b/drivers/gpu/drm/mga/mga_dma.c
@@ -502,31 +502,31 @@ static int mga_do_agp_dma_bootstrap(struct drm_device *dev,
 		return err;
 	}
 
-	/* Make drm_addbufs happy by not trying to create a mapping for less
-	 * than a page.
+	/* Make drm_legacy_addbufs happy by not trying to create a mapping for
+	 * less than a page.
 	 */
 	if (warp_size < PAGE_SIZE)
 		warp_size = PAGE_SIZE;
 
 	offset = 0;
-	err = drm_addmap(dev, offset, warp_size,
-			 _DRM_AGP, _DRM_READ_ONLY, &dev_priv->warp);
+	err = drm_legacy_addmap(dev, offset, warp_size,
+				_DRM_AGP, _DRM_READ_ONLY, &dev_priv->warp);
 	if (err) {
 		DRM_ERROR("Unable to map WARP microcode: %d\n", err);
 		return err;
 	}
 
 	offset += warp_size;
-	err = drm_addmap(dev, offset, dma_bs->primary_size,
-			 _DRM_AGP, _DRM_READ_ONLY, &dev_priv->primary);
+	err = drm_legacy_addmap(dev, offset, dma_bs->primary_size,
+				_DRM_AGP, _DRM_READ_ONLY, &dev_priv->primary);
 	if (err) {
 		DRM_ERROR("Unable to map primary DMA region: %d\n", err);
 		return err;
 	}
 
 	offset += dma_bs->primary_size;
-	err = drm_addmap(dev, offset, secondary_size,
-			 _DRM_AGP, 0, &dev->agp_buffer_map);
+	err = drm_legacy_addmap(dev, offset, secondary_size,
+				_DRM_AGP, 0, &dev->agp_buffer_map);
 	if (err) {
 		DRM_ERROR("Unable to map secondary DMA region: %d\n", err);
 		return err;
@@ -538,7 +538,7 @@ static int mga_do_agp_dma_bootstrap(struct drm_device *dev,
 	req.flags = _DRM_AGP_BUFFER;
 	req.agp_start = offset;
 
-	err = drm_addbufs_agp(dev, &req);
+	err = drm_legacy_addbufs_agp(dev, &req);
 	if (err) {
 		DRM_ERROR("Unable to add secondary DMA buffers: %d\n", err);
 		return err;
@@ -559,16 +559,16 @@ static int mga_do_agp_dma_bootstrap(struct drm_device *dev,
 	}
 
 	offset += secondary_size;
-	err = drm_addmap(dev, offset, agp_size - offset,
-			 _DRM_AGP, 0, &dev_priv->agp_textures);
+	err = drm_legacy_addmap(dev, offset, agp_size - offset,
+				_DRM_AGP, 0, &dev_priv->agp_textures);
 	if (err) {
 		DRM_ERROR("Unable to map AGP texture region %d\n", err);
 		return err;
 	}
 
-	drm_core_ioremap(dev_priv->warp, dev);
-	drm_core_ioremap(dev_priv->primary, dev);
-	drm_core_ioremap(dev->agp_buffer_map, dev);
+	drm_legacy_ioremap(dev_priv->warp, dev);
+	drm_legacy_ioremap(dev_priv->primary, dev);
+	drm_legacy_ioremap(dev->agp_buffer_map, dev);
 
 	if (!dev_priv->warp->handle ||
 	    !dev_priv->primary->handle || !dev->agp_buffer_map->handle) {
@@ -602,7 +602,7 @@ static int mga_do_agp_dma_bootstrap(struct drm_device *dev,
  *
  * \todo
  * Determine whether the maximum address passed to drm_pci_alloc is correct.
- * The same goes for drm_addbufs_pci.
+ * The same goes for drm_legacy_addbufs_pci.
  *
  * \sa mga_do_dma_bootstrap, mga_do_agp_dma_bootstrap
  */
@@ -622,15 +622,15 @@ static int mga_do_pci_dma_bootstrap(struct drm_device *dev,
 		return -EFAULT;
 	}
 
-	/* Make drm_addbufs happy by not trying to create a mapping for less
-	 * than a page.
+	/* Make drm_legacy_addbufs happy by not trying to create a mapping for
+	 * less than a page.
 	 */
 	if (warp_size < PAGE_SIZE)
 		warp_size = PAGE_SIZE;
 
 	/* The proper alignment is 0x100 for this mapping */
-	err = drm_addmap(dev, 0, warp_size, _DRM_CONSISTENT,
-			 _DRM_READ_ONLY, &dev_priv->warp);
+	err = drm_legacy_addmap(dev, 0, warp_size, _DRM_CONSISTENT,
+				_DRM_READ_ONLY, &dev_priv->warp);
 	if (err != 0) {
 		DRM_ERROR("Unable to create mapping for WARP microcode: %d\n",
 			  err);
@@ -645,8 +645,8 @@ static int mga_do_pci_dma_bootstrap(struct drm_device *dev,
 	for (primary_size = dma_bs->primary_size; primary_size != 0;
 	     primary_size >>= 1) {
 		/* The proper alignment for this mapping is 0x04 */
-		err = drm_addmap(dev, 0, primary_size, _DRM_CONSISTENT,
-				 _DRM_READ_ONLY, &dev_priv->primary);
+		err = drm_legacy_addmap(dev, 0, primary_size, _DRM_CONSISTENT,
+					_DRM_READ_ONLY, &dev_priv->primary);
 		if (!err)
 			break;
 	}
@@ -669,7 +669,7 @@ static int mga_do_pci_dma_bootstrap(struct drm_device *dev,
 		req.count = bin_count;
 		req.size = dma_bs->secondary_bin_size;
 
-		err = drm_addbufs_pci(dev, &req);
+		err = drm_legacy_addbufs_pci(dev, &req);
 		if (!err)
 			break;
 	}
@@ -708,15 +708,16 @@ static int mga_do_dma_bootstrap(struct drm_device *dev,
 	/* The first steps are the same for both PCI and AGP based DMA.  Map
 	 * the cards MMIO registers and map a status page.
 	 */
-	err = drm_addmap(dev, dev_priv->mmio_base, dev_priv->mmio_size,
-			 _DRM_REGISTERS, _DRM_READ_ONLY, &dev_priv->mmio);
+	err = drm_legacy_addmap(dev, dev_priv->mmio_base, dev_priv->mmio_size,
+				_DRM_REGISTERS, _DRM_READ_ONLY,
+				&dev_priv->mmio);
 	if (err) {
 		DRM_ERROR("Unable to map MMIO region: %d\n", err);
 		return err;
 	}
 
-	err = drm_addmap(dev, 0, SAREA_MAX, _DRM_SHM,
-			 _DRM_READ_ONLY | _DRM_LOCKED | _DRM_KERNEL,
+	err = drm_legacy_addmap(dev, 0, SAREA_MAX, _DRM_SHM,
+				_DRM_READ_ONLY | _DRM_LOCKED | _DRM_KERNEL,
 			 &dev_priv->status);
 	if (err) {
 		DRM_ERROR("Unable to map status region: %d\n", err);
@@ -809,7 +810,7 @@ static int mga_do_init_dma(struct drm_device *dev, drm_mga_init_t *init)
 	dev_priv->texture_offset = init->texture_offset[0];
 	dev_priv->texture_size = init->texture_size[0];
 
-	dev_priv->sarea = drm_getsarea(dev);
+	dev_priv->sarea = drm_legacy_getsarea(dev);
 	if (!dev_priv->sarea) {
 		DRM_ERROR("failed to find sarea!\n");
 		return -EINVAL;
@@ -820,37 +821,37 @@ static int mga_do_init_dma(struct drm_device *dev, drm_mga_init_t *init)
 		dev_priv->dma_access = MGA_PAGPXFER;
 		dev_priv->wagp_enable = MGA_WAGP_ENABLE;
 
-		dev_priv->status = drm_core_findmap(dev, init->status_offset);
+		dev_priv->status = drm_legacy_findmap(dev, init->status_offset);
 		if (!dev_priv->status) {
 			DRM_ERROR("failed to find status page!\n");
 			return -EINVAL;
 		}
-		dev_priv->mmio = drm_core_findmap(dev, init->mmio_offset);
+		dev_priv->mmio = drm_legacy_findmap(dev, init->mmio_offset);
 		if (!dev_priv->mmio) {
 			DRM_ERROR("failed to find mmio region!\n");
 			return -EINVAL;
 		}
-		dev_priv->warp = drm_core_findmap(dev, init->warp_offset);
+		dev_priv->warp = drm_legacy_findmap(dev, init->warp_offset);
 		if (!dev_priv->warp) {
 			DRM_ERROR("failed to find warp microcode region!\n");
 			return -EINVAL;
 		}
-		dev_priv->primary = drm_core_findmap(dev, init->primary_offset);
+		dev_priv->primary = drm_legacy_findmap(dev, init->primary_offset);
 		if (!dev_priv->primary) {
 			DRM_ERROR("failed to find primary dma region!\n");
 			return -EINVAL;
 		}
 		dev->agp_buffer_token = init->buffers_offset;
 		dev->agp_buffer_map =
-		    drm_core_findmap(dev, init->buffers_offset);
+		    drm_legacy_findmap(dev, init->buffers_offset);
 		if (!dev->agp_buffer_map) {
 			DRM_ERROR("failed to find dma buffer region!\n");
 			return -EINVAL;
 		}
 
-		drm_core_ioremap(dev_priv->warp, dev);
-		drm_core_ioremap(dev_priv->primary, dev);
-		drm_core_ioremap(dev->agp_buffer_map, dev);
+		drm_legacy_ioremap(dev_priv->warp, dev);
+		drm_legacy_ioremap(dev_priv->primary, dev);
+		drm_legacy_ioremap(dev->agp_buffer_map, dev);
 	}
 
 	dev_priv->sarea_priv =
@@ -936,14 +937,14 @@ static int mga_do_cleanup_dma(struct drm_device *dev, int full_cleanup)
 
 		if ((dev_priv->warp != NULL)
 		    && (dev_priv->warp->type != _DRM_CONSISTENT))
-			drm_core_ioremapfree(dev_priv->warp, dev);
+			drm_legacy_ioremapfree(dev_priv->warp, dev);
 
 		if ((dev_priv->primary != NULL)
 		    && (dev_priv->primary->type != _DRM_CONSISTENT))
-			drm_core_ioremapfree(dev_priv->primary, dev);
+			drm_legacy_ioremapfree(dev_priv->primary, dev);
 
 		if (dev->agp_buffer_map != NULL)
-			drm_core_ioremapfree(dev->agp_buffer_map, dev);
+			drm_legacy_ioremapfree(dev->agp_buffer_map, dev);
 
 		if (dev_priv->used_new_dma_init) {
 #if __OS_HAS_AGP
diff --git a/drivers/gpu/drm/mga/mga_drv.c b/drivers/gpu/drm/mga/mga_drv.c
index 6b1a87c8aac5..5e2f131a6a72 100644
--- a/drivers/gpu/drm/mga/mga_drv.c
+++ b/drivers/gpu/drm/mga/mga_drv.c
@@ -48,7 +48,7 @@ static const struct file_operations mga_driver_fops = {
 	.open = drm_open,
 	.release = drm_release,
 	.unlocked_ioctl = drm_ioctl,
-	.mmap = drm_mmap,
+	.mmap = drm_legacy_mmap,
 	.poll = drm_poll,
 #ifdef CONFIG_COMPAT
 	.compat_ioctl = mga_compat_ioctl,
@@ -64,6 +64,7 @@ static struct drm_driver driver = {
 	.load = mga_driver_load,
 	.unload = mga_driver_unload,
 	.lastclose = mga_driver_lastclose,
+	.set_busid = drm_pci_set_busid,
 	.dma_quiescent = mga_driver_dma_quiescent,
 	.device_is_agp = mga_driver_device_is_agp,
 	.get_vblank_counter = mga_get_vblank_counter,
diff --git a/drivers/gpu/drm/mga/mga_drv.h b/drivers/gpu/drm/mga/mga_drv.h
index fe453213600a..b4a2014917e5 100644
--- a/drivers/gpu/drm/mga/mga_drv.h
+++ b/drivers/gpu/drm/mga/mga_drv.h
@@ -31,6 +31,8 @@
 #ifndef __MGA_DRV_H__
 #define __MGA_DRV_H__
 
+#include <drm/drm_legacy.h>
+
 /* General customization:
  */
 
diff --git a/drivers/gpu/drm/mgag200/mgag200_drv.c b/drivers/gpu/drm/mgag200/mgag200_drv.c
index 2d75d6df0789..97745991544d 100644
--- a/drivers/gpu/drm/mgag200/mgag200_drv.c
+++ b/drivers/gpu/drm/mgag200/mgag200_drv.c
@@ -91,6 +91,7 @@ static struct drm_driver driver = {
 	.driver_features = DRIVER_GEM | DRIVER_MODESET,
 	.load = mgag200_driver_load,
 	.unload = mgag200_driver_unload,
+	.set_busid = drm_pci_set_busid,
 	.fops = &mgag200_driver_fops,
 	.name = DRIVER_NAME,
 	.desc = DRIVER_DESC,
diff --git a/drivers/gpu/drm/mgag200/mgag200_drv.h b/drivers/gpu/drm/mgag200/mgag200_drv.h
index 80de23d9b9c9..e9eea1d4e7c3 100644
--- a/drivers/gpu/drm/mgag200/mgag200_drv.h
+++ b/drivers/gpu/drm/mgag200/mgag200_drv.h
@@ -22,6 +22,8 @@
 #include <drm/ttm/ttm_memory.h>
 #include <drm/ttm/ttm_module.h>
 
+#include <drm/drm_gem.h>
+
 #include <linux/i2c.h>
 #include <linux/i2c-algo-bit.h>
 
@@ -190,8 +192,6 @@ struct mga_device {
 	resource_size_t			rmmio_size;
 	void __iomem			*rmmio;
 
-	drm_local_map_t			*framebuffer;
-
 	struct mga_mc			mc;
 	struct mga_mode_info		mode_info;
 
@@ -224,7 +224,7 @@ struct mgag200_bo {
 	struct ttm_placement placement;
 	struct ttm_bo_kmap_obj kmap;
 	struct drm_gem_object gem;
-	u32 placements[3];
+	struct ttm_place placements[3];
 	int pin_count;
 };
 #define gem_to_mga_bo(gobj) container_of((gobj), struct mgag200_bo, gem)
diff --git a/drivers/gpu/drm/mgag200/mgag200_fb.c b/drivers/gpu/drm/mgag200/mgag200_fb.c
index 5451dc58eff1..4415af3666ab 100644
--- a/drivers/gpu/drm/mgag200/mgag200_fb.c
+++ b/drivers/gpu/drm/mgag200/mgag200_fb.c
@@ -158,7 +158,8 @@ static int mgag200fb_create_object(struct mga_fbdev *afbdev,
 static int mgag200fb_create(struct drm_fb_helper *helper,
 			   struct drm_fb_helper_surface_size *sizes)
 {
-	struct mga_fbdev *mfbdev = (struct mga_fbdev *)helper;
+	struct mga_fbdev *mfbdev =
+		container_of(helper, struct mga_fbdev, helper);
 	struct drm_device *dev = mfbdev->helper.dev;
 	struct drm_mode_fb_cmd2 mode_cmd;
 	struct mga_device *mdev = dev->dev_private;
diff --git a/drivers/gpu/drm/mgag200/mgag200_mode.c b/drivers/gpu/drm/mgag200/mgag200_mode.c
index 45f04dea0ac2..83485ab81ce8 100644
--- a/drivers/gpu/drm/mgag200/mgag200_mode.c
+++ b/drivers/gpu/drm/mgag200/mgag200_mode.c
@@ -1483,11 +1483,7 @@ static int mga_vga_mode_valid(struct drm_connector *connector,
 {
 	struct drm_device *dev = connector->dev;
 	struct mga_device *mdev = (struct mga_device*)dev->dev_private;
-	struct mga_fbdev *mfbdev = mdev->mfbdev;
-	struct drm_fb_helper *fb_helper = &mfbdev->helper;
-	struct drm_fb_helper_connector *fb_helper_conn = NULL;
 	int bpp = 32;
-	int i = 0;
 
 	if (IS_G200_SE(mdev)) {
 		if (mdev->unique_rev_id == 0x01) {
@@ -1537,21 +1533,14 @@ static int mga_vga_mode_valid(struct drm_connector *connector,
 	}
 
 	/* Validate the mode input by the user */
-	for (i = 0; i < fb_helper->connector_count; i++) {
-		if (fb_helper->connector_info[i]->connector == connector) {
-			/* Found the helper for this connector */
-			fb_helper_conn = fb_helper->connector_info[i];
-			if (fb_helper_conn->cmdline_mode.specified) {
-				if (fb_helper_conn->cmdline_mode.bpp_specified) {
-					bpp = fb_helper_conn->cmdline_mode.bpp;
-				}
-			}
-		}
+	if (connector->cmdline_mode.specified) {
+		if (connector->cmdline_mode.bpp_specified)
+			bpp = connector->cmdline_mode.bpp;
 	}
 
 	if ((mode->hdisplay * mode->vdisplay * (bpp/8)) > mdev->mc.vram_size) {
-		if (fb_helper_conn)
-			fb_helper_conn->cmdline_mode.specified = false;
+		if (connector->cmdline_mode.specified)
+			connector->cmdline_mode.specified = false;
 		return MODE_BAD;
 	}
 
diff --git a/drivers/gpu/drm/mgag200/mgag200_ttm.c b/drivers/gpu/drm/mgag200/mgag200_ttm.c
index 5a00e90696de..d16964ea0ed4 100644
--- a/drivers/gpu/drm/mgag200/mgag200_ttm.c
+++ b/drivers/gpu/drm/mgag200/mgag200_ttm.c
@@ -293,18 +293,22 @@ void mgag200_mm_fini(struct mga_device *mdev)
 void mgag200_ttm_placement(struct mgag200_bo *bo, int domain)
 {
 	u32 c = 0;
-	bo->placement.fpfn = 0;
-	bo->placement.lpfn = 0;
+	unsigned i;
+
 	bo->placement.placement = bo->placements;
 	bo->placement.busy_placement = bo->placements;
 	if (domain & TTM_PL_FLAG_VRAM)
-		bo->placements[c++] = TTM_PL_FLAG_WC | TTM_PL_FLAG_UNCACHED | TTM_PL_FLAG_VRAM;
+		bo->placements[c++].flags = TTM_PL_FLAG_WC | TTM_PL_FLAG_UNCACHED | TTM_PL_FLAG_VRAM;
 	if (domain & TTM_PL_FLAG_SYSTEM)
-		bo->placements[c++] = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
+		bo->placements[c++].flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
 	if (!c)
-		bo->placements[c++] = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
+		bo->placements[c++].flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
 	bo->placement.num_placement = c;
 	bo->placement.num_busy_placement = c;
+	for (i = 0; i < c; ++i) {
+		bo->placements[i].fpfn = 0;
+		bo->placements[i].lpfn = 0;
+	}
 }
 
 int mgag200_bo_create(struct drm_device *dev, int size, int align,
@@ -335,7 +339,7 @@ int mgag200_bo_create(struct drm_device *dev, int size, int align,
 	ret = ttm_bo_init(&mdev->ttm.bdev, &mgabo->bo, size,
 			  ttm_bo_type_device, &mgabo->placement,
 			  align >> PAGE_SHIFT, false, NULL, acc_size,
-			  NULL, mgag200_bo_ttm_destroy);
+			  NULL, NULL, mgag200_bo_ttm_destroy);
 	if (ret)
 		return ret;
 
@@ -361,7 +365,7 @@ int mgag200_bo_pin(struct mgag200_bo *bo, u32 pl_flag, u64 *gpu_addr)
 
 	mgag200_ttm_placement(bo, pl_flag);
 	for (i = 0; i < bo->placement.num_placement; i++)
-		bo->placements[i] |= TTM_PL_FLAG_NO_EVICT;
+		bo->placements[i].flags |= TTM_PL_FLAG_NO_EVICT;
 	ret = ttm_bo_validate(&bo->bo, &bo->placement, false, false);
 	if (ret)
 		return ret;
@@ -384,7 +388,7 @@ int mgag200_bo_unpin(struct mgag200_bo *bo)
 		return 0;
 
 	for (i = 0; i < bo->placement.num_placement ; i++)
-		bo->placements[i] &= ~TTM_PL_FLAG_NO_EVICT;
+		bo->placements[i].flags &= ~TTM_PL_FLAG_NO_EVICT;
 	ret = ttm_bo_validate(&bo->bo, &bo->placement, false, false);
 	if (ret)
 		return ret;
@@ -408,7 +412,7 @@ int mgag200_bo_push_sysram(struct mgag200_bo *bo)
 
 	mgag200_ttm_placement(bo, TTM_PL_FLAG_SYSTEM);
 	for (i = 0; i < bo->placement.num_placement ; i++)
-		bo->placements[i] |= TTM_PL_FLAG_NO_EVICT;
+		bo->placements[i].flags |= TTM_PL_FLAG_NO_EVICT;
 
 	ret = ttm_bo_validate(&bo->bo, &bo->placement, false, false);
 	if (ret) {
@@ -424,7 +428,7 @@ int mgag200_mmap(struct file *filp, struct vm_area_struct *vma)
 	struct mga_device *mdev;
 
 	if (unlikely(vma->vm_pgoff < DRM_FILE_PAGE_OFFSET))
-		return drm_mmap(filp, vma);
+		return -EINVAL;
 
 	file_priv = filp->private_data;
 	mdev = file_priv->minor->dev->dev_private;
diff --git a/drivers/gpu/drm/msm/Kconfig b/drivers/gpu/drm/msm/Kconfig
index c99c50de3226..9d907c526c94 100644
--- a/drivers/gpu/drm/msm/Kconfig
+++ b/drivers/gpu/drm/msm/Kconfig
@@ -4,6 +4,7 @@ config DRM_MSM
 	depends on DRM
 	depends on ARCH_QCOM || (ARM && COMPILE_TEST)
 	select DRM_KMS_HELPER
+	select DRM_PANEL
 	select SHMEM
 	select TMPFS
 	default y
diff --git a/drivers/gpu/drm/msm/Makefile b/drivers/gpu/drm/msm/Makefile
index 93ca49c8df44..6283dcb96af5 100644
--- a/drivers/gpu/drm/msm/Makefile
+++ b/drivers/gpu/drm/msm/Makefile
@@ -4,6 +4,7 @@ ifeq (, $(findstring -W,$(EXTRA_CFLAGS)))
 endif
 
 msm-y := \
+	adreno/adreno_device.o \
 	adreno/adreno_gpu.o \
 	adreno/a3xx_gpu.o \
 	hdmi/hdmi.o \
@@ -18,6 +19,8 @@ msm-y := \
 	mdp/mdp_kms.o \
 	mdp/mdp4/mdp4_crtc.o \
 	mdp/mdp4/mdp4_dtv_encoder.o \
+	mdp/mdp4/mdp4_lcdc_encoder.o \
+	mdp/mdp4/mdp4_lvds_connector.o \
 	mdp/mdp4/mdp4_irq.o \
 	mdp/mdp4/mdp4_kms.o \
 	mdp/mdp4/mdp4_plane.o \
@@ -39,5 +42,6 @@ msm-y := \
 	msm_ringbuffer.o
 
 msm-$(CONFIG_DRM_MSM_FBDEV) += msm_fbdev.o
+msm-$(CONFIG_COMMON_CLK) += mdp/mdp4/mdp4_lvds_pll.o
 
 obj-$(CONFIG_DRM_MSM)	+= msm.o
diff --git a/drivers/gpu/drm/msm/adreno/a2xx.xml.h b/drivers/gpu/drm/msm/adreno/a2xx.xml.h
index a8a144b38eaa..a3104598c27f 100644
--- a/drivers/gpu/drm/msm/adreno/a2xx.xml.h
+++ b/drivers/gpu/drm/msm/adreno/a2xx.xml.h
@@ -12,9 +12,9 @@ The rules-ng-ng source files this header was generated from are:
 - /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml  (   1453 bytes, from 2013-03-31 16:51:27)
 - /home/robclark/src/freedreno/envytools/rnndb/adreno/a2xx.xml          (  32901 bytes, from 2014-06-02 15:21:30)
 - /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_common.xml (   9859 bytes, from 2014-06-02 15:21:30)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_pm4.xml    (  14477 bytes, from 2014-05-16 11:51:57)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno/a3xx.xml          (  58020 bytes, from 2014-06-25 12:57:16)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno/a4xx.xml          (  26602 bytes, from 2014-06-25 12:57:16)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_pm4.xml    (  14960 bytes, from 2014-07-27 17:22:13)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a3xx.xml          (  58020 bytes, from 2014-08-01 12:22:48)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a4xx.xml          (  41068 bytes, from 2014-08-01 12:22:48)
 
 Copyright (C) 2013-2014 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
diff --git a/drivers/gpu/drm/msm/adreno/a3xx.xml.h b/drivers/gpu/drm/msm/adreno/a3xx.xml.h
index 303e8a9e91a5..82d015279b47 100644
--- a/drivers/gpu/drm/msm/adreno/a3xx.xml.h
+++ b/drivers/gpu/drm/msm/adreno/a3xx.xml.h
@@ -12,9 +12,9 @@ The rules-ng-ng source files this header was generated from are:
 - /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml  (   1453 bytes, from 2013-03-31 16:51:27)
 - /home/robclark/src/freedreno/envytools/rnndb/adreno/a2xx.xml          (  32901 bytes, from 2014-06-02 15:21:30)
 - /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_common.xml (   9859 bytes, from 2014-06-02 15:21:30)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_pm4.xml    (  14477 bytes, from 2014-05-16 11:51:57)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno/a3xx.xml          (  58020 bytes, from 2014-06-25 12:57:16)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno/a4xx.xml          (  26602 bytes, from 2014-06-25 12:57:16)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_pm4.xml    (  14960 bytes, from 2014-07-27 17:22:13)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a3xx.xml          (  58020 bytes, from 2014-08-01 12:22:48)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a4xx.xml          (  41068 bytes, from 2014-08-01 12:22:48)
 
 Copyright (C) 2013-2014 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
@@ -654,7 +654,7 @@ static inline uint32_t A3XX_GRAS_SU_POINT_SIZE(float val)
 #define A3XX_GRAS_SU_POLY_OFFSET_SCALE_VAL__SHIFT		0
 static inline uint32_t A3XX_GRAS_SU_POLY_OFFSET_SCALE_VAL(float val)
 {
-	return ((((uint32_t)(val * 40.0))) << A3XX_GRAS_SU_POLY_OFFSET_SCALE_VAL__SHIFT) & A3XX_GRAS_SU_POLY_OFFSET_SCALE_VAL__MASK;
+	return ((((uint32_t)(val * 28.0))) << A3XX_GRAS_SU_POLY_OFFSET_SCALE_VAL__SHIFT) & A3XX_GRAS_SU_POLY_OFFSET_SCALE_VAL__MASK;
 }
 
 #define REG_A3XX_GRAS_SU_POLY_OFFSET_OFFSET			0x0000206d
@@ -662,7 +662,7 @@ static inline uint32_t A3XX_GRAS_SU_POLY_OFFSET_SCALE_VAL(float val)
 #define A3XX_GRAS_SU_POLY_OFFSET_OFFSET__SHIFT			0
 static inline uint32_t A3XX_GRAS_SU_POLY_OFFSET_OFFSET(float val)
 {
-	return ((((uint32_t)(val * 44.0))) << A3XX_GRAS_SU_POLY_OFFSET_OFFSET__SHIFT) & A3XX_GRAS_SU_POLY_OFFSET_OFFSET__MASK;
+	return ((((uint32_t)(val * 28.0))) << A3XX_GRAS_SU_POLY_OFFSET_OFFSET__SHIFT) & A3XX_GRAS_SU_POLY_OFFSET_OFFSET__MASK;
 }
 
 #define REG_A3XX_GRAS_SU_MODE_CONTROL				0x00002070
@@ -1696,7 +1696,7 @@ static inline uint32_t A3XX_SP_VS_CTRL_REG1_CONSTFOOTPRINT(uint32_t val)
 {
 	return ((val) << A3XX_SP_VS_CTRL_REG1_CONSTFOOTPRINT__SHIFT) & A3XX_SP_VS_CTRL_REG1_CONSTFOOTPRINT__MASK;
 }
-#define A3XX_SP_VS_CTRL_REG1_INITIALOUTSTANDING__MASK		0x3f000000
+#define A3XX_SP_VS_CTRL_REG1_INITIALOUTSTANDING__MASK		0x7f000000
 #define A3XX_SP_VS_CTRL_REG1_INITIALOUTSTANDING__SHIFT		24
 static inline uint32_t A3XX_SP_VS_CTRL_REG1_INITIALOUTSTANDING(uint32_t val)
 {
diff --git a/drivers/gpu/drm/msm/adreno/a3xx_gpu.c b/drivers/gpu/drm/msm/adreno/a3xx_gpu.c
index 2773600c9488..218c5b060398 100644
--- a/drivers/gpu/drm/msm/adreno/a3xx_gpu.c
+++ b/drivers/gpu/drm/msm/adreno/a3xx_gpu.c
@@ -35,10 +35,8 @@
 	 A3XX_INT0_CP_AHB_ERROR_HALT |     \
 	 A3XX_INT0_UCHE_OOB_ACCESS)
 
+extern bool hang_debug;
 
-static bool hang_debug = false;
-MODULE_PARM_DESC(hang_debug, "Dump registers when hang is detected (can be slow!)");
-module_param_named(hang_debug, hang_debug, bool, 0600);
 static void a3xx_dump(struct msm_gpu *gpu);
 
 static void a3xx_me_init(struct msm_gpu *gpu)
@@ -387,58 +385,26 @@ static const unsigned int a3xx_registers[] = {
 	0x2750, 0x2756, 0x2760, 0x2760, 0x300c, 0x300e, 0x301c, 0x301d,
 	0x302a, 0x302a, 0x302c, 0x302d, 0x3030, 0x3031, 0x3034, 0x3036,
 	0x303c, 0x303c, 0x305e, 0x305f,
+	~0   /* sentinel */
 };
 
 #ifdef CONFIG_DEBUG_FS
 static void a3xx_show(struct msm_gpu *gpu, struct seq_file *m)
 {
-	int i;
-
-	adreno_show(gpu, m);
-
 	gpu->funcs->pm_resume(gpu);
-
 	seq_printf(m, "status:   %08x\n",
 			gpu_read(gpu, REG_A3XX_RBBM_STATUS));
-
-	/* dump these out in a form that can be parsed by demsm: */
-	seq_printf(m, "IO:region %s 00000000 00020000\n", gpu->name);
-	for (i = 0; i < ARRAY_SIZE(a3xx_registers); i += 2) {
-		uint32_t start = a3xx_registers[i];
-		uint32_t end   = a3xx_registers[i+1];
-		uint32_t addr;
-
-		for (addr = start; addr <= end; addr++) {
-			uint32_t val = gpu_read(gpu, addr);
-			seq_printf(m, "IO:R %08x %08x\n", addr<<2, val);
-		}
-	}
-
 	gpu->funcs->pm_suspend(gpu);
+	adreno_show(gpu, m);
 }
 #endif
 
 /* would be nice to not have to duplicate the _show() stuff with printk(): */
 static void a3xx_dump(struct msm_gpu *gpu)
 {
-	int i;
-
-	adreno_dump(gpu);
 	printk("status:   %08x\n",
 			gpu_read(gpu, REG_A3XX_RBBM_STATUS));
-
-	/* dump these out in a form that can be parsed by demsm: */
-	printk("IO:region %s 00000000 00020000\n", gpu->name);
-	for (i = 0; i < ARRAY_SIZE(a3xx_registers); i += 2) {
-		uint32_t start = a3xx_registers[i];
-		uint32_t end   = a3xx_registers[i+1];
-		uint32_t addr;
-
-		for (addr = start; addr <= end; addr++) {
-			uint32_t val = gpu_read(gpu, addr);
-			printk("IO:R %08x %08x\n", addr<<2, val);
-		}
-	}
+	adreno_dump(gpu);
 }
 
 static const struct adreno_gpu_funcs funcs = {
@@ -474,7 +440,6 @@ struct msm_gpu *a3xx_gpu_init(struct drm_device *dev)
 	struct msm_gpu *gpu;
 	struct msm_drm_private *priv = dev->dev_private;
 	struct platform_device *pdev = priv->gpu_pdev;
-	struct adreno_platform_config *config;
 	int ret;
 
 	if (!pdev) {
@@ -483,8 +448,6 @@ struct msm_gpu *a3xx_gpu_init(struct drm_device *dev)
 		goto fail;
 	}
 
-	config = pdev->dev.platform_data;
-
 	a3xx_gpu = kzalloc(sizeof(*a3xx_gpu), GFP_KERNEL);
 	if (!a3xx_gpu) {
 		ret = -ENOMEM;
@@ -496,20 +459,12 @@ struct msm_gpu *a3xx_gpu_init(struct drm_device *dev)
 
 	a3xx_gpu->pdev = pdev;
 
-	gpu->fast_rate = config->fast_rate;
-	gpu->slow_rate = config->slow_rate;
-	gpu->bus_freq  = config->bus_freq;
-#ifdef CONFIG_MSM_BUS_SCALING
-	gpu->bus_scale_table = config->bus_scale_table;
-#endif
-
-	DBG("fast_rate=%u, slow_rate=%u, bus_freq=%u",
-			gpu->fast_rate, gpu->slow_rate, gpu->bus_freq);
-
 	gpu->perfcntrs = perfcntrs;
 	gpu->num_perfcntrs = ARRAY_SIZE(perfcntrs);
 
-	ret = adreno_gpu_init(dev, pdev, adreno_gpu, &funcs, config->rev);
+	adreno_gpu->registers = a3xx_registers;
+
+	ret = adreno_gpu_init(dev, pdev, adreno_gpu, &funcs);
 	if (ret)
 		goto fail;
 
@@ -549,158 +504,3 @@ fail:
 
 	return ERR_PTR(ret);
 }
-
-/*
- * The a3xx device:
- */
-
-#if defined(CONFIG_MSM_BUS_SCALING) && !defined(CONFIG_OF)
-#  include <mach/kgsl.h>
-#endif
-
-static void set_gpu_pdev(struct drm_device *dev,
-		struct platform_device *pdev)
-{
-	struct msm_drm_private *priv = dev->dev_private;
-	priv->gpu_pdev = pdev;
-}
-
-static int a3xx_bind(struct device *dev, struct device *master, void *data)
-{
-	static struct adreno_platform_config config = {};
-#ifdef CONFIG_OF
-	struct device_node *child, *node = dev->of_node;
-	u32 val;
-	int ret;
-
-	ret = of_property_read_u32(node, "qcom,chipid", &val);
-	if (ret) {
-		dev_err(dev, "could not find chipid: %d\n", ret);
-		return ret;
-	}
-
-	config.rev = ADRENO_REV((val >> 24) & 0xff,
-			(val >> 16) & 0xff, (val >> 8) & 0xff, val & 0xff);
-
-	/* find clock rates: */
-	config.fast_rate = 0;
-	config.slow_rate = ~0;
-	for_each_child_of_node(node, child) {
-		if (of_device_is_compatible(child, "qcom,gpu-pwrlevels")) {
-			struct device_node *pwrlvl;
-			for_each_child_of_node(child, pwrlvl) {
-				ret = of_property_read_u32(pwrlvl, "qcom,gpu-freq", &val);
-				if (ret) {
-					dev_err(dev, "could not find gpu-freq: %d\n", ret);
-					return ret;
-				}
-				config.fast_rate = max(config.fast_rate, val);
-				config.slow_rate = min(config.slow_rate, val);
-			}
-		}
-	}
-
-	if (!config.fast_rate) {
-		dev_err(dev, "could not find clk rates\n");
-		return -ENXIO;
-	}
-
-#else
-	struct kgsl_device_platform_data *pdata = dev->platform_data;
-	uint32_t version = socinfo_get_version();
-	if (cpu_is_apq8064ab()) {
-		config.fast_rate = 450000000;
-		config.slow_rate = 27000000;
-		config.bus_freq  = 4;
-		config.rev = ADRENO_REV(3, 2, 1, 0);
-	} else if (cpu_is_apq8064()) {
-		config.fast_rate = 400000000;
-		config.slow_rate = 27000000;
-		config.bus_freq  = 4;
-
-		if (SOCINFO_VERSION_MAJOR(version) == 2)
-			config.rev = ADRENO_REV(3, 2, 0, 2);
-		else if ((SOCINFO_VERSION_MAJOR(version) == 1) &&
-				(SOCINFO_VERSION_MINOR(version) == 1))
-			config.rev = ADRENO_REV(3, 2, 0, 1);
-		else
-			config.rev = ADRENO_REV(3, 2, 0, 0);
-
-	} else if (cpu_is_msm8960ab()) {
-		config.fast_rate = 400000000;
-		config.slow_rate = 320000000;
-		config.bus_freq  = 4;
-
-		if (SOCINFO_VERSION_MINOR(version) == 0)
-			config.rev = ADRENO_REV(3, 2, 1, 0);
-		else
-			config.rev = ADRENO_REV(3, 2, 1, 1);
-
-	} else if (cpu_is_msm8930()) {
-		config.fast_rate = 400000000;
-		config.slow_rate = 27000000;
-		config.bus_freq  = 3;
-
-		if ((SOCINFO_VERSION_MAJOR(version) == 1) &&
-			(SOCINFO_VERSION_MINOR(version) == 2))
-			config.rev = ADRENO_REV(3, 0, 5, 2);
-		else
-			config.rev = ADRENO_REV(3, 0, 5, 0);
-
-	}
-#  ifdef CONFIG_MSM_BUS_SCALING
-	config.bus_scale_table = pdata->bus_scale_table;
-#  endif
-#endif
-	dev->platform_data = &config;
-	set_gpu_pdev(dev_get_drvdata(master), to_platform_device(dev));
-	return 0;
-}
-
-static void a3xx_unbind(struct device *dev, struct device *master,
-		void *data)
-{
-	set_gpu_pdev(dev_get_drvdata(master), NULL);
-}
-
-static const struct component_ops a3xx_ops = {
-		.bind   = a3xx_bind,
-		.unbind = a3xx_unbind,
-};
-
-static int a3xx_probe(struct platform_device *pdev)
-{
-	return component_add(&pdev->dev, &a3xx_ops);
-}
-
-static int a3xx_remove(struct platform_device *pdev)
-{
-	component_del(&pdev->dev, &a3xx_ops);
-	return 0;
-}
-
-static const struct of_device_id dt_match[] = {
-	{ .compatible = "qcom,adreno-3xx" },
-	/* for backwards compat w/ downstream kgsl DT files: */
-	{ .compatible = "qcom,kgsl-3d0" },
-	{}
-};
-
-static struct platform_driver a3xx_driver = {
-	.probe = a3xx_probe,
-	.remove = a3xx_remove,
-	.driver = {
-		.name = "kgsl-3d0",
-		.of_match_table = dt_match,
-	},
-};
-
-void __init a3xx_register(void)
-{
-	platform_driver_register(&a3xx_driver);
-}
-
-void __exit a3xx_unregister(void)
-{
-	platform_driver_unregister(&a3xx_driver);
-}
diff --git a/drivers/gpu/drm/msm/adreno/adreno_common.xml.h b/drivers/gpu/drm/msm/adreno/adreno_common.xml.h
index 9de19ac2e86c..cc341bc62b51 100644
--- a/drivers/gpu/drm/msm/adreno/adreno_common.xml.h
+++ b/drivers/gpu/drm/msm/adreno/adreno_common.xml.h
@@ -12,9 +12,9 @@ The rules-ng-ng source files this header was generated from are:
 - /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml  (   1453 bytes, from 2013-03-31 16:51:27)
 - /home/robclark/src/freedreno/envytools/rnndb/adreno/a2xx.xml          (  32901 bytes, from 2014-06-02 15:21:30)
 - /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_common.xml (   9859 bytes, from 2014-06-02 15:21:30)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_pm4.xml    (  14477 bytes, from 2014-05-16 11:51:57)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno/a3xx.xml          (  58020 bytes, from 2014-06-25 12:57:16)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno/a4xx.xml          (  26602 bytes, from 2014-06-25 12:57:16)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_pm4.xml    (  14960 bytes, from 2014-07-27 17:22:13)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a3xx.xml          (  58020 bytes, from 2014-08-01 12:22:48)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a4xx.xml          (  41068 bytes, from 2014-08-01 12:22:48)
 
 Copyright (C) 2013-2014 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
diff --git a/drivers/gpu/drm/msm/adreno/adreno_device.c b/drivers/gpu/drm/msm/adreno/adreno_device.c
new file mode 100644
index 000000000000..7ab85af3a7db
--- /dev/null
+++ b/drivers/gpu/drm/msm/adreno/adreno_device.c
@@ -0,0 +1,285 @@
+/*
+ * Copyright (C) 2013-2014 Red Hat
+ * Author: Rob Clark <robdclark@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include "adreno_gpu.h"
+
+#if defined(CONFIG_MSM_BUS_SCALING) && !defined(CONFIG_OF)
+#  include <mach/kgsl.h>
+#endif
+
+#define ANY_ID 0xff
+
+bool hang_debug = false;
+MODULE_PARM_DESC(hang_debug, "Dump registers when hang is detected (can be slow!)");
+module_param_named(hang_debug, hang_debug, bool, 0600);
+
+struct msm_gpu *a3xx_gpu_init(struct drm_device *dev);
+
+static const struct adreno_info gpulist[] = {
+	{
+		.rev   = ADRENO_REV(3, 0, 5, ANY_ID),
+		.revn  = 305,
+		.name  = "A305",
+		.pm4fw = "a300_pm4.fw",
+		.pfpfw = "a300_pfp.fw",
+		.gmem  = SZ_256K,
+		.init  = a3xx_gpu_init,
+	}, {
+		.rev   = ADRENO_REV(3, 2, ANY_ID, ANY_ID),
+		.revn  = 320,
+		.name  = "A320",
+		.pm4fw = "a300_pm4.fw",
+		.pfpfw = "a300_pfp.fw",
+		.gmem  = SZ_512K,
+		.init  = a3xx_gpu_init,
+	}, {
+		.rev   = ADRENO_REV(3, 3, 0, ANY_ID),
+		.revn  = 330,
+		.name  = "A330",
+		.pm4fw = "a330_pm4.fw",
+		.pfpfw = "a330_pfp.fw",
+		.gmem  = SZ_1M,
+		.init  = a3xx_gpu_init,
+	},
+};
+
+MODULE_FIRMWARE("a300_pm4.fw");
+MODULE_FIRMWARE("a300_pfp.fw");
+MODULE_FIRMWARE("a330_pm4.fw");
+MODULE_FIRMWARE("a330_pfp.fw");
+
+static inline bool _rev_match(uint8_t entry, uint8_t id)
+{
+	return (entry == ANY_ID) || (entry == id);
+}
+
+const struct adreno_info *adreno_info(struct adreno_rev rev)
+{
+	int i;
+
+	/* identify gpu: */
+	for (i = 0; i < ARRAY_SIZE(gpulist); i++) {
+		const struct adreno_info *info = &gpulist[i];
+		if (_rev_match(info->rev.core, rev.core) &&
+				_rev_match(info->rev.major, rev.major) &&
+				_rev_match(info->rev.minor, rev.minor) &&
+				_rev_match(info->rev.patchid, rev.patchid))
+			return info;
+	}
+
+	return NULL;
+}
+
+struct msm_gpu *adreno_load_gpu(struct drm_device *dev)
+{
+	struct msm_drm_private *priv = dev->dev_private;
+	struct platform_device *pdev = priv->gpu_pdev;
+	struct adreno_platform_config *config;
+	struct adreno_rev rev;
+	const struct adreno_info *info;
+	struct msm_gpu *gpu = NULL;
+
+	if (!pdev) {
+		dev_err(dev->dev, "no adreno device\n");
+		return NULL;
+	}
+
+	config = pdev->dev.platform_data;
+	rev = config->rev;
+	info = adreno_info(config->rev);
+
+	if (!info) {
+		dev_warn(dev->dev, "Unknown GPU revision: %u.%u.%u.%u\n",
+				rev.core, rev.major, rev.minor, rev.patchid);
+		return NULL;
+	}
+
+	DBG("Found GPU: %u.%u.%u.%u",  rev.core, rev.major,
+			rev.minor, rev.patchid);
+
+	gpu = info->init(dev);
+	if (IS_ERR(gpu)) {
+		dev_warn(dev->dev, "failed to load adreno gpu\n");
+		gpu = NULL;
+		/* not fatal */
+	}
+
+	if (gpu) {
+		int ret;
+		mutex_lock(&dev->struct_mutex);
+		gpu->funcs->pm_resume(gpu);
+		mutex_unlock(&dev->struct_mutex);
+		ret = gpu->funcs->hw_init(gpu);
+		if (ret) {
+			dev_err(dev->dev, "gpu hw init failed: %d\n", ret);
+			gpu->funcs->destroy(gpu);
+			gpu = NULL;
+		} else {
+			/* give inactive pm a chance to kick in: */
+			msm_gpu_retire(gpu);
+		}
+	}
+
+	return gpu;
+}
+
+static void set_gpu_pdev(struct drm_device *dev,
+		struct platform_device *pdev)
+{
+	struct msm_drm_private *priv = dev->dev_private;
+	priv->gpu_pdev = pdev;
+}
+
+static int adreno_bind(struct device *dev, struct device *master, void *data)
+{
+	static struct adreno_platform_config config = {};
+#ifdef CONFIG_OF
+	struct device_node *child, *node = dev->of_node;
+	u32 val;
+	int ret;
+
+	ret = of_property_read_u32(node, "qcom,chipid", &val);
+	if (ret) {
+		dev_err(dev, "could not find chipid: %d\n", ret);
+		return ret;
+	}
+
+	config.rev = ADRENO_REV((val >> 24) & 0xff,
+			(val >> 16) & 0xff, (val >> 8) & 0xff, val & 0xff);
+
+	/* find clock rates: */
+	config.fast_rate = 0;
+	config.slow_rate = ~0;
+	for_each_child_of_node(node, child) {
+		if (of_device_is_compatible(child, "qcom,gpu-pwrlevels")) {
+			struct device_node *pwrlvl;
+			for_each_child_of_node(child, pwrlvl) {
+				ret = of_property_read_u32(pwrlvl, "qcom,gpu-freq", &val);
+				if (ret) {
+					dev_err(dev, "could not find gpu-freq: %d\n", ret);
+					return ret;
+				}
+				config.fast_rate = max(config.fast_rate, val);
+				config.slow_rate = min(config.slow_rate, val);
+			}
+		}
+	}
+
+	if (!config.fast_rate) {
+		dev_err(dev, "could not find clk rates\n");
+		return -ENXIO;
+	}
+
+#else
+	struct kgsl_device_platform_data *pdata = dev->platform_data;
+	uint32_t version = socinfo_get_version();
+	if (cpu_is_apq8064ab()) {
+		config.fast_rate = 450000000;
+		config.slow_rate = 27000000;
+		config.bus_freq  = 4;
+		config.rev = ADRENO_REV(3, 2, 1, 0);
+	} else if (cpu_is_apq8064()) {
+		config.fast_rate = 400000000;
+		config.slow_rate = 27000000;
+		config.bus_freq  = 4;
+
+		if (SOCINFO_VERSION_MAJOR(version) == 2)
+			config.rev = ADRENO_REV(3, 2, 0, 2);
+		else if ((SOCINFO_VERSION_MAJOR(version) == 1) &&
+				(SOCINFO_VERSION_MINOR(version) == 1))
+			config.rev = ADRENO_REV(3, 2, 0, 1);
+		else
+			config.rev = ADRENO_REV(3, 2, 0, 0);
+
+	} else if (cpu_is_msm8960ab()) {
+		config.fast_rate = 400000000;
+		config.slow_rate = 320000000;
+		config.bus_freq  = 4;
+
+		if (SOCINFO_VERSION_MINOR(version) == 0)
+			config.rev = ADRENO_REV(3, 2, 1, 0);
+		else
+			config.rev = ADRENO_REV(3, 2, 1, 1);
+
+	} else if (cpu_is_msm8930()) {
+		config.fast_rate = 400000000;
+		config.slow_rate = 27000000;
+		config.bus_freq  = 3;
+
+		if ((SOCINFO_VERSION_MAJOR(version) == 1) &&
+			(SOCINFO_VERSION_MINOR(version) == 2))
+			config.rev = ADRENO_REV(3, 0, 5, 2);
+		else
+			config.rev = ADRENO_REV(3, 0, 5, 0);
+
+	}
+#  ifdef CONFIG_MSM_BUS_SCALING
+	config.bus_scale_table = pdata->bus_scale_table;
+#  endif
+#endif
+	dev->platform_data = &config;
+	set_gpu_pdev(dev_get_drvdata(master), to_platform_device(dev));
+	return 0;
+}
+
+static void adreno_unbind(struct device *dev, struct device *master,
+		void *data)
+{
+	set_gpu_pdev(dev_get_drvdata(master), NULL);
+}
+
+static const struct component_ops a3xx_ops = {
+		.bind   = adreno_bind,
+		.unbind = adreno_unbind,
+};
+
+static int adreno_probe(struct platform_device *pdev)
+{
+	return component_add(&pdev->dev, &a3xx_ops);
+}
+
+static int adreno_remove(struct platform_device *pdev)
+{
+	component_del(&pdev->dev, &a3xx_ops);
+	return 0;
+}
+
+static const struct of_device_id dt_match[] = {
+	{ .compatible = "qcom,adreno-3xx" },
+	/* for backwards compat w/ downstream kgsl DT files: */
+	{ .compatible = "qcom,kgsl-3d0" },
+	{}
+};
+
+static struct platform_driver adreno_driver = {
+	.probe = adreno_probe,
+	.remove = adreno_remove,
+	.driver = {
+		.name = "adreno",
+		.of_match_table = dt_match,
+	},
+};
+
+void __init adreno_register(void)
+{
+	platform_driver_register(&adreno_driver);
+}
+
+void __exit adreno_unregister(void)
+{
+	platform_driver_unregister(&adreno_driver);
+}
diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.c b/drivers/gpu/drm/msm/adreno/adreno_gpu.c
index 655ce5b14ad0..6afa29167fee 100644
--- a/drivers/gpu/drm/msm/adreno/adreno_gpu.c
+++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.c
@@ -19,46 +19,6 @@
 #include "msm_gem.h"
 #include "msm_mmu.h"
 
-struct adreno_info {
-	struct adreno_rev rev;
-	uint32_t revn;
-	const char *name;
-	const char *pm4fw, *pfpfw;
-	uint32_t gmem;
-};
-
-#define ANY_ID 0xff
-
-static const struct adreno_info gpulist[] = {
-	{
-		.rev   = ADRENO_REV(3, 0, 5, ANY_ID),
-		.revn  = 305,
-		.name  = "A305",
-		.pm4fw = "a300_pm4.fw",
-		.pfpfw = "a300_pfp.fw",
-		.gmem  = SZ_256K,
-	}, {
-		.rev   = ADRENO_REV(3, 2, ANY_ID, ANY_ID),
-		.revn  = 320,
-		.name  = "A320",
-		.pm4fw = "a300_pm4.fw",
-		.pfpfw = "a300_pfp.fw",
-		.gmem  = SZ_512K,
-	}, {
-		.rev   = ADRENO_REV(3, 3, 0, ANY_ID),
-		.revn  = 330,
-		.name  = "A330",
-		.pm4fw = "a330_pm4.fw",
-		.pfpfw = "a330_pfp.fw",
-		.gmem  = SZ_1M,
-	},
-};
-
-MODULE_FIRMWARE("a300_pm4.fw");
-MODULE_FIRMWARE("a300_pfp.fw");
-MODULE_FIRMWARE("a330_pm4.fw");
-MODULE_FIRMWARE("a330_pfp.fw");
-
 #define RB_SIZE    SZ_32K
 #define RB_BLKSIZE 16
 
@@ -252,6 +212,7 @@ void adreno_idle(struct msm_gpu *gpu)
 void adreno_show(struct msm_gpu *gpu, struct seq_file *m)
 {
 	struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu);
+	int i;
 
 	seq_printf(m, "revision: %d (%d.%d.%d.%d)\n",
 			adreno_gpu->info->revn, adreno_gpu->rev.core,
@@ -263,6 +224,23 @@ void adreno_show(struct msm_gpu *gpu, struct seq_file *m)
 	seq_printf(m, "rptr:     %d\n", adreno_gpu->memptrs->rptr);
 	seq_printf(m, "wptr:     %d\n", adreno_gpu->memptrs->wptr);
 	seq_printf(m, "rb wptr:  %d\n", get_wptr(gpu->rb));
+
+	gpu->funcs->pm_resume(gpu);
+
+	/* dump these out in a form that can be parsed by demsm: */
+	seq_printf(m, "IO:region %s 00000000 00020000\n", gpu->name);
+	for (i = 0; adreno_gpu->registers[i] != ~0; i += 2) {
+		uint32_t start = adreno_gpu->registers[i];
+		uint32_t end   = adreno_gpu->registers[i+1];
+		uint32_t addr;
+
+		for (addr = start; addr <= end; addr++) {
+			uint32_t val = gpu_read(gpu, addr);
+			seq_printf(m, "IO:R %08x %08x\n", addr<<2, val);
+		}
+	}
+
+	gpu->funcs->pm_suspend(gpu);
 }
 #endif
 
@@ -270,6 +248,7 @@ void adreno_show(struct msm_gpu *gpu, struct seq_file *m)
 void adreno_dump(struct msm_gpu *gpu)
 {
 	struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu);
+	int i;
 
 	printk("revision: %d (%d.%d.%d.%d)\n",
 			adreno_gpu->info->revn, adreno_gpu->rev.core,
@@ -282,6 +261,18 @@ void adreno_dump(struct msm_gpu *gpu)
 	printk("wptr:     %d\n", adreno_gpu->memptrs->wptr);
 	printk("rb wptr:  %d\n", get_wptr(gpu->rb));
 
+	/* dump these out in a form that can be parsed by demsm: */
+	printk("IO:region %s 00000000 00020000\n", gpu->name);
+	for (i = 0; adreno_gpu->registers[i] != ~0; i += 2) {
+		uint32_t start = adreno_gpu->registers[i];
+		uint32_t end   = adreno_gpu->registers[i+1];
+		uint32_t addr;
+
+		for (addr = start; addr <= end; addr++) {
+			uint32_t val = gpu_read(gpu, addr);
+			printk("IO:R %08x %08x\n", addr<<2, val);
+		}
+	}
 }
 
 static uint32_t ring_freewords(struct msm_gpu *gpu)
@@ -304,65 +295,51 @@ static const char *iommu_ports[] = {
 		"gfx3d1_user", "gfx3d1_priv",
 };
 
-static inline bool _rev_match(uint8_t entry, uint8_t id)
-{
-	return (entry == ANY_ID) || (entry == id);
-}
-
 int adreno_gpu_init(struct drm_device *drm, struct platform_device *pdev,
-		struct adreno_gpu *gpu, const struct adreno_gpu_funcs *funcs,
-		struct adreno_rev rev)
+		struct adreno_gpu *adreno_gpu, const struct adreno_gpu_funcs *funcs)
 {
+	struct adreno_platform_config *config = pdev->dev.platform_data;
+	struct msm_gpu *gpu = &adreno_gpu->base;
 	struct msm_mmu *mmu;
-	int i, ret;
-
-	/* identify gpu: */
-	for (i = 0; i < ARRAY_SIZE(gpulist); i++) {
-		const struct adreno_info *info = &gpulist[i];
-		if (_rev_match(info->rev.core, rev.core) &&
-				_rev_match(info->rev.major, rev.major) &&
-				_rev_match(info->rev.minor, rev.minor) &&
-				_rev_match(info->rev.patchid, rev.patchid)) {
-			gpu->info = info;
-			gpu->revn = info->revn;
-			break;
-		}
-	}
-
-	if (i == ARRAY_SIZE(gpulist)) {
-		dev_err(drm->dev, "Unknown GPU revision: %u.%u.%u.%u\n",
-				rev.core, rev.major, rev.minor, rev.patchid);
-		return -ENXIO;
-	}
+	int ret;
 
-	DBG("Found GPU: %s (%u.%u.%u.%u)", gpu->info->name,
-			rev.core, rev.major, rev.minor, rev.patchid);
+	adreno_gpu->funcs = funcs;
+	adreno_gpu->info = adreno_info(config->rev);
+	adreno_gpu->gmem = adreno_gpu->info->gmem;
+	adreno_gpu->revn = adreno_gpu->info->revn;
+	adreno_gpu->rev = config->rev;
+
+	gpu->fast_rate = config->fast_rate;
+	gpu->slow_rate = config->slow_rate;
+	gpu->bus_freq  = config->bus_freq;
+#ifdef CONFIG_MSM_BUS_SCALING
+	gpu->bus_scale_table = config->bus_scale_table;
+#endif
 
-	gpu->funcs = funcs;
-	gpu->gmem = gpu->info->gmem;
-	gpu->rev = rev;
+	DBG("fast_rate=%u, slow_rate=%u, bus_freq=%u",
+			gpu->fast_rate, gpu->slow_rate, gpu->bus_freq);
 
-	ret = request_firmware(&gpu->pm4, gpu->info->pm4fw, drm->dev);
+	ret = request_firmware(&adreno_gpu->pm4, adreno_gpu->info->pm4fw, drm->dev);
 	if (ret) {
 		dev_err(drm->dev, "failed to load %s PM4 firmware: %d\n",
-				gpu->info->pm4fw, ret);
+				adreno_gpu->info->pm4fw, ret);
 		return ret;
 	}
 
-	ret = request_firmware(&gpu->pfp, gpu->info->pfpfw, drm->dev);
+	ret = request_firmware(&adreno_gpu->pfp, adreno_gpu->info->pfpfw, drm->dev);
 	if (ret) {
 		dev_err(drm->dev, "failed to load %s PFP firmware: %d\n",
-				gpu->info->pfpfw, ret);
+				adreno_gpu->info->pfpfw, ret);
 		return ret;
 	}
 
-	ret = msm_gpu_init(drm, pdev, &gpu->base, &funcs->base,
-			gpu->info->name, "kgsl_3d0_reg_memory", "kgsl_3d0_irq",
+	ret = msm_gpu_init(drm, pdev, &adreno_gpu->base, &funcs->base,
+			adreno_gpu->info->name, "kgsl_3d0_reg_memory", "kgsl_3d0_irq",
 			RB_SIZE);
 	if (ret)
 		return ret;
 
-	mmu = gpu->base.mmu;
+	mmu = gpu->mmu;
 	if (mmu) {
 		ret = mmu->funcs->attach(mmu, iommu_ports,
 				ARRAY_SIZE(iommu_ports));
@@ -371,24 +348,24 @@ int adreno_gpu_init(struct drm_device *drm, struct platform_device *pdev,
 	}
 
 	mutex_lock(&drm->struct_mutex);
-	gpu->memptrs_bo = msm_gem_new(drm, sizeof(*gpu->memptrs),
+	adreno_gpu->memptrs_bo = msm_gem_new(drm, sizeof(*adreno_gpu->memptrs),
 			MSM_BO_UNCACHED);
 	mutex_unlock(&drm->struct_mutex);
-	if (IS_ERR(gpu->memptrs_bo)) {
-		ret = PTR_ERR(gpu->memptrs_bo);
-		gpu->memptrs_bo = NULL;
+	if (IS_ERR(adreno_gpu->memptrs_bo)) {
+		ret = PTR_ERR(adreno_gpu->memptrs_bo);
+		adreno_gpu->memptrs_bo = NULL;
 		dev_err(drm->dev, "could not allocate memptrs: %d\n", ret);
 		return ret;
 	}
 
-	gpu->memptrs = msm_gem_vaddr(gpu->memptrs_bo);
-	if (!gpu->memptrs) {
+	adreno_gpu->memptrs = msm_gem_vaddr(adreno_gpu->memptrs_bo);
+	if (!adreno_gpu->memptrs) {
 		dev_err(drm->dev, "could not vmap memptrs\n");
 		return -ENOMEM;
 	}
 
-	ret = msm_gem_get_iova(gpu->memptrs_bo, gpu->base.id,
-			&gpu->memptrs_iova);
+	ret = msm_gem_get_iova(adreno_gpu->memptrs_bo, gpu->id,
+			&adreno_gpu->memptrs_iova);
 	if (ret) {
 		dev_err(drm->dev, "could not map memptrs: %d\n", ret);
 		return ret;
diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.h b/drivers/gpu/drm/msm/adreno/adreno_gpu.h
index 63c36ce33020..52f051579753 100644
--- a/drivers/gpu/drm/msm/adreno/adreno_gpu.h
+++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.h
@@ -39,7 +39,16 @@ struct adreno_gpu_funcs {
 	struct msm_gpu_funcs base;
 };
 
-struct adreno_info;
+struct adreno_info {
+	struct adreno_rev rev;
+	uint32_t revn;
+	const char *name;
+	const char *pm4fw, *pfpfw;
+	uint32_t gmem;
+	struct msm_gpu *(*init)(struct drm_device *dev);
+};
+
+const struct adreno_info *adreno_info(struct adreno_rev rev);
 
 struct adreno_rbmemptrs {
 	volatile uint32_t rptr;
@@ -55,6 +64,9 @@ struct adreno_gpu {
 	uint32_t revn;  /* numeric revision name */
 	const struct adreno_gpu_funcs *funcs;
 
+	/* interesting register offsets to dump: */
+	const unsigned int *registers;
+
 	/* firmware: */
 	const struct firmware *pm4, *pfp;
 
@@ -131,8 +143,7 @@ void adreno_dump(struct msm_gpu *gpu);
 void adreno_wait_ring(struct msm_gpu *gpu, uint32_t ndwords);
 
 int adreno_gpu_init(struct drm_device *drm, struct platform_device *pdev,
-		struct adreno_gpu *gpu, const struct adreno_gpu_funcs *funcs,
-		struct adreno_rev rev);
+		struct adreno_gpu *gpu, const struct adreno_gpu_funcs *funcs);
 void adreno_gpu_cleanup(struct adreno_gpu *gpu);
 
 
diff --git a/drivers/gpu/drm/msm/adreno/adreno_pm4.xml.h b/drivers/gpu/drm/msm/adreno/adreno_pm4.xml.h
index 4eee0ec8f069..6ef43f66c30a 100644
--- a/drivers/gpu/drm/msm/adreno/adreno_pm4.xml.h
+++ b/drivers/gpu/drm/msm/adreno/adreno_pm4.xml.h
@@ -12,9 +12,9 @@ The rules-ng-ng source files this header was generated from are:
 - /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml  (   1453 bytes, from 2013-03-31 16:51:27)
 - /home/robclark/src/freedreno/envytools/rnndb/adreno/a2xx.xml          (  32901 bytes, from 2014-06-02 15:21:30)
 - /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_common.xml (   9859 bytes, from 2014-06-02 15:21:30)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_pm4.xml    (  14477 bytes, from 2014-05-16 11:51:57)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno/a3xx.xml          (  58020 bytes, from 2014-06-25 12:57:16)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno/a4xx.xml          (  26602 bytes, from 2014-06-25 12:57:16)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_pm4.xml    (  14960 bytes, from 2014-07-27 17:22:13)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a3xx.xml          (  58020 bytes, from 2014-08-01 12:22:48)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a4xx.xml          (  41068 bytes, from 2014-08-01 12:22:48)
 
 Copyright (C) 2013-2014 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
@@ -163,12 +163,16 @@ enum adreno_pm4_type3_packets {
 	CP_INDIRECT_BUFFER_PFE = 63,
 	CP_SET_BIN = 76,
 	CP_TEST_TWO_MEMS = 113,
+	CP_REG_WR_NO_CTXT = 120,
+	CP_RECORD_PFP_TIMESTAMP = 17,
 	CP_WAIT_FOR_ME = 19,
 	CP_SET_DRAW_STATE = 67,
 	CP_DRAW_INDX_OFFSET = 56,
 	CP_DRAW_INDIRECT = 40,
 	CP_DRAW_INDX_INDIRECT = 41,
 	CP_DRAW_AUTO = 36,
+	CP_UNKNOWN_1A = 26,
+	CP_WIDE_REG_WRITE = 116,
 	IN_IB_PREFETCH_END = 23,
 	IN_SUBBLK_PREFETCH = 31,
 	IN_INSTR_PREFETCH = 32,
diff --git a/drivers/gpu/drm/msm/dsi/dsi.xml.h b/drivers/gpu/drm/msm/dsi/dsi.xml.h
index 0f1f5b9459a5..e965898dfda6 100644
--- a/drivers/gpu/drm/msm/dsi/dsi.xml.h
+++ b/drivers/gpu/drm/msm/dsi/dsi.xml.h
@@ -10,14 +10,14 @@ git clone https://github.com/freedreno/envytools.git
 The rules-ng-ng source files this header was generated from are:
 - /home/robclark/src/freedreno/envytools/rnndb/msm.xml                 (    647 bytes, from 2013-11-30 14:45:35)
 - /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml (   1453 bytes, from 2013-03-31 16:51:27)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp4.xml            (  17996 bytes, from 2013-12-01 19:10:31)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp_common.xml      (   1615 bytes, from 2013-11-30 15:00:52)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp5.xml            (  22517 bytes, from 2014-06-25 12:55:02)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp4.xml            (  20457 bytes, from 2014-08-01 12:22:48)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp_common.xml      (   1615 bytes, from 2014-07-17 15:34:33)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp5.xml            (  22517 bytes, from 2014-07-17 15:34:33)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/dsi.xml             (  11712 bytes, from 2013-08-17 17:13:43)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/sfpb.xml            (    344 bytes, from 2013-08-11 19:26:32)
-- /home/robclark/src/freedreno/envytools/rnndb/dsi/mmss_cc.xml         (   1544 bytes, from 2013-08-16 19:17:05)
+- /home/robclark/src/freedreno/envytools/rnndb/dsi/mmss_cc.xml         (   1686 bytes, from 2014-08-01 12:23:53)
 - /home/robclark/src/freedreno/envytools/rnndb/hdmi/qfprom.xml         (    600 bytes, from 2013-07-05 19:21:12)
-- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  23613 bytes, from 2014-06-25 12:53:44)
+- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  23613 bytes, from 2014-07-17 15:33:30)
 
 Copyright (C) 2013 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
diff --git a/drivers/gpu/drm/msm/dsi/mmss_cc.xml.h b/drivers/gpu/drm/msm/dsi/mmss_cc.xml.h
index d468f86f637c..f2bdda957205 100644
--- a/drivers/gpu/drm/msm/dsi/mmss_cc.xml.h
+++ b/drivers/gpu/drm/msm/dsi/mmss_cc.xml.h
@@ -10,16 +10,16 @@ git clone https://github.com/freedreno/envytools.git
 The rules-ng-ng source files this header was generated from are:
 - /home/robclark/src/freedreno/envytools/rnndb/msm.xml                 (    647 bytes, from 2013-11-30 14:45:35)
 - /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml (   1453 bytes, from 2013-03-31 16:51:27)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp4.xml            (  17996 bytes, from 2013-12-01 19:10:31)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp_common.xml      (   1615 bytes, from 2013-11-30 15:00:52)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp5.xml            (  22517 bytes, from 2014-06-25 12:55:02)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp4.xml            (  20457 bytes, from 2014-08-01 12:22:48)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp_common.xml      (   1615 bytes, from 2014-07-17 15:34:33)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp5.xml            (  22517 bytes, from 2014-07-17 15:34:33)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/dsi.xml             (  11712 bytes, from 2013-08-17 17:13:43)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/sfpb.xml            (    344 bytes, from 2013-08-11 19:26:32)
-- /home/robclark/src/freedreno/envytools/rnndb/dsi/mmss_cc.xml         (   1544 bytes, from 2013-08-16 19:17:05)
+- /home/robclark/src/freedreno/envytools/rnndb/dsi/mmss_cc.xml         (   1686 bytes, from 2014-08-01 12:23:53)
 - /home/robclark/src/freedreno/envytools/rnndb/hdmi/qfprom.xml         (    600 bytes, from 2013-07-05 19:21:12)
-- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  23613 bytes, from 2014-06-25 12:53:44)
+- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  23613 bytes, from 2014-07-17 15:33:30)
 
-Copyright (C) 2013 by the following authors:
+Copyright (C) 2013-2014 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
 
 Permission is hereby granted, free of charge, to any person obtaining
@@ -112,5 +112,11 @@ static inline uint32_t MMSS_CC_CLK_NS_VAL(uint32_t val)
 	return ((val) << MMSS_CC_CLK_NS_VAL__SHIFT) & MMSS_CC_CLK_NS_VAL__MASK;
 }
 
+#define REG_MMSS_CC_DSI2_PIXEL_CC				0x00000094
+
+#define REG_MMSS_CC_DSI2_PIXEL_NS				0x000000e4
+
+#define REG_MMSS_CC_DSI2_PIXEL_CC2				0x00000264
+
 
 #endif /* MMSS_CC_XML */
diff --git a/drivers/gpu/drm/msm/dsi/sfpb.xml.h b/drivers/gpu/drm/msm/dsi/sfpb.xml.h
index da8740054cdf..e5b071ffd865 100644
--- a/drivers/gpu/drm/msm/dsi/sfpb.xml.h
+++ b/drivers/gpu/drm/msm/dsi/sfpb.xml.h
@@ -10,14 +10,14 @@ git clone https://github.com/freedreno/envytools.git
 The rules-ng-ng source files this header was generated from are:
 - /home/robclark/src/freedreno/envytools/rnndb/msm.xml                 (    647 bytes, from 2013-11-30 14:45:35)
 - /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml (   1453 bytes, from 2013-03-31 16:51:27)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp4.xml            (  17996 bytes, from 2013-12-01 19:10:31)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp_common.xml      (   1615 bytes, from 2013-11-30 15:00:52)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp5.xml            (  22517 bytes, from 2014-06-25 12:55:02)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp4.xml            (  20457 bytes, from 2014-08-01 12:22:48)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp_common.xml      (   1615 bytes, from 2014-07-17 15:34:33)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp5.xml            (  22517 bytes, from 2014-07-17 15:34:33)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/dsi.xml             (  11712 bytes, from 2013-08-17 17:13:43)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/sfpb.xml            (    344 bytes, from 2013-08-11 19:26:32)
-- /home/robclark/src/freedreno/envytools/rnndb/dsi/mmss_cc.xml         (   1544 bytes, from 2013-08-16 19:17:05)
+- /home/robclark/src/freedreno/envytools/rnndb/dsi/mmss_cc.xml         (   1686 bytes, from 2014-08-01 12:23:53)
 - /home/robclark/src/freedreno/envytools/rnndb/hdmi/qfprom.xml         (    600 bytes, from 2013-07-05 19:21:12)
-- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  23613 bytes, from 2014-06-25 12:53:44)
+- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  23613 bytes, from 2014-07-17 15:33:30)
 
 Copyright (C) 2013 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
diff --git a/drivers/gpu/drm/msm/hdmi/hdmi.c b/drivers/gpu/drm/msm/hdmi/hdmi.c
index c6c9b02e0ada..9d00dcba6959 100644
--- a/drivers/gpu/drm/msm/hdmi/hdmi.c
+++ b/drivers/gpu/drm/msm/hdmi/hdmi.c
@@ -123,7 +123,7 @@ struct hdmi *hdmi_init(struct drm_device *dev, struct drm_encoder *encoder)
 	for (i = 0; i < config->hpd_reg_cnt; i++) {
 		struct regulator *reg;
 
-		reg = devm_regulator_get_exclusive(&pdev->dev,
+		reg = devm_regulator_get(&pdev->dev,
 				config->hpd_reg_names[i]);
 		if (IS_ERR(reg)) {
 			ret = PTR_ERR(reg);
@@ -139,7 +139,7 @@ struct hdmi *hdmi_init(struct drm_device *dev, struct drm_encoder *encoder)
 	for (i = 0; i < config->pwr_reg_cnt; i++) {
 		struct regulator *reg;
 
-		reg = devm_regulator_get_exclusive(&pdev->dev,
+		reg = devm_regulator_get(&pdev->dev,
 				config->pwr_reg_names[i]);
 		if (IS_ERR(reg)) {
 			ret = PTR_ERR(reg);
diff --git a/drivers/gpu/drm/msm/hdmi/hdmi.xml.h b/drivers/gpu/drm/msm/hdmi/hdmi.xml.h
index e89fe053d375..76fd0cfc6558 100644
--- a/drivers/gpu/drm/msm/hdmi/hdmi.xml.h
+++ b/drivers/gpu/drm/msm/hdmi/hdmi.xml.h
@@ -10,14 +10,14 @@ git clone https://github.com/freedreno/envytools.git
 The rules-ng-ng source files this header was generated from are:
 - /home/robclark/src/freedreno/envytools/rnndb/msm.xml                 (    647 bytes, from 2013-11-30 14:45:35)
 - /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml (   1453 bytes, from 2013-03-31 16:51:27)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp4.xml            (  17996 bytes, from 2013-12-01 19:10:31)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp_common.xml      (   1615 bytes, from 2013-11-30 15:00:52)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp5.xml            (  22517 bytes, from 2014-06-25 12:55:02)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp4.xml            (  20457 bytes, from 2014-08-01 12:22:48)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp_common.xml      (   1615 bytes, from 2014-07-17 15:34:33)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp5.xml            (  22517 bytes, from 2014-07-17 15:34:33)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/dsi.xml             (  11712 bytes, from 2013-08-17 17:13:43)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/sfpb.xml            (    344 bytes, from 2013-08-11 19:26:32)
-- /home/robclark/src/freedreno/envytools/rnndb/dsi/mmss_cc.xml         (   1544 bytes, from 2013-08-16 19:17:05)
+- /home/robclark/src/freedreno/envytools/rnndb/dsi/mmss_cc.xml         (   1686 bytes, from 2014-08-01 12:23:53)
 - /home/robclark/src/freedreno/envytools/rnndb/hdmi/qfprom.xml         (    600 bytes, from 2013-07-05 19:21:12)
-- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  23613 bytes, from 2014-06-25 12:53:44)
+- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  23613 bytes, from 2014-07-17 15:33:30)
 
 Copyright (C) 2013-2014 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
diff --git a/drivers/gpu/drm/msm/hdmi/qfprom.xml.h b/drivers/gpu/drm/msm/hdmi/qfprom.xml.h
index bd81db6a7829..d53c29327df9 100644
--- a/drivers/gpu/drm/msm/hdmi/qfprom.xml.h
+++ b/drivers/gpu/drm/msm/hdmi/qfprom.xml.h
@@ -10,14 +10,14 @@ git clone https://github.com/freedreno/envytools.git
 The rules-ng-ng source files this header was generated from are:
 - /home/robclark/src/freedreno/envytools/rnndb/msm.xml                 (    647 bytes, from 2013-11-30 14:45:35)
 - /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml (   1453 bytes, from 2013-03-31 16:51:27)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp4.xml            (  17996 bytes, from 2013-12-01 19:10:31)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp_common.xml      (   1615 bytes, from 2013-11-30 15:00:52)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp5.xml            (  22517 bytes, from 2014-06-25 12:55:02)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp4.xml            (  20457 bytes, from 2014-08-01 12:22:48)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp_common.xml      (   1615 bytes, from 2014-07-17 15:34:33)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp5.xml            (  22517 bytes, from 2014-07-17 15:34:33)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/dsi.xml             (  11712 bytes, from 2013-08-17 17:13:43)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/sfpb.xml            (    344 bytes, from 2013-08-11 19:26:32)
-- /home/robclark/src/freedreno/envytools/rnndb/dsi/mmss_cc.xml         (   1544 bytes, from 2013-08-16 19:17:05)
+- /home/robclark/src/freedreno/envytools/rnndb/dsi/mmss_cc.xml         (   1686 bytes, from 2014-08-01 12:23:53)
 - /home/robclark/src/freedreno/envytools/rnndb/hdmi/qfprom.xml         (    600 bytes, from 2013-07-05 19:21:12)
-- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  23613 bytes, from 2014-06-25 12:53:44)
+- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  23613 bytes, from 2014-07-17 15:33:30)
 
 Copyright (C) 2013 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
diff --git a/drivers/gpu/drm/msm/mdp/mdp4/mdp4.xml.h b/drivers/gpu/drm/msm/mdp/mdp4/mdp4.xml.h
index 122208e8a2ee..03c0bd9cd5b9 100644
--- a/drivers/gpu/drm/msm/mdp/mdp4/mdp4.xml.h
+++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4.xml.h
@@ -10,16 +10,16 @@ git clone https://github.com/freedreno/envytools.git
 The rules-ng-ng source files this header was generated from are:
 - /home/robclark/src/freedreno/envytools/rnndb/msm.xml                 (    647 bytes, from 2013-11-30 14:45:35)
 - /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml (   1453 bytes, from 2013-03-31 16:51:27)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp4.xml            (  17996 bytes, from 2013-12-01 19:10:31)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp_common.xml      (   1615 bytes, from 2013-11-30 15:00:52)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp5.xml            (  22517 bytes, from 2014-06-25 12:55:02)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp4.xml            (  20457 bytes, from 2014-08-01 12:22:48)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp_common.xml      (   1615 bytes, from 2014-07-17 15:34:33)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp5.xml            (  22517 bytes, from 2014-07-17 15:34:33)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/dsi.xml             (  11712 bytes, from 2013-08-17 17:13:43)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/sfpb.xml            (    344 bytes, from 2013-08-11 19:26:32)
-- /home/robclark/src/freedreno/envytools/rnndb/dsi/mmss_cc.xml         (   1544 bytes, from 2013-08-16 19:17:05)
+- /home/robclark/src/freedreno/envytools/rnndb/dsi/mmss_cc.xml         (   1686 bytes, from 2014-08-01 12:23:53)
 - /home/robclark/src/freedreno/envytools/rnndb/hdmi/qfprom.xml         (    600 bytes, from 2013-07-05 19:21:12)
-- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  23613 bytes, from 2014-06-25 12:53:44)
+- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  23613 bytes, from 2014-07-17 15:33:30)
 
-Copyright (C) 2013 by the following authors:
+Copyright (C) 2013-2014 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
 
 Permission is hereby granted, free of charge, to any person obtaining
@@ -871,6 +871,101 @@ static inline uint32_t MDP4_LCDC_UNDERFLOW_CLR_COLOR(uint32_t val)
 #define MDP4_LCDC_CTRL_POLARITY_VSYNC_LOW			0x00000002
 #define MDP4_LCDC_CTRL_POLARITY_DATA_EN_LOW			0x00000004
 
+#define REG_MDP4_LCDC_LVDS_INTF_CTL				0x000c2000
+#define MDP4_LCDC_LVDS_INTF_CTL_MODE_SEL			0x00000004
+#define MDP4_LCDC_LVDS_INTF_CTL_RGB_OUT				0x00000008
+#define MDP4_LCDC_LVDS_INTF_CTL_CH_SWAP				0x00000010
+#define MDP4_LCDC_LVDS_INTF_CTL_CH1_RES_BIT			0x00000020
+#define MDP4_LCDC_LVDS_INTF_CTL_CH2_RES_BIT			0x00000040
+#define MDP4_LCDC_LVDS_INTF_CTL_ENABLE				0x00000080
+#define MDP4_LCDC_LVDS_INTF_CTL_CH1_DATA_LANE0_EN		0x00000100
+#define MDP4_LCDC_LVDS_INTF_CTL_CH1_DATA_LANE1_EN		0x00000200
+#define MDP4_LCDC_LVDS_INTF_CTL_CH1_DATA_LANE2_EN		0x00000400
+#define MDP4_LCDC_LVDS_INTF_CTL_CH1_DATA_LANE3_EN		0x00000800
+#define MDP4_LCDC_LVDS_INTF_CTL_CH2_DATA_LANE0_EN		0x00001000
+#define MDP4_LCDC_LVDS_INTF_CTL_CH2_DATA_LANE1_EN		0x00002000
+#define MDP4_LCDC_LVDS_INTF_CTL_CH2_DATA_LANE2_EN		0x00004000
+#define MDP4_LCDC_LVDS_INTF_CTL_CH2_DATA_LANE3_EN		0x00008000
+#define MDP4_LCDC_LVDS_INTF_CTL_CH1_CLK_LANE_EN			0x00010000
+#define MDP4_LCDC_LVDS_INTF_CTL_CH2_CLK_LANE_EN			0x00020000
+
+static inline uint32_t REG_MDP4_LCDC_LVDS_MUX_CTL(uint32_t i0) { return 0x000c2014 + 0x8*i0; }
+
+static inline uint32_t REG_MDP4_LCDC_LVDS_MUX_CTL_3_TO_0(uint32_t i0) { return 0x000c2014 + 0x8*i0; }
+#define MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT0__MASK		0x000000ff
+#define MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT0__SHIFT		0
+static inline uint32_t MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT0(uint32_t val)
+{
+	return ((val) << MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT0__SHIFT) & MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT0__MASK;
+}
+#define MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT1__MASK		0x0000ff00
+#define MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT1__SHIFT		8
+static inline uint32_t MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT1(uint32_t val)
+{
+	return ((val) << MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT1__SHIFT) & MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT1__MASK;
+}
+#define MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT2__MASK		0x00ff0000
+#define MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT2__SHIFT		16
+static inline uint32_t MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT2(uint32_t val)
+{
+	return ((val) << MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT2__SHIFT) & MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT2__MASK;
+}
+#define MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT3__MASK		0xff000000
+#define MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT3__SHIFT		24
+static inline uint32_t MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT3(uint32_t val)
+{
+	return ((val) << MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT3__SHIFT) & MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT3__MASK;
+}
+
+static inline uint32_t REG_MDP4_LCDC_LVDS_MUX_CTL_6_TO_4(uint32_t i0) { return 0x000c2018 + 0x8*i0; }
+#define MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT4__MASK		0x000000ff
+#define MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT4__SHIFT		0
+static inline uint32_t MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT4(uint32_t val)
+{
+	return ((val) << MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT4__SHIFT) & MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT4__MASK;
+}
+#define MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT5__MASK		0x0000ff00
+#define MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT5__SHIFT		8
+static inline uint32_t MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT5(uint32_t val)
+{
+	return ((val) << MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT5__SHIFT) & MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT5__MASK;
+}
+#define MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT6__MASK		0x00ff0000
+#define MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT6__SHIFT		16
+static inline uint32_t MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT6(uint32_t val)
+{
+	return ((val) << MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT6__SHIFT) & MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT6__MASK;
+}
+
+#define REG_MDP4_LCDC_LVDS_PHY_RESET				0x000c2034
+
+#define REG_MDP4_LVDS_PHY_PLL_CTRL_0				0x000c3000
+
+#define REG_MDP4_LVDS_PHY_PLL_CTRL_1				0x000c3004
+
+#define REG_MDP4_LVDS_PHY_PLL_CTRL_2				0x000c3008
+
+#define REG_MDP4_LVDS_PHY_PLL_CTRL_3				0x000c300c
+
+#define REG_MDP4_LVDS_PHY_PLL_CTRL_5				0x000c3014
+
+#define REG_MDP4_LVDS_PHY_PLL_CTRL_6				0x000c3018
+
+#define REG_MDP4_LVDS_PHY_PLL_CTRL_7				0x000c301c
+
+#define REG_MDP4_LVDS_PHY_PLL_CTRL_8				0x000c3020
+
+#define REG_MDP4_LVDS_PHY_PLL_CTRL_9				0x000c3024
+
+#define REG_MDP4_LVDS_PHY_PLL_LOCKED				0x000c3080
+
+#define REG_MDP4_LVDS_PHY_CFG2					0x000c3108
+
+#define REG_MDP4_LVDS_PHY_CFG0					0x000c3100
+#define MDP4_LVDS_PHY_CFG0_SERIALIZATION_ENBLE			0x00000010
+#define MDP4_LVDS_PHY_CFG0_CHANNEL0				0x00000040
+#define MDP4_LVDS_PHY_CFG0_CHANNEL1				0x00000080
+
 #define REG_MDP4_DTV						0x000d0000
 
 #define REG_MDP4_DTV_ENABLE					0x000d0000
diff --git a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_crtc.c b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_crtc.c
index c6c80ea28c35..7d00f7fb5773 100644
--- a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_crtc.c
+++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_crtc.c
@@ -273,14 +273,17 @@ static void blend_setup(struct drm_crtc *crtc)
 	};
 	bool alpha[4]= { false, false, false, false };
 
+	/* Don't rely on value read back from hw, but instead use our
+	 * own shadowed value.  Possibly disable/reenable looses the
+	 * previous value and goes back to power-on default?
+	 */
+	mixer_cfg = mdp4_kms->mixer_cfg;
+
 	mdp4_write(mdp4_kms, REG_MDP4_OVLP_TRANSP_LOW0(ovlp), 0);
 	mdp4_write(mdp4_kms, REG_MDP4_OVLP_TRANSP_LOW1(ovlp), 0);
 	mdp4_write(mdp4_kms, REG_MDP4_OVLP_TRANSP_HIGH0(ovlp), 0);
 	mdp4_write(mdp4_kms, REG_MDP4_OVLP_TRANSP_HIGH1(ovlp), 0);
 
-	/* TODO single register for all CRTCs, so this won't work properly
-	 * when multiple CRTCs are active..
-	 */
 	for (i = 0; i < ARRAY_SIZE(mdp4_crtc->planes); i++) {
 		struct drm_plane *plane = mdp4_crtc->planes[i];
 		if (plane) {
@@ -291,7 +294,8 @@ static void blend_setup(struct drm_crtc *crtc)
 					to_mdp_format(msm_framebuffer_format(plane->fb));
 				alpha[idx-1] = format->alpha_enable;
 			}
-			mixer_cfg |= mixercfg(mdp4_crtc->mixer, pipe_id, stages[idx]);
+			mixer_cfg = mixercfg(mixer_cfg, mdp4_crtc->mixer,
+					pipe_id, stages[idx]);
 		}
 	}
 
@@ -320,6 +324,7 @@ static void blend_setup(struct drm_crtc *crtc)
 		mdp4_write(mdp4_kms, REG_MDP4_OVLP_STAGE_TRANSP_HIGH1(ovlp, i), 0);
 	}
 
+	mdp4_kms->mixer_cfg = mixer_cfg;
 	mdp4_write(mdp4_kms, REG_MDP4_LAYERMIXER_IN_CFG, mixer_cfg);
 }
 
@@ -672,7 +677,7 @@ void mdp4_crtc_set_config(struct drm_crtc *crtc, uint32_t config)
 }
 
 /* set interface for routing crtc->encoder: */
-void mdp4_crtc_set_intf(struct drm_crtc *crtc, enum mdp4_intf intf)
+void mdp4_crtc_set_intf(struct drm_crtc *crtc, enum mdp4_intf intf, int mixer)
 {
 	struct mdp4_crtc *mdp4_crtc = to_mdp4_crtc(crtc);
 	struct mdp4_kms *mdp4_kms = get_kms(crtc);
@@ -698,15 +703,13 @@ void mdp4_crtc_set_intf(struct drm_crtc *crtc, enum mdp4_intf intf)
 	if (intf == INTF_DSI_VIDEO) {
 		intf_sel &= ~MDP4_DISP_INTF_SEL_DSI_CMD;
 		intf_sel |= MDP4_DISP_INTF_SEL_DSI_VIDEO;
-		mdp4_crtc->mixer = 0;
 	} else if (intf == INTF_DSI_CMD) {
 		intf_sel &= ~MDP4_DISP_INTF_SEL_DSI_VIDEO;
 		intf_sel |= MDP4_DISP_INTF_SEL_DSI_CMD;
-		mdp4_crtc->mixer = 0;
-	} else if (intf == INTF_LCDC_DTV){
-		mdp4_crtc->mixer = 1;
 	}
 
+	mdp4_crtc->mixer = mixer;
+
 	blend_setup(crtc);
 
 	DBG("%s: intf_sel=%08x", mdp4_crtc->name, intf_sel);
diff --git a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_dtv_encoder.c b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_dtv_encoder.c
index 067ed03b35fe..c3878420180b 100644
--- a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_dtv_encoder.c
+++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_dtv_encoder.c
@@ -233,7 +233,7 @@ static void mdp4_dtv_encoder_commit(struct drm_encoder *encoder)
 			MDP4_DMA_CONFIG_G_BPC(BPC8) |
 			MDP4_DMA_CONFIG_B_BPC(BPC8) |
 			MDP4_DMA_CONFIG_PACK(0x21));
-	mdp4_crtc_set_intf(encoder->crtc, INTF_LCDC_DTV);
+	mdp4_crtc_set_intf(encoder->crtc, INTF_LCDC_DTV, 1);
 	mdp4_dtv_encoder_dpms(encoder, DRM_MODE_DPMS_ON);
 }
 
diff --git a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.c b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.c
index 733646c0d3f8..79d804e61cc4 100644
--- a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.c
+++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.c
@@ -106,6 +106,7 @@ static int mdp4_hw_init(struct msm_kms *kms)
 
 	if (mdp4_kms->rev >= 2)
 		mdp4_write(mdp4_kms, REG_MDP4_LAYERMIXER_IN_CFG_UPDATE_METHOD, 1);
+	mdp4_write(mdp4_kms, REG_MDP4_LAYERMIXER_IN_CFG, 0);
 
 	/* disable CSC matrix / YUV by default: */
 	mdp4_write(mdp4_kms, REG_MDP4_PIPE_OP_MODE(VG1), 0);
@@ -196,6 +197,28 @@ int mdp4_enable(struct mdp4_kms *mdp4_kms)
 	return 0;
 }
 
+#ifdef CONFIG_OF
+static struct drm_panel *detect_panel(struct drm_device *dev, const char *name)
+{
+	struct device_node *n;
+	struct drm_panel *panel = NULL;
+
+	n = of_parse_phandle(dev->dev->of_node, name, 0);
+	if (n) {
+		panel = of_drm_find_panel(n);
+		if (!panel)
+			panel = ERR_PTR(-EPROBE_DEFER);
+	}
+
+	return panel;
+}
+#else
+static struct drm_panel *detect_panel(struct drm_device *dev, const char *name)
+{
+	// ??? maybe use a module param to specify which panel is attached?
+}
+#endif
+
 static int modeset_init(struct mdp4_kms *mdp4_kms)
 {
 	struct drm_device *dev = mdp4_kms->dev;
@@ -203,14 +226,11 @@ static int modeset_init(struct mdp4_kms *mdp4_kms)
 	struct drm_plane *plane;
 	struct drm_crtc *crtc;
 	struct drm_encoder *encoder;
+	struct drm_connector *connector;
+	struct drm_panel *panel;
 	struct hdmi *hdmi;
 	int ret;
 
-	/*
-	 *  NOTE: this is a bit simplistic until we add support
-	 * for more than just RGB1->DMA_E->DTV->HDMI
-	 */
-
 	/* construct non-private planes: */
 	plane = mdp4_plane_init(dev, VG1, false);
 	if (IS_ERR(plane)) {
@@ -228,7 +248,57 @@ static int modeset_init(struct mdp4_kms *mdp4_kms)
 	}
 	priv->planes[priv->num_planes++] = plane;
 
-	/* the CRTCs get constructed with a private plane: */
+	/*
+	 * Setup the LCDC/LVDS path: RGB2 -> DMA_P -> LCDC -> LVDS:
+	 */
+
+	panel = detect_panel(dev, "qcom,lvds-panel");
+	if (IS_ERR(panel)) {
+		ret = PTR_ERR(panel);
+		dev_err(dev->dev, "failed to detect LVDS panel: %d\n", ret);
+		goto fail;
+	}
+
+	plane = mdp4_plane_init(dev, RGB2, true);
+	if (IS_ERR(plane)) {
+		dev_err(dev->dev, "failed to construct plane for RGB2\n");
+		ret = PTR_ERR(plane);
+		goto fail;
+	}
+
+	crtc  = mdp4_crtc_init(dev, plane, priv->num_crtcs, 0, DMA_P);
+	if (IS_ERR(crtc)) {
+		dev_err(dev->dev, "failed to construct crtc for DMA_P\n");
+		ret = PTR_ERR(crtc);
+		goto fail;
+	}
+
+	encoder = mdp4_lcdc_encoder_init(dev, panel);
+	if (IS_ERR(encoder)) {
+		dev_err(dev->dev, "failed to construct LCDC encoder\n");
+		ret = PTR_ERR(encoder);
+		goto fail;
+	}
+
+	/* LCDC can be hooked to DMA_P: */
+	encoder->possible_crtcs = 1 << priv->num_crtcs;
+
+	priv->crtcs[priv->num_crtcs++] = crtc;
+	priv->encoders[priv->num_encoders++] = encoder;
+
+	connector = mdp4_lvds_connector_init(dev, panel, encoder);
+	if (IS_ERR(connector)) {
+		ret = PTR_ERR(connector);
+		dev_err(dev->dev, "failed to initialize LVDS connector: %d\n", ret);
+		goto fail;
+	}
+
+	priv->connectors[priv->num_connectors++] = connector;
+
+	/*
+	 * Setup DTV/HDMI path: RGB1 -> DMA_E -> DTV -> HDMI:
+	 */
+
 	plane = mdp4_plane_init(dev, RGB1, true);
 	if (IS_ERR(plane)) {
 		dev_err(dev->dev, "failed to construct plane for RGB1\n");
@@ -242,7 +312,6 @@ static int modeset_init(struct mdp4_kms *mdp4_kms)
 		ret = PTR_ERR(crtc);
 		goto fail;
 	}
-	priv->crtcs[priv->num_crtcs++] = crtc;
 
 	encoder = mdp4_dtv_encoder_init(dev);
 	if (IS_ERR(encoder)) {
@@ -250,7 +319,11 @@ static int modeset_init(struct mdp4_kms *mdp4_kms)
 		ret = PTR_ERR(encoder);
 		goto fail;
 	}
-	encoder->possible_crtcs = 0x1;     /* DTV can be hooked to DMA_E */
+
+	/* DTV can be hooked to DMA_E: */
+	encoder->possible_crtcs = 1 << priv->num_crtcs;
+
+	priv->crtcs[priv->num_crtcs++] = crtc;
 	priv->encoders[priv->num_encoders++] = encoder;
 
 	hdmi = hdmi_init(dev, encoder);
diff --git a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.h b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.h
index 3225da804c61..9ff6e7ccfe90 100644
--- a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.h
+++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.h
@@ -23,6 +23,8 @@
 #include "mdp/mdp_kms.h"
 #include "mdp4.xml.h"
 
+#include "drm_panel.h"
+
 struct mdp4_kms {
 	struct mdp_kms base;
 
@@ -30,6 +32,13 @@ struct mdp4_kms {
 
 	int rev;
 
+	/* Shadow value for MDP4_LAYERMIXER_IN_CFG.. since setup for all
+	 * crtcs/encoders is in one shared register, we need to update it
+	 * via read/modify/write.  But to avoid getting confused by power-
+	 * on-default values after resume, use this shadow value instead:
+	 */
+	uint32_t mixer_cfg;
+
 	/* mapper-id used to request GEM buffer mapped for scanout: */
 	int id;
 
@@ -74,7 +83,7 @@ static inline uint32_t pipe2flush(enum mdp4_pipe pipe)
 	case VG1:      return MDP4_OVERLAY_FLUSH_VG1;
 	case VG2:      return MDP4_OVERLAY_FLUSH_VG2;
 	case RGB1:     return MDP4_OVERLAY_FLUSH_RGB1;
-	case RGB2:     return MDP4_OVERLAY_FLUSH_RGB1;
+	case RGB2:     return MDP4_OVERLAY_FLUSH_RGB2;
 	default:       return 0;
 	}
 }
@@ -108,38 +117,50 @@ static inline uint32_t dma2err(enum mdp4_dma dma)
 	}
 }
 
-static inline uint32_t mixercfg(int mixer, enum mdp4_pipe pipe,
-		enum mdp_mixer_stage_id stage)
+static inline uint32_t mixercfg(uint32_t mixer_cfg, int mixer,
+		enum mdp4_pipe pipe, enum mdp_mixer_stage_id stage)
 {
-	uint32_t mixer_cfg = 0;
-
 	switch (pipe) {
 	case VG1:
-		mixer_cfg = MDP4_LAYERMIXER_IN_CFG_PIPE0(stage) |
+		mixer_cfg &= ~(MDP4_LAYERMIXER_IN_CFG_PIPE0__MASK |
+				MDP4_LAYERMIXER_IN_CFG_PIPE0_MIXER1);
+		mixer_cfg |= MDP4_LAYERMIXER_IN_CFG_PIPE0(stage) |
 			COND(mixer == 1, MDP4_LAYERMIXER_IN_CFG_PIPE0_MIXER1);
 		break;
 	case VG2:
-		mixer_cfg = MDP4_LAYERMIXER_IN_CFG_PIPE1(stage) |
+		mixer_cfg &= ~(MDP4_LAYERMIXER_IN_CFG_PIPE1__MASK |
+				MDP4_LAYERMIXER_IN_CFG_PIPE1_MIXER1);
+		mixer_cfg |= MDP4_LAYERMIXER_IN_CFG_PIPE1(stage) |
 			COND(mixer == 1, MDP4_LAYERMIXER_IN_CFG_PIPE1_MIXER1);
 		break;
 	case RGB1:
-		mixer_cfg = MDP4_LAYERMIXER_IN_CFG_PIPE2(stage) |
+		mixer_cfg &= ~(MDP4_LAYERMIXER_IN_CFG_PIPE2__MASK |
+				MDP4_LAYERMIXER_IN_CFG_PIPE2_MIXER1);
+		mixer_cfg |= MDP4_LAYERMIXER_IN_CFG_PIPE2(stage) |
 			COND(mixer == 1, MDP4_LAYERMIXER_IN_CFG_PIPE2_MIXER1);
 		break;
 	case RGB2:
-		mixer_cfg = MDP4_LAYERMIXER_IN_CFG_PIPE3(stage) |
+		mixer_cfg &= ~(MDP4_LAYERMIXER_IN_CFG_PIPE3__MASK |
+				MDP4_LAYERMIXER_IN_CFG_PIPE3_MIXER1);
+		mixer_cfg |= MDP4_LAYERMIXER_IN_CFG_PIPE3(stage) |
 			COND(mixer == 1, MDP4_LAYERMIXER_IN_CFG_PIPE3_MIXER1);
 		break;
 	case RGB3:
-		mixer_cfg = MDP4_LAYERMIXER_IN_CFG_PIPE4(stage) |
+		mixer_cfg &= ~(MDP4_LAYERMIXER_IN_CFG_PIPE4__MASK |
+				MDP4_LAYERMIXER_IN_CFG_PIPE4_MIXER1);
+		mixer_cfg |= MDP4_LAYERMIXER_IN_CFG_PIPE4(stage) |
 			COND(mixer == 1, MDP4_LAYERMIXER_IN_CFG_PIPE4_MIXER1);
 		break;
 	case VG3:
-		mixer_cfg = MDP4_LAYERMIXER_IN_CFG_PIPE5(stage) |
+		mixer_cfg &= ~(MDP4_LAYERMIXER_IN_CFG_PIPE5__MASK |
+				MDP4_LAYERMIXER_IN_CFG_PIPE5_MIXER1);
+		mixer_cfg |= MDP4_LAYERMIXER_IN_CFG_PIPE5(stage) |
 			COND(mixer == 1, MDP4_LAYERMIXER_IN_CFG_PIPE5_MIXER1);
 		break;
 	case VG4:
-		mixer_cfg = MDP4_LAYERMIXER_IN_CFG_PIPE6(stage) |
+		mixer_cfg &= ~(MDP4_LAYERMIXER_IN_CFG_PIPE6__MASK |
+				MDP4_LAYERMIXER_IN_CFG_PIPE6_MIXER1);
+		mixer_cfg |= MDP4_LAYERMIXER_IN_CFG_PIPE6(stage) |
 			COND(mixer == 1, MDP4_LAYERMIXER_IN_CFG_PIPE6_MIXER1);
 		break;
 	default:
@@ -188,7 +209,7 @@ struct drm_plane *mdp4_plane_init(struct drm_device *dev,
 uint32_t mdp4_crtc_vblank(struct drm_crtc *crtc);
 void mdp4_crtc_cancel_pending_flip(struct drm_crtc *crtc, struct drm_file *file);
 void mdp4_crtc_set_config(struct drm_crtc *crtc, uint32_t config);
-void mdp4_crtc_set_intf(struct drm_crtc *crtc, enum mdp4_intf intf);
+void mdp4_crtc_set_intf(struct drm_crtc *crtc, enum mdp4_intf intf, int mixer);
 void mdp4_crtc_attach(struct drm_crtc *crtc, struct drm_plane *plane);
 void mdp4_crtc_detach(struct drm_crtc *crtc, struct drm_plane *plane);
 struct drm_crtc *mdp4_crtc_init(struct drm_device *dev,
@@ -198,6 +219,22 @@ struct drm_crtc *mdp4_crtc_init(struct drm_device *dev,
 long mdp4_dtv_round_pixclk(struct drm_encoder *encoder, unsigned long rate);
 struct drm_encoder *mdp4_dtv_encoder_init(struct drm_device *dev);
 
+long mdp4_lcdc_round_pixclk(struct drm_encoder *encoder, unsigned long rate);
+struct drm_encoder *mdp4_lcdc_encoder_init(struct drm_device *dev,
+		struct drm_panel *panel);
+
+struct drm_connector *mdp4_lvds_connector_init(struct drm_device *dev,
+		struct drm_panel *panel, struct drm_encoder *encoder);
+
+#ifdef CONFIG_COMMON_CLK
+struct clk *mpd4_lvds_pll_init(struct drm_device *dev);
+#else
+static inline struct clk *mpd4_lvds_pll_init(struct drm_device *dev)
+{
+	return ERR_PTR(-ENODEV);
+}
+#endif
+
 #ifdef CONFIG_MSM_BUS_SCALING
 static inline int match_dev_name(struct device *dev, void *data)
 {
diff --git a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_lcdc_encoder.c b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_lcdc_encoder.c
new file mode 100644
index 000000000000..41f6436754fc
--- /dev/null
+++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_lcdc_encoder.c
@@ -0,0 +1,506 @@
+/*
+ * Copyright (C) 2014 Red Hat
+ * Author: Rob Clark <robdclark@gmail.com>
+ * Author: Vinay Simha <vinaysimha@inforcecomputing.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include "mdp4_kms.h"
+
+#include "drm_crtc.h"
+#include "drm_crtc_helper.h"
+
+struct mdp4_lcdc_encoder {
+	struct drm_encoder base;
+	struct drm_panel *panel;
+	struct clk *lcdc_clk;
+	unsigned long int pixclock;
+	struct regulator *regs[3];
+	bool enabled;
+	uint32_t bsc;
+};
+#define to_mdp4_lcdc_encoder(x) container_of(x, struct mdp4_lcdc_encoder, base)
+
+static struct mdp4_kms *get_kms(struct drm_encoder *encoder)
+{
+	struct msm_drm_private *priv = encoder->dev->dev_private;
+	return to_mdp4_kms(to_mdp_kms(priv->kms));
+}
+
+#ifdef CONFIG_MSM_BUS_SCALING
+#include <mach/board.h>
+static void bs_init(struct mdp4_lcdc_encoder *mdp4_lcdc_encoder)
+{
+	struct drm_device *dev = mdp4_lcdc_encoder->base.dev;
+	struct lcdc_platform_data *lcdc_pdata = mdp4_find_pdata("lvds.0");
+
+	if (!lcdc_pdata) {
+		dev_err(dev->dev, "could not find lvds pdata\n");
+		return;
+	}
+
+	if (lcdc_pdata->bus_scale_table) {
+		mdp4_lcdc_encoder->bsc = msm_bus_scale_register_client(
+				lcdc_pdata->bus_scale_table);
+		DBG("lvds : bus scale client: %08x", mdp4_lcdc_encoder->bsc);
+	}
+}
+
+static void bs_fini(struct mdp4_lcdc_encoder *mdp4_lcdc_encoder)
+{
+	if (mdp4_lcdc_encoder->bsc) {
+		msm_bus_scale_unregister_client(mdp4_lcdc_encoder->bsc);
+		mdp4_lcdc_encoder->bsc = 0;
+	}
+}
+
+static void bs_set(struct mdp4_lcdc_encoder *mdp4_lcdc_encoder, int idx)
+{
+	if (mdp4_lcdc_encoder->bsc) {
+		DBG("set bus scaling: %d", idx);
+		msm_bus_scale_client_update_request(mdp4_lcdc_encoder->bsc, idx);
+	}
+}
+#else
+static void bs_init(struct mdp4_lcdc_encoder *mdp4_lcdc_encoder) {}
+static void bs_fini(struct mdp4_lcdc_encoder *mdp4_lcdc_encoder) {}
+static void bs_set(struct mdp4_lcdc_encoder *mdp4_lcdc_encoder, int idx) {}
+#endif
+
+static void mdp4_lcdc_encoder_destroy(struct drm_encoder *encoder)
+{
+	struct mdp4_lcdc_encoder *mdp4_lcdc_encoder =
+			to_mdp4_lcdc_encoder(encoder);
+	bs_fini(mdp4_lcdc_encoder);
+	drm_encoder_cleanup(encoder);
+	kfree(mdp4_lcdc_encoder);
+}
+
+static const struct drm_encoder_funcs mdp4_lcdc_encoder_funcs = {
+	.destroy = mdp4_lcdc_encoder_destroy,
+};
+
+/* this should probably be a helper: */
+struct drm_connector *get_connector(struct drm_encoder *encoder)
+{
+	struct drm_device *dev = encoder->dev;
+	struct drm_connector *connector;
+
+	list_for_each_entry(connector, &dev->mode_config.connector_list, head)
+		if (connector->encoder == encoder)
+			return connector;
+
+	return NULL;
+}
+
+static void setup_phy(struct drm_encoder *encoder)
+{
+	struct drm_device *dev = encoder->dev;
+	struct drm_connector *connector = get_connector(encoder);
+	struct mdp4_kms *mdp4_kms = get_kms(encoder);
+	uint32_t lvds_intf = 0, lvds_phy_cfg0 = 0;
+	int bpp, nchan, swap;
+
+	if (!connector)
+		return;
+
+	bpp = 3 * connector->display_info.bpc;
+
+	if (!bpp)
+		bpp = 18;
+
+	/* TODO, these should come from panel somehow: */
+	nchan = 1;
+	swap = 0;
+
+	switch (bpp) {
+	case 24:
+		mdp4_write(mdp4_kms, REG_MDP4_LCDC_LVDS_MUX_CTL_3_TO_0(0),
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT0(0x08) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT1(0x05) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT2(0x04) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT3(0x03));
+		mdp4_write(mdp4_kms, REG_MDP4_LCDC_LVDS_MUX_CTL_6_TO_4(0),
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT4(0x02) |
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT5(0x01) |
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT6(0x00));
+		mdp4_write(mdp4_kms, REG_MDP4_LCDC_LVDS_MUX_CTL_3_TO_0(1),
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT0(0x11) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT1(0x10) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT2(0x0d) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT3(0x0c));
+		mdp4_write(mdp4_kms, REG_MDP4_LCDC_LVDS_MUX_CTL_6_TO_4(1),
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT4(0x0b) |
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT5(0x0a) |
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT6(0x09));
+		mdp4_write(mdp4_kms, REG_MDP4_LCDC_LVDS_MUX_CTL_3_TO_0(2),
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT0(0x1a) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT1(0x19) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT2(0x18) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT3(0x15));
+		mdp4_write(mdp4_kms, REG_MDP4_LCDC_LVDS_MUX_CTL_6_TO_4(2),
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT4(0x14) |
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT5(0x13) |
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT6(0x12));
+		mdp4_write(mdp4_kms, REG_MDP4_LCDC_LVDS_MUX_CTL_3_TO_0(3),
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT0(0x1b) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT1(0x17) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT2(0x16) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT3(0x0f));
+		mdp4_write(mdp4_kms, REG_MDP4_LCDC_LVDS_MUX_CTL_6_TO_4(3),
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT4(0x0e) |
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT5(0x07) |
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT6(0x06));
+		if (nchan == 2) {
+			lvds_intf |= MDP4_LCDC_LVDS_INTF_CTL_CH2_DATA_LANE3_EN |
+					MDP4_LCDC_LVDS_INTF_CTL_CH2_DATA_LANE2_EN |
+					MDP4_LCDC_LVDS_INTF_CTL_CH2_DATA_LANE1_EN |
+					MDP4_LCDC_LVDS_INTF_CTL_CH2_DATA_LANE0_EN |
+					MDP4_LCDC_LVDS_INTF_CTL_CH1_DATA_LANE3_EN |
+					MDP4_LCDC_LVDS_INTF_CTL_CH1_DATA_LANE2_EN |
+					MDP4_LCDC_LVDS_INTF_CTL_CH1_DATA_LANE1_EN |
+					MDP4_LCDC_LVDS_INTF_CTL_CH1_DATA_LANE0_EN;
+		} else {
+			lvds_intf |= MDP4_LCDC_LVDS_INTF_CTL_CH1_DATA_LANE3_EN |
+					MDP4_LCDC_LVDS_INTF_CTL_CH1_DATA_LANE2_EN |
+					MDP4_LCDC_LVDS_INTF_CTL_CH1_DATA_LANE1_EN |
+					MDP4_LCDC_LVDS_INTF_CTL_CH1_DATA_LANE0_EN;
+		}
+		break;
+
+	case 18:
+		mdp4_write(mdp4_kms, REG_MDP4_LCDC_LVDS_MUX_CTL_3_TO_0(0),
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT0(0x0a) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT1(0x07) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT2(0x06) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT3(0x05));
+		mdp4_write(mdp4_kms, REG_MDP4_LCDC_LVDS_MUX_CTL_6_TO_4(0),
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT4(0x04) |
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT5(0x03) |
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT6(0x02));
+		mdp4_write(mdp4_kms, REG_MDP4_LCDC_LVDS_MUX_CTL_3_TO_0(1),
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT0(0x13) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT1(0x12) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT2(0x0f) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT3(0x0e));
+		mdp4_write(mdp4_kms, REG_MDP4_LCDC_LVDS_MUX_CTL_6_TO_4(1),
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT4(0x0d) |
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT5(0x0c) |
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT6(0x0b));
+		mdp4_write(mdp4_kms, REG_MDP4_LCDC_LVDS_MUX_CTL_3_TO_0(2),
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT0(0x1a) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT1(0x19) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT2(0x18) |
+				MDP4_LCDC_LVDS_MUX_CTL_3_TO_0_BIT3(0x17));
+		mdp4_write(mdp4_kms, REG_MDP4_LCDC_LVDS_MUX_CTL_6_TO_4(2),
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT4(0x16) |
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT5(0x15) |
+				MDP4_LCDC_LVDS_MUX_CTL_6_TO_4_BIT6(0x14));
+		if (nchan == 2) {
+			lvds_intf |= MDP4_LCDC_LVDS_INTF_CTL_CH2_DATA_LANE2_EN |
+					MDP4_LCDC_LVDS_INTF_CTL_CH2_DATA_LANE1_EN |
+					MDP4_LCDC_LVDS_INTF_CTL_CH2_DATA_LANE0_EN |
+					MDP4_LCDC_LVDS_INTF_CTL_CH1_DATA_LANE2_EN |
+					MDP4_LCDC_LVDS_INTF_CTL_CH1_DATA_LANE1_EN |
+					MDP4_LCDC_LVDS_INTF_CTL_CH1_DATA_LANE0_EN;
+		} else {
+			lvds_intf |= MDP4_LCDC_LVDS_INTF_CTL_CH1_DATA_LANE2_EN |
+					MDP4_LCDC_LVDS_INTF_CTL_CH1_DATA_LANE1_EN |
+					MDP4_LCDC_LVDS_INTF_CTL_CH1_DATA_LANE0_EN;
+		}
+		lvds_intf |= MDP4_LCDC_LVDS_INTF_CTL_RGB_OUT;
+		break;
+
+	default:
+		dev_err(dev->dev, "unknown bpp: %d\n", bpp);
+		return;
+	}
+
+	switch (nchan) {
+	case 1:
+		lvds_phy_cfg0 = MDP4_LVDS_PHY_CFG0_CHANNEL0;
+		lvds_intf |= MDP4_LCDC_LVDS_INTF_CTL_CH1_CLK_LANE_EN |
+				MDP4_LCDC_LVDS_INTF_CTL_MODE_SEL;
+		break;
+	case 2:
+		lvds_phy_cfg0 = MDP4_LVDS_PHY_CFG0_CHANNEL0 |
+				MDP4_LVDS_PHY_CFG0_CHANNEL1;
+		lvds_intf |= MDP4_LCDC_LVDS_INTF_CTL_CH2_CLK_LANE_EN |
+				MDP4_LCDC_LVDS_INTF_CTL_CH1_CLK_LANE_EN;
+		break;
+	default:
+		dev_err(dev->dev, "unknown # of channels: %d\n", nchan);
+		return;
+	}
+
+	if (swap)
+		lvds_intf |= MDP4_LCDC_LVDS_INTF_CTL_CH_SWAP;
+
+	lvds_intf |= MDP4_LCDC_LVDS_INTF_CTL_ENABLE;
+
+	mdp4_write(mdp4_kms, REG_MDP4_LVDS_PHY_CFG0, lvds_phy_cfg0);
+	mdp4_write(mdp4_kms, REG_MDP4_LCDC_LVDS_INTF_CTL, lvds_intf);
+	mdp4_write(mdp4_kms, REG_MDP4_LVDS_PHY_CFG2, 0x30);
+
+	mb();
+	udelay(1);
+	lvds_phy_cfg0 |= MDP4_LVDS_PHY_CFG0_SERIALIZATION_ENBLE;
+	mdp4_write(mdp4_kms, REG_MDP4_LVDS_PHY_CFG0, lvds_phy_cfg0);
+}
+
+static void mdp4_lcdc_encoder_dpms(struct drm_encoder *encoder, int mode)
+{
+	struct drm_device *dev = encoder->dev;
+	struct mdp4_lcdc_encoder *mdp4_lcdc_encoder =
+			to_mdp4_lcdc_encoder(encoder);
+	struct mdp4_kms *mdp4_kms = get_kms(encoder);
+	struct drm_panel *panel = mdp4_lcdc_encoder->panel;
+	bool enabled = (mode == DRM_MODE_DPMS_ON);
+	int i, ret;
+
+	DBG("mode=%d", mode);
+
+	if (enabled == mdp4_lcdc_encoder->enabled)
+		return;
+
+	if (enabled) {
+		unsigned long pc = mdp4_lcdc_encoder->pixclock;
+		int ret;
+
+		bs_set(mdp4_lcdc_encoder, 1);
+
+		for (i = 0; i < ARRAY_SIZE(mdp4_lcdc_encoder->regs); i++) {
+			ret = regulator_enable(mdp4_lcdc_encoder->regs[i]);
+			if (ret)
+				dev_err(dev->dev, "failed to enable regulator: %d\n", ret);
+		}
+
+		DBG("setting lcdc_clk=%lu", pc);
+		ret = clk_set_rate(mdp4_lcdc_encoder->lcdc_clk, pc);
+		if (ret)
+			dev_err(dev->dev, "failed to configure lcdc_clk: %d\n", ret);
+		ret = clk_prepare_enable(mdp4_lcdc_encoder->lcdc_clk);
+		if (ret)
+			dev_err(dev->dev, "failed to enable lcdc_clk: %d\n", ret);
+
+		if (panel)
+			drm_panel_enable(panel);
+
+		setup_phy(encoder);
+
+		mdp4_write(mdp4_kms, REG_MDP4_LCDC_ENABLE, 1);
+	} else {
+		mdp4_write(mdp4_kms, REG_MDP4_LCDC_ENABLE, 0);
+
+		if (panel)
+			drm_panel_disable(panel);
+
+		/*
+		 * Wait for a vsync so we know the ENABLE=0 latched before
+		 * the (connector) source of the vsync's gets disabled,
+		 * otherwise we end up in a funny state if we re-enable
+		 * before the disable latches, which results that some of
+		 * the settings changes for the new modeset (like new
+		 * scanout buffer) don't latch properly..
+		 */
+		mdp_irq_wait(&mdp4_kms->base, MDP4_IRQ_PRIMARY_VSYNC);
+
+		clk_disable_unprepare(mdp4_lcdc_encoder->lcdc_clk);
+
+		for (i = 0; i < ARRAY_SIZE(mdp4_lcdc_encoder->regs); i++) {
+			ret = regulator_disable(mdp4_lcdc_encoder->regs[i]);
+			if (ret)
+				dev_err(dev->dev, "failed to disable regulator: %d\n", ret);
+		}
+
+		bs_set(mdp4_lcdc_encoder, 0);
+	}
+
+	mdp4_lcdc_encoder->enabled = enabled;
+}
+
+static bool mdp4_lcdc_encoder_mode_fixup(struct drm_encoder *encoder,
+		const struct drm_display_mode *mode,
+		struct drm_display_mode *adjusted_mode)
+{
+	return true;
+}
+
+static void mdp4_lcdc_encoder_mode_set(struct drm_encoder *encoder,
+		struct drm_display_mode *mode,
+		struct drm_display_mode *adjusted_mode)
+{
+	struct mdp4_lcdc_encoder *mdp4_lcdc_encoder =
+			to_mdp4_lcdc_encoder(encoder);
+	struct mdp4_kms *mdp4_kms = get_kms(encoder);
+	uint32_t lcdc_hsync_skew, vsync_period, vsync_len, ctrl_pol;
+	uint32_t display_v_start, display_v_end;
+	uint32_t hsync_start_x, hsync_end_x;
+
+	mode = adjusted_mode;
+
+	DBG("set mode: %d:\"%s\" %d %d %d %d %d %d %d %d %d %d 0x%x 0x%x",
+			mode->base.id, mode->name,
+			mode->vrefresh, mode->clock,
+			mode->hdisplay, mode->hsync_start,
+			mode->hsync_end, mode->htotal,
+			mode->vdisplay, mode->vsync_start,
+			mode->vsync_end, mode->vtotal,
+			mode->type, mode->flags);
+
+	mdp4_lcdc_encoder->pixclock = mode->clock * 1000;
+
+	DBG("pixclock=%lu", mdp4_lcdc_encoder->pixclock);
+
+	ctrl_pol = 0;
+	if (mode->flags & DRM_MODE_FLAG_NHSYNC)
+		ctrl_pol |= MDP4_LCDC_CTRL_POLARITY_HSYNC_LOW;
+	if (mode->flags & DRM_MODE_FLAG_NVSYNC)
+		ctrl_pol |= MDP4_LCDC_CTRL_POLARITY_VSYNC_LOW;
+	/* probably need to get DATA_EN polarity from panel.. */
+
+	lcdc_hsync_skew = 0;  /* get this from panel? */
+
+	hsync_start_x = (mode->htotal - mode->hsync_start);
+	hsync_end_x = mode->htotal - (mode->hsync_start - mode->hdisplay) - 1;
+
+	vsync_period = mode->vtotal * mode->htotal;
+	vsync_len = (mode->vsync_end - mode->vsync_start) * mode->htotal;
+	display_v_start = (mode->vtotal - mode->vsync_start) * mode->htotal + lcdc_hsync_skew;
+	display_v_end = vsync_period - ((mode->vsync_start - mode->vdisplay) * mode->htotal) + lcdc_hsync_skew - 1;
+
+	mdp4_write(mdp4_kms, REG_MDP4_LCDC_HSYNC_CTRL,
+			MDP4_LCDC_HSYNC_CTRL_PULSEW(mode->hsync_end - mode->hsync_start) |
+			MDP4_LCDC_HSYNC_CTRL_PERIOD(mode->htotal));
+	mdp4_write(mdp4_kms, REG_MDP4_LCDC_VSYNC_PERIOD, vsync_period);
+	mdp4_write(mdp4_kms, REG_MDP4_LCDC_VSYNC_LEN, vsync_len);
+	mdp4_write(mdp4_kms, REG_MDP4_LCDC_DISPLAY_HCTRL,
+			MDP4_LCDC_DISPLAY_HCTRL_START(hsync_start_x) |
+			MDP4_LCDC_DISPLAY_HCTRL_END(hsync_end_x));
+	mdp4_write(mdp4_kms, REG_MDP4_LCDC_DISPLAY_VSTART, display_v_start);
+	mdp4_write(mdp4_kms, REG_MDP4_LCDC_DISPLAY_VEND, display_v_end);
+	mdp4_write(mdp4_kms, REG_MDP4_LCDC_BORDER_CLR, 0);
+	mdp4_write(mdp4_kms, REG_MDP4_LCDC_UNDERFLOW_CLR,
+			MDP4_LCDC_UNDERFLOW_CLR_ENABLE_RECOVERY |
+			MDP4_LCDC_UNDERFLOW_CLR_COLOR(0xff));
+	mdp4_write(mdp4_kms, REG_MDP4_LCDC_HSYNC_SKEW, lcdc_hsync_skew);
+	mdp4_write(mdp4_kms, REG_MDP4_LCDC_CTRL_POLARITY, ctrl_pol);
+	mdp4_write(mdp4_kms, REG_MDP4_LCDC_ACTIVE_HCTL,
+			MDP4_LCDC_ACTIVE_HCTL_START(0) |
+			MDP4_LCDC_ACTIVE_HCTL_END(0));
+	mdp4_write(mdp4_kms, REG_MDP4_LCDC_ACTIVE_VSTART, 0);
+	mdp4_write(mdp4_kms, REG_MDP4_LCDC_ACTIVE_VEND, 0);
+}
+
+static void mdp4_lcdc_encoder_prepare(struct drm_encoder *encoder)
+{
+	mdp4_lcdc_encoder_dpms(encoder, DRM_MODE_DPMS_OFF);
+}
+
+static void mdp4_lcdc_encoder_commit(struct drm_encoder *encoder)
+{
+	/* TODO: hard-coded for 18bpp: */
+	mdp4_crtc_set_config(encoder->crtc,
+			MDP4_DMA_CONFIG_R_BPC(BPC6) |
+			MDP4_DMA_CONFIG_G_BPC(BPC6) |
+			MDP4_DMA_CONFIG_B_BPC(BPC6) |
+			MDP4_DMA_CONFIG_PACK_ALIGN_MSB |
+			MDP4_DMA_CONFIG_PACK(0x21) |
+			MDP4_DMA_CONFIG_DEFLKR_EN |
+			MDP4_DMA_CONFIG_DITHER_EN);
+	mdp4_crtc_set_intf(encoder->crtc, INTF_LCDC_DTV, 0);
+	mdp4_lcdc_encoder_dpms(encoder, DRM_MODE_DPMS_ON);
+}
+
+static const struct drm_encoder_helper_funcs mdp4_lcdc_encoder_helper_funcs = {
+	.dpms = mdp4_lcdc_encoder_dpms,
+	.mode_fixup = mdp4_lcdc_encoder_mode_fixup,
+	.mode_set = mdp4_lcdc_encoder_mode_set,
+	.prepare = mdp4_lcdc_encoder_prepare,
+	.commit = mdp4_lcdc_encoder_commit,
+};
+
+long mdp4_lcdc_round_pixclk(struct drm_encoder *encoder, unsigned long rate)
+{
+	struct mdp4_lcdc_encoder *mdp4_lcdc_encoder =
+			to_mdp4_lcdc_encoder(encoder);
+	return clk_round_rate(mdp4_lcdc_encoder->lcdc_clk, rate);
+}
+
+/* initialize encoder */
+struct drm_encoder *mdp4_lcdc_encoder_init(struct drm_device *dev,
+		struct drm_panel *panel)
+{
+	struct drm_encoder *encoder = NULL;
+	struct mdp4_lcdc_encoder *mdp4_lcdc_encoder;
+	struct regulator *reg;
+	int ret;
+
+	mdp4_lcdc_encoder = kzalloc(sizeof(*mdp4_lcdc_encoder), GFP_KERNEL);
+	if (!mdp4_lcdc_encoder) {
+		ret = -ENOMEM;
+		goto fail;
+	}
+
+	mdp4_lcdc_encoder->panel = panel;
+
+	encoder = &mdp4_lcdc_encoder->base;
+
+	drm_encoder_init(dev, encoder, &mdp4_lcdc_encoder_funcs,
+			 DRM_MODE_ENCODER_LVDS);
+	drm_encoder_helper_add(encoder, &mdp4_lcdc_encoder_helper_funcs);
+
+	/* TODO: do we need different pll in other cases? */
+	mdp4_lcdc_encoder->lcdc_clk = mpd4_lvds_pll_init(dev);
+	if (IS_ERR(mdp4_lcdc_encoder->lcdc_clk)) {
+		dev_err(dev->dev, "failed to get lvds_clk\n");
+		ret = PTR_ERR(mdp4_lcdc_encoder->lcdc_clk);
+		goto fail;
+	}
+
+	/* TODO: different regulators in other cases? */
+	reg = devm_regulator_get(dev->dev, "lvds-vccs-3p3v");
+	if (IS_ERR(reg)) {
+		ret = PTR_ERR(reg);
+		dev_err(dev->dev, "failed to get lvds-vccs-3p3v: %d\n", ret);
+		goto fail;
+	}
+	mdp4_lcdc_encoder->regs[0] = reg;
+
+	reg = devm_regulator_get(dev->dev, "lvds-pll-vdda");
+	if (IS_ERR(reg)) {
+		ret = PTR_ERR(reg);
+		dev_err(dev->dev, "failed to get lvds-pll-vdda: %d\n", ret);
+		goto fail;
+	}
+	mdp4_lcdc_encoder->regs[1] = reg;
+
+	reg = devm_regulator_get(dev->dev, "lvds-vdda");
+	if (IS_ERR(reg)) {
+		ret = PTR_ERR(reg);
+		dev_err(dev->dev, "failed to get lvds-vdda: %d\n", ret);
+		goto fail;
+	}
+	mdp4_lcdc_encoder->regs[2] = reg;
+
+	bs_init(mdp4_lcdc_encoder);
+
+	return encoder;
+
+fail:
+	if (encoder)
+		mdp4_lcdc_encoder_destroy(encoder);
+
+	return ERR_PTR(ret);
+}
diff --git a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_lvds_connector.c b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_lvds_connector.c
new file mode 100644
index 000000000000..310034688c15
--- /dev/null
+++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_lvds_connector.c
@@ -0,0 +1,151 @@
+/*
+ * Copyright (C) 2014 Red Hat
+ * Author: Rob Clark <robdclark@gmail.com>
+ * Author: Vinay Simha <vinaysimha@inforcecomputing.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <linux/gpio.h>
+
+#include "mdp4_kms.h"
+
+struct mdp4_lvds_connector {
+	struct drm_connector base;
+	struct drm_encoder *encoder;
+	struct drm_panel *panel;
+};
+#define to_mdp4_lvds_connector(x) container_of(x, struct mdp4_lvds_connector, base)
+
+static enum drm_connector_status mdp4_lvds_connector_detect(
+		struct drm_connector *connector, bool force)
+{
+	struct mdp4_lvds_connector *mdp4_lvds_connector =
+			to_mdp4_lvds_connector(connector);
+
+	return mdp4_lvds_connector->panel ?
+			connector_status_connected :
+			connector_status_disconnected;
+}
+
+static void mdp4_lvds_connector_destroy(struct drm_connector *connector)
+{
+	struct mdp4_lvds_connector *mdp4_lvds_connector =
+			to_mdp4_lvds_connector(connector);
+	struct drm_panel *panel = mdp4_lvds_connector->panel;
+
+	if (panel)
+		drm_panel_detach(panel);
+
+	drm_connector_unregister(connector);
+	drm_connector_cleanup(connector);
+
+	kfree(mdp4_lvds_connector);
+}
+
+static int mdp4_lvds_connector_get_modes(struct drm_connector *connector)
+{
+	struct mdp4_lvds_connector *mdp4_lvds_connector =
+			to_mdp4_lvds_connector(connector);
+	struct drm_panel *panel = mdp4_lvds_connector->panel;
+	int ret = 0;
+
+	if (panel)
+		ret = panel->funcs->get_modes(panel);
+
+	return ret;
+}
+
+static int mdp4_lvds_connector_mode_valid(struct drm_connector *connector,
+				 struct drm_display_mode *mode)
+{
+	struct mdp4_lvds_connector *mdp4_lvds_connector =
+			to_mdp4_lvds_connector(connector);
+	struct drm_encoder *encoder = mdp4_lvds_connector->encoder;
+	long actual, requested;
+
+	requested = 1000 * mode->clock;
+	actual = mdp4_lcdc_round_pixclk(encoder, requested);
+
+	DBG("requested=%ld, actual=%ld", requested, actual);
+
+	if (actual != requested)
+		return MODE_CLOCK_RANGE;
+
+	return MODE_OK;
+}
+
+static struct drm_encoder *
+mdp4_lvds_connector_best_encoder(struct drm_connector *connector)
+{
+	struct mdp4_lvds_connector *mdp4_lvds_connector =
+			to_mdp4_lvds_connector(connector);
+	return mdp4_lvds_connector->encoder;
+}
+
+static const struct drm_connector_funcs mdp4_lvds_connector_funcs = {
+	.dpms = drm_helper_connector_dpms,
+	.detect = mdp4_lvds_connector_detect,
+	.fill_modes = drm_helper_probe_single_connector_modes,
+	.destroy = mdp4_lvds_connector_destroy,
+};
+
+static const struct drm_connector_helper_funcs mdp4_lvds_connector_helper_funcs = {
+	.get_modes = mdp4_lvds_connector_get_modes,
+	.mode_valid = mdp4_lvds_connector_mode_valid,
+	.best_encoder = mdp4_lvds_connector_best_encoder,
+};
+
+/* initialize connector */
+struct drm_connector *mdp4_lvds_connector_init(struct drm_device *dev,
+		struct drm_panel *panel, struct drm_encoder *encoder)
+{
+	struct drm_connector *connector = NULL;
+	struct mdp4_lvds_connector *mdp4_lvds_connector;
+	int ret;
+
+	mdp4_lvds_connector = kzalloc(sizeof(*mdp4_lvds_connector), GFP_KERNEL);
+	if (!mdp4_lvds_connector) {
+		ret = -ENOMEM;
+		goto fail;
+	}
+
+	mdp4_lvds_connector->encoder = encoder;
+	mdp4_lvds_connector->panel = panel;
+
+	connector = &mdp4_lvds_connector->base;
+
+	drm_connector_init(dev, connector, &mdp4_lvds_connector_funcs,
+			DRM_MODE_CONNECTOR_LVDS);
+	drm_connector_helper_add(connector, &mdp4_lvds_connector_helper_funcs);
+
+	connector->polled = 0;
+
+	connector->interlace_allowed = 0;
+	connector->doublescan_allowed = 0;
+
+	drm_connector_register(connector);
+
+	drm_mode_connector_attach_encoder(connector, encoder);
+
+	if (panel)
+		drm_panel_attach(panel, connector);
+
+	return connector;
+
+fail:
+	if (connector)
+		mdp4_lvds_connector_destroy(connector);
+
+	return ERR_PTR(ret);
+}
diff --git a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_lvds_pll.c b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_lvds_pll.c
new file mode 100644
index 000000000000..ce4245971673
--- /dev/null
+++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_lvds_pll.c
@@ -0,0 +1,172 @@
+/*
+ * Copyright (C) 2014 Red Hat
+ * Author: Rob Clark <robdclark@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <linux/clk.h>
+#include <linux/clk-provider.h>
+
+#include "mdp4_kms.h"
+
+struct mdp4_lvds_pll {
+	struct clk_hw pll_hw;
+	struct drm_device *dev;
+	unsigned long pixclk;
+};
+#define to_mdp4_lvds_pll(x) container_of(x, struct mdp4_lvds_pll, pll_hw)
+
+static struct mdp4_kms *get_kms(struct mdp4_lvds_pll *lvds_pll)
+{
+	struct msm_drm_private *priv = lvds_pll->dev->dev_private;
+	return to_mdp4_kms(to_mdp_kms(priv->kms));
+}
+
+struct pll_rate {
+	unsigned long rate;
+	struct {
+		uint32_t val;
+		uint32_t reg;
+	} conf[32];
+};
+
+/* NOTE: keep sorted highest freq to lowest: */
+static const struct pll_rate freqtbl[] = {
+	{ 72000000, {
+		{ 0x8f, REG_MDP4_LVDS_PHY_PLL_CTRL_1 },
+		{ 0x30, REG_MDP4_LVDS_PHY_PLL_CTRL_2 },
+		{ 0xc6, REG_MDP4_LVDS_PHY_PLL_CTRL_3 },
+		{ 0x10, REG_MDP4_LVDS_PHY_PLL_CTRL_5 },
+		{ 0x07, REG_MDP4_LVDS_PHY_PLL_CTRL_6 },
+		{ 0x62, REG_MDP4_LVDS_PHY_PLL_CTRL_7 },
+		{ 0x41, REG_MDP4_LVDS_PHY_PLL_CTRL_8 },
+		{ 0x0d, REG_MDP4_LVDS_PHY_PLL_CTRL_9 },
+		{ 0, 0 } }
+	},
+};
+
+static const struct pll_rate *find_rate(unsigned long rate)
+{
+	int i;
+	for (i = 1; i < ARRAY_SIZE(freqtbl); i++)
+		if (rate > freqtbl[i].rate)
+			return &freqtbl[i-1];
+	return &freqtbl[i-1];
+}
+
+static int mpd4_lvds_pll_enable(struct clk_hw *hw)
+{
+	struct mdp4_lvds_pll *lvds_pll = to_mdp4_lvds_pll(hw);
+	struct mdp4_kms *mdp4_kms = get_kms(lvds_pll);
+	const struct pll_rate *pll_rate = find_rate(lvds_pll->pixclk);
+	int i;
+
+	DBG("pixclk=%lu (%lu)", lvds_pll->pixclk, pll_rate->rate);
+
+	if (WARN_ON(!pll_rate))
+		return -EINVAL;
+
+	mdp4_write(mdp4_kms, REG_MDP4_LCDC_LVDS_PHY_RESET, 0x33);
+
+	for (i = 0; pll_rate->conf[i].reg; i++)
+		mdp4_write(mdp4_kms, pll_rate->conf[i].reg, pll_rate->conf[i].val);
+
+	mdp4_write(mdp4_kms, REG_MDP4_LVDS_PHY_PLL_CTRL_0, 0x01);
+
+	/* Wait until LVDS PLL is locked and ready */
+	while (!mdp4_read(mdp4_kms, REG_MDP4_LVDS_PHY_PLL_LOCKED))
+		cpu_relax();
+
+	return 0;
+}
+
+static void mpd4_lvds_pll_disable(struct clk_hw *hw)
+{
+	struct mdp4_lvds_pll *lvds_pll = to_mdp4_lvds_pll(hw);
+	struct mdp4_kms *mdp4_kms = get_kms(lvds_pll);
+
+	DBG("");
+
+	mdp4_write(mdp4_kms, REG_MDP4_LVDS_PHY_CFG0, 0x0);
+	mdp4_write(mdp4_kms, REG_MDP4_LVDS_PHY_PLL_CTRL_0, 0x0);
+}
+
+static unsigned long mpd4_lvds_pll_recalc_rate(struct clk_hw *hw,
+				unsigned long parent_rate)
+{
+	struct mdp4_lvds_pll *lvds_pll = to_mdp4_lvds_pll(hw);
+	return lvds_pll->pixclk;
+}
+
+static long mpd4_lvds_pll_round_rate(struct clk_hw *hw, unsigned long rate,
+		unsigned long *parent_rate)
+{
+	const struct pll_rate *pll_rate = find_rate(rate);
+	return pll_rate->rate;
+}
+
+static int mpd4_lvds_pll_set_rate(struct clk_hw *hw, unsigned long rate,
+		unsigned long parent_rate)
+{
+	struct mdp4_lvds_pll *lvds_pll = to_mdp4_lvds_pll(hw);
+	lvds_pll->pixclk = rate;
+	return 0;
+}
+
+
+static const struct clk_ops mpd4_lvds_pll_ops = {
+	.enable = mpd4_lvds_pll_enable,
+	.disable = mpd4_lvds_pll_disable,
+	.recalc_rate = mpd4_lvds_pll_recalc_rate,
+	.round_rate = mpd4_lvds_pll_round_rate,
+	.set_rate = mpd4_lvds_pll_set_rate,
+};
+
+static const char *mpd4_lvds_pll_parents[] = {
+	"pxo",
+};
+
+static struct clk_init_data pll_init = {
+	.name = "mpd4_lvds_pll",
+	.ops = &mpd4_lvds_pll_ops,
+	.parent_names = mpd4_lvds_pll_parents,
+	.num_parents = ARRAY_SIZE(mpd4_lvds_pll_parents),
+};
+
+struct clk *mpd4_lvds_pll_init(struct drm_device *dev)
+{
+	struct mdp4_lvds_pll *lvds_pll;
+	struct clk *clk;
+	int ret;
+
+	lvds_pll = devm_kzalloc(dev->dev, sizeof(*lvds_pll), GFP_KERNEL);
+	if (!lvds_pll) {
+		ret = -ENOMEM;
+		goto fail;
+	}
+
+	lvds_pll->dev = dev;
+
+	lvds_pll->pll_hw.init = &pll_init;
+	clk = devm_clk_register(dev->dev, &lvds_pll->pll_hw);
+	if (IS_ERR(clk)) {
+		ret = PTR_ERR(clk);
+		goto fail;
+	}
+
+	return clk;
+
+fail:
+	return ERR_PTR(ret);
+}
diff --git a/drivers/gpu/drm/msm/msm_drv.c b/drivers/gpu/drm/msm/msm_drv.c
index fcf95680413d..b67ef5985125 100644
--- a/drivers/gpu/drm/msm/msm_drv.c
+++ b/drivers/gpu/drm/msm/msm_drv.c
@@ -280,7 +280,7 @@ static int msm_load(struct drm_device *dev, unsigned long flags)
 	dev->mode_config.max_height = 2048;
 	dev->mode_config.funcs = &mode_config_funcs;
 
-	ret = drm_vblank_init(dev, 1);
+	ret = drm_vblank_init(dev, priv->num_crtcs);
 	if (ret < 0) {
 		dev_err(dev->dev, "failed to initialize vblank\n");
 		goto fail;
@@ -315,39 +315,12 @@ static void load_gpu(struct drm_device *dev)
 {
 	static DEFINE_MUTEX(init_lock);
 	struct msm_drm_private *priv = dev->dev_private;
-	struct msm_gpu *gpu;
 
 	mutex_lock(&init_lock);
 
-	if (priv->gpu)
-		goto out;
-
-	gpu = a3xx_gpu_init(dev);
-	if (IS_ERR(gpu)) {
-		dev_warn(dev->dev, "failed to load a3xx gpu\n");
-		gpu = NULL;
-		/* not fatal */
-	}
-
-	if (gpu) {
-		int ret;
-		mutex_lock(&dev->struct_mutex);
-		gpu->funcs->pm_resume(gpu);
-		mutex_unlock(&dev->struct_mutex);
-		ret = gpu->funcs->hw_init(gpu);
-		if (ret) {
-			dev_err(dev->dev, "gpu hw init failed: %d\n", ret);
-			gpu->funcs->destroy(gpu);
-			gpu = NULL;
-		} else {
-			/* give inactive pm a chance to kick in: */
-			msm_gpu_retire(gpu);
-		}
-	}
-
-	priv->gpu = gpu;
+	if (!priv->gpu)
+		priv->gpu = adreno_load_gpu(dev);
 
-out:
 	mutex_unlock(&init_lock);
 }
 
@@ -836,6 +809,7 @@ static struct drm_driver msm_driver = {
 	.open               = msm_open,
 	.preclose           = msm_preclose,
 	.lastclose          = msm_lastclose,
+	.set_busid          = drm_platform_set_busid,
 	.irq_handler        = msm_irq,
 	.irq_preinstall     = msm_irq_preinstall,
 	.irq_postinstall    = msm_irq_postinstall,
@@ -1025,7 +999,7 @@ static int __init msm_drm_register(void)
 {
 	DBG("init");
 	hdmi_register();
-	a3xx_register();
+	adreno_register();
 	return platform_driver_register(&msm_platform_driver);
 }
 
@@ -1034,7 +1008,7 @@ static void __exit msm_drm_unregister(void)
 	DBG("fini");
 	platform_driver_unregister(&msm_platform_driver);
 	hdmi_unregister();
-	a3xx_unregister();
+	adreno_unregister();
 }
 
 module_init(msm_drm_register);
diff --git a/drivers/gpu/drm/msm/msm_drv.h b/drivers/gpu/drm/msm/msm_drv.h
index 8a2c5fd0893e..67f9d0a2332c 100644
--- a/drivers/gpu/drm/msm/msm_drv.h
+++ b/drivers/gpu/drm/msm/msm_drv.h
@@ -51,6 +51,7 @@ static inline struct device *msm_iommu_get_ctx(const char *ctx_name)
 #include <drm/drm_crtc_helper.h>
 #include <drm/drm_fb_helper.h>
 #include <drm/msm_drm.h>
+#include <drm/drm_gem.h>
 
 struct msm_kms;
 struct msm_gpu;
@@ -170,7 +171,7 @@ struct sg_table *msm_gem_prime_get_sg_table(struct drm_gem_object *obj);
 void *msm_gem_prime_vmap(struct drm_gem_object *obj);
 void msm_gem_prime_vunmap(struct drm_gem_object *obj, void *vaddr);
 struct drm_gem_object *msm_gem_prime_import_sg_table(struct drm_device *dev,
-		size_t size, struct sg_table *sg);
+		struct dma_buf_attachment *attach, struct sg_table *sg);
 int msm_gem_prime_pin(struct drm_gem_object *obj);
 void msm_gem_prime_unpin(struct drm_gem_object *obj);
 void *msm_gem_vaddr_locked(struct drm_gem_object *obj);
diff --git a/drivers/gpu/drm/msm/msm_gem_prime.c b/drivers/gpu/drm/msm/msm_gem_prime.c
index d48f9fc5129b..ad772fe36115 100644
--- a/drivers/gpu/drm/msm/msm_gem_prime.c
+++ b/drivers/gpu/drm/msm/msm_gem_prime.c
@@ -18,6 +18,7 @@
 #include "msm_drv.h"
 #include "msm_gem.h"
 
+#include <linux/dma-buf.h>
 
 struct sg_table *msm_gem_prime_get_sg_table(struct drm_gem_object *obj)
 {
@@ -37,9 +38,9 @@ void msm_gem_prime_vunmap(struct drm_gem_object *obj, void *vaddr)
 }
 
 struct drm_gem_object *msm_gem_prime_import_sg_table(struct drm_device *dev,
-		size_t size, struct sg_table *sg)
+		struct dma_buf_attachment *attach, struct sg_table *sg)
 {
-	return msm_gem_import(dev, size, sg);
+	return msm_gem_import(dev, attach->dmabuf->size, sg);
 }
 
 int msm_gem_prime_pin(struct drm_gem_object *obj)
diff --git a/drivers/gpu/drm/msm/msm_gpu.h b/drivers/gpu/drm/msm/msm_gpu.h
index 9b579b792840..fd1e4b4a6d40 100644
--- a/drivers/gpu/drm/msm/msm_gpu.h
+++ b/drivers/gpu/drm/msm/msm_gpu.h
@@ -166,8 +166,8 @@ int msm_gpu_init(struct drm_device *drm, struct platform_device *pdev,
 		const char *name, const char *ioname, const char *irqname, int ringsz);
 void msm_gpu_cleanup(struct msm_gpu *gpu);
 
-struct msm_gpu *a3xx_gpu_init(struct drm_device *dev);
-void __init a3xx_register(void);
-void __exit a3xx_unregister(void);
+struct msm_gpu *adreno_load_gpu(struct drm_device *dev);
+void __init adreno_register(void);
+void __exit adreno_unregister(void);
 
 #endif /* __MSM_GPU_H__ */
diff --git a/drivers/gpu/drm/nouveau/Makefile b/drivers/gpu/drm/nouveau/Makefile
index f5d7f7ce4bc6..12c24c8abf7f 100644
--- a/drivers/gpu/drm/nouveau/Makefile
+++ b/drivers/gpu/drm/nouveau/Makefile
@@ -38,6 +38,7 @@ nouveau-y += core/subdev/bios/dcb.o
 nouveau-y += core/subdev/bios/disp.o
 nouveau-y += core/subdev/bios/dp.o
 nouveau-y += core/subdev/bios/extdev.o
+nouveau-y += core/subdev/bios/fan.o
 nouveau-y += core/subdev/bios/gpio.o
 nouveau-y += core/subdev/bios/i2c.o
 nouveau-y += core/subdev/bios/init.o
@@ -51,6 +52,8 @@ nouveau-y += core/subdev/bios/therm.o
 nouveau-y += core/subdev/bios/vmap.o
 nouveau-y += core/subdev/bios/volt.o
 nouveau-y += core/subdev/bios/xpio.o
+nouveau-y += core/subdev/bios/M0205.o
+nouveau-y += core/subdev/bios/M0209.o
 nouveau-y += core/subdev/bios/P0260.o
 nouveau-y += core/subdev/bus/hwsq.o
 nouveau-y += core/subdev/bus/nv04.o
@@ -124,12 +127,17 @@ nouveau-y += core/subdev/fb/ramnvc0.o
 nouveau-y += core/subdev/fb/ramnve0.o
 nouveau-y += core/subdev/fb/ramgk20a.o
 nouveau-y += core/subdev/fb/ramgm107.o
+nouveau-y += core/subdev/fb/sddr2.o
 nouveau-y += core/subdev/fb/sddr3.o
 nouveau-y += core/subdev/fb/gddr5.o
+nouveau-y += core/subdev/fuse/base.o
+nouveau-y += core/subdev/fuse/g80.o
+nouveau-y += core/subdev/fuse/gf100.o
+nouveau-y += core/subdev/fuse/gm107.o
 nouveau-y += core/subdev/gpio/base.o
 nouveau-y += core/subdev/gpio/nv10.o
 nouveau-y += core/subdev/gpio/nv50.o
-nouveau-y += core/subdev/gpio/nv92.o
+nouveau-y += core/subdev/gpio/nv94.o
 nouveau-y += core/subdev/gpio/nvd0.o
 nouveau-y += core/subdev/gpio/nve0.o
 nouveau-y += core/subdev/i2c/base.o
@@ -190,6 +198,7 @@ nouveau-y += core/subdev/therm/nv50.o
 nouveau-y += core/subdev/therm/nv84.o
 nouveau-y += core/subdev/therm/nva3.o
 nouveau-y += core/subdev/therm/nvd0.o
+nouveau-y += core/subdev/therm/gm107.o
 nouveau-y += core/subdev/timer/base.o
 nouveau-y += core/subdev/timer/nv04.o
 nouveau-y += core/subdev/timer/gk20a.o
@@ -252,6 +261,7 @@ nouveau-y += core/engine/disp/hdanvd0.o
 nouveau-y += core/engine/disp/hdminv84.o
 nouveau-y += core/engine/disp/hdminva3.o
 nouveau-y += core/engine/disp/hdminvd0.o
+nouveau-y += core/engine/disp/hdminve0.o
 nouveau-y += core/engine/disp/piornv50.o
 nouveau-y += core/engine/disp/sornv50.o
 nouveau-y += core/engine/disp/sornv94.o
diff --git a/drivers/gpu/drm/nouveau/core/core/client.c b/drivers/gpu/drm/nouveau/core/core/client.c
index 68bf06768123..e962433294c3 100644
--- a/drivers/gpu/drm/nouveau/core/core/client.c
+++ b/drivers/gpu/drm/nouveau/core/core/client.c
@@ -91,9 +91,10 @@ nvkm_client_notify_del(struct nouveau_client *client, int index)
 }
 
 int
-nvkm_client_notify_new(struct nouveau_client *client,
+nvkm_client_notify_new(struct nouveau_object *object,
 		       struct nvkm_event *event, void *data, u32 size)
 {
+	struct nouveau_client *client = nouveau_client(object);
 	struct nvkm_client_notify *notify;
 	union {
 		struct nvif_notify_req_v0 v0;
@@ -127,8 +128,8 @@ nvkm_client_notify_new(struct nouveau_client *client,
 	}
 
 	if (ret == 0) {
-		ret = nvkm_notify_init(event, nvkm_client_notify, false,
-				       data, size, reply, &notify->n);
+		ret = nvkm_notify_init(object, event, nvkm_client_notify,
+				       false, data, size, reply, &notify->n);
 		if (ret == 0) {
 			client->notify[index] = notify;
 			notify->client = client;
diff --git a/drivers/gpu/drm/nouveau/core/core/event.c b/drivers/gpu/drm/nouveau/core/core/event.c
index 0540a48c5678..ff2b434b3db4 100644
--- a/drivers/gpu/drm/nouveau/core/core/event.c
+++ b/drivers/gpu/drm/nouveau/core/core/event.c
@@ -20,7 +20,7 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  */
 
-#include <core/os.h>
+#include <core/object.h>
 #include <core/event.h>
 
 void
diff --git a/drivers/gpu/drm/nouveau/core/core/gpuobj.c b/drivers/gpu/drm/nouveau/core/core/gpuobj.c
index 560b2214cf1c..daee87702502 100644
--- a/drivers/gpu/drm/nouveau/core/core/gpuobj.c
+++ b/drivers/gpu/drm/nouveau/core/core/gpuobj.c
@@ -115,7 +115,7 @@ nouveau_gpuobj_create_(struct nouveau_object *parent,
 	gpuobj->size = size;
 
 	if (heap) {
-		ret = nouveau_mm_head(heap, 1, size, size,
+		ret = nouveau_mm_head(heap, 0, 1, size, size,
 				      max(align, (u32)1), &gpuobj->node);
 		if (ret)
 			return ret;
diff --git a/drivers/gpu/drm/nouveau/core/core/ioctl.c b/drivers/gpu/drm/nouveau/core/core/ioctl.c
index f7e19bfb489c..692aa92dd850 100644
--- a/drivers/gpu/drm/nouveau/core/core/ioctl.c
+++ b/drivers/gpu/drm/nouveau/core/core/ioctl.c
@@ -349,7 +349,6 @@ nvkm_ioctl_unmap(struct nouveau_handle *handle, void *data, u32 size)
 static int
 nvkm_ioctl_ntfy_new(struct nouveau_handle *handle, void *data, u32 size)
 {
-	struct nouveau_client *client = nouveau_client(handle->object);
 	struct nouveau_object *object = handle->object;
 	struct nouveau_ofuncs *ofuncs = object->oclass->ofuncs;
 	union {
@@ -365,7 +364,7 @@ nvkm_ioctl_ntfy_new(struct nouveau_handle *handle, void *data, u32 size)
 		if (ret = -ENODEV, ofuncs->ntfy)
 			ret = ofuncs->ntfy(object, args->v0.event, &event);
 		if (ret == 0) {
-			ret = nvkm_client_notify_new(client, event, data, size);
+			ret = nvkm_client_notify_new(object, event, data, size);
 			if (ret >= 0) {
 				args->v0.index = ret;
 				ret = 0;
diff --git a/drivers/gpu/drm/nouveau/core/core/mm.c b/drivers/gpu/drm/nouveau/core/core/mm.c
index 7a4e0891c5f8..b4f5db66d5b5 100644
--- a/drivers/gpu/drm/nouveau/core/core/mm.c
+++ b/drivers/gpu/drm/nouveau/core/core/mm.c
@@ -28,6 +28,24 @@
 #define node(root, dir) ((root)->nl_entry.dir == &mm->nodes) ? NULL : \
 	list_entry((root)->nl_entry.dir, struct nouveau_mm_node, nl_entry)
 
+static void
+nouveau_mm_dump(struct nouveau_mm *mm, const char *header)
+{
+	struct nouveau_mm_node *node;
+
+	printk(KERN_ERR "nouveau: %s\n", header);
+	printk(KERN_ERR "nouveau: node list:\n");
+	list_for_each_entry(node, &mm->nodes, nl_entry) {
+		printk(KERN_ERR "nouveau: \t%08x %08x %d\n",
+		       node->offset, node->length, node->type);
+	}
+	printk(KERN_ERR "nouveau: free list:\n");
+	list_for_each_entry(node, &mm->free, fl_entry) {
+		printk(KERN_ERR "nouveau: \t%08x %08x %d\n",
+		       node->offset, node->length, node->type);
+	}
+}
+
 void
 nouveau_mm_free(struct nouveau_mm *mm, struct nouveau_mm_node **pthis)
 {
@@ -37,29 +55,29 @@ nouveau_mm_free(struct nouveau_mm *mm, struct nouveau_mm_node **pthis)
 		struct nouveau_mm_node *prev = node(this, prev);
 		struct nouveau_mm_node *next = node(this, next);
 
-		if (prev && prev->type == 0) {
+		if (prev && prev->type == NVKM_MM_TYPE_NONE) {
 			prev->length += this->length;
 			list_del(&this->nl_entry);
 			kfree(this); this = prev;
 		}
 
-		if (next && next->type == 0) {
+		if (next && next->type == NVKM_MM_TYPE_NONE) {
 			next->offset  = this->offset;
 			next->length += this->length;
-			if (this->type == 0)
+			if (this->type == NVKM_MM_TYPE_NONE)
 				list_del(&this->fl_entry);
 			list_del(&this->nl_entry);
 			kfree(this); this = NULL;
 		}
 
-		if (this && this->type != 0) {
+		if (this && this->type != NVKM_MM_TYPE_NONE) {
 			list_for_each_entry(prev, &mm->free, fl_entry) {
 				if (this->offset < prev->offset)
 					break;
 			}
 
 			list_add_tail(&this->fl_entry, &prev->fl_entry);
-			this->type = 0;
+			this->type = NVKM_MM_TYPE_NONE;
 		}
 	}
 
@@ -80,27 +98,32 @@ region_head(struct nouveau_mm *mm, struct nouveau_mm_node *a, u32 size)
 
 	b->offset = a->offset;
 	b->length = size;
+	b->heap   = a->heap;
 	b->type   = a->type;
 	a->offset += size;
 	a->length -= size;
 	list_add_tail(&b->nl_entry, &a->nl_entry);
-	if (b->type == 0)
+	if (b->type == NVKM_MM_TYPE_NONE)
 		list_add_tail(&b->fl_entry, &a->fl_entry);
 	return b;
 }
 
 int
-nouveau_mm_head(struct nouveau_mm *mm, u8 type, u32 size_max, u32 size_min,
-		u32 align, struct nouveau_mm_node **pnode)
+nouveau_mm_head(struct nouveau_mm *mm, u8 heap, u8 type, u32 size_max,
+		u32 size_min, u32 align, struct nouveau_mm_node **pnode)
 {
 	struct nouveau_mm_node *prev, *this, *next;
 	u32 mask = align - 1;
 	u32 splitoff;
 	u32 s, e;
 
-	BUG_ON(!type);
+	BUG_ON(type == NVKM_MM_TYPE_NONE || type == NVKM_MM_TYPE_HOLE);
 
 	list_for_each_entry(this, &mm->free, fl_entry) {
+		if (unlikely(heap != NVKM_MM_HEAP_ANY)) {
+			if (this->heap != heap)
+				continue;
+		}
 		e = this->offset + this->length;
 		s = this->offset;
 
@@ -149,27 +172,32 @@ region_tail(struct nouveau_mm *mm, struct nouveau_mm_node *a, u32 size)
 	a->length -= size;
 	b->offset  = a->offset + a->length;
 	b->length  = size;
+	b->heap    = a->heap;
 	b->type    = a->type;
 
 	list_add(&b->nl_entry, &a->nl_entry);
-	if (b->type == 0)
+	if (b->type == NVKM_MM_TYPE_NONE)
 		list_add(&b->fl_entry, &a->fl_entry);
 	return b;
 }
 
 int
-nouveau_mm_tail(struct nouveau_mm *mm, u8 type, u32 size_max, u32 size_min,
-		u32 align, struct nouveau_mm_node **pnode)
+nouveau_mm_tail(struct nouveau_mm *mm, u8 heap, u8 type, u32 size_max,
+		u32 size_min, u32 align, struct nouveau_mm_node **pnode)
 {
 	struct nouveau_mm_node *prev, *this, *next;
 	u32 mask = align - 1;
 
-	BUG_ON(!type);
+	BUG_ON(type == NVKM_MM_TYPE_NONE || type == NVKM_MM_TYPE_HOLE);
 
 	list_for_each_entry_reverse(this, &mm->free, fl_entry) {
 		u32 e = this->offset + this->length;
 		u32 s = this->offset;
 		u32 c = 0, a;
+		if (unlikely(heap != NVKM_MM_HEAP_ANY)) {
+			if (this->heap != heap)
+				continue;
+		}
 
 		prev = node(this, prev);
 		if (prev && prev->type != type)
@@ -209,9 +237,23 @@ nouveau_mm_tail(struct nouveau_mm *mm, u8 type, u32 size_max, u32 size_min,
 int
 nouveau_mm_init(struct nouveau_mm *mm, u32 offset, u32 length, u32 block)
 {
-	struct nouveau_mm_node *node;
+	struct nouveau_mm_node *node, *prev;
+	u32 next;
 
-	if (block) {
+	if (nouveau_mm_initialised(mm)) {
+		prev = list_last_entry(&mm->nodes, typeof(*node), nl_entry);
+		next = prev->offset + prev->length;
+		if (next != offset) {
+			BUG_ON(next > offset);
+			if (!(node = kzalloc(sizeof(*node), GFP_KERNEL)))
+				return -ENOMEM;
+			node->type   = NVKM_MM_TYPE_HOLE;
+			node->offset = next;
+			node->length = offset - next;
+			list_add_tail(&node->nl_entry, &mm->nodes);
+		}
+		BUG_ON(block != mm->block_size);
+	} else {
 		INIT_LIST_HEAD(&mm->nodes);
 		INIT_LIST_HEAD(&mm->free);
 		mm->block_size = block;
@@ -230,25 +272,32 @@ nouveau_mm_init(struct nouveau_mm *mm, u32 offset, u32 length, u32 block)
 
 	list_add_tail(&node->nl_entry, &mm->nodes);
 	list_add_tail(&node->fl_entry, &mm->free);
-	mm->heap_nodes++;
+	node->heap = ++mm->heap_nodes;
 	return 0;
 }
 
 int
 nouveau_mm_fini(struct nouveau_mm *mm)
 {
-	if (nouveau_mm_initialised(mm)) {
-		struct nouveau_mm_node *node, *heap =
-			list_first_entry(&mm->nodes, typeof(*heap), nl_entry);
-		int nodes = 0;
+	struct nouveau_mm_node *node, *temp;
+	int nodes = 0;
 
-		list_for_each_entry(node, &mm->nodes, nl_entry) {
-			if (WARN_ON(nodes++ == mm->heap_nodes))
+	if (!nouveau_mm_initialised(mm))
+		return 0;
+
+	list_for_each_entry(node, &mm->nodes, nl_entry) {
+		if (node->type != NVKM_MM_TYPE_HOLE) {
+			if (++nodes > mm->heap_nodes) {
+				nouveau_mm_dump(mm, "mm not clean!");
 				return -EBUSY;
+			}
 		}
-
-		kfree(heap);
 	}
 
+	list_for_each_entry_safe(node, temp, &mm->nodes, nl_entry) {
+		list_del(&node->nl_entry);
+		kfree(node);
+	}
+	mm->heap_nodes = 0;
 	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/core/core/notify.c b/drivers/gpu/drm/nouveau/core/core/notify.c
index 76adb81bdea2..d1bcde55e9d7 100644
--- a/drivers/gpu/drm/nouveau/core/core/notify.c
+++ b/drivers/gpu/drm/nouveau/core/core/notify.c
@@ -134,14 +134,15 @@ nvkm_notify_fini(struct nvkm_notify *notify)
 }
 
 int
-nvkm_notify_init(struct nvkm_event *event, int (*func)(struct nvkm_notify *),
-		 bool work, void *data, u32 size, u32 reply,
+nvkm_notify_init(struct nouveau_object *object, struct nvkm_event *event,
+		 int (*func)(struct nvkm_notify *), bool work,
+		 void *data, u32 size, u32 reply,
 		 struct nvkm_notify *notify)
 {
 	unsigned long flags;
 	int ret = -ENODEV;
 	if ((notify->event = event), event->refs) {
-		ret = event->func->ctor(data, size, notify);
+		ret = event->func->ctor(object, data, size, notify);
 		if (ret == 0 && (ret = -EINVAL, notify->size == reply)) {
 			notify->flags = 0;
 			notify->block = 1;
diff --git a/drivers/gpu/drm/nouveau/core/engine/device/base.c b/drivers/gpu/drm/nouveau/core/engine/device/base.c
index 8928f7981d4a..0ef5a5713182 100644
--- a/drivers/gpu/drm/nouveau/core/engine/device/base.c
+++ b/drivers/gpu/drm/nouveau/core/engine/device/base.c
@@ -505,7 +505,8 @@ nouveau_device_sclass[] = {
 };
 
 static int
-nouveau_device_event_ctor(void *data, u32 size, struct nvkm_notify *notify)
+nouveau_device_event_ctor(struct nouveau_object *object, void *data, u32 size,
+			  struct nvkm_notify *notify)
 {
 	if (!WARN_ON(size != 0)) {
 		notify->size  = 0;
diff --git a/drivers/gpu/drm/nouveau/core/engine/device/gm100.c b/drivers/gpu/drm/nouveau/core/engine/device/gm100.c
index 377ec0b8851e..6295668e29a5 100644
--- a/drivers/gpu/drm/nouveau/core/engine/device/gm100.c
+++ b/drivers/gpu/drm/nouveau/core/engine/device/gm100.c
@@ -26,6 +26,7 @@
 #include <subdev/bus.h>
 #include <subdev/gpio.h>
 #include <subdev/i2c.h>
+#include <subdev/fuse.h>
 #include <subdev/clock.h>
 #include <subdev/therm.h>
 #include <subdev/mxm.h>
@@ -62,10 +63,9 @@ gm100_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
 		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nve0_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nvd0_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] = &gm107_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nve0_clock_oclass;
-#if 0
-		device->oclass[NVDEV_SUBDEV_THERM  ] = &nvd0_therm_oclass;
-#endif
+		device->oclass[NVDEV_SUBDEV_THERM  ] = &gm107_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
 		device->oclass[NVDEV_SUBDEV_DEVINIT] =  gm107_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  gk20a_mc_oclass;
@@ -77,8 +77,9 @@ gm100_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nvc0_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nvc0_bar_oclass;
-#if 0
 		device->oclass[NVDEV_SUBDEV_PWR    ] =  nv108_pwr_oclass;
+
+#if 0
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
 #endif
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] =  nvd0_dmaeng_oclass;
diff --git a/drivers/gpu/drm/nouveau/core/engine/device/nv50.c b/drivers/gpu/drm/nouveau/core/engine/device/nv50.c
index 932f84fae459..96f568d1321b 100644
--- a/drivers/gpu/drm/nouveau/core/engine/device/nv50.c
+++ b/drivers/gpu/drm/nouveau/core/engine/device/nv50.c
@@ -26,6 +26,7 @@
 #include <subdev/bus.h>
 #include <subdev/gpio.h>
 #include <subdev/i2c.h>
+#include <subdev/fuse.h>
 #include <subdev/clock.h>
 #include <subdev/therm.h>
 #include <subdev/mxm.h>
@@ -62,6 +63,7 @@ nv50_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
 		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv50_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv50_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] =  &g80_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nv50_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv50_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -87,6 +89,7 @@ nv50_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
 		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv50_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv50_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] =  &g80_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nv84_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv84_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -115,6 +118,7 @@ nv50_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
 		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv50_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv50_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] =  &g80_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nv84_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv84_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -141,8 +145,9 @@ nv50_identify(struct nouveau_device *device)
 	case 0x92:
 		device->cname = "G92";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv50_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv50_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] =  &g80_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nv84_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv84_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -169,8 +174,9 @@ nv50_identify(struct nouveau_device *device)
 	case 0x94:
 		device->cname = "G94";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv94_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] =  &g80_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nv84_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv84_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -197,8 +203,9 @@ nv50_identify(struct nouveau_device *device)
 	case 0x96:
 		device->cname = "G96";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv94_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] =  &g80_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nv84_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv84_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -225,8 +232,9 @@ nv50_identify(struct nouveau_device *device)
 	case 0x98:
 		device->cname = "G98";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv94_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] =  &g80_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nv84_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv84_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -253,8 +261,9 @@ nv50_identify(struct nouveau_device *device)
 	case 0xa0:
 		device->cname = "G200";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv94_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv50_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] =  &g80_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nv84_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv84_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -281,8 +290,9 @@ nv50_identify(struct nouveau_device *device)
 	case 0xaa:
 		device->cname = "MCP77/MCP78";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv94_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] =  &g80_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nvaa_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv84_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -309,8 +319,9 @@ nv50_identify(struct nouveau_device *device)
 	case 0xac:
 		device->cname = "MCP79/MCP7A";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv94_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] =  &g80_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nvaa_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv84_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -337,8 +348,9 @@ nv50_identify(struct nouveau_device *device)
 	case 0xa3:
 		device->cname = "GT215";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv94_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] =  &g80_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nva3_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -367,8 +379,9 @@ nv50_identify(struct nouveau_device *device)
 	case 0xa5:
 		device->cname = "GT216";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv94_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] =  &g80_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nva3_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -396,8 +409,9 @@ nv50_identify(struct nouveau_device *device)
 	case 0xa8:
 		device->cname = "GT218";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv94_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] =  &g80_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nva3_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -425,8 +439,9 @@ nv50_identify(struct nouveau_device *device)
 	case 0xaf:
 		device->cname = "MCP89";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv94_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] =  &g80_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nva3_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
diff --git a/drivers/gpu/drm/nouveau/core/engine/device/nvc0.c b/drivers/gpu/drm/nouveau/core/engine/device/nvc0.c
index b4a2917ce555..cd05677ad4b7 100644
--- a/drivers/gpu/drm/nouveau/core/engine/device/nvc0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/device/nvc0.c
@@ -26,6 +26,7 @@
 #include <subdev/bus.h>
 #include <subdev/gpio.h>
 #include <subdev/i2c.h>
+#include <subdev/fuse.h>
 #include <subdev/clock.h>
 #include <subdev/therm.h>
 #include <subdev/mxm.h>
@@ -60,8 +61,9 @@ nvc0_identify(struct nouveau_device *device)
 	case 0xc0:
 		device->cname = "GF100";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv94_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] = &gf100_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nvc0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -92,8 +94,9 @@ nvc0_identify(struct nouveau_device *device)
 	case 0xc4:
 		device->cname = "GF104";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv94_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] = &gf100_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nvc0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -124,8 +127,9 @@ nvc0_identify(struct nouveau_device *device)
 	case 0xc3:
 		device->cname = "GF106";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv94_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] = &gf100_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nvc0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -155,8 +159,9 @@ nvc0_identify(struct nouveau_device *device)
 	case 0xce:
 		device->cname = "GF114";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv94_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] = &gf100_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nvc0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -187,8 +192,9 @@ nvc0_identify(struct nouveau_device *device)
 	case 0xcf:
 		device->cname = "GF116";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv94_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] = &gf100_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nvc0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -219,8 +225,9 @@ nvc0_identify(struct nouveau_device *device)
 	case 0xc1:
 		device->cname = "GF108";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv94_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] = &gf100_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nvc0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -250,8 +257,9 @@ nvc0_identify(struct nouveau_device *device)
 	case 0xc8:
 		device->cname = "GF110";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv94_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] = &gf100_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nvc0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -284,6 +292,7 @@ nvc0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
 		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nvd0_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nvd0_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] = &gf100_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nvc0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nvd0_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -315,6 +324,7 @@ nvc0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
 		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nvd0_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  gf117_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] = &gf100_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nvc0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nvd0_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
diff --git a/drivers/gpu/drm/nouveau/core/engine/device/nve0.c b/drivers/gpu/drm/nouveau/core/engine/device/nve0.c
index cdf9147f32a1..b1b2e484ecfa 100644
--- a/drivers/gpu/drm/nouveau/core/engine/device/nve0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/device/nve0.c
@@ -26,6 +26,7 @@
 #include <subdev/bus.h>
 #include <subdev/gpio.h>
 #include <subdev/i2c.h>
+#include <subdev/fuse.h>
 #include <subdev/clock.h>
 #include <subdev/therm.h>
 #include <subdev/mxm.h>
@@ -62,6 +63,7 @@ nve0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
 		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nve0_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nve0_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] = &gf100_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nve0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nvd0_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -95,6 +97,7 @@ nve0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
 		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nve0_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nve0_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] = &gf100_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nve0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nvd0_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -128,6 +131,7 @@ nve0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
 		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nve0_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nve0_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] = &gf100_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nve0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nvd0_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -161,6 +165,7 @@ nve0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &gk20a_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  gk20a_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nvc0_bus_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] = &gf100_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &gk20a_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  gk20a_fb_oclass;
 		device->oclass[NVDEV_SUBDEV_LTC    ] =  gk104_ltc_oclass;
@@ -180,6 +185,7 @@ nve0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
 		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nve0_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nve0_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] = &gf100_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nve0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nvd0_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -213,6 +219,7 @@ nve0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
 		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nve0_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nvd0_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] = &gf100_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nve0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nvd0_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
@@ -246,6 +253,7 @@ nve0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
 		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nve0_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] =  nve0_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_FUSE   ] = &gf100_fuse_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nve0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nvd0_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/base.c b/drivers/gpu/drm/nouveau/core/engine/disp/base.c
index 22d55f6cde50..64b84667f3a5 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/base.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/base.c
@@ -32,7 +32,8 @@
 #include "conn.h"
 
 int
-nouveau_disp_vblank_ctor(void *data, u32 size, struct nvkm_notify *notify)
+nouveau_disp_vblank_ctor(struct nouveau_object *object, void *data, u32 size,
+			 struct nvkm_notify *notify)
 {
 	struct nouveau_disp *disp =
 		container_of(notify->event, typeof(*disp), vblank);
@@ -61,7 +62,8 @@ nouveau_disp_vblank(struct nouveau_disp *disp, int head)
 }
 
 static int
-nouveau_disp_hpd_ctor(void *data, u32 size, struct nvkm_notify *notify)
+nouveau_disp_hpd_ctor(struct nouveau_object *object, void *data, u32 size,
+		      struct nvkm_notify *notify)
 {
 	struct nouveau_disp *disp =
 		container_of(notify->event, typeof(*disp), hpd);
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/conn.c b/drivers/gpu/drm/nouveau/core/engine/disp/conn.c
index 3d1070228977..1496b567dd4a 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/conn.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/conn.c
@@ -126,8 +126,8 @@ nvkm_connector_create_(struct nouveau_object *parent,
 			return 0;
 		}
 
-		ret = nvkm_notify_init(&gpio->event, nvkm_connector_hpd, true,
-				       &(struct nvkm_gpio_ntfy_req) {
+		ret = nvkm_notify_init(NULL, &gpio->event, nvkm_connector_hpd,
+				       true, &(struct nvkm_gpio_ntfy_req) {
 					.mask = NVKM_GPIO_TOGGLED,
 					.line = func.line,
 				       },
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/gm107.c b/drivers/gpu/drm/nouveau/core/engine/disp/gm107.c
index d54da8b5f87e..b3df3fe2dc09 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/gm107.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/gm107.c
@@ -68,6 +68,10 @@ gm107_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	if (ret)
 		return ret;
 
+	ret = nvkm_event_init(&nvd0_disp_chan_uevent, 1, 17, &priv->uevent);
+	if (ret)
+		return ret;
+
 	nv_engine(priv)->sclass = gm107_disp_base_oclass;
 	nv_engine(priv)->cclass = &nv50_disp_cclass;
 	nv_subdev(priv)->intr = nvd0_disp_intr;
@@ -80,7 +84,7 @@ gm107_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	priv->dac.sense = nv50_dac_sense;
 	priv->sor.power = nv50_sor_power;
 	priv->sor.hda_eld = nvd0_hda_eld;
-	priv->sor.hdmi = nvd0_hdmi_ctrl;
+	priv->sor.hdmi = nve0_hdmi_ctrl;
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/hdanva3.c b/drivers/gpu/drm/nouveau/core/engine/disp/hdanva3.c
index 8b4e06abe533..fe9ef5894dd4 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/hdanva3.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/hdanva3.c
@@ -26,6 +26,8 @@
 #include <nvif/unpack.h>
 #include <nvif/class.h>
 
+#include <subdev/timer.h>
+
 #include "nv50.h"
 
 int
@@ -46,16 +48,21 @@ nva3_hda_eld(NV50_DISP_MTHD_V1)
 		return ret;
 
 	if (size && args->v0.data[0]) {
+		if (outp->info.type == DCB_OUTPUT_DP) {
+			nv_mask(priv, 0x61c1e0 + soff, 0x8000000d, 0x80000001);
+			nv_wait(priv, 0x61c1e0 + soff, 0x80000000, 0x00000000);
+		}
 		for (i = 0; i < size; i++)
 			nv_wr32(priv, 0x61c440 + soff, (i << 8) | args->v0.data[0]);
 		for (; i < 0x60; i++)
 			nv_wr32(priv, 0x61c440 + soff, (i << 8));
 		nv_mask(priv, 0x61c448 + soff, 0x80000003, 0x80000003);
-	} else
-	if (size) {
-		nv_mask(priv, 0x61c448 + soff, 0x80000003, 0x80000001);
 	} else {
-		nv_mask(priv, 0x61c448 + soff, 0x80000003, 0x80000000);
+		if (outp->info.type == DCB_OUTPUT_DP) {
+			nv_mask(priv, 0x61c1e0 + soff, 0x80000001, 0x80000000);
+			nv_wait(priv, 0x61c1e0 + soff, 0x80000000, 0x00000000);
+		}
+		nv_mask(priv, 0x61c448 + soff, 0x80000003, 0x80000000 | !!size);
 	}
 
 	return 0;
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/hdanvd0.c b/drivers/gpu/drm/nouveau/core/engine/disp/hdanvd0.c
index baf558fc12fb..1d4e8432d857 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/hdanvd0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/hdanvd0.c
@@ -26,10 +26,7 @@
 #include <nvif/unpack.h>
 #include <nvif/class.h>
 
-#include <subdev/bios.h>
-#include <subdev/bios/dcb.h>
-#include <subdev/bios/dp.h>
-#include <subdev/bios/init.h>
+#include <subdev/timer.h>
 
 #include "nv50.h"
 
@@ -40,6 +37,7 @@ nvd0_hda_eld(NV50_DISP_MTHD_V1)
 		struct nv50_disp_sor_hda_eld_v0 v0;
 	} *args = data;
 	const u32 soff = outp->or * 0x030;
+	const u32 hoff = head * 0x800;
 	int ret, i;
 
 	nv_ioctl(object, "disp sor hda eld size %d\n", size);
@@ -51,16 +49,22 @@ nvd0_hda_eld(NV50_DISP_MTHD_V1)
 		return ret;
 
 	if (size && args->v0.data[0]) {
+		if (outp->info.type == DCB_OUTPUT_DP) {
+			nv_mask(priv, 0x616618 + hoff, 0x8000000c, 0x80000001);
+			nv_wait(priv, 0x616618 + hoff, 0x80000000, 0x00000000);
+		}
+		nv_mask(priv, 0x616548 + hoff, 0x00000070, 0x00000000);
 		for (i = 0; i < size; i++)
 			nv_wr32(priv, 0x10ec00 + soff, (i << 8) | args->v0.data[i]);
 		for (; i < 0x60; i++)
 			nv_wr32(priv, 0x10ec00 + soff, (i << 8));
 		nv_mask(priv, 0x10ec10 + soff, 0x80000003, 0x80000003);
-	} else
-	if (size) {
-		nv_mask(priv, 0x10ec10 + soff, 0x80000003, 0x80000001);
 	} else {
-		nv_mask(priv, 0x10ec10 + soff, 0x80000003, 0x80000000);
+		if (outp->info.type == DCB_OUTPUT_DP) {
+			nv_mask(priv, 0x616618 + hoff, 0x80000001, 0x80000000);
+			nv_wait(priv, 0x616618 + hoff, 0x80000000, 0x00000000);
+		}
+		nv_mask(priv, 0x10ec10 + soff, 0x80000003, 0x80000000 | !!size);
 	}
 
 	return 0;
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/hdminvd0.c b/drivers/gpu/drm/nouveau/core/engine/disp/hdminvd0.c
index 3106d295b48d..bac4fc4570f0 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/hdminvd0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/hdminvd0.c
@@ -75,8 +75,5 @@ nvd0_hdmi_ctrl(NV50_DISP_MTHD_V1)
 
 	/* HDMI_CTRL */
 	nv_mask(priv, 0x616798 + hoff, 0x401f007f, ctrl);
-
-	/* NFI, audio doesn't work without it though.. */
-	nv_mask(priv, 0x616548 + hoff, 0x00000070, 0x00000000);
 	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/hdminve0.c b/drivers/gpu/drm/nouveau/core/engine/disp/hdminve0.c
new file mode 100644
index 000000000000..528d14ec2f7f
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/hdminve0.c
@@ -0,0 +1,83 @@
+/*
+ * Copyright 2014 Red Hat 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: Ben Skeggs
+ */
+
+#include <core/client.h>
+#include <nvif/unpack.h>
+#include <nvif/class.h>
+
+#include "nv50.h"
+
+int
+nve0_hdmi_ctrl(NV50_DISP_MTHD_V1)
+{
+	const u32 hoff = (head * 0x800);
+	const u32 hdmi = (head * 0x400);
+	union {
+		struct nv50_disp_sor_hdmi_pwr_v0 v0;
+	} *args = data;
+	u32 ctrl;
+	int ret;
+
+	nv_ioctl(object, "disp sor hdmi ctrl size %d\n", size);
+	if (nvif_unpack(args->v0, 0, 0, false)) {
+		nv_ioctl(object, "disp sor hdmi ctrl vers %d state %d "
+				 "max_ac_packet %d rekey %d\n",
+			 args->v0.version, args->v0.state,
+			 args->v0.max_ac_packet, args->v0.rekey);
+		if (args->v0.max_ac_packet > 0x1f || args->v0.rekey > 0x7f)
+			return -EINVAL;
+		ctrl  = 0x40000000 * !!args->v0.state;
+		ctrl |= args->v0.max_ac_packet << 16;
+		ctrl |= args->v0.rekey;
+	} else
+		return ret;
+
+	if (!(ctrl & 0x40000000)) {
+		nv_mask(priv, 0x616798 + hoff, 0x40000000, 0x00000000);
+		nv_mask(priv, 0x6900c0 + hdmi, 0x00000001, 0x00000000);
+		nv_mask(priv, 0x690000 + hdmi, 0x00000001, 0x00000000);
+		return 0;
+	}
+
+	/* AVI InfoFrame */
+	nv_mask(priv, 0x690000 + hdmi, 0x00000001, 0x00000000);
+	nv_wr32(priv, 0x690008 + hdmi, 0x000d0282);
+	nv_wr32(priv, 0x69000c + hdmi, 0x0000006f);
+	nv_wr32(priv, 0x690010 + hdmi, 0x00000000);
+	nv_wr32(priv, 0x690014 + hdmi, 0x00000000);
+	nv_wr32(priv, 0x690018 + hdmi, 0x00000000);
+	nv_mask(priv, 0x690000 + hdmi, 0x00000001, 0x00000001);
+
+	/* ??? InfoFrame? */
+	nv_mask(priv, 0x6900c0 + hdmi, 0x00000001, 0x00000000);
+	nv_wr32(priv, 0x6900cc + hdmi, 0x00000010);
+	nv_mask(priv, 0x6900c0 + hdmi, 0x00000001, 0x00000001);
+
+	/* ??? */
+	nv_wr32(priv, 0x690080 + hdmi, 0x82000000);
+
+	/* HDMI_CTRL */
+	nv_mask(priv, 0x616798 + hoff, 0x401f007f, ctrl);
+	return 0;
+}
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nv50.c b/drivers/gpu/drm/nouveau/core/engine/disp/nv50.c
index f8cbb512132f..2df3a937037d 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nv50.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nv50.c
@@ -29,6 +29,7 @@
 #include <core/enum.h>
 #include <nvif/unpack.h>
 #include <nvif/class.h>
+#include <nvif/event.h>
 
 #include <subdev/bios.h>
 #include <subdev/bios/dcb.h>
@@ -82,6 +83,71 @@ nv50_disp_chan_destroy(struct nv50_disp_chan *chan)
 	nouveau_namedb_destroy(&chan->base);
 }
 
+static void
+nv50_disp_chan_uevent_fini(struct nvkm_event *event, int type, int index)
+{
+	struct nv50_disp_priv *priv = container_of(event, typeof(*priv), uevent);
+	nv_mask(priv, 0x610028, 0x00000001 << index, 0x00000000 << index);
+}
+
+static void
+nv50_disp_chan_uevent_init(struct nvkm_event *event, int types, int index)
+{
+	struct nv50_disp_priv *priv = container_of(event, typeof(*priv), uevent);
+	nv_mask(priv, 0x610028, 0x00000001 << index, 0x00000001 << index);
+}
+
+void
+nv50_disp_chan_uevent_send(struct nv50_disp_priv *priv, int chid)
+{
+	struct nvif_notify_uevent_rep {
+	} rep;
+
+	nvkm_event_send(&priv->uevent, 1, chid, &rep, sizeof(rep));
+}
+
+int
+nv50_disp_chan_uevent_ctor(struct nouveau_object *object, void *data, u32 size,
+			   struct nvkm_notify *notify)
+{
+	struct nv50_disp_dmac *dmac = (void *)object;
+	union {
+		struct nvif_notify_uevent_req none;
+	} *args = data;
+	int ret;
+
+	if (nvif_unvers(args->none)) {
+		notify->size  = sizeof(struct nvif_notify_uevent_rep);
+		notify->types = 1;
+		notify->index = dmac->base.chid;
+		return 0;
+	}
+
+	return ret;
+}
+
+const struct nvkm_event_func
+nv50_disp_chan_uevent = {
+	.ctor = nv50_disp_chan_uevent_ctor,
+	.init = nv50_disp_chan_uevent_init,
+	.fini = nv50_disp_chan_uevent_fini,
+};
+
+int
+nv50_disp_chan_ntfy(struct nouveau_object *object, u32 type,
+		    struct nvkm_event **pevent)
+{
+	struct nv50_disp_priv *priv = (void *)object->engine;
+	switch (type) {
+	case NV50_DISP_CORE_CHANNEL_DMA_V0_NTFY_UEVENT:
+		*pevent = &priv->uevent;
+		return 0;
+	default:
+		break;
+	}
+	return -EINVAL;
+}
+
 int
 nv50_disp_chan_map(struct nouveau_object *object, u64 *addr, u32 *size)
 {
@@ -195,7 +261,7 @@ nv50_disp_dmac_init(struct nouveau_object *object)
 		return ret;
 
 	/* enable error reporting */
-	nv_mask(priv, 0x610028, 0x00010001 << chid, 0x00010001 << chid);
+	nv_mask(priv, 0x610028, 0x00010000 << chid, 0x00010000 << chid);
 
 	/* initialise channel for dma command submission */
 	nv_wr32(priv, 0x610204 + (chid * 0x0010), dmac->push);
@@ -232,7 +298,7 @@ nv50_disp_dmac_fini(struct nouveau_object *object, bool suspend)
 			return -EBUSY;
 	}
 
-	/* disable error reporting */
+	/* disable error reporting and completion notifications */
 	nv_mask(priv, 0x610028, 0x00010001 << chid, 0x00000000 << chid);
 
 	return nv50_disp_chan_fini(&dmac->base, suspend);
@@ -454,7 +520,7 @@ nv50_disp_mast_init(struct nouveau_object *object)
 		return ret;
 
 	/* enable error reporting */
-	nv_mask(priv, 0x610028, 0x00010001, 0x00010001);
+	nv_mask(priv, 0x610028, 0x00010000, 0x00010000);
 
 	/* attempt to unstick channel from some unknown state */
 	if ((nv_rd32(priv, 0x610200) & 0x009f0000) == 0x00020000)
@@ -494,7 +560,7 @@ nv50_disp_mast_fini(struct nouveau_object *object, bool suspend)
 			return -EBUSY;
 	}
 
-	/* disable error reporting */
+	/* disable error reporting and completion notifications */
 	nv_mask(priv, 0x610028, 0x00010001, 0x00000000);
 
 	return nv50_disp_chan_fini(&mast->base, suspend);
@@ -507,6 +573,7 @@ nv50_disp_mast_ofuncs = {
 	.base.init = nv50_disp_mast_init,
 	.base.fini = nv50_disp_mast_fini,
 	.base.map  = nv50_disp_chan_map,
+	.base.ntfy = nv50_disp_chan_ntfy,
 	.base.rd32 = nv50_disp_chan_rd32,
 	.base.wr32 = nv50_disp_chan_wr32,
 	.chid = 0,
@@ -607,6 +674,7 @@ nv50_disp_sync_ofuncs = {
 	.base.dtor = nv50_disp_dmac_dtor,
 	.base.init = nv50_disp_dmac_init,
 	.base.fini = nv50_disp_dmac_fini,
+	.base.ntfy = nv50_disp_chan_ntfy,
 	.base.map  = nv50_disp_chan_map,
 	.base.rd32 = nv50_disp_chan_rd32,
 	.base.wr32 = nv50_disp_chan_wr32,
@@ -696,6 +764,7 @@ nv50_disp_ovly_ofuncs = {
 	.base.dtor = nv50_disp_dmac_dtor,
 	.base.init = nv50_disp_dmac_init,
 	.base.fini = nv50_disp_dmac_fini,
+	.base.ntfy = nv50_disp_chan_ntfy,
 	.base.map  = nv50_disp_chan_map,
 	.base.rd32 = nv50_disp_chan_rd32,
 	.base.wr32 = nv50_disp_chan_wr32,
@@ -813,6 +882,7 @@ nv50_disp_oimm_ofuncs = {
 	.base.dtor = nv50_disp_pioc_dtor,
 	.base.init = nv50_disp_pioc_init,
 	.base.fini = nv50_disp_pioc_fini,
+	.base.ntfy = nv50_disp_chan_ntfy,
 	.base.map  = nv50_disp_chan_map,
 	.base.rd32 = nv50_disp_chan_rd32,
 	.base.wr32 = nv50_disp_chan_wr32,
@@ -860,6 +930,7 @@ nv50_disp_curs_ofuncs = {
 	.base.dtor = nv50_disp_pioc_dtor,
 	.base.init = nv50_disp_pioc_init,
 	.base.fini = nv50_disp_pioc_fini,
+	.base.ntfy = nv50_disp_chan_ntfy,
 	.base.map  = nv50_disp_chan_map,
 	.base.rd32 = nv50_disp_chan_rd32,
 	.base.wr32 = nv50_disp_chan_wr32,
@@ -1559,7 +1630,7 @@ nv50_disp_intr_unk20_1(struct nv50_disp_priv *priv, int head)
 }
 
 static void
-nv50_disp_intr_unk20_2_dp(struct nv50_disp_priv *priv,
+nv50_disp_intr_unk20_2_dp(struct nv50_disp_priv *priv, int head,
 			  struct dcb_output *outp, u32 pclk)
 {
 	const int link = !(outp->sorconf.link & 1);
@@ -1568,24 +1639,36 @@ nv50_disp_intr_unk20_2_dp(struct nv50_disp_priv *priv,
 	const u32 loff = (link * 0x080) + soff;
 	const u32 ctrl = nv_rd32(priv, 0x610794 + (or * 8));
 	const u32 symbol = 100000;
-	u32 dpctrl = nv_rd32(priv, 0x61c10c + loff) & 0x0000f0000;
+	const s32 vactive = nv_rd32(priv, 0x610af8 + (head * 0x540)) & 0xffff;
+	const s32 vblanke = nv_rd32(priv, 0x610ae8 + (head * 0x540)) & 0xffff;
+	const s32 vblanks = nv_rd32(priv, 0x610af0 + (head * 0x540)) & 0xffff;
+	u32 dpctrl = nv_rd32(priv, 0x61c10c + loff);
 	u32 clksor = nv_rd32(priv, 0x614300 + soff);
 	int bestTU = 0, bestVTUi = 0, bestVTUf = 0, bestVTUa = 0;
 	int TU, VTUi, VTUf, VTUa;
 	u64 link_data_rate, link_ratio, unk;
 	u32 best_diff = 64 * symbol;
 	u32 link_nr, link_bw, bits;
-
-	/* calculate packed data rate for each lane */
-	if      (dpctrl > 0x00030000) link_nr = 4;
-	else if (dpctrl > 0x00010000) link_nr = 2;
-	else			      link_nr = 1;
-
-	if (clksor & 0x000c0000)
-		link_bw = 270000;
-	else
-		link_bw = 162000;
-
+	u64 value;
+
+	link_bw = (clksor & 0x000c0000) ? 270000 : 162000;
+	link_nr = hweight32(dpctrl & 0x000f0000);
+
+	/* symbols/hblank - algorithm taken from comments in tegra driver */
+	value = vblanke + vactive - vblanks - 7;
+	value = value * link_bw;
+	do_div(value, pclk);
+	value = value - (3 * !!(dpctrl & 0x00004000)) - (12 / link_nr);
+	nv_mask(priv, 0x61c1e8 + soff, 0x0000ffff, value);
+
+	/* symbols/vblank - algorithm taken from comments in tegra driver */
+	value = vblanks - vblanke - 25;
+	value = value * link_bw;
+	do_div(value, pclk);
+	value = value - ((36 / link_nr) + 3) - 1;
+	nv_mask(priv, 0x61c1ec + soff, 0x00ffffff, value);
+
+	/* watermark / activesym */
 	if      ((ctrl & 0xf0000) == 0x60000) bits = 30;
 	else if ((ctrl & 0xf0000) == 0x50000) bits = 24;
 	else                                  bits = 18;
@@ -1731,7 +1814,7 @@ nv50_disp_intr_unk20_2(struct nv50_disp_priv *priv, int head)
 	} else
 	if (!outp->info.location) {
 		if (outp->info.type == DCB_OUTPUT_DP)
-			nv50_disp_intr_unk20_2_dp(priv, &outp->info, pclk);
+			nv50_disp_intr_unk20_2_dp(priv, head, &outp->info, pclk);
 		oreg = 0x614300 + (ffs(outp->info.or) - 1) * 0x800;
 		oval = (conf & 0x0100) ? 0x00000101 : 0x00000000;
 		hval = 0x00000000;
@@ -1847,6 +1930,12 @@ nv50_disp_intr(struct nouveau_subdev *subdev)
 		intr0 &= ~(0x00010000 << chid);
 	}
 
+	while (intr0 & 0x0000001f) {
+		u32 chid = __ffs(intr0 & 0x0000001f);
+		nv50_disp_chan_uevent_send(priv, chid);
+		intr0 &= ~(0x00000001 << chid);
+	}
+
 	if (intr1 & 0x00000004) {
 		nouveau_disp_vblank(&priv->base, 0);
 		nv_wr32(priv, 0x610024, 0x00000004);
@@ -1881,6 +1970,10 @@ nv50_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	if (ret)
 		return ret;
 
+	ret = nvkm_event_init(&nv50_disp_chan_uevent, 1, 9, &priv->uevent);
+	if (ret)
+		return ret;
+
 	nv_engine(priv)->sclass = nv50_disp_base_oclass;
 	nv_engine(priv)->cclass = &nv50_disp_cclass;
 	nv_subdev(priv)->intr = nv50_disp_intr;
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nv50.h b/drivers/gpu/drm/nouveau/core/engine/disp/nv50.h
index 8ab14461f70c..5279feefec06 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nv50.h
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nv50.h
@@ -26,6 +26,8 @@ struct nv50_disp_priv {
 	struct work_struct supervisor;
 	u32 super;
 
+	struct nvkm_event uevent;
+
 	struct {
 		int nr;
 	} head;
@@ -75,6 +77,7 @@ int nvd0_hda_eld(NV50_DISP_MTHD_V1);
 int nv84_hdmi_ctrl(NV50_DISP_MTHD_V1);
 int nva3_hdmi_ctrl(NV50_DISP_MTHD_V1);
 int nvd0_hdmi_ctrl(NV50_DISP_MTHD_V1);
+int nve0_hdmi_ctrl(NV50_DISP_MTHD_V1);
 
 int nv50_sor_power(NV50_DISP_MTHD_V1);
 
@@ -116,9 +119,16 @@ struct nv50_disp_chan {
 	int chid;
 };
 
+int  nv50_disp_chan_ntfy(struct nouveau_object *, u32, struct nvkm_event **);
 int  nv50_disp_chan_map(struct nouveau_object *, u64 *, u32 *);
 u32  nv50_disp_chan_rd32(struct nouveau_object *, u64);
 void nv50_disp_chan_wr32(struct nouveau_object *, u64, u32);
+extern const struct nvkm_event_func nv50_disp_chan_uevent;
+int  nv50_disp_chan_uevent_ctor(struct nouveau_object *, void *, u32,
+				struct nvkm_notify *);
+void nv50_disp_chan_uevent_send(struct nv50_disp_priv *, int);
+
+extern const struct nvkm_event_func nvd0_disp_chan_uevent;
 
 #define nv50_disp_chan_init(a)                                                 \
 	nouveau_namedb_init(&(a)->base)
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nv84.c b/drivers/gpu/drm/nouveau/core/engine/disp/nv84.c
index 788ced1b6182..d36284715b2a 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nv84.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nv84.c
@@ -236,6 +236,10 @@ nv84_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	if (ret)
 		return ret;
 
+	ret = nvkm_event_init(&nv50_disp_chan_uevent, 1, 9, &priv->uevent);
+	if (ret)
+		return ret;
+
 	nv_engine(priv)->sclass = nv84_disp_base_oclass;
 	nv_engine(priv)->cclass = &nv50_disp_cclass;
 	nv_subdev(priv)->intr = nv50_disp_intr;
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nv94.c b/drivers/gpu/drm/nouveau/core/engine/disp/nv94.c
index fa79de906eae..a117064002b1 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nv94.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nv94.c
@@ -95,6 +95,10 @@ nv94_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	if (ret)
 		return ret;
 
+	ret = nvkm_event_init(&nv50_disp_chan_uevent, 1, 9, &priv->uevent);
+	if (ret)
+		return ret;
+
 	nv_engine(priv)->sclass = nv94_disp_base_oclass;
 	nv_engine(priv)->cclass = &nv50_disp_cclass;
 	nv_subdev(priv)->intr = nv50_disp_intr;
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nva0.c b/drivers/gpu/drm/nouveau/core/engine/disp/nva0.c
index 7af15f5d48dc..c67e68aadd45 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nva0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nva0.c
@@ -112,6 +112,10 @@ nva0_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	if (ret)
 		return ret;
 
+	ret = nvkm_event_init(&nv50_disp_chan_uevent, 1, 9, &priv->uevent);
+	if (ret)
+		return ret;
+
 	nv_engine(priv)->sclass = nva0_disp_base_oclass;
 	nv_engine(priv)->cclass = &nv50_disp_cclass;
 	nv_subdev(priv)->intr = nv50_disp_intr;
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nva3.c b/drivers/gpu/drm/nouveau/core/engine/disp/nva3.c
index 6bd39448f8da..22969f355aae 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nva3.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nva3.c
@@ -67,6 +67,10 @@ nva3_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	if (ret)
 		return ret;
 
+	ret = nvkm_event_init(&nv50_disp_chan_uevent, 1, 9, &priv->uevent);
+	if (ret)
+		return ret;
+
 	nv_engine(priv)->sclass = nva3_disp_base_oclass;
 	nv_engine(priv)->cclass = &nv50_disp_cclass;
 	nv_subdev(priv)->intr = nv50_disp_intr;
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nvd0.c b/drivers/gpu/drm/nouveau/core/engine/disp/nvd0.c
index a4bb3c774ee1..747e64bb9c06 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nvd0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nvd0.c
@@ -43,6 +43,31 @@
 #include "nv50.h"
 
 /*******************************************************************************
+ * EVO channel base class
+ ******************************************************************************/
+
+static void
+nvd0_disp_chan_uevent_fini(struct nvkm_event *event, int type, int index)
+{
+	struct nv50_disp_priv *priv = container_of(event, typeof(*priv), uevent);
+	nv_mask(priv, 0x610090, 0x00000001 << index, 0x00000000 << index);
+}
+
+static void
+nvd0_disp_chan_uevent_init(struct nvkm_event *event, int types, int index)
+{
+	struct nv50_disp_priv *priv = container_of(event, typeof(*priv), uevent);
+	nv_mask(priv, 0x610090, 0x00000001 << index, 0x00000001 << index);
+}
+
+const struct nvkm_event_func
+nvd0_disp_chan_uevent = {
+	.ctor = nv50_disp_chan_uevent_ctor,
+	.init = nvd0_disp_chan_uevent_init,
+	.fini = nvd0_disp_chan_uevent_fini,
+};
+
+/*******************************************************************************
  * EVO DMA channel base class
  ******************************************************************************/
 
@@ -77,7 +102,6 @@ nvd0_disp_dmac_init(struct nouveau_object *object)
 		return ret;
 
 	/* enable error reporting */
-	nv_mask(priv, 0x610090, 0x00000001 << chid, 0x00000001 << chid);
 	nv_mask(priv, 0x6100a0, 0x00000001 << chid, 0x00000001 << chid);
 
 	/* initialise channel for dma command submission */
@@ -115,7 +139,7 @@ nvd0_disp_dmac_fini(struct nouveau_object *object, bool suspend)
 			return -EBUSY;
 	}
 
-	/* disable error reporting */
+	/* disable error reporting and completion notification */
 	nv_mask(priv, 0x610090, 0x00000001 << chid, 0x00000000);
 	nv_mask(priv, 0x6100a0, 0x00000001 << chid, 0x00000000);
 
@@ -278,7 +302,6 @@ nvd0_disp_mast_init(struct nouveau_object *object)
 		return ret;
 
 	/* enable error reporting */
-	nv_mask(priv, 0x610090, 0x00000001, 0x00000001);
 	nv_mask(priv, 0x6100a0, 0x00000001, 0x00000001);
 
 	/* initialise channel for dma command submission */
@@ -313,7 +336,7 @@ nvd0_disp_mast_fini(struct nouveau_object *object, bool suspend)
 			return -EBUSY;
 	}
 
-	/* disable error reporting */
+	/* disable error reporting and completion notification */
 	nv_mask(priv, 0x610090, 0x00000001, 0x00000000);
 	nv_mask(priv, 0x6100a0, 0x00000001, 0x00000000);
 
@@ -326,6 +349,7 @@ nvd0_disp_mast_ofuncs = {
 	.base.dtor = nv50_disp_dmac_dtor,
 	.base.init = nvd0_disp_mast_init,
 	.base.fini = nvd0_disp_mast_fini,
+	.base.ntfy = nv50_disp_chan_ntfy,
 	.base.map  = nv50_disp_chan_map,
 	.base.rd32 = nv50_disp_chan_rd32,
 	.base.wr32 = nv50_disp_chan_wr32,
@@ -419,6 +443,7 @@ nvd0_disp_sync_ofuncs = {
 	.base.dtor = nv50_disp_dmac_dtor,
 	.base.init = nvd0_disp_dmac_init,
 	.base.fini = nvd0_disp_dmac_fini,
+	.base.ntfy = nv50_disp_chan_ntfy,
 	.base.map  = nv50_disp_chan_map,
 	.base.rd32 = nv50_disp_chan_rd32,
 	.base.wr32 = nv50_disp_chan_wr32,
@@ -499,6 +524,7 @@ nvd0_disp_ovly_ofuncs = {
 	.base.dtor = nv50_disp_dmac_dtor,
 	.base.init = nvd0_disp_dmac_init,
 	.base.fini = nvd0_disp_dmac_fini,
+	.base.ntfy = nv50_disp_chan_ntfy,
 	.base.map  = nv50_disp_chan_map,
 	.base.rd32 = nv50_disp_chan_rd32,
 	.base.wr32 = nv50_disp_chan_wr32,
@@ -524,7 +550,6 @@ nvd0_disp_pioc_init(struct nouveau_object *object)
 		return ret;
 
 	/* enable error reporting */
-	nv_mask(priv, 0x610090, 0x00000001 << chid, 0x00000001 << chid);
 	nv_mask(priv, 0x6100a0, 0x00000001 << chid, 0x00000001 << chid);
 
 	/* activate channel */
@@ -553,7 +578,7 @@ nvd0_disp_pioc_fini(struct nouveau_object *object, bool suspend)
 			return -EBUSY;
 	}
 
-	/* disable error reporting */
+	/* disable error reporting and completion notification */
 	nv_mask(priv, 0x610090, 0x00000001 << chid, 0x00000000);
 	nv_mask(priv, 0x6100a0, 0x00000001 << chid, 0x00000000);
 
@@ -570,6 +595,7 @@ nvd0_disp_oimm_ofuncs = {
 	.base.dtor = nv50_disp_pioc_dtor,
 	.base.init = nvd0_disp_pioc_init,
 	.base.fini = nvd0_disp_pioc_fini,
+	.base.ntfy = nv50_disp_chan_ntfy,
 	.base.map  = nv50_disp_chan_map,
 	.base.rd32 = nv50_disp_chan_rd32,
 	.base.wr32 = nv50_disp_chan_wr32,
@@ -586,6 +612,7 @@ nvd0_disp_curs_ofuncs = {
 	.base.dtor = nv50_disp_pioc_dtor,
 	.base.init = nvd0_disp_pioc_init,
 	.base.fini = nvd0_disp_pioc_fini,
+	.base.ntfy = nv50_disp_chan_ntfy,
 	.base.map  = nv50_disp_chan_map,
 	.base.rd32 = nv50_disp_chan_rd32,
 	.base.wr32 = nv50_disp_chan_wr32,
@@ -949,6 +976,9 @@ nvd0_disp_intr_unk2_2_tu(struct nv50_disp_priv *priv, int head,
 	const int or = ffs(outp->or) - 1;
 	const u32 ctrl = nv_rd32(priv, 0x660200 + (or   * 0x020));
 	const u32 conf = nv_rd32(priv, 0x660404 + (head * 0x300));
+	const s32 vactive = nv_rd32(priv, 0x660414 + (head * 0x300)) & 0xffff;
+	const s32 vblanke = nv_rd32(priv, 0x66041c + (head * 0x300)) & 0xffff;
+	const s32 vblanks = nv_rd32(priv, 0x660420 + (head * 0x300)) & 0xffff;
 	const u32 pclk = nv_rd32(priv, 0x660450 + (head * 0x300)) / 1000;
 	const u32 link = ((ctrl & 0xf00) == 0x800) ? 0 : 1;
 	const u32 hoff = (head * 0x800);
@@ -956,23 +986,35 @@ nvd0_disp_intr_unk2_2_tu(struct nv50_disp_priv *priv, int head,
 	const u32 loff = (link * 0x080) + soff;
 	const u32 symbol = 100000;
 	const u32 TU = 64;
-	u32 dpctrl = nv_rd32(priv, 0x61c10c + loff) & 0x000f0000;
+	u32 dpctrl = nv_rd32(priv, 0x61c10c + loff);
 	u32 clksor = nv_rd32(priv, 0x612300 + soff);
 	u32 datarate, link_nr, link_bw, bits;
 	u64 ratio, value;
 
+	link_nr  = hweight32(dpctrl & 0x000f0000);
+	link_bw  = (clksor & 0x007c0000) >> 18;
+	link_bw *= 27000;
+
+	/* symbols/hblank - algorithm taken from comments in tegra driver */
+	value = vblanke + vactive - vblanks - 7;
+	value = value * link_bw;
+	do_div(value, pclk);
+	value = value - (3 * !!(dpctrl & 0x00004000)) - (12 / link_nr);
+	nv_mask(priv, 0x616620 + hoff, 0x0000ffff, value);
+
+	/* symbols/vblank - algorithm taken from comments in tegra driver */
+	value = vblanks - vblanke - 25;
+	value = value * link_bw;
+	do_div(value, pclk);
+	value = value - ((36 / link_nr) + 3) - 1;
+	nv_mask(priv, 0x616624 + hoff, 0x00ffffff, value);
+
+	/* watermark */
 	if      ((conf & 0x3c0) == 0x180) bits = 30;
 	else if ((conf & 0x3c0) == 0x140) bits = 24;
 	else                              bits = 18;
 	datarate = (pclk * bits) / 8;
 
-	if      (dpctrl > 0x00030000) link_nr = 4;
-	else if (dpctrl > 0x00010000) link_nr = 2;
-	else			      link_nr = 1;
-
-	link_bw  = (clksor & 0x007c0000) >> 18;
-	link_bw *= 27000;
-
 	ratio  = datarate;
 	ratio *= symbol;
 	do_div(ratio, link_nr * link_bw);
@@ -1153,7 +1195,11 @@ nvd0_disp_intr(struct nouveau_subdev *subdev)
 
 	if (intr & 0x00000001) {
 		u32 stat = nv_rd32(priv, 0x61008c);
-		nv_wr32(priv, 0x61008c, stat);
+		while (stat) {
+			int chid = __ffs(stat); stat &= ~(1 << chid);
+			nv50_disp_chan_uevent_send(priv, chid);
+			nv_wr32(priv, 0x61008c, 1 << chid);
+		}
 		intr &= ~0x00000001;
 	}
 
@@ -1209,6 +1255,10 @@ nvd0_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	if (ret)
 		return ret;
 
+	ret = nvkm_event_init(&nvd0_disp_chan_uevent, 1, 17, &priv->uevent);
+	if (ret)
+		return ret;
+
 	nv_engine(priv)->sclass = nvd0_disp_base_oclass;
 	nv_engine(priv)->cclass = &nv50_disp_cclass;
 	nv_subdev(priv)->intr = nvd0_disp_intr;
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nve0.c b/drivers/gpu/drm/nouveau/core/engine/disp/nve0.c
index 47fef1e398c4..db144b2cf06b 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nve0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nve0.c
@@ -233,6 +233,10 @@ nve0_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	if (ret)
 		return ret;
 
+	ret = nvkm_event_init(&nvd0_disp_chan_uevent, 1, 17, &priv->uevent);
+	if (ret)
+		return ret;
+
 	nv_engine(priv)->sclass = nve0_disp_base_oclass;
 	nv_engine(priv)->cclass = &nv50_disp_cclass;
 	nv_subdev(priv)->intr = nvd0_disp_intr;
@@ -245,7 +249,7 @@ nve0_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	priv->dac.sense = nv50_dac_sense;
 	priv->sor.power = nv50_sor_power;
 	priv->sor.hda_eld = nvd0_hda_eld;
-	priv->sor.hdmi = nvd0_hdmi_ctrl;
+	priv->sor.hdmi = nve0_hdmi_ctrl;
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nvf0.c b/drivers/gpu/drm/nouveau/core/engine/disp/nvf0.c
index 04bda4ac4ed3..402d7d67d806 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nvf0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nvf0.c
@@ -68,6 +68,10 @@ nvf0_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	if (ret)
 		return ret;
 
+	ret = nvkm_event_init(&nvd0_disp_chan_uevent, 1, 17, &priv->uevent);
+	if (ret)
+		return ret;
+
 	nv_engine(priv)->sclass = nvf0_disp_base_oclass;
 	nv_engine(priv)->cclass = &nv50_disp_cclass;
 	nv_subdev(priv)->intr = nvd0_disp_intr;
@@ -80,7 +84,7 @@ nvf0_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	priv->dac.sense = nv50_dac_sense;
 	priv->sor.power = nv50_sor_power;
 	priv->sor.hda_eld = nvd0_hda_eld;
-	priv->sor.hdmi = nvd0_hdmi_ctrl;
+	priv->sor.hdmi = nve0_hdmi_ctrl;
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/outpdp.c b/drivers/gpu/drm/nouveau/core/engine/disp/outpdp.c
index 6f6e2a898270..667a9070e006 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/outpdp.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/outpdp.c
@@ -254,7 +254,7 @@ nvkm_output_dp_create_(struct nouveau_object *parent,
 	atomic_set(&outp->lt.done, 0);
 
 	/* link maintenance */
-	ret = nvkm_notify_init(&i2c->event, nvkm_output_dp_irq, true,
+	ret = nvkm_notify_init(NULL, &i2c->event, nvkm_output_dp_irq, true,
 			       &(struct nvkm_i2c_ntfy_req) {
 				.mask = NVKM_I2C_IRQ,
 				.port = outp->base.edid->index,
@@ -268,7 +268,7 @@ nvkm_output_dp_create_(struct nouveau_object *parent,
 	}
 
 	/* hotplug detect, replaces gpio-based mechanism with aux events */
-	ret = nvkm_notify_init(&i2c->event, nvkm_output_dp_hpd, true,
+	ret = nvkm_notify_init(NULL, &i2c->event, nvkm_output_dp_hpd, true,
 			       &(struct nvkm_i2c_ntfy_req) {
 				.mask = NVKM_I2C_PLUG | NVKM_I2C_UNPLUG,
 				.port = outp->base.edid->index,
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/priv.h b/drivers/gpu/drm/nouveau/core/engine/disp/priv.h
index dbd43ae9df81..6a0511d54ce6 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/priv.h
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/priv.h
@@ -40,7 +40,8 @@ int  _nouveau_disp_fini(struct nouveau_object *, bool);
 extern struct nouveau_oclass *nvkm_output_oclass;
 extern struct nouveau_oclass *nvkm_connector_oclass;
 
-int  nouveau_disp_vblank_ctor(void *data, u32 size, struct nvkm_notify *);
+int  nouveau_disp_vblank_ctor(struct nouveau_object *, void *data, u32 size,
+			      struct nvkm_notify *);
 void nouveau_disp_vblank(struct nouveau_disp *, int head);
 int  nouveau_disp_ntfy(struct nouveau_object *, u32, struct nvkm_event **);
 
diff --git a/drivers/gpu/drm/nouveau/core/engine/fifo/base.c b/drivers/gpu/drm/nouveau/core/engine/fifo/base.c
index 0f999fc45ab9..ac8375cf4eef 100644
--- a/drivers/gpu/drm/nouveau/core/engine/fifo/base.c
+++ b/drivers/gpu/drm/nouveau/core/engine/fifo/base.c
@@ -34,7 +34,8 @@
 #include <engine/fifo.h>
 
 static int
-nouveau_fifo_event_ctor(void *data, u32 size, struct nvkm_notify *notify)
+nouveau_fifo_event_ctor(struct nouveau_object *object, void *data, u32 size,
+			struct nvkm_notify *notify)
 {
 	if (size == 0) {
 		notify->size  = 0;
@@ -170,7 +171,8 @@ _nouveau_fifo_channel_wr32(struct nouveau_object *object, u64 addr, u32 data)
 }
 
 int
-nouveau_fifo_uevent_ctor(void *data, u32 size, struct nvkm_notify *notify)
+nouveau_fifo_uevent_ctor(struct nouveau_object *object, void *data, u32 size,
+			 struct nvkm_notify *notify)
 {
 	union {
 		struct nvif_notify_uevent_req none;
diff --git a/drivers/gpu/drm/nouveau/core/engine/software/nv50.c b/drivers/gpu/drm/nouveau/core/engine/software/nv50.c
index 4d2994d8cc32..a0fec205f9db 100644
--- a/drivers/gpu/drm/nouveau/core/engine/software/nv50.c
+++ b/drivers/gpu/drm/nouveau/core/engine/software/nv50.c
@@ -175,7 +175,8 @@ nv50_software_context_ctor(struct nouveau_object *parent,
 		return ret;
 
 	for (i = 0; pdisp && i < pdisp->vblank.index_nr; i++) {
-		ret = nvkm_notify_init(&pdisp->vblank, pclass->vblank, false,
+		ret = nvkm_notify_init(NULL, &pdisp->vblank, pclass->vblank,
+				       false,
 				       &(struct nvif_notify_head_req_v0) {
 					.head = i,
 				       },
diff --git a/drivers/gpu/drm/nouveau/core/include/core/client.h b/drivers/gpu/drm/nouveau/core/include/core/client.h
index 1794a05205d8..b0ce9f6680b5 100644
--- a/drivers/gpu/drm/nouveau/core/include/core/client.h
+++ b/drivers/gpu/drm/nouveau/core/include/core/client.h
@@ -48,7 +48,7 @@ int  nouveau_client_init(struct nouveau_client *);
 int  nouveau_client_fini(struct nouveau_client *, bool suspend);
 const char *nouveau_client_name(void *obj);
 
-int nvkm_client_notify_new(struct nouveau_client *, struct nvkm_event *,
+int nvkm_client_notify_new(struct nouveau_object *, struct nvkm_event *,
 			   void *data, u32 size);
 int nvkm_client_notify_del(struct nouveau_client *, int index);
 int nvkm_client_notify_get(struct nouveau_client *, int index);
diff --git a/drivers/gpu/drm/nouveau/core/include/core/device.h b/drivers/gpu/drm/nouveau/core/include/core/device.h
index 8743766454a5..1d9d893929bb 100644
--- a/drivers/gpu/drm/nouveau/core/include/core/device.h
+++ b/drivers/gpu/drm/nouveau/core/include/core/device.h
@@ -24,6 +24,7 @@ enum nv_subdev_type {
 	 * been created, and are allowed to assume any subdevs in the
 	 * list above them exist and have been initialised.
 	 */
+	NVDEV_SUBDEV_FUSE,
 	NVDEV_SUBDEV_MXM,
 	NVDEV_SUBDEV_MC,
 	NVDEV_SUBDEV_BUS,
diff --git a/drivers/gpu/drm/nouveau/core/include/core/event.h b/drivers/gpu/drm/nouveau/core/include/core/event.h
index 51e55d03330a..92876528972f 100644
--- a/drivers/gpu/drm/nouveau/core/include/core/event.h
+++ b/drivers/gpu/drm/nouveau/core/include/core/event.h
@@ -4,7 +4,8 @@
 #include <core/notify.h>
 
 struct nvkm_event_func {
-	int  (*ctor)(void *data, u32 size, struct nvkm_notify *);
+	int  (*ctor)(struct nouveau_object *, void *data, u32 size,
+		     struct nvkm_notify *);
 	void (*send)(void *data, u32 size, struct nvkm_notify *);
 	void (*init)(struct nvkm_event *, int type, int index);
 	void (*fini)(struct nvkm_event *, int type, int index);
diff --git a/drivers/gpu/drm/nouveau/core/include/core/mm.h b/drivers/gpu/drm/nouveau/core/include/core/mm.h
index 2bf7d0e32261..bfe6931544fe 100644
--- a/drivers/gpu/drm/nouveau/core/include/core/mm.h
+++ b/drivers/gpu/drm/nouveau/core/include/core/mm.h
@@ -6,6 +6,10 @@ struct nouveau_mm_node {
 	struct list_head fl_entry;
 	struct list_head rl_entry;
 
+#define NVKM_MM_HEAP_ANY 0x00
+	u8  heap;
+#define NVKM_MM_TYPE_NONE 0x00
+#define NVKM_MM_TYPE_HOLE 0xff
 	u8  type;
 	u32 offset;
 	u32 length;
@@ -27,10 +31,10 @@ nouveau_mm_initialised(struct nouveau_mm *mm)
 
 int  nouveau_mm_init(struct nouveau_mm *, u32 offset, u32 length, u32 block);
 int  nouveau_mm_fini(struct nouveau_mm *);
-int  nouveau_mm_head(struct nouveau_mm *, u8 type, u32 size_max, u32 size_min,
-		     u32 align, struct nouveau_mm_node **);
-int  nouveau_mm_tail(struct nouveau_mm *, u8 type, u32 size_max, u32 size_min,
-		     u32 align, struct nouveau_mm_node **);
+int  nouveau_mm_head(struct nouveau_mm *, u8 heap, u8 type, u32 size_max,
+		     u32 size_min, u32 align, struct nouveau_mm_node **);
+int  nouveau_mm_tail(struct nouveau_mm *, u8 heap, u8 type, u32 size_max,
+		     u32 size_min, u32 align, struct nouveau_mm_node **);
 void nouveau_mm_free(struct nouveau_mm *, struct nouveau_mm_node **);
 
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/include/core/notify.h b/drivers/gpu/drm/nouveau/core/include/core/notify.h
index 1262d8f020f3..a7c3c5f578cc 100644
--- a/drivers/gpu/drm/nouveau/core/include/core/notify.h
+++ b/drivers/gpu/drm/nouveau/core/include/core/notify.h
@@ -25,8 +25,9 @@ struct nvkm_notify {
 	const void *data;
 };
 
-int  nvkm_notify_init(struct nvkm_event *, int (*func)(struct nvkm_notify *),
-		      bool work, void *data, u32 size, u32 reply,
+int  nvkm_notify_init(struct nouveau_object *, struct nvkm_event *,
+		      int (*func)(struct nvkm_notify *), bool work,
+		      void *data, u32 size, u32 reply,
 		      struct nvkm_notify *);
 void nvkm_notify_fini(struct nvkm_notify *);
 void nvkm_notify_get(struct nvkm_notify *);
diff --git a/drivers/gpu/drm/nouveau/core/include/engine/fifo.h b/drivers/gpu/drm/nouveau/core/include/engine/fifo.h
index e5e4d930b2c2..2007453f6fce 100644
--- a/drivers/gpu/drm/nouveau/core/include/engine/fifo.h
+++ b/drivers/gpu/drm/nouveau/core/include/engine/fifo.h
@@ -116,7 +116,8 @@ extern struct nouveau_oclass *nve0_fifo_oclass;
 extern struct nouveau_oclass *gk20a_fifo_oclass;
 extern struct nouveau_oclass *nv108_fifo_oclass;
 
-int  nouveau_fifo_uevent_ctor(void *, u32, struct nvkm_notify *);
+int  nouveau_fifo_uevent_ctor(struct nouveau_object *, void *, u32,
+			      struct nvkm_notify *);
 void nouveau_fifo_uevent(struct nouveau_fifo *);
 
 void nv04_fifo_intr(struct nouveau_subdev *);
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/bar.h b/drivers/gpu/drm/nouveau/core/include/subdev/bar.h
index be037fac534c..257ddf6d36d4 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/bar.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/bar.h
@@ -12,7 +12,6 @@ struct nouveau_bar {
 
 	int (*alloc)(struct nouveau_bar *, struct nouveau_object *,
 		     struct nouveau_mem *, struct nouveau_object **);
-	void __iomem *iomem;
 
 	int (*kmap)(struct nouveau_bar *, struct nouveau_mem *,
 		    u32 flags, struct nouveau_vma *);
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/bios/M0205.h b/drivers/gpu/drm/nouveau/core/include/subdev/bios/M0205.h
new file mode 100644
index 000000000000..e171120cec81
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/bios/M0205.h
@@ -0,0 +1,32 @@
+#ifndef __NVBIOS_M0205_H__
+#define __NVBIOS_M0205_H__
+
+struct nvbios_M0205T {
+	u16 freq;
+};
+
+u32 nvbios_M0205Te(struct nouveau_bios *,
+		   u8 *ver, u8 *hdr, u8 *cnt, u8 *len, u8 *snr, u8 *ssz);
+u32 nvbios_M0205Tp(struct nouveau_bios *,
+		   u8 *ver, u8 *hdr, u8 *cnt, u8 *len, u8 *snr, u8 *ssz,
+		   struct nvbios_M0205T *);
+
+struct nvbios_M0205E {
+	u8 type;
+};
+
+u32 nvbios_M0205Ee(struct nouveau_bios *, int idx,
+		   u8 *ver, u8 *hdr, u8 *cnt, u8 *len);
+u32 nvbios_M0205Ep(struct nouveau_bios *, int idx,
+		   u8 *ver, u8 *hdr, u8 *cnt, u8 *len,
+		   struct nvbios_M0205E *);
+
+struct nvbios_M0205S {
+	u8 data;
+};
+
+u32 nvbios_M0205Se(struct nouveau_bios *, int ent, int idx, u8 *ver, u8 *hdr);
+u32 nvbios_M0205Sp(struct nouveau_bios *, int ent, int idx, u8 *ver, u8 *hdr,
+		   struct nvbios_M0205S *);
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/bios/M0209.h b/drivers/gpu/drm/nouveau/core/include/subdev/bios/M0209.h
new file mode 100644
index 000000000000..67dc50d837bc
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/bios/M0209.h
@@ -0,0 +1,30 @@
+#ifndef __NVBIOS_M0209_H__
+#define __NVBIOS_M0209_H__
+
+u32 nvbios_M0209Te(struct nouveau_bios *,
+		   u8 *ver, u8 *hdr, u8 *cnt, u8 *len, u8 *snr, u8 *ssz);
+
+struct nvbios_M0209E {
+	u8 v00_40;
+	u8 bits;
+	u8 modulo;
+	u8 v02_40;
+	u8 v02_07;
+	u8 v03;
+};
+
+u32 nvbios_M0209Ee(struct nouveau_bios *, int idx,
+		   u8 *ver, u8 *hdr, u8 *cnt, u8 *len);
+u32 nvbios_M0209Ep(struct nouveau_bios *, int idx,
+		   u8 *ver, u8 *hdr, u8 *cnt, u8 *len,
+		   struct nvbios_M0209E *);
+
+struct nvbios_M0209S {
+	u32 data[0x200];
+};
+
+u32 nvbios_M0209Se(struct nouveau_bios *, int ent, int idx, u8 *ver, u8 *hdr);
+u32 nvbios_M0209Sp(struct nouveau_bios *, int ent, int idx, u8 *ver, u8 *hdr,
+		   struct nvbios_M0209S *);
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/bios/fan.h b/drivers/gpu/drm/nouveau/core/include/subdev/bios/fan.h
new file mode 100644
index 000000000000..119d0874e041
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/bios/fan.h
@@ -0,0 +1,8 @@
+#ifndef __NVBIOS_FAN_H__
+#define __NVBIOS_FAN_H__
+
+#include <subdev/bios/therm.h>
+
+u16 nvbios_fan_parse(struct nouveau_bios *bios, struct nvbios_therm_fan *fan);
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/bios/ramcfg.h b/drivers/gpu/drm/nouveau/core/include/subdev/bios/ramcfg.h
index c086ac6d677d..a685bbd04568 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/bios/ramcfg.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/bios/ramcfg.h
@@ -4,60 +4,118 @@
 struct nouveau_bios;
 
 struct nvbios_ramcfg {
-	unsigned rammap_11_08_01:1;
-	unsigned rammap_11_08_0c:2;
-	unsigned rammap_11_08_10:1;
-	unsigned rammap_11_11_0c:2;
+	unsigned rammap_ver;
+	unsigned rammap_hdr;
+	unsigned rammap_min;
+	unsigned rammap_max;
+	union {
+		struct {
+			unsigned rammap_10_04_02:1;
+			unsigned rammap_10_04_08:1;
+		};
+		struct {
+			unsigned rammap_11_08_01:1;
+			unsigned rammap_11_08_0c:2;
+			unsigned rammap_11_08_10:1;
+			unsigned rammap_11_09_01ff:9;
+			unsigned rammap_11_0a_03fe:9;
+			unsigned rammap_11_0a_0400:1;
+			unsigned rammap_11_0a_0800:1;
+			unsigned rammap_11_0b_01f0:5;
+			unsigned rammap_11_0b_0200:1;
+			unsigned rammap_11_0b_0400:1;
+			unsigned rammap_11_0b_0800:1;
+			unsigned rammap_11_0d:8;
+			unsigned rammap_11_0e:8;
+			unsigned rammap_11_0f:8;
+			unsigned rammap_11_11_0c:2;
+		};
+	};
 
-	unsigned ramcfg_11_01_01:1;
-	unsigned ramcfg_11_01_02:1;
-	unsigned ramcfg_11_01_04:1;
-	unsigned ramcfg_11_01_08:1;
-	unsigned ramcfg_11_01_10:1;
-	unsigned ramcfg_11_01_20:1;
-	unsigned ramcfg_11_01_40:1;
-	unsigned ramcfg_11_01_80:1;
-	unsigned ramcfg_11_02_03:2;
-	unsigned ramcfg_11_02_04:1;
-	unsigned ramcfg_11_02_08:1;
-	unsigned ramcfg_11_02_10:1;
-	unsigned ramcfg_11_02_40:1;
-	unsigned ramcfg_11_02_80:1;
-	unsigned ramcfg_11_03_0f:4;
-	unsigned ramcfg_11_03_30:2;
-	unsigned ramcfg_11_03_c0:2;
-	unsigned ramcfg_11_03_f0:4;
-	unsigned ramcfg_11_04:8;
-	unsigned ramcfg_11_06:8;
-	unsigned ramcfg_11_07_02:1;
-	unsigned ramcfg_11_07_04:1;
-	unsigned ramcfg_11_07_08:1;
-	unsigned ramcfg_11_07_10:1;
-	unsigned ramcfg_11_07_40:1;
-	unsigned ramcfg_11_07_80:1;
-	unsigned ramcfg_11_08_01:1;
-	unsigned ramcfg_11_08_02:1;
-	unsigned ramcfg_11_08_04:1;
-	unsigned ramcfg_11_08_08:1;
-	unsigned ramcfg_11_08_10:1;
-	unsigned ramcfg_11_08_20:1;
-	unsigned ramcfg_11_09:8;
+	unsigned ramcfg_ver;
+	unsigned ramcfg_hdr;
+	unsigned ramcfg_timing;
+	union {
+		struct {
+			unsigned ramcfg_10_02_01:1;
+			unsigned ramcfg_10_02_02:1;
+			unsigned ramcfg_10_02_04:1;
+			unsigned ramcfg_10_02_08:1;
+			unsigned ramcfg_10_02_10:1;
+			unsigned ramcfg_10_02_20:1;
+			unsigned ramcfg_10_02_40:1;
+			unsigned ramcfg_10_03_0f:4;
+			unsigned ramcfg_10_05:8;
+			unsigned ramcfg_10_06:8;
+			unsigned ramcfg_10_07:8;
+			unsigned ramcfg_10_08:8;
+			unsigned ramcfg_10_09_0f:4;
+			unsigned ramcfg_10_09_f0:4;
+		};
+		struct {
+			unsigned ramcfg_11_01_01:1;
+			unsigned ramcfg_11_01_02:1;
+			unsigned ramcfg_11_01_04:1;
+			unsigned ramcfg_11_01_08:1;
+			unsigned ramcfg_11_01_10:1;
+			unsigned ramcfg_11_01_20:1;
+			unsigned ramcfg_11_01_40:1;
+			unsigned ramcfg_11_01_80:1;
+			unsigned ramcfg_11_02_03:2;
+			unsigned ramcfg_11_02_04:1;
+			unsigned ramcfg_11_02_08:1;
+			unsigned ramcfg_11_02_10:1;
+			unsigned ramcfg_11_02_40:1;
+			unsigned ramcfg_11_02_80:1;
+			unsigned ramcfg_11_03_0f:4;
+			unsigned ramcfg_11_03_30:2;
+			unsigned ramcfg_11_03_c0:2;
+			unsigned ramcfg_11_03_f0:4;
+			unsigned ramcfg_11_04:8;
+			unsigned ramcfg_11_06:8;
+			unsigned ramcfg_11_07_02:1;
+			unsigned ramcfg_11_07_04:1;
+			unsigned ramcfg_11_07_08:1;
+			unsigned ramcfg_11_07_10:1;
+			unsigned ramcfg_11_07_40:1;
+			unsigned ramcfg_11_07_80:1;
+			unsigned ramcfg_11_08_01:1;
+			unsigned ramcfg_11_08_02:1;
+			unsigned ramcfg_11_08_04:1;
+			unsigned ramcfg_11_08_08:1;
+			unsigned ramcfg_11_08_10:1;
+			unsigned ramcfg_11_08_20:1;
+			unsigned ramcfg_11_09:8;
+		};
+	};
 
+	unsigned timing_ver;
+	unsigned timing_hdr;
 	unsigned timing[11];
-	unsigned timing_20_2e_03:2;
-	unsigned timing_20_2e_30:2;
-	unsigned timing_20_2e_c0:2;
-	unsigned timing_20_2f_03:2;
-	unsigned timing_20_2c_003f:6;
-	unsigned timing_20_2c_1fc0:7;
-	unsigned timing_20_30_f8:5;
-	unsigned timing_20_30_07:3;
-	unsigned timing_20_31_0007:3;
-	unsigned timing_20_31_0078:4;
-	unsigned timing_20_31_0780:4;
-	unsigned timing_20_31_0800:1;
-	unsigned timing_20_31_7000:3;
-	unsigned timing_20_31_8000:1;
+	union {
+		struct {
+			unsigned timing_10_WR:8;
+			unsigned timing_10_CL:8;
+			unsigned timing_10_ODT:3;
+			unsigned timing_10_CWL:8;
+		};
+		struct {
+			unsigned timing_20_2e_03:2;
+			unsigned timing_20_2e_30:2;
+			unsigned timing_20_2e_c0:2;
+			unsigned timing_20_2f_03:2;
+			unsigned timing_20_2c_003f:6;
+			unsigned timing_20_2c_1fc0:7;
+			unsigned timing_20_30_f8:5;
+			unsigned timing_20_30_07:3;
+			unsigned timing_20_31_0007:3;
+			unsigned timing_20_31_0078:4;
+			unsigned timing_20_31_0780:4;
+			unsigned timing_20_31_0800:1;
+			unsigned timing_20_31_7000:3;
+			unsigned timing_20_31_8000:1;
+		};
+	};
 };
 
 u8 nvbios_ramcfg_count(struct nouveau_bios *);
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/bios/rammap.h b/drivers/gpu/drm/nouveau/core/include/subdev/bios/rammap.h
index 5bdf8e4db40a..47e021d3e20d 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/bios/rammap.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/bios/rammap.h
@@ -8,9 +8,10 @@ u32 nvbios_rammapTe(struct nouveau_bios *, u8 *ver, u8 *hdr,
 
 u32 nvbios_rammapEe(struct nouveau_bios *, int idx,
 		    u8 *ver, u8 *hdr, u8 *cnt, u8 *len);
+u32 nvbios_rammapEp(struct nouveau_bios *, int idx,
+		    u8 *ver, u8 *hdr, u8 *cnt, u8 *len,
+		    struct nvbios_ramcfg *);
 u32 nvbios_rammapEm(struct nouveau_bios *, u16 mhz,
-		    u8 *ver, u8 *hdr, u8 *cnt, u8 *len);
-u32 nvbios_rammapEp(struct nouveau_bios *, u16 mhz,
 		    u8 *ver, u8 *hdr, u8 *cnt, u8 *len,
 		    struct nvbios_ramcfg *);
 
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/bios/therm.h b/drivers/gpu/drm/nouveau/core/include/subdev/bios/therm.h
index 8dc5051df55d..295d093f3b30 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/bios/therm.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/bios/therm.h
@@ -23,6 +23,12 @@ struct nvbios_therm_sensor {
 	struct nvbios_therm_threshold thrs_shutdown;
 };
 
+enum nvbios_therm_fan_type {
+	NVBIOS_THERM_FAN_UNK = 0,
+	NVBIOS_THERM_FAN_TOGGLE = 1,
+	NVBIOS_THERM_FAN_PWM = 2,
+};
+
 /* no vbios have more than 6 */
 #define NOUVEAU_TEMP_FAN_TRIP_MAX 10
 struct nouveau_therm_trip_point {
@@ -38,7 +44,9 @@ enum nvbios_therm_fan_mode {
 };
 
 struct nvbios_therm_fan {
-	u16 pwm_freq;
+	enum nvbios_therm_fan_type type;
+
+	u32 pwm_freq;
 
 	u8 min_duty;
 	u8 max_duty;
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/clock.h b/drivers/gpu/drm/nouveau/core/include/subdev/clock.h
index a5ca00dd2f61..36ed035d4d42 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/clock.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/clock.h
@@ -29,6 +29,7 @@ enum nv_clk_src {
 	nv_clk_src_mdiv,
 
 	nv_clk_src_core,
+	nv_clk_src_core_intm,
 	nv_clk_src_shader,
 
 	nv_clk_src_mem,
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/fb.h b/drivers/gpu/drm/nouveau/core/include/subdev/fb.h
index 871e73914b24..8d0032f15205 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/fb.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/fb.h
@@ -111,6 +111,7 @@ extern struct nouveau_oclass *gm107_fb_oclass;
 #include <subdev/bios/ramcfg.h>
 
 struct nouveau_ram_data {
+	struct list_head head;
 	struct nvbios_ramcfg bios;
 	u32 freq;
 };
@@ -136,6 +137,7 @@ struct nouveau_ram {
 
 	int ranks;
 	int parts;
+	int part_mask;
 
 	int  (*get)(struct nouveau_fb *, u64 size, u32 align,
 		    u32 size_nc, u32 type, struct nouveau_mem **);
@@ -144,11 +146,6 @@ struct nouveau_ram {
 	int  (*calc)(struct nouveau_fb *, u32 freq);
 	int  (*prog)(struct nouveau_fb *);
 	void (*tidy)(struct nouveau_fb *);
-	struct {
-		u8  version;
-		u32 data;
-		u8  size;
-	} rammap, ramcfg, timing;
 	u32 freq;
 	u32 mr[16];
 	u32 mr1_nuts;
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/fb/regsnv04.h b/drivers/gpu/drm/nouveau/core/include/subdev/fb/regsnv04.h
new file mode 100644
index 000000000000..0f7fc0c52ab2
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/fb/regsnv04.h
@@ -0,0 +1,21 @@
+#ifndef __NOUVEAU_FB_REGS_04_H__
+#define __NOUVEAU_FB_REGS_04_H__
+
+#define NV04_PFB_BOOT_0						0x00100000
+#	define NV04_PFB_BOOT_0_RAM_AMOUNT			0x00000003
+#	define NV04_PFB_BOOT_0_RAM_AMOUNT_32MB			0x00000000
+#	define NV04_PFB_BOOT_0_RAM_AMOUNT_4MB			0x00000001
+#	define NV04_PFB_BOOT_0_RAM_AMOUNT_8MB			0x00000002
+#	define NV04_PFB_BOOT_0_RAM_AMOUNT_16MB			0x00000003
+#	define NV04_PFB_BOOT_0_RAM_WIDTH_128			0x00000004
+#	define NV04_PFB_BOOT_0_RAM_TYPE				0x00000028
+#	define NV04_PFB_BOOT_0_RAM_TYPE_SGRAM_8MBIT		0x00000000
+#	define NV04_PFB_BOOT_0_RAM_TYPE_SGRAM_16MBIT		0x00000008
+#	define NV04_PFB_BOOT_0_RAM_TYPE_SGRAM_16MBIT_4BANK	0x00000010
+#	define NV04_PFB_BOOT_0_RAM_TYPE_SDRAM_16MBIT		0x00000018
+#	define NV04_PFB_BOOT_0_RAM_TYPE_SDRAM_64MBIT		0x00000020
+#	define NV04_PFB_BOOT_0_RAM_TYPE_SDRAM_64MBITX16		0x00000028
+#	define NV04_PFB_BOOT_0_UMA_ENABLE			0x00000100
+#	define NV04_PFB_BOOT_0_UMA_SIZE				0x0000f000
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/fuse.h b/drivers/gpu/drm/nouveau/core/include/subdev/fuse.h
new file mode 100644
index 000000000000..2b1ddb2a9a7d
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/fuse.h
@@ -0,0 +1,30 @@
+#ifndef __NOUVEAU_FUSE_H__
+#define __NOUVEAU_FUSE_H__
+
+#include <core/subdev.h>
+#include <core/device.h>
+
+struct nouveau_fuse {
+	struct nouveau_subdev base;
+};
+
+static inline struct nouveau_fuse *
+nouveau_fuse(void *obj)
+{
+	return (void *)nv_device(obj)->subdev[NVDEV_SUBDEV_FUSE];
+}
+
+#define nouveau_fuse_create(p, e, o, d)                                        \
+	nouveau_fuse_create_((p), (e), (o), sizeof(**d), (void **)d)
+
+int  nouveau_fuse_create_(struct nouveau_object *, struct nouveau_object *,
+			  struct nouveau_oclass *, int, void **);
+void _nouveau_fuse_dtor(struct nouveau_object *);
+int  _nouveau_fuse_init(struct nouveau_object *);
+#define _nouveau_fuse_fini _nouveau_subdev_fini
+
+extern struct nouveau_oclass g80_fuse_oclass;
+extern struct nouveau_oclass gf100_fuse_oclass;
+extern struct nouveau_oclass gm107_fuse_oclass;
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/gpio.h b/drivers/gpu/drm/nouveau/core/include/subdev/gpio.h
index b73733d21cc7..f855140dbcb7 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/gpio.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/gpio.h
@@ -40,7 +40,7 @@ nouveau_gpio(void *obj)
 
 extern struct nouveau_oclass *nv10_gpio_oclass;
 extern struct nouveau_oclass *nv50_gpio_oclass;
-extern struct nouveau_oclass *nv92_gpio_oclass;
+extern struct nouveau_oclass *nv94_gpio_oclass;
 extern struct nouveau_oclass *nvd0_gpio_oclass;
 extern struct nouveau_oclass *nve0_gpio_oclass;
 
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/pwr.h b/drivers/gpu/drm/nouveau/core/include/subdev/pwr.h
index f73feec151db..bf3d1f611333 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/pwr.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/pwr.h
@@ -47,5 +47,8 @@ void nouveau_memx_wr32(struct nouveau_memx *, u32 addr, u32 data);
 void nouveau_memx_wait(struct nouveau_memx *,
 		       u32 addr, u32 mask, u32 data, u32 nsec);
 void nouveau_memx_nsec(struct nouveau_memx *, u32 nsec);
+void nouveau_memx_wait_vblank(struct nouveau_memx *);
+void nouveau_memx_block(struct nouveau_memx *);
+void nouveau_memx_unblock(struct nouveau_memx *);
 
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/therm.h b/drivers/gpu/drm/nouveau/core/include/subdev/therm.h
index d4a68179e586..a437597dcafc 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/therm.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/therm.h
@@ -78,5 +78,6 @@ extern struct nouveau_oclass nv50_therm_oclass;
 extern struct nouveau_oclass nv84_therm_oclass;
 extern struct nouveau_oclass nva3_therm_oclass;
 extern struct nouveau_oclass nvd0_therm_oclass;
+extern struct nouveau_oclass gm107_therm_oclass;
 
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bar/base.c b/drivers/gpu/drm/nouveau/core/subdev/bar/base.c
index 8bcbdf39cfb2..b1adc69efd88 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/bar/base.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/bar/base.c
@@ -38,10 +38,12 @@ struct nouveau_barobj {
 static int
 nouveau_barobj_ctor(struct nouveau_object *parent,
 		    struct nouveau_object *engine,
-		    struct nouveau_oclass *oclass, void *mem, u32 size,
+		    struct nouveau_oclass *oclass, void *data, u32 size,
 		    struct nouveau_object **pobject)
 {
+	struct nouveau_device *device = nv_device(parent);
 	struct nouveau_bar *bar = (void *)engine;
+	struct nouveau_mem *mem = data;
 	struct nouveau_barobj *barobj;
 	int ret;
 
@@ -54,7 +56,13 @@ nouveau_barobj_ctor(struct nouveau_object *parent,
 	if (ret)
 		return ret;
 
-	barobj->iomem = bar->iomem + (u32)barobj->vma.offset;
+	barobj->iomem = ioremap(nv_device_resource_start(device, 3) +
+				(u32)barobj->vma.offset, mem->size << 12);
+	if (!barobj->iomem) {
+		nv_warn(bar, "PRAMIN ioremap failed\n");
+		return -ENOMEM;
+	}
+
 	return 0;
 }
 
@@ -63,8 +71,11 @@ nouveau_barobj_dtor(struct nouveau_object *object)
 {
 	struct nouveau_bar *bar = (void *)object->engine;
 	struct nouveau_barobj *barobj = (void *)object;
-	if (barobj->vma.node)
+	if (barobj->vma.node) {
+		if (barobj->iomem)
+			iounmap(barobj->iomem);
 		bar->unmap(bar, &barobj->vma);
+	}
 	nouveau_object_destroy(&barobj->base);
 }
 
@@ -99,12 +110,11 @@ nouveau_bar_alloc(struct nouveau_bar *bar, struct nouveau_object *parent,
 		  struct nouveau_mem *mem, struct nouveau_object **pobject)
 {
 	struct nouveau_object *engine = nv_object(bar);
-	int ret = -ENOMEM;
-	if (bar->iomem) {
-		ret = nouveau_object_ctor(parent, engine,
-					  &nouveau_barobj_oclass,
-					  mem, 0, pobject);
-	}
+	struct nouveau_object *gpuobj;
+	int ret = nouveau_object_ctor(parent, engine, &nouveau_barobj_oclass,
+				      mem, 0, &gpuobj);
+	if (ret == 0)
+		*pobject = gpuobj;
 	return ret;
 }
 
@@ -113,7 +123,6 @@ nouveau_bar_create_(struct nouveau_object *parent,
 		    struct nouveau_object *engine,
 		    struct nouveau_oclass *oclass, int length, void **pobject)
 {
-	struct nouveau_device *device = nv_device(parent);
 	struct nouveau_bar *bar;
 	int ret;
 
@@ -123,21 +132,12 @@ nouveau_bar_create_(struct nouveau_object *parent,
 	if (ret)
 		return ret;
 
-	if (nv_device_resource_len(device, 3) != 0) {
-		bar->iomem = ioremap(nv_device_resource_start(device, 3),
-				     nv_device_resource_len(device, 3));
-		if (!bar->iomem)
-			nv_warn(bar, "PRAMIN ioremap failed\n");
-	}
-
 	return 0;
 }
 
 void
 nouveau_bar_destroy(struct nouveau_bar *bar)
 {
-	if (bar->iomem)
-		iounmap(bar->iomem);
 	nouveau_subdev_destroy(&bar->base);
 }
 
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/M0205.c b/drivers/gpu/drm/nouveau/core/subdev/bios/M0205.c
new file mode 100644
index 000000000000..ac9617c5fc2a
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/M0205.c
@@ -0,0 +1,136 @@
+/*
+ * Copyright 2013 Red Hat 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: Ben Skeggs
+ */
+
+#include <subdev/bios.h>
+#include <subdev/bios/bit.h>
+#include <subdev/bios/M0205.h>
+
+u32
+nvbios_M0205Te(struct nouveau_bios *bios,
+	       u8 *ver, u8 *hdr, u8 *cnt, u8 *len, u8 *snr, u8 *ssz)
+{
+	struct bit_entry bit_M;
+	u32 data = 0x00000000;
+
+	if (!bit_entry(bios, 'M', &bit_M)) {
+		if (bit_M.version == 2 && bit_M.length > 0x08)
+			data = nv_ro32(bios, bit_M.offset + 0x05);
+		if (data) {
+			*ver = nv_ro08(bios, data + 0x00);
+			switch (*ver) {
+			case 0x10:
+				*hdr = nv_ro08(bios, data + 0x01);
+				*len = nv_ro08(bios, data + 0x02);
+				*ssz = nv_ro08(bios, data + 0x03);
+				*snr = nv_ro08(bios, data + 0x04);
+				*cnt = nv_ro08(bios, data + 0x05);
+				return data;
+			default:
+				break;
+			}
+		}
+	}
+
+	return 0x00000000;
+}
+
+u32
+nvbios_M0205Tp(struct nouveau_bios *bios,
+	       u8 *ver, u8 *hdr, u8 *cnt, u8 *len, u8 *snr, u8 *ssz,
+	       struct nvbios_M0205T *info)
+{
+	u32 data = nvbios_M0205Te(bios, ver, hdr, cnt, len, snr, ssz);
+	memset(info, 0x00, sizeof(*info));
+	switch (!!data * *ver) {
+	case 0x10:
+		info->freq = nv_ro16(bios, data + 0x06);
+		break;
+	default:
+		break;
+	}
+	return data;
+}
+
+u32
+nvbios_M0205Ee(struct nouveau_bios *bios, int idx,
+	       u8 *ver, u8 *hdr, u8 *cnt, u8 *len)
+{
+	u8  snr, ssz;
+	u32 data = nvbios_M0205Te(bios, ver, hdr, cnt, len, &snr, &ssz);
+	if (data && idx < *cnt) {
+		data = data + *hdr + idx * (*len + (snr * ssz));
+		*hdr = *len;
+		*cnt = snr;
+		*len = ssz;
+		return data;
+	}
+	return 0x00000000;
+}
+
+u32
+nvbios_M0205Ep(struct nouveau_bios *bios, int idx,
+	       u8 *ver, u8 *hdr, u8 *cnt, u8 *len,
+	       struct nvbios_M0205E *info)
+{
+	u32 data = nvbios_M0205Ee(bios, idx, ver, hdr, cnt, len);
+	memset(info, 0x00, sizeof(*info));
+	switch (!!data * *ver) {
+	case 0x10:
+		info->type = nv_ro08(bios, data + 0x00) & 0x0f;
+		return data;
+	default:
+		break;
+	}
+	return 0x00000000;
+}
+
+u32
+nvbios_M0205Se(struct nouveau_bios *bios, int ent, int idx, u8 *ver, u8 *hdr)
+{
+
+	u8  cnt, len;
+	u32 data = nvbios_M0205Ee(bios, ent, ver, hdr, &cnt, &len);
+	if (data && idx < cnt) {
+		data = data + *hdr + idx * len;
+		*hdr = len;
+		return data;
+	}
+	return 0x00000000;
+}
+
+u32
+nvbios_M0205Sp(struct nouveau_bios *bios, int ent, int idx, u8 *ver, u8 *hdr,
+	       struct nvbios_M0205S *info)
+{
+	u32 data = nvbios_M0205Se(bios, ent, idx, ver, hdr);
+	memset(info, 0x00, sizeof(*info));
+	switch (!!data * *ver) {
+	case 0x10:
+		info->data = nv_ro08(bios, data + 0x00);
+		return data;
+	default:
+		break;
+	}
+	return 0x00000000;
+}
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/M0209.c b/drivers/gpu/drm/nouveau/core/subdev/bios/M0209.c
new file mode 100644
index 000000000000..b142a510e89f
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/M0209.c
@@ -0,0 +1,137 @@
+/*
+ * Copyright 2013 Red Hat 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: Ben Skeggs
+ */
+
+#include <subdev/bios.h>
+#include <subdev/bios/bit.h>
+#include <subdev/bios/M0209.h>
+
+u32
+nvbios_M0209Te(struct nouveau_bios *bios,
+	       u8 *ver, u8 *hdr, u8 *cnt, u8 *len, u8 *snr, u8 *ssz)
+{
+	struct bit_entry bit_M;
+	u32 data = 0x00000000;
+
+	if (!bit_entry(bios, 'M', &bit_M)) {
+		if (bit_M.version == 2 && bit_M.length > 0x0c)
+			data = nv_ro32(bios, bit_M.offset + 0x09);
+		if (data) {
+			*ver = nv_ro08(bios, data + 0x00);
+			switch (*ver) {
+			case 0x10:
+				*hdr = nv_ro08(bios, data + 0x01);
+				*len = nv_ro08(bios, data + 0x02);
+				*ssz = nv_ro08(bios, data + 0x03);
+				*snr = 1;
+				*cnt = nv_ro08(bios, data + 0x04);
+				return data;
+			default:
+				break;
+			}
+		}
+	}
+
+	return 0x00000000;
+}
+
+u32
+nvbios_M0209Ee(struct nouveau_bios *bios, int idx,
+	       u8 *ver, u8 *hdr, u8 *cnt, u8 *len)
+{
+	u8  snr, ssz;
+	u32 data = nvbios_M0209Te(bios, ver, hdr, cnt, len, &snr, &ssz);
+	if (data && idx < *cnt) {
+		data = data + *hdr + idx * (*len + (snr * ssz));
+		*hdr = *len;
+		*cnt = snr;
+		*len = ssz;
+		return data;
+	}
+	return 0x00000000;
+}
+
+u32
+nvbios_M0209Ep(struct nouveau_bios *bios, int idx,
+	       u8 *ver, u8 *hdr, u8 *cnt, u8 *len,
+	       struct nvbios_M0209E *info)
+{
+	u32 data = nvbios_M0209Ee(bios, idx, ver, hdr, cnt, len);
+	memset(info, 0x00, sizeof(*info));
+	switch (!!data * *ver) {
+	case 0x10:
+		info->v00_40 = (nv_ro08(bios, data + 0x00) & 0x40) >> 6;
+		info->bits   =  nv_ro08(bios, data + 0x00) & 0x3f;
+		info->modulo =  nv_ro08(bios, data + 0x01);
+		info->v02_40 = (nv_ro08(bios, data + 0x02) & 0x40) >> 6;
+		info->v02_07 =  nv_ro08(bios, data + 0x02) & 0x07;
+		info->v03    =  nv_ro08(bios, data + 0x03);
+		return data;
+	default:
+		break;
+	}
+	return 0x00000000;
+}
+
+u32
+nvbios_M0209Se(struct nouveau_bios *bios, int ent, int idx, u8 *ver, u8 *hdr)
+{
+
+	u8  cnt, len;
+	u32 data = nvbios_M0209Ee(bios, ent, ver, hdr, &cnt, &len);
+	if (data && idx < cnt) {
+		data = data + *hdr + idx * len;
+		*hdr = len;
+		return data;
+	}
+	return 0x00000000;
+}
+
+u32
+nvbios_M0209Sp(struct nouveau_bios *bios, int ent, int idx, u8 *ver, u8 *hdr,
+	       struct nvbios_M0209S *info)
+{
+	struct nvbios_M0209E M0209E;
+	u8  cnt, len;
+	u32 data = nvbios_M0209Ep(bios, ent, ver, hdr, &cnt, &len, &M0209E);
+	if (data) {
+		u32 i, data = nvbios_M0209Se(bios, ent, idx, ver, hdr);
+		memset(info, 0x00, sizeof(*info));
+		switch (!!data * *ver) {
+		case 0x10:
+			for (i = 0; i < ARRAY_SIZE(info->data); i++) {
+				u32 bits = (i % M0209E.modulo) * M0209E.bits;
+				u32 mask = (1ULL << M0209E.bits) - 1;
+				u16  off = bits / 8;
+				u8   mod = bits % 8;
+				info->data[i] = nv_ro32(bios, data + off);
+				info->data[i] = info->data[i] >> mod;
+				info->data[i] = info->data[i] & mask;
+			}
+			return data;
+		default:
+			break;
+		}
+	}
+	return 0x00000000;
+}
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/dcb.c b/drivers/gpu/drm/nouveau/core/subdev/bios/dcb.c
index 88606bfaf847..bd8d348385b3 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/bios/dcb.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/dcb.c
@@ -124,6 +124,7 @@ dcb_outp_parse(struct nouveau_bios *bios, u8 idx, u8 *ver, u8 *len,
 	       struct dcb_output *outp)
 {
 	u16 dcb = dcb_outp(bios, idx, ver, len);
+	memset(outp, 0x00, sizeof(*outp));
 	if (dcb) {
 		if (*ver >= 0x20) {
 			u32 conn = nv_ro32(bios, dcb + 0x00);
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/fan.c b/drivers/gpu/drm/nouveau/core/subdev/bios/fan.c
new file mode 100644
index 000000000000..e419892240f5
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/fan.c
@@ -0,0 +1,93 @@
+/*
+ * Copyright 2014 Martin Peres
+ *
+ * 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: Martin Peres
+ */
+
+#include <subdev/bios.h>
+#include <subdev/bios/bit.h>
+#include <subdev/bios/fan.h>
+
+u16
+nvbios_fan_table(struct nouveau_bios *bios, u8 *ver, u8 *hdr, u8 *cnt, u8 *len)
+{
+	struct bit_entry bit_P;
+	u16 fan = 0x0000;
+
+	if (!bit_entry(bios, 'P', &bit_P)) {
+		if (bit_P.version == 2 && bit_P.length >= 0x5a)
+			fan = nv_ro16(bios, bit_P.offset + 0x58);
+
+		if (fan) {
+			*ver = nv_ro08(bios, fan + 0);
+			switch (*ver) {
+			case 0x10:
+				*hdr = nv_ro08(bios, fan + 1);
+				*len = nv_ro08(bios, fan + 2);
+				*cnt = nv_ro08(bios, fan + 3);
+				return fan;
+			default:
+				break;
+			}
+		}
+	}
+
+	return 0x0000;
+}
+
+u16
+nvbios_fan_entry(struct nouveau_bios *bios, int idx, u8 *ver, u8 *hdr,
+		 u8 *cnt, u8 *len)
+{
+	u16 data = nvbios_fan_table(bios, ver, hdr, cnt, len);
+	if (data && idx < *cnt)
+		return data + *hdr + (idx * (*len));
+	return 0x0000;
+}
+
+u16
+nvbios_fan_parse(struct nouveau_bios *bios, struct nvbios_therm_fan *fan)
+{
+	u8 ver, hdr, cnt, len;
+
+	u16 data = nvbios_fan_entry(bios, 0, &ver, &hdr, &cnt, &len);
+	if (data) {
+		u8 type = nv_ro08(bios, data + 0x00);
+		switch (type) {
+		case 0:
+			fan->type = NVBIOS_THERM_FAN_TOGGLE;
+			break;
+		case 1:
+		case 2:
+			/* TODO: Understand the difference between the two! */
+			fan->type = NVBIOS_THERM_FAN_PWM;
+			break;
+		default:
+			fan->type = NVBIOS_THERM_FAN_UNK;
+		}
+
+		fan->min_duty = nv_ro08(bios, data + 0x02);
+		fan->max_duty = nv_ro08(bios, data + 0x03);
+
+		fan->pwm_freq = nv_ro32(bios, data + 0x0b) & 0xffffff;
+	}
+	return data;
+}
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/rammap.c b/drivers/gpu/drm/nouveau/core/subdev/bios/rammap.c
index 1811b2cb0472..585e69331ccc 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/bios/rammap.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/rammap.c
@@ -75,31 +75,39 @@ nvbios_rammapEe(struct nouveau_bios *bios, int idx,
 }
 
 u32
-nvbios_rammapEm(struct nouveau_bios *bios, u16 khz,
-		u8 *ver, u8 *hdr, u8 *cnt, u8 *len)
-{
-	int idx = 0;
-	u32 data;
-	while ((data = nvbios_rammapEe(bios, idx++, ver, hdr, cnt, len))) {
-		if (khz >= nv_ro16(bios, data + 0x00) &&
-		    khz <= nv_ro16(bios, data + 0x02))
-			break;
-	}
-	return data;
-}
-
-u32
-nvbios_rammapEp(struct nouveau_bios *bios, u16 khz,
+nvbios_rammapEp(struct nouveau_bios *bios, int idx,
 		u8 *ver, u8 *hdr, u8 *cnt, u8 *len,
 		struct nvbios_ramcfg *p)
 {
-	u32 data = nvbios_rammapEm(bios, khz, ver, hdr, cnt, len);
+	u32 data = nvbios_rammapEe(bios, idx, ver, hdr, cnt, len), temp;
 	memset(p, 0x00, sizeof(*p));
+	p->rammap_ver = *ver;
+	p->rammap_hdr = *hdr;
 	switch (!!data * *ver) {
+	case 0x10:
+		p->rammap_min      =  nv_ro16(bios, data + 0x00);
+		p->rammap_max      =  nv_ro16(bios, data + 0x02);
+		p->rammap_10_04_02 = (nv_ro08(bios, data + 0x04) & 0x02) >> 1;
+		p->rammap_10_04_08 = (nv_ro08(bios, data + 0x04) & 0x08) >> 3;
+		break;
 	case 0x11:
+		p->rammap_min      =  nv_ro16(bios, data + 0x00);
+		p->rammap_max      =  nv_ro16(bios, data + 0x02);
 		p->rammap_11_08_01 = (nv_ro08(bios, data + 0x08) & 0x01) >> 0;
 		p->rammap_11_08_0c = (nv_ro08(bios, data + 0x08) & 0x0c) >> 2;
 		p->rammap_11_08_10 = (nv_ro08(bios, data + 0x08) & 0x10) >> 4;
+		temp = nv_ro32(bios, data + 0x09);
+		p->rammap_11_09_01ff = (temp & 0x000001ff) >> 0;
+		p->rammap_11_0a_03fe = (temp & 0x0003fe00) >> 9;
+		p->rammap_11_0a_0400 = (temp & 0x00040000) >> 18;
+		p->rammap_11_0a_0800 = (temp & 0x00080000) >> 19;
+		p->rammap_11_0b_01f0 = (temp & 0x01f00000) >> 20;
+		p->rammap_11_0b_0200 = (temp & 0x02000000) >> 25;
+		p->rammap_11_0b_0400 = (temp & 0x04000000) >> 26;
+		p->rammap_11_0b_0800 = (temp & 0x08000000) >> 27;
+		p->rammap_11_0d    =  nv_ro08(bios, data + 0x0d);
+		p->rammap_11_0e    =  nv_ro08(bios, data + 0x0e);
+		p->rammap_11_0f    =  nv_ro08(bios, data + 0x0f);
 		p->rammap_11_11_0c = (nv_ro08(bios, data + 0x11) & 0x0c) >> 2;
 		break;
 	default:
@@ -110,6 +118,20 @@ nvbios_rammapEp(struct nouveau_bios *bios, u16 khz,
 }
 
 u32
+nvbios_rammapEm(struct nouveau_bios *bios, u16 mhz,
+		u8 *ver, u8 *hdr, u8 *cnt, u8 *len,
+		struct nvbios_ramcfg *info)
+{
+	int idx = 0;
+	u32 data;
+	while ((data = nvbios_rammapEp(bios, idx++, ver, hdr, cnt, len, info))) {
+		if (mhz >= info->rammap_min && mhz <= info->rammap_max)
+			break;
+	}
+	return data;
+}
+
+u32
 nvbios_rammapSe(struct nouveau_bios *bios, u32 data,
 		u8 ever, u8 ehdr, u8 ecnt, u8 elen, int idx,
 		u8 *ver, u8 *hdr)
@@ -129,8 +151,28 @@ nvbios_rammapSp(struct nouveau_bios *bios, u32 data,
 		u8 *ver, u8 *hdr, struct nvbios_ramcfg *p)
 {
 	data = nvbios_rammapSe(bios, data, ever, ehdr, ecnt, elen, idx, ver, hdr);
+	p->ramcfg_ver = *ver;
+	p->ramcfg_hdr = *hdr;
 	switch (!!data * *ver) {
+	case 0x10:
+		p->ramcfg_timing   =  nv_ro08(bios, data + 0x01);
+		p->ramcfg_10_02_01 = (nv_ro08(bios, data + 0x02) & 0x01) >> 0;
+		p->ramcfg_10_02_02 = (nv_ro08(bios, data + 0x02) & 0x02) >> 1;
+		p->ramcfg_10_02_04 = (nv_ro08(bios, data + 0x02) & 0x04) >> 2;
+		p->ramcfg_10_02_08 = (nv_ro08(bios, data + 0x02) & 0x08) >> 3;
+		p->ramcfg_10_02_10 = (nv_ro08(bios, data + 0x02) & 0x10) >> 4;
+		p->ramcfg_10_02_20 = (nv_ro08(bios, data + 0x02) & 0x20) >> 5;
+		p->ramcfg_10_02_40 = (nv_ro08(bios, data + 0x02) & 0x40) >> 6;
+		p->ramcfg_10_03_0f = (nv_ro08(bios, data + 0x03) & 0x0f) >> 0;
+		p->ramcfg_10_05    = (nv_ro08(bios, data + 0x05) & 0xff) >> 0;
+		p->ramcfg_10_06    = (nv_ro08(bios, data + 0x06) & 0xff) >> 0;
+		p->ramcfg_10_07    = (nv_ro08(bios, data + 0x07) & 0xff) >> 0;
+		p->ramcfg_10_08    = (nv_ro08(bios, data + 0x08) & 0xff) >> 0;
+		p->ramcfg_10_09_0f = (nv_ro08(bios, data + 0x09) & 0x0f) >> 0;
+		p->ramcfg_10_09_f0 = (nv_ro08(bios, data + 0x09) & 0xf0) >> 4;
+		break;
 	case 0x11:
+		p->ramcfg_timing   =  nv_ro08(bios, data + 0x00);
 		p->ramcfg_11_01_01 = (nv_ro08(bios, data + 0x01) & 0x01) >> 0;
 		p->ramcfg_11_01_02 = (nv_ro08(bios, data + 0x01) & 0x02) >> 1;
 		p->ramcfg_11_01_04 = (nv_ro08(bios, data + 0x01) & 0x04) >> 2;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/timing.c b/drivers/gpu/drm/nouveau/core/subdev/bios/timing.c
index 350d44ab2ba2..46d955eb51eb 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/bios/timing.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/timing.c
@@ -89,7 +89,15 @@ nvbios_timingEp(struct nouveau_bios *bios, int idx,
 		struct nvbios_ramcfg *p)
 {
 	u16 data = nvbios_timingEe(bios, idx, ver, hdr, cnt, len), temp;
+	p->timing_ver = *ver;
+	p->timing_hdr = *hdr;
 	switch (!!data * *ver) {
+	case 0x10:
+		p->timing_10_WR = nv_ro08(bios, data + 0x00);
+		p->timing_10_CL = nv_ro08(bios, data + 0x02);
+		p->timing_10_ODT = nv_ro08(bios, data + 0x0e) & 0x07;
+		p->timing_10_CWL = nv_ro08(bios, data + 0x13);
+		break;
 	case 0x20:
 		p->timing[0] = nv_ro32(bios, data + 0x00);
 		p->timing[1] = nv_ro32(bios, data + 0x04);
diff --git a/drivers/gpu/drm/nouveau/core/subdev/clock/base.c b/drivers/gpu/drm/nouveau/core/subdev/clock/base.c
index a276a711294a..e51b72d47129 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/clock/base.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/clock/base.c
@@ -573,7 +573,7 @@ nouveau_clock_create_(struct nouveau_object *parent,
 
 	clk->allow_reclock = allow_reclock;
 
-	ret = nvkm_notify_init(&device->event, nouveau_clock_pwrsrc, true,
+	ret = nvkm_notify_init(NULL, &device->event, nouveau_clock_pwrsrc, true,
 			       NULL, 0, 0, &clk->pwrsrc_ntfy);
 	if (ret)
 		return ret;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/clock/nva3.c b/drivers/gpu/drm/nouveau/core/subdev/clock/nva3.c
index 087012b18956..094551d8ad9b 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/clock/nva3.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/clock/nva3.c
@@ -20,8 +20,10 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  *
  * Authors: Ben Skeggs
+ *          Roy Spliet
  */
 
+#include <engine/fifo.h>
 #include <subdev/bios.h>
 #include <subdev/bios/pll.h>
 #include <subdev/timer.h>
@@ -42,9 +44,17 @@ static u32
 read_vco(struct nva3_clock_priv *priv, int clk)
 {
 	u32 sctl = nv_rd32(priv, 0x4120 + (clk * 4));
-	if ((sctl & 0x00000030) != 0x00000030)
+
+	switch (sctl & 0x00000030) {
+	case 0x00000000:
+		return nv_device(priv)->crystal;
+	case 0x00000020:
 		return read_pll(priv, 0x41, 0x00e820);
-	return read_pll(priv, 0x42, 0x00e8a0);
+	case 0x00000030:
+		return read_pll(priv, 0x42, 0x00e8a0);
+	default:
+		return 0;
+	}
 }
 
 static u32
@@ -66,14 +76,25 @@ read_clk(struct nva3_clock_priv *priv, int clk, bool ignore_en)
 	if (!ignore_en && !(sctl & 0x00000100))
 		return 0;
 
+	/* out_alt */
+	if (sctl & 0x00000400)
+		return 108000;
+
+	/* vco_out */
 	switch (sctl & 0x00003000) {
 	case 0x00000000:
-		return nv_device(priv)->crystal;
+		if (!(sctl & 0x00000200))
+			return nv_device(priv)->crystal;
+		return 0;
 	case 0x00002000:
 		if (sctl & 0x00000040)
 			return 108000;
 		return 100000;
 	case 0x00003000:
+		/* vco_enable */
+		if (!(sctl & 0x00000001))
+			return 0;
+
 		sclk = read_vco(priv, clk);
 		sdiv = ((sctl & 0x003f0000) >> 16) + 2;
 		return (sclk * 2) / sdiv;
@@ -95,7 +116,9 @@ read_pll(struct nva3_clock_priv *priv, int clk, u32 pll)
 			N = (coef & 0x0000ff00) >> 8;
 			P = (coef & 0x003f0000) >> 16;
 
-			/* no post-divider on these.. */
+			/* no post-divider on these..
+			 * XXX: it looks more like two post-"dividers" that
+			 * cross each other out in the default RPLL config */
 			if ((pll & 0x00ff00) == 0x00e800)
 				P = 1;
 
@@ -114,13 +137,13 @@ static int
 nva3_clock_read(struct nouveau_clock *clk, enum nv_clk_src src)
 {
 	struct nva3_clock_priv *priv = (void *)clk;
+	u32 hsrc;
 
 	switch (src) {
 	case nv_clk_src_crystal:
 		return nv_device(priv)->crystal;
-	case nv_clk_src_href:
-		return 100000;
 	case nv_clk_src_core:
+	case nv_clk_src_core_intm:
 		return read_pll(priv, 0x00, 0x4200);
 	case nv_clk_src_shader:
 		return read_pll(priv, 0x01, 0x4220);
@@ -132,24 +155,33 @@ nva3_clock_read(struct nouveau_clock *clk, enum nv_clk_src src)
 		return read_clk(priv, 0x21, false);
 	case nv_clk_src_daemon:
 		return read_clk(priv, 0x25, false);
+	case nv_clk_src_host:
+		hsrc = (nv_rd32(priv, 0xc040) & 0x30000000) >> 28;
+		switch (hsrc) {
+		case 0:
+			return read_clk(priv, 0x1d, false);
+		case 2:
+		case 3:
+			return 277000;
+		default:
+			nv_error(clk, "unknown HOST clock source %d\n", hsrc);
+			return -EINVAL;
+		}
 	default:
 		nv_error(clk, "invalid clock source %d\n", src);
 		return -EINVAL;
 	}
+
+	return 0;
 }
 
 int
-nva3_clock_info(struct nouveau_clock *clock, int clk, u32 pll, u32 khz,
+nva3_clk_info(struct nouveau_clock *clock, int clk, u32 khz,
 		struct nva3_clock_info *info)
 {
-	struct nouveau_bios *bios = nouveau_bios(clock);
 	struct nva3_clock_priv *priv = (void *)clock;
-	struct nvbios_pll limits;
-	u32 oclk, sclk, sdiv;
-	int P, N, M, diff;
-	int ret;
+	u32 oclk, sclk, sdiv, diff;
 
-	info->pll = 0;
 	info->clk = 0;
 
 	switch (khz) {
@@ -164,43 +196,69 @@ nva3_clock_info(struct nouveau_clock *clock, int clk, u32 pll, u32 khz,
 		return khz;
 	default:
 		sclk = read_vco(priv, clk);
-		sdiv = min((sclk * 2) / (khz - 2999), (u32)65);
-		/* if the clock has a PLL attached, and we can get a within
-		 * [-2, 3) MHz of a divider, we'll disable the PLL and use
-		 * the divider instead.
-		 *
-		 * divider can go as low as 2, limited here because NVIDIA
+		sdiv = min((sclk * 2) / khz, (u32)65);
+		oclk = (sclk * 2) / sdiv;
+		diff = ((khz + 3000) - oclk);
+
+		/* When imprecise, play it safe and aim for a clock lower than
+		 * desired rather than higher */
+		if (diff < 0) {
+			sdiv++;
+			oclk = (sclk * 2) / sdiv;
+		}
+
+		/* divider can go as low as 2, limited here because NVIDIA
 		 * and the VBIOS on my NVA8 seem to prefer using the PLL
 		 * for 810MHz - is there a good reason?
-		 */
+		 * XXX: PLLs with refclk 810MHz?  */
 		if (sdiv > 4) {
-			oclk = (sclk * 2) / sdiv;
-			diff = khz - oclk;
-			if (!pll || (diff >= -2000 && diff < 3000)) {
-				info->clk = (((sdiv - 2) << 16) | 0x00003100);
-				return oclk;
-			}
+			info->clk = (((sdiv - 2) << 16) | 0x00003100);
+			return oclk;
 		}
 
-		if (!pll)
-			return -ERANGE;
 		break;
 	}
 
+	return -ERANGE;
+}
+
+int
+nva3_pll_info(struct nouveau_clock *clock, int clk, u32 pll, u32 khz,
+		struct nva3_clock_info *info)
+{
+	struct nouveau_bios *bios = nouveau_bios(clock);
+	struct nva3_clock_priv *priv = (void *)clock;
+	struct nvbios_pll limits;
+	int P, N, M, diff;
+	int ret;
+
+	info->pll = 0;
+
+	/* If we can get a within [-2, 3) MHz of a divider, we'll disable the
+	 * PLL and use the divider instead. */
+	ret = nva3_clk_info(clock, clk, khz, info);
+	diff = khz - ret;
+	if (!pll || (diff >= -2000 && diff < 3000)) {
+		goto out;
+	}
+
+	/* Try with PLL */
 	ret = nvbios_pll_parse(bios, pll, &limits);
 	if (ret)
 		return ret;
 
-	limits.refclk = read_clk(priv, clk - 0x10, true);
-	if (!limits.refclk)
+	ret = nva3_clk_info(clock, clk - 0x10, limits.refclk, info);
+	if (ret != limits.refclk)
 		return -EINVAL;
 
 	ret = nva3_pll_calc(nv_subdev(priv), &limits, khz, &N, NULL, &M, &P);
 	if (ret >= 0) {
-		info->clk = nv_rd32(priv, 0x4120 + (clk * 4));
 		info->pll = (P << 16) | (N << 8) | M;
 	}
 
+out:
+	info->fb_delay = max(((khz + 7566) / 15133), (u32) 18);
+
 	return ret ? ret : -ERANGE;
 }
 
@@ -208,13 +266,76 @@ static int
 calc_clk(struct nva3_clock_priv *priv, struct nouveau_cstate *cstate,
 	 int clk, u32 pll, int idx)
 {
-	int ret = nva3_clock_info(&priv->base, clk, pll, cstate->domain[idx],
+	int ret = nva3_pll_info(&priv->base, clk, pll, cstate->domain[idx],
 				  &priv->eng[idx]);
 	if (ret >= 0)
 		return 0;
 	return ret;
 }
 
+static int
+calc_host(struct nva3_clock_priv *priv, struct nouveau_cstate *cstate)
+{
+	int ret = 0;
+	u32 kHz = cstate->domain[nv_clk_src_host];
+	struct nva3_clock_info *info = &priv->eng[nv_clk_src_host];
+
+	if (kHz == 277000) {
+		info->clk = 0;
+		info->host_out = NVA3_HOST_277;
+		return 0;
+	}
+
+	info->host_out = NVA3_HOST_CLK;
+
+	ret = nva3_clk_info(&priv->base, 0x1d, kHz, info);
+	if (ret >= 0)
+		return 0;
+	return ret;
+}
+
+int
+nva3_clock_pre(struct nouveau_clock *clk, unsigned long *flags)
+{
+	struct nouveau_fifo *pfifo = nouveau_fifo(clk);
+
+	/* halt and idle execution engines */
+	nv_mask(clk, 0x020060, 0x00070000, 0x00000000);
+	nv_mask(clk, 0x002504, 0x00000001, 0x00000001);
+	/* Wait until the interrupt handler is finished */
+	if (!nv_wait(clk, 0x000100, 0xffffffff, 0x00000000))
+		return -EBUSY;
+
+	if (pfifo)
+		pfifo->pause(pfifo, flags);
+
+	if (!nv_wait(clk, 0x002504, 0x00000010, 0x00000010))
+		return -EIO;
+	if (!nv_wait(clk, 0x00251c, 0x0000003f, 0x0000003f))
+		return -EIO;
+
+	return 0;
+}
+
+void
+nva3_clock_post(struct nouveau_clock *clk, unsigned long *flags)
+{
+	struct nouveau_fifo *pfifo = nouveau_fifo(clk);
+
+	if (pfifo && flags)
+		pfifo->start(pfifo, flags);
+
+	nv_mask(clk, 0x002504, 0x00000001, 0x00000000);
+	nv_mask(clk, 0x020060, 0x00070000, 0x00040000);
+}
+
+static void
+disable_clk_src(struct nva3_clock_priv *priv, u32 src)
+{
+	nv_mask(priv, src, 0x00000100, 0x00000000);
+	nv_mask(priv, src, 0x00000001, 0x00000000);
+}
+
 static void
 prog_pll(struct nva3_clock_priv *priv, int clk, u32 pll, int idx)
 {
@@ -223,24 +344,35 @@ prog_pll(struct nva3_clock_priv *priv, int clk, u32 pll, int idx)
 	const u32 src1 = 0x004160 + (clk * 4);
 	const u32 ctrl = pll + 0;
 	const u32 coef = pll + 4;
+	u32 bypass;
 
 	if (info->pll) {
-		nv_mask(priv, src0, 0x00000101, 0x00000101);
+		/* Always start from a non-PLL clock */
+		bypass = nv_rd32(priv, ctrl)  & 0x00000008;
+		if (!bypass) {
+			nv_mask(priv, src1, 0x00000101, 0x00000101);
+			nv_mask(priv, ctrl, 0x00000008, 0x00000008);
+			udelay(20);
+		}
+
+		nv_mask(priv, src0, 0x003f3141, 0x00000101 | info->clk);
 		nv_wr32(priv, coef, info->pll);
 		nv_mask(priv, ctrl, 0x00000015, 0x00000015);
 		nv_mask(priv, ctrl, 0x00000010, 0x00000000);
-		nv_wait(priv, ctrl, 0x00020000, 0x00020000);
+		if (!nv_wait(priv, ctrl, 0x00020000, 0x00020000)) {
+			nv_mask(priv, ctrl, 0x00000010, 0x00000010);
+			nv_mask(priv, src0, 0x00000101, 0x00000000);
+			return;
+		}
 		nv_mask(priv, ctrl, 0x00000010, 0x00000010);
 		nv_mask(priv, ctrl, 0x00000008, 0x00000000);
-		nv_mask(priv, src1, 0x00000100, 0x00000000);
-		nv_mask(priv, src1, 0x00000001, 0x00000000);
+		disable_clk_src(priv, src1);
 	} else {
 		nv_mask(priv, src1, 0x003f3141, 0x00000101 | info->clk);
 		nv_mask(priv, ctrl, 0x00000018, 0x00000018);
 		udelay(20);
 		nv_mask(priv, ctrl, 0x00000001, 0x00000000);
-		nv_mask(priv, src0, 0x00000100, 0x00000000);
-		nv_mask(priv, src0, 0x00000001, 0x00000000);
+		disable_clk_src(priv, src0);
 	}
 }
 
@@ -251,18 +383,72 @@ prog_clk(struct nva3_clock_priv *priv, int clk, int idx)
 	nv_mask(priv, 0x004120 + (clk * 4), 0x003f3141, 0x00000101 | info->clk);
 }
 
+static void
+prog_host(struct nva3_clock_priv *priv)
+{
+	struct nva3_clock_info *info = &priv->eng[nv_clk_src_host];
+	u32 hsrc = (nv_rd32(priv, 0xc040));
+
+	switch (info->host_out) {
+	case NVA3_HOST_277:
+		if ((hsrc & 0x30000000) == 0) {
+			nv_wr32(priv, 0xc040, hsrc | 0x20000000);
+			disable_clk_src(priv, 0x4194);
+		}
+		break;
+	case NVA3_HOST_CLK:
+		prog_clk(priv, 0x1d, nv_clk_src_host);
+		if ((hsrc & 0x30000000) >= 0x20000000) {
+			nv_wr32(priv, 0xc040, hsrc & ~0x30000000);
+		}
+		break;
+	default:
+		break;
+	}
+
+	/* This seems to be a clock gating factor on idle, always set to 64 */
+	nv_wr32(priv, 0xc044, 0x3e);
+}
+
+static void
+prog_core(struct nva3_clock_priv *priv, int idx)
+{
+	struct nva3_clock_info *info = &priv->eng[idx];
+	u32 fb_delay = nv_rd32(priv, 0x10002c);
+
+	if (fb_delay < info->fb_delay)
+		nv_wr32(priv, 0x10002c, info->fb_delay);
+
+	prog_pll(priv, 0x00, 0x004200, idx);
+
+	if (fb_delay > info->fb_delay)
+		nv_wr32(priv, 0x10002c, info->fb_delay);
+}
+
 static int
 nva3_clock_calc(struct nouveau_clock *clk, struct nouveau_cstate *cstate)
 {
 	struct nva3_clock_priv *priv = (void *)clk;
+	struct nva3_clock_info *core = &priv->eng[nv_clk_src_core];
 	int ret;
 
 	if ((ret = calc_clk(priv, cstate, 0x10, 0x4200, nv_clk_src_core)) ||
 	    (ret = calc_clk(priv, cstate, 0x11, 0x4220, nv_clk_src_shader)) ||
 	    (ret = calc_clk(priv, cstate, 0x20, 0x0000, nv_clk_src_disp)) ||
-	    (ret = calc_clk(priv, cstate, 0x21, 0x0000, nv_clk_src_vdec)))
+	    (ret = calc_clk(priv, cstate, 0x21, 0x0000, nv_clk_src_vdec)) ||
+	    (ret = calc_host(priv, cstate)))
 		return ret;
 
+	/* XXX: Should be reading the highest bit in the VBIOS clock to decide
+	 * whether to use a PLL or not... but using a PLL defeats the purpose */
+	if (core->pll) {
+		ret = nva3_clk_info(clk, 0x10,
+				cstate->domain[nv_clk_src_core_intm],
+				&priv->eng[nv_clk_src_core_intm]);
+		if (ret < 0)
+			return ret;
+	}
+
 	return 0;
 }
 
@@ -270,11 +456,31 @@ static int
 nva3_clock_prog(struct nouveau_clock *clk)
 {
 	struct nva3_clock_priv *priv = (void *)clk;
-	prog_pll(priv, 0x00, 0x004200, nv_clk_src_core);
+	struct nva3_clock_info *core = &priv->eng[nv_clk_src_core];
+	int ret = 0;
+	unsigned long flags;
+	unsigned long *f = &flags;
+
+	ret = nva3_clock_pre(clk, f);
+	if (ret)
+		goto out;
+
+	if (core->pll)
+		prog_core(priv, nv_clk_src_core_intm);
+
+	prog_core(priv,  nv_clk_src_core);
 	prog_pll(priv, 0x01, 0x004220, nv_clk_src_shader);
 	prog_clk(priv, 0x20, nv_clk_src_disp);
 	prog_clk(priv, 0x21, nv_clk_src_vdec);
-	return 0;
+	prog_host(priv);
+
+out:
+	if (ret == -EBUSY)
+		f = NULL;
+
+	nva3_clock_post(clk, f);
+
+	return ret;
 }
 
 static void
@@ -284,13 +490,14 @@ nva3_clock_tidy(struct nouveau_clock *clk)
 
 static struct nouveau_clocks
 nva3_domain[] = {
-	{ nv_clk_src_crystal, 0xff },
-	{ nv_clk_src_href   , 0xff },
-	{ nv_clk_src_core   , 0x00, 0, "core", 1000 },
-	{ nv_clk_src_shader , 0x01, 0, "shader", 1000 },
-	{ nv_clk_src_mem    , 0x02, 0, "memory", 1000 },
-	{ nv_clk_src_vdec   , 0x03 },
-	{ nv_clk_src_disp   , 0x04 },
+	{ nv_clk_src_crystal  , 0xff },
+	{ nv_clk_src_core     , 0x00, 0, "core", 1000 },
+	{ nv_clk_src_shader   , 0x01, 0, "shader", 1000 },
+	{ nv_clk_src_mem      , 0x02, 0, "memory", 1000 },
+	{ nv_clk_src_vdec     , 0x03 },
+	{ nv_clk_src_disp     , 0x04 },
+	{ nv_clk_src_host     , 0x05 },
+	{ nv_clk_src_core_intm, 0x06 },
 	{ nv_clk_src_max }
 };
 
diff --git a/drivers/gpu/drm/nouveau/core/subdev/clock/nva3.h b/drivers/gpu/drm/nouveau/core/subdev/clock/nva3.h
index 6229a509b42e..a45a1038b12f 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/clock/nva3.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/clock/nva3.h
@@ -6,9 +6,15 @@
 struct nva3_clock_info {
 	u32 clk;
 	u32 pll;
+	enum {
+		NVA3_HOST_277,
+		NVA3_HOST_CLK,
+	} host_out;
+	u32 fb_delay;
 };
 
-int nva3_clock_info(struct nouveau_clock *, int, u32, u32,
+int nva3_pll_info(struct nouveau_clock *, int, u32, u32,
 		    struct nva3_clock_info *);
-
+int nva3_clock_pre(struct nouveau_clock *clk, unsigned long *flags);
+void nva3_clock_post(struct nouveau_clock *clk, unsigned long *flags);
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/clock/nvaa.c b/drivers/gpu/drm/nouveau/core/subdev/clock/nvaa.c
index 74e19731b1b7..54aeab8005a0 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/clock/nvaa.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/clock/nvaa.c
@@ -28,6 +28,7 @@
 #include <subdev/timer.h>
 #include <subdev/clock.h>
 
+#include "nva3.h"
 #include "pll.h"
 
 struct nvaa_clock_priv {
@@ -299,25 +300,14 @@ static int
 nvaa_clock_prog(struct nouveau_clock *clk)
 {
 	struct nvaa_clock_priv *priv = (void *)clk;
-	struct nouveau_fifo *pfifo = nouveau_fifo(clk);
+	u32 pllmask = 0, mast;
 	unsigned long flags;
-	u32 pllmask = 0, mast, ptherm_gate;
-	int ret = -EBUSY;
-
-	/* halt and idle execution engines */
-	ptherm_gate = nv_mask(clk, 0x020060, 0x00070000, 0x00000000);
-	nv_mask(clk, 0x002504, 0x00000001, 0x00000001);
-	/* Wait until the interrupt handler is finished */
-	if (!nv_wait(clk, 0x000100, 0xffffffff, 0x00000000))
-		goto resume;
-
-	if (pfifo)
-		pfifo->pause(pfifo, &flags);
+	unsigned long *f = &flags;
+	int ret = 0;
 
-	if (!nv_wait(clk, 0x002504, 0x00000010, 0x00000010))
-		goto resume;
-	if (!nv_wait(clk, 0x00251c, 0x0000003f, 0x0000003f))
-		goto resume;
+	ret = nva3_clock_pre(clk, f);
+	if (ret)
+		goto out;
 
 	/* First switch to safe clocks: href */
 	mast = nv_mask(clk, 0xc054, 0x03400e70, 0x03400640);
@@ -375,15 +365,8 @@ nvaa_clock_prog(struct nouveau_clock *clk)
 	}
 
 	nv_wr32(clk, 0xc054, mast);
-	ret = 0;
 
 resume:
-	if (pfifo)
-		pfifo->start(pfifo, &flags);
-
-	nv_mask(clk, 0x002504, 0x00000001, 0x00000000);
-	nv_wr32(clk, 0x020060, ptherm_gate);
-
 	/* Disable some PLLs and dividers when unused */
 	if (priv->csrc != nv_clk_src_core) {
 		nv_wr32(clk, 0x4040, 0x00000000);
@@ -395,6 +378,12 @@ resume:
 		nv_mask(clk, 0x4020, 0x80000000, 0x00000000);
 	}
 
+out:
+	if (ret == -EBUSY)
+		f = NULL;
+
+	nva3_clock_post(clk, f);
+
 	return ret;
 }
 
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/fbmem.h b/drivers/gpu/drm/nouveau/core/subdev/devinit/fbmem.h
index 4fe49cf4c99a..6103484fea72 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/fbmem.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/fbmem.h
@@ -26,22 +26,8 @@
 
 #include <core/device.h>
 
-#define NV04_PFB_BOOT_0						0x00100000
-#	define NV04_PFB_BOOT_0_RAM_AMOUNT			0x00000003
-#	define NV04_PFB_BOOT_0_RAM_AMOUNT_32MB			0x00000000
-#	define NV04_PFB_BOOT_0_RAM_AMOUNT_4MB			0x00000001
-#	define NV04_PFB_BOOT_0_RAM_AMOUNT_8MB			0x00000002
-#	define NV04_PFB_BOOT_0_RAM_AMOUNT_16MB			0x00000003
-#	define NV04_PFB_BOOT_0_RAM_WIDTH_128			0x00000004
-#	define NV04_PFB_BOOT_0_RAM_TYPE				0x00000028
-#	define NV04_PFB_BOOT_0_RAM_TYPE_SGRAM_8MBIT		0x00000000
-#	define NV04_PFB_BOOT_0_RAM_TYPE_SGRAM_16MBIT		0x00000008
-#	define NV04_PFB_BOOT_0_RAM_TYPE_SGRAM_16MBIT_4BANK	0x00000010
-#	define NV04_PFB_BOOT_0_RAM_TYPE_SDRAM_16MBIT		0x00000018
-#	define NV04_PFB_BOOT_0_RAM_TYPE_SDRAM_64MBIT		0x00000020
-#	define NV04_PFB_BOOT_0_RAM_TYPE_SDRAM_64MBITX16		0x00000028
-#	define NV04_PFB_BOOT_0_UMA_ENABLE			0x00000100
-#	define NV04_PFB_BOOT_0_UMA_SIZE				0x0000f000
+#include <subdev/fb/regsnv04.h>
+
 #define NV04_PFB_DEBUG_0					0x00100080
 #	define NV04_PFB_DEBUG_0_PAGE_MODE			0x00000001
 #	define NV04_PFB_DEBUG_0_REFRESH_OFF			0x00000010
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/gddr5.c b/drivers/gpu/drm/nouveau/core/subdev/fb/gddr5.c
index 66fe959b4f74..7fbbe05d5c60 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/gddr5.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/gddr5.c
@@ -40,7 +40,7 @@ nouveau_gddr5_calc(struct nouveau_ram *ram, bool nuts)
 	int WL, CL, WR, at[2], dt, ds;
 	int rq = ram->freq < 1000000; /* XXX */
 
-	switch (ram->ramcfg.version) {
+	switch (ram->next->bios.ramcfg_ver) {
 	case 0x11:
 		pd =  ram->next->bios.ramcfg_11_01_80;
 		lf =  ram->next->bios.ramcfg_11_01_40;
@@ -54,7 +54,7 @@ nouveau_gddr5_calc(struct nouveau_ram *ram, bool nuts)
 		return -ENOSYS;
 	}
 
-	switch (ram->timing.version) {
+	switch (ram->next->bios.timing_ver) {
 	case 0x20:
 		WL = (ram->next->bios.timing[1] & 0x00000f80) >> 7;
 		CL = (ram->next->bios.timing[1] & 0x0000001f);
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/nv20.c b/drivers/gpu/drm/nouveau/core/subdev/fb/nv20.c
index f003c1b1893f..2209ade63339 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/nv20.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/nv20.c
@@ -45,7 +45,7 @@ nv20_fb_tile_comp(struct nouveau_fb *pfb, int i, u32 size, u32 flags,
 {
 	u32 tiles = DIV_ROUND_UP(size, 0x40);
 	u32 tags  = round_up(tiles / pfb->ram->parts, 0x40);
-	if (!nouveau_mm_head(&pfb->tags, 1, tags, tags, 1, &tile->tag)) {
+	if (!nouveau_mm_head(&pfb->tags, 0, 1, tags, tags, 1, &tile->tag)) {
 		if (!(flags & 2)) tile->zcomp = 0x00000000; /* Z16 */
 		else              tile->zcomp = 0x04000000; /* Z24S8 */
 		tile->zcomp |= tile->tag->offset;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/nv25.c b/drivers/gpu/drm/nouveau/core/subdev/fb/nv25.c
index f34f4223210b..e2a66c355c50 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/nv25.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/nv25.c
@@ -32,7 +32,7 @@ nv25_fb_tile_comp(struct nouveau_fb *pfb, int i, u32 size, u32 flags,
 {
 	u32 tiles = DIV_ROUND_UP(size, 0x40);
 	u32 tags  = round_up(tiles / pfb->ram->parts, 0x40);
-	if (!nouveau_mm_head(&pfb->tags, 1, tags, tags, 1, &tile->tag)) {
+	if (!nouveau_mm_head(&pfb->tags, 0, 1, tags, tags, 1, &tile->tag)) {
 		if (!(flags & 2)) tile->zcomp = 0x00100000; /* Z16 */
 		else              tile->zcomp = 0x00200000; /* Z24S8 */
 		tile->zcomp |= tile->tag->offset;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/nv30.c b/drivers/gpu/drm/nouveau/core/subdev/fb/nv30.c
index 69093f7151f0..cbec402ba5b9 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/nv30.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/nv30.c
@@ -51,7 +51,7 @@ nv30_fb_tile_comp(struct nouveau_fb *pfb, int i, u32 size, u32 flags,
 {
 	u32 tiles = DIV_ROUND_UP(size, 0x40);
 	u32 tags  = round_up(tiles / pfb->ram->parts, 0x40);
-	if (!nouveau_mm_head(&pfb->tags, 1, tags, tags, 1, &tile->tag)) {
+	if (!nouveau_mm_head(&pfb->tags, 0, 1, tags, tags, 1, &tile->tag)) {
 		if (flags & 2) tile->zcomp |= 0x01000000; /* Z16 */
 		else           tile->zcomp |= 0x02000000; /* Z24S8 */
 		tile->zcomp |= ((tile->tag->offset           ) >> 6);
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/nv35.c b/drivers/gpu/drm/nouveau/core/subdev/fb/nv35.c
index 161b06e8fc3f..b2cf8c69fb2e 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/nv35.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/nv35.c
@@ -32,7 +32,7 @@ nv35_fb_tile_comp(struct nouveau_fb *pfb, int i, u32 size, u32 flags,
 {
 	u32 tiles = DIV_ROUND_UP(size, 0x40);
 	u32 tags  = round_up(tiles / pfb->ram->parts, 0x40);
-	if (!nouveau_mm_head(&pfb->tags, 1, tags, tags, 1, &tile->tag)) {
+	if (!nouveau_mm_head(&pfb->tags, 0, 1, tags, tags, 1, &tile->tag)) {
 		if (flags & 2) tile->zcomp |= 0x04000000; /* Z16 */
 		else           tile->zcomp |= 0x08000000; /* Z24S8 */
 		tile->zcomp |= ((tile->tag->offset           ) >> 6);
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/nv36.c b/drivers/gpu/drm/nouveau/core/subdev/fb/nv36.c
index 2dd3d0aab6bb..b4cdae2a3b2f 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/nv36.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/nv36.c
@@ -32,7 +32,7 @@ nv36_fb_tile_comp(struct nouveau_fb *pfb, int i, u32 size, u32 flags,
 {
 	u32 tiles = DIV_ROUND_UP(size, 0x40);
 	u32 tags  = round_up(tiles / pfb->ram->parts, 0x40);
-	if (!nouveau_mm_head(&pfb->tags, 1, tags, tags, 1, &tile->tag)) {
+	if (!nouveau_mm_head(&pfb->tags, 0, 1, tags, tags, 1, &tile->tag)) {
 		if (flags & 2) tile->zcomp |= 0x10000000; /* Z16 */
 		else           tile->zcomp |= 0x20000000; /* Z24S8 */
 		tile->zcomp |= ((tile->tag->offset           ) >> 6);
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/nv40.c b/drivers/gpu/drm/nouveau/core/subdev/fb/nv40.c
index 95a115ab0c86..52814258c212 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/nv40.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/nv40.c
@@ -33,7 +33,7 @@ nv40_fb_tile_comp(struct nouveau_fb *pfb, int i, u32 size, u32 flags,
 	u32 tiles = DIV_ROUND_UP(size, 0x80);
 	u32 tags  = round_up(tiles / pfb->ram->parts, 0x100);
 	if ( (flags & 2) &&
-	    !nouveau_mm_head(&pfb->tags, 1, tags, tags, 1, &tile->tag)) {
+	    !nouveau_mm_head(&pfb->tags, 0, 1, tags, tags, 1, &tile->tag)) {
 		tile->zcomp  = 0x28000000; /* Z24S8_SPLIT_GRAD */
 		tile->zcomp |= ((tile->tag->offset           ) >> 8);
 		tile->zcomp |= ((tile->tag->offset + tags - 1) >> 8) << 13;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/priv.h b/drivers/gpu/drm/nouveau/core/subdev/fb/priv.h
index 82273f832e42..60322e906dd4 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/priv.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/priv.h
@@ -35,6 +35,7 @@ extern struct nouveau_oclass nve0_ram_oclass;
 extern struct nouveau_oclass gk20a_ram_oclass;
 extern struct nouveau_oclass gm107_ram_oclass;
 
+int nouveau_sddr2_calc(struct nouveau_ram *ram);
 int nouveau_sddr3_calc(struct nouveau_ram *ram);
 int nouveau_gddr5_calc(struct nouveau_ram *ram, bool nuts);
 
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/ramfuc.h b/drivers/gpu/drm/nouveau/core/subdev/fb/ramfuc.h
index 2af9cfd2c60f..d1fbbe4b00a2 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/ramfuc.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/ramfuc.h
@@ -12,16 +12,32 @@ struct ramfuc {
 struct ramfuc_reg {
 	int sequence;
 	bool force;
-	u32 addr[2];
+	u32 addr;
+	u32 stride; /* in bytes */
+	u32 mask;
 	u32 data;
 };
 
 static inline struct ramfuc_reg
+ramfuc_stride(u32 addr, u32 stride, u32 mask)
+{
+	return (struct ramfuc_reg) {
+		.sequence = 0,
+		.addr = addr,
+		.stride = stride,
+		.mask = mask,
+		.data = 0xdeadbeef,
+	};
+}
+
+static inline struct ramfuc_reg
 ramfuc_reg2(u32 addr1, u32 addr2)
 {
 	return (struct ramfuc_reg) {
 		.sequence = 0,
-		.addr = { addr1, addr2 },
+		.addr = addr1,
+		.stride = addr2 - addr1,
+		.mask = 0x3,
 		.data = 0xdeadbeef,
 	};
 }
@@ -29,7 +45,13 @@ ramfuc_reg2(u32 addr1, u32 addr2)
 static noinline struct ramfuc_reg
 ramfuc_reg(u32 addr)
 {
-	return ramfuc_reg2(addr, addr);
+	return (struct ramfuc_reg) {
+		.sequence = 0,
+		.addr = addr,
+		.stride = 0,
+		.mask = 0x1,
+		.data = 0xdeadbeef,
+	};
 }
 
 static inline int
@@ -62,18 +84,25 @@ static inline u32
 ramfuc_rd32(struct ramfuc *ram, struct ramfuc_reg *reg)
 {
 	if (reg->sequence != ram->sequence)
-		reg->data = nv_rd32(ram->pfb, reg->addr[0]);
+		reg->data = nv_rd32(ram->pfb, reg->addr);
 	return reg->data;
 }
 
 static inline void
 ramfuc_wr32(struct ramfuc *ram, struct ramfuc_reg *reg, u32 data)
 {
+	unsigned int mask, off = 0;
+
 	reg->sequence = ram->sequence;
 	reg->data = data;
-	if (reg->addr[0] != reg->addr[1])
-		nouveau_memx_wr32(ram->memx, reg->addr[1], reg->data);
-	nouveau_memx_wr32(ram->memx, reg->addr[0], reg->data);
+
+	for (mask = reg->mask; mask > 0; mask = (mask & ~1) >> 1) {
+		if (mask & 1) {
+			nouveau_memx_wr32(ram->memx, reg->addr+off, reg->data);
+		}
+
+		off += reg->stride;
+	}
 }
 
 static inline void
@@ -105,14 +134,35 @@ ramfuc_nsec(struct ramfuc *ram, u32 nsec)
 	nouveau_memx_nsec(ram->memx, nsec);
 }
 
-#define ram_init(s,p)       ramfuc_init(&(s)->base, (p))
-#define ram_exec(s,e)       ramfuc_exec(&(s)->base, (e))
-#define ram_have(s,r)       ((s)->r_##r.addr[0] != 0x000000)
-#define ram_rd32(s,r)       ramfuc_rd32(&(s)->base, &(s)->r_##r)
-#define ram_wr32(s,r,d)     ramfuc_wr32(&(s)->base, &(s)->r_##r, (d))
-#define ram_nuke(s,r)       ramfuc_nuke(&(s)->base, &(s)->r_##r)
-#define ram_mask(s,r,m,d)   ramfuc_mask(&(s)->base, &(s)->r_##r, (m), (d))
-#define ram_wait(s,r,m,d,n) ramfuc_wait(&(s)->base, (r), (m), (d), (n))
-#define ram_nsec(s,n)       ramfuc_nsec(&(s)->base, (n))
+static inline void
+ramfuc_wait_vblank(struct ramfuc *ram)
+{
+	nouveau_memx_wait_vblank(ram->memx);
+}
+
+static inline void
+ramfuc_block(struct ramfuc *ram)
+{
+	nouveau_memx_block(ram->memx);
+}
+
+static inline void
+ramfuc_unblock(struct ramfuc *ram)
+{
+	nouveau_memx_unblock(ram->memx);
+}
+
+#define ram_init(s,p)        ramfuc_init(&(s)->base, (p))
+#define ram_exec(s,e)        ramfuc_exec(&(s)->base, (e))
+#define ram_have(s,r)        ((s)->r_##r.addr != 0x000000)
+#define ram_rd32(s,r)        ramfuc_rd32(&(s)->base, &(s)->r_##r)
+#define ram_wr32(s,r,d)      ramfuc_wr32(&(s)->base, &(s)->r_##r, (d))
+#define ram_nuke(s,r)        ramfuc_nuke(&(s)->base, &(s)->r_##r)
+#define ram_mask(s,r,m,d)    ramfuc_mask(&(s)->base, &(s)->r_##r, (m), (d))
+#define ram_wait(s,r,m,d,n)  ramfuc_wait(&(s)->base, (r), (m), (d), (n))
+#define ram_nsec(s,n)        ramfuc_nsec(&(s)->base, (n))
+#define ram_wait_vblank(s)   ramfuc_wait_vblank(&(s)->base)
+#define ram_block(s)         ramfuc_block(&(s)->base)
+#define ram_unblock(s)       ramfuc_unblock(&(s)->base)
 
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnv04.c b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnv04.c
index e781080d3327..1972268d1410 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnv04.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnv04.c
@@ -22,22 +22,7 @@
  * Authors: Ben Skeggs
  */
 
-#define NV04_PFB_BOOT_0						0x00100000
-#	define NV04_PFB_BOOT_0_RAM_AMOUNT			0x00000003
-#	define NV04_PFB_BOOT_0_RAM_AMOUNT_32MB			0x00000000
-#	define NV04_PFB_BOOT_0_RAM_AMOUNT_4MB			0x00000001
-#	define NV04_PFB_BOOT_0_RAM_AMOUNT_8MB			0x00000002
-#	define NV04_PFB_BOOT_0_RAM_AMOUNT_16MB			0x00000003
-#	define NV04_PFB_BOOT_0_RAM_WIDTH_128			0x00000004
-#	define NV04_PFB_BOOT_0_RAM_TYPE				0x00000028
-#	define NV04_PFB_BOOT_0_RAM_TYPE_SGRAM_8MBIT		0x00000000
-#	define NV04_PFB_BOOT_0_RAM_TYPE_SGRAM_16MBIT		0x00000008
-#	define NV04_PFB_BOOT_0_RAM_TYPE_SGRAM_16MBIT_4BANK	0x00000010
-#	define NV04_PFB_BOOT_0_RAM_TYPE_SDRAM_16MBIT		0x00000018
-#	define NV04_PFB_BOOT_0_RAM_TYPE_SDRAM_64MBIT		0x00000020
-#	define NV04_PFB_BOOT_0_RAM_TYPE_SDRAM_64MBITX16		0x00000028
-#	define NV04_PFB_BOOT_0_UMA_ENABLE			0x00000100
-#	define NV04_PFB_BOOT_0_UMA_SIZE				0x0000f000
+#include <subdev/fb/regsnv04.h>
 
 #include "priv.h"
 
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnv50.c b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnv50.c
index e5d12c24cc43..64a983c96625 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnv50.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnv50.c
@@ -280,7 +280,7 @@ nv50_ram_get(struct nouveau_fb *pfb, u64 size, u32 align, u32 ncmin,
 		if (align == 16) {
 			int n = (max >> 4) * comp;
 
-			ret = nouveau_mm_head(tags, 1, n, n, 1, &mem->tag);
+			ret = nouveau_mm_head(tags, 0, 1, n, n, 1, &mem->tag);
 			if (ret)
 				mem->tag = NULL;
 		}
@@ -296,9 +296,9 @@ nv50_ram_get(struct nouveau_fb *pfb, u64 size, u32 align, u32 ncmin,
 	type = nv50_fb_memtype[type];
 	do {
 		if (back)
-			ret = nouveau_mm_tail(heap, type, max, min, align, &r);
+			ret = nouveau_mm_tail(heap, 0, type, max, min, align, &r);
 		else
-			ret = nouveau_mm_head(heap, type, max, min, align, &r);
+			ret = nouveau_mm_head(heap, 0, type, max, min, align, &r);
 		if (ret) {
 			mutex_unlock(&pfb->base.mutex);
 			pfb->ram->put(pfb, &mem);
@@ -319,27 +319,22 @@ nv50_ram_get(struct nouveau_fb *pfb, u64 size, u32 align, u32 ncmin,
 static u32
 nv50_fb_vram_rblock(struct nouveau_fb *pfb, struct nouveau_ram *ram)
 {
-	int i, parts, colbits, rowbitsa, rowbitsb, banks;
+	int colbits, rowbitsa, rowbitsb, banks;
 	u64 rowsize, predicted;
-	u32 r0, r4, rt, ru, rblock_size;
+	u32 r0, r4, rt, rblock_size;
 
 	r0 = nv_rd32(pfb, 0x100200);
 	r4 = nv_rd32(pfb, 0x100204);
 	rt = nv_rd32(pfb, 0x100250);
-	ru = nv_rd32(pfb, 0x001540);
-	nv_debug(pfb, "memcfg 0x%08x 0x%08x 0x%08x 0x%08x\n", r0, r4, rt, ru);
-
-	for (i = 0, parts = 0; i < 8; i++) {
-		if (ru & (0x00010000 << i))
-			parts++;
-	}
+	nv_debug(pfb, "memcfg 0x%08x 0x%08x 0x%08x 0x%08x\n", r0, r4, rt,
+			nv_rd32(pfb, 0x001540));
 
 	colbits  =  (r4 & 0x0000f000) >> 12;
 	rowbitsa = ((r4 & 0x000f0000) >> 16) + 8;
 	rowbitsb = ((r4 & 0x00f00000) >> 20) + 8;
 	banks    = 1 << (((r4 & 0x03000000) >> 24) + 2);
 
-	rowsize = parts * banks * (1 << colbits) * 8;
+	rowsize = ram->parts * banks * (1 << colbits) * 8;
 	predicted = rowsize << rowbitsa;
 	if (r0 & 0x00000004)
 		predicted += rowsize << rowbitsb;
@@ -376,6 +371,9 @@ nv50_ram_create_(struct nouveau_object *parent, struct nouveau_object *engine,
 	ram->size = nv_rd32(pfb, 0x10020c);
 	ram->size = (ram->size & 0xffffff00) | ((ram->size & 0x000000ff) << 32);
 
+	ram->part_mask = (nv_rd32(pfb, 0x001540) & 0x00ff0000) >> 16;
+	ram->parts = hweight8(ram->part_mask);
+
 	switch (nv_rd32(pfb, 0x100714) & 0x00000007) {
 	case 0: ram->type = NV_MEM_TYPE_DDR1; break;
 	case 1:
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnva3.c b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnva3.c
index 8076fb195dd5..3601deca0bd5 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnva3.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnva3.c
@@ -79,20 +79,27 @@ nva3_ram_calc(struct nouveau_fb *pfb, u32 freq)
 	struct nva3_ram *ram = (void *)pfb->ram;
 	struct nva3_ramfuc *fuc = &ram->fuc;
 	struct nva3_clock_info mclk;
-	u8  ver, cnt, len, strap;
+	struct nouveau_ram_data *next;
+	u8  ver, hdr, cnt, len, strap;
 	u32 data;
-	struct {
-		u32 data;
-		u8  size;
-	} rammap, ramcfg, timing;
 	u32 r004018, r100760, ctrl;
 	u32 unk714, unk718, unk71c;
-	int ret;
+	int ret, i;
+
+	next = &ram->base.target;
+	next->freq = freq;
+	ram->base.next = next;
 
 	/* lookup memory config data relevant to the target frequency */
-	rammap.data = nvbios_rammapEm(bios, freq / 1000, &ver, &rammap.size,
-				     &cnt, &ramcfg.size);
-	if (!rammap.data || ver != 0x10 || rammap.size < 0x0e) {
+	i = 0;
+	while ((data = nvbios_rammapEp(bios, i++, &ver, &hdr, &cnt, &len,
+				      &next->bios))) {
+		if (freq / 1000 >= next->bios.rammap_min &&
+		    freq / 1000 <= next->bios.rammap_max)
+			break;
+	}
+
+	if (!data || ver != 0x10 || hdr < 0x0e) {
 		nv_error(pfb, "invalid/missing rammap entry\n");
 		return -EINVAL;
 	}
@@ -104,26 +111,25 @@ nva3_ram_calc(struct nouveau_fb *pfb, u32 freq)
 		return -EINVAL;
 	}
 
-	ramcfg.data = rammap.data + rammap.size + (strap * ramcfg.size);
-	if (!ramcfg.data || ver != 0x10 || ramcfg.size < 0x0e) {
+	data = nvbios_rammapSp(bios, data, ver, hdr, cnt, len, strap,
+			       &ver, &hdr, &next->bios);
+	if (!data || ver != 0x10 || hdr < 0x0e) {
 		nv_error(pfb, "invalid/missing ramcfg entry\n");
 		return -EINVAL;
 	}
 
 	/* lookup memory timings, if bios says they're present */
-	strap = nv_ro08(bios, ramcfg.data + 0x01);
-	if (strap != 0xff) {
-		timing.data = nvbios_timingEe(bios, strap, &ver, &timing.size,
-					     &cnt, &len);
-		if (!timing.data || ver != 0x10 || timing.size < 0x19) {
+	if (next->bios.ramcfg_timing != 0xff) {
+		data = nvbios_timingEp(bios, next->bios.ramcfg_timing,
+				       &ver, &hdr, &cnt, &len,
+				       &next->bios);
+		if (!data || ver != 0x10 || hdr < 0x19) {
 			nv_error(pfb, "invalid/missing timing entry\n");
 			return -EINVAL;
 		}
-	} else {
-		timing.data = 0;
 	}
 
-	ret = nva3_clock_info(nouveau_clock(pfb), 0x12, 0x4000, freq, &mclk);
+	ret = nva3_pll_info(nouveau_clock(pfb), 0x12, 0x4000, freq, &mclk);
 	if (ret < 0) {
 		nv_error(pfb, "failed mclk calculation\n");
 		return ret;
@@ -163,17 +169,17 @@ nva3_ram_calc(struct nouveau_fb *pfb, u32 freq)
 		ram_mask(fuc, 0x004168, 0x003f3141, ctrl);
 	}
 
-	if ( (nv_ro08(bios, ramcfg.data + 0x02) & 0x10)) {
+	if (next->bios.ramcfg_10_02_10) {
 		ram_mask(fuc, 0x111104, 0x00000600, 0x00000000);
 	} else {
 		ram_mask(fuc, 0x111100, 0x40000000, 0x40000000);
 		ram_mask(fuc, 0x111104, 0x00000180, 0x00000000);
 	}
 
-	if (!(nv_ro08(bios, rammap.data + 0x04) & 0x02))
+	if (!next->bios.rammap_10_04_02)
 		ram_mask(fuc, 0x100200, 0x00000800, 0x00000000);
 	ram_wr32(fuc, 0x611200, 0x00003300);
-	if (!(nv_ro08(bios, ramcfg.data + 0x02) & 0x10))
+	if (!next->bios.ramcfg_10_02_10)
 		ram_wr32(fuc, 0x111100, 0x4c020000); /*XXX*/
 
 	ram_wr32(fuc, 0x1002d4, 0x00000001);
@@ -202,17 +208,16 @@ nva3_ram_calc(struct nouveau_fb *pfb, u32 freq)
 		ram_wr32(fuc, 0x004018, 0x0000d000 | r004018);
 	}
 
-	if ( (nv_ro08(bios, rammap.data + 0x04) & 0x08)) {
-		u32 unk5a0 = (nv_ro16(bios, ramcfg.data + 0x05) << 8) |
-			      nv_ro08(bios, ramcfg.data + 0x05);
-		u32 unk5a4 = (nv_ro16(bios, ramcfg.data + 0x07));
-		u32 unk804 = (nv_ro08(bios, ramcfg.data + 0x09) & 0xf0) << 16 |
-			     (nv_ro08(bios, ramcfg.data + 0x03) & 0x0f) << 16 |
-			     (nv_ro08(bios, ramcfg.data + 0x09) & 0x0f) |
-			     0x80000000;
-		ram_wr32(fuc, 0x1005a0, unk5a0);
-		ram_wr32(fuc, 0x1005a4, unk5a4);
-		ram_wr32(fuc, 0x10f804, unk804);
+	if (next->bios.rammap_10_04_08) {
+		ram_wr32(fuc, 0x1005a0, next->bios.ramcfg_10_06 << 16 |
+					next->bios.ramcfg_10_05 << 8 |
+					next->bios.ramcfg_10_05);
+		ram_wr32(fuc, 0x1005a4, next->bios.ramcfg_10_08 << 8 |
+					next->bios.ramcfg_10_07);
+		ram_wr32(fuc, 0x10f804, next->bios.ramcfg_10_09_f0 << 20 |
+					next->bios.ramcfg_10_03_0f << 16 |
+					next->bios.ramcfg_10_09_0f |
+					0x80000000);
 		ram_mask(fuc, 0x10053c, 0x00001000, 0x00000000);
 	} else {
 		ram_mask(fuc, 0x10053c, 0x00001000, 0x00001000);
@@ -250,27 +255,26 @@ nva3_ram_calc(struct nouveau_fb *pfb, u32 freq)
 	ram_mask(fuc, 0x100220[0], 0x00000000, 0x00000000);
 	ram_mask(fuc, 0x100220[8], 0x00000000, 0x00000000);
 
-	data = (nv_ro08(bios, ramcfg.data + 0x02) & 0x08) ? 0x00000000 : 0x00001000;
-	ram_mask(fuc, 0x100200, 0x00001000, data);
+	ram_mask(fuc, 0x100200, 0x00001000, !next->bios.ramcfg_10_02_08 << 12);
 
 	unk714 = ram_rd32(fuc, 0x100714) & ~0xf0000010;
 	unk718 = ram_rd32(fuc, 0x100718) & ~0x00000100;
 	unk71c = ram_rd32(fuc, 0x10071c) & ~0x00000100;
-	if ( (nv_ro08(bios, ramcfg.data + 0x02) & 0x20))
+	if (next->bios.ramcfg_10_02_20)
 		unk714 |= 0xf0000000;
-	if (!(nv_ro08(bios, ramcfg.data + 0x02) & 0x04))
+	if (!next->bios.ramcfg_10_02_04)
 		unk714 |= 0x00000010;
 	ram_wr32(fuc, 0x100714, unk714);
 
-	if (nv_ro08(bios, ramcfg.data + 0x02) & 0x01)
+	if (next->bios.ramcfg_10_02_01)
 		unk71c |= 0x00000100;
 	ram_wr32(fuc, 0x10071c, unk71c);
 
-	if (nv_ro08(bios, ramcfg.data + 0x02) & 0x02)
+	if (next->bios.ramcfg_10_02_02)
 		unk718 |= 0x00000100;
 	ram_wr32(fuc, 0x100718, unk718);
 
-	if (nv_ro08(bios, ramcfg.data + 0x02) & 0x10)
+	if (next->bios.ramcfg_10_02_10)
 		ram_wr32(fuc, 0x111100, 0x48000000); /*XXX*/
 
 	ram_mask(fuc, mr[0], 0x100, 0x100);
@@ -282,9 +286,9 @@ nva3_ram_calc(struct nouveau_fb *pfb, u32 freq)
 	ram_nsec(fuc, 12000);
 
 	ram_wr32(fuc, 0x611200, 0x00003330);
-	if ( (nv_ro08(bios, rammap.data + 0x04) & 0x02))
+	if (next->bios.rammap_10_04_02)
 		ram_mask(fuc, 0x100200, 0x00000800, 0x00000800);
-	if ( (nv_ro08(bios, ramcfg.data + 0x02) & 0x10)) {
+	if (next->bios.ramcfg_10_02_10) {
 		ram_mask(fuc, 0x111104, 0x00000180, 0x00000180);
 		ram_mask(fuc, 0x111100, 0x40000000, 0x00000000);
 	} else {
@@ -404,11 +408,11 @@ nva3_ram_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	ram->fuc.r_0x100714 = ramfuc_reg(0x100714);
 	ram->fuc.r_0x100718 = ramfuc_reg(0x100718);
 	ram->fuc.r_0x10071c = ramfuc_reg(0x10071c);
-	ram->fuc.r_0x100760 = ramfuc_reg(0x100760);
-	ram->fuc.r_0x1007a0 = ramfuc_reg(0x1007a0);
-	ram->fuc.r_0x1007e0 = ramfuc_reg(0x1007e0);
+	ram->fuc.r_0x100760 = ramfuc_stride(0x100760, 4, ram->base.part_mask);
+	ram->fuc.r_0x1007a0 = ramfuc_stride(0x1007a0, 4, ram->base.part_mask);
+	ram->fuc.r_0x1007e0 = ramfuc_stride(0x1007e0, 4, ram->base.part_mask);
 	ram->fuc.r_0x10f804 = ramfuc_reg(0x10f804);
-	ram->fuc.r_0x1110e0 = ramfuc_reg(0x1110e0);
+	ram->fuc.r_0x1110e0 = ramfuc_stride(0x1110e0, 4, ram->base.part_mask);
 	ram->fuc.r_0x111100 = ramfuc_reg(0x111100);
 	ram->fuc.r_0x111104 = ramfuc_reg(0x111104);
 	ram->fuc.r_0x611200 = ramfuc_reg(0x611200);
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnvc0.c b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnvc0.c
index 2b284b192763..735cb9580abe 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnvc0.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnvc0.c
@@ -133,6 +133,7 @@ nvc0_ram_calc(struct nouveau_fb *pfb, u32 freq)
 	struct nouveau_bios *bios = nouveau_bios(pfb);
 	struct nvc0_ram *ram = (void *)pfb->ram;
 	struct nvc0_ramfuc *fuc = &ram->fuc;
+	struct nvbios_ramcfg cfg;
 	u8  ver, cnt, len, strap;
 	struct {
 		u32 data;
@@ -145,7 +146,7 @@ nvc0_ram_calc(struct nouveau_fb *pfb, u32 freq)
 
 	/* lookup memory config data relevant to the target frequency */
 	rammap.data = nvbios_rammapEm(bios, freq / 1000, &ver, &rammap.size,
-				     &cnt, &ramcfg.size);
+				     &cnt, &ramcfg.size, &cfg);
 	if (!rammap.data || ver != 0x10 || rammap.size < 0x0e) {
 		nv_error(pfb, "invalid/missing rammap entry\n");
 		return -EINVAL;
@@ -483,9 +484,9 @@ nvc0_ram_get(struct nouveau_fb *pfb, u64 size, u32 align, u32 ncmin,
 
 	do {
 		if (back)
-			ret = nouveau_mm_tail(mm, 1, size, ncmin, align, &r);
+			ret = nouveau_mm_tail(mm, 0, 1, size, ncmin, align, &r);
 		else
-			ret = nouveau_mm_head(mm, 1, size, ncmin, align, &r);
+			ret = nouveau_mm_head(mm, 0, 1, size, ncmin, align, &r);
 		if (ret) {
 			mutex_unlock(&pfb->base.mutex);
 			pfb->ram->put(pfb, &mem);
@@ -562,7 +563,7 @@ nvc0_ram_create_(struct nouveau_object *parent, struct nouveau_object *engine,
 		offset = (0x0200000000ULL >> 12) + (bsize << 8);
 		length = (ram->size >> 12) - ((bsize * parts) << 8) - rsvd_tail;
 
-		ret = nouveau_mm_init(&pfb->vram, offset, length, 0);
+		ret = nouveau_mm_init(&pfb->vram, offset, length, 1);
 		if (ret)
 			nouveau_mm_fini(&pfb->vram);
 	}
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnve0.c b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnve0.c
index c5b46e302319..6bae474abb44 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnve0.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnve0.c
@@ -29,6 +29,8 @@
 #include <subdev/bios/init.h>
 #include <subdev/bios/rammap.h>
 #include <subdev/bios/timing.h>
+#include <subdev/bios/M0205.h>
+#include <subdev/bios/M0209.h>
 
 #include <subdev/clock.h>
 #include <subdev/clock/pll.h>
@@ -41,14 +43,6 @@
 
 #include "ramfuc.h"
 
-/* binary driver only executes this path if the condition (a) is true
- * for any configuration (combination of rammap+ramcfg+timing) that
- * can be reached on a given card.  for now, we will execute the branch
- * unconditionally in the hope that a "false everywhere" in the bios
- * tables doesn't actually mean "don't touch this".
- */
-#define NOTE00(a) 1
-
 struct nve0_ramfuc {
 	struct ramfuc base;
 
@@ -134,10 +128,12 @@ struct nve0_ram {
 	struct nouveau_ram base;
 	struct nve0_ramfuc fuc;
 
+	struct list_head cfg;
 	u32 parts;
 	u32 pmask;
 	u32 pnuts;
 
+	struct nvbios_ramcfg diff;
 	int from;
 	int mode;
 	int N1, fN1, M1, P1;
@@ -241,7 +237,7 @@ nve0_ram_nuts(struct nve0_ram *ram, struct ramfuc_reg *reg,
 {
 	struct nve0_fb_priv *priv = (void *)nouveau_fb(ram);
 	struct ramfuc *fuc = &ram->fuc.base;
-	u32 addr = 0x110000 + (reg->addr[0] & 0xfff);
+	u32 addr = 0x110000 + (reg->addr & 0xfff);
 	u32 mask = _mask | _copy;
 	u32 data = (_data & _mask) | (reg->data & _copy);
 	u32 i;
@@ -268,6 +264,7 @@ nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 	u32 mask, data;
 
 	ram_mask(fuc, 0x10f808, 0x40000000, 0x40000000);
+	ram_block(fuc);
 	ram_wr32(fuc, 0x62c000, 0x0f0f0000);
 
 	/* MR1: turn termination on early, for some reason.. */
@@ -478,7 +475,7 @@ nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 	ram_mask(fuc, 0x10f2e8, 0xffffffff, next->bios.timing[9]);
 
 	data = mask = 0x00000000;
-	if (NOTE00(ramcfg_08_20)) {
+	if (ram->diff.ramcfg_11_08_20) {
 		if (next->bios.ramcfg_11_08_20)
 			data |= 0x01000000;
 		mask |= 0x01000000;
@@ -486,11 +483,11 @@ nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 	ram_mask(fuc, 0x10f200, mask, data);
 
 	data = mask = 0x00000000;
-	if (NOTE00(ramcfg_02_03 != 0)) {
+	if (ram->diff.ramcfg_11_02_03) {
 		data |= next->bios.ramcfg_11_02_03 << 8;
 		mask |= 0x00000300;
 	}
-	if (NOTE00(ramcfg_01_10)) {
+	if (ram->diff.ramcfg_11_01_10) {
 		if (next->bios.ramcfg_11_01_10)
 			data |= 0x70000000;
 		mask |= 0x70000000;
@@ -498,11 +495,11 @@ nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 	ram_mask(fuc, 0x10f604, mask, data);
 
 	data = mask = 0x00000000;
-	if (NOTE00(timing_30_07 != 0)) {
+	if (ram->diff.timing_20_30_07) {
 		data |= next->bios.timing_20_30_07 << 28;
 		mask |= 0x70000000;
 	}
-	if (NOTE00(ramcfg_01_01)) {
+	if (ram->diff.ramcfg_11_01_01) {
 		if (next->bios.ramcfg_11_01_01)
 			data |= 0x00000100;
 		mask |= 0x00000100;
@@ -510,11 +507,11 @@ nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 	ram_mask(fuc, 0x10f614, mask, data);
 
 	data = mask = 0x00000000;
-	if (NOTE00(timing_30_07 != 0)) {
+	if (ram->diff.timing_20_30_07) {
 		data |= next->bios.timing_20_30_07 << 28;
 		mask |= 0x70000000;
 	}
-	if (NOTE00(ramcfg_01_02)) {
+	if (ram->diff.ramcfg_11_01_02) {
 		if (next->bios.ramcfg_11_01_02)
 			data |= 0x00000100;
 		mask |= 0x00000100;
@@ -548,11 +545,11 @@ nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 	ram_wr32(fuc, 0x10f870, 0x11111111 * next->bios.ramcfg_11_03_0f);
 
 	data = mask = 0x00000000;
-	if (NOTE00(ramcfg_02_03 != 0)) {
+	if (ram->diff.ramcfg_11_02_03) {
 		data |= next->bios.ramcfg_11_02_03;
 		mask |= 0x00000003;
 	}
-	if (NOTE00(ramcfg_01_10)) {
+	if (ram->diff.ramcfg_11_01_10) {
 		if (next->bios.ramcfg_11_01_10)
 			data |= 0x00000004;
 		mask |= 0x00000004;
@@ -666,6 +663,7 @@ nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 	if (next->bios.ramcfg_11_07_02)
 		nve0_ram_train(fuc, 0x80020000, 0x01000000);
 
+	ram_unblock(fuc);
 	ram_wr32(fuc, 0x62c000, 0x0f0f0f00);
 
 	if (next->bios.rammap_11_08_01)
@@ -695,6 +693,7 @@ nve0_ram_calc_sddr3(struct nouveau_fb *pfb, u32 freq)
 	u32 mask, data;
 
 	ram_mask(fuc, 0x10f808, 0x40000000, 0x40000000);
+	ram_block(fuc);
 	ram_wr32(fuc, 0x62c000, 0x0f0f0000);
 
 	if (vc == 1 && ram_have(fuc, gpio2E)) {
@@ -917,6 +916,7 @@ nve0_ram_calc_sddr3(struct nouveau_fb *pfb, u32 freq)
 	ram_mask(fuc, 0x10f200, 0x80000000, 0x00000000);
 	ram_nsec(fuc, 1000);
 
+	ram_unblock(fuc);
 	ram_wr32(fuc, 0x62c000, 0x0f0f0f00);
 
 	if (next->bios.rammap_11_08_01)
@@ -932,58 +932,24 @@ nve0_ram_calc_sddr3(struct nouveau_fb *pfb, u32 freq)
  ******************************************************************************/
 
 static int
-nve0_ram_calc_data(struct nouveau_fb *pfb, u32 freq,
+nve0_ram_calc_data(struct nouveau_fb *pfb, u32 khz,
 		   struct nouveau_ram_data *data)
 {
-	struct nouveau_bios *bios = nouveau_bios(pfb);
 	struct nve0_ram *ram = (void *)pfb->ram;
-	u8 strap, cnt, len;
-
-	/* lookup memory config data relevant to the target frequency */
-	ram->base.rammap.data = nvbios_rammapEp(bios, freq / 1000,
-					       &ram->base.rammap.version,
-					       &ram->base.rammap.size,
-					       &cnt, &len, &data->bios);
-	if (!ram->base.rammap.data || ram->base.rammap.version != 0x11 ||
-	     ram->base.rammap.size < 0x09) {
-		nv_error(pfb, "invalid/missing rammap entry\n");
-		return -EINVAL;
-	}
-
-	/* locate specific data set for the attached memory */
-	strap = nvbios_ramcfg_index(nv_subdev(pfb));
-	ram->base.ramcfg.data = nvbios_rammapSp(bios, ram->base.rammap.data,
-						ram->base.rammap.version,
-						ram->base.rammap.size,
-						cnt, len, strap,
-						&ram->base.ramcfg.version,
-						&ram->base.ramcfg.size,
-						&data->bios);
-	if (!ram->base.ramcfg.data || ram->base.ramcfg.version != 0x11 ||
-	     ram->base.ramcfg.size < 0x08) {
-		nv_error(pfb, "invalid/missing ramcfg entry\n");
-		return -EINVAL;
-	}
-
-	/* lookup memory timings, if bios says they're present */
-	strap = nv_ro08(bios, ram->base.ramcfg.data + 0x00);
-	if (strap != 0xff) {
-		ram->base.timing.data =
-			nvbios_timingEp(bios, strap, &ram->base.timing.version,
-				       &ram->base.timing.size, &cnt, &len,
-				       &data->bios);
-		if (!ram->base.timing.data ||
-		     ram->base.timing.version != 0x20 ||
-		     ram->base.timing.size < 0x33) {
-			nv_error(pfb, "invalid/missing timing entry\n");
-			return -EINVAL;
+	struct nouveau_ram_data *cfg;
+	u32 mhz = khz / 1000;
+
+	list_for_each_entry(cfg, &ram->cfg, head) {
+		if (mhz >= cfg->bios.rammap_min &&
+		    mhz <= cfg->bios.rammap_max) {
+			*data = *cfg;
+			data->freq = khz;
+			return 0;
 		}
-	} else {
-		ram->base.timing.data = 0;
 	}
 
-	data->freq = freq;
-	return 0;
+	nv_error(ram, "ramcfg data for %dMHz not found\n", mhz);
+	return -EINVAL;
 }
 
 static int
@@ -1106,13 +1072,99 @@ nve0_ram_calc(struct nouveau_fb *pfb, u32 freq)
 	return nve0_ram_calc_xits(pfb, ram->base.next);
 }
 
+static void
+nve0_ram_prog_0(struct nouveau_fb *pfb, u32 freq)
+{
+	struct nve0_ram *ram = (void *)pfb->ram;
+	struct nouveau_ram_data *cfg;
+	u32 mhz = freq / 1000;
+	u32 mask, data;
+
+	list_for_each_entry(cfg, &ram->cfg, head) {
+		if (mhz >= cfg->bios.rammap_min &&
+		    mhz <= cfg->bios.rammap_max)
+			break;
+	}
+
+	if (&cfg->head == &ram->cfg)
+		return;
+
+	if (mask = 0, data = 0, ram->diff.rammap_11_0a_03fe) {
+		data |= cfg->bios.rammap_11_0a_03fe << 12;
+		mask |= 0x001ff000;
+	}
+	if (ram->diff.rammap_11_09_01ff) {
+		data |= cfg->bios.rammap_11_09_01ff;
+		mask |= 0x000001ff;
+	}
+	nv_mask(pfb, 0x10f468, mask, data);
+
+	if (mask = 0, data = 0, ram->diff.rammap_11_0a_0400) {
+		data |= cfg->bios.rammap_11_0a_0400;
+		mask |= 0x00000001;
+	}
+	nv_mask(pfb, 0x10f420, mask, data);
+
+	if (mask = 0, data = 0, ram->diff.rammap_11_0a_0800) {
+		data |= cfg->bios.rammap_11_0a_0800;
+		mask |= 0x00000001;
+	}
+	nv_mask(pfb, 0x10f430, mask, data);
+
+	if (mask = 0, data = 0, ram->diff.rammap_11_0b_01f0) {
+		data |= cfg->bios.rammap_11_0b_01f0;
+		mask |= 0x0000001f;
+	}
+	nv_mask(pfb, 0x10f400, mask, data);
+
+	if (mask = 0, data = 0, ram->diff.rammap_11_0b_0200) {
+		data |= cfg->bios.rammap_11_0b_0200 << 9;
+		mask |= 0x00000200;
+	}
+	nv_mask(pfb, 0x10f410, mask, data);
+
+	if (mask = 0, data = 0, ram->diff.rammap_11_0d) {
+		data |= cfg->bios.rammap_11_0d << 16;
+		mask |= 0x00ff0000;
+	}
+	if (ram->diff.rammap_11_0f) {
+		data |= cfg->bios.rammap_11_0f << 8;
+		mask |= 0x0000ff00;
+	}
+	nv_mask(pfb, 0x10f440, mask, data);
+
+	if (mask = 0, data = 0, ram->diff.rammap_11_0e) {
+		data |= cfg->bios.rammap_11_0e << 8;
+		mask |= 0x0000ff00;
+	}
+	if (ram->diff.rammap_11_0b_0800) {
+		data |= cfg->bios.rammap_11_0b_0800 << 7;
+		mask |= 0x00000080;
+	}
+	if (ram->diff.rammap_11_0b_0400) {
+		data |= cfg->bios.rammap_11_0b_0400 << 5;
+		mask |= 0x00000020;
+	}
+	nv_mask(pfb, 0x10f444, mask, data);
+}
+
 static int
 nve0_ram_prog(struct nouveau_fb *pfb)
 {
 	struct nouveau_device *device = nv_device(pfb);
 	struct nve0_ram *ram = (void *)pfb->ram;
 	struct nve0_ramfuc *fuc = &ram->fuc;
-	ram_exec(fuc, nouveau_boolopt(device->cfgopt, "NvMemExec", true));
+	struct nouveau_ram_data *next = ram->base.next;
+
+	if (!nouveau_boolopt(device->cfgopt, "NvMemExec", true)) {
+		ram_exec(fuc, false);
+		return (ram->base.next == &ram->base.xition);
+	}
+
+	nve0_ram_prog_0(pfb, 1000);
+	ram_exec(fuc, true);
+	nve0_ram_prog_0(pfb, next->freq);
+
 	return (ram->base.next == &ram->base.xition);
 }
 
@@ -1125,24 +1177,147 @@ nve0_ram_tidy(struct nouveau_fb *pfb)
 	ram_exec(fuc, false);
 }
 
+struct nve0_ram_train {
+	u16 mask;
+	struct nvbios_M0209S remap;
+	struct nvbios_M0209S type00;
+	struct nvbios_M0209S type01;
+	struct nvbios_M0209S type04;
+	struct nvbios_M0209S type06;
+	struct nvbios_M0209S type07;
+	struct nvbios_M0209S type08;
+	struct nvbios_M0209S type09;
+};
+
+static int
+nve0_ram_train_type(struct nouveau_fb *pfb, int i, u8 ramcfg,
+		    struct nve0_ram_train *train)
+{
+	struct nouveau_bios *bios = nouveau_bios(pfb);
+	struct nvbios_M0205E M0205E;
+	struct nvbios_M0205S M0205S;
+	struct nvbios_M0209E M0209E;
+	struct nvbios_M0209S *remap = &train->remap;
+	struct nvbios_M0209S *value;
+	u8  ver, hdr, cnt, len;
+	u32 data;
+
+	/* determine type of data for this index */
+	if (!(data = nvbios_M0205Ep(bios, i, &ver, &hdr, &cnt, &len, &M0205E)))
+		return -ENOENT;
+
+	switch (M0205E.type) {
+	case 0x00: value = &train->type00; break;
+	case 0x01: value = &train->type01; break;
+	case 0x04: value = &train->type04; break;
+	case 0x06: value = &train->type06; break;
+	case 0x07: value = &train->type07; break;
+	case 0x08: value = &train->type08; break;
+	case 0x09: value = &train->type09; break;
+	default:
+		return 0;
+	}
+
+	/* training data index determined by ramcfg strap */
+	if (!(data = nvbios_M0205Sp(bios, i, ramcfg, &ver, &hdr, &M0205S)))
+		return -EINVAL;
+	i = M0205S.data;
+
+	/* training data format information */
+	if (!(data = nvbios_M0209Ep(bios, i, &ver, &hdr, &cnt, &len, &M0209E)))
+		return -EINVAL;
+
+	/* ... and the raw data */
+	if (!(data = nvbios_M0209Sp(bios, i, 0, &ver, &hdr, value)))
+		return -EINVAL;
+
+	if (M0209E.v02_07 == 2) {
+		/* of course! why wouldn't we have a pointer to another entry
+		 * in the same table, and use the first one as an array of
+		 * remap indices...
+		 */
+		if (!(data = nvbios_M0209Sp(bios, M0209E.v03, 0, &ver, &hdr,
+					    remap)))
+			return -EINVAL;
+
+		for (i = 0; i < ARRAY_SIZE(value->data); i++)
+			value->data[i] = remap->data[value->data[i]];
+	} else
+	if (M0209E.v02_07 != 1)
+		return -EINVAL;
+
+	train->mask |= 1 << M0205E.type;
+	return 0;
+}
+
+static int
+nve0_ram_train_init_0(struct nouveau_fb *pfb, struct nve0_ram_train *train)
+{
+	int i, j;
+
+	if ((train->mask & 0x03d3) != 0x03d3) {
+		nv_warn(pfb, "missing link training data\n");
+		return -EINVAL;
+	}
+
+	for (i = 0; i < 0x30; i++) {
+		for (j = 0; j < 8; j += 4) {
+			nv_wr32(pfb, 0x10f968 + j, 0x00000000 | (i << 8));
+			nv_wr32(pfb, 0x10f920 + j, 0x00000000 |
+						   train->type08.data[i] << 4 |
+						   train->type06.data[i]);
+			nv_wr32(pfb, 0x10f918 + j, train->type00.data[i]);
+			nv_wr32(pfb, 0x10f920 + j, 0x00000100 |
+						   train->type09.data[i] << 4 |
+						   train->type07.data[i]);
+			nv_wr32(pfb, 0x10f918 + j, train->type01.data[i]);
+		}
+	}
+
+	for (j = 0; j < 8; j += 4) {
+		for (i = 0; i < 0x100; i++) {
+			nv_wr32(pfb, 0x10f968 + j, i);
+			nv_wr32(pfb, 0x10f900 + j, train->type04.data[i]);
+		}
+	}
+
+	return 0;
+}
+
+static int
+nve0_ram_train_init(struct nouveau_fb *pfb)
+{
+	u8 ramcfg = nvbios_ramcfg_index(nv_subdev(pfb));
+	struct nve0_ram_train *train;
+	int ret = -ENOMEM, i;
+
+	if ((train = kzalloc(sizeof(*train), GFP_KERNEL))) {
+		for (i = 0; i < 0x100; i++) {
+			ret = nve0_ram_train_type(pfb, i, ramcfg, train);
+			if (ret && ret != -ENOENT)
+				break;
+		}
+	}
+
+	switch (pfb->ram->type) {
+	case NV_MEM_TYPE_GDDR5:
+		ret = nve0_ram_train_init_0(pfb, train);
+		break;
+	default:
+		ret = 0;
+		break;
+	}
+
+	kfree(train);
+	return ret;
+}
+
 int
 nve0_ram_init(struct nouveau_object *object)
 {
 	struct nouveau_fb *pfb = (void *)object->parent;
 	struct nve0_ram *ram   = (void *)object;
 	struct nouveau_bios *bios = nouveau_bios(pfb);
-	static const u8  train0[] = {
-		0x00, 0xff, 0xff, 0x00, 0xff, 0x00,
-		0x00, 0xff, 0xff, 0x00, 0xff, 0x00,
-	};
-	static const u32 train1[] = {
-		0x00000000, 0xffffffff,
-		0x55555555, 0xaaaaaaaa,
-		0x33333333, 0xcccccccc,
-		0xf0f0f0f0, 0x0f0f0f0f,
-		0x00ff00ff, 0xff00ff00,
-		0x0000ffff, 0xffff0000,
-	};
 	u8  ver, hdr, cnt, len, snr, ssz;
 	u32 data, save;
 	int ret, i;
@@ -1168,51 +1343,107 @@ nve0_ram_init(struct nouveau_object *object)
 
 	cnt  = nv_ro08(bios, data + 0x14); /* guess at count */
 	data = nv_ro32(bios, data + 0x10); /* guess u32... */
-	save = nv_rd32(pfb, 0x10f65c);
-	for (i = 0; i < cnt; i++) {
-		nv_mask(pfb, 0x10f65c, 0x000000f0, i << 4);
-		nvbios_exec(&(struct nvbios_init) {
-				.subdev = nv_subdev(pfb),
-				.bios = bios,
-				.offset = nv_ro32(bios, data), /* guess u32 */
-				.execute = 1,
-			    });
-		data += 4;
-	}
-	nv_wr32(pfb, 0x10f65c, save);
+	save = nv_rd32(pfb, 0x10f65c) & 0x000000f0;
+	for (i = 0; i < cnt; i++, data += 4) {
+		if (i != save >> 4) {
+			nv_mask(pfb, 0x10f65c, 0x000000f0, i << 4);
+			nvbios_exec(&(struct nvbios_init) {
+					.subdev = nv_subdev(pfb),
+					.bios = bios,
+					.offset = nv_ro32(bios, data),
+					.execute = 1,
+				    });
+		}
+	}
+	nv_mask(pfb, 0x10f65c, 0x000000f0, save);
 	nv_mask(pfb, 0x10f584, 0x11000000, 0x00000000);
+	nv_wr32(pfb, 0x10ecc0, 0xffffffff);
+	nv_mask(pfb, 0x10f160, 0x00000010, 0x00000010);
 
-	switch (ram->base.type) {
-	case NV_MEM_TYPE_GDDR5:
-		for (i = 0; i < 0x30; i++) {
-			nv_wr32(pfb, 0x10f968, 0x00000000 | (i << 8));
-			nv_wr32(pfb, 0x10f920, 0x00000000 | train0[i % 12]);
-			nv_wr32(pfb, 0x10f918,              train1[i % 12]);
-			nv_wr32(pfb, 0x10f920, 0x00000100 | train0[i % 12]);
-			nv_wr32(pfb, 0x10f918,              train1[i % 12]);
-
-			nv_wr32(pfb, 0x10f96c, 0x00000000 | (i << 8));
-			nv_wr32(pfb, 0x10f924, 0x00000000 | train0[i % 12]);
-			nv_wr32(pfb, 0x10f91c,              train1[i % 12]);
-			nv_wr32(pfb, 0x10f924, 0x00000100 | train0[i % 12]);
-			nv_wr32(pfb, 0x10f91c,              train1[i % 12]);
-		}
+	return nve0_ram_train_init(pfb);
+}
 
-		for (i = 0; i < 0x100; i++) {
-			nv_wr32(pfb, 0x10f968, i);
-			nv_wr32(pfb, 0x10f900, train1[2 + (i & 1)]);
-		}
+static int
+nve0_ram_ctor_data(struct nve0_ram *ram, u8 ramcfg, int i)
+{
+	struct nouveau_fb *pfb = (void *)nv_object(ram)->parent;
+	struct nouveau_bios *bios = nouveau_bios(pfb);
+	struct nouveau_ram_data *cfg;
+	struct nvbios_ramcfg *d = &ram->diff;
+	struct nvbios_ramcfg *p, *n;
+	u8  ver, hdr, cnt, len;
+	u32 data;
+	int ret;
 
-		for (i = 0; i < 0x100; i++) {
-			nv_wr32(pfb, 0x10f96c, i);
-			nv_wr32(pfb, 0x10f900, train1[2 + (i & 1)]);
-		}
-		break;
-	default:
-		break;
+	if (!(cfg = kmalloc(sizeof(*cfg), GFP_KERNEL)))
+		return -ENOMEM;
+	p = &list_last_entry(&ram->cfg, typeof(*cfg), head)->bios;
+	n = &cfg->bios;
+
+	/* memory config data for a range of target frequencies */
+	data = nvbios_rammapEp(bios, i, &ver, &hdr, &cnt, &len, &cfg->bios);
+	if (ret = -ENOENT, !data)
+		goto done;
+	if (ret = -ENOSYS, ver != 0x11 || hdr < 0x12)
+		goto done;
+
+	/* ... and a portion specific to the attached memory */
+	data = nvbios_rammapSp(bios, data, ver, hdr, cnt, len, ramcfg,
+			       &ver, &hdr, &cfg->bios);
+	if (ret = -EINVAL, !data)
+		goto done;
+	if (ret = -ENOSYS, ver != 0x11 || hdr < 0x0a)
+		goto done;
+
+	/* lookup memory timings, if bios says they're present */
+	if (cfg->bios.ramcfg_timing != 0xff) {
+		data = nvbios_timingEp(bios, cfg->bios.ramcfg_timing,
+				       &ver, &hdr, &cnt, &len,
+				       &cfg->bios);
+		if (ret = -EINVAL, !data)
+			goto done;
+		if (ret = -ENOSYS, ver != 0x20 || hdr < 0x33)
+			goto done;
 	}
 
-	return 0;
+	list_add_tail(&cfg->head, &ram->cfg);
+	if (ret = 0, i == 0)
+		goto done;
+
+	d->rammap_11_0a_03fe |= p->rammap_11_0a_03fe != n->rammap_11_0a_03fe;
+	d->rammap_11_09_01ff |= p->rammap_11_09_01ff != n->rammap_11_09_01ff;
+	d->rammap_11_0a_0400 |= p->rammap_11_0a_0400 != n->rammap_11_0a_0400;
+	d->rammap_11_0a_0800 |= p->rammap_11_0a_0800 != n->rammap_11_0a_0800;
+	d->rammap_11_0b_01f0 |= p->rammap_11_0b_01f0 != n->rammap_11_0b_01f0;
+	d->rammap_11_0b_0200 |= p->rammap_11_0b_0200 != n->rammap_11_0b_0200;
+	d->rammap_11_0d |= p->rammap_11_0d != n->rammap_11_0d;
+	d->rammap_11_0f |= p->rammap_11_0f != n->rammap_11_0f;
+	d->rammap_11_0e |= p->rammap_11_0e != n->rammap_11_0e;
+	d->rammap_11_0b_0800 |= p->rammap_11_0b_0800 != n->rammap_11_0b_0800;
+	d->rammap_11_0b_0400 |= p->rammap_11_0b_0400 != n->rammap_11_0b_0400;
+	d->ramcfg_11_01_01 |= p->ramcfg_11_01_01 != n->ramcfg_11_01_01;
+	d->ramcfg_11_01_02 |= p->ramcfg_11_01_02 != n->ramcfg_11_01_02;
+	d->ramcfg_11_01_10 |= p->ramcfg_11_01_10 != n->ramcfg_11_01_10;
+	d->ramcfg_11_02_03 |= p->ramcfg_11_02_03 != n->ramcfg_11_02_03;
+	d->ramcfg_11_08_20 |= p->ramcfg_11_08_20 != n->ramcfg_11_08_20;
+	d->timing_20_30_07 |= p->timing_20_30_07 != n->timing_20_30_07;
+done:
+	if (ret)
+		kfree(cfg);
+	return ret;
+}
+
+static void
+nve0_ram_dtor(struct nouveau_object *object)
+{
+	struct nve0_ram *ram = (void *)object;
+	struct nouveau_ram_data *cfg, *tmp;
+
+	list_for_each_entry_safe(cfg, tmp, &ram->cfg, head) {
+		kfree(cfg);
+	}
+
+	nouveau_ram_destroy(&ram->base);
 }
 
 static int
@@ -1226,6 +1457,7 @@ nve0_ram_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	struct dcb_gpio_func func;
 	struct nve0_ram *ram;
 	int ret, i;
+	u8  ramcfg = nvbios_ramcfg_index(nv_subdev(pfb));
 	u32 tmp;
 
 	ret = nvc0_ram_create(parent, engine, oclass, 0x022554, &ram);
@@ -1233,6 +1465,8 @@ nve0_ram_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	if (ret)
 		return ret;
 
+	INIT_LIST_HEAD(&ram->cfg);
+
 	switch (ram->base.type) {
 	case NV_MEM_TYPE_DDR3:
 	case NV_MEM_TYPE_GDDR5:
@@ -1264,7 +1498,26 @@ nve0_ram_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 		}
 	}
 
-	// parse bios data for both pll's
+	/* parse bios data for all rammap table entries up-front, and
+	 * build information on whether certain fields differ between
+	 * any of the entries.
+	 *
+	 * the binary driver appears to completely ignore some fields
+	 * when all entries contain the same value.  at first, it was
+	 * hoped that these were mere optimisations and the bios init
+	 * tables had configured as per the values here, but there is
+	 * evidence now to suggest that this isn't the case and we do
+	 * need to treat this condition as a "don't touch" indicator.
+	 */
+	for (i = 0; !ret; i++) {
+		ret = nve0_ram_ctor_data(ram, ramcfg, i);
+		if (ret && ret != -ENOENT) {
+			nv_error(pfb, "failed to parse ramcfg data\n");
+			return ret;
+		}
+	}
+
+	/* parse bios data for both pll's */
 	ret = nvbios_pll_parse(bios, 0x0c, &ram->fuc.refpll);
 	if (ret) {
 		nv_error(pfb, "mclk refpll data not found\n");
@@ -1277,6 +1530,7 @@ nve0_ram_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 		return ret;
 	}
 
+	/* lookup memory voltage gpios */
 	ret = gpio->find(gpio, 0, 0x18, DCB_GPIO_UNUSED, &func);
 	if (ret == 0) {
 		ram->fuc.r_gpioMV = ramfuc_reg(0x00d610 + (func.line * 0x04));
@@ -1385,7 +1639,7 @@ nve0_ram_oclass = {
 	.handle = 0,
 	.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = nve0_ram_ctor,
-		.dtor = _nouveau_ram_dtor,
+		.dtor = nve0_ram_dtor,
 		.init = nve0_ram_init,
 		.fini = _nouveau_ram_fini,
 	}
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/sddr2.c b/drivers/gpu/drm/nouveau/core/subdev/fb/sddr2.c
new file mode 100644
index 000000000000..bb1eb8f3e639
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/sddr2.c
@@ -0,0 +1,94 @@
+/*
+ * Copyright 2014 Roy Spliet
+ *
+ * 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: Roy Spliet <rspliet@eclipso.eu>
+ *          Ben Skeggs
+ */
+
+#include "priv.h"
+
+struct ramxlat {
+	int id;
+	u8 enc;
+};
+
+static inline int
+ramxlat(const struct ramxlat *xlat, int id)
+{
+	while (xlat->id >= 0) {
+		if (xlat->id == id)
+			return xlat->enc;
+		xlat++;
+	}
+	return -EINVAL;
+}
+
+static const struct ramxlat
+ramddr2_cl[] = {
+	{ 2, 2 }, { 3, 3 }, { 4, 4 }, { 5, 5 }, { 6, 6 },
+	/* The following are available in some, but not all DDR2 docs */
+	{ 7, 7 },
+	{ -1 }
+};
+
+static const struct ramxlat
+ramddr2_wr[] = {
+	{ 2, 1 }, { 3, 2 }, { 4, 3 }, { 5, 4 }, { 6, 5 },
+	/* The following are available in some, but not all DDR2 docs */
+	{ 7, 6 },
+	{ -1 }
+};
+
+int
+nouveau_sddr2_calc(struct nouveau_ram *ram)
+{
+	int CL, WR, DLL = 0, ODT = 0;
+
+	switch (ram->next->bios.timing_ver) {
+	case 0x10:
+		CL  = ram->next->bios.timing_10_CL;
+		WR  = ram->next->bios.timing_10_WR;
+		DLL = !ram->next->bios.ramcfg_10_02_40;
+		ODT = ram->next->bios.timing_10_ODT & 3;
+		break;
+	case 0x20:
+		CL  = (ram->next->bios.timing[1] & 0x0000001f);
+		WR  = (ram->next->bios.timing[2] & 0x007f0000) >> 16;
+		break;
+	default:
+		return -ENOSYS;
+	}
+
+	CL  = ramxlat(ramddr2_cl, CL);
+	WR  = ramxlat(ramddr2_wr, WR);
+	if (CL < 0 || WR < 0)
+		return -EINVAL;
+
+	ram->mr[0] &= ~0xf70;
+	ram->mr[0] |= (WR & 0x07) << 9;
+	ram->mr[0] |= (CL & 0x07) << 4;
+
+	ram->mr[1] &= ~0x045;
+	ram->mr[1] |= (ODT & 0x1) << 2;
+	ram->mr[1] |= (ODT & 0x2) << 5;
+	ram->mr[1] |= !DLL;
+	return 0;
+}
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/sddr3.c b/drivers/gpu/drm/nouveau/core/subdev/fb/sddr3.c
index ebd4cd9c35d9..83949b11833a 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/sddr3.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/sddr3.c
@@ -20,9 +20,9 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  *
  * Authors: Ben Skeggs <bskeggs@redhat.com>
+ * 	    Roy Spliet <rspliet@eclipso.eu>
  */
 
-#include <subdev/bios.h>
 #include "priv.h"
 
 struct ramxlat {
@@ -69,31 +69,52 @@ ramddr3_cwl[] = {
 int
 nouveau_sddr3_calc(struct nouveau_ram *ram)
 {
-	struct nouveau_bios *bios = nouveau_bios(ram);
-	int WL, CL, WR;
+	int CWL, CL, WR, DLL = 0, ODT = 0;
 
-	switch (!!ram->timing.data * ram->timing.version) {
+	switch (ram->next->bios.timing_ver) {
+	case 0x10:
+		if (ram->next->bios.timing_hdr < 0x17) {
+			/* XXX: NV50: Get CWL from the timing register */
+			return -ENOSYS;
+		}
+		CWL = ram->next->bios.timing_10_CWL;
+		CL  = ram->next->bios.timing_10_CL;
+		WR  = ram->next->bios.timing_10_WR;
+		DLL = !ram->next->bios.ramcfg_10_02_40;
+		ODT = ram->next->bios.timing_10_ODT;
+		break;
 	case 0x20:
-		WL = (nv_ro16(bios, ram->timing.data + 0x04) & 0x0f80) >> 7;
-		CL =  nv_ro08(bios, ram->timing.data + 0x04) & 0x1f;
-		WR =  nv_ro08(bios, ram->timing.data + 0x0a) & 0x7f;
+		CWL = (ram->next->bios.timing[1] & 0x00000f80) >> 7;
+		CL  = (ram->next->bios.timing[1] & 0x0000001f) >> 0;
+		WR  = (ram->next->bios.timing[2] & 0x007f0000) >> 16;
+		/* XXX: Get these values from the VBIOS instead */
+		DLL = !(ram->mr[1] & 0x1);
+		ODT =   (ram->mr[1] & 0x004) >> 2 |
+			(ram->mr[1] & 0x040) >> 5 |
+		        (ram->mr[1] & 0x200) >> 7;
 		break;
 	default:
 		return -ENOSYS;
 	}
 
-	WL = ramxlat(ramddr3_cwl, WL);
-	CL = ramxlat(ramddr3_cl, CL);
-	WR = ramxlat(ramddr3_wr, WR);
-	if (WL < 0 || CL < 0 || WR < 0)
+	CWL = ramxlat(ramddr3_cwl, CWL);
+	CL  = ramxlat(ramddr3_cl, CL);
+	WR  = ramxlat(ramddr3_wr, WR);
+	if (CL < 0 || CWL < 0 || WR < 0)
 		return -EINVAL;
 
-	ram->mr[0] &= ~0xe74;
+	ram->mr[0] &= ~0xf74;
 	ram->mr[0] |= (WR & 0x07) << 9;
 	ram->mr[0] |= (CL & 0x0e) << 3;
 	ram->mr[0] |= (CL & 0x01) << 2;
 
+	ram->mr[1] &= ~0x245;
+	ram->mr[1] |= (ODT & 0x1) << 2;
+	ram->mr[1] |= (ODT & 0x2) << 5;
+	ram->mr[1] |= (ODT & 0x4) << 7;
+	ram->mr[1] |= !DLL;
+
 	ram->mr[2] &= ~0x038;
-	ram->mr[2] |= (WL & 0x07) << 3;
+	ram->mr[2] |= (CWL & 0x07) << 3;
 	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fuse/base.c b/drivers/gpu/drm/nouveau/core/subdev/fuse/base.c
new file mode 100644
index 000000000000..9e8e92127715
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/fuse/base.c
@@ -0,0 +1,54 @@
+/*
+ * Copyright 2014 Martin Peres
+ *
+ * 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: Martin Peres
+ */
+
+#include <subdev/fuse.h>
+
+int
+_nouveau_fuse_init(struct nouveau_object *object)
+{
+	struct nouveau_fuse *fuse = (void *)object;
+	return nouveau_subdev_init(&fuse->base);
+}
+
+void
+_nouveau_fuse_dtor(struct nouveau_object *object)
+{
+	struct nouveau_fuse *fuse = (void *)object;
+	nouveau_subdev_destroy(&fuse->base);
+}
+
+int
+nouveau_fuse_create_(struct nouveau_object *parent,
+		     struct nouveau_object *engine,
+		     struct nouveau_oclass *oclass, int length, void **pobject)
+{
+	struct nouveau_fuse *fuse;
+	int ret;
+
+	ret = nouveau_subdev_create_(parent, engine, oclass, 0, "FUSE",
+				     "fuse", length, pobject);
+	fuse = *pobject;
+
+	return ret;
+}
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fuse/g80.c b/drivers/gpu/drm/nouveau/core/subdev/fuse/g80.c
new file mode 100644
index 000000000000..a374ade485be
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/fuse/g80.c
@@ -0,0 +1,81 @@
+/*
+ * Copyright 2014 Martin Peres
+ *
+ * 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: Martin Peres
+ */
+
+#include "priv.h"
+
+struct g80_fuse_priv {
+	struct nouveau_fuse base;
+
+	spinlock_t fuse_enable_lock;
+};
+
+static u32
+g80_fuse_rd32(struct nouveau_object *object, u64 addr)
+{
+	struct g80_fuse_priv *priv = (void *)object;
+	unsigned long flags;
+	u32 fuse_enable, val;
+
+	spin_lock_irqsave(&priv->fuse_enable_lock, flags);
+
+	/* racy if another part of nouveau start writing to this reg */
+	fuse_enable = nv_mask(priv, 0x1084, 0x800, 0x800);
+	val = nv_rd32(priv, 0x21000 + addr);
+	nv_wr32(priv, 0x1084, fuse_enable);
+
+	spin_unlock_irqrestore(&priv->fuse_enable_lock, flags);
+
+	return val;
+}
+
+
+static int
+g80_fuse_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
+	       struct nouveau_oclass *oclass, void *data, u32 size,
+	       struct nouveau_object **pobject)
+{
+	struct g80_fuse_priv *priv;
+	int ret;
+
+	ret = nouveau_fuse_create(parent, engine, oclass, &priv);
+	*pobject = nv_object(priv);
+	if (ret)
+		return ret;
+
+	spin_lock_init(&priv->fuse_enable_lock);
+
+	return 0;
+}
+
+struct nouveau_oclass
+g80_fuse_oclass = {
+	.handle = NV_SUBDEV(FUSE, 0x50),
+	.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = g80_fuse_ctor,
+		.dtor = _nouveau_fuse_dtor,
+		.init = _nouveau_fuse_init,
+		.fini = _nouveau_fuse_fini,
+		.rd32 = g80_fuse_rd32,
+	},
+};
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fuse/gf100.c b/drivers/gpu/drm/nouveau/core/subdev/fuse/gf100.c
new file mode 100644
index 000000000000..5ed03f54b3d4
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/fuse/gf100.c
@@ -0,0 +1,83 @@
+/*
+ * Copyright 2014 Martin Peres
+ *
+ * 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: Martin Peres
+ */
+
+#include "priv.h"
+
+struct gf100_fuse_priv {
+	struct nouveau_fuse base;
+
+	spinlock_t fuse_enable_lock;
+};
+
+static u32
+gf100_fuse_rd32(struct nouveau_object *object, u64 addr)
+{
+	struct gf100_fuse_priv *priv = (void *)object;
+	unsigned long flags;
+	u32 fuse_enable, unk, val;
+
+	spin_lock_irqsave(&priv->fuse_enable_lock, flags);
+
+	/* racy if another part of nouveau start writing to these regs */
+	fuse_enable = nv_mask(priv, 0x22400, 0x800, 0x800);
+	unk = nv_mask(priv, 0x21000, 0x1, 0x1);
+	val = nv_rd32(priv, 0x21100 + addr);
+	nv_wr32(priv, 0x21000, unk);
+	nv_wr32(priv, 0x22400, fuse_enable);
+
+	spin_unlock_irqrestore(&priv->fuse_enable_lock, flags);
+
+	return val;
+}
+
+
+static int
+gf100_fuse_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
+	       struct nouveau_oclass *oclass, void *data, u32 size,
+	       struct nouveau_object **pobject)
+{
+	struct gf100_fuse_priv *priv;
+	int ret;
+
+	ret = nouveau_fuse_create(parent, engine, oclass, &priv);
+	*pobject = nv_object(priv);
+	if (ret)
+		return ret;
+
+	spin_lock_init(&priv->fuse_enable_lock);
+
+	return 0;
+}
+
+struct nouveau_oclass
+gf100_fuse_oclass = {
+	.handle = NV_SUBDEV(FUSE, 0xC0),
+	.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = gf100_fuse_ctor,
+		.dtor = _nouveau_fuse_dtor,
+		.init = _nouveau_fuse_init,
+		.fini = _nouveau_fuse_fini,
+		.rd32 = gf100_fuse_rd32,
+	},
+};
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fuse/gm107.c b/drivers/gpu/drm/nouveau/core/subdev/fuse/gm107.c
new file mode 100644
index 000000000000..4f1a636c6538
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/fuse/gm107.c
@@ -0,0 +1,66 @@
+/*
+ * Copyright 2014 Martin Peres
+ *
+ * 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: Martin Peres
+ */
+
+#include "priv.h"
+
+struct gm107_fuse_priv {
+	struct nouveau_fuse base;
+};
+
+static u32
+gm107_fuse_rd32(struct nouveau_object *object, u64 addr)
+{
+	struct gf100_fuse_priv *priv = (void *)object;
+
+	return nv_rd32(priv, 0x21100 + addr);
+}
+
+
+static int
+gm107_fuse_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
+	       struct nouveau_oclass *oclass, void *data, u32 size,
+	       struct nouveau_object **pobject)
+{
+	struct gm107_fuse_priv *priv;
+	int ret;
+
+	ret = nouveau_fuse_create(parent, engine, oclass, &priv);
+	*pobject = nv_object(priv);
+	if (ret)
+		return ret;
+
+	return 0;
+}
+
+struct nouveau_oclass
+gm107_fuse_oclass = {
+	.handle = NV_SUBDEV(FUSE, 0x117),
+	.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = gm107_fuse_ctor,
+		.dtor = _nouveau_fuse_dtor,
+		.init = _nouveau_fuse_init,
+		.fini = _nouveau_fuse_fini,
+		.rd32 = gm107_fuse_rd32,
+	},
+};
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fuse/priv.h b/drivers/gpu/drm/nouveau/core/subdev/fuse/priv.h
new file mode 100644
index 000000000000..d2085411a5cb
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/fuse/priv.h
@@ -0,0 +1,9 @@
+#ifndef __NVKM_FUSE_PRIV_H__
+#define __NVKM_FUSE_PRIV_H__
+
+#include <subdev/fuse.h>
+
+int _nouveau_fuse_init(struct nouveau_object *object);
+void _nouveau_fuse_dtor(struct nouveau_object *object);
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/gpio/base.c b/drivers/gpu/drm/nouveau/core/subdev/gpio/base.c
index b1e3ed7c8beb..7ad99b763f4c 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/gpio/base.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/gpio/base.c
@@ -122,7 +122,8 @@ nouveau_gpio_intr_init(struct nvkm_event *event, int type, int index)
 }
 
 static int
-nouveau_gpio_intr_ctor(void *data, u32 size, struct nvkm_notify *notify)
+nouveau_gpio_intr_ctor(struct nouveau_object *object, void *data, u32 size,
+		       struct nvkm_notify *notify)
 {
 	struct nvkm_gpio_ntfy_req *req = data;
 	if (!WARN_ON(size != sizeof(*req))) {
diff --git a/drivers/gpu/drm/nouveau/core/subdev/gpio/nv92.c b/drivers/gpu/drm/nouveau/core/subdev/gpio/nv94.c
index 252083d376f5..cae404ccadac 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/gpio/nv92.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/gpio/nv94.c
@@ -25,7 +25,7 @@
 #include "priv.h"
 
 void
-nv92_gpio_intr_stat(struct nouveau_gpio *gpio, u32 *hi, u32 *lo)
+nv94_gpio_intr_stat(struct nouveau_gpio *gpio, u32 *hi, u32 *lo)
 {
 	u32 intr0 = nv_rd32(gpio, 0x00e054);
 	u32 intr1 = nv_rd32(gpio, 0x00e074);
@@ -38,7 +38,7 @@ nv92_gpio_intr_stat(struct nouveau_gpio *gpio, u32 *hi, u32 *lo)
 }
 
 void
-nv92_gpio_intr_mask(struct nouveau_gpio *gpio, u32 type, u32 mask, u32 data)
+nv94_gpio_intr_mask(struct nouveau_gpio *gpio, u32 type, u32 mask, u32 data)
 {
 	u32 inte0 = nv_rd32(gpio, 0x00e050);
 	u32 inte1 = nv_rd32(gpio, 0x00e070);
@@ -57,8 +57,8 @@ nv92_gpio_intr_mask(struct nouveau_gpio *gpio, u32 type, u32 mask, u32 data)
 }
 
 struct nouveau_oclass *
-nv92_gpio_oclass = &(struct nouveau_gpio_impl) {
-	.base.handle = NV_SUBDEV(GPIO, 0x92),
+nv94_gpio_oclass = &(struct nouveau_gpio_impl) {
+	.base.handle = NV_SUBDEV(GPIO, 0x94),
 	.base.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = _nouveau_gpio_ctor,
 		.dtor = _nouveau_gpio_dtor,
@@ -66,8 +66,8 @@ nv92_gpio_oclass = &(struct nouveau_gpio_impl) {
 		.fini = _nouveau_gpio_fini,
 	},
 	.lines = 32,
-	.intr_stat = nv92_gpio_intr_stat,
-	.intr_mask = nv92_gpio_intr_mask,
+	.intr_stat = nv94_gpio_intr_stat,
+	.intr_mask = nv94_gpio_intr_mask,
 	.drive = nv50_gpio_drive,
 	.sense = nv50_gpio_sense,
 	.reset = nv50_gpio_reset,
diff --git a/drivers/gpu/drm/nouveau/core/subdev/gpio/nvd0.c b/drivers/gpu/drm/nouveau/core/subdev/gpio/nvd0.c
index a4682b0956ad..480d6d2af770 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/gpio/nvd0.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/gpio/nvd0.c
@@ -77,8 +77,8 @@ nvd0_gpio_oclass = &(struct nouveau_gpio_impl) {
 		.fini = _nouveau_gpio_fini,
 	},
 	.lines = 32,
-	.intr_stat = nv92_gpio_intr_stat,
-	.intr_mask = nv92_gpio_intr_mask,
+	.intr_stat = nv94_gpio_intr_stat,
+	.intr_mask = nv94_gpio_intr_mask,
 	.drive = nvd0_gpio_drive,
 	.sense = nvd0_gpio_sense,
 	.reset = nvd0_gpio_reset,
diff --git a/drivers/gpu/drm/nouveau/core/subdev/gpio/priv.h b/drivers/gpu/drm/nouveau/core/subdev/gpio/priv.h
index e1724dfc86ae..bff98b86e2b5 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/gpio/priv.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/gpio/priv.h
@@ -56,8 +56,8 @@ void nv50_gpio_reset(struct nouveau_gpio *, u8);
 int  nv50_gpio_drive(struct nouveau_gpio *, int, int, int);
 int  nv50_gpio_sense(struct nouveau_gpio *, int);
 
-void nv92_gpio_intr_stat(struct nouveau_gpio *, u32 *, u32 *);
-void nv92_gpio_intr_mask(struct nouveau_gpio *, u32, u32, u32);
+void nv94_gpio_intr_stat(struct nouveau_gpio *, u32 *, u32 *);
+void nv94_gpio_intr_mask(struct nouveau_gpio *, u32, u32, u32);
 
 void nvd0_gpio_reset(struct nouveau_gpio *, u8);
 int  nvd0_gpio_drive(struct nouveau_gpio *, int, int, int);
diff --git a/drivers/gpu/drm/nouveau/core/subdev/i2c/base.c b/drivers/gpu/drm/nouveau/core/subdev/i2c/base.c
index a652cafde3d6..2b1bf545e488 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/i2c/base.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/i2c/base.c
@@ -23,6 +23,7 @@
  */
 
 #include <core/option.h>
+#include <core/object.h>
 #include <core/event.h>
 
 #include <subdev/bios.h>
@@ -346,7 +347,8 @@ nouveau_i2c_intr_init(struct nvkm_event *event, int type, int index)
 }
 
 static int
-nouveau_i2c_intr_ctor(void *data, u32 size, struct nvkm_notify *notify)
+nouveau_i2c_intr_ctor(struct nouveau_object *object, void *data, u32 size,
+		      struct nvkm_notify *notify)
 {
 	struct nvkm_i2c_ntfy_req *req = data;
 	if (!WARN_ON(size != sizeof(*req))) {
diff --git a/drivers/gpu/drm/nouveau/core/subdev/instmem/nv04.c b/drivers/gpu/drm/nouveau/core/subdev/instmem/nv04.c
index 7b64befee48f..e8b1401c59c0 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/instmem/nv04.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/instmem/nv04.c
@@ -69,7 +69,7 @@ nv04_instobj_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	if (ret)
 		return ret;
 
-	ret = nouveau_mm_head(&priv->heap, 1, args->size, args->size,
+	ret = nouveau_mm_head(&priv->heap, 0, 1, args->size, args->size,
 			      args->align, &node->mem);
 	if (ret)
 		return ret;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/ltc/base.c b/drivers/gpu/drm/nouveau/core/subdev/ltc/base.c
index 32ed442c5913..7fa331516f84 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/ltc/base.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/ltc/base.c
@@ -31,7 +31,7 @@ nvkm_ltc_tags_alloc(struct nouveau_ltc *ltc, u32 n,
 	struct nvkm_ltc_priv *priv = (void *)ltc;
 	int ret;
 
-	ret = nouveau_mm_head(&priv->tags, 1, n, n, 1, pnode);
+	ret = nouveau_mm_head(&priv->tags, 0, 1, n, n, 1, pnode);
 	if (ret)
 		*pnode = NULL;
 
diff --git a/drivers/gpu/drm/nouveau/core/subdev/ltc/gf100.c b/drivers/gpu/drm/nouveau/core/subdev/ltc/gf100.c
index d5d65285efe5..2db0977284f8 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/ltc/gf100.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/ltc/gf100.c
@@ -62,16 +62,38 @@ gf100_ltc_zbc_clear_depth(struct nvkm_ltc_priv *priv, int i, const u32 depth)
 	nv_wr32(priv, 0x17ea58, depth);
 }
 
+static const struct nouveau_bitfield
+gf100_ltc_lts_intr_name[] = {
+	{ 0x00000001, "IDLE_ERROR_IQ" },
+	{ 0x00000002, "IDLE_ERROR_CBC" },
+	{ 0x00000004, "IDLE_ERROR_TSTG" },
+	{ 0x00000008, "IDLE_ERROR_DSTG" },
+	{ 0x00000010, "EVICTED_CB" },
+	{ 0x00000020, "ILLEGAL_COMPSTAT" },
+	{ 0x00000040, "BLOCKLINEAR_CB" },
+	{ 0x00000100, "ECC_SEC_ERROR" },
+	{ 0x00000200, "ECC_DED_ERROR" },
+	{ 0x00000400, "DEBUG" },
+	{ 0x00000800, "ATOMIC_TO_Z" },
+	{ 0x00001000, "ILLEGAL_ATOMIC" },
+	{ 0x00002000, "BLKACTIVITY_ERR" },
+	{}
+};
+
 static void
-gf100_ltc_lts_isr(struct nvkm_ltc_priv *priv, int ltc, int lts)
+gf100_ltc_lts_intr(struct nvkm_ltc_priv *priv, int ltc, int lts)
 {
 	u32 base = 0x141000 + (ltc * 0x2000) + (lts * 0x400);
-	u32 stat = nv_rd32(priv, base + 0x020);
+	u32 intr = nv_rd32(priv, base + 0x020);
+	u32 stat = intr & 0x0000ffff;
 
 	if (stat) {
-		nv_info(priv, "LTC%d_LTS%d: 0x%08x\n", ltc, lts, stat);
-		nv_wr32(priv, base + 0x020, stat);
+		nv_info(priv, "LTC%d_LTS%d:", ltc, lts);
+		nouveau_bitfield_print(gf100_ltc_lts_intr_name, stat);
+		pr_cont("\n");
 	}
+
+	nv_wr32(priv, base + 0x020, intr);
 }
 
 void
@@ -84,14 +106,9 @@ gf100_ltc_intr(struct nouveau_subdev *subdev)
 	while (mask) {
 		u32 lts, ltc = __ffs(mask);
 		for (lts = 0; lts < priv->lts_nr; lts++)
-			gf100_ltc_lts_isr(priv, ltc, lts);
+			gf100_ltc_lts_intr(priv, ltc, lts);
 		mask &= ~(1 << ltc);
 	}
-
-	/* we do something horribly wrong and upset PMFB a lot, so mask off
-	 * interrupts from it after the first one until it's fixed
-	 */
-	nv_mask(priv, 0x000640, 0x02000000, 0x00000000);
 }
 
 static int
@@ -153,7 +170,7 @@ gf100_ltc_init_tag_ram(struct nouveau_fb *pfb, struct nvkm_ltc_priv *priv)
 	tag_size += tag_align;
 	tag_size  = (tag_size + 0xfff) >> 12; /* round up */
 
-	ret = nouveau_mm_tail(&pfb->vram, 1, tag_size, tag_size, 1,
+	ret = nouveau_mm_tail(&pfb->vram, 1, 1, tag_size, tag_size, 1,
 	                      &priv->tag_ram);
 	if (ret) {
 		priv->num_tags = 0;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/ltc/gm107.c b/drivers/gpu/drm/nouveau/core/subdev/ltc/gm107.c
index a4de64289762..89fc4238f50c 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/ltc/gm107.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/ltc/gm107.c
@@ -87,11 +87,6 @@ gm107_ltc_intr(struct nouveau_subdev *subdev)
 			gm107_ltc_lts_isr(priv, ltc, lts);
 		mask &= ~(1 << ltc);
 	}
-
-	/* we do something horribly wrong and upset PMFB a lot, so mask off
-	 * interrupts from it after the first one until it's fixed
-	 */
-	nv_mask(priv, 0x000640, 0x02000000, 0x00000000);
 }
 
 static int
diff --git a/drivers/gpu/drm/nouveau/core/subdev/ltc/priv.h b/drivers/gpu/drm/nouveau/core/subdev/ltc/priv.h
index 594924f39126..41f179d93da6 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/ltc/priv.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/ltc/priv.h
@@ -4,6 +4,8 @@
 #include <subdev/ltc.h>
 #include <subdev/fb.h>
 
+#include <core/enum.h>
+
 struct nvkm_ltc_priv {
 	struct nouveau_ltc base;
 	u32 ltc_nr;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/base.c b/drivers/gpu/drm/nouveau/core/subdev/pwr/base.c
index 69f1f34f6931..0ab55f27ec45 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/base.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/base.c
@@ -203,6 +203,8 @@ _nouveau_pwr_init(struct nouveau_object *object)
 	nv_wait(ppwr, 0x10a04c, 0xffffffff, 0x00000000);
 	nv_mask(ppwr, 0x000200, 0x00002000, 0x00000000);
 	nv_mask(ppwr, 0x000200, 0x00002000, 0x00002000);
+	nv_rd32(ppwr, 0x000200);
+	nv_wait(ppwr, 0x10a10c, 0x00000006, 0x00000000);
 
 	/* upload data segment */
 	nv_wr32(ppwr, 0x10a1c0, 0x01000000);
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/arith.fuc b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/arith.fuc
new file mode 100644
index 000000000000..214a6d9e088d
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/arith.fuc
@@ -0,0 +1,94 @@
+/*
+ * Copyright 2014 Martin Peres <martin.peres@free.fr>
+ *
+ * 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 folloing 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: Martin Peres
+ */
+
+/******************************************************************************
+ * arith data segment
+ *****************************************************************************/
+#ifdef INCLUDE_PROC
+#endif
+
+#ifdef INCLUDE_DATA
+#endif
+
+/******************************************************************************
+ * arith code segment
+ *****************************************************************************/
+#ifdef INCLUDE_CODE
+
+// does a 32x32 -> 64 multiplication
+//
+// A * B = A_lo * B_lo
+//        + ( A_hi * B_lo ) << 16
+//        + ( A_lo * B_hi ) << 16
+//        + ( A_hi * B_hi ) << 32
+//
+// $r15 - current
+// $r14 - A
+// $r13 - B
+// $r12 - mul_lo (return)
+// $r11 - mul_hi (return)
+// $r0  - zero
+mulu32_32_64:
+	push $r1 // A_hi
+	push $r2 // B_hi
+	push $r3 // tmp0
+	push $r4 // tmp1
+
+	shr b32 $r1 $r14 16
+	shr b32 $r2 $r13 16
+
+	clear b32 $r12
+	clear b32 $r11
+
+	// A_lo * B_lo
+	mulu $r12 $r14 $r13
+
+	// ( A_hi * B_lo ) << 16
+	mulu $r3 $r1 $r13 // tmp0 = A_hi * B_lo
+	mov b32 $r4 $r3
+	and $r3 0xffff // tmp0 = tmp0_lo
+	shl b32 $r3 16
+	shr b32 $r4 16 // tmp1 = tmp0_hi
+	add b32 $r12 $r3
+	adc b32 $r11 $r4
+
+	// ( A_lo * B_hi ) << 16
+	mulu $r3 $r14 $r2 // tmp0 = A_lo * B_hi
+	mov b32 $r4 $r3
+	and $r3 0xffff // tmp0 = tmp0_lo
+	shl b32 $r3 16
+	shr b32 $r4 16 // tmp1 = tmp0_hi
+	add b32 $r12 $r3
+	adc b32 $r11 $r4
+
+	// ( A_hi * B_hi ) << 32
+	mulu $r3 $r1 $r2 // tmp0 = A_hi * B_hi
+	add b32 $r11 $r3
+
+	pop $r4
+	pop $r3
+	pop $r2
+	pop $r1
+	ret
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/kernel.fuc b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/kernel.fuc
index 8f29badd785f..5cf5be63cbef 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/kernel.fuc
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/kernel.fuc
@@ -98,12 +98,16 @@ wr32:
 // $r14 - ns
 // $r0  - zero
 nsec:
+	push $r9
+	push $r8
 	nv_iord($r8, NV_PPWR_TIMER_LOW)
 	nsec_loop:
 		nv_iord($r9, NV_PPWR_TIMER_LOW)
 		sub b32 $r9 $r8
 		cmp b32 $r9 $r14
 		bra l #nsec_loop
+	pop $r8
+	pop $r9
 	ret
 
 // busy-wait for a period of time
@@ -115,6 +119,8 @@ nsec:
 // $r11 - timeout (ns)
 // $r0  - zero
 wait:
+	push $r9
+	push $r8
 	nv_iord($r8, NV_PPWR_TIMER_LOW)
 	wait_loop:
 		nv_rd32($r10, $r14)
@@ -126,6 +132,8 @@ wait:
 		cmp b32 $r9 $r11
 		bra l #wait_loop
 	wait_done:
+	pop $r8
+	pop $r9
 	ret
 
 // $r15 - current (kern)
@@ -242,12 +250,89 @@ intr:
 	bclr $flags $p0
 	iret
 
-// request the current process be sent a message after a timeout expires
+// calculate the number of ticks in the specified nanoseconds delay
+//
+// $r15 - current
+// $r14 - ns
+// $r14 - ticks (return)
+// $r0  - zero
+ticks_from_ns:
+	push $r12
+	push $r11
+
+	/* try not losing precision (multiply then divide) */
+	imm32($r13, HW_TICKS_PER_US)
+	call #mulu32_32_64
+
+	/* use an immeditate, it's ok because HW_TICKS_PER_US < 16 bits */
+	div $r12 $r12 1000
+
+	/* check if there wasn't any overflow */
+	cmpu b32 $r11 0
+	bra e #ticks_from_ns_quit
+
+	/* let's divide then multiply, too bad for the precision! */
+	div $r14 $r14 1000
+	imm32($r13, HW_TICKS_PER_US)
+	call #mulu32_32_64
+
+	/* this cannot overflow as long as HW_TICKS_PER_US < 1000 */
+
+ticks_from_ns_quit:
+	mov b32 $r14 $r12
+	pop $r11
+	pop $r12
+	ret
+
+// calculate the number of ticks in the specified microsecond delay
+//
+// $r15 - current
+// $r14 - us
+// $r14 - ticks (return)
+// $r0  - zero
+ticks_from_us:
+	push $r12
+	push $r11
+
+	/* simply multiply $us by HW_TICKS_PER_US */
+	imm32($r13, HW_TICKS_PER_US)
+	call #mulu32_32_64
+	mov b32 $r14 $r12
+
+	/* check if there wasn't any overflow */
+	cmpu b32 $r11 0
+	bra e #ticks_from_us_quit
+
+	/* Overflow! */
+	clear b32 $r14
+
+ticks_from_us_quit:
+	pop $r11
+	pop $r12
+	ret
+
+// calculate the number of ticks in the specified microsecond delay
 //
 // $r15 - current
 // $r14 - ticks
+// $r14 - us (return)
+// $r0  - zero
+ticks_to_us:
+	/* simply divide $ticks by HW_TICKS_PER_US */
+	imm32($r13, HW_TICKS_PER_US)
+	div $r14 $r14 $r13
+
+	ret
+
+// request the current process be sent a message after a timeout expires
+//
+// $r15 - current
+// $r14 - ticks (make sure it is < 2^31 to avoid any possible overflow)
 // $r0  - zero
 timer:
+	push $r9
+	push $r8
+
 	// interrupts off to prevent racing with timer isr
 	bclr $flags ie0
 
@@ -255,13 +340,22 @@ timer:
 	ld b32 $r8 D[$r15 + #proc_time]
 	cmp b32 $r8 0
 	bra g #timer_done
-	st b32 D[$r15 + #proc_time] $r14
 
-	// halt watchdog timer temporarily and check for a pending
-	// interrupt.  if there's one already pending, we can just
-	// bail since the timer isr will queue the next soonest
-	// right after it's done
+	// halt watchdog timer temporarily
+	clear b32 $r8
 	nv_iowr(NV_PPWR_WATCHDOG_ENABLE, $r8)
+
+	// find out how much time elapsed since the last update
+	// of the watchdog and add this time to the wanted ticks
+	nv_iord($r8, NV_PPWR_WATCHDOG_TIME)
+	ld b32 $r9 D[$r0 + #time_prev]
+	sub b32 $r9 $r8
+	add b32 $r14 $r9
+	st b32 D[$r15 + #proc_time] $r14
+
+	// check for a pending interrupt.  if there's one already
+	// pending, we can just bail since the timer isr will
+	// queue the next soonest right after it's done
 	nv_iord($r8, NV_PPWR_INTR)
 	and $r8 NV_PPWR_INTR_WATCHDOG
 	bra nz #timer_enable
@@ -272,10 +366,10 @@ timer:
 	cmp b32 $r14 $r0
 	bra e #timer_reset
 	cmp b32 $r14 $r8
-	bra l #timer_done
-	timer_reset:
-	nv_iowr(NV_PPWR_WATCHDOG_TIME, $r14)
-	st b32 D[$r0 + #time_prev] $r14
+	bra g #timer_enable
+		timer_reset:
+		nv_iowr(NV_PPWR_WATCHDOG_TIME, $r14)
+		st b32 D[$r0 + #time_prev] $r14
 
 	// re-enable the watchdog timer
 	timer_enable:
@@ -285,6 +379,9 @@ timer:
 	// interrupts back on
 	timer_done:
 	bset $flags ie0
+
+	pop $r8
+	pop $r9
 	ret
 
 // send message to another process
@@ -371,6 +468,9 @@ send:
 // $r14 - process
 // $r0  - zero
 recv:
+	push $r9
+	push $r8
+
 	ld b32 $r8 D[$r14 + #proc_qget]
 	ld b32 $r9 D[$r14 + #proc_qput]
 	bclr $flags $p1
@@ -403,6 +503,8 @@ recv:
 		bset $flags $p1
 		pop $r15
 	recv_done:
+	pop $r8
+	pop $r9
 	ret
 
 init:
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/macros.fuc b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/macros.fuc
index 5668e045bac1..96fc984dafdc 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/macros.fuc
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/macros.fuc
@@ -250,3 +250,23 @@
 */	st b32 D[$r0] reg /*
 */	clear b32 $r0
 #endif
+
+#define st(size, addr, reg) /*
+*/	movw $r0 addr /*
+*/	st size D[$r0] reg /*
+*/	clear b32 $r0
+
+#define ld(size, reg, addr) /*
+*/	movw $r0 addr /*
+*/	ld size reg D[$r0] /*
+*/	clear b32 $r0
+
+// does a 64+64 -> 64 unsigned addition (C = A + B)
+#define addu64(reg_a_c_hi, reg_a_c_lo, b_hi, b_lo) /*
+*/    add b32 reg_a_c_lo b_lo /*
+*/    adc b32 reg_a_c_hi b_hi
+
+// does a 64+64 -> 64 substraction (C = A - B)
+#define subu64(reg_a_c_hi, reg_a_c_lo, b_hi, b_lo) /*
+*/    sub b32 reg_a_c_lo b_lo /*
+*/    sbb b32 reg_a_c_hi b_hi
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/memx.fuc b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/memx.fuc
index d43741eccb11..e89789a53b80 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/memx.fuc
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/memx.fuc
@@ -43,17 +43,23 @@ process(PROC_MEMX, #memx_init, #memx_recv)
 */	.b32 func
 
 memx_func_head:
-handler(ENTER , 0x0001, 0x0000, #memx_func_enter)
+handler(ENTER , 0x0000, 0x0000, #memx_func_enter)
 memx_func_next:
 handler(LEAVE , 0x0000, 0x0000, #memx_func_leave)
 handler(WR32  , 0x0000, 0x0002, #memx_func_wr32)
 handler(WAIT  , 0x0004, 0x0000, #memx_func_wait)
 handler(DELAY , 0x0001, 0x0000, #memx_func_delay)
+handler(VBLANK, 0x0001, 0x0000, #memx_func_wait_vblank)
 memx_func_tail:
 
 .equ #memx_func_size #memx_func_next - #memx_func_head
 .equ #memx_func_num (#memx_func_tail - #memx_func_head) / #memx_func_size
 
+memx_ts_start:
+.b32 0
+memx_ts_end:
+.b32 0
+
 memx_data_head:
 .skip 0x0800
 memx_data_tail:
@@ -67,19 +73,44 @@ memx_data_tail:
 //
 // $r15 - current (memx)
 // $r4  - packet length
-//	+00: bitmask of heads to wait for vblank on
 // $r3  - opcode desciption
 // $r0  - zero
 memx_func_enter:
+#if NVKM_PPWR_CHIPSET == GT215
+	movw $r8 0x1610
+	nv_rd32($r7, $r8)
+	imm32($r6, 0xfffffffc)
+	and $r7 $r6
+	movw $r6 0x2
+	or $r7 $r6
+	nv_wr32($r8, $r7)
+#else
+	movw $r6 0x001620
+	imm32($r7, ~0x00000aa2);
+	nv_rd32($r8, $r6)
+	and $r8 $r7
+	nv_wr32($r6, $r8)
+
+	imm32($r7, ~0x00000001)
+	nv_rd32($r8, $r6)
+	and $r8 $r7
+	nv_wr32($r6, $r8)
+
+	movw $r6 0x0026f0
+	nv_rd32($r8, $r6)
+	and $r8 $r7
+	nv_wr32($r6, $r8)
+#endif
+
 	mov $r6 NV_PPWR_OUTPUT_SET_FB_PAUSE
 	nv_iowr(NV_PPWR_OUTPUT_SET, $r6)
 	memx_func_enter_wait:
 		nv_iord($r6, NV_PPWR_OUTPUT)
 		and $r6 NV_PPWR_OUTPUT_FB_PAUSE
 		bra z #memx_func_enter_wait
-	//XXX: TODO
-	ld b32 $r6 D[$r1 + 0x00]
-	add b32 $r1 0x04
+
+	nv_iord($r6, NV_PPWR_TIMER_LOW)
+	st b32 D[$r0 + #memx_ts_start] $r6
 	ret
 
 // description
@@ -89,14 +120,93 @@ memx_func_enter:
 // $r3  - opcode desciption
 // $r0  - zero
 memx_func_leave:
+	nv_iord($r6, NV_PPWR_TIMER_LOW)
+	st b32 D[$r0 + #memx_ts_end] $r6
+
 	mov $r6 NV_PPWR_OUTPUT_CLR_FB_PAUSE
 	nv_iowr(NV_PPWR_OUTPUT_CLR, $r6)
 	memx_func_leave_wait:
 		nv_iord($r6, NV_PPWR_OUTPUT)
 		and $r6 NV_PPWR_OUTPUT_FB_PAUSE
 		bra nz #memx_func_leave_wait
+
+#if NVKM_PPWR_CHIPSET == GT215
+	movw $r8 0x1610
+	nv_rd32($r7, $r8)
+	imm32($r6, 0xffffffcc)
+	and $r7 $r6
+	nv_wr32($r8, $r7)
+#else
+	movw $r6 0x0026f0
+	imm32($r7, 0x00000001)
+	nv_rd32($r8, $r6)
+	or $r8 $r7
+	nv_wr32($r6, $r8)
+
+	movw $r6 0x001620
+	nv_rd32($r8, $r6)
+	or $r8 $r7
+	nv_wr32($r6, $r8)
+
+	imm32($r7, 0x00000aa2);
+	nv_rd32($r8, $r6)
+	or $r8 $r7
+	nv_wr32($r6, $r8)
+#endif
+	ret
+
+#if NVKM_PPWR_CHIPSET < GF119
+// description
+//
+// $r15 - current (memx)
+// $r4  - packet length
+//	+00: head to wait for vblank on
+// $r3  - opcode desciption
+// $r0  - zero
+memx_func_wait_vblank:
+	ld b32 $r6 D[$r1 + 0x00]
+	cmp b32 $r6 0x0
+	bra z #memx_func_wait_vblank_head0
+	cmp b32 $r6 0x1
+	bra z #memx_func_wait_vblank_head1
+	bra #memx_func_wait_vblank_fini
+
+	memx_func_wait_vblank_head1:
+	movw $r7 0x20
+	bra #memx_func_wait_vblank_0
+
+	memx_func_wait_vblank_head0:
+	movw $r7 0x8
+
+	memx_func_wait_vblank_0:
+		nv_iord($r6, NV_PPWR_INPUT)
+		and $r6 $r7
+		bra nz #memx_func_wait_vblank_0
+
+	memx_func_wait_vblank_1:
+		nv_iord($r6, NV_PPWR_INPUT)
+		and $r6 $r7
+		bra z #memx_func_wait_vblank_1
+
+	memx_func_wait_vblank_fini:
+	add b32 $r1 0x4
+	ret
+
+#else
+
+// XXX: currently no-op
+//
+// $r15 - current (memx)
+// $r4  - packet length
+//	+00: head to wait for vblank on
+// $r3  - opcode desciption
+// $r0  - zero
+memx_func_wait_vblank:
+	add b32 $r1 0x4
 	ret
 
+#endif
+
 // description
 //
 // $r15 - current (memx)
@@ -160,14 +270,17 @@ memx_exec:
 	push $r13
 	mov b32 $r1 $r12
 	mov b32 $r2 $r11
+
 	memx_exec_next:
-		// fetch the packet header, and locate opcode info
+		// fetch the packet header
 		ld b32 $r3 D[$r1]
 		add b32 $r1 4
-		shr b32 $r4 $r3 16
-		mulu $r3 #memx_func_size
+		extr $r4 $r3 16:31
+		extr $r3 $r3 0:15
 
 		// execute the opcode handler
+		sub b32 $r3 1
+		mulu $r3 #memx_func_size
 		ld b32 $r5 D[$r3 + #memx_func_head + #memx_func]
 		call $r5
 
@@ -176,6 +289,10 @@ memx_exec:
 		bra l #memx_exec_next
 
 	// send completion reply
+	ld b32 $r11 D[$r0 + #memx_ts_start]
+	ld b32 $r12 D[$r0 + #memx_ts_end]
+	sub b32 $r12 $r11
+	nv_iord($r11, NV_PPWR_INPUT)
 	pop $r13
 	pop $r14
 	call(send)
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nv108.fuc b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nv108.fuc
index 17a8a383d91a..b439519ec866 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nv108.fuc
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nv108.fuc
@@ -23,6 +23,7 @@
  */
 
 #define NVKM_PPWR_CHIPSET GK208
+#define HW_TICKS_PER_US 324
 
 #define NVKM_FALCON_PC24
 #define NVKM_FALCON_UNSHIFTED_IO
@@ -34,6 +35,7 @@
 .section #nv108_pwr_data
 #define INCLUDE_PROC
 #include "kernel.fuc"
+#include "arith.fuc"
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
@@ -44,6 +46,7 @@
 
 #define INCLUDE_DATA
 #include "kernel.fuc"
+#include "arith.fuc"
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
@@ -56,6 +59,7 @@
 .section #nv108_pwr_code
 #define INCLUDE_CODE
 #include "kernel.fuc"
+#include "arith.fuc"
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nv108.fuc.h b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nv108.fuc.h
index 986495d533dd..4d278a96b2bb 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nv108.fuc.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nv108.fuc.h
@@ -24,8 +24,8 @@ uint32_t nv108_pwr_data[] = {
 	0x00000000,
 /* 0x0058: proc_list_head */
 	0x54534f48,
-	0x00000379,
-	0x0000032a,
+	0x00000453,
+	0x00000404,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -46,8 +46,8 @@ uint32_t nv108_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x584d454d,
-	0x00000464,
-	0x00000456,
+	0x0000061c,
+	0x0000060e,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -68,8 +68,8 @@ uint32_t nv108_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x46524550,
-	0x00000468,
-	0x00000466,
+	0x00000620,
+	0x0000061e,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -90,8 +90,8 @@ uint32_t nv108_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x5f433249,
-	0x0000086c,
-	0x00000713,
+	0x00000a24,
+	0x000008cb,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -112,8 +112,8 @@ uint32_t nv108_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x54534554,
-	0x0000088d,
-	0x0000086e,
+	0x00000a45,
+	0x00000a26,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -134,8 +134,8 @@ uint32_t nv108_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x454c4449,
-	0x00000898,
-	0x00000896,
+	0x00000a50,
+	0x00000a4e,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -227,25 +227,31 @@ uint32_t nv108_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 /* 0x0370: memx_func_head */
-	0x00010000,
-	0x00000000,
-	0x000003a9,
-/* 0x037c: memx_func_next */
 	0x00000001,
 	0x00000000,
-	0x000003c7,
+	0x00000483,
+/* 0x037c: memx_func_next */
 	0x00000002,
+	0x00000000,
+	0x00000500,
+	0x00000003,
 	0x00000002,
-	0x000003df,
-	0x00040003,
+	0x00000580,
+	0x00040004,
+	0x00000000,
+	0x0000059d,
+	0x00010005,
+	0x00000000,
+	0x000005b7,
+	0x00010006,
 	0x00000000,
-	0x000003fc,
-	0x00010004,
+	0x0000057b,
+/* 0x03b8: memx_func_tail */
+/* 0x03b8: memx_ts_start */
 	0x00000000,
-	0x00000416,
-/* 0x03ac: memx_func_tail */
-/* 0x03ac: memx_data_head */
+/* 0x03bc: memx_ts_end */
 	0x00000000,
+/* 0x03c0: memx_data_head */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -757,8 +763,9 @@ uint32_t nv108_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-/* 0x0bac: memx_data_tail */
-/* 0x0bac: i2c_scl_map */
+	0x00000000,
+/* 0x0bc0: memx_data_tail */
+/* 0x0bc0: i2c_scl_map */
 	0x00000400,
 	0x00000800,
 	0x00001000,
@@ -769,7 +776,7 @@ uint32_t nv108_pwr_data[] = {
 	0x00020000,
 	0x00040000,
 	0x00080000,
-/* 0x0bd4: i2c_sda_map */
+/* 0x0be8: i2c_sda_map */
 	0x00100000,
 	0x00200000,
 	0x00400000,
@@ -781,10 +788,69 @@ uint32_t nv108_pwr_data[] = {
 	0x10000000,
 	0x20000000,
 	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
 };
 
 uint32_t nv108_pwr_code[] = {
-	0x02910ef5,
+	0x031c0ef5,
 /* 0x0004: rd32 */
 	0xf607a040,
 	0x04bd000e,
@@ -812,15 +878,18 @@ uint32_t nv108_pwr_code[] = {
 	0x7000d4f1,
 	0xf8f61bf4,
 /* 0x005d: nsec */
-	0xcf2c0800,
-/* 0x0062: nsec_loop */
+	0xf990f900,
+	0xcf2c0880,
+/* 0x0066: nsec_loop */
 	0x2c090088,
 	0xbb0099cf,
 	0x9ea60298,
-	0xf8f61ef4,
-/* 0x0071: wait */
-	0xcf2c0800,
-/* 0x0076: wait_loop */
+	0xfcf61ef4,
+	0xf890fc80,
+/* 0x0079: wait */
+	0xf990f900,
+	0xcf2c0880,
+/* 0x0082: wait_loop */
 	0xeeb20088,
 	0x0000047e,
 	0xadfddab2,
@@ -828,28 +897,29 @@ uint32_t nv108_pwr_code[] = {
 	0x2c09100b,
 	0xbb0099cf,
 	0x9ba60298,
-/* 0x0093: wait_done */
-	0xf8e61ef4,
-/* 0x0095: intr_watchdog */
+/* 0x009f: wait_done */
+	0xfce61ef4,
+	0xf890fc80,
+/* 0x00a5: intr_watchdog */
 	0x03e99800,
 	0xf40096b0,
 	0x0a98280b,
 	0x029abb9a,
 	0x0d0e1cf4,
-	0x01de7e01,
+	0x02617e01,
 	0xf494bd00,
-/* 0x00b2: intr_watchdog_next_time */
+/* 0x00c2: intr_watchdog_next_time */
 	0x0a98140e,
 	0x00a6b09b,
 	0xa6080bf4,
 	0x061cf49a,
-/* 0x00c0: intr_watchdog_next_time_set */
-/* 0x00c3: intr_watchdog_next_proc */
+/* 0x00d0: intr_watchdog_next_time_set */
+/* 0x00d3: intr_watchdog_next_proc */
 	0xb59b09b5,
 	0xe0b603e9,
 	0x68e6b158,
 	0xc81bf402,
-/* 0x00d2: intr */
+/* 0x00e2: intr */
 	0x00f900f8,
 	0x80f904bd,
 	0xa0f990f9,
@@ -865,13 +935,13 @@ uint32_t nv108_pwr_code[] = {
 	0xc40088cf,
 	0x0bf40289,
 	0x9b00b51f,
-	0x957e580e,
+	0xa57e580e,
 	0x09980000,
 	0x0096b09b,
 	0x000d0bf4,
 	0x0009f634,
 	0x09b504bd,
-/* 0x0125: intr_skip_watchdog */
+/* 0x0135: intr_skip_watchdog */
 	0x0089e49a,
 	0x360bf408,
 	0xcf068849,
@@ -881,20 +951,20 @@ uint32_t nv108_pwr_code[] = {
 	0xc0f900cc,
 	0xf14f484e,
 	0x0d5453e3,
-	0x023f7e00,
+	0x02c27e00,
 	0x40c0fc00,
 	0x0cf604c0,
-/* 0x0157: intr_subintr_skip_fifo */
+/* 0x0167: intr_subintr_skip_fifo */
 	0x4004bd00,
 	0x09f60688,
-/* 0x015f: intr_skip_subintr */
+/* 0x016f: intr_skip_subintr */
 	0xc404bd00,
 	0x0bf42089,
 	0xbfa4f107,
-/* 0x0169: intr_skip_pause */
+/* 0x0179: intr_skip_pause */
 	0x4089c4ff,
 	0xf1070bf4,
-/* 0x0173: intr_skip_user0 */
+/* 0x0183: intr_skip_user0 */
 	0x00ffbfa4,
 	0x0008f604,
 	0x80fc04bd,
@@ -904,304 +974,417 @@ uint32_t nv108_pwr_code[] = {
 	0xfca0fcb0,
 	0xfc80fc90,
 	0x0032f400,
-/* 0x0196: timer */
-	0x32f401f8,
-	0x03f89810,
-	0xf40086b0,
-	0xfeb53a1c,
-	0xf6380003,
+/* 0x01a6: ticks_from_ns */
+	0xc0f901f8,
+	0xd7f1b0f9,
+	0xd3f00144,
+	0x7721f500,
+	0xe8ccec03,
+	0x00b4b003,
+	0xec120bf4,
+	0xf103e8ee,
+	0xf00144d7,
+	0x21f500d3,
+/* 0x01ce: ticks_from_ns_quit */
+	0xceb20377,
+	0xc0fcb0fc,
+/* 0x01d6: ticks_from_us */
+	0xc0f900f8,
+	0xd7f1b0f9,
+	0xd3f00144,
+	0x7721f500,
+	0xb0ceb203,
+	0x0bf400b4,
+/* 0x01ef: ticks_from_us_quit */
+	0xfce4bd05,
+	0xf8c0fcb0,
+/* 0x01f5: ticks_to_us */
+	0x44d7f100,
+	0x00d3f001,
+	0xf8ecedff,
+/* 0x0201: timer */
+	0xf990f900,
+	0x1032f480,
+	0xb003f898,
+	0x1cf40086,
+	0x0084bd4a,
+	0x0008f638,
+	0x340804bd,
+	0x980088cf,
+	0x98bb9a09,
+	0x00e9bb02,
+	0x0803feb5,
+	0x0088cf08,
+	0xf40284f0,
+	0x34081c1b,
+	0xa60088cf,
+	0x080bf4e0,
+	0x1cf4e8a6,
+/* 0x0245: timer_reset */
+	0xf634000d,
+	0x04bd000e,
+/* 0x024f: timer_enable */
+	0x089a0eb5,
+	0xf6380001,
 	0x04bd0008,
-	0x88cf0808,
-	0x0284f000,
-	0x081c1bf4,
-	0x0088cf34,
-	0x0bf4e0a6,
-	0xf4e8a608,
-/* 0x01c6: timer_reset */
-	0x3400161e,
-	0xbd000ef6,
-	0x9a0eb504,
-/* 0x01d0: timer_enable */
-	0x38000108,
-	0xbd0008f6,
-/* 0x01d9: timer_done */
-	0x1031f404,
-/* 0x01de: send_proc */
-	0x80f900f8,
-	0xe89890f9,
-	0x04e99805,
-	0xa60486f0,
-	0x2a0bf489,
-	0x940398c4,
-	0x80b60488,
-	0x008ebb18,
-	0xb500fa98,
-	0x8db5008a,
-	0x028cb501,
-	0xb6038bb5,
-	0x94f00190,
-	0x04e9b507,
-/* 0x0217: send_done */
-	0xfc0231f4,
-	0xf880fc90,
-/* 0x021d: find */
-	0x0880f900,
-	0x0131f458,
-/* 0x0224: find_loop */
-	0xa6008a98,
-	0x100bf4ae,
-	0xb15880b6,
-	0xf4026886,
-	0x32f4f11b,
-/* 0x0239: find_done */
-	0xfc8eb201,
-/* 0x023f: send */
-	0x7e00f880,
-	0xf400021d,
-	0x00f89b01,
-/* 0x0248: recv */
-	0x9805e898,
-	0x32f404e9,
-	0xf489a601,
-	0x89c43c0b,
-	0x0180b603,
-	0xb50784f0,
-	0xea9805e8,
-	0xfef0f902,
-	0xf0f9018f,
-	0x9994efb2,
-	0x00e9bb04,
-	0x9818e0b6,
-	0xec9803eb,
-	0x01ed9802,
-	0xf900ee98,
-	0xfef0fca5,
-	0x31f400f8,
-/* 0x028f: recv_done */
-	0xf8f0fc01,
-/* 0x0291: init */
-	0x01084100,
-	0xe70011cf,
-	0xb6010911,
-	0x14fe0814,
-	0x00e04100,
-	0x000013f0,
-	0x0001f61c,
-	0xff0104bd,
-	0x01f61400,
-	0x0104bd00,
-	0x0015f102,
-	0xf6100008,
-	0x04bd0001,
-	0xf000d241,
-	0x10fe0013,
-	0x1031f400,
-	0x38000101,
+/* 0x0258: timer_done */
+	0xfc1031f4,
+	0xf890fc80,
+/* 0x0261: send_proc */
+	0xf980f900,
+	0x05e89890,
+	0xf004e998,
+	0x89a60486,
+	0xc42a0bf4,
+	0x88940398,
+	0x1880b604,
+	0x98008ebb,
+	0x8ab500fa,
+	0x018db500,
+	0xb5028cb5,
+	0x90b6038b,
+	0x0794f001,
+	0xf404e9b5,
+/* 0x029a: send_done */
+	0x90fc0231,
+	0x00f880fc,
+/* 0x02a0: find */
+	0x580880f9,
+/* 0x02a7: find_loop */
+	0x980131f4,
+	0xaea6008a,
+	0xb6100bf4,
+	0x86b15880,
+	0x1bf40268,
+	0x0132f4f1,
+/* 0x02bc: find_done */
+	0x80fc8eb2,
+/* 0x02c2: send */
+	0xa07e00f8,
+	0x01f40002,
+/* 0x02cb: recv */
+	0xf900f89b,
+	0x9880f990,
+	0xe99805e8,
+	0x0132f404,
+	0x0bf489a6,
+	0x0389c43c,
+	0xf00180b6,
+	0xe8b50784,
+	0x02ea9805,
+	0x8ffef0f9,
+	0xb2f0f901,
+	0x049994ef,
+	0xb600e9bb,
+	0xeb9818e0,
+	0x02ec9803,
+	0x9801ed98,
+	0xa5f900ee,
+	0xf8fef0fc,
+	0x0131f400,
+/* 0x0316: recv_done */
+	0x80fcf0fc,
+	0x00f890fc,
+/* 0x031c: init */
+	0xcf010841,
+	0x11e70011,
+	0x14b60109,
+	0x0014fe08,
+	0xf000e041,
+	0x1c000013,
 	0xbd0001f6,
-/* 0x02db: init_proc */
-	0x98580f04,
-	0x16b001f1,
-	0xfa0bf400,
-	0xf0b615f9,
-	0xf20ef458,
-/* 0x02ec: host_send */
-	0xcf04b041,
-	0xa0420011,
-	0x0022cf04,
-	0x0bf412a6,
-	0x071ec42e,
-	0xb704ee94,
-	0x980270e0,
-	0xec9803eb,
-	0x01ed9802,
-	0x7e00ee98,
-	0xb600023f,
-	0x1ec40110,
-	0x04b0400f,
-	0xbd000ef6,
-	0xc70ef404,
-/* 0x0328: host_send_done */
-/* 0x032a: host_recv */
-	0x494100f8,
-	0x5413f14e,
-	0xf4e1a652,
-/* 0x0336: host_recv_wait */
-	0xcc41b90b,
-	0x0011cf04,
-	0xcf04c842,
-	0x16f00022,
-	0xf412a608,
-	0x23c4ef0b,
-	0x0434b607,
-	0x02f030b7,
-	0xb5033bb5,
-	0x3db5023c,
-	0x003eb501,
-	0xf00120b6,
-	0xc8400f24,
-	0x0002f604,
-	0x400204bd,
-	0x02f60000,
-	0xf804bd00,
-/* 0x0379: host_init */
-	0x00804100,
-	0xf11014b6,
-	0x40027015,
-	0x01f604d0,
+	0x00ff0104,
+	0x0001f614,
+	0x020104bd,
+	0x080015f1,
+	0x01f61000,
 	0x4104bd00,
+	0x13f000e2,
+	0x0010fe00,
+	0x011031f4,
+	0xf6380001,
+	0x04bd0001,
+/* 0x0366: init_proc */
+	0xf198580f,
+	0x0016b001,
+	0xf9fa0bf4,
+	0x58f0b615,
+/* 0x0377: mulu32_32_64 */
+	0xf9f20ef4,
+	0xf920f910,
+	0x9540f930,
+	0xd29510e1,
+	0xbdc4bd10,
+	0xc0edffb4,
+	0xb2301dff,
+	0xff34f134,
+	0x1034b6ff,
+	0xbb1045b6,
+	0xb4bb00c3,
+	0x30e2ff01,
+	0x34f134b2,
+	0x34b6ffff,
+	0x1045b610,
+	0xbb00c3bb,
+	0x12ff01b4,
+	0x00b3bb30,
+	0x30fc40fc,
+	0x10fc20fc,
+/* 0x03c6: host_send */
+	0xb04100f8,
+	0x0011cf04,
+	0xcf04a042,
+	0x12a60022,
+	0xc42e0bf4,
+	0xee94071e,
+	0x70e0b704,
+	0x03eb9802,
+	0x9802ec98,
+	0xee9801ed,
+	0x02c27e00,
+	0x0110b600,
+	0x400f1ec4,
+	0x0ef604b0,
+	0xf404bd00,
+/* 0x0402: host_send_done */
+	0x00f8c70e,
+/* 0x0404: host_recv */
+	0xf14e4941,
+	0xa6525413,
+	0xb90bf4e1,
+/* 0x0410: host_recv_wait */
+	0xcf04cc41,
+	0xc8420011,
+	0x0022cf04,
+	0xa60816f0,
+	0xef0bf412,
+	0xb60723c4,
+	0x30b70434,
+	0x3bb502f0,
+	0x023cb503,
+	0xb5013db5,
+	0x20b6003e,
+	0x0f24f001,
+	0xf604c840,
+	0x04bd0002,
+	0x00004002,
+	0xbd0002f6,
+/* 0x0453: host_init */
+	0x4100f804,
 	0x14b60080,
-	0xf015f110,
-	0x04dc4002,
+	0x7015f110,
+	0x04d04002,
+	0xbd0001f6,
+	0x00804104,
+	0xf11014b6,
+	0x4002f015,
+	0x01f604dc,
+	0x0104bd00,
+	0x04c44001,
 	0xbd0001f6,
-	0x40010104,
-	0x01f604c4,
-	0xf804bd00,
-/* 0x03a9: memx_func_enter */
-	0x40040600,
-	0x06f607e0,
-/* 0x03b3: memx_func_enter_wait */
-	0x4604bd00,
-	0x66cf07c0,
-	0x0464f000,
-	0x98f70bf4,
-	0x10b60016,
-/* 0x03c7: memx_func_leave */
-	0x0600f804,
-	0x07e44004,
-	0xbd0006f6,
-/* 0x03d1: memx_func_leave_wait */
-	0x07c04604,
-	0xf00066cf,
-	0x1bf40464,
-/* 0x03df: memx_func_wr32 */
-	0x9800f8f7,
-	0x15980016,
-	0x0810b601,
-	0x50f960f9,
+/* 0x0483: memx_func_enter */
+	0xf100f804,
+	0xf1162067,
+	0xf1f55d77,
+	0xb2ffff73,
+	0x00047e6e,
+	0xfdd8b200,
+	0x60f90487,
+	0xd0fc80f9,
+	0x2e7ee0fc,
+	0x77f10000,
+	0x73f1fffe,
+	0x6eb2ffff,
+	0x0000047e,
+	0x87fdd8b2,
+	0xf960f904,
+	0xfcd0fc80,
+	0x002e7ee0,
+	0xf067f100,
+	0x7e6eb226,
+	0xb2000004,
+	0x0487fdd8,
+	0x80f960f9,
 	0xe0fcd0fc,
 	0x00002e7e,
-	0xf40242b6,
-	0x00f8e81b,
-/* 0x03fc: memx_func_wait */
-	0x88cf2c08,
-	0x001e9800,
-	0x98011d98,
-	0x1b98021c,
-	0x1010b603,
-	0x0000717e,
-/* 0x0416: memx_func_delay */
-	0x1e9800f8,
-	0x0410b600,
-	0x00005d7e,
-/* 0x0422: memx_exec */
-	0xe0f900f8,
-	0xc1b2d0f9,
-/* 0x042a: memx_exec_next */
-	0x1398b2b2,
-	0x0410b600,
-	0xf0103495,
-	0x35980c30,
-	0xa655f9de,
-	0xed1ef412,
+	0xe0400406,
+	0x0006f607,
+/* 0x04ea: memx_func_enter_wait */
+	0xc04604bd,
+	0x0066cf07,
+	0xf40464f0,
+	0x2c06f70b,
+	0xb50066cf,
+	0x00f8ee06,
+/* 0x0500: memx_func_leave */
+	0x66cf2c06,
+	0xef06b500,
+	0xe4400406,
+	0x0006f607,
+/* 0x0512: memx_func_leave_wait */
+	0xc04604bd,
+	0x0066cf07,
+	0xf40464f0,
+	0x67f1f71b,
+	0x77f126f0,
+	0x73f00001,
+	0x7e6eb200,
+	0xb2000004,
+	0x0587fdd8,
+	0x80f960f9,
 	0xe0fcd0fc,
-	0x00023f7e,
-/* 0x044a: memx_info */
-	0xac4c00f8,
+	0x00002e7e,
+	0x162067f1,
+	0x047e6eb2,
+	0xd8b20000,
+	0xf90587fd,
+	0xfc80f960,
+	0x7ee0fcd0,
+	0xf100002e,
+	0xf00aa277,
+	0x6eb20073,
+	0x0000047e,
+	0x87fdd8b2,
+	0xf960f905,
+	0xfcd0fc80,
+	0x002e7ee0,
+/* 0x057b: memx_func_wait_vblank */
+	0xb600f800,
+	0x00f80410,
+/* 0x0580: memx_func_wr32 */
+	0x98001698,
+	0x10b60115,
+	0xf960f908,
+	0xfcd0fc50,
+	0x002e7ee0,
+	0x0242b600,
+	0xf8e81bf4,
+/* 0x059d: memx_func_wait */
+	0xcf2c0800,
+	0x1e980088,
+	0x011d9800,
+	0x98021c98,
+	0x10b6031b,
+	0x00797e10,
+/* 0x05b7: memx_func_delay */
+	0x9800f800,
+	0x10b6001e,
+	0x005d7e04,
+/* 0x05c3: memx_exec */
+	0xf900f800,
+	0xb2d0f9e0,
+/* 0x05cb: memx_exec_next */
+	0x98b2b2c1,
+	0x10b60013,
+	0xf034e704,
+	0xe033e701,
+	0x0132b601,
+	0x980c30f0,
+	0x55f9de35,
+	0x1ef412a6,
+	0xee0b98e5,
+	0xbbef0c98,
+	0xc44b02cb,
+	0x00bbcf07,
+	0xe0fcd0fc,
+	0x0002c27e,
+/* 0x0602: memx_info */
+	0xc04c00f8,
 	0x08004b03,
-	0x00023f7e,
-/* 0x0456: memx_recv */
+	0x0002c27e,
+/* 0x060e: memx_recv */
 	0xd6b000f8,
-	0xc90bf401,
+	0xb20bf401,
 	0xf400d6b0,
 	0x00f8eb0b,
-/* 0x0464: memx_init */
-/* 0x0466: perf_recv */
+/* 0x061c: memx_init */
+/* 0x061e: perf_recv */
 	0x00f800f8,
-/* 0x0468: perf_init */
-/* 0x046a: i2c_drive_scl */
+/* 0x0620: perf_init */
+/* 0x0622: i2c_drive_scl */
 	0x36b000f8,
 	0x0d0bf400,
 	0xf607e040,
 	0x04bd0001,
-/* 0x047a: i2c_drive_scl_lo */
+/* 0x0632: i2c_drive_scl_lo */
 	0xe44000f8,
 	0x0001f607,
 	0x00f804bd,
-/* 0x0484: i2c_drive_sda */
+/* 0x063c: i2c_drive_sda */
 	0xf40036b0,
 	0xe0400d0b,
 	0x0002f607,
 	0x00f804bd,
-/* 0x0494: i2c_drive_sda_lo */
+/* 0x064c: i2c_drive_sda_lo */
 	0xf607e440,
 	0x04bd0002,
-/* 0x049e: i2c_sense_scl */
+/* 0x0656: i2c_sense_scl */
 	0x32f400f8,
 	0x07c44301,
 	0xfd0033cf,
 	0x0bf40431,
 	0x0131f406,
-/* 0x04b0: i2c_sense_scl_done */
-/* 0x04b2: i2c_sense_sda */
+/* 0x0668: i2c_sense_scl_done */
+/* 0x066a: i2c_sense_sda */
 	0x32f400f8,
 	0x07c44301,
 	0xfd0033cf,
 	0x0bf40432,
 	0x0131f406,
-/* 0x04c4: i2c_sense_sda_done */
-/* 0x04c6: i2c_raise_scl */
+/* 0x067c: i2c_sense_sda_done */
+/* 0x067e: i2c_raise_scl */
 	0x40f900f8,
 	0x03089844,
-	0x046a7e01,
-/* 0x04d1: i2c_raise_scl_wait */
+	0x06227e01,
+/* 0x0689: i2c_raise_scl_wait */
 	0x03e84e00,
 	0x00005d7e,
-	0x00049e7e,
+	0x0006567e,
 	0xb60901f4,
 	0x1bf40142,
-/* 0x04e5: i2c_raise_scl_done */
+/* 0x069d: i2c_raise_scl_done */
 	0xf840fcef,
-/* 0x04e9: i2c_start */
-	0x049e7e00,
+/* 0x06a1: i2c_start */
+	0x06567e00,
 	0x0d11f400,
-	0x0004b27e,
+	0x00066a7e,
 	0xf40611f4,
-/* 0x04fa: i2c_start_rep */
+/* 0x06b2: i2c_start_rep */
 	0x00032e0e,
-	0x00046a7e,
-	0x847e0103,
-	0x76bb0004,
+	0x0006227e,
+	0x3c7e0103,
+	0x76bb0006,
 	0x0465b600,
 	0x659450f9,
 	0x0256bb04,
 	0x75fd50bd,
 	0x7e50fc04,
-	0xb60004c6,
+	0xb600067e,
 	0x11f40464,
-/* 0x0525: i2c_start_send */
+/* 0x06dd: i2c_start_send */
 	0x7e00031d,
-	0x4e000484,
+	0x4e00063c,
 	0x5d7e1388,
 	0x00030000,
-	0x00046a7e,
+	0x0006227e,
 	0x7e13884e,
-/* 0x053f: i2c_start_out */
+/* 0x06f7: i2c_start_out */
 	0xf800005d,
-/* 0x0541: i2c_stop */
+/* 0x06f9: i2c_stop */
 	0x7e000300,
-	0x0300046a,
-	0x04847e00,
+	0x03000622,
+	0x063c7e00,
 	0x03e84e00,
 	0x00005d7e,
-	0x6a7e0103,
-	0x884e0004,
+	0x227e0103,
+	0x884e0006,
 	0x005d7e13,
 	0x7e010300,
-	0x4e000484,
+	0x4e00063c,
 	0x5d7e1388,
 	0x00f80000,
-/* 0x0570: i2c_bitw */
-	0x0004847e,
+/* 0x0728: i2c_bitw */
+	0x00063c7e,
 	0x7e03e84e,
 	0xbb00005d,
 	0x65b60076,
@@ -1209,44 +1392,44 @@ uint32_t nv108_pwr_code[] = {
 	0x56bb0465,
 	0xfd50bd02,
 	0x50fc0475,
-	0x0004c67e,
+	0x00067e7e,
 	0xf40464b6,
 	0x884e1711,
 	0x005d7e13,
 	0x7e000300,
-	0x4e00046a,
+	0x4e000622,
 	0x5d7e1388,
-/* 0x05ae: i2c_bitw_out */
+/* 0x0766: i2c_bitw_out */
 	0x00f80000,
-/* 0x05b0: i2c_bitr */
-	0x847e0103,
-	0xe84e0004,
+/* 0x0768: i2c_bitr */
+	0x3c7e0103,
+	0xe84e0006,
 	0x005d7e03,
 	0x0076bb00,
 	0xf90465b6,
 	0x04659450,
 	0xbd0256bb,
 	0x0475fd50,
-	0xc67e50fc,
-	0x64b60004,
+	0x7e7e50fc,
+	0x64b60006,
 	0x1a11f404,
-	0x0004b27e,
-	0x6a7e0003,
-	0x884e0004,
+	0x00066a7e,
+	0x227e0003,
+	0x884e0006,
 	0x005d7e13,
 	0x013cf000,
-/* 0x05f3: i2c_bitr_done */
+/* 0x07ab: i2c_bitr_done */
 	0xf80131f4,
-/* 0x05f5: i2c_get_byte */
+/* 0x07ad: i2c_get_byte */
 	0x04000500,
-/* 0x05f9: i2c_get_byte_next */
+/* 0x07b1: i2c_get_byte_next */
 	0x0154b608,
 	0xb60076bb,
 	0x50f90465,
 	0xbb046594,
 	0x50bd0256,
 	0xfc0475fd,
-	0x05b07e50,
+	0x07687e50,
 	0x0464b600,
 	0xfd2a11f4,
 	0x42b60553,
@@ -1257,11 +1440,11 @@ uint32_t nv108_pwr_code[] = {
 	0x0256bb04,
 	0x75fd50bd,
 	0x7e50fc04,
-	0xb6000570,
-/* 0x0642: i2c_get_byte_done */
+	0xb6000728,
+/* 0x07fa: i2c_get_byte_done */
 	0x00f80464,
-/* 0x0644: i2c_put_byte */
-/* 0x0646: i2c_put_byte_next */
+/* 0x07fc: i2c_put_byte */
+/* 0x07fe: i2c_put_byte_next */
 	0x42b60804,
 	0x3854ff01,
 	0xb60076bb,
@@ -1269,7 +1452,7 @@ uint32_t nv108_pwr_code[] = {
 	0xbb046594,
 	0x50bd0256,
 	0xfc0475fd,
-	0x05707e50,
+	0x07287e50,
 	0x0464b600,
 	0xb03411f4,
 	0x1bf40046,
@@ -1278,21 +1461,21 @@ uint32_t nv108_pwr_code[] = {
 	0x04659450,
 	0xbd0256bb,
 	0x0475fd50,
-	0xb07e50fc,
-	0x64b60005,
+	0x687e50fc,
+	0x64b60007,
 	0x0f11f404,
 	0xb00076bb,
 	0x1bf40136,
 	0x0132f406,
-/* 0x069c: i2c_put_byte_done */
-/* 0x069e: i2c_addr */
+/* 0x0854: i2c_put_byte_done */
+/* 0x0856: i2c_addr */
 	0x76bb00f8,
 	0x0465b600,
 	0x659450f9,
 	0x0256bb04,
 	0x75fd50bd,
 	0x7e50fc04,
-	0xb60004e9,
+	0xb60006a1,
 	0x11f40464,
 	0x2ec3e729,
 	0x0134b601,
@@ -1302,32 +1485,32 @@ uint32_t nv108_pwr_code[] = {
 	0x56bb0465,
 	0xfd50bd02,
 	0x50fc0475,
-	0x0006447e,
-/* 0x06e3: i2c_addr_done */
+	0x0007fc7e,
+/* 0x089b: i2c_addr_done */
 	0xf80464b6,
-/* 0x06e5: i2c_acquire_addr */
+/* 0x089d: i2c_acquire_addr */
 	0xf8cec700,
 	0xb705e4b6,
 	0xf8d014e0,
-/* 0x06f1: i2c_acquire */
-	0x06e57e00,
+/* 0x08a9: i2c_acquire */
+	0x089d7e00,
 	0x00047e00,
 	0x03d9f000,
 	0x00002e7e,
-/* 0x0702: i2c_release */
-	0xe57e00f8,
-	0x047e0006,
+/* 0x08ba: i2c_release */
+	0x9d7e00f8,
+	0x047e0008,
 	0xdaf00000,
 	0x002e7e03,
-/* 0x0713: i2c_recv */
+/* 0x08cb: i2c_recv */
 	0xf400f800,
 	0xc1c70132,
 	0x0214b6f8,
 	0xf52816b0,
 	0xb801371f,
-	0x000bd413,
+	0x000be813,
 	0xb8003298,
-	0x000bac13,
+	0x000bc013,
 	0xf4003198,
 	0xd0f90231,
 	0xd0f9e0f9,
@@ -1339,7 +1522,7 @@ uint32_t nv108_pwr_code[] = {
 	0x56bb0465,
 	0xfd50bd02,
 	0x50fc0475,
-	0x0006f17e,
+	0x0008a97e,
 	0xfc0464b6,
 	0x00d6b0d0,
 	0x00b01bf5,
@@ -1349,7 +1532,7 @@ uint32_t nv108_pwr_code[] = {
 	0x0256bb04,
 	0x75fd50bd,
 	0x7e50fc04,
-	0xb600069e,
+	0xb6000856,
 	0x11f50464,
 	0xc5c700cc,
 	0x0076bbe0,
@@ -1357,8 +1540,8 @@ uint32_t nv108_pwr_code[] = {
 	0x04659450,
 	0xbd0256bb,
 	0x0475fd50,
-	0x447e50fc,
-	0x64b60006,
+	0xfc7e50fc,
+	0x64b60007,
 	0xa911f504,
 	0xbb010500,
 	0x65b60076,
@@ -1366,7 +1549,7 @@ uint32_t nv108_pwr_code[] = {
 	0x56bb0465,
 	0xfd50bd02,
 	0x50fc0475,
-	0x00069e7e,
+	0x0008567e,
 	0xf50464b6,
 	0xbb008711,
 	0x65b60076,
@@ -1374,7 +1557,7 @@ uint32_t nv108_pwr_code[] = {
 	0x56bb0465,
 	0xfd50bd02,
 	0x50fc0475,
-	0x0005f57e,
+	0x0007ad7e,
 	0xf40464b6,
 	0x5bcb6711,
 	0x0076bbe0,
@@ -1382,37 +1565,37 @@ uint32_t nv108_pwr_code[] = {
 	0x04659450,
 	0xbd0256bb,
 	0x0475fd50,
-	0x417e50fc,
-	0x64b60005,
+	0xf97e50fc,
+	0x64b60006,
 	0xbd5bb204,
 	0x410ef474,
-/* 0x0818: i2c_recv_not_rd08 */
+/* 0x09d0: i2c_recv_not_rd08 */
 	0xf401d6b0,
 	0x00053b1b,
-	0x00069e7e,
+	0x0008567e,
 	0xc73211f4,
-	0x447ee0c5,
-	0x11f40006,
+	0xfc7ee0c5,
+	0x11f40007,
 	0x7e000528,
-	0xf400069e,
+	0xf4000856,
 	0xb5c71f11,
-	0x06447ee0,
+	0x07fc7ee0,
 	0x1511f400,
-	0x0005417e,
+	0x0006f97e,
 	0xc5c774bd,
 	0x091bf408,
 	0xf40232f4,
-/* 0x0856: i2c_recv_not_wr08 */
-/* 0x0856: i2c_recv_done */
+/* 0x0a0e: i2c_recv_not_wr08 */
+/* 0x0a0e: i2c_recv_done */
 	0xcec7030e,
-	0x07027ef8,
+	0x08ba7ef8,
 	0xfce0fc00,
 	0x0912f4d0,
-	0x3f7e7cb2,
-/* 0x086a: i2c_recv_exit */
+	0xc27e7cb2,
+/* 0x0a22: i2c_recv_exit */
 	0x00f80002,
-/* 0x086c: i2c_init */
-/* 0x086e: test_recv */
+/* 0x0a24: i2c_init */
+/* 0x0a26: test_recv */
 	0x584100f8,
 	0x0011cf04,
 	0x400110b6,
@@ -1420,28 +1603,28 @@ uint32_t nv108_pwr_code[] = {
 	0xf104bd00,
 	0xf1d900e7,
 	0x7e134fe3,
-	0xf8000196,
-/* 0x088d: test_init */
+	0xf8000201,
+/* 0x0a45: test_init */
 	0x08004e00,
-	0x0001967e,
-/* 0x0896: idle_recv */
+	0x0002017e,
+/* 0x0a4e: idle_recv */
 	0x00f800f8,
-/* 0x0898: idle */
+/* 0x0a50: idle */
 	0x410031f4,
 	0x11cf0454,
 	0x0110b600,
 	0xf6045440,
 	0x04bd0001,
-/* 0x08ac: idle_loop */
+/* 0x0a64: idle_loop */
 	0x32f45801,
-/* 0x08b1: idle_proc */
-/* 0x08b1: idle_proc_exec */
+/* 0x0a69: idle_proc */
+/* 0x0a69: idle_proc_exec */
 	0xb210f902,
-	0x02487e1e,
+	0x02cb7e1e,
 	0xf410fc00,
 	0x31f40911,
 	0xf00ef402,
-/* 0x08c4: idle_proc_next */
+/* 0x0a7c: idle_proc_next */
 	0xa65810b6,
 	0xe81bf41f,
 	0xf4e002f4,
@@ -1457,4 +1640,22 @@ uint32_t nv108_pwr_code[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
 };
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nva3.fuc b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nva3.fuc
index 6744fcc06151..daa06c1c655e 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nva3.fuc
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nva3.fuc
@@ -23,6 +23,7 @@
  */
 
 #define NVKM_PPWR_CHIPSET GT215
+#define HW_TICKS_PER_US 203 // should be 202.5
 
 //#define NVKM_FALCON_PC24
 //#define NVKM_FALCON_UNSHIFTED_IO
@@ -34,6 +35,7 @@
 .section #nva3_pwr_data
 #define INCLUDE_PROC
 #include "kernel.fuc"
+#include "arith.fuc"
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
@@ -44,6 +46,7 @@
 
 #define INCLUDE_DATA
 #include "kernel.fuc"
+#include "arith.fuc"
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
@@ -56,6 +59,7 @@
 .section #nva3_pwr_code
 #define INCLUDE_CODE
 #include "kernel.fuc"
+#include "arith.fuc"
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nva3.fuc.h b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nva3.fuc.h
index e087ce3041be..64e97baabc3c 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nva3.fuc.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nva3.fuc.h
@@ -24,8 +24,8 @@ uint32_t nva3_pwr_data[] = {
 	0x00000000,
 /* 0x0058: proc_list_head */
 	0x54534f48,
-	0x00000430,
-	0x000003cd,
+	0x00000512,
+	0x000004af,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -46,8 +46,8 @@ uint32_t nva3_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x584d454d,
-	0x00000542,
-	0x00000534,
+	0x000006e0,
+	0x000006d2,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -68,8 +68,8 @@ uint32_t nva3_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x46524550,
-	0x00000546,
-	0x00000544,
+	0x000006e4,
+	0x000006e2,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -90,8 +90,8 @@ uint32_t nva3_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x5f433249,
-	0x00000976,
-	0x00000819,
+	0x00000b14,
+	0x000009b7,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -112,8 +112,8 @@ uint32_t nva3_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x54534554,
-	0x0000099f,
-	0x00000978,
+	0x00000b3d,
+	0x00000b16,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -134,8 +134,8 @@ uint32_t nva3_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x454c4449,
-	0x000009ab,
-	0x000009a9,
+	0x00000b49,
+	0x00000b47,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -227,25 +227,31 @@ uint32_t nva3_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 /* 0x0370: memx_func_head */
-	0x00010000,
-	0x00000000,
-	0x0000046f,
-/* 0x037c: memx_func_next */
 	0x00000001,
 	0x00000000,
-	0x00000496,
+	0x00000551,
+/* 0x037c: memx_func_next */
 	0x00000002,
+	0x00000000,
+	0x000005a8,
+	0x00000003,
 	0x00000002,
-	0x000004b7,
-	0x00040003,
+	0x0000063a,
+	0x00040004,
+	0x00000000,
+	0x00000656,
+	0x00010005,
+	0x00000000,
+	0x00000673,
+	0x00010006,
 	0x00000000,
-	0x000004d3,
-	0x00010004,
+	0x000005f8,
+/* 0x03b8: memx_func_tail */
+/* 0x03b8: memx_ts_start */
 	0x00000000,
-	0x000004f0,
-/* 0x03ac: memx_func_tail */
-/* 0x03ac: memx_data_head */
+/* 0x03bc: memx_ts_end */
 	0x00000000,
+/* 0x03c0: memx_data_head */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -757,8 +763,9 @@ uint32_t nva3_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-/* 0x0bac: memx_data_tail */
-/* 0x0bac: i2c_scl_map */
+	0x00000000,
+/* 0x0bc0: memx_data_tail */
+/* 0x0bc0: i2c_scl_map */
 	0x00001000,
 	0x00004000,
 	0x00010000,
@@ -769,7 +776,7 @@ uint32_t nva3_pwr_data[] = {
 	0x01000000,
 	0x04000000,
 	0x10000000,
-/* 0x0bd4: i2c_sda_map */
+/* 0x0be8: i2c_sda_map */
 	0x00002000,
 	0x00008000,
 	0x00020000,
@@ -780,7 +787,7 @@ uint32_t nva3_pwr_data[] = {
 	0x02000000,
 	0x08000000,
 	0x20000000,
-/* 0x0bfc: i2c_ctrl */
+/* 0x0c10: i2c_ctrl */
 	0x0000e138,
 	0x0000e150,
 	0x0000e168,
@@ -841,15 +848,10 @@ uint32_t nva3_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
 };
 
 uint32_t nva3_pwr_code[] = {
-	0x030d0ef5,
+	0x039e0ef5,
 /* 0x0004: rd32 */
 	0x07a007f1,
 	0xd00604b6,
@@ -885,19 +887,22 @@ uint32_t nva3_pwr_code[] = {
 	0xd4f100dd,
 	0x1bf47000,
 /* 0x007f: nsec */
-	0xf000f8f2,
+	0xf900f8f2,
+	0xf080f990,
 	0x84b62c87,
 	0x0088cf06,
-/* 0x0088: nsec_loop */
+/* 0x008c: nsec_loop */
 	0xb62c97f0,
 	0x99cf0694,
 	0x0298bb00,
 	0xf4069eb8,
-	0x00f8f11e,
-/* 0x009c: wait */
+	0x80fcf11e,
+	0x00f890fc,
+/* 0x00a4: wait */
+	0x80f990f9,
 	0xb62c87f0,
 	0x88cf0684,
-/* 0x00a5: wait_loop */
+/* 0x00b1: wait_loop */
 	0x02eeb900,
 	0xb90421f4,
 	0xadfd02da,
@@ -907,28 +912,29 @@ uint32_t nva3_pwr_code[] = {
 	0x0099cf06,
 	0xb80298bb,
 	0x1ef4069b,
-/* 0x00c9: wait_done */
-/* 0x00cb: intr_watchdog */
-	0x9800f8df,
+/* 0x00d5: wait_done */
+	0xfc80fcdf,
+/* 0x00db: intr_watchdog */
+	0x9800f890,
 	0x96b003e9,
 	0x2a0bf400,
 	0xbb9a0a98,
 	0x1cf4029a,
 	0x01d7f00f,
-	0x025421f5,
+	0x02dd21f5,
 	0x0ef494bd,
-/* 0x00e9: intr_watchdog_next_time */
+/* 0x00f9: intr_watchdog_next_time */
 	0x9b0a9815,
 	0xf400a6b0,
 	0x9ab8090b,
 	0x061cf406,
-/* 0x00f8: intr_watchdog_next_time_set */
-/* 0x00fb: intr_watchdog_next_proc */
+/* 0x0108: intr_watchdog_next_time_set */
+/* 0x010b: intr_watchdog_next_proc */
 	0x809b0980,
 	0xe0b603e9,
 	0x68e6b158,
 	0xc61bf402,
-/* 0x010a: intr */
+/* 0x011a: intr */
 	0x00f900f8,
 	0x80f904bd,
 	0xa0f990f9,
@@ -948,13 +954,13 @@ uint32_t nva3_pwr_code[] = {
 	0xf40289c4,
 	0x0080230b,
 	0x58e7f09b,
-	0x98cb21f4,
+	0x98db21f4,
 	0x96b09b09,
 	0x110bf400,
 	0xb63407f0,
 	0x09d00604,
 	0x8004bd00,
-/* 0x016e: intr_skip_watchdog */
+/* 0x017e: intr_skip_watchdog */
 	0x89e49a09,
 	0x0bf40800,
 	0x8897f148,
@@ -967,22 +973,22 @@ uint32_t nva3_pwr_code[] = {
 	0x48e7f1c0,
 	0x53e3f14f,
 	0x00d7f054,
-	0x02b921f5,
+	0x034221f5,
 	0x07f1c0fc,
 	0x04b604c0,
 	0x000cd006,
-/* 0x01ae: intr_subintr_skip_fifo */
+/* 0x01be: intr_subintr_skip_fifo */
 	0x07f104bd,
 	0x04b60688,
 	0x0009d006,
-/* 0x01ba: intr_skip_subintr */
+/* 0x01ca: intr_skip_subintr */
 	0x89c404bd,
 	0x070bf420,
 	0xffbfa4f1,
-/* 0x01c4: intr_skip_pause */
+/* 0x01d4: intr_skip_pause */
 	0xf44089c4,
 	0xa4f1070b,
-/* 0x01ce: intr_skip_user0 */
+/* 0x01de: intr_skip_user0 */
 	0x07f0ffbf,
 	0x0604b604,
 	0xbd0008d0,
@@ -993,596 +999,732 @@ uint32_t nva3_pwr_code[] = {
 	0x90fca0fc,
 	0x00fc80fc,
 	0xf80032f4,
-/* 0x01f5: timer */
-	0x1032f401,
-	0xb003f898,
-	0x1cf40086,
-	0x03fe8051,
+/* 0x0205: ticks_from_ns */
+	0xf9c0f901,
+	0xcbd7f1b0,
+	0x00d3f000,
+	0x041321f5,
+	0x03e8ccec,
+	0xf400b4b0,
+	0xeeec120b,
+	0xd7f103e8,
+	0xd3f000cb,
+	0x1321f500,
+/* 0x022d: ticks_from_ns_quit */
+	0x02ceb904,
+	0xc0fcb0fc,
+/* 0x0236: ticks_from_us */
+	0xc0f900f8,
+	0xd7f1b0f9,
+	0xd3f000cb,
+	0x1321f500,
+	0x02ceb904,
+	0xf400b4b0,
+	0xe4bd050b,
+/* 0x0250: ticks_from_us_quit */
+	0xc0fcb0fc,
+/* 0x0256: ticks_to_us */
+	0xd7f100f8,
+	0xd3f000cb,
+	0xecedff00,
+/* 0x0262: timer */
+	0x90f900f8,
+	0x32f480f9,
+	0x03f89810,
+	0xf40086b0,
+	0x84bd651c,
 	0xb63807f0,
 	0x08d00604,
 	0xf004bd00,
-	0x84b60887,
+	0x84b63487,
 	0x0088cf06,
-	0xf40284f0,
-	0x87f0261b,
-	0x0684b634,
-	0xb80088cf,
-	0x0bf406e0,
-	0x06e8b809,
-/* 0x0233: timer_reset */
-	0xf01f1ef4,
-	0x04b63407,
-	0x000ed006,
-	0x0e8004bd,
-/* 0x0241: timer_enable */
-	0x0187f09a,
+	0xbb9a0998,
+	0xe9bb0298,
+	0x03fe8000,
+	0xb60887f0,
+	0x88cf0684,
+	0x0284f000,
+	0xf0261bf4,
+	0x84b63487,
+	0x0088cf06,
+	0xf406e0b8,
+	0xe8b8090b,
+	0x111cf406,
+/* 0x02b8: timer_reset */
+	0xb63407f0,
+	0x0ed00604,
+	0x8004bd00,
+/* 0x02c6: timer_enable */
+	0x87f09a0e,
+	0x3807f001,
+	0xd00604b6,
+	0x04bd0008,
+/* 0x02d4: timer_done */
+	0xfc1031f4,
+	0xf890fc80,
+/* 0x02dd: send_proc */
+	0xf980f900,
+	0x05e89890,
+	0xf004e998,
+	0x89b80486,
+	0x2a0bf406,
+	0x940398c4,
+	0x80b60488,
+	0x008ebb18,
+	0x8000fa98,
+	0x8d80008a,
+	0x028c8001,
+	0xb6038b80,
+	0x94f00190,
+	0x04e98007,
+/* 0x0317: send_done */
+	0xfc0231f4,
+	0xf880fc90,
+/* 0x031d: find */
+	0xf080f900,
+	0x31f45887,
+/* 0x0325: find_loop */
+	0x008a9801,
+	0xf406aeb8,
+	0x80b6100b,
+	0x6886b158,
+	0xf01bf402,
+/* 0x033b: find_done */
+	0xb90132f4,
+	0x80fc028e,
+/* 0x0342: send */
+	0x21f500f8,
+	0x01f4031d,
+/* 0x034b: recv */
+	0xf900f897,
+	0x9880f990,
+	0xe99805e8,
+	0x0132f404,
+	0xf40689b8,
+	0x89c43d0b,
+	0x0180b603,
+	0x800784f0,
+	0xea9805e8,
+	0xfef0f902,
+	0xf0f9018f,
+	0x9402efb9,
+	0xe9bb0499,
+	0x18e0b600,
+	0x9803eb98,
+	0xed9802ec,
+	0x00ee9801,
+	0xf0fca5f9,
+	0xf400f8fe,
+	0xf0fc0131,
+/* 0x0398: recv_done */
+	0x90fc80fc,
+/* 0x039e: init */
+	0x17f100f8,
+	0x14b60108,
+	0x0011cf06,
+	0x010911e7,
+	0xfe0814b6,
+	0x17f10014,
+	0x13f000e0,
+	0x1c07f000,
+	0xd00604b6,
+	0x04bd0001,
+	0xf0ff17f0,
+	0x04b61407,
+	0x0001d006,
+	0x17f004bd,
+	0x0015f102,
+	0x1007f008,
+	0xd00604b6,
+	0x04bd0001,
+	0x011a17f1,
+	0xfe0013f0,
+	0x31f40010,
+	0x0117f010,
 	0xb63807f0,
-	0x08d00604,
-/* 0x024f: timer_done */
-	0xf404bd00,
-	0x00f81031,
-/* 0x0254: send_proc */
-	0x90f980f9,
-	0x9805e898,
-	0x86f004e9,
-	0x0689b804,
-	0xc42a0bf4,
-	0x88940398,
-	0x1880b604,
-	0x98008ebb,
-	0x8a8000fa,
-	0x018d8000,
-	0x80028c80,
-	0x90b6038b,
-	0x0794f001,
-	0xf404e980,
-/* 0x028e: send_done */
-	0x90fc0231,
-	0x00f880fc,
-/* 0x0294: find */
-	0x87f080f9,
-	0x0131f458,
-/* 0x029c: find_loop */
-	0xb8008a98,
-	0x0bf406ae,
-	0x5880b610,
-	0x026886b1,
-	0xf4f01bf4,
-/* 0x02b2: find_done */
-	0x8eb90132,
-	0xf880fc02,
-/* 0x02b9: send */
-	0x9421f500,
-	0x9701f402,
-/* 0x02c2: recv */
-	0xe89800f8,
-	0x04e99805,
-	0xb80132f4,
-	0x0bf40689,
-	0x0389c43d,
-	0xf00180b6,
-	0xe8800784,
-	0x02ea9805,
-	0x8ffef0f9,
-	0xb9f0f901,
-	0x999402ef,
-	0x00e9bb04,
-	0x9818e0b6,
-	0xec9803eb,
-	0x01ed9802,
-	0xf900ee98,
-	0xfef0fca5,
-	0x31f400f8,
-/* 0x030b: recv_done */
-	0xf8f0fc01,
-/* 0x030d: init */
-	0x0817f100,
-	0x0614b601,
-	0xe70011cf,
-	0xb6010911,
-	0x14fe0814,
-	0xe017f100,
-	0x0013f000,
-	0xb61c07f0,
 	0x01d00604,
 	0xf004bd00,
-	0x07f0ff17,
-	0x0604b614,
-	0xbd0001d0,
-	0x0217f004,
-	0x080015f1,
-	0xb61007f0,
-	0x01d00604,
-	0xf104bd00,
-	0xf0010a17,
-	0x10fe0013,
-	0x1031f400,
-	0xf00117f0,
-	0x04b63807,
-	0x0001d006,
-	0xf7f004bd,
-/* 0x0371: init_proc */
-	0x01f19858,
-	0xf40016b0,
-	0x15f9fa0b,
-	0xf458f0b6,
-/* 0x0382: host_send */
-	0x17f1f20e,
-	0x14b604b0,
-	0x0011cf06,
-	0x04a027f1,
-	0xcf0624b6,
-	0x12b80022,
-	0x320bf406,
-	0x94071ec4,
-	0xe0b704ee,
-	0xeb980270,
-	0x02ec9803,
-	0x9801ed98,
-	0x21f500ee,
-	0x10b602b9,
-	0x0f1ec401,
-	0x04b007f1,
-	0xd00604b6,
-	0x04bd000e,
-/* 0x03cb: host_send_done */
-	0xf8ba0ef4,
-/* 0x03cd: host_recv */
-	0x4917f100,
-	0x5413f14e,
-	0x06e1b852,
-/* 0x03db: host_recv_wait */
-	0xf1aa0bf4,
-	0xb604cc17,
-	0x11cf0614,
-	0xc827f100,
-	0x0624b604,
-	0xf00022cf,
-	0x12b80816,
-	0xe60bf406,
-	0xb60723c4,
-	0x30b70434,
-	0x3b8002f0,
-	0x023c8003,
-	0x80013d80,
-	0x20b6003e,
-	0x0f24f001,
-	0x04c807f1,
+/* 0x0402: init_proc */
+	0xf19858f7,
+	0x0016b001,
+	0xf9fa0bf4,
+	0x58f0b615,
+/* 0x0413: mulu32_32_64 */
+	0xf9f20ef4,
+	0xf920f910,
+	0x9540f930,
+	0xd29510e1,
+	0xbdc4bd10,
+	0xc0edffb4,
+	0xb9301dff,
+	0x34f10234,
+	0x34b6ffff,
+	0x1045b610,
+	0xbb00c3bb,
+	0xe2ff01b4,
+	0x0234b930,
+	0xffff34f1,
+	0xb61034b6,
+	0xc3bb1045,
+	0x01b4bb00,
+	0xbb3012ff,
+	0x40fc00b3,
+	0x20fc30fc,
+	0x00f810fc,
+/* 0x0464: host_send */
+	0x04b017f1,
+	0xcf0614b6,
+	0x27f10011,
+	0x24b604a0,
+	0x0022cf06,
+	0xf40612b8,
+	0x1ec4320b,
+	0x04ee9407,
+	0x0270e0b7,
+	0x9803eb98,
+	0xed9802ec,
+	0x00ee9801,
+	0x034221f5,
+	0xc40110b6,
+	0x07f10f1e,
+	0x04b604b0,
+	0x000ed006,
+	0x0ef404bd,
+/* 0x04ad: host_send_done */
+/* 0x04af: host_recv */
+	0xf100f8ba,
+	0xf14e4917,
+	0xb8525413,
+	0x0bf406e1,
+/* 0x04bd: host_recv_wait */
+	0xcc17f1aa,
+	0x0614b604,
+	0xf10011cf,
+	0xb604c827,
+	0x22cf0624,
+	0x0816f000,
+	0xf40612b8,
+	0x23c4e60b,
+	0x0434b607,
+	0x02f030b7,
+	0x80033b80,
+	0x3d80023c,
+	0x003e8001,
+	0xf00120b6,
+	0x07f10f24,
+	0x04b604c8,
+	0x0002d006,
+	0x27f004bd,
+	0x0007f040,
 	0xd00604b6,
 	0x04bd0002,
-	0xf04027f0,
-	0x04b60007,
-	0x0002d006,
-	0x00f804bd,
-/* 0x0430: host_init */
-	0x008017f1,
-	0xf11014b6,
-	0xf1027015,
-	0xb604d007,
-	0x01d00604,
-	0xf104bd00,
-	0xb6008017,
-	0x15f11014,
-	0x07f102f0,
-	0x04b604dc,
-	0x0001d006,
-	0x17f004bd,
-	0xc407f101,
+/* 0x0512: host_init */
+	0x17f100f8,
+	0x14b60080,
+	0x7015f110,
+	0xd007f102,
 	0x0604b604,
 	0xbd0001d0,
-/* 0x046f: memx_func_enter */
-	0xf000f804,
+	0x8017f104,
+	0x1014b600,
+	0x02f015f1,
+	0x04dc07f1,
+	0xd00604b6,
+	0x04bd0001,
+	0xf10117f0,
+	0xb604c407,
+	0x01d00604,
+	0xf804bd00,
+/* 0x0551: memx_func_enter */
+	0x1087f100,
+	0x028eb916,
+	0xb90421f4,
+	0x67f102d7,
+	0x63f1fffc,
+	0x76fdffff,
+	0x0267f104,
+	0x0576fd00,
+	0x70f980f9,
+	0xe0fcd0fc,
+	0xf03f21f4,
 	0x07f10467,
 	0x04b607e0,
 	0x0006d006,
-/* 0x047e: memx_func_enter_wait */
+/* 0x058a: memx_func_enter_wait */
 	0x67f104bd,
 	0x64b607c0,
 	0x0066cf06,
 	0xf40464f0,
-	0x1698f30b,
-	0x0410b600,
-/* 0x0496: memx_func_leave */
-	0x67f000f8,
-	0xe407f104,
-	0x0604b607,
-	0xbd0006d0,
-/* 0x04a5: memx_func_leave_wait */
-	0xc067f104,
+	0x67f0f30b,
+	0x0664b62c,
+	0x800066cf,
+	0x00f8ee06,
+/* 0x05a8: memx_func_leave */
+	0xb62c67f0,
+	0x66cf0664,
+	0xef068000,
+	0xf10467f0,
+	0xb607e407,
+	0x06d00604,
+/* 0x05c3: memx_func_leave_wait */
+	0xf104bd00,
+	0xb607c067,
+	0x66cf0664,
+	0x0464f000,
+	0xf1f31bf4,
+	0xb9161087,
+	0x21f4028e,
+	0x02d7b904,
+	0xffcc67f1,
+	0xffff63f1,
+	0xf90476fd,
+	0xfc70f980,
+	0xf4e0fcd0,
+	0x00f83f21,
+/* 0x05f8: memx_func_wait_vblank */
+	0xb0001698,
+	0x0bf40066,
+	0x0166b013,
+	0xf4060bf4,
+/* 0x060a: memx_func_wait_vblank_head1 */
+	0x77f12e0e,
+	0x0ef40020,
+/* 0x0611: memx_func_wait_vblank_head0 */
+	0x0877f107,
+/* 0x0615: memx_func_wait_vblank_0 */
+	0xc467f100,
 	0x0664b607,
-	0xf00066cf,
-	0x1bf40464,
-/* 0x04b7: memx_func_wr32 */
-	0x9800f8f3,
-	0x15980016,
-	0x0810b601,
-	0x50f960f9,
-	0xe0fcd0fc,
-	0xb63f21f4,
-	0x1bf40242,
-/* 0x04d3: memx_func_wait */
-	0xf000f8e9,
-	0x84b62c87,
-	0x0088cf06,
-	0x98001e98,
-	0x1c98011d,
-	0x031b9802,
-	0xf41010b6,
-	0x00f89c21,
-/* 0x04f0: memx_func_delay */
-	0xb6001e98,
-	0x21f40410,
-/* 0x04fb: memx_exec */
-	0xf900f87f,
-	0xb9d0f9e0,
-	0xb2b902c1,
-/* 0x0505: memx_exec_next */
-	0x00139802,
-	0x950410b6,
-	0x30f01034,
-	0xde35980c,
-	0x12b855f9,
-	0xec1ef406,
-	0xe0fcd0fc,
-	0x02b921f5,
-/* 0x0526: memx_info */
-	0xc7f100f8,
-	0xb7f103ac,
-	0x21f50800,
-	0x00f802b9,
-/* 0x0534: memx_recv */
-	0xf401d6b0,
-	0xd6b0c40b,
-	0xe90bf400,
-/* 0x0542: memx_init */
-	0x00f800f8,
-/* 0x0544: perf_recv */
-/* 0x0546: perf_init */
+	0xfd0066cf,
+	0x1bf40467,
+/* 0x0625: memx_func_wait_vblank_1 */
+	0xc467f1f3,
+	0x0664b607,
+	0xfd0066cf,
+	0x0bf40467,
+/* 0x0635: memx_func_wait_vblank_fini */
+	0x0410b6f3,
+/* 0x063a: memx_func_wr32 */
+	0x169800f8,
+	0x01159800,
+	0xf90810b6,
+	0xfc50f960,
+	0xf4e0fcd0,
+	0x42b63f21,
+	0xe91bf402,
+/* 0x0656: memx_func_wait */
+	0x87f000f8,
+	0x0684b62c,
+	0x980088cf,
+	0x1d98001e,
+	0x021c9801,
+	0xb6031b98,
+	0x21f41010,
+/* 0x0673: memx_func_delay */
+	0x9800f8a4,
+	0x10b6001e,
+	0x7f21f404,
+/* 0x067e: memx_exec */
+	0xe0f900f8,
+	0xc1b9d0f9,
+	0x02b2b902,
+/* 0x0688: memx_exec_next */
+	0xb6001398,
+	0x34e70410,
+	0x33e701f0,
+	0x32b601e0,
+	0x0c30f001,
+	0xf9de3598,
+	0x0612b855,
+	0x98e41ef4,
+	0x0c98ee0b,
+	0x02cbbbef,
+	0x07c4b7f1,
+	0xcf06b4b6,
+	0xd0fc00bb,
+	0x21f5e0fc,
+	0x00f80342,
+/* 0x06c4: memx_info */
+	0x03c0c7f1,
+	0x0800b7f1,
+	0x034221f5,
+/* 0x06d2: memx_recv */
+	0xd6b000f8,
+	0xa90bf401,
+	0xf400d6b0,
+	0x00f8e90b,
+/* 0x06e0: memx_init */
+/* 0x06e2: perf_recv */
 	0x00f800f8,
-/* 0x0548: i2c_drive_scl */
-	0xf40036b0,
-	0x07f1110b,
-	0x04b607e0,
-	0x0001d006,
-	0x00f804bd,
-/* 0x055c: i2c_drive_scl_lo */
-	0x07e407f1,
-	0xd00604b6,
-	0x04bd0001,
-/* 0x056a: i2c_drive_sda */
+/* 0x06e4: perf_init */
+/* 0x06e6: i2c_drive_scl */
 	0x36b000f8,
 	0x110bf400,
 	0x07e007f1,
 	0xd00604b6,
-	0x04bd0002,
-/* 0x057e: i2c_drive_sda_lo */
+	0x04bd0001,
+/* 0x06fa: i2c_drive_scl_lo */
 	0x07f100f8,
 	0x04b607e4,
+	0x0001d006,
+	0x00f804bd,
+/* 0x0708: i2c_drive_sda */
+	0xf40036b0,
+	0x07f1110b,
+	0x04b607e0,
 	0x0002d006,
 	0x00f804bd,
-/* 0x058c: i2c_sense_scl */
-	0xf10132f4,
-	0xb607c437,
-	0x33cf0634,
-	0x0431fd00,
-	0xf4060bf4,
-/* 0x05a2: i2c_sense_scl_done */
-	0x00f80131,
-/* 0x05a4: i2c_sense_sda */
-	0xf10132f4,
-	0xb607c437,
-	0x33cf0634,
-	0x0432fd00,
-	0xf4060bf4,
-/* 0x05ba: i2c_sense_sda_done */
-	0x00f80131,
-/* 0x05bc: i2c_raise_scl */
-	0x47f140f9,
-	0x37f00898,
-	0x4821f501,
-/* 0x05c9: i2c_raise_scl_wait */
-	0xe8e7f105,
-	0x7f21f403,
-	0x058c21f5,
-	0xb60901f4,
-	0x1bf40142,
-/* 0x05dd: i2c_raise_scl_done */
-	0xf840fcef,
-/* 0x05e1: i2c_start */
-	0x8c21f500,
-	0x0d11f405,
-	0x05a421f5,
-	0xf40611f4,
-/* 0x05f2: i2c_start_rep */
-	0x37f0300e,
-	0x4821f500,
-	0x0137f005,
-	0x056a21f5,
-	0xb60076bb,
-	0x50f90465,
-	0xbb046594,
-	0x50bd0256,
-	0xfc0475fd,
-	0xbc21f550,
-	0x0464b605,
-/* 0x061f: i2c_start_send */
-	0xf01f11f4,
+/* 0x071c: i2c_drive_sda_lo */
+	0x07e407f1,
+	0xd00604b6,
+	0x04bd0002,
+/* 0x072a: i2c_sense_scl */
+	0x32f400f8,
+	0xc437f101,
+	0x0634b607,
+	0xfd0033cf,
+	0x0bf40431,
+	0x0131f406,
+/* 0x0740: i2c_sense_scl_done */
+/* 0x0742: i2c_sense_sda */
+	0x32f400f8,
+	0xc437f101,
+	0x0634b607,
+	0xfd0033cf,
+	0x0bf40432,
+	0x0131f406,
+/* 0x0758: i2c_sense_sda_done */
+/* 0x075a: i2c_raise_scl */
+	0x40f900f8,
+	0x089847f1,
+	0xf50137f0,
+/* 0x0767: i2c_raise_scl_wait */
+	0xf106e621,
+	0xf403e8e7,
+	0x21f57f21,
+	0x01f4072a,
+	0x0142b609,
+/* 0x077b: i2c_raise_scl_done */
+	0xfcef1bf4,
+/* 0x077f: i2c_start */
+	0xf500f840,
+	0xf4072a21,
+	0x21f50d11,
+	0x11f40742,
+	0x300ef406,
+/* 0x0790: i2c_start_rep */
+	0xf50037f0,
+	0xf006e621,
+	0x21f50137,
+	0x76bb0708,
+	0x0465b600,
+	0x659450f9,
+	0x0256bb04,
+	0x75fd50bd,
+	0xf550fc04,
+	0xb6075a21,
+	0x11f40464,
+/* 0x07bd: i2c_start_send */
+	0x0037f01f,
+	0x070821f5,
+	0x1388e7f1,
+	0xf07f21f4,
 	0x21f50037,
-	0xe7f1056a,
+	0xe7f106e6,
 	0x21f41388,
-	0x0037f07f,
-	0x054821f5,
-	0x1388e7f1,
-/* 0x063b: i2c_start_out */
-	0xf87f21f4,
-/* 0x063d: i2c_stop */
-	0x0037f000,
-	0x054821f5,
-	0xf50037f0,
-	0xf1056a21,
-	0xf403e8e7,
-	0x37f07f21,
-	0x4821f501,
-	0x88e7f105,
-	0x7f21f413,
+/* 0x07d9: i2c_start_out */
+/* 0x07db: i2c_stop */
+	0xf000f87f,
+	0x21f50037,
+	0x37f006e6,
+	0x0821f500,
+	0xe8e7f107,
+	0x7f21f403,
 	0xf50137f0,
-	0xf1056a21,
+	0xf106e621,
 	0xf41388e7,
-	0x00f87f21,
-/* 0x0670: i2c_bitw */
-	0x056a21f5,
-	0x03e8e7f1,
-	0xbb7f21f4,
-	0x65b60076,
-	0x9450f904,
-	0x56bb0465,
-	0xfd50bd02,
-	0x50fc0475,
-	0x05bc21f5,
-	0xf40464b6,
-	0xe7f11811,
-	0x21f41388,
-	0x0037f07f,
-	0x054821f5,
-	0x1388e7f1,
-/* 0x06af: i2c_bitw_out */
-	0xf87f21f4,
-/* 0x06b1: i2c_bitr */
-	0x0137f000,
-	0x056a21f5,
-	0x03e8e7f1,
-	0xbb7f21f4,
-	0x65b60076,
-	0x9450f904,
-	0x56bb0465,
-	0xfd50bd02,
-	0x50fc0475,
-	0x05bc21f5,
-	0xf40464b6,
-	0x21f51b11,
-	0x37f005a4,
-	0x4821f500,
-	0x88e7f105,
+	0x37f07f21,
+	0x0821f501,
+	0x88e7f107,
 	0x7f21f413,
-	0xf4013cf0,
-/* 0x06f6: i2c_bitr_done */
-	0x00f80131,
-/* 0x06f8: i2c_get_byte */
-	0xf00057f0,
-/* 0x06fe: i2c_get_byte_next */
-	0x54b60847,
-	0x0076bb01,
+/* 0x080e: i2c_bitw */
+	0x21f500f8,
+	0xe7f10708,
+	0x21f403e8,
+	0x0076bb7f,
 	0xf90465b6,
 	0x04659450,
 	0xbd0256bb,
 	0x0475fd50,
 	0x21f550fc,
-	0x64b606b1,
-	0x2b11f404,
-	0xb60553fd,
-	0x1bf40142,
-	0x0137f0d8,
-	0xb60076bb,
-	0x50f90465,
-	0xbb046594,
-	0x50bd0256,
-	0xfc0475fd,
-	0x7021f550,
-	0x0464b606,
-/* 0x0748: i2c_get_byte_done */
-/* 0x074a: i2c_put_byte */
-	0x47f000f8,
-/* 0x074d: i2c_put_byte_next */
-	0x0142b608,
-	0xbb3854ff,
+	0x64b6075a,
+	0x1811f404,
+	0x1388e7f1,
+	0xf07f21f4,
+	0x21f50037,
+	0xe7f106e6,
+	0x21f41388,
+/* 0x084d: i2c_bitw_out */
+/* 0x084f: i2c_bitr */
+	0xf000f87f,
+	0x21f50137,
+	0xe7f10708,
+	0x21f403e8,
+	0x0076bb7f,
+	0xf90465b6,
+	0x04659450,
+	0xbd0256bb,
+	0x0475fd50,
+	0x21f550fc,
+	0x64b6075a,
+	0x1b11f404,
+	0x074221f5,
+	0xf50037f0,
+	0xf106e621,
+	0xf41388e7,
+	0x3cf07f21,
+	0x0131f401,
+/* 0x0894: i2c_bitr_done */
+/* 0x0896: i2c_get_byte */
+	0x57f000f8,
+	0x0847f000,
+/* 0x089c: i2c_get_byte_next */
+	0xbb0154b6,
 	0x65b60076,
 	0x9450f904,
 	0x56bb0465,
 	0xfd50bd02,
 	0x50fc0475,
-	0x067021f5,
+	0x084f21f5,
 	0xf40464b6,
-	0x46b03411,
-	0xd81bf400,
-	0xb60076bb,
-	0x50f90465,
-	0xbb046594,
-	0x50bd0256,
-	0xfc0475fd,
-	0xb121f550,
-	0x0464b606,
-	0xbb0f11f4,
-	0x36b00076,
-	0x061bf401,
-/* 0x07a3: i2c_put_byte_done */
-	0xf80132f4,
-/* 0x07a5: i2c_addr */
-	0x0076bb00,
+	0x53fd2b11,
+	0x0142b605,
+	0xf0d81bf4,
+	0x76bb0137,
+	0x0465b600,
+	0x659450f9,
+	0x0256bb04,
+	0x75fd50bd,
+	0xf550fc04,
+	0xb6080e21,
+/* 0x08e6: i2c_get_byte_done */
+	0x00f80464,
+/* 0x08e8: i2c_put_byte */
+/* 0x08eb: i2c_put_byte_next */
+	0xb60847f0,
+	0x54ff0142,
+	0x0076bb38,
 	0xf90465b6,
 	0x04659450,
 	0xbd0256bb,
 	0x0475fd50,
 	0x21f550fc,
-	0x64b605e1,
-	0x2911f404,
-	0x012ec3e7,
-	0xfd0134b6,
-	0x76bb0553,
+	0x64b6080e,
+	0x3411f404,
+	0xf40046b0,
+	0x76bbd81b,
 	0x0465b600,
 	0x659450f9,
 	0x0256bb04,
 	0x75fd50bd,
 	0xf550fc04,
-	0xb6074a21,
-/* 0x07ea: i2c_addr_done */
-	0x00f80464,
-/* 0x07ec: i2c_acquire_addr */
-	0xb6f8cec7,
-	0xe0b702e4,
-	0xee980bfc,
-/* 0x07fb: i2c_acquire */
-	0xf500f800,
-	0xf407ec21,
-	0xd9f00421,
-	0x3f21f403,
-/* 0x080a: i2c_release */
-	0x21f500f8,
-	0x21f407ec,
-	0x03daf004,
-	0xf83f21f4,
-/* 0x0819: i2c_recv */
-	0x0132f400,
-	0xb6f8c1c7,
-	0x16b00214,
-	0x3a1ff528,
-	0xd413a001,
-	0x0032980b,
-	0x0bac13a0,
-	0xf4003198,
-	0xd0f90231,
-	0xd0f9e0f9,
-	0x000067f1,
-	0x100063f1,
-	0xbb016792,
+	0xb6084f21,
+	0x11f40464,
+	0x0076bb0f,
+	0xf40136b0,
+	0x32f4061b,
+/* 0x0941: i2c_put_byte_done */
+/* 0x0943: i2c_addr */
+	0xbb00f801,
 	0x65b60076,
 	0x9450f904,
 	0x56bb0465,
 	0xfd50bd02,
 	0x50fc0475,
-	0x07fb21f5,
-	0xfc0464b6,
-	0x00d6b0d0,
-	0x00b31bf5,
-	0xbb0057f0,
-	0x65b60076,
-	0x9450f904,
-	0x56bb0465,
-	0xfd50bd02,
-	0x50fc0475,
-	0x07a521f5,
-	0xf50464b6,
-	0xc700d011,
-	0x76bbe0c5,
-	0x0465b600,
-	0x659450f9,
-	0x0256bb04,
-	0x75fd50bd,
-	0xf550fc04,
-	0xb6074a21,
-	0x11f50464,
-	0x57f000ad,
+	0x077f21f5,
+	0xf40464b6,
+	0xc3e72911,
+	0x34b6012e,
+	0x0553fd01,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0xe821f550,
+	0x0464b608,
+/* 0x0988: i2c_addr_done */
+/* 0x098a: i2c_acquire_addr */
+	0xcec700f8,
+	0x02e4b6f8,
+	0x0c10e0b7,
+	0xf800ee98,
+/* 0x0999: i2c_acquire */
+	0x8a21f500,
+	0x0421f409,
+	0xf403d9f0,
+	0x00f83f21,
+/* 0x09a8: i2c_release */
+	0x098a21f5,
+	0xf00421f4,
+	0x21f403da,
+/* 0x09b7: i2c_recv */
+	0xf400f83f,
+	0xc1c70132,
+	0x0214b6f8,
+	0xf52816b0,
+	0xa0013a1f,
+	0x980be813,
+	0x13a00032,
+	0x31980bc0,
+	0x0231f400,
+	0xe0f9d0f9,
+	0x67f1d0f9,
+	0x63f10000,
+	0x67921000,
 	0x0076bb01,
 	0xf90465b6,
 	0x04659450,
 	0xbd0256bb,
 	0x0475fd50,
 	0x21f550fc,
-	0x64b607a5,
-	0x8a11f504,
+	0x64b60999,
+	0xb0d0fc04,
+	0x1bf500d6,
+	0x57f000b3,
 	0x0076bb00,
 	0xf90465b6,
 	0x04659450,
 	0xbd0256bb,
 	0x0475fd50,
 	0x21f550fc,
-	0x64b606f8,
-	0x6a11f404,
-	0xbbe05bcb,
+	0x64b60943,
+	0xd011f504,
+	0xe0c5c700,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0xe821f550,
+	0x0464b608,
+	0x00ad11f5,
+	0xbb0157f0,
 	0x65b60076,
 	0x9450f904,
 	0x56bb0465,
 	0xfd50bd02,
 	0x50fc0475,
-	0x063d21f5,
-	0xb90464b6,
-	0x74bd025b,
-/* 0x091f: i2c_recv_not_rd08 */
-	0xb0430ef4,
-	0x1bf401d6,
-	0x0057f03d,
-	0x07a521f5,
-	0xc73311f4,
-	0x21f5e0c5,
-	0x11f4074a,
-	0x0057f029,
-	0x07a521f5,
-	0xc71f11f4,
-	0x21f5e0b5,
-	0x11f4074a,
-	0x3d21f515,
-	0xc774bd06,
-	0x1bf408c5,
-	0x0232f409,
-/* 0x095f: i2c_recv_not_wr08 */
-/* 0x095f: i2c_recv_done */
-	0xc7030ef4,
-	0x21f5f8ce,
-	0xe0fc080a,
-	0x12f4d0fc,
-	0x027cb90a,
-	0x02b921f5,
-/* 0x0974: i2c_recv_exit */
-/* 0x0976: i2c_init */
-	0x00f800f8,
-/* 0x0978: test_recv */
-	0x05d817f1,
+	0x094321f5,
+	0xf50464b6,
+	0xbb008a11,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x089621f5,
+	0xf40464b6,
+	0x5bcb6a11,
+	0x0076bbe0,
+	0xf90465b6,
+	0x04659450,
+	0xbd0256bb,
+	0x0475fd50,
+	0x21f550fc,
+	0x64b607db,
+	0x025bb904,
+	0x0ef474bd,
+/* 0x0abd: i2c_recv_not_rd08 */
+	0x01d6b043,
+	0xf03d1bf4,
+	0x21f50057,
+	0x11f40943,
+	0xe0c5c733,
+	0x08e821f5,
+	0xf02911f4,
+	0x21f50057,
+	0x11f40943,
+	0xe0b5c71f,
+	0x08e821f5,
+	0xf51511f4,
+	0xbd07db21,
+	0x08c5c774,
+	0xf4091bf4,
+	0x0ef40232,
+/* 0x0afd: i2c_recv_not_wr08 */
+/* 0x0afd: i2c_recv_done */
+	0xf8cec703,
+	0x09a821f5,
+	0xd0fce0fc,
+	0xb90a12f4,
+	0x21f5027c,
+/* 0x0b12: i2c_recv_exit */
+	0x00f80342,
+/* 0x0b14: i2c_init */
+/* 0x0b16: test_recv */
+	0x17f100f8,
+	0x14b605d8,
+	0x0011cf06,
+	0xf10110b6,
+	0xb605d807,
+	0x01d00604,
+	0xf104bd00,
+	0xf1d900e7,
+	0xf5134fe3,
+	0xf8026221,
+/* 0x0b3d: test_init */
+	0x00e7f100,
+	0x6221f508,
+/* 0x0b47: idle_recv */
+	0xf800f802,
+/* 0x0b49: idle */
+	0x0031f400,
+	0x05d417f1,
 	0xcf0614b6,
 	0x10b60011,
-	0xd807f101,
+	0xd407f101,
 	0x0604b605,
 	0xbd0001d0,
-	0x00e7f104,
-	0x4fe3f1d9,
-	0xf521f513,
-/* 0x099f: test_init */
-	0xf100f801,
-	0xf50800e7,
-	0xf801f521,
-/* 0x09a9: idle_recv */
-/* 0x09ab: idle */
-	0xf400f800,
-	0x17f10031,
-	0x14b605d4,
-	0x0011cf06,
-	0xf10110b6,
-	0xb605d407,
-	0x01d00604,
-/* 0x09c7: idle_loop */
-	0xf004bd00,
-	0x32f45817,
-/* 0x09cd: idle_proc */
-/* 0x09cd: idle_proc_exec */
-	0xb910f902,
-	0x21f5021e,
-	0x10fc02c2,
-	0xf40911f4,
-	0x0ef40231,
-/* 0x09e1: idle_proc_next */
-	0x5810b6ef,
-	0xf4061fb8,
-	0x02f4e61b,
-	0x0028f4dd,
-	0x00bb0ef4,
+/* 0x0b65: idle_loop */
+	0x5817f004,
+/* 0x0b6b: idle_proc */
+/* 0x0b6b: idle_proc_exec */
+	0xf90232f4,
+	0x021eb910,
+	0x034b21f5,
+	0x11f410fc,
+	0x0231f409,
+/* 0x0b7f: idle_proc_next */
+	0xb6ef0ef4,
+	0x1fb85810,
+	0xe61bf406,
+	0xf4dd02f4,
+	0x0ef40028,
+	0x000000bb,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvc0.fuc b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvc0.fuc
index 48f79434a449..21bf8cc7618f 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvc0.fuc
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvc0.fuc
@@ -23,6 +23,7 @@
  */
 
 #define NVKM_PPWR_CHIPSET GF100
+#define HW_TICKS_PER_US 203 // should be 202.5
 
 //#define NVKM_FALCON_PC24
 //#define NVKM_FALCON_UNSHIFTED_IO
@@ -34,6 +35,7 @@
 .section #nvc0_pwr_data
 #define INCLUDE_PROC
 #include "kernel.fuc"
+#include "arith.fuc"
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
@@ -44,6 +46,7 @@
 
 #define INCLUDE_DATA
 #include "kernel.fuc"
+#include "arith.fuc"
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
@@ -56,6 +59,7 @@
 .section #nvc0_pwr_code
 #define INCLUDE_CODE
 #include "kernel.fuc"
+#include "arith.fuc"
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvc0.fuc.h b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvc0.fuc.h
index 0773ff0e3dc3..ca30fa4011b5 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvc0.fuc.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvc0.fuc.h
@@ -24,8 +24,8 @@ uint32_t nvc0_pwr_data[] = {
 	0x00000000,
 /* 0x0058: proc_list_head */
 	0x54534f48,
-	0x00000430,
-	0x000003cd,
+	0x00000512,
+	0x000004af,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -46,8 +46,8 @@ uint32_t nvc0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x584d454d,
-	0x00000542,
-	0x00000534,
+	0x0000074b,
+	0x0000073d,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -68,8 +68,8 @@ uint32_t nvc0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x46524550,
-	0x00000546,
-	0x00000544,
+	0x0000074f,
+	0x0000074d,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -90,8 +90,8 @@ uint32_t nvc0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x5f433249,
-	0x00000976,
-	0x00000819,
+	0x00000b7f,
+	0x00000a22,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -112,8 +112,8 @@ uint32_t nvc0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x54534554,
-	0x0000099f,
-	0x00000978,
+	0x00000ba8,
+	0x00000b81,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -134,8 +134,8 @@ uint32_t nvc0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x454c4449,
-	0x000009ab,
-	0x000009a9,
+	0x00000bb4,
+	0x00000bb2,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -227,25 +227,31 @@ uint32_t nvc0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 /* 0x0370: memx_func_head */
-	0x00010000,
-	0x00000000,
-	0x0000046f,
-/* 0x037c: memx_func_next */
 	0x00000001,
 	0x00000000,
-	0x00000496,
+	0x00000551,
+/* 0x037c: memx_func_next */
 	0x00000002,
+	0x00000000,
+	0x000005db,
+	0x00000003,
 	0x00000002,
-	0x000004b7,
-	0x00040003,
+	0x000006a5,
+	0x00040004,
+	0x00000000,
+	0x000006c1,
+	0x00010005,
+	0x00000000,
+	0x000006de,
+	0x00010006,
 	0x00000000,
-	0x000004d3,
-	0x00010004,
+	0x00000663,
+/* 0x03b8: memx_func_tail */
+/* 0x03b8: memx_ts_start */
 	0x00000000,
-	0x000004f0,
-/* 0x03ac: memx_func_tail */
-/* 0x03ac: memx_data_head */
+/* 0x03bc: memx_ts_end */
 	0x00000000,
+/* 0x03c0: memx_data_head */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -757,8 +763,9 @@ uint32_t nvc0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-/* 0x0bac: memx_data_tail */
-/* 0x0bac: i2c_scl_map */
+	0x00000000,
+/* 0x0bc0: memx_data_tail */
+/* 0x0bc0: i2c_scl_map */
 	0x00001000,
 	0x00004000,
 	0x00010000,
@@ -769,7 +776,7 @@ uint32_t nvc0_pwr_data[] = {
 	0x01000000,
 	0x04000000,
 	0x10000000,
-/* 0x0bd4: i2c_sda_map */
+/* 0x0be8: i2c_sda_map */
 	0x00002000,
 	0x00008000,
 	0x00020000,
@@ -780,7 +787,7 @@ uint32_t nvc0_pwr_data[] = {
 	0x02000000,
 	0x08000000,
 	0x20000000,
-/* 0x0bfc: i2c_ctrl */
+/* 0x0c10: i2c_ctrl */
 	0x0000e138,
 	0x0000e150,
 	0x0000e168,
@@ -841,15 +848,10 @@ uint32_t nvc0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
 };
 
 uint32_t nvc0_pwr_code[] = {
-	0x030d0ef5,
+	0x039e0ef5,
 /* 0x0004: rd32 */
 	0x07a007f1,
 	0xd00604b6,
@@ -885,19 +887,22 @@ uint32_t nvc0_pwr_code[] = {
 	0xd4f100dd,
 	0x1bf47000,
 /* 0x007f: nsec */
-	0xf000f8f2,
+	0xf900f8f2,
+	0xf080f990,
 	0x84b62c87,
 	0x0088cf06,
-/* 0x0088: nsec_loop */
+/* 0x008c: nsec_loop */
 	0xb62c97f0,
 	0x99cf0694,
 	0x0298bb00,
 	0xf4069eb8,
-	0x00f8f11e,
-/* 0x009c: wait */
+	0x80fcf11e,
+	0x00f890fc,
+/* 0x00a4: wait */
+	0x80f990f9,
 	0xb62c87f0,
 	0x88cf0684,
-/* 0x00a5: wait_loop */
+/* 0x00b1: wait_loop */
 	0x02eeb900,
 	0xb90421f4,
 	0xadfd02da,
@@ -907,28 +912,29 @@ uint32_t nvc0_pwr_code[] = {
 	0x0099cf06,
 	0xb80298bb,
 	0x1ef4069b,
-/* 0x00c9: wait_done */
-/* 0x00cb: intr_watchdog */
-	0x9800f8df,
+/* 0x00d5: wait_done */
+	0xfc80fcdf,
+/* 0x00db: intr_watchdog */
+	0x9800f890,
 	0x96b003e9,
 	0x2a0bf400,
 	0xbb9a0a98,
 	0x1cf4029a,
 	0x01d7f00f,
-	0x025421f5,
+	0x02dd21f5,
 	0x0ef494bd,
-/* 0x00e9: intr_watchdog_next_time */
+/* 0x00f9: intr_watchdog_next_time */
 	0x9b0a9815,
 	0xf400a6b0,
 	0x9ab8090b,
 	0x061cf406,
-/* 0x00f8: intr_watchdog_next_time_set */
-/* 0x00fb: intr_watchdog_next_proc */
+/* 0x0108: intr_watchdog_next_time_set */
+/* 0x010b: intr_watchdog_next_proc */
 	0x809b0980,
 	0xe0b603e9,
 	0x68e6b158,
 	0xc61bf402,
-/* 0x010a: intr */
+/* 0x011a: intr */
 	0x00f900f8,
 	0x80f904bd,
 	0xa0f990f9,
@@ -948,13 +954,13 @@ uint32_t nvc0_pwr_code[] = {
 	0xf40289c4,
 	0x0080230b,
 	0x58e7f09b,
-	0x98cb21f4,
+	0x98db21f4,
 	0x96b09b09,
 	0x110bf400,
 	0xb63407f0,
 	0x09d00604,
 	0x8004bd00,
-/* 0x016e: intr_skip_watchdog */
+/* 0x017e: intr_skip_watchdog */
 	0x89e49a09,
 	0x0bf40800,
 	0x8897f148,
@@ -967,22 +973,22 @@ uint32_t nvc0_pwr_code[] = {
 	0x48e7f1c0,
 	0x53e3f14f,
 	0x00d7f054,
-	0x02b921f5,
+	0x034221f5,
 	0x07f1c0fc,
 	0x04b604c0,
 	0x000cd006,
-/* 0x01ae: intr_subintr_skip_fifo */
+/* 0x01be: intr_subintr_skip_fifo */
 	0x07f104bd,
 	0x04b60688,
 	0x0009d006,
-/* 0x01ba: intr_skip_subintr */
+/* 0x01ca: intr_skip_subintr */
 	0x89c404bd,
 	0x070bf420,
 	0xffbfa4f1,
-/* 0x01c4: intr_skip_pause */
+/* 0x01d4: intr_skip_pause */
 	0xf44089c4,
 	0xa4f1070b,
-/* 0x01ce: intr_skip_user0 */
+/* 0x01de: intr_skip_user0 */
 	0x07f0ffbf,
 	0x0604b604,
 	0xbd0008d0,
@@ -993,597 +999,733 @@ uint32_t nvc0_pwr_code[] = {
 	0x90fca0fc,
 	0x00fc80fc,
 	0xf80032f4,
-/* 0x01f5: timer */
-	0x1032f401,
-	0xb003f898,
-	0x1cf40086,
-	0x03fe8051,
+/* 0x0205: ticks_from_ns */
+	0xf9c0f901,
+	0xcbd7f1b0,
+	0x00d3f000,
+	0x041321f5,
+	0x03e8ccec,
+	0xf400b4b0,
+	0xeeec120b,
+	0xd7f103e8,
+	0xd3f000cb,
+	0x1321f500,
+/* 0x022d: ticks_from_ns_quit */
+	0x02ceb904,
+	0xc0fcb0fc,
+/* 0x0236: ticks_from_us */
+	0xc0f900f8,
+	0xd7f1b0f9,
+	0xd3f000cb,
+	0x1321f500,
+	0x02ceb904,
+	0xf400b4b0,
+	0xe4bd050b,
+/* 0x0250: ticks_from_us_quit */
+	0xc0fcb0fc,
+/* 0x0256: ticks_to_us */
+	0xd7f100f8,
+	0xd3f000cb,
+	0xecedff00,
+/* 0x0262: timer */
+	0x90f900f8,
+	0x32f480f9,
+	0x03f89810,
+	0xf40086b0,
+	0x84bd651c,
 	0xb63807f0,
 	0x08d00604,
 	0xf004bd00,
-	0x84b60887,
+	0x84b63487,
 	0x0088cf06,
-	0xf40284f0,
-	0x87f0261b,
-	0x0684b634,
-	0xb80088cf,
-	0x0bf406e0,
-	0x06e8b809,
-/* 0x0233: timer_reset */
-	0xf01f1ef4,
-	0x04b63407,
-	0x000ed006,
-	0x0e8004bd,
-/* 0x0241: timer_enable */
-	0x0187f09a,
+	0xbb9a0998,
+	0xe9bb0298,
+	0x03fe8000,
+	0xb60887f0,
+	0x88cf0684,
+	0x0284f000,
+	0xf0261bf4,
+	0x84b63487,
+	0x0088cf06,
+	0xf406e0b8,
+	0xe8b8090b,
+	0x111cf406,
+/* 0x02b8: timer_reset */
+	0xb63407f0,
+	0x0ed00604,
+	0x8004bd00,
+/* 0x02c6: timer_enable */
+	0x87f09a0e,
+	0x3807f001,
+	0xd00604b6,
+	0x04bd0008,
+/* 0x02d4: timer_done */
+	0xfc1031f4,
+	0xf890fc80,
+/* 0x02dd: send_proc */
+	0xf980f900,
+	0x05e89890,
+	0xf004e998,
+	0x89b80486,
+	0x2a0bf406,
+	0x940398c4,
+	0x80b60488,
+	0x008ebb18,
+	0x8000fa98,
+	0x8d80008a,
+	0x028c8001,
+	0xb6038b80,
+	0x94f00190,
+	0x04e98007,
+/* 0x0317: send_done */
+	0xfc0231f4,
+	0xf880fc90,
+/* 0x031d: find */
+	0xf080f900,
+	0x31f45887,
+/* 0x0325: find_loop */
+	0x008a9801,
+	0xf406aeb8,
+	0x80b6100b,
+	0x6886b158,
+	0xf01bf402,
+/* 0x033b: find_done */
+	0xb90132f4,
+	0x80fc028e,
+/* 0x0342: send */
+	0x21f500f8,
+	0x01f4031d,
+/* 0x034b: recv */
+	0xf900f897,
+	0x9880f990,
+	0xe99805e8,
+	0x0132f404,
+	0xf40689b8,
+	0x89c43d0b,
+	0x0180b603,
+	0x800784f0,
+	0xea9805e8,
+	0xfef0f902,
+	0xf0f9018f,
+	0x9402efb9,
+	0xe9bb0499,
+	0x18e0b600,
+	0x9803eb98,
+	0xed9802ec,
+	0x00ee9801,
+	0xf0fca5f9,
+	0xf400f8fe,
+	0xf0fc0131,
+/* 0x0398: recv_done */
+	0x90fc80fc,
+/* 0x039e: init */
+	0x17f100f8,
+	0x14b60108,
+	0x0011cf06,
+	0x010911e7,
+	0xfe0814b6,
+	0x17f10014,
+	0x13f000e0,
+	0x1c07f000,
+	0xd00604b6,
+	0x04bd0001,
+	0xf0ff17f0,
+	0x04b61407,
+	0x0001d006,
+	0x17f004bd,
+	0x0015f102,
+	0x1007f008,
+	0xd00604b6,
+	0x04bd0001,
+	0x011a17f1,
+	0xfe0013f0,
+	0x31f40010,
+	0x0117f010,
 	0xb63807f0,
-	0x08d00604,
-/* 0x024f: timer_done */
-	0xf404bd00,
-	0x00f81031,
-/* 0x0254: send_proc */
-	0x90f980f9,
-	0x9805e898,
-	0x86f004e9,
-	0x0689b804,
-	0xc42a0bf4,
-	0x88940398,
-	0x1880b604,
-	0x98008ebb,
-	0x8a8000fa,
-	0x018d8000,
-	0x80028c80,
-	0x90b6038b,
-	0x0794f001,
-	0xf404e980,
-/* 0x028e: send_done */
-	0x90fc0231,
-	0x00f880fc,
-/* 0x0294: find */
-	0x87f080f9,
-	0x0131f458,
-/* 0x029c: find_loop */
-	0xb8008a98,
-	0x0bf406ae,
-	0x5880b610,
-	0x026886b1,
-	0xf4f01bf4,
-/* 0x02b2: find_done */
-	0x8eb90132,
-	0xf880fc02,
-/* 0x02b9: send */
-	0x9421f500,
-	0x9701f402,
-/* 0x02c2: recv */
-	0xe89800f8,
-	0x04e99805,
-	0xb80132f4,
-	0x0bf40689,
-	0x0389c43d,
-	0xf00180b6,
-	0xe8800784,
-	0x02ea9805,
-	0x8ffef0f9,
-	0xb9f0f901,
-	0x999402ef,
-	0x00e9bb04,
-	0x9818e0b6,
-	0xec9803eb,
-	0x01ed9802,
-	0xf900ee98,
-	0xfef0fca5,
-	0x31f400f8,
-/* 0x030b: recv_done */
-	0xf8f0fc01,
-/* 0x030d: init */
-	0x0817f100,
-	0x0614b601,
-	0xe70011cf,
-	0xb6010911,
-	0x14fe0814,
-	0xe017f100,
-	0x0013f000,
-	0xb61c07f0,
 	0x01d00604,
 	0xf004bd00,
-	0x07f0ff17,
-	0x0604b614,
-	0xbd0001d0,
-	0x0217f004,
-	0x080015f1,
-	0xb61007f0,
-	0x01d00604,
-	0xf104bd00,
-	0xf0010a17,
-	0x10fe0013,
-	0x1031f400,
-	0xf00117f0,
-	0x04b63807,
-	0x0001d006,
-	0xf7f004bd,
-/* 0x0371: init_proc */
-	0x01f19858,
-	0xf40016b0,
-	0x15f9fa0b,
-	0xf458f0b6,
-/* 0x0382: host_send */
-	0x17f1f20e,
-	0x14b604b0,
-	0x0011cf06,
-	0x04a027f1,
-	0xcf0624b6,
-	0x12b80022,
-	0x320bf406,
-	0x94071ec4,
-	0xe0b704ee,
-	0xeb980270,
-	0x02ec9803,
-	0x9801ed98,
-	0x21f500ee,
-	0x10b602b9,
-	0x0f1ec401,
-	0x04b007f1,
-	0xd00604b6,
-	0x04bd000e,
-/* 0x03cb: host_send_done */
-	0xf8ba0ef4,
-/* 0x03cd: host_recv */
-	0x4917f100,
-	0x5413f14e,
-	0x06e1b852,
-/* 0x03db: host_recv_wait */
-	0xf1aa0bf4,
-	0xb604cc17,
-	0x11cf0614,
-	0xc827f100,
-	0x0624b604,
-	0xf00022cf,
-	0x12b80816,
-	0xe60bf406,
-	0xb60723c4,
-	0x30b70434,
-	0x3b8002f0,
-	0x023c8003,
-	0x80013d80,
-	0x20b6003e,
-	0x0f24f001,
-	0x04c807f1,
+/* 0x0402: init_proc */
+	0xf19858f7,
+	0x0016b001,
+	0xf9fa0bf4,
+	0x58f0b615,
+/* 0x0413: mulu32_32_64 */
+	0xf9f20ef4,
+	0xf920f910,
+	0x9540f930,
+	0xd29510e1,
+	0xbdc4bd10,
+	0xc0edffb4,
+	0xb9301dff,
+	0x34f10234,
+	0x34b6ffff,
+	0x1045b610,
+	0xbb00c3bb,
+	0xe2ff01b4,
+	0x0234b930,
+	0xffff34f1,
+	0xb61034b6,
+	0xc3bb1045,
+	0x01b4bb00,
+	0xbb3012ff,
+	0x40fc00b3,
+	0x20fc30fc,
+	0x00f810fc,
+/* 0x0464: host_send */
+	0x04b017f1,
+	0xcf0614b6,
+	0x27f10011,
+	0x24b604a0,
+	0x0022cf06,
+	0xf40612b8,
+	0x1ec4320b,
+	0x04ee9407,
+	0x0270e0b7,
+	0x9803eb98,
+	0xed9802ec,
+	0x00ee9801,
+	0x034221f5,
+	0xc40110b6,
+	0x07f10f1e,
+	0x04b604b0,
+	0x000ed006,
+	0x0ef404bd,
+/* 0x04ad: host_send_done */
+/* 0x04af: host_recv */
+	0xf100f8ba,
+	0xf14e4917,
+	0xb8525413,
+	0x0bf406e1,
+/* 0x04bd: host_recv_wait */
+	0xcc17f1aa,
+	0x0614b604,
+	0xf10011cf,
+	0xb604c827,
+	0x22cf0624,
+	0x0816f000,
+	0xf40612b8,
+	0x23c4e60b,
+	0x0434b607,
+	0x02f030b7,
+	0x80033b80,
+	0x3d80023c,
+	0x003e8001,
+	0xf00120b6,
+	0x07f10f24,
+	0x04b604c8,
+	0x0002d006,
+	0x27f004bd,
+	0x0007f040,
 	0xd00604b6,
 	0x04bd0002,
-	0xf04027f0,
-	0x04b60007,
-	0x0002d006,
-	0x00f804bd,
-/* 0x0430: host_init */
-	0x008017f1,
-	0xf11014b6,
-	0xf1027015,
-	0xb604d007,
-	0x01d00604,
-	0xf104bd00,
-	0xb6008017,
-	0x15f11014,
-	0x07f102f0,
-	0x04b604dc,
-	0x0001d006,
-	0x17f004bd,
-	0xc407f101,
+/* 0x0512: host_init */
+	0x17f100f8,
+	0x14b60080,
+	0x7015f110,
+	0xd007f102,
 	0x0604b604,
 	0xbd0001d0,
-/* 0x046f: memx_func_enter */
-	0xf000f804,
+	0x8017f104,
+	0x1014b600,
+	0x02f015f1,
+	0x04dc07f1,
+	0xd00604b6,
+	0x04bd0001,
+	0xf10117f0,
+	0xb604c407,
+	0x01d00604,
+	0xf804bd00,
+/* 0x0551: memx_func_enter */
+	0x2067f100,
+	0x5d77f116,
+	0xff73f1f5,
+	0x026eb9ff,
+	0xb90421f4,
+	0x87fd02d8,
+	0xf960f904,
+	0xfcd0fc80,
+	0x3f21f4e0,
+	0xfffe77f1,
+	0xffff73f1,
+	0xf4026eb9,
+	0xd8b90421,
+	0x0487fd02,
+	0x80f960f9,
+	0xe0fcd0fc,
+	0xf13f21f4,
+	0xb926f067,
+	0x21f4026e,
+	0x02d8b904,
+	0xf90487fd,
+	0xfc80f960,
+	0xf4e0fcd0,
+	0x67f03f21,
+	0xe007f104,
+	0x0604b607,
+	0xbd0006d0,
+/* 0x05bd: memx_func_enter_wait */
+	0xc067f104,
+	0x0664b607,
+	0xf00066cf,
+	0x0bf40464,
+	0x2c67f0f3,
+	0xcf0664b6,
+	0x06800066,
+/* 0x05db: memx_func_leave */
+	0xf000f8ee,
+	0x64b62c67,
+	0x0066cf06,
+	0xf0ef0680,
 	0x07f10467,
-	0x04b607e0,
+	0x04b607e4,
 	0x0006d006,
-/* 0x047e: memx_func_enter_wait */
+/* 0x05f6: memx_func_leave_wait */
 	0x67f104bd,
 	0x64b607c0,
 	0x0066cf06,
 	0xf40464f0,
-	0x1698f30b,
+	0x67f1f31b,
+	0x77f126f0,
+	0x73f00001,
+	0x026eb900,
+	0xb90421f4,
+	0x87fd02d8,
+	0xf960f905,
+	0xfcd0fc80,
+	0x3f21f4e0,
+	0x162067f1,
+	0xf4026eb9,
+	0xd8b90421,
+	0x0587fd02,
+	0x80f960f9,
+	0xe0fcd0fc,
+	0xf13f21f4,
+	0xf00aa277,
+	0x6eb90073,
+	0x0421f402,
+	0xfd02d8b9,
+	0x60f90587,
+	0xd0fc80f9,
+	0x21f4e0fc,
+/* 0x0663: memx_func_wait_vblank */
+	0x9800f83f,
+	0x66b00016,
+	0x130bf400,
+	0xf40166b0,
+	0x0ef4060b,
+/* 0x0675: memx_func_wait_vblank_head1 */
+	0x2077f12e,
+	0x070ef400,
+/* 0x067c: memx_func_wait_vblank_head0 */
+	0x000877f1,
+/* 0x0680: memx_func_wait_vblank_0 */
+	0x07c467f1,
+	0xcf0664b6,
+	0x67fd0066,
+	0xf31bf404,
+/* 0x0690: memx_func_wait_vblank_1 */
+	0x07c467f1,
+	0xcf0664b6,
+	0x67fd0066,
+	0xf30bf404,
+/* 0x06a0: memx_func_wait_vblank_fini */
+	0xf80410b6,
+/* 0x06a5: memx_func_wr32 */
+	0x00169800,
+	0xb6011598,
+	0x60f90810,
+	0xd0fc50f9,
+	0x21f4e0fc,
+	0x0242b63f,
+	0xf8e91bf4,
+/* 0x06c1: memx_func_wait */
+	0x2c87f000,
+	0xcf0684b6,
+	0x1e980088,
+	0x011d9800,
+	0x98021c98,
+	0x10b6031b,
+	0xa421f410,
+/* 0x06de: memx_func_delay */
+	0x1e9800f8,
 	0x0410b600,
-/* 0x0496: memx_func_leave */
-	0x67f000f8,
-	0xe407f104,
+	0xf87f21f4,
+/* 0x06e9: memx_exec */
+	0xf9e0f900,
+	0x02c1b9d0,
+/* 0x06f3: memx_exec_next */
+	0x9802b2b9,
+	0x10b60013,
+	0xf034e704,
+	0xe033e701,
+	0x0132b601,
+	0x980c30f0,
+	0x55f9de35,
+	0xf40612b8,
+	0x0b98e41e,
+	0xef0c98ee,
+	0xf102cbbb,
+	0xb607c4b7,
+	0xbbcf06b4,
+	0xfcd0fc00,
+	0x4221f5e0,
+/* 0x072f: memx_info */
+	0xf100f803,
+	0xf103c0c7,
+	0xf50800b7,
+	0xf8034221,
+/* 0x073d: memx_recv */
+	0x01d6b000,
+	0xb0a90bf4,
+	0x0bf400d6,
+/* 0x074b: memx_init */
+	0xf800f8e9,
+/* 0x074d: perf_recv */
+/* 0x074f: perf_init */
+	0xf800f800,
+/* 0x0751: i2c_drive_scl */
+	0x0036b000,
+	0xf1110bf4,
+	0xb607e007,
+	0x01d00604,
+	0xf804bd00,
+/* 0x0765: i2c_drive_scl_lo */
+	0xe407f100,
 	0x0604b607,
-	0xbd0006d0,
-/* 0x04a5: memx_func_leave_wait */
-	0xc067f104,
-	0x0664b607,
-	0xf00066cf,
-	0x1bf40464,
-/* 0x04b7: memx_func_wr32 */
-	0x9800f8f3,
-	0x15980016,
-	0x0810b601,
-	0x50f960f9,
-	0xe0fcd0fc,
-	0xb63f21f4,
-	0x1bf40242,
-/* 0x04d3: memx_func_wait */
-	0xf000f8e9,
-	0x84b62c87,
-	0x0088cf06,
-	0x98001e98,
-	0x1c98011d,
-	0x031b9802,
-	0xf41010b6,
-	0x00f89c21,
-/* 0x04f0: memx_func_delay */
-	0xb6001e98,
-	0x21f40410,
-/* 0x04fb: memx_exec */
-	0xf900f87f,
-	0xb9d0f9e0,
-	0xb2b902c1,
-/* 0x0505: memx_exec_next */
-	0x00139802,
-	0x950410b6,
-	0x30f01034,
-	0xde35980c,
-	0x12b855f9,
-	0xec1ef406,
-	0xe0fcd0fc,
-	0x02b921f5,
-/* 0x0526: memx_info */
-	0xc7f100f8,
-	0xb7f103ac,
-	0x21f50800,
-	0x00f802b9,
-/* 0x0534: memx_recv */
-	0xf401d6b0,
-	0xd6b0c40b,
-	0xe90bf400,
-/* 0x0542: memx_init */
-	0x00f800f8,
-/* 0x0544: perf_recv */
-/* 0x0546: perf_init */
-	0x00f800f8,
-/* 0x0548: i2c_drive_scl */
-	0xf40036b0,
-	0x07f1110b,
-	0x04b607e0,
-	0x0001d006,
-	0x00f804bd,
-/* 0x055c: i2c_drive_scl_lo */
-	0x07e407f1,
-	0xd00604b6,
-	0x04bd0001,
-/* 0x056a: i2c_drive_sda */
-	0x36b000f8,
-	0x110bf400,
-	0x07e007f1,
-	0xd00604b6,
-	0x04bd0002,
-/* 0x057e: i2c_drive_sda_lo */
-	0x07f100f8,
-	0x04b607e4,
-	0x0002d006,
-	0x00f804bd,
-/* 0x058c: i2c_sense_scl */
-	0xf10132f4,
-	0xb607c437,
-	0x33cf0634,
-	0x0431fd00,
-	0xf4060bf4,
-/* 0x05a2: i2c_sense_scl_done */
-	0x00f80131,
-/* 0x05a4: i2c_sense_sda */
-	0xf10132f4,
-	0xb607c437,
-	0x33cf0634,
-	0x0432fd00,
-	0xf4060bf4,
-/* 0x05ba: i2c_sense_sda_done */
-	0x00f80131,
-/* 0x05bc: i2c_raise_scl */
-	0x47f140f9,
-	0x37f00898,
-	0x4821f501,
-/* 0x05c9: i2c_raise_scl_wait */
-	0xe8e7f105,
-	0x7f21f403,
-	0x058c21f5,
-	0xb60901f4,
-	0x1bf40142,
-/* 0x05dd: i2c_raise_scl_done */
-	0xf840fcef,
-/* 0x05e1: i2c_start */
-	0x8c21f500,
-	0x0d11f405,
-	0x05a421f5,
-	0xf40611f4,
-/* 0x05f2: i2c_start_rep */
-	0x37f0300e,
-	0x4821f500,
-	0x0137f005,
-	0x056a21f5,
-	0xb60076bb,
-	0x50f90465,
-	0xbb046594,
-	0x50bd0256,
-	0xfc0475fd,
-	0xbc21f550,
-	0x0464b605,
-/* 0x061f: i2c_start_send */
-	0xf01f11f4,
+	0xbd0001d0,
+/* 0x0773: i2c_drive_sda */
+	0xb000f804,
+	0x0bf40036,
+	0xe007f111,
+	0x0604b607,
+	0xbd0002d0,
+/* 0x0787: i2c_drive_sda_lo */
+	0xf100f804,
+	0xb607e407,
+	0x02d00604,
+	0xf804bd00,
+/* 0x0795: i2c_sense_scl */
+	0x0132f400,
+	0x07c437f1,
+	0xcf0634b6,
+	0x31fd0033,
+	0x060bf404,
+/* 0x07ab: i2c_sense_scl_done */
+	0xf80131f4,
+/* 0x07ad: i2c_sense_sda */
+	0x0132f400,
+	0x07c437f1,
+	0xcf0634b6,
+	0x32fd0033,
+	0x060bf404,
+/* 0x07c3: i2c_sense_sda_done */
+	0xf80131f4,
+/* 0x07c5: i2c_raise_scl */
+	0xf140f900,
+	0xf0089847,
+	0x21f50137,
+/* 0x07d2: i2c_raise_scl_wait */
+	0xe7f10751,
+	0x21f403e8,
+	0x9521f57f,
+	0x0901f407,
+	0xf40142b6,
+/* 0x07e6: i2c_raise_scl_done */
+	0x40fcef1b,
+/* 0x07ea: i2c_start */
+	0x21f500f8,
+	0x11f40795,
+	0xad21f50d,
+	0x0611f407,
+/* 0x07fb: i2c_start_rep */
+	0xf0300ef4,
 	0x21f50037,
-	0xe7f1056a,
-	0x21f41388,
-	0x0037f07f,
-	0x054821f5,
-	0x1388e7f1,
-/* 0x063b: i2c_start_out */
-	0xf87f21f4,
-/* 0x063d: i2c_stop */
-	0x0037f000,
-	0x054821f5,
+	0x37f00751,
+	0x7321f501,
+	0x0076bb07,
+	0xf90465b6,
+	0x04659450,
+	0xbd0256bb,
+	0x0475fd50,
+	0x21f550fc,
+	0x64b607c5,
+	0x1f11f404,
+/* 0x0828: i2c_start_send */
 	0xf50037f0,
-	0xf1056a21,
-	0xf403e8e7,
+	0xf1077321,
+	0xf41388e7,
 	0x37f07f21,
-	0x4821f501,
-	0x88e7f105,
+	0x5121f500,
+	0x88e7f107,
 	0x7f21f413,
-	0xf50137f0,
-	0xf1056a21,
-	0xf41388e7,
-	0x00f87f21,
-/* 0x0670: i2c_bitw */
-	0x056a21f5,
+/* 0x0844: i2c_start_out */
+/* 0x0846: i2c_stop */
+	0x37f000f8,
+	0x5121f500,
+	0x0037f007,
+	0x077321f5,
 	0x03e8e7f1,
-	0xbb7f21f4,
-	0x65b60076,
-	0x9450f904,
-	0x56bb0465,
-	0xfd50bd02,
-	0x50fc0475,
-	0x05bc21f5,
-	0xf40464b6,
-	0xe7f11811,
+	0xf07f21f4,
+	0x21f50137,
+	0xe7f10751,
 	0x21f41388,
-	0x0037f07f,
-	0x054821f5,
+	0x0137f07f,
+	0x077321f5,
 	0x1388e7f1,
-/* 0x06af: i2c_bitw_out */
 	0xf87f21f4,
-/* 0x06b1: i2c_bitr */
-	0x0137f000,
-	0x056a21f5,
-	0x03e8e7f1,
-	0xbb7f21f4,
-	0x65b60076,
-	0x9450f904,
-	0x56bb0465,
-	0xfd50bd02,
-	0x50fc0475,
-	0x05bc21f5,
-	0xf40464b6,
-	0x21f51b11,
-	0x37f005a4,
-	0x4821f500,
-	0x88e7f105,
+/* 0x0879: i2c_bitw */
+	0x7321f500,
+	0xe8e7f107,
+	0x7f21f403,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0xc521f550,
+	0x0464b607,
+	0xf11811f4,
+	0xf41388e7,
+	0x37f07f21,
+	0x5121f500,
+	0x88e7f107,
 	0x7f21f413,
-	0xf4013cf0,
-/* 0x06f6: i2c_bitr_done */
-	0x00f80131,
-/* 0x06f8: i2c_get_byte */
-	0xf00057f0,
-/* 0x06fe: i2c_get_byte_next */
-	0x54b60847,
+/* 0x08b8: i2c_bitw_out */
+/* 0x08ba: i2c_bitr */
+	0x37f000f8,
+	0x7321f501,
+	0xe8e7f107,
+	0x7f21f403,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0xc521f550,
+	0x0464b607,
+	0xf51b11f4,
+	0xf007ad21,
+	0x21f50037,
+	0xe7f10751,
+	0x21f41388,
+	0x013cf07f,
+/* 0x08ff: i2c_bitr_done */
+	0xf80131f4,
+/* 0x0901: i2c_get_byte */
+	0x0057f000,
+/* 0x0907: i2c_get_byte_next */
+	0xb60847f0,
+	0x76bb0154,
+	0x0465b600,
+	0x659450f9,
+	0x0256bb04,
+	0x75fd50bd,
+	0xf550fc04,
+	0xb608ba21,
+	0x11f40464,
+	0x0553fd2b,
+	0xf40142b6,
+	0x37f0d81b,
 	0x0076bb01,
 	0xf90465b6,
 	0x04659450,
 	0xbd0256bb,
 	0x0475fd50,
 	0x21f550fc,
-	0x64b606b1,
-	0x2b11f404,
-	0xb60553fd,
-	0x1bf40142,
-	0x0137f0d8,
-	0xb60076bb,
-	0x50f90465,
-	0xbb046594,
-	0x50bd0256,
-	0xfc0475fd,
-	0x7021f550,
-	0x0464b606,
-/* 0x0748: i2c_get_byte_done */
-/* 0x074a: i2c_put_byte */
-	0x47f000f8,
-/* 0x074d: i2c_put_byte_next */
-	0x0142b608,
-	0xbb3854ff,
-	0x65b60076,
-	0x9450f904,
-	0x56bb0465,
-	0xfd50bd02,
-	0x50fc0475,
-	0x067021f5,
-	0xf40464b6,
-	0x46b03411,
-	0xd81bf400,
+	0x64b60879,
+/* 0x0951: i2c_get_byte_done */
+/* 0x0953: i2c_put_byte */
+	0xf000f804,
+/* 0x0956: i2c_put_byte_next */
+	0x42b60847,
+	0x3854ff01,
 	0xb60076bb,
 	0x50f90465,
 	0xbb046594,
 	0x50bd0256,
 	0xfc0475fd,
-	0xb121f550,
-	0x0464b606,
-	0xbb0f11f4,
-	0x36b00076,
-	0x061bf401,
-/* 0x07a3: i2c_put_byte_done */
-	0xf80132f4,
-/* 0x07a5: i2c_addr */
-	0x0076bb00,
+	0x7921f550,
+	0x0464b608,
+	0xb03411f4,
+	0x1bf40046,
+	0x0076bbd8,
 	0xf90465b6,
 	0x04659450,
 	0xbd0256bb,
 	0x0475fd50,
 	0x21f550fc,
-	0x64b605e1,
-	0x2911f404,
-	0x012ec3e7,
-	0xfd0134b6,
-	0x76bb0553,
+	0x64b608ba,
+	0x0f11f404,
+	0xb00076bb,
+	0x1bf40136,
+	0x0132f406,
+/* 0x09ac: i2c_put_byte_done */
+/* 0x09ae: i2c_addr */
+	0x76bb00f8,
 	0x0465b600,
 	0x659450f9,
 	0x0256bb04,
 	0x75fd50bd,
 	0xf550fc04,
-	0xb6074a21,
-/* 0x07ea: i2c_addr_done */
-	0x00f80464,
-/* 0x07ec: i2c_acquire_addr */
-	0xb6f8cec7,
-	0xe0b702e4,
-	0xee980bfc,
-/* 0x07fb: i2c_acquire */
-	0xf500f800,
-	0xf407ec21,
-	0xd9f00421,
-	0x3f21f403,
-/* 0x080a: i2c_release */
-	0x21f500f8,
-	0x21f407ec,
-	0x03daf004,
-	0xf83f21f4,
-/* 0x0819: i2c_recv */
-	0x0132f400,
-	0xb6f8c1c7,
-	0x16b00214,
-	0x3a1ff528,
-	0xd413a001,
-	0x0032980b,
-	0x0bac13a0,
-	0xf4003198,
-	0xd0f90231,
-	0xd0f9e0f9,
-	0x000067f1,
-	0x100063f1,
-	0xbb016792,
+	0xb607ea21,
+	0x11f40464,
+	0x2ec3e729,
+	0x0134b601,
+	0xbb0553fd,
 	0x65b60076,
 	0x9450f904,
 	0x56bb0465,
 	0xfd50bd02,
 	0x50fc0475,
-	0x07fb21f5,
-	0xfc0464b6,
-	0x00d6b0d0,
-	0x00b31bf5,
-	0xbb0057f0,
+	0x095321f5,
+/* 0x09f3: i2c_addr_done */
+	0xf80464b6,
+/* 0x09f5: i2c_acquire_addr */
+	0xf8cec700,
+	0xb702e4b6,
+	0x980c10e0,
+	0x00f800ee,
+/* 0x0a04: i2c_acquire */
+	0x09f521f5,
+	0xf00421f4,
+	0x21f403d9,
+/* 0x0a13: i2c_release */
+	0xf500f83f,
+	0xf409f521,
+	0xdaf00421,
+	0x3f21f403,
+/* 0x0a22: i2c_recv */
+	0x32f400f8,
+	0xf8c1c701,
+	0xb00214b6,
+	0x1ff52816,
+	0x13a0013a,
+	0x32980be8,
+	0xc013a000,
+	0x0031980b,
+	0xf90231f4,
+	0xf9e0f9d0,
+	0x0067f1d0,
+	0x0063f100,
+	0x01679210,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0x0421f550,
+	0x0464b60a,
+	0xd6b0d0fc,
+	0xb31bf500,
+	0x0057f000,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0xae21f550,
+	0x0464b609,
+	0x00d011f5,
+	0xbbe0c5c7,
 	0x65b60076,
 	0x9450f904,
 	0x56bb0465,
 	0xfd50bd02,
 	0x50fc0475,
-	0x07a521f5,
+	0x095321f5,
 	0xf50464b6,
-	0xc700d011,
-	0x76bbe0c5,
+	0xf000ad11,
+	0x76bb0157,
 	0x0465b600,
 	0x659450f9,
 	0x0256bb04,
 	0x75fd50bd,
 	0xf550fc04,
-	0xb6074a21,
+	0xb609ae21,
 	0x11f50464,
-	0x57f000ad,
-	0x0076bb01,
-	0xf90465b6,
-	0x04659450,
-	0xbd0256bb,
-	0x0475fd50,
-	0x21f550fc,
-	0x64b607a5,
-	0x8a11f504,
-	0x0076bb00,
-	0xf90465b6,
-	0x04659450,
-	0xbd0256bb,
-	0x0475fd50,
-	0x21f550fc,
-	0x64b606f8,
-	0x6a11f404,
-	0xbbe05bcb,
-	0x65b60076,
-	0x9450f904,
-	0x56bb0465,
-	0xfd50bd02,
-	0x50fc0475,
-	0x063d21f5,
-	0xb90464b6,
-	0x74bd025b,
-/* 0x091f: i2c_recv_not_rd08 */
-	0xb0430ef4,
-	0x1bf401d6,
-	0x0057f03d,
-	0x07a521f5,
-	0xc73311f4,
-	0x21f5e0c5,
-	0x11f4074a,
-	0x0057f029,
-	0x07a521f5,
-	0xc71f11f4,
-	0x21f5e0b5,
-	0x11f4074a,
-	0x3d21f515,
-	0xc774bd06,
-	0x1bf408c5,
-	0x0232f409,
-/* 0x095f: i2c_recv_not_wr08 */
-/* 0x095f: i2c_recv_done */
-	0xc7030ef4,
-	0x21f5f8ce,
-	0xe0fc080a,
-	0x12f4d0fc,
-	0x027cb90a,
-	0x02b921f5,
-/* 0x0974: i2c_recv_exit */
-/* 0x0976: i2c_init */
+	0x76bb008a,
+	0x0465b600,
+	0x659450f9,
+	0x0256bb04,
+	0x75fd50bd,
+	0xf550fc04,
+	0xb6090121,
+	0x11f40464,
+	0xe05bcb6a,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0x4621f550,
+	0x0464b608,
+	0xbd025bb9,
+	0x430ef474,
+/* 0x0b28: i2c_recv_not_rd08 */
+	0xf401d6b0,
+	0x57f03d1b,
+	0xae21f500,
+	0x3311f409,
+	0xf5e0c5c7,
+	0xf4095321,
+	0x57f02911,
+	0xae21f500,
+	0x1f11f409,
+	0xf5e0b5c7,
+	0xf4095321,
+	0x21f51511,
+	0x74bd0846,
+	0xf408c5c7,
+	0x32f4091b,
+	0x030ef402,
+/* 0x0b68: i2c_recv_not_wr08 */
+/* 0x0b68: i2c_recv_done */
+	0xf5f8cec7,
+	0xfc0a1321,
+	0xf4d0fce0,
+	0x7cb90a12,
+	0x4221f502,
+/* 0x0b7d: i2c_recv_exit */
+/* 0x0b7f: i2c_init */
+	0xf800f803,
+/* 0x0b81: test_recv */
+	0xd817f100,
+	0x0614b605,
+	0xb60011cf,
+	0x07f10110,
+	0x04b605d8,
+	0x0001d006,
+	0xe7f104bd,
+	0xe3f1d900,
+	0x21f5134f,
+	0x00f80262,
+/* 0x0ba8: test_init */
+	0x0800e7f1,
+	0x026221f5,
+/* 0x0bb2: idle_recv */
 	0x00f800f8,
-/* 0x0978: test_recv */
-	0x05d817f1,
-	0xcf0614b6,
-	0x10b60011,
-	0xd807f101,
-	0x0604b605,
-	0xbd0001d0,
-	0x00e7f104,
-	0x4fe3f1d9,
-	0xf521f513,
-/* 0x099f: test_init */
-	0xf100f801,
-	0xf50800e7,
-	0xf801f521,
-/* 0x09a9: idle_recv */
-/* 0x09ab: idle */
-	0xf400f800,
-	0x17f10031,
-	0x14b605d4,
-	0x0011cf06,
-	0xf10110b6,
-	0xb605d407,
-	0x01d00604,
-/* 0x09c7: idle_loop */
-	0xf004bd00,
-	0x32f45817,
-/* 0x09cd: idle_proc */
-/* 0x09cd: idle_proc_exec */
-	0xb910f902,
-	0x21f5021e,
-	0x10fc02c2,
-	0xf40911f4,
-	0x0ef40231,
-/* 0x09e1: idle_proc_next */
-	0x5810b6ef,
-	0xf4061fb8,
-	0x02f4e61b,
-	0x0028f4dd,
-	0x00bb0ef4,
-	0x00000000,
-	0x00000000,
+/* 0x0bb4: idle */
+	0xf10031f4,
+	0xb605d417,
+	0x11cf0614,
+	0x0110b600,
+	0x05d407f1,
+	0xd00604b6,
+	0x04bd0001,
+/* 0x0bd0: idle_loop */
+	0xf45817f0,
+/* 0x0bd6: idle_proc */
+/* 0x0bd6: idle_proc_exec */
+	0x10f90232,
+	0xf5021eb9,
+	0xfc034b21,
+	0x0911f410,
+	0xf40231f4,
+/* 0x0bea: idle_proc_next */
+	0x10b6ef0e,
+	0x061fb858,
+	0xf4e61bf4,
+	0x28f4dd02,
+	0xbb0ef400,
 	0x00000000,
 };
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvd0.fuc b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvd0.fuc
index 8a89dfe41ce1..b85443261569 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvd0.fuc
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvd0.fuc
@@ -23,6 +23,7 @@
  */
 
 #define NVKM_PPWR_CHIPSET GF119
+#define HW_TICKS_PER_US 324
 
 //#define NVKM_FALCON_PC24
 #define NVKM_FALCON_UNSHIFTED_IO
@@ -34,6 +35,7 @@
 .section #nvd0_pwr_data
 #define INCLUDE_PROC
 #include "kernel.fuc"
+#include "arith.fuc"
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
@@ -44,6 +46,7 @@
 
 #define INCLUDE_DATA
 #include "kernel.fuc"
+#include "arith.fuc"
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
@@ -56,6 +59,7 @@
 .section #nvd0_pwr_code
 #define INCLUDE_CODE
 #include "kernel.fuc"
+#include "arith.fuc"
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvd0.fuc.h b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvd0.fuc.h
index 8d369b3faaba..12d86f72ad10 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvd0.fuc.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvd0.fuc.h
@@ -24,8 +24,8 @@ uint32_t nvd0_pwr_data[] = {
 	0x00000000,
 /* 0x0058: proc_list_head */
 	0x54534f48,
-	0x000003be,
-	0x00000367,
+	0x0000049d,
+	0x00000446,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -46,8 +46,8 @@ uint32_t nvd0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x584d454d,
-	0x000004b8,
-	0x000004aa,
+	0x00000678,
+	0x0000066a,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -68,8 +68,8 @@ uint32_t nvd0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x46524550,
-	0x000004bc,
-	0x000004ba,
+	0x0000067c,
+	0x0000067a,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -90,8 +90,8 @@ uint32_t nvd0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x5f433249,
-	0x000008d7,
-	0x0000077a,
+	0x00000a97,
+	0x0000093a,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -112,8 +112,8 @@ uint32_t nvd0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x54534554,
-	0x000008fa,
-	0x000008d9,
+	0x00000aba,
+	0x00000a99,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -134,8 +134,8 @@ uint32_t nvd0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x454c4449,
-	0x00000906,
-	0x00000904,
+	0x00000ac6,
+	0x00000ac4,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -227,24 +227,31 @@ uint32_t nvd0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 /* 0x0370: memx_func_head */
-	0x00010000,
-	0x00000000,
-	0x000003f4,
-/* 0x037c: memx_func_next */
 	0x00000001,
 	0x00000000,
-	0x00000415,
+	0x000004d3,
+/* 0x037c: memx_func_next */
 	0x00000002,
+	0x00000000,
+	0x00000554,
+	0x00000003,
 	0x00000002,
-	0x00000430,
-	0x00040003,
+	0x000005d8,
+	0x00040004,
+	0x00000000,
+	0x000005f4,
+	0x00010005,
+	0x00000000,
+	0x0000060e,
+	0x00010006,
+	0x00000000,
+	0x000005d3,
+/* 0x03b8: memx_func_tail */
+/* 0x03b8: memx_ts_start */
 	0x00000000,
-	0x0000044c,
-	0x00010004,
+/* 0x03bc: memx_ts_end */
 	0x00000000,
-	0x00000466,
-/* 0x03ac: memx_func_tail */
-/* 0x03ac: memx_data_head */
+/* 0x03c0: memx_data_head */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -757,8 +764,8 @@ uint32_t nvd0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-/* 0x0bac: memx_data_tail */
-/* 0x0bac: i2c_scl_map */
+/* 0x0bc0: memx_data_tail */
+/* 0x0bc0: i2c_scl_map */
 	0x00000400,
 	0x00000800,
 	0x00001000,
@@ -769,7 +776,7 @@ uint32_t nvd0_pwr_data[] = {
 	0x00020000,
 	0x00040000,
 	0x00080000,
-/* 0x0bd4: i2c_sda_map */
+/* 0x0be8: i2c_sda_map */
 	0x00100000,
 	0x00200000,
 	0x00400000,
@@ -781,10 +788,69 @@ uint32_t nvd0_pwr_data[] = {
 	0x10000000,
 	0x20000000,
 	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
 };
 
 uint32_t nvd0_pwr_code[] = {
-	0x02bf0ef5,
+	0x034d0ef5,
 /* 0x0004: rd32 */
 	0x07a007f1,
 	0xbd000ed0,
@@ -814,17 +880,20 @@ uint32_t nvd0_pwr_code[] = {
 	0xd4f100dd,
 	0x1bf47000,
 /* 0x0067: nsec */
-	0xf000f8f5,
+	0xf900f8f5,
+	0xf080f990,
 	0x88cf2c87,
-/* 0x006d: nsec_loop */
+/* 0x0071: nsec_loop */
 	0x2c97f000,
 	0xbb0099cf,
 	0x9eb80298,
 	0xf41ef406,
-/* 0x007e: wait */
-	0x87f000f8,
+	0x90fc80fc,
+/* 0x0086: wait */
+	0x90f900f8,
+	0x87f080f9,
 	0x0088cf2c,
-/* 0x0084: wait_loop */
+/* 0x0090: wait_loop */
 	0xf402eeb9,
 	0xdab90421,
 	0x04adfd02,
@@ -833,28 +902,29 @@ uint32_t nvd0_pwr_code[] = {
 	0x0099cf2c,
 	0xb80298bb,
 	0x1ef4069b,
-/* 0x00a5: wait_done */
-/* 0x00a7: intr_watchdog */
-	0x9800f8e2,
+/* 0x00b1: wait_done */
+	0xfc80fce2,
+/* 0x00b7: intr_watchdog */
+	0x9800f890,
 	0x96b003e9,
 	0x2a0bf400,
 	0xbb9a0a98,
 	0x1cf4029a,
 	0x01d7f00f,
-	0x020621f5,
+	0x028c21f5,
 	0x0ef494bd,
-/* 0x00c5: intr_watchdog_next_time */
+/* 0x00d5: intr_watchdog_next_time */
 	0x9b0a9815,
 	0xf400a6b0,
 	0x9ab8090b,
 	0x061cf406,
-/* 0x00d4: intr_watchdog_next_time_set */
-/* 0x00d7: intr_watchdog_next_proc */
+/* 0x00e4: intr_watchdog_next_time_set */
+/* 0x00e7: intr_watchdog_next_proc */
 	0x809b0980,
 	0xe0b603e9,
 	0x68e6b158,
 	0xc61bf402,
-/* 0x00e6: intr */
+/* 0x00f6: intr */
 	0x00f900f8,
 	0x80f904bd,
 	0xa0f990f9,
@@ -872,12 +942,12 @@ uint32_t nvd0_pwr_code[] = {
 	0x0bf40289,
 	0x9b008020,
 	0xf458e7f0,
-	0x0998a721,
+	0x0998b721,
 	0x0096b09b,
 	0xf00e0bf4,
 	0x09d03407,
 	0x8004bd00,
-/* 0x013e: intr_skip_watchdog */
+/* 0x014e: intr_skip_watchdog */
 	0x89e49a09,
 	0x0bf40800,
 	0x8897f13c,
@@ -889,20 +959,20 @@ uint32_t nvd0_pwr_code[] = {
 	0xf14f48e7,
 	0xf05453e3,
 	0x21f500d7,
-	0xc0fc026b,
+	0xc0fc02f1,
 	0x04c007f1,
 	0xbd000cd0,
-/* 0x0175: intr_subintr_skip_fifo */
+/* 0x0185: intr_subintr_skip_fifo */
 	0x8807f104,
 	0x0009d006,
-/* 0x017e: intr_skip_subintr */
+/* 0x018e: intr_skip_subintr */
 	0x89c404bd,
 	0x070bf420,
 	0xffbfa4f1,
-/* 0x0188: intr_skip_pause */
+/* 0x0198: intr_skip_pause */
 	0xf44089c4,
 	0xa4f1070b,
-/* 0x0192: intr_skip_user0 */
+/* 0x01a2: intr_skip_user0 */
 	0x07f0ffbf,
 	0x0008d004,
 	0x80fc04bd,
@@ -912,189 +982,298 @@ uint32_t nvd0_pwr_code[] = {
 	0xfca0fcb0,
 	0xfc80fc90,
 	0x0032f400,
-/* 0x01b6: timer */
-	0x32f401f8,
-	0x03f89810,
-	0xf40086b0,
-	0xfe80421c,
-	0x3807f003,
+/* 0x01c6: ticks_from_ns */
+	0xc0f901f8,
+	0xd7f1b0f9,
+	0xd3f00144,
+	0xb321f500,
+	0xe8ccec03,
+	0x00b4b003,
+	0xec120bf4,
+	0xf103e8ee,
+	0xf00144d7,
+	0x21f500d3,
+/* 0x01ee: ticks_from_ns_quit */
+	0xceb903b3,
+	0xfcb0fc02,
+/* 0x01f7: ticks_from_us */
+	0xf900f8c0,
+	0xf1b0f9c0,
+	0xf00144d7,
+	0x21f500d3,
+	0xceb903b3,
+	0x00b4b002,
+	0xbd050bf4,
+/* 0x0211: ticks_from_us_quit */
+	0xfcb0fce4,
+/* 0x0217: ticks_to_us */
+	0xf100f8c0,
+	0xf00144d7,
+	0xedff00d3,
+/* 0x0223: timer */
+	0xf900f8ec,
+	0xf480f990,
+	0xf8981032,
+	0x0086b003,
+	0xbd531cf4,
+	0x3807f084,
 	0xbd0008d0,
-	0x0887f004,
-	0xf00088cf,
-	0x1bf40284,
-	0x3487f020,
-	0xb80088cf,
-	0x0bf406e0,
-	0x06e8b809,
-/* 0x01eb: timer_reset */
-	0xf0191ef4,
-	0x0ed03407,
-	0x8004bd00,
-/* 0x01f6: timer_enable */
-	0x87f09a0e,
-	0x3807f001,
-	0xbd0008d0,
-/* 0x0201: timer_done */
-	0x1031f404,
-/* 0x0206: send_proc */
-	0x80f900f8,
-	0xe89890f9,
+	0x3487f004,
+	0x980088cf,
+	0x98bb9a09,
+	0x00e9bb02,
+	0xf003fe80,
+	0x88cf0887,
+	0x0284f000,
+	0xf0201bf4,
+	0x88cf3487,
+	0x06e0b800,
+	0xb8090bf4,
+	0x1cf406e8,
+/* 0x026d: timer_reset */
+	0x3407f00e,
+	0xbd000ed0,
+	0x9a0e8004,
+/* 0x0278: timer_enable */
+	0xf00187f0,
+	0x08d03807,
+/* 0x0283: timer_done */
+	0xf404bd00,
+	0x80fc1031,
+	0x00f890fc,
+/* 0x028c: send_proc */
+	0x90f980f9,
+	0x9805e898,
+	0x86f004e9,
+	0x0689b804,
+	0xc42a0bf4,
+	0x88940398,
+	0x1880b604,
+	0x98008ebb,
+	0x8a8000fa,
+	0x018d8000,
+	0x80028c80,
+	0x90b6038b,
+	0x0794f001,
+	0xf404e980,
+/* 0x02c6: send_done */
+	0x90fc0231,
+	0x00f880fc,
+/* 0x02cc: find */
+	0x87f080f9,
+	0x0131f458,
+/* 0x02d4: find_loop */
+	0xb8008a98,
+	0x0bf406ae,
+	0x5880b610,
+	0x026886b1,
+	0xf4f01bf4,
+/* 0x02ea: find_done */
+	0x8eb90132,
+	0xf880fc02,
+/* 0x02f1: send */
+	0xcc21f500,
+	0x9701f402,
+/* 0x02fa: recv */
+	0x90f900f8,
+	0xe89880f9,
 	0x04e99805,
-	0xb80486f0,
+	0xb80132f4,
 	0x0bf40689,
-	0x0398c42a,
-	0xb6048894,
-	0x8ebb1880,
-	0x00fa9800,
-	0x80008a80,
-	0x8c80018d,
-	0x038b8002,
-	0xf00190b6,
-	0xe9800794,
-	0x0231f404,
-/* 0x0240: send_done */
-	0x80fc90fc,
-/* 0x0246: find */
-	0x80f900f8,
-	0xf45887f0,
-/* 0x024e: find_loop */
-	0x8a980131,
-	0x06aeb800,
-	0xb6100bf4,
-	0x86b15880,
-	0x1bf40268,
-	0x0132f4f0,
-/* 0x0264: find_done */
-	0xfc028eb9,
-/* 0x026b: send */
-	0xf500f880,
-	0xf4024621,
-	0x00f89701,
-/* 0x0274: recv */
-	0x9805e898,
-	0x32f404e9,
-	0x0689b801,
-	0xc43d0bf4,
-	0x80b60389,
-	0x0784f001,
-	0x9805e880,
-	0xf0f902ea,
-	0xf9018ffe,
-	0x02efb9f0,
-	0xbb049994,
-	0xe0b600e9,
-	0x03eb9818,
-	0x9802ec98,
-	0xee9801ed,
-	0xfca5f900,
-	0x00f8fef0,
-	0xfc0131f4,
-/* 0x02bd: recv_done */
-/* 0x02bf: init */
-	0xf100f8f0,
-	0xcf010817,
-	0x11e70011,
-	0x14b60109,
-	0x0014fe08,
-	0x00e017f1,
-	0xf00013f0,
-	0x01d01c07,
-	0xf004bd00,
-	0x07f0ff17,
-	0x0001d014,
-	0x17f004bd,
-	0x0015f102,
-	0x1007f008,
-	0xbd0001d0,
-	0xe617f104,
-	0x0013f000,
-	0xf40010fe,
-	0x17f01031,
-	0x3807f001,
-	0xbd0001d0,
-	0x58f7f004,
-/* 0x0314: init_proc */
-	0xb001f198,
-	0x0bf40016,
-	0xb615f9fa,
-	0x0ef458f0,
-/* 0x0325: host_send */
-	0xb017f1f2,
-	0x0011cf04,
-	0x04a027f1,
-	0xb80022cf,
-	0x0bf40612,
-	0x071ec42f,
-	0xb704ee94,
-	0x980270e0,
+	0x0389c43d,
+	0xf00180b6,
+	0xe8800784,
+	0x02ea9805,
+	0x8ffef0f9,
+	0xb9f0f901,
+	0x999402ef,
+	0x00e9bb04,
+	0x9818e0b6,
 	0xec9803eb,
 	0x01ed9802,
-	0xf500ee98,
-	0xb6026b21,
-	0x1ec40110,
-	0xb007f10f,
-	0x000ed004,
-	0x0ef404bd,
-/* 0x0365: host_send_done */
-/* 0x0367: host_recv */
-	0xf100f8c3,
-	0xf14e4917,
-	0xb8525413,
-	0x0bf406e1,
-/* 0x0375: host_recv_wait */
-	0xcc17f1b3,
-	0x0011cf04,
-	0x04c827f1,
-	0xf00022cf,
-	0x12b80816,
-	0xec0bf406,
-	0xb60723c4,
-	0x30b70434,
-	0x3b8002f0,
-	0x023c8003,
-	0x80013d80,
-	0x20b6003e,
-	0x0f24f001,
-	0x04c807f1,
-	0xbd0002d0,
-	0x4027f004,
-	0xd00007f0,
-	0x04bd0002,
-/* 0x03be: host_init */
+	0xf900ee98,
+	0xfef0fca5,
+	0x31f400f8,
+/* 0x0347: recv_done */
+	0xfcf0fc01,
+	0xf890fc80,
+/* 0x034d: init */
+	0x0817f100,
+	0x0011cf01,
+	0x010911e7,
+	0xfe0814b6,
+	0x17f10014,
+	0x13f000e0,
+	0x1c07f000,
+	0xbd0001d0,
+	0xff17f004,
+	0xd01407f0,
+	0x04bd0001,
+	0xf10217f0,
+	0xf0080015,
+	0x01d01007,
+	0xf104bd00,
+	0xf000f617,
+	0x10fe0013,
+	0x1031f400,
+	0xf00117f0,
+	0x01d03807,
+	0xf004bd00,
+/* 0x03a2: init_proc */
+	0xf19858f7,
+	0x0016b001,
+	0xf9fa0bf4,
+	0x58f0b615,
+/* 0x03b3: mulu32_32_64 */
+	0xf9f20ef4,
+	0xf920f910,
+	0x9540f930,
+	0xd29510e1,
+	0xbdc4bd10,
+	0xc0edffb4,
+	0xb9301dff,
+	0x34f10234,
+	0x34b6ffff,
+	0x1045b610,
+	0xbb00c3bb,
+	0xe2ff01b4,
+	0x0234b930,
+	0xffff34f1,
+	0xb61034b6,
+	0xc3bb1045,
+	0x01b4bb00,
+	0xbb3012ff,
+	0x40fc00b3,
+	0x20fc30fc,
+	0x00f810fc,
+/* 0x0404: host_send */
+	0x04b017f1,
+	0xf10011cf,
+	0xcf04a027,
+	0x12b80022,
+	0x2f0bf406,
+	0x94071ec4,
+	0xe0b704ee,
+	0xeb980270,
+	0x02ec9803,
+	0x9801ed98,
+	0x21f500ee,
+	0x10b602f1,
+	0x0f1ec401,
+	0x04b007f1,
+	0xbd000ed0,
+	0xc30ef404,
+/* 0x0444: host_send_done */
+/* 0x0446: host_recv */
 	0x17f100f8,
-	0x14b60080,
-	0x7015f110,
-	0xd007f102,
-	0x0001d004,
-	0x17f104bd,
-	0x14b60080,
-	0xf015f110,
-	0xdc07f102,
-	0x0001d004,
-	0x17f004bd,
-	0xc407f101,
-	0x0001d004,
-	0x00f804bd,
-/* 0x03f4: memx_func_enter */
+	0x13f14e49,
+	0xe1b85254,
+	0xb30bf406,
+/* 0x0454: host_recv_wait */
+	0x04cc17f1,
+	0xf10011cf,
+	0xcf04c827,
+	0x16f00022,
+	0x0612b808,
+	0xc4ec0bf4,
+	0x34b60723,
+	0xf030b704,
+	0x033b8002,
+	0x80023c80,
+	0x3e80013d,
+	0x0120b600,
+	0xf10f24f0,
+	0xd004c807,
+	0x04bd0002,
+	0xf04027f0,
+	0x02d00007,
+	0xf804bd00,
+/* 0x049d: host_init */
+	0x8017f100,
+	0x1014b600,
+	0x027015f1,
+	0x04d007f1,
+	0xbd0001d0,
+	0x8017f104,
+	0x1014b600,
+	0x02f015f1,
+	0x04dc07f1,
+	0xbd0001d0,
+	0x0117f004,
+	0x04c407f1,
+	0xbd0001d0,
+/* 0x04d3: memx_func_enter */
+	0xf100f804,
+	0xf1162067,
+	0xf1f55d77,
+	0xb9ffff73,
+	0x21f4026e,
+	0x02d8b904,
+	0xf90487fd,
+	0xfc80f960,
+	0xf4e0fcd0,
+	0x77f13321,
+	0x73f1fffe,
+	0x6eb9ffff,
+	0x0421f402,
+	0xfd02d8b9,
+	0x60f90487,
+	0xd0fc80f9,
+	0x21f4e0fc,
+	0xf067f133,
+	0x026eb926,
+	0xb90421f4,
+	0x87fd02d8,
+	0xf960f904,
+	0xfcd0fc80,
+	0x3321f4e0,
 	0xf10467f0,
 	0xd007e007,
 	0x04bd0006,
-/* 0x0400: memx_func_enter_wait */
+/* 0x053c: memx_func_enter_wait */
 	0x07c067f1,
 	0xf00066cf,
 	0x0bf40464,
-	0x001698f6,
-	0xf80410b6,
-/* 0x0415: memx_func_leave */
-	0x0467f000,
+	0x2c67f0f6,
+	0x800066cf,
+	0x00f8ee06,
+/* 0x0554: memx_func_leave */
+	0xcf2c67f0,
+	0x06800066,
+	0x0467f0ef,
 	0x07e407f1,
 	0xbd0006d0,
-/* 0x0421: memx_func_leave_wait */
+/* 0x0569: memx_func_leave_wait */
 	0xc067f104,
 	0x0066cf07,
 	0xf40464f0,
-	0x00f8f61b,
-/* 0x0430: memx_func_wr32 */
+	0x67f1f61b,
+	0x77f126f0,
+	0x73f00001,
+	0x026eb900,
+	0xb90421f4,
+	0x87fd02d8,
+	0xf960f905,
+	0xfcd0fc80,
+	0x3321f4e0,
+	0x162067f1,
+	0xf4026eb9,
+	0xd8b90421,
+	0x0587fd02,
+	0x80f960f9,
+	0xe0fcd0fc,
+	0xf13321f4,
+	0xf00aa277,
+	0x6eb90073,
+	0x0421f402,
+	0xfd02d8b9,
+	0x60f90587,
+	0xd0fc80f9,
+	0x21f4e0fc,
+/* 0x05d3: memx_func_wait_vblank */
+	0xb600f833,
+	0x00f80410,
+/* 0x05d8: memx_func_wr32 */
 	0x98001698,
 	0x10b60115,
 	0xf960f908,
@@ -1102,131 +1281,137 @@ uint32_t nvd0_pwr_code[] = {
 	0x3321f4e0,
 	0xf40242b6,
 	0x00f8e91b,
-/* 0x044c: memx_func_wait */
+/* 0x05f4: memx_func_wait */
 	0xcf2c87f0,
 	0x1e980088,
 	0x011d9800,
 	0x98021c98,
 	0x10b6031b,
-	0x7e21f410,
-/* 0x0466: memx_func_delay */
+	0x8621f410,
+/* 0x060e: memx_func_delay */
 	0x1e9800f8,
 	0x0410b600,
 	0xf86721f4,
-/* 0x0471: memx_exec */
+/* 0x0619: memx_exec */
 	0xf9e0f900,
 	0x02c1b9d0,
-/* 0x047b: memx_exec_next */
+/* 0x0623: memx_exec_next */
 	0x9802b2b9,
 	0x10b60013,
-	0x10349504,
+	0xf034e704,
+	0xe033e701,
+	0x0132b601,
 	0x980c30f0,
 	0x55f9de35,
 	0xf40612b8,
-	0xd0fcec1e,
+	0x0b98e41e,
+	0xef0c98ee,
+	0xf102cbbb,
+	0xcf07c4b7,
+	0xd0fc00bb,
 	0x21f5e0fc,
-	0x00f8026b,
-/* 0x049c: memx_info */
-	0x03acc7f1,
+	0x00f802f1,
+/* 0x065c: memx_info */
+	0x03c0c7f1,
 	0x0800b7f1,
-	0x026b21f5,
-/* 0x04aa: memx_recv */
+	0x02f121f5,
+/* 0x066a: memx_recv */
 	0xd6b000f8,
-	0xc40bf401,
+	0xac0bf401,
 	0xf400d6b0,
 	0x00f8e90b,
-/* 0x04b8: memx_init */
-/* 0x04ba: perf_recv */
+/* 0x0678: memx_init */
+/* 0x067a: perf_recv */
 	0x00f800f8,
-/* 0x04bc: perf_init */
-/* 0x04be: i2c_drive_scl */
+/* 0x067c: perf_init */
+/* 0x067e: i2c_drive_scl */
 	0x36b000f8,
 	0x0e0bf400,
 	0x07e007f1,
 	0xbd0001d0,
-/* 0x04cf: i2c_drive_scl_lo */
+/* 0x068f: i2c_drive_scl_lo */
 	0xf100f804,
 	0xd007e407,
 	0x04bd0001,
-/* 0x04da: i2c_drive_sda */
+/* 0x069a: i2c_drive_sda */
 	0x36b000f8,
 	0x0e0bf400,
 	0x07e007f1,
 	0xbd0002d0,
-/* 0x04eb: i2c_drive_sda_lo */
+/* 0x06ab: i2c_drive_sda_lo */
 	0xf100f804,
 	0xd007e407,
 	0x04bd0002,
-/* 0x04f6: i2c_sense_scl */
+/* 0x06b6: i2c_sense_scl */
 	0x32f400f8,
 	0xc437f101,
 	0x0033cf07,
 	0xf40431fd,
 	0x31f4060b,
-/* 0x0509: i2c_sense_scl_done */
-/* 0x050b: i2c_sense_sda */
+/* 0x06c9: i2c_sense_scl_done */
+/* 0x06cb: i2c_sense_sda */
 	0xf400f801,
 	0x37f10132,
 	0x33cf07c4,
 	0x0432fd00,
 	0xf4060bf4,
-/* 0x051e: i2c_sense_sda_done */
+/* 0x06de: i2c_sense_sda_done */
 	0x00f80131,
-/* 0x0520: i2c_raise_scl */
+/* 0x06e0: i2c_raise_scl */
 	0x47f140f9,
 	0x37f00898,
-	0xbe21f501,
-/* 0x052d: i2c_raise_scl_wait */
-	0xe8e7f104,
+	0x7e21f501,
+/* 0x06ed: i2c_raise_scl_wait */
+	0xe8e7f106,
 	0x6721f403,
-	0x04f621f5,
+	0x06b621f5,
 	0xb60901f4,
 	0x1bf40142,
-/* 0x0541: i2c_raise_scl_done */
+/* 0x0701: i2c_raise_scl_done */
 	0xf840fcef,
-/* 0x0545: i2c_start */
-	0xf621f500,
-	0x0d11f404,
-	0x050b21f5,
+/* 0x0705: i2c_start */
+	0xb621f500,
+	0x0d11f406,
+	0x06cb21f5,
 	0xf40611f4,
-/* 0x0556: i2c_start_rep */
+/* 0x0716: i2c_start_rep */
 	0x37f0300e,
-	0xbe21f500,
-	0x0137f004,
-	0x04da21f5,
+	0x7e21f500,
+	0x0137f006,
+	0x069a21f5,
 	0xb60076bb,
 	0x50f90465,
 	0xbb046594,
 	0x50bd0256,
 	0xfc0475fd,
-	0x2021f550,
-	0x0464b605,
-/* 0x0583: i2c_start_send */
+	0xe021f550,
+	0x0464b606,
+/* 0x0743: i2c_start_send */
 	0xf01f11f4,
 	0x21f50037,
-	0xe7f104da,
+	0xe7f1069a,
 	0x21f41388,
 	0x0037f067,
-	0x04be21f5,
+	0x067e21f5,
 	0x1388e7f1,
-/* 0x059f: i2c_start_out */
+/* 0x075f: i2c_start_out */
 	0xf86721f4,
-/* 0x05a1: i2c_stop */
+/* 0x0761: i2c_stop */
 	0x0037f000,
-	0x04be21f5,
+	0x067e21f5,
 	0xf50037f0,
-	0xf104da21,
+	0xf1069a21,
 	0xf403e8e7,
 	0x37f06721,
-	0xbe21f501,
-	0x88e7f104,
+	0x7e21f501,
+	0x88e7f106,
 	0x6721f413,
 	0xf50137f0,
-	0xf104da21,
+	0xf1069a21,
 	0xf41388e7,
 	0x00f86721,
-/* 0x05d4: i2c_bitw */
-	0x04da21f5,
+/* 0x0794: i2c_bitw */
+	0x069a21f5,
 	0x03e8e7f1,
 	0xbb6721f4,
 	0x65b60076,
@@ -1234,18 +1419,18 @@ uint32_t nvd0_pwr_code[] = {
 	0x56bb0465,
 	0xfd50bd02,
 	0x50fc0475,
-	0x052021f5,
+	0x06e021f5,
 	0xf40464b6,
 	0xe7f11811,
 	0x21f41388,
 	0x0037f067,
-	0x04be21f5,
+	0x067e21f5,
 	0x1388e7f1,
-/* 0x0613: i2c_bitw_out */
+/* 0x07d3: i2c_bitw_out */
 	0xf86721f4,
-/* 0x0615: i2c_bitr */
+/* 0x07d5: i2c_bitr */
 	0x0137f000,
-	0x04da21f5,
+	0x069a21f5,
 	0x03e8e7f1,
 	0xbb6721f4,
 	0x65b60076,
@@ -1253,19 +1438,19 @@ uint32_t nvd0_pwr_code[] = {
 	0x56bb0465,
 	0xfd50bd02,
 	0x50fc0475,
-	0x052021f5,
+	0x06e021f5,
 	0xf40464b6,
 	0x21f51b11,
-	0x37f0050b,
-	0xbe21f500,
-	0x88e7f104,
+	0x37f006cb,
+	0x7e21f500,
+	0x88e7f106,
 	0x6721f413,
 	0xf4013cf0,
-/* 0x065a: i2c_bitr_done */
+/* 0x081a: i2c_bitr_done */
 	0x00f80131,
-/* 0x065c: i2c_get_byte */
+/* 0x081c: i2c_get_byte */
 	0xf00057f0,
-/* 0x0662: i2c_get_byte_next */
+/* 0x0822: i2c_get_byte_next */
 	0x54b60847,
 	0x0076bb01,
 	0xf90465b6,
@@ -1273,7 +1458,7 @@ uint32_t nvd0_pwr_code[] = {
 	0xbd0256bb,
 	0x0475fd50,
 	0x21f550fc,
-	0x64b60615,
+	0x64b607d5,
 	0x2b11f404,
 	0xb60553fd,
 	0x1bf40142,
@@ -1283,12 +1468,12 @@ uint32_t nvd0_pwr_code[] = {
 	0xbb046594,
 	0x50bd0256,
 	0xfc0475fd,
-	0xd421f550,
-	0x0464b605,
-/* 0x06ac: i2c_get_byte_done */
-/* 0x06ae: i2c_put_byte */
+	0x9421f550,
+	0x0464b607,
+/* 0x086c: i2c_get_byte_done */
+/* 0x086e: i2c_put_byte */
 	0x47f000f8,
-/* 0x06b1: i2c_put_byte_next */
+/* 0x0871: i2c_put_byte_next */
 	0x0142b608,
 	0xbb3854ff,
 	0x65b60076,
@@ -1296,7 +1481,7 @@ uint32_t nvd0_pwr_code[] = {
 	0x56bb0465,
 	0xfd50bd02,
 	0x50fc0475,
-	0x05d421f5,
+	0x079421f5,
 	0xf40464b6,
 	0x46b03411,
 	0xd81bf400,
@@ -1305,21 +1490,21 @@ uint32_t nvd0_pwr_code[] = {
 	0xbb046594,
 	0x50bd0256,
 	0xfc0475fd,
-	0x1521f550,
-	0x0464b606,
+	0xd521f550,
+	0x0464b607,
 	0xbb0f11f4,
 	0x36b00076,
 	0x061bf401,
-/* 0x0707: i2c_put_byte_done */
+/* 0x08c7: i2c_put_byte_done */
 	0xf80132f4,
-/* 0x0709: i2c_addr */
+/* 0x08c9: i2c_addr */
 	0x0076bb00,
 	0xf90465b6,
 	0x04659450,
 	0xbd0256bb,
 	0x0475fd50,
 	0x21f550fc,
-	0x64b60545,
+	0x64b60705,
 	0x2911f404,
 	0x012ec3e7,
 	0xfd0134b6,
@@ -1329,30 +1514,30 @@ uint32_t nvd0_pwr_code[] = {
 	0x0256bb04,
 	0x75fd50bd,
 	0xf550fc04,
-	0xb606ae21,
-/* 0x074e: i2c_addr_done */
+	0xb6086e21,
+/* 0x090e: i2c_addr_done */
 	0x00f80464,
-/* 0x0750: i2c_acquire_addr */
+/* 0x0910: i2c_acquire_addr */
 	0xb6f8cec7,
 	0xe0b705e4,
 	0x00f8d014,
-/* 0x075c: i2c_acquire */
-	0x075021f5,
+/* 0x091c: i2c_acquire */
+	0x091021f5,
 	0xf00421f4,
 	0x21f403d9,
-/* 0x076b: i2c_release */
+/* 0x092b: i2c_release */
 	0xf500f833,
-	0xf4075021,
+	0xf4091021,
 	0xdaf00421,
 	0x3321f403,
-/* 0x077a: i2c_recv */
+/* 0x093a: i2c_recv */
 	0x32f400f8,
 	0xf8c1c701,
 	0xb00214b6,
 	0x1ff52816,
 	0x13a0013a,
-	0x32980bd4,
-	0xac13a000,
+	0x32980be8,
+	0xc013a000,
 	0x0031980b,
 	0xf90231f4,
 	0xf9e0f9d0,
@@ -1364,8 +1549,8 @@ uint32_t nvd0_pwr_code[] = {
 	0xbb046594,
 	0x50bd0256,
 	0xfc0475fd,
-	0x5c21f550,
-	0x0464b607,
+	0x1c21f550,
+	0x0464b609,
 	0xd6b0d0fc,
 	0xb31bf500,
 	0x0057f000,
@@ -1374,8 +1559,8 @@ uint32_t nvd0_pwr_code[] = {
 	0xbb046594,
 	0x50bd0256,
 	0xfc0475fd,
-	0x0921f550,
-	0x0464b607,
+	0xc921f550,
+	0x0464b608,
 	0x00d011f5,
 	0xbbe0c5c7,
 	0x65b60076,
@@ -1383,7 +1568,7 @@ uint32_t nvd0_pwr_code[] = {
 	0x56bb0465,
 	0xfd50bd02,
 	0x50fc0475,
-	0x06ae21f5,
+	0x086e21f5,
 	0xf50464b6,
 	0xf000ad11,
 	0x76bb0157,
@@ -1392,7 +1577,7 @@ uint32_t nvd0_pwr_code[] = {
 	0x0256bb04,
 	0x75fd50bd,
 	0xf550fc04,
-	0xb6070921,
+	0xb608c921,
 	0x11f50464,
 	0x76bb008a,
 	0x0465b600,
@@ -1400,7 +1585,7 @@ uint32_t nvd0_pwr_code[] = {
 	0x0256bb04,
 	0x75fd50bd,
 	0xf550fc04,
-	0xb6065c21,
+	0xb6081c21,
 	0x11f40464,
 	0xe05bcb6a,
 	0xb60076bb,
@@ -1408,38 +1593,38 @@ uint32_t nvd0_pwr_code[] = {
 	0xbb046594,
 	0x50bd0256,
 	0xfc0475fd,
-	0xa121f550,
-	0x0464b605,
+	0x6121f550,
+	0x0464b607,
 	0xbd025bb9,
 	0x430ef474,
-/* 0x0880: i2c_recv_not_rd08 */
+/* 0x0a40: i2c_recv_not_rd08 */
 	0xf401d6b0,
 	0x57f03d1b,
-	0x0921f500,
-	0x3311f407,
+	0xc921f500,
+	0x3311f408,
 	0xf5e0c5c7,
-	0xf406ae21,
+	0xf4086e21,
 	0x57f02911,
-	0x0921f500,
-	0x1f11f407,
+	0xc921f500,
+	0x1f11f408,
 	0xf5e0b5c7,
-	0xf406ae21,
+	0xf4086e21,
 	0x21f51511,
-	0x74bd05a1,
+	0x74bd0761,
 	0xf408c5c7,
 	0x32f4091b,
 	0x030ef402,
-/* 0x08c0: i2c_recv_not_wr08 */
-/* 0x08c0: i2c_recv_done */
+/* 0x0a80: i2c_recv_not_wr08 */
+/* 0x0a80: i2c_recv_done */
 	0xf5f8cec7,
-	0xfc076b21,
+	0xfc092b21,
 	0xf4d0fce0,
 	0x7cb90a12,
-	0x6b21f502,
-/* 0x08d5: i2c_recv_exit */
-/* 0x08d7: i2c_init */
+	0xf121f502,
+/* 0x0a95: i2c_recv_exit */
+/* 0x0a97: i2c_init */
 	0xf800f802,
-/* 0x08d9: test_recv */
+/* 0x0a99: test_recv */
 	0xd817f100,
 	0x0011cf05,
 	0xf10110b6,
@@ -1447,29 +1632,29 @@ uint32_t nvd0_pwr_code[] = {
 	0x04bd0001,
 	0xd900e7f1,
 	0x134fe3f1,
-	0x01b621f5,
-/* 0x08fa: test_init */
+	0x022321f5,
+/* 0x0aba: test_init */
 	0xe7f100f8,
 	0x21f50800,
-	0x00f801b6,
-/* 0x0904: idle_recv */
-/* 0x0906: idle */
+	0x00f80223,
+/* 0x0ac4: idle_recv */
+/* 0x0ac6: idle */
 	0x31f400f8,
 	0xd417f100,
 	0x0011cf05,
 	0xf10110b6,
 	0xd005d407,
 	0x04bd0001,
-/* 0x091c: idle_loop */
+/* 0x0adc: idle_loop */
 	0xf45817f0,
-/* 0x0922: idle_proc */
-/* 0x0922: idle_proc_exec */
+/* 0x0ae2: idle_proc */
+/* 0x0ae2: idle_proc_exec */
 	0x10f90232,
 	0xf5021eb9,
-	0xfc027421,
+	0xfc02fa21,
 	0x0911f410,
 	0xf40231f4,
-/* 0x0936: idle_proc_next */
+/* 0x0af6: idle_proc_next */
 	0x10b6ef0e,
 	0x061fb858,
 	0xf4e61bf4,
@@ -1521,4 +1706,20 @@ uint32_t nvd0_pwr_code[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
 };
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/os.h b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/os.h
index 574acfa44c8c..522e3079f824 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/os.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/os.h
@@ -19,11 +19,12 @@
 #define MEMX_MSG_EXEC 1
 
 /* MEMX: script opcode definitions */
-#define MEMX_ENTER  0
-#define MEMX_LEAVE  1
-#define MEMX_WR32   2
-#define MEMX_WAIT   3
-#define MEMX_DELAY  4
+#define MEMX_ENTER  1
+#define MEMX_LEAVE  2
+#define MEMX_WR32   3
+#define MEMX_WAIT   4
+#define MEMX_DELAY  5
+#define MEMX_VBLANK 6
 
 /* I2C_: message identifiers */
 #define I2C__MSG_RD08 0
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/memx.c b/drivers/gpu/drm/nouveau/core/subdev/pwr/memx.c
index def6a9ac68cf..65eaa2546cad 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/memx.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/memx.c
@@ -20,10 +20,11 @@ memx_out(struct nouveau_memx *memx)
 	struct nouveau_pwr *ppwr = memx->ppwr;
 	int i;
 
-	if (memx->c.size) {
+	if (memx->c.mthd) {
 		nv_wr32(ppwr, 0x10a1c4, (memx->c.size << 16) | memx->c.mthd);
 		for (i = 0; i < memx->c.size; i++)
 			nv_wr32(ppwr, 0x10a1c4, memx->c.data[i]);
+		memx->c.mthd = 0;
 		memx->c.size = 0;
 	}
 }
@@ -32,7 +33,7 @@ static void
 memx_cmd(struct nouveau_memx *memx, u32 mthd, u32 size, u32 data[])
 {
 	if ((memx->c.size + size >= ARRAY_SIZE(memx->c.data)) ||
-	    (memx->c.size && memx->c.mthd != mthd))
+	    (memx->c.mthd && memx->c.mthd != mthd))
 		memx_out(memx);
 	memcpy(&memx->c.data[memx->c.size], data, size * sizeof(data[0]));
 	memx->c.size += size;
@@ -62,8 +63,7 @@ nouveau_memx_init(struct nouveau_pwr *ppwr, struct nouveau_memx **pmemx)
 		nv_wr32(ppwr, 0x10a580, 0x00000003);
 	} while (nv_rd32(ppwr, 0x10a580) != 0x00000003);
 	nv_wr32(ppwr, 0x10a1c0, 0x01000000 | memx->base);
-	nv_wr32(ppwr, 0x10a1c4, 0x00010000 | MEMX_ENTER);
-	nv_wr32(ppwr, 0x10a1c4, 0x00000000);
+
 	return 0;
 }
 
@@ -78,7 +78,6 @@ nouveau_memx_fini(struct nouveau_memx **pmemx, bool exec)
 	memx_out(memx);
 
 	/* release data segment access */
-	nv_wr32(ppwr, 0x10a1c4, 0x00000000 | MEMX_LEAVE);
 	finish = nv_rd32(ppwr, 0x10a1c0) & 0x00ffffff;
 	nv_wr32(ppwr, 0x10a580, 0x00000000);
 
@@ -88,6 +87,8 @@ nouveau_memx_fini(struct nouveau_memx **pmemx, bool exec)
 				 memx->base, finish);
 	}
 
+	nv_debug(memx->ppwr, "Exec took %uns, PPWR_IN %08x\n",
+		 reply[0], reply[1]);
 	kfree(memx);
 	return 0;
 }
@@ -117,4 +118,51 @@ nouveau_memx_nsec(struct nouveau_memx *memx, u32 nsec)
 	memx_out(memx); /* fuc can't handle multiple */
 }
 
+void
+nouveau_memx_wait_vblank(struct nouveau_memx *memx)
+{
+	struct nouveau_pwr *ppwr = memx->ppwr;
+	u32 heads, x, y, px = 0;
+	int i, head_sync;
+
+	if (nv_device(ppwr)->chipset < 0xd0) {
+		heads = nv_rd32(ppwr, 0x610050);
+		for (i = 0; i < 2; i++) {
+			/* Heuristic: sync to head with biggest resolution */
+			if (heads & (2 << (i << 3))) {
+				x = nv_rd32(ppwr, 0x610b40 + (0x540 * i));
+				y = (x & 0xffff0000) >> 16;
+				x &= 0x0000ffff;
+				if ((x * y) > px) {
+					px = (x * y);
+					head_sync = i;
+				}
+			}
+		}
+	}
+
+	if (px == 0) {
+		nv_debug(memx->ppwr, "WAIT VBLANK !NO ACTIVE HEAD\n");
+		return;
+	}
+
+	nv_debug(memx->ppwr, "WAIT VBLANK HEAD%d\n", head_sync);
+	memx_cmd(memx, MEMX_VBLANK, 1, (u32[]){ head_sync });
+	memx_out(memx); /* fuc can't handle multiple */
+}
+
+void
+nouveau_memx_block(struct nouveau_memx *memx)
+{
+	nv_debug(memx->ppwr, "   HOST BLOCKED\n");
+	memx_cmd(memx, MEMX_ENTER, 0, NULL);
+}
+
+void
+nouveau_memx_unblock(struct nouveau_memx *memx)
+{
+	nv_debug(memx->ppwr, "   HOST UNBLOCKED\n");
+	memx_cmd(memx, MEMX_LEAVE, 0, NULL);
+}
+
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/therm/fan.c b/drivers/gpu/drm/nouveau/core/subdev/therm/fan.c
index 016990a8252c..3656d605168f 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/therm/fan.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/therm/fan.c
@@ -31,6 +31,8 @@
 #include <subdev/gpio.h>
 #include <subdev/timer.h>
 
+#include <subdev/bios/fan.h>
+
 static int
 nouveau_fan_update(struct nouveau_fan *fan, bool immediate, int target)
 {
@@ -275,8 +277,11 @@ nouveau_therm_fan_ctor(struct nouveau_therm *therm)
 	/* other random init... */
 	nouveau_therm_fan_set_defaults(therm);
 	nvbios_perf_fan_parse(bios, &priv->fan->perf);
-	if (nvbios_therm_fan_parse(bios, &priv->fan->bios))
-		nv_error(therm, "parsing the thermal table failed\n");
+	if (!nvbios_fan_parse(bios, &priv->fan->bios)) {
+		nv_debug(therm, "parsing the fan table failed\n");
+		if (nvbios_therm_fan_parse(bios, &priv->fan->bios))
+			nv_error(therm, "parsing both fan tables failed\n");
+	}
 	nouveau_therm_fan_safety_checks(therm);
 	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/core/subdev/therm/fanpwm.c b/drivers/gpu/drm/nouveau/core/subdev/therm/fanpwm.c
index 9a5c07340263..c629d7f2a6a4 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/therm/fanpwm.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/therm/fanpwm.c
@@ -25,6 +25,8 @@
 
 #include <core/option.h>
 #include <subdev/gpio.h>
+#include <subdev/bios.h>
+#include <subdev/bios/fan.h>
 
 #include "priv.h"
 
@@ -86,11 +88,15 @@ nouveau_fanpwm_create(struct nouveau_therm *therm, struct dcb_gpio_func *func)
 {
 	struct nouveau_device *device = nv_device(therm);
 	struct nouveau_therm_priv *tpriv = (void *)therm;
+	struct nouveau_bios *bios = nouveau_bios(therm);
 	struct nouveau_fanpwm_priv *priv;
+	struct nvbios_therm_fan fan;
 	u32 divs, duty;
 
+	nvbios_fan_parse(bios, &fan);
+
 	if (!nouveau_boolopt(device->cfgopt, "NvFanPWM", func->param) ||
-	    !therm->pwm_ctrl ||
+	    !therm->pwm_ctrl || fan.type == NVBIOS_THERM_FAN_TOGGLE ||
 	     therm->pwm_get(therm, func->line, &divs, &duty) == -ENODEV)
 		return -ENODEV;
 
diff --git a/drivers/gpu/drm/nouveau/core/subdev/therm/gm107.c b/drivers/gpu/drm/nouveau/core/subdev/therm/gm107.c
new file mode 100644
index 000000000000..668cf3322285
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/therm/gm107.c
@@ -0,0 +1,93 @@
+/*
+ * Copyright 2014 Martin Peres
+ *
+ * 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: Martin Peres
+ */
+
+#include "priv.h"
+
+struct gm107_therm_priv {
+	struct nouveau_therm_priv base;
+};
+
+static int
+gm107_fan_pwm_ctrl(struct nouveau_therm *therm, int line, bool enable)
+{
+	/* nothing to do, it seems hardwired */
+	return 0;
+}
+
+static int
+gm107_fan_pwm_get(struct nouveau_therm *therm, int line, u32 *divs, u32 *duty)
+{
+	*divs = nv_rd32(therm, 0x10eb20) & 0x1fff;
+	*duty = nv_rd32(therm, 0x10eb24) & 0x1fff;
+	return 0;
+}
+
+static int
+gm107_fan_pwm_set(struct nouveau_therm *therm, int line, u32 divs, u32 duty)
+{
+	nv_mask(therm, 0x10eb10, 0x1fff, divs); /* keep the high bits */
+	nv_wr32(therm, 0x10eb14, duty | 0x80000000);
+	return 0;
+}
+
+static int
+gm107_fan_pwm_clock(struct nouveau_therm *therm, int line)
+{
+	return nv_device(therm)->crystal * 1000;
+}
+
+static int
+gm107_therm_ctor(struct nouveau_object *parent,
+		struct nouveau_object *engine,
+		struct nouveau_oclass *oclass, void *data, u32 size,
+		struct nouveau_object **pobject)
+{
+	struct gm107_therm_priv *priv;
+	int ret;
+
+	ret = nouveau_therm_create(parent, engine, oclass, &priv);
+	*pobject = nv_object(priv);
+	if (ret)
+		return ret;
+
+	priv->base.base.pwm_ctrl = gm107_fan_pwm_ctrl;
+	priv->base.base.pwm_get = gm107_fan_pwm_get;
+	priv->base.base.pwm_set = gm107_fan_pwm_set;
+	priv->base.base.pwm_clock = gm107_fan_pwm_clock;
+	priv->base.base.temp_get = nv84_temp_get;
+	priv->base.base.fan_sense = nva3_therm_fan_sense;
+	priv->base.sensor.program_alarms = nouveau_therm_program_alarms_polling;
+	return nouveau_therm_preinit(&priv->base.base);
+}
+
+struct nouveau_oclass
+gm107_therm_oclass = {
+	.handle = NV_SUBDEV(THERM, 0x117),
+	.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = gm107_therm_ctor,
+		.dtor = _nouveau_therm_dtor,
+		.init = nvd0_therm_init,
+		.fini = nv84_therm_fini,
+	},
+};
diff --git a/drivers/gpu/drm/nouveau/core/subdev/therm/nv84.c b/drivers/gpu/drm/nouveau/core/subdev/therm/nv84.c
index 1d15c52fad0c..14e2e09bfc24 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/therm/nv84.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/therm/nv84.c
@@ -24,6 +24,7 @@
  */
 
 #include "priv.h"
+#include <subdev/fuse.h>
 
 struct nv84_therm_priv {
 	struct nouveau_therm_priv base;
@@ -32,7 +33,25 @@ struct nv84_therm_priv {
 int
 nv84_temp_get(struct nouveau_therm *therm)
 {
-	return nv_rd32(therm, 0x20400);
+	struct nouveau_fuse *fuse = nouveau_fuse(therm);
+
+	if (nv_ro32(fuse, 0x1a8) == 1)
+		return nv_rd32(therm, 0x20400);
+	else
+		return -ENODEV;
+}
+
+void
+nv84_sensor_setup(struct nouveau_therm *therm)
+{
+	struct nouveau_fuse *fuse = nouveau_fuse(therm);
+
+	/* enable temperature reading for cards with insane defaults */
+	if (nv_ro32(fuse, 0x1a8) == 1) {
+		nv_mask(therm, 0x20008, 0x80008000, 0x80000000);
+		nv_mask(therm, 0x2000c, 0x80000003, 0x00000000);
+		mdelay(20); /* wait for the temperature to stabilize */
+	}
 }
 
 static void
@@ -171,6 +190,21 @@ nv84_therm_intr(struct nouveau_subdev *subdev)
 }
 
 static int
+nv84_therm_init(struct nouveau_object *object)
+{
+	struct nv84_therm_priv *priv = (void *)object;
+	int ret;
+
+	ret = nouveau_therm_init(&priv->base.base);
+	if (ret)
+		return ret;
+
+	nv84_sensor_setup(&priv->base.base);
+
+	return 0;
+}
+
+static int
 nv84_therm_ctor(struct nouveau_object *parent,
 		struct nouveau_object *engine,
 		struct nouveau_oclass *oclass, void *data, u32 size,
@@ -228,7 +262,7 @@ nv84_therm_oclass = {
 	.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = nv84_therm_ctor,
 		.dtor = _nouveau_therm_dtor,
-		.init = _nouveau_therm_init,
+		.init = nv84_therm_init,
 		.fini = nv84_therm_fini,
 	},
 };
diff --git a/drivers/gpu/drm/nouveau/core/subdev/therm/nva3.c b/drivers/gpu/drm/nouveau/core/subdev/therm/nva3.c
index 0478b2e3fb1d..7893357a7e9f 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/therm/nva3.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/therm/nva3.c
@@ -51,6 +51,8 @@ nva3_therm_init(struct nouveau_object *object)
 	if (ret)
 		return ret;
 
+	nv84_sensor_setup(&priv->base.base);
+
 	/* enable fan tach, count revolutions per-second */
 	nv_mask(priv, 0x00e720, 0x00000003, 0x00000002);
 	if (tach->func != DCB_GPIO_UNUSED) {
diff --git a/drivers/gpu/drm/nouveau/core/subdev/therm/nvd0.c b/drivers/gpu/drm/nouveau/core/subdev/therm/nvd0.c
index bbf117be572f..b70f7cc649b8 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/therm/nvd0.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/therm/nvd0.c
@@ -114,7 +114,7 @@ nvd0_fan_pwm_clock(struct nouveau_therm *therm, int line)
 		return nv_device(therm)->crystal * 1000 / 10;
 }
 
-static int
+int
 nvd0_therm_init(struct nouveau_object *object)
 {
 	struct nvd0_therm_priv *priv = (void *)object;
@@ -150,6 +150,8 @@ nvd0_therm_ctor(struct nouveau_object *parent,
 	if (ret)
 		return ret;
 
+	nv84_sensor_setup(&priv->base.base);
+
 	priv->base.base.pwm_ctrl = nvd0_fan_pwm_ctrl;
 	priv->base.base.pwm_get = nvd0_fan_pwm_get;
 	priv->base.base.pwm_set = nvd0_fan_pwm_set;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/therm/priv.h b/drivers/gpu/drm/nouveau/core/subdev/therm/priv.h
index 916fca5c7816..7dba8c281a0b 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/therm/priv.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/therm/priv.h
@@ -145,10 +145,13 @@ int nv50_fan_pwm_get(struct nouveau_therm *, int, u32 *, u32 *);
 int nv50_fan_pwm_set(struct nouveau_therm *, int, u32, u32);
 int nv50_fan_pwm_clock(struct nouveau_therm *, int);
 int nv84_temp_get(struct nouveau_therm *therm);
+void nv84_sensor_setup(struct nouveau_therm *therm);
 int nv84_therm_fini(struct nouveau_object *object, bool suspend);
 
 int nva3_therm_fan_sense(struct nouveau_therm *);
 
+int nvd0_therm_init(struct nouveau_object *object);
+
 int nouveau_fanpwm_create(struct nouveau_therm *, struct dcb_gpio_func *);
 int nouveau_fantog_create(struct nouveau_therm *, struct dcb_gpio_func *);
 int nouveau_fannil_create(struct nouveau_therm *);
diff --git a/drivers/gpu/drm/nouveau/core/subdev/vm/base.c b/drivers/gpu/drm/nouveau/core/subdev/vm/base.c
index 7dd680ff2f6f..f75a683bd47a 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/vm/base.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/vm/base.c
@@ -296,7 +296,7 @@ nouveau_vm_get(struct nouveau_vm *vm, u64 size, u32 page_shift,
 	int ret;
 
 	mutex_lock(&nv_subdev(vmm)->mutex);
-	ret = nouveau_mm_head(&vm->mm, page_shift, msize, msize, align,
+	ret = nouveau_mm_head(&vm->mm, 0, page_shift, msize, msize, align,
 			     &vma->node);
 	if (unlikely(ret != 0)) {
 		mutex_unlock(&nv_subdev(vmm)->mutex);
diff --git a/drivers/gpu/drm/nouveau/dispnv04/crtc.c b/drivers/gpu/drm/nouveau/dispnv04/crtc.c
index b90aa5c1f90a..fca6a1f9c20c 100644
--- a/drivers/gpu/drm/nouveau/dispnv04/crtc.c
+++ b/drivers/gpu/drm/nouveau/dispnv04/crtc.c
@@ -1127,7 +1127,7 @@ nv04_crtc_create(struct drm_device *dev, int crtc_num)
 	drm_mode_crtc_set_gamma_size(&nv_crtc->base, 256);
 
 	ret = nouveau_bo_new(dev, 64*64*4, 0x100, TTM_PL_FLAG_VRAM,
-			     0, 0x0000, NULL, &nv_crtc->cursor.nvbo);
+			     0, 0x0000, NULL, NULL, &nv_crtc->cursor.nvbo);
 	if (!ret) {
 		ret = nouveau_bo_pin(nv_crtc->cursor.nvbo, TTM_PL_FLAG_VRAM);
 		if (!ret) {
diff --git a/drivers/gpu/drm/nouveau/dispnv04/overlay.c b/drivers/gpu/drm/nouveau/dispnv04/overlay.c
index b36afcbbc83f..1e9056a8df94 100644
--- a/drivers/gpu/drm/nouveau/dispnv04/overlay.c
+++ b/drivers/gpu/drm/nouveau/dispnv04/overlay.c
@@ -97,7 +97,8 @@ nv10_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 		  uint32_t src_w, uint32_t src_h)
 {
 	struct nvif_device *dev = &nouveau_drm(plane->dev)->device;
-	struct nouveau_plane *nv_plane = (struct nouveau_plane *)plane;
+	struct nouveau_plane *nv_plane =
+		container_of(plane, struct nouveau_plane, base);
 	struct nouveau_framebuffer *nv_fb = nouveau_framebuffer(fb);
 	struct nouveau_crtc *nv_crtc = nouveau_crtc(crtc);
 	struct nouveau_bo *cur = nv_plane->cur;
@@ -173,7 +174,8 @@ static int
 nv10_disable_plane(struct drm_plane *plane)
 {
 	struct nvif_device *dev = &nouveau_drm(plane->dev)->device;
-	struct nouveau_plane *nv_plane = (struct nouveau_plane *)plane;
+	struct nouveau_plane *nv_plane =
+		container_of(plane, struct nouveau_plane, base);
 
 	nvif_wr32(dev, NV_PVIDEO_STOP, 1);
 	if (nv_plane->cur) {
@@ -224,7 +226,8 @@ nv_set_property(struct drm_plane *plane,
 		struct drm_property *property,
 		uint64_t value)
 {
-	struct nouveau_plane *nv_plane = (struct nouveau_plane *)plane;
+	struct nouveau_plane *nv_plane =
+		container_of(plane, struct nouveau_plane, base);
 
 	if (property == nv_plane->props.colorkey)
 		nv_plane->colorkey = value;
@@ -344,7 +347,8 @@ nv04_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 		  uint32_t src_w, uint32_t src_h)
 {
 	struct nvif_device *dev = &nouveau_drm(plane->dev)->device;
-	struct nouveau_plane *nv_plane = (struct nouveau_plane *)plane;
+	struct nouveau_plane *nv_plane =
+		container_of(plane, struct nouveau_plane, base);
 	struct nouveau_framebuffer *nv_fb = nouveau_framebuffer(fb);
 	struct nouveau_bo *cur = nv_plane->cur;
 	uint32_t overlay = 1;
@@ -423,7 +427,8 @@ static int
 nv04_disable_plane(struct drm_plane *plane)
 {
 	struct nvif_device *dev = &nouveau_drm(plane->dev)->device;
-	struct nouveau_plane *nv_plane = (struct nouveau_plane *)plane;
+	struct nouveau_plane *nv_plane =
+		container_of(plane, struct nouveau_plane, base);
 
 	nvif_mask(dev, NV_PVIDEO_OVERLAY, 1, 0);
 	nvif_wr32(dev, NV_PVIDEO_OE_STATE, 0);
diff --git a/drivers/gpu/drm/nouveau/nouveau_abi16.c b/drivers/gpu/drm/nouveau/nouveau_abi16.c
index 615714c1727d..a24faa5e2a2a 100644
--- a/drivers/gpu/drm/nouveau/nouveau_abi16.c
+++ b/drivers/gpu/drm/nouveau/nouveau_abi16.c
@@ -448,7 +448,7 @@ nouveau_abi16_ioctl_notifierobj_alloc(ABI16_IOCTL_ARGS)
 	list_add(&ntfy->head, &chan->notifiers);
 	ntfy->handle = info->handle;
 
-	ret = nouveau_mm_head(&chan->heap, 1, info->size, info->size, 1,
+	ret = nouveau_mm_head(&chan->heap, 0, 1, info->size, info->size, 1,
 			      &ntfy->node);
 	if (ret)
 		goto done;
diff --git a/drivers/gpu/drm/nouveau/nouveau_bo.c b/drivers/gpu/drm/nouveau/nouveau_bo.c
index 01da508625f2..3d474ac03f88 100644
--- a/drivers/gpu/drm/nouveau/nouveau_bo.c
+++ b/drivers/gpu/drm/nouveau/nouveau_bo.c
@@ -88,13 +88,13 @@ nv10_bo_get_tile_region(struct drm_device *dev, int i)
 
 static void
 nv10_bo_put_tile_region(struct drm_device *dev, struct nouveau_drm_tile *tile,
-			struct nouveau_fence *fence)
+			struct fence *fence)
 {
 	struct nouveau_drm *drm = nouveau_drm(dev);
 
 	if (tile) {
 		spin_lock(&drm->tile.lock);
-		tile->fence = nouveau_fence_ref(fence);
+		tile->fence = (struct nouveau_fence *)fence_get(fence);
 		tile->used = false;
 		spin_unlock(&drm->tile.lock);
 	}
@@ -181,7 +181,7 @@ nouveau_bo_fixup_align(struct nouveau_bo *nvbo, u32 flags,
 int
 nouveau_bo_new(struct drm_device *dev, int size, int align,
 	       uint32_t flags, uint32_t tile_mode, uint32_t tile_flags,
-	       struct sg_table *sg,
+	       struct sg_table *sg, struct reservation_object *robj,
 	       struct nouveau_bo **pnvbo)
 {
 	struct nouveau_drm *drm = nouveau_drm(dev);
@@ -230,7 +230,7 @@ nouveau_bo_new(struct drm_device *dev, int size, int align,
 	ret = ttm_bo_init(&drm->ttm.bdev, &nvbo->bo, size,
 			  type, &nvbo->placement,
 			  align >> PAGE_SHIFT, false, NULL, acc_size, sg,
-			  nouveau_bo_del_ttm);
+			  robj, nouveau_bo_del_ttm);
 	if (ret) {
 		/* ttm will call nouveau_bo_del_ttm if it fails.. */
 		return ret;
@@ -241,16 +241,16 @@ nouveau_bo_new(struct drm_device *dev, int size, int align,
 }
 
 static void
-set_placement_list(uint32_t *pl, unsigned *n, uint32_t type, uint32_t flags)
+set_placement_list(struct ttm_place *pl, unsigned *n, uint32_t type, uint32_t flags)
 {
 	*n = 0;
 
 	if (type & TTM_PL_FLAG_VRAM)
-		pl[(*n)++] = TTM_PL_FLAG_VRAM | flags;
+		pl[(*n)++].flags = TTM_PL_FLAG_VRAM | flags;
 	if (type & TTM_PL_FLAG_TT)
-		pl[(*n)++] = TTM_PL_FLAG_TT | flags;
+		pl[(*n)++].flags = TTM_PL_FLAG_TT | flags;
 	if (type & TTM_PL_FLAG_SYSTEM)
-		pl[(*n)++] = TTM_PL_FLAG_SYSTEM | flags;
+		pl[(*n)++].flags = TTM_PL_FLAG_SYSTEM | flags;
 }
 
 static void
@@ -258,6 +258,7 @@ set_placement_range(struct nouveau_bo *nvbo, uint32_t type)
 {
 	struct nouveau_drm *drm = nouveau_bdev(nvbo->bo.bdev);
 	u32 vram_pages = drm->device.info.ram_size >> PAGE_SHIFT;
+	unsigned i, fpfn, lpfn;
 
 	if (drm->device.info.family == NV_DEVICE_INFO_V0_CELSIUS &&
 	    nvbo->tile_mode && (type & TTM_PL_FLAG_VRAM) &&
@@ -269,11 +270,19 @@ set_placement_range(struct nouveau_bo *nvbo, uint32_t type)
 		 * at the same time.
 		 */
 		if (nvbo->tile_flags & NOUVEAU_GEM_TILE_ZETA) {
-			nvbo->placement.fpfn = vram_pages / 2;
-			nvbo->placement.lpfn = ~0;
+			fpfn = vram_pages / 2;
+			lpfn = ~0;
 		} else {
-			nvbo->placement.fpfn = 0;
-			nvbo->placement.lpfn = vram_pages / 2;
+			fpfn = 0;
+			lpfn = vram_pages / 2;
+		}
+		for (i = 0; i < nvbo->placement.num_placement; ++i) {
+			nvbo->placements[i].fpfn = fpfn;
+			nvbo->placements[i].lpfn = lpfn;
+		}
+		for (i = 0; i < nvbo->placement.num_busy_placement; ++i) {
+			nvbo->busy_placements[i].fpfn = fpfn;
+			nvbo->busy_placements[i].lpfn = lpfn;
 		}
 	}
 }
@@ -961,13 +970,14 @@ nouveau_bo_move_m2mf(struct ttm_buffer_object *bo, int evict, bool intr,
 	}
 
 	mutex_lock_nested(&cli->mutex, SINGLE_DEPTH_NESTING);
-	ret = nouveau_fence_sync(bo->sync_obj, chan);
+	ret = nouveau_fence_sync(nouveau_bo(bo), chan, true, intr);
 	if (ret == 0) {
 		ret = drm->ttm.move(chan, bo, &bo->mem, new_mem);
 		if (ret == 0) {
 			ret = nouveau_fence_new(chan, false, &fence);
 			if (ret == 0) {
-				ret = ttm_bo_move_accel_cleanup(bo, fence,
+				ret = ttm_bo_move_accel_cleanup(bo,
+								&fence->base,
 								evict,
 								no_wait_gpu,
 								new_mem);
@@ -1041,12 +1051,15 @@ static int
 nouveau_bo_move_flipd(struct ttm_buffer_object *bo, bool evict, bool intr,
 		      bool no_wait_gpu, struct ttm_mem_reg *new_mem)
 {
-	u32 placement_memtype = TTM_PL_FLAG_TT | TTM_PL_MASK_CACHING;
+	struct ttm_place placement_memtype = {
+		.fpfn = 0,
+		.lpfn = 0,
+		.flags = TTM_PL_FLAG_TT | TTM_PL_MASK_CACHING
+	};
 	struct ttm_placement placement;
 	struct ttm_mem_reg tmp_mem;
 	int ret;
 
-	placement.fpfn = placement.lpfn = 0;
 	placement.num_placement = placement.num_busy_placement = 1;
 	placement.placement = placement.busy_placement = &placement_memtype;
 
@@ -1074,12 +1087,15 @@ static int
 nouveau_bo_move_flips(struct ttm_buffer_object *bo, bool evict, bool intr,
 		      bool no_wait_gpu, struct ttm_mem_reg *new_mem)
 {
-	u32 placement_memtype = TTM_PL_FLAG_TT | TTM_PL_MASK_CACHING;
+	struct ttm_place placement_memtype = {
+		.fpfn = 0,
+		.lpfn = 0,
+		.flags = TTM_PL_FLAG_TT | TTM_PL_MASK_CACHING
+	};
 	struct ttm_placement placement;
 	struct ttm_mem_reg tmp_mem;
 	int ret;
 
-	placement.fpfn = placement.lpfn = 0;
 	placement.num_placement = placement.num_busy_placement = 1;
 	placement.placement = placement.busy_placement = &placement_memtype;
 
@@ -1152,8 +1168,9 @@ nouveau_bo_vm_cleanup(struct ttm_buffer_object *bo,
 {
 	struct nouveau_drm *drm = nouveau_bdev(bo->bdev);
 	struct drm_device *dev = drm->dev;
+	struct fence *fence = reservation_object_get_excl(bo->resv);
 
-	nv10_bo_put_tile_region(dev, *old_tile, bo->sync_obj);
+	nv10_bo_put_tile_region(dev, *old_tile, fence);
 	*old_tile = new_tile;
 }
 
@@ -1197,9 +1214,7 @@ nouveau_bo_move(struct ttm_buffer_object *bo, bool evict, bool intr,
 	}
 
 	/* Fallback to software copy. */
-	spin_lock(&bo->bdev->fence_lock);
 	ret = ttm_bo_wait(bo, true, intr, no_wait_gpu);
-	spin_unlock(&bo->bdev->fence_lock);
 	if (ret == 0)
 		ret = ttm_bo_move_memcpy(bo, evict, no_wait_gpu, new_mem);
 
@@ -1294,7 +1309,7 @@ nouveau_ttm_fault_reserve_notify(struct ttm_buffer_object *bo)
 	struct nouveau_bo *nvbo = nouveau_bo(bo);
 	struct nvif_device *device = &drm->device;
 	u32 mappable = nv_device_resource_len(nvkm_device(device), 1) >> PAGE_SHIFT;
-	int ret;
+	int i, ret;
 
 	/* as long as the bo isn't in vram, and isn't tiled, we've got
 	 * nothing to do here.
@@ -1319,9 +1334,16 @@ nouveau_ttm_fault_reserve_notify(struct ttm_buffer_object *bo)
 	    bo->mem.start + bo->mem.num_pages < mappable)
 		return 0;
 
+	for (i = 0; i < nvbo->placement.num_placement; ++i) {
+		nvbo->placements[i].fpfn = 0;
+		nvbo->placements[i].lpfn = mappable;
+	}
+
+	for (i = 0; i < nvbo->placement.num_busy_placement; ++i) {
+		nvbo->busy_placements[i].fpfn = 0;
+		nvbo->busy_placements[i].lpfn = mappable;
+	}
 
-	nvbo->placement.fpfn = 0;
-	nvbo->placement.lpfn = mappable;
 	nouveau_bo_placement_set(nvbo, TTM_PL_FLAG_VRAM, 0);
 	return nouveau_bo_validate(nvbo, false, false);
 }
@@ -1436,47 +1458,14 @@ nouveau_ttm_tt_unpopulate(struct ttm_tt *ttm)
 }
 
 void
-nouveau_bo_fence(struct nouveau_bo *nvbo, struct nouveau_fence *fence)
-{
-	struct nouveau_fence *new_fence = nouveau_fence_ref(fence);
-	struct nouveau_fence *old_fence = NULL;
-
-	spin_lock(&nvbo->bo.bdev->fence_lock);
-	old_fence = nvbo->bo.sync_obj;
-	nvbo->bo.sync_obj = new_fence;
-	spin_unlock(&nvbo->bo.bdev->fence_lock);
-
-	nouveau_fence_unref(&old_fence);
-}
-
-static void
-nouveau_bo_fence_unref(void **sync_obj)
-{
-	nouveau_fence_unref((struct nouveau_fence **)sync_obj);
-}
-
-static void *
-nouveau_bo_fence_ref(void *sync_obj)
+nouveau_bo_fence(struct nouveau_bo *nvbo, struct nouveau_fence *fence, bool exclusive)
 {
-	return nouveau_fence_ref(sync_obj);
-}
+	struct reservation_object *resv = nvbo->bo.resv;
 
-static bool
-nouveau_bo_fence_signalled(void *sync_obj)
-{
-	return nouveau_fence_done(sync_obj);
-}
-
-static int
-nouveau_bo_fence_wait(void *sync_obj, bool lazy, bool intr)
-{
-	return nouveau_fence_wait(sync_obj, lazy, intr);
-}
-
-static int
-nouveau_bo_fence_flush(void *sync_obj)
-{
-	return 0;
+	if (exclusive)
+		reservation_object_add_excl_fence(resv, &fence->base);
+	else if (fence)
+		reservation_object_add_shared_fence(resv, &fence->base);
 }
 
 struct ttm_bo_driver nouveau_bo_driver = {
@@ -1489,11 +1478,6 @@ struct ttm_bo_driver nouveau_bo_driver = {
 	.move_notify = nouveau_bo_move_ntfy,
 	.move = nouveau_bo_move,
 	.verify_access = nouveau_bo_verify_access,
-	.sync_obj_signaled = nouveau_bo_fence_signalled,
-	.sync_obj_wait = nouveau_bo_fence_wait,
-	.sync_obj_flush = nouveau_bo_fence_flush,
-	.sync_obj_unref = nouveau_bo_fence_unref,
-	.sync_obj_ref = nouveau_bo_fence_ref,
 	.fault_reserve_notify = &nouveau_ttm_fault_reserve_notify,
 	.io_mem_reserve = &nouveau_ttm_io_mem_reserve,
 	.io_mem_free = &nouveau_ttm_io_mem_free,
diff --git a/drivers/gpu/drm/nouveau/nouveau_bo.h b/drivers/gpu/drm/nouveau/nouveau_bo.h
index ff17c1f432fc..22d2c764d80b 100644
--- a/drivers/gpu/drm/nouveau/nouveau_bo.h
+++ b/drivers/gpu/drm/nouveau/nouveau_bo.h
@@ -1,6 +1,8 @@
 #ifndef __NOUVEAU_BO_H__
 #define __NOUVEAU_BO_H__
 
+#include <drm/drm_gem.h>
+
 struct nouveau_channel;
 struct nouveau_fence;
 struct nouveau_vma;
@@ -9,8 +11,8 @@ struct nouveau_bo {
 	struct ttm_buffer_object bo;
 	struct ttm_placement placement;
 	u32 valid_domains;
-	u32 placements[3];
-	u32 busy_placements[3];
+	struct ttm_place placements[3];
+	struct ttm_place busy_placements[3];
 	struct ttm_bo_kmap_obj kmap;
 	struct list_head head;
 
@@ -68,6 +70,7 @@ extern struct ttm_bo_driver nouveau_bo_driver;
 void nouveau_bo_move_init(struct nouveau_drm *);
 int  nouveau_bo_new(struct drm_device *, int size, int align, u32 flags,
 		    u32 tile_mode, u32 tile_flags, struct sg_table *sg,
+		    struct reservation_object *robj,
 		    struct nouveau_bo **);
 int  nouveau_bo_pin(struct nouveau_bo *, u32 flags);
 int  nouveau_bo_unpin(struct nouveau_bo *);
@@ -78,7 +81,7 @@ u16  nouveau_bo_rd16(struct nouveau_bo *, unsigned index);
 void nouveau_bo_wr16(struct nouveau_bo *, unsigned index, u16 val);
 u32  nouveau_bo_rd32(struct nouveau_bo *, unsigned index);
 void nouveau_bo_wr32(struct nouveau_bo *, unsigned index, u32 val);
-void nouveau_bo_fence(struct nouveau_bo *, struct nouveau_fence *);
+void nouveau_bo_fence(struct nouveau_bo *, struct nouveau_fence *, bool exclusive);
 int  nouveau_bo_validate(struct nouveau_bo *, bool interruptible,
 			 bool no_wait_gpu);
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_chan.c b/drivers/gpu/drm/nouveau/nouveau_chan.c
index 3440fc999f2f..589dbb582da2 100644
--- a/drivers/gpu/drm/nouveau/nouveau_chan.c
+++ b/drivers/gpu/drm/nouveau/nouveau_chan.c
@@ -36,7 +36,7 @@
 #include "nouveau_abi16.h"
 
 MODULE_PARM_DESC(vram_pushbuf, "Create DMA push buffers in VRAM");
-static int nouveau_vram_pushbuf;
+int nouveau_vram_pushbuf;
 module_param_named(vram_pushbuf, nouveau_vram_pushbuf, int, 0400);
 
 int
@@ -106,7 +106,7 @@ nouveau_channel_prep(struct nouveau_drm *drm, struct nvif_device *device,
 	if (nouveau_vram_pushbuf)
 		target = TTM_PL_FLAG_VRAM;
 
-	ret = nouveau_bo_new(drm->dev, size, 0, target, 0, 0, NULL,
+	ret = nouveau_bo_new(drm->dev, size, 0, target, 0, 0, NULL, NULL,
 			    &chan->push.buffer);
 	if (ret == 0) {
 		ret = nouveau_bo_pin(chan->push.buffer, target);
diff --git a/drivers/gpu/drm/nouveau/nouveau_chan.h b/drivers/gpu/drm/nouveau/nouveau_chan.h
index 20163709d608..8309c24ee698 100644
--- a/drivers/gpu/drm/nouveau/nouveau_chan.h
+++ b/drivers/gpu/drm/nouveau/nouveau_chan.h
@@ -47,4 +47,6 @@ int  nouveau_channel_new(struct nouveau_drm *, struct nvif_device *,
 void nouveau_channel_del(struct nouveau_channel **);
 int  nouveau_channel_idle(struct nouveau_channel *);
 
+extern int nouveau_vram_pushbuf;
+
 #endif
diff --git a/drivers/gpu/drm/nouveau/nouveau_connector.c b/drivers/gpu/drm/nouveau/nouveau_connector.c
index 1ec44c83e919..c8ac9482cf2e 100644
--- a/drivers/gpu/drm/nouveau/nouveau_connector.c
+++ b/drivers/gpu/drm/nouveau/nouveau_connector.c
@@ -45,15 +45,15 @@
 #include <nvif/event.h>
 
 MODULE_PARM_DESC(tv_disable, "Disable TV-out detection");
-static int nouveau_tv_disable = 0;
+int nouveau_tv_disable = 0;
 module_param_named(tv_disable, nouveau_tv_disable, int, 0400);
 
 MODULE_PARM_DESC(ignorelid, "Ignore ACPI lid status");
-static int nouveau_ignorelid = 0;
+int nouveau_ignorelid = 0;
 module_param_named(ignorelid, nouveau_ignorelid, int, 0400);
 
 MODULE_PARM_DESC(duallink, "Allow dual-link TMDS (default: enabled)");
-static int nouveau_duallink = 1;
+int nouveau_duallink = 1;
 module_param_named(duallink, nouveau_duallink, int, 0400);
 
 struct nouveau_encoder *
diff --git a/drivers/gpu/drm/nouveau/nouveau_connector.h b/drivers/gpu/drm/nouveau/nouveau_connector.h
index 68029d041dd2..629a380c7085 100644
--- a/drivers/gpu/drm/nouveau/nouveau_connector.h
+++ b/drivers/gpu/drm/nouveau/nouveau_connector.h
@@ -105,4 +105,8 @@ nouveau_crtc_connector_get(struct nouveau_crtc *nv_crtc)
 struct drm_connector *
 nouveau_connector_create(struct drm_device *, int index);
 
+extern int nouveau_tv_disable;
+extern int nouveau_ignorelid;
+extern int nouveau_duallink;
+
 #endif /* __NOUVEAU_CONNECTOR_H__ */
diff --git a/drivers/gpu/drm/nouveau/nouveau_display.c b/drivers/gpu/drm/nouveau/nouveau_display.c
index 4a21b2b06ce2..a88e6927f571 100644
--- a/drivers/gpu/drm/nouveau/nouveau_display.c
+++ b/drivers/gpu/drm/nouveau/nouveau_display.c
@@ -126,7 +126,7 @@ nouveau_display_scanoutpos_head(struct drm_crtc *crtc, int *vpos, int *hpos,
 	if (etime) *etime = ns_to_ktime(args.scan.time[1]);
 
 	if (*vpos < 0)
-		ret |= DRM_SCANOUTPOS_INVBL;
+		ret |= DRM_SCANOUTPOS_IN_VBLANK;
 	return ret;
 }
 
@@ -657,7 +657,7 @@ nouveau_page_flip_emit(struct nouveau_channel *chan,
 	spin_unlock_irqrestore(&dev->event_lock, flags);
 
 	/* Synchronize with the old framebuffer */
-	ret = nouveau_fence_sync(old_bo->bo.sync_obj, chan);
+	ret = nouveau_fence_sync(old_bo, chan, false, false);
 	if (ret)
 		goto fail;
 
@@ -716,19 +716,24 @@ nouveau_crtc_page_flip(struct drm_crtc *crtc, struct drm_framebuffer *fb,
 	}
 
 	mutex_lock(&cli->mutex);
-
-	/* synchronise rendering channel with the kernel's channel */
-	spin_lock(&new_bo->bo.bdev->fence_lock);
-	fence = nouveau_fence_ref(new_bo->bo.sync_obj);
-	spin_unlock(&new_bo->bo.bdev->fence_lock);
-	ret = nouveau_fence_sync(fence, chan);
-	nouveau_fence_unref(&fence);
+	ret = ttm_bo_reserve(&new_bo->bo, true, false, false, NULL);
 	if (ret)
 		goto fail_unpin;
 
-	ret = ttm_bo_reserve(&old_bo->bo, true, false, false, NULL);
-	if (ret)
+	/* synchronise rendering channel with the kernel's channel */
+	ret = nouveau_fence_sync(new_bo, chan, false, true);
+	if (ret) {
+		ttm_bo_unreserve(&new_bo->bo);
 		goto fail_unpin;
+	}
+
+	if (new_bo != old_bo) {
+		ttm_bo_unreserve(&new_bo->bo);
+
+		ret = ttm_bo_reserve(&old_bo->bo, true, false, false, NULL);
+		if (ret)
+			goto fail_unpin;
+	}
 
 	/* Initialize a page flip struct */
 	*s = (struct nouveau_page_flip_state)
@@ -774,7 +779,7 @@ nouveau_crtc_page_flip(struct drm_crtc *crtc, struct drm_framebuffer *fb,
 	/* Update the crtc struct and cleanup */
 	crtc->primary->fb = fb;
 
-	nouveau_bo_fence(old_bo, fence);
+	nouveau_bo_fence(old_bo, fence, false);
 	ttm_bo_unreserve(&old_bo->bo);
 	if (old_bo != new_bo)
 		nouveau_bo_unpin(old_bo);
diff --git a/drivers/gpu/drm/nouveau/nouveau_drm.c b/drivers/gpu/drm/nouveau/nouveau_drm.c
index 3ed32dd90303..57238076049f 100644
--- a/drivers/gpu/drm/nouveau/nouveau_drm.c
+++ b/drivers/gpu/drm/nouveau/nouveau_drm.c
@@ -51,6 +51,7 @@
 #include "nouveau_fence.h"
 #include "nouveau_debugfs.h"
 #include "nouveau_usif.h"
+#include "nouveau_connector.h"
 
 MODULE_PARM_DESC(config, "option string to pass to driver core");
 static char *nouveau_config;
@@ -73,7 +74,9 @@ MODULE_PARM_DESC(runpm, "disable (0), force enable (1), optimus only default (-1
 int nouveau_runtime_pm = -1;
 module_param_named(runpm, nouveau_runtime_pm, int, 0400);
 
-static struct drm_driver driver;
+static struct drm_driver driver_stub;
+static struct drm_driver driver_pci;
+static struct drm_driver driver_platform;
 
 static u64
 nouveau_pci_name(struct pci_dev *pdev)
@@ -322,7 +325,7 @@ static int nouveau_drm_probe(struct pci_dev *pdev,
 
 	pci_set_master(pdev);
 
-	ret = drm_get_pci_dev(pdev, pent, &driver);
+	ret = drm_get_pci_dev(pdev, pent, &driver_pci);
 	if (ret) {
 		nouveau_object_ref(NULL, (struct nouveau_object **)&device);
 		return ret;
@@ -831,7 +834,7 @@ nouveau_driver_fops = {
 };
 
 static struct drm_driver
-driver = {
+driver_stub = {
 	.driver_features =
 		DRIVER_USE_AGP |
 		DRIVER_GEM | DRIVER_MODESET | DRIVER_PRIME | DRIVER_RENDER,
@@ -1002,6 +1005,23 @@ static int nouveau_pmops_runtime_idle(struct device *dev)
 	return 1;
 }
 
+static void nouveau_display_options(void)
+{
+	DRM_DEBUG_DRIVER("Loading Nouveau with parameters:\n");
+
+	DRM_DEBUG_DRIVER("... tv_disable   : %d\n", nouveau_tv_disable);
+	DRM_DEBUG_DRIVER("... ignorelid    : %d\n", nouveau_ignorelid);
+	DRM_DEBUG_DRIVER("... duallink     : %d\n", nouveau_duallink);
+	DRM_DEBUG_DRIVER("... nofbaccel    : %d\n", nouveau_nofbaccel);
+	DRM_DEBUG_DRIVER("... config       : %s\n", nouveau_config);
+	DRM_DEBUG_DRIVER("... debug        : %s\n", nouveau_debug);
+	DRM_DEBUG_DRIVER("... noaccel      : %d\n", nouveau_noaccel);
+	DRM_DEBUG_DRIVER("... modeset      : %d\n", nouveau_modeset);
+	DRM_DEBUG_DRIVER("... runpm        : %d\n", nouveau_runtime_pm);
+	DRM_DEBUG_DRIVER("... vram_pushbuf : %d\n", nouveau_vram_pushbuf);
+	DRM_DEBUG_DRIVER("... pstate       : %d\n", nouveau_pstate);
+}
+
 static const struct dev_pm_ops nouveau_pm_ops = {
 	.suspend = nouveau_pmops_suspend,
 	.resume = nouveau_pmops_resume,
@@ -1037,7 +1057,7 @@ nouveau_platform_device_create_(struct platform_device *pdev, int size,
 	if (err)
 		return ERR_PTR(err);
 
-	drm = drm_dev_alloc(&driver, &pdev->dev);
+	drm = drm_dev_alloc(&driver_platform, &pdev->dev);
 	if (!drm) {
 		err = -ENOMEM;
 		goto err_free;
@@ -1062,6 +1082,13 @@ EXPORT_SYMBOL(nouveau_platform_device_create_);
 static int __init
 nouveau_drm_init(void)
 {
+	driver_pci = driver_stub;
+	driver_pci.set_busid = drm_pci_set_busid;
+	driver_platform = driver_stub;
+	driver_platform.set_busid = drm_platform_set_busid;
+
+	nouveau_display_options();
+
 	if (nouveau_modeset == -1) {
 #ifdef CONFIG_VGA_CONSOLE
 		if (vgacon_text_force())
@@ -1073,7 +1100,7 @@ nouveau_drm_init(void)
 		return 0;
 
 	nouveau_register_dsm_handler();
-	return drm_pci_init(&driver, &nouveau_drm_pci_driver);
+	return drm_pci_init(&driver_pci, &nouveau_drm_pci_driver);
 }
 
 static void __exit
@@ -1082,7 +1109,7 @@ nouveau_drm_exit(void)
 	if (!nouveau_modeset)
 		return;
 
-	drm_pci_exit(&driver, &nouveau_drm_pci_driver);
+	drm_pci_exit(&driver_pci, &nouveau_drm_pci_driver);
 	nouveau_unregister_dsm_handler();
 }
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_drm.h b/drivers/gpu/drm/nouveau/nouveau_drm.h
index b02b02452c85..8ae36f265fb8 100644
--- a/drivers/gpu/drm/nouveau/nouveau_drm.h
+++ b/drivers/gpu/drm/nouveau/nouveau_drm.h
@@ -10,7 +10,7 @@
 
 #define DRIVER_MAJOR		1
 #define DRIVER_MINOR		2
-#define DRIVER_PATCHLEVEL	0
+#define DRIVER_PATCHLEVEL	1
 
 /*
  * 1.1.1:
@@ -26,6 +26,8 @@
  * 1.2.0:
  * 	- object api exposed to userspace
  * 	- fermi,kepler,maxwell zbc
+ * 1.2.1:
+ *      - allow concurrent access to bo's mapped read/write.
  */
 
 #include <nvif/client.h>
diff --git a/drivers/gpu/drm/nouveau/nouveau_fbcon.c b/drivers/gpu/drm/nouveau/nouveau_fbcon.c
index 49fe6075cc7c..593ef8a2a069 100644
--- a/drivers/gpu/drm/nouveau/nouveau_fbcon.c
+++ b/drivers/gpu/drm/nouveau/nouveau_fbcon.c
@@ -52,7 +52,7 @@
 #include "nouveau_crtc.h"
 
 MODULE_PARM_DESC(nofbaccel, "Disable fbcon acceleration");
-static int nouveau_nofbaccel = 0;
+int nouveau_nofbaccel = 0;
 module_param_named(nofbaccel, nouveau_nofbaccel, int, 0400);
 
 static void
@@ -308,7 +308,8 @@ static int
 nouveau_fbcon_create(struct drm_fb_helper *helper,
 		     struct drm_fb_helper_surface_size *sizes)
 {
-	struct nouveau_fbdev *fbcon = (struct nouveau_fbdev *)helper;
+	struct nouveau_fbdev *fbcon =
+		container_of(helper, struct nouveau_fbdev, helper);
 	struct drm_device *dev = fbcon->dev;
 	struct nouveau_drm *drm = nouveau_drm(dev);
 	struct nvif_device *device = &drm->device;
diff --git a/drivers/gpu/drm/nouveau/nouveau_fbcon.h b/drivers/gpu/drm/nouveau/nouveau_fbcon.h
index 0b465c7d3907..6208e70e4a1c 100644
--- a/drivers/gpu/drm/nouveau/nouveau_fbcon.h
+++ b/drivers/gpu/drm/nouveau/nouveau_fbcon.h
@@ -73,5 +73,8 @@ void nouveau_fbcon_accel_save_disable(struct drm_device *dev);
 void nouveau_fbcon_accel_restore(struct drm_device *dev);
 
 void nouveau_fbcon_output_poll_changed(struct drm_device *dev);
+
+extern int nouveau_nofbaccel;
+
 #endif /* __NV50_FBCON_H__ */
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_fence.c b/drivers/gpu/drm/nouveau/nouveau_fence.c
index 0a93114158cd..515cd9aebb99 100644
--- a/drivers/gpu/drm/nouveau/nouveau_fence.c
+++ b/drivers/gpu/drm/nouveau/nouveau_fence.c
@@ -28,6 +28,7 @@
 
 #include <linux/ktime.h>
 #include <linux/hrtimer.h>
+#include <trace/events/fence.h>
 
 #include <nvif/notify.h>
 #include <nvif/event.h>
@@ -36,123 +37,234 @@
 #include "nouveau_dma.h"
 #include "nouveau_fence.h"
 
-struct fence_work {
-	struct work_struct base;
-	struct list_head head;
-	void (*func)(void *);
-	void *data;
-};
+static const struct fence_ops nouveau_fence_ops_uevent;
+static const struct fence_ops nouveau_fence_ops_legacy;
+
+static inline struct nouveau_fence *
+from_fence(struct fence *fence)
+{
+	return container_of(fence, struct nouveau_fence, base);
+}
+
+static inline struct nouveau_fence_chan *
+nouveau_fctx(struct nouveau_fence *fence)
+{
+	return container_of(fence->base.lock, struct nouveau_fence_chan, lock);
+}
 
 static void
 nouveau_fence_signal(struct nouveau_fence *fence)
 {
-	struct fence_work *work, *temp;
+	fence_signal_locked(&fence->base);
+	list_del(&fence->head);
+
+	if (test_bit(FENCE_FLAG_USER_BITS, &fence->base.flags)) {
+		struct nouveau_fence_chan *fctx = nouveau_fctx(fence);
 
-	list_for_each_entry_safe(work, temp, &fence->work, head) {
-		schedule_work(&work->base);
-		list_del(&work->head);
+		if (!--fctx->notify_ref)
+			nvif_notify_put(&fctx->notify);
 	}
 
-	fence->channel = NULL;
-	list_del(&fence->head);
+	fence_put(&fence->base);
+}
+
+static struct nouveau_fence *
+nouveau_local_fence(struct fence *fence, struct nouveau_drm *drm) {
+	struct nouveau_fence_priv *priv = (void*)drm->fence;
+
+	if (fence->ops != &nouveau_fence_ops_legacy &&
+	    fence->ops != &nouveau_fence_ops_uevent)
+		return NULL;
+
+	if (fence->context < priv->context_base ||
+	    fence->context >= priv->context_base + priv->contexts)
+		return NULL;
+
+	return from_fence(fence);
 }
 
 void
 nouveau_fence_context_del(struct nouveau_fence_chan *fctx)
 {
-	struct nouveau_fence *fence, *fnext;
-	spin_lock(&fctx->lock);
-	list_for_each_entry_safe(fence, fnext, &fctx->pending, head) {
+	struct nouveau_fence *fence;
+
+	nvif_notify_fini(&fctx->notify);
+
+	spin_lock_irq(&fctx->lock);
+	while (!list_empty(&fctx->pending)) {
+		fence = list_entry(fctx->pending.next, typeof(*fence), head);
+
 		nouveau_fence_signal(fence);
+		fence->channel = NULL;
 	}
-	spin_unlock(&fctx->lock);
+	spin_unlock_irq(&fctx->lock);
+}
+
+static void
+nouveau_fence_context_put(struct kref *fence_ref)
+{
+	kfree(container_of(fence_ref, struct nouveau_fence_chan, fence_ref));
 }
 
 void
-nouveau_fence_context_new(struct nouveau_fence_chan *fctx)
+nouveau_fence_context_free(struct nouveau_fence_chan *fctx)
+{
+	kref_put(&fctx->fence_ref, nouveau_fence_context_put);
+}
+
+static void
+nouveau_fence_update(struct nouveau_channel *chan, struct nouveau_fence_chan *fctx)
+{
+	struct nouveau_fence *fence;
+
+	u32 seq = fctx->read(chan);
+
+	while (!list_empty(&fctx->pending)) {
+		fence = list_entry(fctx->pending.next, typeof(*fence), head);
+
+		if ((int)(seq - fence->base.seqno) < 0)
+			return;
+
+		nouveau_fence_signal(fence);
+	}
+}
+
+static int
+nouveau_fence_wait_uevent_handler(struct nvif_notify *notify)
 {
+	struct nouveau_fence_chan *fctx =
+		container_of(notify, typeof(*fctx), notify);
+	unsigned long flags;
+
+	spin_lock_irqsave(&fctx->lock, flags);
+	if (!list_empty(&fctx->pending)) {
+		struct nouveau_fence *fence;
+
+		fence = list_entry(fctx->pending.next, typeof(*fence), head);
+		nouveau_fence_update(fence->channel, fctx);
+	}
+	spin_unlock_irqrestore(&fctx->lock, flags);
+
+	/* Always return keep here. NVIF refcount is handled with nouveau_fence_update */
+	return NVIF_NOTIFY_KEEP;
+}
+
+void
+nouveau_fence_context_new(struct nouveau_channel *chan, struct nouveau_fence_chan *fctx)
+{
+	struct nouveau_fence_priv *priv = (void*)chan->drm->fence;
+	struct nouveau_cli *cli = (void *)nvif_client(chan->object);
+	int ret;
+
 	INIT_LIST_HEAD(&fctx->flip);
 	INIT_LIST_HEAD(&fctx->pending);
 	spin_lock_init(&fctx->lock);
+	fctx->context = priv->context_base + chan->chid;
+
+	if (chan == chan->drm->cechan)
+		strcpy(fctx->name, "copy engine channel");
+	else if (chan == chan->drm->channel)
+		strcpy(fctx->name, "generic kernel channel");
+	else
+		strcpy(fctx->name, nvkm_client(&cli->base)->name);
+
+	kref_init(&fctx->fence_ref);
+	if (!priv->uevent)
+		return;
+
+	ret = nvif_notify_init(chan->object, NULL,
+			 nouveau_fence_wait_uevent_handler, false,
+			 G82_CHANNEL_DMA_V0_NTFY_UEVENT,
+			 &(struct nvif_notify_uevent_req) { },
+			 sizeof(struct nvif_notify_uevent_req),
+			 sizeof(struct nvif_notify_uevent_rep),
+			 &fctx->notify);
+
+	WARN_ON(ret);
 }
 
+struct nouveau_fence_work {
+	struct work_struct work;
+	struct fence_cb cb;
+	void (*func)(void *);
+	void *data;
+};
+
 static void
 nouveau_fence_work_handler(struct work_struct *kwork)
 {
-	struct fence_work *work = container_of(kwork, typeof(*work), base);
+	struct nouveau_fence_work *work = container_of(kwork, typeof(*work), work);
 	work->func(work->data);
 	kfree(work);
 }
 
+static void nouveau_fence_work_cb(struct fence *fence, struct fence_cb *cb)
+{
+	struct nouveau_fence_work *work = container_of(cb, typeof(*work), cb);
+
+	schedule_work(&work->work);
+}
+
 void
-nouveau_fence_work(struct nouveau_fence *fence,
+nouveau_fence_work(struct fence *fence,
 		   void (*func)(void *), void *data)
 {
-	struct nouveau_channel *chan = fence->channel;
-	struct nouveau_fence_chan *fctx;
-	struct fence_work *work = NULL;
+	struct nouveau_fence_work *work;
 
-	if (nouveau_fence_done(fence)) {
-		func(data);
-		return;
-	}
+	if (fence_is_signaled(fence))
+		goto err;
 
-	fctx = chan->fence;
 	work = kmalloc(sizeof(*work), GFP_KERNEL);
 	if (!work) {
-		WARN_ON(nouveau_fence_wait(fence, false, false));
-		func(data);
-		return;
-	}
-
-	spin_lock(&fctx->lock);
-	if (!fence->channel) {
-		spin_unlock(&fctx->lock);
-		kfree(work);
-		func(data);
-		return;
+		/*
+		 * this might not be a nouveau fence any more,
+		 * so force a lazy wait here
+		 */
+		WARN_ON(nouveau_fence_wait((struct nouveau_fence *)fence,
+					   true, false));
+		goto err;
 	}
 
-	INIT_WORK(&work->base, nouveau_fence_work_handler);
+	INIT_WORK(&work->work, nouveau_fence_work_handler);
 	work->func = func;
 	work->data = data;
-	list_add(&work->head, &fence->work);
-	spin_unlock(&fctx->lock);
-}
-
-static void
-nouveau_fence_update(struct nouveau_channel *chan)
-{
-	struct nouveau_fence_chan *fctx = chan->fence;
-	struct nouveau_fence *fence, *fnext;
 
-	spin_lock(&fctx->lock);
-	list_for_each_entry_safe(fence, fnext, &fctx->pending, head) {
-		if (fctx->read(chan) < fence->sequence)
-			break;
+	if (fence_add_callback(fence, &work->cb, nouveau_fence_work_cb) < 0)
+		goto err_free;
+	return;
 
-		nouveau_fence_signal(fence);
-		nouveau_fence_unref(&fence);
-	}
-	spin_unlock(&fctx->lock);
+err_free:
+	kfree(work);
+err:
+	func(data);
 }
 
 int
 nouveau_fence_emit(struct nouveau_fence *fence, struct nouveau_channel *chan)
 {
 	struct nouveau_fence_chan *fctx = chan->fence;
+	struct nouveau_fence_priv *priv = (void*)chan->drm->fence;
 	int ret;
 
 	fence->channel  = chan;
 	fence->timeout  = jiffies + (15 * HZ);
-	fence->sequence = ++fctx->sequence;
 
+	if (priv->uevent)
+		fence_init(&fence->base, &nouveau_fence_ops_uevent,
+			   &fctx->lock, fctx->context, ++fctx->sequence);
+	else
+		fence_init(&fence->base, &nouveau_fence_ops_legacy,
+			   &fctx->lock, fctx->context, ++fctx->sequence);
+	kref_get(&fctx->fence_ref);
+
+	trace_fence_emit(&fence->base);
 	ret = fctx->emit(fence);
 	if (!ret) {
-		kref_get(&fence->kref);
-		spin_lock(&fctx->lock);
+		fence_get(&fence->base);
+		spin_lock_irq(&fctx->lock);
+		nouveau_fence_update(chan, fctx);
 		list_add_tail(&fence->head, &fctx->pending);
-		spin_unlock(&fctx->lock);
+		spin_unlock_irq(&fctx->lock);
 	}
 
 	return ret;
@@ -161,114 +273,70 @@ nouveau_fence_emit(struct nouveau_fence *fence, struct nouveau_channel *chan)
 bool
 nouveau_fence_done(struct nouveau_fence *fence)
 {
-	if (fence->channel)
-		nouveau_fence_update(fence->channel);
-	return !fence->channel;
-}
+	if (fence->base.ops == &nouveau_fence_ops_legacy ||
+	    fence->base.ops == &nouveau_fence_ops_uevent) {
+		struct nouveau_fence_chan *fctx = nouveau_fctx(fence);
+		unsigned long flags;
 
-struct nouveau_fence_wait {
-	struct nouveau_fence_priv *priv;
-	struct nvif_notify notify;
-};
+		if (test_bit(FENCE_FLAG_SIGNALED_BIT, &fence->base.flags))
+			return true;
 
-static int
-nouveau_fence_wait_uevent_handler(struct nvif_notify *notify)
-{
-	struct nouveau_fence_wait *wait =
-		container_of(notify, typeof(*wait), notify);
-	wake_up_all(&wait->priv->waiting);
-	return NVIF_NOTIFY_KEEP;
+		spin_lock_irqsave(&fctx->lock, flags);
+		nouveau_fence_update(fence->channel, fctx);
+		spin_unlock_irqrestore(&fctx->lock, flags);
+	}
+	return fence_is_signaled(&fence->base);
 }
 
-static int
-nouveau_fence_wait_uevent(struct nouveau_fence *fence, bool intr)
-
+static long
+nouveau_fence_wait_legacy(struct fence *f, bool intr, long wait)
 {
-	struct nouveau_channel *chan = fence->channel;
-	struct nouveau_fence_priv *priv = chan->drm->fence;
-	struct nouveau_fence_wait wait = { .priv = priv };
-	int ret = 0;
+	struct nouveau_fence *fence = from_fence(f);
+	unsigned long sleep_time = NSEC_PER_MSEC / 1000;
+	unsigned long t = jiffies, timeout = t + wait;
 
-	ret = nvif_notify_init(chan->object, NULL,
-			       nouveau_fence_wait_uevent_handler, false,
-			       G82_CHANNEL_DMA_V0_NTFY_UEVENT,
-			       &(struct nvif_notify_uevent_req) {
-			       },
-			       sizeof(struct nvif_notify_uevent_req),
-			       sizeof(struct nvif_notify_uevent_rep),
-			       &wait.notify);
-	if (ret)
-		return ret;
+	while (!nouveau_fence_done(fence)) {
+		ktime_t kt;
 
-	nvif_notify_get(&wait.notify);
-
-	if (fence->timeout) {
-		unsigned long timeout = fence->timeout - jiffies;
-
-		if (time_before(jiffies, fence->timeout)) {
-			if (intr) {
-				ret = wait_event_interruptible_timeout(
-						priv->waiting,
-						nouveau_fence_done(fence),
-						timeout);
-			} else {
-				ret = wait_event_timeout(priv->waiting,
-						nouveau_fence_done(fence),
-						timeout);
-			}
-		}
+		t = jiffies;
 
-		if (ret >= 0) {
-			fence->timeout = jiffies + ret;
-			if (time_after_eq(jiffies, fence->timeout))
-				ret = -EBUSY;
-		}
-	} else {
-		if (intr) {
-			ret = wait_event_interruptible(priv->waiting,
-					nouveau_fence_done(fence));
-		} else {
-			wait_event(priv->waiting, nouveau_fence_done(fence));
+		if (wait != MAX_SCHEDULE_TIMEOUT && time_after_eq(t, timeout)) {
+			__set_current_state(TASK_RUNNING);
+			return 0;
 		}
+
+		__set_current_state(intr ? TASK_INTERRUPTIBLE :
+					   TASK_UNINTERRUPTIBLE);
+
+		kt = ktime_set(0, sleep_time);
+		schedule_hrtimeout(&kt, HRTIMER_MODE_REL);
+		sleep_time *= 2;
+		if (sleep_time > NSEC_PER_MSEC)
+			sleep_time = NSEC_PER_MSEC;
+
+		if (intr && signal_pending(current))
+			return -ERESTARTSYS;
 	}
 
-	nvif_notify_fini(&wait.notify);
-	if (unlikely(ret < 0))
-		return ret;
+	__set_current_state(TASK_RUNNING);
 
-	return 0;
+	return timeout - t;
 }
 
-int
-nouveau_fence_wait(struct nouveau_fence *fence, bool lazy, bool intr)
+static int
+nouveau_fence_wait_busy(struct nouveau_fence *fence, bool intr)
 {
-	struct nouveau_channel *chan = fence->channel;
-	struct nouveau_fence_priv *priv = chan ? chan->drm->fence : NULL;
-	unsigned long sleep_time = NSEC_PER_MSEC / 1000;
-	ktime_t t;
 	int ret = 0;
 
-	while (priv && priv->uevent && lazy && !nouveau_fence_done(fence)) {
-		ret = nouveau_fence_wait_uevent(fence, intr);
-		if (ret < 0)
-			return ret;
-	}
-
 	while (!nouveau_fence_done(fence)) {
-		if (fence->timeout && time_after_eq(jiffies, fence->timeout)) {
+		if (time_after_eq(jiffies, fence->timeout)) {
 			ret = -EBUSY;
 			break;
 		}
 
-		__set_current_state(intr ? TASK_INTERRUPTIBLE :
-					   TASK_UNINTERRUPTIBLE);
-		if (lazy) {
-			t = ktime_set(0, sleep_time);
-			schedule_hrtimeout(&t, HRTIMER_MODE_REL);
-			sleep_time *= 2;
-			if (sleep_time > NSEC_PER_MSEC)
-				sleep_time = NSEC_PER_MSEC;
-		}
+		__set_current_state(intr ?
+				    TASK_INTERRUPTIBLE :
+				    TASK_UNINTERRUPTIBLE);
 
 		if (intr && signal_pending(current)) {
 			ret = -ERESTARTSYS;
@@ -281,47 +349,86 @@ nouveau_fence_wait(struct nouveau_fence *fence, bool lazy, bool intr)
 }
 
 int
-nouveau_fence_sync(struct nouveau_fence *fence, struct nouveau_channel *chan)
+nouveau_fence_wait(struct nouveau_fence *fence, bool lazy, bool intr)
 {
-	struct nouveau_fence_chan *fctx = chan->fence;
-	struct nouveau_channel *prev;
-	int ret = 0;
+	long ret;
 
-	prev = fence ? fence->channel : NULL;
-	if (prev) {
-		if (unlikely(prev != chan && !nouveau_fence_done(fence))) {
-			ret = fctx->sync(fence, prev, chan);
-			if (unlikely(ret))
-				ret = nouveau_fence_wait(fence, true, false);
-		}
-	}
+	if (!lazy)
+		return nouveau_fence_wait_busy(fence, intr);
 
-	return ret;
+	ret = fence_wait_timeout(&fence->base, intr, 15 * HZ);
+	if (ret < 0)
+		return ret;
+	else if (!ret)
+		return -EBUSY;
+	else
+		return 0;
 }
 
-static void
-nouveau_fence_del(struct kref *kref)
+int
+nouveau_fence_sync(struct nouveau_bo *nvbo, struct nouveau_channel *chan, bool exclusive, bool intr)
 {
-	struct nouveau_fence *fence = container_of(kref, typeof(*fence), kref);
-	kfree(fence);
+	struct nouveau_fence_chan *fctx = chan->fence;
+	struct fence *fence;
+	struct reservation_object *resv = nvbo->bo.resv;
+	struct reservation_object_list *fobj;
+	struct nouveau_fence *f;
+	int ret = 0, i;
+
+	if (!exclusive) {
+		ret = reservation_object_reserve_shared(resv);
+
+		if (ret)
+			return ret;
+	}
+
+	fobj = reservation_object_get_list(resv);
+	fence = reservation_object_get_excl(resv);
+
+	if (fence && (!exclusive || !fobj || !fobj->shared_count)) {
+		struct nouveau_channel *prev = NULL;
+
+		f = nouveau_local_fence(fence, chan->drm);
+		if (f)
+			prev = f->channel;
+
+		if (!prev || (prev != chan && (ret = fctx->sync(f, prev, chan))))
+			ret = fence_wait(fence, intr);
+
+		return ret;
+	}
+
+	if (!exclusive || !fobj)
+		return ret;
+
+	for (i = 0; i < fobj->shared_count && !ret; ++i) {
+		struct nouveau_channel *prev = NULL;
+
+		fence = rcu_dereference_protected(fobj->shared[i],
+						reservation_object_held(resv));
+
+		f = nouveau_local_fence(fence, chan->drm);
+		if (f)
+			prev = f->channel;
+
+		if (!prev || (prev != chan && (ret = fctx->sync(f, prev, chan))))
+			ret = fence_wait(fence, intr);
+
+		if (ret)
+			break;
+	}
+
+	return ret;
 }
 
 void
 nouveau_fence_unref(struct nouveau_fence **pfence)
 {
 	if (*pfence)
-		kref_put(&(*pfence)->kref, nouveau_fence_del);
+		fence_put(&(*pfence)->base);
 	*pfence = NULL;
 }
 
-struct nouveau_fence *
-nouveau_fence_ref(struct nouveau_fence *fence)
-{
-	if (fence)
-		kref_get(&fence->kref);
-	return fence;
-}
-
 int
 nouveau_fence_new(struct nouveau_channel *chan, bool sysmem,
 		  struct nouveau_fence **pfence)
@@ -336,9 +443,7 @@ nouveau_fence_new(struct nouveau_channel *chan, bool sysmem,
 	if (!fence)
 		return -ENOMEM;
 
-	INIT_LIST_HEAD(&fence->work);
 	fence->sysmem = sysmem;
-	kref_init(&fence->kref);
 
 	ret = nouveau_fence_emit(fence, chan);
 	if (ret)
@@ -347,3 +452,101 @@ nouveau_fence_new(struct nouveau_channel *chan, bool sysmem,
 	*pfence = fence;
 	return ret;
 }
+
+static const char *nouveau_fence_get_get_driver_name(struct fence *fence)
+{
+	return "nouveau";
+}
+
+static const char *nouveau_fence_get_timeline_name(struct fence *f)
+{
+	struct nouveau_fence *fence = from_fence(f);
+	struct nouveau_fence_chan *fctx = nouveau_fctx(fence);
+
+	return fence->channel ? fctx->name : "dead channel";
+}
+
+/*
+ * In an ideal world, read would not assume the channel context is still alive.
+ * This function may be called from another device, running into free memory as a
+ * result. The drm node should still be there, so we can derive the index from
+ * the fence context.
+ */
+static bool nouveau_fence_is_signaled(struct fence *f)
+{
+	struct nouveau_fence *fence = from_fence(f);
+	struct nouveau_fence_chan *fctx = nouveau_fctx(fence);
+	struct nouveau_channel *chan = fence->channel;
+
+	return (int)(fctx->read(chan) - fence->base.seqno) >= 0;
+}
+
+static bool nouveau_fence_no_signaling(struct fence *f)
+{
+	struct nouveau_fence *fence = from_fence(f);
+
+	/*
+	 * caller should have a reference on the fence,
+	 * else fence could get freed here
+	 */
+	WARN_ON(atomic_read(&fence->base.refcount.refcount) <= 1);
+
+	/*
+	 * This needs uevents to work correctly, but fence_add_callback relies on
+	 * being able to enable signaling. It will still get signaled eventually,
+	 * just not right away.
+	 */
+	if (nouveau_fence_is_signaled(f)) {
+		list_del(&fence->head);
+
+		fence_put(&fence->base);
+		return false;
+	}
+
+	return true;
+}
+
+static void nouveau_fence_release(struct fence *f)
+{
+	struct nouveau_fence *fence = from_fence(f);
+	struct nouveau_fence_chan *fctx = nouveau_fctx(fence);
+
+	kref_put(&fctx->fence_ref, nouveau_fence_context_put);
+	fence_free(&fence->base);
+}
+
+static const struct fence_ops nouveau_fence_ops_legacy = {
+	.get_driver_name = nouveau_fence_get_get_driver_name,
+	.get_timeline_name = nouveau_fence_get_timeline_name,
+	.enable_signaling = nouveau_fence_no_signaling,
+	.signaled = nouveau_fence_is_signaled,
+	.wait = nouveau_fence_wait_legacy,
+	.release = nouveau_fence_release
+};
+
+static bool nouveau_fence_enable_signaling(struct fence *f)
+{
+	struct nouveau_fence *fence = from_fence(f);
+	struct nouveau_fence_chan *fctx = nouveau_fctx(fence);
+	bool ret;
+
+	if (!fctx->notify_ref++)
+		nvif_notify_get(&fctx->notify);
+
+	ret = nouveau_fence_no_signaling(f);
+	if (ret)
+		set_bit(FENCE_FLAG_USER_BITS, &fence->base.flags);
+	else if (!--fctx->notify_ref)
+		nvif_notify_put(&fctx->notify);
+
+	return ret;
+}
+
+static const struct fence_ops nouveau_fence_ops_uevent = {
+	.get_driver_name = nouveau_fence_get_get_driver_name,
+	.get_timeline_name = nouveau_fence_get_timeline_name,
+	.enable_signaling = nouveau_fence_enable_signaling,
+	.signaled = nouveau_fence_is_signaled,
+	.wait = fence_default_wait,
+	.release = NULL
+};
diff --git a/drivers/gpu/drm/nouveau/nouveau_fence.h b/drivers/gpu/drm/nouveau/nouveau_fence.h
index c57bb61da58c..943b0b17b1fc 100644
--- a/drivers/gpu/drm/nouveau/nouveau_fence.h
+++ b/drivers/gpu/drm/nouveau/nouveau_fence.h
@@ -1,33 +1,37 @@
 #ifndef __NOUVEAU_FENCE_H__
 #define __NOUVEAU_FENCE_H__
 
+#include <linux/fence.h>
+#include <nvif/notify.h>
+
 struct nouveau_drm;
+struct nouveau_bo;
 
 struct nouveau_fence {
+	struct fence base;
+
 	struct list_head head;
-	struct list_head work;
-	struct kref kref;
 
 	bool sysmem;
 
 	struct nouveau_channel *channel;
 	unsigned long timeout;
-	u32 sequence;
 };
 
 int  nouveau_fence_new(struct nouveau_channel *, bool sysmem,
 		       struct nouveau_fence **);
-struct nouveau_fence *
-nouveau_fence_ref(struct nouveau_fence *);
 void nouveau_fence_unref(struct nouveau_fence **);
 
 int  nouveau_fence_emit(struct nouveau_fence *, struct nouveau_channel *);
 bool nouveau_fence_done(struct nouveau_fence *);
-void nouveau_fence_work(struct nouveau_fence *, void (*)(void *), void *);
+void nouveau_fence_work(struct fence *, void (*)(void *), void *);
 int  nouveau_fence_wait(struct nouveau_fence *, bool lazy, bool intr);
-int  nouveau_fence_sync(struct nouveau_fence *, struct nouveau_channel *);
+int  nouveau_fence_sync(struct nouveau_bo *, struct nouveau_channel *, bool exclusive, bool intr);
 
 struct nouveau_fence_chan {
+	spinlock_t lock;
+	struct kref fence_ref;
+
 	struct list_head pending;
 	struct list_head flip;
 
@@ -38,8 +42,12 @@ struct nouveau_fence_chan {
 	int  (*emit32)(struct nouveau_channel *, u64, u32);
 	int  (*sync32)(struct nouveau_channel *, u64, u32);
 
-	spinlock_t lock;
 	u32 sequence;
+	u32 context;
+	char name[32];
+
+	struct nvif_notify notify;
+	int notify_ref;
 };
 
 struct nouveau_fence_priv {
@@ -49,14 +57,15 @@ struct nouveau_fence_priv {
 	int  (*context_new)(struct nouveau_channel *);
 	void (*context_del)(struct nouveau_channel *);
 
-	wait_queue_head_t waiting;
+	u32 contexts, context_base;
 	bool uevent;
 };
 
 #define nouveau_fence(drm) ((struct nouveau_fence_priv *)(drm)->fence)
 
-void nouveau_fence_context_new(struct nouveau_fence_chan *);
+void nouveau_fence_context_new(struct nouveau_channel *, struct nouveau_fence_chan *);
 void nouveau_fence_context_del(struct nouveau_fence_chan *);
+void nouveau_fence_context_free(struct nouveau_fence_chan *);
 
 int nv04_fence_create(struct nouveau_drm *);
 int nv04_fence_mthd(struct nouveau_channel *, u32, u32, u32);
diff --git a/drivers/gpu/drm/nouveau/nouveau_gem.c b/drivers/gpu/drm/nouveau/nouveau_gem.c
index 292a677bfed4..36951ee4b157 100644
--- a/drivers/gpu/drm/nouveau/nouveau_gem.c
+++ b/drivers/gpu/drm/nouveau/nouveau_gem.c
@@ -98,17 +98,23 @@ static void
 nouveau_gem_object_unmap(struct nouveau_bo *nvbo, struct nouveau_vma *vma)
 {
 	const bool mapped = nvbo->bo.mem.mem_type != TTM_PL_SYSTEM;
-	struct nouveau_fence *fence = NULL;
+	struct reservation_object *resv = nvbo->bo.resv;
+	struct reservation_object_list *fobj;
+	struct fence *fence = NULL;
+
+	fobj = reservation_object_get_list(resv);
 
 	list_del(&vma->head);
 
-	if (mapped) {
-		spin_lock(&nvbo->bo.bdev->fence_lock);
-		fence = nouveau_fence_ref(nvbo->bo.sync_obj);
-		spin_unlock(&nvbo->bo.bdev->fence_lock);
-	}
+	if (fobj && fobj->shared_count > 1)
+		ttm_bo_wait(&nvbo->bo, true, false, false);
+	else if (fobj && fobj->shared_count == 1)
+		fence = rcu_dereference_protected(fobj->shared[0],
+						reservation_object_held(resv));
+	else
+		fence = reservation_object_get_excl(nvbo->bo.resv);
 
-	if (fence) {
+	if (fence && mapped) {
 		nouveau_fence_work(fence, nouveau_gem_object_delete, vma);
 	} else {
 		if (mapped)
@@ -116,7 +122,6 @@ nouveau_gem_object_unmap(struct nouveau_bo *nvbo, struct nouveau_vma *vma)
 		nouveau_vm_put(vma);
 		kfree(vma);
 	}
-	nouveau_fence_unref(&fence);
 }
 
 void
@@ -160,7 +165,7 @@ nouveau_gem_new(struct drm_device *dev, int size, int align, uint32_t domain,
 		flags |= TTM_PL_FLAG_SYSTEM;
 
 	ret = nouveau_bo_new(dev, size, align, flags, tile_mode,
-			     tile_flags, NULL, pnvbo);
+			     tile_flags, NULL, NULL, pnvbo);
 	if (ret)
 		return ret;
 	nvbo = *pnvbo;
@@ -288,24 +293,23 @@ nouveau_gem_set_domain(struct drm_gem_object *gem, uint32_t read_domains,
 }
 
 struct validate_op {
-	struct list_head vram_list;
-	struct list_head gart_list;
-	struct list_head both_list;
+	struct list_head list;
 	struct ww_acquire_ctx ticket;
 };
 
 static void
-validate_fini_list(struct list_head *list, struct nouveau_fence *fence,
-		   struct ww_acquire_ctx *ticket)
+validate_fini_no_ticket(struct validate_op *op, struct nouveau_fence *fence,
+			struct drm_nouveau_gem_pushbuf_bo *pbbo)
 {
-	struct list_head *entry, *tmp;
 	struct nouveau_bo *nvbo;
+	struct drm_nouveau_gem_pushbuf_bo *b;
 
-	list_for_each_safe(entry, tmp, list) {
-		nvbo = list_entry(entry, struct nouveau_bo, entry);
+	while (!list_empty(&op->list)) {
+		nvbo = list_entry(op->list.next, struct nouveau_bo, entry);
+		b = &pbbo[nvbo->pbbo_index];
 
 		if (likely(fence))
-			nouveau_bo_fence(nvbo, fence);
+			nouveau_bo_fence(nvbo, fence, !!b->write_domains);
 
 		if (unlikely(nvbo->validate_mapped)) {
 			ttm_bo_kunmap(&nvbo->kmap);
@@ -314,23 +318,16 @@ validate_fini_list(struct list_head *list, struct nouveau_fence *fence,
 
 		list_del(&nvbo->entry);
 		nvbo->reserved_by = NULL;
-		ttm_bo_unreserve_ticket(&nvbo->bo, ticket);
+		ttm_bo_unreserve_ticket(&nvbo->bo, &op->ticket);
 		drm_gem_object_unreference_unlocked(&nvbo->gem);
 	}
 }
 
 static void
-validate_fini_no_ticket(struct validate_op *op, struct nouveau_fence *fence)
+validate_fini(struct validate_op *op, struct nouveau_fence *fence,
+	      struct drm_nouveau_gem_pushbuf_bo *pbbo)
 {
-	validate_fini_list(&op->vram_list, fence, &op->ticket);
-	validate_fini_list(&op->gart_list, fence, &op->ticket);
-	validate_fini_list(&op->both_list, fence, &op->ticket);
-}
-
-static void
-validate_fini(struct validate_op *op, struct nouveau_fence *fence)
-{
-	validate_fini_no_ticket(op, fence);
+	validate_fini_no_ticket(op, fence, pbbo);
 	ww_acquire_fini(&op->ticket);
 }
 
@@ -344,6 +341,9 @@ validate_init(struct nouveau_channel *chan, struct drm_file *file_priv,
 	int trycnt = 0;
 	int ret, i;
 	struct nouveau_bo *res_bo = NULL;
+	LIST_HEAD(gart_list);
+	LIST_HEAD(vram_list);
+	LIST_HEAD(both_list);
 
 	ww_acquire_init(&op->ticket, &reservation_ww_class);
 retry:
@@ -360,9 +360,8 @@ retry:
 		gem = drm_gem_object_lookup(dev, file_priv, b->handle);
 		if (!gem) {
 			NV_PRINTK(error, cli, "Unknown handle 0x%08x\n", b->handle);
-			ww_acquire_done(&op->ticket);
-			validate_fini(op, NULL);
-			return -ENOENT;
+			ret = -ENOENT;
+			break;
 		}
 		nvbo = nouveau_gem_object(gem);
 		if (nvbo == res_bo) {
@@ -375,14 +374,16 @@ retry:
 			NV_PRINTK(error, cli, "multiple instances of buffer %d on "
 				      "validation list\n", b->handle);
 			drm_gem_object_unreference_unlocked(gem);
-			ww_acquire_done(&op->ticket);
-			validate_fini(op, NULL);
-			return -EINVAL;
+			ret = -EINVAL;
+			break;
 		}
 
 		ret = ttm_bo_reserve(&nvbo->bo, true, false, true, &op->ticket);
 		if (ret) {
-			validate_fini_no_ticket(op, NULL);
+			list_splice_tail_init(&vram_list, &op->list);
+			list_splice_tail_init(&gart_list, &op->list);
+			list_splice_tail_init(&both_list, &op->list);
+			validate_fini_no_ticket(op, NULL, NULL);
 			if (unlikely(ret == -EDEADLK)) {
 				ret = ttm_bo_reserve_slowpath(&nvbo->bo, true,
 							      &op->ticket);
@@ -390,12 +391,9 @@ retry:
 					res_bo = nvbo;
 			}
 			if (unlikely(ret)) {
-				ww_acquire_done(&op->ticket);
-				ww_acquire_fini(&op->ticket);
-				drm_gem_object_unreference_unlocked(gem);
 				if (ret != -ERESTARTSYS)
 					NV_PRINTK(error, cli, "fail reserve\n");
-				return ret;
+				break;
 			}
 		}
 
@@ -404,45 +402,32 @@ retry:
 		nvbo->pbbo_index = i;
 		if ((b->valid_domains & NOUVEAU_GEM_DOMAIN_VRAM) &&
 		    (b->valid_domains & NOUVEAU_GEM_DOMAIN_GART))
-			list_add_tail(&nvbo->entry, &op->both_list);
+			list_add_tail(&nvbo->entry, &both_list);
 		else
 		if (b->valid_domains & NOUVEAU_GEM_DOMAIN_VRAM)
-			list_add_tail(&nvbo->entry, &op->vram_list);
+			list_add_tail(&nvbo->entry, &vram_list);
 		else
 		if (b->valid_domains & NOUVEAU_GEM_DOMAIN_GART)
-			list_add_tail(&nvbo->entry, &op->gart_list);
+			list_add_tail(&nvbo->entry, &gart_list);
 		else {
 			NV_PRINTK(error, cli, "invalid valid domains: 0x%08x\n",
 				 b->valid_domains);
-			list_add_tail(&nvbo->entry, &op->both_list);
-			ww_acquire_done(&op->ticket);
-			validate_fini(op, NULL);
-			return -EINVAL;
+			list_add_tail(&nvbo->entry, &both_list);
+			ret = -EINVAL;
+			break;
 		}
 		if (nvbo == res_bo)
 			goto retry;
 	}
 
 	ww_acquire_done(&op->ticket);
-	return 0;
-}
-
-static int
-validate_sync(struct nouveau_channel *chan, struct nouveau_bo *nvbo)
-{
-	struct nouveau_fence *fence = NULL;
-	int ret = 0;
-
-	spin_lock(&nvbo->bo.bdev->fence_lock);
-	fence = nouveau_fence_ref(nvbo->bo.sync_obj);
-	spin_unlock(&nvbo->bo.bdev->fence_lock);
-
-	if (fence) {
-		ret = nouveau_fence_sync(fence, chan);
-		nouveau_fence_unref(&fence);
-	}
-
+	list_splice_tail(&vram_list, &op->list);
+	list_splice_tail(&gart_list, &op->list);
+	list_splice_tail(&both_list, &op->list);
+	if (ret)
+		validate_fini(op, NULL, NULL);
 	return ret;
+
 }
 
 static int
@@ -474,9 +459,10 @@ validate_list(struct nouveau_channel *chan, struct nouveau_cli *cli,
 			return ret;
 		}
 
-		ret = validate_sync(chan, nvbo);
+		ret = nouveau_fence_sync(nvbo, chan, !!b->write_domains, true);
 		if (unlikely(ret)) {
-			NV_PRINTK(error, cli, "fail post-validate sync\n");
+			if (ret != -ERESTARTSYS)
+				NV_PRINTK(error, cli, "fail post-validate sync\n");
 			return ret;
 		}
 
@@ -513,11 +499,9 @@ nouveau_gem_pushbuf_validate(struct nouveau_channel *chan,
 			     struct validate_op *op, int *apply_relocs)
 {
 	struct nouveau_cli *cli = nouveau_cli(file_priv);
-	int ret, relocs = 0;
+	int ret;
 
-	INIT_LIST_HEAD(&op->vram_list);
-	INIT_LIST_HEAD(&op->gart_list);
-	INIT_LIST_HEAD(&op->both_list);
+	INIT_LIST_HEAD(&op->list);
 
 	if (nr_buffers == 0)
 		return 0;
@@ -529,34 +513,14 @@ nouveau_gem_pushbuf_validate(struct nouveau_channel *chan,
 		return ret;
 	}
 
-	ret = validate_list(chan, cli, &op->vram_list, pbbo, user_buffers);
-	if (unlikely(ret < 0)) {
-		if (ret != -ERESTARTSYS)
-			NV_PRINTK(error, cli, "validate vram_list\n");
-		validate_fini(op, NULL);
-		return ret;
-	}
-	relocs += ret;
-
-	ret = validate_list(chan, cli, &op->gart_list, pbbo, user_buffers);
-	if (unlikely(ret < 0)) {
-		if (ret != -ERESTARTSYS)
-			NV_PRINTK(error, cli, "validate gart_list\n");
-		validate_fini(op, NULL);
-		return ret;
-	}
-	relocs += ret;
-
-	ret = validate_list(chan, cli, &op->both_list, pbbo, user_buffers);
+	ret = validate_list(chan, cli, &op->list, pbbo, user_buffers);
 	if (unlikely(ret < 0)) {
 		if (ret != -ERESTARTSYS)
-			NV_PRINTK(error, cli, "validate both_list\n");
-		validate_fini(op, NULL);
+			NV_PRINTK(error, cli, "validating bo list\n");
+		validate_fini(op, NULL, NULL);
 		return ret;
 	}
-	relocs += ret;
-
-	*apply_relocs = relocs;
+	*apply_relocs = ret;
 	return 0;
 }
 
@@ -659,9 +623,7 @@ nouveau_gem_pushbuf_reloc_apply(struct nouveau_cli *cli,
 				data |= r->vor;
 		}
 
-		spin_lock(&nvbo->bo.bdev->fence_lock);
-		ret = ttm_bo_wait(&nvbo->bo, false, false, false);
-		spin_unlock(&nvbo->bo.bdev->fence_lock);
+		ret = ttm_bo_wait(&nvbo->bo, true, false, false);
 		if (ret) {
 			NV_PRINTK(error, cli, "reloc wait_idle failed: %d\n", ret);
 			break;
@@ -839,7 +801,7 @@ nouveau_gem_ioctl_pushbuf(struct drm_device *dev, void *data,
 	}
 
 out:
-	validate_fini(&op, fence);
+	validate_fini(&op, fence, bo);
 	nouveau_fence_unref(&fence);
 
 out_prevalid:
@@ -884,17 +846,29 @@ nouveau_gem_ioctl_cpu_prep(struct drm_device *dev, void *data,
 	struct drm_gem_object *gem;
 	struct nouveau_bo *nvbo;
 	bool no_wait = !!(req->flags & NOUVEAU_GEM_CPU_PREP_NOWAIT);
-	int ret = -EINVAL;
+	bool write = !!(req->flags & NOUVEAU_GEM_CPU_PREP_WRITE);
+	int ret;
 
 	gem = drm_gem_object_lookup(dev, file_priv, req->handle);
 	if (!gem)
 		return -ENOENT;
 	nvbo = nouveau_gem_object(gem);
 
-	spin_lock(&nvbo->bo.bdev->fence_lock);
-	ret = ttm_bo_wait(&nvbo->bo, true, true, no_wait);
-	spin_unlock(&nvbo->bo.bdev->fence_lock);
+	if (no_wait)
+		ret = reservation_object_test_signaled_rcu(nvbo->bo.resv, write) ? 0 : -EBUSY;
+	else {
+		long lret;
+
+		lret = reservation_object_wait_timeout_rcu(nvbo->bo.resv, write, true, 30 * HZ);
+		if (!lret)
+			ret = -EBUSY;
+		else if (lret > 0)
+			ret = 0;
+		else
+			ret = lret;
+	}
 	drm_gem_object_unreference_unlocked(gem);
+
 	return ret;
 }
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_gem.h b/drivers/gpu/drm/nouveau/nouveau_gem.h
index ddab762d81fe..e4049faca780 100644
--- a/drivers/gpu/drm/nouveau/nouveau_gem.h
+++ b/drivers/gpu/drm/nouveau/nouveau_gem.h
@@ -39,7 +39,7 @@ struct reservation_object *nouveau_gem_prime_res_obj(struct drm_gem_object *);
 extern void nouveau_gem_prime_unpin(struct drm_gem_object *);
 extern struct sg_table *nouveau_gem_prime_get_sg_table(struct drm_gem_object *);
 extern struct drm_gem_object *nouveau_gem_prime_import_sg_table(
-	struct drm_device *, size_t size, struct sg_table *);
+	struct drm_device *, struct dma_buf_attachment *, struct sg_table *);
 extern void *nouveau_gem_prime_vmap(struct drm_gem_object *);
 extern void nouveau_gem_prime_vunmap(struct drm_gem_object *, void *);
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_prime.c b/drivers/gpu/drm/nouveau/nouveau_prime.c
index 1f51008e4d26..228226ab27fc 100644
--- a/drivers/gpu/drm/nouveau/nouveau_prime.c
+++ b/drivers/gpu/drm/nouveau/nouveau_prime.c
@@ -23,6 +23,7 @@
  */
 
 #include <drm/drmP.h>
+#include <linux/dma-buf.h>
 
 #include "nouveau_drm.h"
 #include "nouveau_gem.h"
@@ -56,17 +57,20 @@ void nouveau_gem_prime_vunmap(struct drm_gem_object *obj, void *vaddr)
 }
 
 struct drm_gem_object *nouveau_gem_prime_import_sg_table(struct drm_device *dev,
-							 size_t size,
+							 struct dma_buf_attachment *attach,
 							 struct sg_table *sg)
 {
 	struct nouveau_bo *nvbo;
+	struct reservation_object *robj = attach->dmabuf->resv;
 	u32 flags = 0;
 	int ret;
 
 	flags = TTM_PL_FLAG_TT;
 
-	ret = nouveau_bo_new(dev, size, 0, flags, 0, 0,
-			     sg, &nvbo);
+	ww_mutex_lock(&robj->lock, NULL);
+	ret = nouveau_bo_new(dev, attach->dmabuf->size, 0, flags, 0, 0,
+			     sg, robj, &nvbo);
+	ww_mutex_unlock(&robj->lock);
 	if (ret)
 		return ERR_PTR(ret);
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_sysfs.c b/drivers/gpu/drm/nouveau/nouveau_sysfs.c
index 3c6962d15b26..8fbbf3093d86 100644
--- a/drivers/gpu/drm/nouveau/nouveau_sysfs.c
+++ b/drivers/gpu/drm/nouveau/nouveau_sysfs.c
@@ -29,7 +29,7 @@
 #include "nouveau_sysfs.h"
 
 MODULE_PARM_DESC(pstate, "enable sysfs pstate file, which will be moved in the future");
-static int nouveau_pstate;
+int nouveau_pstate;
 module_param_named(pstate, nouveau_pstate, int, 0400);
 
 static inline struct drm_device *
diff --git a/drivers/gpu/drm/nouveau/nouveau_sysfs.h b/drivers/gpu/drm/nouveau/nouveau_sysfs.h
index f973378160f8..4e5ea9241b28 100644
--- a/drivers/gpu/drm/nouveau/nouveau_sysfs.h
+++ b/drivers/gpu/drm/nouveau/nouveau_sysfs.h
@@ -16,4 +16,6 @@ nouveau_sysfs(struct drm_device *dev)
 int  nouveau_sysfs_init(struct drm_device *);
 void nouveau_sysfs_fini(struct drm_device *);
 
+extern int nouveau_pstate;
+
 #endif
diff --git a/drivers/gpu/drm/nouveau/nouveau_ttm.c b/drivers/gpu/drm/nouveau/nouveau_ttm.c
index 53874b76b031..753a6def61e7 100644
--- a/drivers/gpu/drm/nouveau/nouveau_ttm.c
+++ b/drivers/gpu/drm/nouveau/nouveau_ttm.c
@@ -71,8 +71,7 @@ nouveau_vram_manager_del(struct ttm_mem_type_manager *man,
 static int
 nouveau_vram_manager_new(struct ttm_mem_type_manager *man,
 			 struct ttm_buffer_object *bo,
-			 struct ttm_placement *placement,
-			 uint32_t flags,
+			 const struct ttm_place *place,
 			 struct ttm_mem_reg *mem)
 {
 	struct nouveau_drm *drm = nouveau_bdev(man->bdev);
@@ -158,8 +157,7 @@ nouveau_gart_manager_del(struct ttm_mem_type_manager *man,
 static int
 nouveau_gart_manager_new(struct ttm_mem_type_manager *man,
 			 struct ttm_buffer_object *bo,
-			 struct ttm_placement *placement,
-			 uint32_t flags,
+			 const struct ttm_place *place,
 			 struct ttm_mem_reg *mem)
 {
 	struct nouveau_drm *drm = nouveau_bdev(bo->bdev);
@@ -239,8 +237,7 @@ nv04_gart_manager_del(struct ttm_mem_type_manager *man, struct ttm_mem_reg *mem)
 static int
 nv04_gart_manager_new(struct ttm_mem_type_manager *man,
 		      struct ttm_buffer_object *bo,
-		      struct ttm_placement *placement,
-		      uint32_t flags,
+		      const struct ttm_place *place,
 		      struct ttm_mem_reg *mem)
 {
 	struct nouveau_mem *node;
@@ -284,7 +281,7 @@ nouveau_ttm_mmap(struct file *filp, struct vm_area_struct *vma)
 	struct nouveau_drm *drm = nouveau_drm(file_priv->minor->dev);
 
 	if (unlikely(vma->vm_pgoff < DRM_FILE_PAGE_OFFSET))
-		return drm_mmap(filp, vma);
+		return -EINVAL;
 
 	return ttm_bo_mmap(filp, vma, &drm->ttm.bdev);
 }
diff --git a/drivers/gpu/drm/nouveau/nv04_fence.c b/drivers/gpu/drm/nouveau/nv04_fence.c
index 239c2c5a9615..f9859deb108a 100644
--- a/drivers/gpu/drm/nouveau/nv04_fence.c
+++ b/drivers/gpu/drm/nouveau/nv04_fence.c
@@ -41,7 +41,7 @@ nv04_fence_emit(struct nouveau_fence *fence)
 	int ret = RING_SPACE(chan, 2);
 	if (ret == 0) {
 		BEGIN_NV04(chan, NvSubSw, 0x0150, 1);
-		OUT_RING  (chan, fence->sequence);
+		OUT_RING  (chan, fence->base.seqno);
 		FIRE_RING (chan);
 	}
 	return ret;
@@ -67,7 +67,7 @@ nv04_fence_context_del(struct nouveau_channel *chan)
 	struct nv04_fence_chan *fctx = chan->fence;
 	nouveau_fence_context_del(&fctx->base);
 	chan->fence = NULL;
-	kfree(fctx);
+	nouveau_fence_context_free(&fctx->base);
 }
 
 static int
@@ -75,7 +75,7 @@ nv04_fence_context_new(struct nouveau_channel *chan)
 {
 	struct nv04_fence_chan *fctx = kzalloc(sizeof(*fctx), GFP_KERNEL);
 	if (fctx) {
-		nouveau_fence_context_new(&fctx->base);
+		nouveau_fence_context_new(chan, &fctx->base);
 		fctx->base.emit = nv04_fence_emit;
 		fctx->base.sync = nv04_fence_sync;
 		fctx->base.read = nv04_fence_read;
@@ -105,5 +105,7 @@ nv04_fence_create(struct nouveau_drm *drm)
 	priv->base.dtor = nv04_fence_destroy;
 	priv->base.context_new = nv04_fence_context_new;
 	priv->base.context_del = nv04_fence_context_del;
+	priv->base.contexts = 15;
+	priv->base.context_base = fence_context_alloc(priv->base.contexts);
 	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/nv10_fence.c b/drivers/gpu/drm/nouveau/nv10_fence.c
index 4faaf0acf5d7..5e1ea1cdce75 100644
--- a/drivers/gpu/drm/nouveau/nv10_fence.c
+++ b/drivers/gpu/drm/nouveau/nv10_fence.c
@@ -33,7 +33,7 @@ nv10_fence_emit(struct nouveau_fence *fence)
 	int ret = RING_SPACE(chan, 2);
 	if (ret == 0) {
 		BEGIN_NV04(chan, 0, NV10_SUBCHAN_REF_CNT, 1);
-		OUT_RING  (chan, fence->sequence);
+		OUT_RING  (chan, fence->base.seqno);
 		FIRE_RING (chan);
 	}
 	return ret;
@@ -63,7 +63,7 @@ nv10_fence_context_del(struct nouveau_channel *chan)
 		nvif_object_fini(&fctx->head[i]);
 	nvif_object_fini(&fctx->sema);
 	chan->fence = NULL;
-	kfree(fctx);
+	nouveau_fence_context_free(&fctx->base);
 }
 
 int
@@ -75,7 +75,7 @@ nv10_fence_context_new(struct nouveau_channel *chan)
 	if (!fctx)
 		return -ENOMEM;
 
-	nouveau_fence_context_new(&fctx->base);
+	nouveau_fence_context_new(chan, &fctx->base);
 	fctx->base.emit = nv10_fence_emit;
 	fctx->base.read = nv10_fence_read;
 	fctx->base.sync = nv10_fence_sync;
@@ -106,6 +106,8 @@ nv10_fence_create(struct nouveau_drm *drm)
 	priv->base.dtor = nv10_fence_destroy;
 	priv->base.context_new = nv10_fence_context_new;
 	priv->base.context_del = nv10_fence_context_del;
+	priv->base.contexts = 31;
+	priv->base.context_base = fence_context_alloc(priv->base.contexts);
 	spin_lock_init(&priv->lock);
 	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/nv17_fence.c b/drivers/gpu/drm/nouveau/nv17_fence.c
index ca907479f92f..40b461c7d5c5 100644
--- a/drivers/gpu/drm/nouveau/nv17_fence.c
+++ b/drivers/gpu/drm/nouveau/nv17_fence.c
@@ -84,7 +84,7 @@ nv17_fence_context_new(struct nouveau_channel *chan)
 	if (!fctx)
 		return -ENOMEM;
 
-	nouveau_fence_context_new(&fctx->base);
+	nouveau_fence_context_new(chan, &fctx->base);
 	fctx->base.emit = nv10_fence_emit;
 	fctx->base.read = nv10_fence_read;
 	fctx->base.sync = nv17_fence_sync;
@@ -124,10 +124,12 @@ nv17_fence_create(struct nouveau_drm *drm)
 	priv->base.resume = nv17_fence_resume;
 	priv->base.context_new = nv17_fence_context_new;
 	priv->base.context_del = nv10_fence_context_del;
+	priv->base.contexts = 31;
+	priv->base.context_base = fence_context_alloc(priv->base.contexts);
 	spin_lock_init(&priv->lock);
 
 	ret = nouveau_bo_new(drm->dev, 4096, 0x1000, TTM_PL_FLAG_VRAM,
-			     0, 0x0000, NULL, &priv->bo);
+			     0, 0x0000, NULL, NULL, &priv->bo);
 	if (!ret) {
 		ret = nouveau_bo_pin(priv->bo, TTM_PL_FLAG_VRAM);
 		if (!ret) {
diff --git a/drivers/gpu/drm/nouveau/nv50_display.c b/drivers/gpu/drm/nouveau/nv50_display.c
index 03949eaa629f..ae873d1a8d46 100644
--- a/drivers/gpu/drm/nouveau/nv50_display.c
+++ b/drivers/gpu/drm/nouveau/nv50_display.c
@@ -1066,7 +1066,7 @@ nv50_crtc_mode_set(struct drm_crtc *crtc, struct drm_display_mode *umode,
 	u32 vscan = (mode->flags & DRM_MODE_FLAG_DBLSCAN) ? 2 : 1;
 	u32 hactive, hsynce, hbackp, hfrontp, hblanke, hblanks;
 	u32 vactive, vsynce, vbackp, vfrontp, vblanke, vblanks;
-	u32 vblan2e = 0, vblan2s = 1;
+	u32 vblan2e = 0, vblan2s = 1, vblankus = 0;
 	u32 *push;
 	int ret;
 
@@ -1083,6 +1083,11 @@ nv50_crtc_mode_set(struct drm_crtc *crtc, struct drm_display_mode *umode,
 	vblanke = vsynce + vbackp;
 	vfrontp = (mode->vsync_start - mode->vdisplay) * vscan / ilace;
 	vblanks = vactive - vfrontp - 1;
+	/* XXX: Safe underestimate, even "0" works */
+	vblankus = (vactive - mode->vdisplay - 2) * hactive;
+	vblankus *= 1000;
+	vblankus /= mode->clock;
+
 	if (mode->flags & DRM_MODE_FLAG_INTERLACE) {
 		vblan2e = vactive + vsynce + vbackp;
 		vblan2s = vblan2e + (mode->vdisplay * vscan / ilace);
@@ -1099,14 +1104,14 @@ nv50_crtc_mode_set(struct drm_crtc *crtc, struct drm_display_mode *umode,
 			evo_mthd(push, 0x0804 + (nv_crtc->index * 0x400), 2);
 			evo_data(push, 0x00800000 | mode->clock);
 			evo_data(push, (ilace == 2) ? 2 : 0);
-			evo_mthd(push, 0x0810 + (nv_crtc->index * 0x400), 6);
+			evo_mthd(push, 0x0810 + (nv_crtc->index * 0x400), 8);
 			evo_data(push, 0x00000000);
 			evo_data(push, (vactive << 16) | hactive);
 			evo_data(push, ( vsynce << 16) | hsynce);
 			evo_data(push, (vblanke << 16) | hblanke);
 			evo_data(push, (vblanks << 16) | hblanks);
 			evo_data(push, (vblan2e << 16) | vblan2s);
-			evo_mthd(push, 0x082c + (nv_crtc->index * 0x400), 1);
+			evo_data(push, vblankus);
 			evo_data(push, 0x00000000);
 			evo_mthd(push, 0x0900 + (nv_crtc->index * 0x400), 2);
 			evo_data(push, 0x00000311);
@@ -1378,7 +1383,7 @@ nv50_crtc_create(struct drm_device *dev, int index)
 	drm_mode_crtc_set_gamma_size(crtc, 256);
 
 	ret = nouveau_bo_new(dev, 8192, 0x100, TTM_PL_FLAG_VRAM,
-			     0, 0x0000, NULL, &head->base.lut.nvbo);
+			     0, 0x0000, NULL, NULL, &head->base.lut.nvbo);
 	if (!ret) {
 		ret = nouveau_bo_pin(head->base.lut.nvbo, TTM_PL_FLAG_VRAM);
 		if (!ret) {
@@ -1401,7 +1406,7 @@ nv50_crtc_create(struct drm_device *dev, int index)
 		goto out;
 
 	ret = nouveau_bo_new(dev, 64 * 64 * 4, 0x100, TTM_PL_FLAG_VRAM,
-			     0, 0x0000, NULL, &head->base.cursor.nvbo);
+			     0, 0x0000, NULL, NULL, &head->base.cursor.nvbo);
 	if (!ret) {
 		ret = nouveau_bo_pin(head->base.cursor.nvbo, TTM_PL_FLAG_VRAM);
 		if (!ret) {
@@ -1651,17 +1656,21 @@ static void
 nv50_audio_mode_set(struct drm_encoder *encoder, struct drm_display_mode *mode)
 {
 	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	struct nouveau_crtc *nv_crtc = nouveau_crtc(encoder->crtc);
 	struct nouveau_connector *nv_connector;
 	struct nv50_disp *disp = nv50_disp(encoder->dev);
-	struct {
-		struct nv50_disp_mthd_v1 base;
-		struct nv50_disp_sor_hda_eld_v0 eld;
+	struct __packed {
+		struct {
+			struct nv50_disp_mthd_v1 mthd;
+			struct nv50_disp_sor_hda_eld_v0 eld;
+		} base;
 		u8 data[sizeof(nv_connector->base.eld)];
 	} args = {
-		.base.version = 1,
-		.base.method  = NV50_DISP_MTHD_V1_SOR_HDA_ELD,
-		.base.hasht   = nv_encoder->dcb->hasht,
-		.base.hashm   = nv_encoder->dcb->hashm,
+		.base.mthd.version = 1,
+		.base.mthd.method  = NV50_DISP_MTHD_V1_SOR_HDA_ELD,
+		.base.mthd.hasht   = nv_encoder->dcb->hasht,
+		.base.mthd.hashm   = (0xf0ff & nv_encoder->dcb->hashm) |
+				     (0x0100 << nv_crtc->index),
 	};
 
 	nv_connector = nouveau_encoder_connector_get(nv_encoder);
@@ -1671,11 +1680,11 @@ nv50_audio_mode_set(struct drm_encoder *encoder, struct drm_display_mode *mode)
 	drm_edid_to_eld(&nv_connector->base, nv_connector->edid);
 	memcpy(args.data, nv_connector->base.eld, sizeof(args.data));
 
-	nvif_mthd(disp->disp, 0, &args, sizeof(args));
+	nvif_mthd(disp->disp, 0, &args, sizeof(args.base) + args.data[2] * 4);
 }
 
 static void
-nv50_audio_disconnect(struct drm_encoder *encoder)
+nv50_audio_disconnect(struct drm_encoder *encoder, struct nouveau_crtc *nv_crtc)
 {
 	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
 	struct nv50_disp *disp = nv50_disp(encoder->dev);
@@ -1686,7 +1695,8 @@ nv50_audio_disconnect(struct drm_encoder *encoder)
 		.base.version = 1,
 		.base.method  = NV50_DISP_MTHD_V1_SOR_HDA_ELD,
 		.base.hasht   = nv_encoder->dcb->hasht,
-		.base.hashm   = nv_encoder->dcb->hashm,
+		.base.hashm   = (0xf0ff & nv_encoder->dcb->hashm) |
+				(0x0100 << nv_crtc->index),
 	};
 
 	nvif_mthd(disp->disp, 0, &args, sizeof(args));
@@ -1745,8 +1755,6 @@ nv50_hdmi_disconnect(struct drm_encoder *encoder, struct nouveau_crtc *nv_crtc)
 			       (0x0100 << nv_crtc->index),
 	};
 
-	nv50_audio_disconnect(encoder);
-
 	nvif_mthd(disp->disp, 0, &args, sizeof(args));
 }
 
@@ -1855,6 +1863,7 @@ nv50_sor_disconnect(struct drm_encoder *encoder)
 	if (nv_crtc) {
 		nv50_crtc_prepare(&nv_crtc->base);
 		nv50_sor_ctrl(nv_encoder, 1 << nv_crtc->index, 0);
+		nv50_audio_disconnect(encoder, nv_crtc);
 		nv50_hdmi_disconnect(&nv_encoder->base.base, nv_crtc);
 	}
 }
@@ -1954,6 +1963,7 @@ nv50_sor_mode_set(struct drm_encoder *encoder, struct drm_display_mode *umode,
 			proto = 0x8;
 		else
 			proto = 0x9;
+		nv50_audio_mode_set(encoder, mode);
 		break;
 	default:
 		BUG_ON(1);
@@ -2458,7 +2468,7 @@ nv50_display_create(struct drm_device *dev)
 
 	/* small shared memory area we use for notifiers and semaphores */
 	ret = nouveau_bo_new(dev, 4096, 0x1000, TTM_PL_FLAG_VRAM,
-			     0, 0x0000, NULL, &disp->sync);
+			     0, 0x0000, NULL, NULL, &disp->sync);
 	if (!ret) {
 		ret = nouveau_bo_pin(disp->sync, TTM_PL_FLAG_VRAM);
 		if (!ret) {
diff --git a/drivers/gpu/drm/nouveau/nv50_fence.c b/drivers/gpu/drm/nouveau/nv50_fence.c
index 195cf51a7c31..22d242b37962 100644
--- a/drivers/gpu/drm/nouveau/nv50_fence.c
+++ b/drivers/gpu/drm/nouveau/nv50_fence.c
@@ -46,7 +46,7 @@ nv50_fence_context_new(struct nouveau_channel *chan)
 	if (!fctx)
 		return -ENOMEM;
 
-	nouveau_fence_context_new(&fctx->base);
+	nouveau_fence_context_new(chan, &fctx->base);
 	fctx->base.emit = nv10_fence_emit;
 	fctx->base.read = nv10_fence_read;
 	fctx->base.sync = nv17_fence_sync;
@@ -95,10 +95,12 @@ nv50_fence_create(struct nouveau_drm *drm)
 	priv->base.resume = nv17_fence_resume;
 	priv->base.context_new = nv50_fence_context_new;
 	priv->base.context_del = nv10_fence_context_del;
+	priv->base.contexts = 127;
+	priv->base.context_base = fence_context_alloc(priv->base.contexts);
 	spin_lock_init(&priv->lock);
 
 	ret = nouveau_bo_new(drm->dev, 4096, 0x1000, TTM_PL_FLAG_VRAM,
-			     0, 0x0000, NULL, &priv->bo);
+			     0, 0x0000, NULL, NULL, &priv->bo);
 	if (!ret) {
 		ret = nouveau_bo_pin(priv->bo, TTM_PL_FLAG_VRAM);
 		if (!ret) {
diff --git a/drivers/gpu/drm/nouveau/nv84_fence.c b/drivers/gpu/drm/nouveau/nv84_fence.c
index 933a779c93ab..d6c6c87c3f07 100644
--- a/drivers/gpu/drm/nouveau/nv84_fence.c
+++ b/drivers/gpu/drm/nouveau/nv84_fence.c
@@ -82,7 +82,7 @@ nv84_fence_emit(struct nouveau_fence *fence)
 	else
 		addr += fctx->vma.offset;
 
-	return fctx->base.emit32(chan, addr, fence->sequence);
+	return fctx->base.emit32(chan, addr, fence->base.seqno);
 }
 
 static int
@@ -97,7 +97,7 @@ nv84_fence_sync(struct nouveau_fence *fence,
 	else
 		addr += fctx->vma.offset;
 
-	return fctx->base.sync32(chan, addr, fence->sequence);
+	return fctx->base.sync32(chan, addr, fence->base.seqno);
 }
 
 static u32
@@ -120,11 +120,12 @@ nv84_fence_context_del(struct nouveau_channel *chan)
 		nouveau_bo_vma_del(bo, &fctx->dispc_vma[i]);
 	}
 
+	nouveau_bo_wr32(priv->bo, chan->chid * 16 / 4, fctx->base.sequence);
 	nouveau_bo_vma_del(priv->bo, &fctx->vma_gart);
 	nouveau_bo_vma_del(priv->bo, &fctx->vma);
 	nouveau_fence_context_del(&fctx->base);
 	chan->fence = NULL;
-	kfree(fctx);
+	nouveau_fence_context_free(&fctx->base);
 }
 
 int
@@ -139,12 +140,13 @@ nv84_fence_context_new(struct nouveau_channel *chan)
 	if (!fctx)
 		return -ENOMEM;
 
-	nouveau_fence_context_new(&fctx->base);
+	nouveau_fence_context_new(chan, &fctx->base);
 	fctx->base.emit = nv84_fence_emit;
 	fctx->base.sync = nv84_fence_sync;
 	fctx->base.read = nv84_fence_read;
 	fctx->base.emit32 = nv84_fence_emit32;
 	fctx->base.sync32 = nv84_fence_sync32;
+	fctx->base.sequence = nv84_fence_read(chan);
 
 	ret = nouveau_bo_vma_add(priv->bo, cli->vm, &fctx->vma);
 	if (ret == 0) {
@@ -158,8 +160,6 @@ nv84_fence_context_new(struct nouveau_channel *chan)
 		ret = nouveau_bo_vma_add(bo, cli->vm, &fctx->dispc_vma[i]);
 	}
 
-	nouveau_bo_wr32(priv->bo, chan->chid * 16/4, 0x00000000);
-
 	if (ret)
 		nv84_fence_context_del(chan);
 	return ret;
@@ -168,13 +168,12 @@ nv84_fence_context_new(struct nouveau_channel *chan)
 static bool
 nv84_fence_suspend(struct nouveau_drm *drm)
 {
-	struct nouveau_fifo *pfifo = nvkm_fifo(&drm->device);
 	struct nv84_fence_priv *priv = drm->fence;
 	int i;
 
-	priv->suspend = vmalloc((pfifo->max + 1) * sizeof(u32));
+	priv->suspend = vmalloc(priv->base.contexts * sizeof(u32));
 	if (priv->suspend) {
-		for (i = 0; i <= pfifo->max; i++)
+		for (i = 0; i < priv->base.contexts; i++)
 			priv->suspend[i] = nouveau_bo_rd32(priv->bo, i*4);
 	}
 
@@ -184,12 +183,11 @@ nv84_fence_suspend(struct nouveau_drm *drm)
 static void
 nv84_fence_resume(struct nouveau_drm *drm)
 {
-	struct nouveau_fifo *pfifo = nvkm_fifo(&drm->device);
 	struct nv84_fence_priv *priv = drm->fence;
 	int i;
 
 	if (priv->suspend) {
-		for (i = 0; i <= pfifo->max; i++)
+		for (i = 0; i < priv->base.contexts; i++)
 			nouveau_bo_wr32(priv->bo, i*4, priv->suspend[i]);
 		vfree(priv->suspend);
 		priv->suspend = NULL;
@@ -229,11 +227,12 @@ nv84_fence_create(struct nouveau_drm *drm)
 	priv->base.context_new = nv84_fence_context_new;
 	priv->base.context_del = nv84_fence_context_del;
 
-	init_waitqueue_head(&priv->base.waiting);
+	priv->base.contexts = pfifo->max + 1;
+	priv->base.context_base = fence_context_alloc(priv->base.contexts);
 	priv->base.uevent = true;
 
-	ret = nouveau_bo_new(drm->dev, 16 * (pfifo->max + 1), 0,
-			     TTM_PL_FLAG_VRAM, 0, 0, NULL, &priv->bo);
+	ret = nouveau_bo_new(drm->dev, 16 * priv->base.contexts, 0,
+			     TTM_PL_FLAG_VRAM, 0, 0, NULL, NULL, &priv->bo);
 	if (ret == 0) {
 		ret = nouveau_bo_pin(priv->bo, TTM_PL_FLAG_VRAM);
 		if (ret == 0) {
@@ -246,8 +245,8 @@ nv84_fence_create(struct nouveau_drm *drm)
 	}
 
 	if (ret == 0)
-		ret = nouveau_bo_new(drm->dev, 16 * (pfifo->max + 1), 0,
-				     TTM_PL_FLAG_TT, 0, 0, NULL,
+		ret = nouveau_bo_new(drm->dev, 16 * priv->base.contexts, 0,
+				     TTM_PL_FLAG_TT, 0, 0, NULL, NULL,
 				     &priv->bo_gart);
 	if (ret == 0) {
 		ret = nouveau_bo_pin(priv->bo_gart, TTM_PL_FLAG_TT);
diff --git a/drivers/gpu/drm/nouveau/nvif/class.h b/drivers/gpu/drm/nouveau/nvif/class.h
index 573491f84792..e5a27df0672b 100644
--- a/drivers/gpu/drm/nouveau/nvif/class.h
+++ b/drivers/gpu/drm/nouveau/nvif/class.h
@@ -479,6 +479,8 @@ struct nv50_disp_core_channel_dma_v0 {
 	__u32 pushbuf;
 };
 
+#define NV50_DISP_CORE_CHANNEL_DMA_V0_NTFY_UEVENT                          0x00
+
 /* cursor immediate */
 struct nv50_disp_cursor_v0 {
 	__u8  version;
@@ -486,6 +488,8 @@ struct nv50_disp_cursor_v0 {
 	__u8  pad02[6];
 };
 
+#define NV50_DISP_CURSOR_V0_NTFY_UEVENT                                    0x00
+
 /* base */
 struct nv50_disp_base_channel_dma_v0 {
 	__u8  version;
@@ -494,6 +498,8 @@ struct nv50_disp_base_channel_dma_v0 {
 	__u32 pushbuf;
 };
 
+#define NV50_DISP_BASE_CHANNEL_DMA_V0_NTFY_UEVENT                          0x00
+
 /* overlay */
 struct nv50_disp_overlay_channel_dma_v0 {
 	__u8  version;
@@ -502,6 +508,8 @@ struct nv50_disp_overlay_channel_dma_v0 {
 	__u32 pushbuf;
 };
 
+#define NV50_DISP_OVERLAY_CHANNEL_DMA_V0_NTFY_UEVENT                       0x00
+
 /* overlay immediate */
 struct nv50_disp_overlay_v0 {
 	__u8  version;
@@ -509,6 +517,7 @@ struct nv50_disp_overlay_v0 {
 	__u8  pad02[6];
 };
 
+#define NV50_DISP_OVERLAY_V0_NTFY_UEVENT                                   0x00
 
 /*******************************************************************************
  * fermi
diff --git a/drivers/gpu/drm/omapdrm/omap_drv.c b/drivers/gpu/drm/omapdrm/omap_drv.c
index 002b9721e85a..862ba03c236c 100644
--- a/drivers/gpu/drm/omapdrm/omap_drv.c
+++ b/drivers/gpu/drm/omapdrm/omap_drv.c
@@ -629,6 +629,7 @@ static struct drm_driver omap_drm_driver = {
 		.lastclose = dev_lastclose,
 		.preclose = dev_preclose,
 		.postclose = dev_postclose,
+		.set_busid = drm_platform_set_busid,
 		.get_vblank_counter = drm_vblank_count,
 		.enable_vblank = omap_irq_enable_vblank,
 		.disable_vblank = omap_irq_disable_vblank,
diff --git a/drivers/gpu/drm/omapdrm/omap_drv.h b/drivers/gpu/drm/omapdrm/omap_drv.h
index 84d73a61b34b..60e47b33c801 100644
--- a/drivers/gpu/drm/omapdrm/omap_drv.h
+++ b/drivers/gpu/drm/omapdrm/omap_drv.h
@@ -26,6 +26,7 @@
 #include <drm/drmP.h>
 #include <drm/drm_crtc_helper.h>
 #include <drm/omap_drm.h>
+#include <drm/drm_gem.h>
 #include <linux/platform_data/omap_drm.h>
 
 
diff --git a/drivers/gpu/drm/panel/panel-simple.c b/drivers/gpu/drm/panel/panel-simple.c
index 4ce1db0a68ff..23de22f8c820 100644
--- a/drivers/gpu/drm/panel/panel-simple.c
+++ b/drivers/gpu/drm/panel/panel-simple.c
@@ -352,6 +352,30 @@ static const struct panel_desc auo_b101aw03 = {
 	},
 };
 
+static const struct drm_display_mode auo_b101xtn01_mode = {
+	.clock = 72000,
+	.hdisplay = 1366,
+	.hsync_start = 1366 + 20,
+	.hsync_end = 1366 + 20 + 70,
+	.htotal = 1366 + 20 + 70,
+	.vdisplay = 768,
+	.vsync_start = 768 + 14,
+	.vsync_end = 768 + 14 + 42,
+	.vtotal = 768 + 14 + 42,
+	.vrefresh = 60,
+	.flags = DRM_MODE_FLAG_NVSYNC | DRM_MODE_FLAG_NHSYNC,
+};
+
+static const struct panel_desc auo_b101xtn01 = {
+	.modes = &auo_b101xtn01_mode,
+	.num_modes = 1,
+	.bpc = 6,
+	.size = {
+		.width = 223,
+		.height = 125,
+	},
+};
+
 static const struct drm_display_mode auo_b133xtn01_mode = {
 	.clock = 69500,
 	.hdisplay = 1366,
@@ -616,6 +640,9 @@ static const struct of_device_id platform_of_match[] = {
 		.compatible = "auo,b101aw03",
 		.data = &auo_b101aw03,
 	}, {
+		.compatible = "auo,b101xtn01",
+		.data = &auo_b101xtn01,
+	}, {
 		.compatible = "auo,b133htn01",
 		.data = &auo_b133htn01,
 	}, {
diff --git a/drivers/gpu/drm/qxl/Makefile b/drivers/gpu/drm/qxl/Makefile
index ea046ba691d2..bacc4aff1201 100644
--- a/drivers/gpu/drm/qxl/Makefile
+++ b/drivers/gpu/drm/qxl/Makefile
@@ -4,6 +4,6 @@
 
 ccflags-y := -Iinclude/drm
 
-qxl-y := qxl_drv.o qxl_kms.o qxl_display.o qxl_ttm.o qxl_fb.o qxl_object.o qxl_gem.o qxl_cmd.o qxl_image.o qxl_draw.o qxl_debugfs.o qxl_irq.o qxl_dumb.o qxl_ioctl.o qxl_fence.o qxl_release.o
+qxl-y := qxl_drv.o qxl_kms.o qxl_display.o qxl_ttm.o qxl_fb.o qxl_object.o qxl_gem.o qxl_cmd.o qxl_image.o qxl_draw.o qxl_debugfs.o qxl_irq.o qxl_dumb.o qxl_ioctl.o qxl_release.o qxl_prime.o
 
 obj-$(CONFIG_DRM_QXL)+= qxl.o
diff --git a/drivers/gpu/drm/qxl/qxl_cmd.c b/drivers/gpu/drm/qxl/qxl_cmd.c
index eb89653a7a17..97823644d347 100644
--- a/drivers/gpu/drm/qxl/qxl_cmd.c
+++ b/drivers/gpu/drm/qxl/qxl_cmd.c
@@ -620,17 +620,10 @@ static int qxl_reap_surf(struct qxl_device *qdev, struct qxl_bo *surf, bool stal
 	if (ret == -EBUSY)
 		return -EBUSY;
 
-	if (surf->fence.num_active_releases > 0 && stall == false) {
-		qxl_bo_unreserve(surf);
-		return -EBUSY;
-	}
-
 	if (stall)
 		mutex_unlock(&qdev->surf_evict_mutex);
 
-	spin_lock(&surf->tbo.bdev->fence_lock);
 	ret = ttm_bo_wait(&surf->tbo, true, true, !stall);
-	spin_unlock(&surf->tbo.bdev->fence_lock);
 
 	if (stall)
 		mutex_lock(&qdev->surf_evict_mutex);
diff --git a/drivers/gpu/drm/qxl/qxl_debugfs.c b/drivers/gpu/drm/qxl/qxl_debugfs.c
index c3c2bbdc6674..6911b8c44492 100644
--- a/drivers/gpu/drm/qxl/qxl_debugfs.c
+++ b/drivers/gpu/drm/qxl/qxl_debugfs.c
@@ -58,9 +58,17 @@ qxl_debugfs_buffers_info(struct seq_file *m, void *data)
 	struct qxl_bo *bo;
 
 	list_for_each_entry(bo, &qdev->gem.objects, list) {
-		seq_printf(m, "size %ld, pc %d, sync obj %p, num releases %d\n",
-			   (unsigned long)bo->gem_base.size, bo->pin_count,
-			   bo->tbo.sync_obj, bo->fence.num_active_releases);
+		struct reservation_object_list *fobj;
+		int rel;
+
+		rcu_read_lock();
+		fobj = rcu_dereference(bo->tbo.resv->fence);
+		rel = fobj ? fobj->shared_count : 0;
+		rcu_read_unlock();
+
+		seq_printf(m, "size %ld, pc %d, num releases %d\n",
+			   (unsigned long)bo->gem_base.size,
+			   bo->pin_count, rel);
 	}
 	return 0;
 }
diff --git a/drivers/gpu/drm/qxl/qxl_display.c b/drivers/gpu/drm/qxl/qxl_display.c
index b8ced08b6291..af9e78546688 100644
--- a/drivers/gpu/drm/qxl/qxl_display.c
+++ b/drivers/gpu/drm/qxl/qxl_display.c
@@ -187,6 +187,54 @@ static void qxl_crtc_destroy(struct drm_crtc *crtc)
 	kfree(qxl_crtc);
 }
 
+static int qxl_crtc_page_flip(struct drm_crtc *crtc,
+                              struct drm_framebuffer *fb,
+                              struct drm_pending_vblank_event *event,
+                              uint32_t page_flip_flags)
+{
+	struct drm_device *dev = crtc->dev;
+	struct qxl_device *qdev = dev->dev_private;
+	struct qxl_crtc *qcrtc = to_qxl_crtc(crtc);
+	struct qxl_framebuffer *qfb_src = to_qxl_framebuffer(fb);
+	struct qxl_framebuffer *qfb_old = to_qxl_framebuffer(crtc->primary->fb);
+	struct qxl_bo *bo_old = gem_to_qxl_bo(qfb_old->obj);
+	struct qxl_bo *bo = gem_to_qxl_bo(qfb_src->obj);
+	unsigned long flags;
+	struct drm_clip_rect norect = {
+	    .x1 = 0,
+	    .y1 = 0,
+	    .x2 = fb->width,
+	    .y2 = fb->height
+	};
+	int inc = 1;
+	int one_clip_rect = 1;
+	int ret = 0;
+
+	crtc->primary->fb = fb;
+	bo_old->is_primary = false;
+	bo->is_primary = true;
+
+	ret = qxl_bo_reserve(bo, false);
+	if (ret)
+		return ret;
+
+	qxl_draw_dirty_fb(qdev, qfb_src, bo, 0, 0,
+			  &norect, one_clip_rect, inc);
+
+	drm_vblank_get(dev, qcrtc->index);
+
+	if (event) {
+		spin_lock_irqsave(&dev->event_lock, flags);
+		drm_send_vblank_event(dev, qcrtc->index, event);
+		spin_unlock_irqrestore(&dev->event_lock, flags);
+	}
+	drm_vblank_put(dev, qcrtc->index);
+
+	qxl_bo_unreserve(bo);
+
+	return 0;
+}
+
 static int
 qxl_hide_cursor(struct qxl_device *qdev)
 {
@@ -374,6 +422,7 @@ static const struct drm_crtc_funcs qxl_crtc_funcs = {
 	.cursor_move = qxl_crtc_cursor_move,
 	.set_config = drm_crtc_helper_set_config,
 	.destroy = qxl_crtc_destroy,
+	.page_flip = qxl_crtc_page_flip,
 };
 
 static void qxl_user_framebuffer_destroy(struct drm_framebuffer *fb)
diff --git a/drivers/gpu/drm/qxl/qxl_drv.c b/drivers/gpu/drm/qxl/qxl_drv.c
index a3fd92029a14..1d9b80c91a15 100644
--- a/drivers/gpu/drm/qxl/qxl_drv.c
+++ b/drivers/gpu/drm/qxl/qxl_drv.c
@@ -84,6 +84,7 @@ static const struct file_operations qxl_fops = {
 	.release = drm_release,
 	.unlocked_ioctl = drm_ioctl,
 	.poll = drm_poll,
+	.read = drm_read,
 	.mmap = qxl_mmap,
 };
 
@@ -195,6 +196,20 @@ static int qxl_pm_restore(struct device *dev)
 	return qxl_drm_resume(drm_dev, false);
 }
 
+static u32 qxl_noop_get_vblank_counter(struct drm_device *dev, int crtc)
+{
+	return dev->vblank[crtc].count.counter;
+}
+
+static int qxl_noop_enable_vblank(struct drm_device *dev, int crtc)
+{
+	return 0;
+}
+
+static void qxl_noop_disable_vblank(struct drm_device *dev, int crtc)
+{
+}
+
 static const struct dev_pm_ops qxl_pm_ops = {
 	.suspend = qxl_pm_suspend,
 	.resume = qxl_pm_resume,
@@ -212,10 +227,15 @@ static struct pci_driver qxl_pci_driver = {
 };
 
 static struct drm_driver qxl_driver = {
-	.driver_features = DRIVER_GEM | DRIVER_MODESET |
+	.driver_features = DRIVER_GEM | DRIVER_MODESET | DRIVER_PRIME |
 			   DRIVER_HAVE_IRQ | DRIVER_IRQ_SHARED,
 	.load = qxl_driver_load,
 	.unload = qxl_driver_unload,
+	.get_vblank_counter = qxl_noop_get_vblank_counter,
+	.enable_vblank = qxl_noop_enable_vblank,
+	.disable_vblank = qxl_noop_disable_vblank,
+
+	.set_busid = drm_pci_set_busid,
 
 	.dumb_create = qxl_mode_dumb_create,
 	.dumb_map_offset = qxl_mode_dumb_mmap,
@@ -224,6 +244,17 @@ static struct drm_driver qxl_driver = {
 	.debugfs_init = qxl_debugfs_init,
 	.debugfs_cleanup = qxl_debugfs_takedown,
 #endif
+	.prime_handle_to_fd = drm_gem_prime_handle_to_fd,
+	.prime_fd_to_handle = drm_gem_prime_fd_to_handle,
+	.gem_prime_export = drm_gem_prime_export,
+	.gem_prime_import = drm_gem_prime_import,
+	.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 = qxl_gem_object_free,
 	.gem_open_object = qxl_gem_object_open,
 	.gem_close_object = qxl_gem_object_close,
diff --git a/drivers/gpu/drm/qxl/qxl_drv.h b/drivers/gpu/drm/qxl/qxl_drv.h
index 36ed40ba773f..7c6cafe21f5f 100644
--- a/drivers/gpu/drm/qxl/qxl_drv.h
+++ b/drivers/gpu/drm/qxl/qxl_drv.h
@@ -31,6 +31,7 @@
  * Definitions taken from spice-protocol, plus kernel driver specific bits.
  */
 
+#include <linux/fence.h>
 #include <linux/workqueue.h>
 #include <linux/firmware.h>
 #include <linux/platform_device.h>
@@ -42,6 +43,8 @@
 #include <ttm/ttm_placement.h>
 #include <ttm/ttm_module.h>
 
+#include <drm/drm_gem.h>
+
 /* just for ttm_validate_buffer */
 #include <ttm/ttm_execbuf_util.h>
 
@@ -95,31 +98,24 @@ enum {
 	QXL_INTERRUPT_IO_CMD |\
 	QXL_INTERRUPT_CLIENT_MONITORS_CONFIG)
 
-struct qxl_fence {
-	struct qxl_device *qdev;
-	uint32_t num_active_releases;
-	uint32_t *release_ids;
-	struct radix_tree_root tree;
-};
-
 struct qxl_bo {
 	/* Protected by gem.mutex */
 	struct list_head		list;
 	/* Protected by tbo.reserved */
-	u32				placements[3];
+	struct ttm_place		placements[3];
 	struct ttm_placement		placement;
 	struct ttm_buffer_object	tbo;
 	struct ttm_bo_kmap_obj		kmap;
 	unsigned			pin_count;
 	void				*kptr;
 	int                             type;
+
 	/* Constant after initialization */
 	struct drm_gem_object		gem_base;
 	bool is_primary; /* is this now a primary surface */
 	bool hw_surf_alloc;
 	struct qxl_surface surf;
 	uint32_t surface_id;
-	struct qxl_fence fence; /* per bo fence  - list of releases */
 	struct qxl_release *surf_create;
 };
 #define gem_to_qxl_bo(gobj) container_of((gobj), struct qxl_bo, gem_base)
@@ -191,6 +187,8 @@ enum {
  * spice-protocol/qxl_dev.h */
 #define QXL_MAX_RES 96
 struct qxl_release {
+	struct fence base;
+
 	int id;
 	int type;
 	uint32_t release_offset;
@@ -284,7 +282,9 @@ struct qxl_device {
 	uint8_t		slot_gen_bits;
 	uint64_t	va_slot_mask;
 
+	spinlock_t	release_lock;
 	struct idr	release_idr;
+	uint32_t	release_seqno;
 	spinlock_t release_idr_lock;
 	struct mutex	async_io_mutex;
 	unsigned int last_sent_io_cmd;
@@ -532,6 +532,18 @@ int qxl_garbage_collect(struct qxl_device *qdev);
 int qxl_debugfs_init(struct drm_minor *minor);
 void qxl_debugfs_takedown(struct drm_minor *minor);
 
+/* qxl_prime.c */
+int qxl_gem_prime_pin(struct drm_gem_object *obj);
+void qxl_gem_prime_unpin(struct drm_gem_object *obj);
+struct sg_table *qxl_gem_prime_get_sg_table(struct drm_gem_object *obj);
+struct drm_gem_object *qxl_gem_prime_import_sg_table(
+	struct drm_device *dev, struct dma_buf_attachment *attach,
+	struct sg_table *sgt);
+void *qxl_gem_prime_vmap(struct drm_gem_object *obj);
+void qxl_gem_prime_vunmap(struct drm_gem_object *obj, void *vaddr);
+int qxl_gem_prime_mmap(struct drm_gem_object *obj,
+				struct vm_area_struct *vma);
+
 /* qxl_irq.c */
 int qxl_irq_init(struct qxl_device *qdev);
 irqreturn_t qxl_irq_handler(int irq, void *arg);
@@ -561,10 +573,4 @@ qxl_surface_lookup(struct drm_device *dev, int surface_id);
 void qxl_surface_evict(struct qxl_device *qdev, struct qxl_bo *surf, bool freeing);
 int qxl_update_surface(struct qxl_device *qdev, struct qxl_bo *surf);
 
-/* qxl_fence.c */
-void qxl_fence_add_release_locked(struct qxl_fence *qfence, uint32_t rel_id);
-int qxl_fence_remove_release(struct qxl_fence *qfence, uint32_t rel_id);
-int qxl_fence_init(struct qxl_device *qdev, struct qxl_fence *qfence);
-void qxl_fence_fini(struct qxl_fence *qfence);
-
 #endif
diff --git a/drivers/gpu/drm/qxl/qxl_fb.c b/drivers/gpu/drm/qxl/qxl_fb.c
index df567888bb1e..3d7c1d00a424 100644
--- a/drivers/gpu/drm/qxl/qxl_fb.c
+++ b/drivers/gpu/drm/qxl/qxl_fb.c
@@ -625,7 +625,8 @@ static int qxl_fb_find_or_create_single(
 		struct drm_fb_helper *helper,
 		struct drm_fb_helper_surface_size *sizes)
 {
-	struct qxl_fbdev *qfbdev = (struct qxl_fbdev *)helper;
+	struct qxl_fbdev *qfbdev =
+		container_of(helper, struct qxl_fbdev, helper);
 	int new_fb = 0;
 	int ret;
 
diff --git a/drivers/gpu/drm/qxl/qxl_fence.c b/drivers/gpu/drm/qxl/qxl_fence.c
deleted file mode 100644
index ae59e91cfb9a..000000000000
--- a/drivers/gpu/drm/qxl/qxl_fence.c
+++ /dev/null
@@ -1,91 +0,0 @@
-/*
- * Copyright 2013 Red Hat 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: Dave Airlie
- *          Alon Levy
- */
-
-
-#include "qxl_drv.h"
-
-/* QXL fencing-
-
-   When we submit operations to the GPU we pass a release reference to the GPU
-   with them, the release reference is then added to the release ring when
-   the GPU is finished with that particular operation and has removed it from
-   its tree.
-
-   So we have can have multiple outstanding non linear fences per object.
-
-   From a TTM POV we only care if the object has any outstanding releases on
-   it.
-
-   we wait until all outstanding releases are processeed.
-
-   sync object is just a list of release ids that represent that fence on
-   that buffer.
-
-   we just add new releases onto the sync object attached to the object.
-
-   This currently uses a radix tree to store the list of release ids.
-
-   For some reason every so often qxl hw fails to release, things go wrong.
-*/
-/* must be called with the fence lock held */
-void qxl_fence_add_release_locked(struct qxl_fence *qfence, uint32_t rel_id)
-{
-	radix_tree_insert(&qfence->tree, rel_id, qfence);
-	qfence->num_active_releases++;
-}
-
-int qxl_fence_remove_release(struct qxl_fence *qfence, uint32_t rel_id)
-{
-	void *ret;
-	int retval = 0;
-	struct qxl_bo *bo = container_of(qfence, struct qxl_bo, fence);
-
-	spin_lock(&bo->tbo.bdev->fence_lock);
-
-	ret = radix_tree_delete(&qfence->tree, rel_id);
-	if (ret == qfence)
-		qfence->num_active_releases--;
-	else {
-		DRM_DEBUG("didn't find fence in radix tree for %d\n", rel_id);
-		retval = -ENOENT;
-	}
-	spin_unlock(&bo->tbo.bdev->fence_lock);
-	return retval;
-}
-
-
-int qxl_fence_init(struct qxl_device *qdev, struct qxl_fence *qfence)
-{
-	qfence->qdev = qdev;
-	qfence->num_active_releases = 0;
-	INIT_RADIX_TREE(&qfence->tree, GFP_ATOMIC);
-	return 0;
-}
-
-void qxl_fence_fini(struct qxl_fence *qfence)
-{
-	kfree(qfence->release_ids);
-	qfence->num_active_releases = 0;
-}
diff --git a/drivers/gpu/drm/qxl/qxl_kms.c b/drivers/gpu/drm/qxl/qxl_kms.c
index fd88eb4a3f79..b2977a181935 100644
--- a/drivers/gpu/drm/qxl/qxl_kms.c
+++ b/drivers/gpu/drm/qxl/qxl_kms.c
@@ -223,6 +223,7 @@ static int qxl_device_init(struct qxl_device *qdev,
 
 	idr_init(&qdev->release_idr);
 	spin_lock_init(&qdev->release_idr_lock);
+	spin_lock_init(&qdev->release_lock);
 
 	idr_init(&qdev->surf_id_idr);
 	spin_lock_init(&qdev->surf_id_idr_lock);
@@ -297,6 +298,9 @@ int qxl_driver_unload(struct drm_device *dev)
 
 	if (qdev == NULL)
 		return 0;
+
+	drm_vblank_cleanup(dev);
+
 	qxl_modeset_fini(qdev);
 	qxl_device_fini(qdev);
 
@@ -324,15 +328,20 @@ int qxl_driver_load(struct drm_device *dev, unsigned long flags)
 	if (r)
 		goto out;
 
+	r = drm_vblank_init(dev, 1);
+	if (r)
+		goto unload;
+
 	r = qxl_modeset_init(qdev);
-	if (r) {
-		qxl_driver_unload(dev);
-		goto out;
-	}
+	if (r)
+		goto unload;
 
 	drm_kms_helper_poll_init(qdev->ddev);
 
 	return 0;
+unload:
+	qxl_driver_unload(dev);
+
 out:
 	kfree(qdev);
 	return r;
diff --git a/drivers/gpu/drm/qxl/qxl_object.c b/drivers/gpu/drm/qxl/qxl_object.c
index b95f144f0b49..cdeaf08fdc74 100644
--- a/drivers/gpu/drm/qxl/qxl_object.c
+++ b/drivers/gpu/drm/qxl/qxl_object.c
@@ -36,7 +36,6 @@ static void qxl_ttm_bo_destroy(struct ttm_buffer_object *tbo)
 	qdev = (struct qxl_device *)bo->gem_base.dev->dev_private;
 
 	qxl_surface_evict(qdev, bo, false);
-	qxl_fence_fini(&bo->fence);
 	mutex_lock(&qdev->gem.mutex);
 	list_del_init(&bo->list);
 	mutex_unlock(&qdev->gem.mutex);
@@ -55,21 +54,24 @@ 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;
+	unsigned i;
 
-	qbo->placement.fpfn = 0;
-	qbo->placement.lpfn = 0;
 	qbo->placement.placement = qbo->placements;
 	qbo->placement.busy_placement = qbo->placements;
 	if (domain == QXL_GEM_DOMAIN_VRAM)
-		qbo->placements[c++] = TTM_PL_FLAG_CACHED | TTM_PL_FLAG_VRAM | pflag;
+		qbo->placements[c++].flags = TTM_PL_FLAG_CACHED | TTM_PL_FLAG_VRAM | pflag;
 	if (domain == QXL_GEM_DOMAIN_SURFACE)
-		qbo->placements[c++] = TTM_PL_FLAG_CACHED | TTM_PL_FLAG_PRIV0 | pflag;
+		qbo->placements[c++].flags = TTM_PL_FLAG_CACHED | TTM_PL_FLAG_PRIV0 | pflag;
 	if (domain == QXL_GEM_DOMAIN_CPU)
-		qbo->placements[c++] = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM | pflag;
+		qbo->placements[c++].flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM | pflag;
 	if (!c)
-		qbo->placements[c++] = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
+		qbo->placements[c++].flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
 	qbo->placement.num_placement = c;
 	qbo->placement.num_busy_placement = c;
+	for (i = 0; i < c; ++i) {
+		qbo->placements[i].fpfn = 0;
+		qbo->placements[i].lpfn = 0;
+	}
 }
 
 
@@ -99,7 +101,6 @@ int qxl_bo_create(struct qxl_device *qdev,
 	bo->type = domain;
 	bo->pin_count = pinned ? 1 : 0;
 	bo->surface_id = 0;
-	qxl_fence_init(qdev, &bo->fence);
 	INIT_LIST_HEAD(&bo->list);
 
 	if (surf)
@@ -109,7 +110,7 @@ int qxl_bo_create(struct qxl_device *qdev,
 
 	r = ttm_bo_init(&qdev->mman.bdev, &bo->tbo, size, type,
 			&bo->placement, 0, !kernel, NULL, size,
-			NULL, &qxl_ttm_bo_destroy);
+			NULL, NULL, &qxl_ttm_bo_destroy);
 	if (unlikely(r != 0)) {
 		if (r != -ERESTARTSYS)
 			dev_err(qdev->dev,
@@ -259,7 +260,7 @@ int qxl_bo_unpin(struct qxl_bo *bo)
 	if (bo->pin_count)
 		return 0;
 	for (i = 0; i < bo->placement.num_placement; i++)
-		bo->placements[i] &= ~TTM_PL_FLAG_NO_EVICT;
+		bo->placements[i].flags &= ~TTM_PL_FLAG_NO_EVICT;
 	r = ttm_bo_validate(&bo->tbo, &bo->placement, false, false);
 	if (unlikely(r != 0))
 		dev_err(qdev->dev, "%p validate failed for unpin\n", bo);
diff --git a/drivers/gpu/drm/qxl/qxl_object.h b/drivers/gpu/drm/qxl/qxl_object.h
index 83a423293afd..37af1bc0dd00 100644
--- a/drivers/gpu/drm/qxl/qxl_object.h
+++ b/drivers/gpu/drm/qxl/qxl_object.h
@@ -76,12 +76,10 @@ static inline int qxl_bo_wait(struct qxl_bo *bo, u32 *mem_type,
 		}
 		return r;
 	}
-	spin_lock(&bo->tbo.bdev->fence_lock);
 	if (mem_type)
 		*mem_type = bo->tbo.mem.mem_type;
-	if (bo->tbo.sync_obj)
-		r = ttm_bo_wait(&bo->tbo, true, true, no_wait);
-	spin_unlock(&bo->tbo.bdev->fence_lock);
+
+	r = ttm_bo_wait(&bo->tbo, true, true, no_wait);
 	ttm_bo_unreserve(&bo->tbo);
 	return r;
 }
diff --git a/drivers/gpu/drm/qxl/qxl_prime.c b/drivers/gpu/drm/qxl/qxl_prime.c
new file mode 100644
index 000000000000..3d031b50a8fd
--- /dev/null
+++ b/drivers/gpu/drm/qxl/qxl_prime.c
@@ -0,0 +1,72 @@
+/*
+ * Copyright 2014 Canonical
+ *
+ * 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: Andreas Pokorny
+ */
+
+#include "qxl_drv.h"
+
+/* Empty Implementations as there should not be any other driver for a virtual
+ * device that might share buffers with qxl */
+
+int qxl_gem_prime_pin(struct drm_gem_object *obj)
+{
+	WARN_ONCE(1, "not implemented");
+	return -ENOSYS;
+}
+
+void qxl_gem_prime_unpin(struct drm_gem_object *obj)
+{
+	WARN_ONCE(1, "not implemented");
+}
+
+
+struct sg_table *qxl_gem_prime_get_sg_table(struct drm_gem_object *obj)
+{
+	WARN_ONCE(1, "not implemented");
+	return ERR_PTR(-ENOSYS);
+}
+
+struct drm_gem_object *qxl_gem_prime_import_sg_table(
+	struct drm_device *dev, struct dma_buf_attachment *attach,
+	struct sg_table *table)
+{
+	WARN_ONCE(1, "not implemented");
+	return ERR_PTR(-ENOSYS);
+}
+
+void *qxl_gem_prime_vmap(struct drm_gem_object *obj)
+{
+	WARN_ONCE(1, "not implemented");
+	return ERR_PTR(-ENOSYS);
+}
+
+void qxl_gem_prime_vunmap(struct drm_gem_object *obj, void *vaddr)
+{
+	WARN_ONCE(1, "not implemented");
+}
+
+int qxl_gem_prime_mmap(struct drm_gem_object *obj,
+		       struct vm_area_struct *area)
+{
+	WARN_ONCE(1, "not implemented");
+	return ENOSYS;
+}
diff --git a/drivers/gpu/drm/qxl/qxl_release.c b/drivers/gpu/drm/qxl/qxl_release.c
index 14e776f1d14e..446e71ca36cb 100644
--- a/drivers/gpu/drm/qxl/qxl_release.c
+++ b/drivers/gpu/drm/qxl/qxl_release.c
@@ -21,6 +21,7 @@
  */
 #include "qxl_drv.h"
 #include "qxl_object.h"
+#include <trace/events/fence.h>
 
 /*
  * drawable cmd cache - allocate a bunch of VRAM pages, suballocate
@@ -39,6 +40,88 @@
 static const int release_size_per_bo[] = { RELEASE_SIZE, SURFACE_RELEASE_SIZE, RELEASE_SIZE };
 static const int releases_per_bo[] = { RELEASES_PER_BO, SURFACE_RELEASES_PER_BO, RELEASES_PER_BO };
 
+static const char *qxl_get_driver_name(struct fence *fence)
+{
+	return "qxl";
+}
+
+static const char *qxl_get_timeline_name(struct fence *fence)
+{
+	return "release";
+}
+
+static bool qxl_nop_signaling(struct fence *fence)
+{
+	/* fences are always automatically signaled, so just pretend we did this.. */
+	return true;
+}
+
+static long qxl_fence_wait(struct fence *fence, bool intr, signed long timeout)
+{
+	struct qxl_device *qdev;
+	struct qxl_release *release;
+	int count = 0, sc = 0;
+	bool have_drawable_releases;
+	unsigned long cur, end = jiffies + timeout;
+
+	qdev = container_of(fence->lock, struct qxl_device, release_lock);
+	release = container_of(fence, struct qxl_release, base);
+	have_drawable_releases = release->type == QXL_RELEASE_DRAWABLE;
+
+retry:
+	sc++;
+
+	if (fence_is_signaled(fence))
+		goto signaled;
+
+	qxl_io_notify_oom(qdev);
+
+	for (count = 0; count < 11; count++) {
+		if (!qxl_queue_garbage_collect(qdev, true))
+			break;
+
+		if (fence_is_signaled(fence))
+			goto signaled;
+	}
+
+	if (fence_is_signaled(fence))
+		goto signaled;
+
+	if (have_drawable_releases || sc < 4) {
+		if (sc > 2)
+			/* back off */
+			usleep_range(500, 1000);
+
+		if (time_after(jiffies, end))
+			return 0;
+
+		if (have_drawable_releases && sc > 300) {
+			FENCE_WARN(fence, "failed to wait on release %d "
+					  "after spincount %d\n",
+					  fence->context & ~0xf0000000, sc);
+			goto signaled;
+		}
+		goto retry;
+	}
+	/*
+	 * yeah, original sync_obj_wait gave up after 3 spins when
+	 * have_drawable_releases is not set.
+	 */
+
+signaled:
+	cur = jiffies;
+	if (time_after(cur, end))
+		return 0;
+	return end - cur;
+}
+
+static const struct fence_ops qxl_fence_ops = {
+	.get_driver_name = qxl_get_driver_name,
+	.get_timeline_name = qxl_get_timeline_name,
+	.enable_signaling = qxl_nop_signaling,
+	.wait = qxl_fence_wait,
+};
+
 static uint64_t
 qxl_release_alloc(struct qxl_device *qdev, int type,
 		  struct qxl_release **ret)
@@ -46,13 +129,13 @@ qxl_release_alloc(struct qxl_device *qdev, int type,
 	struct qxl_release *release;
 	int handle;
 	size_t size = sizeof(*release);
-	int idr_ret;
 
 	release = kmalloc(size, GFP_KERNEL);
 	if (!release) {
 		DRM_ERROR("Out of memory\n");
 		return 0;
 	}
+	release->base.ops = NULL;
 	release->type = type;
 	release->release_offset = 0;
 	release->surface_release_id = 0;
@@ -60,44 +143,61 @@ qxl_release_alloc(struct qxl_device *qdev, int type,
 
 	idr_preload(GFP_KERNEL);
 	spin_lock(&qdev->release_idr_lock);
-	idr_ret = idr_alloc(&qdev->release_idr, release, 1, 0, GFP_NOWAIT);
+	handle = idr_alloc(&qdev->release_idr, release, 1, 0, GFP_NOWAIT);
+	release->base.seqno = ++qdev->release_seqno;
 	spin_unlock(&qdev->release_idr_lock);
 	idr_preload_end();
-	handle = idr_ret;
-	if (idr_ret < 0)
-		goto release_fail;
+	if (handle < 0) {
+		kfree(release);
+		*ret = NULL;
+		return handle;
+	}
 	*ret = release;
 	QXL_INFO(qdev, "allocated release %lld\n", handle);
 	release->id = handle;
-release_fail:
-
 	return handle;
 }
 
+static void
+qxl_release_free_list(struct qxl_release *release)
+{
+	while (!list_empty(&release->bos)) {
+		struct qxl_bo_list *entry;
+		struct qxl_bo *bo;
+
+		entry = container_of(release->bos.next,
+				     struct qxl_bo_list, tv.head);
+		bo = to_qxl_bo(entry->tv.bo);
+		qxl_bo_unref(&bo);
+		list_del(&entry->tv.head);
+		kfree(entry);
+	}
+}
+
 void
 qxl_release_free(struct qxl_device *qdev,
 		 struct qxl_release *release)
 {
-	struct qxl_bo_list *entry, *tmp;
 	QXL_INFO(qdev, "release %d, type %d\n", release->id,
 		 release->type);
 
 	if (release->surface_release_id)
 		qxl_surface_id_dealloc(qdev, release->surface_release_id);
 
-	list_for_each_entry_safe(entry, tmp, &release->bos, tv.head) {
-		struct qxl_bo *bo = to_qxl_bo(entry->tv.bo);
-		QXL_INFO(qdev, "release %llx\n",
-			drm_vma_node_offset_addr(&entry->tv.bo->vma_node)
-						- DRM_FILE_OFFSET);
-		qxl_fence_remove_release(&bo->fence, release->id);
-		qxl_bo_unref(&bo);
-		kfree(entry);
-	}
 	spin_lock(&qdev->release_idr_lock);
 	idr_remove(&qdev->release_idr, release->id);
 	spin_unlock(&qdev->release_idr_lock);
-	kfree(release);
+
+	if (release->base.ops) {
+		WARN_ON(list_empty(&release->bos));
+		qxl_release_free_list(release);
+
+		fence_signal(&release->base);
+		fence_put(&release->base);
+	} else {
+		qxl_release_free_list(release);
+		kfree(release);
+	}
 }
 
 static int qxl_release_bo_alloc(struct qxl_device *qdev,
@@ -126,6 +226,7 @@ int qxl_release_list_add(struct qxl_release *release, struct qxl_bo *bo)
 
 	qxl_bo_ref(bo);
 	entry->tv.bo = &bo->tbo;
+	entry->tv.shared = false;
 	list_add_tail(&entry->tv.head, &release->bos);
 	return 0;
 }
@@ -142,6 +243,10 @@ static int qxl_release_validate_bo(struct qxl_bo *bo)
 			return ret;
 	}
 
+	ret = reservation_object_reserve_shared(bo->tbo.resv);
+	if (ret)
+		return ret;
+
 	/* allocate a surface for reserved + validated buffers */
 	ret = qxl_bo_check_id(bo->gem_base.dev->dev_private, bo);
 	if (ret)
@@ -159,7 +264,7 @@ int qxl_release_reserve_list(struct qxl_release *release, bool no_intr)
 	if (list_is_singular(&release->bos))
 		return 0;
 
-	ret = ttm_eu_reserve_buffers(&release->ticket, &release->bos);
+	ret = ttm_eu_reserve_buffers(&release->ticket, &release->bos, !no_intr);
 	if (ret)
 		return ret;
 
@@ -199,6 +304,8 @@ int qxl_alloc_surface_release_reserved(struct qxl_device *qdev,
 
 		/* stash the release after the create command */
 		idr_ret = qxl_release_alloc(qdev, QXL_RELEASE_SURFACE_CMD, release);
+		if (idr_ret < 0)
+			return idr_ret;
 		bo = qxl_bo_ref(to_qxl_bo(entry->tv.bo));
 
 		(*release)->release_offset = create_rel->release_offset + 64;
@@ -239,6 +346,11 @@ int qxl_alloc_release_reserved(struct qxl_device *qdev, unsigned long size,
 	}
 
 	idr_ret = qxl_release_alloc(qdev, type, release);
+	if (idr_ret < 0) {
+		if (rbo)
+			*rbo = NULL;
+		return idr_ret;
+	}
 
 	mutex_lock(&qdev->release_mutex);
 	if (qdev->current_release_bo_offset[cur_idx] + 1 >= releases_per_bo[cur_idx]) {
@@ -319,40 +431,44 @@ void qxl_release_unmap(struct qxl_device *qdev,
 
 void qxl_release_fence_buffer_objects(struct qxl_release *release)
 {
-	struct ttm_validate_buffer *entry;
 	struct ttm_buffer_object *bo;
 	struct ttm_bo_global *glob;
 	struct ttm_bo_device *bdev;
 	struct ttm_bo_driver *driver;
 	struct qxl_bo *qbo;
+	struct ttm_validate_buffer *entry;
+	struct qxl_device *qdev;
 
 	/* if only one object on the release its the release itself
 	   since these objects are pinned no need to reserve */
-	if (list_is_singular(&release->bos))
+	if (list_is_singular(&release->bos) || list_empty(&release->bos))
 		return;
 
 	bo = list_first_entry(&release->bos, struct ttm_validate_buffer, head)->bo;
 	bdev = bo->bdev;
+	qdev = container_of(bdev, struct qxl_device, mman.bdev);
+
+	/*
+	 * Since we never really allocated a context and we don't want to conflict,
+	 * set the highest bits. This will break if we really allow exporting of dma-bufs.
+	 */
+	fence_init(&release->base, &qxl_fence_ops, &qdev->release_lock,
+		   release->id | 0xf0000000, release->base.seqno);
+	trace_fence_emit(&release->base);
+
 	driver = bdev->driver;
 	glob = bo->glob;
 
 	spin_lock(&glob->lru_lock);
-	spin_lock(&bdev->fence_lock);
 
 	list_for_each_entry(entry, &release->bos, head) {
 		bo = entry->bo;
 		qbo = to_qxl_bo(bo);
 
-		if (!entry->bo->sync_obj)
-			entry->bo->sync_obj = &qbo->fence;
-
-		qxl_fence_add_release_locked(&qbo->fence, release->id);
-
+		reservation_object_add_shared_fence(bo->resv, &release->base);
 		ttm_bo_add_to_lru(bo);
 		__ttm_bo_unreserve(bo);
-		entry->reserved = false;
 	}
-	spin_unlock(&bdev->fence_lock);
 	spin_unlock(&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 71a1baeac14e..0cbc4c987164 100644
--- a/drivers/gpu/drm/qxl/qxl_ttm.c
+++ b/drivers/gpu/drm/qxl/qxl_ttm.c
@@ -127,7 +127,7 @@ int qxl_mmap(struct file *filp, struct vm_area_struct *vma)
 	if (unlikely(vma->vm_pgoff < DRM_FILE_PAGE_OFFSET)) {
 		pr_info("%s: vma->vm_pgoff (%ld) < DRM_FILE_PAGE_OFFSET\n",
 			__func__, vma->vm_pgoff);
-		return drm_mmap(filp, vma);
+		return -EINVAL;
 	}
 
 	file_priv = filp->private_data;
@@ -188,11 +188,13 @@ static void qxl_evict_flags(struct ttm_buffer_object *bo,
 				struct ttm_placement *placement)
 {
 	struct qxl_bo *qbo;
-	static u32 placements = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
+	static struct ttm_place placements = {
+		.fpfn = 0,
+		.lpfn = 0,
+		.flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM
+	};
 
 	if (!qxl_ttm_bo_is_qxl_bo(bo)) {
-		placement->fpfn = 0;
-		placement->lpfn = 0;
 		placement->placement = &placements;
 		placement->busy_placement = &placements;
 		placement->num_placement = 1;
@@ -355,92 +357,6 @@ static int qxl_bo_move(struct ttm_buffer_object *bo,
 	return ttm_bo_move_memcpy(bo, evict, no_wait_gpu, new_mem);
 }
 
-
-static int qxl_sync_obj_wait(void *sync_obj,
-			     bool lazy, bool interruptible)
-{
-	struct qxl_fence *qfence = (struct qxl_fence *)sync_obj;
-	int count = 0, sc = 0;
-	struct qxl_bo *bo = container_of(qfence, struct qxl_bo, fence);
-
-	if (qfence->num_active_releases == 0)
-		return 0;
-
-retry:
-	if (sc == 0) {
-		if (bo->type == QXL_GEM_DOMAIN_SURFACE)
-			qxl_update_surface(qfence->qdev, bo);
-	} else if (sc >= 1) {
-		qxl_io_notify_oom(qfence->qdev);
-	}
-
-	sc++;
-
-	for (count = 0; count < 10; count++) {
-		bool ret;
-		ret = qxl_queue_garbage_collect(qfence->qdev, true);
-		if (ret == false)
-			break;
-
-		if (qfence->num_active_releases == 0)
-			return 0;
-	}
-
-	if (qfence->num_active_releases) {
-		bool have_drawable_releases = false;
-		void **slot;
-		struct radix_tree_iter iter;
-		int release_id;
-
-		radix_tree_for_each_slot(slot, &qfence->tree, &iter, 0) {
-			struct qxl_release *release;
-
-			release_id = iter.index;
-			release = qxl_release_from_id_locked(qfence->qdev, release_id);
-			if (release == NULL)
-				continue;
-
-			if (release->type == QXL_RELEASE_DRAWABLE)
-				have_drawable_releases = true;
-		}
-
-		qxl_queue_garbage_collect(qfence->qdev, true);
-
-		if (have_drawable_releases || sc < 4) {
-			if (sc > 2)
-				/* back off */
-				usleep_range(500, 1000);
-			if (have_drawable_releases && sc > 300) {
-				WARN(1, "sync obj %d still has outstanding releases %d %d %d %ld %d\n", sc, bo->surface_id, bo->is_primary, bo->pin_count, (unsigned long)bo->gem_base.size, qfence->num_active_releases);
-				return -EBUSY;
-			}
-			goto retry;
-		}
-	}
-	return 0;
-}
-
-static int qxl_sync_obj_flush(void *sync_obj)
-{
-	return 0;
-}
-
-static void qxl_sync_obj_unref(void **sync_obj)
-{
-	*sync_obj = NULL;
-}
-
-static void *qxl_sync_obj_ref(void *sync_obj)
-{
-	return sync_obj;
-}
-
-static bool qxl_sync_obj_signaled(void *sync_obj)
-{
-	struct qxl_fence *qfence = (struct qxl_fence *)sync_obj;
-	return (qfence->num_active_releases == 0);
-}
-
 static void qxl_bo_move_notify(struct ttm_buffer_object *bo,
 			       struct ttm_mem_reg *new_mem)
 {
@@ -467,16 +383,9 @@ static struct ttm_bo_driver qxl_bo_driver = {
 	.verify_access = &qxl_verify_access,
 	.io_mem_reserve = &qxl_ttm_io_mem_reserve,
 	.io_mem_free = &qxl_ttm_io_mem_free,
-	.sync_obj_signaled = &qxl_sync_obj_signaled,
-	.sync_obj_wait = &qxl_sync_obj_wait,
-	.sync_obj_flush = &qxl_sync_obj_flush,
-	.sync_obj_unref = &qxl_sync_obj_unref,
-	.sync_obj_ref = &qxl_sync_obj_ref,
 	.move_notify = &qxl_bo_move_notify,
 };
 
-
-
 int qxl_ttm_init(struct qxl_device *qdev)
 {
 	int r;
diff --git a/drivers/gpu/drm/r128/r128_cce.c b/drivers/gpu/drm/r128/r128_cce.c
index 59459fe4e8c5..2c45ac9c1dc3 100644
--- a/drivers/gpu/drm/r128/r128_cce.c
+++ b/drivers/gpu/drm/r128/r128_cce.c
@@ -452,7 +452,7 @@ static int r128_do_init_cce(struct drm_device *dev, drm_r128_init_t *init)
 	dev_priv->span_pitch_offset_c = (((dev_priv->depth_pitch / 8) << 21) |
 					 (dev_priv->span_offset >> 5));
 
-	dev_priv->sarea = drm_getsarea(dev);
+	dev_priv->sarea = drm_legacy_getsarea(dev);
 	if (!dev_priv->sarea) {
 		DRM_ERROR("could not find sarea!\n");
 		dev->dev_private = (void *)dev_priv;
@@ -460,21 +460,21 @@ static int r128_do_init_cce(struct drm_device *dev, drm_r128_init_t *init)
 		return -EINVAL;
 	}
 
-	dev_priv->mmio = drm_core_findmap(dev, init->mmio_offset);
+	dev_priv->mmio = drm_legacy_findmap(dev, init->mmio_offset);
 	if (!dev_priv->mmio) {
 		DRM_ERROR("could not find mmio region!\n");
 		dev->dev_private = (void *)dev_priv;
 		r128_do_cleanup_cce(dev);
 		return -EINVAL;
 	}
-	dev_priv->cce_ring = drm_core_findmap(dev, init->ring_offset);
+	dev_priv->cce_ring = drm_legacy_findmap(dev, init->ring_offset);
 	if (!dev_priv->cce_ring) {
 		DRM_ERROR("could not find cce ring region!\n");
 		dev->dev_private = (void *)dev_priv;
 		r128_do_cleanup_cce(dev);
 		return -EINVAL;
 	}
-	dev_priv->ring_rptr = drm_core_findmap(dev, init->ring_rptr_offset);
+	dev_priv->ring_rptr = drm_legacy_findmap(dev, init->ring_rptr_offset);
 	if (!dev_priv->ring_rptr) {
 		DRM_ERROR("could not find ring read pointer!\n");
 		dev->dev_private = (void *)dev_priv;
@@ -482,7 +482,7 @@ static int r128_do_init_cce(struct drm_device *dev, drm_r128_init_t *init)
 		return -EINVAL;
 	}
 	dev->agp_buffer_token = init->buffers_offset;
-	dev->agp_buffer_map = drm_core_findmap(dev, init->buffers_offset);
+	dev->agp_buffer_map = drm_legacy_findmap(dev, init->buffers_offset);
 	if (!dev->agp_buffer_map) {
 		DRM_ERROR("could not find dma buffer region!\n");
 		dev->dev_private = (void *)dev_priv;
@@ -492,7 +492,7 @@ static int r128_do_init_cce(struct drm_device *dev, drm_r128_init_t *init)
 
 	if (!dev_priv->is_pci) {
 		dev_priv->agp_textures =
-		    drm_core_findmap(dev, init->agp_textures_offset);
+		    drm_legacy_findmap(dev, init->agp_textures_offset);
 		if (!dev_priv->agp_textures) {
 			DRM_ERROR("could not find agp texture region!\n");
 			dev->dev_private = (void *)dev_priv;
@@ -507,9 +507,9 @@ static int r128_do_init_cce(struct drm_device *dev, drm_r128_init_t *init)
 
 #if __OS_HAS_AGP
 	if (!dev_priv->is_pci) {
-		drm_core_ioremap_wc(dev_priv->cce_ring, dev);
-		drm_core_ioremap_wc(dev_priv->ring_rptr, dev);
-		drm_core_ioremap_wc(dev->agp_buffer_map, dev);
+		drm_legacy_ioremap_wc(dev_priv->cce_ring, dev);
+		drm_legacy_ioremap_wc(dev_priv->ring_rptr, dev);
+		drm_legacy_ioremap_wc(dev->agp_buffer_map, dev);
 		if (!dev_priv->cce_ring->handle ||
 		    !dev_priv->ring_rptr->handle ||
 		    !dev->agp_buffer_map->handle) {
@@ -603,11 +603,11 @@ int r128_do_cleanup_cce(struct drm_device *dev)
 #if __OS_HAS_AGP
 		if (!dev_priv->is_pci) {
 			if (dev_priv->cce_ring != NULL)
-				drm_core_ioremapfree(dev_priv->cce_ring, dev);
+				drm_legacy_ioremapfree(dev_priv->cce_ring, dev);
 			if (dev_priv->ring_rptr != NULL)
-				drm_core_ioremapfree(dev_priv->ring_rptr, dev);
+				drm_legacy_ioremapfree(dev_priv->ring_rptr, dev);
 			if (dev->agp_buffer_map != NULL) {
-				drm_core_ioremapfree(dev->agp_buffer_map, dev);
+				drm_legacy_ioremapfree(dev->agp_buffer_map, dev);
 				dev->agp_buffer_map = NULL;
 			}
 		} else
diff --git a/drivers/gpu/drm/r128/r128_drv.c b/drivers/gpu/drm/r128/r128_drv.c
index 5bd307cd8da1..c57b4de63caf 100644
--- a/drivers/gpu/drm/r128/r128_drv.c
+++ b/drivers/gpu/drm/r128/r128_drv.c
@@ -46,7 +46,7 @@ static const struct file_operations r128_driver_fops = {
 	.open = drm_open,
 	.release = drm_release,
 	.unlocked_ioctl = drm_ioctl,
-	.mmap = drm_mmap,
+	.mmap = drm_legacy_mmap,
 	.poll = drm_poll,
 #ifdef CONFIG_COMPAT
 	.compat_ioctl = r128_compat_ioctl,
@@ -62,6 +62,7 @@ static struct drm_driver driver = {
 	.load = r128_driver_load,
 	.preclose = r128_driver_preclose,
 	.lastclose = r128_driver_lastclose,
+	.set_busid = drm_pci_set_busid,
 	.get_vblank_counter = r128_get_vblank_counter,
 	.enable_vblank = r128_enable_vblank,
 	.disable_vblank = r128_disable_vblank,
diff --git a/drivers/gpu/drm/r128/r128_drv.h b/drivers/gpu/drm/r128/r128_drv.h
index 5bf3f5ff805d..723e5d6f10a4 100644
--- a/drivers/gpu/drm/r128/r128_drv.h
+++ b/drivers/gpu/drm/r128/r128_drv.h
@@ -35,6 +35,9 @@
 #ifndef __R128_DRV_H__
 #define __R128_DRV_H__
 
+#include <drm/ati_pcigart.h>
+#include <drm/drm_legacy.h>
+
 /* General customization:
  */
 #define DRIVER_AUTHOR		"Gareth Hughes, VA Linux Systems Inc."
diff --git a/drivers/gpu/drm/radeon/Makefile b/drivers/gpu/drm/radeon/Makefile
index f77b7135ee4c..d01b87991422 100644
--- a/drivers/gpu/drm/radeon/Makefile
+++ b/drivers/gpu/drm/radeon/Makefile
@@ -60,7 +60,7 @@ radeon-y := radeon_drv.o
 
 # add UMS driver
 radeon-$(CONFIG_DRM_RADEON_UMS)+= radeon_cp.o radeon_state.o radeon_mem.o \
-	radeon_irq.o r300_cmdbuf.o r600_cp.o r600_blit.o
+	radeon_irq.o r300_cmdbuf.o r600_cp.o r600_blit.o drm_buffer.o
 
 # add KMS driver
 radeon-y += radeon_device.o radeon_asic.o radeon_kms.o \
@@ -72,7 +72,7 @@ radeon-y += radeon_device.o radeon_asic.o radeon_kms.o \
 	radeon_cs.o radeon_bios.o radeon_benchmark.o r100.o r300.o r420.o \
 	rs400.o rs600.o rs690.o rv515.o r520.o r600.o rv770.o radeon_test.o \
 	r200.o radeon_legacy_tv.o r600_cs.o r600_blit_shaders.o \
-	radeon_pm.o atombios_dp.o r600_audio.o r600_hdmi.o dce3_1_afmt.o \
+	radeon_pm.o atombios_dp.o r600_hdmi.o dce3_1_afmt.o \
 	evergreen.o evergreen_cs.o evergreen_blit_shaders.o \
 	evergreen_hdmi.o radeon_trace_points.o ni.o cayman_blit_shaders.o \
 	atombios_encoders.o radeon_semaphore.o radeon_sa.o atombios_i2c.o si.o \
@@ -80,7 +80,7 @@ radeon-y += radeon_device.o radeon_asic.o radeon_kms.o \
 	r600_dpm.o rs780_dpm.o rv6xx_dpm.o rv770_dpm.o rv730_dpm.o rv740_dpm.o \
 	rv770_smc.o cypress_dpm.o btc_dpm.o sumo_dpm.o sumo_smc.o trinity_dpm.o \
 	trinity_smc.o ni_dpm.o si_smc.o si_dpm.o kv_smc.o kv_dpm.o ci_smc.o \
-	ci_dpm.o dce6_afmt.o radeon_vm.o radeon_ucode.o radeon_ib.o
+	ci_dpm.o dce6_afmt.o radeon_vm.o radeon_ucode.o radeon_ib.o radeon_mn.o
 
 # add async DMA block
 radeon-y += \
diff --git a/drivers/gpu/drm/radeon/atombios_dp.c b/drivers/gpu/drm/radeon/atombios_dp.c
index ac14b67621d3..95d5d4ab3335 100644
--- a/drivers/gpu/drm/radeon/atombios_dp.c
+++ b/drivers/gpu/drm/radeon/atombios_dp.c
@@ -232,8 +232,8 @@ void radeon_dp_aux_init(struct radeon_connector *radeon_connector)
 
 /***** general DP utility functions *****/
 
-#define DP_VOLTAGE_MAX         DP_TRAIN_VOLTAGE_SWING_1200
-#define DP_PRE_EMPHASIS_MAX    DP_TRAIN_PRE_EMPHASIS_9_5
+#define DP_VOLTAGE_MAX         DP_TRAIN_VOLTAGE_SWING_LEVEL_3
+#define DP_PRE_EMPHASIS_MAX    DP_TRAIN_PRE_EMPH_LEVEL_3
 
 static void dp_get_adjust_train(u8 link_status[DP_LINK_STATUS_SIZE],
 				int lane_count,
diff --git a/drivers/gpu/drm/radeon/atombios_encoders.c b/drivers/gpu/drm/radeon/atombios_encoders.c
index a7f2ddf09a9d..b8cd7975f797 100644
--- a/drivers/gpu/drm/radeon/atombios_encoders.c
+++ b/drivers/gpu/drm/radeon/atombios_encoders.c
@@ -291,29 +291,6 @@ static void radeon_atom_backlight_exit(struct radeon_encoder *encoder)
 bool radeon_atom_get_tv_timings(struct radeon_device *rdev, int index,
 				struct drm_display_mode *mode);
 
-
-static inline bool radeon_encoder_is_digital(struct drm_encoder *encoder)
-{
-	struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder);
-	switch (radeon_encoder->encoder_id) {
-	case ENCODER_OBJECT_ID_INTERNAL_LVDS:
-	case ENCODER_OBJECT_ID_INTERNAL_TMDS1:
-	case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_TMDS1:
-	case ENCODER_OBJECT_ID_INTERNAL_LVTM1:
-	case ENCODER_OBJECT_ID_INTERNAL_DVO1:
-	case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_DVO1:
-	case ENCODER_OBJECT_ID_INTERNAL_DDI:
-	case ENCODER_OBJECT_ID_INTERNAL_UNIPHY:
-	case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_LVTMA:
-	case ENCODER_OBJECT_ID_INTERNAL_UNIPHY1:
-	case ENCODER_OBJECT_ID_INTERNAL_UNIPHY2:
-	case ENCODER_OBJECT_ID_INTERNAL_UNIPHY3:
-		return true;
-	default:
-		return false;
-	}
-}
-
 static bool radeon_atom_mode_fixup(struct drm_encoder *encoder,
 				   const struct drm_display_mode *mode,
 				   struct drm_display_mode *adjusted_mode)
diff --git a/drivers/gpu/drm/radeon/btc_dpm.c b/drivers/gpu/drm/radeon/btc_dpm.c
index f81d7ca134db..300d971187c4 100644
--- a/drivers/gpu/drm/radeon/btc_dpm.c
+++ b/drivers/gpu/drm/radeon/btc_dpm.c
@@ -1170,23 +1170,6 @@ static const struct radeon_blacklist_clocks btc_blacklist_clocks[] =
         { 25000, 30000, RADEON_SCLK_UP }
 };
 
-void btc_get_max_clock_from_voltage_dependency_table(struct radeon_clock_voltage_dependency_table *table,
-						     u32 *max_clock)
-{
-	u32 i, clock = 0;
-
-	if ((table == NULL) || (table->count == 0)) {
-		*max_clock = clock;
-		return;
-	}
-
-	for (i = 0; i < table->count; i++) {
-		if (clock < table->entries[i].clk)
-			clock = table->entries[i].clk;
-	}
-	*max_clock = clock;
-}
-
 void btc_apply_voltage_dependency_rules(struct radeon_clock_voltage_dependency_table *table,
 					u32 clock, u16 max_voltage, u16 *voltage)
 {
@@ -2099,7 +2082,6 @@ static void btc_apply_state_adjust_rules(struct radeon_device *rdev,
 	bool disable_mclk_switching;
 	u32 mclk, sclk;
 	u16 vddc, vddci;
-	u32 max_sclk_vddc, max_mclk_vddci, max_mclk_vddc;
 
 	if ((rdev->pm.dpm.new_active_crtc_count > 1) ||
 	    btc_dpm_vblank_too_short(rdev))
@@ -2141,39 +2123,6 @@ static void btc_apply_state_adjust_rules(struct radeon_device *rdev,
 			ps->low.vddci = max_limits->vddci;
 	}
 
-	/* limit clocks to max supported clocks based on voltage dependency tables */
-	btc_get_max_clock_from_voltage_dependency_table(&rdev->pm.dpm.dyn_state.vddc_dependency_on_sclk,
-							&max_sclk_vddc);
-	btc_get_max_clock_from_voltage_dependency_table(&rdev->pm.dpm.dyn_state.vddci_dependency_on_mclk,
-							&max_mclk_vddci);
-	btc_get_max_clock_from_voltage_dependency_table(&rdev->pm.dpm.dyn_state.vddc_dependency_on_mclk,
-							&max_mclk_vddc);
-
-	if (max_sclk_vddc) {
-		if (ps->low.sclk > max_sclk_vddc)
-			ps->low.sclk = max_sclk_vddc;
-		if (ps->medium.sclk > max_sclk_vddc)
-			ps->medium.sclk = max_sclk_vddc;
-		if (ps->high.sclk > max_sclk_vddc)
-			ps->high.sclk = max_sclk_vddc;
-	}
-	if (max_mclk_vddci) {
-		if (ps->low.mclk > max_mclk_vddci)
-			ps->low.mclk = max_mclk_vddci;
-		if (ps->medium.mclk > max_mclk_vddci)
-			ps->medium.mclk = max_mclk_vddci;
-		if (ps->high.mclk > max_mclk_vddci)
-			ps->high.mclk = max_mclk_vddci;
-	}
-	if (max_mclk_vddc) {
-		if (ps->low.mclk > max_mclk_vddc)
-			ps->low.mclk = max_mclk_vddc;
-		if (ps->medium.mclk > max_mclk_vddc)
-			ps->medium.mclk = max_mclk_vddc;
-		if (ps->high.mclk > max_mclk_vddc)
-			ps->high.mclk = max_mclk_vddc;
-	}
-
 	/* XXX validate the min clocks required for display */
 
 	if (disable_mclk_switching) {
diff --git a/drivers/gpu/drm/radeon/btc_dpm.h b/drivers/gpu/drm/radeon/btc_dpm.h
index 3b6f12b7760b..1a15e0e41950 100644
--- a/drivers/gpu/drm/radeon/btc_dpm.h
+++ b/drivers/gpu/drm/radeon/btc_dpm.h
@@ -46,8 +46,6 @@ void btc_adjust_clock_combinations(struct radeon_device *rdev,
 				   struct rv7xx_pl *pl);
 void btc_apply_voltage_dependency_rules(struct radeon_clock_voltage_dependency_table *table,
 					u32 clock, u16 max_voltage, u16 *voltage);
-void btc_get_max_clock_from_voltage_dependency_table(struct radeon_clock_voltage_dependency_table *table,
-						     u32 *max_clock);
 void btc_apply_voltage_delta_rules(struct radeon_device *rdev,
 				   u16 max_vddc, u16 max_vddci,
 				   u16 *vddc, u16 *vddci);
diff --git a/drivers/gpu/drm/radeon/ci_dpm.c b/drivers/gpu/drm/radeon/ci_dpm.c
index d416bb2ff48d..f5c8c0445a94 100644
--- a/drivers/gpu/drm/radeon/ci_dpm.c
+++ b/drivers/gpu/drm/radeon/ci_dpm.c
@@ -162,8 +162,6 @@ static const struct ci_pt_config_reg didt_config_ci[] =
 };
 
 extern u8 rv770_get_memory_module_index(struct radeon_device *rdev);
-extern void btc_get_max_clock_from_voltage_dependency_table(struct radeon_clock_voltage_dependency_table *table,
-							    u32 *max_clock);
 extern int ni_copy_and_switch_arb_sets(struct radeon_device *rdev,
 				       u32 arb_freq_src, u32 arb_freq_dest);
 extern u8 si_get_ddr3_mclk_frequency_ratio(u32 memory_clock);
@@ -748,7 +746,6 @@ static void ci_apply_state_adjust_rules(struct radeon_device *rdev,
 	struct radeon_clock_and_voltage_limits *max_limits;
 	bool disable_mclk_switching;
 	u32 sclk, mclk;
-	u32 max_sclk_vddc, max_mclk_vddci, max_mclk_vddc;
 	int i;
 
 	if (rps->vce_active) {
@@ -784,29 +781,6 @@ static void ci_apply_state_adjust_rules(struct radeon_device *rdev,
 		}
 	}
 
-	/* limit clocks to max supported clocks based on voltage dependency tables */
-	btc_get_max_clock_from_voltage_dependency_table(&rdev->pm.dpm.dyn_state.vddc_dependency_on_sclk,
-							&max_sclk_vddc);
-	btc_get_max_clock_from_voltage_dependency_table(&rdev->pm.dpm.dyn_state.vddci_dependency_on_mclk,
-							&max_mclk_vddci);
-	btc_get_max_clock_from_voltage_dependency_table(&rdev->pm.dpm.dyn_state.vddc_dependency_on_mclk,
-							&max_mclk_vddc);
-
-	for (i = 0; i < ps->performance_level_count; i++) {
-		if (max_sclk_vddc) {
-			if (ps->performance_levels[i].sclk > max_sclk_vddc)
-				ps->performance_levels[i].sclk = max_sclk_vddc;
-		}
-		if (max_mclk_vddci) {
-			if (ps->performance_levels[i].mclk > max_mclk_vddci)
-				ps->performance_levels[i].mclk = max_mclk_vddci;
-		}
-		if (max_mclk_vddc) {
-			if (ps->performance_levels[i].mclk > max_mclk_vddc)
-				ps->performance_levels[i].mclk = max_mclk_vddc;
-		}
-	}
-
 	/* XXX validate the min clocks required for display */
 
 	if (disable_mclk_switching) {
@@ -5293,9 +5267,13 @@ int ci_dpm_init(struct radeon_device *rdev)
 void ci_dpm_debugfs_print_current_performance_level(struct radeon_device *rdev,
 						    struct seq_file *m)
 {
+	struct ci_power_info *pi = ci_get_pi(rdev);
+	struct radeon_ps *rps = &pi->current_rps;
 	u32 sclk = ci_get_average_sclk_freq(rdev);
 	u32 mclk = ci_get_average_mclk_freq(rdev);
 
+	seq_printf(m, "uvd    %sabled\n", pi->uvd_enabled ? "en" : "dis");
+	seq_printf(m, "vce    %sabled\n", rps->vce_active ? "en" : "dis");
 	seq_printf(m, "power level avg    sclk: %u mclk: %u\n",
 		   sclk, mclk);
 }
diff --git a/drivers/gpu/drm/radeon/cik.c b/drivers/gpu/drm/radeon/cik.c
index 3d546c606b43..377afa504d2b 100644
--- a/drivers/gpu/drm/radeon/cik.c
+++ b/drivers/gpu/drm/radeon/cik.c
@@ -3959,18 +3959,19 @@ bool cik_semaphore_ring_emit(struct radeon_device *rdev,
  * @src_offset: src GPU address
  * @dst_offset: dst GPU address
  * @num_gpu_pages: number of GPU pages to xfer
- * @fence: radeon fence object
+ * @resv: reservation object to sync to
  *
  * Copy GPU paging using the CP DMA engine (CIK+).
  * Used by the radeon ttm implementation to move pages if
  * registered as the asic copy callback.
  */
-int cik_copy_cpdma(struct radeon_device *rdev,
-		   uint64_t src_offset, uint64_t dst_offset,
-		   unsigned num_gpu_pages,
-		   struct radeon_fence **fence)
+struct radeon_fence *cik_copy_cpdma(struct radeon_device *rdev,
+				    uint64_t src_offset, uint64_t dst_offset,
+				    unsigned num_gpu_pages,
+				    struct reservation_object *resv)
 {
 	struct radeon_semaphore *sem = NULL;
+	struct radeon_fence *fence;
 	int ring_index = rdev->asic->copy.blit_ring_index;
 	struct radeon_ring *ring = &rdev->ring[ring_index];
 	u32 size_in_bytes, cur_size_in_bytes, control;
@@ -3980,7 +3981,7 @@ int cik_copy_cpdma(struct radeon_device *rdev,
 	r = radeon_semaphore_create(rdev, &sem);
 	if (r) {
 		DRM_ERROR("radeon: moving bo (%d).\n", r);
-		return r;
+		return ERR_PTR(r);
 	}
 
 	size_in_bytes = (num_gpu_pages << RADEON_GPU_PAGE_SHIFT);
@@ -3989,10 +3990,10 @@ int cik_copy_cpdma(struct radeon_device *rdev,
 	if (r) {
 		DRM_ERROR("radeon: moving bo (%d).\n", r);
 		radeon_semaphore_free(rdev, &sem, NULL);
-		return r;
+		return ERR_PTR(r);
 	}
 
-	radeon_semaphore_sync_to(sem, *fence);
+	radeon_semaphore_sync_resv(rdev, sem, resv, false);
 	radeon_semaphore_sync_rings(rdev, sem, ring->idx);
 
 	for (i = 0; i < num_loops; i++) {
@@ -4014,17 +4015,17 @@ int cik_copy_cpdma(struct radeon_device *rdev,
 		dst_offset += cur_size_in_bytes;
 	}
 
-	r = radeon_fence_emit(rdev, fence, ring->idx);
+	r = radeon_fence_emit(rdev, &fence, ring->idx);
 	if (r) {
 		radeon_ring_unlock_undo(rdev, ring);
 		radeon_semaphore_free(rdev, &sem, NULL);
-		return r;
+		return ERR_PTR(r);
 	}
 
 	radeon_ring_unlock_commit(rdev, ring, false);
-	radeon_semaphore_free(rdev, &sem, *fence);
+	radeon_semaphore_free(rdev, &sem, fence);
 
-	return r;
+	return fence;
 }
 
 /*
@@ -4234,7 +4235,7 @@ static int cik_cp_gfx_load_microcode(struct radeon_device *rdev)
 		WREG32(CP_PFP_UCODE_ADDR, 0);
 		for (i = 0; i < fw_size; i++)
 			WREG32(CP_PFP_UCODE_DATA, le32_to_cpup(fw_data++));
-		WREG32(CP_PFP_UCODE_ADDR, 0);
+		WREG32(CP_PFP_UCODE_ADDR, le32_to_cpu(pfp_hdr->header.ucode_version));
 
 		/* CE */
 		fw_data = (const __le32 *)
@@ -4243,7 +4244,7 @@ static int cik_cp_gfx_load_microcode(struct radeon_device *rdev)
 		WREG32(CP_CE_UCODE_ADDR, 0);
 		for (i = 0; i < fw_size; i++)
 			WREG32(CP_CE_UCODE_DATA, le32_to_cpup(fw_data++));
-		WREG32(CP_CE_UCODE_ADDR, 0);
+		WREG32(CP_CE_UCODE_ADDR, le32_to_cpu(ce_hdr->header.ucode_version));
 
 		/* ME */
 		fw_data = (const __be32 *)
@@ -4252,7 +4253,8 @@ static int cik_cp_gfx_load_microcode(struct radeon_device *rdev)
 		WREG32(CP_ME_RAM_WADDR, 0);
 		for (i = 0; i < fw_size; i++)
 			WREG32(CP_ME_RAM_DATA, le32_to_cpup(fw_data++));
-		WREG32(CP_ME_RAM_WADDR, 0);
+		WREG32(CP_ME_RAM_WADDR, le32_to_cpu(me_hdr->header.ucode_version));
+		WREG32(CP_ME_RAM_RADDR, le32_to_cpu(me_hdr->header.ucode_version));
 	} else {
 		const __be32 *fw_data;
 
@@ -4278,10 +4280,6 @@ static int cik_cp_gfx_load_microcode(struct radeon_device *rdev)
 		WREG32(CP_ME_RAM_WADDR, 0);
 	}
 
-	WREG32(CP_PFP_UCODE_ADDR, 0);
-	WREG32(CP_CE_UCODE_ADDR, 0);
-	WREG32(CP_ME_RAM_WADDR, 0);
-	WREG32(CP_ME_RAM_RADDR, 0);
 	return 0;
 }
 
@@ -4563,7 +4561,7 @@ static int cik_cp_compute_load_microcode(struct radeon_device *rdev)
 		WREG32(CP_MEC_ME1_UCODE_ADDR, 0);
 		for (i = 0; i < fw_size; i++)
 			WREG32(CP_MEC_ME1_UCODE_DATA, le32_to_cpup(fw_data++));
-		WREG32(CP_MEC_ME1_UCODE_ADDR, 0);
+		WREG32(CP_MEC_ME1_UCODE_ADDR, le32_to_cpu(mec_hdr->header.ucode_version));
 
 		/* MEC2 */
 		if (rdev->family == CHIP_KAVERI) {
@@ -4577,7 +4575,7 @@ static int cik_cp_compute_load_microcode(struct radeon_device *rdev)
 			WREG32(CP_MEC_ME2_UCODE_ADDR, 0);
 			for (i = 0; i < fw_size; i++)
 				WREG32(CP_MEC_ME2_UCODE_DATA, le32_to_cpup(fw_data++));
-			WREG32(CP_MEC_ME2_UCODE_ADDR, 0);
+			WREG32(CP_MEC_ME2_UCODE_ADDR, le32_to_cpu(mec2_hdr->header.ucode_version));
 		}
 	} else {
 		const __be32 *fw_data;
@@ -4689,7 +4687,7 @@ static int cik_mec_init(struct radeon_device *rdev)
 		r = radeon_bo_create(rdev,
 				     rdev->mec.num_mec *rdev->mec.num_pipe * MEC_HPD_SIZE * 2,
 				     PAGE_SIZE, true,
-				     RADEON_GEM_DOMAIN_GTT, 0, NULL,
+				     RADEON_GEM_DOMAIN_GTT, 0, NULL, NULL,
 				     &rdev->mec.hpd_eop_obj);
 		if (r) {
 			dev_warn(rdev->dev, "(%d) create HDP EOP bo failed\n", r);
@@ -4860,7 +4858,7 @@ static int cik_cp_compute_resume(struct radeon_device *rdev)
 					     sizeof(struct bonaire_mqd),
 					     PAGE_SIZE, true,
 					     RADEON_GEM_DOMAIN_GTT, 0, NULL,
-					     &rdev->ring[idx].mqd_obj);
+					     NULL, &rdev->ring[idx].mqd_obj);
 			if (r) {
 				dev_warn(rdev->dev, "(%d) create MQD bo failed\n", r);
 				return r;
@@ -6226,7 +6224,7 @@ static int cik_rlc_resume(struct radeon_device *rdev)
 		WREG32(RLC_GPM_UCODE_ADDR, 0);
 		for (i = 0; i < size; i++)
 			WREG32(RLC_GPM_UCODE_DATA, le32_to_cpup(fw_data++));
-		WREG32(RLC_GPM_UCODE_ADDR, 0);
+		WREG32(RLC_GPM_UCODE_ADDR, le32_to_cpu(hdr->header.ucode_version));
 	} else {
 		const __be32 *fw_data;
 
@@ -8255,8 +8253,10 @@ restart_ih:
 	}
 	if (queue_hotplug)
 		schedule_work(&rdev->hotplug_work);
-	if (queue_reset)
-		schedule_work(&rdev->reset_work);
+	if (queue_reset) {
+		rdev->needs_reset = true;
+		wake_up_all(&rdev->fence_queue);
+	}
 	if (queue_thermal)
 		schedule_work(&rdev->pm.dpm.thermal.work);
 	rdev->ih.rptr = rptr;
diff --git a/drivers/gpu/drm/radeon/cik_sdma.c b/drivers/gpu/drm/radeon/cik_sdma.c
index c4ffa54b1e3d..c77dad1a4576 100644
--- a/drivers/gpu/drm/radeon/cik_sdma.c
+++ b/drivers/gpu/drm/radeon/cik_sdma.c
@@ -530,18 +530,19 @@ void cik_sdma_fini(struct radeon_device *rdev)
  * @src_offset: src GPU address
  * @dst_offset: dst GPU address
  * @num_gpu_pages: number of GPU pages to xfer
- * @fence: radeon fence object
+ * @resv: reservation object to sync to
  *
  * Copy GPU paging using the DMA engine (CIK).
  * Used by the radeon ttm implementation to move pages if
  * registered as the asic copy callback.
  */
-int cik_copy_dma(struct radeon_device *rdev,
-		 uint64_t src_offset, uint64_t dst_offset,
-		 unsigned num_gpu_pages,
-		 struct radeon_fence **fence)
+struct radeon_fence *cik_copy_dma(struct radeon_device *rdev,
+				  uint64_t src_offset, uint64_t dst_offset,
+				  unsigned num_gpu_pages,
+				  struct reservation_object *resv)
 {
 	struct radeon_semaphore *sem = NULL;
+	struct radeon_fence *fence;
 	int ring_index = rdev->asic->copy.dma_ring_index;
 	struct radeon_ring *ring = &rdev->ring[ring_index];
 	u32 size_in_bytes, cur_size_in_bytes;
@@ -551,7 +552,7 @@ int cik_copy_dma(struct radeon_device *rdev,
 	r = radeon_semaphore_create(rdev, &sem);
 	if (r) {
 		DRM_ERROR("radeon: moving bo (%d).\n", r);
-		return r;
+		return ERR_PTR(r);
 	}
 
 	size_in_bytes = (num_gpu_pages << RADEON_GPU_PAGE_SHIFT);
@@ -560,10 +561,10 @@ int cik_copy_dma(struct radeon_device *rdev,
 	if (r) {
 		DRM_ERROR("radeon: moving bo (%d).\n", r);
 		radeon_semaphore_free(rdev, &sem, NULL);
-		return r;
+		return ERR_PTR(r);
 	}
 
-	radeon_semaphore_sync_to(sem, *fence);
+	radeon_semaphore_sync_resv(rdev, sem, resv, false);
 	radeon_semaphore_sync_rings(rdev, sem, ring->idx);
 
 	for (i = 0; i < num_loops; i++) {
@@ -582,17 +583,17 @@ int cik_copy_dma(struct radeon_device *rdev,
 		dst_offset += cur_size_in_bytes;
 	}
 
-	r = radeon_fence_emit(rdev, fence, ring->idx);
+	r = radeon_fence_emit(rdev, &fence, ring->idx);
 	if (r) {
 		radeon_ring_unlock_undo(rdev, ring);
 		radeon_semaphore_free(rdev, &sem, NULL);
-		return r;
+		return ERR_PTR(r);
 	}
 
 	radeon_ring_unlock_commit(rdev, ring, false);
-	radeon_semaphore_free(rdev, &sem, *fence);
+	radeon_semaphore_free(rdev, &sem, fence);
 
-	return r;
+	return fence;
 }
 
 /**
diff --git a/drivers/gpu/drm/radeon/dce3_1_afmt.c b/drivers/gpu/drm/radeon/dce3_1_afmt.c
index 51800e340a57..950af153f30e 100644
--- a/drivers/gpu/drm/radeon/dce3_1_afmt.c
+++ b/drivers/gpu/drm/radeon/dce3_1_afmt.c
@@ -165,7 +165,7 @@ void dce3_1_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode *m
 
 	/* disable audio prior to setting up hw */
 	dig->afmt->pin = r600_audio_get_pin(rdev);
-	r600_audio_enable(rdev, dig->afmt->pin, false);
+	r600_audio_enable(rdev, dig->afmt->pin, 0);
 
 	r600_audio_set_dto(encoder, mode->clock);
 
@@ -240,5 +240,5 @@ void dce3_1_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode *m
 	r600_hdmi_audio_workaround(encoder);
 
 	/* enable audio after to setting up hw */
-	r600_audio_enable(rdev, dig->afmt->pin, true);
+	r600_audio_enable(rdev, dig->afmt->pin, 0xf);
 }
diff --git a/drivers/gpu/drm/radeon/dce6_afmt.c b/drivers/gpu/drm/radeon/dce6_afmt.c
index ab29f953a767..c0bbf68dbc27 100644
--- a/drivers/gpu/drm/radeon/dce6_afmt.c
+++ b/drivers/gpu/drm/radeon/dce6_afmt.c
@@ -284,13 +284,13 @@ static int dce6_audio_chipset_supported(struct radeon_device *rdev)
 
 void dce6_audio_enable(struct radeon_device *rdev,
 		       struct r600_audio_pin *pin,
-		       bool enable)
+		       u8 enable_mask)
 {
 	if (!pin)
 		return;
 
-	WREG32_ENDPOINT(pin->offset, AZ_F0_CODEC_PIN_CONTROL_HOTPLUG_CONTROL,
-			enable ? AUDIO_ENABLED : 0);
+	WREG32_ENDPOINT(pin->offset, AZ_F0_CODEC_PIN_CONTROL_HOT_PLUG_CONTROL,
+			enable_mask ? AUDIO_ENABLED : 0);
 }
 
 static const u32 pin_offsets[7] =
diff --git a/drivers/gpu/drm/drm_buffer.c b/drivers/gpu/drm/radeon/drm_buffer.c
index 86a4a4a60afc..f4e0f3a3d7b1 100644
--- a/drivers/gpu/drm/drm_buffer.c
+++ b/drivers/gpu/drm/radeon/drm_buffer.c
@@ -33,7 +33,7 @@
  */
 
 #include <linux/export.h>
-#include <drm/drm_buffer.h>
+#include "drm_buffer.h"
 
 /**
  * Allocate the drm buffer object.
@@ -86,7 +86,6 @@ error_out:
 	kfree(*buf);
 	return -ENOMEM;
 }
-EXPORT_SYMBOL(drm_buffer_alloc);
 
 /**
  * Copy the user data to the begin of the buffer and reset the processing
@@ -123,7 +122,6 @@ int drm_buffer_copy_from_user(struct drm_buffer *buf,
 	buf->iterator = 0;
 	return 0;
 }
-EXPORT_SYMBOL(drm_buffer_copy_from_user);
 
 /**
  * Free the drm buffer object
@@ -141,7 +139,6 @@ void drm_buffer_free(struct drm_buffer *buf)
 		kfree(buf);
 	}
 }
-EXPORT_SYMBOL(drm_buffer_free);
 
 /**
  * Read an object from buffer that may be split to multiple parts. If object
@@ -178,4 +175,3 @@ void *drm_buffer_read_object(struct drm_buffer *buf,
 	drm_buffer_advance(buf, objsize);
 	return obj;
 }
-EXPORT_SYMBOL(drm_buffer_read_object);
diff --git a/drivers/gpu/drm/radeon/drm_buffer.h b/drivers/gpu/drm/radeon/drm_buffer.h
new file mode 100644
index 000000000000..c80d3a340b94
--- /dev/null
+++ b/drivers/gpu/drm/radeon/drm_buffer.h
@@ -0,0 +1,148 @@
+/**************************************************************************
+ *
+ * Copyright 2010 Pauli Nieminen.
+ * All Rights Reserved.
+ *
+ * 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.
+ *
+ *
+ **************************************************************************/
+/*
+ * Multipart buffer for coping data which is larger than the page size.
+ *
+ * Authors:
+ * Pauli Nieminen <suokkos-at-gmail-dot-com>
+ */
+
+#ifndef _DRM_BUFFER_H_
+#define _DRM_BUFFER_H_
+
+#include <drm/drmP.h>
+
+struct drm_buffer {
+	int iterator;
+	int size;
+	char *data[];
+};
+
+
+/**
+ * Return the index of page that buffer is currently pointing at.
+ */
+static inline int drm_buffer_page(struct drm_buffer *buf)
+{
+	return buf->iterator / PAGE_SIZE;
+}
+/**
+ * Return the index of the current byte in the page
+ */
+static inline int drm_buffer_index(struct drm_buffer *buf)
+{
+	return buf->iterator & (PAGE_SIZE - 1);
+}
+/**
+ * Return number of bytes that is left to process
+ */
+static inline int drm_buffer_unprocessed(struct drm_buffer *buf)
+{
+	return buf->size - buf->iterator;
+}
+
+/**
+ * Advance the buffer iterator number of bytes that is given.
+ */
+static inline void drm_buffer_advance(struct drm_buffer *buf, int bytes)
+{
+	buf->iterator += bytes;
+}
+
+/**
+ * Allocate the drm buffer object.
+ *
+ *   buf: A pointer to a pointer where the object is stored.
+ *   size: The number of bytes to allocate.
+ */
+extern int drm_buffer_alloc(struct drm_buffer **buf, int size);
+
+/**
+ * Copy the user data to the begin of the buffer and reset the processing
+ * iterator.
+ *
+ *   user_data: A pointer the data that is copied to the buffer.
+ *   size: The Number of bytes to copy.
+ */
+extern int drm_buffer_copy_from_user(struct drm_buffer *buf,
+		void __user *user_data, int size);
+
+/**
+ * Free the drm buffer object
+ */
+extern void drm_buffer_free(struct drm_buffer *buf);
+
+/**
+ * Read an object from buffer that may be split to multiple parts. If object
+ * is not split function just returns the pointer to object in buffer. But in
+ * case of split object data is copied to given stack object that is suplied
+ * by caller.
+ *
+ * The processing location of the buffer is also advanced to the next byte
+ * after the object.
+ *
+ *   objsize: The size of the objet in bytes.
+ *   stack_obj: A pointer to a memory location where object can be copied.
+ */
+extern void *drm_buffer_read_object(struct drm_buffer *buf,
+		int objsize, void *stack_obj);
+
+/**
+ * Returns the pointer to the dword which is offset number of elements from the
+ * current processing location.
+ *
+ * Caller must make sure that dword is not split in the buffer. This
+ * requirement is easily met if all the sizes of objects in buffer are
+ * multiples of dword and PAGE_SIZE is multiple dword.
+ *
+ * Call to this function doesn't change the processing location.
+ *
+ *   offset: The index of the dword relative to the internat iterator.
+ */
+static inline void *drm_buffer_pointer_to_dword(struct drm_buffer *buffer,
+		int offset)
+{
+	int iter = buffer->iterator + offset * 4;
+	return &buffer->data[iter / PAGE_SIZE][iter & (PAGE_SIZE - 1)];
+}
+/**
+ * Returns the pointer to the dword which is offset number of elements from
+ * the current processing location.
+ *
+ * Call to this function doesn't change the processing location.
+ *
+ *   offset: The index of the byte relative to the internat iterator.
+ */
+static inline void *drm_buffer_pointer_to_byte(struct drm_buffer *buffer,
+		int offset)
+{
+	int iter = buffer->iterator + offset;
+	return &buffer->data[iter / PAGE_SIZE][iter & (PAGE_SIZE - 1)];
+}
+
+#endif
diff --git a/drivers/gpu/drm/radeon/evergreen.c b/drivers/gpu/drm/radeon/evergreen.c
index e50807c29f69..a31f1ca40c6a 100644
--- a/drivers/gpu/drm/radeon/evergreen.c
+++ b/drivers/gpu/drm/radeon/evergreen.c
@@ -22,7 +22,6 @@
  * Authors: Alex Deucher
  */
 #include <linux/firmware.h>
-#include <linux/platform_device.h>
 #include <linux/slab.h>
 #include <drm/drmP.h>
 #include "radeon.h"
@@ -4023,7 +4022,7 @@ int sumo_rlc_init(struct radeon_device *rdev)
 		if (rdev->rlc.save_restore_obj == NULL) {
 			r = radeon_bo_create(rdev, dws * 4, PAGE_SIZE, true,
 					     RADEON_GEM_DOMAIN_VRAM, 0, NULL,
-					     &rdev->rlc.save_restore_obj);
+					     NULL, &rdev->rlc.save_restore_obj);
 			if (r) {
 				dev_warn(rdev->dev, "(%d) create RLC sr bo failed\n", r);
 				return r;
@@ -4102,7 +4101,7 @@ int sumo_rlc_init(struct radeon_device *rdev)
 		if (rdev->rlc.clear_state_obj == NULL) {
 			r = radeon_bo_create(rdev, dws * 4, PAGE_SIZE, true,
 					     RADEON_GEM_DOMAIN_VRAM, 0, NULL,
-					     &rdev->rlc.clear_state_obj);
+					     NULL, &rdev->rlc.clear_state_obj);
 			if (r) {
 				dev_warn(rdev->dev, "(%d) create RLC c bo failed\n", r);
 				sumo_rlc_fini(rdev);
@@ -4179,7 +4178,7 @@ int sumo_rlc_init(struct radeon_device *rdev)
 			r = radeon_bo_create(rdev, rdev->rlc.cp_table_size,
 					     PAGE_SIZE, true,
 					     RADEON_GEM_DOMAIN_VRAM, 0, NULL,
-					     &rdev->rlc.cp_table_obj);
+					     NULL, &rdev->rlc.cp_table_obj);
 			if (r) {
 				dev_warn(rdev->dev, "(%d) create RLC cp table bo failed\n", r);
 				sumo_rlc_fini(rdev);
diff --git a/drivers/gpu/drm/radeon/evergreen_dma.c b/drivers/gpu/drm/radeon/evergreen_dma.c
index afaba388c36d..66bcfadeedd1 100644
--- a/drivers/gpu/drm/radeon/evergreen_dma.c
+++ b/drivers/gpu/drm/radeon/evergreen_dma.c
@@ -104,12 +104,14 @@ void evergreen_dma_ring_ib_execute(struct radeon_device *rdev,
  * Used by the radeon ttm implementation to move pages if
  * registered as the asic copy callback.
  */
-int evergreen_copy_dma(struct radeon_device *rdev,
-		       uint64_t src_offset, uint64_t dst_offset,
-		       unsigned num_gpu_pages,
-		       struct radeon_fence **fence)
+struct radeon_fence *evergreen_copy_dma(struct radeon_device *rdev,
+					uint64_t src_offset,
+					uint64_t dst_offset,
+					unsigned num_gpu_pages,
+					struct reservation_object *resv)
 {
 	struct radeon_semaphore *sem = NULL;
+	struct radeon_fence *fence;
 	int ring_index = rdev->asic->copy.dma_ring_index;
 	struct radeon_ring *ring = &rdev->ring[ring_index];
 	u32 size_in_dw, cur_size_in_dw;
@@ -119,7 +121,7 @@ int evergreen_copy_dma(struct radeon_device *rdev,
 	r = radeon_semaphore_create(rdev, &sem);
 	if (r) {
 		DRM_ERROR("radeon: moving bo (%d).\n", r);
-		return r;
+		return ERR_PTR(r);
 	}
 
 	size_in_dw = (num_gpu_pages << RADEON_GPU_PAGE_SHIFT) / 4;
@@ -128,10 +130,10 @@ int evergreen_copy_dma(struct radeon_device *rdev,
 	if (r) {
 		DRM_ERROR("radeon: moving bo (%d).\n", r);
 		radeon_semaphore_free(rdev, &sem, NULL);
-		return r;
+		return ERR_PTR(r);
 	}
 
-	radeon_semaphore_sync_to(sem, *fence);
+	radeon_semaphore_sync_resv(rdev, sem, resv, false);
 	radeon_semaphore_sync_rings(rdev, sem, ring->idx);
 
 	for (i = 0; i < num_loops; i++) {
@@ -148,17 +150,17 @@ int evergreen_copy_dma(struct radeon_device *rdev,
 		dst_offset += cur_size_in_dw * 4;
 	}
 
-	r = radeon_fence_emit(rdev, fence, ring->idx);
+	r = radeon_fence_emit(rdev, &fence, ring->idx);
 	if (r) {
 		radeon_ring_unlock_undo(rdev, ring);
 		radeon_semaphore_free(rdev, &sem, NULL);
-		return r;
+		return ERR_PTR(r);
 	}
 
 	radeon_ring_unlock_commit(rdev, ring, false);
-	radeon_semaphore_free(rdev, &sem, *fence);
+	radeon_semaphore_free(rdev, &sem, fence);
 
-	return r;
+	return fence;
 }
 
 /**
diff --git a/drivers/gpu/drm/radeon/evergreen_hdmi.c b/drivers/gpu/drm/radeon/evergreen_hdmi.c
index 278c7a139d74..2514d659b1ba 100644
--- a/drivers/gpu/drm/radeon/evergreen_hdmi.c
+++ b/drivers/gpu/drm/radeon/evergreen_hdmi.c
@@ -38,6 +38,37 @@ extern void dce6_afmt_select_pin(struct drm_encoder *encoder);
 extern void dce6_afmt_write_latency_fields(struct drm_encoder *encoder,
 					   struct drm_display_mode *mode);
 
+/* enable the audio stream */
+static void dce4_audio_enable(struct radeon_device *rdev,
+			      struct r600_audio_pin *pin,
+			      u8 enable_mask)
+{
+	u32 tmp = RREG32(AZ_HOT_PLUG_CONTROL);
+
+	if (!pin)
+		return;
+
+	if (enable_mask) {
+		tmp |= AUDIO_ENABLED;
+		if (enable_mask & 1)
+			tmp |= PIN0_AUDIO_ENABLED;
+		if (enable_mask & 2)
+			tmp |= PIN1_AUDIO_ENABLED;
+		if (enable_mask & 4)
+			tmp |= PIN2_AUDIO_ENABLED;
+		if (enable_mask & 8)
+			tmp |= PIN3_AUDIO_ENABLED;
+	} else {
+		tmp &= ~(AUDIO_ENABLED |
+			 PIN0_AUDIO_ENABLED |
+			 PIN1_AUDIO_ENABLED |
+			 PIN2_AUDIO_ENABLED |
+			 PIN3_AUDIO_ENABLED);
+	}
+
+	WREG32(AZ_HOT_PLUG_CONTROL, tmp);
+}
+
 /*
  * update the N and CTS parameters for a given pixel clock rate
  */
@@ -318,10 +349,10 @@ void evergreen_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode
 	/* disable audio prior to setting up hw */
 	if (ASIC_IS_DCE6(rdev)) {
 		dig->afmt->pin = dce6_audio_get_pin(rdev);
-		dce6_audio_enable(rdev, dig->afmt->pin, false);
+		dce6_audio_enable(rdev, dig->afmt->pin, 0);
 	} else {
 		dig->afmt->pin = r600_audio_get_pin(rdev);
-		r600_audio_enable(rdev, dig->afmt->pin, false);
+		dce4_audio_enable(rdev, dig->afmt->pin, 0);
 	}
 
 	evergreen_audio_set_dto(encoder, mode->clock);
@@ -463,13 +494,15 @@ void evergreen_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode
 
 	/* enable audio after to setting up hw */
 	if (ASIC_IS_DCE6(rdev))
-		dce6_audio_enable(rdev, dig->afmt->pin, true);
+		dce6_audio_enable(rdev, dig->afmt->pin, 1);
 	else
-		r600_audio_enable(rdev, dig->afmt->pin, true);
+		dce4_audio_enable(rdev, dig->afmt->pin, 0xf);
 }
 
 void evergreen_hdmi_enable(struct drm_encoder *encoder, bool enable)
 {
+	struct drm_device *dev = encoder->dev;
+	struct radeon_device *rdev = dev->dev_private;
 	struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder);
 	struct radeon_encoder_atom_dig *dig = radeon_encoder->enc_priv;
 
@@ -482,6 +515,14 @@ void evergreen_hdmi_enable(struct drm_encoder *encoder, bool enable)
 	if (!enable && !dig->afmt->enabled)
 		return;
 
+	if (!enable && dig->afmt->pin) {
+		if (ASIC_IS_DCE6(rdev))
+			dce6_audio_enable(rdev, dig->afmt->pin, 0);
+		else
+			dce4_audio_enable(rdev, dig->afmt->pin, 0);
+		dig->afmt->pin = NULL;
+	}
+
 	dig->afmt->enabled = enable;
 
 	DRM_DEBUG("%sabling HDMI interface @ 0x%04X for encoder 0x%x\n",
diff --git a/drivers/gpu/drm/radeon/kv_dpm.c b/drivers/gpu/drm/radeon/kv_dpm.c
index 67cb472d188c..1dd976f447fa 100644
--- a/drivers/gpu/drm/radeon/kv_dpm.c
+++ b/drivers/gpu/drm/radeon/kv_dpm.c
@@ -2787,6 +2787,8 @@ void kv_dpm_debugfs_print_current_performance_level(struct radeon_device *rdev,
 		tmp = (RREG32_SMC(SMU_VOLTAGE_STATUS) & SMU_VOLTAGE_CURRENT_LEVEL_MASK) >>
 			SMU_VOLTAGE_CURRENT_LEVEL_SHIFT;
 		vddc = kv_convert_8bit_index_to_voltage(rdev, (u16)tmp);
+		seq_printf(m, "uvd    %sabled\n", pi->uvd_power_gated ? "dis" : "en");
+		seq_printf(m, "vce    %sabled\n", pi->vce_power_gated ? "dis" : "en");
 		seq_printf(m, "power level %d    sclk: %u vddc: %u\n",
 			   current_index, sclk, vddc);
 	}
diff --git a/drivers/gpu/drm/radeon/ni_dpm.c b/drivers/gpu/drm/radeon/ni_dpm.c
index 01fc4888e6fe..715b181c6243 100644
--- a/drivers/gpu/drm/radeon/ni_dpm.c
+++ b/drivers/gpu/drm/radeon/ni_dpm.c
@@ -789,7 +789,6 @@ static void ni_apply_state_adjust_rules(struct radeon_device *rdev,
 	bool disable_mclk_switching;
 	u32 mclk;
 	u16 vddci;
-	u32 max_sclk_vddc, max_mclk_vddci, max_mclk_vddc;
 	int i;
 
 	if ((rdev->pm.dpm.new_active_crtc_count > 1) ||
@@ -816,29 +815,6 @@ static void ni_apply_state_adjust_rules(struct radeon_device *rdev,
 		}
 	}
 
-	/* limit clocks to max supported clocks based on voltage dependency tables */
-	btc_get_max_clock_from_voltage_dependency_table(&rdev->pm.dpm.dyn_state.vddc_dependency_on_sclk,
-							&max_sclk_vddc);
-	btc_get_max_clock_from_voltage_dependency_table(&rdev->pm.dpm.dyn_state.vddci_dependency_on_mclk,
-							&max_mclk_vddci);
-	btc_get_max_clock_from_voltage_dependency_table(&rdev->pm.dpm.dyn_state.vddc_dependency_on_mclk,
-							&max_mclk_vddc);
-
-	for (i = 0; i < ps->performance_level_count; i++) {
-		if (max_sclk_vddc) {
-			if (ps->performance_levels[i].sclk > max_sclk_vddc)
-				ps->performance_levels[i].sclk = max_sclk_vddc;
-		}
-		if (max_mclk_vddci) {
-			if (ps->performance_levels[i].mclk > max_mclk_vddci)
-				ps->performance_levels[i].mclk = max_mclk_vddci;
-		}
-		if (max_mclk_vddc) {
-			if (ps->performance_levels[i].mclk > max_mclk_vddc)
-				ps->performance_levels[i].mclk = max_mclk_vddc;
-		}
-	}
-
 	/* XXX validate the min clocks required for display */
 
 	/* adjust low state */
diff --git a/drivers/gpu/drm/radeon/r100.c b/drivers/gpu/drm/radeon/r100.c
index b0098e792e62..10f8be0ee173 100644
--- a/drivers/gpu/drm/radeon/r100.c
+++ b/drivers/gpu/drm/radeon/r100.c
@@ -869,13 +869,14 @@ bool r100_semaphore_ring_emit(struct radeon_device *rdev,
 	return false;
 }
 
-int r100_copy_blit(struct radeon_device *rdev,
-		   uint64_t src_offset,
-		   uint64_t dst_offset,
-		   unsigned num_gpu_pages,
-		   struct radeon_fence **fence)
+struct radeon_fence *r100_copy_blit(struct radeon_device *rdev,
+				    uint64_t src_offset,
+				    uint64_t dst_offset,
+				    unsigned num_gpu_pages,
+				    struct reservation_object *resv)
 {
 	struct radeon_ring *ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX];
+	struct radeon_fence *fence;
 	uint32_t cur_pages;
 	uint32_t stride_bytes = RADEON_GPU_PAGE_SIZE;
 	uint32_t pitch;
@@ -896,7 +897,7 @@ int r100_copy_blit(struct radeon_device *rdev,
 	r = radeon_ring_lock(rdev, ring, ndw);
 	if (r) {
 		DRM_ERROR("radeon: moving bo (%d) asking for %u dw.\n", r, ndw);
-		return -EINVAL;
+		return ERR_PTR(-EINVAL);
 	}
 	while (num_gpu_pages > 0) {
 		cur_pages = num_gpu_pages;
@@ -936,11 +937,13 @@ int r100_copy_blit(struct radeon_device *rdev,
 			  RADEON_WAIT_2D_IDLECLEAN |
 			  RADEON_WAIT_HOST_IDLECLEAN |
 			  RADEON_WAIT_DMA_GUI_IDLE);
-	if (fence) {
-		r = radeon_fence_emit(rdev, fence, RADEON_RING_TYPE_GFX_INDEX);
+	r = radeon_fence_emit(rdev, &fence, RADEON_RING_TYPE_GFX_INDEX);
+	if (r) {
+		radeon_ring_unlock_undo(rdev, ring);
+		return ERR_PTR(r);
 	}
 	radeon_ring_unlock_commit(rdev, ring, false);
-	return r;
+	return fence;
 }
 
 static int r100_cp_wait_for_idle(struct radeon_device *rdev)
diff --git a/drivers/gpu/drm/radeon/r200.c b/drivers/gpu/drm/radeon/r200.c
index 67780374a652..732d4938aab7 100644
--- a/drivers/gpu/drm/radeon/r200.c
+++ b/drivers/gpu/drm/radeon/r200.c
@@ -80,13 +80,14 @@ static int r200_get_vtx_size_0(uint32_t vtx_fmt_0)
 	return vtx_size;
 }
 
-int r200_copy_dma(struct radeon_device *rdev,
-		  uint64_t src_offset,
-		  uint64_t dst_offset,
-		  unsigned num_gpu_pages,
-		  struct radeon_fence **fence)
+struct radeon_fence *r200_copy_dma(struct radeon_device *rdev,
+				   uint64_t src_offset,
+				   uint64_t dst_offset,
+				   unsigned num_gpu_pages,
+				   struct reservation_object *resv)
 {
 	struct radeon_ring *ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX];
+	struct radeon_fence *fence;
 	uint32_t size;
 	uint32_t cur_size;
 	int i, num_loops;
@@ -98,7 +99,7 @@ int r200_copy_dma(struct radeon_device *rdev,
 	r = radeon_ring_lock(rdev, ring, num_loops * 4 + 64);
 	if (r) {
 		DRM_ERROR("radeon: moving bo (%d).\n", r);
-		return r;
+		return ERR_PTR(r);
 	}
 	/* Must wait for 2D idle & clean before DMA or hangs might happen */
 	radeon_ring_write(ring, PACKET0(RADEON_WAIT_UNTIL, 0));
@@ -118,11 +119,13 @@ int r200_copy_dma(struct radeon_device *rdev,
 	}
 	radeon_ring_write(ring, PACKET0(RADEON_WAIT_UNTIL, 0));
 	radeon_ring_write(ring, RADEON_WAIT_DMA_GUI_IDLE);
-	if (fence) {
-		r = radeon_fence_emit(rdev, fence, RADEON_RING_TYPE_GFX_INDEX);
+	r = radeon_fence_emit(rdev, &fence, RADEON_RING_TYPE_GFX_INDEX);
+	if (r) {
+		radeon_ring_unlock_undo(rdev, ring);
+		return ERR_PTR(r);
 	}
 	radeon_ring_unlock_commit(rdev, ring, false);
-	return r;
+	return fence;
 }
 
 
diff --git a/drivers/gpu/drm/radeon/r300_cmdbuf.c b/drivers/gpu/drm/radeon/r300_cmdbuf.c
index 84b1d5367a11..9418e388b045 100644
--- a/drivers/gpu/drm/radeon/r300_cmdbuf.c
+++ b/drivers/gpu/drm/radeon/r300_cmdbuf.c
@@ -34,10 +34,10 @@
  */
 
 #include <drm/drmP.h>
-#include <drm/drm_buffer.h>
 #include <drm/radeon_drm.h>
 #include "radeon_drv.h"
 #include "r300_reg.h"
+#include "drm_buffer.h"
 
 #include <asm/unaligned.h>
 
diff --git a/drivers/gpu/drm/radeon/r600.c b/drivers/gpu/drm/radeon/r600.c
index ea5c9af722ef..56b02927cd3d 100644
--- a/drivers/gpu/drm/radeon/r600.c
+++ b/drivers/gpu/drm/radeon/r600.c
@@ -122,6 +122,94 @@ u32 r600_get_xclk(struct radeon_device *rdev)
 
 int r600_set_uvd_clocks(struct radeon_device *rdev, u32 vclk, u32 dclk)
 {
+	unsigned fb_div = 0, ref_div, vclk_div = 0, dclk_div = 0;
+	int r;
+
+	/* bypass vclk and dclk with bclk */
+	WREG32_P(CG_UPLL_FUNC_CNTL_2,
+		 VCLK_SRC_SEL(1) | DCLK_SRC_SEL(1),
+		 ~(VCLK_SRC_SEL_MASK | DCLK_SRC_SEL_MASK));
+
+	/* assert BYPASS_EN, deassert UPLL_RESET, UPLL_SLEEP and UPLL_CTLREQ */
+	WREG32_P(CG_UPLL_FUNC_CNTL, UPLL_BYPASS_EN_MASK, ~(
+		 UPLL_RESET_MASK | UPLL_SLEEP_MASK | UPLL_CTLREQ_MASK));
+
+	if (rdev->family >= CHIP_RS780)
+		WREG32_P(GFX_MACRO_BYPASS_CNTL, UPLL_BYPASS_CNTL,
+			 ~UPLL_BYPASS_CNTL);
+
+	if (!vclk || !dclk) {
+		/* keep the Bypass mode, put PLL to sleep */
+		WREG32_P(CG_UPLL_FUNC_CNTL, UPLL_SLEEP_MASK, ~UPLL_SLEEP_MASK);
+		return 0;
+	}
+
+	if (rdev->clock.spll.reference_freq == 10000)
+		ref_div = 34;
+	else
+		ref_div = 4;
+
+	r = radeon_uvd_calc_upll_dividers(rdev, vclk, dclk, 50000, 160000,
+					  ref_div + 1, 0xFFF, 2, 30, ~0,
+					  &fb_div, &vclk_div, &dclk_div);
+	if (r)
+		return r;
+
+	if (rdev->family >= CHIP_RV670 && rdev->family < CHIP_RS780)
+		fb_div >>= 1;
+	else
+		fb_div |= 1;
+
+	r = radeon_uvd_send_upll_ctlreq(rdev, CG_UPLL_FUNC_CNTL);
+        if (r)
+                return r;
+
+	/* assert PLL_RESET */
+	WREG32_P(CG_UPLL_FUNC_CNTL, UPLL_RESET_MASK, ~UPLL_RESET_MASK);
+
+	/* For RS780 we have to choose ref clk */
+	if (rdev->family >= CHIP_RS780)
+		WREG32_P(CG_UPLL_FUNC_CNTL, UPLL_REFCLK_SRC_SEL_MASK,
+			 ~UPLL_REFCLK_SRC_SEL_MASK);
+
+	/* set the required fb, ref and post divder values */
+	WREG32_P(CG_UPLL_FUNC_CNTL,
+		 UPLL_FB_DIV(fb_div) |
+		 UPLL_REF_DIV(ref_div),
+		 ~(UPLL_FB_DIV_MASK | UPLL_REF_DIV_MASK));
+	WREG32_P(CG_UPLL_FUNC_CNTL_2,
+		 UPLL_SW_HILEN(vclk_div >> 1) |
+		 UPLL_SW_LOLEN((vclk_div >> 1) + (vclk_div & 1)) |
+		 UPLL_SW_HILEN2(dclk_div >> 1) |
+		 UPLL_SW_LOLEN2((dclk_div >> 1) + (dclk_div & 1)) |
+		 UPLL_DIVEN_MASK | UPLL_DIVEN2_MASK,
+		 ~UPLL_SW_MASK);
+
+	/* give the PLL some time to settle */
+	mdelay(15);
+
+	/* deassert PLL_RESET */
+	WREG32_P(CG_UPLL_FUNC_CNTL, 0, ~UPLL_RESET_MASK);
+
+	mdelay(15);
+
+	/* deassert BYPASS EN */
+	WREG32_P(CG_UPLL_FUNC_CNTL, 0, ~UPLL_BYPASS_EN_MASK);
+
+	if (rdev->family >= CHIP_RS780)
+		WREG32_P(GFX_MACRO_BYPASS_CNTL, 0, ~UPLL_BYPASS_CNTL);
+
+	r = radeon_uvd_send_upll_ctlreq(rdev, CG_UPLL_FUNC_CNTL);
+	if (r)
+		return r;
+
+	/* switch VCLK and DCLK selection */
+	WREG32_P(CG_UPLL_FUNC_CNTL_2,
+		 VCLK_SRC_SEL(2) | DCLK_SRC_SEL(2),
+		 ~(VCLK_SRC_SEL_MASK | DCLK_SRC_SEL_MASK));
+
+	mdelay(100);
+
 	return 0;
 }
 
@@ -992,6 +1080,8 @@ static int r600_pcie_gart_enable(struct radeon_device *rdev)
 	WREG32(MC_VM_L1_TLB_MCB_WR_GFX_CNTL, tmp);
 	WREG32(MC_VM_L1_TLB_MCB_RD_PDMA_CNTL, tmp);
 	WREG32(MC_VM_L1_TLB_MCB_WR_PDMA_CNTL, tmp);
+	WREG32(MC_VM_L1_TLB_MCB_RD_UVD_CNTL, tmp);
+	WREG32(MC_VM_L1_TLB_MCB_WR_UVD_CNTL, tmp);
 	WREG32(MC_VM_L1_TLB_MCB_RD_SEM_CNTL, tmp | ENABLE_SEMAPHORE_MODE);
 	WREG32(MC_VM_L1_TLB_MCB_WR_SEM_CNTL, tmp | ENABLE_SEMAPHORE_MODE);
 	WREG32(VM_CONTEXT0_PAGE_TABLE_START_ADDR, rdev->mc.gtt_start >> 12);
@@ -1042,6 +1132,8 @@ static void r600_pcie_gart_disable(struct radeon_device *rdev)
 	WREG32(MC_VM_L1_TLB_MCB_WR_SYS_CNTL, tmp);
 	WREG32(MC_VM_L1_TLB_MCB_RD_HDP_CNTL, tmp);
 	WREG32(MC_VM_L1_TLB_MCB_WR_HDP_CNTL, tmp);
+	WREG32(MC_VM_L1_TLB_MCB_RD_UVD_CNTL, tmp);
+	WREG32(MC_VM_L1_TLB_MCB_WR_UVD_CNTL, tmp);
 	radeon_gart_table_vram_unpin(rdev);
 }
 
@@ -1338,7 +1430,7 @@ int r600_vram_scratch_init(struct radeon_device *rdev)
 	if (rdev->vram_scratch.robj == NULL) {
 		r = radeon_bo_create(rdev, RADEON_GPU_PAGE_SIZE,
 				     PAGE_SIZE, true, RADEON_GEM_DOMAIN_VRAM,
-				     0, NULL, &rdev->vram_scratch.robj);
+				     0, NULL, NULL, &rdev->vram_scratch.robj);
 		if (r) {
 			return r;
 		}
@@ -2792,12 +2884,13 @@ bool r600_semaphore_ring_emit(struct radeon_device *rdev,
  * Used by the radeon ttm implementation to move pages if
  * registered as the asic copy callback.
  */
-int r600_copy_cpdma(struct radeon_device *rdev,
-		    uint64_t src_offset, uint64_t dst_offset,
-		    unsigned num_gpu_pages,
-		    struct radeon_fence **fence)
+struct radeon_fence *r600_copy_cpdma(struct radeon_device *rdev,
+				     uint64_t src_offset, uint64_t dst_offset,
+				     unsigned num_gpu_pages,
+				     struct reservation_object *resv)
 {
 	struct radeon_semaphore *sem = NULL;
+	struct radeon_fence *fence;
 	int ring_index = rdev->asic->copy.blit_ring_index;
 	struct radeon_ring *ring = &rdev->ring[ring_index];
 	u32 size_in_bytes, cur_size_in_bytes, tmp;
@@ -2807,7 +2900,7 @@ int r600_copy_cpdma(struct radeon_device *rdev,
 	r = radeon_semaphore_create(rdev, &sem);
 	if (r) {
 		DRM_ERROR("radeon: moving bo (%d).\n", r);
-		return r;
+		return ERR_PTR(r);
 	}
 
 	size_in_bytes = (num_gpu_pages << RADEON_GPU_PAGE_SHIFT);
@@ -2816,10 +2909,10 @@ int r600_copy_cpdma(struct radeon_device *rdev,
 	if (r) {
 		DRM_ERROR("radeon: moving bo (%d).\n", r);
 		radeon_semaphore_free(rdev, &sem, NULL);
-		return r;
+		return ERR_PTR(r);
 	}
 
-	radeon_semaphore_sync_to(sem, *fence);
+	radeon_semaphore_sync_resv(rdev, sem, resv, false);
 	radeon_semaphore_sync_rings(rdev, sem, ring->idx);
 
 	radeon_ring_write(ring, PACKET3(PACKET3_SET_CONFIG_REG, 1));
@@ -2846,17 +2939,17 @@ int r600_copy_cpdma(struct radeon_device *rdev,
 	radeon_ring_write(ring, (WAIT_UNTIL - PACKET3_SET_CONFIG_REG_OFFSET) >> 2);
 	radeon_ring_write(ring, WAIT_CP_DMA_IDLE_bit);
 
-	r = radeon_fence_emit(rdev, fence, ring->idx);
+	r = radeon_fence_emit(rdev, &fence, ring->idx);
 	if (r) {
 		radeon_ring_unlock_undo(rdev, ring);
 		radeon_semaphore_free(rdev, &sem, NULL);
-		return r;
+		return ERR_PTR(r);
 	}
 
 	radeon_ring_unlock_commit(rdev, ring, false);
-	radeon_semaphore_free(rdev, &sem, *fence);
+	radeon_semaphore_free(rdev, &sem, fence);
 
-	return r;
+	return fence;
 }
 
 int r600_set_surface_reg(struct radeon_device *rdev, int reg,
@@ -2907,6 +3000,18 @@ static int r600_startup(struct radeon_device *rdev)
 		return r;
 	}
 
+	if (rdev->has_uvd) {
+		r = uvd_v1_0_resume(rdev);
+		if (!r) {
+			r = radeon_fence_driver_start_ring(rdev, R600_RING_TYPE_UVD_INDEX);
+			if (r) {
+				dev_err(rdev->dev, "failed initializing UVD fences (%d).\n", r);
+			}
+		}
+		if (r)
+			rdev->ring[R600_RING_TYPE_UVD_INDEX].ring_size = 0;
+	}
+
 	/* Enable IRQ */
 	if (!rdev->irq.installed) {
 		r = radeon_irq_kms_init(rdev);
@@ -2935,6 +3040,18 @@ static int r600_startup(struct radeon_device *rdev)
 	if (r)
 		return r;
 
+	if (rdev->has_uvd) {
+		ring = &rdev->ring[R600_RING_TYPE_UVD_INDEX];
+		if (ring->ring_size) {
+			r = radeon_ring_init(rdev, ring, ring->ring_size, 0,
+					     RADEON_CP_PACKET2);
+			if (!r)
+				r = uvd_v1_0_init(rdev);
+			if (r)
+				DRM_ERROR("radeon: failed initializing UVD (%d).\n", r);
+		}
+	}
+
 	r = radeon_ib_pool_init(rdev);
 	if (r) {
 		dev_err(rdev->dev, "IB initialization failed (%d).\n", r);
@@ -2994,6 +3111,10 @@ int r600_suspend(struct radeon_device *rdev)
 	radeon_pm_suspend(rdev);
 	r600_audio_fini(rdev);
 	r600_cp_stop(rdev);
+	if (rdev->has_uvd) {
+		uvd_v1_0_fini(rdev);
+		radeon_uvd_suspend(rdev);
+	}
 	r600_irq_suspend(rdev);
 	radeon_wb_disable(rdev);
 	r600_pcie_gart_disable(rdev);
@@ -3073,6 +3194,14 @@ int r600_init(struct radeon_device *rdev)
 	rdev->ring[RADEON_RING_TYPE_GFX_INDEX].ring_obj = NULL;
 	r600_ring_init(rdev, &rdev->ring[RADEON_RING_TYPE_GFX_INDEX], 1024 * 1024);
 
+	if (rdev->has_uvd) {
+		r = radeon_uvd_init(rdev);
+		if (!r) {
+			rdev->ring[R600_RING_TYPE_UVD_INDEX].ring_obj = NULL;
+			r600_ring_init(rdev, &rdev->ring[R600_RING_TYPE_UVD_INDEX], 4096);
+		}
+	}
+
 	rdev->ih.ring_obj = NULL;
 	r600_ih_ring_init(rdev, 64 * 1024);
 
@@ -3102,6 +3231,10 @@ void r600_fini(struct radeon_device *rdev)
 	r600_audio_fini(rdev);
 	r600_cp_fini(rdev);
 	r600_irq_fini(rdev);
+	if (rdev->has_uvd) {
+		uvd_v1_0_fini(rdev);
+		radeon_uvd_fini(rdev);
+	}
 	radeon_wb_fini(rdev);
 	radeon_ib_pool_fini(rdev);
 	radeon_irq_kms_fini(rdev);
@@ -3235,7 +3368,7 @@ int r600_ih_ring_alloc(struct radeon_device *rdev)
 		r = radeon_bo_create(rdev, rdev->ih.ring_size,
 				     PAGE_SIZE, true,
 				     RADEON_GEM_DOMAIN_GTT, 0,
-				     NULL, &rdev->ih.ring_obj);
+				     NULL, NULL, &rdev->ih.ring_obj);
 		if (r) {
 			DRM_ERROR("radeon: failed to create ih ring buffer (%d).\n", r);
 			return r;
diff --git a/drivers/gpu/drm/radeon/r600_audio.c b/drivers/gpu/drm/radeon/r600_audio.c
deleted file mode 100644
index bffac10c4296..000000000000
--- a/drivers/gpu/drm/radeon/r600_audio.c
+++ /dev/null
@@ -1,207 +0,0 @@
-/*
- * Copyright 2008 Advanced Micro Devices, Inc.
- * Copyright 2008 Red Hat Inc.
- * Copyright 2009 Christian König.
- *
- * 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: Christian König
- */
-#include <drm/drmP.h>
-#include "radeon.h"
-#include "radeon_reg.h"
-#include "radeon_asic.h"
-#include "atom.h"
-
-/*
- * check if enc_priv stores radeon_encoder_atom_dig
- */
-static bool radeon_dig_encoder(struct drm_encoder *encoder)
-{
-	struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder);
-	switch (radeon_encoder->encoder_id) {
-	case ENCODER_OBJECT_ID_INTERNAL_LVDS:
-	case ENCODER_OBJECT_ID_INTERNAL_TMDS1:
-	case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_TMDS1:
-	case ENCODER_OBJECT_ID_INTERNAL_LVTM1:
-	case ENCODER_OBJECT_ID_INTERNAL_DVO1:
-	case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_DVO1:
-	case ENCODER_OBJECT_ID_INTERNAL_DDI:
-	case ENCODER_OBJECT_ID_INTERNAL_UNIPHY:
-	case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_LVTMA:
-	case ENCODER_OBJECT_ID_INTERNAL_UNIPHY1:
-	case ENCODER_OBJECT_ID_INTERNAL_UNIPHY2:
-		return true;
-	}
-	return false;
-}
-
-/*
- * check if the chipset is supported
- */
-static int r600_audio_chipset_supported(struct radeon_device *rdev)
-{
-	return ASIC_IS_DCE2(rdev) && !ASIC_IS_NODCE(rdev);
-}
-
-struct r600_audio_pin r600_audio_status(struct radeon_device *rdev)
-{
-	struct r600_audio_pin status;
-	uint32_t value;
-
-	value = RREG32(R600_AUDIO_RATE_BPS_CHANNEL);
-
-	/* number of channels */
-	status.channels = (value & 0x7) + 1;
-
-	/* bits per sample */
-	switch ((value & 0xF0) >> 4) {
-	case 0x0:
-		status.bits_per_sample = 8;
-		break;
-	case 0x1:
-		status.bits_per_sample = 16;
-		break;
-	case 0x2:
-		status.bits_per_sample = 20;
-		break;
-	case 0x3:
-		status.bits_per_sample = 24;
-		break;
-	case 0x4:
-		status.bits_per_sample = 32;
-		break;
-	default:
-		dev_err(rdev->dev, "Unknown bits per sample 0x%x, using 16\n",
-			(int)value);
-		status.bits_per_sample = 16;
-	}
-
-	/* current sampling rate in HZ */
-	if (value & 0x4000)
-		status.rate = 44100;
-	else
-		status.rate = 48000;
-	status.rate *= ((value >> 11) & 0x7) + 1;
-	status.rate /= ((value >> 8) & 0x7) + 1;
-
-	value = RREG32(R600_AUDIO_STATUS_BITS);
-
-	/* iec 60958 status bits */
-	status.status_bits = value & 0xff;
-
-	/* iec 60958 category code */
-	status.category_code = (value >> 8) & 0xff;
-
-	return status;
-}
-
-/*
- * update all hdmi interfaces with current audio parameters
- */
-void r600_audio_update_hdmi(struct work_struct *work)
-{
-	struct radeon_device *rdev = container_of(work, struct radeon_device,
-						  audio_work);
-	struct drm_device *dev = rdev->ddev;
-	struct r600_audio_pin audio_status = r600_audio_status(rdev);
-	struct drm_encoder *encoder;
-	bool changed = false;
-
-	if (rdev->audio.pin[0].channels != audio_status.channels ||
-	    rdev->audio.pin[0].rate != audio_status.rate ||
-	    rdev->audio.pin[0].bits_per_sample != audio_status.bits_per_sample ||
-	    rdev->audio.pin[0].status_bits != audio_status.status_bits ||
-	    rdev->audio.pin[0].category_code != audio_status.category_code) {
-		rdev->audio.pin[0] = audio_status;
-		changed = true;
-	}
-
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) {
-		if (!radeon_dig_encoder(encoder))
-			continue;
-		if (changed || r600_hdmi_buffer_status_changed(encoder))
-			r600_hdmi_update_audio_settings(encoder);
-	}
-}
-
-/* enable the audio stream */
-void r600_audio_enable(struct radeon_device *rdev,
-		       struct r600_audio_pin *pin,
-		       bool enable)
-{
-	u32 value = 0;
-
-	if (!pin)
-		return;
-
-	if (ASIC_IS_DCE4(rdev)) {
-		if (enable) {
-			value |= 0x81000000; /* Required to enable audio */
-			value |= 0x0e1000f0; /* fglrx sets that too */
-		}
-		WREG32(EVERGREEN_AUDIO_ENABLE, value);
-	} else {
-		WREG32_P(R600_AUDIO_ENABLE,
-			 enable ? 0x81000000 : 0x0, ~0x81000000);
-	}
-}
-
-/*
- * initialize the audio vars
- */
-int r600_audio_init(struct radeon_device *rdev)
-{
-	if (!radeon_audio || !r600_audio_chipset_supported(rdev))
-		return 0;
-
-	rdev->audio.enabled = true;
-
-	rdev->audio.num_pins = 1;
-	rdev->audio.pin[0].channels = -1;
-	rdev->audio.pin[0].rate = -1;
-	rdev->audio.pin[0].bits_per_sample = -1;
-	rdev->audio.pin[0].status_bits = 0;
-	rdev->audio.pin[0].category_code = 0;
-	rdev->audio.pin[0].id = 0;
-	/* disable audio.  it will be set up later */
-	r600_audio_enable(rdev, &rdev->audio.pin[0], false);
-
-	return 0;
-}
-
-/*
- * release the audio timer
- * TODO: How to do this correctly on SMP systems?
- */
-void r600_audio_fini(struct radeon_device *rdev)
-{
-	if (!rdev->audio.enabled)
-		return;
-
-	r600_audio_enable(rdev, &rdev->audio.pin[0], false);
-
-	rdev->audio.enabled = false;
-}
-
-struct r600_audio_pin *r600_audio_get_pin(struct radeon_device *rdev)
-{
-	/* only one pin on 6xx-NI */
-	return &rdev->audio.pin[0];
-}
diff --git a/drivers/gpu/drm/radeon/r600_cp.c b/drivers/gpu/drm/radeon/r600_cp.c
index 8c9b7e26533c..09e3f39925fa 100644
--- a/drivers/gpu/drm/radeon/r600_cp.c
+++ b/drivers/gpu/drm/radeon/r600_cp.c
@@ -1949,15 +1949,15 @@ int r600_do_cleanup_cp(struct drm_device *dev)
 #if __OS_HAS_AGP
 	if (dev_priv->flags & RADEON_IS_AGP) {
 		if (dev_priv->cp_ring != NULL) {
-			drm_core_ioremapfree(dev_priv->cp_ring, dev);
+			drm_legacy_ioremapfree(dev_priv->cp_ring, dev);
 			dev_priv->cp_ring = NULL;
 		}
 		if (dev_priv->ring_rptr != NULL) {
-			drm_core_ioremapfree(dev_priv->ring_rptr, dev);
+			drm_legacy_ioremapfree(dev_priv->ring_rptr, dev);
 			dev_priv->ring_rptr = NULL;
 		}
 		if (dev->agp_buffer_map != NULL) {
-			drm_core_ioremapfree(dev->agp_buffer_map, dev);
+			drm_legacy_ioremapfree(dev->agp_buffer_map, dev);
 			dev->agp_buffer_map = NULL;
 		}
 	} else
@@ -1968,7 +1968,7 @@ int r600_do_cleanup_cp(struct drm_device *dev)
 			r600_page_table_cleanup(dev, &dev_priv->gart_info);
 
 		if (dev_priv->gart_info.gart_table_location == DRM_ATI_GART_FB) {
-			drm_core_ioremapfree(&dev_priv->gart_info.mapping, dev);
+			drm_legacy_ioremapfree(&dev_priv->gart_info.mapping, dev);
 			dev_priv->gart_info.addr = NULL;
 		}
 	}
@@ -2052,27 +2052,27 @@ int r600_do_init_cp(struct drm_device *dev, drm_radeon_init_t *init,
 	dev_priv->buffers_offset = init->buffers_offset;
 	dev_priv->gart_textures_offset = init->gart_textures_offset;
 
-	master_priv->sarea = drm_getsarea(dev);
+	master_priv->sarea = drm_legacy_getsarea(dev);
 	if (!master_priv->sarea) {
 		DRM_ERROR("could not find sarea!\n");
 		r600_do_cleanup_cp(dev);
 		return -EINVAL;
 	}
 
-	dev_priv->cp_ring = drm_core_findmap(dev, init->ring_offset);
+	dev_priv->cp_ring = drm_legacy_findmap(dev, init->ring_offset);
 	if (!dev_priv->cp_ring) {
 		DRM_ERROR("could not find cp ring region!\n");
 		r600_do_cleanup_cp(dev);
 		return -EINVAL;
 	}
-	dev_priv->ring_rptr = drm_core_findmap(dev, init->ring_rptr_offset);
+	dev_priv->ring_rptr = drm_legacy_findmap(dev, init->ring_rptr_offset);
 	if (!dev_priv->ring_rptr) {
 		DRM_ERROR("could not find ring read pointer!\n");
 		r600_do_cleanup_cp(dev);
 		return -EINVAL;
 	}
 	dev->agp_buffer_token = init->buffers_offset;
-	dev->agp_buffer_map = drm_core_findmap(dev, init->buffers_offset);
+	dev->agp_buffer_map = drm_legacy_findmap(dev, init->buffers_offset);
 	if (!dev->agp_buffer_map) {
 		DRM_ERROR("could not find dma buffer region!\n");
 		r600_do_cleanup_cp(dev);
@@ -2081,7 +2081,7 @@ int r600_do_init_cp(struct drm_device *dev, drm_radeon_init_t *init,
 
 	if (init->gart_textures_offset) {
 		dev_priv->gart_textures =
-		    drm_core_findmap(dev, init->gart_textures_offset);
+		    drm_legacy_findmap(dev, init->gart_textures_offset);
 		if (!dev_priv->gart_textures) {
 			DRM_ERROR("could not find GART texture region!\n");
 			r600_do_cleanup_cp(dev);
@@ -2092,9 +2092,9 @@ int r600_do_init_cp(struct drm_device *dev, drm_radeon_init_t *init,
 #if __OS_HAS_AGP
 	/* XXX */
 	if (dev_priv->flags & RADEON_IS_AGP) {
-		drm_core_ioremap_wc(dev_priv->cp_ring, dev);
-		drm_core_ioremap_wc(dev_priv->ring_rptr, dev);
-		drm_core_ioremap_wc(dev->agp_buffer_map, dev);
+		drm_legacy_ioremap_wc(dev_priv->cp_ring, dev);
+		drm_legacy_ioremap_wc(dev_priv->ring_rptr, dev);
+		drm_legacy_ioremap_wc(dev->agp_buffer_map, dev);
 		if (!dev_priv->cp_ring->handle ||
 		    !dev_priv->ring_rptr->handle ||
 		    !dev->agp_buffer_map->handle) {
@@ -2235,7 +2235,7 @@ int r600_do_init_cp(struct drm_device *dev, drm_radeon_init_t *init,
 		dev_priv->gart_info.mapping.size =
 			dev_priv->gart_info.table_size;
 
-		drm_core_ioremap_wc(&dev_priv->gart_info.mapping, dev);
+		drm_legacy_ioremap_wc(&dev_priv->gart_info.mapping, dev);
 		if (!dev_priv->gart_info.mapping.handle) {
 			DRM_ERROR("ioremap failed.\n");
 			r600_do_cleanup_cp(dev);
diff --git a/drivers/gpu/drm/radeon/r600_dma.c b/drivers/gpu/drm/radeon/r600_dma.c
index a908daa006d2..100189ec5fa8 100644
--- a/drivers/gpu/drm/radeon/r600_dma.c
+++ b/drivers/gpu/drm/radeon/r600_dma.c
@@ -427,18 +427,19 @@ void r600_dma_ring_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib)
  * @src_offset: src GPU address
  * @dst_offset: dst GPU address
  * @num_gpu_pages: number of GPU pages to xfer
- * @fence: radeon fence object
+ * @resv: reservation object to sync to
  *
  * Copy GPU paging using the DMA engine (r6xx).
  * Used by the radeon ttm implementation to move pages if
  * registered as the asic copy callback.
  */
-int r600_copy_dma(struct radeon_device *rdev,
-		  uint64_t src_offset, uint64_t dst_offset,
-		  unsigned num_gpu_pages,
-		  struct radeon_fence **fence)
+struct radeon_fence *r600_copy_dma(struct radeon_device *rdev,
+				   uint64_t src_offset, uint64_t dst_offset,
+				   unsigned num_gpu_pages,
+				   struct reservation_object *resv)
 {
 	struct radeon_semaphore *sem = NULL;
+	struct radeon_fence *fence;
 	int ring_index = rdev->asic->copy.dma_ring_index;
 	struct radeon_ring *ring = &rdev->ring[ring_index];
 	u32 size_in_dw, cur_size_in_dw;
@@ -448,7 +449,7 @@ int r600_copy_dma(struct radeon_device *rdev,
 	r = radeon_semaphore_create(rdev, &sem);
 	if (r) {
 		DRM_ERROR("radeon: moving bo (%d).\n", r);
-		return r;
+		return ERR_PTR(r);
 	}
 
 	size_in_dw = (num_gpu_pages << RADEON_GPU_PAGE_SHIFT) / 4;
@@ -457,10 +458,10 @@ int r600_copy_dma(struct radeon_device *rdev,
 	if (r) {
 		DRM_ERROR("radeon: moving bo (%d).\n", r);
 		radeon_semaphore_free(rdev, &sem, NULL);
-		return r;
+		return ERR_PTR(r);
 	}
 
-	radeon_semaphore_sync_to(sem, *fence);
+	radeon_semaphore_sync_resv(rdev, sem, resv, false);
 	radeon_semaphore_sync_rings(rdev, sem, ring->idx);
 
 	for (i = 0; i < num_loops; i++) {
@@ -477,15 +478,15 @@ int r600_copy_dma(struct radeon_device *rdev,
 		dst_offset += cur_size_in_dw * 4;
 	}
 
-	r = radeon_fence_emit(rdev, fence, ring->idx);
+	r = radeon_fence_emit(rdev, &fence, ring->idx);
 	if (r) {
 		radeon_ring_unlock_undo(rdev, ring);
 		radeon_semaphore_free(rdev, &sem, NULL);
-		return r;
+		return ERR_PTR(r);
 	}
 
 	radeon_ring_unlock_commit(rdev, ring, false);
-	radeon_semaphore_free(rdev, &sem, *fence);
+	radeon_semaphore_free(rdev, &sem, fence);
 
-	return r;
+	return fence;
 }
diff --git a/drivers/gpu/drm/radeon/r600_hdmi.c b/drivers/gpu/drm/radeon/r600_hdmi.c
index 26ef8ced6f89..b90dc0eb08e6 100644
--- a/drivers/gpu/drm/radeon/r600_hdmi.c
+++ b/drivers/gpu/drm/radeon/r600_hdmi.c
@@ -72,6 +72,169 @@ static const struct radeon_hdmi_acr r600_hdmi_predefined_acr[] = {
 
 
 /*
+ * check if the chipset is supported
+ */
+static int r600_audio_chipset_supported(struct radeon_device *rdev)
+{
+	return ASIC_IS_DCE2(rdev) && !ASIC_IS_NODCE(rdev);
+}
+
+static struct r600_audio_pin r600_audio_status(struct radeon_device *rdev)
+{
+	struct r600_audio_pin status;
+	uint32_t value;
+
+	value = RREG32(R600_AUDIO_RATE_BPS_CHANNEL);
+
+	/* number of channels */
+	status.channels = (value & 0x7) + 1;
+
+	/* bits per sample */
+	switch ((value & 0xF0) >> 4) {
+	case 0x0:
+		status.bits_per_sample = 8;
+		break;
+	case 0x1:
+		status.bits_per_sample = 16;
+		break;
+	case 0x2:
+		status.bits_per_sample = 20;
+		break;
+	case 0x3:
+		status.bits_per_sample = 24;
+		break;
+	case 0x4:
+		status.bits_per_sample = 32;
+		break;
+	default:
+		dev_err(rdev->dev, "Unknown bits per sample 0x%x, using 16\n",
+			(int)value);
+		status.bits_per_sample = 16;
+	}
+
+	/* current sampling rate in HZ */
+	if (value & 0x4000)
+		status.rate = 44100;
+	else
+		status.rate = 48000;
+	status.rate *= ((value >> 11) & 0x7) + 1;
+	status.rate /= ((value >> 8) & 0x7) + 1;
+
+	value = RREG32(R600_AUDIO_STATUS_BITS);
+
+	/* iec 60958 status bits */
+	status.status_bits = value & 0xff;
+
+	/* iec 60958 category code */
+	status.category_code = (value >> 8) & 0xff;
+
+	return status;
+}
+
+/*
+ * update all hdmi interfaces with current audio parameters
+ */
+void r600_audio_update_hdmi(struct work_struct *work)
+{
+	struct radeon_device *rdev = container_of(work, struct radeon_device,
+						  audio_work);
+	struct drm_device *dev = rdev->ddev;
+	struct r600_audio_pin audio_status = r600_audio_status(rdev);
+	struct drm_encoder *encoder;
+	bool changed = false;
+
+	if (rdev->audio.pin[0].channels != audio_status.channels ||
+	    rdev->audio.pin[0].rate != audio_status.rate ||
+	    rdev->audio.pin[0].bits_per_sample != audio_status.bits_per_sample ||
+	    rdev->audio.pin[0].status_bits != audio_status.status_bits ||
+	    rdev->audio.pin[0].category_code != audio_status.category_code) {
+		rdev->audio.pin[0] = audio_status;
+		changed = true;
+	}
+
+	list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) {
+		if (!radeon_encoder_is_digital(encoder))
+			continue;
+		if (changed || r600_hdmi_buffer_status_changed(encoder))
+			r600_hdmi_update_audio_settings(encoder);
+	}
+}
+
+/* enable the audio stream */
+void r600_audio_enable(struct radeon_device *rdev,
+		       struct r600_audio_pin *pin,
+		       u8 enable_mask)
+{
+	u32 tmp = RREG32(AZ_HOT_PLUG_CONTROL);
+
+	if (!pin)
+		return;
+
+	if (enable_mask) {
+		tmp |= AUDIO_ENABLED;
+		if (enable_mask & 1)
+			tmp |= PIN0_AUDIO_ENABLED;
+		if (enable_mask & 2)
+			tmp |= PIN1_AUDIO_ENABLED;
+		if (enable_mask & 4)
+			tmp |= PIN2_AUDIO_ENABLED;
+		if (enable_mask & 8)
+			tmp |= PIN3_AUDIO_ENABLED;
+	} else {
+		tmp &= ~(AUDIO_ENABLED |
+			 PIN0_AUDIO_ENABLED |
+			 PIN1_AUDIO_ENABLED |
+			 PIN2_AUDIO_ENABLED |
+			 PIN3_AUDIO_ENABLED);
+	}
+
+	WREG32(AZ_HOT_PLUG_CONTROL, tmp);
+}
+
+/*
+ * initialize the audio vars
+ */
+int r600_audio_init(struct radeon_device *rdev)
+{
+	if (!radeon_audio || !r600_audio_chipset_supported(rdev))
+		return 0;
+
+	rdev->audio.enabled = true;
+
+	rdev->audio.num_pins = 1;
+	rdev->audio.pin[0].channels = -1;
+	rdev->audio.pin[0].rate = -1;
+	rdev->audio.pin[0].bits_per_sample = -1;
+	rdev->audio.pin[0].status_bits = 0;
+	rdev->audio.pin[0].category_code = 0;
+	rdev->audio.pin[0].id = 0;
+	/* disable audio.  it will be set up later */
+	r600_audio_enable(rdev, &rdev->audio.pin[0], 0);
+
+	return 0;
+}
+
+/*
+ * release the audio timer
+ * TODO: How to do this correctly on SMP systems?
+ */
+void r600_audio_fini(struct radeon_device *rdev)
+{
+	if (!rdev->audio.enabled)
+		return;
+
+	r600_audio_enable(rdev, &rdev->audio.pin[0], 0);
+
+	rdev->audio.enabled = false;
+}
+
+struct r600_audio_pin *r600_audio_get_pin(struct radeon_device *rdev)
+{
+	/* only one pin on 6xx-NI */
+	return &rdev->audio.pin[0];
+}
+
+/*
  * calculate CTS and N values if they are not found in the table
  */
 static void r600_hdmi_calc_cts(uint32_t clock, int *CTS, int *N, int freq)
@@ -357,7 +520,7 @@ void r600_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode *mod
 
 	/* disable audio prior to setting up hw */
 	dig->afmt->pin = r600_audio_get_pin(rdev);
-	r600_audio_enable(rdev, dig->afmt->pin, false);
+	r600_audio_enable(rdev, dig->afmt->pin, 0xf);
 
 	r600_audio_set_dto(encoder, mode->clock);
 
@@ -443,7 +606,7 @@ void r600_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode *mod
 	WREG32(HDMI0_RAMP_CONTROL3 + offset, 0x00000001);
 
 	/* enable audio after to setting up hw */
-	r600_audio_enable(rdev, dig->afmt->pin, true);
+	r600_audio_enable(rdev, dig->afmt->pin, 0xf);
 }
 
 /**
@@ -528,6 +691,11 @@ void r600_hdmi_enable(struct drm_encoder *encoder, bool enable)
 	if (!enable && !dig->afmt->enabled)
 		return;
 
+	if (!enable && dig->afmt->pin) {
+		r600_audio_enable(rdev, dig->afmt->pin, 0);
+		dig->afmt->pin = NULL;
+	}
+
 	/* Older chipsets require setting HDMI and routing manually */
 	if (!ASIC_IS_DCE3(rdev)) {
 		if (enable)
diff --git a/drivers/gpu/drm/radeon/r600d.h b/drivers/gpu/drm/radeon/r600d.h
index 31e1052ad3e3..1e8495cca41e 100644
--- a/drivers/gpu/drm/radeon/r600d.h
+++ b/drivers/gpu/drm/radeon/r600d.h
@@ -323,11 +323,12 @@
 #define	HDP_TILING_CONFIG				0x2F3C
 #define HDP_DEBUG1                                      0x2F34
 
+#define MC_CONFIG					0x2000
 #define MC_VM_AGP_TOP					0x2184
 #define MC_VM_AGP_BOT					0x2188
 #define	MC_VM_AGP_BASE					0x218C
 #define MC_VM_FB_LOCATION				0x2180
-#define MC_VM_L1_TLB_MCD_RD_A_CNTL			0x219C
+#define MC_VM_L1_TLB_MCB_RD_UVD_CNTL			0x2124
 #define 	ENABLE_L1_TLB					(1 << 0)
 #define		ENABLE_L1_FRAGMENT_PROCESSING			(1 << 1)
 #define		ENABLE_L1_STRICT_ORDERING			(1 << 2)
@@ -347,12 +348,14 @@
 #define		EFFECTIVE_L1_QUEUE_SIZE(x)			(((x) & 7) << 15)
 #define		EFFECTIVE_L1_QUEUE_SIZE_MASK			0x00038000
 #define		EFFECTIVE_L1_QUEUE_SIZE_SHIFT			15
+#define MC_VM_L1_TLB_MCD_RD_A_CNTL			0x219C
 #define MC_VM_L1_TLB_MCD_RD_B_CNTL			0x21A0
 #define MC_VM_L1_TLB_MCB_RD_GFX_CNTL			0x21FC
 #define MC_VM_L1_TLB_MCB_RD_HDP_CNTL			0x2204
 #define MC_VM_L1_TLB_MCB_RD_PDMA_CNTL			0x2208
 #define MC_VM_L1_TLB_MCB_RD_SEM_CNTL			0x220C
 #define	MC_VM_L1_TLB_MCB_RD_SYS_CNTL			0x2200
+#define MC_VM_L1_TLB_MCB_WR_UVD_CNTL			0x212c
 #define MC_VM_L1_TLB_MCD_WR_A_CNTL			0x21A4
 #define MC_VM_L1_TLB_MCD_WR_B_CNTL			0x21A8
 #define MC_VM_L1_TLB_MCB_WR_GFX_CNTL			0x2210
@@ -366,6 +369,8 @@
 #define MC_VM_SYSTEM_APERTURE_HIGH_ADDR			0x2194
 #define MC_VM_SYSTEM_APERTURE_DEFAULT_ADDR		0x2198
 
+#define RS_DQ_RD_RET_CONF				0x2348
+
 #define	PA_CL_ENHANCE					0x8A14
 #define		CLIP_VTX_REORDER_ENA				(1 << 0)
 #define		NUM_CLIP_SEQ(x)					((x) << 1)
@@ -922,6 +927,23 @@
 #       define TARGET_LINK_SPEED_MASK                     (0xf << 0)
 #       define SELECTABLE_DEEMPHASIS                      (1 << 6)
 
+/* Audio */
+#define AZ_HOT_PLUG_CONTROL               0x7300
+#       define AZ_FORCE_CODEC_WAKE        (1 << 0)
+#       define JACK_DETECTION_ENABLE      (1 << 4)
+#       define UNSOLICITED_RESPONSE_ENABLE (1 << 8)
+#       define CODEC_HOT_PLUG_ENABLE      (1 << 12)
+#       define AUDIO_ENABLED              (1 << 31)
+/* DCE3 adds */
+#       define PIN0_JACK_DETECTION_ENABLE (1 << 4)
+#       define PIN1_JACK_DETECTION_ENABLE (1 << 5)
+#       define PIN2_JACK_DETECTION_ENABLE (1 << 6)
+#       define PIN3_JACK_DETECTION_ENABLE (1 << 7)
+#       define PIN0_AUDIO_ENABLED         (1 << 24)
+#       define PIN1_AUDIO_ENABLED         (1 << 25)
+#       define PIN2_AUDIO_ENABLED         (1 << 26)
+#       define PIN3_AUDIO_ENABLED         (1 << 27)
+
 /* Audio clocks DCE 2.0/3.0 */
 #define AUDIO_DTO                         0x7340
 #       define AUDIO_DTO_PHASE(x)         (((x) & 0xffff) << 0)
@@ -1476,6 +1498,7 @@
 #define UVD_CGC_GATE					0xf4a8
 #define UVD_LMI_CTRL2					0xf4f4
 #define UVD_MASTINT_EN					0xf500
+#define UVD_FW_START					0xf51C
 #define UVD_LMI_ADDR_EXT				0xf594
 #define UVD_LMI_CTRL					0xf598
 #define UVD_LMI_SWAP_CNTL				0xf5b4
@@ -1488,6 +1511,13 @@
 #define UVD_MPC_SET_MUX					0xf5f4
 #define UVD_MPC_SET_ALU					0xf5f8
 
+#define UVD_VCPU_CACHE_OFFSET0				0xf608
+#define UVD_VCPU_CACHE_SIZE0				0xf60c
+#define UVD_VCPU_CACHE_OFFSET1				0xf610
+#define UVD_VCPU_CACHE_SIZE1				0xf614
+#define UVD_VCPU_CACHE_OFFSET2				0xf618
+#define UVD_VCPU_CACHE_SIZE2				0xf61c
+
 #define UVD_VCPU_CNTL					0xf660
 #define UVD_SOFT_RESET					0xf680
 #define		RBC_SOFT_RESET					(1<<0)
@@ -1517,9 +1547,35 @@
 
 #define UVD_CONTEXT_ID					0xf6f4
 
+/* rs780 only */
+#define	GFX_MACRO_BYPASS_CNTL				0x30c0
+#define		SPLL_BYPASS_CNTL			(1 << 0)
+#define		UPLL_BYPASS_CNTL			(1 << 1)
+
+#define CG_UPLL_FUNC_CNTL				0x7e0
+#	define UPLL_RESET_MASK				0x00000001
+#	define UPLL_SLEEP_MASK				0x00000002
+#	define UPLL_BYPASS_EN_MASK			0x00000004
 #	define UPLL_CTLREQ_MASK				0x00000008
+#	define UPLL_FB_DIV(x)				((x) << 4)
+#	define UPLL_FB_DIV_MASK				0x0000FFF0
+#	define UPLL_REF_DIV(x)				((x) << 16)
+#	define UPLL_REF_DIV_MASK			0x003F0000
+#	define UPLL_REFCLK_SRC_SEL_MASK			0x20000000
 #	define UPLL_CTLACK_MASK				0x40000000
 #	define UPLL_CTLACK2_MASK			0x80000000
+#define CG_UPLL_FUNC_CNTL_2				0x7e4
+#	define UPLL_SW_HILEN(x)				((x) << 0)
+#	define UPLL_SW_LOLEN(x)				((x) << 4)
+#	define UPLL_SW_HILEN2(x)			((x) << 8)
+#	define UPLL_SW_LOLEN2(x)			((x) << 12)
+#	define UPLL_DIVEN_MASK				0x00010000
+#	define UPLL_DIVEN2_MASK				0x00020000
+#	define UPLL_SW_MASK				0x0003FFFF
+#	define VCLK_SRC_SEL(x)				((x) << 20)
+#	define VCLK_SRC_SEL_MASK			0x01F00000
+#	define DCLK_SRC_SEL(x)				((x) << 25)
+#	define DCLK_SRC_SEL_MASK			0x3E000000
 
 /*
  * PM4
diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h
index 3247bfd14410..f7c4b226a284 100644
--- a/drivers/gpu/drm/radeon/radeon.h
+++ b/drivers/gpu/drm/radeon/radeon.h
@@ -65,6 +65,8 @@
 #include <linux/list.h>
 #include <linux/kref.h>
 #include <linux/interval_tree.h>
+#include <linux/hashtable.h>
+#include <linux/fence.h>
 
 #include <ttm/ttm_bo_api.h>
 #include <ttm/ttm_bo_driver.h>
@@ -72,6 +74,8 @@
 #include <ttm/ttm_module.h>
 #include <ttm/ttm_execbuf_util.h>
 
+#include <drm/drm_gem.h>
+
 #include "radeon_family.h"
 #include "radeon_mode.h"
 #include "radeon_reg.h"
@@ -120,9 +124,6 @@ extern int radeon_backlight;
 #define RADEONFB_CONN_LIMIT			4
 #define RADEON_BIOS_NUM_SCRATCH			8
 
-/* fence seq are set to this number when signaled */
-#define RADEON_FENCE_SIGNALED_SEQ		0LL
-
 /* internal ring indices */
 /* r1xx+ has gfx CP ring */
 #define RADEON_RING_TYPE_GFX_INDEX		0
@@ -350,28 +351,32 @@ extern void evergreen_tiling_fields(unsigned tiling_flags, unsigned *bankw,
  * Fences.
  */
 struct radeon_fence_driver {
+	struct radeon_device		*rdev;
 	uint32_t			scratch_reg;
 	uint64_t			gpu_addr;
 	volatile uint32_t		*cpu_addr;
 	/* sync_seq is protected by ring emission lock */
 	uint64_t			sync_seq[RADEON_NUM_RINGS];
 	atomic64_t			last_seq;
-	bool				initialized;
+	bool				initialized, delayed_irq;
+	struct delayed_work		lockup_work;
 };
 
 struct radeon_fence {
+	struct fence base;
+
 	struct radeon_device		*rdev;
-	struct kref			kref;
-	/* protected by radeon_fence.lock */
 	uint64_t			seq;
 	/* RB, DMA, etc. */
 	unsigned			ring;
+
+	wait_queue_t			fence_wake;
 };
 
 int radeon_fence_driver_start_ring(struct radeon_device *rdev, int ring);
 int radeon_fence_driver_init(struct radeon_device *rdev);
 void radeon_fence_driver_fini(struct radeon_device *rdev);
-void radeon_fence_driver_force_completion(struct radeon_device *rdev);
+void radeon_fence_driver_force_completion(struct radeon_device *rdev, int ring);
 int radeon_fence_emit(struct radeon_device *rdev, struct radeon_fence **fence, int ring);
 void radeon_fence_process(struct radeon_device *rdev, int ring);
 bool radeon_fence_signaled(struct radeon_fence *fence);
@@ -469,7 +474,7 @@ struct radeon_bo {
 	struct list_head		list;
 	/* Protected by tbo.reserved */
 	u32				initial_domain;
-	u32				placements[3];
+	struct ttm_place		placements[3];
 	struct ttm_placement		placement;
 	struct ttm_buffer_object	tbo;
 	struct ttm_bo_kmap_obj		kmap;
@@ -489,6 +494,9 @@ struct radeon_bo {
 
 	struct ttm_bo_kmap_obj		dma_buf_vmap;
 	pid_t				pid;
+
+	struct radeon_mn		*mn;
+	struct interval_tree_node	mn_it;
 };
 #define gem_to_radeon_bo(gobj) container_of((gobj), struct radeon_bo, gem_base)
 
@@ -580,8 +588,12 @@ bool radeon_semaphore_emit_signal(struct radeon_device *rdev, int ring,
 				  struct radeon_semaphore *semaphore);
 bool radeon_semaphore_emit_wait(struct radeon_device *rdev, int ring,
 				struct radeon_semaphore *semaphore);
-void radeon_semaphore_sync_to(struct radeon_semaphore *semaphore,
-			      struct radeon_fence *fence);
+void radeon_semaphore_sync_fence(struct radeon_semaphore *semaphore,
+				 struct radeon_fence *fence);
+int radeon_semaphore_sync_resv(struct radeon_device *rdev,
+			       struct radeon_semaphore *semaphore,
+			       struct reservation_object *resv,
+			       bool shared);
 int radeon_semaphore_sync_rings(struct radeon_device *rdev,
 				struct radeon_semaphore *semaphore,
 				int waiting_ring);
@@ -702,7 +714,7 @@ struct radeon_flip_work {
 	uint64_t			base;
 	struct drm_pending_vblank_event *event;
 	struct radeon_bo		*old_rbo;
-	struct radeon_fence		*fence;
+	struct fence			*fence;
 };
 
 struct r500_irq_stat_regs {
@@ -780,6 +792,7 @@ struct radeon_irq {
 int radeon_irq_kms_init(struct radeon_device *rdev);
 void radeon_irq_kms_fini(struct radeon_device *rdev);
 void radeon_irq_kms_sw_irq_get(struct radeon_device *rdev, int ring);
+bool radeon_irq_kms_sw_irq_get_delayed(struct radeon_device *rdev, int ring);
 void radeon_irq_kms_sw_irq_put(struct radeon_device *rdev, int ring);
 void radeon_irq_kms_pflip_irq_get(struct radeon_device *rdev, int crtc);
 void radeon_irq_kms_pflip_irq_put(struct radeon_device *rdev, int crtc);
@@ -1642,7 +1655,8 @@ int radeon_uvd_get_create_msg(struct radeon_device *rdev, int ring,
 			      uint32_t handle, struct radeon_fence **fence);
 int radeon_uvd_get_destroy_msg(struct radeon_device *rdev, int ring,
 			       uint32_t handle, struct radeon_fence **fence);
-void radeon_uvd_force_into_uvd_segment(struct radeon_bo *rbo);
+void radeon_uvd_force_into_uvd_segment(struct radeon_bo *rbo,
+				       uint32_t allowed_domains);
 void radeon_uvd_free_handles(struct radeon_device *rdev,
 			     struct drm_file *filp);
 int radeon_uvd_cs_parse(struct radeon_cs_parser *parser);
@@ -1731,6 +1745,11 @@ void radeon_test_ring_sync(struct radeon_device *rdev,
 			   struct radeon_ring *cpB);
 void radeon_test_syncing(struct radeon_device *rdev);
 
+/*
+ * MMU Notifier
+ */
+int radeon_mn_register(struct radeon_bo *bo, unsigned long addr);
+void radeon_mn_unregister(struct radeon_bo *bo);
 
 /*
  * Debugfs
@@ -1845,24 +1864,24 @@ struct radeon_asic {
 	} display;
 	/* copy functions for bo handling */
 	struct {
-		int (*blit)(struct radeon_device *rdev,
-			    uint64_t src_offset,
-			    uint64_t dst_offset,
-			    unsigned num_gpu_pages,
-			    struct radeon_fence **fence);
+		struct radeon_fence *(*blit)(struct radeon_device *rdev,
+					     uint64_t src_offset,
+					     uint64_t dst_offset,
+					     unsigned num_gpu_pages,
+					     struct reservation_object *resv);
 		u32 blit_ring_index;
-		int (*dma)(struct radeon_device *rdev,
-			   uint64_t src_offset,
-			   uint64_t dst_offset,
-			   unsigned num_gpu_pages,
-			   struct radeon_fence **fence);
+		struct radeon_fence *(*dma)(struct radeon_device *rdev,
+					    uint64_t src_offset,
+					    uint64_t dst_offset,
+					    unsigned num_gpu_pages,
+					    struct reservation_object *resv);
 		u32 dma_ring_index;
 		/* method used for bo copy */
-		int (*copy)(struct radeon_device *rdev,
-			    uint64_t src_offset,
-			    uint64_t dst_offset,
-			    unsigned num_gpu_pages,
-			    struct radeon_fence **fence);
+		struct radeon_fence *(*copy)(struct radeon_device *rdev,
+					     uint64_t src_offset,
+					     uint64_t dst_offset,
+					     unsigned num_gpu_pages,
+					     struct reservation_object *resv);
 		/* ring used for bo copies */
 		u32 copy_ring_index;
 	} copy;
@@ -2144,6 +2163,8 @@ int radeon_gem_info_ioctl(struct drm_device *dev, void *data,
 			  struct drm_file *filp);
 int radeon_gem_create_ioctl(struct drm_device *dev, void *data,
 			    struct drm_file *filp);
+int radeon_gem_userptr_ioctl(struct drm_device *dev, void *data,
+			     struct drm_file *filp);
 int radeon_gem_pin_ioctl(struct drm_device *dev, void *data,
 			 struct drm_file *file_priv);
 int radeon_gem_unpin_ioctl(struct drm_device *dev, void *data,
@@ -2300,6 +2321,7 @@ struct radeon_device {
 	struct radeon_mman		mman;
 	struct radeon_fence_driver	fence_drv[RADEON_NUM_RINGS];
 	wait_queue_head_t		fence_queue;
+	unsigned			fence_context;
 	struct mutex			ring_lock;
 	struct radeon_ring		ring[RADEON_NUM_RINGS];
 	bool				ib_pool_ready;
@@ -2318,7 +2340,7 @@ struct radeon_device {
 	bool				need_dma32;
 	bool				accel_working;
 	bool				fastfb_working; /* IGP feature*/
-	bool				needs_reset;
+	bool				needs_reset, in_reset;
 	struct radeon_surface_reg surface_regs[RADEON_GEM_MAX_SURFACES];
 	const struct firmware *me_fw;	/* all family ME firmware */
 	const struct firmware *pfp_fw;	/* r6/700 PFP firmware */
@@ -2339,7 +2361,6 @@ struct radeon_device {
 	struct radeon_mec mec;
 	struct work_struct hotplug_work;
 	struct work_struct audio_work;
-	struct work_struct reset_work;
 	int num_crtc; /* number of crtcs */
 	struct mutex dc_hw_i2c_mutex; /* display controller hw i2c mutex */
 	bool has_uvd;
@@ -2376,6 +2397,9 @@ struct radeon_device {
 	/* tracking pinned memory */
 	u64 vram_pin_size;
 	u64 gart_pin_size;
+
+	struct mutex	mn_lock;
+	DECLARE_HASHTABLE(mn_hash, 7);
 };
 
 bool radeon_is_px(struct drm_device *dev);
@@ -2431,7 +2455,17 @@ void cik_mm_wdoorbell(struct radeon_device *rdev, u32 index, u32 v);
 /*
  * Cast helper
  */
-#define to_radeon_fence(p) ((struct radeon_fence *)(p))
+extern const struct fence_ops radeon_fence_ops;
+
+static inline struct radeon_fence *to_radeon_fence(struct fence *f)
+{
+	struct radeon_fence *__f = container_of(f, struct radeon_fence, base);
+
+	if (__f->base.ops == &radeon_fence_ops)
+		return __f;
+
+	return NULL;
+}
 
 /*
  * Registers read & write functions.
@@ -2751,18 +2785,25 @@ void radeon_atombios_fini(struct radeon_device *rdev);
 /*
  * RING helpers.
  */
-#if DRM_DEBUG_CODE == 0
+
+/**
+ * radeon_ring_write - write a value to the ring
+ *
+ * @ring: radeon_ring structure holding ring information
+ * @v: dword (dw) value to write
+ *
+ * Write a value to the requested ring buffer (all asics).
+ */
 static inline void radeon_ring_write(struct radeon_ring *ring, uint32_t v)
 {
+	if (ring->count_dw <= 0)
+		DRM_ERROR("radeon: writing more dwords to the ring than expected!\n");
+
 	ring->ring[ring->wptr++] = v;
 	ring->wptr &= ring->ptr_mask;
 	ring->count_dw--;
 	ring->ring_free_dw--;
 }
-#else
-/* With debugging this is just too big to inline */
-void radeon_ring_write(struct radeon_ring *ring, uint32_t v);
-#endif
 
 /*
  * ASICs macro.
@@ -2801,9 +2842,9 @@ void radeon_ring_write(struct radeon_ring *ring, uint32_t v);
 #define radeon_hdmi_setmode(rdev, e, m) (rdev)->asic->display.hdmi_setmode((e), (m))
 #define radeon_fence_ring_emit(rdev, r, fence) (rdev)->asic->ring[(r)]->emit_fence((rdev), (fence))
 #define radeon_semaphore_ring_emit(rdev, r, cp, semaphore, emit_wait) (rdev)->asic->ring[(r)]->emit_semaphore((rdev), (cp), (semaphore), (emit_wait))
-#define radeon_copy_blit(rdev, s, d, np, f) (rdev)->asic->copy.blit((rdev), (s), (d), (np), (f))
-#define radeon_copy_dma(rdev, s, d, np, f) (rdev)->asic->copy.dma((rdev), (s), (d), (np), (f))
-#define radeon_copy(rdev, s, d, np, f) (rdev)->asic->copy.copy((rdev), (s), (d), (np), (f))
+#define radeon_copy_blit(rdev, s, d, np, resv) (rdev)->asic->copy.blit((rdev), (s), (d), (np), (resv))
+#define radeon_copy_dma(rdev, s, d, np, resv) (rdev)->asic->copy.dma((rdev), (s), (d), (np), (resv))
+#define radeon_copy(rdev, s, d, np, resv) (rdev)->asic->copy.copy((rdev), (s), (d), (np), (resv))
 #define radeon_copy_blit_ring_index(rdev) (rdev)->asic->copy.blit_ring_index
 #define radeon_copy_dma_ring_index(rdev) (rdev)->asic->copy.dma_ring_index
 #define radeon_copy_ring_index(rdev) (rdev)->asic->copy.copy_ring_index
@@ -2877,6 +2918,10 @@ extern void radeon_legacy_set_clock_gating(struct radeon_device *rdev, int enabl
 extern void radeon_atom_set_clock_gating(struct radeon_device *rdev, int enable);
 extern void radeon_ttm_placement_from_domain(struct radeon_bo *rbo, u32 domain);
 extern bool radeon_ttm_bo_is_radeon_bo(struct ttm_buffer_object *bo);
+extern int radeon_ttm_tt_set_userptr(struct ttm_tt *ttm, uint64_t addr,
+				     uint32_t flags);
+extern bool radeon_ttm_tt_has_userptr(struct ttm_tt *ttm);
+extern bool radeon_ttm_tt_is_readonly(struct ttm_tt *ttm);
 extern void radeon_vram_location(struct radeon_device *rdev, struct radeon_mc *mc, u64 base);
 extern void radeon_gtt_location(struct radeon_device *rdev, struct radeon_mc *mc);
 extern int radeon_resume_kms(struct drm_device *dev, bool resume, bool fbcon);
@@ -2934,10 +2979,10 @@ struct r600_audio_pin *r600_audio_get_pin(struct radeon_device *rdev);
 struct r600_audio_pin *dce6_audio_get_pin(struct radeon_device *rdev);
 void r600_audio_enable(struct radeon_device *rdev,
 		       struct r600_audio_pin *pin,
-		       bool enable);
+		       u8 enable_mask);
 void dce6_audio_enable(struct radeon_device *rdev,
 		       struct r600_audio_pin *pin,
-		       bool enable);
+		       u8 enable_mask);
 
 /*
  * R600 vram scratch functions
diff --git a/drivers/gpu/drm/radeon/radeon_asic.c b/drivers/gpu/drm/radeon/radeon_asic.c
index 2dd5847f9b98..850de57069be 100644
--- a/drivers/gpu/drm/radeon/radeon_asic.c
+++ b/drivers/gpu/drm/radeon/radeon_asic.c
@@ -963,6 +963,19 @@ static struct radeon_asic r600_asic = {
 	},
 };
 
+static struct radeon_asic_ring rv6xx_uvd_ring = {
+	.ib_execute = &uvd_v1_0_ib_execute,
+	.emit_fence = &uvd_v1_0_fence_emit,
+	.emit_semaphore = &uvd_v1_0_semaphore_emit,
+	.cs_parse = &radeon_uvd_cs_parse,
+	.ring_test = &uvd_v1_0_ring_test,
+	.ib_test = &uvd_v1_0_ib_test,
+	.is_lockup = &radeon_ring_test_lockup,
+	.get_rptr = &uvd_v1_0_get_rptr,
+	.get_wptr = &uvd_v1_0_get_wptr,
+	.set_wptr = &uvd_v1_0_set_wptr,
+};
+
 static struct radeon_asic rv6xx_asic = {
 	.init = &r600_init,
 	.fini = &r600_fini,
@@ -982,6 +995,7 @@ static struct radeon_asic rv6xx_asic = {
 	.ring = {
 		[RADEON_RING_TYPE_GFX_INDEX] = &r600_gfx_ring,
 		[R600_RING_TYPE_DMA_INDEX] = &r600_dma_ring,
+		[R600_RING_TYPE_UVD_INDEX] = &rv6xx_uvd_ring,
 	},
 	.irq = {
 		.set = &r600_irq_set,
@@ -1072,6 +1086,7 @@ static struct radeon_asic rs780_asic = {
 	.ring = {
 		[RADEON_RING_TYPE_GFX_INDEX] = &r600_gfx_ring,
 		[R600_RING_TYPE_DMA_INDEX] = &r600_dma_ring,
+		[R600_RING_TYPE_UVD_INDEX] = &rv6xx_uvd_ring,
 	},
 	.irq = {
 		.set = &r600_irq_set,
@@ -2296,7 +2311,15 @@ int radeon_asic_init(struct radeon_device *rdev)
 	case CHIP_RS780:
 	case CHIP_RS880:
 		rdev->asic = &rs780_asic;
-		rdev->has_uvd = true;
+		/* 760G/780V/880V don't have UVD */
+		if ((rdev->pdev->device == 0x9616)||
+		    (rdev->pdev->device == 0x9611)||
+		    (rdev->pdev->device == 0x9613)||
+		    (rdev->pdev->device == 0x9711)||
+		    (rdev->pdev->device == 0x9713))
+			rdev->has_uvd = false;
+		else
+			rdev->has_uvd = true;
 		break;
 	case CHIP_RV770:
 	case CHIP_RV730:
diff --git a/drivers/gpu/drm/radeon/radeon_asic.h b/drivers/gpu/drm/radeon/radeon_asic.h
index 7756bc1e1cd3..d8ace5b28a5b 100644
--- a/drivers/gpu/drm/radeon/radeon_asic.h
+++ b/drivers/gpu/drm/radeon/radeon_asic.h
@@ -81,11 +81,11 @@ bool r100_semaphore_ring_emit(struct radeon_device *rdev,
 int r100_cs_parse(struct radeon_cs_parser *p);
 void r100_pll_wreg(struct radeon_device *rdev, uint32_t reg, uint32_t v);
 uint32_t r100_pll_rreg(struct radeon_device *rdev, uint32_t reg);
-int r100_copy_blit(struct radeon_device *rdev,
-		   uint64_t src_offset,
-		   uint64_t dst_offset,
-		   unsigned num_gpu_pages,
-		   struct radeon_fence **fence);
+struct radeon_fence *r100_copy_blit(struct radeon_device *rdev,
+				    uint64_t src_offset,
+				    uint64_t dst_offset,
+				    unsigned num_gpu_pages,
+				    struct reservation_object *resv);
 int r100_set_surface_reg(struct radeon_device *rdev, int reg,
 			 uint32_t tiling_flags, uint32_t pitch,
 			 uint32_t offset, uint32_t obj_size);
@@ -152,11 +152,11 @@ void r100_gfx_set_wptr(struct radeon_device *rdev,
 /*
  * r200,rv250,rs300,rv280
  */
-extern int r200_copy_dma(struct radeon_device *rdev,
-			 uint64_t src_offset,
-			 uint64_t dst_offset,
-			 unsigned num_gpu_pages,
-			 struct radeon_fence **fence);
+struct radeon_fence *r200_copy_dma(struct radeon_device *rdev,
+				   uint64_t src_offset,
+				   uint64_t dst_offset,
+				   unsigned num_gpu_pages,
+				   struct reservation_object *resv);
 void r200_set_safe_registers(struct radeon_device *rdev);
 
 /*
@@ -340,12 +340,14 @@ int r600_dma_ib_test(struct radeon_device *rdev, struct radeon_ring *ring);
 void r600_ring_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib);
 int r600_ring_test(struct radeon_device *rdev, struct radeon_ring *cp);
 int r600_dma_ring_test(struct radeon_device *rdev, struct radeon_ring *cp);
-int r600_copy_cpdma(struct radeon_device *rdev,
-		    uint64_t src_offset, uint64_t dst_offset,
-		    unsigned num_gpu_pages, struct radeon_fence **fence);
-int r600_copy_dma(struct radeon_device *rdev,
-		  uint64_t src_offset, uint64_t dst_offset,
-		  unsigned num_gpu_pages, struct radeon_fence **fence);
+struct radeon_fence *r600_copy_cpdma(struct radeon_device *rdev,
+				     uint64_t src_offset, uint64_t dst_offset,
+				     unsigned num_gpu_pages,
+				     struct reservation_object *resv);
+struct radeon_fence *r600_copy_dma(struct radeon_device *rdev,
+				   uint64_t src_offset, uint64_t dst_offset,
+				   unsigned num_gpu_pages,
+				   struct reservation_object *resv);
 void r600_hpd_init(struct radeon_device *rdev);
 void r600_hpd_fini(struct radeon_device *rdev);
 bool r600_hpd_sense(struct radeon_device *rdev, enum radeon_hpd_id hpd);
@@ -389,7 +391,6 @@ void r600_disable_interrupts(struct radeon_device *rdev);
 void r600_rlc_stop(struct radeon_device *rdev);
 /* r600 audio */
 int r600_audio_init(struct radeon_device *rdev);
-struct r600_audio_pin r600_audio_status(struct radeon_device *rdev);
 void r600_audio_fini(struct radeon_device *rdev);
 void r600_audio_set_dto(struct drm_encoder *encoder, u32 clock);
 void r600_hdmi_update_avi_infoframe(struct drm_encoder *encoder, void *buffer,
@@ -461,10 +462,10 @@ bool rv770_page_flip_pending(struct radeon_device *rdev, int crtc);
 void r700_vram_gtt_location(struct radeon_device *rdev, struct radeon_mc *mc);
 void r700_cp_stop(struct radeon_device *rdev);
 void r700_cp_fini(struct radeon_device *rdev);
-int rv770_copy_dma(struct radeon_device *rdev,
-		  uint64_t src_offset, uint64_t dst_offset,
-		  unsigned num_gpu_pages,
-		   struct radeon_fence **fence);
+struct radeon_fence *rv770_copy_dma(struct radeon_device *rdev,
+				    uint64_t src_offset, uint64_t dst_offset,
+				    unsigned num_gpu_pages,
+				    struct reservation_object *resv);
 u32 rv770_get_xclk(struct radeon_device *rdev);
 int rv770_set_uvd_clocks(struct radeon_device *rdev, u32 vclk, u32 dclk);
 int rv770_get_temp(struct radeon_device *rdev);
@@ -535,10 +536,10 @@ void evergreen_dma_fence_ring_emit(struct radeon_device *rdev,
 				   struct radeon_fence *fence);
 void evergreen_dma_ring_ib_execute(struct radeon_device *rdev,
 				   struct radeon_ib *ib);
-int evergreen_copy_dma(struct radeon_device *rdev,
-		       uint64_t src_offset, uint64_t dst_offset,
-		       unsigned num_gpu_pages,
-		       struct radeon_fence **fence);
+struct radeon_fence *evergreen_copy_dma(struct radeon_device *rdev,
+					uint64_t src_offset, uint64_t dst_offset,
+					unsigned num_gpu_pages,
+					struct reservation_object *resv);
 void evergreen_hdmi_enable(struct drm_encoder *encoder, bool enable);
 void evergreen_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode *mode);
 int evergreen_get_temp(struct radeon_device *rdev);
@@ -700,10 +701,10 @@ int si_vm_init(struct radeon_device *rdev);
 void si_vm_fini(struct radeon_device *rdev);
 void si_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm *vm);
 int si_ib_parse(struct radeon_device *rdev, struct radeon_ib *ib);
-int si_copy_dma(struct radeon_device *rdev,
-		uint64_t src_offset, uint64_t dst_offset,
-		unsigned num_gpu_pages,
-		struct radeon_fence **fence);
+struct radeon_fence *si_copy_dma(struct radeon_device *rdev,
+				 uint64_t src_offset, uint64_t dst_offset,
+				 unsigned num_gpu_pages,
+				 struct reservation_object *resv);
 
 void si_dma_vm_copy_pages(struct radeon_device *rdev,
 			  struct radeon_ib *ib,
@@ -759,14 +760,14 @@ bool cik_sdma_semaphore_ring_emit(struct radeon_device *rdev,
 				  struct radeon_semaphore *semaphore,
 				  bool emit_wait);
 void cik_sdma_ring_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib);
-int cik_copy_dma(struct radeon_device *rdev,
-		 uint64_t src_offset, uint64_t dst_offset,
-		 unsigned num_gpu_pages,
-		 struct radeon_fence **fence);
-int cik_copy_cpdma(struct radeon_device *rdev,
-		   uint64_t src_offset, uint64_t dst_offset,
-		   unsigned num_gpu_pages,
-		   struct radeon_fence **fence);
+struct radeon_fence *cik_copy_dma(struct radeon_device *rdev,
+				  uint64_t src_offset, uint64_t dst_offset,
+				  unsigned num_gpu_pages,
+				  struct reservation_object *resv);
+struct radeon_fence *cik_copy_cpdma(struct radeon_device *rdev,
+				    uint64_t src_offset, uint64_t dst_offset,
+				    unsigned num_gpu_pages,
+				    struct reservation_object *resv);
 int cik_sdma_ring_test(struct radeon_device *rdev, struct radeon_ring *ring);
 int cik_sdma_ib_test(struct radeon_device *rdev, struct radeon_ring *ring);
 bool cik_sdma_is_lockup(struct radeon_device *rdev, struct radeon_ring *ring);
@@ -882,6 +883,7 @@ uint32_t uvd_v1_0_get_wptr(struct radeon_device *rdev,
                            struct radeon_ring *ring);
 void uvd_v1_0_set_wptr(struct radeon_device *rdev,
                        struct radeon_ring *ring);
+int uvd_v1_0_resume(struct radeon_device *rdev);
 
 int uvd_v1_0_init(struct radeon_device *rdev);
 void uvd_v1_0_fini(struct radeon_device *rdev);
@@ -889,6 +891,8 @@ int uvd_v1_0_start(struct radeon_device *rdev);
 void uvd_v1_0_stop(struct radeon_device *rdev);
 
 int uvd_v1_0_ring_test(struct radeon_device *rdev, struct radeon_ring *ring);
+void uvd_v1_0_fence_emit(struct radeon_device *rdev,
+			 struct radeon_fence *fence);
 int uvd_v1_0_ib_test(struct radeon_device *rdev, struct radeon_ring *ring);
 bool uvd_v1_0_semaphore_emit(struct radeon_device *rdev,
 			     struct radeon_ring *ring,
diff --git a/drivers/gpu/drm/radeon/radeon_atombios.c b/drivers/gpu/drm/radeon/radeon_atombios.c
index e74c7e387dde..df69b92ba164 100644
--- a/drivers/gpu/drm/radeon/radeon_atombios.c
+++ b/drivers/gpu/drm/radeon/radeon_atombios.c
@@ -458,7 +458,7 @@ static bool radeon_atom_apply_quirks(struct drm_device *dev,
 	return true;
 }
 
-const int supported_devices_connector_convert[] = {
+static const int supported_devices_connector_convert[] = {
 	DRM_MODE_CONNECTOR_Unknown,
 	DRM_MODE_CONNECTOR_VGA,
 	DRM_MODE_CONNECTOR_DVII,
@@ -477,7 +477,7 @@ const int supported_devices_connector_convert[] = {
 	DRM_MODE_CONNECTOR_DisplayPort
 };
 
-const uint16_t supported_devices_connector_object_id_convert[] = {
+static const uint16_t supported_devices_connector_object_id_convert[] = {
 	CONNECTOR_OBJECT_ID_NONE,
 	CONNECTOR_OBJECT_ID_VGA,
 	CONNECTOR_OBJECT_ID_DUAL_LINK_DVI_I, /* not all boards support DL */
@@ -494,7 +494,7 @@ const uint16_t supported_devices_connector_object_id_convert[] = {
 	CONNECTOR_OBJECT_ID_SVIDEO
 };
 
-const int object_connector_convert[] = {
+static const int object_connector_convert[] = {
 	DRM_MODE_CONNECTOR_Unknown,
 	DRM_MODE_CONNECTOR_DVII,
 	DRM_MODE_CONNECTOR_DVII,
diff --git a/drivers/gpu/drm/radeon/radeon_benchmark.c b/drivers/gpu/drm/radeon/radeon_benchmark.c
index 69f5695bdab9..9e7f23dd14bd 100644
--- a/drivers/gpu/drm/radeon/radeon_benchmark.c
+++ b/drivers/gpu/drm/radeon/radeon_benchmark.c
@@ -45,33 +45,29 @@ static int radeon_benchmark_do_move(struct radeon_device *rdev, unsigned size,
 	for (i = 0; i < n; i++) {
 		switch (flag) {
 		case RADEON_BENCHMARK_COPY_DMA:
-			r = radeon_copy_dma(rdev, saddr, daddr,
-					    size / RADEON_GPU_PAGE_SIZE,
-					    &fence);
+			fence = radeon_copy_dma(rdev, saddr, daddr,
+						size / RADEON_GPU_PAGE_SIZE,
+						NULL);
 			break;
 		case RADEON_BENCHMARK_COPY_BLIT:
-			r = radeon_copy_blit(rdev, saddr, daddr,
-					     size / RADEON_GPU_PAGE_SIZE,
-					     &fence);
+			fence = radeon_copy_blit(rdev, saddr, daddr,
+						 size / RADEON_GPU_PAGE_SIZE,
+						 NULL);
 			break;
 		default:
 			DRM_ERROR("Unknown copy method\n");
-			r = -EINVAL;
+			return -EINVAL;
 		}
-		if (r)
-			goto exit_do_move;
+		if (IS_ERR(fence))
+			return PTR_ERR(fence);
+
 		r = radeon_fence_wait(fence, false);
-		if (r)
-			goto exit_do_move;
 		radeon_fence_unref(&fence);
+		if (r)
+			return r;
 	}
 	end_jiffies = jiffies;
-	r = jiffies_to_msecs(end_jiffies - start_jiffies);
-
-exit_do_move:
-	if (fence)
-		radeon_fence_unref(&fence);
-	return r;
+	return jiffies_to_msecs(end_jiffies - start_jiffies);
 }
 
 
@@ -97,7 +93,7 @@ static void radeon_benchmark_move(struct radeon_device *rdev, unsigned size,
 	int time;
 
 	n = RADEON_BENCHMARK_ITERATIONS;
-	r = radeon_bo_create(rdev, size, PAGE_SIZE, true, sdomain, 0, NULL, &sobj);
+	r = radeon_bo_create(rdev, size, PAGE_SIZE, true, sdomain, 0, NULL, NULL, &sobj);
 	if (r) {
 		goto out_cleanup;
 	}
@@ -109,7 +105,7 @@ static void radeon_benchmark_move(struct radeon_device *rdev, unsigned size,
 	if (r) {
 		goto out_cleanup;
 	}
-	r = radeon_bo_create(rdev, size, PAGE_SIZE, true, ddomain, 0, NULL, &dobj);
+	r = radeon_bo_create(rdev, size, PAGE_SIZE, true, ddomain, 0, NULL, NULL, &dobj);
 	if (r) {
 		goto out_cleanup;
 	}
diff --git a/drivers/gpu/drm/radeon/radeon_combios.c b/drivers/gpu/drm/radeon/radeon_combios.c
index 6651177110f0..3e5f6b71f3ad 100644
--- a/drivers/gpu/drm/radeon/radeon_combios.c
+++ b/drivers/gpu/drm/radeon/radeon_combios.c
@@ -116,7 +116,7 @@ enum radeon_combios_connector {
 	CONNECTOR_UNSUPPORTED_LEGACY
 };
 
-const int legacy_connector_convert[] = {
+static const int legacy_connector_convert[] = {
 	DRM_MODE_CONNECTOR_Unknown,
 	DRM_MODE_CONNECTOR_DVID,
 	DRM_MODE_CONNECTOR_VGA,
diff --git a/drivers/gpu/drm/radeon/radeon_cp.c b/drivers/gpu/drm/radeon/radeon_cp.c
index bb0d5c3a8311..ea134a7d51a5 100644
--- a/drivers/gpu/drm/radeon/radeon_cp.c
+++ b/drivers/gpu/drm/radeon/radeon_cp.c
@@ -1298,27 +1298,27 @@ static int radeon_do_init_cp(struct drm_device *dev, drm_radeon_init_t *init,
 	dev_priv->buffers_offset = init->buffers_offset;
 	dev_priv->gart_textures_offset = init->gart_textures_offset;
 
-	master_priv->sarea = drm_getsarea(dev);
+	master_priv->sarea = drm_legacy_getsarea(dev);
 	if (!master_priv->sarea) {
 		DRM_ERROR("could not find sarea!\n");
 		radeon_do_cleanup_cp(dev);
 		return -EINVAL;
 	}
 
-	dev_priv->cp_ring = drm_core_findmap(dev, init->ring_offset);
+	dev_priv->cp_ring = drm_legacy_findmap(dev, init->ring_offset);
 	if (!dev_priv->cp_ring) {
 		DRM_ERROR("could not find cp ring region!\n");
 		radeon_do_cleanup_cp(dev);
 		return -EINVAL;
 	}
-	dev_priv->ring_rptr = drm_core_findmap(dev, init->ring_rptr_offset);
+	dev_priv->ring_rptr = drm_legacy_findmap(dev, init->ring_rptr_offset);
 	if (!dev_priv->ring_rptr) {
 		DRM_ERROR("could not find ring read pointer!\n");
 		radeon_do_cleanup_cp(dev);
 		return -EINVAL;
 	}
 	dev->agp_buffer_token = init->buffers_offset;
-	dev->agp_buffer_map = drm_core_findmap(dev, init->buffers_offset);
+	dev->agp_buffer_map = drm_legacy_findmap(dev, init->buffers_offset);
 	if (!dev->agp_buffer_map) {
 		DRM_ERROR("could not find dma buffer region!\n");
 		radeon_do_cleanup_cp(dev);
@@ -1327,7 +1327,7 @@ static int radeon_do_init_cp(struct drm_device *dev, drm_radeon_init_t *init,
 
 	if (init->gart_textures_offset) {
 		dev_priv->gart_textures =
-		    drm_core_findmap(dev, init->gart_textures_offset);
+		    drm_legacy_findmap(dev, init->gart_textures_offset);
 		if (!dev_priv->gart_textures) {
 			DRM_ERROR("could not find GART texture region!\n");
 			radeon_do_cleanup_cp(dev);
@@ -1337,9 +1337,9 @@ static int radeon_do_init_cp(struct drm_device *dev, drm_radeon_init_t *init,
 
 #if __OS_HAS_AGP
 	if (dev_priv->flags & RADEON_IS_AGP) {
-		drm_core_ioremap_wc(dev_priv->cp_ring, dev);
-		drm_core_ioremap_wc(dev_priv->ring_rptr, dev);
-		drm_core_ioremap_wc(dev->agp_buffer_map, dev);
+		drm_legacy_ioremap_wc(dev_priv->cp_ring, dev);
+		drm_legacy_ioremap_wc(dev_priv->ring_rptr, dev);
+		drm_legacy_ioremap_wc(dev->agp_buffer_map, dev);
 		if (!dev_priv->cp_ring->handle ||
 		    !dev_priv->ring_rptr->handle ||
 		    !dev->agp_buffer_map->handle) {
@@ -1475,7 +1475,7 @@ static int radeon_do_init_cp(struct drm_device *dev, drm_radeon_init_t *init,
 			dev_priv->gart_info.mapping.size =
 			    dev_priv->gart_info.table_size;
 
-			drm_core_ioremap_wc(&dev_priv->gart_info.mapping, dev);
+			drm_legacy_ioremap_wc(&dev_priv->gart_info.mapping, dev);
 			dev_priv->gart_info.addr =
 			    dev_priv->gart_info.mapping.handle;
 
@@ -1569,15 +1569,15 @@ static int radeon_do_cleanup_cp(struct drm_device * dev)
 #if __OS_HAS_AGP
 	if (dev_priv->flags & RADEON_IS_AGP) {
 		if (dev_priv->cp_ring != NULL) {
-			drm_core_ioremapfree(dev_priv->cp_ring, dev);
+			drm_legacy_ioremapfree(dev_priv->cp_ring, dev);
 			dev_priv->cp_ring = NULL;
 		}
 		if (dev_priv->ring_rptr != NULL) {
-			drm_core_ioremapfree(dev_priv->ring_rptr, dev);
+			drm_legacy_ioremapfree(dev_priv->ring_rptr, dev);
 			dev_priv->ring_rptr = NULL;
 		}
 		if (dev->agp_buffer_map != NULL) {
-			drm_core_ioremapfree(dev->agp_buffer_map, dev);
+			drm_legacy_ioremapfree(dev->agp_buffer_map, dev);
 			dev->agp_buffer_map = NULL;
 		}
 	} else
@@ -1597,7 +1597,7 @@ static int radeon_do_cleanup_cp(struct drm_device * dev)
 
 		if (dev_priv->gart_info.gart_table_location == DRM_ATI_GART_FB)
 		{
-			drm_core_ioremapfree(&dev_priv->gart_info.mapping, dev);
+			drm_legacy_ioremapfree(&dev_priv->gart_info.mapping, dev);
 			dev_priv->gart_info.addr = NULL;
 		}
 	}
@@ -2106,9 +2106,9 @@ int radeon_driver_load(struct drm_device *dev, unsigned long flags)
 	else
 		dev_priv->flags |= RADEON_IS_PCI;
 
-	ret = drm_addmap(dev, pci_resource_start(dev->pdev, 2),
-			 pci_resource_len(dev->pdev, 2), _DRM_REGISTERS,
-			 _DRM_READ_ONLY | _DRM_DRIVER, &dev_priv->mmio);
+	ret = drm_legacy_addmap(dev, pci_resource_start(dev->pdev, 2),
+				pci_resource_len(dev->pdev, 2), _DRM_REGISTERS,
+				_DRM_READ_ONLY | _DRM_DRIVER, &dev_priv->mmio);
 	if (ret != 0)
 		return ret;
 
@@ -2135,8 +2135,8 @@ int radeon_master_create(struct drm_device *dev, struct drm_master *master)
 
 	/* prebuild the SAREA */
 	sareapage = max_t(unsigned long, SAREA_MAX, PAGE_SIZE);
-	ret = drm_addmap(dev, 0, sareapage, _DRM_SHM, _DRM_CONTAINS_LOCK,
-			 &master_priv->sarea);
+	ret = drm_legacy_addmap(dev, 0, sareapage, _DRM_SHM, _DRM_CONTAINS_LOCK,
+				&master_priv->sarea);
 	if (ret) {
 		DRM_ERROR("SAREA setup failed\n");
 		kfree(master_priv);
@@ -2162,7 +2162,7 @@ void radeon_master_destroy(struct drm_device *dev, struct drm_master *master)
 
 	master_priv->sarea_priv = NULL;
 	if (master_priv->sarea)
-		drm_rmmap_locked(dev, master_priv->sarea);
+		drm_legacy_rmmap_locked(dev, master_priv->sarea);
 
 	kfree(master_priv);
 
@@ -2181,9 +2181,9 @@ int radeon_driver_firstopen(struct drm_device *dev)
 	dev_priv->gart_info.table_size = RADEON_PCIGART_TABLE_SIZE;
 
 	dev_priv->fb_aper_offset = pci_resource_start(dev->pdev, 0);
-	ret = drm_addmap(dev, dev_priv->fb_aper_offset,
-			 pci_resource_len(dev->pdev, 0), _DRM_FRAME_BUFFER,
-			 _DRM_WRITE_COMBINING, &map);
+	ret = drm_legacy_addmap(dev, dev_priv->fb_aper_offset,
+				pci_resource_len(dev->pdev, 0),
+				_DRM_FRAME_BUFFER, _DRM_WRITE_COMBINING, &map);
 	if (ret != 0)
 		return ret;
 
@@ -2196,7 +2196,7 @@ int radeon_driver_unload(struct drm_device *dev)
 
 	DRM_DEBUG("\n");
 
-	drm_rmmap(dev, dev_priv->mmio);
+	drm_legacy_rmmap(dev, dev_priv->mmio);
 
 	kfree(dev_priv);
 
diff --git a/drivers/gpu/drm/radeon/radeon_cs.c b/drivers/gpu/drm/radeon/radeon_cs.c
index 83f382e8e40e..1c893447d7cd 100644
--- a/drivers/gpu/drm/radeon/radeon_cs.c
+++ b/drivers/gpu/drm/radeon/radeon_cs.c
@@ -78,7 +78,8 @@ static int radeon_cs_parser_relocs(struct radeon_cs_parser *p)
 	struct radeon_cs_chunk *chunk;
 	struct radeon_cs_buckets buckets;
 	unsigned i, j;
-	bool duplicate;
+	bool duplicate, need_mmap_lock = false;
+	int r;
 
 	if (p->chunk_relocs_idx == -1) {
 		return 0;
@@ -136,10 +137,13 @@ static int radeon_cs_parser_relocs(struct radeon_cs_parser *p)
 			   + !!r->write_domain;
 
 		/* the first reloc of an UVD job is the msg and that must be in
-		   VRAM, also but everything into VRAM on AGP cards to avoid
-		   image corruptions */
+		   VRAM, also but everything into VRAM on AGP cards and older
+		   IGP chips to avoid image corruptions */
 		if (p->ring == R600_RING_TYPE_UVD_INDEX &&
-		    (i == 0 || drm_pci_device_is_agp(p->rdev->ddev))) {
+		    (i == 0 || drm_pci_device_is_agp(p->rdev->ddev) ||
+		     p->rdev->family == CHIP_RS780 ||
+		     p->rdev->family == CHIP_RS880)) {
+
 			/* TODO: is this still needed for NI+ ? */
 			p->relocs[i].prefered_domains =
 				RADEON_GEM_DOMAIN_VRAM;
@@ -165,7 +169,21 @@ static int radeon_cs_parser_relocs(struct radeon_cs_parser *p)
 			p->relocs[i].allowed_domains = domain;
 		}
 
+		if (radeon_ttm_tt_has_userptr(p->relocs[i].robj->tbo.ttm)) {
+			uint32_t domain = p->relocs[i].prefered_domains;
+			if (!(domain & RADEON_GEM_DOMAIN_GTT)) {
+				DRM_ERROR("Only RADEON_GEM_DOMAIN_GTT is "
+					  "allowed for userptr BOs\n");
+				return -EINVAL;
+			}
+			need_mmap_lock = true;
+			domain = RADEON_GEM_DOMAIN_GTT;
+			p->relocs[i].prefered_domains = domain;
+			p->relocs[i].allowed_domains = domain;
+		}
+
 		p->relocs[i].tv.bo = &p->relocs[i].robj->tbo;
+		p->relocs[i].tv.shared = !r->write_domain;
 		p->relocs[i].handle = r->handle;
 
 		radeon_cs_buckets_add(&buckets, &p->relocs[i].tv.head,
@@ -177,8 +195,15 @@ static int radeon_cs_parser_relocs(struct radeon_cs_parser *p)
 	if (p->cs_flags & RADEON_CS_USE_VM)
 		p->vm_bos = radeon_vm_get_bos(p->rdev, p->ib.vm,
 					      &p->validated);
+	if (need_mmap_lock)
+		down_read(&current->mm->mmap_sem);
 
-	return radeon_bo_list_validate(p->rdev, &p->ticket, &p->validated, p->ring);
+	r = radeon_bo_list_validate(p->rdev, &p->ticket, &p->validated, p->ring);
+
+	if (need_mmap_lock)
+		up_read(&current->mm->mmap_sem);
+
+	return r;
 }
 
 static int radeon_cs_get_ring(struct radeon_cs_parser *p, u32 ring, s32 priority)
@@ -224,17 +249,24 @@ static int radeon_cs_get_ring(struct radeon_cs_parser *p, u32 ring, s32 priority
 	return 0;
 }
 
-static void radeon_cs_sync_rings(struct radeon_cs_parser *p)
+static int radeon_cs_sync_rings(struct radeon_cs_parser *p)
 {
-	int i;
+	int i, r = 0;
 
 	for (i = 0; i < p->nrelocs; i++) {
+		struct reservation_object *resv;
+
 		if (!p->relocs[i].robj)
 			continue;
 
-		radeon_semaphore_sync_to(p->ib.semaphore,
-					 p->relocs[i].robj->tbo.sync_obj);
+		resv = p->relocs[i].robj->tbo.resv;
+		r = radeon_semaphore_sync_resv(p->rdev, p->ib.semaphore, resv,
+					       p->relocs[i].tv.shared);
+
+		if (r)
+			break;
 	}
+	return r;
 }
 
 /* XXX: note that this is called from the legacy UMS CS ioctl as well */
@@ -403,7 +435,7 @@ static void radeon_cs_parser_fini(struct radeon_cs_parser *parser, int error, bo
 
 		ttm_eu_fence_buffer_objects(&parser->ticket,
 					    &parser->validated,
-					    parser->ib.fence);
+					    &parser->ib.fence->base);
 	} else if (backoff) {
 		ttm_eu_backoff_reservation(&parser->ticket,
 					   &parser->validated);
@@ -444,13 +476,19 @@ static int radeon_cs_ib_chunk(struct radeon_device *rdev,
 		return r;
 	}
 
+	r = radeon_cs_sync_rings(parser);
+	if (r) {
+		if (r != -ERESTARTSYS)
+			DRM_ERROR("Failed to sync rings: %i\n", r);
+		return r;
+	}
+
 	if (parser->ring == R600_RING_TYPE_UVD_INDEX)
 		radeon_uvd_note_usage(rdev);
 	else if ((parser->ring == TN_RING_TYPE_VCE1_INDEX) ||
 		 (parser->ring == TN_RING_TYPE_VCE2_INDEX))
 		radeon_vce_note_usage(rdev);
 
-	radeon_cs_sync_rings(parser);
 	r = radeon_ib_schedule(rdev, &parser->ib, NULL, true);
 	if (r) {
 		DRM_ERROR("Failed to schedule IB !\n");
@@ -537,8 +575,14 @@ static int radeon_cs_ib_vm_chunk(struct radeon_device *rdev,
 	if (r) {
 		goto out;
 	}
-	radeon_cs_sync_rings(parser);
-	radeon_semaphore_sync_to(parser->ib.semaphore, vm->fence);
+
+	r = radeon_cs_sync_rings(parser);
+	if (r) {
+		if (r != -ERESTARTSYS)
+			DRM_ERROR("Failed to sync rings: %i\n", r);
+		goto out;
+	}
+	radeon_semaphore_sync_fence(parser->ib.semaphore, vm->fence);
 
 	if ((rdev->family >= CHIP_TAHITI) &&
 	    (parser->chunk_const_ib_idx != -1)) {
@@ -629,6 +673,13 @@ int radeon_cs_ioctl(struct drm_device *dev, void *data, struct drm_file *filp)
 		up_read(&rdev->exclusive_lock);
 		return -EBUSY;
 	}
+	if (rdev->in_reset) {
+		up_read(&rdev->exclusive_lock);
+		r = radeon_gpu_reset(rdev);
+		if (!r)
+			r = -EAGAIN;
+		return r;
+	}
 	/* initialize parser */
 	memset(&parser, 0, sizeof(struct radeon_cs_parser));
 	parser.filp = filp;
diff --git a/drivers/gpu/drm/radeon/radeon_device.c b/drivers/gpu/drm/radeon/radeon_device.c
index 12c8329644c4..f41cc1538e48 100644
--- a/drivers/gpu/drm/radeon/radeon_device.c
+++ b/drivers/gpu/drm/radeon/radeon_device.c
@@ -434,7 +434,7 @@ int radeon_wb_init(struct radeon_device *rdev)
 
 	if (rdev->wb.wb_obj == NULL) {
 		r = radeon_bo_create(rdev, RADEON_GPU_PAGE_SIZE, PAGE_SIZE, true,
-				     RADEON_GEM_DOMAIN_GTT, 0, NULL,
+				     RADEON_GEM_DOMAIN_GTT, 0, NULL, NULL,
 				     &rdev->wb.wb_obj);
 		if (r) {
 			dev_warn(rdev->dev, "(%d) create WB bo failed\n", r);
@@ -1257,6 +1257,7 @@ int radeon_device_init(struct radeon_device *rdev,
 	for (i = 0; i < RADEON_NUM_RINGS; i++) {
 		rdev->ring[i].idx = i;
 	}
+	rdev->fence_context = fence_context_alloc(RADEON_NUM_RINGS);
 
 	DRM_INFO("initializing kernel modesetting (%s 0x%04X:0x%04X 0x%04X:0x%04X).\n",
 		radeon_family_name[rdev->family], pdev->vendor, pdev->device,
@@ -1274,6 +1275,8 @@ int radeon_device_init(struct radeon_device *rdev,
 	init_rwsem(&rdev->pm.mclk_lock);
 	init_rwsem(&rdev->exclusive_lock);
 	init_waitqueue_head(&rdev->irq.vblank_queue);
+	mutex_init(&rdev->mn_lock);
+	hash_init(rdev->mn_hash);
 	r = radeon_gem_init(rdev);
 	if (r)
 		return r;
@@ -1399,10 +1402,6 @@ int radeon_device_init(struct radeon_device *rdev,
 	if (r)
 		goto failed;
 
-	r = radeon_ib_ring_tests(rdev);
-	if (r)
-		DRM_ERROR("ib ring test failed (%d).\n", r);
-
 	r = radeon_gem_debugfs_init(rdev);
 	if (r) {
 		DRM_ERROR("registering gem debugfs failed (%d).\n", r);
@@ -1420,6 +1419,10 @@ int radeon_device_init(struct radeon_device *rdev,
 			goto failed;
 	}
 
+	r = radeon_ib_ring_tests(rdev);
+	if (r)
+		DRM_ERROR("ib ring test failed (%d).\n", r);
+
 	if ((radeon_testing & 1)) {
 		if (rdev->accel_working)
 			radeon_test_moves(rdev);
@@ -1497,7 +1500,6 @@ int radeon_suspend_kms(struct drm_device *dev, bool suspend, bool fbcon)
 	struct drm_crtc *crtc;
 	struct drm_connector *connector;
 	int i, r;
-	bool force_completion = false;
 
 	if (dev == NULL || dev->dev_private == NULL) {
 		return -ENODEV;
@@ -1541,12 +1543,9 @@ int radeon_suspend_kms(struct drm_device *dev, bool suspend, bool fbcon)
 		r = radeon_fence_wait_empty(rdev, i);
 		if (r) {
 			/* delay GPU reset to resume */
-			force_completion = true;
+			radeon_fence_driver_force_completion(rdev, i);
 		}
 	}
-	if (force_completion) {
-		radeon_fence_driver_force_completion(rdev);
-	}
 
 	radeon_save_bios_scratch_regs(rdev);
 
@@ -1686,8 +1685,6 @@ int radeon_gpu_reset(struct radeon_device *rdev)
 		return 0;
 	}
 
-	rdev->needs_reset = false;
-
 	radeon_save_bios_scratch_regs(rdev);
 	/* block TTM */
 	resched = ttm_bo_lock_delayed_workqueue(&rdev->mman.bdev);
@@ -1704,7 +1701,6 @@ int radeon_gpu_reset(struct radeon_device *rdev)
 		}
 	}
 
-retry:
 	r = radeon_asic_reset(rdev);
 	if (!r) {
 		dev_info(rdev->dev, "GPU reset succeeded, trying to resume\n");
@@ -1713,26 +1709,12 @@ retry:
 
 	radeon_restore_bios_scratch_regs(rdev);
 
-	if (!r) {
-		for (i = 0; i < RADEON_NUM_RINGS; ++i) {
+	for (i = 0; i < RADEON_NUM_RINGS; ++i) {
+		if (!r && ring_data[i]) {
 			radeon_ring_restore(rdev, &rdev->ring[i],
 					    ring_sizes[i], ring_data[i]);
-			ring_sizes[i] = 0;
-			ring_data[i] = NULL;
-		}
-
-		r = radeon_ib_ring_tests(rdev);
-		if (r) {
-			dev_err(rdev->dev, "ib ring test failed (%d).\n", r);
-			if (saved) {
-				saved = false;
-				radeon_suspend(rdev);
-				goto retry;
-			}
-		}
-	} else {
-		radeon_fence_driver_force_completion(rdev);
-		for (i = 0; i < RADEON_NUM_RINGS; ++i) {
+		} else {
+			radeon_fence_driver_force_completion(rdev, i);
 			kfree(ring_data[i]);
 		}
 	}
@@ -1764,19 +1746,32 @@ retry:
 	/* reset hpd state */
 	radeon_hpd_init(rdev);
 
+	ttm_bo_unlock_delayed_workqueue(&rdev->mman.bdev, resched);
+
+	rdev->in_reset = true;
+	rdev->needs_reset = false;
+
+	downgrade_write(&rdev->exclusive_lock);
+
 	drm_helper_resume_force_mode(rdev->ddev);
 
 	/* set the power state here in case we are a PX system or headless */
 	if ((rdev->pm.pm_method == PM_METHOD_DPM) && rdev->pm.dpm_enabled)
 		radeon_pm_compute_clocks(rdev);
 
-	ttm_bo_unlock_delayed_workqueue(&rdev->mman.bdev, resched);
-	if (r) {
+	if (!r) {
+		r = radeon_ib_ring_tests(rdev);
+		if (r && saved)
+			r = -EAGAIN;
+	} else {
 		/* bad news, how to tell it to userspace ? */
 		dev_info(rdev->dev, "GPU reset failed\n");
 	}
 
-	up_write(&rdev->exclusive_lock);
+	rdev->needs_reset = r == -EAGAIN;
+	rdev->in_reset = false;
+
+	up_read(&rdev->exclusive_lock);
 	return r;
 }
 
diff --git a/drivers/gpu/drm/radeon/radeon_display.c b/drivers/gpu/drm/radeon/radeon_display.c
index 3fdf87318069..00ead8c2758a 100644
--- a/drivers/gpu/drm/radeon/radeon_display.c
+++ b/drivers/gpu/drm/radeon/radeon_display.c
@@ -402,12 +402,21 @@ static void radeon_flip_work_func(struct work_struct *__work)
 
         down_read(&rdev->exclusive_lock);
 	if (work->fence) {
-		r = radeon_fence_wait(work->fence, false);
-		if (r == -EDEADLK) {
-			up_read(&rdev->exclusive_lock);
-			r = radeon_gpu_reset(rdev);
-			down_read(&rdev->exclusive_lock);
-		}
+		struct radeon_fence *fence;
+
+		fence = to_radeon_fence(work->fence);
+		if (fence && fence->rdev == rdev) {
+			r = radeon_fence_wait(fence, false);
+			if (r == -EDEADLK) {
+				up_read(&rdev->exclusive_lock);
+				do {
+					r = radeon_gpu_reset(rdev);
+				} while (r == -EAGAIN);
+				down_read(&rdev->exclusive_lock);
+			}
+		} else
+			r = fence_wait(work->fence, false);
+
 		if (r)
 			DRM_ERROR("failed to wait on page flip fence (%d)!\n", r);
 
@@ -416,7 +425,8 @@ static void radeon_flip_work_func(struct work_struct *__work)
 		 * confused about which BO the CRTC is scanning out
 		 */
 
-		radeon_fence_unref(&work->fence);
+		fence_put(work->fence);
+		work->fence = NULL;
 	}
 
 	/* We borrow the event spin lock for protecting flip_status */
@@ -474,11 +484,6 @@ static int radeon_crtc_page_flip(struct drm_crtc *crtc,
 	obj = new_radeon_fb->obj;
 	new_rbo = gem_to_radeon_bo(obj);
 
-	spin_lock(&new_rbo->tbo.bdev->fence_lock);
-	if (new_rbo->tbo.sync_obj)
-		work->fence = radeon_fence_ref(new_rbo->tbo.sync_obj);
-	spin_unlock(&new_rbo->tbo.bdev->fence_lock);
-
 	/* pin the new buffer */
 	DRM_DEBUG_DRIVER("flip-ioctl() cur_rbo = %p, new_rbo = %p\n",
 			 work->old_rbo, new_rbo);
@@ -497,6 +502,7 @@ static int radeon_crtc_page_flip(struct drm_crtc *crtc,
 		DRM_ERROR("failed to pin new rbo buffer before flip\n");
 		goto cleanup;
 	}
+	work->fence = fence_get(reservation_object_get_excl(new_rbo->tbo.resv));
 	radeon_bo_get_tiling_flags(new_rbo, &tiling_flags, NULL);
 	radeon_bo_unreserve(new_rbo);
 
@@ -578,9 +584,8 @@ pflip_cleanup:
 
 cleanup:
 	drm_gem_object_unreference_unlocked(&work->old_rbo->gem_base);
-	radeon_fence_unref(&work->fence);
+	fence_put(work->fence);
 	kfree(work);
-
 	return r;
 }
 
@@ -1917,7 +1922,7 @@ int radeon_get_crtc_scanoutpos(struct drm_device *dev, int crtc, unsigned int fl
 
 	/* In vblank? */
 	if (in_vbl)
-		ret |= DRM_SCANOUTPOS_INVBL;
+		ret |= DRM_SCANOUTPOS_IN_VBLANK;
 
 	/* Is vpos outside nominal vblank area, but less than
 	 * 1/100 of a frame height away from start of vblank?
diff --git a/drivers/gpu/drm/radeon/radeon_drv.c b/drivers/gpu/drm/radeon/radeon_drv.c
index f9d17b29b343..dcffa30ee2db 100644
--- a/drivers/gpu/drm/radeon/radeon_drv.c
+++ b/drivers/gpu/drm/radeon/radeon_drv.c
@@ -38,6 +38,8 @@
 #include <linux/module.h>
 #include <linux/pm_runtime.h>
 #include <linux/vga_switcheroo.h>
+#include <drm/drm_gem.h>
+
 #include "drm_crtc_helper.h"
 /*
  * KMS wrapper.
@@ -114,6 +116,9 @@ int radeon_gem_object_open(struct drm_gem_object *obj,
 				struct drm_file *file_priv);
 void radeon_gem_object_close(struct drm_gem_object *obj,
 				struct drm_file *file_priv);
+struct dma_buf *radeon_gem_prime_export(struct drm_device *dev,
+					struct drm_gem_object *gobj,
+					int flags);
 extern int radeon_get_crtc_scanoutpos(struct drm_device *dev, int crtc,
 				      unsigned int flags,
 				      int *vpos, int *hpos, ktime_t *stime,
@@ -130,7 +135,7 @@ int radeon_mode_dumb_create(struct drm_file *file_priv,
 			    struct drm_mode_create_dumb *args);
 struct sg_table *radeon_gem_prime_get_sg_table(struct drm_gem_object *obj);
 struct drm_gem_object *radeon_gem_prime_import_sg_table(struct drm_device *dev,
-							size_t size,
+							struct dma_buf_attachment *,
 							struct sg_table *sg);
 int radeon_gem_prime_pin(struct drm_gem_object *obj);
 void radeon_gem_prime_unpin(struct drm_gem_object *obj);
@@ -309,7 +314,7 @@ static const struct file_operations radeon_driver_old_fops = {
 	.open = drm_open,
 	.release = drm_release,
 	.unlocked_ioctl = drm_ioctl,
-	.mmap = drm_mmap,
+	.mmap = drm_legacy_mmap,
 	.poll = drm_poll,
 	.read = drm_read,
 #ifdef CONFIG_COMPAT
@@ -329,6 +334,7 @@ static struct drm_driver driver_old = {
 	.preclose = radeon_driver_preclose,
 	.postclose = radeon_driver_postclose,
 	.lastclose = radeon_driver_lastclose,
+	.set_busid = drm_pci_set_busid,
 	.unload = radeon_driver_unload,
 	.suspend = radeon_suspend,
 	.resume = radeon_resume,
@@ -553,6 +559,7 @@ static struct drm_driver kms_driver = {
 	.preclose = radeon_driver_preclose_kms,
 	.postclose = radeon_driver_postclose_kms,
 	.lastclose = radeon_driver_lastclose_kms,
+	.set_busid = drm_pci_set_busid,
 	.unload = radeon_driver_unload_kms,
 	.get_vblank_counter = radeon_get_vblank_counter_kms,
 	.enable_vblank = radeon_enable_vblank_kms,
@@ -578,7 +585,7 @@ static struct drm_driver kms_driver = {
 
 	.prime_handle_to_fd = drm_gem_prime_handle_to_fd,
 	.prime_fd_to_handle = drm_gem_prime_fd_to_handle,
-	.gem_prime_export = drm_gem_prime_export,
+	.gem_prime_export = radeon_gem_prime_export,
 	.gem_prime_import = drm_gem_prime_import,
 	.gem_prime_pin = radeon_gem_prime_pin,
 	.gem_prime_unpin = radeon_gem_prime_unpin,
diff --git a/drivers/gpu/drm/radeon/radeon_drv.h b/drivers/gpu/drm/radeon/radeon_drv.h
index dafd812e4571..46bd3938282c 100644
--- a/drivers/gpu/drm/radeon/radeon_drv.h
+++ b/drivers/gpu/drm/radeon/radeon_drv.h
@@ -33,7 +33,9 @@
 
 #include <linux/firmware.h>
 #include <linux/platform_device.h>
+#include <drm/drm_legacy.h>
 
+#include <drm/ati_pcigart.h>
 #include "radeon_family.h"
 
 /* General customization:
diff --git a/drivers/gpu/drm/radeon/radeon_encoders.c b/drivers/gpu/drm/radeon/radeon_encoders.c
index 15edf23b465c..9a19e52cc655 100644
--- a/drivers/gpu/drm/radeon/radeon_encoders.c
+++ b/drivers/gpu/drm/radeon/radeon_encoders.c
@@ -410,3 +410,24 @@ bool radeon_dig_monitor_is_duallink(struct drm_encoder *encoder,
 	}
 }
 
+bool radeon_encoder_is_digital(struct drm_encoder *encoder)
+{
+	struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder);
+	switch (radeon_encoder->encoder_id) {
+	case ENCODER_OBJECT_ID_INTERNAL_LVDS:
+	case ENCODER_OBJECT_ID_INTERNAL_TMDS1:
+	case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_TMDS1:
+	case ENCODER_OBJECT_ID_INTERNAL_LVTM1:
+	case ENCODER_OBJECT_ID_INTERNAL_DVO1:
+	case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_DVO1:
+	case ENCODER_OBJECT_ID_INTERNAL_DDI:
+	case ENCODER_OBJECT_ID_INTERNAL_UNIPHY:
+	case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_LVTMA:
+	case ENCODER_OBJECT_ID_INTERNAL_UNIPHY1:
+	case ENCODER_OBJECT_ID_INTERNAL_UNIPHY2:
+	case ENCODER_OBJECT_ID_INTERNAL_UNIPHY3:
+		return true;
+	default:
+		return false;
+	}
+}
diff --git a/drivers/gpu/drm/radeon/radeon_fb.c b/drivers/gpu/drm/radeon/radeon_fb.c
index 94b0f2aa3d7c..0ea1db83d573 100644
--- a/drivers/gpu/drm/radeon/radeon_fb.c
+++ b/drivers/gpu/drm/radeon/radeon_fb.c
@@ -189,7 +189,8 @@ out_unref:
 static int radeonfb_create(struct drm_fb_helper *helper,
 			   struct drm_fb_helper_surface_size *sizes)
 {
-	struct radeon_fbdev *rfbdev = (struct radeon_fbdev *)helper;
+	struct radeon_fbdev *rfbdev =
+		container_of(helper, struct radeon_fbdev, helper);
 	struct radeon_device *rdev = rfbdev->rdev;
 	struct fb_info *info;
 	struct drm_framebuffer *fb = NULL;
diff --git a/drivers/gpu/drm/radeon/radeon_fence.c b/drivers/gpu/drm/radeon/radeon_fence.c
index 913787085dfa..995167025282 100644
--- a/drivers/gpu/drm/radeon/radeon_fence.c
+++ b/drivers/gpu/drm/radeon/radeon_fence.c
@@ -98,6 +98,25 @@ static u32 radeon_fence_read(struct radeon_device *rdev, int ring)
 }
 
 /**
+ * radeon_fence_schedule_check - schedule lockup check
+ *
+ * @rdev: radeon_device pointer
+ * @ring: ring index we should work with
+ *
+ * Queues a delayed work item to check for lockups.
+ */
+static void radeon_fence_schedule_check(struct radeon_device *rdev, int ring)
+{
+	/*
+	 * Do not reset the timer here with mod_delayed_work,
+	 * this can livelock in an interaction with TTM delayed destroy.
+	 */
+	queue_delayed_work(system_power_efficient_wq,
+			   &rdev->fence_drv[ring].lockup_work,
+			   RADEON_FENCE_JIFFIES_TIMEOUT);
+}
+
+/**
  * radeon_fence_emit - emit a fence on the requested ring
  *
  * @rdev: radeon_device pointer
@@ -111,30 +130,70 @@ int radeon_fence_emit(struct radeon_device *rdev,
 		      struct radeon_fence **fence,
 		      int ring)
 {
+	u64 seq = ++rdev->fence_drv[ring].sync_seq[ring];
+
 	/* we are protected by the ring emission mutex */
 	*fence = kmalloc(sizeof(struct radeon_fence), GFP_KERNEL);
 	if ((*fence) == NULL) {
 		return -ENOMEM;
 	}
-	kref_init(&((*fence)->kref));
 	(*fence)->rdev = rdev;
-	(*fence)->seq = ++rdev->fence_drv[ring].sync_seq[ring];
+	(*fence)->seq = seq;
 	(*fence)->ring = ring;
+	fence_init(&(*fence)->base, &radeon_fence_ops,
+		   &rdev->fence_queue.lock, rdev->fence_context + ring, seq);
 	radeon_fence_ring_emit(rdev, ring, *fence);
 	trace_radeon_fence_emit(rdev->ddev, ring, (*fence)->seq);
+	radeon_fence_schedule_check(rdev, ring);
 	return 0;
 }
 
 /**
- * radeon_fence_process - process a fence
+ * radeon_fence_check_signaled - callback from fence_queue
+ *
+ * this function is called with fence_queue lock held, which is also used
+ * for the fence locking itself, so unlocked variants are used for
+ * fence_signal, and remove_wait_queue.
+ */
+static int radeon_fence_check_signaled(wait_queue_t *wait, unsigned mode, int flags, void *key)
+{
+	struct radeon_fence *fence;
+	u64 seq;
+
+	fence = container_of(wait, struct radeon_fence, fence_wake);
+
+	/*
+	 * We cannot use radeon_fence_process here because we're already
+	 * in the waitqueue, in a call from wake_up_all.
+	 */
+	seq = atomic64_read(&fence->rdev->fence_drv[fence->ring].last_seq);
+	if (seq >= fence->seq) {
+		int ret = fence_signal_locked(&fence->base);
+
+		if (!ret)
+			FENCE_TRACE(&fence->base, "signaled from irq context\n");
+		else
+			FENCE_TRACE(&fence->base, "was already signaled\n");
+
+		radeon_irq_kms_sw_irq_put(fence->rdev, fence->ring);
+		__remove_wait_queue(&fence->rdev->fence_queue, &fence->fence_wake);
+		fence_put(&fence->base);
+	} else
+		FENCE_TRACE(&fence->base, "pending\n");
+	return 0;
+}
+
+/**
+ * radeon_fence_activity - check for fence activity
  *
  * @rdev: radeon_device pointer
  * @ring: ring index the fence is associated with
  *
- * Checks the current fence value and wakes the fence queue
- * if the sequence number has increased (all asics).
+ * Checks the current fence value and calculates the last
+ * signalled fence value. Returns true if activity occured
+ * on the ring, and the fence_queue should be waken up.
  */
-void radeon_fence_process(struct radeon_device *rdev, int ring)
+static bool radeon_fence_activity(struct radeon_device *rdev, int ring)
 {
 	uint64_t seq, last_seq, last_emitted;
 	unsigned count_loop = 0;
@@ -190,23 +249,77 @@ void radeon_fence_process(struct radeon_device *rdev, int ring)
 		}
 	} while (atomic64_xchg(&rdev->fence_drv[ring].last_seq, seq) > seq);
 
-	if (wake)
-		wake_up_all(&rdev->fence_queue);
+	if (seq < last_emitted)
+		radeon_fence_schedule_check(rdev, ring);
+
+	return wake;
 }
 
 /**
- * radeon_fence_destroy - destroy a fence
+ * radeon_fence_check_lockup - check for hardware lockup
  *
- * @kref: fence kref
+ * @work: delayed work item
  *
- * Frees the fence object (all asics).
+ * Checks for fence activity and if there is none probe
+ * the hardware if a lockup occured.
  */
-static void radeon_fence_destroy(struct kref *kref)
+static void radeon_fence_check_lockup(struct work_struct *work)
 {
-	struct radeon_fence *fence;
+	struct radeon_fence_driver *fence_drv;
+	struct radeon_device *rdev;
+	int ring;
+
+	fence_drv = container_of(work, struct radeon_fence_driver,
+				 lockup_work.work);
+	rdev = fence_drv->rdev;
+	ring = fence_drv - &rdev->fence_drv[0];
+
+	if (!down_read_trylock(&rdev->exclusive_lock)) {
+		/* just reschedule the check if a reset is going on */
+		radeon_fence_schedule_check(rdev, ring);
+		return;
+	}
+
+	if (fence_drv->delayed_irq && rdev->ddev->irq_enabled) {
+		unsigned long irqflags;
+
+		fence_drv->delayed_irq = false;
+		spin_lock_irqsave(&rdev->irq.lock, irqflags);
+		radeon_irq_set(rdev);
+		spin_unlock_irqrestore(&rdev->irq.lock, irqflags);
+	}
+
+	if (radeon_fence_activity(rdev, ring))
+		wake_up_all(&rdev->fence_queue);
 
-	fence = container_of(kref, struct radeon_fence, kref);
-	kfree(fence);
+	else if (radeon_ring_is_lockup(rdev, ring, &rdev->ring[ring])) {
+
+		/* good news we believe it's a lockup */
+		dev_warn(rdev->dev, "GPU lockup (current fence id "
+			 "0x%016llx last fence id 0x%016llx on ring %d)\n",
+			 (uint64_t)atomic64_read(&fence_drv->last_seq),
+			 fence_drv->sync_seq[ring], ring);
+
+		/* remember that we need an reset */
+		rdev->needs_reset = true;
+		wake_up_all(&rdev->fence_queue);
+	}
+	up_read(&rdev->exclusive_lock);
+}
+
+/**
+ * radeon_fence_process - process a fence
+ *
+ * @rdev: radeon_device pointer
+ * @ring: ring index the fence is associated with
+ *
+ * Checks the current fence value and wakes the fence queue
+ * if the sequence number has increased (all asics).
+ */
+void radeon_fence_process(struct radeon_device *rdev, int ring)
+{
+	if (radeon_fence_activity(rdev, ring))
+		wake_up_all(&rdev->fence_queue);
 }
 
 /**
@@ -237,6 +350,75 @@ static bool radeon_fence_seq_signaled(struct radeon_device *rdev,
 	return false;
 }
 
+static bool radeon_fence_is_signaled(struct fence *f)
+{
+	struct radeon_fence *fence = to_radeon_fence(f);
+	struct radeon_device *rdev = fence->rdev;
+	unsigned ring = fence->ring;
+	u64 seq = fence->seq;
+
+	if (atomic64_read(&rdev->fence_drv[ring].last_seq) >= seq) {
+		return true;
+	}
+
+	if (down_read_trylock(&rdev->exclusive_lock)) {
+		radeon_fence_process(rdev, ring);
+		up_read(&rdev->exclusive_lock);
+
+		if (atomic64_read(&rdev->fence_drv[ring].last_seq) >= seq) {
+			return true;
+		}
+	}
+	return false;
+}
+
+/**
+ * radeon_fence_enable_signaling - enable signalling on fence
+ * @fence: fence
+ *
+ * This function is called with fence_queue lock held, and adds a callback
+ * to fence_queue that checks if this fence is signaled, and if so it
+ * signals the fence and removes itself.
+ */
+static bool radeon_fence_enable_signaling(struct fence *f)
+{
+	struct radeon_fence *fence = to_radeon_fence(f);
+	struct radeon_device *rdev = fence->rdev;
+
+	if (atomic64_read(&rdev->fence_drv[fence->ring].last_seq) >= fence->seq)
+		return false;
+
+	if (down_read_trylock(&rdev->exclusive_lock)) {
+		radeon_irq_kms_sw_irq_get(rdev, fence->ring);
+
+		if (radeon_fence_activity(rdev, fence->ring))
+			wake_up_all_locked(&rdev->fence_queue);
+
+		/* did fence get signaled after we enabled the sw irq? */
+		if (atomic64_read(&rdev->fence_drv[fence->ring].last_seq) >= fence->seq) {
+			radeon_irq_kms_sw_irq_put(rdev, fence->ring);
+			up_read(&rdev->exclusive_lock);
+			return false;
+		}
+
+		up_read(&rdev->exclusive_lock);
+	} else {
+		/* we're probably in a lockup, lets not fiddle too much */
+		if (radeon_irq_kms_sw_irq_get_delayed(rdev, fence->ring))
+			rdev->fence_drv[fence->ring].delayed_irq = true;
+		radeon_fence_schedule_check(rdev, fence->ring);
+	}
+
+	fence->fence_wake.flags = 0;
+	fence->fence_wake.private = NULL;
+	fence->fence_wake.func = radeon_fence_check_signaled;
+	__add_wait_queue(&rdev->fence_queue, &fence->fence_wake);
+	fence_get(f);
+
+	FENCE_TRACE(&fence->base, "armed on ring %i!\n", fence->ring);
+	return true;
+}
+
 /**
  * radeon_fence_signaled - check if a fence has signaled
  *
@@ -247,14 +429,15 @@ static bool radeon_fence_seq_signaled(struct radeon_device *rdev,
  */
 bool radeon_fence_signaled(struct radeon_fence *fence)
 {
-	if (!fence) {
+	if (!fence)
 		return true;
-	}
-	if (fence->seq == RADEON_FENCE_SIGNALED_SEQ) {
-		return true;
-	}
+
 	if (radeon_fence_seq_signaled(fence->rdev, fence->seq, fence->ring)) {
-		fence->seq = RADEON_FENCE_SIGNALED_SEQ;
+		int ret;
+
+		ret = fence_signal(&fence->base);
+		if (!ret)
+			FENCE_TRACE(&fence->base, "signaled from radeon_fence_signaled\n");
 		return true;
 	}
 	return false;
@@ -283,110 +466,70 @@ static bool radeon_fence_any_seq_signaled(struct radeon_device *rdev, u64 *seq)
 }
 
 /**
- * radeon_fence_wait_seq - wait for a specific sequence numbers
+ * radeon_fence_wait_seq_timeout - wait for a specific sequence numbers
  *
  * @rdev: radeon device pointer
  * @target_seq: sequence number(s) we want to wait for
  * @intr: use interruptable sleep
+ * @timeout: maximum time to wait, or MAX_SCHEDULE_TIMEOUT for infinite wait
  *
  * Wait for the requested sequence number(s) to be written by any ring
  * (all asics).  Sequnce number array is indexed by ring id.
  * @intr selects whether to use interruptable (true) or non-interruptable
  * (false) sleep when waiting for the sequence number.  Helper function
  * for radeon_fence_wait_*().
- * Returns 0 if the sequence number has passed, error for all other cases.
+ * Returns remaining time if the sequence number has passed, 0 when
+ * the wait timeout, or an error for all other cases.
  * -EDEADLK is returned when a GPU lockup has been detected.
  */
-static int radeon_fence_wait_seq(struct radeon_device *rdev, u64 *target_seq,
-				 bool intr)
+static long radeon_fence_wait_seq_timeout(struct radeon_device *rdev,
+					  u64 *target_seq, bool intr,
+					  long timeout)
 {
-	uint64_t last_seq[RADEON_NUM_RINGS];
-	bool signaled;
-	int i, r;
-
-	while (!radeon_fence_any_seq_signaled(rdev, target_seq)) {
+	long r;
+	int i;
 
-		/* Save current sequence values, used to check for GPU lockups */
-		for (i = 0; i < RADEON_NUM_RINGS; ++i) {
-			if (!target_seq[i])
-				continue;
+	if (radeon_fence_any_seq_signaled(rdev, target_seq))
+		return timeout;
 
-			last_seq[i] = atomic64_read(&rdev->fence_drv[i].last_seq);
-			trace_radeon_fence_wait_begin(rdev->ddev, i, target_seq[i]);
-			radeon_irq_kms_sw_irq_get(rdev, i);
-		}
+	/* enable IRQs and tracing */
+	for (i = 0; i < RADEON_NUM_RINGS; ++i) {
+		if (!target_seq[i])
+			continue;
 
-		if (intr) {
-			r = wait_event_interruptible_timeout(rdev->fence_queue, (
-				(signaled = radeon_fence_any_seq_signaled(rdev, target_seq))
-				 || rdev->needs_reset), RADEON_FENCE_JIFFIES_TIMEOUT);
-		} else {
-			r = wait_event_timeout(rdev->fence_queue, (
-				(signaled = radeon_fence_any_seq_signaled(rdev, target_seq))
-				 || rdev->needs_reset), RADEON_FENCE_JIFFIES_TIMEOUT);
-		}
+		trace_radeon_fence_wait_begin(rdev->ddev, i, target_seq[i]);
+		radeon_irq_kms_sw_irq_get(rdev, i);
+	}
 
-		for (i = 0; i < RADEON_NUM_RINGS; ++i) {
-			if (!target_seq[i])
-				continue;
+	if (intr) {
+		r = wait_event_interruptible_timeout(rdev->fence_queue, (
+			radeon_fence_any_seq_signaled(rdev, target_seq)
+			 || rdev->needs_reset), timeout);
+	} else {
+		r = wait_event_timeout(rdev->fence_queue, (
+			radeon_fence_any_seq_signaled(rdev, target_seq)
+			 || rdev->needs_reset), timeout);
+	}
 
-			radeon_irq_kms_sw_irq_put(rdev, i);
-			trace_radeon_fence_wait_end(rdev->ddev, i, target_seq[i]);
-		}
+	if (rdev->needs_reset)
+		r = -EDEADLK;
 
-		if (unlikely(r < 0))
-			return r;
+	for (i = 0; i < RADEON_NUM_RINGS; ++i) {
+		if (!target_seq[i])
+			continue;
 
-		if (unlikely(!signaled)) {
-			if (rdev->needs_reset)
-				return -EDEADLK;
-
-			/* we were interrupted for some reason and fence
-			 * isn't signaled yet, resume waiting */
-			if (r)
-				continue;
-
-			for (i = 0; i < RADEON_NUM_RINGS; ++i) {
-				if (!target_seq[i])
-					continue;
-
-				if (last_seq[i] != atomic64_read(&rdev->fence_drv[i].last_seq))
-					break;
-			}
-
-			if (i != RADEON_NUM_RINGS)
-				continue;
-
-			for (i = 0; i < RADEON_NUM_RINGS; ++i) {
-				if (!target_seq[i])
-					continue;
-
-				if (radeon_ring_is_lockup(rdev, i, &rdev->ring[i]))
-					break;
-			}
-
-			if (i < RADEON_NUM_RINGS) {
-				/* good news we believe it's a lockup */
-				dev_warn(rdev->dev, "GPU lockup (waiting for "
-					 "0x%016llx last fence id 0x%016llx on"
-					 " ring %d)\n",
-					 target_seq[i], last_seq[i], i);
-
-				/* remember that we need an reset */
-				rdev->needs_reset = true;
-				wake_up_all(&rdev->fence_queue);
-				return -EDEADLK;
-			}
-		}
+		radeon_irq_kms_sw_irq_put(rdev, i);
+		trace_radeon_fence_wait_end(rdev->ddev, i, target_seq[i]);
 	}
-	return 0;
+
+	return r;
 }
 
 /**
  * radeon_fence_wait - wait for a fence to signal
  *
  * @fence: radeon fence object
- * @intr: use interruptable sleep
+ * @intr: use interruptible sleep
  *
  * Wait for the requested fence to signal (all asics).
  * @intr selects whether to use interruptable (true) or non-interruptable
@@ -396,22 +539,26 @@ static int radeon_fence_wait_seq(struct radeon_device *rdev, u64 *target_seq,
 int radeon_fence_wait(struct radeon_fence *fence, bool intr)
 {
 	uint64_t seq[RADEON_NUM_RINGS] = {};
-	int r;
+	long r;
 
-	if (fence == NULL) {
-		WARN(1, "Querying an invalid fence : %p !\n", fence);
-		return -EINVAL;
-	}
+	/*
+	 * This function should not be called on !radeon fences.
+	 * If this is the case, it would mean this function can
+	 * also be called on radeon fences belonging to another card.
+	 * exclusive_lock is not held in that case.
+	 */
+	if (WARN_ON_ONCE(!to_radeon_fence(&fence->base)))
+		return fence_wait(&fence->base, intr);
 
 	seq[fence->ring] = fence->seq;
-	if (seq[fence->ring] == RADEON_FENCE_SIGNALED_SEQ)
-		return 0;
-
-	r = radeon_fence_wait_seq(fence->rdev, seq, intr);
-	if (r)
+	r = radeon_fence_wait_seq_timeout(fence->rdev, seq, intr, MAX_SCHEDULE_TIMEOUT);
+	if (r < 0) {
 		return r;
+	}
 
-	fence->seq = RADEON_FENCE_SIGNALED_SEQ;
+	r = fence_signal(&fence->base);
+	if (!r)
+		FENCE_TRACE(&fence->base, "signaled from fence_wait\n");
 	return 0;
 }
 
@@ -434,7 +581,7 @@ int radeon_fence_wait_any(struct radeon_device *rdev,
 {
 	uint64_t seq[RADEON_NUM_RINGS];
 	unsigned i, num_rings = 0;
-	int r;
+	long r;
 
 	for (i = 0; i < RADEON_NUM_RINGS; ++i) {
 		seq[i] = 0;
@@ -445,18 +592,14 @@ int radeon_fence_wait_any(struct radeon_device *rdev,
 
 		seq[i] = fences[i]->seq;
 		++num_rings;
-
-		/* test if something was allready signaled */
-		if (seq[i] == RADEON_FENCE_SIGNALED_SEQ)
-			return 0;
 	}
 
 	/* nothing to wait for ? */
 	if (num_rings == 0)
 		return -ENOENT;
 
-	r = radeon_fence_wait_seq(rdev, seq, intr);
-	if (r) {
+	r = radeon_fence_wait_seq_timeout(rdev, seq, intr, MAX_SCHEDULE_TIMEOUT);
+	if (r < 0) {
 		return r;
 	}
 	return 0;
@@ -475,6 +618,7 @@ int radeon_fence_wait_any(struct radeon_device *rdev,
 int radeon_fence_wait_next(struct radeon_device *rdev, int ring)
 {
 	uint64_t seq[RADEON_NUM_RINGS] = {};
+	long r;
 
 	seq[ring] = atomic64_read(&rdev->fence_drv[ring].last_seq) + 1ULL;
 	if (seq[ring] >= rdev->fence_drv[ring].sync_seq[ring]) {
@@ -482,7 +626,10 @@ int radeon_fence_wait_next(struct radeon_device *rdev, int ring)
 		   already the last emited fence */
 		return -ENOENT;
 	}
-	return radeon_fence_wait_seq(rdev, seq, false);
+	r = radeon_fence_wait_seq_timeout(rdev, seq, false, MAX_SCHEDULE_TIMEOUT);
+	if (r < 0)
+		return r;
+	return 0;
 }
 
 /**
@@ -498,18 +645,18 @@ int radeon_fence_wait_next(struct radeon_device *rdev, int ring)
 int radeon_fence_wait_empty(struct radeon_device *rdev, int ring)
 {
 	uint64_t seq[RADEON_NUM_RINGS] = {};
-	int r;
+	long r;
 
 	seq[ring] = rdev->fence_drv[ring].sync_seq[ring];
 	if (!seq[ring])
 		return 0;
 
-	r = radeon_fence_wait_seq(rdev, seq, false);
-	if (r) {
+	r = radeon_fence_wait_seq_timeout(rdev, seq, false, MAX_SCHEDULE_TIMEOUT);
+	if (r < 0) {
 		if (r == -EDEADLK)
 			return -EDEADLK;
 
-		dev_err(rdev->dev, "error waiting for ring[%d] to become idle (%d)\n",
+		dev_err(rdev->dev, "error waiting for ring[%d] to become idle (%ld)\n",
 			ring, r);
 	}
 	return 0;
@@ -525,7 +672,7 @@ int radeon_fence_wait_empty(struct radeon_device *rdev, int ring)
  */
 struct radeon_fence *radeon_fence_ref(struct radeon_fence *fence)
 {
-	kref_get(&fence->kref);
+	fence_get(&fence->base);
 	return fence;
 }
 
@@ -542,7 +689,7 @@ void radeon_fence_unref(struct radeon_fence **fence)
 
 	*fence = NULL;
 	if (tmp) {
-		kref_put(&tmp->kref, radeon_fence_destroy);
+		fence_put(&tmp->base);
 	}
 }
 
@@ -711,6 +858,9 @@ static void radeon_fence_driver_init_ring(struct radeon_device *rdev, int ring)
 		rdev->fence_drv[ring].sync_seq[i] = 0;
 	atomic64_set(&rdev->fence_drv[ring].last_seq, 0);
 	rdev->fence_drv[ring].initialized = false;
+	INIT_DELAYED_WORK(&rdev->fence_drv[ring].lockup_work,
+			  radeon_fence_check_lockup);
+	rdev->fence_drv[ring].rdev = rdev;
 }
 
 /**
@@ -758,8 +908,9 @@ void radeon_fence_driver_fini(struct radeon_device *rdev)
 		r = radeon_fence_wait_empty(rdev, ring);
 		if (r) {
 			/* no need to trigger GPU reset as we are unloading */
-			radeon_fence_driver_force_completion(rdev);
+			radeon_fence_driver_force_completion(rdev, ring);
 		}
+		cancel_delayed_work_sync(&rdev->fence_drv[ring].lockup_work);
 		wake_up_all(&rdev->fence_queue);
 		radeon_scratch_free(rdev, rdev->fence_drv[ring].scratch_reg);
 		rdev->fence_drv[ring].initialized = false;
@@ -771,18 +922,16 @@ void radeon_fence_driver_fini(struct radeon_device *rdev)
  * radeon_fence_driver_force_completion - force all fence waiter to complete
  *
  * @rdev: radeon device pointer
+ * @ring: the ring to complete
  *
  * In case of GPU reset failure make sure no process keep waiting on fence
  * that will never complete.
  */
-void radeon_fence_driver_force_completion(struct radeon_device *rdev)
+void radeon_fence_driver_force_completion(struct radeon_device *rdev, int ring)
 {
-	int ring;
-
-	for (ring = 0; ring < RADEON_NUM_RINGS; ring++) {
-		if (!rdev->fence_drv[ring].initialized)
-			continue;
+	if (rdev->fence_drv[ring].initialized) {
 		radeon_fence_write(rdev, rdev->fence_drv[ring].sync_seq[ring], ring);
+		cancel_delayed_work_sync(&rdev->fence_drv[ring].lockup_work);
 	}
 }
 
@@ -833,6 +982,7 @@ static int radeon_debugfs_gpu_reset(struct seq_file *m, void *data)
 	down_read(&rdev->exclusive_lock);
 	seq_printf(m, "%d\n", rdev->needs_reset);
 	rdev->needs_reset = true;
+	wake_up_all(&rdev->fence_queue);
 	up_read(&rdev->exclusive_lock);
 
 	return 0;
@@ -852,3 +1002,72 @@ int radeon_debugfs_fence_init(struct radeon_device *rdev)
 	return 0;
 #endif
 }
+
+static const char *radeon_fence_get_driver_name(struct fence *fence)
+{
+	return "radeon";
+}
+
+static const char *radeon_fence_get_timeline_name(struct fence *f)
+{
+	struct radeon_fence *fence = to_radeon_fence(f);
+	switch (fence->ring) {
+	case RADEON_RING_TYPE_GFX_INDEX: return "radeon.gfx";
+	case CAYMAN_RING_TYPE_CP1_INDEX: return "radeon.cp1";
+	case CAYMAN_RING_TYPE_CP2_INDEX: return "radeon.cp2";
+	case R600_RING_TYPE_DMA_INDEX: return "radeon.dma";
+	case CAYMAN_RING_TYPE_DMA1_INDEX: return "radeon.dma1";
+	case R600_RING_TYPE_UVD_INDEX: return "radeon.uvd";
+	case TN_RING_TYPE_VCE1_INDEX: return "radeon.vce1";
+	case TN_RING_TYPE_VCE2_INDEX: return "radeon.vce2";
+	default: WARN_ON_ONCE(1); return "radeon.unk";
+	}
+}
+
+static inline bool radeon_test_signaled(struct radeon_fence *fence)
+{
+	return test_bit(FENCE_FLAG_SIGNALED_BIT, &fence->base.flags);
+}
+
+static signed long radeon_fence_default_wait(struct fence *f, bool intr,
+					     signed long t)
+{
+	struct radeon_fence *fence = to_radeon_fence(f);
+	struct radeon_device *rdev = fence->rdev;
+	bool signaled;
+
+	fence_enable_sw_signaling(&fence->base);
+
+	/*
+	 * This function has to return -EDEADLK, but cannot hold
+	 * exclusive_lock during the wait because some callers
+	 * may already hold it. This means checking needs_reset without
+	 * lock, and not fiddling with any gpu internals.
+	 *
+	 * The callback installed with fence_enable_sw_signaling will
+	 * run before our wait_event_*timeout call, so we will see
+	 * both the signaled fence and the changes to needs_reset.
+	 */
+
+	if (intr)
+		t = wait_event_interruptible_timeout(rdev->fence_queue,
+			((signaled = radeon_test_signaled(fence)) ||
+			 rdev->needs_reset), t);
+	else
+		t = wait_event_timeout(rdev->fence_queue,
+			((signaled = radeon_test_signaled(fence)) ||
+			 rdev->needs_reset), t);
+
+	if (t > 0 && !signaled)
+		return -EDEADLK;
+	return t;
+}
+
+const struct fence_ops radeon_fence_ops = {
+	.get_driver_name = radeon_fence_get_driver_name,
+	.get_timeline_name = radeon_fence_get_timeline_name,
+	.enable_signaling = radeon_fence_enable_signaling,
+	.signaled = radeon_fence_is_signaled,
+	.wait = radeon_fence_default_wait,
+	.release = NULL,
+};
diff --git a/drivers/gpu/drm/radeon/radeon_gart.c b/drivers/gpu/drm/radeon/radeon_gart.c
index a053a0779aac..84146d5901aa 100644
--- a/drivers/gpu/drm/radeon/radeon_gart.c
+++ b/drivers/gpu/drm/radeon/radeon_gart.c
@@ -128,7 +128,7 @@ int radeon_gart_table_vram_alloc(struct radeon_device *rdev)
 	if (rdev->gart.robj == NULL) {
 		r = radeon_bo_create(rdev, rdev->gart.table_size,
 				     PAGE_SIZE, true, RADEON_GEM_DOMAIN_VRAM,
-				     0, NULL, &rdev->gart.robj);
+				     0, NULL, NULL, &rdev->gart.robj);
 		if (r) {
 			return r;
 		}
diff --git a/drivers/gpu/drm/radeon/radeon_gem.c b/drivers/gpu/drm/radeon/radeon_gem.c
index bfd7e1b0ff3f..c194497aa586 100644
--- a/drivers/gpu/drm/radeon/radeon_gem.c
+++ b/drivers/gpu/drm/radeon/radeon_gem.c
@@ -67,7 +67,7 @@ int radeon_gem_object_create(struct radeon_device *rdev, unsigned long size,
 
 retry:
 	r = radeon_bo_create(rdev, size, alignment, kernel, initial_domain,
-			     flags, NULL, &robj);
+			     flags, NULL, NULL, &robj);
 	if (r) {
 		if (r != -ERESTARTSYS) {
 			if (initial_domain == RADEON_GEM_DOMAIN_VRAM) {
@@ -94,7 +94,7 @@ static int radeon_gem_set_domain(struct drm_gem_object *gobj,
 {
 	struct radeon_bo *robj;
 	uint32_t domain;
-	int r;
+	long r;
 
 	/* FIXME: reeimplement */
 	robj = gem_to_radeon_bo(gobj);
@@ -110,9 +110,12 @@ static int radeon_gem_set_domain(struct drm_gem_object *gobj,
 	}
 	if (domain == RADEON_GEM_DOMAIN_CPU) {
 		/* Asking for cpu access wait for object idle */
-		r = radeon_bo_wait(robj, NULL, false);
-		if (r) {
-			printk(KERN_ERR "Failed to wait for object !\n");
+		r = reservation_object_wait_timeout_rcu(robj->tbo.resv, true, true, 30 * HZ);
+		if (!r)
+			r = -EBUSY;
+
+		if (r < 0 && r != -EINTR) {
+			printk(KERN_ERR "Failed to wait for object: %li\n", r);
 			return r;
 		}
 	}
@@ -272,6 +275,94 @@ int radeon_gem_create_ioctl(struct drm_device *dev, void *data,
 	return 0;
 }
 
+int radeon_gem_userptr_ioctl(struct drm_device *dev, void *data,
+			     struct drm_file *filp)
+{
+	struct radeon_device *rdev = dev->dev_private;
+	struct drm_radeon_gem_userptr *args = data;
+	struct drm_gem_object *gobj;
+	struct radeon_bo *bo;
+	uint32_t handle;
+	int r;
+
+	if (offset_in_page(args->addr | args->size))
+		return -EINVAL;
+
+	/* reject unknown flag values */
+	if (args->flags & ~(RADEON_GEM_USERPTR_READONLY |
+	    RADEON_GEM_USERPTR_ANONONLY | RADEON_GEM_USERPTR_VALIDATE |
+	    RADEON_GEM_USERPTR_REGISTER))
+		return -EINVAL;
+
+	if (args->flags & RADEON_GEM_USERPTR_READONLY) {
+		/* readonly pages not tested on older hardware */
+		if (rdev->family < CHIP_R600)
+			return -EINVAL;
+
+	} else if (!(args->flags & RADEON_GEM_USERPTR_ANONONLY) ||
+		   !(args->flags & RADEON_GEM_USERPTR_REGISTER)) {
+
+		/* if we want to write to it we must require anonymous
+		   memory and install a MMU notifier */
+		return -EACCES;
+	}
+
+	down_read(&rdev->exclusive_lock);
+
+	/* create a gem object to contain this object in */
+	r = radeon_gem_object_create(rdev, args->size, 0,
+				     RADEON_GEM_DOMAIN_CPU, 0,
+				     false, &gobj);
+	if (r)
+		goto handle_lockup;
+
+	bo = gem_to_radeon_bo(gobj);
+	r = radeon_ttm_tt_set_userptr(bo->tbo.ttm, args->addr, args->flags);
+	if (r)
+		goto release_object;
+
+	if (args->flags & RADEON_GEM_USERPTR_REGISTER) {
+		r = radeon_mn_register(bo, args->addr);
+		if (r)
+			goto release_object;
+	}
+
+	if (args->flags & RADEON_GEM_USERPTR_VALIDATE) {
+		down_read(&current->mm->mmap_sem);
+		r = radeon_bo_reserve(bo, true);
+		if (r) {
+			up_read(&current->mm->mmap_sem);
+			goto release_object;
+		}
+
+		radeon_ttm_placement_from_domain(bo, RADEON_GEM_DOMAIN_GTT);
+		r = ttm_bo_validate(&bo->tbo, &bo->placement, true, false);
+		radeon_bo_unreserve(bo);
+		up_read(&current->mm->mmap_sem);
+		if (r)
+			goto release_object;
+	}
+
+	r = drm_gem_handle_create(filp, gobj, &handle);
+	/* drop reference from allocate - handle holds it now */
+	drm_gem_object_unreference_unlocked(gobj);
+	if (r)
+		goto handle_lockup;
+
+	args->handle = handle;
+	up_read(&rdev->exclusive_lock);
+	return 0;
+
+release_object:
+	drm_gem_object_unreference_unlocked(gobj);
+
+handle_lockup:
+	up_read(&rdev->exclusive_lock);
+	r = radeon_gem_handle_lockup(rdev, r);
+
+	return r;
+}
+
 int radeon_gem_set_domain_ioctl(struct drm_device *dev, void *data,
 				struct drm_file *filp)
 {
@@ -315,6 +406,10 @@ int radeon_mode_dumb_mmap(struct drm_file *filp,
 		return -ENOENT;
 	}
 	robj = gem_to_radeon_bo(gobj);
+	if (radeon_ttm_tt_has_userptr(robj->tbo.ttm)) {
+		drm_gem_object_unreference_unlocked(gobj);
+		return -EPERM;
+	}
 	*offset_p = radeon_bo_mmap_offset(robj);
 	drm_gem_object_unreference_unlocked(gobj);
 	return 0;
@@ -357,15 +452,22 @@ int radeon_gem_wait_idle_ioctl(struct drm_device *dev, void *data,
 	struct drm_radeon_gem_wait_idle *args = data;
 	struct drm_gem_object *gobj;
 	struct radeon_bo *robj;
-	int r;
+	int r = 0;
 	uint32_t cur_placement = 0;
+	long ret;
 
 	gobj = drm_gem_object_lookup(dev, filp, args->handle);
 	if (gobj == NULL) {
 		return -ENOENT;
 	}
 	robj = gem_to_radeon_bo(gobj);
-	r = radeon_bo_wait(robj, &cur_placement, false);
+
+	ret = reservation_object_wait_timeout_rcu(robj->tbo.resv, true, true, 30 * HZ);
+	if (ret == 0)
+		r = -EBUSY;
+	else if (ret < 0)
+		r = ret;
+
 	/* Flush HDP cache via MMIO if necessary */
 	if (rdev->asic->mmio_hdp_flush &&
 	    radeon_mem_type_to_domain(cur_placement) == RADEON_GEM_DOMAIN_VRAM)
@@ -532,6 +634,11 @@ int radeon_gem_op_ioctl(struct drm_device *dev, void *data,
 		return -ENOENT;
 	}
 	robj = gem_to_radeon_bo(gobj);
+
+	r = -EPERM;
+	if (radeon_ttm_tt_has_userptr(robj->tbo.ttm))
+		goto out;
+
 	r = radeon_bo_reserve(robj, false);
 	if (unlikely(r))
 		goto out;
diff --git a/drivers/gpu/drm/radeon/radeon_ib.c b/drivers/gpu/drm/radeon/radeon_ib.c
index 5bf2c0a05827..3f39fcca4d07 100644
--- a/drivers/gpu/drm/radeon/radeon_ib.c
+++ b/drivers/gpu/drm/radeon/radeon_ib.c
@@ -145,7 +145,7 @@ int radeon_ib_schedule(struct radeon_device *rdev, struct radeon_ib *ib,
 	if (ib->vm) {
 		struct radeon_fence *vm_id_fence;
 		vm_id_fence = radeon_vm_grab_id(rdev, ib->vm, ib->ring);
-        	radeon_semaphore_sync_to(ib->semaphore, vm_id_fence);
+		radeon_semaphore_sync_fence(ib->semaphore, vm_id_fence);
 	}
 
 	/* sync with other rings */
@@ -269,6 +269,7 @@ int radeon_ib_ring_tests(struct radeon_device *rdev)
 
 		r = radeon_ib_test(rdev, i, ring);
 		if (r) {
+			radeon_fence_driver_force_completion(rdev, i);
 			ring->ready = false;
 			rdev->needs_reset = false;
 
diff --git a/drivers/gpu/drm/radeon/radeon_irq_kms.c b/drivers/gpu/drm/radeon/radeon_irq_kms.c
index 16807afab362..7784911d78ef 100644
--- a/drivers/gpu/drm/radeon/radeon_irq_kms.c
+++ b/drivers/gpu/drm/radeon/radeon_irq_kms.c
@@ -88,23 +88,6 @@ static void radeon_hotplug_work_func(struct work_struct *work)
 }
 
 /**
- * radeon_irq_reset_work_func - execute gpu reset
- *
- * @work: work struct
- *
- * Execute scheduled gpu reset (cayman+).
- * This function is called when the irq handler
- * thinks we need a gpu reset.
- */
-static void radeon_irq_reset_work_func(struct work_struct *work)
-{
-	struct radeon_device *rdev = container_of(work, struct radeon_device,
-						  reset_work);
-
-	radeon_gpu_reset(rdev);
-}
-
-/**
  * radeon_driver_irq_preinstall_kms - drm irq preinstall callback
  *
  * @dev: drm dev pointer
@@ -284,7 +267,6 @@ int radeon_irq_kms_init(struct radeon_device *rdev)
 
 	INIT_WORK(&rdev->hotplug_work, radeon_hotplug_work_func);
 	INIT_WORK(&rdev->audio_work, r600_audio_update_hdmi);
-	INIT_WORK(&rdev->reset_work, radeon_irq_reset_work_func);
 
 	rdev->irq.installed = true;
 	r = drm_irq_install(rdev->ddev, rdev->ddev->pdev->irq);
@@ -342,6 +324,21 @@ void radeon_irq_kms_sw_irq_get(struct radeon_device *rdev, int ring)
 }
 
 /**
+ * radeon_irq_kms_sw_irq_get_delayed - enable software interrupt
+ *
+ * @rdev: radeon device pointer
+ * @ring: ring whose interrupt you want to enable
+ *
+ * Enables the software interrupt for a specific ring (all asics).
+ * The software interrupt is generally used to signal a fence on
+ * a particular ring.
+ */
+bool radeon_irq_kms_sw_irq_get_delayed(struct radeon_device *rdev, int ring)
+{
+	return atomic_inc_return(&rdev->irq.ring_int[ring]) == 1;
+}
+
+/**
  * radeon_irq_kms_sw_irq_put - disable software interrupt
  *
  * @rdev: radeon device pointer
diff --git a/drivers/gpu/drm/radeon/radeon_kms.c b/drivers/gpu/drm/radeon/radeon_kms.c
index eb7164d07985..8309b11e674d 100644
--- a/drivers/gpu/drm/radeon/radeon_kms.c
+++ b/drivers/gpu/drm/radeon/radeon_kms.c
@@ -885,5 +885,6 @@ const struct drm_ioctl_desc radeon_ioctls_kms[] = {
 	DRM_IOCTL_DEF_DRV(RADEON_GEM_BUSY, radeon_gem_busy_ioctl, DRM_AUTH|DRM_UNLOCKED|DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(RADEON_GEM_VA, radeon_gem_va_ioctl, DRM_AUTH|DRM_UNLOCKED|DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(RADEON_GEM_OP, radeon_gem_op_ioctl, DRM_AUTH|DRM_UNLOCKED|DRM_RENDER_ALLOW),
+	DRM_IOCTL_DEF_DRV(RADEON_GEM_USERPTR, radeon_gem_userptr_ioctl, DRM_AUTH|DRM_UNLOCKED|DRM_RENDER_ALLOW),
 };
 int radeon_max_kms_ioctl = ARRAY_SIZE(radeon_ioctls_kms);
diff --git a/drivers/gpu/drm/radeon/radeon_mn.c b/drivers/gpu/drm/radeon/radeon_mn.c
new file mode 100644
index 000000000000..a69bd441dd2d
--- /dev/null
+++ b/drivers/gpu/drm/radeon/radeon_mn.c
@@ -0,0 +1,274 @@
+/*
+ * Copyright 2014 Advanced Micro Devices, Inc.
+ * All Rights Reserved.
+ *
+ * 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 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.
+ *
+ * The above copyright notice and this permission notice (including the
+ * next paragraph) shall be included in all copies or substantial portions
+ * of the Software.
+ *
+ */
+/*
+ * Authors:
+ *    Christian König <christian.koenig@amd.com>
+ */
+
+#include <linux/firmware.h>
+#include <linux/module.h>
+#include <linux/mmu_notifier.h>
+#include <drm/drmP.h>
+#include <drm/drm.h>
+
+#include "radeon.h"
+
+struct radeon_mn {
+	/* constant after initialisation */
+	struct radeon_device	*rdev;
+	struct mm_struct	*mm;
+	struct mmu_notifier	mn;
+
+	/* only used on destruction */
+	struct work_struct	work;
+
+	/* protected by rdev->mn_lock */
+	struct hlist_node	node;
+
+	/* objects protected by lock */
+	struct mutex		lock;
+	struct rb_root		objects;
+};
+
+/**
+ * radeon_mn_destroy - destroy the rmn
+ *
+ * @work: previously sheduled work item
+ *
+ * Lazy destroys the notifier from a work item
+ */
+static void radeon_mn_destroy(struct work_struct *work)
+{
+	struct radeon_mn *rmn = container_of(work, struct radeon_mn, work);
+	struct radeon_device *rdev = rmn->rdev;
+	struct radeon_bo *bo, *next;
+
+	mutex_lock(&rdev->mn_lock);
+	mutex_lock(&rmn->lock);
+	hash_del(&rmn->node);
+	rbtree_postorder_for_each_entry_safe(bo, next, &rmn->objects, mn_it.rb) {
+		interval_tree_remove(&bo->mn_it, &rmn->objects);
+		bo->mn = NULL;
+	}
+	mutex_unlock(&rmn->lock);
+	mutex_unlock(&rdev->mn_lock);
+	mmu_notifier_unregister(&rmn->mn, rmn->mm);
+	kfree(rmn);
+}
+
+/**
+ * radeon_mn_release - callback to notify about mm destruction
+ *
+ * @mn: our notifier
+ * @mn: the mm this callback is about
+ *
+ * Shedule a work item to lazy destroy our notifier.
+ */
+static void radeon_mn_release(struct mmu_notifier *mn,
+			      struct mm_struct *mm)
+{
+	struct radeon_mn *rmn = container_of(mn, struct radeon_mn, mn);
+	INIT_WORK(&rmn->work, radeon_mn_destroy);
+	schedule_work(&rmn->work);
+}
+
+/**
+ * radeon_mn_invalidate_range_start - 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
+ *
+ * We block for all BOs between start and end to be idle and
+ * unmap them by move them into system domain again.
+ */
+static void radeon_mn_invalidate_range_start(struct mmu_notifier *mn,
+					     struct mm_struct *mm,
+					     unsigned long start,
+					     unsigned long end)
+{
+	struct radeon_mn *rmn = container_of(mn, struct radeon_mn, mn);
+	struct interval_tree_node *it;
+
+	/* notification is exclusive, but interval is inclusive */
+	end -= 1;
+
+	mutex_lock(&rmn->lock);
+
+	it = interval_tree_iter_first(&rmn->objects, start, end);
+	while (it) {
+		struct radeon_bo *bo;
+		struct fence *fence;
+		int r;
+
+		bo = container_of(it, struct radeon_bo, mn_it);
+		it = interval_tree_iter_next(it, start, end);
+
+		r = radeon_bo_reserve(bo, true);
+		if (r) {
+			DRM_ERROR("(%d) failed to reserve user bo\n", r);
+			continue;
+		}
+
+		fence = reservation_object_get_excl(bo->tbo.resv);
+		if (fence) {
+			r = radeon_fence_wait((struct radeon_fence *)fence, false);
+			if (r)
+				DRM_ERROR("(%d) 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, false, false);
+		if (r)
+			DRM_ERROR("(%d) failed to validate user bo\n", r);
+
+		radeon_bo_unreserve(bo);
+	}
+	
+	mutex_unlock(&rmn->lock);
+}
+
+static const struct mmu_notifier_ops radeon_mn_ops = {
+	.release = radeon_mn_release,
+	.invalidate_range_start = radeon_mn_invalidate_range_start,
+};
+
+/**
+ * radeon_mn_get - create notifier context
+ *
+ * @rdev: radeon device pointer
+ *
+ * Creates a notifier context for current->mm.
+ */
+static struct radeon_mn *radeon_mn_get(struct radeon_device *rdev)
+{
+	struct mm_struct *mm = current->mm;
+	struct radeon_mn *rmn;
+	int r;
+
+	down_write(&mm->mmap_sem);
+	mutex_lock(&rdev->mn_lock);
+
+	hash_for_each_possible(rdev->mn_hash, rmn, node, (unsigned long)mm)
+		if (rmn->mm == mm)
+			goto release_locks;
+
+	rmn = kzalloc(sizeof(*rmn), GFP_KERNEL);
+	if (!rmn) {
+		rmn = ERR_PTR(-ENOMEM);
+		goto release_locks;
+	}
+
+	rmn->rdev = rdev;
+	rmn->mm = mm;
+	rmn->mn.ops = &radeon_mn_ops;
+	mutex_init(&rmn->lock);
+	rmn->objects = RB_ROOT;
+	
+	r = __mmu_notifier_register(&rmn->mn, mm);
+	if (r)
+		goto free_rmn;
+
+	hash_add(rdev->mn_hash, &rmn->node, (unsigned long)mm);
+
+release_locks:
+	mutex_unlock(&rdev->mn_lock);
+	up_write(&mm->mmap_sem);
+
+	return rmn;
+
+free_rmn:
+	mutex_unlock(&rdev->mn_lock);
+	up_write(&mm->mmap_sem);
+	kfree(rmn);
+
+	return ERR_PTR(r);
+}
+
+/**
+ * radeon_mn_register - register a BO for notifier updates
+ *
+ * @bo: radeon buffer object
+ * @addr: userptr addr we should monitor
+ *
+ * Registers an MMU notifier for the given BO at the specified address.
+ * Returns 0 on success, -ERRNO if anything goes wrong.
+ */
+int radeon_mn_register(struct radeon_bo *bo, unsigned long addr)
+{
+	unsigned long end = addr + radeon_bo_size(bo) - 1;
+	struct radeon_device *rdev = bo->rdev;
+	struct radeon_mn *rmn;
+	struct interval_tree_node *it;
+
+	rmn = radeon_mn_get(rdev);
+	if (IS_ERR(rmn))
+		return PTR_ERR(rmn);
+
+	mutex_lock(&rmn->lock);
+
+	it = interval_tree_iter_first(&rmn->objects, addr, end);
+	if (it) {
+		mutex_unlock(&rmn->lock);
+		return -EEXIST;
+	}
+
+	bo->mn = rmn;
+	bo->mn_it.start = addr;
+	bo->mn_it.last = end;
+	interval_tree_insert(&bo->mn_it, &rmn->objects);
+
+	mutex_unlock(&rmn->lock);
+
+	return 0;
+}
+
+/**
+ * radeon_mn_unregister - unregister a BO for notifier updates
+ *
+ * @bo: radeon buffer object
+ *
+ * Remove any registration of MMU notifier updates from the buffer object.
+ */
+void radeon_mn_unregister(struct radeon_bo *bo)
+{
+	struct radeon_device *rdev = bo->rdev;
+	struct radeon_mn *rmn;
+
+	mutex_lock(&rdev->mn_lock);
+	rmn = bo->mn;
+	if (rmn == NULL) {
+		mutex_unlock(&rdev->mn_lock);
+		return;
+	}
+
+	mutex_lock(&rmn->lock);
+	interval_tree_remove(&bo->mn_it, &rmn->objects);
+	bo->mn = NULL;
+	mutex_unlock(&rmn->lock);
+	mutex_unlock(&rdev->mn_lock);
+}
diff --git a/drivers/gpu/drm/radeon/radeon_mode.h b/drivers/gpu/drm/radeon/radeon_mode.h
index e27608c29c11..04db2fdd8692 100644
--- a/drivers/gpu/drm/radeon/radeon_mode.h
+++ b/drivers/gpu/drm/radeon/radeon_mode.h
@@ -777,6 +777,7 @@ extern void atombios_digital_setup(struct drm_encoder *encoder, int action);
 extern int atombios_get_encoder_mode(struct drm_encoder *encoder);
 extern bool atombios_set_edp_panel_power(struct drm_connector *connector, int action);
 extern void radeon_encoder_set_active_device(struct drm_encoder *encoder);
+extern bool radeon_encoder_is_digital(struct drm_encoder *encoder);
 
 extern void radeon_crtc_load_lut(struct drm_crtc *crtc);
 extern int atombios_crtc_set_base(struct drm_crtc *crtc, int x, int y,
diff --git a/drivers/gpu/drm/radeon/radeon_object.c b/drivers/gpu/drm/radeon/radeon_object.c
index 480c87d8edc5..99a960a4f302 100644
--- a/drivers/gpu/drm/radeon/radeon_object.c
+++ b/drivers/gpu/drm/radeon/radeon_object.c
@@ -75,6 +75,7 @@ static void radeon_ttm_bo_destroy(struct ttm_buffer_object *tbo)
 	bo = container_of(tbo, struct radeon_bo, tbo);
 
 	radeon_update_memory_usage(bo, bo->tbo.mem.mem_type, -1);
+	radeon_mn_unregister(bo);
 
 	mutex_lock(&bo->rdev->gem.mutex);
 	list_del_init(&bo->list);
@@ -96,55 +97,80 @@ void radeon_ttm_placement_from_domain(struct radeon_bo *rbo, u32 domain)
 {
 	u32 c = 0, i;
 
-	rbo->placement.fpfn = 0;
-	rbo->placement.lpfn = 0;
 	rbo->placement.placement = rbo->placements;
 	rbo->placement.busy_placement = rbo->placements;
 	if (domain & RADEON_GEM_DOMAIN_VRAM)
-		rbo->placements[c++] = TTM_PL_FLAG_WC | TTM_PL_FLAG_UNCACHED |
-					TTM_PL_FLAG_VRAM;
+		rbo->placements[c++].flags = TTM_PL_FLAG_WC |
+					     TTM_PL_FLAG_UNCACHED |
+					     TTM_PL_FLAG_VRAM;
+
 	if (domain & RADEON_GEM_DOMAIN_GTT) {
 		if (rbo->flags & RADEON_GEM_GTT_UC) {
-			rbo->placements[c++] = TTM_PL_FLAG_UNCACHED | TTM_PL_FLAG_TT;
+			rbo->placements[c++].flags = TTM_PL_FLAG_UNCACHED |
+				TTM_PL_FLAG_TT;
+
 		} else if ((rbo->flags & RADEON_GEM_GTT_WC) ||
 			   (rbo->rdev->flags & RADEON_IS_AGP)) {
-			rbo->placements[c++] = TTM_PL_FLAG_WC | TTM_PL_FLAG_UNCACHED |
+			rbo->placements[c++].flags = TTM_PL_FLAG_WC |
+				TTM_PL_FLAG_UNCACHED |
 				TTM_PL_FLAG_TT;
 		} else {
-			rbo->placements[c++] = TTM_PL_FLAG_CACHED | TTM_PL_FLAG_TT;
+			rbo->placements[c++].flags = TTM_PL_FLAG_CACHED |
+						     TTM_PL_FLAG_TT;
 		}
 	}
+
 	if (domain & RADEON_GEM_DOMAIN_CPU) {
 		if (rbo->flags & RADEON_GEM_GTT_UC) {
-			rbo->placements[c++] = TTM_PL_FLAG_UNCACHED | TTM_PL_FLAG_SYSTEM;
+			rbo->placements[c++].flags = TTM_PL_FLAG_UNCACHED |
+				TTM_PL_FLAG_SYSTEM;
+
 		} else if ((rbo->flags & RADEON_GEM_GTT_WC) ||
 		    rbo->rdev->flags & RADEON_IS_AGP) {
-			rbo->placements[c++] = TTM_PL_FLAG_WC | TTM_PL_FLAG_UNCACHED |
+			rbo->placements[c++].flags = TTM_PL_FLAG_WC |
+				TTM_PL_FLAG_UNCACHED |
 				TTM_PL_FLAG_SYSTEM;
 		} else {
-			rbo->placements[c++] = TTM_PL_FLAG_CACHED | TTM_PL_FLAG_SYSTEM;
+			rbo->placements[c++].flags = TTM_PL_FLAG_CACHED |
+						     TTM_PL_FLAG_SYSTEM;
 		}
 	}
 	if (!c)
-		rbo->placements[c++] = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
+		rbo->placements[c++].flags = TTM_PL_MASK_CACHING |
+					     TTM_PL_FLAG_SYSTEM;
+
 	rbo->placement.num_placement = c;
 	rbo->placement.num_busy_placement = c;
 
+	for (i = 0; i < c; ++i) {
+		rbo->placements[i].fpfn = 0;
+		if ((rbo->flags & RADEON_GEM_CPU_ACCESS) &&
+		    (rbo->placements[i].flags & TTM_PL_FLAG_VRAM))
+			rbo->placements[i].lpfn =
+				rbo->rdev->mc.visible_vram_size >> PAGE_SHIFT;
+		else
+			rbo->placements[i].lpfn = 0;
+	}
+
 	/*
 	 * Use two-ended allocation depending on the buffer size to
 	 * improve fragmentation quality.
 	 * 512kb was measured as the most optimal number.
 	 */
-	if (rbo->tbo.mem.size > 512 * 1024) {
+	if (!((rbo->flags & RADEON_GEM_CPU_ACCESS) &&
+	      (rbo->placements[i].flags & TTM_PL_FLAG_VRAM)) &&
+	    rbo->tbo.mem.size > 512 * 1024) {
 		for (i = 0; i < c; i++) {
-			rbo->placements[i] |= TTM_PL_FLAG_TOPDOWN;
+			rbo->placements[i].flags |= TTM_PL_FLAG_TOPDOWN;
 		}
 	}
 }
 
 int radeon_bo_create(struct radeon_device *rdev,
-		     unsigned long size, int byte_align, bool kernel, u32 domain,
-		     u32 flags, struct sg_table *sg, struct radeon_bo **bo_ptr)
+		     unsigned long size, int byte_align, bool kernel,
+		     u32 domain, u32 flags, struct sg_table *sg,
+		     struct reservation_object *resv,
+		     struct radeon_bo **bo_ptr)
 {
 	struct radeon_bo *bo;
 	enum ttm_bo_type type;
@@ -192,7 +218,7 @@ int radeon_bo_create(struct radeon_device *rdev,
 	down_read(&rdev->pm.mclk_lock);
 	r = ttm_bo_init(&rdev->mman.bdev, &bo->tbo, size, type,
 			&bo->placement, page_align, !kernel, NULL,
-			acc_size, sg, &radeon_ttm_bo_destroy);
+			acc_size, sg, resv, &radeon_ttm_bo_destroy);
 	up_read(&rdev->pm.mclk_lock);
 	if (unlikely(r != 0)) {
 		return r;
@@ -264,6 +290,9 @@ int radeon_bo_pin_restricted(struct radeon_bo *bo, u32 domain, u64 max_offset,
 {
 	int r, i;
 
+	if (radeon_ttm_tt_has_userptr(bo->tbo.ttm))
+		return -EPERM;
+
 	if (bo->pin_count) {
 		bo->pin_count++;
 		if (gpu_addr)
@@ -283,21 +312,19 @@ int radeon_bo_pin_restricted(struct radeon_bo *bo, u32 domain, u64 max_offset,
 		return 0;
 	}
 	radeon_ttm_placement_from_domain(bo, domain);
-	if (domain == RADEON_GEM_DOMAIN_VRAM) {
+	for (i = 0; i < bo->placement.num_placement; i++) {
 		/* force to pin into visible video ram */
-		bo->placement.lpfn = bo->rdev->mc.visible_vram_size >> PAGE_SHIFT;
-	}
-	if (max_offset) {
-		u64 lpfn = max_offset >> PAGE_SHIFT;
-
-		if (!bo->placement.lpfn)
-			bo->placement.lpfn = bo->rdev->mc.gtt_size >> PAGE_SHIFT;
+		if ((bo->placements[i].flags & TTM_PL_FLAG_VRAM) &&
+		    !(bo->flags & RADEON_GEM_NO_CPU_ACCESS) &&
+		    (!max_offset || max_offset > bo->rdev->mc.visible_vram_size))
+			bo->placements[i].lpfn =
+				bo->rdev->mc.visible_vram_size >> PAGE_SHIFT;
+		else
+			bo->placements[i].lpfn = max_offset >> PAGE_SHIFT;
 
-		if (lpfn < bo->placement.lpfn)
-			bo->placement.lpfn = lpfn;
+		bo->placements[i].flags |= TTM_PL_FLAG_NO_EVICT;
 	}
-	for (i = 0; i < bo->placement.num_placement; i++)
-		bo->placements[i] |= TTM_PL_FLAG_NO_EVICT;
+
 	r = ttm_bo_validate(&bo->tbo, &bo->placement, false, false);
 	if (likely(r == 0)) {
 		bo->pin_count = 1;
@@ -329,8 +356,10 @@ int radeon_bo_unpin(struct radeon_bo *bo)
 	bo->pin_count--;
 	if (bo->pin_count)
 		return 0;
-	for (i = 0; i < bo->placement.num_placement; i++)
-		bo->placements[i] &= ~TTM_PL_FLAG_NO_EVICT;
+	for (i = 0; i < bo->placement.num_placement; i++) {
+		bo->placements[i].lpfn = 0;
+		bo->placements[i].flags &= ~TTM_PL_FLAG_NO_EVICT;
+	}
 	r = ttm_bo_validate(&bo->tbo, &bo->placement, false, false);
 	if (likely(r == 0)) {
 		if (bo->tbo.mem.mem_type == TTM_PL_VRAM)
@@ -459,7 +488,7 @@ int radeon_bo_list_validate(struct radeon_device *rdev,
 	u64 bytes_moved = 0, initial_bytes_moved;
 	u64 bytes_moved_threshold = radeon_bo_get_threshold_for_moves(rdev);
 
-	r = ttm_eu_reserve_buffers(ticket, head);
+	r = ttm_eu_reserve_buffers(ticket, head, true);
 	if (unlikely(r != 0)) {
 		return r;
 	}
@@ -468,6 +497,7 @@ int radeon_bo_list_validate(struct radeon_device *rdev,
 		bo = lobj->robj;
 		if (!bo->pin_count) {
 			u32 domain = lobj->prefered_domains;
+			u32 allowed = lobj->allowed_domains;
 			u32 current_domain =
 				radeon_mem_type_to_domain(bo->tbo.mem.mem_type);
 
@@ -479,7 +509,7 @@ int radeon_bo_list_validate(struct radeon_device *rdev,
 			 * into account. We don't want to disallow buffer moves
 			 * completely.
 			 */
-			if ((lobj->allowed_domains & current_domain) != 0 &&
+			if ((allowed & current_domain) != 0 &&
 			    (domain & current_domain) == 0 && /* will be moved */
 			    bytes_moved > bytes_moved_threshold) {
 				/* don't move it */
@@ -489,7 +519,7 @@ int radeon_bo_list_validate(struct radeon_device *rdev,
 		retry:
 			radeon_ttm_placement_from_domain(bo, domain);
 			if (ring == R600_RING_TYPE_UVD_INDEX)
-				radeon_uvd_force_into_uvd_segment(bo);
+				radeon_uvd_force_into_uvd_segment(bo, allowed);
 
 			initial_bytes_moved = atomic64_read(&rdev->num_bytes_moved);
 			r = ttm_bo_validate(&bo->tbo, &bo->placement, true, false);
@@ -731,7 +761,7 @@ int radeon_bo_fault_reserve_notify(struct ttm_buffer_object *bo)
 
 	/* hurrah the memory is not visible ! */
 	radeon_ttm_placement_from_domain(rbo, RADEON_GEM_DOMAIN_VRAM);
-	rbo->placement.lpfn = rdev->mc.visible_vram_size >> PAGE_SHIFT;
+	rbo->placements[0].lpfn = rdev->mc.visible_vram_size >> PAGE_SHIFT;
 	r = ttm_bo_validate(bo, &rbo->placement, false, false);
 	if (unlikely(r == -ENOMEM)) {
 		radeon_ttm_placement_from_domain(rbo, RADEON_GEM_DOMAIN_GTT);
@@ -755,12 +785,10 @@ int radeon_bo_wait(struct radeon_bo *bo, u32 *mem_type, bool no_wait)
 	r = ttm_bo_reserve(&bo->tbo, true, no_wait, false, NULL);
 	if (unlikely(r != 0))
 		return r;
-	spin_lock(&bo->tbo.bdev->fence_lock);
 	if (mem_type)
 		*mem_type = bo->tbo.mem.mem_type;
-	if (bo->tbo.sync_obj)
-		r = ttm_bo_wait(&bo->tbo, true, true, no_wait);
-	spin_unlock(&bo->tbo.bdev->fence_lock);
+
+	r = ttm_bo_wait(&bo->tbo, true, true, no_wait);
 	ttm_bo_unreserve(&bo->tbo);
 	return r;
 }
diff --git a/drivers/gpu/drm/radeon/radeon_object.h b/drivers/gpu/drm/radeon/radeon_object.h
index 98a47fdf3625..1b8ec7917154 100644
--- a/drivers/gpu/drm/radeon/radeon_object.h
+++ b/drivers/gpu/drm/radeon/radeon_object.h
@@ -126,6 +126,7 @@ extern int radeon_bo_create(struct radeon_device *rdev,
 			    unsigned long size, int byte_align,
 			    bool kernel, u32 domain, u32 flags,
 			    struct sg_table *sg,
+			    struct reservation_object *resv,
 			    struct radeon_bo **bo_ptr);
 extern int radeon_bo_kmap(struct radeon_bo *bo, void **ptr);
 extern void radeon_bo_kunmap(struct radeon_bo *bo);
diff --git a/drivers/gpu/drm/radeon/radeon_pm.c b/drivers/gpu/drm/radeon/radeon_pm.c
index 164898b0010c..32522cc940a1 100644
--- a/drivers/gpu/drm/radeon/radeon_pm.c
+++ b/drivers/gpu/drm/radeon/radeon_pm.c
@@ -1556,7 +1556,7 @@ static bool radeon_pm_in_vbl(struct radeon_device *rdev)
 		if (rdev->pm.active_crtcs & (1 << crtc)) {
 			vbl_status = radeon_get_crtc_scanoutpos(rdev->ddev, crtc, 0, &vpos, &hpos, NULL, NULL);
 			if ((vbl_status & DRM_SCANOUTPOS_VALID) &&
-			    !(vbl_status & DRM_SCANOUTPOS_INVBL))
+			    !(vbl_status & DRM_SCANOUTPOS_IN_VBLANK))
 				in_vbl = false;
 		}
 	}
diff --git a/drivers/gpu/drm/radeon/radeon_prime.c b/drivers/gpu/drm/radeon/radeon_prime.c
index 0b16f2cbcf17..f3609c97496b 100644
--- a/drivers/gpu/drm/radeon/radeon_prime.c
+++ b/drivers/gpu/drm/radeon/radeon_prime.c
@@ -27,6 +27,7 @@
 
 #include "radeon.h"
 #include <drm/radeon_drm.h>
+#include <linux/dma-buf.h>
 
 struct sg_table *radeon_gem_prime_get_sg_table(struct drm_gem_object *obj)
 {
@@ -57,15 +58,18 @@ void radeon_gem_prime_vunmap(struct drm_gem_object *obj, void *vaddr)
 }
 
 struct drm_gem_object *radeon_gem_prime_import_sg_table(struct drm_device *dev,
-							size_t size,
+							struct dma_buf_attachment *attach,
 							struct sg_table *sg)
 {
+	struct reservation_object *resv = attach->dmabuf->resv;
 	struct radeon_device *rdev = dev->dev_private;
 	struct radeon_bo *bo;
 	int ret;
 
-	ret = radeon_bo_create(rdev, size, PAGE_SIZE, false,
-			       RADEON_GEM_DOMAIN_GTT, 0, sg, &bo);
+	ww_mutex_lock(&resv->lock, NULL);
+	ret = radeon_bo_create(rdev, attach->dmabuf->size, PAGE_SIZE, false,
+			       RADEON_GEM_DOMAIN_GTT, 0, sg, resv, &bo);
+	ww_mutex_unlock(&resv->lock);
 	if (ret)
 		return ERR_PTR(ret);
 
@@ -111,3 +115,13 @@ struct reservation_object *radeon_gem_prime_res_obj(struct drm_gem_object *obj)
 
 	return bo->tbo.resv;
 }
+
+struct dma_buf *radeon_gem_prime_export(struct drm_device *dev,
+					struct drm_gem_object *gobj,
+					int flags)
+{
+	struct radeon_bo *bo = gem_to_radeon_bo(gobj);
+	if (radeon_ttm_tt_has_userptr(bo->tbo.ttm))
+		return ERR_PTR(-EPERM);
+	return drm_gem_prime_export(dev, gobj, flags);
+}
diff --git a/drivers/gpu/drm/radeon/radeon_ring.c b/drivers/gpu/drm/radeon/radeon_ring.c
index d65607902537..3d17af34afa7 100644
--- a/drivers/gpu/drm/radeon/radeon_ring.c
+++ b/drivers/gpu/drm/radeon/radeon_ring.c
@@ -45,27 +45,6 @@
 static int radeon_debugfs_ring_init(struct radeon_device *rdev, struct radeon_ring *ring);
 
 /**
- * radeon_ring_write - write a value to the ring
- *
- * @ring: radeon_ring structure holding ring information
- * @v: dword (dw) value to write
- *
- * Write a value to the requested ring buffer (all asics).
- */
-void radeon_ring_write(struct radeon_ring *ring, uint32_t v)
-{
-#if DRM_DEBUG_CODE
-	if (ring->count_dw <= 0) {
-		DRM_ERROR("radeon: writing more dwords to the ring than expected!\n");
-	}
-#endif
-	ring->ring[ring->wptr++] = v;
-	ring->wptr &= ring->ptr_mask;
-	ring->count_dw--;
-	ring->ring_free_dw--;
-}
-
-/**
  * radeon_ring_supports_scratch_reg - check if the ring supports
  * writing to scratch registers
  *
@@ -404,7 +383,7 @@ int radeon_ring_init(struct radeon_device *rdev, struct radeon_ring *ring, unsig
 	/* Allocate ring buffer */
 	if (ring->ring_obj == NULL) {
 		r = radeon_bo_create(rdev, ring->ring_size, PAGE_SIZE, true,
-				     RADEON_GEM_DOMAIN_GTT, 0,
+				     RADEON_GEM_DOMAIN_GTT, 0, NULL,
 				     NULL, &ring->ring_obj);
 		if (r) {
 			dev_err(rdev->dev, "(%d) ring create failed\n", r);
diff --git a/drivers/gpu/drm/radeon/radeon_sa.c b/drivers/gpu/drm/radeon/radeon_sa.c
index b84f97c8718c..c507896aca45 100644
--- a/drivers/gpu/drm/radeon/radeon_sa.c
+++ b/drivers/gpu/drm/radeon/radeon_sa.c
@@ -65,7 +65,7 @@ int radeon_sa_bo_manager_init(struct radeon_device *rdev,
 	}
 
 	r = radeon_bo_create(rdev, size, align, true,
-			     domain, flags, NULL, &sa_manager->bo);
+			     domain, flags, NULL, NULL, &sa_manager->bo);
 	if (r) {
 		dev_err(rdev->dev, "(%d) failed to allocate bo for manager\n", r);
 		return r;
diff --git a/drivers/gpu/drm/radeon/radeon_semaphore.c b/drivers/gpu/drm/radeon/radeon_semaphore.c
index abd6753a570a..6deb08f045b7 100644
--- a/drivers/gpu/drm/radeon/radeon_semaphore.c
+++ b/drivers/gpu/drm/radeon/radeon_semaphore.c
@@ -96,15 +96,15 @@ bool radeon_semaphore_emit_wait(struct radeon_device *rdev, int ridx,
 }
 
 /**
- * radeon_semaphore_sync_to - use the semaphore to sync to a fence
+ * radeon_semaphore_sync_fence - use the semaphore to sync to a fence
  *
  * @semaphore: semaphore object to add fence to
  * @fence: fence to sync to
  *
  * Sync to the fence using this semaphore object
  */
-void radeon_semaphore_sync_to(struct radeon_semaphore *semaphore,
-			      struct radeon_fence *fence)
+void radeon_semaphore_sync_fence(struct radeon_semaphore *semaphore,
+				 struct radeon_fence *fence)
 {
         struct radeon_fence *other;
 
@@ -116,6 +116,53 @@ void radeon_semaphore_sync_to(struct radeon_semaphore *semaphore,
 }
 
 /**
+ * radeon_semaphore_sync_to - use the semaphore to sync to a reservation object
+ *
+ * @sema: semaphore object to add fence from reservation object to
+ * @resv: reservation object with embedded fence
+ * @shared: true if we should onyl sync to the exclusive fence
+ *
+ * Sync to the fence using this semaphore object
+ */
+int radeon_semaphore_sync_resv(struct radeon_device *rdev,
+			       struct radeon_semaphore *sema,
+			       struct reservation_object *resv,
+			       bool shared)
+{
+	struct reservation_object_list *flist;
+	struct fence *f;
+	struct radeon_fence *fence;
+	unsigned i;
+	int r = 0;
+
+	/* always sync to the exclusive fence */
+	f = reservation_object_get_excl(resv);
+	fence = f ? to_radeon_fence(f) : NULL;
+	if (fence && fence->rdev == rdev)
+		radeon_semaphore_sync_fence(sema, fence);
+	else if (f)
+		r = fence_wait(f, true);
+
+	flist = reservation_object_get_list(resv);
+	if (shared || !flist || r)
+		return r;
+
+	for (i = 0; i < flist->shared_count; ++i) {
+		f = rcu_dereference_protected(flist->shared[i],
+					      reservation_object_held(resv));
+		fence = to_radeon_fence(f);
+		if (fence && fence->rdev == rdev)
+			radeon_semaphore_sync_fence(sema, fence);
+		else
+			r = fence_wait(f, true);
+
+		if (r)
+			break;
+	}
+	return r;
+}
+
+/**
  * radeon_semaphore_sync_rings - sync ring to all registered fences
  *
  * @rdev: radeon_device pointer
diff --git a/drivers/gpu/drm/radeon/radeon_state.c b/drivers/gpu/drm/radeon/radeon_state.c
index 23bb64fd775f..535403e0c8a2 100644
--- a/drivers/gpu/drm/radeon/radeon_state.c
+++ b/drivers/gpu/drm/radeon/radeon_state.c
@@ -30,9 +30,9 @@
  */
 
 #include <drm/drmP.h>
-#include <drm/drm_buffer.h>
 #include <drm/radeon_drm.h>
 #include "radeon_drv.h"
+#include "drm_buffer.h"
 
 /* ================================================================
  * Helper functions for client state checking and fixup
diff --git a/drivers/gpu/drm/radeon/radeon_test.c b/drivers/gpu/drm/radeon/radeon_test.c
index 17bc3dced9f1..07b506b41008 100644
--- a/drivers/gpu/drm/radeon/radeon_test.c
+++ b/drivers/gpu/drm/radeon/radeon_test.c
@@ -67,7 +67,7 @@ static void radeon_do_test_moves(struct radeon_device *rdev, int flag)
 	}
 
 	r = radeon_bo_create(rdev, size, PAGE_SIZE, true, RADEON_GEM_DOMAIN_VRAM,
-			     0, NULL, &vram_obj);
+			     0, NULL, NULL, &vram_obj);
 	if (r) {
 		DRM_ERROR("Failed to create VRAM object\n");
 		goto out_cleanup;
@@ -87,7 +87,8 @@ static void radeon_do_test_moves(struct radeon_device *rdev, int flag)
 		struct radeon_fence *fence = NULL;
 
 		r = radeon_bo_create(rdev, size, PAGE_SIZE, true,
-				     RADEON_GEM_DOMAIN_GTT, 0, NULL, gtt_obj + i);
+				     RADEON_GEM_DOMAIN_GTT, 0, NULL, NULL,
+				     gtt_obj + i);
 		if (r) {
 			DRM_ERROR("Failed to create GTT object %d\n", i);
 			goto out_lclean;
@@ -116,11 +117,16 @@ static void radeon_do_test_moves(struct radeon_device *rdev, int flag)
 		radeon_bo_kunmap(gtt_obj[i]);
 
 		if (ring == R600_RING_TYPE_DMA_INDEX)
-			r = radeon_copy_dma(rdev, gtt_addr, vram_addr, size / RADEON_GPU_PAGE_SIZE, &fence);
+			fence = radeon_copy_dma(rdev, gtt_addr, vram_addr,
+						size / RADEON_GPU_PAGE_SIZE,
+						NULL);
 		else
-			r = radeon_copy_blit(rdev, gtt_addr, vram_addr, size / RADEON_GPU_PAGE_SIZE, &fence);
-		if (r) {
+			fence = radeon_copy_blit(rdev, gtt_addr, vram_addr,
+						 size / RADEON_GPU_PAGE_SIZE,
+						 NULL);
+		if (IS_ERR(fence)) {
 			DRM_ERROR("Failed GTT->VRAM copy %d\n", i);
+			r = PTR_ERR(fence);
 			goto out_lclean_unpin;
 		}
 
@@ -162,11 +168,16 @@ static void radeon_do_test_moves(struct radeon_device *rdev, int flag)
 		radeon_bo_kunmap(vram_obj);
 
 		if (ring == R600_RING_TYPE_DMA_INDEX)
-			r = radeon_copy_dma(rdev, vram_addr, gtt_addr, size / RADEON_GPU_PAGE_SIZE, &fence);
+			fence = radeon_copy_dma(rdev, vram_addr, gtt_addr,
+						size / RADEON_GPU_PAGE_SIZE,
+						NULL);
 		else
-			r = radeon_copy_blit(rdev, vram_addr, gtt_addr, size / RADEON_GPU_PAGE_SIZE, &fence);
-		if (r) {
+			fence = radeon_copy_blit(rdev, vram_addr, gtt_addr,
+						 size / RADEON_GPU_PAGE_SIZE,
+						 NULL);
+		if (IS_ERR(fence)) {
 			DRM_ERROR("Failed VRAM->GTT copy %d\n", i);
+			r = PTR_ERR(fence);
 			goto out_lclean_unpin;
 		}
 
@@ -222,7 +233,7 @@ out_lclean:
 			radeon_bo_unreserve(gtt_obj[i]);
 			radeon_bo_unref(&gtt_obj[i]);
 		}
-		if (fence)
+		if (fence && !IS_ERR(fence))
 			radeon_fence_unref(&fence);
 		break;
 	}
diff --git a/drivers/gpu/drm/radeon/radeon_ttm.c b/drivers/gpu/drm/radeon/radeon_ttm.c
index 72afe82a95c9..8624979afb65 100644
--- a/drivers/gpu/drm/radeon/radeon_ttm.c
+++ b/drivers/gpu/drm/radeon/radeon_ttm.c
@@ -39,6 +39,8 @@
 #include <linux/seq_file.h>
 #include <linux/slab.h>
 #include <linux/swiotlb.h>
+#include <linux/swap.h>
+#include <linux/pagemap.h>
 #include <linux/debugfs.h>
 #include "radeon_reg.h"
 #include "radeon.h"
@@ -176,12 +178,15 @@ static int radeon_init_mem_type(struct ttm_bo_device *bdev, uint32_t type,
 static void radeon_evict_flags(struct ttm_buffer_object *bo,
 				struct ttm_placement *placement)
 {
+	static struct ttm_place placements = {
+		.fpfn = 0,
+		.lpfn = 0,
+		.flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM
+	};
+
 	struct radeon_bo *rbo;
-	static u32 placements = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
 
 	if (!radeon_ttm_bo_is_radeon_bo(bo)) {
-		placement->fpfn = 0;
-		placement->lpfn = 0;
 		placement->placement = &placements;
 		placement->busy_placement = &placements;
 		placement->num_placement = 1;
@@ -228,6 +233,7 @@ static int radeon_move_blit(struct ttm_buffer_object *bo,
 	struct radeon_device *rdev;
 	uint64_t old_start, new_start;
 	struct radeon_fence *fence;
+	unsigned num_pages;
 	int r, ridx;
 
 	rdev = radeon_get_rdev(bo->bdev);
@@ -264,13 +270,12 @@ static int radeon_move_blit(struct ttm_buffer_object *bo,
 
 	BUILD_BUG_ON((PAGE_SIZE % RADEON_GPU_PAGE_SIZE) != 0);
 
-	/* sync other rings */
-	fence = bo->sync_obj;
-	r = radeon_copy(rdev, old_start, new_start,
-			new_mem->num_pages * (PAGE_SIZE / RADEON_GPU_PAGE_SIZE), /* GPU pages */
-			&fence);
-	/* FIXME: handle copy error */
-	r = ttm_bo_move_accel_cleanup(bo, (void *)fence,
+	num_pages = new_mem->num_pages * (PAGE_SIZE / RADEON_GPU_PAGE_SIZE);
+	fence = radeon_copy(rdev, old_start, new_start, num_pages, bo->resv);
+	if (IS_ERR(fence))
+		return PTR_ERR(fence);
+
+	r = ttm_bo_move_accel_cleanup(bo, &fence->base,
 				      evict, no_wait_gpu, new_mem);
 	radeon_fence_unref(&fence);
 	return r;
@@ -284,20 +289,20 @@ static int radeon_move_vram_ram(struct ttm_buffer_object *bo,
 	struct radeon_device *rdev;
 	struct ttm_mem_reg *old_mem = &bo->mem;
 	struct ttm_mem_reg tmp_mem;
-	u32 placements;
+	struct ttm_place placements;
 	struct ttm_placement placement;
 	int r;
 
 	rdev = radeon_get_rdev(bo->bdev);
 	tmp_mem = *new_mem;
 	tmp_mem.mm_node = NULL;
-	placement.fpfn = 0;
-	placement.lpfn = 0;
 	placement.num_placement = 1;
 	placement.placement = &placements;
 	placement.num_busy_placement = 1;
 	placement.busy_placement = &placements;
-	placements = TTM_PL_MASK_CACHING | TTM_PL_FLAG_TT;
+	placements.fpfn = 0;
+	placements.lpfn = 0;
+	placements.flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_TT;
 	r = ttm_bo_mem_space(bo, &placement, &tmp_mem,
 			     interruptible, no_wait_gpu);
 	if (unlikely(r)) {
@@ -332,19 +337,19 @@ static int radeon_move_ram_vram(struct ttm_buffer_object *bo,
 	struct ttm_mem_reg *old_mem = &bo->mem;
 	struct ttm_mem_reg tmp_mem;
 	struct ttm_placement placement;
-	u32 placements;
+	struct ttm_place placements;
 	int r;
 
 	rdev = radeon_get_rdev(bo->bdev);
 	tmp_mem = *new_mem;
 	tmp_mem.mm_node = NULL;
-	placement.fpfn = 0;
-	placement.lpfn = 0;
 	placement.num_placement = 1;
 	placement.placement = &placements;
 	placement.num_busy_placement = 1;
 	placement.busy_placement = &placements;
-	placements = TTM_PL_MASK_CACHING | TTM_PL_FLAG_TT;
+	placements.fpfn = 0;
+	placements.lpfn = 0;
+	placements.flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_TT;
 	r = ttm_bo_mem_space(bo, &placement, &tmp_mem,
 			     interruptible, no_wait_gpu);
 	if (unlikely(r)) {
@@ -483,39 +488,108 @@ static void radeon_ttm_io_mem_free(struct ttm_bo_device *bdev, struct ttm_mem_re
 {
 }
 
-static int radeon_sync_obj_wait(void *sync_obj, bool lazy, bool interruptible)
-{
-	return radeon_fence_wait((struct radeon_fence *)sync_obj, interruptible);
-}
+/*
+ * TTM backend functions.
+ */
+struct radeon_ttm_tt {
+	struct ttm_dma_tt		ttm;
+	struct radeon_device		*rdev;
+	u64				offset;
 
-static int radeon_sync_obj_flush(void *sync_obj)
+	uint64_t			userptr;
+	struct mm_struct		*usermm;
+	uint32_t			userflags;
+};
+
+/* prepare the sg table with the user pages */
+static int radeon_ttm_tt_pin_userptr(struct ttm_tt *ttm)
 {
+	struct radeon_device *rdev = radeon_get_rdev(ttm->bdev);
+	struct radeon_ttm_tt *gtt = (void *)ttm;
+	unsigned pinned = 0, nents;
+	int r;
+
+	int write = !(gtt->userflags & RADEON_GEM_USERPTR_READONLY);
+	enum dma_data_direction direction = write ?
+		DMA_BIDIRECTIONAL : DMA_TO_DEVICE;
+
+	if (current->mm != gtt->usermm)
+		return -EPERM;
+
+	if (gtt->userflags & RADEON_GEM_USERPTR_ANONONLY) {
+		/* check that we only pin down anonymous memory
+		   to prevent problems with writeback */
+		unsigned long end = gtt->userptr + ttm->num_pages * PAGE_SIZE;
+		struct vm_area_struct *vma;
+		vma = find_vma(gtt->usermm, gtt->userptr);
+		if (!vma || vma->vm_file || vma->vm_end < end)
+			return -EPERM;
+	}
+
+	do {
+		unsigned num_pages = ttm->num_pages - pinned;
+		uint64_t userptr = gtt->userptr + pinned * PAGE_SIZE;
+		struct page **pages = ttm->pages + pinned;
+
+		r = get_user_pages(current, current->mm, userptr, num_pages,
+				   write, 0, pages, NULL);
+		if (r < 0)
+			goto release_pages;
+
+		pinned += r;
+
+	} while (pinned < ttm->num_pages);
+
+	r = sg_alloc_table_from_pages(ttm->sg, ttm->pages, ttm->num_pages, 0,
+				      ttm->num_pages << PAGE_SHIFT,
+				      GFP_KERNEL);
+	if (r)
+		goto release_sg;
+
+	r = -ENOMEM;
+	nents = dma_map_sg(rdev->dev, ttm->sg->sgl, ttm->sg->nents, direction);
+	if (nents != ttm->sg->nents)
+		goto release_sg;
+
+	drm_prime_sg_to_page_addr_arrays(ttm->sg, ttm->pages,
+					 gtt->ttm.dma_address, ttm->num_pages);
+
 	return 0;
-}
 
-static void radeon_sync_obj_unref(void **sync_obj)
-{
-	radeon_fence_unref((struct radeon_fence **)sync_obj);
-}
+release_sg:
+	kfree(ttm->sg);
 
-static void *radeon_sync_obj_ref(void *sync_obj)
-{
-	return radeon_fence_ref((struct radeon_fence *)sync_obj);
+release_pages:
+	release_pages(ttm->pages, pinned, 0);
+	return r;
 }
 
-static bool radeon_sync_obj_signaled(void *sync_obj)
+static void radeon_ttm_tt_unpin_userptr(struct ttm_tt *ttm)
 {
-	return radeon_fence_signaled((struct radeon_fence *)sync_obj);
-}
+	struct radeon_device *rdev = radeon_get_rdev(ttm->bdev);
+	struct radeon_ttm_tt *gtt = (void *)ttm;
+	struct scatterlist *sg;
+	int i;
 
-/*
- * TTM backend functions.
- */
-struct radeon_ttm_tt {
-	struct ttm_dma_tt		ttm;
-	struct radeon_device		*rdev;
-	u64				offset;
-};
+	int write = !(gtt->userflags & RADEON_GEM_USERPTR_READONLY);
+	enum dma_data_direction direction = write ?
+		DMA_BIDIRECTIONAL : DMA_TO_DEVICE;
+
+	/* free the sg table and pages again */
+	dma_unmap_sg(rdev->dev, ttm->sg->sgl, ttm->sg->nents, direction);
+
+	for_each_sg(ttm->sg->sgl, sg, ttm->sg->nents, i) {
+		struct page *page = sg_page(sg);
+
+		if (!(gtt->userflags & RADEON_GEM_USERPTR_READONLY))
+			set_page_dirty(page);
+
+		mark_page_accessed(page);
+		page_cache_release(page);
+	}
+
+	sg_free_table(ttm->sg);
+}
 
 static int radeon_ttm_backend_bind(struct ttm_tt *ttm,
 				   struct ttm_mem_reg *bo_mem)
@@ -525,6 +599,11 @@ static int radeon_ttm_backend_bind(struct ttm_tt *ttm,
 		RADEON_GART_PAGE_WRITE;
 	int r;
 
+	if (gtt->userptr) {
+		radeon_ttm_tt_pin_userptr(ttm);
+		flags &= ~RADEON_GART_PAGE_WRITE;
+	}
+
 	gtt->offset = (unsigned long)(bo_mem->start << PAGE_SHIFT);
 	if (!ttm->num_pages) {
 		WARN(1, "nothing to bind %lu pages for mreg %p back %p!\n",
@@ -547,6 +626,10 @@ static int radeon_ttm_backend_unbind(struct ttm_tt *ttm)
 	struct radeon_ttm_tt *gtt = (void *)ttm;
 
 	radeon_gart_unbind(gtt->rdev, gtt->offset, ttm->num_pages);
+
+	if (gtt->userptr)
+		radeon_ttm_tt_unpin_userptr(ttm);
+
 	return 0;
 }
 
@@ -592,10 +675,17 @@ static struct ttm_tt *radeon_ttm_tt_create(struct ttm_bo_device *bdev,
 	return &gtt->ttm.ttm;
 }
 
+static struct radeon_ttm_tt *radeon_ttm_tt_to_gtt(struct ttm_tt *ttm)
+{
+	if (!ttm || ttm->func != &radeon_backend_func)
+		return NULL;
+	return (struct radeon_ttm_tt *)ttm;
+}
+
 static int radeon_ttm_tt_populate(struct ttm_tt *ttm)
 {
+	struct radeon_ttm_tt *gtt = radeon_ttm_tt_to_gtt(ttm);
 	struct radeon_device *rdev;
-	struct radeon_ttm_tt *gtt = (void *)ttm;
 	unsigned i;
 	int r;
 	bool slave = !!(ttm->page_flags & TTM_PAGE_FLAG_SG);
@@ -603,6 +693,16 @@ static int radeon_ttm_tt_populate(struct ttm_tt *ttm)
 	if (ttm->state != tt_unpopulated)
 		return 0;
 
+	if (gtt && gtt->userptr) {
+		ttm->sg = kcalloc(1, sizeof(struct sg_table), GFP_KERNEL);
+		if (!ttm->sg)
+			return -ENOMEM;
+
+		ttm->page_flags |= TTM_PAGE_FLAG_SG;
+		ttm->state = tt_unbound;
+		return 0;
+	}
+
 	if (slave && ttm->sg) {
 		drm_prime_sg_to_page_addr_arrays(ttm->sg, ttm->pages,
 						 gtt->ttm.dma_address, ttm->num_pages);
@@ -648,10 +748,16 @@ static int radeon_ttm_tt_populate(struct ttm_tt *ttm)
 static void radeon_ttm_tt_unpopulate(struct ttm_tt *ttm)
 {
 	struct radeon_device *rdev;
-	struct radeon_ttm_tt *gtt = (void *)ttm;
+	struct radeon_ttm_tt *gtt = radeon_ttm_tt_to_gtt(ttm);
 	unsigned i;
 	bool slave = !!(ttm->page_flags & TTM_PAGE_FLAG_SG);
 
+	if (gtt && gtt->userptr) {
+		kfree(ttm->sg);
+		ttm->page_flags &= ~TTM_PAGE_FLAG_SG;
+		return;
+	}
+
 	if (slave)
 		return;
 
@@ -680,6 +786,40 @@ static void radeon_ttm_tt_unpopulate(struct ttm_tt *ttm)
 	ttm_pool_unpopulate(ttm);
 }
 
+int radeon_ttm_tt_set_userptr(struct ttm_tt *ttm, uint64_t addr,
+			      uint32_t flags)
+{
+	struct radeon_ttm_tt *gtt = radeon_ttm_tt_to_gtt(ttm);
+
+	if (gtt == NULL)
+		return -EINVAL;
+
+	gtt->userptr = addr;
+	gtt->usermm = current->mm;
+	gtt->userflags = flags;
+	return 0;
+}
+
+bool radeon_ttm_tt_has_userptr(struct ttm_tt *ttm)
+{
+	struct radeon_ttm_tt *gtt = radeon_ttm_tt_to_gtt(ttm);
+
+	if (gtt == NULL)
+		return false;
+
+	return !!gtt->userptr;
+}
+
+bool radeon_ttm_tt_is_readonly(struct ttm_tt *ttm)
+{
+	struct radeon_ttm_tt *gtt = radeon_ttm_tt_to_gtt(ttm);
+
+	if (gtt == NULL)
+		return false;
+
+	return !!(gtt->userflags & RADEON_GEM_USERPTR_READONLY);
+}
+
 static struct ttm_bo_driver radeon_bo_driver = {
 	.ttm_tt_create = &radeon_ttm_tt_create,
 	.ttm_tt_populate = &radeon_ttm_tt_populate,
@@ -689,11 +829,6 @@ static struct ttm_bo_driver radeon_bo_driver = {
 	.evict_flags = &radeon_evict_flags,
 	.move = &radeon_bo_move,
 	.verify_access = &radeon_verify_access,
-	.sync_obj_signaled = &radeon_sync_obj_signaled,
-	.sync_obj_wait = &radeon_sync_obj_wait,
-	.sync_obj_flush = &radeon_sync_obj_flush,
-	.sync_obj_unref = &radeon_sync_obj_unref,
-	.sync_obj_ref = &radeon_sync_obj_ref,
 	.move_notify = &radeon_bo_move_notify,
 	.fault_reserve_notify = &radeon_bo_fault_reserve_notify,
 	.io_mem_reserve = &radeon_ttm_io_mem_reserve,
@@ -730,7 +865,7 @@ int radeon_ttm_init(struct radeon_device *rdev)
 	radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size);
 
 	r = radeon_bo_create(rdev, 256 * 1024, PAGE_SIZE, true,
-			     RADEON_GEM_DOMAIN_VRAM, 0,
+			     RADEON_GEM_DOMAIN_VRAM, 0, NULL,
 			     NULL, &rdev->stollen_vga_memory);
 	if (r) {
 		return r;
@@ -828,7 +963,7 @@ int radeon_mmap(struct file *filp, struct vm_area_struct *vma)
 	int r;
 
 	if (unlikely(vma->vm_pgoff < DRM_FILE_PAGE_OFFSET)) {
-		return drm_mmap(filp, vma);
+		return -EINVAL;
 	}
 
 	file_priv = filp->private_data;
diff --git a/drivers/gpu/drm/radeon/radeon_uvd.c b/drivers/gpu/drm/radeon/radeon_uvd.c
index 341848a14376..11b662469253 100644
--- a/drivers/gpu/drm/radeon/radeon_uvd.c
+++ b/drivers/gpu/drm/radeon/radeon_uvd.c
@@ -40,12 +40,18 @@
 #define UVD_IDLE_TIMEOUT_MS	1000
 
 /* Firmware Names */
+#define FIRMWARE_R600		"radeon/R600_uvd.bin"
+#define FIRMWARE_RS780		"radeon/RS780_uvd.bin"
+#define FIRMWARE_RV770		"radeon/RV770_uvd.bin"
 #define FIRMWARE_RV710		"radeon/RV710_uvd.bin"
 #define FIRMWARE_CYPRESS	"radeon/CYPRESS_uvd.bin"
 #define FIRMWARE_SUMO		"radeon/SUMO_uvd.bin"
 #define FIRMWARE_TAHITI		"radeon/TAHITI_uvd.bin"
 #define FIRMWARE_BONAIRE	"radeon/BONAIRE_uvd.bin"
 
+MODULE_FIRMWARE(FIRMWARE_R600);
+MODULE_FIRMWARE(FIRMWARE_RS780);
+MODULE_FIRMWARE(FIRMWARE_RV770);
 MODULE_FIRMWARE(FIRMWARE_RV710);
 MODULE_FIRMWARE(FIRMWARE_CYPRESS);
 MODULE_FIRMWARE(FIRMWARE_SUMO);
@@ -63,6 +69,23 @@ int radeon_uvd_init(struct radeon_device *rdev)
 	INIT_DELAYED_WORK(&rdev->uvd.idle_work, radeon_uvd_idle_work_handler);
 
 	switch (rdev->family) {
+	case CHIP_RV610:
+	case CHIP_RV630:
+	case CHIP_RV670:
+	case CHIP_RV620:
+	case CHIP_RV635:
+		fw_name = FIRMWARE_R600;
+		break;
+
+	case CHIP_RS780:
+	case CHIP_RS880:
+		fw_name = FIRMWARE_RS780;
+		break;
+
+	case CHIP_RV770:
+		fw_name = FIRMWARE_RV770;
+		break;
+
 	case CHIP_RV710:
 	case CHIP_RV730:
 	case CHIP_RV740:
@@ -115,9 +138,11 @@ int radeon_uvd_init(struct radeon_device *rdev)
 	}
 
 	bo_size = RADEON_GPU_PAGE_ALIGN(rdev->uvd_fw->size + 8) +
-		  RADEON_UVD_STACK_SIZE + RADEON_UVD_HEAP_SIZE;
+		  RADEON_UVD_STACK_SIZE + RADEON_UVD_HEAP_SIZE +
+		  RADEON_GPU_PAGE_SIZE;
 	r = radeon_bo_create(rdev, bo_size, PAGE_SIZE, true,
-			     RADEON_GEM_DOMAIN_VRAM, 0, NULL, &rdev->uvd.vcpu_bo);
+			     RADEON_GEM_DOMAIN_VRAM, 0, NULL,
+			     NULL, &rdev->uvd.vcpu_bo);
 	if (r) {
 		dev_err(rdev->dev, "(%d) failed to allocate UVD bo\n", r);
 		return r;
@@ -231,10 +256,30 @@ int radeon_uvd_resume(struct radeon_device *rdev)
 	return 0;
 }
 
-void radeon_uvd_force_into_uvd_segment(struct radeon_bo *rbo)
+void radeon_uvd_force_into_uvd_segment(struct radeon_bo *rbo,
+				       uint32_t allowed_domains)
 {
-	rbo->placement.fpfn = 0 >> PAGE_SHIFT;
-	rbo->placement.lpfn = (256 * 1024 * 1024) >> PAGE_SHIFT;
+	int i;
+
+	for (i = 0; i < rbo->placement.num_placement; ++i) {
+		rbo->placements[i].fpfn = 0 >> PAGE_SHIFT;
+		rbo->placements[i].lpfn = (256 * 1024 * 1024) >> PAGE_SHIFT;
+	}
+
+	/* If it must be in VRAM it must be in the first segment as well */
+	if (allowed_domains == RADEON_GEM_DOMAIN_VRAM)
+		return;
+
+	/* abort if we already have more than one placement */
+	if (rbo->placement.num_placement > 1)
+		return;
+
+	/* add another 256MB segment */
+	rbo->placements[1] = rbo->placements[0];
+	rbo->placements[1].fpfn += (256 * 1024 * 1024) >> PAGE_SHIFT;
+	rbo->placements[1].lpfn += (256 * 1024 * 1024) >> PAGE_SHIFT;
+	rbo->placement.num_placement++;
+	rbo->placement.num_busy_placement++;
 }
 
 void radeon_uvd_free_handles(struct radeon_device *rdev, struct drm_file *filp)
@@ -356,6 +401,7 @@ static int radeon_uvd_cs_msg(struct radeon_cs_parser *p, struct radeon_bo *bo,
 {
 	int32_t *msg, msg_type, handle;
 	unsigned img_size = 0;
+	struct fence *f;
 	void *ptr;
 
 	int i, r;
@@ -365,8 +411,9 @@ static int radeon_uvd_cs_msg(struct radeon_cs_parser *p, struct radeon_bo *bo,
 		return -EINVAL;
 	}
 
-	if (bo->tbo.sync_obj) {
-		r = radeon_fence_wait(bo->tbo.sync_obj, false);
+	f = reservation_object_get_excl(bo->tbo.resv);
+	if (f) {
+		r = radeon_fence_wait((struct radeon_fence *)f, false);
 		if (r) {
 			DRM_ERROR("Failed waiting for UVD message (%d)!\n", r);
 			return r;
@@ -604,38 +651,16 @@ int radeon_uvd_cs_parse(struct radeon_cs_parser *p)
 }
 
 static int radeon_uvd_send_msg(struct radeon_device *rdev,
-			       int ring, struct radeon_bo *bo,
+			       int ring, uint64_t addr,
 			       struct radeon_fence **fence)
 {
-	struct ttm_validate_buffer tv;
-	struct ww_acquire_ctx ticket;
-	struct list_head head;
 	struct radeon_ib ib;
-	uint64_t addr;
 	int i, r;
 
-	memset(&tv, 0, sizeof(tv));
-	tv.bo = &bo->tbo;
-
-	INIT_LIST_HEAD(&head);
-	list_add(&tv.head, &head);
-
-	r = ttm_eu_reserve_buffers(&ticket, &head);
-	if (r)
-		return r;
-
-	radeon_ttm_placement_from_domain(bo, RADEON_GEM_DOMAIN_VRAM);
-	radeon_uvd_force_into_uvd_segment(bo);
-
-	r = ttm_bo_validate(&bo->tbo, &bo->placement, true, false);
-	if (r) 
-		goto err;
-
 	r = radeon_ib_get(rdev, ring, &ib, NULL, 64);
 	if (r)
-		goto err;
+		return r;
 
-	addr = radeon_bo_gpu_offset(bo);
 	ib.ptr[0] = PACKET0(UVD_GPCOM_VCPU_DATA0, 0);
 	ib.ptr[1] = addr;
 	ib.ptr[2] = PACKET0(UVD_GPCOM_VCPU_DATA1, 0);
@@ -647,19 +672,11 @@ static int radeon_uvd_send_msg(struct radeon_device *rdev,
 	ib.length_dw = 16;
 
 	r = radeon_ib_schedule(rdev, &ib, NULL, false);
-	if (r)
-		goto err;
-	ttm_eu_fence_buffer_objects(&ticket, &head, ib.fence);
 
 	if (fence)
 		*fence = radeon_fence_ref(ib.fence);
 
 	radeon_ib_free(rdev, &ib);
-	radeon_bo_unref(&bo);
-	return 0;
-
-err:
-	ttm_eu_backoff_reservation(&ticket, &head);
 	return r;
 }
 
@@ -669,27 +686,18 @@ err:
 int radeon_uvd_get_create_msg(struct radeon_device *rdev, int ring,
 			      uint32_t handle, struct radeon_fence **fence)
 {
-	struct radeon_bo *bo;
-	uint32_t *msg;
-	int r, i;
+	/* we use the last page of the vcpu bo for the UVD message */
+	uint64_t offs = radeon_bo_size(rdev->uvd.vcpu_bo) -
+		RADEON_GPU_PAGE_SIZE;
 
-	r = radeon_bo_create(rdev, 1024, PAGE_SIZE, true,
-			     RADEON_GEM_DOMAIN_VRAM, 0, NULL, &bo);
-	if (r)
-		return r;
+	uint32_t *msg = rdev->uvd.cpu_addr + offs;
+	uint64_t addr = rdev->uvd.gpu_addr + offs;
 
-	r = radeon_bo_reserve(bo, false);
-	if (r) {
-		radeon_bo_unref(&bo);
-		return r;
-	}
+	int r, i;
 
-	r = radeon_bo_kmap(bo, (void **)&msg);
-	if (r) {
-		radeon_bo_unreserve(bo);
-		radeon_bo_unref(&bo);
+	r = radeon_bo_reserve(rdev->uvd.vcpu_bo, true);
+	if (r)
 		return r;
-	}
 
 	/* stitch together an UVD create msg */
 	msg[0] = cpu_to_le32(0x00000de4);
@@ -706,36 +714,26 @@ int radeon_uvd_get_create_msg(struct radeon_device *rdev, int ring,
 	for (i = 11; i < 1024; ++i)
 		msg[i] = cpu_to_le32(0x0);
 
-	radeon_bo_kunmap(bo);
-	radeon_bo_unreserve(bo);
-
-	return radeon_uvd_send_msg(rdev, ring, bo, fence);
+	r = radeon_uvd_send_msg(rdev, ring, addr, fence);
+	radeon_bo_unreserve(rdev->uvd.vcpu_bo);
+	return r;
 }
 
 int radeon_uvd_get_destroy_msg(struct radeon_device *rdev, int ring,
 			       uint32_t handle, struct radeon_fence **fence)
 {
-	struct radeon_bo *bo;
-	uint32_t *msg;
-	int r, i;
+	/* we use the last page of the vcpu bo for the UVD message */
+	uint64_t offs = radeon_bo_size(rdev->uvd.vcpu_bo) -
+		RADEON_GPU_PAGE_SIZE;
 
-	r = radeon_bo_create(rdev, 1024, PAGE_SIZE, true,
-			     RADEON_GEM_DOMAIN_VRAM, 0, NULL, &bo);
-	if (r)
-		return r;
+	uint32_t *msg = rdev->uvd.cpu_addr + offs;
+	uint64_t addr = rdev->uvd.gpu_addr + offs;
 
-	r = radeon_bo_reserve(bo, false);
-	if (r) {
-		radeon_bo_unref(&bo);
-		return r;
-	}
+	int r, i;
 
-	r = radeon_bo_kmap(bo, (void **)&msg);
-	if (r) {
-		radeon_bo_unreserve(bo);
-		radeon_bo_unref(&bo);
+	r = radeon_bo_reserve(rdev->uvd.vcpu_bo, true);
+	if (r)
 		return r;
-	}
 
 	/* stitch together an UVD destroy msg */
 	msg[0] = cpu_to_le32(0x00000de4);
@@ -745,10 +743,9 @@ int radeon_uvd_get_destroy_msg(struct radeon_device *rdev, int ring,
 	for (i = 4; i < 1024; ++i)
 		msg[i] = cpu_to_le32(0x0);
 
-	radeon_bo_kunmap(bo);
-	radeon_bo_unreserve(bo);
-
-	return radeon_uvd_send_msg(rdev, ring, bo, fence);
+	r = radeon_uvd_send_msg(rdev, ring, addr, fence);
+	radeon_bo_unreserve(rdev->uvd.vcpu_bo);
+	return r;
 }
 
 /**
diff --git a/drivers/gpu/drm/radeon/radeon_vce.c b/drivers/gpu/drm/radeon/radeon_vce.c
index c7190aadbd89..9e85757d5599 100644
--- a/drivers/gpu/drm/radeon/radeon_vce.c
+++ b/drivers/gpu/drm/radeon/radeon_vce.c
@@ -126,7 +126,8 @@ int radeon_vce_init(struct radeon_device *rdev)
 	size = RADEON_GPU_PAGE_ALIGN(rdev->vce_fw->size) +
 	       RADEON_VCE_STACK_SIZE + RADEON_VCE_HEAP_SIZE;
 	r = radeon_bo_create(rdev, size, PAGE_SIZE, true,
-			     RADEON_GEM_DOMAIN_VRAM, 0, NULL, &rdev->vce.vcpu_bo);
+			     RADEON_GEM_DOMAIN_VRAM, 0, NULL, NULL,
+			     &rdev->vce.vcpu_bo);
 	if (r) {
 		dev_err(rdev->dev, "(%d) failed to allocate VCE bo\n", r);
 		return r;
diff --git a/drivers/gpu/drm/radeon/radeon_vm.c b/drivers/gpu/drm/radeon/radeon_vm.c
index 088ffdc2f577..4532cc76a0a6 100644
--- a/drivers/gpu/drm/radeon/radeon_vm.c
+++ b/drivers/gpu/drm/radeon/radeon_vm.c
@@ -143,6 +143,7 @@ struct radeon_cs_reloc *radeon_vm_get_bos(struct radeon_device *rdev,
 	list[0].prefered_domains = RADEON_GEM_DOMAIN_VRAM;
 	list[0].allowed_domains = RADEON_GEM_DOMAIN_VRAM;
 	list[0].tv.bo = &vm->page_directory->tbo;
+	list[0].tv.shared = false;
 	list[0].tiling_flags = 0;
 	list[0].handle = 0;
 	list_add(&list[0].tv.head, head);
@@ -156,6 +157,7 @@ struct radeon_cs_reloc *radeon_vm_get_bos(struct radeon_device *rdev,
 		list[idx].prefered_domains = RADEON_GEM_DOMAIN_VRAM;
 		list[idx].allowed_domains = RADEON_GEM_DOMAIN_VRAM;
 		list[idx].tv.bo = &list[idx].robj->tbo;
+		list[idx].tv.shared = false;
 		list[idx].tiling_flags = 0;
 		list[idx].handle = 0;
 		list_add(&list[idx++].tv.head, head);
@@ -395,11 +397,12 @@ static int radeon_vm_clear_bo(struct radeon_device *rdev,
 
         memset(&tv, 0, sizeof(tv));
         tv.bo = &bo->tbo;
+	tv.shared = false;
 
         INIT_LIST_HEAD(&head);
         list_add(&tv.head, &head);
 
-        r = ttm_eu_reserve_buffers(&ticket, &head);
+        r = ttm_eu_reserve_buffers(&ticket, &head, true);
         if (r)
 		return r;
 
@@ -424,7 +427,7 @@ static int radeon_vm_clear_bo(struct radeon_device *rdev,
 	if (r)
                 goto error;
 
-	ttm_eu_fence_buffer_objects(&ticket, &head, ib.fence);
+	ttm_eu_fence_buffer_objects(&ticket, &head, &ib.fence->base);
 	radeon_ib_free(rdev, &ib);
 
 	return 0;
@@ -545,7 +548,8 @@ int radeon_vm_bo_set_addr(struct radeon_device *rdev,
 
 		r = radeon_bo_create(rdev, RADEON_VM_PTE_COUNT * 8,
 				     RADEON_GPU_PAGE_SIZE, true,
-				     RADEON_GEM_DOMAIN_VRAM, 0, NULL, &pt);
+				     RADEON_GEM_DOMAIN_VRAM, 0,
+				     NULL, NULL, &pt);
 		if (r)
 			return r;
 
@@ -694,8 +698,9 @@ int radeon_vm_update_page_directory(struct radeon_device *rdev,
 
 	if (ib.length_dw != 0) {
 		radeon_asic_vm_pad_ib(rdev, &ib);
-		radeon_semaphore_sync_to(ib.semaphore, pd->tbo.sync_obj);
-		radeon_semaphore_sync_to(ib.semaphore, vm->last_id_use);
+
+		radeon_semaphore_sync_resv(rdev, ib.semaphore, pd->tbo.resv, false);
+		radeon_semaphore_sync_fence(ib.semaphore, vm->last_id_use);
 		WARN_ON(ib.length_dw > ndw);
 		r = radeon_ib_schedule(rdev, &ib, NULL, false);
 		if (r) {
@@ -821,7 +826,7 @@ static void radeon_vm_update_ptes(struct radeon_device *rdev,
 		unsigned nptes;
 		uint64_t pte;
 
-		radeon_semaphore_sync_to(ib->semaphore, pt->tbo.sync_obj);
+		radeon_semaphore_sync_resv(rdev, ib->semaphore, pt->tbo.resv, false);
 
 		if ((addr & ~mask) == (end & ~mask))
 			nptes = end - addr;
@@ -892,6 +897,9 @@ int radeon_vm_bo_update(struct radeon_device *rdev,
 	bo_va->flags &= ~RADEON_VM_PAGE_VALID;
 	bo_va->flags &= ~RADEON_VM_PAGE_SYSTEM;
 	bo_va->flags &= ~RADEON_VM_PAGE_SNOOPED;
+	if (bo_va->bo && radeon_ttm_tt_is_readonly(bo_va->bo->tbo.ttm))
+		bo_va->flags &= ~RADEON_VM_PAGE_WRITEABLE;
+
 	if (mem) {
 		addr = mem->start << PAGE_SHIFT;
 		if (mem->mem_type != TTM_PL_SYSTEM) {
@@ -960,7 +968,7 @@ int radeon_vm_bo_update(struct radeon_device *rdev,
 	radeon_asic_vm_pad_ib(rdev, &ib);
 	WARN_ON(ib.length_dw > ndw);
 
-	radeon_semaphore_sync_to(ib.semaphore, vm->fence);
+	radeon_semaphore_sync_fence(ib.semaphore, vm->fence);
 	r = radeon_ib_schedule(rdev, &ib, NULL, false);
 	if (r) {
 		radeon_ib_free(rdev, &ib);
@@ -1120,7 +1128,7 @@ int radeon_vm_init(struct radeon_device *rdev, struct radeon_vm *vm)
 
 	r = radeon_bo_create(rdev, pd_size, align, true,
 			     RADEON_GEM_DOMAIN_VRAM, 0, NULL,
-			     &vm->page_directory);
+			     NULL, &vm->page_directory);
 	if (r)
 		return r;
 
diff --git a/drivers/gpu/drm/radeon/rv770.c b/drivers/gpu/drm/radeon/rv770.c
index d9f5ce715c9b..372016e266d0 100644
--- a/drivers/gpu/drm/radeon/rv770.c
+++ b/drivers/gpu/drm/radeon/rv770.c
@@ -26,7 +26,6 @@
  *          Jerome Glisse
  */
 #include <linux/firmware.h>
-#include <linux/platform_device.h>
 #include <linux/slab.h>
 #include <drm/drmP.h>
 #include "radeon.h"
diff --git a/drivers/gpu/drm/radeon/rv770_dma.c b/drivers/gpu/drm/radeon/rv770_dma.c
index 74426ac2bb5c..7f34bad2e724 100644
--- a/drivers/gpu/drm/radeon/rv770_dma.c
+++ b/drivers/gpu/drm/radeon/rv770_dma.c
@@ -33,18 +33,19 @@
  * @src_offset: src GPU address
  * @dst_offset: dst GPU address
  * @num_gpu_pages: number of GPU pages to xfer
- * @fence: radeon fence object
+ * @resv: reservation object to sync to
  *
  * Copy GPU paging using the DMA engine (r7xx).
  * Used by the radeon ttm implementation to move pages if
  * registered as the asic copy callback.
  */
-int rv770_copy_dma(struct radeon_device *rdev,
-		  uint64_t src_offset, uint64_t dst_offset,
-		  unsigned num_gpu_pages,
-		  struct radeon_fence **fence)
+struct radeon_fence *rv770_copy_dma(struct radeon_device *rdev,
+				    uint64_t src_offset, uint64_t dst_offset,
+				    unsigned num_gpu_pages,
+				    struct reservation_object *resv)
 {
 	struct radeon_semaphore *sem = NULL;
+	struct radeon_fence *fence;
 	int ring_index = rdev->asic->copy.dma_ring_index;
 	struct radeon_ring *ring = &rdev->ring[ring_index];
 	u32 size_in_dw, cur_size_in_dw;
@@ -54,7 +55,7 @@ int rv770_copy_dma(struct radeon_device *rdev,
 	r = radeon_semaphore_create(rdev, &sem);
 	if (r) {
 		DRM_ERROR("radeon: moving bo (%d).\n", r);
-		return r;
+		return ERR_PTR(r);
 	}
 
 	size_in_dw = (num_gpu_pages << RADEON_GPU_PAGE_SHIFT) / 4;
@@ -63,10 +64,10 @@ int rv770_copy_dma(struct radeon_device *rdev,
 	if (r) {
 		DRM_ERROR("radeon: moving bo (%d).\n", r);
 		radeon_semaphore_free(rdev, &sem, NULL);
-		return r;
+		return ERR_PTR(r);
 	}
 
-	radeon_semaphore_sync_to(sem, *fence);
+	radeon_semaphore_sync_resv(rdev, sem, resv, false);
 	radeon_semaphore_sync_rings(rdev, sem, ring->idx);
 
 	for (i = 0; i < num_loops; i++) {
@@ -83,15 +84,15 @@ int rv770_copy_dma(struct radeon_device *rdev,
 		dst_offset += cur_size_in_dw * 4;
 	}
 
-	r = radeon_fence_emit(rdev, fence, ring->idx);
+	r = radeon_fence_emit(rdev, &fence, ring->idx);
 	if (r) {
 		radeon_ring_unlock_undo(rdev, ring);
 		radeon_semaphore_free(rdev, &sem, NULL);
-		return r;
+		return ERR_PTR(r);
 	}
 
 	radeon_ring_unlock_commit(rdev, ring, false);
-	radeon_semaphore_free(rdev, &sem, *fence);
+	radeon_semaphore_free(rdev, &sem, fence);
 
-	return r;
+	return fence;
 }
diff --git a/drivers/gpu/drm/radeon/si.c b/drivers/gpu/drm/radeon/si.c
index 3a0b973e8a96..eeea5b6a1775 100644
--- a/drivers/gpu/drm/radeon/si.c
+++ b/drivers/gpu/drm/radeon/si.c
@@ -4684,7 +4684,7 @@ static int si_vm_packet3_compute_check(struct radeon_device *rdev,
 int si_ib_parse(struct radeon_device *rdev, struct radeon_ib *ib)
 {
 	int ret = 0;
-	u32 idx = 0;
+	u32 idx = 0, i;
 	struct radeon_cs_packet pkt;
 
 	do {
@@ -4695,6 +4695,12 @@ int si_ib_parse(struct radeon_device *rdev, struct radeon_ib *ib)
 		switch (pkt.type) {
 		case RADEON_PACKET_TYPE0:
 			dev_err(rdev->dev, "Packet0 not allowed!\n");
+			for (i = 0; i < ib->length_dw; i++) {
+				if (i == idx)
+					printk("\t0x%08x <---\n", ib->ptr[i]);
+				else
+					printk("\t0x%08x\n", ib->ptr[i]);
+			}
 			ret = -EINVAL;
 			break;
 		case RADEON_PACKET_TYPE2:
diff --git a/drivers/gpu/drm/radeon/si_dma.c b/drivers/gpu/drm/radeon/si_dma.c
index 7c22baaf94db..b58f12b762d7 100644
--- a/drivers/gpu/drm/radeon/si_dma.c
+++ b/drivers/gpu/drm/radeon/si_dma.c
@@ -218,18 +218,19 @@ void si_dma_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm *vm)
  * @src_offset: src GPU address
  * @dst_offset: dst GPU address
  * @num_gpu_pages: number of GPU pages to xfer
- * @fence: radeon fence object
+ * @resv: reservation object to sync to
  *
  * Copy GPU paging using the DMA engine (SI).
  * Used by the radeon ttm implementation to move pages if
  * registered as the asic copy callback.
  */
-int si_copy_dma(struct radeon_device *rdev,
-		uint64_t src_offset, uint64_t dst_offset,
-		unsigned num_gpu_pages,
-		struct radeon_fence **fence)
+struct radeon_fence *si_copy_dma(struct radeon_device *rdev,
+				 uint64_t src_offset, uint64_t dst_offset,
+				 unsigned num_gpu_pages,
+				 struct reservation_object *resv)
 {
 	struct radeon_semaphore *sem = NULL;
+	struct radeon_fence *fence;
 	int ring_index = rdev->asic->copy.dma_ring_index;
 	struct radeon_ring *ring = &rdev->ring[ring_index];
 	u32 size_in_bytes, cur_size_in_bytes;
@@ -239,7 +240,7 @@ int si_copy_dma(struct radeon_device *rdev,
 	r = radeon_semaphore_create(rdev, &sem);
 	if (r) {
 		DRM_ERROR("radeon: moving bo (%d).\n", r);
-		return r;
+		return ERR_PTR(r);
 	}
 
 	size_in_bytes = (num_gpu_pages << RADEON_GPU_PAGE_SHIFT);
@@ -248,10 +249,10 @@ int si_copy_dma(struct radeon_device *rdev,
 	if (r) {
 		DRM_ERROR("radeon: moving bo (%d).\n", r);
 		radeon_semaphore_free(rdev, &sem, NULL);
-		return r;
+		return ERR_PTR(r);
 	}
 
-	radeon_semaphore_sync_to(sem, *fence);
+	radeon_semaphore_sync_resv(rdev, sem, resv, false);
 	radeon_semaphore_sync_rings(rdev, sem, ring->idx);
 
 	for (i = 0; i < num_loops; i++) {
@@ -268,16 +269,16 @@ int si_copy_dma(struct radeon_device *rdev,
 		dst_offset += cur_size_in_bytes;
 	}
 
-	r = radeon_fence_emit(rdev, fence, ring->idx);
+	r = radeon_fence_emit(rdev, &fence, ring->idx);
 	if (r) {
 		radeon_ring_unlock_undo(rdev, ring);
 		radeon_semaphore_free(rdev, &sem, NULL);
-		return r;
+		return ERR_PTR(r);
 	}
 
 	radeon_ring_unlock_commit(rdev, ring, false);
-	radeon_semaphore_free(rdev, &sem, *fence);
+	radeon_semaphore_free(rdev, &sem, fence);
 
-	return r;
+	return fence;
 }
 
diff --git a/drivers/gpu/drm/radeon/si_dpm.c b/drivers/gpu/drm/radeon/si_dpm.c
index 70e61ffeace2..9e4d5d7d348f 100644
--- a/drivers/gpu/drm/radeon/si_dpm.c
+++ b/drivers/gpu/drm/radeon/si_dpm.c
@@ -2916,7 +2916,6 @@ static void si_apply_state_adjust_rules(struct radeon_device *rdev,
 	bool disable_sclk_switching = false;
 	u32 mclk, sclk;
 	u16 vddc, vddci;
-	u32 max_sclk_vddc, max_mclk_vddci, max_mclk_vddc;
 	int i;
 
 	if ((rdev->pm.dpm.new_active_crtc_count > 1) ||
@@ -2950,29 +2949,6 @@ static void si_apply_state_adjust_rules(struct radeon_device *rdev,
 		}
 	}
 
-	/* limit clocks to max supported clocks based on voltage dependency tables */
-	btc_get_max_clock_from_voltage_dependency_table(&rdev->pm.dpm.dyn_state.vddc_dependency_on_sclk,
-							&max_sclk_vddc);
-	btc_get_max_clock_from_voltage_dependency_table(&rdev->pm.dpm.dyn_state.vddci_dependency_on_mclk,
-							&max_mclk_vddci);
-	btc_get_max_clock_from_voltage_dependency_table(&rdev->pm.dpm.dyn_state.vddc_dependency_on_mclk,
-							&max_mclk_vddc);
-
-	for (i = 0; i < ps->performance_level_count; i++) {
-		if (max_sclk_vddc) {
-			if (ps->performance_levels[i].sclk > max_sclk_vddc)
-				ps->performance_levels[i].sclk = max_sclk_vddc;
-		}
-		if (max_mclk_vddci) {
-			if (ps->performance_levels[i].mclk > max_mclk_vddci)
-				ps->performance_levels[i].mclk = max_mclk_vddci;
-		}
-		if (max_mclk_vddc) {
-			if (ps->performance_levels[i].mclk > max_mclk_vddc)
-				ps->performance_levels[i].mclk = max_mclk_vddc;
-		}
-	}
-
 	/* XXX validate the min clocks required for display */
 
 	if (disable_mclk_switching) {
diff --git a/drivers/gpu/drm/radeon/sid.h b/drivers/gpu/drm/radeon/sid.h
index fd414d34d885..6635da9ec986 100644
--- a/drivers/gpu/drm/radeon/sid.h
+++ b/drivers/gpu/drm/radeon/sid.h
@@ -736,7 +736,7 @@
 #       define DESCRIPTION16(x)                          (((x) & 0xff) << 0)
 #       define DESCRIPTION17(x)                          (((x) & 0xff) << 8)
 
-#define AZ_F0_CODEC_PIN_CONTROL_HOTPLUG_CONTROL          0x54
+#define AZ_F0_CODEC_PIN_CONTROL_HOT_PLUG_CONTROL         0x54
 #       define AUDIO_ENABLED                             (1 << 31)
 
 #define AZ_F0_CODEC_PIN_CONTROL_RESPONSE_CONFIGURATION_DEFAULT  0x56
diff --git a/drivers/gpu/drm/radeon/uvd_v1_0.c b/drivers/gpu/drm/radeon/uvd_v1_0.c
index cda391347286..e72b3cb59358 100644
--- a/drivers/gpu/drm/radeon/uvd_v1_0.c
+++ b/drivers/gpu/drm/radeon/uvd_v1_0.c
@@ -22,6 +22,7 @@
  * Authors: Christian König <christian.koenig@amd.com>
  */
 
+#include <linux/firmware.h>
 #include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
@@ -70,6 +71,82 @@ void uvd_v1_0_set_wptr(struct radeon_device *rdev,
 }
 
 /**
+ * uvd_v1_0_fence_emit - emit an fence & trap command
+ *
+ * @rdev: radeon_device pointer
+ * @fence: fence to emit
+ *
+ * Write a fence and a trap command to the ring.
+ */
+void uvd_v1_0_fence_emit(struct radeon_device *rdev,
+			 struct radeon_fence *fence)
+{
+	struct radeon_ring *ring = &rdev->ring[fence->ring];
+	uint64_t addr = rdev->fence_drv[fence->ring].gpu_addr;
+
+	radeon_ring_write(ring, PACKET0(UVD_GPCOM_VCPU_DATA0, 0));
+	radeon_ring_write(ring, addr & 0xffffffff);
+	radeon_ring_write(ring, PACKET0(UVD_GPCOM_VCPU_DATA1, 0));
+	radeon_ring_write(ring, fence->seq);
+	radeon_ring_write(ring, PACKET0(UVD_GPCOM_VCPU_CMD, 0));
+	radeon_ring_write(ring, 0);
+
+	radeon_ring_write(ring, PACKET0(UVD_GPCOM_VCPU_DATA0, 0));
+	radeon_ring_write(ring, 0);
+	radeon_ring_write(ring, PACKET0(UVD_GPCOM_VCPU_DATA1, 0));
+	radeon_ring_write(ring, 0);
+	radeon_ring_write(ring, PACKET0(UVD_GPCOM_VCPU_CMD, 0));
+	radeon_ring_write(ring, 2);
+	return;
+}
+
+/**
+ * uvd_v1_0_resume - memory controller programming
+ *
+ * @rdev: radeon_device pointer
+ *
+ * Let the UVD memory controller know it's offsets
+ */
+int uvd_v1_0_resume(struct radeon_device *rdev)
+{
+	uint64_t addr;
+	uint32_t size;
+	int r;
+
+	r = radeon_uvd_resume(rdev);
+	if (r)
+		return r;
+
+	/* programm the VCPU memory controller bits 0-27 */
+	addr = (rdev->uvd.gpu_addr >> 3) + 16;
+	size = RADEON_GPU_PAGE_ALIGN(rdev->uvd_fw->size) >> 3;
+	WREG32(UVD_VCPU_CACHE_OFFSET0, addr);
+	WREG32(UVD_VCPU_CACHE_SIZE0, size);
+
+	addr += size;
+	size = RADEON_UVD_STACK_SIZE >> 3;
+	WREG32(UVD_VCPU_CACHE_OFFSET1, addr);
+	WREG32(UVD_VCPU_CACHE_SIZE1, size);
+
+	addr += size;
+	size = RADEON_UVD_HEAP_SIZE >> 3;
+	WREG32(UVD_VCPU_CACHE_OFFSET2, addr);
+	WREG32(UVD_VCPU_CACHE_SIZE2, size);
+
+	/* bits 28-31 */
+	addr = (rdev->uvd.gpu_addr >> 28) & 0xF;
+	WREG32(UVD_LMI_ADDR_EXT, (addr << 12) | (addr << 0));
+
+	/* bits 32-39 */
+	addr = (rdev->uvd.gpu_addr >> 32) & 0xFF;
+	WREG32(UVD_LMI_EXT40_ADDR, addr | (0x9 << 16) | (0x1 << 31));
+
+	WREG32(UVD_FW_START, *((uint32_t*)rdev->uvd.cpu_addr));
+
+	return 0;
+}
+
+/**
  * uvd_v1_0_init - start and test UVD block
  *
  * @rdev: radeon_device pointer
@@ -130,8 +207,32 @@ done:
 	/* lower clocks again */
 	radeon_set_uvd_clocks(rdev, 0, 0);
 
-	if (!r)
+	if (!r) {
+		switch (rdev->family) {
+		case CHIP_RV610:
+		case CHIP_RV630:
+		case CHIP_RV620:
+			/* 64byte granularity workaround */
+			WREG32(MC_CONFIG, 0);
+			WREG32(MC_CONFIG, 1 << 4);
+			WREG32(RS_DQ_RD_RET_CONF, 0x3f);
+			WREG32(MC_CONFIG, 0x1f);
+
+			/* fall through */
+		case CHIP_RV670:
+		case CHIP_RV635:
+
+			/* write clean workaround */
+			WREG32_P(UVD_VCPU_CNTL, 0x10, ~0x10);
+			break;
+
+		default:
+			/* TODO: Do we need more? */
+			break;
+		}
+
 		DRM_INFO("UVD initialized successfully.\n");
+	}
 
 	return r;
 }
@@ -218,12 +319,12 @@ int uvd_v1_0_start(struct radeon_device *rdev)
 	/* enable UMC */
 	WREG32_P(UVD_LMI_CTRL2, 0, ~(1 << 8));
 
+	WREG32_P(UVD_RB_ARB_CTRL, 0, ~(1 << 3));
+
 	/* boot up the VCPU */
 	WREG32(UVD_SOFT_RESET, 0);
 	mdelay(10);
 
-	WREG32_P(UVD_RB_ARB_CTRL, 0, ~(1 << 3));
-
 	for (i = 0; i < 10; ++i) {
 		uint32_t status;
 		for (j = 0; j < 100; ++j) {
diff --git a/drivers/gpu/drm/radeon/uvd_v2_2.c b/drivers/gpu/drm/radeon/uvd_v2_2.c
index 8bfdadd56598..89193519f8a1 100644
--- a/drivers/gpu/drm/radeon/uvd_v2_2.c
+++ b/drivers/gpu/drm/radeon/uvd_v2_2.c
@@ -72,6 +72,10 @@ int uvd_v2_2_resume(struct radeon_device *rdev)
 	uint32_t chip_id, size;
 	int r;
 
+	/* RV770 uses V1.0 MC */
+	if (rdev->family == CHIP_RV770)
+		return uvd_v1_0_resume(rdev);
+
 	r = radeon_uvd_resume(rdev);
 	if (r)
 		return r;
diff --git a/drivers/gpu/drm/rcar-du/Kconfig b/drivers/gpu/drm/rcar-du/Kconfig
index 2e3d7b5b0ad7..c96f6089f8bf 100644
--- a/drivers/gpu/drm/rcar-du/Kconfig
+++ b/drivers/gpu/drm/rcar-du/Kconfig
@@ -6,6 +6,7 @@ config DRM_RCAR_DU
 	select DRM_KMS_CMA_HELPER
 	select DRM_GEM_CMA_HELPER
 	select DRM_KMS_FB_HELPER
+	select VIDEOMODE_HELPERS
 	help
 	  Choose this option if you have an R-Car chipset.
 	  If M is selected the module will be called rcar-du-drm.
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_crtc.c b/drivers/gpu/drm/rcar-du/rcar_du_crtc.c
index 299267db2898..148b50589181 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_crtc.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_crtc.c
@@ -1,7 +1,7 @@
 /*
  * rcar_du_crtc.c  --  R-Car Display Unit CRTCs
  *
- * Copyright (C) 2013 Renesas Corporation
+ * Copyright (C) 2013-2014 Renesas Electronics Corporation
  *
  * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_crtc.h b/drivers/gpu/drm/rcar-du/rcar_du_crtc.h
index 43e7575c700c..e97ae502dec5 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_crtc.h
+++ b/drivers/gpu/drm/rcar-du/rcar_du_crtc.h
@@ -1,7 +1,7 @@
 /*
  * rcar_du_crtc.h  --  R-Car Display Unit CRTCs
  *
- * Copyright (C) 2013 Renesas Corporation
+ * Copyright (C) 2013-2014 Renesas Electronics Corporation
  *
  * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_drv.c b/drivers/gpu/drm/rcar-du/rcar_du_drv.c
index fda64b7b73e8..d212efa6a495 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_drv.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_drv.c
@@ -1,7 +1,7 @@
 /*
  * rcar_du_drv.c  --  R-Car Display Unit DRM driver
  *
- * Copyright (C) 2013 Renesas Corporation
+ * Copyright (C) 2013-2014 Renesas Electronics Corporation
  *
  * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
@@ -15,6 +15,7 @@
 #include <linux/io.h>
 #include <linux/mm.h>
 #include <linux/module.h>
+#include <linux/of_device.h>
 #include <linux/platform_device.h>
 #include <linux/pm.h>
 #include <linux/slab.h>
@@ -30,6 +31,97 @@
 #include "rcar_du_regs.h"
 
 /* -----------------------------------------------------------------------------
+ * Device Information
+ */
+
+static const struct rcar_du_device_info rcar_du_r8a7779_info = {
+	.features = 0,
+	.num_crtcs = 2,
+	.routes = {
+		/* R8A7779 has two RGB outputs and one (currently unsupported)
+		 * TCON output.
+		 */
+		[RCAR_DU_OUTPUT_DPAD0] = {
+			.possible_crtcs = BIT(0),
+			.encoder_type = DRM_MODE_ENCODER_NONE,
+			.port = 0,
+		},
+		[RCAR_DU_OUTPUT_DPAD1] = {
+			.possible_crtcs = BIT(1) | BIT(0),
+			.encoder_type = DRM_MODE_ENCODER_NONE,
+			.port = 1,
+		},
+	},
+	.num_lvds = 0,
+};
+
+static const struct rcar_du_device_info rcar_du_r8a7790_info = {
+	.features = RCAR_DU_FEATURE_CRTC_IRQ_CLOCK | RCAR_DU_FEATURE_DEFR8,
+	.quirks = RCAR_DU_QUIRK_ALIGN_128B | RCAR_DU_QUIRK_LVDS_LANES,
+	.num_crtcs = 3,
+	.routes = {
+		/* R8A7790 has one RGB output, two LVDS outputs and one
+		 * (currently unsupported) TCON output.
+		 */
+		[RCAR_DU_OUTPUT_DPAD0] = {
+			.possible_crtcs = BIT(2) | BIT(1) | BIT(0),
+			.encoder_type = DRM_MODE_ENCODER_NONE,
+			.port = 0,
+		},
+		[RCAR_DU_OUTPUT_LVDS0] = {
+			.possible_crtcs = BIT(0),
+			.encoder_type = DRM_MODE_ENCODER_LVDS,
+			.port = 1,
+		},
+		[RCAR_DU_OUTPUT_LVDS1] = {
+			.possible_crtcs = BIT(2) | BIT(1),
+			.encoder_type = DRM_MODE_ENCODER_LVDS,
+			.port = 2,
+		},
+	},
+	.num_lvds = 2,
+};
+
+static const struct rcar_du_device_info rcar_du_r8a7791_info = {
+	.features = RCAR_DU_FEATURE_CRTC_IRQ_CLOCK | RCAR_DU_FEATURE_DEFR8,
+	.num_crtcs = 2,
+	.routes = {
+		/* R8A7791 has one RGB output, one LVDS output and one
+		 * (currently unsupported) TCON output.
+		 */
+		[RCAR_DU_OUTPUT_DPAD0] = {
+			.possible_crtcs = BIT(1),
+			.encoder_type = DRM_MODE_ENCODER_NONE,
+			.port = 0,
+		},
+		[RCAR_DU_OUTPUT_LVDS0] = {
+			.possible_crtcs = BIT(0),
+			.encoder_type = DRM_MODE_ENCODER_LVDS,
+			.port = 1,
+		},
+	},
+	.num_lvds = 1,
+};
+
+static const struct platform_device_id rcar_du_id_table[] = {
+	{ "rcar-du-r8a7779", (kernel_ulong_t)&rcar_du_r8a7779_info },
+	{ "rcar-du-r8a7790", (kernel_ulong_t)&rcar_du_r8a7790_info },
+	{ "rcar-du-r8a7791", (kernel_ulong_t)&rcar_du_r8a7791_info },
+	{ }
+};
+
+MODULE_DEVICE_TABLE(platform, rcar_du_id_table);
+
+static const struct of_device_id rcar_du_of_table[] = {
+	{ .compatible = "renesas,du-r8a7779", .data = &rcar_du_r8a7779_info },
+	{ .compatible = "renesas,du-r8a7790", .data = &rcar_du_r8a7790_info },
+	{ .compatible = "renesas,du-r8a7791", .data = &rcar_du_r8a7791_info },
+	{ }
+};
+
+MODULE_DEVICE_TABLE(of, rcar_du_of_table);
+
+/* -----------------------------------------------------------------------------
  * DRM operations
  */
 
@@ -53,12 +145,13 @@ static int rcar_du_unload(struct drm_device *dev)
 static int rcar_du_load(struct drm_device *dev, unsigned long flags)
 {
 	struct platform_device *pdev = dev->platformdev;
+	struct device_node *np = pdev->dev.of_node;
 	struct rcar_du_platform_data *pdata = pdev->dev.platform_data;
 	struct rcar_du_device *rcdu;
 	struct resource *mem;
 	int ret;
 
-	if (pdata == NULL) {
+	if (pdata == NULL && np == NULL) {
 		dev_err(dev->dev, "no platform data\n");
 		return -ENODEV;
 	}
@@ -71,7 +164,8 @@ static int rcar_du_load(struct drm_device *dev, unsigned long flags)
 
 	rcdu->dev = &pdev->dev;
 	rcdu->pdata = pdata;
-	rcdu->info = (struct rcar_du_device_info *)pdev->id_entry->driver_data;
+	rcdu->info = np ? of_match_device(rcar_du_of_table, rcdu->dev)->data
+		   : (void *)platform_get_device_id(pdev)->driver_data;
 	rcdu->ddev = dev;
 	dev->dev_private = rcdu;
 
@@ -158,6 +252,7 @@ static struct drm_driver rcar_du_driver = {
 	.unload			= rcar_du_unload,
 	.preclose		= rcar_du_preclose,
 	.lastclose		= rcar_du_lastclose,
+	.set_busid		= drm_platform_set_busid,
 	.get_vblank_counter	= drm_vblank_count,
 	.enable_vblank		= rcar_du_enable_vblank,
 	.disable_vblank		= rcar_du_disable_vblank,
@@ -231,77 +326,6 @@ static int rcar_du_remove(struct platform_device *pdev)
 	return 0;
 }
 
-static const struct rcar_du_device_info rcar_du_r8a7779_info = {
-	.features = 0,
-	.num_crtcs = 2,
-	.routes = {
-		/* R8A7779 has two RGB outputs and one (currently unsupported)
-		 * TCON output.
-		 */
-		[RCAR_DU_OUTPUT_DPAD0] = {
-			.possible_crtcs = BIT(0),
-			.encoder_type = DRM_MODE_ENCODER_NONE,
-		},
-		[RCAR_DU_OUTPUT_DPAD1] = {
-			.possible_crtcs = BIT(1) | BIT(0),
-			.encoder_type = DRM_MODE_ENCODER_NONE,
-		},
-	},
-	.num_lvds = 0,
-};
-
-static const struct rcar_du_device_info rcar_du_r8a7790_info = {
-	.features = RCAR_DU_FEATURE_CRTC_IRQ_CLOCK | RCAR_DU_FEATURE_DEFR8,
-	.quirks = RCAR_DU_QUIRK_ALIGN_128B | RCAR_DU_QUIRK_LVDS_LANES,
-	.num_crtcs = 3,
-	.routes = {
-		/* R8A7790 has one RGB output, two LVDS outputs and one
-		 * (currently unsupported) TCON output.
-		 */
-		[RCAR_DU_OUTPUT_DPAD0] = {
-			.possible_crtcs = BIT(2) | BIT(1) | BIT(0),
-			.encoder_type = DRM_MODE_ENCODER_NONE,
-		},
-		[RCAR_DU_OUTPUT_LVDS0] = {
-			.possible_crtcs = BIT(0),
-			.encoder_type = DRM_MODE_ENCODER_LVDS,
-		},
-		[RCAR_DU_OUTPUT_LVDS1] = {
-			.possible_crtcs = BIT(2) | BIT(1),
-			.encoder_type = DRM_MODE_ENCODER_LVDS,
-		},
-	},
-	.num_lvds = 2,
-};
-
-static const struct rcar_du_device_info rcar_du_r8a7791_info = {
-	.features = RCAR_DU_FEATURE_CRTC_IRQ_CLOCK | RCAR_DU_FEATURE_DEFR8,
-	.num_crtcs = 2,
-	.routes = {
-		/* R8A7791 has one RGB output, one LVDS output and one
-		 * (currently unsupported) TCON output.
-		 */
-		[RCAR_DU_OUTPUT_DPAD0] = {
-			.possible_crtcs = BIT(1),
-			.encoder_type = DRM_MODE_ENCODER_NONE,
-		},
-		[RCAR_DU_OUTPUT_LVDS0] = {
-			.possible_crtcs = BIT(0),
-			.encoder_type = DRM_MODE_ENCODER_LVDS,
-		},
-	},
-	.num_lvds = 1,
-};
-
-static const struct platform_device_id rcar_du_id_table[] = {
-	{ "rcar-du-r8a7779", (kernel_ulong_t)&rcar_du_r8a7779_info },
-	{ "rcar-du-r8a7790", (kernel_ulong_t)&rcar_du_r8a7790_info },
-	{ "rcar-du-r8a7791", (kernel_ulong_t)&rcar_du_r8a7791_info },
-	{ }
-};
-
-MODULE_DEVICE_TABLE(platform, rcar_du_id_table);
-
 static struct platform_driver rcar_du_platform_driver = {
 	.probe		= rcar_du_probe,
 	.remove		= rcar_du_remove,
@@ -309,6 +333,7 @@ static struct platform_driver rcar_du_platform_driver = {
 		.owner	= THIS_MODULE,
 		.name	= "rcar-du",
 		.pm	= &rcar_du_pm_ops,
+		.of_match_table = rcar_du_of_table,
 	},
 	.id_table	= rcar_du_id_table,
 };
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_drv.h b/drivers/gpu/drm/rcar-du/rcar_du_drv.h
index e31b735d3f25..8e494633c3b3 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_drv.h
+++ b/drivers/gpu/drm/rcar-du/rcar_du_drv.h
@@ -1,7 +1,7 @@
 /*
  * rcar_du_drv.h  --  R-Car Display Unit DRM driver
  *
- * Copyright (C) 2013 Renesas Corporation
+ * Copyright (C) 2013-2014 Renesas Electronics Corporation
  *
  * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
@@ -37,6 +37,7 @@ struct rcar_du_lvdsenc;
  * struct rcar_du_output_routing - Output routing specification
  * @possible_crtcs: bitmask of possible CRTCs for the output
  * @encoder_type: DRM type of the internal encoder associated with the output
+ * @port: device tree port number corresponding to this output route
  *
  * The DU has 5 possible outputs (DPAD0/1, LVDS0/1, TCON). Output routing data
  * specify the valid SoC outputs, which CRTCs can drive the output, and the type
@@ -45,6 +46,7 @@ struct rcar_du_lvdsenc;
 struct rcar_du_output_routing {
 	unsigned int possible_crtcs;
 	unsigned int encoder_type;
+	unsigned int port;
 };
 
 /*
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_encoder.c b/drivers/gpu/drm/rcar-du/rcar_du_encoder.c
index 3daa7a168dc6..7c0ec95915ef 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_encoder.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_encoder.c
@@ -1,7 +1,7 @@
 /*
  * rcar_du_encoder.c  --  R-Car Display Unit Encoder
  *
- * Copyright (C) 2013 Renesas Corporation
+ * Copyright (C) 2013-2014 Renesas Electronics Corporation
  *
  * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
@@ -142,7 +142,8 @@ static const struct drm_encoder_funcs encoder_funcs = {
 int rcar_du_encoder_init(struct rcar_du_device *rcdu,
 			 enum rcar_du_encoder_type type,
 			 enum rcar_du_output output,
-			 const struct rcar_du_encoder_data *data)
+			 const struct rcar_du_encoder_data *data,
+			 struct device_node *np)
 {
 	struct rcar_du_encoder *renc;
 	unsigned int encoder_type;
@@ -189,9 +190,11 @@ int rcar_du_encoder_init(struct rcar_du_device *rcdu,
 	drm_encoder_helper_add(&renc->encoder, &encoder_helper_funcs);
 
 	switch (encoder_type) {
-	case DRM_MODE_ENCODER_LVDS:
-		return rcar_du_lvds_connector_init(rcdu, renc,
-						   &data->connector.lvds.panel);
+	case DRM_MODE_ENCODER_LVDS: {
+		const struct rcar_du_panel_data *pdata =
+			data ? &data->connector.lvds.panel : NULL;
+		return rcar_du_lvds_connector_init(rcdu, renc, pdata, np);
+	}
 
 	case DRM_MODE_ENCODER_DAC:
 		return rcar_du_vga_connector_init(rcdu, renc);
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_encoder.h b/drivers/gpu/drm/rcar-du/rcar_du_encoder.h
index 0e5a65e45d0e..bd624135ef1f 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_encoder.h
+++ b/drivers/gpu/drm/rcar-du/rcar_du_encoder.h
@@ -1,7 +1,7 @@
 /*
  * rcar_du_encoder.h  --  R-Car Display Unit Encoder
  *
- * Copyright (C) 2013 Renesas Corporation
+ * Copyright (C) 2013-2014 Renesas Electronics Corporation
  *
  * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
@@ -44,6 +44,7 @@ rcar_du_connector_best_encoder(struct drm_connector *connector);
 int rcar_du_encoder_init(struct rcar_du_device *rcdu,
 			 enum rcar_du_encoder_type type,
 			 enum rcar_du_output output,
-			 const struct rcar_du_encoder_data *data);
+			 const struct rcar_du_encoder_data *data,
+			 struct device_node *np);
 
 #endif /* __RCAR_DU_ENCODER_H__ */
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_group.c b/drivers/gpu/drm/rcar-du/rcar_du_group.c
index eb53cd97e8c6..4e7614b145db 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_group.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_group.c
@@ -1,7 +1,7 @@
 /*
  * rcar_du_group.c  --  R-Car Display Unit Channels Pair
  *
- * Copyright (C) 2013 Renesas Corporation
+ * Copyright (C) 2013-2014 Renesas Electronics Corporation
  *
  * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_group.h b/drivers/gpu/drm/rcar-du/rcar_du_group.h
index 5025930972ec..0c38cdcda4ca 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_group.h
+++ b/drivers/gpu/drm/rcar-du/rcar_du_group.h
@@ -1,7 +1,7 @@
 /*
  * rcar_du_group.c  --  R-Car Display Unit Planes and CRTCs Group
  *
- * Copyright (C) 2013 Renesas Corporation
+ * Copyright (C) 2013-2014 Renesas Electronics Corporation
  *
  * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_kms.c b/drivers/gpu/drm/rcar-du/rcar_du_kms.c
index 76026104d000..6c24ad7d03ef 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_kms.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_kms.c
@@ -1,7 +1,7 @@
 /*
  * rcar_du_kms.c  --  R-Car Display Unit Mode Setting
  *
- * Copyright (C) 2013 Renesas Corporation
+ * Copyright (C) 2013-2014 Renesas Electronics Corporation
  *
  * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
@@ -17,6 +17,8 @@
 #include <drm/drm_fb_cma_helper.h>
 #include <drm/drm_gem_cma_helper.h>
 
+#include <linux/of_graph.h>
+
 #include "rcar_du_crtc.h"
 #include "rcar_du_drv.h"
 #include "rcar_du_encoder.h"
@@ -188,6 +190,205 @@ static const struct drm_mode_config_funcs rcar_du_mode_config_funcs = {
 	.output_poll_changed = rcar_du_output_poll_changed,
 };
 
+static int rcar_du_encoders_init_pdata(struct rcar_du_device *rcdu)
+{
+	unsigned int num_encoders = 0;
+	unsigned int i;
+	int ret;
+
+	for (i = 0; i < rcdu->pdata->num_encoders; ++i) {
+		const struct rcar_du_encoder_data *pdata =
+			&rcdu->pdata->encoders[i];
+		const struct rcar_du_output_routing *route =
+			&rcdu->info->routes[pdata->output];
+
+		if (pdata->type == RCAR_DU_ENCODER_UNUSED)
+			continue;
+
+		if (pdata->output >= RCAR_DU_OUTPUT_MAX ||
+		    route->possible_crtcs == 0) {
+			dev_warn(rcdu->dev,
+				 "encoder %u references unexisting output %u, skipping\n",
+				 i, pdata->output);
+			continue;
+		}
+
+		ret = rcar_du_encoder_init(rcdu, pdata->type, pdata->output,
+					   pdata, NULL);
+		if (ret < 0)
+			return ret;
+
+		num_encoders++;
+	}
+
+	return num_encoders;
+}
+
+static int rcar_du_encoders_init_dt_one(struct rcar_du_device *rcdu,
+					enum rcar_du_output output,
+					struct of_endpoint *ep)
+{
+	static const struct {
+		const char *compatible;
+		enum rcar_du_encoder_type type;
+	} encoders[] = {
+		{ "adi,adv7123", RCAR_DU_ENCODER_VGA },
+		{ "thine,thc63lvdm83d", RCAR_DU_ENCODER_LVDS },
+	};
+
+	enum rcar_du_encoder_type enc_type = RCAR_DU_ENCODER_NONE;
+	struct device_node *connector = NULL;
+	struct device_node *encoder = NULL;
+	struct device_node *prev = NULL;
+	struct device_node *entity_ep_node;
+	struct device_node *entity;
+	int ret;
+
+	/*
+	 * Locate the connected entity and infer its type from the number of
+	 * endpoints.
+	 */
+	entity = of_graph_get_remote_port_parent(ep->local_node);
+	if (!entity) {
+		dev_dbg(rcdu->dev, "unconnected endpoint %s, skipping\n",
+			ep->local_node->full_name);
+		return 0;
+	}
+
+	entity_ep_node = of_parse_phandle(ep->local_node, "remote-endpoint", 0);
+
+	while (1) {
+		struct device_node *ep_node;
+
+		ep_node = of_graph_get_next_endpoint(entity, prev);
+		of_node_put(prev);
+		prev = ep_node;
+
+		if (!ep_node)
+			break;
+
+		if (ep_node == entity_ep_node)
+			continue;
+
+		/*
+		 * We've found one endpoint other than the input, this must
+		 * be an encoder. Locate the connector.
+		 */
+		encoder = entity;
+		connector = of_graph_get_remote_port_parent(ep_node);
+		of_node_put(ep_node);
+
+		if (!connector) {
+			dev_warn(rcdu->dev,
+				 "no connector for encoder %s, skipping\n",
+				 encoder->full_name);
+			of_node_put(entity_ep_node);
+			of_node_put(encoder);
+			return 0;
+		}
+
+		break;
+	}
+
+	of_node_put(entity_ep_node);
+
+	if (encoder) {
+		/*
+		 * If an encoder has been found, get its type based on its
+		 * compatible string.
+		 */
+		unsigned int i;
+
+		for (i = 0; i < ARRAY_SIZE(encoders); ++i) {
+			if (of_device_is_compatible(encoder,
+						    encoders[i].compatible)) {
+				enc_type = encoders[i].type;
+				break;
+			}
+		}
+
+		if (i == ARRAY_SIZE(encoders)) {
+			dev_warn(rcdu->dev,
+				 "unknown encoder type for %s, skipping\n",
+				 encoder->full_name);
+			of_node_put(encoder);
+			of_node_put(connector);
+			return 0;
+		}
+	} else {
+		/*
+		 * If no encoder has been found the entity must be the
+		 * connector.
+		 */
+		connector = entity;
+	}
+
+	ret = rcar_du_encoder_init(rcdu, enc_type, output, NULL, connector);
+	of_node_put(encoder);
+	of_node_put(connector);
+
+	return ret < 0 ? ret : 1;
+}
+
+static int rcar_du_encoders_init_dt(struct rcar_du_device *rcdu)
+{
+	struct device_node *np = rcdu->dev->of_node;
+	struct device_node *prev = NULL;
+	unsigned int num_encoders = 0;
+
+	/*
+	 * Iterate over the endpoints and create one encoder for each output
+	 * pipeline.
+	 */
+	while (1) {
+		struct device_node *ep_node;
+		enum rcar_du_output output;
+		struct of_endpoint ep;
+		unsigned int i;
+		int ret;
+
+		ep_node = of_graph_get_next_endpoint(np, prev);
+		of_node_put(prev);
+		prev = ep_node;
+
+		if (ep_node == NULL)
+			break;
+
+		ret = of_graph_parse_endpoint(ep_node, &ep);
+		if (ret < 0) {
+			of_node_put(ep_node);
+			return ret;
+		}
+
+		/* Find the output route corresponding to the port number. */
+		for (i = 0; i < RCAR_DU_OUTPUT_MAX; ++i) {
+			if (rcdu->info->routes[i].possible_crtcs &&
+			    rcdu->info->routes[i].port == ep.port) {
+				output = i;
+				break;
+			}
+		}
+
+		if (i == RCAR_DU_OUTPUT_MAX) {
+			dev_warn(rcdu->dev,
+				 "port %u references unexisting output, skipping\n",
+				 ep.port);
+			continue;
+		}
+
+		/* Process the output pipeline. */
+		ret = rcar_du_encoders_init_dt_one(rcdu, output, &ep);
+		if (ret < 0) {
+			of_node_put(ep_node);
+			return ret;
+		}
+
+		num_encoders += ret;
+	}
+
+	return num_encoders;
+}
+
 int rcar_du_modeset_init(struct rcar_du_device *rcdu)
 {
 	static const unsigned int mmio_offsets[] = {
@@ -197,6 +398,7 @@ int rcar_du_modeset_init(struct rcar_du_device *rcdu)
 	struct drm_device *dev = rcdu->ddev;
 	struct drm_encoder *encoder;
 	struct drm_fbdev_cma *fbdev;
+	unsigned int num_encoders;
 	unsigned int num_groups;
 	unsigned int i;
 	int ret;
@@ -240,28 +442,15 @@ int rcar_du_modeset_init(struct rcar_du_device *rcdu)
 	if (ret < 0)
 		return ret;
 
-	for (i = 0; i < rcdu->pdata->num_encoders; ++i) {
-		const struct rcar_du_encoder_data *pdata =
-			&rcdu->pdata->encoders[i];
-		const struct rcar_du_output_routing *route =
-			&rcdu->info->routes[pdata->output];
-
-		if (pdata->type == RCAR_DU_ENCODER_UNUSED)
-			continue;
+	if (rcdu->pdata)
+		ret = rcar_du_encoders_init_pdata(rcdu);
+	else
+		ret = rcar_du_encoders_init_dt(rcdu);
 
-		if (pdata->output >= RCAR_DU_OUTPUT_MAX ||
-		    route->possible_crtcs == 0) {
-			dev_warn(rcdu->dev,
-				 "encoder %u references unexisting output %u, skipping\n",
-				 i, pdata->output);
-			continue;
-		}
+	if (ret < 0)
+		return ret;
 
-		ret = rcar_du_encoder_init(rcdu, pdata->type, pdata->output,
-					   pdata);
-		if (ret < 0)
-			return ret;
-	}
+	num_encoders = ret;
 
 	/* Set the possible CRTCs and possible clones. There's always at least
 	 * one way for all encoders to clone each other, set all bits in the
@@ -273,7 +462,7 @@ int rcar_du_modeset_init(struct rcar_du_device *rcdu)
 			&rcdu->info->routes[renc->output];
 
 		encoder->possible_crtcs = route->possible_crtcs;
-		encoder->possible_clones = (1 << rcdu->pdata->num_encoders) - 1;
+		encoder->possible_clones = (1 << num_encoders) - 1;
 	}
 
 	/* Now that the CRTCs have been initialized register the planes. */
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_kms.h b/drivers/gpu/drm/rcar-du/rcar_du_kms.h
index 5750e6af5655..07951d5fe38b 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_kms.h
+++ b/drivers/gpu/drm/rcar-du/rcar_du_kms.h
@@ -1,7 +1,7 @@
 /*
  * rcar_du_kms.h  --  R-Car Display Unit Mode Setting
  *
- * Copyright (C) 2013 Renesas Corporation
+ * Copyright (C) 2013-2014 Renesas Electronics Corporation
  *
  * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_lvdscon.c b/drivers/gpu/drm/rcar-du/rcar_du_lvdscon.c
index 21426bd234eb..115eed20db12 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_lvdscon.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_lvdscon.c
@@ -1,7 +1,7 @@
 /*
  * rcar_du_lvdscon.c  --  R-Car Display Unit LVDS Connector
  *
- * Copyright (C) 2013 Renesas Corporation
+ * Copyright (C) 2013-2014 Renesas Electronics Corporation
  *
  * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
@@ -15,6 +15,10 @@
 #include <drm/drm_crtc.h>
 #include <drm/drm_crtc_helper.h>
 
+#include <video/display_timing.h>
+#include <video/of_display_timing.h>
+#include <video/videomode.h>
+
 #include "rcar_du_drv.h"
 #include "rcar_du_encoder.h"
 #include "rcar_du_kms.h"
@@ -23,7 +27,7 @@
 struct rcar_du_lvds_connector {
 	struct rcar_du_connector connector;
 
-	const struct rcar_du_panel_data *panel;
+	struct rcar_du_panel_data panel;
 };
 
 #define to_rcar_lvds_connector(c) \
@@ -40,18 +44,9 @@ static int rcar_du_lvds_connector_get_modes(struct drm_connector *connector)
 		return 0;
 
 	mode->type = DRM_MODE_TYPE_PREFERRED | DRM_MODE_TYPE_DRIVER;
-	mode->clock = lvdscon->panel->mode.clock;
-	mode->hdisplay = lvdscon->panel->mode.hdisplay;
-	mode->hsync_start = lvdscon->panel->mode.hsync_start;
-	mode->hsync_end = lvdscon->panel->mode.hsync_end;
-	mode->htotal = lvdscon->panel->mode.htotal;
-	mode->vdisplay = lvdscon->panel->mode.vdisplay;
-	mode->vsync_start = lvdscon->panel->mode.vsync_start;
-	mode->vsync_end = lvdscon->panel->mode.vsync_end;
-	mode->vtotal = lvdscon->panel->mode.vtotal;
-	mode->flags = lvdscon->panel->mode.flags;
-
-	drm_mode_set_name(mode);
+
+	drm_display_mode_from_videomode(&lvdscon->panel.mode, mode);
+
 	drm_mode_probed_add(connector, mode);
 
 	return 1;
@@ -83,7 +78,8 @@ static const struct drm_connector_funcs connector_funcs = {
 
 int rcar_du_lvds_connector_init(struct rcar_du_device *rcdu,
 				struct rcar_du_encoder *renc,
-				const struct rcar_du_panel_data *panel)
+				const struct rcar_du_panel_data *panel,
+				/* TODO const */ struct device_node *np)
 {
 	struct rcar_du_lvds_connector *lvdscon;
 	struct drm_connector *connector;
@@ -93,11 +89,24 @@ int rcar_du_lvds_connector_init(struct rcar_du_device *rcdu,
 	if (lvdscon == NULL)
 		return -ENOMEM;
 
-	lvdscon->panel = panel;
+	if (panel) {
+		lvdscon->panel = *panel;
+	} else {
+		struct display_timing timing;
+
+		ret = of_get_display_timing(np, "panel-timing", &timing);
+		if (ret < 0)
+			return ret;
+
+		videomode_from_timing(&timing, &lvdscon->panel.mode);
+
+		of_property_read_u32(np, "width-mm", &lvdscon->panel.width_mm);
+		of_property_read_u32(np, "height-mm", &lvdscon->panel.height_mm);
+	}
 
 	connector = &lvdscon->connector.connector;
-	connector->display_info.width_mm = panel->width_mm;
-	connector->display_info.height_mm = panel->height_mm;
+	connector->display_info.width_mm = lvdscon->panel.width_mm;
+	connector->display_info.height_mm = lvdscon->panel.height_mm;
 
 	ret = drm_connector_init(rcdu->ddev, connector, &connector_funcs,
 				 DRM_MODE_CONNECTOR_LVDS);
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_lvdscon.h b/drivers/gpu/drm/rcar-du/rcar_du_lvdscon.h
index bff8683699ca..d11424d537f9 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_lvdscon.h
+++ b/drivers/gpu/drm/rcar-du/rcar_du_lvdscon.h
@@ -1,7 +1,7 @@
 /*
  * rcar_du_lvdscon.h  --  R-Car Display Unit LVDS Connector
  *
- * Copyright (C) 2013 Renesas Corporation
+ * Copyright (C) 2013-2014 Renesas Electronics Corporation
  *
  * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
@@ -20,6 +20,7 @@ struct rcar_du_panel_data;
 
 int rcar_du_lvds_connector_init(struct rcar_du_device *rcdu,
 				struct rcar_du_encoder *renc,
-				const struct rcar_du_panel_data *panel);
+				const struct rcar_du_panel_data *panel,
+				struct device_node *np);
 
 #endif /* __RCAR_DU_LVDSCON_H__ */
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_lvdsenc.c b/drivers/gpu/drm/rcar-du/rcar_du_lvdsenc.c
index df30a075d793..7cfb48ce1791 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_lvdsenc.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_lvdsenc.c
@@ -1,7 +1,7 @@
 /*
  * rcar_du_lvdsenc.c  --  R-Car Display Unit LVDS Encoder
  *
- * Copyright (C) 2013 Renesas Corporation
+ * Copyright (C) 2013-2014 Renesas Electronics Corporation
  *
  * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_lvdsenc.h b/drivers/gpu/drm/rcar-du/rcar_du_lvdsenc.h
index 7051c6de19ae..3303a55cec79 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_lvdsenc.h
+++ b/drivers/gpu/drm/rcar-du/rcar_du_lvdsenc.h
@@ -1,7 +1,7 @@
 /*
  * rcar_du_lvdsenc.h  --  R-Car Display Unit LVDS Encoder
  *
- * Copyright (C) 2013 Renesas Corporation
+ * Copyright (C) 2013-2014 Renesas Electronics Corporation
  *
  * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_plane.c b/drivers/gpu/drm/rcar-du/rcar_du_plane.c
index 3fb69d9ae61b..72a7cb47bd9f 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_plane.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_plane.c
@@ -1,7 +1,7 @@
 /*
  * rcar_du_plane.c  --  R-Car Display Unit Planes
  *
- * Copyright (C) 2013 Renesas Corporation
+ * Copyright (C) 2013-2014 Renesas Electronics Corporation
  *
  * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_plane.h b/drivers/gpu/drm/rcar-du/rcar_du_plane.h
index f94f9ce84998..3021288b1a89 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_plane.h
+++ b/drivers/gpu/drm/rcar-du/rcar_du_plane.h
@@ -1,7 +1,7 @@
 /*
  * rcar_du_plane.h  --  R-Car Display Unit Planes
  *
- * Copyright (C) 2013 Renesas Corporation
+ * Copyright (C) 2013-2014 Renesas Electronics Corporation
  *
  * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_vgacon.c b/drivers/gpu/drm/rcar-du/rcar_du_vgacon.c
index 8af3944d31b9..564a723ede03 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_vgacon.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_vgacon.c
@@ -1,7 +1,7 @@
 /*
  * rcar_du_vgacon.c  --  R-Car Display Unit VGA Connector
  *
- * Copyright (C) 2013 Renesas Corporation
+ * Copyright (C) 2013-2014 Renesas Electronics Corporation
  *
  * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_vgacon.h b/drivers/gpu/drm/rcar-du/rcar_du_vgacon.h
index b12b0cf7f117..112f50316e01 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_vgacon.h
+++ b/drivers/gpu/drm/rcar-du/rcar_du_vgacon.h
@@ -1,7 +1,7 @@
 /*
  * rcar_du_vgacon.h  --  R-Car Display Unit VGA Connector
  *
- * Copyright (C) 2013 Renesas Corporation
+ * Copyright (C) 2013-2014 Renesas Electronics Corporation
  *
  * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
diff --git a/drivers/gpu/drm/savage/savage_bci.c b/drivers/gpu/drm/savage/savage_bci.c
index c97cdc9ab239..d47dff95fe52 100644
--- a/drivers/gpu/drm/savage/savage_bci.c
+++ b/drivers/gpu/drm/savage/savage_bci.c
@@ -556,7 +556,7 @@ int savage_driver_load(struct drm_device *dev, unsigned long chipset)
 /*
  * Initialize mappings. On Savage4 and SavageIX the alignment
  * and size of the aperture is not suitable for automatic MTRR setup
- * in drm_addmap. Therefore we add them manually before the maps are
+ * in drm_legacy_addmap. Therefore we add them manually before the maps are
  * initialized, and tear them down on last close.
  */
 int savage_driver_firstopen(struct drm_device *dev)
@@ -624,19 +624,20 @@ int savage_driver_firstopen(struct drm_device *dev)
 		/* Automatic MTRR setup will do the right thing. */
 	}
 
-	ret = drm_addmap(dev, mmio_base, SAVAGE_MMIO_SIZE, _DRM_REGISTERS,
-			 _DRM_READ_ONLY, &dev_priv->mmio);
+	ret = drm_legacy_addmap(dev, mmio_base, SAVAGE_MMIO_SIZE,
+				_DRM_REGISTERS, _DRM_READ_ONLY,
+				&dev_priv->mmio);
 	if (ret)
 		return ret;
 
-	ret = drm_addmap(dev, fb_base, fb_size, _DRM_FRAME_BUFFER,
-			 _DRM_WRITE_COMBINING, &dev_priv->fb);
+	ret = drm_legacy_addmap(dev, fb_base, fb_size, _DRM_FRAME_BUFFER,
+				_DRM_WRITE_COMBINING, &dev_priv->fb);
 	if (ret)
 		return ret;
 
-	ret = drm_addmap(dev, aperture_base, SAVAGE_APERTURE_SIZE,
-			 _DRM_FRAME_BUFFER, _DRM_WRITE_COMBINING,
-			 &dev_priv->aperture);
+	ret = drm_legacy_addmap(dev, aperture_base, SAVAGE_APERTURE_SIZE,
+				_DRM_FRAME_BUFFER, _DRM_WRITE_COMBINING,
+				&dev_priv->aperture);
 	return ret;
 }
 
@@ -698,14 +699,14 @@ static int savage_do_init_bci(struct drm_device * dev, drm_savage_init_t * init)
 	dev_priv->texture_offset = init->texture_offset;
 	dev_priv->texture_size = init->texture_size;
 
-	dev_priv->sarea = drm_getsarea(dev);
+	dev_priv->sarea = drm_legacy_getsarea(dev);
 	if (!dev_priv->sarea) {
 		DRM_ERROR("could not find sarea!\n");
 		savage_do_cleanup_bci(dev);
 		return -EINVAL;
 	}
 	if (init->status_offset != 0) {
-		dev_priv->status = drm_core_findmap(dev, init->status_offset);
+		dev_priv->status = drm_legacy_findmap(dev, init->status_offset);
 		if (!dev_priv->status) {
 			DRM_ERROR("could not find shadow status region!\n");
 			savage_do_cleanup_bci(dev);
@@ -716,14 +717,14 @@ static int savage_do_init_bci(struct drm_device * dev, drm_savage_init_t * init)
 	}
 	if (dev_priv->dma_type == SAVAGE_DMA_AGP && init->buffers_offset) {
 		dev->agp_buffer_token = init->buffers_offset;
-		dev->agp_buffer_map = drm_core_findmap(dev,
+		dev->agp_buffer_map = drm_legacy_findmap(dev,
 						       init->buffers_offset);
 		if (!dev->agp_buffer_map) {
 			DRM_ERROR("could not find DMA buffer region!\n");
 			savage_do_cleanup_bci(dev);
 			return -EINVAL;
 		}
-		drm_core_ioremap(dev->agp_buffer_map, dev);
+		drm_legacy_ioremap(dev->agp_buffer_map, dev);
 		if (!dev->agp_buffer_map->handle) {
 			DRM_ERROR("failed to ioremap DMA buffer region!\n");
 			savage_do_cleanup_bci(dev);
@@ -732,7 +733,7 @@ static int savage_do_init_bci(struct drm_device * dev, drm_savage_init_t * init)
 	}
 	if (init->agp_textures_offset) {
 		dev_priv->agp_textures =
-		    drm_core_findmap(dev, init->agp_textures_offset);
+		    drm_legacy_findmap(dev, init->agp_textures_offset);
 		if (!dev_priv->agp_textures) {
 			DRM_ERROR("could not find agp texture region!\n");
 			savage_do_cleanup_bci(dev);
@@ -755,7 +756,7 @@ static int savage_do_init_bci(struct drm_device * dev, drm_savage_init_t * init)
 			savage_do_cleanup_bci(dev);
 			return -EINVAL;
 		}
-		dev_priv->cmd_dma = drm_core_findmap(dev, init->cmd_dma_offset);
+		dev_priv->cmd_dma = drm_legacy_findmap(dev, init->cmd_dma_offset);
 		if (!dev_priv->cmd_dma) {
 			DRM_ERROR("could not find command DMA region!\n");
 			savage_do_cleanup_bci(dev);
@@ -768,7 +769,7 @@ static int savage_do_init_bci(struct drm_device * dev, drm_savage_init_t * init)
 				savage_do_cleanup_bci(dev);
 				return -EINVAL;
 			}
-			drm_core_ioremap(dev_priv->cmd_dma, dev);
+			drm_legacy_ioremap(dev_priv->cmd_dma, dev);
 			if (!dev_priv->cmd_dma->handle) {
 				DRM_ERROR("failed to ioremap command "
 					  "DMA region!\n");
@@ -894,11 +895,11 @@ static int savage_do_cleanup_bci(struct drm_device * dev)
 	} else if (dev_priv->cmd_dma && dev_priv->cmd_dma->handle &&
 		   dev_priv->cmd_dma->type == _DRM_AGP &&
 		   dev_priv->dma_type == SAVAGE_DMA_AGP)
-		drm_core_ioremapfree(dev_priv->cmd_dma, dev);
+		drm_legacy_ioremapfree(dev_priv->cmd_dma, dev);
 
 	if (dev_priv->dma_type == SAVAGE_DMA_AGP &&
 	    dev->agp_buffer_map && dev->agp_buffer_map->handle) {
-		drm_core_ioremapfree(dev->agp_buffer_map, dev);
+		drm_legacy_ioremapfree(dev->agp_buffer_map, dev);
 		/* make sure the next instance (which may be running
 		 * in PCI mode) doesn't try to use an old
 		 * agp_buffer_map. */
@@ -1050,7 +1051,7 @@ void savage_reclaim_buffers(struct drm_device *dev, struct drm_file *file_priv)
 		return;
 
 	if (file_priv->master && file_priv->master->lock.hw_lock) {
-		drm_idlelock_take(&file_priv->master->lock);
+		drm_legacy_idlelock_take(&file_priv->master->lock);
 		release_idlelock = 1;
 	}
 
@@ -1069,7 +1070,7 @@ void savage_reclaim_buffers(struct drm_device *dev, struct drm_file *file_priv)
 	}
 
 	if (release_idlelock)
-		drm_idlelock_release(&file_priv->master->lock);
+		drm_legacy_idlelock_release(&file_priv->master->lock);
 }
 
 const struct drm_ioctl_desc savage_ioctls[] = {
diff --git a/drivers/gpu/drm/savage/savage_drv.c b/drivers/gpu/drm/savage/savage_drv.c
index 3c030216e888..21aed1febeb4 100644
--- a/drivers/gpu/drm/savage/savage_drv.c
+++ b/drivers/gpu/drm/savage/savage_drv.c
@@ -40,7 +40,7 @@ static const struct file_operations savage_driver_fops = {
 	.open = drm_open,
 	.release = drm_release,
 	.unlocked_ioctl = drm_ioctl,
-	.mmap = drm_mmap,
+	.mmap = drm_legacy_mmap,
 	.poll = drm_poll,
 #ifdef CONFIG_COMPAT
 	.compat_ioctl = drm_compat_ioctl,
@@ -57,6 +57,7 @@ static struct drm_driver driver = {
 	.preclose = savage_reclaim_buffers,
 	.lastclose = savage_driver_lastclose,
 	.unload = savage_driver_unload,
+	.set_busid = drm_pci_set_busid,
 	.ioctls = savage_ioctls,
 	.dma_ioctl = savage_bci_buffers,
 	.fops = &savage_driver_fops,
diff --git a/drivers/gpu/drm/savage/savage_drv.h b/drivers/gpu/drm/savage/savage_drv.h
index 335f8fcf1041..37b699571ad0 100644
--- a/drivers/gpu/drm/savage/savage_drv.h
+++ b/drivers/gpu/drm/savage/savage_drv.h
@@ -26,6 +26,8 @@
 #ifndef __SAVAGE_DRV_H__
 #define __SAVAGE_DRV_H__
 
+#include <drm/drm_legacy.h>
+
 #define DRIVER_AUTHOR	"Felix Kuehling"
 
 #define DRIVER_NAME	"savage"
diff --git a/drivers/gpu/drm/shmobile/shmob_drm_backlight.c b/drivers/gpu/drm/shmobile/shmob_drm_backlight.c
index 463aee18f774..33dd41afea0e 100644
--- a/drivers/gpu/drm/shmobile/shmob_drm_backlight.c
+++ b/drivers/gpu/drm/shmobile/shmob_drm_backlight.c
@@ -1,7 +1,7 @@
 /*
  * shmob_drm_backlight.c  --  SH Mobile DRM Backlight
  *
- * Copyright (C) 2012 Renesas Corporation
+ * Copyright (C) 2012 Renesas Electronics Corporation
  *
  * Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
diff --git a/drivers/gpu/drm/shmobile/shmob_drm_backlight.h b/drivers/gpu/drm/shmobile/shmob_drm_backlight.h
index 9477595d2ff3..bac719ecc301 100644
--- a/drivers/gpu/drm/shmobile/shmob_drm_backlight.h
+++ b/drivers/gpu/drm/shmobile/shmob_drm_backlight.h
@@ -1,7 +1,7 @@
 /*
  * shmob_drm_backlight.h  --  SH Mobile DRM Backlight
  *
- * Copyright (C) 2012 Renesas Corporation
+ * Copyright (C) 2012 Renesas Electronics Corporation
  *
  * Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
diff --git a/drivers/gpu/drm/shmobile/shmob_drm_crtc.c b/drivers/gpu/drm/shmobile/shmob_drm_crtc.c
index 47875de89010..0ddce4d046d9 100644
--- a/drivers/gpu/drm/shmobile/shmob_drm_crtc.c
+++ b/drivers/gpu/drm/shmobile/shmob_drm_crtc.c
@@ -1,7 +1,7 @@
 /*
  * shmob_drm_crtc.c  --  SH Mobile DRM CRTCs
  *
- * Copyright (C) 2012 Renesas Corporation
+ * Copyright (C) 2012 Renesas Electronics Corporation
  *
  * Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
diff --git a/drivers/gpu/drm/shmobile/shmob_drm_crtc.h b/drivers/gpu/drm/shmobile/shmob_drm_crtc.h
index e5bd109c4c38..eddad6dcc88a 100644
--- a/drivers/gpu/drm/shmobile/shmob_drm_crtc.h
+++ b/drivers/gpu/drm/shmobile/shmob_drm_crtc.h
@@ -1,7 +1,7 @@
 /*
  * shmob_drm_crtc.h  --  SH Mobile DRM CRTCs
  *
- * Copyright (C) 2012 Renesas Corporation
+ * Copyright (C) 2012 Renesas Electronics Corporation
  *
  * Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
diff --git a/drivers/gpu/drm/shmobile/shmob_drm_drv.c b/drivers/gpu/drm/shmobile/shmob_drm_drv.c
index ff4ba483b602..e62cbde81e50 100644
--- a/drivers/gpu/drm/shmobile/shmob_drm_drv.c
+++ b/drivers/gpu/drm/shmobile/shmob_drm_drv.c
@@ -1,7 +1,7 @@
 /*
  * shmob_drm_drv.c  --  SH Mobile DRM driver
  *
- * Copyright (C) 2012 Renesas Corporation
+ * Copyright (C) 2012 Renesas Electronics Corporation
  *
  * Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
@@ -267,6 +267,7 @@ static struct drm_driver shmob_drm_driver = {
 	.load			= shmob_drm_load,
 	.unload			= shmob_drm_unload,
 	.preclose		= shmob_drm_preclose,
+	.set_busid		= drm_platform_set_busid,
 	.irq_handler		= shmob_drm_irq,
 	.get_vblank_counter	= drm_vblank_count,
 	.enable_vblank		= shmob_drm_enable_vblank,
diff --git a/drivers/gpu/drm/shmobile/shmob_drm_drv.h b/drivers/gpu/drm/shmobile/shmob_drm_drv.h
index 4d46b811b5a7..02ea315ba69a 100644
--- a/drivers/gpu/drm/shmobile/shmob_drm_drv.h
+++ b/drivers/gpu/drm/shmobile/shmob_drm_drv.h
@@ -1,7 +1,7 @@
 /*
  * shmob_drm.h  --  SH Mobile DRM driver
  *
- * Copyright (C) 2012 Renesas Corporation
+ * Copyright (C) 2012 Renesas Electronics Corporation
  *
  * Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
diff --git a/drivers/gpu/drm/shmobile/shmob_drm_kms.c b/drivers/gpu/drm/shmobile/shmob_drm_kms.c
index fc0ef0ca7d04..aaf98ace4a90 100644
--- a/drivers/gpu/drm/shmobile/shmob_drm_kms.c
+++ b/drivers/gpu/drm/shmobile/shmob_drm_kms.c
@@ -1,7 +1,7 @@
 /*
  * shmob_drm_kms.c  --  SH Mobile DRM Mode Setting
  *
- * Copyright (C) 2012 Renesas Corporation
+ * Copyright (C) 2012 Renesas Electronics Corporation
  *
  * Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
diff --git a/drivers/gpu/drm/shmobile/shmob_drm_kms.h b/drivers/gpu/drm/shmobile/shmob_drm_kms.h
index 9495c9111308..06d5b7caa026 100644
--- a/drivers/gpu/drm/shmobile/shmob_drm_kms.h
+++ b/drivers/gpu/drm/shmobile/shmob_drm_kms.h
@@ -1,7 +1,7 @@
 /*
  * shmob_drm_kms.h  --  SH Mobile DRM Mode Setting
  *
- * Copyright (C) 2012 Renesas Corporation
+ * Copyright (C) 2012 Renesas Electronics Corporation
  *
  * Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
diff --git a/drivers/gpu/drm/shmobile/shmob_drm_plane.c b/drivers/gpu/drm/shmobile/shmob_drm_plane.c
index 060ae03e5f9b..1805bb23b113 100644
--- a/drivers/gpu/drm/shmobile/shmob_drm_plane.c
+++ b/drivers/gpu/drm/shmobile/shmob_drm_plane.c
@@ -1,7 +1,7 @@
 /*
  * shmob_drm_plane.c  --  SH Mobile DRM Planes
  *
- * Copyright (C) 2012 Renesas Corporation
+ * Copyright (C) 2012 Renesas Electronics Corporation
  *
  * Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
diff --git a/drivers/gpu/drm/shmobile/shmob_drm_plane.h b/drivers/gpu/drm/shmobile/shmob_drm_plane.h
index 99623d05e3b0..a58cc1fc3240 100644
--- a/drivers/gpu/drm/shmobile/shmob_drm_plane.h
+++ b/drivers/gpu/drm/shmobile/shmob_drm_plane.h
@@ -1,7 +1,7 @@
 /*
  * shmob_drm_plane.h  --  SH Mobile DRM Planes
  *
- * Copyright (C) 2012 Renesas Corporation
+ * Copyright (C) 2012 Renesas Electronics Corporation
  *
  * Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
diff --git a/drivers/gpu/drm/shmobile/shmob_drm_regs.h b/drivers/gpu/drm/shmobile/shmob_drm_regs.h
index 7923cdd6368e..ea17d4415b9e 100644
--- a/drivers/gpu/drm/shmobile/shmob_drm_regs.h
+++ b/drivers/gpu/drm/shmobile/shmob_drm_regs.h
@@ -1,7 +1,7 @@
 /*
  * shmob_drm_regs.h  --  SH Mobile DRM registers
  *
- * Copyright (C) 2012 Renesas Corporation
+ * Copyright (C) 2012 Renesas Electronics Corporation
  *
  * Laurent Pinchart (laurent.pinchart@ideasonboard.com)
  *
diff --git a/drivers/gpu/drm/sis/sis_drv.c b/drivers/gpu/drm/sis/sis_drv.c
index 756f787b7143..79bce76cb8f7 100644
--- a/drivers/gpu/drm/sis/sis_drv.c
+++ b/drivers/gpu/drm/sis/sis_drv.c
@@ -70,7 +70,7 @@ static const struct file_operations sis_driver_fops = {
 	.open = drm_open,
 	.release = drm_release,
 	.unlocked_ioctl = drm_ioctl,
-	.mmap = drm_mmap,
+	.mmap = drm_legacy_mmap,
 	.poll = drm_poll,
 #ifdef CONFIG_COMPAT
 	.compat_ioctl = drm_compat_ioctl,
@@ -108,6 +108,7 @@ static struct drm_driver driver = {
 	.open = sis_driver_open,
 	.preclose = sis_reclaim_buffers_locked,
 	.postclose = sis_driver_postclose,
+	.set_busid = drm_pci_set_busid,
 	.dma_quiescent = sis_idle,
 	.lastclose = sis_lastclose,
 	.ioctls = sis_ioctls,
diff --git a/drivers/gpu/drm/sis/sis_drv.h b/drivers/gpu/drm/sis/sis_drv.h
index c31c0253054d..16f972b2a76a 100644
--- a/drivers/gpu/drm/sis/sis_drv.h
+++ b/drivers/gpu/drm/sis/sis_drv.h
@@ -28,6 +28,8 @@
 #ifndef _SIS_DRV_H_
 #define _SIS_DRV_H_
 
+#include <drm/drm_legacy.h>
+
 /* General customization:
  */
 
diff --git a/drivers/gpu/drm/sis/sis_mm.c b/drivers/gpu/drm/sis/sis_mm.c
index 77f288e4a0a6..93ad8a5704d1 100644
--- a/drivers/gpu/drm/sis/sis_mm.c
+++ b/drivers/gpu/drm/sis/sis_mm.c
@@ -319,12 +319,12 @@ void sis_reclaim_buffers_locked(struct drm_device *dev,
 	if (!(file->minor->master && file->master->lock.hw_lock))
 		return;
 
-	drm_idlelock_take(&file->master->lock);
+	drm_legacy_idlelock_take(&file->master->lock);
 
 	mutex_lock(&dev->struct_mutex);
 	if (list_empty(&file_priv->obj_list)) {
 		mutex_unlock(&dev->struct_mutex);
-		drm_idlelock_release(&file->master->lock);
+		drm_legacy_idlelock_release(&file->master->lock);
 
 		return;
 	}
@@ -345,7 +345,7 @@ void sis_reclaim_buffers_locked(struct drm_device *dev,
 	}
 	mutex_unlock(&dev->struct_mutex);
 
-	drm_idlelock_release(&file->master->lock);
+	drm_legacy_idlelock_release(&file->master->lock);
 
 	return;
 }
diff --git a/drivers/gpu/drm/sti/sti_vtac.c b/drivers/gpu/drm/sti/sti_vtac.c
index 82a51d488434..97bcdac23ae1 100644
--- a/drivers/gpu/drm/sti/sti_vtac.c
+++ b/drivers/gpu/drm/sti/sti_vtac.c
@@ -56,8 +56,16 @@ struct sti_vtac_mode {
 	u32 phyts_per_pixel;
 };
 
-static const struct sti_vtac_mode vtac_mode_main = {0x2, 0x2, VTAC_5_PPP};
-static const struct sti_vtac_mode vtac_mode_aux = {0x1, 0x0, VTAC_17_PPP};
+static const struct sti_vtac_mode vtac_mode_main = {
+	.vid_in_width = 0x2,
+	.phyts_width = 0x2,
+	.phyts_per_pixel = VTAC_5_PPP,
+};
+static const struct sti_vtac_mode vtac_mode_aux = {
+	.vid_in_width = 0x1,
+	.phyts_width = 0x0,
+	.phyts_per_pixel = VTAC_17_PPP,
+};
 
 /**
  * VTAC structure
diff --git a/drivers/gpu/drm/tdfx/tdfx_drv.c b/drivers/gpu/drm/tdfx/tdfx_drv.c
index 3492ca5c46d3..fab5ebcb0fef 100644
--- a/drivers/gpu/drm/tdfx/tdfx_drv.c
+++ b/drivers/gpu/drm/tdfx/tdfx_drv.c
@@ -36,6 +36,7 @@
 #include "tdfx_drv.h"
 
 #include <drm/drm_pciids.h>
+#include <drm/drm_legacy.h>
 
 static struct pci_device_id pciidlist[] = {
 	tdfx_PCI_IDS
@@ -46,7 +47,7 @@ static const struct file_operations tdfx_driver_fops = {
 	.open = drm_open,
 	.release = drm_release,
 	.unlocked_ioctl = drm_ioctl,
-	.mmap = drm_mmap,
+	.mmap = drm_legacy_mmap,
 	.poll = drm_poll,
 #ifdef CONFIG_COMPAT
 	.compat_ioctl = drm_compat_ioctl,
@@ -55,6 +56,7 @@ static const struct file_operations tdfx_driver_fops = {
 };
 
 static struct drm_driver driver = {
+	.set_busid = drm_pci_set_busid,
 	.fops = &tdfx_driver_fops,
 	.name = DRIVER_NAME,
 	.desc = DRIVER_DESC,
diff --git a/drivers/gpu/drm/tegra/dpaux.c b/drivers/gpu/drm/tegra/dpaux.c
index 708f783ead47..d6b55e3e3716 100644
--- a/drivers/gpu/drm/tegra/dpaux.c
+++ b/drivers/gpu/drm/tegra/dpaux.c
@@ -533,9 +533,9 @@ int tegra_dpaux_train(struct tegra_dpaux *dpaux, struct drm_dp_link *link,
 
 	for (i = 0; i < link->num_lanes; i++)
 		values[i] = DP_TRAIN_MAX_PRE_EMPHASIS_REACHED |
-			    DP_TRAIN_PRE_EMPHASIS_0 |
+			    DP_TRAIN_PRE_EMPH_LEVEL_0 |
 			    DP_TRAIN_MAX_SWING_REACHED |
-			    DP_TRAIN_VOLTAGE_SWING_400;
+			    DP_TRAIN_VOLTAGE_SWING_LEVEL_0;
 
 	err = drm_dp_dpcd_write(&dpaux->aux, DP_TRAINING_LANE0_SET, values,
 				link->num_lanes);
diff --git a/drivers/gpu/drm/tegra/gem.h b/drivers/gpu/drm/tegra/gem.h
index 43a25c853357..6538b56780c2 100644
--- a/drivers/gpu/drm/tegra/gem.h
+++ b/drivers/gpu/drm/tegra/gem.h
@@ -15,6 +15,7 @@
 
 #include <drm/drm.h>
 #include <drm/drmP.h>
+#include <drm/drm_gem.h>
 
 #define TEGRA_BO_BOTTOM_UP (1 << 0)
 
diff --git a/drivers/gpu/drm/tilcdc/tilcdc_drv.c b/drivers/gpu/drm/tilcdc/tilcdc_drv.c
index 6be623b4a86f..79a34cbd29f5 100644
--- a/drivers/gpu/drm/tilcdc/tilcdc_drv.c
+++ b/drivers/gpu/drm/tilcdc/tilcdc_drv.c
@@ -84,6 +84,7 @@ static int modeset_init(struct drm_device *dev)
 	if ((priv->num_encoders == 0) || (priv->num_connectors == 0)) {
 		/* oh nos! */
 		dev_err(dev->dev, "no encoders/connectors found\n");
+		drm_mode_config_cleanup(dev);
 		return -ENXIO;
 	}
 
@@ -172,33 +173,37 @@ static int tilcdc_load(struct drm_device *dev, unsigned long flags)
 	dev->dev_private = priv;
 
 	priv->wq = alloc_ordered_workqueue("tilcdc", 0);
+	if (!priv->wq) {
+		ret = -ENOMEM;
+		goto fail_free_priv;
+	}
 
 	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
 	if (!res) {
 		dev_err(dev->dev, "failed to get memory resource\n");
 		ret = -EINVAL;
-		goto fail;
+		goto fail_free_wq;
 	}
 
 	priv->mmio = ioremap_nocache(res->start, resource_size(res));
 	if (!priv->mmio) {
 		dev_err(dev->dev, "failed to ioremap\n");
 		ret = -ENOMEM;
-		goto fail;
+		goto fail_free_wq;
 	}
 
 	priv->clk = clk_get(dev->dev, "fck");
 	if (IS_ERR(priv->clk)) {
 		dev_err(dev->dev, "failed to get functional clock\n");
 		ret = -ENODEV;
-		goto fail;
+		goto fail_iounmap;
 	}
 
 	priv->disp_clk = clk_get(dev->dev, "dpll_disp_ck");
 	if (IS_ERR(priv->clk)) {
 		dev_err(dev->dev, "failed to get display clock\n");
 		ret = -ENODEV;
-		goto fail;
+		goto fail_put_clk;
 	}
 
 #ifdef CONFIG_CPU_FREQ
@@ -208,7 +213,7 @@ static int tilcdc_load(struct drm_device *dev, unsigned long flags)
 			CPUFREQ_TRANSITION_NOTIFIER);
 	if (ret) {
 		dev_err(dev->dev, "failed to register cpufreq notifier\n");
-		goto fail;
+		goto fail_put_disp_clk;
 	}
 #endif
 
@@ -253,13 +258,13 @@ static int tilcdc_load(struct drm_device *dev, unsigned long flags)
 	ret = modeset_init(dev);
 	if (ret < 0) {
 		dev_err(dev->dev, "failed to initialize mode setting\n");
-		goto fail;
+		goto fail_cpufreq_unregister;
 	}
 
 	ret = drm_vblank_init(dev, 1);
 	if (ret < 0) {
 		dev_err(dev->dev, "failed to initialize vblank\n");
-		goto fail;
+		goto fail_mode_config_cleanup;
 	}
 
 	pm_runtime_get_sync(dev->dev);
@@ -267,7 +272,7 @@ static int tilcdc_load(struct drm_device *dev, unsigned long flags)
 	pm_runtime_put_sync(dev->dev);
 	if (ret < 0) {
 		dev_err(dev->dev, "failed to install IRQ handler\n");
-		goto fail;
+		goto fail_vblank_cleanup;
 	}
 
 	platform_set_drvdata(pdev, dev);
@@ -283,13 +288,48 @@ static int tilcdc_load(struct drm_device *dev, unsigned long flags)
 	priv->fbdev = drm_fbdev_cma_init(dev, bpp,
 			dev->mode_config.num_crtc,
 			dev->mode_config.num_connector);
+	if (IS_ERR(priv->fbdev)) {
+		ret = PTR_ERR(priv->fbdev);
+		goto fail_irq_uninstall;
+	}
 
 	drm_kms_helper_poll_init(dev);
 
 	return 0;
 
-fail:
-	tilcdc_unload(dev);
+fail_irq_uninstall:
+	pm_runtime_get_sync(dev->dev);
+	drm_irq_uninstall(dev);
+	pm_runtime_put_sync(dev->dev);
+
+fail_vblank_cleanup:
+	drm_vblank_cleanup(dev);
+
+fail_mode_config_cleanup:
+	drm_mode_config_cleanup(dev);
+
+fail_cpufreq_unregister:
+	pm_runtime_disable(dev->dev);
+#ifdef CONFIG_CPU_FREQ
+	cpufreq_unregister_notifier(&priv->freq_transition,
+			CPUFREQ_TRANSITION_NOTIFIER);
+fail_put_disp_clk:
+	clk_put(priv->disp_clk);
+#endif
+
+fail_put_clk:
+	clk_put(priv->clk);
+
+fail_iounmap:
+	iounmap(priv->mmio);
+
+fail_free_wq:
+	flush_workqueue(priv->wq);
+	destroy_workqueue(priv->wq);
+
+fail_free_priv:
+	dev->dev_private = NULL;
+	kfree(priv);
 	return ret;
 }
 
@@ -502,6 +542,7 @@ static struct drm_driver tilcdc_driver = {
 	.unload             = tilcdc_unload,
 	.preclose           = tilcdc_preclose,
 	.lastclose          = tilcdc_lastclose,
+	.set_busid          = drm_platform_set_busid,
 	.irq_handler        = tilcdc_irq,
 	.irq_preinstall     = tilcdc_irq_preinstall,
 	.irq_postinstall    = tilcdc_irq_postinstall,
diff --git a/drivers/gpu/drm/tilcdc/tilcdc_panel.c b/drivers/gpu/drm/tilcdc/tilcdc_panel.c
index 4c7aa1d8134f..7a0315855e90 100644
--- a/drivers/gpu/drm/tilcdc/tilcdc_panel.c
+++ b/drivers/gpu/drm/tilcdc/tilcdc_panel.c
@@ -18,6 +18,7 @@
 #include <linux/pinctrl/pinmux.h>
 #include <linux/pinctrl/consumer.h>
 #include <linux/backlight.h>
+#include <linux/gpio/consumer.h>
 #include <video/display_timing.h>
 #include <video/of_display_timing.h>
 #include <video/videomode.h>
@@ -29,6 +30,7 @@ struct panel_module {
 	struct tilcdc_panel_info *info;
 	struct display_timings *timings;
 	struct backlight_device *backlight;
+	struct gpio_desc *enable_gpio;
 };
 #define to_panel_module(x) container_of(x, struct panel_module, base)
 
@@ -55,13 +57,17 @@ static void panel_encoder_dpms(struct drm_encoder *encoder, int mode)
 {
 	struct panel_encoder *panel_encoder = to_panel_encoder(encoder);
 	struct backlight_device *backlight = panel_encoder->mod->backlight;
+	struct gpio_desc *gpio = panel_encoder->mod->enable_gpio;
 
-	if (!backlight)
-		return;
+	if (backlight) {
+		backlight->props.power = mode == DRM_MODE_DPMS_ON ?
+					 FB_BLANK_UNBLANK : FB_BLANK_POWERDOWN;
+		backlight_update_status(backlight);
+	}
 
-	backlight->props.power = mode == DRM_MODE_DPMS_ON
-				     ? FB_BLANK_UNBLANK : FB_BLANK_POWERDOWN;
-	backlight_update_status(backlight);
+	if (gpio)
+		gpiod_set_value_cansleep(gpio,
+					 mode == DRM_MODE_DPMS_ON ? 1 : 0);
 }
 
 static bool panel_encoder_mode_fixup(struct drm_encoder *encoder,
@@ -311,6 +317,7 @@ static struct tilcdc_panel_info *of_get_panel_info(struct device_node *np)
 	info = kzalloc(sizeof(*info), GFP_KERNEL);
 	if (!info) {
 		pr_err("%s: allocation failed\n", __func__);
+		of_node_put(info_np);
 		return NULL;
 	}
 
@@ -331,22 +338,21 @@ static struct tilcdc_panel_info *of_get_panel_info(struct device_node *np)
 	if (ret) {
 		pr_err("%s: error reading panel-info properties\n", __func__);
 		kfree(info);
+		of_node_put(info_np);
 		return NULL;
 	}
+	of_node_put(info_np);
 
 	return info;
 }
 
-static struct of_device_id panel_of_match[];
-
 static int panel_probe(struct platform_device *pdev)
 {
-	struct device_node *node = pdev->dev.of_node;
+	struct device_node *bl_node, *node = pdev->dev.of_node;
 	struct panel_module *panel_mod;
 	struct tilcdc_module *mod;
 	struct pinctrl *pinctrl;
-	int ret = -EINVAL;
-
+	int ret;
 
 	/* bail out early if no DT data: */
 	if (!node) {
@@ -354,10 +360,40 @@ static int panel_probe(struct platform_device *pdev)
 		return -ENXIO;
 	}
 
-	panel_mod = kzalloc(sizeof(*panel_mod), GFP_KERNEL);
+	panel_mod = devm_kzalloc(&pdev->dev, sizeof(*panel_mod), GFP_KERNEL);
 	if (!panel_mod)
 		return -ENOMEM;
 
+	bl_node = of_parse_phandle(node, "backlight", 0);
+	if (bl_node) {
+		panel_mod->backlight = of_find_backlight_by_node(bl_node);
+		of_node_put(bl_node);
+
+		if (!panel_mod->backlight)
+			return -EPROBE_DEFER;
+
+		dev_info(&pdev->dev, "found backlight\n");
+	}
+
+	panel_mod->enable_gpio = devm_gpiod_get(&pdev->dev, "enable");
+	if (IS_ERR(panel_mod->enable_gpio)) {
+		ret = PTR_ERR(panel_mod->enable_gpio);
+		if (ret != -ENOENT) {
+			dev_err(&pdev->dev, "failed to request enable GPIO\n");
+			goto fail_backlight;
+		}
+
+		/* Optional GPIO is not here, continue silently. */
+		panel_mod->enable_gpio = NULL;
+	} else {
+		ret = gpiod_direction_output(panel_mod->enable_gpio, 0);
+		if (ret < 0) {
+			dev_err(&pdev->dev, "failed to setup GPIO\n");
+			goto fail_backlight;
+		}
+		dev_info(&pdev->dev, "found enable GPIO\n");
+	}
+
 	mod = &panel_mod->base;
 	pdev->dev.platform_data = mod;
 
@@ -370,29 +406,30 @@ static int panel_probe(struct platform_device *pdev)
 	panel_mod->timings = of_get_display_timings(node);
 	if (!panel_mod->timings) {
 		dev_err(&pdev->dev, "could not get panel timings\n");
+		ret = -EINVAL;
 		goto fail_free;
 	}
 
 	panel_mod->info = of_get_panel_info(node);
 	if (!panel_mod->info) {
 		dev_err(&pdev->dev, "could not get panel info\n");
+		ret = -EINVAL;
 		goto fail_timings;
 	}
 
 	mod->preferred_bpp = panel_mod->info->bpp;
 
-	panel_mod->backlight = of_find_backlight_by_node(node);
-	if (panel_mod->backlight)
-		dev_info(&pdev->dev, "found backlight\n");
-
 	return 0;
 
 fail_timings:
 	display_timings_release(panel_mod->timings);
 
 fail_free:
-	kfree(panel_mod);
 	tilcdc_module_cleanup(mod);
+
+fail_backlight:
+	if (panel_mod->backlight)
+		put_device(&panel_mod->backlight->dev);
 	return ret;
 }
 
@@ -400,12 +437,15 @@ static int panel_remove(struct platform_device *pdev)
 {
 	struct tilcdc_module *mod = dev_get_platdata(&pdev->dev);
 	struct panel_module *panel_mod = to_panel_module(mod);
+	struct backlight_device *backlight = panel_mod->backlight;
+
+	if (backlight)
+		put_device(&backlight->dev);
 
 	display_timings_release(panel_mod->timings);
 
 	tilcdc_module_cleanup(mod);
 	kfree(panel_mod->info);
-	kfree(panel_mod);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c
index 3da89d5dab60..8f5cec67c47d 100644
--- a/drivers/gpu/drm/ttm/ttm_bo.c
+++ b/drivers/gpu/drm/ttm/ttm_bo.c
@@ -40,6 +40,7 @@
 #include <linux/file.h>
 #include <linux/module.h>
 #include <linux/atomic.h>
+#include <linux/reservation.h>
 
 #define TTM_ASSERT_LOCKED(param)
 #define TTM_DEBUG(fmt, arg...)
@@ -53,12 +54,13 @@ static struct attribute ttm_bo_count = {
 	.mode = S_IRUGO
 };
 
-static inline int ttm_mem_type_from_flags(uint32_t flags, uint32_t *mem_type)
+static inline int ttm_mem_type_from_place(const struct ttm_place *place,
+					  uint32_t *mem_type)
 {
 	int i;
 
 	for (i = 0; i <= TTM_PL_PRIV5; i++)
-		if (flags & (1 << i)) {
+		if (place->flags & (1 << i)) {
 			*mem_type = i;
 			return 0;
 		}
@@ -89,12 +91,12 @@ static void ttm_bo_mem_space_debug(struct ttm_buffer_object *bo,
 	       bo, bo->mem.num_pages, bo->mem.size >> 10,
 	       bo->mem.size >> 20);
 	for (i = 0; i < placement->num_placement; i++) {
-		ret = ttm_mem_type_from_flags(placement->placement[i],
+		ret = ttm_mem_type_from_place(&placement->placement[i],
 						&mem_type);
 		if (ret)
 			return;
 		pr_err("  placement[%d]=0x%08X (%d)\n",
-		       i, placement->placement[i], mem_type);
+		       i, placement->placement[i].flags, mem_type);
 		ttm_mem_type_debug(bo->bdev, mem_type);
 	}
 }
@@ -141,7 +143,6 @@ static void ttm_bo_release_list(struct kref *list_kref)
 	BUG_ON(atomic_read(&bo->list_kref.refcount));
 	BUG_ON(atomic_read(&bo->kref.refcount));
 	BUG_ON(atomic_read(&bo->cpu_writers));
-	BUG_ON(bo->sync_obj != NULL);
 	BUG_ON(bo->mem.mm_node != NULL);
 	BUG_ON(!list_empty(&bo->lru));
 	BUG_ON(!list_empty(&bo->ddestroy));
@@ -402,36 +403,48 @@ static void ttm_bo_cleanup_memtype_use(struct ttm_buffer_object *bo)
 	ww_mutex_unlock (&bo->resv->lock);
 }
 
+static void ttm_bo_flush_all_fences(struct ttm_buffer_object *bo)
+{
+	struct reservation_object_list *fobj;
+	struct fence *fence;
+	int i;
+
+	fobj = reservation_object_get_list(bo->resv);
+	fence = reservation_object_get_excl(bo->resv);
+	if (fence && !fence->ops->signaled)
+		fence_enable_sw_signaling(fence);
+
+	for (i = 0; fobj && i < fobj->shared_count; ++i) {
+		fence = rcu_dereference_protected(fobj->shared[i],
+					reservation_object_held(bo->resv));
+
+		if (!fence->ops->signaled)
+			fence_enable_sw_signaling(fence);
+	}
+}
+
 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 = bo->glob;
-	struct ttm_bo_driver *driver = bdev->driver;
-	void *sync_obj = NULL;
 	int put_count;
 	int ret;
 
 	spin_lock(&glob->lru_lock);
 	ret = __ttm_bo_reserve(bo, false, true, false, NULL);
 
-	spin_lock(&bdev->fence_lock);
-	(void) ttm_bo_wait(bo, false, false, true);
-	if (!ret && !bo->sync_obj) {
-		spin_unlock(&bdev->fence_lock);
-		put_count = ttm_bo_del_from_lru(bo);
-
-		spin_unlock(&glob->lru_lock);
-		ttm_bo_cleanup_memtype_use(bo);
+	if (!ret) {
+		if (!ttm_bo_wait(bo, false, false, true)) {
+			put_count = ttm_bo_del_from_lru(bo);
 
-		ttm_bo_list_ref_sub(bo, put_count, true);
+			spin_unlock(&glob->lru_lock);
+			ttm_bo_cleanup_memtype_use(bo);
 
-		return;
-	}
-	if (bo->sync_obj)
-		sync_obj = driver->sync_obj_ref(bo->sync_obj);
-	spin_unlock(&bdev->fence_lock);
+			ttm_bo_list_ref_sub(bo, put_count, true);
 
-	if (!ret) {
+			return;
+		} else
+			ttm_bo_flush_all_fences(bo);
 
 		/*
 		 * Make NO_EVICT bos immediately available to
@@ -450,10 +463,6 @@ static void ttm_bo_cleanup_refs_or_queue(struct ttm_buffer_object *bo)
 	list_add_tail(&bo->ddestroy, &bdev->ddestroy);
 	spin_unlock(&glob->lru_lock);
 
-	if (sync_obj) {
-		driver->sync_obj_flush(sync_obj);
-		driver->sync_obj_unref(&sync_obj);
-	}
 	schedule_delayed_work(&bdev->wq,
 			      ((HZ / 100) < 1) ? 1 : HZ / 100);
 }
@@ -474,44 +483,26 @@ static int ttm_bo_cleanup_refs_and_unlock(struct ttm_buffer_object *bo,
 					  bool interruptible,
 					  bool no_wait_gpu)
 {
-	struct ttm_bo_device *bdev = bo->bdev;
-	struct ttm_bo_driver *driver = bdev->driver;
 	struct ttm_bo_global *glob = bo->glob;
 	int put_count;
 	int ret;
 
-	spin_lock(&bdev->fence_lock);
 	ret = ttm_bo_wait(bo, false, false, true);
 
 	if (ret && !no_wait_gpu) {
-		void *sync_obj;
-
-		/*
-		 * Take a reference to the fence and unreserve,
-		 * at this point the buffer should be dead, so
-		 * no new sync objects can be attached.
-		 */
-		sync_obj = driver->sync_obj_ref(bo->sync_obj);
-		spin_unlock(&bdev->fence_lock);
-
-		__ttm_bo_unreserve(bo);
+		long lret;
+		ww_mutex_unlock(&bo->resv->lock);
 		spin_unlock(&glob->lru_lock);
 
-		ret = driver->sync_obj_wait(sync_obj, false, interruptible);
-		driver->sync_obj_unref(&sync_obj);
-		if (ret)
-			return ret;
+		lret = reservation_object_wait_timeout_rcu(bo->resv,
+							   true,
+							   interruptible,
+							   30 * HZ);
 
-		/*
-		 * remove sync_obj with ttm_bo_wait, the wait should be
-		 * finished, and no new wait object should have been added.
-		 */
-		spin_lock(&bdev->fence_lock);
-		ret = ttm_bo_wait(bo, false, false, true);
-		WARN_ON(ret);
-		spin_unlock(&bdev->fence_lock);
-		if (ret)
-			return ret;
+		if (lret < 0)
+			return lret;
+		else if (lret == 0)
+			return -EBUSY;
 
 		spin_lock(&glob->lru_lock);
 		ret = __ttm_bo_reserve(bo, false, true, false, NULL);
@@ -528,8 +519,14 @@ static int ttm_bo_cleanup_refs_and_unlock(struct ttm_buffer_object *bo,
 			spin_unlock(&glob->lru_lock);
 			return 0;
 		}
-	} else
-		spin_unlock(&bdev->fence_lock);
+
+		/*
+		 * remove sync_obj with ttm_bo_wait, the wait should be
+		 * finished, and no new wait object should have been added.
+		 */
+		ret = ttm_bo_wait(bo, false, false, true);
+		WARN_ON(ret);
+	}
 
 	if (ret || unlikely(list_empty(&bo->ddestroy))) {
 		__ttm_bo_unreserve(bo);
@@ -667,9 +664,7 @@ static int ttm_bo_evict(struct ttm_buffer_object *bo, bool interruptible,
 	struct ttm_placement placement;
 	int ret = 0;
 
-	spin_lock(&bdev->fence_lock);
 	ret = ttm_bo_wait(bo, false, interruptible, no_wait_gpu);
-	spin_unlock(&bdev->fence_lock);
 
 	if (unlikely(ret != 0)) {
 		if (ret != -ERESTARTSYS) {
@@ -685,8 +680,6 @@ static int ttm_bo_evict(struct ttm_buffer_object *bo, bool interruptible,
 	evict_mem.bus.io_reserved_vm = false;
 	evict_mem.bus.io_reserved_count = 0;
 
-	placement.fpfn = 0;
-	placement.lpfn = 0;
 	placement.num_placement = 0;
 	placement.num_busy_placement = 0;
 	bdev->driver->evict_flags(bo, &placement);
@@ -774,7 +767,7 @@ EXPORT_SYMBOL(ttm_bo_mem_put);
  */
 static int ttm_bo_mem_force_space(struct ttm_buffer_object *bo,
 					uint32_t mem_type,
-					struct ttm_placement *placement,
+					const struct ttm_place *place,
 					struct ttm_mem_reg *mem,
 					bool interruptible,
 					bool no_wait_gpu)
@@ -784,7 +777,7 @@ static int ttm_bo_mem_force_space(struct ttm_buffer_object *bo,
 	int ret;
 
 	do {
-		ret = (*man->func->get_node)(man, bo, placement, 0, mem);
+		ret = (*man->func->get_node)(man, bo, place, mem);
 		if (unlikely(ret != 0))
 			return ret;
 		if (mem->mm_node)
@@ -827,18 +820,18 @@ static uint32_t ttm_bo_select_caching(struct ttm_mem_type_manager *man,
 
 static bool ttm_bo_mt_compatible(struct ttm_mem_type_manager *man,
 				 uint32_t mem_type,
-				 uint32_t proposed_placement,
+				 const struct ttm_place *place,
 				 uint32_t *masked_placement)
 {
 	uint32_t cur_flags = ttm_bo_type_flags(mem_type);
 
-	if ((cur_flags & proposed_placement & TTM_PL_MASK_MEM) == 0)
+	if ((cur_flags & place->flags & TTM_PL_MASK_MEM) == 0)
 		return false;
 
-	if ((proposed_placement & man->available_caching) == 0)
+	if ((place->flags & man->available_caching) == 0)
 		return false;
 
-	cur_flags |= (proposed_placement & man->available_caching);
+	cur_flags |= (place->flags & man->available_caching);
 
 	*masked_placement = cur_flags;
 	return true;
@@ -869,15 +862,14 @@ int ttm_bo_mem_space(struct ttm_buffer_object *bo,
 
 	mem->mm_node = NULL;
 	for (i = 0; i < placement->num_placement; ++i) {
-		ret = ttm_mem_type_from_flags(placement->placement[i],
-						&mem_type);
+		const struct ttm_place *place = &placement->placement[i];
+
+		ret = ttm_mem_type_from_place(place, &mem_type);
 		if (ret)
 			return ret;
 		man = &bdev->man[mem_type];
 
-		type_ok = ttm_bo_mt_compatible(man,
-						mem_type,
-						placement->placement[i],
+		type_ok = ttm_bo_mt_compatible(man, mem_type, place,
 						&cur_flags);
 
 		if (!type_ok)
@@ -889,7 +881,7 @@ int ttm_bo_mem_space(struct ttm_buffer_object *bo,
 		 * Use the access and other non-mapping-related flag bits from
 		 * the memory placement flags to the current flags
 		 */
-		ttm_flag_masked(&cur_flags, placement->placement[i],
+		ttm_flag_masked(&cur_flags, place->flags,
 				~TTM_PL_MASK_MEMTYPE);
 
 		if (mem_type == TTM_PL_SYSTEM)
@@ -897,8 +889,7 @@ int ttm_bo_mem_space(struct ttm_buffer_object *bo,
 
 		if (man->has_type && man->use_type) {
 			type_found = true;
-			ret = (*man->func->get_node)(man, bo, placement,
-						     cur_flags, mem);
+			ret = (*man->func->get_node)(man, bo, place, mem);
 			if (unlikely(ret))
 				return ret;
 		}
@@ -916,17 +907,15 @@ int ttm_bo_mem_space(struct ttm_buffer_object *bo,
 		return -EINVAL;
 
 	for (i = 0; i < placement->num_busy_placement; ++i) {
-		ret = ttm_mem_type_from_flags(placement->busy_placement[i],
-						&mem_type);
+		const struct ttm_place *place = &placement->busy_placement[i];
+
+		ret = ttm_mem_type_from_place(place, &mem_type);
 		if (ret)
 			return ret;
 		man = &bdev->man[mem_type];
 		if (!man->has_type)
 			continue;
-		if (!ttm_bo_mt_compatible(man,
-						mem_type,
-						placement->busy_placement[i],
-						&cur_flags))
+		if (!ttm_bo_mt_compatible(man, mem_type, place, &cur_flags))
 			continue;
 
 		cur_flags = ttm_bo_select_caching(man, bo->mem.placement,
@@ -935,7 +924,7 @@ int ttm_bo_mem_space(struct ttm_buffer_object *bo,
 		 * Use the access and other non-mapping-related flag bits from
 		 * the memory placement flags to the current flags
 		 */
-		ttm_flag_masked(&cur_flags, placement->busy_placement[i],
+		ttm_flag_masked(&cur_flags, place->flags,
 				~TTM_PL_MASK_MEMTYPE);
 
 		if (mem_type == TTM_PL_SYSTEM) {
@@ -945,7 +934,7 @@ int ttm_bo_mem_space(struct ttm_buffer_object *bo,
 			return 0;
 		}
 
-		ret = ttm_bo_mem_force_space(bo, mem_type, placement, mem,
+		ret = ttm_bo_mem_force_space(bo, mem_type, place, mem,
 						interruptible, no_wait_gpu);
 		if (ret == 0 && mem->mm_node) {
 			mem->placement = cur_flags;
@@ -966,7 +955,6 @@ static int ttm_bo_move_buffer(struct ttm_buffer_object *bo,
 {
 	int ret = 0;
 	struct ttm_mem_reg mem;
-	struct ttm_bo_device *bdev = bo->bdev;
 
 	lockdep_assert_held(&bo->resv->lock.base);
 
@@ -975,9 +963,7 @@ static int ttm_bo_move_buffer(struct ttm_buffer_object *bo,
 	 * Have the driver move function wait for idle when necessary,
 	 * instead of doing it here.
 	 */
-	spin_lock(&bdev->fence_lock);
 	ret = ttm_bo_wait(bo, false, interruptible, no_wait_gpu);
-	spin_unlock(&bdev->fence_lock);
 	if (ret)
 		return ret;
 	mem.num_pages = bo->num_pages;
@@ -1006,20 +992,27 @@ static bool ttm_bo_mem_compat(struct ttm_placement *placement,
 {
 	int i;
 
-	if (mem->mm_node && placement->lpfn != 0 &&
-	    (mem->start < placement->fpfn ||
-	     mem->start + mem->num_pages > placement->lpfn))
-		return false;
-
 	for (i = 0; i < placement->num_placement; i++) {
-		*new_flags = placement->placement[i];
+		const struct ttm_place *heap = &placement->placement[i];
+		if (mem->mm_node && heap->lpfn != 0 &&
+		    (mem->start < heap->fpfn ||
+		     mem->start + mem->num_pages > heap->lpfn))
+			continue;
+
+		*new_flags = heap->flags;
 		if ((*new_flags & mem->placement & TTM_PL_MASK_CACHING) &&
 		    (*new_flags & mem->placement & TTM_PL_MASK_MEM))
 			return true;
 	}
 
 	for (i = 0; i < placement->num_busy_placement; i++) {
-		*new_flags = placement->busy_placement[i];
+		const struct ttm_place *heap = &placement->busy_placement[i];
+		if (mem->mm_node && heap->lpfn != 0 &&
+		    (mem->start < heap->fpfn ||
+		     mem->start + mem->num_pages > heap->lpfn))
+			continue;
+
+		*new_flags = heap->flags;
 		if ((*new_flags & mem->placement & TTM_PL_MASK_CACHING) &&
 		    (*new_flags & mem->placement & TTM_PL_MASK_MEM))
 			return true;
@@ -1037,11 +1030,6 @@ int ttm_bo_validate(struct ttm_buffer_object *bo,
 	uint32_t new_flags;
 
 	lockdep_assert_held(&bo->resv->lock.base);
-	/* Check that range is valid */
-	if (placement->lpfn || placement->fpfn)
-		if (placement->fpfn > placement->lpfn ||
-			(placement->lpfn - placement->fpfn) < bo->num_pages)
-			return -EINVAL;
 	/*
 	 * Check whether we need to move buffer.
 	 */
@@ -1070,15 +1058,6 @@ int ttm_bo_validate(struct ttm_buffer_object *bo,
 }
 EXPORT_SYMBOL(ttm_bo_validate);
 
-int ttm_bo_check_placement(struct ttm_buffer_object *bo,
-				struct ttm_placement *placement)
-{
-	BUG_ON((placement->fpfn || placement->lpfn) &&
-	       (bo->mem.num_pages > (placement->lpfn - placement->fpfn)));
-
-	return 0;
-}
-
 int ttm_bo_init(struct ttm_bo_device *bdev,
 		struct ttm_buffer_object *bo,
 		unsigned long size,
@@ -1089,6 +1068,7 @@ int ttm_bo_init(struct ttm_bo_device *bdev,
 		struct file *persistent_swap_storage,
 		size_t acc_size,
 		struct sg_table *sg,
+		struct reservation_object *resv,
 		void (*destroy) (struct ttm_buffer_object *))
 {
 	int ret = 0;
@@ -1142,30 +1122,38 @@ int ttm_bo_init(struct ttm_bo_device *bdev,
 	bo->persistent_swap_storage = persistent_swap_storage;
 	bo->acc_size = acc_size;
 	bo->sg = sg;
-	bo->resv = &bo->ttm_resv;
-	reservation_object_init(bo->resv);
+	if (resv) {
+		bo->resv = resv;
+		lockdep_assert_held(&bo->resv->lock.base);
+	} else {
+		bo->resv = &bo->ttm_resv;
+		reservation_object_init(&bo->ttm_resv);
+	}
 	atomic_inc(&bo->glob->bo_count);
 	drm_vma_node_reset(&bo->vma_node);
 
-	ret = ttm_bo_check_placement(bo, placement);
-
 	/*
 	 * For ttm_bo_type_device buffers, allocate
 	 * address space from the device.
 	 */
-	if (likely(!ret) &&
-	    (bo->type == ttm_bo_type_device ||
-	     bo->type == ttm_bo_type_sg))
+	if (bo->type == ttm_bo_type_device ||
+	    bo->type == ttm_bo_type_sg)
 		ret = drm_vma_offset_add(&bdev->vma_manager, &bo->vma_node,
 					 bo->mem.num_pages);
 
-	locked = ww_mutex_trylock(&bo->resv->lock);
-	WARN_ON(!locked);
+	/* passed reservation objects should already be locked,
+	 * since otherwise lockdep will be angered in radeon.
+	 */
+	if (!resv) {
+		locked = ww_mutex_trylock(&bo->resv->lock);
+		WARN_ON(!locked);
+	}
 
 	if (likely(!ret))
 		ret = ttm_bo_validate(bo, placement, interruptible, false);
 
-	ttm_bo_unreserve(bo);
+	if (!resv)
+		ttm_bo_unreserve(bo);
 
 	if (unlikely(ret))
 		ttm_bo_unref(&bo);
@@ -1223,7 +1211,7 @@ int ttm_bo_create(struct ttm_bo_device *bdev,
 	acc_size = ttm_bo_acc_size(bdev, size, sizeof(struct ttm_buffer_object));
 	ret = ttm_bo_init(bdev, bo, size, type, placement, page_alignment,
 			  interruptible, persistent_swap_storage, acc_size,
-			  NULL, NULL);
+			  NULL, NULL, NULL);
 	if (likely(ret == 0))
 		*p_bo = bo;
 
@@ -1477,7 +1465,6 @@ int ttm_bo_device_init(struct ttm_bo_device *bdev,
 	bdev->glob = glob;
 	bdev->need_dma32 = need_dma32;
 	bdev->val_seq = 0;
-	spin_lock_init(&bdev->fence_lock);
 	mutex_lock(&glob->device_list_mutex);
 	list_add_tail(&bdev->device_list, &glob->device_list);
 	mutex_unlock(&glob->device_list_mutex);
@@ -1530,65 +1517,56 @@ void ttm_bo_unmap_virtual(struct ttm_buffer_object *bo)
 
 EXPORT_SYMBOL(ttm_bo_unmap_virtual);
 
-
 int ttm_bo_wait(struct ttm_buffer_object *bo,
 		bool lazy, bool interruptible, bool no_wait)
 {
-	struct ttm_bo_driver *driver = bo->bdev->driver;
-	struct ttm_bo_device *bdev = bo->bdev;
-	void *sync_obj;
-	int ret = 0;
-
-	if (likely(bo->sync_obj == NULL))
-		return 0;
+	struct reservation_object_list *fobj;
+	struct reservation_object *resv;
+	struct fence *excl;
+	long timeout = 15 * HZ;
+	int i;
 
-	while (bo->sync_obj) {
+	resv = bo->resv;
+	fobj = reservation_object_get_list(resv);
+	excl = reservation_object_get_excl(resv);
+	if (excl) {
+		if (!fence_is_signaled(excl)) {
+			if (no_wait)
+				return -EBUSY;
 
-		if (driver->sync_obj_signaled(bo->sync_obj)) {
-			void *tmp_obj = bo->sync_obj;
-			bo->sync_obj = NULL;
-			clear_bit(TTM_BO_PRIV_FLAG_MOVING, &bo->priv_flags);
-			spin_unlock(&bdev->fence_lock);
-			driver->sync_obj_unref(&tmp_obj);
-			spin_lock(&bdev->fence_lock);
-			continue;
+			timeout = fence_wait_timeout(excl,
+						     interruptible, timeout);
 		}
+	}
 
-		if (no_wait)
-			return -EBUSY;
+	for (i = 0; fobj && timeout > 0 && i < fobj->shared_count; ++i) {
+		struct fence *fence;
+		fence = rcu_dereference_protected(fobj->shared[i],
+						reservation_object_held(resv));
 
-		sync_obj = driver->sync_obj_ref(bo->sync_obj);
-		spin_unlock(&bdev->fence_lock);
-		ret = driver->sync_obj_wait(sync_obj,
-					    lazy, interruptible);
-		if (unlikely(ret != 0)) {
-			driver->sync_obj_unref(&sync_obj);
-			spin_lock(&bdev->fence_lock);
-			return ret;
-		}
-		spin_lock(&bdev->fence_lock);
-		if (likely(bo->sync_obj == sync_obj)) {
-			void *tmp_obj = bo->sync_obj;
-			bo->sync_obj = NULL;
-			clear_bit(TTM_BO_PRIV_FLAG_MOVING,
-				  &bo->priv_flags);
-			spin_unlock(&bdev->fence_lock);
-			driver->sync_obj_unref(&sync_obj);
-			driver->sync_obj_unref(&tmp_obj);
-			spin_lock(&bdev->fence_lock);
-		} else {
-			spin_unlock(&bdev->fence_lock);
-			driver->sync_obj_unref(&sync_obj);
-			spin_lock(&bdev->fence_lock);
+		if (!fence_is_signaled(fence)) {
+			if (no_wait)
+				return -EBUSY;
+
+			timeout = fence_wait_timeout(fence,
+						     interruptible, timeout);
 		}
 	}
+
+	if (timeout < 0)
+		return timeout;
+
+	if (timeout == 0)
+		return -EBUSY;
+
+	reservation_object_add_excl_fence(resv, NULL);
+	clear_bit(TTM_BO_PRIV_FLAG_MOVING, &bo->priv_flags);
 	return 0;
 }
 EXPORT_SYMBOL(ttm_bo_wait);
 
 int ttm_bo_synccpu_write_grab(struct ttm_buffer_object *bo, bool no_wait)
 {
-	struct ttm_bo_device *bdev = bo->bdev;
 	int ret = 0;
 
 	/*
@@ -1598,9 +1576,7 @@ int ttm_bo_synccpu_write_grab(struct ttm_buffer_object *bo, bool no_wait)
 	ret = ttm_bo_reserve(bo, true, no_wait, false, NULL);
 	if (unlikely(ret != 0))
 		return ret;
-	spin_lock(&bdev->fence_lock);
 	ret = ttm_bo_wait(bo, false, true, no_wait);
-	spin_unlock(&bdev->fence_lock);
 	if (likely(ret == 0))
 		atomic_inc(&bo->cpu_writers);
 	ttm_bo_unreserve(bo);
@@ -1657,9 +1633,7 @@ static int ttm_bo_swapout(struct ttm_mem_shrink *shrink)
 	 * Wait for GPU, then move to system cached.
 	 */
 
-	spin_lock(&bo->bdev->fence_lock);
 	ret = ttm_bo_wait(bo, false, false, false);
-	spin_unlock(&bo->bdev->fence_lock);
 
 	if (unlikely(ret != 0))
 		goto out;
diff --git a/drivers/gpu/drm/ttm/ttm_bo_manager.c b/drivers/gpu/drm/ttm/ttm_bo_manager.c
index 9e103a4875c8..964387fc5c8f 100644
--- a/drivers/gpu/drm/ttm/ttm_bo_manager.c
+++ b/drivers/gpu/drm/ttm/ttm_bo_manager.c
@@ -49,8 +49,7 @@ struct ttm_range_manager {
 
 static int ttm_bo_man_get_node(struct ttm_mem_type_manager *man,
 			       struct ttm_buffer_object *bo,
-			       struct ttm_placement *placement,
-			       uint32_t flags,
+			       const struct ttm_place *place,
 			       struct ttm_mem_reg *mem)
 {
 	struct ttm_range_manager *rman = (struct ttm_range_manager *) man->priv;
@@ -60,7 +59,7 @@ static int ttm_bo_man_get_node(struct ttm_mem_type_manager *man,
 	unsigned long lpfn;
 	int ret;
 
-	lpfn = placement->lpfn;
+	lpfn = place->lpfn;
 	if (!lpfn)
 		lpfn = man->size;
 
@@ -68,13 +67,13 @@ static int ttm_bo_man_get_node(struct ttm_mem_type_manager *man,
 	if (!node)
 		return -ENOMEM;
 
-	if (flags & TTM_PL_FLAG_TOPDOWN)
+	if (place->flags & TTM_PL_FLAG_TOPDOWN)
 		aflags = DRM_MM_CREATE_TOP;
 
 	spin_lock(&rman->lock);
 	ret = drm_mm_insert_node_in_range_generic(mm, node, mem->num_pages,
 					  mem->page_alignment, 0,
-					  placement->fpfn, lpfn,
+					  place->fpfn, lpfn,
 					  DRM_MM_SEARCH_BEST,
 					  aflags);
 	spin_unlock(&rman->lock);
diff --git a/drivers/gpu/drm/ttm/ttm_bo_util.c b/drivers/gpu/drm/ttm/ttm_bo_util.c
index 30e5d90cb7bc..882cccdad272 100644
--- a/drivers/gpu/drm/ttm/ttm_bo_util.c
+++ b/drivers/gpu/drm/ttm/ttm_bo_util.c
@@ -37,6 +37,7 @@
 #include <linux/slab.h>
 #include <linux/vmalloc.h>
 #include <linux/module.h>
+#include <linux/reservation.h>
 
 void ttm_bo_free_old_node(struct ttm_buffer_object *bo)
 {
@@ -444,8 +445,6 @@ static int ttm_buffer_object_transfer(struct ttm_buffer_object *bo,
 				      struct ttm_buffer_object **new_obj)
 {
 	struct ttm_buffer_object *fbo;
-	struct ttm_bo_device *bdev = bo->bdev;
-	struct ttm_bo_driver *driver = bdev->driver;
 	int ret;
 
 	fbo = kmalloc(sizeof(*fbo), GFP_KERNEL);
@@ -466,12 +465,6 @@ static int ttm_buffer_object_transfer(struct ttm_buffer_object *bo,
 	drm_vma_node_reset(&fbo->vma_node);
 	atomic_set(&fbo->cpu_writers, 0);
 
-	spin_lock(&bdev->fence_lock);
-	if (bo->sync_obj)
-		fbo->sync_obj = driver->sync_obj_ref(bo->sync_obj);
-	else
-		fbo->sync_obj = NULL;
-	spin_unlock(&bdev->fence_lock);
 	kref_init(&fbo->list_kref);
 	kref_init(&fbo->kref);
 	fbo->destroy = &ttm_transfered_destroy;
@@ -487,28 +480,24 @@ static int ttm_buffer_object_transfer(struct ttm_buffer_object *bo,
 
 pgprot_t ttm_io_prot(uint32_t caching_flags, pgprot_t tmp)
 {
+	/* Cached mappings need no adjustment */
+	if (caching_flags & TTM_PL_FLAG_CACHED)
+		return tmp;
+
 #if defined(__i386__) || defined(__x86_64__)
 	if (caching_flags & TTM_PL_FLAG_WC)
 		tmp = pgprot_writecombine(tmp);
 	else if (boot_cpu_data.x86 > 3)
 		tmp = pgprot_noncached(tmp);
-
-#elif defined(__powerpc__)
-	if (!(caching_flags & TTM_PL_FLAG_CACHED)) {
-		pgprot_val(tmp) |= _PAGE_NO_CACHE;
-		if (caching_flags & TTM_PL_FLAG_UNCACHED)
-			pgprot_val(tmp) |= _PAGE_GUARDED;
-	}
 #endif
-#if defined(__ia64__) || defined(__arm__)
+#if defined(__ia64__) || defined(__arm__) || defined(__powerpc__)
 	if (caching_flags & TTM_PL_FLAG_WC)
 		tmp = pgprot_writecombine(tmp);
 	else
 		tmp = pgprot_noncached(tmp);
 #endif
 #if defined(__sparc__) || defined(__mips__)
-	if (!(caching_flags & TTM_PL_FLAG_CACHED))
-		tmp = pgprot_noncached(tmp);
+	tmp = pgprot_noncached(tmp);
 #endif
 	return tmp;
 }
@@ -567,9 +556,7 @@ static int ttm_bo_kmap_ttm(struct ttm_buffer_object *bo,
 		 * We need to use vmap to get the desired page protection
 		 * or to make the buffer object look contiguous.
 		 */
-		prot = (mem->placement & TTM_PL_FLAG_CACHED) ?
-			PAGE_KERNEL :
-			ttm_io_prot(mem->placement, PAGE_KERNEL);
+		prot = ttm_io_prot(mem->placement, PAGE_KERNEL);
 		map->bo_kmap_type = ttm_bo_map_vmap;
 		map->virtual = vmap(ttm->pages + start_page, num_pages,
 				    0, prot);
@@ -644,30 +631,20 @@ void ttm_bo_kunmap(struct ttm_bo_kmap_obj *map)
 EXPORT_SYMBOL(ttm_bo_kunmap);
 
 int ttm_bo_move_accel_cleanup(struct ttm_buffer_object *bo,
-			      void *sync_obj,
+			      struct fence *fence,
 			      bool evict,
 			      bool no_wait_gpu,
 			      struct ttm_mem_reg *new_mem)
 {
 	struct ttm_bo_device *bdev = bo->bdev;
-	struct ttm_bo_driver *driver = bdev->driver;
 	struct ttm_mem_type_manager *man = &bdev->man[new_mem->mem_type];
 	struct ttm_mem_reg *old_mem = &bo->mem;
 	int ret;
 	struct ttm_buffer_object *ghost_obj;
-	void *tmp_obj = NULL;
 
-	spin_lock(&bdev->fence_lock);
-	if (bo->sync_obj) {
-		tmp_obj = bo->sync_obj;
-		bo->sync_obj = NULL;
-	}
-	bo->sync_obj = driver->sync_obj_ref(sync_obj);
+	reservation_object_add_excl_fence(bo->resv, fence);
 	if (evict) {
 		ret = ttm_bo_wait(bo, false, false, false);
-		spin_unlock(&bdev->fence_lock);
-		if (tmp_obj)
-			driver->sync_obj_unref(&tmp_obj);
 		if (ret)
 			return ret;
 
@@ -688,14 +665,13 @@ int ttm_bo_move_accel_cleanup(struct ttm_buffer_object *bo,
 		 */
 
 		set_bit(TTM_BO_PRIV_FLAG_MOVING, &bo->priv_flags);
-		spin_unlock(&bdev->fence_lock);
-		if (tmp_obj)
-			driver->sync_obj_unref(&tmp_obj);
 
 		ret = ttm_buffer_object_transfer(bo, &ghost_obj);
 		if (ret)
 			return ret;
 
+		reservation_object_add_excl_fence(ghost_obj->resv, fence);
+
 		/**
 		 * If we're not moving to fixed memory, the TTM object
 		 * needs to stay alive. Otherwhise hang it on the ghost
diff --git a/drivers/gpu/drm/ttm/ttm_bo_vm.c b/drivers/gpu/drm/ttm/ttm_bo_vm.c
index 0ce48e5a9cb4..8fb7213277cc 100644
--- a/drivers/gpu/drm/ttm/ttm_bo_vm.c
+++ b/drivers/gpu/drm/ttm/ttm_bo_vm.c
@@ -45,10 +45,8 @@ static int ttm_bo_vm_fault_idle(struct ttm_buffer_object *bo,
 				struct vm_area_struct *vma,
 				struct vm_fault *vmf)
 {
-	struct ttm_bo_device *bdev = bo->bdev;
 	int ret = 0;
 
-	spin_lock(&bdev->fence_lock);
 	if (likely(!test_bit(TTM_BO_PRIV_FLAG_MOVING, &bo->priv_flags)))
 		goto out_unlock;
 
@@ -82,7 +80,6 @@ static int ttm_bo_vm_fault_idle(struct ttm_buffer_object *bo,
 			VM_FAULT_NOPAGE;
 
 out_unlock:
-	spin_unlock(&bdev->fence_lock);
 	return ret;
 }
 
@@ -200,9 +197,8 @@ static int ttm_bo_vm_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 						cvma.vm_page_prot);
 	} else {
 		ttm = bo->ttm;
-		if (!(bo->mem.placement & TTM_PL_FLAG_CACHED))
-			cvma.vm_page_prot = ttm_io_prot(bo->mem.placement,
-							cvma.vm_page_prot);
+		cvma.vm_page_prot = ttm_io_prot(bo->mem.placement,
+						cvma.vm_page_prot);
 
 		/* Allocate all page at once, most common usage */
 		if (ttm->bdev->driver->ttm_tt_populate(ttm)) {
diff --git a/drivers/gpu/drm/ttm/ttm_execbuf_util.c b/drivers/gpu/drm/ttm/ttm_execbuf_util.c
index e8dac8758528..8ce508e76208 100644
--- a/drivers/gpu/drm/ttm/ttm_execbuf_util.c
+++ b/drivers/gpu/drm/ttm/ttm_execbuf_util.c
@@ -32,20 +32,12 @@
 #include <linux/sched.h>
 #include <linux/module.h>
 
-static void ttm_eu_backoff_reservation_locked(struct list_head *list)
+static void ttm_eu_backoff_reservation_reverse(struct list_head *list,
+					      struct ttm_validate_buffer *entry)
 {
-	struct ttm_validate_buffer *entry;
-
-	list_for_each_entry(entry, list, head) {
+	list_for_each_entry_continue_reverse(entry, list, head) {
 		struct ttm_buffer_object *bo = entry->bo;
-		if (!entry->reserved)
-			continue;
 
-		entry->reserved = false;
-		if (entry->removed) {
-			ttm_bo_add_to_lru(bo);
-			entry->removed = false;
-		}
 		__ttm_bo_unreserve(bo);
 	}
 }
@@ -56,27 +48,9 @@ static void ttm_eu_del_from_lru_locked(struct list_head *list)
 
 	list_for_each_entry(entry, list, head) {
 		struct ttm_buffer_object *bo = entry->bo;
-		if (!entry->reserved)
-			continue;
-
-		if (!entry->removed) {
-			entry->put_count = ttm_bo_del_from_lru(bo);
-			entry->removed = true;
-		}
-	}
-}
-
-static void ttm_eu_list_ref_sub(struct list_head *list)
-{
-	struct ttm_validate_buffer *entry;
-
-	list_for_each_entry(entry, list, head) {
-		struct ttm_buffer_object *bo = entry->bo;
+		unsigned put_count = ttm_bo_del_from_lru(bo);
 
-		if (entry->put_count) {
-			ttm_bo_list_ref_sub(bo, entry->put_count, true);
-			entry->put_count = 0;
-		}
+		ttm_bo_list_ref_sub(bo, put_count, true);
 	}
 }
 
@@ -91,11 +65,18 @@ void ttm_eu_backoff_reservation(struct ww_acquire_ctx *ticket,
 
 	entry = list_first_entry(list, struct ttm_validate_buffer, head);
 	glob = entry->bo->glob;
+
 	spin_lock(&glob->lru_lock);
-	ttm_eu_backoff_reservation_locked(list);
+	list_for_each_entry(entry, list, head) {
+		struct ttm_buffer_object *bo = entry->bo;
+
+		ttm_bo_add_to_lru(bo);
+		__ttm_bo_unreserve(bo);
+	}
+	spin_unlock(&glob->lru_lock);
+
 	if (ticket)
 		ww_acquire_fini(ticket);
-	spin_unlock(&glob->lru_lock);
 }
 EXPORT_SYMBOL(ttm_eu_backoff_reservation);
 
@@ -112,7 +93,7 @@ EXPORT_SYMBOL(ttm_eu_backoff_reservation);
  */
 
 int ttm_eu_reserve_buffers(struct ww_acquire_ctx *ticket,
-			   struct list_head *list)
+			   struct list_head *list, bool intr)
 {
 	struct ttm_bo_global *glob;
 	struct ttm_validate_buffer *entry;
@@ -121,60 +102,64 @@ int ttm_eu_reserve_buffers(struct ww_acquire_ctx *ticket,
 	if (list_empty(list))
 		return 0;
 
-	list_for_each_entry(entry, list, head) {
-		entry->reserved = false;
-		entry->put_count = 0;
-		entry->removed = false;
-	}
-
 	entry = list_first_entry(list, struct ttm_validate_buffer, head);
 	glob = entry->bo->glob;
 
 	if (ticket)
 		ww_acquire_init(ticket, &reservation_ww_class);
-retry:
+
 	list_for_each_entry(entry, list, head) {
 		struct ttm_buffer_object *bo = entry->bo;
 
-		/* already slowpath reserved? */
-		if (entry->reserved)
-			continue;
-
-		ret = __ttm_bo_reserve(bo, true, (ticket == NULL), true,
+		ret = __ttm_bo_reserve(bo, intr, (ticket == NULL), true,
 				       ticket);
+		if (!ret && unlikely(atomic_read(&bo->cpu_writers) > 0)) {
+			__ttm_bo_unreserve(bo);
+
+			ret = -EBUSY;
+		}
 
-		if (ret == -EDEADLK) {
-			/* uh oh, we lost out, drop every reservation and try
-			 * to only reserve this buffer, then start over if
-			 * this succeeds.
-			 */
-			BUG_ON(ticket == NULL);
-			spin_lock(&glob->lru_lock);
-			ttm_eu_backoff_reservation_locked(list);
-			spin_unlock(&glob->lru_lock);
-			ttm_eu_list_ref_sub(list);
+		if (!ret) {
+			if (!entry->shared)
+				continue;
+
+			ret = reservation_object_reserve_shared(bo->resv);
+			if (!ret)
+				continue;
+		}
+
+		/* uh oh, we lost out, drop every reservation and try
+		 * to only reserve this buffer, then start over if
+		 * this succeeds.
+		 */
+		ttm_eu_backoff_reservation_reverse(list, entry);
+
+		if (ret == -EDEADLK && intr) {
 			ret = ww_mutex_lock_slow_interruptible(&bo->resv->lock,
 							       ticket);
-			if (unlikely(ret != 0)) {
-				if (ret == -EINTR)
-					ret = -ERESTARTSYS;
-				goto err_fini;
-			}
+		} else if (ret == -EDEADLK) {
+			ww_mutex_lock_slow(&bo->resv->lock, ticket);
+			ret = 0;
+		}
 
-			entry->reserved = true;
-			if (unlikely(atomic_read(&bo->cpu_writers) > 0)) {
-				ret = -EBUSY;
-				goto err;
-			}
-			goto retry;
-		} else if (ret)
-			goto err;
+		if (!ret && entry->shared)
+			ret = reservation_object_reserve_shared(bo->resv);
 
-		entry->reserved = true;
-		if (unlikely(atomic_read(&bo->cpu_writers) > 0)) {
-			ret = -EBUSY;
-			goto err;
+		if (unlikely(ret != 0)) {
+			if (ret == -EINTR)
+				ret = -ERESTARTSYS;
+			if (ticket) {
+				ww_acquire_done(ticket);
+				ww_acquire_fini(ticket);
+			}
+			return ret;
 		}
+
+		/* move this item to the front of the list,
+		 * forces correct iteration of the loop without keeping track
+		 */
+		list_del(&entry->head);
+		list_add(&entry->head, list);
 	}
 
 	if (ticket)
@@ -182,25 +167,12 @@ retry:
 	spin_lock(&glob->lru_lock);
 	ttm_eu_del_from_lru_locked(list);
 	spin_unlock(&glob->lru_lock);
-	ttm_eu_list_ref_sub(list);
 	return 0;
-
-err:
-	spin_lock(&glob->lru_lock);
-	ttm_eu_backoff_reservation_locked(list);
-	spin_unlock(&glob->lru_lock);
-	ttm_eu_list_ref_sub(list);
-err_fini:
-	if (ticket) {
-		ww_acquire_done(ticket);
-		ww_acquire_fini(ticket);
-	}
-	return ret;
 }
 EXPORT_SYMBOL(ttm_eu_reserve_buffers);
 
 void ttm_eu_fence_buffer_objects(struct ww_acquire_ctx *ticket,
-				 struct list_head *list, void *sync_obj)
+				 struct list_head *list, struct fence *fence)
 {
 	struct ttm_validate_buffer *entry;
 	struct ttm_buffer_object *bo;
@@ -217,24 +189,18 @@ void ttm_eu_fence_buffer_objects(struct ww_acquire_ctx *ticket,
 	glob = bo->glob;
 
 	spin_lock(&glob->lru_lock);
-	spin_lock(&bdev->fence_lock);
 
 	list_for_each_entry(entry, list, head) {
 		bo = entry->bo;
-		entry->old_sync_obj = bo->sync_obj;
-		bo->sync_obj = driver->sync_obj_ref(sync_obj);
+		if (entry->shared)
+			reservation_object_add_shared_fence(bo->resv, fence);
+		else
+			reservation_object_add_excl_fence(bo->resv, fence);
 		ttm_bo_add_to_lru(bo);
 		__ttm_bo_unreserve(bo);
-		entry->reserved = false;
 	}
-	spin_unlock(&bdev->fence_lock);
 	spin_unlock(&glob->lru_lock);
 	if (ticket)
 		ww_acquire_fini(ticket);
-
-	list_for_each_entry(entry, list, head) {
-		if (entry->old_sync_obj)
-			driver->sync_obj_unref(&entry->old_sync_obj);
-	}
 }
 EXPORT_SYMBOL(ttm_eu_fence_buffer_objects);
diff --git a/drivers/gpu/drm/ttm/ttm_memory.c b/drivers/gpu/drm/ttm/ttm_memory.c
index dbc2def887cd..a1803fbcc898 100644
--- a/drivers/gpu/drm/ttm/ttm_memory.c
+++ b/drivers/gpu/drm/ttm/ttm_memory.c
@@ -300,7 +300,8 @@ static int ttm_mem_init_highmem_zone(struct ttm_mem_global *glob,
 	zone->glob = glob;
 	glob->zone_highmem = zone;
 	ret = kobject_init_and_add(
-		&zone->kobj, &ttm_mem_zone_kobj_type, &glob->kobj, zone->name);
+		&zone->kobj, &ttm_mem_zone_kobj_type, &glob->kobj, "%s",
+		zone->name);
 	if (unlikely(ret != 0)) {
 		kobject_put(&zone->kobj);
 		return ret;
diff --git a/drivers/gpu/drm/udl/Kconfig b/drivers/gpu/drm/udl/Kconfig
index f02528686cd5..613ab0622d6e 100644
--- a/drivers/gpu/drm/udl/Kconfig
+++ b/drivers/gpu/drm/udl/Kconfig
@@ -1,8 +1,9 @@
 config DRM_UDL
 	tristate "DisplayLink"
 	depends on DRM
+	depends on USB_SUPPORT
 	depends on USB_ARCH_HAS_HCD
-	select DRM_USB
+	select USB
 	select FB_SYS_FILLRECT
 	select FB_SYS_COPYAREA
 	select FB_SYS_IMAGEBLIT
diff --git a/drivers/gpu/drm/udl/udl_connector.c b/drivers/gpu/drm/udl/udl_connector.c
index e026a9e2942a..0110d95522f3 100644
--- a/drivers/gpu/drm/udl/udl_connector.c
+++ b/drivers/gpu/drm/udl/udl_connector.c
@@ -34,8 +34,8 @@ static u8 *udl_get_edid(struct udl_device *udl)
 		goto error;
 
 	for (i = 0; i < EDID_LENGTH; i++) {
-		ret = usb_control_msg(udl->ddev->usbdev,
-				      usb_rcvctrlpipe(udl->ddev->usbdev, 0), (0x02),
+		ret = usb_control_msg(udl->udev,
+				      usb_rcvctrlpipe(udl->udev, 0), (0x02),
 				      (0x80 | (0x02 << 5)), i << 8, 0xA1, rbuf, 2,
 				      HZ);
 		if (ret < 1) {
diff --git a/drivers/gpu/drm/udl/udl_drv.c b/drivers/gpu/drm/udl/udl_drv.c
index 3ddd6cd98ac1..8607e9e513db 100644
--- a/drivers/gpu/drm/udl/udl_drv.c
+++ b/drivers/gpu/drm/udl/udl_drv.c
@@ -7,48 +7,13 @@
  */
 
 #include <linux/module.h>
-#include <drm/drm_usb.h>
+#include <drm/drmP.h>
 #include <drm/drm_crtc_helper.h>
 #include "udl_drv.h"
 
-static struct drm_driver driver;
-
-/*
- * There are many DisplayLink-based graphics products, all with unique PIDs.
- * So we match on DisplayLink's VID + Vendor-Defined Interface Class (0xff)
- * We also require a match on SubClass (0x00) and Protocol (0x00),
- * which is compatible with all known USB 2.0 era graphics chips and firmware,
- * but allows DisplayLink to increment those for any future incompatible chips
- */
-static struct usb_device_id id_table[] = {
-	{.idVendor = 0x17e9, .bInterfaceClass = 0xff,
-	 .bInterfaceSubClass = 0x00,
-	 .bInterfaceProtocol = 0x00,
-	 .match_flags = USB_DEVICE_ID_MATCH_VENDOR |
-			USB_DEVICE_ID_MATCH_INT_CLASS |
-			USB_DEVICE_ID_MATCH_INT_SUBCLASS |
-			USB_DEVICE_ID_MATCH_INT_PROTOCOL,},
-	{},
-};
-MODULE_DEVICE_TABLE(usb, id_table);
-
-MODULE_LICENSE("GPL");
-
-static int udl_usb_probe(struct usb_interface *interface,
-			 const struct usb_device_id *id)
+static int udl_driver_set_busid(struct drm_device *d, struct drm_master *m)
 {
-	return drm_get_usb_dev(interface, id, &driver);
-}
-
-static void udl_usb_disconnect(struct usb_interface *interface)
-{
-	struct drm_device *dev = usb_get_intfdata(interface);
-
-	drm_kms_helper_poll_disable(dev);
-	drm_connector_unplug_all(dev);
-	udl_fbdev_unplug(dev);
-	udl_drop_usb(dev);
-	drm_unplug_dev(dev);
+	return 0;
 }
 
 static const struct vm_operations_struct udl_gem_vm_ops = {
@@ -75,6 +40,7 @@ static struct drm_driver driver = {
 	.driver_features = DRIVER_MODESET | DRIVER_GEM | DRIVER_PRIME,
 	.load = udl_driver_load,
 	.unload = udl_driver_unload,
+	.set_busid = udl_driver_set_busid,
 
 	/* gem hooks */
 	.gem_free_object = udl_gem_free_object,
@@ -96,6 +62,61 @@ static struct drm_driver driver = {
 	.patchlevel = DRIVER_PATCHLEVEL,
 };
 
+static int udl_usb_probe(struct usb_interface *interface,
+			 const struct usb_device_id *id)
+{
+	struct usb_device *udev = interface_to_usbdev(interface);
+	struct drm_device *dev;
+	int r;
+
+	dev = drm_dev_alloc(&driver, &interface->dev);
+	if (!dev)
+		return -ENOMEM;
+
+	r = drm_dev_register(dev, (unsigned long)udev);
+	if (r)
+		goto err_free;
+
+	usb_set_intfdata(interface, dev);
+	DRM_INFO("Initialized udl on minor %d\n", dev->primary->index);
+
+	return 0;
+
+err_free:
+	drm_dev_unref(dev);
+	return r;
+}
+
+static void udl_usb_disconnect(struct usb_interface *interface)
+{
+	struct drm_device *dev = usb_get_intfdata(interface);
+
+	drm_kms_helper_poll_disable(dev);
+	drm_connector_unplug_all(dev);
+	udl_fbdev_unplug(dev);
+	udl_drop_usb(dev);
+	drm_unplug_dev(dev);
+}
+
+/*
+ * There are many DisplayLink-based graphics products, all with unique PIDs.
+ * So we match on DisplayLink's VID + Vendor-Defined Interface Class (0xff)
+ * We also require a match on SubClass (0x00) and Protocol (0x00),
+ * which is compatible with all known USB 2.0 era graphics chips and firmware,
+ * but allows DisplayLink to increment those for any future incompatible chips
+ */
+static struct usb_device_id id_table[] = {
+	{.idVendor = 0x17e9, .bInterfaceClass = 0xff,
+	 .bInterfaceSubClass = 0x00,
+	 .bInterfaceProtocol = 0x00,
+	 .match_flags = USB_DEVICE_ID_MATCH_VENDOR |
+			USB_DEVICE_ID_MATCH_INT_CLASS |
+			USB_DEVICE_ID_MATCH_INT_SUBCLASS |
+			USB_DEVICE_ID_MATCH_INT_PROTOCOL,},
+	{},
+};
+MODULE_DEVICE_TABLE(usb, id_table);
+
 static struct usb_driver udl_driver = {
 	.name = "udl",
 	.probe = udl_usb_probe,
@@ -105,13 +126,14 @@ static struct usb_driver udl_driver = {
 
 static int __init udl_init(void)
 {
-	return drm_usb_init(&driver, &udl_driver);
+	return usb_register(&udl_driver);
 }
 
 static void __exit udl_exit(void)
 {
-	drm_usb_exit(&driver, &udl_driver);
+	usb_deregister(&udl_driver);
 }
 
 module_init(udl_init);
 module_exit(udl_exit);
+MODULE_LICENSE("GPL");
diff --git a/drivers/gpu/drm/udl/udl_drv.h b/drivers/gpu/drm/udl/udl_drv.h
index 1fbf7b357f16..c7490a2489a7 100644
--- a/drivers/gpu/drm/udl/udl_drv.h
+++ b/drivers/gpu/drm/udl/udl_drv.h
@@ -15,6 +15,7 @@
 #define UDL_DRV_H
 
 #include <linux/usb.h>
+#include <drm/drm_gem.h>
 
 #define DRIVER_NAME		"udl"
 #define DRIVER_DESC		"DisplayLink"
@@ -47,6 +48,7 @@ struct udl_fbdev;
 struct udl_device {
 	struct device *dev;
 	struct drm_device *ddev;
+	struct usb_device *udev;
 
 	int sku_pixel_limit;
 
diff --git a/drivers/gpu/drm/udl/udl_fb.c b/drivers/gpu/drm/udl/udl_fb.c
index d1da339843ca..8cbcb4589bd3 100644
--- a/drivers/gpu/drm/udl/udl_fb.c
+++ b/drivers/gpu/drm/udl/udl_fb.c
@@ -472,7 +472,8 @@ udl_framebuffer_init(struct drm_device *dev,
 static int udlfb_create(struct drm_fb_helper *helper,
 			struct drm_fb_helper_surface_size *sizes)
 {
-	struct udl_fbdev *ufbdev = (struct udl_fbdev *)helper;
+	struct udl_fbdev *ufbdev =
+		container_of(helper, struct udl_fbdev, helper);
 	struct drm_device *dev = ufbdev->helper.dev;
 	struct fb_info *info;
 	struct device *device = dev->dev;
diff --git a/drivers/gpu/drm/udl/udl_main.c b/drivers/gpu/drm/udl/udl_main.c
index 42795674bc07..33dbfb2c4748 100644
--- a/drivers/gpu/drm/udl/udl_main.c
+++ b/drivers/gpu/drm/udl/udl_main.c
@@ -202,7 +202,7 @@ static int udl_alloc_urb_list(struct drm_device *dev, int count, size_t size)
 		}
 		unode->urb = urb;
 
-		buf = usb_alloc_coherent(udl->ddev->usbdev, MAX_TRANSFER, GFP_KERNEL,
+		buf = usb_alloc_coherent(udl->udev, MAX_TRANSFER, GFP_KERNEL,
 					 &urb->transfer_dma);
 		if (!buf) {
 			kfree(unode);
@@ -211,7 +211,7 @@ static int udl_alloc_urb_list(struct drm_device *dev, int count, size_t size)
 		}
 
 		/* urb->transfer_buffer_length set to actual before submit */
-		usb_fill_bulk_urb(urb, udl->ddev->usbdev, usb_sndbulkpipe(udl->ddev->usbdev, 1),
+		usb_fill_bulk_urb(urb, udl->udev, usb_sndbulkpipe(udl->udev, 1),
 			buf, size, udl_urb_completion, unode);
 		urb->transfer_flags |= URB_NO_TRANSFER_DMA_MAP;
 
@@ -282,6 +282,7 @@ int udl_submit_urb(struct drm_device *dev, struct urb *urb, size_t len)
 
 int udl_driver_load(struct drm_device *dev, unsigned long flags)
 {
+	struct usb_device *udev = (void*)flags;
 	struct udl_device *udl;
 	int ret = -ENOMEM;
 
@@ -290,10 +291,11 @@ int udl_driver_load(struct drm_device *dev, unsigned long flags)
 	if (!udl)
 		return -ENOMEM;
 
+	udl->udev = udev;
 	udl->ddev = dev;
 	dev->dev_private = udl;
 
-	if (!udl_parse_vendor_descriptor(dev, dev->usbdev)) {
+	if (!udl_parse_vendor_descriptor(dev, udl->udev)) {
 		ret = -ENODEV;
 		DRM_ERROR("firmware not recognized. Assume incompatible device\n");
 		goto err;
diff --git a/drivers/gpu/drm/via/via_dma.c b/drivers/gpu/drm/via/via_dma.c
index 6fc0648dd37f..d17d8f245c1a 100644
--- a/drivers/gpu/drm/via/via_dma.c
+++ b/drivers/gpu/drm/via/via_dma.c
@@ -161,7 +161,7 @@ int via_dma_cleanup(struct drm_device *dev)
 		if (dev_priv->ring.virtual_start) {
 			via_cmdbuf_reset(dev_priv);
 
-			drm_core_ioremapfree(&dev_priv->ring.map, dev);
+			drm_legacy_ioremapfree(&dev_priv->ring.map, dev);
 			dev_priv->ring.virtual_start = NULL;
 		}
 
@@ -200,7 +200,7 @@ static int via_initialize(struct drm_device *dev,
 	dev_priv->ring.map.flags = 0;
 	dev_priv->ring.map.mtrr = 0;
 
-	drm_core_ioremap(&dev_priv->ring.map, dev);
+	drm_legacy_ioremap(&dev_priv->ring.map, dev);
 
 	if (dev_priv->ring.map.handle == NULL) {
 		via_dma_cleanup(dev);
diff --git a/drivers/gpu/drm/via/via_drv.c b/drivers/gpu/drm/via/via_drv.c
index 50abc2adfaee..ed8aa8ff861a 100644
--- a/drivers/gpu/drm/via/via_drv.c
+++ b/drivers/gpu/drm/via/via_drv.c
@@ -62,7 +62,7 @@ static const struct file_operations via_driver_fops = {
 	.open = drm_open,
 	.release = drm_release,
 	.unlocked_ioctl = drm_ioctl,
-	.mmap = drm_mmap,
+	.mmap = drm_legacy_mmap,
 	.poll = drm_poll,
 #ifdef CONFIG_COMPAT
 	.compat_ioctl = drm_compat_ioctl,
@@ -79,6 +79,7 @@ static struct drm_driver driver = {
 	.open = via_driver_open,
 	.preclose = via_reclaim_buffers_locked,
 	.postclose = via_driver_postclose,
+	.set_busid = drm_pci_set_busid,
 	.context_dtor = via_final_context,
 	.get_vblank_counter = via_get_vblank_counter,
 	.enable_vblank = via_enable_vblank,
diff --git a/drivers/gpu/drm/via/via_drv.h b/drivers/gpu/drm/via/via_drv.h
index ad0273256beb..ef8c500b4a00 100644
--- a/drivers/gpu/drm/via/via_drv.h
+++ b/drivers/gpu/drm/via/via_drv.h
@@ -25,6 +25,8 @@
 #define _VIA_DRV_H_
 
 #include <drm/drm_mm.h>
+#include <drm/drm_legacy.h>
+
 #define DRIVER_AUTHOR	"Various"
 
 #define DRIVER_NAME		"via"
diff --git a/drivers/gpu/drm/via/via_map.c b/drivers/gpu/drm/via/via_map.c
index d0ab3fb32acd..0b3522dba6e8 100644
--- a/drivers/gpu/drm/via/via_map.c
+++ b/drivers/gpu/drm/via/via_map.c
@@ -31,7 +31,7 @@ static int via_do_init_map(struct drm_device *dev, drm_via_init_t *init)
 
 	DRM_DEBUG("\n");
 
-	dev_priv->sarea = drm_getsarea(dev);
+	dev_priv->sarea = drm_legacy_getsarea(dev);
 	if (!dev_priv->sarea) {
 		DRM_ERROR("could not find sarea!\n");
 		dev->dev_private = (void *)dev_priv;
@@ -39,14 +39,14 @@ static int via_do_init_map(struct drm_device *dev, drm_via_init_t *init)
 		return -EINVAL;
 	}
 
-	dev_priv->fb = drm_core_findmap(dev, init->fb_offset);
+	dev_priv->fb = drm_legacy_findmap(dev, init->fb_offset);
 	if (!dev_priv->fb) {
 		DRM_ERROR("could not find framebuffer!\n");
 		dev->dev_private = (void *)dev_priv;
 		via_do_cleanup_map(dev);
 		return -EINVAL;
 	}
-	dev_priv->mmio = drm_core_findmap(dev, init->mmio_offset);
+	dev_priv->mmio = drm_legacy_findmap(dev, init->mmio_offset);
 	if (!dev_priv->mmio) {
 		DRM_ERROR("could not find mmio region!\n");
 		dev->dev_private = (void *)dev_priv;
diff --git a/drivers/gpu/drm/via/via_mm.c b/drivers/gpu/drm/via/via_mm.c
index d70b1e1544bf..4f20742e7788 100644
--- a/drivers/gpu/drm/via/via_mm.c
+++ b/drivers/gpu/drm/via/via_mm.c
@@ -211,12 +211,12 @@ void via_reclaim_buffers_locked(struct drm_device *dev,
 	if (!(file->minor->master && file->master->lock.hw_lock))
 		return;
 
-	drm_idlelock_take(&file->master->lock);
+	drm_legacy_idlelock_take(&file->master->lock);
 
 	mutex_lock(&dev->struct_mutex);
 	if (list_empty(&file_priv->obj_list)) {
 		mutex_unlock(&dev->struct_mutex);
-		drm_idlelock_release(&file->master->lock);
+		drm_legacy_idlelock_release(&file->master->lock);
 
 		return;
 	}
@@ -231,7 +231,7 @@ void via_reclaim_buffers_locked(struct drm_device *dev,
 	}
 	mutex_unlock(&dev->struct_mutex);
 
-	drm_idlelock_release(&file->master->lock);
+	drm_legacy_idlelock_release(&file->master->lock);
 
 	return;
 }
diff --git a/drivers/gpu/drm/via/via_verifier.c b/drivers/gpu/drm/via/via_verifier.c
index 9dbc92bd1512..0677bbf4ec7e 100644
--- a/drivers/gpu/drm/via/via_verifier.c
+++ b/drivers/gpu/drm/via/via_verifier.c
@@ -31,6 +31,7 @@
 #include "via_3d_reg.h"
 #include <drm/drmP.h>
 #include <drm/via_drm.h>
+#include <drm/drm_legacy.h>
 #include "via_verifier.h"
 #include "via_drv.h"
 
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_buffer.c b/drivers/gpu/drm/vmwgfx/vmwgfx_buffer.c
index 6327cfc36805..cff2bf9db9d2 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_buffer.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_buffer.c
@@ -30,66 +30,101 @@
 #include <drm/ttm/ttm_placement.h>
 #include <drm/ttm/ttm_page_alloc.h>
 
-static uint32_t vram_placement_flags = TTM_PL_FLAG_VRAM |
-	TTM_PL_FLAG_CACHED;
-
-static uint32_t vram_ne_placement_flags = TTM_PL_FLAG_VRAM |
-	TTM_PL_FLAG_CACHED |
-	TTM_PL_FLAG_NO_EVICT;
+static struct ttm_place vram_placement_flags = {
+	.fpfn = 0,
+	.lpfn = 0,
+	.flags = TTM_PL_FLAG_VRAM | TTM_PL_FLAG_CACHED
+};
 
-static uint32_t sys_placement_flags = TTM_PL_FLAG_SYSTEM |
-	TTM_PL_FLAG_CACHED;
+static struct ttm_place vram_ne_placement_flags = {
+	.fpfn = 0,
+	.lpfn = 0,
+	.flags = TTM_PL_FLAG_VRAM | TTM_PL_FLAG_CACHED | TTM_PL_FLAG_NO_EVICT
+};
 
-static uint32_t sys_ne_placement_flags = TTM_PL_FLAG_SYSTEM |
-	TTM_PL_FLAG_CACHED |
-	TTM_PL_FLAG_NO_EVICT;
+static struct ttm_place sys_placement_flags = {
+	.fpfn = 0,
+	.lpfn = 0,
+	.flags = TTM_PL_FLAG_SYSTEM | TTM_PL_FLAG_CACHED
+};
 
-static uint32_t gmr_placement_flags = VMW_PL_FLAG_GMR |
-	TTM_PL_FLAG_CACHED;
+static struct ttm_place sys_ne_placement_flags = {
+	.fpfn = 0,
+	.lpfn = 0,
+	.flags = TTM_PL_FLAG_SYSTEM | TTM_PL_FLAG_CACHED | TTM_PL_FLAG_NO_EVICT
+};
 
-static uint32_t gmr_ne_placement_flags = VMW_PL_FLAG_GMR |
-	TTM_PL_FLAG_CACHED |
-	TTM_PL_FLAG_NO_EVICT;
+static struct ttm_place gmr_placement_flags = {
+	.fpfn = 0,
+	.lpfn = 0,
+	.flags = VMW_PL_FLAG_GMR | TTM_PL_FLAG_CACHED
+};
 
-static uint32_t mob_placement_flags = VMW_PL_FLAG_MOB |
-	TTM_PL_FLAG_CACHED;
+static struct ttm_place gmr_ne_placement_flags = {
+	.fpfn = 0,
+	.lpfn = 0,
+	.flags = VMW_PL_FLAG_GMR | TTM_PL_FLAG_CACHED | TTM_PL_FLAG_NO_EVICT
+};
 
-struct ttm_placement vmw_vram_placement = {
+static struct ttm_place mob_placement_flags = {
 	.fpfn = 0,
 	.lpfn = 0,
+	.flags = VMW_PL_FLAG_MOB | TTM_PL_FLAG_CACHED
+};
+
+struct ttm_placement vmw_vram_placement = {
 	.num_placement = 1,
 	.placement = &vram_placement_flags,
 	.num_busy_placement = 1,
 	.busy_placement = &vram_placement_flags
 };
 
-static uint32_t vram_gmr_placement_flags[] = {
-	TTM_PL_FLAG_VRAM | TTM_PL_FLAG_CACHED,
-	VMW_PL_FLAG_GMR | TTM_PL_FLAG_CACHED
+static struct ttm_place vram_gmr_placement_flags[] = {
+	{
+		.fpfn = 0,
+		.lpfn = 0,
+		.flags = TTM_PL_FLAG_VRAM | TTM_PL_FLAG_CACHED
+	}, {
+		.fpfn = 0,
+		.lpfn = 0,
+		.flags = VMW_PL_FLAG_GMR | TTM_PL_FLAG_CACHED
+	}
 };
 
-static uint32_t gmr_vram_placement_flags[] = {
-	VMW_PL_FLAG_GMR | TTM_PL_FLAG_CACHED,
-	TTM_PL_FLAG_VRAM | TTM_PL_FLAG_CACHED
+static struct ttm_place gmr_vram_placement_flags[] = {
+	{
+		.fpfn = 0,
+		.lpfn = 0,
+		.flags = VMW_PL_FLAG_GMR | TTM_PL_FLAG_CACHED
+	}, {
+		.fpfn = 0,
+		.lpfn = 0,
+		.flags = TTM_PL_FLAG_VRAM | TTM_PL_FLAG_CACHED
+	}
 };
 
 struct ttm_placement vmw_vram_gmr_placement = {
-	.fpfn = 0,
-	.lpfn = 0,
 	.num_placement = 2,
 	.placement = vram_gmr_placement_flags,
 	.num_busy_placement = 1,
 	.busy_placement = &gmr_placement_flags
 };
 
-static uint32_t vram_gmr_ne_placement_flags[] = {
-	TTM_PL_FLAG_VRAM | TTM_PL_FLAG_CACHED | TTM_PL_FLAG_NO_EVICT,
-	VMW_PL_FLAG_GMR | TTM_PL_FLAG_CACHED | TTM_PL_FLAG_NO_EVICT
+static struct ttm_place vram_gmr_ne_placement_flags[] = {
+	{
+		.fpfn = 0,
+		.lpfn = 0,
+		.flags = TTM_PL_FLAG_VRAM | TTM_PL_FLAG_CACHED |
+			 TTM_PL_FLAG_NO_EVICT
+	}, {
+		.fpfn = 0,
+		.lpfn = 0,
+		.flags = VMW_PL_FLAG_GMR | TTM_PL_FLAG_CACHED |
+			 TTM_PL_FLAG_NO_EVICT
+	}
 };
 
 struct ttm_placement vmw_vram_gmr_ne_placement = {
-	.fpfn = 0,
-	.lpfn = 0,
 	.num_placement = 2,
 	.placement = vram_gmr_ne_placement_flags,
 	.num_busy_placement = 1,
@@ -97,8 +132,6 @@ struct ttm_placement vmw_vram_gmr_ne_placement = {
 };
 
 struct ttm_placement vmw_vram_sys_placement = {
-	.fpfn = 0,
-	.lpfn = 0,
 	.num_placement = 1,
 	.placement = &vram_placement_flags,
 	.num_busy_placement = 1,
@@ -106,8 +139,6 @@ struct ttm_placement vmw_vram_sys_placement = {
 };
 
 struct ttm_placement vmw_vram_ne_placement = {
-	.fpfn = 0,
-	.lpfn = 0,
 	.num_placement = 1,
 	.placement = &vram_ne_placement_flags,
 	.num_busy_placement = 1,
@@ -115,8 +146,6 @@ struct ttm_placement vmw_vram_ne_placement = {
 };
 
 struct ttm_placement vmw_sys_placement = {
-	.fpfn = 0,
-	.lpfn = 0,
 	.num_placement = 1,
 	.placement = &sys_placement_flags,
 	.num_busy_placement = 1,
@@ -124,24 +153,33 @@ struct ttm_placement vmw_sys_placement = {
 };
 
 struct ttm_placement vmw_sys_ne_placement = {
-	.fpfn = 0,
-	.lpfn = 0,
 	.num_placement = 1,
 	.placement = &sys_ne_placement_flags,
 	.num_busy_placement = 1,
 	.busy_placement = &sys_ne_placement_flags
 };
 
-static uint32_t evictable_placement_flags[] = {
-	TTM_PL_FLAG_SYSTEM | TTM_PL_FLAG_CACHED,
-	TTM_PL_FLAG_VRAM | TTM_PL_FLAG_CACHED,
-	VMW_PL_FLAG_GMR | TTM_PL_FLAG_CACHED,
-	VMW_PL_FLAG_MOB | TTM_PL_FLAG_CACHED
+static struct ttm_place evictable_placement_flags[] = {
+	{
+		.fpfn = 0,
+		.lpfn = 0,
+		.flags = TTM_PL_FLAG_SYSTEM | TTM_PL_FLAG_CACHED
+	}, {
+		.fpfn = 0,
+		.lpfn = 0,
+		.flags = TTM_PL_FLAG_VRAM | TTM_PL_FLAG_CACHED
+	}, {
+		.fpfn = 0,
+		.lpfn = 0,
+		.flags = VMW_PL_FLAG_GMR | TTM_PL_FLAG_CACHED
+	}, {
+		.fpfn = 0,
+		.lpfn = 0,
+		.flags = VMW_PL_FLAG_MOB | TTM_PL_FLAG_CACHED
+	}
 };
 
 struct ttm_placement vmw_evictable_placement = {
-	.fpfn = 0,
-	.lpfn = 0,
 	.num_placement = 4,
 	.placement = evictable_placement_flags,
 	.num_busy_placement = 1,
@@ -149,8 +187,6 @@ struct ttm_placement vmw_evictable_placement = {
 };
 
 struct ttm_placement vmw_srf_placement = {
-	.fpfn = 0,
-	.lpfn = 0,
 	.num_placement = 1,
 	.num_busy_placement = 2,
 	.placement = &gmr_placement_flags,
@@ -158,8 +194,6 @@ struct ttm_placement vmw_srf_placement = {
 };
 
 struct ttm_placement vmw_mob_placement = {
-	.fpfn = 0,
-	.lpfn = 0,
 	.num_placement = 1,
 	.num_busy_placement = 1,
 	.placement = &mob_placement_flags,
@@ -768,44 +802,6 @@ static int vmw_ttm_fault_reserve_notify(struct ttm_buffer_object *bo)
 }
 
 /**
- * FIXME: We're using the old vmware polling method to sync.
- * Do this with fences instead.
- */
-
-static void *vmw_sync_obj_ref(void *sync_obj)
-{
-
-	return (void *)
-		vmw_fence_obj_reference((struct vmw_fence_obj *) sync_obj);
-}
-
-static void vmw_sync_obj_unref(void **sync_obj)
-{
-	vmw_fence_obj_unreference((struct vmw_fence_obj **) sync_obj);
-}
-
-static int vmw_sync_obj_flush(void *sync_obj)
-{
-	vmw_fence_obj_flush((struct vmw_fence_obj *) sync_obj);
-	return 0;
-}
-
-static bool vmw_sync_obj_signaled(void *sync_obj)
-{
-	return	vmw_fence_obj_signaled((struct vmw_fence_obj *) sync_obj,
-				       DRM_VMW_FENCE_FLAG_EXEC);
-
-}
-
-static int vmw_sync_obj_wait(void *sync_obj, bool lazy, bool interruptible)
-{
-	return vmw_fence_obj_wait((struct vmw_fence_obj *) sync_obj,
-				  DRM_VMW_FENCE_FLAG_EXEC,
-				  lazy, interruptible,
-				  VMW_FENCE_WAIT_TIMEOUT);
-}
-
-/**
  * vmw_move_notify - TTM move_notify_callback
  *
  * @bo:             The TTM buffer object about to move.
@@ -829,11 +825,7 @@ static void vmw_move_notify(struct ttm_buffer_object *bo,
  */
 static void vmw_swap_notify(struct ttm_buffer_object *bo)
 {
-	struct ttm_bo_device *bdev = bo->bdev;
-
-	spin_lock(&bdev->fence_lock);
 	ttm_bo_wait(bo, false, false, false);
-	spin_unlock(&bdev->fence_lock);
 }
 
 
@@ -846,11 +838,6 @@ struct ttm_bo_driver vmw_bo_driver = {
 	.evict_flags = vmw_evict_flags,
 	.move = NULL,
 	.verify_access = vmw_verify_access,
-	.sync_obj_signaled = vmw_sync_obj_signaled,
-	.sync_obj_wait = vmw_sync_obj_wait,
-	.sync_obj_flush = vmw_sync_obj_flush,
-	.sync_obj_unref = vmw_sync_obj_unref,
-	.sync_obj_ref = vmw_sync_obj_ref,
 	.move_notify = vmw_move_notify,
 	.swap_notify = vmw_swap_notify,
 	.fault_reserve_notify = &vmw_ttm_fault_reserve_notify,
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_dmabuf.c b/drivers/gpu/drm/vmwgfx/vmwgfx_dmabuf.c
index ed1d51006ab1..914b375763dc 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_dmabuf.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_dmabuf.c
@@ -198,13 +198,19 @@ int vmw_dmabuf_to_start_of_vram(struct vmw_private *dev_priv,
 {
 	struct ttm_buffer_object *bo = &buf->base;
 	struct ttm_placement placement;
+	struct ttm_place place;
 	int ret = 0;
 
 	if (pin)
-		placement = vmw_vram_ne_placement;
+		place = vmw_vram_ne_placement.placement[0];
 	else
-		placement = vmw_vram_placement;
-	placement.lpfn = bo->num_pages;
+		place = vmw_vram_placement.placement[0];
+	place.lpfn = bo->num_pages;
+
+	placement.num_placement = 1;
+	placement.placement = &place;
+	placement.num_busy_placement = 1;
+	placement.busy_placement = &place;
 
 	ret = ttm_write_lock(&dev_priv->reservation_sem, interruptible);
 	if (unlikely(ret != 0))
@@ -293,21 +299,23 @@ void vmw_bo_get_guest_ptr(const struct ttm_buffer_object *bo,
  */
 void vmw_bo_pin(struct ttm_buffer_object *bo, bool pin)
 {
-	uint32_t pl_flags;
+	struct ttm_place pl;
 	struct ttm_placement placement;
 	uint32_t old_mem_type = bo->mem.mem_type;
 	int ret;
 
 	lockdep_assert_held(&bo->resv->lock.base);
 
-	pl_flags = TTM_PL_FLAG_VRAM | VMW_PL_FLAG_GMR | VMW_PL_FLAG_MOB
+	pl.fpfn = 0;
+	pl.lpfn = 0;
+	pl.flags = TTM_PL_FLAG_VRAM | VMW_PL_FLAG_GMR | VMW_PL_FLAG_MOB
 		| TTM_PL_FLAG_SYSTEM | TTM_PL_FLAG_CACHED;
 	if (pin)
-		pl_flags |= TTM_PL_FLAG_NO_EVICT;
+		pl.flags |= TTM_PL_FLAG_NO_EVICT;
 
 	memset(&placement, 0, sizeof(placement));
 	placement.num_placement = 1;
-	placement.placement = &pl_flags;
+	placement.placement = &pl;
 
 	ret = ttm_bo_validate(bo, &placement, false, true);
 
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c
index 18b54acacfbb..7197af157313 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c
@@ -1418,6 +1418,7 @@ static struct drm_driver driver = {
 	.open = vmw_driver_open,
 	.preclose = vmw_preclose,
 	.postclose = vmw_postclose,
+	.set_busid = drm_pci_set_busid,
 
 	.dumb_create = vmw_dumb_create,
 	.dumb_map_offset = vmw_dumb_map_offset,
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h
index 99f731757c4b..4ee799b43d5d 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h
@@ -342,7 +342,6 @@ struct vmw_sw_context{
 	uint32_t *cmd_bounce;
 	uint32_t cmd_bounce_size;
 	struct list_head resource_list;
-	uint32_t fence_flags;
 	struct ttm_buffer_object *cur_query_bo;
 	struct list_head res_relocations;
 	uint32_t *buf_start;
@@ -704,6 +703,7 @@ extern void *vmw_fifo_reserve(struct vmw_private *dev_priv, uint32_t bytes);
 extern void vmw_fifo_commit(struct vmw_private *dev_priv, uint32_t bytes);
 extern int vmw_fifo_send_fence(struct vmw_private *dev_priv,
 			       uint32_t *seqno);
+extern void vmw_fifo_ping_host_locked(struct vmw_private *, uint32_t reason);
 extern void vmw_fifo_ping_host(struct vmw_private *dev_priv, uint32_t reason);
 extern bool vmw_fifo_have_3d(struct vmw_private *dev_priv);
 extern bool vmw_fifo_have_pitchlock(struct vmw_private *dev_priv);
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_execbuf.c b/drivers/gpu/drm/vmwgfx/vmwgfx_execbuf.c
index 36b871686d3c..596cd6dafd33 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_execbuf.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_execbuf.c
@@ -346,13 +346,11 @@ static int vmw_bo_to_validate_list(struct vmw_sw_context *sw_context,
 		++sw_context->cur_val_buf;
 		val_buf = &vval_buf->base;
 		val_buf->bo = ttm_bo_reference(bo);
-		val_buf->reserved = false;
+		val_buf->shared = false;
 		list_add_tail(&val_buf->head, &sw_context->validate_nodes);
 		vval_buf->validate_as_mob = validate_as_mob;
 	}
 
-	sw_context->fence_flags |= DRM_VMW_FENCE_FLAG_EXEC;
-
 	if (p_val_node)
 		*p_val_node = val_node;
 
@@ -2337,13 +2335,9 @@ int vmw_execbuf_fence_commands(struct drm_file *file_priv,
 
 	if (p_handle != NULL)
 		ret = vmw_user_fence_create(file_priv, dev_priv->fman,
-					    sequence,
-					    DRM_VMW_FENCE_FLAG_EXEC,
-					    p_fence, p_handle);
+					    sequence, p_fence, p_handle);
 	else
-		ret = vmw_fence_create(dev_priv->fman, sequence,
-				       DRM_VMW_FENCE_FLAG_EXEC,
-				       p_fence);
+		ret = vmw_fence_create(dev_priv->fman, sequence, p_fence);
 
 	if (unlikely(ret != 0 && !synced)) {
 		(void) vmw_fallback_wait(dev_priv, false, false,
@@ -2395,7 +2389,7 @@ vmw_execbuf_copy_fence_user(struct vmw_private *dev_priv,
 		BUG_ON(fence == NULL);
 
 		fence_rep.handle = fence_handle;
-		fence_rep.seqno = fence->seqno;
+		fence_rep.seqno = fence->base.seqno;
 		vmw_update_seqno(dev_priv, &dev_priv->fifo);
 		fence_rep.passed_seqno = dev_priv->last_read_seqno;
 	}
@@ -2416,8 +2410,7 @@ vmw_execbuf_copy_fence_user(struct vmw_private *dev_priv,
 		ttm_ref_object_base_unref(vmw_fp->tfile,
 					  fence_handle, TTM_REF_USAGE);
 		DRM_ERROR("Fence copy error. Syncing.\n");
-		(void) vmw_fence_obj_wait(fence, fence->signal_mask,
-					  false, false,
+		(void) vmw_fence_obj_wait(fence, false, false,
 					  VMW_FENCE_WAIT_TIMEOUT);
 	}
 }
@@ -2469,7 +2462,6 @@ int vmw_execbuf_process(struct drm_file *file_priv,
 	sw_context->fp = vmw_fpriv(file_priv);
 	sw_context->cur_reloc = 0;
 	sw_context->cur_val_buf = 0;
-	sw_context->fence_flags = 0;
 	INIT_LIST_HEAD(&sw_context->resource_list);
 	sw_context->cur_query_bo = dev_priv->pinned_bo;
 	sw_context->last_query_ctx = NULL;
@@ -2495,7 +2487,7 @@ int vmw_execbuf_process(struct drm_file *file_priv,
 	if (unlikely(ret != 0))
 		goto out_err_nores;
 
-	ret = ttm_eu_reserve_buffers(&ticket, &sw_context->validate_nodes);
+	ret = ttm_eu_reserve_buffers(&ticket, &sw_context->validate_nodes, true);
 	if (unlikely(ret != 0))
 		goto out_err;
 
@@ -2678,15 +2670,14 @@ void __vmw_execbuf_release_pinned_bo(struct vmw_private *dev_priv,
 	INIT_LIST_HEAD(&validate_list);
 
 	pinned_val.bo = ttm_bo_reference(dev_priv->pinned_bo);
+	pinned_val.shared = false;
 	list_add_tail(&pinned_val.head, &validate_list);
 
 	query_val.bo = ttm_bo_reference(dev_priv->dummy_query_bo);
+	query_val.shared = false;
 	list_add_tail(&query_val.head, &validate_list);
 
-	do {
-		ret = ttm_eu_reserve_buffers(&ticket, &validate_list);
-	} while (ret == -ERESTARTSYS);
-
+	ret = ttm_eu_reserve_buffers(&ticket, &validate_list, false);
 	if (unlikely(ret != 0)) {
 		vmw_execbuf_unpin_panic(dev_priv);
 		goto out_no_reserve;
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_fb.c b/drivers/gpu/drm/vmwgfx/vmwgfx_fb.c
index b031b48dbb3c..0a474f391fad 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_fb.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_fb.c
@@ -374,10 +374,16 @@ static int vmw_fb_create_bo(struct vmw_private *vmw_priv,
 			    size_t size, struct vmw_dma_buffer **out)
 {
 	struct vmw_dma_buffer *vmw_bo;
-	struct ttm_placement ne_placement = vmw_vram_ne_placement;
+	struct ttm_place ne_place = vmw_vram_ne_placement.placement[0];
+	struct ttm_placement ne_placement;
 	int ret;
 
-	ne_placement.lpfn = (size + PAGE_SIZE - 1) >> PAGE_SHIFT;
+	ne_placement.num_placement = 1;
+	ne_placement.placement = &ne_place;
+	ne_placement.num_busy_placement = 1;
+	ne_placement.busy_placement = &ne_place;
+
+	ne_place.lpfn = (size + PAGE_SIZE - 1) >> PAGE_SHIFT;
 
 	(void) ttm_write_lock(&vmw_priv->reservation_sem, false);
 
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_fence.c b/drivers/gpu/drm/vmwgfx/vmwgfx_fence.c
index 436b013b4231..197164fd7803 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_fence.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_fence.c
@@ -35,7 +35,7 @@ struct vmw_fence_manager {
 	struct vmw_private *dev_priv;
 	spinlock_t lock;
 	struct list_head fence_list;
-	struct work_struct work;
+	struct work_struct work, ping_work;
 	u32 user_fence_size;
 	u32 fence_size;
 	u32 event_fence_action_size;
@@ -46,6 +46,7 @@ struct vmw_fence_manager {
 	bool goal_irq_on; /* Protected by @goal_irq_mutex */
 	bool seqno_valid; /* Protected by @lock, and may not be set to true
 			     without the @goal_irq_mutex held. */
+	unsigned ctx;
 };
 
 struct vmw_user_fence {
@@ -80,6 +81,12 @@ struct vmw_event_fence_action {
 	uint32_t *tv_usec;
 };
 
+static struct vmw_fence_manager *
+fman_from_fence(struct vmw_fence_obj *fence)
+{
+	return container_of(fence->base.lock, struct vmw_fence_manager, lock);
+}
+
 /**
  * Note on fencing subsystem usage of irqs:
  * Typically the vmw_fences_update function is called
@@ -102,25 +109,143 @@ struct vmw_event_fence_action {
  * objects with actions attached to them.
  */
 
-static void vmw_fence_obj_destroy_locked(struct kref *kref)
+static void vmw_fence_obj_destroy(struct fence *f)
 {
 	struct vmw_fence_obj *fence =
-		container_of(kref, struct vmw_fence_obj, kref);
+		container_of(f, struct vmw_fence_obj, base);
 
-	struct vmw_fence_manager *fman = fence->fman;
-	unsigned int num_fences;
+	struct vmw_fence_manager *fman = fman_from_fence(fence);
+	unsigned long irq_flags;
 
+	spin_lock_irqsave(&fman->lock, irq_flags);
 	list_del_init(&fence->head);
-	num_fences = --fman->num_fence_objects;
-	spin_unlock_irq(&fman->lock);
-	if (fence->destroy)
-		fence->destroy(fence);
-	else
-		kfree(fence);
+	--fman->num_fence_objects;
+	spin_unlock_irqrestore(&fman->lock, irq_flags);
+	fence->destroy(fence);
+}
 
-	spin_lock_irq(&fman->lock);
+static const char *vmw_fence_get_driver_name(struct fence *f)
+{
+	return "vmwgfx";
+}
+
+static const char *vmw_fence_get_timeline_name(struct fence *f)
+{
+	return "svga";
+}
+
+static void vmw_fence_ping_func(struct work_struct *work)
+{
+	struct vmw_fence_manager *fman =
+		container_of(work, struct vmw_fence_manager, ping_work);
+
+	vmw_fifo_ping_host(fman->dev_priv, SVGA_SYNC_GENERIC);
+}
+
+static bool vmw_fence_enable_signaling(struct fence *f)
+{
+	struct vmw_fence_obj *fence =
+		container_of(f, struct vmw_fence_obj, base);
+
+	struct vmw_fence_manager *fman = fman_from_fence(fence);
+	struct vmw_private *dev_priv = fman->dev_priv;
+
+	__le32 __iomem *fifo_mem = dev_priv->mmio_virt;
+	u32 seqno = ioread32(fifo_mem + SVGA_FIFO_FENCE);
+	if (seqno - fence->base.seqno < VMW_FENCE_WRAP)
+		return false;
+
+	if (mutex_trylock(&dev_priv->hw_mutex)) {
+		vmw_fifo_ping_host_locked(dev_priv, SVGA_SYNC_GENERIC);
+		mutex_unlock(&dev_priv->hw_mutex);
+	} else
+		schedule_work(&fman->ping_work);
+
+	return true;
+}
+
+struct vmwgfx_wait_cb {
+	struct fence_cb base;
+	struct task_struct *task;
+};
+
+static void
+vmwgfx_wait_cb(struct fence *fence, struct fence_cb *cb)
+{
+	struct vmwgfx_wait_cb *wait =
+		container_of(cb, struct vmwgfx_wait_cb, base);
+
+	wake_up_process(wait->task);
+}
+
+static void __vmw_fences_update(struct vmw_fence_manager *fman);
+
+static long vmw_fence_wait(struct fence *f, bool intr, signed long timeout)
+{
+	struct vmw_fence_obj *fence =
+		container_of(f, struct vmw_fence_obj, base);
+
+	struct vmw_fence_manager *fman = fman_from_fence(fence);
+	struct vmw_private *dev_priv = fman->dev_priv;
+	struct vmwgfx_wait_cb cb;
+	long ret = timeout;
+	unsigned long irq_flags;
+
+	if (likely(vmw_fence_obj_signaled(fence)))
+		return timeout;
+
+	vmw_fifo_ping_host(dev_priv, SVGA_SYNC_GENERIC);
+	vmw_seqno_waiter_add(dev_priv);
+
+	spin_lock_irqsave(f->lock, irq_flags);
+
+	if (intr && signal_pending(current)) {
+		ret = -ERESTARTSYS;
+		goto out;
+	}
+
+	cb.base.func = vmwgfx_wait_cb;
+	cb.task = current;
+	list_add(&cb.base.node, &f->cb_list);
+
+	while (ret > 0) {
+		__vmw_fences_update(fman);
+		if (test_bit(FENCE_FLAG_SIGNALED_BIT, &f->flags))
+			break;
+
+		if (intr)
+			__set_current_state(TASK_INTERRUPTIBLE);
+		else
+			__set_current_state(TASK_UNINTERRUPTIBLE);
+		spin_unlock_irqrestore(f->lock, irq_flags);
+
+		ret = schedule_timeout(ret);
+
+		spin_lock_irqsave(f->lock, irq_flags);
+		if (ret > 0 && intr && signal_pending(current))
+			ret = -ERESTARTSYS;
+	}
+
+	if (!list_empty(&cb.base.node))
+		list_del(&cb.base.node);
+	__set_current_state(TASK_RUNNING);
+
+out:
+	spin_unlock_irqrestore(f->lock, irq_flags);
+
+	vmw_seqno_waiter_remove(dev_priv);
+
+	return ret;
 }
 
+static struct fence_ops vmw_fence_ops = {
+	.get_driver_name = vmw_fence_get_driver_name,
+	.get_timeline_name = vmw_fence_get_timeline_name,
+	.enable_signaling = vmw_fence_enable_signaling,
+	.wait = vmw_fence_wait,
+	.release = vmw_fence_obj_destroy,
+};
+
 
 /**
  * Execute signal actions on fences recently signaled.
@@ -180,12 +305,14 @@ struct vmw_fence_manager *vmw_fence_manager_init(struct vmw_private *dev_priv)
 	INIT_LIST_HEAD(&fman->fence_list);
 	INIT_LIST_HEAD(&fman->cleanup_list);
 	INIT_WORK(&fman->work, &vmw_fence_work_func);
+	INIT_WORK(&fman->ping_work, &vmw_fence_ping_func);
 	fman->fifo_down = true;
 	fman->user_fence_size = ttm_round_pot(sizeof(struct vmw_user_fence));
 	fman->fence_size = ttm_round_pot(sizeof(struct vmw_fence_obj));
 	fman->event_fence_action_size =
 		ttm_round_pot(sizeof(struct vmw_event_fence_action));
 	mutex_init(&fman->goal_irq_mutex);
+	fman->ctx = fence_context_alloc(1);
 
 	return fman;
 }
@@ -196,6 +323,7 @@ void vmw_fence_manager_takedown(struct vmw_fence_manager *fman)
 	bool lists_empty;
 
 	(void) cancel_work_sync(&fman->work);
+	(void) cancel_work_sync(&fman->ping_work);
 
 	spin_lock_irqsave(&fman->lock, irq_flags);
 	lists_empty = list_empty(&fman->fence_list) &&
@@ -207,23 +335,16 @@ void vmw_fence_manager_takedown(struct vmw_fence_manager *fman)
 }
 
 static int vmw_fence_obj_init(struct vmw_fence_manager *fman,
-			      struct vmw_fence_obj *fence,
-			      u32 seqno,
-			      uint32_t mask,
+			      struct vmw_fence_obj *fence, u32 seqno,
 			      void (*destroy) (struct vmw_fence_obj *fence))
 {
 	unsigned long irq_flags;
-	unsigned int num_fences;
 	int ret = 0;
 
-	fence->seqno = seqno;
+	fence_init(&fence->base, &vmw_fence_ops, &fman->lock,
+		   fman->ctx, seqno);
 	INIT_LIST_HEAD(&fence->seq_passed_actions);
-	fence->fman = fman;
-	fence->signaled = 0;
-	fence->signal_mask = mask;
-	kref_init(&fence->kref);
 	fence->destroy = destroy;
-	init_waitqueue_head(&fence->queue);
 
 	spin_lock_irqsave(&fman->lock, irq_flags);
 	if (unlikely(fman->fifo_down)) {
@@ -231,7 +352,7 @@ static int vmw_fence_obj_init(struct vmw_fence_manager *fman,
 		goto out_unlock;
 	}
 	list_add_tail(&fence->head, &fman->fence_list);
-	num_fences = ++fman->num_fence_objects;
+	++fman->num_fence_objects;
 
 out_unlock:
 	spin_unlock_irqrestore(&fman->lock, irq_flags);
@@ -239,38 +360,6 @@ out_unlock:
 
 }
 
-struct vmw_fence_obj *vmw_fence_obj_reference(struct vmw_fence_obj *fence)
-{
-	if (unlikely(fence == NULL))
-		return NULL;
-
-	kref_get(&fence->kref);
-	return fence;
-}
-
-/**
- * vmw_fence_obj_unreference
- *
- * Note that this function may not be entered with disabled irqs since
- * it may re-enable them in the destroy function.
- *
- */
-void vmw_fence_obj_unreference(struct vmw_fence_obj **fence_p)
-{
-	struct vmw_fence_obj *fence = *fence_p;
-	struct vmw_fence_manager *fman;
-
-	if (unlikely(fence == NULL))
-		return;
-
-	fman = fence->fman;
-	*fence_p = NULL;
-	spin_lock_irq(&fman->lock);
-	BUG_ON(atomic_read(&fence->kref.refcount) == 0);
-	kref_put(&fence->kref, vmw_fence_obj_destroy_locked);
-	spin_unlock_irq(&fman->lock);
-}
-
 static void vmw_fences_perform_actions(struct vmw_fence_manager *fman,
 				struct list_head *list)
 {
@@ -326,7 +415,7 @@ static bool vmw_fence_goal_new_locked(struct vmw_fence_manager *fman,
 	list_for_each_entry(fence, &fman->fence_list, head) {
 		if (!list_empty(&fence->seq_passed_actions)) {
 			fman->seqno_valid = true;
-			iowrite32(fence->seqno,
+			iowrite32(fence->base.seqno,
 				  fifo_mem + SVGA_FIFO_FENCE_GOAL);
 			break;
 		}
@@ -353,27 +442,27 @@ static bool vmw_fence_goal_new_locked(struct vmw_fence_manager *fman,
  */
 static bool vmw_fence_goal_check_locked(struct vmw_fence_obj *fence)
 {
+	struct vmw_fence_manager *fman = fman_from_fence(fence);
 	u32 goal_seqno;
 	__le32 __iomem *fifo_mem;
 
-	if (fence->signaled & DRM_VMW_FENCE_FLAG_EXEC)
+	if (fence_is_signaled_locked(&fence->base))
 		return false;
 
-	fifo_mem = fence->fman->dev_priv->mmio_virt;
+	fifo_mem = fman->dev_priv->mmio_virt;
 	goal_seqno = ioread32(fifo_mem + SVGA_FIFO_FENCE_GOAL);
-	if (likely(fence->fman->seqno_valid &&
-		   goal_seqno - fence->seqno < VMW_FENCE_WRAP))
+	if (likely(fman->seqno_valid &&
+		   goal_seqno - fence->base.seqno < VMW_FENCE_WRAP))
 		return false;
 
-	iowrite32(fence->seqno, fifo_mem + SVGA_FIFO_FENCE_GOAL);
-	fence->fman->seqno_valid = true;
+	iowrite32(fence->base.seqno, fifo_mem + SVGA_FIFO_FENCE_GOAL);
+	fman->seqno_valid = true;
 
 	return true;
 }
 
-void vmw_fences_update(struct vmw_fence_manager *fman)
+static void __vmw_fences_update(struct vmw_fence_manager *fman)
 {
-	unsigned long flags;
 	struct vmw_fence_obj *fence, *next_fence;
 	struct list_head action_list;
 	bool needs_rerun;
@@ -382,32 +471,25 @@ void vmw_fences_update(struct vmw_fence_manager *fman)
 
 	seqno = ioread32(fifo_mem + SVGA_FIFO_FENCE);
 rerun:
-	spin_lock_irqsave(&fman->lock, flags);
 	list_for_each_entry_safe(fence, next_fence, &fman->fence_list, head) {
-		if (seqno - fence->seqno < VMW_FENCE_WRAP) {
+		if (seqno - fence->base.seqno < VMW_FENCE_WRAP) {
 			list_del_init(&fence->head);
-			fence->signaled |= DRM_VMW_FENCE_FLAG_EXEC;
+			fence_signal_locked(&fence->base);
 			INIT_LIST_HEAD(&action_list);
 			list_splice_init(&fence->seq_passed_actions,
 					 &action_list);
 			vmw_fences_perform_actions(fman, &action_list);
-			wake_up_all(&fence->queue);
 		} else
 			break;
 	}
 
-	needs_rerun = vmw_fence_goal_new_locked(fman, seqno);
-
-	if (!list_empty(&fman->cleanup_list))
-		(void) schedule_work(&fman->work);
-	spin_unlock_irqrestore(&fman->lock, flags);
-
 	/*
 	 * Rerun if the fence goal seqno was updated, and the
 	 * hardware might have raced with that update, so that
 	 * we missed a fence_goal irq.
 	 */
 
+	needs_rerun = vmw_fence_goal_new_locked(fman, seqno);
 	if (unlikely(needs_rerun)) {
 		new_seqno = ioread32(fifo_mem + SVGA_FIFO_FENCE);
 		if (new_seqno != seqno) {
@@ -415,79 +497,58 @@ rerun:
 			goto rerun;
 		}
 	}
+
+	if (!list_empty(&fman->cleanup_list))
+		(void) schedule_work(&fman->work);
 }
 
-bool vmw_fence_obj_signaled(struct vmw_fence_obj *fence,
-			    uint32_t flags)
+void vmw_fences_update(struct vmw_fence_manager *fman)
 {
-	struct vmw_fence_manager *fman = fence->fman;
 	unsigned long irq_flags;
-	uint32_t signaled;
 
 	spin_lock_irqsave(&fman->lock, irq_flags);
-	signaled = fence->signaled;
+	__vmw_fences_update(fman);
 	spin_unlock_irqrestore(&fman->lock, irq_flags);
+}
 
-	flags &= fence->signal_mask;
-	if ((signaled & flags) == flags)
-		return 1;
+bool vmw_fence_obj_signaled(struct vmw_fence_obj *fence)
+{
+	struct vmw_fence_manager *fman = fman_from_fence(fence);
 
-	if ((signaled & DRM_VMW_FENCE_FLAG_EXEC) == 0)
-		vmw_fences_update(fman);
+	if (test_bit(FENCE_FLAG_SIGNALED_BIT, &fence->base.flags))
+		return 1;
 
-	spin_lock_irqsave(&fman->lock, irq_flags);
-	signaled = fence->signaled;
-	spin_unlock_irqrestore(&fman->lock, irq_flags);
+	vmw_fences_update(fman);
 
-	return ((signaled & flags) == flags);
+	return fence_is_signaled(&fence->base);
 }
 
-int vmw_fence_obj_wait(struct vmw_fence_obj *fence,
-		       uint32_t flags, bool lazy,
+int vmw_fence_obj_wait(struct vmw_fence_obj *fence, bool lazy,
 		       bool interruptible, unsigned long timeout)
 {
-	struct vmw_private *dev_priv = fence->fman->dev_priv;
-	long ret;
+	long ret = fence_wait_timeout(&fence->base, interruptible, timeout);
 
-	if (likely(vmw_fence_obj_signaled(fence, flags)))
+	if (likely(ret > 0))
 		return 0;
-
-	vmw_fifo_ping_host(dev_priv, SVGA_SYNC_GENERIC);
-	vmw_seqno_waiter_add(dev_priv);
-
-	if (interruptible)
-		ret = wait_event_interruptible_timeout
-			(fence->queue,
-			 vmw_fence_obj_signaled(fence, flags),
-			 timeout);
+	else if (ret == 0)
+		return -EBUSY;
 	else
-		ret = wait_event_timeout
-			(fence->queue,
-			 vmw_fence_obj_signaled(fence, flags),
-			 timeout);
-
-	vmw_seqno_waiter_remove(dev_priv);
-
-	if (unlikely(ret == 0))
-		ret = -EBUSY;
-	else if (likely(ret > 0))
-		ret = 0;
-
-	return ret;
+		return ret;
 }
 
 void vmw_fence_obj_flush(struct vmw_fence_obj *fence)
 {
-	struct vmw_private *dev_priv = fence->fman->dev_priv;
+	struct vmw_private *dev_priv = fman_from_fence(fence)->dev_priv;
 
 	vmw_fifo_ping_host(dev_priv, SVGA_SYNC_GENERIC);
 }
 
 static void vmw_fence_destroy(struct vmw_fence_obj *fence)
 {
-	struct vmw_fence_manager *fman = fence->fman;
+	struct vmw_fence_manager *fman = fman_from_fence(fence);
+
+	fence_free(&fence->base);
 
-	kfree(fence);
 	/*
 	 * Free kernel space accounting.
 	 */
@@ -497,7 +558,6 @@ static void vmw_fence_destroy(struct vmw_fence_obj *fence)
 
 int vmw_fence_create(struct vmw_fence_manager *fman,
 		     uint32_t seqno,
-		     uint32_t mask,
 		     struct vmw_fence_obj **p_fence)
 {
 	struct ttm_mem_global *mem_glob = vmw_mem_glob(fman->dev_priv);
@@ -515,7 +575,7 @@ int vmw_fence_create(struct vmw_fence_manager *fman,
 		goto out_no_object;
 	}
 
-	ret = vmw_fence_obj_init(fman, fence, seqno, mask,
+	ret = vmw_fence_obj_init(fman, fence, seqno,
 				 vmw_fence_destroy);
 	if (unlikely(ret != 0))
 		goto out_err_init;
@@ -535,7 +595,7 @@ static void vmw_user_fence_destroy(struct vmw_fence_obj *fence)
 {
 	struct vmw_user_fence *ufence =
 		container_of(fence, struct vmw_user_fence, fence);
-	struct vmw_fence_manager *fman = fence->fman;
+	struct vmw_fence_manager *fman = fman_from_fence(fence);
 
 	ttm_base_object_kfree(ufence, base);
 	/*
@@ -559,7 +619,6 @@ static void vmw_user_fence_base_release(struct ttm_base_object **p_base)
 int vmw_user_fence_create(struct drm_file *file_priv,
 			  struct vmw_fence_manager *fman,
 			  uint32_t seqno,
-			  uint32_t mask,
 			  struct vmw_fence_obj **p_fence,
 			  uint32_t *p_handle)
 {
@@ -586,7 +645,7 @@ int vmw_user_fence_create(struct drm_file *file_priv,
 	}
 
 	ret = vmw_fence_obj_init(fman, &ufence->fence, seqno,
-				 mask, vmw_user_fence_destroy);
+				 vmw_user_fence_destroy);
 	if (unlikely(ret != 0)) {
 		kfree(ufence);
 		goto out_no_object;
@@ -629,7 +688,6 @@ out_no_object:
 
 void vmw_fence_fifo_down(struct vmw_fence_manager *fman)
 {
-	unsigned long irq_flags;
 	struct list_head action_list;
 	int ret;
 
@@ -638,35 +696,32 @@ void vmw_fence_fifo_down(struct vmw_fence_manager *fman)
 	 * restart when we've released the fman->lock.
 	 */
 
-	spin_lock_irqsave(&fman->lock, irq_flags);
+	spin_lock_irq(&fman->lock);
 	fman->fifo_down = true;
 	while (!list_empty(&fman->fence_list)) {
 		struct vmw_fence_obj *fence =
 			list_entry(fman->fence_list.prev, struct vmw_fence_obj,
 				   head);
-		kref_get(&fence->kref);
+		fence_get(&fence->base);
 		spin_unlock_irq(&fman->lock);
 
-		ret = vmw_fence_obj_wait(fence, fence->signal_mask,
-					 false, false,
+		ret = vmw_fence_obj_wait(fence, false, false,
 					 VMW_FENCE_WAIT_TIMEOUT);
 
 		if (unlikely(ret != 0)) {
 			list_del_init(&fence->head);
-			fence->signaled |= DRM_VMW_FENCE_FLAG_EXEC;
+			fence_signal(&fence->base);
 			INIT_LIST_HEAD(&action_list);
 			list_splice_init(&fence->seq_passed_actions,
 					 &action_list);
 			vmw_fences_perform_actions(fman, &action_list);
-			wake_up_all(&fence->queue);
 		}
 
-		spin_lock_irq(&fman->lock);
-
 		BUG_ON(!list_empty(&fence->head));
-		kref_put(&fence->kref, vmw_fence_obj_destroy_locked);
+		fence_put(&fence->base);
+		spin_lock_irq(&fman->lock);
 	}
-	spin_unlock_irqrestore(&fman->lock, irq_flags);
+	spin_unlock_irq(&fman->lock);
 }
 
 void vmw_fence_fifo_up(struct vmw_fence_manager *fman)
@@ -716,14 +771,14 @@ int vmw_fence_obj_wait_ioctl(struct drm_device *dev, void *data,
 
 	timeout = jiffies;
 	if (time_after_eq(timeout, (unsigned long)arg->kernel_cookie)) {
-		ret = ((vmw_fence_obj_signaled(fence, arg->flags)) ?
+		ret = ((vmw_fence_obj_signaled(fence)) ?
 		       0 : -EBUSY);
 		goto out;
 	}
 
 	timeout = (unsigned long)arg->kernel_cookie - timeout;
 
-	ret = vmw_fence_obj_wait(fence, arg->flags, arg->lazy, true, timeout);
+	ret = vmw_fence_obj_wait(fence, arg->lazy, true, timeout);
 
 out:
 	ttm_base_object_unref(&base);
@@ -758,12 +813,12 @@ int vmw_fence_obj_signaled_ioctl(struct drm_device *dev, void *data,
 	}
 
 	fence = &(container_of(base, struct vmw_user_fence, base)->fence);
-	fman = fence->fman;
+	fman = fman_from_fence(fence);
 
-	arg->signaled = vmw_fence_obj_signaled(fence, arg->flags);
-	spin_lock_irq(&fman->lock);
+	arg->signaled = vmw_fence_obj_signaled(fence);
 
-	arg->signaled_flags = fence->signaled;
+	arg->signaled_flags = arg->flags;
+	spin_lock_irq(&fman->lock);
 	arg->passed_seqno = dev_priv->last_read_seqno;
 	spin_unlock_irq(&fman->lock);
 
@@ -876,7 +931,7 @@ static void vmw_event_fence_action_cleanup(struct vmw_fence_action *action)
 {
 	struct vmw_event_fence_action *eaction =
 		container_of(action, struct vmw_event_fence_action, action);
-	struct vmw_fence_manager *fman = eaction->fence->fman;
+	struct vmw_fence_manager *fman = fman_from_fence(eaction->fence);
 	unsigned long irq_flags;
 
 	spin_lock_irqsave(&fman->lock, irq_flags);
@@ -900,7 +955,7 @@ static void vmw_event_fence_action_cleanup(struct vmw_fence_action *action)
 static void vmw_fence_obj_add_action(struct vmw_fence_obj *fence,
 			      struct vmw_fence_action *action)
 {
-	struct vmw_fence_manager *fman = fence->fman;
+	struct vmw_fence_manager *fman = fman_from_fence(fence);
 	unsigned long irq_flags;
 	bool run_update = false;
 
@@ -908,7 +963,7 @@ static void vmw_fence_obj_add_action(struct vmw_fence_obj *fence,
 	spin_lock_irqsave(&fman->lock, irq_flags);
 
 	fman->pending_actions[action->type]++;
-	if (fence->signaled & DRM_VMW_FENCE_FLAG_EXEC) {
+	if (fence_is_signaled_locked(&fence->base)) {
 		struct list_head action_list;
 
 		INIT_LIST_HEAD(&action_list);
@@ -960,7 +1015,7 @@ int vmw_event_fence_action_queue(struct drm_file *file_priv,
 				 bool interruptible)
 {
 	struct vmw_event_fence_action *eaction;
-	struct vmw_fence_manager *fman = fence->fman;
+	struct vmw_fence_manager *fman = fman_from_fence(fence);
 	struct vmw_fpriv *vmw_fp = vmw_fpriv(file_priv);
 	unsigned long irq_flags;
 
@@ -1000,7 +1055,8 @@ static int vmw_event_fence_action_create(struct drm_file *file_priv,
 				  bool interruptible)
 {
 	struct vmw_event_fence_pending *event;
-	struct drm_device *dev = fence->fman->dev_priv->dev;
+	struct vmw_fence_manager *fman = fman_from_fence(fence);
+	struct drm_device *dev = fman->dev_priv->dev;
 	unsigned long irq_flags;
 	int ret;
 
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_fence.h b/drivers/gpu/drm/vmwgfx/vmwgfx_fence.h
index faf2e7873860..26a4add39208 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_fence.h
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_fence.h
@@ -27,6 +27,8 @@
 
 #ifndef _VMWGFX_FENCE_H_
 
+#include <linux/fence.h>
+
 #define VMW_FENCE_WAIT_TIMEOUT (5*HZ)
 
 struct vmw_private;
@@ -50,16 +52,11 @@ struct vmw_fence_action {
 };
 
 struct vmw_fence_obj {
-	struct kref kref;
-	u32 seqno;
+	struct fence base;
 
-	struct vmw_fence_manager *fman;
 	struct list_head head;
-	uint32_t signaled;
-	uint32_t signal_mask;
 	struct list_head seq_passed_actions;
 	void (*destroy)(struct vmw_fence_obj *fence);
-	wait_queue_head_t queue;
 };
 
 extern struct vmw_fence_manager *
@@ -67,17 +64,29 @@ vmw_fence_manager_init(struct vmw_private *dev_priv);
 
 extern void vmw_fence_manager_takedown(struct vmw_fence_manager *fman);
 
-extern void vmw_fence_obj_unreference(struct vmw_fence_obj **fence_p);
+static inline void
+vmw_fence_obj_unreference(struct vmw_fence_obj **fence_p)
+{
+	struct vmw_fence_obj *fence = *fence_p;
+
+	*fence_p = NULL;
+	if (fence)
+		fence_put(&fence->base);
+}
 
-extern struct vmw_fence_obj *
-vmw_fence_obj_reference(struct vmw_fence_obj *fence);
+static inline struct vmw_fence_obj *
+vmw_fence_obj_reference(struct vmw_fence_obj *fence)
+{
+	if (fence)
+		fence_get(&fence->base);
+	return fence;
+}
 
 extern void vmw_fences_update(struct vmw_fence_manager *fman);
 
-extern bool vmw_fence_obj_signaled(struct vmw_fence_obj *fence,
-				   uint32_t flags);
+extern bool vmw_fence_obj_signaled(struct vmw_fence_obj *fence);
 
-extern int vmw_fence_obj_wait(struct vmw_fence_obj *fence, uint32_t flags,
+extern int vmw_fence_obj_wait(struct vmw_fence_obj *fence,
 			      bool lazy,
 			      bool interruptible, unsigned long timeout);
 
@@ -85,13 +94,11 @@ extern void vmw_fence_obj_flush(struct vmw_fence_obj *fence);
 
 extern int vmw_fence_create(struct vmw_fence_manager *fman,
 			    uint32_t seqno,
-			    uint32_t mask,
 			    struct vmw_fence_obj **p_fence);
 
 extern int vmw_user_fence_create(struct drm_file *file_priv,
 				 struct vmw_fence_manager *fman,
 				 uint32_t sequence,
-				 uint32_t mask,
 				 struct vmw_fence_obj **p_fence,
 				 uint32_t *p_handle);
 
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_fifo.c b/drivers/gpu/drm/vmwgfx/vmwgfx_fifo.c
index 6eae14d2a3f7..09e10aefcd8e 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_fifo.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_fifo.c
@@ -160,16 +160,21 @@ int vmw_fifo_init(struct vmw_private *dev_priv, struct vmw_fifo_state *fifo)
 	return vmw_fifo_send_fence(dev_priv, &dummy);
 }
 
-void vmw_fifo_ping_host(struct vmw_private *dev_priv, uint32_t reason)
+void vmw_fifo_ping_host_locked(struct vmw_private *dev_priv, uint32_t reason)
 {
 	__le32 __iomem *fifo_mem = dev_priv->mmio_virt;
 
-	mutex_lock(&dev_priv->hw_mutex);
-
 	if (unlikely(ioread32(fifo_mem + SVGA_FIFO_BUSY) == 0)) {
 		iowrite32(1, fifo_mem + SVGA_FIFO_BUSY);
 		vmw_write(dev_priv, SVGA_REG_SYNC, reason);
 	}
+}
+
+void vmw_fifo_ping_host(struct vmw_private *dev_priv, uint32_t reason)
+{
+	mutex_lock(&dev_priv->hw_mutex);
+
+	vmw_fifo_ping_host_locked(dev_priv, reason);
 
 	mutex_unlock(&dev_priv->hw_mutex);
 }
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_gmrid_manager.c b/drivers/gpu/drm/vmwgfx/vmwgfx_gmrid_manager.c
index 26f8bdde3529..170b61be1e4e 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_gmrid_manager.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_gmrid_manager.c
@@ -46,8 +46,7 @@ struct vmwgfx_gmrid_man {
 
 static int vmw_gmrid_man_get_node(struct ttm_mem_type_manager *man,
 				  struct ttm_buffer_object *bo,
-				  struct ttm_placement *placement,
-				  uint32_t flags,
+				  const struct ttm_place *place,
 				  struct ttm_mem_reg *mem)
 {
 	struct vmwgfx_gmrid_man *gman =
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c b/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c
index a432c0db257c..026de7cea0f6 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c
@@ -133,6 +133,7 @@ static void vmw_resource_release(struct kref *kref)
 			struct ttm_validate_buffer val_buf;
 
 			val_buf.bo = bo;
+			val_buf.shared = false;
 			res->func->unbind(res, false, &val_buf);
 		}
 		res->backup_dirty = false;
@@ -429,7 +430,7 @@ int vmw_dmabuf_init(struct vmw_private *dev_priv,
 	ret = ttm_bo_init(bdev, &vmw_bo->base, size,
 			  ttm_bo_type_device, placement,
 			  0, interruptible,
-			  NULL, acc_size, NULL, bo_free);
+			  NULL, acc_size, NULL, NULL, bo_free);
 	return ret;
 }
 
@@ -567,13 +568,18 @@ static int vmw_user_dmabuf_synccpu_grab(struct vmw_user_dma_buffer *user_bo,
 	int ret;
 
 	if (flags & drm_vmw_synccpu_allow_cs) {
-		struct ttm_bo_device *bdev = bo->bdev;
+		bool nonblock = !!(flags & drm_vmw_synccpu_dontblock);
+		long lret;
 
-		spin_lock(&bdev->fence_lock);
-		ret = ttm_bo_wait(bo, false, true,
-				  !!(flags & drm_vmw_synccpu_dontblock));
-		spin_unlock(&bdev->fence_lock);
-		return ret;
+		if (nonblock)
+			return reservation_object_test_signaled_rcu(bo->resv, true) ? 0 : -EBUSY;
+
+		lret = reservation_object_wait_timeout_rcu(bo->resv, true, true, MAX_SCHEDULE_TIMEOUT);
+		if (!lret)
+			return -EBUSY;
+		else if (lret < 0)
+			return lret;
+		return 0;
 	}
 
 	ret = ttm_bo_synccpu_write_grab
@@ -1214,8 +1220,9 @@ vmw_resource_check_buffer(struct vmw_resource *res,
 
 	INIT_LIST_HEAD(&val_list);
 	val_buf->bo = ttm_bo_reference(&res->backup->base);
+	val_buf->shared = false;
 	list_add_tail(&val_buf->head, &val_list);
-	ret = ttm_eu_reserve_buffers(NULL, &val_list);
+	ret = ttm_eu_reserve_buffers(NULL, &val_list, interruptible);
 	if (unlikely(ret != 0))
 		goto out_no_reserve;
 
@@ -1307,6 +1314,7 @@ int vmw_resource_do_evict(struct vmw_resource *res, bool interruptible)
 	BUG_ON(!func->may_evict);
 
 	val_buf.bo = NULL;
+	val_buf.shared = false;
 	ret = vmw_resource_check_buffer(res, interruptible, &val_buf);
 	if (unlikely(ret != 0))
 		return ret;
@@ -1352,6 +1360,7 @@ int vmw_resource_validate(struct vmw_resource *res)
 		return 0;
 
 	val_buf.bo = NULL;
+	val_buf.shared = false;
 	if (res->backup)
 		val_buf.bo = &res->backup->base;
 	do {
@@ -1419,25 +1428,16 @@ void vmw_fence_single_bo(struct ttm_buffer_object *bo,
 			 struct vmw_fence_obj *fence)
 {
 	struct ttm_bo_device *bdev = bo->bdev;
-	struct ttm_bo_driver *driver = bdev->driver;
-	struct vmw_fence_obj *old_fence_obj;
+
 	struct vmw_private *dev_priv =
 		container_of(bdev, struct vmw_private, bdev);
 
-	if (fence == NULL)
+	if (fence == NULL) {
 		vmw_execbuf_fence_commands(NULL, dev_priv, &fence, NULL);
-	else
-		driver->sync_obj_ref(fence);
-
-	spin_lock(&bdev->fence_lock);
-
-	old_fence_obj = bo->sync_obj;
-	bo->sync_obj = fence;
-
-	spin_unlock(&bdev->fence_lock);
-
-	if (old_fence_obj)
-		vmw_fence_obj_unreference(&old_fence_obj);
+		reservation_object_add_excl_fence(bo->resv, &fence->base);
+		fence_put(&fence->base);
+	} else
+		reservation_object_add_excl_fence(bo->resv, &fence->base);
 }
 
 /**
@@ -1475,10 +1475,10 @@ void vmw_resource_move_notify(struct ttm_buffer_object *bo,
 
 	if (mem->mem_type != VMW_PL_MOB) {
 		struct vmw_resource *res, *n;
-		struct ttm_bo_device *bdev = bo->bdev;
 		struct ttm_validate_buffer val_buf;
 
 		val_buf.bo = bo;
+		val_buf.shared = false;
 
 		list_for_each_entry_safe(res, n, &dma_buf->res_list, mob_head) {
 
@@ -1491,9 +1491,7 @@ void vmw_resource_move_notify(struct ttm_buffer_object *bo,
 			list_del_init(&res->mob_head);
 		}
 
-		spin_lock(&bdev->fence_lock);
 		(void) ttm_bo_wait(bo, false, false, false);
-		spin_unlock(&bdev->fence_lock);
 	}
 }
 
diff --git a/drivers/gpu/ipu-v3/Kconfig b/drivers/gpu/ipu-v3/Kconfig
index 2f228a2f2a48..aefdff95356d 100644
--- a/drivers/gpu/ipu-v3/Kconfig
+++ b/drivers/gpu/ipu-v3/Kconfig
@@ -1,7 +1,8 @@
 config IMX_IPUV3_CORE
 	tristate "IPUv3 core support"
-	depends on SOC_IMX5 || SOC_IMX6Q || SOC_IMX6SL || ARCH_MULTIPLATFORM
+	depends on SOC_IMX5 || SOC_IMX6Q || ARCH_MULTIPLATFORM
 	depends on RESET_CONTROLLER
+	select GENERIC_IRQ_CHIP
 	help
 	  Choose this if you have a i.MX5/6 system and want to use the Image
 	  Processing Unit. This option only enables IPU base support.
diff --git a/drivers/gpu/ipu-v3/Makefile b/drivers/gpu/ipu-v3/Makefile
index 0b42836caae1..107ec236a4a6 100644
--- a/drivers/gpu/ipu-v3/Makefile
+++ b/drivers/gpu/ipu-v3/Makefile
@@ -1,4 +1,4 @@
 obj-$(CONFIG_IMX_IPUV3_CORE) += imx-ipu-v3.o
 
-imx-ipu-v3-objs := ipu-common.o ipu-cpmem.o ipu-dc.o ipu-di.o \
-		ipu-dp.o ipu-dmfc.o ipu-smfc.o
+imx-ipu-v3-objs := ipu-common.o ipu-cpmem.o ipu-csi.o ipu-dc.o ipu-di.o \
+		ipu-dp.o ipu-dmfc.o ipu-ic.o ipu-smfc.o
diff --git a/drivers/gpu/ipu-v3/ipu-common.c b/drivers/gpu/ipu-v3/ipu-common.c
index 5978e7aab8ed..f707d25ae78f 100644
--- a/drivers/gpu/ipu-v3/ipu-common.c
+++ b/drivers/gpu/ipu-v3/ipu-common.c
@@ -74,6 +74,12 @@ enum ipu_color_space ipu_drm_fourcc_to_colorspace(u32 drm_fourcc)
 	case DRM_FORMAT_UYVY:
 	case DRM_FORMAT_YUV420:
 	case DRM_FORMAT_YVU420:
+	case DRM_FORMAT_YUV422:
+	case DRM_FORMAT_YVU422:
+	case DRM_FORMAT_NV12:
+	case DRM_FORMAT_NV21:
+	case DRM_FORMAT_NV16:
+	case DRM_FORMAT_NV61:
 		return IPUV3_COLORSPACE_YUV;
 	default:
 		return IPUV3_COLORSPACE_UNKNOWN;
@@ -86,8 +92,13 @@ enum ipu_color_space ipu_pixelformat_to_colorspace(u32 pixelformat)
 	switch (pixelformat) {
 	case V4L2_PIX_FMT_YUV420:
 	case V4L2_PIX_FMT_YVU420:
+	case V4L2_PIX_FMT_YUV422P:
 	case V4L2_PIX_FMT_UYVY:
 	case V4L2_PIX_FMT_YUYV:
+	case V4L2_PIX_FMT_NV12:
+	case V4L2_PIX_FMT_NV21:
+	case V4L2_PIX_FMT_NV16:
+	case V4L2_PIX_FMT_NV61:
 		return IPUV3_COLORSPACE_YUV;
 	case V4L2_PIX_FMT_RGB32:
 	case V4L2_PIX_FMT_BGR32:
@@ -101,6 +112,135 @@ enum ipu_color_space ipu_pixelformat_to_colorspace(u32 pixelformat)
 }
 EXPORT_SYMBOL_GPL(ipu_pixelformat_to_colorspace);
 
+bool ipu_pixelformat_is_planar(u32 pixelformat)
+{
+	switch (pixelformat) {
+	case V4L2_PIX_FMT_YUV420:
+	case V4L2_PIX_FMT_YVU420:
+	case V4L2_PIX_FMT_YUV422P:
+	case V4L2_PIX_FMT_NV12:
+	case V4L2_PIX_FMT_NV21:
+	case V4L2_PIX_FMT_NV16:
+	case V4L2_PIX_FMT_NV61:
+		return true;
+	}
+
+	return false;
+}
+EXPORT_SYMBOL_GPL(ipu_pixelformat_is_planar);
+
+enum ipu_color_space ipu_mbus_code_to_colorspace(u32 mbus_code)
+{
+	switch (mbus_code & 0xf000) {
+	case 0x1000:
+		return IPUV3_COLORSPACE_RGB;
+	case 0x2000:
+		return IPUV3_COLORSPACE_YUV;
+	default:
+		return IPUV3_COLORSPACE_UNKNOWN;
+	}
+}
+EXPORT_SYMBOL_GPL(ipu_mbus_code_to_colorspace);
+
+int ipu_stride_to_bytes(u32 pixel_stride, u32 pixelformat)
+{
+	switch (pixelformat) {
+	case V4L2_PIX_FMT_YUV420:
+	case V4L2_PIX_FMT_YVU420:
+	case V4L2_PIX_FMT_YUV422P:
+	case V4L2_PIX_FMT_NV12:
+	case V4L2_PIX_FMT_NV21:
+	case V4L2_PIX_FMT_NV16:
+	case V4L2_PIX_FMT_NV61:
+		/*
+		 * for the planar YUV formats, the stride passed to
+		 * cpmem must be the stride in bytes of the Y plane.
+		 * And all the planar YUV formats have an 8-bit
+		 * Y component.
+		 */
+		return (8 * pixel_stride) >> 3;
+	case V4L2_PIX_FMT_RGB565:
+	case V4L2_PIX_FMT_YUYV:
+	case V4L2_PIX_FMT_UYVY:
+		return (16 * pixel_stride) >> 3;
+	case V4L2_PIX_FMT_BGR24:
+	case V4L2_PIX_FMT_RGB24:
+		return (24 * pixel_stride) >> 3;
+	case V4L2_PIX_FMT_BGR32:
+	case V4L2_PIX_FMT_RGB32:
+		return (32 * pixel_stride) >> 3;
+	default:
+		break;
+	}
+
+	return -EINVAL;
+}
+EXPORT_SYMBOL_GPL(ipu_stride_to_bytes);
+
+int ipu_degrees_to_rot_mode(enum ipu_rotate_mode *mode, int degrees,
+			    bool hflip, bool vflip)
+{
+	u32 r90, vf, hf;
+
+	switch (degrees) {
+	case 0:
+		vf = hf = r90 = 0;
+		break;
+	case 90:
+		vf = hf = 0;
+		r90 = 1;
+		break;
+	case 180:
+		vf = hf = 1;
+		r90 = 0;
+		break;
+	case 270:
+		vf = hf = r90 = 1;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	hf ^= (u32)hflip;
+	vf ^= (u32)vflip;
+
+	*mode = (enum ipu_rotate_mode)((r90 << 2) | (hf << 1) | vf);
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ipu_degrees_to_rot_mode);
+
+int ipu_rot_mode_to_degrees(int *degrees, enum ipu_rotate_mode mode,
+			    bool hflip, bool vflip)
+{
+	u32 r90, vf, hf;
+
+	r90 = ((u32)mode >> 2) & 0x1;
+	hf = ((u32)mode >> 1) & 0x1;
+	vf = ((u32)mode >> 0) & 0x1;
+	hf ^= (u32)hflip;
+	vf ^= (u32)vflip;
+
+	switch ((enum ipu_rotate_mode)((r90 << 2) | (hf << 1) | vf)) {
+	case IPU_ROTATE_NONE:
+		*degrees = 0;
+		break;
+	case IPU_ROTATE_90_RIGHT:
+		*degrees = 90;
+		break;
+	case IPU_ROTATE_180:
+		*degrees = 180;
+		break;
+	case IPU_ROTATE_90_LEFT:
+		*degrees = 270;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ipu_rot_mode_to_degrees);
+
 struct ipuv3_channel *ipu_idmac_get(struct ipu_soc *ipu, unsigned num)
 {
 	struct ipuv3_channel *channel;
@@ -143,7 +283,26 @@ void ipu_idmac_put(struct ipuv3_channel *channel)
 }
 EXPORT_SYMBOL_GPL(ipu_idmac_put);
 
-#define idma_mask(ch)			(1 << (ch & 0x1f))
+#define idma_mask(ch)			(1 << ((ch) & 0x1f))
+
+/*
+ * This is an undocumented feature, a write one to a channel bit in
+ * IPU_CHA_CUR_BUF and IPU_CHA_TRIPLE_CUR_BUF will reset the channel's
+ * internal current buffer pointer so that transfers start from buffer
+ * 0 on the next channel enable (that's the theory anyway, the imx6 TRM
+ * only says these are read-only registers). This operation is required
+ * for channel linking to work correctly, for instance video capture
+ * pipelines that carry out image rotations will fail after the first
+ * streaming unless this function is called for each channel before
+ * re-enabling the channels.
+ */
+static void __ipu_idmac_reset_current_buffer(struct ipuv3_channel *channel)
+{
+	struct ipu_soc *ipu = channel->ipu;
+	unsigned int chno = channel->num;
+
+	ipu_cm_write(ipu, idma_mask(chno), IPU_CHA_CUR_BUF(chno));
+}
 
 void ipu_idmac_set_double_buffer(struct ipuv3_channel *channel,
 		bool doublebuffer)
@@ -161,10 +320,81 @@ void ipu_idmac_set_double_buffer(struct ipuv3_channel *channel,
 		reg &= ~idma_mask(channel->num);
 	ipu_cm_write(ipu, reg, IPU_CHA_DB_MODE_SEL(channel->num));
 
+	__ipu_idmac_reset_current_buffer(channel);
+
 	spin_unlock_irqrestore(&ipu->lock, flags);
 }
 EXPORT_SYMBOL_GPL(ipu_idmac_set_double_buffer);
 
+static const struct {
+	int chnum;
+	u32 reg;
+	int shift;
+} idmac_lock_en_info[] = {
+	{ .chnum =  5, .reg = IDMAC_CH_LOCK_EN_1, .shift =  0, },
+	{ .chnum = 11, .reg = IDMAC_CH_LOCK_EN_1, .shift =  2, },
+	{ .chnum = 12, .reg = IDMAC_CH_LOCK_EN_1, .shift =  4, },
+	{ .chnum = 14, .reg = IDMAC_CH_LOCK_EN_1, .shift =  6, },
+	{ .chnum = 15, .reg = IDMAC_CH_LOCK_EN_1, .shift =  8, },
+	{ .chnum = 20, .reg = IDMAC_CH_LOCK_EN_1, .shift = 10, },
+	{ .chnum = 21, .reg = IDMAC_CH_LOCK_EN_1, .shift = 12, },
+	{ .chnum = 22, .reg = IDMAC_CH_LOCK_EN_1, .shift = 14, },
+	{ .chnum = 23, .reg = IDMAC_CH_LOCK_EN_1, .shift = 16, },
+	{ .chnum = 27, .reg = IDMAC_CH_LOCK_EN_1, .shift = 18, },
+	{ .chnum = 28, .reg = IDMAC_CH_LOCK_EN_1, .shift = 20, },
+	{ .chnum = 45, .reg = IDMAC_CH_LOCK_EN_2, .shift =  0, },
+	{ .chnum = 46, .reg = IDMAC_CH_LOCK_EN_2, .shift =  2, },
+	{ .chnum = 47, .reg = IDMAC_CH_LOCK_EN_2, .shift =  4, },
+	{ .chnum = 48, .reg = IDMAC_CH_LOCK_EN_2, .shift =  6, },
+	{ .chnum = 49, .reg = IDMAC_CH_LOCK_EN_2, .shift =  8, },
+	{ .chnum = 50, .reg = IDMAC_CH_LOCK_EN_2, .shift = 10, },
+};
+
+int ipu_idmac_lock_enable(struct ipuv3_channel *channel, int num_bursts)
+{
+	struct ipu_soc *ipu = channel->ipu;
+	unsigned long flags;
+	u32 bursts, regval;
+	int i;
+
+	switch (num_bursts) {
+	case 0:
+	case 1:
+		bursts = 0x00; /* locking disabled */
+		break;
+	case 2:
+		bursts = 0x01;
+		break;
+	case 4:
+		bursts = 0x02;
+		break;
+	case 8:
+		bursts = 0x03;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	for (i = 0; i < ARRAY_SIZE(idmac_lock_en_info); i++) {
+		if (channel->num == idmac_lock_en_info[i].chnum)
+			break;
+	}
+	if (i >= ARRAY_SIZE(idmac_lock_en_info))
+		return -EINVAL;
+
+	spin_lock_irqsave(&ipu->lock, flags);
+
+	regval = ipu_idmac_read(ipu, idmac_lock_en_info[i].reg);
+	regval &= ~(0x03 << idmac_lock_en_info[i].shift);
+	regval |= (bursts << idmac_lock_en_info[i].shift);
+	ipu_idmac_write(ipu, regval, idmac_lock_en_info[i].reg);
+
+	spin_unlock_irqrestore(&ipu->lock, flags);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ipu_idmac_lock_enable);
+
 int ipu_module_enable(struct ipu_soc *ipu, u32 mask)
 {
 	unsigned long lock_flags;
@@ -217,30 +447,6 @@ int ipu_module_disable(struct ipu_soc *ipu, u32 mask)
 }
 EXPORT_SYMBOL_GPL(ipu_module_disable);
 
-int ipu_csi_enable(struct ipu_soc *ipu, int csi)
-{
-	return ipu_module_enable(ipu, csi ? IPU_CONF_CSI1_EN : IPU_CONF_CSI0_EN);
-}
-EXPORT_SYMBOL_GPL(ipu_csi_enable);
-
-int ipu_csi_disable(struct ipu_soc *ipu, int csi)
-{
-	return ipu_module_disable(ipu, csi ? IPU_CONF_CSI1_EN : IPU_CONF_CSI0_EN);
-}
-EXPORT_SYMBOL_GPL(ipu_csi_disable);
-
-int ipu_smfc_enable(struct ipu_soc *ipu)
-{
-	return ipu_module_enable(ipu, IPU_CONF_SMFC_EN);
-}
-EXPORT_SYMBOL_GPL(ipu_smfc_enable);
-
-int ipu_smfc_disable(struct ipu_soc *ipu)
-{
-	return ipu_module_disable(ipu, IPU_CONF_SMFC_EN);
-}
-EXPORT_SYMBOL_GPL(ipu_smfc_disable);
-
 int ipu_idmac_get_current_buffer(struct ipuv3_channel *channel)
 {
 	struct ipu_soc *ipu = channel->ipu;
@@ -250,6 +456,30 @@ int ipu_idmac_get_current_buffer(struct ipuv3_channel *channel)
 }
 EXPORT_SYMBOL_GPL(ipu_idmac_get_current_buffer);
 
+bool ipu_idmac_buffer_is_ready(struct ipuv3_channel *channel, u32 buf_num)
+{
+	struct ipu_soc *ipu = channel->ipu;
+	unsigned long flags;
+	u32 reg = 0;
+
+	spin_lock_irqsave(&ipu->lock, flags);
+	switch (buf_num) {
+	case 0:
+		reg = ipu_cm_read(ipu, IPU_CHA_BUF0_RDY(channel->num));
+		break;
+	case 1:
+		reg = ipu_cm_read(ipu, IPU_CHA_BUF1_RDY(channel->num));
+		break;
+	case 2:
+		reg = ipu_cm_read(ipu, IPU_CHA_BUF2_RDY(channel->num));
+		break;
+	}
+	spin_unlock_irqrestore(&ipu->lock, flags);
+
+	return ((reg & idma_mask(channel->num)) != 0);
+}
+EXPORT_SYMBOL_GPL(ipu_idmac_buffer_is_ready);
+
 void ipu_idmac_select_buffer(struct ipuv3_channel *channel, u32 buf_num)
 {
 	struct ipu_soc *ipu = channel->ipu;
@@ -268,6 +498,34 @@ void ipu_idmac_select_buffer(struct ipuv3_channel *channel, u32 buf_num)
 }
 EXPORT_SYMBOL_GPL(ipu_idmac_select_buffer);
 
+void ipu_idmac_clear_buffer(struct ipuv3_channel *channel, u32 buf_num)
+{
+	struct ipu_soc *ipu = channel->ipu;
+	unsigned int chno = channel->num;
+	unsigned long flags;
+
+	spin_lock_irqsave(&ipu->lock, flags);
+
+	ipu_cm_write(ipu, 0xF0300000, IPU_GPR); /* write one to clear */
+	switch (buf_num) {
+	case 0:
+		ipu_cm_write(ipu, idma_mask(chno), IPU_CHA_BUF0_RDY(chno));
+		break;
+	case 1:
+		ipu_cm_write(ipu, idma_mask(chno), IPU_CHA_BUF1_RDY(chno));
+		break;
+	case 2:
+		ipu_cm_write(ipu, idma_mask(chno), IPU_CHA_BUF2_RDY(chno));
+		break;
+	default:
+		break;
+	}
+	ipu_cm_write(ipu, 0x0, IPU_GPR); /* write one to set */
+
+	spin_unlock_irqrestore(&ipu->lock, flags);
+}
+EXPORT_SYMBOL_GPL(ipu_idmac_clear_buffer);
+
 int ipu_idmac_enable_channel(struct ipuv3_channel *channel)
 {
 	struct ipu_soc *ipu = channel->ipu;
@@ -338,6 +596,8 @@ int ipu_idmac_disable_channel(struct ipuv3_channel *channel)
 	val &= ~idma_mask(channel->num);
 	ipu_idmac_write(ipu, val, IDMAC_CHA_EN(channel->num));
 
+	__ipu_idmac_reset_current_buffer(channel);
+
 	/* Set channel buffers NOT to be ready */
 	ipu_cm_write(ipu, 0xf0000000, IPU_GPR); /* write one to clear */
 
@@ -366,6 +626,31 @@ int ipu_idmac_disable_channel(struct ipuv3_channel *channel)
 }
 EXPORT_SYMBOL_GPL(ipu_idmac_disable_channel);
 
+/*
+ * The imx6 rev. D TRM says that enabling the WM feature will increase
+ * a channel's priority. Refer to Table 36-8 Calculated priority value.
+ * The sub-module that is the sink or source for the channel must enable
+ * watermark signal for this to take effect (SMFC_WM for instance).
+ */
+void ipu_idmac_enable_watermark(struct ipuv3_channel *channel, bool enable)
+{
+	struct ipu_soc *ipu = channel->ipu;
+	unsigned long flags;
+	u32 val;
+
+	spin_lock_irqsave(&ipu->lock, flags);
+
+	val = ipu_idmac_read(ipu, IDMAC_WM_EN(channel->num));
+	if (enable)
+		val |= 1 << (channel->num % 32);
+	else
+		val &= ~(1 << (channel->num % 32));
+	ipu_idmac_write(ipu, val, IDMAC_WM_EN(channel->num));
+
+	spin_unlock_irqrestore(&ipu->lock, flags);
+}
+EXPORT_SYMBOL_GPL(ipu_idmac_enable_watermark);
+
 static int ipu_memory_reset(struct ipu_soc *ipu)
 {
 	unsigned long timeout;
@@ -382,12 +667,66 @@ static int ipu_memory_reset(struct ipu_soc *ipu)
 	return 0;
 }
 
+/*
+ * Set the source mux for the given CSI. Selects either parallel or
+ * MIPI CSI2 sources.
+ */
+void ipu_set_csi_src_mux(struct ipu_soc *ipu, int csi_id, bool mipi_csi2)
+{
+	unsigned long flags;
+	u32 val, mask;
+
+	mask = (csi_id == 1) ? IPU_CONF_CSI1_DATA_SOURCE :
+		IPU_CONF_CSI0_DATA_SOURCE;
+
+	spin_lock_irqsave(&ipu->lock, flags);
+
+	val = ipu_cm_read(ipu, IPU_CONF);
+	if (mipi_csi2)
+		val |= mask;
+	else
+		val &= ~mask;
+	ipu_cm_write(ipu, val, IPU_CONF);
+
+	spin_unlock_irqrestore(&ipu->lock, flags);
+}
+EXPORT_SYMBOL_GPL(ipu_set_csi_src_mux);
+
+/*
+ * Set the source mux for the IC. Selects either CSI[01] or the VDI.
+ */
+void ipu_set_ic_src_mux(struct ipu_soc *ipu, int csi_id, bool vdi)
+{
+	unsigned long flags;
+	u32 val;
+
+	spin_lock_irqsave(&ipu->lock, flags);
+
+	val = ipu_cm_read(ipu, IPU_CONF);
+	if (vdi) {
+		val |= IPU_CONF_IC_INPUT;
+	} else {
+		val &= ~IPU_CONF_IC_INPUT;
+		if (csi_id == 1)
+			val |= IPU_CONF_CSI_SEL;
+		else
+			val &= ~IPU_CONF_CSI_SEL;
+	}
+	ipu_cm_write(ipu, val, IPU_CONF);
+
+	spin_unlock_irqrestore(&ipu->lock, flags);
+}
+EXPORT_SYMBOL_GPL(ipu_set_ic_src_mux);
+
 struct ipu_devtype {
 	const char *name;
 	unsigned long cm_ofs;
 	unsigned long cpmem_ofs;
 	unsigned long srm_ofs;
 	unsigned long tpm_ofs;
+	unsigned long csi0_ofs;
+	unsigned long csi1_ofs;
+	unsigned long ic_ofs;
 	unsigned long disp0_ofs;
 	unsigned long disp1_ofs;
 	unsigned long dc_tmpl_ofs;
@@ -401,6 +740,9 @@ static struct ipu_devtype ipu_type_imx51 = {
 	.cpmem_ofs = 0x1f000000,
 	.srm_ofs = 0x1f040000,
 	.tpm_ofs = 0x1f060000,
+	.csi0_ofs = 0x1f030000,
+	.csi1_ofs = 0x1f038000,
+	.ic_ofs = 0x1f020000,
 	.disp0_ofs = 0x1e040000,
 	.disp1_ofs = 0x1e048000,
 	.dc_tmpl_ofs = 0x1f080000,
@@ -414,6 +756,9 @@ static struct ipu_devtype ipu_type_imx53 = {
 	.cpmem_ofs = 0x07000000,
 	.srm_ofs = 0x07040000,
 	.tpm_ofs = 0x07060000,
+	.csi0_ofs = 0x07030000,
+	.csi1_ofs = 0x07038000,
+	.ic_ofs = 0x07020000,
 	.disp0_ofs = 0x06040000,
 	.disp1_ofs = 0x06048000,
 	.dc_tmpl_ofs = 0x07080000,
@@ -427,6 +772,9 @@ static struct ipu_devtype ipu_type_imx6q = {
 	.cpmem_ofs = 0x00300000,
 	.srm_ofs = 0x00340000,
 	.tpm_ofs = 0x00360000,
+	.csi0_ofs = 0x00230000,
+	.csi1_ofs = 0x00238000,
+	.ic_ofs = 0x00220000,
 	.disp0_ofs = 0x00240000,
 	.disp1_ofs = 0x00248000,
 	.dc_tmpl_ofs = 0x00380000,
@@ -457,8 +805,30 @@ static int ipu_submodules_init(struct ipu_soc *ipu,
 		goto err_cpmem;
 	}
 
+	ret = ipu_csi_init(ipu, dev, 0, ipu_base + devtype->csi0_ofs,
+			   IPU_CONF_CSI0_EN, ipu_clk);
+	if (ret) {
+		unit = "csi0";
+		goto err_csi_0;
+	}
+
+	ret = ipu_csi_init(ipu, dev, 1, ipu_base + devtype->csi1_ofs,
+			   IPU_CONF_CSI1_EN, ipu_clk);
+	if (ret) {
+		unit = "csi1";
+		goto err_csi_1;
+	}
+
+	ret = ipu_ic_init(ipu, dev,
+			  ipu_base + devtype->ic_ofs,
+			  ipu_base + devtype->tpm_ofs);
+	if (ret) {
+		unit = "ic";
+		goto err_ic;
+	}
+
 	ret = ipu_di_init(ipu, dev, 0, ipu_base + devtype->disp0_ofs,
-			IPU_CONF_DI0_EN, ipu_clk);
+			  IPU_CONF_DI0_EN, ipu_clk);
 	if (ret) {
 		unit = "di0";
 		goto err_di_0;
@@ -511,6 +881,12 @@ err_dc:
 err_di_1:
 	ipu_di_exit(ipu, 0);
 err_di_0:
+	ipu_ic_exit(ipu);
+err_ic:
+	ipu_csi_exit(ipu, 1);
+err_csi_1:
+	ipu_csi_exit(ipu, 0);
+err_csi_0:
 	ipu_cpmem_exit(ipu);
 err_cpmem:
 	dev_err(&pdev->dev, "init %s failed with %d\n", unit, ret);
@@ -589,6 +965,9 @@ static void ipu_submodules_exit(struct ipu_soc *ipu)
 	ipu_dc_exit(ipu);
 	ipu_di_exit(ipu, 1);
 	ipu_di_exit(ipu, 0);
+	ipu_ic_exit(ipu);
+	ipu_csi_exit(ipu, 1);
+	ipu_csi_exit(ipu, 0);
 	ipu_cpmem_exit(ipu);
 }
 
@@ -681,8 +1060,10 @@ static int ipu_add_client_devices(struct ipu_soc *ipu, unsigned long ipu_base)
 				id++, &reg->pdata, sizeof(reg->pdata));
 		}
 
-		if (IS_ERR(pdev))
+		if (IS_ERR(pdev)) {
+			ret = PTR_ERR(pdev);
 			goto err_register;
+		}
 	}
 
 	return 0;
@@ -766,6 +1147,44 @@ static void ipu_irq_exit(struct ipu_soc *ipu)
 	irq_domain_remove(ipu->domain);
 }
 
+void ipu_dump(struct ipu_soc *ipu)
+{
+	int i;
+
+	dev_dbg(ipu->dev, "IPU_CONF = \t0x%08X\n",
+		ipu_cm_read(ipu, IPU_CONF));
+	dev_dbg(ipu->dev, "IDMAC_CONF = \t0x%08X\n",
+		ipu_idmac_read(ipu, IDMAC_CONF));
+	dev_dbg(ipu->dev, "IDMAC_CHA_EN1 = \t0x%08X\n",
+		ipu_idmac_read(ipu, IDMAC_CHA_EN(0)));
+	dev_dbg(ipu->dev, "IDMAC_CHA_EN2 = \t0x%08X\n",
+		ipu_idmac_read(ipu, IDMAC_CHA_EN(32)));
+	dev_dbg(ipu->dev, "IDMAC_CHA_PRI1 = \t0x%08X\n",
+		ipu_idmac_read(ipu, IDMAC_CHA_PRI(0)));
+	dev_dbg(ipu->dev, "IDMAC_CHA_PRI2 = \t0x%08X\n",
+		ipu_idmac_read(ipu, IDMAC_CHA_PRI(32)));
+	dev_dbg(ipu->dev, "IDMAC_BAND_EN1 = \t0x%08X\n",
+		ipu_idmac_read(ipu, IDMAC_BAND_EN(0)));
+	dev_dbg(ipu->dev, "IDMAC_BAND_EN2 = \t0x%08X\n",
+		ipu_idmac_read(ipu, IDMAC_BAND_EN(32)));
+	dev_dbg(ipu->dev, "IPU_CHA_DB_MODE_SEL0 = \t0x%08X\n",
+		ipu_cm_read(ipu, IPU_CHA_DB_MODE_SEL(0)));
+	dev_dbg(ipu->dev, "IPU_CHA_DB_MODE_SEL1 = \t0x%08X\n",
+		ipu_cm_read(ipu, IPU_CHA_DB_MODE_SEL(32)));
+	dev_dbg(ipu->dev, "IPU_FS_PROC_FLOW1 = \t0x%08X\n",
+		ipu_cm_read(ipu, IPU_FS_PROC_FLOW1));
+	dev_dbg(ipu->dev, "IPU_FS_PROC_FLOW2 = \t0x%08X\n",
+		ipu_cm_read(ipu, IPU_FS_PROC_FLOW2));
+	dev_dbg(ipu->dev, "IPU_FS_PROC_FLOW3 = \t0x%08X\n",
+		ipu_cm_read(ipu, IPU_FS_PROC_FLOW3));
+	dev_dbg(ipu->dev, "IPU_FS_DISP_FLOW1 = \t0x%08X\n",
+		ipu_cm_read(ipu, IPU_FS_DISP_FLOW1));
+	for (i = 0; i < 15; i++)
+		dev_dbg(ipu->dev, "IPU_INT_CTRL(%d) = \t%08X\n", i,
+			ipu_cm_read(ipu, IPU_INT_CTRL(i)));
+}
+EXPORT_SYMBOL_GPL(ipu_dump);
+
 static int ipu_probe(struct platform_device *pdev)
 {
 	const struct of_device_id *of_id =
@@ -808,6 +1227,12 @@ static int ipu_probe(struct platform_device *pdev)
 			ipu_base + devtype->cm_ofs + IPU_CM_IDMAC_REG_OFS);
 	dev_dbg(&pdev->dev, "cpmem:    0x%08lx\n",
 			ipu_base + devtype->cpmem_ofs);
+	dev_dbg(&pdev->dev, "csi0:    0x%08lx\n",
+			ipu_base + devtype->csi0_ofs);
+	dev_dbg(&pdev->dev, "csi1:    0x%08lx\n",
+			ipu_base + devtype->csi1_ofs);
+	dev_dbg(&pdev->dev, "ic:      0x%08lx\n",
+			ipu_base + devtype->ic_ofs);
 	dev_dbg(&pdev->dev, "disp0:    0x%08lx\n",
 			ipu_base + devtype->disp0_ofs);
 	dev_dbg(&pdev->dev, "disp1:    0x%08lx\n",
diff --git a/drivers/gpu/ipu-v3/ipu-cpmem.c b/drivers/gpu/ipu-v3/ipu-cpmem.c
index 7adfa78a48bc..3bf05bc4ab67 100644
--- a/drivers/gpu/ipu-v3/ipu-cpmem.c
+++ b/drivers/gpu/ipu-v3/ipu-cpmem.c
@@ -64,6 +64,7 @@ struct ipu_cpmem {
 #define IPU_FIELD_BNDM		IPU_CPMEM_WORD(0, 114, 3)
 #define IPU_FIELD_BM		IPU_CPMEM_WORD(0, 117, 2)
 #define IPU_FIELD_ROT		IPU_CPMEM_WORD(0, 119, 1)
+#define IPU_FIELD_ROT_HF_VF	IPU_CPMEM_WORD(0, 119, 3)
 #define IPU_FIELD_HF		IPU_CPMEM_WORD(0, 120, 1)
 #define IPU_FIELD_VF		IPU_CPMEM_WORD(0, 121, 1)
 #define IPU_FIELD_THE		IPU_CPMEM_WORD(0, 122, 1)
@@ -192,8 +193,14 @@ static int v4l2_pix_fmt_to_drm_fourcc(u32 pixelformat)
 		return DRM_FORMAT_YUYV;
 	case V4L2_PIX_FMT_YUV420:
 		return DRM_FORMAT_YUV420;
+	case V4L2_PIX_FMT_YUV422P:
+		return DRM_FORMAT_YUV422;
 	case V4L2_PIX_FMT_YVU420:
 		return DRM_FORMAT_YVU420;
+	case V4L2_PIX_FMT_NV12:
+		return DRM_FORMAT_NV12;
+	case V4L2_PIX_FMT_NV16:
+		return DRM_FORMAT_NV16;
 	}
 
 	return -EINVAL;
@@ -254,12 +261,34 @@ void ipu_cpmem_interlaced_scan(struct ipuv3_channel *ch, int stride)
 };
 EXPORT_SYMBOL_GPL(ipu_cpmem_interlaced_scan);
 
+void ipu_cpmem_set_axi_id(struct ipuv3_channel *ch, u32 id)
+{
+	id &= 0x3;
+	ipu_ch_param_write_field(ch, IPU_FIELD_ID, id);
+}
+EXPORT_SYMBOL_GPL(ipu_cpmem_set_axi_id);
+
 void ipu_cpmem_set_burstsize(struct ipuv3_channel *ch, int burstsize)
 {
 	ipu_ch_param_write_field(ch, IPU_FIELD_NPB, burstsize - 1);
 };
 EXPORT_SYMBOL_GPL(ipu_cpmem_set_burstsize);
 
+void ipu_cpmem_set_block_mode(struct ipuv3_channel *ch)
+{
+	ipu_ch_param_write_field(ch, IPU_FIELD_BM, 1);
+}
+EXPORT_SYMBOL_GPL(ipu_cpmem_set_block_mode);
+
+void ipu_cpmem_set_rotation(struct ipuv3_channel *ch,
+			    enum ipu_rotate_mode rot)
+{
+	u32 temp_rot = bitrev8(rot) >> 5;
+
+	ipu_ch_param_write_field(ch, IPU_FIELD_ROT_HF_VF, temp_rot);
+}
+EXPORT_SYMBOL_GPL(ipu_cpmem_set_rotation);
+
 int ipu_cpmem_set_format_rgb(struct ipuv3_channel *ch,
 			     const struct ipu_rgb *rgb)
 {
@@ -371,6 +400,7 @@ void ipu_cpmem_set_yuv_planar_full(struct ipuv3_channel *ch,
 {
 	switch (pixel_format) {
 	case V4L2_PIX_FMT_YUV420:
+	case V4L2_PIX_FMT_YUV422P:
 		ipu_ch_param_write_field(ch, IPU_FIELD_SLUV, (stride / 2) - 1);
 		ipu_ch_param_write_field(ch, IPU_FIELD_UBO, u_offset / 8);
 		ipu_ch_param_write_field(ch, IPU_FIELD_VBO, v_offset / 8);
@@ -380,6 +410,12 @@ void ipu_cpmem_set_yuv_planar_full(struct ipuv3_channel *ch,
 		ipu_ch_param_write_field(ch, IPU_FIELD_UBO, v_offset / 8);
 		ipu_ch_param_write_field(ch, IPU_FIELD_VBO, u_offset / 8);
 		break;
+	case V4L2_PIX_FMT_NV12:
+	case V4L2_PIX_FMT_NV16:
+		ipu_ch_param_write_field(ch, IPU_FIELD_SLUV, stride - 1);
+		ipu_ch_param_write_field(ch, IPU_FIELD_UBO, u_offset / 8);
+		ipu_ch_param_write_field(ch, IPU_FIELD_VBO, u_offset / 8);
+		break;
 	}
 }
 EXPORT_SYMBOL_GPL(ipu_cpmem_set_yuv_planar_full);
@@ -399,6 +435,19 @@ void ipu_cpmem_set_yuv_planar(struct ipuv3_channel *ch,
 		ipu_cpmem_set_yuv_planar_full(ch, pixel_format, stride,
 					      u_offset, v_offset);
 		break;
+	case V4L2_PIX_FMT_YUV422P:
+		uv_stride = stride / 2;
+		u_offset = stride * height;
+		v_offset = u_offset + (uv_stride * height);
+		ipu_cpmem_set_yuv_planar_full(ch, pixel_format, stride,
+					      u_offset, v_offset);
+		break;
+	case V4L2_PIX_FMT_NV12:
+	case V4L2_PIX_FMT_NV16:
+		u_offset = stride * height;
+		ipu_cpmem_set_yuv_planar_full(ch, pixel_format, stride,
+					      u_offset, 0);
+		break;
 	}
 }
 EXPORT_SYMBOL_GPL(ipu_cpmem_set_yuv_planar);
@@ -452,11 +501,20 @@ static const struct ipu_rgb def_bgr_16 = {
 };
 
 #define Y_OFFSET(pix, x, y)	((x) + pix->width * (y))
-#define U_OFFSET(pix, x, y)	((pix->width * pix->height) + \
-					(pix->width * (y) / 4) + (x) / 2)
-#define V_OFFSET(pix, x, y)	((pix->width * pix->height) + \
-					(pix->width * pix->height / 4) + \
-					(pix->width * (y) / 4) + (x) / 2)
+#define U_OFFSET(pix, x, y)	((pix->width * pix->height) +		\
+				 (pix->width * (y) / 4) + (x) / 2)
+#define V_OFFSET(pix, x, y)	((pix->width * pix->height) +		\
+				 (pix->width * pix->height / 4) +	\
+				 (pix->width * (y) / 4) + (x) / 2)
+#define U2_OFFSET(pix, x, y)	((pix->width * pix->height) +		\
+				 (pix->width * (y) / 2) + (x) / 2)
+#define V2_OFFSET(pix, x, y)	((pix->width * pix->height) +		\
+				 (pix->width * pix->height / 2) +	\
+				 (pix->width * (y) / 2) + (x) / 2)
+#define UV_OFFSET(pix, x, y)	((pix->width * pix->height) +	\
+				 (pix->width * (y) / 2) + (x))
+#define UV2_OFFSET(pix, x, y)	((pix->width * pix->height) +	\
+				 (pix->width * y) + (x))
 
 int ipu_cpmem_set_fmt(struct ipuv3_channel *ch, u32 drm_fourcc)
 {
@@ -468,6 +526,25 @@ int ipu_cpmem_set_fmt(struct ipuv3_channel *ch, u32 drm_fourcc)
 		/* burst size */
 		ipu_ch_param_write_field(ch, IPU_FIELD_NPB, 31);
 		break;
+	case DRM_FORMAT_YUV422:
+	case DRM_FORMAT_YVU422:
+		/* pix format */
+		ipu_ch_param_write_field(ch, IPU_FIELD_PFS, 1);
+		/* burst size */
+		ipu_ch_param_write_field(ch, IPU_FIELD_NPB, 31);
+		break;
+	case DRM_FORMAT_NV12:
+		/* pix format */
+		ipu_ch_param_write_field(ch, IPU_FIELD_PFS, 4);
+		/* burst size */
+		ipu_ch_param_write_field(ch, IPU_FIELD_NPB, 31);
+		break;
+	case DRM_FORMAT_NV16:
+		/* pix format */
+		ipu_ch_param_write_field(ch, IPU_FIELD_PFS, 3);
+		/* burst size */
+		ipu_ch_param_write_field(ch, IPU_FIELD_NPB, 31);
+		break;
 	case DRM_FORMAT_UYVY:
 		/* bits/pixel */
 		ipu_ch_param_write_field(ch, IPU_FIELD_BPP, 3);
@@ -515,7 +592,7 @@ EXPORT_SYMBOL_GPL(ipu_cpmem_set_fmt);
 int ipu_cpmem_set_image(struct ipuv3_channel *ch, struct ipu_image *image)
 {
 	struct v4l2_pix_format *pix = &image->pix;
-	int y_offset, u_offset, v_offset;
+	int offset, u_offset, v_offset;
 
 	pr_debug("%s: resolution: %dx%d stride: %d\n",
 		 __func__, pix->width, pix->height,
@@ -529,47 +606,137 @@ int ipu_cpmem_set_image(struct ipuv3_channel *ch, struct ipu_image *image)
 	switch (pix->pixelformat) {
 	case V4L2_PIX_FMT_YUV420:
 	case V4L2_PIX_FMT_YVU420:
-		y_offset = Y_OFFSET(pix, image->rect.left, image->rect.top);
+		offset = Y_OFFSET(pix, image->rect.left, image->rect.top);
 		u_offset = U_OFFSET(pix, image->rect.left,
-				    image->rect.top) - y_offset;
+				    image->rect.top) - offset;
 		v_offset = V_OFFSET(pix, image->rect.left,
-				    image->rect.top) - y_offset;
+				    image->rect.top) - offset;
+
+		ipu_cpmem_set_yuv_planar_full(ch, pix->pixelformat,
+					      pix->bytesperline,
+					      u_offset, v_offset);
+		break;
+	case V4L2_PIX_FMT_YUV422P:
+		offset = Y_OFFSET(pix, image->rect.left, image->rect.top);
+		u_offset = U2_OFFSET(pix, image->rect.left,
+				     image->rect.top) - offset;
+		v_offset = V2_OFFSET(pix, image->rect.left,
+				     image->rect.top) - offset;
 
 		ipu_cpmem_set_yuv_planar_full(ch, pix->pixelformat,
-				pix->bytesperline, u_offset, v_offset);
-		ipu_cpmem_set_buffer(ch, 0, image->phys + y_offset);
+					      pix->bytesperline,
+					      u_offset, v_offset);
+		break;
+	case V4L2_PIX_FMT_NV12:
+		offset = Y_OFFSET(pix, image->rect.left, image->rect.top);
+		u_offset = UV_OFFSET(pix, image->rect.left,
+				     image->rect.top) - offset;
+		v_offset = 0;
+
+		ipu_cpmem_set_yuv_planar_full(ch, pix->pixelformat,
+					      pix->bytesperline,
+					      u_offset, v_offset);
+		break;
+	case V4L2_PIX_FMT_NV16:
+		offset = Y_OFFSET(pix, image->rect.left, image->rect.top);
+		u_offset = UV2_OFFSET(pix, image->rect.left,
+				      image->rect.top) - offset;
+		v_offset = 0;
+
+		ipu_cpmem_set_yuv_planar_full(ch, pix->pixelformat,
+					      pix->bytesperline,
+					      u_offset, v_offset);
 		break;
 	case V4L2_PIX_FMT_UYVY:
 	case V4L2_PIX_FMT_YUYV:
-		ipu_cpmem_set_buffer(ch, 0, image->phys +
-				     image->rect.left * 2 +
-				     image->rect.top * image->pix.bytesperline);
+	case V4L2_PIX_FMT_RGB565:
+		offset = image->rect.left * 2 +
+			image->rect.top * pix->bytesperline;
 		break;
 	case V4L2_PIX_FMT_RGB32:
 	case V4L2_PIX_FMT_BGR32:
-		ipu_cpmem_set_buffer(ch, 0, image->phys +
-				     image->rect.left * 4 +
-				     image->rect.top * image->pix.bytesperline);
-		break;
-	case V4L2_PIX_FMT_RGB565:
-		ipu_cpmem_set_buffer(ch, 0, image->phys +
-				     image->rect.left * 2 +
-				     image->rect.top * image->pix.bytesperline);
+		offset = image->rect.left * 4 +
+			image->rect.top * pix->bytesperline;
 		break;
 	case V4L2_PIX_FMT_RGB24:
 	case V4L2_PIX_FMT_BGR24:
-		ipu_cpmem_set_buffer(ch, 0, image->phys +
-				     image->rect.left * 3 +
-				     image->rect.top * image->pix.bytesperline);
+		offset = image->rect.left * 3 +
+			image->rect.top * pix->bytesperline;
 		break;
 	default:
 		return -EINVAL;
 	}
 
+	ipu_cpmem_set_buffer(ch, 0, image->phys0 + offset);
+	ipu_cpmem_set_buffer(ch, 1, image->phys1 + offset);
+
 	return 0;
 }
 EXPORT_SYMBOL_GPL(ipu_cpmem_set_image);
 
+void ipu_cpmem_dump(struct ipuv3_channel *ch)
+{
+	struct ipu_ch_param __iomem *p = ipu_get_cpmem(ch);
+	struct ipu_soc *ipu = ch->ipu;
+	int chno = ch->num;
+
+	dev_dbg(ipu->dev, "ch %d word 0 - %08X %08X %08X %08X %08X\n", chno,
+		readl(&p->word[0].data[0]),
+		readl(&p->word[0].data[1]),
+		readl(&p->word[0].data[2]),
+		readl(&p->word[0].data[3]),
+		readl(&p->word[0].data[4]));
+	dev_dbg(ipu->dev, "ch %d word 1 - %08X %08X %08X %08X %08X\n", chno,
+		readl(&p->word[1].data[0]),
+		readl(&p->word[1].data[1]),
+		readl(&p->word[1].data[2]),
+		readl(&p->word[1].data[3]),
+		readl(&p->word[1].data[4]));
+	dev_dbg(ipu->dev, "PFS 0x%x, ",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_PFS));
+	dev_dbg(ipu->dev, "BPP 0x%x, ",
+		ipu_ch_param_read_field(ch, IPU_FIELD_BPP));
+	dev_dbg(ipu->dev, "NPB 0x%x\n",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_NPB));
+
+	dev_dbg(ipu->dev, "FW %d, ",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_FW));
+	dev_dbg(ipu->dev, "FH %d, ",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_FH));
+	dev_dbg(ipu->dev, "EBA0 0x%x\n",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_EBA0) << 3);
+	dev_dbg(ipu->dev, "EBA1 0x%x\n",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_EBA1) << 3);
+	dev_dbg(ipu->dev, "Stride %d\n",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_SL));
+	dev_dbg(ipu->dev, "scan_order %d\n",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_SO));
+	dev_dbg(ipu->dev, "uv_stride %d\n",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_SLUV));
+	dev_dbg(ipu->dev, "u_offset 0x%x\n",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_UBO) << 3);
+	dev_dbg(ipu->dev, "v_offset 0x%x\n",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_VBO) << 3);
+
+	dev_dbg(ipu->dev, "Width0 %d+1, ",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_WID0));
+	dev_dbg(ipu->dev, "Width1 %d+1, ",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_WID1));
+	dev_dbg(ipu->dev, "Width2 %d+1, ",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_WID2));
+	dev_dbg(ipu->dev, "Width3 %d+1, ",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_WID3));
+	dev_dbg(ipu->dev, "Offset0 %d, ",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_OFS0));
+	dev_dbg(ipu->dev, "Offset1 %d, ",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_OFS1));
+	dev_dbg(ipu->dev, "Offset2 %d, ",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_OFS2));
+	dev_dbg(ipu->dev, "Offset3 %d\n",
+		 ipu_ch_param_read_field(ch, IPU_FIELD_OFS3));
+}
+EXPORT_SYMBOL_GPL(ipu_cpmem_dump);
+
 int ipu_cpmem_init(struct ipu_soc *ipu, struct device *dev, unsigned long base)
 {
 	struct ipu_cpmem *cpmem;
diff --git a/drivers/gpu/ipu-v3/ipu-csi.c b/drivers/gpu/ipu-v3/ipu-csi.c
new file mode 100644
index 000000000000..d6f56471bd2a
--- /dev/null
+++ b/drivers/gpu/ipu-v3/ipu-csi.c
@@ -0,0 +1,741 @@
+/*
+ * Copyright (C) 2012-2014 Mentor Graphics Inc.
+ * Copyright (C) 2005-2009 Freescale Semiconductor, Inc.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the
+ * Free Software Foundation; either version 2 of the License, or (at your
+ * option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
+ * or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
+ * for more details.
+ */
+#include <linux/export.h>
+#include <linux/module.h>
+#include <linux/types.h>
+#include <linux/errno.h>
+#include <linux/delay.h>
+#include <linux/io.h>
+#include <linux/err.h>
+#include <linux/platform_device.h>
+#include <linux/videodev2.h>
+#include <uapi/linux/v4l2-mediabus.h>
+#include <linux/clk.h>
+#include <linux/clk-provider.h>
+#include <linux/clkdev.h>
+
+#include "ipu-prv.h"
+
+struct ipu_csi {
+	void __iomem *base;
+	int id;
+	u32 module;
+	struct clk *clk_ipu;	/* IPU bus clock */
+	spinlock_t lock;
+	bool inuse;
+	struct ipu_soc *ipu;
+};
+
+/* CSI Register Offsets */
+#define CSI_SENS_CONF		0x0000
+#define CSI_SENS_FRM_SIZE	0x0004
+#define CSI_ACT_FRM_SIZE	0x0008
+#define CSI_OUT_FRM_CTRL	0x000c
+#define CSI_TST_CTRL		0x0010
+#define CSI_CCIR_CODE_1		0x0014
+#define CSI_CCIR_CODE_2		0x0018
+#define CSI_CCIR_CODE_3		0x001c
+#define CSI_MIPI_DI		0x0020
+#define CSI_SKIP		0x0024
+#define CSI_CPD_CTRL		0x0028
+#define CSI_CPD_RC(n)		(0x002c + ((n)*4))
+#define CSI_CPD_RS(n)		(0x004c + ((n)*4))
+#define CSI_CPD_GRC(n)		(0x005c + ((n)*4))
+#define CSI_CPD_GRS(n)		(0x007c + ((n)*4))
+#define CSI_CPD_GBC(n)		(0x008c + ((n)*4))
+#define CSI_CPD_GBS(n)		(0x00Ac + ((n)*4))
+#define CSI_CPD_BC(n)		(0x00Bc + ((n)*4))
+#define CSI_CPD_BS(n)		(0x00Dc + ((n)*4))
+#define CSI_CPD_OFFSET1		0x00ec
+#define CSI_CPD_OFFSET2		0x00f0
+
+/* CSI Register Fields */
+#define CSI_SENS_CONF_DATA_FMT_SHIFT		8
+#define CSI_SENS_CONF_DATA_FMT_MASK		0x00000700
+#define CSI_SENS_CONF_DATA_FMT_RGB_YUV444	0L
+#define CSI_SENS_CONF_DATA_FMT_YUV422_YUYV	1L
+#define CSI_SENS_CONF_DATA_FMT_YUV422_UYVY	2L
+#define CSI_SENS_CONF_DATA_FMT_BAYER		3L
+#define CSI_SENS_CONF_DATA_FMT_RGB565		4L
+#define CSI_SENS_CONF_DATA_FMT_RGB555		5L
+#define CSI_SENS_CONF_DATA_FMT_RGB444		6L
+#define CSI_SENS_CONF_DATA_FMT_JPEG		7L
+
+#define CSI_SENS_CONF_VSYNC_POL_SHIFT		0
+#define CSI_SENS_CONF_HSYNC_POL_SHIFT		1
+#define CSI_SENS_CONF_DATA_POL_SHIFT		2
+#define CSI_SENS_CONF_PIX_CLK_POL_SHIFT		3
+#define CSI_SENS_CONF_SENS_PRTCL_MASK		0x00000070
+#define CSI_SENS_CONF_SENS_PRTCL_SHIFT		4
+#define CSI_SENS_CONF_PACK_TIGHT_SHIFT		7
+#define CSI_SENS_CONF_DATA_WIDTH_SHIFT		11
+#define CSI_SENS_CONF_EXT_VSYNC_SHIFT		15
+#define CSI_SENS_CONF_DIVRATIO_SHIFT		16
+
+#define CSI_SENS_CONF_DIVRATIO_MASK		0x00ff0000
+#define CSI_SENS_CONF_DATA_DEST_SHIFT		24
+#define CSI_SENS_CONF_DATA_DEST_MASK		0x07000000
+#define CSI_SENS_CONF_JPEG8_EN_SHIFT		27
+#define CSI_SENS_CONF_JPEG_EN_SHIFT		28
+#define CSI_SENS_CONF_FORCE_EOF_SHIFT		29
+#define CSI_SENS_CONF_DATA_EN_POL_SHIFT		31
+
+#define CSI_DATA_DEST_IC			2
+#define CSI_DATA_DEST_IDMAC			4
+
+#define CSI_CCIR_ERR_DET_EN			0x01000000
+#define CSI_HORI_DOWNSIZE_EN			0x80000000
+#define CSI_VERT_DOWNSIZE_EN			0x40000000
+#define CSI_TEST_GEN_MODE_EN			0x01000000
+
+#define CSI_HSC_MASK				0x1fff0000
+#define CSI_HSC_SHIFT				16
+#define CSI_VSC_MASK				0x00000fff
+#define CSI_VSC_SHIFT				0
+
+#define CSI_TEST_GEN_R_MASK			0x000000ff
+#define CSI_TEST_GEN_R_SHIFT			0
+#define CSI_TEST_GEN_G_MASK			0x0000ff00
+#define CSI_TEST_GEN_G_SHIFT			8
+#define CSI_TEST_GEN_B_MASK			0x00ff0000
+#define CSI_TEST_GEN_B_SHIFT			16
+
+#define CSI_MAX_RATIO_SKIP_SMFC_MASK		0x00000007
+#define CSI_MAX_RATIO_SKIP_SMFC_SHIFT		0
+#define CSI_SKIP_SMFC_MASK			0x000000f8
+#define CSI_SKIP_SMFC_SHIFT			3
+#define CSI_ID_2_SKIP_MASK			0x00000300
+#define CSI_ID_2_SKIP_SHIFT			8
+
+#define CSI_COLOR_FIRST_ROW_MASK		0x00000002
+#define CSI_COLOR_FIRST_COMP_MASK		0x00000001
+
+/* MIPI CSI-2 data types */
+#define MIPI_DT_YUV420		0x18 /* YYY.../UYVY.... */
+#define MIPI_DT_YUV420_LEGACY	0x1a /* UYY.../VYY...   */
+#define MIPI_DT_YUV422		0x1e /* UYVY...         */
+#define MIPI_DT_RGB444		0x20
+#define MIPI_DT_RGB555		0x21
+#define MIPI_DT_RGB565		0x22
+#define MIPI_DT_RGB666		0x23
+#define MIPI_DT_RGB888		0x24
+#define MIPI_DT_RAW6		0x28
+#define MIPI_DT_RAW7		0x29
+#define MIPI_DT_RAW8		0x2a
+#define MIPI_DT_RAW10		0x2b
+#define MIPI_DT_RAW12		0x2c
+#define MIPI_DT_RAW14		0x2d
+
+/*
+ * Bitfield of CSI bus signal polarities and modes.
+ */
+struct ipu_csi_bus_config {
+	unsigned data_width:4;
+	unsigned clk_mode:3;
+	unsigned ext_vsync:1;
+	unsigned vsync_pol:1;
+	unsigned hsync_pol:1;
+	unsigned pixclk_pol:1;
+	unsigned data_pol:1;
+	unsigned sens_clksrc:1;
+	unsigned pack_tight:1;
+	unsigned force_eof:1;
+	unsigned data_en_pol:1;
+
+	unsigned data_fmt;
+	unsigned mipi_dt;
+};
+
+/*
+ * Enumeration of CSI data bus widths.
+ */
+enum ipu_csi_data_width {
+	IPU_CSI_DATA_WIDTH_4   = 0,
+	IPU_CSI_DATA_WIDTH_8   = 1,
+	IPU_CSI_DATA_WIDTH_10  = 3,
+	IPU_CSI_DATA_WIDTH_12  = 5,
+	IPU_CSI_DATA_WIDTH_16  = 9,
+};
+
+/*
+ * Enumeration of CSI clock modes.
+ */
+enum ipu_csi_clk_mode {
+	IPU_CSI_CLK_MODE_GATED_CLK,
+	IPU_CSI_CLK_MODE_NONGATED_CLK,
+	IPU_CSI_CLK_MODE_CCIR656_PROGRESSIVE,
+	IPU_CSI_CLK_MODE_CCIR656_INTERLACED,
+	IPU_CSI_CLK_MODE_CCIR1120_PROGRESSIVE_DDR,
+	IPU_CSI_CLK_MODE_CCIR1120_PROGRESSIVE_SDR,
+	IPU_CSI_CLK_MODE_CCIR1120_INTERLACED_DDR,
+	IPU_CSI_CLK_MODE_CCIR1120_INTERLACED_SDR,
+};
+
+static inline u32 ipu_csi_read(struct ipu_csi *csi, unsigned offset)
+{
+	return readl(csi->base + offset);
+}
+
+static inline void ipu_csi_write(struct ipu_csi *csi, u32 value,
+				 unsigned offset)
+{
+	writel(value, csi->base + offset);
+}
+
+/*
+ * Set mclk division ratio for generating test mode mclk. Only used
+ * for test generator.
+ */
+static int ipu_csi_set_testgen_mclk(struct ipu_csi *csi, u32 pixel_clk,
+					u32 ipu_clk)
+{
+	u32 temp;
+	u32 div_ratio;
+
+	div_ratio = (ipu_clk / pixel_clk) - 1;
+
+	if (div_ratio > 0xFF || div_ratio < 0) {
+		dev_err(csi->ipu->dev,
+			"value of pixel_clk extends normal range\n");
+		return -EINVAL;
+	}
+
+	temp = ipu_csi_read(csi, CSI_SENS_CONF);
+	temp &= ~CSI_SENS_CONF_DIVRATIO_MASK;
+	ipu_csi_write(csi, temp | (div_ratio << CSI_SENS_CONF_DIVRATIO_SHIFT),
+			  CSI_SENS_CONF);
+
+	return 0;
+}
+
+/*
+ * Find the CSI data format and data width for the given V4L2 media
+ * bus pixel format code.
+ */
+static int mbus_code_to_bus_cfg(struct ipu_csi_bus_config *cfg, u32 mbus_code)
+{
+	switch (mbus_code) {
+	case V4L2_MBUS_FMT_BGR565_2X8_BE:
+	case V4L2_MBUS_FMT_BGR565_2X8_LE:
+	case V4L2_MBUS_FMT_RGB565_2X8_BE:
+	case V4L2_MBUS_FMT_RGB565_2X8_LE:
+		cfg->data_fmt = CSI_SENS_CONF_DATA_FMT_RGB565;
+		cfg->mipi_dt = MIPI_DT_RGB565;
+		cfg->data_width = IPU_CSI_DATA_WIDTH_8;
+		break;
+	case V4L2_MBUS_FMT_RGB444_2X8_PADHI_BE:
+	case V4L2_MBUS_FMT_RGB444_2X8_PADHI_LE:
+		cfg->data_fmt = CSI_SENS_CONF_DATA_FMT_RGB444;
+		cfg->mipi_dt = MIPI_DT_RGB444;
+		cfg->data_width = IPU_CSI_DATA_WIDTH_8;
+		break;
+	case V4L2_MBUS_FMT_RGB555_2X8_PADHI_BE:
+	case V4L2_MBUS_FMT_RGB555_2X8_PADHI_LE:
+		cfg->data_fmt = CSI_SENS_CONF_DATA_FMT_RGB555;
+		cfg->mipi_dt = MIPI_DT_RGB555;
+		cfg->data_width = IPU_CSI_DATA_WIDTH_8;
+		break;
+	case V4L2_MBUS_FMT_UYVY8_2X8:
+		cfg->data_fmt = CSI_SENS_CONF_DATA_FMT_YUV422_UYVY;
+		cfg->mipi_dt = MIPI_DT_YUV422;
+		cfg->data_width = IPU_CSI_DATA_WIDTH_8;
+		break;
+	case V4L2_MBUS_FMT_YUYV8_2X8:
+		cfg->data_fmt = CSI_SENS_CONF_DATA_FMT_YUV422_YUYV;
+		cfg->mipi_dt = MIPI_DT_YUV422;
+		cfg->data_width = IPU_CSI_DATA_WIDTH_8;
+		break;
+	case V4L2_MBUS_FMT_UYVY8_1X16:
+		cfg->data_fmt = CSI_SENS_CONF_DATA_FMT_YUV422_UYVY;
+		cfg->mipi_dt = MIPI_DT_YUV422;
+		cfg->data_width = IPU_CSI_DATA_WIDTH_16;
+		break;
+	case V4L2_MBUS_FMT_YUYV8_1X16:
+		cfg->data_fmt = CSI_SENS_CONF_DATA_FMT_YUV422_YUYV;
+		cfg->mipi_dt = MIPI_DT_YUV422;
+		cfg->data_width = IPU_CSI_DATA_WIDTH_16;
+		break;
+	case V4L2_MBUS_FMT_SBGGR8_1X8:
+	case V4L2_MBUS_FMT_SGBRG8_1X8:
+	case V4L2_MBUS_FMT_SGRBG8_1X8:
+	case V4L2_MBUS_FMT_SRGGB8_1X8:
+		cfg->data_fmt = CSI_SENS_CONF_DATA_FMT_BAYER;
+		cfg->mipi_dt = MIPI_DT_RAW8;
+		cfg->data_width = IPU_CSI_DATA_WIDTH_8;
+		break;
+	case V4L2_MBUS_FMT_SBGGR10_DPCM8_1X8:
+	case V4L2_MBUS_FMT_SGBRG10_DPCM8_1X8:
+	case V4L2_MBUS_FMT_SGRBG10_DPCM8_1X8:
+	case V4L2_MBUS_FMT_SRGGB10_DPCM8_1X8:
+	case V4L2_MBUS_FMT_SBGGR10_2X8_PADHI_BE:
+	case V4L2_MBUS_FMT_SBGGR10_2X8_PADHI_LE:
+	case V4L2_MBUS_FMT_SBGGR10_2X8_PADLO_BE:
+	case V4L2_MBUS_FMT_SBGGR10_2X8_PADLO_LE:
+		cfg->data_fmt = CSI_SENS_CONF_DATA_FMT_BAYER;
+		cfg->mipi_dt = MIPI_DT_RAW10;
+		cfg->data_width = IPU_CSI_DATA_WIDTH_8;
+		break;
+	case V4L2_MBUS_FMT_SBGGR10_1X10:
+	case V4L2_MBUS_FMT_SGBRG10_1X10:
+	case V4L2_MBUS_FMT_SGRBG10_1X10:
+	case V4L2_MBUS_FMT_SRGGB10_1X10:
+		cfg->data_fmt = CSI_SENS_CONF_DATA_FMT_BAYER;
+		cfg->mipi_dt = MIPI_DT_RAW10;
+		cfg->data_width = IPU_CSI_DATA_WIDTH_10;
+		break;
+	case V4L2_MBUS_FMT_SBGGR12_1X12:
+	case V4L2_MBUS_FMT_SGBRG12_1X12:
+	case V4L2_MBUS_FMT_SGRBG12_1X12:
+	case V4L2_MBUS_FMT_SRGGB12_1X12:
+		cfg->data_fmt = CSI_SENS_CONF_DATA_FMT_BAYER;
+		cfg->mipi_dt = MIPI_DT_RAW12;
+		cfg->data_width = IPU_CSI_DATA_WIDTH_12;
+		break;
+	case V4L2_MBUS_FMT_JPEG_1X8:
+		/* TODO */
+		cfg->data_fmt = CSI_SENS_CONF_DATA_FMT_JPEG;
+		cfg->mipi_dt = MIPI_DT_RAW8;
+		cfg->data_width = IPU_CSI_DATA_WIDTH_8;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+/*
+ * Fill a CSI bus config struct from mbus_config and mbus_framefmt.
+ */
+static void fill_csi_bus_cfg(struct ipu_csi_bus_config *csicfg,
+				 struct v4l2_mbus_config *mbus_cfg,
+				 struct v4l2_mbus_framefmt *mbus_fmt)
+{
+	memset(csicfg, 0, sizeof(*csicfg));
+
+	mbus_code_to_bus_cfg(csicfg, mbus_fmt->code);
+
+	switch (mbus_cfg->type) {
+	case V4L2_MBUS_PARALLEL:
+		csicfg->ext_vsync = 1;
+		csicfg->vsync_pol = (mbus_cfg->flags &
+				     V4L2_MBUS_VSYNC_ACTIVE_LOW) ? 1 : 0;
+		csicfg->hsync_pol = (mbus_cfg->flags &
+				     V4L2_MBUS_HSYNC_ACTIVE_LOW) ? 1 : 0;
+		csicfg->pixclk_pol = (mbus_cfg->flags &
+				      V4L2_MBUS_PCLK_SAMPLE_FALLING) ? 1 : 0;
+		csicfg->clk_mode = IPU_CSI_CLK_MODE_GATED_CLK;
+		break;
+	case V4L2_MBUS_BT656:
+		csicfg->ext_vsync = 0;
+		if (V4L2_FIELD_HAS_BOTH(mbus_fmt->field))
+			csicfg->clk_mode = IPU_CSI_CLK_MODE_CCIR656_INTERLACED;
+		else
+			csicfg->clk_mode = IPU_CSI_CLK_MODE_CCIR656_PROGRESSIVE;
+		break;
+	case V4L2_MBUS_CSI2:
+		/*
+		 * MIPI CSI-2 requires non gated clock mode, all other
+		 * parameters are not applicable for MIPI CSI-2 bus.
+		 */
+		csicfg->clk_mode = IPU_CSI_CLK_MODE_NONGATED_CLK;
+		break;
+	default:
+		/* will never get here, keep compiler quiet */
+		break;
+	}
+}
+
+int ipu_csi_init_interface(struct ipu_csi *csi,
+			   struct v4l2_mbus_config *mbus_cfg,
+			   struct v4l2_mbus_framefmt *mbus_fmt)
+{
+	struct ipu_csi_bus_config cfg;
+	unsigned long flags;
+	u32 data = 0;
+
+	fill_csi_bus_cfg(&cfg, mbus_cfg, mbus_fmt);
+
+	/* Set the CSI_SENS_CONF register remaining fields */
+	data |= cfg.data_width << CSI_SENS_CONF_DATA_WIDTH_SHIFT |
+		cfg.data_fmt << CSI_SENS_CONF_DATA_FMT_SHIFT |
+		cfg.data_pol << CSI_SENS_CONF_DATA_POL_SHIFT |
+		cfg.vsync_pol << CSI_SENS_CONF_VSYNC_POL_SHIFT |
+		cfg.hsync_pol << CSI_SENS_CONF_HSYNC_POL_SHIFT |
+		cfg.pixclk_pol << CSI_SENS_CONF_PIX_CLK_POL_SHIFT |
+		cfg.ext_vsync << CSI_SENS_CONF_EXT_VSYNC_SHIFT |
+		cfg.clk_mode << CSI_SENS_CONF_SENS_PRTCL_SHIFT |
+		cfg.pack_tight << CSI_SENS_CONF_PACK_TIGHT_SHIFT |
+		cfg.force_eof << CSI_SENS_CONF_FORCE_EOF_SHIFT |
+		cfg.data_en_pol << CSI_SENS_CONF_DATA_EN_POL_SHIFT;
+
+	spin_lock_irqsave(&csi->lock, flags);
+
+	ipu_csi_write(csi, data, CSI_SENS_CONF);
+
+	/* Setup sensor frame size */
+	ipu_csi_write(csi,
+		      (mbus_fmt->width - 1) | ((mbus_fmt->height - 1) << 16),
+		      CSI_SENS_FRM_SIZE);
+
+	/* Set CCIR registers */
+
+	switch (cfg.clk_mode) {
+	case IPU_CSI_CLK_MODE_CCIR656_PROGRESSIVE:
+		ipu_csi_write(csi, 0x40030, CSI_CCIR_CODE_1);
+		ipu_csi_write(csi, 0xFF0000, CSI_CCIR_CODE_3);
+		break;
+	case IPU_CSI_CLK_MODE_CCIR656_INTERLACED:
+		if (mbus_fmt->width == 720 && mbus_fmt->height == 576) {
+			/*
+			 * PAL case
+			 *
+			 * Field0BlankEnd = 0x6, Field0BlankStart = 0x2,
+			 * Field0ActiveEnd = 0x4, Field0ActiveStart = 0
+			 * Field1BlankEnd = 0x7, Field1BlankStart = 0x3,
+			 * Field1ActiveEnd = 0x5, Field1ActiveStart = 0x1
+			 */
+			ipu_csi_write(csi, 0x40596 | CSI_CCIR_ERR_DET_EN,
+					  CSI_CCIR_CODE_1);
+			ipu_csi_write(csi, 0xD07DF, CSI_CCIR_CODE_2);
+			ipu_csi_write(csi, 0xFF0000, CSI_CCIR_CODE_3);
+
+		} else if (mbus_fmt->width == 720 && mbus_fmt->height == 480) {
+			/*
+			 * NTSC case
+			 *
+			 * Field0BlankEnd = 0x7, Field0BlankStart = 0x3,
+			 * Field0ActiveEnd = 0x5, Field0ActiveStart = 0x1
+			 * Field1BlankEnd = 0x6, Field1BlankStart = 0x2,
+			 * Field1ActiveEnd = 0x4, Field1ActiveStart = 0
+			 */
+			ipu_csi_write(csi, 0xD07DF | CSI_CCIR_ERR_DET_EN,
+					  CSI_CCIR_CODE_1);
+			ipu_csi_write(csi, 0x40596, CSI_CCIR_CODE_2);
+			ipu_csi_write(csi, 0xFF0000, CSI_CCIR_CODE_3);
+		} else {
+			dev_err(csi->ipu->dev,
+				"Unsupported CCIR656 interlaced video mode\n");
+			spin_unlock_irqrestore(&csi->lock, flags);
+			return -EINVAL;
+		}
+		break;
+	case IPU_CSI_CLK_MODE_CCIR1120_PROGRESSIVE_DDR:
+	case IPU_CSI_CLK_MODE_CCIR1120_PROGRESSIVE_SDR:
+	case IPU_CSI_CLK_MODE_CCIR1120_INTERLACED_DDR:
+	case IPU_CSI_CLK_MODE_CCIR1120_INTERLACED_SDR:
+		ipu_csi_write(csi, 0x40030 | CSI_CCIR_ERR_DET_EN,
+				   CSI_CCIR_CODE_1);
+		ipu_csi_write(csi, 0xFF0000, CSI_CCIR_CODE_3);
+		break;
+	case IPU_CSI_CLK_MODE_GATED_CLK:
+	case IPU_CSI_CLK_MODE_NONGATED_CLK:
+		ipu_csi_write(csi, 0, CSI_CCIR_CODE_1);
+		break;
+	}
+
+	dev_dbg(csi->ipu->dev, "CSI_SENS_CONF = 0x%08X\n",
+		ipu_csi_read(csi, CSI_SENS_CONF));
+	dev_dbg(csi->ipu->dev, "CSI_ACT_FRM_SIZE = 0x%08X\n",
+		ipu_csi_read(csi, CSI_ACT_FRM_SIZE));
+
+	spin_unlock_irqrestore(&csi->lock, flags);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ipu_csi_init_interface);
+
+bool ipu_csi_is_interlaced(struct ipu_csi *csi)
+{
+	unsigned long flags;
+	u32 sensor_protocol;
+
+	spin_lock_irqsave(&csi->lock, flags);
+	sensor_protocol =
+		(ipu_csi_read(csi, CSI_SENS_CONF) &
+		 CSI_SENS_CONF_SENS_PRTCL_MASK) >>
+		CSI_SENS_CONF_SENS_PRTCL_SHIFT;
+	spin_unlock_irqrestore(&csi->lock, flags);
+
+	switch (sensor_protocol) {
+	case IPU_CSI_CLK_MODE_GATED_CLK:
+	case IPU_CSI_CLK_MODE_NONGATED_CLK:
+	case IPU_CSI_CLK_MODE_CCIR656_PROGRESSIVE:
+	case IPU_CSI_CLK_MODE_CCIR1120_PROGRESSIVE_DDR:
+	case IPU_CSI_CLK_MODE_CCIR1120_PROGRESSIVE_SDR:
+		return false;
+	case IPU_CSI_CLK_MODE_CCIR656_INTERLACED:
+	case IPU_CSI_CLK_MODE_CCIR1120_INTERLACED_DDR:
+	case IPU_CSI_CLK_MODE_CCIR1120_INTERLACED_SDR:
+		return true;
+	default:
+		dev_err(csi->ipu->dev,
+			"CSI %d sensor protocol unsupported\n", csi->id);
+		return false;
+	}
+}
+EXPORT_SYMBOL_GPL(ipu_csi_is_interlaced);
+
+void ipu_csi_get_window(struct ipu_csi *csi, struct v4l2_rect *w)
+{
+	unsigned long flags;
+	u32 reg;
+
+	spin_lock_irqsave(&csi->lock, flags);
+
+	reg = ipu_csi_read(csi, CSI_ACT_FRM_SIZE);
+	w->width = (reg & 0xFFFF) + 1;
+	w->height = (reg >> 16 & 0xFFFF) + 1;
+
+	reg = ipu_csi_read(csi, CSI_OUT_FRM_CTRL);
+	w->left = (reg & CSI_HSC_MASK) >> CSI_HSC_SHIFT;
+	w->top = (reg & CSI_VSC_MASK) >> CSI_VSC_SHIFT;
+
+	spin_unlock_irqrestore(&csi->lock, flags);
+}
+EXPORT_SYMBOL_GPL(ipu_csi_get_window);
+
+void ipu_csi_set_window(struct ipu_csi *csi, struct v4l2_rect *w)
+{
+	unsigned long flags;
+	u32 reg;
+
+	spin_lock_irqsave(&csi->lock, flags);
+
+	ipu_csi_write(csi, (w->width - 1) | ((w->height - 1) << 16),
+			  CSI_ACT_FRM_SIZE);
+
+	reg = ipu_csi_read(csi, CSI_OUT_FRM_CTRL);
+	reg &= ~(CSI_HSC_MASK | CSI_VSC_MASK);
+	reg |= ((w->top << CSI_VSC_SHIFT) | (w->left << CSI_HSC_SHIFT));
+	ipu_csi_write(csi, reg, CSI_OUT_FRM_CTRL);
+
+	spin_unlock_irqrestore(&csi->lock, flags);
+}
+EXPORT_SYMBOL_GPL(ipu_csi_set_window);
+
+void ipu_csi_set_test_generator(struct ipu_csi *csi, bool active,
+				u32 r_value, u32 g_value, u32 b_value,
+				u32 pix_clk)
+{
+	unsigned long flags;
+	u32 ipu_clk = clk_get_rate(csi->clk_ipu);
+	u32 temp;
+
+	spin_lock_irqsave(&csi->lock, flags);
+
+	temp = ipu_csi_read(csi, CSI_TST_CTRL);
+
+	if (active == false) {
+		temp &= ~CSI_TEST_GEN_MODE_EN;
+		ipu_csi_write(csi, temp, CSI_TST_CTRL);
+	} else {
+		/* Set sensb_mclk div_ratio */
+		ipu_csi_set_testgen_mclk(csi, pix_clk, ipu_clk);
+
+		temp &= ~(CSI_TEST_GEN_R_MASK | CSI_TEST_GEN_G_MASK |
+			  CSI_TEST_GEN_B_MASK);
+		temp |= CSI_TEST_GEN_MODE_EN;
+		temp |= (r_value << CSI_TEST_GEN_R_SHIFT) |
+			(g_value << CSI_TEST_GEN_G_SHIFT) |
+			(b_value << CSI_TEST_GEN_B_SHIFT);
+		ipu_csi_write(csi, temp, CSI_TST_CTRL);
+	}
+
+	spin_unlock_irqrestore(&csi->lock, flags);
+}
+EXPORT_SYMBOL_GPL(ipu_csi_set_test_generator);
+
+int ipu_csi_set_mipi_datatype(struct ipu_csi *csi, u32 vc,
+			      struct v4l2_mbus_framefmt *mbus_fmt)
+{
+	struct ipu_csi_bus_config cfg;
+	unsigned long flags;
+	u32 temp;
+
+	if (vc > 3)
+		return -EINVAL;
+
+	mbus_code_to_bus_cfg(&cfg, mbus_fmt->code);
+
+	spin_lock_irqsave(&csi->lock, flags);
+
+	temp = ipu_csi_read(csi, CSI_MIPI_DI);
+	temp &= ~(0xff << (vc * 8));
+	temp |= (cfg.mipi_dt << (vc * 8));
+	ipu_csi_write(csi, temp, CSI_MIPI_DI);
+
+	spin_unlock_irqrestore(&csi->lock, flags);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ipu_csi_set_mipi_datatype);
+
+int ipu_csi_set_skip_smfc(struct ipu_csi *csi, u32 skip,
+			  u32 max_ratio, u32 id)
+{
+	unsigned long flags;
+	u32 temp;
+
+	if (max_ratio > 5 || id > 3)
+		return -EINVAL;
+
+	spin_lock_irqsave(&csi->lock, flags);
+
+	temp = ipu_csi_read(csi, CSI_SKIP);
+	temp &= ~(CSI_MAX_RATIO_SKIP_SMFC_MASK | CSI_ID_2_SKIP_MASK |
+		  CSI_SKIP_SMFC_MASK);
+	temp |= (max_ratio << CSI_MAX_RATIO_SKIP_SMFC_SHIFT) |
+		(id << CSI_ID_2_SKIP_SHIFT) |
+		(skip << CSI_SKIP_SMFC_SHIFT);
+	ipu_csi_write(csi, temp, CSI_SKIP);
+
+	spin_unlock_irqrestore(&csi->lock, flags);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ipu_csi_set_skip_smfc);
+
+int ipu_csi_set_dest(struct ipu_csi *csi, enum ipu_csi_dest csi_dest)
+{
+	unsigned long flags;
+	u32 csi_sens_conf, dest;
+
+	if (csi_dest == IPU_CSI_DEST_IDMAC)
+		dest = CSI_DATA_DEST_IDMAC;
+	else
+		dest = CSI_DATA_DEST_IC; /* IC or VDIC */
+
+	spin_lock_irqsave(&csi->lock, flags);
+
+	csi_sens_conf = ipu_csi_read(csi, CSI_SENS_CONF);
+	csi_sens_conf &= ~CSI_SENS_CONF_DATA_DEST_MASK;
+	csi_sens_conf |= (dest << CSI_SENS_CONF_DATA_DEST_SHIFT);
+	ipu_csi_write(csi, csi_sens_conf, CSI_SENS_CONF);
+
+	spin_unlock_irqrestore(&csi->lock, flags);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ipu_csi_set_dest);
+
+int ipu_csi_enable(struct ipu_csi *csi)
+{
+	ipu_module_enable(csi->ipu, csi->module);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ipu_csi_enable);
+
+int ipu_csi_disable(struct ipu_csi *csi)
+{
+	ipu_module_disable(csi->ipu, csi->module);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ipu_csi_disable);
+
+struct ipu_csi *ipu_csi_get(struct ipu_soc *ipu, int id)
+{
+	unsigned long flags;
+	struct ipu_csi *csi, *ret;
+
+	if (id > 1)
+		return ERR_PTR(-EINVAL);
+
+	csi = ipu->csi_priv[id];
+	ret = csi;
+
+	spin_lock_irqsave(&csi->lock, flags);
+
+	if (csi->inuse) {
+		ret = ERR_PTR(-EBUSY);
+		goto unlock;
+	}
+
+	csi->inuse = true;
+unlock:
+	spin_unlock_irqrestore(&csi->lock, flags);
+	return ret;
+}
+EXPORT_SYMBOL_GPL(ipu_csi_get);
+
+void ipu_csi_put(struct ipu_csi *csi)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&csi->lock, flags);
+	csi->inuse = false;
+	spin_unlock_irqrestore(&csi->lock, flags);
+}
+EXPORT_SYMBOL_GPL(ipu_csi_put);
+
+int ipu_csi_init(struct ipu_soc *ipu, struct device *dev, int id,
+		 unsigned long base, u32 module, struct clk *clk_ipu)
+{
+	struct ipu_csi *csi;
+
+	if (id > 1)
+		return -ENODEV;
+
+	csi = devm_kzalloc(dev, sizeof(*csi), GFP_KERNEL);
+	if (!csi)
+		return -ENOMEM;
+
+	ipu->csi_priv[id] = csi;
+
+	spin_lock_init(&csi->lock);
+	csi->module = module;
+	csi->id = id;
+	csi->clk_ipu = clk_ipu;
+	csi->base = devm_ioremap(dev, base, PAGE_SIZE);
+	if (!csi->base)
+		return -ENOMEM;
+
+	dev_dbg(dev, "CSI%d base: 0x%08lx remapped to %p\n",
+		id, base, csi->base);
+	csi->ipu = ipu;
+
+	return 0;
+}
+
+void ipu_csi_exit(struct ipu_soc *ipu, int id)
+{
+}
+
+void ipu_csi_dump(struct ipu_csi *csi)
+{
+	dev_dbg(csi->ipu->dev, "CSI_SENS_CONF:     %08x\n",
+		ipu_csi_read(csi, CSI_SENS_CONF));
+	dev_dbg(csi->ipu->dev, "CSI_SENS_FRM_SIZE: %08x\n",
+		ipu_csi_read(csi, CSI_SENS_FRM_SIZE));
+	dev_dbg(csi->ipu->dev, "CSI_ACT_FRM_SIZE:  %08x\n",
+		ipu_csi_read(csi, CSI_ACT_FRM_SIZE));
+	dev_dbg(csi->ipu->dev, "CSI_OUT_FRM_CTRL:  %08x\n",
+		ipu_csi_read(csi, CSI_OUT_FRM_CTRL));
+	dev_dbg(csi->ipu->dev, "CSI_TST_CTRL:      %08x\n",
+		ipu_csi_read(csi, CSI_TST_CTRL));
+	dev_dbg(csi->ipu->dev, "CSI_CCIR_CODE_1:   %08x\n",
+		ipu_csi_read(csi, CSI_CCIR_CODE_1));
+	dev_dbg(csi->ipu->dev, "CSI_CCIR_CODE_2:   %08x\n",
+		ipu_csi_read(csi, CSI_CCIR_CODE_2));
+	dev_dbg(csi->ipu->dev, "CSI_CCIR_CODE_3:   %08x\n",
+		ipu_csi_read(csi, CSI_CCIR_CODE_3));
+	dev_dbg(csi->ipu->dev, "CSI_MIPI_DI:       %08x\n",
+		ipu_csi_read(csi, CSI_MIPI_DI));
+	dev_dbg(csi->ipu->dev, "CSI_SKIP:          %08x\n",
+		ipu_csi_read(csi, CSI_SKIP));
+}
+EXPORT_SYMBOL_GPL(ipu_csi_dump);
diff --git a/drivers/gpu/ipu-v3/ipu-ic.c b/drivers/gpu/ipu-v3/ipu-ic.c
new file mode 100644
index 000000000000..ad75588e1629
--- /dev/null
+++ b/drivers/gpu/ipu-v3/ipu-ic.c
@@ -0,0 +1,778 @@
+/*
+ * Copyright (C) 2012-2014 Mentor Graphics Inc.
+ * Copyright 2005-2012 Freescale Semiconductor, Inc. All Rights Reserved.
+ *
+ * The code contained herein is licensed under the GNU General Public
+ * License. You may obtain a copy of the GNU General Public License
+ * Version 2 or later at the following locations:
+ *
+ * http://www.opensource.org/licenses/gpl-license.html
+ * http://www.gnu.org/copyleft/gpl.html
+ */
+
+#include <linux/types.h>
+#include <linux/init.h>
+#include <linux/errno.h>
+#include <linux/spinlock.h>
+#include <linux/bitrev.h>
+#include <linux/io.h>
+#include <linux/err.h>
+#include "ipu-prv.h"
+
+/* IC Register Offsets */
+#define IC_CONF                 0x0000
+#define IC_PRP_ENC_RSC          0x0004
+#define IC_PRP_VF_RSC           0x0008
+#define IC_PP_RSC               0x000C
+#define IC_CMBP_1               0x0010
+#define IC_CMBP_2               0x0014
+#define IC_IDMAC_1              0x0018
+#define IC_IDMAC_2              0x001C
+#define IC_IDMAC_3              0x0020
+#define IC_IDMAC_4              0x0024
+
+/* IC Register Fields */
+#define IC_CONF_PRPENC_EN       (1 << 0)
+#define IC_CONF_PRPENC_CSC1     (1 << 1)
+#define IC_CONF_PRPENC_ROT_EN   (1 << 2)
+#define IC_CONF_PRPVF_EN        (1 << 8)
+#define IC_CONF_PRPVF_CSC1      (1 << 9)
+#define IC_CONF_PRPVF_CSC2      (1 << 10)
+#define IC_CONF_PRPVF_CMB       (1 << 11)
+#define IC_CONF_PRPVF_ROT_EN    (1 << 12)
+#define IC_CONF_PP_EN           (1 << 16)
+#define IC_CONF_PP_CSC1         (1 << 17)
+#define IC_CONF_PP_CSC2         (1 << 18)
+#define IC_CONF_PP_CMB          (1 << 19)
+#define IC_CONF_PP_ROT_EN       (1 << 20)
+#define IC_CONF_IC_GLB_LOC_A    (1 << 28)
+#define IC_CONF_KEY_COLOR_EN    (1 << 29)
+#define IC_CONF_RWS_EN          (1 << 30)
+#define IC_CONF_CSI_MEM_WR_EN   (1 << 31)
+
+#define IC_IDMAC_1_CB0_BURST_16         (1 << 0)
+#define IC_IDMAC_1_CB1_BURST_16         (1 << 1)
+#define IC_IDMAC_1_CB2_BURST_16         (1 << 2)
+#define IC_IDMAC_1_CB3_BURST_16         (1 << 3)
+#define IC_IDMAC_1_CB4_BURST_16         (1 << 4)
+#define IC_IDMAC_1_CB5_BURST_16         (1 << 5)
+#define IC_IDMAC_1_CB6_BURST_16         (1 << 6)
+#define IC_IDMAC_1_CB7_BURST_16         (1 << 7)
+#define IC_IDMAC_1_PRPENC_ROT_MASK      (0x7 << 11)
+#define IC_IDMAC_1_PRPENC_ROT_OFFSET    11
+#define IC_IDMAC_1_PRPVF_ROT_MASK       (0x7 << 14)
+#define IC_IDMAC_1_PRPVF_ROT_OFFSET     14
+#define IC_IDMAC_1_PP_ROT_MASK          (0x7 << 17)
+#define IC_IDMAC_1_PP_ROT_OFFSET        17
+#define IC_IDMAC_1_PP_FLIP_RS           (1 << 22)
+#define IC_IDMAC_1_PRPVF_FLIP_RS        (1 << 21)
+#define IC_IDMAC_1_PRPENC_FLIP_RS       (1 << 20)
+
+#define IC_IDMAC_2_PRPENC_HEIGHT_MASK   (0x3ff << 0)
+#define IC_IDMAC_2_PRPENC_HEIGHT_OFFSET 0
+#define IC_IDMAC_2_PRPVF_HEIGHT_MASK    (0x3ff << 10)
+#define IC_IDMAC_2_PRPVF_HEIGHT_OFFSET  10
+#define IC_IDMAC_2_PP_HEIGHT_MASK       (0x3ff << 20)
+#define IC_IDMAC_2_PP_HEIGHT_OFFSET     20
+
+#define IC_IDMAC_3_PRPENC_WIDTH_MASK    (0x3ff << 0)
+#define IC_IDMAC_3_PRPENC_WIDTH_OFFSET  0
+#define IC_IDMAC_3_PRPVF_WIDTH_MASK     (0x3ff << 10)
+#define IC_IDMAC_3_PRPVF_WIDTH_OFFSET   10
+#define IC_IDMAC_3_PP_WIDTH_MASK        (0x3ff << 20)
+#define IC_IDMAC_3_PP_WIDTH_OFFSET      20
+
+struct ic_task_regoffs {
+	u32 rsc;
+	u32 tpmem_csc[2];
+};
+
+struct ic_task_bitfields {
+	u32 ic_conf_en;
+	u32 ic_conf_rot_en;
+	u32 ic_conf_cmb_en;
+	u32 ic_conf_csc1_en;
+	u32 ic_conf_csc2_en;
+	u32 ic_cmb_galpha_bit;
+};
+
+static const struct ic_task_regoffs ic_task_reg[IC_NUM_TASKS] = {
+	[IC_TASK_ENCODER] = {
+		.rsc = IC_PRP_ENC_RSC,
+		.tpmem_csc = {0x2008, 0},
+	},
+	[IC_TASK_VIEWFINDER] = {
+		.rsc = IC_PRP_VF_RSC,
+		.tpmem_csc = {0x4028, 0x4040},
+	},
+	[IC_TASK_POST_PROCESSOR] = {
+		.rsc = IC_PP_RSC,
+		.tpmem_csc = {0x6060, 0x6078},
+	},
+};
+
+static const struct ic_task_bitfields ic_task_bit[IC_NUM_TASKS] = {
+	[IC_TASK_ENCODER] = {
+		.ic_conf_en = IC_CONF_PRPENC_EN,
+		.ic_conf_rot_en = IC_CONF_PRPENC_ROT_EN,
+		.ic_conf_cmb_en = 0,    /* NA */
+		.ic_conf_csc1_en = IC_CONF_PRPENC_CSC1,
+		.ic_conf_csc2_en = 0,   /* NA */
+		.ic_cmb_galpha_bit = 0, /* NA */
+	},
+	[IC_TASK_VIEWFINDER] = {
+		.ic_conf_en = IC_CONF_PRPVF_EN,
+		.ic_conf_rot_en = IC_CONF_PRPVF_ROT_EN,
+		.ic_conf_cmb_en = IC_CONF_PRPVF_CMB,
+		.ic_conf_csc1_en = IC_CONF_PRPVF_CSC1,
+		.ic_conf_csc2_en = IC_CONF_PRPVF_CSC2,
+		.ic_cmb_galpha_bit = 0,
+	},
+	[IC_TASK_POST_PROCESSOR] = {
+		.ic_conf_en = IC_CONF_PP_EN,
+		.ic_conf_rot_en = IC_CONF_PP_ROT_EN,
+		.ic_conf_cmb_en = IC_CONF_PP_CMB,
+		.ic_conf_csc1_en = IC_CONF_PP_CSC1,
+		.ic_conf_csc2_en = IC_CONF_PP_CSC2,
+		.ic_cmb_galpha_bit = 8,
+	},
+};
+
+struct ipu_ic_priv;
+
+struct ipu_ic {
+	enum ipu_ic_task task;
+	const struct ic_task_regoffs *reg;
+	const struct ic_task_bitfields *bit;
+
+	enum ipu_color_space in_cs, g_in_cs;
+	enum ipu_color_space out_cs;
+	bool graphics;
+	bool rotation;
+	bool in_use;
+
+	struct ipu_ic_priv *priv;
+};
+
+struct ipu_ic_priv {
+	void __iomem *base;
+	void __iomem *tpmem_base;
+	spinlock_t lock;
+	struct ipu_soc *ipu;
+	int use_count;
+	struct ipu_ic task[IC_NUM_TASKS];
+};
+
+static inline u32 ipu_ic_read(struct ipu_ic *ic, unsigned offset)
+{
+	return readl(ic->priv->base + offset);
+}
+
+static inline void ipu_ic_write(struct ipu_ic *ic, u32 value, unsigned offset)
+{
+	writel(value, ic->priv->base + offset);
+}
+
+struct ic_csc_params {
+	s16 coeff[3][3];	/* signed 9-bit integer coefficients */
+	s16 offset[3];		/* signed 11+2-bit fixed point offset */
+	u8 scale:2;		/* scale coefficients * 2^(scale-1) */
+	bool sat:1;		/* saturate to (16, 235(Y) / 240(U, V)) */
+};
+
+/*
+ * Y = R *  .299 + G *  .587 + B *  .114;
+ * U = R * -.169 + G * -.332 + B *  .500 + 128.;
+ * V = R *  .500 + G * -.419 + B * -.0813 + 128.;
+ */
+static const struct ic_csc_params ic_csc_rgb2ycbcr = {
+	.coeff = {
+		{ 77, 150, 29 },
+		{ 469, 427, 128 },
+		{ 128, 405, 491 },
+	},
+	.offset = { 0, 512, 512 },
+	.scale = 1,
+};
+
+/* transparent RGB->RGB matrix for graphics combining */
+static const struct ic_csc_params ic_csc_rgb2rgb = {
+	.coeff = {
+		{ 128, 0, 0 },
+		{ 0, 128, 0 },
+		{ 0, 0, 128 },
+	},
+	.scale = 2,
+};
+
+/*
+ * R = (1.164 * (Y - 16)) + (1.596 * (Cr - 128));
+ * G = (1.164 * (Y - 16)) - (0.392 * (Cb - 128)) - (0.813 * (Cr - 128));
+ * B = (1.164 * (Y - 16)) + (2.017 * (Cb - 128);
+ */
+static const struct ic_csc_params ic_csc_ycbcr2rgb = {
+	.coeff = {
+		{ 149, 0, 204 },
+		{ 149, 462, 408 },
+		{ 149, 255, 0 },
+	},
+	.offset = { -446, 266, -554 },
+	.scale = 2,
+};
+
+static int init_csc(struct ipu_ic *ic,
+		    enum ipu_color_space inf,
+		    enum ipu_color_space outf,
+		    int csc_index)
+{
+	struct ipu_ic_priv *priv = ic->priv;
+	const struct ic_csc_params *params;
+	u32 __iomem *base;
+	const u16 (*c)[3];
+	const u16 *a;
+	u32 param;
+
+	base = (u32 __iomem *)
+		(priv->tpmem_base + ic->reg->tpmem_csc[csc_index]);
+
+	if (inf == IPUV3_COLORSPACE_YUV && outf == IPUV3_COLORSPACE_RGB)
+		params = &ic_csc_ycbcr2rgb;
+	else if (inf == IPUV3_COLORSPACE_RGB && outf == IPUV3_COLORSPACE_YUV)
+		params = &ic_csc_rgb2ycbcr;
+	else if (inf == IPUV3_COLORSPACE_RGB && outf == IPUV3_COLORSPACE_RGB)
+		params = &ic_csc_rgb2rgb;
+	else {
+		dev_err(priv->ipu->dev, "Unsupported color space conversion\n");
+		return -EINVAL;
+	}
+
+	/* Cast to unsigned */
+	c = (const u16 (*)[3])params->coeff;
+	a = (const u16 *)params->offset;
+
+	param = ((a[0] & 0x1f) << 27) | ((c[0][0] & 0x1ff) << 18) |
+		((c[1][1] & 0x1ff) << 9) | (c[2][2] & 0x1ff);
+	writel(param, base++);
+
+	param = ((a[0] & 0x1fe0) >> 5) | (params->scale << 8) |
+		(params->sat << 9);
+	writel(param, base++);
+
+	param = ((a[1] & 0x1f) << 27) | ((c[0][1] & 0x1ff) << 18) |
+		((c[1][0] & 0x1ff) << 9) | (c[2][0] & 0x1ff);
+	writel(param, base++);
+
+	param = ((a[1] & 0x1fe0) >> 5);
+	writel(param, base++);
+
+	param = ((a[2] & 0x1f) << 27) | ((c[0][2] & 0x1ff) << 18) |
+		((c[1][2] & 0x1ff) << 9) | (c[2][1] & 0x1ff);
+	writel(param, base++);
+
+	param = ((a[2] & 0x1fe0) >> 5);
+	writel(param, base++);
+
+	return 0;
+}
+
+static int calc_resize_coeffs(struct ipu_ic *ic,
+			      u32 in_size, u32 out_size,
+			      u32 *resize_coeff,
+			      u32 *downsize_coeff)
+{
+	struct ipu_ic_priv *priv = ic->priv;
+	struct ipu_soc *ipu = priv->ipu;
+	u32 temp_size, temp_downsize;
+
+	/*
+	 * Input size cannot be more than 4096, and output size cannot
+	 * be more than 1024
+	 */
+	if (in_size > 4096) {
+		dev_err(ipu->dev, "Unsupported resize (in_size > 4096)\n");
+		return -EINVAL;
+	}
+	if (out_size > 1024) {
+		dev_err(ipu->dev, "Unsupported resize (out_size > 1024)\n");
+		return -EINVAL;
+	}
+
+	/* Cannot downsize more than 8:1 */
+	if ((out_size << 3) < in_size) {
+		dev_err(ipu->dev, "Unsupported downsize\n");
+		return -EINVAL;
+	}
+
+	/* Compute downsizing coefficient */
+	temp_downsize = 0;
+	temp_size = in_size;
+	while (((temp_size > 1024) || (temp_size >= out_size * 2)) &&
+	       (temp_downsize < 2)) {
+		temp_size >>= 1;
+		temp_downsize++;
+	}
+	*downsize_coeff = temp_downsize;
+
+	/*
+	 * compute resizing coefficient using the following equation:
+	 * resize_coeff = M * (SI - 1) / (SO - 1)
+	 * where M = 2^13, SI = input size, SO = output size
+	 */
+	*resize_coeff = (8192L * (temp_size - 1)) / (out_size - 1);
+	if (*resize_coeff >= 16384L) {
+		dev_err(ipu->dev, "Warning! Overflow on resize coeff.\n");
+		*resize_coeff = 0x3FFF;
+	}
+
+	return 0;
+}
+
+void ipu_ic_task_enable(struct ipu_ic *ic)
+{
+	struct ipu_ic_priv *priv = ic->priv;
+	unsigned long flags;
+	u32 ic_conf;
+
+	spin_lock_irqsave(&priv->lock, flags);
+
+	ic_conf = ipu_ic_read(ic, IC_CONF);
+
+	ic_conf |= ic->bit->ic_conf_en;
+
+	if (ic->rotation)
+		ic_conf |= ic->bit->ic_conf_rot_en;
+
+	if (ic->in_cs != ic->out_cs)
+		ic_conf |= ic->bit->ic_conf_csc1_en;
+
+	if (ic->graphics) {
+		ic_conf |= ic->bit->ic_conf_cmb_en;
+		ic_conf |= ic->bit->ic_conf_csc1_en;
+
+		if (ic->g_in_cs != ic->out_cs)
+			ic_conf |= ic->bit->ic_conf_csc2_en;
+	}
+
+	ipu_ic_write(ic, ic_conf, IC_CONF);
+
+	spin_unlock_irqrestore(&priv->lock, flags);
+}
+EXPORT_SYMBOL_GPL(ipu_ic_task_enable);
+
+void ipu_ic_task_disable(struct ipu_ic *ic)
+{
+	struct ipu_ic_priv *priv = ic->priv;
+	unsigned long flags;
+	u32 ic_conf;
+
+	spin_lock_irqsave(&priv->lock, flags);
+
+	ic_conf = ipu_ic_read(ic, IC_CONF);
+
+	ic_conf &= ~(ic->bit->ic_conf_en |
+		     ic->bit->ic_conf_csc1_en |
+		     ic->bit->ic_conf_rot_en);
+	if (ic->bit->ic_conf_csc2_en)
+		ic_conf &= ~ic->bit->ic_conf_csc2_en;
+	if (ic->bit->ic_conf_cmb_en)
+		ic_conf &= ~ic->bit->ic_conf_cmb_en;
+
+	ipu_ic_write(ic, ic_conf, IC_CONF);
+
+	ic->rotation = ic->graphics = false;
+
+	spin_unlock_irqrestore(&priv->lock, flags);
+}
+EXPORT_SYMBOL_GPL(ipu_ic_task_disable);
+
+int ipu_ic_task_graphics_init(struct ipu_ic *ic,
+			      enum ipu_color_space in_g_cs,
+			      bool galpha_en, u32 galpha,
+			      bool colorkey_en, u32 colorkey)
+{
+	struct ipu_ic_priv *priv = ic->priv;
+	unsigned long flags;
+	u32 reg, ic_conf;
+	int ret = 0;
+
+	if (ic->task == IC_TASK_ENCODER)
+		return -EINVAL;
+
+	spin_lock_irqsave(&priv->lock, flags);
+
+	ic_conf = ipu_ic_read(ic, IC_CONF);
+
+	if (!(ic_conf & ic->bit->ic_conf_csc1_en)) {
+		/* need transparent CSC1 conversion */
+		ret = init_csc(ic, IPUV3_COLORSPACE_RGB,
+			       IPUV3_COLORSPACE_RGB, 0);
+		if (ret)
+			goto unlock;
+	}
+
+	ic->g_in_cs = in_g_cs;
+
+	if (ic->g_in_cs != ic->out_cs) {
+		ret = init_csc(ic, ic->g_in_cs, ic->out_cs, 1);
+		if (ret)
+			goto unlock;
+	}
+
+	if (galpha_en) {
+		ic_conf |= IC_CONF_IC_GLB_LOC_A;
+		reg = ipu_ic_read(ic, IC_CMBP_1);
+		reg &= ~(0xff << ic->bit->ic_cmb_galpha_bit);
+		reg |= (galpha << ic->bit->ic_cmb_galpha_bit);
+		ipu_ic_write(ic, reg, IC_CMBP_1);
+	} else
+		ic_conf &= ~IC_CONF_IC_GLB_LOC_A;
+
+	if (colorkey_en) {
+		ic_conf |= IC_CONF_KEY_COLOR_EN;
+		ipu_ic_write(ic, colorkey, IC_CMBP_2);
+	} else
+		ic_conf &= ~IC_CONF_KEY_COLOR_EN;
+
+	ipu_ic_write(ic, ic_conf, IC_CONF);
+
+	ic->graphics = true;
+unlock:
+	spin_unlock_irqrestore(&priv->lock, flags);
+	return ret;
+}
+EXPORT_SYMBOL_GPL(ipu_ic_task_graphics_init);
+
+int ipu_ic_task_init(struct ipu_ic *ic,
+		     int in_width, int in_height,
+		     int out_width, int out_height,
+		     enum ipu_color_space in_cs,
+		     enum ipu_color_space out_cs)
+{
+	struct ipu_ic_priv *priv = ic->priv;
+	u32 reg, downsize_coeff, resize_coeff;
+	unsigned long flags;
+	int ret = 0;
+
+	/* Setup vertical resizing */
+	ret = calc_resize_coeffs(ic, in_height, out_height,
+				 &resize_coeff, &downsize_coeff);
+	if (ret)
+		return ret;
+
+	reg = (downsize_coeff << 30) | (resize_coeff << 16);
+
+	/* Setup horizontal resizing */
+	ret = calc_resize_coeffs(ic, in_width, out_width,
+				 &resize_coeff, &downsize_coeff);
+	if (ret)
+		return ret;
+
+	reg |= (downsize_coeff << 14) | resize_coeff;
+
+	spin_lock_irqsave(&priv->lock, flags);
+
+	ipu_ic_write(ic, reg, ic->reg->rsc);
+
+	/* Setup color space conversion */
+	ic->in_cs = in_cs;
+	ic->out_cs = out_cs;
+
+	if (ic->in_cs != ic->out_cs) {
+		ret = init_csc(ic, ic->in_cs, ic->out_cs, 0);
+		if (ret)
+			goto unlock;
+	}
+
+unlock:
+	spin_unlock_irqrestore(&priv->lock, flags);
+	return ret;
+}
+EXPORT_SYMBOL_GPL(ipu_ic_task_init);
+
+int ipu_ic_task_idma_init(struct ipu_ic *ic, struct ipuv3_channel *channel,
+			  u32 width, u32 height, int burst_size,
+			  enum ipu_rotate_mode rot)
+{
+	struct ipu_ic_priv *priv = ic->priv;
+	struct ipu_soc *ipu = priv->ipu;
+	u32 ic_idmac_1, ic_idmac_2, ic_idmac_3;
+	u32 temp_rot = bitrev8(rot) >> 5;
+	bool need_hor_flip = false;
+	unsigned long flags;
+	int ret = 0;
+
+	if ((burst_size != 8) && (burst_size != 16)) {
+		dev_err(ipu->dev, "Illegal burst length for IC\n");
+		return -EINVAL;
+	}
+
+	width--;
+	height--;
+
+	if (temp_rot & 0x2)	/* Need horizontal flip */
+		need_hor_flip = true;
+
+	spin_lock_irqsave(&priv->lock, flags);
+
+	ic_idmac_1 = ipu_ic_read(ic, IC_IDMAC_1);
+	ic_idmac_2 = ipu_ic_read(ic, IC_IDMAC_2);
+	ic_idmac_3 = ipu_ic_read(ic, IC_IDMAC_3);
+
+	switch (channel->num) {
+	case IPUV3_CHANNEL_IC_PP_MEM:
+		if (burst_size == 16)
+			ic_idmac_1 |= IC_IDMAC_1_CB2_BURST_16;
+		else
+			ic_idmac_1 &= ~IC_IDMAC_1_CB2_BURST_16;
+
+		if (need_hor_flip)
+			ic_idmac_1 |= IC_IDMAC_1_PP_FLIP_RS;
+		else
+			ic_idmac_1 &= ~IC_IDMAC_1_PP_FLIP_RS;
+
+		ic_idmac_2 &= ~IC_IDMAC_2_PP_HEIGHT_MASK;
+		ic_idmac_2 |= height << IC_IDMAC_2_PP_HEIGHT_OFFSET;
+
+		ic_idmac_3 &= ~IC_IDMAC_3_PP_WIDTH_MASK;
+		ic_idmac_3 |= width << IC_IDMAC_3_PP_WIDTH_OFFSET;
+		break;
+	case IPUV3_CHANNEL_MEM_IC_PP:
+		if (burst_size == 16)
+			ic_idmac_1 |= IC_IDMAC_1_CB5_BURST_16;
+		else
+			ic_idmac_1 &= ~IC_IDMAC_1_CB5_BURST_16;
+		break;
+	case IPUV3_CHANNEL_MEM_ROT_PP:
+		ic_idmac_1 &= ~IC_IDMAC_1_PP_ROT_MASK;
+		ic_idmac_1 |= temp_rot << IC_IDMAC_1_PP_ROT_OFFSET;
+		break;
+	case IPUV3_CHANNEL_MEM_IC_PRP_VF:
+		if (burst_size == 16)
+			ic_idmac_1 |= IC_IDMAC_1_CB6_BURST_16;
+		else
+			ic_idmac_1 &= ~IC_IDMAC_1_CB6_BURST_16;
+		break;
+	case IPUV3_CHANNEL_IC_PRP_ENC_MEM:
+		if (burst_size == 16)
+			ic_idmac_1 |= IC_IDMAC_1_CB0_BURST_16;
+		else
+			ic_idmac_1 &= ~IC_IDMAC_1_CB0_BURST_16;
+
+		if (need_hor_flip)
+			ic_idmac_1 |= IC_IDMAC_1_PRPENC_FLIP_RS;
+		else
+			ic_idmac_1 &= ~IC_IDMAC_1_PRPENC_FLIP_RS;
+
+		ic_idmac_2 &= ~IC_IDMAC_2_PRPENC_HEIGHT_MASK;
+		ic_idmac_2 |= height << IC_IDMAC_2_PRPENC_HEIGHT_OFFSET;
+
+		ic_idmac_3 &= ~IC_IDMAC_3_PRPENC_WIDTH_MASK;
+		ic_idmac_3 |= width << IC_IDMAC_3_PRPENC_WIDTH_OFFSET;
+		break;
+	case IPUV3_CHANNEL_MEM_ROT_ENC:
+		ic_idmac_1 &= ~IC_IDMAC_1_PRPENC_ROT_MASK;
+		ic_idmac_1 |= temp_rot << IC_IDMAC_1_PRPENC_ROT_OFFSET;
+		break;
+	case IPUV3_CHANNEL_IC_PRP_VF_MEM:
+		if (burst_size == 16)
+			ic_idmac_1 |= IC_IDMAC_1_CB1_BURST_16;
+		else
+			ic_idmac_1 &= ~IC_IDMAC_1_CB1_BURST_16;
+
+		if (need_hor_flip)
+			ic_idmac_1 |= IC_IDMAC_1_PRPVF_FLIP_RS;
+		else
+			ic_idmac_1 &= ~IC_IDMAC_1_PRPVF_FLIP_RS;
+
+		ic_idmac_2 &= ~IC_IDMAC_2_PRPVF_HEIGHT_MASK;
+		ic_idmac_2 |= height << IC_IDMAC_2_PRPVF_HEIGHT_OFFSET;
+
+		ic_idmac_3 &= ~IC_IDMAC_3_PRPVF_WIDTH_MASK;
+		ic_idmac_3 |= width << IC_IDMAC_3_PRPVF_WIDTH_OFFSET;
+		break;
+	case IPUV3_CHANNEL_MEM_ROT_VF:
+		ic_idmac_1 &= ~IC_IDMAC_1_PRPVF_ROT_MASK;
+		ic_idmac_1 |= temp_rot << IC_IDMAC_1_PRPVF_ROT_OFFSET;
+		break;
+	case IPUV3_CHANNEL_G_MEM_IC_PRP_VF:
+		if (burst_size == 16)
+			ic_idmac_1 |= IC_IDMAC_1_CB3_BURST_16;
+		else
+			ic_idmac_1 &= ~IC_IDMAC_1_CB3_BURST_16;
+		break;
+	case IPUV3_CHANNEL_G_MEM_IC_PP:
+		if (burst_size == 16)
+			ic_idmac_1 |= IC_IDMAC_1_CB4_BURST_16;
+		else
+			ic_idmac_1 &= ~IC_IDMAC_1_CB4_BURST_16;
+		break;
+	case IPUV3_CHANNEL_VDI_MEM_IC_VF:
+		if (burst_size == 16)
+			ic_idmac_1 |= IC_IDMAC_1_CB7_BURST_16;
+		else
+			ic_idmac_1 &= ~IC_IDMAC_1_CB7_BURST_16;
+		break;
+	default:
+		goto unlock;
+	}
+
+	ipu_ic_write(ic, ic_idmac_1, IC_IDMAC_1);
+	ipu_ic_write(ic, ic_idmac_2, IC_IDMAC_2);
+	ipu_ic_write(ic, ic_idmac_3, IC_IDMAC_3);
+
+	if (rot >= IPU_ROTATE_90_RIGHT)
+		ic->rotation = true;
+
+unlock:
+	spin_unlock_irqrestore(&priv->lock, flags);
+	return ret;
+}
+EXPORT_SYMBOL_GPL(ipu_ic_task_idma_init);
+
+int ipu_ic_enable(struct ipu_ic *ic)
+{
+	struct ipu_ic_priv *priv = ic->priv;
+	unsigned long flags;
+	u32 module = IPU_CONF_IC_EN;
+
+	spin_lock_irqsave(&priv->lock, flags);
+
+	if (ic->rotation)
+		module |= IPU_CONF_ROT_EN;
+
+	if (!priv->use_count)
+		ipu_module_enable(priv->ipu, module);
+
+	priv->use_count++;
+
+	spin_unlock_irqrestore(&priv->lock, flags);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ipu_ic_enable);
+
+int ipu_ic_disable(struct ipu_ic *ic)
+{
+	struct ipu_ic_priv *priv = ic->priv;
+	unsigned long flags;
+	u32 module = IPU_CONF_IC_EN | IPU_CONF_ROT_EN;
+
+	spin_lock_irqsave(&priv->lock, flags);
+
+	priv->use_count--;
+
+	if (!priv->use_count)
+		ipu_module_disable(priv->ipu, module);
+
+	if (priv->use_count < 0)
+		priv->use_count = 0;
+
+	spin_unlock_irqrestore(&priv->lock, flags);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ipu_ic_disable);
+
+struct ipu_ic *ipu_ic_get(struct ipu_soc *ipu, enum ipu_ic_task task)
+{
+	struct ipu_ic_priv *priv = ipu->ic_priv;
+	unsigned long flags;
+	struct ipu_ic *ic, *ret;
+
+	if (task >= IC_NUM_TASKS)
+		return ERR_PTR(-EINVAL);
+
+	ic = &priv->task[task];
+
+	spin_lock_irqsave(&priv->lock, flags);
+
+	if (ic->in_use) {
+		ret = ERR_PTR(-EBUSY);
+		goto unlock;
+	}
+
+	ic->in_use = true;
+	ret = ic;
+
+unlock:
+	spin_unlock_irqrestore(&priv->lock, flags);
+	return ret;
+}
+EXPORT_SYMBOL_GPL(ipu_ic_get);
+
+void ipu_ic_put(struct ipu_ic *ic)
+{
+	struct ipu_ic_priv *priv = ic->priv;
+	unsigned long flags;
+
+	spin_lock_irqsave(&priv->lock, flags);
+	ic->in_use = false;
+	spin_unlock_irqrestore(&priv->lock, flags);
+}
+EXPORT_SYMBOL_GPL(ipu_ic_put);
+
+int ipu_ic_init(struct ipu_soc *ipu, struct device *dev,
+		unsigned long base, unsigned long tpmem_base)
+{
+	struct ipu_ic_priv *priv;
+	int i;
+
+	priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL);
+	if (!priv)
+		return -ENOMEM;
+
+	ipu->ic_priv = priv;
+
+	spin_lock_init(&priv->lock);
+	priv->base = devm_ioremap(dev, base, PAGE_SIZE);
+	if (!priv->base)
+		return -ENOMEM;
+	priv->tpmem_base = devm_ioremap(dev, tpmem_base, SZ_64K);
+	if (!priv->tpmem_base)
+		return -ENOMEM;
+
+	dev_dbg(dev, "IC base: 0x%08lx remapped to %p\n", base, priv->base);
+
+	priv->ipu = ipu;
+
+	for (i = 0; i < IC_NUM_TASKS; i++) {
+		priv->task[i].task = i;
+		priv->task[i].priv = priv;
+		priv->task[i].reg = &ic_task_reg[i];
+		priv->task[i].bit = &ic_task_bit[i];
+	}
+
+	return 0;
+}
+
+void ipu_ic_exit(struct ipu_soc *ipu)
+{
+}
+
+void ipu_ic_dump(struct ipu_ic *ic)
+{
+	struct ipu_ic_priv *priv = ic->priv;
+	struct ipu_soc *ipu = priv->ipu;
+
+	dev_dbg(ipu->dev, "IC_CONF = \t0x%08X\n",
+		ipu_ic_read(ic, IC_CONF));
+	dev_dbg(ipu->dev, "IC_PRP_ENC_RSC = \t0x%08X\n",
+		ipu_ic_read(ic, IC_PRP_ENC_RSC));
+	dev_dbg(ipu->dev, "IC_PRP_VF_RSC = \t0x%08X\n",
+		ipu_ic_read(ic, IC_PRP_VF_RSC));
+	dev_dbg(ipu->dev, "IC_PP_RSC = \t0x%08X\n",
+		ipu_ic_read(ic, IC_PP_RSC));
+	dev_dbg(ipu->dev, "IC_CMBP_1 = \t0x%08X\n",
+		ipu_ic_read(ic, IC_CMBP_1));
+	dev_dbg(ipu->dev, "IC_CMBP_2 = \t0x%08X\n",
+		ipu_ic_read(ic, IC_CMBP_2));
+	dev_dbg(ipu->dev, "IC_IDMAC_1 = \t0x%08X\n",
+		ipu_ic_read(ic, IC_IDMAC_1));
+	dev_dbg(ipu->dev, "IC_IDMAC_2 = \t0x%08X\n",
+		ipu_ic_read(ic, IC_IDMAC_2));
+	dev_dbg(ipu->dev, "IC_IDMAC_3 = \t0x%08X\n",
+		ipu_ic_read(ic, IC_IDMAC_3));
+	dev_dbg(ipu->dev, "IC_IDMAC_4 = \t0x%08X\n",
+		ipu_ic_read(ic, IC_IDMAC_4));
+}
+EXPORT_SYMBOL_GPL(ipu_ic_dump);
diff --git a/drivers/gpu/ipu-v3/ipu-prv.h b/drivers/gpu/ipu-v3/ipu-prv.h
index 0a7b2adaba39..bfb1e8a4483f 100644
--- a/drivers/gpu/ipu-v3/ipu-prv.h
+++ b/drivers/gpu/ipu-v3/ipu-prv.h
@@ -24,23 +24,6 @@ struct ipu_soc;
 
 #include <video/imx-ipu-v3.h>
 
-#define IPUV3_CHANNEL_CSI0			 0
-#define IPUV3_CHANNEL_CSI1			 1
-#define IPUV3_CHANNEL_CSI2			 2
-#define IPUV3_CHANNEL_CSI3			 3
-#define IPUV3_CHANNEL_MEM_BG_SYNC		23
-#define IPUV3_CHANNEL_MEM_FG_SYNC		27
-#define IPUV3_CHANNEL_MEM_DC_SYNC		28
-#define IPUV3_CHANNEL_MEM_FG_SYNC_ALPHA		31
-#define IPUV3_CHANNEL_MEM_DC_ASYNC		41
-#define IPUV3_CHANNEL_ROT_ENC_MEM		45
-#define IPUV3_CHANNEL_ROT_VF_MEM		46
-#define IPUV3_CHANNEL_ROT_PP_MEM		47
-#define IPUV3_CHANNEL_ROT_ENC_MEM_OUT		48
-#define IPUV3_CHANNEL_ROT_VF_MEM_OUT		49
-#define IPUV3_CHANNEL_ROT_PP_MEM_OUT		50
-#define IPUV3_CHANNEL_MEM_BG_SYNC_ALPHA		51
-
 #define IPU_MCU_T_DEFAULT	8
 #define IPU_CM_IDMAC_REG_OFS	0x00008000
 #define IPU_CM_IC_REG_OFS	0x00020000
@@ -85,6 +68,7 @@ struct ipu_soc;
 #define IPU_DISP_TASK_STAT		IPU_CM_REG(0x0254)
 #define IPU_CHA_BUF0_RDY(ch)		IPU_CM_REG(0x0268 + 4 * ((ch) / 32))
 #define IPU_CHA_BUF1_RDY(ch)		IPU_CM_REG(0x0270 + 4 * ((ch) / 32))
+#define IPU_CHA_BUF2_RDY(ch)		IPU_CM_REG(0x0288 + 4 * ((ch) / 32))
 #define IPU_ALT_CHA_BUF0_RDY(ch)	IPU_CM_REG(0x0278 + 4 * ((ch) / 32))
 #define IPU_ALT_CHA_BUF1_RDY(ch)	IPU_CM_REG(0x0280 + 4 * ((ch) / 32))
 
@@ -149,9 +133,11 @@ struct ipuv3_channel {
 };
 
 struct ipu_cpmem;
+struct ipu_csi;
 struct ipu_dc_priv;
 struct ipu_dmfc_priv;
 struct ipu_di;
+struct ipu_ic_priv;
 struct ipu_smfc_priv;
 
 struct ipu_devtype;
@@ -181,6 +167,8 @@ struct ipu_soc {
 	struct ipu_dp_priv	*dp_priv;
 	struct ipu_dmfc_priv	*dmfc_priv;
 	struct ipu_di		*di_priv[2];
+	struct ipu_csi		*csi_priv[2];
+	struct ipu_ic_priv	*ic_priv;
 	struct ipu_smfc_priv	*smfc_priv;
 };
 
@@ -203,6 +191,14 @@ int ipu_module_disable(struct ipu_soc *ipu, u32 mask);
 bool ipu_idmac_channel_busy(struct ipu_soc *ipu, unsigned int chno);
 int ipu_wait_interrupt(struct ipu_soc *ipu, int irq, int ms);
 
+int ipu_csi_init(struct ipu_soc *ipu, struct device *dev, int id,
+		 unsigned long base, u32 module, struct clk *clk_ipu);
+void ipu_csi_exit(struct ipu_soc *ipu, int id);
+
+int ipu_ic_init(struct ipu_soc *ipu, struct device *dev,
+		unsigned long base, unsigned long tpmem_base);
+void ipu_ic_exit(struct ipu_soc *ipu);
+
 int ipu_di_init(struct ipu_soc *ipu, struct device *dev, int id,
 		unsigned long base, u32 module, struct clk *ipu_clk);
 void ipu_di_exit(struct ipu_soc *ipu, int id);
diff --git a/drivers/gpu/ipu-v3/ipu-smfc.c b/drivers/gpu/ipu-v3/ipu-smfc.c
index e4f85ad286fc..4ef910991413 100644
--- a/drivers/gpu/ipu-v3/ipu-smfc.c
+++ b/drivers/gpu/ipu-v3/ipu-smfc.c
@@ -8,7 +8,6 @@
  * http://www.opensource.org/licenses/gpl-license.html
  * http://www.gnu.org/copyleft/gpl.html
  */
-#define DEBUG
 #include <linux/export.h>
 #include <linux/types.h>
 #include <linux/init.h>
@@ -21,9 +20,18 @@
 
 #include "ipu-prv.h"
 
+struct ipu_smfc {
+	struct ipu_smfc_priv *priv;
+	int chno;
+	bool inuse;
+};
+
 struct ipu_smfc_priv {
 	void __iomem *base;
 	spinlock_t lock;
+	struct ipu_soc *ipu;
+	struct ipu_smfc channel[4];
+	int use_count;
 };
 
 /*SMFC Registers */
@@ -31,63 +39,166 @@ struct ipu_smfc_priv {
 #define SMFC_WMC	0x0004
 #define SMFC_BS		0x0008
 
-int ipu_smfc_set_burstsize(struct ipu_soc *ipu, int channel, int burstsize)
+int ipu_smfc_set_burstsize(struct ipu_smfc *smfc, int burstsize)
 {
-	struct ipu_smfc_priv *smfc = ipu->smfc_priv;
+	struct ipu_smfc_priv *priv = smfc->priv;
 	unsigned long flags;
 	u32 val, shift;
 
-	spin_lock_irqsave(&smfc->lock, flags);
+	spin_lock_irqsave(&priv->lock, flags);
 
-	shift = channel * 4;
-	val = readl(smfc->base + SMFC_BS);
+	shift = smfc->chno * 4;
+	val = readl(priv->base + SMFC_BS);
 	val &= ~(0xf << shift);
 	val |= burstsize << shift;
-	writel(val, smfc->base + SMFC_BS);
+	writel(val, priv->base + SMFC_BS);
 
-	spin_unlock_irqrestore(&smfc->lock, flags);
+	spin_unlock_irqrestore(&priv->lock, flags);
 
 	return 0;
 }
 EXPORT_SYMBOL_GPL(ipu_smfc_set_burstsize);
 
-int ipu_smfc_map_channel(struct ipu_soc *ipu, int channel, int csi_id, int mipi_id)
+int ipu_smfc_map_channel(struct ipu_smfc *smfc, int csi_id, int mipi_id)
 {
-	struct ipu_smfc_priv *smfc = ipu->smfc_priv;
+	struct ipu_smfc_priv *priv = smfc->priv;
 	unsigned long flags;
 	u32 val, shift;
 
-	spin_lock_irqsave(&smfc->lock, flags);
+	spin_lock_irqsave(&priv->lock, flags);
 
-	shift = channel * 3;
-	val = readl(smfc->base + SMFC_MAP);
+	shift = smfc->chno * 3;
+	val = readl(priv->base + SMFC_MAP);
 	val &= ~(0x7 << shift);
 	val |= ((csi_id << 2) | mipi_id) << shift;
-	writel(val, smfc->base + SMFC_MAP);
+	writel(val, priv->base + SMFC_MAP);
 
-	spin_unlock_irqrestore(&smfc->lock, flags);
+	spin_unlock_irqrestore(&priv->lock, flags);
 
 	return 0;
 }
 EXPORT_SYMBOL_GPL(ipu_smfc_map_channel);
 
+int ipu_smfc_set_watermark(struct ipu_smfc *smfc, u32 set_level, u32 clr_level)
+{
+	struct ipu_smfc_priv *priv = smfc->priv;
+	unsigned long flags;
+	u32 val, shift;
+
+	spin_lock_irqsave(&priv->lock, flags);
+
+	shift = smfc->chno * 6 + (smfc->chno > 1 ? 4 : 0);
+	val = readl(priv->base + SMFC_WMC);
+	val &= ~(0x3f << shift);
+	val |= ((clr_level << 3) | set_level) << shift;
+	writel(val, priv->base + SMFC_WMC);
+
+	spin_unlock_irqrestore(&priv->lock, flags);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ipu_smfc_set_watermark);
+
+int ipu_smfc_enable(struct ipu_smfc *smfc)
+{
+	struct ipu_smfc_priv *priv = smfc->priv;
+	unsigned long flags;
+
+	spin_lock_irqsave(&priv->lock, flags);
+
+	if (!priv->use_count)
+		ipu_module_enable(priv->ipu, IPU_CONF_SMFC_EN);
+
+	priv->use_count++;
+
+	spin_unlock_irqrestore(&priv->lock, flags);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ipu_smfc_enable);
+
+int ipu_smfc_disable(struct ipu_smfc *smfc)
+{
+	struct ipu_smfc_priv *priv = smfc->priv;
+	unsigned long flags;
+
+	spin_lock_irqsave(&priv->lock, flags);
+
+	priv->use_count--;
+
+	if (!priv->use_count)
+		ipu_module_disable(priv->ipu, IPU_CONF_SMFC_EN);
+
+	if (priv->use_count < 0)
+		priv->use_count = 0;
+
+	spin_unlock_irqrestore(&priv->lock, flags);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ipu_smfc_disable);
+
+struct ipu_smfc *ipu_smfc_get(struct ipu_soc *ipu, unsigned int chno)
+{
+	struct ipu_smfc_priv *priv = ipu->smfc_priv;
+	struct ipu_smfc *smfc, *ret;
+	unsigned long flags;
+
+	if (chno >= 4)
+		return ERR_PTR(-EINVAL);
+
+	smfc = &priv->channel[chno];
+	ret = smfc;
+
+	spin_lock_irqsave(&priv->lock, flags);
+
+	if (smfc->inuse) {
+		ret = ERR_PTR(-EBUSY);
+		goto unlock;
+	}
+
+	smfc->inuse = true;
+unlock:
+	spin_unlock_irqrestore(&priv->lock, flags);
+	return ret;
+}
+EXPORT_SYMBOL_GPL(ipu_smfc_get);
+
+void ipu_smfc_put(struct ipu_smfc *smfc)
+{
+	struct ipu_smfc_priv *priv = smfc->priv;
+	unsigned long flags;
+
+	spin_lock_irqsave(&priv->lock, flags);
+	smfc->inuse = false;
+	spin_unlock_irqrestore(&priv->lock, flags);
+}
+EXPORT_SYMBOL_GPL(ipu_smfc_put);
+
 int ipu_smfc_init(struct ipu_soc *ipu, struct device *dev,
 		  unsigned long base)
 {
-	struct ipu_smfc_priv *smfc;
+	struct ipu_smfc_priv *priv;
+	int i;
 
-	smfc = devm_kzalloc(dev, sizeof(*smfc), GFP_KERNEL);
-	if (!smfc)
+	priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL);
+	if (!priv)
 		return -ENOMEM;
 
-	ipu->smfc_priv = smfc;
-	spin_lock_init(&smfc->lock);
+	ipu->smfc_priv = priv;
+	spin_lock_init(&priv->lock);
+	priv->ipu = ipu;
 
-	smfc->base = devm_ioremap(dev, base, PAGE_SIZE);
-	if (!smfc->base)
+	priv->base = devm_ioremap(dev, base, PAGE_SIZE);
+	if (!priv->base)
 		return -ENOMEM;
 
-	pr_debug("%s: ioremap 0x%08lx -> %p\n", __func__, base, smfc->base);
+	for (i = 0; i < 4; i++) {
+		priv->channel[i].priv = priv;
+		priv->channel[i].chno = i;
+	}
+
+	pr_debug("%s: ioremap 0x%08lx -> %p\n", __func__, base, priv->base);
 
 	return 0;
 }
author	Linus Torvalds <torvalds@linux-foundation.org>	2014-10-14 09:39:08 +0200
committer	Linus Torvalds <torvalds@linux-foundation.org>	2014-10-14 09:39:08 +0200
commit	2d65a9f48fcdf7866aab6457bc707ca233e0c791 (patch)
tree	f93e5838d6ac2e59434367f4ff905f7d9c45fc2b /drivers/gpu
parent	Merge branch 'misc' of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild (diff)
parent	drm/mst: rework payload table allocation to conform better. (diff)
download	linux-dev-2d65a9f48fcdf7866aab6457bc707ca233e0c791.tar.xz linux-dev-2d65a9f48fcdf7866aab6457bc707ca233e0c791.zip